Why numpy has two ways to resize viz "numpy.resize" and "ndarray.resize"?

Why it didn't went with a single way with some extra parameters to incorporate the minor difference between both the ways?