Hello everybody,
I noticed a performance problem when using the source-function in R.
When I try to source an r script that is located in the same directory
as the script I execute via command line ('R -f file.r' contents
'source("someOtherScript.r")') it's very fast. But if I
put
'someOtherScript.r' in another directory (say ../../R) R takes moments,
if not minutes to locate and load the script. Is there another way to do
this that is not slow?
Thanks, Jos
--
Jos van Nijnatten, BSc Candidate
Erasmus MC, Rotterdam, NL
Department of Bioinformatics
(ph) +31 1 07 04 45 51
(mob) +31 6 14 92 14 01
(e) j.vannijnatten@erasmusmc.nl
[[alternative HTML version deleted]]
On 19.01.2012 14:25, Jos van Nijnatten wrote:> Hello everybody, > I noticed a performance problem when using the source-function in R. > When I try to source an r script that is located in the same directory > as the script I execute via command line ('R -f file.r' contents > 'source("someOtherScript.r")') it's very fast. But if I put > 'someOtherScript.r' in another directory (say ../../R) R takes moments, > if not minutes to locate and load the script. Is there another way to do > this that is not slow?We are certainly interested in a reproducible example. Or could it be that your "../../R" is on some other possibly network mounted file system? Uwe Ligges> Thanks, Jos >
On 12-01-19 8:25 AM, Jos van Nijnatten wrote:> Hello everybody, > I noticed a performance problem when using the source-function in R. > When I try to source an r script that is located in the same directory > as the script I execute via command line ('R -f file.r' contents > 'source("someOtherScript.r")') it's very fast. But if I put > 'someOtherScript.r' in another directory (say ../../R) R takes moments, > if not minutes to locate and load the script. Is there another way to do > this that is not slow?R uses the same method in both instances. If there's a difference in timing, it's almost certainly because there's something going on with your OS or file system. Duncan Murdoch