com.graphhopper.routing.lm

Class LMApproximator

java.lang.Object

com.graphhopper.routing.lm.LMApproximator

All Implemented Interfaces:

WeightApproximator

public class LMApproximator
extends Object
implements WeightApproximator

This class is a weight approximation based on precalculated landmarks.

Author:

Peter Karich

Constructor Summary

Constructors

Constructor and Description
LMApproximator(Graph graph, Weighting lmWeighting, Weighting routingWeighting, int maxBaseNodes, LandmarkStorage lms, int activeCount, double factor, boolean reverse)

Method Summary

Modifier and Type	Method and Description
double	approximate(int v)
static LMApproximator	forLandmarks(Graph g, Weighting weighting, LandmarkStorage lms, int activeLM)
double	getSlack()
WeightApproximator	reverse() Makes a 'reverse' copy of itself to make it possible using the two objects independent e.g.
LMApproximator	setEpsilon(double epsilon) Increase approximation with higher epsilon
void	setTo(int t)
String	toString()
void	triggerActiveLandmarkRecalculation() This method forces a lazy recalculation of the active landmark set e.g.

Methods inherited from class java.lang.Object

clone, equals, finalize, getClass, hashCode, notify, notifyAll, wait, wait, wait

Constructor Detail

LMApproximator

public LMApproximator(Graph graph,
                      Weighting lmWeighting,
                      Weighting routingWeighting,
                      int maxBaseNodes,
                      LandmarkStorage lms,
                      int activeCount,
                      double factor,
                      boolean reverse)

Method Detail

forLandmarks

public static LMApproximator forLandmarks(Graph g,
                                          Weighting weighting,
                                          LandmarkStorage lms,
                                          int activeLM)

Parameters:

weighting - the weighting used for the current path calculation, not necessarily the same that we used for the LM preparation. All edge weights must be larger or equal compared to those used for the preparation.

setEpsilon

public LMApproximator setEpsilon(double epsilon)

Increase approximation with higher epsilon

approximate

public double approximate(int v)

Specified by:

approximate in interface WeightApproximator

Returns:

minimal weight of the specified currentNode to the 'to' node

setTo

public void setTo(int t)

Specified by:

setTo in interface WeightApproximator

reverse

public WeightApproximator reverse()

Description copied from interface: WeightApproximator

Makes a 'reverse' copy of itself to make it possible using the two objects independent e.g. on different threads. Do not copy state depending on the current approximate calls. 'reverse' means the WeightApproximator should handle approximate calls towards the 'from' instead towards the 'to'.

Specified by:

reverse in interface WeightApproximator

getSlack

public double getSlack()

Specified by:

getSlack in interface WeightApproximator

triggerActiveLandmarkRecalculation

public void triggerActiveLandmarkRecalculation()

This method forces a lazy recalculation of the active landmark set e.g. necessary after the 'to' node changed.

toString

public String toString()

Overrides:

toString in class Object