Efficient pairwise correlation for two matrices of features

Question

In Python I need to find the pairwise correlation between all features in a matrix A and all features in a matrix B. In particular, I am interestin

Answer 1

Adding on to the above answer from personal experience,

p1 = N*np.dot(B.T,A)

worked way faster for me when compared to

p1 = N*np.einsum('ij,ik->kj',A,B)

This was especially true when A and B are large dimensional matrices.