weibozhao commented on a change in pull request #24: URL: https://github.com/apache/flink-ml/pull/24#discussion_r767569310
########## File path: flink-ml-lib/src/main/java/org/apache/flink/ml/classification/knn/Knn.java ########## @@ -0,0 +1,214 @@ +/* + * 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.classification.knn; + +import org.apache.flink.api.common.functions.MapFunction; +import org.apache.flink.api.common.functions.RichMapPartitionFunction; +import org.apache.flink.api.common.typeinfo.TypeInformation; +import org.apache.flink.api.java.tuple.Tuple2; +import org.apache.flink.ml.api.Estimator; +import org.apache.flink.ml.common.datastream.DataStreamUtils; +import org.apache.flink.ml.linalg.DenseMatrix; +import org.apache.flink.ml.linalg.DenseVector; +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.streaming.api.environment.StreamExecutionEnvironment; +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 java.io.IOException; +import java.util.ArrayList; +import java.util.HashMap; +import java.util.List; +import java.util.Map; + +/** + * An Estimator which implements the KNN algorithm. + * + * <p>See: https://en.wikipedia.org/wiki/K-nearest_neighbors_algorithm. + */ +public class Knn implements Estimator<Knn, KnnModel>, KnnParams<Knn> { + + protected Map<Param<?>, Object> params = new HashMap<>(); + + public Knn() { + ParamUtils.initializeMapWithDefaultValues(params, this); + } + + @Override + public KnnModel fit(Table... inputs) { + Preconditions.checkArgument(inputs.length == 1); + StreamTableEnvironment tEnv = + (StreamTableEnvironment) ((TableImpl) inputs[0]).getTableEnvironment(); + /* Tuple2 : <sampleVector, label> */ + DataStream<Tuple2<DenseVector, Double>> inputData = + tEnv.toDataStream(inputs[0]) + .map( + new MapFunction<Row, Tuple2<DenseVector, Double>>() { + @Override + public Tuple2<DenseVector, Double> map(Row value) { + Double label = (Double) value.getField(getLabelCol()); + DenseVector feature = + (DenseVector) value.getField(getFeaturesCol()); + return Tuple2.of(feature, label); + } + }); + DataStream<KnnModelData> distributedModelData = prepareModelData(inputData); + DataStream<KnnModelData> modelData = mergeModelData(distributedModelData); + KnnModel model = new KnnModel().setModelData(tEnv.fromDataStream(modelData)); + ReadWriteUtils.updateExistingParams(model, getParamMap()); + return model; + } + + @Override + public Map<Param<?>, Object> getParamMap() { + return this.params; + } + + @Override + public void save(String path) throws IOException { + ReadWriteUtils.saveMetadata(this, path); + } + + public static Knn load(StreamExecutionEnvironment env, String path) throws IOException { + return ReadWriteUtils.loadStageParam(path); + } + + /** + * Prepares distributed knn model data. Constructs the sample matrix and computes norm of + * vectors. + * + * @param inputData Input vector data with label. + * @return Distributed knn model. + */ + private static DataStream<KnnModelData> prepareModelData( + DataStream<Tuple2<DenseVector, Double>> inputData) { + return DataStreamUtils.mapPartition( + inputData, + new RichMapPartitionFunction<Tuple2<DenseVector, Double>, KnnModelData>() { + @Override + public void mapPartition( + Iterable<Tuple2<DenseVector, Double>> values, + Collector<KnnModelData> out) { + List<Tuple2<DenseVector, Double>> dataPoints = new ArrayList<>(0); + for (Tuple2<DenseVector, Double> tuple2 : values) { + dataPoints.add(tuple2); + } + int featureDim = dataPoints.get(0).f0.size(); + DenseMatrix packedFeatures = new DenseMatrix(featureDim, dataPoints.size()); + DenseVector labels = new DenseVector(dataPoints.size()); + for (int i = 0; i < dataPoints.size(); ++i) { + Tuple2<DenseVector, Double> tuple2 = dataPoints.get(i); + labels.values[i] = tuple2.f1; + System.arraycopy( + tuple2.f0.values, + 0, + packedFeatures.values, + i * featureDim, + featureDim); + } + DenseVector featureNorms = computeNorm(packedFeatures); + if (dataPoints.size() > 0) { + out.collect(new KnnModelData(packedFeatures, featureNorms, labels)); + } + } + }, + TypeInformation.of(KnnModelData.class)); Review comment: OK -- 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-unsubscr...@flink.apache.org For queries about this service, please contact Infrastructure at: us...@infra.apache.org
