[R] strange behavior of matrix

[Phil Spector]

William -
    An interesting feature of matrix indexing in R is that
if you provide a two column matrix as a subscript, you are
refering to the elements whose indices are in the rows
of the matrix.  This is extremely handy for converting 
tables to matrices:

> m = cbind(c(1,1,1,2,2,2,3,3,3),c(1,2,3,1,2,3,1,2,3),c(10,19,8,14,12,6,17,9,2))
> m
       [,1] [,2] [,3]
  [1,]    1    1   10
  [2,]    1    2   19
  [3,]    1    3    8
  [4,]    2    1   14
  [5,]    2    2   12
  [6,]    2    3    6
  [7,]    3    1   17
  [8,]    3    2    9
  [9,]    3    3    2
> newmat = matrix(0,3,3)
> newmat[m[,1:2]] = m[,3]
> newmat
      [,1] [,2] [,3]
[1,]   10   19    8
[2,]   14   12    6
[3,]   17    9    2

It's also handy for extracting a vector with just the elements you want:

> newmat[cbind(c(1,2,3),c(2,3,1))]
[1] 19  6 17  # 1,2 2,3 3,1 elements

So it's a bit surprising when you index a matrix with a 2 column 
matrix, but it is a documented fact.

 					- Phil Spector
 					 Statistical Computing Facility
 					 Department of Statistics
 					 UC Berkeley
 					 spector at stat.berkeley.edu




On Mon, 5 Apr 2010, William Revelle wrote:

> Dear R list,
>
> I have discovered a seemingly peculiar feature when using a matrix to index 
> itself (yes, this is strange code, which I have now modified to be more 
> reasonable).
>
> #this makes sense
> s <- matrix(1:3,nrow=1)
> s[s]    #all three elements are shown
>
> #but when I try
> s <- matrix(1:2,nrow=1)
> s[1]     #fine, the first element is shown
> s[2]     #fine, the second element is shown
> s[s]     #just the second element is shown  -- this is peculiar
>
>
> #But doing it by columns works for both cases
> s <- matrix(1:3,ncol=1)
> s[s]    #all three elements are shown
>
>
> #and when I try the same problem down a column
> s <- matrix(1:2,ncol=1)
> s[1]     #fine
>
> s[2]     #fine
>
> s[s]     #this shows both elements  as would be expected
>
> #clearly since I have just one dimension, it would have been better to
> s <- 1:2
> s[s]   #which works as one would expect.
>
> Or, using the array  function we get the same problem.
>
>>  s <- array(1:2,dim=c(1,2))
>>  s[s]
> [1] 2
>>  s <- array(1:2,dim=c(2,1))
>>  s[s]
> [1] 1 2
>
>
>>  sessionInfo()
> R version 2.11.0 Under development (unstable) (2010-03-24 r51389)
> i386-apple-darwin9.8.0
>
> locale:
> [1] en_US.UTF-8/en_US.UTF-8/C/C/en_US.UTF-8/en_US.UTF-8
>
> attached base packages:
> [1] stats     graphics  grDevices utils     datasets  methods   base
>
> other attached packages:
> [1] psych_1.0-87
>> 
>
> I think this is unexpected behavior.
>
> Best wishes,
>
> Bill
>
>
> -- 
> William Revelle		http://personality-project.org/revelle.html
> Professor			http://personality-project.org
> Department of Psychology             http://www.wcas.northwestern.edu/psych/
> Northwestern University	http://www.northwestern.edu/
> Use R for psychology                       http://personality-project.org/r
> It is 6 minutes to midnight	http://www.thebulletin.org
>
> ______________________________________________
> 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.
>