ValueError: observed and expected need to have the same number of observations, or e needs to add to 1

Question

This ValueError occurs during the implementation of powerdiscrepancy test form statsmodels.stats.gof

Below is a function for printing out results of this test. However, statsmodels throws error on the different sizes of sets even if they are actually the same.

def discrepancy_test(Y,
                     result):
    """
    Calculate discrepancy test for the variables of the classification problem
    
    Parameters
    ----------
    Y : np.array
        Array of true labels
    result : Result instance
        Fitted classification model

    """
    from statsmodels.stats.gof import powerdiscrepancy
    print(Y.shape, result.predict().shape)
    dtest = powerdiscrepancy(Y, result.predict())
    pd.DataFrame(dtest, columns = ['variables', 'p_value'])


discrepancy_test(Y, result_pr)

(140,) (140,)
---------------------------------------------------------------------------
ValueError                                Traceback (most recent call last)
Cell In[120], line 1
----> 1 discrepancy_test(Y, result_pr)

Cell In[119], line 15, in discrepancy_test(Y, result)
      3 """
      4 Calculate discrepancy test for the variables of the classification problem
      5 
   (...)
     12 
     13 """
     14 print(Y.shape, result.predict().shape)
---> 15 dtest = powerdiscrepancy(result.predict(), Y)
     16 pd.DataFrame(dtest, columns = ['variables', 'p_value'])

File c:\Users\user\AppData\Local\Programs\Python\Python310\lib\site-packages\statsmodels\stats\gof.py:153, in powerdiscrepancy(observed, expected, lambd, axis, ddof)
    151     e = nt * e
    152 else:
--> 153     raise ValueError('observed and expected need to have the same '
    154                      'number of observations, or e needs to add to 1')
    155 k = o.shape[axis]
    156 if e.shape[axis] != k:

ValueError: observed and expected need to have the same number of observations, or e needs to add to 1