[R] group by in data.frame

[David Winsemius]

On Feb 25, 2011, at 10:14 AM, zem wrote:

>
> Yeah, you are right
> i want to post an short example what i want to do .. and in the  
> meantime i
> solved the problem ...
> but here is:
> i have something like this dataframe:
> c1<-c(1,2,3,2,2,3,1,2,2,2)
> c2<-c(5,6,7,7,5,7,5,7,6,6)
> c3<-rnorm(10)
> x<-cbind(c1,c2,c3)
>> x
>      c1 c2          c3
> [1,]  1  5  0.08279036
> [2,]  2  6  0.59135988
> [3,]  3  7  1.45520468
> [4,]  2  7 -1.70094640
> [5,]  2  5  0.13065228
> [6,]  3  7 -1.12080980
> [7,]  1  5  0.42779354
> [8,]  2  7 -1.53111972
> [9,]  2  6  0.29299987
> [10,]  2  6 -0.01602095
>
> #whith aggregate i receive this:
>> aggregate(x[,3],list(x[,1],x[,2]),mean)
>  Group.1 Group.2          x
> 1       1       5  0.2552920
> 2       2       5  0.1306523
> 3       2       6  0.2894463
> 4       2       7 -1.6160331
> 5       3       7  0.1671974
>
>
> and the problem was that i was grouping by 2 columns, so i couldn't  
> copy the
> result to x.
>
> the solution was i made another column with paste(x[,1],x[,2],sep="_")
> and then i used the solution from this link:
> http://tolstoy.newcastle.edu.au/R/help/06/07/30184.html
> so i solved my problem

Right. That works and has the virtue that it is reasonably clear what  
is going on. Another approach, possibly even more clear and even more  
R-ish, is to use the interaction() function.

 > aggregate(x[,3], list(interaction(x[,1],x[,2]) ), mean)
   Group.1            x
1     1.5 -0.658932424
2     2.5  0.824756795
3     2.6  0.640471421
4     2.7 -0.008519716
5     3.7 -0.053233855


>
> Ivan, many thanks for your support and quik responses! :)
>
> -- 
> View this message in context: http://r.789695.n4.nabble.com/group-by-in-data-frame-tp3324240p3324608.html
> Sent from the R help mailing list archive at Nabble.com.
>
> ______________________________________________
> R-help at r-project.org mailing list
> https://stat.ethz.ch/mailman/listinfo/r-help
> PLEASE do read the posting guide http://www.R-project.org/posting-guide.html
> and provide commented, minimal, self-contained, reproducible code.

David Winsemius, MD
West Hartford, CT