Constructs exact support metadata of a scalar posterior
distribution, for the support metadata of posterior samples
(posterior_metadata()) and the support of
posterior_density_attribute().
posterior_support_attribute(
bounds,
points = NULL,
type = NULL,
exact = TRUE,
source = NULL
)numeric vector c(lower, upper) with
lower <= upper; bounds may be infinite.
optional finite numeric vector of point-support locations.
support type: "interval", "points", or
"mixed". If NULL, "points" is used when the finite
bounds are exactly (not up to a tolerance) the range of
points and "interval" otherwise.
whether bounds and points describe the true
support. Set exact = FALSE for plotting or integration limits;
non-exact support is not used to exclude a null hypothesis value or to
reflect a kernel density estimate.
optional label of the source of the support metadata.
A BayesTools_posterior_support object.
Posterior support metadata must be created with this constructor or by a BayesTools posterior producer; other objects stored as support metadata are rejected with an error. An unclassed list of such objects, named by column, can hold the support of several columns.
posterior_support_attribute(c(0, Inf))
#> $bounds
#> [1] 0 Inf
#>
#> $points
#> numeric(0)
#>
#> $exact
#> [1] TRUE
#>
#> $source
#> NULL
#>
#> $type
#> [1] "interval"
#>
#> attr(,"class")
#> [1] "BayesTools_posterior_support" "list"
posterior_support_attribute(c(0, 1), points = c(0, 1), type = "points")
#> $bounds
#> [1] 0 1
#>
#> $points
#> [1] 0 1
#>
#> $exact
#> [1] TRUE
#>
#> $source
#> NULL
#>
#> $type
#> [1] "points"
#>
#> attr(,"class")
#> [1] "BayesTools_posterior_support" "list"