/**
* Copyright (c) 2008, The JUNG Authors
*
* All rights reserved.
*
* This software is open-source under the BSD license; see either
* "license.txt" or
* https://github.com/jrtom/jung/blob/master/LICENSE for a description.
* Created on Aug 22, 2008
*
*/
package edu.uci.ics.jung.algorithms.scoring;
import com.google.common.base.Function;
import edu.uci.ics.jung.algorithms.scoring.util.ScoringUtils;
import edu.uci.ics.jung.graph.Hypergraph;
/**
* A special case of {@code PageRankWithPriors} in which the final scores
* represent a probability distribution over position assuming a random (Markovian)
* walk of exactly k steps, based on the initial distribution specified by the priors.
*
* NOTE: The version of {@code KStepMarkov} in {@code algorithms.importance}
* (and in JUNG 1.x) is believed to be incorrect: rather than returning
* a score which represents a probability distribution over position assuming
* a k-step random walk, it returns a score which represents the sum over all steps
* of the probability for each step. If you want that behavior, set the
* 'cumulative' flag as follows before calling {@code evaluate()}:
*
* KStepMarkov ksm = new KStepMarkov(...);
* ksm.setCumulative(true);
* ksm.evaluate();
*
*
* By default, the 'cumulative' flag is set to false.
*
* NOTE: THIS CLASS IS NOT YET COMPLETE. USE AT YOUR OWN RISK. (The original behavior
* is captured by the version still available in {@code algorithms.importance}.)
*
* @see "Algorithms for Estimating Relative Importance in Graphs by Scott White and Padhraic Smyth, 2003"
* @see PageRank
* @see PageRankWithPriors
*/
public class KStepMarkov extends PageRankWithPriors
{
private boolean cumulative;
/**
* Creates an instance based on the specified graph, edge weights, vertex
* priors (initial scores), and number of steps to take.
* @param graph the input graph
* @param edge_weights the edge weights (transition probabilities)
* @param vertex_priors the initial probability distribution (score assignment)
* @param steps the number of times that {@code step()} will be called by {@code evaluate}
*/
public KStepMarkov(Hypergraph graph, Function edge_weights,
Function vertex_priors, int steps)
{
super(graph, edge_weights, vertex_priors, 0);
initialize(steps);
}
/**
* Creates an instance based on the specified graph, vertex
* priors (initial scores), and number of steps to take. The edge
* weights (transition probabilities) are set to default values (a uniform
* distribution over all outgoing edges).
* @param graph the input graph
* @param vertex_priors the initial probability distribution (score assignment)
* @param steps the number of times that {@code step()} will be called by {@code evaluate}
*/
public KStepMarkov(Hypergraph graph, Function vertex_priors, int steps)
{
super(graph, vertex_priors, 0);
initialize(steps);
}
/**
* Creates an instance based on the specified graph and number of steps to
* take. The edge weights (transition probabilities) and vertex initial scores
* (prior probabilities) are set to default values (a uniform
* distribution over all outgoing edges, and a uniform distribution over
* all vertices, respectively).
* @param graph the input graph
* @param steps the number of times that {@code step()} will be called by {@code evaluate}
*/
public KStepMarkov(Hypergraph graph, int steps)
{
super(graph, ScoringUtils.getUniformRootPrior(graph.getVertices()), 0);
initialize(steps);
}
private void initialize(int steps)
{
this.acceptDisconnectedGraph(false);
if (steps <= 0)
throw new IllegalArgumentException("Number of steps must be > 0");
this.max_iterations = steps;
this.tolerance = -1.0;
this.cumulative = false;
}
/**
* Specifies whether this instance should assign a score to each vertex
* based on the sum over all steps of the probability for each step.
* See the class-level documentation for details.
* @param cumulative true if this instance should assign a cumulative score to each vertex
*/
public void setCumulative(boolean cumulative)
{
this.cumulative = cumulative;
}
/**
* Updates the value for this vertex. Called by step().
*/
@Override
public double update(V v)
{
if (!cumulative)
return super.update(v);
collectDisappearingPotential(v);
double v_input = 0;
for (E e : graph.getInEdges(v))
{
// For graphs, the code below is equivalent to
// V w = graph.getOpposite(v, e);
// total_input += (getCurrentValue(w) * getEdgeWeight(w,e).doubleValue());
// For hypergraphs, this divides the potential coming from w
// by the number of vertices in the connecting edge e.
int incident_count = getAdjustedIncidentCount(e);
for (V w : graph.getIncidentVertices(e))
{
if (!w.equals(v) || hyperedges_are_self_loops)
v_input += (getCurrentValue(w) *
getEdgeWeight(w,e).doubleValue() / incident_count);
}
}
// modify total_input according to alpha
double new_value = alpha > 0 ?
v_input * (1 - alpha) + getVertexPrior(v) * alpha :
v_input;
setOutputValue(v, new_value + getCurrentValue(v));
// FIXME: DO WE NEED TO CHANGE HOW DISAPPEARING IS COUNTED? NORMALIZE?
return Math.abs(getCurrentValue(v) - new_value);
}
}
Ads help maintain this website.