ziphsmm (version 2.0.6)

hsmmviterbi_exp: Viterbi algorithm to decode the latent states in hidden semi-Markov models with covariates where the latent state durations have accelerated failure time structure

Description

Viterbi algorithm to decode the latent states in hidden semi-Markov models with covariates where the latent state durations have accelerated failure time structure

Usage

hsmmviterbi_exp(y, M, trunc, dtrate, dtparm, prior, zeroparm, emitparm, tpmparm,
  dt_x, zeroinfl_x, emit_x, tpm_x, plot = FALSE, xlim = NULL, ylim = NULL,
  ...)

Arguments

y

the observed series to be decoded

M

number of latent states

trunc

a vector specifying the truncation at the maximum number of dwelling time in each state.

dtrate

a vector for the scale parameters in the base exponential density for the latent state durations.

dtparm

a matrix of coefficients for the accelerated failure time model in each latent state

prior

a vector of prior probabilities

zeroparm

a vector of regression coefficients for the structural zero proportion in state 1

emitparm

a matrix of regression coefficients for the Poisson regression in each state

tpmparm

a vector of coefficients for the multinomial logistic regression in the transition probabilities

dt_x

a matrix of covariates for the latent state durations

zeroinfl_x

a matrix of covariates for the zero proportion

emit_x

a matrix of covariates for the Poisson means

tpm_x

a matrix of covariates for the transition

plot

whether a plot should be returned

xlim

vector specifying the minimum and maximum on the x-axis in the plot. Default to NULL.

ylim

vector specifying the minimum and maximum on the y-axis in the plot. Default to NULL.

...

further arguments to be passed to the plot() function

Value

decoded series of latent states

References

Walter Zucchini, Iain L. MacDonald, Roland Langrock. Hidden Markov Models for Time Series: An Introduction Using R, Second Edition. Chapman & Hall/CRC

Examples

# NOT RUN {
M <- 3
prior <- c(0.5,0.3,0.2)
dtrate <- c(6,5,4)
dtparm <- matrix(c(0.2,0.1,0.2),nrow=3)
zeroparm <- c(0,-0.2)
emitparm <- matrix(c(4,0.3,5,0.2,6,-0.1),3,2,byrow=TRUE)
tpmparm <- c(1,0.2,0.5,-0.2,0,0.2)

emit_x <- matrix(c(rep(1,1000),rep(0,1000)),nrow=2000,ncol=1)
dt_x <- emit_x
tpm_x <- emit_x
zeroinfl_x <- emit_x
trunc <- c(18,15,10)

re <- hsmmsim2_exp(prior,dtrate,dtparm,zeroparm,emitparm,tpmparm,
                  trunc, M, n, dt_x,tpm_x, emit_x, zeroinfl_x)
y <- re$series

rrr <- hsmmfit_exp(y,M,trunc,dtrate,dtparm,prior,zeroparm,emitparm,tpmparm,
                  dt_x,zeroinfl_x,emit_x,tpm_x,method="BFGS",control=list(trace=1))

decode <- hsmmviterbi_exp(y,M, trunc,dtrate,dtparm,
                          prior,zeroparm,emitparm,tpmparm,
                          dt_x, zeroinfl_x, emit_x, tpm_x)
sum(decode!=re$state)
# }