[BioC] weird problem with ReadAffy and makecdfenv'd CDF package

[Tim]

Sorry to bother the list with what is probably a dumb question.

I looked in the list archives and online, and in the code for the
routines ReadAffy and read.affybatch, and for the life of me, I
can't figure it out.  I used library(makecdfenv) to create an
alternative CDF environment where I mask out some probes in a
probeset, and I try to load it into an affyBatch, and I get...

> library(PHYH)
> ReadAffy(verbose=TRUE,cdfname=PHYH)
1 reading /home/tim/Hacia/cb1.cel ...instantiating an AffyBatch
(intensity a 1354896x55 matrix)...done.
Reading in : /home/tim/Hacia/cb1.cel
Reading in : /home/tim/Hacia/cb2.cel
Reading in : /home/tim/Hacia/cb3.cel
...
Reading in : /home/tim/Hacia/ht6.cel
Error in as.character(args[[i]]) : cannot coerce to vector
>


Any suggestions?  I can't figure out what's going on here.  I have
loaded plenty of affyBatch'es of CEL files before, but never with an
alternatively masked CDF environment... when I do it like so,

> ReadAffy(verbose=TRUE)
1 reading /home/tim/Hacia/cb1.cel ...instantiating an AffyBatch
(intensity a 1354896x55 matrix)...done.
Reading in : /home/tim/Hacia/cb1.cel
Reading in : /home/tim/Hacia/cb2.cel
Reading in : /home/tim/Hacia/cb3.cel
...
* Installing *source* package 'hgu133plus2cdf' ...
** R
** data
** help
 >>> Building/Updating help pages for package 'hgu133plus2cdf'
     Formats: text html latex example
  geometry                          text    html    latex   example
  hgu133plus2cdf                    text    html    latex
** building package indices ...
* DONE (hgu133plus2cdf)

The downloaded packages are in
        /tmp/RtmpIMTi6e/downloaded_packages

Attaching package: 'hgu133plus2cdf'


        The following object(s) are masked from package:PHYH :

         i2xy

        The following object(s) are masked from package:PHYH :

         xy2i

AffyBatch object
size of arrays=1164x1164 features (582203 kb)
cdf=HG-U133_Plus_2 (54675 affyids)
number of samples=55
number of genes=54675
annotation=hgu133plus2

... it works OK.  So should I unpack the hgu133plus2cdf package,
mask out probes (merge into a custom CDF), and then use that?

Seems like I am missing something obvious here.  Need more eyes!


Much obliged, thank you in advance for your helpful advice,

 --tim


-- 
"People seldom do what they believe in.  They do what is convenient,
then repent." 
--Bob Dylan