/usr/include/sofa/component/linearsolver/CompressedRowSparseMatrix.h is in libsofa1-dev 1.0~beta4-10ubuntu2.
This file is owned by root:root, with mode 0o644.
The actual contents of the file can be viewed below.
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 409 410 411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 432 433 434 435 436 437 438 439 440 441 442 443 444 445 446 447 448 449 450 451 452 453 454 455 456 457 458 459 460 461 462 463 464 465 466 467 468 469 470 471 472 473 474 475 476 477 478 479 480 481 482 483 484 485 486 487 488 489 490 491 492 493 494 495 496 497 498 499 500 501 502 503 504 505 506 507 508 509 510 511 512 513 514 515 516 517 518 519 520 521 522 523 524 525 526 527 528 529 530 531 532 533 534 535 536 537 538 539 540 541 542 543 544 545 546 547 548 549 550 551 552 553 554 555 556 557 558 559 560 561 562 563 564 565 566 567 568 569 570 571 572 573 574 575 576 577 578 579 580 581 582 583 584 585 586 587 588 589 590 591 592 593 594 595 596 597 598 599 600 601 602 603 604 605 606 607 608 609 610 611 612 613 614 615 616 617 618 619 620 621 622 623 624 625 626 627 628 629 630 631 632 633 634 635 636 637 638 639 640 641 642 643 644 645 646 647 648 649 650 651 652 653 654 655 656 657 658 659 660 661 662 663 664 665 666 667 668 669 670 671 672 673 674 675 676 677 678 679 680 681 682 683 684 685 686 687 688 689 690 691 692 693 694 695 696 697 698 699 700 701 702 703 704 705 706 707 708 709 710 711 712 713 714 715 716 717 718 719 720 721 722 723 724 725 726 727 728 729 730 731 732 733 734 735 736 737 738 739 740 741 742 743 744 745 746 747 748 749 750 751 752 753 754 755 756 757 758 759 760 761 762 763 764 765 766 767 768 769 770 771 772 773 774 775 776 777 778 779 780 781 782 783 784 785 786 787 788 789 790 791 792 793 794 795 796 797 798 799 800 801 802 803 804 805 806 807 808 809 810 811 812 813 814 815 816 817 818 819 820 821 822 823 824 825 826 827 828 829 830 831 832 833 834 835 836 837 838 839 840 841 842 843 844 845 846 847 848 849 850 851 852 853 854 855 856 857 858 859 860 861 862 863 864 865 866 867 868 869 870 871 872 873 874 875 876 877 878 879 880 881 882 883 884 885 886 887 888 889 890 891 892 893 894 895 896 897 898 899 900 901 902 903 904 905 906 907 908 909 910 911 912 913 914 915 916 917 918 919 920 921 922 923 924 925 926 927 928 929 930 931 932 933 934 935 936 937 938 939 940 941 942 943 944 945 946 947 948 949 950 951 952 953 954 955 956 957 958 959 960 961 962 963 964 965 966 967 968 969 970 971 972 973 974 975 976 977 978 979 980 981 982 983 984 985 986 987 988 989 990 991 992 993 994 995 996 997 998 999 1000 1001 1002 1003 1004 1005 1006 1007 1008 1009 1010 1011 1012 1013 1014 1015 1016 1017 1018 1019 1020 1021 1022 1023 1024 1025 1026 1027 1028 1029 1030 1031 1032 1033 1034 1035 1036 1037 1038 1039 1040 1041 1042 1043 1044 1045 1046 1047 1048 1049 1050 1051 1052 1053 1054 1055 1056 1057 1058 1059 1060 1061 1062 1063 1064 1065 1066 1067 1068 1069 1070 1071 1072 1073 1074 1075 1076 1077 1078 1079 1080 1081 1082 | /******************************************************************************
* SOFA, Simulation Open-Framework Architecture, version 1.0 beta 4 *
* (c) 2006-2009 MGH, INRIA, USTL, UJF, CNRS *
* *
* This library is free software; you can redistribute it and/or modify it *
* under the terms of the GNU Lesser General Public License as published by *
* the Free Software Foundation; either version 2.1 of the License, or (at *
* your option) any later version. *
* *
* This library 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 Lesser General Public License *
* for more details. *
* *
* You should have received a copy of the GNU Lesser General Public License *
* along with this library; if not, write to the Free Software Foundation, *
* Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA. *
*******************************************************************************
* SOFA :: Modules *
* *
* Authors: The SOFA Team and external contributors (see Authors.txt) *
* *
* Contact information: contact@sofa-framework.org *
******************************************************************************/
#ifndef SOFA_COMPONENT_LINEARSOLVER_COMPRESSEDROWSPARSEMATRIX_H
#define SOFA_COMPONENT_LINEARSOLVER_COMPRESSEDROWSPARSEMATRIX_H
#include <sofa/defaulttype/BaseMatrix.h>
#include <sofa/component/linearsolver/MatrixLinearSolver.h>
#include "FullVector.h"
#include <algorithm>
namespace sofa
{
namespace component
{
namespace linearsolver
{
//#define SPARSEMATRIX_CHECK
//#define SPARSEMATRIX_VERBOSE
template<int TN> class bloc_index_func
{
public:
enum { N = TN };
static void split(int& index, int& modulo)
{
modulo = index % N;
index = index / N;
}
};
template<> class bloc_index_func<1>
{
public:
enum { N = 1 };
static void split(int&, int&)
{
}
};
template<> class bloc_index_func<2>
{
public:
enum { N = 2 };
static void split(int& index, int& modulo)
{
modulo = index & 1;
index = index >> 1;
}
};
template<> class bloc_index_func<4>
{
public:
enum { N = 2 };
static void split(int& index, int& modulo)
{
modulo = index & 3;
index = index >> 2;
}
};
template<> class bloc_index_func<8>
{
public:
enum { N = 2 };
static void split(int& index, int& modulo)
{
modulo = index & 7;
index = index >> 3;
}
};
template<class T>
class matrix_bloc_traits;
template <int L, int C, class real>
class matrix_bloc_traits < defaulttype::Mat<L,C,real> >
{
public:
typedef defaulttype::Mat<L,C,real> Bloc;
typedef real Real;
enum { NL = L };
enum { NC = C };
static Real& v(Bloc& b, int row, int col) { return b[row][col]; }
static const Real& v(const Bloc& b, int row, int col) { return b[row][col]; }
static void clear(Bloc& b) { b.clear(); }
static bool empty(const Bloc& b)
{
for (int i=0;i<NL;++i)
for (int j=0;j<NC;++j)
if (b[i][j] != 0) return false;
return true;
}
static const char* Name();
};
template<> inline const char* matrix_bloc_traits<defaulttype::Mat<1,1,float > >::Name() { return "1f"; }
template<> inline const char* matrix_bloc_traits<defaulttype::Mat<1,1,double> >::Name() { return "1d"; }
template<> inline const char* matrix_bloc_traits<defaulttype::Mat<2,2,float > >::Name() { return "2f"; }
template<> inline const char* matrix_bloc_traits<defaulttype::Mat<2,2,double> >::Name() { return "2d"; }
template<> inline const char* matrix_bloc_traits<defaulttype::Mat<3,3,float > >::Name() { return "3f"; }
template<> inline const char* matrix_bloc_traits<defaulttype::Mat<3,3,double> >::Name() { return "3d"; }
template<> inline const char* matrix_bloc_traits<defaulttype::Mat<4,4,float > >::Name() { return "4f"; }
template<> inline const char* matrix_bloc_traits<defaulttype::Mat<4,4,double> >::Name() { return "4d"; }
template<> inline const char* matrix_bloc_traits<defaulttype::Mat<6,6,float > >::Name() { return "6f"; }
template<> inline const char* matrix_bloc_traits<defaulttype::Mat<6,6,double> >::Name() { return "6d"; }
template<> inline const char* matrix_bloc_traits<defaulttype::Mat<8,8,float > >::Name() { return "8f"; }
template<> inline const char* matrix_bloc_traits<defaulttype::Mat<8,8,double> >::Name() { return "8d"; }
template <>
class matrix_bloc_traits < float >
{
public:
typedef float Bloc;
typedef float Real;
enum { NL = 1 };
enum { NC = 1 };
static Real& v(Bloc& b, int, int) { return b; }
static const Real& v(const Bloc& b, int, int) { return b; }
static void clear(Bloc& b) { b = 0; }
static bool empty(const Bloc& b)
{
return b == 0;
}
static const char* Name() { return "f"; }
};
template <>
class matrix_bloc_traits < double >
{
public:
typedef double Bloc;
typedef double Real;
enum { NL = 1 };
enum { NC = 1 };
static Real& v(Bloc& b, int, int) { return b; }
static const Real& v(const Bloc& b, int, int) { return b; }
static void clear(Bloc& b) { b = 0; }
static bool empty(const Bloc& b)
{
return b == 0;
}
static const char* Name() { return "d"; }
};
template<typename TBloc, typename TVecBloc = helper::vector<TBloc>, typename TVecIndex = helper::vector<int> >
class CompressedRowSparseMatrix : public defaulttype::BaseMatrix
{
public:
typedef CompressedRowSparseMatrix<TBloc,TVecBloc,TVecIndex> Matrix;
typedef TBloc Bloc;
typedef matrix_bloc_traits<Bloc> traits;
typedef typename traits::Real Real;
enum { NL = traits::NL };
enum { NC = traits::NC };
typedef int Index;
typedef TVecBloc VecBloc;
typedef TVecIndex VecIndex;
struct IndexedBloc
{
Index l,c;
Bloc value;
IndexedBloc() {}
IndexedBloc(Index i, Index j) : l(i), c(j) {}
IndexedBloc(Index i, Index j, const Bloc& v) : l(i), c(j), value(v) {}
bool operator < (const IndexedBloc& b) const
{
return (l < b.l) || (l == b.l && c < b.c);
}
bool operator <= (const IndexedBloc& b) const
{
return (l < b.l) || (l == b.l && c <= b.c);
}
bool operator > (const IndexedBloc& b) const
{
return (l > b.l) || (l == b.l && c > b.c);
}
bool operator >= (const IndexedBloc& b) const
{
return (l > b.l) || (l == b.l && c >= b.c);
}
bool operator == (const IndexedBloc& b) const
{
return (l == b.l) && (c == b.c);
}
bool operator != (const IndexedBloc& b) const
{
return (l != b.l) || (c != b.c);
}
};
typedef helper::vector<IndexedBloc> VecIndexedBloc;
static void split_row_index(int& index, int& modulo) { bloc_index_func<NL>::split(index, modulo); }
static void split_col_index(int& index, int& modulo) { bloc_index_func<NC>::split(index, modulo); }
class Range : public std::pair<Index, Index>
{
typedef std::pair<Index, Index> Inherit;
public:
Range() : Inherit(0,0) {}
Range(Index begin, Index end) : Inherit(begin,end) {}
Index begin() const { return this->first; }
Index end() const { return this->second; }
void setBegin(Index i) { this->first = i; }
void setEnd(Index i) { this->second = i; }
bool empty() const { return begin() == end(); }
Index size() const { return end()-begin(); }
typename VecBloc::iterator begin(VecBloc& b) const { return b.begin() + begin(); }
typename VecBloc::iterator end (VecBloc& b) const { return b.end () + end (); }
typename VecBloc::const_iterator begin(const VecBloc& b) const { return b.begin() + begin(); }
typename VecBloc::const_iterator end (const VecBloc& b) const { return b.end () + end (); }
typename VecIndex::iterator begin(VecIndex& b) const { return b.begin() + begin(); }
typename VecIndex::iterator end (VecIndex& b) const { return b.end () + end (); }
typename VecIndex::const_iterator begin(const VecIndex& b) const { return b.begin() + begin(); }
typename VecIndex::const_iterator end (const VecIndex& b) const { return b.end () + end (); }
void operator++() { ++first; }
void operator++(int) { ++first; }
};
static bool sortedFind(const VecIndex& v, Range in, Index val, Index& result)
{
if (in.empty()) return false;
Index candidate = (result >= in.begin() && result < in.end()) ? result : ((in.begin() + in.end()) >> 1);
for(;;)
{
Index i = v[candidate];
if (i == val) { result = candidate; return true; }
if (i < val) in.setBegin(candidate+1);
else in.setEnd(candidate);
if (in.empty()) break;
candidate = (in.begin() + in.end()) >> 1;
}
return false;
}
static bool sortedFind(const VecIndex& v, Index val, Index& result)
{
return sortedFind(v, Range(0,v.size()), val, result);
}
protected:
Index nRow,nCol;
Index nBlocRow,nBlocCol;
bool compressed;
VecIndex rowIndex;
VecIndex rowBegin;
VecIndex colsIndex;
VecBloc colsValue;
VecIndexedBloc btemp;
// Temporary vectors used during compression
VecIndex oldRowIndex;
VecIndex oldRowBegin;
VecIndex oldColsIndex;
VecBloc oldColsValue;
public:
CompressedRowSparseMatrix()
: nRow(0), nCol(0), nBlocRow(0), nBlocCol(0), compressed(true)
{
}
CompressedRowSparseMatrix(int nbRow, int nbCol)
: nRow(nbRow), nCol(nbCol), compressed(true)
{
nBlocRow = (nRow + NL-1) / NL;
nBlocCol = (nCol + NC-1) / NC;
}
unsigned int rowBSize() const
{
return nBlocRow;
}
unsigned int colBSize() const
{
return nBlocCol;
}
const VecIndex& getRowIndex() const { return rowIndex; }
const VecIndex& getRowBegin() const { return rowBegin; }
Range getRowRange(int id) const { return Range(rowBegin[id], rowBegin[id+1]); }
const VecIndex& getColsIndex() const { return colsIndex; }
const VecBloc& getColsValue() const { return colsValue; }
void resizeBloc(int nbBRow, int nbBCol)
{
if (nBlocRow == nbBRow && nBlocRow == nbBCol)
{ // just clear the matrix
for (unsigned int i=0; i < colsValue.size(); ++i)
traits::clear(colsValue[i]);
compressed = colsValue.empty();
btemp.clear();
}
else
{
#ifdef SPARSEMATRIX_VERBOSE
std::cout << /* this->Name() << */": resize("<<nbBRow<<"*"<<NL<<","<<nbBCol<<"*"<<NC<<")"<<std::endl;
#endif
nBlocRow = nbBRow;
nBlocCol = nbBCol;
rowIndex.clear();
rowBegin.clear();
colsIndex.clear();
colsValue.clear();
compressed = true;
btemp.clear();
}
}
void compress()
{
if (compressed && btemp.empty()) return;
if (!btemp.empty())
{
#ifdef SPARSEMATRIX_VERBOSE
std::cout << /* this->Name() << */"("<<rowSize()<<","<<colSize()<<"): sort "<<btemp.size()<<" temp blocs."<<std::endl;
#endif
std::sort(btemp.begin(),btemp.end());
#ifdef SPARSEMATRIX_VERBOSE
std::cout << /* this->Name() << */"("<<rowSize()<<","<<colSize()<<"): blocs sorted."<<std::endl;
#endif
}
oldRowIndex.swap(rowIndex);
oldRowBegin.swap(rowBegin);
oldColsIndex.swap(colsIndex);
oldColsValue.swap(colsValue);
rowIndex.clear();
rowBegin.clear();
colsIndex.clear();
colsValue.clear();
rowIndex.reserve(oldRowIndex.empty() ? nBlocRow : oldRowIndex.size());
rowBegin.reserve((oldRowIndex.empty() ? nBlocRow : oldRowIndex.size())+1);
colsIndex.reserve(oldColsIndex.size() + btemp.size());
colsValue.reserve(oldColsIndex.size() + btemp.size());
const Index oldNRow = oldRowIndex.size();
const Index EndRow = nBlocRow;
const Index EndCol = nBlocCol;
//const Index EndVal = oldColsIndex.size();
Index inRowId = 0;
Index inRowIndex = (inRowId < oldNRow ) ? oldRowIndex[inRowId] : EndRow;
typename VecIndexedBloc::const_iterator itbtemp = btemp.begin(), endbtemp = btemp.end();
Index bRowIndex = (itbtemp != endbtemp) ? itbtemp->l : EndRow;
Index outValId = 0;
while (inRowIndex < EndRow || bRowIndex < EndRow)
{
#ifdef SPARSEMATRIX_VERBOSE
std::cout << /* this->Name() << */"("<<rowSize()<<","<<colSize()<<"): inRowIndex = "<<inRowIndex<<" , bRowIndex = "<<bRowIndex<<""<<std::endl;
#endif
if (inRowIndex < bRowIndex)
{ // this row contains values only from old*
rowIndex.push_back(inRowIndex);
rowBegin.push_back(outValId);
Range inRow( oldRowBegin[inRowId], oldRowBegin[inRowId+1] );
while (!inRow.empty())
{
if (!traits::empty(oldColsValue[inRow.begin()]))
{
colsIndex.push_back(oldColsIndex[inRow.begin()]);
colsValue.push_back(oldColsValue[inRow.begin()]);
++outValId;
}
++inRow;
}
//colsIndex.insert(colsIndex.end(), inRow.begin(oldColsIndex), inRow.end(oldColsIndex));
//colsValue.insert(colsValue.end(), inRow.begin(oldColsValue), inRow.end(oldColsValue));
//outValId += inRow.size();
++inRowId;
inRowIndex = (inRowId < oldNRow ) ? oldRowIndex[inRowId] : EndRow;
}
else if (inRowIndex > bRowIndex)
{ // this row contains values only from btemp
rowIndex.push_back(bRowIndex);
rowBegin.push_back(outValId);
while (itbtemp != endbtemp && itbtemp->l == bRowIndex)
{
Index bColIndex = itbtemp->c;
colsIndex.push_back(bColIndex);
colsValue.push_back(itbtemp->value);
++itbtemp;
Bloc& value = colsValue.back();
while (itbtemp != endbtemp && itbtemp->c == bColIndex && itbtemp->l == bRowIndex)
{
value += itbtemp->value;
++itbtemp;
}
++outValId;
}
bRowIndex = (itbtemp != endbtemp) ? itbtemp->l : EndRow;
}
else
{ // this row mixes values from old* and btemp
rowIndex.push_back(inRowIndex);
rowBegin.push_back(outValId);
Range inRow( oldRowBegin[inRowId], oldRowBegin[inRowId+1] );
Index inColIndex = (!inRow.empty()) ? oldColsIndex[inRow.begin()] : EndCol;
Index bColIndex = (itbtemp != endbtemp && itbtemp->l == inRowIndex) ? itbtemp->c : EndCol;
while (inColIndex < EndCol || bColIndex < EndCol)
{
if (inColIndex < bColIndex)
{
if (!traits::empty(oldColsValue[inRow.begin()]))
{
colsIndex.push_back(inColIndex);
colsValue.push_back(oldColsValue[inRow.begin()]);
++outValId;
}
++inRow;
inColIndex = (!inRow.empty()) ? oldColsIndex[inRow.begin()] : EndCol;
}
else if (inColIndex > bColIndex)
{
colsIndex.push_back(bColIndex);
colsValue.push_back(itbtemp->value);
++itbtemp;
Bloc& value = colsValue.back();
while (itbtemp != endbtemp && itbtemp->c == bColIndex && itbtemp->l == bRowIndex)
{
value += itbtemp->value;
++itbtemp;
}
bColIndex = (itbtemp != endbtemp && itbtemp->l == bRowIndex) ? itbtemp->c : EndCol;
++outValId;
}
else
{
colsIndex.push_back(inColIndex);
colsValue.push_back(oldColsValue[inRow.begin()]);
++inRow;
inColIndex = (!inRow.empty()) ? oldColsIndex[inRow.begin()] : EndCol;
Bloc& value = colsValue.back();
while (itbtemp != endbtemp && itbtemp->c == bColIndex && itbtemp->l == bRowIndex)
{
value += itbtemp->value;
++itbtemp;
}
bColIndex = (itbtemp != endbtemp && itbtemp->l == bRowIndex) ? itbtemp->c : EndCol;
++outValId;
}
}
++inRowId;
inRowIndex = (inRowId < oldNRow ) ? oldRowIndex[inRowId] : EndRow;
bRowIndex = (itbtemp != endbtemp) ? itbtemp->l : EndRow;
}
}
rowBegin.push_back(outValId);
//#ifdef SPARSEMATRIX_VERBOSE
std::cout << /* this->Name() << */"("<<rowSize()<<","<<colSize()<<"): compressed " << oldColsIndex.size()<<" old blocs and " << btemp.size() << " temp blocs into " << rowIndex.size() << " lines and " << colsIndex.size() << " blocs."<<std::endl;
//#endif
btemp.clear();
compressed = true;
}
// filtering-out part of a matrix
typedef bool filter_fn (int i , int j , Bloc& val, const Bloc& ref );
static bool nonzeros(int /*i*/, int /*j*/, Bloc& val, const Bloc& /*ref*/) { return (!traits::empty(val)); }
static bool nonsmall(int /*i*/, int /*j*/, Bloc& val, const Bloc& ref )
{
for (int bi = 0; bi < NL; ++bi)
for (int bj = 0; bj < NC; ++bj)
if (helper::rabs(traits::v(val, bi, bj)) >= ref) return true;
return false;
}
static bool upper (int i , int j , Bloc& val, const Bloc& /*ref*/)
{
if (NL>1 && i*NL == j*NC)
{
for (int bi = 1; bi < NL; ++bi)
for (int bj = 0; bj < bi; ++bj)
traits::v(val, bi, bj) = 0;
}
return i*NL <= j*NC;
}
static bool lower (int i , int j , Bloc& val, const Bloc& /*ref*/)
{
if (NL>1 && i*NL == j*NC)
{
for (int bi = 0; bi < NL-1; ++bi)
for (int bj = bi+1; bj < NC; ++bj)
traits::v(val, bi, bj) = 0;
}
return i*NL >= j*NC;
}
static bool upper_nonzeros(int i , int j , Bloc& val, const Bloc& ref ) { return upper(i,j,val,ref) && nonzeros(i,j,val,ref); }
static bool lower_nonzeros(int i , int j , Bloc& val, const Bloc& ref ) { return lower(i,j,val,ref) && nonzeros(i,j,val,ref); }
static bool upper_nonsmall(int i , int j , Bloc& val, const Bloc& ref ) { return upper(i,j,val,ref) && nonsmall(i,j,val,ref); }
static bool lower_nonsmall(int i , int j , Bloc& val, const Bloc& ref ) { return lower(i,j,val,ref) && nonsmall(i,j,val,ref); }
void filterValues(Matrix& M, filter_fn* filter = &nonzeros, const Bloc& ref = Bloc())
{
M.compress();
nRow = M.rowSize();
nCol = M.colSize();
nBlocRow = M.rowBSize();
nBlocCol = M.colBSize();
rowIndex.clear();
rowBegin.clear();
colsIndex.clear();
colsValue.clear();
compressed = true;
btemp.clear();
rowIndex.reserve(M.rowIndex.size());
rowBegin.reserve(M.rowBegin.size());
colsIndex.reserve(M.colsIndex.size());
colsValue.reserve(M.colsValue.size());
int vid = 0;
for (unsigned int rowId = 0; rowId < M.rowIndex.size(); ++rowId)
{
int i = M.rowIndex[rowId];
rowIndex.push_back(i);
rowBegin.push_back(vid);
Range rowRange(M.rowBegin[rowId], M.rowBegin[rowId+1]);
for (int xj = rowRange.begin(); xj < rowRange.end(); ++xj)
{
int j = M.colsIndex[xj];
Bloc b = M.colsValue[xj];
if ((*filter)(i,j,b,ref))
{
colsIndex.push_back(j);
colsValue.push_back(b);
++vid;
}
}
if (rowBegin.back() == vid) // row was empty
{
rowIndex.pop_back();
rowBegin.pop_back();
}
}
rowBegin.push_back(vid); // end of last row
}
void copyNonZeros(Matrix& M)
{
filterValues(M, nonzeros);
}
void copyNonSmall(Matrix& M, const Bloc& ref)
{
filterValues(M, nonzeros, ref);
}
void copyUpper(Matrix& M)
{
filterValues(M, upper);
}
void copyLower(Matrix& M)
{
filterValues(M, lower);
}
void copyUpperNonZeros(Matrix& M)
{
filterValues(M, upper_nonzeros);
}
void copyLowerNonZeros(Matrix& M)
{
filterValues(M, lower_nonzeros);
}
void copyUpperNonSmall(Matrix& M, const Bloc& ref)
{
filterValues(M, upper_nonsmall, ref);
}
void copyLowerNonSmall(Matrix& M, const Bloc& ref)
{
filterValues(M, lower_nonsmall, ref);
}
const Bloc& bloc(int i, int j) const
{
static Bloc empty;
int rowId = i * rowIndex.size() / nBlocRow;
if (sortedFind(rowIndex, i, rowId))
{
Range rowRange(rowBegin[rowId], rowBegin[rowId+1]);
Index colId = rowRange.begin() + j * rowRange.size() / nBlocCol;
if (sortedFind(colsIndex, rowRange, j, colId))
{
return colsValue[colId];
}
}
return empty;
}
Bloc* wbloc(int i, int j, bool create = false)
{
int rowId = i * rowIndex.size() / nBlocRow;
if (sortedFind(rowIndex, i, rowId))
{
Range rowRange(rowBegin[rowId], rowBegin[rowId+1]);
int colId = rowRange.begin() + j * rowRange.size() / nBlocCol;
if (sortedFind(colsIndex, rowRange, j, colId))
{
#ifdef SPARSEMATRIX_VERBOSE
std::cout << /* this->Name() << */"("<<rowBSize()<<"*"<<NL<<","<<colBSize()<<"*"<<NC<<"): bloc("<<i<<","<<j<<") found at "<<colId<<" (line "<<rowId<<")."<<std::endl;
#endif
return &colsValue[colId];
}
}
if (create)
{
if (btemp.empty() || btemp.back().l != i || btemp.back().c != j)
{
#ifdef SPARSEMATRIX_VERBOSE
std::cout << /* this->Name() << */"("<<rowSize()<<","<<colSize()<<"): new temp bloc ("<<i<<","<<j<<")"<<std::endl;
#endif
btemp.push_back(IndexedBloc(i,j));
traits::clear(btemp.back().value);
}
return &btemp.back().value;
}
return NULL;
}
unsigned int rowSize() const
{
return nRow;
}
unsigned int colSize() const
{
return nCol;
}
void resize(int nbRow, int nbCol)
{
#ifdef SPARSEMATRIX_VERBOSE
if (nbRow != (int)rowSize() || nbCol != (int)colSize())
std::cout << /* this->Name() << */": resize("<<nbRow<<","<<nbCol<<")"<<std::endl;
#endif
nRow = nbRow;
nCol = nbCol;
resizeBloc((nRow + NL-1) / NL, (nCol + NC-1) / NC);
}
SReal element(int i, int j) const
{
#ifdef SPARSEMATRIX_CHECK
if ((unsigned)i >= (unsigned)rowSize() || (unsigned)j >= (unsigned)colSize())
{
std::cerr << "ERROR: invalid read access to element ("<<i<<","<<j<<") in "<</* this->Name() <<*/" of size ("<<rowSize()<<","<<colSize()<<")"<<std::endl;
return 0.0;
}
#endif
int bi=0, bj=0; split_row_index(i, bi); split_col_index(j, bj);
((Matrix*)this)->compress();
return traits::v(bloc(i, j), bi, bj);
}
void set(int i, int j, double v)
{
#ifdef SPARSEMATRIX_VERBOSE
std::cout << /* this->Name() << */"("<<rowSize()<<","<<colSize()<<"): element("<<i<<","<<j<<") = "<<v<<std::endl;
#endif
#ifdef SPARSEMATRIX_CHECK
if ((unsigned)i >= (unsigned)rowSize() || (unsigned)j >= (unsigned)colSize())
{
std::cerr << "ERROR: invalid write access to element ("<<i<<","<<j<<") in "<</* this->Name() <<*/" of size ("<<rowSize()<<","<<colSize()<<")"<<std::endl;
return;
}
#endif
int bi=0, bj=0; split_row_index(i, bi); split_col_index(j, bj);
#ifdef SPARSEMATRIX_VERBOSE
std::cout << /* this->Name() << */"("<<rowBSize()<<"*"<<NL<<","<<colBSize()<<"*"<<NC<<"): bloc("<<i<<","<<j<<")["<<bi<<","<<bj<<"] = "<<v<<std::endl;
#endif
traits::v(*wbloc(i,j,true), bi, bj) = (Real)v;
}
void add(int i, int j, double v)
{
#ifdef SPARSEMATRIX_VERBOSE
std::cout << /* this->Name() << */"("<<rowSize()<<","<<colSize()<<"): element("<<i<<","<<j<<") += "<<v<<std::endl;
#endif
#ifdef SPARSEMATRIX_CHECK
if ((unsigned)i >= (unsigned)rowSize() || (unsigned)j >= (unsigned)colSize())
{
std::cerr << "ERROR: invalid write access to element ("<<i<<","<<j<<") in "<</* this->Name() <<*/" of size ("<<rowSize()<<","<<colSize()<<")"<<std::endl;
return;
}
#endif
int bi=0, bj=0; split_row_index(i, bi); split_col_index(j, bj);
#ifdef SPARSEMATRIX_VERBOSE
std::cout << /* this->Name() << */"("<<rowBSize()<<"*"<<NL<<","<<colBSize()<<"*"<<NC<<"): bloc("<<i<<","<<j<<")["<<bi<<","<<bj<<"] += "<<v<<std::endl;
#endif
traits::v(*wbloc(i,j,true), bi, bj) += (Real)v;
}
void clear(int i, int j)
{
#ifdef SPARSEMATRIX_VERBOSE
std::cout << /* this->Name() << */"("<<rowSize()<<","<<colSize()<<"): element("<<i<<","<<j<<") = 0"<<std::endl;
#endif
#ifdef SPARSEMATRIX_CHECK
if ((unsigned)i >= (unsigned)rowSize() || (unsigned)j >= (unsigned)colSize())
{
std::cerr << "ERROR: invalid write access to element ("<<i<<","<<j<<") in "<</* this->Name() <<*/" of size ("<<rowSize()<<","<<colSize()<<")"<<std::endl;
return;
}
#endif
int bi=0, bj=0; split_row_index(i, bi); split_col_index(j, bj);
compress();
Bloc* b = wbloc(i,j,false);
if (b)
traits::v(*b, bi, bj) = 0;
}
void clearRow(int i)
{
#ifdef SPARSEMATRIX_VERBOSE
std::cout << /* this->Name() << */"("<<rowSize()<<","<<colSize()<<"): row("<<i<<") = 0"<<std::endl;
#endif
#ifdef SPARSEMATRIX_CHECK
if ((unsigned)i >= (unsigned)rowSize())
{
std::cerr << "ERROR: invalid write access to row "<<i<<" in "<</* this->Name() <<*/" of size ("<<rowSize()<<","<<colSize()<<")"<<std::endl;
return;
}
#endif
int bi=0; split_row_index(i, bi);
compress();
/*
for (int j=0; j<nBlocCol; ++j)
{
Bloc* b = wbloc(i,j,false);
if (b)
{
for (int bj = 0; bj < NC; ++bj)
traits::v(*b, bi, bj) = 0;
}
}
*/
int rowId = i * rowIndex.size() / nBlocRow;
if (sortedFind(rowIndex, i, rowId))
{
Range rowRange(rowBegin[rowId], rowBegin[rowId+1]);
for (int xj = rowRange.begin(); xj < rowRange.end(); ++xj)
{
Bloc& b = colsValue[xj];
for (int bj = 0; bj < NC; ++bj)
traits::v(b, bi, bj) = 0;
}
}
}
void clearCol(int j)
{
#ifdef SPARSEMATRIX_VERBOSE
std::cout << /* this->Name() << */"("<<rowSize()<<","<<colSize()<<"): col("<<j<<") = 0"<<std::endl;
#endif
#ifdef SPARSEMATRIX_CHECK
if ((unsigned)j >= (unsigned)colSize())
{
std::cerr << "ERROR: invalid write access to column "<<j<<" in "<</* this->Name() <<*/" of size ("<<rowSize()<<","<<colSize()<<")"<<std::endl;
return;
}
#endif
int bj=0; split_col_index(j, bj);
compress();
for (int i=0; i<nBlocRow; ++i)
{
Bloc* b = wbloc(i,j,false);
if (b)
{
for (int bi = 0; bi < NL; ++bi)
traits::v(*b, bi, bj) = 0;
}
}
}
void clearRowCol(int i)
{
#ifdef SPARSEMATRIX_VERBOSE
std::cout << /* this->Name() << */"("<<rowSize()<<","<<colSize()<<"): row("<<i<<") = 0 and col("<<i<<") = 0"<<std::endl;
#endif
#ifdef SPARSEMATRIX_CHECK
if ((unsigned)i >= (unsigned)rowSize() || (unsigned)i >= (unsigned)colSize())
{
std::cerr << "ERROR: invalid write access to row and column "<<i<<" in "<</* this->Name() <<*/" of size ("<<rowSize()<<","<<colSize()<<")"<<std::endl;
return;
}
#endif
if ((int)NL != (int)NC || nRow != nCol)
{
clearRow(i);
clearCol(i);
}
else
{
//std::cout << /* this->Name() << */"("<<rowSize()<<","<<colSize()<<"): sparse row("<<i<<") = 0 and col("<<i<<") = 0"<<std::endl;
// Here we assume the matrix is symmetric
int bi=0; split_row_index(i, bi);
compress();
int rowId = i * rowIndex.size() / nBlocRow;
if (sortedFind(rowIndex, i, rowId))
{
Range rowRange(rowBegin[rowId], rowBegin[rowId+1]);
for (int xj = rowRange.begin(); xj < rowRange.end(); ++xj)
{
Bloc& b = colsValue[xj];
for (int bj = 0; bj < NC; ++bj)
traits::v(b, bi, bj) = 0;
int j = colsIndex[xj];
if (j != i)
{ // non diagonal bloc
Bloc* b = wbloc(j,i,false);
if (b)
{
for (int bj = 0; bj < NL; ++bj)
traits::v(*b, bj, bi) = 0;
}
}
}
}
}
}
void clear()
{
for (unsigned int i=0; i < colsValue.size(); ++i)
traits::clear(colsValue[i]);
compressed = colsValue.empty();
btemp.clear();
}
protected:
template<class Real2>
static Real vget(const defaulttype::BaseVector& vec, int i) { return vec.element(i); }
template<class Real2> static Real2 vget(const FullVector<Real2>& vec, int i) { return vec[i]; }
static void vset(defaulttype::BaseVector& vec, int i, Real v) { vec.set(i, v); }
template<class Real2> static void vset(FullVector<Real2>& vec, int i, Real2 v) { vec[i] = v; }
static void vadd(defaulttype::BaseVector& vec, int i, Real v) { vec.add(i, v); }
template<class Real2> static void vadd(FullVector<Real2>& vec, int i, Real2 v) { vec[i] += v; }
public:
template<class Real2, class V1, class V2>
void tmul(V1& res, const V2& vec) const
{
((Matrix*)this)->compress();
res.resize(rowSize());
for (unsigned int xi = 0; xi < rowIndex.size(); ++xi)
{
Index iN = rowIndex[xi] * NL;
Range rowRange(rowBegin[xi], rowBegin[xi+1]);
defaulttype::Vec<NL,Real2> r;
for (int xj = rowRange.begin(); xj < rowRange.end(); ++xj)
{
Index jN = colsIndex[xj] * NC;
const Bloc& b = colsValue[xj];
defaulttype::Vec<NC,Real2> v;
for (int bj = 0; bj < NC; ++bj)
v[bj] = vget(vec,jN + bj);
for (int bi = 0; bi < NL; ++bi)
for (int bj = 0; bj < NC; ++bj)
r[bi] += traits::v(b, bi, bj) * v[bj];
}
for (int bi = 0; bi < NL; ++bi)
vset(res, iN + bi, r[bi]);
}
}
template<class Real2, class V1, class V2>
void tmulTranspose(V1& res, const V2& vec) const
{
((Matrix*)this)->compress();
res.resize(colSize());
for (unsigned int xi = 0; xi < rowIndex.size(); ++xi)
{
Index iN = rowIndex[xi] * NL;
Range rowRange(rowBegin[xi], rowBegin[xi+1]);
defaulttype::Vec<NL,Real2> v;
for (int bi = 0; bi < NL; ++bi)
v[bi] = vget(vec, iN + bi);
for (int xj = rowRange.begin(); xj < rowRange.end(); ++xj)
{
Index jN = colsIndex[xj] * NC;
const Bloc& b = colsValue[xj];
defaulttype::Vec<NC,Real2> r;
for (int bj = 0; bj < NC; ++bj)
r[bj] = traits::v(b, 0, bj) * v[0];
for (int bi = 1; bi < NL; ++bi)
for (int bj = 0; bj < NC; ++bj)
r[bj] += traits::v(b, bi, bj) * v[bi];
for (int bj = 0; bj < NC; ++bj)
vadd(res, jN + bj, r[bj]);
}
}
}
template<class Real2>
void mul(FullVector<Real2>& res, const FullVector<Real2>& v) const
{
tmul< Real2, FullVector<Real2>, FullVector<Real2> >(res, v);
}
template<class Real2>
void mulTranspose(FullVector<Real2>& res, const FullVector<Real2>& v) const
{
tmulTranspose< Real2, FullVector<Real2>, FullVector<Real2> >(res, v);
}
template<class Real2>
void mul(FullVector<Real2>& res, const defaulttype::BaseVector* v) const
{
tmul< Real2, FullVector<Real2>, defaulttype::BaseVector >(res, *v);
}
template<class Real2>
void mulTranspose(FullVector<Real2>& res, const defaulttype::BaseVector* v) const
{
tmulTranspose< Real2, FullVector<Real2>, defaulttype::BaseVector >(res, *v);
}
template<class Real2>
void mul(defaulttype::BaseVector* res, const FullVector<Real2>& v) const
{
tmul< Real2, defaulttype::BaseVector, FullVector<Real2> >(*res, v);
}
template<class Real2>
void mulTranspose(defaulttype::BaseVector*& res, const FullVector<Real2>& v) const
{
tmulTranspose< Real2, defaulttype::BaseVector, FullVector<Real2> >(*res, v);
}
template<class Real2>
void mul(defaulttype::BaseVector* res, const defaulttype::BaseVector* v) const
{
tmul< Real, defaulttype::BaseVector, defaulttype::BaseVector >(*res, *v);
}
void mulTranspose(defaulttype::BaseVector* res, const defaulttype::BaseVector* v) const
{
tmul< Real, defaulttype::BaseVector, defaulttype::BaseVector >(*res, *v);
}
template<class Real2>
FullVector<Real2> operator*(const FullVector<Real2>& v) const
{
FullVector<Real2> res;
mul(res,v);
return res;
}
friend std::ostream& operator << (std::ostream& out, const Matrix& v )
{
int nx = v.colSize();
int ny = v.rowSize();
out << "[";
for (int y=0;y<ny;++y)
{
out << "\n[";
for (int x=0;x<nx;++x)
{
out << " " << v.element(y,x);
}
out << " ]";
}
out << " ]";
return out;
}
static const char* Name()
{
static std::string name = std::string("CompressedRowSparseMatrix") + std::string(traits::Name());
return name.c_str();
}
bool check_matrix() {
return check_matrix(
this->getColsValue().size(),
this->rowBSize(),
this->colBSize(),
(int *) &(this->getRowBegin()[0]),
(int *) &(this->getColsIndex()[0]),
(double *) &(this->getColsValue()[0])
);
}
static bool check_matrix(
int nzmax,// nb values
int m,// number of row
int n,// number of columns
int * a_p,// column pointers (size n+1) or col indices (size nzmax)
int * a_i,// row indices, size nzmax
double * a_x// numerical values, size nzmax
) {
// check ap, size m beecause ther is at least the diagonal value wich is different of 0
if (a_p[0]!=0) {
std::cerr << "CompressedRowSparseMatrix:First value of row indices (a_p) should be 0" << std::endl;
return false;
}
for (int i=1;i<=m;i++) {
if (a_p[i]<=a_p[i-1]) {
std::cerr << "CompressedRowSparseMatrix:Row (a_p) indices are not sorted indice " << i-1 << " : " << a_p[i-1] << " , " << i << " : " << a_p[i] << std::endl;
return false;
}
}
if (nzmax == -1) {
nzmax = a_p[m];
} else if (a_p[m]!=nzmax) {
std::cerr << "CompressedRowSparseMatrix:Last value of row indices (a_p) should be " << nzmax << " and is " << a_p[m] << std::endl;
return false;
}
int k=1;
for (int i=0;i<nzmax;i++) {
i++;
for (;i<a_p[k];i++) {
if (a_i[i] <= a_i[i-1]) {
std::cerr << "CompressedRowSparseMatrix:Column (a_i) indices are not sorted indice " << i-1 << " : " << a_i[i-1] << " , " << i << " : " << a_p[i] << std::endl;
return false;
}
if (a_i[i]<0 || a_i[i]>=n) {
std::cerr << "CompressedRowSparseMatrix:Column (a_i) indices are not correct " << i << " : " << a_i[i] << std::endl;
return false;
}
}
k++;
}
for (int i=0;i<nzmax;i++) {
if (a_x[i]==0) {
std::cerr << "CompressedRowSparseMatrix:Warning , matrix contains 0 , indice " << i << std::endl;
return false;
}
}
if (n!=m) {
std::cerr << "CompressedRowSparseMatrix:the matrix is not square" << std::endl;
return false;
}
std::cerr << "Check_matrix passed succefull" << std::endl;
return true;
}
};
#ifdef SPARSEMATRIX_CHECK
#undef SPARSEMATRIX_CHECK
#endif
#ifdef SPARSEMATRIX_VERBOSE
#undef SPARSEMATRIX_VERBOSE
#endif
} // namespace linearsolver
} // namespace component
} // namespace sofa
#endif
|