[R] Odp: Cumulative row sums, row differences

[Petr PIKAL]

Hi

r-help-bounces at r-project.org napsal dne 28.07.2009 09:18:43:

> 
> 
> I tried searching but I couldn't quite find what I was looking for.
> 
> Here's a dummy data matrix (with row and column labels):
> > y
>    0 1  2 3  4
> 21 3 4  8 5  5
> 22 3 6  8 6 NA
> 23 4 5 11 4  3
> 24 4 2  1 4  6
> 25 6 4  4 6  6
> 
> I can get cumulative row sums as follows:
> > cy<-t(apply(y,1,cumsum))
> > cy
>    0  1  2  3  4
> 21 3  7 15 20 25
> 22 3  9 17 23 NA
> 23 4  9 20 24 27
> 24 4  6  7 11 17
> 25 6 10 14 20 26
> 
> Which works, but this seems rather clumsy, especially the need for t().
> 
> Is there a better way? One that still retains row and/or column labels?
> (that will also work for data frames, if possible - though of course one 
can
> always as.data.frame() )
> 
> Row differences present a different problem. Here's one way to get back 
the
> original data:
> 
> > cbind(cy[,1],cy[,-1]-cy[,-nrow(cy)])
>      1  2 3  4
> 21 3 4  8 5  5
> 22 3 6  8 6 NA
> 23 4 5 11 4  3
> 24 4 2  1 4  6
> 25 6 4  4 6  6
> 
> However, if I use that I lose the first column label. Is there a way to 
do
> something like this without
> losing that label? (again, if possible, that also works for data 
frames?)

You can use apply approach with diff

cbind(vvv[,1, drop=F],t(apply(vvv,1, diff)))

or your construction

cbind(cy[,1, drop=F],cy[,-1]-cy[,-nrow(cy)])

Do not forget drop argument which prevents losing label!!!

Regards
Petr

And with data frames do not forget they need not have only numeric 
columns, even if they look like numeric.


> 
> thanks!
> 
> Glen_B.
> 
> -- 
> View this message in context: 
http://www.nabble.com/Cumulative-row-sums%2C-
> row-differences-tp24692986p24692986.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.