Aliases: mclapply mcmapply mcMap
Keywords: interface
### ** Examples ## No test: simplify2array(mclapply(rep(4, 5), rnorm))
[,1] [,2] [,3] [,4] [,5] [1,] 1.5436178 2.2195551 -0.59771234 0.9284825 -1.17773758 [2,] -0.3127122 0.1154679 -2.64483554 -0.3497437 -1.84956446 [3,] 0.4986590 1.7314462 0.83552972 0.3157886 0.34313444 [4,] 0.3620856 0.1221489 -0.03274486 1.1421857 0.01718078
# use the same random numbers for all values set.seed(1) simplify2array(mclapply(rep(4, 5), rnorm, mc.preschedule = FALSE, mc.set.seed = FALSE))
[,1] [,2] [,3] [,4] [,5] [1,] -0.6264538 -0.6264538 -0.6264538 -0.6264538 -0.6264538 [2,] 0.1836433 0.1836433 0.1836433 0.1836433 0.1836433 [3,] -0.8356286 -0.8356286 -0.8356286 -0.8356286 -0.8356286 [4,] 1.5952808 1.5952808 1.5952808 1.5952808 1.5952808
## Contrast this with the examples for clusterCall library(boot) cd4.rg <- function(data, mle) MASS::mvrnorm(nrow(data), mle$m, mle$v) cd4.mle <- list(m = colMeans(cd4), v = var(cd4)) mc <- getOption("mc.cores", 2) run1 <- function(...) boot(cd4, corr, R = 500, sim = "parametric", ran.gen = cd4.rg, mle = cd4.mle) ## To make this reproducible: set.seed(123, "L'Ecuyer") res <- mclapply(seq_len(mc), run1) cd4.boot <- do.call(c, res) boot.ci(cd4.boot, type = c("norm", "basic", "perc"), conf = 0.9, h = atanh, hinv = tanh)
BOOTSTRAP CONFIDENCE INTERVAL CALCULATIONS
Based on 1000 bootstrap replicates
CALL :
boot.ci(boot.out = cd4.boot, conf = 0.9, type = c("norm", "basic",
"perc"), h = atanh, hinv = tanh)
Intervals :
Level Normal Basic Percentile
90% ( 0.4575, 0.8547 ) ( 0.4703, 0.8575 ) ( 0.4964, 0.8663 )
Calculations on Transformed Scale; Intervals on Original Scale
## Usage of the affinity.list parameter A <- runif(2500000,0,100) B <- runif(2500000,0,100) C <- runif(5000000,0,100) first <- function(i) head(sort(i), n = 1) # Restict all elements of X to run on CPU 1 and 2 affL <- list(c(1,2), c(1,2), c(1,2)) mclapply(list(A, A, A), first, mc.preschedule = FALSE, affinity.list = affL)
[[1]] [1] 0.0001111999 [[2]] [1] 0.0001111999 [[3]] [1] 0.0001111999
# Completion times are assumed to have a high variance # To optimize the overall execution time elements of X are scheduled to suitable CPUs # Assuming that the runtime for C is as long as the runtime of A plus B # mapping: A to 1 , B to 1, C to 2 X <- list(A, B, C) affL <- c(1, 1, 2) mclapply(X, first, mc.preschedule = FALSE, affinity.list = affL)
[[1]] [1] 0.0001111999 [[2]] [1] 1.373701e-05 [[3]] [1] 2.18628e-05
## End(No test)