Re: [Rcpp-devel] Matrix assignment with inline
Thanks for the advice Dirk, > >Kaveh, > >If you are having problems with things like matrix indexing I would recommend >a somewhat more incremental approach: > > pass a matrix from R, find the maximum element and return it > > pass a matrix from R, find the min and maximum elements and return both > > multiply two (small) matrices by doing the looping yourself and return > the new matrix > >etc pp before trying to port code. > >Cheers, Dirk > >-- >Dirk Eddelbuettel | e...@debian.org | http://dirk.eddelbuettel.com > > ___ Rcpp-devel mailing list Rcpp-devel@lists.r-forge.r-project.org https://lists.r-forge.r-project.org/cgi-bin/mailman/listinfo/rcpp-devel
Re: [Rcpp-devel] Matrix assignment with inline
Kaveh, If you are having problems with things like matrix indexing I would recommend a somewhat more incremental approach: pass a matrix from R, find the maximum element and return it pass a matrix from R, find the min and maximum elements and return both multiply two (small) matrices by doing the looping yourself and return the new matrix etc pp before trying to port code. Cheers, Dirk -- Dirk Eddelbuettel | e...@debian.org | http://dirk.eddelbuettel.com ___ Rcpp-devel mailing list Rcpp-devel@lists.r-forge.r-project.org https://lists.r-forge.r-project.org/cgi-bin/mailman/listinfo/rcpp-devel
Re: [Rcpp-devel] Matrix assignment with inline
Le 03/11/10 12:16, Kaveh Vakili a écrit :
Hi list,
I'm trying to learn inline by translating snippets of c++ code from
the book "Numerical Recipes".
in page 44-45, there is a c++ code to perform Gauss-Jordan elimination of a
matrix (see below). I understand that they have a large template of additional
operator/function/class (it is online: www.nr.com/codefile.php?nr3 ) which
means the codes in the book have to be slightly 'translated' (correct me if i'm
wrong) unto proper inline inputs. I think i have a problem with translating the
matrix functions. I have searched the devel archives, but have only found codes
to assign full row/column (not individual matrix entries).
Anyhow, my use of 'a[j][k]' in line 44 (and similar threats later) seem to be
an issue: none of the examples in Rcpp seems to use it. My question is how do
you translate this construct (i.e. a[j][k]) so that inline can 'eat' it ?
Hi,
You can use a(j,k)
There are a lot of places in this code where Rcpp sugar might help. I
don't want to go through all of it right now, but for example this line:
for(l=0;l system.file( "unitTests/runit.sugar.R", package = "Rcpp" )
Also, I'm not sure that this code is valid:
throw ("gaussj: Singular Matrix");
Maybe you mean :
throw std::range_error("gaussj: Singular Matrix");
or perhaps using some other exception class instead of range_error.
HTH,
Romain
library(Rcpp)
library(inline)
fun3<-'
NumericMatrix a(x) ;
NumericMatrix b = clone(a);
int i,icol,irow,j,k,l,ll,n=a.nrow(),m=a.ncol();
std::vector indxc(n);
std::vector indxr(n);
std::vector ipiv(n);
double big,dum,piinv;
for(j=0;j= big) {
big = abs(a[j][k]);
irow = j;
icol = k;
}
}
}
++(ipiv[icol]);
if(irow != icol){
for (l=0;l=0;l--){
if(indxr[l] != indxc[l])
for(k=0;khttps://lists.r-forge.r-project.org/cgi-bin/mailman/listinfo/rcpp-devel
--
Romain Francois
Professional R Enthusiast
+33(0) 6 28 91 30 30
http://romainfrancois.blog.free.fr
|- http://bit.ly/czHPM7 : Rcpp Google tech talk on youtube
|- http://bit.ly/9P0eF9 : Google slides
`- http://bit.ly/cVPjPe : Chicago R Meetup slides
___
Rcpp-devel mailing list
Rcpp-devel@lists.r-forge.r-project.org
https://lists.r-forge.r-project.org/cgi-bin/mailman/listinfo/rcpp-devel
[Rcpp-devel] Matrix assignment with inline
Hi list,
I'm trying to learn inline by translating snippets of c++ code from
the book "Numerical Recipes".
in page 44-45, there is a c++ code to perform Gauss-Jordan elimination of a
matrix (see below). I understand that they have a large template of additional
operator/function/class (it is online: www.nr.com/codefile.php?nr3 ) which
means the codes in the book have to be slightly 'translated' (correct me if i'm
wrong) unto proper inline inputs. I think i have a problem with translating the
matrix functions. I have searched the devel archives, but have only found codes
to assign full row/column (not individual matrix entries).
Anyhow, my use of 'a[j][k]' in line 44 (and similar threats later) seem to be
an issue: none of the examples in Rcpp seems to use it. My question is how do
you translate this construct (i.e. a[j][k]) so that inline can 'eat' it ?
library(Rcpp)
library(inline)
fun3<-'
NumericMatrix a(x) ;
NumericMatrix b = clone(a);
int i,icol,irow,j,k,l,ll,n=a.nrow(),m=a.ncol();
std::vector indxc(n);
std::vector indxr(n);
std::vector ipiv(n);
double big,dum,piinv;
for(j=0;j= big) {
big = abs(a[j][k]);
irow = j;
icol = k;
}
}
}
++(ipiv[icol]);
if(irow != icol){
for (l=0;l=0;l--){
if(indxr[l] != indxc[l])
for(k=0;khttps://lists.r-forge.r-project.org/cgi-bin/mailman/listinfo/rcpp-devel
