Skip to contents

Summarises draws from the predictive distribution

Source: R/tidy.R
tidy_forecast.Rd

Transforms draws from the joint predictive distribution into a table containing posterior predictive means, medians, standard deviations, and equal-tailed intervals, or retains the individual draws when requested.

Usage

tidy_forecast(object, ...)

# Default S3 method
tidy_forecast(object, ...)

Arguments

object

A posterior model object or a Forecasts object.

...

Additional arguments passed to computation methods.

Value

A bsvar_post_tbl (tibble subclass) with columns model, object_type, variable, horizon, mean, median, sd, lower, and upper. When draws = TRUE, columns draw and value replace the summary statistics.

Examples

# Small posterior (S = 5 draws)
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.
post <- bsvars::estimate(spec, S = 5, show_progress = FALSE)

# Summaries of the predictive distribution
result <- tidy_forecast(post, horizon = 3)
head(result)
#> # A tibble: 6 × 19
#>   model  object_type variable horizon   mean median    sd  lower  upper
#>   <chr>  <chr>       <chr>      <dbl>  <dbl>  <dbl> <dbl>  <dbl>  <dbl>
#> 1 model1 forecast    ttr            1  -8.70  -8.79 0.218  -8.88  -8.41
#> 2 model1 forecast    ttr            2  -8.73  -8.77 0.145  -8.86  -8.54
#> 3 model1 forecast    ttr            3  -8.50  -8.50 0.137  -8.62  -8.33
#> 4 model1 forecast    gs             1 -10.3  -10.1  0.612 -11.0   -9.71
#> 5 model1 forecast    gs             2 -10.2  -10.0  0.371 -10.7   -9.88
#> 6 model1 forecast    gs             3 -10.8  -10.9  0.482 -11.3  -10.2 
#> # ℹ 10 more variables: 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>