The following code

import numpy as np import pandas as pd hor= [str(k) for k in range(3)] columns=[ 'A','B','C', 'D' ] ind1=np.repeat(hor, len(columns)) ind2=np.tile(columns,4) df=pd.DataFrame(index=[ind1,ind2],columns=columns) print(df) print(df.xs('A',level=1)) 

generates an error when trying to trying to slice the dataframe. The slicing syntax is correct, but the issue appears because the dataframe is invalid:
len(ind1)= 12 and len(ind2)=16.
I think an error should be generated upon creation of the dataframe.