hmm.discnp (version 0.2-4)

logLikHmm: Log likelihood of a hidden Markov model

Description

Calculate the log likelihood of a hidden Markov model with discrete non-parametric observation distributions.

Usage

logLikHmm(y, par)

Arguments

y

A sequence, or list of sequences, of observations from a hidden Markov model with discrete non-parametric observation distributions.

par

A list with entries

• Rho: The matrix whose columns consist of the probabilities specifying the distributions of the observations.

• tpm: The transition probability matrix of the Markov chain.

• (Optionally) ispd: The vector of probabilities specifying the initial state probability distribution, or a matrix each of whose columns are trivial (“delta function”) vectors specifying the “most probable” initial state for each observation sequence. If the component ispd of par is not present then ispd is calculated as the stationary distribution determined by tpm.

This list may be an object returned by hmm().

Value

The loglikehood of y given the parameter values specified in par.

Details

The observation values (the entries of y, or of the sequences which constitute the entries of y if y is a list) must be members of the set of row names of Rho. More specifically, if Rho has no row names then these observation values must all be integers between 1 and nrow(Rho). If Rho has row names then these observations must be character strings and must be members of row.names(Rho).

References

See hmm() for references.

See Also

hmm(), pr(), sp()

Examples

# NOT RUN {
P <- matrix(c(0.7,0.3,0.1,0.9),2,2,byrow=TRUE)
R <- matrix(c(0.5,0,0.1,0.1,0.3,
              0.1,0.1,0,0.3,0.5),5,2)
set.seed(42)
lll   <- sample(250:350,20,TRUE)
y.num <- sim.hmm(lll,P,R)
y.let <- sim.hmm(lll,P,R,yval=letters[1:5])
ll1 <- logLikHmm(y.num,list(tpm=P,Rho=R))
row.names(R) <- letters[1:5]
ll2 <- logLikHmm(y.let,list(tpm=P,Rho=R))
# Note: ll2 differs from ll1 since y.num and y.let were
# simulated from different random seeds.
ll3 <- logLikHmm(y.let,list(tpm=P,Rho=R,ispd=c(0.5,0.5)))
fit <- hmm(y.num,K=2)
ll4 <- logLikHmm(y.num,fit) # Use the fitted rather than the "true" parameters.
# }