Description Usage Arguments Value Examples
View source: R/hybrid_simultanee.R
It is an algorithm which combines dynamic programming and the EM algorithm to calculate the MLE of phi and T, which are the mixture parameters and the change point instants. this algorithm is run for a given number of clusters, and estimates the parameters for a segmentation/clustering model with P clusters and 1:Kmax segments
1 | hybrid_simultanee(x, P, Kmax, lmin, sameSigma = TRUE)
|
x |
the two-dimensionnal signal, one line per dimension |
P |
the number of classes |
Kmax |
the maximal number of segments |
a list with tau posterior probability
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 | K <- 5; rupt <- sample(1:20, K+1, replace=TRUE); rupt <- cumsum(rupt);
n <- max(rupt)
muSim <- matrix(rnorm(2*K+2, mean=20, sd=5), nrow=2)
muSim <- apply(muSim,1, cumsum)
muSim <- t(muSim)
sdSim <- matrix(sqrt(1/rgamma(2*K+2, shape = 10, rate = 10)), nrow=2)
print(muSim)
pos <- lapply(1:(K+1), function(d) 1*(rupt[d]<(1:n )))
pos <- Reduce('+', x=pos)+1
x <- matrix(rnorm(2*length(pos), mean=muSim[,pos], sd=sdSim[,pos]), nrow=2)
bisig_plot(x = x)
n = dim(x)[2]
res <- hybrid_simultanee(x, P=2, Kmax=10)
Kopt=5
param <- res$param[[Kopt]]
bisig_plot(x = x, rupt = param$rupt, mu=param$phi$mu )
|
Add the following code to your website.
For more information on customizing the embed code, read Embedding Snippets.