1

When I clear the Global environment in R, and then invoke garbage collection, R continues to hog memory (see code below). Is there a way to clean up memory "even more" without restarting R?

> rm(list = ls())
> gc()
          used (Mb) gc trigger   (Mb)   max used    (Mb)
Ncells 1352699 72.3    2419890  129.3    3886542   207.6
Vcells 4231877 32.3  910657255 6947.8 1412741750 10778.4
> unloadlibs() # wrapper to remove all packages in 'sessionInfo()$otherPkgs'
[1] "Unloaded packages:"
 [1] "magrittr"   "zoo"        "reshape"    "stringi"    "openxlsx"   "lubridate"  "dtplyr"     "dplyr"      "RODBC"     
[11] "data.table"
> gc()
          used (Mb) gc trigger   (Mb)   max used    (Mb)
Ncells 1349320 72.1    2419890  129.3    3886542   207.6
Vcells 4230328 32.3  728525804 5558.3 1412741750 10778.4

In this case, there is still more than 700mb of memory used after cleaning everything.

Is it a known issue, or am I doing something wrong (in this case, what is it)? What may be the culprit and faulty code that may result in this? I already :

  • Dropped dplyr usage,
  • Try to use data.table in any case possible,
  • And use copy() in the code when I am afraid that data.table references will somehow confuse garbage collection (not sure how a single reference to a data.table will behave).

Thanks for your inputs.

Additional info:

> R.version

platform       x86_64-w64-mingw32          
arch           x86_64                      
os             mingw32                     
system         x86_64, mingw32             
status                                     
major          3                           
minor          4.3                         
year           2017                        
month          11                          
day            30                          
svn rev        73796                       
language       R                           
version.string R version 3.4.3 (2017-11-30)
nickname       Kite-Eating Tree
Konrad Rudolph
  • 530,221
  • 131
  • 937
  • 1,214
itarill
  • 323
  • 1
  • 14
  • What is Ctrl+Shift+F10 supposed to do? This doesn’t mean anything in the context of pure R. Anyway, to answer your question it’s kind of necessary to know what you did in the R session *before* the `gc()` call. – Konrad Rudolph Apr 04 '19 at 09:52
  • 3
    @KonradRudolph It's RStudio's keyboard shortcut to restart R. I have added the rstudio tag. – Rui Barradas Apr 04 '19 at 09:53
  • @RuiBarradas Thanks. However, with your permission I’ll remove that tag again and fix the text, since the question doesn’t seem to be specific to RStudio. – Konrad Rudolph Apr 04 '19 at 09:54
  • @KonradRudolph Ok, it seems to be an R question, not a RStudio one. – Rui Barradas Apr 04 '19 at 09:55
  • @KonradRudolph: A lot of data wrangling actually- merging, subsetting rows and columns. It is indeed duplicate (I could not find that "original" topic on my own), I think the phenomenon is indeed related to the lack of memory compaction of R. – itarill Apr 04 '19 at 11:15

0 Answers0