def test_groupby1(self):
df = load_iris()
train_out = xgb_regression_train(
df,
feature_cols=['sepal_length', 'sepal_width', 'petal_length'],
label_col='petal_width',
group_by=['species'])
predict_out = xgb_regression_predict(df, train_out['model'])
def test(self):
xgb_train = xgb_regression_train(
self.testdata,
feature_cols=['sepal_length', 'sepal_width', 'petal_length'],
label_col='petal_width')['model']
np.testing.assert_array_almost_equal(
xgb_train['feature_importance'],
[0.2591911852359772, 0.2077205926179886, 0.533088207244873], 10)
predict = xgb_regression_predict(self.testdata,
xgb_train)['out_table']['prediction']
np.testing.assert_array_almost_equal(predict[:5], [
0.2634012401, 0.1740619838, 0.1951409280, 0.1879036427,
0.2634012401
], 10)
def groupby1(self):
df = get_iris()
train_out = xgb_regression_train(df, feature_cols=['sepal_length', 'sepal_width', 'petal_length'], label_col='petal_width', group_by=['species'])
predict_out = xgb_regression_predict(df, train_out['model'])
print(predict_out['out_table'][['petal_width', 'prediction']])