Label Encoding with LabelBinarizer in Scikit-Learn
Python
Transforms the column 'target' that contains three distinct classes to three columns with a binary value indicating whether which class the row in question belongs to.
1| from sklearn.preprocessing import LabelBinarizer 2| 3| # Step 1: Initialise and fit label binarizer 4| binarizer = LabelBinarizer(neg_label=0, pos_label=1) 5| binarizer.fit(df['target']) 6| print(binarizer.classes_) 7| 8| # Step 2: Transforms target column and assigns to array 9| binarized_target = binarizer.transform(df['target']) 10| 11| # Step 3: Convert array to DataFrame with column names and join back to 12| # original DataFrame 13| column_names = ['class_1', 'class_2', 'class_3'] 14| binarized_target = pd.DataFrame(binarized_target, columns=column_names) 15| df = pd.concat([df,binarized_target],axis=1) 16| 17| """ 18| 19| BEFORE TRANSFORMATION: 20| Target 21| 3 22| 2 23| 2 24| 25| AFTER TRANSFORMATION: 26| class_1 class_2 class_3 27| 0 0 1 28| 0 1 0 29| 0 1 0 30| """
133
121
117
109