This line
breaks when y is 1d, does not even throw the error but breaks because y.shape[1] does not exist (index error)
#%% do MGC stuff...
---------------------------------------------------------------------------
IndexError Traceback (most recent call last)
in
8 node_latent = np.concatenate((node_latent_pop1, node_latent_pop2), axis=0)
9 pop_indicator = np.array(n_graphs * [0] + n_graphs * [1])
---> 10 u, v = k_sample_transform(node_latent, pop_indicator, is_y_categorical=True)
11 mgc = MGC()
12 p_val, meta = mgc.p_value(u, v, 100)
/miniconda3/envs/dos_and_donts/lib/python3.7/site-packages/mgcpy/hypothesis_tests/transforms.py in k_sample_transform(x, y, is_y_categorical)
34 assert x.shape == y.shape, "Matrices X and Y need to be of same dimensions [n, p]"
35 else:
---> 36 assert x.shape[0] == y.shape[0] and y.shape[1] == 1, "Matrices X and Y need to be of dimensions [n, p], [n, 1]"
37
38 if not is_y_categorical:
IndexError: tuple index out of range
This line breaks when
y
is 1d, does not even throw the error but breaks becausey.shape[1]
does not exist (index error)I'd consider using this or something more standardized to do input checking, (n,) dimensional should be fine IMO https://scikit-learn.org/stable/modules/generated/sklearn.utils.check_X_y.html