ELC REPORT

submitted by

KOENA (1024030831)

B.TECH (COE)

THAPAR INSTITUTE OF ENGINEERING &

TECHNOLOGY
 TABLE OF CONTENT
1. PANDAS
2. NUMPY
3. TENSORFLOW
4. MATPLOTLIB

PANDAS
PANDAS is a software library written for the Python programming language for data
manipulation and analysis. It was released on 11th January 2008.

It has functions for analysing, cleaning, exploring, and manipulating data.

Pandas library is built on top of Numpy library, it is expected that Numpy will work with Pandas.

Key features:

 It is a DataFrame object that is quick and effective, with both standard and custom
indexing.
 The library is used for reshaping and turning of the informational indexes.
 It is used to align the data and integrate the missing data.
 Process a variety of data sets in various formats, such as matrix data, heterogeneous
tabular data, and time series.
 It incorporates with different libraries like SciPy, and scikit-learn.

Application:

Data Cleaning and Preprocessing: Handle missing values, convert data types,remove duplicates
and reshape data .

Data Exploration: Perform descriptive statistics, filter/subset data, group and aggregate and
sort data .

Data Visualization: Create plots using Pandas or integrate with Matplotlib and Seaborn.

Machine Learning: Engineer features,prepare data and integrate with ML libraries like Scikit-
learn and TensorFlow .

Time Series Analysis: Work with timestamps,resample data and perform time-specific
operations .

NUMPY
NumPy is the fundamental package for scientific computing in Python. It is a Python library that
provides a multidimensional array object, various derived objects and an assortment of
routines for fast operations on arrays

NumPy arrays facilitate advanced mathematical and other types of operations on large
numbers of data

Key features:

 Vectorized code is more concise and easier to read.

 Fewer lines of code generally mean fewer bugs.
 The code more closely resembles standard mathematical notation,making it easier to
correctly code mathematical constructs.
 Vectorization results in more “Pythonic” code. Without vectorization, our code would be
littered with inefficient and difficult to read for loops.

Application:

Data Creation: Generate arrays using functions like arrange(), zeros(),ones() and random().

Data Manipulation: Reshape, flatten and transpose arrays.

Mathematical Operations: Perform element-wise calculations,matrix operations and linear

algebra.

Statistical Functions: Compute mean,median,standard deviation and other statistical measures.

Indexing and Slicing: Access specific elements,rows or columns in arrays.

Logical Operations: Apply conditions to filter or manipulate data.

Broadcasting: Perform operations on arrays of different shapes.

TensorFlow
TensorFlow is an open-source software library for high performance numerical computation. Its
flexible architecture allows easy deployment of computation across a variety of platforms
(CPUs, GPUs, TPUs), and from desktops to clusters of servers to mobile and edge devices.

It is a library which is released on 9 November 2015 and developed by Google Brain Team. It is
entirely based on Python programming language and use for numerical computation and data
flow, which makes machine learning faster and easier.

Key features:

 It’s flexible and promotes easy modularity.

 It is great for quick debugging.
 It allows users to train several neural networks and GPUs.
 The visualizer is it enables verification of representation models.

Application:

Model Building: Create machine learning models using Sequential API or Functional API.

Tensor Operations: Perform mathematical operations on multidimensional arrays (tensors).

Model Evaluation: Evaluate performance using metrics on validation/test datasets.

Model Deployment: Export models for deployment with TensorFlow Serving or TensorFlow
Lite.

GPU Acceleration: Utilize GPUs for faster training and inference.

Custom Operations: Write custom layers, loss functions, and optimizers for specific tasks.

Integration: Easily integrate with Keras and other ML frameworks.

Scalability: Scale training across multiple GPUs/TPUs or distributed systems.

Matplotlib
Matplotlib is a comprehensive library for creating static, animated, and interactive
visualizations in Python. Matplotlib makes easy things easy and hard things possible.

It was released in 2003 by John Hunter, it enables users to graphically represent data,
facilitation easier analysis and understanding.

Key features:

 Creates a wide variety of visualization including line plots,scatterplots,bar charts and

histograms.
 It seamlessly integrates with NUMPY and plots arrays
 It can be used on several platforms like Windows,macOS and Linux etc.
 Generates aesthetic publication-ready graphs.

Application:

Diverse Plotting: Create various plots like line, scatter, bar, and histograms.

Customization: Enhance visuals with tailored titles, labels, legends, and colors.

Subplots and Layouts: Organize multiple plots within a single figure.

3D Visualizations: Represent multidimensional data with 3D plots.

Integration: Seamlessly work with Pandas, NumPy, and Seaborn.

Export Options: Save high-quality plots in formats like PNG, PDF, and SVG.

NUMPY
 A PROJECT REPORT
submitted by

KOENA (1024030831)

B.TECH (COE)

THAPAR INSTITUTE OF ENGINEERING &

TECHNOLOGY
 TABLE OF CONTENT
5. INTRODUCTION
6. FEATURES
7. APPLICATIONS
8. STRENGTHS AND WEAKNESSES
9. CONCLUSION
10. REFRENCES

INTRODUCTION
WHAT IS NUMPY?
NUMPY is an open-source python library which provides a multidimensional array
object, various derived objects and an assortment of routines for fast operations
on arrays (including mathematical, logical, shape manipulation, sorting, selecting
etc.). NUMPY easily integrates with other libraries like pandas, matplotlib, and
SciPy. Use of NUMPY is high prevalent in data science training projects and
mathematical fields.

PANDAS is a powerful, open-source python library that is used for data analysis
and manipulation. Built on top of NumPy, PANDAS provides data structures and
operations like Series and DataFrame for efficiently handling of structured data
sets (eg: TABLES) very easily. Widely used in the fields of Data Science, Artificial
Intelligence and Machine learning and various other places where data
manipulation is required.

FEATURES
 Multi-dimensional arrays
NumPy supports multi-dimensional arrays, masked arrays, and matrices.

 Mathematical functions

NumPy includes a variety of mathematical functions, such as linear algebra, statistical

operations, and random number generation.

 Vectorized operations

NumPy allows you to perform mathematical computations on entire arrays without

explicit looping. This leads to faster execution times and cleaner code.

 Broadcasting

NumPy's broadcasting mechanism allows arithmetic operations between arrays of

different shapes and sizes.

 Open source

NumPy is developed and maintained publicly on GitHub.

 Interoperable

NumPy integrates with other libraries and tools in the Python ecosystem, such as SciPy,
Matplotlib, pandas, and scikit-learn.

 Efficient

NumPy is implemented in C and Fortran, making it highly efficient for numerical

operations.

 Easy to use

NumPy has a clear and intuitive syntax, making it accessible for both beginners and
experienced developers.

APPLICATION
NUMPY is widely used in various domains

 Data Science: It’s used to handle many numerical analyses and functions.
 Machine Learning: NUMPY helps in multiple mathematical tasks and helps
in data manipulation.
 Finance Modelling: Time series and risk analysis, Excel integration and Data
visualization.
 Scientific Research: Statistical analysis, interdisciplinary research and
statistical analysis.
 Web Analytics: Mathematical optimization and linear programming.

STRENGTHS and LIMITATIONS

ADVANTAGES:
 Better speed and performance: Quick at great at number-crunching tasks.
 Efficient memory usage: It has more compact and efficient storage.
 Large Community and Documentation: Extensive documentation.
 Support with other Python Libraries: NUMPY integrates seamlessly with
many other Python libraries, making it highly versatile for a variety of tasks
like Pandas, Scikit-Learn, Matplotlib and TensorFlow.
LIMITATIONS:
 No native GPU support: Doesn’t provide direct support for GPUs and is
restricted to only CPU processing.
 Hard learning curve: For beginners to programming, NUMPY’s array
structures can be a bit daunting and overwhelming.
 Relatively small community: It has few followers compared to Pandas and
TensorFlow.
CONCLUSION
NUMPY is an essential library in Python for data analysis. NumPy’s ability to
perform complex mathematical operations on large datasets makes it an
essential tool across fields that rely on extensive numerical calculations. It has
a lower flexibility and makes it difficult to work with large datasets; however,
its vast feature sets and interoperability with other languages and libraries
make it ideal for most data-related tasks.

REFERENCES
 https://www.geeksforgeeks.org/python-numpy/
 https://www.altexsoft.com/blog/numpy-pros-and-cons/
 https://numpy.org/doc/stable/user/whatisnumpy.html
 https://medium.com/@techwithjulles/introduction-to-numpy-advantages-
and-applications-a1977b288ca5