Skip to contents

Fit INLA model

Source: R/fit_inla_model.R
fit_inla_model.Rd

Fit an INLA model based on a constructed data stack and formula

Usage

fit_inla_model(
  formula,
  data_stack,
  spde,
  samplesize_vec = 1,
  precision_vec = 1,
  family = "binomial",
  link = "logit",
  fixed_effects_pc_prior = list(threshold = 3, prob_above = 0.05),
  verbose = TRUE
)

Arguments

formula

(character) INLA formula to fit. Generated in prepare_inla_data_stack(), then interpreted using stats::as.formula() within the call to INLA::inla().

data_stack

Stacked data, covariates, and spatial index. Generated in prepare_inla_data_stack().

spde

SPDE object generated by prepare_inla_data_stack().

samplesize_vec

(integer(N), default 1) Sample sizes for each data observation. Only used for binomial data models.

precision_vec

(numeric(N), default 1) Precision for each data observation. Only used for gaussian data models.

family

(character, default 'binomial') GLM family to use. For more information, see stats::family().

(character, default 'logit') Link function to use, typically related to the GLM family.

fixed_effects_pc_prior

A named list specifying the penalized complexity prior for all fixed effects except for the intercept. The two named items are "threshold", the test threshold for the size of each fixed effect, and "prob_above", the prior probability that the beta for each covariate will EXCEED that threshold.

verbose

(logical(1), default TRUE) Log progress for INLA model fitting?

Value

A fitted INLA model object created by INLA::inla()

Details

Using INLA::inla() with reasonable defaults and settings tuned to predict across a grid.

See also

MbgModelRunner