How do I remove outliers from a pandas DataFrame that has both numerical and non-numerical data

Question

I have a dataframe (cgf) that looks as follows and I want to remove the outliers for only the numerical columns:

    Product          ob