R help - May 2003 - rank correlation and distance between two different matrices

Dear all,
in package Hmisc  `rcorr' computes a matrix of Spearman's `rho' rank
correlation coefficients for all possible pairs of columns of a matrix.
What if I want a matrix of  rank correlation coefficients for pair of
columns of two different matrices?

I have the same question about distance metrics in package Vegan. The
function 'vegdist' computes distance indexes for all possible pairs of
row of a matrix. What if I am interested in comparing each single row of
matrix A with the correspondent row of matrix B?

Any help would be appreciated.

Luca

Apply these to c/rbind(A,B) and extract the part of the result which you 
want.

On Mon, 12 May 2003, Luca De Benedictis wrote:
> in package Hmisc  `rcorr' computes a matrix of Spearman's `rho'
rank
> correlation coefficients for all possible pairs of columns of a matrix.
> What if I want a matrix of  rank correlation coefficients for pair of
> columns of two different matrices?
> 
> I have the same question about distance metrics in package Vegan. The
> function 'vegdist' computes distance indexes for all possible pairs
of
> row of a matrix. What if I am interested in comparing each single row of
> matrix A with the correspondent row of matrix B?
-- 
Brian D. Ripley,                  ripley at stats.ox.ac.uk
Professor of Applied Statistics,  http://www.stats.ox.ac.uk/~ripley/
University of Oxford,             Tel:  +44 1865 272861 (self)
1 South Parks Road,                     +44 1865 272866 (PA)
Oxford OX1 3TG, UK                Fax:  +44 1865 272595

On Mon, 2003-05-12 at 01:36, Luca De Benedictis wrote:
> Dear all,
> in package Hmisc  `rcorr' computes a matrix of Spearman's `rho'
rank
> correlation coefficients for all possible pairs of columns of a matrix.
> What if I want a matrix of  rank correlation coefficients for pair of
> columns of two different matrices?
> 
> I have the same question about distance metrics in package Vegan. The
> function 'vegdist' computes distance indexes for all possible pairs
of
> row of a matrix. What if I am interested in comparing each single row of
> matrix A with the correspondent row of matrix B?
Wasteful, but may do what you asked for:
> A <- matrix(runif(120), nrow=10, ncol=12)
> B <- matrix(runif(120), nrow=10, ncol=12)
> diag(as.matrix(dist(rbind(A,B)))[11:20, 1:10])
 [1] 1.362722 1.381713 1.526918 1.565419 1.103431 1.384486 1.186895
 [8] 1.512283 1.401761 1.817098

cheers, jari oksanen

-- 
Jari Oksanen <jarioksa at sun3.oulu.fi>