为什么我不能在for循环中直接操作pandas DataFrame？

Question

full_data = [train, test]
for dataset in full_data:
     dataset = pd.get_dummies(dataset)

I am trying to do one-hot-encoding of both my train & test pandas DataFrames. It turned out that neither the train or the test DataFrame was changed. Could I get some help here on why this does not work? Thanks a lot!

Answer 1

You need:

full_data = pd.concat([train,test])

df_new = pd.get_dummies(full_data, 
                        columns=categorical_columns, 
                        drop_first=True)