All Downloads are FREE. Search and download functionalities are using the official Maven repository.

org.neo4j.gds.allshortestpaths.WeightedAllShortestPaths Maven / Gradle / Ivy

There is a newer version: 2.15.0
Show newest version
/*
 * Copyright (c) "Neo4j"
 * Neo4j Sweden AB [http://neo4j.com]
 *
 * This file is part of Neo4j.
 *
 * Neo4j is free software: you can redistribute it and/or modify
 * it under the terms of the GNU General Public License as published by
 * the Free Software Foundation, either version 3 of the License, or
 * (at your option) any later version.
 *
 * This program is distributed in the hope that it will be useful,
 * but WITHOUT ANY WARRANTY; without even the implied warranty of
 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
 * GNU General Public License for more details.
 *
 * You should have received a copy of the GNU General Public License
 * along with this program.  If not, see .
 */
package org.neo4j.gds.allshortestpaths;

import org.neo4j.gds.api.Graph;
import org.neo4j.gds.api.properties.relationships.RelationshipIterator;
import org.neo4j.gds.core.concurrency.Concurrency;
import org.neo4j.gds.core.utils.progress.tasks.ProgressTracker;
import org.neo4j.gds.termination.TerminationFlag;

import java.util.Arrays;
import java.util.concurrent.BlockingQueue;
import java.util.concurrent.ExecutorService;
import java.util.concurrent.LinkedBlockingQueue;
import java.util.concurrent.atomic.AtomicInteger;
import java.util.stream.Stream;

import static org.neo4j.gds.Converters.longToIntConsumer;

/**
 * WeightedAllShortestPaths:
 * 

* multi-source parallel dijkstra algorithm for computing the shortest path between * each pair of nodes. *

* Since all nodeId's have already been ordered by the idMap we can use an integer * instead of a queue which just count's up for each startNodeId as long as it is * {@code < nodeCount}. Each thread tries to take one int from the counter at one time and * starts its computation on it. *

* The {@link WeightedAllShortestPaths#concurrency} value determines the count of workers * that should be spawned. *

* Due to the high memory footprint the result set would have we emit each result into * a blocking queue. The result stream takes elements from the queue while the workers * add elements to it. The result stream is limited by N^2. If the stream gets closed * prematurely the workers get closed too. */ public class WeightedAllShortestPaths extends MSBFSASPAlgorithm { private final BlockingQueue resultQueue = new LinkedBlockingQueue<>(); private final int nodeCount; private final Concurrency concurrency; // maximum number of workers private final ExecutorService executorService; private final Graph graph; private final AtomicInteger counter; // nodeId counter (init with nodeCount, counts down for each node) private volatile boolean outputStreamOpen; public WeightedAllShortestPaths(Graph graph, ExecutorService executorService, Concurrency concurrency, TerminationFlag terminationFlag) { super(ProgressTracker.NULL_TRACKER); if (!graph.hasRelationshipProperty()) { throw new UnsupportedOperationException("WeightedAllShortestPaths is not supported on graphs without a weight property"); } this.graph = graph; this.nodeCount = Math.toIntExact(graph.nodeCount()); this.executorService = executorService; this.concurrency = concurrency; this.counter = new AtomicInteger(); this.terminationFlag = terminationFlag; } /** * the compute(..) method starts the computation and * returns a Stream of SP-Tuples (source, target, minDist) * * @return the result stream */ @Override public Stream compute() { progressTracker.beginSubTask(); counter.set(0); outputStreamOpen = true; for (int i = 0; i < concurrency.value(); i++) { executorService.submit(new ShortestPathTask()); } return AllShortestPathsStream.stream(resultQueue, () -> { outputStreamOpen = false; progressTracker.endSubTask(); }) .limit((long) nodeCount * nodeCount) .filter(result -> result.distance != Double.POSITIVE_INFINITY); } /** * Dijkstra Task. Takes one element of the counter at a time * and starts dijkstra on it. */ private final class ShortestPathTask implements Runnable { private final IntPriorityQueue queue; private final double[] distance; private final RelationshipIterator threadLocalGraph; private ShortestPathTask() { distance = new double[nodeCount]; queue = IntPriorityQueue.min(); this.threadLocalGraph = graph.concurrentCopy(); } @Override public void run() { int startNode; while (outputStreamOpen && terminationFlag.running() && (startNode = counter.getAndIncrement()) < nodeCount) { compute(startNode); for (int i = 0; i < nodeCount; i++) { var result = AllShortestPathsStreamResult.result( graph.toOriginalNodeId(startNode), graph.toOriginalNodeId(i), distance[i] ); try { resultQueue.put(result); } catch (InterruptedException e) { Thread.currentThread().interrupt(); throw new RuntimeException(e); } } progressTracker.logProgress(); } } void compute(int startNode) { Arrays.fill(distance, Double.POSITIVE_INFINITY); distance[startNode] = 0D; queue.add(startNode, 0D); while (outputStreamOpen && !queue.isEmpty()) { final int node = queue.pop(); final double sourceDistance = distance[node]; threadLocalGraph.forEachRelationship( node, Double.NaN, longToIntConsumer((source, target, weight) -> { // relax final double targetDistance = weight + sourceDistance; if (targetDistance < distance[target]) { distance[target] = targetDistance; queue.set(target, targetDistance); } return true; })); } } } }





© 2015 - 2025 Weber Informatics LLC | Privacy Policy