[kth-package] Re: SD?(standard devation)
- From: Valtteri Wirta <valtteri@xxxxxxxxxxxxxx>
- To: kth-package@xxxxxxxxxxxxx
- Date: Thu, 17 Nov 2005 14:12:30 +0100
Hi again,
A late answer is better than no answer.....
First you need the SD values (assume you have them in the sd.vec
vector)... and assuming they are in the same order as your gpr/ma
objects, then do this:
idx.remove <- which( sd.vec > YourSDCutOff ) # YourSDCutOff is the
cutoff value for SD that you want to use
... then kick out the data:
numSlides <- nbrOfSlides(ma) # ma is your ma-object
ma.obj$M[idx.remove, 1:numSlides] <- NA # removes genes that have SD
larger than your cutoff by setting their value to NA
OR
ma.obj$M[idx.remove, slideVec] <- NA # removes genes that have SD
larger than your cutoff by setting their value to NA and only for
those slides that are defined in slideVec
regards,
Valtteri
At 23:51 2005-10-30, you wrote:
Hi,
I wonder how to write if you want to remove genes from slides which have a
SD-value over a given SD-value?
H/Monica
****************************************************************************
Mailing list for KTH-package. For more information see:
www.biotech.kth.se/molbio/microarray
To change the settings or unsubscribe:
//www.freelists.org/cgi-bin/list?list_id=kth-package
To send mail to the list use: kth-package@xxxxxxxxxxxxx
To get information on how the mailing list works, send an empty
e-mail with the word 'faq' to kth-package-request@xxxxxxxxxxxxx
***************************************************************************
****************************************************************************
Mailing list for KTH-package. For more information see: www.biotech.kth.se/molbio/microarray
To change the settings or unsubscribe: //www.freelists.org/cgi-bin/list?list_id=kth-package
To send mail to the list use: kth-package@xxxxxxxxxxxxx
To get information on how the mailing list works, send an empty e-mail with the word 'faq' to kth-package-request@xxxxxxxxxxxxx
***************************************************************************
Other related posts:
