[R] Ranking within a classification variable.

[Liaw, Andy]

Here's one way:

> dat <- data.frame(x=runif(10), f=sample(1:2, 10, replace=TRUE))
> dat
            x f
1  0.06928792 1
2  0.55972448 1
3  0.39285280 1
4  0.72547126 1
5  0.69027628 1
6  0.51918727 1
7  0.62048040 2
8  0.42240036 1
9  0.90938702 1
10 0.16013661 2
> dat$rank <- with(dat, ave(x, f, FUN=rank))
> dat[order(dat$f, dat$x),]  ## Just checking...
            x f rank
1  0.06928792 1    1
3  0.39285280 1    2
8  0.42240036 1    3
6  0.51918727 1    4
2  0.55972448 1    5
5  0.69027628 1    6
4  0.72547126 1    7
9  0.90938702 1    8
10 0.16013661 2    1
7  0.62048040 2    2

Don't know how this rates on the slick-quick-sexy scale...

Andy


> From: Rolf Turner
> 
> 
> Suppose I have a data frame with two columns ``district'' and
> ``score'' --- score is numeric; district may be considered
> categorical.
> 
> I wish to append to this data frame a third column whose entries are
> the ranks of ``score'' ***within*** district.
> 
> I've tried fiddling about with tapply() and by() but the result is a
> list whose i-th component consists of the ranks of the scores within
> the i-th district.  I then have trouble figuring out how to put these
> results into a column of the data frame in the proper order.
> 
> Is there a slick-quick-sexy way of doing this (without resorting to
> looping)?  Am I missing something obvious?
> 
> Thanks for any help bestowed.
> 
> 				cheers,
> 
> 					Rolf Turner
> 					rolf at math.unb.ca
> 
> ______________________________________________
> R-help at stat.math.ethz.ch mailing list
> https://stat.ethz.ch/mailman/listinfo/r-help
> PLEASE do read the posting guide! 
> http://www.R-project.org/posting-guide.html
> 
> 
>