test_sklearn.py

# coding: utf-8
import itertools
import joblib
import math
import os
import unittest
import warnings

import lightgbm as lgb
import numpy as np
from sklearn import __version__ as sk_version
from sklearn.base import clone
from sklearn.datasets import (load_boston, load_breast_cancer, load_digits,
                              load_iris, load_svmlight_file)
from sklearn.exceptions import SkipTestWarning
from sklearn.metrics import log_loss, mean_squared_error
from sklearn.model_selection import GridSearchCV, train_test_split
from sklearn.utils.estimator_checks import (_yield_all_checks, SkipTest,
                                            check_parameters_default_constructible)


decreasing_generator = itertools.count(0, -1)


def custom_asymmetric_obj(y_true, y_pred):
    residual = (y_true - y_pred).astype("float")
    grad = np.where(residual < 0, -2 * 10.0 * residual, -2 * residual)
    hess = np.where(residual < 0, 2 * 10.0, 2.0)
    return grad, hess


def objective_ls(y_true, y_pred):
    grad = (y_pred - y_true)
    hess = np.ones(len(y_true))
    return grad, hess


def logregobj(y_true, y_pred):
    y_pred = 1.0 / (1.0 + np.exp(-y_pred))
    grad = y_pred - y_true
    hess = y_pred * (1.0 - y_pred)
    return grad, hess


def custom_dummy_obj(y_true, y_pred):
    return np.ones(y_true.shape), np.ones(y_true.shape)


def constant_metric(y_true, y_pred):
    return 'error', 0, False


def decreasing_metric(y_true, y_pred):
    return ('decreasing_metric', next(decreasing_generator), False)


def mse(y_true, y_pred):
    return 'custom MSE', mean_squared_error(y_true, y_pred), False


def binary_error(y_true, y_pred):
    return np.mean((y_pred > 0.5) != y_true)


def multi_error(y_true, y_pred):
    return np.mean(y_true != y_pred)


def multi_logloss(y_true, y_pred):
    return np.mean([-math.log(y_pred[i][y]) for i, y in enumerate(y_true)])


class TestSklearn(unittest.TestCase):

    def test_binary(self):
        X, y = load_breast_cancer(True)
        X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.1, random_state=42)
        gbm = lgb.LGBMClassifier(n_estimators=50, silent=True)
        gbm.fit(X_train, y_train, eval_set=[(X_test, y_test)], early_stopping_rounds=5, verbose=False)
        ret = log_loss(y_test, gbm.predict_proba(X_test))
        self.assertLess(ret, 0.12)
        self.assertAlmostEqual(ret, gbm.evals_result_['valid_0']['binary_logloss'][gbm.best_iteration_ - 1], places=5)

    def test_regression(self):
        X, y = load_boston(True)
        X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.1, random_state=42)
        gbm = lgb.LGBMRegressor(n_estimators=50, silent=True)
        gbm.fit(X_train, y_train, eval_set=[(X_test, y_test)], early_stopping_rounds=5, verbose=False)
        ret = mean_squared_error(y_test, gbm.predict(X_test))
        self.assertLess(ret, 7)
        self.assertAlmostEqual(ret, gbm.evals_result_['valid_0']['l2'][gbm.best_iteration_ - 1], places=5)

    def test_multiclass(self):
        X, y = load_digits(10, True)
        X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.1, random_state=42)
        gbm = lgb.LGBMClassifier(n_estimators=50, silent=True)
        gbm.fit(X_train, y_train, eval_set=[(X_test, y_test)], early_stopping_rounds=5, verbose=False)
        ret = multi_error(y_test, gbm.predict(X_test))
        self.assertLess(ret, 0.05)
        ret = multi_logloss(y_test, gbm.predict_proba(X_test))
        self.assertLess(ret, 0.16)
        self.assertAlmostEqual(ret, gbm.evals_result_['valid_0']['multi_logloss'][gbm.best_iteration_ - 1], places=5)

    def test_lambdarank(self):
        X_train, y_train = load_svmlight_file(os.path.join(os.path.dirname(os.path.realpath(__file__)),
                                                           '../../examples/lambdarank/rank.train'))
        X_test, y_test = load_svmlight_file(os.path.join(os.path.dirname(os.path.realpath(__file__)),
                                                         '../../examples/lambdarank/rank.test'))
        q_train = np.loadtxt(os.path.join(os.path.dirname(os.path.realpath(__file__)),
                                          '../../examples/lambdarank/rank.train.query'))
        q_test = np.loadtxt(os.path.join(os.path.dirname(os.path.realpath(__file__)),
                                         '../../examples/lambdarank/rank.test.query'))
        gbm = lgb.LGBMRanker(n_estimators=50)
        gbm.fit(X_train, y_train, group=q_train, eval_set=[(X_test, y_test)],
                eval_group=[q_test], eval_at=[1, 3], early_stopping_rounds=10, verbose=False,
                callbacks=[lgb.reset_parameter(learning_rate=lambda x: max(0.01, 0.1 - 0.01 * x))])
        self.assertLessEqual(gbm.best_iteration_, 24)
        self.assertGreater(gbm.best_score_['valid_0']['ndcg@1'], 0.5769)
        self.assertGreater(gbm.best_score_['valid_0']['ndcg@3'], 0.5920)

    def test_xendcg(self):
        dir_path = os.path.dirname(os.path.realpath(__file__))
        X_train, y_train = load_svmlight_file(os.path.join(dir_path, '../../examples/xendcg/rank.train'))
        X_test, y_test = load_svmlight_file(os.path.join(dir_path, '../../examples/xendcg/rank.test'))
        q_train = np.loadtxt(os.path.join(dir_path, '../../examples/xendcg/rank.train.query'))
        q_test = np.loadtxt(os.path.join(dir_path, '../../examples/xendcg/rank.test.query'))
        gbm = lgb.LGBMRanker(n_estimators=50, objective='rank_xendcg', random_state=5, n_jobs=1)
        gbm.fit(X_train, y_train, group=q_train, eval_set=[(X_test, y_test)],
                eval_group=[q_test], eval_at=[1, 3], early_stopping_rounds=10, verbose=False,
                eval_metric='ndcg',
                callbacks=[lgb.reset_parameter(learning_rate=lambda x: max(0.01, 0.1 - 0.01 * x))])
        self.assertLessEqual(gbm.best_iteration_, 24)
        self.assertGreater(gbm.best_score_['valid_0']['ndcg@1'], 0.6382)
        self.assertGreater(gbm.best_score_['valid_0']['ndcg@3'], 0.6319)

    def test_regression_with_custom_objective(self):
        X, y = load_boston(True)
        X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.1, random_state=42)
        gbm = lgb.LGBMRegressor(n_estimators=50, silent=True, objective=objective_ls)
        gbm.fit(X_train, y_train, eval_set=[(X_test, y_test)], early_stopping_rounds=5, verbose=False)
        ret = mean_squared_error(y_test, gbm.predict(X_test))
        self.assertLess(ret, 7.0)
        self.assertAlmostEqual(ret, gbm.evals_result_['valid_0']['l2'][gbm.best_iteration_ - 1], places=5)

    def test_binary_classification_with_custom_objective(self):
        X, y = load_digits(2, True)
        X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.1, random_state=42)
        gbm = lgb.LGBMClassifier(n_estimators=50, silent=True, objective=logregobj)
        gbm.fit(X_train, y_train, eval_set=[(X_test, y_test)], early_stopping_rounds=5, verbose=False)
        # prediction result is actually not transformed (is raw) due to custom objective
        y_pred_raw = gbm.predict_proba(X_test)
        self.assertFalse(np.all(y_pred_raw >= 0))
        y_pred = 1.0 / (1.0 + np.exp(-y_pred_raw))
        ret = binary_error(y_test, y_pred)
        self.assertLess(ret, 0.05)

    def test_dart(self):
        X, y = load_boston(True)
        X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.1, random_state=42)
        gbm = lgb.LGBMRegressor(boosting_type='dart', n_estimators=50)
        gbm.fit(X_train, y_train)
        score = gbm.score(X_test, y_test)
        self.assertGreaterEqual(score, 0.8)
        self.assertLessEqual(score, 1.)

    def test_grid_search(self):
        X, y = load_iris(True)
        y = np.array(list(map(str, y)))  # utilize label encoder at it's max power
        X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.1, random_state=42)
        params = {'subsample': 0.8,
                  'subsample_freq': 1}
        grid_params = {'boosting_type': ['rf', 'gbdt'],
                       'n_estimators': [4, 6],
                       'reg_alpha': [0.01, 0.005]}
        fit_params = {'verbose': False,
                      'eval_set': [(X_test, y_test)],
                      'eval_metric': constant_metric,
                      'early_stopping_rounds': 2}
        grid = GridSearchCV(lgb.LGBMClassifier(**params), grid_params, cv=2)
        grid.fit(X, y, **fit_params)
        self.assertIn(grid.best_params_['boosting_type'], ['rf', 'gbdt'])
        self.assertIn(grid.best_params_['n_estimators'], [4, 6])
        self.assertIn(grid.best_params_['reg_alpha'], [0.01, 0.005])
        self.assertLess(grid.best_score_, 0.9)
        self.assertEqual(grid.best_estimator_.best_iteration_, 1)
        self.assertLess(grid.best_estimator_.best_score_['valid_0']['multi_logloss'], 0.25)
        self.assertEqual(grid.best_estimator_.best_score_['valid_0']['error'], 0)

    def test_clone_and_property(self):
        X, y = load_boston(True)
        gbm = lgb.LGBMRegressor(n_estimators=10, silent=True)
        gbm.fit(X, y, verbose=False)

        gbm_clone = clone(gbm)
        self.assertIsInstance(gbm.booster_, lgb.Booster)
        self.assertIsInstance(gbm.feature_importances_, np.ndarray)

        X, y = load_digits(2, True)
        clf = lgb.LGBMClassifier(n_estimators=10, silent=True)
        clf.fit(X, y, verbose=False)
        self.assertListEqual(sorted(clf.classes_), [0, 1])
        self.assertEqual(clf.n_classes_, 2)
        self.assertIsInstance(clf.booster_, lgb.Booster)
        self.assertIsInstance(clf.feature_importances_, np.ndarray)

    def test_joblib(self):
        X, y = load_boston(True)
        X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.1, random_state=42)
        gbm = lgb.LGBMRegressor(n_estimators=10, objective=custom_asymmetric_obj,
                                silent=True, importance_type='split')
        gbm.fit(X_train, y_train, eval_set=[(X_train, y_train), (X_test, y_test)],
                eval_metric=mse, early_stopping_rounds=5, verbose=False,
                callbacks=[lgb.reset_parameter(learning_rate=list(np.arange(1, 0, -0.1)))])

        joblib.dump(gbm, 'lgb.pkl')  # test model with custom functions
        gbm_pickle = joblib.load('lgb.pkl')
        self.assertIsInstance(gbm_pickle.booster_, lgb.Booster)
        self.assertDictEqual(gbm.get_params(), gbm_pickle.get_params())
        np.testing.assert_array_equal(gbm.feature_importances_, gbm_pickle.feature_importances_)
        self.assertAlmostEqual(gbm_pickle.learning_rate, 0.1)
        self.assertTrue(callable(gbm_pickle.objective))

        for eval_set in gbm.evals_result_:
            for metric in gbm.evals_result_[eval_set]:
                np.testing.assert_allclose(gbm.evals_result_[eval_set][metric],
                                           gbm_pickle.evals_result_[eval_set][metric])
        pred_origin = gbm.predict(X_test)
        pred_pickle = gbm_pickle.predict(X_test)
        np.testing.assert_allclose(pred_origin, pred_pickle)

    def test_random_state_object(self):
        X, y = load_iris(True)
        X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.1, random_state=42)
        state1 = np.random.RandomState(123)
        state2 = np.random.RandomState(123)
        clf1 = lgb.LGBMClassifier(n_estimators=10, subsample=0.5, subsample_freq=1, random_state=state1)
        clf2 = lgb.LGBMClassifier(n_estimators=10, subsample=0.5, subsample_freq=1, random_state=state2)
        # Test if random_state is properly stored
        self.assertIs(clf1.random_state, state1)
        self.assertIs(clf2.random_state, state2)
        # Test if two random states produce identical models
        clf1.fit(X_train, y_train)
        clf2.fit(X_train, y_train)
        y_pred1 = clf1.predict(X_test, raw_score=True)
        y_pred2 = clf2.predict(X_test, raw_score=True)
        np.testing.assert_allclose(y_pred1, y_pred2)
        np.testing.assert_array_equal(clf1.feature_importances_, clf2.feature_importances_)
        df1 = clf1.booster_.model_to_string(num_iteration=0)
        df2 = clf2.booster_.model_to_string(num_iteration=0)
        self.assertMultiLineEqual(df1, df2)
        # Test if subsequent fits sample from random_state object and produce different models
        clf1.fit(X_train, y_train)
        y_pred1_refit = clf1.predict(X_test, raw_score=True)
        df3 = clf1.booster_.model_to_string(num_iteration=0)
        self.assertIs(clf1.random_state, state1)
        self.assertIs(clf2.random_state, state2)
        self.assertRaises(AssertionError,
                          np.testing.assert_allclose,
                          y_pred1, y_pred1_refit)
        self.assertRaises(AssertionError,
                          self.assertMultiLineEqual,
                          df1, df3)

    def test_feature_importances_single_leaf(self):
        data = load_iris()
        clf = lgb.LGBMClassifier(n_estimators=10)
        clf.fit(data.data, data.target)
        importances = clf.feature_importances_
        self.assertEqual(len(importances), 4)

    def test_feature_importances_type(self):
        data = load_iris()
        clf = lgb.LGBMClassifier(n_estimators=10)
        clf.fit(data.data, data.target)
        clf.set_params(importance_type='split')
        importances_split = clf.feature_importances_
        clf.set_params(importance_type='gain')
        importances_gain = clf.feature_importances_
        # Test that the largest element is NOT the same, the smallest can be the same, i.e. zero
        importance_split_top1 = sorted(importances_split, reverse=True)[0]
        importance_gain_top1 = sorted(importances_gain, reverse=True)[0]
        self.assertNotEqual(importance_split_top1, importance_gain_top1)

    # sklearn <0.19 cannot accept instance, but many tests could be passed only with min_data=1 and min_data_in_bin=1
    @unittest.skipIf(sk_version < '0.19.0', 'scikit-learn version is less than 0.19')
    def test_sklearn_integration(self):
        # we cannot use `check_estimator` directly since there is no skip test mechanism
        for name, estimator in ((lgb.sklearn.LGBMClassifier.__name__, lgb.sklearn.LGBMClassifier),
                                (lgb.sklearn.LGBMRegressor.__name__, lgb.sklearn.LGBMRegressor)):
            check_parameters_default_constructible(name, estimator)
            # we cannot leave default params (see https://github.com/microsoft/LightGBM/issues/833)
            estimator = estimator(min_child_samples=1, min_data_in_bin=1)
            for check in _yield_all_checks(name, estimator):
                check_name = check.func.__name__ if hasattr(check, 'func') else check.__name__
                if check_name == 'check_estimators_nan_inf':
                    continue  # skip test because LightGBM deals with nan
                elif check_name == "check_no_attributes_set_in_init":
                    # skip test because scikit-learn incorrectly asserts that
                    # private attributes cannot be set in __init__
                    # (see https://github.com/microsoft/LightGBM/issues/2628)
                    continue
                try:
                    check(name, estimator)
                except SkipTest as message:
                    warnings.warn(message, SkipTestWarning)

    @unittest.skipIf(not lgb.compat.PANDAS_INSTALLED, 'pandas is not installed')
    def test_pandas_categorical(self):
        import pandas as pd
        np.random.seed(42)  # sometimes there is no difference how cols are treated (cat or not cat)
        X = pd.DataFrame({"A": np.random.permutation(['a', 'b', 'c', 'd'] * 75),  # str
                          "B": np.random.permutation([1, 2, 3] * 100),  # int
                          "C": np.random.permutation([0.1, 0.2, -0.1, -0.1, 0.2] * 60),  # float
                          "D": np.random.permutation([True, False] * 150),  # bool
                          "E": pd.Categorical(np.random.permutation(['z', 'y', 'x', 'w', 'v'] * 60),
                                              ordered=True)})  # str and ordered categorical
        y = np.random.permutation([0, 1] * 150)
        X_test = pd.DataFrame({"A": np.random.permutation(['a', 'b', 'e'] * 20),  # unseen category
                               "B": np.random.permutation([1, 3] * 30),
                               "C": np.random.permutation([0.1, -0.1, 0.2, 0.2] * 15),
                               "D": np.random.permutation([True, False] * 30),
                               "E": pd.Categorical(np.random.permutation(['z', 'y'] * 30),
                                                   ordered=True)})
        np.random.seed()  # reset seed
        cat_cols_actual = ["A", "B", "C", "D"]
        cat_cols_to_store = cat_cols_actual + ["E"]
        X[cat_cols_actual] = X[cat_cols_actual].astype('category')
        X_test[cat_cols_actual] = X_test[cat_cols_actual].astype('category')
        cat_values = [X[col].cat.categories.tolist() for col in cat_cols_to_store]
        gbm0 = lgb.sklearn.LGBMClassifier(n_estimators=10).fit(X, y)
        pred0 = gbm0.predict(X_test, raw_score=True)
        pred_prob = gbm0.predict_proba(X_test)[:, 1]
        gbm1 = lgb.sklearn.LGBMClassifier(n_estimators=10).fit(X, pd.Series(y), categorical_feature=[0])
        pred1 = gbm1.predict(X_test, raw_score=True)
        gbm2 = lgb.sklearn.LGBMClassifier(n_estimators=10).fit(X, y, categorical_feature=['A'])
        pred2 = gbm2.predict(X_test, raw_score=True)
        gbm3 = lgb.sklearn.LGBMClassifier(n_estimators=10).fit(X, y, categorical_feature=['A', 'B', 'C', 'D'])
        pred3 = gbm3.predict(X_test, raw_score=True)
        gbm3.booster_.save_model('categorical.model')
        gbm4 = lgb.Booster(model_file='categorical.model')
        pred4 = gbm4.predict(X_test)
        gbm5 = lgb.sklearn.LGBMClassifier(n_estimators=10).fit(X, y, categorical_feature=['A', 'B', 'C', 'D', 'E'])
        pred5 = gbm5.predict(X_test, raw_score=True)
        gbm6 = lgb.sklearn.LGBMClassifier(n_estimators=10).fit(X, y, categorical_feature=[])
        pred6 = gbm6.predict(X_test, raw_score=True)
        self.assertRaises(AssertionError,
                          np.testing.assert_allclose,
                          pred0, pred1)
        self.assertRaises(AssertionError,
                          np.testing.assert_allclose,
                          pred0, pred2)
        np.testing.assert_allclose(pred1, pred2)
        np.testing.assert_allclose(pred0, pred3)
        np.testing.assert_allclose(pred_prob, pred4)
        self.assertRaises(AssertionError,
                          np.testing.assert_allclose,
                          pred0, pred5)  # ordered cat features aren't treated as cat features by default
        self.assertRaises(AssertionError,
                          np.testing.assert_allclose,
                          pred0, pred6)
        self.assertListEqual(gbm0.booster_.pandas_categorical, cat_values)
        self.assertListEqual(gbm1.booster_.pandas_categorical, cat_values)
        self.assertListEqual(gbm2.booster_.pandas_categorical, cat_values)
        self.assertListEqual(gbm3.booster_.pandas_categorical, cat_values)
        self.assertListEqual(gbm4.pandas_categorical, cat_values)
        self.assertListEqual(gbm5.booster_.pandas_categorical, cat_values)
        self.assertListEqual(gbm6.booster_.pandas_categorical, cat_values)

    @unittest.skipIf(not lgb.compat.PANDAS_INSTALLED, 'pandas is not installed')
    def test_pandas_sparse(self):
        import pandas as pd
        try:
            from pandas.arrays import SparseArray
        except ImportError:  # support old versions
            from pandas import SparseArray
        X = pd.DataFrame({"A": SparseArray(np.random.permutation([0, 1, 2] * 100)),
                          "B": SparseArray(np.random.permutation([0.0, 0.1, 0.2, -0.1, 0.2] * 60)),
                          "C": SparseArray(np.random.permutation([True, False] * 150))})
        y = pd.Series(SparseArray(np.random.permutation([0, 1] * 150)))
        X_test = pd.DataFrame({"A": SparseArray(np.random.permutation([0, 2] * 30)),
                               "B": SparseArray(np.random.permutation([0.0, 0.1, 0.2, -0.1] * 15)),
                               "C": SparseArray(np.random.permutation([True, False] * 30))})
        if pd.__version__ >= '0.24.0':
            for dtype in pd.concat([X.dtypes, X_test.dtypes, pd.Series(y.dtypes)]):
                self.assertTrue(pd.api.types.is_sparse(dtype))
        gbm = lgb.sklearn.LGBMClassifier(n_estimators=10).fit(X, y)
        pred_sparse = gbm.predict(X_test, raw_score=True)
        if hasattr(X_test, 'sparse'):
            pred_dense = gbm.predict(X_test.sparse.to_dense(), raw_score=True)
        else:
            pred_dense = gbm.predict(X_test.to_dense(), raw_score=True)
        np.testing.assert_allclose(pred_sparse, pred_dense)

    def test_predict(self):
        # With default params
        iris = load_iris()
        X_train, X_test, y_train, y_test = train_test_split(iris.data, iris.target,
                                                            test_size=0.2, random_state=42)

        gbm = lgb.train({'objective': 'multiclass',
                         'num_class': 3,
                         'verbose': -1},
                        lgb.Dataset(X_train, y_train))
        clf = lgb.LGBMClassifier(verbose=-1).fit(X_train, y_train)

        # Tests same probabilities
        res_engine = gbm.predict(X_test)
        res_sklearn = clf.predict_proba(X_test)
        np.testing.assert_allclose(res_engine, res_sklearn)

        # Tests same predictions
        res_engine = np.argmax(gbm.predict(X_test), axis=1)
        res_sklearn = clf.predict(X_test)
        np.testing.assert_equal(res_engine, res_sklearn)

        # Tests same raw scores
        res_engine = gbm.predict(X_test, raw_score=True)
        res_sklearn = clf.predict(X_test, raw_score=True)
        np.testing.assert_allclose(res_engine, res_sklearn)

        # Tests same leaf indices
        res_engine = gbm.predict(X_test, pred_leaf=True)
        res_sklearn = clf.predict(X_test, pred_leaf=True)
        np.testing.assert_equal(res_engine, res_sklearn)

        # Tests same feature contributions
        res_engine = gbm.predict(X_test, pred_contrib=True)
        res_sklearn = clf.predict(X_test, pred_contrib=True)
        np.testing.assert_allclose(res_engine, res_sklearn)

        # Tests other parameters for the prediction works
        res_engine = gbm.predict(X_test)
        res_sklearn_params = clf.predict_proba(X_test,
                                               pred_early_stop=True,
                                               pred_early_stop_margin=1.0)
        self.assertRaises(AssertionError,
                          np.testing.assert_allclose,
                          res_engine, res_sklearn_params)

    def test_evaluate_train_set(self):
        X, y = load_boston(True)
        X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.1, random_state=42)
        gbm = lgb.LGBMRegressor(n_estimators=10, silent=True)
        gbm.fit(X_train, y_train, eval_set=[(X_train, y_train), (X_test, y_test)], verbose=False)
        self.assertEqual(len(gbm.evals_result_), 2)
        self.assertIn('training', gbm.evals_result_)
        self.assertEqual(len(gbm.evals_result_['training']), 1)
        self.assertIn('l2', gbm.evals_result_['training'])
        self.assertIn('valid_1', gbm.evals_result_)
        self.assertEqual(len(gbm.evals_result_['valid_1']), 1)
        self.assertIn('l2', gbm.evals_result_['valid_1'])

    def test_metrics(self):
        X, y = load_boston(True)
        params = {'n_estimators': 2, 'verbose': -1}
        params_fit = {'X': X, 'y': y, 'eval_set': (X, y), 'verbose': False}

        # no custom objective, no custom metric
        # default metric
        gbm = lgb.LGBMRegressor(**params).fit(**params_fit)
        self.assertEqual(len(gbm.evals_result_['training']), 1)
        self.assertIn('l2', gbm.evals_result_['training'])

        # non-default metric
        gbm = lgb.LGBMRegressor(metric='mape', **params).fit(**params_fit)
        self.assertEqual(len(gbm.evals_result_['training']), 1)
        self.assertIn('mape', gbm.evals_result_['training'])

        # no metric
        gbm = lgb.LGBMRegressor(metric='None', **params).fit(**params_fit)
        self.assertIs(gbm.evals_result_, None)

        # non-default metric in eval_metric
        gbm = lgb.LGBMRegressor(**params).fit(eval_metric='mape', **params_fit)
        self.assertEqual(len(gbm.evals_result_['training']), 2)
        self.assertIn('l2', gbm.evals_result_['training'])
        self.assertIn('mape', gbm.evals_result_['training'])

        # non-default metric with non-default metric in eval_metric
        gbm = lgb.LGBMRegressor(metric='gamma', **params).fit(eval_metric='mape', **params_fit)
        self.assertEqual(len(gbm.evals_result_['training']), 2)
        self.assertIn('gamma', gbm.evals_result_['training'])
        self.assertIn('mape', gbm.evals_result_['training'])

        # non-default metric with multiple metrics in eval_metric
        gbm = lgb.LGBMRegressor(metric='gamma',
                                **params).fit(eval_metric=['l2', 'mape'], **params_fit)
        self.assertEqual(len(gbm.evals_result_['training']), 3)
        self.assertIn('gamma', gbm.evals_result_['training'])
        self.assertIn('l2', gbm.evals_result_['training'])
        self.assertIn('mape', gbm.evals_result_['training'])

        # default metric for non-default objective
        gbm = lgb.LGBMRegressor(objective='regression_l1', **params).fit(**params_fit)
        self.assertEqual(len(gbm.evals_result_['training']), 1)
        self.assertIn('l1', gbm.evals_result_['training'])

        # non-default metric for non-default objective
        gbm = lgb.LGBMRegressor(objective='regression_l1', metric='mape',
                                **params).fit(**params_fit)
        self.assertEqual(len(gbm.evals_result_['training']), 1)
        self.assertIn('mape', gbm.evals_result_['training'])

        # no metric
        gbm = lgb.LGBMRegressor(objective='regression_l1', metric='None',
                                **params).fit(**params_fit)
        self.assertIs(gbm.evals_result_, None)

        # non-default metric in eval_metric for non-default objective
        gbm = lgb.LGBMRegressor(objective='regression_l1',
                                **params).fit(eval_metric='mape', **params_fit)
        self.assertEqual(len(gbm.evals_result_['training']), 2)
        self.assertIn('l1', gbm.evals_result_['training'])
        self.assertIn('mape', gbm.evals_result_['training'])

        # non-default metric with non-default metric in eval_metric for non-default objective
        gbm = lgb.LGBMRegressor(objective='regression_l1', metric='gamma',
                                **params).fit(eval_metric='mape', **params_fit)
        self.assertEqual(len(gbm.evals_result_['training']), 2)
        self.assertIn('gamma', gbm.evals_result_['training'])
        self.assertIn('mape', gbm.evals_result_['training'])

        # non-default metric with multiple metrics in eval_metric for non-default objective
        gbm = lgb.LGBMRegressor(objective='regression_l1', metric='gamma',
                                **params).fit(eval_metric=['l2', 'mape'], **params_fit)
        self.assertEqual(len(gbm.evals_result_['training']), 3)
        self.assertIn('gamma', gbm.evals_result_['training'])
        self.assertIn('l2', gbm.evals_result_['training'])
        self.assertIn('mape', gbm.evals_result_['training'])

        # custom objective, no custom metric
        # default regression metric for custom objective
        gbm = lgb.LGBMRegressor(objective=custom_dummy_obj, **params).fit(**params_fit)
        self.assertEqual(len(gbm.evals_result_['training']), 1)
        self.assertIn('l2', gbm.evals_result_['training'])

        # non-default regression metric for custom objective
        gbm = lgb.LGBMRegressor(objective=custom_dummy_obj, metric='mape', **params).fit(**params_fit)
        self.assertEqual(len(gbm.evals_result_['training']), 1)
        self.assertIn('mape', gbm.evals_result_['training'])

        # multiple regression metrics for custom objective
        gbm = lgb.LGBMRegressor(objective=custom_dummy_obj, metric=['l1', 'gamma'],
                                **params).fit(**params_fit)
        self.assertEqual(len(gbm.evals_result_['training']), 2)
        self.assertIn('l1', gbm.evals_result_['training'])
        self.assertIn('gamma', gbm.evals_result_['training'])

        # no metric
        gbm = lgb.LGBMRegressor(objective=custom_dummy_obj, metric='None',
                                **params).fit(**params_fit)
        self.assertIs(gbm.evals_result_, None)

        # default regression metric with non-default metric in eval_metric for custom objective
        gbm = lgb.LGBMRegressor(objective=custom_dummy_obj,
                                **params).fit(eval_metric='mape', **params_fit)
        self.assertEqual(len(gbm.evals_result_['training']), 2)
        self.assertIn('l2', gbm.evals_result_['training'])
        self.assertIn('mape', gbm.evals_result_['training'])

        # non-default regression metric with metric in eval_metric for custom objective
        gbm = lgb.LGBMRegressor(objective=custom_dummy_obj, metric='mape',
                                **params).fit(eval_metric='gamma', **params_fit)
        self.assertEqual(len(gbm.evals_result_['training']), 2)
        self.assertIn('mape', gbm.evals_result_['training'])
        self.assertIn('gamma', gbm.evals_result_['training'])

        # multiple regression metrics with metric in eval_metric for custom objective
        gbm = lgb.LGBMRegressor(objective=custom_dummy_obj, metric=['l1', 'gamma'],
                                **params).fit(eval_metric='l2', **params_fit)
        self.assertEqual(len(gbm.evals_result_['training']), 3)
        self.assertIn('l1', gbm.evals_result_['training'])
        self.assertIn('gamma', gbm.evals_result_['training'])
        self.assertIn('l2', gbm.evals_result_['training'])

        # multiple regression metrics with multiple metrics in eval_metric for custom objective
        gbm = lgb.LGBMRegressor(objective=custom_dummy_obj, metric=['l1', 'gamma'],
                                **params).fit(eval_metric=['l2', 'mape'], **params_fit)
        self.assertEqual(len(gbm.evals_result_['training']), 4)
        self.assertIn('l1', gbm.evals_result_['training'])
        self.assertIn('gamma', gbm.evals_result_['training'])
        self.assertIn('l2', gbm.evals_result_['training'])
        self.assertIn('mape', gbm.evals_result_['training'])

        # no custom objective, custom metric
        # default metric with custom metric
        gbm = lgb.LGBMRegressor(**params).fit(eval_metric=constant_metric, **params_fit)
        self.assertEqual(len(gbm.evals_result_['training']), 2)
        self.assertIn('l2', gbm.evals_result_['training'])
        self.assertIn('error', gbm.evals_result_['training'])

        # non-default metric with custom metric
        gbm = lgb.LGBMRegressor(metric='mape',
                                **params).fit(eval_metric=constant_metric, **params_fit)
        self.assertEqual(len(gbm.evals_result_['training']), 2)
        self.assertIn('mape', gbm.evals_result_['training'])
        self.assertIn('error', gbm.evals_result_['training'])

        # multiple metrics with custom metric
        gbm = lgb.LGBMRegressor(metric=['l1', 'gamma'],
                                **params).fit(eval_metric=constant_metric, **params_fit)
        self.assertEqual(len(gbm.evals_result_['training']), 3)
        self.assertIn('l1', gbm.evals_result_['training'])
        self.assertIn('gamma', gbm.evals_result_['training'])
        self.assertIn('error', gbm.evals_result_['training'])

        # custom metric (disable default metric)
        gbm = lgb.LGBMRegressor(metric='None',
                                **params).fit(eval_metric=constant_metric, **params_fit)
        self.assertEqual(len(gbm.evals_result_['training']), 1)
        self.assertIn('error', gbm.evals_result_['training'])

        # default metric for non-default objective with custom metric
        gbm = lgb.LGBMRegressor(objective='regression_l1',
                                **params).fit(eval_metric=constant_metric, **params_fit)
        self.assertEqual(len(gbm.evals_result_['training']), 2)
        self.assertIn('l1', gbm.evals_result_['training'])
        self.assertIn('error', gbm.evals_result_['training'])

        # non-default metric for non-default objective with custom metric
        gbm = lgb.LGBMRegressor(objective='regression_l1', metric='mape',
                                **params).fit(eval_metric=constant_metric, **params_fit)
        self.assertEqual(len(gbm.evals_result_['training']), 2)
        self.assertIn('mape', gbm.evals_result_['training'])
        self.assertIn('error', gbm.evals_result_['training'])

        # multiple metrics for non-default objective with custom metric
        gbm = lgb.LGBMRegressor(objective='regression_l1', metric=['l1', 'gamma'],
                                **params).fit(eval_metric=constant_metric, **params_fit)
        self.assertEqual(len(gbm.evals_result_['training']), 3)
        self.assertIn('l1', gbm.evals_result_['training'])
        self.assertIn('gamma', gbm.evals_result_['training'])
        self.assertIn('error', gbm.evals_result_['training'])

        # custom metric (disable default metric for non-default objective)
        gbm = lgb.LGBMRegressor(objective='regression_l1', metric='None',
                                **params).fit(eval_metric=constant_metric, **params_fit)
        self.assertEqual(len(gbm.evals_result_['training']), 1)
        self.assertIn('error', gbm.evals_result_['training'])

        # custom objective, custom metric
        # custom metric for custom objective
        gbm = lgb.LGBMRegressor(objective=custom_dummy_obj,
                                **params).fit(eval_metric=constant_metric, **params_fit)
        self.assertEqual(len(gbm.evals_result_['training']), 1)
        self.assertIn('error', gbm.evals_result_['training'])

        # non-default regression metric with custom metric for custom objective
        gbm = lgb.LGBMRegressor(objective=custom_dummy_obj, metric='mape',
                                **params).fit(eval_metric=constant_metric, **params_fit)
        self.assertEqual(len(gbm.evals_result_['training']), 2)
        self.assertIn('mape', gbm.evals_result_['training'])
        self.assertIn('error', gbm.evals_result_['training'])

        # multiple regression metrics with custom metric for custom objective
        gbm = lgb.LGBMRegressor(objective=custom_dummy_obj, metric=['l2', 'mape'],
                                **params).fit(eval_metric=constant_metric, **params_fit)
        self.assertEqual(len(gbm.evals_result_['training']), 3)
        self.assertIn('l2', gbm.evals_result_['training'])
        self.assertIn('mape', gbm.evals_result_['training'])
        self.assertIn('error', gbm.evals_result_['training'])

        X, y = load_digits(3, True)
        params_fit = {'X': X, 'y': y, 'eval_set': (X, y), 'verbose': False}

        # default metric and invalid binary metric is replaced with multiclass alternative
        gbm = lgb.LGBMClassifier(**params).fit(eval_metric='binary_error', **params_fit)
        self.assertEqual(len(gbm.evals_result_['training']), 2)
        self.assertIn('multi_logloss', gbm.evals_result_['training'])
        self.assertIn('multi_error', gbm.evals_result_['training'])

        # invalid objective is replaced with default multiclass one
        # and invalid binary metric is replaced with multiclass alternative
        gbm = lgb.LGBMClassifier(objective='invalid_obj',
                                 **params).fit(eval_metric='binary_error', **params_fit)
        self.assertEqual(gbm.objective_, 'multiclass')
        self.assertEqual(len(gbm.evals_result_['training']), 2)
        self.assertIn('multi_logloss', gbm.evals_result_['training'])
        self.assertIn('multi_error', gbm.evals_result_['training'])

        # default metric for non-default multiclass objective
        # and invalid binary metric is replaced with multiclass alternative
        gbm = lgb.LGBMClassifier(objective='ovr',
                                 **params).fit(eval_metric='binary_error', **params_fit)
        self.assertEqual(gbm.objective_, 'ovr')
        self.assertEqual(len(gbm.evals_result_['training']), 2)
        self.assertIn('multi_logloss', gbm.evals_result_['training'])
        self.assertIn('multi_error', gbm.evals_result_['training'])

        X, y = load_digits(2, True)
        params_fit = {'X': X, 'y': y, 'eval_set': (X, y), 'verbose': False}

        # default metric and invalid multiclass metric is replaced with binary alternative
        gbm = lgb.LGBMClassifier(**params).fit(eval_metric='multi_error', **params_fit)
        self.assertEqual(len(gbm.evals_result_['training']), 2)
        self.assertIn('binary_logloss', gbm.evals_result_['training'])
        self.assertIn('binary_error', gbm.evals_result_['training'])

        # invalid multiclass metric is replaced with binary alternative for custom objective
        gbm = lgb.LGBMClassifier(objective=custom_dummy_obj,
                                 **params).fit(eval_metric='multi_logloss', **params_fit)
        self.assertEqual(len(gbm.evals_result_['training']), 1)
        self.assertIn('binary_logloss', gbm.evals_result_['training'])

    def test_inf_handle(self):
        nrows = 100
        ncols = 10
        X = np.random.randn(nrows, ncols)
        y = np.random.randn(nrows) + np.full(nrows, 1e30)
        weight = np.full(nrows, 1e10)
        params = {'n_estimators': 20, 'verbose': -1}
        params_fit = {'X': X, 'y': y, 'sample_weight': weight, 'eval_set': (X, y),
                      'verbose': False, 'early_stopping_rounds': 5}
        gbm = lgb.LGBMRegressor(**params).fit(**params_fit)
        np.testing.assert_allclose(gbm.evals_result_['training']['l2'], np.inf)

    def test_nan_handle(self):
        nrows = 100
        ncols = 10
        X = np.random.randn(nrows, ncols)
        y = np.random.randn(nrows) + np.full(nrows, 1e30)
        weight = np.zeros(nrows)
        params = {'n_estimators': 20, 'verbose': -1}
        params_fit = {'X': X, 'y': y, 'sample_weight': weight, 'eval_set': (X, y),
                      'verbose': False, 'early_stopping_rounds': 5}
        gbm = lgb.LGBMRegressor(**params).fit(**params_fit)
        np.testing.assert_allclose(gbm.evals_result_['training']['l2'], np.nan)

    def test_first_metric_only(self):

        def fit_and_check(eval_set_names, metric_names, assumed_iteration, first_metric_only):
            params['first_metric_only'] = first_metric_only
            gbm = lgb.LGBMRegressor(**params).fit(**params_fit)
            self.assertEqual(len(gbm.evals_result_), len(eval_set_names))
            for eval_set_name in eval_set_names:
                self.assertIn(eval_set_name, gbm.evals_result_)
                self.assertEqual(len(gbm.evals_result_[eval_set_name]), len(metric_names))
                for metric_name in metric_names:
                    self.assertIn(metric_name, gbm.evals_result_[eval_set_name])

                    actual = len(gbm.evals_result_[eval_set_name][metric_name])
                    expected = assumed_iteration + (params_fit['early_stopping_rounds']
                                                    if eval_set_name != 'training'
                                                    and assumed_iteration != gbm.n_estimators else 0)
                    self.assertEqual(expected, actual)
                    self.assertEqual(assumed_iteration if eval_set_name != 'training' else gbm.n_estimators,
                                     gbm.best_iteration_)

        X, y = load_boston(True)
        X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.2, random_state=42)
        X_test1, X_test2, y_test1, y_test2 = train_test_split(X_test, y_test, test_size=0.5, random_state=72)
        params = {'n_estimators': 30,
                  'learning_rate': 0.8,
                  'num_leaves': 15,
                  'verbose': -1,
                  'seed': 123}
        params_fit = {'X': X_train,
                      'y': y_train,
                      'early_stopping_rounds': 5,
                      'verbose': False}

        iter_valid1_l1 = 3
        iter_valid1_l2 = 18
        iter_valid2_l1 = 11
        iter_valid2_l2 = 7
        self.assertEqual(len(set([iter_valid1_l1, iter_valid1_l2, iter_valid2_l1, iter_valid2_l2])), 4)
        iter_min_l1 = min([iter_valid1_l1, iter_valid2_l1])
        iter_min_l2 = min([iter_valid1_l2, iter_valid2_l2])
        iter_min = min([iter_min_l1, iter_min_l2])
        iter_min_valid1 = min([iter_valid1_l1, iter_valid1_l2])

        # training data as eval_set
        params_fit['eval_set'] = (X_train, y_train)
        fit_and_check(['training'], ['l2'], 30, False)
        fit_and_check(['training'], ['l2'], 30, True)

        # feval
        params['metric'] = 'None'
        params_fit['eval_metric'] = lambda preds, train_data: [decreasing_metric(preds, train_data),
                                                               constant_metric(preds, train_data)]
        params_fit['eval_set'] = (X_test1, y_test1)
        fit_and_check(['valid_0'], ['decreasing_metric', 'error'], 1, False)
        fit_and_check(['valid_0'], ['decreasing_metric', 'error'], 30, True)
        params_fit['eval_metric'] = lambda preds, train_data: [constant_metric(preds, train_data),
                                                               decreasing_metric(preds, train_data)]
        fit_and_check(['valid_0'], ['decreasing_metric', 'error'], 1, True)

        # single eval_set
        params.pop('metric')
        params_fit.pop('eval_metric')
        fit_and_check(['valid_0'], ['l2'], iter_valid1_l2, False)
        fit_and_check(['valid_0'], ['l2'], iter_valid1_l2, True)

        params_fit['eval_metric'] = "l2"
        fit_and_check(['valid_0'], ['l2'], iter_valid1_l2, False)
        fit_and_check(['valid_0'], ['l2'], iter_valid1_l2, True)

        params_fit['eval_metric'] = "l1"
        fit_and_check(['valid_0'], ['l1', 'l2'], iter_min_valid1, False)
        fit_and_check(['valid_0'], ['l1', 'l2'], iter_valid1_l1, True)

        params_fit['eval_metric'] = ["l1", "l2"]
        fit_and_check(['valid_0'], ['l1', 'l2'], iter_min_valid1, False)
        fit_and_check(['valid_0'], ['l1', 'l2'], iter_valid1_l1, True)

        params_fit['eval_metric'] = ["l2", "l1"]
        fit_and_check(['valid_0'], ['l1', 'l2'], iter_min_valid1, False)
        fit_and_check(['valid_0'], ['l1', 'l2'], iter_valid1_l2, True)

        params_fit['eval_metric'] = ["l2", "regression", "mse"]  # test aliases
        fit_and_check(['valid_0'], ['l2'], iter_valid1_l2, False)
        fit_and_check(['valid_0'], ['l2'], iter_valid1_l2, True)

        # two eval_set
        params_fit['eval_set'] = [(X_test1, y_test1), (X_test2, y_test2)]
        params_fit['eval_metric'] = ["l1", "l2"]
        fit_and_check(['valid_0', 'valid_1'], ['l1', 'l2'], iter_min_l1, True)
        params_fit['eval_metric'] = ["l2", "l1"]
        fit_and_check(['valid_0', 'valid_1'], ['l1', 'l2'], iter_min_l2, True)

        params_fit['eval_set'] = [(X_test2, y_test2), (X_test1, y_test1)]
        params_fit['eval_metric'] = ["l1", "l2"]
        fit_and_check(['valid_0', 'valid_1'], ['l1', 'l2'], iter_min, False)
        fit_and_check(['valid_0', 'valid_1'], ['l1', 'l2'], iter_min_l1, True)
        params_fit['eval_metric'] = ["l2", "l1"]
        fit_and_check(['valid_0', 'valid_1'], ['l1', 'l2'], iter_min, False)
        fit_and_check(['valid_0', 'valid_1'], ['l1', 'l2'], iter_min_l2, True)

    def test_class_weight(self):
        X, y = load_digits(10, True)
        X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.2, random_state=42)
        y_train_str = y_train.astype('str')
        y_test_str = y_test.astype('str')
        gbm = lgb.LGBMClassifier(n_estimators=10, class_weight='balanced', silent=True)
        gbm.fit(X_train, y_train,
                eval_set=[(X_train, y_train), (X_test, y_test), (X_test, y_test),
                          (X_test, y_test), (X_test, y_test)],
                eval_class_weight=['balanced', None, 'balanced', {1: 10, 4: 20}, {5: 30, 2: 40}],
                verbose=False)
        for eval_set1, eval_set2 in itertools.combinations(gbm.evals_result_.keys(), 2):
            for metric in gbm.evals_result_[eval_set1]:
                np.testing.assert_raises(AssertionError,
                                         np.testing.assert_allclose,
                                         gbm.evals_result_[eval_set1][metric],
                                         gbm.evals_result_[eval_set2][metric])
        gbm_str = lgb.LGBMClassifier(n_estimators=10, class_weight='balanced', silent=True)
        gbm_str.fit(X_train, y_train_str,
                    eval_set=[(X_train, y_train_str), (X_test, y_test_str),
                              (X_test, y_test_str), (X_test, y_test_str), (X_test, y_test_str)],
                    eval_class_weight=['balanced', None, 'balanced', {'1': 10, '4': 20}, {'5': 30, '2': 40}],
                    verbose=False)
        for eval_set1, eval_set2 in itertools.combinations(gbm_str.evals_result_.keys(), 2):
            for metric in gbm_str.evals_result_[eval_set1]:
                np.testing.assert_raises(AssertionError,
                                         np.testing.assert_allclose,
                                         gbm_str.evals_result_[eval_set1][metric],
                                         gbm_str.evals_result_[eval_set2][metric])
        for eval_set in gbm.evals_result_:
            for metric in gbm.evals_result_[eval_set]:
                np.testing.assert_allclose(gbm.evals_result_[eval_set][metric],
                                           gbm_str.evals_result_[eval_set][metric])

    def test_continue_training_with_model(self):
        X, y = load_digits(3, True)
        X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.1, random_state=42)
        init_gbm = lgb.LGBMClassifier(n_estimators=5).fit(X_train, y_train, eval_set=(X_test, y_test),
                                                          verbose=False)
        gbm = lgb.LGBMClassifier(n_estimators=5).fit(X_train, y_train, eval_set=(X_test, y_test),
                                                     verbose=False, init_model=init_gbm)
        self.assertEqual(len(init_gbm.evals_result_['valid_0']['multi_logloss']),
                         len(gbm.evals_result_['valid_0']['multi_logloss']))
        self.assertEqual(len(init_gbm.evals_result_['valid_0']['multi_logloss']), 5)
        self.assertLess(gbm.evals_result_['valid_0']['multi_logloss'][-1],
                        init_gbm.evals_result_['valid_0']['multi_logloss'][-1])