Java example source code file (WeightedRandomWalkGraphIteratorProvider.java)

This example Java source code file (WeightedRandomWalkGraphIteratorProvider.java) is included in the alvinalexander.com "Java Source Code Warehouse" project. The intent of this project is to help you "Learn Java by Example" ^TM.

Learn more about this Java project at its project page.

Java - Java tags/keywords

arraylist, graphwalkiterator, graphwalkiteratorprovider, igraph, list, noedgehandling, number, override, random, util, weightedrandomwalkgraphiteratorprovider

The WeightedRandomWalkGraphIteratorProvider.java Java example source code

package org.deeplearning4j.graph.iterator.parallel;

import org.deeplearning4j.graph.api.IGraph;
import org.deeplearning4j.graph.api.NoEdgeHandling;
import org.deeplearning4j.graph.iterator.GraphWalkIterator;
import org.deeplearning4j.graph.iterator.RandomWalkIterator;
import org.deeplearning4j.graph.iterator.WeightedRandomWalkIterator;

import java.util.ArrayList;
import java.util.List;
import java.util.Random;

/**Weighted random walk graph iterator provider: given a weighted graph (of type {@code IGraph<?,? extends Number>}),
 * split up the generation of weighted random walks for parallel learning. Specifically: with N threads and V vertices:
 * - First iterator generates weighted random walks starting at vertices 0 to V/N
 * - Second iterator generates weighted random walks starting at vertices V/N+1 to 2*V/N
 * - and so on
 * @param <V> Vertex type
 * @see WeightedRandomWalkIterator
 */
public class WeightedRandomWalkGraphIteratorProvider<V> implements GraphWalkIteratorProvider {

    private IGraph<V,? extends Number> graph;
    private int walkLength;
    private Random rng;
    private NoEdgeHandling mode;

    public WeightedRandomWalkGraphIteratorProvider(IGraph<V, ? extends Number> graph, int walkLength){
        this(graph, walkLength, System.currentTimeMillis(), NoEdgeHandling.EXCEPTION_ON_DISCONNECTED);
    }

    public WeightedRandomWalkGraphIteratorProvider(IGraph<V, ? extends Number> graph, int walkLength, long seed,
                                                   NoEdgeHandling mode){
        this.graph = graph;
        this.walkLength = walkLength;
        this.rng = new Random(seed);
        this.mode = mode;
    }


    @Override
    public List<GraphWalkIterator getGraphWalkIterators(int numIterators) {
        int nVertices = graph.numVertices();
        if(numIterators > nVertices) numIterators = nVertices;

        int verticesPerIter = nVertices / numIterators;

        List<GraphWalkIterator list = new ArrayList<>(numIterators);
        int last = 0;
        for( int i=0; i<numIterators; i++ ){
            int from = last;
            int to = Math.min(nVertices,from+verticesPerIter);
            if(i == numIterators - 1) to = nVertices;

            GraphWalkIterator<V> iter = new WeightedRandomWalkIterator(graph,walkLength,rng.nextLong(),mode,from,to);
            list.add(iter);
            last = to;
        }

        return list;
    }
}