Sample points in a domain

Draws training points uniformly at random or with Latin hypercube sampling. Variables listed in fixed are held at the given values, which is how boundary and initial points are generated.

Usage

pinn_sample(n, domain, method = c("lhs", "uniform"), fixed = list())

Arguments

n

Number of points.

domain

A pinn_domain().

method

"lhs" (Latin hypercube, the default) or "uniform".

fixed

Named list of numeric values. When a variable has several values, the points are split evenly between them. With several fixed variables, every combination of their values is used, so list(x = c(0, 1), y = c(0, 1)) gives the four corners. Each value (or combination) gets at least one point, so the result can have more than n rows when n is smaller than the number of combinations.

Value

A data frame with one column per domain variable.

Examples

dom <- pinn_domain(x = c(0, 1), t = c(0, 2))
head(pinn_sample(5, dom))
#>           x          t
#> 1 0.7723993 1.97610760
#> 2 0.3928102 0.77896715
#> 3 0.1289966 1.21089190
#> 4 0.9242108 0.08215696
#> 5 0.4770447 1.18733292
pinn_sample(4, dom, fixed = list(x = c(0, 1)))
#>   x         t
#> 1 0 0.4947259
#> 2 0 0.5910430
#> 3 1 1.0718271
#> 4 1 1.7888452