Add unittest

function/python/brightics/function/classification/test/ada_boost_classification_test.py

@@ -0,0 +1,54 @@
+"""
+    Copyright 2019 Samsung SDS
+
+    Licensed under the Apache License, Version 2.0 (the "License");
+    you may not use this file except in compliance with the License.
+    You may obtain a copy of the License at
+
+        http://www.apache.org/licenses/LICENSE-2.0
+
+    Unless required by applicable law or agreed to in writing, software
+    distributed under the License is distributed on an "AS IS" BASIS,
+    WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+    See the License for the specific language governing permissions and
+    limitations under the License.
+"""
+
+
+from brightics.function.classification.ada_boost_classification import ada_boost_classification_train
+from brightics.function.classification.ada_boost_classification import ada_boost_classification_predict
+from brightics.common.datasets import load_iris
+import unittest
+import pandas as pd
+import numpy as np
+
+
+class ADABoostClassification(unittest.TestCase):
+
+    def setUp(self):
+        print("*** ADA Boost Classification UnitTest Start ***")
+        self.testdata = load_iris()
+
+    def tearDown(self):
+        print("*** ADA Boost Classification UnitTest End ***")
+
+    def test(self):
+        ada_train = ada_boost_classification_train(self.testdata, feature_cols=['sepal_length', 'sepal_width', 'petal_length', 'petal_width'], label_col='species', random_state=12345)
+        ada_model = ada_train['model']['classifier']
+        estimators = ada_model.estimators_ if hasattr(ada_model, 'estimators_') else None
+        classes = ada_model.classes_ if hasattr(ada_model, 'classes_') else None
+        n_classes = ada_model.n_classes_ if hasattr(ada_model, 'n_classes_') else None
+        estimator_weights = ada_model.estimator_weights_ if hasattr(ada_model, 'estimator_weights_') else None
+        estimator_errors = ada_model.estimator_errors_ if hasattr(ada_model, 'estimator_errors_') else None
+        feature_importances = ada_model.feature_importances_ if hasattr(ada_model, 'feature_importances_') else None
+
+        np.testing.assert_array_equal([round(x, 15) for x in estimator_weights], [1.000000000000000,1.000000000000000,1.000000000000000,1.000000000000000,1.000000000000000,1.000000000000000,1.000000000000000,1.000000000000000,1.000000000000000,1.000000000000000,1.000000000000000,1.000000000000000,1.000000000000000,1.000000000000000,1.000000000000000,1.000000000000000,1.000000000000000,1.000000000000000,1.000000000000000,1.000000000000000,1.000000000000000,1.000000000000000,1.000000000000000,1.000000000000000,1.000000000000000,1.000000000000000,1.000000000000000,1.000000000000000,1.000000000000000,1.000000000000000,1.000000000000000,1.000000000000000,1.000000000000000,1.000000000000000,1.000000000000000,1.000000000000000,1.000000000000000,1.000000000000000,1.000000000000000,1.000000000000000,1.000000000000000,1.000000000000000,1.000000000000000,1.000000000000000,1.000000000000000,1.000000000000000,1.000000000000000,1.000000000000000,1.000000000000000,1.000000000000000])
+        np.testing.assert_array_equal([round(x, 15) for x in estimator_errors], [0.333333333333333,0.060002258916742,0.333355372258786,0.120397413941043,0.333354186061486,0.000246496771993,0.333379977227275,0.000324696290229,0.333365707421006,0.000377301321863,0.333379703755078,0.000349084998612,0.333368068990284,0.000377542088862,0.333377933238813,0.000352106375872,0.333369432842104,0.000372712089944,0.333376615861411,0.000354056001133,0.333370427040033,0.000369265051501,0.333375673592510,0.000355694330200,0.333371161245900,0.000366896338034,0.333374996123643,0.000357017448219,0.333371702726590,0.000365249477557,0.333374506534519,0.000358050974252,0.333372101321344,0.000364091056983,0.333374151478965,0.000358841484973,0.333372394265748,0.000363268581665,0.333373893375635,0.000359437901881,0.333372609287275,0.000362680377430,0.333373705442854,0.000359883750645,0.333372766955686,0.000362257368360,0.333373568446543,0.000360214927717,0.333372882480536,0.000361951872769])
+        np.testing.assert_array_equal([round(x, 15) for x in feature_importances], [0.000000000000000,0.000000000000000,0.440000000000000,0.560000000000000])
+
+        predict = ada_boost_classification_predict(self.testdata, ada_train['model'])
+        prob1 = predict['out_table']['probability_1']
+        prob2 = predict['out_table']['probability_2']
+        np.testing.assert_array_equal([round(x, 15) for x in prob1], [0.000006788589536,0.000006788589536,0.000006788589536,0.000006788589536,0.000006788589536,0.000006788589536,0.000006788589536,0.000006788589536,0.000006788589536,0.000006788589536,0.000006788589536,0.000006788589536,0.000006788589536,0.000006788589536,0.000006788589536,0.000006788589536,0.000006788589536,0.000006788589536,0.000006788589536,0.000006788589536,0.000006788589536,0.000006788589536,0.000006788589536,0.000006788589536,0.000006788589536,0.000006788589536,0.000006788589536,0.000006788589536,0.000006788589536,0.000006788589536,0.000006788589536,0.000006788589536,0.000006788589536,0.000006788589536,0.000006788589536,0.000006788589536,0.000006788589536,0.000006788589536,0.000006788589536,0.000006788589536,0.000006788589536,0.000006788589536,0.000006788589536,0.000006788589536,0.000006788589536,0.000006788589536,0.000006788589536,0.000006788589536,0.000006788589536,0.000006788589536,0.976767518408969,0.976767518408969,0.976767518408969,0.976767518408969,0.976767518408969,0.976767518408969,0.976767518408969,0.976767518408969,0.976767518408969,0.976767518408969,0.976767518408969,0.976767518408969,0.976767518408969,0.976767518408969,0.976767518408969,0.976767518408969,0.976767518408969,0.976767518408969,0.976767518408969,0.976767518408969,0.449747463798886,0.976767518408969,0.976767518408969,0.976767518408969,0.976767518408969,0.976767518408969,0.976767518408969,0.024737382237678,0.976767518408969,0.976767518408969,0.976767518408969,0.976767518408969,0.976767518408969,0.535991707392255,0.976767518408969,0.976767518408969,0.976767518408969,0.976767518408969,0.976767518408969,0.976767518408969,0.976767518408969,0.976767518408969,0.976767518408969,0.976767518408969,0.976767518408969,0.976767518408969,0.976767518408969,0.976767518408969,0.976767518408969,0.976767518408969,0.021963346892192,0.021963346892192,0.021963346892192,0.021963346892192,0.021963346892192,0.021963346892192,0.480033074405363,0.021963346892192,0.021963346892192,0.021963346892192,0.021963346892192,0.021963346892192,0.021963346892192,0.021963346892192,0.021963346892192,0.021963346892192,0.021963346892192,0.021963346892192,0.021963346892192,0.535991707392255,0.021963346892192,0.449747463798886,0.021963346892192,0.449747463798886,0.021963346892192,0.021963346892192,0.449747463798886,0.449747463798886,0.021963346892192,0.535991707392255,0.021963346892192,0.021963346892192,0.021963346892192,0.535991707392255,0.535991707392255,0.021963346892192,0.021963346892192,0.021963346892192,0.449747463798886,0.021963346892192,0.021963346892192,0.021963346892192,0.021963346892192,0.021963346892192,0.021963346892192,0.021963346892192,0.021963346892192,0.021963346892192,0.021963346892192,0.021963346892192])
+        np.testing.assert_array_equal([round(x, 15) for x in prob2], [0.000000161467062,0.000000161467062,0.000000161467062,0.000000161467062,0.000000161467062,0.000000161467062,0.000000161467062,0.000000161467062,0.000000161467062,0.000000161467062,0.000000161467062,0.000000161467062,0.000000161467062,0.000000161467062,0.000000161467062,0.000000161467062,0.000000161467062,0.000000161467062,0.000000161467062,0.000000161467062,0.000000161467062,0.000000161467062,0.000000161467062,0.000000161467062,0.000000161467062,0.000000161467062,0.000000161467062,0.000000161467062,0.000000161467062,0.000000161467062,0.000000161467062,0.000000161467062,0.000000161467062,0.000000161467062,0.000000161467062,0.000000161467062,0.000000161467062,0.000000161467062,0.000000161467062,0.000000161467062,0.000000161467062,0.000000161467062,0.000000161467062,0.000000161467062,0.000000161467062,0.000000161467062,0.000000161467062,0.000000161467062,0.000000161467062,0.000000161467062,0.023232481545849,0.023232481545849,0.023232481545849,0.023232481545849,0.023232481545849,0.023232481545849,0.023232481545849,0.023232481545849,0.023232481545849,0.023232481545849,0.023232481545849,0.023232481545849,0.023232481545849,0.023232481545849,0.023232481545849,0.023232481545849,0.023232481545849,0.023232481545849,0.023232481545849,0.023232481545849,0.550252536200809,0.023232481545849,0.023232481545849,0.023232481545849,0.023232481545849,0.023232481545849,0.023232481545849,0.975262617762322,0.023232481545849,0.023232481545849,0.023232481545849,0.023232481545849,0.023232481545849,0.464008292607232,0.023232481545849,0.023232481545849,0.023232481545849,0.023232481545849,0.023232481545849,0.023232481545849,0.023232481545849,0.023232481545849,0.023232481545849,0.023232481545849,0.023232481545849,0.023232481545849,0.023232481545849,0.023232481545849,0.023232481545849,0.023232481545849,0.978036653107808,0.978036653107808,0.978036653107808,0.978036653107808,0.978036653107808,0.978036653107808,0.519966925594152,0.978036653107808,0.978036653107808,0.978036653107808,0.978036653107808,0.978036653107808,0.978036653107808,0.978036653107808,0.978036653107808,0.978036653107808,0.978036653107808,0.978036653107808,0.978036653107808,0.464008292607232,0.978036653107808,0.550252536200809,0.978036653107808,0.550252536200809,0.978036653107808,0.978036653107808,0.550252536200809,0.550252536200809,0.978036653107808,0.464008292607232,0.978036653107808,0.978036653107808,0.978036653107808,0.464008292607232,0.464008292607232,0.978036653107808,0.978036653107808,0.978036653107808,0.550252536200809,0.978036653107808,0.978036653107808,0.978036653107808,0.978036653107808,0.978036653107808,0.978036653107808,0.978036653107808,0.978036653107808,0.978036653107808,0.978036653107808,0.978036653107808])
+

function/python/brightics/function/classification/test/decision_tree_classification_test.py

@@ -0,0 +1,65 @@
+"""
+    Copyright 2019 Samsung SDS
+    
+    Licensed under the Apache License, Version 2.0 (the "License");
+    you may not use this file except in compliance with the License.
+    You may obtain a copy of the License at
+    
+        http://www.apache.org/licenses/LICENSE-2.0
+    
+    Unless required by applicable law or agreed to in writing, software
+    distributed under the License is distributed on an "AS IS" BASIS,
+    WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+    See the License for the specific language governing permissions and
+    limitations under the License.
+"""
+
+import unittest
+from brightics.common.datasets import load_iris
+from brightics.function.classification.decision_tree_classification import decision_tree_classification_train, \
+    decision_tree_classification_predict
+
+
+class DecisionTreeClassificationTest(unittest.TestCase):
+
+    def setUp(self):
+        print("*** Decision Tree Classification UnitTest Start ***")
+        self.iris = load_iris()
+
+    def tearDown(self):
+        print("*** Decision Tree Classification UnitTest End ***")
+
+    def test_decision_tree_classification1(self):
+        train_out = decision_tree_classification_train(table=self.iris, feature_cols=['sepal_length', 'sepal_width', 'petal_length', 'petal_width'], label_col='species', random_state=12345)
+
+        table = train_out['model']['feature_importance']
+        self.assertAlmostEqual(table[0], 0.02666667, 6)
+        self.assertAlmostEqual(table[1], 0.0, 1)
+        self.assertAlmostEqual(table[2], 0.55072262, 6)
+        self.assertAlmostEqual(table[3], 0.42261071, 6)
+
+    def test_decision_tree_classification2(self):
+        train_out = decision_tree_classification_train(table=self.iris, feature_cols=['sepal_length', 'sepal_width', 'petal_length', 'petal_width'], label_col='species', random_state=12345)
+        predict_out = decision_tree_classification_predict(table=self.iris, model=train_out['model'])
+
+        table = predict_out['out_table'].values.tolist()
+        self.assertListEqual(table[0], [5.1, 3.5, 1.4, 0.2, 'setosa', 'setosa'])
+        self.assertListEqual(table[1], [4.9, 3.0, 1.4, 0.2, 'setosa', 'setosa'])
+        self.assertListEqual(table[2], [4.7, 3.2, 1.3, 0.2, 'setosa', 'setosa'])
+        self.assertListEqual(table[3], [4.6, 3.1, 1.5, 0.2, 'setosa', 'setosa'])
+        self.assertListEqual(table[4], [5.0, 3.6, 1.4, 0.2, 'setosa', 'setosa'])
+
+    def test_decision_tree_classification3(self):
+        train_out = decision_tree_classification_train(table=self.iris, feature_cols=['sepal_length', 'sepal_width', 'petal_length', 'petal_width'], label_col='species', random_state=12345, criterion='entropy', max_leaf_nodes=2, group_by=['species'])
+        predict_out = decision_tree_classification_predict(table=self.iris, model=train_out['model'])
+
+        table = predict_out['out_table'].values.tolist()
+        self.assertListEqual(table[0], [5.1, 3.5, 1.4, 0.2, 'setosa', 'setosa'])
+        self.assertListEqual(table[1], [4.9, 3.0, 1.4, 0.2, 'setosa', 'setosa'])
+        self.assertListEqual(table[2], [4.7, 3.2, 1.3, 0.2, 'setosa', 'setosa'])
+        self.assertListEqual(table[3], [4.6, 3.1, 1.5, 0.2, 'setosa', 'setosa'])
+        self.assertListEqual(table[4], [5.0, 3.6, 1.4, 0.2, 'setosa', 'setosa'])
+
+
+if __name__ == '__main__':
+    unittest.main()

function/python/brightics/function/classification/test/knn_classification_test.py

@@ -0,0 +1,48 @@
+"""
+    Copyright 2019 Samsung SDS
+    
+    Licensed under the Apache License, Version 2.0 (the "License");
+    you may not use this file except in compliance with the License.
+    You may obtain a copy of the License at
+    
+        http://www.apache.org/licenses/LICENSE-2.0
+    
+    Unless required by applicable law or agreed to in writing, software
+    distributed under the License is distributed on an "AS IS" BASIS,
+    WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+    See the License for the specific language governing permissions and
+    limitations under the License.
+"""
+
+import unittest
+import numpy as np
+from sklearn.model_selection import train_test_split
+from brightics.common.datasets import load_iris
+from brightics.function.classification import knn_classification
+
+
+class TestKNNClassification(unittest.TestCase):
+
+    def test_default(self):
+        df_iris = load_iris()
+        df_train, df_test = train_test_split(df_iris, random_state=12345)
+        df_res = knn_classification(train_table=df_train, test_table=df_test,
+                                    feature_cols=['sepal_length', 'sepal_width'], label_col='species',
+                                    k=5, algorithm='auto', leaf_size=30, p=2)['out_table']
+
+        self.assertListEqual(['versicolor', 'setosa', 'virginica', 'setosa', 'setosa'], df_res['prediction'].tolist()[:5], 'incorrect prediction')                                    
+        np.testing.assert_array_almost_equal([0.0, 1.0, 0.0, 1.0, 1.0], df_res['probability_0'].values[:5], 7, 'incorrect probability_0')
+        np.testing.assert_array_almost_equal([0.8, 0.0, 0.2, 0.0, 0.0], df_res['probability_1'].values[:5], 7, 'incorrect probability_1')
+        np.testing.assert_array_almost_equal([0.2, 0.0, 0.8, 0.0, 0.0], df_res['probability_2'].values[:5], 7, 'incorrect probability_2')
+
+    def test_optional(self):
+        df_iris = load_iris()
+        df_train, df_test = train_test_split(df_iris, random_state=12345)
+        df_res = knn_classification(train_table=df_train, test_table=df_test,
+                                    feature_cols=['sepal_length', 'sepal_width', 'petal_length'], label_col='species',
+                                    k=10, algorithm='auto', leaf_size=30, p=2)['out_table']
+
+        self.assertListEqual(['versicolor', 'setosa', 'versicolor', 'setosa', 'setosa'], df_res['prediction'].tolist()[:5], 'incorrect prediction')                                    
+        np.testing.assert_array_almost_equal([0.0, 1.0, 0.0, 1.0, 1.0], df_res['probability_0'].values[:5], 7, 'incorrect probability_0')
+        np.testing.assert_array_almost_equal([1.0, 0.0, 0.7, 0.0, 0.0], df_res['probability_1'].values[:5], 7, 'incorrect probability_1')
+        np.testing.assert_array_almost_equal([0.0, 0.0, 0.3, 0.0, 0.0], df_res['probability_2'].values[:5], 7, 'incorrect probability_2')