[R] Help Needed on Merging Columns by Summation

[Gabor Grothendieck]

On Wed, Jun 22, 2011 at 2:35 AM, Nabila Binte Zahur <nabila at nus.edu.sg> wrote:
> Dear Sirs/Madam,
>
> I am a beginner to R, and I am currently working on a data matrix which looks like this:
>
>> head(oligo)
>
>        ko:K00001 ko:K00003 ko:K00005 ko:K00008 ko:K00009 ko:K00010 ko:K00012
> AAA       370          631                  365               67           164             455       491
> KAA       603         1208                 170              157        68                495       922
> NAA        60         110                     10                7            44               51        94
> DAA       183         802                    41               62          26              166       263
> CAA        58            12                    58                1               1              23         8
> TAA       451          468                  201                90       131              320       518
>
>
> For certain specific columns, I need to merge the columns in such a way that the elements get added together.
>
> For example, in order to merge the first and the second columns, my output for the first row shld be
>
>           ko:K00001/ko:K00003
> AAA       1001 (i.e. 370+631)
> KAA        1811 (i.e. 603+1208)
>
>
> and so on for the whole column.
>

Define a by vector (called collapse here) that has equal numbers for
columns to be collapsed and distinct numbers for those that are not to
be collapsed.    We then transpose,  use rowsum and transpose back.
Assuming DF is the data frame we have:

collapse <- c(1, 1, 3:7)
setNames(as.data.frame(t(rowsum(t(DF), collapse))),
    tapply(names(DF), collapse, paste, collapse = "/"))

-- 
Statistics & Software Consulting
GKX Group, GKX Associates Inc.
tel: 1-877-GKX-GROUP
email: ggrothendieck at gmail.com