An embedded and charset-unspecified text was scrubbed... Name: not available Url: https://stat.ethz.ch/pipermail/bioconductor/attachments/20051107/ 007450f4/attachment.pl

fortuitous. look at indexProbes. we use whatever order ls gives us. -r On Mon, 7 Nov 2005, Jamain, Adrien J wrote: > Dear BioC people, > > I have a little technical question about how the order of the probesets > in an exprSet is defined. > Is this order given by the order of the probesets in the corresponding > cdf environment (e.g. ls(mouse4302cdf))? > Or is it alphabetical? > Or is it "unconstrained"? > > In the examples I've looked at both the order of the probesets in the > cdf environment and the exprSets are the same (and alphabetical). > > But is it a characteristic which is desired by development or is it just > fortuitous? > > Thank you in advance for your insight... > Adrien Jamain > > > [[alternative HTML version deleted]] > > _______________________________________________ > Bioconductor mailing list > Bioconductor at stat.math.ethz.ch > https://stat.ethz.ch/mailman/listinfo/bioconductor >

> Dear BioC people, > > I have a little technical question about how the order of the probesets > in an exprSet is defined. > Is this order given by the order of the probesets in the corresponding > cdf environment (e.g. ls(mouse4302cdf))? > Or is it alphabetical? > Or is it "unconstrained"? > > In the examples I've looked at both the order of the probesets in the > cdf environment and the exprSets are the same (and alphabetical). There is probably no "order" in a cdf environment (in the sense people think of it). This is a hash table, and the order in which elements are returned depends on the implementation of it, and eventually on the keys in the hash table. I would suspect that R returns keys in alphabetical order (but I am not sure of that, and therefore never -or should not- assume it blindly in my code). Having the same order (at least on the same platform) is implicit if you used 'expresso' (or wrappers) without specifying a list of probe set IDs to work with: an exprSet is then obtained by looping through all keys in the cdfenv (and in the order these keys are returned, AFAIR). Hoping this helps, L. > But is it a characteristic which is desired by development or is it just > fortuitous? > > Thank you in advance for your insight... > Adrien Jamain > > > [[alternative HTML version deleted]] > > _______________________________________________ > Bioconductor mailing list > Bioconductor at stat.math.ethz.ch > https://stat.ethz.ch/mailman/listinfo/bioconductor >