GofHMMGen {GenHMM1d} | R Documentation |
Goodness-of-fit of univariate hidden Markov model
Description
This function performs a goodness-of-fit test for a univariate hidden Markov model
Usage
GofHMMGen(
y,
ZI = 0,
reg,
family,
start = 0,
max_iter = 10000,
eps = 1e-04,
size = 0,
n_samples = 1000,
n_cores = 1,
useFest = TRUE
)
Arguments
y |
observations |
ZI |
1 if zero-inflated, 0 otherwise (default) |
reg |
number of regimes |
family |
distribution name; run the function distributions() for help |
start |
starting parameter for the estimation |
max_iter |
maximum number of iterations of the EM algorithm; suggestion 10000 |
eps |
precision (stopping criteria); suggestion 0.0001. |
size |
additional parameter for some discrete distributions; run the command distributions() for help |
n_samples |
number of bootstrap samples; suggestion 1000 |
n_cores |
number of cores to use in the parallel computing |
useFest |
TRUE (default) to use the first estimated parameters as starting value for the bootstrap, FALSE otherwise |
Value
pvalue |
pvalue of the Cramer-von Mises statistic in percent |
theta |
Estimated parameters; (r x p) |
Q |
estimated transition matrix; ; (r x r) |
eta |
(conditional probabilities of being in regime k at time t given observations up to time t; (n x r) |
lambda |
conditional probabilities of being in regime k at time t given all observations; (n x r) |
U |
matrix of Rosenblatt transforms; (n x r) |
cvm |
Cramer-von-Mises statistic for goodness-of-fit |
W |
pseudo-observations that should be uniformly distributed under the null hypothesis |
LL |
log-likelihood |
nu |
stationary distribution |
AIC |
Akaike information criterion |
BIC |
bayesian information criterion |
CAIC |
consistent Akaike information criterion |
AICcorrected |
Akaike information criterion corrected |
HQC |
Hannan-Quinn information criterion |
stats |
Empirical means and standard deviation of each regimes using lambda |
pred_l |
Estimated regime using lambda |
pred_e |
Estimated regime using eta |
runs_l |
Estimated number of runs using lambda |
runs_e |
Estimated number of runs using eta |
Examples
family = "gaussian"
Q = matrix(c(0.8, 0.3, 0.2, 0.7), 2, 2) ; theta = matrix(c(0, 1.7, 0, 1),2,2) ;
y = SimHMMGen(theta, size=0, Q, ZI=1, family, 100)$SimData
out=GofHMMGen(y,1,2,family,n_samples=10)