How do I decode the multiclass when I use a LabelEncoder for my target variable in sklearn?

Question

I have a multi classification problem which I am trying to solve and the target variable contains genders(male, female). I have used a LabelEncoder from sklearn to implement the one hot encoding. I plotted a confusion matrix, and I got an output containing 0's and 1's instead of classes. How would I be able to get back which class was mapped to which binary number(0,1 etc.)? Any suggestions would be very much helpful.

from sklearn.preprocessing import LabelEncoder

encoder = LabelEncoder()

y = encoder.fit_transform(y)

score 0 · Answer 1 · answered May 24 '20 at 22:22

0

Just use inverse_transform:

from sklearn import preprocessing

le = preprocessing.LabelEncoder()
data = ['apple', 'orange', 'pinaple', 'orange']
le.fit(data)

encoded = le.transform(data)

>>> [0 1 2 1]

decoded = le.inverse_transform(encoded)

>>> ['apple' 'orange' 'pinaple' 'orange']

Also try to read the documentation before asking, there are a lot of examples: https://scikit-learn.org/stable/modules/generated/sklearn.preprocessing.LabelEncoder.html

answered May 24 '20 at 22:22

Sebastian

551
2
8

someone could also say try to search for duplicates before answering, there are a lot of [examples](https://stackoverflow.com/questions/58217005/how-to-reverse-label-encoder-from-sklearn-for-multiple-columns) ;) – Nicolas Gervais May 24 '20 at 22:36
that's exactly what you should do! – Nicolas Gervais May 24 '20 at 22:41
Thanks guys for the help :) – Ravi Vemuri May 24 '20 at 22:50

How do I decode the multiclass when I use a LabelEncoder for my target variable in sklearn?

1 Answers1