Compute the expected frequencies from a contingency table.
Given an ndimensional contingency table of observed frequencies, compute the expected frequencies for the table based on the marginal sums under the assumption that the groups associated with each dimension are independent.
Parameters :  observed : array_like


Returns :  expected : ndarray of type numpy.float64, same shape as observed.

Examples
>>> observed = np.array([[10, 10, 20],[20, 20, 20]])
>>> expected_freq(observed)
array([[ 12., 12., 16.],
[ 18., 18., 24.]])