Skip to contents

Compute Savege-Dickey Bayes factors

Source: R/bayes_factor.R
bf_table.Rd

Computes Savage-Dickey density ratio Bayes factors for each (requested set of) parameter in the model. By default, all applicable parameters are tested and returned in a tidy data frame. The type argument controls which parameter groups are included; the variable argument can be used to filter to effects involving specific variables. The log_BF10 argument allows including the natural log of the Bayes factor in the output, and round controls numeric rounding of the results.

Usage

bf_table(
  object,
  type = "all",
  lag = 1L,
  null_value = 0,
  variable = NULL,
  log_BF10 = FALSE,
  round = 5L
)

Arguments

object

A bvarnet object returned by bvar().

type

Character vector specifying which parameter groups to test. Use "all" (default) to include all applicable groups automatically. Available options:

"ar"

Autoregressive effects (self-loops). Per-cell BFs for the lag specified by lag, plus a joint BF.

"cl"

Cross-lagged effects. Same structure as "ar".

"intercepts"

Intercept parameters. Skipped automatically for ordinal outcomes.

"fe"

Non-intercept fixed effects (covariates).

"lag_fe"

Joint BFs for lag \(\times\) covariate interaction terms. Only available when the model was fitted with fe_interactions containing lag terms.

"temporal"

Joint BF for the entire temporal structure (all AR + CL parameters across all lags). When lag \(\times\) covariate interactions are present, additional omnibus rows are included.

lag

Integer; which lag block to use (default 1). Applies to "ar" and "cl" types.

null_value

Numeric scalar; the null hypothesis value (default 0).

variable

Character vector or NULL (default). One or more variable names. When set, only effects involving these variables are included.

log_BF10

Logical; if TRUE, an additional log_BF10 column (natural log of BF10) is appended to the output. Default is FALSE.

round

Integer or NULL; number of decimal places to round numeric output columns. Default is 5. Set to NULL to disable rounding.

Value

A data frame with columns: type, predictor, outcome, BF10 (and optionally log_BF10).

Which prior the Bayes factor divides by

The Savage-Dickey density ratio is only valid against the prior the model was actually fitted with, so bf_table() evaluates the prior density using object$priors_effective — resolved per outcome, since Gaussian nodes scale their default priors by the outcome SD (see set_priors). For a Gaussian outcome with default priors this makes intercepts and fe Bayes factors differ from what the unscaled object$priors would give, by roughly a factor of sd(y). ar, cl and temporal tests are unaffected: phi priors are never scaled.