Advanced Applications of Python Data

Structures and Algorithms

Mohammad Gouse Galety

Department of Computer Science, Samarkand International University of

Technology, Uzbekistan

Arul Kumar Natarajan

CHRIST University (Deemed), India

A. V. Sriharsha

MB University, India

A volume in the Advances in Systems Analysis, Software Engineering, and

High Performance Computing (ASASEHPC) Book Series Book Series

Published in the United States of America by IGI Global (an imprint of IGI

Global) 701 E. Chocolate Avenue Hershey PA 17033 Tel: 717-533-8845

Fax: 717-533-8661 E-mail: [email protected] Web site: http://www.igi-

global.com/reference

Copyright © 2023 by IGI Global. All rights reserved. No part of this

publication may be reproduced, stored or distributed in any form or by any

means, electronic or mechanical, including photocopying, without written

permission from the publisher.

Product or company names used in this set are for identification purposes

only. Inclusion of the names of the products or companies does not indicate

a claim of ownership by IGI Global of the trademark or registered

trademark.

Library of Congress Cataloging-in-Publication Data

Names: Galety, Mohammad Gouse, 1976- editor. | Natarajan, Arulkumar,

1986-

editor. | AV, Sriharsha, editor.

Title: Advanced applications of Python data structures and algorithms /

edited by Mohammad Galety, Arulkumar Natarajan, Sriharsha AV.

Description: Hershey, PA : Engineering Science Reference, [2023] |

Includes

bibliographical references and index. | Summary: "Assist in

understanding and applying the fundamentals of data structures and their

many implementations. Data structures play a critical role in creating

efficient applications. To improve one's ability to solve problems by

utilizing proper data formats. Upon successful completion of reading

this textbook, students will have gained the ability to: Recognize and

use the most fundamental types of data structures and algorithms. Be

familiar with the advantages and disadvantages of various data

structures. Use the proper data structure in the given problem's

solution context. Improve your ability to program to address a

particular problem. To distinguish between programmers and

non-programmers, it is necessary to know about Data Structures and

Algorithms"-- Provided by publisher.

Identifiers: LCCN 2022050718 (print) | LCCN 2022050719 (ebook) | ISBN

9781668471005 (h/c) | ISBN 9781668471012 (s/c) | ISBN 9781668471029

(eISBN)

Subjects: LCSH: Python (Computer program language) | Data structures

(Computer science) | Computer algorithms.

Classification: LCC QA76.73.P98 A37 2023 (print) | LCC QA76.73.P98

(ebook) | DDC 005.13/3--dc23/eng/20230104

LC record available at https://lccn.loc.gov/2022050718

LC ebook record available at https://lccn.loc.gov/2022050719

This book is published under the IGI Global book series Advances in

Systems Analysis, Software Engineering, and High Performance

Computing (ASASEHPC) (ISSN: 2327-3453 eISSN: 2327-3461)

British Cataloguing in Publication Data

A Cataloguing in Publication record for this book is available from the

British Library.

All work contributed to this book is new, previously-unpublished material.

The views expressed in this book are those of the authors, but not

necessarily of the publisher.

 Advances in Systems Analysis, Software

Engineering, and High Performance

Computing (ASASEHPC) Book Series

Vijayan Sugumaran (Oakland University, USA)

ISSN: 2327-3453

Mission

The theory and practice of computing applications and distributed systems

has emerged as one of the key areas of research driving innovations in

business, engineering, and science. The fields of software engineering,

systems analysis, and high performance computing offer a wide range of

applications and solutions in solving computational problems for any

modern organization.

The Advances in Systems Analysis, Software Engineering, and High

Performance Computing (ASASEHPC) Book Series brings together

research in the areas of distributed computing, systems and software

engineering, high performance computing, and service science. This

collection of publications is useful for academics, researchers, and

practitioners seeking the latest practices and knowledge in this field.

IGI Global is
Coverage
currently accepting

manuscripts for

Storage publications within

Systems
this series. To submit
Engineering
a proposal for a
Environments
volume in this series
Computer
please contact our
System
 Analysis Acquisition Editors
Distributed
at Acquisitions@igi-
Cloud
global.com or visit
Computing
http://www.igi-
Network
global.com/publish .
Management

Virtual Data

Systems

Software

Engineering

Enterprise

Information

Systems

Performance

Modelling

Human-

Computer

Interaction

The Advances in Systems Analysis, Software Engineering, and High

Performance Computing (ASASEHPC) Book Series(ISSN 2327-3453) is

published by IGI Global, 701 E. Chocolate Avenue, Hershey, PA 17033-

1240, USA, www.igi-global.com. This series is composed of titlesavailable

for purchase individually; each title is edited to be contextually exclusive

from any other title within the series. For pricing and ordering information

please visit http://www.igi-global.com/book-series/advances-systems-

analysis-software-engineering/73689. Postmaster: send all address changes

to above address. Copyright © 2023 IGI Global. All rights, including

translation in other languages reserved by the pulisher. No part of this series

may be reproduced or used in any form or by any means - graphics,

electronic, or mechanical, including photocopying, recoreding, taping, or

information and retrieval systems - without written permission from the

publisher, except for non commercial, educational use, including classroom

teaching purposes. The views expressed in this series are those of the

authors, but not necessarily of IGI Global

Titles in this Series

Neuromorphic Computing Systems for Industry 4.0

S. Dhanasekar (Department of ECE, Sree Eshwar Engineering College,

India) K. Martin Sagayam (Karunya Institute of Technology and Sciences,

India) Surbhi Vijh (KIET Group of Institutions, India) Vipin Tyagi (Jaypee

University of Engineering and Technology, India) and Alex Norta (Tallinn

University of Technology, Estonia)

Engineering Science Reference • copyright 2023 • 300pp • H/C (ISBN:

9781668465967) • US $270.00 (our price)

Business Models and Strategies for Open Source Projects

Francisco José Monaco (Universidade de São Paulo, Brazil)

Business Science Reference • copyright 2023 • 300pp • H/C (ISBN:

9781668447857) • US $270.00 (our price)

Principles, Policies, and Applications of Kotlin Programming

Duy Thanh Tran (University of Economics and Law, Ho Chi Minh City,

Vietnam & Vietnam National University, Ho Chi Minh City, Vietnam) and

Jun-Ho Huh (Korea Maritime and Ocean University, South Korea)

Engineering Science Reference • copyright 2023 • 457pp • H/C (ISBN:

9781668466872) • US $215.00 (our price)

Concepts and Techniques of Graph Neural Networks

Vinod Kumar (Koneru Lakshmaiah Education Foundation (Deemed), India)

and Dharmendra Singh Rajput (VIT University, India)

Engineering Science Reference • copyright 2023 • 247pp • H/C (ISBN:

9781668469033) • US $270.00 (our price)

Cyber-Physical System Solutions for Smart Cities

Vanamoorthy Muthumanikandan (Vellore Institute of Technology, Chennai,

India) Anbalagan Bhuvaneswari (Vellore Institute of Technology, Chennai,

India) Balamurugan Easwaran (University of Africa, Toru-Orua, Nigeria)

and T. Sudarson Rama Perumal (ROHINI College of Engineering and

Technology, India)

Engineering Science Reference • copyright 2023 • 300pp • H/C (ISBN:

9781668477564) • US $270.00 (our price)

Adaptive Security and Cyber Assurance for Risk-Based Decision Making

Tyson T. Brooks (Syracuse University, USA)

Engineering Science Reference • copyright 2023 • 243pp • H/C (ISBN:

9781668477663) • US $225.00 (our price)

Novel Research and Development Approaches in Heterogeneous Systems

and Algorithms

Santanu Koley (Haldia Institute of Technology, India) Subhabrata Barman

(Haldia Institute of Technology, India) and Subhankar Joardar (Haldia

Institute of Technology, India)

Engineering Science Reference • copyright 2023 • 323pp • H/C (ISBN:

9781668475249) • US $270.00 (our price)

Handbook of Research on Quantum Computing for Smart Environments

Amit Kumar Tyagi (National Institute of Fashion Technology, New Delhi,

India)

Engineering Science Reference • copyright 2023 • 564pp • H/C (ISBN:

9781668466971) • US $325.00 (our price)

Developing Linear Algebra Codes on Modern Processors Emerging

Research and Opportunities

Sandra Catalán Pallarés (Universidad Complutense de Madrid, Spain) Pedro

Valero-Lara (Oak Ridge National Laboratory, USA) Leonel Antonio Toledo

Díaz (Barcelona Supercomputing Center, Spain) and Rocío Carratalá Sáez

(Universidad de Valladolid, Spain)

Engineering Science Reference • copyright 2023 • 266pp • H/C (ISBN:

9781799870821) • US $215.00 (our price)

New Approaches to Data Analytics and Internet of Things Through Digital

Twin

P. Karthikeyan (National Chung Cheng University, Chiayi, Taiwan)

Polinpapilinho F. Katina (University of South Carolina Upstate, USA) and

S.P. Anandaraj (Presidency University, India)

Engineering Science Reference • copyright 2023 • 307pp • H/C (ISBN:

9781668457221) • US $270.00 (our price)

Futuristic Trends for Sustainable Development and Sustainable Ecosystems

Fernando Ortiz-Rodriguez (Tamaulipas Autonomous University, Mexico)

Sanju Tiwari (Tamaulipas Autonomous University, Mexico) Sailesh Iyer

(Rai University, India) and José Melchor Medina-Quintero (Tamaulipas

Autonomous University, Mexico)

Engineering Science Reference • copyright 2022 • 320pp • H/C (ISBN:

9781668442258) • US $270.00 (our price)

Emerging Technologies for Innovation Management in the Software

Industry

Varun Gupta (Universidad de Alcalá, Madrid, Spain) and Chetna Gupta

(Jaypee Institute of Information Technology, Noida, India)

Engineering Science Reference • copyright 2022 • 282pp • H/C (ISBN:

9781799890591) • US $270.00 (our price)

Technology Road Mapping for Quantum Computing and Engineering

Brojo Kishore Mishra (GIET University, India)

Engineering Science Reference • copyright 2022 • 243pp • H/C (ISBN:

9781799891833) • US $250.00 (our price)

Designing User Interfaces With a Data Science Approach

Abhijit Narayanrao Banubakode (MET Institute of Computer Science,

India) Ganesh Dattatray Bhutkar (Vishwakarma Institute of Technology,

India) Yohannes Kurniawan (Bina Nusantara University, Indonesia) and

Chhaya Santosh Gosavi (MKSSS's Cummins College of Engineering, India)

Engineering Science Reference • copyright 2022 • 325pp • H/C (ISBN:

9781799891215) • US $270.00 (our price)

Implementation of Machine Learning Algorithms Using Control-Flow and

Dataflow Paradigms

Veljko Milutinović (Indiana University, Bloomington, USA) Nenad Mitić

(University of Belgrade, Serbia) Aleksandar Kartelj (University of

Belgrade, Serbia) and Miloš Kotlar (University of Belgrade, Serbia)

Engineering Science Reference • copyright 2022 • 296pp • H/C (ISBN:

9781799883500) • US $270.00 (our price)

Advancing Smarter and More Secure Industrial Applications Using AI, IoT,

and Blockchain Technology

Kavita Saini (Galgotias University, India) and Pethuru Raj (Reliance Jio

Platforms Ltd., Bangalore, India)

Engineering Science Reference • copyright 2022 • 309pp • H/C (ISBN:

9781799883678) • US $270.00 (our price)

Deep Learning Applications for Cyber-Physical Systems

Monica R. Mundada (M.S. Ramaiah Institute of Technology, India) S.

Seema (M.S. Ramaiah Institute of Technology, India) Srinivasa K.G.

(National Institute of Technical Teachers Training and Research,

Chandigarh, India) and M. Shilpa (M.S. Ramaiah Institute of Technology,

India)

Engineering Science Reference • copyright 2022 • 293pp • H/C (ISBN:

9781799881612) • US $270.00 (our price)

Design, Applications, and Maintenance of Cyber-Physical Systems

Pierluigi Rea (University of Cagliari, Italy) Erika Ottaviano (University of

Cassino and Southern Lazio, Italy) José Machado (University of Minho,

Portugal) and Katarzyna Antosz (Rzeszow University of Technology,

Poland)

Engineering Science Reference • copyright 2021 • 314pp • H/C (ISBN:

9781799867210) • US $250.00 (our price)

Methodologies and Applications of Computational Statistics for Machine

Intelligence

Debabrata Samanta (Christ University (Deemed), India) Raghavendra Rao

Althar (QMS, First American India, Bangalore, India) Sabyasachi Pramanik

(Haldia Institute of Technology, India) and Soumi Dutta (Institute of

Engineering and Management, Kolkata, India)

Engineering Science Reference • copyright 2021 • 277pp • H/C (ISBN:

9781799877011) • US $270.00 (our price)

Handbook of Research on Software Quality Innovation in Interactive

Systems

Francisco Vicente Cipolla-Ficarra (Latin Association of Human-Computer

Interaction, Spain & International Association of Interactive

Communication, Italy)

Engineering Science Reference • copyright 2021 • 501pp • H/C (ISBN:

9781799870104) • US $325.00 (our price)

Handbook of Research on Methodologies and Applications of

Supercomputing

Veljko Milutinović (Indiana University, Bloomington, USA) and Miloš

Kotlar (University of Belgrade, Serbia)

Engineering Science Reference • copyright 2021 • 393pp • H/C (ISBN:

9781799871569) • US $380.00 (our price)

MATLAB® With Applications in Mechanics and Tribology

Leonid Burstein (Independent Researcher, Israel)

Engineering Science Reference • copyright 2021 • 368pp • H/C (ISBN:

9781799870784) • US $215.00 (our price)

Advancements in Fuzzy Reliability Theory

Akshay Kumar (Graphic Era Hill University, India) Mangey Ram

(Department of Mathematics, Graphic Era (Deemed to be University),

India) and Om Prakash Yadav (North Dakota State University, USA)

Engineering Science Reference • copyright 2021 • 322pp • H/C (ISBN:

9781799875642) • US $270.00 (our price)

Impacts and Challenges of Cloud Business Intelligence

Shadi Aljawarneh (Jordan University of Science and Technology, Jordan)

and Manisha Malhotra (Chandigarh University, India)

Business Science Reference • copyright 2021 • 263pp • H/C (ISBN:

9781799850403) • US $215.00 (our price)

Handbook of Research on Modeling, Analysis, and Control of Complex

Systems

Ahmad Taher Azar (Faculty of Computers and Artificial Intelligence,

Benha University, Benha, Egypt & College of Computer and Information

Sciences, Prince Sultan University, Riyadh, Saudi Arabia) and Nashwa

Ahmad Kamal (Faculty of Engineering, Cairo University, Giza, Egypt)

Engineering Science Reference • copyright 2021 • 685pp • H/C (ISBN:

9781799857884) • US $325.00 (our price)

Artificial Intelligence Paradigms for Smart Cyber-Physical Systems

Ashish Kumar Luhach (The PNG University of Technology, Papua New

Guinea) and Atilla Elçi (Hasan Kalyoncu University, Turkey)

Engineering Science Reference • copyright 2021 • 392pp • H/C (ISBN:

9781799851011) • US $250.00 (our price)

Advancements in Model-Driven Architecture in Software Engineering

Yassine Rhazali (Moulay Ismail University of Meknes, Morocco)

Engineering Science Reference • copyright 2021 • 287pp • H/C (ISBN:

9781799836612) • US $240.00 (our price)

Cloud-Based Big Data Analytics in Vehicular Ad-Hoc Networks

Ram Shringar Rao (Ambedkar Institute of Advanced Communication

Technologies and Research, India) Nanhay Singh (Ambedkar Institute of

Advanced Communication Technologies and Research, India) Omprakash

Kaiwartya (School of Science and Technology, Nottingham Trent

University, UK) and Sanjoy Das (Indira Gandhi National Tribal University,

India)

Engineering Science Reference • copyright 2021 • 312pp • H/C (ISBN:

9781799827641) • US $270.00 (our price)

Formal and Adaptive Methods for Automation of Parallel Programs

Construction Emerging Research and Opportunities

Anatoliy Doroshenko (Institute of Software Systems, Ukraine) and Olena

Yatsenko (Institute of Software Systems, Ukraine)

Engineering Science Reference • copyright 2021 • 279pp • H/C (ISBN:

9781522593843) • US $215.00 (our price)

Agile Scrum Implementation and Its Long-Term Impact on Organizations

Kenneth R. Walsh (University of New Orleans, USA) Sathiadev Mahesh

(University of New Orleans, USA) and Cherie C. Trumbach (University of

New Orleans, USA)

Engineering Science Reference • copyright 2021 • 268pp • H/C (ISBN:

9781799848851) • US $240.00 (our price)

Balancing Agile and Disciplined Engineering and Management Approaches

for IT Services and Software Products

Manuel Mora (Universidad Autónoma de Aguascalientes, Mexico) Jorge

Marx Gómez (University of Oldenburg, Germany) Rory V. O'Connor

(Dublin City University, Ireland) and Alena Buchalcevová (University of

Economics, Prague, Czech Republic)

Engineering Science Reference • copyright 2021 • 354pp • H/C (ISBN:

9781799841654) • US $250.00 (our price)

Urban Spatial Data Handling and Computing

Mainak Bandyopadhyay (DIT University-Dehradun, India) and Varun Singh

(MNNIT-Allahabad, India)

Engineering Science Reference • copyright 2020 • 300pp • H/C (ISBN:

9781799801221) • US $245.00 (our price)

FPGA Algorithms and Applications for the Internet of Things

Preeti Sharma (Bansal College of Engineering, Mandideep, India) and Rajit

Nair (Jagran Lakecity University, Bhopal, India)

Engineering Science Reference • copyright 2020 • 257pp • H/C (ISBN:

9781522598060) • US $240.00 (our price)

Advancements in Instrumentation and Control in Applied System

Applications

Srijan Bhattacharya (RCC Institute of Information Technology, India)

Engineering Science Reference • copyright 2020 • 298pp • H/C (ISBN:

9781799825845) • US $250.00 (our price)

Cloud Computing Applications and Techniques for E-Commerce

Saikat Gochhait (Symbiosis Institute of Digital and Telecom Management,

Symbiosis International University, India) David Tawei Shou (University of

Taipei, Taiwan) and Sabiha Fazalbhoy (Symbiosis Centre for Management

Studies, Symbiosis International University, India)

Engineering Science Reference • copyright 2020 • 185pp • H/C (ISBN:

9781799812944) • US $240.00 (our price)

Soft Computing Methods for System Dependability

Mohamed Arezki Mellal (M'Hamed Bougara University, Algeria)

Engineering Science Reference • copyright 2020 • 293pp • H/C (ISBN:

9781799817185) • US $250.00 (our price)

Grammatical and Syntactical Approaches in Architecture Emerging

Research and Opportunities

Ju Hyun Lee (University of New South Wales, Australia) and Michael J.

Ostwald (University of New South Wales, Australia)

Engineering Science Reference • copyright 2020 • 351pp • H/C (ISBN:

9781799816980) • US $215.00 (our price)

Tools and Techniques for Software Development in Large Organizations

Emerging Research and Opportunities

Vishnu Pendyala (Cisco Systems Inc., USA)

Engineering Science Reference • copyright 2020 • 223pp • H/C (ISBN:

9781799818632) • US $215.00 (our price)

Fundamental and Supportive Technologies for 5G Mobile Networks

Sherine Mohamed Abd El-Kader (Electronics Research Institute, Egypt)

and Hanan Hussein (Electronics Research Institute, Egypt)

Information Science Reference • copyright 2020 • 360pp • H/C (ISBN:

9781799811527) • US $250.00 (our price)

Deep Learning Techniques and Optimization Strategies in Big Data

Analytics

J. Joshua Thomas (UOW Malaysia KDU Penang University College,

Malaysia) Pinar Karagoz (Middle East Technical University, Turkey) B.

Bazeer Ahamed (Balaji Institute of Technology and Science, Warangal,

India) and Pandian Vasant (University of Technology Petronas, Malaysia)

Engineering Science Reference • copyright 2020 • 355pp • H/C (ISBN:

9781799811923) • US $270.00 (our price)

Applications and Approaches to Object-Oriented Software Design

Emerging Research and Opportunities

Zeynep Altan (Beykent University, Turkey)

Engineering Science Reference • copyright 2020 • 238pp • H/C (ISBN:

9781799821427) • US $215.00 (our price)

Crowdsourcing and Probabilistic Decision-Making in Software

Engineering Emerging Research and Opportunities

Varun Gupta (University of Beira Interior, Covilha, Portugal)

Engineering Science Reference • copyright 2020 • 182pp • H/C (ISBN:

9781522596592) • US $220.00 (our price)

Metrics and Models for Evaluating the Quality and Effectiveness of ERP

Software

Geoffrey Muchiri Muketha (Murang'a University of Technology, Kenya)

and Elyjoy Muthoni Micheni (Technical University of Kenya, Kenya)

Engineering Science Reference • copyright 2020 • 391pp • H/C (ISBN:

9781522576785) • US $250.00 (our price)

User-Centered Software Development for the Blind and Visually Impaired

Emerging Research and Opportunities

Teresita de Jesús Álvarez Robles (Universidad Veracruzana, Mexico)

Francisco Javier Álvarez Rodríguez (Universidad Autónoma de

Aguascalientes, Mexico) and Edgard Benítez-Guerrero (Universidad

Veracruzana, Mexico)

Engineering Science Reference • copyright 2020 • 173pp • H/C (ISBN:

9781522585398) • US $215.00 (our price)

Architectures and Frameworks for Developing and Applying Blockchain

Technology

Nansi Shi (Logic International Consultants, Singapore)

Engineering Science Reference • copyright 2019 • 337pp • H/C (ISBN:

9781522592570) • US $245.00 (our price)

Human Factors in Global Software Engineering

Mobashar Rehman (Universiti Tunku Abdul Rahman, Malaysia) Aamir

Amin (Universiti Tunku Abdul Rahman, Malaysia) Abdul Rehman Gilal

(Sukkur IBA University, Pakistan) and Manzoor Ahmed Hashmani

(University Technology PETRONAS, Malaysia)

Engineering Science Reference • copyright 2019 • 381pp • H/C (ISBN:

9781522594482) • US $245.00 (our price)

Interdisciplinary Approaches to Information Systems and Software

Engineering

Alok Bhushan Mukherjee (North-Eastern Hill University Shillong, India)

and Akhouri Pramod Krishna (Birla Institute of Technology Mesra, India)

Engineering Science Reference • copyright 2019 • 299pp • H/C (ISBN:

9781522577843) • US $215.00 (our price)

Cyber-Physical Systems for Social Applications

Maya Dimitrova (Bulgarian Academy of Sciences, Bulgaria) and Hiroaki

Wagatsuma (Kyushu Institute of Technology, Japan)

Engineering Science Reference • copyright 2019 • 440pp • H/C (ISBN:

9781522578796) • US $265.00 (our price)

Integrating the Internet of Things Into Software Engineering Practices

D. Jeya Mala (Thiagarajar College of Engineering, India)

Engineering Science Reference • copyright 2019 • 293pp • H/C (ISBN:

9781522577904) • US $215.00 (our price)

Analyzing the Role of Risk Mitigation and Monitoring in Software

Development

Rohit Kumar (Chandigarh University, India) Anjali Tayal (Infosys

Technologies, India) and Sargam Kapil (C-DAC, India)

Engineering Science Reference • copyright 2018 • 308pp • H/C (ISBN:

9781522560296) • US $225.00 (our price)

Handbook of Research on Pattern Engineering System Development for Big

Data Analytics

Vivek Tiwari (International Institute of Information Technology, India)

Ramjeevan Singh Thakur (Maulana Azad National Institute of Technology,

India) Basant Tiwari (Hawassa University, Ethiopia) and Shailendra Gupta

(AISECT University, India)

Engineering Science Reference • copyright 2018 • 396pp • H/C (ISBN:

9781522538707) • US $320.00 (our price)

Incorporating Nature-Inspired Paradigms in Computational Applications

Mehdi Khosrow-Pour, D.B.A. (Information Resources Management

Association, USA)

Engineering Science Reference • copyright 2018 • 385pp • H/C (ISBN:

9781522550204) • US $195.00 (our price)

Innovations in Software-Defined Networking and Network Functions

Virtualization

Ankur Dumka (University of Petroleum and Energy Studies, India)

Engineering Science Reference • copyright 2018 • 364pp • H/C (ISBN:

9781522536406) • US $235.00 (our price)

Advances in System Dynamics and Control

Ahmad Taher Azar (College of Computer & Information Sciences, Prince

Sultan University, Riyadh, Saudi Arabia & Faculty of Computers and

Artificial Intelligence, Benha University, Benha, Egypt) and

Sundarapandian Vaidyanathan (Vel Tech University, India)

Engineering Science Reference • copyright 2018 • 680pp • H/C (ISBN:

9781522540779) • US $235.00 (our price)

Green Computing Strategies for Competitive Advantage and Business

Sustainability

Mehdi Khosrow-Pour, D.B.A. (Information Resources Management

Association, USA)

Business Science Reference • copyright 2018 • 324pp • H/C (ISBN:

9781522550174) • US $185.00 (our price)

Optimizing Contemporary Application and Processes in Open Source

Software

Mehdi Khosrow-Pour, D.B.A. (Information Resources Management

Association, USA)
Engineering Science Reference • copyright 2018 • 318pp • H/C (ISBN:

9781522553144) • US $215.00 (our price)

Formation Methods, Models, and Hardware Implementation of

Pseudorandom Number Generators Emerging Research and Opportunities

Stepan Mykolayovych Bilan (State University of Infrastructure and

Technology, Ukraine)

Engineering Science Reference • copyright 2018 • 301pp • H/C (ISBN:

9781522527732) • US $180.00 (our price)

Aligning Perceptual and Conceptual Information for Cognitive Contextual

System Development Emerging Research and Opportunities

Gary Kuvich (IBM, USA)

Engineering Science Reference • copyright 2018 • 172pp • H/C (ISBN:

9781522524311) • US $165.00 (our price)

Applied Computational Intelligence and Soft Computing in Engineering

Saifullah Khalid (CCSI Airport, India)

Engineering Science Reference • copyright 2018 • 340pp • H/C (ISBN:

9781522531296) • US $225.00 (our price)

Enhancing Software Fault Prediction With Machine Learning Emerging

Research and Opportunities

Ekbal Rashid (Aurora's Technological and Research Institute, India)

Engineering Science Reference • copyright 2018 • 129pp • H/C (ISBN:

9781522531852) • US $165.00 (our price)

Solutions for Cyber-Physical Systems Ubiquity

Norbert Druml (Independent Researcher, Austria) Andreas Genser

(Independent Researcher, Austria) Armin Krieg (Independent Researcher,

Austria) Manuel Menghin (Independent Researcher, Austria) and Andrea

Hoeller (Independent Researcher, Austria)

Engineering Science Reference • copyright 2018 • 482pp • H/C (ISBN:

9781522528456) • US $225.00 (our price)

Large-Scale Fuzzy Interconnected Control Systems Design and Analysis

Zhixiong Zhong (Xiamen University of Technology, China) and Chih-Min

Lin (Yuan Ze University, Taiwan)

Information Science Reference • copyright 2017 • 223pp • H/C (ISBN:

9781522523857) • US $175.00 (our price)

Microcontroller System Design Using PIC18F Processors

Nicolas K. Haddad (University of Balamand, Lebanon)

Information Science Reference • copyright 2017 • 428pp • H/C (ISBN:

9781683180005) • US $195.00 (our price)

Probabilistic Nodes Combination (PNC) for Object Modeling and Contour

Reconstruction

Dariusz Jacek Jakóbczak (Technical University of Koszalin, Poland)

Information Science Reference • copyright 2017 • 312pp • H/C (ISBN:

9781522525318) • US $175.00 (our price)

Model-Based Design for Effective Control System Development

Wei Wu (Independent Researcher, USA)

Information Science Reference • copyright 2017 • 299pp • H/C (ISBN:

9781522523031) • US $185.00 (our price)

Comparative Approaches to Using R and Python for Statistical Data

Analysis

Rui Sarmento (University of Porto, Portugal) and Vera Costa (University of

Porto, Portugal)

Information Science Reference • copyright 2017 • 197pp • H/C (ISBN:

9781683180166) • US $180.00 (our price)

Developing Service-Oriented Applications Using the Windows

Communication Foundation (WCF) Framework

Chirag Patel (Charotar University of Science and Technology, India)

Information Science Reference • copyright 2017 • 487pp • H/C (ISBN:

9781522519973) • US $200.00 (our price)

Resource Management and Efficiency in Cloud Computing Environments

Ashok Kumar Turuk (National Institute of Technology Rourkela, India)

Bibhudatta Sahoo (National Institute of Technology Rourkela, India) and

Sourav Kanti Addya (National Institute of Technology Rourkela, India)

Information Science Reference • copyright 2017 • 352pp • H/C (ISBN:

9781522517214) • US $205.00 (our price)

Handbook of Research on End-to-End Cloud Computing Architecture

Design

Jianwen “Wendy” Chen (IBM, Australia) Yan Zhang (Western Sydney

University, Australia) and Ron Gottschalk (IBM, Australia)

Information Science Reference • copyright 2017 • 507pp • H/C (ISBN:

9781522507598) • US $325.00 (our price)

Innovative Research and Applications in Next-Generation High

Performance Computing

Qusay F. Hassan (Mansoura University, Egypt)

Information Science Reference • copyright 2016 • 488pp • H/C (ISBN:

9781522502876) • US $205.00 (our price)

Developing Interoperable and Federated Cloud Architecture

Gabor Kecskemeti (University of Miskolc, Hungary) Attila Kertesz

(University of Szeged, Hungary) and Zsolt Nemeth (MTA SZTAKI,

Hungary)

Information Science Reference • copyright 2016 • 398pp • H/C (ISBN:

9781522501534) • US $210.00 (our price)

Managing Big Data in Cloud Computing Environments

Zongmin Ma (Nanjing University of Aeronautics and Astronautics, China)

Information Science Reference • copyright 2016 • 314pp • H/C (ISBN:

9781466698345) • US $195.00 (our price)

Emerging Innovations in Agile Software Development

Imran Ghani (Universiti Teknologi Malaysia, Malaysia) Dayang Norhayati

Abang Jawawi (Universiti Teknologi Malaysia, Malaysia) Siva Dorairaj

(Software Education, New Zealand) and Ahmed Sidky (ICAgile, USA)

Information Science Reference • copyright 2016 • 323pp • H/C (ISBN:

9781466698581) • US $205.00 (our price)

Modern Software Engineering Methodologies for Mobile and Cloud

Environments
António Miguel Rosado da Cruz (Instituto Politécnico de Viana do Castelo,

Portugal) and Sara Paiva (Instituto Politécnico de Viana do Castelo,

Portugal)

Information Science Reference • copyright 2016 • 355pp • H/C (ISBN:

9781466699168) • US $210.00 (our price)

Emerging Research Surrounding Power Consumption and Performance

Issues in Utility Computing

Ganesh Chandra Deka (Regional Vocational Training Institute (RVTI) for

Women, India) G.M. Siddesh (Ramaiah Institute of Technology, India) K.

G. Srinivasa (M S Ramaiah Institute of Technology, Bangalore, India) and

L.M. Patnaik (IISc, Bangalore, India)

Information Science Reference • copyright 2016 • 460pp • H/C (ISBN:

9781466688537) • US $215.00 (our price)

Advanced Research on Cloud Computing Design and Applications

Shadi Aljawarneh (Jordan University of Science and Technology, Jordan)

Information Science Reference • copyright 2015 • 388pp • H/C (ISBN:

9781466686762) • US $205.00 (our price)

Handbook of Research on Computational Simulation and Modeling in

Engineering

Francisco Miranda (Instituto Politécnico de Viana do Castelo and CIDMA

of University of Aveiro, Portugal) and Carlos Abreu (Instituto Politécnico

de Viana do Castelo, Portugal)

Engineering Science Reference • copyright 2016 • 824pp • H/C (ISBN:

9781466688230) • US $420.00 (our price)

Intelligent Applications for Heterogeneous System Modeling and Design

Kandarpa Kumar Sarma (Gauhati University, India) Manash Pratim Sarma

(Gauhati University, India) and Mousmita Sarma (SpeecHWareNet (I) Pvt.

Ltd, India)

Information Science Reference • copyright 2015 • 407pp • H/C (ISBN:

9781466684935) • US $255.00 (our price)

Achieving Enterprise Agility through Innovative Software Development

Amitoj Singh (Chitkara University, Punjab, India)

Information Science Reference • copyright 2015 • 349pp • H/C (ISBN:

9781466685109) • US $225.00 (our price)

Delivery and Adoption of Cloud Computing Services in Contemporary

Organizations

Victor Chang (Computing, Creative Technologies and Engineering, Leeds

Beckett University, UK) Robert John Walters (Electronics and Computer

Science, University of Southampton, UK) and Gary Wills (Electronics and

Computer Science, University of Southampton, UK)

Information Science Reference • copyright 2015 • 519pp • H/C (ISBN:

9781466682108) • US $225.00 (our price)

Emerging Research in Cloud Distributed Computing Systems

Susmit Bagchi (Gyeongsang National University, South Korea)

Information Science Reference • copyright 2015 • 446pp • H/C (ISBN:

9781466682139) • US $200.00 (our price)

Resource Management of Mobile Cloud Computing Networks and

Environments

George Mastorakis (Technological Educational Institute of Crete, Greece)

Constandinos X. Mavromoustakis (University of Nicosia, Cyprus) and

Evangelos Pallis (Technological Educational Institute of Crete, Greece)

Information Science Reference • copyright 2015 • 432pp • H/C (ISBN:

9781466682252) • US $215.00 (our price)

Research and Applications in Global Supercomputing

Richard S. Segall (Arkansas State University, USA) Jeffrey S. Cook

(Independent Researcher, USA) and Qingyu Zhang (Shenzhen University,

China)

Information Science Reference • copyright 2015 • 672pp • H/C (ISBN:

9781466674615) • US $265.00 (our price)

Challenges, Opportunities, and Dimensions of Cyber-Physical Systems

P. Venkata Krishna (VIT University, India) V. Saritha (VIT University,

India) and H. P. Sultana (VIT University, India)

Information Science Reference • copyright 2015 • 314pp • H/C (ISBN:

9781466673120) • US $200.00 (our price)

Human Factors in Software Development and Design

Saqib Saeed (University of Dammam, Saudi Arabia) Imran Sarwar Bajwa

(The Islamia University of Bahawalpur, Pakistan) and Zaigham Mahmood

(University of Derby, UK & North West University, South Africa)

Information Science Reference • copyright 2015 • 354pp • H/C (ISBN:

9781466664852) • US $195.00 (our price)

Handbook of Research on Innovations in Systems and Software Engineering

Vicente García Díaz (University of Oviedo, Spain) Juan Manuel Cueva

Lovelle (University of Oviedo, Spain) and B. Cristina Pelayo García-

Bustelo (University of Oviedo, Spain)

Information Science Reference • copyright 2015 • 745pp • H/C (ISBN:

9781466663596) • US $515.00 (our price)

Handbook of Research on Architectural Trends in Service-Driven

Computing

Raja Ramanathan (Independent Researcher, USA) and Kirtana Raja (IBM,

USA)

Information Science Reference • copyright 2014 • 759pp • H/C (ISBN:

9781466661783) • US $515.00 (our price)

Handbook of Research on Embedded Systems Design

Alessandra Bagnato (Softeam R&D, France) Leandro Soares Indrusiak

(University of York, UK) Imran Rafiq Quadri (Softeam R&D, France) and

Matteo Rossi (Politecnico di Milano, Italy)

Information Science Reference • copyright 2014 • 520pp • H/C (ISBN:

9781466661943) • US $345.00 (our price)

Contemporary Advancements in Information Technology Development in

Dynamic Environments

Mehdi Khosrow-Pour, D.B.A. (Information Resources Management

Association, USA)
Information Science Reference • copyright 2014 • 410pp • H/C (ISBN:

9781466662520) • US $205.00 (our price)

Systems and Software Development, Modeling, and Analysis New

Perspectives and Methodologies

Mehdi Khosrow-Pour, D.B.A. (Information Resources Management

Association, USA)

Information Science Reference • copyright 2014 • 365pp • H/C (ISBN:

9781466660984) • US $215.00 (our price)

Handbook of Research on Emerging Advancements and Technologies in

Software Engineering

Imran Ghani (Universiti Teknologi Malaysia, Malaysia) Wan Mohd Nasir

Wan Kadir (Universiti Teknologi Malaysia, Malaysia) and Mohammad

Nazir Ahmad (Universiti Teknologi Malaysia, Malaysia)

Engineering Science Reference • copyright 2014 • 686pp • H/C (ISBN:

9781466660267) • US $395.00 (our price)

Advancing Embedded Systems and Real-Time Communications with

Emerging Technologies

Seppo Virtanen (University of Turku, Finland)

Information Science Reference • copyright 2014 • 502pp • H/C (ISBN:

9781466660342) • US $235.00 (our price)

Handbook of Research on High Performance and Cloud Computing in

Scientific Research and Education

Marijana Despotović-Zrakić (University of Belgrade, Serbia) Veljko

Milutinović (University of Belgrade, Serbia) and Aleksandar Belić

(University of Belgrade, Serbia)

Information Science Reference • copyright 2014 • 476pp • H/C (ISBN:

9781466657847) • US $325.00 (our price)

Agile Estimation Techniques and Innovative Approaches to Software

Process Improvement

Ricardo Colomo-Palacios (Østfold University College, Norway) Jose

Antonio Calvo-Manzano Villalón (Universidad Politécnica De Madrid,

Spain) Antonio de Amescua Seco (Universidad Carlos III de Madrid,

Spain) and Tomás San Feliu Gilabert (Universidad Politécnica De Madrid,

Spain)

Information Science Reference • copyright 2014 • 399pp • H/C (ISBN:

9781466651821) • US $215.00 (our price)

Enabling the New Era of Cloud Computing Data Security, Transfer, and

Management

Yushi Shen (Microsoft, USA) Yale Li (Microsoft, USA) Ling Wu (EMC,

USA) Shaofeng Liu (Microsoft, USA) and Qian Wen (Endronic Corp,

USA)

Information Science Reference • copyright 2014 • 336pp • H/C (ISBN:

9781466648012) • US $195.00 (our price)

Theory and Application of Multi-Formalism Modeling

Marco Gribaudo (Politecnico di Milano, Italy) and Mauro Iacono (Seconda

Università degli Studi di Napoli, Italy)

Information Science Reference • copyright 2014 • 314pp • H/C (ISBN:

9781466646599) • US $195.00 (our price)

Pervasive Cloud Computing Technologies Future Outlooks and

Interdisciplinary Perspectives

Lucio Grandinetti (University of Calabria, Italy) Ornella Pisacane

(Polytechnic University of Marche, Italy) and Mehdi Sheikhalishahi

(University of Calabria, Italy)

Information Science Reference • copyright 2014 • 325pp • H/C (ISBN:

9781466646834) • US $190.00 (our price)

Communication Infrastructures for Cloud Computing

Hussein T. Mouftah (University of Ottawa, Canada) and Burak Kantarci

(University of Ottawa, Canada)

Information Science Reference • copyright 2014 • 583pp • H/C (ISBN:

9781466645226) • US $195.00 (our price)

Organizational, Legal, and Technological Dimensions of Information

System Administration
Irene Maria Portela (Polytechnic Institute of Cávado and Ave, Portugal) and

Fernando Almeida (Polytechnic Institute of Gaya, Portugal)

Information Science Reference • copyright 2014 • 321pp • H/C (ISBN:

9781466645264) • US $195.00 (our price)

Advances and Applications in Model-Driven Engineering

Vicente García Díaz (University of Oviedo, Spain) Juan Manuel Cueva

Lovelle (University of Oviedo, Spain) B. Cristina Pelayo García-Bustelo

(University of Oviedo, Spain) and Oscar Sanjuán Martinez (University of

Carlos III, Spain)

Information Science Reference • copyright 2014 • 426pp • H/C (ISBN:

9781466644946) • US $195.00 (our price)

Service-Driven Approaches to Architecture and Enterprise Integration

Raja Ramanathan (Independent Researcher, USA) and Kirtana Raja

(Independent Researcher, USA)

Information Science Reference • copyright 2013 • 411pp • H/C (ISBN:

9781466641938) • US $195.00 (our price)

Progressions and Innovations in Model-Driven Software Engineering

Vicente García Díaz (University of Oviedo, Spain) Juan Manuel Cueva

Lovelle (University of Oviedo, Spain) B. Cristina Pelayo García-Bustelo

(University of Oviedo, Spain) and Oscar Sanjuán Martínez (University of

Oviedo, Spain)

Engineering Science Reference • copyright 2013 • 388pp • H/C (ISBN:

9781466642171) • US $195.00 (our price)

Knowledge-Based Processes in Software Development

Saqib Saeed (Department of Computer Information Systems, College of

Computer Science and Information Technology, Imam Abdulrahman Bin

Faisal University, Dammam, Saudi Arabia) and Izzat Alsmadi (Prince

Sultan University, Saudi Arabia)

Information Science Reference • copyright 2013 • 296pp • H/C (ISBN:

9781466642294) • US $195.00 (our price)

Distributed Computing Innovations for Business, Engineering, and Science

Alfred Waising Loo (Lingnan University, Hong Kong)

Information Science Reference • copyright 2013 • 369pp • H/C (ISBN:

9781466625334) • US $195.00 (our price)

Data Intensive Distributed Computing Challenges and Solutions for Large-

scale Information Management

Tevfik Kosar (University at Buffalo, USA)

Information Science Reference • copyright 2012 • 352pp • H/C (ISBN:

9781615209712) • US $180.00 (our price)

Achieving Real-Time in Distributed Computing From Grids to Clouds

Dimosthenis Kyriazis (National Technical University of Athens, Greece)

Theodora Varvarigou (National Technical University of Athens, Greece)

and Kleopatra G. Konstanteli (National Technical University of Athens,

Greece)

Information Science Reference • copyright 2012 • 330pp • H/C (ISBN:

9781609608279) • US $195.00 (our price)

Principles and Applications of Distributed Event-Based Systems

Annika M. Hinze (University of Waikato, New Zealand) and Alejandro

Buchmann (University of Waikato, New Zealand)

Information Science Reference • copyright 2010 • 538pp • H/C (ISBN:

9781605666976) • US $180.00 (our price)

Large-Scale Distributed Computing and Applications Models and Trends

Valentin Cristea (Politehnica University of Bucharest, Romania) Ciprian

Dobre (Politehnica University of Bucharest, Romania) Corina Stratan

(Politehnica University of Bucharest, Romania) Florin Pop (Politehnica

University of Bucharest, Romania) and Alexandru Costan (Politehnica

University of Bucharest, Romania)

Information Science Reference • copyright 2010 • 276pp • H/C (ISBN:

9781615207039) • US $180.00 (our price)

 701 E. Chocolate Ave., Hershey, PA 17033

Order online at http://www.igi-global.com or call 717-533-8845 x100

To place a standing order for titles released in this series, contact:

[email protected]

Mon-Fri 8:30 - 5:00 pm (est) or fax 24 hours a day 717-533-8661

 Preface
Mohammad Gouse Galety

Department of Computer Science, Samarkand International University of

Technology, Uzbekistan

Arul Kumar Natarajan

CHRIST University (Deemed), India

A. V. Sriharsha

MB University, India

In today's rapidly evolving technological landscape, efficient data

management and analysis are pivotal in numerous domains, ranging from

scientific research and finance to healthcare and beyond. As the demand for

sophisticated data processing and algorithmic solutions grows, Python has

emerged as a powerful programming language, providing a rich ecosystem

of data structures and algorithms. It is with great pleasure that we present

Advanced Applications of Python Data Structures and Algorithms, a

comprehensive compilation of cutting-edge techniques and practices that

harness the full potential of Python to tackle complex data challenges.

This book delves into the depths of Python's data structures and algorithms,

exploring their applications in diverse fields and showcasing the versatility

of this language. It is designed for both seasoned professionals seeking to

enhance their skills and beginners aspiring to master the art of data

manipulation and algorithmic problem-solving.

Chapter 1, “Fundamentals of Data Structures: Stacks, Queues, Linked Lists,

and Graphs,” provides a solid foundation by introducing fundamental data

structures and their applications.

Chapter 2, “Organizing Data Using List: A Sequential Data Structure,”

explores the intricacies of list data structures and their sequential nature,
highlighting their significance in efficient data organization.

Chapter 3, “Linear Data Structures and Their Applications,” takes a deeper

dive into linear data structures such as arrays and explores their practical

applications in various domains.

Chapter 4, “A Comprehensive Analysis of Stack and Queue Data Structures

and Their Uses,” delves into the nuances of stack and queue data structures,

showcasing their importance and usage scenarios in different contexts.

Chapter 5, “Dynamic Programming With Python,” offers an in-depth

exploration of dynamic programming techniques, empowering readers to

solve complex optimization problems efficiently.

Chapter 6, “Exploratory Data Analysis in Python,” equips readers with

essential skills to analyze and derive insights from complex datasets,

employing Python's versatile libraries.

Chapter 7, “Empowering Scientific Computing and Data Manipulation With

Numerical Python (NumPy),” emphasizes the role of NumPy in scientific

computing, highlighting its powerful data manipulation and analysis

capabilities.

Chapter 8, “Exploring Python's Powerful Data Collections,” dives into

Python's extensive collection of data structures, showcasing their

applications and benefits in diverse scenarios.

Chapter 9, “Interactive Visualization With Plotly Express,” introduces

readers to the world of interactive data visualization using Plotly Express,

enabling them to create captivating visual representations of their data.

Chapter 10, “Enhancing Big Data Analytics and Recommendation Systems

With Python: Hierarchical Clustering and Filtering Using Slicing

Technique,” focuses on applying Python in advanced analytics, particularly

hierarchical clustering and filtering techniques for big data and

recommendation systems.
Chapter 11, “Optimizing Energy Consumption in Wireless Sensor Networks

Using Python Libraries,” explores how Python libraries can be leveraged to

optimize energy consumption in wireless sensor networks, contributing to

sustainable and efficient systems.

Chapter 12, “An Exploratory Study of Python's Role in the Advancement of

Cryptocurrency and Blockchain Ecosystems,” investigates Python's role in

the evolution of cryptocurrency and blockchain technologies, shedding light

on their potential applications and challenges.

Chapter 13, “Exploring the Role of Python in Self-Supervised Contrastive

Learning for Generating Medical Imaging Reports,” examines how Python

facilitates self-supervised contrastive learning techniques in medical

imaging, paving the way for automated report generation and improved

healthcare diagnostics.

Chapter 14, “Python's Role in Predicting Type 2 Diabetes Using Insulin

DNA Sequence,” delves into the use of Python to predict and analyze type 2

diabetes using DNA sequence data, offering insights into the potential of

Python in personalized medicine and healthcare.

This book aims to contribute to data structures and algorithms by exploring

Python's capabilities and their applications in diverse domains. Delving into

advanced topics and showcasing practical implementations equips readers

with the tools to tackle complex data challenges and make informed

decisions.

We hope that Advanced Applications of Python Data Structures and

Algorithm serves as a valuable resource for researchers, data scientists,

software engineers, and enthusiasts alike, empowering them to unlock the

full potential of Python and revolutionize their approach to data

management and algorithmic problem-solving. We invite you to embark on

this enlightening journey through the pages of this book and explore the

endless possibilities that await you.

 Acknowledgment
This book is based on the research conducted on Data Structures using

Python programming. We are grateful to each other for encouraging us and

starting the book editing, preserve with it, and finally publishing it.

We would not be able to get our work done without the continual support

and vision of Elizabeth Barrantes, Assistant Development Editor; IGI

Global is just incredible.

Finally, we would like to acknowledge with gratitude the support of (the

editors) each other and the authors. This book would not have been possible

without Dr Arul Kumar.

 Introduction
This foreword serves as an introduction to Advanced Applications of Python

Data Structures and Algorithms, an extensive guide that explores the vast

capabilities of Python in the realm of data structures and algorithms. The

book exemplifies Python's growing technological significance, presenting

practical solutions to intricate problems.

Within the following pages, readers will embark on an engaging journey

encompassing a diverse range of topics that vividly showcase the power and

versatility of Python. The chapters cover various subjects, including Plotly

Express for interactive visualization, utilizing lists for data organization,

applying NumPy for scientific computing, and exploring Python's robust

data collections. This well-structured progression of knowledge empowers

readers with a profound understanding of the subject matter.

Advanced Applications of Python Data Structures and Algorithms

effectively highlights Python's applicability in various domains, such as

self-supervised contrastive learning, blockchain ecosystems, big data

analytics, and medical imaging. Moreover, the book addresses critical

concerns, including the prediction of type 2 diabetes and the optimization of

energy consumption in wireless sensor networks, thus exemplifying

Python's potential to tackle significant societal challenges.

The authors' meticulous crafting of each chapter seamlessly bridges

theoretical foundations with practical applications. This collaborative

endeavour presents readers with an invaluable resource to enhance their

comprehension of Python, data structures, and algorithms. By delving into

the depths of this book, readers will be inspired to unlock novel possibilities

and achieve excellence in their pursuits. Unlock new options and excel in

their endeavours.
 CHAPTER 1

Fundamentals of Data Structures:

Stacks, Queues, Linked Lists, and Graphs

D. Varshaa

Coimbatore Institute of Technology, India

A. Keerthana Devi

Coimbatore Institute of Technology, India

M. Sujithra

Coimbatore Institute of Technology, India

ABSTRACT

Computer science encompasses various subfields, with data structures and

algorithms being important. This chapter aims to enhance proficiency in

constructing and analyzing data structures effectively. Studying data

structures equips students with the skills to solve real-world problems using

suitable structures. The study of data structures offers diverse programming

techniques for complex problems yet introduces challenges due to intricate

complexities that reshape program architecture. Covering all aspects in a

single semester is impractical, as data structures receive extensive attention

in graduate, upper-division, and lower-division programs. This text is an

introductory resource for the stack, queue, linked list, graph, trees,

searching, and sorting algorithms. It also offers insights into their Python

implementation and its complexities, applications, and sample code.

INTRODUCTION
A data structure is a meticulously designed framework for organizing,

manipulating, retrieving, and storing data. It encompasses a variety of

simple and intricate forms, all devised to arrange data in a manner suited to

specific use cases. Users can conveniently access and employ the required

information using data structures. These structures facilitate data

organization in a manner comprehensible to machines and individuals. A

data structure can be either developed or selected to facilitate data storage

for various operations. In certain instances, the core operations of an

algorithm and the design of the data structure are closely intertwined. Each

data structure incorporates details about the values and connections of the

data, and in rare cases, it may include functions capable of modifying the

data.

Conventional elementary data types, such as integers or floating-point

numbers, often prove inadequate for effectively expressing the logical intent

underlying data processing and application. However, programs involved in

receiving, manipulating, and outputting information must be aware of how

data should be organized to streamline the processing tasks. Data structures

facilitate efficient utilization, persistence, and sharing by logically

amalgamating individual data elements. They provide a formal model that

delineates the arrangement of data elements. Data structures serve as the

foundation for increasingly complex applications. These structures are

constructed by aggregating data elements into logical units that represent

abstract data types relevant to the algorithm or application at hand, thereby

enabling the development of sophisticated systems.

STACK

A stack is a linear data structure that embodies the Last-In-First-Out (LIFO)

principle (Carullo, 2020). Under this principle, the item most recently

added to the stack is the first to be removed. The stack can be likened to a

collection of plates, wherein the plate most recently placed on top is the one

to be taken when selecting. Moreover, accessing the plate at the bottom

necessitates removing all the plates above it. The stack data structure

operates similarly, adhering to this analogy.

Operations On Stack

The following are vital operations associated with a stack data structure:

● Push: This operation involves adding an element to the top of the

stack.

● Pop: This operation entails removing an element from the top of the

stack.

● IsEmpty: This operation allows checking whether the stack is

empty or not.

● IsFull: This operation verifies whether the stack has reached its

maximum capacity.

● Peek: This operation permits retrieving the value of the top element

without removing it.

Stack Implementation Using Python's Built-in List

Python's inherent data structure, the list, can be employed as a stack by

utilizing specific methods. In place of the conventional “push()” operation,

the “append()” method is utilized to add elements to the top of the stack,

ensuring adherence to the Last-In-First-Out (LIFO) principle. Similarly, the

“pop()” method facilitates the removal of elements from the stack in the

prescribed LIFO order. The stack implementation using Python is given in

Figure 1.

Figure 1. Stack Implementation using Python

QUEUE

A queue is defined as a linear data structure characterized by open ends and

the execution of operations according to the First In, First Out (FIFO)

principle (Alcoz et al., 2020). Within this definition, a queue is a list where

all additions occur at one end, and all deletions occur at the other. The

operation is initiated on the element first pushed into the Queue. The

general representation of Queue operations is given in Figure 2.

Figure 2. Queue: Operations

FIFO Principle of Queue

A queue exhibits similarities to a line of individuals awaiting ticket

purchase, where the individual at the front of the line is the first to receive

service, adhering to the “first come, first served” principle. The front of the

Queue, also known as the head, represents the position of the entry prepared

to be served, while the rear, referred to as the tail, denotes the position of

the most recently added entry. The First-In-First-Out (FIFO) characteristic

of a queue is visually illustrated in Figure 3.

Figure 3. FIFO property of Queue

Characteristics of Queue
The characteristics of a queue data structure encompass the following key

aspects:

● Versatility: Queues can handle a wide range of data types.

● Accessibility: Both ends of the Queue, the front, and the rear, are

easily accessible for efficient operations.

● Efficiency: Queues demonstrate swift and smooth movement,

facilitating effective data management.

Queue Representation Using Array

Queues can be represented using an array, similar to how stacks are

implemented (Streib et al., 2017) (Patel, 2018). In this particular

representation, the following variables are employed:

• Queue: This is the name assigned to the array that holds the

elements of the Queue.

• Front: This variable indicates the index in the array where the first

element of the Queue is stored.

• Rear: This variable denotes the index in the array where the last

element of the Queue is stored.

LINKED LIST

A linked list comprises a sequence of interconnected data structures

(Karimov, 2020) (Wiener, 2022). It is a collection of links with associated

data referred to as elements. The relationship between one link and another

is established through a connection known as “Next,” whereby each link

references the following link in the list. The linked list represents the second

most commonly utilized data structure, surpassed only by the array. To

comprehend the concept of a linked list, it is essential to familiarize oneself

with the following key terms:

• Link: Each link within the linked list can store data elements.

• Next: This attribute of each link denotes the reference to the

following link in the list, referred to as “Next.”

 • LinkedList: A Linked List encompasses the connection link to the

initial link in the sequence, designated as “First.”

Linked List Representation

The linked list can be perceived as a succession of nodes, wherein each

node is connected to the next node. This visual representation of the linked

list is illustrated in Figure 4, showcasing the overall structure.

Figure 4. Linked list representation

The linked list representation is given in Figure 4. Here, the primary

constituent of Linked Lists is the initial link element, denoted as “first.”

Each link within the Linked List encompasses a data field or fields, and a

link field is known as “next.” Every link establishes a connection with the

subsequent link in the sequence by utilizing its following link. To signify

the termination point of the list, the final link is endowed with a null link.

Types of Linked Lists (Medjedovic & Tahirovic, 2022):

• Simple Linked List: Only forward item navigation is feasible in this

type of linked list. It allows traversal in a single direction.

• Doubly Linked List: Unlike a simple linked list, a doubly linked list

enables movement in both forward and backward directions. Each item

within the list links to the next and previous elements.

• Circular Linked List: In a circular linked list, the first element is

connected to the last element as its predecessor, while the last element

is linked to the first element as its successor. This circular arrangement

allows for seamless cyclic traversal of the list.

Linked List: Fundamental Operations

The following are the essential operations supported by a linked list:

• Insertion: This operation involves adding a new element to the

beginning of the list, effectively updating the list's starting position.

• Deletion: The deletion operation removes the first member from the

list, resulting in the adjustment of the list's structure.

• Display: The display operation presents the complete contents of the

list, enabling comprehensive visualization of its elements.

• Search: By utilizing a specified key, the search operation aims to

locate a particular element within the list.

• Delete: This operation involves the removal of an element from the

list using a specified key, thereby eliminating the corresponding entry

from the list's structure.

SINGLY LINKED LIST

Insertion Operation

Adding a new node to a linked list necessitates multiple steps. In this

context, the process will be examined through diagrams (Almadhoun &

Parham-Mocello, 2023). A node is to be created employing the identical

structure, followed by identifying the appropriate position for its placement.

The procedure for inserting a node in a singly linked list is illustrated in

Figure 5.

Figure 5. Singly linked list: Insertion operation

Deletion Operation

The deletion operation involves a sequence of steps that will be explained

through visual representations. Initially, search algorithms are utilized to

locate the target node within the linked list for removal. Subsequently, the

necessary adjustments to the list's links are made. The node preceding the

target node, commonly called the “left” node, is redirected to the node

succeeding the target node. This ensures the continuity of the list's

connectivity. Furthermore, the link associated with the target node is

eradicated, eliminating the target node from the list. It becomes inaccessible

via the preceding or succeeding nodes. The subsequent consideration

pertains to the object stored within the target node. Based on the

requirements, appropriate measures, such as memory deallocation, can be

undertaken to remove the target node altogether. Lastly, after the deletion

process, the removed node no longer exists within the linked list, and its

memory allocation can be managed accordingly. Figure 6 illustrates the

deletion operation in the singly linked list.

Figure 6. Singly linked list: Deletion operation

Reverse Operation

Reversing the linked list is executed with meticulous attention to detail,

ensuring a complete reversal of the elements' order. To accomplish this, the

last node in the list is modified to point back to the head node, effectively

reversing the direction of the linked list. To commence the reverse

operation, the traversal begins at the head and continues until the list ends.

At this point, the pointer of the last node is adjusted to reference its

preceding node, facilitating the reversal. Verifying that the last node is not
the final node in the original list is imperative. To ensure this, a temporary

node is introduced, resembling the head node and pointing to the last node.

Subsequently, all nodes on the left side of the list are iteratively adjusted to

point to their preceding nodes, sequentially reversing the pointers.

Except for the first node adjacent to the head node, all subsequent nodes are

modified to point to their predecessor, effectively updating their successor.

Consequently, the first node is assigned a NULL pointer. To complete the

reversal process, the head node is updated to point to the new first node,

utilizing the temporary node to facilitate this adjustment. As a result of

these particular operations, the linked list undergoes a complete reversal,

with the elements appearing in the reverse order compared to their original

configuration. Figure 7 demonstrates the reverse operation in the linked list.

Figure 7. Singly linked list: Reverse operation

DOUBLY LINKED LIST

A variant of a linked list, known as a doubly linked list, provides enhanced

navigation capabilities in both forward and backward directions compared

to a single linked list (Besta et al., 2019). The fundamental terminology

associated with the concept of a doubly linked list is outlined below.

● Link: Each link is a container for storing data elements within a

linked list.

o Next Link: Every link in a linked list references the following

link.

o Prev Link: Each link in a linked list references the preceding

link.
 ● LinkedList: A LinkedList comprises a connecting link to the First

and Last links, representing the initial and final links in the list,

respectively.

Doubly Linked List: Representation

The doubly linked list representation is given in Figure 8. The following are

the essential considerations:

● The doubly linked list consists of the first and last link elements.

● Each link encompasses one or more data fields beside the next and

prev link fields.

● Each link establishes a connection with the following link through

its next link.

● Each link establishes a connection with the preceding link through

its prev link.

● The concluding link in the list is indicated by a null link, denoting

the end of the list.

Figure 8. Doubly linked list representation

CIRCULAR LINKED LIST

A circular linked list is a variation of a linked list in which the first element

points to the last element, and the last element points back to the first

element. This circular structure can be formed from a singly linked or

doubly linked list (Yang & Li, 2020; Sher, 2004).

Singly Linked List as Circular

In the context of a singly linked list, it is noteworthy that the next pointer of

the final node is directed toward the initial node. Figure 9 depicts the Singly

Linked List represented as a circular structure.

Figure 9. Singly linked list as circular

Doubly Linked List as Circular

Within a doubly linked list, the establishment of a circular direction is

achieved through the assignment of the previous pointer of the initial node

to the last node, and correspondingly, the next pointer of the last node is

directed towards the first node. Figure 10 showcases the Doubly Linked List

represented as a circular structure.

The following are the key points to consider based on the given an example:

● In both a singly linked list and a doubly linked list, the “next”

pointer of the last link points to the first link in the list.

● In the case of a doubly linked list, the “previous” pointer of the first

link points to the last item.

Figure 10. Doubly linked list as circular

Fundamental Operations of a Circular List

A circular list supports the following significant operations:

● Insertion: This operation adds a new element to the beginning of

the list.

● Removal: This operation removes an element from the beginning of

the list.

● Display: This operation displays the contents of the list when

selected.

GRAPH

A group of nodes with data and connections to other nodes makes up a

graph data structure (Chen et al., 2020). Let us use an illustration to try to

comprehend this. Everything on Facebook is a node. Everything with data

is a node, including User, Photo, Album, Event, Group, Page, Comment,

Story, Video, Link, and Note. Every connection between two nodes is an

edge. If you share a photo, join a group, “like” a page, etc., that relationship

gains a new edge.

Figure 11. Example of graph data structure

An example of graph data structure is given in Figure 11. The entirety of

Facebook is represented by the collection of nodes and edges, constituting a

graph data structure utilized for data storage (Souravlas et al., 2021). A

graph (V, E) is a data structure comprising the following components:

• A collection of vertices denoted as V.

• A collection of edges is denoted as E, wherein each edge is

represented as an ordered pair of vertices (u, v).

Terms in Graph data structure

 ● Adjacent: If there is an edge between two vertices, they are said to

be adjacent.

● Path: A path is a sequence of edges that allows us to move from

vertex A to vertex B.

● Directed Graph: A graph in which an edge (u,v) does not always

imply the presence of an edge (v, u). Arrows represent edges in such a

graph to indicate their direction.

Graph Representation

Graphs can be represented in two ways:

1. Adjacency Matrix: An adjacency matrix is a two-dimensional array

of size V x V, where V represents the number of vertices in the graph.

Each row and column within the matrix correspond to a vertex in the

graph. If the value of any element a[i][j] is 1, it signifies the existence

of an edge between vertex i and vertex j.

2. Adjacency List: An adjacency list is an array of linked lists

representing a graph. The index of the array corresponds to a vertex,

and each element within the linked list associated with that index

represents the other vertices connected to the corresponding vertex by

an edge.

Graph Operations

Graph operations encompass the following functionalities:

• Element Existence Check: This operation verifies the presence of

an element within the graph.

• Graph Traversal: This operation involves traversing the graph to

visit and explore its vertices and edges.

• Element Addition: This operation facilitates the addition of

elements, including vertices and edges, to the graph.

• Path Finding: This operation determines the path between two

specified vertices within the graph.

TREE

A tree is a nonlinear hierarchical data structure comprising nodes linked

together by edges (Dutta, 2020) (Piatkowski, 2020). An example of tree data

structure is given in Figure 12.

Figure 12. A tree data structure

Binary Tree

A binary tree is a tree data structure in which each parent node has no more

than two children. A binary tree node comprises three components: data

item, address of the left child, and address of the right child.

Full Binary Tree

A full Binary tree is a binary tree in which each parent node/internal node

has two or no children. Figure 13 displays a full binary tree.

Figure 13. Full binary tree

Python Implementation

Figure 14 showcases the Python implementation of a full binary tree.

Figure 14. Python implementation of full binary tree

Perfect Binary Tree

A perfect binary tree is one in which each internal node has exactly two

child nodes, and all leaf nodes are on the same level. An example of perfect

binary tree is given in Figure 15.

Figure 15. Perfect binary tree

Python Implementation

The Python implementation of a perfect binary tree is presented in Figure

16.

Figure 16. Python implementation of perfect binary tree

Complete Binary Tree

In contrast to a full binary tree, a complete binary tree exhibits two

significant distinctions. Firstly, a complete binary tree mandates that each

level be filled, necessitating the presence of all nodes at each respective

level. Secondly, each leaf component of a complete binary tree must lean

towards the left, signifying that the left subtree of every node is populated

before the right subtree. It is crucial to acknowledge that while a complete

binary tree satisfies the conditions above, it does not invariably qualify as a
full binary tree. The absence of a right sibling for the last leaf element in a

complete binary tree indicates the potential absence of a right-sided child

node. Figure 17 displays the complete binary tree.

Figure 17. Complete binary tree

Python Implementation

Figure 18 presents the Python implementation of a complete binary tree.

Figure 18. Python implementation of complete binary tree

Balanced Binary Tree

It is a binary tree in which each node's left and right subtree height

differences are either 0 or 1. Balanced binary tree is given in Figure 19.

Figure 19. Balanced binary tree with depth represented by df

Python Implementation

Figure 20 describes the Python implementation of a Balanced Binary Tree.

Figure 20. Python implementation of balanced binary tree

Binary Search Tree

Binary search trees are a data structure designed for efficiently managing

sorted lists of numbers. The term “binary” stems from the fact that each tree

node can have a maximum of two children. This structure enables quick

searching of numbers, with a time complexity of O(log(n)), hence the name

“search tree.” By leveraging the properties of binary search trees, one can

determine the presence or absence of a number with significant efficiency.

The following are the properties of the Binary Search Tree:

 ● The first characteristic of a binary search tree (BST) is that the root

node is greater than all nodes of the left subtree.

● The second characteristic states that every node in the right subtree

of a BST exceeds the root node in value.

● The third characteristic of a BST is that each node's two subtrees are

also BSTs, meaning they possess the first two characteristics.

AVL Tree

The AVL tree, recognized as a self-balancing binary search tree, includes

supplementary data within each node, referred to as the balance factor. This

factor assumes one of three possible values: -1, 0, or +1. The nomenclature

“AVL” originates from its creators, Georgy Adelson-Velsky and Landis.

The AVL tree is depicted in Figure 21.

Figure 21. AVL tree

B-Tree

A distinctive type of self-balancing search tree, referred to as a “B-tree,”

incorporates the capability for each node to accommodate multiple keys and

children. It is also recognized by the term “height-balanced m-way tree” as

an alternative designation. The graphical depiction of the B-Tree is

presented in Figure 22.

Figure 22. B-tree

Properties:

● For every node x, the keys are stored in ascending order.

● A Boolean value x. The leaf is present within each node,

representing its status as a leaf node when set to true.

● In the case of a tree with an order of n, an internal node can hold a

maximum of n - 1 keys, accompanied by a pointer to each child.

● Each node can possess a minimum of n/2 children and a maximum

of n children except for the root.

● All leaves share an equal depth, also known as the height (h) of the

tree.

● The root must have at least two children and contain one key.

● If n is greater than or equal to 1, then for any B-tree with n keys, a

height (h) and a minimum degree (t) of at least 2, the inequality h ≥

log t(n+1)/2 holds true.

B+Tree

A B+ tree constitutes an advanced variant of a self-balancing tree structure

wherein all the values reside within the leaf level.

Properties:

1. All leaves are at the same level.

2. The root has at least two children.

 3. Each node except the root can have a maximum of m children and at

least m/2 children.

4. Each node can contain a maximum of m - 1 key and a minimum of

⌈m/2⌉ - 1 key.

Red-Black Tree

The Red-Black tree is a self-balancing binary search tree that incorporates

an additional bit within each node to indicate its color, either red or black.

Properties:

● The Red/Black Property states that each node is assigned a red or

black color.

● According to the Root Property, the tree's root node is always black.

● The Leaf Property specifies that all leaf nodes (NIL nodes) in the

tree are black.

● As per the Red Property, if a node is colored red, all its children

must be colored black.

● The Depth Property asserts that for any given node, every direct

path from that node to any of its descendant leaves contains an equal

number of black nodes.

SEARCHING

Searching is finding a specific element within a data collection (Mehlhorn,

2013; Kochenderfer & Wheeler, 2019).

Linear Search

In a sequential search technique such as linear search, the search process

initiates from one end of the list and progressively examines each element

until the target element is encountered. Linear search is recognized as the

most elementary and direct search algorithm.

Binary Search
The process of determining the position of an element within a sorted array

can be accomplished by utilizing a search technique known as binary

search. This approach involves consistently searching the middle element of

the array for the desired element. The binary search algorithm can be

implemented through two distinct methods, elucidated as follows:

● Iterative Method

● Recursive Method

Python Implementation

The Python implementation of the Binary Search using the iterative method

is illustrated in Figure 23.

Figure 23. Binary search

SORTING
Sorting is arranging elements in a specific order within a data collection.

Various sorting algorithms are available to achieve this task efficiently.

Bubble Sort

Bubble Sort is a sorting algorithm that analyzes adjacent elements and

swaps them iteratively until the desired order is achieved (Inayat, 2021).

Python Implementation

Figure 24 provides the Python implementation of the bubble sort algorithm.

Figure 24. Bubble sort

Selection Sort

The selection sort, a well-known sorting algorithm, follows a procedure

where the smallest element from an unsorted list is chosen during each

iteration and subsequently inserted at the beginning of the list (Salihu et al.,

2022).

Python Implementation
Figure 25 presents the Python implementation of the Selection sort

algorithm.

Figure 25. Selection sort

Insertion Sort

Insertion sort is an efficient sorting algorithm that, following each iteration,

strategically inserts an unsorted element into its correct position within the

sorted portion of the array (Arslan, 2022).

Python Implementation

The Python implementation of the Insertion sort algorithm can be found in

Figure 26.

Figure 26. Insertion sort

Merge Sort

Merge sort is a highly utilized sorting algorithm that is founded upon the

fundamental concept of the Divide and Conquer algorithm. It involves

decomposing a complex problem into multiple smaller subproblems,

individually resolving each subproblem and subsequently integrating the

sub-solutions to obtain the final complete solution.

Python Implementation

Figure 27 presents the Python implementation of the Merge sort algorithm.

Figure 27. Merge sort

Quick Sort

Quicksort is a divide-and-conquer sorting algorithm that selects a pivot

element from the array to divide it into subarrays. During the partitioning

process, the pivot element is positioned such that elements greater than the

pivot are placed on the right side, while elements smaller than the pivot are
placed on the left side. The same partitioning method is applied recursively

to the left and right subarrays until each subarray contains only one

element. At this stage, the individual elements are sorted, and combining

them results in a sorted array.

Python Implementation

Figure 28 presents the Python implementation of the Quick Sort algorithm.

Figure 28. Quick sort

Counting Sort

Counting Sort is a sorting technique that organizes the elements of an array

based on the frequency of each distinct element. The sorting process is

facilitated by mapping the count of each element to an index in an auxiliary

array, which serves as a container for the counts.

Python Implementation

The Python implementation of the Counting Sort algorithm is illustrated in

Figure 29.

Figure 29. Counting sort

Radix Sort

Radix Sort is a sorting technique that involves grouping the individual digits

of the same place value before sorting the elements. Subsequently, the

components are arranged in either ascending or descending order.

Python Implementation
Figure 30 showcases the Python implementation of the Radix Sort

algorithm.

Figure 30. Radix sort

Bucket Sort

The bucket sort technique involves partitioning the elements of an unsorted

array into multiple buckets. Subsequently, each bucket is sorted either by

employing the recursive application of the bucket algorithm or by utilizing

any suitable sorting algorithm.

Python Implementation

The Python implementation of the Bucket Sort algorithm is presented in

Figure 31.

Figure 31. Bucket sort

Heap Sort

The heap sort algorithm is prominent in computer programming due to its

popularity and efficiency. This sorting technique necessitates

comprehending two distinct data structures: arrays and trees.

Python Implementation

Figure 32 provides the Python implementation of the Heap Sort algorithm.

Figure 32. Heap sort

Shell Sort

Shell Sort is a well-known sorting algorithm that can be considered a

generalized version of the insertion sort algorithm. It operates by sorting

elements that are initially far apart from each other, gradually decreasing the

interval between the components to be sorted.

Python Implementation

The Python implementation of the Shell Sort algorithm is illustrated in

Figure 33.

Figure 33. Shell sort

CONCLUSION

The chapter introduces the concept of data structures, which are frameworks

used for organizing, manipulating, retrieving, and storing data. Data

structures facilitate efficient access and utilization of data by machines and

individuals. The chapter focuses on three types of data structures: stacks,

queues, and linked lists. A stack is a last-in-first-out (LIFO) data structure

with operations like push, pop, isEmpty, isFull, and peek. A queue is a first-

in-first-out (FIFO) data structure with similar operations. Linked lists

consist of interconnected nodes, and there are different types like singly

linked lists, doubly linked lists, and circular linked lists. The chapter also

briefly mentions the graph data structure, consisting of nodes and edges

representing relationships between data elements.

REFERENCES

Alcoz, A. G., Dietmüller, A., & Vanbever, L. (2020, February). SP-

PIFO: Approximating Push-In First-Out Behaviors using Strict-

Priority Queues. In NSDI (pp. 59-76). Academic Press.

Almadhoun, E., & Parham-Mocello, J. (2023). Students’ difficulties

with inserting and deleting nodes in a singly linked list in the C

programming language. Journal of Computer Languages, 74,

101184. doi:10.1016/j.cola.2022.101184
Arslan, B. (2022). Search and Sort Algorithms for Big Data

Structures. Academic Press.

Besta, M., Peter, E., Gerstenberger, R., Fischer, M., Podstawski,

M., Barthels, C., . . . Hoefler, T. (2019). Demystifying graph

databases: Analysis and taxonomy of data organization, system

designs, and graph queries. arXiv preprint arXiv:1910.09017.

Carullo, G. (2020). Data Structures. Implementing Effective Code

Reviews: How to Build and Maintain Clean Code, 27-42.

Chen, Z., Wang, Y., Zhao, B., Cheng, J., Zhao, X., & Duan, Z.

(2020). Knowledge graph completion: A review. IEEE Access :

Practical Innovations, Open Solutions, 8, 192435–192456.

doi:10.1109/ACCESS.2020.3030076

Dutta, S., Chowdhury, A., Debnath, I., Sarkar, R., Dey, H., &

Dutta, A. (2020). The role of data structures in different fields of

computer-science: A review. Journal of Mathematical Sciences &

Computational Mathematics, 1(3), 363–373.

doi:10.15864/jmscm.1310

InayatZ.SajjadR.AnamM.YounasA.HussainM. (2021, November).

Analysis of Comparison-Based Sorting Algorithms. In 2021

International Conference on Innovative Computing (ICIC) (pp. 1-

8). IEEE.

Karimov, E. (2020). Linked Lists. Data Structures and Algorithms

in Swift: Implement Stacks, Queues, Dictionaries, and Lists in Your

Apps, 41-54.

Kochenderfer, M. J., & Wheeler, T. A. (2019). Algorithms for

optimization. MIT Press.

Medjedovic, D., & Tahirovic, E. (2022). Algorithms and data

structures for massive datasets. Simon and Schuster.

Mehlhorn, K. (2013). Data structures and algorithms 1: Sorting and

searching (Vol. 1). Springer Science & Business Media.

Patel, M. (2018). Data Structure and Algorithm With C.

Educreation Publishing.

Piatkowski, J. (2020). The Conditional Multiway Mapped Tree:

Modeling and Analysis of Hierarchical Data Dependencies. IEEE

Access : Practical Innovations, Open Solutions, 8, 74083–74092.

doi:10.1109/ACCESS.2020.2988358

SalihuA.HotiM.HotiA. (2022, December). A Review of

Performance and Complexity on Sorting Algorithms. In 2022

International Conference on Computing, Networking,

Telecommunications & Engineering Sciences Applications

(CoNTESA) (pp. 45-50). IEEE.

doi:10.1109/CoNTESA57046.2022.10011382

Sher, D. B. (2004). A simple implementation of a queue with a

circularly linked list. SIGCSE Bulletin, 36(3), 274–274.

doi:10.1145/1026487.1008112

Souravlas, S., Anastasiadou, S., & Katsavounis, S. (2021). A survey

on the recent advances of deep community detection. Applied

Sciences (Basel, Switzerland), 11(16), 7179.

doi:10.3390/app11167179

Streib, J. T., Soma, T., Streib, J. T., & Soma, T. (2017). Stacks and

Queues Using References. Guide to Data Structures: A Concise

Introduction Using Java, 173-198.

Wiener, R. (2022). Queues and Lists. In Generic Data Structures

and Algorithms in Go: An Applied Approach Using Concurrency,

Genericity and Heuristics (pp. 187–236). Apress. doi:10.1007/978-

1-4842-8191-8_6

YangY.LiX. (2020, October). Design and Implementation of

Sliding Window with Circular Linked List for Dynamic Data Flow.

In Proceedings of the 4th International Conference on Computer

Science and Application Engineering (pp. 1-5).

doi:10.1145/3424978.3425152
 CHAPTER 2

Organizing Data Using Lists:

A Sequential Data Structure

Saleem Raja Abdul Samad

University of Technology and Applied Sciences, Shinas, Oman

N. Arulkumar

CHRIST University (Deemed), India

Justin Rajasekaran

University of Technology and Applied Sciences, Shinas, Oman

R. Vinodini

Little Angels Institute, Karur, India

Pradeepa Ganesan

University of Technology and Applied Sciences, Shinas, Oman

ABSTRACT

Computer programming aims to organize and process data to get the desired

result. Software developer chooses a programming language for application

development based on the data processing capabilities of the language. The list

is one of the sequential data structures in Python. A list is limited to a particular

data type, such as numbers or strings. Occasionally, a list may include data of

mixed types, including numbers and strings. Elements in the list can be

accessed by using the index. Usually, a list's elements are enclosed within

square brackets and divided using commas. The list may be referred to as a

dynamic-sized array, which denotes that its size increases as additional data is

added and that its size is not predefined. List data structure allows repetition;

hence, a single data item may appear several times in a list. The list assists us in

solving several real-world problems. This chapter deals with the list's creation
and manipulation, the complexity of processing the list, sorting, stack, and

queue operations.

THE BASICS OF LIST

Python has a sequence data type called List. It can ably hold multiple elements

of different data types. However, it may have multiple elements of the same data

type in practice. The elements in a list are separated by commas (Erciyes,

2021). All the elements in a list must be placed inside the square bracket [], as

shown in Figure 1.

Figure 1. Three items list

Accessing an Item in the List

A list is an ordered sequence of items, and each element is accessed by its

index, as shown in Figure 2. The index must be an integer. The first item is

indexed with [0], the second with [1], and so on.

Figure 2. Accessing an item in the list using an index

Python allows negative indexing for its sequences, as shown in Figure 3. The

last item is represented by the index of -1, the second last by -2, and so on.

Figure 3. Accessing an item in the list using a negative index

Changing an Item in the List

The List is mutable (Oliphant, 2007). That is, items in the List can be modified

even after the creation of the List. After creating a list, objects may be added,

removed, shifted, and moved about at will. Python provides several methods for

modifying lists. To change an item or set of items in a list, we can use an

assignment operation on the specific index or range of index positions (Schwarz

et al., 2020). Figure 4 illustrates modifying an item or a set of items within a list

by employing an index. This visualization showcases the steps involved in

altering the content of the List, providing valuable insights into the

manipulation of data using indexing techniques.

Figure 4. Changing an item/set of items in a list using an index

Adding Items to the List

The append() method is a popular option for adding a single item to an existing

list. Using the append() function, adding an item to the end of a list is

straightforward. This procedure is efficient and uncomplicated, preserving the

original List's order. Attach() is particularly useful when adding a single

element without modifying the existing objects.

In situations where multiple items are added to a list, the extend() method is

applicable. extend(), unlike append(), permits the addition of numerous items to

a list, effectively expanding its size. With the extend() function, it is possible to

append multiple elements in a single operation, providing flexibility and

efficiency. This method is advantageous when sequentially adding multiple

items or combining two listings into one.

Although both methods can add items to a list, their functionality and

application are distinct. The append() method excels at adding a single item to

the end of a list, making the operation straightforward and concise. On the other

hand, the extend() method is more versatile because it permits adding multiple

elements at once. It is especially advantageous when multiple items or merge

lists must be appended (Martelli, 2006).

Figure 5. Adding an item to a list using the append() method

In Figure 5, the item 'srilanka' is appended to the end of list 'country'. The

extend() adds a specified set of items at the end of the existing List, as given in

Figure 6.

Figure 6. Adding a set of items into an existing list using extend() method

Here the list “country_2” items are added to the end of the existing list

'country'. Table 1 showcases the Membership Operator in Python, illustrating its

functionality and usage in different scenarios.

Table 1. Membership operator in Python

in Returns True if the specified value is found in the List.

not in Returns True if the specified value is not found in the List.

Searching and Sorting the Items

Python's membership operator can be used to find the item in the List, as shown

in Figure 7. Python has two membership operators (Dubois et al., 1996).

Figure 7. Searching an item in a list

Figure 8 illustrates how the default ordering of the List's contents by Python's

built-in sort() method is alphanumerical and ascending (Lutz Mark, 2013).

Figure 8. Sorting the items in a list

Time Complexity

Index lookup and assignment require a constant amount of time regardless of

how long the List is; hence they are O(1). Append has constant time

complexity, i.e., O(1). Extend has a time complexity of O(k). Here, k is the list

length that needs to be increased. O(n log n) time complexity is required for

sorting (Navghare et al., 2020).

STACK

A fundamental data structure for effective data organization is the Stack. It is an

ordered collection of items (data) that are accessed on a LAST IN FIRST OUT

(LIFO) basis (Saabith et al., 2019). The end of the Stack, where items are added

or removed, is considered the TOP of the Stack. The other end of the Stack is
known as the BASE or BOTTOM. The most recently inserted item is always on

top of the Stack and can be removed first. Push and Pop are the terms for adding

items to stacks and removing them from stacks.

Figure 9. PUSH and POP operation in stack

The operations PUSH and POP in the Stack are shown in Figure 9 (Hetland,

2017). The top of the plate rack is for adding the new plates. Plates are removed

from the bottom of the plate rack. Table 2 contains the most fundamental

operations for a Stack.

The push(n) operation is an essential list function, as it permits the insertion of

an item on top of the stack. The argument 'n' when utilizing push(n) represents
the item to be added to the list. This operation guarantees that the new item is

placed at the top of the array, where it can be accessed or modified as necessary.

The pop() operation, on the other hand, functions to remove the uppermost

element from the stack. This operation is beneficial when it is necessary to

extract or delete a list item. The prior element at the top of the list is removed

when pop() is executed, making the element below it the new element at the

top.

The isempty() function is utilized to determine whether a stack is empty or not.

This operation determines whether or not the array contains elements. If the

array is empty, this function returns true to indicate that no items are present. In

contrast, if the stack is not empty, the function returns false, indicating the

presence of list elements. When working with collections, it is frequently

necessary to know how many items they comprise. This is where the size()

function is useful. By utilizing size(), programmers can retrieve the number of

stack elements. This information is especially helpful for memory management

and determining the overall extent of the data structure. Lastly, the top()

function provides a convenient method for accessing the stack's uppermost

element without modifying the list. This operation retrieves the value of the

item at the top of the stack, allowing developers to investigate or use it without

modifying the structure of the list.

Table 2. Stack operations

Time
No Operation Description
Complexity

To insert an item on top of the Stack. The

1 push(n) O(1)
argument 'n' is an item to push.

It removes the Stack's topmost element of the

2 pop() O(1)
Stack.

3 isempty() To check if the Stack is empty. O(1)

4 size() It returns the number of items in the Stack. O(1)

5 top() Returns the Stack's topmost element. O(1)

Code for stack operations using the List

# Empty List
country=[]
# Add an item on top of the Stack.
def push(x):
country.append(x)
print(“push operation result:”, country)
# Remove an item from the top of the Stack.
def pop():
if (len(country)>0):
x=country.pop()
print(“removed element:”, x)
print(“pop operation result:”, country)
# Check whether or not the Stack is empty.
def empty():
if (len(country)==0):
print(“Stack is empty”)
else:
print(“Stack is NOT empty”)
# Find the number of items in the Stack.
def size():
print(“size() operation result:”, len(country))
# Find the top item in the Stack
def top():
if (len(country)>0):
x=country[-1]
print(“top of the stack is:”, x)
while(1):
print (“Stack Operations”)
print (“****************”)
print (“1. PUSH”)
print (“2. POP”)
print (“3. IS EMPTY”)
print (“4. SIZE”)
print (“5. TOP”)
print (“6. EXIT”)
 inp=int(input(“Enter your choice number”))
if(inp > 6 | inp < 1):
print(“Wrong choice so try again “)
else:
if(inp == 1):
name=input(“Enter the country name to
added in the stack:”)
push(name)
if(inp == 2):
pop()
if(inp == 3):
empty()
if(inp == 4):
size()
if(inp == 5):
top()
if (inp == 6):
break
print(“Completed >>>>”)

QUEUE

A queue is a sequential data structure that stores items (data) in an order where

new items are added at one end, generally called the “rear.” The removal of an

existing item occurs at the other end, commonly called the “front.”. A newly

added item to a queue will be at the rear of that queue. The first item added to

the queue will be at the front. This concept is FIFO, which stands for “first in,

first out” (Popescu & Bîră, 2022). ENQUEUE and DEQUEUE are used to add

and delete items from the queue (Kobbaey et al., 2022) (Hunt, 2019).

Figure 10. ENQUEUE operation

Figure 11. DEQUEUE operation
Figure 10 depicts the queue's ENQUEUE, and Figure 11 presents the

DEQUEUE operations. At a ticket counter, people are standing in a queue. The

first person to join the queue will also be the first to depart it. Table 3 lists the

most frequent queue operations.

Enqueue(n), dequeue(), isempty(), size(), front(), and rear() are the operations

conducted on a queue utilizing a list. The enqueue(n) operation appends a new

item to the end of the queue, as indicated by the 'n' operand. It ensures that the

newly added item is placed at the end of the sequence, preserving the order in

which elements were added.

The dequeue() operation, on the other hand, removes the first item from the

queue. By eliminating the item from the front of the queue, the FIFO principle

is maintained by promoting the next item to the front. To determine whether the

queue is vacant, we can use the isempty() function. It returns a Boolean value

signifying whether the queue is vacant or contains elements. The size() function

returns the quantity of elements presently in the queue. It provides valuable

information regarding the queue's size and capacity, allowing us to manage and

optimize the storage accordingly.

Moreover, the front() function permits access to the first item in the queue. It

returns the element without removing it, which is beneficial when we need a

sneak look at the subsequent item to be processed. Similarly, the rear()

operation permits access to the final item in the queue. It returns the element

without removing it and provides details about the most recently inserted item.

Table 3. The most common queue operations

Time
No Operation Description
Complexity

It adds a new item to the queue's end. Its 'n'

1 enqueue(n) O(1)
argument denotes an item for addition.

It removes the first item from the queue. It

2 dequeue() O(1)
takes no parameters and returns the item.

3 isempty() To check if the queue is empty. O(1)

It has the details of the number of items in

4 size() O(1)
the queue.

5 front() The first item in the queue is returned. O(1)

6 rear() The last item in the queue is returned. O(1)

Code for Queue operations using the List

# Empty List
country=[]
# Add an item on rear of the queue.
def enqueue(x):
country.append(x)
print(“Enqueue operation result:”, country)
# Remove an item from the front of the queue.
def dequeue():
if (len(country)>0):
x=country.pop(0)
 print(“removed element:”, x)
print(“Dequeue operation result:”, country)
# Check whether or not the queue is empty.
def empty():
if (len(country)==0):
print(“Queue is empty”)
else:
print(“Queue is NOT empty”)
# Find the number of items in the queue.
def size():
print(“size() operation result:”, len(country))
# Find the top item in the Stack
def front():
if (len(country)>0):
x=country[0]
print(“Front item in the queue is:”, x)
def rear():
if (len(country)>0):
y=country[-1]
print(“Rear item in the queue is:”, y)
while(1):
print (“Queue Operations”)
print (“****************”)
print (“1. ENQUEUE”)
print (“2. DEQUEUE”)
print (“3. IS EMPTY”)
print (“4. SIZE”)
print (“5. FRONT”)
print (“6. RARE”)
print (“7. EXIT”)
inp=int(input(“Enter your choice number”))
if(inp > 7 | inp < 1):
print(“Wrong choice so try again “)
else:
if(inp == 1):
name=input(“Enter the country name to
added in the queue:”)
 enqueue(name)
if(inp == 2):
pop()
if(inp == 3):
empty()
if(inp == 4):
size()
if(inp == 5):
front()
if(inp == 6):
rear()
if (inp == 7):
break
print(“Completed >>>>”)

RESEARCH DIRECTIONS AND CHALLENGES

These research challenges serve as a springboard for exploring advanced topics

and stretching the limits of data organization with Python lists. Researchers can

delve deeper into these areas to advance list-based data administration and

propose novel solutions to the challenges below.

Develop and investigate efficient collection management algorithms. Consider

scenarios in which the size of the List exceeds the quantity of available

memory, and study Lists optimization techniques such as disk-based storage,

cache, and parallel processing. Examine optimization techniques such as

appending, extending, and modifying list items. Analyze techniques such as

algorithmic enhancements, memory management strategies, and specialized

data structures created for specific use cases.

Explore and develop advanced list indexing and searching techniques. Consider

instances in which the List contains intricate data structures or nested lists.

Explore techniques such as multidimensional indexing, indexing based on

particular properties or attributes, and the implementation of effective search

algorithms for diverse data patterns. List Compression and Encoding

Techniques: Analyze techniques for compressing and encoding lists to reduce

storage requirements and improve data transfer efficiency. Explore list-specific

compression algorithms, encoding schemes, and data representations. Consider

the tradeoffs between compression ratios, encoding/decoding rates, and

memory consumption.

List Visualization and Interactive Exploration: An investigation into techniques

for effectively visualizing large lists. Create interactive frameworks or tools that

enable users to investigate and analyze lists efficiently. Consider using data

visualizations like heat maps, scatter graphs, and hierarchical representations to

identify patterns, relationships, and outliers. Security and Privacy of Lists:

Examine methods for ensuring the security and privacy of lists, particularly in

scenarios involving sensitive or confidential data. Examine techniques such as

data anonymization, secure storage, access control mechanisms, and encryption

algorithms to safeguard list data's confidentiality and integrity.

Examine approaches for integrating and interoperating lists with other data

structures and systems. Analyze techniques for data exchange between

databases and Lists, cloud storage, and distributed computing frameworks.

Consider obstacles such as data consistency, synchronization, and the

maintenance of efficient data flow between systems. Examine List data

structures for tasks involving machine learning and data extraction. Develop

techniques and algorithms for data analysis and extraction from collections.

Examine obstacles associated with list data, such as feature extraction,

dimensionality reduction, outlier detection, and pattern recognition.

Preprocessing and Cleaning of List-Based Data: Examine inventory

preprocessing and cleansing techniques to ensure data quality and

dependability. Formulate algorithms for absent values, outliers, duplicate

entries, and inconsistent data management in Lists. Explore list-specific

techniques for data imputation, outlier detection, and data validation. Examine

methods for collaborating and sharing listings in a decentralized or distributed

environment. In collaborative list editing scenarios, investigate techniques for

versioning, conflict resolution, access control, and data consistency. Consider

obstacles such as efficient synchronization, change merging, and data integrity

maintenance in distributed list environments.

CONCLUSION

In this chapter, the organization of data using the list was explored, a sequential

data structure in Python. The basics of lists and their ability to hold multiple
elements of different data types were learned. The access of elements in a list

using indices was discovered, allowing retrieval of specific items based on their

position. Python's support for negative indexing provided a convenient way to

access items from the end of the list. Furthermore, the mutability of lists was

discussed, enabling modification of their content even after creation. Various

methods for modifying lists were explored, including changing items at specific

indices or ranges of indices. Additionally, the append() and extend() methods

were examined for adding items to lists, understanding their differences and use

cases.

The chapter also covered searching and sorting operations on lists. The use of

Python's membership operator to find items in a list and the sorting of a list

using the built-in sort() method were learned. Furthermore, the time complexity

of different list operations, including index lookup, assignment, append, extend,

and sorting, was discussed. Next, the stack data structure was introduced as a

fundamental tool for effective data organization. The concept of a stack as a

Last-In-First-Out (LIFO) structure was understood, and essential operations

such as push and pop were learned. The implementation of a stack using a list

in Python was explored, and code examples for stack operations were provided.

Lastly, the queue data structure, which follows the First-In-First-Out (FIFO)

principle, was discussed. The enqueue and dequeue operations for adding and

removing items from a queue were examined. Similar to the stack, code

examples for queue operations implemented using a list were presented.

Throughout this chapter, various research directions and challenges for further

exploration were identified. These include the development of efficient

collection management algorithms, investigation of advanced list indexing and

searching techniques, exploration of optimization techniques for large lists, and

consideration of scenarios with complex data structures or nested lists.

In conclusion, this chapter has provided a comprehensive overview of

organizing data using lists as a sequential data structure in Python. By

understanding the fundamentals, operations, and time complexity of lists, as

well as exploring the stack and queue data structures implemented with lists,

readers are equipped with valuable knowledge for effective data organization

and manipulation. The presented research directions and challenges serve as an

invitation for researchers to delve deeper into advanced topics and propose

innovative solutions to enhance list-based data administration

REFERENCES

Dubois, P. F., Hinsen, K., & Hugunin, J. (1996). Numerical Python.

Computers in Physics, 10(3), 262–267. doi:10.1063/1.4822400

Erciyes, K. (2021). A Short Review of Python. In Algebraic Graph

Algorithms. Springer. doi:10.1007/978-3-030-87886-3

Hetland, M. L. (2017). Beginning Python: From novice to

professional. Apress. doi:10.1007/978-1-4842-0028-5

Hunt, J. (2019). ADTs, Queues and Stacks. In A Beginners Guide to

Python 3 Programming (pp. 407–414). Springer. doi:10.1007/978-3-

030-25943-3_34

Kobbaey, T., Xanthidis, D., & Bilquise, G. (2022). Data Structures and

Algorithms with Python. In Handbook of Computer Programming with

Python (pp. 207-272). Chapman and Hall/CRC.

doi:10.1201/9781003139010-6

Lutz, M. (2013). Learning Python: Powerful object-oriented

programming. O'Reilly Media, Inc.

Martelli, A. (2006). Python in a Nutshell. O'Reilly Media, Inc.

Navghare, N., Kedar, P., Sangamkar, P., & Mahajan, M. (2020). Python

and OpenCV in automation of live Surveillance. In Machine Learning

and Information Processing (pp. 235–243). Springer. doi:10.1007/978-

981-15-1884-3_22

Oliphant, T. E. (2007). Python for scientific computing. Computing in

Science & Engineering, 9(3), 10–20. doi:10.1109/MCSE.2007.58

Popescu, G. V., & Bîră, C. (2022, June). Python-Based Programming

Framework for a Heterogeneous MapReduce Architecture. In 2022

14th International Conference on Communications (COMM) (pp. 1-6).

IEEE. doi:10.1109/COMM54429.2022.9817183

Saabith, A. S., Fareez, M. M. M., & Vinothraj, T. (2019). Python

current trend applications-an overview. International Journal of

Advance Engineering and Research Development, 6(10).

Schwarz, J. S., Chapman, C., & Feit, E. M. (2020). An Overview of

Python. Python for Marketing Research and Analytics. Springer.

doi:10.1007/978-3-030-49720-0
 CHAPTER 3

Linear Data Structures and Their Applications

Kavita Srivastava

Institute of Information Technology and Management, GGSIP University,

India

ABSTRACT

In a linear data structure, the elements are arranged in a specific order. Each

element on a linear data structure is connected to the elements that proceed

it as well as the element that follows it. In other words, the elements are

placed in sequential order. Hence, it is possible to traverse the elements in

sequential order in a single pass. The objective of the chapter is to provide a

basic understanding of linear data structures and their implementation in

Python. The chapter is organized as follows. The chapter starts with an

introduction to the linear data structure. The objective is to provide a basic

understanding of linear data structures. A comparison of linear data

structures and non-linear data structures is given next. Next, all of the

above-mentioned data structures are explained in brief. The Python

implementation of all of these data structures is provided. After that, their

applications and their implementation in Python is discussed. The chapter

concludes by highlighting the essential concepts.

INTRODUCTION

In the realm of data organization and efficient data processing, data

structures play a vital role. They enable us to arrange and manage data

effectively, allowing for the seamless execution of related operations. Data

structures can be broadly categorized into Linear Data Structures and Non-

Linear Data Structures. The linear arrangement facilitates sequential

traversal in a single pass, distinguishing them from non-linear systems.

Common examples of linear data structures include arrays, linked lists,

stacks, and queues (Domkin, 2021). Let’s explore these data structures in

detail, along with their applications.

Arrays

An array is a collection of elements identified by an index or a key. Details

in an array are stored contiguously in memory, allowing random access.

Furthermore, all elements in an array must belong to the same data type.

Arrays find applications in various software systems that require storing

similar records, such as employee management systems, library

management systems, flight booking systems, and video rental applications

(McKinney, 2010). They are also extensively used in image processing,

where two-dimensional arrays represent image pixels.

Linked Lists

A linked list consists of interconnected nodes. Each node contains a data

value and a pointer to the next node. Linked lists can be either singly linked

or doubly linked. In a singly linked list, each node has a pointer to the next

node, while in a doubly linked list, nodes have pointers to both the next and

previous nodes, enabling bidirectional traversal. Circular linked lists, a

variant of singly linked lists, have the last node pointing back to the first

node. Linked lists are employed when storing records in non-contiguous

memory locations and when dynamic memory allocation is required

(Boockmann & Lüttgen, 2020). Applications of linked lists include web

page navigation through hyperlinks, organization of social media content

and feeds, symbol table management in IDEs and compilers, multiplayer

games, and the implementation of various operating system algorithms.

Stacks

A stack follows the Last-In-First-Out (LIFO) principle, where elements are

added and removed from the top of the Stack. The push() operation adds an

element to the top, while the pop() function removes the topmost element.

Stacks find numerous applications, such as evaluating postfix expressions,

maintaining history and logs, creating playlists, and implementing recursion

(Buckley & Buckley, 2017).

Queues

A queue operates on the First-In-First-Out (FIFO) principle, where elements

are added at the back and removed from the front. The primary application

of queues lies in CPU scheduling. Queue data structures are employed

whenever maintaining records in a FIFO order is crucial. They are also used

in communication networks for message sending, email systems, congestion

control, data packet forwarding, and more. Additionally, queues play a role

in request and response systems, printer queues, multiplayer games, and

other scenarios requiring the maintenance of a queue-like structure (Zhang,

2013).

Linear data structures offer various ways to manipulate and organize data,

including sorting, searching, and traversing elements. The choice of a data

structure depends on specific requirements, such as the need for fast random

access or efficient insertion and deletion operations. Implementing linear

data structures is straightforward, resembling the arrangement of memory

cells, and they efficiently utilize memory resources. Primarily employed in

software application development, these structures play a pivotal role. In the

following sections, we will delve into a comparison between linear and non-

linear data structures.

How Do Linear Data Structures Differ From Other Data

Structures?

Linear data structures, as compared to other data structures, exhibit distinct

characteristics and play a vital role in various applications. The

dissimilarities between linear data structures and other data structures are

noteworthy. Here are some key differences:

• Accessing Elements: Linear data structures enforce a sequential

order for elements, necessitating specific access patterns. For instance,

a stack follows the LIFO (Last-In-First-Out) order, while a queue

 adheres to the FIFO (First-In-First-Out) order. In contrast, data

structures like trees and graphs allow random access to elements.

• Storage and Manipulation: Linear data structures store elements in

a contiguous memory block, enabling straightforward traversal and

manipulation. On the other hand, data structures such as trees and

graphs involve intricate relationships between elements, demanding

more sophisticated algorithms for manipulation.

• Memory Usage: Linear data structures exhibit linear memory usage,

meaning that memory consumption increases proportionally with the

number of elements stored. Conversely, data structures like trees and

graphs can have non-linear memory usage patterns.

• Performance: Linear data structures offer rapid access and

manipulation times for specific operations, like popping an element

from a stack or inserting an element into a queue. However, certain

operations, such as searching for an element, might be slower than data

structures like trees and hash tables.

When selecting a data structure, the choice depends on the specific

requirements of the problem. Factors like the necessity for fast random

access or efficient insertion and deletion operations come into play. Linear

data structures are beneficial for simple problems with direct element

relationships. However, data structures such as trees and graphs may prove

more suitable for more complex situations. The following section will detail

the characteristics of Linear Data Structures.

CHARACTERISTICS OF LINEAR DATA STRUCTURES

Linear data structures possess distinctive features that make them versatile

tools in programming (Millman & Aivazis, 2011):

• Sequential Access: Elements within linear data structures are

accessed in a specific order, following a Last-In-First-Out (LIFO)

pattern like in a stack or a First-In-First-Out (FIFO) pattern like in a

queue.

• Linear Memory Usage: As the data size increases, linear data

structures consume memory proportionally, maintaining a linear

 relationship between data and memory usage.

• Straightforward Manipulation: Elements are stored in contiguous

memory locations, enabling effortless traversal and manipulation.

• Swift Operations for Specific Tasks: Linear data structures offer

rapid access and manipulation times for specialized operations such as

element popping from a stack or element insertion into a queue.

• Restricted Element Access: Due to the enforced order, accessing

elements randomly becomes more challenging than other data

structures like trees and hash tables.

• Efficient Insertion and Deletion: Linear data structures provide

efficient insertion and deletion times for specific operations like

popping an element from a stack or inserting a part into a queue.

These characteristics bestow linear data structures with the ability to solve

straightforward problems efficiently. However, alternative data structures

such as trees and graphs might prove more suitable for more complex

issues. In the subsequent section, we delve into significant applications

where linear data structures find practical use.

APPLICATIONS OF LINEAR DATA STRUCTURES

Linear data structures offer many practical applications, showcasing their

versatility and usefulness in various domains. Here are some special

applications (Gibbons & Matias, 1999):

• Storing and Retrieving Data: Linear data structures efficiently store

and retrieve information. Numerous storage systems leverage these

structures to ensure swift access to data whenever needed.

• Dynamic Memory Allocation: Linked lists provide an effective

means of implementing dynamic memory allocation. This feature is

precious when allocating memory for objects on the heap.

• Stack-Based Algorithms: Stack data structures prove invaluable in

implementing stack-based algorithms. These structures can efficiently

accomplish tasks such as expression evaluation and maze traversal.

• Queue-Based Algorithms: Queue-based algorithms are highly

efficient for tasks like breadth-first search and simulating real-life

 queues. Queues are a fundamental component for implementing such

algorithms.

• History Management: Linear data structures, like stacks and

queues, find practical use in applications such as web browsers and text

editors. They enable tracking recent actions, facilitating the essential

undo and redo functionalities.

• Computer Networks: Queues, as linear data structures, are crucial

in managing communication within computer networks. They ensure

the orderly transmission of data between devices.

• Data Compression: Linear data structures, such as stacks, prove

valuable in data compression algorithms. These structures enable

efficient data encoding and decoding, contributing to effective

compression techniques.

These examples provide a glimpse into the vast realm of applications for

linear data structures. When selecting a data structure, specific problem

requirements, such as the need for rapid random access or efficient insertion

and deletion operations, guide the decision-making process. The following

section will delve into the advantages of employing linear data structures.

ADVANTAGES OF LINEAR DATA STRUCTURES

In the realm of Python programming, linear data structures serve many

purposes and offer numerous benefits. Here, we delve into the advantages

that make them invaluable tools (Goodrich et al., 2013):

• Simplicity at Its Core: Linear data structures possess an innate

simplicity that renders them easily understandable and implementable.

Consequently, they are optimal for introductory computer science

courses, providing a smooth learning curve.

• Swift and Targeted Operations: With linear data structures,

specific operations are executed remarkably quickly and efficiently.

Whether it involves extracting an element from a stack or inserting an

element into a queue, these structures excel at rapid access and

manipulation.
 • Efficiency in Insertion and Deletion: Seamlessly accommodating

insertion and deletion operations, linear data structures facilitate

optimal efficiency. These structures offer streamlined performance,

whether it entails popping an element from a stack or inserting an

element into a queue.

• Seamless Sequential Access: Applications that require sequential

access to elements find an ideal ally in linear data structures. Whether

storing and retrieving data, managing historical records, or navigating

computer networks, these structures exhibit exceptional suitability.

• Minimal Resource Overhead: Linear data structures boast a

remarkable advantage in terms of resource utilization. The overhead

they impose remains minimal, sparing unnecessary memory

consumption and computational burden associated with maintaining

relationships between elements.

• Dynamic Memory Allocation Possibilities: The prowess of

dynamic memory allocation finds a proper implementation through

linked lists, a type of linear data structure. This allows efficient

memory allocation for objects on the heap, further expanding the

capabilities of these structures.

The advantages above position linear data structures as a powerful solution

for tackling simpler problems with direct element relationships. However,

alternative data structures such as trees and graphs may be more adept for

complex challenges. It is essential to recognize the drawbacks associated

with linear data structures, which we will explore in the subsequent

discussion.

DRAWBACKS OF LINEAR DATA STRUCTURES

Linear data structures, while immensely useful, do come with certain

limitations. These limitations include the following (Othman, 2023):

• Limited Access to Elements: One drawback of linear data structures

is that elements can only be accessed in a specific order. This makes it

more challenging to access elements randomly compared to other data

structures like trees and hash tables.

 • Slow Random Access: Accessing elements randomly in linear data

structures can be slow. Each element must be traversed sequentially to

reach a specific location, hindering efficiency.

• Fixed-Size: Linear data structures like arrays have a predetermined

size. They cannot dynamically grow or shrink to accommodate

changes, which may pose constraints in flexible storage scenarios.

• Memory Usage: Linear data structures exhibit linear memory usage.

This means that memory consumption increases linearly with data

size. While this can be advantageous for smaller datasets, it becomes a

drawback when dealing with applications that demand extensive data

storage.

• Complexity: Managing large-sized data and complex relationships

between elements can introduce complexity in implementing and

maintaining linear data structures. As the data grows in size and

complexity, handling and organizing it becomes more challenging.

• Inefficiency for Complex Operations: Linear data structures can

prove inefficient for operations involving intricate relationships

between elements. Tasks like finding the shortest path between two

nodes in a graph may require more optimized data structures.

Considering these drawbacks when selecting a data structure for a

specific problem is crucial. Complex problems may be better suited to

alternative data structures that provide more efficient and effective

solutions.

ARRAYS

In linear data structures, arrays are a powerful tool, enabling us to refer to

multiple elements by a single name and significantly reducing the amount

of code required. Each piece within an array possesses a unique index that

facilitates easy access. It’s important to note that Python arrays should not

be confused with lists or numpy arrays, as they are distinct entities. While

Python lists also store values akin to arrays, they possess the flexibility of

accommodating elements of different types, unlike arrays which mandate

uniformity.
The following example demonstrates using Python lists and arrays for

storing values.

mylist=[1,2,3,'A',”some text”, True, [11,12,13],

'X', 3.5]
print(mylist)

Output

Figure 1. Python list

In the context of linear data structures and their practical applications, it

becomes evident that a list in Python can store a diverse range of values,

including integers, characters, strings, Booleans, floating-point numbers,

and even other lists. On the other hand, when working with arrays, all

elements need to share the same data type. Fortunately, Python’s array

module facilitates the creation of arrays of values belonging to basic types.

This functionality allows for the creation of arrays with integers ranging

from one to eight bytes in size, floating-point values with sizes of four and

eight bytes, and arrays containing Unicode characters with a size of two

bytes. When using the array() function, the first parameter represents a code

that specifies the desired type. To illustrate the practical usage of arrays in

Python, consider the following program demonstration.

import array as my_array

a1=my_array.array(‘i’, [10, 20, 30])
print('An Integer Array: ')
for i in range (0, 3):
print (a1[i], end =” “)
print()
a2=my_array.array('d', [2.99, 1.34377, 12.70,
0.125])
print('An Floating Point Array: ')
for i in range (0, 4):
print (a2[i], end =” “)
print()

Output

Figure 2. Python arrays

When to mix the values belonging to different data types, Python doesn’t

allow it, as shown in the following code.

import array as my_array

a1=my_array.array(‘i’, [10, 20, 3.5, ‘ABC’, 30])
print('An Integer Array: ')
for i in range (0, 3):
print (a1[i], end =” “)
print()

Output

Figure 3. Creating Python arrays with different types

SINGLY-LINKED LIST

Linear data structures, an essential concept in programming, resemble

arrays as they consist of a collection of elements. However, unlike arrays,

linked lists dynamically generate elements. Consequently, linked lists

allocate only the necessary elements at a given moment, preventing

memory wastage encountered with arrays. The advantage lies in the fact

that linked lists eliminate the need for pre-allocated memory and allow non-

contiguous locations for data storage. In a linked list, nodes store elements

and are interconnected, with each node linked to the next one. Specifically,

a singly linked list comprises nodes with a data field and a pointer called

“next,” which points to the subsequent node. The last node in the List points

to a null value, indicating the end. The initial node, the head node (refer to

Figure 4), marks the beginning of the linked List (Baka, 2017).

Figure 4. Nodes in linked list

The process of traversing a linear data structure is simple. It begins by

starting at the head node and then scanning each subsequent node through

the next pointer until a null value is encountered. When creating a new list,

two classes are required: “node” and “linked_list.” The “node” class

represents an individual node in the List and contains two essential

members: “info” and “next.” “info” stores the data value of the node, while

“next” holds the reference to the next node in the List. In Python, the

keyword “None” is used to indicate a null value, so the constructor assigns

“None” to the “next” member. The data for a node is passed as a parameter

to the constructor and then transferred to the “info” field. In the

“linked_list” class, the single data member required is “head,” which is

initialized as “None” in the constructor. The code snippet below

demonstrates the usage of the “create_list()” function for list creation and
the “traverse_list()” function for list traversal. The Python code for creating

a singly linked list is given below.

class node:
def __init__(self, info):
self.info=info
self.next=None
class linked_list:
def __init__(self):
self.head=None
def create_list(self, info):
if self.head is None:
self.head=node(info)
self.head.next=None
else:
p=node(info)
q=self.head
while q.next is not None:
q=q.next
q.next=p

p.next=None

def traverse_list(self):
p=self.head
while p is not None:
print(p.info, end=' ')
p=p.next
mylist=linked_list()
n=int(input('How many elements you want to enter
in the list? '))
for i in range(n):
info=int(input('Enter value: '))
mylist.create_list(info)
print('The Elements in the Linked List are....',
end=' ')
mylist.traverse_list()

Output

Figure 5. Elements in the linked list

create_list() Method

This method demonstrates its proficiency when provided with a parameter

named “info,” housing the cherished value of a data element. Its initial

undertaking involves scrutinizing whether the List is devoid of any content,

accomplished by assessing the status of the head node. Should an absence

be detected, a node is promptly forged, launched as the head node, and

bestowed with the virtue of having its next pointer set to None. Conversely,

if the List possesses prior occupants, a deliberate traversal commences,

tracing the path from the head node until the very last node. Herein lies the

pivotal moment wherein the new value is warmly embraced by appending it

to the List, as the previous node's next pointer seamlessly interlocks with

the newly minted node. Alas, to maintain the structure, the following field

of the new node graciously assumes the value of None, signifying the end

of this union.

traverse_list() Method
the traverse_list() method displays all the nodes in the List, starting from the

head and following the subsequent nodes' next field until the List ends.

Insertion

There are three situations for insertion in the List.

• Insertion at the first position

• Insertion at the last position

• Insertion at any Valid Position in the List

Deleting a Node

Like insertion, there are three situations for deletion.

• Deletion at the Beginning

• Deletion at the End

• Deletion at any Valid Position

DOUBLY LINKED LIST

A doubly linked list is an essential linear data structure incorporating an

information field and two pointers - next and previous. This fundamental

difference from a singly linked list enables seamless traversal in both the

forward and reverse directions. By maintaining two-pointers, a doubly

linked list provides increased flexibility. Using the diagram below, let’s

visualize a doubly linked list (Bandopadhyay & Shang, 2022).

Figure 6. Doubly linked list

To implement linear data structures, we create a class called “node” to

represent a node within the List. The “node” class consists of three
members: “info”, “next”, and “previous”. The “info” field stores the data

value of the node, while the “next” and “previous” pointers are used to

navigate between nodes. Additionally, the “doubly_linked_list” class

contains a ” head ” field,” which refers to the first node in the doubly-linked

List. This class also defines two methods: “create_list()” and

“traverse_list().” These methods are responsible for creating a new list and

traversing through the existing List.

Furthermore, the “doubly_linked_list” class will later include node

insertion and deletion methods. When adding a node to the List, we must

ensure that the “next” and “previous” pointers are set correctly. If the List is

empty, the new node becomes the head, so we assign the “head” field to the

new node, and both the “previous” and “next” fields are set to None.

However, if the List is not empty, adding a node involves traversing to the

end of the List and making the last node point to the new node. This is

achieved by assigning the “next” pointer of the last node to the new node,

while the “previous” field of the new node should point to the node at the

end of the existing List. The Python code for the Doubly Linked List is

given below.

class node:
def __init__(self, info):
self.info=info
self.next=None
self.previous=None
class doubly_linked_list:
def __init__(self):
self.head=None
def create_list(self, info):
p=node(info)
if self.head is None:

self.head=p
self.head.next=None
self.head.previous=None
else:
 q=self.head
while q.next is not None:
q=q.next
p.previous=q
p.next=None
q.next=p
def traverse_list(self):
p=self.head
while p is not None:
print(p.info, end=' ')
p=p.next
mylist=doubly_linked_list()
n=int(input('How many elements you want to enter
in the list? '))
for i in range(n):
info=int(input('Enter value: '))
mylist.create_list(info)
print('\nThe Elements in the Linked List are
(Traversing Forward)....', end=' ')
mylist.traverse_list()

Output

Figure 7. Elements in the doubly linked list

CIRCULAR LINKED LIST

This List also maintains a single pointer and each node's data field.

However, the last node points to the head node (Guzdial & Ericson, 2012).

The Python code for the Circular linked list is given below.

class node:
def __init__(self, info):
self.info=info
self.next=None
class circular_linked_list:
def __init__(self):
self.head=None
self.last=None
def append(self, info):
p=node(info)
if self.head is None:
self.head=p
self.last=p
self.last.next=self.head
else:
self.last.next=p
self.last=p
self.last.next=self.head
def display_circular(self):
p=self.head
print(p.info, end=' ')
while p.next is not self.head:
p=p.next
print(p.info, end=' ')
print('\nThe Next Element in the
Circular Linked List...', end=' ')
print(p.next.info, end=' ')
circularlist=circular_linked_list()
n=int(input('How many elements you want to enter
in the list? '))
for i in range(n):
info=int(input('Enter value: '))
 circularlist.append(info)
print('\nThe Elements in the Circular Linked List
are....', end=' ')
circularlist.display_circular()

STACK

The Stack stands out as it adheres to a Last In First Out (LIFO) order. This

implies that elements are added and removed from the last position. Adding

an element is known as the push() operation, while the removal is called the

pop() operation. Moreover, the Stack includes a peek() operation that

retrieves the topmost element from the Stack without deleting it. Presented

below is the Python code for executing stack operations.

class node:
def __init__(self, info):
self.info=info
self.next=None
class Stack:
def __init__(self):
self.head=None
self.top=None
def push(self, info):
p=node(info)
if self.isEmpty():
self.head=p
self.top=p
self.head.next=None
self.top.next=None
else:
self.top.next=p
p.next=None
self.top=p
def pop(self):
if self.isEmpty():
print(“\nStack is
Empty!!!\nStack Underflow!!!\n”);
return -9999
else:
n=self.top.info
if self.top==self.head:
self.top=None
self.head=None
else:
p=self.head
while p.next is
not self.top:
p=p.next
p.next=None
self.top=p
return n
def peek(self):
if self.isEmpty():
print(“\nStack is
Empty!!!\nStack Underflow!!!\n”);
return -9999
else:

n=self.top.info
return n
def stack_traversal(self):
if not self.isEmpty():
p=self.head
while p is not None:
print(p.info,
end=' ')

p=p.next
else:
print(“\nStack is
Empty!!!\n”);
def isEmpty(self):
 if self.top is not None:
return False
else:
return
True
mystack=Stack()
n=int(input('How many elements you want to enter
in the stack? '))
for i in range(n):
info=int(input('Enter value: '))
mystack.push(info)
print('\nStack Traversal\n')
mystack.stack_traversal()
print('\nElement at the top of stack: ')
print(mystack.peek())
print('\nAdding an element: ')
info=int(input('Enter value: '))
mystack.push(info)
print('\nElement at the top of stack: ')
print(mystack.peek())
print('\nRetrieving the elements of Stack...',
end=' ')
while not mystack.isEmpty():
x=mystack.pop()
print(x, end=' ')
print('\nStack Traversal\n')
mystack.stack_traversal()

Output

Figure 8. Stack operations

QUEUE

A queue data structure follows the FIFO (First-In-First-Out) principle. This

means elements are added to the queue from the last position and removed

from the front. The two ends of a queue are the front and rear. Elements are

removed from the front end and inserted at the rear end of the queue. The

queue supports two essential operations: enqueue() and dequeue(). The

enqueue() process adds an element to the queue, while the dequeue()

operation removes an element from the queue. Additionally, the queue

provides other functions such as isEmpty() and isFull().

In implementing a queue, the isEmpty() operation determines whether the

queue contains any elements. It returns a valid Boolean value when the

queue is empty. Similarly, the isFull() operation indicates whether the queue
has reached its maximum capacity and produces an excellent Boolean value

in such cases. However, if the queue is implemented using a linked list, the

isFull() operation is unnecessary. Below is an example of implementing a

Queue using a linked list in Python programming.

class node:
def __init__(self, info):
self.info=info
self.next=None
class Queue:
def __init__(self):
self.front=None
self.rear=None
def enqueue(self, info):
my_node=node(info)
if self.front is None:
self.front=my_node
self.rear=my_node
self.front.next=None
self.rear.next=None
else:
if self.front==self.rear:

self.rear.next=my_node

self.rear=self.rear.next

self.rear.next=None

self.front.next=self.rear
else:

this_node=self.rear

self.rear.next=my_node
self.rear=my_node

self.rear.next=None
def dequeue(self):
if self.front is None and
self.rear is None:
print('\nQueue is
Empty!!!\n');
else:
if self.front==self.rear:
self.front=None
self.rear=None
else:

this_node=self.front

self.front=this_node.next

this_node=None
def isEmpty(self):
if self.front is None and
self.rear is None:
return True
else:
return False
def traverse_queue(self):
if self.front is None and
self.rear is None:
print('\nQueue is
Empty!!!\n');
else:
this_node=self.front
while this_node is not
None:

print(this_node.info, end=' ')

this_node=this_node.next
new_queue=Queue()
n=int(input('What is the size of the Queue?'))
for i in range(n):
info=int(input('Enter value: '))
new_queue.enqueue(info)
print('\nQueue Elements....\n', end=' ')
new_queue.traverse_queue()
print('\nRemoving element from queue....\n', end='
')
new_queue.dequeue();
print('\nQueue Elements....\n', end=' ')
new_queue.traverse_queue()
print('\nRemoving element from queue....\n', end='
')
new_queue.dequeue();
print('\nQueue Elements....\n', end=' ')
new_queue.traverse_queue()
print('\nAdding an element in queue....\n', end='
')
new_queue.enqueue(int(input('Enter value: ')));
print('\nQueue Elements....\n', end=' ')
new_queue.traverse_queue()

Output

Figure 9. Operations on queue

PRIORITY QUEUE

Each node is assigned a priority in a priority queue, and the nodes are

organized based on their priority. The node with the highest priority is

placed at the front, while the node with the lowest priority is at the rear.

Consequently, the node with the highest priority is permanently removed

when performing the Dequeue operation. Priority queues find practical use

in various real-life scenarios (Üçoluk et al., 2012). For instance, in CPU job

scheduling, jobs are arranged according to their priority, and higher-priority

jobs are executed before lower-priority jobs. It is important to note that the

node class includes an additional attribute called “priority.” During the

Enqueue operation, the program determines the element with a higher

priority than the new node and inserts the new node before it. In this
context, priority is represented as an integer ranging from 1 to 10, where 1

signifies the highest priority, and 10 describes the lowest.

CONCLUSION

In conclusion, this research work has comprehensively explored linear data

structures and their applications, with a specific focus on arrays, linked lists,

stacks, and queues. These data structures offer efficient ways to organize

and manipulate data, making them invaluable tools in software

development. The advantages of using linear data structures in Python

programming have been highlighted, including their simplicity, swift and

targeted operations, efficient insertion and deletion, seamless sequential

access, minimal resource overhead, and dynamic memory allocation

possibilities. It is important to note that this research has specifically

emphasized implementing linear data structures using the Python

programming language. Python offers a user-friendly and versatile

environment for working with these data structures, enabling programmers

to leverage Python's built-in features and libraries for efficient data

processing.

The drawbacks of linear data structures have also been discussed, such as

limited access to elements, slow random access, fixed size, memory usage,

complexity, and inefficiency for complex operations. Considering these

limitations when selecting a data structure for a specific problem is

essential, especially when dealing with complex scenarios that may require

alternative data structures like trees and graphs. Overall, the research has

shed light on linear data structures' characteristics, applications, advantages,

and drawbacks, emphasizing their significance in solving more

straightforward problems and their limitations in more complex scenarios.

By implementing these data structures using Python programming,

developers can harness the power of these structures to efficiently organize

and process data in various domains, ranging from storage systems and

dynamic memory allocation to stack-based and queue-based algorithms,

history management, computer networks, and data compression. The

knowledge gained from this research work can serve as a solid foundation

for further exploration and application of linear data structures in Python

programming, enabling developers to make informed decisions when

choosing the most appropriate data structure for a given problem.

REFERENCES

Baka, B. (2017). Python Data Structures and Algorithms. Packt

Publishing Ltd.

Bandopadhyay, S., & Shang, H. (2022). SADHANA: A Doubly

Linked List-based Multidimensional Adaptive Mesh Refinement

Framework for Solving Hyperbolic Conservation Laws with

Application to Astrophysical Hydrodynamics and

Magnetohydrodynamics. The Astrophysical Journal. Supplement

Series, 263(2), 32. doi:10.3847/1538-4365/ac9279

Boockmann, J. H., & Lüttgen, G. (2020, May). Learning Data

Structure Shapes from Memory Graphs. LPAR.

Buckley, I. A., & Buckley, W. S. (2017). Teaching software testing

using data structures. International Journal of Advanced Computer

Science and Applications, 8(4).

Domkin, V. (2021). Linked Lists. Programming Algorithms in

Lisp: Writing Efficient Programs with Examples in ANSI Common

Lisp, 75-99.

Gibbons, P. B., & Matias, Y. (1999, January). Synopsis Data

Structures for Massive Data Sets. In SODA (Vol. 10, pp. 909-910).

doi:10.1090/dimacs/050/02

Goodrich, M. T., Tamassia, R., & Goldwasser, M. H. (2013). Data

structures and algorithms in Python. John Wiley & Sons Ltd.

GuzdialM.EricsonB. (2012, February). Listening to linked lists:

Using multimedia to learn data structures. In Proceedings of the

43rd ACM technical symposium on Computer Science Education

(pp. 663-663). doi:10.1145/2157136.2157358

McKinneyW. (2010, June). Data structures for statistical computing

in Python. In Proceedings of the 9th Python in Science Conference

(Vol. 445, No. 1, pp. 51-56). doi:10.25080/Majora-92bf1922-00a

Millman, K. J., & Aivazis, M. (2011). Python for scientists and

engineers. Computing in Science & Engineering, 13(2), 9–12.

doi:10.1109/MCSE.2011.36

Othman, J. (2023). Data structure uniqueness in Python

programming language. Enhancing Innovations in E-Learning for

Future Preparation, 68.

Üçoluk, G., Kalkan, S., Üçoluk, G., & Kalkan, S. (2012).

Organizing Data. Introduction to Programming Concepts with Case

Studies in Python, 165-194.

Zhang, H. L., Han, B. T., Xu, J. C., & Feng, Y. (2013). Preparation

of the public bus schedule based on fcfs. In Proceedings of 20th

International Conference on Industrial Engineering and

Engineering Management: Theory and Apply of Industrial

Engineering (pp. 1021-1029). Springer Berlin Heidelberg.

doi:10.1007/978-3-642-40063-6_100
 CHAPTER 4

A Comprehensive Analysis of Stack and

Queue Data Structures and Their Uses
S. Rajasekaran

https://orcid.org/0000-0002-7893-9072

University of Technology and Applied Sciences, Ibri, Oman

Mastan Vali Shaik

https://orcid.org/0000-0002-0068-5519

University of Technology and Applied Sciences, Ibri, Oman

ABSTRACT

The chapter provides a comprehensive and in-depth exploration of two

fundamental linear data structures in computer science: Stack and Queue. It

begins with an introduction to these data structures, highlighting their key

features and fundamental differences. It then explains various operations

that can be performed on both data structures, such as push, pop, enqueue,

and dequeue. These implementations are thoroughly compared and

contrasted, highlighting their advantages and disadvantages. These

applications are discussed in detail, showcasing the versatility of Stack and

Queue data structures in different scenarios. The chapter includes Python

code snippets demonstrating Stack and Queue data structures further to aid

the reader's understanding of the material. This chapter provides a

comprehensive and practical guide to Stack and Queue data structures,

making it a valuable resource for computer science students, researchers,

and practitioners.

INTRODUCTION TO STACK
A stack is a simple linear abstract data structure with finite (predefined)

capacity that follows the Last In First Out (LIFO) principle. That means the

element inserted last into the Stack will be removed first. Conversely, it can

say that First In Last Out (FILO). That means the element inserted first into

the Stack will be removed last. Both insert and delete operations are

performed only on the Top of the Stack. TOP is a marker pointing to the

current position in the Stack (Goodrich et al., 2013; Vasudevan et al., 2021).

The initial value of Top is -1. Figure 1 provides examples of stack data

structure.

Figure 1. Examples of stack data structure

FEATURES OF STACK

The following features characterize the Stack.

1. A Stack consists of an ordered collection of homogeneous (same

type) data elements.

2. The Stack accesses its data in LIFO or FILO order.

3. The push() method inserts new items into the Stack, while the drop()

function removes components from the Stack. Only at one end of the

Stack, termed the Top, are insert and delete operations done.
 4. When a stack is filled, and no additional components can be added,

it is in the Overflow state. Underflow describes a situation with no

items, and the collection is empty.

STACK OPERATIONS

Table 1 provides details on different stack operations and their descriptions.

(Dhruv et al., 2021)

Table 1. Stack operations

Basic
Description
Operations

push() Adds a new element at the Top of the Stack.

Deletes an existing element from the Top of the

pop()
Stack.

Additional
Description
Operations

It just returns the top element of the Stack without

top()
removing it.

is_empty() It returns true if the Stack is empty; otherwise, false.

is_full() It returns true if the Stack is full; otherwise, false.

It searches whether the particular element available

search(data)
in the Stack or not.

display() It displays the stack contents.

It returns the total number of elements available in

len()
the Stack.

PUSH Operation-Algorithm
 1. First, check whether the Stack is full or has space to insert more

elements.

2. Print the message “Stack Full” and exit if the Stack is full.

3. Increase the top value and add the new element if the Stack is

incomplete.

POP Operation-Algorithm

1. The pop operation will check whether the Stack has data or is empty.

2. Displays “Stack underflow” and exits the application if the Stack is

empty or contains no data.

3. Returns the element at the Top of the Stack and decreases the Top of

the Stack containing data or is not empty.

STACK IMPLEMENTATION

Two different methods are used to implement the stack operations. A Stack

implementation using arrays is called a static implementation (Harris et al.,

2020). The array implementation of the Stack is illustrated in Figure 2.

Stack implementation using a linked list is called a dynamic

implementation. Figure 3 outlines the Linked List implementation of Stack.

Figure 2. Array implementation of Stack

Figure 3. Linked List implementation of Stack

Exercise: Use the below Stack to find the answer to the following questions.

Table 2. Sample Stack for exercise

Element H @ M # C T % P D $

Index 0 1 2 3 4 5 6 7 8 9

• The number of elements present in the Stack?

• What will Stack's content be after executing the algorithm below on

the above Stack?

The algorithm for the Stack exercise is presented in Table 3.

Table 3. Stack exercise algorithm

1. Start

2. Push(“^”)

3. y=4

4. While y!=0 then

 5. Pop()

6. y=y-1

7. Push(“&”)

8. End

APPLICATIONS OF STACK

Stacks are used for several purposes (Lambert, 2019).

1. Reverse a String: A given string can be reversed using Stack.

2. Balancing Symbols in an Expression: The compiler checks your

program for syntax errors, but often a missing symbol (such as a

missing parenthesis or a missing comment start character) causes the

compiler to generate hundreds of diagnostic lines without identifying

the actual error. Output is possible. To ensure that each right

parenthesis, braces, and brackets must match their left counterparts,

e.g., [()] is legal, but [(]) is illegal.

3. Function Calls: Most programming tool runtime systems use stacks

to follow the chain of active functions. The runtime system pushes a

frame onto the Stack with the following information whenever the

process is called:

▪ Local variables and return values.

▪ A program counter that monitors instructions performed.

▪ After the function has finished executing, its frame will be taken

off the Stack, and the method at the Top of the Stack will assume

control.

4. Web Browsers: Stack can store the addresses of recently visited

sites in Web browsers. Every time a user visits a new website, the

website address is “pushed” onto the address stack. Browsers allow

users to return to previously visited pages using the back button.

5. Text Editors: Text editors often have a “undo” function to undo

recent changes and restore the document to its former state. This

“undo” action may be accomplished by storing text modifications in

batch.
 6. Recursion: Recursion is another use case of stack solutions often

preferred and implemented using stacks.

7. Backtracking (Game Playing, Discovering Pathways, Exhaustive

Searching): The Backtracking approach is employed in algorithms

when there are steps along a path (state) from a starting point to a goal.

▪ Go through the maze

▪ Find the path from one point to another on a diagram (roadmap)

▪ Play games with movement (chess, checkers)

In all of these scenarios, multiple options necessitate a system to recall

decision points if a choice must be attempted. For example, in navigating

through a maze, there comes a juncture where a decision needs to be made,

and it is possible to discover that the chosen path leads to a dead-end. The

objective is to return to the decision point and pursue the next alternative.

Stacks serve as a viable component in the solution.

8. Evaluation of Postfix Expressions: Converting expressions from

infix to postfix notation is another issue with expression manipulation.

The operations of an expression are put between the operands in infix

notation.

The algorithm to evaluate postfix expression is depicted in Figure 4.

Figure 4. Evaluate postfix expression: Algorithm

Evaluate a postfix expression 3 6 4 * + 5 - using Stack (Table 4).

Table 4. Evaluate postfix expression: Example

Stack
Symbol/Token Operation Calculation
Content

3 PUSH(3) 3

6 PUSH(6) 3 6

4 PUSH(4) 3 6 4

A = POP() B =
A = 4
POP()
* 3 24 B = 6
C = B * A
C = 6 * 4 = 24
PUSH(C)
 + A = POP() B = 27 A = 24

POP() B = 3

C = B + A C = 3 + 24 =

PUSH(C) 27

5 PUSH(5) 27 5

A = POP() B =
A = 5
POP()
- 22 B = 27
C = B - A
C = 27 - 5 = 22
PUSH(C)

) POP() Result = 22

9. Expression Transformations (Postfix to Prefix, Infix to Postfix,

etc.): The Stack transforms infix expressions into postfix expressions

and postfix expressions into prefix expressions.

The algorithm to convert the infix to a postfix expression is given in Figure

5.

Figure 5. Convert infix expression to postfix expression: Algorithm

Using Stack, convert the following infix expression into its corresponding

postfix expression.

P / (R + S) – N * M

Table 5 outlines the infix to postfix expression.

Table 5. Convert infix expression to postfix expression: Example

Symbol/Token Stack Content Output

(
P ( P

/ (/ P

( (/ ( P

R (/ ( P R

+ (/ (+ P R

S (/ (+ P R S

) (/ P R S +

- (- P R S + /

N (- P R S + / N

* (- * P R S + / N

M (- * P R S + / N M

) P R S + / N M * -

Python Code for Stack Operations

The Stack data structure helps solve various computer science problems,

and having a good understanding of its operations is crucial for any

software developer. The following is the Python code for stack operations in

Figures 6 and 7.

Figure 6. Stack data structure Python implementation Part 1

Figure 7. Stack data structure Python implementation Part 2
QUEUE: INTRODUCTION
A queue is a simple linear abstract data structure with a bounded

(predefined) capacity that follows the First In First Out (FIFO) principle.

This means the first element inserted into the Queue will be removed first.

In reverse, it can be written as Last In Last Out (LILO). That means the last

element inserted into the Queue will be removed last. Front and Rear are the

two markers used to perform the queue operations. All the insertions will be

done at the Rear, and the deletions at the Front. The initial value of both

Front and Rear are -1, and both the values are reset to 0 when the first data

item is inserted into the Queue. Only the rear value will be increased for the

subsequent insertion, and only the rear value will be raised for the deletions

(Thomas, 2021). The examples of Queue data structure are given in Figure

8.

Figure 8. Examples of Queue data structure

FEATURES OF QUEUE

The Queue has the following features.

1. A Queue consists of an ordered collection of homogeneous (same

type) data elements.

2. The Queue accesses its data in FIFO or LILO order.

 3. To insert new elements into the Queue enqueue() function and to

delete an element from the queue dequeue() function is used. All the

insertions will be done at the Rear, and the deletions will be done at

the Front.

4. The Front refers to the first element of the Queue, and the Rear

refers to the last element of the Queue.

5. The overflow condition occurs when the Queue is complete, and the

underflow condition is when it is empty.

QUEUE OPERATIONS

Table 6 provides the various queue operations.

Table 6. Queue operations

Basic Operations Description

Enqueue() It will add an element at the Rear of the Queue.

It will delete an element from the Front of the

Dequeue()
Queue.

Additional
Description
Operations

It will return the front element of the Queue without

Front()
removing it.

It will return the rear element of the Queue without

Rear()
removing it.

It returns true if the Queue is empty; otherwise

Is_empty()
false.

Is_full() It returns true if the Queue is full; otherwise false.

Search(data It will search whether the particular data item exists

item) in the Queue.

Display() Displays the queue contents.

It returns the total number of elements present in the

len()
Queue.

Algorithm for Creating and Initializing Queue

The algorithm for creating and initializing a Queue is presented below.

1. Create an empty Queue.

2. Initialize two Front and Rear pointers with 0 [Front=rear=0].

Enqueue Operation

The following is the algorithm for Enqueue operation

1. First, check whether the Queue is complete or if there is a place to

insert more items.

2. Print the message “Queue is full” and exit if the Queue is complete.

3. If the Queue is incomplete, set the Front and rear values to 0 when

inserting the first item into the Queue. Only increase the rear value

when enqueuing other items.

Dequeue Operation: Algorithm

The Queue operations are given in Figure 9. The following is the algorithm

for Dequeue.

1. First, determine if the Queue is vacant or if items must be deleted.

2. If the Queue is empty, print the message “Queue is vacant” and

terminate the program.

3. If the Queue is not empty, return the front element and increment its

front value by 1.

Figure 9. Queue operations

Example: Queue operations and their effects are explained in Table 7.

Consider an empty at the beginning and do the operations (Karumanchi,

2015).

Table 7. Queue operations: Example

Operations Output Value Queue Elements

Dequeue() Error [ ]

Enqueue(8) - [8]

Enqueue(3) - [8, 3]

Enqueue(5) - [8, 3, 5]

Enqueue(2) - [8, 3, 5, 2]

Empty() False

Len() 4 [8, 3, 5, 2]

Dequeue() 8 [3, 5, 2]

Dequeue() 3 [5, 2]

Enqueue(7) - [5, 2, 7]

Dequeue() 5 [5, 2, 7]

Len() 2 [2, 7]

Empty() False

Front() 2 [2, 7]

Rear() 7 [2, 7]

Dequeue() 2 [7]

Dequeue() 7 [ ]

Dequeue() Error [ ]

Empty() True [ ]

Len() 0 [ ]

QUEUE IMPLEMENTATION

Queues can be implemented in two different ways (Cutler et al., 2020). A

queue implementation using arrays is called a static implementation. The

array implementation of Queue is represented in Figure 10.

Figure 10. Array implementation of Queue

Queue implementation using a linked list is called a dynamic

implementation. The linked list implementation of Queue is shown in

Figure 11.

Figure 11. Linked list implementation of Queue

TYPES OF QUEUES

Queues are classified into four different types, which are mentioned below.

• Linear Queue/Simple Queue

• Circular Queue

• Double Ended Queue

• Priority Queue

Simple/Linear Queue

The Simple/Linear Queue possesses the following features.

 • Linear queues are the most basic types of queue data structure.

• Insert (Enqueue) operation is performed at REAR.

• Delete (Dequeue) operation is performed at FRONT.

• FRONT and REAR will point to -1 when Queue is empty.

• FRONT and REAR will be the same when a single element exists.

• A queue will be full when the REAR reaches the last index of the

array.

• Always FRONT <= REAR.

The linear queue operations are given in Figure 12.

Figure 12. Linear Queue operations

Circular Queue

Figure 13 displays the Linear Queue along with its disadvantage.

Figure 13. Linear Queue with its disadvantage

Disadvantages of this linear Queue are enqueuer operation in the above

Queue will generate an error message “Queue is full.” However, the Queue

has empty spots and cannot be used to save the element. To resolve this

disadvantage, a circular Queue can be introduced (Charatan & Kans, 2022).

Features of Circular Queue

The following features are associated with the Circular Queue.

• The last node of the Queue will be connected to the first node of the

Queue to form a circle shape, which is known as a circular Queue.

• It is more efficient than the linear queue data structure.

• Circular Queue accesses its data in the principle of First In First Out.

• Insert (Enqueue) operation is performed at FRONT at REAR &

Delete (Dequeue) operation.

• FRONT and REAR will point to -1 when Queue is empty.

• FRONT and REAR will be the same when a single element exists.

• FRONT can be > REAR, unlike the Linear Queue.

• Full only when either FRONT = 0 and REAR = last index or

FRONT = REAR+1.

The Circular Queue is specified in Figure 14, and the circular queue

examples are presented in Figure 15.

Figure 14. Circular Queue

Figure 15. Circular Queue examples

Sample Operations

Figure 16 provides a set of sample circular queue operations. These

operations are used to manipulate a circular queue data structure, a linear

data structure in which the first and last elements are connected to form a

circle. These operations are essential for implementing and working with a

circular queue data structure. By using these operations, we can add,

remove, and manipulate elements in a circular queue. They also enable us to

check the status of the circular Queue, such as if it is full or empty, and

access the Front and Rear elements of the Queue (Aggarwal, 2021).

Figure 16. Circular Queue operations

Double Ended Queue
A double-ended queue, a deque, is an ordered collection of queue-like

items. It has two ends, Front and Rear, so the item stays in the collection,

and the operations can be done at both ends.

Features of Double-Ended Queue

Here are the features that define the Double-Ended Queue.

• Double-ended Queue is also known as Deque (Not to be confused

with Dequeue).

• FRONT and REAR will point to -1 when Queue is empty.

• FRONT and REAR will be the same when there exists a single

element.

• Insert and delete can be performed on both ends (FRONT and

REAR)

• EnqueFront will perform insert at FRONT, and EnqueueRear will

perform insert at REAR.

• DequeueFront will perform delete at FRONT, and DequeRear will

perform delete at REAR.

The Double-ended Queue operations are represented in Figure 17.

Figure 17. Double-ended Queue operations

Priority Queue

A priority queue is an abstract data type resembling a simple queue data

structure, with the addition of a “priority” value for each constituent of the

Queue. Priority channels provide service to elements with a higher priority

before those with a lower priority (Baka, 2017).

Features of Double-Ended Queue

The following features can describe the Double-Ended Queue.

• A priority queue is a type of Queue in which each data item is

assigned an extra priority value, and items are accessed according to

their priority.

• Data elements with a higher priority are accessed antecedent to those

with a lower priority.

 • If more than one data items have a similar priority value, it can be

accessed in the order they were inserted into the priority queue.

Figure 18 depicts an example of a priority queue.

Figure 18. Priority Queue example

Python Code for Queue Operations

The Python code featured in Figure 19 and 20 is used for Queue operations.

Figure 19. Queue data structure Python implementation Part 1

Figure 20. Queue data structure Python implementation Part 2
APPLICATIONS OF QUEUE
Queue finds its applications in various domains, and some of them are listed

below.

1. Service requests from a single shared resource (CPU, Disk, Printer,

etc.) can be handled with the help of the Queue.

2. Queues are used for asynchronous data transfer between two

processes. Examples: File I/O, Pipes, and Sockets.

3. A call center phone system that uses queues to wait for a service

agent to become available.

4. Frequently, computer systems must “hold” messages between two

processes, programs, or even systems. This holding area is commonly

known as a “buffer.” It is implementable with a queue data structure.

5. The Round-robin processor scheduling algorithm is implemented

using queues.

To discover the shortest path between two vertices in a weighted graph

using search algorithms, artificial intelligence first attempts the most

probable paths. A priority queue keeps track of undiscovered routes,

prioritizing those with the minimum total path length.

DIFFERENCES BETWEEN STACK AND QUEUE DATA

STRUCTURES

A few differences between Stack and Queue are discussed below for better

understanding. The characteristics that distinguish Stack and Queue data

structures are outlined in Table 8.

Table 8. Differences between Stack and Queue data structures

S.No. Stack Queue

Data items can be accessed in a Data items can be accessed in a

1.
Last In First Out manner. First In First Out manner.
 2. Only one marker is used [Top]. Two markers are used [Front

and Rear].

The Top is the only marker that The Rear helps to insert data

3. helps do insertion and deletion items, and the Front helps to

operations. delete data items.

Insert operation done through Insert operation done through

4.
push(). Enqueue().

Delete operation done through Delete operation done through

5.
pop(). Dequeue().

There is a chance for memory

6. There is no memory wastage.
wastage.

CONCLUSION

The chapter thoroughly explained two essential linear data structures in

computer science, namely Stack and Queue, by demonstrating their key

features, operations, and implementation methods. Additionally, it has

showcased the versatility of Stack and Queue data structures by discussing

their real-world applications, such as infix to postfix conversion, postfix

expression evaluation, and finding the shortest path in a graph. The chapter

has included clear and concise Python code snippets to make it easier for

readers to comprehend and implement Stack and Queue data structures.

This chapter has been an excellent resource for computer science students,

researchers, and practitioners, providing a practical guide to understanding

and utilizing these data structures. By exploring the different

implementations and applications of Stack and Queue data structures, this

chapter has offered readers a comprehensive understanding of how to apply

them in various scenarios. Moreover, it has emphasized the critical

differences between Stack and Queue data structures, enabling readers to

choose the most appropriate structure for their needs. Overall, this chapter

has provided a valuable resource for anyone looking to deepen their

knowledge of Stack and Queue data structures.

REFERENCES

Aggarwal, S., & Kumar, N. (2021). Data structures. Advances in

Computers, 121, 43–81. doi:10.1016/bs.adcom.2020.08.002

Baka, B. (2017). Python Data Structures and Algorithms. Packt

Publishing Ltd.

Charatan, Q., & Kans, A. (2022). Python Collections: Lists and

Tuples. In Programming in Two Semesters: Using Python and Java

(pp. 105–128). Springer International Publishing. doi:10.1007/978-

3-031-01326-3_6

Cutler, J., Dickenson, M., Cutler, J., & Dickenson, M. (2020).

Introduction to Data Structures. Computational Frameworks for

Political and Social Research with Python, 59-71.

Dhruv, A. J., Patel, R., & Doshi, N. (2021). Python: the most

advanced programming language for computer science

applications. In Proceedings of the international conference on

culture heritage, education, sustainable tourism, and innovation

technologies (CESIT 2020) (pp. 292-299). Academic Press.

Goodrich, M. T., Tamassia, I. R., & Goldwasser, M. H. (2013).

Data Structures and Algorithms in Python. John Wiley & Sons.

Harris, C. R., Millman, K. J., Van Der Walt, S. J., Gommers, R.,

Virtanen, P., Cournapeau, D., Wieser, E., Taylor, J., Berg, S.,

Smith, N. J., Kern, R., Picus, M., Hoyer, S., van Kerkwijk, M. H.,

Brett, M., Haldane, A., del Río, J. F., Wiebe, M., Peterson, P., &

Oliphant, T. E. (2020). Array programming with NumPy. Nature,

585(7825), 357–362. doi:10.1038/s41586-020-2649-2

Lambert, K. (2019). Fundamentals of Python: Data Structures.

Cengage Learning.

Mailund, T. (2021). Stacks and Queues. Introduction to

Computational Thinking: Problem Solving, Algorithms, Data

Structures, and More, 547-582.

Vasudevan, S. K., Nagarajan, A. S., & Nanmaran, K. (2021). Data

Structures using Python. Oxford University Press.

 CHAPTER 5

Dynamic Programming With Python

Gurram Sunitha

School of Computing, Mohan Babu University, Tirupati, India

Arman Abouali

Technical University Clausthal, Germany

Mohammad Gouse Galety

https://orcid.org/0000-0003-1666-2001

Samarkand International University of Technology, Uzbekistan

A. V. Sriharsha

Mohan Babu University, India

ABSTRACT

Algorithms are at the heart of computer programming. They form the basis

of all software applications and help to solve complex computational

problems. Various problem-solving strategies involve divide and conquer,

greedy, recursion, dynamic programming, backtracking, etc. This can be

used to solve optimization problems that have overlapping subproblems. It

aims to find an optimal solution by breaking down a problem into sub-

problems in order to manage the complexity of the problem-solving and

remembering their solutions in order to avoid repeating computation time in

future steps. Mathematical optimization is a crucial component of dynamic

programming, and it allows us to efficiently solve a wide range of problems

that involve making decisions over time. This chapter discusses dynamic

programming's relevance, mathematical optimization, ability to solve a

wide range of issues, important qualities, and top-down and bottom-up

problem-solving methodologies. Dynamic programming solves some

typical computational problems effectively, and Python code is supplied for

reference.

INTRODUCTION

Algorithms are at the heart of computer programming. They form the basis

of all software applications and help to solve complex computational

problems. Various problem-solving strategies involve divide & conquer,

greedy, recursion, dynamic programming, backtracking, etc. In optimization

problems, greedy algorithms shall be used to find the best outcome given a

set of conditions. The backtracking algorithm is used for problems that

require exploring multiple solutions before finding the best one. The divide

& conquer technique breaks a problem into smaller parts so that each part

can be solved separately before being combined for a final solution. Finally,

dynamic programming aims to find an optimal solution by breaking down a

problem into sub-problems to manage the complexity of the problem-

solving and remembering their answers to avoid repeating computation time

in future steps. This chapter explores the fundamentals of dynamic

programming and presents a detailed discussion of various computational

problems solved using dynamic programming. Also, Python code is

provided for reference.

DYNAMIC PROGRAMMING

Dynamic Programming is an essential algorithmic technique used to solve

complex problems efficiently. It is widely used in various fields, such as

computer science, operations research, and economics. Dynamic

programming can solve optimization problems with overlapping

subproblems by breaking them down into smaller subproblems and storing

their solutions in a table for later use. This technique helps reduce the time

complexity of these problems and makes them easier to solve (Kool et al.,

2022). Dynamic programming is beneficial when dealing with issues that

depend on previously calculated solutions or where multiple decisions must

be made simultaneously (Şenaras, et al., 2021).

“Dynamic Programming” makes sense as dynamic means' changing' and

programming means 'the set of instructions to be executed.' This was

introduced by “Richard Bellman” in the 1950s, which is a mathematical

optimization and a computer programming technique. This has been used in

various disciplines, from economics to aeronautical engineering. Let us

imagine what happens if a non-optimized code consumes much memory

and computing time. Consequently, the processor may need help allocating

resources for other services. Most of these problems can be controlled by

using dynamic programming (Bertsekas, 2022).

Expressed, this is a technique for resolving complex problems by breaking

them down into smaller, more manageable sub-problems and archiving the

solutions for use when the smaller sub-problems come up later. The

subproblems are computed to find the optimal solution to the complex

problem. In brief words, this is an optimization technique over plain

recursion.

Mathematical Optimization in Dynamic Programming

Mathematical optimization plays a significant role in dynamic

programming, a powerful technique for solving problems involving making

decisions over time. In dynamic programming, a problem is typically

broken down into smaller sub-problems that can be solved recursively. The

optimization part of dynamic programming involves finding the best

decision at each sequence step to maximize some objective function. The

objective function being maximized could be anything, depending on the

problem (Taylor, 2019). For instance, the goal of a scheduling task can be to

reduce the overall time complexity of the execution of a batch of jobs.

Minimizing the overall cost of retaining inventory over time may be the

goal function in an inventory management problem.

Optimization in dynamic programming can be done in several ways,

depending on the problem being solved. One common technique is using a

dynamic programming algorithm such as the Bellman-Ford or Dijkstra

algorithms. These algorithms work by recursively computing the optimal

decisions at each step in the sequence and using these decisions to build up
a solution to the overall problem. Another technique for optimization in

dynamic programming is to use linear programming. Linear programming

is an essential problem-solving technique with applications in various fields.

It can maximize or minimize operational resources, allocate resources

efficiently, develop production and distribution plans, and solve

optimization problems. Through linear programming, organizations can

make more informed decisions that result in improved profits and outcomes.

This technique is beneficial when there are multiple possible solutions with

uncertain variables, as it enables organizations to make data-driven

decisions based on strict parameters. Overall, mathematical optimization is

a crucial component of dynamic programming, and it allows us to

efficiently solve a wide range of problems that involve making decisions

over time (Liu et al., 2020).

Key Attributes of Dynamic Programming

Dynamic programming is an algorithm-based approach to finding a

problem's optimal solution. It involves breaking down a complex problem

into smaller subproblems and solving each separately. With dynamic

programming, reducing the time needed to solve a problem significantly by

storing intermediate results in memory is possible. It can be applied to

various issues, from optimization and planning to game theory and machine

learning. It should be emphasized that the best solution to each subproblem

affects the overall situation. The critical component of the optimization

process in dynamic programming is the multistage aspect of problem-

solving (Cooper, 2016). Dynamic programming offers a more all-

encompassing framework for problem analysis than the other optimization

methods.

Ingenuity is usually required before a particular challenge can be effectively

comprehended as a dynamic programme. Delicate insights are frequently

needed to reconstruct the formulation to solve it successfully. The steps in

solving a problem using dynamic programming are as follows (Kennedy,

2012):
 • Define the Problem and Identify the Subproblems: Dynamic

programming is often used to solve problems that can be broken down

into smaller subproblems. Define the problem you want to solve and

identify the subproblems that will help you reach the solution.

• Formulate a Recursive Relation: Once you have identified the

subproblems, you can formulate a recursive relation that describes the

solution to each subproblem in terms of the solutions to smaller

subproblems. This is often called the “optimal substructure” property.

• Create a Memoization Table: To avoid redundant computations,

create a table to store the solutions to the subproblems as you compute

them. This is called memoization, a key feature of dynamic

programming.

• Compute the Solutions to the Subproblems: Starting with the

smallest subproblems, use the recursive relation to compute the

solutions to larger subproblems, filling in the memoization table as you

go.

• Build the Solution to the Original Problem: Once you have

computed the solutions to all of the subproblems, you can use them to

build the solution to the actual problem.

• Analyze the Time and Space Complexity: As with any algorithm,

it is essential to analyze your dynamic programming solution's time

and space complexity to ensure it is efficient enough for your needs.

Its subproblem should possess the following key attributes to solve any

problem using dynamic programming. They include:

• Overlapping Subproblems: Subproblems are said to be overlapping

in nature if the subproblems are calculated multiple times to compute

the original solution of the problem.

• Optimal Substructure: Any problem is said to have an optimal

substructure if the complex problem's optimal solution is obtained by

recursively solving the subproblem.

To solve a problem using a dynamic programming technique, it must have

both overlapping subproblems and optimal substructure properties.

Other key attributes of Dynamic Programming include:

• Memoization stores the solutions to subproblems in a table or array

for future reference. Memoization is used to avoid the repeated

computation of subproblems and speed up the overall problem-solving

process.

• Computational Complexity: Dynamic Programming can solve

problems that would otherwise be computationally infeasible. By

breaking the problem down into smaller subproblems and storing the

solutions to these subproblems, Dynamic Programming can reduce the

computational complexity of a problem from exponential to

polynomial.

• Greedy Approach: In some cases, Dynamic Programming can be

used to solve problems using a Greedy method. This involves making

locally optimal choices at each step of the problem, which leads to a

globally optimal solution. However, not all problems can be solved

using a greedy approach, and in these cases, Dynamic Programming

can provide an alternative solution.

Top-Down Approach of Dynamic Programming

Bottom-up dynamic programming is an efficient and effective way to solve

programming problems. It is based on breaking down complex issues into

simpler subproblems and then systematically solving them. This approach

enables the programmer to efficiently solve a problem by solving its smaller

subproblems and combining their solutions. Doing so eliminates the need

for recomputing the key multiple times, thus saving time and memory

space. This makes it ideal for many real-world applications, such as solving

complex optimization problems, parsing natural language texts, AI planning

and scheduling tasks, etc. When the function is called with a particular

problem instance, it first checks whether the solution for that instance is

already in the memoization table. If it is, the function returns the

precomputed solution. Otherwise, it recursively computes the answer for

that instance while storing the result in the memoization table for future use

(Tarek et al., 2020).

The top-down approach of dynamic programming involves the following

general steps:

1. Identify the Subproblems: Break down the problem into smaller

subproblems. Identify the recurring patterns and overlapping

subproblems.

2. Define the Memoization Table: Create a table that stores the

solutions to the subproblems. This table can be a 2D array or a hash

table, depending on the problem.

3. Define the Base Cases: Define the base cases for the subproblems.

The base cases are the simplest subproblems that can be solved directly

without further recursion.

4. Define the Recursive Function: Define a recursive function that

solves the subproblems using the memoization table. This function

should check the memoization table first to see if the solution to the

subproblem has already been computed. If it has not, then the function

should solve the subproblem recursively.

5. Populate the Memoization Table: Call the recursive function with

the initial parameters and populate the memoization table with the

solutions to the subproblems.

6. Return the Final Solution: Once the memoization table has been

populated, return the solution to the original problem, which will be

stored in the memoization table at the appropriate position.

Overall, the top-down approach involves breaking down the problem into

smaller subproblems, solving them recursively using memoization to avoid

redundant computations, and returning the solution to the original problem

based on the solutions to the subproblems.

The essential advantage of the top-down approach is that it allows us to

avoid recomputing the same subproblems multiple times. By caching the

intermediate results, the exponential time complexity of the algorithm can

be managed to a polynomial level.

Bottom-Up Approach of Dynamic Programming

The bottom-up approach of dynamic programming is a technique used to

solve optimization problems by breaking them down into smaller

subproblems and solving them iteratively. In this approach, we start by

solving the subproblems and then combine their solutions to find the

solution to the original problem.

We typically use a table to store the solutions to the subproblems to

implement the bottom-up approach. The table is usually filled

systematically, starting from the smallest subproblems and gradually

building up to the larger ones until we solve the original problem.

Here are the general steps involved in implementing the bottom-up

approach of dynamic programming:

1. Identify the Subproblems: Break down the problem into smaller

subproblems. Identify the recurring patterns and overlapping

subproblems.

2. Define the Memoization Table: Create a table that stores the

solutions to the subproblems. This table can be a 2D array or a hash

table, depending on the problem.

3. Define the Base Cases: Define the base cases for the subproblems.

The base cases are the simplest subproblems that can be solved directly

without further recursion.

4. Populate the Memoization Table: Fill in the memoization table

with the solutions to the subproblems in a bottom-up manner. Start

with the base cases and work up to the final solution.

5. Return the Final Solution: Once the memoization table has been

fully populated, return the solution to the original problem, which will

be stored in the memoization table at the appropriate position.

Overall, the bottom-up approach involves breaking down the problem into

smaller subproblems, solving them iteratively using memoization to avoid

redundant computations, and returning the solution to the original problem

based on the solutions to the subproblems. The bottom-up approach of

dynamic programming is usually more efficient than the top-down approach

because it avoids the overhead of recursion and memoization. However, it

may require more memory to store the table of solutions.

Applications of Dynamic Programming

Dynamic programming is a powerful algorithmic technique that can be

applied to various problems. Here are some examples of its applications

(Bellman, 1957; Rosso & Venturino, 2023; Terek, 2020):

• Optimization Problems: Dynamic programming can solve

optimization problems where the goal is to maximize or minimize

some function. Examples include the knapsack, traveling salesman,

and most extended common subsequence problems.

• Path-Finding Problems: Dynamic programming can be used to find

the shortest or longest path in a graph or grid. Examples include

Dijkstra's algorithm and the Floyd-Warshall algorithm.

• Game Theory: Dynamic programming can be used to analyze and

solve games, such as the minimax algorithm for two-player games.

• Machine Learning: Dynamic programming is used in various

machine learning techniques, such as reinforcement learning, where an

agent learns to make decisions by maximizing a reward function.

• Image Processing: Dynamic programming can be used for image

analysis and processing tasks such as object tracking, segmentation,

and feature detection.

• Natural Language Processing: Dynamic programming can be used

in natural language processing tasks such as speech recognition,

machine translation, and text summarization.

• Bioinformatics: Dynamic programming is widely used for sequence

alignment, gene prediction, and protein structure prediction.

• Financial Modeling: Dynamic programming can model financial

decisions such as portfolio optimization, option pricing, and risk

management.

Dynamic programming is a versatile and powerful tool for solving complex

problems across many domains. Further, this chapter presents a detailed

discussion of various computational problems solved using dynamic

programming – Fibonacci Series, Towers of Hanoi Puzzle, Checkered

Board Problem, and Dijkstra's Algorithm (Li et al., 2020; Senaras et al.,

2020; Li et al., 2020).

FIBONACCI SERIES

The Fibonacci series is an essential mathematical concept in computer

science. It can solve many problems, such as computing the nth Fibonacci

number and more.

Dynamic programming can be understood with a simple example of the

Fibonacci series. Let us consider generating the “nth” term of the Fibonacci
th
series as a complex problem to be solved. The N term of the Fibonacci
th th
series is equal to the sum of (N-1) Fibonacci term and (N-2) Fibonacci

term. The Fibonacci series problem and the inherent recurrence of the

subproblems are shown in Figure 1. Observe the subproblems being

repeated (similarly colored for visual reference). In the figure showcasing

the first two levels of problem division into smaller tasks or subproblems, it

can be observed that subproblem-2 is repeated, and subproblem-11 is

repeated.

Figure 1. Fibonacci series problem and the inherent recurrence of the

subproblems
The Fibonacci series problem inherently comprehends two features – its

subproblem's solutions depend on previously calculated solutions, and

subproblems repeatedly occur over time. Let us understand with the help of

an example. For example, the complex problem is F(5), where F is the

th
Fibonacci function (the task is to generate the 5 term of the Fibonacci

series). The process of F(5) is described as follows.

a. To calculate F(5), we need to calculate F(4) and F(3) terms as F(5)

is expressed as the sum of F(4) and F(3). (F(4) and F(3) are

subproblems)

b. To calculate F(4), we need to calculate F(3) and F(2) terms as F(4)

is expressed as the sum of F(3) and F(2). Here the subproblem F(3) is

repeated. (F(3) and F(2) are subproblems)

c. To calculate F(3), we need to calculate F(2) and F(1) terms as F(3)

is expressed as the sum of F(2) and F(1). Here the subproblem F(2) is

repeated. (F(2) and F(1) are subproblems)

and so on.
This process of calculating the subproblems multiple times to obtain the

optimal solution to obtain F(5) is known as the overlapping of subproblems.

Also, observe that solution of a Fibonacci series problem (or subproblem)

depends on the resolution of its subproblems. Hence, the Fibonacci series

problem inherently exhibits both the required features; therefore, dynamic

programming is the best technique to solve the problem.

Generating Fibonacci Series Using Dynamic

Programming With Python

One way to solve the problem of generating the first N Fibonacci terms is to

use dynamic programming with a top-down approach. This approach

utilizes memoization to store intermediate results and can significantly

reduce the time complexity of a given problem. This section will explore

using this approach to solve Fibonacci series-related problems. Dynamic

programming for solving the Fibonacci series problem using a top-down

approach works as follows - the concept is similar to the recursive

technique. Still, the answers to any of the subproblems are preserved when

the subproblems are encountered and solved for the first time. The held

result will be used in the top-down approach whenever the subproblem is

reencountered. As a result, each subproblem has to be solved precisely

once. This provides savings in the computational time for solving the

problem (Forootani et al., 2021; Xu & Wu, 2021; Xie et al., 2021).

The following is the Python code for generating the first N Fibonacci terms

using dynamic programming (top-down approach). Sample input and output

are also presented. We will start with a top-down approach to dynamic

programming.

# Initializing the first two terms of the

Fibonacci Series
fib_list = [0, 1]
# Defining a function fibonac() to generate the
Fibonacci Series
def fibonac(n):
# If the number 'n' is less than or equal to
two, return (n-1)th
# Fibonacci term
if n <= len(fib_list):
return fib_list[n-1]
# If the number 'n' is more significant than
two, generate the Fibonacci
# Series
else:
# Computing the nth term in Fibonacci
Series
next = fibonac(n-1) + fibonac(n-2)
# Storing the next term in the list
fib_list.append(next)
# Displaying the next term
print((next), end = ' ')
# Returning the Fibonacci term
return next
# Prompting user to enter his choice for the
number of Fibonacci terms to # be generated
n = int (input (“How many Fibonacci terms are to
be generated? \
(Enter any number > 2): “))
# Displaying the first two terms of the Fibonacci
Series
print (0, 1, end = ' ')
# Calling function fibonac() to generate Fibonacci
Series
fibonac(n)

Input: How many Fibonacci terms are to be generated? (Enter any number

> 2): 5

Output: 0 1 1 2 3

In dynamic programming using the bottom-up approach, the subproblems

are reordered for solving. Only two intermediate results are to be stored at a
time, which will allow computing the solution to each subproblem only

once. For instance, all that is required are the answers to F(1) and F(0)

when trying to generate F(2). Similarly, generating F(2) and F(3) requires

that we have the answers to F(2) and F(1). As a result, the solution will

consume less memory. The following is the Python code for generating the

Fibonacci series using dynamic programming (bottom-up approach).

# Prompting user to enter his choice for the

number of Fibonacci
# terms to be generated
N = int (input (“How many Fibonacci terms will be
generated?
(Enter any number > 2): “))
# If the number 'N' is equal to zero or one,
display the first
# ' N' Fibonacci terms and exit; If the number 'N'
is greater than # or equal to two, display the
first 'N' Fibonacci terms
if N == 0:
exit()
elif N == 1:
print (0, end = ' ')
exit()
elif N >= 2:
print (0, 1, end = ' ')
# initializing the first two Fibonacci terms
A = 0
B = 1
# Generating the remaining Fibonacci terms
for i in range(3, N+1):
# Calculate the next Fibonacci Term
temp = A + B
# Displaying Fibonacci term
print (temp, end = ' ')
# storing the most recent two Fibonacci terms
 A = B
B = temp

Input: How many Fibonacci terms are to be generated? (Enter any number

> 2): 8

Output: 0 1 1 2 3 5 8 13

For the top-down approach, each subproblem is solved only once. Since

each subproblem takes a constant amount of time to solve, this gives an

O(N) time complexity. However, since an array of size N + 1 needs to be

maintained to save all the intermediate results, the space complexity is

O(N). In the bottom-up approach, each subproblem is solved only once. So,

the time complexity of the algorithm is O(N). However, only the most

recent two Fibonacci terms are only to be maintained; the space complexity

is a constant, which is O(1).

TOWERS OF HANOI PUZZLE

The Towers of Hanoi problem can be solved using dynamic programming

using a memoization approach. The problem statement is as follows: given

three towers and n disks (each of the disks of a different size), the objective

is to move the entire stack of disks from one tower to another tower with the

following restrictions:

• Only one disk can be moved at a time.

• Each move consists of taking the upper disk from one of the stacks

and placing it on top of another.

• No disk may be placed on top of a smaller disk.

To solve this problem using dynamic programming, a function Hanoi(n,

source, target, aux) that takes four arguments can be defined where:

• n: is the number of disks to move

• source: is the tower where the disks are initially located

• aux: is an auxiliary tower to help with the move

• target: is the tower where the disks should be moved to

The base case for this function is when n = 1. In this case, the top disk is

moved from the source tower to the target tower. For n > 1, memoization

can be used to avoid redundant computations. The result of each call to

Hanoi (n, source, target, aux) shall be stored in a table, indexed by the

values of n, source, target, aux. If a call to Hanoi (n, source, target, aux)

already in the table is encountered, simply the stored result shall be

returned. The recursive algorithm for Hanoi (n, source, target, aux) is as

follows:

1. If n = 1, move the top disk from source to target.

2. Otherwise, recursively move n-1 disks from source to aux, using

target as the auxiliary tower.

3. Move the largest disk from source to target.

4. Recursively move n-1 disks from aux to target, using the source as

the auxiliary tower.

The Python code for the dynamic programming solution (top-down

approach) to the Towers of Hanoi problem is as follows:

# define a memoization table to store the results

of subproblems
memo = {}
def Hanoi(n, source, target, aux):
# Check if we have already computed the
solution for this
# subproblem
if (n, source, target, aux) in memo:
return memo[(n, source, target, aux)]
# In the situation of the base case, the top
disk has to be moved
# from the source tower to the target tower
if n == 1:
move(source, target)
memo[(n, source, target, aux)] = 1
return 1
# In the situation of Recursive case, n-1
number of disks have to
# moved from the source tower to the auxiliary
tower
else:
Hanoi(n-1, source, aux, target)
move(source, target)
Hanoi(n-1, aux, target, source)
# Update the memoization table with the
solution to this
# subproblem
memo[(n, source, target, aux)] =
memo.get((n-1, source, aux,
target), 0) + 1 + memo.get((n-1, aux,
target, source), 0)
return memo[(n, source, target, aux)]
def move(source, target):
print(“Disk moved “, source, “->”, target)
# example usage:
n = int(input(“How many number of disks are to be
moved? “))
print(“Total number of moves required:, “Hanoi(n,
'A', 'C', 'B'))

Input:

How many number of disks are to be moved? 3

Output:

Disk moved A -> C

Disk moved A -> B

Disk moved C -> B

Disk moved A -> C

Disk moved B -> A

Disk moved B -> C

Total Number of Moves Required: 7

This function takes four arguments: the number of disks to be moved (n),

source tower (source), auxiliary tower (aux), and target tower (target). It

also takes an optional memo dictionary, which stores the results of previous

function calls. The function first checks whether the result of this particular

call to Hanoi is already in the memo dictionary. If it is, it returns the stored

result. Otherwise, it checks whether n is equal to 1. If it is, it simply moves

the top disk from the source tower to the target tower and stores this move

in the memo dictionary. If n is greater than 1, the function recursively calls

itself to move n-1 disks from the source tower to the auxiliary tower, using

the target tower as the auxiliary tower. It then stores the result of this

recursive call in the memo dictionary. It then moves the largest disk from

the source tower to the target tower and adds this move to the result list.

Finally, it recursively calls itself to move n-1 disks from the auxiliary tower

to the target tower, using the source tower as the auxiliary tower.

DIJKSTRA'S ALGORITHM

Dijkstra's algorithm is famous for finding the shortest path in a weighted

graph with non-negative weights. Dynamic programming can be used to

implement this algorithm efficiently, which involves breaking down a more

significant problem into smaller subproblems and solving each subproblem

only once. Dijkstra's algorithm is a popular approach for finding the shortest

path from a source node to all other nodes in a weighted graph. The basic

idea behind Dijkstra's algorithm is to start at the source node and repeatedly

visit the unvisited neighbor with the smallest tentative distance until all

nodes have been processed. It is commonly implemented using a priority

queue to maintain the vertices that have not yet been processed.

A top-down approach to implementing Dijkstra's algorithm using dynamic

programming involves the following steps:

1. Initialize the distance to the source node as 0 and the distance to all

other nodes as infinity.

 2. Create a priority queue (min heap) to maintain the set of unvisited

nodes.

3. Insert the source node into the priority queue with a priority of 0.

4. While the priority queue is not empty, do the following:

a. Extract the node with the smallest tentative distance from the

priority queue.

b. For each unvisited neighbor of the extracted node, do the

following:

i. Calculate the tentative distance to the neighbor by adding

the weight of the edge between the two nodes to the distance

of the extracted node.

ii. If the tentative distance is smaller than the neighbor's

current distance, update the neighbor's distance in the

distance array and insert the neighbor into the priority queue

with the new distance as its priority.

5. After all, nodes have been processed; the distance array will contain

the shortest distance from the source node to all other nodes.

This top-down approach uses memoization to avoid redundant calculations.

It is beneficial when the graph is large and sparse, allowing us to focus on

the subproblems needed to find the shortest path. A Python code

implementation top-down dynamic programming solution to find the

shortest distance between a start node to all the other nodes in the weighted

graph using Dijkstra's algorithm is as follows:

import heapq as hp
# Dijkstra() function returns a list of shortest
distances from the
# start node to all the other nodes in the graph
def dijkstra(graph, start):
# Initialize distances to all nodes
distances = {node: float('inf') for node in
graph}
distances[start] = 0
# Create a heap to store nodes that need to be
processed
 heap = [(0, start)]
# Iterate until all nodes have been processed
while heap:
# Get the node with the smallest distance
from the heap
(curr_dist, curr_node) = hp.heappop(heap)
# If the current distance is > than the
previously
# calculated distance, skip it
if curr_dist > distances[curr_node]:
continue
# Update the distances of all the
neighbors of current node
for neighbor, weight in
graph[curr_node].items():
distance = curr_dist + weight
# Update the distance of current node
if required
if distance < distances[neighbor]:
distances[neighbor] = distance
# Add the neighbor to the heap to
be processed
hp.heappush(heap, (distance,
neighbor))
# Return the list of shortest distances
return distances
# Sample graph
graph = {
1: {2: 2, 3:4, 4:2},
2: {3: 3, 4:1},
3: {1: 4},
4: {3:1}
}
# Sample function call to find shortest routes
from node 1 to all
# other nodes
d = dijkstra(graph, 1)
print(d)

Output:

{1: 0, 2: 2, 3: 3, 4: 2}

The graph parameter in the above code is assumed to be a list of adjacency

lists, where graph[i] is a list of pairs {j: w} representing the edges going out

from node i, where j is the index of the neighbor node and w is the weight

of the edge. 'start' is the starting node for the algorithm. The algorithm uses

a heap to store the graph nodes that need processing. The heap is

maintained with the nodes with the smallest distance from the start node at

the top. The algorithm iterates until all nodes have been processed, updating

the distances to each neighbor of the current node if the new distance is

shorter than the current distance (Gouse et al., 2019; Reshma et al., 2022).

CHECKERED BOARD PROBLEM

The Checkered Board Problem is a classic puzzle in mathematics. The

problem can be stated as follows:

Given a checkered board of size n x n, where n is an even

number, and two types of tiles - black and white, how

many ways are there to fill the checkered board with

these tiles such that no two adjacent tiles have the same

color?

The Checkered board problem can be solved using the top-down dynamic

programming approach. Let us define a function countWays(i, j, color) that

calculates the number of ways to fill the checkered board of size “i x j” with

the given color of the top-left cell. The color parameter can be either “B” or

“W” for black or white.

The base cases of the function are as follows:

1. If the size of the board is 1 x 1, return 1.

 2. If the board size is 1 x 2, return 2 if the top-left cell has the same

color as the second cell; otherwise, return 1.

3. If the board size is 2 x 1, return 2 if the top-left cell has the same

color as the second cell; otherwise, return 1.

For the recursive case, we need to consider two issues:

1. If the top-left cell has the same color as the second cell, we can only

fill the rest of the board with alternating colors. So, we need to

calculate the number of ways to fill the remaining board with the

opposite color. This can be done by calling the countWays(i-2, j,

oppositeColor) function, where oppositeColor is the contrasting color

of the top-left cell.

2. If the top-left cell has a different color than the second cell, we can

either fill the second cell with the same color as the top-left cell or

with the opposite color. So, we need to calculate the number of ways to

fill the remaining board in both cases and add them up. This can be

done by calling the countWays(i-2, j, sameColor) and countWays(i-2, j,

oppositeColor) functions, where sameColor is the same color as the

top-left cell and oppositeColor is the opposite color of the top-left cell.

Finally, we need to memoize the results of the function to avoid redundant

computations. Here is the Python code for the solution:

memo = {}
def countWays(rows, cols, color):
if (rows, cols, color) in memo:
return memo[(rows, cols, color)]
if rows == 1 and cols == 1:
return 1
elif rows == 1 and cols == 2:
return 2 if color == “B” else 1
elif rows == 2 and cols == 1:
return 2 if color == “B” else 1
sameColor = color
oppositeColor = “B” if color == “W” else “W”
 if color == “B”:
oppositeColor = “W”
else:
oppositeColor = “B”
if color == “B”:
if (rows-2, cols, oppositeColor) not in
memo:
memo[(rows-2, cols, oppositeColor)] =
countWays(rows-2, cols,
oppositeColor)
return memo[(rows-2, cols, oppositeColor)]
else:
if (rows-2, cols, sameColor) not in memo:
memo[(rows-2,cols,sameColor)]=
countWays(rows-
2,cols,sameColor)
print(countWays(2,1,”B”))

Output:

Since they make it possible to solve challenging computational problems,

algorithms are crucial to computer programming. Dynamic programming is

an algorithmic technique that addresses optimization challenges with

overlapping subproblems by dividing issues into smaller ones and

memorizing the solutions. Mathematical optimization is a crucial element

of dynamic programming that enables long-term problem solutions. The

importance of dynamic programming, its versatility, and how it may be

used to solve problems are all emphasized in this chapter. Additionally, the

chapter covers top-down and bottom-up problem-solving strategies utilizing

dynamic programming and offers Python code as a resource (Gouse et al.,

2018; Samanta et al., 2021; Gouse et al., 2021).

CONCLUSION
Dynamic programming is an algorithmic approach that enables us to solve

complex problems by breaking them into smaller sub-problems and solving

each optimally. It can identify which parts of the problem can be reused

from previous attempts at solving it, thus saving time and resources. In this

perspective, this chapter presented a detailed discussion of dynamic

programming - its importance, mathematical optimization, skill in solving

various problems, key attributes, and the top-down and bottom-up

approaches for problem-solving. A few of the famous computational

problems are undertaken to solve them efficiently using dynamic

programming, and the Python code is provided for reference. Dynamic

programming is a powerful tool for tackling complex issues, particularly

involving large amounts of data or multiple inputs. This method is often

used in robotics and artificial intelligence to improve efficiency and

accuracy. It also has applications in finance, economics, logistics, computer

vision, and natural language processing.

REFERENCES

Bellman, R. E. (1957). Dynamic programming, ser. Cambridge

Studies in Speech Science and Communication. Princeton

University Press.

Bertsekas, D. (2022). Abstract dynamic programming. Athena

Scientific.

Cooper, L., & Cooper, M. W. (2016). Introduction to Dynamic

Programming: International Series in Modern Applied

Mathematics and Computer Science (Vol. 1). Elsevier.

Forootani, A., Tipaldi, M., Ghaniee Zarch, M., Liuzza, D., &

Glielmo, L. (2021). Modelling and solving resource allocation

problems via a dynamic programming approach. International

Journal of Control, 94(6), 1544–1555.

doi:10.1080/00207179.2019.1661521
Gouse, G. M., & Ahmed, A. N. (2019). Ensuring the public cloud

security on scalability feature. Journal of Advanced Research in

Dynamical and Control Systems, 11(1), 132–137.

Gouse, M. (2018). Optimized Secure Scan Flip Flop to Thwart Side

Channel Attack in Crypto-Chip. Academic Press.

Gouse, M. (2021). Deep Neural Network Concepts for

Classification using Convolutional Neural Network: A Systematic

Review and Evaluation. Technium Romanian Journal of Applied

Sciences and Technology., 3(8), 58–70.

doi:10.47577/technium.v3i8.4554

Kennedy, J. O. (Ed.). (2012). Dynamic programming: applications

to agriculture and natural resources. Springer Science & Business

Media.

Kool, W., van Hoof, H., Gromicho, J., & Welling, M. (2022, June).

Deep policy dynamic programming for vehicle routing problems.

In Integration of Constraint Programming, Artificial Intelligence,

and Operations Research: 19th International Conference, CPAIOR

2022, Los Angeles, CA, USA, June 20-23, 2022, Proceedings (pp.

190-213). Cham: Springer International Publishing.

doi:10.1007/978-3-031-08011-1_14

Li, G., Rana, M. N., Sun, J., Song, Y., & Qu, J. (2020). Real-time

image enhancement with efficient dynamic programming.

Multimedia Tools and Applications, 79(41-42), 30883–30903.

doi:10.1007/s11042-020-09586-y

Li, X., Ye, X., & Lu, L. (2020). Dynamic Programming

Approaches for Solving Shortest Path Problem in Transportation:

Comparison and Application. In Green, Smart and Connected

Transportation Systems: Proceedings of the 9th International

Conference on Green Intelligent Transportation Systems and Safety

(pp. 141-160). Springer Singapore.

Liu, D., Xue, S., Zhao, B., Luo, B., & Wei, Q. (2020). Adaptive

dynamic programming for control: A survey and recent advances.

IEEE Transactions on Systems, Man, and Cybernetics. Systems,

51(1), 142–160. doi:10.1109/TSMC.2020.3042876

Reshma, G., Al-Atroshi, C., Nassa, V. K., Geetha, B. T., Sunitha,

G., Galety, M. G., & Neelakandan, S. (2022). Deep Learning-

Based Skin Lesion Diagnosis Model Using Dermoscopic Images.

Intelligent Automation & Soft Computing, 31(1).

Rosso, A., & Venturino, E. (2023). A Dynamic Programming

Approach to Ecosystem Management. Algorithms, 16(3), 139.

doi:10.3390/a16030139

Samanta, D., Dutta, S., Gouse, M., & Pramanik, S. (2021). A Novel

Approach for Web Mining Taxonomy for High-Performance

Computing. doi:10.1007/978-981-16-4284-5_37

Şenaras, A. E., İnanç, Ş., Sezen, H. K., & Şenaras, O. M. (2021).

Shortest Route Application via Dynamic Programming in the

Transportation Networks. In Handbook of Research on Decision

Sciences and Applications in the Transportation Sector (pp. 362–

371). IGI Global. doi:10.4018/978-1-7998-8040-0.ch017

Tarek, A., Elsayed, H., Rashad, M., & Hassan, M. (2020, October).

Dynamic programming applications: a suvrvey. In 2020 2nd Novel

Intelligent and Leading Emerging Sciences Conference (NILES)

(pp. 380-385). IEEE. doi:10.1109/NILES50944.2020.9257968

Taylor, C. R. (2019). Applications of dynamic programming to

agricultural decision problems. CRC Press.

doi:10.1201/9780429040917

Xie, F., Li, H., & Xu, Z. (2021). An approximate dynamic

programming approach to project scheduling with uncertain

resource availabilities. Applied Mathematical Modelling, 97, 226–

243. doi:10.1016/j.apm.2021.03.048

Xu, J., & Wu, S. (2021, April). Analysis and Application of

Dynamic Programming. Journal of Physics: Conference Series,

1865(4), 042023. doi:10.1088/1742-6596/1865/4/042023

 CHAPTER 6

Exploratory Data Analysis in Python

R. Sruthi

Coimbatore Institute of Technology, India

G. B. Anuvarshini

Coimbatore Institute of Technology, India

M. Sujithra

Coimbatore Institute of Technology, India

ABSTRACT

Data science is extremely important because of the immense value of data.

Python provides extensive library support for data science and analytics,

which has functions, tools, and methods to manage and analyze data.

Python Libraries are used for exploratory data analysis. Libraries in Python

such as Numpy, Pandas, Matplotlib, SciPy, etc. are used for the same. Data

visualization's major objective is to make it simpler to spot patterns, trends,

and outliers in big data sets. One of the processes in the data science

process is data visualization, which asserts that after data has been gathered,

processed, and modelled, it must be represented to draw conclusions. As a

result, it is crucial to have systems in place for managing and regulating the

quality of corporate data, metadata, and data sources. So, this chapter

focuses on the libraries used in Python, their properties, functions, how few

data structures are related to them, and a detailed explanation about their

purpose serving as a better foundation for learning them.

INTRODUCTION
The vital process of performing primary analyses on data to explore new

trends, patterns, identify outliers, hypotheses, and cross-check assumptions

using summary of data's statistics and graphical representations is known as

exploratory data analysis. Exploratory Data Analysis (EDA) is not a formal

procedure with rigid guidelines. The approach to EDA emphasizes open-

mindedness and the exploration of various notions in the early stages of

analysis. Some concepts utilized may succeed while others may fail,

highlighting the beauty of EDA as a means of exploring uncertainty.

Despite the availability of predefined methods and techniques, exploratory

data analysis remains a crucial component in any data analysis. As research

progresses, certain fruitful areas will be discovered, documented, and

shared with others.

PYTHON LIBRARIES TO PERFORM EDA

Numpy

NumPy is a powerful Python library for efficient array manipulation and

numerical computing (Harris et al., 2022). Originally known as “Numeric,”

it evolved into NumPy with enhanced capabilities. Its optimized C

programming and extensive functions make it the standard for array

computation, supporting tasks like linear algebra and Fourier

transformations. With a vibrant community and easy accessibility, NumPy

is widely used in data analysis, machine learning, and more.

Pandas

Pandas is a Python library that efficiently manipulates diverse data,

collaborating seamlessly with other data science libraries like NumPy,

Matplotlib, SciPy, and scikit-learn. It plays a vital role in machine learning

by accurately exploring, cleaning, transforming, and visualizing data.

Jupyter notebook facilitates easy execution of Pandas programs, enabling

data visualization and analysis. Pandas expand Python's data analysis

capabilities with essential procedures: data loading, manipulation,

preparation, modeling, and analysis (Ateeq & Afzal, 2023).

SciPy

SciPy is a powerful Python library that builds upon NumPy, providing

multidimensional arrays for scientific and mathematical problem-solving

(Khandare et al., 2023). It eliminates the need for separate NumPy imports

and is widely used in Machine Learning, particularly for image processing.

With modules for optimization, linear algebra, parallel programming, and

integration, SciPy is favored by data scientists, analysts, and engineers,

enabling efficient numerical computations and faster data processing with

enhanced expressiveness compared to other tools and libraries.

Matplotlib

Matplotlib is a popular library that simplifies complex tasks and offers

advanced capabilities for statisticians and analysts (Lafuente et al., 2021). It

is widely used to create charts, such as bar graphs, histograms, and scatter

plots. Matplotlib is versatile and can be used by beginners, experts, and

professionals in Python scripts, shells, and Jupyter Notebooks. It also

enables the creation of 3D charts and facilitates animated and interactive

visualizations in Machine Learning.

Sci-Kit

Scikit-learn is a widely used Python library for machine learning, offering

mathematical and statistical algorithms and specialized toolkits like Scikit-

image (Ahmed et al., 2022). It is essential for Python-based machine

learning in the software industry, assisting with data preparation, analysis,

model development, and post-model analysis. Scikit-learn represents data as

tabular datasets, utilizing x and y variables for supervised learning and only

x variables for unsupervised learning, where x variables are independent,

and y variable is dependent, describing samples in qualitative or

quantitative manners.

Seaborn
Statistical analysis involves understanding relationships and their

dependencies on other variables in a dataset, aided by visualizations that

enable the human eye to detect patterns. Seaborn, a Python library built on

Matplotlib, provides an intuitive interface for creating aesthetically pleasing

and informative statistical graphs, making it easier to observe and interpret

correlations within the data (Bisong, 2019).

PYTHON FUNCTIONS FOR EDA

These functions run only if the panda's library is imported to the Python

coding platform (Mukhiya & Ahmed, 2020).

To obtain the needed information from our data:

Data.info()

This Python code helps obtain the attributes of our dataset or data frame (if

in CSV format), their total non-null elements, and their data types for each

attribute, field, or column name. Also, this .info() function in Python helps

identify the number of instances or rows present in our data for further

analysis.

data.describe()

This function in Python describes the data, which returns the descriptive

statistics values of our data frame such as count, minimum, and maximum

values out of all the elements under each attribute, mean, standard

deviation, 25 percentiles, 50 percentiles, and 75 percentiles.

Duplicate Values
data.duplicated().sum()

This function is crucial in exploratory data analysis, finding duplicate

values. When the values are duplicated, those can negatively impact our

predictions and analysis because these duplicated values occur due to

human error or misinterpretation. It is good when all the values are unique
in the data and there are null duplicate values. (However, sometimes,

duplicate values might occur based on the attribute type and data). Hence,

the .duplicated().sum() function returns the total number of duplicated

values in our data.

Representing Unique Values in the Data

data ['attribute_name']. unique()

The unique function in Python returns all the unique values for the specified

attribute of the data frame. This function not only returns the unique values

but also specifies the datatype of those unique values.

Unique Counts Visualization

To visualize the unique value counts for each data field, initially import the

seaborn library.

sns. countplot (data ['attribute_name']). unique()

The above function displays a simple normal graph plotted to visualize the

unique value counts for each field or column.

Capturing Null Values

Enhancing the quality of our data is very necessary. To check for the null

values present in any of the fields, the following function is used:

data. isnull (). sum()

This function returns the total number of null values under each attribute. If

there are no null values, the function returns zero. That does not mean the

attribute has one null value, but the attribute has no null values.

Null Values Replacement

Capturing the null values alone is not essential, but we must also take

measures to replace them.

data.replace (np. nan,' 0', inplace = True)

Here, nan refers to the null values. This function replaces each null value in

the whole data with zero. There are many optimized methods for replacing

null values, like using the mean or median of the entire column and

replacing it with the null value.

Attribute Datatypes

data. Dtypes

Use the above function to specifically know the data's datatype for all the

columns (attributes).

Data Filtration

data [ data ['attribute_name'] = = specify_value

]. head()

When the data is extensive, but the required data is on a small scale, data

filtration is one of the best choices. This function returns the values

belonging to the class of that particular attribute. So, the required data is

filtered from the actual data.

Plotting
data [ [‘attribute_name ] ]. type_of_plot()

This function displays the graph of the variables of the mentioned type.

Correlation

data. corr()
A table of the correlated values for each attribute is displayed in matrix

format under the function mentioned earlier. Correlation is a statistical

concept that estimates the degree to which two defined variables move

concerning each other. The correlation matrix values will usually lie

between -1 and +1, where -1 denotes a high negative correlation, and +1

denotes a high positive correlation. The methods mentioned earlier are the

fundamental and most important methods and functions in exploratory data

analysis.

UNIVARIATE GRAPHICAL METHOD

Bar charts are excellent for categorical data. Histogram is a bar graph where

each of the bars reflects the number of cases for a specific range of values.

They can therefore be used for categorical or continuous data. They aid in

comprehending the outliers, spread, and center trend. Another way to

examine univariate data is through box plots, which can be graphically

st rd
represented by 5 number summary of the fir quartile, median, thi

quartile, minimum and maximum. From quartile to quartile, a box is drawn,

with whiskers extending to the minimum and maximum values. Therefore,

they use reliable statistics like the median and interquartile range to provide

information about the central tendency, spread, symmetry, and outliers.

UNIVARIATE NON-GRAPHICAL METHOD

Univariate EDA involves analyzing sample distributions to understand

population characteristics, including outlier detection and examining

categorical and quantitative data. Measures such as mean, median, mode,

variance, standard deviation, and interquartile range are used to assess

central tendency, spread, and asymmetry. Tabulating frequency helps

analyze categorical variables, while summary statistics provide insights into

numerical data's center, spread, and outliers.

MULTIVARIATE NON-GRAPHICAL METHOD

Multivariate non-graphical EDA techniques employ cross-tabulation or

statistics to analyze relationships between variables. This includes

comparing statistics across levels of categorical variables and computing

covariances or correlations for numerical variables. Cross-tabulated reports

help understand counts or percentages of two categorical variables.

Correlation, ranging from -1 to +1, helps determine the strength and

direction of association between two quantitative variables, aiding feature

selection in machine learning (Chemmalar Selvi, & Lakshmi Priya, 2020).

MULTIVARIATE GRAPHICAL METHOD

Bar plots are commonly used to represent categorical multivariate values,

with each group representing a level of one variable and each bar indicating

levels of another variable. Conversely, scatterplots effectively visualize two

quantitative multivariate variables, with points representing data samples

and variables plotted on the x and y axes. Side-by-side box plots and bars

help display relationships between quantitative and categorical variables,

while scatter plots can accommodate multiple variables using different sizes

and shapes.

DATA VISUALISATION IN EDA

Exploratory data analysis (EDA) is crucial in understanding and

formulating questions for data science research (Shamim & Rihan, 2023).

Visual methods, such as various types of graphs, are emphasized in EDA to

evaluate and extract meaningful information from the data. Python libraries

like matplotlib and Seaborn provide tools for data visualization, enabling us

to comprehend and analyze the data through visual representations quickly.

Univariate Enumerative Plots

These lists of plots display every observation in the data and reveal how the

observations are distributed across a particular data variable. Now let us

examine various enumerative plots. Let us consider the “Iris dataset” for

understanding the types. The dataset columns are sepal width, length, and

petal width and length.

Univariate Scatter Plot

Univariate scatter plot helps plot various observations of the same variable

in correspondence to the index number of that data. A univariate scatter

diagram with the matplotlib's plt.scatter() method can also be plotted. To

plot, the scatter() function needs two parameters. Therefore, plotting the

“sepal.width” variable in this example contradicts matching the number of

observations kept as the data frame's index (df.index). Then, using the

seaborn library's sns.scatterplot() function, depict the same plot while

considering its diversity. Let us consider that we are plotting the 'sepal

length(cm)' variable:

Figure 1. Scatter plot

Line-Plots (Using Markers)

The data is visualized in a line plot using line segments to connect the data

points. It looks like a scatter plot, but the measurement points are joined by
straight-line segments and are organized differently (usually by their x-axis

value). The line plot is given in Figure 2.

Figure 2. Line plot

By default, the matplotlib plt.plot() method uses a line plot to plot the data.

Although matplotlib does not offer such an automated option, one can

utilize pandas' groupby() function to plot such a graph. The line plot can

also be seen using the sns.lineplot() function.

Strip-Plots

Both scatter and strip plots are comparable. Strip plots are frequently used

with a few other plot types for more accurate analysis. These plots display

the data points distribution of a specified variable. Use the function

sns.striplot () to plot a strip plot. The strip plot is given in Figure 3.

Figure 3. Strip plot

Figure 4. Swarm plot
Swarm-Plots

A way for visualizing univariate data to see the range of values in a

continuous variable is the swarm plot, which resembles a strip plot. The

sole difference between the swarm-automatic plot and the strip plot, which

produces a better visual representation of the data, is spreading the

variable's data points to prevent overlap. Use the sns.swarmplot() function

to plot a swarm-plot. An example of swam plot is presented in Figure 4.

Univariate Summary Plots

Compared with an enumerative plot, these plots give a more succinct

representation of a variable's dispersion, position, diversity, and

distribution. Though attaining every data value in a summary plot is

impractical, it effectively portrays all data to draw more accurate inferences

about the complete data set.

Histograms

Histograms resemble bar charts in appearance because they display counts

or relative frequencies of values falling within specific class intervals or

ranges. A histogram displays the shape and distribution of continuous

sample data. It also aids in our comprehension of the data's skewness and

kurtosis. In the matplotlib, plt.hist() function is used to plot a histogram.

Figure 5 depicts the standard representation of the histogram.

Figure 5. Histogram
Density Plots

The density plot mirrors a smooth histogram. The probability density

function of the variable is typically displayed in density plots using the

kernel density estimate. A continuous curve—the kernel—is created to

estimate the density of the entire data set smoothly. Plotting the density plot

for the 'petal.length' variable is given in Figure 6.

Figure 6. Density plot

Rug Plots

Rug plots are a straightforward yet excellent legal technique to depict a

distribution. Each data point has vertical lines in it. The height, in this case,

is arbitrary. The density of the tick marks can be used to determine the

distribution's density. The rug plot and histogram are directly related since a

histogram makes bins with data range, and then a bar is drawn with a height

equal to the ticks count in each bin. In the rug plot, each data point is

represented by a single tick mark or line on a single axis. The rug plot's

readability of the distribution reduces slightly compared to a marginal

histogram. Figure 7 illustrates the representation of the Rug plot.

Figure 7. Rug plot

Box Plots

A box plot is a beneficial and standard method when depicting the data

diversity by a 5-number description or summary (first quartile, second

quartile (median), third quartile, maximum, and minimum). It greatly aids

in identifying outliers and aids in comprehending these characteristics of

the data distribution. Figure 8 illustrates the representation of the Box Plot.

Figure 8. Box plot

Distplot

Earlier, the seaborn library's distplot() function was mentioned in the rug

plot portion. The seaborn kdeplot() and rugplot() functions are combined

with the matplotlib hist() function in this function. Figure 9 showcases the

representation of the Dist Plot.

Figure 9. Dist plot

Violin Plots

With a rotating kernel density plot on either side, the box plot resembles the

violin plot. In order to allow for comparison, it displays the diversity of data

(quantitative) across several levels with a single (or more) categorical

variable. The function plt.violinplot() is utilized, setting the Boolean

parameter “show medians” to “True” to mark the medians for each variable.

The violin plot aids in comprehending the estimated density of the variable.

Figure 10. Violin plot

Bar Charts

A 2-D axis representation of univariate data is a Bar Plot. The category

axis, which displays the category, is the first axis. The value axis, which

displays the numerical value of the category as indicated by the length of

the bar, is the second axis. A category variable is shown on the plot. Bar()

produces a bar graph. A series holding the counts of the variable's unique

values is returned by the function value counts().The bar plot is a two-

dimensional axis depiction of univariate data. The first axis is the category

axis, which displays the category, and the second axis is the value axis,

which displays the category's numerical value as represented by the length

of the bar.A categorical variable is plotted using the plot.bar() function as a

bar plot.

Figure 11. Bar chart

Pie Chart

The most popular method for representing the numerical percentages held

by each category is a pie-plot or pie-chart. To simulate a pie-chart, use the

plt.pie() function. Separate the parts equally in pie-chart because each

category will be spread equally. Passing values in an array to the 'labels'

option will then add the labels. The DataFrame.sample() function can be

used to generate a random sample. The division of items that are to be

returned in axis is indicated by frac argument of sample() function. The

pie() function spins literally everything anti-clockwise at a specified angle

using 'startangle' option. Additionally, startangle's default value is 0. One

can use Python string formatting to display the percentage value using the

'autopct' argument. Figure 12 illustrates the representation of the Pie Chart.

Figure 12. Pie chart

BIVARIATE ANALYSIS

When it's necessary to investigate the link between two distinct variables,

bivariate analysis is utilized (Dugard et al., 2022). In the end, this is what

needed to accomplish in order to develop a strong model. Multivariate

analysis is the process of combining more than two variables for analysis.

For both Bivariate and Multivariate Analysis, will work on several charts.

More variables are taken into account at once in multivariate analysis. Only

two variables at a time are considered and concentrate on bivariate analysis.

For visualization random datasets are considered.

Three categories of visual methods can be taken into account for bivariate

analysis:

• Numerical and Numerical

• Numerical and Categorical

• Categorical and Categorical

NUMERICAL AND NUMERICAL

In this instance, two number variables are compared. The scatter plot, which

shows the location where the two variables overlap, is the most typical

graph used to show how they relate (Cooksey, 2020).

Scatter Plot

Using a scatter graph or plot to show the association between two numerical

variables is easy. Let's use a scatter plot to visualise the relationship

between the overall cost and the tip given. Figure 13 illustrates the

representation of the scatter plot.

Figure 13. Scatter plot

NUMERICAL AND CATEGORICAL

There are several plots that can be used for bivariate and multivariate

analysis if one variable is numerical and the other is categorical (Kannan et

al., 2019).
Bar Plot

By showing a numerical variable on the y-axis and a category variable on

the x-axis, a bar plot is a simple graphic that may be used to investigate the

relationship between two variables. Figure 14 illustrates the general

representation of the Bar plot.

Figure 14. Bar plot

Boxplot

In the aforementioned Univariate analysis, boxplots is discussed in detail.

For each variable, a separate boxplot can be created. A boxplot to

investigate gender and age is given in Figure 15.

Figure 15. Box plot

Distplot

By applying kernel density estimation, Distplot explains the PDF function

(given in Figure 16). There is no hue option in Distplot, but can add one.

Assume if want to determine whose survival probability is high compared

to the age range of death ratio and want to see the chance of persons with an

age range that of survival probability. The graph is clearly quite intriguing,

as can see in Figure 16. The orange plot depicts the likelihood of survival,

whereas the blue one displays the likelihood of dying. If observed, can see

that children have a larger chance of surviving than dying, however it is the

contrary for elderly individuals. This short analysis occasionally reveals

important data insights and is helpful for crafting data narrative.

Figure 16. Dist plot

CATEGORICAL AND CATEGORICAL

There are several plots which can be used for bivariate and multivariate

analysis if one variable is categorical and the other is categorical (Denis,

2020; Adams et al., 2021).

Heatmap

Heatmap is a visual depiction of only that, if you've ever used the crosstab

function of pandas. It essentially displays how prevalent a particular

category is in relation to another category in the dataset. To determine how

many people survived and perished can use the heatmap. Figure 17 provides

a visual depiction of the Heat map in a general form.

Figure 17. Heat map

Cluster Map

To comprehend the link between two categorical variables, can also use a

cluster map. In essence, a cluster map is a dendrogram that groups

comparable behaviour categories together. Figure 18 visually represents a

cluster map. If one is familiar with clustering methods, specifically

DBSCAN, dendrograms are commonly used graphs in exploratory analysis.

Figure 18. Cluster map

CONCLUSION

Exploratory data analysis is useful if a data scientist needs to uncover fresh

information from a vast array of data sets. EDA can be advantageous in this

regard for industries like data science, engineering, and research and

development. As a result, current analytics are supported together with

access to powerful computing today. For academics or data scientists, using

EDA can be an exciting and interesting way to discover unanticipated value

in a large number of complex data sets. From this chapter the various

libraries used in EDA could be known and various visualization could also

be learnt. The steps in EDA could also be known step by step.

REFERENCES

Adams, B., Baller, D., Jonas, B., Joseph, A. C., & Cummiskey, K.

(2021). Computational skills for multivariable thinking in

introductory statistics. Journal of Statistics and Data Science

Education, 29(sup1), S123-S131.

AhmedT.AlamM. A.PaulR. R.HasanM. T.RabR. (2022, February).

Machine Learning and Deep Learning Techniques For Genre

Classification of Bangla Music. In 2022 International Conference

on Advancement in Electrical and Electronic Engineering

(ICAEEE) (pp. 1-6). IEEE.

doi:10.1109/ICAEEE54957.2022.9836434

Ateeq, M., & Afzal, M. K. (2023). 10 Programming Languages,

Tools, and Techniques. Data-Driven Intelligence in Wireless

Networks: Concepts, Solutions, and Applications, 237.

Bisong, E. (2019). Matplotlib and seaborn. Building Machine

Learning and Deep Learning Models on Google Cloud Platform: A

Comprehensive Guide for Beginners, 151-165.

Chemmalar Selvi, G., & Lakshmi Priya, G. G. (2020). An

Epidemic Analysis of COVID-19 using Exploratory Data Analysis

Approach. Predictive Analytics Using Statistics and Big Data:

Concepts and Modeling, 99.

Cooksey, R. W. (2020). Descriptive statistics for summarising data.

Illustrating statistical procedures: Finding meaning in quantitative

data, 61-139.

Denis, D. J. (2020). Univariate, bivariate, and multivariate statistics

using R: quantitative tools for data analysis and data science. John

Wiley & Sons. doi:10.1002/9781119549963

Dugard, P., Todman, J., & Staines, H. (2022). Approaching

multivariate analysis: A practical introduction. Taylor & Francis.

doi:10.4324/9781003343097
Harris, C. R., Millman, K. J., Van Der Walt, S. J., Gommers, R.,

Virtanen, P., Cournapeau, D., Wieser, E., Taylor, J., Berg, S.,

Smith, N. J., Kern, R., Picus, M., Hoyer, S., van Kerkwijk, M. H.,

Brett, M., Haldane, A., del Río, J. F., Wiebe, M., Peterson, P., &

Oliphant, T. E. (2020). Array programming with NumPy. Nature,

585(7825), 357–362. doi:10.1038/s41586-020-2649-2

Kannan, S., Dongare, P. A., Garg, R., & Harsoor, S. S. (2019).

Describing and displaying numerical and categorical data. Airway,

2(2), 64. doi:10.4103/ARWY.ARWY_24_19

Khandare, A., Agarwal, N., Bodhankar, A., Kulkarni, A., & Mane,

I. (2023). Analysis of Python Libraries for Artificial Intelligence. In

Intelligent Computing and Networking. Proceedings of IC-ICN,

2022, 157–177.

Lafuente, D., Cohen, B., Fiorini, G., García, A. A., Bringas, M.,

Morzan, E., & Onna, D. (2021). A Gentle introduction to machine

learning for chemists: An undergraduate workshop using python

notebooks for visualization, data processing, analysis, and

modeling. Journal of Chemical Education, 98(9), 2892–2898.

doi:10.1021/acs.jchemed.1c00142

Mukhiya, S. K., & Ahmed, U. (2020). Hands-On Exploratory Data

Analysis with Python: Perform EDA techniques to understand,

summarize, and investigate your data. Packt Publishing Ltd.

Shamim, G., & Rihan, M. (2023). Exploratory Data Analytics and

PCA-Based Dimensionality Reduction for Improvement in Smart

Meter Data Clustering. Journal of the Institution of Electronics and

Telecommunication Engineers, 1–10.

doi:10.1080/03772063.2023.2218317
 CHAPTER 7

Empowering Scientific Computing and Data

Manipulation With Numerical Python (NumPy)
Tesfaye Fufa Gedefa

https://orcid.org/0000-0001-5405-5320

Space Science and Geospatial Institute, Ethiopia

Galety Mohammed Gouse

https://orcid.org/0000-0003-1666-2001

Samarkand International University of Technology, Uzbekistan

Garamu Tilahun Iticha

Debre Markos University, Ethiopia

ABSTRACT

NumPy is a Python library for performing numerical data structures and

specialized computing. It improves the components of N-dimensional arrays

and provides operations and tools to interface with these arrays. NumPy

implements the N-dimensional array, or ndarray, and provides Python-

specific scientific methods for performing realistic array and matrix

operations. When compared to array programming in other languages, it

will allow us to do a wide range of mathematical operations and data

manipulations. NumPy can be used with other Python packages and

programming languages such as C and C++. NumPy now supports object-

oriented programming as well. For example, a class called ndarray may be

an N-dimensional array with multiple ways of performing various data

structure operations and characteristics.

INTRODUCTION
NumPy is an essential package for systematic computing in Python. The

library gives an N-dimensional array question and different inferred objects

(such as masked arrays and matrices) (NumPy Developers, 2008-2022). It is

a mixture of sequences for fast procedures on arrays, with mathematical,

logical, and shape manipulation; sorting; selecting; input/output; discrete

Fourier transforms; introductory linear algebra; basic statistical techniques;

arbitrary simulation; and more (Johansson, 2019). NumPy delivers the

numerical backend for nearly every scientific or technical library for

Python. It provides an N-dimensional array object, a robust data structure

that has become Python's standard numerical data representation (Millman,

2011). It is, therefore, a vital part of the scientific Python ecosystem

(Johansson, 2019). Its goal is to create the cornerstone of a suitable

environment for scientific computing. As stated (Agathiya Raja, 2022), the

Python library plays a significant role in different application areas for

analyzing complex networks, data science, and big data; analyzing and

visualizing the networks using Python offers good insights about the

networks to end-users using the python library. To better understand the

people surrounding NumPy, at the core of the NumPy package is the

ndarray object (Alammar, 2022).

This encapsulates n-dimensional arrays of the same data types, with many

operations performed in the accumulated code for performance. NumPy is

an essential bundle for scientific computing in Python, and it provides an n-

dimensional array object and numerous derived things (such as screened

arrays and matrices). The ndarray object is at the heart of the NumPy

bundle. However, NumPy arrays and standard Python sequences differ in

several ways (Johansson, 2019; Albenis Pérez-Alarcón, 2021).

Python extends Python with robust mathematical operations that provide

practical guarantees in calculations with arrays and matrices. NumPy stores

a massive library of high-level mathematical functions that work on these

arrays and matrices. NumPy aims to deliver an array object faster than

traditional Python lists. As a result, Python is the fastest to read fewer bytes

of memory or has contiguous memory. It assists with different operations

like slicing, indexing, and broadcasting; matrix computational functions;

and has the benefits of single instruction multiple data vector processes
(SIMD). In addition, NumPy completely supports an object-oriented

approach, and it is an N-dimensional array named ndarray, a class

possessing various techniques and qualities. Functions mirror several

methods in the outermost NumPy namespace. This lets the languages code

in whatever makes the most sense.

Installing Numpy

This chapter strongly recommends using a technical Python source to install

NumPy. In this chapter, looking for the complete commands to mount

NumPy on the operating system, see Connecting NumPy and if there is

already installed Python, easily connect NumPy with the following:

pip install NumPy, or > conda install NumPy

Importing Numpy

Importing Modules: To use the NumPy library in this chapter, import it into

the program. The NumPy module is often imported using the alias np.

NumPy should be imported as np.

import NumPy as np

After this, this chapter can access functions and classes in the NumPy

module using the np namespace. It shortens the imported name to np for

better readability (alias) of the code using NumPy. This is a widely adopted

convention, so anyone working with code can easily understand it.

NUMPY DATATYPES

The most common way to manipulate values in NumPy is through ndarray

objects (Johansson, 2019). Because ndarrays can hold any number of

quantities, they can represent any data type, such as scalars, vectors,

matrices, or tensors. They are similar to Python lists but can have multiple

dimensions. The basic numeric information types saved by NumPy are

listed in Figure 1. Non-numerical data sortings, such as text, objects, and

custom compound sorts, are also accessible. Datasets for Basic Numeric

Data that may be accessed using NumPy Scalar: A scalar is a single integer

with zero dimension, and NumPy has one. As a result, instead of Python int,

NumPy data types utilize collations such as uint8, int8, uint16, and int16.

Strings, objects, and complicated data types the user creates are also

supported.

Figure 1. Basic numerical data types NumPy

Source: NumPy Developers (2008-2022)

Scalar

Scalar and NumPy are terms for single numbers with a dimension of zero.

As a result, this book chapter may pass the value to NumPy's array function

to generate a NumPy array that holds a scalar instead of Python's int types,

such as uint8, int8, uint16, and int16. It returns an empty parenthesis to

indicate that it has no dimensions (Mulani, 2022)

Vectors

It is a one-dimensional list of numbers. To create a vector, pass a Python list

to the array function (Mulani, 2022).

Matrices

In Numpy, matrices are represented as lists of lists or 2D vectors. Like

vectors, this chapter generates matrices using NumPy's array function.

However, one must provide a list where each A list denotes a row rather

than merely passing through a single list. Consequently, to make a three-by-

three matrix using digits 1 through 9 (McKinney, 2012).

Tensors

N-dimensional ndarrays are tensors. They can include additional

dimensions similar to vectors (Alammar, 2022).

NUMPY ARRAYS

A data structure that effectively stores and accesses multidimensional arrays

is the NumPy array. A pointer to memory is included in Numpy, along with

metadata used to convert the data stored there—particularly information

sorting, shaping, and steps. The data sort shows the kinds of parts stored in

an array. The information sorting process depicts the personalities of

components organized into an array. A data sort is included in an array, and

the memory requirements for each cluster element are the same (Harris, 16

September 2020). In Python, NumPy arrays are the standard way to store

numerical data and allow numerical operations to be done in a high-level

language. As stated by (Oliphant, 2006), an array is a form of data that

stores multiple values using just two identifiers.

An array is a collection of data pieces that can be accessed by their

directories and are organized sequentially. Lists of numerical data, vectors,

and matrices are stored in NumPy arrays. The NumPy library has numerous

procedures (built-in functions) for creating, modifying, and modifying

NumPy arrays. Even though the Python programming language has an array

data structure, it is less flexible, practical, and powerful than the NumPy

array. The NumPy Connect memory location: Each position is only

assigned to one piece of data, and the memory space needs to be partitioned

into precisely sized areas (Stefan van der Walt, 2021). Contiguous memory

allocation is now available: To ensure that memory space can be used, split

the data into pieces and put them in different places.

Figure 2. The NumPy array incorporates various fundamental

features of the Array

Source: Harris (2020)

Figure 2 shows that the metadata areas are connected to the NumPy array

data structure. Ordering an array using cuts is what b is about. These forms

give back a summary of the initial data. c. arrange an array using masks,

scalar aids, or other clusters in a way that duplicates the initial data. An

array is ordered with different arrays in the bottom example, broadcasting

the ordering disputes before the lookup of the original data. The bottom

example discloses the indexing arguments before the lookup by indexing an

array with several arrays. d. The use of vectorization allows operations to be

applied to groups of elements. e.As two-dimensional arrays increase, so

does broadcasting. f, acting in conjunction with one or more tomahawks

while reducing operations. In this scenario, an array is added to a few

tomahawks to produce a vector continuously or two tomahawks to produce

a scalar. Using two consecutive axes to construct a scalar is explained in g,

an illustration of NumPy code. An example, the NumPy code in g,

demonstrates some of these ideas.

THE DIFFERENCE BETWEEN A NUMPY ARRAY AND A

LIST
There are three benefits to using a Python NumPy array instead of a list.

Less memory, faster, and more convenient. The Python NumPy array should

be the first option because it uses less memory than a list. Then, it is

relatively quick to implement and works well with NumPy simultaneously.

Figure 3. The difference between a NumPy array and a list

Source: Alammar (2022)

In Figure 3, the list and Numpy array store data in memory. However, there

is a significant difference. Performing these operations on a list can be

tedious and time-consuming when searching, updating, and manipulating

data from memory. On the other hand, using a NumPy array is

advantageous as it allows for efficient data manipulation directly from

memory due to its sequential data storage.

Difference between List and ArrayTable

• Changeability: An array can keep the features of several different

data types, but it can only store the basics of one data type.

• Data Types Storage: The list can store various data, while the Array

provides equivalent data values.

• Importing Packages: Bringing in parcels No container is necessary

before utilizing a list because the Python programming language is an

inbuilt data structure. The Array, however, is not a feature of the

NumPy language. Therefore, we must ingest the “collection”

component before creating and utilizing arrays.

 • Arithmetical Process: The NumPy module allows an array

assembly to save data values and modify them quickly, which makes

using an array advantageous during mathematical operations in

Python. On the other hand, the outcomes are not shown in the list. Due

to this, math operations with the list are still possible but less practical

than those with the collection.

• Alter Competences: The list, on the other hand, is an in-built data

construction and hence can be resized and adjusted very rapidly and

efficiently, whereas the Array consumes very little memory, changing

the memory position.

ARRAY OPERATIONS

When an array is declared, we can access its elements or carry out

particular operations. The last session taught us how to access items. The

various functions that can be used with arrays are covered in this section.

Arithmetic Operations

Mathematical operations on NumPy arrays are quick and straightforward.

An essential arithmetic operation, such as addition, subtraction,

multiplication, division, etc., is performed on each pair of components when

applied to two collections. For example, combine two arrays, the first

element from the first Array will be added to the first element of the second

Array, and so on. Consider the subsequent element-wise operations on two

arrays:

Array Transpose

To reverse the tensor swap using the keyword argument, use transpose (an

argsort (axes)). Transposing a one-dimensional array produces an

unchanged copy of the original Array.

Array Sorting
This involves pushing components in logical order. Any sequence with an

element-based order, such as numeric or alphabetical, ascending or

descending, is considered well-ordered. For example, a function called sort

() on the NumPy n-dimensional array element sorts an array.

Concatenating Arrays and Splitting

Every code that came before it only operated on individual arrays. On the

other hand, it is also possible to join many arrays into a single array, and it

is also possible to break a single array into several arrays. The

concatenation process refers to joining two or more arrays together. When

arrays are concatenated, sequences are added one after the other in

sequential order.

Array Reshaping

Reshaping entails changing an array's shape. The number of necessary

elements in each length determines the form of an array. By reshaping, we

can change the size of the Array, the number of elements in each, or both.

Figure 4. Data reshaping

Source: Java Tutorial Developers (2022)

As shown in Figure 4, reshaping the Array with the help of NumPy to

change the dimension as we want using rows and columns by changing the

size of the Array.

Slicing and Indexing

This approach makes it possible to choose any item in an array depending

on its N-dimensional index, allowing for more flexibility in the selection

process. Every integer array has a representation of the total number of

indexes for its respective dimension. When the index is composed of the

same number of integer arrays as the dimensions of the ndarray being

targeted, completing the operation becomes quite simple. NumPy uses the

Python language's sophisticated indexing (McKinney, 2012) capability.

Python's indexing features were initially developed in response to the

demands of users who worked with numbers. Slicing is removing a chosen

group of components from an array. This division process is comparable to

the one used in the list. Think about the following:

Figure 5. Slicing a NumPy array

Source: Oliphant (2006)

As described in Figure 5, the slicing of an array in NumPy is the way of

removing a chosen group of components from a given an array.

Broadcasting

The ability of NumPy to handle arrays with varying dimensions

appropriately while performing arithmetic operations is referred to as

broadcasting. Arithmetic operations are often performed on items

corresponding to one another when working with arrays. These procedures

may be carried out without a hitch so long as the two arrays being worked

with have an identical structure. Vital processes on NumPy arrays are

element-based. This works on arrays of the same degree. In any case, it is

imaginable to perform actions on arrays of various sizes if NumPy can

change them; this change is called broadcasting. Figure 6 gives an example

of broadcasting.

Figure 6. Numpy array broadcasting

Source: McKinney (2012)

In any case, it is also imaginable to perform actions on arrays of various

sizes on the off chance that NumPy can change them/

SPECIAL FUNCTION OF NUMPY PYTHON

Numerous numerical capabilities in NumPy allow for a variety of numerical

operations. Trigonometric aptitudes, arithmetic skills, and the capacity for

working with complex numbers are all examples of numerical aptitudes

(NumPy Developers, 2008-2022). Let us talk about the mathematical

functions first, in any case.

Trigonometric Functions

Figure 7. Trigonometric functions

Source: Mulani (2022)

Figure 7 shows that the trigonometric formulas for computing the sine,

cosine, and tangents of various angles in radians are available in Numpy.

The sin, cos, and tan functions return the trigonometric ratio for the

supplied angles.

Rounding Functions

NumPy offers a variety of routines for truncating the value of a decimal

number tuned to the accuracy of decimal digits. First, let us discuss the

adjusting features—the numpy. The around () method works as intended.

This task returns the corrected decimal value at the desired decimal

location. Figure 8 contains the work's syntax.

Figure 8. Numpy rounding functions

Statistical Functions in NumPy

Figure 9. Numpy statistical functions for maximum and minimum

Source: Millman (2011)

As shown in Figure 9, Numerous statistical functions offered by Numpy can

be utilized to analyze statistical data. For instance, identifying the Array's

most minor and extreme elements using NumPy.amin() and NumPy.amax()

can find each array element's minimum and maximum values along the

required axis.
NumPy Sorting and Searching

Numerous sorting and searching functions are available in Numpy. The sort

() method is used to build several sorting algorithms, including quicksort,

merge sort, and heapsort. The function call must specify the sorting

algorithm utilized in the sort operation. Let us talk about the sorting

algorithm used by NumPy. Sort () is given in Figure 10.

Figure 10. NumPy sorting and search algorithms

Linear Algebra in NumPy

NumPy gives various capabilities to perform multiple algebraic calculations

on the input data. Numpy linear algebra is shown in Figure 11.

Figure 11. NumPy linear algebra

Numpy Copies and Views

While executing the functions, some return a copy of the input array, while

some return the view. When the contents are physically stored in another

location, it is called Copy. If, on the other hand, a different view of the same

memory content is provided, we call it a View (NumPy Developers, 2008-

2022). A replica of the array object is not created by simply assigning

values to its elements. Instead, it accesses it using the same id () used for the

first Array. The id () method in Python is analogous to the pointer function

in C in that it returns an object's universal identification. In addition, any

changes that occur in one are immediately reflected in the other. For

example, if one changes its form, this will cause the other to alter its shape

as well.

CONCLUSION

Most of this chapter is devoted to examining the importance of NumPy and

the applications that may use its various packages. When using Python for

technical and scientific computing, grasping the NumPy package ecosystem

is essential. How the packages are employed for computing is necessary to

get the NumPy package ecosystem and how they are employed. This is

because of the nature of the computing tasks that Python is used for. It was

underlined that matrices, vectors, ndarray, tensors, reshaping array

expressions, and arithmetic operators should all be employed to perform a

realistic array analysis. This chapter briefly introduces the NumPy library,

emphasizing its mathematical operations and benefits when dealing with

data structures. NumPy is a library that is important for doing computations

in Python, and it has the potential to serve as the basis for all other Python

computational libraries. The Python Scientific Community first developed

NumPy. The ndarray, which is NumPy's data structure for N-dimensional

arrays, was the topic of our discussion at the outset of this subsection. Then,

we looked at how the ndarray may be used to generate and manage arrays,

count orders, solve issues with data structures, carry out mathematical

calculations, design applications for data science, and clip off sections of

arrays.

NumPy can be used with other Python packages and programming

languages such as C and C++. In addition, NumPy now supports object-

oriented programming as well. For example, a class called ndarray may be

an N-dimensional array with multiple ways of performing various data

structure operations and characteristics. In conclusion, NumPy gives users

access to various functions that can execute basic mathematical operations.

Users can design applications based on data structures, generate arrays with

N dimensions with the aid of this package, and adjust the geometry of

existing arrays. Users must consider every detail seriously to get any use out

of computers.

REFERENCES

Agathiya Raja, G. K. (2022). Introduction To Python for Social

Network Analysis. Wiley Online Library.

Alammar, J. (2022). NumPy: the absolute basics for beginners.

Retrieved 10 25, 25, from http://jalammar.github.io/

Albenis Pérez-Alarcón, J. C.-A. (2021). Alarconpy: a Python

Package for Meteorologists.

Harris, C. M. (2020). Array Programming with NumPy. Nature,

357–362.

Java Tutorial Developers. (2022). Retrieved from NumPy Tutorial:

https://www.javatpoint.com/

Johansson, R. (2019). 1. Scientific Computing and Data Science

Applications with Numpy, SciPy, and Matplotlib (2nd ed.).

Urayasu-shi.

McKinney, W. (2012). NumPy Basics:Arrays and Vectorized

Computation. In W. McKinney, Python for Data Analysis. O'Reilly

Media, Inc.

Millman, K. J., & Aivazis, M. (2011). Python for Scientists and

Engineers. Computing in Science & Engineering, 13(2), 9–12.

doi:10.1109/MCSE.2011.36

Mulani, S. (2022). Python Advanced. Retrieved from Vectors in

Python:

https://www.digitalocean.com/community/tutorials/vectors-in-

python

NumPy Developers. (2008-2022). Retrieved from Numpy User

Guide: https://numpy.org/doc/stable/user/whatisnumpy.html

Oliphant, T. E. (2006). Guide to NumPy. Academic Press.

Stefan van der Walt, S. C. (2021). The NumPy Array: A Structure

for Efficient Numerical Computation. IEEE, 13(2), 22-30.

 CHAPTER 8

Exploring Python's Powerful Data Collections

Mastan Vali Shaik

University of Technology and Applied Sciences, Ibri, Oman

Rajasrkaran Selvaraju

University of Technology and Applied Sciences, Ibri, Oman

ABSTRACT

One of the powerful data types for programmers that a Python supports is a

list. A list is a mutable data structure that holds an ordered collection of

items. Lists are useful to work with related data. A tuple is an immutable

data type that stores values in a sequence. Slicing can be used to get a range

of items. In Python, a set is variable that can store a variety of data, and

different data types perform different operations. A set data type in Python

is used to store multiple values in a single variable. In Python, dictionary is

one of the built-in data types where elements are key: value pairs. In other

programming languages, these are called associative memory or associative

arrays. Dictionaries are faster in execution, provide easy lookups, and are

implemented as hash tables. This chapter focuses on data collection in

Python.

INTRODUCTION TO LIST

A list is a sequence of values and can be of any type. A list is a data

structure that holds an ordered collection of items i.e., and you can store a

sequence of items in a list. The values in the list are called elements or

items. List values are accessed by using an index. The index value can be

positive from 0 or negative from -1. Once you have created a list, you can

add, remove or search for items in the list. Since we can add and remove
items, we say that a list is a mutable data type, i.e., this type can be altered

(Learn to code for IoT with python essentials course, 2022). Figure 1

illustrates the list representation, while Table 1 provides information on

accessing list values.

Example
List1 = [10, 20,15.7, 30, 40.2]
List2 = ['Chocolate', 'Juice', 'Burger', ’Pepsi’]
List3 = [] #empty list

Accessing List Values

num = [ 2, 4, 1, 10, 5]

Figure 1. List representation

Table 1. Accessing list values

print(num[0]) → 2 print(num[-1]) → 5

print(num[2]) → 1 print(num[-3]) → 1
print(num[4]) → 5 print(num[-5]) → 2

print(num[5]) → IndexError print(num[-6]) → IndexError

Syntax: list_name[index value]

Slicing a List

The slicing operation takes a specific range of elements from the list

(Moruzzi, 2020). Slice operation is performed on Lists with the use of a

colon.

Syntax
List_name[ start: end: step ]

Start is the starting index, End is the ending index (end value will be always

the previous value of end) Step is the increment or decrement value for

index. Table 2 outlines the slice operation.

Example

Consider the list with the name myList = ['a','b','c','d','e','f']

Table 2. Slice operation

List Slicing and Output Description

myList[0: 3] = ['a', 'b', 'c'] Elements from index 0 to index 2

myList[ 3: ] = ['d', 'e', 'f'] Elements from index 3 till the end

myList[: 4 ] = ['a', 'b', 'c', 'd'] starting index(0) to index 3

Starting index 1, last index 4 and step

myList[1: 5: 2] = ['b', 'd']
by 2
myList[: ] = ['a', 'b', 'c', 'd', 'e', From first element to the last element

'f']

myList[:: 2] = ['a', 'c', 'e'] From start to last, step by 2

myList[:: -1] = ['f', 'e', 'd', 'c', 'b',

From the last element to the first one
'a']

Input to a List

Quite often, we need to give some input to a list. An input() can be used to

read values to a List.

The following code inputs four values to a list and print.

Example
newlist=[] #Empty list
for idx in range(4):
val = int(input('Enter list value '))
#Take input
newlist.append(val) #append
value to list
print(newlist) #Print all
list values

NESTED LIST

A List can have another list. The following code illustrates nested list and

printing values.

Example
list_one = [10, 20, 12, [5, 9, 4], 17]
print(list_one)
print(list_one[3])
print(list_one[3][1])
Output

print(list_one) prints all values of a list

print(list_one[3])prints [5,9,4]
print(list_one[3][1]) prints 9

LIST METHODS

Table 3 provides the list methods in Python (Bisong, 2019).

Table 3. List methods

Method Description Example

add one item at the end

append(val) myList.append('M')
of the list

add sequence of items

extend(seq) myList.extend('N',’T’)
at end

insert one item ‘val’ at

insert(idx,val) myList.pop(3,’z’)
index idx

remove first occurrence

remove(val) myList.remove('Z')
with value val

remove & return item at

pop(idx) Res = myList.pop(4)
index idx (default last)

sort and store in

sort() myList.sort()
ascending order

sort and store in

sort(reverse=True) myList.sort(reverse=True)
descending

len(list_name) return items count siz = len(myList)

Exercise

Consider the following list declaration and write Python statements.

mylist = [10, 2, 15, 17, 25, 8, 11]

Table 4 presents the exercise for Lists, including the associated questions.

Table 4. List exercise

Question Answer

Print all values of list

nd th
Print 2 to 5 Value

Find length of list

Find maximum value

th nd
Print 6 to 2 value from reverse

Print all values in reverse order

LIST AND FUNCTIONS

We can have a user-defined function that takes the list as an argument and

perform some operation. For example, the following code Initializes a list

with some values. Write a user-defined function to find a number of positive

numbers in that list (Kong et al., 2020).

Example

#List And User Defined Function

def mytest(mylist):
pos=0
 n = len(mylist)
for idx in range(n):
if mylist[idx] > 0:
pos = pos+1
return pos
newList = [10, 2, -5, 17, -11, 8]
result = mytest(newList)
print(' Number of positive numbers in the list are
', result)

Practical Exercises

1. Read five values to the list and print the square root of each value.

2. Read seven integer values to the list and count number even

numbers.

3. Read six integer values to the list and count the number of positive

numbers.

4. Read seven integer values. Find average. Find and print the number

of values below the average value.

INTRODUCTION TO TUPLE

A Tuple data structure is used when we have data that will not change over

some time. A tuple is an ordered sequence of values separated by a comma.

Tuples are helpful for group-related information such as student name,

level, specialization, and grade. A tuple is heterogeneous Tuples are

immutable. Once created can’t be modified. Whereas a list is mutable. Since

it is immutable, it faster execution. Tuple is represented in an optional

parenthesis ‘(‘ and ‘)’. The three features of the tuple are immutable,

indexed, and heterogeneous. Tuple can have duplicate values (Guttag, 2021;

Johansson & John, 2019).

Creating and Printing a Tuple

Example
#working with tuples
student = ('76J1235',”Ahmed”, 'Bachelor',3.2)
print(“Student ID : “, student[0])
print(“Student name : “, student[1])
print(“Student level : “, student[2])
print(“Student grade score : “, student[3])
print(“Length of tuple = “, len(student))

Output:

The above code prints the following output

Student ID : 76J1235
Student name : Ahmed
Student level : Bachelor
Student grade score : 3.2
Length of tuple = 4

Figure 2. Tuple representation

In this example, student is the name of the tuple. The above example creates

a tuple with four values. Tuple values are accessed using index student[0]

refers to ‘76J1235’ and student[3] refers to 3.2. Positive index starts from 0,

whereas negative indexing starts from -1, shown in Figure 2.

A for could be used to access tuple values as shown below.

Example
student = ('76J1235',”Ahmed”, 'Bachelor',3.2)
for idx in student:
 print (idx)

TUPLE SLICING

Slicing operator:: is used to retrieve some elements of a tuple (Beri, 2019;

Arulkumar, 2020).

Syntax: tuple_name(start: stop:

step)

Example:

student = ('76J1235',”Ahmed”, 'Bachelor',3.2)

Table 5 provides examples of tuple slicing, including the input, output, and

description for each example.

Table 5. Slicing a tuple

Slice Output Description

student[1] Ahmed Second value of the tuple

Using negative index. -1 represents last

student[-1] 3.2
value of the tuple.

Starting from index 1 (Second value of

student[1::2] ('Ahmed', 3.2) the tuple), move until end of tuple with

step 2.

(3.2, 'Bachelor',

student[::-1] 'Ahmed', Tuples values in reverse order

'76J1235')

student[0::3] ('76J1235', 3.2) Starting from index 0 (Second value of

the tuple), move until end of tuple with

 step 3.

('Bachelor',
student[-2::-2] Starting index -2, step 2
'76J1235')

TUPLE PACKING AND UNPACKING

Tuple packing refers to assigning multiple values to a tuple. Tuple

unpacking refers to assigning a tuple to various variables (Unpingco, 2021).

Example:

#Tuple packing
student = ('76J1235',”Ahmed”, 'Bachelor',3.2)
#Tuple unpacking
(sid, name, level, grade) = student
print('student name is: ', name)
print('student grade is: ', grade)

Tuple Examples

mark1 = (7,8,6,17,42)
print(“maximum value in tuple is “, max(mark1))
print(“Minimum value in tuple is “, min(mark1))
print(“Length of tuple is “, len(mark1))
mark2 = sorted(mark1)
print(“Before sorting: “,mark1, ” After sorting:”,
mark2)
print(type(mark1), type(mark2))
name1 = (“Ahmed”, “Rashid”, “Omar”)
name2 = (“ahmed”, “rashid”, “omar”)
print(“NAME1 = “, name1, ” NAME2 = “, name2)
print(name1 == name2)
print(name1 < name2)
print(name1 > name2)
NESTED TUPLES

A tuple can be defined inside another tuple.

Example

emp = (('E101','Rashid', 3000), ('E105',

'Mohammed', 2900), ('E104','Omar',2810))
print(“Emp ID \t Emp Name \t Emp Salary”)
for k in range(len(emp)):
print(emp[k][0],'\t', emp[k][1],'\t', emp[k]
[2])

The above course code generates the following output.

Emp ID Emp Name Emp Salary

E101 Rashid 3000

E105 Mohammed 2900

E104 Omar 2810

COUNT AND INDEX METHODS OF A TUPLE

count() returns the number of times a specified value is present. index ()

Searches the tuple for a specified value and returns the position of the value.

index () generates ValueError if the value is absent in the tuple (Hill

Christian, 2020).

Example

#initialize a tuple
tp = (2, 8, 5, 8, 7, 3, 6, 5, 1)
print(“count(8) = “, tp.count(8))
print(“count(3) = “, tp.count(3))
print(“count(4) = “, tp.count(4))
print(“index(7) = “, tp.index(7))
#Value to search, starting index, End index
print(“index(5) = “, tp.index(5,0,4))

The above code generates the following output:

count(8) = 2
count(3) = 1
count(4) = 0
index(7) = 4
index(5) = 2

INTRODUCTION TO SET

The word set is predominately used in set theory of mathematics

(Haslwanter, Thomas, 2016). The implementation of set theory can

efficiently carry through set data type in Python. A set is represented in

curly braces{ }. Basically, set has the following properties:

1. A set is an unordered collection of data (The order of the set changes

every time it is used)

2. set is an unindexed data type (Set items can appear in a different

order every time you use them and cannot be referred to by index or

key)

3. set doesn’t have duplicate values (If items appear multiple times,

only one will be recognized in the set)

4. set values are unchanged (we cannot change the items after the set

has been created) set items are unchangeable, but you can remove and

add new items.

INITIALIZING AND PRINTING A SET

Example
#initializing and printing a set
myset = {10, 8, 12, 7, 11}
myset2 = {'Oman', 'Iraq', 'USA', 'India'}
print(myset)
print(myset2)

The above code prints the following output

{7, 8, 10, 11, 12}

{'India', 'Iraq', 'USA', 'Oman'}

Example:

myset2 = {'Iraq', 'Oman', 'Iraq', 'USA', 'India',

'iraq', 'China'}
print(myset2)

The above code prints the following output:

{'iraq', 'USA', 'Iraq', 'China', 'Oman',

'India'}

The initialization statement myset2 has multiple entries for ‘Iraq’. The

property of set is ‘No duplicate values’. However, ‘Iraq’ and ‘iraq’ are

different.

A for loop can also be used to print set values.

for inp in myset2:

print (inp)

SET METHODS

Table 6 provides an overview of the set methods along with their

descriptions and an illustrative example.

Table 6. SET methods

Method Description Example

add() Add a value to the set. myset2.add('Dubai')

discard() Remove a value from given set myset2.discard('iraq')

pop() Removes an element from the set myset2.pop()

clear() Removes all elements from the set myset2.clear()

Removes a specified element from

remove() myset2.remove('USA')
the set

Update the set with another set, or

update() myset2.update(myset3)
any other iterable

Example: Union, Intersection, Difference

Below is an example demonstrating the concepts of Union, Intersection, and

Difference.

#set operations (union, intersection, difference)

itcourses =
{'html','python','office','java','php'}
buscourses = {'office','tally','html'}
course1 = itcourses.union(buscourses)
course2 = itcourses.intersection(buscourses)
course3 = itcourses.difference(buscourses)
course4 =
itcourses.symmetric_difference(buscourses)
print(course1)
print(course2)
print(course3)
print(course4)
print(itcourses.isdisjoint(buscourses))
print(itcourses.issubset(buscourses))
print(itcourses.issuperset(buscourses))

The above code generates the following output:

{'office', 'php', 'java', 'python', 'html',

'tally'}
{'html', 'office'}
{'php', 'java', 'python'}
{'java', 'tally', 'php', 'python'}
False
False
False

Let’s have a look on above methods. Table 7 presents a variety of set

methods along with their descriptions and examples.

Table 7. SET methods

Method Description Example

Return a set.
course1 =
union Combination (union) of
itcourses.union(buscourses)
two given sets

Returns a set, that is the

course2 =
intersection intersection of two or
itcourses.intersection(buscourses)
more sets

Returns a set. Contains

course3 =
difference difference between two
itcourses.difference(buscourses)
sets

isdisjoint Returns Boolean Value. itcourses.isdisjoint(buscourses)

Checks whether two

 sets have intersection or

not.

Returns Boolean Value.

issubset Checks whether set-1 is itcourses.issubset(buscourses)

subset of set-2

Returns Boolean Value.

issuperset Checks whether set-1 is itcourses.issuperset(buscourses)

superset of set-2

Example: Test and Analyze the Following Code

#Test the following code

st = {'Apple', 'Banana'}
while True:
print(“ 1. Add to set \n 2. Print Set \n 3.
Delete from set \n Other to Exit”)
ch = int(input(“ Select one choice”))
if ch==1:
v = input('Enter fruit name to insert ')
st.add(v)
elif ch==2:
print(' **** Fruit Set has ....')
print(st)
elif ch==3:
a = input(' Enter fruit name to delete ')
if a in st:
st.discard(a)
print(' **** delete success ')
else:
print(“ *** No fruit available in
set”)
else:
print(' ****** Exit choice ')
break
INTRODUCTION TO DICTIONARY

A dictionary is a collection which is ordered, changeable and does not allow

duplicates. It is a Key: Value pair (Martelli et al., 2023) (Neelakandan,

2022). The following are some of the properties of a dictionary.

• Keys are immutable, case sensitive and unique.

• Dictionary values are accessed by using key.

• Dictionaries are used to store data values in key: value pairs.

• Dictionary items are ordered, changeable, and does not allow

duplicates. Dictionary can be referred to by using the key name.

• Dictionary is created using { }

Example:

strength = {
“java” : 25,
“python”: 24,
“sql” : 19,
“php” : 13
}

Here, strength is the name of our dictionary with four Key: Value pairs.

Now, it is very easy for us to access dictionary values. Even the value can

be changed using key. A key: value can be deleted from the dictionary. A

new key: value pair can be added dynamically.

CREATING A DICTIONARY

There are several methods to create a dictionary (McKinney Wes, 2022).

Example

#Empty dictionary
course = { }
#Empty dictionay using dict method
courseone = dict()
#dictionary with pre inserted values
coursetwo = {'math':30, 'physics': 24}
#dictionary with integer keys
mymark = {7: 'quiz', 10: 'test', 40: 'final'}
#dictionary using dict method
myaddress = dict({'street': 'alaqdar',
'city':'iraq'})

ACCESS DICTIONARY ITEMS

Using print() to print dictionary values.

mycar={
“brand”:”Hyundai”,
“model”:2022,
“color”:”red”
}
print(“car details “, mycar)
print(“car has “, len(mycar), ” items”)
print(“data type is “, type(mycar))

mycar is the name of the dictionary with three Key: Value pairs.

print(mycar) is used to print all dictionary values. len(mycar) is used to get

length of dictionary.

The above code generates the following output.

car details {'brand': 'Hyundai', 'model': 2022,

'color': 'red'}
car has 3 items
data type is <class 'dict'>

To get the value of a mycar key, we can use the following ways.

M = mycar['brand']
M = mycar.get('brand')
However, dictionary values can be changed as mycar['brand'] = 'Toyota'

A for loop is more effective to print keys and values.

Example

#Working with dictionary

mycar={
“brand”:”Hyundai”,
“model”:2022,
“color”:”red”
}
print(“---> Key names are “)
for k in mycar:
print(k)
print(“---> Values in dictionary are “)
for k in mycar:
print(mycar[k])

The above code generates the following output. Off course, it prints all keys

and values.

---> Key names are

brand
model
color
---> Values in dictionary are
Hyundai
2022
Red

Besides, the following methods are used to get keys and values of

dictionary.

ky = mycar.keys()
val = mycar.values()
print(“Dictionary keys are “, ky)
print(“Dictionary values are “, val)

A simplest way to check whether the key is present in dictionary is:

if “model” in mycar:
print(“Model is one of the key in dictionary”)
else:
print(“Model is not a key “)

ADDING AND REMOVING ITEMS

We can add a new item as dictionaryname[key] = value

mycar[“price”]=2700
mycar[“engine”]=”1600CC”

There are several methods to remove:

Method-1:

mycar.pop(“price”)

Method-2:

mycar.popitem(“price”)

Method-3:

del mycar[“brand”]

Method-4: Deleted entire dictionary

del mycar

DICTIONARY METHODS
Table 8 showcases a range of dictionary methods along with accompanying

descriptions and examples.

Table 8. Dictionary methods

Method Example Description

keys() ky = mycar.keys() Get all keys

values() val = mycar.values() Get all values associated with keys

Return number of items in a

len() siz = len(mycar)
dictionary

del() del mycar Delete entire dictionary

popitem() mycar.popitem(“price”) Remove most recently added item

pop() mycar.pop(“price”) Remove the specified item

Empty all values without deleting

clear() mycar.clear()
dictionary

copy() Newcar = mycar.copy() Return a copy of dictionary

CONCLUSION

A list is a flexible data type that can have heterogeneous items. List items

are accessed through index number. This chapter emphasizes list operations.

A list can be nested within another list. Tuples are used to group

information such as employees, students, and books. Tuple slicing allows

retrieval of some parts of required items. A tuple can be nested in another

tuple. Tuple values are accessed by index. This chapter covers tuple

initialization, accessing tuple values, functions, and the nesting of tuples.

This chapter concisely explains set properties and practical methods in our

daily programming life. Although the set has advantages, it has a few

limitations too. For example, a set can’t be embedded in another set. Also, a
set can’t have a list. A dictionary is a Key: Value pair that doesn’t allow

duplicate keys. There are several ways to access dictionary values.

Dictionaries are sometimes called associative arrays. The dictionary

provides faster execution than other data structures, such as lists.

REFERENCES

Arulkumar, N., Paulose, J., Galety, M. G., Manimaran, A.,

Saravanan, S., & Saleem, R. A. (2022). Exploring Social

Networking Data Sets. Social Network Analysis: Theory and

Applications, 205-228.

Beri, R. (2019). Python Made Simple: Learn Python programming

in easy steps with examples. Bpb Publications.

Bisong, E. (2019). Python. Building Machine Learning and Deep

Learning Models on Google Cloud Platform: A Comprehensive

Guide for Beginners, 71-89.

Guttag, J. V. (2021). Introduction to Computation and

Programming Using Python: With Application to Computational

Modeling and Understanding Data. MIT Press.

Haslwanter, T. (2016). An Introduction to Statistics with Python. In

With Applications in the Life Sciences. Springer International

Publishing. doi:10.1007/978-3-319-28316-6

Hill, C. (2020). Learning scientific programming with Python.

Cambridge University Press. doi:10.1017/9781108778039

Johansson, R., & John, S. (2019). Numerical Python (Vol. 1).

Apress. doi:10.1007/978-1-4842-4246-9
Kong, Q., Siauw, T., & Bayen, A. (2020). Python programming and

numerical methods: A guide for engineers and scientists. Academic

Press.

Learn To Code For IoT With Python Essentials Course. (2022).

Networking Academy.

https://www.netacad.com/courses/programming/pcap-

programming-essentials-python

Martelli, A., Ravenscroft, A. M., Holden, S., & McGuire, P. (2023).

Python in a Nutshell. O'Reilly Media, Inc.

McKinney. (2022). Python for Data Analysis. O'Reilly Media, Inc.

Moruzzi, G. (2020). Python basics and the interactive mode.

Essential python for the physicist, 1-39.

Neelakandan, S., Rene Beulah, J., Prathiba, L., Murthy, G. L. N.,

Fantin Irudaya Raj, E., & Arulkumar, N. (2022). Blockchain with

deep learning-enabled secure healthcare data transmission and

diagnostic model. International Journal of Modeling, Simulation,

and Scientific Computing, 13(04), 2241006.

doi:10.1142/S1793962322410069

Unpingco, J. (2021). Python Programming for Data Analysis.

Springer. doi:10.1007/978-3-030-68952-0
 CHAPTER 9

Interactive Visualization With Plotly Express

Gurram Sunitha

School of Computing, Mohan Babu University, Tirupati, India

A. V. Sriharsha

https://orcid.org/0000-0003-4244-4243

Mohan Babu University, India

Olimjon Yalgashev

Samarkand International University of Technology, Uzbekistan

Islom Mamatov

https://orcid.org/0000-0002-7029-5890

Samarkand International University of Technology, Uzbekistan

ABSTRACT

Data visualisation is the process of visualizing data for the purpose of

analyzing and correlating data to confirm complicated ideas and to spot

visual patterns after an investigation. Data visualization allows data

scientists to quickly identify visual patterns and trends that may not be

immediately apparent from raw data. Visualization also enables us to build

models that are easier for stakeholders to understand and apply. Data

visualization is an essential tool for exploratory data analysis (EDA). Plotly

Express is a terse, high-level Python visualization library. It is a wrapper

module for Plotly library that provides simple interface for visualizing data

in the form of various plots, maps, graph objects, layouts, traces, figures,

etc. Plotly Express is a user-friendly, high-level interface to Plotly that

generates easily styleable, interactive, beautiful, and informative figures

visualizing various types of data. Plotly Express provides interactive

visualizations.

IMPORTANCE OF DATA VISUALIZATION

Data visualization is converting data into visual representations, such as

plots, graphs, charts, and maps, that may be used to explain the data and

provide new perspectives. Data visualization is visualizing data to analyze

and correlate data to confirm complicated ideas and spot visual patterns

after an investigation. Data visualization plays a critical role in

understanding and interpreting data. It can be used to uncover relationships

in large, complex datasets and to make sense of the vast amounts of

information available today. Data visualization allows data scientists to

quickly identify visual patterns and trends that may not be immediately

apparent from raw data. By presenting data visually, data scientists can

promptly recognize outliers, patterns, and correlations that might otherwise

be difficult to spot. Visualization also enables us to build more accessible

models for stakeholders to understand and apply.

The main goal of data visualization is to make it easier for viewers to

identify patterns, styles, and vendors in large datasets. By creating visual

representations of data points such as charts, graphs, or maps, the most

critical aspects of the data set can be highlighted while still conveying its

complete story. This makes it easier for business analysts to interpret the

underlying trends or correlations within the dataset, which ultimately helps

them make better business decisions. Visual communication is a necessary

talent for all managers in this era. Data visualization is an essential tool for

Exploratory Data Analysis (EDA). It provides a way to gain insight into

data by visually representing it and helps reveal patterns, correlations,

outliers, and anomalies that may otherwise remain hidden. Data

visualization allows business analysts to quickly observe trends in data,

identify relationships between variables and understand the underlying

structure of the data. This helps us form hypotheses about what might be

causing specific patterns or behaviors in our data. Using data visualization

techniques, we can more quickly and easily explore large data sets than

possible by manually examining each observation. This makes it easier to

discover meaningful insights from our data that can inform decisions and

help guide further exploration.

PLOTLY EXPRESS FOR DATA VISUALIZATION

Plotly Express is a terse, high-level Python visualization library. It is a

wrapper module for the Plotly library that provides a simple interface for

visualizing data in various plots, maps, graph objects, layouts, traces,

figures, etc. Plotly Express is a user-friendly, high-level interface to Plotly,

which generates easily style able, interactive, beautiful, and informative

figures visualizing various types of data. Plotly Express provides interactive

visualizations; for example, the Plotly figures are by default appended with

hover text which the programmer can customize. An overview of Plotly

Express can be referred to at (Plotly Express in Python, 2022). Table 1

illustrates the different types of plots supported by Plotly Express.

Table 1. Types of plots supported by Plotly Express

Various Types of Plots Supported by Plotly

Plot Type
Express

Basic Plots scatter, dot, line, area, bar, funnel, timeline

Part-of- Whole pie, sunburst, treemap. icicle, funnel_area

1D Distributions histogram, box, violin, strip, ecdf

2D Distributions density_heatmap. density_contour

Matrix or Image
imshow
Input

3-Dimensional scatter_3d, line_3d

scatter matrix, parallel_coordinates,

Multidimensional
parallel_categories
Tile Maps scatter_mapbox, line_mapbox, choropleth_mapbox,

density_mapbox

Outline Maps scatter_geo, line geo, choropleth

Polar Charts scatter_polar, line_polar, bar_polar

Ternary Charts scatter_ternary, line_ternary

Plotly Express can create entire figures at once, thus reducing the burden on

programmers in creating figures. All the figures created using Plotly

Express are instances of plotly.graph_objects.Figure class. More than 30

functions are available in Plotly Express to develop various types of plots,

maps, graph objects, layouts, traces, and figures. Throughout a data

exploration session, it is simple to transition between multiple kinds of plots

since the API for these functions was deliberately created to be as consistent

and straightforward to understand as possible. Table 1 shows various types

of plots that Plotly Express supports. The Plotly Express API for multiple

plots, maps, graph objects, layouts, traces, and figures can be found at

(Python API reference for Plotly Express, 2022). Plotly Express has various

amicable features such as low code, user-friendly, simple API, automatic

figure labeling, sensible & overridable default parameter settings, flexible

input formats, automatic trace and layout configuration, uniform color

handling, faceting, animations, etc.

INTERACTIVE VISUALIZATION WITH PLOTLY EXPRESS

This section presents a brief overview and sample Python code snippets for

various types of plot visualizations – Scatter plot, Dot plot, Line plot, Area

plot, Pie chart, Bar chart, Grouped bar chart, Stacked bar chart, Sunburst

chart, Density heatmap.

Scatter Plot

The scatter plot is one of the types of relationship plot. A scatter plot

primarily shows the relationship between two variables (bivariate data). A

scatter plot can visually inspect the linear correlation between any two
variables. Also, it allows us to visualize when one of the two variables is

categorized. Clustered markers in a scatter plot, if any, enable the detection

of data patterns. It is simple to determine whether there is a positive or

negative or no correlation, strong or weak or no correlation, by simply

looking at the visualization of the scatter plot. It frequently provides

mathematicians with a solid beginning point for their research when they

can quickly see how the two variables relate. The following Table 2 is the

sample code snippet for visualizing the scatter plot for the considered data.

Table 2. Scatter plot

1 # Importing the “pandas” and “plotly” libraries

2 import plotly.express as px

3 import pandas as pd

4 # Creating a dataset of individuals “Gender”, “Height” and

5 “Weight”.

data = [[“Male”,180,84],[“Female”,150,60],[“Female”,154,55],

[“Male”,169,78],[“Male”,170,84], [“Female”,157,56],

[“Female”,164,65],[“Male”,165,69], [“Male”,172,73],

[“Female”,156,67], [“Female”,184,55], [“Male”,169,81],

[“Male”,178,90], [“Female”,156,61], [“Female”,168,55],

[“Male”,149,50], [“Male”,181,80], [“Female”,159,74],

6 [“Female”,169,85],[“Male”,151,48]]

# Creating a dataframe with data and column names

8 health = pd.DataFrame(data, columns=[“Gender”, “Height”,

9 “Weight”])

# Creating a scatter plot

10 fig = px.scatter(health, x=”Height”, y= “Weight”, color =

11 “Gender”,

symbol = “Gender”, title = “Scatter Plot - Height vs Weight”)

12 # Setting marker size, marker border width, and marker border-color

13 fig.update_traces(marker=dict(size=12,line=dict(width=2,

color= “DarkSlateGrey”)))

# Displaying the scatter plot

fig.show()

The Python libraries “plotly.express” and “pandas” have been imported

(Lines 2 & 3). A dataset has been created of twenty individuals with their

Gender (“Male”, “Female”), Height (in centimeters), and Weight (in

kilograms) (Line 5). A dataframe “health” has been created; it is initialized

with the dataset “data”, and the column names have been set for the

dataframe (Line 7). “px.scatter()” function creates an entire figure with a

scatter plot (Line 9).

The parameters for the “px.scatter()” function are set as follows

the dataset “health”,

the x-axis variable (x=”Height”),
the y-axis variable (y=”Weight”),
the color of the marker (color = “Gender”)
the symbol of the marker (symbol = “Gender”)
the title of the scatter plot (title = “Scatter
Plot - Height vs Weight”)

The marker color and the market shape are set differently based on Gender

by setting the parameters as “color = “Gender” and symbol = “Gender”.

This trick allows to include of a third dimension, “Gender” into the scatter

plot. The update_traces() function is used here to set marker size, marker

border line width, and marker border line color (Line 11). The show()

function will visualize the figure (Line 13). The Plotly Express API for the

“scatter()” function can be referred to at (Plotly Express for scatter plot,

2022) (Plotly Express API for scatter plot, 2022). The Plotly Express API

for the “update_traces()” function can be referred to at (Plotly Express API

for Creating and Updating Figures, 2022). Figure 1 shows the scatter plot
visualized using the above Python code snippet. Visual inspection shows

that the attributes “Height” and “Weight” are positively correlated. Once

the Plot is visualized, one can interact with the Plot by hovering the cursor

over the Plot (Figure 2).

Figure 1. Data visualization using scatter plot

Figure 2. Interaction with data visualized using scatter plot

Dot Plot

The dot plot is one of the variations of the scatter plot. A scatter plot

primarily shows the relationship between two numerical variables. In

contrast, a dot plot's primary use is to observe and establish relationships

between numeric and categorical variables. The linear correlation between a

numeric variable and a categorical variable can be visually inspected using

a scatter plot. The following is the sample code snippet for visualizing the

dot plot for the considered data. The categorical variable “Gender” is

considered on the x-axis of the Plot, and the numerical variable “Weight” is

considered on the y-axis of the Plot. Figure 3 shows the dot plot visualized

using the below Python code snippet. Table 3 shows the code for Dot Plot.

Table 3. Dot plot

1 # Importing the “pandas” and “plotly” libraries

2 import plotly.express as px

3 import pandas as pd
4 # Creating a dataset of individuals “Gender”, “Height” and

5 “Weight”.

data = [[“Male”,180,84],[“Female”,150,60],[“Female”,154,55],

[“Male”,169,78], [“Male”,170,84], [“Female”,157,56],

[“Female”,164,65],[“Male”,165,69], [“Male”,172,73],

[“Female”,156,67], [“Female”,184,55], [“Male”,169,81],

[“Male”,178,90], [“Female”,156,61], [“Female”,168,55],

[“Male”,149,50], [“Male”,181,80], [“Female”,159,74],

[“Female”,169,85],[“Male”,151,48]]

7 # Creating a dataframe with data and column names

health = pd.DataFrame(data, columns=[“Gender”, “Height”,

8 “Weight”])

# Creating a dot plot

fig = px.scatter(health, x=”Gender”, y= “Weight”, color =

10 “Gender”,

11 symbol = “Gender”, title = “Dot Plot - Gender vs Weight”)

# Setting marker size, marker border width, and marker border-color

12 fig.update_traces(marker=dict(size=12,line=dict(width=2,

13 color= “DarkSlateGrey”)))

# Displaying the dot plot

fig.show()

Figure 3. Data visualization using dot plot

Line Plot

A line plot plots quantitative values of a subject/concept/entity/theme/class

over a predetermined period (Dabbas, 2021). It is generally used to visually

inspect the trend of any subject/concept/entity/theme/class over some time.

It is a plot between any two variables, and it shows the movement of change

of one variable concerning the difference in the other variable. The

following Table 4 is the sample code snippet for visualizing a line plot for

the considered data.

Table 4. Line plot

1 # Importing the “pandas” and “plotly” libraries

2 import plotly.express as px

3 import pandas as pd

4 # Creating a dataset of students “Test” and “Marks in Biology”.

5 data = [[“T1”,48],[“T2”,51],[“T3”,56],[“T4”,64],[“T5”,59],
 [“T6”,69],

[“T7”,72],[“T8”,71], [“T9”,73],[“T10”,79]]

7 # Creating a dataframe with data and column names

marks = pd.DataFrame(data, columns=[“Test”, “Marks”])

9 # Creating a line plot

fig = px.line(marks, x=”Test”, y=”Marks”,

title = “Student 1 - Improvement in Biology”)

10

11 # Displaying the line plot

fig.show()

A dataset has been created of “Student 1” who has attempted ten tests (T1

to T10) in a biology course (Line 5). A dataframe “marks” has been

created; it is initialized with the dataset “data”, and the column names have

been set for the dataframe (Line 7). “px.line()” function creates an entire

figure with line plot (Line 9).

The parameters for the “px.line()” function are set as follows –

the dataset “marks”,

the x-axis variable (x= “Test”),
the y-axis variable (y= “Marks”),
the title of the line plot (title = “Student 1 -
Improvement in Biology”)

The show() function will visualize the figure (Line 11). The Plotly Express

API for the “line()” function can be referred to at (Plotly Express for line

plot, 2022) (Plotly Express API for line plot, 2022). Figure 4 shows the line

plot visualized using the above Python code snippet. By visual inspection, it

can be observed that the student has clearly improved his knowledge,

performed well, and shown considerable improvement in the biology

course.

Figure 4. Data visualization using line plot

The line plot can include a third dimension by setting the “color” parameter.

This parameter sets a different color for each line drawn on the line plot.

The following is the sample code snippet for visualizing a line plot with the

third dimension for the considered data. The Table 5 code snippet draws

multiple lines based on the “Subject”. Figure 5 shows the line plot with the

encoding column color visualized using the below Python code snippet.

Table 5. Line plot with the encoding column color

1 # Importing the “pandas” and “plotly” libraries

2 import plotly.express as px

3 import pandas as pd

4 # Creating a dataset of students “Subject”, “Test” and “Marks”.

5 data= [[“Biology”,”T1”,48],[“Biology”,”T2”,51],

[“Biology”,”T3”,56],

[“Biology”,”T4”,64], [“Biology”,”T5”,59],[“Biology”,”T6”,69],

[“Biology”,”T7”,72], [“Biology”,”T8”,71],[“Biology”,”T9”,73],
 [“Biology”,”T10”,79],[“Chemistry”,”T1”,58],

[“Chemistry”,”T2”,51],

[“Chemistry”,”T3”,56],[“Chemistry”,”T4”,62],

[“Chemistry”,”T5”,64],

6 [“Chemistry”,”T6”,70],[“Chemistry”,”T7”,78],

7 [“Chemistry”,”T8”,75],

[“Chemistry”,”T9”,79], [“Chemistry”,”T10”,84]]

9 # Creating a dataframe with data and column names

marks = pd.DataFrame(data, columns=[“Subject”, “Test”,

“Marks”])

10

11 # Creating a line plot

fig = px.line(marks, x=”Test”, y=”Marks”, color = “Subject”,

title = “Student 1 - Improvement in Biology & Chemistry”)

# Displaying the line plot

fig.show()

Figure 5. Data visualization using line plot with column encoding

color
Area Plot

Area plots are similar to line plots. First, the lines are drawn, and then the

area between the lines and the axes is filled. Hence, these plots are called

filled area plots. Generally, these plots are used to visualize multiple

variables' data trends over time. The filled areas can be overlapped or

stacked one over the other. These graphs are the standard approach to show

stacked lines and often indicate accumulated totals over time. The following

Table 6 is the sample code snippet for visualizing a filled area plot for the

considered data.

Table 6. Area plot

1 # Importing the “pandas” and “plotly” libraries

2 import plotly.express as px

3 import pandas as pd

4 #Creating a dataset of student's “Subject”, “Test” and “Marks”.

5 data= [[“Biology”,”T1”,48],[“Biology”,”T2”,51],

[“Biology”,”T3”,56],[“Biology”,”T4”,64],[“Biology”,”T5”,59],

[“Biology”,”T6”,69],[“Biology”,”T7”,72],[“Biology”,”T8”,71],

[“Biology”,”T9”,73],[“Biology”,”T10”,79],[“Chemistry”,”T1”,58],

[“Chemistry”,”T2”,51],[“Chemistry”,”T3”,56],

[“Chemistry”,”T4”,62],

[“Chemistry”,”T5”,64],[“Chemistry”,”T6”,70],

[“Chemistry”,”T7”,78],

[“Chemistry”,”T8”,75],[“Chemistry”,”T9”,79],

6 [“Chemistry”,”T10”,84]]

8 # Creating a dataframe with data and column names

9 marks = pd.DataFrame(data, columns=[“Subject”,”Test”, “Marks”])

 # Creating an area plot

10 fig = px.area(marks, x=”Test”, y=”Marks”, color = “Subject”,

11 title = “Student 1 - Improvement in Biology & Chemistry”)

# Displaying the area plot

fig.show()

Figure 6. Data visualization using area plot

A dataset has been created of “Student 1” who has attempted ten tests (T1

to T10) in biology and chemistry courses (Line 5). A dataframe “marks”

has been created, it is initialized with the dataset “data”, and the column

names have been set for the dataframe (Line 7). “px.area()” function creates

an entire figure with filled area plot (Line 9).

The parameters for the “px.area()” function are set as follows –

the dataset “marks”,

the x-axis variable (x= “Test”),
the y-axis variable (y= “Marks”),
the color for the filling area (color = “Subject”)
the title of the line plot (title = “Student 1 -
Improvement in Biology & Chemistry”)

The Plotly Express API for the “area()” function can be referred to at (Plotly

Express for filled area plot, 2022) (Plotly Express API for filled area plot,

2022). Figure 6 shows the filled area plot visualized using the above Python

code snippet. By visual inspection, it can be observed that the student has

clearly improved his knowledge, performed well, and has shown

considerable improvement in both aggregated courses.

Pie Chart

A pie chart visualizes a part-in-the-whole relationship (Pattanaik, S. N., &

Wiegand, R. P., 2021). If the data comprises categories where each

contributes to a whole, then such partial contributions can be visualized

using a pie chart. A pie chart is graphically visualized as a pie (circular-

shaped) comprising slices, each representing a category in the data. In the

pie chart, the size of a slice of a particular variety represents the

proportional contribution of that category to the whole, i.e., each slice in the

pie chart represents a proportion of data of each type. Using a pie chart, the

importance/contribution of each category to the whole. The following Table

7 is the sample code snippet for visualizing a pie chart for the considered

data.

Table 7. Pie chart

1 # Importing the “pandas” and “plotly” libraries

2 import plotly.express as px

3 import pandas as pd

4 # Creating a dataset of population distribution in a country as per

# gender (in millions).

5 data = [[“Male”,180], [“Female”,150],[“Others”,25]]

6 # Creating a dataframe with data and column names

7 gender_distribution = pd.DataFrame(data,

columns=[“Gender”, “Population”])

8 # Creating a pie chart

9 fig = px.pie(gender_distribution, values=”Population”,

names=”Gender”,

title = “Pie Chart - Gender Distribution”)

10

11 # Displaying the pie chart

fig.show()

A dataset has been created representing the population distribution in a

country as per Gender (in millions) (Line 5). A dataframe

“gender_distribution” has been completed; it is initialized with the dataset

“data”, and the column names have been set for the dataframe (Line 7).

“px.pie()” function creates an entire figure with a pie chart (Line 9).

The parameters for the “px.pie()” function are set as follows –

the dataset “gender_distribution”,

categories in the pie (the name for each slice in
the pie) (names=”Gender”),
size of each slice in the pie
(values=”Population”),
the color for filling area (color = “Subject”)
the title of the pie chart (title = “Pie Chart -
Gender Distribution”)

The Plotly Express API for the “pie()” function can be referred to at (Plotly

Express for pie chart, 2022) (Plotly Express API for pie chart, 2022). Figure

7 shows the pie chart visualized using the above Python code snippet. By

visual inspection, it can be observed that the male population in the country

contributes 50.7%, and the female population contributes 42.3%, which is

less when compared to the male population.

Figure 7. Data visualization using pie chart

Bar Chart

Bar charts can be used for displaying metrics of categories of data (Sievert,

2020). Bar charts can help compare multiple types of data or discrete data,

provided there aren't too many categories to compare. The values are

compared over time between the various groups using the bar chart. To

track changes over time, bar graphs are utilized. For time series data, they

are highly beneficial. The bars on the bar chart can be drawn horizontally or

vertically. On the bar chart, each bar on the x-axis (or y-axis) represents a

category. All the bars have uniform width (or height). The bar's height (or

length) represents the category's value or metric on the y-axis (or x-axis). A

bar chart is a statistical approach to defining the statistics of each type of

data. Bar charts show the relationship between any two dimensions. On one

axis, it represents the categories; on the other, it means the category's

discrete values. The following is the sample code snippet for visualizing a

bar chart for the considered data. There is one category in the data – “Year,”

and the statistical value for each class is the “Population.” In the following

code, the category “Year” is taken on the x-axis, and the y-axis represents

the statistical data of each “Year.” The height of the bar on the chart denotes
the “Population” value for each “Year,” while the width of the bar indicates

the “Year” itself. Table 8 has the code for bar chart.

Table 8. Bar chart

# Importing the “pandas” and “plotly” libraries

1
import plotly.express as px
2
import pandas as pd
3

# Creating a dataset of population distribution in the USA during

4
2019-

# 2022 (in millions).

5
data = [[“2019”,328.24],[“2020”,331.5],[“2021”,331.89],

[“2022”,333.39]]

6
# Creating a dataframe with data and column names
7
population = pd.DataFrame(data, columns=[“Year”, “Population”])

8
# Creating the bar chart
9
fig = px.bar(population, x=”Year”, y=”Population”,

title = “Bar Chart - Population”)

10
# Setting the scale of the y-axis
11
fig.update_layout(yaxis_range=[325,335])

12
# Displaying the bar chart
13
fig.show()

A dataset has been created representing the population distribution in a

country per year (in millions) (Line 5). A dataframe “Population” has been

completed; it is initialized with the dataset “data”, and the column names

have been set for the dataframe (Line 7). “px.bar()” function creates an

entire figure with bar chart (Line 9).

The parameters for the “px.bar()” function are set as follows –

the dataset “Population”,

the x-axis variable representing the category (x =
“Year”),
the y-axis variable representing the contribution
of categories to the whole (y = “Population”),
the title of the pie chart (title = “Bar Chart -
Population”)

Figure 8. Data visualization using bar chart

The Plotly Express API for the “bar()” function can be referred to at (Plotly

Express for bar chart, 2022) (Plotly Express API for bar chart, 2022). The

update_layout() function is used to set the y-axis range. The y-axis scale

ranges from a minimum value of 325 to a maximum of 335. Figure 8 shows

the bar chart visualized using the above Python code snippet. By visual

inspection, it can be how the population in a country has increased over the

years 2019 to 2022.

Grouped Bar Chart

If the data contains two categories, a third dimension can be added to the

bar chart, visualized as a grouped bar chart. The following is the sample

code snippet for visualizing a bar chart for the considered data. There are

two categories in the data – “Year” and “Gender,” and the statistical data is

the “Population.” In the following code, the category “Year” is taken on the

x-axis, and the y-axis represents the statistical data of each “Year.” The

height of the bar on the chart denotes the “Population” value for each

“Year,” while the width of the bar indicates the “Year” itself. The other

category, “Gender,” is considered the third dimension where for each

“Gender” subcategory, multiple bars will be visualized. They will be

grouped for each “Year” on the x-axis, i.e., in the grouped bar chart, the

bars representing the third dimension “Gender” are grouped for each x-axis

value “Year”. Table 9 has the code for Grouped Bar chart.

Table 9. Grouped bar chart

1 # Importing the “pandas” and “plotly” libraries

2 import plotly.express as px

3 import pandas as pd

4 # Creating a population distribution dataset as per Gender in the

USA

5 during 2019-2022 (in millions.

data = [[“2019”,”Male”,161.66], [“2019”,”Female”,166.58],

[“2020”,”Male”,164.21], [“2020”,”Female”,167.29],

[“2021”,”Male”,164.38], [“2021”,”Female”,167.51],

[“2022”,”Male”,165.13], [“2022”,”Female”,168.26] ]

7 # Creating a dataframe with data and column names

gender_distribution = pd.DataFrame(data,

columns=[“Year”, “Gender”, “Population”])

9 # Creating the grouped bar chart

 fig = px.bar(gender_distribution, x=”Year”, y=”Population”,

color=”Gender”, barmode='group', text_auto=True,

title = “Grouped Bar Chart - Gender Distribution”)

10

11 # Setting the scale of the y-axis

fig.update_layout(yaxis_range=[160,170])

12

13 # Displaying the grouped bar chart

fig.show()

A dataset has been created representing the population distribution in a

country per year (in millions) and Gender (Line 5). A dataframe

“gender_distribution” has been completed; it is initialized with the dataset

“data”, and the column names have been set for the dataframe (Line 7).

“px.bar()” function creates an entire figure with a bar chart (Line 9).

The parameters for the “px.bar()” function are set as follows –

the dataset “gender_distribution”,

the x-axis variable representing the first
category (x = “Year”),
the y-axis variable representing the contribution
of categories to the whole (y = “Population”),
the second category is taken as a third dimension
(color= “Gender”)
the mode of the bars of the third dimension
(barmode = “group”)
display of text on the bars (“text_auto=True”)
the title of the pie chart (title = “Grouped Bar
Chart - Population”)

Figure 9 shows the grouped bar chart visualized using the above Python

code snippet. Observe that there are two bars visualized for each “Year” on

the x-axis (one bar representing one subcategory (“Male”, “Female”) of the

third dimension “Gender”). By visual inspection, it can be how various

genders are contributing to the population in a country over the years 2019

to 2022.

Figure 9. Data visualization using grouped bar chart

Stacked Bar Chart

With one aesthetic distinction, the stacked and grouped bar charts are

comparable (Zhao, 2023). The bars representing the third dimension,

“Gender,” are stacked one over another rather than being grouped side-by-

side. The following Table 10 has the sample code snippet for visualizing a

bar chart for the considered data.

Table 10. Stacked bar chart

1 # Importing the “pandas” and “plotly” libraries

2 import plotly.express as px

3 import pandas as pd
4 # Creating a population distribution dataset per genders in USA #

during 2019-2022 (in millions).

5 data = [[“2019”,”Male”,161.66], [“2019”,”Female”,166.58],

[“2020”,”Male”,164.21], [“2020”,”Female”,167.29],

[“2021”,”Male”,164.38], [“2021”,”Female”,167.51],

[“2022”,”Male”,165.13], [“2022”,”Female”,168.26] ]

6 # Creating a dataframe with data and column names

7 gender_distribution = pd.DataFrame(data,

columns=[“Year”, “Gender”, “Population”])

8 # Creating the stacked bar chart

9 fig = px.bar(gender_distribution, x=”Year”, y=”Population”,

color=”Gender”, title = “Stacked Bar Chart - Gender Distribution”,

pattern_shape=”Gender”, pattern_shape_sequence=[“.”, “x”],

text_auto=True)

10 # Displaying the stacked bar chart

11 fig.show()

Observe that in the bar() function, the mode of the bars (barmode=

“group”) is not mentioned. If “barmode” is not mentioned, the bars

representing the third dimension will be stacked rather than grouped. Also,

different patterns and shapes can be used to fill the bars along with the

color, if appropriate.

Two more parameters may be mentioned in the bar() function –

the filling patterns for the bars (pattern_shape =

“Gender”)
the filling shapes for the bars
(pattern_shape_sequence = [“.”, “x”])

Figure 10. Data visualization using stacked bar chart

Figure 10 shows the stacked bar chart visualized using the above Python

code snippet. The third dimension, “Gender,” is pictured as stacked bars.

Only one bar represents each “Year” on the x-axis. Each subcategory

(“Male”, “Female”) of the third dimension, “Gender” is visualized as part

of the stacked bar.

Sunburst Chart

The sunburst chart is a variation of the doughnut plot used to show a

hierarchical structure (Rees & Laramee, 2019). Hierarchical data is best

displayed with a sunburst chart. Without hierarchical information (only one

level of categories), a sunburst chart resembles a doughnut. A sunburst plot

shows discrete hierarchical data from the roots to the leaves. The root forms

the core (innermost circle) of the Plot. Then, outer rings are added to the

core as needed. One ring or circle represents each hierarchy level, with the

innermost circle acting as the parent or root for all the data. Subsequent

rings are divided into slices, and the size of each slice is proportional to the

data value. Each ring demonstrates how the outer rings relate to the interior

rings.
The following Table 11 has the sample code snippet for visualizing a

sunburst chart for the considered data.

Table 11. Sunburst chart

#Importing the “pandas” and “plotly” libraries

1
import plotly.express as px
2
import pandas as pd
3

# Creating a dataset of weekend planners.

4
data = dict(event=[“Breakfast”,”Movie”,”Lunch”,”Shopping”,
5
“Drinks”,

“Concert”,”Gym”,”Brunch”,”Family Meet”],

day=[“Saturday”,”Saturday”,”Saturday”,

“Saturday”,”Saturday”,”Saturday”,”Sunday”,

“Sunday”,”Sunday”],

hours=[1, 3, 2, 2, 1, 6, 2, 4, 4])

6
# Creating the sunburst plot
7
fig = px.sunburst(data, path = [“day”, “event”],values=”hours”)

8
# Displaying the sunburst plot
9
fig.show()

Figure 11. Data visualization using sunburst chart

A dataset represents a two-level hierarchy of the weekend planner (Line 5).

The attribute “day” is the sunburst plot's root (or core). The attribute
nd
“event” forms the outer ring (2 level). The attribute “day” includes the
nd
data values that decide the size of each slice in the 2 level ring.

“px.sunburst()” function creates an entire figure with a bar chart (Line 7).

The parameters for the “px.sunburst ()” function are set as follows –

the dataset “data”,

the path specifying the hierarchy of the
hierarchical structure (path = [“day”, “event”]),
the data values (values = “hours”),
The Plotly Express API for “bar()” function can be referred to at (Plotly

Express for sunburst chart, 2022) (Plotly Express API for sunburst chart,

2022). Figure 11 shows the sunburst chart visualized using the above

Python code snippet.

Density Heatmap

In a heatmap, values for a primary variable plotted against two-axis

variables are shown as a grid of colored squares (Berinato, 2019). The

continuous color scale is used to represent the density of the primary

variable, thus providing visual clues on the density or sparsity of the

primary variable when compared against the two-axis variables. Darker

colors on the color scale represent richer density levels, and sparse density

levels are represented by lighter colors on the color scale. The following

Table 12 has the sample code snippet for visualizing a density heatmap for

the considered data.

Table 12. Density heatmap

1 # Importing the “pandas” and “plotly” libraries

2 import plotly.express as px

3 import pandas as pd

4 # Creating a dataset of students “Subject”, “Test” and “Marks”.

5 data= [[“Biology”,”T1”,48],[“Biology”,”T2”,51],

[“Biology”,”T3”,56], [“Biology”,”T4”,64],[“Biology”,”T5”,59],

[“Biology”,”T6”,69], [“Biology”,”T7”,72],[“Biology”,”T8”,71],

[“Biology”,”T9”,73], [“Biology”,”T10”,79],[“Chemistry”,”T1”,58],

[“Chemistry”,”T2”,51],

[“Chemistry”,”T3”,56],[“Chemistry”,”T4”,62],

[“Chemistry”,”T5”,64],

[“Chemistry”,”T6”,70],[“Chemistry”,”T7”,78],

[“Chemistry”,”T8”,75], [“Chemistry”,”T9”,79],
6 [“Chemistry”,”T10”,84]]

8 # Creating a dataframe with data and column names

9 marks = pd.DataFrame(data, columns=[“Subject”,”Test”, “Marks”])

# Creating an area plot

10 fig = px.density_heatmap(marks, x=”Subject”, y=”Test”, z =

11 “Marks”,

title = “Student 1 - Improvement in Biology & Chemistry”)

# Displaying the area plot

fig.show()

A dataset has been created of “Student 1” who has attempted ten tests (T1

to T10) in biology and chemistry courses (Line 5). A dataframe “marks”

has been created, it is initialized with the dataset “data”, and the column

names have been set for the dataframe (Line 7). “px.density_heatmap()”

function creates an entire figure with density heatmap (Line 9).

The parameters for the “px.density_heatmap()” function are set as follows –

the dataset “marks”,

the x-axis variable (x= “Subject”),
the y-axis variable (y= “Test”),
the z-axis variable, which will be visualized as
density represented by variation of density on the
color scale (z = “Marks”),
the title of the density heatmap(title=“Student 1
- Improvement in Biology & Chemistry”)

The Plotly Express API for “density_heatmap()” function can be referred to

at (Plotly Express for density heatmap, 2022) (Plotly Express API for

density heatmap, 2022). Figure 12 shows the density heatmap visualized

using the above Python code snippet. By visual inspection, it can be

observed that the student has clearly improved his knowledge, performed
well, and has shown considerable improvement in both aggregated courses.

Also, it can be observed that the student has performed well in the

“Biology” approach than in “Chemistry” over tests.

Figure 12. Data visualization using density heatmap

CONCLUSION

Data visualization is visualizing data to analyze and correlate data to

confirm complicated ideas and spot visual patterns after an investigation.

By presenting data visually, data scientists can quickly recognize outliers,

patterns, and correlations that might otherwise be difficult to spot.

Visualization also enables us to build more accessible models for

stakeholders to understand and apply. Data visualization is an essential tool

for Exploratory Data Analysis (EDA).

Plotly Express is a terse, high-level Python visualization library. It is a

wrapper module for the Plotly library that provides a simple interface for

visualizing data in various plots, maps, graph objects, layouts, traces,

figures, etc. Plotly Express is a user-friendly, high-level interface to Plotly,

which generates easily designable, interactive, beautiful, and informative

figures visualizing various types of data. Plotly Express provides interactive

visualizations; for example, the Plotly figures are by default appended with

hover text which the programmer can customize. More than 30 functions

are available in Plotly Express to create various types of plots, maps, graph

objects, layouts, traces, and figures. Throughout a data exploration session,

it is simple to transition between multiple kinds of plots since the API for

these functions was deliberately created to be as consistent and

straightforward to understand as possible. Plotly Express has various

amicable features such as low code, user-friendly, simple API, automatic

figure labeling, sensible & overridable default parameter settings, flexible

input formats, automatic trace and layout configuration, uniform color

handling, faceting, animations, etc.

This chapter presents a brief overview of the importance of data

visualization as part of Exploratory Data Analysis (EDA). Briefly, it

explains the capabilities of the Plotly Express Python graphing library.

Various types of plot visualizations such as Scatter plots, Dot plots, Line

plots, Area plots, Pie charts, Bar charts, Grouped bar charts, Stacked bar

charts, Sunburst charts, and Density heatmaps have been presented and

discussed along with sample Python code snippets.

REFERENCES

Berinato, S. (2019). Good charts workbook: Tips, tools, and

exercises for making better data visualizations. Harvard Business

Press.

Dabbas, E. (2021). Interactive Dashboards and Data Apps with

Plotly and Dash: Harness the power of a fully fledged frontend web

framework in Python–no JavaScript required. Packt Publishing Ltd.

Pattanaik, S. N., & Wiegand, R. P. (2021). Data visualization.

Handbook of Human Factors and Ergonomics, 893-946.

Plotly Express API for bar chart. (n.d.). Retrieved September 2022,

from https://plotly.com/python-api-

reference/generated/plotly.express.bar.html

Plotly Express for bar chart. (n.d.). Retrieved September 2022,

from https://plotly.com/python/bar-charts/

Plotly Express API for Creating and Updating Figures. (n.d.).

Retrieved September 2022, from

https://plotly.com/python/creating-and-updating-figures/

Plotly Express API for density heatmap. (n.d.). Retrieved

September 2022, from https://plotly.com/python-api-

reference/generated/plotly.express.density_heatmap.html

Plotly Express API for filled area plot. (n.d.). Retrieved September

2022, from https://plotly.com/python-api-

reference/generated/plotly.express.area.html

Plotly Express API for line plot. (n.d.). Retrieved September 2022,

from https://plotly.com/python-api-

reference/generated/plotly.express.line.html

Plotly Express API for pie chart. (n.d.). Retrieved September 2022,

from https://plotly.com/python-api-

reference/generated/plotly.express.pie.html

Plotly Express API for scatter plot. (n.d.). Retrieved September

2022, from https://plotly.com/python-api-

reference/generated/plotly.express.scatter.html/

Plotly Express API for sunburst chart. (n.d.). Retrieved September

2022, from https://plotly.com/python-api-

reference/generated/plotly.express.sunburst.html

Plotly Express in Python. (n.d.). Retrieved September 2022, from

https://plotly.com/python/plotly-express/

Python API reference for Plotly Express. (n.d.). Retrieved

September 2022, from https://plotly.com/python-api-reference/

Rees, D., & Laramee, R. S. (2019, February). A survey of

information visualization books. Computer Graphics Forum, 38(1),

610–646. doi:10.1111/cgf.13595

Sievert, C. (2020). Interactive web-based data visualization with R,

plotly, and shiny. CRC Press. doi:10.1201/9780429447273

Zhao, P. (2023). Visualizing Data. In Working with Data in Public

Health: A Practical Pathway with R (pp. 131–158). Springer.

doi:10.1007/978-981-99-0135-7_8
 CHAPTER 10

Enhancing Big Data Analytics and

Recommendation Systems With Python:

Hierarchical Clustering and Filtering Using Slicing

Technique
P. Thamilselvan

Bishop Heber College, India

ABSTRACT

The advancement of technology has led to an exponential increase in the

volume, velocity, and variety of data generated, necessitating the

development of effective methods for analyzing and extracting valuable

insights from large datasets. This research focuses on enhancing big data

analytics and recommendation systems using Python, specifically

employing hierarchical clustering and a filtering approach with the slicing

technique. This study proposes a novel approach to leverage Python's

capabilities in processing and analyzing big data. Hierarchical clustering

algorithms organize and structure data into hierarchical groups, enabling

efficient exploration and extraction of relevant information. Additionally, a

filtering mechanism is integrated with the slicing technique, allowing for

identifying and extracting specific subsets of data based on predefined

criteria. Experiments are conducted using real-world datasets in the context

of recommendation systems to evaluate the approach's effectiveness.

INTRODUCTION
Big Data consists of vast data from diverse and independent sources,

making traditional software difficult to manage (Beyer & Laney, 2012).

Analyzing and extracting helpful information for future operations is a

crucial challenge in Big Data applications. Recommendation Systems (RSs)

are sophisticated instruments that aid consumers in making product or

service selections. Recommendation systems confront two formidable

challenges in applications containing enormous amounts of data. The first

challenge is making decisions promptly. It is crucial that recommendation

systems efficiently process and provide user-satisfying suggestions in light

of the immense quantities of available data. The second challenge is

generating optimal service recommendations from various alternatives. As

the number of services and options increases, it becomes more challenging

to identify the most relevant and customized offers for users. These

obstacles must be surmounted for recommendation systems to provide

timely and high-quality recommendations within the context of big data

applications.

Clustering to integrate similar services has proven highly effective at

reducing aggregate data size (Rajaraman & Ullman, 2012). Clustering and

collaborative filtering are the two primary phases of this procedure (Zheng

et al., 2013). Each subset corresponds to a discrete cluster consisting of

services with similar characteristics. It is possible to considerably reduce

computational time by limiting the number of services subject to the

collaborative filtering phase to a subset within a particular cluster. Notably,

this approach leverages the inherent relevance of ratings ascribed to similar

services within a given cluster, significantly improving the accuracy of

recommendations derived from user ratings (Bellogín et al., 2013).

CLUSTERING

Clustering has the potential to significantly reduce data volume by

clustering together similar services (Rajaraman & Ullman, 2012).

Clustering and collaborative filtration are the two fundamental phases of

this procedure. Clustering is an essential preparatory step that facilitates the

subdivision of enormous datasets into smaller, more manageable subsets

(Zheng et al., 2013). Each cluster consists of services with comparable

characteristics, enabling a more precise analysis. It is possible to

significantly reduce computational time during the collaborative filtering

phase by limiting the number of services within a cluster relative to the total

benefits. Moreover, given the inherent relevance of ratings about services

within a cluster as opposed to ratings about disparate services, the accuracy

of recommendations based on user ratings can be substantially improved

(Bellogín et al., 2013).

Clustering is a crucial element of data analysis and data mining applications

(Zeng et al., 2010). Objects from the same category are more comparable

than those from different types. A cluster is a collection of data objects

distinguished by their unique attributes and inherent properties. Cluster

analysis organizes data objects into meaningful groups by identifying

similar features (Li et al., 2009).

Document clustering is advantageous because it eliminates the need for

human intervention in the classification process (Havens et al., 2012).

Instead, cluster membership is determined based on the distribution and

composition of the data. A competent clustering method produces high-

quality clusters characterized by high similarity within each cluster and low

similarity between clusters. The effectiveness of a clustering result is

exceptionally dependent on the similarity measure employed and its

implementation (Liu et al., 2009). In addition, a clustering technique's

efficacy is contingent on its ability to reveal underlying patterns, thereby

illuminating the proximity of clusters through a distance function (Liu et al.,

2009). When aggregating data using data mining techniques, adhering to a

set of requirements is essential. Clustering-based strategies for collaborative

filtering typically employ two distinct clustering algorithms.

RELATED WORK

Density-based clustering, a widely employed method for data partitioning

based on density, involves separating high- and low-density points using a

defined threshold (Ester et al., 1996). This technique serves as the

foundation for density-based clustering algorithms. However, these

algorithms lack inherent support for variable density, necessitating the

development of alternative approaches to address this limitation.

An example of such an algorithm is OPTICS, proposed by Ankerst et al.,

which aims to overcome a significant constraint of DBSCAN by identifying

meaningful clusters in datasets with varying densities. It achieves this by

expanding the ordering of data points (Parimala et al., 2011). Another

algorithm, VDBSCAN, was introduced by (Liu et al., 2011) to tackle the

issue of variable density datasets within the DBSCAN framework. Before

employing the conventional DBSCAN algorithm, VDBSCAN utilizes

several techniques to determine the boundary parameter Eps based on a k-

dist plot those accounts for varying densities. This enables the simultaneous

discovery of clusters with diverse densities.

The VDBSCAN algorithm comprises five phases, including:

• Computation and storage of the k-dist for each data item,

constructing k-dist plots

• Determination of the number of densities from the k-dist plot

• Automatic selection of Eps values for each density

• Scanning the dataset and clustering data points with different

densities using the corresponding Eps values; and

• Visualization of the resulting valid clusters (Liu et al., 2007).

LDBSCAN, a density-based clustering algorithm based on local density,

uses the Local Outlier Factor (LOF) to distinguish noise, setting it apart

from other density-based clustering techniques (Duan et al., 2007). As

proposed by Ram and Jalal, DVBSCAN introduces the density-variable

DBSCAN algorithm (Ram et al., 2013) that can account for variations in

local cluster density. AUTOEPSDBSCAN is an enhanced algorithm that

chooses input boundaries autonomously, enabling the identification of

clusters with variable densities, shapes, and sizes in large datasets

comprising anomalies and noise. AUTOEPSDBSCAN outperforms the

conventional DBSCAN algorithm by only requiring a single input boundary

and producing superior results.

Borah and Bhattacharyya's DDSC algorithm is an effective method for

detecting clusters with varying densities (Borah & Bhattacharyya, 2008).

Even when discrete regions do not serve as apparent separators, DDSC

attempts to identify cohesive groupings based on density by distributing

local densities uniformly within a cluster and dividing local regions into

separate clusters in the presence of significant density changes.

VMDBSCAN, an extension of the DBSCAN algorithm, concentrates on

detecting clusters with variable sizes, geometries, and local densities

(Elbatta et al., 2011). This strategy entails identifying the “center” of each

cluster generated by DBSCAN and then introducing “vibrations” to the

cluster characteristics that impact these pivotal nodes most. VMDBSCAN

facilitates the determination of the optimal number of clusters. These

examples illustrate a variety of density-based clustering algorithms.

Despite the emergence of the big data paradigm (Ishwarappa, 2011), it has

been discovered that the DBSCAN algorithm is inefficient when processing

large data volumes on a single machine. Therefore, the most effective

method for analyzing such massive datasets is to employ distributed

computing technology (Akbar et al., 2015). Hadoop (Hadoop, 2009), an

open-source platform, offers distributed storage and processing capabilities

for administering vast amounts of data. It is founded on the MapReduce

programming model (He et al., 2014), ideally suited for massive data

processing. Mahran and Mahar (2008) developed the GriDBSCAN

algorithm to resolve the deficiencies of DBSCAN. This algorithm builds

multiple regular grids and assigns data points to similar grids, thereby

producing partitions with distinct borders. Subsequently, DBSCAN is

utilized to independently examine each section and its respective boundary

points to merge. Nonetheless, regular grids in GriDBSCAN may divide

datasets into high-density regions, producing many duplicate boundary

points.

Given the obstacles mentioned earlier, He et al. (2014) proposed the MR-

DBSCAN approach, which employs the Hadoop platform and MapReduce

to facilitate distributed DBSCAN processing. Load balancing in large

datasets is given special consideration to accelerate and efficiently scale

skewed big data. The algorithm consists of three primary stages: data
partitioning, local clustering, and global merging. In addition, Dai and Li

developed the PRBP algorithm (Dai & Lin, 2012) to accomplish load

balancing on individual nodes by choosing partition boundaries based on

the distribution of data points. DBSCAN-MR, a MapReduce-based method,

enables cloud-based parallel processing without requiring a global index. In

addition, the inclusion of PRBP optimizes data partitioning. Bhardwaj and

Das (Bhardwaj & Dash, 2015) developed a density-based approach for the

DBSCAN-MR and VDMR-DBSCAN algorithms. Incorporating eps_diff

into their merging strategy can prevent clusters of varying densities from

sharing boundary points. Utilizing PRBP, the algorithm identifies clusters

with varying densities within each partition, albeit at the expense of some

inter-partition connectivity.

PROPOSED WORK

This research paper presents a HACE framework that proposes a data

mining-based paradigm for managing Big Data. The framework

encompasses multiple stages, including data acquisition from diverse

sources, data extraction and analysis, modeling of user interests, and

considerations for security and privacy. The advent of the data-driven model

and the ongoing revolution in Big Data present unique challenges that can

be effectively addressed. The capability to extract valuable insights from

these vast datasets or data streams, referred to as Big Data mining, has

emerged as a previously inconceivable prospect owing to the sheer volume,

variability, and velocity of the data. The Big Data challenge represents an

exciting opportunity to shape the future landscape. The term “Big Data”

denotes datasets that exceed the capacity of our current systems and

traditional data mining software tools in terms of size and complexity. This

technology finds application in diverse domains, including credit card fraud

detection and enterprise risk management, where large databases are

utilized, owing to its significant performance enhancements over

conventional data extraction and analysis techniques. The overall process of

Big Data management is illustrated in Figure 1.

Figure 1. Process of bigdata

Data Preprocessing

It is impossible to exaggerate the significance of data preprocessing in

preparing unstructured data for effective data mining analytics. Frequently,

primary data are insufficient and formatted inconsistently, making

subsequent analysis difficult. The quality of data formatting directly impacts

the success of any data analytics endeavor. Data preprocessing entails

essential operations like data purification and validation. Data validation

aims to assess the accuracy and completeness of the data, thereby ensuring

its reliability for further analysis. Data cleansing involves the elimination of

errors and the administration of missing values through the use of manual

intervention or automated software tools. Both database-driven and rule-

based applications require data preprocessing to ensure data integrity and

usability. Data preprocessing is essential in artificial intelligence for

structuring large datasets in a manner that facilitates interpretation and

processing by learning algorithms.

MULTI-DIMENSIONAL CONTENT DIFFICULTY RATING

Users may become frustrated and forsake their learning sessions if they find

the content too difficult to comprehend. A dynamic strategy to regulate the

difficulty of content based on the performance of each user across multiple

dimensions is proposed as a solution to this problem. The first dimension

focuses on classifying the difficulty of Big Data content by identifying

repetitively complex or straightforward inquiries. The proficiency

requirements for these queries are then adjusted for all users. The second

dimension attempts to balance intended difficulty levels and learning

progress by altering the content difficulty within each dataset based on the

performance of individual users.

The development of distributed computing infrastructures has significantly

accelerated the storage, processing, and distribution of enormous quantities

of data. However, implementing reliable ranking mechanisms in search

engine results and effectively supporting keyword-based queries present

formidable challenges. Existing research primarily focuses on single-

keyword questions and lacks appropriate ranking systems. We propose a

flexible multi-keyword query method to circumvent this restriction. This

algorithm considers keyword weights and user access history when

compiling search results. The search results displayed on search engine

results pages are generated by a multi-keyword search engine, a software

application designed to retrieve data from immense data collections. These

outcomes may contain web pages, images, and other relevant files.

Figure 2. Process of multi-dimensional content difficulty rating

EXPERIMENTAL RESULT
The term “stem” in phonetics pertains to a constituent element of a word,

and its meaning exhibits slight variations. In one usage, a stem is a

structural foundation to which affixes can be attached. For instance, in the

English word “fellowships,” the stem “companion” provides the base form,

to which the derivational suffix “-ship” is appended, resulting in the

formation of the derived stem “friendship.” Subsequently, the inflectional

suffix “-s” is attached, forming the final stem “friendships.” However, it is

essential to note that the original root of the word (in this case,

“companion”) is not considered a stem in this particular interpretation.

Within this article, the term “stem” refers to the shared portion of a word in

all its inflected forms. Hence, in this usage, all derivational affixes

contribute to the stem. For example, in the case of “friendships,” the stem is

“friendship,” and the inflectional suffix “-s” is added. Cluster analysis,

commonly called clustering, encompasses categorizing a collection of

objects so that objects within the same group, known as clusters, exhibit

more significant similarity to one another than to objects in other clusters.

This analytical technique constitutes a fundamental aspect of exploratory

data mining and serves as a standard procedure for statistical data analysis.

Cluster analysis finds application in various domains, including machine

learning, pattern recognition, image analysis, information retrieval, and

bioinformatics.

Collaborative filtering (CF) is widely employed in multiple recommender

systems. CF can be classified into two categories: extensive and restricted.

Collaborative filtering generally involves extracting information or patterns

through the interactions among various agents, perspectives, and data

sources. Typically, the implementation of collaborative filtering entails the

utilization of large-scale datasets. It has found applications in diverse

domains such as sensing and monitoring data (e.g., mineral analysis,

environmental sensing employing multiple sensors), financial data (e.g.,

financial institutions aggregating data from various financial sources), and

e-commerce and web applications focused on user information.

Performance Analysis
The Collaborative Filtering (CF) method encompasses various strategies,

such as item-based, user-based, and cluster-based approaches, which

operate on the premise that individuals who have demonstrated agreement

in the past are likely to continue exhibiting understanding. In item-based

CF, the algorithm generates recommendations by identifying items similar

to the user's previous choices. The CF methodology comprises two distinct

stages: clustering and collaborative filtering.

Nevertheless, it is essential to note that the similarity calculation between

each pair of users or services may prove to be computationally intensive,

potentially surpassing the available processing capacity. Consequently, the

timely relevance and feasibility of service recommendations based on

similar users or services may be compromised. Furthermore, conventional

CF algorithms often evaluate all services during the computation of service

rating similarity, even though most of these services exhibit dissimilarities

to the target service. Considering these different services in the rating

calculation can impact the accuracy of the predicted ratings. For an

overview of the performance analysis, please refer to Figure 3.

Figure 3. Performance analysis diagram

Clustering-based Collaborative Filtering (ClubCF) has been incorporated

into the framework. Using the Markov chain model, it is feasible to

distinguish between images based on unique keywords. The system

retrieves and displays all associated images when a user searches for a

specific keyword. The comparison between the current and proposed

systems is illustrated in Figure 4.

Figure 4. Existing vs. proposed system

It is possible to decrease response time and expedite decision-making.

Creating optimal recommendations from an extensive assortment of

services can be challenging. A simple solution is to reduce the number of

services requiring real-time processing. In addition, the accuracy of

recommendations based on user ratings can be enhanced by giving similar

services within a cluster a greater weight than other services.

Role of Python in Hierarchical Clustering and Filtering

Python is a flexible programming language that is frequently used for big

data analytics and machine learning tasks, making it an ideal choice for

hierarchical clustering and filtering algorithms (Perret et al., 2019). Python

can be used to implement the hierarchical clustering algorithm in the

research endeavor. Numerous Python libraries and frameworks, including

scikit-learn and SciPy, provide efficient and versatile clustering analysis

tools. These libraries provide ready-to-use hierarchical clustering functions

and methods and can manage large-scale datasets, which are both necessary

for big data analytics. Python can also be used to implement the research-

described filtering strategies (de Aguiar Neto et al., 2020). It is possible to

implement collaborative filtering algorithms using Python's libraries and

frameworks, such as the clustering-based collaborative filtering technique

mentioned. NumPy and other robust Python numerical computation utilities

facilitate the efficient data manipulation and processing necessary for

filtering operations (Subasi, 2020).

In addition, Python can be used to construct the segmentation method and

CF algorithm. The array slicing capabilities of Python can be used to

accomplish the goal of the slicing technique, which is to reduce

computation time. In addition, Python's adaptability and extensive library

ecosystem provide numerous implementation options for the CF algorithm,

allowing the research to utilize Python's vast resources. Python is

advantageous for implementing the hierarchical clustering and filtering

methods described in the research paper. Its libraries, frameworks, and data

processing capabilities make it ideal for big data analytics, enhancing search

engine relevance, and enhancing user experience (Stančin & Jović, 2019).

The built-in data structures of Python, such as hierarchical clustering and

filtering, play an important role in data analysis and processing, including

big data analytics. These data structures facilitate the efficient storage,

manipulation, and retrieval of data, thereby enhancing Python's ability to

perform these tasks.

• Lists: Lists are one of Python's most versatile and pervasive data

structures. They allow for the storage of heterogeneous data types and

provide flexibility for dynamic resizing, appending, and indexing

 operations. In the context of hierarchical clustering and filtering, data

elements, feature vectors, and cluster assignments can be retained and

organized using lists.

• Arrays: Python arrays, in particular those provided by the NumPy

library, offer efficient storage and operations for large homogenous

numerical datasets. They enable vectorized operations, which

facilitates the high-performance computations required to process vast

quantities of data. In algorithms for hierarchical clustering and

filtering, arrays are commonly used to store numerical data.

• Dictionaries: The key-value mapping provided by dictionaries

enables efficient data access and retrieval based on unique keys. They

facilitate the storage and retrieval of metadata or additional data

associated with data points, clusters, or filtering criteria. Dictionaries

can facilitate the organization and retrieval of pertinent data in

hierarchical clustering and filtering processes.

• Sets: Sets are unordered collections of distinct elements that

facilitate efficient membership testing and duplication elimination.

They can be utilized for data preprocessing duties, such as eradicating

redundant data points or filtering out irrelevant elements according to

predetermined criteria. In hierarchical clustering and filtering

algorithms for data cleansing and optimization, sets are advantageous.

• Trees and Graphs: Despite not being built-in, Python's libraries,

such as networkx, provide efficient data structures for representing

trees and graphs. Because they can represent hierarchical relationships

between clusters or data elements, these data structures are

advantageous for hierarchical clustering. Additionally, they can

facilitate graph-based filtering techniques, such as collaborative

filtering based on similarity graphs.

The leveraging python's robust data structures for efficient hierarchical

clustering and filtering in big data analytics is given in Figure 5.

Figure 5. Python's robust data structures

By leveraging these built-in and specialized data structures, Python

enhances the efficiency and effectiveness of hierarchical clustering and

filtering algorithms in big data analytics. The adaptability and capabilities

of these data structures enable the efficient storage, retrieval, and

manipulation of data, thereby facilitating the processing and analysis of

enormous datasets.

CONCLUSION

This chapter focused on the application of hierarchical clustering and

filtering techniques for slicing-based big data analytics. The study

emphasized the importance of advanced techniques such as hierarchical

clustering and collaborative filtering for analyzing vast amounts of data and

augmenting search engine relevance and user experience. The research

offers a data mining-based method for managing vast data sets. The

framework incorporated data acquisition, extraction, analysis, user interest

modelling, and security considerations. It was emphasized that hierarchical

clustering and filtering algorithms can be implemented using Python's

flexibility and vast library resources. The chapter emphasized the

significance of data preprocessing in preparing unstructured data for data

mining analytics. In addition, it introduced the concept of multivariate

content difficulty rating to regulate content difficulty based on user

performance across multiple dimensions.

Clustering and collaborative filtering were proposed as effective solutions to

the problems encountered by recommendation systems when dealing with

large amounts of data. It emphasized the benefits of clustering in terms of

reducing data volume and increasing the accuracy of recommendations

based on user ratings within clusters. In addition, the study evaluated the

effectiveness of collaborative filtering methods and introduced the Cluster-

based Collaborative Filtering (ClubCF) approach, which demonstrated

substantial reductions in online processing computational time.

In addition, the conclusion identified two potential future research avenues.

Inclusion of historical preferences derived from user-centric approaches in

the absence of explicit ratings to generate recommendations. By

emphasizing the significance of hierarchical clustering and filtering

techniques, this article contributes to the field of big data analytics. It

explores various aspects of data preprocessing, content difficulty

classification, and collaborative filtering techniques. The findings and

suggested future research directions offer valuable insights for enhancing

recommendation systems and addressing the challenges posed by big data.

REFERENCES

Akbar, I., Hashem, T., Yaqoob, I., Anuar, N.B., Mokhtar, S., Gani,

A., & Khan, S.A. (2015). The rise of “big data” on cloud

computing: review and open research issues. Inf Syst., 47, 98–115.

Bellogín, A., Cantador, I., Díez, F., Castells, P., & Chavarriaga, E.

(2013). An empirical comparison of social, collaborative filtering,

and hybrid recommenders. ACM Transactions on Intelligent

Systems and Technology, 4(1), 1–37.

doi:10.1145/2414425.2414439

Beyer & Laney. (2012). The importance of “big data”: A

definition. Gartner, Tech. Rep.

BhardwajS.DashS. K. (2015). VDMR-DBSCAN: varied density

Mapreduce DBSCAN. International conference on Big Data

analytics. doi:10.1007/978-3-319-27057-9_10

Borah, B., & Bhattacharyya, D. K. (2008). DDSC: A density

differentiated spatial clustering technique. Journal of Computers,

3(2), 72–79. doi:10.4304/jcp.3.2.72-79

Dai, B. R., & Lin, I.-C. (2012). Efficient Map/Reduce-based

DBSCAN algorithm with optimized data partition. Fifth

International Conference on Cloud Computing.

de Aguiar Neto, F. S., da Costa, A. F., Manzato, M. G., &

Campello, R. J. (2020). Pre-processing approaches for collaborative

filtering based on hierarchical clustering. Information Sciences,

534, 172–191. doi:10.1016/j.ins.2020.05.021

Duan, L., Xu, L., Guo, F., Lee, J., & Yan, B. (2007). A local-

density based spatial clustering algorithm with noise. Information

Systems, 32(7), 978–986. doi:10.1016/j.is.2006.10.006

Elbatta, M. T., Bolbol, R. M., & Ashur, W. M. (2011). A vibration

method for discovering density varied clusters. Int Sch Res Not.

Ester, M., Kriegel, H. P., Sander, J., & Xu, X. (1996). A density-

based algorithm for discovering clusters in large spatial databases

with noise. In 2nd international conference on knowledge discovery

and data mining. Academic Press.

Gaonkar, M.N., & Sawant, K. (2013). AutoEpsDBSCAN:

DBSCAN with Eps automatic for large dataset. Int J Adv Comput

Theory Eng., 2(2), 2319–526.

Hadoop, W. T. (2009). The definitive guide. O'Reilly Media.

Havens, T. C., Bezdek, J. C., Leckie, C., Hall, L. O., &

Palaniswami, M. (2012). Fuzzy c-Means Algorithms for Very

Large Data. IEEE Transactions on Fuzzy Systems, 20(6), 1130–

1146. doi:10.1109/TFUZZ.2012.2201485
He, Y., Tan, H., Luo, W., Feng, S., & Fan, J. (2014). MR-

DBSCAN: A scalable Map Reduce-based DBSCAN algorithm for

heavily skewed data. Frontiers of Computer Science, 8(1), 83–99.

doi:10.1007/s11704-013-3158-3

Ishwarappa, A. J., & Anuradha, J. (2011). A brief introduction on

Big Data 5Vs characteristics and Hadoop technology. Procedia

Computer Science, 48, 319–324. doi:10.1016/j.procs.2015.04.188

Li, H. H., Du, X. Y., & Tian, X. (2009). A review-based reputation

evaluation approach for Web services. Journal of Computer Science

and Technology, 24(5), 893–900. doi:10.1007/s11390-009-9280-x

Liu, P., Zhou, D., & Wu, N. (2007, June). VDBSCAN: varied

density based spatial clustering of applications with noise. In 2007

International conference on service systems and service

management (pp. 1-4). IEEE. doi:10.1109/ICSSSM.2007.4280175

Liu, X., Huang, G., & Mei, H. (2009). Discovering homogeneous

web service community in the user-centric web environment. IEEE

Transactions on Services Computing, 2(2), 167–181.

doi:10.1109/TSC.2009.11

LiuZ.LiP.ZhengY. (2009). Clustering to find exemplar terms for

keyphrase extraction. Proc. 2009 Conf. on Empirical Methods in

Natural Language Processing, 257-266.

doi:10.3115/1699510.1699544

Mahran, S., & Mahar, K. (2008). Using grid for accelerating

density-based clustering. CIT 2008 IEEE international conference.

doi:10.1109/CIT.2008.4594646
Parimala M, Lopez D, Senthilkumar N, (2011). A survey on

density based clustring algorithms for mining large spatial

databases. Int J Adv Sci Technol., 31-59.

Perret, B., Chierchia, G., Cousty, J., Guimaraes, S. J. F., Kenmochi,

Y., & Najman, L. (2019). Higra: Hierarchical graph analysis.

SoftwareX, 10, 100335. doi:10.1016/j.softx.2019.100335

Rajaraman, A., & Ullman, J. D. (2012). Mining of massive

datasets. Cambridge University Press.

Ram, A., Jalal, S., & Kumar, M. (2013). A density based algorithm

for discovering density varied clusters in large spatial databases.

IJCA, 3(6), 1–4.

Stančin, I., & Jović, A. (2019, May). An overview and comparison

of free Python libraries for data mining and big data analysis. In

2019 42nd International convention on information and

communication technology, electronics and microelectronics

(MIPRO) (pp. 977-982). IEEE.

doi:10.23919/MIPRO.2019.8757088

Subasi, A. (2020). Practical machine learning for data analysis

using python. Academic Press.

Wu, X., Zhu, X., & Wu, G. Q. (2014). Data mining with big data.

IEEE Transactions on Knowledge and Data Engineering, 26(1),

97–107. doi:10.1109/TKDE.2013.109

Zeng, W., Shang, M. S., Zhang, Q. M., Lü, L., & Zhou, T. (2010).

Can Dissimilar Users Contribute to Accuracy and Diversity of

Personalized Recommendation. International Journal of Modern

Physics C, 21(10), 1217–1227. doi:10.1142/S0129183110015786

ZhengZ.ZhuJ.LyuM. R. (2013). Service-generated Big Data and

Big Data-as-a-Service: An Overview. Proc. IEEE BigData, 403-

410. doi:10.1109/BigData.Congress.2013.60
 CHAPTER 11

Optimizing Energy Consumption in Wireless

Sensor Networks Using Python Libraries
Jency Jose

CHRIST University (Deemed), India

N. Arulkumar

CHRIST University (Deemed), India

ABSTRACT

Wireless sensor networks (WSNs) are widely utilized in various fields,

including environmental monitoring, healthcare, and industrial automation.

Optimizing energy consumption is one of the most challenging aspects of

WSNs due to the limited capacity of the batteries that power the sensors.

This chapter explores using Python libraries to optimize the energy

consumption of WSNs. In WSNs, various nodes, including sensor, relay,

and sink nodes, are introduced. How Python libraries such as NumPy,

Pandas, Scikit-Learn, and Matplotlib can be used to optimize energy

consumption is discussed. Techniques for optimizing energy consumption,

such as data aggregation, duty cycling, and power management, are also

presented. By employing these techniques and Python libraries, the energy

consumption of WSNs can be drastically decreased, thereby extending

battery life and boosting performance.

INTRODUCTION

Wireless Sensor Networks (WSNs) encounter several obstacles, such as data

transmission and energy consumption restrictions, which immediately affect

operational longevity (Abu-Baker et al., 2023). These obstacles can impede

the extensive deployment and efficacy of WSNs in numerous application

domains. Existing clustering techniques, such as LEACH, cannot resolve

these obstacles effectively. To surmount these limitations and optimize

energy consumption in WSNs, this Study introduces a novel Python library-

based method. By strategically deploying sensor and relay nodes throughout

the network, the proposed approach seeks to improve the energy efficiency

of WSNs (Abdulzahra et al., 2023). The research uses relay sites to extend

the data transmission range while reducing energy consumption. This

method addresses the difficulties posed by limited transmission distance and

the requirement for energy-efficient communication protocols in WSNs.

This Study demonstrates, through extensive testing and analysis, that the

proposed technology reduces energy consumption without compromising

the integrity of data transmission. The findings provide valuable insights

into the optimization potential of Python libraries in wireless sensor

networks. This research has enormous application-domain significance by

contributing to developing energy-efficient wireless sensor networks. It

enables the creation of more dependable and sustainable WSN solutions

that can be deployed in various industries, such as environmental

monitoring, smart cities, agriculture, healthcare, and industrial automation.

WIRELESS SENSOR NETWORKS

Wireless Sensor Networks (WSNs) are battery-operated data collection

devices with sensors and communication capabilities. They are used

extensively in environmental monitoring and industrial automation

applications. Energy consumption is a crucial aspect of WSNs due to the

deployment of numerous tiny sensors that rely on battery power for

extended periods (Jondhale et al., 2022). It is crucial to effectively manage

energy consumption in WSNs to maximize network lifetime and reduce

maintenance costs.

Several protocols have been developed to address the challenges posed by

energy consumption. These protocols aim to reduce network energy

consumption and lifespan. Examples of notable protocols include Sensor

Medium Access Control (SMAC), Energy-Efficient Medium Access

Control (E-MAC), Threshold-sensitive Energy Efficient sensor Network

protocol (TEEN), Berkeley Medium Access Control (B-MAC), and Traffic

Adaptive Medium Access Control (T-MAC).

Using Python libraries to optimize the energy consumption of wireless

sensor networks presents the following difficulties.

• Developing energy-saving techniques and algorithms that do not

degrade network performance.

• Ensuring accurate and timely data acquisition by balancing the trade-

off between transmission quality and energy consumption.

• Managing node sleep-wake cycles and duty cycling strategies for

energy conservation and network upkeep.

• The optimizing data transmission paths by implementing efficient

routing protocols for energy awareness.

• As the number of sensor nodes increases, the network must address

scalability issues while maintaining energy efficiency.

By addressing these challenges and leveraging the capabilities of Python

libraries, researchers can further refine the energy consumption of wireless

sensor networks, leading to more sustainable and efficient deployments.

LITERATURE REVIEW

Specific constraints and needs determine the protocol choice for precision

agriculture in wireless sensor networks—variables like transmission range,

data rate, and deployment density significantly impact protocol selection.

Several protocols have been proposed to optimize energy consumption in

wireless sensor networks. SMAC (Sensor-MAC) is a protocol that uses a

synchronized sleep schedule. By alternating between active and inactive

states, SMAC reduces passive listening and overhearing, conserving energy

(Chen et al., 2020).

E-MAC (Energy-efficient MAC) is another protocol that modifies node duty

cycles in response to traffic demand (Soni & Selvaradjou, 2020). This

adaptive approach helps conserve energy in wireless sensor networks. The

TEEN (Threshold-sensitive Energy Efficient sensor Network) protocol is a

threshold-based method created to minimize unnecessary data transfer and

save energy. By establishing appropriate thresholds, TEEN decreases the

energy consumption of wireless sensor networks.

B-MAC (Battery MAC) employs a randomized listening schedule and duty

cycling to reduce idle listening and overhearing, thus conserving energy

(Bria, 2019; Roopitha & Pai, 2020). The T-MAC (Traffic-aware MAC)

protocol modifies the responsibility cycle of nodes based on their traffic

volume in a dynamic fashion. This adaptive mechanism optimizes the

energy consumption of sensor networks (Kim, 2019).

Several obstacles exist when refining energy consumption in wireless sensor

networks utilizing Python libraries. Compatibility issues between the

selected protocols and the employed Python libraries, efficient

implementation of the protocols in Python, and the requirement for accurate

energy modeling and measurement techniques within the Python

environment may be among these obstacles. These obstacles must be

overcome to optimize the energy consumption of wireless sensor networks

with Python libraries.

Problem Identified From Literature Review

The existing energy-efficient routing protocols are not entirely effective

based on parameters and factors for the dynamic environment in WSN. The

efficiency of the current protocols can be enhanced in terms of being more

adaptive and dynamic to sustain the node's lifetime, as they currently do not

meet the energy efficiency standards for WSN in precision agriculture.

RESEARCH GAP IDENTIFIED

• Fixed cluster heads consume more energy to transmit data to the sink

node.

• Sleep/wake cycles are not regulated in the current scenario for data

transmission.

• We need a solution to adapt to the change in the field size because

the additional sensors lead to multi-path routing.

 • Existing protocols need improvement based on various unpredictable

parameters and factors.

• Improper supplement of factors to the crops to be monitored to

improve the quantity and quantity of the crop

Problem Statement

The existing energy-efficient routing protocols are not entirely effective

based on parameters and factors for the dynamic environment in WSN. The

efficiency of the current protocols can be enhanced in terms of being more

adaptive and dynamic to sustain the node's lifetime, as they currently do not

meet the energy efficiency standards for WSN in precision agriculture.

OBJECTIVES OF THE STUDY

Primary Objectives

• To design a framework that improves the energy efficiency of WSN,

which reduces energy consumption and thus increases the lifetime of

the network

• To compare the performance of the proposed model with the existing

models

Secondary Objectives

• To implement the proposed model in the field of WSN for precision

agriculture

• To perform experiments on a diverse set of large-scale scenarios

METHODOLOGY

To optimize the energy consumption of wireless sensor networks (WSNs), a

thorough methodology was developed. Identifying the research voids in

enhancing the energy efficacy of WSNs was the initial step. This analysis

assisted in comprehending the extant obstacles and areas requiring

additional research. Using this information as a foundation, the next phase

centered on the development of a WSN specifically tailored for precision

agriculture. This required the meticulous selection and deployment of

sensors in strategic locations, taking into account crop types, terrain, and

environmental conditions, among other factors.

After designing the WSN, the model was implemented and data collection

began. Utilizing Python libraries streamlined the process and facilitated

efficient data management. The collected data were then uploaded to a

multi-tenant platform, enabling simple access and analysis by multiple

parties. This platform not only enabled real-time monitoring, but also

facilitated the energy efficiency comparison of various WSN protocols.

Extensive testing and experimentation were conducted to validate and

enhance the model. The collected data was analyzed to assess the efficacy

of various protocols and identify areas for enhancement. The data analysis

tools provided by Python libraries enabled researchers to extract meaningful

insights and draw accurate conclusions.

On the basis of the findings, an energy efficiency framework was proposed.

This framework incorporated the knowledge acquired in the preceding

stages and offered guidelines for optimizing energy consumption in WSNs.

It included sensor placement, data transmission protocols, and power

management strategies, among others. By implementing this framework,

researchers and practitioners in the field of wireless sensor networks could

increase the energy efficiency of their systems, thereby decreasing

operational expenses and extending the network's lifecycle.

In conclusion, the above-described methodology addressed the optimization

of energy consumption in wireless sensor networks. This study sought to

contribute to the development of more sustainable and efficient WSN

solutions by identifying research voids, designing a WSN for precision

agriculture, comparing protocols, and proposing a framework for energy

efficiency. Utilizing Python libraries throughout the process facilitated data

manipulation, analysis, and model enhancement, demonstrating the strength

of programming tools in advancing research in this field. The complete

methodology is given below.

• Identify the research gaps in improving the energy efficiency of a

WSN

• Designing a WSN for precision agriculture

• Implement the model and collect data into a multi-tenant platform

• Compare the energy efficiency of WSN protocols

• Validate and improve the model

• Propose the framework for energy efficiency

LIMITATIONS OF LEACH PROTOCOL

Its random cluster structure and cluster head rotation may help evenly

distribute energy consumption among nodes. This is particularly crucial in

farming environments where node failure due to weather, pests, or other

factors is widespread. Due to the frequent transmission of control messages

between nodes, the LEACH protocol incurs significant communication

overhead, reducing network performance. The LEACH protocol uses a

randomized technique to create clusters, which may result in some nodes

being selected as cluster leaders more often than others, resulting in uneven

energy usage and a shorter network lifetime (Daanoune et al., 2021). It is

not scalable since the protocol requires centralized control and

communication, which may result in a bottleneck as the number of nodes

increases (Liang, 2019).

IMPORTANCE OF RELAY NODES

Relay nodes are essential in wireless sensor networks (WSN). Relay nodes

are indispensable for extending the network's data transmission range and

conserving energy (Su et al., 2019). By strategically positioning cluster

heads and employing directional antennas, relay nodes in WSNs enable

more reliable data transmission over greater distances with reduced energy

consumption. The following are the importance of relay nodes in wireless

sensor networks (WSNs):

 • Network data transmission range depends on cluster head

arrangement

• Data transmission range may be increased by carefully positioning

cluster heads in areas with a high node density or a clear line of sight

to the base station

• It can relay data between out-of-range nodes, extending the network's

data transmission range and reducing energy consumption

• This strategy optimizes the energy consumption of WSNs while

maintaining a reliable data transmission range

• Relay nodes can be used to extend the network's data transmission

range and reduce energy consumption by employing directional

antennas to increase the broadcast range

ENERGY CONSUMPTION OPTIMISATION IN WSN USING

PYTHON

There are numerous ways to reduce the energy consumption of Wireless

Sensor Networks (WSN). Increasing the LEACH protocol's data

transmission range is one way to improve energy efficiency (Bharany et al.,

2021). However, this increase in transmission capacity must be carefully

weighed against energy consumption and network lifespan considerations.

Adding more cluster leaders to the network is an additional strategy. Cluster

heads capture and transmit data from member nodes to the base station. The

network's transmission range can be increased by strategically placing

cluster heads in areas with a high node density or a direct line of sight to the

base station.

Additionally, directional antennas can be advantageous for extending

broadcast range while simultaneously decreasing transmission power. In

addition, relay nodes can facilitate data transmission between out-of-range

nodes, thereby improving network efficiency. However, it is necessary to

evaluate the benefits and drawbacks of each strategy in the context of the

particular wireless sensor network application, considering the challenges

associated with optimizing energy consumption (Thiagarajan, 2020).

Figure 1 presents a general sequence diagram illustrating the data

transmission process in a WSN. The diagram encompasses the interactions

among the Sensor Node, Base Station, and Relay Node, providing an

overview of the data flow within the network.

Figure 1. Sequence diagram: WSN with sensor and relay nodes

Figure 2. Proposed model with relay and sensor nodes

Steps to Implement Relay Nodes in WSN

The following steps are to implement relay nodes in WSN (Chauhan &

Soni, 2021).

• Locate nodes out of range

• Place relay nodes strategically to efficiently transmit data between

out-of-range nodes

• Identify the communication protocols to receive and transfer data

• Assess relay node performance

• Test and evaluate to optimize relay node locations to improve

network performance

The flow diagram depicts the functioning of a wireless sensor network in

Figure 2, including Sensor Nodes, Relay Nodes, and Base Stations.

 • A SensorNode, represented by the color blue, initiates data

transmission to a RelayNode and waits for an acknowledgment (ACK)

in return.

• The RelayNode, depicted in red, receives the data from the

SensorNode, forwards it to the base station, and waits for an

acknowledgment (ACK) from the base station. Subsequently, the

RelayNode sends an ACK back to the SensorNode.

• The base station receives the data from the RelayNode and responds

with an acknowledgment (ACK).

ADAPTIVE DUTY CYCLING WITH RELAY NODES

Duty cycling is a strategy for conserving energy that involves the

intermittent deactivation of node radio transceivers (Khriji et al., 2022).

However, data delays or loss may result if neighboring nodes are

unavailable for relaying. To address this issue, strategically locating network

relay nodes can assist in overcoming connectivity issues. These relay nodes

function as intermediaries, transmitting data between out-of-range nodes.

Using relay nodes increases network connectivity and decreases data loss

caused by duty cycling (Sarwar et al., 2020).

When relay nodes are in use, adaptive duty cycling can be used to optimize

energy consumption further. This technique allows nodes to dynamically

adjust their duty cycle based on the proximity of relay nodes. When a relay

node is present, the duty cycle of a node can be decreased to conserve

energy. If a relay node is unavailable, its duty cycle can be increased to

maintain network connectivity. It is possible to balance network

connectivity and energy conservation by integrating duty cycling and relay

nodes. This strategy extends the network's lifetime and reduces the nodes'

aggregate energy consumption (Abdalkafor & Aliesawi, 2022).

Figure 3. Implementation of the nodes and relay levels

In Figure 3, sensor node and relay node coordinates are generated arbitrarily

in a 2D plane representing an agricultural field. Also specified is the

maximum transmission range for sensor and relay nodes. Based on their

relative positions and transmission ranges, the code identifies all feasible

data transmission lines between sensor nodes and relay nodes and between

them. These data transmission paths are saved as tuples comprising the

coordinates of the transmitting nodes. The code then depicts the nodes and

data transmission lines using Matplotlib and displays the resulting graph.

This code is intended to simulate a wireless sensor network employed in

precision agriculture, in which relay nodes extend the transmission range of

sensor nodes, enabling them to cover a more significant portion of the

agricultural field.

Figure 4. Energy consumption using relay nodes

Figure 5. Total energy consumption without using relay nodes
Figure 4 depicts the Energy Consumption of a Python system built with

Relay Nodes. Figure 5 illustrates the system's Total Energy Consumption

without Relay Nodes, also implemented using Python.

CONCLUSION

This research work describes a novel Python library-based method for

maximizing energy efficiency in Wireless Sensor Networks (WSN). The

proposed method strategically employs sensor and relay nodes to improve

energy efficiency. The way overcomes the limitations of existing clustering

techniques, such as LEACH, by extending the data transmission range and

reducing relay sites' energy consumption. The research demonstrates

through extensive testing and analysis that the proposed technology

effectively reduces energy consumption while preserving the integrity of

data transmission. The findings highlight the potential of Python libraries

for optimizing WSNs and their implementations in numerous industries,

including environmental monitoring, smart cities, agriculture, healthcare,

and industrial automation. The Study emphasizes the importance of relay

nodes in WSNs for extending the data transmission range and conserving
energy. Also discussed is the significance of adaptive duty cycling in

conjunction with relay nodes for optimizing energy consumption. This

Study compares the energy consumption of WSNs with and without relay

nodes and provides implementation guidelines for relay node integration

into WSNs. Results indicate that integrating relay nodes into the network

significantly reduces energy consumption, highlighting the energy-

optimization benefits of relay node integration in WSNs. In the given

scenario, the network's total energy consumption is 67.86 when relay nodes

are not used, which is higher than the energy consumption with relay nodes.

This indicates that incorporating relay nodes in the wireless sensor network

(WSN) assists in optimizing energy consumption by the sensor nodes in the

network.

REFERENCES

Abdalkafor, A. S., & Aliesawi, S. A. (2022, October). Data

aggregation techniques in wireless sensors networks (WSNs):

Taxonomy and an accurate literature survey. In AIP Conference

Proceedings (Vol. 2400, No. 1, p. 020012). AIP Publishing LLC.

Abdulzahra, A. M. K., Al-Qurabat, A. K. M., & Abdulzahra, S. A.

(2023). Optimizing energy consumption in WSN-based IoT using

unequal clustering and sleep scheduling methods. Internet of

Things, 22, 100765. doi:10.1016/j.iot.2023.100765

Abu-Baker, A., Shakhatreh, H., Sawalmeh, A., & Alenezi, A. H.

(2023). Efficient Data Collection in UAV-Assisted Cluster-Based

Wireless Sensor Networks for 3D Environment: Optimization

Study. Journal of Sensors, 2023, 2023. doi:10.1155/2023/9513868

Bharany, S., Sharma, S., Badotra, S., Khalaf, O. I., Alotaibi, Y.,

Alghamdi, S., & Alassery, F. (2021). Energy-efficient clustering

scheme for flying ad-hoc networks using an optimized LEACH

protocol. Energies, 14(19), 6016. doi:10.3390/en14196016

BriaR.WahabA.AlaydrusM. (2019, October). Energy efficiency

analysis of TEEN routing protocol with isolated nodes. In 2019

Fourth International Conference on Informatics and Computing

(ICIC) (pp. 1-5). IEEE. doi:10.1109/ICIC47613.2019.8985668

Chauhan, V., & Soni, S. (2021). Energy aware unequal clustering

algorithm with multi-hop routing via low degree relay nodes for

wireless sensor networks. Journal of Ambient Intelligence and

Humanized Computing, 12(2), 2469–2482. doi:10.1007/s12652-

020-02385-1

Chen, S., Zhang, L., Tang, Y., Shen, C., Kumar, R., Yu, K., Tariq,

U., & Bashir, A. K. (2020). Indoor temperature monitoring using

wireless sensor networks: A SMAC application in smart cities.

Sustainable Cities and Society, 61, 102333.

doi:10.1016/j.scs.2020.102333

Daanoune, I., Abdennaceur, B., & Ballouk, A. (2021). A

comprehensive survey on LEACH-based clustering routing

protocols in Wireless Sensor Networks. Ad Hoc Networks, 114,

102409. doi:10.1016/j.adhoc.2020.102409

Jondhale, S. R., Maheswar, R., Lloret, J., Jondhale, S. R.,

Maheswar, R., & Lloret, J. (2022). Fundamentals of wireless sensor

networks. Received Signal Strength Based Target Localization and

Tracking Using Wireless Sensor Networks, 1-19.

Khriji, S., Chéour, R., & Kanoun, O. (2022). Dynamic Voltage and

Frequency Scaling and Duty-Cycling for Ultra Low-Power

Wireless Sensor Nodes. Electronics (Basel), 11(24), 4071.

doi:10.3390/electronics11244071
Kim, D. S., Tran-Dang, H., Kim, D. S., & Tran-Dang, H. (2019).

Mac protocols for energy-efficient wireless sensor networks.

Industrial Sensors and Controls in Communication Networks: From

Wired Technologies to Cloud Computing and the Internet of

Things, 141-159.

Liang, H., Yang, S., Li, L., & Gao, J. (2019). Research on routing

optimization of WSNs based on improved LEACH protocol.

EURASIP Journal on Wireless Communications and Networking,

2019(1), 1–12. doi:10.1186/s13638-019-1509-y

RoopithaC. H.PaiV. (2020, January). Survey On Media Access

Control Protocols In Wireless Sensor Network. In 2020 Fourth

International Conference on Inventive Systems and Control

(ICISC) (pp. 163-168). IEEE.

doi:10.1109/ICISC47916.2020.9171085

Sarwar, S., Sirhindi, R., Aslam, L., Mustafa, G., Yousaf, M. M., &

Jaffry, S. W. U. Q. (2020). Reinforcement learning based adaptive

duty cycling in LR-WPANs. IEEE Access : Practical Innovations,

Open Solutions, 8, 161157–161174.

doi:10.1109/ACCESS.2020.3021016

Soni, G., & Selvaradjou, K. (2020). Performance evaluation of

wireless sensor network MAC protocols with early sleep problem.

International Journal of Communication Networks and Distributed

Systems, 25(2), 123–144. doi:10.1504/IJCNDS.2020.108875

Su, Y., Lu, X., Zhao, Y., Huang, L., & Du, X. (2019). Cooperative

communications with relay selection based on deep reinforcement

learning in wireless sensor networks. IEEE Sensors Journal,

19(20), 9561–9569. doi:10.1109/JSEN.2019.2925719

Thiagarajan, R. (2020). Energy consumption and network

connectivity based on Novel-LEACH-POS protocol networks.

Computer Communications, 149, 90–98.

doi:10.1016/j.comcom.2019.10.006
 CHAPTER 12

An Exploratory Study of Python's Role in the

Advancement of Cryptocurrency and
Blockchain Ecosystems
Agrata Gupta

CHRIST University (Deemed), India

N. Arulkumar

https://orcid.org/0000-0002-9728-477X

CHRIST University (Deemed), India

ABSTRACT

Blockchain is the foundation of cryptocurrency and enables decentralized

transactions through its immutable ledger. The technology uses hashing to

ensure secure transactions and is becoming increasingly popular due to its

wide range of applications. Python is a performant, secure, scalable

language well-suited for blockchain applications. It provides developers free

tools for faster code writing and simplifies crypto analysis. Python allows

developers to code blockchains quickly and efficiently as it is a completely

scripted language that does not require compilation. Different models such

as SVR, ARIMA, and LSTM can be used to predict cryptocurrency prices,

and many Python packages are available for seamlessly pulling

cryptocurrency data. Python can also create one's cryptocurrency version,

as seen with Facebook's proposed cryptocurrency, Libra. Finally, a versatile

and speedy language is needed for blockchain applications that enable chain

addition without parallel processing, so Python is a suitable choice.

INTRODUCTION
In recent years, blockchain technology, the foundation of cryptocurrencies,

has attracted a great deal of interest. It functions as an immutable

distributed ledger that enables decentralized transactions. Transactions are

distributed across an interconnected network of nodes and stored in

segments. The owner's digital signature on the ledger provides security and

prevents tampering. Consequently, cryptocurrency utilizes Blockchain

technology, functioning as a decentralized and secure medium of exchange

independent of centralized authorities like governments and institutions.

Coin ownership records are stored in a database that employs cryptography

to ensure transaction security, regulate coin production, and validate

ownership transfers.

Current monetary systems are exposed to significant hazards and require

sophisticated financial management. The fact that industries such as

healthcare, supply chain, and the Internet of Things (IoT) use centralized

processes makes decentralization essential. Numerous industries, such as

agriculture, banking, healthcare administration, IoT, and governance, have

the potential to be revolutionized by blockchain technology. It is the

technology of the future, and nations including Japan, Germany, and France

have already employed it in a variety of fields. The Indian state of Andhra

Pradesh has already implemented blockchain technology in agribusiness.

Despite Bitcoin's emergence in 2009, the financial applications of

currencies and blockchain are still developing and will continue to do so.

Due to its high efficacy, robust security, and scalability, Python is the

preferable programming language for Blockchain applications. It has a

moderate learning curve, allowing programmers to acquire proficiency

rapidly. Even inexperienced developers are able to contribute rapidly to

blockchain projects. Python is a scripting language, unlike C++, which

eliminates the need for compilation and increases developer productivity.

Python also provides a large number of gratis programming utilities. Python

makes it possible for developers to construct Blockchain applications with

minimal code. In addition, Python facilitates the analysis, visualization, and

forecasting of the prices and volatility of cryptocurrencies. Various models,

including SVR, ARIMA, and LSTM, can be utilized to predict the price of

cryptocurrencies.
However, the volatility and unpredictability of cryptocurrencies can make it

difficult to make accurate predictions. Python packages such as Alpha

Vantage, Pandas DataReader, YFinance, CrytpoCompare, IEX Cloud, and

Twelve Data make it simple to retrieve cryptocurrency data. Additionally,

Python enables us to construct our own cryptocurrency, as Facebook's

recent introduction of Libra, which ignited global outrage, demonstrates.

When developing a Blockchain application, it is crucial to enable anyone to

contribute to the chain without requiring parallel transaction processing. For

this reason, Python, a rapid and flexible programming language, is an

excellent option. This article explores Python's key roles in Blockchain

Technology and Cryptocurrency.

LITERATURE REVIEW

The blockchain technology was described in 1991 by the research scientist

Stuart Haber and W. Scott Stornetta. They wanted to introduce a

computationally practical solution for time-stamping digital documents so

that they could not be backdated or tampered. In 1992, Merkle Trees were

incorporated into the design, which made blockchain more efficient by

allowing several documents to be collected into one block. In 2004,

computer scientist and cryptographic activist Hal Finney introduced a

system called Reusable Proof of Work (RPoW) as a prototype for digital

cash. It was a significant early step in the history of cryptocurrencies. In

2008, Satoshi Nakamoto conceptualized the theory of distributed

blockchains. The words block and chain were used separately in Satoshi

Nakamoto's original paper but were eventually popularized as a single

word, the Blockchain, by 2016.

In recent time, the file size of cryptocurrency blockchain containing records

of all transactions that occurred on the network has grown from 20 GB to

100 GB (Larrier, 2021). On the other hand, the idea for cryptocurrency first

began in the late 1980's, the idea was for a currency that could be sent

untraceable and in a manner that did not require centralized entities (i.e.,

Banks). In 1995, American cryptographer David Chaum implemented an

anonymous cryptographic electronic money called Digicash. Bit Gold,

often called a direct precursor to Bitcoin, was designed in 1998 by Nick

Szabo. It required a participant to dedicate computer power to solving

cryptographic puzzles, and those who solved the puzzle received the reward

(Aggarwal & Kumar, 2021). The focus on cryptocurrencies in the finance

and banking sectors is gaining momentum.

The Bitcoin system only provides eventual consistency. For everyday life,

the time to confirm a Bitcoin transaction is prohibitively slow (Decker,

Seidel, & Wattenhofer, 2014). In this paper we propose a new system, built

on the Bitcoin block chain, which enables strong consistency. Our system,

Peer Census, acts as a certification authority, manages peer identities in a

peer- to-peer network, and ultimately enhances Bitcoin and similar systems

with strong consistency. Our extensive analysis shows that Peer Census is in

a secure state with high probability. We also show how Discoin, a Bitcoin

variant that decouples block creation and transaction confirmation, can be

built on top of Peer Census, enabling real-time payments. Unlike Bitcoin,

once transactions in Discoin are committed, they stay committed.

A purely peer-to-peer version of electronic cash would allow online

payments to be sent directly from one party to another without going

through a financial institution. Digital signatures provide part of the

solution, but the main benefits are lost if a trusted third party is still required

to prevent double-spending. A solution is proposed to the double-spending

problem using a peer-to-peer network. the objective of this research paper is

to focuses on working framework of Block chain Technology and enlightens

the security of Block chain technology through vulnerability (Nakamoto,

2008; Massias et al., 1999).

One of the first studies to attempt to predict cryptocurrency using non-linear

neural networks—conducted by Maiti, Vyklyuk, and Vukovic and published

in 2020—found that the non-linear models were preferable to linear models

as the former's error was lower than the latter, giving more accurate

predictions. This study also found that some elements of the

cryptocurrency's historical data were not helpful in prediction, such as the

volume (Maiti et al., 2020).

Another study released in 2022 to expand the prediction beyond merely

yielding a direction; instead of reporting whether the price would go up or

down, the model would report “the magnitude of the price change” (Critien

et al., 2022). The study leveraged not only historical price data for Bitcoin,

but also Twitter posts to glean public sentiment about the currency. An

important distinction of this study is that it used a bidirectional LSTM,

which is a model composed of two LSTMs: one for historical price data and

one for Twitter posts. Its accuracy of 64.18% was achieved by using 450

days of historical data.

Furthermore, a 2022 study by Sarkodie, Ahmed, and Owusu used COVID-

19 data, such as the number of cases and deaths, to determine whether

cryptocurrencies such as Bitcoin would rise or fall in price. The study made

use of the Romano–Wolf algorithm to test this correlation. In particular, it

deemed Bitcoin to be very highly correlated with deaths due to COVID-19,

correspondingly fluctuating 90.50% of the time. Lastly, a 2019 study by

Azari utilized an autoregressive integrated moving average (ARIMA)

approach to predict the price of Bitcoin. In this study, the model's effective-

ness was evaluated by examining the mean-squared error; the findings

reported a very low level of prediction error (Critien, et al., 2022).

Bitcoin was the first cryptocurrency introduced in the market by

pseudonymous entity named Satoshi Nakamoto in 2009 but the real market

for cryptocurrency started in 2013. Today there is more than 1650 different

altcoins in the market. Bitcoin, the pioneer cryptocurrency, was conceived

in 2008 by an individual or a group of researchers under the pseudonymous

of Nakamoto (2008) and implemented in 2009 as an electronic payment

system via a breakthrough application of blockchain technology.

THE FACTORS AND PARAMETERS

The study looks towards creating a comprehensive framework for predicting

cryptocurrency prices using deep learning models. Several factors influence

cryptocurrency prices. Market demand and supply dynamics play a crucial

role, with increased demand driving prices up and oversupply causing them

to decline. Investor sentiment and market psychology can significantly

impact prices, with positive news and confidence-boosting prices, while

negative developments and uncertainties can lead to declines. The

regulatory environment, including government policies and regulations, can

have a profound effect on cryptocurrency prices, as supportive regulations

and institutional adoption tend to drive prices higher, while restrictive

measures or bans can lead to price drops.

Technological advancements, upgrades, and innovations in blockchain

technology or specific cryptocurrencies also influence prices, as improved

features and utility attract more demand. Market liquidity, the ease of

buying and selling cryptocurrencies, can affect prices, with higher liquidity

generally leading to more stable prices. Moreover, market manipulation,

integration and adoption into mainstream systems, macroeconomic factors,

and global events all contribute to the complex dynamics that shape

cryptocurrency prices. The parameters like MAE, RMSE, MSE, and MAPE

are commonly used to evaluate the performance of predictive models; it's

important to note that the choice of evaluation metrics may depend on the

specific context and requirements of the cryptocurrency price prediction

task or dataset. These metrics assess different aspects of prediction accuracy

and can help identify the best-fit model.

PROBLEM STATEMENT

The accurate prediction of cryptocurrency prices and volatility poses a

significant challenge in the market, with existing methods showing limited

accuracy. To address this challenge, this research aims to develop a

comprehensive framework using deep learning models to enhance the

accuracy of cryptocurrency price and volatility prediction, evaluate

volatility levels of different cryptocurrencies, and empower investors to

make informed decisions in the dynamic and volatile cryptocurrency

market.

OBJECTIVES OF THE STUDY

Primary Objective
 • To develop an integrated framework of multiple deep learning

models to accurately predict cryptocurrency prices, identify the most

effective deep learning approach for enhanced accuracy, and forecast

the volatility of various cryptocurrencies

Secondary Objectives

• To evaluate and compare the effectiveness of Deep Learning

techniques in cryptocurrency price and volatility prediction against

traditional statistical methods

• To develop a live dashboard that provides real-time tracking of daily

cryptocurrency prices, incorporating the integrated framework of

multiple Deep Learning models.

SCOPE OF THE STUDY

• The aim of this research is to develop a comprehensive framework

that utilizes deep-learning models for predicting cryptocurrency prices

and volatility more effectively.

• The study aims to provide investors with more accurate predictions,

enabling them to make informed investment decisions and potentially

increase their profits.

• Additionally, the research will focus on volatility prediction, aiding

investors in making wise investment choices based on the expected

fluctuations in cryptocurrency prices.

METHODOLOGY

The section that follows describes the research methodology used to achieve

the primary and secondary goals of this study, which is to develop a

comprehensive framework using deep learning models.

1. Identify and Review: Review the existing models and methods for

predicting the price of the cryptocurrency.

2. Data Collection: Gather historical data on cryptocurrency prices,

including the relevant features such as trading volume, market

capitalization, and other market indicators. This data can be obtained

from various sources like cryptocurrency exchanges, APIs, or online

databases.

3. Data Preprocessing: Clean the collected data by removing any

inconsistencies, missing values, or outliers. Perform necessary

transformations, such as normalization or scaling, to ensure the data is

in a suitable format for modeling.

4. Feature Selection/Engineering: Identify and select the most

relevant features that can potentially influence cryptocurrency prices.

Additionally, create new features or derive meaningful representations

from the existing ones to improve the predictive power of the models.

Consider factors like market sentiment, technical indicators, or

macroeconomic data that might impact cryptocurrency prices.

5. Split the Data: Divide the dataset into training, validation, and

testing sets. The training set will be used to train the prediction models,

the validation set to tune model hyperparameters, and the testing set to

evaluate the final model's performance.

6. Model Selection: Choose appropriate prediction models for

cryptocurrency price forecasting. This can include traditional statistical

models, machine learning algorithms (e.g., linear regression, random

forests, support vector machines), or deep learning models (e.g.,

recurrent neural networks, long short-term memory networks, or

transformer-based models).

7. Model Training: Train the selected models using the training

dataset. Adjust the model parameters or hyperparameters through

techniques like grid search, cross-validation, or Bayesian optimization

to achieve optimal performance.

8. Model Evaluation: Evaluate the trained models using appropriate

evaluation metrics such as MAE, RMSE, MSE, or MAPE. Compare

the performance of different models to select the best-performing one.

9. Model Validation: Validate the selected model using the validation

dataset. Assess its performance on unseen data to ensure its

generalizability and robustness.

10. Prediction and Analysis: Apply the validated model to make

predictions on new or future data. Monitor the predictions and analyze

 the results, comparing them with actual cryptocurrency prices to

assess the accuracy and reliability of the model.

11. Model Refinement: Iterate and refine the model based on the

feedback obtained from prediction analysis. Fine-tune the model

parameters, update the feature selection/engineering process, or

consider ensemble techniques to improve the prediction accuracy

further.

TOOLS

Python Programming Language

Python is a widely used programming language in the field of data science

and has become a popular tool for predicting cryptocurrency prices. Python

offers several libraries, frameworks, and tools that facilitate data analysis,

modelling, and visualization, making it suitable for developing

cryptocurrency price prediction models. Python plays a crucial role in

predicting cryptocurrency prices using deep learning methods. Here's how

Python is involved: Here are some key Python tools commonly used in this

context:

1. Deep Learning Libraries: Python offers popular deep learning

libraries like TensorFlow, Keras, and PyTorch, which provide high-

level APIs for building and training deep neural networks. These

libraries enable the implementation of complex architectures such as

recurrent neural networks (RNNs), long short-term memory networks

(LSTMs), and convolutional neural networks (CNNs), which are

commonly used in cryptocurrency price prediction.

2. Data Pre-Processing: Python, along with libraries like NumPy and

Pandas, is used for data pre-processing tasks. This involves cleaning

and transforming the cryptocurrency price data to prepare it for

training deep learning models. Python's efficient data manipulation

capabilities enable handling large datasets and performing necessary

transformations, such as normalization or scaling.

3. Feature Engineering: Python allows for feature engineering tasks,

where additional relevant features can be derived from the raw data.
 This may include technical indicators, sentiment analysis scores, or

lagged values. Python libraries like Pandas make it easier to create new

features or extract meaningful representations that enhance the

predictive power of deep learning models.

4. Model Development and Training: Python's deep learning

libraries provide a rich set of tools for building and training deep

learning models. Developers can define the architecture of the model,

configure its layers, and set hyperparameters. Python libraries facilitate

the efficient training of models on large datasets using techniques such

as mini-batch processing, regularization, and optimization algorithms.

5. Hyperparameter Tuning: Python libraries, like scikit-learn,

provide tools for hyperparameter tuning, which involves searching for

the optimal configuration of hyperparameters to improve model

performance. Techniques like grid search or random search can be

employed to find the best combination of hyperparameters for deep

learning models.

6. Model Evaluation and Validation: Python enables the evaluation

and validation of deep learning models. Various metrics, such as mean

squared error (MSE), root mean squared error (RMSE), or mean

absolute error (MAE), can be calculated to assess the model's

performance. Python libraries for data visualization, such as Matplotlib

and Seaborn, can be used to visualize model predictions and compare

them with actual cryptocurrency prices.

7. Deployment and Integration: Python allows for easy deployment

and integration of deep learning models into production systems or

trading platforms. Python frameworks like Flask or Django can be

used to develop web-based applications or APIs to serve the trained

models. This enables real-time predictions and integration with other

systems or trading algorithms.

Python's versatility, extensive deep learning libraries, and robust ecosystem

make it a preferred choice for cryptocurrency price prediction using deep

learning methods. It provides a flexible and efficient environment for model

development, training, evaluation, and deployment, facilitating the entire

prediction pipeline.
ANALYZING CRYPTO MARKET DATA WITH PYTHON

The code is separated into six portions: downloading the data and loading

the necessary libraries, exploratory analysis, return series, Volatility,

correlation, and moving average, respectively.

Figure 1 provides the Python code necessary for loading the required

libraries.

Figure 1. Python code: Loading the necessary libraries

For instance, the desired Crypto data for Binance, Bitcoin, Ethereum, and

XRP (Ripple) will be downloaded using their respective ticker symbols:

BNB, BTC, ETH, and XRP (Shrivas, et al., 2022; Jamision et al., 2018).

Figure 2, which presents Python code demonstrating how to convert ticker

arguments to a list of crypto, providing readers with a practical and helpful

example.

Figure 2. Python code: List of crypto as ticker arguments

To observe the fluctuations in the cryptocurrency market during the Covid-

19 pandemic, the analysis begins on January 1, 2020.

Figure 3 presents Python code that demonstrates how to load crypto data

starting from the beginning of the Covid-19 pandemic.

Figure 3. Python code: Loading crypto data from the beginning of the

COVID-19 pandemic

• After inspecting the data, it is determined that there are no missing

values in the data frame. Consequently, the analysis can proceed, but it

is necessary to understand the meaning of the features in the dataset

first (Mohapatra et al., 2019).

• Through exploratory analysis, the time series plots reveal the

fluctuation of stock prices over time. Due to the varying scales, direct

comparison of the charts is not possible. Instead, the cumulative

returns chart will be utilized, providing relative changes.

Figure 4, showcasing the graph with time-series plots, offering a visual

representation of data trends and patterns.

Figure 4. Graph: Time-series plots

 • The returns of an asset are more manageable compared to the price

series, providing a comprehensive overview of the investment

opportunity. Calculating the returns will be done using the pandas

pct_change() function.

Figure 5. Python code: Relative price changes of the cryptocurrencies

Figure 6. Graph: Return plot

Figure 5, which presents the Python code displaying the relative price

changes of cryptocurrencies, offering a practical and insightful approach to

understanding the dynamic nature of these digital assets. Figure 6,

showcasing the Graph with Return Plot, enhances the chapter by providing

visual representation and analysis of the data.

• XRP exhibits the highest volatility, with Binance being the next most

volatile, while Bitcoin demonstrates the lowest volatility. A significant

market crash can be observed across all cryptocurrencies around

March 2020. Volatility, as measured by standard deviation, indicates

the extent of price fluctuations in an asset (Molin & Jee, 2021). Figure

7, which presents the Python code for calculating the volatility and

standard deviation of returns.

Figure 7. Python code: Volatility, standard deviation of the returns

 • XRP (Ripple) exhibits the highest volatility among the four assets,

whereas Bitcoin demonstrates the lowest volatility. Correlation will be

computed based on return data to avoid biased results that can occur

when using raw price data.

• Moving average (rolling average) is used to smooth out short-term

fluctuations to identify long-term trends or cycles.

Figure 8. Python code: Correlation of cryptocurrencies

Figure 8, illustrating how to represent the correlation of cryptocurrencies,

offering a practical and informative guide for analyzing and understanding

the relationships between different digital currencies. The Python code for

computing moving average is presented in Figure 9, providing a helpful

resource for understanding and implementing this calculation.

Figure 9. Python code: Computing moving average

Figure 10. Plotting moving average

• From the chart in Figure 10, it is evident that in mid-May 2021, the

price falls below the 50-day MA, indicating a downward trend.

Similarly, in mid-August, the price rises above the MA, signaling an

upward trend.
CREATING OUR OWN CRYPTOCURRENCY USING
PYTHON

Governments, organizations, and individuals have been adopting blockchain

technology to develop their own cryptocurrencies, aiming to stay relevant in

the evolving landscape. An example that garnered significant attention

worldwide was Facebook's introduction of its cryptocurrency, known as

Libra (Opidi, 2019; Z, 2023; Yogeshwaran, 2019).

Figure 11. Python code: To create crypto coin (fccCoin)

Figure 11 presents the fundamental design of the blockchain code in Python

for creating the cryptocurrency (fccCoin), providing a clear and practical

example for understanding the implementation of a blockchain.

CONCLUSION

This article demonstrates that Python is an excellent language for

Blockchain projects. It offers a gentle learning curve, allowing developers to

master it efficiently. With Python, developers can code a blockchain with

minimal effort, and it simplifies cryptoanalysis. Utilizing Python packages

like Alpha Vantage, Pandas DataReader, YFinance, CryptoCompare, IEX

Cloud, Twelve Data, etc., facilitates seamless retrieval of cryptocurrency

data. Python finds various applications in the cryptocurrency realm,

including the ability to create custom crypto coins. Although an initially

created coin may not meet the current market demands for stability,

security, and user-friendliness, it can be enhanced by adding features for

mining and financial transactions. In conclusion, Python is a versatile and

efficient language.

REFERENCES
Aggarwal, S., & Kumar, N. (2021). History of blockchain-

blockchain 1.0: Currency. Advances in Computers, 121, 147–169.

Critien, J. V., Gatt, A., & Ellul, J. (2022). Bitcoin price change and

trend prediction through twitter sentiment and data volume.

Financial Innovation, 8(1), 1–20.

Decker, C., Seidel, J., & Wattenhofer, R. (2016, January). Bitcoin

meets strong consistency. In Proceedings of the 17th International

Conference on Distributed Computing and Networking (pp. 1-10).

Academic Press.

Jamison, M. A., & Tariq, P. (2018). Five things regulators should

know about blockchain (and three myths to forget). The Electricity

Journal, 31(9), 20–23.

Larrier, J. H. (2021). A Brief History of Blockchain. Transforming

Scholarly Publishing With Blockchain Technologies and AI, 85-

100.

Maiti, M., Vyklyuk, Y., & Vuković, D. (2020). Cryptocurrencies

chaotic co‐movement forecasting with neural networks. Internet

Technology Letters, 3(3), e157.

Massias, H., Avila, X. S., & Quisquarter, J. J. (1999). Design of a

secure timestamping service with minimal trust requirement. In

Symposium on Information theory in the Benelux (pp. 79-86).

Werkgemeenschap voor Informatie-en Communicatietheorie.

Mohapatra, S., Ahmed, N., & Alencar, P. (2019, December).

KryptoOracle: A real-time cryptocurrency price prediction

platform using twitter sentiments. 2019 IEEE international

conference on big data, 5544–5551.

Molin, S., & Jee, K. (2021). Hands-On Data Analysis with Pandas:

A Python data science handbook for data collection, wrangling,

analysis, and visualization. Packt Publishing Ltd.

Nakamoto, S. (2008). Bitcoin: A peer-to-peer electronic cash

system. Decentralized Business Review, 21260.

Opidi, A. (2019, October 8). How to Create Your Own

Cryptocurrency Using Python. freeCodeCamp.org.

https://www.freecodecamp.org/news/create-cryptocurrency-using-

python/

S, R. A. (2023, March 31). What is Blockchain Technology? How

Does Blockchain Work? [Updated]. Simplilearn.com.

https://www.simplilearn.com/tutorials/blockchain-

tutorial/blockchain-technology

Shrivas, M. K., Hiran, K. K., Bhansali, A., & Doshi, R. (2022).

Advancements in Quantum Blockchain With Real-Time

Applications. IGI Global.

YogeshwaranS.KaurM. J.MaheshwariP. (2019, April). Project

based learning: predicting bitcoin prices using deep learning. In

2019 IEEE Global Engineering Education Conference (EDUCON)

(pp. 1449-1454). IEEE.

Z. (2023, February 22). Top 10 Useful Python Packages to Pull

Cryptocurrency Data Seamlessly. Analytics Insight.

https://www.analyticsinsight.net/top-10-useful-python-packages-to-

pull-cryptocurrency-data-seamlessly/
 CHAPTER 13

Exploring the Role of Python in Self-

Supervised Contrastive Learning for
Generating Medical Imaging Reports
Rahul Kumar

CHRIST University (Deemed), India

N. Arulkumar

CHRIST University (Deemed), India

ABSTRACT

This chapter investigates Python's involvement in self-supervised

contrastive learning (SSCL) for medical imagery with report generation.

The research highlights the relevance of SSCL as a method for creating

medical imaging reports and the benefits of implementing it using Python.

The literature review gives a complete overview of SSCL approaches in

medical imaging and shows the advantages of SSCL implementation using

Python libraries such as PyTorch, TensorFlow, and Keras. The study's

methodology describes the research topics, survey design, methods of data

gathering, and analytic procedures. The study named SSCL-GMIR findings

indicate that several practitioners utilize SSCL in medical imaging using

Python modules. This study highlights Python's significance in

implementing SSCL for creating medical imaging report documents,

offering researchers and practitioners a more efficient and effective method

for producing accurate and informative reports and diagnoses.

INTRODUCTION
Self-Supervised Contrastive Learning for Medical Images with Report

Generation heavily relies on implementing Python data structures and

algorithms. Self-Supervised Contrastive Learning is a machine learning

technique that constructs data representations by comparing various

perspectives of the same data, which helps generate relevant reports or

diagnoses in medical imaging. Python data structures such as dictionaries,

lists, and arrays store and process the data. To extract valuable insights from

these structures, it is necessary to use algorithms such as k-means clustering

or PCA. Other techniques, such as gradient descent or backpropagation,

may be used to train a neural network for representation learning (Li

Zhiyuan et al., 2023; Mohan et al., 2023; Upadhye et al., 2023).

Once the data representations have been learned, they can generate reports

or diagnoses. For instance, decision trees or random forests can categorize

images and provide relevant reports. Python's implementation is highly

advantageous for Self-Supervised Contrastive Learning for Medical Images

with Report Generation. Python's data structures and algorithms facilitate

efficient data manipulation, information extraction, model training, and

evaluation, resulting in more accurate and informative reports and

diagnoses (Mishra et al., 2023; Saravanan et al., 2023; Neelakandan et al.,

2022).

LITERATURE REVIEW

Medical imaging is a potent, non-invasive diagnostic tool essential for

diagnosing and treating various diseases because it permits the observation

of the internal structures of human organs, tissues, and bones. SSCL, a

methodology that trains visual representations via unsupervised Learning,

has been a potential method for producing accurate medical reports and

diagnoses. TensorFlow and Keras are some of the tools and frameworks the

popular programming language Python provides for SSCL applications in

medical imaging (Najma et al., 2023). Python's simplicity, readability, and

scalability make it popular among medical imaging academics and industry

professionals. Using Python for SSCL implementation in medical imaging

presents unique problems, such as managing massive datasets and

enhancing algorithm performance, but these obstacles may be solved with

proper planning and design.

Chen et al. (2019) analyzed the challenges of finding enough labeled

medical images to train deep-learning models and the need to include

unlabeled data to improve model performance. The authors propose a

context restoration-based self-supervised learning method to maximize

medical image analysis using unlabeled images. The suggested context

restoration strategy improved classification, localization, and segmentation

performance. The proposed self-supervised learning strategy, which uses

context restoration, may enhance machine learning models for medical

imaging applications by acquiring semantic information.

In a unique self-supervised pre-training strategy for medical picture

analysis, extra pixel-level information and scale information are directly

included in high-level semantics (Zhou et al., 2023). PCRLv2 integrates

multi-scale pixel restoration with siamese feature comparison in a feature

pyramid for 3D medical imaging using a non-skip U-Net and sub-crop. The

system outperforms self-supervised algorithms for segmenting brain

tumors, identifying chest pathologies, detecting lung nodules, and

segmenting abdominal organs, sometimes with little annotations.

CheSS, a publicly available chest X-ray pre-trained model, employs self-

supervised contrastive Learning to learn various radiograph representations

(Cho et al., 2023). Disease categorization, bone suppression, and nodule

creation verify the model trained on 4.8 million CXR pictures. CheSS

enhanced accuracy by 28.5% on the 6-class classification test dataset and

1.3% on the CheXpert dataset. The article concludes that CheSS may help

researchers overcome data imbalance, inadequacy, and medical image-

collecting inaccessibility.

Problem Identified From Literature Review

The review of relevant literature reveals a number of significant findings in

the field of self-supervised contrastive learning for medical image texture

generation. Despite the progress made, the following challenges and

improvement opportunities have been identified:

• Traditional supervised methods rely heavily on manually annotated

data, which can be time-consuming and costly to acquire. The

difficulty lies in reducing this dependence and effectively utilising

unlabelled data to enhance the learning process.

• Integration of diverse multi-organ datasets: In order to increase the

efficacy of learned features, it is essential to integrate diverse multi-

organ datasets with varying staining and resolution characteristics. The

difficulty lies in developing methodologies that can handle the

heterogeneity and complexity of such datasets.

• Clinical metadata, when incorporated into the self-supervised

learning framework, can provide valuable contextual information and

enhance the comprehension of medical images. However, additional

research and the development of techniques to effectively utilise

clinical metadata in the learning process are required.

• Self-supervised learning utilising contrastive learning and clustering

methods has demonstrated superior performance in comparison to

supervised learning, but there is room for improvement. Potentially

enhancing the learning of visual features is the investigation of

additional techniques or novel strategies beyond those currently

employed.

• The effectiveness of attention mechanisms in medical image analysis,

such as the detection of COVID-19, can be further optimised.

Potentially increasing diagnostic precision and robustness is the

investigation of alternative attention mechanisms or the refinement of

the existing ones.

• Exploration of alternative ensemble strategies: The hybrid ensemble

model proposed for detecting COVID-19 has shown promise.

However, alternative ensemble strategies could be investigated to

improve the overall performance of diagnostic models.

• While self-supervised learning has demonstrated promising results, it

is necessary to investigate the potential advantages of using smaller

sample sizes and fewer training steps. For practical implementation, it

 is essential to comprehend the trade-offs between computational

requirements and output quality.

• Enhanced pre-training strategies for medical image classification

Self-supervised learning has proven to be an effective pre-training

strategy for medical image classification. In order to improve the

accuracy of medical image classifiers, future research can concentrate

on refining self-supervised learning techniques and exploring novel

methods for constructing more informative positive pairs for self-

supervised learning.

• The proposed self-supervised learning technique for abnormality

detection in chest radiography has improved performance through

model optimisation and training diversity. To improve its robustness

and generalizability, the model must undergo additional optimisation

and training on a more diverse set of images.

Reducing reliance on manually annotated data, handling diverse multi-

organ datasets, incorporating clinical metadata, refining self-supervised

learning techniques, optimising attention mechanisms, examining

alternative ensemble strategies, analysing computational requirements,

enhancing pre-training strategies for medical image classification, and

enhancing model optimisation and training diversity are some of the

identified problem areas. Addressing these obstacles will advance self-

supervised contrastive learning with texture generation for medical images.

RESEARCH GAP

• If the number of instances is proportional to the intractable number

of classes, labelling unlabelled images will become challenging.

• The existing research with K-Means clustering with contrastive

learning has resulted in a only 3% improvement in accuracy, which

falls short of a perfect 100% improvement.

• A novel approach is needed to enhance the effectiveness of

contrastive learning with text, specifically for the purpose of detecting

abnormalities.

• Despite the vast amount of research on contrastive learning for

detecting abnormalities in text, there is a gap in the literature on

 incorporating the generation of texture to facilitate a seamless

experience for end-users.

RESEARCH QUESTIONS

• What are the core techniques for increasing accuracy of medical

image annotation

• How can medical texture generation be achieved using Self-

Supervised learning or any other methods?

• What additional augmentation technique can be applied to generate

multiple images from the same image?

• What is the importance of improving self-supervised contrastive

learning in healthcare?

THE FACTORS AND PARAMETERS

The study looks towards annotation for medical images using self-

supervised contrastive learning. It also aims to provide the texture

generation of medical images. Data augmentation techniques including

cropping, rotation, flipping, and color jittering are high factors. The

parameters could be the softmax function with vector of numbers as input,

the size of the contrastive loss batch, the learning rate, the quantity of

training data, the size of the embedding space, and the regularization

techniques like dropout, weight decay, or early stopping. The suggested

self-supervised contrastive learning framework for medical image analysis

and texture generation performs more effectively due to these characteristics

and parameters, which are essential in predicting sections of the input data

from other portions of the same input data.

PROBLEM STATEMENT

The lack of accurate and robust labelling for medical images has been a

significant challenge in the field of medical image analysis. To address this

challenge, this research aims to design a self-supervised method and model

that improves the classification of unlabelled medical images. This research

aims to implement texture description about disease and anomalies in

medical images and to perform integration of self-supervised method with

texture generation method. Overall, this research will contribute to the

development of more accurate and reliable medical image analysis methods,

which can help in the diagnosis and treatment of various medical

conditions.

OBJECTIVES OF THE STUDY

Primary Objectives:

• To design a Self-Supervised method and model that improves the

classification of unlabelled medical images

• To compare the performance of the proposed method and model with

the existing models

• To provide accurate and robust labelling for medical images

Secondary Objectives:

• To implement the texture description of diseases and anomalies in

medical images

• To perform integration of Self-Supervised method with texture

generation method in API

SCOPE OF THE STUDY

• The study aims to develop a framework called the Advanced

Contrastive Learning Approach with Text (ACLAT) for accurately

labelling unlabelled images.

• The proposed framework intends to improve the accuracy of

labelling and to enhance the performance of existing labelling methods

and models.

• Upon implementation, the framework will generate a model capable

of identifying anomalies and generating textures for end-users.

• This methodology can be applied to various domains, including the

medical domain, to improve the accuracy of labelling and texture

 generation for unlabelled images.

METHODOLOGY

The section that follows describes the research methodology used to achieve

the primary and secondary goals of this study, which is to develop a self-

supervised contrastive learning framework for medical images with texture

generation.

• Identify and review existing self-supervised methods and models for

classification of unlabelled medical images.

• Designing a self-supervised method, ACLAT Framework, to enhance

the accuracy and robustness of labelling with texture description.

• Implementation of the proposed framework within an API to

generate the texture about diseases and abnormalities in images

• Compare the performance of the proposed ACLAT framework with

existing models identified in the literature review.

• The proposed ACLAT framework is validated and assessed with a

large dataset of unlabelled medical images to ensure accurate labelling.

• Model Enhancement: Based on the findings from the comparative

analysis and validation, the ACLAT framework will undergo

enhancements and optimizations to further improve its performance.

Application: The proposed methodology has the potential to enhance the

accuracy of labeling and improve text generation for unlabeled images

across different domains.

A STUDY ON EXITING VGG16 MODEL TO PREDICT

LABELS

The code is separated into two portions. The first phase consists of

downloading and preparing chest X-ray images for deep-learning models,

while the second step employs the pre-trained VGG16 model to predict

picture labels. The code searches the internet for photos, downloads them,

adjusts their proportions, and eliminates broken images in the first stage. In

the second stage, the pre-trained VGG16 model is loaded, pictures are
processed for the model, predictions are generated, and probabilities are

displayed next to labels (Sharma et al., 2023; Jaishankar et al., 2022; Kiran

et al., 2021; Moussaid et al., 2023).

Downloading and Preprocessing

The following steps gather chest X-ray images and prepare them for use in

deep-learning models. The code searches the internet for images, downloads

them, modifies their dimensions, and removes broken images. The sample

code for this task is given in Figure 1.

Steps

• Use pip to install required packages for image processing.

• Import necessary packages.

• Define a function to search for images online and retrieve their

URLs.

• Download the images using the retrieved URLs and save them in the

directory.

• Resize the downloaded and save them in the same directory.

• Verify the downloaded images and delete any broken or corrupt

images from the directory.

VGG16 to Predict Labels

The provided code utilizes the pre-trained VGG16 model to predict labels

for a collection of chest X-ray images. The sample code is given in Figure 2.

Steps

• Load the pre-trained VGG16 model.

• Set the path to the directory containing the chest X-ray images.

• Get a list of all the image file names in the directory.

• Create an empty list to store the images.

• Loop over each image file in the directory and load the image.
 • Preprocess the image by resizing it to 224x224 and converting it to

an array.

• Append the image array to the list of images.

• Preprocess the images for the VGG16 model.

• Make predictions with the VGG16 model on the preprocessed images

and decode the predictions.

• Print the predicted labels and their probabilities.

Figure 1. Python code: Downloading and preprocessing the images

Figure 2. Python code: VGG16 to predict labels
ROLE OF PYTHON TO PREDICT LABELS

Python is the programming language used to create the code. The code

involves loading a pre-trained VGG16 model, preprocessing images,

making predictions using the model, and printing the predicted labels and

probabilities. Additionally, Python is used to install the necessary packages

for image processing, import the required packages, define functions to

search for and download images, process the downloaded images, and load

the pre-trained model. These tasks are performed using Python (Alghamdi

et al., 2023). The various role of Python in predicting labels is given in

Figure 3.

Figure 3. Python to predict labels

Image Processing: The chest X-ray pictures may be preprocessed using

Python's many image processing packages, such as OpenCV, sci-kit-image,

and Pillow. Image enhancement, noise reduction, and segmentation

techniques may be used to segment the picture's areas of interest.

Deep Learning: Python is home to several deep learning frameworks,

including TensorFlow, Keras, and PyTorch, which may be used to train a

neural network model to recognize abnormalities in chest X-rays. This can

include looking for pneumonia, pneumonia consolidation, or lung nodules.

Statistical Analysis: Python provides several libraries that may be used to

analyze data statistically, such as NumPy, pandas, and SciPy. This may

include examining the dataset's distribution of anomalies or comparing the

effectiveness of various models.

Report Generation: To create PDF reports summarizing the results of the

chest X-ray examination, Python offers several packages, including

ReportLab and PyPDF2. The created reports may include images, graphs,

and statistical summaries of the findings.

CONCLUSION

This chapter's conclusion emphasizes Python's significance in executing

Self-Supervised Contrastive Learning for Medical Images with Report

Generation. The research findings demonstrate that Python's data structures

and algorithms allow fast data manipulation, information extraction, model

training, and evaluation, resulting in more precise and comprehensive

reports and diagnoses. The literature review highlights the benefits of

utilizing Python frameworks like PyTorch, TensorFlow, and Keras for Self-

Supervised Contrastive Learning and reinforces its applicability to medical

imaging.

The potential impact of this study is immense, as medical imaging is a

critical diagnostic tool that enables non-invasive examination of internal

structures of human organs, tissues, and bones. Improving the accuracy of

medical imaging reports and diagnoses can significantly impact patient care

and treatment outcomes. Combining Python with Self-Supervised

Contrastive Learning can enhance patient care by enabling medical

professionals and researchers to generate more accurate and informative

reports.

Future research could improve algorithm performance while overcoming

challenges associated with managing large datasets and integrating self-

supervised contrastive Learning with other machine learning techniques to

enhance the accuracy and effectiveness of medical imaging reports and

diagnoses. These findings suggest that Python and self-supervised

contrastive learning have tremendous potential to improve patient care and

treatment outcomes by developing more accurate and informative medical

imaging reports and diagnoses. Continued research and improvements in

this method could lead to significant advancements in medical care.

REFERENCES
Alghamdi, M. M. M., Dahab, M. Y. H., & Alazwary, N. H. A.

(2023). Enhancing deep learning techniques for the diagnosis of the

novel coronavirus (COVID-19) using X-ray images. Cogent

Engineering, 10(1). doi:10.1080/23311916.2023.2181917

Chen, L., Bentley, P., Mori, K., Misawa, K., Fujiwara, M., &

Rueckert, D. (2019). Self-supervised learning for medical image

analysis using image context restoration. Medical Image Analysis,

58, 101539. doi:10.1016/j.media.2019.101539

Cho, K., Kim, K. D., Nam, Y., Jeong, J., Kim, J., Choi, C., Lee, S.,

Lee, J. S., Woo, S., Hong, G.-S., Seo, J. B., & Kim, N. (2023).

CheSS: Chest X-Ray Pre-trained Model via Self-supervised

Contrastive Learning. Journal of Digital Imaging, 1–9.

doi:10.1007/s10278-023-00782-4

Jaishankar, B., Vishwakarma, S., Mohan, P., Singh Pundir, A. K.,

Patel, I., & Arulkumar, N. (2022). Blockchain for Securing

Healthcare Data Using Squirrel Search Optimization Algorithm.

Intelligent Automation & Soft Computing, 32(3).

Kiran, S. V., Kaur, I., Thangaraj, K., Saveetha, V., Kingsy Grace,

R., & Arulkumar, N. (2021). Machine Learning with Data Science-

Enabled Lung Cancer Diagnosis and Classification Using

Computed Tomography Images. International Journal of Image and

Graphics.

Li, Z., Li, H., Ralescu, A. L., Dillman, J. R., Parikh, N. A., & He,

L. (2023). A Novel Collaborative Self-Supervised Learning Method

for Radiomic Data. arXiv preprint arXiv:2302.09807.

Mishra, A., & Soni, S. (2023, March). Summarization of

Unstructured Medical Data for Accurate Medical Prognosis—A

Learning Approach. In Advances in Cognitive Science and

Communications: Selected Articles from the 5th International

Conference on Communications and Cyber-Physical Engineering

(ICCCE 2022), Hyderabad, India (pp. 825-838). Singapore:

Springer Nature Singapore.

Mohan, P., Neelakandan, S., Mardani, A., Maurya, S., Arulkumar,

N., & Thangaraj, K. (2023). Eagle Strategy Arithmetic

Optimisation Algorithm with Optimal Deep Convolutional Forest

Based FinTech Application for Hyper-automation. Enterprise

Information Systems. doi:10.1080/17517575.2023.2188123

Moussaid, A., Zrira, N., Benmiloud, I., Farahat, Z., Karmoun, Y.,

Benzidia, Y., . . . Ngote, N. (2023, February). On the

Implementation of a Post-Pandemic Deep Learning Algorithm

Based on a Hybrid CT-Scan/X-ray Images Classification Applied to

Pneumonia Categories. In Healthcare (Vol. 11, No. 5, p. 662).

MDPI.

Najma, M., & Sekar, G. (2023, February). Novel CNN Approach

(YOLO v5) to Detect Plant Diseases and Estimation of Nutritional

Facts for Raw and Cooked Foods. In Proceedings of Third

International Conference on Sustainable Expert Systems: ICSES

2022 (pp. 305-329). Singapore: Springer Nature Singapore.

doi:10.1007/978-981-19-7874-6_23

Neelakandan, S., Beulah, J. R., Prathiba, L., Murthy, G. L. N.,

Irudaya Raj, E. F., & Arulkumar, N. (2022). Blockchain with deep

learning-enabled secure healthcare data transmission and

diagnostic model. International Journal of Modeling, Simulation,

and Scientific Computing, 13(04), 2241006.

doi:10.1142/S1793962322410069
Saravanan, S., Lavanya, M., Srinivas, C. M. V., Arunadevi, M., &

Arulkumar, N. (n.d.). Secure IoT Protocol for Implementing

Classified Electroencephalogram (EEG) Signals in the Field of

Smart Healthcare. In Cyber Security Applications for Industry 4.0

(pp. 111-129). Chapman and Hall/CRC.

Sharma, S., & Guleria, K. (2023). A Deep Learning based model

for the Detection of Pneumonia from Chest X-Ray Images using

VGG-16 and Neural Networks. Procedia Computer Science, 218,

357–366. doi:10.1016/j.procs.2023.01.018

Upadhye, S., Neelakandan, S., Thangaraj, K., Babu, D. V.,

Arulkumar, N., & Qureshi, K. (2023). Modeling of Real Time

Traffic Flow Monitoring System Using Deep Learning and

Unmanned Aerial Vehicles. Journal of Mobile Multimedia, 477-

496.

Zhou, H. Y., Lu, C., Chen, C., Yang, S., & Yu, Y. (2023). A Unified

Visual Information Preservation Framework for Self-supervised

Pre-training in Medical Image Analysis. IEEE Transactions on

Pattern Analysis and Machine Intelligence, 1–16.

doi:10.1109/TPAMI.2023.3234002
 CHAPTER 14

Python's Role in Predicting Type 2 Diabetes

Using Insulin DNA Sequence
Aswathi Sasidharan

CHRIST University (Deemed), India

N. Arulkumar

https://orcid.org/0000-0002-9728-477X

CHRIST University (Deemed), India

ABSTRACT

This chapter examines how Python can assist in predicting type 2 diabetes

using insulin DNA sequences, given the substantial problem that biologists

face in objectively evaluating diverse biological characteristics of DNA

sequences. The chapter highlights Python's various libraries, such as

NumPy, Pandas, and Scikit-learn, for data handling, analysis, and machine

learning, as well as visualization tools, such as Matplotlib and Seaborn, to

help researchers understand the relationship between different DNA

sequences and type 2 diabetes. Additionally, Python's ease of integration

with other bioinformatics tools, like BLAST, EMBOSS, and ClustalW, can

help identify DNA markers that could aid in predicting type 2 diabetes. In

addition, the initiative tries to identify unique gene variants of insulin

protein that contribute to diabetes prognosis and investigates the risk factors

connected with the discovered gene variants. In conclusion, Python's

versatility and functionality make it a valuable tool for researchers studying

insulin DNA sequences and type 2 diabetes prediction.

INTRODUCTION
Diabetes is a chronic disease that affects millions of people worldwide and

is characterized by elevated blood sugar levels that can lead to various

health complications. The two most common types of diabetes are type 1

and type 2 (Eizirik et al., 2020). Insulin-producing pancreatic cells are

attacked and eliminated by the immune system in type 1 diabetes, an

autoimmune condition. Type 2 diabetes, on the other hand, is marked by

insulin resistance or inadequate insulin production to maintain normal blood

sugar levels. The sequencing of DNA is a powerful tool that allows

scientists to investigate the genetic basis of numerous diseases, such as

diabetes. Researchers can identify genetic variants associated with an

increased risk of developing type 2 diabetes by analyzing DNA sequences

(Jurgens et al., 2020). The identification of these genetic markers could

assist healthcare professionals in devising more individualized treatment

plans for diabetic patients and in preventing the onset of diabetes in those at

risk.

PREDICTING TYPE 2 DIABETES

Important for early intervention and disease prevention is the ability to

predict type 2 diabetes. By identifying individuals at risk for developing

type 2 diabetes, healthcare professionals can provide targeted interventions,

such as lifestyle modifications or medication, to prevent or better manage

the disease. Alterations in the insulin gene sequence can contribute to the

development of type 2 diabetes, making the sequencing of insulin DNA a

crucial component of predicting type 2 diabetes (Batista et al, 2020).

Understanding the genetic basis of type 2 diabetes can lead to more

effective interventions and therapies.

Predicting Type 2 Diabetes Using Insulin DNA Sequence

Multiple methods exist in which Python can contribute to research on

predicting type 2 diabetes using insulin DNA sequences.

• Data Administration and Analysis: NumPy, Pandas, and Scikit-

learn are Python libraries that can preprocess and analyze DNA

sequence data. These libraries provide data manipulation, purification,

 analysis functions, and methods essential for predicting type 2 diabetes

based on the insulin DNA sequence (Li et al., 2023).

• Automatic Learning: Python has become the most popular

programming language for machine learning and artificial intelligence

research due to its simplicity and easy learning syntax. Using Python's

machine learning libraries like Scikit-learn, TensorFlow, and Keras, it

is possible to develop predictive models for identifying DNA markers

that could be used to predict type 2 diabetes (Li et al., 2021).

• Visualization The powerful visualization libraries of Python, such as

Matplotlib and Seaborn, can facilitate data visualization, such as DNA

sequence data. With visualization, researchers can better comprehend

and investigate the relationship between the numerous DNA sequences

and type 2 diabetes.

• Integration with other tools Python is easily compatible with

standard bioinformatics tools for DNA sequence analysis, such as

BLAST, EMBOSS, and ClustalW. These tools can be used to compare

the DNA sequence data for insulin with other DNA sequences in a

database to identify similarities and differences that could aid in

predicting type 2 diabetes (Rather et al., 2023).

Python provides data analysis, machine learning, visualization, and

integration with other bioinformatics tools, thereby playing a crucial role in

research on predicting type 2 diabetes based on the insulin DNA sequence.

THE MAIN OBJECTIVES OF THE RESEARCH

This chapter's primary objective is to study the existing machine-learning

model that classifies the DNA sequence of the insulin gene to identify type

2 diabetes based on gene sequence transformation. This study will evaluate

this machine-learning model's effectiveness compared to other extant

models. In addition, the initiative endeavors to identify unique insulin gene

variants that contribute to the prognosis of diabetes. In addition, the

associated risk factors will be investigated. The study will highlight

Python's role in facilitating this research and demonstrate its value as a tool

for predicting type 2 diabetes based on the insulin DNA sequence.

Python is an integral component of each phase of this methodology.

• Input: Python can retrieve the DNA sequence for the insulin gene

from Genebank using Biopython libraries.

• Python can handle one-hot encoding of DNA sequences,

representing the DNA sequence as a binary code to make it compatible

with machine learning algorithms.

• Python can divide data into train and test sets, a model training and

evaluation requirement.

• Model-Construction: Using the training set, Python's machine

learning libraries, such as Scikit-learn and Keras, can be used to

construct the model (Liu, 2020).

• Python can be used to evaluate the model's performance in

predicting type 2 diabetes and validate the model using metrics such as

precision, recall, F1 score, and accuracy. Additionally, it can be used to

visualize results with the Matplotlib and Seaborn libraries.

• The adaptability and functionality of Python make it an

indispensable tool for each phase of this methodology.

THE ROLE OF PYTHON

Python is a popular programming language in various scientific fields,

including genomics and bioinformatics. NumPy, Pandas, and Scikit-learn

are just a few of Python's libraries for data manipulation, analysis, and

machine learning. In addition, visualization tools such as Matplotlib and

Seaborn can aid researchers in analyzing and comprehending the

relationship between different DNA sequences and type 2 diabetes. The

straightforward incorporation of Python with other bioinformatics tools,

such as BLAST, EMBOSS, and ClustalW, enables the identification of

DNA markers that may aid in predicting type 2 diabetes. Python's

adaptability and functionality make it a valuable resource for scientists

studying insulin DNA sequences and type 2 diabetes prediction.

Python Libraries for Data Handling, Analysis, and

Machine Learning
Due to its extensive libraries that provide data manipulation, analysis, and

visualization tools, Python is now the preferred language for data analysis

and machine learning. This article concentrates on NumPy, Pandas, and

Scikit-learn, the three most commonly used Python libraries for data

manipulation, research, and machine learning. In the context of type 2

diabetes prediction based on the insulin DNA sequence, these Python

libraries are indispensable for managing and manipulating DNA sequence

data and constructing and evaluating predictive models. NumPy and Pandas

are useful for preprocessing and analyzing DNA sequence data, whereas

Scikit-learn helps build predictive models. Python and its libraries provide a

robust and adaptable environment for analyzing and predicting complex

biological data.

NumPy

NumPy (Numerical Python) is a Python library for scientific computing. It

supports enormous multidimensional arrays and matrices and has a vast

library of mathematical functions that operate on these arrays. Numerous

other Python scientific libraries, including Pandas and Scikit-learn,

significantly rely on NumPy. NumPy is a handy tool for manipulating and

managing numerical data. Array-oriented programming simplifies the

composition and execution of complex algorithms. NumPy also offers tools

for linear algebra, Fourier transforms, and random number generation

(Yalçın, 2021).

Pandas

Pandas is a Python data manipulation and analysis utility. It provides tools

for reading and writing data in CSV, Excel, SQL databases, and JSON

formats. Pandas is based on NumPy and offers the DataFrame object, a

table-like data structure with rows and columns. Pandas provides tools for

data purification, integration, filtering, and aggregation. Additionally, it

offers robust data exploration and comprehension tools through its

visualization capabilities. Pandas is particularly useful for working with

time-series data because it allows for tools for resampling, time shifting,

and window rotation (Betancourt, R. et al., 2019).

Scikit-Learn

Scikit-learn is a Python library for machine learning. It includes supervised

and unsupervised learning techniques, data preprocessing, and model

evaluation (Hao, J. et al., 2019). Scikit-learn is built upon NumPy and

provides a standard API for working with different machine-learning

algorithms. Scikit-learn includes classification, regression, clustering,

dimensionality reduction, and model selection software. In addition, it

provides tools for feature selection, extraction, model evaluation, and

validation. Scikit-learn is especially useful for developing predictive

models, as it allows for an extensive selection of machine learning

algorithms and optimization tools.

VISUALIZATION TOOLS

Using Python's extensive library of visualization tools, DNA sequences

associated with type 2 diabetes can be analyzed and comprehended.

Matplotlib and Seaborn are well-known libraries that provide various

visualization tools, including line charts, scatter plots, bar charts, and

heatmaps. Using these tools, researchers can better understand the genetic

factors contributing to type 2 diabetes and develop more effective

treatments. As bioinformatics continues to develop, the significance of

visualization tools for analyzing and understanding DNA sequences

increases. Python offers vast libraries for data visualization, analysis, and

interpretation. This article examines how two popular libraries, Matplotlib

and Seaborn, can be utilized to comprehend the DNA sequences associated

with type 2 diabetes.

Matplotlib

Matplotlib is a popular Python library for creating static, interactive, and

animated plots. It provides, among other visualizations, line charts, scatter

graphs, bar charts, and heatmaps (Klasen et al., 2023). Matplotlib can be

used to visualize DNA sequences and identify patterns and anomalies. A

scatter plot, for instance, can be used to depict the distribution of

nucleotides in a DNA sequence. A line chart can be used to illustrate the

evolution of the nucleotide sequence over time or to compare different DNA

sequences. Heatmaps can depict associations between other nucleotides and

their propensity to induce type 2 diabetes.

Seaborn

Seaborn is a Matplotlib-based Python library for data visualization that

offers a high-level interface for creating informative and aesthetically

appealing statistical visualizations (DSouza et al., 2020). Seaborn is notably

useful for examining the relationships between variables and visualizing

DNA sequence data. Seaborn can be utilized, for example, to visualize the

distribution of different nucleotides in the DNA sequence and identify

patterns indicative of type 2 diabetes. Seaborn can also generate heatmaps

illustrating the associations between various nucleotides in the DNA

sequence and the likelihood of developing type 2 diabetes.

ALGORITHMS IN PYTHON

This chapter examines five distinct machine learning approaches and how

they could be used to predict type 2 diabetes using the insulin DNA

sequence. In this context, the different algorithms are investigated

convolutional neural networks, decision trees, random forests, logistic

regression, and k-nearest neighbours. Decision tree classifiers can pinpoint

the most important genetic factors associated with diabetes risk, unlike

logistic regression, which may detect genetic markers and provide a

prediction model for risk assessment. To improve predicted accuracy, the

ensemble learning method known as random forest combines the

predictions of several decision trees. Convolutional neural networks may

extract pertinent traits from DNA sequence data to predict the likelihood of

developing diabetes, unlike K-nearest neighbors, which may categorize

people's DNA sequences into diabetic or non-diabetic groups based on

similarity.

Logistic Regression
Logistic regression is a standard statistical technique for predicting results

from a collection of independent variables or predictors. Logistic regression

may be used to determine the relationship between the DNA sequence and a

person's chance of acquiring diabetes when predicting type 2 diabetes using

the insulin DNA sequence (Klarin et al., 2019). The existence or absence of

type 2 diabetes would result from the dependent variable in this situation,

while the insulin DNA sequence would be the predictor variable. To

establish a prediction model for future occurrences, logistic regression may

be performed to determine if specific DNA sequence changes or areas are

linked to a greater risk of acquiring diabetes.

By looking at several measures like accuracy, precision, and recall, logistic

regression may also be used to assess the model's efficacy. These criteria

may be used to assess the model's propensity for prediction and clinical

utility. Overall, by assisting in identifying pertinent genetic markers and

developing a predictive model for risk assessment, logistic regression may

significantly contribute to the prediction of type 2 diabetes by utilizing the

insulin DNA sequence.

Decision Tree Classifier

Decision tree classifier, another well-liked machine learning method, uses

the insulin DNA sequence as a predictor (Malik et al., 2022). A decision

tree is a visual depiction of a collection of decision rules and their results.

By using a variety of inputs, such as information on the insulin DNA

sequence, the decision tree may be used to predict a person's chance of

acquiring diabetes. The most significant traits or genetic variables linked to

a higher risk of acquiring diabetes may be found using the decision tree

classifier. It could be possible to comprehend the hereditary causes of

diabetes by pointing forth potential points of intervention.

A prediction model that divides incoming instances into groups related to or

unrelated to diabetes may be developed using the decision tree classifier.

The right drugs may be given to patients with a high risk of acquiring

diabetes by identifying them. Decision tree classifiers may also be used to

calculate accuracy, precision, and recall metrics to assess the model's

efficacy. These criteria may be used to assess the model's propensity for

prediction and clinical utility. By selecting significant genetic markers,

developing a predictive model, and assessing the model's performance, the

decision tree classifier may significantly predict type 2 diabetes by utilizing

insulin DNA sequence data.

Random Forest

Random forest is a machine-learning technique that may be used for

classification and regression issues (Feng et al., 2021). It is a kind of

ensemble learning approach that incorporates several decision trees to

improve the model's predictive accuracy. The forecasts of all the individual

decision trees, each trained using a random sample of the data and

attributes, are combined to create the final forecast.

In the context of disease prediction using the Insulin DNA sequence, the

Random forest may be utilized to find important DNA sequence

characteristics linked to Type 2 Diabetes. The algorithm may be trained on

a dataset including the DNA sequences of persons with and without Type 2

Diabetes and other relevant clinical data. A set of traits in the DNA

sequence may point to the existence or absence of a particular mutation or

variant. The Random Forest approach may then be used to identify the

primary symptoms of the condition, and based on a patient's DNA

sequence, it is possible to predict if they are likely to develop Type 2

Diabetes. This knowledge may be used to develop customized treatment

plans for those more likely to get the ailment. Random forests are typically

helpful in identifying relevant components in challenging datasets and may

be used in various applications, including the analysis of genetic data and

medical diagnostics.

K-Nearest Neighbours (K-NN)

Machine learning techniques like the K-NN algorithm are used for

classification and regression analysis (Mezquita et al., 2021). It is often used

for predictive modeling in many disciplines, including biology and genetics.

The K-NN technique may be used to categorize an individual's DNA

sequence as either belonging to a diabetic or non-diabetic group based on

the similarity of their DNA sequence to other known DNA sequences in the

context of predicting Type 2 Diabetes using insulin DNA sequence. The

first step is to gather a large dataset of DNA sequences from people who are

either diabetic or not to employ the k-NN technique for this. The next step

would be to identify relevant qualities from these DNA sequences that may

be used to predict the chance of acquiring diabetes.

The K-NN approach trains a model on the dataset after collecting features.

The method locates a new DNA sequence's K-Nearest neighbors in the

dataset and then utilizes the neighbors' class labels to predict the class label

for the new sequence. When determining the value of k, the size of the

dataset and the needed degree of precision are often considered. The K-NN

algorithm may, in general, be a valuable tool for predicting the likelihood of

developing Type 2 Diabetes based on knowledge of the insulin DNA

sequence. However, it is essential to remember that the accuracy of the

predictions will depend on the quality and volume of the dataset, the

features used, and the value of k.

Convolutional Neural Networks (CNNs)

CNNs, a deep learning technique, are often employed in image recognition

and classification applications (Alzubaidi et al., 2021). However, they may

also be utilized in other domains, including bioinformatics, to predict

outcomes based on DNA sequence data.

In order to predict Type 2 Diabetes using the insulin DNA sequence, CNNs

may be utilized to extract relevant traits from DNA sequence data. Using a

CNN, the system may automatically find connections and patterns in DNA

sequence data that would not be apparent to the human eye. The DNA

sequence data may be represented as a one-dimensional sequence of

nucleotides (A, C, G, and T) as an input to the CNN. The CNN will next

employ convolutional and pooling layers to extract features from the

sequence data. Fully connected layers may further analyze these data to

provide a forecast of the presence or absence of type 2 diabetes.

Overall, CNNs may help improve the accuracy of predictions generated

using insulin DNA sequence data to predict Type 2 Diabetes by spotting

important patterns and links in the data that may not be immediately

obvious using traditional methodologies.

Long Short-Term Memory (LSTM)

Finding patterns and correlations in sequential data is a task that is

especially well suited to recurrent neural networks (RNNs), specifically

LSTM. LSTM algorithms may analyze the sequential data and provide

predictions for Type 2 Diabetes based on insulin DNA sequence data

analysis. When dealing with sequential data, such as DNA sequences,

LSTMs can potentially develop long-term dependencies. DNA sequences

may have significant patterns or links spread across the sequence at various

locations. They are often relatively lengthy and complex. These tendencies

may be recognized by LSTMs, which can then be utilized to make precise

forecasts (Sansano et al., 2020).

The insulin DNA sequence may be seen as a time series using the

information it contains, with each point in the sequence denoting a different

time step. The patterns and connections it discovers in the sequential data

may then be used to train the LSTM to predict whether or not a person has

Type 2 Diabetes. The ability of an LSTM technique to capture correlations

and patterns over many time steps may be pretty helpful for investigating

lengthy DNA sequences. Additionally, LSTMs can make predictions even

in the presence of missing data, which often happens with data from DNA

sequencing. Overall, by efficiently gathering patterns and correlations

throughout the whole sequence of DNA data, the application of LSTM

algorithms in predicting Type 2 Diabetes utilizing insulin DNA sequence

data may assist in improving prediction accuracy.

BIOINFORMATICS TOOLS

Because of its flexibility and versatility, Python is the ideal programming

language for integrating various bioinformatics tools to identify DNA

markers associated with type 2 diabetes. This section will explain the three
well-known bioinformatics software BLAST, EMBOSS, and ClustalW, and

how they may be used with Python to find DNA markers (Lo, 2023).

BLAST

A bioinformatics method known as BLAST (Basic Local Alignment Search

Technique) compares DNA sequences. The most comparable sequences are

shown when a query sequence is compared to a database of known DNA

sequences. Users may utilize BLAST capability inside Python by using one

of the many Python libraries that includes a BLAST module, such as

Biopython. It is feasible to compare and contrast the insulin DNA sequences

of people with and without type 2 diabetes using BLAST, which one day

might be used as DNA markers for the condition. Researchers may quickly

analyze and compare enormous DNA sequence databases by combining

Python with BLAST.

EMBOSS

A package of bioinformatics tools called EMBOSS is utilized for molecular

biology tasks, including protein and sequence analysis. Among the more

than 200 tools EMBOSS offers are sequence alignment, protein structure

prediction, and sequence editing. Users may access EMBOSS capabilities

from Python using the PyEMBOSS module for EMBOSS. In the context of

predicting type 2 diabetes, EMBOSS may be used to evaluate insulin DNA

sequences and identify any alterations or mutations that may potentially

serve as DNA markers for the disease. By integrating Python with

EMBOSS, researchers may speed up and increase the effectiveness of DNA

sequence analysis.

ClustalW

ClustalW, a bioinformatics technique, may align multiple DNA or protein

sequences. A multiple sequence alignment is produced by comparing

sequences and identifying similarities and differences, which may be used

to examine genetic linkages. The language supports the Biopython module

of Python, which may be used to access ClustalW's functionality. ClustalW

may be used to compare the insulin DNA sequences of people with and

without type 2 diabetes to find similarities and differences that potentially

act as DNA markers for the condition. Researchers can quickly compare and

analyze enormous DNA sequence datasets using Python and ClustalW,

making it simpler to find DNA markers for type 2 diabetes.

CONCLUSION

Python's ability to differentiate specific gene changes and risk variables

makes it a helpful tool for forecasting diseases, notably type 2 diabetes.

Through statistical analysis and machine learning algorithms, researchers

may develop exact models for predicting the chance of developing diabetes

and for producing customized medications. Python is the best programming

language for integrating with several bioinformatics tools, such as BLAST,

EMBOSS, and ClustalW, to swiftly analyze and compare huge DNA

sequence datasets in search of DNA markers connected to type 2 diabetes.

This insight may make the development of more potent treatments and

drugs possible. To predict a patient's chance of getting type 2 diabetes,

machine learning models created using Scikit-learn, TensorFlow, and Keras,

three of the machine learning libraries available with Python, may be

utilized. These models are based on the insulin DNA sequences.

Additionally, by analyzing and displaying data using Python tools like

Seaborn and Matplotlib, researchers may find patterns in patient data and

improve the accuracy of the machine-learning model.

REFERENCES

Alzubaidi, L., Zhang, J., Humaidi, A. J., Al-Dujaili, A., Duan, Y.,

Al-Shamma, O., Santamaría, J., Fadhel, M. A., Al-Amidie, M., &

Farhan, L. (2021). Review of deep learning: Concepts, CNN

architectures, challenges, applications, future directions. Journal of

Big Data, 8(1), 1–74. doi:10.1186/s40537-021-00444-8

Batista, T. M., Jayavelu, A. K., Albrechtsen, N. J. W., Iovino, S.,

Lebastchi, J., Pan, H., & Kahn, C. R. (2020). A cell-autonomous

signature of dysregulated protein phosphorylation underlies muscle

insulin resistance in type 2 diabetes. Cell Metabolism, 32(5), 844–

859. doi:10.1016/j.cmet.2020.08.007

Eizirik, D. L., Pasquali, L., & Cnop, M. (2020). Pancreatic β-cells

in type 1 and type 2 diabetes mellitus: Different pathways to failure.

Nature Reviews. Endocrinology, 16(7), 349–362.

doi:10.1038/s41574-020-0355-7

Hao, J., & Ho, T. K. (2019). Machine learning made easy: A review

of scikit-learn package in python programming language. Journal

of Educational and Behavioral Statistics, 44(3), 348–361.

doi:10.3102/1076998619832248

Jurgens, S. J., Choi, S. H., Morrill, V. N., Chaffin, M., Pirruccello,

J. P., Halford, J. L., Weng, L.-C., Nauffal, V., Roselli, C., Hall, A.

W., Oetjens, M. T., Lagerman, B., vanMaanen, D. P., Abecasis, G.,

Bai, X., Balasubramanian, S., Baras, A., Beechert, C., Boutkov, B.,

& Ellinor, P. T. (2022). Analysis of rare genetic variation

underlying cardiometabolic diseases and traits among 200,000

individuals in the UK Biobank. Nature Genetics, 54(3), 240–250.

doi:10.1038/s41588-021-01011-w

Klarin, D., Lynch, J., Aragam, K., Chaffin, M., Assimes, T. L.,

Huang, J., Lee, K. M., Shao, Q., Huffman, J. E., Natarajan, P.,

Arya, S., Small, A., Sun, Y. V., Vujkovic, M., Freiberg, M. S.,

Wang, L., Chen, J., Saleheen, D., Lee, J. S., & Damrauer, S. M.

(2019). Genome-wide association study of peripheral artery disease

in the Million Veteran Program. Nature Medicine, 25(8), 1274–

1279. doi:10.1038/s41591-019-0492-5
Klasen, V., Bogucka, E. P., Meng, L., & Krisp, J. M. (2023). How

we see time–the evolution and current state of visualizations of

temporal data. International Journal of Cartography, 1-18.

doi:10.1080/23729333.2022.2156316

Li, J., Chen, Q., Hu, X., Yuan, P., Cui, L., Tu, L., Cui, J., Huang, J.,

Jiang, T., Ma, X., Yao, X., Zhou, C., Lu, H., & Xu, J. (2021).

Establishment of noninvasive diabetes risk prediction model based

on tongue features and machine learning techniques. International

Journal of Medical Informatics, 149, 104429.

doi:10.1016/j.ijmedinf.2021.104429

Li, X., Zhang, Y., Leung, J., Sun, C., & Zhao, J. (2023).

EDAssistant: Supporting Exploratory Data Analysis in

Computational Notebooks with In Situ Code Search and

Recommendation. ACM Transactions on Interactive Intelligent

Systems, 13(1), 1–27. doi:10.1145/3545995

Liu, Y. H. (2020). Python Machine Learning By Example: Build

intelligent systems using Python, TensorFlow 2, PyTorch, and

scikit-learn. Packt Publishing Ltd.

Lo, Y. T. (2023). Bioinformatics for DNA Sequencing Data

Analysis. In Authentication of Chinese Medicinal Materials by

DNA Technology: Techniques and Applications (pp. 89-125).

Academic Press.

Malik, M., Kaushik, A., & Khatana, R. (2022). Classification-

Based Prediction Techniques Using ML: A Perspective for Health

Care. In Data Science for Effective Healthcare Systems (pp. 43-56).

Chapman and Hall/CRC.

Mezquita, Y., Alonso, R. S., Casado-Vara, R., Prieto, J., &

Corchado, J. M. (2021). A review of k-nn algorithm based on

classical and quantum machine learning. In Distributed Computing

and Artificial Intelligence, Special Sessions, 17th International

Conference (pp. 189-198). Springer International Publishing.

doi:10.1007/978-3-030-53829-3_20

Rather, M. A., Agarwal, D., Bhat, T. A., Khan, I. A., Zafar, I.,

Kumar, S., Amin, A., Sundaray, J. K., & Qadri, T. (2023).

Bioinformatics approaches and big data analytics opportunities in

improving fisheries and aquaculture. International Journal of

Biological Macromolecules, 233, 123549.

doi:10.1016/j.ijbiomac.2023.123549

Sansano, E., Montoliu, R., & Belmonte Fernandez, O. (2020). A

study of deep neural networks for human activity recognition.

Computational Intelligence, 36(3), 1113–1139.

doi:10.1111/coin.12318

Yalçın, O. G. (2021). Complementary Libraries to TensorFlow 2.x.

Applied Neural Networks with TensorFlow 2: API Oriented Deep

Learning with Python, 81-94.

 Compilation of References
Abdalkafor, A. S., & Aliesawi, S. A. (2022, October). Data

aggregation techniques in wireless sensors networks (WSNs):

Taxonomy and an accurate literature survey. In AIP Conference

Proceedings (Vol. 2400, No. 1, p. 020012). AIP Publishing LLC.

Abdulzahra, A. M. K., Al-Qurabat, A. K. M., & Abdulzahra, S. A.

(2023). Optimizing energy consumption in WSN-based IoT using

unequal clustering and sleep scheduling methods. Internet of

Things, 22, 100765. doi:10.1016/j.iot.2023.100765

Abu-Baker, A., Shakhatreh, H., Sawalmeh, A., & Alenezi, A. H.

(2023). Efficient Data Collection in UAV-Assisted Cluster-Based

Wireless Sensor Networks for 3D Environment: Optimization

Study. Journal of Sensors, 2023, 2023. doi:10.1155/2023/9513868

Adams, B., Baller, D., Jonas, B., Joseph, A. C., & Cummiskey, K.

(2021). Computational skills for multivariable thinking in

introductory statistics. Journal of Statistics and Data Science

Education, 29(sup1), S123-S131.

Agathiya Raja, G. K. (2022). Introduction To Python for Social

Network Analysis. Wiley Online Library.

Aggarwal, S., & Kumar, N. (2021). History of blockchain-

blockchain 1.0: Currency. Advances in Computers, 121, 147–169.

Ahmed T. Alam M. A. Paul R. R. Hasan M. T. Rab R. (2022,

February). Machine Learning and Deep Learning Techniques For

Genre Classification of Bangla Music. In 2022 International

Conference on Advancement in Electrical and Electronic

Engineering (ICAEEE) (pp. 1-6). IEEE.

doi:10.1109/ICAEEE54957.2022.9836434

Akbar, I., Hashem, T., Yaqoob, I., Anuar, N.B., Mokhtar, S., Gani,

A., & Khan, S.A. (2015). The rise of “big data” on cloud

computing: review and open research issues. Inf Syst., 47, 98–115.

Alammar, J. (2022). NumPy: the absolute basics for beginners.

Retrieved 10 25, 25, from http://jalammar.github.io/

Albenis Pérez-Alarcón, J. C.-A. (2021). Alarconpy: a Python

Package for Meteorologists.

Alcoz, A. G., Dietmüller, A., & Vanbever, L. (2020, February). SP-

PIFO: Approximating Push-In First-Out Behaviors using Strict-

Priority Queues. In NSDI (pp. 59-76). Academic Press.

Alghamdi, M. M. M., Dahab, M. Y. H., & Alazwary, N. H. A.

(2023). Enhancing deep learning techniques for the diagnosis of the

novel coronavirus (COVID-19) using X-ray images. Cogent

Engineering, 10(1). doi:10.1080/23311916.2023.2181917

Almadhoun, E., & Parham-Mocello, J. (2023). Students’ difficulties

with inserting and deleting nodes in a singly linked list in the C

programming language. Journal of Computer Languages, 74,

101184. doi:10.1016/j.cola.2022.101184

Alzubaidi, L., Zhang, J., Humaidi, A. J., Al-Dujaili, A., Duan, Y.,

Al-Shamma, O., Santamaría, J., Fadhel, M. A., Al-Amidie, M., &

Farhan, L. (2021). Review of deep learning: Concepts, CNN

architectures, challenges, applications, future directions. Journal of

Big Data, 8(1), 1–74. doi:10.1186/s40537-021-00444-8

Arslan, B. (2022). Search and Sort Algorithms for Big Data

Structures. Academic Press.

Arulkumar, N., Paulose, J., Galety, M. G., Manimaran, A.,

Saravanan, S., & Saleem, R. A. (2022). Exploring Social

Networking Data Sets. Social Network Analysis: Theory and

Applications, 205-228.

Ateeq, M., & Afzal, M. K. (2023). 10 Programming Languages,

Tools, and Techniques. Data-Driven Intelligence in Wireless

Networks: Concepts, Solutions, and Applications, 237.

Baka, B. (2017). Python Data Structures and Algorithms. Packt

Publishing Ltd.

Bandopadhyay, S., & Shang, H. (2022). SADHANA: A Doubly

Linked List-based Multidimensional Adaptive Mesh Refinement

Framework for Solving Hyperbolic Conservation Laws with

Application to Astrophysical Hydrodynamics and

Magnetohydrodynamics. The Astrophysical Journal. Supplement

Series, 263(2), 32. doi:10.3847/1538-4365/ac9279

Batista, T. M., Jayavelu, A. K., Albrechtsen, N. J. W., Iovino, S.,

Lebastchi, J., Pan, H., & Kahn, C. R. (2020). A cell-autonomous

signature of dysregulated protein phosphorylation underlies muscle

insulin resistance in type 2 diabetes. Cell Metabolism, 32(5), 844–

859. doi:10.1016/j.cmet.2020.08.007

Bellogín, A., Cantador, I., Díez, F., Castells, P., & Chavarriaga, E.

(2013). An empirical comparison of social, collaborative filtering,

and hybrid recommenders. ACM Transactions on Intelligent

Systems and Technology, 4(1), 1–37.

doi:10.1145/2414425.2414439
Beri, R. (2019). Python Made Simple: Learn Python programming

in easy steps with examples. Bpb Publications.

Besta, M., Peter, E., Gerstenberger, R., Fischer, M., Podstawski,

M., Barthels, C., . . . Hoefler, T. (2019). Demystifying graph

databases: Analysis and taxonomy of data organization, system

designs, and graph queries. arXiv preprint arXiv:1910.09017.

Beyer & Laney. (2012). The importance of “big data”: A

definition. Gartner, Tech. Rep.

Bharany, S., Sharma, S., Badotra, S., Khalaf, O. I., Alotaibi, Y.,

Alghamdi, S., & Alassery, F. (2021). Energy-efficient clustering

scheme for flying ad-hoc networks using an optimized LEACH

protocol. Energies, 14(19), 6016. doi:10.3390/en14196016

Bhardwaj S. Dash S. K. (2015). VDMR-DBSCAN: varied density

Mapreduce DBSCAN. International conference on Big Data

analytics. doi:10.1007/978-3-319-27057-9_10

Bisong, E. (2019). Matplotlib and seaborn. Building Machine

Learning and Deep Learning Models on Google Cloud Platform: A

Comprehensive Guide for Beginners, 151-165.

Bisong, E. (2019). Python. Building Machine Learning and Deep

Learning Models on Google Cloud Platform: A Comprehensive

Guide for Beginners, 71-89.

Boockmann, J. H., & Lüttgen, G. (2020, May). Learning Data

Structure Shapes from Memory Graphs. LPAR.

Borah, B., & Bhattacharyya, D. K. (2008). DDSC: A density

differentiated spatial clustering technique. Journal of Computers,

3(2), 72–79. doi:10.4304/jcp.3.2.72-79

Bria R. Wahab A. Alaydrus M. (2019, October). Energy efficiency

analysis of TEEN routing protocol with isolated nodes. In 2019

Fourth International Conference on Informatics and Computing

(ICIC) (pp. 1-5). IEEE. doi:10.1109/ICIC47613.2019.8985668

Buckley, I. A., & Buckley, W. S. (2017). Teaching software testing

using data structures. International Journal of Advanced Computer

Science and Applications, 8(4).

Carullo, G. (2020). Data Structures. Implementing Effective Code

Reviews: How to Build and Maintain Clean Code, 27-42.

Chauhan, V., & Soni, S. (2021). Energy aware unequal clustering

algorithm with multi-hop routing via low degree relay nodes for

wireless sensor networks. Journal of Ambient Intelligence and

Humanized Computing, 12(2), 2469–2482. doi:10.1007/s12652-

020-02385-1

Chemmalar Selvi, G., & Lakshmi Priya, G. G. (2020). An

Epidemic Analysis of COVID-19 using Exploratory Data Analysis

Approach. Predictive Analytics Using Statistics and Big Data:

Concepts and Modeling, 99.

Chen, L., Bentley, P., Mori, K., Misawa, K., Fujiwara, M., &

Rueckert, D. (2019). Self-supervised learning for medical image

analysis using image context restoration. Medical Image Analysis,

58, 101539. doi:10.1016/j.media.2019.101539

Chen, S., Zhang, L., Tang, Y., Shen, C., Kumar, R., Yu, K., Tariq,

U., & Bashir, A. K. (2020). Indoor temperature monitoring using

wireless sensor networks: A SMAC application in smart cities.

Sustainable Cities and Society, 61, 102333.

doi:10.1016/j.scs.2020.102333

Chen, Z., Wang, Y., Zhao, B., Cheng, J., Zhao, X., & Duan, Z.

(2020). Knowledge graph completion: A review. IEEE Access :

Practical Innovations, Open Solutions, 8, 192435–192456.

doi:10.1109/ACCESS.2020.3030076

Cho, K., Kim, K. D., Nam, Y., Jeong, J., Kim, J., Choi, C., Lee, S.,

Lee, J. S., Woo, S., Hong, G.-S., Seo, J. B., & Kim, N. (2023).

CheSS: Chest X-Ray Pre-trained Model via Self-supervised

Contrastive Learning. Journal of Digital Imaging, 1–9.

doi:10.1007/s10278-023-00782-4

Cooksey, R. W. (2020). Descriptive statistics for summarising data.

Illustrating statistical procedures: Finding meaning in quantitative

data, 61-139.

Critien, J. V., Gatt, A., & Ellul, J. (2022). Bitcoin price change and

trend prediction through twitter sentiment and data volume.

Financial Innovation, 8(1), 1–20.

Daanoune, I., Abdennaceur, B., & Ballouk, A. (2021). A

comprehensive survey on LEACH-based clustering routing

protocols in Wireless Sensor Networks. Ad Hoc Networks, 114,

102409. doi:10.1016/j.adhoc.2020.102409

Dai, B. R., & Lin, I.-C. (2012). Efficient Map/Reduce-based

DBSCAN algorithm with optimized data partition. Fifth

International Conference on Cloud Computing.

de Aguiar Neto, F. S., da Costa, A. F., Manzato, M. G., &

Campello, R. J. (2020). Pre-processing approaches for collaborative

filtering based on hierarchical clustering. Information Sciences,

534, 172–191. doi:10.1016/j.ins.2020.05.021

Decker, C., Seidel, J., & Wattenhofer, R. (2016, January). Bitcoin

meets strong consistency. In Proceedings of the 17th International

Conference on Distributed Computing and Networking (pp. 1-10).

Academic Press.

Denis, D. J. (2020). Univariate, bivariate, and multivariate statistics

using R: quantitative tools for data analysis and data science. John

Wiley & Sons. doi:10.1002/9781119549963

Domkin, V. (2021). Linked Lists. Programming Algorithms in

Lisp: Writing Efficient Programs with Examples in ANSI Common

Lisp, 75-99.

Duan, L., Xu, L., Guo, F., Lee, J., & Yan, B. (2007). A local-

density based spatial clustering algorithm with noise. Information

Systems, 32(7), 978–986. doi:10.1016/j.is.2006.10.006

Dubois, P. F., Hinsen, K., & Hugunin, J. (1996). Numerical Python.

Computers in Physics, 10(3), 262–267. doi:10.1063/1.4822400

Dugard, P., Todman, J., & Staines, H. (2022). Approaching

multivariate analysis: A practical introduction. Taylor & Francis.

doi:10.4324/9781003343097

Dutta, S., Chowdhury, A., Debnath, I., Sarkar, R., Dey, H., &

Dutta, A. (2020). The role of data structures in different fields of

computer-science: A review. Journal of Mathematical Sciences &

Computational Mathematics, 1(3), 363–373.

doi:10.15864/jmscm.1310
Eizirik, D. L., Pasquali, L., & Cnop, M. (2020). Pancreatic β-cells

in type 1 and type 2 diabetes mellitus: Different pathways to failure.

Nature Reviews. Endocrinology, 16(7), 349–362.

doi:10.1038/s41574-020-0355-7

Elbatta, M. T., Bolbol, R. M., & Ashur, W. M. (2011). A vibration

method for discovering density varied clusters. Int Sch Res Not.

Erciyes, K. (2021). A Short Review of Python. In Algebraic Graph

Algorithms. Springer. doi:10.1007/978-3-030-87886-3

Ester, M., Kriegel, H. P., Sander, J., & Xu, X. (1996). A density-

based algorithm for discovering clusters in large spatial databases

with noise. In 2nd international conference on knowledge discovery

and data mining. Academic Press.

Gaonkar, M.N., & Sawant, K. (2013). AutoEpsDBSCAN:

DBSCAN with Eps automatic for large dataset. Int J Adv Comput

Theory Eng., 2(2), 2319–526.

Gibbons, P. B., & Matias, Y. (1999, January). Synopsis Data

Structures for Massive Data Sets. In SODA (Vol. 10, pp. 909-910).

doi:10.1090/dimacs/050/02

Goodrich, M. T., Tamassia, R., & Goldwasser, M. H. (2013). Data

structures and algorithms in Python. John Wiley & Sons Ltd.

Guttag, J. V. (2021). Introduction to Computation and

Programming Using Python: With Application to Computational

Modeling and Understanding Data. MIT Press.

Guzdial M. Ericson B. (2012, February). Listening to linked lists:

Using multimedia to learn data structures. In Proceedings of the

43rd ACM technical symposium on Computer Science Education

(pp. 663-663). doi:10.1145/2157136.2157358

Hadoop, W. T. (2009). The definitive guide. O'Reilly Media.

Hao, J., & Ho, T. K. (2019). Machine learning made easy: A review

of scikit-learn package in python programming language. Journal

of Educational and Behavioral Statistics, 44(3), 348–361.

doi:10.3102/1076998619832248

Harris, C. M. (2020). Array Programming with NumPy. Nature,

357–362.

Harris, C. R., Millman, K. J., Van Der Walt, S. J., Gommers, R.,

Virtanen, P., Cournapeau, D., Wieser, E., Taylor, J., Berg, S.,

Smith, N. J., Kern, R., Picus, M., Hoyer, S., van Kerkwijk, M. H.,

Brett, M., Haldane, A., del Río, J. F., Wiebe, M., Peterson, P., &

Oliphant, T. E. (2020). Array programming with NumPy. Nature,

585(7825), 357–362. doi:10.1038/s41586-020-2649-2

Haslwanter, T. (2016). An Introduction to Statistics with Python. In

With Applications in the Life Sciences. Springer International

Publishing. doi:10.1007/978-3-319-28316-6

Havens, T. C., Bezdek, J. C., Leckie, C., Hall, L. O., &

Palaniswami, M. (2012). Fuzzy c-Means Algorithms for Very

Large Data. IEEE Transactions on Fuzzy Systems, 20(6), 1130–

1146. doi:10.1109/TFUZZ.2012.2201485

Hetland, M. L. (2017). Beginning Python: From novice to

professional. Apress. doi:10.1007/978-1-4842-0028-5

He, Y., Tan, H., Luo, W., Feng, S., & Fan, J. (2014). MR-

DBSCAN: A scalable Map Reduce-based DBSCAN algorithm for

heavily skewed data. Frontiers of Computer Science, 8(1), 83–99.

doi:10.1007/s11704-013-3158-3

Hill, C. (2020). Learning scientific programming with Python.

Cambridge University Press. doi:10.1017/9781108778039

Hunt, J. (2019). ADTs, Queues and Stacks. In A Beginners Guide

to Python 3 Programming (pp. 407–414). Springer.

doi:10.1007/978-3-030-25943-3_34

Inayat Z. Sajjad R. Anam M. Younas A. Hussain M. (2021,

November). Analysis of Comparison-Based Sorting Algorithms. In

2021 International Conference on Innovative Computing (ICIC)

(pp. 1-8). IEEE.

Ishwarappa, A. J., & Anuradha, J. (2011). A brief introduction on

Big Data 5Vs characteristics and Hadoop technology. Procedia

Computer Science, 48, 319–324. doi:10.1016/j.procs.2015.04.188

Jaishankar, B., Vishwakarma, S., Mohan, P., Singh Pundir, A. K.,

Patel, I., & Arulkumar, N. (2022). Blockchain for Securing

Healthcare Data Using Squirrel Search Optimization Algorithm.

Intelligent Automation & Soft Computing, 32(3).

Jamison, M. A., & Tariq, P. (2018). Five things regulators should

know about blockchain (and three myths to forget). The Electricity

Journal, 31(9), 20–23.

Java Tutorial Developers. (2022). Retrieved from NumPy Tutorial:

https://www.javatpoint.com/
Johansson, R. (2019). 1. Scientific Computing and Data Science

Applications with Numpy, SciPy, and Matplotlib (2nd ed.).

Urayasu-shi.

Johansson, R., & John, S. (2019). Numerical Python (Vol. 1).

Apress. doi:10.1007/978-1-4842-4246-9

Jondhale, S. R., Maheswar, R., Lloret, J., Jondhale, S. R.,

Maheswar, R., & Lloret, J. (2022). Fundamentals of wireless sensor

networks. Received Signal Strength Based Target Localization and

Tracking Using Wireless Sensor Networks, 1-19.

Jurgens, S. J., Choi, S. H., Morrill, V. N., Chaffin, M., Pirruccello,

J. P., Halford, J. L., Weng, L.-C., Nauffal, V., Roselli, C., Hall, A.

W., Oetjens, M. T., Lagerman, B., vanMaanen, D. P., Abecasis, G.,

Bai, X., Balasubramanian, S., Baras, A., Beechert, C., Boutkov, B.,

& Ellinor, P. T. (2022). Analysis of rare genetic variation

underlying cardiometabolic diseases and traits among 200,000

individuals in the UK Biobank. Nature Genetics, 54(3), 240–250.

doi:10.1038/s41588-021-01011-w

Kannan, S., Dongare, P. A., Garg, R., & Harsoor, S. S. (2019).

Describing and displaying numerical and categorical data. Airway,

2(2), 64. doi:10.4103/ARWY.ARWY_24_19

Karimov, E. (2020). Linked Lists. Data Structures and Algorithms

in Swift: Implement Stacks, Queues, Dictionaries, and Lists in Your

Apps, 41-54.

Khandare, A., Agarwal, N., Bodhankar, A., Kulkarni, A., & Mane,

I. (2023). Analysis of Python Libraries for Artificial Intelligence. In

Intelligent Computing and Networking. Proceedings of IC-ICN,

2022, 157–177.
Khriji, S., Chéour, R., & Kanoun, O. (2022). Dynamic Voltage and

Frequency Scaling and Duty-Cycling for Ultra Low-Power

Wireless Sensor Nodes. Electronics (Basel), 11(24), 4071.

doi:10.3390/electronics11244071

Kim, D. S., Tran-Dang, H., Kim, D. S., & Tran-Dang, H. (2019).

Mac protocols for energy-efficient wireless sensor networks.

Industrial Sensors and Controls in Communication Networks: From

Wired Technologies to Cloud Computing and the Internet of

Things, 141-159.

Kiran, S. V., Kaur, I., Thangaraj, K., Saveetha, V., Kingsy Grace,

R., & Arulkumar, N. (2021). Machine Learning with Data Science-

Enabled Lung Cancer Diagnosis and Classification Using

Computed Tomography Images. International Journal of Image and

Graphics.

Klarin, D., Lynch, J., Aragam, K., Chaffin, M., Assimes, T. L.,

Huang, J., Lee, K. M., Shao, Q., Huffman, J. E., Natarajan, P.,

Arya, S., Small, A., Sun, Y. V., Vujkovic, M., Freiberg, M. S.,

Wang, L., Chen, J., Saleheen, D., Lee, J. S., & Damrauer, S. M.

(2019). Genome-wide association study of peripheral artery disease

in the Million Veteran Program. Nature Medicine, 25(8), 1274–

1279. doi:10.1038/s41591-019-0492-5

Klasen, V., Bogucka, E. P., Meng, L., & Krisp, J. M. (2023). How

we see time–the evolution and current state of visualizations of

temporal data. International Journal of Cartography, 1-18.

doi:10.1080/23729333.2022.2156316

Kobbaey, T., Xanthidis, D., & Bilquise, G. (2022). Data Structures

and Algorithms with Python. In Handbook of Computer

Programming with Python (pp. 207-272). Chapman and Hall/CRC.

doi:10.1201/9781003139010-6

Kochenderfer, M. J., & Wheeler, T. A. (2019). Algorithms for

optimization. MIT Press.

Kong, Q., Siauw, T., & Bayen, A. (2020). Python programming and

numerical methods: A guide for engineers and scientists. Academic

Press.

Lafuente, D., Cohen, B., Fiorini, G., García, A. A., Bringas, M.,

Morzan, E., & Onna, D. (2021). A Gentle introduction to machine

learning for chemists: An undergraduate workshop using python

notebooks for visualization, data processing, analysis, and

modeling. Journal of Chemical Education, 98(9), 2892–2898.

doi:10.1021/acs.jchemed.1c00142

Larrier, J. H. (2021). A Brief History of Blockchain. Transforming

Scholarly Publishing With Blockchain Technologies and AI, 85-

100.

Learn To Code For IoT With Python Essentials Course. (2022).

Networking Academy.

https://www.netacad.com/courses/programming/pcap-

programming-essentials-python

Li, Z., Li, H., Ralescu, A. L., Dillman, J. R., Parikh, N. A., & He,

L. (2023). A Novel Collaborative Self-Supervised Learning Method

for Radiomic Data. arXiv preprint arXiv:2302.09807.

Liang, H., Yang, S., Li, L., & Gao, J. (2019). Research on routing

optimization of WSNs based on improved LEACH protocol.

EURASIP Journal on Wireless Communications and Networking,

2019(1), 1–12. doi:10.1186/s13638-019-1509-y

Li, H. H., Du, X. Y., & Tian, X. (2009). A review-based reputation

evaluation approach for Web services. Journal of Computer Science

and Technology, 24(5), 893–900. doi:10.1007/s11390-009-9280-x

Li, J., Chen, Q., Hu, X., Yuan, P., Cui, L., Tu, L., Cui, J., Huang, J.,

Jiang, T., Ma, X., Yao, X., Zhou, C., Lu, H., & Xu, J. (2021).

Establishment of noninvasive diabetes risk prediction model based

on tongue features and machine learning techniques. International

Journal of Medical Informatics, 149, 104429.

doi:10.1016/j.ijmedinf.2021.104429

Liu, P., Zhou, D., & Wu, N. (2007, June). VDBSCAN: varied

density based spatial clustering of applications with noise. In 2007

International conference on service systems and service

management (pp. 1-4). IEEE. doi:10.1109/ICSSSM.2007.4280175

Liu, X., Huang, G., & Mei, H. (2009). Discovering homogeneous

web service community in the user-centric web environment. IEEE

Transactions on Services Computing, 2(2), 167–181.

doi:10.1109/TSC.2009.11

Liu, Y. H. (2020). Python Machine Learning By Example: Build

intelligent systems using Python, TensorFlow 2, PyTorch, and

scikit-learn. Packt Publishing Ltd.

Liu Z. Li P. Zheng Y. (2009). Clustering to find exemplar terms for

keyphrase extraction. Proc. 2009 Conf. on Empirical Methods in

Natural Language Processing, 257-266.

doi:10.3115/1699510.1699544
Li, X., Zhang, Y., Leung, J., Sun, C., & Zhao, J. (2023).

EDAssistant: Supporting Exploratory Data Analysis in

Computational Notebooks with In Situ Code Search and

Recommendation. ACM Transactions on Interactive Intelligent

Systems, 13(1), 1–27. doi:10.1145/3545995

Lo, Y. T. (2023). Bioinformatics for DNA Sequencing Data

Analysis. In Authentication of Chinese Medicinal Materials by

DNA Technology: Techniques and Applications (pp. 89-125).

Academic Press.

Lutz, M. (2013). Learning Python: Powerful object-oriented

programming. O'Reilly Media, Inc.

Mahran, S., & Mahar, K. (2008). Using grid for accelerating

density-based clustering. CIT 2008 IEEE international conference.

doi:10.1109/CIT.2008.4594646

Maiti, M., Vyklyuk, Y., & Vuković, D. (2020). Cryptocurrencies

chaotic co‐movement forecasting with neural networks. Internet

Technology Letters, 3(3), e157.

Malik, M., Kaushik, A., & Khatana, R. (2022). Classification-

Based Prediction Techniques Using ML: A Perspective for Health

Care. In Data Science for Effective Healthcare Systems (pp. 43-56).

Chapman and Hall/CRC.

Martelli, A. (2006). Python in a Nutshell. O'Reilly Media, Inc.

Massias, H., Avila, X. S., & Quisquarter, J. J. (1999). Design of a

secure timestamping service with minimal trust requirement. In

Symposium on Information theory in the Benelux (pp. 79-86).

Werkgemeenschap voor Informatie-en Communicatietheorie.

McKinney. (2022). Python for Data Analysis. O'Reilly Media, Inc.

McKinney W. (2010, June). Data structures for statistical

computing in Python. In Proceedings of the 9th Python in Science

Conference (Vol. 445, No. 1, pp. 51-56). doi:10.25080/Majora-

92bf1922-00a

McKinney, W. (2012). NumPy Basics:Arrays and Vectorized

Computation. In W. McKinney, Python for Data Analysis. O'Reilly

Media, Inc.

Medjedovic, D., & Tahirovic, E. (2022). Algorithms and data

structures for massive datasets. Simon and Schuster.

Mehlhorn, K. (2013). Data structures and algorithms 1: Sorting and

searching (Vol. 1). Springer Science & Business Media.

Mezquita, Y., Alonso, R. S., Casado-Vara, R., Prieto, J., &

Corchado, J. M. (2021). A review of k-nn algorithm based on

classical and quantum machine learning. In Distributed Computing

and Artificial Intelligence, Special Sessions, 17th International

Conference (pp. 189-198). Springer International Publishing.

doi:10.1007/978-3-030-53829-3_20

Millman, K. J., & Aivazis, M. (2011). Python for scientists and

engineers. Computing in Science & Engineering, 13(2), 9–12.

doi:10.1109/MCSE.2011.36

Mishra, A., & Soni, S. (2023, March). Summarization of

Unstructured Medical Data for Accurate Medical Prognosis—A

Learning Approach. In Advances in Cognitive Science and

Communications: Selected Articles from the 5th International

Conference on Communications and Cyber-Physical Engineering

(ICCCE 2022), Hyderabad, India (pp. 825-838). Singapore:

Springer Nature Singapore.

Mohan, P., Neelakandan, S., Mardani, A., Maurya, S., Arulkumar,

N., & Thangaraj, K. (2023). Eagle Strategy Arithmetic

Optimisation Algorithm with Optimal Deep Convolutional Forest

Based FinTech Application for Hyper-automation. Enterprise

Information Systems. doi:10.1080/17517575.2023.2188123

Mohapatra, S., Ahmed, N., & Alencar, P. (2019, December).

KryptoOracle: A real-time cryptocurrency price prediction

platform using twitter sentiments. 2019 IEEE international

conference on big data, 5544–5551.

Molin, S., & Jee, K. (2021). Hands-On Data Analysis with Pandas:

A Python data science handbook for data collection, wrangling,

analysis, and visualization. Packt Publishing Ltd.

Moruzzi, G. (2020). Python basics and the interactive mode.

Essential python for the physicist, 1-39.

Moussaid, A., Zrira, N., Benmiloud, I., Farahat, Z., Karmoun, Y.,

Benzidia, Y., . . . Ngote, N. (2023, February). On the

Implementation of a Post-Pandemic Deep Learning Algorithm

Based on a Hybrid CT-Scan/X-ray Images Classification Applied to

Pneumonia Categories. In Healthcare (Vol. 11, No. 5, p. 662).

MDPI.

Mukhiya, S. K., & Ahmed, U. (2020). Hands-On Exploratory Data

Analysis with Python: Perform EDA techniques to understand,

summarize, and investigate your data. Packt Publishing Ltd.

Mulani, S. (2022). Python Advanced. Retrieved from Vectors in

Python:

https://www.digitalocean.com/community/tutorials/vectors-in-

python

Najma, M., & Sekar, G. (2023, February). Novel CNN Approach

(YOLO v5) to Detect Plant Diseases and Estimation of Nutritional

Facts for Raw and Cooked Foods. In Proceedings of Third

International Conference on Sustainable Expert Systems: ICSES

2022 (pp. 305-329). Singapore: Springer Nature Singapore.

doi:10.1007/978-981-19-7874-6_23

Nakamoto, S. (2008). Bitcoin: A peer-to-peer electronic cash

system. Decentralized Business Review, 21260.

Navghare, N., Kedar, P., Sangamkar, P., & Mahajan, M. (2020).

Python and OpenCV in automation of live Surveillance. In

Machine Learning and Information Processing (pp. 235–243).

Springer. doi:10.1007/978-981-15-1884-3_22

Neelakandan, S., Rene Beulah, J., Prathiba, L., Murthy, G. L. N.,

Fantin Irudaya Raj, E., & Arulkumar, N. (2022). Blockchain with

deep learning-enabled secure healthcare data transmission and

diagnostic model. International Journal of Modeling, Simulation,

and Scientific Computing, 13(04), 2241006.

doi:10.1142/S1793962322410069

NumPy Developers. (2008-2022). Retrieved from Numpy User

Guide: https://numpy.org/doc/stable/user/whatisnumpy.html

Oliphant, T. E. (2006). Guide to NumPy. Academic Press.

Oliphant, T. E. (2007). Python for scientific computing. Computing

in Science & Engineering, 9(3), 10–20.

doi:10.1109/MCSE.2007.58

Opidi, A. (2019, October 8). How to Create Your Own

Cryptocurrency Using Python. freeCodeCamp.org.

https://www.freecodecamp.org/news/create-cryptocurrency-using-

python/

Othman, J. (2023). Data structure uniqueness in Python

programming language. Enhancing Innovations in E-Learning for

Future Preparation, 68.

Parimala M, Lopez D, Senthilkumar N, (2011). A survey on

density based clustring algorithms for mining large spatial

databases. Int J Adv Sci Technol., 31-59.

Patel, M. (2018). Data Structure and Algorithm With C.

Educreation Publishing.

Perret, B., Chierchia, G., Cousty, J., Guimaraes, S. J. F., Kenmochi,

Y., & Najman, L. (2019). Higra: Hierarchical graph analysis.

SoftwareX, 10, 100335. doi:10.1016/j.softx.2019.100335

Piatkowski, J. (2020). The Conditional Multiway Mapped Tree:

Modeling and Analysis of Hierarchical Data Dependencies. IEEE

Access : Practical Innovations, Open Solutions, 8, 74083–74092.

doi:10.1109/ACCESS.2020.2988358

Popescu, G. V., & Bîră, C. (2022, June). Python-Based

Programming Framework for a Heterogeneous MapReduce

Architecture. In 2022 14th International Conference on

Communications (COMM) (pp. 1-6). IEEE.

doi:10.1109/COMM54429.2022.9817183

Rajaraman, A., & Ullman, J. D. (2012). Mining of massive

datasets. Cambridge University Press.

Ram, A., Jalal, S., & Kumar, M. (2013). A density based algorithm

for discovering density varied clusters in large spatial databases.

IJCA, 3(6), 1–4.

Rather, M. A., Agarwal, D., Bhat, T. A., Khan, I. A., Zafar, I.,

Kumar, S., Amin, A., Sundaray, J. K., & Qadri, T. (2023).

Bioinformatics approaches and big data analytics opportunities in

improving fisheries and aquaculture. International Journal of

Biological Macromolecules, 233, 123549.

doi:10.1016/j.ijbiomac.2023.123549

Roopitha C. H. Pai V. (2020, January). Survey On Media Access

Control Protocols In Wireless Sensor Network. In 2020 Fourth

International Conference on Inventive Systems and Control

(ICISC) (pp. 163-168). IEEE.

doi:10.1109/ICISC47916.2020.9171085

S, R. A. (2023, March 31). What is Blockchain Technology? How

Does Blockchain Work? [Updated]. Simplilearn.com.

https://www.simplilearn.com/tutorials/blockchain-

tutorial/blockchain-technology

Saabith, A. S., Fareez, M. M. M., & Vinothraj, T. (2019). Python

current trend applications-an overview. International Journal of

Advance Engineering and Research Development, 6(10).

Salihu A. Hoti M. Hoti A. (2022, December). A Review of

Performance and Complexity on Sorting Algorithms. In 2022

International Conference on Computing, Networking,

Telecommunications & Engineering Sciences Applications

(CoNTESA) (pp. 45-50). IEEE.

doi:10.1109/CoNTESA57046.2022.10011382

Sansano, E., Montoliu, R., & Belmonte Fernandez, O. (2020). A

study of deep neural networks for human activity recognition.

Computational Intelligence, 36(3), 1113–1139.

doi:10.1111/coin.12318

Saravanan, S., Lavanya, M., Srinivas, C. M. V., Arunadevi, M., &

Arulkumar, N. (n.d.). Secure IoT Protocol for Implementing

Classified Electroencephalogram (EEG) Signals in the Field of

Smart Healthcare. In Cyber Security Applications for Industry 4.0

(pp. 111-129). Chapman and Hall/CRC.

Sarwar, S., Sirhindi, R., Aslam, L., Mustafa, G., Yousaf, M. M., &

Jaffry, S. W. U. Q. (2020). Reinforcement learning based adaptive

duty cycling in LR-WPANs. IEEE Access : Practical Innovations,

Open Solutions, 8, 161157–161174.

doi:10.1109/ACCESS.2020.3021016

Schwarz, J. S., Chapman, C., & Feit, E. M. (2020). An Overview of

Python. Python for Marketing Research and Analytics. Springer.

doi:10.1007/978-3-030-49720-0

Shamim, G., & Rihan, M. (2023). Exploratory Data Analytics and

PCA-Based Dimensionality Reduction for Improvement in Smart

Meter Data Clustering. Journal of the Institution of Electronics and

Telecommunication Engineers, 1–10.

doi:10.1080/03772063.2023.2218317

Sharma, S., & Guleria, K. (2023). A Deep Learning based model

for the Detection of Pneumonia from Chest X-Ray Images using

VGG-16 and Neural Networks. Procedia Computer Science, 218,

357–366. doi:10.1016/j.procs.2023.01.018

Sher, D. B. (2004). A simple implementation of a queue with a

circularly linked list. SIGCSE Bulletin, 36(3), 274–274.

doi:10.1145/1026487.1008112

Shrivas, M. K., Hiran, K. K., Bhansali, A., & Doshi, R. (2022).

Advancements in Quantum Blockchain With Real-Time

Applications. IGI Global.

Soni, G., & Selvaradjou, K. (2020). Performance evaluation of

wireless sensor network MAC protocols with early sleep problem.

International Journal of Communication Networks and Distributed

Systems, 25(2), 123–144. doi:10.1504/IJCNDS.2020.108875

Souravlas, S., Anastasiadou, S., & Katsavounis, S. (2021). A survey

on the recent advances of deep community detection. Applied

Sciences (Basel, Switzerland), 11(16), 7179.

doi:10.3390/app11167179

Stančin, I., & Jović, A. (2019, May). An overview and comparison

of free Python libraries for data mining and big data analysis. In

2019 42nd International convention on information and

communication technology, electronics and microelectronics

(MIPRO) (pp. 977-982). IEEE.

doi:10.23919/MIPRO.2019.8757088
Stefan van der Walt, S. C. (2021). The NumPy Array: A Structure

for Efficient Numerical Computation. IEEE, 13(2), 22-30.

Streib, J. T., Soma, T., Streib, J. T., & Soma, T. (2017). Stacks and

Queues Using References. Guide to Data Structures: A Concise

Introduction Using Java, 173-198.

Subasi, A. (2020). Practical machine learning for data analysis

using python. Academic Press.

Su, Y., Lu, X., Zhao, Y., Huang, L., & Du, X. (2019). Cooperative

communications with relay selection based on deep reinforcement

learning in wireless sensor networks. IEEE Sensors Journal,

19(20), 9561–9569. doi:10.1109/JSEN.2019.2925719

Thiagarajan, R. (2020). Energy consumption and network

connectivity based on Novel-LEACH-POS protocol networks.

Computer Communications, 149, 90–98.

doi:10.1016/j.comcom.2019.10.006

Üçoluk, G., Kalkan, S., Üçoluk, G., & Kalkan, S. (2012).

Organizing Data. Introduction to Programming Concepts with Case

Studies in Python, 165-194.

Unpingco, J. (2021). Python Programming for Data Analysis.

Springer. doi:10.1007/978-3-030-68952-0

Upadhye, S., Neelakandan, S., Thangaraj, K., Babu, D. V.,

Arulkumar, N., & Qureshi, K. (2023). Modeling of Real Time

Traffic Flow Monitoring System Using Deep Learning and

Unmanned Aerial Vehicles. Journal of Mobile Multimedia, 477-

496.
Wiener, R. (2022). Queues and Lists. In Generic Data Structures

and Algorithms in Go: An Applied Approach Using Concurrency,

Genericity and Heuristics (pp. 187–236). Apress. doi:10.1007/978-

1-4842-8191-8_6

Wu, X., Zhu, X., & Wu, G. Q. (2014). Data mining with big data.

IEEE Transactions on Knowledge and Data Engineering, 26(1),

97–107. doi:10.1109/TKDE.2013.109

Yalçın, O. G. (2021). Complementary Libraries to TensorFlow 2.x.

Applied Neural Networks with TensorFlow 2: API Oriented Deep

Learning with Python, 81-94.

Yang Y. Li X. (2020, October). Design and Implementation of

Sliding Window with Circular Linked List for Dynamic Data Flow.

In Proceedings of the 4th International Conference on Computer

Science and Application Engineering (pp. 1-5).

doi:10.1145/3424978.3425152

Yogeshwaran S. Kaur M. J. Maheshwari P. (2019, April). Project

based learning: predicting bitcoin prices using deep learning. In

2019 IEEE Global Engineering Education Conference (EDUCON)

(pp. 1449-1454). IEEE.

Z. (2023, February 22). Top 10 Useful Python Packages to Pull

Cryptocurrency Data Seamlessly. Analytics Insight.

https://www.analyticsinsight.net/top-10-useful-python-packages-to-

pull-cryptocurrency-data-seamlessly/

Zeng, W., Shang, M. S., Zhang, Q. M., Lü, L., & Zhou, T. (2010).

Can Dissimilar Users Contribute to Accuracy and Diversity of

Personalized Recommendation. International Journal of Modern

Physics C, 21(10), 1217–1227. doi:10.1142/S0129183110015786

Zhang, H. L., Han, B. T., Xu, J. C., & Feng, Y. (2013). Preparation

of the public bus schedule based on fcfs. In Proceedings of 20th

International Conference on Industrial Engineering and

Engineering Management: Theory and Apply of Industrial

Engineering (pp. 1021-1029). Springer Berlin Heidelberg.

doi:10.1007/978-3-642-40063-6_100

Zheng Z. Zhu J. Lyu M. R. (2013). Service-generated Big Data and

Big Data-as-a-Service: An Overview. Proc. IEEE BigData, 403-

410. doi:10.1109/BigData.Congress.2013.60

Zhou, H. Y., Lu, C., Chen, C., Yang, S., & Yu, Y. (2023). A Unified

Visual Information Preservation Framework for Self-supervised

Pre-training in Medical Image Analysis. IEEE Transactions on

Pattern Analysis and Machine Intelligence, 1–16.

doi:10.1109/TPAMI.2023.3234002
 About the Contributors
|MOHAMMAD GALETY - Editor|Mohammad Gouse Galety received his

PhD in Computer Science in 2012. His research interests span both

computer science and information science. Dr Mohammad Gouse Galety

(co‐) authorizes more than 50 journal papers and publications in

international conference proceedings indexed by Springer, Scopus, two

patents, and five books. Since 1999, he has served in different national and

international organizations like Sree Vidyanikethan Degree College,

Tirupati, India; Emeralds Degree College, Tirupati, India; Brindavan

College of Engineering, Bangalore, India; Kuwait Educational Center,

Kuwait; Ambo University, Ambo, Ethiopia, Debre Berhan University,

Debre Berhan, Ethiopia, Lebanese French University, Erbil, Iraq and

Catholic University in Erbil, Erbil, Iraq. He is a professor in the Computer

Science Department at Samarkand International University of Technology,

Samarkand, Uzbekistan. In addition, he teaches several computer sciences

and information technology/science engineering to undergraduate and

postgraduate students. His research interests cover several aspects of

computer science, mainly web mining, IoT, and Networks. He is a Fellow of

the IEEE. As of 2020, Google Scholar reports over 250 citations to his

work.e502d01f-6393-42c4-8526-9d8a33b2bb9b

|Arulkumar Natarajan - Editor|Arul Kumar Natarajan is currently

working as an Assistant Professor at Christ (Deemed to be University),

Bangalore, INDIA. He received a Ph.D. degree in Computer Science from

Bharathidasan University, India, in 2019. His research areas are Computer

Networks, Cyber Security, and the Internet of Things (IoT). He published

more than 30 research papers in both journals and conferences. He

published 4 patents in physics, communication, and computer science. He

has chaired many technical sessions and delivered more than 15 invited

talks at the national and international levels. He has completed over 33

certifications from IBM, Google, Amazon, etc. He passed the CCNA:

Routing and Switching Exam in 2017. Additionally, he also passed the

Networking Fundamentals in the year 2017 exam from Microsoft.f2d007a7-

a87d-488f-882e-c486da9905f7

|Sriharsha AV - Editor|A. V. Sriharsha completed his B.Tech from AU,

M.Tech from Sathyabama University, and Ph.D. from SV University and

SCSVMV University. He holds Honorary Diploma in Computer

Programming and Applications moderated by National Computing Center,

United Kingdom. He is a member and designated as a chartered engineer by

the Institution of Engineers India, Calcutta. Member of ISOC, “The Internet

Society”, Life Member in IAENG. He has been in IT, IT-enabled services,

and Education for about 25+ years. Have 7 years of industrial experience on

strategic servers and 19 years of teaching and research. Around 40+ articles

of research have been published in journals and conferences of national and

international repute where some of them are indexed in Scopus and ESCI.

His research interests are data science, privacy-preserving data mining,

differential privacy and machine learning, big data, software architectures,

and robotics. Freelance programming and consulting for industrial

applications, web service frameworks, and cyber security solutions are the

regular track of his occupation while building research models adaptable for

industry and academia. His academic profile fairs with funded projects and

seminar grants from AICTE, ISRO, etc.. He has 3 (three) patents filed, 1

(one) design patent in the amendment, and 1 (one) copyright to his

credits.160b6754-0ced-4621-9013-07fa4118aaf1

* * *

|Keerthana Devi A - Contributing Author|Keerthana Devi A. is pursuing

MSc in Data Science at Coimbatore Institute of Technology,

Coimbatore.05401693-0e93-4b08-94f9-878f99796ad5

|Saleem Abdul Samad - Contributing Author|411020c2-a4e9-4c20-909d-

de4e83218624

|Arman Abouali - Contributing Author|30baddab-d402-4d1c-ac93-

057895edd4d8

|Varshaa D - Contributing Author|ef2be387-92a4-4a28-8ad6-f962725be4b6

|Tesfaye Fufa - Contributing Author|Tesfaye Fufa graduated from Jima

University with a Bachelor's Degree in Information Technology in July

2012 and received a Master of Science in Software Engineering from

Adama Science and Technology University. Previously, he worked from

2013–2020 as a lecturer at Debre Berhan University, Ethiopia. He is

currently working as an Associate Researcher II, Satellite Operator and

Development at the Space Science and Geospatial Institute. His research

interests include artificial intelligence, network security, cloud computing,

big data, image processing, and data science. He has written more than four

journal papers.b8a944a9-579a-4a5c-8b85-dd9ef6728b9b

|Pradeepa Ganesan - Contributing Author|Pradeepa Ganesan holds her

MCA degree in 2003 from Madurai Kamaraj University, Tamilnadu, India,

and her M.Phil degree in 2006 from Bharathidasan University, Tamilnadu,

India. She has 18 years of teaching experience. Her research interests

include cybersecurity, machine learning, and deep learning.9c197f95-1ae9-

4e3b-a695-137c30599fd3

|Agrata Gupta - Contributing Author|Agrata Gupta is currently working as

a Data Science Research scholar at Christ (Deemed to be University),

Bengaluru, India. She completed her master’s degree in Data Analytics

from Christ (Deemed to be University), Bengaluru, India, in 2022 and her

bachelor’s degree in Statistics (Hons.) from the University of Delhi, India,

in 2020. Her research areas are Blockchain technology, Cryptocurrency and

Security. She has completed nine certifications in the field of Statistics and

Data Science from different IITs and from other institutions. She also works

as a Financial Analyst in a Las Vegas-based software development

company, AXEL.789f5444-6334-4814-8b44-4ac27541623f

|Sunitha Gurram - Contributing Author|Sunitha Gurram is currently

working as Professor at School of Computing, Mohan Babu University,

Tirupati, India. Her research interests include Data Mining, Spatio-

Temporal Analytics, Machine Learning and Artificial Intelligence. She has

13 patents and 6 books published to her credit. She has published around 70

papers in reputed journals and conferences. She has been serving as

reviewer for several journals; and has served on the program committees
and co-chaired for various international conferences.81bfef00-ef2d-4b46-

b07c-904dc1dd98f7

|Garamu Iticha - Contributing Author|Garamu Iticha graduated from Jima

University with a Bachelor's Degree in Information Technology in July

2012 and received a Master of Science in Software Engineering from

Adama Science and Technology University. He is currently a lecturer at

Debre Markos University, Ethiopia. His research interests include artificial

intelligence, cloud computing, networking, and data science. He has written

more than two journal papers.6b535057-ddd9-412f-879a-3790ea55a39c

|Jency Jose - Contributing Author|3f8fe8c4-0451-4007-a6c3-a9ee2e67e57f

|Rahul Kumar - Contributing Author|554de65e-c54d-4463-bad9-

7d8113d2765c

|Anuvarshini M - Contributing Author|cdf920b3-c305-4970-9433-

c21438cd94c8

|Justin M - Contributing Author|Justin Rajasekaran M. is a Lecturer in the

Department of Information Technology at University of Technology and

Applied Sciences, Shinas, Sultanate of Oman where he has been since

2011. His research interests span both computer networking and software

engineering and quality assurance. Much of his work has been on improving

the understanding, design, and performance of computer application

systems. He also involved in developing course curriculum in the area of

Data Science and Artificial Intelligence.7f94678d-1afe-4152-9428-

983e3382f5df

|Sujithra M - Contributing Author|Sujithra M. is currently working as an

Assistant Professor, Department of Computing -Data Science at Coimbatore

Institute of Technology, Coimbatore, India. Dr.M.Sujithra received the

MCA degree from Bharathiyar University and the BSC degree in Computer

Science from Bharathiyar University and also received the Ph.D. degree in

computer science from Avinashilingam University, India. She has published

more than 85 Research articles in International Journals/Conference

proceedings .Moreover, she has published 3 books, few book chapters and 1
patent under a reputed publishers. Her research interests include the cloud

computing, Data Security, Mobile Application Development, Machine

Learning and Data Analytics and Visualization.961d97f1-faa7-4951-b36f-

8ba2c50e05a9

|Islom Mamatov - Contributing Author|ca988e57-7d6f-4b92-90f4-

3584674d0559

|Sruthi R - Contributing Author|e65d4b86-4b3c-43c3-94b0-be6916beb861

|Vinodini R - Contributing Author|65094adc-461a-4de8-927e-

1fc4204dbc2b

|Aswathi Sasidharan - Contributing Author|f79bb9ee-e608-40bd-b034-

475a51e88e04

|Rajasrkaran Selvaraju - Contributing Author|Rajasrkaran Selvaraju has

22+ years of teaching and research experience in the fields of Computer

Science and Information Technology. Presently, he is working at the

University of Technology and Applied Sciences - Ibri in the Sultanate of

Oman. He has acted as a resource person for various international

conferences, webinars and workshops. Also, serving as a reviewer for many

reputed international journals and conferences. His area of interest is not

limited to Data Science, Artificial Intelligence and Machine Learning,

Internet of Things and Software Engineering.cb1761ad-e9ea-4654-8e63-

fca913ed9b71

|Mastan Vali Shaik - Contributing Author|Mastan Vali Shaik has 20+

years of teaching and research experience in the fields of Computer Science

and Information Technology. Presently, he is working at the University of

Technology and Applied Sciences - Ibri in the Sultanate of Oman. He has

acted as a resource person for various international conferences, webinars

and workshops. Also, serving as a reviewer for many reputed international

journals and conferences. His area of interest is not limited to Data Science,

Artificial Intelligence and Machine Learning, Internet of Things and

Software Engineering.67674f66-a42b-40e9-bbb2-7b468eeb631c
|A.V. Sriharsha - Contributing Author|9457aed5-c4f0-4d13-bb2b-

bb8d71fc4b25

|Kavita Srivastava - Contributing Author|fd2ad491-a9df-4c3e-b02b-

d3fface024a6

|P Thamilselvan - Contributing Author|a0cd0ccf-2274-498a-a5a5-

4d3eaaf29cd3

|Olimjon Yalgashev - Contributing Author|Olimjon Yalgashev works at

Samarkand International University of Technology at Department of

Computer Science. He received his doctorate in France, at the

Technological University of Belfort-Montbéliard. He is the author of many

international and local scientific publications.82cdbcee-a5e0-4ed2-8166-

8731ca2dc25b