Question

I have a dataframe with 100 variables, of which I would like a subset, say dataframename[,30:50] to be converted to their original numeric values (1,2,3,4,5).

I know that I should use as.numeric(levels(f))[f] when I convert a factor, but I can only make this work when I convert the factors one at a time. I would like to convert them all at once.

This wont work:

as.numeric(levels(dataframename[,30:50]))[dataframename[,30:50]]  

neither will this:

sapply(dataframename[,30:50],as.numeric(levels(dataframename[,30:50]))    
[dataframename[,30:50]]  

Any ideas or stuff I should read?

Answer 1

This is a smaller example but the idea should hold. You can use lapply to apply your conversion to each column of your data frame and then just replace those columns directly.

# make example data
dat <- as.data.frame(lapply(as.data.frame(matrix(seq(2*3), ncol = 3)), factor))

factorconvert <- function(f){as.numeric(levels(f))[f]}
dat[, 2:3] <- lapply(dat[, 2:3], factorconvert)
dat
#  V1 V2 V3
#1  1  3  5
#2  2  4  6
#str(dat)
#'data.frame':  2 obs. of  3 variables:
# $ V1: Factor w/ 2 levels "1","2": 1 2
# $ V2: num  3 4
# $ V3: num  5 6