change matlab linear index to numpy.

I have a three-dimensional matrix X with dimensions 16 x 10 x 1200. The original program used

    for i = 3876:10000
        My(:,:,i) = X(:,i)*X(:,i)';  
    end

Obviously, this program uses linear indexing in Matlab. I have read the relevant documentation for Matlab and NumPy, but I still do not know how to convert it to the corresponding operation in NumPy. Can anyone help me? Thank you very much.

Answer 1

[Kautuk Raj]

This can be the equivalent NumPy code for the given MATLAB code:

import numpy as np
X = np.random.rand(16, 10, 1200)
My = np.zeros((16, 16, 10000-3876+1))
for i in range(3876, 10001):
    My[:, :, i-3876] = np.outer(X[:, i-1], X[:, i-1])

Note that we use i-1 as the index for X since Python uses 0-based indexing.