Hi James, You can use basename and dirname to pull these apart. Try: a<-"c:/progra~1/test" basename(a) dirname(a) sub(dirname(a), "", a) sub(paste(dirname(a), "/", sep=""), "", a) Would this do it? Aedin Culhane -----Original Message----- From: bioconductor-bounces@stat.math.ethz.ch [mailto:bioconductor-bounces@stat.math.ethz.ch]On Behalf Of James MacDonald Sent: 11 February 2004 17:53 To: yenlin@stanford.edu; bioconductor@stat.math.ethz.ch Subject: Re: [BioC] Edit sample Names I am sure you could extract the sampleNames, hit them with sub and then put them back in, but I don't know regular expressions well enough to tell you how to do it (I just tried and failed miserably). My ugly hack would be to do something like this: filenames <- list.files(pattern="\\.[cC][eE][lL]$") #Gets all .cel or .CEL files in current directory Then, if you used all these files you can do: sampleNames(eset) <- filenames # Where eset is the exprSet (or AffyBatch) you need to fix. otherwise you can subset the filenames to choose the ones you used, e.g., filenames[c(1,3,4,5,6)] or whatever. My preferred way to set up the AffyBatch is to use the above list of filenames in the call to ReadAffy so you don't have the long names: Data <- ReadAffy(filenames=filenames) HTH, Jim James W. MacDonald Affymetrix and cDNA Microarray Core University of Michigan Cancer Center 1500 E. Medical Center Drive 7410 CCGC Ann Arbor MI 48109 734-647-5623 >>> Yen Lin Chia <yenlin@stanford.edu> 02/10/04 06:25PM >>> Hi, When I do Data<-ReadAffy(widget=TRUE), the sample names appeared to be what I want, i.e. 10E, 2D, etc. But, after the data are read, when I type sampleNames(Data), the sample names is the whole path of the file, like c:\my documents\....\....CEL, the names are simply too long for histogram. Anybody know how to fix the sample names? Thanks. Yen Lin _______________________________________________ Bioconductor mailing list Bioconductor@stat.math.ethz.ch https://www.stat.math.ethz.ch/mailman/listinfo/bioconductor _______________________________________________ Bioconductor mailing list Bioconductor@stat.math.ethz.ch https://www.stat.math.ethz.ch/mailman/listinfo/bioconductor

Hi James Yes you are right, its a forward and back lash effect, > a <- "c:\my documents\otherdirs\celfile.CEL" > a [1] "c:my documentsotherdirscelfile.CEL" > a <- "c:/my documents/otherdirs/celfile.CEL" > basename(a) [1] "celfile.CEL" Aedin -----Original Message----- From: bioconductor-bounces+aedin.culhane=ucd.ie@stat.math.ethz.ch [mailto:bioconductor-bounces+aedin.culhane=ucd.ie@stat.math.ethz.ch]On Behalf Of James MacDonald Sent: 11 February 2004 18:16 To: yenlin@stanford.edu; bioconductor@stat.math.ethz.ch; aedin.culhane@ucd.ie Subject: RE: [BioC] Edit sample Names It would on a *nix box, but not on win32. Note that the names are c:\my documents\otherdirs\celfile.CEL > a <- "c:\my documents\otherdirs\celfile.CEL" > dirname(a) [1] "." > basename(a) [1] "c:my documentsotherdirscelfile.CEL" Blame it on Bill G. ;-D Best, Jim James W. MacDonald Affymetrix and cDNA Microarray Core University of Michigan Cancer Center 1500 E. Medical Center Drive 7410 CCGC Ann Arbor MI 48109 734-647-5623 >>> Aedin <aedin.culhane@ucd.ie> 02/11/04 01:06PM >>> Hi James, You can use basename and dirname to pull these apart. Try: a<-"c:/progra~1/test" basename(a) dirname(a) sub(dirname(a), "", a) sub(paste(dirname(a), "/", sep=""), "", a) Would this do it? Aedin Culhane -----Original Message----- From: bioconductor-bounces@stat.math.ethz.ch [mailto:bioconductor-bounces@stat.math.ethz.ch]On Behalf Of James MacDonald Sent: 11 February 2004 17:53 To: yenlin@stanford.edu; bioconductor@stat.math.ethz.ch Subject: Re: [BioC] Edit sample Names I am sure you could extract the sampleNames, hit them with sub and then put them back in, but I don't know regular expressions well enough to tell you how to do it (I just tried and failed miserably). My ugly hack would be to do something like this: filenames <- list.files(pattern="\\.[cC][eE][lL]$") #Gets all .cel or .CEL files in current directory Then, if you used all these files you can do: sampleNames(eset) <- filenames # Where eset is the exprSet (or AffyBatch) you need to fix. otherwise you can subset the filenames to choose the ones you used, e.g., filenames[c(1,3,4,5,6)] or whatever. My preferred way to set up the AffyBatch is to use the above list of filenames in the call to ReadAffy so you don't have the long names: Data <- ReadAffy(filenames=filenames) HTH, Jim James W. MacDonald Affymetrix and cDNA Microarray Core University of Michigan Cancer Center 1500 E. Medical Center Drive 7410 CCGC Ann Arbor MI 48109 734-647-5623 >>> Yen Lin Chia <yenlin@stanford.edu> 02/10/04 06:25PM >>> Hi, When I do Data<-ReadAffy(widget=TRUE), the sample names appeared to be what I want, i.e. 10E, 2D, etc. But, after the data are read, when I type sampleNames(Data), the sample names is the whole path of the file, like c:\my documents\....\....CEL, the names are simply too long for histogram. Anybody know how to fix the sample names? Thanks. Yen Lin _______________________________________________ Bioconductor mailing list Bioconductor@stat.math.ethz.ch https://www.stat.math.ethz.ch/mailman/listinfo/bioconductor _______________________________________________ Bioconductor mailing list Bioconductor@stat.math.ethz.ch https://www.stat.math.ethz.ch/mailman/listinfo/bioconductor _______________________________________________ Bioconductor mailing list Bioconductor@stat.math.ethz.ch https://www.stat.math.ethz.ch/mailman/listinfo/bioconductor