net.maizegenetics.analysis.imputation

Class BackwardForwardAlgorithm

java.lang.Object

net.maizegenetics.analysis.imputation.BackwardForwardAlgorithm

public class BackwardForwardAlgorithm
extends Object

Author:

pbradbury The BackwardForward algorithm is an HMM that is used to estimate the probability of every possible state at each position in a sequence. This implementation is based on the description in LR Rabiner (1986) A tutorial on hidden Markov models and select applications in speech recognition. Proceedings of the IEEE 77(2):257-286. To use this class, first supply observations, positions, a TransitionProbability object, an EmissionProbability object, and the initial state probabilities. The calculate alpha and beta. Typical usage will be: BackwardForwardAlgorithm myBackwardForward = new BackwardForwardAlgorithm() .observations(obs) .positions(pos) .transition(transition) .emission(emission) .initialStateProbability(probs) .calculateAlpha() .calculateBeta() The either use gamma() to retrieve a List of values or use writeGamma(String outputFile) to save the values to an output file.

Constructor Summary

Constructors

Constructor and Description
BackwardForwardAlgorithm() The BackwardForward algorithm is used to calculate the probability of each state at each position.

Method Summary

Modifier and Type	Method and Description
List<double[]>	alpha()
List<double[]>	beta()
BackwardForwardAlgorithm	calculateAlpha()
BackwardForwardAlgorithm	calculateBeta()
BackwardForwardAlgorithm	emission(EmissionProbability emission)
List<double[]>	gamma()
BackwardForwardAlgorithm	initialStateProbability(double[] probs)
BackwardForwardAlgorithm	observations(int[] observations)
BackwardForwardAlgorithm	positions(int[] positions)
BackwardForwardAlgorithm	transition(TransitionProbability transition)
void	writeGamma(String outputFile)
void	writeGamma(String outputFile, String formatString)

Methods inherited from class java.lang.Object

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

Constructor Detail

BackwardForwardAlgorithm

public BackwardForwardAlgorithm()

The BackwardForward algorithm is used to calculate the probability of each state at each position.

Method Detail

calculateAlpha

public BackwardForwardAlgorithm calculateAlpha()

calculateBeta

public BackwardForwardAlgorithm calculateBeta()

gamma

public List<double[]> gamma()

writeGamma

public void writeGamma(String outputFile,
                       String formatString)

writeGamma

public void writeGamma(String outputFile)

emission

public BackwardForwardAlgorithm emission(EmissionProbability emission)

transition

public BackwardForwardAlgorithm transition(TransitionProbability transition)

observations

public BackwardForwardAlgorithm observations(int[] observations)

positions

public BackwardForwardAlgorithm positions(int[] positions)

initialStateProbability

public BackwardForwardAlgorithm initialStateProbability(double[] probs)

alpha

public List<double[]> alpha()

beta

public List<double[]> beta()