Segmentor3IsBack (version 2.0)

BestSegmentation: BestSegmentation

Description

This function is used to compute the cost of the best segmentation in K segments given the position of a change-point, and to return the optimal segmenation having a change-point at location t.

Usage

BestSegmentation(x,K,t=numeric(),compress=TRUE)

Arguments

x

An object of class Segmentor returned by function Segmentor

K

The number of segments of the segmentation for which the cost or best segmentation is wanted

t

The position for which the best segmentation with t as change-point is wanted

compress

A boolean stating whether data should be compressed prior to segmentation

Value

bestCost

A matrix of size n*K: the cost of the optimal segmentation with jth change-point i

bestSeg

If a t has been specified, a vector of size K+1 containing values of indicating the optimal segmentation with t as a change-point

References

PDPA: Rigaill, G. Pruned dynamic programming for optimal multiple change-point detection: Submitted http://arxiv.org/abs/1004.0887

PDPA: Cleynen, A. and Koskas, M. and Lebarbier, E. and Rigaill, G. and Robin, S. Segmentor3IsBack (2014): an R package for the fast and exact segmentation of Seq-data Algorithms for Molecular Biology

Examples

require(Segmentor3IsBack);
N=2000 
x=rnbinom(3*N,size=1.3,prob=rep(c(0.7,0.2,0.01),each=N));
res=Segmentor(data=x,model=3,Kmax=10, keep=TRUE);  
# Finds the optimal segmentation in up to 10 segments with respect to 
#the negative binomial model.
K<-3
Best<-BestSegmentation(res,K=3,t=3000,compress=FALSE)
matplot(Best$bestCost, type='l', lty=2)
points(apply(Best$bestCost,2,which.min),apply(Best$bestCost,2,min),pch=20,col=1:(K-1))
apply(Best$bestCost, 2,which.min)
getBreaks(res)[K,1:(K-1)]
#computes and plots cost of best segmentation in 3 segments with 
#change-point t, and compares result with change-point estimates.
Best$bestSeg
#returns the optimal segmentation in 3 segments with 3000 as a
#change-point
Best<-BestSegmentation(res,K=3,t=3000,compress=TRUE)
Best$bestSeg
# this segmentation usually does not make sense because of the
# compress option. t has to be adapted consequently