HPC_optim: Computes an optimized repartition of the HPC computing

In hrvg/RiverML: Machine Learning for River Science

Description

This function uses the limits of maximum concurrent jobs, the input/output time and computng time. The time are derived using profvis package.

Usage

HPC_optim(
  len,
  hpc_lim = 10000,
  io_time = 20000,
  computing_time = 1830,
  AssocGrpCpuLimit = 258,
  pl = TRUE,
  core_per_task = 1
)

Arguments

len	the number of time the computation needs to be iterated
hpc_lim	the maximum number of subjobs per user
io_time	the input/output time in ms, default to 20e3
computing_time	the computing time in ms, default to 1830
AssocGrpCpuLimit	the maximum number of CPU per user, default to 258
pl	logical, if TRUE a plot is produced, default to TRUE
core_per_task	number of CPU per task, default to 1.

Value

a data.frame with columns task_per_array, exec_time_min and n_array

hrvg/RiverML documentation built on Oct. 12, 2020, 10:40 a.m.