xref: /petsc/src/mat/interface/matrix.c (revision ef7bb5aa64da98453fa5fb4571451d73f643854f)
1 #define PETSCMAT_DLL
2 
3 /*
4    This is where the abstract matrix operations are defined
5 */
6 
7 #include "private/matimpl.h"        /*I "petscmat.h" I*/
8 #include "private/vecimpl.h"
9 
10 /* Logging support */
11 PetscCookie PETSCMAT_DLLEXPORT MAT_COOKIE;
12 PetscCookie PETSCMAT_DLLEXPORT MAT_FDCOLORING_COOKIE;
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_Relax, 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_ZeroEntries, MAT_Load, MAT_View, MAT_AXPY, MAT_FDColoringCreate;
21 PetscLogEvent  MAT_FDColoringApply,MAT_Transpose,MAT_FDColoringFunction;
22 PetscLogEvent  MAT_MatMult, MAT_MatMultSymbolic, MAT_MatMultNumeric;
23 PetscLogEvent  MAT_PtAP, MAT_PtAPSymbolic, MAT_PtAPNumeric;
24 PetscLogEvent  MAT_MatMultTranspose, MAT_MatMultTransposeSymbolic, MAT_MatMultTransposeNumeric;
25 PetscLogEvent  MAT_Getsymtranspose, MAT_Getsymtransreduced, MAT_Transpose_SeqAIJ, MAT_GetBrowsOfAcols;
26 PetscLogEvent  MAT_GetBrowsOfAocols, MAT_Getlocalmat, MAT_Getlocalmatcondensed, MAT_Seqstompi, MAT_Seqstompinum, MAT_Seqstompisym;
27 PetscLogEvent  MAT_Applypapt, MAT_Applypapt_numeric, MAT_Applypapt_symbolic, MAT_GetSequentialNonzeroStructure;
28 PetscLogEvent  MAT_ILUDTFactorSymbolic, MAT_ILUDTFactorNumeric;
29 
30 /* nasty global values for MatSetValue() */
31 PetscInt    PETSCMAT_DLLEXPORT MatSetValue_Row = 0;
32 PetscInt    PETSCMAT_DLLEXPORT MatSetValue_Column = 0;
33 PetscScalar PETSCMAT_DLLEXPORT MatSetValue_Value = 0.0;
34 
35 #undef __FUNCT__
36 #define __FUNCT__ "MatGetDiagonalBlock"
37 /*@
38    MatGetDiagonalBlock - Returns the part of the matrix associated with the on-process coupling
39 
40    Not Collective
41 
42    Input Parameters:
43 +  mat - the matrix
44 -  reuse - indicates you are passing in the a matrix and want it reused
45 
46    Output Parameters:
47 +   iscopy - indicates a copy of the diagonal matrix was created and you should use MatDestroy() on it
48 -   a - the diagonal part (which is a SEQUENTIAL matrix)
49 
50    Notes: see the manual page for MatCreateMPIAIJ() for more information on the "diagonal part" of the matrix
51 
52    Level: advanced
53 
54 @*/
55 PetscErrorCode PETSCMAT_DLLEXPORT MatGetDiagonalBlock(Mat A,PetscTruth *iscopy,MatReuse reuse,Mat *a)
56 {
57   PetscErrorCode ierr,(*f)(Mat,PetscTruth*,MatReuse,Mat*);
58   PetscMPIInt    size;
59 
60   PetscFunctionBegin;
61   PetscValidHeaderSpecific(A,MAT_COOKIE,1);
62   PetscValidType(A,1);
63   PetscValidPointer(iscopy,2);
64   PetscValidPointer(a,3);
65   if (!A->assembled) SETERRQ(PETSC_ERR_ARG_WRONGSTATE,"Not for unassembled matrix");
66   if (A->factor) SETERRQ(PETSC_ERR_ARG_WRONGSTATE,"Not for factored matrix");
67   ierr = MPI_Comm_size(((PetscObject)A)->comm,&size);CHKERRQ(ierr);
68   ierr = PetscObjectQueryFunction((PetscObject)A,"MatGetDiagonalBlock_C",(void (**)(void))&f);CHKERRQ(ierr);
69   if (f) {
70     ierr = (*f)(A,iscopy,reuse,a);CHKERRQ(ierr);
71   } else if (size == 1) {
72     *a = A;
73     *iscopy = PETSC_FALSE;
74   } else {
75     SETERRQ(PETSC_ERR_SUP,"Cannot get diagonal part for this matrix");
76   }
77   PetscFunctionReturn(0);
78 }
79 
80 #undef __FUNCT__
81 #define __FUNCT__ "MatRealPart"
82 /*@
83    MatRealPart - Zeros out the imaginary part of the matrix
84 
85    Collective on Mat
86 
87    Input Parameters:
88 .  mat - the matrix
89 
90    Level: advanced
91 
92 
93 .seealso: MatImaginaryPart()
94 @*/
95 PetscErrorCode PETSCMAT_DLLEXPORT MatRealPart(Mat mat)
96 {
97   PetscErrorCode ierr;
98 
99   PetscFunctionBegin;
100   PetscValidHeaderSpecific(mat,MAT_COOKIE,1);
101   PetscValidType(mat,1);
102   if (!mat->assembled) SETERRQ(PETSC_ERR_ARG_WRONGSTATE,"Not for unassembled matrix");
103   if (mat->factor) SETERRQ(PETSC_ERR_ARG_WRONGSTATE,"Not for factored matrix");
104   if (!mat->ops->realpart) SETERRQ1(PETSC_ERR_SUP,"Mat type %s",((PetscObject)mat)->type_name);
105   ierr = MatPreallocated(mat);CHKERRQ(ierr);
106   ierr = (*mat->ops->realpart)(mat);CHKERRQ(ierr);
107   PetscFunctionReturn(0);
108 }
109 
110 #undef __FUNCT__
111 #define __FUNCT__ "MatGetGhosts"
112 /*@C
113    MatGetGhosts - Get the global index of all ghost nodes defined by the sparse matrix
114 
115    Collective on Mat
116 
117    Input Parameter:
118 .  mat - the matrix
119 
120    Output Parameters:
121 +   nghosts - number of ghosts (note for BAIJ matrices there is one ghost for each block)
122 -   ghosts - the global indices of the ghost points
123 
124    Notes: the nghosts and ghosts are suitable to pass into VecCreateGhost()
125 
126    Level: advanced
127 
128 @*/
129 PetscErrorCode PETSCMAT_DLLEXPORT MatGetGhosts(Mat mat,PetscInt *nghosts,const PetscInt *ghosts[])
130 {
131   PetscErrorCode ierr;
132 
133   PetscFunctionBegin;
134   PetscValidHeaderSpecific(mat,MAT_COOKIE,1);
135   PetscValidType(mat,1);
136   if (!mat->assembled) SETERRQ(PETSC_ERR_ARG_WRONGSTATE,"Not for unassembled matrix");
137   if (mat->factor) SETERRQ(PETSC_ERR_ARG_WRONGSTATE,"Not for factored matrix");
138   if (!mat->ops->getghosts) {
139     if (nghosts) *nghosts = 0;
140     if (ghosts) *ghosts = 0;
141   } else {
142     ierr = (*mat->ops->getghosts)(mat,nghosts,ghosts);CHKERRQ(ierr);
143   }
144   PetscFunctionReturn(0);
145 }
146 
147 
148 #undef __FUNCT__
149 #define __FUNCT__ "MatImaginaryPart"
150 /*@
151    MatImaginaryPart - Moves the imaginary part of the matrix to the real part and zeros the imaginary part
152 
153    Collective on Mat
154 
155    Input Parameters:
156 .  mat - the matrix
157 
158    Level: advanced
159 
160 
161 .seealso: MatRealPart()
162 @*/
163 PetscErrorCode PETSCMAT_DLLEXPORT MatImaginaryPart(Mat mat)
164 {
165   PetscErrorCode ierr;
166 
167   PetscFunctionBegin;
168   PetscValidHeaderSpecific(mat,MAT_COOKIE,1);
169   PetscValidType(mat,1);
170   if (!mat->assembled) SETERRQ(PETSC_ERR_ARG_WRONGSTATE,"Not for unassembled matrix");
171   if (mat->factor) SETERRQ(PETSC_ERR_ARG_WRONGSTATE,"Not for factored matrix");
172   if (!mat->ops->imaginarypart) SETERRQ1(PETSC_ERR_SUP,"Mat type %s",((PetscObject)mat)->type_name);
173   ierr = MatPreallocated(mat);CHKERRQ(ierr);
174   ierr = (*mat->ops->imaginarypart)(mat);CHKERRQ(ierr);
175   PetscFunctionReturn(0);
176 }
177 
178 #undef __FUNCT__
179 #define __FUNCT__ "MatMissingDiagonal"
180 /*@
181    MatMissingDiagonal - Determine if sparse matrix is missing a diagonal entry (or block entry for BAIJ matrices)
182 
183    Collective on Mat
184 
185    Input Parameter:
186 .  mat - the matrix
187 
188    Output Parameters:
189 +  missing - is any diagonal missing
190 -  dd - first diagonal entry that is missing (optional)
191 
192    Level: advanced
193 
194 
195 .seealso: MatRealPart()
196 @*/
197 PetscErrorCode PETSCMAT_DLLEXPORT MatMissingDiagonal(Mat mat,PetscTruth *missing,PetscInt *dd)
198 {
199   PetscErrorCode ierr;
200 
201   PetscFunctionBegin;
202   PetscValidHeaderSpecific(mat,MAT_COOKIE,1);
203   PetscValidType(mat,1);
204   if (!mat->assembled) SETERRQ(PETSC_ERR_ARG_WRONGSTATE,"Not for unassembled matrix");
205   if (mat->factor) SETERRQ(PETSC_ERR_ARG_WRONGSTATE,"Not for factored matrix");
206   if (!mat->ops->missingdiagonal) SETERRQ1(PETSC_ERR_SUP,"Mat type %s",((PetscObject)mat)->type_name);
207   ierr = (*mat->ops->missingdiagonal)(mat,missing,dd);CHKERRQ(ierr);
208   PetscFunctionReturn(0);
209 }
210 
211 #undef __FUNCT__
212 #define __FUNCT__ "MatGetRow"
213 /*@C
214    MatGetRow - Gets a row of a matrix.  You MUST call MatRestoreRow()
215    for each row that you get to ensure that your application does
216    not bleed memory.
217 
218    Not Collective
219 
220    Input Parameters:
221 +  mat - the matrix
222 -  row - the row to get
223 
224    Output Parameters:
225 +  ncols -  if not NULL, the number of nonzeros in the row
226 .  cols - if not NULL, the column numbers
227 -  vals - if not NULL, the values
228 
229    Notes:
230    This routine is provided for people who need to have direct access
231    to the structure of a matrix.  We hope that we provide enough
232    high-level matrix routines that few users will need it.
233 
234    MatGetRow() always returns 0-based column indices, regardless of
235    whether the internal representation is 0-based (default) or 1-based.
236 
237    For better efficiency, set cols and/or vals to PETSC_NULL if you do
238    not wish to extract these quantities.
239 
240    The user can only examine the values extracted with MatGetRow();
241    the values cannot be altered.  To change the matrix entries, one
242    must use MatSetValues().
243 
244    You can only have one call to MatGetRow() outstanding for a particular
245    matrix at a time, per processor. MatGetRow() can only obtain rows
246    associated with the given processor, it cannot get rows from the
247    other processors; for that we suggest using MatGetSubMatrices(), then
248    MatGetRow() on the submatrix. The row indix passed to MatGetRows()
249    is in the global number of rows.
250 
251    Fortran Notes:
252    The calling sequence from Fortran is
253 .vb
254    MatGetRow(matrix,row,ncols,cols,values,ierr)
255          Mat     matrix (input)
256          integer row    (input)
257          integer ncols  (output)
258          integer cols(maxcols) (output)
259          double precision (or double complex) values(maxcols) output
260 .ve
261    where maxcols >= maximum nonzeros in any row of the matrix.
262 
263 
264    Caution:
265    Do not try to change the contents of the output arrays (cols and vals).
266    In some cases, this may corrupt the matrix.
267 
268    Level: advanced
269 
270    Concepts: matrices^row access
271 
272 .seealso: MatRestoreRow(), MatSetValues(), MatGetValues(), MatGetSubMatrices(), MatGetDiagonal()
273 @*/
274 PetscErrorCode PETSCMAT_DLLEXPORT MatGetRow(Mat mat,PetscInt row,PetscInt *ncols,const PetscInt *cols[],const PetscScalar *vals[])
275 {
276   PetscErrorCode ierr;
277   PetscInt       incols;
278 
279   PetscFunctionBegin;
280   PetscValidHeaderSpecific(mat,MAT_COOKIE,1);
281   PetscValidType(mat,1);
282   if (!mat->assembled) SETERRQ(PETSC_ERR_ARG_WRONGSTATE,"Not for unassembled matrix");
283   if (mat->factor) SETERRQ(PETSC_ERR_ARG_WRONGSTATE,"Not for factored matrix");
284   if (!mat->ops->getrow) SETERRQ1(PETSC_ERR_SUP,"Mat type %s",((PetscObject)mat)->type_name);
285   ierr = MatPreallocated(mat);CHKERRQ(ierr);
286   ierr = PetscLogEventBegin(MAT_GetRow,mat,0,0,0);CHKERRQ(ierr);
287   ierr = (*mat->ops->getrow)(mat,row,&incols,(PetscInt **)cols,(PetscScalar **)vals);CHKERRQ(ierr);
288   if (ncols) *ncols = incols;
289   ierr = PetscLogEventEnd(MAT_GetRow,mat,0,0,0);CHKERRQ(ierr);
290   PetscFunctionReturn(0);
291 }
292 
293 #undef __FUNCT__
294 #define __FUNCT__ "MatConjugate"
295 /*@
296    MatConjugate - replaces the matrix values with their complex conjugates
297 
298    Collective on Mat
299 
300    Input Parameters:
301 .  mat - the matrix
302 
303    Level: advanced
304 
305 .seealso:  VecConjugate()
306 @*/
307 PetscErrorCode PETSCMAT_DLLEXPORT MatConjugate(Mat mat)
308 {
309   PetscErrorCode ierr;
310 
311   PetscFunctionBegin;
312   PetscValidHeaderSpecific(mat,MAT_COOKIE,1);
313   if (!mat->assembled) SETERRQ(PETSC_ERR_ARG_WRONGSTATE,"Not for unassembled matrix");
314   if (!mat->ops->conjugate) SETERRQ(PETSC_ERR_SUP,"Not provided for this matrix format, send email to petsc-maint@mcs.anl.gov");
315   ierr = (*mat->ops->conjugate)(mat);CHKERRQ(ierr);
316   PetscFunctionReturn(0);
317 }
318 
319 #undef __FUNCT__
320 #define __FUNCT__ "MatRestoreRow"
321 /*@C
322    MatRestoreRow - Frees any temporary space allocated by MatGetRow().
323 
324    Not Collective
325 
326    Input Parameters:
327 +  mat - the matrix
328 .  row - the row to get
329 .  ncols, cols - the number of nonzeros and their columns
330 -  vals - if nonzero the column values
331 
332    Notes:
333    This routine should be called after you have finished examining the entries.
334 
335    Fortran Notes:
336    The calling sequence from Fortran is
337 .vb
338    MatRestoreRow(matrix,row,ncols,cols,values,ierr)
339       Mat     matrix (input)
340       integer row    (input)
341       integer ncols  (output)
342       integer cols(maxcols) (output)
343       double precision (or double complex) values(maxcols) output
344 .ve
345    Where maxcols >= maximum nonzeros in any row of the matrix.
346 
347    In Fortran MatRestoreRow() MUST be called after MatGetRow()
348    before another call to MatGetRow() can be made.
349 
350    Level: advanced
351 
352 .seealso:  MatGetRow()
353 @*/
354 PetscErrorCode PETSCMAT_DLLEXPORT MatRestoreRow(Mat mat,PetscInt row,PetscInt *ncols,const PetscInt *cols[],const PetscScalar *vals[])
355 {
356   PetscErrorCode ierr;
357 
358   PetscFunctionBegin;
359   PetscValidHeaderSpecific(mat,MAT_COOKIE,1);
360   PetscValidIntPointer(ncols,3);
361   if (!mat->assembled) SETERRQ(PETSC_ERR_ARG_WRONGSTATE,"Not for unassembled matrix");
362   if (!mat->ops->restorerow) PetscFunctionReturn(0);
363   ierr = (*mat->ops->restorerow)(mat,row,ncols,(PetscInt **)cols,(PetscScalar **)vals);CHKERRQ(ierr);
364   PetscFunctionReturn(0);
365 }
366 
367 #undef __FUNCT__
368 #define __FUNCT__ "MatGetRowUpperTriangular"
369 /*@
370    MatGetRowUpperTriangular - Sets a flag to enable calls to MatGetRow() for matrix in MATSBAIJ format.
371    You should call MatRestoreRowUpperTriangular() after calling MatGetRow/MatRestoreRow() to disable the flag.
372 
373    Not Collective
374 
375    Input Parameters:
376 +  mat - the matrix
377 
378    Notes:
379    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.
380 
381    Level: advanced
382 
383    Concepts: matrices^row access
384 
385 .seealso: MatRestoreRowRowUpperTriangular()
386 @*/
387 PetscErrorCode PETSCMAT_DLLEXPORT MatGetRowUpperTriangular(Mat mat)
388 {
389   PetscErrorCode ierr;
390 
391   PetscFunctionBegin;
392   PetscValidHeaderSpecific(mat,MAT_COOKIE,1);
393   PetscValidType(mat,1);
394   if (!mat->assembled) SETERRQ(PETSC_ERR_ARG_WRONGSTATE,"Not for unassembled matrix");
395   if (mat->factor) SETERRQ(PETSC_ERR_ARG_WRONGSTATE,"Not for factored matrix");
396   if (!mat->ops->getrowuppertriangular) SETERRQ1(PETSC_ERR_SUP,"Mat type %s",((PetscObject)mat)->type_name);
397   ierr = MatPreallocated(mat);CHKERRQ(ierr);
398   ierr = (*mat->ops->getrowuppertriangular)(mat);CHKERRQ(ierr);
399   PetscFunctionReturn(0);
400 }
401 
402 #undef __FUNCT__
403 #define __FUNCT__ "MatRestoreRowUpperTriangular"
404 /*@
405    MatRestoreRowUpperTriangular - Disable calls to MatGetRow() for matrix in MATSBAIJ format.
406 
407    Not Collective
408 
409    Input Parameters:
410 +  mat - the matrix
411 
412    Notes:
413    This routine should be called after you have finished MatGetRow/MatRestoreRow().
414 
415 
416    Level: advanced
417 
418 .seealso:  MatGetRowUpperTriangular()
419 @*/
420 PetscErrorCode PETSCMAT_DLLEXPORT MatRestoreRowUpperTriangular(Mat mat)
421 {
422   PetscErrorCode ierr;
423 
424   PetscFunctionBegin;
425   PetscValidHeaderSpecific(mat,MAT_COOKIE,1);
426   if (!mat->assembled) SETERRQ(PETSC_ERR_ARG_WRONGSTATE,"Not for unassembled matrix");
427   if (!mat->ops->restorerowuppertriangular) PetscFunctionReturn(0);
428   ierr = (*mat->ops->restorerowuppertriangular)(mat);CHKERRQ(ierr);
429   PetscFunctionReturn(0);
430 }
431 
432 #undef __FUNCT__
433 #define __FUNCT__ "MatSetOptionsPrefix"
434 /*@C
435    MatSetOptionsPrefix - Sets the prefix used for searching for all
436    Mat options in the database.
437 
438    Collective on Mat
439 
440    Input Parameter:
441 +  A - the Mat context
442 -  prefix - the prefix to prepend to all option names
443 
444    Notes:
445    A hyphen (-) must NOT be given at the beginning of the prefix name.
446    The first character of all runtime options is AUTOMATICALLY the hyphen.
447 
448    Level: advanced
449 
450 .keywords: Mat, set, options, prefix, database
451 
452 .seealso: MatSetFromOptions()
453 @*/
454 PetscErrorCode PETSCMAT_DLLEXPORT MatSetOptionsPrefix(Mat A,const char prefix[])
455 {
456   PetscErrorCode ierr;
457 
458   PetscFunctionBegin;
459   PetscValidHeaderSpecific(A,MAT_COOKIE,1);
460   ierr = PetscObjectSetOptionsPrefix((PetscObject)A,prefix);CHKERRQ(ierr);
461   PetscFunctionReturn(0);
462 }
463 
464 #undef __FUNCT__
465 #define __FUNCT__ "MatAppendOptionsPrefix"
466 /*@C
467    MatAppendOptionsPrefix - Appends to the prefix used for searching for all
468    Mat options in the database.
469 
470    Collective on Mat
471 
472    Input Parameters:
473 +  A - the Mat context
474 -  prefix - the prefix to prepend to all option names
475 
476    Notes:
477    A hyphen (-) must NOT be given at the beginning of the prefix name.
478    The first character of all runtime options is AUTOMATICALLY the hyphen.
479 
480    Level: advanced
481 
482 .keywords: Mat, append, options, prefix, database
483 
484 .seealso: MatGetOptionsPrefix()
485 @*/
486 PetscErrorCode PETSCMAT_DLLEXPORT MatAppendOptionsPrefix(Mat A,const char prefix[])
487 {
488   PetscErrorCode ierr;
489 
490   PetscFunctionBegin;
491   PetscValidHeaderSpecific(A,MAT_COOKIE,1);
492   ierr = PetscObjectAppendOptionsPrefix((PetscObject)A,prefix);CHKERRQ(ierr);
493   PetscFunctionReturn(0);
494 }
495 
496 #undef __FUNCT__
497 #define __FUNCT__ "MatGetOptionsPrefix"
498 /*@C
499    MatGetOptionsPrefix - Sets the prefix used for searching for all
500    Mat options in the database.
501 
502    Not Collective
503 
504    Input Parameter:
505 .  A - the Mat context
506 
507    Output Parameter:
508 .  prefix - pointer to the prefix string used
509 
510    Notes: On the fortran side, the user should pass in a string 'prefix' of
511    sufficient length to hold the prefix.
512 
513    Level: advanced
514 
515 .keywords: Mat, get, options, prefix, database
516 
517 .seealso: MatAppendOptionsPrefix()
518 @*/
519 PetscErrorCode PETSCMAT_DLLEXPORT MatGetOptionsPrefix(Mat A,const char *prefix[])
520 {
521   PetscErrorCode ierr;
522 
523   PetscFunctionBegin;
524   PetscValidHeaderSpecific(A,MAT_COOKIE,1);
525   ierr = PetscObjectGetOptionsPrefix((PetscObject)A,prefix);CHKERRQ(ierr);
526   PetscFunctionReturn(0);
527 }
528 
529 #undef __FUNCT__
530 #define __FUNCT__ "MatSetUp"
531 /*@
532    MatSetUp - Sets up the internal matrix data structures for the later use.
533 
534    Collective on Mat
535 
536    Input Parameters:
537 .  A - the Mat context
538 
539    Notes:
540    For basic use of the Mat classes the user need not explicitly call
541    MatSetUp(), since these actions will happen automatically.
542 
543    Level: advanced
544 
545 .keywords: Mat, setup
546 
547 .seealso: MatCreate(), MatDestroy()
548 @*/
549 PetscErrorCode PETSCMAT_DLLEXPORT MatSetUp(Mat A)
550 {
551   PetscMPIInt    size;
552   PetscErrorCode ierr;
553 
554   PetscFunctionBegin;
555   PetscValidHeaderSpecific(A,MAT_COOKIE,1);
556   if (!((PetscObject)A)->type_name) {
557     ierr = MPI_Comm_size(((PetscObject)A)->comm, &size);CHKERRQ(ierr);
558     if (size == 1) {
559       ierr = MatSetType(A, MATSEQAIJ);CHKERRQ(ierr);
560     } else {
561       ierr = MatSetType(A, MATMPIAIJ);CHKERRQ(ierr);
562     }
563   }
564   ierr = MatSetUpPreallocation(A);CHKERRQ(ierr);
565   PetscFunctionReturn(0);
566 }
567 
568 #undef __FUNCT__
569 #define __FUNCT__ "MatView"
570 /*@C
571    MatView - Visualizes a matrix object.
572 
573    Collective on Mat
574 
575    Input Parameters:
576 +  mat - the matrix
577 -  viewer - visualization context
578 
579   Notes:
580   The available visualization contexts include
581 +    PETSC_VIEWER_STDOUT_SELF - standard output (default)
582 .    PETSC_VIEWER_STDOUT_WORLD - synchronized standard
583         output where only the first processor opens
584         the file.  All other processors send their
585         data to the first processor to print.
586 -     PETSC_VIEWER_DRAW_WORLD - graphical display of nonzero structure
587 
588    The user can open alternative visualization contexts with
589 +    PetscViewerASCIIOpen() - Outputs matrix to a specified file
590 .    PetscViewerBinaryOpen() - Outputs matrix in binary to a
591          specified file; corresponding input uses MatLoad()
592 .    PetscViewerDrawOpen() - Outputs nonzero matrix structure to
593          an X window display
594 -    PetscViewerSocketOpen() - Outputs matrix to Socket viewer.
595          Currently only the sequential dense and AIJ
596          matrix types support the Socket viewer.
597 
598    The user can call PetscViewerSetFormat() to specify the output
599    format of ASCII printed objects (when using PETSC_VIEWER_STDOUT_SELF,
600    PETSC_VIEWER_STDOUT_WORLD and PetscViewerASCIIOpen).  Available formats include
601 +    PETSC_VIEWER_DEFAULT - default, prints matrix contents
602 .    PETSC_VIEWER_ASCII_MATLAB - prints matrix contents in Matlab format
603 .    PETSC_VIEWER_ASCII_DENSE - prints entire matrix including zeros
604 .    PETSC_VIEWER_ASCII_COMMON - prints matrix contents, using a sparse
605          format common among all matrix types
606 .    PETSC_VIEWER_ASCII_IMPL - prints matrix contents, using an implementation-specific
607          format (which is in many cases the same as the default)
608 .    PETSC_VIEWER_ASCII_INFO - prints basic information about the matrix
609          size and structure (not the matrix entries)
610 .    PETSC_VIEWER_ASCII_INFO_DETAIL - prints more detailed information about
611          the matrix structure
612 
613    Options Database Keys:
614 +  -mat_view_info - Prints info on matrix at conclusion of MatEndAssembly()
615 .  -mat_view_info_detailed - Prints more detailed info
616 .  -mat_view - Prints matrix in ASCII format
617 .  -mat_view_matlab - Prints matrix in Matlab format
618 .  -mat_view_draw - PetscDraws nonzero structure of matrix, using MatView() and PetscDrawOpenX().
619 .  -display <name> - Sets display name (default is host)
620 .  -draw_pause <sec> - Sets number of seconds to pause after display
621 .  -mat_view_socket - Sends matrix to socket, can be accessed from Matlab (see users manual)
622 .  -viewer_socket_machine <machine>
623 .  -viewer_socket_port <port>
624 .  -mat_view_binary - save matrix to file in binary format
625 -  -viewer_binary_filename <name>
626    Level: beginner
627 
628    Notes: see the manual page for MatLoad() for the exact format of the binary file when the binary
629       viewer is used.
630 
631       See bin/matlab/PetscBinaryRead.m for a Matlab code that can read in the binary file when the binary
632       viewer is used.
633 
634    Concepts: matrices^viewing
635    Concepts: matrices^plotting
636    Concepts: matrices^printing
637 
638 .seealso: PetscViewerSetFormat(), PetscViewerASCIIOpen(), PetscViewerDrawOpen(),
639           PetscViewerSocketOpen(), PetscViewerBinaryOpen(), MatLoad()
640 @*/
641 PetscErrorCode PETSCMAT_DLLEXPORT MatView(Mat mat,PetscViewer viewer)
642 {
643   PetscErrorCode    ierr;
644   PetscInt          rows,cols;
645   PetscTruth        iascii;
646   const MatType     cstr;
647   PetscViewerFormat format;
648 
649   PetscFunctionBegin;
650   PetscValidHeaderSpecific(mat,MAT_COOKIE,1);
651   PetscValidType(mat,1);
652   if (!viewer) {
653     ierr = PetscViewerASCIIGetStdout(((PetscObject)mat)->comm,&viewer);CHKERRQ(ierr);
654   }
655   PetscValidHeaderSpecific(viewer,PETSC_VIEWER_COOKIE,2);
656   PetscCheckSameComm(mat,1,viewer,2);
657   if (!mat->assembled) SETERRQ(PETSC_ERR_ORDER,"Must call MatAssemblyBegin/End() before viewing matrix");
658   ierr = MatPreallocated(mat);CHKERRQ(ierr);
659 
660   ierr = PetscLogEventBegin(MAT_View,mat,viewer,0,0);CHKERRQ(ierr);
661   ierr = PetscTypeCompare((PetscObject)viewer,PETSC_VIEWER_ASCII,&iascii);CHKERRQ(ierr);
662   if (iascii) {
663     ierr = PetscViewerGetFormat(viewer,&format);CHKERRQ(ierr);
664     if (format == PETSC_VIEWER_ASCII_INFO || format == PETSC_VIEWER_ASCII_INFO_DETAIL) {
665       if (((PetscObject)mat)->prefix) {
666         ierr = PetscViewerASCIIPrintf(viewer,"Matrix Object:(%s)\n",((PetscObject)mat)->prefix);CHKERRQ(ierr);
667       } else {
668         ierr = PetscViewerASCIIPrintf(viewer,"Matrix Object:\n");CHKERRQ(ierr);
669       }
670       ierr = PetscViewerASCIIPushTab(viewer);CHKERRQ(ierr);
671       ierr = MatGetType(mat,&cstr);CHKERRQ(ierr);
672       ierr = MatGetSize(mat,&rows,&cols);CHKERRQ(ierr);
673       ierr = PetscViewerASCIIPrintf(viewer,"type=%s, rows=%D, cols=%D\n",cstr,rows,cols);CHKERRQ(ierr);
674       if (mat->factor) {
675         const MatSolverPackage solver;
676         ierr = MatFactorGetSolverPackage(mat,&solver);CHKERRQ(ierr);
677         ierr = PetscViewerASCIIPrintf(viewer,"package used to perform factorization: %s\n",solver);CHKERRQ(ierr);
678       }
679       if (mat->ops->getinfo) {
680         MatInfo info;
681         ierr = MatGetInfo(mat,MAT_GLOBAL_SUM,&info);CHKERRQ(ierr);
682         ierr = PetscViewerASCIIPrintf(viewer,"total: nonzeros=%D, allocated nonzeros=%D\n",(PetscInt)info.nz_used,(PetscInt)info.nz_allocated);CHKERRQ(ierr);
683       }
684     }
685   }
686   if (mat->ops->view) {
687     ierr = PetscViewerASCIIPushTab(viewer);CHKERRQ(ierr);
688     ierr = (*mat->ops->view)(mat,viewer);CHKERRQ(ierr);
689     ierr = PetscViewerASCIIPopTab(viewer);CHKERRQ(ierr);
690   } else if (!iascii) {
691     SETERRQ1(PETSC_ERR_SUP,"Viewer type %s not supported",((PetscObject)viewer)->type_name);
692   }
693   if (iascii) {
694     ierr = PetscViewerGetFormat(viewer,&format);CHKERRQ(ierr);
695     if (format == PETSC_VIEWER_ASCII_INFO || format == PETSC_VIEWER_ASCII_INFO_DETAIL) {
696       ierr = PetscViewerASCIIPopTab(viewer);CHKERRQ(ierr);
697     }
698   }
699   ierr = PetscLogEventEnd(MAT_View,mat,viewer,0,0);CHKERRQ(ierr);
700   PetscFunctionReturn(0);
701 }
702 
703 #undef __FUNCT__
704 #define __FUNCT__ "MatScaleSystem"
705 /*@
706    MatScaleSystem - Scale a vector solution and right hand side to
707    match the scaling of a scaled matrix.
708 
709    Collective on Mat
710 
711    Input Parameter:
712 +  mat - the matrix
713 .  b - right hand side vector (or PETSC_NULL)
714 -  x - solution vector (or PETSC_NULL)
715 
716 
717    Notes:
718    For AIJ, and BAIJ matrix formats, the matrices are not
719    internally scaled, so this does nothing. For MPIROWBS it
720    permutes and diagonally scales.
721 
722    The KSP methods automatically call this routine when required
723    (via PCPreSolve()) so it is rarely used directly.
724 
725    Level: Developer
726 
727    Concepts: matrices^scaling
728 
729 .seealso: MatUseScaledForm(), MatUnScaleSystem()
730 @*/
731 PetscErrorCode PETSCMAT_DLLEXPORT MatScaleSystem(Mat mat,Vec b,Vec x)
732 {
733   PetscErrorCode ierr;
734 
735   PetscFunctionBegin;
736   PetscValidHeaderSpecific(mat,MAT_COOKIE,1);
737   PetscValidType(mat,1);
738   ierr = MatPreallocated(mat);CHKERRQ(ierr);
739   if (x) {PetscValidHeaderSpecific(x,VEC_COOKIE,2);PetscCheckSameComm(mat,1,x,2);}
740   if (b) {PetscValidHeaderSpecific(b,VEC_COOKIE,3);PetscCheckSameComm(mat,1,b,3);}
741 
742   if (mat->ops->scalesystem) {
743     ierr = (*mat->ops->scalesystem)(mat,b,x);CHKERRQ(ierr);
744   }
745   PetscFunctionReturn(0);
746 }
747 
748 #undef __FUNCT__
749 #define __FUNCT__ "MatUnScaleSystem"
750 /*@
751    MatUnScaleSystem - Unscales a vector solution and right hand side to
752    match the original scaling of a scaled matrix.
753 
754    Collective on Mat
755 
756    Input Parameter:
757 +  mat - the matrix
758 .  b - right hand side vector (or PETSC_NULL)
759 -  x - solution vector (or PETSC_NULL)
760 
761 
762    Notes:
763    For AIJ and BAIJ matrix formats, the matrices are not
764    internally scaled, so this does nothing. For MPIROWBS it
765    permutes and diagonally scales.
766 
767    The KSP methods automatically call this routine when required
768    (via PCPreSolve()) so it is rarely used directly.
769 
770    Level: Developer
771 
772 .seealso: MatUseScaledForm(), MatScaleSystem()
773 @*/
774 PetscErrorCode PETSCMAT_DLLEXPORT MatUnScaleSystem(Mat mat,Vec b,Vec x)
775 {
776   PetscErrorCode ierr;
777 
778   PetscFunctionBegin;
779   PetscValidHeaderSpecific(mat,MAT_COOKIE,1);
780   PetscValidType(mat,1);
781   ierr = MatPreallocated(mat);CHKERRQ(ierr);
782   if (x) {PetscValidHeaderSpecific(x,VEC_COOKIE,2);PetscCheckSameComm(mat,1,x,2);}
783   if (b) {PetscValidHeaderSpecific(b,VEC_COOKIE,3);PetscCheckSameComm(mat,1,b,3);}
784   if (mat->ops->unscalesystem) {
785     ierr = (*mat->ops->unscalesystem)(mat,b,x);CHKERRQ(ierr);
786   }
787   PetscFunctionReturn(0);
788 }
789 
790 #undef __FUNCT__
791 #define __FUNCT__ "MatUseScaledForm"
792 /*@
793    MatUseScaledForm - For matrix storage formats that scale the
794    matrix (for example MPIRowBS matrices are diagonally scaled on
795    assembly) indicates matrix operations (MatMult() etc) are
796    applied using the scaled matrix.
797 
798    Collective on Mat
799 
800    Input Parameter:
801 +  mat - the matrix
802 -  scaled - PETSC_TRUE for applying the scaled, PETSC_FALSE for
803             applying the original matrix
804 
805    Notes:
806    For scaled matrix formats, applying the original, unscaled matrix
807    will be slightly more expensive
808 
809    Level: Developer
810 
811 .seealso: MatScaleSystem(), MatUnScaleSystem()
812 @*/
813 PetscErrorCode PETSCMAT_DLLEXPORT MatUseScaledForm(Mat mat,PetscTruth scaled)
814 {
815   PetscErrorCode ierr;
816 
817   PetscFunctionBegin;
818   PetscValidHeaderSpecific(mat,MAT_COOKIE,1);
819   PetscValidType(mat,1);
820   ierr = MatPreallocated(mat);CHKERRQ(ierr);
821   if (mat->ops->usescaledform) {
822     ierr = (*mat->ops->usescaledform)(mat,scaled);CHKERRQ(ierr);
823   }
824   PetscFunctionReturn(0);
825 }
826 
827 #undef __FUNCT__
828 #define __FUNCT__ "MatDestroy"
829 /*@
830    MatDestroy - Frees space taken by a matrix.
831 
832    Collective on Mat
833 
834    Input Parameter:
835 .  A - the matrix
836 
837    Level: beginner
838 
839 @*/
840 PetscErrorCode PETSCMAT_DLLEXPORT MatDestroy(Mat A)
841 {
842   PetscErrorCode ierr;
843   PetscFunctionBegin;
844   PetscValidHeaderSpecific(A,MAT_COOKIE,1);
845   if (--((PetscObject)A)->refct > 0) PetscFunctionReturn(0);
846   ierr = MatPreallocated(A);CHKERRQ(ierr);
847   /* if memory was published with AMS then destroy it */
848   ierr = PetscObjectDepublish(A);CHKERRQ(ierr);
849   if (A->ops->destroy) {
850     ierr = (*A->ops->destroy)(A);CHKERRQ(ierr);
851   }
852   if (A->mapping) {
853     ierr = ISLocalToGlobalMappingDestroy(A->mapping);CHKERRQ(ierr);
854   }
855   if (A->bmapping) {
856     ierr = ISLocalToGlobalMappingDestroy(A->bmapping);CHKERRQ(ierr);
857   }
858 
859   if (A->spptr){ierr = PetscFree(A->spptr);CHKERRQ(ierr);}
860   ierr = PetscMapDestroy(A->rmap);CHKERRQ(ierr);
861   ierr = PetscMapDestroy(A->cmap);CHKERRQ(ierr);
862   ierr = PetscHeaderDestroy(A);CHKERRQ(ierr);
863   PetscFunctionReturn(0);
864 }
865 
866 #undef __FUNCT__
867 #define __FUNCT__ "MatValid"
868 /*@
869    MatValid - Checks whether a matrix object is valid.
870 
871    Collective on Mat
872 
873    Input Parameter:
874 .  m - the matrix to check
875 
876    Output Parameter:
877    flg - flag indicating matrix status, either
878    PETSC_TRUE if matrix is valid, or PETSC_FALSE otherwise.
879 
880    Level: developer
881 
882    Concepts: matrices^validity
883 @*/
884 PetscErrorCode PETSCMAT_DLLEXPORT MatValid(Mat m,PetscTruth *flg)
885 {
886   PetscFunctionBegin;
887   PetscValidIntPointer(flg,1);
888   if (!m)                                          *flg = PETSC_FALSE;
889   else if (((PetscObject)m)->cookie != MAT_COOKIE) *flg = PETSC_FALSE;
890   else                                             *flg = PETSC_TRUE;
891   PetscFunctionReturn(0);
892 }
893 
894 #undef __FUNCT__
895 #define __FUNCT__ "MatSetValues"
896 /*@
897    MatSetValues - Inserts or adds a block of values into a matrix.
898    These values may be cached, so MatAssemblyBegin() and MatAssemblyEnd()
899    MUST be called after all calls to MatSetValues() have been completed.
900 
901    Not Collective
902 
903    Input Parameters:
904 +  mat - the matrix
905 .  v - a logically two-dimensional array of values
906 .  m, idxm - the number of rows and their global indices
907 .  n, idxn - the number of columns and their global indices
908 -  addv - either ADD_VALUES or INSERT_VALUES, where
909    ADD_VALUES adds values to any existing entries, and
910    INSERT_VALUES replaces existing entries with new values
911 
912    Notes:
913    By default the values, v, are row-oriented and unsorted.
914    See MatSetOption() for other options.
915 
916    Calls to MatSetValues() with the INSERT_VALUES and ADD_VALUES
917    options cannot be mixed without intervening calls to the assembly
918    routines.
919 
920    MatSetValues() uses 0-based row and column numbers in Fortran
921    as well as in C.
922 
923    Negative indices may be passed in idxm and idxn, these rows and columns are
924    simply ignored. This allows easily inserting element stiffness matrices
925    with homogeneous Dirchlet boundary conditions that you don't want represented
926    in the matrix.
927 
928    Efficiency Alert:
929    The routine MatSetValuesBlocked() may offer much better efficiency
930    for users of block sparse formats (MATSEQBAIJ and MATMPIBAIJ).
931 
932    Level: beginner
933 
934    Concepts: matrices^putting entries in
935 
936 .seealso: MatSetOption(), MatAssemblyBegin(), MatAssemblyEnd(), MatSetValuesBlocked(), MatSetValuesLocal(),
937           InsertMode, INSERT_VALUES, ADD_VALUES
938 @*/
939 PetscErrorCode PETSCMAT_DLLEXPORT MatSetValues(Mat mat,PetscInt m,const PetscInt idxm[],PetscInt n,const PetscInt idxn[],const PetscScalar v[],InsertMode addv)
940 {
941   PetscErrorCode ierr;
942 
943   PetscFunctionBegin;
944   PetscValidHeaderSpecific(mat,MAT_COOKIE,1);
945   PetscValidType(mat,1);
946   if (!m || !n) PetscFunctionReturn(0); /* no values to insert */
947   PetscValidIntPointer(idxm,3);
948   PetscValidIntPointer(idxn,5);
949   ierr = MatPreallocated(mat);CHKERRQ(ierr);
950   if (mat->insertmode == NOT_SET_VALUES) {
951     mat->insertmode = addv;
952   }
953 #if defined(PETSC_USE_DEBUG)
954   else if (mat->insertmode != addv) {
955     SETERRQ(PETSC_ERR_ARG_WRONGSTATE,"Cannot mix add values and insert values");
956   }
957   if (mat->factor) SETERRQ(PETSC_ERR_ARG_WRONGSTATE,"Not for factored matrix");
958 #endif
959 
960   if (mat->assembled) {
961     mat->was_assembled = PETSC_TRUE;
962     mat->assembled     = PETSC_FALSE;
963   }
964   ierr = PetscLogEventBegin(MAT_SetValues,mat,0,0,0);CHKERRQ(ierr);
965   if (!mat->ops->setvalues) SETERRQ1(PETSC_ERR_SUP,"Mat type %s",((PetscObject)mat)->type_name);
966   ierr = (*mat->ops->setvalues)(mat,m,idxm,n,idxn,v,addv);CHKERRQ(ierr);
967   ierr = PetscLogEventEnd(MAT_SetValues,mat,0,0,0);CHKERRQ(ierr);
968   PetscFunctionReturn(0);
969 }
970 
971 
972 #undef __FUNCT__
973 #define __FUNCT__ "MatSetValuesRowLocal"
974 /*@
975    MatSetValuesRowLocal - Inserts a row (block row for BAIJ matrices) of nonzero
976         values into a matrix
977 
978    Not Collective
979 
980    Input Parameters:
981 +  mat - the matrix
982 .  row - the (block) row to set
983 -  v - a logically two-dimensional array of values
984 
985    Notes:
986    By the values, v, are column-oriented (for the block version) and sorted
987 
988    All the nonzeros in the row must be provided
989 
990    The matrix must have previously had its column indices set
991 
992    The row must belong to this process
993 
994    Level: intermediate
995 
996    Concepts: matrices^putting entries in
997 
998 .seealso: MatSetOption(), MatAssemblyBegin(), MatAssemblyEnd(), MatSetValuesBlocked(), MatSetValuesLocal(),
999           InsertMode, INSERT_VALUES, ADD_VALUES, MatSetValues(), MatSetValuesRow(), MatSetLocalToGlobalMapping()
1000 @*/
1001 PetscErrorCode PETSCMAT_DLLEXPORT MatSetValuesRowLocal(Mat mat,PetscInt row,const PetscScalar v[])
1002 {
1003   PetscErrorCode ierr;
1004 
1005   PetscFunctionBegin;
1006   PetscValidHeaderSpecific(mat,MAT_COOKIE,1);
1007   PetscValidType(mat,1);
1008   PetscValidScalarPointer(v,2);
1009   ierr = MatSetValuesRow(mat, mat->mapping->indices[row],v);CHKERRQ(ierr);
1010   PetscFunctionReturn(0);
1011 }
1012 
1013 #undef __FUNCT__
1014 #define __FUNCT__ "MatSetValuesRow"
1015 /*@
1016    MatSetValuesRow - Inserts a row (block row for BAIJ matrices) of nonzero
1017         values into a matrix
1018 
1019    Not Collective
1020 
1021    Input Parameters:
1022 +  mat - the matrix
1023 .  row - the (block) row to set
1024 -  v - a logically two-dimensional array of values
1025 
1026    Notes:
1027    By the values, v, are column-oriented (for the block version) and sorted
1028 
1029    All the nonzeros in the row must be provided
1030 
1031    The matrix must have previously had its column indices set
1032 
1033    The row must belong to this process
1034 
1035    Level: intermediate
1036 
1037    Concepts: matrices^putting entries in
1038 
1039 .seealso: MatSetOption(), MatAssemblyBegin(), MatAssemblyEnd(), MatSetValuesBlocked(), MatSetValuesLocal(),
1040           InsertMode, INSERT_VALUES, ADD_VALUES, MatSetValues()
1041 @*/
1042 PetscErrorCode PETSCMAT_DLLEXPORT MatSetValuesRow(Mat mat,PetscInt row,const PetscScalar v[])
1043 {
1044   PetscErrorCode ierr;
1045 
1046   PetscFunctionBegin;
1047   PetscValidHeaderSpecific(mat,MAT_COOKIE,1);
1048   PetscValidType(mat,1);
1049   PetscValidScalarPointer(v,2);
1050 #if defined(PETSC_USE_DEBUG)
1051   if (mat->insertmode == ADD_VALUES) SETERRQ(PETSC_ERR_ARG_WRONGSTATE,"Cannot mix add and insert values");
1052   if (mat->factor) SETERRQ(PETSC_ERR_ARG_WRONGSTATE,"Not for factored matrix");
1053 #endif
1054   mat->insertmode = INSERT_VALUES;
1055 
1056   if (mat->assembled) {
1057     mat->was_assembled = PETSC_TRUE;
1058     mat->assembled     = PETSC_FALSE;
1059   }
1060   ierr = PetscLogEventBegin(MAT_SetValues,mat,0,0,0);CHKERRQ(ierr);
1061   if (!mat->ops->setvaluesrow) SETERRQ1(PETSC_ERR_SUP,"Mat type %s",((PetscObject)mat)->type_name);
1062   ierr = (*mat->ops->setvaluesrow)(mat,row,v);CHKERRQ(ierr);
1063   ierr = PetscLogEventEnd(MAT_SetValues,mat,0,0,0);CHKERRQ(ierr);
1064   PetscFunctionReturn(0);
1065 }
1066 
1067 #undef __FUNCT__
1068 #define __FUNCT__ "MatSetValuesStencil"
1069 /*@
1070    MatSetValuesStencil - Inserts or adds a block of values into a matrix.
1071      Using structured grid indexing
1072 
1073    Not Collective
1074 
1075    Input Parameters:
1076 +  mat - the matrix
1077 .  m - number of rows being entered
1078 .  idxm - grid coordinates (and component number when dof > 1) for matrix rows being entered
1079 .  n - number of columns being entered
1080 .  idxn - grid coordinates (and component number when dof > 1) for matrix columns being entered
1081 .  v - a logically two-dimensional array of values
1082 -  addv - either ADD_VALUES or INSERT_VALUES, where
1083    ADD_VALUES adds values to any existing entries, and
1084    INSERT_VALUES replaces existing entries with new values
1085 
1086    Notes:
1087    By default the values, v, are row-oriented.  See MatSetOption() for other options.
1088 
1089    Calls to MatSetValuesStencil() with the INSERT_VALUES and ADD_VALUES
1090    options cannot be mixed without intervening calls to the assembly
1091    routines.
1092 
1093    The grid coordinates are across the entire grid, not just the local portion
1094 
1095    MatSetValuesStencil() uses 0-based row and column numbers in Fortran
1096    as well as in C.
1097 
1098    For setting/accessing vector values via array coordinates you can use the DAVecGetArray() routine
1099 
1100    In order to use this routine you must either obtain the matrix with DAGetMatrix()
1101    or call MatSetLocalToGlobalMapping() and MatSetStencil() first.
1102 
1103    The columns and rows in the stencil passed in MUST be contained within the
1104    ghost region of the given process as set with DACreateXXX() or MatSetStencil(). For example,
1105    if you create a DA with an overlap of one grid level and on a particular process its first
1106    local nonghost x logical coordinate is 6 (so its first ghost x logical coordinate is 5) the
1107    first i index you can use in your column and row indices in MatSetStencil() is 5.
1108 
1109    In Fortran idxm and idxn should be declared as
1110 $     MatStencil idxm(4,m),idxn(4,n)
1111    and the values inserted using
1112 $    idxm(MatStencil_i,1) = i
1113 $    idxm(MatStencil_j,1) = j
1114 $    idxm(MatStencil_k,1) = k
1115 $    idxm(MatStencil_c,1) = c
1116    etc
1117 
1118    For periodic boundary conditions use negative indices for values to the left (below 0; that are to be
1119    obtained by wrapping values from right edge). For values to the right of the last entry using that index plus one
1120    etc to obtain values that obtained by wrapping the values from the left edge. This does not work for the DA_NONPERIODIC
1121    wrap.
1122 
1123    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
1124    a single value per point) you can skip filling those indices.
1125 
1126    Inspired by the structured grid interface to the HYPRE package
1127    (http://www.llnl.gov/CASC/hypre)
1128 
1129    Efficiency Alert:
1130    The routine MatSetValuesBlockedStencil() may offer much better efficiency
1131    for users of block sparse formats (MATSEQBAIJ and MATMPIBAIJ).
1132 
1133    Level: beginner
1134 
1135    Concepts: matrices^putting entries in
1136 
1137 .seealso: MatSetOption(), MatAssemblyBegin(), MatAssemblyEnd(), MatSetValuesBlocked(), MatSetValuesLocal()
1138           MatSetValues(), MatSetValuesBlockedStencil(), MatSetStencil(), DAGetMatrix(), DAVecGetArray(), MatStencil
1139 @*/
1140 PetscErrorCode PETSCMAT_DLLEXPORT MatSetValuesStencil(Mat mat,PetscInt m,const MatStencil idxm[],PetscInt n,const MatStencil idxn[],const PetscScalar v[],InsertMode addv)
1141 {
1142   PetscErrorCode ierr;
1143   PetscInt       j,i,jdxm[128],jdxn[256],dim = mat->stencil.dim,*dims = mat->stencil.dims+1,tmp;
1144   PetscInt       *starts = mat->stencil.starts,*dxm = (PetscInt*)idxm,*dxn = (PetscInt*)idxn,sdim = dim - (1 - (PetscInt)mat->stencil.noc);
1145 
1146   PetscFunctionBegin;
1147   if (!m || !n) PetscFunctionReturn(0); /* no values to insert */
1148   PetscValidHeaderSpecific(mat,MAT_COOKIE,1);
1149   PetscValidType(mat,1);
1150   PetscValidIntPointer(idxm,3);
1151   PetscValidIntPointer(idxn,5);
1152   PetscValidScalarPointer(v,6);
1153 
1154   if (m > 128) SETERRQ1(PETSC_ERR_SUP,"Can only set 128 rows at a time; trying to set %D",m);
1155   if (n > 256) SETERRQ1(PETSC_ERR_SUP,"Can only set 256 columns at a time; trying to set %D",n);
1156 
1157   for (i=0; i<m; i++) {
1158     for (j=0; j<3-sdim; j++) dxm++;
1159     tmp = *dxm++ - starts[0];
1160     for (j=0; j<dim-1; j++) {
1161       if ((*dxm++ - starts[j+1]) < 0 || tmp < 0) tmp = PETSC_MIN_INT;
1162       else                                       tmp = tmp*dims[j] + *(dxm-1) - starts[j+1];
1163     }
1164     if (mat->stencil.noc) dxm++;
1165     jdxm[i] = tmp;
1166   }
1167   for (i=0; i<n; i++) {
1168     for (j=0; j<3-sdim; j++) dxn++;
1169     tmp = *dxn++ - starts[0];
1170     for (j=0; j<dim-1; j++) {
1171       if ((*dxn++ - starts[j+1]) < 0 || tmp < 0) tmp = PETSC_MIN_INT;
1172       else                                       tmp = tmp*dims[j] + *(dxn-1) - starts[j+1];
1173     }
1174     if (mat->stencil.noc) dxn++;
1175     jdxn[i] = tmp;
1176   }
1177   ierr = MatSetValuesLocal(mat,m,jdxm,n,jdxn,v,addv);CHKERRQ(ierr);
1178   PetscFunctionReturn(0);
1179 }
1180 
1181 #undef __FUNCT__
1182 #define __FUNCT__ "MatSetValuesBlockedStencil"
1183 /*@C
1184    MatSetValuesBlockedStencil - Inserts or adds a block of values into a matrix.
1185      Using structured grid indexing
1186 
1187    Not Collective
1188 
1189    Input Parameters:
1190 +  mat - the matrix
1191 .  m - number of rows being entered
1192 .  idxm - grid coordinates for matrix rows being entered
1193 .  n - number of columns being entered
1194 .  idxn - grid coordinates for matrix columns being entered
1195 .  v - a logically two-dimensional array of values
1196 -  addv - either ADD_VALUES or INSERT_VALUES, where
1197    ADD_VALUES adds values to any existing entries, and
1198    INSERT_VALUES replaces existing entries with new values
1199 
1200    Notes:
1201    By default the values, v, are row-oriented and unsorted.
1202    See MatSetOption() for other options.
1203 
1204    Calls to MatSetValuesBlockedStencil() with the INSERT_VALUES and ADD_VALUES
1205    options cannot be mixed without intervening calls to the assembly
1206    routines.
1207 
1208    The grid coordinates are across the entire grid, not just the local portion
1209 
1210    MatSetValuesBlockedStencil() uses 0-based row and column numbers in Fortran
1211    as well as in C.
1212 
1213    For setting/accessing vector values via array coordinates you can use the DAVecGetArray() routine
1214 
1215    In order to use this routine you must either obtain the matrix with DAGetMatrix()
1216    or call MatSetBlockSize(), MatSetLocalToGlobalMapping() and MatSetStencil() first.
1217 
1218    The columns and rows in the stencil passed in MUST be contained within the
1219    ghost region of the given process as set with DACreateXXX() or MatSetStencil(). For example,
1220    if you create a DA with an overlap of one grid level and on a particular process its first
1221    local nonghost x logical coordinate is 6 (so its first ghost x logical coordinate is 5) the
1222    first i index you can use in your column and row indices in MatSetStencil() is 5.
1223 
1224    In Fortran idxm and idxn should be declared as
1225 $     MatStencil idxm(4,m),idxn(4,n)
1226    and the values inserted using
1227 $    idxm(MatStencil_i,1) = i
1228 $    idxm(MatStencil_j,1) = j
1229 $    idxm(MatStencil_k,1) = k
1230    etc
1231 
1232    Negative indices may be passed in idxm and idxn, these rows and columns are
1233    simply ignored. This allows easily inserting element stiffness matrices
1234    with homogeneous Dirchlet boundary conditions that you don't want represented
1235    in the matrix.
1236 
1237    Inspired by the structured grid interface to the HYPRE package
1238    (http://www.llnl.gov/CASC/hypre)
1239 
1240    Level: beginner
1241 
1242    Concepts: matrices^putting entries in
1243 
1244 .seealso: MatSetOption(), MatAssemblyBegin(), MatAssemblyEnd(), MatSetValuesBlocked(), MatSetValuesLocal()
1245           MatSetValues(), MatSetValuesStencil(), MatSetStencil(), DAGetMatrix(), DAVecGetArray(), MatStencil,
1246           MatSetBlockSize(), MatSetLocalToGlobalMapping()
1247 @*/
1248 PetscErrorCode PETSCMAT_DLLEXPORT MatSetValuesBlockedStencil(Mat mat,PetscInt m,const MatStencil idxm[],PetscInt n,const MatStencil idxn[],const PetscScalar v[],InsertMode addv)
1249 {
1250   PetscErrorCode ierr;
1251   PetscInt       j,i,jdxm[128],jdxn[256],dim = mat->stencil.dim,*dims = mat->stencil.dims+1,tmp;
1252   PetscInt       *starts = mat->stencil.starts,*dxm = (PetscInt*)idxm,*dxn = (PetscInt*)idxn,sdim = dim - (1 - (PetscInt)mat->stencil.noc);
1253 
1254   PetscFunctionBegin;
1255   if (!m || !n) PetscFunctionReturn(0); /* no values to insert */
1256   PetscValidHeaderSpecific(mat,MAT_COOKIE,1);
1257   PetscValidType(mat,1);
1258   PetscValidIntPointer(idxm,3);
1259   PetscValidIntPointer(idxn,5);
1260   PetscValidScalarPointer(v,6);
1261 
1262   if (m > 128) SETERRQ1(PETSC_ERR_SUP,"Can only set 128 rows at a time; trying to set %D",m);
1263   if (n > 128) SETERRQ1(PETSC_ERR_SUP,"Can only set 256 columns at a time; trying to set %D",n);
1264 
1265   for (i=0; i<m; i++) {
1266     for (j=0; j<3-sdim; j++) dxm++;
1267     tmp = *dxm++ - starts[0];
1268     for (j=0; j<sdim-1; j++) {
1269       if ((*dxm++ - starts[j+1]) < 0 || tmp < 0) tmp = PETSC_MIN_INT;
1270       else                                      tmp = tmp*dims[j] + *(dxm-1) - starts[j+1];
1271     }
1272     dxm++;
1273     jdxm[i] = tmp;
1274   }
1275   for (i=0; i<n; i++) {
1276     for (j=0; j<3-sdim; j++) dxn++;
1277     tmp = *dxn++ - starts[0];
1278     for (j=0; j<sdim-1; j++) {
1279       if ((*dxn++ - starts[j+1]) < 0 || tmp < 0) tmp = PETSC_MIN_INT;
1280       else                                       tmp = tmp*dims[j] + *(dxn-1) - starts[j+1];
1281     }
1282     dxn++;
1283     jdxn[i] = tmp;
1284   }
1285   ierr = MatSetValuesBlockedLocal(mat,m,jdxm,n,jdxn,v,addv);CHKERRQ(ierr);
1286   PetscFunctionReturn(0);
1287 }
1288 
1289 #undef __FUNCT__
1290 #define __FUNCT__ "MatSetStencil"
1291 /*@
1292    MatSetStencil - Sets the grid information for setting values into a matrix via
1293         MatSetValuesStencil()
1294 
1295    Not Collective
1296 
1297    Input Parameters:
1298 +  mat - the matrix
1299 .  dim - dimension of the grid 1, 2, or 3
1300 .  dims - number of grid points in x, y, and z direction, including ghost points on your processor
1301 .  starts - starting point of ghost nodes on your processor in x, y, and z direction
1302 -  dof - number of degrees of freedom per node
1303 
1304 
1305    Inspired by the structured grid interface to the HYPRE package
1306    (www.llnl.gov/CASC/hyper)
1307 
1308    For matrices generated with DAGetMatrix() this routine is automatically called and so not needed by the
1309    user.
1310 
1311    Level: beginner
1312 
1313    Concepts: matrices^putting entries in
1314 
1315 .seealso: MatSetOption(), MatAssemblyBegin(), MatAssemblyEnd(), MatSetValuesBlocked(), MatSetValuesLocal()
1316           MatSetValues(), MatSetValuesBlockedStencil(), MatSetValuesStencil()
1317 @*/
1318 PetscErrorCode PETSCMAT_DLLEXPORT MatSetStencil(Mat mat,PetscInt dim,const PetscInt dims[],const PetscInt starts[],PetscInt dof)
1319 {
1320   PetscInt i;
1321 
1322   PetscFunctionBegin;
1323   PetscValidHeaderSpecific(mat,MAT_COOKIE,1);
1324   PetscValidIntPointer(dims,3);
1325   PetscValidIntPointer(starts,4);
1326 
1327   mat->stencil.dim = dim + (dof > 1);
1328   for (i=0; i<dim; i++) {
1329     mat->stencil.dims[i]   = dims[dim-i-1];      /* copy the values in backwards */
1330     mat->stencil.starts[i] = starts[dim-i-1];
1331   }
1332   mat->stencil.dims[dim]   = dof;
1333   mat->stencil.starts[dim] = 0;
1334   mat->stencil.noc         = (PetscTruth)(dof == 1);
1335   PetscFunctionReturn(0);
1336 }
1337 
1338 #undef __FUNCT__
1339 #define __FUNCT__ "MatSetValuesBlocked"
1340 /*@
1341    MatSetValuesBlocked - Inserts or adds a block of values into a matrix.
1342 
1343    Not Collective
1344 
1345    Input Parameters:
1346 +  mat - the matrix
1347 .  v - a logically two-dimensional array of values
1348 .  m, idxm - the number of block rows and their global block indices
1349 .  n, idxn - the number of block columns and their global block indices
1350 -  addv - either ADD_VALUES or INSERT_VALUES, where
1351    ADD_VALUES adds values to any existing entries, and
1352    INSERT_VALUES replaces existing entries with new values
1353 
1354    Notes:
1355    The m and n count the NUMBER of blocks in the row direction and column direction,
1356    NOT the total number of rows/columns; for example, if the block size is 2 and
1357    you are passing in values for rows 2,3,4,5  then m would be 2 (not 4).
1358    The values in idxm would be 1 2; that is the first index for each block divided by
1359    the block size.
1360 
1361    Note that you must call MatSetBlockSize() when constructing this matrix (and before
1362    preallocating it).
1363 
1364    By default the values, v, are row-oriented and unsorted. So the layout of
1365    v is the same as for MatSetValues(). See MatSetOption() for other options.
1366 
1367    Calls to MatSetValuesBlocked() with the INSERT_VALUES and ADD_VALUES
1368    options cannot be mixed without intervening calls to the assembly
1369    routines.
1370 
1371    MatSetValuesBlocked() uses 0-based row and column numbers in Fortran
1372    as well as in C.
1373 
1374    Negative indices may be passed in idxm and idxn, these rows and columns are
1375    simply ignored. This allows easily inserting element stiffness matrices
1376    with homogeneous Dirchlet boundary conditions that you don't want represented
1377    in the matrix.
1378 
1379    Each time an entry is set within a sparse matrix via MatSetValues(),
1380    internal searching must be done to determine where to place the the
1381    data in the matrix storage space.  By instead inserting blocks of
1382    entries via MatSetValuesBlocked(), the overhead of matrix assembly is
1383    reduced.
1384 
1385    Example:
1386 $   Suppose m=n=2 and block size(bs) = 2 The array is
1387 $
1388 $   1  2  | 3  4
1389 $   5  6  | 7  8
1390 $   - - - | - - -
1391 $   9  10 | 11 12
1392 $   13 14 | 15 16
1393 $
1394 $   v[] should be passed in like
1395 $   v[] = [1,2,3,4,5,6,7,8,9,10,11,12,13,14,15,16]
1396 $
1397 $  If you are not using row oriented storage of v (that is you called MatSetOption(mat,MAT_ROW_ORIENTED,PETSC_FALSE)) then
1398 $   v[] = [1,5,9,13,2,6,10,14,3,7,11,15,4,8,12,16]
1399 
1400    Level: intermediate
1401 
1402    Concepts: matrices^putting entries in blocked
1403 
1404 .seealso: MatSetBlockSize(), MatSetOption(), MatAssemblyBegin(), MatAssemblyEnd(), MatSetValues(), MatSetValuesBlockedLocal()
1405 @*/
1406 PetscErrorCode PETSCMAT_DLLEXPORT MatSetValuesBlocked(Mat mat,PetscInt m,const PetscInt idxm[],PetscInt n,const PetscInt idxn[],const PetscScalar v[],InsertMode addv)
1407 {
1408   PetscErrorCode ierr;
1409 
1410   PetscFunctionBegin;
1411   PetscValidHeaderSpecific(mat,MAT_COOKIE,1);
1412   PetscValidType(mat,1);
1413   if (!m || !n) PetscFunctionReturn(0); /* no values to insert */
1414   PetscValidIntPointer(idxm,3);
1415   PetscValidIntPointer(idxn,5);
1416   PetscValidScalarPointer(v,6);
1417   ierr = MatPreallocated(mat);CHKERRQ(ierr);
1418   if (mat->insertmode == NOT_SET_VALUES) {
1419     mat->insertmode = addv;
1420   }
1421 #if defined(PETSC_USE_DEBUG)
1422   else if (mat->insertmode != addv) {
1423     SETERRQ(PETSC_ERR_ARG_WRONGSTATE,"Cannot mix add values and insert values");
1424   }
1425   if (mat->factor) SETERRQ(PETSC_ERR_ARG_WRONGSTATE,"Not for factored matrix");
1426 #endif
1427 
1428   if (mat->assembled) {
1429     mat->was_assembled = PETSC_TRUE;
1430     mat->assembled     = PETSC_FALSE;
1431   }
1432   ierr = PetscLogEventBegin(MAT_SetValues,mat,0,0,0);CHKERRQ(ierr);
1433   if (mat->ops->setvaluesblocked) {
1434     ierr = (*mat->ops->setvaluesblocked)(mat,m,idxm,n,idxn,v,addv);CHKERRQ(ierr);
1435   } else {
1436     PetscInt buf[4096],*ibufm=0,*ibufn=0;
1437     PetscInt i,j,*iidxm,*iidxn,bs=mat->rmap->bs;
1438     if ((m+n)*bs <= 4096) {
1439       iidxm = buf; iidxn = buf + m*bs;
1440     } else {
1441       ierr = PetscMalloc2(m*bs,PetscInt,&ibufm,n*bs,PetscInt,&ibufn);CHKERRQ(ierr);
1442       iidxm = ibufm; iidxn = ibufn;
1443     }
1444     for (i=0; i<m; i++) {
1445       for (j=0; j<bs; j++) {
1446 	iidxm[i*bs+j] = bs*idxm[i] + j;
1447       }
1448     }
1449     for (i=0; i<n; i++) {
1450       for (j=0; j<bs; j++) {
1451 	iidxn[i*bs+j] = bs*idxn[i] + j;
1452       }
1453     }
1454     ierr = MatSetValues(mat,bs*m,iidxm,bs*n,iidxn,v,addv);CHKERRQ(ierr);
1455     ierr = PetscFree2(ibufm,ibufn);CHKERRQ(ierr);
1456   }
1457   ierr = PetscLogEventEnd(MAT_SetValues,mat,0,0,0);CHKERRQ(ierr);
1458   PetscFunctionReturn(0);
1459 }
1460 
1461 #undef __FUNCT__
1462 #define __FUNCT__ "MatGetValues"
1463 /*@
1464    MatGetValues - Gets a block of values from a matrix.
1465 
1466    Not Collective; currently only returns a local block
1467 
1468    Input Parameters:
1469 +  mat - the matrix
1470 .  v - a logically two-dimensional array for storing the values
1471 .  m, idxm - the number of rows and their global indices
1472 -  n, idxn - the number of columns and their global indices
1473 
1474    Notes:
1475    The user must allocate space (m*n PetscScalars) for the values, v.
1476    The values, v, are then returned in a row-oriented format,
1477    analogous to that used by default in MatSetValues().
1478 
1479    MatGetValues() uses 0-based row and column numbers in
1480    Fortran as well as in C.
1481 
1482    MatGetValues() requires that the matrix has been assembled
1483    with MatAssemblyBegin()/MatAssemblyEnd().  Thus, calls to
1484    MatSetValues() and MatGetValues() CANNOT be made in succession
1485    without intermediate matrix assembly.
1486 
1487    Negative row or column indices will be ignored and those locations in v[] will be
1488    left unchanged.
1489 
1490    Level: advanced
1491 
1492    Concepts: matrices^accessing values
1493 
1494 .seealso: MatGetRow(), MatGetSubMatrices(), MatSetValues()
1495 @*/
1496 PetscErrorCode PETSCMAT_DLLEXPORT MatGetValues(Mat mat,PetscInt m,const PetscInt idxm[],PetscInt n,const PetscInt idxn[],PetscScalar v[])
1497 {
1498   PetscErrorCode ierr;
1499 
1500   PetscFunctionBegin;
1501   PetscValidHeaderSpecific(mat,MAT_COOKIE,1);
1502   PetscValidType(mat,1);
1503   PetscValidIntPointer(idxm,3);
1504   PetscValidIntPointer(idxn,5);
1505   PetscValidScalarPointer(v,6);
1506   if (!mat->assembled) SETERRQ(PETSC_ERR_ARG_WRONGSTATE,"Not for unassembled matrix");
1507   if (mat->factor) SETERRQ(PETSC_ERR_ARG_WRONGSTATE,"Not for factored matrix");
1508   if (!mat->ops->getvalues) SETERRQ1(PETSC_ERR_SUP,"Mat type %s",((PetscObject)mat)->type_name);
1509   ierr = MatPreallocated(mat);CHKERRQ(ierr);
1510 
1511   ierr = PetscLogEventBegin(MAT_GetValues,mat,0,0,0);CHKERRQ(ierr);
1512   ierr = (*mat->ops->getvalues)(mat,m,idxm,n,idxn,v);CHKERRQ(ierr);
1513   ierr = PetscLogEventEnd(MAT_GetValues,mat,0,0,0);CHKERRQ(ierr);
1514   PetscFunctionReturn(0);
1515 }
1516 
1517 #undef __FUNCT__
1518 #define __FUNCT__ "MatSetLocalToGlobalMapping"
1519 /*@
1520    MatSetLocalToGlobalMapping - Sets a local-to-global numbering for use by
1521    the routine MatSetValuesLocal() to allow users to insert matrix entries
1522    using a local (per-processor) numbering.
1523 
1524    Not Collective
1525 
1526    Input Parameters:
1527 +  x - the matrix
1528 -  mapping - mapping created with ISLocalToGlobalMappingCreate()
1529              or ISLocalToGlobalMappingCreateIS()
1530 
1531    Level: intermediate
1532 
1533    Concepts: matrices^local to global mapping
1534    Concepts: local to global mapping^for matrices
1535 
1536 .seealso:  MatAssemblyBegin(), MatAssemblyEnd(), MatSetValues(), MatSetValuesLocal()
1537 @*/
1538 PetscErrorCode PETSCMAT_DLLEXPORT MatSetLocalToGlobalMapping(Mat x,ISLocalToGlobalMapping mapping)
1539 {
1540   PetscErrorCode ierr;
1541   PetscFunctionBegin;
1542   PetscValidHeaderSpecific(x,MAT_COOKIE,1);
1543   PetscValidType(x,1);
1544   PetscValidHeaderSpecific(mapping,IS_LTOGM_COOKIE,2);
1545   if (x->mapping) {
1546     SETERRQ(PETSC_ERR_ARG_WRONGSTATE,"Mapping already set for matrix");
1547   }
1548   ierr = MatPreallocated(x);CHKERRQ(ierr);
1549 
1550   if (x->ops->setlocaltoglobalmapping) {
1551     ierr = (*x->ops->setlocaltoglobalmapping)(x,mapping);CHKERRQ(ierr);
1552   } else {
1553     ierr = PetscObjectReference((PetscObject)mapping);CHKERRQ(ierr);
1554     if (x->mapping) { ierr = ISLocalToGlobalMappingDestroy(x->mapping);CHKERRQ(ierr); }
1555     x->mapping = mapping;
1556   }
1557   PetscFunctionReturn(0);
1558 }
1559 
1560 #undef __FUNCT__
1561 #define __FUNCT__ "MatSetLocalToGlobalMappingBlock"
1562 /*@
1563    MatSetLocalToGlobalMappingBlock - Sets a local-to-global numbering for use
1564    by the routine MatSetValuesBlockedLocal() to allow users to insert matrix
1565    entries using a local (per-processor) numbering.
1566 
1567    Not Collective
1568 
1569    Input Parameters:
1570 +  x - the matrix
1571 -  mapping - mapping created with ISLocalToGlobalMappingCreate() or
1572              ISLocalToGlobalMappingCreateIS()
1573 
1574    Level: intermediate
1575 
1576    Concepts: matrices^local to global mapping blocked
1577    Concepts: local to global mapping^for matrices, blocked
1578 
1579 .seealso:  MatAssemblyBegin(), MatAssemblyEnd(), MatSetValues(), MatSetValuesBlockedLocal(),
1580            MatSetValuesBlocked(), MatSetValuesLocal()
1581 @*/
1582 PetscErrorCode PETSCMAT_DLLEXPORT MatSetLocalToGlobalMappingBlock(Mat x,ISLocalToGlobalMapping mapping)
1583 {
1584   PetscErrorCode ierr;
1585   PetscFunctionBegin;
1586   PetscValidHeaderSpecific(x,MAT_COOKIE,1);
1587   PetscValidType(x,1);
1588   PetscValidHeaderSpecific(mapping,IS_LTOGM_COOKIE,2);
1589   if (x->bmapping) {
1590     SETERRQ(PETSC_ERR_ARG_WRONGSTATE,"Mapping already set for matrix");
1591   }
1592   ierr = PetscObjectReference((PetscObject)mapping);CHKERRQ(ierr);
1593   if (x->bmapping) { ierr = ISLocalToGlobalMappingDestroy(x->bmapping);CHKERRQ(ierr); }
1594   x->bmapping = mapping;
1595   PetscFunctionReturn(0);
1596 }
1597 
1598 #undef __FUNCT__
1599 #define __FUNCT__ "MatSetValuesLocal"
1600 /*@
1601    MatSetValuesLocal - Inserts or adds values into certain locations of a matrix,
1602    using a local ordering of the nodes.
1603 
1604    Not Collective
1605 
1606    Input Parameters:
1607 +  x - the matrix
1608 .  nrow, irow - number of rows and their local indices
1609 .  ncol, icol - number of columns and their local indices
1610 .  y -  a logically two-dimensional array of values
1611 -  addv - either INSERT_VALUES or ADD_VALUES, where
1612    ADD_VALUES adds values to any existing entries, and
1613    INSERT_VALUES replaces existing entries with new values
1614 
1615    Notes:
1616    Before calling MatSetValuesLocal(), the user must first set the
1617    local-to-global mapping by calling MatSetLocalToGlobalMapping().
1618 
1619    Calls to MatSetValuesLocal() with the INSERT_VALUES and ADD_VALUES
1620    options cannot be mixed without intervening calls to the assembly
1621    routines.
1622 
1623    These values may be cached, so MatAssemblyBegin() and MatAssemblyEnd()
1624    MUST be called after all calls to MatSetValuesLocal() have been completed.
1625 
1626    Level: intermediate
1627 
1628    Concepts: matrices^putting entries in with local numbering
1629 
1630 .seealso:  MatAssemblyBegin(), MatAssemblyEnd(), MatSetValues(), MatSetLocalToGlobalMapping(),
1631            MatSetValueLocal()
1632 @*/
1633 PetscErrorCode PETSCMAT_DLLEXPORT MatSetValuesLocal(Mat mat,PetscInt nrow,const PetscInt irow[],PetscInt ncol,const PetscInt icol[],const PetscScalar y[],InsertMode addv)
1634 {
1635   PetscErrorCode ierr;
1636   PetscInt       irowm[2048],icolm[2048];
1637 
1638   PetscFunctionBegin;
1639   PetscValidHeaderSpecific(mat,MAT_COOKIE,1);
1640   PetscValidType(mat,1);
1641   if (!nrow || !ncol) PetscFunctionReturn(0); /* no values to insert */
1642   PetscValidIntPointer(irow,3);
1643   PetscValidIntPointer(icol,5);
1644   PetscValidScalarPointer(y,6);
1645   ierr = MatPreallocated(mat);CHKERRQ(ierr);
1646   if (mat->insertmode == NOT_SET_VALUES) {
1647     mat->insertmode = addv;
1648   }
1649 #if defined(PETSC_USE_DEBUG)
1650   else if (mat->insertmode != addv) {
1651     SETERRQ(PETSC_ERR_ARG_WRONGSTATE,"Cannot mix add values and insert values");
1652   }
1653   if (!mat->ops->setvalueslocal && (nrow > 2048 || ncol > 2048)) {
1654     SETERRQ2(PETSC_ERR_SUP,"Number column/row indices must be <= 2048: are %D %D",nrow,ncol);
1655   }
1656   if (mat->factor) SETERRQ(PETSC_ERR_ARG_WRONGSTATE,"Not for factored matrix");
1657 #endif
1658 
1659   if (mat->assembled) {
1660     mat->was_assembled = PETSC_TRUE;
1661     mat->assembled     = PETSC_FALSE;
1662   }
1663   ierr = PetscLogEventBegin(MAT_SetValues,mat,0,0,0);CHKERRQ(ierr);
1664   if (!mat->ops->setvalueslocal) {
1665     ierr = ISLocalToGlobalMappingApply(mat->mapping,nrow,irow,irowm);CHKERRQ(ierr);
1666     ierr = ISLocalToGlobalMappingApply(mat->mapping,ncol,icol,icolm);CHKERRQ(ierr);
1667     ierr = (*mat->ops->setvalues)(mat,nrow,irowm,ncol,icolm,y,addv);CHKERRQ(ierr);
1668   } else {
1669     ierr = (*mat->ops->setvalueslocal)(mat,nrow,irow,ncol,icol,y,addv);CHKERRQ(ierr);
1670   }
1671   mat->same_nonzero = PETSC_FALSE;
1672   ierr = PetscLogEventEnd(MAT_SetValues,mat,0,0,0);CHKERRQ(ierr);
1673   PetscFunctionReturn(0);
1674 }
1675 
1676 #undef __FUNCT__
1677 #define __FUNCT__ "MatSetValuesBlockedLocal"
1678 /*@
1679    MatSetValuesBlockedLocal - Inserts or adds values into certain locations of a matrix,
1680    using a local ordering of the nodes a block at a time.
1681 
1682    Not Collective
1683 
1684    Input Parameters:
1685 +  x - the matrix
1686 .  nrow, irow - number of rows and their local indices
1687 .  ncol, icol - number of columns and their local indices
1688 .  y -  a logically two-dimensional array of values
1689 -  addv - either INSERT_VALUES or ADD_VALUES, where
1690    ADD_VALUES adds values to any existing entries, and
1691    INSERT_VALUES replaces existing entries with new values
1692 
1693    Notes:
1694    Before calling MatSetValuesBlockedLocal(), the user must first set the
1695    block size using MatSetBlockSize(), and the local-to-global mapping by
1696    calling MatSetLocalToGlobalMappingBlock(), where the mapping MUST be
1697    set for matrix blocks, not for matrix elements.
1698 
1699    Calls to MatSetValuesBlockedLocal() with the INSERT_VALUES and ADD_VALUES
1700    options cannot be mixed without intervening calls to the assembly
1701    routines.
1702 
1703    These values may be cached, so MatAssemblyBegin() and MatAssemblyEnd()
1704    MUST be called after all calls to MatSetValuesBlockedLocal() have been completed.
1705 
1706    Level: intermediate
1707 
1708    Concepts: matrices^putting blocked values in with local numbering
1709 
1710 .seealso:  MatSetBlockSize(), MatSetLocalToGlobalMappingBlock(), MatAssemblyBegin(), MatAssemblyEnd(),
1711            MatSetValuesLocal(), MatSetLocalToGlobalMappingBlock(), MatSetValuesBlocked()
1712 @*/
1713 PetscErrorCode PETSCMAT_DLLEXPORT MatSetValuesBlockedLocal(Mat mat,PetscInt nrow,const PetscInt irow[],PetscInt ncol,const PetscInt icol[],const PetscScalar y[],InsertMode addv)
1714 {
1715   PetscErrorCode ierr;
1716   PetscInt       irowm[2048],icolm[2048];
1717 
1718   PetscFunctionBegin;
1719   PetscValidHeaderSpecific(mat,MAT_COOKIE,1);
1720   PetscValidType(mat,1);
1721   if (!nrow || !ncol) PetscFunctionReturn(0); /* no values to insert */
1722   PetscValidIntPointer(irow,3);
1723   PetscValidIntPointer(icol,5);
1724   PetscValidScalarPointer(y,6);
1725   ierr = MatPreallocated(mat);CHKERRQ(ierr);
1726   if (mat->insertmode == NOT_SET_VALUES) {
1727     mat->insertmode = addv;
1728   }
1729 #if defined(PETSC_USE_DEBUG)
1730   else if (mat->insertmode != addv) {
1731     SETERRQ(PETSC_ERR_ARG_WRONGSTATE,"Cannot mix add values and insert values");
1732   }
1733   if (!mat->bmapping) {
1734     SETERRQ(PETSC_ERR_ARG_WRONGSTATE,"Local to global never set with MatSetLocalToGlobalMappingBlock()");
1735   }
1736   if (nrow > 2048 || ncol > 2048) {
1737     SETERRQ2(PETSC_ERR_SUP,"Number column/row indices must be <= 2048: are %D %D",nrow,ncol);
1738   }
1739   if (mat->factor) SETERRQ(PETSC_ERR_ARG_WRONGSTATE,"Not for factored matrix");
1740 #endif
1741 
1742   if (mat->assembled) {
1743     mat->was_assembled = PETSC_TRUE;
1744     mat->assembled     = PETSC_FALSE;
1745   }
1746   ierr = PetscLogEventBegin(MAT_SetValues,mat,0,0,0);CHKERRQ(ierr);
1747   ierr = ISLocalToGlobalMappingApply(mat->bmapping,nrow,irow,irowm);CHKERRQ(ierr);
1748   ierr = ISLocalToGlobalMappingApply(mat->bmapping,ncol,icol,icolm);CHKERRQ(ierr);
1749   if (mat->ops->setvaluesblocked) {
1750   ierr = (*mat->ops->setvaluesblocked)(mat,nrow,irowm,ncol,icolm,y,addv);CHKERRQ(ierr);
1751   } else {
1752     PetscInt buf[4096],*ibufm=0,*ibufn=0;
1753     PetscInt i,j,*iirowm,*iicolm,bs=mat->rmap->bs;
1754     if ((nrow+ncol)*bs <= 4096) {
1755       iirowm = buf; iicolm = buf + nrow*bs;
1756     } else {
1757       ierr = PetscMalloc2(nrow*bs,PetscInt,&ibufm,ncol*bs,PetscInt,&ibufn);CHKERRQ(ierr);
1758       iirowm = ibufm; iicolm = ibufn;
1759     }
1760     for (i=0; i<nrow; i++) {
1761       for (j=0; j<bs; j++) {
1762 	iirowm[i*bs+j] = bs*irowm[i] + j;
1763       }
1764     }
1765     for (i=0; i<ncol; i++) {
1766       for (j=0; j<bs; j++) {
1767 	iicolm[i*bs+j] = bs*icolm[i] + j;
1768       }
1769     }
1770     ierr = MatSetValues(mat,bs*nrow,iirowm,bs*ncol,iicolm,y,addv);CHKERRQ(ierr);
1771     ierr = PetscFree2(ibufm,ibufn);CHKERRQ(ierr);
1772   }
1773   ierr = PetscLogEventEnd(MAT_SetValues,mat,0,0,0);CHKERRQ(ierr);
1774   PetscFunctionReturn(0);
1775 }
1776 
1777 #undef __FUNCT__
1778 #define __FUNCT__ "MatMultDiagonalBlock"
1779 /*@
1780    MatMultDiagonalBlock - Computes the matrix-vector product, y = Dx. Where D is defined by the inode or block structure of the diagonal
1781 
1782    Collective on Mat and Vec
1783 
1784    Input Parameters:
1785 +  mat - the matrix
1786 -  x   - the vector to be multiplied
1787 
1788    Output Parameters:
1789 .  y - the result
1790 
1791    Notes:
1792    The vectors x and y cannot be the same.  I.e., one cannot
1793    call MatMult(A,y,y).
1794 
1795    Level: developer
1796 
1797    Concepts: matrix-vector product
1798 
1799 .seealso: MatMultTranspose(), MatMultAdd(), MatMultTransposeAdd()
1800 @*/
1801 PetscErrorCode PETSCMAT_DLLEXPORT MatMultDiagonalBlock(Mat mat,Vec x,Vec y)
1802 {
1803   PetscErrorCode ierr;
1804 
1805   PetscFunctionBegin;
1806   PetscValidHeaderSpecific(mat,MAT_COOKIE,1);
1807   PetscValidType(mat,1);
1808   PetscValidHeaderSpecific(x,VEC_COOKIE,2);
1809   PetscValidHeaderSpecific(y,VEC_COOKIE,3);
1810 
1811   if (!mat->assembled) SETERRQ(PETSC_ERR_ARG_WRONGSTATE,"Not for unassembled matrix");
1812   if (mat->factor) SETERRQ(PETSC_ERR_ARG_WRONGSTATE,"Not for factored matrix");
1813   if (x == y) SETERRQ(PETSC_ERR_ARG_WRONGSTATE,"x and y must be different vectors");
1814   ierr = MatPreallocated(mat);CHKERRQ(ierr);
1815 
1816   if (!mat->ops->multdiagonalblock) SETERRQ(PETSC_ERR_SUP,"This matrix type does not have a multiply defined");
1817   ierr = (*mat->ops->multdiagonalblock)(mat,x,y);CHKERRQ(ierr);
1818   ierr = PetscObjectStateIncrease((PetscObject)y);CHKERRQ(ierr);
1819   PetscFunctionReturn(0);
1820 }
1821 
1822 /* --------------------------------------------------------*/
1823 #undef __FUNCT__
1824 #define __FUNCT__ "MatMult"
1825 /*@
1826    MatMult - Computes the matrix-vector product, y = Ax.
1827 
1828    Collective on Mat and Vec
1829 
1830    Input Parameters:
1831 +  mat - the matrix
1832 -  x   - the vector to be multiplied
1833 
1834    Output Parameters:
1835 .  y - the result
1836 
1837    Notes:
1838    The vectors x and y cannot be the same.  I.e., one cannot
1839    call MatMult(A,y,y).
1840 
1841    Level: beginner
1842 
1843    Concepts: matrix-vector product
1844 
1845 .seealso: MatMultTranspose(), MatMultAdd(), MatMultTransposeAdd()
1846 @*/
1847 PetscErrorCode PETSCMAT_DLLEXPORT MatMult(Mat mat,Vec x,Vec y)
1848 {
1849   PetscErrorCode ierr;
1850 
1851   PetscFunctionBegin;
1852   PetscValidHeaderSpecific(mat,MAT_COOKIE,1);
1853   PetscValidType(mat,1);
1854   PetscValidHeaderSpecific(x,VEC_COOKIE,2);
1855   PetscValidHeaderSpecific(y,VEC_COOKIE,3);
1856 
1857   if (!mat->assembled) SETERRQ(PETSC_ERR_ARG_WRONGSTATE,"Not for unassembled matrix");
1858   if (mat->factor) SETERRQ(PETSC_ERR_ARG_WRONGSTATE,"Not for factored matrix");
1859   if (x == y) SETERRQ(PETSC_ERR_ARG_WRONGSTATE,"x and y must be different vectors");
1860 #ifndef PETSC_HAVE_CONSTRAINTS
1861   if (mat->cmap->N != x->map->N) SETERRQ2(PETSC_ERR_ARG_SIZ,"Mat mat,Vec x: global dim %D %D",mat->cmap->N,x->map->N);
1862   if (mat->rmap->N != y->map->N) SETERRQ2(PETSC_ERR_ARG_SIZ,"Mat mat,Vec y: global dim %D %D",mat->rmap->N,y->map->N);
1863   if (mat->rmap->n != y->map->n) SETERRQ2(PETSC_ERR_ARG_SIZ,"Mat mat,Vec y: local dim %D %D",mat->rmap->n,y->map->n);
1864 #endif
1865   ierr = MatPreallocated(mat);CHKERRQ(ierr);
1866 
1867   if (mat->nullsp) {
1868     ierr = MatNullSpaceRemove(mat->nullsp,x,&x);CHKERRQ(ierr);
1869   }
1870 
1871   if (!mat->ops->mult) SETERRQ(PETSC_ERR_SUP,"This matrix type does not have a multiply defined");
1872   ierr = PetscLogEventBegin(MAT_Mult,mat,x,y,0);CHKERRQ(ierr);
1873   ierr = (*mat->ops->mult)(mat,x,y);CHKERRQ(ierr);
1874   ierr = PetscLogEventEnd(MAT_Mult,mat,x,y,0);CHKERRQ(ierr);
1875 
1876   if (mat->nullsp) {
1877     ierr = MatNullSpaceRemove(mat->nullsp,y,PETSC_NULL);CHKERRQ(ierr);
1878   }
1879   ierr = PetscObjectStateIncrease((PetscObject)y);CHKERRQ(ierr);
1880   PetscFunctionReturn(0);
1881 }
1882 
1883 #undef __FUNCT__
1884 #define __FUNCT__ "MatMultTranspose"
1885 /*@
1886    MatMultTranspose - Computes matrix transpose times a vector.
1887 
1888    Collective on Mat and Vec
1889 
1890    Input Parameters:
1891 +  mat - the matrix
1892 -  x   - the vector to be multilplied
1893 
1894    Output Parameters:
1895 .  y - the result
1896 
1897    Notes:
1898    The vectors x and y cannot be the same.  I.e., one cannot
1899    call MatMultTranspose(A,y,y).
1900 
1901    Level: beginner
1902 
1903    Concepts: matrix vector product^transpose
1904 
1905 .seealso: MatMult(), MatMultAdd(), MatMultTransposeAdd()
1906 @*/
1907 PetscErrorCode PETSCMAT_DLLEXPORT MatMultTranspose(Mat mat,Vec x,Vec y)
1908 {
1909   PetscErrorCode ierr;
1910 
1911   PetscFunctionBegin;
1912   PetscValidHeaderSpecific(mat,MAT_COOKIE,1);
1913   PetscValidType(mat,1);
1914   PetscValidHeaderSpecific(x,VEC_COOKIE,2);
1915   PetscValidHeaderSpecific(y,VEC_COOKIE,3);
1916 
1917   if (!mat->assembled) SETERRQ(PETSC_ERR_ARG_WRONGSTATE,"Not for unassembled matrix");
1918   if (mat->factor) SETERRQ(PETSC_ERR_ARG_WRONGSTATE,"Not for factored matrix");
1919   if (x == y) SETERRQ(PETSC_ERR_ARG_WRONGSTATE,"x and y must be different vectors");
1920 #ifndef PETSC_HAVE_CONSTRAINTS
1921   if (mat->rmap->N != x->map->N) SETERRQ2(PETSC_ERR_ARG_SIZ,"Mat mat,Vec x: global dim %D %D",mat->rmap->N,x->map->N);
1922   if (mat->cmap->N != y->map->N) SETERRQ2(PETSC_ERR_ARG_SIZ,"Mat mat,Vec y: global dim %D %D",mat->cmap->N,y->map->N);
1923 #endif
1924   ierr = MatPreallocated(mat);CHKERRQ(ierr);
1925 
1926   if (!mat->ops->multtranspose) SETERRQ(PETSC_ERR_SUP,"This matrix type does not have a multiply tranpose defined");
1927   ierr = PetscLogEventBegin(MAT_MultTranspose,mat,x,y,0);CHKERRQ(ierr);
1928   ierr = (*mat->ops->multtranspose)(mat,x,y);CHKERRQ(ierr);
1929   ierr = PetscLogEventEnd(MAT_MultTranspose,mat,x,y,0);CHKERRQ(ierr);
1930   ierr = PetscObjectStateIncrease((PetscObject)y);CHKERRQ(ierr);
1931   PetscFunctionReturn(0);
1932 }
1933 
1934 #undef __FUNCT__
1935 #define __FUNCT__ "MatMultAdd"
1936 /*@
1937     MatMultAdd -  Computes v3 = v2 + A * v1.
1938 
1939     Collective on Mat and Vec
1940 
1941     Input Parameters:
1942 +   mat - the matrix
1943 -   v1, v2 - the vectors
1944 
1945     Output Parameters:
1946 .   v3 - the result
1947 
1948     Notes:
1949     The vectors v1 and v3 cannot be the same.  I.e., one cannot
1950     call MatMultAdd(A,v1,v2,v1).
1951 
1952     Level: beginner
1953 
1954     Concepts: matrix vector product^addition
1955 
1956 .seealso: MatMultTranspose(), MatMult(), MatMultTransposeAdd()
1957 @*/
1958 PetscErrorCode PETSCMAT_DLLEXPORT MatMultAdd(Mat mat,Vec v1,Vec v2,Vec v3)
1959 {
1960   PetscErrorCode ierr;
1961 
1962   PetscFunctionBegin;
1963   PetscValidHeaderSpecific(mat,MAT_COOKIE,1);
1964   PetscValidType(mat,1);
1965   PetscValidHeaderSpecific(v1,VEC_COOKIE,2);
1966   PetscValidHeaderSpecific(v2,VEC_COOKIE,3);
1967   PetscValidHeaderSpecific(v3,VEC_COOKIE,4);
1968 
1969   if (!mat->assembled) SETERRQ(PETSC_ERR_ARG_WRONGSTATE,"Not for unassembled matrix");
1970   if (mat->factor) SETERRQ(PETSC_ERR_ARG_WRONGSTATE,"Not for factored matrix");
1971   if (mat->cmap->N != v1->map->N) SETERRQ2(PETSC_ERR_ARG_SIZ,"Mat mat,Vec v1: global dim %D %D",mat->cmap->N,v1->map->N);
1972   /* if (mat->rmap->N != v2->map->N) SETERRQ2(PETSC_ERR_ARG_SIZ,"Mat mat,Vec v2: global dim %D %D",mat->rmap->N,v2->map->N);
1973      if (mat->rmap->N != v3->map->N) SETERRQ2(PETSC_ERR_ARG_SIZ,"Mat mat,Vec v3: global dim %D %D",mat->rmap->N,v3->map->N); */
1974   if (mat->rmap->n != v3->map->n) SETERRQ2(PETSC_ERR_ARG_SIZ,"Mat mat,Vec v3: local dim %D %D",mat->rmap->n,v3->map->n);
1975   if (mat->rmap->n != v2->map->n) SETERRQ2(PETSC_ERR_ARG_SIZ,"Mat mat,Vec v2: local dim %D %D",mat->rmap->n,v2->map->n);
1976   if (v1 == v3) SETERRQ(PETSC_ERR_ARG_IDN,"v1 and v3 must be different vectors");
1977   ierr = MatPreallocated(mat);CHKERRQ(ierr);
1978 
1979   ierr = PetscLogEventBegin(MAT_MultAdd,mat,v1,v2,v3);CHKERRQ(ierr);
1980   ierr = (*mat->ops->multadd)(mat,v1,v2,v3);CHKERRQ(ierr);
1981   ierr = PetscLogEventEnd(MAT_MultAdd,mat,v1,v2,v3);CHKERRQ(ierr);
1982   ierr = PetscObjectStateIncrease((PetscObject)v3);CHKERRQ(ierr);
1983   PetscFunctionReturn(0);
1984 }
1985 
1986 #undef __FUNCT__
1987 #define __FUNCT__ "MatMultTransposeAdd"
1988 /*@
1989    MatMultTransposeAdd - Computes v3 = v2 + A' * v1.
1990 
1991    Collective on Mat and Vec
1992 
1993    Input Parameters:
1994 +  mat - the matrix
1995 -  v1, v2 - the vectors
1996 
1997    Output Parameters:
1998 .  v3 - the result
1999 
2000    Notes:
2001    The vectors v1 and v3 cannot be the same.  I.e., one cannot
2002    call MatMultTransposeAdd(A,v1,v2,v1).
2003 
2004    Level: beginner
2005 
2006    Concepts: matrix vector product^transpose and addition
2007 
2008 .seealso: MatMultTranspose(), MatMultAdd(), MatMult()
2009 @*/
2010 PetscErrorCode PETSCMAT_DLLEXPORT MatMultTransposeAdd(Mat mat,Vec v1,Vec v2,Vec v3)
2011 {
2012   PetscErrorCode ierr;
2013 
2014   PetscFunctionBegin;
2015   PetscValidHeaderSpecific(mat,MAT_COOKIE,1);
2016   PetscValidType(mat,1);
2017   PetscValidHeaderSpecific(v1,VEC_COOKIE,2);
2018   PetscValidHeaderSpecific(v2,VEC_COOKIE,3);
2019   PetscValidHeaderSpecific(v3,VEC_COOKIE,4);
2020 
2021   if (!mat->assembled) SETERRQ(PETSC_ERR_ARG_WRONGSTATE,"Not for unassembled matrix");
2022   if (mat->factor) SETERRQ(PETSC_ERR_ARG_WRONGSTATE,"Not for factored matrix");
2023   if (!mat->ops->multtransposeadd) SETERRQ1(PETSC_ERR_SUP,"Mat type %s",((PetscObject)mat)->type_name);
2024   if (v1 == v3) SETERRQ(PETSC_ERR_ARG_IDN,"v1 and v3 must be different vectors");
2025   if (mat->rmap->N != v1->map->N) SETERRQ2(PETSC_ERR_ARG_SIZ,"Mat mat,Vec v1: global dim %D %D",mat->rmap->N,v1->map->N);
2026   if (mat->cmap->N != v2->map->N) SETERRQ2(PETSC_ERR_ARG_SIZ,"Mat mat,Vec v2: global dim %D %D",mat->cmap->N,v2->map->N);
2027   if (mat->cmap->N != v3->map->N) SETERRQ2(PETSC_ERR_ARG_SIZ,"Mat mat,Vec v3: global dim %D %D",mat->cmap->N,v3->map->N);
2028   ierr = MatPreallocated(mat);CHKERRQ(ierr);
2029 
2030   ierr = PetscLogEventBegin(MAT_MultTransposeAdd,mat,v1,v2,v3);CHKERRQ(ierr);
2031   ierr = (*mat->ops->multtransposeadd)(mat,v1,v2,v3);CHKERRQ(ierr);
2032   ierr = PetscLogEventEnd(MAT_MultTransposeAdd,mat,v1,v2,v3);CHKERRQ(ierr);
2033   ierr = PetscObjectStateIncrease((PetscObject)v3);CHKERRQ(ierr);
2034   PetscFunctionReturn(0);
2035 }
2036 
2037 #undef __FUNCT__
2038 #define __FUNCT__ "MatMultConstrained"
2039 /*@
2040    MatMultConstrained - The inner multiplication routine for a
2041    constrained matrix P^T A P.
2042 
2043    Collective on Mat and Vec
2044 
2045    Input Parameters:
2046 +  mat - the matrix
2047 -  x   - the vector to be multilplied
2048 
2049    Output Parameters:
2050 .  y - the result
2051 
2052    Notes:
2053    The vectors x and y cannot be the same.  I.e., one cannot
2054    call MatMult(A,y,y).
2055 
2056    Level: beginner
2057 
2058 .keywords: matrix, multiply, matrix-vector product, constraint
2059 .seealso: MatMult(), MatMultTranspose(), MatMultAdd(), MatMultTransposeAdd()
2060 @*/
2061 PetscErrorCode PETSCMAT_DLLEXPORT MatMultConstrained(Mat mat,Vec x,Vec y)
2062 {
2063   PetscErrorCode ierr;
2064 
2065   PetscFunctionBegin;
2066   PetscValidHeaderSpecific(mat,MAT_COOKIE,1);
2067   PetscValidHeaderSpecific(x,VEC_COOKIE,2);
2068   PetscValidHeaderSpecific(y,VEC_COOKIE,3);
2069   if (!mat->assembled) SETERRQ(PETSC_ERR_ARG_WRONGSTATE,"Not for unassembled matrix");
2070   if (mat->factor) SETERRQ(PETSC_ERR_ARG_WRONGSTATE,"Not for factored matrix");
2071   if (x == y) SETERRQ(PETSC_ERR_ARG_WRONGSTATE,"x and y must be different vectors");
2072   if (mat->cmap->N != x->map->N) SETERRQ2(PETSC_ERR_ARG_SIZ,"Mat mat,Vec x: global dim %D %D",mat->cmap->N,x->map->N);
2073   if (mat->rmap->N != y->map->N) SETERRQ2(PETSC_ERR_ARG_SIZ,"Mat mat,Vec y: global dim %D %D",mat->rmap->N,y->map->N);
2074   if (mat->rmap->n != y->map->n) SETERRQ2(PETSC_ERR_ARG_SIZ,"Mat mat,Vec y: local dim %D %D",mat->rmap->n,y->map->n);
2075 
2076   ierr = PetscLogEventBegin(MAT_MultConstrained,mat,x,y,0);CHKERRQ(ierr);
2077   ierr = (*mat->ops->multconstrained)(mat,x,y);CHKERRQ(ierr);
2078   ierr = PetscLogEventEnd(MAT_MultConstrained,mat,x,y,0);CHKERRQ(ierr);
2079   ierr = PetscObjectStateIncrease((PetscObject)y);CHKERRQ(ierr);
2080 
2081   PetscFunctionReturn(0);
2082 }
2083 
2084 #undef __FUNCT__
2085 #define __FUNCT__ "MatMultTransposeConstrained"
2086 /*@
2087    MatMultTransposeConstrained - The inner multiplication routine for a
2088    constrained matrix P^T A^T P.
2089 
2090    Collective on Mat and Vec
2091 
2092    Input Parameters:
2093 +  mat - the matrix
2094 -  x   - the vector to be multilplied
2095 
2096    Output Parameters:
2097 .  y - the result
2098 
2099    Notes:
2100    The vectors x and y cannot be the same.  I.e., one cannot
2101    call MatMult(A,y,y).
2102 
2103    Level: beginner
2104 
2105 .keywords: matrix, multiply, matrix-vector product, constraint
2106 .seealso: MatMult(), MatMultTranspose(), MatMultAdd(), MatMultTransposeAdd()
2107 @*/
2108 PetscErrorCode PETSCMAT_DLLEXPORT MatMultTransposeConstrained(Mat mat,Vec x,Vec y)
2109 {
2110   PetscErrorCode ierr;
2111 
2112   PetscFunctionBegin;
2113   PetscValidHeaderSpecific(mat,MAT_COOKIE,1);
2114   PetscValidHeaderSpecific(x,VEC_COOKIE,2);
2115   PetscValidHeaderSpecific(y,VEC_COOKIE,3);
2116   if (!mat->assembled) SETERRQ(PETSC_ERR_ARG_WRONGSTATE,"Not for unassembled matrix");
2117   if (mat->factor) SETERRQ(PETSC_ERR_ARG_WRONGSTATE,"Not for factored matrix");
2118   if (x == y) SETERRQ(PETSC_ERR_ARG_WRONGSTATE,"x and y must be different vectors");
2119   if (mat->rmap->N != x->map->N) SETERRQ2(PETSC_ERR_ARG_SIZ,"Mat mat,Vec x: global dim %D %D",mat->cmap->N,x->map->N);
2120   if (mat->cmap->N != y->map->N) SETERRQ2(PETSC_ERR_ARG_SIZ,"Mat mat,Vec y: global dim %D %D",mat->rmap->N,y->map->N);
2121 
2122   ierr = PetscLogEventBegin(MAT_MultConstrained,mat,x,y,0);CHKERRQ(ierr);
2123   ierr = (*mat->ops->multtransposeconstrained)(mat,x,y);CHKERRQ(ierr);
2124   ierr = PetscLogEventEnd(MAT_MultConstrained,mat,x,y,0);CHKERRQ(ierr);
2125   ierr = PetscObjectStateIncrease((PetscObject)y);CHKERRQ(ierr);
2126 
2127   PetscFunctionReturn(0);
2128 }
2129 /* ------------------------------------------------------------*/
2130 #undef __FUNCT__
2131 #define __FUNCT__ "MatGetInfo"
2132 /*@C
2133    MatGetInfo - Returns information about matrix storage (number of
2134    nonzeros, memory, etc.).
2135 
2136    Collective on Mat if MAT_GLOBAL_MAX or MAT_GLOBAL_SUM is used
2137    as the flag
2138 
2139    Input Parameters:
2140 .  mat - the matrix
2141 
2142    Output Parameters:
2143 +  flag - flag indicating the type of parameters to be returned
2144    (MAT_LOCAL - local matrix, MAT_GLOBAL_MAX - maximum over all processors,
2145    MAT_GLOBAL_SUM - sum over all processors)
2146 -  info - matrix information context
2147 
2148    Notes:
2149    The MatInfo context contains a variety of matrix data, including
2150    number of nonzeros allocated and used, number of mallocs during
2151    matrix assembly, etc.  Additional information for factored matrices
2152    is provided (such as the fill ratio, number of mallocs during
2153    factorization, etc.).  Much of this info is printed to PETSC_STDOUT
2154    when using the runtime options
2155 $       -info -mat_view_info
2156 
2157    Example for C/C++ Users:
2158    See the file ${PETSC_DIR}/include/petscmat.h for a complete list of
2159    data within the MatInfo context.  For example,
2160 .vb
2161       MatInfo info;
2162       Mat     A;
2163       double  mal, nz_a, nz_u;
2164 
2165       MatGetInfo(A,MAT_LOCAL,&info);
2166       mal  = info.mallocs;
2167       nz_a = info.nz_allocated;
2168 .ve
2169 
2170    Example for Fortran Users:
2171    Fortran users should declare info as a double precision
2172    array of dimension MAT_INFO_SIZE, and then extract the parameters
2173    of interest.  See the file ${PETSC_DIR}/include/finclude/petscmat.h
2174    a complete list of parameter names.
2175 .vb
2176       double  precision info(MAT_INFO_SIZE)
2177       double  precision mal, nz_a
2178       Mat     A
2179       integer ierr
2180 
2181       call MatGetInfo(A,MAT_LOCAL,info,ierr)
2182       mal = info(MAT_INFO_MALLOCS)
2183       nz_a = info(MAT_INFO_NZ_ALLOCATED)
2184 .ve
2185 
2186     Level: intermediate
2187 
2188     Concepts: matrices^getting information on
2189 
2190     Developer Note: fortran interface is not autogenerated as the f90
2191     interface defintion cannot be generated correctly [due to MatInfo]
2192 
2193 @*/
2194 PetscErrorCode PETSCMAT_DLLEXPORT MatGetInfo(Mat mat,MatInfoType flag,MatInfo *info)
2195 {
2196   PetscErrorCode ierr;
2197 
2198   PetscFunctionBegin;
2199   PetscValidHeaderSpecific(mat,MAT_COOKIE,1);
2200   PetscValidType(mat,1);
2201   PetscValidPointer(info,3);
2202   if (!mat->ops->getinfo) SETERRQ1(PETSC_ERR_SUP,"Mat type %s",((PetscObject)mat)->type_name);
2203   ierr = MatPreallocated(mat);CHKERRQ(ierr);
2204   ierr = (*mat->ops->getinfo)(mat,flag,info);CHKERRQ(ierr);
2205   PetscFunctionReturn(0);
2206 }
2207 
2208 /* ----------------------------------------------------------*/
2209 #undef __FUNCT__
2210 #define __FUNCT__ "MatILUDTFactor"
2211 /*@C
2212    MatILUDTFactor - Performs a drop tolerance ILU factorization.
2213 
2214    Collective on Mat
2215 
2216    Input Parameters:
2217 +  mat - the matrix
2218 .  row - row permutation
2219 .  col - column permutation
2220 -  info - information about the factorization to be done
2221 
2222    Output Parameters:
2223 .  fact - the factored matrix
2224 
2225    Level: developer
2226 
2227    Notes:
2228    Most users should employ the simplified KSP interface for linear solvers
2229    instead of working directly with matrix algebra routines such as this.
2230    See, e.g., KSPCreate().
2231 
2232     Concepts: matrices^ILUDT factorization
2233 
2234 .seealso: MatLUFactorSymbolic(), MatLUFactorNumeric(), MatCholeskyFactor(), MatFactorInfo
2235 @*/
2236 PetscErrorCode PETSCMAT_DLLEXPORT MatILUDTFactor(Mat mat,IS row,IS col,const MatFactorInfo *info,Mat *fact)
2237 {
2238   PetscErrorCode ierr;
2239 
2240   PetscFunctionBegin;
2241   PetscValidHeaderSpecific(mat,MAT_COOKIE,1);
2242   PetscValidType(mat,1);
2243   if (row) PetscValidHeaderSpecific(row,IS_COOKIE,2);
2244   if (col) PetscValidHeaderSpecific(col,IS_COOKIE,3);
2245   PetscValidPointer(info,4);
2246   PetscValidPointer(fact,5);
2247   if (!mat->assembled) SETERRQ(PETSC_ERR_ARG_WRONGSTATE,"Not for unassembled matrix");
2248   if (mat->factor) SETERRQ(PETSC_ERR_ARG_WRONGSTATE,"Not for factored matrix");
2249   if (!mat->ops->iludtfactor) SETERRQ1(PETSC_ERR_SUP,"Mat type %s",((PetscObject)mat)->type_name);
2250   if (mat->rmap->N != mat->cmap->N) SETERRQ(PETSC_ERR_ARG_WRONG,"matrix must be square");
2251   ierr = MatPreallocated(mat);CHKERRQ(ierr);
2252   ierr = PetscLogEventBegin(MAT_ILUFactor,mat,row,col,0);CHKERRQ(ierr);
2253   ierr = (*mat->ops->iludtfactor)(mat,row,col,info,fact);CHKERRQ(ierr);
2254   ierr = PetscLogEventEnd(MAT_ILUFactor,mat,row,col,0);CHKERRQ(ierr);
2255   ierr = PetscObjectStateIncrease((PetscObject)*fact);CHKERRQ(ierr);
2256 
2257   PetscFunctionReturn(0);
2258 }
2259 
2260 #undef __FUNCT__
2261 #define __FUNCT__ "MatLUFactor"
2262 /*@C
2263    MatLUFactor - Performs in-place LU factorization of matrix.
2264 
2265    Collective on Mat
2266 
2267    Input Parameters:
2268 +  mat - the matrix
2269 .  row - row permutation
2270 .  col - column permutation
2271 -  info - options for factorization, includes
2272 $          fill - expected fill as ratio of original fill.
2273 $          dtcol - pivot tolerance (0 no pivot, 1 full column pivoting)
2274 $                   Run with the option -info to determine an optimal value to use
2275 
2276    Notes:
2277    Most users should employ the simplified KSP interface for linear solvers
2278    instead of working directly with matrix algebra routines such as this.
2279    See, e.g., KSPCreate().
2280 
2281    This changes the state of the matrix to a factored matrix; it cannot be used
2282    for example with MatSetValues() unless one first calls MatSetUnfactored().
2283 
2284    Level: developer
2285 
2286    Concepts: matrices^LU factorization
2287 
2288 .seealso: MatLUFactorSymbolic(), MatLUFactorNumeric(), MatCholeskyFactor(),
2289           MatGetOrdering(), MatSetUnfactored(), MatFactorInfo
2290 
2291     Developer Note: fortran interface is not autogenerated as the f90
2292     interface defintion cannot be generated correctly [due to MatFactorInfo]
2293 
2294 @*/
2295 PetscErrorCode PETSCMAT_DLLEXPORT MatLUFactor(Mat mat,IS row,IS col,const MatFactorInfo *info)
2296 {
2297   PetscErrorCode ierr;
2298 
2299   PetscFunctionBegin;
2300   PetscValidHeaderSpecific(mat,MAT_COOKIE,1);
2301   if (row) PetscValidHeaderSpecific(row,IS_COOKIE,2);
2302   if (col) PetscValidHeaderSpecific(col,IS_COOKIE,3);
2303   PetscValidPointer(info,4);
2304   PetscValidType(mat,1);
2305   if (!mat->assembled) SETERRQ(PETSC_ERR_ARG_WRONGSTATE,"Not for unassembled matrix");
2306   if (mat->factor) SETERRQ(PETSC_ERR_ARG_WRONGSTATE,"Not for factored matrix");
2307   if (!mat->ops->lufactor) SETERRQ1(PETSC_ERR_SUP,"Mat type %s",((PetscObject)mat)->type_name);
2308   ierr = MatPreallocated(mat);CHKERRQ(ierr);
2309 
2310   ierr = PetscLogEventBegin(MAT_LUFactor,mat,row,col,0);CHKERRQ(ierr);
2311   ierr = (*mat->ops->lufactor)(mat,row,col,info);CHKERRQ(ierr);
2312   ierr = PetscLogEventEnd(MAT_LUFactor,mat,row,col,0);CHKERRQ(ierr);
2313   ierr = PetscObjectStateIncrease((PetscObject)mat);CHKERRQ(ierr);
2314   PetscFunctionReturn(0);
2315 }
2316 
2317 #undef __FUNCT__
2318 #define __FUNCT__ "MatILUFactor"
2319 /*@C
2320    MatILUFactor - Performs in-place ILU factorization of matrix.
2321 
2322    Collective on Mat
2323 
2324    Input Parameters:
2325 +  mat - the matrix
2326 .  row - row permutation
2327 .  col - column permutation
2328 -  info - structure containing
2329 $      levels - number of levels of fill.
2330 $      expected fill - as ratio of original fill.
2331 $      1 or 0 - indicating force fill on diagonal (improves robustness for matrices
2332                 missing diagonal entries)
2333 
2334    Notes:
2335    Probably really in-place only when level of fill is zero, otherwise allocates
2336    new space to store factored matrix and deletes previous memory.
2337 
2338    Most users should employ the simplified KSP interface for linear solvers
2339    instead of working directly with matrix algebra routines such as this.
2340    See, e.g., KSPCreate().
2341 
2342    Level: developer
2343 
2344    Concepts: matrices^ILU factorization
2345 
2346 .seealso: MatILUFactorSymbolic(), MatLUFactorNumeric(), MatCholeskyFactor(), MatFactorInfo
2347 
2348     Developer Note: fortran interface is not autogenerated as the f90
2349     interface defintion cannot be generated correctly [due to MatFactorInfo]
2350 
2351 @*/
2352 PetscErrorCode PETSCMAT_DLLEXPORT MatILUFactor(Mat mat,IS row,IS col,const MatFactorInfo *info)
2353 {
2354   PetscErrorCode ierr;
2355 
2356   PetscFunctionBegin;
2357   PetscValidHeaderSpecific(mat,MAT_COOKIE,1);
2358   if (row) PetscValidHeaderSpecific(row,IS_COOKIE,2);
2359   if (col) PetscValidHeaderSpecific(col,IS_COOKIE,3);
2360   PetscValidPointer(info,4);
2361   PetscValidType(mat,1);
2362   if (mat->rmap->N != mat->cmap->N) SETERRQ(PETSC_ERR_ARG_WRONG,"matrix must be square");
2363   if (!mat->assembled) SETERRQ(PETSC_ERR_ARG_WRONGSTATE,"Not for unassembled matrix");
2364   if (mat->factor) SETERRQ(PETSC_ERR_ARG_WRONGSTATE,"Not for factored matrix");
2365   if (!mat->ops->ilufactor) SETERRQ1(PETSC_ERR_SUP,"Mat type %s",((PetscObject)mat)->type_name);
2366   ierr = MatPreallocated(mat);CHKERRQ(ierr);
2367 
2368   ierr = PetscLogEventBegin(MAT_ILUFactor,mat,row,col,0);CHKERRQ(ierr);
2369   ierr = (*mat->ops->ilufactor)(mat,row,col,info);CHKERRQ(ierr);
2370   ierr = PetscLogEventEnd(MAT_ILUFactor,mat,row,col,0);CHKERRQ(ierr);
2371   ierr = PetscObjectStateIncrease((PetscObject)mat);CHKERRQ(ierr);
2372   PetscFunctionReturn(0);
2373 }
2374 
2375 #undef __FUNCT__
2376 #define __FUNCT__ "MatLUFactorSymbolic"
2377 /*@C
2378    MatLUFactorSymbolic - Performs symbolic LU factorization of matrix.
2379    Call this routine before calling MatLUFactorNumeric().
2380 
2381    Collective on Mat
2382 
2383    Input Parameters:
2384 +  fact - the factor matrix obtained with MatGetFactor()
2385 .  mat - the matrix
2386 .  row, col - row and column permutations
2387 -  info - options for factorization, includes
2388 $          fill - expected fill as ratio of original fill.
2389 $          dtcol - pivot tolerance (0 no pivot, 1 full column pivoting)
2390 $                   Run with the option -info to determine an optimal value to use
2391 
2392 
2393    Notes:
2394    See the users manual for additional information about
2395    choosing the fill factor for better efficiency.
2396 
2397    Most users should employ the simplified KSP interface for linear solvers
2398    instead of working directly with matrix algebra routines such as this.
2399    See, e.g., KSPCreate().
2400 
2401    Level: developer
2402 
2403    Concepts: matrices^LU symbolic factorization
2404 
2405 .seealso: MatLUFactor(), MatLUFactorNumeric(), MatCholeskyFactor(), MatFactorInfo
2406 
2407     Developer Note: fortran interface is not autogenerated as the f90
2408     interface defintion cannot be generated correctly [due to MatFactorInfo]
2409 
2410 @*/
2411 PetscErrorCode PETSCMAT_DLLEXPORT MatLUFactorSymbolic(Mat fact,Mat mat,IS row,IS col,const MatFactorInfo *info)
2412 {
2413   PetscErrorCode ierr;
2414 
2415   PetscFunctionBegin;
2416   PetscValidHeaderSpecific(mat,MAT_COOKIE,1);
2417   if (row) PetscValidHeaderSpecific(row,IS_COOKIE,2);
2418   if (col) PetscValidHeaderSpecific(col,IS_COOKIE,3);
2419   PetscValidPointer(info,4);
2420   PetscValidType(mat,1);
2421   PetscValidPointer(fact,5);
2422   if (!mat->assembled) SETERRQ(PETSC_ERR_ARG_WRONGSTATE,"Not for unassembled matrix");
2423   if (mat->factor) SETERRQ(PETSC_ERR_ARG_WRONGSTATE,"Not for factored matrix");
2424   if (!(fact)->ops->lufactorsymbolic) SETERRQ1(PETSC_ERR_SUP,"Matrix type %s  symbolic LU",((PetscObject)mat)->type_name);
2425   ierr = MatPreallocated(mat);CHKERRQ(ierr);
2426 
2427   ierr = PetscLogEventBegin(MAT_LUFactorSymbolic,mat,row,col,0);CHKERRQ(ierr);
2428   ierr = (fact->ops->lufactorsymbolic)(fact,mat,row,col,info);CHKERRQ(ierr);
2429   ierr = PetscLogEventEnd(MAT_LUFactorSymbolic,mat,row,col,0);CHKERRQ(ierr);
2430   ierr = PetscObjectStateIncrease((PetscObject)fact);CHKERRQ(ierr);
2431   PetscFunctionReturn(0);
2432 }
2433 
2434 #undef __FUNCT__
2435 #define __FUNCT__ "MatLUFactorNumeric"
2436 /*@C
2437    MatLUFactorNumeric - Performs numeric LU factorization of a matrix.
2438    Call this routine after first calling MatLUFactorSymbolic().
2439 
2440    Collective on Mat
2441 
2442    Input Parameters:
2443 +  fact - the factor matrix obtained with MatGetFactor()
2444 .  mat - the matrix
2445 -  info - options for factorization
2446 
2447    Notes:
2448    See MatLUFactor() for in-place factorization.  See
2449    MatCholeskyFactorNumeric() for the symmetric, positive definite case.
2450 
2451    Most users should employ the simplified KSP interface for linear solvers
2452    instead of working directly with matrix algebra routines such as this.
2453    See, e.g., KSPCreate().
2454 
2455    Level: developer
2456 
2457    Concepts: matrices^LU numeric factorization
2458 
2459 .seealso: MatLUFactorSymbolic(), MatLUFactor(), MatCholeskyFactor()
2460 
2461     Developer Note: fortran interface is not autogenerated as the f90
2462     interface defintion cannot be generated correctly [due to MatFactorInfo]
2463 
2464 @*/
2465 PetscErrorCode PETSCMAT_DLLEXPORT MatLUFactorNumeric(Mat fact,Mat mat,const MatFactorInfo *info)
2466 {
2467   PetscErrorCode ierr;
2468 
2469   PetscFunctionBegin;
2470   PetscValidHeaderSpecific(mat,MAT_COOKIE,1);
2471   PetscValidType(mat,1);
2472   PetscValidPointer(fact,2);
2473   PetscValidHeaderSpecific(fact,MAT_COOKIE,2);
2474   if (!mat->assembled) SETERRQ(PETSC_ERR_ARG_WRONGSTATE,"Not for unassembled matrix");
2475   if (mat->rmap->N != (fact)->rmap->N || mat->cmap->N != (fact)->cmap->N) {
2476     SETERRQ4(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);
2477   }
2478   if (!(fact)->ops->lufactornumeric) SETERRQ1(PETSC_ERR_SUP,"Mat type %s",((PetscObject)mat)->type_name);
2479   ierr = MatPreallocated(mat);CHKERRQ(ierr);
2480   ierr = PetscLogEventBegin(MAT_LUFactorNumeric,mat,fact,0,0);CHKERRQ(ierr);
2481   ierr = (fact->ops->lufactornumeric)(fact,mat,info);CHKERRQ(ierr);
2482   ierr = PetscLogEventEnd(MAT_LUFactorNumeric,mat,fact,0,0);CHKERRQ(ierr);
2483 
2484   ierr = MatView_Private(fact);CHKERRQ(ierr);
2485   ierr = PetscObjectStateIncrease((PetscObject)fact);CHKERRQ(ierr);
2486   PetscFunctionReturn(0);
2487 }
2488 
2489 #undef __FUNCT__
2490 #define __FUNCT__ "MatCholeskyFactor"
2491 /*@C
2492    MatCholeskyFactor - Performs in-place Cholesky factorization of a
2493    symmetric matrix.
2494 
2495    Collective on Mat
2496 
2497    Input Parameters:
2498 +  mat - the matrix
2499 .  perm - row and column permutations
2500 -  f - expected fill as ratio of original fill
2501 
2502    Notes:
2503    See MatLUFactor() for the nonsymmetric case.  See also
2504    MatCholeskyFactorSymbolic(), and MatCholeskyFactorNumeric().
2505 
2506    Most users should employ the simplified KSP interface for linear solvers
2507    instead of working directly with matrix algebra routines such as this.
2508    See, e.g., KSPCreate().
2509 
2510    Level: developer
2511 
2512    Concepts: matrices^Cholesky factorization
2513 
2514 .seealso: MatLUFactor(), MatCholeskyFactorSymbolic(), MatCholeskyFactorNumeric()
2515           MatGetOrdering()
2516 
2517     Developer Note: fortran interface is not autogenerated as the f90
2518     interface defintion cannot be generated correctly [due to MatFactorInfo]
2519 
2520 @*/
2521 PetscErrorCode PETSCMAT_DLLEXPORT MatCholeskyFactor(Mat mat,IS perm,const MatFactorInfo *info)
2522 {
2523   PetscErrorCode ierr;
2524 
2525   PetscFunctionBegin;
2526   PetscValidHeaderSpecific(mat,MAT_COOKIE,1);
2527   PetscValidType(mat,1);
2528   PetscValidHeaderSpecific(perm,IS_COOKIE,2);
2529   PetscValidPointer(info,3);
2530   if (mat->rmap->N != mat->cmap->N) SETERRQ(PETSC_ERR_ARG_WRONG,"Matrix must be square");
2531   if (!mat->assembled) SETERRQ(PETSC_ERR_ARG_WRONGSTATE,"Not for unassembled matrix");
2532   if (mat->factor) SETERRQ(PETSC_ERR_ARG_WRONGSTATE,"Not for factored matrix");
2533   if (!mat->ops->choleskyfactor) SETERRQ1(PETSC_ERR_SUP,"Mat type %s",((PetscObject)mat)->type_name);
2534   ierr = MatPreallocated(mat);CHKERRQ(ierr);
2535 
2536   ierr = PetscLogEventBegin(MAT_CholeskyFactor,mat,perm,0,0);CHKERRQ(ierr);
2537   ierr = (*mat->ops->choleskyfactor)(mat,perm,info);CHKERRQ(ierr);
2538   ierr = PetscLogEventEnd(MAT_CholeskyFactor,mat,perm,0,0);CHKERRQ(ierr);
2539   ierr = PetscObjectStateIncrease((PetscObject)mat);CHKERRQ(ierr);
2540   PetscFunctionReturn(0);
2541 }
2542 
2543 #undef __FUNCT__
2544 #define __FUNCT__ "MatCholeskyFactorSymbolic"
2545 /*@C
2546    MatCholeskyFactorSymbolic - Performs symbolic Cholesky factorization
2547    of a symmetric matrix.
2548 
2549    Collective on Mat
2550 
2551    Input Parameters:
2552 +  fact - the factor matrix obtained with MatGetFactor()
2553 .  mat - the matrix
2554 .  perm - row and column permutations
2555 -  info - options for factorization, includes
2556 $          fill - expected fill as ratio of original fill.
2557 $          dtcol - pivot tolerance (0 no pivot, 1 full column pivoting)
2558 $                   Run with the option -info to determine an optimal value to use
2559 
2560    Notes:
2561    See MatLUFactorSymbolic() for the nonsymmetric case.  See also
2562    MatCholeskyFactor() and MatCholeskyFactorNumeric().
2563 
2564    Most users should employ the simplified KSP interface for linear solvers
2565    instead of working directly with matrix algebra routines such as this.
2566    See, e.g., KSPCreate().
2567 
2568    Level: developer
2569 
2570    Concepts: matrices^Cholesky symbolic factorization
2571 
2572 .seealso: MatLUFactorSymbolic(), MatCholeskyFactor(), MatCholeskyFactorNumeric()
2573           MatGetOrdering()
2574 
2575     Developer Note: fortran interface is not autogenerated as the f90
2576     interface defintion cannot be generated correctly [due to MatFactorInfo]
2577 
2578 @*/
2579 PetscErrorCode PETSCMAT_DLLEXPORT MatCholeskyFactorSymbolic(Mat fact,Mat mat,IS perm,const MatFactorInfo *info)
2580 {
2581   PetscErrorCode ierr;
2582 
2583   PetscFunctionBegin;
2584   PetscValidHeaderSpecific(mat,MAT_COOKIE,1);
2585   PetscValidType(mat,1);
2586   if (perm) PetscValidHeaderSpecific(perm,IS_COOKIE,2);
2587   PetscValidPointer(info,3);
2588   PetscValidPointer(fact,4);
2589   if (mat->rmap->N != mat->cmap->N) SETERRQ(PETSC_ERR_ARG_WRONG,"Matrix must be square");
2590   if (!mat->assembled) SETERRQ(PETSC_ERR_ARG_WRONGSTATE,"Not for unassembled matrix");
2591   if (mat->factor) SETERRQ(PETSC_ERR_ARG_WRONGSTATE,"Not for factored matrix");
2592   if (!(fact)->ops->choleskyfactorsymbolic) SETERRQ1(PETSC_ERR_SUP,"Mat type %s",((PetscObject)mat)->type_name);
2593   ierr = MatPreallocated(mat);CHKERRQ(ierr);
2594 
2595   ierr = PetscLogEventBegin(MAT_CholeskyFactorSymbolic,mat,perm,0,0);CHKERRQ(ierr);
2596   ierr = (fact->ops->choleskyfactorsymbolic)(fact,mat,perm,info);CHKERRQ(ierr);
2597   ierr = PetscLogEventEnd(MAT_CholeskyFactorSymbolic,mat,perm,0,0);CHKERRQ(ierr);
2598   ierr = PetscObjectStateIncrease((PetscObject)fact);CHKERRQ(ierr);
2599   PetscFunctionReturn(0);
2600 }
2601 
2602 #undef __FUNCT__
2603 #define __FUNCT__ "MatCholeskyFactorNumeric"
2604 /*@C
2605    MatCholeskyFactorNumeric - Performs numeric Cholesky factorization
2606    of a symmetric matrix. Call this routine after first calling
2607    MatCholeskyFactorSymbolic().
2608 
2609    Collective on Mat
2610 
2611    Input Parameters:
2612 +  fact - the factor matrix obtained with MatGetFactor()
2613 .  mat - the initial matrix
2614 .  info - options for factorization
2615 -  fact - the symbolic factor of mat
2616 
2617 
2618    Notes:
2619    Most users should employ the simplified KSP interface for linear solvers
2620    instead of working directly with matrix algebra routines such as this.
2621    See, e.g., KSPCreate().
2622 
2623    Level: developer
2624 
2625    Concepts: matrices^Cholesky numeric factorization
2626 
2627 .seealso: MatCholeskyFactorSymbolic(), MatCholeskyFactor(), MatLUFactorNumeric()
2628 
2629     Developer Note: fortran interface is not autogenerated as the f90
2630     interface defintion cannot be generated correctly [due to MatFactorInfo]
2631 
2632 @*/
2633 PetscErrorCode PETSCMAT_DLLEXPORT MatCholeskyFactorNumeric(Mat fact,Mat mat,const MatFactorInfo *info)
2634 {
2635   PetscErrorCode ierr;
2636 
2637   PetscFunctionBegin;
2638   PetscValidHeaderSpecific(mat,MAT_COOKIE,1);
2639   PetscValidType(mat,1);
2640   PetscValidPointer(fact,2);
2641   PetscValidHeaderSpecific(fact,MAT_COOKIE,2);
2642   if (!mat->assembled) SETERRQ(PETSC_ERR_ARG_WRONGSTATE,"Not for unassembled matrix");
2643   if (!(fact)->ops->choleskyfactornumeric) SETERRQ1(PETSC_ERR_SUP,"Mat type %s",((PetscObject)mat)->type_name);
2644   if (mat->rmap->N != (fact)->rmap->N || mat->cmap->N != (fact)->cmap->N) {
2645     SETERRQ4(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);
2646   }
2647   ierr = MatPreallocated(mat);CHKERRQ(ierr);
2648 
2649   ierr = PetscLogEventBegin(MAT_CholeskyFactorNumeric,mat,fact,0,0);CHKERRQ(ierr);
2650   ierr = (fact->ops->choleskyfactornumeric)(fact,mat,info);CHKERRQ(ierr);
2651   ierr = PetscLogEventEnd(MAT_CholeskyFactorNumeric,mat,fact,0,0);CHKERRQ(ierr);
2652 
2653   ierr = MatView_Private(fact);CHKERRQ(ierr);
2654   ierr = PetscObjectStateIncrease((PetscObject)fact);CHKERRQ(ierr);
2655   PetscFunctionReturn(0);
2656 }
2657 
2658 /* ----------------------------------------------------------------*/
2659 #undef __FUNCT__
2660 #define __FUNCT__ "MatSolve"
2661 /*@
2662    MatSolve - Solves A x = b, given a factored matrix.
2663 
2664    Collective on Mat and Vec
2665 
2666    Input Parameters:
2667 +  mat - the factored matrix
2668 -  b - the right-hand-side vector
2669 
2670    Output Parameter:
2671 .  x - the result vector
2672 
2673    Notes:
2674    The vectors b and x cannot be the same.  I.e., one cannot
2675    call MatSolve(A,x,x).
2676 
2677    Notes:
2678    Most users should employ the simplified KSP interface for linear solvers
2679    instead of working directly with matrix algebra routines such as this.
2680    See, e.g., KSPCreate().
2681 
2682    Level: developer
2683 
2684    Concepts: matrices^triangular solves
2685 
2686 .seealso: MatSolveAdd(), MatSolveTranspose(), MatSolveTransposeAdd()
2687 @*/
2688 PetscErrorCode PETSCMAT_DLLEXPORT MatSolve(Mat mat,Vec b,Vec x)
2689 {
2690   PetscErrorCode ierr;
2691 
2692   PetscFunctionBegin;
2693   PetscValidHeaderSpecific(mat,MAT_COOKIE,1);
2694   PetscValidType(mat,1);
2695   PetscValidHeaderSpecific(b,VEC_COOKIE,2);
2696   PetscValidHeaderSpecific(x,VEC_COOKIE,3);
2697   PetscCheckSameComm(mat,1,b,2);
2698   PetscCheckSameComm(mat,1,x,3);
2699   if (x == b) SETERRQ(PETSC_ERR_ARG_IDN,"x and b must be different vectors");
2700   if (!mat->factor) SETERRQ(PETSC_ERR_ARG_WRONGSTATE,"Unfactored matrix");
2701   if (mat->cmap->N != x->map->N) SETERRQ2(PETSC_ERR_ARG_SIZ,"Mat mat,Vec x: global dim %D %D",mat->cmap->N,x->map->N);
2702   if (mat->rmap->N != b->map->N) SETERRQ2(PETSC_ERR_ARG_SIZ,"Mat mat,Vec b: global dim %D %D",mat->rmap->N,b->map->N);
2703   if (mat->rmap->n != b->map->n) SETERRQ2(PETSC_ERR_ARG_SIZ,"Mat mat,Vec b: local dim %D %D",mat->rmap->n,b->map->n);
2704   if (!mat->rmap->N && !mat->cmap->N) PetscFunctionReturn(0);
2705   if (!mat->ops->solve) SETERRQ1(PETSC_ERR_SUP,"Mat type %s",((PetscObject)mat)->type_name);
2706   ierr = MatPreallocated(mat);CHKERRQ(ierr);
2707 
2708   ierr = PetscLogEventBegin(MAT_Solve,mat,b,x,0);CHKERRQ(ierr);
2709   ierr = (*mat->ops->solve)(mat,b,x);CHKERRQ(ierr);
2710   ierr = PetscLogEventEnd(MAT_Solve,mat,b,x,0);CHKERRQ(ierr);
2711   ierr = PetscObjectStateIncrease((PetscObject)x);CHKERRQ(ierr);
2712   PetscFunctionReturn(0);
2713 }
2714 
2715 #undef __FUNCT__
2716 #define __FUNCT__ "MatMatSolve_Basic"
2717 PetscErrorCode PETSCMAT_DLLEXPORT MatMatSolve_Basic(Mat A,Mat B,Mat X)
2718 {
2719   PetscErrorCode ierr;
2720   Vec            b,x;
2721   PetscInt       m,N,i;
2722   PetscScalar    *bb,*xx;
2723 
2724   PetscFunctionBegin;
2725   ierr = MatGetArray(B,&bb);CHKERRQ(ierr);
2726   ierr = MatGetArray(X,&xx);CHKERRQ(ierr);
2727   ierr = MatGetLocalSize(B,&m,PETSC_NULL);CHKERRQ(ierr);  /* number local rows */
2728   ierr = MatGetSize(B,PETSC_NULL,&N);CHKERRQ(ierr);       /* total columns in dense matrix */
2729   ierr = MatGetVecs(B,&x,&b);CHKERRQ(ierr);
2730   for (i=0; i<N; i++) {
2731     ierr = VecPlaceArray(b,bb + i*m);CHKERRQ(ierr);
2732     ierr = VecPlaceArray(x,xx + i*m);CHKERRQ(ierr);
2733     ierr = MatSolve(A,b,x);CHKERRQ(ierr);
2734     ierr = VecResetArray(x);CHKERRQ(ierr);
2735     ierr = VecResetArray(b);CHKERRQ(ierr);
2736   }
2737   ierr = VecDestroy(b);CHKERRQ(ierr);
2738   ierr = VecDestroy(x);CHKERRQ(ierr);
2739   ierr = MatRestoreArray(B,&bb);CHKERRQ(ierr);
2740   ierr = MatRestoreArray(X,&xx);CHKERRQ(ierr);
2741   PetscFunctionReturn(0);
2742 }
2743 
2744 #undef __FUNCT__
2745 #define __FUNCT__ "MatMatSolve"
2746 /*@
2747    MatMatSolve - Solves A X = B, given a factored matrix.
2748 
2749    Collective on Mat
2750 
2751    Input Parameters:
2752 +  mat - the factored matrix
2753 -  B - the right-hand-side matrix  (dense matrix)
2754 
2755    Output Parameter:
2756 .  X - the result matrix (dense matrix)
2757 
2758    Notes:
2759    The matrices b and x cannot be the same.  I.e., one cannot
2760    call MatMatSolve(A,x,x).
2761 
2762    Notes:
2763    Most users should usually employ the simplified KSP interface for linear solvers
2764    instead of working directly with matrix algebra routines such as this.
2765    See, e.g., KSPCreate(). However KSP can only solve for one vector (column of X)
2766    at a time.
2767 
2768    Level: developer
2769 
2770    Concepts: matrices^triangular solves
2771 
2772 .seealso: MatMatSolveAdd(), MatMatSolveTranspose(), MatMatSolveTransposeAdd(), MatLUFactor(), MatCholeskyFactor()
2773 @*/
2774 PetscErrorCode PETSCMAT_DLLEXPORT MatMatSolve(Mat A,Mat B,Mat X)
2775 {
2776   PetscErrorCode ierr;
2777 
2778   PetscFunctionBegin;
2779   PetscValidHeaderSpecific(A,MAT_COOKIE,1);
2780   PetscValidType(A,1);
2781   PetscValidHeaderSpecific(B,MAT_COOKIE,2);
2782   PetscValidHeaderSpecific(X,MAT_COOKIE,3);
2783   PetscCheckSameComm(A,1,B,2);
2784   PetscCheckSameComm(A,1,X,3);
2785   if (X == B) SETERRQ(PETSC_ERR_ARG_IDN,"X and B must be different matrices");
2786   if (!A->factor) SETERRQ(PETSC_ERR_ARG_WRONGSTATE,"Unfactored matrix");
2787   if (A->cmap->N != X->rmap->N) SETERRQ2(PETSC_ERR_ARG_SIZ,"Mat A,Mat X: global dim %D %D",A->cmap->N,X->rmap->N);
2788   if (A->rmap->N != B->rmap->N) SETERRQ2(PETSC_ERR_ARG_SIZ,"Mat A,Mat B: global dim %D %D",A->rmap->N,B->rmap->N);
2789   if (A->rmap->n != B->rmap->n) SETERRQ2(PETSC_ERR_ARG_SIZ,"Mat A,Mat B: local dim %D %D",A->rmap->n,B->rmap->n);
2790   if (!A->rmap->N && !A->cmap->N) PetscFunctionReturn(0);
2791   ierr = MatPreallocated(A);CHKERRQ(ierr);
2792 
2793   ierr = PetscLogEventBegin(MAT_MatSolve,A,B,X,0);CHKERRQ(ierr);
2794   if (!A->ops->matsolve) {
2795     ierr = PetscInfo1(A,"Mat type %s using basic MatMatSolve",((PetscObject)A)->type_name);CHKERRQ(ierr);
2796     ierr = MatMatSolve_Basic(A,B,X);CHKERRQ(ierr);
2797   } else {
2798     ierr = (*A->ops->matsolve)(A,B,X);CHKERRQ(ierr);
2799   }
2800   ierr = PetscLogEventEnd(MAT_MatSolve,A,B,X,0);CHKERRQ(ierr);
2801   ierr = PetscObjectStateIncrease((PetscObject)X);CHKERRQ(ierr);
2802   PetscFunctionReturn(0);
2803 }
2804 
2805 
2806 #undef __FUNCT__
2807 #define __FUNCT__ "MatForwardSolve"
2808 /* @
2809    MatForwardSolve - Solves L x = b, given a factored matrix, A = LU, or
2810                             U^T*D^(1/2) x = b, given a factored symmetric matrix, A = U^T*D*U,
2811 
2812    Collective on Mat and Vec
2813 
2814    Input Parameters:
2815 +  mat - the factored matrix
2816 -  b - the right-hand-side vector
2817 
2818    Output Parameter:
2819 .  x - the result vector
2820 
2821    Notes:
2822    MatSolve() should be used for most applications, as it performs
2823    a forward solve followed by a backward solve.
2824 
2825    The vectors b and x cannot be the same,  i.e., one cannot
2826    call MatForwardSolve(A,x,x).
2827 
2828    For matrix in seqsbaij format with block size larger than 1,
2829    the diagonal blocks are not implemented as D = D^(1/2) * D^(1/2) yet.
2830    MatForwardSolve() solves U^T*D y = b, and
2831    MatBackwardSolve() solves U x = y.
2832    Thus they do not provide a symmetric preconditioner.
2833 
2834    Most users should employ the simplified KSP interface for linear solvers
2835    instead of working directly with matrix algebra routines such as this.
2836    See, e.g., KSPCreate().
2837 
2838    Level: developer
2839 
2840    Concepts: matrices^forward solves
2841 
2842 .seealso: MatSolve(), MatBackwardSolve()
2843 @ */
2844 PetscErrorCode PETSCMAT_DLLEXPORT MatForwardSolve(Mat mat,Vec b,Vec x)
2845 {
2846   PetscErrorCode ierr;
2847 
2848   PetscFunctionBegin;
2849   PetscValidHeaderSpecific(mat,MAT_COOKIE,1);
2850   PetscValidType(mat,1);
2851   PetscValidHeaderSpecific(b,VEC_COOKIE,2);
2852   PetscValidHeaderSpecific(x,VEC_COOKIE,3);
2853   PetscCheckSameComm(mat,1,b,2);
2854   PetscCheckSameComm(mat,1,x,3);
2855   if (x == b) SETERRQ(PETSC_ERR_ARG_IDN,"x and b must be different vectors");
2856   if (!mat->factor) SETERRQ(PETSC_ERR_ARG_WRONGSTATE,"Unfactored matrix");
2857   if (!mat->ops->forwardsolve) SETERRQ1(PETSC_ERR_SUP,"Mat type %s",((PetscObject)mat)->type_name);
2858   if (mat->cmap->N != x->map->N) SETERRQ2(PETSC_ERR_ARG_SIZ,"Mat mat,Vec x: global dim %D %D",mat->cmap->N,x->map->N);
2859   if (mat->rmap->N != b->map->N) SETERRQ2(PETSC_ERR_ARG_SIZ,"Mat mat,Vec b: global dim %D %D",mat->rmap->N,b->map->N);
2860   if (mat->rmap->n != b->map->n) SETERRQ2(PETSC_ERR_ARG_SIZ,"Mat mat,Vec b: local dim %D %D",mat->rmap->n,b->map->n);
2861   ierr = MatPreallocated(mat);CHKERRQ(ierr);
2862   ierr = PetscLogEventBegin(MAT_ForwardSolve,mat,b,x,0);CHKERRQ(ierr);
2863   ierr = (*mat->ops->forwardsolve)(mat,b,x);CHKERRQ(ierr);
2864   ierr = PetscLogEventEnd(MAT_ForwardSolve,mat,b,x,0);CHKERRQ(ierr);
2865   ierr = PetscObjectStateIncrease((PetscObject)x);CHKERRQ(ierr);
2866   PetscFunctionReturn(0);
2867 }
2868 
2869 #undef __FUNCT__
2870 #define __FUNCT__ "MatBackwardSolve"
2871 /* @
2872    MatBackwardSolve - Solves U x = b, given a factored matrix, A = LU.
2873                              D^(1/2) U x = b, given a factored symmetric matrix, A = U^T*D*U,
2874 
2875    Collective on Mat and Vec
2876 
2877    Input Parameters:
2878 +  mat - the factored matrix
2879 -  b - the right-hand-side vector
2880 
2881    Output Parameter:
2882 .  x - the result vector
2883 
2884    Notes:
2885    MatSolve() should be used for most applications, as it performs
2886    a forward solve followed by a backward solve.
2887 
2888    The vectors b and x cannot be the same.  I.e., one cannot
2889    call MatBackwardSolve(A,x,x).
2890 
2891    For matrix in seqsbaij format with block size larger than 1,
2892    the diagonal blocks are not implemented as D = D^(1/2) * D^(1/2) yet.
2893    MatForwardSolve() solves U^T*D y = b, and
2894    MatBackwardSolve() solves U x = y.
2895    Thus they do not provide a symmetric preconditioner.
2896 
2897    Most users should employ the simplified KSP interface for linear solvers
2898    instead of working directly with matrix algebra routines such as this.
2899    See, e.g., KSPCreate().
2900 
2901    Level: developer
2902 
2903    Concepts: matrices^backward solves
2904 
2905 .seealso: MatSolve(), MatForwardSolve()
2906 @ */
2907 PetscErrorCode PETSCMAT_DLLEXPORT MatBackwardSolve(Mat mat,Vec b,Vec x)
2908 {
2909   PetscErrorCode ierr;
2910 
2911   PetscFunctionBegin;
2912   PetscValidHeaderSpecific(mat,MAT_COOKIE,1);
2913   PetscValidType(mat,1);
2914   PetscValidHeaderSpecific(b,VEC_COOKIE,2);
2915   PetscValidHeaderSpecific(x,VEC_COOKIE,3);
2916   PetscCheckSameComm(mat,1,b,2);
2917   PetscCheckSameComm(mat,1,x,3);
2918   if (x == b) SETERRQ(PETSC_ERR_ARG_IDN,"x and b must be different vectors");
2919   if (!mat->factor) SETERRQ(PETSC_ERR_ARG_WRONGSTATE,"Unfactored matrix");
2920   if (!mat->ops->backwardsolve) SETERRQ1(PETSC_ERR_SUP,"Mat type %s",((PetscObject)mat)->type_name);
2921   if (mat->cmap->N != x->map->N) SETERRQ2(PETSC_ERR_ARG_SIZ,"Mat mat,Vec x: global dim %D %D",mat->cmap->N,x->map->N);
2922   if (mat->rmap->N != b->map->N) SETERRQ2(PETSC_ERR_ARG_SIZ,"Mat mat,Vec b: global dim %D %D",mat->rmap->N,b->map->N);
2923   if (mat->rmap->n != b->map->n) SETERRQ2(PETSC_ERR_ARG_SIZ,"Mat mat,Vec b: local dim %D %D",mat->rmap->n,b->map->n);
2924   ierr = MatPreallocated(mat);CHKERRQ(ierr);
2925 
2926   ierr = PetscLogEventBegin(MAT_BackwardSolve,mat,b,x,0);CHKERRQ(ierr);
2927   ierr = (*mat->ops->backwardsolve)(mat,b,x);CHKERRQ(ierr);
2928   ierr = PetscLogEventEnd(MAT_BackwardSolve,mat,b,x,0);CHKERRQ(ierr);
2929   ierr = PetscObjectStateIncrease((PetscObject)x);CHKERRQ(ierr);
2930   PetscFunctionReturn(0);
2931 }
2932 
2933 #undef __FUNCT__
2934 #define __FUNCT__ "MatSolveAdd"
2935 /*@
2936    MatSolveAdd - Computes x = y + inv(A)*b, given a factored matrix.
2937 
2938    Collective on Mat and Vec
2939 
2940    Input Parameters:
2941 +  mat - the factored matrix
2942 .  b - the right-hand-side vector
2943 -  y - the vector to be added to
2944 
2945    Output Parameter:
2946 .  x - the result vector
2947 
2948    Notes:
2949    The vectors b and x cannot be the same.  I.e., one cannot
2950    call MatSolveAdd(A,x,y,x).
2951 
2952    Most users should employ the simplified KSP interface for linear solvers
2953    instead of working directly with matrix algebra routines such as this.
2954    See, e.g., KSPCreate().
2955 
2956    Level: developer
2957 
2958    Concepts: matrices^triangular solves
2959 
2960 .seealso: MatSolve(), MatSolveTranspose(), MatSolveTransposeAdd()
2961 @*/
2962 PetscErrorCode PETSCMAT_DLLEXPORT MatSolveAdd(Mat mat,Vec b,Vec y,Vec x)
2963 {
2964   PetscScalar    one = 1.0;
2965   Vec            tmp;
2966   PetscErrorCode ierr;
2967 
2968   PetscFunctionBegin;
2969   PetscValidHeaderSpecific(mat,MAT_COOKIE,1);
2970   PetscValidType(mat,1);
2971   PetscValidHeaderSpecific(y,VEC_COOKIE,2);
2972   PetscValidHeaderSpecific(b,VEC_COOKIE,3);
2973   PetscValidHeaderSpecific(x,VEC_COOKIE,4);
2974   PetscCheckSameComm(mat,1,b,2);
2975   PetscCheckSameComm(mat,1,y,2);
2976   PetscCheckSameComm(mat,1,x,3);
2977   if (x == b) SETERRQ(PETSC_ERR_ARG_IDN,"x and b must be different vectors");
2978   if (!mat->factor) SETERRQ(PETSC_ERR_ARG_WRONGSTATE,"Unfactored matrix");
2979   if (mat->cmap->N != x->map->N) SETERRQ2(PETSC_ERR_ARG_SIZ,"Mat mat,Vec x: global dim %D %D",mat->cmap->N,x->map->N);
2980   if (mat->rmap->N != b->map->N) SETERRQ2(PETSC_ERR_ARG_SIZ,"Mat mat,Vec b: global dim %D %D",mat->rmap->N,b->map->N);
2981   if (mat->rmap->N != y->map->N) SETERRQ2(PETSC_ERR_ARG_SIZ,"Mat mat,Vec y: global dim %D %D",mat->rmap->N,y->map->N);
2982   if (mat->rmap->n != b->map->n) SETERRQ2(PETSC_ERR_ARG_SIZ,"Mat mat,Vec b: local dim %D %D",mat->rmap->n,b->map->n);
2983   if (x->map->n != y->map->n) SETERRQ2(PETSC_ERR_ARG_SIZ,"Vec x,Vec y: local dim %D %D",x->map->n,y->map->n);
2984   ierr = MatPreallocated(mat);CHKERRQ(ierr);
2985 
2986   ierr = PetscLogEventBegin(MAT_SolveAdd,mat,b,x,y);CHKERRQ(ierr);
2987   if (mat->ops->solveadd)  {
2988     ierr = (*mat->ops->solveadd)(mat,b,y,x);CHKERRQ(ierr);
2989   } else {
2990     /* do the solve then the add manually */
2991     if (x != y) {
2992       ierr = MatSolve(mat,b,x);CHKERRQ(ierr);
2993       ierr = VecAXPY(x,one,y);CHKERRQ(ierr);
2994     } else {
2995       ierr = VecDuplicate(x,&tmp);CHKERRQ(ierr);
2996       ierr = PetscLogObjectParent(mat,tmp);CHKERRQ(ierr);
2997       ierr = VecCopy(x,tmp);CHKERRQ(ierr);
2998       ierr = MatSolve(mat,b,x);CHKERRQ(ierr);
2999       ierr = VecAXPY(x,one,tmp);CHKERRQ(ierr);
3000       ierr = VecDestroy(tmp);CHKERRQ(ierr);
3001     }
3002   }
3003   ierr = PetscLogEventEnd(MAT_SolveAdd,mat,b,x,y);CHKERRQ(ierr);
3004   ierr = PetscObjectStateIncrease((PetscObject)x);CHKERRQ(ierr);
3005   PetscFunctionReturn(0);
3006 }
3007 
3008 #undef __FUNCT__
3009 #define __FUNCT__ "MatSolveTranspose"
3010 /*@
3011    MatSolveTranspose - Solves A' x = b, given a factored matrix.
3012 
3013    Collective on Mat and Vec
3014 
3015    Input Parameters:
3016 +  mat - the factored matrix
3017 -  b - the right-hand-side vector
3018 
3019    Output Parameter:
3020 .  x - the result vector
3021 
3022    Notes:
3023    The vectors b and x cannot be the same.  I.e., one cannot
3024    call MatSolveTranspose(A,x,x).
3025 
3026    Most users should employ the simplified KSP interface for linear solvers
3027    instead of working directly with matrix algebra routines such as this.
3028    See, e.g., KSPCreate().
3029 
3030    Level: developer
3031 
3032    Concepts: matrices^triangular solves
3033 
3034 .seealso: MatSolve(), MatSolveAdd(), MatSolveTransposeAdd()
3035 @*/
3036 PetscErrorCode PETSCMAT_DLLEXPORT MatSolveTranspose(Mat mat,Vec b,Vec x)
3037 {
3038   PetscErrorCode ierr;
3039 
3040   PetscFunctionBegin;
3041   PetscValidHeaderSpecific(mat,MAT_COOKIE,1);
3042   PetscValidType(mat,1);
3043   PetscValidHeaderSpecific(b,VEC_COOKIE,2);
3044   PetscValidHeaderSpecific(x,VEC_COOKIE,3);
3045   PetscCheckSameComm(mat,1,b,2);
3046   PetscCheckSameComm(mat,1,x,3);
3047   if (!mat->factor) SETERRQ(PETSC_ERR_ARG_WRONGSTATE,"Unfactored matrix");
3048   if (x == b) SETERRQ(PETSC_ERR_ARG_IDN,"x and b must be different vectors");
3049   if (!mat->ops->solvetranspose) SETERRQ1(PETSC_ERR_SUP,"Matrix type %s",((PetscObject)mat)->type_name);
3050   if (mat->rmap->N != x->map->N) SETERRQ2(PETSC_ERR_ARG_SIZ,"Mat mat,Vec x: global dim %D %D",mat->rmap->N,x->map->N);
3051   if (mat->cmap->N != b->map->N) SETERRQ2(PETSC_ERR_ARG_SIZ,"Mat mat,Vec b: global dim %D %D",mat->cmap->N,b->map->N);
3052   ierr = MatPreallocated(mat);CHKERRQ(ierr);
3053   ierr = PetscLogEventBegin(MAT_SolveTranspose,mat,b,x,0);CHKERRQ(ierr);
3054   ierr = (*mat->ops->solvetranspose)(mat,b,x);CHKERRQ(ierr);
3055   ierr = PetscLogEventEnd(MAT_SolveTranspose,mat,b,x,0);CHKERRQ(ierr);
3056   ierr = PetscObjectStateIncrease((PetscObject)x);CHKERRQ(ierr);
3057   PetscFunctionReturn(0);
3058 }
3059 
3060 #undef __FUNCT__
3061 #define __FUNCT__ "MatSolveTransposeAdd"
3062 /*@
3063    MatSolveTransposeAdd - Computes x = y + inv(Transpose(A)) b, given a
3064                       factored matrix.
3065 
3066    Collective on Mat and Vec
3067 
3068    Input Parameters:
3069 +  mat - the factored matrix
3070 .  b - the right-hand-side vector
3071 -  y - the vector to be added to
3072 
3073    Output Parameter:
3074 .  x - the result vector
3075 
3076    Notes:
3077    The vectors b and x cannot be the same.  I.e., one cannot
3078    call MatSolveTransposeAdd(A,x,y,x).
3079 
3080    Most users should employ the simplified KSP interface for linear solvers
3081    instead of working directly with matrix algebra routines such as this.
3082    See, e.g., KSPCreate().
3083 
3084    Level: developer
3085 
3086    Concepts: matrices^triangular solves
3087 
3088 .seealso: MatSolve(), MatSolveAdd(), MatSolveTranspose()
3089 @*/
3090 PetscErrorCode PETSCMAT_DLLEXPORT MatSolveTransposeAdd(Mat mat,Vec b,Vec y,Vec x)
3091 {
3092   PetscScalar    one = 1.0;
3093   PetscErrorCode ierr;
3094   Vec            tmp;
3095 
3096   PetscFunctionBegin;
3097   PetscValidHeaderSpecific(mat,MAT_COOKIE,1);
3098   PetscValidType(mat,1);
3099   PetscValidHeaderSpecific(y,VEC_COOKIE,2);
3100   PetscValidHeaderSpecific(b,VEC_COOKIE,3);
3101   PetscValidHeaderSpecific(x,VEC_COOKIE,4);
3102   PetscCheckSameComm(mat,1,b,2);
3103   PetscCheckSameComm(mat,1,y,3);
3104   PetscCheckSameComm(mat,1,x,4);
3105   if (x == b) SETERRQ(PETSC_ERR_ARG_IDN,"x and b must be different vectors");
3106   if (!mat->factor) SETERRQ(PETSC_ERR_ARG_WRONGSTATE,"Unfactored matrix");
3107   if (mat->rmap->N != x->map->N) SETERRQ2(PETSC_ERR_ARG_SIZ,"Mat mat,Vec x: global dim %D %D",mat->rmap->N,x->map->N);
3108   if (mat->cmap->N != b->map->N) SETERRQ2(PETSC_ERR_ARG_SIZ,"Mat mat,Vec b: global dim %D %D",mat->cmap->N,b->map->N);
3109   if (mat->cmap->N != y->map->N) SETERRQ2(PETSC_ERR_ARG_SIZ,"Mat mat,Vec y: global dim %D %D",mat->cmap->N,y->map->N);
3110   if (x->map->n != y->map->n)   SETERRQ2(PETSC_ERR_ARG_SIZ,"Vec x,Vec y: local dim %D %D",x->map->n,y->map->n);
3111   ierr = MatPreallocated(mat);CHKERRQ(ierr);
3112 
3113   ierr = PetscLogEventBegin(MAT_SolveTransposeAdd,mat,b,x,y);CHKERRQ(ierr);
3114   if (mat->ops->solvetransposeadd) {
3115     ierr = (*mat->ops->solvetransposeadd)(mat,b,y,x);CHKERRQ(ierr);
3116   } else {
3117     /* do the solve then the add manually */
3118     if (x != y) {
3119       ierr = MatSolveTranspose(mat,b,x);CHKERRQ(ierr);
3120       ierr = VecAXPY(x,one,y);CHKERRQ(ierr);
3121     } else {
3122       ierr = VecDuplicate(x,&tmp);CHKERRQ(ierr);
3123       ierr = PetscLogObjectParent(mat,tmp);CHKERRQ(ierr);
3124       ierr = VecCopy(x,tmp);CHKERRQ(ierr);
3125       ierr = MatSolveTranspose(mat,b,x);CHKERRQ(ierr);
3126       ierr = VecAXPY(x,one,tmp);CHKERRQ(ierr);
3127       ierr = VecDestroy(tmp);CHKERRQ(ierr);
3128     }
3129   }
3130   ierr = PetscLogEventEnd(MAT_SolveTransposeAdd,mat,b,x,y);CHKERRQ(ierr);
3131   ierr = PetscObjectStateIncrease((PetscObject)x);CHKERRQ(ierr);
3132   PetscFunctionReturn(0);
3133 }
3134 /* ----------------------------------------------------------------*/
3135 
3136 #undef __FUNCT__
3137 #define __FUNCT__ "MatRelax"
3138 /*@
3139    MatRelax - Computes relaxation (SOR, Gauss-Seidel) sweeps.
3140 
3141    Collective on Mat and Vec
3142 
3143    Input Parameters:
3144 +  mat - the matrix
3145 .  b - the right hand side
3146 .  omega - the relaxation factor
3147 .  flag - flag indicating the type of SOR (see below)
3148 .  shift -  diagonal shift
3149 .  its - the number of iterations
3150 -  lits - the number of local iterations
3151 
3152    Output Parameters:
3153 .  x - the solution (can contain an initial guess)
3154 
3155    SOR Flags:
3156 .     SOR_FORWARD_SWEEP - forward SOR
3157 .     SOR_BACKWARD_SWEEP - backward SOR
3158 .     SOR_SYMMETRIC_SWEEP - SSOR (symmetric SOR)
3159 .     SOR_LOCAL_FORWARD_SWEEP - local forward SOR
3160 .     SOR_LOCAL_BACKWARD_SWEEP - local forward SOR
3161 .     SOR_LOCAL_SYMMETRIC_SWEEP - local SSOR
3162 .     SOR_APPLY_UPPER, SOR_APPLY_LOWER - applies
3163          upper/lower triangular part of matrix to
3164          vector (with omega)
3165 .     SOR_ZERO_INITIAL_GUESS - zero initial guess
3166 
3167    Notes:
3168    SOR_LOCAL_FORWARD_SWEEP, SOR_LOCAL_BACKWARD_SWEEP, and
3169    SOR_LOCAL_SYMMETRIC_SWEEP perform separate independent smoothings
3170    on each processor.
3171 
3172    Application programmers will not generally use MatRelax() directly,
3173    but instead will employ the KSP/PC interface.
3174 
3175    Notes for Advanced Users:
3176    The flags are implemented as bitwise inclusive or operations.
3177    For example, use (SOR_ZERO_INITIAL_GUESS | SOR_SYMMETRIC_SWEEP)
3178    to specify a zero initial guess for SSOR.
3179 
3180    Most users should employ the simplified KSP interface for linear solvers
3181    instead of working directly with matrix algebra routines such as this.
3182    See, e.g., KSPCreate().
3183 
3184    See also, MatPBRelax(). This routine will automatically call the point block
3185    version if the point version is not available.
3186 
3187    Level: developer
3188 
3189    Concepts: matrices^relaxation
3190    Concepts: matrices^SOR
3191    Concepts: matrices^Gauss-Seidel
3192 
3193 @*/
3194 PetscErrorCode PETSCMAT_DLLEXPORT MatRelax(Mat mat,Vec b,PetscReal omega,MatSORType flag,PetscReal shift,PetscInt its,PetscInt lits,Vec x)
3195 {
3196   PetscErrorCode ierr;
3197 
3198   PetscFunctionBegin;
3199   PetscValidHeaderSpecific(mat,MAT_COOKIE,1);
3200   PetscValidType(mat,1);
3201   PetscValidHeaderSpecific(b,VEC_COOKIE,2);
3202   PetscValidHeaderSpecific(x,VEC_COOKIE,8);
3203   PetscCheckSameComm(mat,1,b,2);
3204   PetscCheckSameComm(mat,1,x,8);
3205   if (!mat->ops->relax && !mat->ops->pbrelax) SETERRQ1(PETSC_ERR_SUP,"Mat type %s",((PetscObject)mat)->type_name);
3206   if (!mat->assembled) SETERRQ(PETSC_ERR_ARG_WRONGSTATE,"Not for unassembled matrix");
3207   if (mat->factor) SETERRQ(PETSC_ERR_ARG_WRONGSTATE,"Not for factored matrix");
3208   if (mat->cmap->N != x->map->N) SETERRQ2(PETSC_ERR_ARG_SIZ,"Mat mat,Vec x: global dim %D %D",mat->cmap->N,x->map->N);
3209   if (mat->rmap->N != b->map->N) SETERRQ2(PETSC_ERR_ARG_SIZ,"Mat mat,Vec b: global dim %D %D",mat->rmap->N,b->map->N);
3210   if (mat->rmap->n != b->map->n) SETERRQ2(PETSC_ERR_ARG_SIZ,"Mat mat,Vec b: local dim %D %D",mat->rmap->n,b->map->n);
3211   if (its <= 0) SETERRQ1(PETSC_ERR_ARG_WRONG,"Relaxation requires global its %D positive",its);
3212   if (lits <= 0) SETERRQ1(PETSC_ERR_ARG_WRONG,"Relaxation requires local its %D positive",lits);
3213 
3214   ierr = MatPreallocated(mat);CHKERRQ(ierr);
3215   ierr = PetscLogEventBegin(MAT_Relax,mat,b,x,0);CHKERRQ(ierr);
3216   if (mat->ops->relax) {
3217     ierr =(*mat->ops->relax)(mat,b,omega,flag,shift,its,lits,x);CHKERRQ(ierr);
3218   } else {
3219     ierr =(*mat->ops->pbrelax)(mat,b,omega,flag,shift,its,lits,x);CHKERRQ(ierr);
3220   }
3221   ierr = PetscLogEventEnd(MAT_Relax,mat,b,x,0);CHKERRQ(ierr);
3222   ierr = PetscObjectStateIncrease((PetscObject)x);CHKERRQ(ierr);
3223   PetscFunctionReturn(0);
3224 }
3225 
3226 #undef __FUNCT__
3227 #define __FUNCT__ "MatPBRelax"
3228 /*@
3229    MatPBRelax - Computes relaxation (SOR, Gauss-Seidel) sweeps.
3230 
3231    Collective on Mat and Vec
3232 
3233    See MatRelax() for usage. This is called by MatRelax() when appropriate so need not be called by users.
3234 
3235    For multi-component PDEs where the Jacobian is stored in a point block format
3236    (with the PETSc BAIJ matrix formats) the relaxation is done one point block at
3237    a time. That is, the small (for example, 4 by 4) blocks along the diagonal are solved
3238    simultaneously (that is a 4 by 4 linear solve is done) to update all the values at a point.
3239 
3240    Level: developer
3241 
3242 @*/
3243 PetscErrorCode PETSCMAT_DLLEXPORT MatPBRelax(Mat mat,Vec b,PetscReal omega,MatSORType flag,PetscReal shift,PetscInt its,PetscInt lits,Vec x)
3244 {
3245   PetscErrorCode ierr;
3246 
3247   PetscFunctionBegin;
3248   PetscValidHeaderSpecific(mat,MAT_COOKIE,1);
3249   PetscValidType(mat,1);
3250   PetscValidHeaderSpecific(b,VEC_COOKIE,2);
3251   PetscValidHeaderSpecific(x,VEC_COOKIE,8);
3252   PetscCheckSameComm(mat,1,b,2);
3253   PetscCheckSameComm(mat,1,x,8);
3254   if (!mat->ops->pbrelax) SETERRQ1(PETSC_ERR_SUP,"Mat type %s",((PetscObject)mat)->type_name);
3255   if (!mat->assembled) SETERRQ(PETSC_ERR_ARG_WRONGSTATE,"Not for unassembled matrix");
3256   if (mat->factor) SETERRQ(PETSC_ERR_ARG_WRONGSTATE,"Not for factored matrix");
3257   if (mat->cmap->N != x->map->N) SETERRQ2(PETSC_ERR_ARG_SIZ,"Mat mat,Vec x: global dim %D %D",mat->cmap->N,x->map->N);
3258   if (mat->rmap->N != b->map->N) SETERRQ2(PETSC_ERR_ARG_SIZ,"Mat mat,Vec b: global dim %D %D",mat->rmap->N,b->map->N);
3259   if (mat->rmap->n != b->map->n) SETERRQ2(PETSC_ERR_ARG_SIZ,"Mat mat,Vec b: local dim %D %D",mat->rmap->n,b->map->n);
3260   ierr = MatPreallocated(mat);CHKERRQ(ierr);
3261 
3262   ierr = PetscLogEventBegin(MAT_Relax,mat,b,x,0);CHKERRQ(ierr);
3263   ierr =(*mat->ops->pbrelax)(mat,b,omega,flag,shift,its,lits,x);CHKERRQ(ierr);
3264   ierr = PetscLogEventEnd(MAT_Relax,mat,b,x,0);CHKERRQ(ierr);
3265   ierr = PetscObjectStateIncrease((PetscObject)x);CHKERRQ(ierr);
3266   PetscFunctionReturn(0);
3267 }
3268 
3269 #undef __FUNCT__
3270 #define __FUNCT__ "MatCopy_Basic"
3271 /*
3272       Default matrix copy routine.
3273 */
3274 PetscErrorCode MatCopy_Basic(Mat A,Mat B,MatStructure str)
3275 {
3276   PetscErrorCode    ierr;
3277   PetscInt          i,rstart,rend,nz;
3278   const PetscInt    *cwork;
3279   const PetscScalar *vwork;
3280 
3281   PetscFunctionBegin;
3282   if (B->assembled) {
3283     ierr = MatZeroEntries(B);CHKERRQ(ierr);
3284   }
3285   ierr = MatGetOwnershipRange(A,&rstart,&rend);CHKERRQ(ierr);
3286   for (i=rstart; i<rend; i++) {
3287     ierr = MatGetRow(A,i,&nz,&cwork,&vwork);CHKERRQ(ierr);
3288     ierr = MatSetValues(B,1,&i,nz,cwork,vwork,INSERT_VALUES);CHKERRQ(ierr);
3289     ierr = MatRestoreRow(A,i,&nz,&cwork,&vwork);CHKERRQ(ierr);
3290   }
3291   ierr = MatAssemblyBegin(B,MAT_FINAL_ASSEMBLY);CHKERRQ(ierr);
3292   ierr = MatAssemblyEnd(B,MAT_FINAL_ASSEMBLY);CHKERRQ(ierr);
3293   ierr = PetscObjectStateIncrease((PetscObject)B);CHKERRQ(ierr);
3294   PetscFunctionReturn(0);
3295 }
3296 
3297 #undef __FUNCT__
3298 #define __FUNCT__ "MatCopy"
3299 /*@
3300    MatCopy - Copys a matrix to another matrix.
3301 
3302    Collective on Mat
3303 
3304    Input Parameters:
3305 +  A - the matrix
3306 -  str - SAME_NONZERO_PATTERN or DIFFERENT_NONZERO_PATTERN
3307 
3308    Output Parameter:
3309 .  B - where the copy is put
3310 
3311    Notes:
3312    If you use SAME_NONZERO_PATTERN then the two matrices had better have the
3313    same nonzero pattern or the routine will crash.
3314 
3315    MatCopy() copies the matrix entries of a matrix to another existing
3316    matrix (after first zeroing the second matrix).  A related routine is
3317    MatConvert(), which first creates a new matrix and then copies the data.
3318 
3319    Level: intermediate
3320 
3321    Concepts: matrices^copying
3322 
3323 .seealso: MatConvert(), MatDuplicate()
3324 
3325 @*/
3326 PetscErrorCode PETSCMAT_DLLEXPORT MatCopy(Mat A,Mat B,MatStructure str)
3327 {
3328   PetscErrorCode ierr;
3329   PetscInt       i;
3330 
3331   PetscFunctionBegin;
3332   PetscValidHeaderSpecific(A,MAT_COOKIE,1);
3333   PetscValidHeaderSpecific(B,MAT_COOKIE,2);
3334   PetscValidType(A,1);
3335   PetscValidType(B,2);
3336   PetscCheckSameComm(A,1,B,2);
3337   ierr = MatPreallocated(B);CHKERRQ(ierr);
3338   if (!A->assembled) SETERRQ(PETSC_ERR_ARG_WRONGSTATE,"Not for unassembled matrix");
3339   if (A->factor) SETERRQ(PETSC_ERR_ARG_WRONGSTATE,"Not for factored matrix");
3340   if (A->rmap->N != B->rmap->N || A->cmap->N != B->cmap->N) SETERRQ4(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);
3341   ierr = MatPreallocated(A);CHKERRQ(ierr);
3342 
3343   ierr = PetscLogEventBegin(MAT_Copy,A,B,0,0);CHKERRQ(ierr);
3344   if (A->ops->copy) {
3345     ierr = (*A->ops->copy)(A,B,str);CHKERRQ(ierr);
3346   } else { /* generic conversion */
3347     ierr = MatCopy_Basic(A,B,str);CHKERRQ(ierr);
3348   }
3349   if (A->mapping) {
3350     if (B->mapping) {ierr = ISLocalToGlobalMappingDestroy(B->mapping);CHKERRQ(ierr);B->mapping = 0;}
3351     ierr = MatSetLocalToGlobalMapping(B,A->mapping);CHKERRQ(ierr);
3352   }
3353   if (A->bmapping) {
3354     if (B->bmapping) {ierr = ISLocalToGlobalMappingDestroy(B->bmapping);CHKERRQ(ierr);B->bmapping = 0;}
3355     ierr = MatSetLocalToGlobalMappingBlock(B,A->mapping);CHKERRQ(ierr);
3356   }
3357 
3358   B->stencil.dim = A->stencil.dim;
3359   B->stencil.noc = A->stencil.noc;
3360   for (i=0; i<=A->stencil.dim; i++) {
3361     B->stencil.dims[i]   = A->stencil.dims[i];
3362     B->stencil.starts[i] = A->stencil.starts[i];
3363   }
3364 
3365   ierr = PetscLogEventEnd(MAT_Copy,A,B,0,0);CHKERRQ(ierr);
3366   ierr = PetscObjectStateIncrease((PetscObject)B);CHKERRQ(ierr);
3367   PetscFunctionReturn(0);
3368 }
3369 
3370 #undef __FUNCT__
3371 #define __FUNCT__ "MatConvert"
3372 /*@C
3373    MatConvert - Converts a matrix to another matrix, either of the same
3374    or different type.
3375 
3376    Collective on Mat
3377 
3378    Input Parameters:
3379 +  mat - the matrix
3380 .  newtype - new matrix type.  Use MATSAME to create a new matrix of the
3381    same type as the original matrix.
3382 -  reuse - denotes if the destination matrix is to be created or reused.  Currently
3383    MAT_REUSE_MATRIX is only supported for inplace conversion, otherwise use
3384    MAT_INITIAL_MATRIX.
3385 
3386    Output Parameter:
3387 .  M - pointer to place new matrix
3388 
3389    Notes:
3390    MatConvert() first creates a new matrix and then copies the data from
3391    the first matrix.  A related routine is MatCopy(), which copies the matrix
3392    entries of one matrix to another already existing matrix context.
3393 
3394    Cannot be used to convert a sequential matrix to parallel or parallel to sequential,
3395    the MPI communicator of the generated matrix is always the same as the communicator
3396    of the input matrix.
3397 
3398    Level: intermediate
3399 
3400    Concepts: matrices^converting between storage formats
3401 
3402 .seealso: MatCopy(), MatDuplicate()
3403 @*/
3404 PetscErrorCode PETSCMAT_DLLEXPORT MatConvert(Mat mat, const MatType newtype,MatReuse reuse,Mat *M)
3405 {
3406   PetscErrorCode         ierr;
3407   PetscTruth             sametype,issame,flg;
3408   char                   convname[256],mtype[256];
3409   Mat                    B;
3410 
3411   PetscFunctionBegin;
3412   PetscValidHeaderSpecific(mat,MAT_COOKIE,1);
3413   PetscValidType(mat,1);
3414   PetscValidPointer(M,3);
3415   if (!mat->assembled) SETERRQ(PETSC_ERR_ARG_WRONGSTATE,"Not for unassembled matrix");
3416   if (mat->factor) SETERRQ(PETSC_ERR_ARG_WRONGSTATE,"Not for factored matrix");
3417   ierr = MatPreallocated(mat);CHKERRQ(ierr);
3418 
3419   ierr = PetscOptionsGetString(((PetscObject)mat)->prefix,"-matconvert_type",mtype,256,&flg);CHKERRQ(ierr);
3420   if (flg) {
3421     newtype = mtype;
3422   }
3423   ierr = PetscTypeCompare((PetscObject)mat,newtype,&sametype);CHKERRQ(ierr);
3424   ierr = PetscStrcmp(newtype,"same",&issame);CHKERRQ(ierr);
3425   if ((reuse == MAT_REUSE_MATRIX) && (mat != *M)) {
3426     SETERRQ(PETSC_ERR_SUP,"MAT_REUSE_MATRIX only supported for in-place conversion currently");
3427   }
3428 
3429   if ((reuse == MAT_REUSE_MATRIX) && (issame || sametype)) PetscFunctionReturn(0);
3430 
3431   if ((sametype || issame) && (reuse==MAT_INITIAL_MATRIX) && mat->ops->duplicate) {
3432     ierr = (*mat->ops->duplicate)(mat,MAT_COPY_VALUES,M);CHKERRQ(ierr);
3433   } else {
3434     PetscErrorCode (*conv)(Mat, const MatType,MatReuse,Mat*)=PETSC_NULL;
3435     const char     *prefix[3] = {"seq","mpi",""};
3436     PetscInt       i;
3437     /*
3438        Order of precedence:
3439        1) See if a specialized converter is known to the current matrix.
3440        2) See if a specialized converter is known to the desired matrix class.
3441        3) See if a good general converter is registered for the desired class
3442           (as of 6/27/03 only MATMPIADJ falls into this category).
3443        4) See if a good general converter is known for the current matrix.
3444        5) Use a really basic converter.
3445     */
3446 
3447     /* 1) See if a specialized converter is known to the current matrix and the desired class */
3448     for (i=0; i<3; i++) {
3449       ierr = PetscStrcpy(convname,"MatConvert_");CHKERRQ(ierr);
3450       ierr = PetscStrcat(convname,((PetscObject)mat)->type_name);CHKERRQ(ierr);
3451       ierr = PetscStrcat(convname,"_");CHKERRQ(ierr);
3452       ierr = PetscStrcat(convname,prefix[i]);CHKERRQ(ierr);
3453       ierr = PetscStrcat(convname,newtype);CHKERRQ(ierr);
3454       ierr = PetscStrcat(convname,"_C");CHKERRQ(ierr);
3455       ierr = PetscObjectQueryFunction((PetscObject)mat,convname,(void (**)(void))&conv);CHKERRQ(ierr);
3456       if (conv) goto foundconv;
3457     }
3458 
3459     /* 2)  See if a specialized converter is known to the desired matrix class. */
3460     ierr = MatCreate(((PetscObject)mat)->comm,&B);CHKERRQ(ierr);
3461     ierr = MatSetSizes(B,mat->rmap->n,mat->cmap->n,mat->rmap->N,mat->cmap->N);CHKERRQ(ierr);
3462     ierr = MatSetType(B,newtype);CHKERRQ(ierr);
3463     for (i=0; i<3; i++) {
3464       ierr = PetscStrcpy(convname,"MatConvert_");CHKERRQ(ierr);
3465       ierr = PetscStrcat(convname,((PetscObject)mat)->type_name);CHKERRQ(ierr);
3466       ierr = PetscStrcat(convname,"_");CHKERRQ(ierr);
3467       ierr = PetscStrcat(convname,prefix[i]);CHKERRQ(ierr);
3468       ierr = PetscStrcat(convname,newtype);CHKERRQ(ierr);
3469       ierr = PetscStrcat(convname,"_C");CHKERRQ(ierr);
3470       ierr = PetscObjectQueryFunction((PetscObject)B,convname,(void (**)(void))&conv);CHKERRQ(ierr);
3471       if (conv) {
3472         ierr = MatDestroy(B);CHKERRQ(ierr);
3473         goto foundconv;
3474       }
3475     }
3476 
3477     /* 3) See if a good general converter is registered for the desired class */
3478     conv = B->ops->convertfrom;
3479     ierr = MatDestroy(B);CHKERRQ(ierr);
3480     if (conv) goto foundconv;
3481 
3482     /* 4) See if a good general converter is known for the current matrix */
3483     if (mat->ops->convert) {
3484       conv = mat->ops->convert;
3485     }
3486     if (conv) goto foundconv;
3487 
3488     /* 5) Use a really basic converter. */
3489     conv = MatConvert_Basic;
3490 
3491     foundconv:
3492     ierr = PetscLogEventBegin(MAT_Convert,mat,0,0,0);CHKERRQ(ierr);
3493     ierr = (*conv)(mat,newtype,reuse,M);CHKERRQ(ierr);
3494     ierr = PetscLogEventEnd(MAT_Convert,mat,0,0,0);CHKERRQ(ierr);
3495   }
3496   ierr = PetscObjectStateIncrease((PetscObject)*M);CHKERRQ(ierr);
3497   PetscFunctionReturn(0);
3498 }
3499 
3500 #undef __FUNCT__
3501 #define __FUNCT__ "MatFactorGetSolverPackage"
3502 /*@C
3503    MatFactorGetSolverPackage - Returns name of the package providing the factorization routines
3504 
3505    Not Collective
3506 
3507    Input Parameter:
3508 .  mat - the matrix, must be a factored matrix
3509 
3510    Output Parameter:
3511 .   type - the string name of the package (do not free this string)
3512 
3513    Notes:
3514       In Fortran you pass in a empty string and the package name will be copied into it.
3515     (Make sure the string is long enough)
3516 
3517    Level: intermediate
3518 
3519 .seealso: MatCopy(), MatDuplicate(), MatGetFactorAvailable(), MatGetFactor()
3520 @*/
3521 PetscErrorCode PETSCMAT_DLLEXPORT MatFactorGetSolverPackage(Mat mat, const MatSolverPackage *type)
3522 {
3523   PetscErrorCode         ierr;
3524   PetscErrorCode         (*conv)(Mat,const MatSolverPackage*);
3525 
3526   PetscFunctionBegin;
3527   PetscValidHeaderSpecific(mat,MAT_COOKIE,1);
3528   PetscValidType(mat,1);
3529   if (!mat->factor) SETERRQ(PETSC_ERR_ARG_WRONGSTATE,"Only for factored matrix");
3530   ierr = PetscObjectQueryFunction((PetscObject)mat,"MatFactorGetSolverPackage_C",(void (**)(void))&conv);CHKERRQ(ierr);
3531   if (!conv) {
3532     *type = MAT_SOLVER_PETSC;
3533   } else {
3534     ierr = (*conv)(mat,type);CHKERRQ(ierr);
3535   }
3536   PetscFunctionReturn(0);
3537 }
3538 
3539 #undef __FUNCT__
3540 #define __FUNCT__ "MatGetFactor"
3541 /*@C
3542    MatGetFactor - Returns a matrix suitable to calls to MatXXFactorSymbolic()
3543 
3544    Collective on Mat
3545 
3546    Input Parameters:
3547 +  mat - the matrix
3548 .  type - name of solver type, for example, spooles, superlu, plapack, petsc (to use PETSc's default)
3549 -  ftype - factor type, MAT_FACTOR_LU, MAT_FACTOR_CHOLESKY, MAT_FACTOR_ICC, MAT_FACTOR_ILU,
3550 
3551    Output Parameters:
3552 .  f - the factor matrix used with MatXXFactorSymbolic() calls
3553 
3554    Notes:
3555       Some PETSc matrix formats have alternative solvers available that are contained in alternative packages
3556      such as pastix, superlu, mumps, spooles etc.
3557 
3558       PETSc must have been config/configure.py to use the external solver, using the option --download-package
3559 
3560    Level: intermediate
3561 
3562 .seealso: MatCopy(), MatDuplicate(), MatGetFactorAvailable()
3563 @*/
3564 PetscErrorCode PETSCMAT_DLLEXPORT MatGetFactor(Mat mat, const MatSolverPackage type,MatFactorType ftype,Mat *f)
3565 {
3566   PetscErrorCode         ierr;
3567   char                   convname[256];
3568   PetscErrorCode         (*conv)(Mat,MatFactorType,Mat*);
3569 
3570   PetscFunctionBegin;
3571   PetscValidHeaderSpecific(mat,MAT_COOKIE,1);
3572   PetscValidType(mat,1);
3573 
3574   if (mat->factor) SETERRQ(PETSC_ERR_ARG_WRONGSTATE,"Not for factored matrix");
3575   ierr = MatPreallocated(mat);CHKERRQ(ierr);
3576 
3577   ierr = PetscStrcpy(convname,"MatGetFactor_");CHKERRQ(ierr);
3578   ierr = PetscStrcat(convname,type);CHKERRQ(ierr);
3579   ierr = PetscStrcat(convname,"_C");CHKERRQ(ierr);
3580   ierr = PetscObjectQueryFunction((PetscObject)mat,convname,(void (**)(void))&conv);CHKERRQ(ierr);
3581   if (!conv) {
3582     PetscTruth flag;
3583     ierr = PetscStrcasecmp(MAT_SOLVER_PETSC,type,&flag);CHKERRQ(ierr);
3584     if (flag) {
3585       SETERRQ1(PETSC_ERR_SUP,"Matrix format %s does not have a built-in PETSc direct solver",((PetscObject)mat)->type_name);
3586     } else {
3587       SETERRQ3(PETSC_ERR_SUP,"Matrix format %s does not have a solver %s. Perhaps you must config/configure.py with --download-%s",((PetscObject)mat)->type_name,type,type);
3588     }
3589   }
3590   ierr = (*conv)(mat,ftype,f);CHKERRQ(ierr);
3591   PetscFunctionReturn(0);
3592 }
3593 
3594 #undef __FUNCT__
3595 #define __FUNCT__ "MatGetFactorAvailable"
3596 /*@C
3597    MatGetFactorAvailable - Returns a a flag if matrix supports particular package and factor type
3598 
3599    Collective on Mat
3600 
3601    Input Parameters:
3602 +  mat - the matrix
3603 .  type - name of solver type, for example, spooles, superlu, plapack, petsc (to use PETSc's default)
3604 -  ftype - factor type, MAT_FACTOR_LU, MAT_FACTOR_CHOLESKY, MAT_FACTOR_ICC, MAT_FACTOR_ILU,
3605 
3606    Output Parameter:
3607 .    flg - PETSC_TRUE if the factorization is available
3608 
3609    Notes:
3610       Some PETSc matrix formats have alternative solvers available that are contained in alternative packages
3611      such as pastix, superlu, mumps, spooles etc.
3612 
3613       PETSc must have been config/configure.py to use the external solver, using the option --download-package
3614 
3615    Level: intermediate
3616 
3617 .seealso: MatCopy(), MatDuplicate(), MatGetFactor()
3618 @*/
3619 PetscErrorCode PETSCMAT_DLLEXPORT MatGetFactorAvailable(Mat mat, const MatSolverPackage type,MatFactorType ftype,PetscTruth *flg)
3620 {
3621   PetscErrorCode         ierr;
3622   char                   convname[256];
3623   PetscErrorCode         (*conv)(Mat,MatFactorType,PetscTruth*);
3624 
3625   PetscFunctionBegin;
3626   PetscValidHeaderSpecific(mat,MAT_COOKIE,1);
3627   PetscValidType(mat,1);
3628 
3629   if (mat->factor) SETERRQ(PETSC_ERR_ARG_WRONGSTATE,"Not for factored matrix");
3630   ierr = MatPreallocated(mat);CHKERRQ(ierr);
3631 
3632   ierr = PetscStrcpy(convname,"MatGetFactorAvailable_");CHKERRQ(ierr);
3633   ierr = PetscStrcat(convname,type);CHKERRQ(ierr);
3634   ierr = PetscStrcat(convname,"_C");CHKERRQ(ierr);
3635   ierr = PetscObjectQueryFunction((PetscObject)mat,convname,(void (**)(void))&conv);CHKERRQ(ierr);
3636   if (!conv) {
3637     *flg = PETSC_FALSE;
3638   } else {
3639     ierr = (*conv)(mat,ftype,flg);CHKERRQ(ierr);
3640   }
3641   PetscFunctionReturn(0);
3642 }
3643 
3644 
3645 #undef __FUNCT__
3646 #define __FUNCT__ "MatDuplicate"
3647 /*@
3648    MatDuplicate - Duplicates a matrix including the non-zero structure.
3649 
3650    Collective on Mat
3651 
3652    Input Parameters:
3653 +  mat - the matrix
3654 -  op - either MAT_DO_NOT_COPY_VALUES or MAT_COPY_VALUES, cause it to copy the numerical values in the matrix
3655         MAT_SHARE_NONZERO_PATTERN to share the nonzero patterns with the previous matrix and not copy them.
3656 
3657    Output Parameter:
3658 .  M - pointer to place new matrix
3659 
3660    Level: intermediate
3661 
3662    Concepts: matrices^duplicating
3663 
3664     Notes: You cannot change the nonzero pattern for the parent or child matrix if you use MAT_SHARE_NONZERO_PATTERN.
3665 
3666 .seealso: MatCopy(), MatConvert()
3667 @*/
3668 PetscErrorCode PETSCMAT_DLLEXPORT MatDuplicate(Mat mat,MatDuplicateOption op,Mat *M)
3669 {
3670   PetscErrorCode ierr;
3671   Mat            B;
3672   PetscInt       i;
3673 
3674   PetscFunctionBegin;
3675   PetscValidHeaderSpecific(mat,MAT_COOKIE,1);
3676   PetscValidType(mat,1);
3677   PetscValidPointer(M,3);
3678   if (!mat->assembled) SETERRQ(PETSC_ERR_ARG_WRONGSTATE,"Not for unassembled matrix");
3679   if (mat->factor) SETERRQ(PETSC_ERR_ARG_WRONGSTATE,"Not for factored matrix");
3680   ierr = MatPreallocated(mat);CHKERRQ(ierr);
3681 
3682   *M  = 0;
3683   if (!mat->ops->duplicate) {
3684     SETERRQ(PETSC_ERR_SUP,"Not written for this matrix type");
3685   }
3686   ierr = PetscLogEventBegin(MAT_Convert,mat,0,0,0);CHKERRQ(ierr);
3687   ierr = (*mat->ops->duplicate)(mat,op,M);CHKERRQ(ierr);
3688   B = *M;
3689   if (mat->mapping) {
3690     ierr = MatSetLocalToGlobalMapping(B,mat->mapping);CHKERRQ(ierr);
3691   }
3692   if (mat->bmapping) {
3693     ierr = MatSetLocalToGlobalMappingBlock(B,mat->bmapping);CHKERRQ(ierr);
3694   }
3695   ierr = PetscMapCopy(((PetscObject)mat)->comm,mat->rmap,B->rmap);CHKERRQ(ierr);
3696   ierr = PetscMapCopy(((PetscObject)mat)->comm,mat->cmap,B->cmap);CHKERRQ(ierr);
3697 
3698   B->stencil.dim = mat->stencil.dim;
3699   B->stencil.noc = mat->stencil.noc;
3700   for (i=0; i<=mat->stencil.dim; i++) {
3701     B->stencil.dims[i]   = mat->stencil.dims[i];
3702     B->stencil.starts[i] = mat->stencil.starts[i];
3703   }
3704 
3705   ierr = PetscLogEventEnd(MAT_Convert,mat,0,0,0);CHKERRQ(ierr);
3706   ierr = PetscObjectStateIncrease((PetscObject)B);CHKERRQ(ierr);
3707   PetscFunctionReturn(0);
3708 }
3709 
3710 #undef __FUNCT__
3711 #define __FUNCT__ "MatGetDiagonal"
3712 /*@
3713    MatGetDiagonal - Gets the diagonal of a matrix.
3714 
3715    Collective on Mat and Vec
3716 
3717    Input Parameters:
3718 +  mat - the matrix
3719 -  v - the vector for storing the diagonal
3720 
3721    Output Parameter:
3722 .  v - the diagonal of the matrix
3723 
3724    Level: intermediate
3725 
3726    Concepts: matrices^accessing diagonals
3727 
3728 .seealso: MatGetRow(), MatGetSubMatrices(), MatGetSubmatrix(), MatGetRowMaxAbs()
3729 @*/
3730 PetscErrorCode PETSCMAT_DLLEXPORT MatGetDiagonal(Mat mat,Vec v)
3731 {
3732   PetscErrorCode ierr;
3733 
3734   PetscFunctionBegin;
3735   PetscValidHeaderSpecific(mat,MAT_COOKIE,1);
3736   PetscValidType(mat,1);
3737   PetscValidHeaderSpecific(v,VEC_COOKIE,2);
3738   if (!mat->assembled) SETERRQ(PETSC_ERR_ARG_WRONGSTATE,"Not for unassembled matrix");
3739   if (!mat->ops->getdiagonal) SETERRQ1(PETSC_ERR_SUP,"Mat type %s",((PetscObject)mat)->type_name);
3740   ierr = MatPreallocated(mat);CHKERRQ(ierr);
3741 
3742   ierr = (*mat->ops->getdiagonal)(mat,v);CHKERRQ(ierr);
3743   ierr = PetscObjectStateIncrease((PetscObject)v);CHKERRQ(ierr);
3744   PetscFunctionReturn(0);
3745 }
3746 
3747 #undef __FUNCT__
3748 #define __FUNCT__ "MatGetRowMin"
3749 /*@
3750    MatGetRowMin - Gets the minimum value (of the real part) of each
3751         row of the matrix
3752 
3753    Collective on Mat and Vec
3754 
3755    Input Parameters:
3756 .  mat - the matrix
3757 
3758    Output Parameter:
3759 +  v - the vector for storing the maximums
3760 -  idx - the indices of the column found for each row (optional)
3761 
3762    Level: intermediate
3763 
3764    Notes: The result of this call are the same as if one converted the matrix to dense format
3765       and found the minimum value in each row (i.e. the implicit zeros are counted as zeros).
3766 
3767     This code is only implemented for a couple of matrix formats.
3768 
3769    Concepts: matrices^getting row maximums
3770 
3771 .seealso: MatGetDiagonal(), MatGetSubMatrices(), MatGetSubmatrix(), MatGetRowMaxAbs(),
3772           MatGetRowMax()
3773 @*/
3774 PetscErrorCode PETSCMAT_DLLEXPORT MatGetRowMin(Mat mat,Vec v,PetscInt idx[])
3775 {
3776   PetscErrorCode ierr;
3777 
3778   PetscFunctionBegin;
3779   PetscValidHeaderSpecific(mat,MAT_COOKIE,1);
3780   PetscValidType(mat,1);
3781   PetscValidHeaderSpecific(v,VEC_COOKIE,2);
3782   if (!mat->assembled) SETERRQ(PETSC_ERR_ARG_WRONGSTATE,"Not for unassembled matrix");
3783   if (!mat->ops->getrowmax) SETERRQ1(PETSC_ERR_SUP,"Mat type %s",((PetscObject)mat)->type_name);
3784   ierr = MatPreallocated(mat);CHKERRQ(ierr);
3785 
3786   ierr = (*mat->ops->getrowmin)(mat,v,idx);CHKERRQ(ierr);
3787   ierr = PetscObjectStateIncrease((PetscObject)v);CHKERRQ(ierr);
3788   PetscFunctionReturn(0);
3789 }
3790 
3791 #undef __FUNCT__
3792 #define __FUNCT__ "MatGetRowMinAbs"
3793 /*@
3794    MatGetRowMinAbs - Gets the minimum value (in absolute value) of each
3795         row of the matrix
3796 
3797    Collective on Mat and Vec
3798 
3799    Input Parameters:
3800 .  mat - the matrix
3801 
3802    Output Parameter:
3803 +  v - the vector for storing the minimums
3804 -  idx - the indices of the column found for each row (optional)
3805 
3806    Level: intermediate
3807 
3808    Notes: if a row is completely empty or has only 0.0 values then the idx[] value for that
3809     row is 0 (the first column).
3810 
3811     This code is only implemented for a couple of matrix formats.
3812 
3813    Concepts: matrices^getting row maximums
3814 
3815 .seealso: MatGetDiagonal(), MatGetSubMatrices(), MatGetSubmatrix(), MatGetRowMax(), MatGetRowMaxAbs(), MatGetRowMin()
3816 @*/
3817 PetscErrorCode PETSCMAT_DLLEXPORT MatGetRowMinAbs(Mat mat,Vec v,PetscInt idx[])
3818 {
3819   PetscErrorCode ierr;
3820 
3821   PetscFunctionBegin;
3822   PetscValidHeaderSpecific(mat,MAT_COOKIE,1);
3823   PetscValidType(mat,1);
3824   PetscValidHeaderSpecific(v,VEC_COOKIE,2);
3825   if (!mat->assembled) SETERRQ(PETSC_ERR_ARG_WRONGSTATE,"Not for unassembled matrix");
3826   if (!mat->ops->getrowminabs) SETERRQ1(PETSC_ERR_SUP,"Mat type %s",((PetscObject)mat)->type_name);
3827   ierr = MatPreallocated(mat);CHKERRQ(ierr);
3828   if (idx) {ierr = PetscMemzero(idx,mat->rmap->n*sizeof(PetscInt));CHKERRQ(ierr);}
3829 
3830   ierr = (*mat->ops->getrowminabs)(mat,v,idx);CHKERRQ(ierr);
3831   ierr = PetscObjectStateIncrease((PetscObject)v);CHKERRQ(ierr);
3832   PetscFunctionReturn(0);
3833 }
3834 
3835 #undef __FUNCT__
3836 #define __FUNCT__ "MatGetRowMax"
3837 /*@
3838    MatGetRowMax - Gets the maximum value (of the real part) of each
3839         row of the matrix
3840 
3841    Collective on Mat and Vec
3842 
3843    Input Parameters:
3844 .  mat - the matrix
3845 
3846    Output Parameter:
3847 +  v - the vector for storing the maximums
3848 -  idx - the indices of the column found for each row (optional)
3849 
3850    Level: intermediate
3851 
3852    Notes: The result of this call are the same as if one converted the matrix to dense format
3853       and found the minimum value in each row (i.e. the implicit zeros are counted as zeros).
3854 
3855     This code is only implemented for a couple of matrix formats.
3856 
3857    Concepts: matrices^getting row maximums
3858 
3859 .seealso: MatGetDiagonal(), MatGetSubMatrices(), MatGetSubmatrix(), MatGetRowMaxAbs(), MatGetRowMin()
3860 @*/
3861 PetscErrorCode PETSCMAT_DLLEXPORT MatGetRowMax(Mat mat,Vec v,PetscInt idx[])
3862 {
3863   PetscErrorCode ierr;
3864 
3865   PetscFunctionBegin;
3866   PetscValidHeaderSpecific(mat,MAT_COOKIE,1);
3867   PetscValidType(mat,1);
3868   PetscValidHeaderSpecific(v,VEC_COOKIE,2);
3869   if (!mat->assembled) SETERRQ(PETSC_ERR_ARG_WRONGSTATE,"Not for unassembled matrix");
3870   if (!mat->ops->getrowmax) SETERRQ1(PETSC_ERR_SUP,"Mat type %s",((PetscObject)mat)->type_name);
3871   ierr = MatPreallocated(mat);CHKERRQ(ierr);
3872 
3873   ierr = (*mat->ops->getrowmax)(mat,v,idx);CHKERRQ(ierr);
3874   ierr = PetscObjectStateIncrease((PetscObject)v);CHKERRQ(ierr);
3875   PetscFunctionReturn(0);
3876 }
3877 
3878 #undef __FUNCT__
3879 #define __FUNCT__ "MatGetRowMaxAbs"
3880 /*@
3881    MatGetRowMaxAbs - Gets the maximum value (in absolute value) of each
3882         row of the matrix
3883 
3884    Collective on Mat and Vec
3885 
3886    Input Parameters:
3887 .  mat - the matrix
3888 
3889    Output Parameter:
3890 +  v - the vector for storing the maximums
3891 -  idx - the indices of the column found for each row (optional)
3892 
3893    Level: intermediate
3894 
3895    Notes: if a row is completely empty or has only 0.0 values then the idx[] value for that
3896     row is 0 (the first column).
3897 
3898     This code is only implemented for a couple of matrix formats.
3899 
3900    Concepts: matrices^getting row maximums
3901 
3902 .seealso: MatGetDiagonal(), MatGetSubMatrices(), MatGetSubmatrix(), MatGetRowMax(), MatGetRowMin()
3903 @*/
3904 PetscErrorCode PETSCMAT_DLLEXPORT MatGetRowMaxAbs(Mat mat,Vec v,PetscInt idx[])
3905 {
3906   PetscErrorCode ierr;
3907 
3908   PetscFunctionBegin;
3909   PetscValidHeaderSpecific(mat,MAT_COOKIE,1);
3910   PetscValidType(mat,1);
3911   PetscValidHeaderSpecific(v,VEC_COOKIE,2);
3912   if (!mat->assembled) SETERRQ(PETSC_ERR_ARG_WRONGSTATE,"Not for unassembled matrix");
3913   if (!mat->ops->getrowmaxabs) SETERRQ1(PETSC_ERR_SUP,"Mat type %s",((PetscObject)mat)->type_name);
3914   ierr = MatPreallocated(mat);CHKERRQ(ierr);
3915   if (idx) {ierr = PetscMemzero(idx,mat->rmap->n*sizeof(PetscInt));CHKERRQ(ierr);}
3916 
3917   ierr = (*mat->ops->getrowmaxabs)(mat,v,idx);CHKERRQ(ierr);
3918   ierr = PetscObjectStateIncrease((PetscObject)v);CHKERRQ(ierr);
3919   PetscFunctionReturn(0);
3920 }
3921 
3922 #undef __FUNCT__
3923 #define __FUNCT__ "MatGetRowSum"
3924 /*@
3925    MatGetRowSum - Gets the sum of each row of the matrix
3926 
3927    Collective on Mat and Vec
3928 
3929    Input Parameters:
3930 .  mat - the matrix
3931 
3932    Output Parameter:
3933 .  v - the vector for storing the maximums
3934 
3935    Level: intermediate
3936 
3937    Notes: This code is slow since it is not currently specialized for different formats
3938 
3939    Concepts: matrices^getting row sums
3940 
3941 .seealso: MatGetDiagonal(), MatGetSubMatrices(), MatGetSubmatrix(), MatGetRowMax(), MatGetRowMin()
3942 @*/
3943 PetscErrorCode PETSCMAT_DLLEXPORT MatGetRowSum(Mat mat, Vec v)
3944 {
3945   PetscInt       start, end, row;
3946   PetscScalar   *array;
3947   PetscErrorCode ierr;
3948 
3949   PetscFunctionBegin;
3950   PetscValidHeaderSpecific(mat,MAT_COOKIE,1);
3951   PetscValidType(mat,1);
3952   PetscValidHeaderSpecific(v,VEC_COOKIE,2);
3953   if (!mat->assembled) SETERRQ(PETSC_ERR_ARG_WRONGSTATE,"Not for unassembled matrix");
3954   ierr = MatPreallocated(mat);CHKERRQ(ierr);
3955   ierr = MatGetOwnershipRange(mat, &start, &end);CHKERRQ(ierr);
3956   ierr = VecGetArray(v, &array);CHKERRQ(ierr);
3957   for(row = start; row < end; ++row) {
3958     PetscInt           ncols, col;
3959     const PetscInt    *cols;
3960     const PetscScalar *vals;
3961 
3962     array[row - start] = 0.0;
3963     ierr = MatGetRow(mat, row, &ncols, &cols, &vals);CHKERRQ(ierr);
3964     for(col = 0; col < ncols; col++) {
3965       array[row - start] += vals[col];
3966     }
3967     ierr = MatRestoreRow(mat, row, &ncols, &cols, &vals);CHKERRQ(ierr);
3968   }
3969   ierr = VecRestoreArray(v, &array);CHKERRQ(ierr);
3970   ierr = PetscObjectStateIncrease((PetscObject) v);CHKERRQ(ierr);
3971   PetscFunctionReturn(0);
3972 }
3973 
3974 #undef __FUNCT__
3975 #define __FUNCT__ "MatTranspose"
3976 /*@
3977    MatTranspose - Computes an in-place or out-of-place transpose of a matrix.
3978 
3979    Collective on Mat
3980 
3981    Input Parameter:
3982 +  mat - the matrix to transpose
3983 -  reuse - store the transpose matrix in the provided B
3984 
3985    Output Parameters:
3986 .  B - the transpose
3987 
3988    Notes:
3989      If you  pass in &mat for B the transpose will be done in place
3990 
3991    Level: intermediate
3992 
3993    Concepts: matrices^transposing
3994 
3995 .seealso: MatMultTranspose(), MatMultTransposeAdd(), MatIsTranspose(), MatReuse
3996 @*/
3997 PetscErrorCode PETSCMAT_DLLEXPORT MatTranspose(Mat mat,MatReuse reuse,Mat *B)
3998 {
3999   PetscErrorCode ierr;
4000 
4001   PetscFunctionBegin;
4002   PetscValidHeaderSpecific(mat,MAT_COOKIE,1);
4003   PetscValidType(mat,1);
4004   if (!mat->assembled) SETERRQ(PETSC_ERR_ARG_WRONGSTATE,"Not for unassembled matrix");
4005   if (mat->factor) SETERRQ(PETSC_ERR_ARG_WRONGSTATE,"Not for factored matrix");
4006   if (!mat->ops->transpose) SETERRQ1(PETSC_ERR_SUP,"Mat type %s",((PetscObject)mat)->type_name);
4007   ierr = MatPreallocated(mat);CHKERRQ(ierr);
4008 
4009   ierr = PetscLogEventBegin(MAT_Transpose,mat,0,0,0);CHKERRQ(ierr);
4010   ierr = (*mat->ops->transpose)(mat,reuse,B);CHKERRQ(ierr);
4011   ierr = PetscLogEventEnd(MAT_Transpose,mat,0,0,0);CHKERRQ(ierr);
4012   if (B) {ierr = PetscObjectStateIncrease((PetscObject)*B);CHKERRQ(ierr);}
4013   PetscFunctionReturn(0);
4014 }
4015 
4016 #undef __FUNCT__
4017 #define __FUNCT__ "MatIsTranspose"
4018 /*@
4019    MatIsTranspose - Test whether a matrix is another one's transpose,
4020         or its own, in which case it tests symmetry.
4021 
4022    Collective on Mat
4023 
4024    Input Parameter:
4025 +  A - the matrix to test
4026 -  B - the matrix to test against, this can equal the first parameter
4027 
4028    Output Parameters:
4029 .  flg - the result
4030 
4031    Notes:
4032    Only available for SeqAIJ/MPIAIJ matrices. The sequential algorithm
4033    has a running time of the order of the number of nonzeros; the parallel
4034    test involves parallel copies of the block-offdiagonal parts of the matrix.
4035 
4036    Level: intermediate
4037 
4038    Concepts: matrices^transposing, matrix^symmetry
4039 
4040 .seealso: MatTranspose(), MatIsSymmetric(), MatIsHermitian()
4041 @*/
4042 PetscErrorCode PETSCMAT_DLLEXPORT MatIsTranspose(Mat A,Mat B,PetscReal tol,PetscTruth *flg)
4043 {
4044   PetscErrorCode ierr,(*f)(Mat,Mat,PetscReal,PetscTruth*),(*g)(Mat,Mat,PetscReal,PetscTruth*);
4045 
4046   PetscFunctionBegin;
4047   PetscValidHeaderSpecific(A,MAT_COOKIE,1);
4048   PetscValidHeaderSpecific(B,MAT_COOKIE,2);
4049   PetscValidPointer(flg,3);
4050   ierr = PetscObjectQueryFunction((PetscObject)A,"MatIsTranspose_C",(void (**)(void))&f);CHKERRQ(ierr);
4051   ierr = PetscObjectQueryFunction((PetscObject)B,"MatIsTranspose_C",(void (**)(void))&g);CHKERRQ(ierr);
4052   if (f && g) {
4053     if (f==g) {
4054       ierr = (*f)(A,B,tol,flg);CHKERRQ(ierr);
4055     } else {
4056       SETERRQ(PETSC_ERR_ARG_NOTSAMETYPE,"Matrices do not have the same comparator for symmetry test");
4057     }
4058   }
4059   PetscFunctionReturn(0);
4060 }
4061 
4062 #undef __FUNCT__
4063 #define __FUNCT__ "MatIsHermitianTranspose"
4064 /*@
4065    MatIsHermitianTranspose - Test whether a matrix is another one's Hermitian transpose,
4066 
4067    Collective on Mat
4068 
4069    Input Parameter:
4070 +  A - the matrix to test
4071 -  B - the matrix to test against, this can equal the first parameter
4072 
4073    Output Parameters:
4074 .  flg - the result
4075 
4076    Notes:
4077    Only available for SeqAIJ/MPIAIJ matrices. The sequential algorithm
4078    has a running time of the order of the number of nonzeros; the parallel
4079    test involves parallel copies of the block-offdiagonal parts of the matrix.
4080 
4081    Level: intermediate
4082 
4083    Concepts: matrices^transposing, matrix^symmetry
4084 
4085 .seealso: MatTranspose(), MatIsSymmetric(), MatIsHermitian(), MatIsTranspose()
4086 @*/
4087 PetscErrorCode PETSCMAT_DLLEXPORT MatIsHermitianTranspose(Mat A,Mat B,PetscReal tol,PetscTruth *flg)
4088 {
4089   PetscErrorCode ierr,(*f)(Mat,Mat,PetscReal,PetscTruth*),(*g)(Mat,Mat,PetscReal,PetscTruth*);
4090 
4091   PetscFunctionBegin;
4092   PetscValidHeaderSpecific(A,MAT_COOKIE,1);
4093   PetscValidHeaderSpecific(B,MAT_COOKIE,2);
4094   PetscValidPointer(flg,3);
4095   ierr = PetscObjectQueryFunction((PetscObject)A,"MatIsHermitianTranspose_C",(void (**)(void))&f);CHKERRQ(ierr);
4096   ierr = PetscObjectQueryFunction((PetscObject)B,"MatIsHermitianTranspose_C",(void (**)(void))&g);CHKERRQ(ierr);
4097   if (f && g) {
4098     if (f==g) {
4099       ierr = (*f)(A,B,tol,flg);CHKERRQ(ierr);
4100     } else {
4101       SETERRQ(PETSC_ERR_ARG_NOTSAMETYPE,"Matrices do not have the same comparator for Hermitian test");
4102     }
4103   }
4104   PetscFunctionReturn(0);
4105 }
4106 
4107 #undef __FUNCT__
4108 #define __FUNCT__ "MatPermute"
4109 /*@
4110    MatPermute - Creates a new matrix with rows and columns permuted from the
4111    original.
4112 
4113    Collective on Mat
4114 
4115    Input Parameters:
4116 +  mat - the matrix to permute
4117 .  row - row permutation, each processor supplies only the permutation for its rows
4118 -  col - column permutation, each processor needs the entire column permutation, that is
4119          this is the same size as the total number of columns in the matrix. It can often
4120          be obtained with ISAllGather() on the row permutation
4121 
4122    Output Parameters:
4123 .  B - the permuted matrix
4124 
4125    Level: advanced
4126 
4127    Concepts: matrices^permuting
4128 
4129 .seealso: MatGetOrdering(), ISAllGather()
4130 
4131 @*/
4132 PetscErrorCode PETSCMAT_DLLEXPORT MatPermute(Mat mat,IS row,IS col,Mat *B)
4133 {
4134   PetscErrorCode ierr;
4135 
4136   PetscFunctionBegin;
4137   PetscValidHeaderSpecific(mat,MAT_COOKIE,1);
4138   PetscValidType(mat,1);
4139   PetscValidHeaderSpecific(row,IS_COOKIE,2);
4140   PetscValidHeaderSpecific(col,IS_COOKIE,3);
4141   PetscValidPointer(B,4);
4142   if (!mat->assembled) SETERRQ(PETSC_ERR_ARG_WRONGSTATE,"Not for unassembled matrix");
4143   if (mat->factor) SETERRQ(PETSC_ERR_ARG_WRONGSTATE,"Not for factored matrix");
4144   if (!mat->ops->permute) SETERRQ1(PETSC_ERR_SUP,"MatPermute not available for Mat type %s",((PetscObject)mat)->type_name);
4145   ierr = MatPreallocated(mat);CHKERRQ(ierr);
4146 
4147   ierr = (*mat->ops->permute)(mat,row,col,B);CHKERRQ(ierr);
4148   ierr = PetscObjectStateIncrease((PetscObject)*B);CHKERRQ(ierr);
4149   PetscFunctionReturn(0);
4150 }
4151 
4152 #undef __FUNCT__
4153 #define __FUNCT__ "MatPermuteSparsify"
4154 /*@
4155   MatPermuteSparsify - Creates a new matrix with rows and columns permuted from the
4156   original and sparsified to the prescribed tolerance.
4157 
4158   Collective on Mat
4159 
4160   Input Parameters:
4161 + A    - The matrix to permute
4162 . band - The half-bandwidth of the sparsified matrix, or PETSC_DECIDE
4163 . frac - The half-bandwidth as a fraction of the total size, or 0.0
4164 . tol  - The drop tolerance
4165 . rowp - The row permutation
4166 - colp - The column permutation
4167 
4168   Output Parameter:
4169 . B    - The permuted, sparsified matrix
4170 
4171   Level: advanced
4172 
4173   Note:
4174   The default behavior (band = PETSC_DECIDE and frac = 0.0) is to
4175   restrict the half-bandwidth of the resulting matrix to 5% of the
4176   total matrix size.
4177 
4178 .keywords: matrix, permute, sparsify
4179 
4180 .seealso: MatGetOrdering(), MatPermute()
4181 @*/
4182 PetscErrorCode PETSCMAT_DLLEXPORT MatPermuteSparsify(Mat A, PetscInt band, PetscReal frac, PetscReal tol, IS rowp, IS colp, Mat *B)
4183 {
4184   IS                irowp, icolp;
4185   const PetscInt    *rows, *cols;
4186   PetscInt          M, N, locRowStart, locRowEnd;
4187   PetscInt          nz, newNz;
4188   const PetscInt    *cwork;
4189   PetscInt          *cnew;
4190   const PetscScalar *vwork;
4191   PetscScalar       *vnew;
4192   PetscInt          bw, issize;
4193   PetscInt          row, locRow, newRow, col, newCol;
4194   PetscErrorCode    ierr;
4195 
4196   PetscFunctionBegin;
4197   PetscValidHeaderSpecific(A,    MAT_COOKIE,1);
4198   PetscValidHeaderSpecific(rowp, IS_COOKIE,5);
4199   PetscValidHeaderSpecific(colp, IS_COOKIE,6);
4200   PetscValidPointer(B,7);
4201   if (!A->assembled) SETERRQ(PETSC_ERR_ARG_WRONGSTATE, "Not for unassembled matrix");
4202   if (A->factor)     SETERRQ(PETSC_ERR_ARG_WRONGSTATE, "Not for factored matrix");
4203   if (!A->ops->permutesparsify) {
4204     ierr = MatGetSize(A, &M, &N);CHKERRQ(ierr);
4205     ierr = MatGetOwnershipRange(A, &locRowStart, &locRowEnd);CHKERRQ(ierr);
4206     ierr = ISGetSize(rowp, &issize);CHKERRQ(ierr);
4207     if (issize != M) SETERRQ2(PETSC_ERR_ARG_WRONG, "Wrong size %D for row permutation, should be %D", issize, M);
4208     ierr = ISGetSize(colp, &issize);CHKERRQ(ierr);
4209     if (issize != N) SETERRQ2(PETSC_ERR_ARG_WRONG, "Wrong size %D for column permutation, should be %D", issize, N);
4210     ierr = ISInvertPermutation(rowp, 0, &irowp);CHKERRQ(ierr);
4211     ierr = ISGetIndices(irowp, &rows);CHKERRQ(ierr);
4212     ierr = ISInvertPermutation(colp, 0, &icolp);CHKERRQ(ierr);
4213     ierr = ISGetIndices(icolp, &cols);CHKERRQ(ierr);
4214     ierr = PetscMalloc(N * sizeof(PetscInt),         &cnew);CHKERRQ(ierr);
4215     ierr = PetscMalloc(N * sizeof(PetscScalar), &vnew);CHKERRQ(ierr);
4216 
4217     /* Setup bandwidth to include */
4218     if (band == PETSC_DECIDE) {
4219       if (frac <= 0.0)
4220         bw = (PetscInt) (M * 0.05);
4221       else
4222         bw = (PetscInt) (M * frac);
4223     } else {
4224       if (band <= 0) SETERRQ(PETSC_ERR_ARG_WRONG, "Bandwidth must be a positive integer");
4225       bw = band;
4226     }
4227 
4228     /* Put values into new matrix */
4229     ierr = MatDuplicate(A, MAT_DO_NOT_COPY_VALUES, B);CHKERRQ(ierr);
4230     for(row = locRowStart, locRow = 0; row < locRowEnd; row++, locRow++) {
4231       ierr = MatGetRow(A, row, &nz, &cwork, &vwork);CHKERRQ(ierr);
4232       newRow   = rows[locRow]+locRowStart;
4233       for(col = 0, newNz = 0; col < nz; col++) {
4234         newCol = cols[cwork[col]];
4235         if ((newCol >= newRow - bw) && (newCol < newRow + bw) && (PetscAbsScalar(vwork[col]) >= tol)) {
4236           cnew[newNz] = newCol;
4237           vnew[newNz] = vwork[col];
4238           newNz++;
4239         }
4240       }
4241       ierr = MatSetValues(*B, 1, &newRow, newNz, cnew, vnew, INSERT_VALUES);CHKERRQ(ierr);
4242       ierr = MatRestoreRow(A, row, &nz, &cwork, &vwork);CHKERRQ(ierr);
4243     }
4244     ierr = PetscFree(cnew);CHKERRQ(ierr);
4245     ierr = PetscFree(vnew);CHKERRQ(ierr);
4246     ierr = MatAssemblyBegin(*B, MAT_FINAL_ASSEMBLY);CHKERRQ(ierr);
4247     ierr = MatAssemblyEnd(*B, MAT_FINAL_ASSEMBLY);CHKERRQ(ierr);
4248     ierr = ISRestoreIndices(irowp, &rows);CHKERRQ(ierr);
4249     ierr = ISRestoreIndices(icolp, &cols);CHKERRQ(ierr);
4250     ierr = ISDestroy(irowp);CHKERRQ(ierr);
4251     ierr = ISDestroy(icolp);CHKERRQ(ierr);
4252   } else {
4253     ierr = (*A->ops->permutesparsify)(A, band, frac, tol, rowp, colp, B);CHKERRQ(ierr);
4254   }
4255   ierr = PetscObjectStateIncrease((PetscObject)*B);CHKERRQ(ierr);
4256   PetscFunctionReturn(0);
4257 }
4258 
4259 #undef __FUNCT__
4260 #define __FUNCT__ "MatEqual"
4261 /*@
4262    MatEqual - Compares two matrices.
4263 
4264    Collective on Mat
4265 
4266    Input Parameters:
4267 +  A - the first matrix
4268 -  B - the second matrix
4269 
4270    Output Parameter:
4271 .  flg - PETSC_TRUE if the matrices are equal; PETSC_FALSE otherwise.
4272 
4273    Level: intermediate
4274 
4275    Concepts: matrices^equality between
4276 @*/
4277 PetscErrorCode PETSCMAT_DLLEXPORT MatEqual(Mat A,Mat B,PetscTruth *flg)
4278 {
4279   PetscErrorCode ierr;
4280 
4281   PetscFunctionBegin;
4282   PetscValidHeaderSpecific(A,MAT_COOKIE,1);
4283   PetscValidHeaderSpecific(B,MAT_COOKIE,2);
4284   PetscValidType(A,1);
4285   PetscValidType(B,2);
4286   PetscValidIntPointer(flg,3);
4287   PetscCheckSameComm(A,1,B,2);
4288   ierr = MatPreallocated(B);CHKERRQ(ierr);
4289   if (!A->assembled) SETERRQ(PETSC_ERR_ARG_WRONGSTATE,"Not for unassembled matrix");
4290   if (!B->assembled) SETERRQ(PETSC_ERR_ARG_WRONGSTATE,"Not for unassembled matrix");
4291   if (A->rmap->N != B->rmap->N || A->cmap->N != B->cmap->N) SETERRQ4(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);
4292   if (!A->ops->equal) SETERRQ1(PETSC_ERR_SUP,"Mat type %s",((PetscObject)A)->type_name);
4293   if (!B->ops->equal) SETERRQ1(PETSC_ERR_SUP,"Mat type %s",((PetscObject)B)->type_name);
4294   if (A->ops->equal != B->ops->equal) SETERRQ2(PETSC_ERR_ARG_INCOMP,"A is type: %s\nB is type: %s",((PetscObject)A)->type_name,((PetscObject)B)->type_name);
4295   ierr = MatPreallocated(A);CHKERRQ(ierr);
4296 
4297   ierr = (*A->ops->equal)(A,B,flg);CHKERRQ(ierr);
4298   PetscFunctionReturn(0);
4299 }
4300 
4301 #undef __FUNCT__
4302 #define __FUNCT__ "MatDiagonalScale"
4303 /*@
4304    MatDiagonalScale - Scales a matrix on the left and right by diagonal
4305    matrices that are stored as vectors.  Either of the two scaling
4306    matrices can be PETSC_NULL.
4307 
4308    Collective on Mat
4309 
4310    Input Parameters:
4311 +  mat - the matrix to be scaled
4312 .  l - the left scaling vector (or PETSC_NULL)
4313 -  r - the right scaling vector (or PETSC_NULL)
4314 
4315    Notes:
4316    MatDiagonalScale() computes A = LAR, where
4317    L = a diagonal matrix (stored as a vector), R = a diagonal matrix (stored as a vector)
4318 
4319    Level: intermediate
4320 
4321    Concepts: matrices^diagonal scaling
4322    Concepts: diagonal scaling of matrices
4323 
4324 .seealso: MatScale()
4325 @*/
4326 PetscErrorCode PETSCMAT_DLLEXPORT MatDiagonalScale(Mat mat,Vec l,Vec r)
4327 {
4328   PetscErrorCode ierr;
4329 
4330   PetscFunctionBegin;
4331   PetscValidHeaderSpecific(mat,MAT_COOKIE,1);
4332   PetscValidType(mat,1);
4333   if (!mat->ops->diagonalscale) SETERRQ1(PETSC_ERR_SUP,"Mat type %s",((PetscObject)mat)->type_name);
4334   if (l) {PetscValidHeaderSpecific(l,VEC_COOKIE,2);PetscCheckSameComm(mat,1,l,2);}
4335   if (r) {PetscValidHeaderSpecific(r,VEC_COOKIE,3);PetscCheckSameComm(mat,1,r,3);}
4336   if (!mat->assembled) SETERRQ(PETSC_ERR_ARG_WRONGSTATE,"Not for unassembled matrix");
4337   if (mat->factor) SETERRQ(PETSC_ERR_ARG_WRONGSTATE,"Not for factored matrix");
4338   ierr = MatPreallocated(mat);CHKERRQ(ierr);
4339 
4340   ierr = PetscLogEventBegin(MAT_Scale,mat,0,0,0);CHKERRQ(ierr);
4341   ierr = (*mat->ops->diagonalscale)(mat,l,r);CHKERRQ(ierr);
4342   ierr = PetscLogEventEnd(MAT_Scale,mat,0,0,0);CHKERRQ(ierr);
4343   ierr = PetscObjectStateIncrease((PetscObject)mat);CHKERRQ(ierr);
4344   PetscFunctionReturn(0);
4345 }
4346 
4347 #undef __FUNCT__
4348 #define __FUNCT__ "MatScale"
4349 /*@
4350     MatScale - Scales all elements of a matrix by a given number.
4351 
4352     Collective on Mat
4353 
4354     Input Parameters:
4355 +   mat - the matrix to be scaled
4356 -   a  - the scaling value
4357 
4358     Output Parameter:
4359 .   mat - the scaled matrix
4360 
4361     Level: intermediate
4362 
4363     Concepts: matrices^scaling all entries
4364 
4365 .seealso: MatDiagonalScale()
4366 @*/
4367 PetscErrorCode PETSCMAT_DLLEXPORT MatScale(Mat mat,PetscScalar a)
4368 {
4369   PetscErrorCode ierr;
4370 
4371   PetscFunctionBegin;
4372   PetscValidHeaderSpecific(mat,MAT_COOKIE,1);
4373   PetscValidType(mat,1);
4374   if (a != 1.0 && !mat->ops->scale) SETERRQ1(PETSC_ERR_SUP,"Mat type %s",((PetscObject)mat)->type_name);
4375   if (!mat->assembled) SETERRQ(PETSC_ERR_ARG_WRONGSTATE,"Not for unassembled matrix");
4376   if (mat->factor) SETERRQ(PETSC_ERR_ARG_WRONGSTATE,"Not for factored matrix");
4377   ierr = MatPreallocated(mat);CHKERRQ(ierr);
4378 
4379   ierr = PetscLogEventBegin(MAT_Scale,mat,0,0,0);CHKERRQ(ierr);
4380   if (a != 1.0) {
4381     ierr = (*mat->ops->scale)(mat,a);CHKERRQ(ierr);
4382     ierr = PetscObjectStateIncrease((PetscObject)mat);CHKERRQ(ierr);
4383   }
4384   ierr = PetscLogEventEnd(MAT_Scale,mat,0,0,0);CHKERRQ(ierr);
4385   PetscFunctionReturn(0);
4386 }
4387 
4388 #undef __FUNCT__
4389 #define __FUNCT__ "MatNorm"
4390 /*@
4391    MatNorm - Calculates various norms of a matrix.
4392 
4393    Collective on Mat
4394 
4395    Input Parameters:
4396 +  mat - the matrix
4397 -  type - the type of norm, NORM_1, NORM_FROBENIUS, NORM_INFINITY
4398 
4399    Output Parameters:
4400 .  nrm - the resulting norm
4401 
4402    Level: intermediate
4403 
4404    Concepts: matrices^norm
4405    Concepts: norm^of matrix
4406 @*/
4407 PetscErrorCode PETSCMAT_DLLEXPORT MatNorm(Mat mat,NormType type,PetscReal *nrm)
4408 {
4409   PetscErrorCode ierr;
4410 
4411   PetscFunctionBegin;
4412   PetscValidHeaderSpecific(mat,MAT_COOKIE,1);
4413   PetscValidType(mat,1);
4414   PetscValidScalarPointer(nrm,3);
4415 
4416   if (!mat->assembled) SETERRQ(PETSC_ERR_ARG_WRONGSTATE,"Not for unassembled matrix");
4417   if (mat->factor) SETERRQ(PETSC_ERR_ARG_WRONGSTATE,"Not for factored matrix");
4418   if (!mat->ops->norm) SETERRQ1(PETSC_ERR_SUP,"Mat type %s",((PetscObject)mat)->type_name);
4419   ierr = MatPreallocated(mat);CHKERRQ(ierr);
4420 
4421   ierr = (*mat->ops->norm)(mat,type,nrm);CHKERRQ(ierr);
4422   PetscFunctionReturn(0);
4423 }
4424 
4425 /*
4426      This variable is used to prevent counting of MatAssemblyBegin() that
4427    are called from within a MatAssemblyEnd().
4428 */
4429 static PetscInt MatAssemblyEnd_InUse = 0;
4430 #undef __FUNCT__
4431 #define __FUNCT__ "MatAssemblyBegin"
4432 /*@
4433    MatAssemblyBegin - Begins assembling the matrix.  This routine should
4434    be called after completing all calls to MatSetValues().
4435 
4436    Collective on Mat
4437 
4438    Input Parameters:
4439 +  mat - the matrix
4440 -  type - type of assembly, either MAT_FLUSH_ASSEMBLY or MAT_FINAL_ASSEMBLY
4441 
4442    Notes:
4443    MatSetValues() generally caches the values.  The matrix is ready to
4444    use only after MatAssemblyBegin() and MatAssemblyEnd() have been called.
4445    Use MAT_FLUSH_ASSEMBLY when switching between ADD_VALUES and INSERT_VALUES
4446    in MatSetValues(); use MAT_FINAL_ASSEMBLY for the final assembly before
4447    using the matrix.
4448 
4449    Level: beginner
4450 
4451    Concepts: matrices^assembling
4452 
4453 .seealso: MatAssemblyEnd(), MatSetValues(), MatAssembled()
4454 @*/
4455 PetscErrorCode PETSCMAT_DLLEXPORT MatAssemblyBegin(Mat mat,MatAssemblyType type)
4456 {
4457   PetscErrorCode ierr;
4458 
4459   PetscFunctionBegin;
4460   PetscValidHeaderSpecific(mat,MAT_COOKIE,1);
4461   PetscValidType(mat,1);
4462   ierr = MatPreallocated(mat);CHKERRQ(ierr);
4463   if (mat->factor) SETERRQ(PETSC_ERR_ARG_WRONGSTATE,"Not for factored matrix.\nDid you forget to call MatSetUnfactored()?");
4464   if (mat->assembled) {
4465     mat->was_assembled = PETSC_TRUE;
4466     mat->assembled     = PETSC_FALSE;
4467   }
4468   if (!MatAssemblyEnd_InUse) {
4469     ierr = PetscLogEventBegin(MAT_AssemblyBegin,mat,0,0,0);CHKERRQ(ierr);
4470     if (mat->ops->assemblybegin){ierr = (*mat->ops->assemblybegin)(mat,type);CHKERRQ(ierr);}
4471     ierr = PetscLogEventEnd(MAT_AssemblyBegin,mat,0,0,0);CHKERRQ(ierr);
4472   } else {
4473     if (mat->ops->assemblybegin){ierr = (*mat->ops->assemblybegin)(mat,type);CHKERRQ(ierr);}
4474   }
4475   PetscFunctionReturn(0);
4476 }
4477 
4478 #undef __FUNCT__
4479 #define __FUNCT__ "MatAssembed"
4480 /*@
4481    MatAssembled - Indicates if a matrix has been assembled and is ready for
4482      use; for example, in matrix-vector product.
4483 
4484    Collective on Mat
4485 
4486    Input Parameter:
4487 .  mat - the matrix
4488 
4489    Output Parameter:
4490 .  assembled - PETSC_TRUE or PETSC_FALSE
4491 
4492    Level: advanced
4493 
4494    Concepts: matrices^assembled?
4495 
4496 .seealso: MatAssemblyEnd(), MatSetValues(), MatAssemblyBegin()
4497 @*/
4498 PetscErrorCode PETSCMAT_DLLEXPORT MatAssembled(Mat mat,PetscTruth *assembled)
4499 {
4500   PetscFunctionBegin;
4501   PetscValidHeaderSpecific(mat,MAT_COOKIE,1);
4502   PetscValidType(mat,1);
4503   PetscValidPointer(assembled,2);
4504   *assembled = mat->assembled;
4505   PetscFunctionReturn(0);
4506 }
4507 
4508 #undef __FUNCT__
4509 #define __FUNCT__ "MatView_Private"
4510 /*
4511     Processes command line options to determine if/how a matrix
4512   is to be viewed. Called by MatAssemblyEnd() and MatLoad().
4513 */
4514 PetscErrorCode MatView_Private(Mat mat)
4515 {
4516   PetscErrorCode    ierr;
4517   PetscTruth        flg1 = PETSC_FALSE,flg2 = PETSC_FALSE,flg3 = PETSC_FALSE,flg4 = PETSC_FALSE,flg6 = PETSC_FALSE,flg7 = PETSC_FALSE,flg8 = PETSC_FALSE;
4518   static PetscTruth incall = PETSC_FALSE;
4519 #if defined(PETSC_USE_SOCKET_VIEWER)
4520   PetscTruth        flg5 = PETSC_FALSE;
4521 #endif
4522 
4523   PetscFunctionBegin;
4524   if (incall) PetscFunctionReturn(0);
4525   incall = PETSC_TRUE;
4526   ierr = PetscOptionsBegin(((PetscObject)mat)->comm,((PetscObject)mat)->prefix,"Matrix Options","Mat");CHKERRQ(ierr);
4527     ierr = PetscOptionsTruth("-mat_view_info","Information on matrix size","MatView",flg1,&flg1,PETSC_NULL);CHKERRQ(ierr);
4528     ierr = PetscOptionsTruth("-mat_view_info_detailed","Nonzeros in the matrix","MatView",flg2,&flg2,PETSC_NULL);CHKERRQ(ierr);
4529     ierr = PetscOptionsTruth("-mat_view","Print matrix to stdout","MatView",flg3,&flg3,PETSC_NULL);CHKERRQ(ierr);
4530     ierr = PetscOptionsTruth("-mat_view_matlab","Print matrix to stdout in a format Matlab can read","MatView",flg4,&flg4,PETSC_NULL);CHKERRQ(ierr);
4531 #if defined(PETSC_USE_SOCKET_VIEWER)
4532     ierr = PetscOptionsTruth("-mat_view_socket","Send matrix to socket (can be read from matlab)","MatView",flg5,&flg5,PETSC_NULL);CHKERRQ(ierr);
4533 #endif
4534     ierr = PetscOptionsTruth("-mat_view_binary","Save matrix to file in binary format","MatView",flg6,&flg6,PETSC_NULL);CHKERRQ(ierr);
4535     ierr = PetscOptionsTruth("-mat_view_draw","Draw the matrix nonzero structure","MatView",flg7,&flg7,PETSC_NULL);CHKERRQ(ierr);
4536   ierr = PetscOptionsEnd();CHKERRQ(ierr);
4537 
4538   if (flg1) {
4539     PetscViewer viewer;
4540 
4541     ierr = PetscViewerASCIIGetStdout(((PetscObject)mat)->comm,&viewer);CHKERRQ(ierr);
4542     ierr = PetscViewerPushFormat(viewer,PETSC_VIEWER_ASCII_INFO);CHKERRQ(ierr);
4543     ierr = MatView(mat,viewer);CHKERRQ(ierr);
4544     ierr = PetscViewerPopFormat(viewer);CHKERRQ(ierr);
4545   }
4546   if (flg2) {
4547     PetscViewer viewer;
4548 
4549     ierr = PetscViewerASCIIGetStdout(((PetscObject)mat)->comm,&viewer);CHKERRQ(ierr);
4550     ierr = PetscViewerPushFormat(viewer,PETSC_VIEWER_ASCII_INFO_DETAIL);CHKERRQ(ierr);
4551     ierr = MatView(mat,viewer);CHKERRQ(ierr);
4552     ierr = PetscViewerPopFormat(viewer);CHKERRQ(ierr);
4553   }
4554   if (flg3) {
4555     PetscViewer viewer;
4556 
4557     ierr = PetscViewerASCIIGetStdout(((PetscObject)mat)->comm,&viewer);CHKERRQ(ierr);
4558     ierr = MatView(mat,viewer);CHKERRQ(ierr);
4559   }
4560   if (flg4) {
4561     PetscViewer viewer;
4562 
4563     ierr = PetscViewerASCIIGetStdout(((PetscObject)mat)->comm,&viewer);CHKERRQ(ierr);
4564     ierr = PetscViewerPushFormat(viewer,PETSC_VIEWER_ASCII_MATLAB);CHKERRQ(ierr);
4565     ierr = MatView(mat,viewer);CHKERRQ(ierr);
4566     ierr = PetscViewerPopFormat(viewer);CHKERRQ(ierr);
4567   }
4568 #if defined(PETSC_USE_SOCKET_VIEWER)
4569   if (flg5) {
4570     ierr = MatView(mat,PETSC_VIEWER_SOCKET_(((PetscObject)mat)->comm));CHKERRQ(ierr);
4571     ierr = PetscViewerFlush(PETSC_VIEWER_SOCKET_(((PetscObject)mat)->comm));CHKERRQ(ierr);
4572   }
4573 #endif
4574   if (flg6) {
4575     ierr = MatView(mat,PETSC_VIEWER_BINARY_(((PetscObject)mat)->comm));CHKERRQ(ierr);
4576     ierr = PetscViewerFlush(PETSC_VIEWER_BINARY_(((PetscObject)mat)->comm));CHKERRQ(ierr);
4577   }
4578   if (flg7) {
4579     ierr = PetscOptionsGetTruth(((PetscObject)mat)->prefix,"-mat_view_contour",&flg8,PETSC_NULL);CHKERRQ(ierr);
4580     if (flg8) {
4581       PetscViewerPushFormat(PETSC_VIEWER_DRAW_(((PetscObject)mat)->comm),PETSC_VIEWER_DRAW_CONTOUR);CHKERRQ(ierr);
4582     }
4583     ierr = MatView(mat,PETSC_VIEWER_DRAW_(((PetscObject)mat)->comm));CHKERRQ(ierr);
4584     ierr = PetscViewerFlush(PETSC_VIEWER_DRAW_(((PetscObject)mat)->comm));CHKERRQ(ierr);
4585     if (flg8) {
4586       PetscViewerPopFormat(PETSC_VIEWER_DRAW_(((PetscObject)mat)->comm));CHKERRQ(ierr);
4587     }
4588   }
4589   incall = PETSC_FALSE;
4590   PetscFunctionReturn(0);
4591 }
4592 
4593 #undef __FUNCT__
4594 #define __FUNCT__ "MatAssemblyEnd"
4595 /*@
4596    MatAssemblyEnd - Completes assembling the matrix.  This routine should
4597    be called after MatAssemblyBegin().
4598 
4599    Collective on Mat
4600 
4601    Input Parameters:
4602 +  mat - the matrix
4603 -  type - type of assembly, either MAT_FLUSH_ASSEMBLY or MAT_FINAL_ASSEMBLY
4604 
4605    Options Database Keys:
4606 +  -mat_view_info - Prints info on matrix at conclusion of MatEndAssembly()
4607 .  -mat_view_info_detailed - Prints more detailed info
4608 .  -mat_view - Prints matrix in ASCII format
4609 .  -mat_view_matlab - Prints matrix in Matlab format
4610 .  -mat_view_draw - PetscDraws nonzero structure of matrix, using MatView() and PetscDrawOpenX().
4611 .  -display <name> - Sets display name (default is host)
4612 .  -draw_pause <sec> - Sets number of seconds to pause after display
4613 .  -mat_view_socket - Sends matrix to socket, can be accessed from Matlab (see users manual)
4614 .  -viewer_socket_machine <machine>
4615 .  -viewer_socket_port <port>
4616 .  -mat_view_binary - save matrix to file in binary format
4617 -  -viewer_binary_filename <name>
4618 
4619    Notes:
4620    MatSetValues() generally caches the values.  The matrix is ready to
4621    use only after MatAssemblyBegin() and MatAssemblyEnd() have been called.
4622    Use MAT_FLUSH_ASSEMBLY when switching between ADD_VALUES and INSERT_VALUES
4623    in MatSetValues(); use MAT_FINAL_ASSEMBLY for the final assembly before
4624    using the matrix.
4625 
4626    Level: beginner
4627 
4628 .seealso: MatAssemblyBegin(), MatSetValues(), PetscDrawOpenX(), MatView(), MatAssembled(), PetscViewerSocketOpen()
4629 @*/
4630 PetscErrorCode PETSCMAT_DLLEXPORT MatAssemblyEnd(Mat mat,MatAssemblyType type)
4631 {
4632   PetscErrorCode  ierr;
4633   static PetscInt inassm = 0;
4634   PetscTruth      flg = PETSC_FALSE;
4635 
4636   PetscFunctionBegin;
4637   PetscValidHeaderSpecific(mat,MAT_COOKIE,1);
4638   PetscValidType(mat,1);
4639 
4640   inassm++;
4641   MatAssemblyEnd_InUse++;
4642   if (MatAssemblyEnd_InUse == 1) { /* Do the logging only the first time through */
4643     ierr = PetscLogEventBegin(MAT_AssemblyEnd,mat,0,0,0);CHKERRQ(ierr);
4644     if (mat->ops->assemblyend) {
4645       ierr = (*mat->ops->assemblyend)(mat,type);CHKERRQ(ierr);
4646     }
4647     ierr = PetscLogEventEnd(MAT_AssemblyEnd,mat,0,0,0);CHKERRQ(ierr);
4648   } else {
4649     if (mat->ops->assemblyend) {
4650       ierr = (*mat->ops->assemblyend)(mat,type);CHKERRQ(ierr);
4651     }
4652   }
4653 
4654   /* Flush assembly is not a true assembly */
4655   if (type != MAT_FLUSH_ASSEMBLY) {
4656     mat->assembled  = PETSC_TRUE; mat->num_ass++;
4657   }
4658   mat->insertmode = NOT_SET_VALUES;
4659   MatAssemblyEnd_InUse--;
4660   ierr = PetscObjectStateIncrease((PetscObject)mat);CHKERRQ(ierr);
4661   if (!mat->symmetric_eternal) {
4662     mat->symmetric_set              = PETSC_FALSE;
4663     mat->hermitian_set              = PETSC_FALSE;
4664     mat->structurally_symmetric_set = PETSC_FALSE;
4665   }
4666   if (inassm == 1 && type != MAT_FLUSH_ASSEMBLY) {
4667     ierr = MatView_Private(mat);CHKERRQ(ierr);
4668     ierr = PetscOptionsGetTruth(((PetscObject)mat)->prefix,"-mat_is_symmetric",&flg,PETSC_NULL);CHKERRQ(ierr);
4669     if (flg) {
4670       PetscReal tol = 0.0;
4671       ierr = PetscOptionsGetReal(((PetscObject)mat)->prefix,"-mat_is_symmetric",&tol,PETSC_NULL);CHKERRQ(ierr);
4672       ierr = MatIsSymmetric(mat,tol,&flg);CHKERRQ(ierr);
4673       if (flg) {
4674         ierr = PetscPrintf(((PetscObject)mat)->comm,"Matrix is symmetric (tolerance %G)\n",tol);CHKERRQ(ierr);
4675       } else {
4676         ierr = PetscPrintf(((PetscObject)mat)->comm,"Matrix is not symmetric (tolerance %G)\n",tol);CHKERRQ(ierr);
4677       }
4678     }
4679   }
4680   inassm--;
4681   PetscFunctionReturn(0);
4682 }
4683 
4684 #undef __FUNCT__
4685 #define __FUNCT__ "MatSetOption"
4686 /*@
4687    MatSetOption - Sets a parameter option for a matrix. Some options
4688    may be specific to certain storage formats.  Some options
4689    determine how values will be inserted (or added). Sorted,
4690    row-oriented input will generally assemble the fastest. The default
4691    is row-oriented, nonsorted input.
4692 
4693    Collective on Mat
4694 
4695    Input Parameters:
4696 +  mat - the matrix
4697 .  option - the option, one of those listed below (and possibly others),
4698 -  flg - turn the option on (PETSC_TRUE) or off (PETSC_FALSE)
4699 
4700   Options Describing Matrix Structure:
4701 +    MAT_SYMMETRIC - symmetric in terms of both structure and value
4702 .    MAT_HERMITIAN - transpose is the complex conjugation
4703 .    MAT_STRUCTURALLY_SYMMETRIC - symmetric nonzero structure
4704 -    MAT_SYMMETRY_ETERNAL - if you would like the symmetry/Hermitian flag
4705                             you set to be kept with all future use of the matrix
4706                             including after MatAssemblyBegin/End() which could
4707                             potentially change the symmetry structure, i.e. you
4708                             KNOW the matrix will ALWAYS have the property you set.
4709 
4710 
4711    Options For Use with MatSetValues():
4712    Insert a logically dense subblock, which can be
4713 .    MAT_ROW_ORIENTED - row-oriented (default)
4714 
4715    Note these options reflect the data you pass in with MatSetValues(); it has
4716    nothing to do with how the data is stored internally in the matrix
4717    data structure.
4718 
4719    When (re)assembling a matrix, we can restrict the input for
4720    efficiency/debugging purposes.  These options include
4721 +    MAT_NEW_NONZERO_LOCATIONS - additional insertions will be
4722         allowed if they generate a new nonzero
4723 .    MAT_NEW_DIAGONALS - new diagonals will be allowed (for block diagonal format only)
4724 .    MAT_IGNORE_OFF_PROC_ENTRIES - drops off-processor entries
4725 .    MAT_NEW_NONZERO_LOCATION_ERR - generates an error for new matrix entry
4726 -    MAT_USE_HASH_TABLE - uses a hash table to speed up matrix assembly
4727 
4728    Notes:
4729    Some options are relevant only for particular matrix types and
4730    are thus ignored by others.  Other options are not supported by
4731    certain matrix types and will generate an error message if set.
4732 
4733    If using a Fortran 77 module to compute a matrix, one may need to
4734    use the column-oriented option (or convert to the row-oriented
4735    format).
4736 
4737    MAT_NEW_NONZERO_LOCATIONS set to PETSC_FALSE indicates that any add or insertion
4738    that would generate a new entry in the nonzero structure is instead
4739    ignored.  Thus, if memory has not alredy been allocated for this particular
4740    data, then the insertion is ignored. For dense matrices, in which
4741    the entire array is allocated, no entries are ever ignored.
4742    Set after the first MatAssemblyEnd()
4743 
4744    MAT_NEW_NONZERO_LOCATION_ERR indicates that any add or insertion
4745    that would generate a new entry in the nonzero structure instead produces
4746    an error. (Currently supported for AIJ and BAIJ formats only.)
4747    This is a useful flag when using SAME_NONZERO_PATTERN in calling
4748    KSPSetOperators() to ensure that the nonzero pattern truely does
4749    remain unchanged. Set after the first MatAssemblyEnd()
4750 
4751    MAT_NEW_NONZERO_ALLOCATION_ERR indicates that any add or insertion
4752    that would generate a new entry that has not been preallocated will
4753    instead produce an error. (Currently supported for AIJ and BAIJ formats
4754    only.) This is a useful flag when debugging matrix memory preallocation.
4755 
4756    MAT_IGNORE_OFF_PROC_ENTRIES indicates entries destined for
4757    other processors should be dropped, rather than stashed.
4758    This is useful if you know that the "owning" processor is also
4759    always generating the correct matrix entries, so that PETSc need
4760    not transfer duplicate entries generated on another processor.
4761 
4762    MAT_USE_HASH_TABLE indicates that a hash table be used to improve the
4763    searches during matrix assembly. When this flag is set, the hash table
4764    is created during the first Matrix Assembly. This hash table is
4765    used the next time through, during MatSetVaules()/MatSetVaulesBlocked()
4766    to improve the searching of indices. MAT_NEW_NONZERO_LOCATIONS flag
4767    should be used with MAT_USE_HASH_TABLE flag. This option is currently
4768    supported by MATMPIBAIJ format only.
4769 
4770    MAT_KEEP_NONZERO_PATTERN indicates when MatZeroRows() is called the zeroed entries
4771    are kept in the nonzero structure
4772 
4773    MAT_IGNORE_ZERO_ENTRIES - for AIJ/IS matrices this will stop zero values from creating
4774    a zero location in the matrix
4775 
4776    MAT_USE_INODES - indicates using inode version of the code - works with AIJ and
4777    ROWBS matrix types
4778 
4779    Level: intermediate
4780 
4781    Concepts: matrices^setting options
4782 
4783 @*/
4784 PetscErrorCode PETSCMAT_DLLEXPORT MatSetOption(Mat mat,MatOption op,PetscTruth flg)
4785 {
4786   PetscErrorCode ierr;
4787 
4788   PetscFunctionBegin;
4789   PetscValidHeaderSpecific(mat,MAT_COOKIE,1);
4790   PetscValidType(mat,1);
4791   if (((int) op) < 0 || ((int) op) >= NUM_MAT_OPTIONS) SETERRQ1(PETSC_ERR_ARG_OUTOFRANGE,"Options %d is out of range",(int)op);
4792   ierr = MatPreallocated(mat);CHKERRQ(ierr);
4793   switch (op) {
4794   case MAT_SYMMETRIC:
4795     mat->symmetric                  = flg;
4796     if (flg) mat->structurally_symmetric     = PETSC_TRUE;
4797     mat->symmetric_set              = PETSC_TRUE;
4798     mat->structurally_symmetric_set = flg;
4799     break;
4800   case MAT_HERMITIAN:
4801     mat->hermitian                  = flg;
4802     if (flg) mat->structurally_symmetric     = PETSC_TRUE;
4803     mat->hermitian_set              = PETSC_TRUE;
4804     mat->structurally_symmetric_set = flg;
4805     break;
4806   case MAT_STRUCTURALLY_SYMMETRIC:
4807     mat->structurally_symmetric     = flg;
4808     mat->structurally_symmetric_set = PETSC_TRUE;
4809     break;
4810   case MAT_SYMMETRY_ETERNAL:
4811     mat->symmetric_eternal          = flg;
4812     break;
4813   default:
4814     break;
4815   }
4816   if (mat->ops->setoption) {
4817     ierr = (*mat->ops->setoption)(mat,op,flg);CHKERRQ(ierr);
4818   }
4819   PetscFunctionReturn(0);
4820 }
4821 
4822 #undef __FUNCT__
4823 #define __FUNCT__ "MatZeroEntries"
4824 /*@
4825    MatZeroEntries - Zeros all entries of a matrix.  For sparse matrices
4826    this routine retains the old nonzero structure.
4827 
4828    Collective on Mat
4829 
4830    Input Parameters:
4831 .  mat - the matrix
4832 
4833    Level: intermediate
4834 
4835    Concepts: matrices^zeroing
4836 
4837 .seealso: MatZeroRows()
4838 @*/
4839 PetscErrorCode PETSCMAT_DLLEXPORT MatZeroEntries(Mat mat)
4840 {
4841   PetscErrorCode ierr;
4842 
4843   PetscFunctionBegin;
4844   PetscValidHeaderSpecific(mat,MAT_COOKIE,1);
4845   PetscValidType(mat,1);
4846   if (mat->factor) SETERRQ(PETSC_ERR_ARG_WRONGSTATE,"Not for factored matrix");
4847   if (mat->insertmode != NOT_SET_VALUES) SETERRQ(PETSC_ERR_ARG_WRONGSTATE,"Not for matrices where you have set values but not yet assembled");
4848   if (!mat->ops->zeroentries) SETERRQ1(PETSC_ERR_SUP,"Mat type %s",((PetscObject)mat)->type_name);
4849   ierr = MatPreallocated(mat);CHKERRQ(ierr);
4850 
4851   ierr = PetscLogEventBegin(MAT_ZeroEntries,mat,0,0,0);CHKERRQ(ierr);
4852   ierr = (*mat->ops->zeroentries)(mat);CHKERRQ(ierr);
4853   ierr = PetscLogEventEnd(MAT_ZeroEntries,mat,0,0,0);CHKERRQ(ierr);
4854   ierr = PetscObjectStateIncrease((PetscObject)mat);CHKERRQ(ierr);
4855   PetscFunctionReturn(0);
4856 }
4857 
4858 #undef __FUNCT__
4859 #define __FUNCT__ "MatZeroRows"
4860 /*@C
4861    MatZeroRows - Zeros all entries (except possibly the main diagonal)
4862    of a set of rows of a matrix.
4863 
4864    Collective on Mat
4865 
4866    Input Parameters:
4867 +  mat - the matrix
4868 .  numRows - the number of rows to remove
4869 .  rows - the global row indices
4870 -  diag - value put in all diagonals of eliminated rows (0.0 will even eliminate diagonal entry)
4871 
4872    Notes:
4873    For the AIJ and BAIJ matrix formats this removes the old nonzero structure,
4874    but does not release memory.  For the dense and block diagonal
4875    formats this does not alter the nonzero structure.
4876 
4877    If the option MatSetOption(mat,MAT_KEEP_NONZERO_PATTERN,PETSC_TRUE) the nonzero structure
4878    of the matrix is not changed (even for AIJ and BAIJ matrices) the values are
4879    merely zeroed.
4880 
4881    The user can set a value in the diagonal entry (or for the AIJ and
4882    row formats can optionally remove the main diagonal entry from the
4883    nonzero structure as well, by passing 0.0 as the final argument).
4884 
4885    For the parallel case, all processes that share the matrix (i.e.,
4886    those in the communicator used for matrix creation) MUST call this
4887    routine, regardless of whether any rows being zeroed are owned by
4888    them.
4889 
4890    Each processor can indicate any rows in the entire matrix to be zeroed (i.e. each process does NOT have to
4891    list only rows local to itself).
4892 
4893    Level: intermediate
4894 
4895    Concepts: matrices^zeroing rows
4896 
4897 .seealso: MatZeroRowsIS(), MatZeroEntries(), MatZeroRowsLocal(), MatSetOption()
4898 @*/
4899 PetscErrorCode PETSCMAT_DLLEXPORT MatZeroRows(Mat mat,PetscInt numRows,const PetscInt rows[],PetscScalar diag)
4900 {
4901   PetscErrorCode ierr;
4902 
4903   PetscFunctionBegin;
4904   PetscValidHeaderSpecific(mat,MAT_COOKIE,1);
4905   PetscValidType(mat,1);
4906   if (numRows) PetscValidIntPointer(rows,3);
4907   if (!mat->assembled) SETERRQ(PETSC_ERR_ARG_WRONGSTATE,"Not for unassembled matrix");
4908   if (mat->factor) SETERRQ(PETSC_ERR_ARG_WRONGSTATE,"Not for factored matrix");
4909   if (!mat->ops->zerorows) SETERRQ1(PETSC_ERR_SUP,"Mat type %s",((PetscObject)mat)->type_name);
4910   ierr = MatPreallocated(mat);CHKERRQ(ierr);
4911 
4912   ierr = (*mat->ops->zerorows)(mat,numRows,rows,diag);CHKERRQ(ierr);
4913   ierr = MatView_Private(mat);CHKERRQ(ierr);
4914   ierr = PetscObjectStateIncrease((PetscObject)mat);CHKERRQ(ierr);
4915   PetscFunctionReturn(0);
4916 }
4917 
4918 #undef __FUNCT__
4919 #define __FUNCT__ "MatZeroRowsIS"
4920 /*@C
4921    MatZeroRowsIS - Zeros all entries (except possibly the main diagonal)
4922    of a set of rows of a matrix.
4923 
4924    Collective on Mat
4925 
4926    Input Parameters:
4927 +  mat - the matrix
4928 .  is - index set of rows to remove
4929 -  diag - value put in all diagonals of eliminated rows
4930 
4931    Notes:
4932    For the AIJ and BAIJ matrix formats this removes the old nonzero structure,
4933    but does not release memory.  For the dense and block diagonal
4934    formats this does not alter the nonzero structure.
4935 
4936    If the option MatSetOption(mat,MAT_KEEP_NONZERO_PATTERN,PETSC_TRUE) the nonzero structure
4937    of the matrix is not changed (even for AIJ and BAIJ matrices) the values are
4938    merely zeroed.
4939 
4940    The user can set a value in the diagonal entry (or for the AIJ and
4941    row formats can optionally remove the main diagonal entry from the
4942    nonzero structure as well, by passing 0.0 as the final argument).
4943 
4944    For the parallel case, all processes that share the matrix (i.e.,
4945    those in the communicator used for matrix creation) MUST call this
4946    routine, regardless of whether any rows being zeroed are owned by
4947    them.
4948 
4949    Each processor should list the rows that IT wants zeroed
4950 
4951    Level: intermediate
4952 
4953    Concepts: matrices^zeroing rows
4954 
4955 .seealso: MatZeroRows(), MatZeroEntries(), MatZeroRowsLocal(), MatSetOption()
4956 @*/
4957 PetscErrorCode PETSCMAT_DLLEXPORT MatZeroRowsIS(Mat mat,IS is,PetscScalar diag)
4958 {
4959   PetscInt       numRows;
4960   const PetscInt *rows;
4961   PetscErrorCode ierr;
4962 
4963   PetscFunctionBegin;
4964   PetscValidHeaderSpecific(mat,MAT_COOKIE,1);
4965   PetscValidType(mat,1);
4966   PetscValidHeaderSpecific(is,IS_COOKIE,2);
4967   ierr = ISGetLocalSize(is,&numRows);CHKERRQ(ierr);
4968   ierr = ISGetIndices(is,&rows);CHKERRQ(ierr);
4969   ierr = MatZeroRows(mat,numRows,rows,diag);CHKERRQ(ierr);
4970   ierr = ISRestoreIndices(is,&rows);CHKERRQ(ierr);
4971   PetscFunctionReturn(0);
4972 }
4973 
4974 #undef __FUNCT__
4975 #define __FUNCT__ "MatZeroRowsLocal"
4976 /*@C
4977    MatZeroRowsLocal - Zeros all entries (except possibly the main diagonal)
4978    of a set of rows of a matrix; using local numbering of rows.
4979 
4980    Collective on Mat
4981 
4982    Input Parameters:
4983 +  mat - the matrix
4984 .  numRows - the number of rows to remove
4985 .  rows - the global row indices
4986 -  diag - value put in all diagonals of eliminated rows
4987 
4988    Notes:
4989    Before calling MatZeroRowsLocal(), the user must first set the
4990    local-to-global mapping by calling MatSetLocalToGlobalMapping().
4991 
4992    For the AIJ matrix formats this removes the old nonzero structure,
4993    but does not release memory.  For the dense and block diagonal
4994    formats this does not alter the nonzero structure.
4995 
4996    If the option MatSetOption(mat,MAT_KEEP_NONZERO_PATTERN,PETSC_TRUE) the nonzero structure
4997    of the matrix is not changed (even for AIJ and BAIJ matrices) the values are
4998    merely zeroed.
4999 
5000    The user can set a value in the diagonal entry (or for the AIJ and
5001    row formats can optionally remove the main diagonal entry from the
5002    nonzero structure as well, by passing 0.0 as the final argument).
5003 
5004    Level: intermediate
5005 
5006    Concepts: matrices^zeroing
5007 
5008 .seealso: MatZeroRows(), MatZeroRowsLocalIS(), MatZeroEntries(), MatZeroRows(), MatSetLocalToGlobalMapping
5009 @*/
5010 PetscErrorCode PETSCMAT_DLLEXPORT MatZeroRowsLocal(Mat mat,PetscInt numRows,const PetscInt rows[],PetscScalar diag)
5011 {
5012   PetscErrorCode ierr;
5013 
5014   PetscFunctionBegin;
5015   PetscValidHeaderSpecific(mat,MAT_COOKIE,1);
5016   PetscValidType(mat,1);
5017   if (numRows) PetscValidIntPointer(rows,3);
5018   if (!mat->assembled) SETERRQ(PETSC_ERR_ARG_WRONGSTATE,"Not for unassembled matrix");
5019   if (mat->factor) SETERRQ(PETSC_ERR_ARG_WRONGSTATE,"Not for factored matrix");
5020   ierr = MatPreallocated(mat);CHKERRQ(ierr);
5021 
5022   if (mat->ops->zerorowslocal) {
5023     ierr = (*mat->ops->zerorowslocal)(mat,numRows,rows,diag);CHKERRQ(ierr);
5024   } else {
5025     IS             is, newis;
5026     const PetscInt *newRows;
5027 
5028     if (!mat->mapping) SETERRQ(PETSC_ERR_ARG_WRONGSTATE,"Need to provide local to global mapping to matrix first");
5029     ierr = ISCreateGeneral(PETSC_COMM_SELF,numRows,rows,&is);CHKERRQ(ierr);
5030     ierr = ISLocalToGlobalMappingApplyIS(mat->mapping,is,&newis);CHKERRQ(ierr);
5031     ierr = ISGetIndices(newis,&newRows);CHKERRQ(ierr);
5032     ierr = (*mat->ops->zerorows)(mat,numRows,newRows,diag);CHKERRQ(ierr);
5033     ierr = ISRestoreIndices(newis,&newRows);CHKERRQ(ierr);
5034     ierr = ISDestroy(newis);CHKERRQ(ierr);
5035     ierr = ISDestroy(is);CHKERRQ(ierr);
5036   }
5037   ierr = PetscObjectStateIncrease((PetscObject)mat);CHKERRQ(ierr);
5038   PetscFunctionReturn(0);
5039 }
5040 
5041 #undef __FUNCT__
5042 #define __FUNCT__ "MatZeroRowsLocalIS"
5043 /*@C
5044    MatZeroRowsLocalIS - Zeros all entries (except possibly the main diagonal)
5045    of a set of rows of a matrix; using local numbering of rows.
5046 
5047    Collective on Mat
5048 
5049    Input Parameters:
5050 +  mat - the matrix
5051 .  is - index set of rows to remove
5052 -  diag - value put in all diagonals of eliminated rows
5053 
5054    Notes:
5055    Before calling MatZeroRowsLocalIS(), the user must first set the
5056    local-to-global mapping by calling MatSetLocalToGlobalMapping().
5057 
5058    For the AIJ matrix formats this removes the old nonzero structure,
5059    but does not release memory.  For the dense and block diagonal
5060    formats this does not alter the nonzero structure.
5061 
5062    If the option MatSetOption(mat,MAT_KEEP_NONZERO_PATTERN,PETSC_TRUE) the nonzero structure
5063    of the matrix is not changed (even for AIJ and BAIJ matrices) the values are
5064    merely zeroed.
5065 
5066    The user can set a value in the diagonal entry (or for the AIJ and
5067    row formats can optionally remove the main diagonal entry from the
5068    nonzero structure as well, by passing 0.0 as the final argument).
5069 
5070    Level: intermediate
5071 
5072    Concepts: matrices^zeroing
5073 
5074 .seealso: MatZeroRows(), MatZeroRowsLocal(), MatZeroEntries(), MatZeroRows(), MatSetLocalToGlobalMapping
5075 @*/
5076 PetscErrorCode PETSCMAT_DLLEXPORT MatZeroRowsLocalIS(Mat mat,IS is,PetscScalar diag)
5077 {
5078   PetscErrorCode ierr;
5079   PetscInt       numRows;
5080   const PetscInt *rows;
5081 
5082   PetscFunctionBegin;
5083   PetscValidHeaderSpecific(mat,MAT_COOKIE,1);
5084   PetscValidType(mat,1);
5085   PetscValidHeaderSpecific(is,IS_COOKIE,2);
5086   if (!mat->assembled) SETERRQ(PETSC_ERR_ARG_WRONGSTATE,"Not for unassembled matrix");
5087   if (mat->factor) SETERRQ(PETSC_ERR_ARG_WRONGSTATE,"Not for factored matrix");
5088   ierr = MatPreallocated(mat);CHKERRQ(ierr);
5089 
5090   ierr = ISGetLocalSize(is,&numRows);CHKERRQ(ierr);
5091   ierr = ISGetIndices(is,&rows);CHKERRQ(ierr);
5092   ierr = MatZeroRowsLocal(mat,numRows,rows,diag);CHKERRQ(ierr);
5093   ierr = ISRestoreIndices(is,&rows);CHKERRQ(ierr);
5094   PetscFunctionReturn(0);
5095 }
5096 
5097 #undef __FUNCT__
5098 #define __FUNCT__ "MatGetSize"
5099 /*@
5100    MatGetSize - Returns the numbers of rows and columns in a matrix.
5101 
5102    Not Collective
5103 
5104    Input Parameter:
5105 .  mat - the matrix
5106 
5107    Output Parameters:
5108 +  m - the number of global rows
5109 -  n - the number of global columns
5110 
5111    Note: both output parameters can be PETSC_NULL on input.
5112 
5113    Level: beginner
5114 
5115    Concepts: matrices^size
5116 
5117 .seealso: MatGetLocalSize()
5118 @*/
5119 PetscErrorCode PETSCMAT_DLLEXPORT MatGetSize(Mat mat,PetscInt *m,PetscInt* n)
5120 {
5121   PetscFunctionBegin;
5122   PetscValidHeaderSpecific(mat,MAT_COOKIE,1);
5123   if (m) *m = mat->rmap->N;
5124   if (n) *n = mat->cmap->N;
5125   PetscFunctionReturn(0);
5126 }
5127 
5128 #undef __FUNCT__
5129 #define __FUNCT__ "MatGetLocalSize"
5130 /*@
5131    MatGetLocalSize - Returns the number of rows and columns in a matrix
5132    stored locally.  This information may be implementation dependent, so
5133    use with care.
5134 
5135    Not Collective
5136 
5137    Input Parameters:
5138 .  mat - the matrix
5139 
5140    Output Parameters:
5141 +  m - the number of local rows
5142 -  n - the number of local columns
5143 
5144    Note: both output parameters can be PETSC_NULL on input.
5145 
5146    Level: beginner
5147 
5148    Concepts: matrices^local size
5149 
5150 .seealso: MatGetSize()
5151 @*/
5152 PetscErrorCode PETSCMAT_DLLEXPORT MatGetLocalSize(Mat mat,PetscInt *m,PetscInt* n)
5153 {
5154   PetscFunctionBegin;
5155   PetscValidHeaderSpecific(mat,MAT_COOKIE,1);
5156   if (m) PetscValidIntPointer(m,2);
5157   if (n) PetscValidIntPointer(n,3);
5158   if (m) *m = mat->rmap->n;
5159   if (n) *n = mat->cmap->n;
5160   PetscFunctionReturn(0);
5161 }
5162 
5163 #undef __FUNCT__
5164 #define __FUNCT__ "MatGetOwnershipRangeColumn"
5165 /*@
5166    MatGetOwnershipRangeColumn - Returns the range of matrix columns owned by
5167    this processor.
5168 
5169    Not Collective, unless matrix has not been allocated, then collective on Mat
5170 
5171    Input Parameters:
5172 .  mat - the matrix
5173 
5174    Output Parameters:
5175 +  m - the global index of the first local column
5176 -  n - one more than the global index of the last local column
5177 
5178    Notes: both output parameters can be PETSC_NULL on input.
5179 
5180    Level: developer
5181 
5182    Concepts: matrices^column ownership
5183 
5184 .seealso:  MatGetOwnershipRange(), MatGetOwnershipRanges(), MatGetOwnershipRangesColumn()
5185 
5186 @*/
5187 PetscErrorCode PETSCMAT_DLLEXPORT MatGetOwnershipRangeColumn(Mat mat,PetscInt *m,PetscInt* n)
5188 {
5189   PetscErrorCode ierr;
5190 
5191   PetscFunctionBegin;
5192   PetscValidHeaderSpecific(mat,MAT_COOKIE,1);
5193   PetscValidType(mat,1);
5194   if (m) PetscValidIntPointer(m,2);
5195   if (n) PetscValidIntPointer(n,3);
5196   ierr = MatPreallocated(mat);CHKERRQ(ierr);
5197   if (m) *m = mat->cmap->rstart;
5198   if (n) *n = mat->cmap->rend;
5199   PetscFunctionReturn(0);
5200 }
5201 
5202 #undef __FUNCT__
5203 #define __FUNCT__ "MatGetOwnershipRange"
5204 /*@
5205    MatGetOwnershipRange - Returns the range of matrix rows owned by
5206    this processor, assuming that the matrix is laid out with the first
5207    n1 rows on the first processor, the next n2 rows on the second, etc.
5208    For certain parallel layouts this range may not be well defined.
5209 
5210    Not Collective, unless matrix has not been allocated, then collective on Mat
5211 
5212    Input Parameters:
5213 .  mat - the matrix
5214 
5215    Output Parameters:
5216 +  m - the global index of the first local row
5217 -  n - one more than the global index of the last local row
5218 
5219    Note: both output parameters can be PETSC_NULL on input.
5220 
5221    Level: beginner
5222 
5223    Concepts: matrices^row ownership
5224 
5225 .seealso:   MatGetOwnershipRanges(), MatGetOwnershipRangeColumn(), MatGetOwnershipRangesColumn()
5226 
5227 @*/
5228 PetscErrorCode PETSCMAT_DLLEXPORT MatGetOwnershipRange(Mat mat,PetscInt *m,PetscInt* n)
5229 {
5230   PetscErrorCode ierr;
5231 
5232   PetscFunctionBegin;
5233   PetscValidHeaderSpecific(mat,MAT_COOKIE,1);
5234   PetscValidType(mat,1);
5235   if (m) PetscValidIntPointer(m,2);
5236   if (n) PetscValidIntPointer(n,3);
5237   ierr = MatPreallocated(mat);CHKERRQ(ierr);
5238   if (m) *m = mat->rmap->rstart;
5239   if (n) *n = mat->rmap->rend;
5240   PetscFunctionReturn(0);
5241 }
5242 
5243 #undef __FUNCT__
5244 #define __FUNCT__ "MatGetOwnershipRanges"
5245 /*@C
5246    MatGetOwnershipRanges - Returns the range of matrix rows owned by
5247    each process
5248 
5249    Not Collective, unless matrix has not been allocated, then collective on Mat
5250 
5251    Input Parameters:
5252 .  mat - the matrix
5253 
5254    Output Parameters:
5255 .  ranges - start of each processors portion plus one more then the total length at the end
5256 
5257    Level: beginner
5258 
5259    Concepts: matrices^row ownership
5260 
5261 .seealso:   MatGetOwnershipRange(), MatGetOwnershipRangeColumn(), MatGetOwnershipRangesColumn()
5262 
5263 @*/
5264 PetscErrorCode PETSCMAT_DLLEXPORT MatGetOwnershipRanges(Mat mat,const PetscInt **ranges)
5265 {
5266   PetscErrorCode ierr;
5267 
5268   PetscFunctionBegin;
5269   PetscValidHeaderSpecific(mat,MAT_COOKIE,1);
5270   PetscValidType(mat,1);
5271   ierr = MatPreallocated(mat);CHKERRQ(ierr);
5272   ierr = PetscMapGetRanges(mat->rmap,ranges);CHKERRQ(ierr);
5273   PetscFunctionReturn(0);
5274 }
5275 
5276 #undef __FUNCT__
5277 #define __FUNCT__ "MatGetOwnershipRangesColumn"
5278 /*@C
5279    MatGetOwnershipRangesColumn - Returns the range of local columns for each process
5280 
5281    Not Collective, unless matrix has not been allocated, then collective on Mat
5282 
5283    Input Parameters:
5284 .  mat - the matrix
5285 
5286    Output Parameters:
5287 .  ranges - start of each processors portion plus one more then the total length at the end
5288 
5289    Level: beginner
5290 
5291    Concepts: matrices^column ownership
5292 
5293 .seealso:   MatGetOwnershipRange(), MatGetOwnershipRangeColumn(), MatGetOwnershipRanges()
5294 
5295 @*/
5296 PetscErrorCode PETSCMAT_DLLEXPORT MatGetOwnershipRangesColumn(Mat mat,const PetscInt **ranges)
5297 {
5298   PetscErrorCode ierr;
5299 
5300   PetscFunctionBegin;
5301   PetscValidHeaderSpecific(mat,MAT_COOKIE,1);
5302   PetscValidType(mat,1);
5303   ierr = MatPreallocated(mat);CHKERRQ(ierr);
5304   ierr = PetscMapGetRanges(mat->cmap,ranges);CHKERRQ(ierr);
5305   PetscFunctionReturn(0);
5306 }
5307 
5308 #undef __FUNCT__
5309 #define __FUNCT__ "MatILUFactorSymbolic"
5310 /*@C
5311    MatILUFactorSymbolic - Performs symbolic ILU factorization of a matrix.
5312    Uses levels of fill only, not drop tolerance. Use MatLUFactorNumeric()
5313    to complete the factorization.
5314 
5315    Collective on Mat
5316 
5317    Input Parameters:
5318 +  mat - the matrix
5319 .  row - row permutation
5320 .  column - column permutation
5321 -  info - structure containing
5322 $      levels - number of levels of fill.
5323 $      expected fill - as ratio of original fill.
5324 $      1 or 0 - indicating force fill on diagonal (improves robustness for matrices
5325                 missing diagonal entries)
5326 
5327    Output Parameters:
5328 .  fact - new matrix that has been symbolically factored
5329 
5330    Notes:
5331    See the users manual for additional information about
5332    choosing the fill factor for better efficiency.
5333 
5334    Most users should employ the simplified KSP interface for linear solvers
5335    instead of working directly with matrix algebra routines such as this.
5336    See, e.g., KSPCreate().
5337 
5338    Level: developer
5339 
5340   Concepts: matrices^symbolic LU factorization
5341   Concepts: matrices^factorization
5342   Concepts: LU^symbolic factorization
5343 
5344 .seealso: MatLUFactorSymbolic(), MatLUFactorNumeric(), MatCholeskyFactor()
5345           MatGetOrdering(), MatFactorInfo
5346 
5347     Developer Note: fortran interface is not autogenerated as the f90
5348     interface defintion cannot be generated correctly [due to MatFactorInfo]
5349 
5350 @*/
5351 PetscErrorCode PETSCMAT_DLLEXPORT MatILUFactorSymbolic(Mat fact,Mat mat,IS row,IS col,const MatFactorInfo *info)
5352 {
5353   PetscErrorCode ierr;
5354 
5355   PetscFunctionBegin;
5356   PetscValidHeaderSpecific(mat,MAT_COOKIE,1);
5357   PetscValidType(mat,1);
5358   PetscValidHeaderSpecific(row,IS_COOKIE,2);
5359   PetscValidHeaderSpecific(col,IS_COOKIE,3);
5360   PetscValidPointer(info,4);
5361   PetscValidPointer(fact,5);
5362   if (info->levels < 0) SETERRQ1(PETSC_ERR_ARG_OUTOFRANGE,"Levels of fill negative %D",(PetscInt)info->levels);
5363   if (info->fill < 1.0) SETERRQ1(PETSC_ERR_ARG_OUTOFRANGE,"Expected fill less than 1.0 %G",info->fill);
5364   if (!(fact)->ops->ilufactorsymbolic) SETERRQ1(PETSC_ERR_SUP,"Matrix type %s  symbolic ILU",((PetscObject)mat)->type_name);
5365   if (!mat->assembled) SETERRQ(PETSC_ERR_ARG_WRONGSTATE,"Not for unassembled matrix");
5366   if (mat->factor) SETERRQ(PETSC_ERR_ARG_WRONGSTATE,"Not for factored matrix");
5367   ierr = MatPreallocated(mat);CHKERRQ(ierr);
5368 
5369   ierr = PetscLogEventBegin(MAT_ILUFactorSymbolic,mat,row,col,0);CHKERRQ(ierr);
5370   ierr = (fact->ops->ilufactorsymbolic)(fact,mat,row,col,info);CHKERRQ(ierr);
5371   ierr = PetscLogEventEnd(MAT_ILUFactorSymbolic,mat,row,col,0);CHKERRQ(ierr);
5372   PetscFunctionReturn(0);
5373 }
5374 
5375 #undef __FUNCT__
5376 #define __FUNCT__ "MatICCFactorSymbolic"
5377 /*@C
5378    MatICCFactorSymbolic - Performs symbolic incomplete
5379    Cholesky factorization for a symmetric matrix.  Use
5380    MatCholeskyFactorNumeric() to complete the factorization.
5381 
5382    Collective on Mat
5383 
5384    Input Parameters:
5385 +  mat - the matrix
5386 .  perm - row and column permutation
5387 -  info - structure containing
5388 $      levels - number of levels of fill.
5389 $      expected fill - as ratio of original fill.
5390 
5391    Output Parameter:
5392 .  fact - the factored matrix
5393 
5394    Notes:
5395    Most users should employ the KSP interface for linear solvers
5396    instead of working directly with matrix algebra routines such as this.
5397    See, e.g., KSPCreate().
5398 
5399    Level: developer
5400 
5401   Concepts: matrices^symbolic incomplete Cholesky factorization
5402   Concepts: matrices^factorization
5403   Concepts: Cholsky^symbolic factorization
5404 
5405 .seealso: MatCholeskyFactorNumeric(), MatCholeskyFactor(), MatFactorInfo
5406 
5407     Developer Note: fortran interface is not autogenerated as the f90
5408     interface defintion cannot be generated correctly [due to MatFactorInfo]
5409 
5410 @*/
5411 PetscErrorCode PETSCMAT_DLLEXPORT MatICCFactorSymbolic(Mat fact,Mat mat,IS perm,const MatFactorInfo *info)
5412 {
5413   PetscErrorCode ierr;
5414 
5415   PetscFunctionBegin;
5416   PetscValidHeaderSpecific(mat,MAT_COOKIE,1);
5417   PetscValidType(mat,1);
5418   PetscValidHeaderSpecific(perm,IS_COOKIE,2);
5419   PetscValidPointer(info,3);
5420   PetscValidPointer(fact,4);
5421   if (mat->factor) SETERRQ(PETSC_ERR_ARG_WRONGSTATE,"Not for factored matrix");
5422   if (info->levels < 0) SETERRQ1(PETSC_ERR_ARG_OUTOFRANGE,"Levels negative %D",(PetscInt) info->levels);
5423   if (info->fill < 1.0) SETERRQ1(PETSC_ERR_ARG_OUTOFRANGE,"Expected fill less than 1.0 %G",info->fill);
5424   if (!(fact)->ops->iccfactorsymbolic) SETERRQ1(PETSC_ERR_SUP,"Matrix type %s  symbolic ICC",((PetscObject)mat)->type_name);
5425   if (!mat->assembled) SETERRQ(PETSC_ERR_ARG_WRONGSTATE,"Not for unassembled matrix");
5426   ierr = MatPreallocated(mat);CHKERRQ(ierr);
5427 
5428   ierr = PetscLogEventBegin(MAT_ICCFactorSymbolic,mat,perm,0,0);CHKERRQ(ierr);
5429   ierr = (fact->ops->iccfactorsymbolic)(fact,mat,perm,info);CHKERRQ(ierr);
5430   ierr = PetscLogEventEnd(MAT_ICCFactorSymbolic,mat,perm,0,0);CHKERRQ(ierr);
5431   PetscFunctionReturn(0);
5432 }
5433 
5434 #undef __FUNCT__
5435 #define __FUNCT__ "MatILUDTFactorSymbolic"
5436 /*@
5437    MatILUDTFactorSymbolic - Performs symbolic pivoting drop-tolerance ILU factorization of a matrix.
5438    User provides the drop tolerance(dt) and the maximum nonzeros to be allowed per row(dtcount).
5439    Use MatILUDTFactorNumeric() to complete the factorization.
5440 
5441    Collective on Mat
5442 
5443    Input Parameters:
5444 +  mat - the matrix
5445 .  row - row permutation
5446 .  column - column permutation
5447 -  info - structure containing
5448 $      dt - drop tolerance.
5449 $      dtcount - maximum nonzeros to be allowed per row.
5450 
5451    Output Parameters:
5452 .  fact - factor matrix with memory preallocated
5453 
5454    Notes:
5455    See the ILUT algorithm written by Yousef Saad.
5456 
5457    Most users should employ the simplified KSP interface for linear solvers
5458    instead of working directly with matrix algebra routines such as this.
5459    See, e.g., KSPCreate().
5460 
5461    Level: developer
5462 
5463   Concepts: matrices^symbolic ILU factorization
5464   Concepts: matrices^factorization
5465 
5466 .seealso: MatILUDTFactorNumeric()
5467           MatGetOrdering(), MatGetFactor(), MatFactorInfo
5468 
5469 @*/
5470 PetscErrorCode PETSCMAT_DLLEXPORT MatILUDTFactorSymbolic(Mat fact,Mat mat,IS row,IS col,const MatFactorInfo *info)
5471 {
5472   PetscErrorCode ierr;
5473 
5474   PetscFunctionBegin;
5475   PetscValidHeaderSpecific(mat,MAT_COOKIE,1);
5476   PetscValidType(mat,1);
5477   PetscValidHeaderSpecific(row,IS_COOKIE,2);
5478   PetscValidHeaderSpecific(col,IS_COOKIE,3);
5479   PetscValidPointer(info,4);
5480   PetscValidPointer(fact,5);
5481   if (info->dt < 0) SETERRQ1(PETSC_ERR_ARG_OUTOFRANGE,"drop tolerance negative %G",(PetscInt)info->dt);
5482   if (info->dtcount < 0) SETERRQ1(PETSC_ERR_ARG_OUTOFRANGE,"nonzeros per row %D <0",info->dtcount);
5483   if (!(fact)->ops->iludtfactorsymbolic) SETERRQ1(PETSC_ERR_SUP,"Matrix type %s  symbolic ILUDT",((PetscObject)mat)->type_name);
5484   if (mat->rmap->N != mat->cmap->N) SETERRQ(PETSC_ERR_ARG_WRONG,"matrix must be square");
5485   if (!mat->assembled) SETERRQ(PETSC_ERR_ARG_WRONGSTATE,"Not for unassembled matrix");
5486   if (mat->factor) SETERRQ(PETSC_ERR_ARG_WRONGSTATE,"Not for factored matrix");
5487   ierr = MatPreallocated(mat);CHKERRQ(ierr);
5488 
5489   ierr = PetscLogEventBegin(MAT_ILUDTFactorSymbolic,mat,row,col,0);CHKERRQ(ierr);
5490   ierr = (fact->ops->iludtfactorsymbolic)(fact,mat,row,col,info);CHKERRQ(ierr);
5491   ierr = PetscLogEventEnd(MAT_ILUDTFactorSymbolic,mat,row,col,0);CHKERRQ(ierr);
5492   PetscFunctionReturn(0);
5493 }
5494 
5495 #undef __FUNCT__
5496 #define __FUNCT__ "MatILUDTFactorNumeric"
5497 /*@
5498    MatILUDTFactorNumeric - Performs numeric pivoting drop-tolerance ILU factorization of a matrix.
5499    Call this routine after first calling MatILUDTFactorSymbolic().
5500 
5501    Collective on Mat
5502 
5503    Input Parameters:
5504 +  fact - the factor matrix obtained with MatGetFactor()
5505 .  mat - the matrix
5506 -  info - options for factorization
5507 
5508    Output Parameters:
5509 .  fact - assembled factor matrix
5510 
5511    Notes:
5512    Most users should employ the simplified KSP interface for linear solvers
5513    instead of working directly with matrix algebra routines such as this.
5514    See, e.g., KSPCreate().
5515 
5516    Level: developer
5517 
5518 .seealso: MatILUDTFactorSymbolic()
5519 @*/
5520 PetscErrorCode PETSCMAT_DLLEXPORT MatILUDTFactorNumeric(Mat fact,Mat mat,const MatFactorInfo *info)
5521 {
5522   PetscErrorCode ierr;
5523 
5524   PetscFunctionBegin;
5525   PetscValidHeaderSpecific(mat,MAT_COOKIE,1);
5526   PetscValidType(mat,1);
5527   PetscValidPointer(fact,2);
5528   PetscValidHeaderSpecific(fact,MAT_COOKIE,2);
5529   if (!mat->assembled) SETERRQ(PETSC_ERR_ARG_WRONGSTATE,"Not for unassembled matrix");
5530   if (mat->rmap->N != (fact)->rmap->N || mat->cmap->N != (fact)->cmap->N) {
5531     SETERRQ4(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);
5532   }
5533   if (!(fact)->ops->iludtfactornumeric) SETERRQ1(PETSC_ERR_SUP,"Mat type %s",((PetscObject)mat)->type_name);
5534   ierr = MatPreallocated(mat);CHKERRQ(ierr);
5535   ierr = PetscLogEventBegin(MAT_ILUDTFactorNumeric,mat,fact,0,0);CHKERRQ(ierr);
5536   ierr = (fact->ops->iludtfactornumeric)(fact,mat,info);CHKERRQ(ierr);
5537   ierr = PetscLogEventEnd(MAT_ILUDTFactorNumeric,mat,fact,0,0);CHKERRQ(ierr);
5538 
5539   ierr = MatView_Private(fact);CHKERRQ(ierr);
5540   ierr = PetscObjectStateIncrease((PetscObject)fact);CHKERRQ(ierr);
5541   PetscFunctionReturn(0);
5542 }
5543 
5544 #undef __FUNCT__
5545 #define __FUNCT__ "MatGetArray"
5546 /*@C
5547    MatGetArray - Returns a pointer to the element values in the matrix.
5548    The result of this routine is dependent on the underlying matrix data
5549    structure, and may not even work for certain matrix types.  You MUST
5550    call MatRestoreArray() when you no longer need to access the array.
5551 
5552    Not Collective
5553 
5554    Input Parameter:
5555 .  mat - the matrix
5556 
5557    Output Parameter:
5558 .  v - the location of the values
5559 
5560 
5561    Fortran Note:
5562    This routine is used differently from Fortran, e.g.,
5563 .vb
5564         Mat         mat
5565         PetscScalar mat_array(1)
5566         PetscOffset i_mat
5567         PetscErrorCode ierr
5568         call MatGetArray(mat,mat_array,i_mat,ierr)
5569 
5570   C  Access first local entry in matrix; note that array is
5571   C  treated as one dimensional
5572         value = mat_array(i_mat + 1)
5573 
5574         [... other code ...]
5575         call MatRestoreArray(mat,mat_array,i_mat,ierr)
5576 .ve
5577 
5578    See the Fortran chapter of the users manual and
5579    petsc/src/mat/examples/tests for details.
5580 
5581    Level: advanced
5582 
5583    Concepts: matrices^access array
5584 
5585 .seealso: MatRestoreArray(), MatGetArrayF90(), MatGetRowIJ()
5586 @*/
5587 PetscErrorCode PETSCMAT_DLLEXPORT MatGetArray(Mat mat,PetscScalar *v[])
5588 {
5589   PetscErrorCode ierr;
5590 
5591   PetscFunctionBegin;
5592   PetscValidHeaderSpecific(mat,MAT_COOKIE,1);
5593   PetscValidType(mat,1);
5594   PetscValidPointer(v,2);
5595   if (!mat->ops->getarray) SETERRQ1(PETSC_ERR_SUP,"Mat type %s",((PetscObject)mat)->type_name);
5596   ierr = MatPreallocated(mat);CHKERRQ(ierr);
5597   ierr = (*mat->ops->getarray)(mat,v);CHKERRQ(ierr);
5598   CHKMEMQ;
5599   PetscFunctionReturn(0);
5600 }
5601 
5602 #undef __FUNCT__
5603 #define __FUNCT__ "MatRestoreArray"
5604 /*@C
5605    MatRestoreArray - Restores the matrix after MatGetArray() has been called.
5606 
5607    Not Collective
5608 
5609    Input Parameter:
5610 +  mat - the matrix
5611 -  v - the location of the values
5612 
5613    Fortran Note:
5614    This routine is used differently from Fortran, e.g.,
5615 .vb
5616         Mat         mat
5617         PetscScalar mat_array(1)
5618         PetscOffset i_mat
5619         PetscErrorCode ierr
5620         call MatGetArray(mat,mat_array,i_mat,ierr)
5621 
5622   C  Access first local entry in matrix; note that array is
5623   C  treated as one dimensional
5624         value = mat_array(i_mat + 1)
5625 
5626         [... other code ...]
5627         call MatRestoreArray(mat,mat_array,i_mat,ierr)
5628 .ve
5629 
5630    See the Fortran chapter of the users manual and
5631    petsc/src/mat/examples/tests for details
5632 
5633    Level: advanced
5634 
5635 .seealso: MatGetArray(), MatRestoreArrayF90()
5636 @*/
5637 PetscErrorCode PETSCMAT_DLLEXPORT MatRestoreArray(Mat mat,PetscScalar *v[])
5638 {
5639   PetscErrorCode ierr;
5640 
5641   PetscFunctionBegin;
5642   PetscValidHeaderSpecific(mat,MAT_COOKIE,1);
5643   PetscValidType(mat,1);
5644   PetscValidPointer(v,2);
5645 #if defined(PETSC_USE_DEBUG)
5646   CHKMEMQ;
5647 #endif
5648   if (!mat->ops->restorearray) SETERRQ1(PETSC_ERR_SUP,"Mat type %s",((PetscObject)mat)->type_name);
5649   ierr = (*mat->ops->restorearray)(mat,v);CHKERRQ(ierr);
5650   ierr = PetscObjectStateIncrease((PetscObject)mat);CHKERRQ(ierr);
5651   PetscFunctionReturn(0);
5652 }
5653 
5654 #undef __FUNCT__
5655 #define __FUNCT__ "MatGetSubMatrices"
5656 /*@C
5657    MatGetSubMatrices - Extracts several submatrices from a matrix. If submat
5658    points to an array of valid matrices, they may be reused to store the new
5659    submatrices.
5660 
5661    Collective on Mat
5662 
5663    Input Parameters:
5664 +  mat - the matrix
5665 .  n   - the number of submatrixes to be extracted (on this processor, may be zero)
5666 .  irow, icol - index sets of rows and columns to extract
5667 -  scall - either MAT_INITIAL_MATRIX or MAT_REUSE_MATRIX
5668 
5669    Output Parameter:
5670 .  submat - the array of submatrices
5671 
5672    Notes:
5673    MatGetSubMatrices() can extract ONLY sequential submatrices
5674    (from both sequential and parallel matrices). Use MatGetSubMatrix()
5675    to extract a parallel submatrix.
5676 
5677    When extracting submatrices from a parallel matrix, each processor can
5678    form a different submatrix by setting the rows and columns of its
5679    individual index sets according to the local submatrix desired.
5680 
5681    When finished using the submatrices, the user should destroy
5682    them with MatDestroyMatrices().
5683 
5684    MAT_REUSE_MATRIX can only be used when the nonzero structure of the
5685    original matrix has not changed from that last call to MatGetSubMatrices().
5686 
5687    This routine creates the matrices in submat; you should NOT create them before
5688    calling it. It also allocates the array of matrix pointers submat.
5689 
5690    For BAIJ matrices the index sets must respect the block structure, that is if they
5691    request one row/column in a block, they must request all rows/columns that are in
5692    that block. For example, if the block size is 2 you cannot request just row 0 and
5693    column 0.
5694 
5695    Fortran Note:
5696    The Fortran interface is slightly different from that given below; it
5697    requires one to pass in  as submat a Mat (integer) array of size at least m.
5698 
5699    Level: advanced
5700 
5701    Concepts: matrices^accessing submatrices
5702    Concepts: submatrices
5703 
5704 .seealso: MatDestroyMatrices(), MatGetSubMatrix(), MatGetRow(), MatGetDiagonal(), MatReuse
5705 @*/
5706 PetscErrorCode PETSCMAT_DLLEXPORT MatGetSubMatrices(Mat mat,PetscInt n,const IS irow[],const IS icol[],MatReuse scall,Mat *submat[])
5707 {
5708   PetscErrorCode ierr;
5709   PetscInt        i;
5710   PetscTruth      eq;
5711 
5712   PetscFunctionBegin;
5713   PetscValidHeaderSpecific(mat,MAT_COOKIE,1);
5714   PetscValidType(mat,1);
5715   if (n) {
5716     PetscValidPointer(irow,3);
5717     PetscValidHeaderSpecific(*irow,IS_COOKIE,3);
5718     PetscValidPointer(icol,4);
5719     PetscValidHeaderSpecific(*icol,IS_COOKIE,4);
5720   }
5721   PetscValidPointer(submat,6);
5722   if (n && scall == MAT_REUSE_MATRIX) {
5723     PetscValidPointer(*submat,6);
5724     PetscValidHeaderSpecific(**submat,MAT_COOKIE,6);
5725   }
5726   if (!mat->ops->getsubmatrices) SETERRQ1(PETSC_ERR_SUP,"Mat type %s",((PetscObject)mat)->type_name);
5727   if (!mat->assembled) SETERRQ(PETSC_ERR_ARG_WRONGSTATE,"Not for unassembled matrix");
5728   if (mat->factor) SETERRQ(PETSC_ERR_ARG_WRONGSTATE,"Not for factored matrix");
5729   ierr = MatPreallocated(mat);CHKERRQ(ierr);
5730 
5731   ierr = PetscLogEventBegin(MAT_GetSubMatrices,mat,0,0,0);CHKERRQ(ierr);
5732   ierr = (*mat->ops->getsubmatrices)(mat,n,irow,icol,scall,submat);CHKERRQ(ierr);
5733   ierr = PetscLogEventEnd(MAT_GetSubMatrices,mat,0,0,0);CHKERRQ(ierr);
5734   for (i=0; i<n; i++) {
5735     if (mat->symmetric || mat->structurally_symmetric || mat->hermitian) {
5736       ierr = ISEqual(irow[i],icol[i],&eq);CHKERRQ(ierr);
5737       if (eq) {
5738 	if (mat->symmetric){
5739 	  ierr = MatSetOption((*submat)[i],MAT_SYMMETRIC,PETSC_TRUE);CHKERRQ(ierr);
5740 	} else if (mat->hermitian) {
5741 	  ierr = MatSetOption((*submat)[i],MAT_HERMITIAN,PETSC_TRUE);CHKERRQ(ierr);
5742 	} else if (mat->structurally_symmetric) {
5743 	  ierr = MatSetOption((*submat)[i],MAT_STRUCTURALLY_SYMMETRIC,PETSC_TRUE);CHKERRQ(ierr);
5744 	}
5745       }
5746     }
5747   }
5748   PetscFunctionReturn(0);
5749 }
5750 
5751 #undef __FUNCT__
5752 #define __FUNCT__ "MatDestroyMatrices"
5753 /*@C
5754    MatDestroyMatrices - Destroys a set of matrices obtained with MatGetSubMatrices().
5755 
5756    Collective on Mat
5757 
5758    Input Parameters:
5759 +  n - the number of local matrices
5760 -  mat - the matrices (note that this is a pointer to the array of matrices, just to match the calling
5761                        sequence of MatGetSubMatrices())
5762 
5763    Level: advanced
5764 
5765     Notes: Frees not only the matrices, but also the array that contains the matrices
5766            In Fortran will not free the array.
5767 
5768 .seealso: MatGetSubMatrices()
5769 @*/
5770 PetscErrorCode PETSCMAT_DLLEXPORT MatDestroyMatrices(PetscInt n,Mat *mat[])
5771 {
5772   PetscErrorCode ierr;
5773   PetscInt       i;
5774 
5775   PetscFunctionBegin;
5776   if (n < 0) SETERRQ1(PETSC_ERR_ARG_OUTOFRANGE,"Trying to destroy negative number of matrices %D",n);
5777   PetscValidPointer(mat,2);
5778   for (i=0; i<n; i++) {
5779     ierr = MatDestroy((*mat)[i]);CHKERRQ(ierr);
5780   }
5781   /* memory is allocated even if n = 0 */
5782   ierr = PetscFree(*mat);CHKERRQ(ierr);
5783   PetscFunctionReturn(0);
5784 }
5785 
5786 #undef __FUNCT__
5787 #define __FUNCT__ "MatGetSeqNonzeroStructure"
5788 /*@C
5789    MatGetSeqNonzeroStructure - Extracts the sequential nonzero structure from a matrix.
5790 
5791    Collective on Mat
5792 
5793    Input Parameters:
5794 .  mat - the matrix
5795 
5796    Output Parameter:
5797 .  matstruct - the sequential matrix with the nonzero structure of mat
5798 
5799   Level: intermediate
5800 
5801 .seealso: MatDestroySeqNonzeroStructure(), MatGetSubMatrices(), MatDestroyMatrices()
5802 @*/
5803 PetscErrorCode PETSCMAT_DLLEXPORT MatGetSeqNonzeroStructure(Mat mat,Mat *matstruct)
5804 {
5805   PetscErrorCode ierr;
5806 
5807   PetscFunctionBegin;
5808   PetscValidHeaderSpecific(mat,MAT_COOKIE,1);
5809   PetscValidPointer(matstruct,2);
5810 
5811   PetscValidType(mat,1);
5812   if (mat->factor) SETERRQ(PETSC_ERR_ARG_WRONGSTATE,"Not for factored matrix");
5813   ierr = MatPreallocated(mat);CHKERRQ(ierr);
5814 
5815   if (!mat->ops->getseqnonzerostructure) SETERRQ1(PETSC_ERR_SUP,"Not for matrix type %s\n",((PetscObject)mat)->type_name);
5816   ierr = PetscLogEventBegin(MAT_GetSeqNonzeroStructure,mat,0,0,0);CHKERRQ(ierr);
5817   ierr = (*mat->ops->getseqnonzerostructure)(mat,matstruct);CHKERRQ(ierr);
5818   ierr = PetscLogEventEnd(MAT_GetSeqNonzeroStructure,mat,0,0,0);CHKERRQ(ierr);
5819   PetscFunctionReturn(0);
5820 }
5821 
5822 #undef __FUNCT__
5823 #define __FUNCT__ "MatDestroySeqNonzeroStructure"
5824 /*@C
5825    MatDestroySeqNonzeroStructure - Destroys matrix obtained with MatGetSeqNonzeroStructure().
5826 
5827    Collective on Mat
5828 
5829    Input Parameters:
5830 .  mat - the matrix (note that this is a pointer to the array of matrices, just to match the calling
5831                        sequence of MatGetSequentialNonzeroStructure())
5832 
5833    Level: advanced
5834 
5835     Notes: Frees not only the matrices, but also the array that contains the matrices
5836 
5837 .seealso: MatGetSeqNonzeroStructure()
5838 @*/
5839 PetscErrorCode PETSCMAT_DLLEXPORT MatDestroySeqNonzeroStructure(Mat *mat)
5840 {
5841   PetscErrorCode ierr;
5842 
5843   PetscFunctionBegin;
5844   PetscValidPointer(mat,1);
5845   ierr = MatDestroy(*mat);CHKERRQ(ierr);
5846   PetscFunctionReturn(0);
5847 }
5848 
5849 #undef __FUNCT__
5850 #define __FUNCT__ "MatIncreaseOverlap"
5851 /*@
5852    MatIncreaseOverlap - Given a set of submatrices indicated by index sets,
5853    replaces the index sets by larger ones that represent submatrices with
5854    additional overlap.
5855 
5856    Collective on Mat
5857 
5858    Input Parameters:
5859 +  mat - the matrix
5860 .  n   - the number of index sets
5861 .  is  - the array of index sets (these index sets will changed during the call)
5862 -  ov  - the additional overlap requested
5863 
5864    Level: developer
5865 
5866    Concepts: overlap
5867    Concepts: ASM^computing overlap
5868 
5869 .seealso: MatGetSubMatrices()
5870 @*/
5871 PetscErrorCode PETSCMAT_DLLEXPORT MatIncreaseOverlap(Mat mat,PetscInt n,IS is[],PetscInt ov)
5872 {
5873   PetscErrorCode ierr;
5874 
5875   PetscFunctionBegin;
5876   PetscValidHeaderSpecific(mat,MAT_COOKIE,1);
5877   PetscValidType(mat,1);
5878   if (n < 0) SETERRQ1(PETSC_ERR_ARG_OUTOFRANGE,"Must have one or more domains, you have %D",n);
5879   if (n) {
5880     PetscValidPointer(is,3);
5881     PetscValidHeaderSpecific(*is,IS_COOKIE,3);
5882   }
5883   if (!mat->assembled) SETERRQ(PETSC_ERR_ARG_WRONGSTATE,"Not for unassembled matrix");
5884   if (mat->factor)     SETERRQ(PETSC_ERR_ARG_WRONGSTATE,"Not for factored matrix");
5885   ierr = MatPreallocated(mat);CHKERRQ(ierr);
5886 
5887   if (!ov) PetscFunctionReturn(0);
5888   if (!mat->ops->increaseoverlap) SETERRQ1(PETSC_ERR_SUP,"Mat type %s",((PetscObject)mat)->type_name);
5889   ierr = PetscLogEventBegin(MAT_IncreaseOverlap,mat,0,0,0);CHKERRQ(ierr);
5890   ierr = (*mat->ops->increaseoverlap)(mat,n,is,ov);CHKERRQ(ierr);
5891   ierr = PetscLogEventEnd(MAT_IncreaseOverlap,mat,0,0,0);CHKERRQ(ierr);
5892   PetscFunctionReturn(0);
5893 }
5894 
5895 #undef __FUNCT__
5896 #define __FUNCT__ "MatGetBlockSize"
5897 /*@
5898    MatGetBlockSize - Returns the matrix block size; useful especially for the
5899    block row and block diagonal formats.
5900 
5901    Not Collective
5902 
5903    Input Parameter:
5904 .  mat - the matrix
5905 
5906    Output Parameter:
5907 .  bs - block size
5908 
5909    Notes:
5910    Block row formats are MATSEQBAIJ, MATMPIBAIJ, MATSEQSBAIJ, MATMPISBAIJ
5911 
5912    Level: intermediate
5913 
5914    Concepts: matrices^block size
5915 
5916 .seealso: MatCreateSeqBAIJ(), MatCreateMPIBAIJ()
5917 @*/
5918 PetscErrorCode PETSCMAT_DLLEXPORT MatGetBlockSize(Mat mat,PetscInt *bs)
5919 {
5920   PetscErrorCode ierr;
5921 
5922   PetscFunctionBegin;
5923   PetscValidHeaderSpecific(mat,MAT_COOKIE,1);
5924   PetscValidType(mat,1);
5925   PetscValidIntPointer(bs,2);
5926   ierr = MatPreallocated(mat);CHKERRQ(ierr);
5927   *bs = mat->rmap->bs;
5928   PetscFunctionReturn(0);
5929 }
5930 
5931 #undef __FUNCT__
5932 #define __FUNCT__ "MatSetBlockSize"
5933 /*@
5934    MatSetBlockSize - Sets the matrix block size; for many matrix types you
5935      cannot use this and MUST set the blocksize when you preallocate the matrix
5936 
5937    Collective on Mat
5938 
5939    Input Parameters:
5940 +  mat - the matrix
5941 -  bs - block size
5942 
5943    Notes:
5944      Only works for shell and AIJ matrices
5945 
5946    Level: intermediate
5947 
5948    Concepts: matrices^block size
5949 
5950 .seealso: MatCreateSeqBAIJ(), MatCreateMPIBAIJ(), MatGetBlockSize()
5951 @*/
5952 PetscErrorCode PETSCMAT_DLLEXPORT MatSetBlockSize(Mat mat,PetscInt bs)
5953 {
5954   PetscErrorCode ierr;
5955 
5956   PetscFunctionBegin;
5957   PetscValidHeaderSpecific(mat,MAT_COOKIE,1);
5958   PetscValidType(mat,1);
5959   ierr = MatPreallocated(mat);CHKERRQ(ierr);
5960   if (mat->ops->setblocksize) {
5961     /* XXX should check if (bs < 1) ??? */
5962     ierr = PetscMapSetBlockSize(mat->rmap,bs);CHKERRQ(ierr);
5963     ierr = PetscMapSetBlockSize(mat->cmap,bs);CHKERRQ(ierr);
5964     ierr = (*mat->ops->setblocksize)(mat,bs);CHKERRQ(ierr);
5965   } else {
5966     SETERRQ1(PETSC_ERR_ARG_INCOMP,"Cannot set the blocksize for matrix type %s",((PetscObject)mat)->type_name);
5967   }
5968   PetscFunctionReturn(0);
5969 }
5970 
5971 #undef __FUNCT__
5972 #define __FUNCT__ "MatGetRowIJ"
5973 /*@C
5974     MatGetRowIJ - Returns the compressed row storage i and j indices for sequential matrices.
5975 
5976    Collective on Mat
5977 
5978     Input Parameters:
5979 +   mat - the matrix
5980 .   shift -  0 or 1 indicating we want the indices starting at 0 or 1
5981 .   symmetric - PETSC_TRUE or PETSC_FALSE indicating the matrix data structure should be
5982                 symmetrized
5983 -   inodecompressed - PETSC_TRUE or PETSC_FALSE  indicating if the nonzero structure of the
5984                  inodes or the nonzero elements is wanted. For BAIJ matrices the compressed version is
5985                  always used.
5986 
5987     Output Parameters:
5988 +   n - number of rows in the (possibly compressed) matrix
5989 .   ia - the row pointers [of length n+1]
5990 .   ja - the column indices
5991 -   done - indicates if the routine actually worked and returned appropriate ia[] and ja[] arrays; callers
5992            are responsible for handling the case when done == PETSC_FALSE and ia and ja are not set
5993 
5994     Level: developer
5995 
5996     Notes: You CANNOT change any of the ia[] or ja[] values.
5997 
5998            Use MatRestoreRowIJ() when you are finished accessing the ia[] and ja[] values
5999 
6000     Fortran Node
6001 
6002            In Fortran use
6003 $           PetscInt ia(1), ja(1)
6004 $           PetscOffset iia, jja
6005 $      call MatGetRowIJ(mat,shift,symmetric,inodecompressed,n,ia,iia,ja,jja,done,ierr)
6006 $
6007 $          or
6008 $
6009 $           PetscScalar, pointer :: xx_v(:)
6010 $    call  MatGetRowIJF90(mat,shift,symmetric,inodecompressed,n,ia,ja,done,ierr)
6011 
6012 
6013        Acess the ith and jth entries via ia(iia + i) and ja(jja + j)
6014 
6015 .seealso: MatGetColumnIJ(), MatRestoreRowIJ(), MatGetArray()
6016 @*/
6017 PetscErrorCode PETSCMAT_DLLEXPORT MatGetRowIJ(Mat mat,PetscInt shift,PetscTruth symmetric,PetscTruth inodecompressed,PetscInt *n,PetscInt *ia[],PetscInt* ja[],PetscTruth *done)
6018 {
6019   PetscErrorCode ierr;
6020 
6021   PetscFunctionBegin;
6022   PetscValidHeaderSpecific(mat,MAT_COOKIE,1);
6023   PetscValidType(mat,1);
6024   PetscValidIntPointer(n,4);
6025   if (ia) PetscValidIntPointer(ia,5);
6026   if (ja) PetscValidIntPointer(ja,6);
6027   PetscValidIntPointer(done,7);
6028   ierr = MatPreallocated(mat);CHKERRQ(ierr);
6029   if (!mat->ops->getrowij) *done = PETSC_FALSE;
6030   else {
6031     *done = PETSC_TRUE;
6032     ierr = PetscLogEventBegin(MAT_GetRowIJ,mat,0,0,0);CHKERRQ(ierr);
6033     ierr  = (*mat->ops->getrowij)(mat,shift,symmetric,inodecompressed,n,ia,ja,done);CHKERRQ(ierr);
6034     ierr = PetscLogEventEnd(MAT_GetRowIJ,mat,0,0,0);CHKERRQ(ierr);
6035   }
6036   PetscFunctionReturn(0);
6037 }
6038 
6039 #undef __FUNCT__
6040 #define __FUNCT__ "MatGetColumnIJ"
6041 /*@C
6042     MatGetColumnIJ - Returns the compressed column storage i and j indices for sequential matrices.
6043 
6044     Collective on Mat
6045 
6046     Input Parameters:
6047 +   mat - the matrix
6048 .   shift - 1 or zero indicating we want the indices starting at 0 or 1
6049 .   symmetric - PETSC_TRUE or PETSC_FALSE indicating the matrix data structure should be
6050                 symmetrized
6051 -   inodecompressed - PETSC_TRUE or PETSC_FALSE indicating if the nonzero structure of the
6052                  inodes or the nonzero elements is wanted. For BAIJ matrices the compressed version is
6053                  always used.
6054 
6055     Output Parameters:
6056 +   n - number of columns in the (possibly compressed) matrix
6057 .   ia - the column pointers
6058 .   ja - the row indices
6059 -   done - PETSC_TRUE or PETSC_FALSE, indicating whether the values have been returned
6060 
6061     Level: developer
6062 
6063 .seealso: MatGetRowIJ(), MatRestoreColumnIJ()
6064 @*/
6065 PetscErrorCode PETSCMAT_DLLEXPORT MatGetColumnIJ(Mat mat,PetscInt shift,PetscTruth symmetric,PetscTruth inodecompressed,PetscInt *n,PetscInt *ia[],PetscInt* ja[],PetscTruth *done)
6066 {
6067   PetscErrorCode ierr;
6068 
6069   PetscFunctionBegin;
6070   PetscValidHeaderSpecific(mat,MAT_COOKIE,1);
6071   PetscValidType(mat,1);
6072   PetscValidIntPointer(n,4);
6073   if (ia) PetscValidIntPointer(ia,5);
6074   if (ja) PetscValidIntPointer(ja,6);
6075   PetscValidIntPointer(done,7);
6076   ierr = MatPreallocated(mat);CHKERRQ(ierr);
6077   if (!mat->ops->getcolumnij) *done = PETSC_FALSE;
6078   else {
6079     *done = PETSC_TRUE;
6080     ierr  = (*mat->ops->getcolumnij)(mat,shift,symmetric,inodecompressed,n,ia,ja,done);CHKERRQ(ierr);
6081   }
6082   PetscFunctionReturn(0);
6083 }
6084 
6085 #undef __FUNCT__
6086 #define __FUNCT__ "MatRestoreRowIJ"
6087 /*@C
6088     MatRestoreRowIJ - Call after you are completed with the ia,ja indices obtained with
6089     MatGetRowIJ().
6090 
6091     Collective on Mat
6092 
6093     Input Parameters:
6094 +   mat - the matrix
6095 .   shift - 1 or zero indicating we want the indices starting at 0 or 1
6096 .   symmetric - PETSC_TRUE or PETSC_FALSE indicating the matrix data structure should be
6097                 symmetrized
6098 -   inodecompressed -  PETSC_TRUE or PETSC_FALSE indicating if the nonzero structure of the
6099                  inodes or the nonzero elements is wanted. For BAIJ matrices the compressed version is
6100                  always used.
6101 
6102     Output Parameters:
6103 +   n - size of (possibly compressed) matrix
6104 .   ia - the row pointers
6105 .   ja - the column indices
6106 -   done - PETSC_TRUE or PETSC_FALSE indicated that the values have been returned
6107 
6108     Level: developer
6109 
6110 .seealso: MatGetRowIJ(), MatRestoreColumnIJ()
6111 @*/
6112 PetscErrorCode PETSCMAT_DLLEXPORT MatRestoreRowIJ(Mat mat,PetscInt shift,PetscTruth symmetric,PetscTruth inodecompressed,PetscInt *n,PetscInt *ia[],PetscInt* ja[],PetscTruth *done)
6113 {
6114   PetscErrorCode ierr;
6115 
6116   PetscFunctionBegin;
6117   PetscValidHeaderSpecific(mat,MAT_COOKIE,1);
6118   PetscValidType(mat,1);
6119   if (ia) PetscValidIntPointer(ia,5);
6120   if (ja) PetscValidIntPointer(ja,6);
6121   PetscValidIntPointer(done,7);
6122   ierr = MatPreallocated(mat);CHKERRQ(ierr);
6123 
6124   if (!mat->ops->restorerowij) *done = PETSC_FALSE;
6125   else {
6126     *done = PETSC_TRUE;
6127     ierr  = (*mat->ops->restorerowij)(mat,shift,symmetric,inodecompressed,n,ia,ja,done);CHKERRQ(ierr);
6128   }
6129   PetscFunctionReturn(0);
6130 }
6131 
6132 #undef __FUNCT__
6133 #define __FUNCT__ "MatRestoreColumnIJ"
6134 /*@C
6135     MatRestoreColumnIJ - Call after you are completed with the ia,ja indices obtained with
6136     MatGetColumnIJ().
6137 
6138     Collective on Mat
6139 
6140     Input Parameters:
6141 +   mat - the matrix
6142 .   shift - 1 or zero indicating we want the indices starting at 0 or 1
6143 -   symmetric - PETSC_TRUE or PETSC_FALSE indicating the matrix data structure should be
6144                 symmetrized
6145 -   inodecompressed - PETSC_TRUE or PETSC_FALSE indicating if the nonzero structure of the
6146                  inodes or the nonzero elements is wanted. For BAIJ matrices the compressed version is
6147                  always used.
6148 
6149     Output Parameters:
6150 +   n - size of (possibly compressed) matrix
6151 .   ia - the column pointers
6152 .   ja - the row indices
6153 -   done - PETSC_TRUE or PETSC_FALSE indicated that the values have been returned
6154 
6155     Level: developer
6156 
6157 .seealso: MatGetColumnIJ(), MatRestoreRowIJ()
6158 @*/
6159 PetscErrorCode PETSCMAT_DLLEXPORT MatRestoreColumnIJ(Mat mat,PetscInt shift,PetscTruth symmetric,PetscTruth inodecompressed,PetscInt *n,PetscInt *ia[],PetscInt* ja[],PetscTruth *done)
6160 {
6161   PetscErrorCode ierr;
6162 
6163   PetscFunctionBegin;
6164   PetscValidHeaderSpecific(mat,MAT_COOKIE,1);
6165   PetscValidType(mat,1);
6166   if (ia) PetscValidIntPointer(ia,5);
6167   if (ja) PetscValidIntPointer(ja,6);
6168   PetscValidIntPointer(done,7);
6169   ierr = MatPreallocated(mat);CHKERRQ(ierr);
6170 
6171   if (!mat->ops->restorecolumnij) *done = PETSC_FALSE;
6172   else {
6173     *done = PETSC_TRUE;
6174     ierr  = (*mat->ops->restorecolumnij)(mat,shift,symmetric,inodecompressed,n,ia,ja,done);CHKERRQ(ierr);
6175   }
6176   PetscFunctionReturn(0);
6177 }
6178 
6179 #undef __FUNCT__
6180 #define __FUNCT__ "MatColoringPatch"
6181 /*@C
6182     MatColoringPatch -Used inside matrix coloring routines that
6183     use MatGetRowIJ() and/or MatGetColumnIJ().
6184 
6185     Collective on Mat
6186 
6187     Input Parameters:
6188 +   mat - the matrix
6189 .   ncolors - max color value
6190 .   n   - number of entries in colorarray
6191 -   colorarray - array indicating color for each column
6192 
6193     Output Parameters:
6194 .   iscoloring - coloring generated using colorarray information
6195 
6196     Level: developer
6197 
6198 .seealso: MatGetRowIJ(), MatGetColumnIJ()
6199 
6200 @*/
6201 PetscErrorCode PETSCMAT_DLLEXPORT MatColoringPatch(Mat mat,PetscInt ncolors,PetscInt n,ISColoringValue colorarray[],ISColoring *iscoloring)
6202 {
6203   PetscErrorCode ierr;
6204 
6205   PetscFunctionBegin;
6206   PetscValidHeaderSpecific(mat,MAT_COOKIE,1);
6207   PetscValidType(mat,1);
6208   PetscValidIntPointer(colorarray,4);
6209   PetscValidPointer(iscoloring,5);
6210   ierr = MatPreallocated(mat);CHKERRQ(ierr);
6211 
6212   if (!mat->ops->coloringpatch){
6213     ierr = ISColoringCreate(((PetscObject)mat)->comm,ncolors,n,colorarray,iscoloring);CHKERRQ(ierr);
6214   } else {
6215     ierr = (*mat->ops->coloringpatch)(mat,ncolors,n,colorarray,iscoloring);CHKERRQ(ierr);
6216   }
6217   PetscFunctionReturn(0);
6218 }
6219 
6220 
6221 #undef __FUNCT__
6222 #define __FUNCT__ "MatSetUnfactored"
6223 /*@
6224    MatSetUnfactored - Resets a factored matrix to be treated as unfactored.
6225 
6226    Collective on Mat
6227 
6228    Input Parameter:
6229 .  mat - the factored matrix to be reset
6230 
6231    Notes:
6232    This routine should be used only with factored matrices formed by in-place
6233    factorization via ILU(0) (or by in-place LU factorization for the MATSEQDENSE
6234    format).  This option can save memory, for example, when solving nonlinear
6235    systems with a matrix-free Newton-Krylov method and a matrix-based, in-place
6236    ILU(0) preconditioner.
6237 
6238    Note that one can specify in-place ILU(0) factorization by calling
6239 .vb
6240      PCType(pc,PCILU);
6241      PCFactorSeUseInPlace(pc);
6242 .ve
6243    or by using the options -pc_type ilu -pc_factor_in_place
6244 
6245    In-place factorization ILU(0) can also be used as a local
6246    solver for the blocks within the block Jacobi or additive Schwarz
6247    methods (runtime option: -sub_pc_factor_in_place).  See the discussion
6248    of these preconditioners in the users manual for details on setting
6249    local solver options.
6250 
6251    Most users should employ the simplified KSP interface for linear solvers
6252    instead of working directly with matrix algebra routines such as this.
6253    See, e.g., KSPCreate().
6254 
6255    Level: developer
6256 
6257 .seealso: PCFactorSetUseInPlace()
6258 
6259    Concepts: matrices^unfactored
6260 
6261 @*/
6262 PetscErrorCode PETSCMAT_DLLEXPORT MatSetUnfactored(Mat mat)
6263 {
6264   PetscErrorCode ierr;
6265 
6266   PetscFunctionBegin;
6267   PetscValidHeaderSpecific(mat,MAT_COOKIE,1);
6268   PetscValidType(mat,1);
6269   ierr = MatPreallocated(mat);CHKERRQ(ierr);
6270   mat->factor = MAT_FACTOR_NONE;
6271   if (!mat->ops->setunfactored) PetscFunctionReturn(0);
6272   ierr = (*mat->ops->setunfactored)(mat);CHKERRQ(ierr);
6273   PetscFunctionReturn(0);
6274 }
6275 
6276 /*MC
6277     MatGetArrayF90 - Accesses a matrix array from Fortran90.
6278 
6279     Synopsis:
6280     MatGetArrayF90(Mat x,{Scalar, pointer :: xx_v(:)},integer ierr)
6281 
6282     Not collective
6283 
6284     Input Parameter:
6285 .   x - matrix
6286 
6287     Output Parameters:
6288 +   xx_v - the Fortran90 pointer to the array
6289 -   ierr - error code
6290 
6291     Example of Usage:
6292 .vb
6293       PetscScalar, pointer xx_v(:)
6294       ....
6295       call MatGetArrayF90(x,xx_v,ierr)
6296       a = xx_v(3)
6297       call MatRestoreArrayF90(x,xx_v,ierr)
6298 .ve
6299 
6300     Notes:
6301     Not yet supported for all F90 compilers
6302 
6303     Level: advanced
6304 
6305 .seealso:  MatRestoreArrayF90(), MatGetArray(), MatRestoreArray()
6306 
6307     Concepts: matrices^accessing array
6308 
6309 M*/
6310 
6311 /*MC
6312     MatRestoreArrayF90 - Restores a matrix array that has been
6313     accessed with MatGetArrayF90().
6314 
6315     Synopsis:
6316     MatRestoreArrayF90(Mat x,{Scalar, pointer :: xx_v(:)},integer ierr)
6317 
6318     Not collective
6319 
6320     Input Parameters:
6321 +   x - matrix
6322 -   xx_v - the Fortran90 pointer to the array
6323 
6324     Output Parameter:
6325 .   ierr - error code
6326 
6327     Example of Usage:
6328 .vb
6329        PetscScalar, pointer xx_v(:)
6330        ....
6331        call MatGetArrayF90(x,xx_v,ierr)
6332        a = xx_v(3)
6333        call MatRestoreArrayF90(x,xx_v,ierr)
6334 .ve
6335 
6336     Notes:
6337     Not yet supported for all F90 compilers
6338 
6339     Level: advanced
6340 
6341 .seealso:  MatGetArrayF90(), MatGetArray(), MatRestoreArray()
6342 
6343 M*/
6344 
6345 
6346 #undef __FUNCT__
6347 #define __FUNCT__ "MatGetSubMatrix"
6348 /*@
6349     MatGetSubMatrix - Gets a single submatrix on the same number of processors
6350                       as the original matrix.
6351 
6352     Collective on Mat
6353 
6354     Input Parameters:
6355 +   mat - the original matrix
6356 .   isrow - parallel IS containing the rows this processor should obtain
6357 .   iscol - parallel IS containing all columns you wish to keep
6358 -   cll - either MAT_INITIAL_MATRIX or MAT_REUSE_MATRIX
6359 
6360     Output Parameter:
6361 .   newmat - the new submatrix, of the same type as the old
6362 
6363     Level: advanced
6364 
6365     Notes:
6366     The submatrix will be able to be multiplied with vectors using the same layout as iscol.
6367 
6368     The rows is isrow will be sorted into the same order as the original matrix.
6369 
6370       The first time this is called you should use a cll of MAT_INITIAL_MATRIX,
6371    the MatGetSubMatrix() routine will create the newmat for you. Any additional calls
6372    to this routine with a mat of the same nonzero structure and with a call of MAT_REUSE_MATRIX
6373    will reuse the matrix generated the first time.  You should call MatDestroy() on newmat when
6374    you are finished using it.
6375 
6376     The communicator of the newly obtained matrix is ALWAYS the same as the communicator of
6377     the input matrix.
6378 
6379     If iscol is PETSC_NULL then all columns are obtained (not supported in Fortran).
6380 
6381    Example usage:
6382    Consider the following 8x8 matrix with 34 non-zero values, that is
6383    assembled across 3 processors. Lets assume that proc0 owns 3 rows,
6384    proc1 owns 3 rows, proc2 owns 2 rows. This division can be shown
6385    as follows:
6386 
6387 .vb
6388             1  2  0  |  0  3  0  |  0  4
6389     Proc0   0  5  6  |  7  0  0  |  8  0
6390             9  0 10  | 11  0  0  | 12  0
6391     -------------------------------------
6392            13  0 14  | 15 16 17  |  0  0
6393     Proc1   0 18  0  | 19 20 21  |  0  0
6394             0  0  0  | 22 23  0  | 24  0
6395     -------------------------------------
6396     Proc2  25 26 27  |  0  0 28  | 29  0
6397            30  0  0  | 31 32 33  |  0 34
6398 .ve
6399 
6400     Suppose isrow = [0 1 | 4 | 5 6] and iscol = [1 2 | 3 4 5 | 6].  The resulting submatrix is
6401 
6402 .vb
6403             2  0  |  0  3  0  |  0
6404     Proc0   5  6  |  7  0  0  |  8
6405     -------------------------------
6406     Proc1  18  0  | 19 20 21  |  0
6407     -------------------------------
6408     Proc2  26 27  |  0  0 28  | 29
6409             0  0  | 31 32 33  |  0
6410 .ve
6411 
6412 
6413     Concepts: matrices^submatrices
6414 
6415 .seealso: MatGetSubMatrices()
6416 @*/
6417 PetscErrorCode PETSCMAT_DLLEXPORT MatGetSubMatrix(Mat mat,IS isrow,IS iscol,MatReuse cll,Mat *newmat)
6418 {
6419   PetscErrorCode ierr;
6420   PetscMPIInt    size;
6421   Mat            *local;
6422   IS             iscoltmp;
6423 
6424   PetscFunctionBegin;
6425   PetscValidHeaderSpecific(mat,MAT_COOKIE,1);
6426   PetscValidHeaderSpecific(isrow,IS_COOKIE,2);
6427   if (iscol) PetscValidHeaderSpecific(iscol,IS_COOKIE,3);
6428   PetscValidPointer(newmat,6);
6429   if (cll == MAT_REUSE_MATRIX) PetscValidHeaderSpecific(*newmat,MAT_COOKIE,6);
6430   PetscValidType(mat,1);
6431   if (mat->factor) SETERRQ(PETSC_ERR_ARG_WRONGSTATE,"Not for factored matrix");
6432   ierr = MatPreallocated(mat);CHKERRQ(ierr);
6433   ierr = MPI_Comm_size(((PetscObject)mat)->comm,&size);CHKERRQ(ierr);
6434 
6435   if (!iscol) {
6436     ierr = ISCreateStride(((PetscObject)mat)->comm,mat->cmap->n,mat->cmap->rstart,1,&iscoltmp);CHKERRQ(ierr);
6437   } else {
6438     iscoltmp = iscol;
6439   }
6440 
6441   /* if original matrix is on just one processor then use submatrix generated */
6442   if (mat->ops->getsubmatrices && !mat->ops->getsubmatrix && size == 1 && cll == MAT_REUSE_MATRIX) {
6443     ierr = MatGetSubMatrices(mat,1,&isrow,&iscoltmp,MAT_REUSE_MATRIX,&newmat);CHKERRQ(ierr);
6444     if (!iscol) {ierr = ISDestroy(iscoltmp);CHKERRQ(ierr);}
6445     PetscFunctionReturn(0);
6446   } else if (mat->ops->getsubmatrices && !mat->ops->getsubmatrix && size == 1) {
6447     ierr    = MatGetSubMatrices(mat,1,&isrow,&iscoltmp,MAT_INITIAL_MATRIX,&local);CHKERRQ(ierr);
6448     *newmat = *local;
6449     ierr    = PetscFree(local);CHKERRQ(ierr);
6450     if (!iscol) {ierr = ISDestroy(iscoltmp);CHKERRQ(ierr);}
6451     PetscFunctionReturn(0);
6452   } else if (!mat->ops->getsubmatrix) {
6453     /* Create a new matrix type that implements the operation using the full matrix */
6454     switch (cll) {
6455       case MAT_INITIAL_MATRIX:
6456         ierr = MatCreateSubMatrix(mat,isrow,iscoltmp,newmat);CHKERRQ(ierr);
6457         break;
6458       case MAT_REUSE_MATRIX:
6459         ierr = MatSubMatrixUpdate(*newmat,mat,isrow,iscoltmp);CHKERRQ(ierr);
6460         break;
6461       default: SETERRQ(PETSC_ERR_ARG_OUTOFRANGE,"Invalid MatReuse, must be either MAT_INITIAL_MATRIX or MAT_REUSE_MATRIX");
6462     }
6463     if (!iscol) {ierr = ISDestroy(iscoltmp);CHKERRQ(ierr);}
6464     PetscFunctionReturn(0);
6465   }
6466 
6467   if (!mat->ops->getsubmatrix) SETERRQ1(PETSC_ERR_SUP,"Mat type %s",((PetscObject)mat)->type_name);
6468   ierr = (*mat->ops->getsubmatrix)(mat,isrow,iscoltmp,cll,newmat);CHKERRQ(ierr);
6469   if (!iscol) {ierr = ISDestroy(iscoltmp);CHKERRQ(ierr);}
6470   ierr = PetscObjectStateIncrease((PetscObject)*newmat);CHKERRQ(ierr);
6471   PetscFunctionReturn(0);
6472 }
6473 
6474 #undef __FUNCT__
6475 #define __FUNCT__ "MatStashSetInitialSize"
6476 /*@
6477    MatStashSetInitialSize - sets the sizes of the matrix stash, that is
6478    used during the assembly process to store values that belong to
6479    other processors.
6480 
6481    Not Collective
6482 
6483    Input Parameters:
6484 +  mat   - the matrix
6485 .  size  - the initial size of the stash.
6486 -  bsize - the initial size of the block-stash(if used).
6487 
6488    Options Database Keys:
6489 +   -matstash_initial_size <size> or <size0,size1,...sizep-1>
6490 -   -matstash_block_initial_size <bsize>  or <bsize0,bsize1,...bsizep-1>
6491 
6492    Level: intermediate
6493 
6494    Notes:
6495      The block-stash is used for values set with MatSetValuesBlocked() while
6496      the stash is used for values set with MatSetValues()
6497 
6498      Run with the option -info and look for output of the form
6499      MatAssemblyBegin_MPIXXX:Stash has MM entries, uses nn mallocs.
6500      to determine the appropriate value, MM, to use for size and
6501      MatAssemblyBegin_MPIXXX:Block-Stash has BMM entries, uses nn mallocs.
6502      to determine the value, BMM to use for bsize
6503 
6504    Concepts: stash^setting matrix size
6505    Concepts: matrices^stash
6506 
6507 .seealso: MatAssemblyBegin(), MatAssemblyEnd(), Mat, MatStashGetInfo()
6508 
6509 @*/
6510 PetscErrorCode PETSCMAT_DLLEXPORT MatStashSetInitialSize(Mat mat,PetscInt size, PetscInt bsize)
6511 {
6512   PetscErrorCode ierr;
6513 
6514   PetscFunctionBegin;
6515   PetscValidHeaderSpecific(mat,MAT_COOKIE,1);
6516   PetscValidType(mat,1);
6517   ierr = MatStashSetInitialSize_Private(&mat->stash,size);CHKERRQ(ierr);
6518   ierr = MatStashSetInitialSize_Private(&mat->bstash,bsize);CHKERRQ(ierr);
6519   PetscFunctionReturn(0);
6520 }
6521 
6522 #undef __FUNCT__
6523 #define __FUNCT__ "MatInterpolateAdd"
6524 /*@
6525    MatInterpolateAdd - w = y + A*x or A'*x depending on the shape of
6526      the matrix
6527 
6528    Collective on Mat
6529 
6530    Input Parameters:
6531 +  mat   - the matrix
6532 .  x,y - the vectors
6533 -  w - where the result is stored
6534 
6535    Level: intermediate
6536 
6537    Notes:
6538     w may be the same vector as y.
6539 
6540     This allows one to use either the restriction or interpolation (its transpose)
6541     matrix to do the interpolation
6542 
6543     Concepts: interpolation
6544 
6545 .seealso: MatMultAdd(), MatMultTransposeAdd(), MatRestrict()
6546 
6547 @*/
6548 PetscErrorCode PETSCMAT_DLLEXPORT MatInterpolateAdd(Mat A,Vec x,Vec y,Vec w)
6549 {
6550   PetscErrorCode ierr;
6551   PetscInt       M,N;
6552 
6553   PetscFunctionBegin;
6554   PetscValidHeaderSpecific(A,MAT_COOKIE,1);
6555   PetscValidHeaderSpecific(x,VEC_COOKIE,2);
6556   PetscValidHeaderSpecific(y,VEC_COOKIE,3);
6557   PetscValidHeaderSpecific(w,VEC_COOKIE,4);
6558   PetscValidType(A,1);
6559   ierr = MatPreallocated(A);CHKERRQ(ierr);
6560   ierr = MatGetSize(A,&M,&N);CHKERRQ(ierr);
6561   if (N > M) {
6562     ierr = MatMultTransposeAdd(A,x,y,w);CHKERRQ(ierr);
6563   } else {
6564     ierr = MatMultAdd(A,x,y,w);CHKERRQ(ierr);
6565   }
6566   PetscFunctionReturn(0);
6567 }
6568 
6569 #undef __FUNCT__
6570 #define __FUNCT__ "MatInterpolate"
6571 /*@
6572    MatInterpolate - y = A*x or A'*x depending on the shape of
6573      the matrix
6574 
6575    Collective on Mat
6576 
6577    Input Parameters:
6578 +  mat   - the matrix
6579 -  x,y - the vectors
6580 
6581    Level: intermediate
6582 
6583    Notes:
6584     This allows one to use either the restriction or interpolation (its transpose)
6585     matrix to do the interpolation
6586 
6587    Concepts: matrices^interpolation
6588 
6589 .seealso: MatMultAdd(), MatMultTransposeAdd(), MatRestrict()
6590 
6591 @*/
6592 PetscErrorCode PETSCMAT_DLLEXPORT MatInterpolate(Mat A,Vec x,Vec y)
6593 {
6594   PetscErrorCode ierr;
6595   PetscInt       M,N;
6596 
6597   PetscFunctionBegin;
6598   PetscValidHeaderSpecific(A,MAT_COOKIE,1);
6599   PetscValidHeaderSpecific(x,VEC_COOKIE,2);
6600   PetscValidHeaderSpecific(y,VEC_COOKIE,3);
6601   PetscValidType(A,1);
6602   ierr = MatPreallocated(A);CHKERRQ(ierr);
6603   ierr = MatGetSize(A,&M,&N);CHKERRQ(ierr);
6604   if (N > M) {
6605     ierr = MatMultTranspose(A,x,y);CHKERRQ(ierr);
6606   } else {
6607     ierr = MatMult(A,x,y);CHKERRQ(ierr);
6608   }
6609   PetscFunctionReturn(0);
6610 }
6611 
6612 #undef __FUNCT__
6613 #define __FUNCT__ "MatRestrict"
6614 /*@
6615    MatRestrict - y = A*x or A'*x
6616 
6617    Collective on Mat
6618 
6619    Input Parameters:
6620 +  mat   - the matrix
6621 -  x,y - the vectors
6622 
6623    Level: intermediate
6624 
6625    Notes:
6626     This allows one to use either the restriction or interpolation (its transpose)
6627     matrix to do the restriction
6628 
6629    Concepts: matrices^restriction
6630 
6631 .seealso: MatMultAdd(), MatMultTransposeAdd(), MatInterpolate()
6632 
6633 @*/
6634 PetscErrorCode PETSCMAT_DLLEXPORT MatRestrict(Mat A,Vec x,Vec y)
6635 {
6636   PetscErrorCode ierr;
6637   PetscInt       M,N;
6638 
6639   PetscFunctionBegin;
6640   PetscValidHeaderSpecific(A,MAT_COOKIE,1);
6641   PetscValidHeaderSpecific(x,VEC_COOKIE,2);
6642   PetscValidHeaderSpecific(y,VEC_COOKIE,3);
6643   PetscValidType(A,1);
6644   ierr = MatPreallocated(A);CHKERRQ(ierr);
6645 
6646   ierr = MatGetSize(A,&M,&N);CHKERRQ(ierr);
6647   if (N > M) {
6648     ierr = MatMult(A,x,y);CHKERRQ(ierr);
6649   } else {
6650     ierr = MatMultTranspose(A,x,y);CHKERRQ(ierr);
6651   }
6652   PetscFunctionReturn(0);
6653 }
6654 
6655 #undef __FUNCT__
6656 #define __FUNCT__ "MatNullSpaceAttach"
6657 /*@
6658    MatNullSpaceAttach - attaches a null space to a matrix.
6659         This null space will be removed from the resulting vector whenever
6660         MatMult() is called
6661 
6662    Collective on Mat
6663 
6664    Input Parameters:
6665 +  mat - the matrix
6666 -  nullsp - the null space object
6667 
6668    Level: developer
6669 
6670    Notes:
6671       Overwrites any previous null space that may have been attached
6672 
6673    Concepts: null space^attaching to matrix
6674 
6675 .seealso: MatCreate(), MatNullSpaceCreate()
6676 @*/
6677 PetscErrorCode PETSCMAT_DLLEXPORT MatNullSpaceAttach(Mat mat,MatNullSpace nullsp)
6678 {
6679   PetscErrorCode ierr;
6680 
6681   PetscFunctionBegin;
6682   PetscValidHeaderSpecific(mat,MAT_COOKIE,1);
6683   PetscValidType(mat,1);
6684   PetscValidHeaderSpecific(nullsp,MAT_NULLSPACE_COOKIE,2);
6685   ierr = MatPreallocated(mat);CHKERRQ(ierr);
6686   ierr = PetscObjectReference((PetscObject)nullsp);CHKERRQ(ierr);
6687   if (mat->nullsp) { ierr = MatNullSpaceDestroy(mat->nullsp);CHKERRQ(ierr); }
6688   mat->nullsp = nullsp;
6689   PetscFunctionReturn(0);
6690 }
6691 
6692 #undef __FUNCT__
6693 #define __FUNCT__ "MatICCFactor"
6694 /*@C
6695    MatICCFactor - Performs in-place incomplete Cholesky factorization of matrix.
6696 
6697    Collective on Mat
6698 
6699    Input Parameters:
6700 +  mat - the matrix
6701 .  row - row/column permutation
6702 .  fill - expected fill factor >= 1.0
6703 -  level - level of fill, for ICC(k)
6704 
6705    Notes:
6706    Probably really in-place only when level of fill is zero, otherwise allocates
6707    new space to store factored matrix and deletes previous memory.
6708 
6709    Most users should employ the simplified KSP interface for linear solvers
6710    instead of working directly with matrix algebra routines such as this.
6711    See, e.g., KSPCreate().
6712 
6713    Level: developer
6714 
6715    Concepts: matrices^incomplete Cholesky factorization
6716    Concepts: Cholesky factorization
6717 
6718 .seealso: MatICCFactorSymbolic(), MatLUFactorNumeric(), MatCholeskyFactor()
6719 
6720     Developer Note: fortran interface is not autogenerated as the f90
6721     interface defintion cannot be generated correctly [due to MatFactorInfo]
6722 
6723 @*/
6724 PetscErrorCode PETSCMAT_DLLEXPORT MatICCFactor(Mat mat,IS row,const MatFactorInfo* info)
6725 {
6726   PetscErrorCode ierr;
6727 
6728   PetscFunctionBegin;
6729   PetscValidHeaderSpecific(mat,MAT_COOKIE,1);
6730   PetscValidType(mat,1);
6731   if (row) PetscValidHeaderSpecific(row,IS_COOKIE,2);
6732   PetscValidPointer(info,3);
6733   if (mat->rmap->N != mat->cmap->N) SETERRQ(PETSC_ERR_ARG_WRONG,"matrix must be square");
6734   if (!mat->assembled) SETERRQ(PETSC_ERR_ARG_WRONGSTATE,"Not for unassembled matrix");
6735   if (mat->factor) SETERRQ(PETSC_ERR_ARG_WRONGSTATE,"Not for factored matrix");
6736   if (!mat->ops->iccfactor) SETERRQ1(PETSC_ERR_SUP,"Mat type %s",((PetscObject)mat)->type_name);
6737   ierr = MatPreallocated(mat);CHKERRQ(ierr);
6738   ierr = (*mat->ops->iccfactor)(mat,row,info);CHKERRQ(ierr);
6739   ierr = PetscObjectStateIncrease((PetscObject)mat);CHKERRQ(ierr);
6740   PetscFunctionReturn(0);
6741 }
6742 
6743 #undef __FUNCT__
6744 #define __FUNCT__ "MatSetValuesAdic"
6745 /*@
6746    MatSetValuesAdic - Sets values computed with ADIC automatic differentiation into a matrix.
6747 
6748    Not Collective
6749 
6750    Input Parameters:
6751 +  mat - the matrix
6752 -  v - the values compute with ADIC
6753 
6754    Level: developer
6755 
6756    Notes:
6757      Must call MatSetColoring() before using this routine. Also this matrix must already
6758      have its nonzero pattern determined.
6759 
6760 .seealso: MatSetOption(), MatAssemblyBegin(), MatAssemblyEnd(), MatSetValuesBlocked(), MatSetValuesLocal(),
6761           MatSetValues(), MatSetColoring(), MatSetValuesAdifor()
6762 @*/
6763 PetscErrorCode PETSCMAT_DLLEXPORT MatSetValuesAdic(Mat mat,void *v)
6764 {
6765   PetscErrorCode ierr;
6766 
6767   PetscFunctionBegin;
6768   PetscValidHeaderSpecific(mat,MAT_COOKIE,1);
6769   PetscValidType(mat,1);
6770   PetscValidPointer(mat,2);
6771 
6772   if (!mat->assembled) {
6773     SETERRQ(PETSC_ERR_ARG_WRONGSTATE,"Matrix must be already assembled");
6774   }
6775   ierr = PetscLogEventBegin(MAT_SetValues,mat,0,0,0);CHKERRQ(ierr);
6776   if (!mat->ops->setvaluesadic) SETERRQ1(PETSC_ERR_SUP,"Mat type %s",((PetscObject)mat)->type_name);
6777   ierr = (*mat->ops->setvaluesadic)(mat,v);CHKERRQ(ierr);
6778   ierr = PetscLogEventEnd(MAT_SetValues,mat,0,0,0);CHKERRQ(ierr);
6779   ierr = MatView_Private(mat);CHKERRQ(ierr);
6780   ierr = PetscObjectStateIncrease((PetscObject)mat);CHKERRQ(ierr);
6781   PetscFunctionReturn(0);
6782 }
6783 
6784 
6785 #undef __FUNCT__
6786 #define __FUNCT__ "MatSetColoring"
6787 /*@
6788    MatSetColoring - Sets a coloring used by calls to MatSetValuesAdic()
6789 
6790    Not Collective
6791 
6792    Input Parameters:
6793 +  mat - the matrix
6794 -  coloring - the coloring
6795 
6796    Level: developer
6797 
6798 .seealso: MatSetOption(), MatAssemblyBegin(), MatAssemblyEnd(), MatSetValuesBlocked(), MatSetValuesLocal(),
6799           MatSetValues(), MatSetValuesAdic()
6800 @*/
6801 PetscErrorCode PETSCMAT_DLLEXPORT MatSetColoring(Mat mat,ISColoring coloring)
6802 {
6803   PetscErrorCode ierr;
6804 
6805   PetscFunctionBegin;
6806   PetscValidHeaderSpecific(mat,MAT_COOKIE,1);
6807   PetscValidType(mat,1);
6808   PetscValidPointer(coloring,2);
6809 
6810   if (!mat->assembled) {
6811     SETERRQ(PETSC_ERR_ARG_WRONGSTATE,"Matrix must be already assembled");
6812   }
6813   if (!mat->ops->setcoloring) SETERRQ1(PETSC_ERR_SUP,"Mat type %s",((PetscObject)mat)->type_name);
6814   ierr = (*mat->ops->setcoloring)(mat,coloring);CHKERRQ(ierr);
6815   PetscFunctionReturn(0);
6816 }
6817 
6818 #undef __FUNCT__
6819 #define __FUNCT__ "MatSetValuesAdifor"
6820 /*@
6821    MatSetValuesAdifor - Sets values computed with automatic differentiation into a matrix.
6822 
6823    Not Collective
6824 
6825    Input Parameters:
6826 +  mat - the matrix
6827 .  nl - leading dimension of v
6828 -  v - the values compute with ADIFOR
6829 
6830    Level: developer
6831 
6832    Notes:
6833      Must call MatSetColoring() before using this routine. Also this matrix must already
6834      have its nonzero pattern determined.
6835 
6836 .seealso: MatSetOption(), MatAssemblyBegin(), MatAssemblyEnd(), MatSetValuesBlocked(), MatSetValuesLocal(),
6837           MatSetValues(), MatSetColoring()
6838 @*/
6839 PetscErrorCode PETSCMAT_DLLEXPORT MatSetValuesAdifor(Mat mat,PetscInt nl,void *v)
6840 {
6841   PetscErrorCode ierr;
6842 
6843   PetscFunctionBegin;
6844   PetscValidHeaderSpecific(mat,MAT_COOKIE,1);
6845   PetscValidType(mat,1);
6846   PetscValidPointer(v,3);
6847 
6848   if (!mat->assembled) {
6849     SETERRQ(PETSC_ERR_ARG_WRONGSTATE,"Matrix must be already assembled");
6850   }
6851   ierr = PetscLogEventBegin(MAT_SetValues,mat,0,0,0);CHKERRQ(ierr);
6852   if (!mat->ops->setvaluesadifor) SETERRQ1(PETSC_ERR_SUP,"Mat type %s",((PetscObject)mat)->type_name);
6853   ierr = (*mat->ops->setvaluesadifor)(mat,nl,v);CHKERRQ(ierr);
6854   ierr = PetscLogEventEnd(MAT_SetValues,mat,0,0,0);CHKERRQ(ierr);
6855   ierr = PetscObjectStateIncrease((PetscObject)mat);CHKERRQ(ierr);
6856   PetscFunctionReturn(0);
6857 }
6858 
6859 #undef __FUNCT__
6860 #define __FUNCT__ "MatDiagonalScaleLocal"
6861 /*@
6862    MatDiagonalScaleLocal - Scales columns of a matrix given the scaling values including the
6863          ghosted ones.
6864 
6865    Not Collective
6866 
6867    Input Parameters:
6868 +  mat - the matrix
6869 -  diag = the diagonal values, including ghost ones
6870 
6871    Level: developer
6872 
6873    Notes: Works only for MPIAIJ and MPIBAIJ matrices
6874 
6875 .seealso: MatDiagonalScale()
6876 @*/
6877 PetscErrorCode PETSCMAT_DLLEXPORT MatDiagonalScaleLocal(Mat mat,Vec diag)
6878 {
6879   PetscErrorCode ierr;
6880   PetscMPIInt    size;
6881 
6882   PetscFunctionBegin;
6883   PetscValidHeaderSpecific(mat,MAT_COOKIE,1);
6884   PetscValidHeaderSpecific(diag,VEC_COOKIE,2);
6885   PetscValidType(mat,1);
6886 
6887   if (!mat->assembled) {
6888     SETERRQ(PETSC_ERR_ARG_WRONGSTATE,"Matrix must be already assembled");
6889   }
6890   ierr = PetscLogEventBegin(MAT_Scale,mat,0,0,0);CHKERRQ(ierr);
6891   ierr = MPI_Comm_size(((PetscObject)mat)->comm,&size);CHKERRQ(ierr);
6892   if (size == 1) {
6893     PetscInt n,m;
6894     ierr = VecGetSize(diag,&n);CHKERRQ(ierr);
6895     ierr = MatGetSize(mat,0,&m);CHKERRQ(ierr);
6896     if (m == n) {
6897       ierr = MatDiagonalScale(mat,0,diag);CHKERRQ(ierr);
6898     } else {
6899       SETERRQ(PETSC_ERR_SUP,"Only supported for sequential matrices when no ghost points/periodic conditions");
6900     }
6901   } else {
6902     PetscErrorCode (*f)(Mat,Vec);
6903     ierr = PetscObjectQueryFunction((PetscObject)mat,"MatDiagonalScaleLocal_C",(void (**)(void))&f);CHKERRQ(ierr);
6904     if (f) {
6905       ierr = (*f)(mat,diag);CHKERRQ(ierr);
6906     } else {
6907       SETERRQ(PETSC_ERR_SUP,"Only supported for MPIAIJ and MPIBAIJ parallel matrices");
6908     }
6909   }
6910   ierr = PetscLogEventEnd(MAT_Scale,mat,0,0,0);CHKERRQ(ierr);
6911   ierr = PetscObjectStateIncrease((PetscObject)mat);CHKERRQ(ierr);
6912   PetscFunctionReturn(0);
6913 }
6914 
6915 #undef __FUNCT__
6916 #define __FUNCT__ "MatGetInertia"
6917 /*@
6918    MatGetInertia - Gets the inertia from a factored matrix
6919 
6920    Collective on Mat
6921 
6922    Input Parameter:
6923 .  mat - the matrix
6924 
6925    Output Parameters:
6926 +   nneg - number of negative eigenvalues
6927 .   nzero - number of zero eigenvalues
6928 -   npos - number of positive eigenvalues
6929 
6930    Level: advanced
6931 
6932    Notes: Matrix must have been factored by MatCholeskyFactor()
6933 
6934 
6935 @*/
6936 PetscErrorCode PETSCMAT_DLLEXPORT MatGetInertia(Mat mat,PetscInt *nneg,PetscInt *nzero,PetscInt *npos)
6937 {
6938   PetscErrorCode ierr;
6939 
6940   PetscFunctionBegin;
6941   PetscValidHeaderSpecific(mat,MAT_COOKIE,1);
6942   PetscValidType(mat,1);
6943   if (!mat->factor)    SETERRQ(PETSC_ERR_ARG_WRONGSTATE,"Unfactored matrix");
6944   if (!mat->assembled) SETERRQ(PETSC_ERR_ARG_WRONGSTATE,"Numeric factor mat is not assembled");
6945   if (!mat->ops->getinertia) SETERRQ1(PETSC_ERR_SUP,"Mat type %s",((PetscObject)mat)->type_name);
6946   ierr = (*mat->ops->getinertia)(mat,nneg,nzero,npos);CHKERRQ(ierr);
6947   PetscFunctionReturn(0);
6948 }
6949 
6950 /* ----------------------------------------------------------------*/
6951 #undef __FUNCT__
6952 #define __FUNCT__ "MatSolves"
6953 /*@C
6954    MatSolves - Solves A x = b, given a factored matrix, for a collection of vectors
6955 
6956    Collective on Mat and Vecs
6957 
6958    Input Parameters:
6959 +  mat - the factored matrix
6960 -  b - the right-hand-side vectors
6961 
6962    Output Parameter:
6963 .  x - the result vectors
6964 
6965    Notes:
6966    The vectors b and x cannot be the same.  I.e., one cannot
6967    call MatSolves(A,x,x).
6968 
6969    Notes:
6970    Most users should employ the simplified KSP interface for linear solvers
6971    instead of working directly with matrix algebra routines such as this.
6972    See, e.g., KSPCreate().
6973 
6974    Level: developer
6975 
6976    Concepts: matrices^triangular solves
6977 
6978 .seealso: MatSolveAdd(), MatSolveTranspose(), MatSolveTransposeAdd(), MatSolve()
6979 @*/
6980 PetscErrorCode PETSCMAT_DLLEXPORT MatSolves(Mat mat,Vecs b,Vecs x)
6981 {
6982   PetscErrorCode ierr;
6983 
6984   PetscFunctionBegin;
6985   PetscValidHeaderSpecific(mat,MAT_COOKIE,1);
6986   PetscValidType(mat,1);
6987   if (x == b) SETERRQ(PETSC_ERR_ARG_IDN,"x and b must be different vectors");
6988   if (!mat->factor) SETERRQ(PETSC_ERR_ARG_WRONGSTATE,"Unfactored matrix");
6989   if (!mat->rmap->N && !mat->cmap->N) PetscFunctionReturn(0);
6990 
6991   if (!mat->ops->solves) SETERRQ1(PETSC_ERR_SUP,"Mat type %s",((PetscObject)mat)->type_name);
6992   ierr = MatPreallocated(mat);CHKERRQ(ierr);
6993   ierr = PetscLogEventBegin(MAT_Solves,mat,0,0,0);CHKERRQ(ierr);
6994   ierr = (*mat->ops->solves)(mat,b,x);CHKERRQ(ierr);
6995   ierr = PetscLogEventEnd(MAT_Solves,mat,0,0,0);CHKERRQ(ierr);
6996   PetscFunctionReturn(0);
6997 }
6998 
6999 #undef __FUNCT__
7000 #define __FUNCT__ "MatIsSymmetric"
7001 /*@
7002    MatIsSymmetric - Test whether a matrix is symmetric
7003 
7004    Collective on Mat
7005 
7006    Input Parameter:
7007 +  A - the matrix to test
7008 -  tol - difference between value and its transpose less than this amount counts as equal (use 0.0 for exact transpose)
7009 
7010    Output Parameters:
7011 .  flg - the result
7012 
7013    Level: intermediate
7014 
7015    Concepts: matrix^symmetry
7016 
7017 .seealso: MatTranspose(), MatIsTranspose(), MatIsHermitian(), MatIsStructurallySymmetric(), MatSetOption(), MatIsSymmetricKnown()
7018 @*/
7019 PetscErrorCode PETSCMAT_DLLEXPORT MatIsSymmetric(Mat A,PetscReal tol,PetscTruth *flg)
7020 {
7021   PetscErrorCode ierr;
7022 
7023   PetscFunctionBegin;
7024   PetscValidHeaderSpecific(A,MAT_COOKIE,1);
7025   PetscValidPointer(flg,2);
7026   if (!A->symmetric_set) {
7027     if (!A->ops->issymmetric) {
7028       const MatType mattype;
7029       ierr = MatGetType(A,&mattype);CHKERRQ(ierr);
7030       SETERRQ1(PETSC_ERR_SUP,"Matrix of type <%s> does not support checking for symmetric",mattype);
7031     }
7032     ierr = (*A->ops->issymmetric)(A,tol,&A->symmetric);CHKERRQ(ierr);
7033     A->symmetric_set = PETSC_TRUE;
7034     if (A->symmetric) {
7035       A->structurally_symmetric_set = PETSC_TRUE;
7036       A->structurally_symmetric     = PETSC_TRUE;
7037     }
7038   }
7039   *flg = A->symmetric;
7040   PetscFunctionReturn(0);
7041 }
7042 
7043 #undef __FUNCT__
7044 #define __FUNCT__ "MatIsHermitian"
7045 /*@
7046    MatIsHermitian - Test whether a matrix is Hermitian
7047 
7048    Collective on Mat
7049 
7050    Input Parameter:
7051 +  A - the matrix to test
7052 -  tol - difference between value and its transpose less than this amount counts as equal (use 0.0 for exact Hermitian)
7053 
7054    Output Parameters:
7055 .  flg - the result
7056 
7057    Level: intermediate
7058 
7059    Concepts: matrix^symmetry
7060 
7061 .seealso: MatTranspose(), MatIsTranspose(), MatIsHermitian(), MatIsStructurallySymmetric(), MatSetOption(), MatIsSymmetricKnown()
7062 @*/
7063 PetscErrorCode PETSCMAT_DLLEXPORT MatIsHermitian(Mat A,PetscReal tol,PetscTruth *flg)
7064 {
7065   PetscErrorCode ierr;
7066 
7067   PetscFunctionBegin;
7068   PetscValidHeaderSpecific(A,MAT_COOKIE,1);
7069   PetscValidPointer(flg,2);
7070   if (!A->hermitian_set) {
7071     if (!A->ops->ishermitian) {
7072       const MatType mattype;
7073       ierr = MatGetType(A,&mattype);CHKERRQ(ierr);
7074       SETERRQ1(PETSC_ERR_SUP,"Matrix of type <%s> does not support checking for Hermitian",mattype);
7075     }
7076     ierr = (*A->ops->ishermitian)(A,tol,&A->hermitian);CHKERRQ(ierr);
7077     A->hermitian_set = PETSC_TRUE;
7078     if (A->hermitian) {
7079       A->structurally_symmetric_set = PETSC_TRUE;
7080       A->structurally_symmetric     = PETSC_TRUE;
7081     }
7082   }
7083   *flg = A->hermitian;
7084   PetscFunctionReturn(0);
7085 }
7086 
7087 #undef __FUNCT__
7088 #define __FUNCT__ "MatIsSymmetricKnown"
7089 /*@
7090    MatIsSymmetricKnown - Checks the flag on the matrix to see if it is symmetric.
7091 
7092    Collective on Mat
7093 
7094    Input Parameter:
7095 .  A - the matrix to check
7096 
7097    Output Parameters:
7098 +  set - if the symmetric flag is set (this tells you if the next flag is valid)
7099 -  flg - the result
7100 
7101    Level: advanced
7102 
7103    Concepts: matrix^symmetry
7104 
7105    Note: Does not check the matrix values directly, so this may return unknown (set = PETSC_FALSE). Use MatIsSymmetric()
7106          if you want it explicitly checked
7107 
7108 .seealso: MatTranspose(), MatIsTranspose(), MatIsHermitian(), MatIsStructurallySymmetric(), MatSetOption(), MatIsSymmetric()
7109 @*/
7110 PetscErrorCode PETSCMAT_DLLEXPORT MatIsSymmetricKnown(Mat A,PetscTruth *set,PetscTruth *flg)
7111 {
7112   PetscFunctionBegin;
7113   PetscValidHeaderSpecific(A,MAT_COOKIE,1);
7114   PetscValidPointer(set,2);
7115   PetscValidPointer(flg,3);
7116   if (A->symmetric_set) {
7117     *set = PETSC_TRUE;
7118     *flg = A->symmetric;
7119   } else {
7120     *set = PETSC_FALSE;
7121   }
7122   PetscFunctionReturn(0);
7123 }
7124 
7125 #undef __FUNCT__
7126 #define __FUNCT__ "MatIsHermitianKnown"
7127 /*@
7128    MatIsHermitianKnown - Checks the flag on the matrix to see if it is hermitian.
7129 
7130    Collective on Mat
7131 
7132    Input Parameter:
7133 .  A - the matrix to check
7134 
7135    Output Parameters:
7136 +  set - if the hermitian flag is set (this tells you if the next flag is valid)
7137 -  flg - the result
7138 
7139    Level: advanced
7140 
7141    Concepts: matrix^symmetry
7142 
7143    Note: Does not check the matrix values directly, so this may return unknown (set = PETSC_FALSE). Use MatIsHermitian()
7144          if you want it explicitly checked
7145 
7146 .seealso: MatTranspose(), MatIsTranspose(), MatIsHermitian(), MatIsStructurallySymmetric(), MatSetOption(), MatIsSymmetric()
7147 @*/
7148 PetscErrorCode PETSCMAT_DLLEXPORT MatIsHermitianKnown(Mat A,PetscTruth *set,PetscTruth *flg)
7149 {
7150   PetscFunctionBegin;
7151   PetscValidHeaderSpecific(A,MAT_COOKIE,1);
7152   PetscValidPointer(set,2);
7153   PetscValidPointer(flg,3);
7154   if (A->hermitian_set) {
7155     *set = PETSC_TRUE;
7156     *flg = A->hermitian;
7157   } else {
7158     *set = PETSC_FALSE;
7159   }
7160   PetscFunctionReturn(0);
7161 }
7162 
7163 #undef __FUNCT__
7164 #define __FUNCT__ "MatIsStructurallySymmetric"
7165 /*@
7166    MatIsStructurallySymmetric - Test whether a matrix is structurally symmetric
7167 
7168    Collective on Mat
7169 
7170    Input Parameter:
7171 .  A - the matrix to test
7172 
7173    Output Parameters:
7174 .  flg - the result
7175 
7176    Level: intermediate
7177 
7178    Concepts: matrix^symmetry
7179 
7180 .seealso: MatTranspose(), MatIsTranspose(), MatIsHermitian(), MatIsSymmetric(), MatSetOption()
7181 @*/
7182 PetscErrorCode PETSCMAT_DLLEXPORT MatIsStructurallySymmetric(Mat A,PetscTruth *flg)
7183 {
7184   PetscErrorCode ierr;
7185 
7186   PetscFunctionBegin;
7187   PetscValidHeaderSpecific(A,MAT_COOKIE,1);
7188   PetscValidPointer(flg,2);
7189   if (!A->structurally_symmetric_set) {
7190     if (!A->ops->isstructurallysymmetric) SETERRQ(PETSC_ERR_SUP,"Matrix does not support checking for structural symmetric");
7191     ierr = (*A->ops->isstructurallysymmetric)(A,&A->structurally_symmetric);CHKERRQ(ierr);
7192     A->structurally_symmetric_set = PETSC_TRUE;
7193   }
7194   *flg = A->structurally_symmetric;
7195   PetscFunctionReturn(0);
7196 }
7197 
7198 #undef __FUNCT__
7199 #define __FUNCT__ "MatStashGetInfo"
7200 extern PetscErrorCode MatStashGetInfo_Private(MatStash*,PetscInt*,PetscInt*);
7201 /*@
7202    MatStashGetInfo - Gets how many values are currently in the vector stash, i.e. need
7203        to be communicated to other processors during the MatAssemblyBegin/End() process
7204 
7205     Not collective
7206 
7207    Input Parameter:
7208 .   vec - the vector
7209 
7210    Output Parameters:
7211 +   nstash   - the size of the stash
7212 .   reallocs - the number of additional mallocs incurred.
7213 .   bnstash   - the size of the block stash
7214 -   breallocs - the number of additional mallocs incurred.in the block stash
7215 
7216    Level: advanced
7217 
7218 .seealso: MatAssemblyBegin(), MatAssemblyEnd(), Mat, MatStashSetInitialSize()
7219 
7220 @*/
7221 PetscErrorCode PETSCMAT_DLLEXPORT MatStashGetInfo(Mat mat,PetscInt *nstash,PetscInt *reallocs,PetscInt *bnstash,PetscInt *breallocs)
7222 {
7223   PetscErrorCode ierr;
7224   PetscFunctionBegin;
7225   ierr = MatStashGetInfo_Private(&mat->stash,nstash,reallocs);CHKERRQ(ierr);
7226   ierr = MatStashGetInfo_Private(&mat->bstash,bnstash,breallocs);CHKERRQ(ierr);
7227   PetscFunctionReturn(0);
7228 }
7229 
7230 #undef __FUNCT__
7231 #define __FUNCT__ "MatGetVecs"
7232 /*@C
7233    MatGetVecs - Get vector(s) compatible with the matrix, i.e. with the same
7234      parallel layout
7235 
7236    Collective on Mat
7237 
7238    Input Parameter:
7239 .  mat - the matrix
7240 
7241    Output Parameter:
7242 +   right - (optional) vector that the matrix can be multiplied against
7243 -   left - (optional) vector that the matrix vector product can be stored in
7244 
7245   Level: advanced
7246 
7247 .seealso: MatCreate()
7248 @*/
7249 PetscErrorCode PETSCMAT_DLLEXPORT MatGetVecs(Mat mat,Vec *right,Vec *left)
7250 {
7251   PetscErrorCode ierr;
7252 
7253   PetscFunctionBegin;
7254   PetscValidHeaderSpecific(mat,MAT_COOKIE,1);
7255   PetscValidType(mat,1);
7256   ierr = MatPreallocated(mat);CHKERRQ(ierr);
7257   if (mat->ops->getvecs) {
7258     ierr = (*mat->ops->getvecs)(mat,right,left);CHKERRQ(ierr);
7259   } else {
7260     PetscMPIInt size;
7261     ierr = MPI_Comm_size(((PetscObject)mat)->comm, &size);CHKERRQ(ierr);
7262     if (right) {
7263       ierr = VecCreate(((PetscObject)mat)->comm,right);CHKERRQ(ierr);
7264       ierr = VecSetSizes(*right,mat->cmap->n,PETSC_DETERMINE);CHKERRQ(ierr);
7265       ierr = VecSetBlockSize(*right,mat->rmap->bs);CHKERRQ(ierr);
7266       if (size > 1) {
7267         /* New vectors uses Mat cmap and does not create a new one */
7268 	ierr = PetscMapDestroy((*right)->map);CHKERRQ(ierr);
7269 	(*right)->map = mat->cmap;
7270 	mat->cmap->refcnt++;
7271 
7272         ierr = VecSetType(*right,VECMPI);CHKERRQ(ierr);
7273       } else {ierr = VecSetType(*right,VECSEQ);CHKERRQ(ierr);}
7274     }
7275     if (left) {
7276       ierr = VecCreate(((PetscObject)mat)->comm,left);CHKERRQ(ierr);
7277       ierr = VecSetSizes(*left,mat->rmap->n,PETSC_DETERMINE);CHKERRQ(ierr);
7278       ierr = VecSetBlockSize(*left,mat->rmap->bs);CHKERRQ(ierr);
7279       if (size > 1) {
7280         /* New vectors uses Mat rmap and does not create a new one */
7281 	ierr = PetscMapDestroy((*left)->map);CHKERRQ(ierr);
7282 	(*left)->map = mat->rmap;
7283 	mat->rmap->refcnt++;
7284 
7285         ierr = VecSetType(*left,VECMPI);CHKERRQ(ierr);
7286       } else {ierr = VecSetType(*left,VECSEQ);CHKERRQ(ierr);}
7287     }
7288   }
7289   if (mat->mapping) {
7290     if (right) {ierr = VecSetLocalToGlobalMapping(*right,mat->mapping);CHKERRQ(ierr);}
7291     if (left) {ierr = VecSetLocalToGlobalMapping(*left,mat->mapping);CHKERRQ(ierr);}
7292   }
7293   if (mat->bmapping) {
7294     if (right) {ierr = VecSetLocalToGlobalMappingBlock(*right,mat->bmapping);CHKERRQ(ierr);}
7295     if (left) {ierr = VecSetLocalToGlobalMappingBlock(*left,mat->bmapping);CHKERRQ(ierr);}
7296   }
7297   PetscFunctionReturn(0);
7298 }
7299 
7300 #undef __FUNCT__
7301 #define __FUNCT__ "MatFactorInfoInitialize"
7302 /*@C
7303    MatFactorInfoInitialize - Initializes a MatFactorInfo data structure
7304      with default values.
7305 
7306    Not Collective
7307 
7308    Input Parameters:
7309 .    info - the MatFactorInfo data structure
7310 
7311 
7312    Notes: The solvers are generally used through the KSP and PC objects, for example
7313           PCLU, PCILU, PCCHOLESKY, PCICC
7314 
7315    Level: developer
7316 
7317 .seealso: MatFactorInfo
7318 
7319     Developer Note: fortran interface is not autogenerated as the f90
7320     interface defintion cannot be generated correctly [due to MatFactorInfo]
7321 
7322 @*/
7323 
7324 PetscErrorCode PETSCMAT_DLLEXPORT MatFactorInfoInitialize(MatFactorInfo *info)
7325 {
7326   PetscErrorCode ierr;
7327 
7328   PetscFunctionBegin;
7329   ierr = PetscMemzero(info,sizeof(MatFactorInfo));CHKERRQ(ierr);
7330   PetscFunctionReturn(0);
7331 }
7332 
7333 #undef __FUNCT__
7334 #define __FUNCT__ "MatPtAP"
7335 /*@
7336    MatPtAP - Creates the matrix projection C = P^T * A * P
7337 
7338    Collective on Mat
7339 
7340    Input Parameters:
7341 +  A - the matrix
7342 .  P - the projection matrix
7343 .  scall - either MAT_INITIAL_MATRIX or MAT_REUSE_MATRIX
7344 -  fill - expected fill as ratio of nnz(C)/nnz(A)
7345 
7346    Output Parameters:
7347 .  C - the product matrix
7348 
7349    Notes:
7350    C will be created and must be destroyed by the user with MatDestroy().
7351 
7352    This routine is currently only implemented for pairs of AIJ matrices and classes
7353    which inherit from AIJ.
7354 
7355    Level: intermediate
7356 
7357 .seealso: MatPtAPSymbolic(), MatPtAPNumeric(), MatMatMult()
7358 @*/
7359 PetscErrorCode PETSCMAT_DLLEXPORT MatPtAP(Mat A,Mat P,MatReuse scall,PetscReal fill,Mat *C)
7360 {
7361   PetscErrorCode ierr;
7362 
7363   PetscFunctionBegin;
7364   PetscValidHeaderSpecific(A,MAT_COOKIE,1);
7365   PetscValidType(A,1);
7366   if (!A->assembled) SETERRQ(PETSC_ERR_ARG_WRONGSTATE,"Not for unassembled matrix");
7367   if (A->factor) SETERRQ(PETSC_ERR_ARG_WRONGSTATE,"Not for factored matrix");
7368   PetscValidHeaderSpecific(P,MAT_COOKIE,2);
7369   PetscValidType(P,2);
7370   ierr = MatPreallocated(P);CHKERRQ(ierr);
7371   if (!P->assembled) SETERRQ(PETSC_ERR_ARG_WRONGSTATE,"Not for unassembled matrix");
7372   if (P->factor) SETERRQ(PETSC_ERR_ARG_WRONGSTATE,"Not for factored matrix");
7373   PetscValidPointer(C,3);
7374   if (P->rmap->N!=A->cmap->N) SETERRQ2(PETSC_ERR_ARG_SIZ,"Matrix dimensions are incompatible, %D != %D",P->rmap->N,A->cmap->N);
7375   if (fill < 1.0) SETERRQ1(PETSC_ERR_ARG_SIZ,"Expected fill=%G must be >= 1.0",fill);
7376   ierr = MatPreallocated(A);CHKERRQ(ierr);
7377 
7378   ierr = PetscLogEventBegin(MAT_PtAP,A,P,0,0);CHKERRQ(ierr);
7379   ierr = (*A->ops->ptap)(A,P,scall,fill,C);CHKERRQ(ierr);
7380   ierr = PetscLogEventEnd(MAT_PtAP,A,P,0,0);CHKERRQ(ierr);
7381 
7382   PetscFunctionReturn(0);
7383 }
7384 
7385 #undef __FUNCT__
7386 #define __FUNCT__ "MatPtAPNumeric"
7387 /*@
7388    MatPtAPNumeric - Computes the matrix projection C = P^T * A * P
7389 
7390    Collective on Mat
7391 
7392    Input Parameters:
7393 +  A - the matrix
7394 -  P - the projection matrix
7395 
7396    Output Parameters:
7397 .  C - the product matrix
7398 
7399    Notes:
7400    C must have been created by calling MatPtAPSymbolic and must be destroyed by
7401    the user using MatDeatroy().
7402 
7403    This routine is currently only implemented for pairs of AIJ matrices and classes
7404    which inherit from AIJ.  C will be of type MATAIJ.
7405 
7406    Level: intermediate
7407 
7408 .seealso: MatPtAP(), MatPtAPSymbolic(), MatMatMultNumeric()
7409 @*/
7410 PetscErrorCode PETSCMAT_DLLEXPORT MatPtAPNumeric(Mat A,Mat P,Mat C)
7411 {
7412   PetscErrorCode ierr;
7413 
7414   PetscFunctionBegin;
7415   PetscValidHeaderSpecific(A,MAT_COOKIE,1);
7416   PetscValidType(A,1);
7417   if (!A->assembled) SETERRQ(PETSC_ERR_ARG_WRONGSTATE,"Not for unassembled matrix");
7418   if (A->factor) SETERRQ(PETSC_ERR_ARG_WRONGSTATE,"Not for factored matrix");
7419   PetscValidHeaderSpecific(P,MAT_COOKIE,2);
7420   PetscValidType(P,2);
7421   ierr = MatPreallocated(P);CHKERRQ(ierr);
7422   if (!P->assembled) SETERRQ(PETSC_ERR_ARG_WRONGSTATE,"Not for unassembled matrix");
7423   if (P->factor) SETERRQ(PETSC_ERR_ARG_WRONGSTATE,"Not for factored matrix");
7424   PetscValidHeaderSpecific(C,MAT_COOKIE,3);
7425   PetscValidType(C,3);
7426   ierr = MatPreallocated(C);CHKERRQ(ierr);
7427   if (C->factor) SETERRQ(PETSC_ERR_ARG_WRONGSTATE,"Not for factored matrix");
7428   if (P->cmap->N!=C->rmap->N) SETERRQ2(PETSC_ERR_ARG_SIZ,"Matrix dimensions are incompatible, %D != %D",P->cmap->N,C->rmap->N);
7429   if (P->rmap->N!=A->cmap->N) SETERRQ2(PETSC_ERR_ARG_SIZ,"Matrix dimensions are incompatible, %D != %D",P->rmap->N,A->cmap->N);
7430   if (A->rmap->N!=A->cmap->N) SETERRQ2(PETSC_ERR_ARG_SIZ,"Matrix 'A' must be square, %D != %D",A->rmap->N,A->cmap->N);
7431   if (P->cmap->N!=C->cmap->N) SETERRQ2(PETSC_ERR_ARG_SIZ,"Matrix dimensions are incompatible, %D != %D",P->cmap->N,C->cmap->N);
7432   ierr = MatPreallocated(A);CHKERRQ(ierr);
7433 
7434   ierr = PetscLogEventBegin(MAT_PtAPNumeric,A,P,0,0);CHKERRQ(ierr);
7435   ierr = (*A->ops->ptapnumeric)(A,P,C);CHKERRQ(ierr);
7436   ierr = PetscLogEventEnd(MAT_PtAPNumeric,A,P,0,0);CHKERRQ(ierr);
7437   PetscFunctionReturn(0);
7438 }
7439 
7440 #undef __FUNCT__
7441 #define __FUNCT__ "MatPtAPSymbolic"
7442 /*@
7443    MatPtAPSymbolic - Creates the (i,j) structure of the matrix projection C = P^T * A * P
7444 
7445    Collective on Mat
7446 
7447    Input Parameters:
7448 +  A - the matrix
7449 -  P - the projection matrix
7450 
7451    Output Parameters:
7452 .  C - the (i,j) structure of the product matrix
7453 
7454    Notes:
7455    C will be created and must be destroyed by the user with MatDestroy().
7456 
7457    This routine is currently only implemented for pairs of SeqAIJ matrices and classes
7458    which inherit from SeqAIJ.  C will be of type MATSEQAIJ.  The product is computed using
7459    this (i,j) structure by calling MatPtAPNumeric().
7460 
7461    Level: intermediate
7462 
7463 .seealso: MatPtAP(), MatPtAPNumeric(), MatMatMultSymbolic()
7464 @*/
7465 PetscErrorCode PETSCMAT_DLLEXPORT MatPtAPSymbolic(Mat A,Mat P,PetscReal fill,Mat *C)
7466 {
7467   PetscErrorCode ierr;
7468 
7469   PetscFunctionBegin;
7470   PetscValidHeaderSpecific(A,MAT_COOKIE,1);
7471   PetscValidType(A,1);
7472   if (!A->assembled) SETERRQ(PETSC_ERR_ARG_WRONGSTATE,"Not for unassembled matrix");
7473   if (A->factor) SETERRQ(PETSC_ERR_ARG_WRONGSTATE,"Not for factored matrix");
7474   if (fill <1.0) SETERRQ1(PETSC_ERR_ARG_SIZ,"Expected fill=%G must be >= 1.0",fill);
7475   PetscValidHeaderSpecific(P,MAT_COOKIE,2);
7476   PetscValidType(P,2);
7477   ierr = MatPreallocated(P);CHKERRQ(ierr);
7478   if (!P->assembled) SETERRQ(PETSC_ERR_ARG_WRONGSTATE,"Not for unassembled matrix");
7479   if (P->factor) SETERRQ(PETSC_ERR_ARG_WRONGSTATE,"Not for factored matrix");
7480   PetscValidPointer(C,3);
7481 
7482   if (P->rmap->N!=A->cmap->N) SETERRQ2(PETSC_ERR_ARG_SIZ,"Matrix dimensions are incompatible, %D != %D",P->rmap->N,A->cmap->N);
7483   if (A->rmap->N!=A->cmap->N) SETERRQ2(PETSC_ERR_ARG_SIZ,"Matrix 'A' must be square, %D != %D",A->rmap->N,A->cmap->N);
7484   ierr = MatPreallocated(A);CHKERRQ(ierr);
7485   ierr = PetscLogEventBegin(MAT_PtAPSymbolic,A,P,0,0);CHKERRQ(ierr);
7486   ierr = (*A->ops->ptapsymbolic)(A,P,fill,C);CHKERRQ(ierr);
7487   ierr = PetscLogEventEnd(MAT_PtAPSymbolic,A,P,0,0);CHKERRQ(ierr);
7488 
7489   ierr = MatSetBlockSize(*C,A->rmap->bs);CHKERRQ(ierr);
7490 
7491   PetscFunctionReturn(0);
7492 }
7493 
7494 #undef __FUNCT__
7495 #define __FUNCT__ "MatMatMult"
7496 /*@
7497    MatMatMult - Performs Matrix-Matrix Multiplication C=A*B.
7498 
7499    Collective on Mat
7500 
7501    Input Parameters:
7502 +  A - the left matrix
7503 .  B - the right matrix
7504 .  scall - either MAT_INITIAL_MATRIX or MAT_REUSE_MATRIX
7505 -  fill - expected fill as ratio of nnz(C)/(nnz(A) + nnz(B)), use PETSC_DEFAULT if you do not have a good estimate
7506           if the result is a dense matrix this is irrelevent
7507 
7508    Output Parameters:
7509 .  C - the product matrix
7510 
7511    Notes:
7512    Unless scall is MAT_REUSE_MATRIX C will be created.
7513 
7514    MAT_REUSE_MATRIX can only be used if the matrices A and B have the same nonzero pattern as in the previous call
7515 
7516    To determine the correct fill value, run with -info and search for the string "Fill ratio" to see the value
7517    actually needed.
7518 
7519    If you have many matrices with the same non-zero structure to multiply, you
7520    should either
7521 $   1) use MAT_REUSE_MATRIX in all calls but the first or
7522 $   2) call MatMatMultSymbolic() once and then MatMatMultNumeric() for each product needed
7523 
7524    Level: intermediate
7525 
7526 .seealso: MatMatMultSymbolic(), MatMatMultNumeric(), MatPtAP()
7527 @*/
7528 PetscErrorCode PETSCMAT_DLLEXPORT MatMatMult(Mat A,Mat B,MatReuse scall,PetscReal fill,Mat *C)
7529 {
7530   PetscErrorCode ierr;
7531   PetscErrorCode (*fA)(Mat,Mat,MatReuse,PetscReal,Mat*);
7532   PetscErrorCode (*fB)(Mat,Mat,MatReuse,PetscReal,Mat*);
7533   PetscErrorCode (*mult)(Mat,Mat,MatReuse,PetscReal,Mat *)=PETSC_NULL;
7534 
7535   PetscFunctionBegin;
7536   PetscValidHeaderSpecific(A,MAT_COOKIE,1);
7537   PetscValidType(A,1);
7538   if (!A->assembled) SETERRQ(PETSC_ERR_ARG_WRONGSTATE,"Not for unassembled matrix");
7539   if (A->factor) SETERRQ(PETSC_ERR_ARG_WRONGSTATE,"Not for factored matrix");
7540   PetscValidHeaderSpecific(B,MAT_COOKIE,2);
7541   PetscValidType(B,2);
7542   ierr = MatPreallocated(B);CHKERRQ(ierr);
7543   if (!B->assembled) SETERRQ(PETSC_ERR_ARG_WRONGSTATE,"Not for unassembled matrix");
7544   if (B->factor) SETERRQ(PETSC_ERR_ARG_WRONGSTATE,"Not for factored matrix");
7545   PetscValidPointer(C,3);
7546   if (B->rmap->N!=A->cmap->N) SETERRQ2(PETSC_ERR_ARG_SIZ,"Matrix dimensions are incompatible, %D != %D",B->rmap->N,A->cmap->N);
7547   if (scall == MAT_REUSE_MATRIX){
7548     PetscValidPointer(*C,5);
7549     PetscValidHeaderSpecific(*C,MAT_COOKIE,5);
7550   }
7551   if (fill == PETSC_DEFAULT || fill == PETSC_DECIDE) fill = 2.0;
7552   if (fill < 1.0) SETERRQ1(PETSC_ERR_ARG_SIZ,"Expected fill=%G must be >= 1.0",fill);
7553   ierr = MatPreallocated(A);CHKERRQ(ierr);
7554 
7555   fA = A->ops->matmult;
7556   fB = B->ops->matmult;
7557   if (fB == fA) {
7558     if (!fB) SETERRQ1(PETSC_ERR_SUP,"MatMatMult not supported for B of type %s",((PetscObject)B)->type_name);
7559     mult = fB;
7560   } else {
7561     /* dispatch based on the type of A and B */
7562     char  multname[256];
7563     ierr = PetscStrcpy(multname,"MatMatMult_");CHKERRQ(ierr);
7564     ierr = PetscStrcat(multname,((PetscObject)A)->type_name);CHKERRQ(ierr);
7565     ierr = PetscStrcat(multname,"_");CHKERRQ(ierr);
7566     ierr = PetscStrcat(multname,((PetscObject)B)->type_name);CHKERRQ(ierr);
7567     ierr = PetscStrcat(multname,"_C");CHKERRQ(ierr); /* e.g., multname = "MatMatMult_seqdense_seqaij_C" */
7568     ierr = PetscObjectQueryFunction((PetscObject)B,multname,(void (**)(void))&mult);CHKERRQ(ierr);
7569     if (!mult) SETERRQ2(PETSC_ERR_ARG_INCOMP,"MatMatMult requires A, %s, to be compatible with B, %s",((PetscObject)A)->type_name,((PetscObject)B)->type_name);
7570   }
7571   ierr = PetscLogEventBegin(MAT_MatMult,A,B,0,0);CHKERRQ(ierr);
7572   ierr = (*mult)(A,B,scall,fill,C);CHKERRQ(ierr);
7573   ierr = PetscLogEventEnd(MAT_MatMult,A,B,0,0);CHKERRQ(ierr);
7574   PetscFunctionReturn(0);
7575 }
7576 
7577 #undef __FUNCT__
7578 #define __FUNCT__ "MatMatMultSymbolic"
7579 /*@
7580    MatMatMultSymbolic - Performs construction, preallocation, and computes the ij structure
7581    of the matrix-matrix product C=A*B.  Call this routine before calling MatMatMultNumeric().
7582 
7583    Collective on Mat
7584 
7585    Input Parameters:
7586 +  A - the left matrix
7587 .  B - the right matrix
7588 -  fill - expected fill as ratio of nnz(C)/(nnz(A) + nnz(B)), use PETSC_DEFAULT if you do not have a good estimate,
7589       if C is a dense matrix this is irrelevent
7590 
7591    Output Parameters:
7592 .  C - the product matrix
7593 
7594    Notes:
7595    Unless scall is MAT_REUSE_MATRIX C will be created.
7596 
7597    To determine the correct fill value, run with -info and search for the string "Fill ratio" to see the value
7598    actually needed.
7599 
7600    This routine is currently implemented for
7601     - pairs of AIJ matrices and classes which inherit from AIJ, C will be of type AIJ
7602     - pairs of AIJ (A) and Dense (B) matrix, C will be of type Dense.
7603     - pairs of Dense (A) and AIJ (B) matrix, C will be of type Dense.
7604 
7605    Level: intermediate
7606 
7607 .seealso: MatMatMult(), MatMatMultNumeric()
7608 @*/
7609 PetscErrorCode PETSCMAT_DLLEXPORT MatMatMultSymbolic(Mat A,Mat B,PetscReal fill,Mat *C)
7610 {
7611   PetscErrorCode ierr;
7612   PetscErrorCode (*Asymbolic)(Mat,Mat,PetscReal,Mat *);
7613   PetscErrorCode (*Bsymbolic)(Mat,Mat,PetscReal,Mat *);
7614   PetscErrorCode (*symbolic)(Mat,Mat,PetscReal,Mat *)=PETSC_NULL;
7615 
7616   PetscFunctionBegin;
7617   PetscValidHeaderSpecific(A,MAT_COOKIE,1);
7618   PetscValidType(A,1);
7619   if (!A->assembled) SETERRQ(PETSC_ERR_ARG_WRONGSTATE,"Not for unassembled matrix");
7620   if (A->factor) SETERRQ(PETSC_ERR_ARG_WRONGSTATE,"Not for factored matrix");
7621 
7622   PetscValidHeaderSpecific(B,MAT_COOKIE,2);
7623   PetscValidType(B,2);
7624   ierr = MatPreallocated(B);CHKERRQ(ierr);
7625   if (!B->assembled) SETERRQ(PETSC_ERR_ARG_WRONGSTATE,"Not for unassembled matrix");
7626   if (B->factor) SETERRQ(PETSC_ERR_ARG_WRONGSTATE,"Not for factored matrix");
7627   PetscValidPointer(C,3);
7628 
7629   if (B->rmap->N!=A->cmap->N) SETERRQ2(PETSC_ERR_ARG_SIZ,"Matrix dimensions are incompatible, %D != %D",B->rmap->N,A->cmap->N);
7630   if (fill == PETSC_DEFAULT) fill = 2.0;
7631   if (fill < 1.0) SETERRQ1(PETSC_ERR_ARG_SIZ,"Expected fill=%G must be > 1.0",fill);
7632   ierr = MatPreallocated(A);CHKERRQ(ierr);
7633 
7634   Asymbolic = A->ops->matmultsymbolic;
7635   Bsymbolic = B->ops->matmultsymbolic;
7636   if (Asymbolic == Bsymbolic){
7637     if (!Bsymbolic) SETERRQ1(PETSC_ERR_SUP,"C=A*B not implemented for B of type %s",((PetscObject)B)->type_name);
7638     symbolic = Bsymbolic;
7639   } else { /* dispatch based on the type of A and B */
7640     char  symbolicname[256];
7641     ierr = PetscStrcpy(symbolicname,"MatMatMultSymbolic_");CHKERRQ(ierr);
7642     ierr = PetscStrcat(symbolicname,((PetscObject)A)->type_name);CHKERRQ(ierr);
7643     ierr = PetscStrcat(symbolicname,"_");CHKERRQ(ierr);
7644     ierr = PetscStrcat(symbolicname,((PetscObject)B)->type_name);CHKERRQ(ierr);
7645     ierr = PetscStrcat(symbolicname,"_C");CHKERRQ(ierr);
7646     ierr = PetscObjectQueryFunction((PetscObject)B,symbolicname,(void (**)(void))&symbolic);CHKERRQ(ierr);
7647     if (!symbolic) SETERRQ2(PETSC_ERR_ARG_INCOMP,"MatMatMultSymbolic requires A, %s, to be compatible with B, %s",((PetscObject)A)->type_name,((PetscObject)B)->type_name);
7648   }
7649   ierr = PetscLogEventBegin(MAT_MatMultSymbolic,A,B,0,0);CHKERRQ(ierr);
7650   ierr = (*symbolic)(A,B,fill,C);CHKERRQ(ierr);
7651   ierr = PetscLogEventEnd(MAT_MatMultSymbolic,A,B,0,0);CHKERRQ(ierr);
7652   PetscFunctionReturn(0);
7653 }
7654 
7655 #undef __FUNCT__
7656 #define __FUNCT__ "MatMatMultNumeric"
7657 /*@
7658    MatMatMultNumeric - Performs the numeric matrix-matrix product.
7659    Call this routine after first calling MatMatMultSymbolic().
7660 
7661    Collective on Mat
7662 
7663    Input Parameters:
7664 +  A - the left matrix
7665 -  B - the right matrix
7666 
7667    Output Parameters:
7668 .  C - the product matrix, which was created by from MatMatMultSymbolic() or a call to MatMatMult().
7669 
7670    Notes:
7671    C must have been created with MatMatMultSymbolic().
7672 
7673    This routine is currently implemented for
7674     - pairs of AIJ matrices and classes which inherit from AIJ, C will be of type MATAIJ.
7675     - pairs of AIJ (A) and Dense (B) matrix, C will be of type Dense.
7676     - pairs of Dense (A) and AIJ (B) matrix, C will be of type Dense.
7677 
7678    Level: intermediate
7679 
7680 .seealso: MatMatMult(), MatMatMultSymbolic()
7681 @*/
7682 PetscErrorCode PETSCMAT_DLLEXPORT MatMatMultNumeric(Mat A,Mat B,Mat C)
7683 {
7684   PetscErrorCode ierr;
7685   PetscErrorCode (*Anumeric)(Mat,Mat,Mat);
7686   PetscErrorCode (*Bnumeric)(Mat,Mat,Mat);
7687   PetscErrorCode (*numeric)(Mat,Mat,Mat)=PETSC_NULL;
7688 
7689   PetscFunctionBegin;
7690   PetscValidHeaderSpecific(A,MAT_COOKIE,1);
7691   PetscValidType(A,1);
7692   if (!A->assembled) SETERRQ(PETSC_ERR_ARG_WRONGSTATE,"Not for unassembled matrix");
7693   if (A->factor) SETERRQ(PETSC_ERR_ARG_WRONGSTATE,"Not for factored matrix");
7694 
7695   PetscValidHeaderSpecific(B,MAT_COOKIE,2);
7696   PetscValidType(B,2);
7697   ierr = MatPreallocated(B);CHKERRQ(ierr);
7698   if (!B->assembled) SETERRQ(PETSC_ERR_ARG_WRONGSTATE,"Not for unassembled matrix");
7699   if (B->factor) SETERRQ(PETSC_ERR_ARG_WRONGSTATE,"Not for factored matrix");
7700 
7701   PetscValidHeaderSpecific(C,MAT_COOKIE,3);
7702   PetscValidType(C,3);
7703   ierr = MatPreallocated(C);CHKERRQ(ierr);
7704   if (!C->assembled) SETERRQ(PETSC_ERR_ARG_WRONGSTATE,"Not for unassembled matrix");
7705   if (C->factor) SETERRQ(PETSC_ERR_ARG_WRONGSTATE,"Not for factored matrix");
7706 
7707   if (B->cmap->N!=C->cmap->N) SETERRQ2(PETSC_ERR_ARG_SIZ,"Matrix dimensions are incompatible, %D != %D",B->cmap->N,C->cmap->N);
7708   if (B->rmap->N!=A->cmap->N) SETERRQ2(PETSC_ERR_ARG_SIZ,"Matrix dimensions are incompatible, %D != %D",B->rmap->N,A->cmap->N);
7709   if (A->rmap->N!=C->rmap->N) SETERRQ2(PETSC_ERR_ARG_SIZ,"Matrix dimensions are incompatible, %D != %D",A->rmap->N,C->rmap->N);
7710   ierr = MatPreallocated(A);CHKERRQ(ierr);
7711 
7712   Anumeric = A->ops->matmultnumeric;
7713   Bnumeric = B->ops->matmultnumeric;
7714   if (Anumeric == Bnumeric){
7715     if (!Bnumeric) SETERRQ1(PETSC_ERR_SUP,"MatMatMultNumeric not supported for B of type %s",((PetscObject)B)->type_name);
7716     numeric = Bnumeric;
7717   } else {
7718     char  numericname[256];
7719     ierr = PetscStrcpy(numericname,"MatMatMultNumeric_");CHKERRQ(ierr);
7720     ierr = PetscStrcat(numericname,((PetscObject)A)->type_name);CHKERRQ(ierr);
7721     ierr = PetscStrcat(numericname,"_");CHKERRQ(ierr);
7722     ierr = PetscStrcat(numericname,((PetscObject)B)->type_name);CHKERRQ(ierr);
7723     ierr = PetscStrcat(numericname,"_C");CHKERRQ(ierr);
7724     ierr = PetscObjectQueryFunction((PetscObject)B,numericname,(void (**)(void))&numeric);CHKERRQ(ierr);
7725     if (!numeric)
7726       SETERRQ2(PETSC_ERR_ARG_INCOMP,"MatMatMultNumeric requires A, %s, to be compatible with B, %s",((PetscObject)A)->type_name,((PetscObject)B)->type_name);
7727   }
7728   ierr = PetscLogEventBegin(MAT_MatMultNumeric,A,B,0,0);CHKERRQ(ierr);
7729   ierr = (*numeric)(A,B,C);CHKERRQ(ierr);
7730   ierr = PetscLogEventEnd(MAT_MatMultNumeric,A,B,0,0);CHKERRQ(ierr);
7731   PetscFunctionReturn(0);
7732 }
7733 
7734 #undef __FUNCT__
7735 #define __FUNCT__ "MatMatMultTranspose"
7736 /*@
7737    MatMatMultTranspose - Performs Matrix-Matrix Multiplication C=A^T*B.
7738 
7739    Collective on Mat
7740 
7741    Input Parameters:
7742 +  A - the left matrix
7743 .  B - the right matrix
7744 .  scall - either MAT_INITIAL_MATRIX or MAT_REUSE_MATRIX
7745 -  fill - expected fill as ratio of nnz(C)/(nnz(A) + nnz(B)), use PETSC_DEFAULT if not known
7746 
7747    Output Parameters:
7748 .  C - the product matrix
7749 
7750    Notes:
7751    C will be created if MAT_INITIAL_MATRIX and must be destroyed by the user with MatDestroy().
7752 
7753    MAT_REUSE_MATRIX can only be used if the matrices A and B have the same nonzero pattern as in the previous call
7754 
7755   To determine the correct fill value, run with -info and search for the string "Fill ratio" to see the value
7756    actually needed.
7757 
7758    This routine is currently only implemented for pairs of SeqAIJ matrices and pairs of SeqDense matrices and classes
7759    which inherit from SeqAIJ.  C will be of type MATSEQAIJ.
7760 
7761    Level: intermediate
7762 
7763 .seealso: MatMatMultTransposeSymbolic(), MatMatMultTransposeNumeric(), MatPtAP()
7764 @*/
7765 PetscErrorCode PETSCMAT_DLLEXPORT MatMatMultTranspose(Mat A,Mat B,MatReuse scall,PetscReal fill,Mat *C)
7766 {
7767   PetscErrorCode ierr;
7768   PetscErrorCode (*fA)(Mat,Mat,MatReuse,PetscReal,Mat*);
7769   PetscErrorCode (*fB)(Mat,Mat,MatReuse,PetscReal,Mat*);
7770 
7771   PetscFunctionBegin;
7772   PetscValidHeaderSpecific(A,MAT_COOKIE,1);
7773   PetscValidType(A,1);
7774   if (!A->assembled) SETERRQ(PETSC_ERR_ARG_WRONGSTATE,"Not for unassembled matrix");
7775   if (A->factor) SETERRQ(PETSC_ERR_ARG_WRONGSTATE,"Not for factored matrix");
7776   PetscValidHeaderSpecific(B,MAT_COOKIE,2);
7777   PetscValidType(B,2);
7778   ierr = MatPreallocated(B);CHKERRQ(ierr);
7779   if (!B->assembled) SETERRQ(PETSC_ERR_ARG_WRONGSTATE,"Not for unassembled matrix");
7780   if (B->factor) SETERRQ(PETSC_ERR_ARG_WRONGSTATE,"Not for factored matrix");
7781   PetscValidPointer(C,3);
7782   if (B->rmap->N!=A->rmap->N) SETERRQ2(PETSC_ERR_ARG_SIZ,"Matrix dimensions are incompatible, %D != %D",B->rmap->N,A->rmap->N);
7783   if (fill < 1.0) SETERRQ1(PETSC_ERR_ARG_SIZ,"Expected fill=%G must be > 1.0",fill);
7784   ierr = MatPreallocated(A);CHKERRQ(ierr);
7785 
7786   fA = A->ops->matmulttranspose;
7787   if (!fA) SETERRQ1(PETSC_ERR_SUP,"MatMatMultTranspose not supported for A of type %s",((PetscObject)A)->type_name);
7788   fB = B->ops->matmulttranspose;
7789   if (!fB) SETERRQ1(PETSC_ERR_SUP,"MatMatMultTranspose not supported for B of type %s",((PetscObject)B)->type_name);
7790   if (fB!=fA) SETERRQ2(PETSC_ERR_ARG_INCOMP,"MatMatMultTranspose requires A, %s, to be compatible with B, %s",((PetscObject)A)->type_name,((PetscObject)B)->type_name);
7791 
7792   ierr = PetscLogEventBegin(MAT_MatMultTranspose,A,B,0,0);CHKERRQ(ierr);
7793   ierr = (*A->ops->matmulttranspose)(A,B,scall,fill,C);CHKERRQ(ierr);
7794   ierr = PetscLogEventEnd(MAT_MatMultTranspose,A,B,0,0);CHKERRQ(ierr);
7795 
7796   PetscFunctionReturn(0);
7797 }
7798 
7799 #undef __FUNCT__
7800 #define __FUNCT__ "MatGetRedundantMatrix"
7801 /*@C
7802    MatGetRedundantMatrix - Create redundant matrices and put them into processors of subcommunicators.
7803 
7804    Collective on Mat
7805 
7806    Input Parameters:
7807 +  mat - the matrix
7808 .  nsubcomm - the number of subcommunicators (= number of redundant pareallel or sequential matrices)
7809 .  subcomm - MPI communicator split from the communicator where mat resides in
7810 .  mlocal_red - number of local rows of the redundant matrix
7811 -  reuse - either MAT_INITIAL_MATRIX or MAT_REUSE_MATRIX
7812 
7813    Output Parameter:
7814 .  matredundant - redundant matrix
7815 
7816    Notes:
7817    MAT_REUSE_MATRIX can only be used when the nonzero structure of the
7818    original matrix has not changed from that last call to MatGetRedundantMatrix().
7819 
7820    This routine creates the duplicated matrices in subcommunicators; you should NOT create them before
7821    calling it.
7822 
7823    Only MPIAIJ matrix is supported.
7824 
7825    Level: advanced
7826 
7827    Concepts: subcommunicator
7828    Concepts: duplicate matrix
7829 
7830 .seealso: MatDestroy()
7831 @*/
7832 PetscErrorCode PETSCMAT_DLLEXPORT MatGetRedundantMatrix(Mat mat,PetscInt nsubcomm,MPI_Comm subcomm,PetscInt mlocal_red,MatReuse reuse,Mat *matredundant)
7833 {
7834   PetscErrorCode ierr;
7835 
7836   PetscFunctionBegin;
7837   PetscValidHeaderSpecific(mat,MAT_COOKIE,1);
7838   if (nsubcomm && reuse == MAT_REUSE_MATRIX) {
7839     PetscValidPointer(*matredundant,6);
7840     PetscValidHeaderSpecific(*matredundant,MAT_COOKIE,6);
7841   }
7842   if (!mat->ops->getredundantmatrix) SETERRQ1(PETSC_ERR_SUP,"Mat type %s",((PetscObject)mat)->type_name);
7843   if (!mat->assembled) SETERRQ(PETSC_ERR_ARG_WRONGSTATE,"Not for unassembled matrix");
7844   if (mat->factor) SETERRQ(PETSC_ERR_ARG_WRONGSTATE,"Not for factored matrix");
7845   ierr = MatPreallocated(mat);CHKERRQ(ierr);
7846 
7847   ierr = PetscLogEventBegin(MAT_GetRedundantMatrix,mat,0,0,0);CHKERRQ(ierr);
7848   ierr = (*mat->ops->getredundantmatrix)(mat,nsubcomm,subcomm,mlocal_red,reuse,matredundant);CHKERRQ(ierr);
7849   ierr = PetscLogEventEnd(MAT_GetRedundantMatrix,mat,0,0,0);CHKERRQ(ierr);
7850   PetscFunctionReturn(0);
7851 }
7852