Check Bayesian models fitted with brms
Usage
dh_check_brms(
model,
resp = NULL,
integer = FALSE,
plot = TRUE,
nsamples = 1000,
ntrys = 5,
...
)Arguments
- model
A fitted model
brms::brmsfit-class(). Categorical and ordinal models not supported by now.- resp
Optional name of response variable (for multivariate models).
- integer
Logical (TRUE/FALSE), indicating if response is an integer, as in Poisson and binomial models
- plot
Logical. Plot residual checks? Default is TRUE.
- nsamples
Integer. Number of samples to draw from the posterior.
- ntrys
Integer. Number of trys to use for truncated distributions. See
brms::posterior_predict().- ...
Further arguments for
DHARMa::plotResiduals()
Value
An object of type DHARMa. See DHARMa::createDHARMa() for more details.
Examples
if (FALSE) { # interactive()
#' # Example models taken from brms::brm()
# Poisson regression for the number of seizures in epileptic patients
fit1 <- brm(count ~ zAge + zBase * Trt + (1|patient),
data = epilepsy, family = poisson())
simres <- dh_check_brms(fit1, integer = TRUE)
plot(simres, form = epilepsy$zAge)
testDispersion(simres)
# Probit regression using the binomial family
ntrials <- sample(1:10, 100, TRUE)
success <- rbinom(100, size = ntrials, prob = 0.4)
x <- rnorm(100)
data4 <- data.frame(ntrials, success, x)
fit4 <- brm(success | trials(ntrials) ~ x, data = data4,
family = binomial("probit"))
summary(fit4)
simres <- dh_check_brms(fit4, integer = TRUE)
plot(simres, form = data4$x)
# Multivariate (multiresponse) model
data("BTdata", package = "MCMCglmm")
bf_tarsus <- bf(tarsus ~ sex + (1|p|fosternest) + (1|q|dam))
bf_back <- bf(back ~ hatchdate + (1|p|fosternest) + (1|q|dam))
fit <- brm(bf_tarsus + bf_back + set_rescor(TRUE),
data = BTdata, chains = 2, cores = 2)
dh_check_brms(fit, resp = "tarsus")
brms::pp_check(fit, resp = "tarsus")
dh_check_brms(fit, resp = "back")
brms::pp_check(fit, resp = "back")
}