nauf models.When nauf_pmmeans is used with Bayesian regressions, the
elements of the resulting nauf.pmm.list have class
nauf.pmm.stan.
# S3 method for nauf.pmm.stan
summary(object, probs = c(0.025, 0.975),
type = c("link", "response"), ...)# S3 method for summ.nauf.pmm.stan
print(x, row.names = FALSE, mcse = FALSE,
rhat = NULL, ...)
# S3 method for nauf.pmm.stan
as.array(x, ...)
# S3 method for nauf.pmm.stan
as.matrix(x, ...)
# S3 method for nauf.pmm.stan
print(x, ...)
# S3 method for nauf.pmm.stan
as.data.frame(x, row.names = NULL, optional = FALSE,
...)
# S3 method for summ.nauf.pmm.stan
as.data.frame(x, row.names = NULL,
optional = FALSE, ...)
A nauf.pmm.stan object.
A vector of quantiles to calculate for the estimates. The default is a 95 The median (0.5) is always added regardless of whether it is specified.
If "link" (the default), then the estimates are not
transformed prior to summary; if "response", then the inverse link
function in the object's family element is applied to the
samples element prior to summary.
See the Methods section.
Either a nauf.pmm.stan object, or the summ.nauf.pmm.stan
object returned by calling summary on a nauf.pmm.stan object.
Changes from the defaults are ignored.
A logical indicating whether the Monte Carlo standard errors
should be printed (default FALSE since the column is always recoverable
by dividing the SD column by the square root of the ESS column).
An optional logical indicating whether or not to print the
Gelman-Rubin R-hat statistic. If rhat = NULL (the default), then
the Rhat column of the summary is only printed if any of the
statistics are greater that 1.1. Regardless of the value of the
rhat argument, a warning is issued if any of the statistics are
greater than 1.1.
The returned object depends on the function. See the 'Methods' section.
Returns a data frame with class
summ.nauf.pmm.stan, with means, standard deviations, medians, mean
absolute differences, the posterior probability that the estimate is
greater than zero, quantiles specified in probs, the effective
sample size, Monte Carlo standard error, and Gelman-Rubin R-hat statistic.
Additional arguments in ... are ignored. Quantiles, effective
samples size, and Monte Carlo standard error, and R-hat statistics are
computed using monitor.
For summ.nauf.pmm.stan objects, the data frame is
printed, omitting the Monte Carlo standard error and
R-hat statistic based on mcse and rhat. If the inverse
link function was used, then a message indicating the response type
is also printed. Additional arguments in ... are passed to
print.data.frame. For nauf.pmm.stan objects,
first summary is called, passing along ...,
and then print is called on the resulting summ.nauf.pmm.stan, also
passing along ... arguments.
For summ.nauf.pmm.stan objects, removes the
data.frame contained in a summ.nauf.pmm.stan. Additional
arguments in ... are ignored. For nauf.pmm.stan objects,
first summary is called, passing along ..., and then
as.data.frame is called on the resulting summ.nauf.pmm.stan
object, ignoring ... arugments.
Returns the samples element of a
nauf.pmm.stan object. Additional arguments in ... are
ignored.
Returns the samples element of a
namf.pmm.stan object, flattening the array into a matrix such that
the rows represent iterations ordered by chain and the columns represent
parameters.
The nauf.pmm.stan object is a list with the following elements.
A data frame with the levels of each factor (or 'inc_1' for
covariates), with one row for each element in the third dimension of
samples.
The fixed effects model matrix showing the contrasts
applied to the regression coefficients (rows correspond to the
names element).
An array with three dimensions. The first corresponds to
iterations, the second to chains, and the third to parameters (the same
structure returned by as.array.stanfit).
The regression family.
The label of the inverse link (e.g. probability, rate, etc.).
A list with additional information based on the type of model.