Compares response half-lives across model specifications — compare_half_life

Summarises across model specifications the first horizons at which posterior response draws fall to a specified fraction of a reference response.

Usage

compare_half_life_response(
  ...,
  horizon = NULL,
  type = c("irf", "cdm"),
  variables = NULL,
  shocks = NULL,
  variable = NULL,
  shock = NULL,
  fraction = 0.5,
  baseline = c("peak", "initial"),
  absolute = TRUE,
  probability = 0.9,
  scale_by = c("none", "shock_sd"),
  scale_var = NULL
)

Arguments

...: Posterior model objects or a named list of model objects.
horizon: Maximum horizon used when object is a posterior model object. If NULL (default), resolves to 20 periods.
type: Response type for posterior model objects: "irf" or "cdm".
variables: Response variables to include.
shocks: Structural shocks to include.
variable: Deprecated. Use variables instead.
shock: Deprecated. Use shocks instead.
fraction: Fraction of the reference level used to define the half-life.
baseline: Reference level: "peak" or "initial".
absolute: If TRUE, compute half-lives using absolute responses.
probability: Probability mass of the equal-tailed credible interval.
scale_by: Optional scaling mode for CDMs.
scale_var: Optional scaling variable specification.

Value

A bsvar_post_tbl combining half-life summary results across models, with a model column identifying each model specification.

Examples

data(us_fiscal_lsuw, package = "bsvars")
spec <- bsvars::specify_bsvar$new(us_fiscal_lsuw, p = 1)
#> The identification is set to the default option of lower-triangular structural matrix.
post1 <- bsvars::estimate(spec, S = 5, show_progress = FALSE)
post2 <- bsvars::estimate(spec, S = 5, show_progress = FALSE)

comp <- compare_half_life_response(m1 = post1, m2 = post2, horizon = 3)
head(comp)
#> # A tibble: 6 × 22
#>   model object_type   variable shock fraction baseline mean_half_life
#>   <chr> <chr>         <chr>    <chr>    <dbl> <chr>             <dbl>
#> 1 m1    half_life_irf ttr      ttr        0.5 peak               1   
#> 2 m1    half_life_irf ttr      gs         0.5 peak              NA   
#> 3 m1    half_life_irf ttr      gdp        0.5 peak               1.5 
#> 4 m1    half_life_irf gs       ttr        0.5 peak               1.75
#> 5 m1    half_life_irf gs       gs         0.5 peak               1   
#> 6 m1    half_life_irf gs       gdp        0.5 peak              NA   
#> # ℹ 15 more variables: median_half_life <dbl>, sd_half_life <dbl>,
#> #   lower_half_life <dbl>, upper_half_life <dbl>, reached_prob <dbl>,
#> #   analysis_type <chr>, interval_probability <dbl>, interval_method <chr>,
#> #   scale_by <chr>, scale_var <chr>, scale_factor <dbl>, units <chr>,
#> #   source_package <chr>, source_version <chr>, n_draws <int>