predict
gives empirical Bayes predictions of the response, while
sigma
gives the dispersion parameter.
# S3 method for mhglm
predict(object, newdata = NULL, type = c("link", "response"),
se.fit = FALSE, na.action = na.pass, ...)# S3 method for mhglm
sigma(object, ...)
an mhglm
object
these arguments behave as in predict.glm
. See Details, below.
further arguments passed to or from other methods.
The predict
function gives empirical Bayes posterior mean estimates of
response values. If se.fit = TRUE
, then the conditional variances of
the random effects are used along with the fixed effect variance-covariance
matrix to estimate the standard errors.
The sigma
function gives the square root of the dispersion parameter
or the model; for linear models, this is the error standard deviation.