Adding new variable to dataframe equally

Question

using RStudio, I have this :

GROUP NUM
A     45
A     78
A     79  
B     45
B     47
B     99
C     28
C     78
C     54

I want to add a new variable, named AGENT, which is:

AGENT=c("John", "Maria", "Pamela")

But the problem is that, I want each of my Agent to be equally spreaded amongsts the initial dataframe according to the ID. Basically, I want this:

GROUP NUM AGENT
A     45  John
A     78  Maria
A     79  Pamela
B     45  John 
B     47  Maria
B     99  Pamela
C     28  John
C     78  Maria
C     54  Pamela

My example here is basic because I have as many groups as I have agents. However in my case, I might have 70 of each letter (70 A, 70 B and 70 C) and still only 3 agents. I still want them to be spread as equally as possible....

For example, if I had 6 A, I would have :

GROUP NUM AGENT
A     45  John
A     78  Maria
A     79  Pamela
A     48  John
A     97  Maria
A     59  Pamela
...

And if I had 7, then 7th would be assigned randomly, or just the next on the list.

Any ideas? I've been torturing myself over this. Thanks in advance! :P

Answer 1

If "or just the next on the list." is appropriate for any overflow when the group is larger, you can take advantage of vector recycling and just do it in one assignment:

dat$newvar <- with(dat, ave(1:nrow(dat), GROUP, FUN=function(x) AGENT) )
dat
#  GROUP NUM newvar
#1     A  45   John
#2     A  78  Maria
#3     A  79 Pamela
#4     B  45   John
#5     B  47  Maria
#6     B  99 Pamela
#7     C  28   John
#8     C  78  Maria
#9     C  54 Pamela

Just ignore any warnings you might get when the groups are not neatly matched to the size of AGENT

data.table could be used too, in a similar fashion:

library(data.table)
setDT(dat)
dat[, newvar2 := AGENT, by=GROUP]