Extract with np. where

Question

Why can\'t you do complex functions with np where?

This leads to succesful replacement:

df[\'helper_column\'] = df.column2.str.extract(\'([a-zA-Z ]+)\')
df