A helper method to find the best multithreading configuration for your computer

Usage

optimizeCoreGx(sample_data, set = FALSE, report = !set)

Arguments

sample_data: TreatmentResponseExperiment
set: logical(1) Should the function modify your R environment with the predicted optimal settings? This changes the global state of your R session!
report: logical(1) Should a data.frame of results be returned by number of threads and operation be returned? Defaults to !set.

Value

If set=TRUE, modifies data.table threads via setDTthreads(), otherwise displays a message indicating the optimal number of threads. If report=TRUE, also returns a data.frame of the benchmark results.

Examples

# \donttest{
  data(merckLongTable)
  optimizeCoreGx(merckLongTable)
#> Benchmarking assay(sample_data, withDimnames=TRUE)...
#> Running with:
#>   nthread
#> 1       1
#> 2       2
#> Benchmarking assays(sample_data)...
#> Running with:
#>   nthread
#> 1       1
#> 2       2
#> Benchmarking reindex(sample_data)...
#> Running with:
#>   nthread
#> 1       1
#> 2       2
#> Optimal cores for your machine are: 2
#>                                          expression nthread     min_sec
#> 1 { assay(sample_data, 1, withDimnames = TRUE) NA }       1 0.004809363
#> 2 { assay(sample_data, 1, withDimnames = TRUE) NA }       2 0.004651678
#> 3                        { assays(sample_data) NA }       1 0.006402510
#> 4                        { assays(sample_data) NA }       2 0.006333161
#> 5                       { reindex(sample_data) NA }       1 0.007267264
#> 6                       { reindex(sample_data) NA }       2 0.007330232
#>    median_sec total_sec mem_alloc_mb  itr/sec   gc/sec
#> 1 0.005027693 0.4759264     3.318424 193.3072 4.202330
#> 2 0.004806819 0.4769783     2.181288 207.5566 4.193062
#> 3 0.006555127 0.4684473     2.738928 151.5645 4.269423
#> 4 0.006459779 0.2018707     2.738928 153.5636 4.953666
#> 5 0.007471156 0.5051070     1.845168 132.6452 0.000000
#> 6 0.007523981 0.5002275     1.537576 131.9400 0.000000
# }