You are viewing a plain text version of this content. The canonical link for it is here.
Posted to issues@flink.apache.org by GitBox <gi...@apache.org> on 2022/04/13 10:33:51 UTC
[GitHub] [flink-ml] lindong28 commented on a diff in pull request #82: [FLINK-27072] Add Transformer of Bucketizer
lindong28 commented on code in PR #82:
URL: https://github.com/apache/flink-ml/pull/82#discussion_r849333000
##########
flink-ml-lib/src/main/java/org/apache/flink/ml/feature/bucketizer/Bucketizer.java:
##########
@@ -0,0 +1,179 @@
+/*
+ * 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.bucketizer;
+
+import org.apache.flink.api.common.functions.FlatMapFunction;
+import org.apache.flink.api.common.typeinfo.BasicTypeInfo;
+import org.apache.flink.api.common.typeinfo.TypeInformation;
+import org.apache.flink.api.java.typeutils.RowTypeInfo;
+import org.apache.flink.ml.api.Transformer;
+import org.apache.flink.ml.common.datastream.TableUtils;
+import org.apache.flink.ml.param.Param;
+import org.apache.flink.ml.util.ParamUtils;
+import org.apache.flink.ml.util.ReadWriteUtils;
+import org.apache.flink.streaming.api.datastream.DataStream;
+import org.apache.flink.table.api.Table;
+import org.apache.flink.table.api.bridge.java.StreamTableEnvironment;
+import org.apache.flink.table.api.internal.TableImpl;
+import org.apache.flink.types.Row;
+import org.apache.flink.util.Collector;
+import org.apache.flink.util.Preconditions;
+
+import org.apache.commons.lang3.ArrayUtils;
+
+import java.io.IOException;
+import java.util.Arrays;
+import java.util.HashMap;
+import java.util.Map;
+
+/**
+ * Bucketizer is a transformer that maps multiple columns of continuous features to multiple columns
+ * of discrete features, i.e., buckets IDs.
+ */
+public class Bucketizer implements Transformer<Bucketizer>, BucketizerParams<Bucketizer> {
+ private final Map<Param<?>, Object> paramMap = new HashMap<>();
+
+ public Bucketizer() {
+ ParamUtils.initializeMapWithDefaultValues(paramMap, this);
+ }
+
+ @Override
+ public Table[] transform(Table... inputs) {
+ Preconditions.checkArgument(inputs.length == 1);
+ String[] inputCols = getInputCols();
+ String[] outputCols = getOutputCols();
+ Double[][] splitsArray = getSplitsArray();
+
+ Preconditions.checkArgument(inputCols.length == outputCols.length);
+ Preconditions.checkArgument(inputCols.length == splitsArray.length);
+ for (Double[] splits : splitsArray) {
+ Preconditions.checkArgument(
+ splits.length >= 3,
+ "Illegal value for "
+ + BucketizerParams.SPLITS_ARRAY
+ + ". See param "
+ + BucketizerParams.SPLITS_ARRAY
+ + " for details.");
+ for (int j = 1; j < splits.length; j++) {
+ Preconditions.checkArgument(
+ splits[j] > splits[j - 1],
+ "Illegal value for "
+ + BucketizerParams.SPLITS_ARRAY
+ + ". See param "
+ + BucketizerParams.SPLITS_ARRAY
+ + " for details.");
+ }
+ }
+
+ StreamTableEnvironment tEnv =
+ (StreamTableEnvironment) ((TableImpl) inputs[0]).getTableEnvironment();
+
+ RowTypeInfo inputTypeInfo = TableUtils.getRowTypeInfo(inputs[0].getResolvedSchema());
+ TypeInformation<?>[] outputTypes = new TypeInformation[outputCols.length];
+ Arrays.fill(outputTypes, BasicTypeInfo.INT_TYPE_INFO);
+ RowTypeInfo outputTypeInfo =
+ new RowTypeInfo(
+ ArrayUtils.addAll(inputTypeInfo.getFieldTypes(), outputTypes),
+ ArrayUtils.addAll(inputTypeInfo.getFieldNames(), getOutputCols()));
+
+ DataStream<Row> result =
+ tEnv.toDataStream(inputs[0])
+ .flatMap(
+ new FindBucketFunction(inputCols, splitsArray, getHandleInvalid()),
+ outputTypeInfo);
+ return new Table[] {tEnv.fromDataStream(result)};
+ }
+
+ /** Finds the bucket index for each continuous feature of an input data point. */
+ private static class FindBucketFunction implements FlatMapFunction<Row, Row> {
+ private final String[] inputCols;
+ private final String handleInvalid;
+ private final Double[][] splitsArray;
+
+ public FindBucketFunction(
+ String[] inputCols, Double[][] splitsArray, String handleInvalid) {
+ this.inputCols = inputCols;
+ this.splitsArray = splitsArray;
+ this.handleInvalid = handleInvalid;
+ }
+
+ @Override
+ public void flatMap(Row value, Collector<Row> out) {
+ Row outputRow = new Row(inputCols.length);
+
+ for (int i = 0; i < inputCols.length; i++) {
+ double feature = ((Number) value.getField(inputCols[i])).doubleValue();
+ Double[] splits = splitsArray[i];
+ boolean isInvalid = false;
+
+ if (!Double.isNaN(feature)) {
+ int index = Arrays.binarySearch(splits, feature);
+ if (index >= 0) {
+ if (index == inputCols.length - 1) {
+ index--;
+ }
+ outputRow.setField(i, index);
+ } else {
+ index = -index - 1;
+ if (index == 0 || index == inputCols.length) {
+ isInvalid = true;
+ } else {
+ outputRow.setField(i, index - 1);
+ }
+ }
+ } else {
+ isInvalid = true;
+ }
+
+ if (isInvalid) {
+ switch (handleInvalid) {
+ case BucketizerParams.ERROR_INVALID:
+ throw new RuntimeException(
+ "The input contains invalid value. See "
+ + BucketizerParams.HANDLE_INVALID
+ + " parameter for more options.");
+ case BucketizerParams.SKIP_INVALID:
+ return;
+ case BucketizerParams.KEEP_INVALID:
+ outputRow.setField(i, splits.length - 1);
+ break;
Review Comment:
@yunfengzhou-hub by the `break` here, the code breaks out of the switch loop, and the for loop continues. So in the case described above, the current implementation would still output `[a*, splits.length-1, c*]`.
--
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: issues-unsubscribe@flink.apache.org
For queries about this service, please contact Infrastructure at:
users@infra.apache.org