contHMM-access {tileHMM} | R Documentation |
Access to model parameters and densities of emission distributions.
## S3 method for class 'contHMM': x[i, j, transition = TRUE, log = FALSE, sum = TRUE, ...] ## S4 method for signature 'hmm': length(x)
x |
Object of class contHMM |
i |
State for which parameter values should be retrieved. This can either a numeric value giving the state index or a character string with the state name. |
j |
Second index identifying parameter (see Details). |
transition |
Logical indicating whether transition probabilities or density values should be returned. |
log |
Logical indicating whether values should be log transformed before they are returned. |
sum |
Logical indicating whether densities of mixture components should be summed up.
This is ignored if transition = TRUE . |
... |
Futher arguments to be passed to and from other methods. |
The ‘[’ function allows access to the transition probability matrix of the model as well as the emission
distributions. If transition = TRUE
the transition probability matrix is accessed. In this case
i
and j
identify rows and columns of the matrix respectively. Both can be given as either
numeric index or name of the respective states. Either or both of i
and j
may be missing
to indicate that an entire row or column should be selected.
If transition = FALSE
the emission distribution of state i
is accessed instead. In this case
the density function is evaluated at point j
.
For ‘[’ either a subset of the transition probability matrix of x
or the probability density of state i
evaluated at point j
(see Details).
For ‘length’ the number of states in the model.
Peter Humburg
## create two state HMM with t distributions state.names <- c("one","two") transition <- c(0.1, 0.02) location <- c(1, 2) scale <- c(1, 1) df <- c(4, 6) hmm <- getHMM(list(a=transition, mu=location, sigma=scale, nu=df), state.names) ## number of states in the model length(hmm) ## transition probability from state 'one' to state 'two' hmm["one", "two"] ## or equivalently hmm[1, 2] ## get the transition probability matrix hmm[ , ] ## evaluate emission distribution function of state 'one' at 0 hmm["one", 0, transition = FALSE] ## again, this time using log transformation hmm["one", 0, transition = FALSE, log = TRUE]