lindong28 commented on a change in pull request #52: URL: https://github.com/apache/flink-ml/pull/52#discussion_r831837017
########## File path: flink-ml-lib/src/test/java/org/apache/flink/ml/feature/stringindexer/StringIndexerTest.java ########## @@ -0,0 +1,319 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one + * or more contributor license agreements. See the NOTICE file + * distributed with this work for additional information + * regarding copyright ownership. The ASF licenses this file + * to you 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. + */ + +package org.apache.flink.ml.feature.stringindexer; + +import org.apache.flink.api.common.restartstrategy.RestartStrategies; +import org.apache.flink.configuration.Configuration; +import org.apache.flink.ml.util.ReadWriteUtils; +import org.apache.flink.ml.util.StageTestUtils; +import org.apache.flink.streaming.api.environment.ExecutionCheckpointingOptions; +import org.apache.flink.streaming.api.environment.StreamExecutionEnvironment; +import org.apache.flink.table.api.Table; +import org.apache.flink.table.api.bridge.java.StreamTableEnvironment; +import org.apache.flink.test.util.TestBaseUtils; +import org.apache.flink.types.Row; + +import org.apache.commons.collections.IteratorUtils; +import org.junit.Before; +import org.junit.Rule; +import org.junit.Test; +import org.junit.rules.TemporaryFolder; + +import java.util.Arrays; +import java.util.Collections; +import java.util.HashSet; +import java.util.List; +import java.util.Set; + +import static org.junit.Assert.assertArrayEquals; +import static org.junit.Assert.assertEquals; +import static org.junit.Assert.assertTrue; +import static org.junit.Assert.fail; + +/** Tests the {@link StringIndexer} and {@link StringIndexerModel}. */ +public class StringIndexerTest { + @Rule public final TemporaryFolder tempFolder = new TemporaryFolder(); + + private StreamExecutionEnvironment env; + private StreamTableEnvironment tEnv; + private Table trainTable; + private Table predictTable; + + private String[][] expectedAlphabeticAscModelData; + private List<Row> expectedAlphabeticAscPredictData; + private List<Row> expectedAlphabeticDescPredictData; + private List<Row> expectedFreqAscPredictData; + private List<Row> expectedFreqDescPredictData; + + @Before + public void before() { + Configuration config = new Configuration(); + config.set(ExecutionCheckpointingOptions.ENABLE_CHECKPOINTS_AFTER_TASKS_FINISH, true); + env = StreamExecutionEnvironment.getExecutionEnvironment(config); + env.setParallelism(4); + env.enableCheckpointing(100); + env.setRestartStrategy(RestartStrategies.noRestart()); + tEnv = StreamTableEnvironment.create(env); + + List<Row> trainData = + Arrays.asList( + Row.of("a", 1.), + Row.of("b", 1.), + Row.of("b", 2.0), + Row.of("c", 0.0), + Row.of("d", 2.), + Row.of("a", 2.), + Row.of("b", 2.), + Row.of("b", -1.), + Row.of("a", -1.), + Row.of("c", -1.)); + trainTable = + tEnv.fromDataStream(env.fromCollection(trainData)).as("inputCol1", "inputCol2"); + + List<Row> predictData = Arrays.asList(Row.of("a", 2.), Row.of("b", 1.), Row.of("e", 2.)); + predictTable = + tEnv.fromDataStream(env.fromCollection(predictData)).as("inputCol1", "inputCol2"); + + expectedAlphabeticAscModelData = + new String[][] {{"a", "b", "c", "d"}, {"-1.0", "0.0", "1.0", "2.0"}}; + expectedAlphabeticAscPredictData = + Arrays.asList(Row.of("a", 2., 0, 3), Row.of("b", 1., 1, 2), Row.of("e", 2.0, 4, 3)); + expectedAlphabeticDescPredictData = + Arrays.asList(Row.of("a", 2., 3, 0), Row.of("b", 1., 2, 1), Row.of("e", 2.0, 4, 0)); + expectedFreqAscPredictData = + Arrays.asList(Row.of("a", 2., 2, 3), Row.of("b", 1., 3, 1), Row.of("e", 2.0, 4, 3)); + expectedFreqDescPredictData = + Arrays.asList(Row.of("a", 2., 1, 0), Row.of("b", 1., 0, 2), Row.of("e", 2.0, 4, 0)); + } + + @Test + public void testFitParam() { + StringIndexer stringIndexer = new StringIndexer(); + assertEquals(stringIndexer.getStringOrderType(), StringIndexerParams.RANDOM_ORDER); + assertEquals(stringIndexer.getHandleInvalid(), StringIndexerParams.ERROR_INVALID); + + stringIndexer + .setInputCols("inputCol1", "inputCol2") + .setOutputCols("outputCol1", "outputCol1") + .setStringOrderType(StringIndexerParams.ALPHABET_ASC_ORDER) + .setHandleInvalid(StringIndexerParams.SKIP_INVALID); + + assertArrayEquals(new String[] {"inputCol1", "inputCol2"}, stringIndexer.getInputCols()); + assertArrayEquals(new String[] {"outputCol1", "outputCol1"}, stringIndexer.getOutputCols()); + assertEquals(stringIndexer.getStringOrderType(), StringIndexerParams.ALPHABET_ASC_ORDER); + assertEquals(stringIndexer.getHandleInvalid(), StringIndexerParams.SKIP_INVALID); + } + + @Test + public void testPredictParam() { + StringIndexer stringIndexer = + new StringIndexer() + .setInputCols("inputCol1", "inputCol2") + .setOutputCols("outputCol1", "outputCol2") + .setStringOrderType(StringIndexerParams.ALPHABET_ASC_ORDER) + .setHandleInvalid(StringIndexerParams.SKIP_INVALID); + Table output = stringIndexer.fit(trainTable).transform(predictTable)[0]; + assertEquals( + Arrays.asList("inputCol1", "inputCol2", "outputCol1", "outputCol2"), + output.getResolvedSchema().getColumnNames()); + } + + @Test + public void testStringOrderType() throws Exception { + StringIndexer stringIndexer = + new StringIndexer() + .setInputCols("inputCol1", "inputCol2") + .setOutputCols("outputCol1", "outputCol2") + .setHandleInvalid(StringIndexerParams.KEEP_INVALID); + + Table output; + List<Row> predictedResult; + + // alphabetAsc + stringIndexer.setStringOrderType(StringIndexerParams.ALPHABET_ASC_ORDER); + output = stringIndexer.fit(trainTable).transform(predictTable)[0]; + predictedResult = IteratorUtils.toList(tEnv.toDataStream(output).executeAndCollect()); + verifyPredictionResult(expectedAlphabeticAscPredictData, predictedResult); + + // alphabetDesc + stringIndexer.setStringOrderType(StringIndexerParams.ALPHABET_DESC_ORDER); + output = stringIndexer.fit(trainTable).transform(predictTable)[0]; + predictedResult = IteratorUtils.toList(tEnv.toDataStream(output).executeAndCollect()); + verifyPredictionResult(expectedAlphabeticDescPredictData, predictedResult); + + // frequencyAsc + stringIndexer.setStringOrderType(StringIndexerParams.FREQUENCY_ASC_ORDER); + output = stringIndexer.fit(trainTable).transform(predictTable)[0]; + predictedResult = IteratorUtils.toList(tEnv.toDataStream(output).executeAndCollect()); + verifyPredictionResult(expectedFreqAscPredictData, predictedResult); + + // frequencyDesc + stringIndexer.setStringOrderType(StringIndexerParams.FREQUENCY_DESC_ORDER); + output = stringIndexer.fit(trainTable).transform(predictTable)[0]; + predictedResult = IteratorUtils.toList(tEnv.toDataStream(output).executeAndCollect()); + verifyPredictionResult(expectedFreqDescPredictData, predictedResult); + + // random + stringIndexer.setStringOrderType(StringIndexerParams.RANDOM_ORDER); + output = stringIndexer.fit(trainTable).transform(predictTable)[0]; + predictedResult = IteratorUtils.toList(tEnv.toDataStream(output).executeAndCollect()); + Set<Integer> distinctStringsCol1 = new HashSet<>(); + Set<Integer> distinctStringsCol2 = new HashSet<>(); + int index; + for (Row r : predictedResult) { + index = (Integer) r.getField(2); + distinctStringsCol1.add(index); + assertTrue(index >= 0 && index <= 4); + index = (Integer) r.getField(3); + assertTrue(index >= 0 && index <= 3); + distinctStringsCol2.add(index); + } + assertEquals(3, distinctStringsCol1.size()); + assertEquals(2, distinctStringsCol2.size()); + } + + @Test + public void testHandleInvalid() throws Exception { + StringIndexer stringIndexer = + new StringIndexer() + .setInputCols("inputCol1", "inputCol2") + .setOutputCols("outputCol1", "outputCol2") + .setStringOrderType(StringIndexerParams.ALPHABET_ASC_ORDER); + + Table output; + List<Row> expectedResult; + + // keep invalid data + stringIndexer.setHandleInvalid(StringIndexerParams.KEEP_INVALID); + output = stringIndexer.fit(trainTable).transform(predictTable)[0]; + List<Row> predictedResult = + IteratorUtils.toList(tEnv.toDataStream(output).executeAndCollect()); + verifyPredictionResult(expectedAlphabeticAscPredictData, predictedResult); + + // skip invalid data + stringIndexer.setHandleInvalid(StringIndexerParams.SKIP_INVALID); + output = stringIndexer.fit(trainTable).transform(predictTable)[0]; + predictedResult = IteratorUtils.toList(tEnv.toDataStream(output).executeAndCollect()); + expectedResult = Arrays.asList(Row.of("a", 2., 0, 3), Row.of("b", 1., 1, 2)); + verifyPredictionResult(expectedResult, predictedResult); + + // throw exception on invalid data + stringIndexer.setHandleInvalid(StringIndexerParams.ERROR_INVALID); + try { + output = stringIndexer.fit(trainTable).transform(predictTable)[0]; + IteratorUtils.toList(tEnv.toDataStream(output).executeAndCollect()); + fail(); + } catch (Exception e) { + assertEquals( + "The input contains unseen string: e. " + + "To handle unseen strings, set Param handleInvalid to keep.", + e.getCause().getCause().getCause().getCause().getCause().getMessage()); + } + } + + @Test + public void testFitAndPredict() throws Exception { + StringIndexer stringIndexer = + new StringIndexer() + .setInputCols("inputCol1", "inputCol2") + .setOutputCols("outputCol1", "outputCol2") + .setStringOrderType(StringIndexerParams.ALPHABET_ASC_ORDER) + .setHandleInvalid(StringIndexerParams.KEEP_INVALID); + Table output = stringIndexer.fit(trainTable).transform(predictTable)[0]; + List<Row> predictedResult = + IteratorUtils.toList(tEnv.toDataStream(output).executeAndCollect()); + verifyPredictionResult(expectedAlphabeticAscPredictData, predictedResult); + } + + @Test + public void testSaveLoadAndPredict() throws Exception { + StringIndexer stringIndexer = + new StringIndexer() + .setInputCols("inputCol1", "inputCol2") + .setOutputCols("outputCol1", "outputCol2") + .setStringOrderType(StringIndexerParams.ALPHABET_ASC_ORDER) + .setHandleInvalid(StringIndexerParams.KEEP_INVALID); + stringIndexer = + StageTestUtils.saveAndReload( + env, stringIndexer, tempFolder.newFolder().getAbsolutePath()); + StringIndexerModel model = stringIndexer.fit(trainTable); + model = StageTestUtils.saveAndReload(env, model, tempFolder.newFolder().getAbsolutePath()); + assertEquals( + Collections.singletonList("stringsArray"), + model.getModelData()[0].getResolvedSchema().getColumnNames()); + Table output = model.transform(predictTable)[0]; + List<Row> predictedResult = + IteratorUtils.toList(tEnv.toDataStream(output).executeAndCollect()); + verifyPredictionResult(expectedAlphabeticAscPredictData, predictedResult); + } + + @Test + public void testGetModelData() throws Exception { + StringIndexerModel model = + new StringIndexer() + .setInputCols("inputCol1", "inputCol2") + .setOutputCols("outputCol1", "outputCol2") + .setStringOrderType(StringIndexerParams.ALPHABET_ASC_ORDER) + .fit(trainTable); + StringIndexerModelData modelData = + StringIndexerModelData.getModelDataStream(model.getModelData()[0]) Review comment: Sounds good. Thanks. -- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. To unsubscribe, e-mail: [email protected] For queries about this service, please contact Infrastructure at: [email protected]
