Sample an uncertain paired curve

Mirrors the C# UncertainOrderedPairedData.CurveSample()/CurveSample(double): collapses a curve whose Y-coordinate is a whole distribution at each x down to a plain x-y curve, either at the distributions' means (probability = NULL, the default) or at a shared quantile (probability in [0, 1]).

Usage

uncertain_curve_sample(
  x,
  distributions,
  probability = NULL,
  order_x = "ascending",
  order_y = "ascending",
  strict_x = TRUE,
  strict_y = TRUE
)

Arguments

x

numeric vector of x positions, at least one element.

distributions

a distribution() object, or a list of them the same length as x – one distribution per x position. A single distribution is recycled across every x.

probability

quantile in [0, 1] to sample at; NULL (default) samples the mean. Values outside [0, 1] are rejected – the underlying C# CurveSample(double) silently clamps them instead, so this range check is enforced here rather than core-side.

order_x, order_y

one of "ascending" (default), "descending", or "none".

strict_x, strict_y

require x/y to strictly increase/decrease (per order_x/order_y) between consecutive ordinates. Default TRUE.

Value

a data frame with columns x and y.

Examples

x <- c(1, 2, 3, 5)
d <- list(distribution("Triangular", c(1, 2, 3)), distribution("Triangular", c(2, 4, 5)),
          distribution("Triangular", c(6, 8, 12)), distribution("Triangular", c(13, 19, 20)))
uncertain_curve_sample(x, d, probability = 0.5)
#>   x         y
#> 1 1  2.000000
#> 2 2  3.732051
#> 3 3  8.535898
#> 4 5 17.582576