BHARATIYA VIDYA BHAVAN’S

SARDAR PATEL INSTITUTE OF TECHNOLOGY

Bhavan’s Campus, Munshi Nagar, Andheri (West), Mumbai – 400058-India
Department of Computer Science and Engineering

Name Sanchit Ravindra Raut

UID no. 2023800095

Experiment No. 3

AIM: Perform operations using NUMPY on given dataset

Program 1

PROBLEM a. Use NumPy to load the dataset into an array for processing.
STATEMENT :

PROGRAM: import numpy as np

# Load dataset as a NumPy array

arr = np.loadtxt("/content/data.txt",
delimiter=",")
print('loaded dataset \n',arr)

RESULT :
np.loadtxt() reads numerical data from a file, using , as the delimiter to separate values.

Output:

Program 2

PROBLEM # Calculate mean, median, standard deviation, and variance

STATEMENT :

PROGRAM: # Calculate mean, median, standard deviation, and

variance
mean = np.mean(arr) # Average value
median = np.median(arr) # Middle value in sorted
 BHARATIYA VIDYA BHAVAN’S
SARDAR PATEL INSTITUTE OF TECHNOLOGY
Bhavan’s Campus, Munshi Nagar, Andheri (West), Mumbai – 400058-India
Department of Computer Science and Engineering

data
std = np.std(arr) # Spread of data
var = np.var(arr) # Variance (square of std
deviation)

print("Mean:", mean)
print("Median:", median)
print("Standard Deviation:", std)
print("Variance:", var)

RESULT :
Built-in NumPy functions compute key statistical measures to analyze data distribution.

Output:

Program 3

PROBLEM # Change the shape of the array to a specified dimension, such as converting a
STATEMENT: 1D

PROGRAM: # Print current shape of array

print(arr.shape)

# Convert array into a 1D array with 60 elements

arr_1d = arr.reshape(60)

# Copy the reshaped array

arr = arr_1d.copy()

print(arr_1d)

RESULT :
The reshape() function modifies the array shape without changing data, and copy() ensures the
original array remains unchanged.
OUTPUT:
 BHARATIYA VIDYA BHAVAN’S
SARDAR PATEL INSTITUTE OF TECHNOLOGY
Bhavan’s Campus, Munshi Nagar, Andheri (West), Mumbai – 400058-India
Department of Computer Science and Engineering

Program 4

PROBLEM # Perform element-wise addition, subtraction, multiplication, and division on

STATEMENT: the array

PROGRAM: # Create a copy of the array

arr2 = arr.copy()

# Perform element-wise operations

print('sum : ', arr + arr2) # Addition
print('sub : ', arr - arr2) # Subtraction
print('mul : ', arr * arr2) # Multiplication
print('div : ', arr / arr2) # Division

RESULT:
NumPy applies arithmetic operations element-wise, meaning each element interacts with its
corresponding element.
OUTPUT:

Program 5

PROBLEM # Extract specific elements, rows, or columns from the array using slicing
STATEMENT:
 BHARATIYA VIDYA BHAVAN’S
SARDAR PATEL INSTITUTE OF TECHNOLOGY
Bhavan’s Campus, Munshi Nagar, Andheri (West), Mumbai – 400058-India
Department of Computer Science and Engineering

PROGRAM: # Reshape array into 4 rows and 15 columns

arr = arr.reshape(4, 15)

# Extract the first row (all columns of row index

0)
row = arr[0, :]

# Extract the first column (all rows of column

index 0)
col = arr[:, 0]

print('row 1 :', row)

print('col 1 :', col)

RESULT:
NumPy slicing (:) is used to extract specific rows and columns from a reshaped 2D array.
OUTPUT:

Program 6

PROBLEM # Apply a condition to the array to filter out elements that meet certain criteria.
STATEMENT:

PROGRAM: # Select elements that are less than 4

print('elements less than 4 :', arr[arr < 4])

RESULT:
Boolean indexing (arr < 4) filters elements based on a condition, returning values that meet the
criteria.

OUTPUT:
 BHARATIYA VIDYA BHAVAN’S
SARDAR PATEL INSTITUTE OF TECHNOLOGY
Bhavan’s Campus, Munshi Nagar, Andheri (West), Mumbai – 400058-India
Department of Computer Science and Engineering

Program 7

PROBLEM # Sort the array in ascending or descending order.

STATEMENT:

PROGRAM: # Convert back to 1D and sort in ascending order

arr = arr.reshape(60)
arr = np.sort(arr)

print(arr)

RESULT:
The np.sort() function arranges the array elements in increasing order.

OUTPUT:

Program 8

PROBLEM # Identify and count the unique elements present in the array.
STATEMENT:

PROGRAM: # Get unique elements in the array

print(np.unique(arr))

# Get unique elements and their counts

uni = np.unique(arr, return_counts=True)
print('unique elements : ', uni[0])
print('unique count : ', uni[1])

RESULT:
np.unique() identifies distinct elements, and return_counts=True returns the frequency of each
unique value.
 BHARATIYA VIDYA BHAVAN’S
SARDAR PATEL INSTITUTE OF TECHNOLOGY
Bhavan’s Campus, Munshi Nagar, Andheri (West), Mumbai – 400058-India
Department of Computer Science and Engineering

OUTPUT:

Program 9

PROBLEM # If applicable, perform matrix multiplication between two arrays.

STATEMENT:

PROGRAM: # Reshape array to 15x4 for matrix multiplication

arr = arr.reshape(15, 4)

# Multiply the transposed array with itself

print('multiplication :', np.dot(arr.T, arr))

RESULT:
np.dot() performs matrix multiplication. The array is transposed (arr.T) before multiplication to
match dimensions.

OUTPUT:

Program 10

PROBLEM # Demonstrate NumPy's broadcasting by performing operations between arrays

STATEMENT: of different shapes.

PROGRAM: # Add a smaller array (first 4 elements of arr_1d)

using broadcasting
print(arr + arr_1d[:4])
 BHARATIYA VIDYA BHAVAN’S
SARDAR PATEL INSTITUTE OF TECHNOLOGY
Bhavan’s Campus, Munshi Nagar, Andheri (West), Mumbai – 400058-India
Department of Computer Science and Engineering

RESULT:
Broadcasting allows NumPy to expand smaller arrays automatically to match larger arrays for
element-wise operations.

OUTPUT:

Program 11

PROBLEM # Given stock prices for a week, calculate the daily percentage returns and find
STATEMENT: the highest gain and worst loss. Sample stock prices (closing prices for 7 days)

PROGRAM: # Stock prices over a week

prices = np.array([150, 153, 149, 155, 160, 162,
158])

# Compute daily percentage returns

daily_returns = (prices[1:] - prices[:-1]) /
prices[:-1] * 100

# Find highest gain and worst loss

highest_gain = np.max(daily_returns)
 BHARATIYA VIDYA BHAVAN’S
SARDAR PATEL INSTITUTE OF TECHNOLOGY
Bhavan’s Campus, Munshi Nagar, Andheri (West), Mumbai – 400058-India
Department of Computer Science and Engineering

worst_loss = np.min(daily_returns)

print("Daily Returns:", daily_returns)

print("Highest Gain:", highest_gain)
print("Worst Loss:", worst_loss)

RESULT:
Stock returns are calculated using (new - old) / old * 100. np.max() and np.min() find the highest
and lowest daily changes.

OUTPUT:

Program 12

PROBLEM # solve equations

STATEMENT: # 3x + 2y -z =-2
# 12x -2y + 4z = -2
# -x + 0.5y - z = 0

PROGRAM: # Define the coefficient matrix

A = np.array([[3, 2, -1],
[12, -2, 4],
[-1, 0.5, -1]])

# Define the constants

b = np.array([-2, -2, 0])

# Solve for x, y, and z

x = np.linalg.solve(A, b)

print("Solution:", x)

RESULT:
np.linalg.solve(A, b) uses matrix algebra to solve the system of linear equations Ax = b.
 BHARATIYA VIDYA BHAVAN’S
SARDAR PATEL INSTITUTE OF TECHNOLOGY
Bhavan’s Campus, Munshi Nagar, Andheri (West), Mumbai – 400058-India
Department of Computer Science and Engineering

OUTPUT:

Program 13

PROBLEM # Problem Statement: You have two numpy arrays representing the scores of
STATEMENT: two different exams for the same group of students. Your tasks are:
# a. Create the Arrays: Define two arrays exam1 and exam2 with the given
scores.
# b. Compute the Average Scores: Calculate the average score for each exam.
# c. Combine the Scores: Create a new array where each student's combined
score is the sum of their scores in both exams.
# d. Find the Number of Students with a Combined Score Greater Than 160:
Count how many students have a combined score exceeding 160.

PROGRAM: # Exam scores of students

exam1 = np.array([78, 85, 92, 88, 76, 95, 89, 84])
exam2 = np.array([82, 90, 85, 91, 80, 92, 87, 86])

# Compute average scores

average_exam1 = np.mean(exam1)
average_exam2 = np.mean(exam2)

# Compute total scores for each student

combined_scores = exam1 + exam2

# Count students with total score > 160

students_above_160 = np.sum(combined_scores > 160)

print("Average Exam 1 Score:", average_exam1)

print("Average Exam 2 Score:", average_exam2)
print("Combined Scores:", combined_scores)
print("Students with combined score > 160:",
students_above_160)
 BHARATIYA VIDYA BHAVAN’S
SARDAR PATEL INSTITUTE OF TECHNOLOGY
Bhavan’s Campus, Munshi Nagar, Andheri (West), Mumbai – 400058-India
Department of Computer Science and Engineering

RESULT:
np.mean() calculates averages, and np.sum(condition) counts students scoring above 160.

OUTPUT:

Program 14

PROBLEM You are working with data from multiple sensors recorded over time. The data
STATEMENT: is in a 1D array, but you need to reshape it to analyze it as a matrix of sensor
readings over multiple time steps.
# a. Create a 1D Sensor Data Array: Define a 1D array of length 60, where each
value # represents a sensor reading.
# b. Reshape to 3x4x5: Reshape this 1D array into a 3D array with shape
3x4x5, where 3 represents time steps, 4 represents different sensors, and 5
represents readings at each sensor.
# c. Display the Reshaped Array: Show the resulting 3D array.

PROGRAM: # Create a 1D array of sensor readings

sensor_data = np.arange(1, 61)

# Reshape into a 3D array with dimensions (3,4,5)

reshaped_data = sensor_data.reshape(3, 4, 5)

print("Reshaped Sensor Data (3x4x5):")

print(reshaped_data)

RESULT:
np.arange(1, 61) generates values, and reshape(3,4,5) organizes them into a 3D array.

OUTPUT:
 BHARATIYA VIDYA BHAVAN’S
SARDAR PATEL INSTITUTE OF TECHNOLOGY
Bhavan’s Campus, Munshi Nagar, Andheri (West), Mumbai – 400058-India
Department of Computer Science and Engineering

Program 15

PROBLEM # Write a NumPy program to create a 3x4 matrix filled with values from 10 to
STATEMENT: 21.

PROGRAM: # Generate a 3x4 matrix with values 10 to 21

matrix = np.arange(10, 22).reshape(3, 4)

print("Matrix (3x4) with values from 10 to 21:")

print(matrix)

RESULT:
np.arange(10, 22) creates numbers from 10 to 21, and reshape(3,4) organizes them into a matrix.

OUTPUT:

Program 16
 BHARATIYA VIDYA BHAVAN’S
SARDAR PATEL INSTITUTE OF TECHNOLOGY
Bhavan’s Campus, Munshi Nagar, Andheri (West), Mumbai – 400058-India
Department of Computer Science and Engineering

PROBLEM # Problem Statement: You have an array of temperatures recorded over a week.
STATEMENT: # a. Create the Array: Define a numpy array with the given temperatures.
# b. Find the Average Temperature: Calculate the average temperature over the
week.
# c. Filter Temperatures Above Average: Create an array with temperatures that
are above # the average temperature.
# d. Find the Number of Days with Temperatures Below Freezing: Count how
many days # had temperatures below 0°C.

PROGRAM: # Weekly temperature readings

temperatures = np.array([12, 15, 11, 10, 9, 14,
13])

# Compute average temperature

average_temperature = np.mean(temperatures)

# Filter temperatures above average

above_average = temperatures[temperatures >
average_temperature]

# Count days with temperatures below freezing

(0°C)
below_freezing_count = np.sum(temperatures < 0)

print("Average Temperature:", average_temperature)

print("Temperatures Above Average:",
above_average)
print("Number of Days Below Freezing:",
below_freezing_count)

RESULT:
np.mean() computes the average temperature, Boolean indexing selects values above the average,
and np.sum() counts freezing days.

OUTPUT:
 BHARATIYA VIDYA BHAVAN’S
SARDAR PATEL INSTITUTE OF TECHNOLOGY
Bhavan’s Campus, Munshi Nagar, Andheri (West), Mumbai – 400058-India
Department of Computer Science and Engineering

Program 17

PROBLEM # 8) Problem Statement: You are working with grayscale images in a 2D matrix
STATEMENT: format. You need to prepare these images for machine learning models, which
often require the data in a flattened 1D format.
# a. Create a Grayscale Image Matrix: Define a 4x4 grayscale image matrix
with random
# values between 0 and 255.
# b. Flatten the Image Matrix: Convert the 2D image matrix into a 1D array.
# c. Display the Results: Show the original 2D image matrix and the flattened
1D array.

PROGRAM: # Create a 4x4 grayscale image matrix with random

values (0-255)
image_matrix = np.random.randint(0, 256, (4, 4))

# Flatten the 2D image into a 1D array

flattened_image = image_matrix.flatten()

print("Original 2D Image Matrix:")

print(image_matrix)
print("\nFlattened 1D Image Array:")
print(flattened_image)

RESULT:
np.random.randint(0, 256, (4, 4)) generates random grayscale pixel values, and flatten() converts
the 2D matrix into a 1D array.

OUTPUT:
 BHARATIYA VIDYA BHAVAN’S
SARDAR PATEL INSTITUTE OF TECHNOLOGY
Bhavan’s Campus, Munshi Nagar, Andheri (West), Mumbai – 400058-India
Department of Computer Science and Engineering

Program 18

PROBLEM Write a NumPy program to append values to the end of an array.

STATEMENT:

PROGRAM: # Define an array

arr = np.array([1, 2, 3, 4])

# Append new values to the array

arr_appended = np.append(arr, [5, 6, 7])

print("Original Array:", arr)

print("Array After Appending:", arr_appended)

RESULT:
np.append() adds new elements to an existing array without modifying the original.

OUTPUT:

Program 19

PROBLEM # Given a 2D array of shape (3, 5) and a 1D array of shape (3, ). Write a Numpy
STATEMENT: program that transposes the 2D array and add the 1D array to each row of the
transposed array

PROGRAM: # Define a 3x5 matrix

array_2d = np.array([[1, 2, 3, 4, 5],
 BHARATIYA VIDYA BHAVAN’S
SARDAR PATEL INSTITUTE OF TECHNOLOGY
Bhavan’s Campus, Munshi Nagar, Andheri (West), Mumbai – 400058-India
Department of Computer Science and Engineering

[6, 7, 8, 9, 10],
[11, 12, 13, 14, 15]])

# Define a 1D array
array_1d = np.array([10, 20, 30])

# Transpose the 2D array

transposed_array = array_2d.T

# Add the 1D array to each row of the transposed

array
result_array = transposed_array +
array_1d.reshape(1, -1)

print("Transposed Array:")
print(transposed_array)
print("\nResult after Adding 1D Array to Each
Row:")
print(result_array)

RESULT:
The 2D array is transposed using .T, and reshape(1,-1) adjusts the 1D array shape for broadcasting.

OUTPUT:

Program 20
 BHARATIYA VIDYA BHAVAN’S
SARDAR PATEL INSTITUTE OF TECHNOLOGY
Bhavan’s Campus, Munshi Nagar, Andheri (West), Mumbai – 400058-India
Department of Computer Science and Engineering

PROBLEM Write a NumPy program to create a 3D array of shape (3, 3, 3), then reshape it
STATEMENT: into a 1D array using ravel () and finally print the result.

PROGRAM: # Create a 3x3x3 array with random values (0-9)

array_3d = np.random.randint(0, 10, (3, 3, 3))

# Flatten the 3D array to 1D

reshaped_array = array_3d.ravel()

print("Original 3D Array:")
print(array_3d)
print("\nReshaped 1D Array:")
print(reshaped_array)

RESULT:
np.random.randint() generates random numbers, and ravel() flattens the 3D array into a 1D array.

OUTPUT:

Program 21

PROBLEM # 12) Write a NumPy program that creates a 1D array of 16 elements, reshape it
STATEMENT: to (4,4), then change its shape to (2, 8) without changing the underlying data.

PROGRAM: # Create a 1D array with values from 1 to 16

array_1d = np.arange(1, 17)
 BHARATIYA VIDYA BHAVAN’S
SARDAR PATEL INSTITUTE OF TECHNOLOGY
Bhavan’s Campus, Munshi Nagar, Andheri (West), Mumbai – 400058-India
Department of Computer Science and Engineering

# Reshape to (4x4)
array_2d = array_1d.reshape(4, 4)

# Reshape to (2x8) without changing data

reshaped_2d = array_1d.reshape(2, 8)

print("Original 1D Array:")
print(array_1d)
print("\nReshaped to (4, 4):")
print(array_2d)
print("\nReshaped to (2, 8):")
print(reshaped_2d)

RESULT:
reshape(4,4) and reshape(2,8) change the structure of the array without modifying its values.

OUTPUT:

Program 22

PROBLEM 13) A company records daily sales for a week. Compute the total revenue, find
STATEMENT: which day had the highest and lowest sales, and calculate the rolling 3-day
average.

PROGRAM: # Daily sales data (in dollars)

sales = np.array([200, 450, 700, 300, 500, 650,
900])
 BHARATIYA VIDYA BHAVAN’S
SARDAR PATEL INSTITUTE OF TECHNOLOGY
Bhavan’s Campus, Munshi Nagar, Andheri (West), Mumbai – 400058-India
Department of Computer Science and Engineering

# Calculate total revenue

total_revenue = np.sum(sales)

# Find day with highest and lowest sales

highest_sales_day = np.argmax(sales)
lowest_sales_day = np.argmin(sales)

# Compute rolling 3-day average sales

rolling_3_day_avg = np.convolve(sales,
np.ones(3)/3, mode='valid')

print("Total Revenue:", total_revenue)

print("Highest Sales Day (index):",
highest_sales_day)
print("Lowest Sales Day (index):",
lowest_sales_day)
print("Rolling 3-Day Average Sales:",
rolling_3_day_avg)

RESULT:
np.sum() calculates total sales, np.argmax() and np.argmin() identify the highest and lowest sales
days, and np.convolve() computes a rolling average.

OUTPUT:

CONCLUSION: In this Python lab experiment, I learned how to perform numerical

operations using NumPy. I gained hands-on experience with array manipulation,
mathematical calculations, and handling multidimensional arrays, which are essential for
data analysis. Performed various built-in functions and imported functions of Numpy
Library.