[BioC] Kooperberg

Matthew Ritchie mritchie at wehi.edu.au
Thu Sep 2 02:18:50 CEST 2004


Hi Jason,

Are the grids in your .gpr file all of the same size as specified in the 
layout argument?  The kooperberg function assumes this and will produce 
an error like the one you are getting if they aren't.  You might also 
have a problem doing print-tip loess normalization on this data if the 
layout argument and number of spots on the array don't agree.  If this 
isn't the problem, perhaps you can send me one of your .gpr files and 
I'll take a more careful look.  Best wishes,

Matt

Jason Skelton wrote:

> Hi all
> I guess this question is mainly for matt richie but if anyone can tell 
> me what I'm doing wrong...
> been using kooperberg correction for quite a while and had alot of 
> success with it
> however with some new data I recieve the following error:
>
> > RG <- kooperberg(FILES, fg="median", bg="median", layout=LAYOUT)
> Error in .varaux2(x, i, j, ncols, nrows) :
>        subscript out of bounds
> In addition: Warning message:
> data length [118] is not a sub-multiple or multiple of the number of 
> rows [8] in matrix
>
> doesn't look like a problem with limma/kooperberg so its more likely 
> to be my files however
> they read perfectly fine with read.maimages
> had a look at the gpr files and nothing is obviously wrong
> any ideas would be greatfully recieved
>
> using R (1.9.0)
> limma(1.6.7)



More information about the Bioconductor mailing list