'<' not supported between instances of 'float' and 'str' Error for Tukey HSD Test

Question

I get a strange error when running the Tukey test. I hope somebody is able to help me with this as I tried a lot. This is my dataframe:

    Name      Score
1         


        

Answer 1

You have the problem because df['Name'] contains both floats and strings AND df['Name'] is of type pandas.core.series.Series. This combination leads to an error with numpy.unique() as seen from traceback. You can fix the problem with 2 ways.

tukey = pairwise_tukeyhsd(endog=df['Score'].astype('float'),
                          groups=list(df['Name']),  # list instead of a Series
                          alpha=0.05)

OR

Make sure df['Name'] contains only numbers or only strings.