[BioC] readGenericHeader problem in limma 2.7.8?

[Marcus Davy]

Relating to your question,

> I don't know how to attempt an install of 1.7 to see if that affects
> anything. 

You want to use limma which does not have any critical package dependencies
so you don't need BioC 1.7 unless you want to use other classes and methods
than that provided in limma.

Using OS X, to examine different versions of limma quickly, one option is to
reinstall a recent previous R.app dmg package major release of R (e.g.
Compare limma between R-2.2.1, and R-2.3.1) at the system level and install
a previous version of limma that you think works to test if it is a limma
version problem. Then you can just quickly change R versions to examine
different limma versions on a test script. Reinstallation does not effect
any contributed packages you have previously installed for a particular
version of R which are located in;

/Library/Frameworks/R.framework/Versions/Current/Resources/library
(Current directory is a symbolic pointer)

You just need to make sure that the contributed package your are installing
(limma version) is compatible across the recent major versions of R you want
to test with. If you were installing lots of packages this would not be
recommended, but since it is only one which has no critical dependencies you
should be ok.

The releases of R will go into;

ls -la /Library/Frameworks/R.framework/Versions/
drwxrwxr-x   9 root  admin   306 Jul 27 16:44 .
drwxrwxr-x   8 root  admin   272 Jul 27 16:44 ..
-rw-rw-r--   1 root  admin  6148 Oct 17  2005 .DS_Store
lrwxr-xr-x   1 root  admin     5 Aug 16  2005 2.1 -> 2.1.1
lrwxr-xr-x   1 root  admin     5 Aug 16  2005 2.1.0 -> 2.1.1
drwxrwxr-x   4 root  admin   136 Dec 14  2005 2.1.1
drwxrwxr-x   4 root  admin   136 Jul 27 16:44 2.2
drwxrwxr-x   6 root  admin   204 Jul 27 16:44 2.3
lrwxr-xr-x   1 root  admin     3 Jul 27 16:44 Current -> 2.3

The 'Current' symbolic pointer tells you which version is currently working.
So if you have a previous installation working (where 'Current' points to
say version 2.2) and the latest version of R is installed but not pointed
to, then it *should* just be a case of changing the symbolic link of
'Current' back to 2.3 again, e.g.

sudo -s
#[PASSWORD]
rm /Library/Frameworks/R.framework/Versions/Current
ln -s 2.3 Current

but I have found that reinstallation of that particular R.app dma version is
required sometimes (when the executable is not found, problem with some DMG
installations perhaps?). Once again this does not effect any previously
contributed package libraries you have installed for that R version, as long
as you haven't specifically deleted them.


In R use 'version', and 'sessionInfo()'to gather information about the
current R session if errors still occur.


Marcus


On 8/22/06 7:12 AM, "Bobby Prill" <rprill at jhu.edu> wrote:

> There seems to be a problem with read.maimages() in the current
> release version of bioconductor (1.8).  I need load a "generic" gpr-
> like file using the column headers.
> 
> The following code used to work, but now is broken:
> 
> library(limma)
> cols = list(Rf="F635 Median",Gf="F532 Median",Rb="B635
> Median",Gb="B532 Median")
> RG = read.maimages("mydata.gpr", columns=cols)
> 
> Is this a bug?  How can I get around this problem?
> 
> I have tried
> limma version 2.7.8 and 2.7.9
> 
> I'm on a Mac OSX 10.4.7
> Running R.app 1.16
> 
> The error message is:
> Error in readGenericHeader(fullname, columns = columns, sep = sep) :
>          Specified column headings not found in file
> In addition: Warning message:
> input string 1 is invalid in this locale in: grep(pattern, x,
> ignore.case, extended, value, fixed, useBytes)
> 
> Even though it says it can't find the headers, I know they are
> there.  This error started to occur when I installed a fresh copy of
> Bioconductor 1.8 after my disk crashed.  I was probably running 1.7
> before that.
> 
> I don't know how to attempt an install of 1.7 to see if that affects
> anything.
> 
> Thanks for the help.
> 
> _______________________________________________
> Bioconductor mailing list
> Bioconductor at stat.math.ethz.ch
> https://stat.ethz.ch/mailman/listinfo/bioconductor
> Search the archives:
> http://news.gmane.org/gmane.science.biology.informatics.conductor


______________________________________________________

The contents of this e-mail are privileged and/or confidenti...{{dropped}}