xref: /petsc/src/mat/interface/matrix.c (revision 5d973c197c8b7d308ce7b868381cbe50f94f9f77)
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 PetscClassId PETSCMAT_DLLEXPORT MAT_CLASSID;
12 PetscClassId PETSCMAT_DLLEXPORT MAT_FDCOLORING_CLASSID;
13 
14 PetscLogEvent  MAT_Mult, MAT_Mults, MAT_MultConstrained, MAT_MultAdd, MAT_MultTranspose;
15 PetscLogEvent  MAT_MultTransposeConstrained, MAT_MultTransposeAdd, MAT_Solve, MAT_Solves, MAT_SolveAdd, MAT_SolveTranspose, MAT_MatSolve;
16 PetscLogEvent  MAT_SolveTransposeAdd, MAT_SOR, MAT_ForwardSolve, MAT_BackwardSolve, MAT_LUFactor, MAT_LUFactorSymbolic;
17 PetscLogEvent  MAT_LUFactorNumeric, MAT_CholeskyFactor, MAT_CholeskyFactorSymbolic, MAT_CholeskyFactorNumeric, MAT_ILUFactor;
18 PetscLogEvent  MAT_ILUFactorSymbolic, MAT_ICCFactorSymbolic, MAT_Copy, MAT_Convert, MAT_Scale, MAT_AssemblyBegin;
19 PetscLogEvent  MAT_AssemblyEnd, MAT_SetValues, MAT_GetValues, MAT_GetRow, MAT_GetRowIJ, MAT_GetSubMatrices, MAT_GetColoring, MAT_GetOrdering, MAT_GetRedundantMatrix, MAT_GetSeqNonzeroStructure;
20 PetscLogEvent  MAT_IncreaseOverlap, MAT_Partitioning, MAT_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_MultHermitianTranspose,MAT_MultHermitianTransposeAdd;
26 PetscLogEvent  MAT_Getsymtranspose, MAT_Getsymtransreduced, MAT_Transpose_SeqAIJ, MAT_GetBrowsOfAcols;
27 PetscLogEvent  MAT_GetBrowsOfAocols, MAT_Getlocalmat, MAT_Getlocalmatcondensed, MAT_Seqstompi, MAT_Seqstompinum, MAT_Seqstompisym;
28 PetscLogEvent  MAT_Applypapt, MAT_Applypapt_numeric, MAT_Applypapt_symbolic, MAT_GetSequentialNonzeroStructure;
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 const char *const MatFactorTypes[] = {"NONE","LU","CHOLESKY","ILU","ICC","ILUDT","MatFactorType","MAT_FACTOR_",0};
36 
37 #undef __FUNCT__
38 #define __FUNCT__ "MatGetDiagonalBlock"
39 /*@
40    MatGetDiagonalBlock - Returns the part of the matrix associated with the on-process coupling
41 
42    Not Collective
43 
44    Input Parameters:
45 +  mat - the matrix
46 -  reuse - indicates you are passing in the a matrix and want it reused
47 
48    Output Parameters:
49 +   iscopy - indicates a copy of the diagonal matrix was created and you should use MatDestroy() on it
50 -   a - the diagonal part (which is a SEQUENTIAL matrix)
51 
52    Notes: see the manual page for MatCreateMPIAIJ() for more information on the "diagonal part" of the matrix
53 
54    Level: advanced
55 
56 @*/
57 PetscErrorCode PETSCMAT_DLLEXPORT MatGetDiagonalBlock(Mat A,PetscTruth *iscopy,MatReuse reuse,Mat *a)
58 {
59   PetscErrorCode ierr,(*f)(Mat,PetscTruth*,MatReuse,Mat*);
60   PetscMPIInt    size;
61 
62   PetscFunctionBegin;
63   PetscValidHeaderSpecific(A,MAT_CLASSID,1);
64   PetscValidType(A,1);
65   PetscValidPointer(iscopy,2);
66   PetscValidPointer(a,3);
67   if (!A->assembled) SETERRQ(((PetscObject)A)->comm,PETSC_ERR_ARG_WRONGSTATE,"Not for unassembled matrix");
68   if (A->factortype) SETERRQ(((PetscObject)A)->comm,PETSC_ERR_ARG_WRONGSTATE,"Not for factored matrix");
69   ierr = MPI_Comm_size(((PetscObject)A)->comm,&size);CHKERRQ(ierr);
70   ierr = PetscObjectQueryFunction((PetscObject)A,"MatGetDiagonalBlock_C",(void (**)(void))&f);CHKERRQ(ierr);
71   if (f) {
72     ierr = (*f)(A,iscopy,reuse,a);CHKERRQ(ierr);
73   } else if (size == 1) {
74     *a = A;
75     *iscopy = PETSC_FALSE;
76   } else SETERRQ(((PetscObject)A)->comm,PETSC_ERR_SUP,"Cannot get diagonal part for this matrix");
77   PetscFunctionReturn(0);
78 }
79 
80 #undef __FUNCT__
81 #define __FUNCT__ "MatGetTrace"
82 /*@
83    MatGetTrace - Gets the trace of a matrix. The sum of the diagonal entries.
84 
85    Collective on Mat
86 
87    Input Parameters:
88 .  mat - the matrix
89 
90    Output Parameter:
91 .   trace - the sum of the diagonal entries
92 
93    Level: advanced
94 
95 @*/
96 PetscErrorCode  MatGetTrace(Mat mat,PetscScalar *trace)
97 {
98    PetscErrorCode ierr;
99    Vec            diag;
100 
101    PetscFunctionBegin;
102    ierr = MatGetVecs(mat,&diag,PETSC_NULL);CHKERRQ(ierr);
103    ierr = MatGetDiagonal(mat,diag);CHKERRQ(ierr);
104    ierr = VecSum(diag,trace);CHKERRQ(ierr);
105    ierr = VecDestroy(diag);CHKERRQ(ierr);
106    PetscFunctionReturn(0);
107 }
108 
109 #undef __FUNCT__
110 #define __FUNCT__ "MatRealPart"
111 /*@
112    MatRealPart - Zeros out the imaginary part of the matrix
113 
114    Collective on Mat
115 
116    Input Parameters:
117 .  mat - the matrix
118 
119    Level: advanced
120 
121 
122 .seealso: MatImaginaryPart()
123 @*/
124 PetscErrorCode PETSCMAT_DLLEXPORT MatRealPart(Mat mat)
125 {
126   PetscErrorCode ierr;
127 
128   PetscFunctionBegin;
129   PetscValidHeaderSpecific(mat,MAT_CLASSID,1);
130   PetscValidType(mat,1);
131   if (!mat->assembled) SETERRQ(((PetscObject)mat)->comm,PETSC_ERR_ARG_WRONGSTATE,"Not for unassembled matrix");
132   if (mat->factortype) SETERRQ(((PetscObject)mat)->comm,PETSC_ERR_ARG_WRONGSTATE,"Not for factored matrix");
133   if (!mat->ops->realpart) SETERRQ1(((PetscObject)mat)->comm,PETSC_ERR_SUP,"Mat type %s",((PetscObject)mat)->type_name);
134   ierr = MatPreallocated(mat);CHKERRQ(ierr);
135   ierr = (*mat->ops->realpart)(mat);CHKERRQ(ierr);
136   PetscFunctionReturn(0);
137 }
138 
139 #undef __FUNCT__
140 #define __FUNCT__ "MatGetGhosts"
141 /*@C
142    MatGetGhosts - Get the global index of all ghost nodes defined by the sparse matrix
143 
144    Collective on Mat
145 
146    Input Parameter:
147 .  mat - the matrix
148 
149    Output Parameters:
150 +   nghosts - number of ghosts (note for BAIJ matrices there is one ghost for each block)
151 -   ghosts - the global indices of the ghost points
152 
153    Notes: the nghosts and ghosts are suitable to pass into VecCreateGhost()
154 
155    Level: advanced
156 
157 @*/
158 PetscErrorCode PETSCMAT_DLLEXPORT MatGetGhosts(Mat mat,PetscInt *nghosts,const PetscInt *ghosts[])
159 {
160   PetscErrorCode ierr;
161 
162   PetscFunctionBegin;
163   PetscValidHeaderSpecific(mat,MAT_CLASSID,1);
164   PetscValidType(mat,1);
165   if (!mat->assembled) SETERRQ(((PetscObject)mat)->comm,PETSC_ERR_ARG_WRONGSTATE,"Not for unassembled matrix");
166   if (mat->factortype) SETERRQ(((PetscObject)mat)->comm,PETSC_ERR_ARG_WRONGSTATE,"Not for factored matrix");
167   if (!mat->ops->getghosts) {
168     if (nghosts) *nghosts = 0;
169     if (ghosts) *ghosts = 0;
170   } else {
171     ierr = (*mat->ops->getghosts)(mat,nghosts,ghosts);CHKERRQ(ierr);
172   }
173   PetscFunctionReturn(0);
174 }
175 
176 
177 #undef __FUNCT__
178 #define __FUNCT__ "MatImaginaryPart"
179 /*@
180    MatImaginaryPart - Moves the imaginary part of the matrix to the real part and zeros the imaginary part
181 
182    Collective on Mat
183 
184    Input Parameters:
185 .  mat - the matrix
186 
187    Level: advanced
188 
189 
190 .seealso: MatRealPart()
191 @*/
192 PetscErrorCode PETSCMAT_DLLEXPORT MatImaginaryPart(Mat mat)
193 {
194   PetscErrorCode ierr;
195 
196   PetscFunctionBegin;
197   PetscValidHeaderSpecific(mat,MAT_CLASSID,1);
198   PetscValidType(mat,1);
199   if (!mat->assembled) SETERRQ(((PetscObject)mat)->comm,PETSC_ERR_ARG_WRONGSTATE,"Not for unassembled matrix");
200   if (mat->factortype) SETERRQ(((PetscObject)mat)->comm,PETSC_ERR_ARG_WRONGSTATE,"Not for factored matrix");
201   if (!mat->ops->imaginarypart) SETERRQ1(((PetscObject)mat)->comm,PETSC_ERR_SUP,"Mat type %s",((PetscObject)mat)->type_name);
202   ierr = MatPreallocated(mat);CHKERRQ(ierr);
203   ierr = (*mat->ops->imaginarypart)(mat);CHKERRQ(ierr);
204   PetscFunctionReturn(0);
205 }
206 
207 #undef __FUNCT__
208 #define __FUNCT__ "MatMissingDiagonal"
209 /*@
210    MatMissingDiagonal - Determine if sparse matrix is missing a diagonal entry (or block entry for BAIJ matrices)
211 
212    Collective on Mat
213 
214    Input Parameter:
215 .  mat - the matrix
216 
217    Output Parameters:
218 +  missing - is any diagonal missing
219 -  dd - first diagonal entry that is missing (optional)
220 
221    Level: advanced
222 
223 
224 .seealso: MatRealPart()
225 @*/
226 PetscErrorCode PETSCMAT_DLLEXPORT MatMissingDiagonal(Mat mat,PetscTruth *missing,PetscInt *dd)
227 {
228   PetscErrorCode ierr;
229 
230   PetscFunctionBegin;
231   PetscValidHeaderSpecific(mat,MAT_CLASSID,1);
232   PetscValidType(mat,1);
233   if (!mat->assembled) SETERRQ(((PetscObject)mat)->comm,PETSC_ERR_ARG_WRONGSTATE,"Not for unassembled matrix");
234   if (mat->factortype) SETERRQ(((PetscObject)mat)->comm,PETSC_ERR_ARG_WRONGSTATE,"Not for factored matrix");
235   if (!mat->ops->missingdiagonal) SETERRQ1(((PetscObject)mat)->comm,PETSC_ERR_SUP,"Mat type %s",((PetscObject)mat)->type_name);
236   ierr = (*mat->ops->missingdiagonal)(mat,missing,dd);CHKERRQ(ierr);
237   PetscFunctionReturn(0);
238 }
239 
240 #undef __FUNCT__
241 #define __FUNCT__ "MatGetRow"
242 /*@C
243    MatGetRow - Gets a row of a matrix.  You MUST call MatRestoreRow()
244    for each row that you get to ensure that your application does
245    not bleed memory.
246 
247    Not Collective
248 
249    Input Parameters:
250 +  mat - the matrix
251 -  row - the row to get
252 
253    Output Parameters:
254 +  ncols -  if not NULL, the number of nonzeros in the row
255 .  cols - if not NULL, the column numbers
256 -  vals - if not NULL, the values
257 
258    Notes:
259    This routine is provided for people who need to have direct access
260    to the structure of a matrix.  We hope that we provide enough
261    high-level matrix routines that few users will need it.
262 
263    MatGetRow() always returns 0-based column indices, regardless of
264    whether the internal representation is 0-based (default) or 1-based.
265 
266    For better efficiency, set cols and/or vals to PETSC_NULL if you do
267    not wish to extract these quantities.
268 
269    The user can only examine the values extracted with MatGetRow();
270    the values cannot be altered.  To change the matrix entries, one
271    must use MatSetValues().
272 
273    You can only have one call to MatGetRow() outstanding for a particular
274    matrix at a time, per processor. MatGetRow() can only obtain rows
275    associated with the given processor, it cannot get rows from the
276    other processors; for that we suggest using MatGetSubMatrices(), then
277    MatGetRow() on the submatrix. The row indix passed to MatGetRows()
278    is in the global number of rows.
279 
280    Fortran Notes:
281    The calling sequence from Fortran is
282 .vb
283    MatGetRow(matrix,row,ncols,cols,values,ierr)
284          Mat     matrix (input)
285          integer row    (input)
286          integer ncols  (output)
287          integer cols(maxcols) (output)
288          double precision (or double complex) values(maxcols) output
289 .ve
290    where maxcols >= maximum nonzeros in any row of the matrix.
291 
292 
293    Caution:
294    Do not try to change the contents of the output arrays (cols and vals).
295    In some cases, this may corrupt the matrix.
296 
297    Level: advanced
298 
299    Concepts: matrices^row access
300 
301 .seealso: MatRestoreRow(), MatSetValues(), MatGetValues(), MatGetSubMatrices(), MatGetDiagonal()
302 @*/
303 PetscErrorCode PETSCMAT_DLLEXPORT MatGetRow(Mat mat,PetscInt row,PetscInt *ncols,const PetscInt *cols[],const PetscScalar *vals[])
304 {
305   PetscErrorCode ierr;
306   PetscInt       incols;
307 
308   PetscFunctionBegin;
309   PetscValidHeaderSpecific(mat,MAT_CLASSID,1);
310   PetscValidType(mat,1);
311   if (!mat->assembled) SETERRQ(PETSC_COMM_SELF,PETSC_ERR_ARG_WRONGSTATE,"Not for unassembled matrix");
312   if (mat->factortype) SETERRQ(PETSC_COMM_SELF,PETSC_ERR_ARG_WRONGSTATE,"Not for factored matrix");
313   if (!mat->ops->getrow) SETERRQ1(PETSC_COMM_SELF,PETSC_ERR_SUP,"Mat type %s",((PetscObject)mat)->type_name);
314   ierr = MatPreallocated(mat);CHKERRQ(ierr);
315   ierr = PetscLogEventBegin(MAT_GetRow,mat,0,0,0);CHKERRQ(ierr);
316   ierr = (*mat->ops->getrow)(mat,row,&incols,(PetscInt **)cols,(PetscScalar **)vals);CHKERRQ(ierr);
317   if (ncols) *ncols = incols;
318   ierr = PetscLogEventEnd(MAT_GetRow,mat,0,0,0);CHKERRQ(ierr);
319   PetscFunctionReturn(0);
320 }
321 
322 #undef __FUNCT__
323 #define __FUNCT__ "MatConjugate"
324 /*@
325    MatConjugate - replaces the matrix values with their complex conjugates
326 
327    Collective on Mat
328 
329    Input Parameters:
330 .  mat - the matrix
331 
332    Level: advanced
333 
334 .seealso:  VecConjugate()
335 @*/
336 PetscErrorCode PETSCMAT_DLLEXPORT MatConjugate(Mat mat)
337 {
338   PetscErrorCode ierr;
339 
340   PetscFunctionBegin;
341   PetscValidHeaderSpecific(mat,MAT_CLASSID,1);
342   if (!mat->assembled) SETERRQ(((PetscObject)mat)->comm,PETSC_ERR_ARG_WRONGSTATE,"Not for unassembled matrix");
343   if (!mat->ops->conjugate) SETERRQ(((PetscObject)mat)->comm,PETSC_ERR_SUP,"Not provided for this matrix format, send email to petsc-maint@mcs.anl.gov");
344   ierr = (*mat->ops->conjugate)(mat);CHKERRQ(ierr);
345   PetscFunctionReturn(0);
346 }
347 
348 #undef __FUNCT__
349 #define __FUNCT__ "MatRestoreRow"
350 /*@C
351    MatRestoreRow - Frees any temporary space allocated by MatGetRow().
352 
353    Not Collective
354 
355    Input Parameters:
356 +  mat - the matrix
357 .  row - the row to get
358 .  ncols, cols - the number of nonzeros and their columns
359 -  vals - if nonzero the column values
360 
361    Notes:
362    This routine should be called after you have finished examining the entries.
363 
364    Fortran Notes:
365    The calling sequence from Fortran is
366 .vb
367    MatRestoreRow(matrix,row,ncols,cols,values,ierr)
368       Mat     matrix (input)
369       integer row    (input)
370       integer ncols  (output)
371       integer cols(maxcols) (output)
372       double precision (or double complex) values(maxcols) output
373 .ve
374    Where maxcols >= maximum nonzeros in any row of the matrix.
375 
376    In Fortran MatRestoreRow() MUST be called after MatGetRow()
377    before another call to MatGetRow() can be made.
378 
379    Level: advanced
380 
381 .seealso:  MatGetRow()
382 @*/
383 PetscErrorCode PETSCMAT_DLLEXPORT MatRestoreRow(Mat mat,PetscInt row,PetscInt *ncols,const PetscInt *cols[],const PetscScalar *vals[])
384 {
385   PetscErrorCode ierr;
386 
387   PetscFunctionBegin;
388   PetscValidHeaderSpecific(mat,MAT_CLASSID,1);
389   PetscValidIntPointer(ncols,3);
390   if (!mat->assembled) SETERRQ(PETSC_COMM_SELF,PETSC_ERR_ARG_WRONGSTATE,"Not for unassembled matrix");
391   if (!mat->ops->restorerow) PetscFunctionReturn(0);
392   ierr = (*mat->ops->restorerow)(mat,row,ncols,(PetscInt **)cols,(PetscScalar **)vals);CHKERRQ(ierr);
393   PetscFunctionReturn(0);
394 }
395 
396 #undef __FUNCT__
397 #define __FUNCT__ "MatGetRowUpperTriangular"
398 /*@
399    MatGetRowUpperTriangular - Sets a flag to enable calls to MatGetRow() for matrix in MATSBAIJ format.
400    You should call MatRestoreRowUpperTriangular() after calling MatGetRow/MatRestoreRow() to disable the flag.
401 
402    Not Collective
403 
404    Input Parameters:
405 +  mat - the matrix
406 
407    Notes:
408    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.
409 
410    Level: advanced
411 
412    Concepts: matrices^row access
413 
414 .seealso: MatRestoreRowRowUpperTriangular()
415 @*/
416 PetscErrorCode PETSCMAT_DLLEXPORT MatGetRowUpperTriangular(Mat mat)
417 {
418   PetscErrorCode ierr;
419 
420   PetscFunctionBegin;
421   PetscValidHeaderSpecific(mat,MAT_CLASSID,1);
422   PetscValidType(mat,1);
423   if (!mat->assembled) SETERRQ(PETSC_COMM_SELF,PETSC_ERR_ARG_WRONGSTATE,"Not for unassembled matrix");
424   if (mat->factortype) SETERRQ(PETSC_COMM_SELF,PETSC_ERR_ARG_WRONGSTATE,"Not for factored matrix");
425   if (!mat->ops->getrowuppertriangular) SETERRQ1(PETSC_COMM_SELF,PETSC_ERR_SUP,"Mat type %s",((PetscObject)mat)->type_name);
426   ierr = MatPreallocated(mat);CHKERRQ(ierr);
427   ierr = (*mat->ops->getrowuppertriangular)(mat);CHKERRQ(ierr);
428   PetscFunctionReturn(0);
429 }
430 
431 #undef __FUNCT__
432 #define __FUNCT__ "MatRestoreRowUpperTriangular"
433 /*@
434    MatRestoreRowUpperTriangular - Disable calls to MatGetRow() for matrix in MATSBAIJ format.
435 
436    Not Collective
437 
438    Input Parameters:
439 +  mat - the matrix
440 
441    Notes:
442    This routine should be called after you have finished MatGetRow/MatRestoreRow().
443 
444 
445    Level: advanced
446 
447 .seealso:  MatGetRowUpperTriangular()
448 @*/
449 PetscErrorCode PETSCMAT_DLLEXPORT MatRestoreRowUpperTriangular(Mat mat)
450 {
451   PetscErrorCode ierr;
452 
453   PetscFunctionBegin;
454   PetscValidHeaderSpecific(mat,MAT_CLASSID,1);
455   if (!mat->assembled) SETERRQ(PETSC_COMM_SELF,PETSC_ERR_ARG_WRONGSTATE,"Not for unassembled matrix");
456   if (!mat->ops->restorerowuppertriangular) PetscFunctionReturn(0);
457   ierr = (*mat->ops->restorerowuppertriangular)(mat);CHKERRQ(ierr);
458   PetscFunctionReturn(0);
459 }
460 
461 #undef __FUNCT__
462 #define __FUNCT__ "MatSetOptionsPrefix"
463 /*@C
464    MatSetOptionsPrefix - Sets the prefix used for searching for all
465    Mat options in the database.
466 
467    Collective on Mat
468 
469    Input Parameter:
470 +  A - the Mat context
471 -  prefix - the prefix to prepend to all option names
472 
473    Notes:
474    A hyphen (-) must NOT be given at the beginning of the prefix name.
475    The first character of all runtime options is AUTOMATICALLY the hyphen.
476 
477    Level: advanced
478 
479 .keywords: Mat, set, options, prefix, database
480 
481 .seealso: MatSetFromOptions()
482 @*/
483 PetscErrorCode PETSCMAT_DLLEXPORT MatSetOptionsPrefix(Mat A,const char prefix[])
484 {
485   PetscErrorCode ierr;
486 
487   PetscFunctionBegin;
488   PetscValidHeaderSpecific(A,MAT_CLASSID,1);
489   ierr = PetscObjectSetOptionsPrefix((PetscObject)A,prefix);CHKERRQ(ierr);
490   PetscFunctionReturn(0);
491 }
492 
493 #undef __FUNCT__
494 #define __FUNCT__ "MatAppendOptionsPrefix"
495 /*@C
496    MatAppendOptionsPrefix - Appends to the prefix used for searching for all
497    Mat options in the database.
498 
499    Collective on Mat
500 
501    Input Parameters:
502 +  A - the Mat context
503 -  prefix - the prefix to prepend to all option names
504 
505    Notes:
506    A hyphen (-) must NOT be given at the beginning of the prefix name.
507    The first character of all runtime options is AUTOMATICALLY the hyphen.
508 
509    Level: advanced
510 
511 .keywords: Mat, append, options, prefix, database
512 
513 .seealso: MatGetOptionsPrefix()
514 @*/
515 PetscErrorCode PETSCMAT_DLLEXPORT MatAppendOptionsPrefix(Mat A,const char prefix[])
516 {
517   PetscErrorCode ierr;
518 
519   PetscFunctionBegin;
520   PetscValidHeaderSpecific(A,MAT_CLASSID,1);
521   ierr = PetscObjectAppendOptionsPrefix((PetscObject)A,prefix);CHKERRQ(ierr);
522   PetscFunctionReturn(0);
523 }
524 
525 #undef __FUNCT__
526 #define __FUNCT__ "MatGetOptionsPrefix"
527 /*@C
528    MatGetOptionsPrefix - Sets the prefix used for searching for all
529    Mat options in the database.
530 
531    Not Collective
532 
533    Input Parameter:
534 .  A - the Mat context
535 
536    Output Parameter:
537 .  prefix - pointer to the prefix string used
538 
539    Notes: On the fortran side, the user should pass in a string 'prefix' of
540    sufficient length to hold the prefix.
541 
542    Level: advanced
543 
544 .keywords: Mat, get, options, prefix, database
545 
546 .seealso: MatAppendOptionsPrefix()
547 @*/
548 PetscErrorCode PETSCMAT_DLLEXPORT MatGetOptionsPrefix(Mat A,const char *prefix[])
549 {
550   PetscErrorCode ierr;
551 
552   PetscFunctionBegin;
553   PetscValidHeaderSpecific(A,MAT_CLASSID,1);
554   ierr = PetscObjectGetOptionsPrefix((PetscObject)A,prefix);CHKERRQ(ierr);
555   PetscFunctionReturn(0);
556 }
557 
558 #undef __FUNCT__
559 #define __FUNCT__ "MatSetUp"
560 /*@
561    MatSetUp - Sets up the internal matrix data structures for the later use.
562 
563    Collective on Mat
564 
565    Input Parameters:
566 .  A - the Mat context
567 
568    Notes:
569    For basic use of the Mat classes the user need not explicitly call
570    MatSetUp(), since these actions will happen automatically.
571 
572    Level: advanced
573 
574 .keywords: Mat, setup
575 
576 .seealso: MatCreate(), MatDestroy()
577 @*/
578 PetscErrorCode PETSCMAT_DLLEXPORT MatSetUp(Mat A)
579 {
580   PetscMPIInt    size;
581   PetscErrorCode ierr;
582 
583   PetscFunctionBegin;
584   PetscValidHeaderSpecific(A,MAT_CLASSID,1);
585   if (!((PetscObject)A)->type_name) {
586     ierr = MPI_Comm_size(((PetscObject)A)->comm, &size);CHKERRQ(ierr);
587     if (size == 1) {
588       ierr = MatSetType(A, MATSEQAIJ);CHKERRQ(ierr);
589     } else {
590       ierr = MatSetType(A, MATMPIAIJ);CHKERRQ(ierr);
591     }
592   }
593   ierr = MatSetUpPreallocation(A);CHKERRQ(ierr);
594   PetscFunctionReturn(0);
595 }
596 
597 #undef __FUNCT__
598 #define __FUNCT__ "MatView"
599 /*@C
600    MatView - Visualizes a matrix object.
601 
602    Collective on Mat
603 
604    Input Parameters:
605 +  mat - the matrix
606 -  viewer - visualization context
607 
608   Notes:
609   The available visualization contexts include
610 +    PETSC_VIEWER_STDOUT_SELF - standard output (default)
611 .    PETSC_VIEWER_STDOUT_WORLD - synchronized standard
612         output where only the first processor opens
613         the file.  All other processors send their
614         data to the first processor to print.
615 -     PETSC_VIEWER_DRAW_WORLD - graphical display of nonzero structure
616 
617    The user can open alternative visualization contexts with
618 +    PetscViewerASCIIOpen() - Outputs matrix to a specified file
619 .    PetscViewerBinaryOpen() - Outputs matrix in binary to a
620          specified file; corresponding input uses MatLoad()
621 .    PetscViewerDrawOpen() - Outputs nonzero matrix structure to
622          an X window display
623 -    PetscViewerSocketOpen() - Outputs matrix to Socket viewer.
624          Currently only the sequential dense and AIJ
625          matrix types support the Socket viewer.
626 
627    The user can call PetscViewerSetFormat() to specify the output
628    format of ASCII printed objects (when using PETSC_VIEWER_STDOUT_SELF,
629    PETSC_VIEWER_STDOUT_WORLD and PetscViewerASCIIOpen).  Available formats include
630 +    PETSC_VIEWER_DEFAULT - default, prints matrix contents
631 .    PETSC_VIEWER_ASCII_MATLAB - prints matrix contents in Matlab format
632 .    PETSC_VIEWER_ASCII_DENSE - prints entire matrix including zeros
633 .    PETSC_VIEWER_ASCII_COMMON - prints matrix contents, using a sparse
634          format common among all matrix types
635 .    PETSC_VIEWER_ASCII_IMPL - prints matrix contents, using an implementation-specific
636          format (which is in many cases the same as the default)
637 .    PETSC_VIEWER_ASCII_INFO - prints basic information about the matrix
638          size and structure (not the matrix entries)
639 .    PETSC_VIEWER_ASCII_INFO_DETAIL - prints more detailed information about
640          the matrix structure
641 
642    Options Database Keys:
643 +  -mat_view_info - Prints info on matrix at conclusion of MatEndAssembly()
644 .  -mat_view_info_detailed - Prints more detailed info
645 .  -mat_view - Prints matrix in ASCII format
646 .  -mat_view_matlab - Prints matrix in Matlab format
647 .  -mat_view_draw - PetscDraws nonzero structure of matrix, using MatView() and PetscDrawOpenX().
648 .  -display <name> - Sets display name (default is host)
649 .  -draw_pause <sec> - Sets number of seconds to pause after display
650 .  -mat_view_socket - Sends matrix to socket, can be accessed from Matlab (see users manual)
651 .  -viewer_socket_machine <machine>
652 .  -viewer_socket_port <port>
653 .  -mat_view_binary - save matrix to file in binary format
654 -  -viewer_binary_filename <name>
655    Level: beginner
656 
657    Notes: see the manual page for MatLoad() for the exact format of the binary file when the binary
658       viewer is used.
659 
660       See bin/matlab/PetscBinaryRead.m for a Matlab code that can read in the binary file when the binary
661       viewer is used.
662 
663    Concepts: matrices^viewing
664    Concepts: matrices^plotting
665    Concepts: matrices^printing
666 
667 .seealso: PetscViewerSetFormat(), PetscViewerASCIIOpen(), PetscViewerDrawOpen(),
668           PetscViewerSocketOpen(), PetscViewerBinaryOpen(), MatLoad()
669 @*/
670 PetscErrorCode PETSCMAT_DLLEXPORT MatView(Mat mat,PetscViewer viewer)
671 {
672   PetscErrorCode    ierr;
673   PetscInt          rows,cols;
674   PetscTruth        iascii;
675   const MatType     cstr;
676   PetscViewerFormat format;
677 
678   PetscFunctionBegin;
679   PetscValidHeaderSpecific(mat,MAT_CLASSID,1);
680   PetscValidType(mat,1);
681   if (!viewer) {
682     ierr = PetscViewerASCIIGetStdout(((PetscObject)mat)->comm,&viewer);CHKERRQ(ierr);
683   }
684   PetscValidHeaderSpecific(viewer,PETSC_VIEWER_CLASSID,2);
685   PetscCheckSameComm(mat,1,viewer,2);
686   if (!mat->assembled) SETERRQ(((PetscObject)mat)->comm,PETSC_ERR_ORDER,"Must call MatAssemblyBegin/End() before viewing matrix");
687   ierr = MatPreallocated(mat);CHKERRQ(ierr);
688 
689   ierr = PetscLogEventBegin(MAT_View,mat,viewer,0,0);CHKERRQ(ierr);
690   ierr = PetscTypeCompare((PetscObject)viewer,PETSCVIEWERASCII,&iascii);CHKERRQ(ierr);
691   if (iascii) {
692     ierr = PetscViewerGetFormat(viewer,&format);CHKERRQ(ierr);
693     if (format == PETSC_VIEWER_ASCII_INFO || format == PETSC_VIEWER_ASCII_INFO_DETAIL) {
694       if (((PetscObject)mat)->prefix) {
695         ierr = PetscViewerASCIIPrintf(viewer,"Matrix Object:(%s)\n",((PetscObject)mat)->prefix);CHKERRQ(ierr);
696       } else {
697         ierr = PetscViewerASCIIPrintf(viewer,"Matrix Object:\n");CHKERRQ(ierr);
698       }
699       ierr = PetscViewerASCIIPushTab(viewer);CHKERRQ(ierr);
700       ierr = MatGetType(mat,&cstr);CHKERRQ(ierr);
701       ierr = MatGetSize(mat,&rows,&cols);CHKERRQ(ierr);
702       ierr = PetscViewerASCIIPrintf(viewer,"type=%s, rows=%D, cols=%D\n",cstr,rows,cols);CHKERRQ(ierr);
703       if (mat->factortype) {
704         const MatSolverPackage solver;
705         ierr = MatFactorGetSolverPackage(mat,&solver);CHKERRQ(ierr);
706         ierr = PetscViewerASCIIPrintf(viewer,"package used to perform factorization: %s\n",solver);CHKERRQ(ierr);
707       }
708       if (mat->ops->getinfo) {
709         MatInfo info;
710         ierr = MatGetInfo(mat,MAT_GLOBAL_SUM,&info);CHKERRQ(ierr);
711         ierr = PetscViewerASCIIPrintf(viewer,"total: nonzeros=%D, allocated nonzeros=%D\n",(PetscInt)info.nz_used,(PetscInt)info.nz_allocated);CHKERRQ(ierr);
712         ierr = PetscViewerASCIIPrintf(viewer,"total number of mallocs used during MatSetValues calls =%D\n",(PetscInt)info.mallocs);CHKERRQ(ierr);
713       }
714     }
715   }
716   if (mat->ops->view) {
717     ierr = PetscViewerASCIIPushTab(viewer);CHKERRQ(ierr);
718     ierr = (*mat->ops->view)(mat,viewer);CHKERRQ(ierr);
719     ierr = PetscViewerASCIIPopTab(viewer);CHKERRQ(ierr);
720   } else if (!iascii) {
721     SETERRQ1(PETSC_COMM_SELF,PETSC_ERR_SUP,"Viewer type %s not supported",((PetscObject)viewer)->type_name);
722   }
723   if (iascii) {
724     ierr = PetscViewerGetFormat(viewer,&format);CHKERRQ(ierr);
725     if (format == PETSC_VIEWER_ASCII_INFO || format == PETSC_VIEWER_ASCII_INFO_DETAIL) {
726       ierr = PetscViewerASCIIPopTab(viewer);CHKERRQ(ierr);
727     }
728   }
729   ierr = PetscLogEventEnd(MAT_View,mat,viewer,0,0);CHKERRQ(ierr);
730   PetscFunctionReturn(0);
731 }
732 
733 #if defined(PETSC_USE_DEBUG)
734 #include "../src/sys/totalview/tv_data_display.h"
735 PETSC_UNUSED static int TV_display_type(const struct _p_Mat *mat)
736 {
737   TV_add_row("Local rows", "int", &mat->rmap->n);
738   TV_add_row("Local columns", "int", &mat->cmap->n);
739   TV_add_row("Global rows", "int", &mat->rmap->N);
740   TV_add_row("Global columns", "int", &mat->cmap->N);
741   TV_add_row("Typename", TV_ascii_string_type, ((PetscObject)mat)->type_name);
742   return TV_format_OK;
743 }
744 #endif
745 
746 #undef __FUNCT__
747 #define __FUNCT__ "MatLoad"
748 /*@C
749    MatLoad - Loads a matrix that has been stored in binary format
750    with MatView().  The matrix format is determined from the options database.
751    Generates a parallel MPI matrix if the communicator has more than one
752    processor.  The default matrix type is AIJ.
753 
754    Collective on PetscViewer
755 
756    Input Parameters:
757 +  viewer - binary file viewer, created with PetscViewerBinaryOpen()
758 -  outtype - type of matrix desired, for example MATSEQAIJ,
759               MATMPISBAIJ etc.  See types in petsc/include/petscmat.h.
760 
761    Output Parameters:
762 .  newmat - new matrix
763 
764    Basic Options Database Keys:
765 +    -matload_type seqaij   - AIJ type
766 .    -matload_type mpiaij   - parallel AIJ type
767 .    -matload_type seqbaij  - block AIJ type
768 .    -matload_type mpibaij  - parallel block AIJ type
769 .    -matload_type seqsbaij - block symmetric AIJ type
770 .    -matload_type mpisbaij - parallel block symmetric AIJ type
771 .    -matload_type seqdense - dense type
772 .    -matload_type mpidense - parallel dense type
773 -    -matload_symmetric - matrix in file is symmetric
774 
775    More Options Database Keys:
776    Used with block matrix formats (MATSEQBAIJ,  ...) to specify
777    block size
778 .    -matload_block_size <bs>
779 
780    Level: beginner
781 
782    Notes:
783    MatLoad() automatically loads into the options database any options
784    given in the file filename.info where filename is the name of the file
785    that was passed to the PetscViewerBinaryOpen(). The options in the info
786    file will be ignored if you use the -viewer_binary_skip_info option.
787 
788    If the type or size of newmat is not set before a call to MatLoad, PETSc
789    sets the default matrix type AIJ and sets the local and global sizes.
790    If type and/or size is already set, then the same are used.
791 
792    In parallel, each processor can load a subset of rows (or the
793    entire matrix).  This routine is especially useful when a large
794    matrix is stored on disk and only part of it existsis desired on each
795    processor.  For example, a parallel solver may access only some of
796    the rows from each processor.  The algorithm used here reads
797    relatively small blocks of data rather than reading the entire
798    matrix and then subsetting it.
799 
800    Notes for advanced users:
801    Most users should not need to know the details of the binary storage
802    format, since MatLoad() and MatView() completely hide these details.
803    But for anyone who's interested, the standard binary matrix storage
804    format is
805 
806 $    int    MAT_FILE_CLASSID
807 $    int    number of rows
808 $    int    number of columns
809 $    int    total number of nonzeros
810 $    int    *number nonzeros in each row
811 $    int    *column indices of all nonzeros (starting index is zero)
812 $    PetscScalar *values of all nonzeros
813 
814    PETSc automatically does the byte swapping for
815 machines that store the bytes reversed, e.g.  DEC alpha, freebsd,
816 linux, Windows and the paragon; thus if you write your own binary
817 read/write routines you have to swap the bytes; see PetscBinaryRead()
818 and PetscBinaryWrite() to see how this may be done.
819 
820 .keywords: matrix, load, binary, input
821 
822 .seealso: PetscViewerBinaryOpen(), MatView(), VecLoad()
823 
824  @*/
825 PetscErrorCode PETSCMAT_DLLEXPORT MatLoad(PetscViewer viewer, Mat newmat)
826 {
827   PetscErrorCode ierr;
828   PetscTruth     isbinary,flg;
829   char           mtype[256];
830   const char     *prefix;
831   const MatType  outtype=0;
832 
833   PetscFunctionBegin;
834   PetscValidHeaderSpecific(viewer,PETSC_VIEWER_CLASSID,1);
835   PetscValidPointer(newmat,3);
836   ierr = PetscTypeCompare((PetscObject)viewer,PETSCVIEWERBINARY,&isbinary);CHKERRQ(ierr);
837   if (!isbinary) SETERRQ(PETSC_COMM_SELF,PETSC_ERR_ARG_WRONG,"Invalid viewer; open viewer with PetscViewerBinaryOpen()");
838 
839 
840   if (((PetscObject)newmat)->type_name) outtype = ((PetscObject)newmat)->type_name;
841   if (!outtype) {
842     ierr = PetscObjectGetOptionsPrefix((PetscObject)viewer,(const char **)&prefix);CHKERRQ(ierr);
843     ierr = PetscOptionsGetString(prefix,"-mat_type",mtype,256,&flg);CHKERRQ(ierr);
844     if (flg) {
845       outtype = mtype;
846     }
847     ierr = PetscOptionsGetString(prefix,"-matload_type",mtype,256,&flg);CHKERRQ(ierr);
848     if (flg) {
849       outtype = mtype;
850     }
851     if (!outtype) outtype = MATAIJ;
852     ierr = MatSetType(newmat,outtype);CHKERRQ(ierr);
853   }
854 
855   if (!newmat->ops->load) SETERRQ1(PETSC_COMM_SELF,PETSC_ERR_SUP,"MatLoad is not supported for type: %s",outtype);
856 
857   ierr = PetscLogEventBegin(MAT_Load,viewer,0,0,0);CHKERRQ(ierr);
858   ierr = (*newmat->ops->load)(viewer,newmat);CHKERRQ(ierr);
859   ierr = PetscLogEventEnd(MAT_Load,viewer,0,0,0);CHKERRQ(ierr);
860 
861   flg  = PETSC_FALSE;
862   ierr = PetscOptionsGetTruth(prefix,"-matload_symmetric",&flg,PETSC_NULL);CHKERRQ(ierr);
863   if (flg) {
864     ierr = MatSetOption(newmat,MAT_SYMMETRIC,PETSC_TRUE);CHKERRQ(ierr);
865     ierr = MatSetOption(newmat,MAT_SYMMETRY_ETERNAL,PETSC_TRUE);CHKERRQ(ierr);
866   }
867   PetscFunctionReturn(0);
868 }
869 
870 #undef __FUNCT__
871 #define __FUNCT__ "MatScaleSystem"
872 /*@
873    MatScaleSystem - Scale a vector solution and right hand side to
874    match the scaling of a scaled matrix.
875 
876    Collective on Mat
877 
878    Input Parameter:
879 +  mat - the matrix
880 .  b - right hand side vector (or PETSC_NULL)
881 -  x - solution vector (or PETSC_NULL)
882 
883 
884    Notes:
885    For AIJ, and BAIJ matrix formats, the matrices are not
886    internally scaled, so this does nothing.
887 
888    The KSP methods automatically call this routine when required
889    (via PCPreSolve()) so it is rarely used directly.
890 
891    Level: Developer
892 
893    Concepts: matrices^scaling
894 
895 .seealso: MatUseScaledForm(), MatUnScaleSystem()
896 @*/
897 PetscErrorCode PETSCMAT_DLLEXPORT MatScaleSystem(Mat mat,Vec b,Vec x)
898 {
899   PetscErrorCode ierr;
900 
901   PetscFunctionBegin;
902   PetscValidHeaderSpecific(mat,MAT_CLASSID,1);
903   PetscValidType(mat,1);
904   ierr = MatPreallocated(mat);CHKERRQ(ierr);
905   if (x) {PetscValidHeaderSpecific(x,VEC_CLASSID,3);PetscCheckSameComm(mat,1,x,3);}
906   if (b) {PetscValidHeaderSpecific(b,VEC_CLASSID,2);PetscCheckSameComm(mat,1,b,2);}
907 
908   if (mat->ops->scalesystem) {
909     ierr = (*mat->ops->scalesystem)(mat,b,x);CHKERRQ(ierr);
910   }
911   PetscFunctionReturn(0);
912 }
913 
914 #undef __FUNCT__
915 #define __FUNCT__ "MatUnScaleSystem"
916 /*@
917    MatUnScaleSystem - Unscales a vector solution and right hand side to
918    match the original scaling of a scaled matrix.
919 
920    Collective on Mat
921 
922    Input Parameter:
923 +  mat - the matrix
924 .  b - right hand side vector (or PETSC_NULL)
925 -  x - solution vector (or PETSC_NULL)
926 
927 
928    Notes:
929    For AIJ and BAIJ matrix formats, the matrices are not
930    internally scaled, so this does nothing.
931 
932    The KSP methods automatically call this routine when required
933    (via PCPreSolve()) so it is rarely used directly.
934 
935    Level: Developer
936 
937 .seealso: MatUseScaledForm(), MatScaleSystem()
938 @*/
939 PetscErrorCode PETSCMAT_DLLEXPORT MatUnScaleSystem(Mat mat,Vec b,Vec x)
940 {
941   PetscErrorCode ierr;
942 
943   PetscFunctionBegin;
944   PetscValidHeaderSpecific(mat,MAT_CLASSID,1);
945   PetscValidType(mat,1);
946   ierr = MatPreallocated(mat);CHKERRQ(ierr);
947   if (x) {PetscValidHeaderSpecific(x,VEC_CLASSID,3);PetscCheckSameComm(mat,1,x,3);}
948   if (b) {PetscValidHeaderSpecific(b,VEC_CLASSID,2);PetscCheckSameComm(mat,1,b,2);}
949   if (mat->ops->unscalesystem) {
950     ierr = (*mat->ops->unscalesystem)(mat,b,x);CHKERRQ(ierr);
951   }
952   PetscFunctionReturn(0);
953 }
954 
955 #undef __FUNCT__
956 #define __FUNCT__ "MatUseScaledForm"
957 /*@
958    MatUseScaledForm - For matrix storage formats that scale the
959    matrix indicates matrix operations (MatMult() etc) are
960    applied using the scaled matrix.
961 
962    Collective on Mat
963 
964    Input Parameter:
965 +  mat - the matrix
966 -  scaled - PETSC_TRUE for applying the scaled, PETSC_FALSE for
967             applying the original matrix
968 
969    Notes:
970    For scaled matrix formats, applying the original, unscaled matrix
971    will be slightly more expensive
972 
973    Level: Developer
974 
975 .seealso: MatScaleSystem(), MatUnScaleSystem()
976 @*/
977 PetscErrorCode PETSCMAT_DLLEXPORT MatUseScaledForm(Mat mat,PetscTruth scaled)
978 {
979   PetscErrorCode ierr;
980 
981   PetscFunctionBegin;
982   PetscValidHeaderSpecific(mat,MAT_CLASSID,1);
983   PetscValidType(mat,1);
984   ierr = MatPreallocated(mat);CHKERRQ(ierr);
985   if (mat->ops->usescaledform) {
986     ierr = (*mat->ops->usescaledform)(mat,scaled);CHKERRQ(ierr);
987   }
988   PetscFunctionReturn(0);
989 }
990 
991 #undef __FUNCT__
992 #define __FUNCT__ "MatDestroy"
993 /*@
994    MatDestroy - Frees space taken by a matrix.
995 
996    Collective on Mat
997 
998    Input Parameter:
999 .  A - the matrix
1000 
1001    Level: beginner
1002 
1003 @*/
1004 PetscErrorCode PETSCMAT_DLLEXPORT MatDestroy(Mat A)
1005 {
1006   PetscErrorCode ierr;
1007   PetscFunctionBegin;
1008   PetscValidHeaderSpecific(A,MAT_CLASSID,1);
1009   if (--((PetscObject)A)->refct > 0) PetscFunctionReturn(0);
1010   ierr = MatPreallocated(A);CHKERRQ(ierr);
1011   /* if memory was published with AMS then destroy it */
1012   ierr = PetscObjectDepublish(A);CHKERRQ(ierr);
1013   if (A->ops->destroy) {
1014     ierr = (*A->ops->destroy)(A);CHKERRQ(ierr);
1015   }
1016   if (A->mapping) {
1017     ierr = ISLocalToGlobalMappingDestroy(A->mapping);CHKERRQ(ierr);
1018   }
1019   if (A->bmapping) {
1020     ierr = ISLocalToGlobalMappingDestroy(A->bmapping);CHKERRQ(ierr);
1021   }
1022 
1023   if (A->spptr){ierr = PetscFree(A->spptr);CHKERRQ(ierr);}
1024   ierr = PetscLayoutDestroy(A->rmap);CHKERRQ(ierr);
1025   ierr = PetscLayoutDestroy(A->cmap);CHKERRQ(ierr);
1026   ierr = PetscHeaderDestroy(A);CHKERRQ(ierr);
1027   PetscFunctionReturn(0);
1028 }
1029 
1030 #undef __FUNCT__
1031 #define __FUNCT__ "MatValid"
1032 /*@
1033    MatValid - Checks whether a matrix object is valid.
1034 
1035    Collective on Mat
1036 
1037    Input Parameter:
1038 .  m - the matrix to check
1039 
1040    Output Parameter:
1041    flg - flag indicating matrix status, either
1042    PETSC_TRUE if matrix is valid, or PETSC_FALSE otherwise.
1043 
1044    Level: developer
1045 
1046    Concepts: matrices^validity
1047 @*/
1048 PetscErrorCode PETSCMAT_DLLEXPORT MatValid(Mat m,PetscTruth *flg)
1049 {
1050   PetscFunctionBegin;
1051   PetscValidIntPointer(flg,1);
1052   if (!m)                                          *flg = PETSC_FALSE;
1053   else if (((PetscObject)m)->classid != MAT_CLASSID) *flg = PETSC_FALSE;
1054   else                                             *flg = PETSC_TRUE;
1055   PetscFunctionReturn(0);
1056 }
1057 
1058 #undef __FUNCT__
1059 #define __FUNCT__ "MatSetValues"
1060 /*@
1061    MatSetValues - Inserts or adds a block of values into a matrix.
1062    These values may be cached, so MatAssemblyBegin() and MatAssemblyEnd()
1063    MUST be called after all calls to MatSetValues() have been completed.
1064 
1065    Not Collective
1066 
1067    Input Parameters:
1068 +  mat - the matrix
1069 .  v - a logically two-dimensional array of values
1070 .  m, idxm - the number of rows and their global indices
1071 .  n, idxn - the number of columns and their global indices
1072 -  addv - either ADD_VALUES or INSERT_VALUES, where
1073    ADD_VALUES adds values to any existing entries, and
1074    INSERT_VALUES replaces existing entries with new values
1075 
1076    Notes:
1077    By default the values, v, are row-oriented. See MatSetOption() for other options.
1078 
1079    Calls to MatSetValues() with the INSERT_VALUES and ADD_VALUES
1080    options cannot be mixed without intervening calls to the assembly
1081    routines.
1082 
1083    MatSetValues() uses 0-based row and column numbers in Fortran
1084    as well as in C.
1085 
1086    Negative indices may be passed in idxm and idxn, these rows and columns are
1087    simply ignored. This allows easily inserting element stiffness matrices
1088    with homogeneous Dirchlet boundary conditions that you don't want represented
1089    in the matrix.
1090 
1091    Efficiency Alert:
1092    The routine MatSetValuesBlocked() may offer much better efficiency
1093    for users of block sparse formats (MATSEQBAIJ and MATMPIBAIJ).
1094 
1095    Level: beginner
1096 
1097    Concepts: matrices^putting entries in
1098 
1099 .seealso: MatSetOption(), MatAssemblyBegin(), MatAssemblyEnd(), MatSetValuesBlocked(), MatSetValuesLocal(),
1100           InsertMode, INSERT_VALUES, ADD_VALUES
1101 @*/
1102 PetscErrorCode PETSCMAT_DLLEXPORT MatSetValues(Mat mat,PetscInt m,const PetscInt idxm[],PetscInt n,const PetscInt idxn[],const PetscScalar v[],InsertMode addv)
1103 {
1104   PetscErrorCode ierr;
1105 
1106   PetscFunctionBegin;
1107   PetscValidHeaderSpecific(mat,MAT_CLASSID,1);
1108   PetscValidType(mat,1);
1109   if (!m || !n) PetscFunctionReturn(0); /* no values to insert */
1110   PetscValidIntPointer(idxm,3);
1111   PetscValidIntPointer(idxn,5);
1112   if (v) PetscValidDoublePointer(v,6);
1113   ierr = MatPreallocated(mat);CHKERRQ(ierr);
1114   if (mat->insertmode == NOT_SET_VALUES) {
1115     mat->insertmode = addv;
1116   }
1117 #if defined(PETSC_USE_DEBUG)
1118   else if (mat->insertmode != addv) SETERRQ(PETSC_COMM_SELF,PETSC_ERR_ARG_WRONGSTATE,"Cannot mix add values and insert values");
1119   if (mat->factortype) SETERRQ(PETSC_COMM_SELF,PETSC_ERR_ARG_WRONGSTATE,"Not for factored matrix");
1120 #endif
1121 
1122   if (mat->assembled) {
1123     mat->was_assembled = PETSC_TRUE;
1124     mat->assembled     = PETSC_FALSE;
1125   }
1126   ierr = PetscLogEventBegin(MAT_SetValues,mat,0,0,0);CHKERRQ(ierr);
1127   if (!mat->ops->setvalues) SETERRQ1(PETSC_COMM_SELF,PETSC_ERR_SUP,"Mat type %s",((PetscObject)mat)->type_name);
1128   ierr = (*mat->ops->setvalues)(mat,m,idxm,n,idxn,v,addv);CHKERRQ(ierr);
1129   ierr = PetscLogEventEnd(MAT_SetValues,mat,0,0,0);CHKERRQ(ierr);
1130   PetscFunctionReturn(0);
1131 }
1132 
1133 
1134 #undef __FUNCT__
1135 #define __FUNCT__ "MatSetValuesRowLocal"
1136 /*@
1137    MatSetValuesRowLocal - Inserts a row (block row for BAIJ matrices) of nonzero
1138         values into a matrix
1139 
1140    Not Collective
1141 
1142    Input Parameters:
1143 +  mat - the matrix
1144 .  row - the (block) row to set
1145 -  v - a logically two-dimensional array of values
1146 
1147    Notes:
1148    By the values, v, are column-oriented (for the block version) and sorted
1149 
1150    All the nonzeros in the row must be provided
1151 
1152    The matrix must have previously had its column indices set
1153 
1154    The row must belong to this process
1155 
1156    Level: intermediate
1157 
1158    Concepts: matrices^putting entries in
1159 
1160 .seealso: MatSetOption(), MatAssemblyBegin(), MatAssemblyEnd(), MatSetValuesBlocked(), MatSetValuesLocal(),
1161           InsertMode, INSERT_VALUES, ADD_VALUES, MatSetValues(), MatSetValuesRow(), MatSetLocalToGlobalMapping()
1162 @*/
1163 PetscErrorCode PETSCMAT_DLLEXPORT MatSetValuesRowLocal(Mat mat,PetscInt row,const PetscScalar v[])
1164 {
1165   PetscErrorCode ierr;
1166 
1167   PetscFunctionBegin;
1168   PetscValidHeaderSpecific(mat,MAT_CLASSID,1);
1169   PetscValidType(mat,1);
1170   PetscValidScalarPointer(v,2);
1171   ierr = MatSetValuesRow(mat, mat->mapping->indices[row],v);CHKERRQ(ierr);
1172   PetscFunctionReturn(0);
1173 }
1174 
1175 #undef __FUNCT__
1176 #define __FUNCT__ "MatSetValuesRow"
1177 /*@
1178    MatSetValuesRow - Inserts a row (block row for BAIJ matrices) of nonzero
1179         values into a matrix
1180 
1181    Not Collective
1182 
1183    Input Parameters:
1184 +  mat - the matrix
1185 .  row - the (block) row to set
1186 -  v - a logically two-dimensional array of values
1187 
1188    Notes:
1189    The values, v, are column-oriented for the block version.
1190 
1191    All the nonzeros in the row must be provided
1192 
1193    THE MATRIX MUSAT HAVE PREVIOUSLY HAD ITS COLUMN INDICES SET. IT IS RARE THAT THIS ROUTINE IS USED, usually MatSetValues() is used.
1194 
1195    The row must belong to this process
1196 
1197    Level: advanced
1198 
1199    Concepts: matrices^putting entries in
1200 
1201 .seealso: MatSetOption(), MatAssemblyBegin(), MatAssemblyEnd(), MatSetValuesBlocked(), MatSetValuesLocal(),
1202           InsertMode, INSERT_VALUES, ADD_VALUES, MatSetValues()
1203 @*/
1204 PetscErrorCode PETSCMAT_DLLEXPORT MatSetValuesRow(Mat mat,PetscInt row,const PetscScalar v[])
1205 {
1206   PetscErrorCode ierr;
1207 
1208   PetscFunctionBegin;
1209   PetscValidHeaderSpecific(mat,MAT_CLASSID,1);
1210   PetscValidType(mat,1);
1211   PetscValidScalarPointer(v,2);
1212 #if defined(PETSC_USE_DEBUG)
1213   if (mat->insertmode == ADD_VALUES) SETERRQ(PETSC_COMM_SELF,PETSC_ERR_ARG_WRONGSTATE,"Cannot mix add and insert values");
1214   if (mat->factortype) SETERRQ(PETSC_COMM_SELF,PETSC_ERR_ARG_WRONGSTATE,"Not for factored matrix");
1215 #endif
1216   mat->insertmode = INSERT_VALUES;
1217 
1218   if (mat->assembled) {
1219     mat->was_assembled = PETSC_TRUE;
1220     mat->assembled     = PETSC_FALSE;
1221   }
1222   ierr = PetscLogEventBegin(MAT_SetValues,mat,0,0,0);CHKERRQ(ierr);
1223   if (!mat->ops->setvaluesrow) SETERRQ1(PETSC_COMM_SELF,PETSC_ERR_SUP,"Mat type %s",((PetscObject)mat)->type_name);
1224   ierr = (*mat->ops->setvaluesrow)(mat,row,v);CHKERRQ(ierr);
1225   ierr = PetscLogEventEnd(MAT_SetValues,mat,0,0,0);CHKERRQ(ierr);
1226   PetscFunctionReturn(0);
1227 }
1228 
1229 #undef __FUNCT__
1230 #define __FUNCT__ "MatSetValuesStencil"
1231 /*@
1232    MatSetValuesStencil - Inserts or adds a block of values into a matrix.
1233      Using structured grid indexing
1234 
1235    Not Collective
1236 
1237    Input Parameters:
1238 +  mat - the matrix
1239 .  m - number of rows being entered
1240 .  idxm - grid coordinates (and component number when dof > 1) for matrix rows being entered
1241 .  n - number of columns being entered
1242 .  idxn - grid coordinates (and component number when dof > 1) for matrix columns being entered
1243 .  v - a logically two-dimensional array of values
1244 -  addv - either ADD_VALUES or INSERT_VALUES, where
1245    ADD_VALUES adds values to any existing entries, and
1246    INSERT_VALUES replaces existing entries with new values
1247 
1248    Notes:
1249    By default the values, v, are row-oriented.  See MatSetOption() for other options.
1250 
1251    Calls to MatSetValuesStencil() with the INSERT_VALUES and ADD_VALUES
1252    options cannot be mixed without intervening calls to the assembly
1253    routines.
1254 
1255    The grid coordinates are across the entire grid, not just the local portion
1256 
1257    MatSetValuesStencil() uses 0-based row and column numbers in Fortran
1258    as well as in C.
1259 
1260    For setting/accessing vector values via array coordinates you can use the DAVecGetArray() routine
1261 
1262    In order to use this routine you must either obtain the matrix with DAGetMatrix()
1263    or call MatSetLocalToGlobalMapping() and MatSetStencil() first.
1264 
1265    The columns and rows in the stencil passed in MUST be contained within the
1266    ghost region of the given process as set with DACreateXXX() or MatSetStencil(). For example,
1267    if you create a DA with an overlap of one grid level and on a particular process its first
1268    local nonghost x logical coordinate is 6 (so its first ghost x logical coordinate is 5) the
1269    first i index you can use in your column and row indices in MatSetStencil() is 5.
1270 
1271    In Fortran idxm and idxn should be declared as
1272 $     MatStencil idxm(4,m),idxn(4,n)
1273    and the values inserted using
1274 $    idxm(MatStencil_i,1) = i
1275 $    idxm(MatStencil_j,1) = j
1276 $    idxm(MatStencil_k,1) = k
1277 $    idxm(MatStencil_c,1) = c
1278    etc
1279 
1280    For periodic boundary conditions use negative indices for values to the left (below 0; that are to be
1281    obtained by wrapping values from right edge). For values to the right of the last entry using that index plus one
1282    etc to obtain values that obtained by wrapping the values from the left edge. This does not work for the DA_NONPERIODIC
1283    wrap.
1284 
1285    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
1286    a single value per point) you can skip filling those indices.
1287 
1288    Inspired by the structured grid interface to the HYPRE package
1289    (http://www.llnl.gov/CASC/hypre)
1290 
1291    Efficiency Alert:
1292    The routine MatSetValuesBlockedStencil() may offer much better efficiency
1293    for users of block sparse formats (MATSEQBAIJ and MATMPIBAIJ).
1294 
1295    Level: beginner
1296 
1297    Concepts: matrices^putting entries in
1298 
1299 .seealso: MatSetOption(), MatAssemblyBegin(), MatAssemblyEnd(), MatSetValuesBlocked(), MatSetValuesLocal()
1300           MatSetValues(), MatSetValuesBlockedStencil(), MatSetStencil(), DAGetMatrix(), DAVecGetArray(), MatStencil
1301 @*/
1302 PetscErrorCode PETSCMAT_DLLEXPORT MatSetValuesStencil(Mat mat,PetscInt m,const MatStencil idxm[],PetscInt n,const MatStencil idxn[],const PetscScalar v[],InsertMode addv)
1303 {
1304   PetscErrorCode ierr;
1305   PetscInt       j,i,jdxm[128],jdxn[256],dim = mat->stencil.dim,*dims = mat->stencil.dims+1,tmp;
1306   PetscInt       *starts = mat->stencil.starts,*dxm = (PetscInt*)idxm,*dxn = (PetscInt*)idxn,sdim = dim - (1 - (PetscInt)mat->stencil.noc);
1307 
1308   PetscFunctionBegin;
1309   if (!m || !n) PetscFunctionReturn(0); /* no values to insert */
1310   PetscValidHeaderSpecific(mat,MAT_CLASSID,1);
1311   PetscValidType(mat,1);
1312   PetscValidIntPointer(idxm,3);
1313   PetscValidIntPointer(idxn,5);
1314   PetscValidScalarPointer(v,6);
1315 
1316   if (m > 128) SETERRQ1(PETSC_COMM_SELF,PETSC_ERR_SUP,"Can only set 128 rows at a time; trying to set %D",m);
1317   if (n > 256) SETERRQ1(PETSC_COMM_SELF,PETSC_ERR_SUP,"Can only set 256 columns at a time; trying to set %D",n);
1318 
1319   for (i=0; i<m; i++) {
1320     for (j=0; j<3-sdim; j++) dxm++;
1321     tmp = *dxm++ - starts[0];
1322     for (j=0; j<dim-1; j++) {
1323       if ((*dxm++ - starts[j+1]) < 0 || tmp < 0) tmp = PETSC_MIN_INT;
1324       else                                       tmp = tmp*dims[j] + *(dxm-1) - starts[j+1];
1325     }
1326     if (mat->stencil.noc) dxm++;
1327     jdxm[i] = tmp;
1328   }
1329   for (i=0; i<n; i++) {
1330     for (j=0; j<3-sdim; j++) dxn++;
1331     tmp = *dxn++ - starts[0];
1332     for (j=0; j<dim-1; j++) {
1333       if ((*dxn++ - starts[j+1]) < 0 || tmp < 0) tmp = PETSC_MIN_INT;
1334       else                                       tmp = tmp*dims[j] + *(dxn-1) - starts[j+1];
1335     }
1336     if (mat->stencil.noc) dxn++;
1337     jdxn[i] = tmp;
1338   }
1339   ierr = MatSetValuesLocal(mat,m,jdxm,n,jdxn,v,addv);CHKERRQ(ierr);
1340   PetscFunctionReturn(0);
1341 }
1342 
1343 #undef __FUNCT__
1344 #define __FUNCT__ "MatSetValuesBlockedStencil"
1345 /*@C
1346    MatSetValuesBlockedStencil - Inserts or adds a block of values into a matrix.
1347      Using structured grid indexing
1348 
1349    Not Collective
1350 
1351    Input Parameters:
1352 +  mat - the matrix
1353 .  m - number of rows being entered
1354 .  idxm - grid coordinates for matrix rows being entered
1355 .  n - number of columns being entered
1356 .  idxn - grid coordinates for matrix columns being entered
1357 .  v - a logically two-dimensional array of values
1358 -  addv - either ADD_VALUES or INSERT_VALUES, where
1359    ADD_VALUES adds values to any existing entries, and
1360    INSERT_VALUES replaces existing entries with new values
1361 
1362    Notes:
1363    By default the values, v, are row-oriented and unsorted.
1364    See MatSetOption() for other options.
1365 
1366    Calls to MatSetValuesBlockedStencil() with the INSERT_VALUES and ADD_VALUES
1367    options cannot be mixed without intervening calls to the assembly
1368    routines.
1369 
1370    The grid coordinates are across the entire grid, not just the local portion
1371 
1372    MatSetValuesBlockedStencil() uses 0-based row and column numbers in Fortran
1373    as well as in C.
1374 
1375    For setting/accessing vector values via array coordinates you can use the DAVecGetArray() routine
1376 
1377    In order to use this routine you must either obtain the matrix with DAGetMatrix()
1378    or call MatSetBlockSize(), MatSetLocalToGlobalMapping() and MatSetStencil() first.
1379 
1380    The columns and rows in the stencil passed in MUST be contained within the
1381    ghost region of the given process as set with DACreateXXX() or MatSetStencil(). For example,
1382    if you create a DA with an overlap of one grid level and on a particular process its first
1383    local nonghost x logical coordinate is 6 (so its first ghost x logical coordinate is 5) the
1384    first i index you can use in your column and row indices in MatSetStencil() is 5.
1385 
1386    In Fortran idxm and idxn should be declared as
1387 $     MatStencil idxm(4,m),idxn(4,n)
1388    and the values inserted using
1389 $    idxm(MatStencil_i,1) = i
1390 $    idxm(MatStencil_j,1) = j
1391 $    idxm(MatStencil_k,1) = k
1392    etc
1393 
1394    Negative indices may be passed in idxm and idxn, these rows and columns are
1395    simply ignored. This allows easily inserting element stiffness matrices
1396    with homogeneous Dirchlet boundary conditions that you don't want represented
1397    in the matrix.
1398 
1399    Inspired by the structured grid interface to the HYPRE package
1400    (http://www.llnl.gov/CASC/hypre)
1401 
1402    Level: beginner
1403 
1404    Concepts: matrices^putting entries in
1405 
1406 .seealso: MatSetOption(), MatAssemblyBegin(), MatAssemblyEnd(), MatSetValuesBlocked(), MatSetValuesLocal()
1407           MatSetValues(), MatSetValuesStencil(), MatSetStencil(), DAGetMatrix(), DAVecGetArray(), MatStencil,
1408           MatSetBlockSize(), MatSetLocalToGlobalMapping()
1409 @*/
1410 PetscErrorCode PETSCMAT_DLLEXPORT MatSetValuesBlockedStencil(Mat mat,PetscInt m,const MatStencil idxm[],PetscInt n,const MatStencil idxn[],const PetscScalar v[],InsertMode addv)
1411 {
1412   PetscErrorCode ierr;
1413   PetscInt       j,i,jdxm[128],jdxn[256],dim = mat->stencil.dim,*dims = mat->stencil.dims+1,tmp;
1414   PetscInt       *starts = mat->stencil.starts,*dxm = (PetscInt*)idxm,*dxn = (PetscInt*)idxn,sdim = dim - (1 - (PetscInt)mat->stencil.noc);
1415 
1416   PetscFunctionBegin;
1417   if (!m || !n) PetscFunctionReturn(0); /* no values to insert */
1418   PetscValidHeaderSpecific(mat,MAT_CLASSID,1);
1419   PetscValidType(mat,1);
1420   PetscValidIntPointer(idxm,3);
1421   PetscValidIntPointer(idxn,5);
1422   PetscValidScalarPointer(v,6);
1423 
1424   if (m > 128) SETERRQ1(PETSC_COMM_SELF,PETSC_ERR_SUP,"Can only set 128 rows at a time; trying to set %D",m);
1425   if (n > 128) SETERRQ1(PETSC_COMM_SELF,PETSC_ERR_SUP,"Can only set 256 columns at a time; trying to set %D",n);
1426 
1427   for (i=0; i<m; i++) {
1428     for (j=0; j<3-sdim; j++) dxm++;
1429     tmp = *dxm++ - starts[0];
1430     for (j=0; j<sdim-1; j++) {
1431       if ((*dxm++ - starts[j+1]) < 0 || tmp < 0) tmp = PETSC_MIN_INT;
1432       else                                      tmp = tmp*dims[j] + *(dxm-1) - starts[j+1];
1433     }
1434     dxm++;
1435     jdxm[i] = tmp;
1436   }
1437   for (i=0; i<n; i++) {
1438     for (j=0; j<3-sdim; j++) dxn++;
1439     tmp = *dxn++ - starts[0];
1440     for (j=0; j<sdim-1; j++) {
1441       if ((*dxn++ - starts[j+1]) < 0 || tmp < 0) tmp = PETSC_MIN_INT;
1442       else                                       tmp = tmp*dims[j] + *(dxn-1) - starts[j+1];
1443     }
1444     dxn++;
1445     jdxn[i] = tmp;
1446   }
1447   ierr = MatSetValuesBlockedLocal(mat,m,jdxm,n,jdxn,v,addv);CHKERRQ(ierr);
1448   PetscFunctionReturn(0);
1449 }
1450 
1451 #undef __FUNCT__
1452 #define __FUNCT__ "MatSetStencil"
1453 /*@
1454    MatSetStencil - Sets the grid information for setting values into a matrix via
1455         MatSetValuesStencil()
1456 
1457    Not Collective
1458 
1459    Input Parameters:
1460 +  mat - the matrix
1461 .  dim - dimension of the grid 1, 2, or 3
1462 .  dims - number of grid points in x, y, and z direction, including ghost points on your processor
1463 .  starts - starting point of ghost nodes on your processor in x, y, and z direction
1464 -  dof - number of degrees of freedom per node
1465 
1466 
1467    Inspired by the structured grid interface to the HYPRE package
1468    (www.llnl.gov/CASC/hyper)
1469 
1470    For matrices generated with DAGetMatrix() this routine is automatically called and so not needed by the
1471    user.
1472 
1473    Level: beginner
1474 
1475    Concepts: matrices^putting entries in
1476 
1477 .seealso: MatSetOption(), MatAssemblyBegin(), MatAssemblyEnd(), MatSetValuesBlocked(), MatSetValuesLocal()
1478           MatSetValues(), MatSetValuesBlockedStencil(), MatSetValuesStencil()
1479 @*/
1480 PetscErrorCode PETSCMAT_DLLEXPORT MatSetStencil(Mat mat,PetscInt dim,const PetscInt dims[],const PetscInt starts[],PetscInt dof)
1481 {
1482   PetscInt i;
1483 
1484   PetscFunctionBegin;
1485   PetscValidHeaderSpecific(mat,MAT_CLASSID,1);
1486   PetscValidIntPointer(dims,3);
1487   PetscValidIntPointer(starts,4);
1488 
1489   mat->stencil.dim = dim + (dof > 1);
1490   for (i=0; i<dim; i++) {
1491     mat->stencil.dims[i]   = dims[dim-i-1];      /* copy the values in backwards */
1492     mat->stencil.starts[i] = starts[dim-i-1];
1493   }
1494   mat->stencil.dims[dim]   = dof;
1495   mat->stencil.starts[dim] = 0;
1496   mat->stencil.noc         = (PetscTruth)(dof == 1);
1497   PetscFunctionReturn(0);
1498 }
1499 
1500 #undef __FUNCT__
1501 #define __FUNCT__ "MatSetValuesBlocked"
1502 /*@
1503    MatSetValuesBlocked - Inserts or adds a block of values into a matrix.
1504 
1505    Not Collective
1506 
1507    Input Parameters:
1508 +  mat - the matrix
1509 .  v - a logically two-dimensional array of values
1510 .  m, idxm - the number of block rows and their global block indices
1511 .  n, idxn - the number of block columns and their global block indices
1512 -  addv - either ADD_VALUES or INSERT_VALUES, where
1513    ADD_VALUES adds values to any existing entries, and
1514    INSERT_VALUES replaces existing entries with new values
1515 
1516    Notes:
1517    The m and n count the NUMBER of blocks in the row direction and column direction,
1518    NOT the total number of rows/columns; for example, if the block size is 2 and
1519    you are passing in values for rows 2,3,4,5  then m would be 2 (not 4).
1520    The values in idxm would be 1 2; that is the first index for each block divided by
1521    the block size.
1522 
1523    Note that you must call MatSetBlockSize() when constructing this matrix (after
1524    preallocating it).
1525 
1526    By default the values, v, are row-oriented, so the layout of
1527    v is the same as for MatSetValues(). See MatSetOption() for other options.
1528 
1529    Calls to MatSetValuesBlocked() with the INSERT_VALUES and ADD_VALUES
1530    options cannot be mixed without intervening calls to the assembly
1531    routines.
1532 
1533    MatSetValuesBlocked() uses 0-based row and column numbers in Fortran
1534    as well as in C.
1535 
1536    Negative indices may be passed in idxm and idxn, these rows and columns are
1537    simply ignored. This allows easily inserting element stiffness matrices
1538    with homogeneous Dirchlet boundary conditions that you don't want represented
1539    in the matrix.
1540 
1541    Each time an entry is set within a sparse matrix via MatSetValues(),
1542    internal searching must be done to determine where to place the the
1543    data in the matrix storage space.  By instead inserting blocks of
1544    entries via MatSetValuesBlocked(), the overhead of matrix assembly is
1545    reduced.
1546 
1547    Example:
1548 $   Suppose m=n=2 and block size(bs) = 2 The array is
1549 $
1550 $   1  2  | 3  4
1551 $   5  6  | 7  8
1552 $   - - - | - - -
1553 $   9  10 | 11 12
1554 $   13 14 | 15 16
1555 $
1556 $   v[] should be passed in like
1557 $   v[] = [1,2,3,4,5,6,7,8,9,10,11,12,13,14,15,16]
1558 $
1559 $  If you are not using row oriented storage of v (that is you called MatSetOption(mat,MAT_ROW_ORIENTED,PETSC_FALSE)) then
1560 $   v[] = [1,5,9,13,2,6,10,14,3,7,11,15,4,8,12,16]
1561 
1562    Level: intermediate
1563 
1564    Concepts: matrices^putting entries in blocked
1565 
1566 .seealso: MatSetBlockSize(), MatSetOption(), MatAssemblyBegin(), MatAssemblyEnd(), MatSetValues(), MatSetValuesBlockedLocal()
1567 @*/
1568 PetscErrorCode PETSCMAT_DLLEXPORT MatSetValuesBlocked(Mat mat,PetscInt m,const PetscInt idxm[],PetscInt n,const PetscInt idxn[],const PetscScalar v[],InsertMode addv)
1569 {
1570   PetscErrorCode ierr;
1571 
1572   PetscFunctionBegin;
1573   PetscValidHeaderSpecific(mat,MAT_CLASSID,1);
1574   PetscValidType(mat,1);
1575   if (!m || !n) PetscFunctionReturn(0); /* no values to insert */
1576   PetscValidIntPointer(idxm,3);
1577   PetscValidIntPointer(idxn,5);
1578   PetscValidScalarPointer(v,6);
1579   ierr = MatPreallocated(mat);CHKERRQ(ierr);
1580   if (mat->insertmode == NOT_SET_VALUES) {
1581     mat->insertmode = addv;
1582   }
1583 #if defined(PETSC_USE_DEBUG)
1584   else if (mat->insertmode != addv) SETERRQ(PETSC_COMM_SELF,PETSC_ERR_ARG_WRONGSTATE,"Cannot mix add values and insert values");
1585   if (mat->factortype) SETERRQ(PETSC_COMM_SELF,PETSC_ERR_ARG_WRONGSTATE,"Not for factored matrix");
1586 #endif
1587 
1588   if (mat->assembled) {
1589     mat->was_assembled = PETSC_TRUE;
1590     mat->assembled     = PETSC_FALSE;
1591   }
1592   ierr = PetscLogEventBegin(MAT_SetValues,mat,0,0,0);CHKERRQ(ierr);
1593   if (mat->ops->setvaluesblocked) {
1594     ierr = (*mat->ops->setvaluesblocked)(mat,m,idxm,n,idxn,v,addv);CHKERRQ(ierr);
1595   } else {
1596     PetscInt buf[4096],*ibufm=0,*ibufn=0;
1597     PetscInt i,j,*iidxm,*iidxn,bs=mat->rmap->bs;
1598     if ((m+n)*bs <= 4096) {
1599       iidxm = buf; iidxn = buf + m*bs;
1600     } else {
1601       ierr = PetscMalloc2(m*bs,PetscInt,&ibufm,n*bs,PetscInt,&ibufn);CHKERRQ(ierr);
1602       iidxm = ibufm; iidxn = ibufn;
1603     }
1604     for (i=0; i<m; i++) {
1605       for (j=0; j<bs; j++) {
1606 	iidxm[i*bs+j] = bs*idxm[i] + j;
1607       }
1608     }
1609     for (i=0; i<n; i++) {
1610       for (j=0; j<bs; j++) {
1611 	iidxn[i*bs+j] = bs*idxn[i] + j;
1612       }
1613     }
1614     ierr = MatSetValues(mat,bs*m,iidxm,bs*n,iidxn,v,addv);CHKERRQ(ierr);
1615     ierr = PetscFree2(ibufm,ibufn);CHKERRQ(ierr);
1616   }
1617   ierr = PetscLogEventEnd(MAT_SetValues,mat,0,0,0);CHKERRQ(ierr);
1618   PetscFunctionReturn(0);
1619 }
1620 
1621 #undef __FUNCT__
1622 #define __FUNCT__ "MatGetValues"
1623 /*@
1624    MatGetValues - Gets a block of values from a matrix.
1625 
1626    Not Collective; currently only returns a local block
1627 
1628    Input Parameters:
1629 +  mat - the matrix
1630 .  v - a logically two-dimensional array for storing the values
1631 .  m, idxm - the number of rows and their global indices
1632 -  n, idxn - the number of columns and their global indices
1633 
1634    Notes:
1635    The user must allocate space (m*n PetscScalars) for the values, v.
1636    The values, v, are then returned in a row-oriented format,
1637    analogous to that used by default in MatSetValues().
1638 
1639    MatGetValues() uses 0-based row and column numbers in
1640    Fortran as well as in C.
1641 
1642    MatGetValues() requires that the matrix has been assembled
1643    with MatAssemblyBegin()/MatAssemblyEnd().  Thus, calls to
1644    MatSetValues() and MatGetValues() CANNOT be made in succession
1645    without intermediate matrix assembly.
1646 
1647    Negative row or column indices will be ignored and those locations in v[] will be
1648    left unchanged.
1649 
1650    Level: advanced
1651 
1652    Concepts: matrices^accessing values
1653 
1654 .seealso: MatGetRow(), MatGetSubMatrices(), MatSetValues()
1655 @*/
1656 PetscErrorCode PETSCMAT_DLLEXPORT MatGetValues(Mat mat,PetscInt m,const PetscInt idxm[],PetscInt n,const PetscInt idxn[],PetscScalar v[])
1657 {
1658   PetscErrorCode ierr;
1659 
1660   PetscFunctionBegin;
1661   PetscValidHeaderSpecific(mat,MAT_CLASSID,1);
1662   PetscValidType(mat,1);
1663   if (!m || !n) PetscFunctionReturn(0);
1664   PetscValidIntPointer(idxm,3);
1665   PetscValidIntPointer(idxn,5);
1666   PetscValidScalarPointer(v,6);
1667   if (!mat->assembled) SETERRQ(PETSC_COMM_SELF,PETSC_ERR_ARG_WRONGSTATE,"Not for unassembled matrix");
1668   if (mat->factortype) SETERRQ(PETSC_COMM_SELF,PETSC_ERR_ARG_WRONGSTATE,"Not for factored matrix");
1669   if (!mat->ops->getvalues) SETERRQ1(PETSC_COMM_SELF,PETSC_ERR_SUP,"Mat type %s",((PetscObject)mat)->type_name);
1670   ierr = MatPreallocated(mat);CHKERRQ(ierr);
1671 
1672   ierr = PetscLogEventBegin(MAT_GetValues,mat,0,0,0);CHKERRQ(ierr);
1673   ierr = (*mat->ops->getvalues)(mat,m,idxm,n,idxn,v);CHKERRQ(ierr);
1674   ierr = PetscLogEventEnd(MAT_GetValues,mat,0,0,0);CHKERRQ(ierr);
1675   PetscFunctionReturn(0);
1676 }
1677 
1678 #undef __FUNCT__
1679 #define __FUNCT__ "MatSetLocalToGlobalMapping"
1680 /*@
1681    MatSetLocalToGlobalMapping - Sets a local-to-global numbering for use by
1682    the routine MatSetValuesLocal() to allow users to insert matrix entries
1683    using a local (per-processor) numbering.
1684 
1685    Not Collective
1686 
1687    Input Parameters:
1688 +  x - the matrix
1689 -  mapping - mapping created with ISLocalToGlobalMappingCreate()
1690              or ISLocalToGlobalMappingCreateIS()
1691 
1692    Level: intermediate
1693 
1694    Concepts: matrices^local to global mapping
1695    Concepts: local to global mapping^for matrices
1696 
1697 .seealso:  MatAssemblyBegin(), MatAssemblyEnd(), MatSetValues(), MatSetValuesLocal()
1698 @*/
1699 PetscErrorCode PETSCMAT_DLLEXPORT MatSetLocalToGlobalMapping(Mat x,ISLocalToGlobalMapping mapping)
1700 {
1701   PetscErrorCode ierr;
1702   PetscFunctionBegin;
1703   PetscValidHeaderSpecific(x,MAT_CLASSID,1);
1704   PetscValidType(x,1);
1705   PetscValidHeaderSpecific(mapping,IS_LTOGM_CLASSID,2);
1706   if (x->mapping) SETERRQ(PETSC_COMM_SELF,PETSC_ERR_ARG_WRONGSTATE,"Mapping already set for matrix");
1707   ierr = MatPreallocated(x);CHKERRQ(ierr);
1708 
1709   if (x->ops->setlocaltoglobalmapping) {
1710     ierr = (*x->ops->setlocaltoglobalmapping)(x,mapping);CHKERRQ(ierr);
1711   } else {
1712     ierr = PetscObjectReference((PetscObject)mapping);CHKERRQ(ierr);
1713     if (x->mapping) { ierr = ISLocalToGlobalMappingDestroy(x->mapping);CHKERRQ(ierr); }
1714     x->mapping = mapping;
1715   }
1716   PetscFunctionReturn(0);
1717 }
1718 
1719 #undef __FUNCT__
1720 #define __FUNCT__ "MatSetLocalToGlobalMappingBlock"
1721 /*@
1722    MatSetLocalToGlobalMappingBlock - Sets a local-to-global numbering for use
1723    by the routine MatSetValuesBlockedLocal() to allow users to insert matrix
1724    entries using a local (per-processor) numbering.
1725 
1726    Not Collective
1727 
1728    Input Parameters:
1729 +  x - the matrix
1730 -  mapping - mapping created with ISLocalToGlobalMappingCreate() or
1731              ISLocalToGlobalMappingCreateIS()
1732 
1733    Level: intermediate
1734 
1735    Concepts: matrices^local to global mapping blocked
1736    Concepts: local to global mapping^for matrices, blocked
1737 
1738 .seealso:  MatAssemblyBegin(), MatAssemblyEnd(), MatSetValues(), MatSetValuesBlockedLocal(),
1739            MatSetValuesBlocked(), MatSetValuesLocal()
1740 @*/
1741 PetscErrorCode PETSCMAT_DLLEXPORT MatSetLocalToGlobalMappingBlock(Mat x,ISLocalToGlobalMapping mapping)
1742 {
1743   PetscErrorCode ierr;
1744   PetscFunctionBegin;
1745   PetscValidHeaderSpecific(x,MAT_CLASSID,1);
1746   PetscValidType(x,1);
1747   PetscValidHeaderSpecific(mapping,IS_LTOGM_CLASSID,2);
1748   if (x->bmapping) SETERRQ(PETSC_COMM_SELF,PETSC_ERR_ARG_WRONGSTATE,"Mapping already set for matrix");
1749   ierr = PetscObjectReference((PetscObject)mapping);CHKERRQ(ierr);
1750   if (x->bmapping) { ierr = ISLocalToGlobalMappingDestroy(x->bmapping);CHKERRQ(ierr); }
1751   x->bmapping = mapping;
1752   PetscFunctionReturn(0);
1753 }
1754 
1755 #undef __FUNCT__
1756 #define __FUNCT__ "MatSetValuesLocal"
1757 /*@
1758    MatSetValuesLocal - Inserts or adds values into certain locations of a matrix,
1759    using a local ordering of the nodes.
1760 
1761    Not Collective
1762 
1763    Input Parameters:
1764 +  x - the matrix
1765 .  nrow, irow - number of rows and their local indices
1766 .  ncol, icol - number of columns and their local indices
1767 .  y -  a logically two-dimensional array of values
1768 -  addv - either INSERT_VALUES or ADD_VALUES, where
1769    ADD_VALUES adds values to any existing entries, and
1770    INSERT_VALUES replaces existing entries with new values
1771 
1772    Notes:
1773    Before calling MatSetValuesLocal(), the user must first set the
1774    local-to-global mapping by calling MatSetLocalToGlobalMapping().
1775 
1776    Calls to MatSetValuesLocal() with the INSERT_VALUES and ADD_VALUES
1777    options cannot be mixed without intervening calls to the assembly
1778    routines.
1779 
1780    These values may be cached, so MatAssemblyBegin() and MatAssemblyEnd()
1781    MUST be called after all calls to MatSetValuesLocal() have been completed.
1782 
1783    Level: intermediate
1784 
1785    Concepts: matrices^putting entries in with local numbering
1786 
1787 .seealso:  MatAssemblyBegin(), MatAssemblyEnd(), MatSetValues(), MatSetLocalToGlobalMapping(),
1788            MatSetValueLocal()
1789 @*/
1790 PetscErrorCode PETSCMAT_DLLEXPORT MatSetValuesLocal(Mat mat,PetscInt nrow,const PetscInt irow[],PetscInt ncol,const PetscInt icol[],const PetscScalar y[],InsertMode addv)
1791 {
1792   PetscErrorCode ierr;
1793   PetscInt       irowm[2048],icolm[2048];
1794 
1795   PetscFunctionBegin;
1796   PetscValidHeaderSpecific(mat,MAT_CLASSID,1);
1797   PetscValidType(mat,1);
1798   if (!nrow || !ncol) PetscFunctionReturn(0); /* no values to insert */
1799   PetscValidIntPointer(irow,3);
1800   PetscValidIntPointer(icol,5);
1801   PetscValidScalarPointer(y,6);
1802   ierr = MatPreallocated(mat);CHKERRQ(ierr);
1803   if (mat->insertmode == NOT_SET_VALUES) {
1804     mat->insertmode = addv;
1805   }
1806 #if defined(PETSC_USE_DEBUG)
1807   else if (mat->insertmode != addv) SETERRQ(PETSC_COMM_SELF,PETSC_ERR_ARG_WRONGSTATE,"Cannot mix add values and insert values");
1808   if (!mat->ops->setvalueslocal && (nrow > 2048 || ncol > 2048)) SETERRQ2(PETSC_COMM_SELF,PETSC_ERR_SUP,"Number column/row indices must be <= 2048: are %D %D",nrow,ncol);
1809   if (mat->factortype) SETERRQ(PETSC_COMM_SELF,PETSC_ERR_ARG_WRONGSTATE,"Not for factored matrix");
1810 #endif
1811 
1812   if (mat->assembled) {
1813     mat->was_assembled = PETSC_TRUE;
1814     mat->assembled     = PETSC_FALSE;
1815   }
1816   ierr = PetscLogEventBegin(MAT_SetValues,mat,0,0,0);CHKERRQ(ierr);
1817   if (!mat->ops->setvalueslocal) {
1818     ierr = ISLocalToGlobalMappingApply(mat->mapping,nrow,irow,irowm);CHKERRQ(ierr);
1819     ierr = ISLocalToGlobalMappingApply(mat->mapping,ncol,icol,icolm);CHKERRQ(ierr);
1820     ierr = (*mat->ops->setvalues)(mat,nrow,irowm,ncol,icolm,y,addv);CHKERRQ(ierr);
1821   } else {
1822     ierr = (*mat->ops->setvalueslocal)(mat,nrow,irow,ncol,icol,y,addv);CHKERRQ(ierr);
1823   }
1824   mat->same_nonzero = PETSC_FALSE;
1825   ierr = PetscLogEventEnd(MAT_SetValues,mat,0,0,0);CHKERRQ(ierr);
1826   PetscFunctionReturn(0);
1827 }
1828 
1829 #undef __FUNCT__
1830 #define __FUNCT__ "MatSetValuesBlockedLocal"
1831 /*@
1832    MatSetValuesBlockedLocal - Inserts or adds values into certain locations of a matrix,
1833    using a local ordering of the nodes a block at a time.
1834 
1835    Not Collective
1836 
1837    Input Parameters:
1838 +  x - the matrix
1839 .  nrow, irow - number of rows and their local indices
1840 .  ncol, icol - number of columns and their local indices
1841 .  y -  a logically two-dimensional array of values
1842 -  addv - either INSERT_VALUES or ADD_VALUES, where
1843    ADD_VALUES adds values to any existing entries, and
1844    INSERT_VALUES replaces existing entries with new values
1845 
1846    Notes:
1847    Before calling MatSetValuesBlockedLocal(), the user must first set the
1848    block size using MatSetBlockSize(), and the local-to-global mapping by
1849    calling MatSetLocalToGlobalMappingBlock(), where the mapping MUST be
1850    set for matrix blocks, not for matrix elements.
1851 
1852    Calls to MatSetValuesBlockedLocal() with the INSERT_VALUES and ADD_VALUES
1853    options cannot be mixed without intervening calls to the assembly
1854    routines.
1855 
1856    These values may be cached, so MatAssemblyBegin() and MatAssemblyEnd()
1857    MUST be called after all calls to MatSetValuesBlockedLocal() have been completed.
1858 
1859    Level: intermediate
1860 
1861    Concepts: matrices^putting blocked values in with local numbering
1862 
1863 .seealso:  MatSetBlockSize(), MatSetLocalToGlobalMappingBlock(), MatAssemblyBegin(), MatAssemblyEnd(),
1864            MatSetValuesLocal(), MatSetLocalToGlobalMappingBlock(), MatSetValuesBlocked()
1865 @*/
1866 PetscErrorCode PETSCMAT_DLLEXPORT MatSetValuesBlockedLocal(Mat mat,PetscInt nrow,const PetscInt irow[],PetscInt ncol,const PetscInt icol[],const PetscScalar y[],InsertMode addv)
1867 {
1868   PetscErrorCode ierr;
1869   PetscInt       irowm[2048],icolm[2048];
1870 
1871   PetscFunctionBegin;
1872   PetscValidHeaderSpecific(mat,MAT_CLASSID,1);
1873   PetscValidType(mat,1);
1874   if (!nrow || !ncol) PetscFunctionReturn(0); /* no values to insert */
1875   PetscValidIntPointer(irow,3);
1876   PetscValidIntPointer(icol,5);
1877   PetscValidScalarPointer(y,6);
1878   ierr = MatPreallocated(mat);CHKERRQ(ierr);
1879   if (mat->insertmode == NOT_SET_VALUES) {
1880     mat->insertmode = addv;
1881   }
1882 #if defined(PETSC_USE_DEBUG)
1883   else if (mat->insertmode != addv) SETERRQ(PETSC_COMM_SELF,PETSC_ERR_ARG_WRONGSTATE,"Cannot mix add values and insert values");
1884   if (!mat->bmapping) SETERRQ(PETSC_COMM_SELF,PETSC_ERR_ARG_WRONGSTATE,"Local to global never set with MatSetLocalToGlobalMappingBlock()");
1885   if (nrow > 2048 || ncol > 2048) SETERRQ2(PETSC_COMM_SELF,PETSC_ERR_SUP,"Number column/row indices must be <= 2048: are %D %D",nrow,ncol);
1886   if (mat->factortype) SETERRQ(PETSC_COMM_SELF,PETSC_ERR_ARG_WRONGSTATE,"Not for factored matrix");
1887 #endif
1888 
1889   if (mat->assembled) {
1890     mat->was_assembled = PETSC_TRUE;
1891     mat->assembled     = PETSC_FALSE;
1892   }
1893   ierr = PetscLogEventBegin(MAT_SetValues,mat,0,0,0);CHKERRQ(ierr);
1894   ierr = ISLocalToGlobalMappingApply(mat->bmapping,nrow,irow,irowm);CHKERRQ(ierr);
1895   ierr = ISLocalToGlobalMappingApply(mat->bmapping,ncol,icol,icolm);CHKERRQ(ierr);
1896   if (mat->ops->setvaluesblocked) {
1897     ierr = (*mat->ops->setvaluesblocked)(mat,nrow,irowm,ncol,icolm,y,addv);CHKERRQ(ierr);
1898   } else {
1899     PetscInt buf[4096],*ibufm=0,*ibufn=0;
1900     PetscInt i,j,*iirowm,*iicolm,bs=mat->rmap->bs;
1901     if ((nrow+ncol)*bs <= 4096) {
1902       iirowm = buf; iicolm = buf + nrow*bs;
1903     } else {
1904       ierr = PetscMalloc2(nrow*bs,PetscInt,&ibufm,ncol*bs,PetscInt,&ibufn);CHKERRQ(ierr);
1905       iirowm = ibufm; iicolm = ibufn;
1906     }
1907     for (i=0; i<nrow; i++) {
1908       for (j=0; j<bs; j++) {
1909 	iirowm[i*bs+j] = bs*irowm[i] + j;
1910       }
1911     }
1912     for (i=0; i<ncol; i++) {
1913       for (j=0; j<bs; j++) {
1914 	iicolm[i*bs+j] = bs*icolm[i] + j;
1915       }
1916     }
1917     ierr = MatSetValues(mat,bs*nrow,iirowm,bs*ncol,iicolm,y,addv);CHKERRQ(ierr);
1918     ierr = PetscFree2(ibufm,ibufn);CHKERRQ(ierr);
1919   }
1920   ierr = PetscLogEventEnd(MAT_SetValues,mat,0,0,0);CHKERRQ(ierr);
1921   PetscFunctionReturn(0);
1922 }
1923 
1924 #undef __FUNCT__
1925 #define __FUNCT__ "MatMultDiagonalBlock"
1926 /*@
1927    MatMultDiagonalBlock - Computes the matrix-vector product, y = Dx. Where D is defined by the inode or block structure of the diagonal
1928 
1929    Collective on Mat and Vec
1930 
1931    Input Parameters:
1932 +  mat - the matrix
1933 -  x   - the vector to be multiplied
1934 
1935    Output Parameters:
1936 .  y - the result
1937 
1938    Notes:
1939    The vectors x and y cannot be the same.  I.e., one cannot
1940    call MatMult(A,y,y).
1941 
1942    Level: developer
1943 
1944    Concepts: matrix-vector product
1945 
1946 .seealso: MatMultTranspose(), MatMultAdd(), MatMultTransposeAdd()
1947 @*/
1948 PetscErrorCode PETSCMAT_DLLEXPORT MatMultDiagonalBlock(Mat mat,Vec x,Vec y)
1949 {
1950   PetscErrorCode ierr;
1951 
1952   PetscFunctionBegin;
1953   PetscValidHeaderSpecific(mat,MAT_CLASSID,1);
1954   PetscValidType(mat,1);
1955   PetscValidHeaderSpecific(x,VEC_CLASSID,2);
1956   PetscValidHeaderSpecific(y,VEC_CLASSID,3);
1957 
1958   if (!mat->assembled) SETERRQ(((PetscObject)mat)->comm,PETSC_ERR_ARG_WRONGSTATE,"Not for unassembled matrix");
1959   if (mat->factortype) SETERRQ(((PetscObject)mat)->comm,PETSC_ERR_ARG_WRONGSTATE,"Not for factored matrix");
1960   if (x == y) SETERRQ(((PetscObject)mat)->comm,PETSC_ERR_ARG_WRONGSTATE,"x and y must be different vectors");
1961   ierr = MatPreallocated(mat);CHKERRQ(ierr);
1962 
1963   if (!mat->ops->multdiagonalblock) SETERRQ(((PetscObject)mat)->comm,PETSC_ERR_SUP,"This matrix type does not have a multiply defined");
1964   ierr = (*mat->ops->multdiagonalblock)(mat,x,y);CHKERRQ(ierr);
1965   ierr = PetscObjectStateIncrease((PetscObject)y);CHKERRQ(ierr);
1966   PetscFunctionReturn(0);
1967 }
1968 
1969 /* --------------------------------------------------------*/
1970 #undef __FUNCT__
1971 #define __FUNCT__ "MatMult"
1972 /*@
1973    MatMult - Computes the matrix-vector product, y = Ax.
1974 
1975    Collective on Mat and Vec
1976 
1977    Input Parameters:
1978 +  mat - the matrix
1979 -  x   - the vector to be multiplied
1980 
1981    Output Parameters:
1982 .  y - the result
1983 
1984    Notes:
1985    The vectors x and y cannot be the same.  I.e., one cannot
1986    call MatMult(A,y,y).
1987 
1988    Level: beginner
1989 
1990    Concepts: matrix-vector product
1991 
1992 .seealso: MatMultTranspose(), MatMultAdd(), MatMultTransposeAdd()
1993 @*/
1994 PetscErrorCode PETSCMAT_DLLEXPORT MatMult(Mat mat,Vec x,Vec y)
1995 {
1996   PetscErrorCode ierr;
1997 
1998   PetscFunctionBegin;
1999   PetscValidHeaderSpecific(mat,MAT_CLASSID,1);
2000   PetscValidType(mat,1);
2001   PetscValidHeaderSpecific(x,VEC_CLASSID,2);
2002   PetscValidHeaderSpecific(y,VEC_CLASSID,3);
2003 
2004   if (!mat->assembled) SETERRQ(((PetscObject)mat)->comm,PETSC_ERR_ARG_WRONGSTATE,"Not for unassembled matrix");
2005   if (mat->factortype) SETERRQ(((PetscObject)mat)->comm,PETSC_ERR_ARG_WRONGSTATE,"Not for factored matrix");
2006   if (x == y) SETERRQ(((PetscObject)mat)->comm,PETSC_ERR_ARG_WRONGSTATE,"x and y must be different vectors");
2007 #ifndef PETSC_HAVE_CONSTRAINTS
2008   if (mat->cmap->N != x->map->N) SETERRQ2(PETSC_COMM_SELF,PETSC_ERR_ARG_SIZ,"Mat mat,Vec x: global dim %D %D",mat->cmap->N,x->map->N);
2009   if (mat->rmap->N != y->map->N) SETERRQ2(PETSC_COMM_SELF,PETSC_ERR_ARG_SIZ,"Mat mat,Vec y: global dim %D %D",mat->rmap->N,y->map->N);
2010   if (mat->rmap->n != y->map->n) SETERRQ2(PETSC_COMM_SELF,PETSC_ERR_ARG_SIZ,"Mat mat,Vec y: local dim %D %D",mat->rmap->n,y->map->n);
2011 #endif
2012   ierr = MatPreallocated(mat);CHKERRQ(ierr);
2013 
2014   if (mat->nullsp) {
2015     ierr = MatNullSpaceRemove(mat->nullsp,x,&x);CHKERRQ(ierr);
2016   }
2017 
2018   if (!mat->ops->mult) SETERRQ(((PetscObject)mat)->comm,PETSC_ERR_SUP,"This matrix type does not have a multiply defined");
2019   ierr = PetscLogEventBegin(MAT_Mult,mat,x,y,0);CHKERRQ(ierr);
2020   ierr = (*mat->ops->mult)(mat,x,y);CHKERRQ(ierr);
2021   ierr = PetscLogEventEnd(MAT_Mult,mat,x,y,0);CHKERRQ(ierr);
2022 
2023   if (mat->nullsp) {
2024     ierr = MatNullSpaceRemove(mat->nullsp,y,PETSC_NULL);CHKERRQ(ierr);
2025   }
2026   ierr = PetscObjectStateIncrease((PetscObject)y);CHKERRQ(ierr);
2027   PetscFunctionReturn(0);
2028 }
2029 
2030 #undef __FUNCT__
2031 #define __FUNCT__ "MatMultTranspose"
2032 /*@
2033    MatMultTranspose - Computes matrix transpose times a vector.
2034 
2035    Collective on Mat and Vec
2036 
2037    Input Parameters:
2038 +  mat - the matrix
2039 -  x   - the vector to be multilplied
2040 
2041    Output Parameters:
2042 .  y - the result
2043 
2044    Notes:
2045    The vectors x and y cannot be the same.  I.e., one cannot
2046    call MatMultTranspose(A,y,y).
2047 
2048    Level: beginner
2049 
2050    Concepts: matrix vector product^transpose
2051 
2052 .seealso: MatMult(), MatMultAdd(), MatMultTransposeAdd()
2053 @*/
2054 PetscErrorCode PETSCMAT_DLLEXPORT MatMultTranspose(Mat mat,Vec x,Vec y)
2055 {
2056   PetscErrorCode ierr;
2057 
2058   PetscFunctionBegin;
2059   PetscValidHeaderSpecific(mat,MAT_CLASSID,1);
2060   PetscValidType(mat,1);
2061   PetscValidHeaderSpecific(x,VEC_CLASSID,2);
2062   PetscValidHeaderSpecific(y,VEC_CLASSID,3);
2063 
2064   if (!mat->assembled) SETERRQ(((PetscObject)mat)->comm,PETSC_ERR_ARG_WRONGSTATE,"Not for unassembled matrix");
2065   if (mat->factortype) SETERRQ(((PetscObject)mat)->comm,PETSC_ERR_ARG_WRONGSTATE,"Not for factored matrix");
2066   if (x == y) SETERRQ(((PetscObject)mat)->comm,PETSC_ERR_ARG_WRONGSTATE,"x and y must be different vectors");
2067 #ifndef PETSC_HAVE_CONSTRAINTS
2068   if (mat->rmap->N != x->map->N) SETERRQ2(PETSC_COMM_SELF,PETSC_ERR_ARG_SIZ,"Mat mat,Vec x: global dim %D %D",mat->rmap->N,x->map->N);
2069   if (mat->cmap->N != y->map->N) SETERRQ2(PETSC_COMM_SELF,PETSC_ERR_ARG_SIZ,"Mat mat,Vec y: global dim %D %D",mat->cmap->N,y->map->N);
2070 #endif
2071   ierr = MatPreallocated(mat);CHKERRQ(ierr);
2072 
2073   if (!mat->ops->multtranspose) SETERRQ(((PetscObject)mat)->comm,PETSC_ERR_SUP,"This matrix type does not have a multiply tranpose defined");
2074   ierr = PetscLogEventBegin(MAT_MultTranspose,mat,x,y,0);CHKERRQ(ierr);
2075   ierr = (*mat->ops->multtranspose)(mat,x,y);CHKERRQ(ierr);
2076   ierr = PetscLogEventEnd(MAT_MultTranspose,mat,x,y,0);CHKERRQ(ierr);
2077   ierr = PetscObjectStateIncrease((PetscObject)y);CHKERRQ(ierr);
2078   PetscFunctionReturn(0);
2079 }
2080 
2081 #undef __FUNCT__
2082 #define __FUNCT__ "MatMultHermitianTranspose"
2083 /*@
2084    MatMultHermitianTranspose - Computes matrix Hermitian transpose times a vector.
2085 
2086    Collective on Mat and Vec
2087 
2088    Input Parameters:
2089 +  mat - the matrix
2090 -  x   - the vector to be multilplied
2091 
2092    Output Parameters:
2093 .  y - the result
2094 
2095    Notes:
2096    The vectors x and y cannot be the same.  I.e., one cannot
2097    call MatMultHermitianTranspose(A,y,y).
2098 
2099    Level: beginner
2100 
2101    Concepts: matrix vector product^transpose
2102 
2103 .seealso: MatMult(), MatMultAdd(), MatMultHermitianTransposeAdd(), MatMultTranspose()
2104 @*/
2105 PetscErrorCode PETSCMAT_DLLEXPORT MatMultHermitianTranspose(Mat mat,Vec x,Vec y)
2106 {
2107   PetscErrorCode ierr;
2108 
2109   PetscFunctionBegin;
2110   PetscValidHeaderSpecific(mat,MAT_CLASSID,1);
2111   PetscValidType(mat,1);
2112   PetscValidHeaderSpecific(x,VEC_CLASSID,2);
2113   PetscValidHeaderSpecific(y,VEC_CLASSID,3);
2114 
2115   if (!mat->assembled) SETERRQ(((PetscObject)mat)->comm,PETSC_ERR_ARG_WRONGSTATE,"Not for unassembled matrix");
2116   if (mat->factortype) SETERRQ(((PetscObject)mat)->comm,PETSC_ERR_ARG_WRONGSTATE,"Not for factored matrix");
2117   if (x == y) SETERRQ(((PetscObject)mat)->comm,PETSC_ERR_ARG_WRONGSTATE,"x and y must be different vectors");
2118 #ifndef PETSC_HAVE_CONSTRAINTS
2119   if (mat->rmap->N != x->map->N) SETERRQ2(PETSC_COMM_SELF,PETSC_ERR_ARG_SIZ,"Mat mat,Vec x: global dim %D %D",mat->rmap->N,x->map->N);
2120   if (mat->cmap->N != y->map->N) SETERRQ2(PETSC_COMM_SELF,PETSC_ERR_ARG_SIZ,"Mat mat,Vec y: global dim %D %D",mat->cmap->N,y->map->N);
2121 #endif
2122   ierr = MatPreallocated(mat);CHKERRQ(ierr);
2123 
2124   if (!mat->ops->multtranspose) SETERRQ(((PetscObject)mat)->comm,PETSC_ERR_SUP,"This matrix type does not have a multiply tranpose defined");
2125   ierr = PetscLogEventBegin(MAT_MultHermitianTranspose,mat,x,y,0);CHKERRQ(ierr);
2126   ierr = (*mat->ops->multhermitiantranspose)(mat,x,y);CHKERRQ(ierr);
2127   ierr = PetscLogEventEnd(MAT_MultHermitianTranspose,mat,x,y,0);CHKERRQ(ierr);
2128   ierr = PetscObjectStateIncrease((PetscObject)y);CHKERRQ(ierr);
2129   PetscFunctionReturn(0);
2130 }
2131 
2132 #undef __FUNCT__
2133 #define __FUNCT__ "MatMultAdd"
2134 /*@
2135     MatMultAdd -  Computes v3 = v2 + A * v1.
2136 
2137     Collective on Mat and Vec
2138 
2139     Input Parameters:
2140 +   mat - the matrix
2141 -   v1, v2 - the vectors
2142 
2143     Output Parameters:
2144 .   v3 - the result
2145 
2146     Notes:
2147     The vectors v1 and v3 cannot be the same.  I.e., one cannot
2148     call MatMultAdd(A,v1,v2,v1).
2149 
2150     Level: beginner
2151 
2152     Concepts: matrix vector product^addition
2153 
2154 .seealso: MatMultTranspose(), MatMult(), MatMultTransposeAdd()
2155 @*/
2156 PetscErrorCode PETSCMAT_DLLEXPORT MatMultAdd(Mat mat,Vec v1,Vec v2,Vec v3)
2157 {
2158   PetscErrorCode ierr;
2159 
2160   PetscFunctionBegin;
2161   PetscValidHeaderSpecific(mat,MAT_CLASSID,1);
2162   PetscValidType(mat,1);
2163   PetscValidHeaderSpecific(v1,VEC_CLASSID,2);
2164   PetscValidHeaderSpecific(v2,VEC_CLASSID,3);
2165   PetscValidHeaderSpecific(v3,VEC_CLASSID,4);
2166 
2167   if (!mat->assembled) SETERRQ(((PetscObject)mat)->comm,PETSC_ERR_ARG_WRONGSTATE,"Not for unassembled matrix");
2168   if (mat->factortype) SETERRQ(((PetscObject)mat)->comm,PETSC_ERR_ARG_WRONGSTATE,"Not for factored matrix");
2169   if (mat->cmap->N != v1->map->N) SETERRQ2(((PetscObject)mat)->comm,PETSC_ERR_ARG_SIZ,"Mat mat,Vec v1: global dim %D %D",mat->cmap->N,v1->map->N);
2170   /* if (mat->rmap->N != v2->map->N) SETERRQ2(PETSC_COMM_SELF,PETSC_ERR_ARG_SIZ,"Mat mat,Vec v2: global dim %D %D",mat->rmap->N,v2->map->N);
2171      if (mat->rmap->N != v3->map->N) SETERRQ2(PETSC_COMM_SELF,PETSC_ERR_ARG_SIZ,"Mat mat,Vec v3: global dim %D %D",mat->rmap->N,v3->map->N); */
2172   if (mat->rmap->n != v3->map->n) SETERRQ2(PETSC_COMM_SELF,PETSC_ERR_ARG_SIZ,"Mat mat,Vec v3: local dim %D %D",mat->rmap->n,v3->map->n);
2173   if (mat->rmap->n != v2->map->n) SETERRQ2(PETSC_COMM_SELF,PETSC_ERR_ARG_SIZ,"Mat mat,Vec v2: local dim %D %D",mat->rmap->n,v2->map->n);
2174   if (v1 == v3) SETERRQ(((PetscObject)mat)->comm,PETSC_ERR_ARG_IDN,"v1 and v3 must be different vectors");
2175   ierr = MatPreallocated(mat);CHKERRQ(ierr);
2176 
2177   ierr = PetscLogEventBegin(MAT_MultAdd,mat,v1,v2,v3);CHKERRQ(ierr);
2178   ierr = (*mat->ops->multadd)(mat,v1,v2,v3);CHKERRQ(ierr);
2179   ierr = PetscLogEventEnd(MAT_MultAdd,mat,v1,v2,v3);CHKERRQ(ierr);
2180   ierr = PetscObjectStateIncrease((PetscObject)v3);CHKERRQ(ierr);
2181   PetscFunctionReturn(0);
2182 }
2183 
2184 #undef __FUNCT__
2185 #define __FUNCT__ "MatMultTransposeAdd"
2186 /*@
2187    MatMultTransposeAdd - Computes v3 = v2 + A' * v1.
2188 
2189    Collective on Mat and Vec
2190 
2191    Input Parameters:
2192 +  mat - the matrix
2193 -  v1, v2 - the vectors
2194 
2195    Output Parameters:
2196 .  v3 - the result
2197 
2198    Notes:
2199    The vectors v1 and v3 cannot be the same.  I.e., one cannot
2200    call MatMultTransposeAdd(A,v1,v2,v1).
2201 
2202    Level: beginner
2203 
2204    Concepts: matrix vector product^transpose and addition
2205 
2206 .seealso: MatMultTranspose(), MatMultAdd(), MatMult()
2207 @*/
2208 PetscErrorCode PETSCMAT_DLLEXPORT MatMultTransposeAdd(Mat mat,Vec v1,Vec v2,Vec v3)
2209 {
2210   PetscErrorCode ierr;
2211 
2212   PetscFunctionBegin;
2213   PetscValidHeaderSpecific(mat,MAT_CLASSID,1);
2214   PetscValidType(mat,1);
2215   PetscValidHeaderSpecific(v1,VEC_CLASSID,2);
2216   PetscValidHeaderSpecific(v2,VEC_CLASSID,3);
2217   PetscValidHeaderSpecific(v3,VEC_CLASSID,4);
2218 
2219   if (!mat->assembled) SETERRQ(((PetscObject)mat)->comm,PETSC_ERR_ARG_WRONGSTATE,"Not for unassembled matrix");
2220   if (mat->factortype) SETERRQ(((PetscObject)mat)->comm,PETSC_ERR_ARG_WRONGSTATE,"Not for factored matrix");
2221   if (!mat->ops->multtransposeadd) SETERRQ1(((PetscObject)mat)->comm,PETSC_ERR_SUP,"Mat type %s",((PetscObject)mat)->type_name);
2222   if (v1 == v3) SETERRQ(((PetscObject)mat)->comm,PETSC_ERR_ARG_IDN,"v1 and v3 must be different vectors");
2223   if (mat->rmap->N != v1->map->N) SETERRQ2(((PetscObject)mat)->comm,PETSC_ERR_ARG_SIZ,"Mat mat,Vec v1: global dim %D %D",mat->rmap->N,v1->map->N);
2224   if (mat->cmap->N != v2->map->N) SETERRQ2(((PetscObject)mat)->comm,PETSC_ERR_ARG_SIZ,"Mat mat,Vec v2: global dim %D %D",mat->cmap->N,v2->map->N);
2225   if (mat->cmap->N != v3->map->N) SETERRQ2(((PetscObject)mat)->comm,PETSC_ERR_ARG_SIZ,"Mat mat,Vec v3: global dim %D %D",mat->cmap->N,v3->map->N);
2226   ierr = MatPreallocated(mat);CHKERRQ(ierr);
2227 
2228   ierr = PetscLogEventBegin(MAT_MultTransposeAdd,mat,v1,v2,v3);CHKERRQ(ierr);
2229   ierr = (*mat->ops->multtransposeadd)(mat,v1,v2,v3);CHKERRQ(ierr);
2230   ierr = PetscLogEventEnd(MAT_MultTransposeAdd,mat,v1,v2,v3);CHKERRQ(ierr);
2231   ierr = PetscObjectStateIncrease((PetscObject)v3);CHKERRQ(ierr);
2232   PetscFunctionReturn(0);
2233 }
2234 
2235 #undef __FUNCT__
2236 #define __FUNCT__ "MatMultHermitianTransposeAdd"
2237 /*@
2238    MatMultHermitianTransposeAdd - Computes v3 = v2 + A^H * v1.
2239 
2240    Collective on Mat and Vec
2241 
2242    Input Parameters:
2243 +  mat - the matrix
2244 -  v1, v2 - the vectors
2245 
2246    Output Parameters:
2247 .  v3 - the result
2248 
2249    Notes:
2250    The vectors v1 and v3 cannot be the same.  I.e., one cannot
2251    call MatMultHermitianTransposeAdd(A,v1,v2,v1).
2252 
2253    Level: beginner
2254 
2255    Concepts: matrix vector product^transpose and addition
2256 
2257 .seealso: MatMultHermitianTranspose(), MatMultTranspose(), MatMultAdd(), MatMult()
2258 @*/
2259 PetscErrorCode PETSCMAT_DLLEXPORT MatMultHermitianTransposeAdd(Mat mat,Vec v1,Vec v2,Vec v3)
2260 {
2261   PetscErrorCode ierr;
2262 
2263   PetscFunctionBegin;
2264   PetscValidHeaderSpecific(mat,MAT_CLASSID,1);
2265   PetscValidType(mat,1);
2266   PetscValidHeaderSpecific(v1,VEC_CLASSID,2);
2267   PetscValidHeaderSpecific(v2,VEC_CLASSID,3);
2268   PetscValidHeaderSpecific(v3,VEC_CLASSID,4);
2269 
2270   if (!mat->assembled) SETERRQ(((PetscObject)mat)->comm,PETSC_ERR_ARG_WRONGSTATE,"Not for unassembled matrix");
2271   if (mat->factortype) SETERRQ(((PetscObject)mat)->comm,PETSC_ERR_ARG_WRONGSTATE,"Not for factored matrix");
2272   if (!mat->ops->multtransposeadd) SETERRQ1(((PetscObject)mat)->comm,PETSC_ERR_SUP,"Mat type %s",((PetscObject)mat)->type_name);
2273   if (v1 == v3) SETERRQ(((PetscObject)mat)->comm,PETSC_ERR_ARG_IDN,"v1 and v3 must be different vectors");
2274   if (mat->rmap->N != v1->map->N) SETERRQ2(((PetscObject)mat)->comm,PETSC_ERR_ARG_SIZ,"Mat mat,Vec v1: global dim %D %D",mat->rmap->N,v1->map->N);
2275   if (mat->cmap->N != v2->map->N) SETERRQ2(((PetscObject)mat)->comm,PETSC_ERR_ARG_SIZ,"Mat mat,Vec v2: global dim %D %D",mat->cmap->N,v2->map->N);
2276   if (mat->cmap->N != v3->map->N) SETERRQ2(((PetscObject)mat)->comm,PETSC_ERR_ARG_SIZ,"Mat mat,Vec v3: global dim %D %D",mat->cmap->N,v3->map->N);
2277   ierr = MatPreallocated(mat);CHKERRQ(ierr);
2278 
2279   ierr = PetscLogEventBegin(MAT_MultHermitianTransposeAdd,mat,v1,v2,v3);CHKERRQ(ierr);
2280   ierr = (*mat->ops->multhermitiantransposeadd)(mat,v1,v2,v3);CHKERRQ(ierr);
2281   ierr = PetscLogEventEnd(MAT_MultHermitianTransposeAdd,mat,v1,v2,v3);CHKERRQ(ierr);
2282   ierr = PetscObjectStateIncrease((PetscObject)v3);CHKERRQ(ierr);
2283   PetscFunctionReturn(0);
2284 }
2285 
2286 #undef __FUNCT__
2287 #define __FUNCT__ "MatMultConstrained"
2288 /*@
2289    MatMultConstrained - The inner multiplication routine for a
2290    constrained matrix P^T A P.
2291 
2292    Collective on Mat and Vec
2293 
2294    Input Parameters:
2295 +  mat - the matrix
2296 -  x   - the vector to be multilplied
2297 
2298    Output Parameters:
2299 .  y - the result
2300 
2301    Notes:
2302    The vectors x and y cannot be the same.  I.e., one cannot
2303    call MatMult(A,y,y).
2304 
2305    Level: beginner
2306 
2307 .keywords: matrix, multiply, matrix-vector product, constraint
2308 .seealso: MatMult(), MatMultTranspose(), MatMultAdd(), MatMultTransposeAdd()
2309 @*/
2310 PetscErrorCode PETSCMAT_DLLEXPORT MatMultConstrained(Mat mat,Vec x,Vec y)
2311 {
2312   PetscErrorCode ierr;
2313 
2314   PetscFunctionBegin;
2315   PetscValidHeaderSpecific(mat,MAT_CLASSID,1);
2316   PetscValidHeaderSpecific(x,VEC_CLASSID,2);
2317   PetscValidHeaderSpecific(y,VEC_CLASSID,3);
2318   if (!mat->assembled) SETERRQ(PETSC_COMM_SELF,PETSC_ERR_ARG_WRONGSTATE,"Not for unassembled matrix");
2319   if (mat->factortype) SETERRQ(PETSC_COMM_SELF,PETSC_ERR_ARG_WRONGSTATE,"Not for factored matrix");
2320   if (x == y) SETERRQ(PETSC_COMM_SELF,PETSC_ERR_ARG_WRONGSTATE,"x and y must be different vectors");
2321   if (mat->cmap->N != x->map->N) SETERRQ2(PETSC_COMM_SELF,PETSC_ERR_ARG_SIZ,"Mat mat,Vec x: global dim %D %D",mat->cmap->N,x->map->N);
2322   if (mat->rmap->N != y->map->N) SETERRQ2(PETSC_COMM_SELF,PETSC_ERR_ARG_SIZ,"Mat mat,Vec y: global dim %D %D",mat->rmap->N,y->map->N);
2323   if (mat->rmap->n != y->map->n) SETERRQ2(PETSC_COMM_SELF,PETSC_ERR_ARG_SIZ,"Mat mat,Vec y: local dim %D %D",mat->rmap->n,y->map->n);
2324 
2325   ierr = PetscLogEventBegin(MAT_MultConstrained,mat,x,y,0);CHKERRQ(ierr);
2326   ierr = (*mat->ops->multconstrained)(mat,x,y);CHKERRQ(ierr);
2327   ierr = PetscLogEventEnd(MAT_MultConstrained,mat,x,y,0);CHKERRQ(ierr);
2328   ierr = PetscObjectStateIncrease((PetscObject)y);CHKERRQ(ierr);
2329 
2330   PetscFunctionReturn(0);
2331 }
2332 
2333 #undef __FUNCT__
2334 #define __FUNCT__ "MatMultTransposeConstrained"
2335 /*@
2336    MatMultTransposeConstrained - The inner multiplication routine for a
2337    constrained matrix P^T A^T P.
2338 
2339    Collective on Mat and Vec
2340 
2341    Input Parameters:
2342 +  mat - the matrix
2343 -  x   - the vector to be multilplied
2344 
2345    Output Parameters:
2346 .  y - the result
2347 
2348    Notes:
2349    The vectors x and y cannot be the same.  I.e., one cannot
2350    call MatMult(A,y,y).
2351 
2352    Level: beginner
2353 
2354 .keywords: matrix, multiply, matrix-vector product, constraint
2355 .seealso: MatMult(), MatMultTranspose(), MatMultAdd(), MatMultTransposeAdd()
2356 @*/
2357 PetscErrorCode PETSCMAT_DLLEXPORT MatMultTransposeConstrained(Mat mat,Vec x,Vec y)
2358 {
2359   PetscErrorCode ierr;
2360 
2361   PetscFunctionBegin;
2362   PetscValidHeaderSpecific(mat,MAT_CLASSID,1);
2363   PetscValidHeaderSpecific(x,VEC_CLASSID,2);
2364   PetscValidHeaderSpecific(y,VEC_CLASSID,3);
2365   if (!mat->assembled) SETERRQ(PETSC_COMM_SELF,PETSC_ERR_ARG_WRONGSTATE,"Not for unassembled matrix");
2366   if (mat->factortype) SETERRQ(PETSC_COMM_SELF,PETSC_ERR_ARG_WRONGSTATE,"Not for factored matrix");
2367   if (x == y) SETERRQ(PETSC_COMM_SELF,PETSC_ERR_ARG_WRONGSTATE,"x and y must be different vectors");
2368   if (mat->rmap->N != x->map->N) SETERRQ2(PETSC_COMM_SELF,PETSC_ERR_ARG_SIZ,"Mat mat,Vec x: global dim %D %D",mat->cmap->N,x->map->N);
2369   if (mat->cmap->N != y->map->N) SETERRQ2(PETSC_COMM_SELF,PETSC_ERR_ARG_SIZ,"Mat mat,Vec y: global dim %D %D",mat->rmap->N,y->map->N);
2370 
2371   ierr = PetscLogEventBegin(MAT_MultConstrained,mat,x,y,0);CHKERRQ(ierr);
2372   ierr = (*mat->ops->multtransposeconstrained)(mat,x,y);CHKERRQ(ierr);
2373   ierr = PetscLogEventEnd(MAT_MultConstrained,mat,x,y,0);CHKERRQ(ierr);
2374   ierr = PetscObjectStateIncrease((PetscObject)y);CHKERRQ(ierr);
2375 
2376   PetscFunctionReturn(0);
2377 }
2378 
2379 #undef __FUNCT__
2380 #define __FUNCT__ "MatGetFactorType"
2381 /*@C
2382    MatGetFactorType - gets the type of factorization it is
2383 
2384    Note Collective
2385    as the flag
2386 
2387    Input Parameters:
2388 .  mat - the matrix
2389 
2390    Output Parameters:
2391 .  t - the type, one of MAT_FACTOR_NONE, MAT_FACTOR_LU, MAT_FACTOR_CHOLESKY, MAT_FACTOR_ILU, MAT_FACTOR_ICC,MAT_FACTOR_ILUDT
2392 
2393     Level: intermediate
2394 
2395 .seealso:    MatFactorType, MatGetFactor()
2396 @*/
2397 PetscErrorCode PETSCMAT_DLLEXPORT MatGetFactorType(Mat mat,MatFactorType *t)
2398 {
2399   PetscFunctionBegin;
2400   PetscValidHeaderSpecific(mat,MAT_CLASSID,1);
2401   PetscValidType(mat,1);
2402   *t = mat->factortype;
2403   PetscFunctionReturn(0);
2404 }
2405 
2406 /* ------------------------------------------------------------*/
2407 #undef __FUNCT__
2408 #define __FUNCT__ "MatGetInfo"
2409 /*@C
2410    MatGetInfo - Returns information about matrix storage (number of
2411    nonzeros, memory, etc.).
2412 
2413    Collective on Mat if MAT_GLOBAL_MAX or MAT_GLOBAL_SUM is used
2414    as the flag
2415 
2416    Input Parameters:
2417 .  mat - the matrix
2418 
2419    Output Parameters:
2420 +  flag - flag indicating the type of parameters to be returned
2421    (MAT_LOCAL - local matrix, MAT_GLOBAL_MAX - maximum over all processors,
2422    MAT_GLOBAL_SUM - sum over all processors)
2423 -  info - matrix information context
2424 
2425    Notes:
2426    The MatInfo context contains a variety of matrix data, including
2427    number of nonzeros allocated and used, number of mallocs during
2428    matrix assembly, etc.  Additional information for factored matrices
2429    is provided (such as the fill ratio, number of mallocs during
2430    factorization, etc.).  Much of this info is printed to PETSC_STDOUT
2431    when using the runtime options
2432 $       -info -mat_view_info
2433 
2434    Example for C/C++ Users:
2435    See the file ${PETSC_DIR}/include/petscmat.h for a complete list of
2436    data within the MatInfo context.  For example,
2437 .vb
2438       MatInfo info;
2439       Mat     A;
2440       double  mal, nz_a, nz_u;
2441 
2442       MatGetInfo(A,MAT_LOCAL,&info);
2443       mal  = info.mallocs;
2444       nz_a = info.nz_allocated;
2445 .ve
2446 
2447    Example for Fortran Users:
2448    Fortran users should declare info as a double precision
2449    array of dimension MAT_INFO_SIZE, and then extract the parameters
2450    of interest.  See the file ${PETSC_DIR}/include/finclude/petscmat.h
2451    a complete list of parameter names.
2452 .vb
2453       double  precision info(MAT_INFO_SIZE)
2454       double  precision mal, nz_a
2455       Mat     A
2456       integer ierr
2457 
2458       call MatGetInfo(A,MAT_LOCAL,info,ierr)
2459       mal = info(MAT_INFO_MALLOCS)
2460       nz_a = info(MAT_INFO_NZ_ALLOCATED)
2461 .ve
2462 
2463     Level: intermediate
2464 
2465     Concepts: matrices^getting information on
2466 
2467     Developer Note: fortran interface is not autogenerated as the f90
2468     interface defintion cannot be generated correctly [due to MatInfo]
2469 
2470 @*/
2471 PetscErrorCode PETSCMAT_DLLEXPORT MatGetInfo(Mat mat,MatInfoType flag,MatInfo *info)
2472 {
2473   PetscErrorCode ierr;
2474 
2475   PetscFunctionBegin;
2476   PetscValidHeaderSpecific(mat,MAT_CLASSID,1);
2477   PetscValidType(mat,1);
2478   PetscValidPointer(info,3);
2479   if (!mat->ops->getinfo) SETERRQ1(PETSC_COMM_SELF,PETSC_ERR_SUP,"Mat type %s",((PetscObject)mat)->type_name);
2480   ierr = MatPreallocated(mat);CHKERRQ(ierr);
2481   ierr = (*mat->ops->getinfo)(mat,flag,info);CHKERRQ(ierr);
2482   PetscFunctionReturn(0);
2483 }
2484 
2485 /* ----------------------------------------------------------*/
2486 
2487 #undef __FUNCT__
2488 #define __FUNCT__ "MatLUFactor"
2489 /*@C
2490    MatLUFactor - Performs in-place LU factorization of matrix.
2491 
2492    Collective on Mat
2493 
2494    Input Parameters:
2495 +  mat - the matrix
2496 .  row - row permutation
2497 .  col - column permutation
2498 -  info - options for factorization, includes
2499 $          fill - expected fill as ratio of original fill.
2500 $          dtcol - pivot tolerance (0 no pivot, 1 full column pivoting)
2501 $                   Run with the option -info to determine an optimal value to use
2502 
2503    Notes:
2504    Most users should employ the simplified KSP interface for linear solvers
2505    instead of working directly with matrix algebra routines such as this.
2506    See, e.g., KSPCreate().
2507 
2508    This changes the state of the matrix to a factored matrix; it cannot be used
2509    for example with MatSetValues() unless one first calls MatSetUnfactored().
2510 
2511    Level: developer
2512 
2513    Concepts: matrices^LU factorization
2514 
2515 .seealso: MatLUFactorSymbolic(), MatLUFactorNumeric(), MatCholeskyFactor(),
2516           MatGetOrdering(), MatSetUnfactored(), MatFactorInfo
2517 
2518     Developer Note: fortran interface is not autogenerated as the f90
2519     interface defintion cannot be generated correctly [due to MatFactorInfo]
2520 
2521 @*/
2522 PetscErrorCode PETSCMAT_DLLEXPORT MatLUFactor(Mat mat,IS row,IS col,const MatFactorInfo *info)
2523 {
2524   PetscErrorCode ierr;
2525 
2526   PetscFunctionBegin;
2527   PetscValidHeaderSpecific(mat,MAT_CLASSID,1);
2528   if (row) PetscValidHeaderSpecific(row,IS_CLASSID,2);
2529   if (col) PetscValidHeaderSpecific(col,IS_CLASSID,3);
2530   PetscValidPointer(info,4);
2531   PetscValidType(mat,1);
2532   if (!mat->assembled) SETERRQ(((PetscObject)mat)->comm,PETSC_ERR_ARG_WRONGSTATE,"Not for unassembled matrix");
2533   if (mat->factortype) SETERRQ(((PetscObject)mat)->comm,PETSC_ERR_ARG_WRONGSTATE,"Not for factored matrix");
2534   if (!mat->ops->lufactor) SETERRQ1(((PetscObject)mat)->comm,PETSC_ERR_SUP,"Mat type %s",((PetscObject)mat)->type_name);
2535   ierr = MatPreallocated(mat);CHKERRQ(ierr);
2536 
2537   ierr = PetscLogEventBegin(MAT_LUFactor,mat,row,col,0);CHKERRQ(ierr);
2538   ierr = (*mat->ops->lufactor)(mat,row,col,info);CHKERRQ(ierr);
2539   ierr = PetscLogEventEnd(MAT_LUFactor,mat,row,col,0);CHKERRQ(ierr);
2540   ierr = PetscObjectStateIncrease((PetscObject)mat);CHKERRQ(ierr);
2541   PetscFunctionReturn(0);
2542 }
2543 
2544 #undef __FUNCT__
2545 #define __FUNCT__ "MatILUFactor"
2546 /*@C
2547    MatILUFactor - Performs in-place ILU factorization of matrix.
2548 
2549    Collective on Mat
2550 
2551    Input Parameters:
2552 +  mat - the matrix
2553 .  row - row permutation
2554 .  col - column permutation
2555 -  info - structure containing
2556 $      levels - number of levels of fill.
2557 $      expected fill - as ratio of original fill.
2558 $      1 or 0 - indicating force fill on diagonal (improves robustness for matrices
2559                 missing diagonal entries)
2560 
2561    Notes:
2562    Probably really in-place only when level of fill is zero, otherwise allocates
2563    new space to store factored matrix and deletes previous memory.
2564 
2565    Most users should employ the simplified KSP interface for linear solvers
2566    instead of working directly with matrix algebra routines such as this.
2567    See, e.g., KSPCreate().
2568 
2569    Level: developer
2570 
2571    Concepts: matrices^ILU factorization
2572 
2573 .seealso: MatILUFactorSymbolic(), MatLUFactorNumeric(), MatCholeskyFactor(), MatFactorInfo
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 MatILUFactor(Mat mat,IS row,IS col,const MatFactorInfo *info)
2580 {
2581   PetscErrorCode ierr;
2582 
2583   PetscFunctionBegin;
2584   PetscValidHeaderSpecific(mat,MAT_CLASSID,1);
2585   if (row) PetscValidHeaderSpecific(row,IS_CLASSID,2);
2586   if (col) PetscValidHeaderSpecific(col,IS_CLASSID,3);
2587   PetscValidPointer(info,4);
2588   PetscValidType(mat,1);
2589   if (mat->rmap->N != mat->cmap->N) SETERRQ(((PetscObject)mat)->comm,PETSC_ERR_ARG_WRONG,"matrix must be square");
2590   if (!mat->assembled) SETERRQ(((PetscObject)mat)->comm,PETSC_ERR_ARG_WRONGSTATE,"Not for unassembled matrix");
2591   if (mat->factortype) SETERRQ(((PetscObject)mat)->comm,PETSC_ERR_ARG_WRONGSTATE,"Not for factored matrix");
2592   if (!mat->ops->ilufactor) SETERRQ1(((PetscObject)mat)->comm,PETSC_ERR_SUP,"Mat type %s",((PetscObject)mat)->type_name);
2593   ierr = MatPreallocated(mat);CHKERRQ(ierr);
2594 
2595   ierr = PetscLogEventBegin(MAT_ILUFactor,mat,row,col,0);CHKERRQ(ierr);
2596   ierr = (*mat->ops->ilufactor)(mat,row,col,info);CHKERRQ(ierr);
2597   ierr = PetscLogEventEnd(MAT_ILUFactor,mat,row,col,0);CHKERRQ(ierr);
2598   ierr = PetscObjectStateIncrease((PetscObject)mat);CHKERRQ(ierr);
2599   PetscFunctionReturn(0);
2600 }
2601 
2602 #undef __FUNCT__
2603 #define __FUNCT__ "MatLUFactorSymbolic"
2604 /*@C
2605    MatLUFactorSymbolic - Performs symbolic LU factorization of matrix.
2606    Call this routine before calling MatLUFactorNumeric().
2607 
2608    Collective on Mat
2609 
2610    Input Parameters:
2611 +  fact - the factor matrix obtained with MatGetFactor()
2612 .  mat - the matrix
2613 .  row, col - row and column permutations
2614 -  info - options for factorization, includes
2615 $          fill - expected fill as ratio of original fill.
2616 $          dtcol - pivot tolerance (0 no pivot, 1 full column pivoting)
2617 $                   Run with the option -info to determine an optimal value to use
2618 
2619 
2620    Notes:
2621    See the users manual for additional information about
2622    choosing the fill factor for better efficiency.
2623 
2624    Most users should employ the simplified KSP interface for linear solvers
2625    instead of working directly with matrix algebra routines such as this.
2626    See, e.g., KSPCreate().
2627 
2628    Level: developer
2629 
2630    Concepts: matrices^LU symbolic factorization
2631 
2632 .seealso: MatLUFactor(), MatLUFactorNumeric(), MatCholeskyFactor(), MatFactorInfo
2633 
2634     Developer Note: fortran interface is not autogenerated as the f90
2635     interface defintion cannot be generated correctly [due to MatFactorInfo]
2636 
2637 @*/
2638 PetscErrorCode PETSCMAT_DLLEXPORT MatLUFactorSymbolic(Mat fact,Mat mat,IS row,IS col,const MatFactorInfo *info)
2639 {
2640   PetscErrorCode ierr;
2641 
2642   PetscFunctionBegin;
2643   PetscValidHeaderSpecific(mat,MAT_CLASSID,1);
2644   if (row) PetscValidHeaderSpecific(row,IS_CLASSID,2);
2645   if (col) PetscValidHeaderSpecific(col,IS_CLASSID,3);
2646   PetscValidPointer(info,4);
2647   PetscValidType(mat,1);
2648   PetscValidPointer(fact,5);
2649   if (!mat->assembled) SETERRQ(((PetscObject)mat)->comm,PETSC_ERR_ARG_WRONGSTATE,"Not for unassembled matrix");
2650   if (mat->factortype) SETERRQ(((PetscObject)mat)->comm,PETSC_ERR_ARG_WRONGSTATE,"Not for factored matrix");
2651   if (!(fact)->ops->lufactorsymbolic) SETERRQ1(((PetscObject)mat)->comm,PETSC_ERR_SUP,"Matrix type %s  symbolic LU",((PetscObject)mat)->type_name);
2652   ierr = MatPreallocated(mat);CHKERRQ(ierr);
2653 
2654   ierr = PetscLogEventBegin(MAT_LUFactorSymbolic,mat,row,col,0);CHKERRQ(ierr);
2655   ierr = (fact->ops->lufactorsymbolic)(fact,mat,row,col,info);CHKERRQ(ierr);
2656   ierr = PetscLogEventEnd(MAT_LUFactorSymbolic,mat,row,col,0);CHKERRQ(ierr);
2657   ierr = PetscObjectStateIncrease((PetscObject)fact);CHKERRQ(ierr);
2658   PetscFunctionReturn(0);
2659 }
2660 
2661 #undef __FUNCT__
2662 #define __FUNCT__ "MatLUFactorNumeric"
2663 /*@C
2664    MatLUFactorNumeric - Performs numeric LU factorization of a matrix.
2665    Call this routine after first calling MatLUFactorSymbolic().
2666 
2667    Collective on Mat
2668 
2669    Input Parameters:
2670 +  fact - the factor matrix obtained with MatGetFactor()
2671 .  mat - the matrix
2672 -  info - options for factorization
2673 
2674    Notes:
2675    See MatLUFactor() for in-place factorization.  See
2676    MatCholeskyFactorNumeric() for the symmetric, positive definite case.
2677 
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^LU numeric factorization
2685 
2686 .seealso: MatLUFactorSymbolic(), MatLUFactor(), MatCholeskyFactor()
2687 
2688     Developer Note: fortran interface is not autogenerated as the f90
2689     interface defintion cannot be generated correctly [due to MatFactorInfo]
2690 
2691 @*/
2692 PetscErrorCode PETSCMAT_DLLEXPORT MatLUFactorNumeric(Mat fact,Mat mat,const MatFactorInfo *info)
2693 {
2694   PetscErrorCode ierr;
2695 
2696   PetscFunctionBegin;
2697   PetscValidHeaderSpecific(mat,MAT_CLASSID,1);
2698   PetscValidType(mat,1);
2699   PetscValidPointer(fact,2);
2700   PetscValidHeaderSpecific(fact,MAT_CLASSID,2);
2701   if (!mat->assembled) SETERRQ(((PetscObject)mat)->comm,PETSC_ERR_ARG_WRONGSTATE,"Not for unassembled matrix");
2702   if (mat->rmap->N != (fact)->rmap->N || mat->cmap->N != (fact)->cmap->N) {
2703     SETERRQ4(((PetscObject)mat)->comm,PETSC_ERR_ARG_SIZ,"Mat mat,Mat fact: global dimensions are different %D should = %D %D should = %D",mat->rmap->N,(fact)->rmap->N,mat->cmap->N,(fact)->cmap->N);
2704   }
2705   if (!(fact)->ops->lufactornumeric) SETERRQ1(((PetscObject)mat)->comm,PETSC_ERR_SUP,"Mat type %s",((PetscObject)mat)->type_name);
2706   ierr = MatPreallocated(mat);CHKERRQ(ierr);
2707   ierr = PetscLogEventBegin(MAT_LUFactorNumeric,mat,fact,0,0);CHKERRQ(ierr);
2708   ierr = (fact->ops->lufactornumeric)(fact,mat,info);CHKERRQ(ierr);
2709   ierr = PetscLogEventEnd(MAT_LUFactorNumeric,mat,fact,0,0);CHKERRQ(ierr);
2710 
2711   ierr = MatView_Private(fact);CHKERRQ(ierr);
2712   ierr = PetscObjectStateIncrease((PetscObject)fact);CHKERRQ(ierr);
2713   PetscFunctionReturn(0);
2714 }
2715 
2716 #undef __FUNCT__
2717 #define __FUNCT__ "MatCholeskyFactor"
2718 /*@C
2719    MatCholeskyFactor - Performs in-place Cholesky factorization of a
2720    symmetric matrix.
2721 
2722    Collective on Mat
2723 
2724    Input Parameters:
2725 +  mat - the matrix
2726 .  perm - row and column permutations
2727 -  f - expected fill as ratio of original fill
2728 
2729    Notes:
2730    See MatLUFactor() for the nonsymmetric case.  See also
2731    MatCholeskyFactorSymbolic(), and MatCholeskyFactorNumeric().
2732 
2733    Most users should employ the simplified KSP interface for linear solvers
2734    instead of working directly with matrix algebra routines such as this.
2735    See, e.g., KSPCreate().
2736 
2737    Level: developer
2738 
2739    Concepts: matrices^Cholesky factorization
2740 
2741 .seealso: MatLUFactor(), MatCholeskyFactorSymbolic(), MatCholeskyFactorNumeric()
2742           MatGetOrdering()
2743 
2744     Developer Note: fortran interface is not autogenerated as the f90
2745     interface defintion cannot be generated correctly [due to MatFactorInfo]
2746 
2747 @*/
2748 PetscErrorCode PETSCMAT_DLLEXPORT MatCholeskyFactor(Mat mat,IS perm,const MatFactorInfo *info)
2749 {
2750   PetscErrorCode ierr;
2751 
2752   PetscFunctionBegin;
2753   PetscValidHeaderSpecific(mat,MAT_CLASSID,1);
2754   PetscValidType(mat,1);
2755   PetscValidHeaderSpecific(perm,IS_CLASSID,2);
2756   PetscValidPointer(info,3);
2757   if (mat->rmap->N != mat->cmap->N) SETERRQ(((PetscObject)mat)->comm,PETSC_ERR_ARG_WRONG,"Matrix must be square");
2758   if (!mat->assembled) SETERRQ(((PetscObject)mat)->comm,PETSC_ERR_ARG_WRONGSTATE,"Not for unassembled matrix");
2759   if (mat->factortype) SETERRQ(((PetscObject)mat)->comm,PETSC_ERR_ARG_WRONGSTATE,"Not for factored matrix");
2760   if (!mat->ops->choleskyfactor) SETERRQ1(((PetscObject)mat)->comm,PETSC_ERR_SUP,"Mat type %s",((PetscObject)mat)->type_name);
2761   ierr = MatPreallocated(mat);CHKERRQ(ierr);
2762 
2763   ierr = PetscLogEventBegin(MAT_CholeskyFactor,mat,perm,0,0);CHKERRQ(ierr);
2764   ierr = (*mat->ops->choleskyfactor)(mat,perm,info);CHKERRQ(ierr);
2765   ierr = PetscLogEventEnd(MAT_CholeskyFactor,mat,perm,0,0);CHKERRQ(ierr);
2766   ierr = PetscObjectStateIncrease((PetscObject)mat);CHKERRQ(ierr);
2767   PetscFunctionReturn(0);
2768 }
2769 
2770 #undef __FUNCT__
2771 #define __FUNCT__ "MatCholeskyFactorSymbolic"
2772 /*@C
2773    MatCholeskyFactorSymbolic - Performs symbolic Cholesky factorization
2774    of a symmetric matrix.
2775 
2776    Collective on Mat
2777 
2778    Input Parameters:
2779 +  fact - the factor matrix obtained with MatGetFactor()
2780 .  mat - the matrix
2781 .  perm - row and column permutations
2782 -  info - options for factorization, includes
2783 $          fill - expected fill as ratio of original fill.
2784 $          dtcol - pivot tolerance (0 no pivot, 1 full column pivoting)
2785 $                   Run with the option -info to determine an optimal value to use
2786 
2787    Notes:
2788    See MatLUFactorSymbolic() for the nonsymmetric case.  See also
2789    MatCholeskyFactor() and MatCholeskyFactorNumeric().
2790 
2791    Most users should employ the simplified KSP interface for linear solvers
2792    instead of working directly with matrix algebra routines such as this.
2793    See, e.g., KSPCreate().
2794 
2795    Level: developer
2796 
2797    Concepts: matrices^Cholesky symbolic factorization
2798 
2799 .seealso: MatLUFactorSymbolic(), MatCholeskyFactor(), MatCholeskyFactorNumeric()
2800           MatGetOrdering()
2801 
2802     Developer Note: fortran interface is not autogenerated as the f90
2803     interface defintion cannot be generated correctly [due to MatFactorInfo]
2804 
2805 @*/
2806 PetscErrorCode PETSCMAT_DLLEXPORT MatCholeskyFactorSymbolic(Mat fact,Mat mat,IS perm,const MatFactorInfo *info)
2807 {
2808   PetscErrorCode ierr;
2809 
2810   PetscFunctionBegin;
2811   PetscValidHeaderSpecific(mat,MAT_CLASSID,1);
2812   PetscValidType(mat,1);
2813   if (perm) PetscValidHeaderSpecific(perm,IS_CLASSID,2);
2814   PetscValidPointer(info,3);
2815   PetscValidPointer(fact,4);
2816   if (mat->rmap->N != mat->cmap->N) SETERRQ(((PetscObject)mat)->comm,PETSC_ERR_ARG_WRONG,"Matrix must be square");
2817   if (!mat->assembled) SETERRQ(((PetscObject)mat)->comm,PETSC_ERR_ARG_WRONGSTATE,"Not for unassembled matrix");
2818   if (mat->factortype) SETERRQ(((PetscObject)mat)->comm,PETSC_ERR_ARG_WRONGSTATE,"Not for factored matrix");
2819   if (!(fact)->ops->choleskyfactorsymbolic) SETERRQ1(((PetscObject)mat)->comm,PETSC_ERR_SUP,"Mat type %s",((PetscObject)mat)->type_name);
2820   ierr = MatPreallocated(mat);CHKERRQ(ierr);
2821 
2822   ierr = PetscLogEventBegin(MAT_CholeskyFactorSymbolic,mat,perm,0,0);CHKERRQ(ierr);
2823   ierr = (fact->ops->choleskyfactorsymbolic)(fact,mat,perm,info);CHKERRQ(ierr);
2824   ierr = PetscLogEventEnd(MAT_CholeskyFactorSymbolic,mat,perm,0,0);CHKERRQ(ierr);
2825   ierr = PetscObjectStateIncrease((PetscObject)fact);CHKERRQ(ierr);
2826   PetscFunctionReturn(0);
2827 }
2828 
2829 #undef __FUNCT__
2830 #define __FUNCT__ "MatCholeskyFactorNumeric"
2831 /*@C
2832    MatCholeskyFactorNumeric - Performs numeric Cholesky factorization
2833    of a symmetric matrix. Call this routine after first calling
2834    MatCholeskyFactorSymbolic().
2835 
2836    Collective on Mat
2837 
2838    Input Parameters:
2839 +  fact - the factor matrix obtained with MatGetFactor()
2840 .  mat - the initial matrix
2841 .  info - options for factorization
2842 -  fact - the symbolic factor of mat
2843 
2844 
2845    Notes:
2846    Most users should employ the simplified KSP interface for linear solvers
2847    instead of working directly with matrix algebra routines such as this.
2848    See, e.g., KSPCreate().
2849 
2850    Level: developer
2851 
2852    Concepts: matrices^Cholesky numeric factorization
2853 
2854 .seealso: MatCholeskyFactorSymbolic(), MatCholeskyFactor(), MatLUFactorNumeric()
2855 
2856     Developer Note: fortran interface is not autogenerated as the f90
2857     interface defintion cannot be generated correctly [due to MatFactorInfo]
2858 
2859 @*/
2860 PetscErrorCode PETSCMAT_DLLEXPORT MatCholeskyFactorNumeric(Mat fact,Mat mat,const MatFactorInfo *info)
2861 {
2862   PetscErrorCode ierr;
2863 
2864   PetscFunctionBegin;
2865   PetscValidHeaderSpecific(mat,MAT_CLASSID,1);
2866   PetscValidType(mat,1);
2867   PetscValidPointer(fact,2);
2868   PetscValidHeaderSpecific(fact,MAT_CLASSID,2);
2869   if (!mat->assembled) SETERRQ(((PetscObject)mat)->comm,PETSC_ERR_ARG_WRONGSTATE,"Not for unassembled matrix");
2870   if (!(fact)->ops->choleskyfactornumeric) SETERRQ1(((PetscObject)mat)->comm,PETSC_ERR_SUP,"Mat type %s",((PetscObject)mat)->type_name);
2871   if (mat->rmap->N != (fact)->rmap->N || mat->cmap->N != (fact)->cmap->N) {
2872     SETERRQ4(((PetscObject)mat)->comm,PETSC_ERR_ARG_SIZ,"Mat mat,Mat fact: global dim %D should = %D %D should = %D",mat->rmap->N,(fact)->rmap->N,mat->cmap->N,(fact)->cmap->N);
2873   }
2874   ierr = MatPreallocated(mat);CHKERRQ(ierr);
2875 
2876   ierr = PetscLogEventBegin(MAT_CholeskyFactorNumeric,mat,fact,0,0);CHKERRQ(ierr);
2877   ierr = (fact->ops->choleskyfactornumeric)(fact,mat,info);CHKERRQ(ierr);
2878   ierr = PetscLogEventEnd(MAT_CholeskyFactorNumeric,mat,fact,0,0);CHKERRQ(ierr);
2879 
2880   ierr = MatView_Private(fact);CHKERRQ(ierr);
2881   ierr = PetscObjectStateIncrease((PetscObject)fact);CHKERRQ(ierr);
2882   PetscFunctionReturn(0);
2883 }
2884 
2885 /* ----------------------------------------------------------------*/
2886 #undef __FUNCT__
2887 #define __FUNCT__ "MatSolve"
2888 /*@
2889    MatSolve - Solves A x = b, given a factored matrix.
2890 
2891    Collective on Mat and Vec
2892 
2893    Input Parameters:
2894 +  mat - the factored matrix
2895 -  b - the right-hand-side vector
2896 
2897    Output Parameter:
2898 .  x - the result vector
2899 
2900    Notes:
2901    The vectors b and x cannot be the same.  I.e., one cannot
2902    call MatSolve(A,x,x).
2903 
2904    Notes:
2905    Most users should employ the simplified KSP interface for linear solvers
2906    instead of working directly with matrix algebra routines such as this.
2907    See, e.g., KSPCreate().
2908 
2909    Level: developer
2910 
2911    Concepts: matrices^triangular solves
2912 
2913 .seealso: MatSolveAdd(), MatSolveTranspose(), MatSolveTransposeAdd()
2914 @*/
2915 PetscErrorCode PETSCMAT_DLLEXPORT MatSolve(Mat mat,Vec b,Vec x)
2916 {
2917   PetscErrorCode ierr;
2918 
2919   PetscFunctionBegin;
2920   PetscValidHeaderSpecific(mat,MAT_CLASSID,1);
2921   PetscValidType(mat,1);
2922   PetscValidHeaderSpecific(b,VEC_CLASSID,2);
2923   PetscValidHeaderSpecific(x,VEC_CLASSID,3);
2924   PetscCheckSameComm(mat,1,b,2);
2925   PetscCheckSameComm(mat,1,x,3);
2926   if (x == b) SETERRQ(((PetscObject)mat)->comm,PETSC_ERR_ARG_IDN,"x and b must be different vectors");
2927   if (!mat->factortype) SETERRQ(((PetscObject)mat)->comm,PETSC_ERR_ARG_WRONGSTATE,"Unfactored matrix");
2928   if (mat->cmap->N != x->map->N) SETERRQ2(((PetscObject)mat)->comm,PETSC_ERR_ARG_SIZ,"Mat mat,Vec x: global dim %D %D",mat->cmap->N,x->map->N);
2929   if (mat->rmap->N != b->map->N) SETERRQ2(((PetscObject)mat)->comm,PETSC_ERR_ARG_SIZ,"Mat mat,Vec b: global dim %D %D",mat->rmap->N,b->map->N);
2930   if (mat->rmap->n != b->map->n) SETERRQ2(PETSC_COMM_SELF,PETSC_ERR_ARG_SIZ,"Mat mat,Vec b: local dim %D %D",mat->rmap->n,b->map->n);
2931   if (!mat->rmap->N && !mat->cmap->N) PetscFunctionReturn(0);
2932   if (!mat->ops->solve) SETERRQ1(((PetscObject)mat)->comm,PETSC_ERR_SUP,"Mat type %s",((PetscObject)mat)->type_name);
2933   ierr = MatPreallocated(mat);CHKERRQ(ierr);
2934 
2935   ierr = PetscLogEventBegin(MAT_Solve,mat,b,x,0);CHKERRQ(ierr);
2936   ierr = (*mat->ops->solve)(mat,b,x);CHKERRQ(ierr);
2937   ierr = PetscLogEventEnd(MAT_Solve,mat,b,x,0);CHKERRQ(ierr);
2938   ierr = PetscObjectStateIncrease((PetscObject)x);CHKERRQ(ierr);
2939   PetscFunctionReturn(0);
2940 }
2941 
2942 #undef __FUNCT__
2943 #define __FUNCT__ "MatMatSolve_Basic"
2944 PetscErrorCode PETSCMAT_DLLEXPORT MatMatSolve_Basic(Mat A,Mat B,Mat X)
2945 {
2946   PetscErrorCode ierr;
2947   Vec            b,x;
2948   PetscInt       m,N,i;
2949   PetscScalar    *bb,*xx;
2950 
2951   PetscFunctionBegin;
2952   ierr = MatGetArray(B,&bb);CHKERRQ(ierr);
2953   ierr = MatGetArray(X,&xx);CHKERRQ(ierr);
2954   ierr = MatGetLocalSize(B,&m,PETSC_NULL);CHKERRQ(ierr);  /* number local rows */
2955   ierr = MatGetSize(B,PETSC_NULL,&N);CHKERRQ(ierr);       /* total columns in dense matrix */
2956   ierr = MatGetVecs(A,&x,&b);CHKERRQ(ierr);
2957   for (i=0; i<N; i++) {
2958     ierr = VecPlaceArray(b,bb + i*m);CHKERRQ(ierr);
2959     ierr = VecPlaceArray(x,xx + i*m);CHKERRQ(ierr);
2960     ierr = MatSolve(A,b,x);CHKERRQ(ierr);
2961     ierr = VecResetArray(x);CHKERRQ(ierr);
2962     ierr = VecResetArray(b);CHKERRQ(ierr);
2963   }
2964   ierr = VecDestroy(b);CHKERRQ(ierr);
2965   ierr = VecDestroy(x);CHKERRQ(ierr);
2966   ierr = MatRestoreArray(B,&bb);CHKERRQ(ierr);
2967   ierr = MatRestoreArray(X,&xx);CHKERRQ(ierr);
2968   PetscFunctionReturn(0);
2969 }
2970 
2971 #undef __FUNCT__
2972 #define __FUNCT__ "MatMatSolve"
2973 /*@
2974    MatMatSolve - Solves A X = B, given a factored matrix.
2975 
2976    Collective on Mat
2977 
2978    Input Parameters:
2979 +  mat - the factored matrix
2980 -  B - the right-hand-side matrix  (dense matrix)
2981 
2982    Output Parameter:
2983 .  X - the result matrix (dense matrix)
2984 
2985    Notes:
2986    The matrices b and x cannot be the same.  I.e., one cannot
2987    call MatMatSolve(A,x,x).
2988 
2989    Notes:
2990    Most users should usually employ the simplified KSP interface for linear solvers
2991    instead of working directly with matrix algebra routines such as this.
2992    See, e.g., KSPCreate(). However KSP can only solve for one vector (column of X)
2993    at a time.
2994 
2995    Level: developer
2996 
2997    Concepts: matrices^triangular solves
2998 
2999 .seealso: MatMatSolveAdd(), MatMatSolveTranspose(), MatMatSolveTransposeAdd(), MatLUFactor(), MatCholeskyFactor()
3000 @*/
3001 PetscErrorCode PETSCMAT_DLLEXPORT MatMatSolve(Mat A,Mat B,Mat X)
3002 {
3003   PetscErrorCode ierr;
3004 
3005   PetscFunctionBegin;
3006   PetscValidHeaderSpecific(A,MAT_CLASSID,1);
3007   PetscValidType(A,1);
3008   PetscValidHeaderSpecific(B,MAT_CLASSID,2);
3009   PetscValidHeaderSpecific(X,MAT_CLASSID,3);
3010   PetscCheckSameComm(A,1,B,2);
3011   PetscCheckSameComm(A,1,X,3);
3012   if (X == B) SETERRQ(((PetscObject)A)->comm,PETSC_ERR_ARG_IDN,"X and B must be different matrices");
3013   if (!A->factortype) SETERRQ(((PetscObject)A)->comm,PETSC_ERR_ARG_WRONGSTATE,"Unfactored matrix");
3014   if (A->cmap->N != X->rmap->N) SETERRQ2(((PetscObject)A)->comm,PETSC_ERR_ARG_SIZ,"Mat A,Mat X: global dim %D %D",A->cmap->N,X->rmap->N);
3015   if (A->rmap->N != B->rmap->N) SETERRQ2(((PetscObject)A)->comm,PETSC_ERR_ARG_SIZ,"Mat A,Mat B: global dim %D %D",A->rmap->N,B->rmap->N);
3016   if (A->rmap->n != B->rmap->n) SETERRQ2(PETSC_COMM_SELF,PETSC_ERR_ARG_SIZ,"Mat A,Mat B: local dim %D %D",A->rmap->n,B->rmap->n);
3017   if (!A->rmap->N && !A->cmap->N) PetscFunctionReturn(0);
3018   ierr = MatPreallocated(A);CHKERRQ(ierr);
3019 
3020   ierr = PetscLogEventBegin(MAT_MatSolve,A,B,X,0);CHKERRQ(ierr);
3021   if (!A->ops->matsolve) {
3022     ierr = PetscInfo1(A,"Mat type %s using basic MatMatSolve",((PetscObject)A)->type_name);CHKERRQ(ierr);
3023     ierr = MatMatSolve_Basic(A,B,X);CHKERRQ(ierr);
3024   } else {
3025     ierr = (*A->ops->matsolve)(A,B,X);CHKERRQ(ierr);
3026   }
3027   ierr = PetscLogEventEnd(MAT_MatSolve,A,B,X,0);CHKERRQ(ierr);
3028   ierr = PetscObjectStateIncrease((PetscObject)X);CHKERRQ(ierr);
3029   PetscFunctionReturn(0);
3030 }
3031 
3032 
3033 #undef __FUNCT__
3034 #define __FUNCT__ "MatForwardSolve"
3035 /*@
3036    MatForwardSolve - Solves L x = b, given a factored matrix, A = LU, or
3037                             U^T*D^(1/2) x = b, given a factored symmetric matrix, A = U^T*D*U,
3038 
3039    Collective on Mat and Vec
3040 
3041    Input Parameters:
3042 +  mat - the factored matrix
3043 -  b - the right-hand-side vector
3044 
3045    Output Parameter:
3046 .  x - the result vector
3047 
3048    Notes:
3049    MatSolve() should be used for most applications, as it performs
3050    a forward solve followed by a backward solve.
3051 
3052    The vectors b and x cannot be the same,  i.e., one cannot
3053    call MatForwardSolve(A,x,x).
3054 
3055    For matrix in seqsbaij format with block size larger than 1,
3056    the diagonal blocks are not implemented as D = D^(1/2) * D^(1/2) yet.
3057    MatForwardSolve() solves U^T*D y = b, and
3058    MatBackwardSolve() solves U x = y.
3059    Thus they do not provide a symmetric preconditioner.
3060 
3061    Most users should employ the simplified KSP interface for linear solvers
3062    instead of working directly with matrix algebra routines such as this.
3063    See, e.g., KSPCreate().
3064 
3065    Level: developer
3066 
3067    Concepts: matrices^forward solves
3068 
3069 .seealso: MatSolve(), MatBackwardSolve()
3070 @*/
3071 PetscErrorCode PETSCMAT_DLLEXPORT MatForwardSolve(Mat mat,Vec b,Vec x)
3072 {
3073   PetscErrorCode ierr;
3074 
3075   PetscFunctionBegin;
3076   PetscValidHeaderSpecific(mat,MAT_CLASSID,1);
3077   PetscValidType(mat,1);
3078   PetscValidHeaderSpecific(b,VEC_CLASSID,2);
3079   PetscValidHeaderSpecific(x,VEC_CLASSID,3);
3080   PetscCheckSameComm(mat,1,b,2);
3081   PetscCheckSameComm(mat,1,x,3);
3082   if (x == b) SETERRQ(((PetscObject)mat)->comm,PETSC_ERR_ARG_IDN,"x and b must be different vectors");
3083   if (!mat->factortype) SETERRQ(((PetscObject)mat)->comm,PETSC_ERR_ARG_WRONGSTATE,"Unfactored matrix");
3084   if (!mat->ops->forwardsolve) SETERRQ1(((PetscObject)mat)->comm,PETSC_ERR_SUP,"Mat type %s",((PetscObject)mat)->type_name);
3085   if (mat->cmap->N != x->map->N) SETERRQ2(((PetscObject)mat)->comm,PETSC_ERR_ARG_SIZ,"Mat mat,Vec x: global dim %D %D",mat->cmap->N,x->map->N);
3086   if (mat->rmap->N != b->map->N) SETERRQ2(((PetscObject)mat)->comm,PETSC_ERR_ARG_SIZ,"Mat mat,Vec b: global dim %D %D",mat->rmap->N,b->map->N);
3087   if (mat->rmap->n != b->map->n) SETERRQ2(PETSC_COMM_SELF,PETSC_ERR_ARG_SIZ,"Mat mat,Vec b: local dim %D %D",mat->rmap->n,b->map->n);
3088   ierr = MatPreallocated(mat);CHKERRQ(ierr);
3089   ierr = PetscLogEventBegin(MAT_ForwardSolve,mat,b,x,0);CHKERRQ(ierr);
3090   ierr = (*mat->ops->forwardsolve)(mat,b,x);CHKERRQ(ierr);
3091   ierr = PetscLogEventEnd(MAT_ForwardSolve,mat,b,x,0);CHKERRQ(ierr);
3092   ierr = PetscObjectStateIncrease((PetscObject)x);CHKERRQ(ierr);
3093   PetscFunctionReturn(0);
3094 }
3095 
3096 #undef __FUNCT__
3097 #define __FUNCT__ "MatBackwardSolve"
3098 /*@
3099    MatBackwardSolve - Solves U x = b, given a factored matrix, A = LU.
3100                              D^(1/2) U x = b, given a factored symmetric matrix, A = U^T*D*U,
3101 
3102    Collective on Mat and Vec
3103 
3104    Input Parameters:
3105 +  mat - the factored matrix
3106 -  b - the right-hand-side vector
3107 
3108    Output Parameter:
3109 .  x - the result vector
3110 
3111    Notes:
3112    MatSolve() should be used for most applications, as it performs
3113    a forward solve followed by a backward solve.
3114 
3115    The vectors b and x cannot be the same.  I.e., one cannot
3116    call MatBackwardSolve(A,x,x).
3117 
3118    For matrix in seqsbaij format with block size larger than 1,
3119    the diagonal blocks are not implemented as D = D^(1/2) * D^(1/2) yet.
3120    MatForwardSolve() solves U^T*D y = b, and
3121    MatBackwardSolve() solves U x = y.
3122    Thus they do not provide a symmetric preconditioner.
3123 
3124    Most users should employ the simplified KSP interface for linear solvers
3125    instead of working directly with matrix algebra routines such as this.
3126    See, e.g., KSPCreate().
3127 
3128    Level: developer
3129 
3130    Concepts: matrices^backward solves
3131 
3132 .seealso: MatSolve(), MatForwardSolve()
3133 @*/
3134 PetscErrorCode PETSCMAT_DLLEXPORT MatBackwardSolve(Mat mat,Vec b,Vec x)
3135 {
3136   PetscErrorCode ierr;
3137 
3138   PetscFunctionBegin;
3139   PetscValidHeaderSpecific(mat,MAT_CLASSID,1);
3140   PetscValidType(mat,1);
3141   PetscValidHeaderSpecific(b,VEC_CLASSID,2);
3142   PetscValidHeaderSpecific(x,VEC_CLASSID,3);
3143   PetscCheckSameComm(mat,1,b,2);
3144   PetscCheckSameComm(mat,1,x,3);
3145   if (x == b) SETERRQ(((PetscObject)mat)->comm,PETSC_ERR_ARG_IDN,"x and b must be different vectors");
3146   if (!mat->factortype) SETERRQ(((PetscObject)mat)->comm,PETSC_ERR_ARG_WRONGSTATE,"Unfactored matrix");
3147   if (!mat->ops->backwardsolve) SETERRQ1(((PetscObject)mat)->comm,PETSC_ERR_SUP,"Mat type %s",((PetscObject)mat)->type_name);
3148   if (mat->cmap->N != x->map->N) SETERRQ2(((PetscObject)mat)->comm,PETSC_ERR_ARG_SIZ,"Mat mat,Vec x: global dim %D %D",mat->cmap->N,x->map->N);
3149   if (mat->rmap->N != b->map->N) SETERRQ2(((PetscObject)mat)->comm,PETSC_ERR_ARG_SIZ,"Mat mat,Vec b: global dim %D %D",mat->rmap->N,b->map->N);
3150   if (mat->rmap->n != b->map->n) SETERRQ2(PETSC_COMM_SELF,PETSC_ERR_ARG_SIZ,"Mat mat,Vec b: local dim %D %D",mat->rmap->n,b->map->n);
3151   ierr = MatPreallocated(mat);CHKERRQ(ierr);
3152 
3153   ierr = PetscLogEventBegin(MAT_BackwardSolve,mat,b,x,0);CHKERRQ(ierr);
3154   ierr = (*mat->ops->backwardsolve)(mat,b,x);CHKERRQ(ierr);
3155   ierr = PetscLogEventEnd(MAT_BackwardSolve,mat,b,x,0);CHKERRQ(ierr);
3156   ierr = PetscObjectStateIncrease((PetscObject)x);CHKERRQ(ierr);
3157   PetscFunctionReturn(0);
3158 }
3159 
3160 #undef __FUNCT__
3161 #define __FUNCT__ "MatSolveAdd"
3162 /*@
3163    MatSolveAdd - Computes x = y + inv(A)*b, given a factored matrix.
3164 
3165    Collective on Mat and Vec
3166 
3167    Input Parameters:
3168 +  mat - the factored matrix
3169 .  b - the right-hand-side vector
3170 -  y - the vector to be added to
3171 
3172    Output Parameter:
3173 .  x - the result vector
3174 
3175    Notes:
3176    The vectors b and x cannot be the same.  I.e., one cannot
3177    call MatSolveAdd(A,x,y,x).
3178 
3179    Most users should employ the simplified KSP interface for linear solvers
3180    instead of working directly with matrix algebra routines such as this.
3181    See, e.g., KSPCreate().
3182 
3183    Level: developer
3184 
3185    Concepts: matrices^triangular solves
3186 
3187 .seealso: MatSolve(), MatSolveTranspose(), MatSolveTransposeAdd()
3188 @*/
3189 PetscErrorCode PETSCMAT_DLLEXPORT MatSolveAdd(Mat mat,Vec b,Vec y,Vec x)
3190 {
3191   PetscScalar    one = 1.0;
3192   Vec            tmp;
3193   PetscErrorCode ierr;
3194 
3195   PetscFunctionBegin;
3196   PetscValidHeaderSpecific(mat,MAT_CLASSID,1);
3197   PetscValidType(mat,1);
3198   PetscValidHeaderSpecific(y,VEC_CLASSID,2);
3199   PetscValidHeaderSpecific(b,VEC_CLASSID,3);
3200   PetscValidHeaderSpecific(x,VEC_CLASSID,4);
3201   PetscCheckSameComm(mat,1,b,2);
3202   PetscCheckSameComm(mat,1,y,2);
3203   PetscCheckSameComm(mat,1,x,3);
3204   if (x == b) SETERRQ(((PetscObject)mat)->comm,PETSC_ERR_ARG_IDN,"x and b must be different vectors");
3205   if (!mat->factortype) SETERRQ(((PetscObject)mat)->comm,PETSC_ERR_ARG_WRONGSTATE,"Unfactored matrix");
3206   if (mat->cmap->N != x->map->N) SETERRQ2(((PetscObject)mat)->comm,PETSC_ERR_ARG_SIZ,"Mat mat,Vec x: global dim %D %D",mat->cmap->N,x->map->N);
3207   if (mat->rmap->N != b->map->N) SETERRQ2(((PetscObject)mat)->comm,PETSC_ERR_ARG_SIZ,"Mat mat,Vec b: global dim %D %D",mat->rmap->N,b->map->N);
3208   if (mat->rmap->N != y->map->N) SETERRQ2(((PetscObject)mat)->comm,PETSC_ERR_ARG_SIZ,"Mat mat,Vec y: global dim %D %D",mat->rmap->N,y->map->N);
3209   if (mat->rmap->n != b->map->n) SETERRQ2(PETSC_COMM_SELF,PETSC_ERR_ARG_SIZ,"Mat mat,Vec b: local dim %D %D",mat->rmap->n,b->map->n);
3210   if (x->map->n != y->map->n) SETERRQ2(PETSC_COMM_SELF,PETSC_ERR_ARG_SIZ,"Vec x,Vec y: local dim %D %D",x->map->n,y->map->n);
3211   ierr = MatPreallocated(mat);CHKERRQ(ierr);
3212 
3213   ierr = PetscLogEventBegin(MAT_SolveAdd,mat,b,x,y);CHKERRQ(ierr);
3214   if (mat->ops->solveadd)  {
3215     ierr = (*mat->ops->solveadd)(mat,b,y,x);CHKERRQ(ierr);
3216   } else {
3217     /* do the solve then the add manually */
3218     if (x != y) {
3219       ierr = MatSolve(mat,b,x);CHKERRQ(ierr);
3220       ierr = VecAXPY(x,one,y);CHKERRQ(ierr);
3221     } else {
3222       ierr = VecDuplicate(x,&tmp);CHKERRQ(ierr);
3223       ierr = PetscLogObjectParent(mat,tmp);CHKERRQ(ierr);
3224       ierr = VecCopy(x,tmp);CHKERRQ(ierr);
3225       ierr = MatSolve(mat,b,x);CHKERRQ(ierr);
3226       ierr = VecAXPY(x,one,tmp);CHKERRQ(ierr);
3227       ierr = VecDestroy(tmp);CHKERRQ(ierr);
3228     }
3229   }
3230   ierr = PetscLogEventEnd(MAT_SolveAdd,mat,b,x,y);CHKERRQ(ierr);
3231   ierr = PetscObjectStateIncrease((PetscObject)x);CHKERRQ(ierr);
3232   PetscFunctionReturn(0);
3233 }
3234 
3235 #undef __FUNCT__
3236 #define __FUNCT__ "MatSolveTranspose"
3237 /*@
3238    MatSolveTranspose - Solves A' x = b, given a factored matrix.
3239 
3240    Collective on Mat and Vec
3241 
3242    Input Parameters:
3243 +  mat - the factored matrix
3244 -  b - the right-hand-side vector
3245 
3246    Output Parameter:
3247 .  x - the result vector
3248 
3249    Notes:
3250    The vectors b and x cannot be the same.  I.e., one cannot
3251    call MatSolveTranspose(A,x,x).
3252 
3253    Most users should employ the simplified KSP interface for linear solvers
3254    instead of working directly with matrix algebra routines such as this.
3255    See, e.g., KSPCreate().
3256 
3257    Level: developer
3258 
3259    Concepts: matrices^triangular solves
3260 
3261 .seealso: MatSolve(), MatSolveAdd(), MatSolveTransposeAdd()
3262 @*/
3263 PetscErrorCode PETSCMAT_DLLEXPORT MatSolveTranspose(Mat mat,Vec b,Vec x)
3264 {
3265   PetscErrorCode ierr;
3266 
3267   PetscFunctionBegin;
3268   PetscValidHeaderSpecific(mat,MAT_CLASSID,1);
3269   PetscValidType(mat,1);
3270   PetscValidHeaderSpecific(b,VEC_CLASSID,2);
3271   PetscValidHeaderSpecific(x,VEC_CLASSID,3);
3272   PetscCheckSameComm(mat,1,b,2);
3273   PetscCheckSameComm(mat,1,x,3);
3274   if (!mat->factortype) SETERRQ(((PetscObject)mat)->comm,PETSC_ERR_ARG_WRONGSTATE,"Unfactored matrix");
3275   if (x == b) SETERRQ(((PetscObject)mat)->comm,PETSC_ERR_ARG_IDN,"x and b must be different vectors");
3276   if (!mat->ops->solvetranspose) SETERRQ1(((PetscObject)mat)->comm,PETSC_ERR_SUP,"Matrix type %s",((PetscObject)mat)->type_name);
3277   if (mat->rmap->N != x->map->N) SETERRQ2(((PetscObject)mat)->comm,PETSC_ERR_ARG_SIZ,"Mat mat,Vec x: global dim %D %D",mat->rmap->N,x->map->N);
3278   if (mat->cmap->N != b->map->N) SETERRQ2(((PetscObject)mat)->comm,PETSC_ERR_ARG_SIZ,"Mat mat,Vec b: global dim %D %D",mat->cmap->N,b->map->N);
3279   ierr = MatPreallocated(mat);CHKERRQ(ierr);
3280   ierr = PetscLogEventBegin(MAT_SolveTranspose,mat,b,x,0);CHKERRQ(ierr);
3281   ierr = (*mat->ops->solvetranspose)(mat,b,x);CHKERRQ(ierr);
3282   ierr = PetscLogEventEnd(MAT_SolveTranspose,mat,b,x,0);CHKERRQ(ierr);
3283   ierr = PetscObjectStateIncrease((PetscObject)x);CHKERRQ(ierr);
3284   PetscFunctionReturn(0);
3285 }
3286 
3287 #undef __FUNCT__
3288 #define __FUNCT__ "MatSolveTransposeAdd"
3289 /*@
3290    MatSolveTransposeAdd - Computes x = y + inv(Transpose(A)) b, given a
3291                       factored matrix.
3292 
3293    Collective on Mat and Vec
3294 
3295    Input Parameters:
3296 +  mat - the factored matrix
3297 .  b - the right-hand-side vector
3298 -  y - the vector to be added to
3299 
3300    Output Parameter:
3301 .  x - the result vector
3302 
3303    Notes:
3304    The vectors b and x cannot be the same.  I.e., one cannot
3305    call MatSolveTransposeAdd(A,x,y,x).
3306 
3307    Most users should employ the simplified KSP interface for linear solvers
3308    instead of working directly with matrix algebra routines such as this.
3309    See, e.g., KSPCreate().
3310 
3311    Level: developer
3312 
3313    Concepts: matrices^triangular solves
3314 
3315 .seealso: MatSolve(), MatSolveAdd(), MatSolveTranspose()
3316 @*/
3317 PetscErrorCode PETSCMAT_DLLEXPORT MatSolveTransposeAdd(Mat mat,Vec b,Vec y,Vec x)
3318 {
3319   PetscScalar    one = 1.0;
3320   PetscErrorCode ierr;
3321   Vec            tmp;
3322 
3323   PetscFunctionBegin;
3324   PetscValidHeaderSpecific(mat,MAT_CLASSID,1);
3325   PetscValidType(mat,1);
3326   PetscValidHeaderSpecific(y,VEC_CLASSID,2);
3327   PetscValidHeaderSpecific(b,VEC_CLASSID,3);
3328   PetscValidHeaderSpecific(x,VEC_CLASSID,4);
3329   PetscCheckSameComm(mat,1,b,2);
3330   PetscCheckSameComm(mat,1,y,3);
3331   PetscCheckSameComm(mat,1,x,4);
3332   if (x == b) SETERRQ(((PetscObject)mat)->comm,PETSC_ERR_ARG_IDN,"x and b must be different vectors");
3333   if (!mat->factortype) SETERRQ(((PetscObject)mat)->comm,PETSC_ERR_ARG_WRONGSTATE,"Unfactored matrix");
3334   if (mat->rmap->N != x->map->N) SETERRQ2(((PetscObject)mat)->comm,PETSC_ERR_ARG_SIZ,"Mat mat,Vec x: global dim %D %D",mat->rmap->N,x->map->N);
3335   if (mat->cmap->N != b->map->N) SETERRQ2(((PetscObject)mat)->comm,PETSC_ERR_ARG_SIZ,"Mat mat,Vec b: global dim %D %D",mat->cmap->N,b->map->N);
3336   if (mat->cmap->N != y->map->N) SETERRQ2(((PetscObject)mat)->comm,PETSC_ERR_ARG_SIZ,"Mat mat,Vec y: global dim %D %D",mat->cmap->N,y->map->N);
3337   if (x->map->n != y->map->n)   SETERRQ2(PETSC_COMM_SELF,PETSC_ERR_ARG_SIZ,"Vec x,Vec y: local dim %D %D",x->map->n,y->map->n);
3338   ierr = MatPreallocated(mat);CHKERRQ(ierr);
3339 
3340   ierr = PetscLogEventBegin(MAT_SolveTransposeAdd,mat,b,x,y);CHKERRQ(ierr);
3341   if (mat->ops->solvetransposeadd) {
3342     ierr = (*mat->ops->solvetransposeadd)(mat,b,y,x);CHKERRQ(ierr);
3343   } else {
3344     /* do the solve then the add manually */
3345     if (x != y) {
3346       ierr = MatSolveTranspose(mat,b,x);CHKERRQ(ierr);
3347       ierr = VecAXPY(x,one,y);CHKERRQ(ierr);
3348     } else {
3349       ierr = VecDuplicate(x,&tmp);CHKERRQ(ierr);
3350       ierr = PetscLogObjectParent(mat,tmp);CHKERRQ(ierr);
3351       ierr = VecCopy(x,tmp);CHKERRQ(ierr);
3352       ierr = MatSolveTranspose(mat,b,x);CHKERRQ(ierr);
3353       ierr = VecAXPY(x,one,tmp);CHKERRQ(ierr);
3354       ierr = VecDestroy(tmp);CHKERRQ(ierr);
3355     }
3356   }
3357   ierr = PetscLogEventEnd(MAT_SolveTransposeAdd,mat,b,x,y);CHKERRQ(ierr);
3358   ierr = PetscObjectStateIncrease((PetscObject)x);CHKERRQ(ierr);
3359   PetscFunctionReturn(0);
3360 }
3361 /* ----------------------------------------------------------------*/
3362 
3363 #undef __FUNCT__
3364 #define __FUNCT__ "MatSOR"
3365 /*@
3366    MatSOR - Computes relaxation (SOR, Gauss-Seidel) sweeps.
3367 
3368    Collective on Mat and Vec
3369 
3370    Input Parameters:
3371 +  mat - the matrix
3372 .  b - the right hand side
3373 .  omega - the relaxation factor
3374 .  flag - flag indicating the type of SOR (see below)
3375 .  shift -  diagonal shift
3376 .  its - the number of iterations
3377 -  lits - the number of local iterations
3378 
3379    Output Parameters:
3380 .  x - the solution (can contain an initial guess, use option SOR_ZERO_INITIAL_GUESS to indicate no guess)
3381 
3382    SOR Flags:
3383 .     SOR_FORWARD_SWEEP - forward SOR
3384 .     SOR_BACKWARD_SWEEP - backward SOR
3385 .     SOR_SYMMETRIC_SWEEP - SSOR (symmetric SOR)
3386 .     SOR_LOCAL_FORWARD_SWEEP - local forward SOR
3387 .     SOR_LOCAL_BACKWARD_SWEEP - local forward SOR
3388 .     SOR_LOCAL_SYMMETRIC_SWEEP - local SSOR
3389 .     SOR_APPLY_UPPER, SOR_APPLY_LOWER - applies
3390          upper/lower triangular part of matrix to
3391          vector (with omega)
3392 .     SOR_ZERO_INITIAL_GUESS - zero initial guess
3393 
3394    Notes:
3395    SOR_LOCAL_FORWARD_SWEEP, SOR_LOCAL_BACKWARD_SWEEP, and
3396    SOR_LOCAL_SYMMETRIC_SWEEP perform separate independent smoothings
3397    on each processor.
3398 
3399    Application programmers will not generally use MatSOR() directly,
3400    but instead will employ the KSP/PC interface.
3401 
3402    Notes: for BAIJ, SBAIJ, and AIJ matrices with Inodes this does a block SOR smoothing, otherwise it does a pointwise smoothing
3403 
3404    Notes for Advanced Users:
3405    The flags are implemented as bitwise inclusive or operations.
3406    For example, use (SOR_ZERO_INITIAL_GUESS | SOR_SYMMETRIC_SWEEP)
3407    to specify a zero initial guess for SSOR.
3408 
3409    Most users should employ the simplified KSP interface for linear solvers
3410    instead of working directly with matrix algebra routines such as this.
3411    See, e.g., KSPCreate().
3412 
3413    Vectors x and b CANNOT be the same
3414 
3415    Level: developer
3416 
3417    Concepts: matrices^relaxation
3418    Concepts: matrices^SOR
3419    Concepts: matrices^Gauss-Seidel
3420 
3421 @*/
3422 PetscErrorCode PETSCMAT_DLLEXPORT MatSOR(Mat mat,Vec b,PetscReal omega,MatSORType flag,PetscReal shift,PetscInt its,PetscInt lits,Vec x)
3423 {
3424   PetscErrorCode ierr;
3425 
3426   PetscFunctionBegin;
3427   PetscValidHeaderSpecific(mat,MAT_CLASSID,1);
3428   PetscValidType(mat,1);
3429   PetscValidHeaderSpecific(b,VEC_CLASSID,2);
3430   PetscValidHeaderSpecific(x,VEC_CLASSID,8);
3431   PetscCheckSameComm(mat,1,b,2);
3432   PetscCheckSameComm(mat,1,x,8);
3433   if (!mat->ops->sor) SETERRQ1(((PetscObject)mat)->comm,PETSC_ERR_SUP,"Mat type %s",((PetscObject)mat)->type_name);
3434   if (!mat->assembled) SETERRQ(((PetscObject)mat)->comm,PETSC_ERR_ARG_WRONGSTATE,"Not for unassembled matrix");
3435   if (mat->factortype) SETERRQ(((PetscObject)mat)->comm,PETSC_ERR_ARG_WRONGSTATE,"Not for factored matrix");
3436   if (mat->cmap->N != x->map->N) SETERRQ2(((PetscObject)mat)->comm,PETSC_ERR_ARG_SIZ,"Mat mat,Vec x: global dim %D %D",mat->cmap->N,x->map->N);
3437   if (mat->rmap->N != b->map->N) SETERRQ2(((PetscObject)mat)->comm,PETSC_ERR_ARG_SIZ,"Mat mat,Vec b: global dim %D %D",mat->rmap->N,b->map->N);
3438   if (mat->rmap->n != b->map->n) SETERRQ2(PETSC_COMM_SELF,PETSC_ERR_ARG_SIZ,"Mat mat,Vec b: local dim %D %D",mat->rmap->n,b->map->n);
3439   if (its <= 0) SETERRQ1(PETSC_COMM_SELF,PETSC_ERR_ARG_WRONG,"Relaxation requires global its %D positive",its);
3440   if (lits <= 0) SETERRQ1(PETSC_COMM_SELF,PETSC_ERR_ARG_WRONG,"Relaxation requires local its %D positive",lits);
3441   if (b == x) SETERRQ(PETSC_COMM_SELF,PETSC_ERR_ARG_IDN,"b and x vector cannot be the same");
3442 
3443   ierr = MatPreallocated(mat);CHKERRQ(ierr);
3444   ierr = PetscLogEventBegin(MAT_SOR,mat,b,x,0);CHKERRQ(ierr);
3445   ierr =(*mat->ops->sor)(mat,b,omega,flag,shift,its,lits,x);CHKERRQ(ierr);
3446   ierr = PetscLogEventEnd(MAT_SOR,mat,b,x,0);CHKERRQ(ierr);
3447   ierr = PetscObjectStateIncrease((PetscObject)x);CHKERRQ(ierr);
3448   PetscFunctionReturn(0);
3449 }
3450 
3451 #undef __FUNCT__
3452 #define __FUNCT__ "MatCopy_Basic"
3453 /*
3454       Default matrix copy routine.
3455 */
3456 PetscErrorCode MatCopy_Basic(Mat A,Mat B,MatStructure str)
3457 {
3458   PetscErrorCode    ierr;
3459   PetscInt          i,rstart = 0,rend = 0,nz;
3460   const PetscInt    *cwork;
3461   const PetscScalar *vwork;
3462 
3463   PetscFunctionBegin;
3464   if (B->assembled) {
3465     ierr = MatZeroEntries(B);CHKERRQ(ierr);
3466   }
3467   ierr = MatGetOwnershipRange(A,&rstart,&rend);CHKERRQ(ierr);
3468   for (i=rstart; i<rend; i++) {
3469     ierr = MatGetRow(A,i,&nz,&cwork,&vwork);CHKERRQ(ierr);
3470     ierr = MatSetValues(B,1,&i,nz,cwork,vwork,INSERT_VALUES);CHKERRQ(ierr);
3471     ierr = MatRestoreRow(A,i,&nz,&cwork,&vwork);CHKERRQ(ierr);
3472   }
3473   ierr = MatAssemblyBegin(B,MAT_FINAL_ASSEMBLY);CHKERRQ(ierr);
3474   ierr = MatAssemblyEnd(B,MAT_FINAL_ASSEMBLY);CHKERRQ(ierr);
3475   ierr = PetscObjectStateIncrease((PetscObject)B);CHKERRQ(ierr);
3476   PetscFunctionReturn(0);
3477 }
3478 
3479 #undef __FUNCT__
3480 #define __FUNCT__ "MatCopy"
3481 /*@
3482    MatCopy - Copys a matrix to another matrix.
3483 
3484    Collective on Mat
3485 
3486    Input Parameters:
3487 +  A - the matrix
3488 -  str - SAME_NONZERO_PATTERN or DIFFERENT_NONZERO_PATTERN
3489 
3490    Output Parameter:
3491 .  B - where the copy is put
3492 
3493    Notes:
3494    If you use SAME_NONZERO_PATTERN then the two matrices had better have the
3495    same nonzero pattern or the routine will crash.
3496 
3497    MatCopy() copies the matrix entries of a matrix to another existing
3498    matrix (after first zeroing the second matrix).  A related routine is
3499    MatConvert(), which first creates a new matrix and then copies the data.
3500 
3501    Level: intermediate
3502 
3503    Concepts: matrices^copying
3504 
3505 .seealso: MatConvert(), MatDuplicate()
3506 
3507 @*/
3508 PetscErrorCode PETSCMAT_DLLEXPORT MatCopy(Mat A,Mat B,MatStructure str)
3509 {
3510   PetscErrorCode ierr;
3511   PetscInt       i;
3512 
3513   PetscFunctionBegin;
3514   PetscValidHeaderSpecific(A,MAT_CLASSID,1);
3515   PetscValidHeaderSpecific(B,MAT_CLASSID,2);
3516   PetscValidType(A,1);
3517   PetscValidType(B,2);
3518   PetscCheckSameComm(A,1,B,2);
3519   ierr = MatPreallocated(B);CHKERRQ(ierr);
3520   if (!A->assembled) SETERRQ(((PetscObject)A)->comm,PETSC_ERR_ARG_WRONGSTATE,"Not for unassembled matrix");
3521   if (A->factortype) SETERRQ(((PetscObject)A)->comm,PETSC_ERR_ARG_WRONGSTATE,"Not for factored matrix");
3522   if (A->rmap->N != B->rmap->N || A->cmap->N != B->cmap->N) SETERRQ4(((PetscObject)A)->comm,PETSC_ERR_ARG_SIZ,"Mat A,Mat B: global dim (%D,%D) (%D,%D)",A->rmap->N,B->rmap->N,A->cmap->N,B->cmap->N);
3523   ierr = MatPreallocated(A);CHKERRQ(ierr);
3524 
3525   ierr = PetscLogEventBegin(MAT_Copy,A,B,0,0);CHKERRQ(ierr);
3526   if (A->ops->copy) {
3527     ierr = (*A->ops->copy)(A,B,str);CHKERRQ(ierr);
3528   } else { /* generic conversion */
3529     ierr = MatCopy_Basic(A,B,str);CHKERRQ(ierr);
3530   }
3531   if (A->mapping) {
3532     if (B->mapping) {ierr = ISLocalToGlobalMappingDestroy(B->mapping);CHKERRQ(ierr);B->mapping = 0;}
3533     ierr = MatSetLocalToGlobalMapping(B,A->mapping);CHKERRQ(ierr);
3534   }
3535   if (A->bmapping) {
3536     if (B->bmapping) {ierr = ISLocalToGlobalMappingDestroy(B->bmapping);CHKERRQ(ierr);B->bmapping = 0;}
3537     ierr = MatSetLocalToGlobalMappingBlock(B,A->mapping);CHKERRQ(ierr);
3538   }
3539 
3540   B->stencil.dim = A->stencil.dim;
3541   B->stencil.noc = A->stencil.noc;
3542   for (i=0; i<=A->stencil.dim; i++) {
3543     B->stencil.dims[i]   = A->stencil.dims[i];
3544     B->stencil.starts[i] = A->stencil.starts[i];
3545   }
3546 
3547   ierr = PetscLogEventEnd(MAT_Copy,A,B,0,0);CHKERRQ(ierr);
3548   ierr = PetscObjectStateIncrease((PetscObject)B);CHKERRQ(ierr);
3549   PetscFunctionReturn(0);
3550 }
3551 
3552 #undef __FUNCT__
3553 #define __FUNCT__ "MatConvert"
3554 /*@C
3555    MatConvert - Converts a matrix to another matrix, either of the same
3556    or different type.
3557 
3558    Collective on Mat
3559 
3560    Input Parameters:
3561 +  mat - the matrix
3562 .  newtype - new matrix type.  Use MATSAME to create a new matrix of the
3563    same type as the original matrix.
3564 -  reuse - denotes if the destination matrix is to be created or reused.  Currently
3565    MAT_REUSE_MATRIX is only supported for inplace conversion, otherwise use
3566    MAT_INITIAL_MATRIX.
3567 
3568    Output Parameter:
3569 .  M - pointer to place new matrix
3570 
3571    Notes:
3572    MatConvert() first creates a new matrix and then copies the data from
3573    the first matrix.  A related routine is MatCopy(), which copies the matrix
3574    entries of one matrix to another already existing matrix context.
3575 
3576    Cannot be used to convert a sequential matrix to parallel or parallel to sequential,
3577    the MPI communicator of the generated matrix is always the same as the communicator
3578    of the input matrix.
3579 
3580    Level: intermediate
3581 
3582    Concepts: matrices^converting between storage formats
3583 
3584 .seealso: MatCopy(), MatDuplicate()
3585 @*/
3586 PetscErrorCode PETSCMAT_DLLEXPORT MatConvert(Mat mat, const MatType newtype,MatReuse reuse,Mat *M)
3587 {
3588   PetscErrorCode         ierr;
3589   PetscTruth             sametype,issame,flg;
3590   char                   convname[256],mtype[256];
3591   Mat                    B;
3592 
3593   PetscFunctionBegin;
3594   PetscValidHeaderSpecific(mat,MAT_CLASSID,1);
3595   PetscValidType(mat,1);
3596   PetscValidPointer(M,3);
3597   if (!mat->assembled) SETERRQ(((PetscObject)mat)->comm,PETSC_ERR_ARG_WRONGSTATE,"Not for unassembled matrix");
3598   if (mat->factortype) SETERRQ(((PetscObject)mat)->comm,PETSC_ERR_ARG_WRONGSTATE,"Not for factored matrix");
3599   ierr = MatPreallocated(mat);CHKERRQ(ierr);
3600 
3601   ierr = PetscOptionsGetString(((PetscObject)mat)->prefix,"-matconvert_type",mtype,256,&flg);CHKERRQ(ierr);
3602   if (flg) {
3603     newtype = mtype;
3604   }
3605   ierr = PetscTypeCompare((PetscObject)mat,newtype,&sametype);CHKERRQ(ierr);
3606   ierr = PetscStrcmp(newtype,"same",&issame);CHKERRQ(ierr);
3607   if ((reuse == MAT_REUSE_MATRIX) && (mat != *M)) {
3608     SETERRQ(((PetscObject)mat)->comm,PETSC_ERR_SUP,"MAT_REUSE_MATRIX only supported for in-place conversion currently");
3609   }
3610 
3611   if ((reuse == MAT_REUSE_MATRIX) && (issame || sametype)) PetscFunctionReturn(0);
3612 
3613   if ((sametype || issame) && (reuse==MAT_INITIAL_MATRIX) && mat->ops->duplicate) {
3614     ierr = (*mat->ops->duplicate)(mat,MAT_COPY_VALUES,M);CHKERRQ(ierr);
3615   } else {
3616     PetscErrorCode (*conv)(Mat, const MatType,MatReuse,Mat*)=PETSC_NULL;
3617     const char     *prefix[3] = {"seq","mpi",""};
3618     PetscInt       i;
3619     /*
3620        Order of precedence:
3621        1) See if a specialized converter is known to the current matrix.
3622        2) See if a specialized converter is known to the desired matrix class.
3623        3) See if a good general converter is registered for the desired class
3624           (as of 6/27/03 only MATMPIADJ falls into this category).
3625        4) See if a good general converter is known for the current matrix.
3626        5) Use a really basic converter.
3627     */
3628 
3629     /* 1) See if a specialized converter is known to the current matrix and the desired class */
3630     for (i=0; i<3; i++) {
3631       ierr = PetscStrcpy(convname,"MatConvert_");CHKERRQ(ierr);
3632       ierr = PetscStrcat(convname,((PetscObject)mat)->type_name);CHKERRQ(ierr);
3633       ierr = PetscStrcat(convname,"_");CHKERRQ(ierr);
3634       ierr = PetscStrcat(convname,prefix[i]);CHKERRQ(ierr);
3635       ierr = PetscStrcat(convname,newtype);CHKERRQ(ierr);
3636       ierr = PetscStrcat(convname,"_C");CHKERRQ(ierr);
3637       ierr = PetscObjectQueryFunction((PetscObject)mat,convname,(void (**)(void))&conv);CHKERRQ(ierr);
3638       if (conv) goto foundconv;
3639     }
3640 
3641     /* 2)  See if a specialized converter is known to the desired matrix class. */
3642     ierr = MatCreate(((PetscObject)mat)->comm,&B);CHKERRQ(ierr);
3643     ierr = MatSetSizes(B,mat->rmap->n,mat->cmap->n,mat->rmap->N,mat->cmap->N);CHKERRQ(ierr);
3644     ierr = MatSetType(B,newtype);CHKERRQ(ierr);
3645     for (i=0; i<3; i++) {
3646       ierr = PetscStrcpy(convname,"MatConvert_");CHKERRQ(ierr);
3647       ierr = PetscStrcat(convname,((PetscObject)mat)->type_name);CHKERRQ(ierr);
3648       ierr = PetscStrcat(convname,"_");CHKERRQ(ierr);
3649       ierr = PetscStrcat(convname,prefix[i]);CHKERRQ(ierr);
3650       ierr = PetscStrcat(convname,newtype);CHKERRQ(ierr);
3651       ierr = PetscStrcat(convname,"_C");CHKERRQ(ierr);
3652       ierr = PetscObjectQueryFunction((PetscObject)B,convname,(void (**)(void))&conv);CHKERRQ(ierr);
3653       if (conv) {
3654         ierr = MatDestroy(B);CHKERRQ(ierr);
3655         goto foundconv;
3656       }
3657     }
3658 
3659     /* 3) See if a good general converter is registered for the desired class */
3660     conv = B->ops->convertfrom;
3661     ierr = MatDestroy(B);CHKERRQ(ierr);
3662     if (conv) goto foundconv;
3663 
3664     /* 4) See if a good general converter is known for the current matrix */
3665     if (mat->ops->convert) {
3666       conv = mat->ops->convert;
3667     }
3668     if (conv) goto foundconv;
3669 
3670     /* 5) Use a really basic converter. */
3671     conv = MatConvert_Basic;
3672 
3673     foundconv:
3674     ierr = PetscLogEventBegin(MAT_Convert,mat,0,0,0);CHKERRQ(ierr);
3675     ierr = (*conv)(mat,newtype,reuse,M);CHKERRQ(ierr);
3676     ierr = PetscLogEventEnd(MAT_Convert,mat,0,0,0);CHKERRQ(ierr);
3677   }
3678   ierr = PetscObjectStateIncrease((PetscObject)*M);CHKERRQ(ierr);
3679   PetscFunctionReturn(0);
3680 }
3681 
3682 #undef __FUNCT__
3683 #define __FUNCT__ "MatFactorGetSolverPackage"
3684 /*@C
3685    MatFactorGetSolverPackage - Returns name of the package providing the factorization routines
3686 
3687    Not Collective
3688 
3689    Input Parameter:
3690 .  mat - the matrix, must be a factored matrix
3691 
3692    Output Parameter:
3693 .   type - the string name of the package (do not free this string)
3694 
3695    Notes:
3696       In Fortran you pass in a empty string and the package name will be copied into it.
3697     (Make sure the string is long enough)
3698 
3699    Level: intermediate
3700 
3701 .seealso: MatCopy(), MatDuplicate(), MatGetFactorAvailable(), MatGetFactor()
3702 @*/
3703 PetscErrorCode PETSCMAT_DLLEXPORT MatFactorGetSolverPackage(Mat mat, const MatSolverPackage *type)
3704 {
3705   PetscErrorCode         ierr;
3706   PetscErrorCode         (*conv)(Mat,const MatSolverPackage*);
3707 
3708   PetscFunctionBegin;
3709   PetscValidHeaderSpecific(mat,MAT_CLASSID,1);
3710   PetscValidType(mat,1);
3711   if (!mat->factortype) SETERRQ(((PetscObject)mat)->comm,PETSC_ERR_ARG_WRONGSTATE,"Only for factored matrix");
3712   ierr = PetscObjectQueryFunction((PetscObject)mat,"MatFactorGetSolverPackage_C",(void (**)(void))&conv);CHKERRQ(ierr);
3713   if (!conv) {
3714     *type = MATSOLVERPETSC;
3715   } else {
3716     ierr = (*conv)(mat,type);CHKERRQ(ierr);
3717   }
3718   PetscFunctionReturn(0);
3719 }
3720 
3721 #undef __FUNCT__
3722 #define __FUNCT__ "MatGetFactor"
3723 /*@C
3724    MatGetFactor - Returns a matrix suitable to calls to MatXXFactorSymbolic()
3725 
3726    Collective on Mat
3727 
3728    Input Parameters:
3729 +  mat - the matrix
3730 .  type - name of solver type, for example, spooles, superlu, plapack, petsc (to use PETSc's default)
3731 -  ftype - factor type, MAT_FACTOR_LU, MAT_FACTOR_CHOLESKY, MAT_FACTOR_ICC, MAT_FACTOR_ILU,
3732 
3733    Output Parameters:
3734 .  f - the factor matrix used with MatXXFactorSymbolic() calls
3735 
3736    Notes:
3737       Some PETSc matrix formats have alternative solvers available that are contained in alternative packages
3738      such as pastix, superlu, mumps, spooles etc.
3739 
3740       PETSc must have been ./configure to use the external solver, using the option --download-package
3741 
3742    Level: intermediate
3743 
3744 .seealso: MatCopy(), MatDuplicate(), MatGetFactorAvailable()
3745 @*/
3746 PetscErrorCode PETSCMAT_DLLEXPORT MatGetFactor(Mat mat, const MatSolverPackage type,MatFactorType ftype,Mat *f)
3747 {
3748   PetscErrorCode  ierr,(*conv)(Mat,MatFactorType,Mat*);
3749   char            convname[256];
3750 
3751   PetscFunctionBegin;
3752   PetscValidHeaderSpecific(mat,MAT_CLASSID,1);
3753   PetscValidType(mat,1);
3754 
3755   if (mat->factortype) SETERRQ(((PetscObject)mat)->comm,PETSC_ERR_ARG_WRONGSTATE,"Not for factored matrix");
3756   ierr = MatPreallocated(mat);CHKERRQ(ierr);
3757 
3758   ierr = PetscStrcpy(convname,"MatGetFactor_");CHKERRQ(ierr);
3759   ierr = PetscStrcat(convname,type);CHKERRQ(ierr);
3760   ierr = PetscStrcat(convname,"_C");CHKERRQ(ierr);
3761   ierr = PetscObjectQueryFunction((PetscObject)mat,convname,(void (**)(void))&conv);CHKERRQ(ierr);
3762   if (!conv) {
3763     PetscTruth flag;
3764     MPI_Comm   comm;
3765 
3766     ierr = PetscObjectGetComm((PetscObject)mat,&comm);CHKERRQ(ierr);
3767     ierr = PetscStrcasecmp(MATSOLVERPETSC,type,&flag);CHKERRQ(ierr);
3768     if (flag) {
3769       SETERRQ2(comm,PETSC_ERR_SUP,"Matrix format %s does not have a built-in PETSc %s",((PetscObject)mat)->type_name,MatFactorTypes[ftype]);
3770     } else {
3771       SETERRQ4(comm,PETSC_ERR_SUP,"Matrix format %s does not have a solver package %s for %s. Perhaps you must ./configure with --download-%s",((PetscObject)mat)->type_name,type,MatFactorTypes[ftype],type);
3772     }
3773   }
3774   ierr = (*conv)(mat,ftype,f);CHKERRQ(ierr);
3775   PetscFunctionReturn(0);
3776 }
3777 
3778 #undef __FUNCT__
3779 #define __FUNCT__ "MatGetFactorAvailable"
3780 /*@C
3781    MatGetFactorAvailable - Returns a a flag if matrix supports particular package and factor type
3782 
3783    Collective on Mat
3784 
3785    Input Parameters:
3786 +  mat - the matrix
3787 .  type - name of solver type, for example, spooles, superlu, plapack, petsc (to use PETSc's default)
3788 -  ftype - factor type, MAT_FACTOR_LU, MAT_FACTOR_CHOLESKY, MAT_FACTOR_ICC, MAT_FACTOR_ILU,
3789 
3790    Output Parameter:
3791 .    flg - PETSC_TRUE if the factorization is available
3792 
3793    Notes:
3794       Some PETSc matrix formats have alternative solvers available that are contained in alternative packages
3795      such as pastix, superlu, mumps, spooles etc.
3796 
3797       PETSc must have been ./configure to use the external solver, using the option --download-package
3798 
3799    Level: intermediate
3800 
3801 .seealso: MatCopy(), MatDuplicate(), MatGetFactor()
3802 @*/
3803 PetscErrorCode PETSCMAT_DLLEXPORT MatGetFactorAvailable(Mat mat, const MatSolverPackage type,MatFactorType ftype,PetscTruth *flg)
3804 {
3805   PetscErrorCode         ierr;
3806   char                   convname[256];
3807   PetscErrorCode         (*conv)(Mat,MatFactorType,PetscTruth*);
3808 
3809   PetscFunctionBegin;
3810   PetscValidHeaderSpecific(mat,MAT_CLASSID,1);
3811   PetscValidType(mat,1);
3812 
3813   if (mat->factortype) SETERRQ(((PetscObject)mat)->comm,PETSC_ERR_ARG_WRONGSTATE,"Not for factored matrix");
3814   ierr = MatPreallocated(mat);CHKERRQ(ierr);
3815 
3816   ierr = PetscStrcpy(convname,"MatGetFactorAvailable_");CHKERRQ(ierr);
3817   ierr = PetscStrcat(convname,type);CHKERRQ(ierr);
3818   ierr = PetscStrcat(convname,"_C");CHKERRQ(ierr);
3819   ierr = PetscObjectQueryFunction((PetscObject)mat,convname,(void (**)(void))&conv);CHKERRQ(ierr);
3820   if (!conv) {
3821     *flg = PETSC_FALSE;
3822   } else {
3823     ierr = (*conv)(mat,ftype,flg);CHKERRQ(ierr);
3824   }
3825   PetscFunctionReturn(0);
3826 }
3827 
3828 
3829 #undef __FUNCT__
3830 #define __FUNCT__ "MatDuplicate"
3831 /*@
3832    MatDuplicate - Duplicates a matrix including the non-zero structure.
3833 
3834    Collective on Mat
3835 
3836    Input Parameters:
3837 +  mat - the matrix
3838 -  op - either MAT_DO_NOT_COPY_VALUES or MAT_COPY_VALUES, cause it to copy the numerical values in the matrix
3839         MAT_SHARE_NONZERO_PATTERN to share the nonzero patterns with the previous matrix and not copy them.
3840 
3841    Output Parameter:
3842 .  M - pointer to place new matrix
3843 
3844    Level: intermediate
3845 
3846    Concepts: matrices^duplicating
3847 
3848     Notes: You cannot change the nonzero pattern for the parent or child matrix if you use MAT_SHARE_NONZERO_PATTERN.
3849 
3850 .seealso: MatCopy(), MatConvert()
3851 @*/
3852 PetscErrorCode PETSCMAT_DLLEXPORT MatDuplicate(Mat mat,MatDuplicateOption op,Mat *M)
3853 {
3854   PetscErrorCode ierr;
3855   Mat            B;
3856   PetscInt       i;
3857 
3858   PetscFunctionBegin;
3859   PetscValidHeaderSpecific(mat,MAT_CLASSID,1);
3860   PetscValidType(mat,1);
3861   PetscValidPointer(M,3);
3862   if (!mat->assembled) SETERRQ(((PetscObject)mat)->comm,PETSC_ERR_ARG_WRONGSTATE,"Not for unassembled matrix");
3863   if (mat->factortype) SETERRQ(((PetscObject)mat)->comm,PETSC_ERR_ARG_WRONGSTATE,"Not for factored matrix");
3864   ierr = MatPreallocated(mat);CHKERRQ(ierr);
3865 
3866   *M  = 0;
3867   if (!mat->ops->duplicate) SETERRQ(((PetscObject)mat)->comm,PETSC_ERR_SUP,"Not written for this matrix type");
3868   ierr = PetscLogEventBegin(MAT_Convert,mat,0,0,0);CHKERRQ(ierr);
3869   ierr = (*mat->ops->duplicate)(mat,op,M);CHKERRQ(ierr);
3870   B = *M;
3871   if (mat->mapping) {
3872     ierr = MatSetLocalToGlobalMapping(B,mat->mapping);CHKERRQ(ierr);
3873   }
3874   if (mat->bmapping) {
3875     ierr = MatSetLocalToGlobalMappingBlock(B,mat->bmapping);CHKERRQ(ierr);
3876   }
3877   ierr = PetscLayoutCopy(mat->rmap,&B->rmap);CHKERRQ(ierr);
3878   ierr = PetscLayoutCopy(mat->cmap,&B->cmap);CHKERRQ(ierr);
3879 
3880   B->stencil.dim = mat->stencil.dim;
3881   B->stencil.noc = mat->stencil.noc;
3882   for (i=0; i<=mat->stencil.dim; i++) {
3883     B->stencil.dims[i]   = mat->stencil.dims[i];
3884     B->stencil.starts[i] = mat->stencil.starts[i];
3885   }
3886 
3887   ierr = PetscLogEventEnd(MAT_Convert,mat,0,0,0);CHKERRQ(ierr);
3888   ierr = PetscObjectStateIncrease((PetscObject)B);CHKERRQ(ierr);
3889   PetscFunctionReturn(0);
3890 }
3891 
3892 #undef __FUNCT__
3893 #define __FUNCT__ "MatGetDiagonal"
3894 /*@
3895    MatGetDiagonal - Gets the diagonal of a matrix.
3896 
3897    Collective on Mat and Vec
3898 
3899    Input Parameters:
3900 +  mat - the matrix
3901 -  v - the vector for storing the diagonal
3902 
3903    Output Parameter:
3904 .  v - the diagonal of the matrix
3905 
3906    Level: intermediate
3907 
3908    Concepts: matrices^accessing diagonals
3909 
3910 .seealso: MatGetRow(), MatGetSubMatrices(), MatGetSubmatrix(), MatGetRowMaxAbs()
3911 @*/
3912 PetscErrorCode PETSCMAT_DLLEXPORT MatGetDiagonal(Mat mat,Vec v)
3913 {
3914   PetscErrorCode ierr;
3915 
3916   PetscFunctionBegin;
3917   PetscValidHeaderSpecific(mat,MAT_CLASSID,1);
3918   PetscValidType(mat,1);
3919   PetscValidHeaderSpecific(v,VEC_CLASSID,2);
3920   if (!mat->assembled) SETERRQ(((PetscObject)mat)->comm,PETSC_ERR_ARG_WRONGSTATE,"Not for unassembled matrix");
3921   if (!mat->ops->getdiagonal) SETERRQ1(((PetscObject)mat)->comm,PETSC_ERR_SUP,"Mat type %s",((PetscObject)mat)->type_name);
3922   ierr = MatPreallocated(mat);CHKERRQ(ierr);
3923 
3924   ierr = (*mat->ops->getdiagonal)(mat,v);CHKERRQ(ierr);
3925   ierr = PetscObjectStateIncrease((PetscObject)v);CHKERRQ(ierr);
3926   PetscFunctionReturn(0);
3927 }
3928 
3929 #undef __FUNCT__
3930 #define __FUNCT__ "MatGetRowMin"
3931 /*@
3932    MatGetRowMin - Gets the minimum value (of the real part) of each
3933         row of the matrix
3934 
3935    Collective on Mat and Vec
3936 
3937    Input Parameters:
3938 .  mat - the matrix
3939 
3940    Output Parameter:
3941 +  v - the vector for storing the maximums
3942 -  idx - the indices of the column found for each row (optional)
3943 
3944    Level: intermediate
3945 
3946    Notes: The result of this call are the same as if one converted the matrix to dense format
3947       and found the minimum value in each row (i.e. the implicit zeros are counted as zeros).
3948 
3949     This code is only implemented for a couple of matrix formats.
3950 
3951    Concepts: matrices^getting row maximums
3952 
3953 .seealso: MatGetDiagonal(), MatGetSubMatrices(), MatGetSubmatrix(), MatGetRowMaxAbs(),
3954           MatGetRowMax()
3955 @*/
3956 PetscErrorCode PETSCMAT_DLLEXPORT MatGetRowMin(Mat mat,Vec v,PetscInt idx[])
3957 {
3958   PetscErrorCode ierr;
3959 
3960   PetscFunctionBegin;
3961   PetscValidHeaderSpecific(mat,MAT_CLASSID,1);
3962   PetscValidType(mat,1);
3963   PetscValidHeaderSpecific(v,VEC_CLASSID,2);
3964   if (!mat->assembled) SETERRQ(((PetscObject)mat)->comm,PETSC_ERR_ARG_WRONGSTATE,"Not for unassembled matrix");
3965   if (!mat->ops->getrowmax) SETERRQ1(PETSC_COMM_SELF,PETSC_ERR_SUP,"Mat type %s",((PetscObject)mat)->type_name);
3966   ierr = MatPreallocated(mat);CHKERRQ(ierr);
3967 
3968   ierr = (*mat->ops->getrowmin)(mat,v,idx);CHKERRQ(ierr);
3969   ierr = PetscObjectStateIncrease((PetscObject)v);CHKERRQ(ierr);
3970   PetscFunctionReturn(0);
3971 }
3972 
3973 #undef __FUNCT__
3974 #define __FUNCT__ "MatGetRowMinAbs"
3975 /*@
3976    MatGetRowMinAbs - Gets the minimum value (in absolute value) of each
3977         row of the matrix
3978 
3979    Collective on Mat and Vec
3980 
3981    Input Parameters:
3982 .  mat - the matrix
3983 
3984    Output Parameter:
3985 +  v - the vector for storing the minimums
3986 -  idx - the indices of the column found for each row (optional)
3987 
3988    Level: intermediate
3989 
3990    Notes: if a row is completely empty or has only 0.0 values then the idx[] value for that
3991     row is 0 (the first column).
3992 
3993     This code is only implemented for a couple of matrix formats.
3994 
3995    Concepts: matrices^getting row maximums
3996 
3997 .seealso: MatGetDiagonal(), MatGetSubMatrices(), MatGetSubmatrix(), MatGetRowMax(), MatGetRowMaxAbs(), MatGetRowMin()
3998 @*/
3999 PetscErrorCode PETSCMAT_DLLEXPORT MatGetRowMinAbs(Mat mat,Vec v,PetscInt idx[])
4000 {
4001   PetscErrorCode ierr;
4002 
4003   PetscFunctionBegin;
4004   PetscValidHeaderSpecific(mat,MAT_CLASSID,1);
4005   PetscValidType(mat,1);
4006   PetscValidHeaderSpecific(v,VEC_CLASSID,2);
4007   if (!mat->assembled) SETERRQ(((PetscObject)mat)->comm,PETSC_ERR_ARG_WRONGSTATE,"Not for unassembled matrix");
4008   if (!mat->ops->getrowminabs) SETERRQ1(((PetscObject)mat)->comm,PETSC_ERR_SUP,"Mat type %s",((PetscObject)mat)->type_name);
4009   ierr = MatPreallocated(mat);CHKERRQ(ierr);
4010   if (idx) {ierr = PetscMemzero(idx,mat->rmap->n*sizeof(PetscInt));CHKERRQ(ierr);}
4011 
4012   ierr = (*mat->ops->getrowminabs)(mat,v,idx);CHKERRQ(ierr);
4013   ierr = PetscObjectStateIncrease((PetscObject)v);CHKERRQ(ierr);
4014   PetscFunctionReturn(0);
4015 }
4016 
4017 #undef __FUNCT__
4018 #define __FUNCT__ "MatGetRowMax"
4019 /*@
4020    MatGetRowMax - Gets the maximum value (of the real part) of each
4021         row of the matrix
4022 
4023    Collective on Mat and Vec
4024 
4025    Input Parameters:
4026 .  mat - the matrix
4027 
4028    Output Parameter:
4029 +  v - the vector for storing the maximums
4030 -  idx - the indices of the column found for each row (optional)
4031 
4032    Level: intermediate
4033 
4034    Notes: The result of this call are the same as if one converted the matrix to dense format
4035       and found the minimum value in each row (i.e. the implicit zeros are counted as zeros).
4036 
4037     This code is only implemented for a couple of matrix formats.
4038 
4039    Concepts: matrices^getting row maximums
4040 
4041 .seealso: MatGetDiagonal(), MatGetSubMatrices(), MatGetSubmatrix(), MatGetRowMaxAbs(), MatGetRowMin()
4042 @*/
4043 PetscErrorCode PETSCMAT_DLLEXPORT MatGetRowMax(Mat mat,Vec v,PetscInt idx[])
4044 {
4045   PetscErrorCode ierr;
4046 
4047   PetscFunctionBegin;
4048   PetscValidHeaderSpecific(mat,MAT_CLASSID,1);
4049   PetscValidType(mat,1);
4050   PetscValidHeaderSpecific(v,VEC_CLASSID,2);
4051   if (!mat->assembled) SETERRQ(((PetscObject)mat)->comm,PETSC_ERR_ARG_WRONGSTATE,"Not for unassembled matrix");
4052   if (!mat->ops->getrowmax) SETERRQ1(((PetscObject)mat)->comm,PETSC_ERR_SUP,"Mat type %s",((PetscObject)mat)->type_name);
4053   ierr = MatPreallocated(mat);CHKERRQ(ierr);
4054 
4055   ierr = (*mat->ops->getrowmax)(mat,v,idx);CHKERRQ(ierr);
4056   ierr = PetscObjectStateIncrease((PetscObject)v);CHKERRQ(ierr);
4057   PetscFunctionReturn(0);
4058 }
4059 
4060 #undef __FUNCT__
4061 #define __FUNCT__ "MatGetRowMaxAbs"
4062 /*@
4063    MatGetRowMaxAbs - Gets the maximum value (in absolute value) of each
4064         row of the matrix
4065 
4066    Collective on Mat and Vec
4067 
4068    Input Parameters:
4069 .  mat - the matrix
4070 
4071    Output Parameter:
4072 +  v - the vector for storing the maximums
4073 -  idx - the indices of the column found for each row (optional)
4074 
4075    Level: intermediate
4076 
4077    Notes: if a row is completely empty or has only 0.0 values then the idx[] value for that
4078     row is 0 (the first column).
4079 
4080     This code is only implemented for a couple of matrix formats.
4081 
4082    Concepts: matrices^getting row maximums
4083 
4084 .seealso: MatGetDiagonal(), MatGetSubMatrices(), MatGetSubmatrix(), MatGetRowMax(), MatGetRowMin()
4085 @*/
4086 PetscErrorCode PETSCMAT_DLLEXPORT MatGetRowMaxAbs(Mat mat,Vec v,PetscInt idx[])
4087 {
4088   PetscErrorCode ierr;
4089 
4090   PetscFunctionBegin;
4091   PetscValidHeaderSpecific(mat,MAT_CLASSID,1);
4092   PetscValidType(mat,1);
4093   PetscValidHeaderSpecific(v,VEC_CLASSID,2);
4094   if (!mat->assembled) SETERRQ(((PetscObject)mat)->comm,PETSC_ERR_ARG_WRONGSTATE,"Not for unassembled matrix");
4095   if (!mat->ops->getrowmaxabs) SETERRQ1(PETSC_COMM_SELF,PETSC_ERR_SUP,"Mat type %s",((PetscObject)mat)->type_name);
4096   ierr = MatPreallocated(mat);CHKERRQ(ierr);
4097   if (idx) {ierr = PetscMemzero(idx,mat->rmap->n*sizeof(PetscInt));CHKERRQ(ierr);}
4098 
4099   ierr = (*mat->ops->getrowmaxabs)(mat,v,idx);CHKERRQ(ierr);
4100   ierr = PetscObjectStateIncrease((PetscObject)v);CHKERRQ(ierr);
4101   PetscFunctionReturn(0);
4102 }
4103 
4104 #undef __FUNCT__
4105 #define __FUNCT__ "MatGetRowSum"
4106 /*@
4107    MatGetRowSum - Gets the sum of each row of the matrix
4108 
4109    Collective on Mat and Vec
4110 
4111    Input Parameters:
4112 .  mat - the matrix
4113 
4114    Output Parameter:
4115 .  v - the vector for storing the sum of rows
4116 
4117    Level: intermediate
4118 
4119    Notes: This code is slow since it is not currently specialized for different formats
4120 
4121    Concepts: matrices^getting row sums
4122 
4123 .seealso: MatGetDiagonal(), MatGetSubMatrices(), MatGetSubmatrix(), MatGetRowMax(), MatGetRowMin()
4124 @*/
4125 PetscErrorCode PETSCMAT_DLLEXPORT MatGetRowSum(Mat mat, Vec v)
4126 {
4127   PetscInt       start = 0, end = 0, row;
4128   PetscScalar   *array;
4129   PetscErrorCode ierr;
4130 
4131   PetscFunctionBegin;
4132   PetscValidHeaderSpecific(mat,MAT_CLASSID,1);
4133   PetscValidType(mat,1);
4134   PetscValidHeaderSpecific(v,VEC_CLASSID,2);
4135   if (!mat->assembled) SETERRQ(((PetscObject)mat)->comm,PETSC_ERR_ARG_WRONGSTATE,"Not for unassembled matrix");
4136   ierr = MatPreallocated(mat);CHKERRQ(ierr);
4137   ierr = MatGetOwnershipRange(mat, &start, &end);CHKERRQ(ierr);
4138   ierr = VecGetArray(v, &array);CHKERRQ(ierr);
4139   for(row = start; row < end; ++row) {
4140     PetscInt           ncols, col;
4141     const PetscInt    *cols;
4142     const PetscScalar *vals;
4143 
4144     array[row - start] = 0.0;
4145     ierr = MatGetRow(mat, row, &ncols, &cols, &vals);CHKERRQ(ierr);
4146     for(col = 0; col < ncols; col++) {
4147       array[row - start] += vals[col];
4148     }
4149     ierr = MatRestoreRow(mat, row, &ncols, &cols, &vals);CHKERRQ(ierr);
4150   }
4151   ierr = VecRestoreArray(v, &array);CHKERRQ(ierr);
4152   ierr = PetscObjectStateIncrease((PetscObject) v);CHKERRQ(ierr);
4153   PetscFunctionReturn(0);
4154 }
4155 
4156 #undef __FUNCT__
4157 #define __FUNCT__ "MatTranspose"
4158 /*@
4159    MatTranspose - Computes an in-place or out-of-place transpose of a matrix.
4160 
4161    Collective on Mat
4162 
4163    Input Parameter:
4164 +  mat - the matrix to transpose
4165 -  reuse - store the transpose matrix in the provided B
4166 
4167    Output Parameters:
4168 .  B - the transpose
4169 
4170    Notes:
4171      If you  pass in &mat for B the transpose will be done in place
4172 
4173    Level: intermediate
4174 
4175    Concepts: matrices^transposing
4176 
4177 .seealso: MatMultTranspose(), MatMultTransposeAdd(), MatIsTranspose(), MatReuse
4178 @*/
4179 PetscErrorCode PETSCMAT_DLLEXPORT MatTranspose(Mat mat,MatReuse reuse,Mat *B)
4180 {
4181   PetscErrorCode ierr;
4182 
4183   PetscFunctionBegin;
4184   PetscValidHeaderSpecific(mat,MAT_CLASSID,1);
4185   PetscValidType(mat,1);
4186   if (!mat->assembled) SETERRQ(((PetscObject)mat)->comm,PETSC_ERR_ARG_WRONGSTATE,"Not for unassembled matrix");
4187   if (mat->factortype) SETERRQ(((PetscObject)mat)->comm,PETSC_ERR_ARG_WRONGSTATE,"Not for factored matrix");
4188   if (!mat->ops->transpose) SETERRQ1(((PetscObject)mat)->comm,PETSC_ERR_SUP,"Mat type %s",((PetscObject)mat)->type_name);
4189   ierr = MatPreallocated(mat);CHKERRQ(ierr);
4190 
4191   ierr = PetscLogEventBegin(MAT_Transpose,mat,0,0,0);CHKERRQ(ierr);
4192   ierr = (*mat->ops->transpose)(mat,reuse,B);CHKERRQ(ierr);
4193   ierr = PetscLogEventEnd(MAT_Transpose,mat,0,0,0);CHKERRQ(ierr);
4194   if (B) {ierr = PetscObjectStateIncrease((PetscObject)*B);CHKERRQ(ierr);}
4195   PetscFunctionReturn(0);
4196 }
4197 
4198 #undef __FUNCT__
4199 #define __FUNCT__ "MatIsTranspose"
4200 /*@
4201    MatIsTranspose - Test whether a matrix is another one's transpose,
4202         or its own, in which case it tests symmetry.
4203 
4204    Collective on Mat
4205 
4206    Input Parameter:
4207 +  A - the matrix to test
4208 -  B - the matrix to test against, this can equal the first parameter
4209 
4210    Output Parameters:
4211 .  flg - the result
4212 
4213    Notes:
4214    Only available for SeqAIJ/MPIAIJ matrices. The sequential algorithm
4215    has a running time of the order of the number of nonzeros; the parallel
4216    test involves parallel copies of the block-offdiagonal parts of the matrix.
4217 
4218    Level: intermediate
4219 
4220    Concepts: matrices^transposing, matrix^symmetry
4221 
4222 .seealso: MatTranspose(), MatIsSymmetric(), MatIsHermitian()
4223 @*/
4224 PetscErrorCode PETSCMAT_DLLEXPORT MatIsTranspose(Mat A,Mat B,PetscReal tol,PetscTruth *flg)
4225 {
4226   PetscErrorCode ierr,(*f)(Mat,Mat,PetscReal,PetscTruth*),(*g)(Mat,Mat,PetscReal,PetscTruth*);
4227 
4228   PetscFunctionBegin;
4229   PetscValidHeaderSpecific(A,MAT_CLASSID,1);
4230   PetscValidHeaderSpecific(B,MAT_CLASSID,2);
4231   PetscValidPointer(flg,3);
4232   ierr = PetscObjectQueryFunction((PetscObject)A,"MatIsTranspose_C",(void (**)(void))&f);CHKERRQ(ierr);
4233   ierr = PetscObjectQueryFunction((PetscObject)B,"MatIsTranspose_C",(void (**)(void))&g);CHKERRQ(ierr);
4234   if (f && g) {
4235     if (f==g) {
4236       ierr = (*f)(A,B,tol,flg);CHKERRQ(ierr);
4237     } else {
4238       SETERRQ(((PetscObject)A)->comm,PETSC_ERR_ARG_NOTSAMETYPE,"Matrices do not have the same comparator for symmetry test");
4239     }
4240   }
4241   PetscFunctionReturn(0);
4242 }
4243 
4244 #undef __FUNCT__
4245 #define __FUNCT__ "MatHermitianTranspose"
4246 /*@
4247    MatHermitianTranspose - Computes an in-place or out-of-place transpose of a matrix in complex conjugate.
4248 
4249    Collective on Mat
4250 
4251    Input Parameter:
4252 +  mat - the matrix to transpose and complex conjugate
4253 -  reuse - store the transpose matrix in the provided B
4254 
4255    Output Parameters:
4256 .  B - the Hermitian
4257 
4258    Notes:
4259      If you  pass in &mat for B the Hermitian will be done in place
4260 
4261    Level: intermediate
4262 
4263    Concepts: matrices^transposing, complex conjugatex
4264 
4265 .seealso: MatTranspose(), MatMultTranspose(), MatMultTransposeAdd(), MatIsTranspose(), MatReuse
4266 @*/
4267 PetscErrorCode PETSCMAT_DLLEXPORT MatHermitianTranspose(Mat mat,MatReuse reuse,Mat *B)
4268 {
4269   PetscErrorCode ierr;
4270 
4271   PetscFunctionBegin;
4272   ierr = MatTranspose(mat,reuse,B);CHKERRQ(ierr);
4273 #if defined(PETSC_USE_COMPLEX)
4274   ierr = MatConjugate(*B);CHKERRQ(ierr);
4275 #endif
4276   PetscFunctionReturn(0);
4277 }
4278 
4279 #undef __FUNCT__
4280 #define __FUNCT__ "MatIsHermitianTranspose"
4281 /*@
4282    MatIsHermitianTranspose - Test whether a matrix is another one's Hermitian transpose,
4283 
4284    Collective on Mat
4285 
4286    Input Parameter:
4287 +  A - the matrix to test
4288 -  B - the matrix to test against, this can equal the first parameter
4289 
4290    Output Parameters:
4291 .  flg - the result
4292 
4293    Notes:
4294    Only available for SeqAIJ/MPIAIJ matrices. The sequential algorithm
4295    has a running time of the order of the number of nonzeros; the parallel
4296    test involves parallel copies of the block-offdiagonal parts of the matrix.
4297 
4298    Level: intermediate
4299 
4300    Concepts: matrices^transposing, matrix^symmetry
4301 
4302 .seealso: MatTranspose(), MatIsSymmetric(), MatIsHermitian(), MatIsTranspose()
4303 @*/
4304 PetscErrorCode PETSCMAT_DLLEXPORT MatIsHermitianTranspose(Mat A,Mat B,PetscReal tol,PetscTruth *flg)
4305 {
4306   PetscErrorCode ierr,(*f)(Mat,Mat,PetscReal,PetscTruth*),(*g)(Mat,Mat,PetscReal,PetscTruth*);
4307 
4308   PetscFunctionBegin;
4309   PetscValidHeaderSpecific(A,MAT_CLASSID,1);
4310   PetscValidHeaderSpecific(B,MAT_CLASSID,2);
4311   PetscValidPointer(flg,3);
4312   ierr = PetscObjectQueryFunction((PetscObject)A,"MatIsHermitianTranspose_C",(void (**)(void))&f);CHKERRQ(ierr);
4313   ierr = PetscObjectQueryFunction((PetscObject)B,"MatIsHermitianTranspose_C",(void (**)(void))&g);CHKERRQ(ierr);
4314   if (f && g) {
4315     if (f==g) {
4316       ierr = (*f)(A,B,tol,flg);CHKERRQ(ierr);
4317     } else {
4318       SETERRQ(((PetscObject)A)->comm,PETSC_ERR_ARG_NOTSAMETYPE,"Matrices do not have the same comparator for Hermitian test");
4319     }
4320   }
4321   PetscFunctionReturn(0);
4322 }
4323 
4324 #undef __FUNCT__
4325 #define __FUNCT__ "MatPermute"
4326 /*@
4327    MatPermute - Creates a new matrix with rows and columns permuted from the
4328    original.
4329 
4330    Collective on Mat
4331 
4332    Input Parameters:
4333 +  mat - the matrix to permute
4334 .  row - row permutation, each processor supplies only the permutation for its rows
4335 -  col - column permutation, each processor needs the entire column permutation, that is
4336          this is the same size as the total number of columns in the matrix. It can often
4337          be obtained with ISAllGather() on the row permutation
4338 
4339    Output Parameters:
4340 .  B - the permuted matrix
4341 
4342    Level: advanced
4343 
4344    Concepts: matrices^permuting
4345 
4346 .seealso: MatGetOrdering(), ISAllGather()
4347 
4348 @*/
4349 PetscErrorCode PETSCMAT_DLLEXPORT MatPermute(Mat mat,IS row,IS col,Mat *B)
4350 {
4351   PetscErrorCode ierr;
4352 
4353   PetscFunctionBegin;
4354   PetscValidHeaderSpecific(mat,MAT_CLASSID,1);
4355   PetscValidType(mat,1);
4356   PetscValidHeaderSpecific(row,IS_CLASSID,2);
4357   PetscValidHeaderSpecific(col,IS_CLASSID,3);
4358   PetscValidPointer(B,4);
4359   if (!mat->assembled) SETERRQ(((PetscObject)mat)->comm,PETSC_ERR_ARG_WRONGSTATE,"Not for unassembled matrix");
4360   if (mat->factortype) SETERRQ(((PetscObject)mat)->comm,PETSC_ERR_ARG_WRONGSTATE,"Not for factored matrix");
4361   if (!mat->ops->permute) SETERRQ1(PETSC_COMM_SELF,PETSC_ERR_SUP,"MatPermute not available for Mat type %s",((PetscObject)mat)->type_name);
4362   ierr = MatPreallocated(mat);CHKERRQ(ierr);
4363 
4364   ierr = (*mat->ops->permute)(mat,row,col,B);CHKERRQ(ierr);
4365   ierr = PetscObjectStateIncrease((PetscObject)*B);CHKERRQ(ierr);
4366   PetscFunctionReturn(0);
4367 }
4368 
4369 #undef __FUNCT__
4370 #define __FUNCT__ "MatPermuteSparsify"
4371 /*@
4372   MatPermuteSparsify - Creates a new matrix with rows and columns permuted from the
4373   original and sparsified to the prescribed tolerance.
4374 
4375   Collective on Mat
4376 
4377   Input Parameters:
4378 + A    - The matrix to permute
4379 . band - The half-bandwidth of the sparsified matrix, or PETSC_DECIDE
4380 . frac - The half-bandwidth as a fraction of the total size, or 0.0
4381 . tol  - The drop tolerance
4382 . rowp - The row permutation
4383 - colp - The column permutation
4384 
4385   Output Parameter:
4386 . B    - The permuted, sparsified matrix
4387 
4388   Level: advanced
4389 
4390   Note:
4391   The default behavior (band = PETSC_DECIDE and frac = 0.0) is to
4392   restrict the half-bandwidth of the resulting matrix to 5% of the
4393   total matrix size.
4394 
4395 .keywords: matrix, permute, sparsify
4396 
4397 .seealso: MatGetOrdering(), MatPermute()
4398 @*/
4399 PetscErrorCode PETSCMAT_DLLEXPORT MatPermuteSparsify(Mat A, PetscInt band, PetscReal frac, PetscReal tol, IS rowp, IS colp, Mat *B)
4400 {
4401   IS                irowp, icolp;
4402   const PetscInt    *rows, *cols;
4403   PetscInt          M, N, locRowStart = 0, locRowEnd = 0;
4404   PetscInt          nz, newNz;
4405   const PetscInt    *cwork;
4406   PetscInt          *cnew;
4407   const PetscScalar *vwork;
4408   PetscScalar       *vnew;
4409   PetscInt          bw, issize;
4410   PetscInt          row, locRow, newRow, col, newCol;
4411   PetscErrorCode    ierr;
4412 
4413   PetscFunctionBegin;
4414   PetscValidHeaderSpecific(A,    MAT_CLASSID,1);
4415   PetscValidHeaderSpecific(rowp, IS_CLASSID,5);
4416   PetscValidHeaderSpecific(colp, IS_CLASSID,6);
4417   PetscValidPointer(B,7);
4418   if (!A->assembled) SETERRQ(((PetscObject)A)->comm,PETSC_ERR_ARG_WRONGSTATE, "Not for unassembled matrix");
4419   if (A->factortype) SETERRQ(((PetscObject)A)->comm,PETSC_ERR_ARG_WRONGSTATE, "Not for factored matrix");
4420   if (!A->ops->permutesparsify) {
4421     ierr = MatGetSize(A, &M, &N);CHKERRQ(ierr);
4422     ierr = MatGetOwnershipRange(A, &locRowStart, &locRowEnd);CHKERRQ(ierr);
4423     ierr = ISGetSize(rowp, &issize);CHKERRQ(ierr);
4424     if (issize != M) SETERRQ2(PETSC_COMM_SELF,PETSC_ERR_ARG_WRONG, "Wrong size %D for row permutation, should be %D", issize, M);
4425     ierr = ISGetSize(colp, &issize);CHKERRQ(ierr);
4426     if (issize != N) SETERRQ2(PETSC_COMM_SELF,PETSC_ERR_ARG_WRONG, "Wrong size %D for column permutation, should be %D", issize, N);
4427     ierr = ISInvertPermutation(rowp, 0, &irowp);CHKERRQ(ierr);
4428     ierr = ISGetIndices(irowp, &rows);CHKERRQ(ierr);
4429     ierr = ISInvertPermutation(colp, 0, &icolp);CHKERRQ(ierr);
4430     ierr = ISGetIndices(icolp, &cols);CHKERRQ(ierr);
4431     ierr = PetscMalloc(N*sizeof(PetscInt),&cnew);CHKERRQ(ierr);
4432     ierr = PetscMalloc(N*sizeof(PetscScalar),&vnew);CHKERRQ(ierr);
4433 
4434     /* Setup bandwidth to include */
4435     if (band == PETSC_DECIDE) {
4436       if (frac <= 0.0)
4437         bw = (PetscInt) (M * 0.05);
4438       else
4439         bw = (PetscInt) (M * frac);
4440     } else {
4441       if (band <= 0) SETERRQ(PETSC_COMM_SELF,PETSC_ERR_ARG_WRONG, "Bandwidth must be a positive integer");
4442       bw = band;
4443     }
4444 
4445     /* Put values into new matrix */
4446     ierr = MatDuplicate(A, MAT_DO_NOT_COPY_VALUES, B);CHKERRQ(ierr);
4447     for(row = locRowStart, locRow = 0; row < locRowEnd; row++, locRow++) {
4448       ierr = MatGetRow(A, row, &nz, &cwork, &vwork);CHKERRQ(ierr);
4449       newRow   = rows[locRow]+locRowStart;
4450       for(col = 0, newNz = 0; col < nz; col++) {
4451         newCol = cols[cwork[col]];
4452         if ((newCol >= newRow - bw) && (newCol < newRow + bw) && (PetscAbsScalar(vwork[col]) >= tol)) {
4453           cnew[newNz] = newCol;
4454           vnew[newNz] = vwork[col];
4455           newNz++;
4456         }
4457       }
4458       ierr = MatSetValues(*B, 1, &newRow, newNz, cnew, vnew, INSERT_VALUES);CHKERRQ(ierr);
4459       ierr = MatRestoreRow(A, row, &nz, &cwork, &vwork);CHKERRQ(ierr);
4460     }
4461     ierr = PetscFree(cnew);CHKERRQ(ierr);
4462     ierr = PetscFree(vnew);CHKERRQ(ierr);
4463     ierr = MatAssemblyBegin(*B, MAT_FINAL_ASSEMBLY);CHKERRQ(ierr);
4464     ierr = MatAssemblyEnd(*B, MAT_FINAL_ASSEMBLY);CHKERRQ(ierr);
4465     ierr = ISRestoreIndices(irowp, &rows);CHKERRQ(ierr);
4466     ierr = ISRestoreIndices(icolp, &cols);CHKERRQ(ierr);
4467     ierr = ISDestroy(irowp);CHKERRQ(ierr);
4468     ierr = ISDestroy(icolp);CHKERRQ(ierr);
4469   } else {
4470     ierr = (*A->ops->permutesparsify)(A, band, frac, tol, rowp, colp, B);CHKERRQ(ierr);
4471   }
4472   ierr = PetscObjectStateIncrease((PetscObject)*B);CHKERRQ(ierr);
4473   PetscFunctionReturn(0);
4474 }
4475 
4476 #undef __FUNCT__
4477 #define __FUNCT__ "MatEqual"
4478 /*@
4479    MatEqual - Compares two matrices.
4480 
4481    Collective on Mat
4482 
4483    Input Parameters:
4484 +  A - the first matrix
4485 -  B - the second matrix
4486 
4487    Output Parameter:
4488 .  flg - PETSC_TRUE if the matrices are equal; PETSC_FALSE otherwise.
4489 
4490    Level: intermediate
4491 
4492    Concepts: matrices^equality between
4493 @*/
4494 PetscErrorCode PETSCMAT_DLLEXPORT MatEqual(Mat A,Mat B,PetscTruth *flg)
4495 {
4496   PetscErrorCode ierr;
4497 
4498   PetscFunctionBegin;
4499   PetscValidHeaderSpecific(A,MAT_CLASSID,1);
4500   PetscValidHeaderSpecific(B,MAT_CLASSID,2);
4501   PetscValidType(A,1);
4502   PetscValidType(B,2);
4503   PetscValidIntPointer(flg,3);
4504   PetscCheckSameComm(A,1,B,2);
4505   ierr = MatPreallocated(B);CHKERRQ(ierr);
4506   if (!A->assembled) SETERRQ(((PetscObject)A)->comm,PETSC_ERR_ARG_WRONGSTATE,"Not for unassembled matrix");
4507   if (!B->assembled) SETERRQ(((PetscObject)A)->comm,PETSC_ERR_ARG_WRONGSTATE,"Not for unassembled matrix");
4508   if (A->rmap->N != B->rmap->N || A->cmap->N != B->cmap->N) SETERRQ4(((PetscObject)A)->comm,PETSC_ERR_ARG_SIZ,"Mat A,Mat B: global dim %D %D %D %D",A->rmap->N,B->rmap->N,A->cmap->N,B->cmap->N);
4509   if (!A->ops->equal) SETERRQ1(((PetscObject)A)->comm,PETSC_ERR_SUP,"Mat type %s",((PetscObject)A)->type_name);
4510   if (!B->ops->equal) SETERRQ1(((PetscObject)A)->comm,PETSC_ERR_SUP,"Mat type %s",((PetscObject)B)->type_name);
4511   if (A->ops->equal != B->ops->equal) SETERRQ2(((PetscObject)A)->comm,PETSC_ERR_ARG_INCOMP,"A is type: %s\nB is type: %s",((PetscObject)A)->type_name,((PetscObject)B)->type_name);
4512   ierr = MatPreallocated(A);CHKERRQ(ierr);
4513 
4514   ierr = (*A->ops->equal)(A,B,flg);CHKERRQ(ierr);
4515   PetscFunctionReturn(0);
4516 }
4517 
4518 #undef __FUNCT__
4519 #define __FUNCT__ "MatDiagonalScale"
4520 /*@
4521    MatDiagonalScale - Scales a matrix on the left and right by diagonal
4522    matrices that are stored as vectors.  Either of the two scaling
4523    matrices can be PETSC_NULL.
4524 
4525    Collective on Mat
4526 
4527    Input Parameters:
4528 +  mat - the matrix to be scaled
4529 .  l - the left scaling vector (or PETSC_NULL)
4530 -  r - the right scaling vector (or PETSC_NULL)
4531 
4532    Notes:
4533    MatDiagonalScale() computes A = LAR, where
4534    L = a diagonal matrix (stored as a vector), R = a diagonal matrix (stored as a vector)
4535    The L scales the rows of the matrix, the R scales the columns of the matrix.
4536 
4537    Level: intermediate
4538 
4539    Concepts: matrices^diagonal scaling
4540    Concepts: diagonal scaling of matrices
4541 
4542 .seealso: MatScale()
4543 @*/
4544 PetscErrorCode PETSCMAT_DLLEXPORT MatDiagonalScale(Mat mat,Vec l,Vec r)
4545 {
4546   PetscErrorCode ierr;
4547 
4548   PetscFunctionBegin;
4549   PetscValidHeaderSpecific(mat,MAT_CLASSID,1);
4550   PetscValidType(mat,1);
4551   if (!mat->ops->diagonalscale) SETERRQ1(((PetscObject)mat)->comm,PETSC_ERR_SUP,"Mat type %s",((PetscObject)mat)->type_name);
4552   if (l) {PetscValidHeaderSpecific(l,VEC_CLASSID,2);PetscCheckSameComm(mat,1,l,2);}
4553   if (r) {PetscValidHeaderSpecific(r,VEC_CLASSID,3);PetscCheckSameComm(mat,1,r,3);}
4554   if (!mat->assembled) SETERRQ(((PetscObject)mat)->comm,PETSC_ERR_ARG_WRONGSTATE,"Not for unassembled matrix");
4555   if (mat->factortype) SETERRQ(((PetscObject)mat)->comm,PETSC_ERR_ARG_WRONGSTATE,"Not for factored matrix");
4556   ierr = MatPreallocated(mat);CHKERRQ(ierr);
4557 
4558   ierr = PetscLogEventBegin(MAT_Scale,mat,0,0,0);CHKERRQ(ierr);
4559   ierr = (*mat->ops->diagonalscale)(mat,l,r);CHKERRQ(ierr);
4560   ierr = PetscLogEventEnd(MAT_Scale,mat,0,0,0);CHKERRQ(ierr);
4561   ierr = PetscObjectStateIncrease((PetscObject)mat);CHKERRQ(ierr);
4562   PetscFunctionReturn(0);
4563 }
4564 
4565 #undef __FUNCT__
4566 #define __FUNCT__ "MatScale"
4567 /*@
4568     MatScale - Scales all elements of a matrix by a given number.
4569 
4570     Collective on Mat
4571 
4572     Input Parameters:
4573 +   mat - the matrix to be scaled
4574 -   a  - the scaling value
4575 
4576     Output Parameter:
4577 .   mat - the scaled matrix
4578 
4579     Level: intermediate
4580 
4581     Concepts: matrices^scaling all entries
4582 
4583 .seealso: MatDiagonalScale()
4584 @*/
4585 PetscErrorCode PETSCMAT_DLLEXPORT MatScale(Mat mat,PetscScalar a)
4586 {
4587   PetscErrorCode ierr;
4588 
4589   PetscFunctionBegin;
4590   PetscValidHeaderSpecific(mat,MAT_CLASSID,1);
4591   PetscValidType(mat,1);
4592   if (a != 1.0 && !mat->ops->scale) SETERRQ1(PETSC_COMM_SELF,PETSC_ERR_SUP,"Mat type %s",((PetscObject)mat)->type_name);
4593   if (!mat->assembled) SETERRQ(((PetscObject)mat)->comm,PETSC_ERR_ARG_WRONGSTATE,"Not for unassembled matrix");
4594   if (mat->factortype) SETERRQ(((PetscObject)mat)->comm,PETSC_ERR_ARG_WRONGSTATE,"Not for factored matrix");
4595   ierr = MatPreallocated(mat);CHKERRQ(ierr);
4596 
4597   ierr = PetscLogEventBegin(MAT_Scale,mat,0,0,0);CHKERRQ(ierr);
4598   if (a != 1.0) {
4599     ierr = (*mat->ops->scale)(mat,a);CHKERRQ(ierr);
4600     ierr = PetscObjectStateIncrease((PetscObject)mat);CHKERRQ(ierr);
4601   }
4602   ierr = PetscLogEventEnd(MAT_Scale,mat,0,0,0);CHKERRQ(ierr);
4603   PetscFunctionReturn(0);
4604 }
4605 
4606 #undef __FUNCT__
4607 #define __FUNCT__ "MatNorm"
4608 /*@
4609    MatNorm - Calculates various norms of a matrix.
4610 
4611    Collective on Mat
4612 
4613    Input Parameters:
4614 +  mat - the matrix
4615 -  type - the type of norm, NORM_1, NORM_FROBENIUS, NORM_INFINITY
4616 
4617    Output Parameters:
4618 .  nrm - the resulting norm
4619 
4620    Level: intermediate
4621 
4622    Concepts: matrices^norm
4623    Concepts: norm^of matrix
4624 @*/
4625 PetscErrorCode PETSCMAT_DLLEXPORT MatNorm(Mat mat,NormType type,PetscReal *nrm)
4626 {
4627   PetscErrorCode ierr;
4628 
4629   PetscFunctionBegin;
4630   PetscValidHeaderSpecific(mat,MAT_CLASSID,1);
4631   PetscValidType(mat,1);
4632   PetscValidScalarPointer(nrm,3);
4633 
4634   if (!mat->assembled) SETERRQ(((PetscObject)mat)->comm,PETSC_ERR_ARG_WRONGSTATE,"Not for unassembled matrix");
4635   if (mat->factortype) SETERRQ(((PetscObject)mat)->comm,PETSC_ERR_ARG_WRONGSTATE,"Not for factored matrix");
4636   if (!mat->ops->norm) SETERRQ1(((PetscObject)mat)->comm,PETSC_ERR_SUP,"Mat type %s",((PetscObject)mat)->type_name);
4637   ierr = MatPreallocated(mat);CHKERRQ(ierr);
4638 
4639   ierr = (*mat->ops->norm)(mat,type,nrm);CHKERRQ(ierr);
4640   PetscFunctionReturn(0);
4641 }
4642 
4643 /*
4644      This variable is used to prevent counting of MatAssemblyBegin() that
4645    are called from within a MatAssemblyEnd().
4646 */
4647 static PetscInt MatAssemblyEnd_InUse = 0;
4648 #undef __FUNCT__
4649 #define __FUNCT__ "MatAssemblyBegin"
4650 /*@
4651    MatAssemblyBegin - Begins assembling the matrix.  This routine should
4652    be called after completing all calls to MatSetValues().
4653 
4654    Collective on Mat
4655 
4656    Input Parameters:
4657 +  mat - the matrix
4658 -  type - type of assembly, either MAT_FLUSH_ASSEMBLY or MAT_FINAL_ASSEMBLY
4659 
4660    Notes:
4661    MatSetValues() generally caches the values.  The matrix is ready to
4662    use only after MatAssemblyBegin() and MatAssemblyEnd() have been called.
4663    Use MAT_FLUSH_ASSEMBLY when switching between ADD_VALUES and INSERT_VALUES
4664    in MatSetValues(); use MAT_FINAL_ASSEMBLY for the final assembly before
4665    using the matrix.
4666 
4667    Level: beginner
4668 
4669    Concepts: matrices^assembling
4670 
4671 .seealso: MatAssemblyEnd(), MatSetValues(), MatAssembled()
4672 @*/
4673 PetscErrorCode PETSCMAT_DLLEXPORT MatAssemblyBegin(Mat mat,MatAssemblyType type)
4674 {
4675   PetscErrorCode ierr;
4676 
4677   PetscFunctionBegin;
4678   PetscValidHeaderSpecific(mat,MAT_CLASSID,1);
4679   PetscValidType(mat,1);
4680   ierr = MatPreallocated(mat);CHKERRQ(ierr);
4681   if (mat->factortype) SETERRQ(((PetscObject)mat)->comm,PETSC_ERR_ARG_WRONGSTATE,"Not for factored matrix.\nDid you forget to call MatSetUnfactored()?");
4682   if (mat->assembled) {
4683     mat->was_assembled = PETSC_TRUE;
4684     mat->assembled     = PETSC_FALSE;
4685   }
4686   if (!MatAssemblyEnd_InUse) {
4687     ierr = PetscLogEventBegin(MAT_AssemblyBegin,mat,0,0,0);CHKERRQ(ierr);
4688     if (mat->ops->assemblybegin){ierr = (*mat->ops->assemblybegin)(mat,type);CHKERRQ(ierr);}
4689     ierr = PetscLogEventEnd(MAT_AssemblyBegin,mat,0,0,0);CHKERRQ(ierr);
4690   } else {
4691     if (mat->ops->assemblybegin){ierr = (*mat->ops->assemblybegin)(mat,type);CHKERRQ(ierr);}
4692   }
4693   PetscFunctionReturn(0);
4694 }
4695 
4696 #undef __FUNCT__
4697 #define __FUNCT__ "MatAssembled"
4698 /*@
4699    MatAssembled - Indicates if a matrix has been assembled and is ready for
4700      use; for example, in matrix-vector product.
4701 
4702    Collective on Mat
4703 
4704    Input Parameter:
4705 .  mat - the matrix
4706 
4707    Output Parameter:
4708 .  assembled - PETSC_TRUE or PETSC_FALSE
4709 
4710    Level: advanced
4711 
4712    Concepts: matrices^assembled?
4713 
4714 .seealso: MatAssemblyEnd(), MatSetValues(), MatAssemblyBegin()
4715 @*/
4716 PetscErrorCode PETSCMAT_DLLEXPORT MatAssembled(Mat mat,PetscTruth *assembled)
4717 {
4718   PetscFunctionBegin;
4719   PetscValidHeaderSpecific(mat,MAT_CLASSID,1);
4720   PetscValidType(mat,1);
4721   PetscValidPointer(assembled,2);
4722   *assembled = mat->assembled;
4723   PetscFunctionReturn(0);
4724 }
4725 
4726 #undef __FUNCT__
4727 #define __FUNCT__ "MatView_Private"
4728 /*
4729     Processes command line options to determine if/how a matrix
4730   is to be viewed. Called by MatAssemblyEnd() and MatLoad().
4731 */
4732 PetscErrorCode MatView_Private(Mat mat)
4733 {
4734   PetscErrorCode    ierr;
4735   PetscTruth        flg1 = PETSC_FALSE,flg2 = PETSC_FALSE,flg3 = PETSC_FALSE,flg4 = PETSC_FALSE,flg6 = PETSC_FALSE,flg7 = PETSC_FALSE,flg8 = PETSC_FALSE;
4736   static PetscTruth incall = PETSC_FALSE;
4737 #if defined(PETSC_USE_SOCKET_VIEWER)
4738   PetscTruth        flg5 = PETSC_FALSE;
4739 #endif
4740 
4741   PetscFunctionBegin;
4742   if (incall) PetscFunctionReturn(0);
4743   incall = PETSC_TRUE;
4744   ierr = PetscOptionsBegin(((PetscObject)mat)->comm,((PetscObject)mat)->prefix,"Matrix Options","Mat");CHKERRQ(ierr);
4745     ierr = PetscOptionsTruth("-mat_view_info","Information on matrix size","MatView",flg1,&flg1,PETSC_NULL);CHKERRQ(ierr);
4746     ierr = PetscOptionsTruth("-mat_view_info_detailed","Nonzeros in the matrix","MatView",flg2,&flg2,PETSC_NULL);CHKERRQ(ierr);
4747     ierr = PetscOptionsTruth("-mat_view","Print matrix to stdout","MatView",flg3,&flg3,PETSC_NULL);CHKERRQ(ierr);
4748     ierr = PetscOptionsTruth("-mat_view_matlab","Print matrix to stdout in a format Matlab can read","MatView",flg4,&flg4,PETSC_NULL);CHKERRQ(ierr);
4749 #if defined(PETSC_USE_SOCKET_VIEWER)
4750     ierr = PetscOptionsTruth("-mat_view_socket","Send matrix to socket (can be read from matlab)","MatView",flg5,&flg5,PETSC_NULL);CHKERRQ(ierr);
4751 #endif
4752     ierr = PetscOptionsTruth("-mat_view_binary","Save matrix to file in binary format","MatView",flg6,&flg6,PETSC_NULL);CHKERRQ(ierr);
4753     ierr = PetscOptionsTruth("-mat_view_draw","Draw the matrix nonzero structure","MatView",flg7,&flg7,PETSC_NULL);CHKERRQ(ierr);
4754   ierr = PetscOptionsEnd();CHKERRQ(ierr);
4755 
4756   if (flg1) {
4757     PetscViewer viewer;
4758 
4759     ierr = PetscViewerASCIIGetStdout(((PetscObject)mat)->comm,&viewer);CHKERRQ(ierr);
4760     ierr = PetscViewerPushFormat(viewer,PETSC_VIEWER_ASCII_INFO);CHKERRQ(ierr);
4761     ierr = MatView(mat,viewer);CHKERRQ(ierr);
4762     ierr = PetscViewerPopFormat(viewer);CHKERRQ(ierr);
4763   }
4764   if (flg2) {
4765     PetscViewer viewer;
4766 
4767     ierr = PetscViewerASCIIGetStdout(((PetscObject)mat)->comm,&viewer);CHKERRQ(ierr);
4768     ierr = PetscViewerPushFormat(viewer,PETSC_VIEWER_ASCII_INFO_DETAIL);CHKERRQ(ierr);
4769     ierr = MatView(mat,viewer);CHKERRQ(ierr);
4770     ierr = PetscViewerPopFormat(viewer);CHKERRQ(ierr);
4771   }
4772   if (flg3) {
4773     PetscViewer viewer;
4774 
4775     ierr = PetscViewerASCIIGetStdout(((PetscObject)mat)->comm,&viewer);CHKERRQ(ierr);
4776     ierr = MatView(mat,viewer);CHKERRQ(ierr);
4777   }
4778   if (flg4) {
4779     PetscViewer viewer;
4780 
4781     ierr = PetscViewerASCIIGetStdout(((PetscObject)mat)->comm,&viewer);CHKERRQ(ierr);
4782     ierr = PetscViewerPushFormat(viewer,PETSC_VIEWER_ASCII_MATLAB);CHKERRQ(ierr);
4783     ierr = MatView(mat,viewer);CHKERRQ(ierr);
4784     ierr = PetscViewerPopFormat(viewer);CHKERRQ(ierr);
4785   }
4786 #if defined(PETSC_USE_SOCKET_VIEWER)
4787   if (flg5) {
4788     ierr = MatView(mat,PETSC_VIEWER_SOCKET_(((PetscObject)mat)->comm));CHKERRQ(ierr);
4789     ierr = PetscViewerFlush(PETSC_VIEWER_SOCKET_(((PetscObject)mat)->comm));CHKERRQ(ierr);
4790   }
4791 #endif
4792   if (flg6) {
4793     ierr = MatView(mat,PETSC_VIEWER_BINARY_(((PetscObject)mat)->comm));CHKERRQ(ierr);
4794     ierr = PetscViewerFlush(PETSC_VIEWER_BINARY_(((PetscObject)mat)->comm));CHKERRQ(ierr);
4795   }
4796   if (flg7) {
4797     ierr = PetscOptionsGetTruth(((PetscObject)mat)->prefix,"-mat_view_contour",&flg8,PETSC_NULL);CHKERRQ(ierr);
4798     if (flg8) {
4799       PetscViewerPushFormat(PETSC_VIEWER_DRAW_(((PetscObject)mat)->comm),PETSC_VIEWER_DRAW_CONTOUR);CHKERRQ(ierr);
4800     }
4801     ierr = MatView(mat,PETSC_VIEWER_DRAW_(((PetscObject)mat)->comm));CHKERRQ(ierr);
4802     ierr = PetscViewerFlush(PETSC_VIEWER_DRAW_(((PetscObject)mat)->comm));CHKERRQ(ierr);
4803     if (flg8) {
4804       PetscViewerPopFormat(PETSC_VIEWER_DRAW_(((PetscObject)mat)->comm));CHKERRQ(ierr);
4805     }
4806   }
4807   incall = PETSC_FALSE;
4808   PetscFunctionReturn(0);
4809 }
4810 
4811 #undef __FUNCT__
4812 #define __FUNCT__ "MatAssemblyEnd"
4813 /*@
4814    MatAssemblyEnd - Completes assembling the matrix.  This routine should
4815    be called after MatAssemblyBegin().
4816 
4817    Collective on Mat
4818 
4819    Input Parameters:
4820 +  mat - the matrix
4821 -  type - type of assembly, either MAT_FLUSH_ASSEMBLY or MAT_FINAL_ASSEMBLY
4822 
4823    Options Database Keys:
4824 +  -mat_view_info - Prints info on matrix at conclusion of MatEndAssembly()
4825 .  -mat_view_info_detailed - Prints more detailed info
4826 .  -mat_view - Prints matrix in ASCII format
4827 .  -mat_view_matlab - Prints matrix in Matlab format
4828 .  -mat_view_draw - PetscDraws nonzero structure of matrix, using MatView() and PetscDrawOpenX().
4829 .  -display <name> - Sets display name (default is host)
4830 .  -draw_pause <sec> - Sets number of seconds to pause after display
4831 .  -mat_view_socket - Sends matrix to socket, can be accessed from Matlab (see users manual)
4832 .  -viewer_socket_machine <machine>
4833 .  -viewer_socket_port <port>
4834 .  -mat_view_binary - save matrix to file in binary format
4835 -  -viewer_binary_filename <name>
4836 
4837    Notes:
4838    MatSetValues() generally caches the values.  The matrix is ready to
4839    use only after MatAssemblyBegin() and MatAssemblyEnd() have been called.
4840    Use MAT_FLUSH_ASSEMBLY when switching between ADD_VALUES and INSERT_VALUES
4841    in MatSetValues(); use MAT_FINAL_ASSEMBLY for the final assembly before
4842    using the matrix.
4843 
4844    Level: beginner
4845 
4846 .seealso: MatAssemblyBegin(), MatSetValues(), PetscDrawOpenX(), MatView(), MatAssembled(), PetscViewerSocketOpen()
4847 @*/
4848 PetscErrorCode PETSCMAT_DLLEXPORT MatAssemblyEnd(Mat mat,MatAssemblyType type)
4849 {
4850   PetscErrorCode  ierr;
4851   static PetscInt inassm = 0;
4852   PetscTruth      flg = PETSC_FALSE;
4853 
4854   PetscFunctionBegin;
4855   PetscValidHeaderSpecific(mat,MAT_CLASSID,1);
4856   PetscValidType(mat,1);
4857 
4858   inassm++;
4859   MatAssemblyEnd_InUse++;
4860   if (MatAssemblyEnd_InUse == 1) { /* Do the logging only the first time through */
4861     ierr = PetscLogEventBegin(MAT_AssemblyEnd,mat,0,0,0);CHKERRQ(ierr);
4862     if (mat->ops->assemblyend) {
4863       ierr = (*mat->ops->assemblyend)(mat,type);CHKERRQ(ierr);
4864     }
4865     ierr = PetscLogEventEnd(MAT_AssemblyEnd,mat,0,0,0);CHKERRQ(ierr);
4866   } else {
4867     if (mat->ops->assemblyend) {
4868       ierr = (*mat->ops->assemblyend)(mat,type);CHKERRQ(ierr);
4869     }
4870   }
4871 
4872   /* Flush assembly is not a true assembly */
4873   if (type != MAT_FLUSH_ASSEMBLY) {
4874     mat->assembled  = PETSC_TRUE; mat->num_ass++;
4875   }
4876   mat->insertmode = NOT_SET_VALUES;
4877   MatAssemblyEnd_InUse--;
4878   ierr = PetscObjectStateIncrease((PetscObject)mat);CHKERRQ(ierr);
4879   if (!mat->symmetric_eternal) {
4880     mat->symmetric_set              = PETSC_FALSE;
4881     mat->hermitian_set              = PETSC_FALSE;
4882     mat->structurally_symmetric_set = PETSC_FALSE;
4883   }
4884   if (inassm == 1 && type != MAT_FLUSH_ASSEMBLY) {
4885     ierr = MatView_Private(mat);CHKERRQ(ierr);
4886     ierr = PetscOptionsGetTruth(((PetscObject)mat)->prefix,"-mat_is_symmetric",&flg,PETSC_NULL);CHKERRQ(ierr);
4887     if (flg) {
4888       PetscReal tol = 0.0;
4889       ierr = PetscOptionsGetReal(((PetscObject)mat)->prefix,"-mat_is_symmetric",&tol,PETSC_NULL);CHKERRQ(ierr);
4890       ierr = MatIsSymmetric(mat,tol,&flg);CHKERRQ(ierr);
4891       if (flg) {
4892         ierr = PetscPrintf(((PetscObject)mat)->comm,"Matrix is symmetric (tolerance %G)\n",tol);CHKERRQ(ierr);
4893       } else {
4894         ierr = PetscPrintf(((PetscObject)mat)->comm,"Matrix is not symmetric (tolerance %G)\n",tol);CHKERRQ(ierr);
4895       }
4896     }
4897   }
4898   inassm--;
4899   PetscFunctionReturn(0);
4900 }
4901 
4902 #undef __FUNCT__
4903 #define __FUNCT__ "MatSetOption"
4904 /*@
4905    MatSetOption - Sets a parameter option for a matrix. Some options
4906    may be specific to certain storage formats.  Some options
4907    determine how values will be inserted (or added). Sorted,
4908    row-oriented input will generally assemble the fastest. The default
4909    is row-oriented, nonsorted input.
4910 
4911    Collective on Mat
4912 
4913    Input Parameters:
4914 +  mat - the matrix
4915 .  option - the option, one of those listed below (and possibly others),
4916 -  flg - turn the option on (PETSC_TRUE) or off (PETSC_FALSE)
4917 
4918   Options Describing Matrix Structure:
4919 +    MAT_SYMMETRIC - symmetric in terms of both structure and value
4920 .    MAT_HERMITIAN - transpose is the complex conjugation
4921 .    MAT_STRUCTURALLY_SYMMETRIC - symmetric nonzero structure
4922 -    MAT_SYMMETRY_ETERNAL - if you would like the symmetry/Hermitian flag
4923                             you set to be kept with all future use of the matrix
4924                             including after MatAssemblyBegin/End() which could
4925                             potentially change the symmetry structure, i.e. you
4926                             KNOW the matrix will ALWAYS have the property you set.
4927 
4928 
4929    Options For Use with MatSetValues():
4930    Insert a logically dense subblock, which can be
4931 .    MAT_ROW_ORIENTED - row-oriented (default)
4932 
4933    Note these options reflect the data you pass in with MatSetValues(); it has
4934    nothing to do with how the data is stored internally in the matrix
4935    data structure.
4936 
4937    When (re)assembling a matrix, we can restrict the input for
4938    efficiency/debugging purposes.  These options include
4939 +    MAT_NEW_NONZERO_LOCATIONS - additional insertions will be
4940         allowed if they generate a new nonzero
4941 .    MAT_NEW_DIAGONALS - new diagonals will be allowed (for block diagonal format only)
4942 .    MAT_IGNORE_OFF_PROC_ENTRIES - drops off-processor entries
4943 .    MAT_NEW_NONZERO_LOCATION_ERR - generates an error for new matrix entry
4944 -    MAT_USE_HASH_TABLE - uses a hash table to speed up matrix assembly
4945 
4946    Notes:
4947    Some options are relevant only for particular matrix types and
4948    are thus ignored by others.  Other options are not supported by
4949    certain matrix types and will generate an error message if set.
4950 
4951    If using a Fortran 77 module to compute a matrix, one may need to
4952    use the column-oriented option (or convert to the row-oriented
4953    format).
4954 
4955    MAT_NEW_NONZERO_LOCATIONS set to PETSC_FALSE indicates that any add or insertion
4956    that would generate a new entry in the nonzero structure is instead
4957    ignored.  Thus, if memory has not alredy been allocated for this particular
4958    data, then the insertion is ignored. For dense matrices, in which
4959    the entire array is allocated, no entries are ever ignored.
4960    Set after the first MatAssemblyEnd()
4961 
4962    MAT_NEW_NONZERO_LOCATION_ERR indicates that any add or insertion
4963    that would generate a new entry in the nonzero structure instead produces
4964    an error. (Currently supported for AIJ and BAIJ formats only.)
4965    This is a useful flag when using SAME_NONZERO_PATTERN in calling
4966    KSPSetOperators() to ensure that the nonzero pattern truely does
4967    remain unchanged. Set after the first MatAssemblyEnd()
4968 
4969    MAT_NEW_NONZERO_ALLOCATION_ERR indicates that any add or insertion
4970    that would generate a new entry that has not been preallocated will
4971    instead produce an error. (Currently supported for AIJ and BAIJ formats
4972    only.) This is a useful flag when debugging matrix memory preallocation.
4973 
4974    MAT_IGNORE_OFF_PROC_ENTRIES indicates entries destined for
4975    other processors should be dropped, rather than stashed.
4976    This is useful if you know that the "owning" processor is also
4977    always generating the correct matrix entries, so that PETSc need
4978    not transfer duplicate entries generated on another processor.
4979 
4980    MAT_USE_HASH_TABLE indicates that a hash table be used to improve the
4981    searches during matrix assembly. When this flag is set, the hash table
4982    is created during the first Matrix Assembly. This hash table is
4983    used the next time through, during MatSetVaules()/MatSetVaulesBlocked()
4984    to improve the searching of indices. MAT_NEW_NONZERO_LOCATIONS flag
4985    should be used with MAT_USE_HASH_TABLE flag. This option is currently
4986    supported by MATMPIBAIJ format only.
4987 
4988    MAT_KEEP_NONZERO_PATTERN indicates when MatZeroRows() is called the zeroed entries
4989    are kept in the nonzero structure
4990 
4991    MAT_IGNORE_ZERO_ENTRIES - for AIJ/IS matrices this will stop zero values from creating
4992    a zero location in the matrix
4993 
4994    MAT_USE_INODES - indicates using inode version of the code - works with AIJ and
4995    ROWBS matrix types
4996 
4997    Level: intermediate
4998 
4999    Concepts: matrices^setting options
5000 
5001 @*/
5002 PetscErrorCode PETSCMAT_DLLEXPORT MatSetOption(Mat mat,MatOption op,PetscTruth flg)
5003 {
5004   PetscErrorCode ierr;
5005 
5006   PetscFunctionBegin;
5007   PetscValidHeaderSpecific(mat,MAT_CLASSID,1);
5008   PetscValidType(mat,1);
5009   if (((int) op) < 0 || ((int) op) >= NUM_MAT_OPTIONS) SETERRQ1(((PetscObject)mat)->comm,PETSC_ERR_ARG_OUTOFRANGE,"Options %d is out of range",(int)op);
5010   if (!((PetscObject)mat)->type_name) SETERRQ(((PetscObject)mat)->comm,PETSC_ERR_ARG_TYPENOTSET,"Cannot set options until type and size have been set, see MatSetType() and MatSetSizes()");
5011   ierr = MatPreallocated(mat);CHKERRQ(ierr);
5012   switch (op) {
5013   case MAT_SYMMETRIC:
5014     mat->symmetric                  = flg;
5015     if (flg) mat->structurally_symmetric = PETSC_TRUE;
5016     mat->symmetric_set              = PETSC_TRUE;
5017     mat->structurally_symmetric_set = flg;
5018     break;
5019   case MAT_HERMITIAN:
5020     mat->hermitian                  = flg;
5021     if (flg) mat->structurally_symmetric = PETSC_TRUE;
5022     mat->hermitian_set              = PETSC_TRUE;
5023     mat->structurally_symmetric_set = flg;
5024     break;
5025   case MAT_STRUCTURALLY_SYMMETRIC:
5026     mat->structurally_symmetric     = flg;
5027     mat->structurally_symmetric_set = PETSC_TRUE;
5028     break;
5029   case MAT_SYMMETRY_ETERNAL:
5030     mat->symmetric_eternal          = flg;
5031     break;
5032   default:
5033     break;
5034   }
5035   if (mat->ops->setoption) {
5036     ierr = (*mat->ops->setoption)(mat,op,flg);CHKERRQ(ierr);
5037   }
5038   PetscFunctionReturn(0);
5039 }
5040 
5041 #undef __FUNCT__
5042 #define __FUNCT__ "MatZeroEntries"
5043 /*@
5044    MatZeroEntries - Zeros all entries of a matrix.  For sparse matrices
5045    this routine retains the old nonzero structure.
5046 
5047    Collective on Mat
5048 
5049    Input Parameters:
5050 .  mat - the matrix
5051 
5052    Level: intermediate
5053 
5054    Concepts: matrices^zeroing
5055 
5056 .seealso: MatZeroRows()
5057 @*/
5058 PetscErrorCode PETSCMAT_DLLEXPORT MatZeroEntries(Mat mat)
5059 {
5060   PetscErrorCode ierr;
5061 
5062   PetscFunctionBegin;
5063   PetscValidHeaderSpecific(mat,MAT_CLASSID,1);
5064   PetscValidType(mat,1);
5065   if (mat->factortype) SETERRQ(((PetscObject)mat)->comm,PETSC_ERR_ARG_WRONGSTATE,"Not for factored matrix");
5066   if (mat->insertmode != NOT_SET_VALUES) SETERRQ(PETSC_COMM_SELF,PETSC_ERR_ARG_WRONGSTATE,"Not for matrices where you have set values but not yet assembled");
5067   if (!mat->ops->zeroentries) SETERRQ1(((PetscObject)mat)->comm,PETSC_ERR_SUP,"Mat type %s",((PetscObject)mat)->type_name);
5068   ierr = MatPreallocated(mat);CHKERRQ(ierr);
5069 
5070   ierr = PetscLogEventBegin(MAT_ZeroEntries,mat,0,0,0);CHKERRQ(ierr);
5071   ierr = (*mat->ops->zeroentries)(mat);CHKERRQ(ierr);
5072   ierr = PetscLogEventEnd(MAT_ZeroEntries,mat,0,0,0);CHKERRQ(ierr);
5073   ierr = PetscObjectStateIncrease((PetscObject)mat);CHKERRQ(ierr);
5074   PetscFunctionReturn(0);
5075 }
5076 
5077 #undef __FUNCT__
5078 #define __FUNCT__ "MatZeroRows"
5079 /*@C
5080    MatZeroRows - Zeros all entries (except possibly the main diagonal)
5081    of a set of rows of a matrix.
5082 
5083    Collective on Mat
5084 
5085    Input Parameters:
5086 +  mat - the matrix
5087 .  numRows - the number of rows to remove
5088 .  rows - the global row indices
5089 -  diag - value put in all diagonals of eliminated rows (0.0 will even eliminate diagonal entry)
5090 
5091    Notes:
5092    For the AIJ and BAIJ matrix formats this removes the old nonzero structure,
5093    but does not release memory.  For the dense and block diagonal
5094    formats this does not alter the nonzero structure.
5095 
5096    If the option MatSetOption(mat,MAT_KEEP_NONZERO_PATTERN,PETSC_TRUE) the nonzero structure
5097    of the matrix is not changed (even for AIJ and BAIJ matrices) the values are
5098    merely zeroed.
5099 
5100    The user can set a value in the diagonal entry (or for the AIJ and
5101    row formats can optionally remove the main diagonal entry from the
5102    nonzero structure as well, by passing 0.0 as the final argument).
5103 
5104    For the parallel case, all processes that share the matrix (i.e.,
5105    those in the communicator used for matrix creation) MUST call this
5106    routine, regardless of whether any rows being zeroed are owned by
5107    them.
5108 
5109    Each processor can indicate any rows in the entire matrix to be zeroed (i.e. each process does NOT have to
5110    list only rows local to itself).
5111 
5112    Level: intermediate
5113 
5114    Concepts: matrices^zeroing rows
5115 
5116 .seealso: MatZeroRowsIS(), MatZeroEntries(), MatZeroRowsLocal(), MatSetOption()
5117 @*/
5118 PetscErrorCode PETSCMAT_DLLEXPORT MatZeroRows(Mat mat,PetscInt numRows,const PetscInt rows[],PetscScalar diag)
5119 {
5120   PetscErrorCode ierr;
5121 
5122   PetscFunctionBegin;
5123   PetscValidHeaderSpecific(mat,MAT_CLASSID,1);
5124   PetscValidType(mat,1);
5125   if (numRows) PetscValidIntPointer(rows,3);
5126   if (!mat->assembled) SETERRQ(((PetscObject)mat)->comm,PETSC_ERR_ARG_WRONGSTATE,"Not for unassembled matrix");
5127   if (mat->factortype) SETERRQ(((PetscObject)mat)->comm,PETSC_ERR_ARG_WRONGSTATE,"Not for factored matrix");
5128   if (!mat->ops->zerorows) SETERRQ1(((PetscObject)mat)->comm,PETSC_ERR_SUP,"Mat type %s",((PetscObject)mat)->type_name);
5129   ierr = MatPreallocated(mat);CHKERRQ(ierr);
5130 
5131   ierr = (*mat->ops->zerorows)(mat,numRows,rows,diag);CHKERRQ(ierr);
5132   ierr = MatView_Private(mat);CHKERRQ(ierr);
5133   ierr = PetscObjectStateIncrease((PetscObject)mat);CHKERRQ(ierr);
5134   PetscFunctionReturn(0);
5135 }
5136 
5137 #undef __FUNCT__
5138 #define __FUNCT__ "MatZeroRowsIS"
5139 /*@C
5140    MatZeroRowsIS - Zeros all entries (except possibly the main diagonal)
5141    of a set of rows of a matrix.
5142 
5143    Collective on Mat
5144 
5145    Input Parameters:
5146 +  mat - the matrix
5147 .  is - index set of rows to remove
5148 -  diag - value put in all diagonals of eliminated rows
5149 
5150    Notes:
5151    For the AIJ and BAIJ matrix formats this removes the old nonzero structure,
5152    but does not release memory.  For the dense and block diagonal
5153    formats this does not alter the nonzero structure.
5154 
5155    If the option MatSetOption(mat,MAT_KEEP_NONZERO_PATTERN,PETSC_TRUE) the nonzero structure
5156    of the matrix is not changed (even for AIJ and BAIJ matrices) the values are
5157    merely zeroed.
5158 
5159    The user can set a value in the diagonal entry (or for the AIJ and
5160    row formats can optionally remove the main diagonal entry from the
5161    nonzero structure as well, by passing 0.0 as the final argument).
5162 
5163    For the parallel case, all processes that share the matrix (i.e.,
5164    those in the communicator used for matrix creation) MUST call this
5165    routine, regardless of whether any rows being zeroed are owned by
5166    them.
5167 
5168    Each processor should list the rows that IT wants zeroed
5169 
5170    Level: intermediate
5171 
5172    Concepts: matrices^zeroing rows
5173 
5174 .seealso: MatZeroRows(), MatZeroEntries(), MatZeroRowsLocal(), MatSetOption()
5175 @*/
5176 PetscErrorCode PETSCMAT_DLLEXPORT MatZeroRowsIS(Mat mat,IS is,PetscScalar diag)
5177 {
5178   PetscInt       numRows;
5179   const PetscInt *rows;
5180   PetscErrorCode ierr;
5181 
5182   PetscFunctionBegin;
5183   PetscValidHeaderSpecific(mat,MAT_CLASSID,1);
5184   PetscValidType(mat,1);
5185   PetscValidHeaderSpecific(is,IS_CLASSID,2);
5186   ierr = ISGetLocalSize(is,&numRows);CHKERRQ(ierr);
5187   ierr = ISGetIndices(is,&rows);CHKERRQ(ierr);
5188   ierr = MatZeroRows(mat,numRows,rows,diag);CHKERRQ(ierr);
5189   ierr = ISRestoreIndices(is,&rows);CHKERRQ(ierr);
5190   PetscFunctionReturn(0);
5191 }
5192 
5193 #undef __FUNCT__
5194 #define __FUNCT__ "MatZeroRowsLocal"
5195 /*@C
5196    MatZeroRowsLocal - Zeros all entries (except possibly the main diagonal)
5197    of a set of rows of a matrix; using local numbering of rows.
5198 
5199    Collective on Mat
5200 
5201    Input Parameters:
5202 +  mat - the matrix
5203 .  numRows - the number of rows to remove
5204 .  rows - the global row indices
5205 -  diag - value put in all diagonals of eliminated rows
5206 
5207    Notes:
5208    Before calling MatZeroRowsLocal(), the user must first set the
5209    local-to-global mapping by calling MatSetLocalToGlobalMapping().
5210 
5211    For the AIJ matrix formats this removes the old nonzero structure,
5212    but does not release memory.  For the dense and block diagonal
5213    formats this does not alter the nonzero structure.
5214 
5215    If the option MatSetOption(mat,MAT_KEEP_NONZERO_PATTERN,PETSC_TRUE) the nonzero structure
5216    of the matrix is not changed (even for AIJ and BAIJ matrices) the values are
5217    merely zeroed.
5218 
5219    The user can set a value in the diagonal entry (or for the AIJ and
5220    row formats can optionally remove the main diagonal entry from the
5221    nonzero structure as well, by passing 0.0 as the final argument).
5222 
5223    Level: intermediate
5224 
5225    Concepts: matrices^zeroing
5226 
5227 .seealso: MatZeroRows(), MatZeroRowsLocalIS(), MatZeroEntries(), MatZeroRows(), MatSetLocalToGlobalMapping
5228 @*/
5229 PetscErrorCode PETSCMAT_DLLEXPORT MatZeroRowsLocal(Mat mat,PetscInt numRows,const PetscInt rows[],PetscScalar diag)
5230 {
5231   PetscErrorCode ierr;
5232 
5233   PetscFunctionBegin;
5234   PetscValidHeaderSpecific(mat,MAT_CLASSID,1);
5235   PetscValidType(mat,1);
5236   if (numRows) PetscValidIntPointer(rows,3);
5237   if (!mat->assembled) SETERRQ(((PetscObject)mat)->comm,PETSC_ERR_ARG_WRONGSTATE,"Not for unassembled matrix");
5238   if (mat->factortype) SETERRQ(((PetscObject)mat)->comm,PETSC_ERR_ARG_WRONGSTATE,"Not for factored matrix");
5239   ierr = MatPreallocated(mat);CHKERRQ(ierr);
5240 
5241   if (mat->ops->zerorowslocal) {
5242     ierr = (*mat->ops->zerorowslocal)(mat,numRows,rows,diag);CHKERRQ(ierr);
5243   } else {
5244     IS             is, newis;
5245     const PetscInt *newRows;
5246 
5247     if (!mat->mapping) SETERRQ(((PetscObject)mat)->comm,PETSC_ERR_ARG_WRONGSTATE,"Need to provide local to global mapping to matrix first");
5248     ierr = ISCreateGeneral(PETSC_COMM_SELF,numRows,rows,&is);CHKERRQ(ierr);
5249     ierr = ISLocalToGlobalMappingApplyIS(mat->mapping,is,&newis);CHKERRQ(ierr);
5250     ierr = ISGetIndices(newis,&newRows);CHKERRQ(ierr);
5251     ierr = (*mat->ops->zerorows)(mat,numRows,newRows,diag);CHKERRQ(ierr);
5252     ierr = ISRestoreIndices(newis,&newRows);CHKERRQ(ierr);
5253     ierr = ISDestroy(newis);CHKERRQ(ierr);
5254     ierr = ISDestroy(is);CHKERRQ(ierr);
5255   }
5256   ierr = PetscObjectStateIncrease((PetscObject)mat);CHKERRQ(ierr);
5257   PetscFunctionReturn(0);
5258 }
5259 
5260 #undef __FUNCT__
5261 #define __FUNCT__ "MatZeroRowsLocalIS"
5262 /*@C
5263    MatZeroRowsLocalIS - Zeros all entries (except possibly the main diagonal)
5264    of a set of rows of a matrix; using local numbering of rows.
5265 
5266    Collective on Mat
5267 
5268    Input Parameters:
5269 +  mat - the matrix
5270 .  is - index set of rows to remove
5271 -  diag - value put in all diagonals of eliminated rows
5272 
5273    Notes:
5274    Before calling MatZeroRowsLocalIS(), the user must first set the
5275    local-to-global mapping by calling MatSetLocalToGlobalMapping().
5276 
5277    For the AIJ matrix formats this removes the old nonzero structure,
5278    but does not release memory.  For the dense and block diagonal
5279    formats this does not alter the nonzero structure.
5280 
5281    If the option MatSetOption(mat,MAT_KEEP_NONZERO_PATTERN,PETSC_TRUE) the nonzero structure
5282    of the matrix is not changed (even for AIJ and BAIJ matrices) the values are
5283    merely zeroed.
5284 
5285    The user can set a value in the diagonal entry (or for the AIJ and
5286    row formats can optionally remove the main diagonal entry from the
5287    nonzero structure as well, by passing 0.0 as the final argument).
5288 
5289    Level: intermediate
5290 
5291    Concepts: matrices^zeroing
5292 
5293 .seealso: MatZeroRows(), MatZeroRowsLocal(), MatZeroEntries(), MatZeroRows(), MatSetLocalToGlobalMapping
5294 @*/
5295 PetscErrorCode PETSCMAT_DLLEXPORT MatZeroRowsLocalIS(Mat mat,IS is,PetscScalar diag)
5296 {
5297   PetscErrorCode ierr;
5298   PetscInt       numRows;
5299   const PetscInt *rows;
5300 
5301   PetscFunctionBegin;
5302   PetscValidHeaderSpecific(mat,MAT_CLASSID,1);
5303   PetscValidType(mat,1);
5304   PetscValidHeaderSpecific(is,IS_CLASSID,2);
5305   if (!mat->assembled) SETERRQ(((PetscObject)mat)->comm,PETSC_ERR_ARG_WRONGSTATE,"Not for unassembled matrix");
5306   if (mat->factortype) SETERRQ(((PetscObject)mat)->comm,PETSC_ERR_ARG_WRONGSTATE,"Not for factored matrix");
5307   ierr = MatPreallocated(mat);CHKERRQ(ierr);
5308 
5309   ierr = ISGetLocalSize(is,&numRows);CHKERRQ(ierr);
5310   ierr = ISGetIndices(is,&rows);CHKERRQ(ierr);
5311   ierr = MatZeroRowsLocal(mat,numRows,rows,diag);CHKERRQ(ierr);
5312   ierr = ISRestoreIndices(is,&rows);CHKERRQ(ierr);
5313   PetscFunctionReturn(0);
5314 }
5315 
5316 #undef __FUNCT__
5317 #define __FUNCT__ "MatGetSize"
5318 /*@
5319    MatGetSize - Returns the numbers of rows and columns in a matrix.
5320 
5321    Not Collective
5322 
5323    Input Parameter:
5324 .  mat - the matrix
5325 
5326    Output Parameters:
5327 +  m - the number of global rows
5328 -  n - the number of global columns
5329 
5330    Note: both output parameters can be PETSC_NULL on input.
5331 
5332    Level: beginner
5333 
5334    Concepts: matrices^size
5335 
5336 .seealso: MatGetLocalSize()
5337 @*/
5338 PetscErrorCode PETSCMAT_DLLEXPORT MatGetSize(Mat mat,PetscInt *m,PetscInt* n)
5339 {
5340   PetscFunctionBegin;
5341   PetscValidHeaderSpecific(mat,MAT_CLASSID,1);
5342   if (m) *m = mat->rmap->N;
5343   if (n) *n = mat->cmap->N;
5344   PetscFunctionReturn(0);
5345 }
5346 
5347 #undef __FUNCT__
5348 #define __FUNCT__ "MatGetLocalSize"
5349 /*@
5350    MatGetLocalSize - Returns the number of rows and columns in a matrix
5351    stored locally.  This information may be implementation dependent, so
5352    use with care.
5353 
5354    Not Collective
5355 
5356    Input Parameters:
5357 .  mat - the matrix
5358 
5359    Output Parameters:
5360 +  m - the number of local rows
5361 -  n - the number of local columns
5362 
5363    Note: both output parameters can be PETSC_NULL on input.
5364 
5365    Level: beginner
5366 
5367    Concepts: matrices^local size
5368 
5369 .seealso: MatGetSize()
5370 @*/
5371 PetscErrorCode PETSCMAT_DLLEXPORT MatGetLocalSize(Mat mat,PetscInt *m,PetscInt* n)
5372 {
5373   PetscFunctionBegin;
5374   PetscValidHeaderSpecific(mat,MAT_CLASSID,1);
5375   if (m) PetscValidIntPointer(m,2);
5376   if (n) PetscValidIntPointer(n,3);
5377   if (m) *m = mat->rmap->n;
5378   if (n) *n = mat->cmap->n;
5379   PetscFunctionReturn(0);
5380 }
5381 
5382 #undef __FUNCT__
5383 #define __FUNCT__ "MatGetOwnershipRangeColumn"
5384 /*@
5385    MatGetOwnershipRangeColumn - Returns the range of matrix columns owned by
5386    this processor.
5387 
5388    Not Collective, unless matrix has not been allocated, then collective on Mat
5389 
5390    Input Parameters:
5391 .  mat - the matrix
5392 
5393    Output Parameters:
5394 +  m - the global index of the first local column
5395 -  n - one more than the global index of the last local column
5396 
5397    Notes: both output parameters can be PETSC_NULL on input.
5398 
5399    Level: developer
5400 
5401    Concepts: matrices^column ownership
5402 
5403 .seealso:  MatGetOwnershipRange(), MatGetOwnershipRanges(), MatGetOwnershipRangesColumn()
5404 
5405 @*/
5406 PetscErrorCode PETSCMAT_DLLEXPORT MatGetOwnershipRangeColumn(Mat mat,PetscInt *m,PetscInt* n)
5407 {
5408   PetscErrorCode ierr;
5409 
5410   PetscFunctionBegin;
5411   PetscValidHeaderSpecific(mat,MAT_CLASSID,1);
5412   PetscValidType(mat,1);
5413   if (m) PetscValidIntPointer(m,2);
5414   if (n) PetscValidIntPointer(n,3);
5415   ierr = MatPreallocated(mat);CHKERRQ(ierr);
5416   if (m) *m = mat->cmap->rstart;
5417   if (n) *n = mat->cmap->rend;
5418   PetscFunctionReturn(0);
5419 }
5420 
5421 #undef __FUNCT__
5422 #define __FUNCT__ "MatGetOwnershipRange"
5423 /*@
5424    MatGetOwnershipRange - Returns the range of matrix rows owned by
5425    this processor, assuming that the matrix is laid out with the first
5426    n1 rows on the first processor, the next n2 rows on the second, etc.
5427    For certain parallel layouts this range may not be well defined.
5428 
5429    Not Collective, unless matrix has not been allocated, then collective on Mat
5430 
5431    Input Parameters:
5432 .  mat - the matrix
5433 
5434    Output Parameters:
5435 +  m - the global index of the first local row
5436 -  n - one more than the global index of the last local row
5437 
5438    Note: both output parameters can be PETSC_NULL on input.
5439 
5440    Level: beginner
5441 
5442    Concepts: matrices^row ownership
5443 
5444 .seealso:   MatGetOwnershipRanges(), MatGetOwnershipRangeColumn(), MatGetOwnershipRangesColumn()
5445 
5446 @*/
5447 PetscErrorCode PETSCMAT_DLLEXPORT MatGetOwnershipRange(Mat mat,PetscInt *m,PetscInt* n)
5448 {
5449   PetscErrorCode ierr;
5450 
5451   PetscFunctionBegin;
5452   PetscValidHeaderSpecific(mat,MAT_CLASSID,1);
5453   PetscValidType(mat,1);
5454   if (m) PetscValidIntPointer(m,2);
5455   if (n) PetscValidIntPointer(n,3);
5456   ierr = MatPreallocated(mat);CHKERRQ(ierr);
5457   if (m) *m = mat->rmap->rstart;
5458   if (n) *n = mat->rmap->rend;
5459   PetscFunctionReturn(0);
5460 }
5461 
5462 #undef __FUNCT__
5463 #define __FUNCT__ "MatGetOwnershipRanges"
5464 /*@C
5465    MatGetOwnershipRanges - Returns the range of matrix rows owned by
5466    each process
5467 
5468    Not Collective, unless matrix has not been allocated, then collective on Mat
5469 
5470    Input Parameters:
5471 .  mat - the matrix
5472 
5473    Output Parameters:
5474 .  ranges - start of each processors portion plus one more then the total length at the end
5475 
5476    Level: beginner
5477 
5478    Concepts: matrices^row ownership
5479 
5480 .seealso:   MatGetOwnershipRange(), MatGetOwnershipRangeColumn(), MatGetOwnershipRangesColumn()
5481 
5482 @*/
5483 PetscErrorCode PETSCMAT_DLLEXPORT MatGetOwnershipRanges(Mat mat,const PetscInt **ranges)
5484 {
5485   PetscErrorCode ierr;
5486 
5487   PetscFunctionBegin;
5488   PetscValidHeaderSpecific(mat,MAT_CLASSID,1);
5489   PetscValidType(mat,1);
5490   ierr = MatPreallocated(mat);CHKERRQ(ierr);
5491   ierr = PetscLayoutGetRanges(mat->rmap,ranges);CHKERRQ(ierr);
5492   PetscFunctionReturn(0);
5493 }
5494 
5495 #undef __FUNCT__
5496 #define __FUNCT__ "MatGetOwnershipRangesColumn"
5497 /*@C
5498    MatGetOwnershipRangesColumn - Returns the range of local columns for each process
5499 
5500    Not Collective, unless matrix has not been allocated, then collective on Mat
5501 
5502    Input Parameters:
5503 .  mat - the matrix
5504 
5505    Output Parameters:
5506 .  ranges - start of each processors portion plus one more then the total length at the end
5507 
5508    Level: beginner
5509 
5510    Concepts: matrices^column ownership
5511 
5512 .seealso:   MatGetOwnershipRange(), MatGetOwnershipRangeColumn(), MatGetOwnershipRanges()
5513 
5514 @*/
5515 PetscErrorCode PETSCMAT_DLLEXPORT MatGetOwnershipRangesColumn(Mat mat,const PetscInt **ranges)
5516 {
5517   PetscErrorCode ierr;
5518 
5519   PetscFunctionBegin;
5520   PetscValidHeaderSpecific(mat,MAT_CLASSID,1);
5521   PetscValidType(mat,1);
5522   ierr = MatPreallocated(mat);CHKERRQ(ierr);
5523   ierr = PetscLayoutGetRanges(mat->cmap,ranges);CHKERRQ(ierr);
5524   PetscFunctionReturn(0);
5525 }
5526 
5527 #undef __FUNCT__
5528 #define __FUNCT__ "MatILUFactorSymbolic"
5529 /*@C
5530    MatILUFactorSymbolic - Performs symbolic ILU factorization of a matrix.
5531    Uses levels of fill only, not drop tolerance. Use MatLUFactorNumeric()
5532    to complete the factorization.
5533 
5534    Collective on Mat
5535 
5536    Input Parameters:
5537 +  mat - the matrix
5538 .  row - row permutation
5539 .  column - column permutation
5540 -  info - structure containing
5541 $      levels - number of levels of fill.
5542 $      expected fill - as ratio of original fill.
5543 $      1 or 0 - indicating force fill on diagonal (improves robustness for matrices
5544                 missing diagonal entries)
5545 
5546    Output Parameters:
5547 .  fact - new matrix that has been symbolically factored
5548 
5549    Notes:
5550    See the users manual for additional information about
5551    choosing the fill factor for better efficiency.
5552 
5553    Most users should employ the simplified KSP interface for linear solvers
5554    instead of working directly with matrix algebra routines such as this.
5555    See, e.g., KSPCreate().
5556 
5557    Level: developer
5558 
5559   Concepts: matrices^symbolic LU factorization
5560   Concepts: matrices^factorization
5561   Concepts: LU^symbolic factorization
5562 
5563 .seealso: MatLUFactorSymbolic(), MatLUFactorNumeric(), MatCholeskyFactor()
5564           MatGetOrdering(), MatFactorInfo
5565 
5566     Developer Note: fortran interface is not autogenerated as the f90
5567     interface defintion cannot be generated correctly [due to MatFactorInfo]
5568 
5569 @*/
5570 PetscErrorCode PETSCMAT_DLLEXPORT MatILUFactorSymbolic(Mat fact,Mat mat,IS row,IS col,const MatFactorInfo *info)
5571 {
5572   PetscErrorCode ierr;
5573 
5574   PetscFunctionBegin;
5575   PetscValidHeaderSpecific(mat,MAT_CLASSID,1);
5576   PetscValidType(mat,1);
5577   PetscValidHeaderSpecific(row,IS_CLASSID,2);
5578   PetscValidHeaderSpecific(col,IS_CLASSID,3);
5579   PetscValidPointer(info,4);
5580   PetscValidPointer(fact,5);
5581   if (info->levels < 0) SETERRQ1(((PetscObject)mat)->comm,PETSC_ERR_ARG_OUTOFRANGE,"Levels of fill negative %D",(PetscInt)info->levels);
5582   if (info->fill < 1.0) SETERRQ1(((PetscObject)mat)->comm,PETSC_ERR_ARG_OUTOFRANGE,"Expected fill less than 1.0 %G",info->fill);
5583   if (!(fact)->ops->ilufactorsymbolic) SETERRQ1(((PetscObject)mat)->comm,PETSC_ERR_SUP,"Matrix type %s  symbolic ILU",((PetscObject)mat)->type_name);
5584   if (!mat->assembled) SETERRQ(((PetscObject)mat)->comm,PETSC_ERR_ARG_WRONGSTATE,"Not for unassembled matrix");
5585   if (mat->factortype) SETERRQ(((PetscObject)mat)->comm,PETSC_ERR_ARG_WRONGSTATE,"Not for factored matrix");
5586   ierr = MatPreallocated(mat);CHKERRQ(ierr);
5587 
5588   ierr = PetscLogEventBegin(MAT_ILUFactorSymbolic,mat,row,col,0);CHKERRQ(ierr);
5589   ierr = (fact->ops->ilufactorsymbolic)(fact,mat,row,col,info);CHKERRQ(ierr);
5590   ierr = PetscLogEventEnd(MAT_ILUFactorSymbolic,mat,row,col,0);CHKERRQ(ierr);
5591   PetscFunctionReturn(0);
5592 }
5593 
5594 #undef __FUNCT__
5595 #define __FUNCT__ "MatICCFactorSymbolic"
5596 /*@C
5597    MatICCFactorSymbolic - Performs symbolic incomplete
5598    Cholesky factorization for a symmetric matrix.  Use
5599    MatCholeskyFactorNumeric() to complete the factorization.
5600 
5601    Collective on Mat
5602 
5603    Input Parameters:
5604 +  mat - the matrix
5605 .  perm - row and column permutation
5606 -  info - structure containing
5607 $      levels - number of levels of fill.
5608 $      expected fill - as ratio of original fill.
5609 
5610    Output Parameter:
5611 .  fact - the factored matrix
5612 
5613    Notes:
5614    Most users should employ the KSP interface for linear solvers
5615    instead of working directly with matrix algebra routines such as this.
5616    See, e.g., KSPCreate().
5617 
5618    Level: developer
5619 
5620   Concepts: matrices^symbolic incomplete Cholesky factorization
5621   Concepts: matrices^factorization
5622   Concepts: Cholsky^symbolic factorization
5623 
5624 .seealso: MatCholeskyFactorNumeric(), MatCholeskyFactor(), MatFactorInfo
5625 
5626     Developer Note: fortran interface is not autogenerated as the f90
5627     interface defintion cannot be generated correctly [due to MatFactorInfo]
5628 
5629 @*/
5630 PetscErrorCode PETSCMAT_DLLEXPORT MatICCFactorSymbolic(Mat fact,Mat mat,IS perm,const MatFactorInfo *info)
5631 {
5632   PetscErrorCode ierr;
5633 
5634   PetscFunctionBegin;
5635   PetscValidHeaderSpecific(mat,MAT_CLASSID,1);
5636   PetscValidType(mat,1);
5637   PetscValidHeaderSpecific(perm,IS_CLASSID,2);
5638   PetscValidPointer(info,3);
5639   PetscValidPointer(fact,4);
5640   if (mat->factortype) SETERRQ(((PetscObject)mat)->comm,PETSC_ERR_ARG_WRONGSTATE,"Not for factored matrix");
5641   if (info->levels < 0) SETERRQ1(((PetscObject)mat)->comm,PETSC_ERR_ARG_OUTOFRANGE,"Levels negative %D",(PetscInt) info->levels);
5642   if (info->fill < 1.0) SETERRQ1(((PetscObject)mat)->comm,PETSC_ERR_ARG_OUTOFRANGE,"Expected fill less than 1.0 %G",info->fill);
5643   if (!(fact)->ops->iccfactorsymbolic) SETERRQ1(((PetscObject)mat)->comm,PETSC_ERR_SUP,"Matrix type %s  symbolic ICC",((PetscObject)mat)->type_name);
5644   if (!mat->assembled) SETERRQ(((PetscObject)mat)->comm,PETSC_ERR_ARG_WRONGSTATE,"Not for unassembled matrix");
5645   ierr = MatPreallocated(mat);CHKERRQ(ierr);
5646 
5647   ierr = PetscLogEventBegin(MAT_ICCFactorSymbolic,mat,perm,0,0);CHKERRQ(ierr);
5648   ierr = (fact->ops->iccfactorsymbolic)(fact,mat,perm,info);CHKERRQ(ierr);
5649   ierr = PetscLogEventEnd(MAT_ICCFactorSymbolic,mat,perm,0,0);CHKERRQ(ierr);
5650   PetscFunctionReturn(0);
5651 }
5652 
5653 #undef __FUNCT__
5654 #define __FUNCT__ "MatGetArray"
5655 /*@C
5656    MatGetArray - Returns a pointer to the element values in the matrix.
5657    The result of this routine is dependent on the underlying matrix data
5658    structure, and may not even work for certain matrix types.  You MUST
5659    call MatRestoreArray() when you no longer need to access the array.
5660 
5661    Not Collective
5662 
5663    Input Parameter:
5664 .  mat - the matrix
5665 
5666    Output Parameter:
5667 .  v - the location of the values
5668 
5669 
5670    Fortran Note:
5671    This routine is used differently from Fortran, e.g.,
5672 .vb
5673         Mat         mat
5674         PetscScalar mat_array(1)
5675         PetscOffset i_mat
5676         PetscErrorCode ierr
5677         call MatGetArray(mat,mat_array,i_mat,ierr)
5678 
5679   C  Access first local entry in matrix; note that array is
5680   C  treated as one dimensional
5681         value = mat_array(i_mat + 1)
5682 
5683         [... other code ...]
5684         call MatRestoreArray(mat,mat_array,i_mat,ierr)
5685 .ve
5686 
5687    See the Fortran chapter of the users manual and
5688    petsc/src/mat/examples/tests for details.
5689 
5690    Level: advanced
5691 
5692    Concepts: matrices^access array
5693 
5694 .seealso: MatRestoreArray(), MatGetArrayF90(), MatGetRowIJ()
5695 @*/
5696 PetscErrorCode PETSCMAT_DLLEXPORT MatGetArray(Mat mat,PetscScalar *v[])
5697 {
5698   PetscErrorCode ierr;
5699 
5700   PetscFunctionBegin;
5701   PetscValidHeaderSpecific(mat,MAT_CLASSID,1);
5702   PetscValidType(mat,1);
5703   PetscValidPointer(v,2);
5704   if (!mat->ops->getarray) SETERRQ1(PETSC_COMM_SELF,PETSC_ERR_SUP,"Mat type %s",((PetscObject)mat)->type_name);
5705   ierr = MatPreallocated(mat);CHKERRQ(ierr);
5706   ierr = (*mat->ops->getarray)(mat,v);CHKERRQ(ierr);
5707   CHKMEMQ;
5708   PetscFunctionReturn(0);
5709 }
5710 
5711 #undef __FUNCT__
5712 #define __FUNCT__ "MatRestoreArray"
5713 /*@C
5714    MatRestoreArray - Restores the matrix after MatGetArray() has been called.
5715 
5716    Not Collective
5717 
5718    Input Parameter:
5719 +  mat - the matrix
5720 -  v - the location of the values
5721 
5722    Fortran Note:
5723    This routine is used differently from Fortran, e.g.,
5724 .vb
5725         Mat         mat
5726         PetscScalar mat_array(1)
5727         PetscOffset i_mat
5728         PetscErrorCode ierr
5729         call MatGetArray(mat,mat_array,i_mat,ierr)
5730 
5731   C  Access first local entry in matrix; note that array is
5732   C  treated as one dimensional
5733         value = mat_array(i_mat + 1)
5734 
5735         [... other code ...]
5736         call MatRestoreArray(mat,mat_array,i_mat,ierr)
5737 .ve
5738 
5739    See the Fortran chapter of the users manual and
5740    petsc/src/mat/examples/tests for details
5741 
5742    Level: advanced
5743 
5744 .seealso: MatGetArray(), MatRestoreArrayF90()
5745 @*/
5746 PetscErrorCode PETSCMAT_DLLEXPORT MatRestoreArray(Mat mat,PetscScalar *v[])
5747 {
5748   PetscErrorCode ierr;
5749 
5750   PetscFunctionBegin;
5751   PetscValidHeaderSpecific(mat,MAT_CLASSID,1);
5752   PetscValidType(mat,1);
5753   PetscValidPointer(v,2);
5754 #if defined(PETSC_USE_DEBUG)
5755   CHKMEMQ;
5756 #endif
5757   if (!mat->ops->restorearray) SETERRQ1(PETSC_COMM_SELF,PETSC_ERR_SUP,"Mat type %s",((PetscObject)mat)->type_name);
5758   ierr = (*mat->ops->restorearray)(mat,v);CHKERRQ(ierr);
5759   ierr = PetscObjectStateIncrease((PetscObject)mat);CHKERRQ(ierr);
5760   PetscFunctionReturn(0);
5761 }
5762 
5763 #undef __FUNCT__
5764 #define __FUNCT__ "MatGetSubMatrices"
5765 /*@C
5766    MatGetSubMatrices - Extracts several submatrices from a matrix. If submat
5767    points to an array of valid matrices, they may be reused to store the new
5768    submatrices.
5769 
5770    Collective on Mat
5771 
5772    Input Parameters:
5773 +  mat - the matrix
5774 .  n   - the number of submatrixes to be extracted (on this processor, may be zero)
5775 .  irow, icol - index sets of rows and columns to extract
5776 -  scall - either MAT_INITIAL_MATRIX or MAT_REUSE_MATRIX
5777 
5778    Output Parameter:
5779 .  submat - the array of submatrices
5780 
5781    Notes:
5782    MatGetSubMatrices() can extract ONLY sequential submatrices
5783    (from both sequential and parallel matrices). Use MatGetSubMatrix()
5784    to extract a parallel submatrix.
5785 
5786    When extracting submatrices from a parallel matrix, each processor can
5787    form a different submatrix by setting the rows and columns of its
5788    individual index sets according to the local submatrix desired.
5789 
5790    When finished using the submatrices, the user should destroy
5791    them with MatDestroyMatrices().
5792 
5793    MAT_REUSE_MATRIX can only be used when the nonzero structure of the
5794    original matrix has not changed from that last call to MatGetSubMatrices().
5795 
5796    This routine creates the matrices in submat; you should NOT create them before
5797    calling it. It also allocates the array of matrix pointers submat.
5798 
5799    For BAIJ matrices the index sets must respect the block structure, that is if they
5800    request one row/column in a block, they must request all rows/columns that are in
5801    that block. For example, if the block size is 2 you cannot request just row 0 and
5802    column 0.
5803 
5804    Fortran Note:
5805    The Fortran interface is slightly different from that given below; it
5806    requires one to pass in  as submat a Mat (integer) array of size at least m.
5807 
5808    Level: advanced
5809 
5810    Concepts: matrices^accessing submatrices
5811    Concepts: submatrices
5812 
5813 .seealso: MatDestroyMatrices(), MatGetSubMatrix(), MatGetRow(), MatGetDiagonal(), MatReuse
5814 @*/
5815 PetscErrorCode PETSCMAT_DLLEXPORT MatGetSubMatrices(Mat mat,PetscInt n,const IS irow[],const IS icol[],MatReuse scall,Mat *submat[])
5816 {
5817   PetscErrorCode ierr;
5818   PetscInt        i;
5819   PetscTruth      eq;
5820 
5821   PetscFunctionBegin;
5822   PetscValidHeaderSpecific(mat,MAT_CLASSID,1);
5823   PetscValidType(mat,1);
5824   if (n) {
5825     PetscValidPointer(irow,3);
5826     PetscValidHeaderSpecific(*irow,IS_CLASSID,3);
5827     PetscValidPointer(icol,4);
5828     PetscValidHeaderSpecific(*icol,IS_CLASSID,4);
5829   }
5830   PetscValidPointer(submat,6);
5831   if (n && scall == MAT_REUSE_MATRIX) {
5832     PetscValidPointer(*submat,6);
5833     PetscValidHeaderSpecific(**submat,MAT_CLASSID,6);
5834   }
5835   if (!mat->ops->getsubmatrices) SETERRQ1(PETSC_COMM_SELF,PETSC_ERR_SUP,"Mat type %s",((PetscObject)mat)->type_name);
5836   if (!mat->assembled) SETERRQ(((PetscObject)mat)->comm,PETSC_ERR_ARG_WRONGSTATE,"Not for unassembled matrix");
5837   if (mat->factortype) SETERRQ(((PetscObject)mat)->comm,PETSC_ERR_ARG_WRONGSTATE,"Not for factored matrix");
5838   ierr = MatPreallocated(mat);CHKERRQ(ierr);
5839 
5840   ierr = PetscLogEventBegin(MAT_GetSubMatrices,mat,0,0,0);CHKERRQ(ierr);
5841   ierr = (*mat->ops->getsubmatrices)(mat,n,irow,icol,scall,submat);CHKERRQ(ierr);
5842   ierr = PetscLogEventEnd(MAT_GetSubMatrices,mat,0,0,0);CHKERRQ(ierr);
5843   for (i=0; i<n; i++) {
5844     if (mat->symmetric || mat->structurally_symmetric || mat->hermitian) {
5845       ierr = ISEqual(irow[i],icol[i],&eq);CHKERRQ(ierr);
5846       if (eq) {
5847 	if (mat->symmetric){
5848 	  ierr = MatSetOption((*submat)[i],MAT_SYMMETRIC,PETSC_TRUE);CHKERRQ(ierr);
5849 	} else if (mat->hermitian) {
5850 	  ierr = MatSetOption((*submat)[i],MAT_HERMITIAN,PETSC_TRUE);CHKERRQ(ierr);
5851 	} else if (mat->structurally_symmetric) {
5852 	  ierr = MatSetOption((*submat)[i],MAT_STRUCTURALLY_SYMMETRIC,PETSC_TRUE);CHKERRQ(ierr);
5853 	}
5854       }
5855     }
5856   }
5857   PetscFunctionReturn(0);
5858 }
5859 
5860 #undef __FUNCT__
5861 #define __FUNCT__ "MatDestroyMatrices"
5862 /*@C
5863    MatDestroyMatrices - Destroys a set of matrices obtained with MatGetSubMatrices().
5864 
5865    Collective on Mat
5866 
5867    Input Parameters:
5868 +  n - the number of local matrices
5869 -  mat - the matrices (note that this is a pointer to the array of matrices, just to match the calling
5870                        sequence of MatGetSubMatrices())
5871 
5872    Level: advanced
5873 
5874     Notes: Frees not only the matrices, but also the array that contains the matrices
5875            In Fortran will not free the array.
5876 
5877 .seealso: MatGetSubMatrices()
5878 @*/
5879 PetscErrorCode PETSCMAT_DLLEXPORT MatDestroyMatrices(PetscInt n,Mat *mat[])
5880 {
5881   PetscErrorCode ierr;
5882   PetscInt       i;
5883 
5884   PetscFunctionBegin;
5885   if (n < 0) SETERRQ1(PETSC_COMM_SELF,PETSC_ERR_ARG_OUTOFRANGE,"Trying to destroy negative number of matrices %D",n);
5886   PetscValidPointer(mat,2);
5887   for (i=0; i<n; i++) {
5888     ierr = MatDestroy((*mat)[i]);CHKERRQ(ierr);
5889   }
5890   /* memory is allocated even if n = 0 */
5891   ierr = PetscFree(*mat);CHKERRQ(ierr);
5892   PetscFunctionReturn(0);
5893 }
5894 
5895 #undef __FUNCT__
5896 #define __FUNCT__ "MatGetSeqNonzeroStructure"
5897 /*@C
5898    MatGetSeqNonzeroStructure - Extracts the sequential nonzero structure from a matrix.
5899 
5900    Collective on Mat
5901 
5902    Input Parameters:
5903 .  mat - the matrix
5904 
5905    Output Parameter:
5906 .  matstruct - the sequential matrix with the nonzero structure of mat
5907 
5908   Level: intermediate
5909 
5910 .seealso: MatDestroySeqNonzeroStructure(), MatGetSubMatrices(), MatDestroyMatrices()
5911 @*/
5912 PetscErrorCode PETSCMAT_DLLEXPORT MatGetSeqNonzeroStructure(Mat mat,Mat *matstruct)
5913 {
5914   PetscErrorCode ierr;
5915 
5916   PetscFunctionBegin;
5917   PetscValidHeaderSpecific(mat,MAT_CLASSID,1);
5918   PetscValidPointer(matstruct,2);
5919 
5920   PetscValidType(mat,1);
5921   if (mat->factortype) SETERRQ(((PetscObject)mat)->comm,PETSC_ERR_ARG_WRONGSTATE,"Not for factored matrix");
5922   ierr = MatPreallocated(mat);CHKERRQ(ierr);
5923 
5924   if (!mat->ops->getseqnonzerostructure) SETERRQ1(((PetscObject)mat)->comm,PETSC_ERR_SUP,"Not for matrix type %s\n",((PetscObject)mat)->type_name);
5925   ierr = PetscLogEventBegin(MAT_GetSeqNonzeroStructure,mat,0,0,0);CHKERRQ(ierr);
5926   ierr = (*mat->ops->getseqnonzerostructure)(mat,matstruct);CHKERRQ(ierr);
5927   ierr = PetscLogEventEnd(MAT_GetSeqNonzeroStructure,mat,0,0,0);CHKERRQ(ierr);
5928   PetscFunctionReturn(0);
5929 }
5930 
5931 #undef __FUNCT__
5932 #define __FUNCT__ "MatDestroySeqNonzeroStructure"
5933 /*@C
5934    MatDestroySeqNonzeroStructure - Destroys matrix obtained with MatGetSeqNonzeroStructure().
5935 
5936    Collective on Mat
5937 
5938    Input Parameters:
5939 .  mat - the matrix (note that this is a pointer to the array of matrices, just to match the calling
5940                        sequence of MatGetSequentialNonzeroStructure())
5941 
5942    Level: advanced
5943 
5944     Notes: Frees not only the matrices, but also the array that contains the matrices
5945 
5946 .seealso: MatGetSeqNonzeroStructure()
5947 @*/
5948 PetscErrorCode PETSCMAT_DLLEXPORT MatDestroySeqNonzeroStructure(Mat *mat)
5949 {
5950   PetscErrorCode ierr;
5951 
5952   PetscFunctionBegin;
5953   PetscValidPointer(mat,1);
5954   ierr = MatDestroy(*mat);CHKERRQ(ierr);
5955   PetscFunctionReturn(0);
5956 }
5957 
5958 #undef __FUNCT__
5959 #define __FUNCT__ "MatIncreaseOverlap"
5960 /*@
5961    MatIncreaseOverlap - Given a set of submatrices indicated by index sets,
5962    replaces the index sets by larger ones that represent submatrices with
5963    additional overlap.
5964 
5965    Collective on Mat
5966 
5967    Input Parameters:
5968 +  mat - the matrix
5969 .  n   - the number of index sets
5970 .  is  - the array of index sets (these index sets will changed during the call)
5971 -  ov  - the additional overlap requested
5972 
5973    Level: developer
5974 
5975    Concepts: overlap
5976    Concepts: ASM^computing overlap
5977 
5978 .seealso: MatGetSubMatrices()
5979 @*/
5980 PetscErrorCode PETSCMAT_DLLEXPORT MatIncreaseOverlap(Mat mat,PetscInt n,IS is[],PetscInt ov)
5981 {
5982   PetscErrorCode ierr;
5983 
5984   PetscFunctionBegin;
5985   PetscValidHeaderSpecific(mat,MAT_CLASSID,1);
5986   PetscValidType(mat,1);
5987   if (n < 0) SETERRQ1(PETSC_COMM_SELF,PETSC_ERR_ARG_OUTOFRANGE,"Must have one or more domains, you have %D",n);
5988   if (n) {
5989     PetscValidPointer(is,3);
5990     PetscValidHeaderSpecific(*is,IS_CLASSID,3);
5991   }
5992   if (!mat->assembled) SETERRQ(((PetscObject)mat)->comm,PETSC_ERR_ARG_WRONGSTATE,"Not for unassembled matrix");
5993   if (mat->factortype)     SETERRQ(((PetscObject)mat)->comm,PETSC_ERR_ARG_WRONGSTATE,"Not for factored matrix");
5994   ierr = MatPreallocated(mat);CHKERRQ(ierr);
5995 
5996   if (!ov) PetscFunctionReturn(0);
5997   if (!mat->ops->increaseoverlap) SETERRQ1(((PetscObject)mat)->comm,PETSC_ERR_SUP,"Mat type %s",((PetscObject)mat)->type_name);
5998   ierr = PetscLogEventBegin(MAT_IncreaseOverlap,mat,0,0,0);CHKERRQ(ierr);
5999   ierr = (*mat->ops->increaseoverlap)(mat,n,is,ov);CHKERRQ(ierr);
6000   ierr = PetscLogEventEnd(MAT_IncreaseOverlap,mat,0,0,0);CHKERRQ(ierr);
6001   PetscFunctionReturn(0);
6002 }
6003 
6004 #undef __FUNCT__
6005 #define __FUNCT__ "MatGetBlockSize"
6006 /*@
6007    MatGetBlockSize - Returns the matrix block size; useful especially for the
6008    block row and block diagonal formats.
6009 
6010    Not Collective
6011 
6012    Input Parameter:
6013 .  mat - the matrix
6014 
6015    Output Parameter:
6016 .  bs - block size
6017 
6018    Notes:
6019    Block row formats are MATSEQBAIJ, MATMPIBAIJ, MATSEQSBAIJ, MATMPISBAIJ
6020 
6021    Level: intermediate
6022 
6023    Concepts: matrices^block size
6024 
6025 .seealso: MatCreateSeqBAIJ(), MatCreateMPIBAIJ()
6026 @*/
6027 PetscErrorCode PETSCMAT_DLLEXPORT MatGetBlockSize(Mat mat,PetscInt *bs)
6028 {
6029   PetscErrorCode ierr;
6030 
6031   PetscFunctionBegin;
6032   PetscValidHeaderSpecific(mat,MAT_CLASSID,1);
6033   PetscValidType(mat,1);
6034   PetscValidIntPointer(bs,2);
6035   ierr = MatPreallocated(mat);CHKERRQ(ierr);
6036   *bs = mat->rmap->bs;
6037   PetscFunctionReturn(0);
6038 }
6039 
6040 #undef __FUNCT__
6041 #define __FUNCT__ "MatSetBlockSize"
6042 /*@
6043    MatSetBlockSize - Sets the matrix block size; for many matrix types you
6044      cannot use this and MUST set the blocksize when you preallocate the matrix
6045 
6046    Collective on Mat
6047 
6048    Input Parameters:
6049 +  mat - the matrix
6050 -  bs - block size
6051 
6052    Notes:
6053      For BAIJ matrices, this just checks that the block size agrees with the BAIJ size,
6054      it is not possible to change BAIJ block sizes after preallocation.
6055 
6056    Level: intermediate
6057 
6058    Concepts: matrices^block size
6059 
6060 .seealso: MatCreateSeqBAIJ(), MatCreateMPIBAIJ(), MatGetBlockSize()
6061 @*/
6062 PetscErrorCode PETSCMAT_DLLEXPORT MatSetBlockSize(Mat mat,PetscInt bs)
6063 {
6064   PetscErrorCode ierr;
6065 
6066   PetscFunctionBegin;
6067   PetscValidHeaderSpecific(mat,MAT_CLASSID,1);
6068   PetscValidType(mat,1);
6069   ierr = MatPreallocated(mat);CHKERRQ(ierr);
6070   if (bs < 1) SETERRQ1(PETSC_COMM_SELF,PETSC_ERR_ARG_OUTOFRANGE,"Block size %d, must be positive",bs);
6071   if (mat->ops->setblocksize) {
6072     ierr = (*mat->ops->setblocksize)(mat,bs);CHKERRQ(ierr);
6073   } else {
6074     SETERRQ1(PETSC_COMM_SELF,PETSC_ERR_ARG_INCOMP,"Cannot set the blocksize for matrix type %s",((PetscObject)mat)->type_name);
6075   }
6076   PetscFunctionReturn(0);
6077 }
6078 
6079 #undef __FUNCT__
6080 #define __FUNCT__ "MatGetRowIJ"
6081 /*@C
6082     MatGetRowIJ - Returns the compressed row storage i and j indices for sequential matrices.
6083 
6084    Collective on Mat
6085 
6086     Input Parameters:
6087 +   mat - the matrix
6088 .   shift -  0 or 1 indicating we want the indices starting at 0 or 1
6089 .   symmetric - PETSC_TRUE or PETSC_FALSE indicating the matrix data structure should be
6090                 symmetrized
6091 -   inodecompressed - PETSC_TRUE or PETSC_FALSE  indicating if the nonzero structure of the
6092                  inodes or the nonzero elements is wanted. For BAIJ matrices the compressed version is
6093                  always used.
6094 
6095     Output Parameters:
6096 +   n - number of rows in the (possibly compressed) matrix
6097 .   ia - the row pointers [of length n+1]
6098 .   ja - the column indices
6099 -   done - indicates if the routine actually worked and returned appropriate ia[] and ja[] arrays; callers
6100            are responsible for handling the case when done == PETSC_FALSE and ia and ja are not set
6101 
6102     Level: developer
6103 
6104     Notes: You CANNOT change any of the ia[] or ja[] values.
6105 
6106            Use MatRestoreRowIJ() when you are finished accessing the ia[] and ja[] values
6107 
6108     Fortran Node
6109 
6110            In Fortran use
6111 $           PetscInt ia(1), ja(1)
6112 $           PetscOffset iia, jja
6113 $      call MatGetRowIJ(mat,shift,symmetric,inodecompressed,n,ia,iia,ja,jja,done,ierr)
6114 $
6115 $          or
6116 $
6117 $           PetscScalar, pointer :: xx_v(:)
6118 $    call  MatGetRowIJF90(mat,shift,symmetric,inodecompressed,n,ia,ja,done,ierr)
6119 
6120 
6121        Acess the ith and jth entries via ia(iia + i) and ja(jja + j)
6122 
6123 .seealso: MatGetColumnIJ(), MatRestoreRowIJ(), MatGetArray()
6124 @*/
6125 PetscErrorCode PETSCMAT_DLLEXPORT MatGetRowIJ(Mat mat,PetscInt shift,PetscTruth symmetric,PetscTruth inodecompressed,PetscInt *n,PetscInt *ia[],PetscInt* ja[],PetscTruth *done)
6126 {
6127   PetscErrorCode ierr;
6128 
6129   PetscFunctionBegin;
6130   PetscValidHeaderSpecific(mat,MAT_CLASSID,1);
6131   PetscValidType(mat,1);
6132   PetscValidIntPointer(n,4);
6133   if (ia) PetscValidIntPointer(ia,5);
6134   if (ja) PetscValidIntPointer(ja,6);
6135   PetscValidIntPointer(done,7);
6136   ierr = MatPreallocated(mat);CHKERRQ(ierr);
6137   if (!mat->ops->getrowij) *done = PETSC_FALSE;
6138   else {
6139     *done = PETSC_TRUE;
6140     ierr = PetscLogEventBegin(MAT_GetRowIJ,mat,0,0,0);CHKERRQ(ierr);
6141     ierr  = (*mat->ops->getrowij)(mat,shift,symmetric,inodecompressed,n,ia,ja,done);CHKERRQ(ierr);
6142     ierr = PetscLogEventEnd(MAT_GetRowIJ,mat,0,0,0);CHKERRQ(ierr);
6143   }
6144   PetscFunctionReturn(0);
6145 }
6146 
6147 #undef __FUNCT__
6148 #define __FUNCT__ "MatGetColumnIJ"
6149 /*@C
6150     MatGetColumnIJ - Returns the compressed column storage i and j indices for sequential matrices.
6151 
6152     Collective on Mat
6153 
6154     Input Parameters:
6155 +   mat - the matrix
6156 .   shift - 1 or zero indicating we want the indices starting at 0 or 1
6157 .   symmetric - PETSC_TRUE or PETSC_FALSE indicating the matrix data structure should be
6158                 symmetrized
6159 -   inodecompressed - PETSC_TRUE or PETSC_FALSE indicating if the nonzero structure of the
6160                  inodes or the nonzero elements is wanted. For BAIJ matrices the compressed version is
6161                  always used.
6162 
6163     Output Parameters:
6164 +   n - number of columns in the (possibly compressed) matrix
6165 .   ia - the column pointers
6166 .   ja - the row indices
6167 -   done - PETSC_TRUE or PETSC_FALSE, indicating whether the values have been returned
6168 
6169     Level: developer
6170 
6171 .seealso: MatGetRowIJ(), MatRestoreColumnIJ()
6172 @*/
6173 PetscErrorCode PETSCMAT_DLLEXPORT MatGetColumnIJ(Mat mat,PetscInt shift,PetscTruth symmetric,PetscTruth inodecompressed,PetscInt *n,PetscInt *ia[],PetscInt* ja[],PetscTruth *done)
6174 {
6175   PetscErrorCode ierr;
6176 
6177   PetscFunctionBegin;
6178   PetscValidHeaderSpecific(mat,MAT_CLASSID,1);
6179   PetscValidType(mat,1);
6180   PetscValidIntPointer(n,4);
6181   if (ia) PetscValidIntPointer(ia,5);
6182   if (ja) PetscValidIntPointer(ja,6);
6183   PetscValidIntPointer(done,7);
6184   ierr = MatPreallocated(mat);CHKERRQ(ierr);
6185   if (!mat->ops->getcolumnij) *done = PETSC_FALSE;
6186   else {
6187     *done = PETSC_TRUE;
6188     ierr  = (*mat->ops->getcolumnij)(mat,shift,symmetric,inodecompressed,n,ia,ja,done);CHKERRQ(ierr);
6189   }
6190   PetscFunctionReturn(0);
6191 }
6192 
6193 #undef __FUNCT__
6194 #define __FUNCT__ "MatRestoreRowIJ"
6195 /*@C
6196     MatRestoreRowIJ - Call after you are completed with the ia,ja indices obtained with
6197     MatGetRowIJ().
6198 
6199     Collective on Mat
6200 
6201     Input Parameters:
6202 +   mat - the matrix
6203 .   shift - 1 or zero indicating we want the indices starting at 0 or 1
6204 .   symmetric - PETSC_TRUE or PETSC_FALSE indicating the matrix data structure should be
6205                 symmetrized
6206 -   inodecompressed -  PETSC_TRUE or PETSC_FALSE indicating if the nonzero structure of the
6207                  inodes or the nonzero elements is wanted. For BAIJ matrices the compressed version is
6208                  always used.
6209 
6210     Output Parameters:
6211 +   n - size of (possibly compressed) matrix
6212 .   ia - the row pointers
6213 .   ja - the column indices
6214 -   done - PETSC_TRUE or PETSC_FALSE indicated that the values have been returned
6215 
6216     Level: developer
6217 
6218 .seealso: MatGetRowIJ(), MatRestoreColumnIJ()
6219 @*/
6220 PetscErrorCode PETSCMAT_DLLEXPORT MatRestoreRowIJ(Mat mat,PetscInt shift,PetscTruth symmetric,PetscTruth inodecompressed,PetscInt *n,PetscInt *ia[],PetscInt* ja[],PetscTruth *done)
6221 {
6222   PetscErrorCode ierr;
6223 
6224   PetscFunctionBegin;
6225   PetscValidHeaderSpecific(mat,MAT_CLASSID,1);
6226   PetscValidType(mat,1);
6227   if (ia) PetscValidIntPointer(ia,5);
6228   if (ja) PetscValidIntPointer(ja,6);
6229   PetscValidIntPointer(done,7);
6230   ierr = MatPreallocated(mat);CHKERRQ(ierr);
6231 
6232   if (!mat->ops->restorerowij) *done = PETSC_FALSE;
6233   else {
6234     *done = PETSC_TRUE;
6235     ierr  = (*mat->ops->restorerowij)(mat,shift,symmetric,inodecompressed,n,ia,ja,done);CHKERRQ(ierr);
6236   }
6237   PetscFunctionReturn(0);
6238 }
6239 
6240 #undef __FUNCT__
6241 #define __FUNCT__ "MatRestoreColumnIJ"
6242 /*@C
6243     MatRestoreColumnIJ - Call after you are completed with the ia,ja indices obtained with
6244     MatGetColumnIJ().
6245 
6246     Collective on Mat
6247 
6248     Input Parameters:
6249 +   mat - the matrix
6250 .   shift - 1 or zero indicating we want the indices starting at 0 or 1
6251 -   symmetric - PETSC_TRUE or PETSC_FALSE indicating the matrix data structure should be
6252                 symmetrized
6253 -   inodecompressed - PETSC_TRUE or PETSC_FALSE indicating if the nonzero structure of the
6254                  inodes or the nonzero elements is wanted. For BAIJ matrices the compressed version is
6255                  always used.
6256 
6257     Output Parameters:
6258 +   n - size of (possibly compressed) matrix
6259 .   ia - the column pointers
6260 .   ja - the row indices
6261 -   done - PETSC_TRUE or PETSC_FALSE indicated that the values have been returned
6262 
6263     Level: developer
6264 
6265 .seealso: MatGetColumnIJ(), MatRestoreRowIJ()
6266 @*/
6267 PetscErrorCode PETSCMAT_DLLEXPORT MatRestoreColumnIJ(Mat mat,PetscInt shift,PetscTruth symmetric,PetscTruth inodecompressed,PetscInt *n,PetscInt *ia[],PetscInt* ja[],PetscTruth *done)
6268 {
6269   PetscErrorCode ierr;
6270 
6271   PetscFunctionBegin;
6272   PetscValidHeaderSpecific(mat,MAT_CLASSID,1);
6273   PetscValidType(mat,1);
6274   if (ia) PetscValidIntPointer(ia,5);
6275   if (ja) PetscValidIntPointer(ja,6);
6276   PetscValidIntPointer(done,7);
6277   ierr = MatPreallocated(mat);CHKERRQ(ierr);
6278 
6279   if (!mat->ops->restorecolumnij) *done = PETSC_FALSE;
6280   else {
6281     *done = PETSC_TRUE;
6282     ierr  = (*mat->ops->restorecolumnij)(mat,shift,symmetric,inodecompressed,n,ia,ja,done);CHKERRQ(ierr);
6283   }
6284   PetscFunctionReturn(0);
6285 }
6286 
6287 #undef __FUNCT__
6288 #define __FUNCT__ "MatColoringPatch"
6289 /*@C
6290     MatColoringPatch -Used inside matrix coloring routines that
6291     use MatGetRowIJ() and/or MatGetColumnIJ().
6292 
6293     Collective on Mat
6294 
6295     Input Parameters:
6296 +   mat - the matrix
6297 .   ncolors - max color value
6298 .   n   - number of entries in colorarray
6299 -   colorarray - array indicating color for each column
6300 
6301     Output Parameters:
6302 .   iscoloring - coloring generated using colorarray information
6303 
6304     Level: developer
6305 
6306 .seealso: MatGetRowIJ(), MatGetColumnIJ()
6307 
6308 @*/
6309 PetscErrorCode PETSCMAT_DLLEXPORT MatColoringPatch(Mat mat,PetscInt ncolors,PetscInt n,ISColoringValue colorarray[],ISColoring *iscoloring)
6310 {
6311   PetscErrorCode ierr;
6312 
6313   PetscFunctionBegin;
6314   PetscValidHeaderSpecific(mat,MAT_CLASSID,1);
6315   PetscValidType(mat,1);
6316   PetscValidIntPointer(colorarray,4);
6317   PetscValidPointer(iscoloring,5);
6318   ierr = MatPreallocated(mat);CHKERRQ(ierr);
6319 
6320   if (!mat->ops->coloringpatch){
6321     ierr = ISColoringCreate(((PetscObject)mat)->comm,ncolors,n,colorarray,iscoloring);CHKERRQ(ierr);
6322   } else {
6323     ierr = (*mat->ops->coloringpatch)(mat,ncolors,n,colorarray,iscoloring);CHKERRQ(ierr);
6324   }
6325   PetscFunctionReturn(0);
6326 }
6327 
6328 
6329 #undef __FUNCT__
6330 #define __FUNCT__ "MatSetUnfactored"
6331 /*@
6332    MatSetUnfactored - Resets a factored matrix to be treated as unfactored.
6333 
6334    Collective on Mat
6335 
6336    Input Parameter:
6337 .  mat - the factored matrix to be reset
6338 
6339    Notes:
6340    This routine should be used only with factored matrices formed by in-place
6341    factorization via ILU(0) (or by in-place LU factorization for the MATSEQDENSE
6342    format).  This option can save memory, for example, when solving nonlinear
6343    systems with a matrix-free Newton-Krylov method and a matrix-based, in-place
6344    ILU(0) preconditioner.
6345 
6346    Note that one can specify in-place ILU(0) factorization by calling
6347 .vb
6348      PCType(pc,PCILU);
6349      PCFactorSeUseInPlace(pc);
6350 .ve
6351    or by using the options -pc_type ilu -pc_factor_in_place
6352 
6353    In-place factorization ILU(0) can also be used as a local
6354    solver for the blocks within the block Jacobi or additive Schwarz
6355    methods (runtime option: -sub_pc_factor_in_place).  See the discussion
6356    of these preconditioners in the users manual for details on setting
6357    local solver options.
6358 
6359    Most users should employ the simplified KSP interface for linear solvers
6360    instead of working directly with matrix algebra routines such as this.
6361    See, e.g., KSPCreate().
6362 
6363    Level: developer
6364 
6365 .seealso: PCFactorSetUseInPlace()
6366 
6367    Concepts: matrices^unfactored
6368 
6369 @*/
6370 PetscErrorCode PETSCMAT_DLLEXPORT MatSetUnfactored(Mat mat)
6371 {
6372   PetscErrorCode ierr;
6373 
6374   PetscFunctionBegin;
6375   PetscValidHeaderSpecific(mat,MAT_CLASSID,1);
6376   PetscValidType(mat,1);
6377   ierr = MatPreallocated(mat);CHKERRQ(ierr);
6378   mat->factortype = MAT_FACTOR_NONE;
6379   if (!mat->ops->setunfactored) PetscFunctionReturn(0);
6380   ierr = (*mat->ops->setunfactored)(mat);CHKERRQ(ierr);
6381   PetscFunctionReturn(0);
6382 }
6383 
6384 /*MC
6385     MatGetArrayF90 - Accesses a matrix array from Fortran90.
6386 
6387     Synopsis:
6388     MatGetArrayF90(Mat x,{Scalar, pointer :: xx_v(:)},integer ierr)
6389 
6390     Not collective
6391 
6392     Input Parameter:
6393 .   x - matrix
6394 
6395     Output Parameters:
6396 +   xx_v - the Fortran90 pointer to the array
6397 -   ierr - error code
6398 
6399     Example of Usage:
6400 .vb
6401       PetscScalar, pointer xx_v(:)
6402       ....
6403       call MatGetArrayF90(x,xx_v,ierr)
6404       a = xx_v(3)
6405       call MatRestoreArrayF90(x,xx_v,ierr)
6406 .ve
6407 
6408     Notes:
6409     Not yet supported for all F90 compilers
6410 
6411     Level: advanced
6412 
6413 .seealso:  MatRestoreArrayF90(), MatGetArray(), MatRestoreArray()
6414 
6415     Concepts: matrices^accessing array
6416 
6417 M*/
6418 
6419 /*MC
6420     MatRestoreArrayF90 - Restores a matrix array that has been
6421     accessed with MatGetArrayF90().
6422 
6423     Synopsis:
6424     MatRestoreArrayF90(Mat x,{Scalar, pointer :: xx_v(:)},integer ierr)
6425 
6426     Not collective
6427 
6428     Input Parameters:
6429 +   x - matrix
6430 -   xx_v - the Fortran90 pointer to the array
6431 
6432     Output Parameter:
6433 .   ierr - error code
6434 
6435     Example of Usage:
6436 .vb
6437        PetscScalar, pointer xx_v(:)
6438        ....
6439        call MatGetArrayF90(x,xx_v,ierr)
6440        a = xx_v(3)
6441        call MatRestoreArrayF90(x,xx_v,ierr)
6442 .ve
6443 
6444     Notes:
6445     Not yet supported for all F90 compilers
6446 
6447     Level: advanced
6448 
6449 .seealso:  MatGetArrayF90(), MatGetArray(), MatRestoreArray()
6450 
6451 M*/
6452 
6453 
6454 #undef __FUNCT__
6455 #define __FUNCT__ "MatGetSubMatrix"
6456 /*@
6457     MatGetSubMatrix - Gets a single submatrix on the same number of processors
6458                       as the original matrix.
6459 
6460     Collective on Mat
6461 
6462     Input Parameters:
6463 +   mat - the original matrix
6464 .   isrow - parallel IS containing the rows this processor should obtain
6465 .   iscol - parallel IS containing all columns you wish to keep. Each process should list the columns that will be in IT's "diagonal part" in the new matrix.
6466 -   cll - either MAT_INITIAL_MATRIX or MAT_REUSE_MATRIX
6467 
6468     Output Parameter:
6469 .   newmat - the new submatrix, of the same type as the old
6470 
6471     Level: advanced
6472 
6473     Notes:
6474     The submatrix will be able to be multiplied with vectors using the same layout as iscol.
6475 
6476     The rows in isrow will be sorted into the same order as the original matrix on each process.
6477 
6478       The first time this is called you should use a cll of MAT_INITIAL_MATRIX,
6479    the MatGetSubMatrix() routine will create the newmat for you. Any additional calls
6480    to this routine with a mat of the same nonzero structure and with a call of MAT_REUSE_MATRIX
6481    will reuse the matrix generated the first time.  You should call MatDestroy() on newmat when
6482    you are finished using it.
6483 
6484     The communicator of the newly obtained matrix is ALWAYS the same as the communicator of
6485     the input matrix.
6486 
6487     If iscol is PETSC_NULL then all columns are obtained (not supported in Fortran).
6488 
6489    Example usage:
6490    Consider the following 8x8 matrix with 34 non-zero values, that is
6491    assembled across 3 processors. Lets assume that proc0 owns 3 rows,
6492    proc1 owns 3 rows, proc2 owns 2 rows. This division can be shown
6493    as follows:
6494 
6495 .vb
6496             1  2  0  |  0  3  0  |  0  4
6497     Proc0   0  5  6  |  7  0  0  |  8  0
6498             9  0 10  | 11  0  0  | 12  0
6499     -------------------------------------
6500            13  0 14  | 15 16 17  |  0  0
6501     Proc1   0 18  0  | 19 20 21  |  0  0
6502             0  0  0  | 22 23  0  | 24  0
6503     -------------------------------------
6504     Proc2  25 26 27  |  0  0 28  | 29  0
6505            30  0  0  | 31 32 33  |  0 34
6506 .ve
6507 
6508     Suppose isrow = [0 1 | 4 | 6 7] and iscol = [1 2 | 3 4 5 | 6].  The resulting submatrix is
6509 
6510 .vb
6511             2  0  |  0  3  0  |  0
6512     Proc0   5  6  |  7  0  0  |  8
6513     -------------------------------
6514     Proc1  18  0  | 19 20 21  |  0
6515     -------------------------------
6516     Proc2  26 27  |  0  0 28  | 29
6517             0  0  | 31 32 33  |  0
6518 .ve
6519 
6520 
6521     Concepts: matrices^submatrices
6522 
6523 .seealso: MatGetSubMatrices()
6524 @*/
6525 PetscErrorCode PETSCMAT_DLLEXPORT MatGetSubMatrix(Mat mat,IS isrow,IS iscol,MatReuse cll,Mat *newmat)
6526 {
6527   PetscErrorCode ierr;
6528   PetscMPIInt    size;
6529   Mat            *local;
6530   IS             iscoltmp;
6531 
6532   PetscFunctionBegin;
6533   PetscValidHeaderSpecific(mat,MAT_CLASSID,1);
6534   PetscValidHeaderSpecific(isrow,IS_CLASSID,2);
6535   if (iscol) PetscValidHeaderSpecific(iscol,IS_CLASSID,3);
6536   PetscValidPointer(newmat,5);
6537   if (cll == MAT_REUSE_MATRIX) PetscValidHeaderSpecific(*newmat,MAT_CLASSID,5);
6538   PetscValidType(mat,1);
6539   if (mat->factortype) SETERRQ(((PetscObject)mat)->comm,PETSC_ERR_ARG_WRONGSTATE,"Not for factored matrix");
6540   ierr = MatPreallocated(mat);CHKERRQ(ierr);
6541   ierr = MPI_Comm_size(((PetscObject)mat)->comm,&size);CHKERRQ(ierr);
6542 
6543   if (!iscol) {
6544     ierr = ISCreateStride(((PetscObject)mat)->comm,mat->cmap->n,mat->cmap->rstart,1,&iscoltmp);CHKERRQ(ierr);
6545   } else {
6546     iscoltmp = iscol;
6547   }
6548 
6549   /* if original matrix is on just one processor then use submatrix generated */
6550   if (mat->ops->getsubmatrices && !mat->ops->getsubmatrix && size == 1 && cll == MAT_REUSE_MATRIX) {
6551     ierr = MatGetSubMatrices(mat,1,&isrow,&iscoltmp,MAT_REUSE_MATRIX,&newmat);CHKERRQ(ierr);
6552     if (!iscol) {ierr = ISDestroy(iscoltmp);CHKERRQ(ierr);}
6553     PetscFunctionReturn(0);
6554   } else if (mat->ops->getsubmatrices && !mat->ops->getsubmatrix && size == 1) {
6555     ierr    = MatGetSubMatrices(mat,1,&isrow,&iscoltmp,MAT_INITIAL_MATRIX,&local);CHKERRQ(ierr);
6556     *newmat = *local;
6557     ierr    = PetscFree(local);CHKERRQ(ierr);
6558     if (!iscol) {ierr = ISDestroy(iscoltmp);CHKERRQ(ierr);}
6559     PetscFunctionReturn(0);
6560   } else if (!mat->ops->getsubmatrix) {
6561     /* Create a new matrix type that implements the operation using the full matrix */
6562     switch (cll) {
6563       case MAT_INITIAL_MATRIX:
6564         ierr = MatCreateSubMatrix(mat,isrow,iscoltmp,newmat);CHKERRQ(ierr);
6565         break;
6566       case MAT_REUSE_MATRIX:
6567         ierr = MatSubMatrixUpdate(*newmat,mat,isrow,iscoltmp);CHKERRQ(ierr);
6568         break;
6569       default: SETERRQ(((PetscObject)mat)->comm,PETSC_ERR_ARG_OUTOFRANGE,"Invalid MatReuse, must be either MAT_INITIAL_MATRIX or MAT_REUSE_MATRIX");
6570     }
6571     if (!iscol) {ierr = ISDestroy(iscoltmp);CHKERRQ(ierr);}
6572     PetscFunctionReturn(0);
6573   }
6574 
6575   if (!mat->ops->getsubmatrix) SETERRQ1(PETSC_COMM_SELF,PETSC_ERR_SUP,"Mat type %s",((PetscObject)mat)->type_name);
6576   ierr = (*mat->ops->getsubmatrix)(mat,isrow,iscoltmp,cll,newmat);CHKERRQ(ierr);
6577   if (!iscol) {ierr = ISDestroy(iscoltmp);CHKERRQ(ierr);}
6578   ierr = PetscObjectStateIncrease((PetscObject)*newmat);CHKERRQ(ierr);
6579   PetscFunctionReturn(0);
6580 }
6581 
6582 #undef __FUNCT__
6583 #define __FUNCT__ "MatStashSetInitialSize"
6584 /*@
6585    MatStashSetInitialSize - sets the sizes of the matrix stash, that is
6586    used during the assembly process to store values that belong to
6587    other processors.
6588 
6589    Not Collective
6590 
6591    Input Parameters:
6592 +  mat   - the matrix
6593 .  size  - the initial size of the stash.
6594 -  bsize - the initial size of the block-stash(if used).
6595 
6596    Options Database Keys:
6597 +   -matstash_initial_size <size> or <size0,size1,...sizep-1>
6598 -   -matstash_block_initial_size <bsize>  or <bsize0,bsize1,...bsizep-1>
6599 
6600    Level: intermediate
6601 
6602    Notes:
6603      The block-stash is used for values set with MatSetValuesBlocked() while
6604      the stash is used for values set with MatSetValues()
6605 
6606      Run with the option -info and look for output of the form
6607      MatAssemblyBegin_MPIXXX:Stash has MM entries, uses nn mallocs.
6608      to determine the appropriate value, MM, to use for size and
6609      MatAssemblyBegin_MPIXXX:Block-Stash has BMM entries, uses nn mallocs.
6610      to determine the value, BMM to use for bsize
6611 
6612    Concepts: stash^setting matrix size
6613    Concepts: matrices^stash
6614 
6615 .seealso: MatAssemblyBegin(), MatAssemblyEnd(), Mat, MatStashGetInfo()
6616 
6617 @*/
6618 PetscErrorCode PETSCMAT_DLLEXPORT MatStashSetInitialSize(Mat mat,PetscInt size, PetscInt bsize)
6619 {
6620   PetscErrorCode ierr;
6621 
6622   PetscFunctionBegin;
6623   PetscValidHeaderSpecific(mat,MAT_CLASSID,1);
6624   PetscValidType(mat,1);
6625   ierr = MatStashSetInitialSize_Private(&mat->stash,size);CHKERRQ(ierr);
6626   ierr = MatStashSetInitialSize_Private(&mat->bstash,bsize);CHKERRQ(ierr);
6627   PetscFunctionReturn(0);
6628 }
6629 
6630 #undef __FUNCT__
6631 #define __FUNCT__ "MatInterpolateAdd"
6632 /*@
6633    MatInterpolateAdd - w = y + A*x or A'*x depending on the shape of
6634      the matrix
6635 
6636    Collective on Mat
6637 
6638    Input Parameters:
6639 +  mat   - the matrix
6640 .  x,y - the vectors
6641 -  w - where the result is stored
6642 
6643    Level: intermediate
6644 
6645    Notes:
6646     w may be the same vector as y.
6647 
6648     This allows one to use either the restriction or interpolation (its transpose)
6649     matrix to do the interpolation
6650 
6651     Concepts: interpolation
6652 
6653 .seealso: MatMultAdd(), MatMultTransposeAdd(), MatRestrict()
6654 
6655 @*/
6656 PetscErrorCode PETSCMAT_DLLEXPORT MatInterpolateAdd(Mat A,Vec x,Vec y,Vec w)
6657 {
6658   PetscErrorCode ierr;
6659   PetscInt       M,N;
6660 
6661   PetscFunctionBegin;
6662   PetscValidHeaderSpecific(A,MAT_CLASSID,1);
6663   PetscValidHeaderSpecific(x,VEC_CLASSID,2);
6664   PetscValidHeaderSpecific(y,VEC_CLASSID,3);
6665   PetscValidHeaderSpecific(w,VEC_CLASSID,4);
6666   PetscValidType(A,1);
6667   ierr = MatPreallocated(A);CHKERRQ(ierr);
6668   ierr = MatGetSize(A,&M,&N);CHKERRQ(ierr);
6669   if (N > M) {
6670     ierr = MatMultTransposeAdd(A,x,y,w);CHKERRQ(ierr);
6671   } else {
6672     ierr = MatMultAdd(A,x,y,w);CHKERRQ(ierr);
6673   }
6674   PetscFunctionReturn(0);
6675 }
6676 
6677 #undef __FUNCT__
6678 #define __FUNCT__ "MatInterpolate"
6679 /*@
6680    MatInterpolate - y = A*x or A'*x depending on the shape of
6681      the matrix
6682 
6683    Collective on Mat
6684 
6685    Input Parameters:
6686 +  mat   - the matrix
6687 -  x,y - the vectors
6688 
6689    Level: intermediate
6690 
6691    Notes:
6692     This allows one to use either the restriction or interpolation (its transpose)
6693     matrix to do the interpolation
6694 
6695    Concepts: matrices^interpolation
6696 
6697 .seealso: MatMultAdd(), MatMultTransposeAdd(), MatRestrict()
6698 
6699 @*/
6700 PetscErrorCode PETSCMAT_DLLEXPORT MatInterpolate(Mat A,Vec x,Vec y)
6701 {
6702   PetscErrorCode ierr;
6703   PetscInt       M,N;
6704 
6705   PetscFunctionBegin;
6706   PetscValidHeaderSpecific(A,MAT_CLASSID,1);
6707   PetscValidHeaderSpecific(x,VEC_CLASSID,2);
6708   PetscValidHeaderSpecific(y,VEC_CLASSID,3);
6709   PetscValidType(A,1);
6710   ierr = MatPreallocated(A);CHKERRQ(ierr);
6711   ierr = MatGetSize(A,&M,&N);CHKERRQ(ierr);
6712   if (N > M) {
6713     ierr = MatMultTranspose(A,x,y);CHKERRQ(ierr);
6714   } else {
6715     ierr = MatMult(A,x,y);CHKERRQ(ierr);
6716   }
6717   PetscFunctionReturn(0);
6718 }
6719 
6720 #undef __FUNCT__
6721 #define __FUNCT__ "MatRestrict"
6722 /*@
6723    MatRestrict - y = A*x or A'*x
6724 
6725    Collective on Mat
6726 
6727    Input Parameters:
6728 +  mat   - the matrix
6729 -  x,y - the vectors
6730 
6731    Level: intermediate
6732 
6733    Notes:
6734     This allows one to use either the restriction or interpolation (its transpose)
6735     matrix to do the restriction
6736 
6737    Concepts: matrices^restriction
6738 
6739 .seealso: MatMultAdd(), MatMultTransposeAdd(), MatInterpolate()
6740 
6741 @*/
6742 PetscErrorCode PETSCMAT_DLLEXPORT MatRestrict(Mat A,Vec x,Vec y)
6743 {
6744   PetscErrorCode ierr;
6745   PetscInt       M,N;
6746 
6747   PetscFunctionBegin;
6748   PetscValidHeaderSpecific(A,MAT_CLASSID,1);
6749   PetscValidHeaderSpecific(x,VEC_CLASSID,2);
6750   PetscValidHeaderSpecific(y,VEC_CLASSID,3);
6751   PetscValidType(A,1);
6752   ierr = MatPreallocated(A);CHKERRQ(ierr);
6753 
6754   ierr = MatGetSize(A,&M,&N);CHKERRQ(ierr);
6755   if (N > M) {
6756     ierr = MatMult(A,x,y);CHKERRQ(ierr);
6757   } else {
6758     ierr = MatMultTranspose(A,x,y);CHKERRQ(ierr);
6759   }
6760   PetscFunctionReturn(0);
6761 }
6762 
6763 #undef __FUNCT__
6764 #define __FUNCT__ "MatNullSpaceAttach"
6765 /*@
6766    MatNullSpaceAttach - attaches a null space to a matrix.
6767         This null space will be removed from the resulting vector whenever
6768         MatMult() is called
6769 
6770    Collective on Mat
6771 
6772    Input Parameters:
6773 +  mat - the matrix
6774 -  nullsp - the null space object
6775 
6776    Level: developer
6777 
6778    Notes:
6779       Overwrites any previous null space that may have been attached
6780 
6781    Concepts: null space^attaching to matrix
6782 
6783 .seealso: MatCreate(), MatNullSpaceCreate()
6784 @*/
6785 PetscErrorCode PETSCMAT_DLLEXPORT MatNullSpaceAttach(Mat mat,MatNullSpace nullsp)
6786 {
6787   PetscErrorCode ierr;
6788 
6789   PetscFunctionBegin;
6790   PetscValidHeaderSpecific(mat,MAT_CLASSID,1);
6791   PetscValidType(mat,1);
6792   PetscValidHeaderSpecific(nullsp,MAT_NULLSPACE_CLASSID,2);
6793   ierr = MatPreallocated(mat);CHKERRQ(ierr);
6794   ierr = PetscObjectReference((PetscObject)nullsp);CHKERRQ(ierr);
6795   if (mat->nullsp) { ierr = MatNullSpaceDestroy(mat->nullsp);CHKERRQ(ierr); }
6796   mat->nullsp = nullsp;
6797   PetscFunctionReturn(0);
6798 }
6799 
6800 #undef __FUNCT__
6801 #define __FUNCT__ "MatICCFactor"
6802 /*@C
6803    MatICCFactor - Performs in-place incomplete Cholesky factorization of matrix.
6804 
6805    Collective on Mat
6806 
6807    Input Parameters:
6808 +  mat - the matrix
6809 .  row - row/column permutation
6810 .  fill - expected fill factor >= 1.0
6811 -  level - level of fill, for ICC(k)
6812 
6813    Notes:
6814    Probably really in-place only when level of fill is zero, otherwise allocates
6815    new space to store factored matrix and deletes previous memory.
6816 
6817    Most users should employ the simplified KSP interface for linear solvers
6818    instead of working directly with matrix algebra routines such as this.
6819    See, e.g., KSPCreate().
6820 
6821    Level: developer
6822 
6823    Concepts: matrices^incomplete Cholesky factorization
6824    Concepts: Cholesky factorization
6825 
6826 .seealso: MatICCFactorSymbolic(), MatLUFactorNumeric(), MatCholeskyFactor()
6827 
6828     Developer Note: fortran interface is not autogenerated as the f90
6829     interface defintion cannot be generated correctly [due to MatFactorInfo]
6830 
6831 @*/
6832 PetscErrorCode PETSCMAT_DLLEXPORT MatICCFactor(Mat mat,IS row,const MatFactorInfo* info)
6833 {
6834   PetscErrorCode ierr;
6835 
6836   PetscFunctionBegin;
6837   PetscValidHeaderSpecific(mat,MAT_CLASSID,1);
6838   PetscValidType(mat,1);
6839   if (row) PetscValidHeaderSpecific(row,IS_CLASSID,2);
6840   PetscValidPointer(info,3);
6841   if (mat->rmap->N != mat->cmap->N) SETERRQ(((PetscObject)mat)->comm,PETSC_ERR_ARG_WRONG,"matrix must be square");
6842   if (!mat->assembled) SETERRQ(((PetscObject)mat)->comm,PETSC_ERR_ARG_WRONGSTATE,"Not for unassembled matrix");
6843   if (mat->factortype) SETERRQ(((PetscObject)mat)->comm,PETSC_ERR_ARG_WRONGSTATE,"Not for factored matrix");
6844   if (!mat->ops->iccfactor) SETERRQ1(((PetscObject)mat)->comm,PETSC_ERR_SUP,"Mat type %s",((PetscObject)mat)->type_name);
6845   ierr = MatPreallocated(mat);CHKERRQ(ierr);
6846   ierr = (*mat->ops->iccfactor)(mat,row,info);CHKERRQ(ierr);
6847   ierr = PetscObjectStateIncrease((PetscObject)mat);CHKERRQ(ierr);
6848   PetscFunctionReturn(0);
6849 }
6850 
6851 #undef __FUNCT__
6852 #define __FUNCT__ "MatSetValuesAdic"
6853 /*@
6854    MatSetValuesAdic - Sets values computed with ADIC automatic differentiation into a matrix.
6855 
6856    Not Collective
6857 
6858    Input Parameters:
6859 +  mat - the matrix
6860 -  v - the values compute with ADIC
6861 
6862    Level: developer
6863 
6864    Notes:
6865      Must call MatSetColoring() before using this routine. Also this matrix must already
6866      have its nonzero pattern determined.
6867 
6868 .seealso: MatSetOption(), MatAssemblyBegin(), MatAssemblyEnd(), MatSetValuesBlocked(), MatSetValuesLocal(),
6869           MatSetValues(), MatSetColoring(), MatSetValuesAdifor()
6870 @*/
6871 PetscErrorCode PETSCMAT_DLLEXPORT MatSetValuesAdic(Mat mat,void *v)
6872 {
6873   PetscErrorCode ierr;
6874 
6875   PetscFunctionBegin;
6876   PetscValidHeaderSpecific(mat,MAT_CLASSID,1);
6877   PetscValidType(mat,1);
6878   PetscValidPointer(mat,2);
6879 
6880   if (!mat->assembled) {
6881     SETERRQ(PETSC_COMM_SELF,PETSC_ERR_ARG_WRONGSTATE,"Matrix must be already assembled");
6882   }
6883   ierr = PetscLogEventBegin(MAT_SetValues,mat,0,0,0);CHKERRQ(ierr);
6884   if (!mat->ops->setvaluesadic) SETERRQ1(PETSC_COMM_SELF,PETSC_ERR_SUP,"Mat type %s",((PetscObject)mat)->type_name);
6885   ierr = (*mat->ops->setvaluesadic)(mat,v);CHKERRQ(ierr);
6886   ierr = PetscLogEventEnd(MAT_SetValues,mat,0,0,0);CHKERRQ(ierr);
6887   ierr = MatView_Private(mat);CHKERRQ(ierr);
6888   ierr = PetscObjectStateIncrease((PetscObject)mat);CHKERRQ(ierr);
6889   PetscFunctionReturn(0);
6890 }
6891 
6892 
6893 #undef __FUNCT__
6894 #define __FUNCT__ "MatSetColoring"
6895 /*@
6896    MatSetColoring - Sets a coloring used by calls to MatSetValuesAdic()
6897 
6898    Not Collective
6899 
6900    Input Parameters:
6901 +  mat - the matrix
6902 -  coloring - the coloring
6903 
6904    Level: developer
6905 
6906 .seealso: MatSetOption(), MatAssemblyBegin(), MatAssemblyEnd(), MatSetValuesBlocked(), MatSetValuesLocal(),
6907           MatSetValues(), MatSetValuesAdic()
6908 @*/
6909 PetscErrorCode PETSCMAT_DLLEXPORT MatSetColoring(Mat mat,ISColoring coloring)
6910 {
6911   PetscErrorCode ierr;
6912 
6913   PetscFunctionBegin;
6914   PetscValidHeaderSpecific(mat,MAT_CLASSID,1);
6915   PetscValidType(mat,1);
6916   PetscValidPointer(coloring,2);
6917 
6918   if (!mat->assembled) SETERRQ(((PetscObject)mat)->comm,PETSC_ERR_ARG_WRONGSTATE,"Matrix must be already assembled");
6919   if (!mat->ops->setcoloring) SETERRQ1(((PetscObject)mat)->comm,PETSC_ERR_SUP,"Mat type %s",((PetscObject)mat)->type_name);
6920   ierr = (*mat->ops->setcoloring)(mat,coloring);CHKERRQ(ierr);
6921   PetscFunctionReturn(0);
6922 }
6923 
6924 #undef __FUNCT__
6925 #define __FUNCT__ "MatSetValuesAdifor"
6926 /*@
6927    MatSetValuesAdifor - Sets values computed with automatic differentiation into a matrix.
6928 
6929    Not Collective
6930 
6931    Input Parameters:
6932 +  mat - the matrix
6933 .  nl - leading dimension of v
6934 -  v - the values compute with ADIFOR
6935 
6936    Level: developer
6937 
6938    Notes:
6939      Must call MatSetColoring() before using this routine. Also this matrix must already
6940      have its nonzero pattern determined.
6941 
6942 .seealso: MatSetOption(), MatAssemblyBegin(), MatAssemblyEnd(), MatSetValuesBlocked(), MatSetValuesLocal(),
6943           MatSetValues(), MatSetColoring()
6944 @*/
6945 PetscErrorCode PETSCMAT_DLLEXPORT MatSetValuesAdifor(Mat mat,PetscInt nl,void *v)
6946 {
6947   PetscErrorCode ierr;
6948 
6949   PetscFunctionBegin;
6950   PetscValidHeaderSpecific(mat,MAT_CLASSID,1);
6951   PetscValidType(mat,1);
6952   PetscValidPointer(v,3);
6953 
6954   if (!mat->assembled) SETERRQ(((PetscObject)mat)->comm,PETSC_ERR_ARG_WRONGSTATE,"Matrix must be already assembled");
6955   ierr = PetscLogEventBegin(MAT_SetValues,mat,0,0,0);CHKERRQ(ierr);
6956   if (!mat->ops->setvaluesadifor) SETERRQ1(((PetscObject)mat)->comm,PETSC_ERR_SUP,"Mat type %s",((PetscObject)mat)->type_name);
6957   ierr = (*mat->ops->setvaluesadifor)(mat,nl,v);CHKERRQ(ierr);
6958   ierr = PetscLogEventEnd(MAT_SetValues,mat,0,0,0);CHKERRQ(ierr);
6959   ierr = PetscObjectStateIncrease((PetscObject)mat);CHKERRQ(ierr);
6960   PetscFunctionReturn(0);
6961 }
6962 
6963 #undef __FUNCT__
6964 #define __FUNCT__ "MatDiagonalScaleLocal"
6965 /*@
6966    MatDiagonalScaleLocal - Scales columns of a matrix given the scaling values including the
6967          ghosted ones.
6968 
6969    Not Collective
6970 
6971    Input Parameters:
6972 +  mat - the matrix
6973 -  diag = the diagonal values, including ghost ones
6974 
6975    Level: developer
6976 
6977    Notes: Works only for MPIAIJ and MPIBAIJ matrices
6978 
6979 .seealso: MatDiagonalScale()
6980 @*/
6981 PetscErrorCode PETSCMAT_DLLEXPORT MatDiagonalScaleLocal(Mat mat,Vec diag)
6982 {
6983   PetscErrorCode ierr;
6984   PetscMPIInt    size;
6985 
6986   PetscFunctionBegin;
6987   PetscValidHeaderSpecific(mat,MAT_CLASSID,1);
6988   PetscValidHeaderSpecific(diag,VEC_CLASSID,2);
6989   PetscValidType(mat,1);
6990 
6991   if (!mat->assembled) SETERRQ(((PetscObject)mat)->comm,PETSC_ERR_ARG_WRONGSTATE,"Matrix must be already assembled");
6992   ierr = PetscLogEventBegin(MAT_Scale,mat,0,0,0);CHKERRQ(ierr);
6993   ierr = MPI_Comm_size(((PetscObject)mat)->comm,&size);CHKERRQ(ierr);
6994   if (size == 1) {
6995     PetscInt n,m;
6996     ierr = VecGetSize(diag,&n);CHKERRQ(ierr);
6997     ierr = MatGetSize(mat,0,&m);CHKERRQ(ierr);
6998     if (m == n) {
6999       ierr = MatDiagonalScale(mat,0,diag);CHKERRQ(ierr);
7000     } else {
7001       SETERRQ(PETSC_COMM_SELF,PETSC_ERR_SUP,"Only supported for sequential matrices when no ghost points/periodic conditions");
7002     }
7003   } else {
7004     PetscErrorCode (*f)(Mat,Vec);
7005     ierr = PetscObjectQueryFunction((PetscObject)mat,"MatDiagonalScaleLocal_C",(void (**)(void))&f);CHKERRQ(ierr);
7006     if (f) {
7007       ierr = (*f)(mat,diag);CHKERRQ(ierr);
7008     } else {
7009       SETERRQ(PETSC_COMM_SELF,PETSC_ERR_SUP,"Only supported for MPIAIJ and MPIBAIJ parallel matrices");
7010     }
7011   }
7012   ierr = PetscLogEventEnd(MAT_Scale,mat,0,0,0);CHKERRQ(ierr);
7013   ierr = PetscObjectStateIncrease((PetscObject)mat);CHKERRQ(ierr);
7014   PetscFunctionReturn(0);
7015 }
7016 
7017 #undef __FUNCT__
7018 #define __FUNCT__ "MatGetInertia"
7019 /*@
7020    MatGetInertia - Gets the inertia from a factored matrix
7021 
7022    Collective on Mat
7023 
7024    Input Parameter:
7025 .  mat - the matrix
7026 
7027    Output Parameters:
7028 +   nneg - number of negative eigenvalues
7029 .   nzero - number of zero eigenvalues
7030 -   npos - number of positive eigenvalues
7031 
7032    Level: advanced
7033 
7034    Notes: Matrix must have been factored by MatCholeskyFactor()
7035 
7036 
7037 @*/
7038 PetscErrorCode PETSCMAT_DLLEXPORT MatGetInertia(Mat mat,PetscInt *nneg,PetscInt *nzero,PetscInt *npos)
7039 {
7040   PetscErrorCode ierr;
7041 
7042   PetscFunctionBegin;
7043   PetscValidHeaderSpecific(mat,MAT_CLASSID,1);
7044   PetscValidType(mat,1);
7045   if (!mat->factortype) SETERRQ(((PetscObject)mat)->comm,PETSC_ERR_ARG_WRONGSTATE,"Unfactored matrix");
7046   if (!mat->assembled) SETERRQ(((PetscObject)mat)->comm,PETSC_ERR_ARG_WRONGSTATE,"Numeric factor mat is not assembled");
7047   if (!mat->ops->getinertia) SETERRQ1(PETSC_COMM_SELF,PETSC_ERR_SUP,"Mat type %s",((PetscObject)mat)->type_name);
7048   ierr = (*mat->ops->getinertia)(mat,nneg,nzero,npos);CHKERRQ(ierr);
7049   PetscFunctionReturn(0);
7050 }
7051 
7052 /* ----------------------------------------------------------------*/
7053 #undef __FUNCT__
7054 #define __FUNCT__ "MatSolves"
7055 /*@C
7056    MatSolves - Solves A x = b, given a factored matrix, for a collection of vectors
7057 
7058    Collective on Mat and Vecs
7059 
7060    Input Parameters:
7061 +  mat - the factored matrix
7062 -  b - the right-hand-side vectors
7063 
7064    Output Parameter:
7065 .  x - the result vectors
7066 
7067    Notes:
7068    The vectors b and x cannot be the same.  I.e., one cannot
7069    call MatSolves(A,x,x).
7070 
7071    Notes:
7072    Most users should employ the simplified KSP interface for linear solvers
7073    instead of working directly with matrix algebra routines such as this.
7074    See, e.g., KSPCreate().
7075 
7076    Level: developer
7077 
7078    Concepts: matrices^triangular solves
7079 
7080 .seealso: MatSolveAdd(), MatSolveTranspose(), MatSolveTransposeAdd(), MatSolve()
7081 @*/
7082 PetscErrorCode PETSCMAT_DLLEXPORT MatSolves(Mat mat,Vecs b,Vecs x)
7083 {
7084   PetscErrorCode ierr;
7085 
7086   PetscFunctionBegin;
7087   PetscValidHeaderSpecific(mat,MAT_CLASSID,1);
7088   PetscValidType(mat,1);
7089   if (x == b) SETERRQ(((PetscObject)mat)->comm,PETSC_ERR_ARG_IDN,"x and b must be different vectors");
7090   if (!mat->factortype) SETERRQ(((PetscObject)mat)->comm,PETSC_ERR_ARG_WRONGSTATE,"Unfactored matrix");
7091   if (!mat->rmap->N && !mat->cmap->N) PetscFunctionReturn(0);
7092 
7093   if (!mat->ops->solves) SETERRQ1(((PetscObject)mat)->comm,PETSC_ERR_SUP,"Mat type %s",((PetscObject)mat)->type_name);
7094   ierr = MatPreallocated(mat);CHKERRQ(ierr);
7095   ierr = PetscLogEventBegin(MAT_Solves,mat,0,0,0);CHKERRQ(ierr);
7096   ierr = (*mat->ops->solves)(mat,b,x);CHKERRQ(ierr);
7097   ierr = PetscLogEventEnd(MAT_Solves,mat,0,0,0);CHKERRQ(ierr);
7098   PetscFunctionReturn(0);
7099 }
7100 
7101 #undef __FUNCT__
7102 #define __FUNCT__ "MatIsSymmetric"
7103 /*@
7104    MatIsSymmetric - Test whether a matrix is symmetric
7105 
7106    Collective on Mat
7107 
7108    Input Parameter:
7109 +  A - the matrix to test
7110 -  tol - difference between value and its transpose less than this amount counts as equal (use 0.0 for exact transpose)
7111 
7112    Output Parameters:
7113 .  flg - the result
7114 
7115    Level: intermediate
7116 
7117    Concepts: matrix^symmetry
7118 
7119 .seealso: MatTranspose(), MatIsTranspose(), MatIsHermitian(), MatIsStructurallySymmetric(), MatSetOption(), MatIsSymmetricKnown()
7120 @*/
7121 PetscErrorCode PETSCMAT_DLLEXPORT MatIsSymmetric(Mat A,PetscReal tol,PetscTruth *flg)
7122 {
7123   PetscErrorCode ierr;
7124 
7125   PetscFunctionBegin;
7126   PetscValidHeaderSpecific(A,MAT_CLASSID,1);
7127   PetscValidPointer(flg,2);
7128 
7129   if (!A->symmetric_set) {
7130     if (!A->ops->issymmetric) {
7131       const MatType mattype;
7132       ierr = MatGetType(A,&mattype);CHKERRQ(ierr);
7133       SETERRQ1(PETSC_COMM_SELF,PETSC_ERR_SUP,"Matrix of type <%s> does not support checking for symmetric",mattype);
7134     }
7135     ierr = (*A->ops->issymmetric)(A,tol,flg);CHKERRQ(ierr);
7136     if (!tol) {
7137       A->symmetric_set = PETSC_TRUE;
7138       A->symmetric = *flg;
7139       if (A->symmetric) {
7140 	A->structurally_symmetric_set = PETSC_TRUE;
7141 	A->structurally_symmetric     = PETSC_TRUE;
7142       }
7143     }
7144   } else if (A->symmetric) {
7145     *flg = PETSC_TRUE;
7146   } else if (!tol) {
7147     *flg = PETSC_FALSE;
7148   } else {
7149     if (!A->ops->issymmetric) {
7150       const MatType mattype;
7151       ierr = MatGetType(A,&mattype);CHKERRQ(ierr);
7152       SETERRQ1(PETSC_COMM_SELF,PETSC_ERR_SUP,"Matrix of type <%s> does not support checking for symmetric",mattype);
7153     }
7154     ierr = (*A->ops->issymmetric)(A,tol,flg);CHKERRQ(ierr);
7155   }
7156   PetscFunctionReturn(0);
7157 }
7158 
7159 #undef __FUNCT__
7160 #define __FUNCT__ "MatIsHermitian"
7161 /*@
7162    MatIsHermitian - Test whether a matrix is Hermitian
7163 
7164    Collective on Mat
7165 
7166    Input Parameter:
7167 +  A - the matrix to test
7168 -  tol - difference between value and its transpose less than this amount counts as equal (use 0.0 for exact Hermitian)
7169 
7170    Output Parameters:
7171 .  flg - the result
7172 
7173    Level: intermediate
7174 
7175    Concepts: matrix^symmetry
7176 
7177 .seealso: MatTranspose(), MatIsTranspose(), MatIsHermitian(), MatIsStructurallySymmetric(), MatSetOption(), MatIsSymmetricKnown()
7178 @*/
7179 PetscErrorCode PETSCMAT_DLLEXPORT MatIsHermitian(Mat A,PetscReal tol,PetscTruth *flg)
7180 {
7181   PetscErrorCode ierr;
7182 
7183   PetscFunctionBegin;
7184   PetscValidHeaderSpecific(A,MAT_CLASSID,1);
7185   PetscValidPointer(flg,2);
7186 
7187   if (!A->hermitian_set) {
7188     if (!A->ops->ishermitian) {
7189       const MatType mattype;
7190       ierr = MatGetType(A,&mattype);CHKERRQ(ierr);
7191       SETERRQ1(PETSC_COMM_SELF,PETSC_ERR_SUP,"Matrix of type <%s> does not support checking for hermitian",mattype);
7192     }
7193     ierr = (*A->ops->ishermitian)(A,tol,flg);CHKERRQ(ierr);
7194     if (!tol) {
7195       A->hermitian_set = PETSC_TRUE;
7196       A->hermitian = *flg;
7197       if (A->hermitian) {
7198 	A->structurally_symmetric_set = PETSC_TRUE;
7199 	A->structurally_symmetric     = PETSC_TRUE;
7200       }
7201     }
7202   } else if (A->hermitian) {
7203     *flg = PETSC_TRUE;
7204   } else if (!tol) {
7205     *flg = PETSC_FALSE;
7206   } else {
7207     if (!A->ops->ishermitian) {
7208       const MatType mattype;
7209       ierr = MatGetType(A,&mattype);CHKERRQ(ierr);
7210       SETERRQ1(PETSC_COMM_SELF,PETSC_ERR_SUP,"Matrix of type <%s> does not support checking for hermitian",mattype);
7211     }
7212     ierr = (*A->ops->ishermitian)(A,tol,flg);CHKERRQ(ierr);
7213   }
7214   PetscFunctionReturn(0);
7215 }
7216 
7217 #undef __FUNCT__
7218 #define __FUNCT__ "MatIsSymmetricKnown"
7219 /*@
7220    MatIsSymmetricKnown - Checks the flag on the matrix to see if it is symmetric.
7221 
7222    Collective on Mat
7223 
7224    Input Parameter:
7225 .  A - the matrix to check
7226 
7227    Output Parameters:
7228 +  set - if the symmetric flag is set (this tells you if the next flag is valid)
7229 -  flg - the result
7230 
7231    Level: advanced
7232 
7233    Concepts: matrix^symmetry
7234 
7235    Note: Does not check the matrix values directly, so this may return unknown (set = PETSC_FALSE). Use MatIsSymmetric()
7236          if you want it explicitly checked
7237 
7238 .seealso: MatTranspose(), MatIsTranspose(), MatIsHermitian(), MatIsStructurallySymmetric(), MatSetOption(), MatIsSymmetric()
7239 @*/
7240 PetscErrorCode PETSCMAT_DLLEXPORT MatIsSymmetricKnown(Mat A,PetscTruth *set,PetscTruth *flg)
7241 {
7242   PetscFunctionBegin;
7243   PetscValidHeaderSpecific(A,MAT_CLASSID,1);
7244   PetscValidPointer(set,2);
7245   PetscValidPointer(flg,3);
7246   if (A->symmetric_set) {
7247     *set = PETSC_TRUE;
7248     *flg = A->symmetric;
7249   } else {
7250     *set = PETSC_FALSE;
7251   }
7252   PetscFunctionReturn(0);
7253 }
7254 
7255 #undef __FUNCT__
7256 #define __FUNCT__ "MatIsHermitianKnown"
7257 /*@
7258    MatIsHermitianKnown - Checks the flag on the matrix to see if it is hermitian.
7259 
7260    Collective on Mat
7261 
7262    Input Parameter:
7263 .  A - the matrix to check
7264 
7265    Output Parameters:
7266 +  set - if the hermitian flag is set (this tells you if the next flag is valid)
7267 -  flg - the result
7268 
7269    Level: advanced
7270 
7271    Concepts: matrix^symmetry
7272 
7273    Note: Does not check the matrix values directly, so this may return unknown (set = PETSC_FALSE). Use MatIsHermitian()
7274          if you want it explicitly checked
7275 
7276 .seealso: MatTranspose(), MatIsTranspose(), MatIsHermitian(), MatIsStructurallySymmetric(), MatSetOption(), MatIsSymmetric()
7277 @*/
7278 PetscErrorCode PETSCMAT_DLLEXPORT MatIsHermitianKnown(Mat A,PetscTruth *set,PetscTruth *flg)
7279 {
7280   PetscFunctionBegin;
7281   PetscValidHeaderSpecific(A,MAT_CLASSID,1);
7282   PetscValidPointer(set,2);
7283   PetscValidPointer(flg,3);
7284   if (A->hermitian_set) {
7285     *set = PETSC_TRUE;
7286     *flg = A->hermitian;
7287   } else {
7288     *set = PETSC_FALSE;
7289   }
7290   PetscFunctionReturn(0);
7291 }
7292 
7293 #undef __FUNCT__
7294 #define __FUNCT__ "MatIsStructurallySymmetric"
7295 /*@
7296    MatIsStructurallySymmetric - Test whether a matrix is structurally symmetric
7297 
7298    Collective on Mat
7299 
7300    Input Parameter:
7301 .  A - the matrix to test
7302 
7303    Output Parameters:
7304 .  flg - the result
7305 
7306    Level: intermediate
7307 
7308    Concepts: matrix^symmetry
7309 
7310 .seealso: MatTranspose(), MatIsTranspose(), MatIsHermitian(), MatIsSymmetric(), MatSetOption()
7311 @*/
7312 PetscErrorCode PETSCMAT_DLLEXPORT MatIsStructurallySymmetric(Mat A,PetscTruth *flg)
7313 {
7314   PetscErrorCode ierr;
7315 
7316   PetscFunctionBegin;
7317   PetscValidHeaderSpecific(A,MAT_CLASSID,1);
7318   PetscValidPointer(flg,2);
7319   if (!A->structurally_symmetric_set) {
7320     if (!A->ops->isstructurallysymmetric) SETERRQ(((PetscObject)A)->comm,PETSC_ERR_SUP,"Matrix does not support checking for structural symmetric");
7321     ierr = (*A->ops->isstructurallysymmetric)(A,&A->structurally_symmetric);CHKERRQ(ierr);
7322     A->structurally_symmetric_set = PETSC_TRUE;
7323   }
7324   *flg = A->structurally_symmetric;
7325   PetscFunctionReturn(0);
7326 }
7327 
7328 #undef __FUNCT__
7329 #define __FUNCT__ "MatStashGetInfo"
7330 extern PetscErrorCode MatStashGetInfo_Private(MatStash*,PetscInt*,PetscInt*);
7331 /*@
7332    MatStashGetInfo - Gets how many values are currently in the vector stash, i.e. need
7333        to be communicated to other processors during the MatAssemblyBegin/End() process
7334 
7335     Not collective
7336 
7337    Input Parameter:
7338 .   vec - the vector
7339 
7340    Output Parameters:
7341 +   nstash   - the size of the stash
7342 .   reallocs - the number of additional mallocs incurred.
7343 .   bnstash   - the size of the block stash
7344 -   breallocs - the number of additional mallocs incurred.in the block stash
7345 
7346    Level: advanced
7347 
7348 .seealso: MatAssemblyBegin(), MatAssemblyEnd(), Mat, MatStashSetInitialSize()
7349 
7350 @*/
7351 PetscErrorCode PETSCMAT_DLLEXPORT MatStashGetInfo(Mat mat,PetscInt *nstash,PetscInt *reallocs,PetscInt *bnstash,PetscInt *breallocs)
7352 {
7353   PetscErrorCode ierr;
7354   PetscFunctionBegin;
7355   ierr = MatStashGetInfo_Private(&mat->stash,nstash,reallocs);CHKERRQ(ierr);
7356   ierr = MatStashGetInfo_Private(&mat->bstash,bnstash,breallocs);CHKERRQ(ierr);
7357   PetscFunctionReturn(0);
7358 }
7359 
7360 #undef __FUNCT__
7361 #define __FUNCT__ "MatGetVecs"
7362 /*@C
7363    MatGetVecs - Get vector(s) compatible with the matrix, i.e. with the same
7364      parallel layout
7365 
7366    Collective on Mat
7367 
7368    Input Parameter:
7369 .  mat - the matrix
7370 
7371    Output Parameter:
7372 +   right - (optional) vector that the matrix can be multiplied against
7373 -   left - (optional) vector that the matrix vector product can be stored in
7374 
7375   Level: advanced
7376 
7377 .seealso: MatCreate()
7378 @*/
7379 PetscErrorCode PETSCMAT_DLLEXPORT MatGetVecs(Mat mat,Vec *right,Vec *left)
7380 {
7381   PetscErrorCode ierr;
7382 
7383   PetscFunctionBegin;
7384   PetscValidHeaderSpecific(mat,MAT_CLASSID,1);
7385   PetscValidType(mat,1);
7386   ierr = MatPreallocated(mat);CHKERRQ(ierr);
7387   if (mat->ops->getvecs) {
7388     ierr = (*mat->ops->getvecs)(mat,right,left);CHKERRQ(ierr);
7389   } else {
7390     PetscMPIInt size;
7391     ierr = MPI_Comm_size(((PetscObject)mat)->comm, &size);CHKERRQ(ierr);
7392     if (right) {
7393       ierr = VecCreate(((PetscObject)mat)->comm,right);CHKERRQ(ierr);
7394       ierr = VecSetSizes(*right,mat->cmap->n,PETSC_DETERMINE);CHKERRQ(ierr);
7395       ierr = VecSetBlockSize(*right,mat->rmap->bs);CHKERRQ(ierr);
7396       if (size > 1) {
7397         /* New vectors uses Mat cmap and does not create a new one */
7398 	ierr = PetscLayoutDestroy((*right)->map);CHKERRQ(ierr);
7399 	(*right)->map = mat->cmap;
7400 	mat->cmap->refcnt++;
7401 
7402         ierr = VecSetType(*right,VECMPI);CHKERRQ(ierr);
7403       } else {ierr = VecSetType(*right,VECSEQ);CHKERRQ(ierr);}
7404     }
7405     if (left) {
7406       ierr = VecCreate(((PetscObject)mat)->comm,left);CHKERRQ(ierr);
7407       ierr = VecSetSizes(*left,mat->rmap->n,PETSC_DETERMINE);CHKERRQ(ierr);
7408       ierr = VecSetBlockSize(*left,mat->rmap->bs);CHKERRQ(ierr);
7409       if (size > 1) {
7410         /* New vectors uses Mat rmap and does not create a new one */
7411 	ierr = PetscLayoutDestroy((*left)->map);CHKERRQ(ierr);
7412 	(*left)->map = mat->rmap;
7413 	mat->rmap->refcnt++;
7414 
7415         ierr = VecSetType(*left,VECMPI);CHKERRQ(ierr);
7416       } else {ierr = VecSetType(*left,VECSEQ);CHKERRQ(ierr);}
7417     }
7418   }
7419   if (mat->mapping) {
7420     if (right) {ierr = VecSetLocalToGlobalMapping(*right,mat->mapping);CHKERRQ(ierr);}
7421     if (left) {ierr = VecSetLocalToGlobalMapping(*left,mat->mapping);CHKERRQ(ierr);}
7422   }
7423   if (mat->bmapping) {
7424     if (right) {ierr = VecSetLocalToGlobalMappingBlock(*right,mat->bmapping);CHKERRQ(ierr);}
7425     if (left) {ierr = VecSetLocalToGlobalMappingBlock(*left,mat->bmapping);CHKERRQ(ierr);}
7426   }
7427   PetscFunctionReturn(0);
7428 }
7429 
7430 #undef __FUNCT__
7431 #define __FUNCT__ "MatFactorInfoInitialize"
7432 /*@C
7433    MatFactorInfoInitialize - Initializes a MatFactorInfo data structure
7434      with default values.
7435 
7436    Not Collective
7437 
7438    Input Parameters:
7439 .    info - the MatFactorInfo data structure
7440 
7441 
7442    Notes: The solvers are generally used through the KSP and PC objects, for example
7443           PCLU, PCILU, PCCHOLESKY, PCICC
7444 
7445    Level: developer
7446 
7447 .seealso: MatFactorInfo
7448 
7449     Developer Note: fortran interface is not autogenerated as the f90
7450     interface defintion cannot be generated correctly [due to MatFactorInfo]
7451 
7452 @*/
7453 
7454 PetscErrorCode PETSCMAT_DLLEXPORT MatFactorInfoInitialize(MatFactorInfo *info)
7455 {
7456   PetscErrorCode ierr;
7457 
7458   PetscFunctionBegin;
7459   ierr = PetscMemzero(info,sizeof(MatFactorInfo));CHKERRQ(ierr);
7460   PetscFunctionReturn(0);
7461 }
7462 
7463 #undef __FUNCT__
7464 #define __FUNCT__ "MatPtAP"
7465 /*@
7466    MatPtAP - Creates the matrix projection C = P^T * A * P
7467 
7468    Collective on Mat
7469 
7470    Input Parameters:
7471 +  A - the matrix
7472 .  P - the projection matrix
7473 .  scall - either MAT_INITIAL_MATRIX or MAT_REUSE_MATRIX
7474 -  fill - expected fill as ratio of nnz(C)/nnz(A)
7475 
7476    Output Parameters:
7477 .  C - the product matrix
7478 
7479    Notes:
7480    C will be created and must be destroyed by the user with MatDestroy().
7481 
7482    This routine is currently only implemented for pairs of AIJ matrices and classes
7483    which inherit from AIJ.
7484 
7485    Level: intermediate
7486 
7487 .seealso: MatPtAPSymbolic(), MatPtAPNumeric(), MatMatMult()
7488 @*/
7489 PetscErrorCode PETSCMAT_DLLEXPORT MatPtAP(Mat A,Mat P,MatReuse scall,PetscReal fill,Mat *C)
7490 {
7491   PetscErrorCode ierr;
7492 
7493   PetscFunctionBegin;
7494   PetscValidHeaderSpecific(A,MAT_CLASSID,1);
7495   PetscValidType(A,1);
7496   if (!A->assembled) SETERRQ(((PetscObject)A)->comm,PETSC_ERR_ARG_WRONGSTATE,"Not for unassembled matrix");
7497   if (A->factortype) SETERRQ(((PetscObject)A)->comm,PETSC_ERR_ARG_WRONGSTATE,"Not for factored matrix");
7498   PetscValidHeaderSpecific(P,MAT_CLASSID,2);
7499   PetscValidType(P,2);
7500   ierr = MatPreallocated(P);CHKERRQ(ierr);
7501   if (!P->assembled) SETERRQ(((PetscObject)A)->comm,PETSC_ERR_ARG_WRONGSTATE,"Not for unassembled matrix");
7502   if (P->factortype) SETERRQ(((PetscObject)A)->comm,PETSC_ERR_ARG_WRONGSTATE,"Not for factored matrix");
7503   PetscValidPointer(C,3);
7504   if (P->rmap->N!=A->cmap->N) SETERRQ2(((PetscObject)A)->comm,PETSC_ERR_ARG_SIZ,"Matrix dimensions are incompatible, %D != %D",P->rmap->N,A->cmap->N);
7505   if (fill < 1.0) SETERRQ1(((PetscObject)A)->comm,PETSC_ERR_ARG_SIZ,"Expected fill=%G must be >= 1.0",fill);
7506   ierr = MatPreallocated(A);CHKERRQ(ierr);
7507 
7508   if (!A->ops->ptap) {
7509     const MatType mattype;
7510     ierr = MatGetType(A,&mattype);CHKERRQ(ierr);
7511     SETERRQ1(((PetscObject)A)->comm,PETSC_ERR_SUP,"Matrix of type <%s> does not support PtAP",mattype);
7512   }
7513   ierr = PetscLogEventBegin(MAT_PtAP,A,P,0,0);CHKERRQ(ierr);
7514   ierr = (*A->ops->ptap)(A,P,scall,fill,C);CHKERRQ(ierr);
7515   ierr = PetscLogEventEnd(MAT_PtAP,A,P,0,0);CHKERRQ(ierr);
7516 
7517   PetscFunctionReturn(0);
7518 }
7519 
7520 #undef __FUNCT__
7521 #define __FUNCT__ "MatPtAPNumeric"
7522 /*@
7523    MatPtAPNumeric - Computes the matrix projection C = P^T * A * P
7524 
7525    Collective on Mat
7526 
7527    Input Parameters:
7528 +  A - the matrix
7529 -  P - the projection matrix
7530 
7531    Output Parameters:
7532 .  C - the product matrix
7533 
7534    Notes:
7535    C must have been created by calling MatPtAPSymbolic and must be destroyed by
7536    the user using MatDeatroy().
7537 
7538    This routine is currently only implemented for pairs of AIJ matrices and classes
7539    which inherit from AIJ.  C will be of type MATAIJ.
7540 
7541    Level: intermediate
7542 
7543 .seealso: MatPtAP(), MatPtAPSymbolic(), MatMatMultNumeric()
7544 @*/
7545 PetscErrorCode PETSCMAT_DLLEXPORT MatPtAPNumeric(Mat A,Mat P,Mat C)
7546 {
7547   PetscErrorCode ierr;
7548 
7549   PetscFunctionBegin;
7550   PetscValidHeaderSpecific(A,MAT_CLASSID,1);
7551   PetscValidType(A,1);
7552   if (!A->assembled) SETERRQ(((PetscObject)A)->comm,PETSC_ERR_ARG_WRONGSTATE,"Not for unassembled matrix");
7553   if (A->factortype) SETERRQ(((PetscObject)A)->comm,PETSC_ERR_ARG_WRONGSTATE,"Not for factored matrix");
7554   PetscValidHeaderSpecific(P,MAT_CLASSID,2);
7555   PetscValidType(P,2);
7556   ierr = MatPreallocated(P);CHKERRQ(ierr);
7557   if (!P->assembled) SETERRQ(((PetscObject)A)->comm,PETSC_ERR_ARG_WRONGSTATE,"Not for unassembled matrix");
7558   if (P->factortype) SETERRQ(((PetscObject)A)->comm,PETSC_ERR_ARG_WRONGSTATE,"Not for factored matrix");
7559   PetscValidHeaderSpecific(C,MAT_CLASSID,3);
7560   PetscValidType(C,3);
7561   ierr = MatPreallocated(C);CHKERRQ(ierr);
7562   if (C->factortype) SETERRQ(((PetscObject)A)->comm,PETSC_ERR_ARG_WRONGSTATE,"Not for factored matrix");
7563   if (P->cmap->N!=C->rmap->N) SETERRQ2(((PetscObject)A)->comm,PETSC_ERR_ARG_SIZ,"Matrix dimensions are incompatible, %D != %D",P->cmap->N,C->rmap->N);
7564   if (P->rmap->N!=A->cmap->N) SETERRQ2(((PetscObject)A)->comm,PETSC_ERR_ARG_SIZ,"Matrix dimensions are incompatible, %D != %D",P->rmap->N,A->cmap->N);
7565   if (A->rmap->N!=A->cmap->N) SETERRQ2(((PetscObject)A)->comm,PETSC_ERR_ARG_SIZ,"Matrix 'A' must be square, %D != %D",A->rmap->N,A->cmap->N);
7566   if (P->cmap->N!=C->cmap->N) SETERRQ2(((PetscObject)A)->comm,PETSC_ERR_ARG_SIZ,"Matrix dimensions are incompatible, %D != %D",P->cmap->N,C->cmap->N);
7567   ierr = MatPreallocated(A);CHKERRQ(ierr);
7568 
7569   ierr = PetscLogEventBegin(MAT_PtAPNumeric,A,P,0,0);CHKERRQ(ierr);
7570   ierr = (*A->ops->ptapnumeric)(A,P,C);CHKERRQ(ierr);
7571   ierr = PetscLogEventEnd(MAT_PtAPNumeric,A,P,0,0);CHKERRQ(ierr);
7572   PetscFunctionReturn(0);
7573 }
7574 
7575 #undef __FUNCT__
7576 #define __FUNCT__ "MatPtAPSymbolic"
7577 /*@
7578    MatPtAPSymbolic - Creates the (i,j) structure of the matrix projection C = P^T * A * P
7579 
7580    Collective on Mat
7581 
7582    Input Parameters:
7583 +  A - the matrix
7584 -  P - the projection matrix
7585 
7586    Output Parameters:
7587 .  C - the (i,j) structure of the product matrix
7588 
7589    Notes:
7590    C will be created and must be destroyed by the user with MatDestroy().
7591 
7592    This routine is currently only implemented for pairs of SeqAIJ matrices and classes
7593    which inherit from SeqAIJ.  C will be of type MATSEQAIJ.  The product is computed using
7594    this (i,j) structure by calling MatPtAPNumeric().
7595 
7596    Level: intermediate
7597 
7598 .seealso: MatPtAP(), MatPtAPNumeric(), MatMatMultSymbolic()
7599 @*/
7600 PetscErrorCode PETSCMAT_DLLEXPORT MatPtAPSymbolic(Mat A,Mat P,PetscReal fill,Mat *C)
7601 {
7602   PetscErrorCode ierr;
7603 
7604   PetscFunctionBegin;
7605   PetscValidHeaderSpecific(A,MAT_CLASSID,1);
7606   PetscValidType(A,1);
7607   if (!A->assembled) SETERRQ(((PetscObject)A)->comm,PETSC_ERR_ARG_WRONGSTATE,"Not for unassembled matrix");
7608   if (A->factortype) SETERRQ(((PetscObject)A)->comm,PETSC_ERR_ARG_WRONGSTATE,"Not for factored matrix");
7609   if (fill <1.0) SETERRQ1(((PetscObject)A)->comm,PETSC_ERR_ARG_SIZ,"Expected fill=%G must be >= 1.0",fill);
7610   PetscValidHeaderSpecific(P,MAT_CLASSID,2);
7611   PetscValidType(P,2);
7612   ierr = MatPreallocated(P);CHKERRQ(ierr);
7613   if (!P->assembled) SETERRQ(((PetscObject)A)->comm,PETSC_ERR_ARG_WRONGSTATE,"Not for unassembled matrix");
7614   if (P->factortype) SETERRQ(((PetscObject)A)->comm,PETSC_ERR_ARG_WRONGSTATE,"Not for factored matrix");
7615   PetscValidPointer(C,3);
7616 
7617   if (P->rmap->N!=A->cmap->N) SETERRQ2(((PetscObject)A)->comm,PETSC_ERR_ARG_SIZ,"Matrix dimensions are incompatible, %D != %D",P->rmap->N,A->cmap->N);
7618   if (A->rmap->N!=A->cmap->N) SETERRQ2(((PetscObject)A)->comm,PETSC_ERR_ARG_SIZ,"Matrix 'A' must be square, %D != %D",A->rmap->N,A->cmap->N);
7619   ierr = MatPreallocated(A);CHKERRQ(ierr);
7620   ierr = PetscLogEventBegin(MAT_PtAPSymbolic,A,P,0,0);CHKERRQ(ierr);
7621   ierr = (*A->ops->ptapsymbolic)(A,P,fill,C);CHKERRQ(ierr);
7622   ierr = PetscLogEventEnd(MAT_PtAPSymbolic,A,P,0,0);CHKERRQ(ierr);
7623 
7624   ierr = MatSetBlockSize(*C,A->rmap->bs);CHKERRQ(ierr);
7625 
7626   PetscFunctionReturn(0);
7627 }
7628 
7629 #undef __FUNCT__
7630 #define __FUNCT__ "MatMatMult"
7631 /*@
7632    MatMatMult - Performs Matrix-Matrix Multiplication C=A*B.
7633 
7634    Collective on Mat
7635 
7636    Input Parameters:
7637 +  A - the left matrix
7638 .  B - the right matrix
7639 .  scall - either MAT_INITIAL_MATRIX or MAT_REUSE_MATRIX
7640 -  fill - expected fill as ratio of nnz(C)/(nnz(A) + nnz(B)), use PETSC_DEFAULT if you do not have a good estimate
7641           if the result is a dense matrix this is irrelevent
7642 
7643    Output Parameters:
7644 .  C - the product matrix
7645 
7646    Notes:
7647    Unless scall is MAT_REUSE_MATRIX C will be created.
7648 
7649    MAT_REUSE_MATRIX can only be used if the matrices A and B have the same nonzero pattern as in the previous call
7650 
7651    To determine the correct fill value, run with -info and search for the string "Fill ratio" to see the value
7652    actually needed.
7653 
7654    If you have many matrices with the same non-zero structure to multiply, you
7655    should either
7656 $   1) use MAT_REUSE_MATRIX in all calls but the first or
7657 $   2) call MatMatMultSymbolic() once and then MatMatMultNumeric() for each product needed
7658 
7659    Level: intermediate
7660 
7661 .seealso: MatMatMultSymbolic(), MatMatMultNumeric(), MatPtAP()
7662 @*/
7663 PetscErrorCode PETSCMAT_DLLEXPORT MatMatMult(Mat A,Mat B,MatReuse scall,PetscReal fill,Mat *C)
7664 {
7665   PetscErrorCode ierr;
7666   PetscErrorCode (*fA)(Mat,Mat,MatReuse,PetscReal,Mat*);
7667   PetscErrorCode (*fB)(Mat,Mat,MatReuse,PetscReal,Mat*);
7668   PetscErrorCode (*mult)(Mat,Mat,MatReuse,PetscReal,Mat *)=PETSC_NULL;
7669 
7670   PetscFunctionBegin;
7671   PetscValidHeaderSpecific(A,MAT_CLASSID,1);
7672   PetscValidType(A,1);
7673   if (!A->assembled) SETERRQ(((PetscObject)A)->comm,PETSC_ERR_ARG_WRONGSTATE,"Not for unassembled matrix");
7674   if (A->factortype) SETERRQ(((PetscObject)A)->comm,PETSC_ERR_ARG_WRONGSTATE,"Not for factored matrix");
7675   PetscValidHeaderSpecific(B,MAT_CLASSID,2);
7676   PetscValidType(B,2);
7677   ierr = MatPreallocated(B);CHKERRQ(ierr);
7678   if (!B->assembled) SETERRQ(((PetscObject)A)->comm,PETSC_ERR_ARG_WRONGSTATE,"Not for unassembled matrix");
7679   if (B->factortype) SETERRQ(((PetscObject)A)->comm,PETSC_ERR_ARG_WRONGSTATE,"Not for factored matrix");
7680   PetscValidPointer(C,3);
7681   if (B->rmap->N!=A->cmap->N) SETERRQ2(((PetscObject)A)->comm,PETSC_ERR_ARG_SIZ,"Matrix dimensions are incompatible, %D != %D",B->rmap->N,A->cmap->N);
7682   if (scall == MAT_REUSE_MATRIX){
7683     PetscValidPointer(*C,5);
7684     PetscValidHeaderSpecific(*C,MAT_CLASSID,5);
7685   }
7686   if (fill == PETSC_DEFAULT || fill == PETSC_DECIDE) fill = 2.0;
7687   if (fill < 1.0) SETERRQ1(((PetscObject)A)->comm,PETSC_ERR_ARG_SIZ,"Expected fill=%G must be >= 1.0",fill);
7688   ierr = MatPreallocated(A);CHKERRQ(ierr);
7689 
7690   fA = A->ops->matmult;
7691   fB = B->ops->matmult;
7692   if (fB == fA) {
7693     if (!fB) SETERRQ1(((PetscObject)A)->comm,PETSC_ERR_SUP,"MatMatMult not supported for B of type %s",((PetscObject)B)->type_name);
7694     mult = fB;
7695   } else {
7696     /* dispatch based on the type of A and B */
7697     char  multname[256];
7698     ierr = PetscStrcpy(multname,"MatMatMult_");CHKERRQ(ierr);
7699     ierr = PetscStrcat(multname,((PetscObject)A)->type_name);CHKERRQ(ierr);
7700     ierr = PetscStrcat(multname,"_");CHKERRQ(ierr);
7701     ierr = PetscStrcat(multname,((PetscObject)B)->type_name);CHKERRQ(ierr);
7702     ierr = PetscStrcat(multname,"_C");CHKERRQ(ierr); /* e.g., multname = "MatMatMult_seqdense_seqaij_C" */
7703     ierr = PetscObjectQueryFunction((PetscObject)B,multname,(void (**)(void))&mult);CHKERRQ(ierr);
7704     if (!mult) SETERRQ2(((PetscObject)A)->comm,PETSC_ERR_ARG_INCOMP,"MatMatMult requires A, %s, to be compatible with B, %s",((PetscObject)A)->type_name,((PetscObject)B)->type_name);
7705   }
7706   ierr = PetscLogEventBegin(MAT_MatMult,A,B,0,0);CHKERRQ(ierr);
7707   ierr = (*mult)(A,B,scall,fill,C);CHKERRQ(ierr);
7708   ierr = PetscLogEventEnd(MAT_MatMult,A,B,0,0);CHKERRQ(ierr);
7709   PetscFunctionReturn(0);
7710 }
7711 
7712 #undef __FUNCT__
7713 #define __FUNCT__ "MatMatMultSymbolic"
7714 /*@
7715    MatMatMultSymbolic - Performs construction, preallocation, and computes the ij structure
7716    of the matrix-matrix product C=A*B.  Call this routine before calling MatMatMultNumeric().
7717 
7718    Collective on Mat
7719 
7720    Input Parameters:
7721 +  A - the left matrix
7722 .  B - the right matrix
7723 -  fill - expected fill as ratio of nnz(C)/(nnz(A) + nnz(B)), use PETSC_DEFAULT if you do not have a good estimate,
7724       if C is a dense matrix this is irrelevent
7725 
7726    Output Parameters:
7727 .  C - the product matrix
7728 
7729    Notes:
7730    Unless scall is MAT_REUSE_MATRIX C will be created.
7731 
7732    To determine the correct fill value, run with -info and search for the string "Fill ratio" to see the value
7733    actually needed.
7734 
7735    This routine is currently implemented for
7736     - pairs of AIJ matrices and classes which inherit from AIJ, C will be of type AIJ
7737     - pairs of AIJ (A) and Dense (B) matrix, C will be of type Dense.
7738     - pairs of Dense (A) and AIJ (B) matrix, C will be of type Dense.
7739 
7740    Level: intermediate
7741 
7742    Developers Note: There are ways to estimate the number of nonzeros in the resulting product, see for example, http://arxiv.org/abs/1006.4173
7743      We should incorporate them into PETSc.
7744 
7745 .seealso: MatMatMult(), MatMatMultNumeric()
7746 @*/
7747 PetscErrorCode PETSCMAT_DLLEXPORT MatMatMultSymbolic(Mat A,Mat B,PetscReal fill,Mat *C)
7748 {
7749   PetscErrorCode ierr;
7750   PetscErrorCode (*Asymbolic)(Mat,Mat,PetscReal,Mat *);
7751   PetscErrorCode (*Bsymbolic)(Mat,Mat,PetscReal,Mat *);
7752   PetscErrorCode (*symbolic)(Mat,Mat,PetscReal,Mat *)=PETSC_NULL;
7753 
7754   PetscFunctionBegin;
7755   PetscValidHeaderSpecific(A,MAT_CLASSID,1);
7756   PetscValidType(A,1);
7757   if (!A->assembled) SETERRQ(((PetscObject)A)->comm,PETSC_ERR_ARG_WRONGSTATE,"Not for unassembled matrix");
7758   if (A->factortype) SETERRQ(((PetscObject)A)->comm,PETSC_ERR_ARG_WRONGSTATE,"Not for factored matrix");
7759 
7760   PetscValidHeaderSpecific(B,MAT_CLASSID,2);
7761   PetscValidType(B,2);
7762   ierr = MatPreallocated(B);CHKERRQ(ierr);
7763   if (!B->assembled) SETERRQ(((PetscObject)A)->comm,PETSC_ERR_ARG_WRONGSTATE,"Not for unassembled matrix");
7764   if (B->factortype) SETERRQ(((PetscObject)A)->comm,PETSC_ERR_ARG_WRONGSTATE,"Not for factored matrix");
7765   PetscValidPointer(C,3);
7766 
7767   if (B->rmap->N!=A->cmap->N) SETERRQ2(((PetscObject)A)->comm,PETSC_ERR_ARG_SIZ,"Matrix dimensions are incompatible, %D != %D",B->rmap->N,A->cmap->N);
7768   if (fill == PETSC_DEFAULT) fill = 2.0;
7769   if (fill < 1.0) SETERRQ1(((PetscObject)A)->comm,PETSC_ERR_ARG_SIZ,"Expected fill=%G must be > 1.0",fill);
7770   ierr = MatPreallocated(A);CHKERRQ(ierr);
7771 
7772   Asymbolic = A->ops->matmultsymbolic;
7773   Bsymbolic = B->ops->matmultsymbolic;
7774   if (Asymbolic == Bsymbolic){
7775     if (!Bsymbolic) SETERRQ1(((PetscObject)A)->comm,PETSC_ERR_SUP,"C=A*B not implemented for B of type %s",((PetscObject)B)->type_name);
7776     symbolic = Bsymbolic;
7777   } else { /* dispatch based on the type of A and B */
7778     char  symbolicname[256];
7779     ierr = PetscStrcpy(symbolicname,"MatMatMultSymbolic_");CHKERRQ(ierr);
7780     ierr = PetscStrcat(symbolicname,((PetscObject)A)->type_name);CHKERRQ(ierr);
7781     ierr = PetscStrcat(symbolicname,"_");CHKERRQ(ierr);
7782     ierr = PetscStrcat(symbolicname,((PetscObject)B)->type_name);CHKERRQ(ierr);
7783     ierr = PetscStrcat(symbolicname,"_C");CHKERRQ(ierr);
7784     ierr = PetscObjectQueryFunction((PetscObject)B,symbolicname,(void (**)(void))&symbolic);CHKERRQ(ierr);
7785     if (!symbolic) SETERRQ2(((PetscObject)A)->comm,PETSC_ERR_ARG_INCOMP,"MatMatMultSymbolic requires A, %s, to be compatible with B, %s",((PetscObject)A)->type_name,((PetscObject)B)->type_name);
7786   }
7787   ierr = PetscLogEventBegin(MAT_MatMultSymbolic,A,B,0,0);CHKERRQ(ierr);
7788   ierr = (*symbolic)(A,B,fill,C);CHKERRQ(ierr);
7789   ierr = PetscLogEventEnd(MAT_MatMultSymbolic,A,B,0,0);CHKERRQ(ierr);
7790   PetscFunctionReturn(0);
7791 }
7792 
7793 #undef __FUNCT__
7794 #define __FUNCT__ "MatMatMultNumeric"
7795 /*@
7796    MatMatMultNumeric - Performs the numeric matrix-matrix product.
7797    Call this routine after first calling MatMatMultSymbolic().
7798 
7799    Collective on Mat
7800 
7801    Input Parameters:
7802 +  A - the left matrix
7803 -  B - the right matrix
7804 
7805    Output Parameters:
7806 .  C - the product matrix, which was created by from MatMatMultSymbolic() or a call to MatMatMult().
7807 
7808    Notes:
7809    C must have been created with MatMatMultSymbolic().
7810 
7811    This routine is currently implemented for
7812     - pairs of AIJ matrices and classes which inherit from AIJ, C will be of type MATAIJ.
7813     - pairs of AIJ (A) and Dense (B) matrix, C will be of type Dense.
7814     - pairs of Dense (A) and AIJ (B) matrix, C will be of type Dense.
7815 
7816    Level: intermediate
7817 
7818 .seealso: MatMatMult(), MatMatMultSymbolic()
7819 @*/
7820 PetscErrorCode PETSCMAT_DLLEXPORT MatMatMultNumeric(Mat A,Mat B,Mat C)
7821 {
7822   PetscErrorCode ierr;
7823   PetscErrorCode (*Anumeric)(Mat,Mat,Mat);
7824   PetscErrorCode (*Bnumeric)(Mat,Mat,Mat);
7825   PetscErrorCode (*numeric)(Mat,Mat,Mat)=PETSC_NULL;
7826 
7827   PetscFunctionBegin;
7828   PetscValidHeaderSpecific(A,MAT_CLASSID,1);
7829   PetscValidType(A,1);
7830   if (!A->assembled) SETERRQ(((PetscObject)A)->comm,PETSC_ERR_ARG_WRONGSTATE,"Not for unassembled matrix");
7831   if (A->factortype) SETERRQ(((PetscObject)A)->comm,PETSC_ERR_ARG_WRONGSTATE,"Not for factored matrix");
7832 
7833   PetscValidHeaderSpecific(B,MAT_CLASSID,2);
7834   PetscValidType(B,2);
7835   ierr = MatPreallocated(B);CHKERRQ(ierr);
7836   if (!B->assembled) SETERRQ(((PetscObject)A)->comm,PETSC_ERR_ARG_WRONGSTATE,"Not for unassembled matrix");
7837   if (B->factortype) SETERRQ(((PetscObject)A)->comm,PETSC_ERR_ARG_WRONGSTATE,"Not for factored matrix");
7838 
7839   PetscValidHeaderSpecific(C,MAT_CLASSID,3);
7840   PetscValidType(C,3);
7841   ierr = MatPreallocated(C);CHKERRQ(ierr);
7842   if (!C->assembled) SETERRQ(((PetscObject)A)->comm,PETSC_ERR_ARG_WRONGSTATE,"Not for unassembled matrix");
7843   if (C->factortype) SETERRQ(((PetscObject)A)->comm,PETSC_ERR_ARG_WRONGSTATE,"Not for factored matrix");
7844 
7845   if (B->cmap->N!=C->cmap->N) SETERRQ2(((PetscObject)A)->comm,PETSC_ERR_ARG_SIZ,"Matrix dimensions are incompatible, %D != %D",B->cmap->N,C->cmap->N);
7846   if (B->rmap->N!=A->cmap->N) SETERRQ2(((PetscObject)A)->comm,PETSC_ERR_ARG_SIZ,"Matrix dimensions are incompatible, %D != %D",B->rmap->N,A->cmap->N);
7847   if (A->rmap->N!=C->rmap->N) SETERRQ2(((PetscObject)A)->comm,PETSC_ERR_ARG_SIZ,"Matrix dimensions are incompatible, %D != %D",A->rmap->N,C->rmap->N);
7848   ierr = MatPreallocated(A);CHKERRQ(ierr);
7849 
7850   Anumeric = A->ops->matmultnumeric;
7851   Bnumeric = B->ops->matmultnumeric;
7852   if (Anumeric == Bnumeric){
7853     if (!Bnumeric) SETERRQ1(((PetscObject)A)->comm,PETSC_ERR_SUP,"MatMatMultNumeric not supported for B of type %s",((PetscObject)B)->type_name);
7854     numeric = Bnumeric;
7855   } else {
7856     char  numericname[256];
7857     ierr = PetscStrcpy(numericname,"MatMatMultNumeric_");CHKERRQ(ierr);
7858     ierr = PetscStrcat(numericname,((PetscObject)A)->type_name);CHKERRQ(ierr);
7859     ierr = PetscStrcat(numericname,"_");CHKERRQ(ierr);
7860     ierr = PetscStrcat(numericname,((PetscObject)B)->type_name);CHKERRQ(ierr);
7861     ierr = PetscStrcat(numericname,"_C");CHKERRQ(ierr);
7862     ierr = PetscObjectQueryFunction((PetscObject)B,numericname,(void (**)(void))&numeric);CHKERRQ(ierr);
7863     if (!numeric)
7864       SETERRQ2(((PetscObject)A)->comm,PETSC_ERR_ARG_INCOMP,"MatMatMultNumeric requires A, %s, to be compatible with B, %s",((PetscObject)A)->type_name,((PetscObject)B)->type_name);
7865   }
7866   ierr = PetscLogEventBegin(MAT_MatMultNumeric,A,B,0,0);CHKERRQ(ierr);
7867   ierr = (*numeric)(A,B,C);CHKERRQ(ierr);
7868   ierr = PetscLogEventEnd(MAT_MatMultNumeric,A,B,0,0);CHKERRQ(ierr);
7869   PetscFunctionReturn(0);
7870 }
7871 
7872 #undef __FUNCT__
7873 #define __FUNCT__ "MatMatMultTranspose"
7874 /*@
7875    MatMatMultTranspose - Performs Matrix-Matrix Multiplication C=A^T*B.
7876 
7877    Collective on Mat
7878 
7879    Input Parameters:
7880 +  A - the left matrix
7881 .  B - the right matrix
7882 .  scall - either MAT_INITIAL_MATRIX or MAT_REUSE_MATRIX
7883 -  fill - expected fill as ratio of nnz(C)/(nnz(A) + nnz(B)), use PETSC_DEFAULT if not known
7884 
7885    Output Parameters:
7886 .  C - the product matrix
7887 
7888    Notes:
7889    C will be created if MAT_INITIAL_MATRIX and must be destroyed by the user with MatDestroy().
7890 
7891    MAT_REUSE_MATRIX can only be used if the matrices A and B have the same nonzero pattern as in the previous call
7892 
7893   To determine the correct fill value, run with -info and search for the string "Fill ratio" to see the value
7894    actually needed.
7895 
7896    This routine is currently only implemented for pairs of SeqAIJ matrices and pairs of SeqDense matrices and classes
7897    which inherit from SeqAIJ.  C will be of type MATSEQAIJ.
7898 
7899    Level: intermediate
7900 
7901 .seealso: MatMatMultTransposeSymbolic(), MatMatMultTransposeNumeric(), MatPtAP()
7902 @*/
7903 PetscErrorCode PETSCMAT_DLLEXPORT MatMatMultTranspose(Mat A,Mat B,MatReuse scall,PetscReal fill,Mat *C)
7904 {
7905   PetscErrorCode ierr;
7906   PetscErrorCode (*fA)(Mat,Mat,MatReuse,PetscReal,Mat*);
7907   PetscErrorCode (*fB)(Mat,Mat,MatReuse,PetscReal,Mat*);
7908 
7909   PetscFunctionBegin;
7910   PetscValidHeaderSpecific(A,MAT_CLASSID,1);
7911   PetscValidType(A,1);
7912   if (!A->assembled) SETERRQ(((PetscObject)A)->comm,PETSC_ERR_ARG_WRONGSTATE,"Not for unassembled matrix");
7913   if (A->factortype) SETERRQ(((PetscObject)A)->comm,PETSC_ERR_ARG_WRONGSTATE,"Not for factored matrix");
7914   PetscValidHeaderSpecific(B,MAT_CLASSID,2);
7915   PetscValidType(B,2);
7916   ierr = MatPreallocated(B);CHKERRQ(ierr);
7917   if (!B->assembled) SETERRQ(((PetscObject)A)->comm,PETSC_ERR_ARG_WRONGSTATE,"Not for unassembled matrix");
7918   if (B->factortype) SETERRQ(((PetscObject)A)->comm,PETSC_ERR_ARG_WRONGSTATE,"Not for factored matrix");
7919   PetscValidPointer(C,3);
7920   if (B->rmap->N!=A->rmap->N) SETERRQ2(((PetscObject)A)->comm,PETSC_ERR_ARG_SIZ,"Matrix dimensions are incompatible, %D != %D",B->rmap->N,A->rmap->N);
7921   if (fill == PETSC_DEFAULT || fill == PETSC_DECIDE) fill = 2.0;
7922   if (fill < 1.0) SETERRQ1(((PetscObject)A)->comm,PETSC_ERR_ARG_SIZ,"Expected fill=%G must be > 1.0",fill);
7923   ierr = MatPreallocated(A);CHKERRQ(ierr);
7924 
7925   fA = A->ops->matmulttranspose;
7926   if (!fA) SETERRQ1(((PetscObject)A)->comm,PETSC_ERR_SUP,"MatMatMultTranspose not supported for A of type %s",((PetscObject)A)->type_name);
7927   fB = B->ops->matmulttranspose;
7928   if (!fB) SETERRQ1(((PetscObject)A)->comm,PETSC_ERR_SUP,"MatMatMultTranspose not supported for B of type %s",((PetscObject)B)->type_name);
7929   if (fB!=fA) SETERRQ2(((PetscObject)A)->comm,PETSC_ERR_ARG_INCOMP,"MatMatMultTranspose requires A, %s, to be compatible with B, %s",((PetscObject)A)->type_name,((PetscObject)B)->type_name);
7930 
7931   ierr = PetscLogEventBegin(MAT_MatMultTranspose,A,B,0,0);CHKERRQ(ierr);
7932   ierr = (*A->ops->matmulttranspose)(A,B,scall,fill,C);CHKERRQ(ierr);
7933   ierr = PetscLogEventEnd(MAT_MatMultTranspose,A,B,0,0);CHKERRQ(ierr);
7934 
7935   PetscFunctionReturn(0);
7936 }
7937 
7938 #undef __FUNCT__
7939 #define __FUNCT__ "MatGetRedundantMatrix"
7940 /*@C
7941    MatGetRedundantMatrix - Create redundant matrices and put them into processors of subcommunicators.
7942 
7943    Collective on Mat
7944 
7945    Input Parameters:
7946 +  mat - the matrix
7947 .  nsubcomm - the number of subcommunicators (= number of redundant pareallel or sequential matrices)
7948 .  subcomm - MPI communicator split from the communicator where mat resides in
7949 .  mlocal_red - number of local rows of the redundant matrix
7950 -  reuse - either MAT_INITIAL_MATRIX or MAT_REUSE_MATRIX
7951 
7952    Output Parameter:
7953 .  matredundant - redundant matrix
7954 
7955    Notes:
7956    MAT_REUSE_MATRIX can only be used when the nonzero structure of the
7957    original matrix has not changed from that last call to MatGetRedundantMatrix().
7958 
7959    This routine creates the duplicated matrices in subcommunicators; you should NOT create them before
7960    calling it.
7961 
7962    Only MPIAIJ matrix is supported.
7963 
7964    Level: advanced
7965 
7966    Concepts: subcommunicator
7967    Concepts: duplicate matrix
7968 
7969 .seealso: MatDestroy()
7970 @*/
7971 PetscErrorCode PETSCMAT_DLLEXPORT MatGetRedundantMatrix(Mat mat,PetscInt nsubcomm,MPI_Comm subcomm,PetscInt mlocal_red,MatReuse reuse,Mat *matredundant)
7972 {
7973   PetscErrorCode ierr;
7974 
7975   PetscFunctionBegin;
7976   PetscValidHeaderSpecific(mat,MAT_CLASSID,1);
7977   if (nsubcomm && reuse == MAT_REUSE_MATRIX) {
7978     PetscValidPointer(*matredundant,6);
7979     PetscValidHeaderSpecific(*matredundant,MAT_CLASSID,6);
7980   }
7981   if (!mat->ops->getredundantmatrix) SETERRQ1(((PetscObject)mat)->comm,PETSC_ERR_SUP,"Mat type %s",((PetscObject)mat)->type_name);
7982   if (!mat->assembled) SETERRQ(((PetscObject)mat)->comm,PETSC_ERR_ARG_WRONGSTATE,"Not for unassembled matrix");
7983   if (mat->factortype) SETERRQ(((PetscObject)mat)->comm,PETSC_ERR_ARG_WRONGSTATE,"Not for factored matrix");
7984   ierr = MatPreallocated(mat);CHKERRQ(ierr);
7985 
7986   ierr = PetscLogEventBegin(MAT_GetRedundantMatrix,mat,0,0,0);CHKERRQ(ierr);
7987   ierr = (*mat->ops->getredundantmatrix)(mat,nsubcomm,subcomm,mlocal_red,reuse,matredundant);CHKERRQ(ierr);
7988   ierr = PetscLogEventEnd(MAT_GetRedundantMatrix,mat,0,0,0);CHKERRQ(ierr);
7989   PetscFunctionReturn(0);
7990 }
7991