How to find most frequent values in numpy ndarray?

Question

I have a numpy ndarray with shape of (30,480,640), the 1th and 2th axis representing locations(latitude and longitute), the 0th axis contains actual data points.I want to us

Answer 1

Explaining @ecatmurs part

u[np.argmax(np.apply_along_axis(np.bincount, axis, indices.reshape(arr.shape),
                                None, np.max(indices) + 1), axis=axis)]

a little bit more and restructuring it to be more concise when re-reading it (because I used this solution and after a few weeks I was wondering what had happened in this function):

axis = 0
uniques, indices = np.unique(arr, return_inverse=True)

args_for_bincount_fn = None, np.max(indices) + 1
binned_indices = np.apply_along_axis(np.bincount,
                            last_axis, 
                            indices.reshape(arr.shape),
                            *args_for_bincount_fn)

most_common = uniques[np.argmax(binned_indices,axis=axis)]