Add relative frequency by factor to a data frame - r

Question

I want to add a column to a data frame that looks like this with the relative frequency by factor (Var2)
X = structure(list(Var1 = structure(c(1L, 2L, 3L, 4L, 5L, 6L, 7L,
8L, 9L, 10L, 11L, 1L, 2L, 3L, 4L, 5L, 6L, 7L, 8L, 9L, 10L, 11L
), .Label = c("0", "1", "2", "3", "4", "5", "6", "7", "8", "9",
"10"), class = "factor"), Var2 = structure(c(1L, 1L, 1L, 1L,
1L, 1L, 1L, 1L, 1L, 1L, 1L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L,
2L, 2L), .Label = c("No Treatment", "Any Treatment"), class = "factor"),
Freq = c(1L, 3L, 6L, 13L, 30L, 53L, 69L, 123L, 198L, 270L,
1324L, 1L, 0L, 4L, 10L, 16L, 33L, 44L, 75L, 113L, 159L, 630L
)), .Names = c("Var1", "Var2", "Freq"), row.names = c(NA,
-22L), class = "data.frame")
The solution that I have in mind is very complicated, and not very flexible. This is what I'm doing right now:
library(data.table)
DT =data.table(X)
myfun <- function (freq, group, total1, total2)
{
if(group[[1]] == "No Treatment"){
relfreq = freq/total1
}else{
relfreq = freq/total2
}
return(relfreq)
}
DT[,relfreq:=myfun(Freq,Var2,sum(DT$Freq[DT$Var2=="No Treatment"]), sum(DT$Freq[DT$Var2=="Any Treatment"]))]
Can somebody show me a better solution that is more flexible and allows Var2 to take more than 2 values?
Thanks!

Answer 1

Here is a data table solution, since you started out that way.
DT[,relfreq:=Freq/sum(Freq),by=Var2]
This will be faster if your dataset is extremely large, mostly because data table adds the new column by reference, rather than copying the whole dataset.

Answer 2

You can get a vector of the sum by factor with ave and divide X$Freq by this vector:
X$relfreq <- X$Freq / ave(X$Freq, X$Var2, FUN=sum)
Or even:
X$relfreq <- ave(X$Freq, X$Var2, FUN=function(x) x/sum(x))
Note that your function is incorrect, and divides each Freq by 2090 in your example, rather than dividing by the sum of the Freq of each factor level.