You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
If you used either R or R CMD BATCH for ENTRYPOINT instead of RScript (you may have to pass some additional arguments with CMD) then your R log files would be the stdout that you see in the R Console when using R.
Currently users are recording log files using sink() and those log files are much harder to tell where a script is failing if you get an error message.
I have used a work around where my project.yaml action calls a 1 line R script which calls the script I want the nicer log file for, the 1 line R script being
I'm open for changing this, now we can easily do so. The only problem I can see is if it breaks b/w compatibility with old invocations of the R image. AFAIK, all run: commands are either r:latest script.R or r:latest -e ..., both of which I think would work with either R or Rscript.
I think we should definitely consider this for next version of R image.
A small point for consideration.
If you used either
R
orR CMD BATCH
forENTRYPOINT
instead ofRScript
(you may have to pass some additional arguments withCMD
) then your R log files would be the stdout that you see in the R Console when using R.Currently users are recording log files using
sink()
and those log files are much harder to tell where a script is failing if you get an error message.I have used a work around where my project.yaml action calls a 1 line R script which calls the script I want the nicer log file for, the 1 line R script being
system("R CMD BATCH --no-save --no-restore ./analysis/2-basic-rd.R ./logs/2-basic-rd.log")
The text was updated successfully, but these errors were encountered: