Thank you very much indeed. PD's code is exactly what I was looking for, and unfortunately it is so obvious, that I could kick myself for not thinking of it in the first place, as I feared would happen. WM: thanks for the additional suggestion. I'll check it in detail to see how my codes can benefit from it.
As a general note, I would like to express my gratitude for all contributors for their very fast and most helpful replies. For me it always made a huge difference to a) spending frustrating hours looking for the solution, knowing it is there but not finding it, and therefore b) having to work around it in a most inelegant manner. Thanks again! On 1 March 2017 15:38:15 CET, William Michels <w...@caa.columbia.edu> wrote: >Hello Wolfgang, > >Building on Peter Dalgaard's code, are you just trying to take a sample >of >a random column from each row? You don't need to use apply: > >> array[cbind(1:nrow(array), sample.int(ncol(array), nrow(array), >replace=TRUE ))] > >Just a general note, since you're sampling one-column-per-row from an >array >with more rows than columns, you'll have to set replace=TRUE. However, >there may be other datasets where you have more columns than rows and >never >want to sample each column more than once, in which case you would set >replace=FALSE. > >Best Regards. > > > >On Wed, Mar 1, 2017 at 5:38 AM, peter dalgaard <pda...@gmail.com> >wrote: > >> >> array[cbind(1:999,vector)] >> >> -pd >> >> On 01 Mar 2017, at 14:28 , Wolfgang Waser ><wa...@frankenfoerder-fg.de> >> wrote: >> >> > Dear all, >> > >> > I have to pick one value per row from an array, but from row to row >from >> > a different column. The column positions of the values for each row >are >> > stored in a vector. >> > >> > array: 999 rows, 48 columns >> > >> > vector: 999 values (each between 1 and 48) indicating for each row >which >> > value to pick from that row. >> > >> > Is there a non-loop way to pick the 999 values from the array, >probably >> > using some form of ?apply? >> > >> > >> > Thank you very much for help and suggestions! >> > >> > Wolfgang >> > >> > ______________________________________________ >> > R-help@r-project.org mailing list -- To UNSUBSCRIBE and more, see >> > https://stat.ethz.ch/mailman/listinfo/r-help >> > PLEASE do read the posting guide http://www.R-project.org/ >> posting-guide.html >> > and provide commented, minimal, self-contained, reproducible code. >> >> -- >> Peter Dalgaard, Professor, >> Center for Statistics, Copenhagen Business School >> Solbjerg Plads 3, 2000 Frederiksberg, Denmark >> Phone: (+45)38153501 >> Office: A 4.23 >> Email: pd....@cbs.dk Priv: pda...@gmail.com >> >> ______________________________________________ >> R-help@r-project.org mailing list -- To UNSUBSCRIBE and more, see >> https://stat.ethz.ch/mailman/listinfo/r-help >> PLEASE do read the posting guide http://www.R-project.org/ >> posting-guide.html >> and provide commented, minimal, self-contained, reproducible code. >> -- Sent from my touchy-wipy-thing. Please excuse the typos. [[alternative HTML version deleted]] ______________________________________________ R-help@r-project.org mailing list -- To UNSUBSCRIBE and more, see https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
