source("incl/start.R") strategies <- future:::supportedStrategies() message("*** times() ...") res0 <- NULL for (strategy in strategies) { message(sprintf("- plan('%s') ...", strategy)) plan(strategy) mu <- 1.0 sigma <- 2.0 ## NOTE: Using times() with %dopar% without proper parallel RNG will likely ## produce unreliable results. Here we don't produce random numbers so it ## is ok, but it's a toy example because just like base::replicate(), ## foreach::times() is commonly used for resampling purposes. res <- times(3L) %dopar% { dnorm(2L, mean = mu, sd = sigma) } print(res) if (is.null(res0)) { res0 <- res } else { stopifnot(all.equal(res, res0)) } # Shutdown current plan plan(sequential) message(sprintf("- plan('%s') ... DONE", strategy)) } ## for (strategy ...) message("*** times() ... DONE") source("incl/end.R")