cc.mallet.fst.semi_supervised.constraints
Class TwoLabelGEConstraints

java.lang.Object
  cc.mallet.fst.semi_supervised.constraints.TwoLabelGEConstraints

All Implemented Interfaces:

GEConstraint

Direct Known Subclasses:

TwoLabelKLGEConstraints, TwoLabelL2GEConstraints

public abstract class TwoLabelGEConstraints
extends java.lang.Object
implements GEConstraint

A set of constraints on distributions over pairs of consecutive labels conditioned on the presence of input features. Subclasses are to be used with GE. Multiple constraints are grouped together here to make things more efficient.

Author:

Gregory Druck

Nested Class Summary

protected class

TwoLabelGEConstraints.TwoLabelGEConstraint

Field Summary

protected gnu.trove.TIntArrayList	cache
protected java.util.ArrayList<TwoLabelGEConstraints.TwoLabelGEConstraint>	constraintsList
protected gnu.trove.TIntIntHashMap	constraintsMap
protected StateLabelMap	map

Constructor Summary

	TwoLabelGEConstraints()
protected	TwoLabelGEConstraints(java.util.ArrayList<TwoLabelGEConstraints.TwoLabelGEConstraint> constraintsList, gnu.trove.TIntIntHashMap constraintsMap, StateLabelMap map)

Method Summary

abstract void	addConstraint(int fi, double[][] target, double weight)
void	computeExpectations(java.util.ArrayList<SumLattice> lattices) Compute expectations using cached lattices.
double	getCompositeConstraintFeatureValue(FeatureVector fv, int ip, int si1, int si2) Computes the composite constraint feature value (over all constraint features) for FeatureVector fv and a transition from state li1 to li2.
abstract double	getValue() Returns the total constraint value.
boolean	isOneStateConstraint()
void	preProcess(FeatureVector fv) Gives the constraint the option to do some caching using only the FeatureVector.
java.util.BitSet	preProcess(InstanceList data)
void	setStateLabelMap(StateLabelMap map) Sets that map between the state indices and label indices.
void	zeroExpectations() Zero expectation values.

Methods inherited from class java.lang.Object

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

Methods inherited from interface cc.mallet.fst.semi_supervised.constraints.GEConstraint

copy

Field Detail

constraintsList

protected java.util.ArrayList<TwoLabelGEConstraints.TwoLabelGEConstraint> constraintsList

constraintsMap

protected gnu.trove.TIntIntHashMap constraintsMap

map

protected StateLabelMap map

cache

protected gnu.trove.TIntArrayList cache

Constructor Detail

TwoLabelGEConstraints

public TwoLabelGEConstraints()

TwoLabelGEConstraints

protected TwoLabelGEConstraints(java.util.ArrayList<TwoLabelGEConstraints.TwoLabelGEConstraint> constraintsList,
                                gnu.trove.TIntIntHashMap constraintsMap,
                                StateLabelMap map)

Method Detail

addConstraint

public abstract void addConstraint(int fi,
                                   double[][] target,
                                   double weight)

Parameters:

fi - Input feature index

target - Target distribution over pairs of labels

weight - Weight of this constraint

isOneStateConstraint

public boolean isOneStateConstraint()

Specified by:

isOneStateConstraint in interface GEConstraint

Returns:

true if constraint feature only considers one state

setStateLabelMap

public void setStateLabelMap(StateLabelMap map)

Description copied from interface: GEConstraint

Sets that map between the state indices and label indices.

Specified by:

setStateLabelMap in interface GEConstraint

Parameters:

map - StateLabelMap

preProcess

public void preProcess(FeatureVector fv)

Description copied from interface: GEConstraint

Gives the constraint the option to do some caching using only the FeatureVector. For example, the constrained input features could be cached.

Specified by:

preProcess in interface GEConstraint

Parameters:

fv - FeatureVector input

preProcess

public java.util.BitSet preProcess(InstanceList data)

Specified by:

preProcess in interface GEConstraint

Parameters:

data - Unlabeled data

Returns:

Returns a bitset of the size of the data, with the bit set if a constraint feature fires in that instance.

getCompositeConstraintFeatureValue

public double getCompositeConstraintFeatureValue(FeatureVector fv,
                                                 int ip,
                                                 int si1,
                                                 int si2)

Description copied from interface: GEConstraint

Computes the composite constraint feature value (over all constraint features) for FeatureVector fv and a transition from state li1 to li2.

Specified by:

getCompositeConstraintFeatureValue in interface GEConstraint

Parameters:

fv - FeatureVector on transition

ip - Position of input in sequence

si1 - Source state index for transition

si2 - Destination state index for transition

Returns:

Constraint feature value

getValue

public abstract double getValue()

Description copied from interface: GEConstraint

Returns the total constraint value.

Specified by:

getValue in interface GEConstraint

Returns:

Constraint value

zeroExpectations

public void zeroExpectations()

Description copied from interface: GEConstraint

Zero expectation values. Called before re-computing gradient.

Specified by:

zeroExpectations in interface GEConstraint

computeExpectations

public void computeExpectations(java.util.ArrayList<SumLattice> lattices)

Description copied from interface: GEConstraint

Compute expectations using cached lattices.

Specified by:

computeExpectations in interface GEConstraint

Parameters:

lattices - Cached SumLattices