Time comparison of a NumPy operation and a corresponding Python code

Let us create the below function (multiply_loops) which takes two arrays as input and computes their multiplication using normal Python way.

def multiply_loops(A, B):
    c=np.zeros((A.shape[0], B.shape[1]))
    for i in range(A.shape[0]):
        for k in range(B.shape[1]):
            c[i,k] = 0
            for j in range(B.shape[0]):
                n = A[i,j] * B[j,k]
                c[i,k] += n
    return c

Now, let us create the below function (multiply_vector) which takes two arrays as input and computes their multiplication using NumPy's vector multiplication way.

def multiply_vector(A, B):
     return A @ B

Let us create two randomly generated 100x100 matrices - X and Y - to test the above functions

X = np.random.random((100, 100))
Y = np.random.random((100, 100))

Now execute the below command (timeit) in Jupyter, which will output you the time taken by each of these functions

%timeit multiply_loops(X, Y)
%timeit multiply_vector(X, Y)