summary methods there are also some standard
extraction methods defined for objects of class
from a call to
The implementation is illustrated in Meyer et al. (2017, Section 5),
# S3 method for hhh4 print(x, digits = max(3, getOption("digits") - 3), ...) # S3 method for hhh4 summary(object, maxEV = FALSE, ...) # S3 method for hhh4 coef(object, se = FALSE, reparamPsi = TRUE, idx2Exp = NULL, amplitudeShift = FALSE, ...) # S3 method for hhh4 fixef(object, ...) # S3 method for hhh4 ranef(object, tomatrix = FALSE, intercept = FALSE, ...) # S3 method for hhh4 coeflist(x, ...) # S3 method for hhh4 formula(x, ...) # S3 method for hhh4 nobs(object, ...) # S3 method for hhh4 logLik(object, ...) # S3 method for hhh4 vcov(object, reparamPsi = TRUE, idx2Exp = NULL, amplitudeShift = FALSE, ...) # S3 method for hhh4 confint(object, parm, level = 0.95, reparamPsi = TRUE, idx2Exp = NULL, amplitudeShift = FALSE, ...) # S3 method for hhh4 residuals(object, type = c("deviance", "response"), ...)
an object of class
the number of significant digits to use when printing
logical indicating if the summary should contain the
(range of the) dominant eigenvalue as a measure of the importance of
the epidemic components. By default, the value is not calculated as
this may take some seconds depending on the number of time points
and units in
coeflist methods: arguments passed to
For the remaining methods: unused (argument of the generic).
TRUE (default), the overdispersion parameter from the
negative binomial distribution is transformed from internal scale (-log)
to standard scale, where zero corresponds to a Poisson distribution.
logical switch indicating if standard errors are required
integer vector selecting the parameters
which should be returned on exp-scale.
idx2Exp = TRUE will exp-transform all
parameters except for those associated with
or already affected by
logical switch indicating whether the parameters
for sine/cosine terms modelling seasonal patterns
addSeason2formula) should be transformed
to an amplitude/shift formulation.
FALSE (default), the vector of
all random effects is returned (as used internally). However, for
random intercepts of
type="car", the number of parameters is
one less than the number of regions and the individual parameters are
not obviously linked to specific regions. Setting
TRUE returns a more useful representation of random effects
in a matrix with as many rows as there are regions and as many
columns as there are random effects. Here, any CAR-effects are
transformed to region-specific effects.
FALSE (default), the returned
random effects represent zero-mean deviations around the
corresponding global intercepts of the log-linear predictors.
intercept=TRUE adds these global intercepts to the
result (and implies
a vector of numbers or names, specifying which parameters are to be given confidence intervals. If missing, all parameters are considered.
the confidence level required.
the type of residuals which should be returned. The
"deviance" (default) and
coef-method returns all estimated (regression)
parameters from a
If the model includes random effects, those can be extracted with
fixef returns the fixed parameters.
coeflist-method extracts the model coefficients in a list
(by parameter group).
formula-method returns the formulae used for the
three log-linear predictors in a list with elements
nobs-method returns the number of observations used
for model fitting.
logLik-method returns an object of class
For a random effects model, the value of the penalized
log-likelihood at the MLE is returned, but degrees of freedom are
not available (
As a consequence,
BIC are only
well defined for models without random effects;
otherwise these functions return
vcov-method returns the estimated
variance-covariance matrix of the regression parameters.
The estimated variance-covariance matrix of random effects is
confint-method returns Wald-type confidence
intervals (assuming asymptotic normality).
Michaela Paul and Sebastian Meyer
Meyer, S., Held, L. and Höhle, M. (2017): Spatio-temporal analysis of epidemic phenomena using the R package surveillance. Journal of Statistical Software, 77 (11), 1-55. doi: 10.18637/jss.v077.i11