Label Encoder - Use of Inverse_transform function

Question

I'm trying to figure it out how to use the inverse_transform function from LabelEncoder(). For example, in the below code,

from sklearn.preprocessing import LabelEncoder
le = LabelEncoder()
df['Label'] = le.fit_transform(df[['Actual']]

If i want to reverse, i can simply call:

le.inverse_transform(df['Label'])

However, i need to apply that same transformation/inverse into a new dataset, which might be predicted from the model above. I.e, it is been done in a new notebook, so, it seems like i have to store the labels. Any ideas how to do this? My only idea is to export a dataframe with 2 columns, and use pd.merge.

score 0 · Answer 1 · answered Jun 10 '21 at 18:09

0

Make a dictionary containing the inverse transform of that LabelEncoder that you used in 1st notebook. here
And then use that dictionary to remap values in 2nd notebook. here

answered Jun 10 '21 at 18:09

Adarsh Wase

1,727
3
12
26

Label Encoder - Use of Inverse_transform function

1 Answers1