cheops_lapply: Make a parallel load balancing lapply call on the cluster

In bonartm/cheopsr: Talking to the Hight Performance CHEOPS Cluster at the University of Cologne

Description

Make a parallel load balancing lapply call on the cluster

Usage

cheops_lapply(x, fun, options, jobname, args = list(), packages = c(),
  load.balancing = TRUE, module = getOption("cheopsr.module"),
  account = getOption("cheopsr.account"),
  lib = getOption("cheopsr.libloc"))

Arguments

x	a vector or list, see clusterApply
fun	a function, see clusterApply
options	named list of cluster options generated by cheops_slurmcontrol
jobname	name of the job as character string
args	optionally a list of additional arguments passed to fun via do.call
packages	character vector of package names to be loaded on the cluster
load.balancing	logical indicating if clusterApplyLB or clusterApply should be used
module	name of r module to be loaded
account	name of the account to be charged
lib	location of the r lib on the cluster

Details

starts a slurm job on the cluster. after the job is terminated, results have to be retrieved manually with the command: cheops_readRDS(<filepath>).

Value

a named list with the name and id of the job and the file path on the cluster for the results

bonartm/cheopsr documentation built on May 3, 2019, 9:37 p.m.