tests/testthat/helper_solve_fixed_seed.R

In prioritizr: Systematic Conservation Prioritization in R

#' Solve with a fixed seed
#'
#' This function is a wrapper for [prioritizr::solve()] that imposes
#' a fixed seed when solving the problem.
#'
#' @param x [problem()] object.
#'
#' @param ... passed to [prioritizr::solve()].
#'
#' @param seed `integer` denoting the random number state. Defaults to 500.
#'
#' @details
#' This function is useful because, by default, the optimization problem will
#' shuffled prior to solving it (because the default portfolio
#' [add_shuffle_portfolio()]. This means that solving the exact same
#' problem multiple times with a slightly relaxed optimality gap can
#' produce different results. As such, when running tests, this wrapper
#' function is useful to promote consistency between runs.
#'
#' @return The output from running [prioritizr::solve()] on the `x`.
solve_fixed_seed <- function(x, ..., seed = 500) {
  withr::with_seed(500, solve(x, ...))
}