Hello!
I have a very large experimental design space (all possible combinations of
all possible levels of several factors). For example, 'allcand' below has
1,875,000 possible combinations of 9 factors.
allcand<-expand.grid(a1=as.factor(1:5),a2=as.factor(1:5),a3=as.factor(1:5),a4=as.factor(1:5),
a5=as.factor(1:5),a6=as.factor(1:3),a7=as.factor(1:5),a8=as.factor(1:5),a9=as.factor(1:8))
dim(allcand)
My ultimate goal is to grab a subset of 10,000 out of those 1.875 million
candidates, such that the resulting 10,000 are as orthogonal as possible.
Usually, I use package "AlgDesign" for such tasks. However, my design space
is so large that it is taking too long even to grab 100 out 1,875,000 -
like that:
library(AlgDesign)
system.time(mydes<-optFederov(~.,data=allcand,nTrials=100))
# It took me on my machine 38 min.
Is there a package that could do something like this faster?
Thank you very much!
--
Dimitri Liakhovitski
[[alternative HTML version deleted]]
______________________________________________
[email protected] mailing list
https://stat.ethz.ch/mailman/listinfo/r-help
PLEASE do read the posting guide http://www.R-project.org/posting-guide.html
and provide commented, minimal, self-contained, reproducible code.
