In the assignment code template, it converts the list of document vectors into a 2D array with np.vstack. Is this more efficient than using np.array directly? The results are the same.
np.stack is a dedicated function for exactly what we want and it eliminates some possible questions what to do when arrays have different dimensions.