OneHotEncoder_example.py

import pandas as pd
import numpy as np
from sklearn.preprocessing import OneHotEncoder

# https://scikit-learn.org/stable/modules/generated/sklearn.preprocessing.OneHotEncoder.html#sklearn.preprocessing.OneHotEncoder
# https://datascienceparichay.com/article/numpy-hstack/

if __name__ == '__main__':

    array = [['Male',   1, 'Green',  12, 5],
             ['Female', 2, 'Yellow', 13, 6],
             ['Female', 3, 'Red',    14, 8],
             ['Male',   4, 'Red',    16, 9]
            ]

    df = pd.DataFrame(array, columns = ['Sex', 'Id', 'Color', 'Age', 'Class'])
    print (df)

    bin_features = ['Sex', 'Color']
    num_features = ['Id', 'Age']
    col_labels = ['Class']

    # https: // www.statology.org / numpy - get - column /
    enc = OneHotEncoder(handle_unknown='ignore', sparse = False)
    enc.fit(df[bin_features].to_numpy())

    print(f"Binary categories: {enc.categories_}")

    X_bin = enc.transform(df[bin_features].to_numpy())
    X_num = df[num_features].to_numpy()
    X_lab = df[col_labels].to_numpy()

    print(f"Num columns:   {X_num}")
    print(f"Bin columns:   {X_bin}")
    print(f"Label columns: {X_lab}")

    ar_h = np.hstack((X_num, X_bin, X_lab))
    print(f"Converted:     {ar_h}")