R aggregate multiple variables using cbind

Question

I'm aggregating all of the variables in the TEF data frame by the variable TEF using the following code :

TEF2<-aggregate(cbind(column2, column3)~TEF, data=TEF, sum, na.rm=TRUE)

This works great for the variables in columns 2 and 3. However the trouble is I have 338 variables to aggregate and don't want to type out all their names.

I've tried using

TEF2<-aggregate(cbind(2:339)~TEF, data=TEF, sum, na.rm=TRUE) 

but this doesn't work, does anyone have any suggestions?

Thanks

Answer 1

Use the . notation in the formula. That takes all columns except the grouping column:

TEF <- data.frame(matrix(rnorm(100),nrow=10),TEF=1:2)

TEF2 <- aggregate(.~TEF, data=TEF, sum, na.rm=TRUE)