[BioC] CPU time of Anova in R

Arne.Muller@sanofi-aventis.com Arne.Muller at sanofi-aventis.com
Mon Jul 11 13:38:24 CEST 2005


Marcin,

the only dramatic speedup I can think about is to use limma instead of R's lm. Otherwise 20 min for 45k anovas in R is not too bad, isn't it?

	kind regards,

	Arne

-----Original Message-----
From: bioconductor-bounces at stat.math.ethz.ch
[mailto:bioconductor-bounces at stat.math.ethz.ch]On Behalf Of Marcin
Piechota
Sent: Monday, July 11, 2005 12:19
To: bioconductor at stat.math.ethz.ch
Subject: [BioC] CPU time of Anova in R


Hi

I have problem with very long lasting time of anova in R.

My script:
Exprs <- exprs(eset)
Anno <- geneNames(eset)
scheme <- gl(4,3,12, labels=c("a", "b","c", "d"))
x<-matrix(1:90202,nrow=45101, ncol=2) 
for (i in 1:45101) {
Data = Exprs[i,]
results <- lm(Data~schemat)
nextstep <- anova(results)
x[i,] <- c(Anno[i],nextstep[1,5])
}

where "eset' is rma normalized affy data from m430v2

this script lasts for about 20 minutes
on P4 3GHz, 4GB RAM

How to make it shorter?

Marcin Piechota
Institute of Pharmacology PAS Krakow
	[[alternative HTML version deleted]]

_______________________________________________
Bioconductor mailing list
Bioconductor at stat.math.ethz.ch
https://stat.ethz.ch/mailman/listinfo/bioconductor



More information about the Bioconductor mailing list