Dear Laurent, Thanks for the email. I got the problem solved with help from Wolfgang. It works great. I am getting more familiar with the packages now. Some the docs are a little bit confusing, but still pretty good. If somebody could put together a tutorial, that would be even better. I like these packages very much and will continue to explore. I am posting the response from Wolfgang in this list in case somebody else has similar questions. Best regards -hui ********************************************** On Wed, 12 Mar 2003, Wang, Hui wrote: > Dear Wolfgang, > Thanks for your reply. I tried as you suggested. I sucessfully made > the cdf file. However, I could not find "testcdf" package on the web > site. I still get errors. With "make.cdf.package", you actually make your own package! I suggested you call it "testcdf" since that what seems that be mentioned in your cel file (from your earlier post). Now there is one catch: makeing packages is easy on a linux/unix system, but requires some additional installation of tools under windows. So, if you use windows, it depends how serious you are about using R. If you are you using windows, you may just want yo use the "make.cdf.env" function from the makecdfenv package, for example: res = make.cdf.env("my.cdf", cdf.path="C:/temp") testcdf = res$env This will create an environemt (=hastable) in your current workspace and ReadAffy should work with this. Please make sure to give it the proper name as mentioned in the CEL files. Best Wolfgang > -----Original Message----- > From: Laurent Gautier [mailto:laurent@cbs.dtu.dk] > Sent: Wednesday, March 12, 2003 7:31 PM > To: Wang, Hui > Cc: 'bioconductor@stat.math.ethz.ch' > Subject: Re: [BioC] Questions on getting information from affy's CEL, CDF or 1lq file s > > On Wed, Mar 12, 2003 at 02:14:37PM -0800, Wang, Hui wrote: > > Hi, > > > > I am trying to use read.affybatch, however encountering the below error. > > > > Since the cdf I was using doesnot belong to any of the standard ones, could > > that be the problem? If yes, how do I attach my own cdf? > > > > Thanks > > > > -hui > > > > > > > cdf.env<-read.cdffile("test.CDF") > > >read.affybatch("030228_01_GL.CEL", "030228_02_GL.CEL") > > AffyBatch object > > size of arrays=712x712 features (7924 kb) > > Error in .local(object, ...) : Information about probe locations for test > > could not be found. Try downloading the testcdf package from > > http://www.bioconductor.org/data/cdfenvs/cdfenvs.html > > In addition: Warning message: > > Incompatible phenoData object. Created a new one. > > in: read.affybatch("030228_01_GL.CEL", "030228_02_GL.CEL") > > cdf=test (0 affyids) > > number of samples=2 > > Error in .local(object, ...) : Information about probe locations for test > > could not be found. Try downloading the testcdf package from > > http://www.bioconductor.org/data/cdfenvs/cdfenvs.html > > > > > > > Dear Hui, > > > The function 'read.cdffile' reads .CDF files, and store the results > in an object of class Cdf (doing 'help(read.cdffile)' tells it, and > even more, like to refer to the doc for 'Cdf' to know more). > > The doc for read.affybatch leads to the one for AffyBatch, in which > I admit things could be clearer (you will however note that it points > to the vignette, in which things are (a bit) clear(er)). > > The class 'Cdf' and the cdf environments are two different structures. > > For what you seems wanting to do (just guessing), you may have to > work with both structures at the same time. For example, the Cdf are > better for dealing with what exactly is on a Cdf. In your previous > post, the neat way to get 'non-control/non-qc' probes is probably to > use the fuction 'pmormm'. You may also want to generate cdfenvironments > a quick way. Check the function 'getLocationsData.Cdf' for that. > > > Hopin' it helps, > > > > L. > > > > (the environments could also have their own class soon (if no one is fast enough > to stop me ;) ) > > > > > -- > -------------------------------------------------------------- > currently at the National Yang-Ming University in Taipei, Taiwan > -------------------------------------------------------------- > Laurent Gautier CBS, Building 208, DTU > PhD. Student DK-2800 Lyngby,Denmark > tel: +45 45 25 24 89 http://www.cbs.dtu.dk/laurent [[alternate HTML version deleted]]