encoder = OneHotEncoder(sparse=False)
encoded_tr_feat = encoder.fit_transform(X_train[CATEGORY_FEAT_COLS])
encoded_te_feat = encoder.transform(X_test[CATEGORY_FEAT_COLS])
scaler = MinMaxScaler()
scaled_tr_feat = scaler.fit_transform(X_train[NUMERIC_FEAT_COLS])
scaled_te_feat = scaler.transform(X_test[NUMERIC_FEAT_COLS])
X_train_proc = np.hstack((encoded_tr_feat,scaled_tr_feat))
X_test_proc = np.hstack(((encoded_te_feat,scaled_te_feat)))
model = LinearRegression()
model.fit(X_train_proc,y_train)
r2_score2 = model.score(X_test_proc,y_test)
davidkorea / house_price_pred_feat_preprocessing Goto Github PK
View Code? Open in Web Editor NEWFeature Pre-processing