1 2 /* 3 This is where the abstract matrix operations are defined 4 */ 5 6 #include <petsc-private/matimpl.h> /*I "petscmat.h" I*/ 7 #include <petsc-private/vecimpl.h> 8 9 /* Logging support */ 10 PetscClassId MAT_CLASSID; 11 PetscClassId MAT_FDCOLORING_CLASSID; 12 PetscClassId MAT_TRANSPOSECOLORING_CLASSID; 13 14 PetscLogEvent MAT_Mult, MAT_Mults, MAT_MultConstrained, MAT_MultAdd, MAT_MultTranspose; 15 PetscLogEvent MAT_MultTransposeConstrained, MAT_MultTransposeAdd, MAT_Solve, MAT_Solves, MAT_SolveAdd, MAT_SolveTranspose, MAT_MatSolve; 16 PetscLogEvent MAT_SolveTransposeAdd, MAT_SOR, MAT_ForwardSolve, MAT_BackwardSolve, MAT_LUFactor, MAT_LUFactorSymbolic; 17 PetscLogEvent MAT_LUFactorNumeric, MAT_CholeskyFactor, MAT_CholeskyFactorSymbolic, MAT_CholeskyFactorNumeric, MAT_ILUFactor; 18 PetscLogEvent MAT_ILUFactorSymbolic, MAT_ICCFactorSymbolic, MAT_Copy, MAT_Convert, MAT_Scale, MAT_AssemblyBegin; 19 PetscLogEvent MAT_AssemblyEnd, MAT_SetValues, MAT_GetValues, MAT_GetRow, MAT_GetRowIJ, MAT_GetSubMatrices, MAT_GetColoring, MAT_GetOrdering, MAT_GetRedundantMatrix, MAT_GetSeqNonzeroStructure; 20 PetscLogEvent MAT_IncreaseOverlap, MAT_Partitioning, MAT_Coarsen, MAT_ZeroEntries, MAT_Load, MAT_View, MAT_AXPY, MAT_FDColoringCreate; 21 PetscLogEvent MAT_FDColoringApply,MAT_Transpose,MAT_FDColoringFunction; 22 PetscLogEvent MAT_TransposeColoringCreate; 23 PetscLogEvent MAT_MatMult, MAT_MatMultSymbolic, MAT_MatMultNumeric; 24 PetscLogEvent MAT_PtAP, MAT_PtAPSymbolic, MAT_PtAPNumeric,MAT_RARt, MAT_RARtSymbolic, MAT_RARtNumeric; 25 PetscLogEvent MAT_MatTransposeMult, MAT_MatTransposeMultSymbolic, MAT_MatTransposeMultNumeric; 26 PetscLogEvent MAT_TransposeMatMult, MAT_TransposeMatMultSymbolic, MAT_TransposeMatMultNumeric; 27 PetscLogEvent MAT_MultHermitianTranspose,MAT_MultHermitianTransposeAdd; 28 PetscLogEvent MAT_Getsymtranspose, MAT_Getsymtransreduced, MAT_Transpose_SeqAIJ, MAT_GetBrowsOfAcols; 29 PetscLogEvent MAT_GetBrowsOfAocols, MAT_Getlocalmat, MAT_Getlocalmatcondensed, MAT_Seqstompi, MAT_Seqstompinum, MAT_Seqstompisym; 30 PetscLogEvent MAT_Applypapt, MAT_Applypapt_numeric, MAT_Applypapt_symbolic, MAT_GetSequentialNonzeroStructure; 31 PetscLogEvent MAT_GetMultiProcBlock; 32 PetscLogEvent MAT_CUSPCopyToGPU, MAT_CUSPARSECopyToGPU, MAT_SetValuesBatch, MAT_SetValuesBatchI, MAT_SetValuesBatchII, MAT_SetValuesBatchIII, MAT_SetValuesBatchIV; 33 PetscLogEvent MAT_Merge; 34 35 const char *const MatFactorTypes[] = {"NONE","LU","CHOLESKY","ILU","ICC","ILUDT","MatFactorType","MAT_FACTOR_",0}; 36 37 #undef __FUNCT__ 38 #define __FUNCT__ "MatFindNonzeroRows" 39 /*@ 40 MatFindNonzeroRows - Locate all rows that are not completely zero in the matrix 41 42 Input Parameter: 43 . A - the matrix 44 45 Output Parameter: 46 . keptrows - the rows that are not completely zero 47 48 Level: intermediate 49 50 @*/ 51 PetscErrorCode MatFindNonzeroRows(Mat mat,IS *keptrows) 52 { 53 PetscErrorCode ierr; 54 55 PetscValidHeaderSpecific(mat,MAT_CLASSID,1); 56 PetscValidType(mat,1); 57 if (!mat->assembled) SETERRQ(((PetscObject)mat)->comm,PETSC_ERR_ARG_WRONGSTATE,"Not for unassembled matrix"); 58 if (mat->factortype) SETERRQ(((PetscObject)mat)->comm,PETSC_ERR_ARG_WRONGSTATE,"Not for factored matrix"); 59 if (!mat->ops->findnonzerorows) SETERRQ(((PetscObject)mat)->comm,PETSC_ERR_SUP,"Not coded for this matrix type"); 60 ierr = (*mat->ops->findnonzerorows)(mat,keptrows);CHKERRQ(ierr); 61 PetscFunctionReturn(0); 62 } 63 64 #undef __FUNCT__ 65 #define __FUNCT__ "MatGetDiagonalBlock" 66 /*@ 67 MatGetDiagonalBlock - Returns the part of the matrix associated with the on-process coupling 68 69 Not Collective 70 71 Input Parameters: 72 . A - the matrix 73 74 Output Parameters: 75 . a - the diagonal part (which is a SEQUENTIAL matrix) 76 77 Notes: see the manual page for MatCreateAIJ() for more information on the "diagonal part" of the matrix. 78 79 Level: advanced 80 81 @*/ 82 PetscErrorCode MatGetDiagonalBlock(Mat A,Mat *a) 83 { 84 PetscErrorCode ierr,(*f)(Mat,Mat*); 85 PetscMPIInt size; 86 87 PetscFunctionBegin; 88 PetscValidHeaderSpecific(A,MAT_CLASSID,1); 89 PetscValidType(A,1); 90 PetscValidPointer(a,3); 91 if (!A->assembled) SETERRQ(((PetscObject)A)->comm,PETSC_ERR_ARG_WRONGSTATE,"Not for unassembled matrix"); 92 if (A->factortype) SETERRQ(((PetscObject)A)->comm,PETSC_ERR_ARG_WRONGSTATE,"Not for factored matrix"); 93 ierr = MPI_Comm_size(((PetscObject)A)->comm,&size);CHKERRQ(ierr); 94 ierr = PetscObjectQueryFunction((PetscObject)A,"MatGetDiagonalBlock_C",(void (**)(void))&f);CHKERRQ(ierr); 95 if (f) { 96 ierr = (*f)(A,a);CHKERRQ(ierr); 97 PetscFunctionReturn(0); 98 } else if (size == 1) { 99 *a = A; 100 } else { 101 const MatType mattype; 102 ierr = MatGetType(A,&mattype);CHKERRQ(ierr); 103 SETERRQ1(((PetscObject)A)->comm,PETSC_ERR_SUP,"Matrix type %s does not support getting diagonal block",mattype); 104 } 105 PetscFunctionReturn(0); 106 } 107 108 #undef __FUNCT__ 109 #define __FUNCT__ "MatGetTrace" 110 /*@ 111 MatGetTrace - Gets the trace of a matrix. The sum of the diagonal entries. 112 113 Collective on Mat 114 115 Input Parameters: 116 . mat - the matrix 117 118 Output Parameter: 119 . trace - the sum of the diagonal entries 120 121 Level: advanced 122 123 @*/ 124 PetscErrorCode MatGetTrace(Mat mat,PetscScalar *trace) 125 { 126 PetscErrorCode ierr; 127 Vec diag; 128 129 PetscFunctionBegin; 130 ierr = MatGetVecs(mat,&diag,PETSC_NULL);CHKERRQ(ierr); 131 ierr = MatGetDiagonal(mat,diag);CHKERRQ(ierr); 132 ierr = VecSum(diag,trace);CHKERRQ(ierr); 133 ierr = VecDestroy(&diag);CHKERRQ(ierr); 134 PetscFunctionReturn(0); 135 } 136 137 #undef __FUNCT__ 138 #define __FUNCT__ "MatRealPart" 139 /*@ 140 MatRealPart - Zeros out the imaginary part of the matrix 141 142 Logically Collective on Mat 143 144 Input Parameters: 145 . mat - the matrix 146 147 Level: advanced 148 149 150 .seealso: MatImaginaryPart() 151 @*/ 152 PetscErrorCode MatRealPart(Mat mat) 153 { 154 PetscErrorCode ierr; 155 156 PetscFunctionBegin; 157 PetscValidHeaderSpecific(mat,MAT_CLASSID,1); 158 PetscValidType(mat,1); 159 if (!mat->assembled) SETERRQ(((PetscObject)mat)->comm,PETSC_ERR_ARG_WRONGSTATE,"Not for unassembled matrix"); 160 if (mat->factortype) SETERRQ(((PetscObject)mat)->comm,PETSC_ERR_ARG_WRONGSTATE,"Not for factored matrix"); 161 if (!mat->ops->realpart) SETERRQ1(((PetscObject)mat)->comm,PETSC_ERR_SUP,"Mat type %s",((PetscObject)mat)->type_name); 162 MatCheckPreallocated(mat,1); 163 ierr = (*mat->ops->realpart)(mat);CHKERRQ(ierr); 164 #if defined(PETSC_HAVE_CUSP) 165 if (mat->valid_GPU_matrix != PETSC_CUSP_UNALLOCATED) { 166 mat->valid_GPU_matrix = PETSC_CUSP_CPU; 167 } 168 #endif 169 PetscFunctionReturn(0); 170 } 171 172 #undef __FUNCT__ 173 #define __FUNCT__ "MatGetGhosts" 174 /*@C 175 MatGetGhosts - Get the global index of all ghost nodes defined by the sparse matrix 176 177 Collective on Mat 178 179 Input Parameter: 180 . mat - the matrix 181 182 Output Parameters: 183 + nghosts - number of ghosts (note for BAIJ matrices there is one ghost for each block) 184 - ghosts - the global indices of the ghost points 185 186 Notes: the nghosts and ghosts are suitable to pass into VecCreateGhost() 187 188 Level: advanced 189 190 @*/ 191 PetscErrorCode MatGetGhosts(Mat mat,PetscInt *nghosts,const PetscInt *ghosts[]) 192 { 193 PetscErrorCode ierr; 194 195 PetscFunctionBegin; 196 PetscValidHeaderSpecific(mat,MAT_CLASSID,1); 197 PetscValidType(mat,1); 198 if (!mat->assembled) SETERRQ(((PetscObject)mat)->comm,PETSC_ERR_ARG_WRONGSTATE,"Not for unassembled matrix"); 199 if (mat->factortype) SETERRQ(((PetscObject)mat)->comm,PETSC_ERR_ARG_WRONGSTATE,"Not for factored matrix"); 200 if (!mat->ops->getghosts) { 201 if (nghosts) *nghosts = 0; 202 if (ghosts) *ghosts = 0; 203 } else { 204 ierr = (*mat->ops->getghosts)(mat,nghosts,ghosts);CHKERRQ(ierr); 205 } 206 PetscFunctionReturn(0); 207 } 208 209 210 #undef __FUNCT__ 211 #define __FUNCT__ "MatImaginaryPart" 212 /*@ 213 MatImaginaryPart - Moves the imaginary part of the matrix to the real part and zeros the imaginary part 214 215 Logically Collective on Mat 216 217 Input Parameters: 218 . mat - the matrix 219 220 Level: advanced 221 222 223 .seealso: MatRealPart() 224 @*/ 225 PetscErrorCode MatImaginaryPart(Mat mat) 226 { 227 PetscErrorCode ierr; 228 229 PetscFunctionBegin; 230 PetscValidHeaderSpecific(mat,MAT_CLASSID,1); 231 PetscValidType(mat,1); 232 if (!mat->assembled) SETERRQ(((PetscObject)mat)->comm,PETSC_ERR_ARG_WRONGSTATE,"Not for unassembled matrix"); 233 if (mat->factortype) SETERRQ(((PetscObject)mat)->comm,PETSC_ERR_ARG_WRONGSTATE,"Not for factored matrix"); 234 if (!mat->ops->imaginarypart) SETERRQ1(((PetscObject)mat)->comm,PETSC_ERR_SUP,"Mat type %s",((PetscObject)mat)->type_name); 235 MatCheckPreallocated(mat,1); 236 ierr = (*mat->ops->imaginarypart)(mat);CHKERRQ(ierr); 237 #if defined(PETSC_HAVE_CUSP) 238 if (mat->valid_GPU_matrix != PETSC_CUSP_UNALLOCATED) { 239 mat->valid_GPU_matrix = PETSC_CUSP_CPU; 240 } 241 #endif 242 PetscFunctionReturn(0); 243 } 244 245 #undef __FUNCT__ 246 #define __FUNCT__ "MatMissingDiagonal" 247 /*@ 248 MatMissingDiagonal - Determine if sparse matrix is missing a diagonal entry (or block entry for BAIJ matrices) 249 250 Collective on Mat 251 252 Input Parameter: 253 . mat - the matrix 254 255 Output Parameters: 256 + missing - is any diagonal missing 257 - dd - first diagonal entry that is missing (optional) 258 259 Level: advanced 260 261 262 .seealso: MatRealPart() 263 @*/ 264 PetscErrorCode MatMissingDiagonal(Mat mat,PetscBool *missing,PetscInt *dd) 265 { 266 PetscErrorCode ierr; 267 268 PetscFunctionBegin; 269 PetscValidHeaderSpecific(mat,MAT_CLASSID,1); 270 PetscValidType(mat,1); 271 if (!mat->assembled) SETERRQ(((PetscObject)mat)->comm,PETSC_ERR_ARG_WRONGSTATE,"Not for unassembled matrix"); 272 if (mat->factortype) SETERRQ(((PetscObject)mat)->comm,PETSC_ERR_ARG_WRONGSTATE,"Not for factored matrix"); 273 if (!mat->ops->missingdiagonal) SETERRQ1(((PetscObject)mat)->comm,PETSC_ERR_SUP,"Mat type %s",((PetscObject)mat)->type_name); 274 ierr = (*mat->ops->missingdiagonal)(mat,missing,dd);CHKERRQ(ierr); 275 PetscFunctionReturn(0); 276 } 277 278 #undef __FUNCT__ 279 #define __FUNCT__ "MatGetRow" 280 /*@C 281 MatGetRow - Gets a row of a matrix. You MUST call MatRestoreRow() 282 for each row that you get to ensure that your application does 283 not bleed memory. 284 285 Not Collective 286 287 Input Parameters: 288 + mat - the matrix 289 - row - the row to get 290 291 Output Parameters: 292 + ncols - if not NULL, the number of nonzeros in the row 293 . cols - if not NULL, the column numbers 294 - vals - if not NULL, the values 295 296 Notes: 297 This routine is provided for people who need to have direct access 298 to the structure of a matrix. We hope that we provide enough 299 high-level matrix routines that few users will need it. 300 301 MatGetRow() always returns 0-based column indices, regardless of 302 whether the internal representation is 0-based (default) or 1-based. 303 304 For better efficiency, set cols and/or vals to PETSC_NULL if you do 305 not wish to extract these quantities. 306 307 The user can only examine the values extracted with MatGetRow(); 308 the values cannot be altered. To change the matrix entries, one 309 must use MatSetValues(). 310 311 You can only have one call to MatGetRow() outstanding for a particular 312 matrix at a time, per processor. MatGetRow() can only obtain rows 313 associated with the given processor, it cannot get rows from the 314 other processors; for that we suggest using MatGetSubMatrices(), then 315 MatGetRow() on the submatrix. The row indix passed to MatGetRows() 316 is in the global number of rows. 317 318 Fortran Notes: 319 The calling sequence from Fortran is 320 .vb 321 MatGetRow(matrix,row,ncols,cols,values,ierr) 322 Mat matrix (input) 323 integer row (input) 324 integer ncols (output) 325 integer cols(maxcols) (output) 326 double precision (or double complex) values(maxcols) output 327 .ve 328 where maxcols >= maximum nonzeros in any row of the matrix. 329 330 331 Caution: 332 Do not try to change the contents of the output arrays (cols and vals). 333 In some cases, this may corrupt the matrix. 334 335 Level: advanced 336 337 Concepts: matrices^row access 338 339 .seealso: MatRestoreRow(), MatSetValues(), MatGetValues(), MatGetSubMatrices(), MatGetDiagonal() 340 @*/ 341 PetscErrorCode MatGetRow(Mat mat,PetscInt row,PetscInt *ncols,const PetscInt *cols[],const PetscScalar *vals[]) 342 { 343 PetscErrorCode ierr; 344 PetscInt incols; 345 346 PetscFunctionBegin; 347 PetscValidHeaderSpecific(mat,MAT_CLASSID,1); 348 PetscValidType(mat,1); 349 if (!mat->assembled) SETERRQ(PETSC_COMM_SELF,PETSC_ERR_ARG_WRONGSTATE,"Not for unassembled matrix"); 350 if (mat->factortype) SETERRQ(PETSC_COMM_SELF,PETSC_ERR_ARG_WRONGSTATE,"Not for factored matrix"); 351 if (!mat->ops->getrow) SETERRQ1(PETSC_COMM_SELF,PETSC_ERR_SUP,"Mat type %s",((PetscObject)mat)->type_name); 352 MatCheckPreallocated(mat,1); 353 ierr = PetscLogEventBegin(MAT_GetRow,mat,0,0,0);CHKERRQ(ierr); 354 ierr = (*mat->ops->getrow)(mat,row,&incols,(PetscInt **)cols,(PetscScalar **)vals);CHKERRQ(ierr); 355 if (ncols) *ncols = incols; 356 ierr = PetscLogEventEnd(MAT_GetRow,mat,0,0,0);CHKERRQ(ierr); 357 PetscFunctionReturn(0); 358 } 359 360 #undef __FUNCT__ 361 #define __FUNCT__ "MatConjugate" 362 /*@ 363 MatConjugate - replaces the matrix values with their complex conjugates 364 365 Logically Collective on Mat 366 367 Input Parameters: 368 . mat - the matrix 369 370 Level: advanced 371 372 .seealso: VecConjugate() 373 @*/ 374 PetscErrorCode MatConjugate(Mat mat) 375 { 376 #ifdef PETSC_USE_COMPLEX 377 PetscErrorCode ierr; 378 379 PetscFunctionBegin; 380 PetscValidHeaderSpecific(mat,MAT_CLASSID,1); 381 if (!mat->assembled) SETERRQ(((PetscObject)mat)->comm,PETSC_ERR_ARG_WRONGSTATE,"Not for unassembled matrix"); 382 if (!mat->ops->conjugate) SETERRQ(((PetscObject)mat)->comm,PETSC_ERR_SUP,"Not provided for this matrix format, send email to petsc-maint@mcs.anl.gov"); 383 ierr = (*mat->ops->conjugate)(mat);CHKERRQ(ierr); 384 #if defined(PETSC_HAVE_CUSP) 385 if (mat->valid_GPU_matrix != PETSC_CUSP_UNALLOCATED) { 386 mat->valid_GPU_matrix = PETSC_CUSP_CPU; 387 } 388 #endif 389 PetscFunctionReturn(0); 390 #else 391 return 0; 392 #endif 393 } 394 395 #undef __FUNCT__ 396 #define __FUNCT__ "MatRestoreRow" 397 /*@C 398 MatRestoreRow - Frees any temporary space allocated by MatGetRow(). 399 400 Not Collective 401 402 Input Parameters: 403 + mat - the matrix 404 . row - the row to get 405 . ncols, cols - the number of nonzeros and their columns 406 - vals - if nonzero the column values 407 408 Notes: 409 This routine should be called after you have finished examining the entries. 410 411 Fortran Notes: 412 The calling sequence from Fortran is 413 .vb 414 MatRestoreRow(matrix,row,ncols,cols,values,ierr) 415 Mat matrix (input) 416 integer row (input) 417 integer ncols (output) 418 integer cols(maxcols) (output) 419 double precision (or double complex) values(maxcols) output 420 .ve 421 Where maxcols >= maximum nonzeros in any row of the matrix. 422 423 In Fortran MatRestoreRow() MUST be called after MatGetRow() 424 before another call to MatGetRow() can be made. 425 426 Level: advanced 427 428 .seealso: MatGetRow() 429 @*/ 430 PetscErrorCode MatRestoreRow(Mat mat,PetscInt row,PetscInt *ncols,const PetscInt *cols[],const PetscScalar *vals[]) 431 { 432 PetscErrorCode ierr; 433 434 PetscFunctionBegin; 435 PetscValidHeaderSpecific(mat,MAT_CLASSID,1); 436 if (ncols) {PetscValidIntPointer(ncols,3);} 437 if (!mat->assembled) SETERRQ(PETSC_COMM_SELF,PETSC_ERR_ARG_WRONGSTATE,"Not for unassembled matrix"); 438 if (!mat->ops->restorerow) PetscFunctionReturn(0); 439 ierr = (*mat->ops->restorerow)(mat,row,ncols,(PetscInt **)cols,(PetscScalar **)vals);CHKERRQ(ierr); 440 PetscFunctionReturn(0); 441 } 442 443 #undef __FUNCT__ 444 #define __FUNCT__ "MatGetRowUpperTriangular" 445 /*@ 446 MatGetRowUpperTriangular - Sets a flag to enable calls to MatGetRow() for matrix in MATSBAIJ format. 447 You should call MatRestoreRowUpperTriangular() after calling MatGetRow/MatRestoreRow() to disable the flag. 448 449 Not Collective 450 451 Input Parameters: 452 + mat - the matrix 453 454 Notes: 455 The flag is to ensure that users are aware of MatGetRow() only provides the upper trianglular part of the row for the matrices in MATSBAIJ format. 456 457 Level: advanced 458 459 Concepts: matrices^row access 460 461 .seealso: MatRestoreRowRowUpperTriangular() 462 @*/ 463 PetscErrorCode MatGetRowUpperTriangular(Mat mat) 464 { 465 PetscErrorCode ierr; 466 467 PetscFunctionBegin; 468 PetscValidHeaderSpecific(mat,MAT_CLASSID,1); 469 PetscValidType(mat,1); 470 if (!mat->assembled) SETERRQ(PETSC_COMM_SELF,PETSC_ERR_ARG_WRONGSTATE,"Not for unassembled matrix"); 471 if (mat->factortype) SETERRQ(PETSC_COMM_SELF,PETSC_ERR_ARG_WRONGSTATE,"Not for factored matrix"); 472 if (!mat->ops->getrowuppertriangular) SETERRQ1(PETSC_COMM_SELF,PETSC_ERR_SUP,"Mat type %s",((PetscObject)mat)->type_name); 473 MatCheckPreallocated(mat,1); 474 ierr = (*mat->ops->getrowuppertriangular)(mat);CHKERRQ(ierr); 475 PetscFunctionReturn(0); 476 } 477 478 #undef __FUNCT__ 479 #define __FUNCT__ "MatRestoreRowUpperTriangular" 480 /*@ 481 MatRestoreRowUpperTriangular - Disable calls to MatGetRow() for matrix in MATSBAIJ format. 482 483 Not Collective 484 485 Input Parameters: 486 + mat - the matrix 487 488 Notes: 489 This routine should be called after you have finished MatGetRow/MatRestoreRow(). 490 491 492 Level: advanced 493 494 .seealso: MatGetRowUpperTriangular() 495 @*/ 496 PetscErrorCode MatRestoreRowUpperTriangular(Mat mat) 497 { 498 PetscErrorCode ierr; 499 500 PetscFunctionBegin; 501 PetscValidHeaderSpecific(mat,MAT_CLASSID,1); 502 if (!mat->assembled) SETERRQ(PETSC_COMM_SELF,PETSC_ERR_ARG_WRONGSTATE,"Not for unassembled matrix"); 503 if (!mat->ops->restorerowuppertriangular) PetscFunctionReturn(0); 504 ierr = (*mat->ops->restorerowuppertriangular)(mat);CHKERRQ(ierr); 505 PetscFunctionReturn(0); 506 } 507 508 #undef __FUNCT__ 509 #define __FUNCT__ "MatSetOptionsPrefix" 510 /*@C 511 MatSetOptionsPrefix - Sets the prefix used for searching for all 512 Mat options in the database. 513 514 Logically Collective on Mat 515 516 Input Parameter: 517 + A - the Mat context 518 - prefix - the prefix to prepend to all option names 519 520 Notes: 521 A hyphen (-) must NOT be given at the beginning of the prefix name. 522 The first character of all runtime options is AUTOMATICALLY the hyphen. 523 524 Level: advanced 525 526 .keywords: Mat, set, options, prefix, database 527 528 .seealso: MatSetFromOptions() 529 @*/ 530 PetscErrorCode MatSetOptionsPrefix(Mat A,const char prefix[]) 531 { 532 PetscErrorCode ierr; 533 534 PetscFunctionBegin; 535 PetscValidHeaderSpecific(A,MAT_CLASSID,1); 536 ierr = PetscObjectSetOptionsPrefix((PetscObject)A,prefix);CHKERRQ(ierr); 537 PetscFunctionReturn(0); 538 } 539 540 #undef __FUNCT__ 541 #define __FUNCT__ "MatAppendOptionsPrefix" 542 /*@C 543 MatAppendOptionsPrefix - Appends to the prefix used for searching for all 544 Mat options in the database. 545 546 Logically Collective on Mat 547 548 Input Parameters: 549 + A - the Mat context 550 - prefix - the prefix to prepend to all option names 551 552 Notes: 553 A hyphen (-) must NOT be given at the beginning of the prefix name. 554 The first character of all runtime options is AUTOMATICALLY the hyphen. 555 556 Level: advanced 557 558 .keywords: Mat, append, options, prefix, database 559 560 .seealso: MatGetOptionsPrefix() 561 @*/ 562 PetscErrorCode MatAppendOptionsPrefix(Mat A,const char prefix[]) 563 { 564 PetscErrorCode ierr; 565 566 PetscFunctionBegin; 567 PetscValidHeaderSpecific(A,MAT_CLASSID,1); 568 ierr = PetscObjectAppendOptionsPrefix((PetscObject)A,prefix);CHKERRQ(ierr); 569 PetscFunctionReturn(0); 570 } 571 572 #undef __FUNCT__ 573 #define __FUNCT__ "MatGetOptionsPrefix" 574 /*@C 575 MatGetOptionsPrefix - Sets the prefix used for searching for all 576 Mat options in the database. 577 578 Not Collective 579 580 Input Parameter: 581 . A - the Mat context 582 583 Output Parameter: 584 . prefix - pointer to the prefix string used 585 586 Notes: On the fortran side, the user should pass in a string 'prefix' of 587 sufficient length to hold the prefix. 588 589 Level: advanced 590 591 .keywords: Mat, get, options, prefix, database 592 593 .seealso: MatAppendOptionsPrefix() 594 @*/ 595 PetscErrorCode MatGetOptionsPrefix(Mat A,const char *prefix[]) 596 { 597 PetscErrorCode ierr; 598 599 PetscFunctionBegin; 600 PetscValidHeaderSpecific(A,MAT_CLASSID,1); 601 ierr = PetscObjectGetOptionsPrefix((PetscObject)A,prefix);CHKERRQ(ierr); 602 PetscFunctionReturn(0); 603 } 604 605 #undef __FUNCT__ 606 #define __FUNCT__ "MatSetUp" 607 /*@ 608 MatSetUp - Sets up the internal matrix data structures for the later use. 609 610 Collective on Mat 611 612 Input Parameters: 613 . A - the Mat context 614 615 Notes: 616 If the user has not set preallocation for this matrix then a default preallocation that is likely to be inefficient is used. 617 618 If a suitable preallocation routine is used, this function does not need to be called. 619 620 See the Performance chapter of the PETSc users manual for how to preallocate matrices 621 622 Level: beginner 623 624 .keywords: Mat, setup 625 626 .seealso: MatCreate(), MatDestroy() 627 @*/ 628 PetscErrorCode MatSetUp(Mat A) 629 { 630 PetscMPIInt size; 631 PetscErrorCode ierr; 632 633 PetscFunctionBegin; 634 PetscValidHeaderSpecific(A,MAT_CLASSID,1); 635 if (!((PetscObject)A)->type_name) { 636 ierr = MPI_Comm_size(((PetscObject)A)->comm, &size);CHKERRQ(ierr); 637 if (size == 1) { 638 ierr = MatSetType(A, MATSEQAIJ);CHKERRQ(ierr); 639 } else { 640 ierr = MatSetType(A, MATMPIAIJ);CHKERRQ(ierr); 641 } 642 } 643 if (!A->preallocated && A->ops->setup) { 644 ierr = PetscInfo(A,"Warning not preallocating matrix storage\n");CHKERRQ(ierr); 645 ierr = (*A->ops->setup)(A);CHKERRQ(ierr); 646 } 647 A->preallocated = PETSC_TRUE; 648 PetscFunctionReturn(0); 649 } 650 651 652 #undef __FUNCT__ 653 #define __FUNCT__ "MatView" 654 /*@C 655 MatView - Visualizes a matrix object. 656 657 Collective on Mat 658 659 Input Parameters: 660 + mat - the matrix 661 - viewer - visualization context 662 663 Notes: 664 The available visualization contexts include 665 + PETSC_VIEWER_STDOUT_SELF - standard output (default) 666 . PETSC_VIEWER_STDOUT_WORLD - synchronized standard 667 output where only the first processor opens 668 the file. All other processors send their 669 data to the first processor to print. 670 - PETSC_VIEWER_DRAW_WORLD - graphical display of nonzero structure 671 672 The user can open alternative visualization contexts with 673 + PetscViewerASCIIOpen() - Outputs matrix to a specified file 674 . PetscViewerBinaryOpen() - Outputs matrix in binary to a 675 specified file; corresponding input uses MatLoad() 676 . PetscViewerDrawOpen() - Outputs nonzero matrix structure to 677 an X window display 678 - PetscViewerSocketOpen() - Outputs matrix to Socket viewer. 679 Currently only the sequential dense and AIJ 680 matrix types support the Socket viewer. 681 682 The user can call PetscViewerSetFormat() to specify the output 683 format of ASCII printed objects (when using PETSC_VIEWER_STDOUT_SELF, 684 PETSC_VIEWER_STDOUT_WORLD and PetscViewerASCIIOpen). Available formats include 685 + PETSC_VIEWER_DEFAULT - default, prints matrix contents 686 . PETSC_VIEWER_ASCII_MATLAB - prints matrix contents in Matlab format 687 . PETSC_VIEWER_ASCII_DENSE - prints entire matrix including zeros 688 . PETSC_VIEWER_ASCII_COMMON - prints matrix contents, using a sparse 689 format common among all matrix types 690 . PETSC_VIEWER_ASCII_IMPL - prints matrix contents, using an implementation-specific 691 format (which is in many cases the same as the default) 692 . PETSC_VIEWER_ASCII_INFO - prints basic information about the matrix 693 size and structure (not the matrix entries) 694 . PETSC_VIEWER_ASCII_INFO_DETAIL - prints more detailed information about 695 the matrix structure 696 697 Options Database Keys: 698 + -mat_view_info - Prints info on matrix at conclusion of MatEndAssembly() 699 . -mat_view_info_detailed - Prints more detailed info 700 . -mat_view - Prints matrix in ASCII format 701 . -mat_view_matlab - Prints matrix in Matlab format 702 . -mat_view_draw - PetscDraws nonzero structure of matrix, using MatView() and PetscDrawOpenX(). 703 . -display <name> - Sets display name (default is host) 704 . -draw_pause <sec> - Sets number of seconds to pause after display 705 . -mat_view_socket - Sends matrix to socket, can be accessed from Matlab (see the <a href="../../docs/manual.pdf">users manual</a> for details). 706 . -viewer_socket_machine <machine> 707 . -viewer_socket_port <port> 708 . -mat_view_binary - save matrix to file in binary format 709 - -viewer_binary_filename <name> 710 Level: beginner 711 712 Notes: see the manual page for MatLoad() for the exact format of the binary file when the binary 713 viewer is used. 714 715 See bin/matlab/PetscBinaryRead.m for a Matlab code that can read in the binary file when the binary 716 viewer is used. 717 718 Concepts: matrices^viewing 719 Concepts: matrices^plotting 720 Concepts: matrices^printing 721 722 .seealso: PetscViewerSetFormat(), PetscViewerASCIIOpen(), PetscViewerDrawOpen(), 723 PetscViewerSocketOpen(), PetscViewerBinaryOpen(), MatLoad() 724 @*/ 725 PetscErrorCode MatView(Mat mat,PetscViewer viewer) 726 { 727 PetscErrorCode ierr; 728 PetscInt rows,cols,bs; 729 PetscBool iascii; 730 PetscViewerFormat format; 731 732 PetscFunctionBegin; 733 PetscValidHeaderSpecific(mat,MAT_CLASSID,1); 734 PetscValidType(mat,1); 735 if (!viewer) { 736 ierr = PetscViewerASCIIGetStdout(((PetscObject)mat)->comm,&viewer);CHKERRQ(ierr); 737 } 738 PetscValidHeaderSpecific(viewer,PETSC_VIEWER_CLASSID,2); 739 PetscCheckSameComm(mat,1,viewer,2); 740 if (!mat->assembled) SETERRQ(((PetscObject)mat)->comm,PETSC_ERR_ORDER,"Must call MatAssemblyBegin/End() before viewing matrix"); 741 MatCheckPreallocated(mat,1); 742 743 ierr = PetscLogEventBegin(MAT_View,mat,viewer,0,0);CHKERRQ(ierr); 744 ierr = PetscObjectTypeCompare((PetscObject)viewer,PETSCVIEWERASCII,&iascii);CHKERRQ(ierr); 745 if (iascii) { 746 ierr = PetscViewerGetFormat(viewer,&format);CHKERRQ(ierr); 747 if (format == PETSC_VIEWER_ASCII_INFO || format == PETSC_VIEWER_ASCII_INFO_DETAIL) { 748 ierr = PetscObjectPrintClassNamePrefixType((PetscObject)mat,viewer,"Matrix Object");CHKERRQ(ierr); 749 ierr = PetscViewerASCIIPushTab(viewer);CHKERRQ(ierr); 750 ierr = MatGetSize(mat,&rows,&cols);CHKERRQ(ierr); 751 ierr = MatGetBlockSize(mat,&bs);CHKERRQ(ierr); 752 if (bs != 1) { 753 ierr = PetscViewerASCIIPrintf(viewer,"rows=%D, cols=%D, bs=%D\n",rows,cols,bs);CHKERRQ(ierr); 754 } else { 755 ierr = PetscViewerASCIIPrintf(viewer,"rows=%D, cols=%D\n",rows,cols);CHKERRQ(ierr); 756 } 757 if (mat->factortype) { 758 const MatSolverPackage solver; 759 ierr = MatFactorGetSolverPackage(mat,&solver);CHKERRQ(ierr); 760 ierr = PetscViewerASCIIPrintf(viewer,"package used to perform factorization: %s\n",solver);CHKERRQ(ierr); 761 } 762 if (mat->ops->getinfo) { 763 MatInfo info; 764 ierr = MatGetInfo(mat,MAT_GLOBAL_SUM,&info);CHKERRQ(ierr); 765 ierr = PetscViewerASCIIPrintf(viewer,"total: nonzeros=%D, allocated nonzeros=%D\n",(PetscInt)info.nz_used,(PetscInt)info.nz_allocated);CHKERRQ(ierr); 766 ierr = PetscViewerASCIIPrintf(viewer,"total number of mallocs used during MatSetValues calls =%D\n",(PetscInt)info.mallocs);CHKERRQ(ierr); 767 } 768 } 769 } 770 if (mat->ops->view) { 771 ierr = PetscViewerASCIIPushTab(viewer);CHKERRQ(ierr); 772 ierr = (*mat->ops->view)(mat,viewer);CHKERRQ(ierr); 773 ierr = PetscViewerASCIIPopTab(viewer);CHKERRQ(ierr); 774 } else if (!iascii) { 775 SETERRQ1(PETSC_COMM_SELF,PETSC_ERR_SUP,"Viewer type %s not supported",((PetscObject)viewer)->type_name); 776 } 777 if (iascii) { 778 ierr = PetscViewerGetFormat(viewer,&format);CHKERRQ(ierr); 779 if (format == PETSC_VIEWER_ASCII_INFO || format == PETSC_VIEWER_ASCII_INFO_DETAIL) { 780 ierr = PetscViewerASCIIPopTab(viewer);CHKERRQ(ierr); 781 } 782 } 783 ierr = PetscLogEventEnd(MAT_View,mat,viewer,0,0);CHKERRQ(ierr); 784 PetscFunctionReturn(0); 785 } 786 787 #if defined(PETSC_USE_DEBUG) 788 #include <../src/sys/totalview/tv_data_display.h> 789 PETSC_UNUSED static int TV_display_type(const struct _p_Mat *mat) 790 { 791 TV_add_row("Local rows", "int", &mat->rmap->n); 792 TV_add_row("Local columns", "int", &mat->cmap->n); 793 TV_add_row("Global rows", "int", &mat->rmap->N); 794 TV_add_row("Global columns", "int", &mat->cmap->N); 795 TV_add_row("Typename", TV_ascii_string_type, ((PetscObject)mat)->type_name); 796 return TV_format_OK; 797 } 798 #endif 799 800 #undef __FUNCT__ 801 #define __FUNCT__ "MatLoad" 802 /*@C 803 MatLoad - Loads a matrix that has been stored in binary format 804 with MatView(). The matrix format is determined from the options database. 805 Generates a parallel MPI matrix if the communicator has more than one 806 processor. The default matrix type is AIJ. 807 808 Collective on PetscViewer 809 810 Input Parameters: 811 + newmat - the newly loaded matrix, this needs to have been created with MatCreate() 812 or some related function before a call to MatLoad() 813 - viewer - binary file viewer, created with PetscViewerBinaryOpen() 814 815 Options Database Keys: 816 Used with block matrix formats (MATSEQBAIJ, ...) to specify 817 block size 818 . -matload_block_size <bs> 819 820 Level: beginner 821 822 Notes: 823 If the Mat type has not yet been given then MATAIJ is used, call MatSetFromOptions() on the 824 Mat before calling this routine if you wish to set it from the options database. 825 826 MatLoad() automatically loads into the options database any options 827 given in the file filename.info where filename is the name of the file 828 that was passed to the PetscViewerBinaryOpen(). The options in the info 829 file will be ignored if you use the -viewer_binary_skip_info option. 830 831 If the type or size of newmat is not set before a call to MatLoad, PETSc 832 sets the default matrix type AIJ and sets the local and global sizes. 833 If type and/or size is already set, then the same are used. 834 835 In parallel, each processor can load a subset of rows (or the 836 entire matrix). This routine is especially useful when a large 837 matrix is stored on disk and only part of it is desired on each 838 processor. For example, a parallel solver may access only some of 839 the rows from each processor. The algorithm used here reads 840 relatively small blocks of data rather than reading the entire 841 matrix and then subsetting it. 842 843 Notes for advanced users: 844 Most users should not need to know the details of the binary storage 845 format, since MatLoad() and MatView() completely hide these details. 846 But for anyone who's interested, the standard binary matrix storage 847 format is 848 849 $ int MAT_FILE_CLASSID 850 $ int number of rows 851 $ int number of columns 852 $ int total number of nonzeros 853 $ int *number nonzeros in each row 854 $ int *column indices of all nonzeros (starting index is zero) 855 $ PetscScalar *values of all nonzeros 856 857 PETSc automatically does the byte swapping for 858 machines that store the bytes reversed, e.g. DEC alpha, freebsd, 859 linux, Windows and the paragon; thus if you write your own binary 860 read/write routines you have to swap the bytes; see PetscBinaryRead() 861 and PetscBinaryWrite() to see how this may be done. 862 863 .keywords: matrix, load, binary, input 864 865 .seealso: PetscViewerBinaryOpen(), MatView(), VecLoad() 866 867 @*/ 868 PetscErrorCode MatLoad(Mat newmat,PetscViewer viewer) 869 { 870 PetscErrorCode ierr; 871 PetscBool isbinary,flg; 872 873 PetscFunctionBegin; 874 PetscValidHeaderSpecific(newmat,MAT_CLASSID,1); 875 PetscValidHeaderSpecific(viewer,PETSC_VIEWER_CLASSID,2); 876 ierr = PetscObjectTypeCompare((PetscObject)viewer,PETSCVIEWERBINARY,&isbinary);CHKERRQ(ierr); 877 if (!isbinary) SETERRQ(PETSC_COMM_SELF,PETSC_ERR_ARG_WRONG,"Invalid viewer; open viewer with PetscViewerBinaryOpen()"); 878 879 if (!((PetscObject)newmat)->type_name) { 880 ierr = MatSetType(newmat,MATAIJ);CHKERRQ(ierr); 881 } 882 883 if (!newmat->ops->load) SETERRQ(PETSC_COMM_SELF,PETSC_ERR_SUP,"MatLoad is not supported for type"); 884 ierr = PetscLogEventBegin(MAT_Load,viewer,0,0,0);CHKERRQ(ierr); 885 ierr = (*newmat->ops->load)(newmat,viewer);CHKERRQ(ierr); 886 ierr = PetscLogEventEnd(MAT_Load,viewer,0,0,0);CHKERRQ(ierr); 887 888 flg = PETSC_FALSE; 889 ierr = PetscOptionsGetBool(((PetscObject)newmat)->prefix,"-matload_symmetric",&flg,PETSC_NULL);CHKERRQ(ierr); 890 if (flg) { 891 ierr = MatSetOption(newmat,MAT_SYMMETRIC,PETSC_TRUE);CHKERRQ(ierr); 892 ierr = MatSetOption(newmat,MAT_SYMMETRY_ETERNAL,PETSC_TRUE);CHKERRQ(ierr); 893 } 894 flg = PETSC_FALSE; 895 ierr = PetscOptionsGetBool(((PetscObject)newmat)->prefix,"-matload_spd",&flg,PETSC_NULL);CHKERRQ(ierr); 896 if (flg) { 897 ierr = MatSetOption(newmat,MAT_SPD,PETSC_TRUE);CHKERRQ(ierr); 898 } 899 PetscFunctionReturn(0); 900 } 901 902 #undef __FUNCT__ 903 #define __FUNCT__ "MatDestroy" 904 /*@ 905 MatDestroy - Frees space taken by a matrix. 906 907 Collective on Mat 908 909 Input Parameter: 910 . A - the matrix 911 912 Level: beginner 913 914 @*/ 915 PetscErrorCode MatDestroy(Mat *A) 916 { 917 PetscErrorCode ierr; 918 919 PetscFunctionBegin; 920 if (!*A) PetscFunctionReturn(0); 921 PetscValidHeaderSpecific(*A,MAT_CLASSID,1); 922 if (--((PetscObject)(*A))->refct > 0) {*A = PETSC_NULL; PetscFunctionReturn(0);} 923 /* if memory was published with AMS then destroy it */ 924 ierr = PetscObjectDepublish(*A);CHKERRQ(ierr); 925 if ((*A)->ops->destroy) { 926 ierr = (*(*A)->ops->destroy)(*A);CHKERRQ(ierr); 927 } 928 ierr = MatNullSpaceDestroy(&(*A)->nullsp);CHKERRQ(ierr); 929 ierr = MatNullSpaceDestroy(&(*A)->nearnullsp);CHKERRQ(ierr); 930 ierr = PetscLayoutDestroy(&(*A)->rmap);CHKERRQ(ierr); 931 ierr = PetscLayoutDestroy(&(*A)->cmap);CHKERRQ(ierr); 932 ierr = PetscHeaderDestroy(A);CHKERRQ(ierr); 933 PetscFunctionReturn(0); 934 } 935 936 #undef __FUNCT__ 937 #define __FUNCT__ "MatSetValues" 938 /*@ 939 MatSetValues - Inserts or adds a block of values into a matrix. 940 These values may be cached, so MatAssemblyBegin() and MatAssemblyEnd() 941 MUST be called after all calls to MatSetValues() have been completed. 942 943 Not Collective 944 945 Input Parameters: 946 + mat - the matrix 947 . v - a logically two-dimensional array of values 948 . m, idxm - the number of rows and their global indices 949 . n, idxn - the number of columns and their global indices 950 - addv - either ADD_VALUES or INSERT_VALUES, where 951 ADD_VALUES adds values to any existing entries, and 952 INSERT_VALUES replaces existing entries with new values 953 954 Notes: 955 If you create the matrix yourself (that is not with a call to DMCreateMatrix()) then you MUST call MatXXXXSetPreallocation() or 956 MatSetUp() before using this routine 957 958 By default the values, v, are row-oriented. See MatSetOption() for other options. 959 960 Calls to MatSetValues() with the INSERT_VALUES and ADD_VALUES 961 options cannot be mixed without intervening calls to the assembly 962 routines. 963 964 MatSetValues() uses 0-based row and column numbers in Fortran 965 as well as in C. 966 967 Negative indices may be passed in idxm and idxn, these rows and columns are 968 simply ignored. This allows easily inserting element stiffness matrices 969 with homogeneous Dirchlet boundary conditions that you don't want represented 970 in the matrix. 971 972 Efficiency Alert: 973 The routine MatSetValuesBlocked() may offer much better efficiency 974 for users of block sparse formats (MATSEQBAIJ and MATMPIBAIJ). 975 976 Level: beginner 977 978 Concepts: matrices^putting entries in 979 980 .seealso: MatSetOption(), MatAssemblyBegin(), MatAssemblyEnd(), MatSetValuesBlocked(), MatSetValuesLocal(), 981 InsertMode, INSERT_VALUES, ADD_VALUES 982 @*/ 983 PetscErrorCode MatSetValues(Mat mat,PetscInt m,const PetscInt idxm[],PetscInt n,const PetscInt idxn[],const PetscScalar v[],InsertMode addv) 984 { 985 PetscErrorCode ierr; 986 #if defined(PETSC_USE_DEBUG) 987 PetscInt i,j; 988 #endif 989 990 PetscFunctionBegin; 991 PetscValidHeaderSpecific(mat,MAT_CLASSID,1); 992 PetscValidType(mat,1); 993 if (!m || !n) PetscFunctionReturn(0); /* no values to insert */ 994 PetscValidIntPointer(idxm,3); 995 PetscValidIntPointer(idxn,5); 996 if (v) PetscValidScalarPointer(v,6); 997 MatCheckPreallocated(mat,1); 998 if (mat->insertmode == NOT_SET_VALUES) { 999 mat->insertmode = addv; 1000 } 1001 #if defined(PETSC_USE_DEBUG) 1002 else if (mat->insertmode != addv) SETERRQ(PETSC_COMM_SELF,PETSC_ERR_ARG_WRONGSTATE,"Cannot mix add values and insert values"); 1003 if (mat->factortype) SETERRQ(PETSC_COMM_SELF,PETSC_ERR_ARG_WRONGSTATE,"Not for factored matrix"); 1004 if (!mat->ops->setvalues) SETERRQ1(PETSC_COMM_SELF,PETSC_ERR_SUP,"Mat type %s",((PetscObject)mat)->type_name); 1005 1006 if (v) { 1007 for (i=0; i<m; i++) { 1008 for (j=0; j<n; j++) { 1009 if (PetscIsInfOrNanScalar(v[i*n+j])) 1010 #if defined(PETSC_USE_COMPLEX) 1011 SETERRQ4(PETSC_COMM_SELF,PETSC_ERR_FP,"Inserting %G+iG at matrix entry (%D,%D)",PetscRealPart(v[i*n+j]),PetscImaginaryPart(v[i*n+j]),idxm[i],idxn[j]); 1012 #else 1013 SETERRQ3(PETSC_COMM_SELF,PETSC_ERR_FP,"Inserting %G at matrix entry (%D,%D)",(PetscReal)v[i*n+j],idxm[i],idxn[j]); 1014 #endif 1015 } 1016 } 1017 } 1018 #endif 1019 1020 if (mat->assembled) { 1021 mat->was_assembled = PETSC_TRUE; 1022 mat->assembled = PETSC_FALSE; 1023 } 1024 ierr = PetscLogEventBegin(MAT_SetValues,mat,0,0,0);CHKERRQ(ierr); 1025 ierr = (*mat->ops->setvalues)(mat,m,idxm,n,idxn,v,addv);CHKERRQ(ierr); 1026 ierr = PetscLogEventEnd(MAT_SetValues,mat,0,0,0);CHKERRQ(ierr); 1027 #if defined(PETSC_HAVE_CUSP) 1028 if (mat->valid_GPU_matrix != PETSC_CUSP_UNALLOCATED) { 1029 mat->valid_GPU_matrix = PETSC_CUSP_CPU; 1030 } 1031 #endif 1032 PetscFunctionReturn(0); 1033 } 1034 1035 1036 #undef __FUNCT__ 1037 #define __FUNCT__ "MatSetValuesRowLocal" 1038 /*@ 1039 MatSetValuesRowLocal - Inserts a row (block row for BAIJ matrices) of nonzero 1040 values into a matrix 1041 1042 Not Collective 1043 1044 Input Parameters: 1045 + mat - the matrix 1046 . row - the (block) row to set 1047 - v - a logically two-dimensional array of values 1048 1049 Notes: 1050 By the values, v, are column-oriented (for the block version) and sorted 1051 1052 All the nonzeros in the row must be provided 1053 1054 The matrix must have previously had its column indices set 1055 1056 The row must belong to this process 1057 1058 Level: intermediate 1059 1060 Concepts: matrices^putting entries in 1061 1062 .seealso: MatSetOption(), MatAssemblyBegin(), MatAssemblyEnd(), MatSetValuesBlocked(), MatSetValuesLocal(), 1063 InsertMode, INSERT_VALUES, ADD_VALUES, MatSetValues(), MatSetValuesRow(), MatSetLocalToGlobalMapping() 1064 @*/ 1065 PetscErrorCode MatSetValuesRowLocal(Mat mat,PetscInt row,const PetscScalar v[]) 1066 { 1067 PetscErrorCode ierr; 1068 1069 PetscFunctionBegin; 1070 PetscValidHeaderSpecific(mat,MAT_CLASSID,1); 1071 PetscValidType(mat,1); 1072 PetscValidScalarPointer(v,2); 1073 ierr = MatSetValuesRow(mat, mat->rmap->mapping->indices[row],v);CHKERRQ(ierr); 1074 #if defined(PETSC_HAVE_CUSP) 1075 if (mat->valid_GPU_matrix != PETSC_CUSP_UNALLOCATED) { 1076 mat->valid_GPU_matrix = PETSC_CUSP_CPU; 1077 } 1078 #endif 1079 PetscFunctionReturn(0); 1080 } 1081 1082 #undef __FUNCT__ 1083 #define __FUNCT__ "MatSetValuesRow" 1084 /*@ 1085 MatSetValuesRow - Inserts a row (block row for BAIJ matrices) of nonzero 1086 values into a matrix 1087 1088 Not Collective 1089 1090 Input Parameters: 1091 + mat - the matrix 1092 . row - the (block) row to set 1093 - v - a logically two-dimensional array of values 1094 1095 Notes: 1096 The values, v, are column-oriented for the block version. 1097 1098 All the nonzeros in the row must be provided 1099 1100 THE MATRIX MUSAT HAVE PREVIOUSLY HAD ITS COLUMN INDICES SET. IT IS RARE THAT THIS ROUTINE IS USED, usually MatSetValues() is used. 1101 1102 The row must belong to this process 1103 1104 Level: advanced 1105 1106 Concepts: matrices^putting entries in 1107 1108 .seealso: MatSetOption(), MatAssemblyBegin(), MatAssemblyEnd(), MatSetValuesBlocked(), MatSetValuesLocal(), 1109 InsertMode, INSERT_VALUES, ADD_VALUES, MatSetValues() 1110 @*/ 1111 PetscErrorCode MatSetValuesRow(Mat mat,PetscInt row,const PetscScalar v[]) 1112 { 1113 PetscErrorCode ierr; 1114 1115 PetscFunctionBegin; 1116 PetscValidHeaderSpecific(mat,MAT_CLASSID,1); 1117 PetscValidType(mat,1); 1118 MatCheckPreallocated(mat,1); 1119 PetscValidScalarPointer(v,2); 1120 #if defined(PETSC_USE_DEBUG) 1121 if (mat->insertmode == ADD_VALUES) SETERRQ(PETSC_COMM_SELF,PETSC_ERR_ARG_WRONGSTATE,"Cannot mix add and insert values"); 1122 if (mat->factortype) SETERRQ(PETSC_COMM_SELF,PETSC_ERR_ARG_WRONGSTATE,"Not for factored matrix"); 1123 #endif 1124 mat->insertmode = INSERT_VALUES; 1125 1126 if (mat->assembled) { 1127 mat->was_assembled = PETSC_TRUE; 1128 mat->assembled = PETSC_FALSE; 1129 } 1130 ierr = PetscLogEventBegin(MAT_SetValues,mat,0,0,0);CHKERRQ(ierr); 1131 if (!mat->ops->setvaluesrow) SETERRQ1(PETSC_COMM_SELF,PETSC_ERR_SUP,"Mat type %s",((PetscObject)mat)->type_name); 1132 ierr = (*mat->ops->setvaluesrow)(mat,row,v);CHKERRQ(ierr); 1133 ierr = PetscLogEventEnd(MAT_SetValues,mat,0,0,0);CHKERRQ(ierr); 1134 #if defined(PETSC_HAVE_CUSP) 1135 if (mat->valid_GPU_matrix != PETSC_CUSP_UNALLOCATED) { 1136 mat->valid_GPU_matrix = PETSC_CUSP_CPU; 1137 } 1138 #endif 1139 PetscFunctionReturn(0); 1140 } 1141 1142 #undef __FUNCT__ 1143 #define __FUNCT__ "MatSetValuesStencil" 1144 /*@ 1145 MatSetValuesStencil - Inserts or adds a block of values into a matrix. 1146 Using structured grid indexing 1147 1148 Not Collective 1149 1150 Input Parameters: 1151 + mat - the matrix 1152 . m - number of rows being entered 1153 . idxm - grid coordinates (and component number when dof > 1) for matrix rows being entered 1154 . n - number of columns being entered 1155 . idxn - grid coordinates (and component number when dof > 1) for matrix columns being entered 1156 . v - a logically two-dimensional array of values 1157 - addv - either ADD_VALUES or INSERT_VALUES, where 1158 ADD_VALUES adds values to any existing entries, and 1159 INSERT_VALUES replaces existing entries with new values 1160 1161 Notes: 1162 By default the values, v, are row-oriented. See MatSetOption() for other options. 1163 1164 Calls to MatSetValuesStencil() with the INSERT_VALUES and ADD_VALUES 1165 options cannot be mixed without intervening calls to the assembly 1166 routines. 1167 1168 The grid coordinates are across the entire grid, not just the local portion 1169 1170 MatSetValuesStencil() uses 0-based row and column numbers in Fortran 1171 as well as in C. 1172 1173 For setting/accessing vector values via array coordinates you can use the DMDAVecGetArray() routine 1174 1175 In order to use this routine you must either obtain the matrix with DMCreateMatrix() 1176 or call MatSetLocalToGlobalMapping() and MatSetStencil() first. 1177 1178 The columns and rows in the stencil passed in MUST be contained within the 1179 ghost region of the given process as set with DMDACreateXXX() or MatSetStencil(). For example, 1180 if you create a DMDA with an overlap of one grid level and on a particular process its first 1181 local nonghost x logical coordinate is 6 (so its first ghost x logical coordinate is 5) the 1182 first i index you can use in your column and row indices in MatSetStencil() is 5. 1183 1184 In Fortran idxm and idxn should be declared as 1185 $ MatStencil idxm(4,m),idxn(4,n) 1186 and the values inserted using 1187 $ idxm(MatStencil_i,1) = i 1188 $ idxm(MatStencil_j,1) = j 1189 $ idxm(MatStencil_k,1) = k 1190 $ idxm(MatStencil_c,1) = c 1191 etc 1192 1193 For periodic boundary conditions use negative indices for values to the left (below 0; that are to be 1194 obtained by wrapping values from right edge). For values to the right of the last entry using that index plus one 1195 etc to obtain values that obtained by wrapping the values from the left edge. This does not work for anything but the 1196 DMDA_BOUNDARY_PERIODIC boundary type. 1197 1198 For indices that don't mean anything for your case (like the k index when working in 2d) or the c index when you have 1199 a single value per point) you can skip filling those indices. 1200 1201 Inspired by the structured grid interface to the HYPRE package 1202 (http://www.llnl.gov/CASC/hypre) 1203 1204 Efficiency Alert: 1205 The routine MatSetValuesBlockedStencil() may offer much better efficiency 1206 for users of block sparse formats (MATSEQBAIJ and MATMPIBAIJ). 1207 1208 Level: beginner 1209 1210 Concepts: matrices^putting entries in 1211 1212 .seealso: MatSetOption(), MatAssemblyBegin(), MatAssemblyEnd(), MatSetValuesBlocked(), MatSetValuesLocal() 1213 MatSetValues(), MatSetValuesBlockedStencil(), MatSetStencil(), DMCreateMatrix(), DMDAVecGetArray(), MatStencil 1214 @*/ 1215 PetscErrorCode MatSetValuesStencil(Mat mat,PetscInt m,const MatStencil idxm[],PetscInt n,const MatStencil idxn[],const PetscScalar v[],InsertMode addv) 1216 { 1217 PetscErrorCode ierr; 1218 PetscInt buf[8192],*bufm=0,*bufn=0,*jdxm,*jdxn; 1219 PetscInt j,i,dim = mat->stencil.dim,*dims = mat->stencil.dims+1,tmp; 1220 PetscInt *starts = mat->stencil.starts,*dxm = (PetscInt*)idxm,*dxn = (PetscInt*)idxn,sdim = dim - (1 - (PetscInt)mat->stencil.noc); 1221 1222 PetscFunctionBegin; 1223 if (!m || !n) PetscFunctionReturn(0); /* no values to insert */ 1224 PetscValidHeaderSpecific(mat,MAT_CLASSID,1); 1225 PetscValidType(mat,1); 1226 PetscValidIntPointer(idxm,3); 1227 PetscValidIntPointer(idxn,5); 1228 PetscValidScalarPointer(v,6); 1229 1230 if ((m+n) <= (PetscInt)(sizeof(buf)/sizeof(PetscInt))) { 1231 jdxm = buf; jdxn = buf+m; 1232 } else { 1233 ierr = PetscMalloc2(m,PetscInt,&bufm,n,PetscInt,&bufn);CHKERRQ(ierr); 1234 jdxm = bufm; jdxn = bufn; 1235 } 1236 for (i=0; i<m; i++) { 1237 for (j=0; j<3-sdim; j++) dxm++; 1238 tmp = *dxm++ - starts[0]; 1239 for (j=0; j<dim-1; j++) { 1240 if ((*dxm++ - starts[j+1]) < 0 || tmp < 0) tmp = PETSC_MIN_INT; 1241 else tmp = tmp*dims[j] + *(dxm-1) - starts[j+1]; 1242 } 1243 if (mat->stencil.noc) dxm++; 1244 jdxm[i] = tmp; 1245 } 1246 for (i=0; i<n; i++) { 1247 for (j=0; j<3-sdim; j++) dxn++; 1248 tmp = *dxn++ - starts[0]; 1249 for (j=0; j<dim-1; j++) { 1250 if ((*dxn++ - starts[j+1]) < 0 || tmp < 0) tmp = PETSC_MIN_INT; 1251 else tmp = tmp*dims[j] + *(dxn-1) - starts[j+1]; 1252 } 1253 if (mat->stencil.noc) dxn++; 1254 jdxn[i] = tmp; 1255 } 1256 ierr = MatSetValuesLocal(mat,m,jdxm,n,jdxn,v,addv);CHKERRQ(ierr); 1257 ierr = PetscFree2(bufm,bufn);CHKERRQ(ierr); 1258 PetscFunctionReturn(0); 1259 } 1260 1261 #undef __FUNCT__ 1262 #define __FUNCT__ "MatSetValuesBlockedStencil" 1263 /*@C 1264 MatSetValuesBlockedStencil - Inserts or adds a block of values into a matrix. 1265 Using structured grid indexing 1266 1267 Not Collective 1268 1269 Input Parameters: 1270 + mat - the matrix 1271 . m - number of rows being entered 1272 . idxm - grid coordinates for matrix rows being entered 1273 . n - number of columns being entered 1274 . idxn - grid coordinates for matrix columns being entered 1275 . v - a logically two-dimensional array of values 1276 - addv - either ADD_VALUES or INSERT_VALUES, where 1277 ADD_VALUES adds values to any existing entries, and 1278 INSERT_VALUES replaces existing entries with new values 1279 1280 Notes: 1281 By default the values, v, are row-oriented and unsorted. 1282 See MatSetOption() for other options. 1283 1284 Calls to MatSetValuesBlockedStencil() with the INSERT_VALUES and ADD_VALUES 1285 options cannot be mixed without intervening calls to the assembly 1286 routines. 1287 1288 The grid coordinates are across the entire grid, not just the local portion 1289 1290 MatSetValuesBlockedStencil() uses 0-based row and column numbers in Fortran 1291 as well as in C. 1292 1293 For setting/accessing vector values via array coordinates you can use the DMDAVecGetArray() routine 1294 1295 In order to use this routine you must either obtain the matrix with DMCreateMatrix() 1296 or call MatSetBlockSize(), MatSetLocalToGlobalMapping() and MatSetStencil() first. 1297 1298 The columns and rows in the stencil passed in MUST be contained within the 1299 ghost region of the given process as set with DMDACreateXXX() or MatSetStencil(). For example, 1300 if you create a DMDA with an overlap of one grid level and on a particular process its first 1301 local nonghost x logical coordinate is 6 (so its first ghost x logical coordinate is 5) the 1302 first i index you can use in your column and row indices in MatSetStencil() is 5. 1303 1304 In Fortran idxm and idxn should be declared as 1305 $ MatStencil idxm(4,m),idxn(4,n) 1306 and the values inserted using 1307 $ idxm(MatStencil_i,1) = i 1308 $ idxm(MatStencil_j,1) = j 1309 $ idxm(MatStencil_k,1) = k 1310 etc 1311 1312 Negative indices may be passed in idxm and idxn, these rows and columns are 1313 simply ignored. This allows easily inserting element stiffness matrices 1314 with homogeneous Dirchlet boundary conditions that you don't want represented 1315 in the matrix. 1316 1317 Inspired by the structured grid interface to the HYPRE package 1318 (http://www.llnl.gov/CASC/hypre) 1319 1320 Level: beginner 1321 1322 Concepts: matrices^putting entries in 1323 1324 .seealso: MatSetOption(), MatAssemblyBegin(), MatAssemblyEnd(), MatSetValuesBlocked(), MatSetValuesLocal() 1325 MatSetValues(), MatSetValuesStencil(), MatSetStencil(), DMCreateMatrix(), DMDAVecGetArray(), MatStencil, 1326 MatSetBlockSize(), MatSetLocalToGlobalMapping() 1327 @*/ 1328 PetscErrorCode MatSetValuesBlockedStencil(Mat mat,PetscInt m,const MatStencil idxm[],PetscInt n,const MatStencil idxn[],const PetscScalar v[],InsertMode addv) 1329 { 1330 PetscErrorCode ierr; 1331 PetscInt buf[8192],*bufm=0,*bufn=0,*jdxm,*jdxn; 1332 PetscInt j,i,dim = mat->stencil.dim,*dims = mat->stencil.dims+1,tmp; 1333 PetscInt *starts = mat->stencil.starts,*dxm = (PetscInt*)idxm,*dxn = (PetscInt*)idxn,sdim = dim - (1 - (PetscInt)mat->stencil.noc); 1334 1335 PetscFunctionBegin; 1336 if (!m || !n) PetscFunctionReturn(0); /* no values to insert */ 1337 PetscValidHeaderSpecific(mat,MAT_CLASSID,1); 1338 PetscValidType(mat,1); 1339 PetscValidIntPointer(idxm,3); 1340 PetscValidIntPointer(idxn,5); 1341 PetscValidScalarPointer(v,6); 1342 1343 if ((m+n) <= (PetscInt)(sizeof(buf)/sizeof(PetscInt))) { 1344 jdxm = buf; jdxn = buf+m; 1345 } else { 1346 ierr = PetscMalloc2(m,PetscInt,&bufm,n,PetscInt,&bufn);CHKERRQ(ierr); 1347 jdxm = bufm; jdxn = bufn; 1348 } 1349 for (i=0; i<m; i++) { 1350 for (j=0; j<3-sdim; j++) dxm++; 1351 tmp = *dxm++ - starts[0]; 1352 for (j=0; j<sdim-1; j++) { 1353 if ((*dxm++ - starts[j+1]) < 0 || tmp < 0) tmp = PETSC_MIN_INT; 1354 else tmp = tmp*dims[j] + *(dxm-1) - starts[j+1]; 1355 } 1356 dxm++; 1357 jdxm[i] = tmp; 1358 } 1359 for (i=0; i<n; i++) { 1360 for (j=0; j<3-sdim; j++) dxn++; 1361 tmp = *dxn++ - starts[0]; 1362 for (j=0; j<sdim-1; j++) { 1363 if ((*dxn++ - starts[j+1]) < 0 || tmp < 0) tmp = PETSC_MIN_INT; 1364 else tmp = tmp*dims[j] + *(dxn-1) - starts[j+1]; 1365 } 1366 dxn++; 1367 jdxn[i] = tmp; 1368 } 1369 ierr = MatSetValuesBlockedLocal(mat,m,jdxm,n,jdxn,v,addv);CHKERRQ(ierr); 1370 ierr = PetscFree2(bufm,bufn);CHKERRQ(ierr); 1371 #if defined(PETSC_HAVE_CUSP) 1372 if (mat->valid_GPU_matrix != PETSC_CUSP_UNALLOCATED) { 1373 mat->valid_GPU_matrix = PETSC_CUSP_CPU; 1374 } 1375 #endif 1376 PetscFunctionReturn(0); 1377 } 1378 1379 #undef __FUNCT__ 1380 #define __FUNCT__ "MatSetStencil" 1381 /*@ 1382 MatSetStencil - Sets the grid information for setting values into a matrix via 1383 MatSetValuesStencil() 1384 1385 Not Collective 1386 1387 Input Parameters: 1388 + mat - the matrix 1389 . dim - dimension of the grid 1, 2, or 3 1390 . dims - number of grid points in x, y, and z direction, including ghost points on your processor 1391 . starts - starting point of ghost nodes on your processor in x, y, and z direction 1392 - dof - number of degrees of freedom per node 1393 1394 1395 Inspired by the structured grid interface to the HYPRE package 1396 (www.llnl.gov/CASC/hyper) 1397 1398 For matrices generated with DMCreateMatrix() this routine is automatically called and so not needed by the 1399 user. 1400 1401 Level: beginner 1402 1403 Concepts: matrices^putting entries in 1404 1405 .seealso: MatSetOption(), MatAssemblyBegin(), MatAssemblyEnd(), MatSetValuesBlocked(), MatSetValuesLocal() 1406 MatSetValues(), MatSetValuesBlockedStencil(), MatSetValuesStencil() 1407 @*/ 1408 PetscErrorCode MatSetStencil(Mat mat,PetscInt dim,const PetscInt dims[],const PetscInt starts[],PetscInt dof) 1409 { 1410 PetscInt i; 1411 1412 PetscFunctionBegin; 1413 PetscValidHeaderSpecific(mat,MAT_CLASSID,1); 1414 PetscValidIntPointer(dims,3); 1415 PetscValidIntPointer(starts,4); 1416 1417 mat->stencil.dim = dim + (dof > 1); 1418 for (i=0; i<dim; i++) { 1419 mat->stencil.dims[i] = dims[dim-i-1]; /* copy the values in backwards */ 1420 mat->stencil.starts[i] = starts[dim-i-1]; 1421 } 1422 mat->stencil.dims[dim] = dof; 1423 mat->stencil.starts[dim] = 0; 1424 mat->stencil.noc = (PetscBool)(dof == 1); 1425 PetscFunctionReturn(0); 1426 } 1427 1428 #undef __FUNCT__ 1429 #define __FUNCT__ "MatSetValuesBlocked" 1430 /*@ 1431 MatSetValuesBlocked - Inserts or adds a block of values into a matrix. 1432 1433 Not Collective 1434 1435 Input Parameters: 1436 + mat - the matrix 1437 . v - a logically two-dimensional array of values 1438 . m, idxm - the number of block rows and their global block indices 1439 . n, idxn - the number of block columns and their global block indices 1440 - addv - either ADD_VALUES or INSERT_VALUES, where 1441 ADD_VALUES adds values to any existing entries, and 1442 INSERT_VALUES replaces existing entries with new values 1443 1444 Notes: 1445 If you create the matrix yourself (that is not with a call to DMCreateMatrix()) then you MUST call 1446 MatXXXXSetPreallocation() or MatSetUp() before using this routine. 1447 1448 The m and n count the NUMBER of blocks in the row direction and column direction, 1449 NOT the total number of rows/columns; for example, if the block size is 2 and 1450 you are passing in values for rows 2,3,4,5 then m would be 2 (not 4). 1451 The values in idxm would be 1 2; that is the first index for each block divided by 1452 the block size. 1453 1454 Note that you must call MatSetBlockSize() when constructing this matrix (after 1455 preallocating it). 1456 1457 By default the values, v, are row-oriented, so the layout of 1458 v is the same as for MatSetValues(). See MatSetOption() for other options. 1459 1460 Calls to MatSetValuesBlocked() with the INSERT_VALUES and ADD_VALUES 1461 options cannot be mixed without intervening calls to the assembly 1462 routines. 1463 1464 MatSetValuesBlocked() uses 0-based row and column numbers in Fortran 1465 as well as in C. 1466 1467 Negative indices may be passed in idxm and idxn, these rows and columns are 1468 simply ignored. This allows easily inserting element stiffness matrices 1469 with homogeneous Dirchlet boundary conditions that you don't want represented 1470 in the matrix. 1471 1472 Each time an entry is set within a sparse matrix via MatSetValues(), 1473 internal searching must be done to determine where to place the the 1474 data in the matrix storage space. By instead inserting blocks of 1475 entries via MatSetValuesBlocked(), the overhead of matrix assembly is 1476 reduced. 1477 1478 Example: 1479 $ Suppose m=n=2 and block size(bs) = 2 The array is 1480 $ 1481 $ 1 2 | 3 4 1482 $ 5 6 | 7 8 1483 $ - - - | - - - 1484 $ 9 10 | 11 12 1485 $ 13 14 | 15 16 1486 $ 1487 $ v[] should be passed in like 1488 $ v[] = [1,2,3,4,5,6,7,8,9,10,11,12,13,14,15,16] 1489 $ 1490 $ If you are not using row oriented storage of v (that is you called MatSetOption(mat,MAT_ROW_ORIENTED,PETSC_FALSE)) then 1491 $ v[] = [1,5,9,13,2,6,10,14,3,7,11,15,4,8,12,16] 1492 1493 Level: intermediate 1494 1495 Concepts: matrices^putting entries in blocked 1496 1497 .seealso: MatSetBlockSize(), MatSetOption(), MatAssemblyBegin(), MatAssemblyEnd(), MatSetValues(), MatSetValuesBlockedLocal() 1498 @*/ 1499 PetscErrorCode MatSetValuesBlocked(Mat mat,PetscInt m,const PetscInt idxm[],PetscInt n,const PetscInt idxn[],const PetscScalar v[],InsertMode addv) 1500 { 1501 PetscErrorCode ierr; 1502 1503 PetscFunctionBegin; 1504 PetscValidHeaderSpecific(mat,MAT_CLASSID,1); 1505 PetscValidType(mat,1); 1506 if (!m || !n) PetscFunctionReturn(0); /* no values to insert */ 1507 PetscValidIntPointer(idxm,3); 1508 PetscValidIntPointer(idxn,5); 1509 PetscValidScalarPointer(v,6); 1510 MatCheckPreallocated(mat,1); 1511 if (mat->insertmode == NOT_SET_VALUES) { 1512 mat->insertmode = addv; 1513 } 1514 #if defined(PETSC_USE_DEBUG) 1515 else if (mat->insertmode != addv) SETERRQ(PETSC_COMM_SELF,PETSC_ERR_ARG_WRONGSTATE,"Cannot mix add values and insert values"); 1516 if (mat->factortype) SETERRQ(PETSC_COMM_SELF,PETSC_ERR_ARG_WRONGSTATE,"Not for factored matrix"); 1517 if (!mat->ops->setvaluesblocked && !mat->ops->setvalues) SETERRQ1(PETSC_COMM_SELF,PETSC_ERR_SUP,"Mat type %s",((PetscObject)mat)->type_name); 1518 #endif 1519 1520 if (mat->assembled) { 1521 mat->was_assembled = PETSC_TRUE; 1522 mat->assembled = PETSC_FALSE; 1523 } 1524 ierr = PetscLogEventBegin(MAT_SetValues,mat,0,0,0);CHKERRQ(ierr); 1525 if (mat->ops->setvaluesblocked) { 1526 ierr = (*mat->ops->setvaluesblocked)(mat,m,idxm,n,idxn,v,addv);CHKERRQ(ierr); 1527 } else { 1528 PetscInt buf[8192],*bufr=0,*bufc=0,*iidxm,*iidxn; 1529 PetscInt i,j,bs=mat->rmap->bs; 1530 if ((m+n)*bs <= (PetscInt)(sizeof(buf)/sizeof(PetscInt))) { 1531 iidxm = buf; iidxn = buf + m*bs; 1532 } else { 1533 ierr = PetscMalloc2(m*bs,PetscInt,&bufr,n*bs,PetscInt,&bufc);CHKERRQ(ierr); 1534 iidxm = bufr; iidxn = bufc; 1535 } 1536 for (i=0; i<m; i++) 1537 for (j=0; j<bs; j++) 1538 iidxm[i*bs+j] = bs*idxm[i] + j; 1539 for (i=0; i<n; i++) 1540 for (j=0; j<bs; j++) 1541 iidxn[i*bs+j] = bs*idxn[i] + j; 1542 ierr = MatSetValues(mat,m*bs,iidxm,n*bs,iidxn,v,addv);CHKERRQ(ierr); 1543 ierr = PetscFree2(bufr,bufc);CHKERRQ(ierr); 1544 } 1545 ierr = PetscLogEventEnd(MAT_SetValues,mat,0,0,0);CHKERRQ(ierr); 1546 #if defined(PETSC_HAVE_CUSP) 1547 if (mat->valid_GPU_matrix != PETSC_CUSP_UNALLOCATED) { 1548 mat->valid_GPU_matrix = PETSC_CUSP_CPU; 1549 } 1550 #endif 1551 PetscFunctionReturn(0); 1552 } 1553 1554 #undef __FUNCT__ 1555 #define __FUNCT__ "MatGetValues" 1556 /*@ 1557 MatGetValues - Gets a block of values from a matrix. 1558 1559 Not Collective; currently only returns a local block 1560 1561 Input Parameters: 1562 + mat - the matrix 1563 . v - a logically two-dimensional array for storing the values 1564 . m, idxm - the number of rows and their global indices 1565 - n, idxn - the number of columns and their global indices 1566 1567 Notes: 1568 The user must allocate space (m*n PetscScalars) for the values, v. 1569 The values, v, are then returned in a row-oriented format, 1570 analogous to that used by default in MatSetValues(). 1571 1572 MatGetValues() uses 0-based row and column numbers in 1573 Fortran as well as in C. 1574 1575 MatGetValues() requires that the matrix has been assembled 1576 with MatAssemblyBegin()/MatAssemblyEnd(). Thus, calls to 1577 MatSetValues() and MatGetValues() CANNOT be made in succession 1578 without intermediate matrix assembly. 1579 1580 Negative row or column indices will be ignored and those locations in v[] will be 1581 left unchanged. 1582 1583 Level: advanced 1584 1585 Concepts: matrices^accessing values 1586 1587 .seealso: MatGetRow(), MatGetSubMatrices(), MatSetValues() 1588 @*/ 1589 PetscErrorCode MatGetValues(Mat mat,PetscInt m,const PetscInt idxm[],PetscInt n,const PetscInt idxn[],PetscScalar v[]) 1590 { 1591 PetscErrorCode ierr; 1592 1593 PetscFunctionBegin; 1594 PetscValidHeaderSpecific(mat,MAT_CLASSID,1); 1595 PetscValidType(mat,1); 1596 if (!m || !n) PetscFunctionReturn(0); 1597 PetscValidIntPointer(idxm,3); 1598 PetscValidIntPointer(idxn,5); 1599 PetscValidScalarPointer(v,6); 1600 if (!mat->assembled) SETERRQ(PETSC_COMM_SELF,PETSC_ERR_ARG_WRONGSTATE,"Not for unassembled matrix"); 1601 if (mat->factortype) SETERRQ(PETSC_COMM_SELF,PETSC_ERR_ARG_WRONGSTATE,"Not for factored matrix"); 1602 if (!mat->ops->getvalues) SETERRQ1(PETSC_COMM_SELF,PETSC_ERR_SUP,"Mat type %s",((PetscObject)mat)->type_name); 1603 MatCheckPreallocated(mat,1); 1604 1605 ierr = PetscLogEventBegin(MAT_GetValues,mat,0,0,0);CHKERRQ(ierr); 1606 ierr = (*mat->ops->getvalues)(mat,m,idxm,n,idxn,v);CHKERRQ(ierr); 1607 ierr = PetscLogEventEnd(MAT_GetValues,mat,0,0,0);CHKERRQ(ierr); 1608 PetscFunctionReturn(0); 1609 } 1610 1611 #undef __FUNCT__ 1612 #define __FUNCT__ "MatSetValuesBatch" 1613 /*@ 1614 MatSetValuesBatch - Adds (ADD_VALUES) many blocks of values into a matrix at once. The blocks must all be square and 1615 the same size. Currently, this can only be called once and creates the given matrix. 1616 1617 Not Collective 1618 1619 Input Parameters: 1620 + mat - the matrix 1621 . nb - the number of blocks 1622 . bs - the number of rows (and columns) in each block 1623 . rows - a concatenation of the rows for each block 1624 - v - a concatenation of logically two-dimensional arrays of values 1625 1626 Notes: 1627 In the future, we will extend this routine to handle rectangular blocks, and to allow multiple calls for a given matrix. 1628 1629 Level: advanced 1630 1631 Concepts: matrices^putting entries in 1632 1633 .seealso: MatSetOption(), MatAssemblyBegin(), MatAssemblyEnd(), MatSetValuesBlocked(), MatSetValuesLocal(), 1634 InsertMode, INSERT_VALUES, ADD_VALUES, MatSetValues() 1635 @*/ 1636 PetscErrorCode MatSetValuesBatch(Mat mat, PetscInt nb, PetscInt bs, PetscInt rows[], const PetscScalar v[]) 1637 { 1638 PetscErrorCode ierr; 1639 1640 PetscFunctionBegin; 1641 PetscValidHeaderSpecific(mat,MAT_CLASSID,1); 1642 PetscValidType(mat,1); 1643 PetscValidScalarPointer(rows,4); 1644 PetscValidScalarPointer(v,5); 1645 #if defined(PETSC_USE_DEBUG) 1646 if (mat->factortype) SETERRQ(PETSC_COMM_SELF,PETSC_ERR_ARG_WRONGSTATE,"Not for factored matrix"); 1647 #endif 1648 1649 ierr = PetscLogEventBegin(MAT_SetValuesBatch,mat,0,0,0);CHKERRQ(ierr); 1650 if (mat->ops->setvaluesbatch) { 1651 ierr = (*mat->ops->setvaluesbatch)(mat,nb,bs,rows,v);CHKERRQ(ierr); 1652 } else { 1653 PetscInt b; 1654 for(b = 0; b < nb; ++b) { 1655 ierr = MatSetValues(mat, bs, &rows[b*bs], bs, &rows[b*bs], &v[b*bs*bs], ADD_VALUES);CHKERRQ(ierr); 1656 } 1657 } 1658 ierr = PetscLogEventEnd(MAT_SetValuesBatch,mat,0,0,0);CHKERRQ(ierr); 1659 PetscFunctionReturn(0); 1660 } 1661 1662 #undef __FUNCT__ 1663 #define __FUNCT__ "MatSetLocalToGlobalMapping" 1664 /*@ 1665 MatSetLocalToGlobalMapping - Sets a local-to-global numbering for use by 1666 the routine MatSetValuesLocal() to allow users to insert matrix entries 1667 using a local (per-processor) numbering. 1668 1669 Not Collective 1670 1671 Input Parameters: 1672 + x - the matrix 1673 . rmapping - row mapping created with ISLocalToGlobalMappingCreate() 1674 or ISLocalToGlobalMappingCreateIS() 1675 - cmapping - column mapping 1676 1677 Level: intermediate 1678 1679 Concepts: matrices^local to global mapping 1680 Concepts: local to global mapping^for matrices 1681 1682 .seealso: MatAssemblyBegin(), MatAssemblyEnd(), MatSetValues(), MatSetValuesLocal() 1683 @*/ 1684 PetscErrorCode MatSetLocalToGlobalMapping(Mat x,ISLocalToGlobalMapping rmapping,ISLocalToGlobalMapping cmapping) 1685 { 1686 PetscErrorCode ierr; 1687 PetscFunctionBegin; 1688 PetscValidHeaderSpecific(x,MAT_CLASSID,1); 1689 PetscValidType(x,1); 1690 PetscValidHeaderSpecific(rmapping,IS_LTOGM_CLASSID,2); 1691 PetscValidHeaderSpecific(cmapping,IS_LTOGM_CLASSID,3); 1692 1693 if (x->ops->setlocaltoglobalmapping) { 1694 ierr = (*x->ops->setlocaltoglobalmapping)(x,rmapping,cmapping);CHKERRQ(ierr); 1695 } else { 1696 ierr = PetscLayoutSetISLocalToGlobalMapping(x->rmap,rmapping);CHKERRQ(ierr); 1697 ierr = PetscLayoutSetISLocalToGlobalMapping(x->cmap,cmapping);CHKERRQ(ierr); 1698 } 1699 PetscFunctionReturn(0); 1700 } 1701 1702 #undef __FUNCT__ 1703 #define __FUNCT__ "MatSetLocalToGlobalMappingBlock" 1704 /*@ 1705 MatSetLocalToGlobalMappingBlock - Sets a local-to-global numbering for use 1706 by the routine MatSetValuesBlockedLocal() to allow users to insert matrix 1707 entries using a local (per-processor) numbering. 1708 1709 Not Collective 1710 1711 Input Parameters: 1712 + x - the matrix 1713 . rmapping - row mapping created with ISLocalToGlobalMappingCreate() or 1714 ISLocalToGlobalMappingCreateIS() 1715 - cmapping - column mapping 1716 1717 Level: intermediate 1718 1719 Concepts: matrices^local to global mapping blocked 1720 Concepts: local to global mapping^for matrices, blocked 1721 1722 .seealso: MatAssemblyBegin(), MatAssemblyEnd(), MatSetValues(), MatSetValuesBlockedLocal(), 1723 MatSetValuesBlocked(), MatSetValuesLocal() 1724 @*/ 1725 PetscErrorCode MatSetLocalToGlobalMappingBlock(Mat x,ISLocalToGlobalMapping rmapping,ISLocalToGlobalMapping cmapping) 1726 { 1727 PetscErrorCode ierr; 1728 PetscFunctionBegin; 1729 PetscValidHeaderSpecific(x,MAT_CLASSID,1); 1730 PetscValidType(x,1); 1731 PetscValidHeaderSpecific(rmapping,IS_LTOGM_CLASSID,2); 1732 PetscValidHeaderSpecific(cmapping,IS_LTOGM_CLASSID,3); 1733 1734 ierr = PetscLayoutSetISLocalToGlobalMappingBlock(x->rmap,rmapping);CHKERRQ(ierr); 1735 ierr = PetscLayoutSetISLocalToGlobalMappingBlock(x->cmap,cmapping);CHKERRQ(ierr); 1736 PetscFunctionReturn(0); 1737 } 1738 1739 #undef __FUNCT__ 1740 #define __FUNCT__ "MatGetLocalToGlobalMapping" 1741 /*@ 1742 MatGetLocalToGlobalMapping - Gets the local-to-global numbering set by MatSetLocalToGlobalMapping() 1743 1744 Not Collective 1745 1746 Input Parameters: 1747 . A - the matrix 1748 1749 Output Parameters: 1750 + rmapping - row mapping 1751 - cmapping - column mapping 1752 1753 Level: advanced 1754 1755 Concepts: matrices^local to global mapping 1756 Concepts: local to global mapping^for matrices 1757 1758 .seealso: MatSetValuesLocal(), MatGetLocalToGlobalMappingBlock() 1759 @*/ 1760 PetscErrorCode MatGetLocalToGlobalMapping(Mat A,ISLocalToGlobalMapping *rmapping,ISLocalToGlobalMapping *cmapping) 1761 { 1762 PetscFunctionBegin; 1763 PetscValidHeaderSpecific(A,MAT_CLASSID,1); 1764 PetscValidType(A,1); 1765 if (rmapping) PetscValidPointer(rmapping,2); 1766 if (cmapping) PetscValidPointer(cmapping,3); 1767 if (rmapping) *rmapping = A->rmap->mapping; 1768 if (cmapping) *cmapping = A->cmap->mapping; 1769 PetscFunctionReturn(0); 1770 } 1771 1772 #undef __FUNCT__ 1773 #define __FUNCT__ "MatGetLocalToGlobalMappingBlock" 1774 /*@ 1775 MatGetLocalToGlobalMappingBlock - Gets the local-to-global numbering set by MatSetLocalToGlobalMappingBlock() 1776 1777 Not Collective 1778 1779 Input Parameters: 1780 . A - the matrix 1781 1782 Output Parameters: 1783 + rmapping - row mapping 1784 - cmapping - column mapping 1785 1786 Level: advanced 1787 1788 Concepts: matrices^local to global mapping blocked 1789 Concepts: local to global mapping^for matrices, blocked 1790 1791 .seealso: MatSetValuesBlockedLocal(), MatGetLocalToGlobalMapping() 1792 @*/ 1793 PetscErrorCode MatGetLocalToGlobalMappingBlock(Mat A,ISLocalToGlobalMapping *rmapping,ISLocalToGlobalMapping *cmapping) 1794 { 1795 PetscFunctionBegin; 1796 PetscValidHeaderSpecific(A,MAT_CLASSID,1); 1797 PetscValidType(A,1); 1798 if (rmapping) PetscValidPointer(rmapping,2); 1799 if (cmapping) PetscValidPointer(cmapping,3); 1800 if (rmapping) *rmapping = A->rmap->bmapping; 1801 if (cmapping) *cmapping = A->cmap->bmapping; 1802 PetscFunctionReturn(0); 1803 } 1804 1805 #undef __FUNCT__ 1806 #define __FUNCT__ "MatSetValuesLocal" 1807 /*@ 1808 MatSetValuesLocal - Inserts or adds values into certain locations of a matrix, 1809 using a local ordering of the nodes. 1810 1811 Not Collective 1812 1813 Input Parameters: 1814 + x - the matrix 1815 . nrow, irow - number of rows and their local indices 1816 . ncol, icol - number of columns and their local indices 1817 . y - a logically two-dimensional array of values 1818 - addv - either INSERT_VALUES or ADD_VALUES, where 1819 ADD_VALUES adds values to any existing entries, and 1820 INSERT_VALUES replaces existing entries with new values 1821 1822 Notes: 1823 If you create the matrix yourself (that is not with a call to DMCreateMatrix()) then you MUST call MatXXXXSetPreallocation() or 1824 MatSetUp() before using this routine 1825 1826 If you create the matrix yourself (that is not with a call to DMCreateMatrix()) then you MUST call MatSetLocalToGlobalMapping() before using this routine 1827 1828 Calls to MatSetValuesLocal() with the INSERT_VALUES and ADD_VALUES 1829 options cannot be mixed without intervening calls to the assembly 1830 routines. 1831 1832 These values may be cached, so MatAssemblyBegin() and MatAssemblyEnd() 1833 MUST be called after all calls to MatSetValuesLocal() have been completed. 1834 1835 Level: intermediate 1836 1837 Concepts: matrices^putting entries in with local numbering 1838 1839 .seealso: MatAssemblyBegin(), MatAssemblyEnd(), MatSetValues(), MatSetLocalToGlobalMapping(), 1840 MatSetValueLocal() 1841 @*/ 1842 PetscErrorCode MatSetValuesLocal(Mat mat,PetscInt nrow,const PetscInt irow[],PetscInt ncol,const PetscInt icol[],const PetscScalar y[],InsertMode addv) 1843 { 1844 PetscErrorCode ierr; 1845 1846 PetscFunctionBegin; 1847 PetscValidHeaderSpecific(mat,MAT_CLASSID,1); 1848 PetscValidType(mat,1); 1849 MatCheckPreallocated(mat,1); 1850 if (!nrow || !ncol) PetscFunctionReturn(0); /* no values to insert */ 1851 PetscValidIntPointer(irow,3); 1852 PetscValidIntPointer(icol,5); 1853 PetscValidScalarPointer(y,6); 1854 if (mat->insertmode == NOT_SET_VALUES) { 1855 mat->insertmode = addv; 1856 } 1857 #if defined(PETSC_USE_DEBUG) 1858 else if (mat->insertmode != addv) SETERRQ(PETSC_COMM_SELF,PETSC_ERR_ARG_WRONGSTATE,"Cannot mix add values and insert values"); 1859 if (mat->factortype) SETERRQ(PETSC_COMM_SELF,PETSC_ERR_ARG_WRONGSTATE,"Not for factored matrix"); 1860 if (!mat->ops->setvalueslocal && !mat->ops->setvalues) SETERRQ1(PETSC_COMM_SELF,PETSC_ERR_SUP,"Mat type %s",((PetscObject)mat)->type_name); 1861 #endif 1862 1863 if (mat->assembled) { 1864 mat->was_assembled = PETSC_TRUE; 1865 mat->assembled = PETSC_FALSE; 1866 } 1867 ierr = PetscLogEventBegin(MAT_SetValues,mat,0,0,0);CHKERRQ(ierr); 1868 if (mat->ops->setvalueslocal) { 1869 ierr = (*mat->ops->setvalueslocal)(mat,nrow,irow,ncol,icol,y,addv);CHKERRQ(ierr); 1870 } else { 1871 PetscInt buf[8192],*bufr=0,*bufc=0,*irowm,*icolm; 1872 if ((nrow+ncol) <= (PetscInt)(sizeof(buf)/sizeof(PetscInt))) { 1873 irowm = buf; icolm = buf+nrow; 1874 } else { 1875 ierr = PetscMalloc2(nrow,PetscInt,&bufr,ncol,PetscInt,&bufc);CHKERRQ(ierr); 1876 irowm = bufr; icolm = bufc; 1877 } 1878 ierr = ISLocalToGlobalMappingApply(mat->rmap->mapping,nrow,irow,irowm);CHKERRQ(ierr); 1879 ierr = ISLocalToGlobalMappingApply(mat->cmap->mapping,ncol,icol,icolm);CHKERRQ(ierr); 1880 ierr = MatSetValues(mat,nrow,irowm,ncol,icolm,y,addv);CHKERRQ(ierr); 1881 ierr = PetscFree2(bufr,bufc);CHKERRQ(ierr); 1882 } 1883 ierr = PetscLogEventEnd(MAT_SetValues,mat,0,0,0);CHKERRQ(ierr); 1884 #if defined(PETSC_HAVE_CUSP) 1885 if (mat->valid_GPU_matrix != PETSC_CUSP_UNALLOCATED) { 1886 mat->valid_GPU_matrix = PETSC_CUSP_CPU; 1887 } 1888 #endif 1889 PetscFunctionReturn(0); 1890 } 1891 1892 #undef __FUNCT__ 1893 #define __FUNCT__ "MatSetValuesBlockedLocal" 1894 /*@ 1895 MatSetValuesBlockedLocal - Inserts or adds values into certain locations of a matrix, 1896 using a local ordering of the nodes a block at a time. 1897 1898 Not Collective 1899 1900 Input Parameters: 1901 + x - the matrix 1902 . nrow, irow - number of rows and their local indices 1903 . ncol, icol - number of columns and their local indices 1904 . y - a logically two-dimensional array of values 1905 - addv - either INSERT_VALUES or ADD_VALUES, where 1906 ADD_VALUES adds values to any existing entries, and 1907 INSERT_VALUES replaces existing entries with new values 1908 1909 Notes: 1910 If you create the matrix yourself (that is not with a call to DMCreateMatrix()) then you MUST call MatXXXXSetPreallocation() or 1911 MatSetUp() before using this routine 1912 1913 If you create the matrix yourself (that is not with a call to DMCreateMatrix()) then you MUST call MatSetBlockSize() and MatSetLocalToGlobalMappingBlock() 1914 before using this routineBefore calling MatSetValuesLocal(), the user must first set the 1915 1916 Calls to MatSetValuesBlockedLocal() with the INSERT_VALUES and ADD_VALUES 1917 options cannot be mixed without intervening calls to the assembly 1918 routines. 1919 1920 These values may be cached, so MatAssemblyBegin() and MatAssemblyEnd() 1921 MUST be called after all calls to MatSetValuesBlockedLocal() have been completed. 1922 1923 Level: intermediate 1924 1925 Concepts: matrices^putting blocked values in with local numbering 1926 1927 .seealso: MatSetBlockSize(), MatSetLocalToGlobalMappingBlock(), MatAssemblyBegin(), MatAssemblyEnd(), 1928 MatSetValuesLocal(), MatSetLocalToGlobalMappingBlock(), MatSetValuesBlocked() 1929 @*/ 1930 PetscErrorCode MatSetValuesBlockedLocal(Mat mat,PetscInt nrow,const PetscInt irow[],PetscInt ncol,const PetscInt icol[],const PetscScalar y[],InsertMode addv) 1931 { 1932 PetscErrorCode ierr; 1933 1934 PetscFunctionBegin; 1935 PetscValidHeaderSpecific(mat,MAT_CLASSID,1); 1936 PetscValidType(mat,1); 1937 MatCheckPreallocated(mat,1); 1938 if (!nrow || !ncol) PetscFunctionReturn(0); /* no values to insert */ 1939 PetscValidIntPointer(irow,3); 1940 PetscValidIntPointer(icol,5); 1941 PetscValidScalarPointer(y,6); 1942 if (mat->insertmode == NOT_SET_VALUES) { 1943 mat->insertmode = addv; 1944 } 1945 #if defined(PETSC_USE_DEBUG) 1946 else if (mat->insertmode != addv) SETERRQ(PETSC_COMM_SELF,PETSC_ERR_ARG_WRONGSTATE,"Cannot mix add values and insert values"); 1947 if (mat->factortype) SETERRQ(PETSC_COMM_SELF,PETSC_ERR_ARG_WRONGSTATE,"Not for factored matrix"); 1948 if (!mat->ops->setvaluesblockedlocal && !mat->ops->setvaluesblocked && !mat->ops->setvalueslocal && !mat->ops->setvalues) SETERRQ1(PETSC_COMM_SELF,PETSC_ERR_SUP,"Mat type %s",((PetscObject)mat)->type_name); 1949 #endif 1950 1951 if (mat->assembled) { 1952 mat->was_assembled = PETSC_TRUE; 1953 mat->assembled = PETSC_FALSE; 1954 } 1955 ierr = PetscLogEventBegin(MAT_SetValues,mat,0,0,0);CHKERRQ(ierr); 1956 if (mat->ops->setvaluesblockedlocal) { 1957 ierr = (*mat->ops->setvaluesblockedlocal)(mat,nrow,irow,ncol,icol,y,addv);CHKERRQ(ierr); 1958 } else { 1959 PetscInt buf[8192],*bufr=0,*bufc=0,*irowm,*icolm; 1960 if (mat->rmap->bmapping && mat->cmap->bmapping) { 1961 if ((nrow+ncol) <= (PetscInt)(sizeof(buf)/sizeof(PetscInt))) { 1962 irowm = buf; icolm = buf + nrow; 1963 } else { 1964 ierr = PetscMalloc2(nrow,PetscInt,&bufr,ncol,PetscInt,&bufc);CHKERRQ(ierr); 1965 irowm = bufr; icolm = bufc; 1966 } 1967 ierr = ISLocalToGlobalMappingApply(mat->rmap->bmapping,nrow,irow,irowm);CHKERRQ(ierr); 1968 ierr = ISLocalToGlobalMappingApply(mat->cmap->bmapping,ncol,icol,icolm);CHKERRQ(ierr); 1969 ierr = MatSetValuesBlocked(mat,nrow,irowm,ncol,icolm,y,addv);CHKERRQ(ierr); 1970 ierr = PetscFree2(bufr,bufc);CHKERRQ(ierr); 1971 } else { 1972 PetscInt i,j,bs=mat->rmap->bs; 1973 if ((nrow+ncol)*bs <=(PetscInt)(sizeof(buf)/sizeof(PetscInt))) { 1974 irowm = buf; icolm = buf + nrow; 1975 } else { 1976 ierr = PetscMalloc2(nrow*bs,PetscInt,&bufr,ncol*bs,PetscInt,&bufc);CHKERRQ(ierr); 1977 irowm = bufr; icolm = bufc; 1978 } 1979 for (i=0; i<nrow; i++) 1980 for (j=0; j<bs; j++) 1981 irowm[i*bs+j] = irow[i]*bs+j; 1982 for (i=0; i<ncol; i++) 1983 for (j=0; j<bs; j++) 1984 icolm[i*bs+j] = icol[i]*bs+j; 1985 ierr = MatSetValuesLocal(mat,nrow*bs,irowm,ncol*bs,icolm,y,addv);CHKERRQ(ierr); 1986 ierr = PetscFree2(bufr,bufc);CHKERRQ(ierr); 1987 } 1988 } 1989 ierr = PetscLogEventEnd(MAT_SetValues,mat,0,0,0);CHKERRQ(ierr); 1990 #if defined(PETSC_HAVE_CUSP) 1991 if (mat->valid_GPU_matrix != PETSC_CUSP_UNALLOCATED) { 1992 mat->valid_GPU_matrix = PETSC_CUSP_CPU; 1993 } 1994 #endif 1995 PetscFunctionReturn(0); 1996 } 1997 1998 #undef __FUNCT__ 1999 #define __FUNCT__ "MatMultDiagonalBlock" 2000 /*@ 2001 MatMultDiagonalBlock - Computes the matrix-vector product, y = Dx. Where D is defined by the inode or block structure of the diagonal 2002 2003 Collective on Mat and Vec 2004 2005 Input Parameters: 2006 + mat - the matrix 2007 - x - the vector to be multiplied 2008 2009 Output Parameters: 2010 . y - the result 2011 2012 Notes: 2013 The vectors x and y cannot be the same. I.e., one cannot 2014 call MatMult(A,y,y). 2015 2016 Level: developer 2017 2018 Concepts: matrix-vector product 2019 2020 .seealso: MatMultTranspose(), MatMultAdd(), MatMultTransposeAdd() 2021 @*/ 2022 PetscErrorCode MatMultDiagonalBlock(Mat mat,Vec x,Vec y) 2023 { 2024 PetscErrorCode ierr; 2025 2026 PetscFunctionBegin; 2027 PetscValidHeaderSpecific(mat,MAT_CLASSID,1); 2028 PetscValidType(mat,1); 2029 PetscValidHeaderSpecific(x,VEC_CLASSID,2); 2030 PetscValidHeaderSpecific(y,VEC_CLASSID,3); 2031 2032 if (!mat->assembled) SETERRQ(((PetscObject)mat)->comm,PETSC_ERR_ARG_WRONGSTATE,"Not for unassembled matrix"); 2033 if (mat->factortype) SETERRQ(((PetscObject)mat)->comm,PETSC_ERR_ARG_WRONGSTATE,"Not for factored matrix"); 2034 if (x == y) SETERRQ(((PetscObject)mat)->comm,PETSC_ERR_ARG_WRONGSTATE,"x and y must be different vectors"); 2035 MatCheckPreallocated(mat,1); 2036 2037 if (!mat->ops->multdiagonalblock) SETERRQ(((PetscObject)mat)->comm,PETSC_ERR_SUP,"This matrix type does not have a multiply defined"); 2038 ierr = (*mat->ops->multdiagonalblock)(mat,x,y);CHKERRQ(ierr); 2039 ierr = PetscObjectStateIncrease((PetscObject)y);CHKERRQ(ierr); 2040 PetscFunctionReturn(0); 2041 } 2042 2043 /* --------------------------------------------------------*/ 2044 #undef __FUNCT__ 2045 #define __FUNCT__ "MatMult" 2046 /*@ 2047 MatMult - Computes the matrix-vector product, y = Ax. 2048 2049 Neighbor-wise Collective on Mat and Vec 2050 2051 Input Parameters: 2052 + mat - the matrix 2053 - x - the vector to be multiplied 2054 2055 Output Parameters: 2056 . y - the result 2057 2058 Notes: 2059 The vectors x and y cannot be the same. I.e., one cannot 2060 call MatMult(A,y,y). 2061 2062 Level: beginner 2063 2064 Concepts: matrix-vector product 2065 2066 .seealso: MatMultTranspose(), MatMultAdd(), MatMultTransposeAdd() 2067 @*/ 2068 PetscErrorCode MatMult(Mat mat,Vec x,Vec y) 2069 { 2070 PetscErrorCode ierr; 2071 2072 PetscFunctionBegin; 2073 PetscValidHeaderSpecific(mat,MAT_CLASSID,1); 2074 PetscValidType(mat,1); 2075 PetscValidHeaderSpecific(x,VEC_CLASSID,2); 2076 PetscValidHeaderSpecific(y,VEC_CLASSID,3); 2077 if (!mat->assembled) SETERRQ(((PetscObject)mat)->comm,PETSC_ERR_ARG_WRONGSTATE,"Not for unassembled matrix"); 2078 if (mat->factortype) SETERRQ(((PetscObject)mat)->comm,PETSC_ERR_ARG_WRONGSTATE,"Not for factored matrix"); 2079 if (x == y) SETERRQ(((PetscObject)mat)->comm,PETSC_ERR_ARG_WRONGSTATE,"x and y must be different vectors"); 2080 #ifndef PETSC_HAVE_CONSTRAINTS 2081 if (mat->cmap->N != x->map->N) SETERRQ2(PETSC_COMM_SELF,PETSC_ERR_ARG_SIZ,"Mat mat,Vec x: global dim %D %D",mat->cmap->N,x->map->N); 2082 if (mat->rmap->N != y->map->N) SETERRQ2(PETSC_COMM_SELF,PETSC_ERR_ARG_SIZ,"Mat mat,Vec y: global dim %D %D",mat->rmap->N,y->map->N); 2083 if (mat->rmap->n != y->map->n) SETERRQ2(PETSC_COMM_SELF,PETSC_ERR_ARG_SIZ,"Mat mat,Vec y: local dim %D %D",mat->rmap->n,y->map->n); 2084 #endif 2085 VecValidValues(x,2,PETSC_TRUE); 2086 MatCheckPreallocated(mat,1); 2087 2088 if (!mat->ops->mult) SETERRQ(((PetscObject)mat)->comm,PETSC_ERR_SUP,"This matrix type does not have a multiply defined"); 2089 ierr = PetscLogEventBegin(MAT_Mult,mat,x,y,0);CHKERRQ(ierr); 2090 ierr = (*mat->ops->mult)(mat,x,y);CHKERRQ(ierr); 2091 ierr = PetscLogEventEnd(MAT_Mult,mat,x,y,0);CHKERRQ(ierr); 2092 VecValidValues(y,3,PETSC_FALSE); 2093 PetscFunctionReturn(0); 2094 } 2095 2096 #undef __FUNCT__ 2097 #define __FUNCT__ "MatMultTranspose" 2098 /*@ 2099 MatMultTranspose - Computes matrix transpose times a vector. 2100 2101 Neighbor-wise Collective on Mat and Vec 2102 2103 Input Parameters: 2104 + mat - the matrix 2105 - x - the vector to be multilplied 2106 2107 Output Parameters: 2108 . y - the result 2109 2110 Notes: 2111 The vectors x and y cannot be the same. I.e., one cannot 2112 call MatMultTranspose(A,y,y). 2113 2114 For complex numbers this does NOT compute the Hermitian (complex conjugate) transpose multiple, 2115 use MatMultHermitianTranspose() 2116 2117 Level: beginner 2118 2119 Concepts: matrix vector product^transpose 2120 2121 .seealso: MatMult(), MatMultAdd(), MatMultTransposeAdd(), MatMultHermitianTranspose(), MatTranspose() 2122 @*/ 2123 PetscErrorCode MatMultTranspose(Mat mat,Vec x,Vec y) 2124 { 2125 PetscErrorCode ierr; 2126 2127 PetscFunctionBegin; 2128 PetscValidHeaderSpecific(mat,MAT_CLASSID,1); 2129 PetscValidType(mat,1); 2130 PetscValidHeaderSpecific(x,VEC_CLASSID,2); 2131 PetscValidHeaderSpecific(y,VEC_CLASSID,3); 2132 2133 if (!mat->assembled) SETERRQ(((PetscObject)mat)->comm,PETSC_ERR_ARG_WRONGSTATE,"Not for unassembled matrix"); 2134 if (mat->factortype) SETERRQ(((PetscObject)mat)->comm,PETSC_ERR_ARG_WRONGSTATE,"Not for factored matrix"); 2135 if (x == y) SETERRQ(((PetscObject)mat)->comm,PETSC_ERR_ARG_WRONGSTATE,"x and y must be different vectors"); 2136 #ifndef PETSC_HAVE_CONSTRAINTS 2137 if (mat->rmap->N != x->map->N) SETERRQ2(PETSC_COMM_SELF,PETSC_ERR_ARG_SIZ,"Mat mat,Vec x: global dim %D %D",mat->rmap->N,x->map->N); 2138 if (mat->cmap->N != y->map->N) SETERRQ2(PETSC_COMM_SELF,PETSC_ERR_ARG_SIZ,"Mat mat,Vec y: global dim %D %D",mat->cmap->N,y->map->N); 2139 #endif 2140 VecValidValues(x,2,PETSC_TRUE); 2141 MatCheckPreallocated(mat,1); 2142 2143 if (!mat->ops->multtranspose) SETERRQ(((PetscObject)mat)->comm,PETSC_ERR_SUP,"This matrix type does not have a multiply tranpose defined"); 2144 ierr = PetscLogEventBegin(MAT_MultTranspose,mat,x,y,0);CHKERRQ(ierr); 2145 ierr = (*mat->ops->multtranspose)(mat,x,y);CHKERRQ(ierr); 2146 ierr = PetscLogEventEnd(MAT_MultTranspose,mat,x,y,0);CHKERRQ(ierr); 2147 ierr = PetscObjectStateIncrease((PetscObject)y);CHKERRQ(ierr); 2148 VecValidValues(y,3,PETSC_FALSE); 2149 PetscFunctionReturn(0); 2150 } 2151 2152 #undef __FUNCT__ 2153 #define __FUNCT__ "MatMultHermitianTranspose" 2154 /*@ 2155 MatMultHermitianTranspose - Computes matrix Hermitian transpose times a vector. 2156 2157 Neighbor-wise Collective on Mat and Vec 2158 2159 Input Parameters: 2160 + mat - the matrix 2161 - x - the vector to be multilplied 2162 2163 Output Parameters: 2164 . y - the result 2165 2166 Notes: 2167 The vectors x and y cannot be the same. I.e., one cannot 2168 call MatMultHermitianTranspose(A,y,y). 2169 2170 Also called the conjugate transpose, complex conjugate transpose, or adjoint. 2171 2172 For real numbers MatMultTranspose() and MatMultHermitianTranspose() are identical. 2173 2174 Level: beginner 2175 2176 Concepts: matrix vector product^transpose 2177 2178 .seealso: MatMult(), MatMultAdd(), MatMultHermitianTransposeAdd(), MatMultTranspose() 2179 @*/ 2180 PetscErrorCode MatMultHermitianTranspose(Mat mat,Vec x,Vec y) 2181 { 2182 PetscErrorCode ierr; 2183 Vec w; 2184 2185 PetscFunctionBegin; 2186 PetscValidHeaderSpecific(mat,MAT_CLASSID,1); 2187 PetscValidType(mat,1); 2188 PetscValidHeaderSpecific(x,VEC_CLASSID,2); 2189 PetscValidHeaderSpecific(y,VEC_CLASSID,3); 2190 2191 if (!mat->assembled) SETERRQ(((PetscObject)mat)->comm,PETSC_ERR_ARG_WRONGSTATE,"Not for unassembled matrix"); 2192 if (mat->factortype) SETERRQ(((PetscObject)mat)->comm,PETSC_ERR_ARG_WRONGSTATE,"Not for factored matrix"); 2193 if (x == y) SETERRQ(((PetscObject)mat)->comm,PETSC_ERR_ARG_WRONGSTATE,"x and y must be different vectors"); 2194 #ifndef PETSC_HAVE_CONSTRAINTS 2195 if (mat->rmap->N != x->map->N) SETERRQ2(PETSC_COMM_SELF,PETSC_ERR_ARG_SIZ,"Mat mat,Vec x: global dim %D %D",mat->rmap->N,x->map->N); 2196 if (mat->cmap->N != y->map->N) SETERRQ2(PETSC_COMM_SELF,PETSC_ERR_ARG_SIZ,"Mat mat,Vec y: global dim %D %D",mat->cmap->N,y->map->N); 2197 #endif 2198 MatCheckPreallocated(mat,1); 2199 2200 ierr = PetscLogEventBegin(MAT_MultHermitianTranspose,mat,x,y,0);CHKERRQ(ierr); 2201 if (mat->ops->multhermitiantranspose) { 2202 ierr = (*mat->ops->multhermitiantranspose)(mat,x,y);CHKERRQ(ierr); 2203 } else { 2204 ierr = VecDuplicate(x,&w);CHKERRQ(ierr); 2205 ierr = VecCopy(x,w);CHKERRQ(ierr); 2206 ierr = VecConjugate(w);CHKERRQ(ierr); 2207 ierr = MatMultTranspose(mat,w,y);CHKERRQ(ierr); 2208 ierr = VecDestroy(&w);CHKERRQ(ierr); 2209 ierr = VecConjugate(y);CHKERRQ(ierr); 2210 } 2211 ierr = PetscLogEventEnd(MAT_MultHermitianTranspose,mat,x,y,0);CHKERRQ(ierr); 2212 ierr = PetscObjectStateIncrease((PetscObject)y);CHKERRQ(ierr); 2213 PetscFunctionReturn(0); 2214 } 2215 2216 #undef __FUNCT__ 2217 #define __FUNCT__ "MatMultAdd" 2218 /*@ 2219 MatMultAdd - Computes v3 = v2 + A * v1. 2220 2221 Neighbor-wise Collective on Mat and Vec 2222 2223 Input Parameters: 2224 + mat - the matrix 2225 - v1, v2 - the vectors 2226 2227 Output Parameters: 2228 . v3 - the result 2229 2230 Notes: 2231 The vectors v1 and v3 cannot be the same. I.e., one cannot 2232 call MatMultAdd(A,v1,v2,v1). 2233 2234 Level: beginner 2235 2236 Concepts: matrix vector product^addition 2237 2238 .seealso: MatMultTranspose(), MatMult(), MatMultTransposeAdd() 2239 @*/ 2240 PetscErrorCode MatMultAdd(Mat mat,Vec v1,Vec v2,Vec v3) 2241 { 2242 PetscErrorCode ierr; 2243 2244 PetscFunctionBegin; 2245 PetscValidHeaderSpecific(mat,MAT_CLASSID,1); 2246 PetscValidType(mat,1); 2247 PetscValidHeaderSpecific(v1,VEC_CLASSID,2); 2248 PetscValidHeaderSpecific(v2,VEC_CLASSID,3); 2249 PetscValidHeaderSpecific(v3,VEC_CLASSID,4); 2250 2251 if (!mat->assembled) SETERRQ(((PetscObject)mat)->comm,PETSC_ERR_ARG_WRONGSTATE,"Not for unassembled matrix"); 2252 if (mat->factortype) SETERRQ(((PetscObject)mat)->comm,PETSC_ERR_ARG_WRONGSTATE,"Not for factored matrix"); 2253 if (mat->cmap->N != v1->map->N) SETERRQ2(((PetscObject)mat)->comm,PETSC_ERR_ARG_SIZ,"Mat mat,Vec v1: global dim %D %D",mat->cmap->N,v1->map->N); 2254 /* if (mat->rmap->N != v2->map->N) SETERRQ2(PETSC_COMM_SELF,PETSC_ERR_ARG_SIZ,"Mat mat,Vec v2: global dim %D %D",mat->rmap->N,v2->map->N); 2255 if (mat->rmap->N != v3->map->N) SETERRQ2(PETSC_COMM_SELF,PETSC_ERR_ARG_SIZ,"Mat mat,Vec v3: global dim %D %D",mat->rmap->N,v3->map->N); */ 2256 if (mat->rmap->n != v3->map->n) SETERRQ2(PETSC_COMM_SELF,PETSC_ERR_ARG_SIZ,"Mat mat,Vec v3: local dim %D %D",mat->rmap->n,v3->map->n); 2257 if (mat->rmap->n != v2->map->n) SETERRQ2(PETSC_COMM_SELF,PETSC_ERR_ARG_SIZ,"Mat mat,Vec v2: local dim %D %D",mat->rmap->n,v2->map->n); 2258 if (v1 == v3) SETERRQ(((PetscObject)mat)->comm,PETSC_ERR_ARG_IDN,"v1 and v3 must be different vectors"); 2259 MatCheckPreallocated(mat,1); 2260 2261 if (!mat->ops->multadd) SETERRQ1(((PetscObject)mat)->comm,PETSC_ERR_SUP,"No MatMultAdd() for matrix type '%s'",((PetscObject)mat)->type_name); 2262 ierr = PetscLogEventBegin(MAT_MultAdd,mat,v1,v2,v3);CHKERRQ(ierr); 2263 ierr = (*mat->ops->multadd)(mat,v1,v2,v3);CHKERRQ(ierr); 2264 ierr = PetscLogEventEnd(MAT_MultAdd,mat,v1,v2,v3);CHKERRQ(ierr); 2265 ierr = PetscObjectStateIncrease((PetscObject)v3);CHKERRQ(ierr); 2266 PetscFunctionReturn(0); 2267 } 2268 2269 #undef __FUNCT__ 2270 #define __FUNCT__ "MatMultTransposeAdd" 2271 /*@ 2272 MatMultTransposeAdd - Computes v3 = v2 + A' * v1. 2273 2274 Neighbor-wise Collective on Mat and Vec 2275 2276 Input Parameters: 2277 + mat - the matrix 2278 - v1, v2 - the vectors 2279 2280 Output Parameters: 2281 . v3 - the result 2282 2283 Notes: 2284 The vectors v1 and v3 cannot be the same. I.e., one cannot 2285 call MatMultTransposeAdd(A,v1,v2,v1). 2286 2287 Level: beginner 2288 2289 Concepts: matrix vector product^transpose and addition 2290 2291 .seealso: MatMultTranspose(), MatMultAdd(), MatMult() 2292 @*/ 2293 PetscErrorCode MatMultTransposeAdd(Mat mat,Vec v1,Vec v2,Vec v3) 2294 { 2295 PetscErrorCode ierr; 2296 2297 PetscFunctionBegin; 2298 PetscValidHeaderSpecific(mat,MAT_CLASSID,1); 2299 PetscValidType(mat,1); 2300 PetscValidHeaderSpecific(v1,VEC_CLASSID,2); 2301 PetscValidHeaderSpecific(v2,VEC_CLASSID,3); 2302 PetscValidHeaderSpecific(v3,VEC_CLASSID,4); 2303 2304 if (!mat->assembled) SETERRQ(((PetscObject)mat)->comm,PETSC_ERR_ARG_WRONGSTATE,"Not for unassembled matrix"); 2305 if (mat->factortype) SETERRQ(((PetscObject)mat)->comm,PETSC_ERR_ARG_WRONGSTATE,"Not for factored matrix"); 2306 if (!mat->ops->multtransposeadd) SETERRQ1(((PetscObject)mat)->comm,PETSC_ERR_SUP,"Mat type %s",((PetscObject)mat)->type_name); 2307 if (v1 == v3) SETERRQ(((PetscObject)mat)->comm,PETSC_ERR_ARG_IDN,"v1 and v3 must be different vectors"); 2308 if (mat->rmap->N != v1->map->N) SETERRQ2(((PetscObject)mat)->comm,PETSC_ERR_ARG_SIZ,"Mat mat,Vec v1: global dim %D %D",mat->rmap->N,v1->map->N); 2309 if (mat->cmap->N != v2->map->N) SETERRQ2(((PetscObject)mat)->comm,PETSC_ERR_ARG_SIZ,"Mat mat,Vec v2: global dim %D %D",mat->cmap->N,v2->map->N); 2310 if (mat->cmap->N != v3->map->N) SETERRQ2(((PetscObject)mat)->comm,PETSC_ERR_ARG_SIZ,"Mat mat,Vec v3: global dim %D %D",mat->cmap->N,v3->map->N); 2311 MatCheckPreallocated(mat,1); 2312 2313 ierr = PetscLogEventBegin(MAT_MultTransposeAdd,mat,v1,v2,v3);CHKERRQ(ierr); 2314 ierr = (*mat->ops->multtransposeadd)(mat,v1,v2,v3);CHKERRQ(ierr); 2315 ierr = PetscLogEventEnd(MAT_MultTransposeAdd,mat,v1,v2,v3);CHKERRQ(ierr); 2316 ierr = PetscObjectStateIncrease((PetscObject)v3);CHKERRQ(ierr); 2317 PetscFunctionReturn(0); 2318 } 2319 2320 #undef __FUNCT__ 2321 #define __FUNCT__ "MatMultHermitianTransposeAdd" 2322 /*@ 2323 MatMultHermitianTransposeAdd - Computes v3 = v2 + A^H * v1. 2324 2325 Neighbor-wise Collective on Mat and Vec 2326 2327 Input Parameters: 2328 + mat - the matrix 2329 - v1, v2 - the vectors 2330 2331 Output Parameters: 2332 . v3 - the result 2333 2334 Notes: 2335 The vectors v1 and v3 cannot be the same. I.e., one cannot 2336 call MatMultHermitianTransposeAdd(A,v1,v2,v1). 2337 2338 Level: beginner 2339 2340 Concepts: matrix vector product^transpose and addition 2341 2342 .seealso: MatMultHermitianTranspose(), MatMultTranspose(), MatMultAdd(), MatMult() 2343 @*/ 2344 PetscErrorCode MatMultHermitianTransposeAdd(Mat mat,Vec v1,Vec v2,Vec v3) 2345 { 2346 PetscErrorCode ierr; 2347 2348 PetscFunctionBegin; 2349 PetscValidHeaderSpecific(mat,MAT_CLASSID,1); 2350 PetscValidType(mat,1); 2351 PetscValidHeaderSpecific(v1,VEC_CLASSID,2); 2352 PetscValidHeaderSpecific(v2,VEC_CLASSID,3); 2353 PetscValidHeaderSpecific(v3,VEC_CLASSID,4); 2354 2355 if (!mat->assembled) SETERRQ(((PetscObject)mat)->comm,PETSC_ERR_ARG_WRONGSTATE,"Not for unassembled matrix"); 2356 if (mat->factortype) SETERRQ(((PetscObject)mat)->comm,PETSC_ERR_ARG_WRONGSTATE,"Not for factored matrix"); 2357 if (!mat->ops->multhermitiantransposeadd) SETERRQ1(((PetscObject)mat)->comm,PETSC_ERR_SUP,"Mat type %s",((PetscObject)mat)->type_name); 2358 if (v1 == v3) SETERRQ(((PetscObject)mat)->comm,PETSC_ERR_ARG_IDN,"v1 and v3 must be different vectors"); 2359 if (mat->rmap->N != v1->map->N) SETERRQ2(((PetscObject)mat)->comm,PETSC_ERR_ARG_SIZ,"Mat mat,Vec v1: global dim %D %D",mat->rmap->N,v1->map->N); 2360 if (mat->cmap->N != v2->map->N) SETERRQ2(((PetscObject)mat)->comm,PETSC_ERR_ARG_SIZ,"Mat mat,Vec v2: global dim %D %D",mat->cmap->N,v2->map->N); 2361 if (mat->cmap->N != v3->map->N) SETERRQ2(((PetscObject)mat)->comm,PETSC_ERR_ARG_SIZ,"Mat mat,Vec v3: global dim %D %D",mat->cmap->N,v3->map->N); 2362 MatCheckPreallocated(mat,1); 2363 2364 ierr = PetscLogEventBegin(MAT_MultHermitianTransposeAdd,mat,v1,v2,v3);CHKERRQ(ierr); 2365 ierr = (*mat->ops->multhermitiantransposeadd)(mat,v1,v2,v3);CHKERRQ(ierr); 2366 ierr = PetscLogEventEnd(MAT_MultHermitianTransposeAdd,mat,v1,v2,v3);CHKERRQ(ierr); 2367 ierr = PetscObjectStateIncrease((PetscObject)v3);CHKERRQ(ierr); 2368 PetscFunctionReturn(0); 2369 } 2370 2371 #undef __FUNCT__ 2372 #define __FUNCT__ "MatMultConstrained" 2373 /*@ 2374 MatMultConstrained - The inner multiplication routine for a 2375 constrained matrix P^T A P. 2376 2377 Neighbor-wise Collective on Mat and Vec 2378 2379 Input Parameters: 2380 + mat - the matrix 2381 - x - the vector to be multilplied 2382 2383 Output Parameters: 2384 . y - the result 2385 2386 Notes: 2387 The vectors x and y cannot be the same. I.e., one cannot 2388 call MatMult(A,y,y). 2389 2390 Level: beginner 2391 2392 .keywords: matrix, multiply, matrix-vector product, constraint 2393 .seealso: MatMult(), MatMultTranspose(), MatMultAdd(), MatMultTransposeAdd() 2394 @*/ 2395 PetscErrorCode MatMultConstrained(Mat mat,Vec x,Vec y) 2396 { 2397 PetscErrorCode ierr; 2398 2399 PetscFunctionBegin; 2400 PetscValidHeaderSpecific(mat,MAT_CLASSID,1); 2401 PetscValidHeaderSpecific(x,VEC_CLASSID,2); 2402 PetscValidHeaderSpecific(y,VEC_CLASSID,3); 2403 if (!mat->assembled) SETERRQ(PETSC_COMM_SELF,PETSC_ERR_ARG_WRONGSTATE,"Not for unassembled matrix"); 2404 if (mat->factortype) SETERRQ(PETSC_COMM_SELF,PETSC_ERR_ARG_WRONGSTATE,"Not for factored matrix"); 2405 if (x == y) SETERRQ(PETSC_COMM_SELF,PETSC_ERR_ARG_WRONGSTATE,"x and y must be different vectors"); 2406 if (mat->cmap->N != x->map->N) SETERRQ2(PETSC_COMM_SELF,PETSC_ERR_ARG_SIZ,"Mat mat,Vec x: global dim %D %D",mat->cmap->N,x->map->N); 2407 if (mat->rmap->N != y->map->N) SETERRQ2(PETSC_COMM_SELF,PETSC_ERR_ARG_SIZ,"Mat mat,Vec y: global dim %D %D",mat->rmap->N,y->map->N); 2408 if (mat->rmap->n != y->map->n) SETERRQ2(PETSC_COMM_SELF,PETSC_ERR_ARG_SIZ,"Mat mat,Vec y: local dim %D %D",mat->rmap->n,y->map->n); 2409 2410 ierr = PetscLogEventBegin(MAT_MultConstrained,mat,x,y,0);CHKERRQ(ierr); 2411 ierr = (*mat->ops->multconstrained)(mat,x,y);CHKERRQ(ierr); 2412 ierr = PetscLogEventEnd(MAT_MultConstrained,mat,x,y,0);CHKERRQ(ierr); 2413 ierr = PetscObjectStateIncrease((PetscObject)y);CHKERRQ(ierr); 2414 2415 PetscFunctionReturn(0); 2416 } 2417 2418 #undef __FUNCT__ 2419 #define __FUNCT__ "MatMultTransposeConstrained" 2420 /*@ 2421 MatMultTransposeConstrained - The inner multiplication routine for a 2422 constrained matrix P^T A^T P. 2423 2424 Neighbor-wise Collective on Mat and Vec 2425 2426 Input Parameters: 2427 + mat - the matrix 2428 - x - the vector to be multilplied 2429 2430 Output Parameters: 2431 . y - the result 2432 2433 Notes: 2434 The vectors x and y cannot be the same. I.e., one cannot 2435 call MatMult(A,y,y). 2436 2437 Level: beginner 2438 2439 .keywords: matrix, multiply, matrix-vector product, constraint 2440 .seealso: MatMult(), MatMultTranspose(), MatMultAdd(), MatMultTransposeAdd() 2441 @*/ 2442 PetscErrorCode MatMultTransposeConstrained(Mat mat,Vec x,Vec y) 2443 { 2444 PetscErrorCode ierr; 2445 2446 PetscFunctionBegin; 2447 PetscValidHeaderSpecific(mat,MAT_CLASSID,1); 2448 PetscValidHeaderSpecific(x,VEC_CLASSID,2); 2449 PetscValidHeaderSpecific(y,VEC_CLASSID,3); 2450 if (!mat->assembled) SETERRQ(PETSC_COMM_SELF,PETSC_ERR_ARG_WRONGSTATE,"Not for unassembled matrix"); 2451 if (mat->factortype) SETERRQ(PETSC_COMM_SELF,PETSC_ERR_ARG_WRONGSTATE,"Not for factored matrix"); 2452 if (x == y) SETERRQ(PETSC_COMM_SELF,PETSC_ERR_ARG_WRONGSTATE,"x and y must be different vectors"); 2453 if (mat->rmap->N != x->map->N) SETERRQ2(PETSC_COMM_SELF,PETSC_ERR_ARG_SIZ,"Mat mat,Vec x: global dim %D %D",mat->cmap->N,x->map->N); 2454 if (mat->cmap->N != y->map->N) SETERRQ2(PETSC_COMM_SELF,PETSC_ERR_ARG_SIZ,"Mat mat,Vec y: global dim %D %D",mat->rmap->N,y->map->N); 2455 2456 ierr = PetscLogEventBegin(MAT_MultConstrained,mat,x,y,0);CHKERRQ(ierr); 2457 ierr = (*mat->ops->multtransposeconstrained)(mat,x,y);CHKERRQ(ierr); 2458 ierr = PetscLogEventEnd(MAT_MultConstrained,mat,x,y,0);CHKERRQ(ierr); 2459 ierr = PetscObjectStateIncrease((PetscObject)y);CHKERRQ(ierr); 2460 2461 PetscFunctionReturn(0); 2462 } 2463 2464 #undef __FUNCT__ 2465 #define __FUNCT__ "MatGetFactorType" 2466 /*@C 2467 MatGetFactorType - gets the type of factorization it is 2468 2469 Note Collective 2470 as the flag 2471 2472 Input Parameters: 2473 . mat - the matrix 2474 2475 Output Parameters: 2476 . t - the type, one of MAT_FACTOR_NONE, MAT_FACTOR_LU, MAT_FACTOR_CHOLESKY, MAT_FACTOR_ILU, MAT_FACTOR_ICC,MAT_FACTOR_ILUDT 2477 2478 Level: intermediate 2479 2480 .seealso: MatFactorType, MatGetFactor() 2481 @*/ 2482 PetscErrorCode MatGetFactorType(Mat mat,MatFactorType *t) 2483 { 2484 PetscFunctionBegin; 2485 PetscValidHeaderSpecific(mat,MAT_CLASSID,1); 2486 PetscValidType(mat,1); 2487 *t = mat->factortype; 2488 PetscFunctionReturn(0); 2489 } 2490 2491 /* ------------------------------------------------------------*/ 2492 #undef __FUNCT__ 2493 #define __FUNCT__ "MatGetInfo" 2494 /*@C 2495 MatGetInfo - Returns information about matrix storage (number of 2496 nonzeros, memory, etc.). 2497 2498 Collective on Mat if MAT_GLOBAL_MAX or MAT_GLOBAL_SUM is used as the flag 2499 2500 Input Parameters: 2501 . mat - the matrix 2502 2503 Output Parameters: 2504 + flag - flag indicating the type of parameters to be returned 2505 (MAT_LOCAL - local matrix, MAT_GLOBAL_MAX - maximum over all processors, 2506 MAT_GLOBAL_SUM - sum over all processors) 2507 - info - matrix information context 2508 2509 Notes: 2510 The MatInfo context contains a variety of matrix data, including 2511 number of nonzeros allocated and used, number of mallocs during 2512 matrix assembly, etc. Additional information for factored matrices 2513 is provided (such as the fill ratio, number of mallocs during 2514 factorization, etc.). Much of this info is printed to PETSC_STDOUT 2515 when using the runtime options 2516 $ -info -mat_view_info 2517 2518 Example for C/C++ Users: 2519 See the file ${PETSC_DIR}/include/petscmat.h for a complete list of 2520 data within the MatInfo context. For example, 2521 .vb 2522 MatInfo info; 2523 Mat A; 2524 double mal, nz_a, nz_u; 2525 2526 MatGetInfo(A,MAT_LOCAL,&info); 2527 mal = info.mallocs; 2528 nz_a = info.nz_allocated; 2529 .ve 2530 2531 Example for Fortran Users: 2532 Fortran users should declare info as a double precision 2533 array of dimension MAT_INFO_SIZE, and then extract the parameters 2534 of interest. See the file ${PETSC_DIR}/include/finclude/petscmat.h 2535 a complete list of parameter names. 2536 .vb 2537 double precision info(MAT_INFO_SIZE) 2538 double precision mal, nz_a 2539 Mat A 2540 integer ierr 2541 2542 call MatGetInfo(A,MAT_LOCAL,info,ierr) 2543 mal = info(MAT_INFO_MALLOCS) 2544 nz_a = info(MAT_INFO_NZ_ALLOCATED) 2545 .ve 2546 2547 Level: intermediate 2548 2549 Concepts: matrices^getting information on 2550 2551 Developer Note: fortran interface is not autogenerated as the f90 2552 interface defintion cannot be generated correctly [due to MatInfo] 2553 2554 .seealso: MatStashGetInfo() 2555 2556 @*/ 2557 PetscErrorCode MatGetInfo(Mat mat,MatInfoType flag,MatInfo *info) 2558 { 2559 PetscErrorCode ierr; 2560 2561 PetscFunctionBegin; 2562 PetscValidHeaderSpecific(mat,MAT_CLASSID,1); 2563 PetscValidType(mat,1); 2564 PetscValidPointer(info,3); 2565 if (!mat->ops->getinfo) SETERRQ1(PETSC_COMM_SELF,PETSC_ERR_SUP,"Mat type %s",((PetscObject)mat)->type_name); 2566 MatCheckPreallocated(mat,1); 2567 ierr = (*mat->ops->getinfo)(mat,flag,info);CHKERRQ(ierr); 2568 PetscFunctionReturn(0); 2569 } 2570 2571 /* ----------------------------------------------------------*/ 2572 2573 #undef __FUNCT__ 2574 #define __FUNCT__ "MatLUFactor" 2575 /*@C 2576 MatLUFactor - Performs in-place LU factorization of matrix. 2577 2578 Collective on Mat 2579 2580 Input Parameters: 2581 + mat - the matrix 2582 . row - row permutation 2583 . col - column permutation 2584 - info - options for factorization, includes 2585 $ fill - expected fill as ratio of original fill. 2586 $ dtcol - pivot tolerance (0 no pivot, 1 full column pivoting) 2587 $ Run with the option -info to determine an optimal value to use 2588 2589 Notes: 2590 Most users should employ the simplified KSP interface for linear solvers 2591 instead of working directly with matrix algebra routines such as this. 2592 See, e.g., KSPCreate(). 2593 2594 This changes the state of the matrix to a factored matrix; it cannot be used 2595 for example with MatSetValues() unless one first calls MatSetUnfactored(). 2596 2597 Level: developer 2598 2599 Concepts: matrices^LU factorization 2600 2601 .seealso: MatLUFactorSymbolic(), MatLUFactorNumeric(), MatCholeskyFactor(), 2602 MatGetOrdering(), MatSetUnfactored(), MatFactorInfo, MatGetFactor() 2603 2604 Developer Note: fortran interface is not autogenerated as the f90 2605 interface defintion cannot be generated correctly [due to MatFactorInfo] 2606 2607 @*/ 2608 PetscErrorCode MatLUFactor(Mat mat,IS row,IS col,const MatFactorInfo *info) 2609 { 2610 PetscErrorCode ierr; 2611 MatFactorInfo tinfo; 2612 2613 PetscFunctionBegin; 2614 PetscValidHeaderSpecific(mat,MAT_CLASSID,1); 2615 if (row) PetscValidHeaderSpecific(row,IS_CLASSID,2); 2616 if (col) PetscValidHeaderSpecific(col,IS_CLASSID,3); 2617 if (info) PetscValidPointer(info,4); 2618 PetscValidType(mat,1); 2619 if (!mat->assembled) SETERRQ(((PetscObject)mat)->comm,PETSC_ERR_ARG_WRONGSTATE,"Not for unassembled matrix"); 2620 if (mat->factortype) SETERRQ(((PetscObject)mat)->comm,PETSC_ERR_ARG_WRONGSTATE,"Not for factored matrix"); 2621 if (!mat->ops->lufactor) SETERRQ1(((PetscObject)mat)->comm,PETSC_ERR_SUP,"Mat type %s",((PetscObject)mat)->type_name); 2622 MatCheckPreallocated(mat,1); 2623 if (!info) { 2624 ierr = MatFactorInfoInitialize(&tinfo);CHKERRQ(ierr); 2625 info = &tinfo; 2626 } 2627 2628 ierr = PetscLogEventBegin(MAT_LUFactor,mat,row,col,0);CHKERRQ(ierr); 2629 ierr = (*mat->ops->lufactor)(mat,row,col,info);CHKERRQ(ierr); 2630 ierr = PetscLogEventEnd(MAT_LUFactor,mat,row,col,0);CHKERRQ(ierr); 2631 ierr = PetscObjectStateIncrease((PetscObject)mat);CHKERRQ(ierr); 2632 PetscFunctionReturn(0); 2633 } 2634 2635 #undef __FUNCT__ 2636 #define __FUNCT__ "MatILUFactor" 2637 /*@C 2638 MatILUFactor - Performs in-place ILU factorization of matrix. 2639 2640 Collective on Mat 2641 2642 Input Parameters: 2643 + mat - the matrix 2644 . row - row permutation 2645 . col - column permutation 2646 - info - structure containing 2647 $ levels - number of levels of fill. 2648 $ expected fill - as ratio of original fill. 2649 $ 1 or 0 - indicating force fill on diagonal (improves robustness for matrices 2650 missing diagonal entries) 2651 2652 Notes: 2653 Probably really in-place only when level of fill is zero, otherwise allocates 2654 new space to store factored matrix and deletes previous memory. 2655 2656 Most users should employ the simplified KSP interface for linear solvers 2657 instead of working directly with matrix algebra routines such as this. 2658 See, e.g., KSPCreate(). 2659 2660 Level: developer 2661 2662 Concepts: matrices^ILU factorization 2663 2664 .seealso: MatILUFactorSymbolic(), MatLUFactorNumeric(), MatCholeskyFactor(), MatFactorInfo 2665 2666 Developer Note: fortran interface is not autogenerated as the f90 2667 interface defintion cannot be generated correctly [due to MatFactorInfo] 2668 2669 @*/ 2670 PetscErrorCode MatILUFactor(Mat mat,IS row,IS col,const MatFactorInfo *info) 2671 { 2672 PetscErrorCode ierr; 2673 2674 PetscFunctionBegin; 2675 PetscValidHeaderSpecific(mat,MAT_CLASSID,1); 2676 if (row) PetscValidHeaderSpecific(row,IS_CLASSID,2); 2677 if (col) PetscValidHeaderSpecific(col,IS_CLASSID,3); 2678 PetscValidPointer(info,4); 2679 PetscValidType(mat,1); 2680 if (mat->rmap->N != mat->cmap->N) SETERRQ(((PetscObject)mat)->comm,PETSC_ERR_ARG_WRONG,"matrix must be square"); 2681 if (!mat->assembled) SETERRQ(((PetscObject)mat)->comm,PETSC_ERR_ARG_WRONGSTATE,"Not for unassembled matrix"); 2682 if (mat->factortype) SETERRQ(((PetscObject)mat)->comm,PETSC_ERR_ARG_WRONGSTATE,"Not for factored matrix"); 2683 if (!mat->ops->ilufactor) SETERRQ1(((PetscObject)mat)->comm,PETSC_ERR_SUP,"Mat type %s",((PetscObject)mat)->type_name); 2684 MatCheckPreallocated(mat,1); 2685 2686 ierr = PetscLogEventBegin(MAT_ILUFactor,mat,row,col,0);CHKERRQ(ierr); 2687 ierr = (*mat->ops->ilufactor)(mat,row,col,info);CHKERRQ(ierr); 2688 ierr = PetscLogEventEnd(MAT_ILUFactor,mat,row,col,0);CHKERRQ(ierr); 2689 ierr = PetscObjectStateIncrease((PetscObject)mat);CHKERRQ(ierr); 2690 PetscFunctionReturn(0); 2691 } 2692 2693 #undef __FUNCT__ 2694 #define __FUNCT__ "MatLUFactorSymbolic" 2695 /*@C 2696 MatLUFactorSymbolic - Performs symbolic LU factorization of matrix. 2697 Call this routine before calling MatLUFactorNumeric(). 2698 2699 Collective on Mat 2700 2701 Input Parameters: 2702 + fact - the factor matrix obtained with MatGetFactor() 2703 . mat - the matrix 2704 . row, col - row and column permutations 2705 - info - options for factorization, includes 2706 $ fill - expected fill as ratio of original fill. 2707 $ dtcol - pivot tolerance (0 no pivot, 1 full column pivoting) 2708 $ Run with the option -info to determine an optimal value to use 2709 2710 2711 Notes: 2712 See the <a href="../../docs/manual.pdf">users manual</a> for additional information about 2713 choosing the fill factor for better efficiency. 2714 2715 Most users should employ the simplified KSP interface for linear solvers 2716 instead of working directly with matrix algebra routines such as this. 2717 See, e.g., KSPCreate(). 2718 2719 Level: developer 2720 2721 Concepts: matrices^LU symbolic factorization 2722 2723 .seealso: MatLUFactor(), MatLUFactorNumeric(), MatCholeskyFactor(), MatFactorInfo 2724 2725 Developer Note: fortran interface is not autogenerated as the f90 2726 interface defintion cannot be generated correctly [due to MatFactorInfo] 2727 2728 @*/ 2729 PetscErrorCode MatLUFactorSymbolic(Mat fact,Mat mat,IS row,IS col,const MatFactorInfo *info) 2730 { 2731 PetscErrorCode ierr; 2732 2733 PetscFunctionBegin; 2734 PetscValidHeaderSpecific(mat,MAT_CLASSID,1); 2735 if (row) PetscValidHeaderSpecific(row,IS_CLASSID,2); 2736 if (col) PetscValidHeaderSpecific(col,IS_CLASSID,3); 2737 if (info) PetscValidPointer(info,4); 2738 PetscValidType(mat,1); 2739 PetscValidPointer(fact,5); 2740 if (!mat->assembled) SETERRQ(((PetscObject)mat)->comm,PETSC_ERR_ARG_WRONGSTATE,"Not for unassembled matrix"); 2741 if (mat->factortype) SETERRQ(((PetscObject)mat)->comm,PETSC_ERR_ARG_WRONGSTATE,"Not for factored matrix"); 2742 if (!(fact)->ops->lufactorsymbolic) { 2743 const MatSolverPackage spackage; 2744 ierr = MatFactorGetSolverPackage(fact,&spackage);CHKERRQ(ierr); 2745 SETERRQ2(((PetscObject)mat)->comm,PETSC_ERR_SUP,"Matrix type %s symbolic LU using solver package %s",((PetscObject)mat)->type_name,spackage); 2746 } 2747 MatCheckPreallocated(mat,2); 2748 2749 ierr = PetscLogEventBegin(MAT_LUFactorSymbolic,mat,row,col,0);CHKERRQ(ierr); 2750 ierr = (fact->ops->lufactorsymbolic)(fact,mat,row,col,info);CHKERRQ(ierr); 2751 ierr = PetscLogEventEnd(MAT_LUFactorSymbolic,mat,row,col,0);CHKERRQ(ierr); 2752 ierr = PetscObjectStateIncrease((PetscObject)fact);CHKERRQ(ierr); 2753 PetscFunctionReturn(0); 2754 } 2755 2756 #undef __FUNCT__ 2757 #define __FUNCT__ "MatLUFactorNumeric" 2758 /*@C 2759 MatLUFactorNumeric - Performs numeric LU factorization of a matrix. 2760 Call this routine after first calling MatLUFactorSymbolic(). 2761 2762 Collective on Mat 2763 2764 Input Parameters: 2765 + fact - the factor matrix obtained with MatGetFactor() 2766 . mat - the matrix 2767 - info - options for factorization 2768 2769 Notes: 2770 See MatLUFactor() for in-place factorization. See 2771 MatCholeskyFactorNumeric() for the symmetric, positive definite case. 2772 2773 Most users should employ the simplified KSP interface for linear solvers 2774 instead of working directly with matrix algebra routines such as this. 2775 See, e.g., KSPCreate(). 2776 2777 Level: developer 2778 2779 Concepts: matrices^LU numeric factorization 2780 2781 .seealso: MatLUFactorSymbolic(), MatLUFactor(), MatCholeskyFactor() 2782 2783 Developer Note: fortran interface is not autogenerated as the f90 2784 interface defintion cannot be generated correctly [due to MatFactorInfo] 2785 2786 @*/ 2787 PetscErrorCode MatLUFactorNumeric(Mat fact,Mat mat,const MatFactorInfo *info) 2788 { 2789 PetscErrorCode ierr; 2790 2791 PetscFunctionBegin; 2792 PetscValidHeaderSpecific(mat,MAT_CLASSID,1); 2793 PetscValidType(mat,1); 2794 PetscValidPointer(fact,2); 2795 PetscValidHeaderSpecific(fact,MAT_CLASSID,2); 2796 if (!mat->assembled) SETERRQ(((PetscObject)mat)->comm,PETSC_ERR_ARG_WRONGSTATE,"Not for unassembled matrix"); 2797 if (mat->rmap->N != (fact)->rmap->N || mat->cmap->N != (fact)->cmap->N) { 2798 SETERRQ4(((PetscObject)mat)->comm,PETSC_ERR_ARG_SIZ,"Mat mat,Mat fact: global dimensions are different %D should = %D %D should = %D",mat->rmap->N,(fact)->rmap->N,mat->cmap->N,(fact)->cmap->N); 2799 } 2800 if (!(fact)->ops->lufactornumeric) SETERRQ1(((PetscObject)mat)->comm,PETSC_ERR_SUP,"Mat type %s numeric LU",((PetscObject)mat)->type_name); 2801 MatCheckPreallocated(mat,2); 2802 ierr = PetscLogEventBegin(MAT_LUFactorNumeric,mat,fact,0,0);CHKERRQ(ierr); 2803 ierr = (fact->ops->lufactornumeric)(fact,mat,info);CHKERRQ(ierr); 2804 ierr = PetscLogEventEnd(MAT_LUFactorNumeric,mat,fact,0,0);CHKERRQ(ierr); 2805 2806 ierr = MatView_Private(fact);CHKERRQ(ierr); 2807 ierr = PetscObjectStateIncrease((PetscObject)fact);CHKERRQ(ierr); 2808 PetscFunctionReturn(0); 2809 } 2810 2811 #undef __FUNCT__ 2812 #define __FUNCT__ "MatCholeskyFactor" 2813 /*@C 2814 MatCholeskyFactor - Performs in-place Cholesky factorization of a 2815 symmetric matrix. 2816 2817 Collective on Mat 2818 2819 Input Parameters: 2820 + mat - the matrix 2821 . perm - row and column permutations 2822 - f - expected fill as ratio of original fill 2823 2824 Notes: 2825 See MatLUFactor() for the nonsymmetric case. See also 2826 MatCholeskyFactorSymbolic(), and MatCholeskyFactorNumeric(). 2827 2828 Most users should employ the simplified KSP interface for linear solvers 2829 instead of working directly with matrix algebra routines such as this. 2830 See, e.g., KSPCreate(). 2831 2832 Level: developer 2833 2834 Concepts: matrices^Cholesky factorization 2835 2836 .seealso: MatLUFactor(), MatCholeskyFactorSymbolic(), MatCholeskyFactorNumeric() 2837 MatGetOrdering() 2838 2839 Developer Note: fortran interface is not autogenerated as the f90 2840 interface defintion cannot be generated correctly [due to MatFactorInfo] 2841 2842 @*/ 2843 PetscErrorCode MatCholeskyFactor(Mat mat,IS perm,const MatFactorInfo *info) 2844 { 2845 PetscErrorCode ierr; 2846 2847 PetscFunctionBegin; 2848 PetscValidHeaderSpecific(mat,MAT_CLASSID,1); 2849 PetscValidType(mat,1); 2850 if (perm) PetscValidHeaderSpecific(perm,IS_CLASSID,2); 2851 if (info) PetscValidPointer(info,3); 2852 if (mat->rmap->N != mat->cmap->N) SETERRQ(((PetscObject)mat)->comm,PETSC_ERR_ARG_WRONG,"Matrix must be square"); 2853 if (!mat->assembled) SETERRQ(((PetscObject)mat)->comm,PETSC_ERR_ARG_WRONGSTATE,"Not for unassembled matrix"); 2854 if (mat->factortype) SETERRQ(((PetscObject)mat)->comm,PETSC_ERR_ARG_WRONGSTATE,"Not for factored matrix"); 2855 if (!mat->ops->choleskyfactor) SETERRQ1(((PetscObject)mat)->comm,PETSC_ERR_SUP,"Mat type %s",((PetscObject)mat)->type_name); 2856 MatCheckPreallocated(mat,1); 2857 2858 ierr = PetscLogEventBegin(MAT_CholeskyFactor,mat,perm,0,0);CHKERRQ(ierr); 2859 ierr = (*mat->ops->choleskyfactor)(mat,perm,info);CHKERRQ(ierr); 2860 ierr = PetscLogEventEnd(MAT_CholeskyFactor,mat,perm,0,0);CHKERRQ(ierr); 2861 ierr = PetscObjectStateIncrease((PetscObject)mat);CHKERRQ(ierr); 2862 PetscFunctionReturn(0); 2863 } 2864 2865 #undef __FUNCT__ 2866 #define __FUNCT__ "MatCholeskyFactorSymbolic" 2867 /*@C 2868 MatCholeskyFactorSymbolic - Performs symbolic Cholesky factorization 2869 of a symmetric matrix. 2870 2871 Collective on Mat 2872 2873 Input Parameters: 2874 + fact - the factor matrix obtained with MatGetFactor() 2875 . mat - the matrix 2876 . perm - row and column permutations 2877 - info - options for factorization, includes 2878 $ fill - expected fill as ratio of original fill. 2879 $ dtcol - pivot tolerance (0 no pivot, 1 full column pivoting) 2880 $ Run with the option -info to determine an optimal value to use 2881 2882 Notes: 2883 See MatLUFactorSymbolic() for the nonsymmetric case. See also 2884 MatCholeskyFactor() and MatCholeskyFactorNumeric(). 2885 2886 Most users should employ the simplified KSP interface for linear solvers 2887 instead of working directly with matrix algebra routines such as this. 2888 See, e.g., KSPCreate(). 2889 2890 Level: developer 2891 2892 Concepts: matrices^Cholesky symbolic factorization 2893 2894 .seealso: MatLUFactorSymbolic(), MatCholeskyFactor(), MatCholeskyFactorNumeric() 2895 MatGetOrdering() 2896 2897 Developer Note: fortran interface is not autogenerated as the f90 2898 interface defintion cannot be generated correctly [due to MatFactorInfo] 2899 2900 @*/ 2901 PetscErrorCode MatCholeskyFactorSymbolic(Mat fact,Mat mat,IS perm,const MatFactorInfo *info) 2902 { 2903 PetscErrorCode ierr; 2904 2905 PetscFunctionBegin; 2906 PetscValidHeaderSpecific(mat,MAT_CLASSID,1); 2907 PetscValidType(mat,1); 2908 if (perm) PetscValidHeaderSpecific(perm,IS_CLASSID,2); 2909 if (info) PetscValidPointer(info,3); 2910 PetscValidPointer(fact,4); 2911 if (mat->rmap->N != mat->cmap->N) SETERRQ(((PetscObject)mat)->comm,PETSC_ERR_ARG_WRONG,"Matrix must be square"); 2912 if (!mat->assembled) SETERRQ(((PetscObject)mat)->comm,PETSC_ERR_ARG_WRONGSTATE,"Not for unassembled matrix"); 2913 if (mat->factortype) SETERRQ(((PetscObject)mat)->comm,PETSC_ERR_ARG_WRONGSTATE,"Not for factored matrix"); 2914 if (!(fact)->ops->choleskyfactorsymbolic) { 2915 const MatSolverPackage spackage; 2916 ierr = MatFactorGetSolverPackage(fact,&spackage);CHKERRQ(ierr); 2917 SETERRQ2(((PetscObject)mat)->comm,PETSC_ERR_SUP,"Mat type %s symbolic factor Cholesky using solver package %s",((PetscObject)mat)->type_name,spackage); 2918 } 2919 MatCheckPreallocated(mat,2); 2920 2921 ierr = PetscLogEventBegin(MAT_CholeskyFactorSymbolic,mat,perm,0,0);CHKERRQ(ierr); 2922 ierr = (fact->ops->choleskyfactorsymbolic)(fact,mat,perm,info);CHKERRQ(ierr); 2923 ierr = PetscLogEventEnd(MAT_CholeskyFactorSymbolic,mat,perm,0,0);CHKERRQ(ierr); 2924 ierr = PetscObjectStateIncrease((PetscObject)fact);CHKERRQ(ierr); 2925 PetscFunctionReturn(0); 2926 } 2927 2928 #undef __FUNCT__ 2929 #define __FUNCT__ "MatCholeskyFactorNumeric" 2930 /*@C 2931 MatCholeskyFactorNumeric - Performs numeric Cholesky factorization 2932 of a symmetric matrix. Call this routine after first calling 2933 MatCholeskyFactorSymbolic(). 2934 2935 Collective on Mat 2936 2937 Input Parameters: 2938 + fact - the factor matrix obtained with MatGetFactor() 2939 . mat - the initial matrix 2940 . info - options for factorization 2941 - fact - the symbolic factor of mat 2942 2943 2944 Notes: 2945 Most users should employ the simplified KSP interface for linear solvers 2946 instead of working directly with matrix algebra routines such as this. 2947 See, e.g., KSPCreate(). 2948 2949 Level: developer 2950 2951 Concepts: matrices^Cholesky numeric factorization 2952 2953 .seealso: MatCholeskyFactorSymbolic(), MatCholeskyFactor(), MatLUFactorNumeric() 2954 2955 Developer Note: fortran interface is not autogenerated as the f90 2956 interface defintion cannot be generated correctly [due to MatFactorInfo] 2957 2958 @*/ 2959 PetscErrorCode MatCholeskyFactorNumeric(Mat fact,Mat mat,const MatFactorInfo *info) 2960 { 2961 PetscErrorCode ierr; 2962 2963 PetscFunctionBegin; 2964 PetscValidHeaderSpecific(mat,MAT_CLASSID,1); 2965 PetscValidType(mat,1); 2966 PetscValidPointer(fact,2); 2967 PetscValidHeaderSpecific(fact,MAT_CLASSID,2); 2968 if (!mat->assembled) SETERRQ(((PetscObject)mat)->comm,PETSC_ERR_ARG_WRONGSTATE,"Not for unassembled matrix"); 2969 if (!(fact)->ops->choleskyfactornumeric) SETERRQ1(((PetscObject)mat)->comm,PETSC_ERR_SUP,"Mat type %s numeric factor Cholesky",((PetscObject)mat)->type_name); 2970 if (mat->rmap->N != (fact)->rmap->N || mat->cmap->N != (fact)->cmap->N) { 2971 SETERRQ4(((PetscObject)mat)->comm,PETSC_ERR_ARG_SIZ,"Mat mat,Mat fact: global dim %D should = %D %D should = %D",mat->rmap->N,(fact)->rmap->N,mat->cmap->N,(fact)->cmap->N); 2972 } 2973 MatCheckPreallocated(mat,2); 2974 2975 ierr = PetscLogEventBegin(MAT_CholeskyFactorNumeric,mat,fact,0,0);CHKERRQ(ierr); 2976 ierr = (fact->ops->choleskyfactornumeric)(fact,mat,info);CHKERRQ(ierr); 2977 ierr = PetscLogEventEnd(MAT_CholeskyFactorNumeric,mat,fact,0,0);CHKERRQ(ierr); 2978 2979 ierr = MatView_Private(fact);CHKERRQ(ierr); 2980 ierr = PetscObjectStateIncrease((PetscObject)fact);CHKERRQ(ierr); 2981 PetscFunctionReturn(0); 2982 } 2983 2984 /* ----------------------------------------------------------------*/ 2985 #undef __FUNCT__ 2986 #define __FUNCT__ "MatSolve" 2987 /*@ 2988 MatSolve - Solves A x = b, given a factored matrix. 2989 2990 Neighbor-wise Collective on Mat and Vec 2991 2992 Input Parameters: 2993 + mat - the factored matrix 2994 - b - the right-hand-side vector 2995 2996 Output Parameter: 2997 . x - the result vector 2998 2999 Notes: 3000 The vectors b and x cannot be the same. I.e., one cannot 3001 call MatSolve(A,x,x). 3002 3003 Notes: 3004 Most users should employ the simplified KSP interface for linear solvers 3005 instead of working directly with matrix algebra routines such as this. 3006 See, e.g., KSPCreate(). 3007 3008 Level: developer 3009 3010 Concepts: matrices^triangular solves 3011 3012 .seealso: MatSolveAdd(), MatSolveTranspose(), MatSolveTransposeAdd() 3013 @*/ 3014 PetscErrorCode MatSolve(Mat mat,Vec b,Vec x) 3015 { 3016 PetscErrorCode ierr; 3017 3018 PetscFunctionBegin; 3019 PetscValidHeaderSpecific(mat,MAT_CLASSID,1); 3020 PetscValidType(mat,1); 3021 PetscValidHeaderSpecific(b,VEC_CLASSID,2); 3022 PetscValidHeaderSpecific(x,VEC_CLASSID,3); 3023 PetscCheckSameComm(mat,1,b,2); 3024 PetscCheckSameComm(mat,1,x,3); 3025 if (x == b) SETERRQ(((PetscObject)mat)->comm,PETSC_ERR_ARG_IDN,"x and b must be different vectors"); 3026 if (!mat->factortype) SETERRQ(((PetscObject)mat)->comm,PETSC_ERR_ARG_WRONGSTATE,"Unfactored matrix"); 3027 if (mat->cmap->N != x->map->N) SETERRQ2(((PetscObject)mat)->comm,PETSC_ERR_ARG_SIZ,"Mat mat,Vec x: global dim %D %D",mat->cmap->N,x->map->N); 3028 if (mat->rmap->N != b->map->N) SETERRQ2(((PetscObject)mat)->comm,PETSC_ERR_ARG_SIZ,"Mat mat,Vec b: global dim %D %D",mat->rmap->N,b->map->N); 3029 if (mat->rmap->n != b->map->n) SETERRQ2(PETSC_COMM_SELF,PETSC_ERR_ARG_SIZ,"Mat mat,Vec b: local dim %D %D",mat->rmap->n,b->map->n); 3030 if (!mat->rmap->N && !mat->cmap->N) PetscFunctionReturn(0); 3031 if (!mat->ops->solve) SETERRQ1(((PetscObject)mat)->comm,PETSC_ERR_SUP,"Mat type %s",((PetscObject)mat)->type_name); 3032 MatCheckPreallocated(mat,1); 3033 3034 ierr = PetscLogEventBegin(MAT_Solve,mat,b,x,0);CHKERRQ(ierr); 3035 ierr = (*mat->ops->solve)(mat,b,x);CHKERRQ(ierr); 3036 ierr = PetscLogEventEnd(MAT_Solve,mat,b,x,0);CHKERRQ(ierr); 3037 ierr = PetscObjectStateIncrease((PetscObject)x);CHKERRQ(ierr); 3038 PetscFunctionReturn(0); 3039 } 3040 3041 #undef __FUNCT__ 3042 #define __FUNCT__ "MatMatSolve_Basic" 3043 PetscErrorCode MatMatSolve_Basic(Mat A,Mat B,Mat X) 3044 { 3045 PetscErrorCode ierr; 3046 Vec b,x; 3047 PetscInt m,N,i; 3048 PetscScalar *bb,*xx; 3049 PetscBool flg; 3050 3051 PetscFunctionBegin; 3052 ierr = PetscObjectTypeCompareAny((PetscObject)B,&flg,MATSEQDENSE,MATMPIDENSE,PETSC_NULL);CHKERRQ(ierr); 3053 if (!flg) SETERRQ(((PetscObject)A)->comm,PETSC_ERR_ARG_WRONG,"Matrix B must be MATDENSE matrix"); 3054 ierr = PetscObjectTypeCompareAny((PetscObject)X,&flg,MATSEQDENSE,MATMPIDENSE,PETSC_NULL);CHKERRQ(ierr); 3055 if (!flg) SETERRQ(((PetscObject)A)->comm,PETSC_ERR_ARG_WRONG,"Matrix X must be MATDENSE matrix"); 3056 3057 ierr = MatGetArray(B,&bb);CHKERRQ(ierr); 3058 ierr = MatGetArray(X,&xx);CHKERRQ(ierr); 3059 ierr = MatGetLocalSize(B,&m,PETSC_NULL);CHKERRQ(ierr); /* number local rows */ 3060 ierr = MatGetSize(B,PETSC_NULL,&N);CHKERRQ(ierr); /* total columns in dense matrix */ 3061 ierr = MatGetVecs(A,&x,&b);CHKERRQ(ierr); 3062 for (i=0; i<N; i++) { 3063 ierr = VecPlaceArray(b,bb + i*m);CHKERRQ(ierr); 3064 ierr = VecPlaceArray(x,xx + i*m);CHKERRQ(ierr); 3065 ierr = MatSolve(A,b,x);CHKERRQ(ierr); 3066 ierr = VecResetArray(x);CHKERRQ(ierr); 3067 ierr = VecResetArray(b);CHKERRQ(ierr); 3068 } 3069 ierr = VecDestroy(&b);CHKERRQ(ierr); 3070 ierr = VecDestroy(&x);CHKERRQ(ierr); 3071 ierr = MatRestoreArray(B,&bb);CHKERRQ(ierr); 3072 ierr = MatRestoreArray(X,&xx);CHKERRQ(ierr); 3073 PetscFunctionReturn(0); 3074 } 3075 3076 #undef __FUNCT__ 3077 #define __FUNCT__ "MatMatSolve" 3078 /*@ 3079 MatMatSolve - Solves A X = B, given a factored matrix. 3080 3081 Neighbor-wise Collective on Mat 3082 3083 Input Parameters: 3084 + mat - the factored matrix 3085 - B - the right-hand-side matrix (dense matrix) 3086 3087 Output Parameter: 3088 . X - the result matrix (dense matrix) 3089 3090 Notes: 3091 The matrices b and x cannot be the same. I.e., one cannot 3092 call MatMatSolve(A,x,x). 3093 3094 Notes: 3095 Most users should usually employ the simplified KSP interface for linear solvers 3096 instead of working directly with matrix algebra routines such as this. 3097 See, e.g., KSPCreate(). However KSP can only solve for one vector (column of X) 3098 at a time. 3099 3100 When using SuperLU_Dist as a parallel solver PETSc will use the SuperLU_Dist functionality to solve multiple right hand sides simultaneously. For MUMPS 3101 it calls a separate solve for each right hand side since MUMPS does not yet support distributed right hand sides. 3102 3103 Since the resulting matrix X must always be dense we do not support sparse representation of the matrix B. 3104 3105 Level: developer 3106 3107 Concepts: matrices^triangular solves 3108 3109 .seealso: MatMatSolveAdd(), MatMatSolveTranspose(), MatMatSolveTransposeAdd(), MatLUFactor(), MatCholeskyFactor() 3110 @*/ 3111 PetscErrorCode MatMatSolve(Mat A,Mat B,Mat X) 3112 { 3113 PetscErrorCode ierr; 3114 3115 PetscFunctionBegin; 3116 PetscValidHeaderSpecific(A,MAT_CLASSID,1); 3117 PetscValidType(A,1); 3118 PetscValidHeaderSpecific(B,MAT_CLASSID,2); 3119 PetscValidHeaderSpecific(X,MAT_CLASSID,3); 3120 PetscCheckSameComm(A,1,B,2); 3121 PetscCheckSameComm(A,1,X,3); 3122 if (X == B) SETERRQ(((PetscObject)A)->comm,PETSC_ERR_ARG_IDN,"X and B must be different matrices"); 3123 if (!A->factortype) SETERRQ(((PetscObject)A)->comm,PETSC_ERR_ARG_WRONGSTATE,"Unfactored matrix"); 3124 if (A->cmap->N != X->rmap->N) SETERRQ2(((PetscObject)A)->comm,PETSC_ERR_ARG_SIZ,"Mat A,Mat X: global dim %D %D",A->cmap->N,X->rmap->N); 3125 if (A->rmap->N != B->rmap->N) SETERRQ2(((PetscObject)A)->comm,PETSC_ERR_ARG_SIZ,"Mat A,Mat B: global dim %D %D",A->rmap->N,B->rmap->N); 3126 if (A->rmap->n != B->rmap->n) SETERRQ2(PETSC_COMM_SELF,PETSC_ERR_ARG_SIZ,"Mat A,Mat B: local dim %D %D",A->rmap->n,B->rmap->n); 3127 if (!A->rmap->N && !A->cmap->N) PetscFunctionReturn(0); 3128 MatCheckPreallocated(A,1); 3129 3130 ierr = PetscLogEventBegin(MAT_MatSolve,A,B,X,0);CHKERRQ(ierr); 3131 if (!A->ops->matsolve) { 3132 ierr = PetscInfo1(A,"Mat type %s using basic MatMatSolve",((PetscObject)A)->type_name);CHKERRQ(ierr); 3133 ierr = MatMatSolve_Basic(A,B,X);CHKERRQ(ierr); 3134 } else { 3135 ierr = (*A->ops->matsolve)(A,B,X);CHKERRQ(ierr); 3136 } 3137 ierr = PetscLogEventEnd(MAT_MatSolve,A,B,X,0);CHKERRQ(ierr); 3138 ierr = PetscObjectStateIncrease((PetscObject)X);CHKERRQ(ierr); 3139 PetscFunctionReturn(0); 3140 } 3141 3142 3143 #undef __FUNCT__ 3144 #define __FUNCT__ "MatForwardSolve" 3145 /*@ 3146 MatForwardSolve - Solves L x = b, given a factored matrix, A = LU, or 3147 U^T*D^(1/2) x = b, given a factored symmetric matrix, A = U^T*D*U, 3148 3149 Neighbor-wise Collective on Mat and Vec 3150 3151 Input Parameters: 3152 + mat - the factored matrix 3153 - b - the right-hand-side vector 3154 3155 Output Parameter: 3156 . x - the result vector 3157 3158 Notes: 3159 MatSolve() should be used for most applications, as it performs 3160 a forward solve followed by a backward solve. 3161 3162 The vectors b and x cannot be the same, i.e., one cannot 3163 call MatForwardSolve(A,x,x). 3164 3165 For matrix in seqsbaij format with block size larger than 1, 3166 the diagonal blocks are not implemented as D = D^(1/2) * D^(1/2) yet. 3167 MatForwardSolve() solves U^T*D y = b, and 3168 MatBackwardSolve() solves U x = y. 3169 Thus they do not provide a symmetric preconditioner. 3170 3171 Most users should employ the simplified KSP interface for linear solvers 3172 instead of working directly with matrix algebra routines such as this. 3173 See, e.g., KSPCreate(). 3174 3175 Level: developer 3176 3177 Concepts: matrices^forward solves 3178 3179 .seealso: MatSolve(), MatBackwardSolve() 3180 @*/ 3181 PetscErrorCode MatForwardSolve(Mat mat,Vec b,Vec x) 3182 { 3183 PetscErrorCode ierr; 3184 3185 PetscFunctionBegin; 3186 PetscValidHeaderSpecific(mat,MAT_CLASSID,1); 3187 PetscValidType(mat,1); 3188 PetscValidHeaderSpecific(b,VEC_CLASSID,2); 3189 PetscValidHeaderSpecific(x,VEC_CLASSID,3); 3190 PetscCheckSameComm(mat,1,b,2); 3191 PetscCheckSameComm(mat,1,x,3); 3192 if (x == b) SETERRQ(((PetscObject)mat)->comm,PETSC_ERR_ARG_IDN,"x and b must be different vectors"); 3193 if (!mat->factortype) SETERRQ(((PetscObject)mat)->comm,PETSC_ERR_ARG_WRONGSTATE,"Unfactored matrix"); 3194 if (!mat->ops->forwardsolve) SETERRQ1(((PetscObject)mat)->comm,PETSC_ERR_SUP,"Mat type %s",((PetscObject)mat)->type_name); 3195 if (mat->cmap->N != x->map->N) SETERRQ2(((PetscObject)mat)->comm,PETSC_ERR_ARG_SIZ,"Mat mat,Vec x: global dim %D %D",mat->cmap->N,x->map->N); 3196 if (mat->rmap->N != b->map->N) SETERRQ2(((PetscObject)mat)->comm,PETSC_ERR_ARG_SIZ,"Mat mat,Vec b: global dim %D %D",mat->rmap->N,b->map->N); 3197 if (mat->rmap->n != b->map->n) SETERRQ2(PETSC_COMM_SELF,PETSC_ERR_ARG_SIZ,"Mat mat,Vec b: local dim %D %D",mat->rmap->n,b->map->n); 3198 MatCheckPreallocated(mat,1); 3199 ierr = PetscLogEventBegin(MAT_ForwardSolve,mat,b,x,0);CHKERRQ(ierr); 3200 ierr = (*mat->ops->forwardsolve)(mat,b,x);CHKERRQ(ierr); 3201 ierr = PetscLogEventEnd(MAT_ForwardSolve,mat,b,x,0);CHKERRQ(ierr); 3202 ierr = PetscObjectStateIncrease((PetscObject)x);CHKERRQ(ierr); 3203 PetscFunctionReturn(0); 3204 } 3205 3206 #undef __FUNCT__ 3207 #define __FUNCT__ "MatBackwardSolve" 3208 /*@ 3209 MatBackwardSolve - Solves U x = b, given a factored matrix, A = LU. 3210 D^(1/2) U x = b, given a factored symmetric matrix, A = U^T*D*U, 3211 3212 Neighbor-wise Collective on Mat and Vec 3213 3214 Input Parameters: 3215 + mat - the factored matrix 3216 - b - the right-hand-side vector 3217 3218 Output Parameter: 3219 . x - the result vector 3220 3221 Notes: 3222 MatSolve() should be used for most applications, as it performs 3223 a forward solve followed by a backward solve. 3224 3225 The vectors b and x cannot be the same. I.e., one cannot 3226 call MatBackwardSolve(A,x,x). 3227 3228 For matrix in seqsbaij format with block size larger than 1, 3229 the diagonal blocks are not implemented as D = D^(1/2) * D^(1/2) yet. 3230 MatForwardSolve() solves U^T*D y = b, and 3231 MatBackwardSolve() solves U x = y. 3232 Thus they do not provide a symmetric preconditioner. 3233 3234 Most users should employ the simplified KSP interface for linear solvers 3235 instead of working directly with matrix algebra routines such as this. 3236 See, e.g., KSPCreate(). 3237 3238 Level: developer 3239 3240 Concepts: matrices^backward solves 3241 3242 .seealso: MatSolve(), MatForwardSolve() 3243 @*/ 3244 PetscErrorCode MatBackwardSolve(Mat mat,Vec b,Vec x) 3245 { 3246 PetscErrorCode ierr; 3247 3248 PetscFunctionBegin; 3249 PetscValidHeaderSpecific(mat,MAT_CLASSID,1); 3250 PetscValidType(mat,1); 3251 PetscValidHeaderSpecific(b,VEC_CLASSID,2); 3252 PetscValidHeaderSpecific(x,VEC_CLASSID,3); 3253 PetscCheckSameComm(mat,1,b,2); 3254 PetscCheckSameComm(mat,1,x,3); 3255 if (x == b) SETERRQ(((PetscObject)mat)->comm,PETSC_ERR_ARG_IDN,"x and b must be different vectors"); 3256 if (!mat->factortype) SETERRQ(((PetscObject)mat)->comm,PETSC_ERR_ARG_WRONGSTATE,"Unfactored matrix"); 3257 if (!mat->ops->backwardsolve) SETERRQ1(((PetscObject)mat)->comm,PETSC_ERR_SUP,"Mat type %s",((PetscObject)mat)->type_name); 3258 if (mat->cmap->N != x->map->N) SETERRQ2(((PetscObject)mat)->comm,PETSC_ERR_ARG_SIZ,"Mat mat,Vec x: global dim %D %D",mat->cmap->N,x->map->N); 3259 if (mat->rmap->N != b->map->N) SETERRQ2(((PetscObject)mat)->comm,PETSC_ERR_ARG_SIZ,"Mat mat,Vec b: global dim %D %D",mat->rmap->N,b->map->N); 3260 if (mat->rmap->n != b->map->n) SETERRQ2(PETSC_COMM_SELF,PETSC_ERR_ARG_SIZ,"Mat mat,Vec b: local dim %D %D",mat->rmap->n,b->map->n); 3261 MatCheckPreallocated(mat,1); 3262 3263 ierr = PetscLogEventBegin(MAT_BackwardSolve,mat,b,x,0);CHKERRQ(ierr); 3264 ierr = (*mat->ops->backwardsolve)(mat,b,x);CHKERRQ(ierr); 3265 ierr = PetscLogEventEnd(MAT_BackwardSolve,mat,b,x,0);CHKERRQ(ierr); 3266 ierr = PetscObjectStateIncrease((PetscObject)x);CHKERRQ(ierr); 3267 PetscFunctionReturn(0); 3268 } 3269 3270 #undef __FUNCT__ 3271 #define __FUNCT__ "MatSolveAdd" 3272 /*@ 3273 MatSolveAdd - Computes x = y + inv(A)*b, given a factored matrix. 3274 3275 Neighbor-wise Collective on Mat and Vec 3276 3277 Input Parameters: 3278 + mat - the factored matrix 3279 . b - the right-hand-side vector 3280 - y - the vector to be added to 3281 3282 Output Parameter: 3283 . x - the result vector 3284 3285 Notes: 3286 The vectors b and x cannot be the same. I.e., one cannot 3287 call MatSolveAdd(A,x,y,x). 3288 3289 Most users should employ the simplified KSP interface for linear solvers 3290 instead of working directly with matrix algebra routines such as this. 3291 See, e.g., KSPCreate(). 3292 3293 Level: developer 3294 3295 Concepts: matrices^triangular solves 3296 3297 .seealso: MatSolve(), MatSolveTranspose(), MatSolveTransposeAdd() 3298 @*/ 3299 PetscErrorCode MatSolveAdd(Mat mat,Vec b,Vec y,Vec x) 3300 { 3301 PetscScalar one = 1.0; 3302 Vec tmp; 3303 PetscErrorCode ierr; 3304 3305 PetscFunctionBegin; 3306 PetscValidHeaderSpecific(mat,MAT_CLASSID,1); 3307 PetscValidType(mat,1); 3308 PetscValidHeaderSpecific(y,VEC_CLASSID,2); 3309 PetscValidHeaderSpecific(b,VEC_CLASSID,3); 3310 PetscValidHeaderSpecific(x,VEC_CLASSID,4); 3311 PetscCheckSameComm(mat,1,b,2); 3312 PetscCheckSameComm(mat,1,y,2); 3313 PetscCheckSameComm(mat,1,x,3); 3314 if (x == b) SETERRQ(((PetscObject)mat)->comm,PETSC_ERR_ARG_IDN,"x and b must be different vectors"); 3315 if (!mat->factortype) SETERRQ(((PetscObject)mat)->comm,PETSC_ERR_ARG_WRONGSTATE,"Unfactored matrix"); 3316 if (mat->cmap->N != x->map->N) SETERRQ2(((PetscObject)mat)->comm,PETSC_ERR_ARG_SIZ,"Mat mat,Vec x: global dim %D %D",mat->cmap->N,x->map->N); 3317 if (mat->rmap->N != b->map->N) SETERRQ2(((PetscObject)mat)->comm,PETSC_ERR_ARG_SIZ,"Mat mat,Vec b: global dim %D %D",mat->rmap->N,b->map->N); 3318 if (mat->rmap->N != y->map->N) SETERRQ2(((PetscObject)mat)->comm,PETSC_ERR_ARG_SIZ,"Mat mat,Vec y: global dim %D %D",mat->rmap->N,y->map->N); 3319 if (mat->rmap->n != b->map->n) SETERRQ2(PETSC_COMM_SELF,PETSC_ERR_ARG_SIZ,"Mat mat,Vec b: local dim %D %D",mat->rmap->n,b->map->n); 3320 if (x->map->n != y->map->n) SETERRQ2(PETSC_COMM_SELF,PETSC_ERR_ARG_SIZ,"Vec x,Vec y: local dim %D %D",x->map->n,y->map->n); 3321 MatCheckPreallocated(mat,1); 3322 3323 ierr = PetscLogEventBegin(MAT_SolveAdd,mat,b,x,y);CHKERRQ(ierr); 3324 if (mat->ops->solveadd) { 3325 ierr = (*mat->ops->solveadd)(mat,b,y,x);CHKERRQ(ierr); 3326 } else { 3327 /* do the solve then the add manually */ 3328 if (x != y) { 3329 ierr = MatSolve(mat,b,x);CHKERRQ(ierr); 3330 ierr = VecAXPY(x,one,y);CHKERRQ(ierr); 3331 } else { 3332 ierr = VecDuplicate(x,&tmp);CHKERRQ(ierr); 3333 ierr = PetscLogObjectParent(mat,tmp);CHKERRQ(ierr); 3334 ierr = VecCopy(x,tmp);CHKERRQ(ierr); 3335 ierr = MatSolve(mat,b,x);CHKERRQ(ierr); 3336 ierr = VecAXPY(x,one,tmp);CHKERRQ(ierr); 3337 ierr = VecDestroy(&tmp);CHKERRQ(ierr); 3338 } 3339 } 3340 ierr = PetscLogEventEnd(MAT_SolveAdd,mat,b,x,y);CHKERRQ(ierr); 3341 ierr = PetscObjectStateIncrease((PetscObject)x);CHKERRQ(ierr); 3342 PetscFunctionReturn(0); 3343 } 3344 3345 #undef __FUNCT__ 3346 #define __FUNCT__ "MatSolveTranspose" 3347 /*@ 3348 MatSolveTranspose - Solves A' x = b, given a factored matrix. 3349 3350 Neighbor-wise Collective on Mat and Vec 3351 3352 Input Parameters: 3353 + mat - the factored matrix 3354 - b - the right-hand-side vector 3355 3356 Output Parameter: 3357 . x - the result vector 3358 3359 Notes: 3360 The vectors b and x cannot be the same. I.e., one cannot 3361 call MatSolveTranspose(A,x,x). 3362 3363 Most users should employ the simplified KSP interface for linear solvers 3364 instead of working directly with matrix algebra routines such as this. 3365 See, e.g., KSPCreate(). 3366 3367 Level: developer 3368 3369 Concepts: matrices^triangular solves 3370 3371 .seealso: MatSolve(), MatSolveAdd(), MatSolveTransposeAdd() 3372 @*/ 3373 PetscErrorCode MatSolveTranspose(Mat mat,Vec b,Vec x) 3374 { 3375 PetscErrorCode ierr; 3376 3377 PetscFunctionBegin; 3378 PetscValidHeaderSpecific(mat,MAT_CLASSID,1); 3379 PetscValidType(mat,1); 3380 PetscValidHeaderSpecific(b,VEC_CLASSID,2); 3381 PetscValidHeaderSpecific(x,VEC_CLASSID,3); 3382 PetscCheckSameComm(mat,1,b,2); 3383 PetscCheckSameComm(mat,1,x,3); 3384 if (!mat->factortype) SETERRQ(((PetscObject)mat)->comm,PETSC_ERR_ARG_WRONGSTATE,"Unfactored matrix"); 3385 if (x == b) SETERRQ(((PetscObject)mat)->comm,PETSC_ERR_ARG_IDN,"x and b must be different vectors"); 3386 if (!mat->ops->solvetranspose) SETERRQ1(((PetscObject)mat)->comm,PETSC_ERR_SUP,"Matrix type %s",((PetscObject)mat)->type_name); 3387 if (mat->rmap->N != x->map->N) SETERRQ2(((PetscObject)mat)->comm,PETSC_ERR_ARG_SIZ,"Mat mat,Vec x: global dim %D %D",mat->rmap->N,x->map->N); 3388 if (mat->cmap->N != b->map->N) SETERRQ2(((PetscObject)mat)->comm,PETSC_ERR_ARG_SIZ,"Mat mat,Vec b: global dim %D %D",mat->cmap->N,b->map->N); 3389 MatCheckPreallocated(mat,1); 3390 ierr = PetscLogEventBegin(MAT_SolveTranspose,mat,b,x,0);CHKERRQ(ierr); 3391 ierr = (*mat->ops->solvetranspose)(mat,b,x);CHKERRQ(ierr); 3392 ierr = PetscLogEventEnd(MAT_SolveTranspose,mat,b,x,0);CHKERRQ(ierr); 3393 ierr = PetscObjectStateIncrease((PetscObject)x);CHKERRQ(ierr); 3394 PetscFunctionReturn(0); 3395 } 3396 3397 #undef __FUNCT__ 3398 #define __FUNCT__ "MatSolveTransposeAdd" 3399 /*@ 3400 MatSolveTransposeAdd - Computes x = y + inv(Transpose(A)) b, given a 3401 factored matrix. 3402 3403 Neighbor-wise Collective on Mat and Vec 3404 3405 Input Parameters: 3406 + mat - the factored matrix 3407 . b - the right-hand-side vector 3408 - y - the vector to be added to 3409 3410 Output Parameter: 3411 . x - the result vector 3412 3413 Notes: 3414 The vectors b and x cannot be the same. I.e., one cannot 3415 call MatSolveTransposeAdd(A,x,y,x). 3416 3417 Most users should employ the simplified KSP interface for linear solvers 3418 instead of working directly with matrix algebra routines such as this. 3419 See, e.g., KSPCreate(). 3420 3421 Level: developer 3422 3423 Concepts: matrices^triangular solves 3424 3425 .seealso: MatSolve(), MatSolveAdd(), MatSolveTranspose() 3426 @*/ 3427 PetscErrorCode MatSolveTransposeAdd(Mat mat,Vec b,Vec y,Vec x) 3428 { 3429 PetscScalar one = 1.0; 3430 PetscErrorCode ierr; 3431 Vec tmp; 3432 3433 PetscFunctionBegin; 3434 PetscValidHeaderSpecific(mat,MAT_CLASSID,1); 3435 PetscValidType(mat,1); 3436 PetscValidHeaderSpecific(y,VEC_CLASSID,2); 3437 PetscValidHeaderSpecific(b,VEC_CLASSID,3); 3438 PetscValidHeaderSpecific(x,VEC_CLASSID,4); 3439 PetscCheckSameComm(mat,1,b,2); 3440 PetscCheckSameComm(mat,1,y,3); 3441 PetscCheckSameComm(mat,1,x,4); 3442 if (x == b) SETERRQ(((PetscObject)mat)->comm,PETSC_ERR_ARG_IDN,"x and b must be different vectors"); 3443 if (!mat->factortype) SETERRQ(((PetscObject)mat)->comm,PETSC_ERR_ARG_WRONGSTATE,"Unfactored matrix"); 3444 if (mat->rmap->N != x->map->N) SETERRQ2(((PetscObject)mat)->comm,PETSC_ERR_ARG_SIZ,"Mat mat,Vec x: global dim %D %D",mat->rmap->N,x->map->N); 3445 if (mat->cmap->N != b->map->N) SETERRQ2(((PetscObject)mat)->comm,PETSC_ERR_ARG_SIZ,"Mat mat,Vec b: global dim %D %D",mat->cmap->N,b->map->N); 3446 if (mat->cmap->N != y->map->N) SETERRQ2(((PetscObject)mat)->comm,PETSC_ERR_ARG_SIZ,"Mat mat,Vec y: global dim %D %D",mat->cmap->N,y->map->N); 3447 if (x->map->n != y->map->n) SETERRQ2(PETSC_COMM_SELF,PETSC_ERR_ARG_SIZ,"Vec x,Vec y: local dim %D %D",x->map->n,y->map->n); 3448 MatCheckPreallocated(mat,1); 3449 3450 ierr = PetscLogEventBegin(MAT_SolveTransposeAdd,mat,b,x,y);CHKERRQ(ierr); 3451 if (mat->ops->solvetransposeadd) { 3452 ierr = (*mat->ops->solvetransposeadd)(mat,b,y,x);CHKERRQ(ierr); 3453 } else { 3454 /* do the solve then the add manually */ 3455 if (x != y) { 3456 ierr = MatSolveTranspose(mat,b,x);CHKERRQ(ierr); 3457 ierr = VecAXPY(x,one,y);CHKERRQ(ierr); 3458 } else { 3459 ierr = VecDuplicate(x,&tmp);CHKERRQ(ierr); 3460 ierr = PetscLogObjectParent(mat,tmp);CHKERRQ(ierr); 3461 ierr = VecCopy(x,tmp);CHKERRQ(ierr); 3462 ierr = MatSolveTranspose(mat,b,x);CHKERRQ(ierr); 3463 ierr = VecAXPY(x,one,tmp);CHKERRQ(ierr); 3464 ierr = VecDestroy(&tmp);CHKERRQ(ierr); 3465 } 3466 } 3467 ierr = PetscLogEventEnd(MAT_SolveTransposeAdd,mat,b,x,y);CHKERRQ(ierr); 3468 ierr = PetscObjectStateIncrease((PetscObject)x);CHKERRQ(ierr); 3469 PetscFunctionReturn(0); 3470 } 3471 /* ----------------------------------------------------------------*/ 3472 3473 #undef __FUNCT__ 3474 #define __FUNCT__ "MatSOR" 3475 /*@ 3476 MatSOR - Computes relaxation (SOR, Gauss-Seidel) sweeps. 3477 3478 Neighbor-wise Collective on Mat and Vec 3479 3480 Input Parameters: 3481 + mat - the matrix 3482 . b - the right hand side 3483 . omega - the relaxation factor 3484 . flag - flag indicating the type of SOR (see below) 3485 . shift - diagonal shift 3486 . its - the number of iterations 3487 - lits - the number of local iterations 3488 3489 Output Parameters: 3490 . x - the solution (can contain an initial guess, use option SOR_ZERO_INITIAL_GUESS to indicate no guess) 3491 3492 SOR Flags: 3493 . SOR_FORWARD_SWEEP - forward SOR 3494 . SOR_BACKWARD_SWEEP - backward SOR 3495 . SOR_SYMMETRIC_SWEEP - SSOR (symmetric SOR) 3496 . SOR_LOCAL_FORWARD_SWEEP - local forward SOR 3497 . SOR_LOCAL_BACKWARD_SWEEP - local forward SOR 3498 . SOR_LOCAL_SYMMETRIC_SWEEP - local SSOR 3499 . SOR_APPLY_UPPER, SOR_APPLY_LOWER - applies 3500 upper/lower triangular part of matrix to 3501 vector (with omega) 3502 . SOR_ZERO_INITIAL_GUESS - zero initial guess 3503 3504 Notes: 3505 SOR_LOCAL_FORWARD_SWEEP, SOR_LOCAL_BACKWARD_SWEEP, and 3506 SOR_LOCAL_SYMMETRIC_SWEEP perform separate independent smoothings 3507 on each processor. 3508 3509 Application programmers will not generally use MatSOR() directly, 3510 but instead will employ the KSP/PC interface. 3511 3512 Notes: for BAIJ, SBAIJ, and AIJ matrices with Inodes this does a block SOR smoothing, otherwise it does a pointwise smoothing 3513 3514 Notes for Advanced Users: 3515 The flags are implemented as bitwise inclusive or operations. 3516 For example, use (SOR_ZERO_INITIAL_GUESS | SOR_SYMMETRIC_SWEEP) 3517 to specify a zero initial guess for SSOR. 3518 3519 Most users should employ the simplified KSP interface for linear solvers 3520 instead of working directly with matrix algebra routines such as this. 3521 See, e.g., KSPCreate(). 3522 3523 Vectors x and b CANNOT be the same 3524 3525 Developer Note: We should add block SOR support for AIJ matrices with block size set to great than one and no inodes 3526 3527 Level: developer 3528 3529 Concepts: matrices^relaxation 3530 Concepts: matrices^SOR 3531 Concepts: matrices^Gauss-Seidel 3532 3533 @*/ 3534 PetscErrorCode MatSOR(Mat mat,Vec b,PetscReal omega,MatSORType flag,PetscReal shift,PetscInt its,PetscInt lits,Vec x) 3535 { 3536 PetscErrorCode ierr; 3537 3538 PetscFunctionBegin; 3539 PetscValidHeaderSpecific(mat,MAT_CLASSID,1); 3540 PetscValidType(mat,1); 3541 PetscValidHeaderSpecific(b,VEC_CLASSID,2); 3542 PetscValidHeaderSpecific(x,VEC_CLASSID,8); 3543 PetscCheckSameComm(mat,1,b,2); 3544 PetscCheckSameComm(mat,1,x,8); 3545 if (!mat->ops->sor) SETERRQ1(((PetscObject)mat)->comm,PETSC_ERR_SUP,"Mat type %s",((PetscObject)mat)->type_name); 3546 if (!mat->assembled) SETERRQ(((PetscObject)mat)->comm,PETSC_ERR_ARG_WRONGSTATE,"Not for unassembled matrix"); 3547 if (mat->factortype) SETERRQ(((PetscObject)mat)->comm,PETSC_ERR_ARG_WRONGSTATE,"Not for factored matrix"); 3548 if (mat->cmap->N != x->map->N) SETERRQ2(((PetscObject)mat)->comm,PETSC_ERR_ARG_SIZ,"Mat mat,Vec x: global dim %D %D",mat->cmap->N,x->map->N); 3549 if (mat->rmap->N != b->map->N) SETERRQ2(((PetscObject)mat)->comm,PETSC_ERR_ARG_SIZ,"Mat mat,Vec b: global dim %D %D",mat->rmap->N,b->map->N); 3550 if (mat->rmap->n != b->map->n) SETERRQ2(PETSC_COMM_SELF,PETSC_ERR_ARG_SIZ,"Mat mat,Vec b: local dim %D %D",mat->rmap->n,b->map->n); 3551 if (its <= 0) SETERRQ1(PETSC_COMM_SELF,PETSC_ERR_ARG_WRONG,"Relaxation requires global its %D positive",its); 3552 if (lits <= 0) SETERRQ1(PETSC_COMM_SELF,PETSC_ERR_ARG_WRONG,"Relaxation requires local its %D positive",lits); 3553 if (b == x) SETERRQ(PETSC_COMM_SELF,PETSC_ERR_ARG_IDN,"b and x vector cannot be the same"); 3554 3555 MatCheckPreallocated(mat,1); 3556 ierr = PetscLogEventBegin(MAT_SOR,mat,b,x,0);CHKERRQ(ierr); 3557 ierr =(*mat->ops->sor)(mat,b,omega,flag,shift,its,lits,x);CHKERRQ(ierr); 3558 ierr = PetscLogEventEnd(MAT_SOR,mat,b,x,0);CHKERRQ(ierr); 3559 ierr = PetscObjectStateIncrease((PetscObject)x);CHKERRQ(ierr); 3560 PetscFunctionReturn(0); 3561 } 3562 3563 #undef __FUNCT__ 3564 #define __FUNCT__ "MatCopy_Basic" 3565 /* 3566 Default matrix copy routine. 3567 */ 3568 PetscErrorCode MatCopy_Basic(Mat A,Mat B,MatStructure str) 3569 { 3570 PetscErrorCode ierr; 3571 PetscInt i,rstart = 0,rend = 0,nz; 3572 const PetscInt *cwork; 3573 const PetscScalar *vwork; 3574 3575 PetscFunctionBegin; 3576 if (B->assembled) { 3577 ierr = MatZeroEntries(B);CHKERRQ(ierr); 3578 } 3579 ierr = MatGetOwnershipRange(A,&rstart,&rend);CHKERRQ(ierr); 3580 for (i=rstart; i<rend; i++) { 3581 ierr = MatGetRow(A,i,&nz,&cwork,&vwork);CHKERRQ(ierr); 3582 ierr = MatSetValues(B,1,&i,nz,cwork,vwork,INSERT_VALUES);CHKERRQ(ierr); 3583 ierr = MatRestoreRow(A,i,&nz,&cwork,&vwork);CHKERRQ(ierr); 3584 } 3585 ierr = MatAssemblyBegin(B,MAT_FINAL_ASSEMBLY);CHKERRQ(ierr); 3586 ierr = MatAssemblyEnd(B,MAT_FINAL_ASSEMBLY);CHKERRQ(ierr); 3587 ierr = PetscObjectStateIncrease((PetscObject)B);CHKERRQ(ierr); 3588 PetscFunctionReturn(0); 3589 } 3590 3591 #undef __FUNCT__ 3592 #define __FUNCT__ "MatCopy" 3593 /*@ 3594 MatCopy - Copys a matrix to another matrix. 3595 3596 Collective on Mat 3597 3598 Input Parameters: 3599 + A - the matrix 3600 - str - SAME_NONZERO_PATTERN or DIFFERENT_NONZERO_PATTERN 3601 3602 Output Parameter: 3603 . B - where the copy is put 3604 3605 Notes: 3606 If you use SAME_NONZERO_PATTERN then the two matrices had better have the 3607 same nonzero pattern or the routine will crash. 3608 3609 MatCopy() copies the matrix entries of a matrix to another existing 3610 matrix (after first zeroing the second matrix). A related routine is 3611 MatConvert(), which first creates a new matrix and then copies the data. 3612 3613 Level: intermediate 3614 3615 Concepts: matrices^copying 3616 3617 .seealso: MatConvert(), MatDuplicate() 3618 3619 @*/ 3620 PetscErrorCode MatCopy(Mat A,Mat B,MatStructure str) 3621 { 3622 PetscErrorCode ierr; 3623 PetscInt i; 3624 3625 PetscFunctionBegin; 3626 PetscValidHeaderSpecific(A,MAT_CLASSID,1); 3627 PetscValidHeaderSpecific(B,MAT_CLASSID,2); 3628 PetscValidType(A,1); 3629 PetscValidType(B,2); 3630 PetscCheckSameComm(A,1,B,2); 3631 MatCheckPreallocated(B,2); 3632 if (!A->assembled) SETERRQ(((PetscObject)A)->comm,PETSC_ERR_ARG_WRONGSTATE,"Not for unassembled matrix"); 3633 if (A->factortype) SETERRQ(((PetscObject)A)->comm,PETSC_ERR_ARG_WRONGSTATE,"Not for factored matrix"); 3634 if (A->rmap->N != B->rmap->N || A->cmap->N != B->cmap->N) SETERRQ4(((PetscObject)A)->comm,PETSC_ERR_ARG_SIZ,"Mat A,Mat B: global dim (%D,%D) (%D,%D)",A->rmap->N,B->rmap->N,A->cmap->N,B->cmap->N); 3635 MatCheckPreallocated(A,1); 3636 3637 ierr = PetscLogEventBegin(MAT_Copy,A,B,0,0);CHKERRQ(ierr); 3638 if (A->ops->copy) { 3639 ierr = (*A->ops->copy)(A,B,str);CHKERRQ(ierr); 3640 } else { /* generic conversion */ 3641 ierr = MatCopy_Basic(A,B,str);CHKERRQ(ierr); 3642 } 3643 3644 B->stencil.dim = A->stencil.dim; 3645 B->stencil.noc = A->stencil.noc; 3646 for (i=0; i<=A->stencil.dim; i++) { 3647 B->stencil.dims[i] = A->stencil.dims[i]; 3648 B->stencil.starts[i] = A->stencil.starts[i]; 3649 } 3650 3651 ierr = PetscLogEventEnd(MAT_Copy,A,B,0,0);CHKERRQ(ierr); 3652 ierr = PetscObjectStateIncrease((PetscObject)B);CHKERRQ(ierr); 3653 PetscFunctionReturn(0); 3654 } 3655 3656 #undef __FUNCT__ 3657 #define __FUNCT__ "MatConvert" 3658 /*@C 3659 MatConvert - Converts a matrix to another matrix, either of the same 3660 or different type. 3661 3662 Collective on Mat 3663 3664 Input Parameters: 3665 + mat - the matrix 3666 . newtype - new matrix type. Use MATSAME to create a new matrix of the 3667 same type as the original matrix. 3668 - reuse - denotes if the destination matrix is to be created or reused. Currently 3669 MAT_REUSE_MATRIX is only supported for inplace conversion, otherwise use 3670 MAT_INITIAL_MATRIX. 3671 3672 Output Parameter: 3673 . M - pointer to place new matrix 3674 3675 Notes: 3676 MatConvert() first creates a new matrix and then copies the data from 3677 the first matrix. A related routine is MatCopy(), which copies the matrix 3678 entries of one matrix to another already existing matrix context. 3679 3680 Cannot be used to convert a sequential matrix to parallel or parallel to sequential, 3681 the MPI communicator of the generated matrix is always the same as the communicator 3682 of the input matrix. 3683 3684 Level: intermediate 3685 3686 Concepts: matrices^converting between storage formats 3687 3688 .seealso: MatCopy(), MatDuplicate() 3689 @*/ 3690 PetscErrorCode MatConvert(Mat mat, const MatType newtype,MatReuse reuse,Mat *M) 3691 { 3692 PetscErrorCode ierr; 3693 PetscBool sametype,issame,flg; 3694 char convname[256],mtype[256]; 3695 Mat B; 3696 3697 PetscFunctionBegin; 3698 PetscValidHeaderSpecific(mat,MAT_CLASSID,1); 3699 PetscValidType(mat,1); 3700 PetscValidPointer(M,3); 3701 if (!mat->assembled) SETERRQ(((PetscObject)mat)->comm,PETSC_ERR_ARG_WRONGSTATE,"Not for unassembled matrix"); 3702 if (mat->factortype) SETERRQ(((PetscObject)mat)->comm,PETSC_ERR_ARG_WRONGSTATE,"Not for factored matrix"); 3703 MatCheckPreallocated(mat,1); 3704 ierr = MatSetOption(mat,MAT_NEW_NONZERO_LOCATION_ERR,PETSC_FALSE);CHKERRQ(ierr); 3705 3706 ierr = PetscOptionsGetString(((PetscObject)mat)->prefix,"-matconvert_type",mtype,256,&flg);CHKERRQ(ierr); 3707 if (flg) { 3708 newtype = mtype; 3709 } 3710 ierr = PetscObjectTypeCompare((PetscObject)mat,newtype,&sametype);CHKERRQ(ierr); 3711 ierr = PetscStrcmp(newtype,"same",&issame);CHKERRQ(ierr); 3712 if ((reuse == MAT_REUSE_MATRIX) && (mat != *M)) SETERRQ(((PetscObject)mat)->comm,PETSC_ERR_SUP,"MAT_REUSE_MATRIX only supported for in-place conversion currently"); 3713 3714 if ((reuse == MAT_REUSE_MATRIX) && (issame || sametype)) PetscFunctionReturn(0); 3715 3716 if ((sametype || issame) && (reuse==MAT_INITIAL_MATRIX) && mat->ops->duplicate) { 3717 ierr = (*mat->ops->duplicate)(mat,MAT_COPY_VALUES,M);CHKERRQ(ierr); 3718 } else { 3719 PetscErrorCode (*conv)(Mat, const MatType,MatReuse,Mat*)=PETSC_NULL; 3720 const char *prefix[3] = {"seq","mpi",""}; 3721 PetscInt i; 3722 /* 3723 Order of precedence: 3724 1) See if a specialized converter is known to the current matrix. 3725 2) See if a specialized converter is known to the desired matrix class. 3726 3) See if a good general converter is registered for the desired class 3727 (as of 6/27/03 only MATMPIADJ falls into this category). 3728 4) See if a good general converter is known for the current matrix. 3729 5) Use a really basic converter. 3730 */ 3731 3732 /* 1) See if a specialized converter is known to the current matrix and the desired class */ 3733 for (i=0; i<3; i++) { 3734 ierr = PetscStrcpy(convname,"MatConvert_");CHKERRQ(ierr); 3735 ierr = PetscStrcat(convname,((PetscObject)mat)->type_name);CHKERRQ(ierr); 3736 ierr = PetscStrcat(convname,"_");CHKERRQ(ierr); 3737 ierr = PetscStrcat(convname,prefix[i]);CHKERRQ(ierr); 3738 ierr = PetscStrcat(convname,issame?((PetscObject)mat)->type_name:newtype);CHKERRQ(ierr); 3739 ierr = PetscStrcat(convname,"_C");CHKERRQ(ierr); 3740 ierr = PetscObjectQueryFunction((PetscObject)mat,convname,(void (**)(void))&conv);CHKERRQ(ierr); 3741 if (conv) goto foundconv; 3742 } 3743 3744 /* 2) See if a specialized converter is known to the desired matrix class. */ 3745 ierr = MatCreate(((PetscObject)mat)->comm,&B);CHKERRQ(ierr); 3746 ierr = MatSetSizes(B,mat->rmap->n,mat->cmap->n,mat->rmap->N,mat->cmap->N);CHKERRQ(ierr); 3747 ierr = MatSetType(B,newtype);CHKERRQ(ierr); 3748 for (i=0; i<3; i++) { 3749 ierr = PetscStrcpy(convname,"MatConvert_");CHKERRQ(ierr); 3750 ierr = PetscStrcat(convname,((PetscObject)mat)->type_name);CHKERRQ(ierr); 3751 ierr = PetscStrcat(convname,"_");CHKERRQ(ierr); 3752 ierr = PetscStrcat(convname,prefix[i]);CHKERRQ(ierr); 3753 ierr = PetscStrcat(convname,newtype);CHKERRQ(ierr); 3754 ierr = PetscStrcat(convname,"_C");CHKERRQ(ierr); 3755 ierr = PetscObjectQueryFunction((PetscObject)B,convname,(void (**)(void))&conv);CHKERRQ(ierr); 3756 if (conv) { 3757 ierr = MatDestroy(&B);CHKERRQ(ierr); 3758 goto foundconv; 3759 } 3760 } 3761 3762 /* 3) See if a good general converter is registered for the desired class */ 3763 conv = B->ops->convertfrom; 3764 ierr = MatDestroy(&B);CHKERRQ(ierr); 3765 if (conv) goto foundconv; 3766 3767 /* 4) See if a good general converter is known for the current matrix */ 3768 if (mat->ops->convert) { 3769 conv = mat->ops->convert; 3770 } 3771 if (conv) goto foundconv; 3772 3773 /* 5) Use a really basic converter. */ 3774 conv = MatConvert_Basic; 3775 3776 foundconv: 3777 ierr = PetscLogEventBegin(MAT_Convert,mat,0,0,0);CHKERRQ(ierr); 3778 ierr = (*conv)(mat,newtype,reuse,M);CHKERRQ(ierr); 3779 ierr = PetscLogEventEnd(MAT_Convert,mat,0,0,0);CHKERRQ(ierr); 3780 } 3781 ierr = PetscObjectStateIncrease((PetscObject)*M);CHKERRQ(ierr); 3782 3783 /* Copy Mat options */ 3784 if (mat->symmetric){ierr = MatSetOption(*M,MAT_SYMMETRIC,PETSC_TRUE);CHKERRQ(ierr);} 3785 if (mat->hermitian){ierr = MatSetOption(*M,MAT_HERMITIAN,PETSC_TRUE);CHKERRQ(ierr);} 3786 PetscFunctionReturn(0); 3787 } 3788 3789 #undef __FUNCT__ 3790 #define __FUNCT__ "MatFactorGetSolverPackage" 3791 /*@C 3792 MatFactorGetSolverPackage - Returns name of the package providing the factorization routines 3793 3794 Not Collective 3795 3796 Input Parameter: 3797 . mat - the matrix, must be a factored matrix 3798 3799 Output Parameter: 3800 . type - the string name of the package (do not free this string) 3801 3802 Notes: 3803 In Fortran you pass in a empty string and the package name will be copied into it. 3804 (Make sure the string is long enough) 3805 3806 Level: intermediate 3807 3808 .seealso: MatCopy(), MatDuplicate(), MatGetFactorAvailable(), MatGetFactor() 3809 @*/ 3810 PetscErrorCode MatFactorGetSolverPackage(Mat mat, const MatSolverPackage *type) 3811 { 3812 PetscErrorCode ierr; 3813 PetscErrorCode (*conv)(Mat,const MatSolverPackage*); 3814 3815 PetscFunctionBegin; 3816 PetscValidHeaderSpecific(mat,MAT_CLASSID,1); 3817 PetscValidType(mat,1); 3818 if (!mat->factortype) SETERRQ(((PetscObject)mat)->comm,PETSC_ERR_ARG_WRONGSTATE,"Only for factored matrix"); 3819 ierr = PetscObjectQueryFunction((PetscObject)mat,"MatFactorGetSolverPackage_C",(void (**)(void))&conv);CHKERRQ(ierr); 3820 if (!conv) { 3821 *type = MATSOLVERPETSC; 3822 } else { 3823 ierr = (*conv)(mat,type);CHKERRQ(ierr); 3824 } 3825 PetscFunctionReturn(0); 3826 } 3827 3828 #undef __FUNCT__ 3829 #define __FUNCT__ "MatGetFactor" 3830 /*@C 3831 MatGetFactor - Returns a matrix suitable to calls to MatXXFactorSymbolic() 3832 3833 Collective on Mat 3834 3835 Input Parameters: 3836 + mat - the matrix 3837 . type - name of solver type, for example, spooles, superlu, plapack, petsc (to use PETSc's default) 3838 - ftype - factor type, MAT_FACTOR_LU, MAT_FACTOR_CHOLESKY, MAT_FACTOR_ICC, MAT_FACTOR_ILU, 3839 3840 Output Parameters: 3841 . f - the factor matrix used with MatXXFactorSymbolic() calls 3842 3843 Notes: 3844 Some PETSc matrix formats have alternative solvers available that are contained in alternative packages 3845 such as pastix, superlu, mumps, spooles etc. 3846 3847 PETSc must have been ./configure to use the external solver, using the option --download-package 3848 3849 Level: intermediate 3850 3851 .seealso: MatCopy(), MatDuplicate(), MatGetFactorAvailable() 3852 @*/ 3853 PetscErrorCode MatGetFactor(Mat mat, const MatSolverPackage type,MatFactorType ftype,Mat *f) 3854 { 3855 PetscErrorCode ierr,(*conv)(Mat,MatFactorType,Mat*); 3856 char convname[256]; 3857 3858 PetscFunctionBegin; 3859 PetscValidHeaderSpecific(mat,MAT_CLASSID,1); 3860 PetscValidType(mat,1); 3861 3862 if (mat->factortype) SETERRQ(((PetscObject)mat)->comm,PETSC_ERR_ARG_WRONGSTATE,"Not for factored matrix"); 3863 MatCheckPreallocated(mat,1); 3864 3865 ierr = PetscStrcpy(convname,"MatGetFactor_");CHKERRQ(ierr); 3866 ierr = PetscStrcat(convname,type);CHKERRQ(ierr); 3867 ierr = PetscStrcat(convname,"_C");CHKERRQ(ierr); 3868 ierr = PetscObjectQueryFunction((PetscObject)mat,convname,(void (**)(void))&conv);CHKERRQ(ierr); 3869 if (!conv) { 3870 PetscBool flag; 3871 MPI_Comm comm; 3872 3873 ierr = PetscObjectGetComm((PetscObject)mat,&comm);CHKERRQ(ierr); 3874 ierr = PetscStrcasecmp(MATSOLVERPETSC,type,&flag);CHKERRQ(ierr); 3875 if (flag) { 3876 SETERRQ2(comm,PETSC_ERR_SUP,"Matrix format %s does not have a built-in PETSc %s",((PetscObject)mat)->type_name,MatFactorTypes[ftype]); 3877 } else { 3878 SETERRQ4(comm,PETSC_ERR_SUP,"Matrix format %s does not have a solver package %s for %s. Perhaps you must ./configure with --download-%s",((PetscObject)mat)->type_name,type,MatFactorTypes[ftype],type); 3879 } 3880 } 3881 ierr = (*conv)(mat,ftype,f);CHKERRQ(ierr); 3882 PetscFunctionReturn(0); 3883 } 3884 3885 #undef __FUNCT__ 3886 #define __FUNCT__ "MatGetFactorAvailable" 3887 /*@C 3888 MatGetFactorAvailable - Returns a a flag if matrix supports particular package and factor type 3889 3890 Not Collective 3891 3892 Input Parameters: 3893 + mat - the matrix 3894 . type - name of solver type, for example, spooles, superlu, plapack, petsc (to use PETSc's default) 3895 - ftype - factor type, MAT_FACTOR_LU, MAT_FACTOR_CHOLESKY, MAT_FACTOR_ICC, MAT_FACTOR_ILU, 3896 3897 Output Parameter: 3898 . flg - PETSC_TRUE if the factorization is available 3899 3900 Notes: 3901 Some PETSc matrix formats have alternative solvers available that are contained in alternative packages 3902 such as pastix, superlu, mumps, spooles etc. 3903 3904 PETSc must have been ./configure to use the external solver, using the option --download-package 3905 3906 Level: intermediate 3907 3908 .seealso: MatCopy(), MatDuplicate(), MatGetFactor() 3909 @*/ 3910 PetscErrorCode MatGetFactorAvailable(Mat mat, const MatSolverPackage type,MatFactorType ftype,PetscBool *flg) 3911 { 3912 PetscErrorCode ierr; 3913 char convname[256]; 3914 PetscErrorCode (*conv)(Mat,MatFactorType,PetscBool *); 3915 3916 PetscFunctionBegin; 3917 PetscValidHeaderSpecific(mat,MAT_CLASSID,1); 3918 PetscValidType(mat,1); 3919 3920 if (mat->factortype) SETERRQ(((PetscObject)mat)->comm,PETSC_ERR_ARG_WRONGSTATE,"Not for factored matrix"); 3921 MatCheckPreallocated(mat,1); 3922 3923 ierr = PetscStrcpy(convname,"MatGetFactorAvailable_");CHKERRQ(ierr); 3924 ierr = PetscStrcat(convname,type);CHKERRQ(ierr); 3925 ierr = PetscStrcat(convname,"_C");CHKERRQ(ierr); 3926 ierr = PetscObjectQueryFunction((PetscObject)mat,convname,(void (**)(void))&conv);CHKERRQ(ierr); 3927 if (!conv) { 3928 *flg = PETSC_FALSE; 3929 } else { 3930 ierr = (*conv)(mat,ftype,flg);CHKERRQ(ierr); 3931 } 3932 PetscFunctionReturn(0); 3933 } 3934 3935 3936 #undef __FUNCT__ 3937 #define __FUNCT__ "MatDuplicate" 3938 /*@ 3939 MatDuplicate - Duplicates a matrix including the non-zero structure. 3940 3941 Collective on Mat 3942 3943 Input Parameters: 3944 + mat - the matrix 3945 - op - either MAT_DO_NOT_COPY_VALUES or MAT_COPY_VALUES, cause it to copy the numerical values in the matrix 3946 MAT_SHARE_NONZERO_PATTERN to share the nonzero patterns with the previous matrix and not copy them. 3947 3948 Output Parameter: 3949 . M - pointer to place new matrix 3950 3951 Level: intermediate 3952 3953 Concepts: matrices^duplicating 3954 3955 Notes: You cannot change the nonzero pattern for the parent or child matrix if you use MAT_SHARE_NONZERO_PATTERN. 3956 3957 .seealso: MatCopy(), MatConvert() 3958 @*/ 3959 PetscErrorCode MatDuplicate(Mat mat,MatDuplicateOption op,Mat *M) 3960 { 3961 PetscErrorCode ierr; 3962 Mat B; 3963 PetscInt i; 3964 3965 PetscFunctionBegin; 3966 PetscValidHeaderSpecific(mat,MAT_CLASSID,1); 3967 PetscValidType(mat,1); 3968 PetscValidPointer(M,3); 3969 if (!mat->assembled) SETERRQ(((PetscObject)mat)->comm,PETSC_ERR_ARG_WRONGSTATE,"Not for unassembled matrix"); 3970 if (mat->factortype) SETERRQ(((PetscObject)mat)->comm,PETSC_ERR_ARG_WRONGSTATE,"Not for factored matrix"); 3971 MatCheckPreallocated(mat,1); 3972 3973 *M = 0; 3974 if (!mat->ops->duplicate) SETERRQ(((PetscObject)mat)->comm,PETSC_ERR_SUP,"Not written for this matrix type"); 3975 ierr = PetscLogEventBegin(MAT_Convert,mat,0,0,0);CHKERRQ(ierr); 3976 ierr = (*mat->ops->duplicate)(mat,op,M);CHKERRQ(ierr); 3977 B = *M; 3978 3979 B->stencil.dim = mat->stencil.dim; 3980 B->stencil.noc = mat->stencil.noc; 3981 for (i=0; i<=mat->stencil.dim; i++) { 3982 B->stencil.dims[i] = mat->stencil.dims[i]; 3983 B->stencil.starts[i] = mat->stencil.starts[i]; 3984 } 3985 3986 B->nooffproczerorows = mat->nooffproczerorows; 3987 B->nooffprocentries = mat->nooffprocentries; 3988 ierr = PetscLogEventEnd(MAT_Convert,mat,0,0,0);CHKERRQ(ierr); 3989 ierr = PetscObjectStateIncrease((PetscObject)B);CHKERRQ(ierr); 3990 PetscFunctionReturn(0); 3991 } 3992 3993 #undef __FUNCT__ 3994 #define __FUNCT__ "MatGetDiagonal" 3995 /*@ 3996 MatGetDiagonal - Gets the diagonal of a matrix. 3997 3998 Logically Collective on Mat and Vec 3999 4000 Input Parameters: 4001 + mat - the matrix 4002 - v - the vector for storing the diagonal 4003 4004 Output Parameter: 4005 . v - the diagonal of the matrix 4006 4007 Level: intermediate 4008 4009 Note: 4010 Currently only correct in parallel for square matrices. 4011 4012 Concepts: matrices^accessing diagonals 4013 4014 .seealso: MatGetRow(), MatGetSubMatrices(), MatGetSubmatrix(), MatGetRowMaxAbs() 4015 @*/ 4016 PetscErrorCode MatGetDiagonal(Mat mat,Vec v) 4017 { 4018 PetscErrorCode ierr; 4019 4020 PetscFunctionBegin; 4021 PetscValidHeaderSpecific(mat,MAT_CLASSID,1); 4022 PetscValidType(mat,1); 4023 PetscValidHeaderSpecific(v,VEC_CLASSID,2); 4024 if (!mat->assembled) SETERRQ(((PetscObject)mat)->comm,PETSC_ERR_ARG_WRONGSTATE,"Not for unassembled matrix"); 4025 if (!mat->ops->getdiagonal) SETERRQ1(((PetscObject)mat)->comm,PETSC_ERR_SUP,"Mat type %s",((PetscObject)mat)->type_name); 4026 MatCheckPreallocated(mat,1); 4027 4028 ierr = (*mat->ops->getdiagonal)(mat,v);CHKERRQ(ierr); 4029 ierr = PetscObjectStateIncrease((PetscObject)v);CHKERRQ(ierr); 4030 PetscFunctionReturn(0); 4031 } 4032 4033 #undef __FUNCT__ 4034 #define __FUNCT__ "MatGetRowMin" 4035 /*@ 4036 MatGetRowMin - Gets the minimum value (of the real part) of each 4037 row of the matrix 4038 4039 Logically Collective on Mat and Vec 4040 4041 Input Parameters: 4042 . mat - the matrix 4043 4044 Output Parameter: 4045 + v - the vector for storing the maximums 4046 - idx - the indices of the column found for each row (optional) 4047 4048 Level: intermediate 4049 4050 Notes: The result of this call are the same as if one converted the matrix to dense format 4051 and found the minimum value in each row (i.e. the implicit zeros are counted as zeros). 4052 4053 This code is only implemented for a couple of matrix formats. 4054 4055 Concepts: matrices^getting row maximums 4056 4057 .seealso: MatGetDiagonal(), MatGetSubMatrices(), MatGetSubmatrix(), MatGetRowMaxAbs(), 4058 MatGetRowMax() 4059 @*/ 4060 PetscErrorCode MatGetRowMin(Mat mat,Vec v,PetscInt idx[]) 4061 { 4062 PetscErrorCode ierr; 4063 4064 PetscFunctionBegin; 4065 PetscValidHeaderSpecific(mat,MAT_CLASSID,1); 4066 PetscValidType(mat,1); 4067 PetscValidHeaderSpecific(v,VEC_CLASSID,2); 4068 if (!mat->assembled) SETERRQ(((PetscObject)mat)->comm,PETSC_ERR_ARG_WRONGSTATE,"Not for unassembled matrix"); 4069 if (!mat->ops->getrowmax) SETERRQ1(PETSC_COMM_SELF,PETSC_ERR_SUP,"Mat type %s",((PetscObject)mat)->type_name); 4070 MatCheckPreallocated(mat,1); 4071 4072 ierr = (*mat->ops->getrowmin)(mat,v,idx);CHKERRQ(ierr); 4073 ierr = PetscObjectStateIncrease((PetscObject)v);CHKERRQ(ierr); 4074 PetscFunctionReturn(0); 4075 } 4076 4077 #undef __FUNCT__ 4078 #define __FUNCT__ "MatGetRowMinAbs" 4079 /*@ 4080 MatGetRowMinAbs - Gets the minimum value (in absolute value) of each 4081 row of the matrix 4082 4083 Logically Collective on Mat and Vec 4084 4085 Input Parameters: 4086 . mat - the matrix 4087 4088 Output Parameter: 4089 + v - the vector for storing the minimums 4090 - idx - the indices of the column found for each row (or PETSC_NULL if not needed) 4091 4092 Level: intermediate 4093 4094 Notes: if a row is completely empty or has only 0.0 values then the idx[] value for that 4095 row is 0 (the first column). 4096 4097 This code is only implemented for a couple of matrix formats. 4098 4099 Concepts: matrices^getting row maximums 4100 4101 .seealso: MatGetDiagonal(), MatGetSubMatrices(), MatGetSubmatrix(), MatGetRowMax(), MatGetRowMaxAbs(), MatGetRowMin() 4102 @*/ 4103 PetscErrorCode MatGetRowMinAbs(Mat mat,Vec v,PetscInt idx[]) 4104 { 4105 PetscErrorCode ierr; 4106 4107 PetscFunctionBegin; 4108 PetscValidHeaderSpecific(mat,MAT_CLASSID,1); 4109 PetscValidType(mat,1); 4110 PetscValidHeaderSpecific(v,VEC_CLASSID,2); 4111 if (!mat->assembled) SETERRQ(((PetscObject)mat)->comm,PETSC_ERR_ARG_WRONGSTATE,"Not for unassembled matrix"); 4112 if (!mat->ops->getrowminabs) SETERRQ1(((PetscObject)mat)->comm,PETSC_ERR_SUP,"Mat type %s",((PetscObject)mat)->type_name); 4113 MatCheckPreallocated(mat,1); 4114 if (idx) {ierr = PetscMemzero(idx,mat->rmap->n*sizeof(PetscInt));CHKERRQ(ierr);} 4115 4116 ierr = (*mat->ops->getrowminabs)(mat,v,idx);CHKERRQ(ierr); 4117 ierr = PetscObjectStateIncrease((PetscObject)v);CHKERRQ(ierr); 4118 PetscFunctionReturn(0); 4119 } 4120 4121 #undef __FUNCT__ 4122 #define __FUNCT__ "MatGetRowMax" 4123 /*@ 4124 MatGetRowMax - Gets the maximum value (of the real part) of each 4125 row of the matrix 4126 4127 Logically Collective on Mat and Vec 4128 4129 Input Parameters: 4130 . mat - the matrix 4131 4132 Output Parameter: 4133 + v - the vector for storing the maximums 4134 - idx - the indices of the column found for each row (optional) 4135 4136 Level: intermediate 4137 4138 Notes: The result of this call are the same as if one converted the matrix to dense format 4139 and found the minimum value in each row (i.e. the implicit zeros are counted as zeros). 4140 4141 This code is only implemented for a couple of matrix formats. 4142 4143 Concepts: matrices^getting row maximums 4144 4145 .seealso: MatGetDiagonal(), MatGetSubMatrices(), MatGetSubmatrix(), MatGetRowMaxAbs(), MatGetRowMin() 4146 @*/ 4147 PetscErrorCode MatGetRowMax(Mat mat,Vec v,PetscInt idx[]) 4148 { 4149 PetscErrorCode ierr; 4150 4151 PetscFunctionBegin; 4152 PetscValidHeaderSpecific(mat,MAT_CLASSID,1); 4153 PetscValidType(mat,1); 4154 PetscValidHeaderSpecific(v,VEC_CLASSID,2); 4155 if (!mat->assembled) SETERRQ(((PetscObject)mat)->comm,PETSC_ERR_ARG_WRONGSTATE,"Not for unassembled matrix"); 4156 if (!mat->ops->getrowmax) SETERRQ1(((PetscObject)mat)->comm,PETSC_ERR_SUP,"Mat type %s",((PetscObject)mat)->type_name); 4157 MatCheckPreallocated(mat,1); 4158 4159 ierr = (*mat->ops->getrowmax)(mat,v,idx);CHKERRQ(ierr); 4160 ierr = PetscObjectStateIncrease((PetscObject)v);CHKERRQ(ierr); 4161 PetscFunctionReturn(0); 4162 } 4163 4164 #undef __FUNCT__ 4165 #define __FUNCT__ "MatGetRowMaxAbs" 4166 /*@ 4167 MatGetRowMaxAbs - Gets the maximum value (in absolute value) of each 4168 row of the matrix 4169 4170 Logically Collective on Mat and Vec 4171 4172 Input Parameters: 4173 . mat - the matrix 4174 4175 Output Parameter: 4176 + v - the vector for storing the maximums 4177 - idx - the indices of the column found for each row (or PETSC_NULL if not needed) 4178 4179 Level: intermediate 4180 4181 Notes: if a row is completely empty or has only 0.0 values then the idx[] value for that 4182 row is 0 (the first column). 4183 4184 This code is only implemented for a couple of matrix formats. 4185 4186 Concepts: matrices^getting row maximums 4187 4188 .seealso: MatGetDiagonal(), MatGetSubMatrices(), MatGetSubmatrix(), MatGetRowMax(), MatGetRowMin() 4189 @*/ 4190 PetscErrorCode MatGetRowMaxAbs(Mat mat,Vec v,PetscInt idx[]) 4191 { 4192 PetscErrorCode ierr; 4193 4194 PetscFunctionBegin; 4195 PetscValidHeaderSpecific(mat,MAT_CLASSID,1); 4196 PetscValidType(mat,1); 4197 PetscValidHeaderSpecific(v,VEC_CLASSID,2); 4198 if (!mat->assembled) SETERRQ(((PetscObject)mat)->comm,PETSC_ERR_ARG_WRONGSTATE,"Not for unassembled matrix"); 4199 if (!mat->ops->getrowmaxabs) SETERRQ1(PETSC_COMM_SELF,PETSC_ERR_SUP,"Mat type %s",((PetscObject)mat)->type_name); 4200 MatCheckPreallocated(mat,1); 4201 if (idx) {ierr = PetscMemzero(idx,mat->rmap->n*sizeof(PetscInt));CHKERRQ(ierr);} 4202 4203 ierr = (*mat->ops->getrowmaxabs)(mat,v,idx);CHKERRQ(ierr); 4204 ierr = PetscObjectStateIncrease((PetscObject)v);CHKERRQ(ierr); 4205 PetscFunctionReturn(0); 4206 } 4207 4208 #undef __FUNCT__ 4209 #define __FUNCT__ "MatGetRowSum" 4210 /*@ 4211 MatGetRowSum - Gets the sum of each row of the matrix 4212 4213 Logically Collective on Mat and Vec 4214 4215 Input Parameters: 4216 . mat - the matrix 4217 4218 Output Parameter: 4219 . v - the vector for storing the sum of rows 4220 4221 Level: intermediate 4222 4223 Notes: This code is slow since it is not currently specialized for different formats 4224 4225 Concepts: matrices^getting row sums 4226 4227 .seealso: MatGetDiagonal(), MatGetSubMatrices(), MatGetSubmatrix(), MatGetRowMax(), MatGetRowMin() 4228 @*/ 4229 PetscErrorCode MatGetRowSum(Mat mat, Vec v) 4230 { 4231 PetscInt start = 0, end = 0, row; 4232 PetscScalar *array; 4233 PetscErrorCode ierr; 4234 4235 PetscFunctionBegin; 4236 PetscValidHeaderSpecific(mat,MAT_CLASSID,1); 4237 PetscValidType(mat,1); 4238 PetscValidHeaderSpecific(v,VEC_CLASSID,2); 4239 if (!mat->assembled) SETERRQ(((PetscObject)mat)->comm,PETSC_ERR_ARG_WRONGSTATE,"Not for unassembled matrix"); 4240 MatCheckPreallocated(mat,1); 4241 ierr = MatGetOwnershipRange(mat, &start, &end);CHKERRQ(ierr); 4242 ierr = VecGetArray(v, &array);CHKERRQ(ierr); 4243 for(row = start; row < end; ++row) { 4244 PetscInt ncols, col; 4245 const PetscInt *cols; 4246 const PetscScalar *vals; 4247 4248 array[row - start] = 0.0; 4249 ierr = MatGetRow(mat, row, &ncols, &cols, &vals);CHKERRQ(ierr); 4250 for(col = 0; col < ncols; col++) { 4251 array[row - start] += vals[col]; 4252 } 4253 ierr = MatRestoreRow(mat, row, &ncols, &cols, &vals);CHKERRQ(ierr); 4254 } 4255 ierr = VecRestoreArray(v, &array);CHKERRQ(ierr); 4256 ierr = PetscObjectStateIncrease((PetscObject) v);CHKERRQ(ierr); 4257 PetscFunctionReturn(0); 4258 } 4259 4260 #undef __FUNCT__ 4261 #define __FUNCT__ "MatTranspose" 4262 /*@ 4263 MatTranspose - Computes an in-place or out-of-place transpose of a matrix. 4264 4265 Collective on Mat 4266 4267 Input Parameter: 4268 + mat - the matrix to transpose 4269 - reuse - either MAT_INITIAL_MATRIX or MAT_REUSE_MATRIX 4270 4271 Output Parameters: 4272 . B - the transpose 4273 4274 Notes: 4275 If you pass in &mat for B the transpose will be done in place, for example MatTranspose(mat,MAT_REUSE_MATRIX,&mat); 4276 4277 Consider using MatCreateTranspose() instead if you only need a matrix that behaves like the transpose, but don't need the storage to be changed. 4278 4279 Level: intermediate 4280 4281 Concepts: matrices^transposing 4282 4283 .seealso: MatMultTranspose(), MatMultTransposeAdd(), MatIsTranspose(), MatReuse 4284 @*/ 4285 PetscErrorCode MatTranspose(Mat mat,MatReuse reuse,Mat *B) 4286 { 4287 PetscErrorCode ierr; 4288 4289 PetscFunctionBegin; 4290 PetscValidHeaderSpecific(mat,MAT_CLASSID,1); 4291 PetscValidType(mat,1); 4292 if (!mat->assembled) SETERRQ(((PetscObject)mat)->comm,PETSC_ERR_ARG_WRONGSTATE,"Not for unassembled matrix"); 4293 if (mat->factortype) SETERRQ(((PetscObject)mat)->comm,PETSC_ERR_ARG_WRONGSTATE,"Not for factored matrix"); 4294 if (!mat->ops->transpose) SETERRQ1(((PetscObject)mat)->comm,PETSC_ERR_SUP,"Mat type %s",((PetscObject)mat)->type_name); 4295 MatCheckPreallocated(mat,1); 4296 4297 ierr = PetscLogEventBegin(MAT_Transpose,mat,0,0,0);CHKERRQ(ierr); 4298 ierr = (*mat->ops->transpose)(mat,reuse,B);CHKERRQ(ierr); 4299 ierr = PetscLogEventEnd(MAT_Transpose,mat,0,0,0);CHKERRQ(ierr); 4300 if (B) {ierr = PetscObjectStateIncrease((PetscObject)*B);CHKERRQ(ierr);} 4301 PetscFunctionReturn(0); 4302 } 4303 4304 #undef __FUNCT__ 4305 #define __FUNCT__ "MatIsTranspose" 4306 /*@ 4307 MatIsTranspose - Test whether a matrix is another one's transpose, 4308 or its own, in which case it tests symmetry. 4309 4310 Collective on Mat 4311 4312 Input Parameter: 4313 + A - the matrix to test 4314 - B - the matrix to test against, this can equal the first parameter 4315 4316 Output Parameters: 4317 . flg - the result 4318 4319 Notes: 4320 Only available for SeqAIJ/MPIAIJ matrices. The sequential algorithm 4321 has a running time of the order of the number of nonzeros; the parallel 4322 test involves parallel copies of the block-offdiagonal parts of the matrix. 4323 4324 Level: intermediate 4325 4326 Concepts: matrices^transposing, matrix^symmetry 4327 4328 .seealso: MatTranspose(), MatIsSymmetric(), MatIsHermitian() 4329 @*/ 4330 PetscErrorCode MatIsTranspose(Mat A,Mat B,PetscReal tol,PetscBool *flg) 4331 { 4332 PetscErrorCode ierr,(*f)(Mat,Mat,PetscReal,PetscBool *),(*g)(Mat,Mat,PetscReal,PetscBool *); 4333 4334 PetscFunctionBegin; 4335 PetscValidHeaderSpecific(A,MAT_CLASSID,1); 4336 PetscValidHeaderSpecific(B,MAT_CLASSID,2); 4337 PetscValidPointer(flg,3); 4338 ierr = PetscObjectQueryFunction((PetscObject)A,"MatIsTranspose_C",(void (**)(void))&f);CHKERRQ(ierr); 4339 ierr = PetscObjectQueryFunction((PetscObject)B,"MatIsTranspose_C",(void (**)(void))&g);CHKERRQ(ierr); 4340 *flg = PETSC_FALSE; 4341 if (f && g) { 4342 if (f == g) { 4343 ierr = (*f)(A,B,tol,flg);CHKERRQ(ierr); 4344 } else { 4345 SETERRQ(((PetscObject)A)->comm,PETSC_ERR_ARG_NOTSAMETYPE,"Matrices do not have the same comparator for symmetry test"); 4346 } 4347 } else { 4348 const MatType mattype; 4349 if (!f) {ierr = MatGetType(A,&mattype);CHKERRQ(ierr);} 4350 else {ierr = MatGetType(B,&mattype);CHKERRQ(ierr);} 4351 SETERRQ1(PETSC_COMM_SELF,PETSC_ERR_SUP,"Matrix of type <%s> does not support checking for transpose",mattype); 4352 } 4353 PetscFunctionReturn(0); 4354 } 4355 4356 #undef __FUNCT__ 4357 #define __FUNCT__ "MatHermitianTranspose" 4358 /*@ 4359 MatHermitianTranspose - Computes an in-place or out-of-place transpose of a matrix in complex conjugate. 4360 4361 Collective on Mat 4362 4363 Input Parameter: 4364 + mat - the matrix to transpose and complex conjugate 4365 - reuse - store the transpose matrix in the provided B 4366 4367 Output Parameters: 4368 . B - the Hermitian 4369 4370 Notes: 4371 If you pass in &mat for B the Hermitian will be done in place 4372 4373 Level: intermediate 4374 4375 Concepts: matrices^transposing, complex conjugatex 4376 4377 .seealso: MatTranspose(), MatMultTranspose(), MatMultTransposeAdd(), MatIsTranspose(), MatReuse 4378 @*/ 4379 PetscErrorCode MatHermitianTranspose(Mat mat,MatReuse reuse,Mat *B) 4380 { 4381 PetscErrorCode ierr; 4382 4383 PetscFunctionBegin; 4384 ierr = MatTranspose(mat,reuse,B);CHKERRQ(ierr); 4385 #if defined(PETSC_USE_COMPLEX) 4386 ierr = MatConjugate(*B);CHKERRQ(ierr); 4387 #endif 4388 PetscFunctionReturn(0); 4389 } 4390 4391 #undef __FUNCT__ 4392 #define __FUNCT__ "MatIsHermitianTranspose" 4393 /*@ 4394 MatIsHermitianTranspose - Test whether a matrix is another one's Hermitian transpose, 4395 4396 Collective on Mat 4397 4398 Input Parameter: 4399 + A - the matrix to test 4400 - B - the matrix to test against, this can equal the first parameter 4401 4402 Output Parameters: 4403 . flg - the result 4404 4405 Notes: 4406 Only available for SeqAIJ/MPIAIJ matrices. The sequential algorithm 4407 has a running time of the order of the number of nonzeros; the parallel 4408 test involves parallel copies of the block-offdiagonal parts of the matrix. 4409 4410 Level: intermediate 4411 4412 Concepts: matrices^transposing, matrix^symmetry 4413 4414 .seealso: MatTranspose(), MatIsSymmetric(), MatIsHermitian(), MatIsTranspose() 4415 @*/ 4416 PetscErrorCode MatIsHermitianTranspose(Mat A,Mat B,PetscReal tol,PetscBool *flg) 4417 { 4418 PetscErrorCode ierr,(*f)(Mat,Mat,PetscReal,PetscBool *),(*g)(Mat,Mat,PetscReal,PetscBool *); 4419 4420 PetscFunctionBegin; 4421 PetscValidHeaderSpecific(A,MAT_CLASSID,1); 4422 PetscValidHeaderSpecific(B,MAT_CLASSID,2); 4423 PetscValidPointer(flg,3); 4424 ierr = PetscObjectQueryFunction((PetscObject)A,"MatIsHermitianTranspose_C",(void (**)(void))&f);CHKERRQ(ierr); 4425 ierr = PetscObjectQueryFunction((PetscObject)B,"MatIsHermitianTranspose_C",(void (**)(void))&g);CHKERRQ(ierr); 4426 if (f && g) { 4427 if (f==g) { 4428 ierr = (*f)(A,B,tol,flg);CHKERRQ(ierr); 4429 } else { 4430 SETERRQ(((PetscObject)A)->comm,PETSC_ERR_ARG_NOTSAMETYPE,"Matrices do not have the same comparator for Hermitian test"); 4431 } 4432 } 4433 PetscFunctionReturn(0); 4434 } 4435 4436 #undef __FUNCT__ 4437 #define __FUNCT__ "MatPermute" 4438 /*@ 4439 MatPermute - Creates a new matrix with rows and columns permuted from the 4440 original. 4441 4442 Collective on Mat 4443 4444 Input Parameters: 4445 + mat - the matrix to permute 4446 . row - row permutation, each processor supplies only the permutation for its rows 4447 - col - column permutation, each processor needs the entire column permutation, that is 4448 this is the same size as the total number of columns in the matrix. It can often 4449 be obtained with ISAllGather() on the row permutation 4450 4451 Output Parameters: 4452 . B - the permuted matrix 4453 4454 Level: advanced 4455 4456 Concepts: matrices^permuting 4457 4458 .seealso: MatGetOrdering(), ISAllGather() 4459 4460 @*/ 4461 PetscErrorCode MatPermute(Mat mat,IS row,IS col,Mat *B) 4462 { 4463 PetscErrorCode ierr; 4464 4465 PetscFunctionBegin; 4466 PetscValidHeaderSpecific(mat,MAT_CLASSID,1); 4467 PetscValidType(mat,1); 4468 PetscValidHeaderSpecific(row,IS_CLASSID,2); 4469 PetscValidHeaderSpecific(col,IS_CLASSID,3); 4470 PetscValidPointer(B,4); 4471 if (!mat->assembled) SETERRQ(((PetscObject)mat)->comm,PETSC_ERR_ARG_WRONGSTATE,"Not for unassembled matrix"); 4472 if (mat->factortype) SETERRQ(((PetscObject)mat)->comm,PETSC_ERR_ARG_WRONGSTATE,"Not for factored matrix"); 4473 if (!mat->ops->permute) SETERRQ1(PETSC_COMM_SELF,PETSC_ERR_SUP,"MatPermute not available for Mat type %s",((PetscObject)mat)->type_name); 4474 MatCheckPreallocated(mat,1); 4475 4476 ierr = (*mat->ops->permute)(mat,row,col,B);CHKERRQ(ierr); 4477 ierr = PetscObjectStateIncrease((PetscObject)*B);CHKERRQ(ierr); 4478 PetscFunctionReturn(0); 4479 } 4480 4481 #undef __FUNCT__ 4482 #define __FUNCT__ "MatEqual" 4483 /*@ 4484 MatEqual - Compares two matrices. 4485 4486 Collective on Mat 4487 4488 Input Parameters: 4489 + A - the first matrix 4490 - B - the second matrix 4491 4492 Output Parameter: 4493 . flg - PETSC_TRUE if the matrices are equal; PETSC_FALSE otherwise. 4494 4495 Level: intermediate 4496 4497 Concepts: matrices^equality between 4498 @*/ 4499 PetscErrorCode MatEqual(Mat A,Mat B,PetscBool *flg) 4500 { 4501 PetscErrorCode ierr; 4502 4503 PetscFunctionBegin; 4504 PetscValidHeaderSpecific(A,MAT_CLASSID,1); 4505 PetscValidHeaderSpecific(B,MAT_CLASSID,2); 4506 PetscValidType(A,1); 4507 PetscValidType(B,2); 4508 PetscValidIntPointer(flg,3); 4509 PetscCheckSameComm(A,1,B,2); 4510 MatCheckPreallocated(B,2); 4511 if (!A->assembled) SETERRQ(((PetscObject)A)->comm,PETSC_ERR_ARG_WRONGSTATE,"Not for unassembled matrix"); 4512 if (!B->assembled) SETERRQ(((PetscObject)A)->comm,PETSC_ERR_ARG_WRONGSTATE,"Not for unassembled matrix"); 4513 if (A->rmap->N != B->rmap->N || A->cmap->N != B->cmap->N) SETERRQ4(((PetscObject)A)->comm,PETSC_ERR_ARG_SIZ,"Mat A,Mat B: global dim %D %D %D %D",A->rmap->N,B->rmap->N,A->cmap->N,B->cmap->N); 4514 if (!A->ops->equal) SETERRQ1(((PetscObject)A)->comm,PETSC_ERR_SUP,"Mat type %s",((PetscObject)A)->type_name); 4515 if (!B->ops->equal) SETERRQ1(((PetscObject)A)->comm,PETSC_ERR_SUP,"Mat type %s",((PetscObject)B)->type_name); 4516 if (A->ops->equal != B->ops->equal) SETERRQ2(((PetscObject)A)->comm,PETSC_ERR_ARG_INCOMP,"A is type: %s\nB is type: %s",((PetscObject)A)->type_name,((PetscObject)B)->type_name); 4517 MatCheckPreallocated(A,1); 4518 4519 ierr = (*A->ops->equal)(A,B,flg);CHKERRQ(ierr); 4520 PetscFunctionReturn(0); 4521 } 4522 4523 #undef __FUNCT__ 4524 #define __FUNCT__ "MatDiagonalScale" 4525 /*@ 4526 MatDiagonalScale - Scales a matrix on the left and right by diagonal 4527 matrices that are stored as vectors. Either of the two scaling 4528 matrices can be PETSC_NULL. 4529 4530 Collective on Mat 4531 4532 Input Parameters: 4533 + mat - the matrix to be scaled 4534 . l - the left scaling vector (or PETSC_NULL) 4535 - r - the right scaling vector (or PETSC_NULL) 4536 4537 Notes: 4538 MatDiagonalScale() computes A = LAR, where 4539 L = a diagonal matrix (stored as a vector), R = a diagonal matrix (stored as a vector) 4540 The L scales the rows of the matrix, the R scales the columns of the matrix. 4541 4542 Level: intermediate 4543 4544 Concepts: matrices^diagonal scaling 4545 Concepts: diagonal scaling of matrices 4546 4547 .seealso: MatScale() 4548 @*/ 4549 PetscErrorCode MatDiagonalScale(Mat mat,Vec l,Vec r) 4550 { 4551 PetscErrorCode ierr; 4552 4553 PetscFunctionBegin; 4554 PetscValidHeaderSpecific(mat,MAT_CLASSID,1); 4555 PetscValidType(mat,1); 4556 if (!mat->ops->diagonalscale) SETERRQ1(((PetscObject)mat)->comm,PETSC_ERR_SUP,"Mat type %s",((PetscObject)mat)->type_name); 4557 if (l) {PetscValidHeaderSpecific(l,VEC_CLASSID,2);PetscCheckSameComm(mat,1,l,2);} 4558 if (r) {PetscValidHeaderSpecific(r,VEC_CLASSID,3);PetscCheckSameComm(mat,1,r,3);} 4559 if (!mat->assembled) SETERRQ(((PetscObject)mat)->comm,PETSC_ERR_ARG_WRONGSTATE,"Not for unassembled matrix"); 4560 if (mat->factortype) SETERRQ(((PetscObject)mat)->comm,PETSC_ERR_ARG_WRONGSTATE,"Not for factored matrix"); 4561 MatCheckPreallocated(mat,1); 4562 4563 ierr = PetscLogEventBegin(MAT_Scale,mat,0,0,0);CHKERRQ(ierr); 4564 ierr = (*mat->ops->diagonalscale)(mat,l,r);CHKERRQ(ierr); 4565 ierr = PetscLogEventEnd(MAT_Scale,mat,0,0,0);CHKERRQ(ierr); 4566 ierr = PetscObjectStateIncrease((PetscObject)mat);CHKERRQ(ierr); 4567 #if defined(PETSC_HAVE_CUSP) 4568 if (mat->valid_GPU_matrix != PETSC_CUSP_UNALLOCATED) { 4569 mat->valid_GPU_matrix = PETSC_CUSP_CPU; 4570 } 4571 #endif 4572 PetscFunctionReturn(0); 4573 } 4574 4575 #undef __FUNCT__ 4576 #define __FUNCT__ "MatScale" 4577 /*@ 4578 MatScale - Scales all elements of a matrix by a given number. 4579 4580 Logically Collective on Mat 4581 4582 Input Parameters: 4583 + mat - the matrix to be scaled 4584 - a - the scaling value 4585 4586 Output Parameter: 4587 . mat - the scaled matrix 4588 4589 Level: intermediate 4590 4591 Concepts: matrices^scaling all entries 4592 4593 .seealso: MatDiagonalScale() 4594 @*/ 4595 PetscErrorCode MatScale(Mat mat,PetscScalar a) 4596 { 4597 PetscErrorCode ierr; 4598 4599 PetscFunctionBegin; 4600 PetscValidHeaderSpecific(mat,MAT_CLASSID,1); 4601 PetscValidType(mat,1); 4602 if (a != (PetscScalar)1.0 && !mat->ops->scale) SETERRQ1(PETSC_COMM_SELF,PETSC_ERR_SUP,"Mat type %s",((PetscObject)mat)->type_name); 4603 if (!mat->assembled) SETERRQ(((PetscObject)mat)->comm,PETSC_ERR_ARG_WRONGSTATE,"Not for unassembled matrix"); 4604 if (mat->factortype) SETERRQ(((PetscObject)mat)->comm,PETSC_ERR_ARG_WRONGSTATE,"Not for factored matrix"); 4605 PetscValidLogicalCollectiveScalar(mat,a,2); 4606 MatCheckPreallocated(mat,1); 4607 4608 ierr = PetscLogEventBegin(MAT_Scale,mat,0,0,0);CHKERRQ(ierr); 4609 if (a != (PetscScalar)1.0) { 4610 ierr = (*mat->ops->scale)(mat,a);CHKERRQ(ierr); 4611 ierr = PetscObjectStateIncrease((PetscObject)mat);CHKERRQ(ierr); 4612 } 4613 ierr = PetscLogEventEnd(MAT_Scale,mat,0,0,0);CHKERRQ(ierr); 4614 #if defined(PETSC_HAVE_CUSP) 4615 if (mat->valid_GPU_matrix != PETSC_CUSP_UNALLOCATED) { 4616 mat->valid_GPU_matrix = PETSC_CUSP_CPU; 4617 } 4618 #endif 4619 PetscFunctionReturn(0); 4620 } 4621 4622 #undef __FUNCT__ 4623 #define __FUNCT__ "MatNorm" 4624 /*@ 4625 MatNorm - Calculates various norms of a matrix. 4626 4627 Collective on Mat 4628 4629 Input Parameters: 4630 + mat - the matrix 4631 - type - the type of norm, NORM_1, NORM_FROBENIUS, NORM_INFINITY 4632 4633 Output Parameters: 4634 . nrm - the resulting norm 4635 4636 Level: intermediate 4637 4638 Concepts: matrices^norm 4639 Concepts: norm^of matrix 4640 @*/ 4641 PetscErrorCode MatNorm(Mat mat,NormType type,PetscReal *nrm) 4642 { 4643 PetscErrorCode ierr; 4644 4645 PetscFunctionBegin; 4646 PetscValidHeaderSpecific(mat,MAT_CLASSID,1); 4647 PetscValidType(mat,1); 4648 PetscValidScalarPointer(nrm,3); 4649 4650 if (!mat->assembled) SETERRQ(((PetscObject)mat)->comm,PETSC_ERR_ARG_WRONGSTATE,"Not for unassembled matrix"); 4651 if (mat->factortype) SETERRQ(((PetscObject)mat)->comm,PETSC_ERR_ARG_WRONGSTATE,"Not for factored matrix"); 4652 if (!mat->ops->norm) SETERRQ1(((PetscObject)mat)->comm,PETSC_ERR_SUP,"Mat type %s",((PetscObject)mat)->type_name); 4653 MatCheckPreallocated(mat,1); 4654 4655 ierr = (*mat->ops->norm)(mat,type,nrm);CHKERRQ(ierr); 4656 PetscFunctionReturn(0); 4657 } 4658 4659 /* 4660 This variable is used to prevent counting of MatAssemblyBegin() that 4661 are called from within a MatAssemblyEnd(). 4662 */ 4663 static PetscInt MatAssemblyEnd_InUse = 0; 4664 #undef __FUNCT__ 4665 #define __FUNCT__ "MatAssemblyBegin" 4666 /*@ 4667 MatAssemblyBegin - Begins assembling the matrix. This routine should 4668 be called after completing all calls to MatSetValues(). 4669 4670 Collective on Mat 4671 4672 Input Parameters: 4673 + mat - the matrix 4674 - type - type of assembly, either MAT_FLUSH_ASSEMBLY or MAT_FINAL_ASSEMBLY 4675 4676 Notes: 4677 MatSetValues() generally caches the values. The matrix is ready to 4678 use only after MatAssemblyBegin() and MatAssemblyEnd() have been called. 4679 Use MAT_FLUSH_ASSEMBLY when switching between ADD_VALUES and INSERT_VALUES 4680 in MatSetValues(); use MAT_FINAL_ASSEMBLY for the final assembly before 4681 using the matrix. 4682 4683 Level: beginner 4684 4685 Concepts: matrices^assembling 4686 4687 .seealso: MatAssemblyEnd(), MatSetValues(), MatAssembled() 4688 @*/ 4689 PetscErrorCode MatAssemblyBegin(Mat mat,MatAssemblyType type) 4690 { 4691 PetscErrorCode ierr; 4692 4693 PetscFunctionBegin; 4694 PetscValidHeaderSpecific(mat,MAT_CLASSID,1); 4695 PetscValidType(mat,1); 4696 MatCheckPreallocated(mat,1); 4697 if (mat->factortype) SETERRQ(((PetscObject)mat)->comm,PETSC_ERR_ARG_WRONGSTATE,"Not for factored matrix.\nDid you forget to call MatSetUnfactored()?"); 4698 if (mat->assembled) { 4699 mat->was_assembled = PETSC_TRUE; 4700 mat->assembled = PETSC_FALSE; 4701 } 4702 if (!MatAssemblyEnd_InUse) { 4703 ierr = PetscLogEventBegin(MAT_AssemblyBegin,mat,0,0,0);CHKERRQ(ierr); 4704 if (mat->ops->assemblybegin){ierr = (*mat->ops->assemblybegin)(mat,type);CHKERRQ(ierr);} 4705 ierr = PetscLogEventEnd(MAT_AssemblyBegin,mat,0,0,0);CHKERRQ(ierr); 4706 } else { 4707 if (mat->ops->assemblybegin){ierr = (*mat->ops->assemblybegin)(mat,type);CHKERRQ(ierr);} 4708 } 4709 PetscFunctionReturn(0); 4710 } 4711 4712 #undef __FUNCT__ 4713 #define __FUNCT__ "MatAssembled" 4714 /*@ 4715 MatAssembled - Indicates if a matrix has been assembled and is ready for 4716 use; for example, in matrix-vector product. 4717 4718 Not Collective 4719 4720 Input Parameter: 4721 . mat - the matrix 4722 4723 Output Parameter: 4724 . assembled - PETSC_TRUE or PETSC_FALSE 4725 4726 Level: advanced 4727 4728 Concepts: matrices^assembled? 4729 4730 .seealso: MatAssemblyEnd(), MatSetValues(), MatAssemblyBegin() 4731 @*/ 4732 PetscErrorCode MatAssembled(Mat mat,PetscBool *assembled) 4733 { 4734 PetscFunctionBegin; 4735 PetscValidHeaderSpecific(mat,MAT_CLASSID,1); 4736 PetscValidType(mat,1); 4737 PetscValidPointer(assembled,2); 4738 *assembled = mat->assembled; 4739 PetscFunctionReturn(0); 4740 } 4741 4742 #undef __FUNCT__ 4743 #define __FUNCT__ "MatView_Private" 4744 /* 4745 Processes command line options to determine if/how a matrix 4746 is to be viewed. Called by MatAssemblyEnd() and MatLoad(). 4747 */ 4748 PetscErrorCode MatView_Private(Mat mat) 4749 { 4750 PetscErrorCode ierr; 4751 PetscBool flg1 = PETSC_FALSE,flg2 = PETSC_FALSE,flg3 = PETSC_FALSE,flg4 = PETSC_FALSE,flg6 = PETSC_FALSE,flg7 = PETSC_FALSE,flg8 = PETSC_FALSE; 4752 static PetscBool incall = PETSC_FALSE; 4753 #if defined(PETSC_USE_SOCKET_VIEWER) 4754 PetscBool flg5 = PETSC_FALSE; 4755 #endif 4756 4757 PetscFunctionBegin; 4758 if (incall) PetscFunctionReturn(0); 4759 incall = PETSC_TRUE; 4760 ierr = PetscObjectOptionsBegin((PetscObject)mat);CHKERRQ(ierr); 4761 ierr = PetscOptionsBool("-mat_view_info","Information on matrix size","MatView",flg1,&flg1,PETSC_NULL);CHKERRQ(ierr); 4762 ierr = PetscOptionsBool("-mat_view_info_detailed","Nonzeros in the matrix","MatView",flg2,&flg2,PETSC_NULL);CHKERRQ(ierr); 4763 ierr = PetscOptionsBool("-mat_view","Print matrix to stdout","MatView",flg3,&flg3,PETSC_NULL);CHKERRQ(ierr); 4764 ierr = PetscOptionsBool("-mat_view_matlab","Print matrix to stdout in a format Matlab can read","MatView",flg4,&flg4,PETSC_NULL);CHKERRQ(ierr); 4765 #if defined(PETSC_USE_SOCKET_VIEWER) 4766 ierr = PetscOptionsBool("-mat_view_socket","Send matrix to socket (can be read from matlab)","MatView",flg5,&flg5,PETSC_NULL);CHKERRQ(ierr); 4767 #endif 4768 ierr = PetscOptionsBool("-mat_view_binary","Save matrix to file in binary format","MatView",flg6,&flg6,PETSC_NULL);CHKERRQ(ierr); 4769 ierr = PetscOptionsBool("-mat_view_draw","Draw the matrix nonzero structure","MatView",flg7,&flg7,PETSC_NULL);CHKERRQ(ierr); 4770 ierr = PetscOptionsEnd();CHKERRQ(ierr); 4771 4772 if (flg1) { 4773 PetscViewer viewer; 4774 4775 ierr = PetscViewerASCIIGetStdout(((PetscObject)mat)->comm,&viewer);CHKERRQ(ierr); 4776 ierr = PetscViewerPushFormat(viewer,PETSC_VIEWER_ASCII_INFO);CHKERRQ(ierr); 4777 ierr = MatView(mat,viewer);CHKERRQ(ierr); 4778 ierr = PetscViewerPopFormat(viewer);CHKERRQ(ierr); 4779 } 4780 if (flg2) { 4781 PetscViewer viewer; 4782 4783 ierr = PetscViewerASCIIGetStdout(((PetscObject)mat)->comm,&viewer);CHKERRQ(ierr); 4784 ierr = PetscViewerPushFormat(viewer,PETSC_VIEWER_ASCII_INFO_DETAIL);CHKERRQ(ierr); 4785 ierr = MatView(mat,viewer);CHKERRQ(ierr); 4786 ierr = PetscViewerPopFormat(viewer);CHKERRQ(ierr); 4787 } 4788 if (flg3) { 4789 PetscViewer viewer; 4790 4791 ierr = PetscViewerASCIIGetStdout(((PetscObject)mat)->comm,&viewer);CHKERRQ(ierr); 4792 ierr = MatView(mat,viewer);CHKERRQ(ierr); 4793 } 4794 if (flg4) { 4795 PetscViewer viewer; 4796 4797 ierr = PetscViewerASCIIGetStdout(((PetscObject)mat)->comm,&viewer);CHKERRQ(ierr); 4798 ierr = PetscViewerPushFormat(viewer,PETSC_VIEWER_ASCII_MATLAB);CHKERRQ(ierr); 4799 ierr = MatView(mat,viewer);CHKERRQ(ierr); 4800 ierr = PetscViewerPopFormat(viewer);CHKERRQ(ierr); 4801 } 4802 #if defined(PETSC_USE_SOCKET_VIEWER) 4803 if (flg5) { 4804 ierr = MatView(mat,PETSC_VIEWER_SOCKET_(((PetscObject)mat)->comm));CHKERRQ(ierr); 4805 ierr = PetscViewerFlush(PETSC_VIEWER_SOCKET_(((PetscObject)mat)->comm));CHKERRQ(ierr); 4806 } 4807 #endif 4808 if (flg6) { 4809 ierr = MatView(mat,PETSC_VIEWER_BINARY_(((PetscObject)mat)->comm));CHKERRQ(ierr); 4810 ierr = PetscViewerFlush(PETSC_VIEWER_BINARY_(((PetscObject)mat)->comm));CHKERRQ(ierr); 4811 } 4812 if (flg7) { 4813 ierr = PetscOptionsGetBool(((PetscObject)mat)->prefix,"-mat_view_contour",&flg8,PETSC_NULL);CHKERRQ(ierr); 4814 if (flg8) { 4815 PetscViewerPushFormat(PETSC_VIEWER_DRAW_(((PetscObject)mat)->comm),PETSC_VIEWER_DRAW_CONTOUR);CHKERRQ(ierr); 4816 } 4817 ierr = MatView(mat,PETSC_VIEWER_DRAW_(((PetscObject)mat)->comm));CHKERRQ(ierr); 4818 ierr = PetscViewerFlush(PETSC_VIEWER_DRAW_(((PetscObject)mat)->comm));CHKERRQ(ierr); 4819 if (flg8) { 4820 PetscViewerPopFormat(PETSC_VIEWER_DRAW_(((PetscObject)mat)->comm));CHKERRQ(ierr); 4821 } 4822 } 4823 incall = PETSC_FALSE; 4824 PetscFunctionReturn(0); 4825 } 4826 4827 #undef __FUNCT__ 4828 #define __FUNCT__ "MatAssemblyEnd" 4829 /*@ 4830 MatAssemblyEnd - Completes assembling the matrix. This routine should 4831 be called after MatAssemblyBegin(). 4832 4833 Collective on Mat 4834 4835 Input Parameters: 4836 + mat - the matrix 4837 - type - type of assembly, either MAT_FLUSH_ASSEMBLY or MAT_FINAL_ASSEMBLY 4838 4839 Options Database Keys: 4840 + -mat_view_info - Prints info on matrix at conclusion of MatEndAssembly() 4841 . -mat_view_info_detailed - Prints more detailed info 4842 . -mat_view - Prints matrix in ASCII format 4843 . -mat_view_matlab - Prints matrix in Matlab format 4844 . -mat_view_draw - PetscDraws nonzero structure of matrix, using MatView() and PetscDrawOpenX(). 4845 . -display <name> - Sets display name (default is host) 4846 . -draw_pause <sec> - Sets number of seconds to pause after display 4847 . -mat_view_socket - Sends matrix to socket, can be accessed from Matlab (See the <a href="../../docs/manual.pdf">users manual</a>) 4848 . -viewer_socket_machine <machine> 4849 . -viewer_socket_port <port> 4850 . -mat_view_binary - save matrix to file in binary format 4851 - -viewer_binary_filename <name> 4852 4853 Notes: 4854 MatSetValues() generally caches the values. The matrix is ready to 4855 use only after MatAssemblyBegin() and MatAssemblyEnd() have been called. 4856 Use MAT_FLUSH_ASSEMBLY when switching between ADD_VALUES and INSERT_VALUES 4857 in MatSetValues(); use MAT_FINAL_ASSEMBLY for the final assembly before 4858 using the matrix. 4859 4860 Level: beginner 4861 4862 .seealso: MatAssemblyBegin(), MatSetValues(), PetscDrawOpenX(), MatView(), MatAssembled(), PetscViewerSocketOpen() 4863 @*/ 4864 PetscErrorCode MatAssemblyEnd(Mat mat,MatAssemblyType type) 4865 { 4866 PetscErrorCode ierr; 4867 static PetscInt inassm = 0; 4868 PetscBool flg = PETSC_FALSE; 4869 4870 PetscFunctionBegin; 4871 PetscValidHeaderSpecific(mat,MAT_CLASSID,1); 4872 PetscValidType(mat,1); 4873 4874 inassm++; 4875 MatAssemblyEnd_InUse++; 4876 if (MatAssemblyEnd_InUse == 1) { /* Do the logging only the first time through */ 4877 ierr = PetscLogEventBegin(MAT_AssemblyEnd,mat,0,0,0);CHKERRQ(ierr); 4878 if (mat->ops->assemblyend) { 4879 ierr = (*mat->ops->assemblyend)(mat,type);CHKERRQ(ierr); 4880 } 4881 ierr = PetscLogEventEnd(MAT_AssemblyEnd,mat,0,0,0);CHKERRQ(ierr); 4882 } else { 4883 if (mat->ops->assemblyend) { 4884 ierr = (*mat->ops->assemblyend)(mat,type);CHKERRQ(ierr); 4885 } 4886 } 4887 4888 /* Flush assembly is not a true assembly */ 4889 if (type != MAT_FLUSH_ASSEMBLY) { 4890 mat->assembled = PETSC_TRUE; mat->num_ass++; 4891 } 4892 mat->insertmode = NOT_SET_VALUES; 4893 MatAssemblyEnd_InUse--; 4894 ierr = PetscObjectStateIncrease((PetscObject)mat);CHKERRQ(ierr); 4895 if (!mat->symmetric_eternal) { 4896 mat->symmetric_set = PETSC_FALSE; 4897 mat->hermitian_set = PETSC_FALSE; 4898 mat->structurally_symmetric_set = PETSC_FALSE; 4899 } 4900 #if defined(PETSC_HAVE_CUSP) 4901 if (mat->valid_GPU_matrix != PETSC_CUSP_UNALLOCATED) { 4902 mat->valid_GPU_matrix = PETSC_CUSP_CPU; 4903 } 4904 #endif 4905 if (inassm == 1 && type != MAT_FLUSH_ASSEMBLY) { 4906 ierr = MatView_Private(mat);CHKERRQ(ierr); 4907 ierr = PetscOptionsHasName(((PetscObject)mat)->prefix,"-mat_is_symmetric",&flg);CHKERRQ(ierr); 4908 if (flg) { 4909 PetscReal tol = 0.0; 4910 ierr = PetscOptionsGetReal(((PetscObject)mat)->prefix,"-mat_is_symmetric",&tol,PETSC_NULL);CHKERRQ(ierr); 4911 ierr = MatIsSymmetric(mat,tol,&flg);CHKERRQ(ierr); 4912 if (flg) { 4913 ierr = PetscPrintf(((PetscObject)mat)->comm,"Matrix is symmetric (tolerance %G)\n",tol);CHKERRQ(ierr); 4914 } else { 4915 ierr = PetscPrintf(((PetscObject)mat)->comm,"Matrix is not symmetric (tolerance %G)\n",tol);CHKERRQ(ierr); 4916 } 4917 } 4918 if (mat->nullsp) { 4919 ierr = PetscOptionsGetBool(PETSC_NULL,"-mat_null_space_test",&flg,PETSC_NULL);CHKERRQ(ierr); 4920 if (flg) { 4921 ierr = MatNullSpaceTest(mat->nullsp,mat,PETSC_NULL);CHKERRQ(ierr); 4922 } 4923 } 4924 } 4925 inassm--; 4926 PetscFunctionReturn(0); 4927 } 4928 4929 #undef __FUNCT__ 4930 #define __FUNCT__ "MatSetOption" 4931 /*@ 4932 MatSetOption - Sets a parameter option for a matrix. Some options 4933 may be specific to certain storage formats. Some options 4934 determine how values will be inserted (or added). Sorted, 4935 row-oriented input will generally assemble the fastest. The default 4936 is row-oriented. 4937 4938 Logically Collective on Mat 4939 4940 Input Parameters: 4941 + mat - the matrix 4942 . option - the option, one of those listed below (and possibly others), 4943 - flg - turn the option on (PETSC_TRUE) or off (PETSC_FALSE) 4944 4945 Options Describing Matrix Structure: 4946 + MAT_SPD - symmetric positive definite 4947 - MAT_SYMMETRIC - symmetric in terms of both structure and value 4948 . MAT_HERMITIAN - transpose is the complex conjugation 4949 . MAT_STRUCTURALLY_SYMMETRIC - symmetric nonzero structure 4950 - MAT_SYMMETRY_ETERNAL - if you would like the symmetry/Hermitian flag 4951 you set to be kept with all future use of the matrix 4952 including after MatAssemblyBegin/End() which could 4953 potentially change the symmetry structure, i.e. you 4954 KNOW the matrix will ALWAYS have the property you set. 4955 4956 4957 Options For Use with MatSetValues(): 4958 Insert a logically dense subblock, which can be 4959 . MAT_ROW_ORIENTED - row-oriented (default) 4960 4961 Note these options reflect the data you pass in with MatSetValues(); it has 4962 nothing to do with how the data is stored internally in the matrix 4963 data structure. 4964 4965 When (re)assembling a matrix, we can restrict the input for 4966 efficiency/debugging purposes. These options include 4967 + MAT_NEW_NONZERO_LOCATIONS - additional insertions will be 4968 allowed if they generate a new nonzero 4969 . MAT_NEW_DIAGONALS - new diagonals will be allowed (for block diagonal format only) 4970 . MAT_IGNORE_OFF_PROC_ENTRIES - drops off-processor entries 4971 . MAT_NEW_NONZERO_LOCATION_ERR - generates an error for new matrix entry 4972 . MAT_USE_HASH_TABLE - uses a hash table to speed up matrix assembly 4973 + MAT_NO_OFF_PROC_ENTRIES - you know each process will only set values for its own rows, will generate an error if 4974 any process sets values for another process. This avoids all reductions in the MatAssembly routines and thus improves 4975 performance for very large process counts. 4976 4977 Notes: 4978 Some options are relevant only for particular matrix types and 4979 are thus ignored by others. Other options are not supported by 4980 certain matrix types and will generate an error message if set. 4981 4982 If using a Fortran 77 module to compute a matrix, one may need to 4983 use the column-oriented option (or convert to the row-oriented 4984 format). 4985 4986 MAT_NEW_NONZERO_LOCATIONS set to PETSC_FALSE indicates that any add or insertion 4987 that would generate a new entry in the nonzero structure is instead 4988 ignored. Thus, if memory has not alredy been allocated for this particular 4989 data, then the insertion is ignored. For dense matrices, in which 4990 the entire array is allocated, no entries are ever ignored. 4991 Set after the first MatAssemblyEnd() 4992 4993 MAT_NEW_NONZERO_LOCATION_ERR indicates that any add or insertion 4994 that would generate a new entry in the nonzero structure instead produces 4995 an error. (Currently supported for AIJ and BAIJ formats only.) 4996 This is a useful flag when using SAME_NONZERO_PATTERN in calling 4997 KSPSetOperators() to ensure that the nonzero pattern truely does 4998 remain unchanged. Set after the first MatAssemblyEnd() 4999 5000 MAT_NEW_NONZERO_ALLOCATION_ERR indicates that any add or insertion 5001 that would generate a new entry that has not been preallocated will 5002 instead produce an error. (Currently supported for AIJ and BAIJ formats 5003 only.) This is a useful flag when debugging matrix memory preallocation. 5004 5005 MAT_IGNORE_OFF_PROC_ENTRIES indicates entries destined for 5006 other processors should be dropped, rather than stashed. 5007 This is useful if you know that the "owning" processor is also 5008 always generating the correct matrix entries, so that PETSc need 5009 not transfer duplicate entries generated on another processor. 5010 5011 MAT_USE_HASH_TABLE indicates that a hash table be used to improve the 5012 searches during matrix assembly. When this flag is set, the hash table 5013 is created during the first Matrix Assembly. This hash table is 5014 used the next time through, during MatSetVaules()/MatSetVaulesBlocked() 5015 to improve the searching of indices. MAT_NEW_NONZERO_LOCATIONS flag 5016 should be used with MAT_USE_HASH_TABLE flag. This option is currently 5017 supported by MATMPIBAIJ format only. 5018 5019 MAT_KEEP_NONZERO_PATTERN indicates when MatZeroRows() is called the zeroed entries 5020 are kept in the nonzero structure 5021 5022 MAT_IGNORE_ZERO_ENTRIES - for AIJ/IS matrices this will stop zero values from creating 5023 a zero location in the matrix 5024 5025 MAT_USE_INODES - indicates using inode version of the code - works with AIJ and 5026 ROWBS matrix types 5027 5028 MAT_NO_OFF_PROC_ZERO_ROWS - you know each process will only zero its own rows. This avoids all reductions in the 5029 zero row routines and thus improves performance for very large process counts. 5030 5031 MAT_IGNORE_LOWER_TRIANGULAR - For SBAIJ matrices will ignore any insertions you make in the lower triangular 5032 part of the matrix (since they should match the upper triangular part). 5033 5034 Notes: Can only be called after MatSetSizes() and MatSetType() have been set. 5035 5036 Level: intermediate 5037 5038 Concepts: matrices^setting options 5039 5040 @*/ 5041 PetscErrorCode MatSetOption(Mat mat,MatOption op,PetscBool flg) 5042 { 5043 PetscErrorCode ierr; 5044 5045 PetscFunctionBegin; 5046 PetscValidHeaderSpecific(mat,MAT_CLASSID,1); 5047 PetscValidType(mat,1); 5048 PetscValidLogicalCollectiveEnum(mat,op,2); 5049 PetscValidLogicalCollectiveBool(mat,flg,3); 5050 5051 if (((int) op) < 0 || ((int) op) >= NUM_MAT_OPTIONS) SETERRQ1(((PetscObject)mat)->comm,PETSC_ERR_ARG_OUTOFRANGE,"Options %d is out of range",(int)op); 5052 if (!((PetscObject)mat)->type_name) SETERRQ(((PetscObject)mat)->comm,PETSC_ERR_ARG_TYPENOTSET,"Cannot set options until type and size have been set, see MatSetType() and MatSetSizes()"); 5053 5054 switch (op) { 5055 case MAT_NO_OFF_PROC_ENTRIES: 5056 mat->nooffprocentries = flg; 5057 PetscFunctionReturn(0); 5058 break; 5059 case MAT_NO_OFF_PROC_ZERO_ROWS: 5060 mat->nooffproczerorows = flg; 5061 PetscFunctionReturn(0); 5062 break; 5063 case MAT_SPD: 5064 mat->spd_set = PETSC_TRUE; 5065 mat->spd = flg; 5066 if (flg) { 5067 mat->symmetric = PETSC_TRUE; 5068 mat->structurally_symmetric = PETSC_TRUE; 5069 mat->symmetric_set = PETSC_TRUE; 5070 mat->structurally_symmetric_set = PETSC_TRUE; 5071 } 5072 break; 5073 case MAT_SYMMETRIC: 5074 mat->symmetric = flg; 5075 if (flg) mat->structurally_symmetric = PETSC_TRUE; 5076 mat->symmetric_set = PETSC_TRUE; 5077 mat->structurally_symmetric_set = flg; 5078 break; 5079 case MAT_HERMITIAN: 5080 mat->hermitian = flg; 5081 if (flg) mat->structurally_symmetric = PETSC_TRUE; 5082 mat->hermitian_set = PETSC_TRUE; 5083 mat->structurally_symmetric_set = flg; 5084 break; 5085 case MAT_STRUCTURALLY_SYMMETRIC: 5086 mat->structurally_symmetric = flg; 5087 mat->structurally_symmetric_set = PETSC_TRUE; 5088 break; 5089 case MAT_SYMMETRY_ETERNAL: 5090 mat->symmetric_eternal = flg; 5091 break; 5092 default: 5093 break; 5094 } 5095 if (mat->ops->setoption) { 5096 ierr = (*mat->ops->setoption)(mat,op,flg);CHKERRQ(ierr); 5097 } 5098 PetscFunctionReturn(0); 5099 } 5100 5101 #undef __FUNCT__ 5102 #define __FUNCT__ "MatZeroEntries" 5103 /*@ 5104 MatZeroEntries - Zeros all entries of a matrix. For sparse matrices 5105 this routine retains the old nonzero structure. 5106 5107 Logically Collective on Mat 5108 5109 Input Parameters: 5110 . mat - the matrix 5111 5112 Level: intermediate 5113 5114 Notes: If the matrix was not preallocated then a default, likely poor preallocation will be set in the matrix, so this should be called after the preallocation phase. 5115 See the Performance chapter of the users manual for information on preallocating matrices. 5116 5117 Concepts: matrices^zeroing 5118 5119 .seealso: MatZeroRows() 5120 @*/ 5121 PetscErrorCode MatZeroEntries(Mat mat) 5122 { 5123 PetscErrorCode ierr; 5124 5125 PetscFunctionBegin; 5126 PetscValidHeaderSpecific(mat,MAT_CLASSID,1); 5127 PetscValidType(mat,1); 5128 if (mat->factortype) SETERRQ(((PetscObject)mat)->comm,PETSC_ERR_ARG_WRONGSTATE,"Not for factored matrix"); 5129 if (mat->insertmode != NOT_SET_VALUES) SETERRQ(PETSC_COMM_SELF,PETSC_ERR_ARG_WRONGSTATE,"Not for matrices where you have set values but not yet assembled"); 5130 if (!mat->ops->zeroentries) SETERRQ1(((PetscObject)mat)->comm,PETSC_ERR_SUP,"Mat type %s",((PetscObject)mat)->type_name); 5131 MatCheckPreallocated(mat,1); 5132 5133 ierr = PetscLogEventBegin(MAT_ZeroEntries,mat,0,0,0);CHKERRQ(ierr); 5134 ierr = (*mat->ops->zeroentries)(mat);CHKERRQ(ierr); 5135 ierr = PetscLogEventEnd(MAT_ZeroEntries,mat,0,0,0);CHKERRQ(ierr); 5136 ierr = PetscObjectStateIncrease((PetscObject)mat);CHKERRQ(ierr); 5137 #if defined(PETSC_HAVE_CUSP) 5138 if (mat->valid_GPU_matrix != PETSC_CUSP_UNALLOCATED) { 5139 mat->valid_GPU_matrix = PETSC_CUSP_CPU; 5140 } 5141 #endif 5142 PetscFunctionReturn(0); 5143 } 5144 5145 #undef __FUNCT__ 5146 #define __FUNCT__ "MatZeroRowsColumns" 5147 /*@C 5148 MatZeroRowsColumns - Zeros all entries (except possibly the main diagonal) 5149 of a set of rows and columns of a matrix. 5150 5151 Collective on Mat 5152 5153 Input Parameters: 5154 + mat - the matrix 5155 . numRows - the number of rows to remove 5156 . rows - the global row indices 5157 . diag - value put in all diagonals of eliminated rows (0.0 will even eliminate diagonal entry) 5158 . x - optional vector of solutions for zeroed rows (other entries in vector are not used) 5159 - b - optional vector of right hand side, that will be adjusted by provided solution 5160 5161 Notes: 5162 This does not change the nonzero structure of the matrix, it merely zeros those entries in the matrix. 5163 5164 The user can set a value in the diagonal entry (or for the AIJ and 5165 row formats can optionally remove the main diagonal entry from the 5166 nonzero structure as well, by passing 0.0 as the final argument). 5167 5168 For the parallel case, all processes that share the matrix (i.e., 5169 those in the communicator used for matrix creation) MUST call this 5170 routine, regardless of whether any rows being zeroed are owned by 5171 them. 5172 5173 Each processor can indicate any rows in the entire matrix to be zeroed (i.e. each process does NOT have to 5174 list only rows local to itself). 5175 5176 The option MAT_NO_OFF_PROC_ZERO_ROWS does not apply to this routine. 5177 5178 Level: intermediate 5179 5180 Concepts: matrices^zeroing rows 5181 5182 .seealso: MatZeroRowsIS(), MatZeroRowsStencil(), MatZeroEntries(), MatZeroRowsLocal(), MatSetOption(), MatZeroRowsColumnsIS() 5183 @*/ 5184 PetscErrorCode MatZeroRowsColumns(Mat mat,PetscInt numRows,const PetscInt rows[],PetscScalar diag,Vec x,Vec b) 5185 { 5186 PetscErrorCode ierr; 5187 5188 PetscFunctionBegin; 5189 PetscValidHeaderSpecific(mat,MAT_CLASSID,1); 5190 PetscValidType(mat,1); 5191 if (numRows) PetscValidIntPointer(rows,3); 5192 if (!mat->assembled) SETERRQ(((PetscObject)mat)->comm,PETSC_ERR_ARG_WRONGSTATE,"Not for unassembled matrix"); 5193 if (mat->factortype) SETERRQ(((PetscObject)mat)->comm,PETSC_ERR_ARG_WRONGSTATE,"Not for factored matrix"); 5194 if (!mat->ops->zerorowscolumns) SETERRQ1(((PetscObject)mat)->comm,PETSC_ERR_SUP,"Mat type %s",((PetscObject)mat)->type_name); 5195 MatCheckPreallocated(mat,1); 5196 5197 ierr = (*mat->ops->zerorowscolumns)(mat,numRows,rows,diag,x,b);CHKERRQ(ierr); 5198 ierr = MatView_Private(mat);CHKERRQ(ierr); 5199 ierr = PetscObjectStateIncrease((PetscObject)mat);CHKERRQ(ierr); 5200 #if defined(PETSC_HAVE_CUSP) 5201 if (mat->valid_GPU_matrix != PETSC_CUSP_UNALLOCATED) { 5202 mat->valid_GPU_matrix = PETSC_CUSP_CPU; 5203 } 5204 #endif 5205 PetscFunctionReturn(0); 5206 } 5207 5208 #undef __FUNCT__ 5209 #define __FUNCT__ "MatZeroRowsColumnsIS" 5210 /*@C 5211 MatZeroRowsColumnsIS - Zeros all entries (except possibly the main diagonal) 5212 of a set of rows and columns of a matrix. 5213 5214 Collective on Mat 5215 5216 Input Parameters: 5217 + mat - the matrix 5218 . is - the rows to zero 5219 . diag - value put in all diagonals of eliminated rows (0.0 will even eliminate diagonal entry) 5220 . x - optional vector of solutions for zeroed rows (other entries in vector are not used) 5221 - b - optional vector of right hand side, that will be adjusted by provided solution 5222 5223 Notes: 5224 This does not change the nonzero structure of the matrix, it merely zeros those entries in the matrix. 5225 5226 The user can set a value in the diagonal entry (or for the AIJ and 5227 row formats can optionally remove the main diagonal entry from the 5228 nonzero structure as well, by passing 0.0 as the final argument). 5229 5230 For the parallel case, all processes that share the matrix (i.e., 5231 those in the communicator used for matrix creation) MUST call this 5232 routine, regardless of whether any rows being zeroed are owned by 5233 them. 5234 5235 Each processor can indicate any rows in the entire matrix to be zeroed (i.e. each process does NOT have to 5236 list only rows local to itself). 5237 5238 The option MAT_NO_OFF_PROC_ZERO_ROWS does not apply to this routine. 5239 5240 Level: intermediate 5241 5242 Concepts: matrices^zeroing rows 5243 5244 .seealso: MatZeroRowsIS(), MatZeroRowsStencil(), MatZeroEntries(), MatZeroRowsLocal(), MatSetOption(), MatZeroRowsColumns() 5245 @*/ 5246 PetscErrorCode MatZeroRowsColumnsIS(Mat mat,IS is,PetscScalar diag,Vec x,Vec b) 5247 { 5248 PetscErrorCode ierr; 5249 PetscInt numRows; 5250 const PetscInt *rows; 5251 5252 PetscFunctionBegin; 5253 PetscValidHeaderSpecific(mat,MAT_CLASSID,1); 5254 PetscValidHeaderSpecific(is,IS_CLASSID,2); 5255 PetscValidType(mat,1); 5256 PetscValidType(is,2); 5257 ierr = ISGetLocalSize(is,&numRows);CHKERRQ(ierr); 5258 ierr = ISGetIndices(is,&rows);CHKERRQ(ierr); 5259 ierr = MatZeroRowsColumns(mat,numRows,rows,diag,x,b);CHKERRQ(ierr); 5260 ierr = ISRestoreIndices(is,&rows);CHKERRQ(ierr); 5261 PetscFunctionReturn(0); 5262 } 5263 5264 #undef __FUNCT__ 5265 #define __FUNCT__ "MatZeroRows" 5266 /*@C 5267 MatZeroRows - Zeros all entries (except possibly the main diagonal) 5268 of a set of rows of a matrix. 5269 5270 Collective on Mat 5271 5272 Input Parameters: 5273 + mat - the matrix 5274 . numRows - the number of rows to remove 5275 . rows - the global row indices 5276 . diag - value put in all diagonals of eliminated rows (0.0 will even eliminate diagonal entry) 5277 . x - optional vector of solutions for zeroed rows (other entries in vector are not used) 5278 - b - optional vector of right hand side, that will be adjusted by provided solution 5279 5280 Notes: 5281 For the AIJ and BAIJ matrix formats this removes the old nonzero structure, 5282 but does not release memory. For the dense and block diagonal 5283 formats this does not alter the nonzero structure. 5284 5285 If the option MatSetOption(mat,MAT_KEEP_NONZERO_PATTERN,PETSC_TRUE) the nonzero structure 5286 of the matrix is not changed (even for AIJ and BAIJ matrices) the values are 5287 merely zeroed. 5288 5289 The user can set a value in the diagonal entry (or for the AIJ and 5290 row formats can optionally remove the main diagonal entry from the 5291 nonzero structure as well, by passing 0.0 as the final argument). 5292 5293 For the parallel case, all processes that share the matrix (i.e., 5294 those in the communicator used for matrix creation) MUST call this 5295 routine, regardless of whether any rows being zeroed are owned by 5296 them. 5297 5298 Each processor can indicate any rows in the entire matrix to be zeroed (i.e. each process does NOT have to 5299 list only rows local to itself). 5300 5301 You can call MatSetOption(mat,MAT_NO_OFF_PROC_ZERO_ROWS,PETSC_TRUE) if each process indicates only rows it 5302 owns that are to be zeroed. This saves a global synchronization in the implementation. 5303 5304 Level: intermediate 5305 5306 Concepts: matrices^zeroing rows 5307 5308 .seealso: MatZeroRowsIS(), MatZeroRowsStencil(), MatZeroEntries(), MatZeroRowsLocal(), MatSetOption() 5309 @*/ 5310 PetscErrorCode MatZeroRows(Mat mat,PetscInt numRows,const PetscInt rows[],PetscScalar diag,Vec x,Vec b) 5311 { 5312 PetscErrorCode ierr; 5313 5314 PetscFunctionBegin; 5315 PetscValidHeaderSpecific(mat,MAT_CLASSID,1); 5316 PetscValidType(mat,1); 5317 if (numRows) PetscValidIntPointer(rows,3); 5318 if (!mat->assembled) SETERRQ(((PetscObject)mat)->comm,PETSC_ERR_ARG_WRONGSTATE,"Not for unassembled matrix"); 5319 if (mat->factortype) SETERRQ(((PetscObject)mat)->comm,PETSC_ERR_ARG_WRONGSTATE,"Not for factored matrix"); 5320 if (!mat->ops->zerorows) SETERRQ1(((PetscObject)mat)->comm,PETSC_ERR_SUP,"Mat type %s",((PetscObject)mat)->type_name); 5321 MatCheckPreallocated(mat,1); 5322 5323 ierr = (*mat->ops->zerorows)(mat,numRows,rows,diag,x,b);CHKERRQ(ierr); 5324 ierr = MatView_Private(mat);CHKERRQ(ierr); 5325 ierr = PetscObjectStateIncrease((PetscObject)mat);CHKERRQ(ierr); 5326 #if defined(PETSC_HAVE_CUSP) 5327 if (mat->valid_GPU_matrix != PETSC_CUSP_UNALLOCATED) { 5328 mat->valid_GPU_matrix = PETSC_CUSP_CPU; 5329 } 5330 #endif 5331 PetscFunctionReturn(0); 5332 } 5333 5334 #undef __FUNCT__ 5335 #define __FUNCT__ "MatZeroRowsIS" 5336 /*@C 5337 MatZeroRowsIS - Zeros all entries (except possibly the main diagonal) 5338 of a set of rows of a matrix. 5339 5340 Collective on Mat 5341 5342 Input Parameters: 5343 + mat - the matrix 5344 . is - index set of rows to remove 5345 . diag - value put in all diagonals of eliminated rows 5346 . x - optional vector of solutions for zeroed rows (other entries in vector are not used) 5347 - b - optional vector of right hand side, that will be adjusted by provided solution 5348 5349 Notes: 5350 For the AIJ and BAIJ matrix formats this removes the old nonzero structure, 5351 but does not release memory. For the dense and block diagonal 5352 formats this does not alter the nonzero structure. 5353 5354 If the option MatSetOption(mat,MAT_KEEP_NONZERO_PATTERN,PETSC_TRUE) the nonzero structure 5355 of the matrix is not changed (even for AIJ and BAIJ matrices) the values are 5356 merely zeroed. 5357 5358 The user can set a value in the diagonal entry (or for the AIJ and 5359 row formats can optionally remove the main diagonal entry from the 5360 nonzero structure as well, by passing 0.0 as the final argument). 5361 5362 For the parallel case, all processes that share the matrix (i.e., 5363 those in the communicator used for matrix creation) MUST call this 5364 routine, regardless of whether any rows being zeroed are owned by 5365 them. 5366 5367 Each processor can indicate any rows in the entire matrix to be zeroed (i.e. each process does NOT have to 5368 list only rows local to itself). 5369 5370 You can call MatSetOption(mat,MAT_NO_OFF_PROC_ZERO_ROWS,PETSC_TRUE) if each process indicates only rows it 5371 owns that are to be zeroed. This saves a global synchronization in the implementation. 5372 5373 Level: intermediate 5374 5375 Concepts: matrices^zeroing rows 5376 5377 .seealso: MatZeroRows(), MatZeroRowsStencil(), MatZeroEntries(), MatZeroRowsLocal(), MatSetOption() 5378 @*/ 5379 PetscErrorCode MatZeroRowsIS(Mat mat,IS is,PetscScalar diag,Vec x,Vec b) 5380 { 5381 PetscInt numRows; 5382 const PetscInt *rows; 5383 PetscErrorCode ierr; 5384 5385 PetscFunctionBegin; 5386 PetscValidHeaderSpecific(mat,MAT_CLASSID,1); 5387 PetscValidType(mat,1); 5388 PetscValidHeaderSpecific(is,IS_CLASSID,2); 5389 ierr = ISGetLocalSize(is,&numRows);CHKERRQ(ierr); 5390 ierr = ISGetIndices(is,&rows);CHKERRQ(ierr); 5391 ierr = MatZeroRows(mat,numRows,rows,diag,x,b);CHKERRQ(ierr); 5392 ierr = ISRestoreIndices(is,&rows);CHKERRQ(ierr); 5393 PetscFunctionReturn(0); 5394 } 5395 5396 #undef __FUNCT__ 5397 #define __FUNCT__ "MatZeroRowsStencil" 5398 /*@C 5399 MatZeroRowsStencil - Zeros all entries (except possibly the main diagonal) 5400 of a set of rows of a matrix. These rows must be local to the process. 5401 5402 Collective on Mat 5403 5404 Input Parameters: 5405 + mat - the matrix 5406 . numRows - the number of rows to remove 5407 . rows - the grid coordinates (and component number when dof > 1) for matrix rows 5408 . diag - value put in all diagonals of eliminated rows (0.0 will even eliminate diagonal entry) 5409 . x - optional vector of solutions for zeroed rows (other entries in vector are not used) 5410 - b - optional vector of right hand side, that will be adjusted by provided solution 5411 5412 Notes: 5413 For the AIJ and BAIJ matrix formats this removes the old nonzero structure, 5414 but does not release memory. For the dense and block diagonal 5415 formats this does not alter the nonzero structure. 5416 5417 If the option MatSetOption(mat,MAT_KEEP_NONZERO_PATTERN,PETSC_TRUE) the nonzero structure 5418 of the matrix is not changed (even for AIJ and BAIJ matrices) the values are 5419 merely zeroed. 5420 5421 The user can set a value in the diagonal entry (or for the AIJ and 5422 row formats can optionally remove the main diagonal entry from the 5423 nonzero structure as well, by passing 0.0 as the final argument). 5424 5425 For the parallel case, all processes that share the matrix (i.e., 5426 those in the communicator used for matrix creation) MUST call this 5427 routine, regardless of whether any rows being zeroed are owned by 5428 them. 5429 5430 Each processor can indicate any rows in the entire matrix to be zeroed (i.e. each process does NOT have to 5431 list only rows local to itself). 5432 5433 The grid coordinates are across the entire grid, not just the local portion 5434 5435 In Fortran idxm and idxn should be declared as 5436 $ MatStencil idxm(4,m) 5437 and the values inserted using 5438 $ idxm(MatStencil_i,1) = i 5439 $ idxm(MatStencil_j,1) = j 5440 $ idxm(MatStencil_k,1) = k 5441 $ idxm(MatStencil_c,1) = c 5442 etc 5443 5444 For periodic boundary conditions use negative indices for values to the left (below 0; that are to be 5445 obtained by wrapping values from right edge). For values to the right of the last entry using that index plus one 5446 etc to obtain values that obtained by wrapping the values from the left edge. This does not work for anything but the 5447 DMDA_BOUNDARY_PERIODIC boundary type. 5448 5449 For indices that don't mean anything for your case (like the k index when working in 2d) or the c index when you have 5450 a single value per point) you can skip filling those indices. 5451 5452 Level: intermediate 5453 5454 Concepts: matrices^zeroing rows 5455 5456 .seealso: MatZeroRows(), MatZeroRowsIS(), MatZeroEntries(), MatZeroRowsLocal(), MatSetOption() 5457 @*/ 5458 PetscErrorCode MatZeroRowsStencil(Mat mat,PetscInt numRows,const MatStencil rows[],PetscScalar diag,Vec x,Vec b) 5459 { 5460 PetscInt dim = mat->stencil.dim; 5461 PetscInt sdim = dim - (1 - (PetscInt) mat->stencil.noc); 5462 PetscInt *dims = mat->stencil.dims+1; 5463 PetscInt *starts = mat->stencil.starts; 5464 PetscInt *dxm = (PetscInt *) rows; 5465 PetscInt *jdxm, i, j, tmp, numNewRows = 0; 5466 PetscErrorCode ierr; 5467 5468 PetscFunctionBegin; 5469 PetscValidHeaderSpecific(mat,MAT_CLASSID,1); 5470 PetscValidType(mat,1); 5471 if (numRows) PetscValidIntPointer(rows,3); 5472 5473 ierr = PetscMalloc(numRows*sizeof(PetscInt), &jdxm);CHKERRQ(ierr); 5474 for(i = 0; i < numRows; ++i) { 5475 /* Skip unused dimensions (they are ordered k, j, i, c) */ 5476 for(j = 0; j < 3-sdim; ++j) dxm++; 5477 /* Local index in X dir */ 5478 tmp = *dxm++ - starts[0]; 5479 /* Loop over remaining dimensions */ 5480 for(j = 0; j < dim-1; ++j) { 5481 /* If nonlocal, set index to be negative */ 5482 if ((*dxm++ - starts[j+1]) < 0 || tmp < 0) tmp = PETSC_MIN_INT; 5483 /* Update local index */ 5484 else tmp = tmp*dims[j] + *(dxm-1) - starts[j+1]; 5485 } 5486 /* Skip component slot if necessary */ 5487 if (mat->stencil.noc) dxm++; 5488 /* Local row number */ 5489 if (tmp >= 0) { 5490 jdxm[numNewRows++] = tmp; 5491 } 5492 } 5493 ierr = MatZeroRowsLocal(mat,numNewRows,jdxm,diag,x,b);CHKERRQ(ierr); 5494 ierr = PetscFree(jdxm);CHKERRQ(ierr); 5495 PetscFunctionReturn(0); 5496 } 5497 5498 #undef __FUNCT__ 5499 #define __FUNCT__ "MatZeroRowsColumnsStencil" 5500 /*@C 5501 MatZeroRowsColumnsStencil - Zeros all row and column entries (except possibly the main diagonal) 5502 of a set of rows and columns of a matrix. 5503 5504 Collective on Mat 5505 5506 Input Parameters: 5507 + mat - the matrix 5508 . numRows - the number of rows/columns to remove 5509 . rows - the grid coordinates (and component number when dof > 1) for matrix rows 5510 . diag - value put in all diagonals of eliminated rows (0.0 will even eliminate diagonal entry) 5511 . x - optional vector of solutions for zeroed rows (other entries in vector are not used) 5512 - b - optional vector of right hand side, that will be adjusted by provided solution 5513 5514 Notes: 5515 For the AIJ and BAIJ matrix formats this removes the old nonzero structure, 5516 but does not release memory. For the dense and block diagonal 5517 formats this does not alter the nonzero structure. 5518 5519 If the option MatSetOption(mat,MAT_KEEP_NONZERO_PATTERN,PETSC_TRUE) the nonzero structure 5520 of the matrix is not changed (even for AIJ and BAIJ matrices) the values are 5521 merely zeroed. 5522 5523 The user can set a value in the diagonal entry (or for the AIJ and 5524 row formats can optionally remove the main diagonal entry from the 5525 nonzero structure as well, by passing 0.0 as the final argument). 5526 5527 For the parallel case, all processes that share the matrix (i.e., 5528 those in the communicator used for matrix creation) MUST call this 5529 routine, regardless of whether any rows being zeroed are owned by 5530 them. 5531 5532 Each processor can indicate any rows in the entire matrix to be zeroed (i.e. each process does NOT have to 5533 list only rows local to itself, but the row/column numbers are given in local numbering). 5534 5535 The grid coordinates are across the entire grid, not just the local portion 5536 5537 In Fortran idxm and idxn should be declared as 5538 $ MatStencil idxm(4,m) 5539 and the values inserted using 5540 $ idxm(MatStencil_i,1) = i 5541 $ idxm(MatStencil_j,1) = j 5542 $ idxm(MatStencil_k,1) = k 5543 $ idxm(MatStencil_c,1) = c 5544 etc 5545 5546 For periodic boundary conditions use negative indices for values to the left (below 0; that are to be 5547 obtained by wrapping values from right edge). For values to the right of the last entry using that index plus one 5548 etc to obtain values that obtained by wrapping the values from the left edge. This does not work for anything but the 5549 DMDA_BOUNDARY_PERIODIC boundary type. 5550 5551 For indices that don't mean anything for your case (like the k index when working in 2d) or the c index when you have 5552 a single value per point) you can skip filling those indices. 5553 5554 Level: intermediate 5555 5556 Concepts: matrices^zeroing rows 5557 5558 .seealso: MatZeroRows(), MatZeroRowsIS(), MatZeroEntries(), MatZeroRowsLocal(), MatSetOption() 5559 @*/ 5560 PetscErrorCode MatZeroRowsColumnsStencil(Mat mat,PetscInt numRows,const MatStencil rows[],PetscScalar diag,Vec x,Vec b) 5561 { 5562 PetscInt dim = mat->stencil.dim; 5563 PetscInt sdim = dim - (1 - (PetscInt) mat->stencil.noc); 5564 PetscInt *dims = mat->stencil.dims+1; 5565 PetscInt *starts = mat->stencil.starts; 5566 PetscInt *dxm = (PetscInt *) rows; 5567 PetscInt *jdxm, i, j, tmp, numNewRows = 0; 5568 PetscErrorCode ierr; 5569 5570 PetscFunctionBegin; 5571 PetscValidHeaderSpecific(mat,MAT_CLASSID,1); 5572 PetscValidType(mat,1); 5573 if (numRows) PetscValidIntPointer(rows,3); 5574 5575 ierr = PetscMalloc(numRows*sizeof(PetscInt), &jdxm);CHKERRQ(ierr); 5576 for(i = 0; i < numRows; ++i) { 5577 /* Skip unused dimensions (they are ordered k, j, i, c) */ 5578 for(j = 0; j < 3-sdim; ++j) dxm++; 5579 /* Local index in X dir */ 5580 tmp = *dxm++ - starts[0]; 5581 /* Loop over remaining dimensions */ 5582 for(j = 0; j < dim-1; ++j) { 5583 /* If nonlocal, set index to be negative */ 5584 if ((*dxm++ - starts[j+1]) < 0 || tmp < 0) tmp = PETSC_MIN_INT; 5585 /* Update local index */ 5586 else tmp = tmp*dims[j] + *(dxm-1) - starts[j+1]; 5587 } 5588 /* Skip component slot if necessary */ 5589 if (mat->stencil.noc) dxm++; 5590 /* Local row number */ 5591 if (tmp >= 0) { 5592 jdxm[numNewRows++] = tmp; 5593 } 5594 } 5595 ierr = MatZeroRowsColumnsLocal(mat,numNewRows,jdxm,diag,x,b);CHKERRQ(ierr); 5596 ierr = PetscFree(jdxm);CHKERRQ(ierr); 5597 PetscFunctionReturn(0); 5598 } 5599 5600 #undef __FUNCT__ 5601 #define __FUNCT__ "MatZeroRowsLocal" 5602 /*@C 5603 MatZeroRowsLocal - Zeros all entries (except possibly the main diagonal) 5604 of a set of rows of a matrix; using local numbering of rows. 5605 5606 Collective on Mat 5607 5608 Input Parameters: 5609 + mat - the matrix 5610 . numRows - the number of rows to remove 5611 . rows - the global row indices 5612 . diag - value put in all diagonals of eliminated rows 5613 . x - optional vector of solutions for zeroed rows (other entries in vector are not used) 5614 - b - optional vector of right hand side, that will be adjusted by provided solution 5615 5616 Notes: 5617 Before calling MatZeroRowsLocal(), the user must first set the 5618 local-to-global mapping by calling MatSetLocalToGlobalMapping(). 5619 5620 For the AIJ matrix formats this removes the old nonzero structure, 5621 but does not release memory. For the dense and block diagonal 5622 formats this does not alter the nonzero structure. 5623 5624 If the option MatSetOption(mat,MAT_KEEP_NONZERO_PATTERN,PETSC_TRUE) the nonzero structure 5625 of the matrix is not changed (even for AIJ and BAIJ matrices) the values are 5626 merely zeroed. 5627 5628 The user can set a value in the diagonal entry (or for the AIJ and 5629 row formats can optionally remove the main diagonal entry from the 5630 nonzero structure as well, by passing 0.0 as the final argument). 5631 5632 You can call MatSetOption(mat,MAT_NO_OFF_PROC_ZERO_ROWS,PETSC_TRUE) if each process indicates only rows it 5633 owns that are to be zeroed. This saves a global synchronization in the implementation. 5634 5635 Level: intermediate 5636 5637 Concepts: matrices^zeroing 5638 5639 .seealso: MatZeroRows(), MatZeroRowsLocalIS(), MatZeroEntries(), MatZeroRows(), MatSetLocalToGlobalMapping 5640 @*/ 5641 PetscErrorCode MatZeroRowsLocal(Mat mat,PetscInt numRows,const PetscInt rows[],PetscScalar diag,Vec x,Vec b) 5642 { 5643 PetscErrorCode ierr; 5644 PetscMPIInt size; 5645 5646 PetscFunctionBegin; 5647 PetscValidHeaderSpecific(mat,MAT_CLASSID,1); 5648 PetscValidType(mat,1); 5649 if (numRows) PetscValidIntPointer(rows,3); 5650 if (!mat->assembled) SETERRQ(((PetscObject)mat)->comm,PETSC_ERR_ARG_WRONGSTATE,"Not for unassembled matrix"); 5651 if (mat->factortype) SETERRQ(((PetscObject)mat)->comm,PETSC_ERR_ARG_WRONGSTATE,"Not for factored matrix"); 5652 MatCheckPreallocated(mat,1); 5653 5654 ierr = MPI_Comm_size(((PetscObject)mat)->comm,&size);CHKERRQ(ierr); 5655 if (mat->ops->zerorowslocal) { 5656 ierr = (*mat->ops->zerorowslocal)(mat,numRows,rows,diag,x,b);CHKERRQ(ierr); 5657 } else if (size == 1) { 5658 ierr = (*mat->ops->zerorows)(mat,numRows,rows,diag,x,b);CHKERRQ(ierr); 5659 } else { 5660 IS is, newis; 5661 const PetscInt *newRows; 5662 5663 if (!mat->rmap->mapping) SETERRQ(((PetscObject)mat)->comm,PETSC_ERR_ARG_WRONGSTATE,"Need to provide local to global mapping to matrix first"); 5664 ierr = ISCreateGeneral(PETSC_COMM_SELF,numRows,rows,PETSC_COPY_VALUES,&is);CHKERRQ(ierr); 5665 ierr = ISLocalToGlobalMappingApplyIS(mat->rmap->mapping,is,&newis);CHKERRQ(ierr); 5666 ierr = ISGetIndices(newis,&newRows);CHKERRQ(ierr); 5667 ierr = (*mat->ops->zerorows)(mat,numRows,newRows,diag,x,b);CHKERRQ(ierr); 5668 ierr = ISRestoreIndices(newis,&newRows);CHKERRQ(ierr); 5669 ierr = ISDestroy(&newis);CHKERRQ(ierr); 5670 ierr = ISDestroy(&is);CHKERRQ(ierr); 5671 } 5672 ierr = PetscObjectStateIncrease((PetscObject)mat);CHKERRQ(ierr); 5673 #if defined(PETSC_HAVE_CUSP) 5674 if (mat->valid_GPU_matrix != PETSC_CUSP_UNALLOCATED) { 5675 mat->valid_GPU_matrix = PETSC_CUSP_CPU; 5676 } 5677 #endif 5678 PetscFunctionReturn(0); 5679 } 5680 5681 #undef __FUNCT__ 5682 #define __FUNCT__ "MatZeroRowsLocalIS" 5683 /*@C 5684 MatZeroRowsLocalIS - Zeros all entries (except possibly the main diagonal) 5685 of a set of rows of a matrix; using local numbering of rows. 5686 5687 Collective on Mat 5688 5689 Input Parameters: 5690 + mat - the matrix 5691 . is - index set of rows to remove 5692 . diag - value put in all diagonals of eliminated rows 5693 . x - optional vector of solutions for zeroed rows (other entries in vector are not used) 5694 - b - optional vector of right hand side, that will be adjusted by provided solution 5695 5696 Notes: 5697 Before calling MatZeroRowsLocalIS(), the user must first set the 5698 local-to-global mapping by calling MatSetLocalToGlobalMapping(). 5699 5700 For the AIJ matrix formats this removes the old nonzero structure, 5701 but does not release memory. For the dense and block diagonal 5702 formats this does not alter the nonzero structure. 5703 5704 If the option MatSetOption(mat,MAT_KEEP_NONZERO_PATTERN,PETSC_TRUE) the nonzero structure 5705 of the matrix is not changed (even for AIJ and BAIJ matrices) the values are 5706 merely zeroed. 5707 5708 The user can set a value in the diagonal entry (or for the AIJ and 5709 row formats can optionally remove the main diagonal entry from the 5710 nonzero structure as well, by passing 0.0 as the final argument). 5711 5712 You can call MatSetOption(mat,MAT_NO_OFF_PROC_ZERO_ROWS,PETSC_TRUE) if each process indicates only rows it 5713 owns that are to be zeroed. This saves a global synchronization in the implementation. 5714 5715 Level: intermediate 5716 5717 Concepts: matrices^zeroing 5718 5719 .seealso: MatZeroRows(), MatZeroRowsLocal(), MatZeroEntries(), MatZeroRows(), MatSetLocalToGlobalMapping 5720 @*/ 5721 PetscErrorCode MatZeroRowsLocalIS(Mat mat,IS is,PetscScalar diag,Vec x,Vec b) 5722 { 5723 PetscErrorCode ierr; 5724 PetscInt numRows; 5725 const PetscInt *rows; 5726 5727 PetscFunctionBegin; 5728 PetscValidHeaderSpecific(mat,MAT_CLASSID,1); 5729 PetscValidType(mat,1); 5730 PetscValidHeaderSpecific(is,IS_CLASSID,2); 5731 if (!mat->assembled) SETERRQ(((PetscObject)mat)->comm,PETSC_ERR_ARG_WRONGSTATE,"Not for unassembled matrix"); 5732 if (mat->factortype) SETERRQ(((PetscObject)mat)->comm,PETSC_ERR_ARG_WRONGSTATE,"Not for factored matrix"); 5733 MatCheckPreallocated(mat,1); 5734 5735 ierr = ISGetLocalSize(is,&numRows);CHKERRQ(ierr); 5736 ierr = ISGetIndices(is,&rows);CHKERRQ(ierr); 5737 ierr = MatZeroRowsLocal(mat,numRows,rows,diag,x,b);CHKERRQ(ierr); 5738 ierr = ISRestoreIndices(is,&rows);CHKERRQ(ierr); 5739 PetscFunctionReturn(0); 5740 } 5741 5742 #undef __FUNCT__ 5743 #define __FUNCT__ "MatZeroRowsColumnsLocal" 5744 /*@C 5745 MatZeroRowsColumnsLocal - Zeros all entries (except possibly the main diagonal) 5746 of a set of rows and columns of a matrix; using local numbering of rows. 5747 5748 Collective on Mat 5749 5750 Input Parameters: 5751 + mat - the matrix 5752 . numRows - the number of rows to remove 5753 . rows - the global row indices 5754 . diag - value put in all diagonals of eliminated rows 5755 . x - optional vector of solutions for zeroed rows (other entries in vector are not used) 5756 - b - optional vector of right hand side, that will be adjusted by provided solution 5757 5758 Notes: 5759 Before calling MatZeroRowsColumnsLocal(), the user must first set the 5760 local-to-global mapping by calling MatSetLocalToGlobalMapping(). 5761 5762 The user can set a value in the diagonal entry (or for the AIJ and 5763 row formats can optionally remove the main diagonal entry from the 5764 nonzero structure as well, by passing 0.0 as the final argument). 5765 5766 Level: intermediate 5767 5768 Concepts: matrices^zeroing 5769 5770 .seealso: MatZeroRows(), MatZeroRowsLocalIS(), MatZeroEntries(), MatZeroRows(), MatSetLocalToGlobalMapping 5771 @*/ 5772 PetscErrorCode MatZeroRowsColumnsLocal(Mat mat,PetscInt numRows,const PetscInt rows[],PetscScalar diag,Vec x,Vec b) 5773 { 5774 PetscErrorCode ierr; 5775 PetscMPIInt size; 5776 5777 PetscFunctionBegin; 5778 PetscValidHeaderSpecific(mat,MAT_CLASSID,1); 5779 PetscValidType(mat,1); 5780 if (numRows) PetscValidIntPointer(rows,3); 5781 if (!mat->assembled) SETERRQ(((PetscObject)mat)->comm,PETSC_ERR_ARG_WRONGSTATE,"Not for unassembled matrix"); 5782 if (mat->factortype) SETERRQ(((PetscObject)mat)->comm,PETSC_ERR_ARG_WRONGSTATE,"Not for factored matrix"); 5783 MatCheckPreallocated(mat,1); 5784 5785 ierr = MPI_Comm_size(((PetscObject)mat)->comm,&size);CHKERRQ(ierr); 5786 if (size == 1) { 5787 ierr = (*mat->ops->zerorowscolumns)(mat,numRows,rows,diag,x,b);CHKERRQ(ierr); 5788 } else { 5789 IS is, newis; 5790 const PetscInt *newRows; 5791 5792 if (!mat->cmap->mapping) SETERRQ(((PetscObject)mat)->comm,PETSC_ERR_ARG_WRONGSTATE,"Need to provide local to global mapping to matrix first"); 5793 ierr = ISCreateGeneral(PETSC_COMM_SELF,numRows,rows,PETSC_COPY_VALUES,&is);CHKERRQ(ierr); 5794 ierr = ISLocalToGlobalMappingApplyIS(mat->cmap->mapping,is,&newis);CHKERRQ(ierr); 5795 ierr = ISGetIndices(newis,&newRows);CHKERRQ(ierr); 5796 ierr = (*mat->ops->zerorowscolumns)(mat,numRows,newRows,diag,x,b);CHKERRQ(ierr); 5797 ierr = ISRestoreIndices(newis,&newRows);CHKERRQ(ierr); 5798 ierr = ISDestroy(&newis);CHKERRQ(ierr); 5799 ierr = ISDestroy(&is);CHKERRQ(ierr); 5800 } 5801 ierr = PetscObjectStateIncrease((PetscObject)mat);CHKERRQ(ierr); 5802 #if defined(PETSC_HAVE_CUSP) 5803 if (mat->valid_GPU_matrix != PETSC_CUSP_UNALLOCATED) { 5804 mat->valid_GPU_matrix = PETSC_CUSP_CPU; 5805 } 5806 #endif 5807 PetscFunctionReturn(0); 5808 } 5809 5810 #undef __FUNCT__ 5811 #define __FUNCT__ "MatZeroRowsColumnsLocalIS" 5812 /*@C 5813 MatZeroRowsColumnsLocalIS - Zeros all entries (except possibly the main diagonal) 5814 of a set of rows and columns of a matrix; using local numbering of rows. 5815 5816 Collective on Mat 5817 5818 Input Parameters: 5819 + mat - the matrix 5820 . is - index set of rows to remove 5821 . diag - value put in all diagonals of eliminated rows 5822 . x - optional vector of solutions for zeroed rows (other entries in vector are not used) 5823 - b - optional vector of right hand side, that will be adjusted by provided solution 5824 5825 Notes: 5826 Before calling MatZeroRowsColumnsLocalIS(), the user must first set the 5827 local-to-global mapping by calling MatSetLocalToGlobalMapping(). 5828 5829 The user can set a value in the diagonal entry (or for the AIJ and 5830 row formats can optionally remove the main diagonal entry from the 5831 nonzero structure as well, by passing 0.0 as the final argument). 5832 5833 Level: intermediate 5834 5835 Concepts: matrices^zeroing 5836 5837 .seealso: MatZeroRows(), MatZeroRowsLocal(), MatZeroEntries(), MatZeroRows(), MatSetLocalToGlobalMapping 5838 @*/ 5839 PetscErrorCode MatZeroRowsColumnsLocalIS(Mat mat,IS is,PetscScalar diag,Vec x,Vec b) 5840 { 5841 PetscErrorCode ierr; 5842 PetscInt numRows; 5843 const PetscInt *rows; 5844 5845 PetscFunctionBegin; 5846 PetscValidHeaderSpecific(mat,MAT_CLASSID,1); 5847 PetscValidType(mat,1); 5848 PetscValidHeaderSpecific(is,IS_CLASSID,2); 5849 if (!mat->assembled) SETERRQ(((PetscObject)mat)->comm,PETSC_ERR_ARG_WRONGSTATE,"Not for unassembled matrix"); 5850 if (mat->factortype) SETERRQ(((PetscObject)mat)->comm,PETSC_ERR_ARG_WRONGSTATE,"Not for factored matrix"); 5851 MatCheckPreallocated(mat,1); 5852 5853 ierr = ISGetLocalSize(is,&numRows);CHKERRQ(ierr); 5854 ierr = ISGetIndices(is,&rows);CHKERRQ(ierr); 5855 ierr = MatZeroRowsColumnsLocal(mat,numRows,rows,diag,x,b);CHKERRQ(ierr); 5856 ierr = ISRestoreIndices(is,&rows);CHKERRQ(ierr); 5857 PetscFunctionReturn(0); 5858 } 5859 5860 #undef __FUNCT__ 5861 #define __FUNCT__ "MatGetSize" 5862 /*@ 5863 MatGetSize - Returns the numbers of rows and columns in a matrix. 5864 5865 Not Collective 5866 5867 Input Parameter: 5868 . mat - the matrix 5869 5870 Output Parameters: 5871 + m - the number of global rows 5872 - n - the number of global columns 5873 5874 Note: both output parameters can be PETSC_NULL on input. 5875 5876 Level: beginner 5877 5878 Concepts: matrices^size 5879 5880 .seealso: MatGetLocalSize() 5881 @*/ 5882 PetscErrorCode MatGetSize(Mat mat,PetscInt *m,PetscInt* n) 5883 { 5884 PetscFunctionBegin; 5885 PetscValidHeaderSpecific(mat,MAT_CLASSID,1); 5886 if (m) *m = mat->rmap->N; 5887 if (n) *n = mat->cmap->N; 5888 PetscFunctionReturn(0); 5889 } 5890 5891 #undef __FUNCT__ 5892 #define __FUNCT__ "MatGetLocalSize" 5893 /*@ 5894 MatGetLocalSize - Returns the number of rows and columns in a matrix 5895 stored locally. This information may be implementation dependent, so 5896 use with care. 5897 5898 Not Collective 5899 5900 Input Parameters: 5901 . mat - the matrix 5902 5903 Output Parameters: 5904 + m - the number of local rows 5905 - n - the number of local columns 5906 5907 Note: both output parameters can be PETSC_NULL on input. 5908 5909 Level: beginner 5910 5911 Concepts: matrices^local size 5912 5913 .seealso: MatGetSize() 5914 @*/ 5915 PetscErrorCode MatGetLocalSize(Mat mat,PetscInt *m,PetscInt* n) 5916 { 5917 PetscFunctionBegin; 5918 PetscValidHeaderSpecific(mat,MAT_CLASSID,1); 5919 if (m) PetscValidIntPointer(m,2); 5920 if (n) PetscValidIntPointer(n,3); 5921 if (m) *m = mat->rmap->n; 5922 if (n) *n = mat->cmap->n; 5923 PetscFunctionReturn(0); 5924 } 5925 5926 #undef __FUNCT__ 5927 #define __FUNCT__ "MatGetOwnershipRangeColumn" 5928 /*@ 5929 MatGetOwnershipRangeColumn - Returns the range of matrix columns associated with rows of a vector one multiplies by that owned by 5930 this processor. (The columns of the "diagonal block") 5931 5932 Not Collective, unless matrix has not been allocated, then collective on Mat 5933 5934 Input Parameters: 5935 . mat - the matrix 5936 5937 Output Parameters: 5938 + m - the global index of the first local column 5939 - n - one more than the global index of the last local column 5940 5941 Notes: both output parameters can be PETSC_NULL on input. 5942 5943 Level: developer 5944 5945 Concepts: matrices^column ownership 5946 5947 .seealso: MatGetOwnershipRange(), MatGetOwnershipRanges(), MatGetOwnershipRangesColumn() 5948 5949 @*/ 5950 PetscErrorCode MatGetOwnershipRangeColumn(Mat mat,PetscInt *m,PetscInt* n) 5951 { 5952 5953 PetscFunctionBegin; 5954 PetscValidHeaderSpecific(mat,MAT_CLASSID,1); 5955 PetscValidType(mat,1); 5956 if (m) PetscValidIntPointer(m,2); 5957 if (n) PetscValidIntPointer(n,3); 5958 MatCheckPreallocated(mat,1); 5959 if (m) *m = mat->cmap->rstart; 5960 if (n) *n = mat->cmap->rend; 5961 PetscFunctionReturn(0); 5962 } 5963 5964 #undef __FUNCT__ 5965 #define __FUNCT__ "MatGetOwnershipRange" 5966 /*@ 5967 MatGetOwnershipRange - Returns the range of matrix rows owned by 5968 this processor, assuming that the matrix is laid out with the first 5969 n1 rows on the first processor, the next n2 rows on the second, etc. 5970 For certain parallel layouts this range may not be well defined. 5971 5972 Not Collective, unless matrix has not been allocated, then collective on Mat 5973 5974 Input Parameters: 5975 . mat - the matrix 5976 5977 Output Parameters: 5978 + m - the global index of the first local row 5979 - n - one more than the global index of the last local row 5980 5981 Note: both output parameters can be PETSC_NULL on input. 5982 5983 Level: beginner 5984 5985 Concepts: matrices^row ownership 5986 5987 .seealso: MatGetOwnershipRanges(), MatGetOwnershipRangeColumn(), MatGetOwnershipRangesColumn() 5988 5989 @*/ 5990 PetscErrorCode MatGetOwnershipRange(Mat mat,PetscInt *m,PetscInt* n) 5991 { 5992 5993 PetscFunctionBegin; 5994 PetscValidHeaderSpecific(mat,MAT_CLASSID,1); 5995 PetscValidType(mat,1); 5996 if (m) PetscValidIntPointer(m,2); 5997 if (n) PetscValidIntPointer(n,3); 5998 MatCheckPreallocated(mat,1); 5999 if (m) *m = mat->rmap->rstart; 6000 if (n) *n = mat->rmap->rend; 6001 PetscFunctionReturn(0); 6002 } 6003 6004 #undef __FUNCT__ 6005 #define __FUNCT__ "MatGetOwnershipRanges" 6006 /*@C 6007 MatGetOwnershipRanges - Returns the range of matrix rows owned by 6008 each process 6009 6010 Not Collective, unless matrix has not been allocated, then collective on Mat 6011 6012 Input Parameters: 6013 . mat - the matrix 6014 6015 Output Parameters: 6016 . ranges - start of each processors portion plus one more then the total length at the end 6017 6018 Level: beginner 6019 6020 Concepts: matrices^row ownership 6021 6022 .seealso: MatGetOwnershipRange(), MatGetOwnershipRangeColumn(), MatGetOwnershipRangesColumn() 6023 6024 @*/ 6025 PetscErrorCode MatGetOwnershipRanges(Mat mat,const PetscInt **ranges) 6026 { 6027 PetscErrorCode ierr; 6028 6029 PetscFunctionBegin; 6030 PetscValidHeaderSpecific(mat,MAT_CLASSID,1); 6031 PetscValidType(mat,1); 6032 MatCheckPreallocated(mat,1); 6033 ierr = PetscLayoutGetRanges(mat->rmap,ranges);CHKERRQ(ierr); 6034 PetscFunctionReturn(0); 6035 } 6036 6037 #undef __FUNCT__ 6038 #define __FUNCT__ "MatGetOwnershipRangesColumn" 6039 /*@C 6040 MatGetOwnershipRangesColumn - Returns the range of matrix columns associated with rows of a vector one multiplies by that owned by 6041 this processor. (The columns of the "diagonal blocks" for each process) 6042 6043 Not Collective, unless matrix has not been allocated, then collective on Mat 6044 6045 Input Parameters: 6046 . mat - the matrix 6047 6048 Output Parameters: 6049 . ranges - start of each processors portion plus one more then the total length at the end 6050 6051 Level: beginner 6052 6053 Concepts: matrices^column ownership 6054 6055 .seealso: MatGetOwnershipRange(), MatGetOwnershipRangeColumn(), MatGetOwnershipRanges() 6056 6057 @*/ 6058 PetscErrorCode MatGetOwnershipRangesColumn(Mat mat,const PetscInt **ranges) 6059 { 6060 PetscErrorCode ierr; 6061 6062 PetscFunctionBegin; 6063 PetscValidHeaderSpecific(mat,MAT_CLASSID,1); 6064 PetscValidType(mat,1); 6065 MatCheckPreallocated(mat,1); 6066 ierr = PetscLayoutGetRanges(mat->cmap,ranges);CHKERRQ(ierr); 6067 PetscFunctionReturn(0); 6068 } 6069 6070 #undef __FUNCT__ 6071 #define __FUNCT__ "MatGetOwnershipIS" 6072 /*@C 6073 MatGetOwnershipIS - Get row and column ownership as index sets 6074 6075 Not Collective 6076 6077 Input Arguments: 6078 . A - matrix of type Elemental 6079 6080 Output Arguments: 6081 + rows - rows in which this process owns elements 6082 . cols - columns in which this process owns elements 6083 6084 Level: intermediate 6085 6086 .seealso: MatGetOwnershipRange(), MatGetOwnershipRangeColumn(), MatSetValues(), MATELEMENTAL, MatSetValues() 6087 @*/ 6088 PetscErrorCode MatGetOwnershipIS(Mat A,IS *rows,IS *cols) 6089 { 6090 PetscErrorCode ierr,(*f)(Mat,IS*,IS*); 6091 6092 PetscFunctionBegin; 6093 MatCheckPreallocated(A,1); 6094 ierr = PetscObjectQueryFunction((PetscObject)A,"MatGetOwnershipIS_C",(PetscVoidStarFunction)&f);CHKERRQ(ierr); 6095 if (f) { 6096 ierr = (*f)(A,rows,cols);CHKERRQ(ierr); 6097 } else { /* Create a standard row-based partition, each process is responsible for ALL columns in their row block */ 6098 if (rows) {ierr = ISCreateStride(PETSC_COMM_SELF,A->rmap->n,A->rmap->rstart,1,rows);CHKERRQ(ierr);} 6099 if (cols) {ierr = ISCreateStride(PETSC_COMM_SELF,A->cmap->N,0,1,cols);CHKERRQ(ierr);} 6100 } 6101 PetscFunctionReturn(0); 6102 } 6103 6104 #undef __FUNCT__ 6105 #define __FUNCT__ "MatILUFactorSymbolic" 6106 /*@C 6107 MatILUFactorSymbolic - Performs symbolic ILU factorization of a matrix. 6108 Uses levels of fill only, not drop tolerance. Use MatLUFactorNumeric() 6109 to complete the factorization. 6110 6111 Collective on Mat 6112 6113 Input Parameters: 6114 + mat - the matrix 6115 . row - row permutation 6116 . column - column permutation 6117 - info - structure containing 6118 $ levels - number of levels of fill. 6119 $ expected fill - as ratio of original fill. 6120 $ 1 or 0 - indicating force fill on diagonal (improves robustness for matrices 6121 missing diagonal entries) 6122 6123 Output Parameters: 6124 . fact - new matrix that has been symbolically factored 6125 6126 Notes: 6127 See the <a href="../../docs/manual.pdf">users manual</a> for additional information about 6128 choosing the fill factor for better efficiency. 6129 6130 Most users should employ the simplified KSP interface for linear solvers 6131 instead of working directly with matrix algebra routines such as this. 6132 See, e.g., KSPCreate(). 6133 6134 Level: developer 6135 6136 Concepts: matrices^symbolic LU factorization 6137 Concepts: matrices^factorization 6138 Concepts: LU^symbolic factorization 6139 6140 .seealso: MatLUFactorSymbolic(), MatLUFactorNumeric(), MatCholeskyFactor() 6141 MatGetOrdering(), MatFactorInfo 6142 6143 Developer Note: fortran interface is not autogenerated as the f90 6144 interface defintion cannot be generated correctly [due to MatFactorInfo] 6145 6146 @*/ 6147 PetscErrorCode MatILUFactorSymbolic(Mat fact,Mat mat,IS row,IS col,const MatFactorInfo *info) 6148 { 6149 PetscErrorCode ierr; 6150 6151 PetscFunctionBegin; 6152 PetscValidHeaderSpecific(mat,MAT_CLASSID,1); 6153 PetscValidType(mat,1); 6154 PetscValidHeaderSpecific(row,IS_CLASSID,2); 6155 PetscValidHeaderSpecific(col,IS_CLASSID,3); 6156 PetscValidPointer(info,4); 6157 PetscValidPointer(fact,5); 6158 if (info->levels < 0) SETERRQ1(((PetscObject)mat)->comm,PETSC_ERR_ARG_OUTOFRANGE,"Levels of fill negative %D",(PetscInt)info->levels); 6159 if (info->fill < 1.0) SETERRQ1(((PetscObject)mat)->comm,PETSC_ERR_ARG_OUTOFRANGE,"Expected fill less than 1.0 %G",info->fill); 6160 if (!(fact)->ops->ilufactorsymbolic) { 6161 const MatSolverPackage spackage; 6162 ierr = MatFactorGetSolverPackage(fact,&spackage);CHKERRQ(ierr); 6163 SETERRQ2(((PetscObject)mat)->comm,PETSC_ERR_SUP,"Matrix type %s symbolic ILU using solver package %s",((PetscObject)mat)->type_name,spackage); 6164 } 6165 if (!mat->assembled) SETERRQ(((PetscObject)mat)->comm,PETSC_ERR_ARG_WRONGSTATE,"Not for unassembled matrix"); 6166 if (mat->factortype) SETERRQ(((PetscObject)mat)->comm,PETSC_ERR_ARG_WRONGSTATE,"Not for factored matrix"); 6167 MatCheckPreallocated(mat,2); 6168 6169 ierr = PetscLogEventBegin(MAT_ILUFactorSymbolic,mat,row,col,0);CHKERRQ(ierr); 6170 ierr = (fact->ops->ilufactorsymbolic)(fact,mat,row,col,info);CHKERRQ(ierr); 6171 ierr = PetscLogEventEnd(MAT_ILUFactorSymbolic,mat,row,col,0);CHKERRQ(ierr); 6172 PetscFunctionReturn(0); 6173 } 6174 6175 #undef __FUNCT__ 6176 #define __FUNCT__ "MatICCFactorSymbolic" 6177 /*@C 6178 MatICCFactorSymbolic - Performs symbolic incomplete 6179 Cholesky factorization for a symmetric matrix. Use 6180 MatCholeskyFactorNumeric() to complete the factorization. 6181 6182 Collective on Mat 6183 6184 Input Parameters: 6185 + mat - the matrix 6186 . perm - row and column permutation 6187 - info - structure containing 6188 $ levels - number of levels of fill. 6189 $ expected fill - as ratio of original fill. 6190 6191 Output Parameter: 6192 . fact - the factored matrix 6193 6194 Notes: 6195 Most users should employ the KSP interface for linear solvers 6196 instead of working directly with matrix algebra routines such as this. 6197 See, e.g., KSPCreate(). 6198 6199 Level: developer 6200 6201 Concepts: matrices^symbolic incomplete Cholesky factorization 6202 Concepts: matrices^factorization 6203 Concepts: Cholsky^symbolic factorization 6204 6205 .seealso: MatCholeskyFactorNumeric(), MatCholeskyFactor(), MatFactorInfo 6206 6207 Developer Note: fortran interface is not autogenerated as the f90 6208 interface defintion cannot be generated correctly [due to MatFactorInfo] 6209 6210 @*/ 6211 PetscErrorCode MatICCFactorSymbolic(Mat fact,Mat mat,IS perm,const MatFactorInfo *info) 6212 { 6213 PetscErrorCode ierr; 6214 6215 PetscFunctionBegin; 6216 PetscValidHeaderSpecific(mat,MAT_CLASSID,1); 6217 PetscValidType(mat,1); 6218 PetscValidHeaderSpecific(perm,IS_CLASSID,2); 6219 PetscValidPointer(info,3); 6220 PetscValidPointer(fact,4); 6221 if (mat->factortype) SETERRQ(((PetscObject)mat)->comm,PETSC_ERR_ARG_WRONGSTATE,"Not for factored matrix"); 6222 if (info->levels < 0) SETERRQ1(((PetscObject)mat)->comm,PETSC_ERR_ARG_OUTOFRANGE,"Levels negative %D",(PetscInt) info->levels); 6223 if (info->fill < 1.0) SETERRQ1(((PetscObject)mat)->comm,PETSC_ERR_ARG_OUTOFRANGE,"Expected fill less than 1.0 %G",info->fill); 6224 if (!(fact)->ops->iccfactorsymbolic) { 6225 const MatSolverPackage spackage; 6226 ierr = MatFactorGetSolverPackage(fact,&spackage);CHKERRQ(ierr); 6227 SETERRQ2(((PetscObject)mat)->comm,PETSC_ERR_SUP,"Matrix type %s symbolic ICC using solver package %s",((PetscObject)mat)->type_name,spackage); 6228 } 6229 if (!mat->assembled) SETERRQ(((PetscObject)mat)->comm,PETSC_ERR_ARG_WRONGSTATE,"Not for unassembled matrix"); 6230 MatCheckPreallocated(mat,2); 6231 6232 ierr = PetscLogEventBegin(MAT_ICCFactorSymbolic,mat,perm,0,0);CHKERRQ(ierr); 6233 ierr = (fact->ops->iccfactorsymbolic)(fact,mat,perm,info);CHKERRQ(ierr); 6234 ierr = PetscLogEventEnd(MAT_ICCFactorSymbolic,mat,perm,0,0);CHKERRQ(ierr); 6235 PetscFunctionReturn(0); 6236 } 6237 6238 #undef __FUNCT__ 6239 #define __FUNCT__ "MatGetArray" 6240 /*@C 6241 MatGetArray - Returns a pointer to the element values in the matrix. 6242 The result of this routine is dependent on the underlying matrix data 6243 structure, and may not even work for certain matrix types. You MUST 6244 call MatRestoreArray() when you no longer need to access the array. 6245 6246 Not Collective 6247 6248 Input Parameter: 6249 . mat - the matrix 6250 6251 Output Parameter: 6252 . v - the location of the values 6253 6254 6255 Fortran Note: 6256 This routine is used differently from Fortran, e.g., 6257 .vb 6258 Mat mat 6259 PetscScalar mat_array(1) 6260 PetscOffset i_mat 6261 PetscErrorCode ierr 6262 call MatGetArray(mat,mat_array,i_mat,ierr) 6263 6264 C Access first local entry in matrix; note that array is 6265 C treated as one dimensional 6266 value = mat_array(i_mat + 1) 6267 6268 [... other code ...] 6269 call MatRestoreArray(mat,mat_array,i_mat,ierr) 6270 .ve 6271 6272 See the <a href="../../docs/manual.pdf#ch_fortran">Fortran chapter of the users manual</a> and 6273 src/mat/examples/tests for details. 6274 6275 Level: advanced 6276 6277 Concepts: matrices^access array 6278 6279 .seealso: MatRestoreArray(), MatGetArrayF90(), MatGetRowIJ() 6280 @*/ 6281 PetscErrorCode MatGetArray(Mat mat,PetscScalar *v[]) 6282 { 6283 PetscErrorCode ierr; 6284 6285 PetscFunctionBegin; 6286 PetscValidHeaderSpecific(mat,MAT_CLASSID,1); 6287 PetscValidType(mat,1); 6288 PetscValidPointer(v,2); 6289 if (!mat->ops->getarray) SETERRQ1(PETSC_COMM_SELF,PETSC_ERR_SUP,"Mat type %s",((PetscObject)mat)->type_name); 6290 MatCheckPreallocated(mat,1); 6291 ierr = (*mat->ops->getarray)(mat,v);CHKERRQ(ierr); 6292 CHKMEMQ; 6293 PetscFunctionReturn(0); 6294 } 6295 6296 #undef __FUNCT__ 6297 #define __FUNCT__ "MatRestoreArray" 6298 /*@C 6299 MatRestoreArray - Restores the matrix after MatGetArray() has been called. 6300 6301 Not Collective 6302 6303 Input Parameter: 6304 + mat - the matrix 6305 - v - the location of the values 6306 6307 Fortran Note: 6308 This routine is used differently from Fortran, e.g., 6309 .vb 6310 Mat mat 6311 PetscScalar mat_array(1) 6312 PetscOffset i_mat 6313 PetscErrorCode ierr 6314 call MatGetArray(mat,mat_array,i_mat,ierr) 6315 6316 C Access first local entry in matrix; note that array is 6317 C treated as one dimensional 6318 value = mat_array(i_mat + 1) 6319 6320 [... other code ...] 6321 call MatRestoreArray(mat,mat_array,i_mat,ierr) 6322 .ve 6323 6324 See the <a href="../../docs/manual.pdf#ch_fortran">Fortran chapter of the users manual</a> 6325 src/mat/examples/tests for details 6326 6327 Level: advanced 6328 6329 .seealso: MatGetArray(), MatRestoreArrayF90() 6330 @*/ 6331 PetscErrorCode MatRestoreArray(Mat mat,PetscScalar *v[]) 6332 { 6333 PetscErrorCode ierr; 6334 6335 PetscFunctionBegin; 6336 PetscValidHeaderSpecific(mat,MAT_CLASSID,1); 6337 PetscValidType(mat,1); 6338 PetscValidPointer(v,2); 6339 CHKMEMQ; 6340 if (!mat->ops->restorearray) SETERRQ1(PETSC_COMM_SELF,PETSC_ERR_SUP,"Mat type %s",((PetscObject)mat)->type_name); 6341 ierr = (*mat->ops->restorearray)(mat,v);CHKERRQ(ierr); 6342 ierr = PetscObjectStateIncrease((PetscObject)mat);CHKERRQ(ierr); 6343 #if defined(PETSC_HAVE_CUSP) 6344 if (mat->valid_GPU_matrix != PETSC_CUSP_UNALLOCATED) { 6345 mat->valid_GPU_matrix = PETSC_CUSP_CPU; 6346 } 6347 #endif 6348 PetscFunctionReturn(0); 6349 } 6350 6351 #undef __FUNCT__ 6352 #define __FUNCT__ "MatGetSubMatrices" 6353 /*@C 6354 MatGetSubMatrices - Extracts several submatrices from a matrix. If submat 6355 points to an array of valid matrices, they may be reused to store the new 6356 submatrices. 6357 6358 Collective on Mat 6359 6360 Input Parameters: 6361 + mat - the matrix 6362 . n - the number of submatrixes to be extracted (on this processor, may be zero) 6363 . irow, icol - index sets of rows and columns to extract (must be sorted) 6364 - scall - either MAT_INITIAL_MATRIX or MAT_REUSE_MATRIX 6365 6366 Output Parameter: 6367 . submat - the array of submatrices 6368 6369 Notes: 6370 MatGetSubMatrices() can extract ONLY sequential submatrices 6371 (from both sequential and parallel matrices). Use MatGetSubMatrix() 6372 to extract a parallel submatrix. 6373 6374 Currently both row and column indices must be sorted to guarantee 6375 correctness with all matrix types. 6376 6377 When extracting submatrices from a parallel matrix, each processor can 6378 form a different submatrix by setting the rows and columns of its 6379 individual index sets according to the local submatrix desired. 6380 6381 When finished using the submatrices, the user should destroy 6382 them with MatDestroyMatrices(). 6383 6384 MAT_REUSE_MATRIX can only be used when the nonzero structure of the 6385 original matrix has not changed from that last call to MatGetSubMatrices(). 6386 6387 This routine creates the matrices in submat; you should NOT create them before 6388 calling it. It also allocates the array of matrix pointers submat. 6389 6390 For BAIJ matrices the index sets must respect the block structure, that is if they 6391 request one row/column in a block, they must request all rows/columns that are in 6392 that block. For example, if the block size is 2 you cannot request just row 0 and 6393 column 0. 6394 6395 Fortran Note: 6396 The Fortran interface is slightly different from that given below; it 6397 requires one to pass in as submat a Mat (integer) array of size at least m. 6398 6399 Level: advanced 6400 6401 Concepts: matrices^accessing submatrices 6402 Concepts: submatrices 6403 6404 .seealso: MatDestroyMatrices(), MatGetSubMatrix(), MatGetRow(), MatGetDiagonal(), MatReuse 6405 @*/ 6406 PetscErrorCode MatGetSubMatrices(Mat mat,PetscInt n,const IS irow[],const IS icol[],MatReuse scall,Mat *submat[]) 6407 { 6408 PetscErrorCode ierr; 6409 PetscInt i; 6410 PetscBool eq; 6411 6412 PetscFunctionBegin; 6413 PetscValidHeaderSpecific(mat,MAT_CLASSID,1); 6414 PetscValidType(mat,1); 6415 if (n) { 6416 PetscValidPointer(irow,3); 6417 PetscValidHeaderSpecific(*irow,IS_CLASSID,3); 6418 PetscValidPointer(icol,4); 6419 PetscValidHeaderSpecific(*icol,IS_CLASSID,4); 6420 } 6421 PetscValidPointer(submat,6); 6422 if (n && scall == MAT_REUSE_MATRIX) { 6423 PetscValidPointer(*submat,6); 6424 PetscValidHeaderSpecific(**submat,MAT_CLASSID,6); 6425 } 6426 if (!mat->ops->getsubmatrices) SETERRQ1(PETSC_COMM_SELF,PETSC_ERR_SUP,"Mat type %s",((PetscObject)mat)->type_name); 6427 if (!mat->assembled) SETERRQ(((PetscObject)mat)->comm,PETSC_ERR_ARG_WRONGSTATE,"Not for unassembled matrix"); 6428 if (mat->factortype) SETERRQ(((PetscObject)mat)->comm,PETSC_ERR_ARG_WRONGSTATE,"Not for factored matrix"); 6429 MatCheckPreallocated(mat,1); 6430 6431 ierr = PetscLogEventBegin(MAT_GetSubMatrices,mat,0,0,0);CHKERRQ(ierr); 6432 ierr = (*mat->ops->getsubmatrices)(mat,n,irow,icol,scall,submat);CHKERRQ(ierr); 6433 ierr = PetscLogEventEnd(MAT_GetSubMatrices,mat,0,0,0);CHKERRQ(ierr); 6434 for (i=0; i<n; i++) { 6435 if (mat->symmetric || mat->structurally_symmetric || mat->hermitian) { 6436 ierr = ISEqual(irow[i],icol[i],&eq);CHKERRQ(ierr); 6437 if (eq) { 6438 if (mat->symmetric){ 6439 ierr = MatSetOption((*submat)[i],MAT_SYMMETRIC,PETSC_TRUE);CHKERRQ(ierr); 6440 } else if (mat->hermitian) { 6441 ierr = MatSetOption((*submat)[i],MAT_HERMITIAN,PETSC_TRUE);CHKERRQ(ierr); 6442 } else if (mat->structurally_symmetric) { 6443 ierr = MatSetOption((*submat)[i],MAT_STRUCTURALLY_SYMMETRIC,PETSC_TRUE);CHKERRQ(ierr); 6444 } 6445 } 6446 } 6447 } 6448 PetscFunctionReturn(0); 6449 } 6450 6451 #undef __FUNCT__ 6452 #define __FUNCT__ "MatGetSubMatricesParallel" 6453 PetscErrorCode MatGetSubMatricesParallel(Mat mat,PetscInt n,const IS irow[],const IS icol[],MatReuse scall,Mat *submat[]) 6454 { 6455 PetscErrorCode ierr; 6456 PetscInt i; 6457 PetscBool eq; 6458 6459 PetscFunctionBegin; 6460 PetscValidHeaderSpecific(mat,MAT_CLASSID,1); 6461 PetscValidType(mat,1); 6462 if (n) { 6463 PetscValidPointer(irow,3); 6464 PetscValidHeaderSpecific(*irow,IS_CLASSID,3); 6465 PetscValidPointer(icol,4); 6466 PetscValidHeaderSpecific(*icol,IS_CLASSID,4); 6467 } 6468 PetscValidPointer(submat,6); 6469 if (n && scall == MAT_REUSE_MATRIX) { 6470 PetscValidPointer(*submat,6); 6471 PetscValidHeaderSpecific(**submat,MAT_CLASSID,6); 6472 } 6473 if (!mat->ops->getsubmatricesparallel) SETERRQ1(PETSC_COMM_SELF,PETSC_ERR_SUP,"Mat type %s",((PetscObject)mat)->type_name); 6474 if (!mat->assembled) SETERRQ(((PetscObject)mat)->comm,PETSC_ERR_ARG_WRONGSTATE,"Not for unassembled matrix"); 6475 if (mat->factortype) SETERRQ(((PetscObject)mat)->comm,PETSC_ERR_ARG_WRONGSTATE,"Not for factored matrix"); 6476 MatCheckPreallocated(mat,1); 6477 6478 ierr = PetscLogEventBegin(MAT_GetSubMatrices,mat,0,0,0);CHKERRQ(ierr); 6479 ierr = (*mat->ops->getsubmatricesparallel)(mat,n,irow,icol,scall,submat);CHKERRQ(ierr); 6480 ierr = PetscLogEventEnd(MAT_GetSubMatrices,mat,0,0,0);CHKERRQ(ierr); 6481 for (i=0; i<n; i++) { 6482 if (mat->symmetric || mat->structurally_symmetric || mat->hermitian) { 6483 ierr = ISEqual(irow[i],icol[i],&eq);CHKERRQ(ierr); 6484 if (eq) { 6485 if (mat->symmetric){ 6486 ierr = MatSetOption((*submat)[i],MAT_SYMMETRIC,PETSC_TRUE);CHKERRQ(ierr); 6487 } else if (mat->hermitian) { 6488 ierr = MatSetOption((*submat)[i],MAT_HERMITIAN,PETSC_TRUE);CHKERRQ(ierr); 6489 } else if (mat->structurally_symmetric) { 6490 ierr = MatSetOption((*submat)[i],MAT_STRUCTURALLY_SYMMETRIC,PETSC_TRUE);CHKERRQ(ierr); 6491 } 6492 } 6493 } 6494 } 6495 PetscFunctionReturn(0); 6496 } 6497 6498 #undef __FUNCT__ 6499 #define __FUNCT__ "MatDestroyMatrices" 6500 /*@C 6501 MatDestroyMatrices - Destroys a set of matrices obtained with MatGetSubMatrices(). 6502 6503 Collective on Mat 6504 6505 Input Parameters: 6506 + n - the number of local matrices 6507 - mat - the matrices (note that this is a pointer to the array of matrices, just to match the calling 6508 sequence of MatGetSubMatrices()) 6509 6510 Level: advanced 6511 6512 Notes: Frees not only the matrices, but also the array that contains the matrices 6513 In Fortran will not free the array. 6514 6515 .seealso: MatGetSubMatrices() 6516 @*/ 6517 PetscErrorCode MatDestroyMatrices(PetscInt n,Mat *mat[]) 6518 { 6519 PetscErrorCode ierr; 6520 PetscInt i; 6521 6522 PetscFunctionBegin; 6523 if (!*mat) PetscFunctionReturn(0); 6524 if (n < 0) SETERRQ1(PETSC_COMM_SELF,PETSC_ERR_ARG_OUTOFRANGE,"Trying to destroy negative number of matrices %D",n); 6525 PetscValidPointer(mat,2); 6526 for (i=0; i<n; i++) { 6527 ierr = MatDestroy(&(*mat)[i]);CHKERRQ(ierr); 6528 } 6529 /* memory is allocated even if n = 0 */ 6530 ierr = PetscFree(*mat);CHKERRQ(ierr); 6531 *mat = PETSC_NULL; 6532 PetscFunctionReturn(0); 6533 } 6534 6535 #undef __FUNCT__ 6536 #define __FUNCT__ "MatGetSeqNonzeroStructure" 6537 /*@C 6538 MatGetSeqNonzeroStructure - Extracts the sequential nonzero structure from a matrix. 6539 6540 Collective on Mat 6541 6542 Input Parameters: 6543 . mat - the matrix 6544 6545 Output Parameter: 6546 . matstruct - the sequential matrix with the nonzero structure of mat 6547 6548 Level: intermediate 6549 6550 .seealso: MatDestroySeqNonzeroStructure(), MatGetSubMatrices(), MatDestroyMatrices() 6551 @*/ 6552 PetscErrorCode MatGetSeqNonzeroStructure(Mat mat,Mat *matstruct) 6553 { 6554 PetscErrorCode ierr; 6555 6556 PetscFunctionBegin; 6557 PetscValidHeaderSpecific(mat,MAT_CLASSID,1); 6558 PetscValidPointer(matstruct,2); 6559 6560 PetscValidType(mat,1); 6561 if (mat->factortype) SETERRQ(((PetscObject)mat)->comm,PETSC_ERR_ARG_WRONGSTATE,"Not for factored matrix"); 6562 MatCheckPreallocated(mat,1); 6563 6564 if (!mat->ops->getseqnonzerostructure) SETERRQ1(((PetscObject)mat)->comm,PETSC_ERR_SUP,"Not for matrix type %s\n",((PetscObject)mat)->type_name); 6565 ierr = PetscLogEventBegin(MAT_GetSeqNonzeroStructure,mat,0,0,0);CHKERRQ(ierr); 6566 ierr = (*mat->ops->getseqnonzerostructure)(mat,matstruct);CHKERRQ(ierr); 6567 ierr = PetscLogEventEnd(MAT_GetSeqNonzeroStructure,mat,0,0,0);CHKERRQ(ierr); 6568 PetscFunctionReturn(0); 6569 } 6570 6571 #undef __FUNCT__ 6572 #define __FUNCT__ "MatDestroySeqNonzeroStructure" 6573 /*@C 6574 MatDestroySeqNonzeroStructure - Destroys matrix obtained with MatGetSeqNonzeroStructure(). 6575 6576 Collective on Mat 6577 6578 Input Parameters: 6579 . mat - the matrix (note that this is a pointer to the array of matrices, just to match the calling 6580 sequence of MatGetSequentialNonzeroStructure()) 6581 6582 Level: advanced 6583 6584 Notes: Frees not only the matrices, but also the array that contains the matrices 6585 6586 .seealso: MatGetSeqNonzeroStructure() 6587 @*/ 6588 PetscErrorCode MatDestroySeqNonzeroStructure(Mat *mat) 6589 { 6590 PetscErrorCode ierr; 6591 6592 PetscFunctionBegin; 6593 PetscValidPointer(mat,1); 6594 ierr = MatDestroy(mat);CHKERRQ(ierr); 6595 PetscFunctionReturn(0); 6596 } 6597 6598 #undef __FUNCT__ 6599 #define __FUNCT__ "MatIncreaseOverlap" 6600 /*@ 6601 MatIncreaseOverlap - Given a set of submatrices indicated by index sets, 6602 replaces the index sets by larger ones that represent submatrices with 6603 additional overlap. 6604 6605 Collective on Mat 6606 6607 Input Parameters: 6608 + mat - the matrix 6609 . n - the number of index sets 6610 . is - the array of index sets (these index sets will changed during the call) 6611 - ov - the additional overlap requested 6612 6613 Level: developer 6614 6615 Concepts: overlap 6616 Concepts: ASM^computing overlap 6617 6618 .seealso: MatGetSubMatrices() 6619 @*/ 6620 PetscErrorCode MatIncreaseOverlap(Mat mat,PetscInt n,IS is[],PetscInt ov) 6621 { 6622 PetscErrorCode ierr; 6623 6624 PetscFunctionBegin; 6625 PetscValidHeaderSpecific(mat,MAT_CLASSID,1); 6626 PetscValidType(mat,1); 6627 if (n < 0) SETERRQ1(PETSC_COMM_SELF,PETSC_ERR_ARG_OUTOFRANGE,"Must have one or more domains, you have %D",n); 6628 if (n) { 6629 PetscValidPointer(is,3); 6630 PetscValidHeaderSpecific(*is,IS_CLASSID,3); 6631 } 6632 if (!mat->assembled) SETERRQ(((PetscObject)mat)->comm,PETSC_ERR_ARG_WRONGSTATE,"Not for unassembled matrix"); 6633 if (mat->factortype) SETERRQ(((PetscObject)mat)->comm,PETSC_ERR_ARG_WRONGSTATE,"Not for factored matrix"); 6634 MatCheckPreallocated(mat,1); 6635 6636 if (!ov) PetscFunctionReturn(0); 6637 if (!mat->ops->increaseoverlap) SETERRQ1(((PetscObject)mat)->comm,PETSC_ERR_SUP,"Mat type %s",((PetscObject)mat)->type_name); 6638 ierr = PetscLogEventBegin(MAT_IncreaseOverlap,mat,0,0,0);CHKERRQ(ierr); 6639 ierr = (*mat->ops->increaseoverlap)(mat,n,is,ov);CHKERRQ(ierr); 6640 ierr = PetscLogEventEnd(MAT_IncreaseOverlap,mat,0,0,0);CHKERRQ(ierr); 6641 PetscFunctionReturn(0); 6642 } 6643 6644 #undef __FUNCT__ 6645 #define __FUNCT__ "MatGetBlockSize" 6646 /*@ 6647 MatGetBlockSize - Returns the matrix block size; useful especially for the 6648 block row and block diagonal formats. 6649 6650 Not Collective 6651 6652 Input Parameter: 6653 . mat - the matrix 6654 6655 Output Parameter: 6656 . bs - block size 6657 6658 Notes: 6659 Block row formats are MATSEQBAIJ, MATMPIBAIJ, MATSEQSBAIJ, MATMPISBAIJ 6660 6661 Level: intermediate 6662 6663 Concepts: matrices^block size 6664 6665 .seealso: MatCreateSeqBAIJ(), MatCreateBAIJ(), MatGetBlockSizes() 6666 @*/ 6667 PetscErrorCode MatGetBlockSize(Mat mat,PetscInt *bs) 6668 { 6669 6670 PetscFunctionBegin; 6671 PetscValidHeaderSpecific(mat,MAT_CLASSID,1); 6672 PetscValidType(mat,1); 6673 PetscValidIntPointer(bs,2); 6674 MatCheckPreallocated(mat,1); 6675 *bs = mat->rmap->bs; 6676 PetscFunctionReturn(0); 6677 } 6678 6679 #undef __FUNCT__ 6680 #define __FUNCT__ "MatGetBlockSizes" 6681 /*@ 6682 MatGetBlockSizes - Returns the matrix block row and column sizes; 6683 useful especially for the block row and block diagonal formats. 6684 6685 Not Collective 6686 6687 Input Parameter: 6688 . mat - the matrix 6689 6690 Output Parameter: 6691 . rbs - row block size 6692 . cbs - coumn block size 6693 6694 Notes: 6695 Block row formats are MATSEQBAIJ, MATMPIBAIJ, MATSEQSBAIJ, MATMPISBAIJ 6696 6697 Level: intermediate 6698 6699 Concepts: matrices^block size 6700 6701 .seealso: MatCreateSeqBAIJ(), MatCreateBAIJ(), MatGetBlockSize() 6702 @*/ 6703 PetscErrorCode MatGetBlockSizes(Mat mat,PetscInt *rbs, PetscInt *cbs) 6704 { 6705 6706 PetscFunctionBegin; 6707 PetscValidHeaderSpecific(mat,MAT_CLASSID,1); 6708 PetscValidType(mat,1); 6709 if(rbs) PetscValidIntPointer(rbs,2); 6710 if(cbs) PetscValidIntPointer(cbs,3); 6711 MatCheckPreallocated(mat,1); 6712 if(rbs) *rbs = mat->rmap->bs; 6713 if(cbs) *cbs = mat->cmap->bs; 6714 PetscFunctionReturn(0); 6715 } 6716 6717 #undef __FUNCT__ 6718 #define __FUNCT__ "MatSetBlockSize" 6719 /*@ 6720 MatSetBlockSize - Sets the matrix block size. 6721 6722 Logically Collective on Mat 6723 6724 Input Parameters: 6725 + mat - the matrix 6726 - bs - block size 6727 6728 Notes: 6729 This must be called before MatSetUp() or MatXXXSetPreallocation() (or will default to 1) and the block size cannot be changed later 6730 6731 Level: intermediate 6732 6733 Concepts: matrices^block size 6734 6735 .seealso: MatCreateSeqBAIJ(), MatCreateBAIJ(), MatGetBlockSize() 6736 @*/ 6737 PetscErrorCode MatSetBlockSize(Mat mat,PetscInt bs) 6738 { 6739 PetscErrorCode ierr; 6740 6741 PetscFunctionBegin; 6742 PetscValidHeaderSpecific(mat,MAT_CLASSID,1); 6743 PetscValidLogicalCollectiveInt(mat,bs,2); 6744 ierr = PetscLayoutSetBlockSize(mat->rmap,bs);CHKERRQ(ierr); 6745 ierr = PetscLayoutSetBlockSize(mat->cmap,bs);CHKERRQ(ierr); 6746 PetscFunctionReturn(0); 6747 } 6748 6749 #undef __FUNCT__ 6750 #define __FUNCT__ "MatSetBlockSizes" 6751 /*@ 6752 MatSetBlockSizes - Sets the matrix block row and column sizes. 6753 6754 Logically Collective on Mat 6755 6756 Input Parameters: 6757 + mat - the matrix 6758 - rbs - row block size 6759 - cbs - column block size 6760 6761 Notes: 6762 This must be called before MatSetUp() or MatXXXSetPreallocation() (or will default to 1) and the block size cannot be changed later 6763 6764 Level: intermediate 6765 6766 Concepts: matrices^block size 6767 6768 .seealso: MatCreateSeqBAIJ(), MatCreateBAIJ(), MatGetBlockSize() 6769 @*/ 6770 PetscErrorCode MatSetBlockSizes(Mat mat,PetscInt rbs,PetscInt cbs) 6771 { 6772 PetscErrorCode ierr; 6773 6774 PetscFunctionBegin; 6775 PetscValidHeaderSpecific(mat,MAT_CLASSID,1); 6776 PetscValidLogicalCollectiveInt(mat,rbs,2); 6777 ierr = PetscLayoutSetBlockSize(mat->rmap,rbs);CHKERRQ(ierr); 6778 ierr = PetscLayoutSetBlockSize(mat->cmap,cbs);CHKERRQ(ierr); 6779 PetscFunctionReturn(0); 6780 } 6781 6782 #undef __FUNCT__ 6783 #define __FUNCT__ "MatGetRowIJ" 6784 /*@C 6785 MatGetRowIJ - Returns the compressed row storage i and j indices for sequential matrices. 6786 6787 Collective on Mat 6788 6789 Input Parameters: 6790 + mat - the matrix 6791 . shift - 0 or 1 indicating we want the indices starting at 0 or 1 6792 . symmetric - PETSC_TRUE or PETSC_FALSE indicating the matrix data structure should be symmetrized 6793 - inodecompressed - PETSC_TRUE or PETSC_FALSE indicating if the nonzero structure of the 6794 inodes or the nonzero elements is wanted. For BAIJ matrices the compressed version is 6795 always used. 6796 6797 Output Parameters: 6798 + n - number of rows in the (possibly compressed) matrix 6799 . ia - the row pointers [of length n+1] 6800 . ja - the column indices 6801 - done - indicates if the routine actually worked and returned appropriate ia[] and ja[] arrays; callers 6802 are responsible for handling the case when done == PETSC_FALSE and ia and ja are not set 6803 6804 Level: developer 6805 6806 Notes: You CANNOT change any of the ia[] or ja[] values. 6807 6808 Use MatRestoreRowIJ() when you are finished accessing the ia[] and ja[] values 6809 6810 Fortran Node 6811 6812 In Fortran use 6813 $ PetscInt ia(1), ja(1) 6814 $ PetscOffset iia, jja 6815 $ call MatGetRowIJ(mat,shift,symmetric,inodecompressed,n,ia,iia,ja,jja,done,ierr) 6816 $ 6817 $ or 6818 $ 6819 $ PetscScalar, pointer :: xx_v(:) 6820 $ call MatGetRowIJF90(mat,shift,symmetric,inodecompressed,n,ia,ja,done,ierr) 6821 6822 6823 Acess the ith and jth entries via ia(iia + i) and ja(jja + j) 6824 6825 .seealso: MatGetColumnIJ(), MatRestoreRowIJ(), MatGetArray() 6826 @*/ 6827 PetscErrorCode MatGetRowIJ(Mat mat,PetscInt shift,PetscBool symmetric,PetscBool inodecompressed,PetscInt *n,PetscInt *ia[],PetscInt* ja[],PetscBool *done) 6828 { 6829 PetscErrorCode ierr; 6830 6831 PetscFunctionBegin; 6832 PetscValidHeaderSpecific(mat,MAT_CLASSID,1); 6833 PetscValidType(mat,1); 6834 PetscValidIntPointer(n,4); 6835 if (ia) PetscValidIntPointer(ia,5); 6836 if (ja) PetscValidIntPointer(ja,6); 6837 PetscValidIntPointer(done,7); 6838 MatCheckPreallocated(mat,1); 6839 if (!mat->ops->getrowij) *done = PETSC_FALSE; 6840 else { 6841 *done = PETSC_TRUE; 6842 ierr = PetscLogEventBegin(MAT_GetRowIJ,mat,0,0,0);CHKERRQ(ierr); 6843 ierr = (*mat->ops->getrowij)(mat,shift,symmetric,inodecompressed,n,ia,ja,done);CHKERRQ(ierr); 6844 ierr = PetscLogEventEnd(MAT_GetRowIJ,mat,0,0,0);CHKERRQ(ierr); 6845 } 6846 PetscFunctionReturn(0); 6847 } 6848 6849 #undef __FUNCT__ 6850 #define __FUNCT__ "MatGetColumnIJ" 6851 /*@C 6852 MatGetColumnIJ - Returns the compressed column storage i and j indices for sequential matrices. 6853 6854 Collective on Mat 6855 6856 Input Parameters: 6857 + mat - the matrix 6858 . shift - 1 or zero indicating we want the indices starting at 0 or 1 6859 . symmetric - PETSC_TRUE or PETSC_FALSE indicating the matrix data structure should be 6860 symmetrized 6861 - inodecompressed - PETSC_TRUE or PETSC_FALSE indicating if the nonzero structure of the 6862 inodes or the nonzero elements is wanted. For BAIJ matrices the compressed version is 6863 always used. 6864 6865 Output Parameters: 6866 + n - number of columns in the (possibly compressed) matrix 6867 . ia - the column pointers 6868 . ja - the row indices 6869 - done - PETSC_TRUE or PETSC_FALSE, indicating whether the values have been returned 6870 6871 Level: developer 6872 6873 .seealso: MatGetRowIJ(), MatRestoreColumnIJ() 6874 @*/ 6875 PetscErrorCode MatGetColumnIJ(Mat mat,PetscInt shift,PetscBool symmetric,PetscBool inodecompressed,PetscInt *n,PetscInt *ia[],PetscInt* ja[],PetscBool *done) 6876 { 6877 PetscErrorCode ierr; 6878 6879 PetscFunctionBegin; 6880 PetscValidHeaderSpecific(mat,MAT_CLASSID,1); 6881 PetscValidType(mat,1); 6882 PetscValidIntPointer(n,4); 6883 if (ia) PetscValidIntPointer(ia,5); 6884 if (ja) PetscValidIntPointer(ja,6); 6885 PetscValidIntPointer(done,7); 6886 MatCheckPreallocated(mat,1); 6887 if (!mat->ops->getcolumnij) *done = PETSC_FALSE; 6888 else { 6889 *done = PETSC_TRUE; 6890 ierr = (*mat->ops->getcolumnij)(mat,shift,symmetric,inodecompressed,n,ia,ja,done);CHKERRQ(ierr); 6891 } 6892 PetscFunctionReturn(0); 6893 } 6894 6895 #undef __FUNCT__ 6896 #define __FUNCT__ "MatRestoreRowIJ" 6897 /*@C 6898 MatRestoreRowIJ - Call after you are completed with the ia,ja indices obtained with 6899 MatGetRowIJ(). 6900 6901 Collective on Mat 6902 6903 Input Parameters: 6904 + mat - the matrix 6905 . shift - 1 or zero indicating we want the indices starting at 0 or 1 6906 . symmetric - PETSC_TRUE or PETSC_FALSE indicating the matrix data structure should be 6907 symmetrized 6908 - inodecompressed - PETSC_TRUE or PETSC_FALSE indicating if the nonzero structure of the 6909 inodes or the nonzero elements is wanted. For BAIJ matrices the compressed version is 6910 always used. 6911 6912 Output Parameters: 6913 + n - size of (possibly compressed) matrix 6914 . ia - the row pointers 6915 . ja - the column indices 6916 - done - PETSC_TRUE or PETSC_FALSE indicated that the values have been returned 6917 6918 Level: developer 6919 6920 .seealso: MatGetRowIJ(), MatRestoreColumnIJ() 6921 @*/ 6922 PetscErrorCode MatRestoreRowIJ(Mat mat,PetscInt shift,PetscBool symmetric,PetscBool inodecompressed,PetscInt *n,PetscInt *ia[],PetscInt* ja[],PetscBool *done) 6923 { 6924 PetscErrorCode ierr; 6925 6926 PetscFunctionBegin; 6927 PetscValidHeaderSpecific(mat,MAT_CLASSID,1); 6928 PetscValidType(mat,1); 6929 if (ia) PetscValidIntPointer(ia,5); 6930 if (ja) PetscValidIntPointer(ja,6); 6931 PetscValidIntPointer(done,7); 6932 MatCheckPreallocated(mat,1); 6933 6934 if (!mat->ops->restorerowij) *done = PETSC_FALSE; 6935 else { 6936 *done = PETSC_TRUE; 6937 ierr = (*mat->ops->restorerowij)(mat,shift,symmetric,inodecompressed,n,ia,ja,done);CHKERRQ(ierr); 6938 } 6939 PetscFunctionReturn(0); 6940 } 6941 6942 #undef __FUNCT__ 6943 #define __FUNCT__ "MatRestoreColumnIJ" 6944 /*@C 6945 MatRestoreColumnIJ - Call after you are completed with the ia,ja indices obtained with 6946 MatGetColumnIJ(). 6947 6948 Collective on Mat 6949 6950 Input Parameters: 6951 + mat - the matrix 6952 . shift - 1 or zero indicating we want the indices starting at 0 or 1 6953 - symmetric - PETSC_TRUE or PETSC_FALSE indicating the matrix data structure should be 6954 symmetrized 6955 - inodecompressed - PETSC_TRUE or PETSC_FALSE indicating if the nonzero structure of the 6956 inodes or the nonzero elements is wanted. For BAIJ matrices the compressed version is 6957 always used. 6958 6959 Output Parameters: 6960 + n - size of (possibly compressed) matrix 6961 . ia - the column pointers 6962 . ja - the row indices 6963 - done - PETSC_TRUE or PETSC_FALSE indicated that the values have been returned 6964 6965 Level: developer 6966 6967 .seealso: MatGetColumnIJ(), MatRestoreRowIJ() 6968 @*/ 6969 PetscErrorCode MatRestoreColumnIJ(Mat mat,PetscInt shift,PetscBool symmetric,PetscBool inodecompressed,PetscInt *n,PetscInt *ia[],PetscInt* ja[],PetscBool *done) 6970 { 6971 PetscErrorCode ierr; 6972 6973 PetscFunctionBegin; 6974 PetscValidHeaderSpecific(mat,MAT_CLASSID,1); 6975 PetscValidType(mat,1); 6976 if (ia) PetscValidIntPointer(ia,5); 6977 if (ja) PetscValidIntPointer(ja,6); 6978 PetscValidIntPointer(done,7); 6979 MatCheckPreallocated(mat,1); 6980 6981 if (!mat->ops->restorecolumnij) *done = PETSC_FALSE; 6982 else { 6983 *done = PETSC_TRUE; 6984 ierr = (*mat->ops->restorecolumnij)(mat,shift,symmetric,inodecompressed,n,ia,ja,done);CHKERRQ(ierr); 6985 } 6986 PetscFunctionReturn(0); 6987 } 6988 6989 #undef __FUNCT__ 6990 #define __FUNCT__ "MatColoringPatch" 6991 /*@C 6992 MatColoringPatch -Used inside matrix coloring routines that 6993 use MatGetRowIJ() and/or MatGetColumnIJ(). 6994 6995 Collective on Mat 6996 6997 Input Parameters: 6998 + mat - the matrix 6999 . ncolors - max color value 7000 . n - number of entries in colorarray 7001 - colorarray - array indicating color for each column 7002 7003 Output Parameters: 7004 . iscoloring - coloring generated using colorarray information 7005 7006 Level: developer 7007 7008 .seealso: MatGetRowIJ(), MatGetColumnIJ() 7009 7010 @*/ 7011 PetscErrorCode MatColoringPatch(Mat mat,PetscInt ncolors,PetscInt n,ISColoringValue colorarray[],ISColoring *iscoloring) 7012 { 7013 PetscErrorCode ierr; 7014 7015 PetscFunctionBegin; 7016 PetscValidHeaderSpecific(mat,MAT_CLASSID,1); 7017 PetscValidType(mat,1); 7018 PetscValidIntPointer(colorarray,4); 7019 PetscValidPointer(iscoloring,5); 7020 MatCheckPreallocated(mat,1); 7021 7022 if (!mat->ops->coloringpatch){ 7023 ierr = ISColoringCreate(((PetscObject)mat)->comm,ncolors,n,colorarray,iscoloring);CHKERRQ(ierr); 7024 } else { 7025 ierr = (*mat->ops->coloringpatch)(mat,ncolors,n,colorarray,iscoloring);CHKERRQ(ierr); 7026 } 7027 PetscFunctionReturn(0); 7028 } 7029 7030 7031 #undef __FUNCT__ 7032 #define __FUNCT__ "MatSetUnfactored" 7033 /*@ 7034 MatSetUnfactored - Resets a factored matrix to be treated as unfactored. 7035 7036 Logically Collective on Mat 7037 7038 Input Parameter: 7039 . mat - the factored matrix to be reset 7040 7041 Notes: 7042 This routine should be used only with factored matrices formed by in-place 7043 factorization via ILU(0) (or by in-place LU factorization for the MATSEQDENSE 7044 format). This option can save memory, for example, when solving nonlinear 7045 systems with a matrix-free Newton-Krylov method and a matrix-based, in-place 7046 ILU(0) preconditioner. 7047 7048 Note that one can specify in-place ILU(0) factorization by calling 7049 .vb 7050 PCType(pc,PCILU); 7051 PCFactorSeUseInPlace(pc); 7052 .ve 7053 or by using the options -pc_type ilu -pc_factor_in_place 7054 7055 In-place factorization ILU(0) can also be used as a local 7056 solver for the blocks within the block Jacobi or additive Schwarz 7057 methods (runtime option: -sub_pc_factor_in_place). See the discussion 7058 of these preconditioners in the <a href="../../docs/manual.pdf#ch_pc">PC chapter of the users manual</a> for details on setting 7059 local solver options. 7060 7061 Most users should employ the simplified KSP interface for linear solvers 7062 instead of working directly with matrix algebra routines such as this. 7063 See, e.g., KSPCreate(). 7064 7065 Level: developer 7066 7067 .seealso: PCFactorSetUseInPlace() 7068 7069 Concepts: matrices^unfactored 7070 7071 @*/ 7072 PetscErrorCode MatSetUnfactored(Mat mat) 7073 { 7074 PetscErrorCode ierr; 7075 7076 PetscFunctionBegin; 7077 PetscValidHeaderSpecific(mat,MAT_CLASSID,1); 7078 PetscValidType(mat,1); 7079 MatCheckPreallocated(mat,1); 7080 mat->factortype = MAT_FACTOR_NONE; 7081 if (!mat->ops->setunfactored) PetscFunctionReturn(0); 7082 ierr = (*mat->ops->setunfactored)(mat);CHKERRQ(ierr); 7083 PetscFunctionReturn(0); 7084 } 7085 7086 /*MC 7087 MatGetArrayF90 - Accesses a matrix array from Fortran90. 7088 7089 Synopsis: 7090 MatGetArrayF90(Mat x,{Scalar, pointer :: xx_v(:,:)},integer ierr) 7091 7092 Not collective 7093 7094 Input Parameter: 7095 . x - matrix 7096 7097 Output Parameters: 7098 + xx_v - the Fortran90 pointer to the array 7099 - ierr - error code 7100 7101 Example of Usage: 7102 .vb 7103 PetscScalar, pointer xx_v(:,:) 7104 .... 7105 call MatGetArrayF90(x,xx_v,ierr) 7106 a = xx_v(3) 7107 call MatRestoreArrayF90(x,xx_v,ierr) 7108 .ve 7109 7110 Notes: 7111 Not yet supported for all F90 compilers 7112 7113 Level: advanced 7114 7115 .seealso: MatRestoreArrayF90(), MatGetArray(), MatRestoreArray() 7116 7117 Concepts: matrices^accessing array 7118 7119 M*/ 7120 7121 /*MC 7122 MatRestoreArrayF90 - Restores a matrix array that has been 7123 accessed with MatGetArrayF90(). 7124 7125 Synopsis: 7126 MatRestoreArrayF90(Mat x,{Scalar, pointer :: xx_v(:)},integer ierr) 7127 7128 Not collective 7129 7130 Input Parameters: 7131 + x - matrix 7132 - xx_v - the Fortran90 pointer to the array 7133 7134 Output Parameter: 7135 . ierr - error code 7136 7137 Example of Usage: 7138 .vb 7139 PetscScalar, pointer xx_v(:) 7140 .... 7141 call MatGetArrayF90(x,xx_v,ierr) 7142 a = xx_v(3) 7143 call MatRestoreArrayF90(x,xx_v,ierr) 7144 .ve 7145 7146 Notes: 7147 Not yet supported for all F90 compilers 7148 7149 Level: advanced 7150 7151 .seealso: MatGetArrayF90(), MatGetArray(), MatRestoreArray() 7152 7153 M*/ 7154 7155 7156 #undef __FUNCT__ 7157 #define __FUNCT__ "MatGetSubMatrix" 7158 /*@ 7159 MatGetSubMatrix - Gets a single submatrix on the same number of processors 7160 as the original matrix. 7161 7162 Collective on Mat 7163 7164 Input Parameters: 7165 + mat - the original matrix 7166 . isrow - parallel IS containing the rows this processor should obtain 7167 . iscol - parallel IS containing all columns you wish to keep. Each process should list the columns that will be in IT's "diagonal part" in the new matrix. 7168 - cll - either MAT_INITIAL_MATRIX or MAT_REUSE_MATRIX 7169 7170 Output Parameter: 7171 . newmat - the new submatrix, of the same type as the old 7172 7173 Level: advanced 7174 7175 Notes: 7176 The submatrix will be able to be multiplied with vectors using the same layout as iscol. 7177 7178 The rows in isrow will be sorted into the same order as the original matrix on each process. 7179 7180 The first time this is called you should use a cll of MAT_INITIAL_MATRIX, 7181 the MatGetSubMatrix() routine will create the newmat for you. Any additional calls 7182 to this routine with a mat of the same nonzero structure and with a call of MAT_REUSE_MATRIX 7183 will reuse the matrix generated the first time. You should call MatDestroy() on newmat when 7184 you are finished using it. 7185 7186 The communicator of the newly obtained matrix is ALWAYS the same as the communicator of 7187 the input matrix. 7188 7189 If iscol is PETSC_NULL then all columns are obtained (not supported in Fortran). 7190 7191 Example usage: 7192 Consider the following 8x8 matrix with 34 non-zero values, that is 7193 assembled across 3 processors. Let's assume that proc0 owns 3 rows, 7194 proc1 owns 3 rows, proc2 owns 2 rows. This division can be shown 7195 as follows: 7196 7197 .vb 7198 1 2 0 | 0 3 0 | 0 4 7199 Proc0 0 5 6 | 7 0 0 | 8 0 7200 9 0 10 | 11 0 0 | 12 0 7201 ------------------------------------- 7202 13 0 14 | 15 16 17 | 0 0 7203 Proc1 0 18 0 | 19 20 21 | 0 0 7204 0 0 0 | 22 23 0 | 24 0 7205 ------------------------------------- 7206 Proc2 25 26 27 | 0 0 28 | 29 0 7207 30 0 0 | 31 32 33 | 0 34 7208 .ve 7209 7210 Suppose isrow = [0 1 | 4 | 6 7] and iscol = [1 2 | 3 4 5 | 6]. The resulting submatrix is 7211 7212 .vb 7213 2 0 | 0 3 0 | 0 7214 Proc0 5 6 | 7 0 0 | 8 7215 ------------------------------- 7216 Proc1 18 0 | 19 20 21 | 0 7217 ------------------------------- 7218 Proc2 26 27 | 0 0 28 | 29 7219 0 0 | 31 32 33 | 0 7220 .ve 7221 7222 7223 Concepts: matrices^submatrices 7224 7225 .seealso: MatGetSubMatrices() 7226 @*/ 7227 PetscErrorCode MatGetSubMatrix(Mat mat,IS isrow,IS iscol,MatReuse cll,Mat *newmat) 7228 { 7229 PetscErrorCode ierr; 7230 PetscMPIInt size; 7231 Mat *local; 7232 IS iscoltmp; 7233 7234 PetscFunctionBegin; 7235 PetscValidHeaderSpecific(mat,MAT_CLASSID,1); 7236 PetscValidHeaderSpecific(isrow,IS_CLASSID,2); 7237 if (iscol) PetscValidHeaderSpecific(iscol,IS_CLASSID,3); 7238 PetscValidPointer(newmat,5); 7239 if (cll == MAT_REUSE_MATRIX) PetscValidHeaderSpecific(*newmat,MAT_CLASSID,5); 7240 PetscValidType(mat,1); 7241 if (mat->factortype) SETERRQ(((PetscObject)mat)->comm,PETSC_ERR_ARG_WRONGSTATE,"Not for factored matrix"); 7242 MatCheckPreallocated(mat,1); 7243 ierr = MPI_Comm_size(((PetscObject)mat)->comm,&size);CHKERRQ(ierr); 7244 7245 if (!iscol) { 7246 ierr = ISCreateStride(((PetscObject)mat)->comm,mat->cmap->n,mat->cmap->rstart,1,&iscoltmp);CHKERRQ(ierr); 7247 } else { 7248 iscoltmp = iscol; 7249 } 7250 7251 /* if original matrix is on just one processor then use submatrix generated */ 7252 if (mat->ops->getsubmatrices && !mat->ops->getsubmatrix && size == 1 && cll == MAT_REUSE_MATRIX) { 7253 ierr = MatGetSubMatrices(mat,1,&isrow,&iscoltmp,MAT_REUSE_MATRIX,&newmat);CHKERRQ(ierr); 7254 if (!iscol) {ierr = ISDestroy(&iscoltmp);CHKERRQ(ierr);} 7255 PetscFunctionReturn(0); 7256 } else if (mat->ops->getsubmatrices && !mat->ops->getsubmatrix && size == 1) { 7257 ierr = MatGetSubMatrices(mat,1,&isrow,&iscoltmp,MAT_INITIAL_MATRIX,&local);CHKERRQ(ierr); 7258 *newmat = *local; 7259 ierr = PetscFree(local);CHKERRQ(ierr); 7260 if (!iscol) {ierr = ISDestroy(&iscoltmp);CHKERRQ(ierr);} 7261 PetscFunctionReturn(0); 7262 } else if (!mat->ops->getsubmatrix) { 7263 /* Create a new matrix type that implements the operation using the full matrix */ 7264 switch (cll) { 7265 case MAT_INITIAL_MATRIX: 7266 ierr = MatCreateSubMatrix(mat,isrow,iscoltmp,newmat);CHKERRQ(ierr); 7267 break; 7268 case MAT_REUSE_MATRIX: 7269 ierr = MatSubMatrixUpdate(*newmat,mat,isrow,iscoltmp);CHKERRQ(ierr); 7270 break; 7271 default: SETERRQ(((PetscObject)mat)->comm,PETSC_ERR_ARG_OUTOFRANGE,"Invalid MatReuse, must be either MAT_INITIAL_MATRIX or MAT_REUSE_MATRIX"); 7272 } 7273 if (!iscol) {ierr = ISDestroy(&iscoltmp);CHKERRQ(ierr);} 7274 PetscFunctionReturn(0); 7275 } 7276 7277 if (!mat->ops->getsubmatrix) SETERRQ1(PETSC_COMM_SELF,PETSC_ERR_SUP,"Mat type %s",((PetscObject)mat)->type_name); 7278 ierr = (*mat->ops->getsubmatrix)(mat,isrow,iscoltmp,cll,newmat);CHKERRQ(ierr); 7279 if (!iscol) {ierr = ISDestroy(&iscoltmp);CHKERRQ(ierr);} 7280 if (*newmat && cll == MAT_INITIAL_MATRIX) {ierr = PetscObjectStateIncrease((PetscObject)*newmat);CHKERRQ(ierr);} 7281 PetscFunctionReturn(0); 7282 } 7283 7284 #undef __FUNCT__ 7285 #define __FUNCT__ "MatStashSetInitialSize" 7286 /*@ 7287 MatStashSetInitialSize - sets the sizes of the matrix stash, that is 7288 used during the assembly process to store values that belong to 7289 other processors. 7290 7291 Not Collective 7292 7293 Input Parameters: 7294 + mat - the matrix 7295 . size - the initial size of the stash. 7296 - bsize - the initial size of the block-stash(if used). 7297 7298 Options Database Keys: 7299 + -matstash_initial_size <size> or <size0,size1,...sizep-1> 7300 - -matstash_block_initial_size <bsize> or <bsize0,bsize1,...bsizep-1> 7301 7302 Level: intermediate 7303 7304 Notes: 7305 The block-stash is used for values set with MatSetValuesBlocked() while 7306 the stash is used for values set with MatSetValues() 7307 7308 Run with the option -info and look for output of the form 7309 MatAssemblyBegin_MPIXXX:Stash has MM entries, uses nn mallocs. 7310 to determine the appropriate value, MM, to use for size and 7311 MatAssemblyBegin_MPIXXX:Block-Stash has BMM entries, uses nn mallocs. 7312 to determine the value, BMM to use for bsize 7313 7314 Concepts: stash^setting matrix size 7315 Concepts: matrices^stash 7316 7317 .seealso: MatAssemblyBegin(), MatAssemblyEnd(), Mat, MatStashGetInfo() 7318 7319 @*/ 7320 PetscErrorCode MatStashSetInitialSize(Mat mat,PetscInt size, PetscInt bsize) 7321 { 7322 PetscErrorCode ierr; 7323 7324 PetscFunctionBegin; 7325 PetscValidHeaderSpecific(mat,MAT_CLASSID,1); 7326 PetscValidType(mat,1); 7327 ierr = MatStashSetInitialSize_Private(&mat->stash,size);CHKERRQ(ierr); 7328 ierr = MatStashSetInitialSize_Private(&mat->bstash,bsize);CHKERRQ(ierr); 7329 PetscFunctionReturn(0); 7330 } 7331 7332 #undef __FUNCT__ 7333 #define __FUNCT__ "MatInterpolateAdd" 7334 /*@ 7335 MatInterpolateAdd - w = y + A*x or A'*x depending on the shape of 7336 the matrix 7337 7338 Neighbor-wise Collective on Mat 7339 7340 Input Parameters: 7341 + mat - the matrix 7342 . x,y - the vectors 7343 - w - where the result is stored 7344 7345 Level: intermediate 7346 7347 Notes: 7348 w may be the same vector as y. 7349 7350 This allows one to use either the restriction or interpolation (its transpose) 7351 matrix to do the interpolation 7352 7353 Concepts: interpolation 7354 7355 .seealso: MatMultAdd(), MatMultTransposeAdd(), MatRestrict() 7356 7357 @*/ 7358 PetscErrorCode MatInterpolateAdd(Mat A,Vec x,Vec y,Vec w) 7359 { 7360 PetscErrorCode ierr; 7361 PetscInt M,N,Ny; 7362 7363 PetscFunctionBegin; 7364 PetscValidHeaderSpecific(A,MAT_CLASSID,1); 7365 PetscValidHeaderSpecific(x,VEC_CLASSID,2); 7366 PetscValidHeaderSpecific(y,VEC_CLASSID,3); 7367 PetscValidHeaderSpecific(w,VEC_CLASSID,4); 7368 PetscValidType(A,1); 7369 MatCheckPreallocated(A,1); 7370 ierr = MatGetSize(A,&M,&N);CHKERRQ(ierr); 7371 ierr = VecGetSize(y,&Ny);CHKERRQ(ierr); 7372 if (M == Ny) { 7373 ierr = MatMultAdd(A,x,y,w);CHKERRQ(ierr); 7374 } else { 7375 ierr = MatMultTransposeAdd(A,x,y,w);CHKERRQ(ierr); 7376 } 7377 PetscFunctionReturn(0); 7378 } 7379 7380 #undef __FUNCT__ 7381 #define __FUNCT__ "MatInterpolate" 7382 /*@ 7383 MatInterpolate - y = A*x or A'*x depending on the shape of 7384 the matrix 7385 7386 Neighbor-wise Collective on Mat 7387 7388 Input Parameters: 7389 + mat - the matrix 7390 - x,y - the vectors 7391 7392 Level: intermediate 7393 7394 Notes: 7395 This allows one to use either the restriction or interpolation (its transpose) 7396 matrix to do the interpolation 7397 7398 Concepts: matrices^interpolation 7399 7400 .seealso: MatMultAdd(), MatMultTransposeAdd(), MatRestrict() 7401 7402 @*/ 7403 PetscErrorCode MatInterpolate(Mat A,Vec x,Vec y) 7404 { 7405 PetscErrorCode ierr; 7406 PetscInt M,N,Ny; 7407 7408 PetscFunctionBegin; 7409 PetscValidHeaderSpecific(A,MAT_CLASSID,1); 7410 PetscValidHeaderSpecific(x,VEC_CLASSID,2); 7411 PetscValidHeaderSpecific(y,VEC_CLASSID,3); 7412 PetscValidType(A,1); 7413 MatCheckPreallocated(A,1); 7414 ierr = MatGetSize(A,&M,&N);CHKERRQ(ierr); 7415 ierr = VecGetSize(y,&Ny);CHKERRQ(ierr); 7416 if (M == Ny) { 7417 ierr = MatMult(A,x,y);CHKERRQ(ierr); 7418 } else { 7419 ierr = MatMultTranspose(A,x,y);CHKERRQ(ierr); 7420 } 7421 PetscFunctionReturn(0); 7422 } 7423 7424 #undef __FUNCT__ 7425 #define __FUNCT__ "MatRestrict" 7426 /*@ 7427 MatRestrict - y = A*x or A'*x 7428 7429 Neighbor-wise Collective on Mat 7430 7431 Input Parameters: 7432 + mat - the matrix 7433 - x,y - the vectors 7434 7435 Level: intermediate 7436 7437 Notes: 7438 This allows one to use either the restriction or interpolation (its transpose) 7439 matrix to do the restriction 7440 7441 Concepts: matrices^restriction 7442 7443 .seealso: MatMultAdd(), MatMultTransposeAdd(), MatInterpolate() 7444 7445 @*/ 7446 PetscErrorCode MatRestrict(Mat A,Vec x,Vec y) 7447 { 7448 PetscErrorCode ierr; 7449 PetscInt M,N,Ny; 7450 7451 PetscFunctionBegin; 7452 PetscValidHeaderSpecific(A,MAT_CLASSID,1); 7453 PetscValidHeaderSpecific(x,VEC_CLASSID,2); 7454 PetscValidHeaderSpecific(y,VEC_CLASSID,3); 7455 PetscValidType(A,1); 7456 MatCheckPreallocated(A,1); 7457 7458 ierr = MatGetSize(A,&M,&N);CHKERRQ(ierr); 7459 ierr = VecGetSize(y,&Ny);CHKERRQ(ierr); 7460 if (M == Ny) { 7461 ierr = MatMult(A,x,y);CHKERRQ(ierr); 7462 } else { 7463 ierr = MatMultTranspose(A,x,y);CHKERRQ(ierr); 7464 } 7465 PetscFunctionReturn(0); 7466 } 7467 7468 #undef __FUNCT__ 7469 #define __FUNCT__ "MatGetNullSpace" 7470 /*@ 7471 MatGetNullSpace - retrieves the null space to a matrix. 7472 7473 Logically Collective on Mat and MatNullSpace 7474 7475 Input Parameters: 7476 + mat - the matrix 7477 - nullsp - the null space object 7478 7479 Level: developer 7480 7481 Notes: 7482 This null space is used by solvers. Overwrites any previous null space that may have been attached 7483 7484 Concepts: null space^attaching to matrix 7485 7486 .seealso: MatCreate(), MatNullSpaceCreate(), MatSetNearNullSpace() 7487 @*/ 7488 PetscErrorCode MatGetNullSpace(Mat mat, MatNullSpace *nullsp) 7489 { 7490 PetscFunctionBegin; 7491 PetscValidHeaderSpecific(mat,MAT_CLASSID,1); 7492 PetscValidType(mat,1); 7493 PetscValidPointer(nullsp,2); 7494 *nullsp = mat->nullsp; 7495 PetscFunctionReturn(0); 7496 } 7497 7498 #undef __FUNCT__ 7499 #define __FUNCT__ "MatSetNullSpace" 7500 /*@ 7501 MatSetNullSpace - attaches a null space to a matrix. 7502 This null space will be removed from the resulting vector whenever 7503 MatMult() is called 7504 7505 Logically Collective on Mat and MatNullSpace 7506 7507 Input Parameters: 7508 + mat - the matrix 7509 - nullsp - the null space object 7510 7511 Level: advanced 7512 7513 Notes: 7514 This null space is used by solvers. Overwrites any previous null space that may have been attached 7515 7516 Concepts: null space^attaching to matrix 7517 7518 .seealso: MatCreate(), MatNullSpaceCreate(), MatSetNearNullSpace() 7519 @*/ 7520 PetscErrorCode MatSetNullSpace(Mat mat,MatNullSpace nullsp) 7521 { 7522 PetscErrorCode ierr; 7523 7524 PetscFunctionBegin; 7525 PetscValidHeaderSpecific(mat,MAT_CLASSID,1); 7526 PetscValidType(mat,1); 7527 PetscValidHeaderSpecific(nullsp,MAT_NULLSPACE_CLASSID,2); 7528 MatCheckPreallocated(mat,1); 7529 ierr = PetscObjectReference((PetscObject)nullsp);CHKERRQ(ierr); 7530 ierr = MatNullSpaceDestroy(&mat->nullsp);CHKERRQ(ierr); 7531 mat->nullsp = nullsp; 7532 PetscFunctionReturn(0); 7533 } 7534 7535 #undef __FUNCT__ 7536 #define __FUNCT__ "MatSetNearNullSpace" 7537 /*@ 7538 MatSetNearNullSpace - attaches a null space to a matrix. 7539 This null space will be used to provide near null space vectors to a multigrid preconditioner built from this matrix. 7540 7541 Logically Collective on Mat and MatNullSpace 7542 7543 Input Parameters: 7544 + mat - the matrix 7545 - nullsp - the null space object 7546 7547 Level: advanced 7548 7549 Notes: 7550 Overwrites any previous near null space that may have been attached 7551 7552 Concepts: null space^attaching to matrix 7553 7554 .seealso: MatCreate(), MatNullSpaceCreate(), MatSetNullSpace() 7555 @*/ 7556 PetscErrorCode MatSetNearNullSpace(Mat mat,MatNullSpace nullsp) 7557 { 7558 PetscErrorCode ierr; 7559 7560 PetscFunctionBegin; 7561 PetscValidHeaderSpecific(mat,MAT_CLASSID,1); 7562 PetscValidType(mat,1); 7563 PetscValidHeaderSpecific(nullsp,MAT_NULLSPACE_CLASSID,2); 7564 MatCheckPreallocated(mat,1); 7565 ierr = PetscObjectReference((PetscObject)nullsp);CHKERRQ(ierr); 7566 ierr = MatNullSpaceDestroy(&mat->nearnullsp);CHKERRQ(ierr); 7567 mat->nearnullsp = nullsp; 7568 PetscFunctionReturn(0); 7569 } 7570 7571 #undef __FUNCT__ 7572 #define __FUNCT__ "MatGetNearNullSpace" 7573 /*@ 7574 MatGetNearNullSpace -Get null space attached with MatSetNearNullSpace() 7575 7576 Not Collective 7577 7578 Input Parameters: 7579 . mat - the matrix 7580 7581 Output Parameters: 7582 . nullsp - the null space object, PETSC_NULL if not set 7583 7584 Level: developer 7585 7586 Concepts: null space^attaching to matrix 7587 7588 .seealso: MatSetNearNullSpace(), MatGetNullSpace() 7589 @*/ 7590 PetscErrorCode MatGetNearNullSpace(Mat mat,MatNullSpace *nullsp) 7591 { 7592 7593 PetscFunctionBegin; 7594 PetscValidHeaderSpecific(mat,MAT_CLASSID,1); 7595 PetscValidType(mat,1); 7596 PetscValidPointer(nullsp,2); 7597 MatCheckPreallocated(mat,1); 7598 *nullsp = mat->nearnullsp; 7599 PetscFunctionReturn(0); 7600 } 7601 7602 #undef __FUNCT__ 7603 #define __FUNCT__ "MatICCFactor" 7604 /*@C 7605 MatICCFactor - Performs in-place incomplete Cholesky factorization of matrix. 7606 7607 Collective on Mat 7608 7609 Input Parameters: 7610 + mat - the matrix 7611 . row - row/column permutation 7612 . fill - expected fill factor >= 1.0 7613 - level - level of fill, for ICC(k) 7614 7615 Notes: 7616 Probably really in-place only when level of fill is zero, otherwise allocates 7617 new space to store factored matrix and deletes previous memory. 7618 7619 Most users should employ the simplified KSP interface for linear solvers 7620 instead of working directly with matrix algebra routines such as this. 7621 See, e.g., KSPCreate(). 7622 7623 Level: developer 7624 7625 Concepts: matrices^incomplete Cholesky factorization 7626 Concepts: Cholesky factorization 7627 7628 .seealso: MatICCFactorSymbolic(), MatLUFactorNumeric(), MatCholeskyFactor() 7629 7630 Developer Note: fortran interface is not autogenerated as the f90 7631 interface defintion cannot be generated correctly [due to MatFactorInfo] 7632 7633 @*/ 7634 PetscErrorCode MatICCFactor(Mat mat,IS row,const MatFactorInfo* info) 7635 { 7636 PetscErrorCode ierr; 7637 7638 PetscFunctionBegin; 7639 PetscValidHeaderSpecific(mat,MAT_CLASSID,1); 7640 PetscValidType(mat,1); 7641 if (row) PetscValidHeaderSpecific(row,IS_CLASSID,2); 7642 PetscValidPointer(info,3); 7643 if (mat->rmap->N != mat->cmap->N) SETERRQ(((PetscObject)mat)->comm,PETSC_ERR_ARG_WRONG,"matrix must be square"); 7644 if (!mat->assembled) SETERRQ(((PetscObject)mat)->comm,PETSC_ERR_ARG_WRONGSTATE,"Not for unassembled matrix"); 7645 if (mat->factortype) SETERRQ(((PetscObject)mat)->comm,PETSC_ERR_ARG_WRONGSTATE,"Not for factored matrix"); 7646 if (!mat->ops->iccfactor) SETERRQ1(((PetscObject)mat)->comm,PETSC_ERR_SUP,"Mat type %s",((PetscObject)mat)->type_name); 7647 MatCheckPreallocated(mat,1); 7648 ierr = (*mat->ops->iccfactor)(mat,row,info);CHKERRQ(ierr); 7649 ierr = PetscObjectStateIncrease((PetscObject)mat);CHKERRQ(ierr); 7650 PetscFunctionReturn(0); 7651 } 7652 7653 #undef __FUNCT__ 7654 #define __FUNCT__ "MatSetValuesAdic" 7655 /*@ 7656 MatSetValuesAdic - Sets values computed with ADIC automatic differentiation into a matrix. 7657 7658 Not Collective 7659 7660 Input Parameters: 7661 + mat - the matrix 7662 - v - the values compute with ADIC 7663 7664 Level: developer 7665 7666 Notes: 7667 Must call MatSetColoring() before using this routine. Also this matrix must already 7668 have its nonzero pattern determined. 7669 7670 .seealso: MatSetOption(), MatAssemblyBegin(), MatAssemblyEnd(), MatSetValuesBlocked(), MatSetValuesLocal(), 7671 MatSetValues(), MatSetColoring(), MatSetValuesAdifor() 7672 @*/ 7673 PetscErrorCode MatSetValuesAdic(Mat mat,void *v) 7674 { 7675 PetscErrorCode ierr; 7676 7677 PetscFunctionBegin; 7678 PetscValidHeaderSpecific(mat,MAT_CLASSID,1); 7679 PetscValidType(mat,1); 7680 PetscValidPointer(mat,2); 7681 7682 if (!mat->assembled) { 7683 SETERRQ(PETSC_COMM_SELF,PETSC_ERR_ARG_WRONGSTATE,"Matrix must be already assembled"); 7684 } 7685 ierr = PetscLogEventBegin(MAT_SetValues,mat,0,0,0);CHKERRQ(ierr); 7686 if (!mat->ops->setvaluesadic) SETERRQ1(PETSC_COMM_SELF,PETSC_ERR_SUP,"Mat type %s",((PetscObject)mat)->type_name); 7687 ierr = (*mat->ops->setvaluesadic)(mat,v);CHKERRQ(ierr); 7688 ierr = PetscLogEventEnd(MAT_SetValues,mat,0,0,0);CHKERRQ(ierr); 7689 ierr = MatView_Private(mat);CHKERRQ(ierr); 7690 ierr = PetscObjectStateIncrease((PetscObject)mat);CHKERRQ(ierr); 7691 PetscFunctionReturn(0); 7692 } 7693 7694 7695 #undef __FUNCT__ 7696 #define __FUNCT__ "MatSetColoring" 7697 /*@ 7698 MatSetColoring - Sets a coloring used by calls to MatSetValuesAdic() 7699 7700 Not Collective 7701 7702 Input Parameters: 7703 + mat - the matrix 7704 - coloring - the coloring 7705 7706 Level: developer 7707 7708 .seealso: MatSetOption(), MatAssemblyBegin(), MatAssemblyEnd(), MatSetValuesBlocked(), MatSetValuesLocal(), 7709 MatSetValues(), MatSetValuesAdic() 7710 @*/ 7711 PetscErrorCode MatSetColoring(Mat mat,ISColoring coloring) 7712 { 7713 PetscErrorCode ierr; 7714 7715 PetscFunctionBegin; 7716 PetscValidHeaderSpecific(mat,MAT_CLASSID,1); 7717 PetscValidType(mat,1); 7718 PetscValidPointer(coloring,2); 7719 7720 if (!mat->assembled) SETERRQ(((PetscObject)mat)->comm,PETSC_ERR_ARG_WRONGSTATE,"Matrix must be already assembled"); 7721 if (!mat->ops->setcoloring) SETERRQ1(((PetscObject)mat)->comm,PETSC_ERR_SUP,"Mat type %s",((PetscObject)mat)->type_name); 7722 ierr = (*mat->ops->setcoloring)(mat,coloring);CHKERRQ(ierr); 7723 PetscFunctionReturn(0); 7724 } 7725 7726 #undef __FUNCT__ 7727 #define __FUNCT__ "MatSetValuesAdifor" 7728 /*@ 7729 MatSetValuesAdifor - Sets values computed with automatic differentiation into a matrix. 7730 7731 Not Collective 7732 7733 Input Parameters: 7734 + mat - the matrix 7735 . nl - leading dimension of v 7736 - v - the values compute with ADIFOR 7737 7738 Level: developer 7739 7740 Notes: 7741 Must call MatSetColoring() before using this routine. Also this matrix must already 7742 have its nonzero pattern determined. 7743 7744 .seealso: MatSetOption(), MatAssemblyBegin(), MatAssemblyEnd(), MatSetValuesBlocked(), MatSetValuesLocal(), 7745 MatSetValues(), MatSetColoring() 7746 @*/ 7747 PetscErrorCode MatSetValuesAdifor(Mat mat,PetscInt nl,void *v) 7748 { 7749 PetscErrorCode ierr; 7750 7751 PetscFunctionBegin; 7752 PetscValidHeaderSpecific(mat,MAT_CLASSID,1); 7753 PetscValidType(mat,1); 7754 PetscValidPointer(v,3); 7755 7756 if (!mat->assembled) SETERRQ(((PetscObject)mat)->comm,PETSC_ERR_ARG_WRONGSTATE,"Matrix must be already assembled"); 7757 ierr = PetscLogEventBegin(MAT_SetValues,mat,0,0,0);CHKERRQ(ierr); 7758 if (!mat->ops->setvaluesadifor) SETERRQ1(((PetscObject)mat)->comm,PETSC_ERR_SUP,"Mat type %s",((PetscObject)mat)->type_name); 7759 ierr = (*mat->ops->setvaluesadifor)(mat,nl,v);CHKERRQ(ierr); 7760 ierr = PetscLogEventEnd(MAT_SetValues,mat,0,0,0);CHKERRQ(ierr); 7761 ierr = PetscObjectStateIncrease((PetscObject)mat);CHKERRQ(ierr); 7762 PetscFunctionReturn(0); 7763 } 7764 7765 #undef __FUNCT__ 7766 #define __FUNCT__ "MatDiagonalScaleLocal" 7767 /*@ 7768 MatDiagonalScaleLocal - Scales columns of a matrix given the scaling values including the 7769 ghosted ones. 7770 7771 Not Collective 7772 7773 Input Parameters: 7774 + mat - the matrix 7775 - diag = the diagonal values, including ghost ones 7776 7777 Level: developer 7778 7779 Notes: Works only for MPIAIJ and MPIBAIJ matrices 7780 7781 .seealso: MatDiagonalScale() 7782 @*/ 7783 PetscErrorCode MatDiagonalScaleLocal(Mat mat,Vec diag) 7784 { 7785 PetscErrorCode ierr; 7786 PetscMPIInt size; 7787 7788 PetscFunctionBegin; 7789 PetscValidHeaderSpecific(mat,MAT_CLASSID,1); 7790 PetscValidHeaderSpecific(diag,VEC_CLASSID,2); 7791 PetscValidType(mat,1); 7792 7793 if (!mat->assembled) SETERRQ(((PetscObject)mat)->comm,PETSC_ERR_ARG_WRONGSTATE,"Matrix must be already assembled"); 7794 ierr = PetscLogEventBegin(MAT_Scale,mat,0,0,0);CHKERRQ(ierr); 7795 ierr = MPI_Comm_size(((PetscObject)mat)->comm,&size);CHKERRQ(ierr); 7796 if (size == 1) { 7797 PetscInt n,m; 7798 ierr = VecGetSize(diag,&n);CHKERRQ(ierr); 7799 ierr = MatGetSize(mat,0,&m);CHKERRQ(ierr); 7800 if (m == n) { 7801 ierr = MatDiagonalScale(mat,0,diag);CHKERRQ(ierr); 7802 } else { 7803 SETERRQ(PETSC_COMM_SELF,PETSC_ERR_SUP,"Only supported for sequential matrices when no ghost points/periodic conditions"); 7804 } 7805 } else { 7806 ierr = PetscUseMethod(mat,"MatDiagonalScaleLocal_C",(Mat,Vec),(mat,diag));CHKERRQ(ierr); 7807 } 7808 ierr = PetscLogEventEnd(MAT_Scale,mat,0,0,0);CHKERRQ(ierr); 7809 ierr = PetscObjectStateIncrease((PetscObject)mat);CHKERRQ(ierr); 7810 PetscFunctionReturn(0); 7811 } 7812 7813 #undef __FUNCT__ 7814 #define __FUNCT__ "MatGetInertia" 7815 /*@ 7816 MatGetInertia - Gets the inertia from a factored matrix 7817 7818 Collective on Mat 7819 7820 Input Parameter: 7821 . mat - the matrix 7822 7823 Output Parameters: 7824 + nneg - number of negative eigenvalues 7825 . nzero - number of zero eigenvalues 7826 - npos - number of positive eigenvalues 7827 7828 Level: advanced 7829 7830 Notes: Matrix must have been factored by MatCholeskyFactor() 7831 7832 7833 @*/ 7834 PetscErrorCode MatGetInertia(Mat mat,PetscInt *nneg,PetscInt *nzero,PetscInt *npos) 7835 { 7836 PetscErrorCode ierr; 7837 7838 PetscFunctionBegin; 7839 PetscValidHeaderSpecific(mat,MAT_CLASSID,1); 7840 PetscValidType(mat,1); 7841 if (!mat->factortype) SETERRQ(((PetscObject)mat)->comm,PETSC_ERR_ARG_WRONGSTATE,"Unfactored matrix"); 7842 if (!mat->assembled) SETERRQ(((PetscObject)mat)->comm,PETSC_ERR_ARG_WRONGSTATE,"Numeric factor mat is not assembled"); 7843 if (!mat->ops->getinertia) SETERRQ1(PETSC_COMM_SELF,PETSC_ERR_SUP,"Mat type %s",((PetscObject)mat)->type_name); 7844 ierr = (*mat->ops->getinertia)(mat,nneg,nzero,npos);CHKERRQ(ierr); 7845 PetscFunctionReturn(0); 7846 } 7847 7848 /* ----------------------------------------------------------------*/ 7849 #undef __FUNCT__ 7850 #define __FUNCT__ "MatSolves" 7851 /*@C 7852 MatSolves - Solves A x = b, given a factored matrix, for a collection of vectors 7853 7854 Neighbor-wise Collective on Mat and Vecs 7855 7856 Input Parameters: 7857 + mat - the factored matrix 7858 - b - the right-hand-side vectors 7859 7860 Output Parameter: 7861 . x - the result vectors 7862 7863 Notes: 7864 The vectors b and x cannot be the same. I.e., one cannot 7865 call MatSolves(A,x,x). 7866 7867 Notes: 7868 Most users should employ the simplified KSP interface for linear solvers 7869 instead of working directly with matrix algebra routines such as this. 7870 See, e.g., KSPCreate(). 7871 7872 Level: developer 7873 7874 Concepts: matrices^triangular solves 7875 7876 .seealso: MatSolveAdd(), MatSolveTranspose(), MatSolveTransposeAdd(), MatSolve() 7877 @*/ 7878 PetscErrorCode MatSolves(Mat mat,Vecs b,Vecs x) 7879 { 7880 PetscErrorCode ierr; 7881 7882 PetscFunctionBegin; 7883 PetscValidHeaderSpecific(mat,MAT_CLASSID,1); 7884 PetscValidType(mat,1); 7885 if (x == b) SETERRQ(((PetscObject)mat)->comm,PETSC_ERR_ARG_IDN,"x and b must be different vectors"); 7886 if (!mat->factortype) SETERRQ(((PetscObject)mat)->comm,PETSC_ERR_ARG_WRONGSTATE,"Unfactored matrix"); 7887 if (!mat->rmap->N && !mat->cmap->N) PetscFunctionReturn(0); 7888 7889 if (!mat->ops->solves) SETERRQ1(((PetscObject)mat)->comm,PETSC_ERR_SUP,"Mat type %s",((PetscObject)mat)->type_name); 7890 MatCheckPreallocated(mat,1); 7891 ierr = PetscLogEventBegin(MAT_Solves,mat,0,0,0);CHKERRQ(ierr); 7892 ierr = (*mat->ops->solves)(mat,b,x);CHKERRQ(ierr); 7893 ierr = PetscLogEventEnd(MAT_Solves,mat,0,0,0);CHKERRQ(ierr); 7894 PetscFunctionReturn(0); 7895 } 7896 7897 #undef __FUNCT__ 7898 #define __FUNCT__ "MatIsSymmetric" 7899 /*@ 7900 MatIsSymmetric - Test whether a matrix is symmetric 7901 7902 Collective on Mat 7903 7904 Input Parameter: 7905 + A - the matrix to test 7906 - tol - difference between value and its transpose less than this amount counts as equal (use 0.0 for exact transpose) 7907 7908 Output Parameters: 7909 . flg - the result 7910 7911 Notes: For real numbers MatIsSymmetric() and MatIsHermitian() return identical results 7912 7913 Level: intermediate 7914 7915 Concepts: matrix^symmetry 7916 7917 .seealso: MatTranspose(), MatIsTranspose(), MatIsHermitian(), MatIsStructurallySymmetric(), MatSetOption(), MatIsSymmetricKnown() 7918 @*/ 7919 PetscErrorCode MatIsSymmetric(Mat A,PetscReal tol,PetscBool *flg) 7920 { 7921 PetscErrorCode ierr; 7922 7923 PetscFunctionBegin; 7924 PetscValidHeaderSpecific(A,MAT_CLASSID,1); 7925 PetscValidPointer(flg,2); 7926 7927 if (!A->symmetric_set) { 7928 if (!A->ops->issymmetric) { 7929 const MatType mattype; 7930 ierr = MatGetType(A,&mattype);CHKERRQ(ierr); 7931 SETERRQ1(PETSC_COMM_SELF,PETSC_ERR_SUP,"Matrix of type <%s> does not support checking for symmetric",mattype); 7932 } 7933 ierr = (*A->ops->issymmetric)(A,tol,flg);CHKERRQ(ierr); 7934 if (!tol) { 7935 A->symmetric_set = PETSC_TRUE; 7936 A->symmetric = *flg; 7937 if (A->symmetric) { 7938 A->structurally_symmetric_set = PETSC_TRUE; 7939 A->structurally_symmetric = PETSC_TRUE; 7940 } 7941 } 7942 } else if (A->symmetric) { 7943 *flg = PETSC_TRUE; 7944 } else if (!tol) { 7945 *flg = PETSC_FALSE; 7946 } else { 7947 if (!A->ops->issymmetric) { 7948 const MatType mattype; 7949 ierr = MatGetType(A,&mattype);CHKERRQ(ierr); 7950 SETERRQ1(PETSC_COMM_SELF,PETSC_ERR_SUP,"Matrix of type <%s> does not support checking for symmetric",mattype); 7951 } 7952 ierr = (*A->ops->issymmetric)(A,tol,flg);CHKERRQ(ierr); 7953 } 7954 PetscFunctionReturn(0); 7955 } 7956 7957 #undef __FUNCT__ 7958 #define __FUNCT__ "MatIsHermitian" 7959 /*@ 7960 MatIsHermitian - Test whether a matrix is Hermitian 7961 7962 Collective on Mat 7963 7964 Input Parameter: 7965 + A - the matrix to test 7966 - tol - difference between value and its transpose less than this amount counts as equal (use 0.0 for exact Hermitian) 7967 7968 Output Parameters: 7969 . flg - the result 7970 7971 Level: intermediate 7972 7973 Concepts: matrix^symmetry 7974 7975 .seealso: MatTranspose(), MatIsTranspose(), MatIsHermitian(), MatIsStructurallySymmetric(), MatSetOption(), 7976 MatIsSymmetricKnown(), MatIsSymmetric() 7977 @*/ 7978 PetscErrorCode MatIsHermitian(Mat A,PetscReal tol,PetscBool *flg) 7979 { 7980 PetscErrorCode ierr; 7981 7982 PetscFunctionBegin; 7983 PetscValidHeaderSpecific(A,MAT_CLASSID,1); 7984 PetscValidPointer(flg,2); 7985 7986 if (!A->hermitian_set) { 7987 if (!A->ops->ishermitian) { 7988 const MatType mattype; 7989 ierr = MatGetType(A,&mattype);CHKERRQ(ierr); 7990 SETERRQ1(PETSC_COMM_SELF,PETSC_ERR_SUP,"Matrix of type <%s> does not support checking for hermitian",mattype); 7991 } 7992 ierr = (*A->ops->ishermitian)(A,tol,flg);CHKERRQ(ierr); 7993 if (!tol) { 7994 A->hermitian_set = PETSC_TRUE; 7995 A->hermitian = *flg; 7996 if (A->hermitian) { 7997 A->structurally_symmetric_set = PETSC_TRUE; 7998 A->structurally_symmetric = PETSC_TRUE; 7999 } 8000 } 8001 } else if (A->hermitian) { 8002 *flg = PETSC_TRUE; 8003 } else if (!tol) { 8004 *flg = PETSC_FALSE; 8005 } else { 8006 if (!A->ops->ishermitian) { 8007 const MatType mattype; 8008 ierr = MatGetType(A,&mattype);CHKERRQ(ierr); 8009 SETERRQ1(PETSC_COMM_SELF,PETSC_ERR_SUP,"Matrix of type <%s> does not support checking for hermitian",mattype); 8010 } 8011 ierr = (*A->ops->ishermitian)(A,tol,flg);CHKERRQ(ierr); 8012 } 8013 PetscFunctionReturn(0); 8014 } 8015 8016 #undef __FUNCT__ 8017 #define __FUNCT__ "MatIsSymmetricKnown" 8018 /*@ 8019 MatIsSymmetricKnown - Checks the flag on the matrix to see if it is symmetric. 8020 8021 Not Collective 8022 8023 Input Parameter: 8024 . A - the matrix to check 8025 8026 Output Parameters: 8027 + set - if the symmetric flag is set (this tells you if the next flag is valid) 8028 - flg - the result 8029 8030 Level: advanced 8031 8032 Concepts: matrix^symmetry 8033 8034 Note: Does not check the matrix values directly, so this may return unknown (set = PETSC_FALSE). Use MatIsSymmetric() 8035 if you want it explicitly checked 8036 8037 .seealso: MatTranspose(), MatIsTranspose(), MatIsHermitian(), MatIsStructurallySymmetric(), MatSetOption(), MatIsSymmetric() 8038 @*/ 8039 PetscErrorCode MatIsSymmetricKnown(Mat A,PetscBool *set,PetscBool *flg) 8040 { 8041 PetscFunctionBegin; 8042 PetscValidHeaderSpecific(A,MAT_CLASSID,1); 8043 PetscValidPointer(set,2); 8044 PetscValidPointer(flg,3); 8045 if (A->symmetric_set) { 8046 *set = PETSC_TRUE; 8047 *flg = A->symmetric; 8048 } else { 8049 *set = PETSC_FALSE; 8050 } 8051 PetscFunctionReturn(0); 8052 } 8053 8054 #undef __FUNCT__ 8055 #define __FUNCT__ "MatIsHermitianKnown" 8056 /*@ 8057 MatIsHermitianKnown - Checks the flag on the matrix to see if it is hermitian. 8058 8059 Not Collective 8060 8061 Input Parameter: 8062 . A - the matrix to check 8063 8064 Output Parameters: 8065 + set - if the hermitian flag is set (this tells you if the next flag is valid) 8066 - flg - the result 8067 8068 Level: advanced 8069 8070 Concepts: matrix^symmetry 8071 8072 Note: Does not check the matrix values directly, so this may return unknown (set = PETSC_FALSE). Use MatIsHermitian() 8073 if you want it explicitly checked 8074 8075 .seealso: MatTranspose(), MatIsTranspose(), MatIsHermitian(), MatIsStructurallySymmetric(), MatSetOption(), MatIsSymmetric() 8076 @*/ 8077 PetscErrorCode MatIsHermitianKnown(Mat A,PetscBool *set,PetscBool *flg) 8078 { 8079 PetscFunctionBegin; 8080 PetscValidHeaderSpecific(A,MAT_CLASSID,1); 8081 PetscValidPointer(set,2); 8082 PetscValidPointer(flg,3); 8083 if (A->hermitian_set) { 8084 *set = PETSC_TRUE; 8085 *flg = A->hermitian; 8086 } else { 8087 *set = PETSC_FALSE; 8088 } 8089 PetscFunctionReturn(0); 8090 } 8091 8092 #undef __FUNCT__ 8093 #define __FUNCT__ "MatIsStructurallySymmetric" 8094 /*@ 8095 MatIsStructurallySymmetric - Test whether a matrix is structurally symmetric 8096 8097 Collective on Mat 8098 8099 Input Parameter: 8100 . A - the matrix to test 8101 8102 Output Parameters: 8103 . flg - the result 8104 8105 Level: intermediate 8106 8107 Concepts: matrix^symmetry 8108 8109 .seealso: MatTranspose(), MatIsTranspose(), MatIsHermitian(), MatIsSymmetric(), MatSetOption() 8110 @*/ 8111 PetscErrorCode MatIsStructurallySymmetric(Mat A,PetscBool *flg) 8112 { 8113 PetscErrorCode ierr; 8114 8115 PetscFunctionBegin; 8116 PetscValidHeaderSpecific(A,MAT_CLASSID,1); 8117 PetscValidPointer(flg,2); 8118 if (!A->structurally_symmetric_set) { 8119 if (!A->ops->isstructurallysymmetric) SETERRQ(((PetscObject)A)->comm,PETSC_ERR_SUP,"Matrix does not support checking for structural symmetric"); 8120 ierr = (*A->ops->isstructurallysymmetric)(A,&A->structurally_symmetric);CHKERRQ(ierr); 8121 A->structurally_symmetric_set = PETSC_TRUE; 8122 } 8123 *flg = A->structurally_symmetric; 8124 PetscFunctionReturn(0); 8125 } 8126 8127 #undef __FUNCT__ 8128 #define __FUNCT__ "MatStashGetInfo" 8129 extern PetscErrorCode MatStashGetInfo_Private(MatStash*,PetscInt*,PetscInt*); 8130 /*@ 8131 MatStashGetInfo - Gets how many values are currently in the matrix stash, i.e. need 8132 to be communicated to other processors during the MatAssemblyBegin/End() process 8133 8134 Not collective 8135 8136 Input Parameter: 8137 . vec - the vector 8138 8139 Output Parameters: 8140 + nstash - the size of the stash 8141 . reallocs - the number of additional mallocs incurred. 8142 . bnstash - the size of the block stash 8143 - breallocs - the number of additional mallocs incurred.in the block stash 8144 8145 Level: advanced 8146 8147 .seealso: MatAssemblyBegin(), MatAssemblyEnd(), Mat, MatStashSetInitialSize() 8148 8149 @*/ 8150 PetscErrorCode MatStashGetInfo(Mat mat,PetscInt *nstash,PetscInt *reallocs,PetscInt *bnstash,PetscInt *breallocs) 8151 { 8152 PetscErrorCode ierr; 8153 PetscFunctionBegin; 8154 ierr = MatStashGetInfo_Private(&mat->stash,nstash,reallocs);CHKERRQ(ierr); 8155 ierr = MatStashGetInfo_Private(&mat->bstash,bnstash,breallocs);CHKERRQ(ierr); 8156 PetscFunctionReturn(0); 8157 } 8158 8159 #undef __FUNCT__ 8160 #define __FUNCT__ "MatGetVecs" 8161 /*@C 8162 MatGetVecs - Get vector(s) compatible with the matrix, i.e. with the same 8163 parallel layout 8164 8165 Collective on Mat 8166 8167 Input Parameter: 8168 . mat - the matrix 8169 8170 Output Parameter: 8171 + right - (optional) vector that the matrix can be multiplied against 8172 - left - (optional) vector that the matrix vector product can be stored in 8173 8174 Level: advanced 8175 8176 .seealso: MatCreate() 8177 @*/ 8178 PetscErrorCode MatGetVecs(Mat mat,Vec *right,Vec *left) 8179 { 8180 PetscErrorCode ierr; 8181 8182 PetscFunctionBegin; 8183 PetscValidHeaderSpecific(mat,MAT_CLASSID,1); 8184 PetscValidType(mat,1); 8185 MatCheckPreallocated(mat,1); 8186 if (mat->ops->getvecs) { 8187 ierr = (*mat->ops->getvecs)(mat,right,left);CHKERRQ(ierr); 8188 } else { 8189 PetscMPIInt size; 8190 ierr = MPI_Comm_size(((PetscObject)mat)->comm, &size);CHKERRQ(ierr); 8191 if (right) { 8192 ierr = VecCreate(((PetscObject)mat)->comm,right);CHKERRQ(ierr); 8193 ierr = VecSetSizes(*right,mat->cmap->n,PETSC_DETERMINE);CHKERRQ(ierr); 8194 ierr = VecSetBlockSize(*right,mat->rmap->bs);CHKERRQ(ierr); 8195 ierr = VecSetType(*right,VECSTANDARD);CHKERRQ(ierr); 8196 ierr = PetscLayoutReference(mat->cmap,&(*right)->map);CHKERRQ(ierr); 8197 } 8198 if (left) { 8199 ierr = VecCreate(((PetscObject)mat)->comm,left);CHKERRQ(ierr); 8200 ierr = VecSetSizes(*left,mat->rmap->n,PETSC_DETERMINE);CHKERRQ(ierr); 8201 ierr = VecSetBlockSize(*left,mat->rmap->bs);CHKERRQ(ierr); 8202 ierr = VecSetType(*left,VECSTANDARD);CHKERRQ(ierr); 8203 ierr = PetscLayoutReference(mat->rmap,&(*left)->map);CHKERRQ(ierr); 8204 } 8205 } 8206 PetscFunctionReturn(0); 8207 } 8208 8209 #undef __FUNCT__ 8210 #define __FUNCT__ "MatFactorInfoInitialize" 8211 /*@C 8212 MatFactorInfoInitialize - Initializes a MatFactorInfo data structure 8213 with default values. 8214 8215 Not Collective 8216 8217 Input Parameters: 8218 . info - the MatFactorInfo data structure 8219 8220 8221 Notes: The solvers are generally used through the KSP and PC objects, for example 8222 PCLU, PCILU, PCCHOLESKY, PCICC 8223 8224 Level: developer 8225 8226 .seealso: MatFactorInfo 8227 8228 Developer Note: fortran interface is not autogenerated as the f90 8229 interface defintion cannot be generated correctly [due to MatFactorInfo] 8230 8231 @*/ 8232 8233 PetscErrorCode MatFactorInfoInitialize(MatFactorInfo *info) 8234 { 8235 PetscErrorCode ierr; 8236 8237 PetscFunctionBegin; 8238 ierr = PetscMemzero(info,sizeof(MatFactorInfo));CHKERRQ(ierr); 8239 PetscFunctionReturn(0); 8240 } 8241 8242 #undef __FUNCT__ 8243 #define __FUNCT__ "MatPtAP" 8244 /*@ 8245 MatPtAP - Creates the matrix product C = P^T * A * P 8246 8247 Neighbor-wise Collective on Mat 8248 8249 Input Parameters: 8250 + A - the matrix 8251 . P - the projection matrix 8252 . scall - either MAT_INITIAL_MATRIX or MAT_REUSE_MATRIX 8253 - fill - expected fill as ratio of nnz(C)/(nnz(A) + nnz(P)) 8254 8255 Output Parameters: 8256 . C - the product matrix 8257 8258 Notes: 8259 C will be created and must be destroyed by the user with MatDestroy(). 8260 8261 This routine is currently only implemented for pairs of AIJ matrices and classes 8262 which inherit from AIJ. 8263 8264 Level: intermediate 8265 8266 .seealso: MatPtAPSymbolic(), MatPtAPNumeric(), MatMatMult(), MatRARt() 8267 @*/ 8268 PetscErrorCode MatPtAP(Mat A,Mat P,MatReuse scall,PetscReal fill,Mat *C) 8269 { 8270 PetscErrorCode ierr; 8271 8272 PetscFunctionBegin; 8273 PetscValidHeaderSpecific(A,MAT_CLASSID,1); 8274 PetscValidType(A,1); 8275 if (!A->assembled) SETERRQ(((PetscObject)A)->comm,PETSC_ERR_ARG_WRONGSTATE,"Not for unassembled matrix"); 8276 if (A->factortype) SETERRQ(((PetscObject)A)->comm,PETSC_ERR_ARG_WRONGSTATE,"Not for factored matrix"); 8277 PetscValidHeaderSpecific(P,MAT_CLASSID,2); 8278 PetscValidType(P,2); 8279 MatCheckPreallocated(P,2); 8280 if (!P->assembled) SETERRQ(((PetscObject)A)->comm,PETSC_ERR_ARG_WRONGSTATE,"Not for unassembled matrix"); 8281 if (P->factortype) SETERRQ(((PetscObject)A)->comm,PETSC_ERR_ARG_WRONGSTATE,"Not for factored matrix"); 8282 PetscValidPointer(C,3); 8283 if (P->rmap->N!=A->cmap->N) SETERRQ2(((PetscObject)A)->comm,PETSC_ERR_ARG_SIZ,"Matrix dimensions are incompatible, %D != %D",P->rmap->N,A->cmap->N); 8284 if (fill < 1.0) SETERRQ1(((PetscObject)A)->comm,PETSC_ERR_ARG_SIZ,"Expected fill=%G must be >= 1.0",fill); 8285 MatCheckPreallocated(A,1); 8286 8287 if (!A->ops->ptap) { 8288 const MatType mattype; 8289 ierr = MatGetType(A,&mattype);CHKERRQ(ierr); 8290 SETERRQ1(((PetscObject)A)->comm,PETSC_ERR_SUP,"Matrix of type <%s> does not support PtAP",mattype); 8291 } 8292 ierr = PetscLogEventBegin(MAT_PtAP,A,P,0,0);CHKERRQ(ierr); 8293 ierr = (*A->ops->ptap)(A,P,scall,fill,C);CHKERRQ(ierr); 8294 ierr = PetscLogEventEnd(MAT_PtAP,A,P,0,0);CHKERRQ(ierr); 8295 PetscFunctionReturn(0); 8296 } 8297 8298 #undef __FUNCT__ 8299 #define __FUNCT__ "MatPtAPNumeric" 8300 /*@ 8301 MatPtAPNumeric - Computes the matrix product C = P^T * A * P 8302 8303 Neighbor-wise Collective on Mat 8304 8305 Input Parameters: 8306 + A - the matrix 8307 - P - the projection matrix 8308 8309 Output Parameters: 8310 . C - the product matrix 8311 8312 Notes: 8313 C must have been created by calling MatPtAPSymbolic and must be destroyed by 8314 the user using MatDeatroy(). 8315 8316 This routine is currently only implemented for pairs of AIJ matrices and classes 8317 which inherit from AIJ. C will be of type MATAIJ. 8318 8319 Level: intermediate 8320 8321 .seealso: MatPtAP(), MatPtAPSymbolic(), MatMatMultNumeric() 8322 @*/ 8323 PetscErrorCode MatPtAPNumeric(Mat A,Mat P,Mat C) 8324 { 8325 PetscErrorCode ierr; 8326 8327 PetscFunctionBegin; 8328 PetscValidHeaderSpecific(A,MAT_CLASSID,1); 8329 PetscValidType(A,1); 8330 if (!A->assembled) SETERRQ(((PetscObject)A)->comm,PETSC_ERR_ARG_WRONGSTATE,"Not for unassembled matrix"); 8331 if (A->factortype) SETERRQ(((PetscObject)A)->comm,PETSC_ERR_ARG_WRONGSTATE,"Not for factored matrix"); 8332 PetscValidHeaderSpecific(P,MAT_CLASSID,2); 8333 PetscValidType(P,2); 8334 MatCheckPreallocated(P,2); 8335 if (!P->assembled) SETERRQ(((PetscObject)A)->comm,PETSC_ERR_ARG_WRONGSTATE,"Not for unassembled matrix"); 8336 if (P->factortype) SETERRQ(((PetscObject)A)->comm,PETSC_ERR_ARG_WRONGSTATE,"Not for factored matrix"); 8337 PetscValidHeaderSpecific(C,MAT_CLASSID,3); 8338 PetscValidType(C,3); 8339 MatCheckPreallocated(C,3); 8340 if (C->factortype) SETERRQ(((PetscObject)A)->comm,PETSC_ERR_ARG_WRONGSTATE,"Not for factored matrix"); 8341 if (P->cmap->N!=C->rmap->N) SETERRQ2(((PetscObject)A)->comm,PETSC_ERR_ARG_SIZ,"Matrix dimensions are incompatible, %D != %D",P->cmap->N,C->rmap->N); 8342 if (P->rmap->N!=A->cmap->N) SETERRQ2(((PetscObject)A)->comm,PETSC_ERR_ARG_SIZ,"Matrix dimensions are incompatible, %D != %D",P->rmap->N,A->cmap->N); 8343 if (A->rmap->N!=A->cmap->N) SETERRQ2(((PetscObject)A)->comm,PETSC_ERR_ARG_SIZ,"Matrix 'A' must be square, %D != %D",A->rmap->N,A->cmap->N); 8344 if (P->cmap->N!=C->cmap->N) SETERRQ2(((PetscObject)A)->comm,PETSC_ERR_ARG_SIZ,"Matrix dimensions are incompatible, %D != %D",P->cmap->N,C->cmap->N); 8345 MatCheckPreallocated(A,1); 8346 8347 ierr = PetscLogEventBegin(MAT_PtAPNumeric,A,P,0,0);CHKERRQ(ierr); 8348 ierr = (*A->ops->ptapnumeric)(A,P,C);CHKERRQ(ierr); 8349 ierr = PetscLogEventEnd(MAT_PtAPNumeric,A,P,0,0);CHKERRQ(ierr); 8350 PetscFunctionReturn(0); 8351 } 8352 8353 #undef __FUNCT__ 8354 #define __FUNCT__ "MatPtAPSymbolic" 8355 /*@ 8356 MatPtAPSymbolic - Creates the (i,j) structure of the matrix product C = P^T * A * P 8357 8358 Neighbor-wise Collective on Mat 8359 8360 Input Parameters: 8361 + A - the matrix 8362 - P - the projection matrix 8363 8364 Output Parameters: 8365 . C - the (i,j) structure of the product matrix 8366 8367 Notes: 8368 C will be created and must be destroyed by the user with MatDestroy(). 8369 8370 This routine is currently only implemented for pairs of SeqAIJ matrices and classes 8371 which inherit from SeqAIJ. C will be of type MATSEQAIJ. The product is computed using 8372 this (i,j) structure by calling MatPtAPNumeric(). 8373 8374 Level: intermediate 8375 8376 .seealso: MatPtAP(), MatPtAPNumeric(), MatMatMultSymbolic() 8377 @*/ 8378 PetscErrorCode MatPtAPSymbolic(Mat A,Mat P,PetscReal fill,Mat *C) 8379 { 8380 PetscErrorCode ierr; 8381 8382 PetscFunctionBegin; 8383 PetscValidHeaderSpecific(A,MAT_CLASSID,1); 8384 PetscValidType(A,1); 8385 if (!A->assembled) SETERRQ(((PetscObject)A)->comm,PETSC_ERR_ARG_WRONGSTATE,"Not for unassembled matrix"); 8386 if (A->factortype) SETERRQ(((PetscObject)A)->comm,PETSC_ERR_ARG_WRONGSTATE,"Not for factored matrix"); 8387 if (fill <1.0) SETERRQ1(((PetscObject)A)->comm,PETSC_ERR_ARG_SIZ,"Expected fill=%G must be >= 1.0",fill); 8388 PetscValidHeaderSpecific(P,MAT_CLASSID,2); 8389 PetscValidType(P,2); 8390 MatCheckPreallocated(P,2); 8391 if (!P->assembled) SETERRQ(((PetscObject)A)->comm,PETSC_ERR_ARG_WRONGSTATE,"Not for unassembled matrix"); 8392 if (P->factortype) SETERRQ(((PetscObject)A)->comm,PETSC_ERR_ARG_WRONGSTATE,"Not for factored matrix"); 8393 PetscValidPointer(C,3); 8394 8395 if (P->rmap->N!=A->cmap->N) SETERRQ2(((PetscObject)A)->comm,PETSC_ERR_ARG_SIZ,"Matrix dimensions are incompatible, %D != %D",P->rmap->N,A->cmap->N); 8396 if (A->rmap->N!=A->cmap->N) SETERRQ2(((PetscObject)A)->comm,PETSC_ERR_ARG_SIZ,"Matrix 'A' must be square, %D != %D",A->rmap->N,A->cmap->N); 8397 MatCheckPreallocated(A,1); 8398 ierr = PetscLogEventBegin(MAT_PtAPSymbolic,A,P,0,0);CHKERRQ(ierr); 8399 ierr = (*A->ops->ptapsymbolic)(A,P,fill,C);CHKERRQ(ierr); 8400 ierr = PetscLogEventEnd(MAT_PtAPSymbolic,A,P,0,0);CHKERRQ(ierr); 8401 8402 /* ierr = MatSetBlockSize(*C,A->rmap->bs);CHKERRQ(ierr); NO! this is not always true -ma */ 8403 8404 PetscFunctionReturn(0); 8405 } 8406 8407 #undef __FUNCT__ 8408 #define __FUNCT__ "MatRARt" 8409 /*@ 8410 MatRARt - Creates the matrix product C = R * A * R^T 8411 8412 Neighbor-wise Collective on Mat 8413 8414 Input Parameters: 8415 + A - the matrix 8416 . R - the projection matrix 8417 . scall - either MAT_INITIAL_MATRIX or MAT_REUSE_MATRIX 8418 - fill - expected fill as ratio of nnz(C)/nnz(A) 8419 8420 Output Parameters: 8421 . C - the product matrix 8422 8423 Notes: 8424 C will be created and must be destroyed by the user with MatDestroy(). 8425 8426 This routine is currently only implemented for pairs of AIJ matrices and classes 8427 which inherit from AIJ. 8428 8429 Level: intermediate 8430 8431 .seealso: MatRARtSymbolic(), MatRARtNumeric(), MatMatMult(), MatPtAP() 8432 @*/ 8433 PetscErrorCode MatRARt(Mat A,Mat R,MatReuse scall,PetscReal fill,Mat *C) 8434 { 8435 PetscErrorCode ierr; 8436 8437 PetscFunctionBegin; 8438 PetscValidHeaderSpecific(A,MAT_CLASSID,1); 8439 PetscValidType(A,1); 8440 if (!A->assembled) SETERRQ(((PetscObject)A)->comm,PETSC_ERR_ARG_WRONGSTATE,"Not for unassembled matrix"); 8441 if (A->factortype) SETERRQ(((PetscObject)A)->comm,PETSC_ERR_ARG_WRONGSTATE,"Not for factored matrix"); 8442 PetscValidHeaderSpecific(R,MAT_CLASSID,2); 8443 PetscValidType(R,2); 8444 MatCheckPreallocated(R,2); 8445 if (!R->assembled) SETERRQ(((PetscObject)A)->comm,PETSC_ERR_ARG_WRONGSTATE,"Not for unassembled matrix"); 8446 if (R->factortype) SETERRQ(((PetscObject)A)->comm,PETSC_ERR_ARG_WRONGSTATE,"Not for factored matrix"); 8447 PetscValidPointer(C,3); 8448 if (R->cmap->N!=A->rmap->N) SETERRQ2(((PetscObject)R)->comm,PETSC_ERR_ARG_SIZ,"Matrix dimensions are incompatible, %D != %D",R->cmap->N,A->rmap->N); 8449 if (fill < 1.0) SETERRQ1(((PetscObject)A)->comm,PETSC_ERR_ARG_SIZ,"Expected fill=%G must be >= 1.0",fill); 8450 MatCheckPreallocated(A,1); 8451 8452 if (!A->ops->rart) { 8453 const MatType mattype; 8454 ierr = MatGetType(A,&mattype);CHKERRQ(ierr); 8455 SETERRQ1(((PetscObject)A)->comm,PETSC_ERR_SUP,"Matrix of type <%s> does not support RARt",mattype); 8456 } 8457 ierr = PetscLogEventBegin(MAT_RARt,A,R,0,0);CHKERRQ(ierr); 8458 ierr = (*A->ops->rart)(A,R,scall,fill,C);CHKERRQ(ierr); 8459 ierr = PetscLogEventEnd(MAT_RARt,A,R,0,0);CHKERRQ(ierr); 8460 PetscFunctionReturn(0); 8461 } 8462 8463 #undef __FUNCT__ 8464 #define __FUNCT__ "MatRARtNumeric" 8465 /*@ 8466 MatRARtNumeric - Computes the matrix product C = R * A * R^T 8467 8468 Neighbor-wise Collective on Mat 8469 8470 Input Parameters: 8471 + A - the matrix 8472 - R - the projection matrix 8473 8474 Output Parameters: 8475 . C - the product matrix 8476 8477 Notes: 8478 C must have been created by calling MatRARtSymbolic and must be destroyed by 8479 the user using MatDeatroy(). 8480 8481 This routine is currently only implemented for pairs of AIJ matrices and classes 8482 which inherit from AIJ. C will be of type MATAIJ. 8483 8484 Level: intermediate 8485 8486 .seealso: MatRARt(), MatRARtSymbolic(), MatMatMultNumeric() 8487 @*/ 8488 PetscErrorCode MatRARtNumeric(Mat A,Mat R,Mat C) 8489 { 8490 PetscErrorCode ierr; 8491 8492 PetscFunctionBegin; 8493 PetscValidHeaderSpecific(A,MAT_CLASSID,1); 8494 PetscValidType(A,1); 8495 if (!A->assembled) SETERRQ(((PetscObject)A)->comm,PETSC_ERR_ARG_WRONGSTATE,"Not for unassembled matrix"); 8496 if (A->factortype) SETERRQ(((PetscObject)A)->comm,PETSC_ERR_ARG_WRONGSTATE,"Not for factored matrix"); 8497 PetscValidHeaderSpecific(R,MAT_CLASSID,2); 8498 PetscValidType(R,2); 8499 MatCheckPreallocated(R,2); 8500 if (!R->assembled) SETERRQ(((PetscObject)A)->comm,PETSC_ERR_ARG_WRONGSTATE,"Not for unassembled matrix"); 8501 if (R->factortype) SETERRQ(((PetscObject)A)->comm,PETSC_ERR_ARG_WRONGSTATE,"Not for factored matrix"); 8502 PetscValidHeaderSpecific(C,MAT_CLASSID,3); 8503 PetscValidType(C,3); 8504 MatCheckPreallocated(C,3); 8505 if (C->factortype) SETERRQ(((PetscObject)A)->comm,PETSC_ERR_ARG_WRONGSTATE,"Not for factored matrix"); 8506 if (R->rmap->N!=C->rmap->N) SETERRQ2(((PetscObject)A)->comm,PETSC_ERR_ARG_SIZ,"Matrix dimensions are incompatible, %D != %D",R->rmap->N,C->rmap->N); 8507 if (R->cmap->N!=A->rmap->N) SETERRQ2(((PetscObject)A)->comm,PETSC_ERR_ARG_SIZ,"Matrix dimensions are incompatible, %D != %D",R->cmap->N,A->rmap->N); 8508 if (A->rmap->N!=A->cmap->N) SETERRQ2(((PetscObject)A)->comm,PETSC_ERR_ARG_SIZ,"Matrix 'A' must be square, %D != %D",A->rmap->N,A->cmap->N); 8509 if (R->rmap->N!=C->cmap->N) SETERRQ2(((PetscObject)A)->comm,PETSC_ERR_ARG_SIZ,"Matrix dimensions are incompatible, %D != %D",R->rmap->N,C->cmap->N); 8510 MatCheckPreallocated(A,1); 8511 8512 ierr = PetscLogEventBegin(MAT_RARtNumeric,A,R,0,0);CHKERRQ(ierr); 8513 ierr = (*A->ops->rartnumeric)(A,R,C);CHKERRQ(ierr); 8514 ierr = PetscLogEventEnd(MAT_RARtNumeric,A,R,0,0);CHKERRQ(ierr); 8515 PetscFunctionReturn(0); 8516 } 8517 8518 #undef __FUNCT__ 8519 #define __FUNCT__ "MatRARtSymbolic" 8520 /*@ 8521 MatRARtSymbolic - Creates the (i,j) structure of the matrix product C = R * A * R^T 8522 8523 Neighbor-wise Collective on Mat 8524 8525 Input Parameters: 8526 + A - the matrix 8527 - R - the projection matrix 8528 8529 Output Parameters: 8530 . C - the (i,j) structure of the product matrix 8531 8532 Notes: 8533 C will be created and must be destroyed by the user with MatDestroy(). 8534 8535 This routine is currently only implemented for pairs of SeqAIJ matrices and classes 8536 which inherit from SeqAIJ. C will be of type MATSEQAIJ. The product is computed using 8537 this (i,j) structure by calling MatRARtNumeric(). 8538 8539 Level: intermediate 8540 8541 .seealso: MatRARt(), MatRARtNumeric(), MatMatMultSymbolic() 8542 @*/ 8543 PetscErrorCode MatRARtSymbolic(Mat A,Mat R,PetscReal fill,Mat *C) 8544 { 8545 PetscErrorCode ierr; 8546 8547 PetscFunctionBegin; 8548 PetscValidHeaderSpecific(A,MAT_CLASSID,1); 8549 PetscValidType(A,1); 8550 if (!A->assembled) SETERRQ(((PetscObject)A)->comm,PETSC_ERR_ARG_WRONGSTATE,"Not for unassembled matrix"); 8551 if (A->factortype) SETERRQ(((PetscObject)A)->comm,PETSC_ERR_ARG_WRONGSTATE,"Not for factored matrix"); 8552 if (fill <1.0) SETERRQ1(((PetscObject)A)->comm,PETSC_ERR_ARG_SIZ,"Expected fill=%G must be >= 1.0",fill); 8553 PetscValidHeaderSpecific(R,MAT_CLASSID,2); 8554 PetscValidType(R,2); 8555 MatCheckPreallocated(R,2); 8556 if (!R->assembled) SETERRQ(((PetscObject)A)->comm,PETSC_ERR_ARG_WRONGSTATE,"Not for unassembled matrix"); 8557 if (R->factortype) SETERRQ(((PetscObject)A)->comm,PETSC_ERR_ARG_WRONGSTATE,"Not for factored matrix"); 8558 PetscValidPointer(C,3); 8559 8560 if (R->cmap->N!=A->rmap->N) SETERRQ2(((PetscObject)A)->comm,PETSC_ERR_ARG_SIZ,"Matrix dimensions are incompatible, %D != %D",R->cmap->N,A->rmap->N); 8561 if (A->rmap->N!=A->cmap->N) SETERRQ2(((PetscObject)A)->comm,PETSC_ERR_ARG_SIZ,"Matrix 'A' must be square, %D != %D",A->rmap->N,A->cmap->N); 8562 MatCheckPreallocated(A,1); 8563 ierr = PetscLogEventBegin(MAT_RARtSymbolic,A,R,0,0);CHKERRQ(ierr); 8564 ierr = (*A->ops->rartsymbolic)(A,R,fill,C);CHKERRQ(ierr); 8565 ierr = PetscLogEventEnd(MAT_RARtSymbolic,A,R,0,0);CHKERRQ(ierr); 8566 8567 ierr = MatSetBlockSize(*C,A->rmap->bs);CHKERRQ(ierr); 8568 PetscFunctionReturn(0); 8569 } 8570 8571 extern PetscErrorCode MatQueryOp(MPI_Comm comm, void (**function)(void), const char op[], PetscInt numArgs, ...); 8572 8573 #undef __FUNCT__ 8574 #define __FUNCT__ "MatMatMult" 8575 /*@ 8576 MatMatMult - Performs Matrix-Matrix Multiplication C=A*B. 8577 8578 Neighbor-wise Collective on Mat 8579 8580 Input Parameters: 8581 + A - the left matrix 8582 . B - the right matrix 8583 . scall - either MAT_INITIAL_MATRIX or MAT_REUSE_MATRIX 8584 - fill - expected fill as ratio of nnz(C)/(nnz(A) + nnz(B)), use PETSC_DEFAULT if you do not have a good estimate 8585 if the result is a dense matrix this is irrelevent 8586 8587 Output Parameters: 8588 . C - the product matrix 8589 8590 Notes: 8591 Unless scall is MAT_REUSE_MATRIX C will be created. 8592 8593 MAT_REUSE_MATRIX can only be used if the matrices A and B have the same nonzero pattern as in the previous call 8594 8595 To determine the correct fill value, run with -info and search for the string "Fill ratio" to see the value 8596 actually needed. 8597 8598 If you have many matrices with the same non-zero structure to multiply, you 8599 should either 8600 $ 1) use MAT_REUSE_MATRIX in all calls but the first or 8601 $ 2) call MatMatMultSymbolic() once and then MatMatMultNumeric() for each product needed 8602 8603 Level: intermediate 8604 8605 .seealso: MatMatMultSymbolic(), MatMatMultNumeric(), MatTransposeMatMult(), MatMatTransposeMult(), MatPtAP() 8606 @*/ 8607 PetscErrorCode MatMatMult(Mat A,Mat B,MatReuse scall,PetscReal fill,Mat *C) 8608 { 8609 PetscErrorCode ierr; 8610 PetscErrorCode (*fA)(Mat,Mat,MatReuse,PetscReal,Mat*); 8611 PetscErrorCode (*fB)(Mat,Mat,MatReuse,PetscReal,Mat*); 8612 PetscErrorCode (*mult)(Mat,Mat,MatReuse,PetscReal,Mat *)=PETSC_NULL; 8613 8614 PetscFunctionBegin; 8615 PetscValidHeaderSpecific(A,MAT_CLASSID,1); 8616 PetscValidType(A,1); 8617 if (!A->assembled) SETERRQ(((PetscObject)A)->comm,PETSC_ERR_ARG_WRONGSTATE,"Not for unassembled matrix"); 8618 if (A->factortype) SETERRQ(((PetscObject)A)->comm,PETSC_ERR_ARG_WRONGSTATE,"Not for factored matrix"); 8619 PetscValidHeaderSpecific(B,MAT_CLASSID,2); 8620 PetscValidType(B,2); 8621 MatCheckPreallocated(B,2); 8622 if (!B->assembled) SETERRQ(((PetscObject)A)->comm,PETSC_ERR_ARG_WRONGSTATE,"Not for unassembled matrix"); 8623 if (B->factortype) SETERRQ(((PetscObject)A)->comm,PETSC_ERR_ARG_WRONGSTATE,"Not for factored matrix"); 8624 PetscValidPointer(C,3); 8625 if (B->rmap->N!=A->cmap->N) SETERRQ2(((PetscObject)A)->comm,PETSC_ERR_ARG_SIZ,"Matrix dimensions are incompatible, %D != %D",B->rmap->N,A->cmap->N); 8626 if (scall == MAT_REUSE_MATRIX){ 8627 PetscValidPointer(*C,5); 8628 PetscValidHeaderSpecific(*C,MAT_CLASSID,5); 8629 ierr = PetscLogEventBegin(MAT_MatMult,A,B,0,0);CHKERRQ(ierr); 8630 ierr = (*(*C)->ops->matmult)(A,B,scall,fill,C);CHKERRQ(ierr); 8631 ierr = PetscLogEventEnd(MAT_MatMult,A,B,0,0);CHKERRQ(ierr); 8632 } 8633 if (fill == PETSC_DEFAULT || fill == PETSC_DECIDE) fill = 2.0; 8634 if (fill < 1.0) SETERRQ1(((PetscObject)A)->comm,PETSC_ERR_ARG_SIZ,"Expected fill=%G must be >= 1.0",fill); 8635 MatCheckPreallocated(A,1); 8636 8637 fA = A->ops->matmult; 8638 fB = B->ops->matmult; 8639 if (fB == fA) { 8640 if (!fB) SETERRQ1(((PetscObject)A)->comm,PETSC_ERR_SUP,"MatMatMult not supported for B of type %s",((PetscObject)B)->type_name); 8641 mult = fB; 8642 } else { 8643 /* dispatch based on the type of A and B from their PetscObject's PetscFLists. */ 8644 char multname[256]; 8645 ierr = PetscStrcpy(multname,"MatMatMult_");CHKERRQ(ierr); 8646 ierr = PetscStrcat(multname,((PetscObject)A)->type_name);CHKERRQ(ierr); 8647 ierr = PetscStrcat(multname,"_");CHKERRQ(ierr); 8648 ierr = PetscStrcat(multname,((PetscObject)B)->type_name);CHKERRQ(ierr); 8649 ierr = PetscStrcat(multname,"_C");CHKERRQ(ierr); /* e.g., multname = "MatMatMult_seqdense_seqaij_C" */ 8650 ierr = PetscObjectQueryFunction((PetscObject)B,multname,(void (**)(void))&mult);CHKERRQ(ierr); 8651 if(!mult){ 8652 /* dual dispatch using MatQueryOp */ 8653 ierr = MatQueryOp(((PetscObject)A)->comm, (PetscVoidFunction*)(&mult), "MatMatMult",2,((PetscObject)A)->type_name,((PetscObject)B)->type_name); CHKERRQ(ierr); 8654 if (!mult) SETERRQ2(((PetscObject)A)->comm,PETSC_ERR_ARG_INCOMP,"MatMatMult requires A, %s, to be compatible with B, %s",((PetscObject)A)->type_name,((PetscObject)B)->type_name); 8655 } 8656 } 8657 ierr = PetscLogEventBegin(MAT_MatMult,A,B,0,0);CHKERRQ(ierr); 8658 ierr = (*mult)(A,B,scall,fill,C);CHKERRQ(ierr); 8659 ierr = PetscLogEventEnd(MAT_MatMult,A,B,0,0);CHKERRQ(ierr); 8660 PetscFunctionReturn(0); 8661 } 8662 8663 #undef __FUNCT__ 8664 #define __FUNCT__ "MatMatMultSymbolic" 8665 /*@ 8666 MatMatMultSymbolic - Performs construction, preallocation, and computes the ij structure 8667 of the matrix-matrix product C=A*B. Call this routine before calling MatMatMultNumeric(). 8668 8669 Neighbor-wise Collective on Mat 8670 8671 Input Parameters: 8672 + A - the left matrix 8673 . B - the right matrix 8674 - fill - expected fill as ratio of nnz(C)/(nnz(A) + nnz(B)), use PETSC_DEFAULT if you do not have a good estimate, 8675 if C is a dense matrix this is irrelevent 8676 8677 Output Parameters: 8678 . C - the product matrix 8679 8680 Notes: 8681 Unless scall is MAT_REUSE_MATRIX C will be created. 8682 8683 To determine the correct fill value, run with -info and search for the string "Fill ratio" to see the value 8684 actually needed. 8685 8686 This routine is currently implemented for 8687 - pairs of AIJ matrices and classes which inherit from AIJ, C will be of type AIJ 8688 - pairs of AIJ (A) and Dense (B) matrix, C will be of type Dense. 8689 - pairs of Dense (A) and AIJ (B) matrix, C will be of type Dense. 8690 8691 Level: intermediate 8692 8693 Developers Note: There are ways to estimate the number of nonzeros in the resulting product, see for example, http://arxiv.org/abs/1006.4173 8694 We should incorporate them into PETSc. 8695 8696 .seealso: MatMatMult(), MatMatMultNumeric() 8697 @*/ 8698 PetscErrorCode MatMatMultSymbolic(Mat A,Mat B,PetscReal fill,Mat *C) 8699 { 8700 PetscErrorCode ierr; 8701 PetscErrorCode (*Asymbolic)(Mat,Mat,PetscReal,Mat *); 8702 PetscErrorCode (*Bsymbolic)(Mat,Mat,PetscReal,Mat *); 8703 PetscErrorCode (*symbolic)(Mat,Mat,PetscReal,Mat *)=PETSC_NULL; 8704 8705 PetscFunctionBegin; 8706 PetscValidHeaderSpecific(A,MAT_CLASSID,1); 8707 PetscValidType(A,1); 8708 if (!A->assembled) SETERRQ(((PetscObject)A)->comm,PETSC_ERR_ARG_WRONGSTATE,"Not for unassembled matrix"); 8709 if (A->factortype) SETERRQ(((PetscObject)A)->comm,PETSC_ERR_ARG_WRONGSTATE,"Not for factored matrix"); 8710 8711 PetscValidHeaderSpecific(B,MAT_CLASSID,2); 8712 PetscValidType(B,2); 8713 MatCheckPreallocated(B,2); 8714 if (!B->assembled) SETERRQ(((PetscObject)A)->comm,PETSC_ERR_ARG_WRONGSTATE,"Not for unassembled matrix"); 8715 if (B->factortype) SETERRQ(((PetscObject)A)->comm,PETSC_ERR_ARG_WRONGSTATE,"Not for factored matrix"); 8716 PetscValidPointer(C,3); 8717 8718 if (B->rmap->N!=A->cmap->N) SETERRQ2(((PetscObject)A)->comm,PETSC_ERR_ARG_SIZ,"Matrix dimensions are incompatible, %D != %D",B->rmap->N,A->cmap->N); 8719 if (fill == PETSC_DEFAULT) fill = 2.0; 8720 if (fill < 1.0) SETERRQ1(((PetscObject)A)->comm,PETSC_ERR_ARG_SIZ,"Expected fill=%G must be > 1.0",fill); 8721 MatCheckPreallocated(A,1); 8722 8723 Asymbolic = A->ops->matmultsymbolic; 8724 Bsymbolic = B->ops->matmultsymbolic; 8725 if (Asymbolic == Bsymbolic){ 8726 if (!Bsymbolic) SETERRQ1(((PetscObject)A)->comm,PETSC_ERR_SUP,"C=A*B not implemented for B of type %s",((PetscObject)B)->type_name); 8727 symbolic = Bsymbolic; 8728 } else { /* dispatch based on the type of A and B */ 8729 char symbolicname[256]; 8730 ierr = PetscStrcpy(symbolicname,"MatMatMultSymbolic_");CHKERRQ(ierr); 8731 ierr = PetscStrcat(symbolicname,((PetscObject)A)->type_name);CHKERRQ(ierr); 8732 ierr = PetscStrcat(symbolicname,"_");CHKERRQ(ierr); 8733 ierr = PetscStrcat(symbolicname,((PetscObject)B)->type_name);CHKERRQ(ierr); 8734 ierr = PetscStrcat(symbolicname,"_C");CHKERRQ(ierr); 8735 ierr = PetscObjectQueryFunction((PetscObject)B,symbolicname,(void (**)(void))&symbolic);CHKERRQ(ierr); 8736 if (!symbolic) SETERRQ2(((PetscObject)A)->comm,PETSC_ERR_ARG_INCOMP,"MatMatMultSymbolic requires A, %s, to be compatible with B, %s",((PetscObject)A)->type_name,((PetscObject)B)->type_name); 8737 } 8738 ierr = PetscLogEventBegin(MAT_MatMultSymbolic,A,B,0,0);CHKERRQ(ierr); 8739 ierr = (*symbolic)(A,B,fill,C);CHKERRQ(ierr); 8740 ierr = PetscLogEventEnd(MAT_MatMultSymbolic,A,B,0,0);CHKERRQ(ierr); 8741 PetscFunctionReturn(0); 8742 } 8743 8744 #undef __FUNCT__ 8745 #define __FUNCT__ "MatMatMultNumeric" 8746 /*@ 8747 MatMatMultNumeric - Performs the numeric matrix-matrix product. 8748 Call this routine after first calling MatMatMultSymbolic(). 8749 8750 Neighbor-wise Collective on Mat 8751 8752 Input Parameters: 8753 + A - the left matrix 8754 - B - the right matrix 8755 8756 Output Parameters: 8757 . C - the product matrix, which was created by from MatMatMultSymbolic() or a call to MatMatMult(). 8758 8759 Notes: 8760 C must have been created with MatMatMultSymbolic(). 8761 8762 This routine is currently implemented for 8763 - pairs of AIJ matrices and classes which inherit from AIJ, C will be of type MATAIJ. 8764 - pairs of AIJ (A) and Dense (B) matrix, C will be of type Dense. 8765 - pairs of Dense (A) and AIJ (B) matrix, C will be of type Dense. 8766 8767 Level: intermediate 8768 8769 .seealso: MatMatMult(), MatMatMultSymbolic() 8770 @*/ 8771 PetscErrorCode MatMatMultNumeric(Mat A,Mat B,Mat C) 8772 { 8773 PetscErrorCode ierr; 8774 PetscErrorCode (*Anumeric)(Mat,Mat,Mat); 8775 PetscErrorCode (*Bnumeric)(Mat,Mat,Mat); 8776 PetscErrorCode (*numeric)(Mat,Mat,Mat)=PETSC_NULL; 8777 8778 PetscFunctionBegin; 8779 PetscValidHeaderSpecific(A,MAT_CLASSID,1); 8780 PetscValidType(A,1); 8781 if (!A->assembled) SETERRQ(((PetscObject)A)->comm,PETSC_ERR_ARG_WRONGSTATE,"Not for unassembled matrix"); 8782 if (A->factortype) SETERRQ(((PetscObject)A)->comm,PETSC_ERR_ARG_WRONGSTATE,"Not for factored matrix"); 8783 8784 PetscValidHeaderSpecific(B,MAT_CLASSID,2); 8785 PetscValidType(B,2); 8786 MatCheckPreallocated(B,2); 8787 if (!B->assembled) SETERRQ(((PetscObject)A)->comm,PETSC_ERR_ARG_WRONGSTATE,"Not for unassembled matrix"); 8788 if (B->factortype) SETERRQ(((PetscObject)A)->comm,PETSC_ERR_ARG_WRONGSTATE,"Not for factored matrix"); 8789 8790 PetscValidHeaderSpecific(C,MAT_CLASSID,3); 8791 PetscValidType(C,3); 8792 MatCheckPreallocated(C,3); 8793 if (!C->assembled) SETERRQ(((PetscObject)A)->comm,PETSC_ERR_ARG_WRONGSTATE,"Not for unassembled matrix"); 8794 if (C->factortype) SETERRQ(((PetscObject)A)->comm,PETSC_ERR_ARG_WRONGSTATE,"Not for factored matrix"); 8795 8796 if (B->cmap->N!=C->cmap->N) SETERRQ2(((PetscObject)A)->comm,PETSC_ERR_ARG_SIZ,"Matrix dimensions are incompatible, %D != %D",B->cmap->N,C->cmap->N); 8797 if (B->rmap->N!=A->cmap->N) SETERRQ2(((PetscObject)A)->comm,PETSC_ERR_ARG_SIZ,"Matrix dimensions are incompatible, %D != %D",B->rmap->N,A->cmap->N); 8798 if (A->rmap->N!=C->rmap->N) SETERRQ2(((PetscObject)A)->comm,PETSC_ERR_ARG_SIZ,"Matrix dimensions are incompatible, %D != %D",A->rmap->N,C->rmap->N); 8799 MatCheckPreallocated(A,1); 8800 8801 Anumeric = A->ops->matmultnumeric; 8802 Bnumeric = B->ops->matmultnumeric; 8803 if (Anumeric == Bnumeric){ 8804 if (!Bnumeric) SETERRQ1(((PetscObject)A)->comm,PETSC_ERR_SUP,"MatMatMultNumeric not supported for B of type %s",((PetscObject)B)->type_name); 8805 numeric = Bnumeric; 8806 } else { 8807 char numericname[256]; 8808 ierr = PetscStrcpy(numericname,"MatMatMultNumeric_");CHKERRQ(ierr); 8809 ierr = PetscStrcat(numericname,((PetscObject)A)->type_name);CHKERRQ(ierr); 8810 ierr = PetscStrcat(numericname,"_");CHKERRQ(ierr); 8811 ierr = PetscStrcat(numericname,((PetscObject)B)->type_name);CHKERRQ(ierr); 8812 ierr = PetscStrcat(numericname,"_C");CHKERRQ(ierr); 8813 ierr = PetscObjectQueryFunction((PetscObject)B,numericname,(void (**)(void))&numeric);CHKERRQ(ierr); 8814 if (!numeric) 8815 SETERRQ2(((PetscObject)A)->comm,PETSC_ERR_ARG_INCOMP,"MatMatMultNumeric requires A, %s, to be compatible with B, %s",((PetscObject)A)->type_name,((PetscObject)B)->type_name); 8816 } 8817 ierr = PetscLogEventBegin(MAT_MatMultNumeric,A,B,0,0);CHKERRQ(ierr); 8818 ierr = (*numeric)(A,B,C);CHKERRQ(ierr); 8819 ierr = PetscLogEventEnd(MAT_MatMultNumeric,A,B,0,0);CHKERRQ(ierr); 8820 PetscFunctionReturn(0); 8821 } 8822 8823 #undef __FUNCT__ 8824 #define __FUNCT__ "MatMatTransposeMult" 8825 /*@ 8826 MatMatTransposeMult - Performs Matrix-Matrix Multiplication C=A*B^T. 8827 8828 Neighbor-wise Collective on Mat 8829 8830 Input Parameters: 8831 + A - the left matrix 8832 . B - the right matrix 8833 . scall - either MAT_INITIAL_MATRIX or MAT_REUSE_MATRIX 8834 - fill - expected fill as ratio of nnz(C)/(nnz(A) + nnz(B)), use PETSC_DEFAULT if not known 8835 8836 Output Parameters: 8837 . C - the product matrix 8838 8839 Notes: 8840 C will be created if MAT_INITIAL_MATRIX and must be destroyed by the user with MatDestroy(). 8841 8842 MAT_REUSE_MATRIX can only be used if the matrices A and B have the same nonzero pattern as in the previous call 8843 8844 To determine the correct fill value, run with -info and search for the string "Fill ratio" to see the value 8845 actually needed. 8846 8847 This routine is currently only implemented for pairs of SeqAIJ matrices. C will be of type MATSEQAIJ. 8848 8849 Level: intermediate 8850 8851 .seealso: MatMatTransposeMultSymbolic(), MatMatTransposeMultNumeric(), MatMatMult(), MatTransposeMatMult() MatPtAP() 8852 @*/ 8853 PetscErrorCode MatMatTransposeMult(Mat A,Mat B,MatReuse scall,PetscReal fill,Mat *C) 8854 { 8855 PetscErrorCode ierr; 8856 PetscErrorCode (*fA)(Mat,Mat,MatReuse,PetscReal,Mat*); 8857 PetscErrorCode (*fB)(Mat,Mat,MatReuse,PetscReal,Mat*); 8858 8859 PetscFunctionBegin; 8860 PetscValidHeaderSpecific(A,MAT_CLASSID,1); 8861 PetscValidType(A,1); 8862 if (!A->assembled) SETERRQ(((PetscObject)A)->comm,PETSC_ERR_ARG_WRONGSTATE,"Not for unassembled matrix"); 8863 if (A->factortype) SETERRQ(((PetscObject)A)->comm,PETSC_ERR_ARG_WRONGSTATE,"Not for factored matrix"); 8864 PetscValidHeaderSpecific(B,MAT_CLASSID,2); 8865 PetscValidType(B,2); 8866 MatCheckPreallocated(B,2); 8867 if (!B->assembled) SETERRQ(((PetscObject)A)->comm,PETSC_ERR_ARG_WRONGSTATE,"Not for unassembled matrix"); 8868 if (B->factortype) SETERRQ(((PetscObject)A)->comm,PETSC_ERR_ARG_WRONGSTATE,"Not for factored matrix"); 8869 PetscValidPointer(C,3); 8870 if (B->cmap->N!=A->cmap->N) SETERRQ2(((PetscObject)A)->comm,PETSC_ERR_ARG_SIZ,"Matrix dimensions are incompatible, AN %D != BN %D",A->cmap->N,B->cmap->N); 8871 if (fill == PETSC_DEFAULT || fill == PETSC_DECIDE) fill = 2.0; 8872 if (fill < 1.0) SETERRQ1(((PetscObject)A)->comm,PETSC_ERR_ARG_SIZ,"Expected fill=%G must be > 1.0",fill); 8873 MatCheckPreallocated(A,1); 8874 8875 fA = A->ops->mattransposemult; 8876 if (!fA) SETERRQ1(((PetscObject)A)->comm,PETSC_ERR_SUP,"MatMatTransposeMult not supported for A of type %s",((PetscObject)A)->type_name); 8877 fB = B->ops->mattransposemult; 8878 if (!fB) SETERRQ1(((PetscObject)A)->comm,PETSC_ERR_SUP,"MatMatTransposeMult not supported for B of type %s",((PetscObject)B)->type_name); 8879 if (fB!=fA) SETERRQ2(((PetscObject)A)->comm,PETSC_ERR_ARG_INCOMP,"MatMatTransposeMult requires A, %s, to be compatible with B, %s",((PetscObject)A)->type_name,((PetscObject)B)->type_name); 8880 8881 if (scall == MAT_INITIAL_MATRIX){ 8882 ierr = PetscLogEventBegin(MAT_MatTransposeMultSymbolic,A,B,0,0);CHKERRQ(ierr); 8883 ierr = (*A->ops->mattransposemultsymbolic)(A,B,fill,C);CHKERRQ(ierr); 8884 ierr = PetscLogEventEnd(MAT_MatTransposeMultSymbolic,A,B,0,0);CHKERRQ(ierr); 8885 } 8886 ierr = PetscLogEventBegin(MAT_MatTransposeMultNumeric,A,B,0,0);CHKERRQ(ierr); 8887 ierr = (*A->ops->mattransposemultnumeric)(A,B,*C);CHKERRQ(ierr); 8888 ierr = PetscLogEventEnd(MAT_MatTransposeMultNumeric,A,B,0,0);CHKERRQ(ierr); 8889 PetscFunctionReturn(0); 8890 } 8891 8892 #undef __FUNCT__ 8893 #define __FUNCT__ "MatTransposeMatMult" 8894 /*@ 8895 MatTransposeMatMult - Performs Matrix-Matrix Multiplication C=A^T*B. 8896 8897 Neighbor-wise Collective on Mat 8898 8899 Input Parameters: 8900 + A - the left matrix 8901 . B - the right matrix 8902 . scall - either MAT_INITIAL_MATRIX or MAT_REUSE_MATRIX 8903 - fill - expected fill as ratio of nnz(C)/(nnz(A) + nnz(B)), use PETSC_DEFAULT if not known 8904 8905 Output Parameters: 8906 . C - the product matrix 8907 8908 Notes: 8909 C will be created if MAT_INITIAL_MATRIX and must be destroyed by the user with MatDestroy(). 8910 8911 MAT_REUSE_MATRIX can only be used if the matrices A and B have the same nonzero pattern as in the previous call 8912 8913 To determine the correct fill value, run with -info and search for the string "Fill ratio" to see the value 8914 actually needed. 8915 8916 This routine is currently implemented for pairs of AIJ matrices and pairs of SeqDense matrices and classes 8917 which inherit from SeqAIJ. C will be of same type as the input matrices. 8918 8919 Level: intermediate 8920 8921 .seealso: MatTransposeMatMultSymbolic(), MatTransposeMatMultNumeric(), MatMatMult(), MatMatTransposeMult(), MatPtAP() 8922 @*/ 8923 PetscErrorCode MatTransposeMatMult(Mat A,Mat B,MatReuse scall,PetscReal fill,Mat *C) 8924 { 8925 PetscErrorCode ierr; 8926 PetscErrorCode (*fA)(Mat,Mat,MatReuse,PetscReal,Mat*); 8927 PetscErrorCode (*fB)(Mat,Mat,MatReuse,PetscReal,Mat*); 8928 PetscErrorCode (*transposematmult)(Mat,Mat,MatReuse,PetscReal,Mat*); 8929 8930 PetscFunctionBegin; 8931 PetscValidHeaderSpecific(A,MAT_CLASSID,1); 8932 PetscValidType(A,1); 8933 if (!A->assembled) SETERRQ(((PetscObject)A)->comm,PETSC_ERR_ARG_WRONGSTATE,"Not for unassembled matrix"); 8934 if (A->factortype) SETERRQ(((PetscObject)A)->comm,PETSC_ERR_ARG_WRONGSTATE,"Not for factored matrix"); 8935 PetscValidHeaderSpecific(B,MAT_CLASSID,2); 8936 PetscValidType(B,2); 8937 MatCheckPreallocated(B,2); 8938 if (!B->assembled) SETERRQ(((PetscObject)A)->comm,PETSC_ERR_ARG_WRONGSTATE,"Not for unassembled matrix"); 8939 if (B->factortype) SETERRQ(((PetscObject)A)->comm,PETSC_ERR_ARG_WRONGSTATE,"Not for factored matrix"); 8940 PetscValidPointer(C,3); 8941 if (B->rmap->N!=A->rmap->N) SETERRQ2(((PetscObject)A)->comm,PETSC_ERR_ARG_SIZ,"Matrix dimensions are incompatible, %D != %D",B->rmap->N,A->rmap->N); 8942 if (fill == PETSC_DEFAULT || fill == PETSC_DECIDE) fill = 2.0; 8943 if (fill < 1.0) SETERRQ1(((PetscObject)A)->comm,PETSC_ERR_ARG_SIZ,"Expected fill=%G must be > 1.0",fill); 8944 MatCheckPreallocated(A,1); 8945 8946 fA = A->ops->transposematmult; 8947 if (!fA) SETERRQ1(((PetscObject)A)->comm,PETSC_ERR_SUP,"MatTransposeMatMult not supported for A of type %s",((PetscObject)A)->type_name); 8948 fB = B->ops->transposematmult; 8949 if (!fB) SETERRQ1(((PetscObject)A)->comm,PETSC_ERR_SUP,"MatTransposeMatMult not supported for B of type %s",((PetscObject)B)->type_name); 8950 if (fB==fA) { 8951 transposematmult = fA; 8952 } 8953 else { 8954 /* dual dispatch using MatQueryOp */ 8955 ierr = MatQueryOp(((PetscObject)A)->comm, (PetscVoidFunction*)(&transposematmult), "MatTansposeMatMult",2,((PetscObject)A)->type_name,((PetscObject)B)->type_name); CHKERRQ(ierr); 8956 if(!transposematmult) 8957 SETERRQ2(((PetscObject)A)->comm,PETSC_ERR_ARG_INCOMP,"MatTransposeMatMult requires A, %s, to be compatible with B, %s",((PetscObject)A)->type_name,((PetscObject)B)->type_name); 8958 } 8959 ierr = PetscLogEventBegin(MAT_TransposeMatMult,A,B,0,0);CHKERRQ(ierr); 8960 ierr = (*transposematmult)(A,B,scall,fill,C);CHKERRQ(ierr); 8961 ierr = PetscLogEventEnd(MAT_TransposeMatMult,A,B,0,0);CHKERRQ(ierr); 8962 PetscFunctionReturn(0); 8963 } 8964 8965 #undef __FUNCT__ 8966 #define __FUNCT__ "MatGetRedundantMatrix" 8967 /*@C 8968 MatGetRedundantMatrix - Create redundant matrices and put them into processors of subcommunicators. 8969 8970 Collective on Mat 8971 8972 Input Parameters: 8973 + mat - the matrix 8974 . nsubcomm - the number of subcommunicators (= number of redundant parallel or sequential matrices) 8975 . subcomm - MPI communicator split from the communicator where mat resides in 8976 . mlocal_red - number of local rows of the redundant matrix 8977 - reuse - either MAT_INITIAL_MATRIX or MAT_REUSE_MATRIX 8978 8979 Output Parameter: 8980 . matredundant - redundant matrix 8981 8982 Notes: 8983 MAT_REUSE_MATRIX can only be used when the nonzero structure of the 8984 original matrix has not changed from that last call to MatGetRedundantMatrix(). 8985 8986 This routine creates the duplicated matrices in subcommunicators; you should NOT create them before 8987 calling it. 8988 8989 Only MPIAIJ matrix is supported. 8990 8991 Level: advanced 8992 8993 Concepts: subcommunicator 8994 Concepts: duplicate matrix 8995 8996 .seealso: MatDestroy() 8997 @*/ 8998 PetscErrorCode MatGetRedundantMatrix(Mat mat,PetscInt nsubcomm,MPI_Comm subcomm,PetscInt mlocal_red,MatReuse reuse,Mat *matredundant) 8999 { 9000 PetscErrorCode ierr; 9001 9002 PetscFunctionBegin; 9003 PetscValidHeaderSpecific(mat,MAT_CLASSID,1); 9004 if (nsubcomm && reuse == MAT_REUSE_MATRIX) { 9005 PetscValidPointer(*matredundant,6); 9006 PetscValidHeaderSpecific(*matredundant,MAT_CLASSID,6); 9007 } 9008 if (!mat->ops->getredundantmatrix) SETERRQ1(((PetscObject)mat)->comm,PETSC_ERR_SUP,"Mat type %s",((PetscObject)mat)->type_name); 9009 if (!mat->assembled) SETERRQ(((PetscObject)mat)->comm,PETSC_ERR_ARG_WRONGSTATE,"Not for unassembled matrix"); 9010 if (mat->factortype) SETERRQ(((PetscObject)mat)->comm,PETSC_ERR_ARG_WRONGSTATE,"Not for factored matrix"); 9011 MatCheckPreallocated(mat,1); 9012 9013 ierr = PetscLogEventBegin(MAT_GetRedundantMatrix,mat,0,0,0);CHKERRQ(ierr); 9014 ierr = (*mat->ops->getredundantmatrix)(mat,nsubcomm,subcomm,mlocal_red,reuse,matredundant);CHKERRQ(ierr); 9015 ierr = PetscLogEventEnd(MAT_GetRedundantMatrix,mat,0,0,0);CHKERRQ(ierr); 9016 PetscFunctionReturn(0); 9017 } 9018 9019 #undef __FUNCT__ 9020 #define __FUNCT__ "MatGetMultiProcBlock" 9021 /*@C 9022 MatGetMultiProcBlock - Create multiple [bjacobi] 'parallel submatrices' from 9023 a given 'mat' object. Each submatrix can span multiple procs. 9024 9025 Collective on Mat 9026 9027 Input Parameters: 9028 + mat - the matrix 9029 . subcomm - the subcommunicator obtained by com_split(comm) 9030 - scall - either MAT_INITIAL_MATRIX or MAT_REUSE_MATRIX 9031 9032 Output Parameter: 9033 . subMat - 'parallel submatrices each spans a given subcomm 9034 9035 Notes: 9036 The submatrix partition across processors is dicated by 'subComm' a 9037 communicator obtained by com_split(comm). The comm_split 9038 is not restriced to be grouped with consequitive original ranks. 9039 9040 Due the comm_split() usage, the parallel layout of the submatrices 9041 map directly to the layout of the original matrix [wrt the local 9042 row,col partitioning]. So the original 'DiagonalMat' naturally maps 9043 into the 'DiagonalMat' of the subMat, hence it is used directly from 9044 the subMat. However the offDiagMat looses some columns - and this is 9045 reconstructed with MatSetValues() 9046 9047 Level: advanced 9048 9049 Concepts: subcommunicator 9050 Concepts: submatrices 9051 9052 .seealso: MatGetSubMatrices() 9053 @*/ 9054 PetscErrorCode MatGetMultiProcBlock(Mat mat, MPI_Comm subComm, MatReuse scall,Mat* subMat) 9055 { 9056 PetscErrorCode ierr; 9057 PetscMPIInt commsize,subCommSize; 9058 9059 PetscFunctionBegin; 9060 ierr = MPI_Comm_size(((PetscObject)mat)->comm,&commsize);CHKERRQ(ierr); 9061 ierr = MPI_Comm_size(subComm,&subCommSize);CHKERRQ(ierr); 9062 if (subCommSize > commsize) SETERRQ2(((PetscObject)mat)->comm,PETSC_ERR_ARG_OUTOFRANGE,"CommSize %D < SubCommZize %D",commsize,subCommSize); 9063 9064 ierr = PetscLogEventBegin(MAT_GetMultiProcBlock,mat,0,0,0);CHKERRQ(ierr); 9065 ierr = (*mat->ops->getmultiprocblock)(mat,subComm,scall,subMat);CHKERRQ(ierr); 9066 ierr = PetscLogEventEnd(MAT_GetMultiProcBlock,mat,0,0,0);CHKERRQ(ierr); 9067 PetscFunctionReturn(0); 9068 } 9069 9070 #undef __FUNCT__ 9071 #define __FUNCT__ "MatGetLocalSubMatrix" 9072 /*@ 9073 MatGetLocalSubMatrix - Gets a reference to a submatrix specified in local numbering 9074 9075 Not Collective 9076 9077 Input Arguments: 9078 mat - matrix to extract local submatrix from 9079 isrow - local row indices for submatrix 9080 iscol - local column indices for submatrix 9081 9082 Output Arguments: 9083 submat - the submatrix 9084 9085 Level: intermediate 9086 9087 Notes: 9088 The submat should be returned with MatRestoreLocalSubMatrix(). 9089 9090 Depending on the format of mat, the returned submat may not implement MatMult(). Its communicator may be 9091 the same as mat, it may be PETSC_COMM_SELF, or some other subcomm of mat's. 9092 9093 The submat always implements MatSetValuesLocal(). If isrow and iscol have the same block size, then 9094 MatSetValuesBlockedLocal() will also be implemented. 9095 9096 .seealso: MatRestoreLocalSubMatrix(), MatCreateLocalRef() 9097 @*/ 9098 PetscErrorCode MatGetLocalSubMatrix(Mat mat,IS isrow,IS iscol,Mat *submat) 9099 { 9100 PetscErrorCode ierr; 9101 9102 PetscFunctionBegin; 9103 PetscValidHeaderSpecific(mat,MAT_CLASSID,1); 9104 PetscValidHeaderSpecific(isrow,IS_CLASSID,2); 9105 PetscValidHeaderSpecific(iscol,IS_CLASSID,3); 9106 PetscCheckSameComm(isrow,2,iscol,3); 9107 PetscValidPointer(submat,4); 9108 9109 if (mat->ops->getlocalsubmatrix) { 9110 ierr = (*mat->ops->getlocalsubmatrix)(mat,isrow,iscol,submat);CHKERRQ(ierr); 9111 } else { 9112 ierr = MatCreateLocalRef(mat,isrow,iscol,submat);CHKERRQ(ierr); 9113 } 9114 PetscFunctionReturn(0); 9115 } 9116 9117 #undef __FUNCT__ 9118 #define __FUNCT__ "MatRestoreLocalSubMatrix" 9119 /*@ 9120 MatRestoreLocalSubMatrix - Restores a reference to a submatrix specified in local numbering 9121 9122 Not Collective 9123 9124 Input Arguments: 9125 mat - matrix to extract local submatrix from 9126 isrow - local row indices for submatrix 9127 iscol - local column indices for submatrix 9128 submat - the submatrix 9129 9130 Level: intermediate 9131 9132 .seealso: MatGetLocalSubMatrix() 9133 @*/ 9134 PetscErrorCode MatRestoreLocalSubMatrix(Mat mat,IS isrow,IS iscol,Mat *submat) 9135 { 9136 PetscErrorCode ierr; 9137 9138 PetscFunctionBegin; 9139 PetscValidHeaderSpecific(mat,MAT_CLASSID,1); 9140 PetscValidHeaderSpecific(isrow,IS_CLASSID,2); 9141 PetscValidHeaderSpecific(iscol,IS_CLASSID,3); 9142 PetscCheckSameComm(isrow,2,iscol,3); 9143 PetscValidPointer(submat,4); 9144 if (*submat) {PetscValidHeaderSpecific(*submat,MAT_CLASSID,4);} 9145 9146 if (mat->ops->restorelocalsubmatrix) { 9147 ierr = (*mat->ops->restorelocalsubmatrix)(mat,isrow,iscol,submat);CHKERRQ(ierr); 9148 } else { 9149 ierr = MatDestroy(submat);CHKERRQ(ierr); 9150 } 9151 *submat = PETSC_NULL; 9152 PetscFunctionReturn(0); 9153 } 9154 9155 /* --------------------------------------------------------*/ 9156 #undef __FUNCT__ 9157 #define __FUNCT__ "MatFindZeroDiagonals" 9158 /*@ 9159 MatFindZeroDiagonals - Finds all the rows of a matrix that have zero or no entry in the matrix 9160 9161 Collective on Mat 9162 9163 Input Parameter: 9164 . mat - the matrix 9165 9166 Output Parameter: 9167 . is - if any rows have zero diagonals this contains the list of them 9168 9169 Level: developer 9170 9171 Concepts: matrix-vector product 9172 9173 .seealso: MatMultTranspose(), MatMultAdd(), MatMultTransposeAdd() 9174 @*/ 9175 PetscErrorCode MatFindZeroDiagonals(Mat mat,IS *is) 9176 { 9177 PetscErrorCode ierr; 9178 9179 PetscFunctionBegin; 9180 PetscValidHeaderSpecific(mat,MAT_CLASSID,1); 9181 PetscValidType(mat,1); 9182 if (!mat->assembled) SETERRQ(((PetscObject)mat)->comm,PETSC_ERR_ARG_WRONGSTATE,"Not for unassembled matrix"); 9183 if (mat->factortype) SETERRQ(((PetscObject)mat)->comm,PETSC_ERR_ARG_WRONGSTATE,"Not for factored matrix"); 9184 9185 if (!mat->ops->findzerodiagonals) SETERRQ(((PetscObject)mat)->comm,PETSC_ERR_SUP,"This matrix type does not have a find zero diagonals defined"); 9186 ierr = (*mat->ops->findzerodiagonals)(mat,is);CHKERRQ(ierr); 9187 PetscFunctionReturn(0); 9188 } 9189 9190 #undef __FUNCT__ 9191 #define __FUNCT__ "MatInvertBlockDiagonal" 9192 /*@C 9193 MatInvertBlockDiagonal - Inverts the block diagonal entries. 9194 9195 Collective on Mat 9196 9197 Input Parameters: 9198 . mat - the matrix 9199 9200 Output Parameters: 9201 . values - the block inverses in column major order (FORTRAN-like) 9202 9203 Note: 9204 This routine is not available from Fortran. 9205 9206 Level: advanced 9207 @*/ 9208 PetscErrorCode MatInvertBlockDiagonal(Mat mat,const PetscScalar **values) 9209 { 9210 PetscErrorCode ierr; 9211 9212 PetscFunctionBegin; 9213 PetscValidHeaderSpecific(mat,MAT_CLASSID,1); 9214 if (!mat->assembled) SETERRQ(PETSC_COMM_SELF,PETSC_ERR_ARG_WRONGSTATE,"Not for unassembled matrix"); 9215 if (mat->factortype) SETERRQ(PETSC_COMM_SELF,PETSC_ERR_ARG_WRONGSTATE,"Not for factored matrix"); 9216 if (!mat->ops->invertblockdiagonal) SETERRQ(PETSC_COMM_SELF,PETSC_ERR_SUP,"Not supported"); 9217 ierr = (*mat->ops->invertblockdiagonal)(mat,values);CHKERRQ(ierr); 9218 PetscFunctionReturn(0); 9219 } 9220 9221 #undef __FUNCT__ 9222 #define __FUNCT__ "MatTransposeColoringDestroy" 9223 /*@C 9224 MatTransposeColoringDestroy - Destroys a coloring context for matrix product C=A*B^T that was created 9225 via MatTransposeColoringCreate(). 9226 9227 Collective on MatTransposeColoring 9228 9229 Input Parameter: 9230 . c - coloring context 9231 9232 Level: intermediate 9233 9234 .seealso: MatTransposeColoringCreate() 9235 @*/ 9236 PetscErrorCode MatTransposeColoringDestroy(MatTransposeColoring *c) 9237 { 9238 PetscErrorCode ierr; 9239 MatTransposeColoring matcolor=*c; 9240 9241 PetscFunctionBegin; 9242 if (!matcolor) PetscFunctionReturn(0); 9243 if (--((PetscObject)matcolor)->refct > 0) {matcolor = 0; PetscFunctionReturn(0);} 9244 9245 ierr = PetscFree(matcolor->ncolumns);CHKERRQ(ierr); 9246 ierr = PetscFree(matcolor->nrows);CHKERRQ(ierr); 9247 ierr = PetscFree(matcolor->colorforrow);CHKERRQ(ierr); 9248 ierr = PetscFree2(matcolor->rows,matcolor->columnsforspidx);CHKERRQ(ierr); 9249 ierr = PetscFree(matcolor->colorforcol);CHKERRQ(ierr); 9250 ierr = PetscFree(matcolor->columns);CHKERRQ(ierr); 9251 ierr = PetscHeaderDestroy(c);CHKERRQ(ierr); 9252 PetscFunctionReturn(0); 9253 } 9254 9255 #undef __FUNCT__ 9256 #define __FUNCT__ "MatTransColoringApplySpToDen" 9257 /*@C 9258 MatTransColoringApplySpToDen - Given a symbolic matrix product C=A*B^T for which 9259 a MatTransposeColoring context has been created, computes a dense B^T by Apply 9260 MatTransposeColoring to sparse B. 9261 9262 Collective on MatTransposeColoring 9263 9264 Input Parameters: 9265 + B - sparse matrix B 9266 . Btdense - symbolic dense matrix B^T 9267 - coloring - coloring context created with MatTransposeColoringCreate() 9268 9269 Output Parameter: 9270 . Btdense - dense matrix B^T 9271 9272 Options Database Keys: 9273 + -mat_transpose_coloring_view - Activates basic viewing or coloring 9274 . -mat_transpose_coloring_view_draw - Activates drawing of coloring 9275 - -mat_transpose_coloring_view_info - Activates viewing of coloring info 9276 9277 Level: intermediate 9278 9279 .seealso: MatTransposeColoringCreate(), MatTransposeColoringDestroy() 9280 9281 .keywords: coloring 9282 @*/ 9283 PetscErrorCode MatTransColoringApplySpToDen(MatTransposeColoring coloring,Mat B,Mat Btdense) 9284 { 9285 PetscErrorCode ierr; 9286 9287 PetscFunctionBegin; 9288 PetscValidHeaderSpecific(B,MAT_CLASSID,1); 9289 PetscValidHeaderSpecific(Btdense,MAT_CLASSID,2); 9290 PetscValidHeaderSpecific(coloring,MAT_TRANSPOSECOLORING_CLASSID,3); 9291 9292 if (!B->ops->transcoloringapplysptoden) SETERRQ1(PETSC_COMM_SELF,PETSC_ERR_SUP,"Not supported for this matrix type %s",((PetscObject)B)->type_name); 9293 ierr = (B->ops->transcoloringapplysptoden)(coloring,B,Btdense);CHKERRQ(ierr); 9294 PetscFunctionReturn(0); 9295 } 9296 9297 #undef __FUNCT__ 9298 #define __FUNCT__ "MatTransColoringApplyDenToSp" 9299 /*@C 9300 MatTransColoringApplyDenToSp - Given a symbolic matrix product Csp=A*B^T for which 9301 a MatTransposeColoring context has been created and a dense matrix Cden=A*Btdense 9302 in which Btdens is obtained from MatTransColoringApplySpToDen(), recover sparse matrix 9303 Csp from Cden. 9304 9305 Collective on MatTransposeColoring 9306 9307 Input Parameters: 9308 + coloring - coloring context created with MatTransposeColoringCreate() 9309 - Cden - matrix product of a sparse matrix and a dense matrix Btdense 9310 9311 Output Parameter: 9312 . Csp - sparse matrix 9313 9314 Options Database Keys: 9315 + -mat_multtranspose_coloring_view - Activates basic viewing or coloring 9316 . -mat_multtranspose_coloring_view_draw - Activates drawing of coloring 9317 - -mat_multtranspose_coloring_view_info - Activates viewing of coloring info 9318 9319 Level: intermediate 9320 9321 .seealso: MatTransposeColoringCreate(), MatTransposeColoringDestroy(), MatTransColoringApplySpToDen() 9322 9323 .keywords: coloring 9324 @*/ 9325 PetscErrorCode MatTransColoringApplyDenToSp(MatTransposeColoring matcoloring,Mat Cden,Mat Csp) 9326 { 9327 PetscErrorCode ierr; 9328 9329 PetscFunctionBegin; 9330 PetscValidHeaderSpecific(matcoloring,MAT_TRANSPOSECOLORING_CLASSID,1); 9331 PetscValidHeaderSpecific(Cden,MAT_CLASSID,2); 9332 PetscValidHeaderSpecific(Csp,MAT_CLASSID,3); 9333 9334 if (!Csp->ops->transcoloringapplydentosp) SETERRQ1(PETSC_COMM_SELF,PETSC_ERR_SUP,"Not supported for this matrix type %s",((PetscObject)Csp)->type_name); 9335 ierr = (Csp->ops->transcoloringapplydentosp)(matcoloring,Cden,Csp);CHKERRQ(ierr); 9336 PetscFunctionReturn(0); 9337 } 9338 9339 #undef __FUNCT__ 9340 #define __FUNCT__ "MatTransposeColoringCreate" 9341 /*@C 9342 MatTransposeColoringCreate - Creates a matrix coloring context for matrix product C=A*B^T. 9343 9344 Collective on Mat 9345 9346 Input Parameters: 9347 + mat - the matrix product C 9348 - iscoloring - the coloring of the matrix; usually obtained with MatGetColoring() or DMCreateColoring() 9349 9350 Output Parameter: 9351 . color - the new coloring context 9352 9353 Level: intermediate 9354 9355 .seealso: MatTransposeColoringDestroy(), MatTransposeColoringSetFromOptions(), MatTransColoringApplySpToDen(), 9356 MatTransColoringApplyDen()ToSp, MatTransposeColoringView(), 9357 @*/ 9358 PetscErrorCode MatTransposeColoringCreate(Mat mat,ISColoring iscoloring,MatTransposeColoring *color) 9359 { 9360 MatTransposeColoring c; 9361 MPI_Comm comm; 9362 PetscErrorCode ierr; 9363 9364 PetscFunctionBegin; 9365 ierr = PetscLogEventBegin(MAT_TransposeColoringCreate,mat,0,0,0);CHKERRQ(ierr); 9366 ierr = PetscObjectGetComm((PetscObject)mat,&comm);CHKERRQ(ierr); 9367 ierr = PetscHeaderCreate(c,_p_MatTransposeColoring,int,MAT_TRANSPOSECOLORING_CLASSID,0,"MatTransposeColoring","Matrix product C=A*B^T via coloring","Mat",comm,MatTransposeColoringDestroy,0);CHKERRQ(ierr); 9368 9369 c->ctype = iscoloring->ctype; 9370 if (mat->ops->transposecoloringcreate) { 9371 ierr = (*mat->ops->transposecoloringcreate)(mat,iscoloring,c);CHKERRQ(ierr); 9372 } else SETERRQ(((PetscObject)mat)->comm,PETSC_ERR_SUP,"Code not yet written for this matrix type"); 9373 9374 *color = c; 9375 ierr = PetscLogEventEnd(MAT_TransposeColoringCreate,mat,0,0,0);CHKERRQ(ierr); 9376 PetscFunctionReturn(0); 9377 } 9378