R help - Jun 2012 - How to index a matrix with different row-number for each column?

here's my question: suppose I have a matrix:

    mt<-matrix(1:12,ncol=6)

now I have a vector

    vt<-c(1,2,2,2,1,2)

which means I want to get:
the 1st row for column1;
the 2nd row for column2;
the 2nd row for column3;
the 2nd row for column4;
...

that what I want is this vector:
1,4,6,8,9,12

Does anyone know how to do this fast?
I know I can use for-loop to travel all columns,but that's not what I want.

I think you are looking for

mt[cbind(vt, seq_along(vt))]

which uses some trickines known as matrix indexing (basically, the two
columns created by cbind() are considered row & col indices -- then
seq_along() gives the col1,col2, bit as needed.

HTH,

Michael

On Tue, Jun 12, 2012 at 10:28 AM, Zhenjiang Lan <lan.zhenjiang at
gmail.com> wrote:
> here's my question: suppose I have a matrix:
>
> ? ?mt<-matrix(1:12,ncol=6)
>
> now I have a vector
>
> ? ?vt<-c(1,2,2,2,1,2)
>
> which means I want to get:
> the 1st row for column1;
> the 2nd row for column2;
> the 2nd row for column3;
> the 2nd row for column4;
> ...
>
> that what I want is this vector:
> 1,4,6,8,9,12
>
> Does anyone know how to do this fast?
> I know I can use for-loop to travel all columns,but that's not what I
want.
>
> ______________________________________________
> 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.

Use  a 2-column matrix of integer subscripts, the first column
being the row indexes and the the second the column indexes:
  > mt<-matrix(1:12,ncol=6)
  > vt<-c(1,2,2,2,1,2)
  > mt[cbind(vt,1:6)]
  [1]  1  4  6  8  9 12

Bill Dunlap
Spotfire, TIBCO Software
wdunlap tibco.com

> -----Original Message-----
> From: r-help-bounces at r-project.org [mailto:r-help-bounces at
r-project.org] On Behalf
> Of Zhenjiang Lan
> Sent: Tuesday, June 12, 2012 8:28 AM
> To: r-help at r-project.org
> Subject: [R] How to index a matrix with different row-number for each
column?
> 
> here's my question: suppose I have a matrix:
> 
>     mt<-matrix(1:12,ncol=6)
> 
> now I have a vector
> 
>     vt<-c(1,2,2,2,1,2)
> 
> which means I want to get:
> the 1st row for column1;
> the 2nd row for column2;
> the 2nd row for column3;
> the 2nd row for column4;
> ...
> 
> that what I want is this vector:
> 1,4,6,8,9,12
> 
> Does anyone know how to do this fast?
> I know I can use for-loop to travel all columns,but that's not what I
want.
> 
> ______________________________________________
> 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.

> -----Original Message-----
> From: r-help-bounces at r-project.org [mailto:r-help-bounces at r-
> project.org] On Behalf Of Zhenjiang Lan
> Sent: Tuesday, June 12, 2012 8:28 AM
> To: r-help at r-project.org
> Subject: [R] How to index a matrix with different row-number for each
> column?
> 
> here's my question: suppose I have a matrix:
> 
>     mt<-matrix(1:12,ncol=6)
> 
> now I have a vector
> 
>     vt<-c(1,2,2,2,1,2)
> 
> which means I want to get:
> the 1st row for column1;
> the 2nd row for column2;
> the 2nd row for column3;
> the 2nd row for column4;
> ...
> 
> that what I want is this vector:
> 1,4,6,8,9,12
> 
> Does anyone know how to do this fast?
> I know I can use for-loop to travel all columns,but that's not what I
> want.
> 
There may be a more memory efficient method, but this will get the job done.

diag(mt[vt,1:6])


Hope this is helpful,

Dan

Daniel J. Nordlund
Washington State Department of Social and Health Services
Planning, Performance, and Accountability
Research and Data Analysis Division
Olympia, WA 98504-5204

Thank you, it works.
by the way, sometime I guess *apply functions act like loop, because
when I apply a complicated function, they run slowly and one by one.

On Tue, Jun 12, 2012 at 3:39 PM, Rui Barradas <ruipbarradas at sapo.pt>
wrote:
> Hello,
>
> Try
>
> sapply(seq_len(ncol(mt2)), function(j) mt1[mt2[, j], j])
>
>
> Hope this helps,
>
> Rui Barradas
>
> Em 12-06-2012 17:19, Zhenjiang Lan escreveu:
>
>> mt1:
>> ? ? ?[,1] [,2] [,3]
>> [1,] "G" ?"A" ?"A"
>> [2,] "C" ?"T" ?"C"
>>
>> mt2:
>> ? ? ?[,1] [,2] [,3]
>> [1,] ? ?1 ? ?2 ? ?2
>> [2,] ? ?2 ? ?1 ? ?2
>> [3,] ? ?2 ? ?2 ? ?2
>> [4,] ? ?1 ? ?1 ? ?1
>
>