Z-residuals for hurdle negative binomial models fitted with brms

Zresidual.hurdle_negbinomial.brms() is the S3 method for Zresidual() when applied to hurdle negative binomial models fitted with brms::brm() and family = hurdle_negbinomial(). Objects are dispatched here when the fitted object is a "brmsfit" with brms::family(object)$family == "hurdle_negbinomial" and has been internally tagged with the class "hurdle_negbinomial.brms" by Zresidual().

In normal use, users should call Zresidual() directly on the brmsfit object (for example Zresidual(fit)), rather than calling Zresidual.hurdle_negbinomial.brms() explicitly.

Usage

# S3 method for class 'hurdle_negbinomial.brms'
Zresidual(
  object,
  nrep = 1,
  data,
  type = c("hurdle", "count", "zero"),
  method = "iscv",
  ...
)

Arguments

object

A brmsfit object with hurdle negative binomial family (brms::family(object)$family == "hurdle_negbinomial").

nrep

Integer; number of replicated Z-residual sets to generate. Default is 1.

data

Optional data frame used for prediction or residual computation. If NULL, the data stored inside the brmsfit object are used.

type

Character string specifying which part of the model to compute Z-residuals for:

• "zero" — the hurdle/zero part;

• "count" — the truncated negative binomial count part;

• "hurdle" — the full hurdle negative binomial model.

The default is "hurdle".

method

Character string specifying the residual calculation method: "iscv" for importance-sampled cross-validated randomized predictive p-values, "rpost" for randomized posterior predictive p-values, or "mpost" for middle-value posterior predictive p-values. Default is "iscv".

...

Further arguments passed to the underlying implementation function Zresidual_hurdle_negbinomial_brms().

Value

A numeric matrix of Z-residuals (one column per replication) as returned by Zresidual_hurdle_negbinomial_brms(), with the class "zresid" added to its class vector.

Examples

if (FALSE) { # \dontrun{
  library(brms)
  fit_hnb <- brm(y ~ x1 + x2, data = df,
                 family = hurdle_negbinomial())

  ## Counts part only
  z_count <- Zresidual(fit_hnb, type = "count", method = "iscv")

  ## Full hurdle model with 2 replicates
  z_hurdle <- Zresidual(fit_hnb, type = "hurdle", nrep = 2)
} # }