tags:

views:

36

answers:

1

Using R's builtin ToothGrowth example dataset, this works:

ddply(ToothGrowth, .(supp,dose), function(df) mean(df$len)) 

But I would like to have the subsetting factors be variables, something like

factor1 = 'supp'
factor2 = 'dose'
ddply(ToothGrowth, .(factor1,factor2), function(df) mean(df$len)) 

That doesn't work. How should this be done?

I thought perhaps something like this:

factorCombo = paste('.(',factor1,',',factor2,')', sep='') 
ddply(ToothGrowth, factorCombo, function(df) mean(df$len)) 

But it doesn't work either. I think I am close, but not sure the proper way to do it. I suppose the entire command could be put into a string, followed by an eval() call of the string, but hopefully is there a more elegant way?

+1  A: 

Try:

x <- c("supp", "dose") 
ddply(ToothGrowth, x, function(df) mean(df$len)) 
Brandon Bertelsen
thanks! and to use exactly the format I was asking about, the first line would be: x <- c(factor1, factor2)
Alex Holcombe
No problem! ddply is godly, once you get it you can pretty much do anything you want.
Brandon Bertelsen