Skip to contents

Default Priors for a Model Family

Source: R/priors.R
priors_default.Rd

Display the priors a fit uses for every component ratiod_priors() leaves unspecified, as a starting point for customization.

Usage

priors_default(family = NULL, spatial = FALSE, temporal = FALSE)

Arguments

family

A tulpaRatio family object (e.g., ratiod_negbin_negbin()). If NULL (default), shows the dispersion default of every family.

spatial

Logical; if TRUE, include the BYM2 mixing prior.

temporal

Logical; if TRUE, include the AR1 correlation prior.

Value

Invisibly, a ratiod_priors object holding the defaults (for family = NULL, with the negative binomial dispersion default).

Details

  • Fixed effects (beta): Normal(0, 10) on the link scale.

  • Random effect SD (sigma): Half-Cauchy(2.5); the Laplace backend uses a PC prior with P(sigma > 1) = 0.01.

  • Dispersion (phi): Gamma(2, 2) for the lognormal log-scale SD, Gamma(2, 0.5) for a Gamma shape (gamma_gamma, poisson_gamma), Gamma(2, 0.1) for a negative binomial size or beta-binomial precision.

  • Temporal correlation (rho): Beta(2, 2) on (rho + 1) / 2, symmetric about rho = 0.

  • Spatial mixing (rho_spatial): Beta(1, 1), uniform on (0, 1).

See also

ratiod_priors() for creating custom priors

Examples

priors_default(ratiod_negbin_negbin())
#> Default priors for the negbin_negbin family
#> ===========================================
#> 
#> Fixed effects (beta):
#>   Normal(0.00, 10.00)
#> Random effect SD (sigma):
#>   Half-Cauchy(2.50)
#>   (Laplace backend: PC prior, P(sigma > 1) = 0.01)
#> Dispersion (phi):
#>   Gamma(2.00, 0.10)  [mean = 20.00]
priors_default(ratiod_poisson_gamma(), spatial = TRUE, temporal = TRUE)
#> Default priors for the poisson_gamma family
#> ===========================================
#> 
#> Fixed effects (beta):
#>   Normal(0.00, 10.00)
#> Random effect SD (sigma):
#>   Half-Cauchy(2.50)
#>   (Laplace backend: PC prior, P(sigma > 1) = 0.01)
#> Dispersion (phi):
#>   Gamma(2.00, 0.50)  [mean = 4.00]
#> Temporal autocorrelation (rho_temporal):
#>   Beta(2.00, 2.00)  [mean = 0.50]
#> Spatial mixing (rho_spatial):
#>   Beta(1.00, 1.00)  [mean = 0.50]
priors_default()
#> Default priors
#> ==============
#> 
#> Fixed effects (beta):
#>   Normal(0.00, 10.00)
#> Random effect SD (sigma):
#>   Half-Cauchy(2.50)
#>   (Laplace backend: PC prior, P(sigma > 1) = 0.01)
#> Dispersion (phi):
#>   lognormal: Gamma(2.00, 2.00)  [mean = 1.00]
#>   gamma_gamma, poisson_gamma: Gamma(2.00, 0.50)  [mean = 4.00]
#>   negbin_negbin, negbin_gamma, beta_binomial: Gamma(2.00, 0.10)  [mean = 20.00]