0

I am trying to reproduce the example of Spatial interpolation / prediction using Ensemble Machine Learning

for which I need to use the "mlr" package I have installed it and everything is fine, but when trying to load it I have the following error:

> library(mlr)
#Error: package or namespace load failed for ‘mlr’ in get(Info[i, 1], envir = env):cannot allocate memory block of size 2.5 Gb

I understand that this is because my system is 32 bits (plattform: i386-w64-mingw32 / i386) it is also worth mentioning that I use version 4.1.1 of R, Rstudio Version 1.0.153

What I don't understand is that when I see the memory limit with:

> memory.limit()
[1] 4095

which is clearly higher than 2.5 Gb and yet I can't continue to load the package

and modifying this limit is not possible due to my 32-bit system as I have investigated and in the same way when I try I get this:

> memory.limit(5000)
Error in memory.limit(5000) : 
  don't be silly!: your machine has a 4Gb address limit

know how to make this package work?

dovat_
  • 1
  • 3
  • Have you tried garbage collection? Run `gc()` and try again. – Kat Aug 30 '21 at 04:08
  • 32-bit support will be dropped soon and has many memory restrictions. Also RStudio 1.0 is a few years old and might have some issues with the recent R version eventually. It is hard to help you here as most people won't have access to a 32-bit machine anymore. Make sure you are in a new session, the package should be loadable at least. However, you might run into troubles doing the linked analysis then even if you get the package to load. – pat-s Aug 30 '21 at 06:31

0 Answers0