What is the fastest way to multiply a matrix with its transpose (AA^T) in Python?
I don't think NumPy/SciPy take into account the symmetries involved when using e.g. np.dot or np.matmul. The resulting matrix is always symmetric so I can imagine there is a faster solution.