Page MenuHomec4science
Files F122190482

symbchol.c
No OneTemporary
Actions

File Metadata

Created
Wed, Jul 16, 11:58

symbchol.c
View Options

/*% L = symbchol(X)
% SYMBCHOL Symbolic block sparse Cholesky factorization.
% L = symbchol(X) returns a structure L that can be used
% by the efficient block sparse Cholesky solver SPARCHOL.
% The fields in L have the following meaning:
%
% L.perm - Multiple minimum degree ordering.
%
% L.L - Sparse lower triangular matrix, has sparsity structure
% of Cholesky factor of X(L.perm,L.perm).
%
% L.xsuper - Supernode partition. Supernode jsup consists of
% the nodes L.xsuper(jsup) : L.xsuper(jsup)-1.
%
% L.split - Splitting of supernodes. Recommends to split supernode
% in blocks of sizes L.split(xsuper(jsup):L.xsuper(jsup)-1).
%
% L.tmpsiz - Quantity used by SPARCHOL, to allocated enough working
% storage.
%
% L = symbchol(X,cachsz) optimizes L.split for a computer cache
% of size CACHSZ * 1024 byte. Default cachsz = 16.
% SEE ALSO sparchol, fwblkslv, bwblkslv
% This file is part of SeDuMi 1.1 by Imre Polik and Oleksandr Romanko
% Copyright (C) 2005 McMaster University, Hamilton, CANADA (since 1.1)
%
% Copyright (C) 2001 Jos F. Sturm (up to 1.05R5)
% Dept. Econometrics & O.R., Tilburg University, the Netherlands.
% Supported by the Netherlands Organization for Scientific Research (NWO).
%
% Affiliation SeDuMi 1.03 and 1.04Beta (2000):
% Dept. Quantitative Economics, Maastricht University, the Netherlands.
%
% Affiliations up to SeDuMi 1.02 (AUG1998):
% CRL, McMaster University, Canada.
% Supported by the Netherlands Organization for Scientific Research (NWO).
%
% This program is free software; you can redistribute it and/or modify
% it under the terms of the GNU General Public License as published by
% the Free Software Foundation; either version 2 of the License, or
% (at your option) any later version.
%
% This program is distributed in the hope that it will be useful,
% but WITHOUT ANY WARRANTY; without even the implied warranty of
% MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
% GNU General Public License for more details.
%
% You should have received a copy of the GNU General Public License
% along with this program; if not, write to the Free Software
% Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA
% 02110-1301, USA
*/
#include <string.h>
#include "mex.h"
#include "blksdp.h"
#define L_OUT plhs[0]
#define NPAROUT 1
#define X_IN prhs[0]
#define cachsz_IN prhs[1]
#define NPARINMIN 1
#define NPARIN 2
/* ============================================================
MAIN: MEXFUNCTION
============================================================ */
/* ************************************************************
PROCEDURE mexFunction - Entry for Matlab
************************************************************ */
void mexFunction(const int nlhs, mxArray *plhs[],
const int nrhs, const mxArray *prhs[])
{
mwIndex m, i,j, iwsiz, flag, nofsub, cachesiz, nsuper, nsub, nnzl;
mwIndex *Xjc,*Xir, *snode, *xsuper, *invp, *colcnt, *xadj, *adjncy;
mxArray *L_FIELD;
const char *LFieldnames[] = {"L", "perm", "xsuper", "tmpsiz", "split"};
/* ------------------------------------------------------------
Check for proper number of arguments
------------------------------------------------------------ */
if(nrhs < NPARINMIN)
mexErrMsgTxt("symbchol requires 1 or 2 input arguments");
if(nlhs > NPAROUT)
mexErrMsgTxt("symbchol produces 1 output argument");
/* ------------------------------------------------------------
Check input X
------------------------------------------------------------ */
if(!mxIsSparse(X_IN))
mexErrMsgTxt("Input matrix must be sparse");
if( (m = mxGetM(X_IN)) != mxGetN(X_IN) )
mexErrMsgTxt("X should be square.");
/* ------------------------------------------------------------
Get input X
------------------------------------------------------------ */
Xjc = mxGetJc(X_IN);
Xir = mxGetIr(X_IN);
/* ------------------------------------------------------------
Allocate working arrays:
int xadj(m+1), adjncy(Xnnz), perm(m), invp(m), iwork(iwsiz)
------------------------------------------------------------ */
xadj = (mwIndex *) mxCalloc(m+1,sizeof(mwIndex));
adjncy = (mwIndex *) mxCalloc(Xjc[m],sizeof(mwIndex));
perm = (mwIndex *) mxCalloc(m,sizeof(mwIndex));
invp = (mwIndex *) mxCalloc(m,sizeof(mwIndex));
iwsiz = 4 * m;
iwork = (mwIndex *) mxCalloc(iwsiz,sizeof(mwIndex));
/* ------------------------------------------------------------
Convert C-style symmetric matrix to adjacency structure
(xadj,adjncy) in Fortran-style.
------------------------------------------------------------ */
getadj(xadj,adjncy, Xjc,Xir,m);
/* ------------------------------------------------------------
Compute multiple minimum degree ordering (J. Liu, in Fortran)
------------------------------------------------------------ */
ordmmd_(&m,xadj,adjncy, invp,perm, &iwsiz,iwork, &nofsub, &flag);
if(flag == -1)
mexErrMsgTxt("Error in ordmmd.");
mxFree(iwork);
iwsiz = 7*m + 3;
iwork = (mwIndex *) mxCalloc(iwsiz, sizeof(mwIndex));
colcnt = (mwIndex *) mxCalloc(m, sizeof(mwIndex));
snode = (mwIndex *) mxCalloc(m, sizeof(mwIndex));
xsuper = (mwIndex *) mxCalloc(m+1,sizeof(mwIndex));
xlindx = (mwIndex *) mxCalloc(m+1,sizeof(mwIndex));
sfinit_(&m, Xjc+m, xadj,adjncy, perm, invp, colcnt,
&nnzl, &nsub, &nsuper, snode, xsuper, &iwsiz, iwork, &flag);
if(flag == -1)
mexErrMsgTxt("sfinit error.");
/* ------------------------------------------------------------
Do symbolic factorization
------------------------------------------------------------ */
symfct_(&m, Xjc+m,xadj,adjncy, perm,invp,colcnt,
&nsuper,xsuper,snode, &nsub, xlindx, Lir, Ljc,
&iwsiz,iwork, &flag);
if(flag == -1)
mexErrMsgTxt("Insufficient working space.");
if(flag == -2)
mexErrMsgTxt("Input error symfct.");
#ifdef DO_BFINIT
/* ------------------------------------------------------------
Compute memory needs and cache-supernode-splitting for
sparse block Cholesky
------------------------------------------------------------ */
bfinit_(&m, &nsuper, xsuper,snode,xlindx, Lir,
&cachsz, &tmpsiz, split);
#endif
}
/* ============================================================
SUBROUTINES:
============================================================ */
/* ------------------------------------------------------------
GETADJ - Copies off-diagonal entries from C-style sparse
matrix (cjc,cir) to Fortran style sparse matrix (forjc,forir).
On input, n is number of columns.
------------------------------------------------------------ */
void getadj(mwIndex *forjc,mwIndex *forir,const mwIndex *cjc,const mwIndex *cir,const mwIndex n)
{
mwIndex i,j,inz,ix;
inz = 0;
for(j = 0; j < n; j++){
forjc[j] = inz + 1;
for(ix = cjc[j]; ix < cjc[j+1]; ix++)
if((i = cir[ix]) != j)
forir[inz++] = ++i;
}
forjc[n] = ++inz;
}

Event Timeline

c4science · Help