CRFCliqueTree (Stanford JavaNLP API)

java.lang.Object
- edu.stanford.nlp.ie.crf.CRFCliqueTree<E>

Type Parameters:

E - The type of the label (usually String in our uses)

All Implemented Interfaces:

ListeningSequenceModel, SequenceListener, SequenceModel
```
public class CRFCliqueTree<E>
extends java.lang.Object
implements ListeningSequenceModel
```
Builds a CliqueTree (an array of FactorTable) and does message passing inference along it.

Author:

Jenny Finkel

Constructor Summary

Constructors
Constructor and Description

CRFCliqueTree(FactorTable[] factorTables, Index<E> classIndex, E backgroundSymbol)
Initialize a clique tree.

Constructors
Constructor and Description
`CRFCliqueTree(FactorTable[] factorTables, Index<E> classIndex, E backgroundSymbol)` Initialize a clique tree.

Method Summary

All Methods Static Methods Instance Methods Concrete Methods
Modifier and Type	Method and Description
`int`	`backgroundIndex()`
`E`	`backgroundSymbol()`
`Index<E>`	`classIndex()`
`double`	`condLogProbGivenNext(int position, E label, E[] nextLabels)`
`double`	`condLogProbGivenNext(int position, int label, int[] nextLabels)`
`double`	`condLogProbGivenPrevious(int position, E label, E[] prevLabels)`
`double`	`condLogProbGivenPrevious(int position, int label, int[] prevLabels)` Gives the probability of a tag at a single position conditioned on a sequence of previous labels.
`Counter<E>`	`condLogProbsGivenNext(int position, E[] nextlabels)`
`Counter<E>`	`condLogProbsGivenNext(int position, int[] nextlabels)`
`Counter<E>`	`condLogProbsGivenPrevious(int position, E[] prevlabels)`
`Counter<E>`	`condLogProbsGivenPrevious(int position, int[] prevlabels)`
`double`	`condProbGivenNext(int position, E label, E[] nextLabels)`
`double`	`condProbGivenNext(int position, int label, int[] nextLabels)`
`double`	`condProbGivenPrevious(int position, E label, E[] prevLabels)`
`double`	`condProbGivenPrevious(int position, int label, int[] prevLabels)`
`static <E> CRFCliqueTree<E>`	`getCalibratedCliqueTree(double[] weights, double wscale, int[][] weightIndices, int[][][] data, java.util.List<Index<CRFLabel>> labelIndices, int numClasses, Index<E> classIndex, E backgroundSymbol)` This function assumes a LinearCliquePotentialFunction is used for wrapping the weights
`static <E> CRFCliqueTree<E>`	`getCalibratedCliqueTree(int[][][] data, java.util.List<Index<CRFLabel>> labelIndices, int numClasses, Index<E> classIndex, E backgroundSymbol, CliquePotentialFunction cliquePotentialFunc, double[][][] featureVals)`
`double[]`	`getConditionalDistribution(int[] sequence, int position)` Computes the distribution over values of the element at position pos in the sequence, conditioned on the values of the elements in all other positions of the provided sequence.
`FactorTable[]`	`getFactorTables()`
`int`	`getNumClasses()`
`int`	`getNumValues()`
`int[]`	`getPossibleValues(int position)` Return the valid sequence labels (as integer indices) for a particular position in the sequence.
`int`	`leftWindow()` How many label positions to the left influence the label assignment at a particular position.
`int`	`length()`
`double`	`logProb(int position, E label)`
`double`	`logProb(int position, E[] labels)` returns the log probability for the given labels, where the last label corresponds to the label at the specified position.
`double`	`logProb(int position, int label)`
`double`	`logProb(int position, int[] labels)` returns the log probability for the given labels (indexed using classIndex), where the last label corresponds to the label at the specified position.
`Counter<E>`	`logProbs(int position)`
`GeneralizedCounter<E>`	`logProbs(int position, int window)`
`double`	`logProbStartPos()`
`double[]`	`logProbsToDoubleArr(int position)`
`double[][]`	`logProbTable()`
`double`	`prob(int position, E label)`
`double`	`prob(int position, E[] labels)` returns the probability for the given labels, where the last label corresponds to the label at the specified position.
`double`	`prob(int position, int label)`
`double`	`prob(int position, int[] labels)` Returns the probability for the given labels (indexed using classIndex), where the last label corresponds to the label at the specified position.
`Counter<E>`	`probs(int position)`
`GeneralizedCounter<E>`	`probs(int position, int window)`
`double[]`	`probsToDoubleArr(int position)`
`int`	`rightWindow()` How many label positions to the right influence the label assignment at a particular position.
`double`	`scoreOf(int[] sequence)` Returns the log probability of this sequence given the CRF.
`double`	`scoreOf(int[] sequence, int pos)` Computes the score of the element at the given position in the sequence, conditioned on the values of the elements in all other positions of the provided sequence.
`double[]`	`scoresOf(int[] sequence, int position)` Computes the unnormalized log conditional distribution over values of the element at position pos in the sequence, conditioned on the values of the elements in all other positions of the provided sequence.
`void`	`setInitialSequence(int[] sequence)` Informs this sequence model that the value of the whole sequence is initialized to sequence
`double`	`totalMass()`
`void`	`updateSequenceElement(int[] sequence, int pos, int oldVal)` Informs this sequence model that the value of the element at position pos has changed.
`int`	`window()`

Methods inherited from class java.lang.Object
clone, equals, finalize, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait

- Constructor Detail
  - CRFCliqueTree
```
public CRFCliqueTree(FactorTable[] factorTables,
                     Index<E> classIndex,
                     E backgroundSymbol)
```
    Initialize a clique tree.
- Method Detail
  - getFactorTables
```
public FactorTable[] getFactorTables()
```
  - classIndex
```
public Index<E> classIndex()
```
  - length
```
public int length()
```
    Specified by:
    
    length in interface SequenceModel
    
    Returns:
    
    The length of the sequences modeled by this SequenceModel. This is its real length, not including the leftWindow() + rightWindow() of extra padding positions.
  - leftWindow
```
public int leftWindow()
```
    Description copied from interface: SequenceModel
    
    How many label positions to the left influence the label assignment at a particular position.
    
    Specified by:
    
    leftWindow in interface SequenceModel
    
    Returns:
    
    The size of the left window used by this sequence model
  - rightWindow
```
public int rightWindow()
```
    Description copied from interface: SequenceModel
    
    How many label positions to the right influence the label assignment at a particular position.
    
    Specified by:
    
    rightWindow in interface SequenceModel
    
    Returns:
    
    the size of the right window used by this sequence model
  - getPossibleValues
```
public int[] getPossibleValues(int position)
```
    Description copied from interface: SequenceModel
    
    Return the valid sequence labels (as integer indices) for a particular position in the sequence. Since the sequence is padded at each end, typically sequence items 0...leftWindow-1 are null, leftWindow...length+leftWindow-1 are words, length+leftWindow...length+leftWindow+rightWindow-1 are null.
    
    Specified by:
    
    getPossibleValues in interface SequenceModel
    
    Parameters:
    
    position - The position
    
    Returns:
    
    The set of possible int values at this position, as an int array
  - scoreOf
```
public double scoreOf(int[] sequence,
                      int pos)
```
    Description copied from interface: SequenceModel
    
    Computes the score of the element at the given position in the sequence, conditioned on the values of the elements in all other positions of the provided sequence. Typically, this is an unnormalized log conditional probability of the label at the given position in the sequence, given the input data nd the other labels.
    
    Specified by:
    
    scoreOf in interface SequenceModel
    
    Parameters:
    
    sequence - The sequence containing the prediction and the rest of the labels to condition on
    
    pos - The position of the element to give a score for
    
    Returns:
    
    The score of the label at the specified position in the sequence
  - scoresOf
```
public double[] scoresOf(int[] sequence,
                         int position)
```
    Computes the unnormalized log conditional distribution over values of the element at position pos in the sequence, conditioned on the values of the elements in all other positions of the provided sequence.
    
    Specified by:
    
    scoresOf in interface SequenceModel
    
    Parameters:
    
    sequence - the sequence containing the rest of the values to condition on
    
    position - the position of the element to give a distribution for
    
    Returns:
    
    an array of type double, representing a probability distribution; sums to 1.0
  - scoreOf
```
public double scoreOf(int[] sequence)
```
    Returns the log probability of this sequence given the CRF. Does so by computing the marginal of the first windowSize tags, and then computing the conditional probability for the rest of them, conditioned on the previous tags.
    
    Specified by:
    
    scoreOf in interface SequenceModel
    
    Parameters:
    
    sequence - The sequence to compute a score for
    
    Returns:
    
    the score for the sequence
  - window
```
public int window()
```
  - getNumClasses
```
public int getNumClasses()
```
  - totalMass
```
public double totalMass()
```
  - backgroundIndex
```
public int backgroundIndex()
```
  - backgroundSymbol
```
public E backgroundSymbol()
```
  - logProbTable
```
public double[][] logProbTable()
```
  - logProbStartPos
```
public double logProbStartPos()
```
  - logProb
```
public double logProb(int position,
                      int label)
```
  - prob
```
public double prob(int position,
                   int label)
```
  - logProb
```
public double logProb(int position,
                      E label)
```
  - prob
```
public double prob(int position,
                   E label)
```
  - probsToDoubleArr
```
public double[] probsToDoubleArr(int position)
```
  - logProbsToDoubleArr
```
public double[] logProbsToDoubleArr(int position)
```
  - probs
```
public Counter<E> probs(int position)
```
  - logProbs
```
public Counter<E> logProbs(int position)
```
  - logProb
```
public double logProb(int position,
                      int[] labels)
```
    returns the log probability for the given labels (indexed using classIndex), where the last label corresponds to the label at the specified position. For instance if you called logProb(5, {1,2,3}) it will return the marginal log prob that the label at position 3 is 1, the label at position 4 is 2 and the label at position 5 is 3.
  - prob
```
public double prob(int position,
                   int[] labels)
```
    Returns the probability for the given labels (indexed using classIndex), where the last label corresponds to the label at the specified position. For instance if you called prob(5, {1,2,3}) it will return the marginal prob that the label at position 3 is 1, the label at position 4 is 2 and the label at position 5 is 3.
  - logProb
```
public double logProb(int position,
                      E[] labels)
```
    returns the log probability for the given labels, where the last label corresponds to the label at the specified position. For instance if you called logProb(5, {"O", "PER", "ORG"}) it will return the marginal log prob that the label at position 3 is "O", the label at position 4 is "PER" and the label at position 5 is "ORG".
  - prob
```
public double prob(int position,
                   E[] labels)
```
    returns the probability for the given labels, where the last label corresponds to the label at the specified position. For instance if you called logProb(5, {"O", "PER", "ORG"}) it will return the marginal prob that the label at position 3 is "O", the label at position 4 is "PER" and the label at position 5 is "ORG".
  - logProbs
```
public GeneralizedCounter<E> logProbs(int position,
                                      int window)
```
  - probs
```
public GeneralizedCounter<E> probs(int position,
                                   int window)
```
  - condLogProbGivenPrevious
```
public double condLogProbGivenPrevious(int position,
                                       int label,
                                       int[] prevLabels)
```
    Gives the probability of a tag at a single position conditioned on a sequence of previous labels.
    
    Parameters:
    
    position - Index in sequence
    
    label - Label of item at index
    
    prevLabels - Indices of labels in previous positions
    
    Returns:
    
    conditional log probability
  - condLogProbGivenPrevious
```
public double condLogProbGivenPrevious(int position,
                                       E label,
                                       E[] prevLabels)
```
  - condProbGivenPrevious
```
public double condProbGivenPrevious(int position,
                                    int label,
                                    int[] prevLabels)
```
  - condProbGivenPrevious
```
public double condProbGivenPrevious(int position,
                                    E label,
                                    E[] prevLabels)
```
  - condLogProbsGivenPrevious
```
public Counter<E> condLogProbsGivenPrevious(int position,
                                            int[] prevlabels)
```
  - condLogProbsGivenPrevious
```
public Counter<E> condLogProbsGivenPrevious(int position,
                                            E[] prevlabels)
```
  - condLogProbGivenNext
```
public double condLogProbGivenNext(int position,
                                   int label,
                                   int[] nextLabels)
```
  - condLogProbGivenNext
```
public double condLogProbGivenNext(int position,
                                   E label,
                                   E[] nextLabels)
```
  - condProbGivenNext
```
public double condProbGivenNext(int position,
                                int label,
                                int[] nextLabels)
```
  - condProbGivenNext
```
public double condProbGivenNext(int position,
                                E label,
                                E[] nextLabels)
```
  - condLogProbsGivenNext
```
public Counter<E> condLogProbsGivenNext(int position,
                                        int[] nextlabels)
```
  - condLogProbsGivenNext
```
public Counter<E> condLogProbsGivenNext(int position,
                                        E[] nextlabels)
```
  - getCalibratedCliqueTree
```
public static <E> CRFCliqueTree<E> getCalibratedCliqueTree(int[][][] data,
                                                           java.util.List<Index<CRFLabel>> labelIndices,
                                                           int numClasses,
                                                           Index<E> classIndex,
                                                           E backgroundSymbol,
                                                           CliquePotentialFunction cliquePotentialFunc,
                                                           double[][][] featureVals)
```
    Returns:
    
    a new CRFCliqueTree for the weights on the data
  - getCalibratedCliqueTree
```
public static <E> CRFCliqueTree<E> getCalibratedCliqueTree(double[] weights,
                                                           double wscale,
                                                           int[][] weightIndices,
                                                           int[][][] data,
                                                           java.util.List<Index<CRFLabel>> labelIndices,
                                                           int numClasses,
                                                           Index<E> classIndex,
                                                           E backgroundSymbol)
```
    This function assumes a LinearCliquePotentialFunction is used for wrapping the weights
    
    Returns:
    
    a new CRFCliqueTree for the weights on the data
  - getConditionalDistribution
```
public double[] getConditionalDistribution(int[] sequence,
                                           int position)
```
    Computes the distribution over values of the element at position pos in the sequence, conditioned on the values of the elements in all other positions of the provided sequence.
    
    Parameters:
    
    sequence - the sequence containing the rest of the values to condition on
    
    position - the position of the element to give a distribution for
    
    Returns:
    
    an array of type double, representing a probability distribution; sums to 1.0
  - updateSequenceElement
```
public void updateSequenceElement(int[] sequence,
                                  int pos,
                                  int oldVal)
```
    Informs this sequence model that the value of the element at position pos has changed. This allows this sequence model to update its internal model if desired.
    
    Specified by:
    
    updateSequenceElement in interface SequenceListener
  - setInitialSequence
```
public void setInitialSequence(int[] sequence)
```
    Informs this sequence model that the value of the whole sequence is initialized to sequence
    
    Specified by:
    
    setInitialSequence in interface SequenceListener
  - getNumValues
```
public int getNumValues()
```
    Returns:
    
    the number of possible values for each element; it is assumed to be the same for the element at each position

Class CRFCliqueTree<E>

Constructor Summary

Method Summary

Methods inherited from class java.lang.Object

Constructor Detail

CRFCliqueTree

Method Detail

getFactorTables

classIndex

length

leftWindow

rightWindow

getPossibleValues

scoreOf

scoresOf

scoreOf

window

getNumClasses

totalMass

backgroundIndex

backgroundSymbol

logProbTable

logProbStartPos

logProb

prob

logProb

prob

probsToDoubleArr

logProbsToDoubleArr

probs

logProbs

logProb

prob

logProb

prob

logProbs

probs

condLogProbGivenPrevious

condLogProbGivenPrevious

condProbGivenPrevious

condProbGivenPrevious

condLogProbsGivenPrevious

condLogProbsGivenPrevious

condLogProbGivenNext

condLogProbGivenNext

condProbGivenNext

condProbGivenNext

condLogProbsGivenNext

condLogProbsGivenNext

getCalibratedCliqueTree

getCalibratedCliqueTree

getConditionalDistribution

updateSequenceElement

setInitialSequence

getNumValues