RE: pm or mm methods for AffyBatch object
1
0
Entering edit mode
Wang, Hui ▴ 170
@wang-hui-219
Last seen 9.6 years ago
Dear list, It turned out the slowness and out of memory has nothing to do with bioconductor package. I tested it by the following: test.intensity <- test@exprs #where test.intensity is a 0.5 miliion x 5 matrix plot(data.frame(test.intenisty)) # scatter plot for replicate comparison. it took >10 mins after I increased the memory size. (I was running the window version of R). As you can see that have nothing to do with biocondunctor package. Best regards -h -----Original Message----- From: Wang, Hui Sent: Saturday, March 15, 2003 1:35 PM To: bioconductor@stat.math.ethz.ch Subject: pm or mm methods for AffyBatch object Hi List, I have two questions: 1) I have a chip which is the same feature as u133A, however have much less mm probes. When using pm or mm method, the matrices have the same dimension. I wonder, do the methods assume the number of probes for pm or mm are the sames by default? How do you determine pm or mm probes? 2) The cdf I am using is the same size as u133A(I wish I could contribute my cdf), I only read in 5 chips. However running into memory problem quickly. For example, I did a boxplot on 5 chips, it gave memory error (I allocated 1G) after 10mins of plotting. Could somebody give me some tips besides breaking down the files (since I only have 5 files)? Best regards -h [[alternate HTML version deleted]]
cdf cdf • 778 views
ADD COMMENT
0
Entering edit mode
@wolfgang-huber-3550
Last seen 3 months ago
EMBL European Molecular Biology Laborat…
Hi, > > test.intensity <- test@exprs #where test.intensity is a 0.5 miliion x 5 > matrix > plot(data.frame(test.intenisty)) # scatter plot for replicate comparison. > it took >10 mins after I increased the memory size. (I was running the > window version of R). As you can see that have nothing to do with > biocondunctor package. Plotting a scatter plot with half a million points will certainly take forever; if you tried to save the plot or copy/paste it across to other applications, you would also see that it becomes huge! And finally, you will need see very much except for a big black mess. Consider: - using the hexbin package - subsampling, i.e. plot(test.intensity[sample(nrow(intensity), 1e4), ]) Cheers Wolfgang
ADD COMMENT

Login before adding your answer.

Traffic: 865 users visited in the last hour
Help About
FAQ
Access RSS
API
Stats

Use of this site constitutes acceptance of our User Agreement and Privacy Policy.

Powered by the version 2.3.6