Frequency table by group in data.table (R)

Question

I want to get for each group the frequencies of the values of a factor/categorial variable.

The following does not work:

library(data.table)
dt<-data.table(fac=c("l1","l1","l2"),grp=c("A","B","B"))
dt[,fac:=as.factor(fac)]
dt[,list( table(fac) ),by=grp]

The error message is:

Error in `[.data.table`(dt, , list(table(fac)), by = grp) : 
  All items in j=list(...) should be atomic vectors or lists. If you are trying something like j=list(.SD,newcol=mean(colA)) then use := by group instead (much quicker), or cbind or merge afterwards.

Is there an easy way to accomblish this task? Thanks.

`dt[,as.list(table(fac)),by=grp]` maybe – David Arenburg Aug 14 '16 at 11:13 — David Arenburg, Aug 14 '16 at 11:13
Oh wow. That's it! Thank you :) – Funkwecker Aug 14 '16 at 12:14 — Funkwecker, Aug 14 '16 at 12:14
Or `with(dt, table(grp, fac))`... – Frank Aug 14 '16 at 14:32 — Frank, Aug 14 '16 at 14:32

akrun · Answer 1 · 2016-08-14T12:31:54.390

4

We can use dcast and bypass the second and third line of OP's code.

dcast(dt, grp~fac, length)
#   grp l1 l2
#1:   A  1  0
#2:   B  1  1

edited Aug 14 '16 at 12:31

answered Aug 14 '16 at 12:25

akrun

874,273
37
540
662

1

for dcast function, load firstly data.table library: require("data.table") – maycca Apr 24 '17 at 13:46
@maycca You are right. As the OP's post showed the loading of library, I didn't show it – akrun Apr 24 '17 at 14:28

Frequency table by group in data.table (R)

1 Answers1