Hello all, I am having an issue with using an aggregator. My goal is to perform cross validation on several different datasets. I would like to use an aggregator to keep track of the results from a single dataset, but not across datasets. Here is the code that I have written: for(i in 1:niters) { agg<-new("aggregator") fs<-fsCV( data[,i], gfun, class[,i], agg) rm( agg ) } The problem is that every time I run the line agg<-new("aggregator") it returns the old aggregator. So I end up aggregating the results across datasets, which I do not want to do. What am I missing here? Is there a function for clearing the aggregator? I haven't been able to find one. Thanks, Cameron

Hi Richard -- Not really a pro on using aggregator, but I'm guessing that what you'd like to do is agg <- new("aggregator", aggenv=new.env(parent=emptyenv())) 'aggenv' is where the aggregated data gets stored. It's a data structure called an 'environment'. What the line above does is create a new environment each time you create a 'new' aggregator. Hope this helps. Martin "Craddock, Richard C. (CDC/CCID/NCZVED) (CTR)" <cmi5 at="" cdc.gov=""> writes: > Hello all, > > I am having an issue with using an aggregator. My goal is to perform > cross validation on several different datasets. I would like to use an > aggregator to keep track of the results from a single dataset, but not > across datasets. > > Here is the code that I have written: > > for(i in 1:niters) > { > agg<-new("aggregator") > fs<-fsCV( data[,i], gfun, class[,i], agg) > rm( agg ) > } > > The problem is that every time I run the line agg<-new("aggregator") it > returns the old aggregator. So I end up aggregating the results across > datasets, which I do not want to do. > > What am I missing here? Is there a function for clearing the > aggregator? I haven't been able to find one. > > Thanks, > Cameron > > _______________________________________________ > 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 -- Martin T. Morgan Bioconductor / Computational Biology http://bioconductor.org