Numpy - Compute column-wise Mean of large 2D array using For loop and Optimization

Last update on December 21 2024 07:37:04 (UTC/GMT +8 hours)

Write a NumPy program that generates a large 2D NumPy array and write a function to compute the column-wise mean using a for loop. Optimize it using NumPy's built-in functions.

Sample Solution:

Python Code:

import numpy as np

# Generate a large 2D NumPy array with random integers
large_2d_array = np.random.randint(1, 1000, size=(1000, 1000))

# Function to compute column-wise mean using a for loop
def column_wise_mean_with_loop(arr):
    num_rows, num_cols = arr.shape
    col_means = np.empty(num_cols)
    for col in range(num_cols):
        col_sum = 0
        for row in range(num_rows):
            col_sum += arr[row, col]
        col_means[col] = col_sum / num_rows
    return col_means

# Compute column-wise mean using the for loop method
col_mean_with_loop = column_wise_mean_with_loop(large_2d_array)

# Compute column-wise mean using NumPy's built-in functions
col_mean_with_numpy = np.mean(large_2d_array, axis=0)

# Display first 10 column means to verify
print("First 10 column means using for loop:")
print(col_mean_with_loop[:10])

print("First 10 column means using NumPy:")
print(col_mean_with_numpy[:10])

Output:

First 10 column means using for loop:
[487.134 519.27  502.076 486.791 506.818 505.29  507.058 495.108 508.137
 501.575]
First 10 column means using NumPy:
[487.134 519.27  502.076 486.791 506.818 505.29  507.058 495.108 508.137
 501.575]

Explanation:

• Importing numpy: We first import the numpy library for array manipulations.
• Generating a large array: A large 2D NumPy array with random integers is generated.
• Defining the function: A function column_wise_mean_with_loop is defined to compute the column-wise mean using a for loop.
• Computing with loop: The column-wise mean is calculated using the for loop method by summing each column and dividing by the number of rows.
• Computing with numpy: The column-wise mean is calculated using NumPy's built-in mean() function with axis=0.
• Displaying results: The first 10 column means from both methods are printed out to verify correctness.

Python-Numpy Code Editor: