Actual source code: itfunc.c
petsc-3.11.4 2019-09-28
2: /*
3: Interface KSP routines that the user calls.
4: */
6: #include <petsc/private/kspimpl.h>
7: #include <petscdm.h>
9: PETSC_STATIC_INLINE PetscErrorCode ObjectView(PetscObject obj, PetscViewer viewer, PetscViewerFormat format)
10: {
13: PetscViewerPushFormat(viewer, format);
14: PetscObjectView(obj, viewer);
15: PetscViewerPopFormat(viewer);
16: return(0);
17: }
19: /*@
20: KSPComputeExtremeSingularValues - Computes the extreme singular values
21: for the preconditioned operator. Called after or during KSPSolve().
23: Not Collective
25: Input Parameter:
26: . ksp - iterative context obtained from KSPCreate()
28: Output Parameters:
29: . emin, emax - extreme singular values
31: Options Database Keys:
32: . -ksp_compute_singularvalues - compute extreme singular values and print when KSPSolve completes.
34: Notes:
35: One must call KSPSetComputeSingularValues() before calling KSPSetUp()
36: (or use the option -ksp_compute_eigenvalues) in order for this routine to work correctly.
38: Many users may just want to use the monitoring routine
39: KSPMonitorSingularValue() (which can be set with option -ksp_monitor_singular_value)
40: to print the extreme singular values at each iteration of the linear solve.
42: Estimates of the smallest singular value may be very inaccurate, especially if the Krylov method has not converged.
43: The largest singular value is usually accurate to within a few percent if the method has converged, but is still not
44: intended for eigenanalysis.
46: Disable restarts if using KSPGMRES, otherwise this estimate will only be using those iterations after the last
47: restart. See KSPGMRESSetRestart() for more details.
49: Level: advanced
51: .keywords: compute, extreme, singular, values
53: .seealso: KSPSetComputeSingularValues(), KSPMonitorSingularValue(), KSPComputeEigenvalues(), KSP
54: @*/
55: PetscErrorCode KSPComputeExtremeSingularValues(KSP ksp,PetscReal *emax,PetscReal *emin)
56: {
63: if (!ksp->calc_sings) SETERRQ(PetscObjectComm((PetscObject)ksp),4,"Singular values not requested before KSPSetUp()");
65: if (ksp->ops->computeextremesingularvalues) {
66: (*ksp->ops->computeextremesingularvalues)(ksp,emax,emin);
67: } else {
68: *emin = -1.0;
69: *emax = -1.0;
70: }
71: return(0);
72: }
74: /*@
75: KSPComputeEigenvalues - Computes the extreme eigenvalues for the
76: preconditioned operator. Called after or during KSPSolve().
78: Not Collective
80: Input Parameter:
81: + ksp - iterative context obtained from KSPCreate()
82: - n - size of arrays r and c. The number of eigenvalues computed (neig) will, in
83: general, be less than this.
85: Output Parameters:
86: + r - real part of computed eigenvalues, provided by user with a dimension of at least n
87: . c - complex part of computed eigenvalues, provided by user with a dimension of at least n
88: - neig - actual number of eigenvalues computed (will be less than or equal to n)
90: Options Database Keys:
91: + -ksp_compute_eigenvalues - Prints eigenvalues to stdout
92: - -ksp_plot_eigenvalues - Plots eigenvalues in an x-window display
94: Notes:
95: The number of eigenvalues estimated depends on the size of the Krylov space
96: generated during the KSPSolve() ; for example, with
97: CG it corresponds to the number of CG iterations, for GMRES it is the number
98: of GMRES iterations SINCE the last restart. Any extra space in r[] and c[]
99: will be ignored.
101: KSPComputeEigenvalues() does not usually provide accurate estimates; it is
102: intended only for assistance in understanding the convergence of iterative
103: methods, not for eigenanalysis. For accurate computation of eigenvalues we recommend using
104: the excellent package SLEPc.
106: One must call KSPSetComputeEigenvalues() before calling KSPSetUp()
107: in order for this routine to work correctly.
109: Many users may just want to use the monitoring routine
110: KSPMonitorSingularValue() (which can be set with option -ksp_monitor_singular_value)
111: to print the singular values at each iteration of the linear solve.
113: Level: advanced
115: .keywords: compute, extreme, singular, values
117: .seealso: KSPSetComputeSingularValues(), KSPMonitorSingularValue(), KSPComputeExtremeSingularValues(), KSP
118: @*/
119: PetscErrorCode KSPComputeEigenvalues(KSP ksp,PetscInt n,PetscReal r[],PetscReal c[],PetscInt *neig)
120: {
127: if (n<0) SETERRQ(PetscObjectComm((PetscObject)ksp),PETSC_ERR_ARG_OUTOFRANGE,"Requested < 0 Eigenvalues");
129: if (!ksp->calc_sings) SETERRQ(PetscObjectComm((PetscObject)ksp),4,"Eigenvalues not requested before KSPSetUp()");
131: if (n && ksp->ops->computeeigenvalues) {
132: (*ksp->ops->computeeigenvalues)(ksp,n,r,c,neig);
133: } else {
134: *neig = 0;
135: }
136: return(0);
137: }
139: /*@
140: KSPComputeRitz - Computes the Ritz or harmonic Ritz pairs associated to the
141: smallest or largest in modulus, for the preconditioned operator.
142: Called after KSPSolve().
144: Not Collective
146: Input Parameter:
147: + ksp - iterative context obtained from KSPCreate()
148: . ritz - PETSC_TRUE or PETSC_FALSE for ritz pairs or harmonic Ritz pairs, respectively
149: . small - PETSC_TRUE or PETSC_FALSE for smallest or largest (harmonic) Ritz values, respectively
150: . nrit - number of (harmonic) Ritz pairs to compute
152: Output Parameters:
153: + nrit - actual number of computed (harmonic) Ritz pairs
154: . S - multidimensional vector with Ritz vectors
155: . tetar - real part of the Ritz values
156: . tetai - imaginary part of the Ritz values
158: Notes:
159: -For GMRES, the (harmonic) Ritz pairs are computed from the Hessenberg matrix obtained during
160: the last complete cycle, or obtained at the end of the solution if the method is stopped before
161: a restart. Then, the number of actual (harmonic) Ritz pairs computed is less or equal to the restart
162: parameter for GMRES if a complete cycle has been performed or less or equal to the number of GMRES
163: iterations.
164: -Moreover, for real matrices, the (harmonic) Ritz pairs are possibly complex-valued. In such a case,
165: the routine selects the complex (harmonic) Ritz value and its conjugate, and two successive columns of S
166: are equal to the real and the imaginary parts of the associated vectors.
167: -the (harmonic) Ritz pairs are given in order of increasing (harmonic) Ritz values in modulus
168: -this is currently not implemented when PETSc is built with complex numbers
170: One must call KSPSetComputeRitz() before calling KSPSetUp()
171: in order for this routine to work correctly.
173: Level: advanced
175: .keywords: compute, ritz, values
177: .seealso: KSPSetComputeRitz(), KSP
178: @*/
179: PetscErrorCode KSPComputeRitz(KSP ksp,PetscBool ritz,PetscBool small,PetscInt *nrit,Vec S[],PetscReal tetar[],PetscReal tetai[])
180: {
185: if (!ksp->calc_ritz) SETERRQ(PetscObjectComm((PetscObject)ksp),4,"Ritz pairs not requested before KSPSetUp()");
186: if (ksp->ops->computeritz) {(*ksp->ops->computeritz)(ksp,ritz,small,nrit,S,tetar,tetai);}
187: return(0);
188: }
189: /*@
190: KSPSetUpOnBlocks - Sets up the preconditioner for each block in
191: the block Jacobi, block Gauss-Seidel, and overlapping Schwarz
192: methods.
194: Collective on KSP
196: Input Parameter:
197: . ksp - the KSP context
199: Notes:
200: KSPSetUpOnBlocks() is a routine that the user can optinally call for
201: more precise profiling (via -log_view) of the setup phase for these
202: block preconditioners. If the user does not call KSPSetUpOnBlocks(),
203: it will automatically be called from within KSPSolve().
205: Calling KSPSetUpOnBlocks() is the same as calling PCSetUpOnBlocks()
206: on the PC context within the KSP context.
208: Level: advanced
210: .keywords: setup, blocks
212: .seealso: PCSetUpOnBlocks(), KSPSetUp(), PCSetUp(), KSP
213: @*/
214: PetscErrorCode KSPSetUpOnBlocks(KSP ksp)
215: {
216: PC pc;
218: PCFailedReason pcreason;
222: KSPGetPC(ksp,&pc);
223: PCSetUpOnBlocks(pc);
224: PCGetFailedReason(pc,&pcreason);
225: if (pcreason) {
226: ksp->reason = KSP_DIVERGED_PC_FAILED;
227: }
228: return(0);
229: }
231: /*@
232: KSPSetReusePreconditioner - reuse the current preconditioner, do not construct a new one even if the operator changes
234: Collective on KSP
236: Input Parameters:
237: + ksp - iterative context obtained from KSPCreate()
238: - flag - PETSC_TRUE to reuse the current preconditioner
240: Level: intermediate
242: .keywords: setup
244: .seealso: KSPCreate(), KSPSolve(), KSPDestroy(), PCSetReusePreconditioner(), KSP
245: @*/
246: PetscErrorCode KSPSetReusePreconditioner(KSP ksp,PetscBool flag)
247: {
248: PC pc;
253: KSPGetPC(ksp,&pc);
254: PCSetReusePreconditioner(pc,flag);
255: return(0);
256: }
258: /*@
259: KSPSetSkipPCSetFromOptions - prevents KSPSetFromOptions() from call PCSetFromOptions(). This is used if the same PC is shared by more than one KSP so its options are not resetable for each KSP
261: Collective on KSP
263: Input Parameters:
264: + ksp - iterative context obtained from KSPCreate()
265: - flag - PETSC_TRUE to skip calling the PCSetFromOptions()
267: Level: intermediate
269: .keywords: setup
271: .seealso: KSPCreate(), KSPSolve(), KSPDestroy(), PCSetReusePreconditioner(), KSP
272: @*/
273: PetscErrorCode KSPSetSkipPCSetFromOptions(KSP ksp,PetscBool flag)
274: {
277: ksp->skippcsetfromoptions = flag;
278: return(0);
279: }
281: /*@
282: KSPSetUp - Sets up the internal data structures for the
283: later use of an iterative solver.
285: Collective on KSP
287: Input Parameter:
288: . ksp - iterative context obtained from KSPCreate()
290: Level: developer
292: .keywords: setup
294: .seealso: KSPCreate(), KSPSolve(), KSPDestroy(), KSP
295: @*/
296: PetscErrorCode KSPSetUp(KSP ksp)
297: {
299: Mat A,B;
300: Mat mat,pmat;
301: MatNullSpace nullsp;
302: PCFailedReason pcreason;
303:
307: /* reset the convergence flag from the previous solves */
308: ksp->reason = KSP_CONVERGED_ITERATING;
310: if (!((PetscObject)ksp)->type_name) {
311: KSPSetType(ksp,KSPGMRES);
312: }
313: KSPSetUpNorms_Private(ksp,PETSC_TRUE,&ksp->normtype,&ksp->pc_side);
315: if (ksp->dmActive && !ksp->setupstage) {
316: /* first time in so build matrix and vector data structures using DM */
317: if (!ksp->vec_rhs) {DMCreateGlobalVector(ksp->dm,&ksp->vec_rhs);}
318: if (!ksp->vec_sol) {DMCreateGlobalVector(ksp->dm,&ksp->vec_sol);}
319: DMCreateMatrix(ksp->dm,&A);
320: KSPSetOperators(ksp,A,A);
321: PetscObjectDereference((PetscObject)A);
322: }
324: if (ksp->dmActive) {
325: DMKSP kdm;
326: DMGetDMKSP(ksp->dm,&kdm);
328: if (kdm->ops->computeinitialguess && ksp->setupstage != KSP_SETUP_NEWRHS) {
329: /* only computes initial guess the first time through */
330: (*kdm->ops->computeinitialguess)(ksp,ksp->vec_sol,kdm->initialguessctx);
331: KSPSetInitialGuessNonzero(ksp,PETSC_TRUE);
332: }
333: if (kdm->ops->computerhs) {
334: (*kdm->ops->computerhs)(ksp,ksp->vec_rhs,kdm->rhsctx);
335: }
337: if (ksp->setupstage != KSP_SETUP_NEWRHS) {
338: if (kdm->ops->computeoperators) {
339: KSPGetOperators(ksp,&A,&B);
340: (*kdm->ops->computeoperators)(ksp,A,B,kdm->operatorsctx);
341: } else SETERRQ(PetscObjectComm((PetscObject)ksp),PETSC_ERR_ARG_WRONGSTATE,"You called KSPSetDM() but did not use DMKSPSetComputeOperators() or KSPSetDMActive(ksp,PETSC_FALSE);");
342: }
343: }
345: if (ksp->setupstage == KSP_SETUP_NEWRHS) return(0);
346: PetscLogEventBegin(KSP_SetUp,ksp,ksp->vec_rhs,ksp->vec_sol,0);
348: switch (ksp->setupstage) {
349: case KSP_SETUP_NEW:
350: (*ksp->ops->setup)(ksp);
351: break;
352: case KSP_SETUP_NEWMATRIX: { /* This should be replaced with a more general mechanism */
353: if (ksp->setupnewmatrix) {
354: (*ksp->ops->setup)(ksp);
355: }
356: } break;
357: default: break;
358: }
360: if (!ksp->pc) {KSPGetPC(ksp,&ksp->pc);}
361: PCGetOperators(ksp->pc,&mat,&pmat);
362: /* scale the matrix if requested */
363: if (ksp->dscale) {
364: PetscScalar *xx;
365: PetscInt i,n;
366: PetscBool zeroflag = PETSC_FALSE;
367: if (!ksp->pc) {KSPGetPC(ksp,&ksp->pc);}
368: if (!ksp->diagonal) { /* allocate vector to hold diagonal */
369: MatCreateVecs(pmat,&ksp->diagonal,0);
370: }
371: MatGetDiagonal(pmat,ksp->diagonal);
372: VecGetLocalSize(ksp->diagonal,&n);
373: VecGetArray(ksp->diagonal,&xx);
374: for (i=0; i<n; i++) {
375: if (xx[i] != 0.0) xx[i] = 1.0/PetscSqrtReal(PetscAbsScalar(xx[i]));
376: else {
377: xx[i] = 1.0;
378: zeroflag = PETSC_TRUE;
379: }
380: }
381: VecRestoreArray(ksp->diagonal,&xx);
382: if (zeroflag) {
383: PetscInfo(ksp,"Zero detected in diagonal of matrix, using 1 at those locations\n");
384: }
385: MatDiagonalScale(pmat,ksp->diagonal,ksp->diagonal);
386: if (mat != pmat) {MatDiagonalScale(mat,ksp->diagonal,ksp->diagonal);}
387: ksp->dscalefix2 = PETSC_FALSE;
388: }
389: PetscLogEventEnd(KSP_SetUp,ksp,ksp->vec_rhs,ksp->vec_sol,0);
390: PCSetErrorIfFailure(ksp->pc,ksp->errorifnotconverged);
391: PCSetUp(ksp->pc);
392: PCGetFailedReason(ksp->pc,&pcreason);
393: if (pcreason) {
394: ksp->reason = KSP_DIVERGED_PC_FAILED;
395: }
397: MatGetNullSpace(mat,&nullsp);
398: if (nullsp) {
399: PetscBool test = PETSC_FALSE;
400: PetscOptionsGetBool(((PetscObject)ksp)->options,((PetscObject)ksp)->prefix,"-ksp_test_null_space",&test,NULL);
401: if (test) {
402: MatNullSpaceTest(nullsp,mat,NULL);
403: }
404: }
405: ksp->setupstage = KSP_SETUP_NEWRHS;
406: return(0);
407: }
409: static PetscErrorCode KSPReasonView_Internal(KSP ksp, PetscViewer viewer, PetscViewerFormat format)
410: {
412: PetscBool isAscii;
415: if (format != PETSC_VIEWER_DEFAULT) {PetscViewerPushFormat(viewer,format);}
416: PetscObjectTypeCompare((PetscObject)viewer,PETSCVIEWERASCII,&isAscii);
417: if (isAscii) {
418: PetscViewerASCIIAddTab(viewer,((PetscObject)ksp)->tablevel);
419: if (ksp->reason > 0) {
420: if (((PetscObject) ksp)->prefix) {
421: PetscViewerASCIIPrintf(viewer,"Linear %s solve converged due to %s iterations %D\n",((PetscObject) ksp)->prefix,KSPConvergedReasons[ksp->reason],ksp->its);
422: } else {
423: PetscViewerASCIIPrintf(viewer,"Linear solve converged due to %s iterations %D\n",KSPConvergedReasons[ksp->reason],ksp->its);
424: }
425: } else {
426: if (((PetscObject) ksp)->prefix) {
427: PetscViewerASCIIPrintf(viewer,"Linear %s solve did not converge due to %s iterations %D\n",((PetscObject) ksp)->prefix,KSPConvergedReasons[ksp->reason],ksp->its);
428: } else {
429: PetscViewerASCIIPrintf(viewer,"Linear solve did not converge due to %s iterations %D\n",KSPConvergedReasons[ksp->reason],ksp->its);
430: }
431: if (ksp->reason == KSP_DIVERGED_PC_FAILED) {
432: PCFailedReason reason;
433: PCGetFailedReason(ksp->pc,&reason);
434: PetscViewerASCIIPrintf(viewer," PC_FAILED due to %s \n",PCFailedReasons[reason]);
435: }
436: }
437: PetscViewerASCIISubtractTab(viewer,((PetscObject)ksp)->tablevel);
438: }
439: if (format != PETSC_VIEWER_DEFAULT) {PetscViewerPopFormat(viewer);}
440: return(0);
441: }
443: /*@
444: KSPReasonView - Displays the reason a KSP solve converged or diverged to a viewer
446: Collective on KSP
448: Parameter:
449: + ksp - iterative context obtained from KSPCreate()
450: - viewer - the viewer to display the reason
453: Options Database Keys:
454: . -ksp_converged_reason - print reason for converged or diverged, also prints number of iterations
456: Level: beginner
458: .keywords: solve, linear system
460: .seealso: KSPCreate(), KSPSetUp(), KSPDestroy(), KSPSetTolerances(), KSPConvergedDefault(),
461: KSPSolveTranspose(), KSPGetIterationNumber(), KSP
462: @*/
463: PetscErrorCode KSPReasonView(KSP ksp,PetscViewer viewer)
464: {
468: KSPReasonView_Internal(ksp, viewer, PETSC_VIEWER_DEFAULT);
469: return(0);
470: }
472: #if defined(PETSC_HAVE_THREADSAFETY)
473: #define KSPReasonViewFromOptions KSPReasonViewFromOptionsUnsafe
474: #else
475: #endif
476: /*@C
477: KSPReasonViewFromOptions - Processes command line options to determine if/how a KSPReason is to be viewed.
479: Collective on KSP
481: Input Parameters:
482: . ksp - the KSP object
484: Level: intermediate
486: @*/
487: PetscErrorCode KSPReasonViewFromOptions(KSP ksp)
488: {
489: PetscViewer viewer;
490: PetscBool flg;
491: PetscViewerFormat format;
492: PetscErrorCode ierr;
495: PetscOptionsGetViewer(PetscObjectComm((PetscObject)ksp),((PetscObject)ksp)->options,((PetscObject)ksp)->prefix,"-ksp_converged_reason",&viewer,&format,&flg);
496: if (flg) {
497: KSPReasonView_Internal(ksp, viewer, format);
498: PetscViewerDestroy(&viewer);
499: }
500: return(0);
501: }
503: #include <petscdraw.h>
505: static PetscErrorCode KSPViewEigenvalues_Internal(KSP ksp, PetscBool isExplicit, PetscViewer viewer, PetscViewerFormat format)
506: {
507: PetscReal *r, *c;
508: PetscInt n, i, neig;
509: PetscBool isascii, isdraw;
510: PetscMPIInt rank;
514: MPI_Comm_rank(PetscObjectComm((PetscObject) ksp), &rank);
515: PetscObjectTypeCompare((PetscObject) viewer, PETSCVIEWERASCII, &isascii);
516: PetscObjectTypeCompare((PetscObject) viewer, PETSCVIEWERDRAW, &isdraw);
517: if (isExplicit) {
518: VecGetSize(ksp->vec_sol,&n);
519: PetscMalloc2(n, &r, n, &c);
520: KSPComputeEigenvaluesExplicitly(ksp, n, r, c);
521: neig = n;
522: } else {
523: PetscInt nits;
525: KSPGetIterationNumber(ksp, &nits);
526: n = nits+2;
527: if (!nits) {PetscViewerASCIIPrintf(viewer, "Zero iterations in solver, cannot approximate any eigenvalues\n");return(0);}
528: PetscMalloc2(n, &r, n, &c);
529: KSPComputeEigenvalues(ksp, n, r, c, &neig);
530: }
531: if (isascii) {
532: PetscViewerASCIIPrintf(viewer, "%s computed eigenvalues\n", isExplicit ? "Explicitly" : "Iteratively");
533: for (i = 0; i < neig; ++i) {
534: if (c[i] >= 0.0) {PetscViewerASCIIPrintf(viewer, "%g + %gi\n", (double) r[i], (double) c[i]);}
535: else {PetscViewerASCIIPrintf(viewer, "%g - %gi\n", (double) r[i], -(double) c[i]);}
536: }
537: } else if (isdraw && !rank) {
538: PetscDraw draw;
539: PetscDrawSP drawsp;
541: if (format == PETSC_VIEWER_DRAW_CONTOUR) {
542: KSPPlotEigenContours_Private(ksp,neig,r,c);
543: } else {
544: if (!ksp->eigviewer) {PetscViewerDrawOpen(PETSC_COMM_SELF,0,isExplicit ? "Explicitly Computed Eigenvalues" : "Iteratively Computed Eigenvalues",PETSC_DECIDE,PETSC_DECIDE,400,400,&ksp->eigviewer);}
545: PetscViewerDrawGetDraw(ksp->eigviewer,0,&draw);
546: PetscDrawSPCreate(draw,1,&drawsp);
547: PetscDrawSPReset(drawsp);
548: for (i = 0; i < neig; ++i) {PetscDrawSPAddPoint(drawsp,r+i,c+i);}
549: PetscDrawSPDraw(drawsp,PETSC_TRUE);
550: PetscDrawSPSave(drawsp);
551: PetscDrawSPDestroy(&drawsp);
552: }
553: }
554: PetscFree2(r, c);
555: return(0);
556: }
558: static PetscErrorCode KSPViewSingularvalues_Internal(KSP ksp, PetscViewer viewer, PetscViewerFormat format)
559: {
560: PetscReal smax, smin;
561: PetscInt nits;
562: PetscBool isascii;
566: PetscObjectTypeCompare((PetscObject) viewer, PETSCVIEWERASCII, &isascii);
567: KSPGetIterationNumber(ksp, &nits);
568: if (!nits) {PetscViewerASCIIPrintf(viewer, "Zero iterations in solver, cannot approximate any singular values\n");return(0);}
569: KSPComputeExtremeSingularValues(ksp, &smax, &smin);
570: if (isascii) {PetscViewerASCIIPrintf(viewer, "Iteratively computed extreme singular values: max %g min %g max/min %g\n",(double)smax,(double)smin,(double)(smax/smin));}
571: return(0);
572: }
574: static PetscErrorCode KSPViewFinalResidual_Internal(KSP ksp, PetscViewer viewer, PetscViewerFormat format)
575: {
576: PetscBool isascii;
580: PetscObjectTypeCompare((PetscObject) viewer, PETSCVIEWERASCII, &isascii);
581: if (ksp->dscale && !ksp->dscalefix) SETERRQ(PetscObjectComm((PetscObject) ksp), PETSC_ERR_ARG_WRONGSTATE, "Cannot compute final scale with -ksp_diagonal_scale except also with -ksp_diagonal_scale_fix");
582: if (isascii) {
583: Mat A;
584: Vec t;
585: PetscReal norm;
587: PCGetOperators(ksp->pc, &A, NULL);
588: VecDuplicate(ksp->vec_rhs, &t);
589: KSP_MatMult(ksp, A, ksp->vec_sol, t);
590: VecAYPX(t, -1.0, ksp->vec_rhs);
591: VecNorm(t, NORM_2, &norm);
592: VecDestroy(&t);
593: PetscViewerASCIIPrintf(viewer, "KSP final norm of residual %g\n", (double) norm);
594: }
595: return(0);
596: }
598: /*@
599: KSPSolve - Solves linear system.
601: Collective on KSP
603: Parameter:
604: + ksp - iterative context obtained from KSPCreate()
605: . b - the right hand side vector
606: - x - the solution (this may be the same vector as b, then b will be overwritten with answer)
608: Options Database Keys:
609: + -ksp_compute_eigenvalues - compute preconditioned operators eigenvalues
610: . -ksp_plot_eigenvalues - plot the computed eigenvalues in an X-window
611: . -ksp_plot_eigencontours - plot the computed eigenvalues in an X-window with contours
612: . -ksp_compute_eigenvalues_explicitly - compute the eigenvalues by forming the dense operator and using LAPACK
613: . -ksp_plot_eigenvalues_explicitly - plot the explicitly computing eigenvalues
614: . -ksp_view_mat binary - save matrix to the default binary viewer
615: . -ksp_view_pmat binary - save matrix used to build preconditioner to the default binary viewer
616: . -ksp_view_rhs binary - save right hand side vector to the default binary viewer
617: . -ksp_view_solution binary - save computed solution vector to the default binary viewer
618: (can be read later with src/ksp/examples/tutorials/ex10.c for testing solvers)
619: . -ksp_view_mat_explicit - for matrix-free operators, computes the matrix entries and views them
620: . -ksp_view_preconditioned_operator_explicit - computes the product of the preconditioner and matrix as an explicit matrix and views it
621: . -ksp_converged_reason - print reason for converged or diverged, also prints number of iterations
622: . -ksp_final_residual - print 2-norm of true linear system residual at the end of the solution process
623: - -ksp_view - print the ksp data structure at the end of the system solution
625: Notes:
627: If one uses KSPSetDM() then x or b need not be passed. Use KSPGetSolution() to access the solution in this case.
629: The operator is specified with KSPSetOperators().
631: Call KSPGetConvergedReason() to determine if the solver converged or failed and
632: why. The number of iterations can be obtained from KSPGetIterationNumber().
634: If you provide a matrix that has a MatSetNullSpace() and MatSetTransposeNullSpace() this will use that information to solve singular systems
635: in the least squares sense with a norm minimizing solution.
636: $
637: $ A x = b where b = b_p + b_t where b_t is not in the range of A (and hence by the fundamental theorem of linear algebra is in the nullspace(A') see MatSetNullSpace()
638: $
639: $ KSP first removes b_t producing the linear system A x = b_p (which has multiple solutions) and solves this to find the ||x|| minimizing solution (and hence
640: $ it finds the solution x orthogonal to the nullspace(A). The algorithm is simply in each iteration of the Krylov method we remove the nullspace(A) from the search
641: $ direction thus the solution which is a linear combination of the search directions has no component in the nullspace(A).
642: $
643: $ We recommend always using GMRES for such singular systems.
644: $ If nullspace(A) = nullspace(A') (note symmetric matrices always satisfy this property) then both left and right preconditioning will work
645: $ If nullspace(A) != nullspace(A') then left preconditioning will work but right preconditioning may not work (or it may).
647: Developer Note: The reason we cannot always solve nullspace(A) != nullspace(A') systems with right preconditioning is because we need to remove at each iteration
648: the nullspace(AB) from the search direction. While we know the nullspace(A) the nullspace(AB) equals B^-1 times the nullspace(A) but except for trivial preconditioners
649: such as diagonal scaling we cannot apply the inverse of the preconditioner to a vector and thus cannot compute the nullspace(AB).
652: If using a direct method (e.g., via the KSP solver
653: KSPPREONLY and a preconditioner such as PCLU/PCILU),
654: then its=1. See KSPSetTolerances() and KSPConvergedDefault()
655: for more details.
657: Understanding Convergence:
658: The routines KSPMonitorSet(), KSPComputeEigenvalues(), and
659: KSPComputeEigenvaluesExplicitly() provide information on additional
660: options to monitor convergence and print eigenvalue information.
662: Level: beginner
664: .keywords: solve, linear system
666: .seealso: KSPCreate(), KSPSetUp(), KSPDestroy(), KSPSetTolerances(), KSPConvergedDefault(),
667: KSPSolveTranspose(), KSPGetIterationNumber(), MatNullSpaceCreate(), MatSetNullSpace(), MatSetTransposeNullSpace(), KSP
668: @*/
669: PetscErrorCode KSPSolve(KSP ksp,Vec b,Vec x)
670: {
671: PetscErrorCode ierr;
672: PetscBool flg = PETSC_FALSE,inXisinB=PETSC_FALSE,guess_zero;
673: Mat mat,pmat;
674: MPI_Comm comm;
675: MatNullSpace nullsp;
676: Vec btmp,vec_rhs=0;
682: comm = PetscObjectComm((PetscObject)ksp);
683: if (x && x == b) {
684: if (!ksp->guess_zero) SETERRQ(comm,PETSC_ERR_ARG_INCOMP,"Cannot use x == b with nonzero initial guess");
685: VecDuplicate(b,&x);
686: inXisinB = PETSC_TRUE;
687: }
688: if (b) {
689: PetscObjectReference((PetscObject)b);
690: VecDestroy(&ksp->vec_rhs);
691: ksp->vec_rhs = b;
692: }
693: if (x) {
694: PetscObjectReference((PetscObject)x);
695: VecDestroy(&ksp->vec_sol);
696: ksp->vec_sol = x;
697: }
698: if (ksp->viewPre) {ObjectView((PetscObject) ksp, ksp->viewerPre, ksp->formatPre);}
700: ksp->transpose_solve = PETSC_FALSE;
702: if (ksp->presolve) {(*ksp->presolve)(ksp,ksp->vec_rhs,ksp->vec_sol,ksp->prectx);}
704: /* reset the residual history list if requested */
705: if (ksp->res_hist_reset) ksp->res_hist_len = 0;
707: PetscLogEventBegin(KSP_Solve,ksp,ksp->vec_rhs,ksp->vec_sol,0);
709: if (ksp->guess) {
710: PetscObjectState ostate,state;
712: KSPGuessSetUp(ksp->guess);
713: PetscObjectStateGet((PetscObject)ksp->vec_sol,&ostate);
714: KSPGuessFormGuess(ksp->guess,ksp->vec_rhs,ksp->vec_sol);
715: PetscObjectStateGet((PetscObject)ksp->vec_sol,&state);
716: if (state != ostate) {
717: ksp->guess_zero = PETSC_FALSE;
718: } else {
719: PetscInfo(ksp,"Using zero initial guess since the KSPGuess object did not change the vector\n");
720: ksp->guess_zero = PETSC_TRUE;
721: }
722: }
724: /* KSPSetUp() scales the matrix if needed */
725: KSPSetUp(ksp);
726: KSPSetUpOnBlocks(ksp);
728: VecSetErrorIfLocked(ksp->vec_sol,3);
730: PCGetOperators(ksp->pc,&mat,&pmat);
731: /* diagonal scale RHS if called for */
732: if (ksp->dscale) {
733: VecPointwiseMult(ksp->vec_rhs,ksp->vec_rhs,ksp->diagonal);
734: /* second time in, but matrix was scaled back to original */
735: if (ksp->dscalefix && ksp->dscalefix2) {
736: Mat mat,pmat;
738: PCGetOperators(ksp->pc,&mat,&pmat);
739: MatDiagonalScale(pmat,ksp->diagonal,ksp->diagonal);
740: if (mat != pmat) {MatDiagonalScale(mat,ksp->diagonal,ksp->diagonal);}
741: }
743: /* scale initial guess */
744: if (!ksp->guess_zero) {
745: if (!ksp->truediagonal) {
746: VecDuplicate(ksp->diagonal,&ksp->truediagonal);
747: VecCopy(ksp->diagonal,ksp->truediagonal);
748: VecReciprocal(ksp->truediagonal);
749: }
750: VecPointwiseMult(ksp->vec_sol,ksp->vec_sol,ksp->truediagonal);
751: }
752: }
753: PCPreSolve(ksp->pc,ksp);
755: if (ksp->guess_zero) { VecSet(ksp->vec_sol,0.0);}
756: if (ksp->guess_knoll) { /* The Knoll trick is independent on the KSPGuess specified */
757: PCApply(ksp->pc,ksp->vec_rhs,ksp->vec_sol);
758: KSP_RemoveNullSpace(ksp,ksp->vec_sol);
759: ksp->guess_zero = PETSC_FALSE;
760: }
762: /* can we mark the initial guess as zero for this solve? */
763: guess_zero = ksp->guess_zero;
764: if (!ksp->guess_zero) {
765: PetscReal norm;
767: VecNormAvailable(ksp->vec_sol,NORM_2,&flg,&norm);
768: if (flg && !norm) ksp->guess_zero = PETSC_TRUE;
769: }
770: MatGetTransposeNullSpace(pmat,&nullsp);
771: if (nullsp) {
772: VecDuplicate(ksp->vec_rhs,&btmp);
773: VecCopy(ksp->vec_rhs,btmp);
774: MatNullSpaceRemove(nullsp,btmp);
775: vec_rhs = ksp->vec_rhs;
776: ksp->vec_rhs = btmp;
777: }
778: VecLockReadPush(ksp->vec_rhs);
779: if (ksp->reason == KSP_DIVERGED_PC_FAILED) {
780: VecSetInf(ksp->vec_sol);
781: }
782: (*ksp->ops->solve)(ksp);
784: VecLockReadPop(ksp->vec_rhs);
785: if (nullsp) {
786: ksp->vec_rhs = vec_rhs;
787: VecDestroy(&btmp);
788: }
790: ksp->guess_zero = guess_zero;
792: if (!ksp->reason) SETERRQ(comm,PETSC_ERR_PLIB,"Internal error, solver returned without setting converged reason");
793: ksp->totalits += ksp->its;
795: if (ksp->viewReason) {KSPReasonView_Internal(ksp, ksp->viewerReason, ksp->formatReason);}
796: PCPostSolve(ksp->pc,ksp);
798: /* diagonal scale solution if called for */
799: if (ksp->dscale) {
800: VecPointwiseMult(ksp->vec_sol,ksp->vec_sol,ksp->diagonal);
801: /* unscale right hand side and matrix */
802: if (ksp->dscalefix) {
803: Mat mat,pmat;
805: VecReciprocal(ksp->diagonal);
806: VecPointwiseMult(ksp->vec_rhs,ksp->vec_rhs,ksp->diagonal);
807: PCGetOperators(ksp->pc,&mat,&pmat);
808: MatDiagonalScale(pmat,ksp->diagonal,ksp->diagonal);
809: if (mat != pmat) {MatDiagonalScale(mat,ksp->diagonal,ksp->diagonal);}
810: VecReciprocal(ksp->diagonal);
811: ksp->dscalefix2 = PETSC_TRUE;
812: }
813: }
814: PetscLogEventEnd(KSP_Solve,ksp,ksp->vec_rhs,ksp->vec_sol,0);
815: if (ksp->guess) {
816: KSPGuessUpdate(ksp->guess,ksp->vec_rhs,ksp->vec_sol);
817: }
818: if (ksp->postsolve) {
819: (*ksp->postsolve)(ksp,ksp->vec_rhs,ksp->vec_sol,ksp->postctx);
820: }
822: PCGetOperators(ksp->pc,&mat,&pmat);
823: if (ksp->viewEV) {KSPViewEigenvalues_Internal(ksp, PETSC_FALSE, ksp->viewerEV, ksp->formatEV);}
824: if (ksp->viewEVExp) {KSPViewEigenvalues_Internal(ksp, PETSC_TRUE, ksp->viewerEVExp, ksp->formatEVExp);}
825: if (ksp->viewSV) {KSPViewSingularvalues_Internal(ksp, ksp->viewerSV, ksp->formatSV);}
826: if (ksp->viewFinalRes) {KSPViewFinalResidual_Internal(ksp, ksp->viewerFinalRes, ksp->formatFinalRes);}
827: if (ksp->viewMat) {ObjectView((PetscObject) mat, ksp->viewerMat, ksp->formatMat);}
828: if (ksp->viewPMat) {ObjectView((PetscObject) pmat, ksp->viewerPMat, ksp->formatPMat);}
829: if (ksp->viewRhs) {ObjectView((PetscObject) ksp->vec_rhs, ksp->viewerRhs, ksp->formatRhs);}
830: if (ksp->viewSol) {ObjectView((PetscObject) ksp->vec_sol, ksp->viewerSol, ksp->formatSol);}
831: if (ksp->view) {ObjectView((PetscObject) ksp, ksp->viewer, ksp->format);}
832: if (ksp->viewDScale) {ObjectView((PetscObject) ksp->diagonal, ksp->viewerDScale, ksp->formatDScale);}
833: if (ksp->viewMatExp) {
834: Mat A, B;
836: PCGetOperators(ksp->pc, &A, NULL);
837: MatComputeExplicitOperator(A, &B);
838: ObjectView((PetscObject) B, ksp->viewerMatExp, ksp->formatMatExp);
839: MatDestroy(&B);
840: }
841: if (ksp->viewPOpExp) {
842: Mat B;
844: KSPComputeExplicitOperator(ksp, &B);
845: ObjectView((PetscObject) B, ksp->viewerPOpExp, ksp->formatPOpExp);
846: MatDestroy(&B);
847: }
849: if (inXisinB) {
850: VecCopy(x,b);
851: VecDestroy(&x);
852: }
853: PetscObjectSAWsBlock((PetscObject)ksp);
854: if (ksp->errorifnotconverged && ksp->reason < 0 && ksp->reason != KSP_DIVERGED_ITS) SETERRQ1(comm,PETSC_ERR_NOT_CONVERGED,"KSPSolve has not converged, reason %s",KSPConvergedReasons[ksp->reason]);
855: return(0);
856: }
858: /*@
859: KSPSolveTranspose - Solves the transpose of a linear system.
861: Collective on KSP
863: Input Parameter:
864: + ksp - iterative context obtained from KSPCreate()
865: . b - right hand side vector
866: - x - solution vector
868: Notes:
869: For complex numbers this solve the non-Hermitian transpose system.
871: This currently does NOT correctly use the null space of the operator and its transpose for solving singular systems.
873: Developer Notes:
874: We need to implement a KSPSolveHermitianTranspose()
876: Level: developer
878: .keywords: solve, linear system
880: .seealso: KSPCreate(), KSPSetUp(), KSPDestroy(), KSPSetTolerances(), KSPConvergedDefault(),
881: KSPSolve(), KSP
882: @*/
884: PetscErrorCode KSPSolveTranspose(KSP ksp,Vec b,Vec x)
885: {
887: PetscBool inXisinB=PETSC_FALSE;
888: Vec vec_rhs = 0,btmp;
889: Mat mat,pmat;
890: MatNullSpace nullsp;
896: if (x == b) {
897: VecDuplicate(b,&x);
898: inXisinB = PETSC_TRUE;
899: }
900: PetscObjectReference((PetscObject)b);
901: PetscObjectReference((PetscObject)x);
902: VecDestroy(&ksp->vec_rhs);
903: VecDestroy(&ksp->vec_sol);
905: ksp->vec_rhs = b;
906: ksp->vec_sol = x;
907: ksp->transpose_solve = PETSC_TRUE;
909: if (ksp->presolve) {(*ksp->presolve)(ksp,ksp->vec_rhs,ksp->vec_sol,ksp->prectx);}
911: if (ksp->guess) {
912: PetscObjectState ostate,state;
914: KSPGuessSetUp(ksp->guess);
915: PetscObjectStateGet((PetscObject)ksp->vec_sol,&ostate);
916: KSPGuessFormGuess(ksp->guess,ksp->vec_rhs,ksp->vec_sol);
917: PetscObjectStateGet((PetscObject)ksp->vec_sol,&state);
918: if (state != ostate) {
919: ksp->guess_zero = PETSC_FALSE;
920: } else {
921: PetscInfo(ksp,"Using zero initial guess since the KSPGuess object did not change the vector\n");
922: ksp->guess_zero = PETSC_TRUE;
923: }
924: }
926: KSPSetUp(ksp);
927: KSPSetUpOnBlocks(ksp);
928: if (ksp->guess_zero) { VecSet(ksp->vec_sol,0.0);}
930: PCGetOperators(ksp->pc,&mat,&pmat);
931: MatGetNullSpace(pmat,&nullsp);
932: if (nullsp) {
933: VecDuplicate(ksp->vec_rhs,&btmp);
934: VecCopy(ksp->vec_rhs,btmp);
935: MatNullSpaceRemove(nullsp,btmp);
936: vec_rhs = ksp->vec_rhs;
937: ksp->vec_rhs = btmp;
938: }
940: (*ksp->ops->solve)(ksp);
941: ksp->totalits += ksp->its;
942: if (nullsp) {
943: ksp->vec_rhs = vec_rhs;
944: VecDestroy(&btmp);
945: }
946: if (!ksp->reason) SETERRQ(PetscObjectComm((PetscObject)ksp),PETSC_ERR_PLIB,"Internal error, solver returned without setting converged reason");
947: if (ksp->viewReason) {KSPReasonView_Internal(ksp, ksp->viewerReason, ksp->formatReason);}
948: if (ksp->guess) {
949: KSPGuessUpdate(ksp->guess,ksp->vec_rhs,ksp->vec_sol);
950: }
951: if (ksp->postsolve) {
952: (*ksp->postsolve)(ksp,ksp->vec_rhs,ksp->vec_sol,ksp->postctx);
953: }
955: if (ksp->viewMat) {ObjectView((PetscObject) mat, ksp->viewerMat, ksp->formatMat);}
956: if (ksp->viewPMat) {ObjectView((PetscObject) pmat, ksp->viewerPMat, ksp->formatPMat);}
957: if (ksp->viewRhs) {ObjectView((PetscObject) ksp->vec_rhs, ksp->viewerRhs, ksp->formatRhs);}
958: if (ksp->viewSol) {ObjectView((PetscObject) ksp->vec_sol, ksp->viewerSol, ksp->formatSol);}
959: if (ksp->view) {ObjectView((PetscObject) ksp, ksp->viewer, ksp->format);}
961: if (inXisinB) {
962: VecCopy(x,b);
963: VecDestroy(&x);
964: }
965: if (ksp->errorifnotconverged && ksp->reason < 0) SETERRQ(PetscObjectComm((PetscObject)ksp),PETSC_ERR_NOT_CONVERGED,"KSPSolve has not converged");
966: return(0);
967: }
969: /*@
970: KSPResetViewers - Resets all the viewers set from the options database during KSPSetFromOptions()
972: Collective on KSP
974: Input Parameter:
975: . ksp - iterative context obtained from KSPCreate()
977: Level: beginner
979: .keywords: destroy
981: .seealso: KSPCreate(), KSPSetUp(), KSPSolve(), KSPSetFromOptions(), KSP
982: @*/
983: PetscErrorCode KSPResetViewers(KSP ksp)
984: {
989: if (!ksp) return(0);
990: PetscViewerDestroy(&ksp->viewer);
991: PetscViewerDestroy(&ksp->viewerPre);
992: PetscViewerDestroy(&ksp->viewerReason);
993: PetscViewerDestroy(&ksp->viewerMat);
994: PetscViewerDestroy(&ksp->viewerPMat);
995: PetscViewerDestroy(&ksp->viewerRhs);
996: PetscViewerDestroy(&ksp->viewerSol);
997: PetscViewerDestroy(&ksp->viewerMatExp);
998: PetscViewerDestroy(&ksp->viewerEV);
999: PetscViewerDestroy(&ksp->viewerSV);
1000: PetscViewerDestroy(&ksp->viewerEVExp);
1001: PetscViewerDestroy(&ksp->viewerFinalRes);
1002: PetscViewerDestroy(&ksp->viewerPOpExp);
1003: PetscViewerDestroy(&ksp->viewerDScale);
1004: ksp->view = PETSC_FALSE;
1005: ksp->viewPre = PETSC_FALSE;
1006: ksp->viewReason = PETSC_FALSE;
1007: ksp->viewMat = PETSC_FALSE;
1008: ksp->viewPMat = PETSC_FALSE;
1009: ksp->viewRhs = PETSC_FALSE;
1010: ksp->viewSol = PETSC_FALSE;
1011: ksp->viewMatExp = PETSC_FALSE;
1012: ksp->viewEV = PETSC_FALSE;
1013: ksp->viewSV = PETSC_FALSE;
1014: ksp->viewEVExp = PETSC_FALSE;
1015: ksp->viewFinalRes = PETSC_FALSE;
1016: ksp->viewPOpExp = PETSC_FALSE;
1017: ksp->viewDScale = PETSC_FALSE;
1018: return(0);
1019: }
1021: /*@
1022: KSPReset - Resets a KSP context to the kspsetupcalled = 0 state and removes any allocated Vecs and Mats
1024: Collective on KSP
1026: Input Parameter:
1027: . ksp - iterative context obtained from KSPCreate()
1029: Level: beginner
1031: .keywords: destroy
1033: .seealso: KSPCreate(), KSPSetUp(), KSPSolve(), KSP
1034: @*/
1035: PetscErrorCode KSPReset(KSP ksp)
1036: {
1041: if (!ksp) return(0);
1042: if (ksp->ops->reset) {
1043: (*ksp->ops->reset)(ksp);
1044: }
1045: if (ksp->pc) {PCReset(ksp->pc);}
1046: if (ksp->guess) {
1047: KSPGuess guess = ksp->guess;
1048: if (guess->ops->reset) { (*guess->ops->reset)(guess); }
1049: }
1050: VecDestroyVecs(ksp->nwork,&ksp->work);
1051: VecDestroy(&ksp->vec_rhs);
1052: VecDestroy(&ksp->vec_sol);
1053: VecDestroy(&ksp->diagonal);
1054: VecDestroy(&ksp->truediagonal);
1056: KSPResetViewers(ksp);
1058: ksp->setupstage = KSP_SETUP_NEW;
1059: return(0);
1060: }
1062: /*@
1063: KSPDestroy - Destroys KSP context.
1065: Collective on KSP
1067: Input Parameter:
1068: . ksp - iterative context obtained from KSPCreate()
1070: Level: beginner
1072: .keywords: destroy
1074: .seealso: KSPCreate(), KSPSetUp(), KSPSolve(), KSP
1075: @*/
1076: PetscErrorCode KSPDestroy(KSP *ksp)
1077: {
1079: PC pc;
1082: if (!*ksp) return(0);
1084: if (--((PetscObject)(*ksp))->refct > 0) {*ksp = 0; return(0);}
1086: PetscObjectSAWsViewOff((PetscObject)*ksp);
1088: /*
1089: Avoid a cascading call to PCReset(ksp->pc) from the following call:
1090: PCReset() shouldn't be called from KSPDestroy() as it is unprotected by pc's
1091: refcount (and may be shared, e.g., by other ksps).
1092: */
1093: pc = (*ksp)->pc;
1094: (*ksp)->pc = NULL;
1095: KSPReset((*ksp));
1096: (*ksp)->pc = pc;
1097: if ((*ksp)->ops->destroy) {(*(*ksp)->ops->destroy)(*ksp);}
1099: KSPGuessDestroy(&(*ksp)->guess);
1100: DMDestroy(&(*ksp)->dm);
1101: PCDestroy(&(*ksp)->pc);
1102: PetscFree((*ksp)->res_hist_alloc);
1103: if ((*ksp)->convergeddestroy) {
1104: (*(*ksp)->convergeddestroy)((*ksp)->cnvP);
1105: }
1106: KSPMonitorCancel((*ksp));
1107: PetscViewerDestroy(&(*ksp)->eigviewer);
1108: PetscHeaderDestroy(ksp);
1109: return(0);
1110: }
1112: /*@
1113: KSPSetPCSide - Sets the preconditioning side.
1115: Logically Collective on KSP
1117: Input Parameter:
1118: . ksp - iterative context obtained from KSPCreate()
1120: Output Parameter:
1121: . side - the preconditioning side, where side is one of
1122: .vb
1123: PC_LEFT - left preconditioning (default)
1124: PC_RIGHT - right preconditioning
1125: PC_SYMMETRIC - symmetric preconditioning
1126: .ve
1128: Options Database Keys:
1129: . -ksp_pc_side <right,left,symmetric>
1131: Notes:
1132: Left preconditioning is used by default for most Krylov methods except KSPFGMRES which only supports right preconditioning.
1134: For methods changing the side of the preconditioner changes the norm type that is used, see KSPSetNormType().
1136: Symmetric preconditioning is currently available only for the KSPQCG method. Note, however, that
1137: symmetric preconditioning can be emulated by using either right or left
1138: preconditioning and a pre or post processing step.
1140: Setting the PC side often affects the default norm type. See KSPSetNormType() for details.
1142: Level: intermediate
1144: .keywords: set, right, left, symmetric, side, preconditioner, flag
1146: .seealso: KSPGetPCSide(), KSPSetNormType(), KSPGetNormType(), KSP
1147: @*/
1148: PetscErrorCode KSPSetPCSide(KSP ksp,PCSide side)
1149: {
1153: ksp->pc_side = ksp->pc_side_set = side;
1154: return(0);
1155: }
1157: /*@
1158: KSPGetPCSide - Gets the preconditioning side.
1160: Not Collective
1162: Input Parameter:
1163: . ksp - iterative context obtained from KSPCreate()
1165: Output Parameter:
1166: . side - the preconditioning side, where side is one of
1167: .vb
1168: PC_LEFT - left preconditioning (default)
1169: PC_RIGHT - right preconditioning
1170: PC_SYMMETRIC - symmetric preconditioning
1171: .ve
1173: Level: intermediate
1175: .keywords: get, right, left, symmetric, side, preconditioner, flag
1177: .seealso: KSPSetPCSide(), KSP
1178: @*/
1179: PetscErrorCode KSPGetPCSide(KSP ksp,PCSide *side)
1180: {
1186: KSPSetUpNorms_Private(ksp,PETSC_TRUE,&ksp->normtype,&ksp->pc_side);
1187: *side = ksp->pc_side;
1188: return(0);
1189: }
1191: /*@
1192: KSPGetTolerances - Gets the relative, absolute, divergence, and maximum
1193: iteration tolerances used by the default KSP convergence tests.
1195: Not Collective
1197: Input Parameter:
1198: . ksp - the Krylov subspace context
1200: Output Parameters:
1201: + rtol - the relative convergence tolerance
1202: . abstol - the absolute convergence tolerance
1203: . dtol - the divergence tolerance
1204: - maxits - maximum number of iterations
1206: Notes:
1207: The user can specify NULL for any parameter that is not needed.
1209: Level: intermediate
1211: .keywords: get, tolerance, absolute, relative, divergence, convergence,
1212: maximum, iterations
1214: .seealso: KSPSetTolerances(), KSP
1215: @*/
1216: PetscErrorCode KSPGetTolerances(KSP ksp,PetscReal *rtol,PetscReal *abstol,PetscReal *dtol,PetscInt *maxits)
1217: {
1220: if (abstol) *abstol = ksp->abstol;
1221: if (rtol) *rtol = ksp->rtol;
1222: if (dtol) *dtol = ksp->divtol;
1223: if (maxits) *maxits = ksp->max_it;
1224: return(0);
1225: }
1227: /*@
1228: KSPSetTolerances - Sets the relative, absolute, divergence, and maximum
1229: iteration tolerances used by the default KSP convergence testers.
1231: Logically Collective on KSP
1233: Input Parameters:
1234: + ksp - the Krylov subspace context
1235: . rtol - the relative convergence tolerance, relative decrease in the (possibly preconditioned) residual norm
1236: . abstol - the absolute convergence tolerance absolute size of the (possibly preconditioned) residual norm
1237: . dtol - the divergence tolerance, amount (possibly preconditioned) residual norm can increase before KSPConvergedDefault() concludes that the method is diverging
1238: - maxits - maximum number of iterations to use
1240: Options Database Keys:
1241: + -ksp_atol <abstol> - Sets abstol
1242: . -ksp_rtol <rtol> - Sets rtol
1243: . -ksp_divtol <dtol> - Sets dtol
1244: - -ksp_max_it <maxits> - Sets maxits
1246: Notes:
1247: Use PETSC_DEFAULT to retain the default value of any of the tolerances.
1249: See KSPConvergedDefault() for details how these parameters are used in the default convergence test. See also KSPSetConvergenceTest()
1250: for setting user-defined stopping criteria.
1252: Level: intermediate
1254: .keywords: set, tolerance, absolute, relative, divergence,
1255: convergence, maximum, iterations
1257: .seealso: KSPGetTolerances(), KSPConvergedDefault(), KSPSetConvergenceTest(), KSP
1258: @*/
1259: PetscErrorCode KSPSetTolerances(KSP ksp,PetscReal rtol,PetscReal abstol,PetscReal dtol,PetscInt maxits)
1260: {
1268: if (rtol != PETSC_DEFAULT) {
1269: if (rtol < 0.0 || 1.0 <= rtol) SETERRQ1(PetscObjectComm((PetscObject)ksp),PETSC_ERR_ARG_OUTOFRANGE,"Relative tolerance %g must be non-negative and less than 1.0",(double)rtol);
1270: ksp->rtol = rtol;
1271: }
1272: if (abstol != PETSC_DEFAULT) {
1273: if (abstol < 0.0) SETERRQ1(PetscObjectComm((PetscObject)ksp),PETSC_ERR_ARG_OUTOFRANGE,"Absolute tolerance %g must be non-negative",(double)abstol);
1274: ksp->abstol = abstol;
1275: }
1276: if (dtol != PETSC_DEFAULT) {
1277: if (dtol < 0.0) SETERRQ1(PetscObjectComm((PetscObject)ksp),PETSC_ERR_ARG_OUTOFRANGE,"Divergence tolerance %g must be larger than 1.0",(double)dtol);
1278: ksp->divtol = dtol;
1279: }
1280: if (maxits != PETSC_DEFAULT) {
1281: if (maxits < 0) SETERRQ1(PetscObjectComm((PetscObject)ksp),PETSC_ERR_ARG_OUTOFRANGE,"Maximum number of iterations %D must be non-negative",maxits);
1282: ksp->max_it = maxits;
1283: }
1284: return(0);
1285: }
1287: /*@
1288: KSPSetInitialGuessNonzero - Tells the iterative solver that the
1289: initial guess is nonzero; otherwise KSP assumes the initial guess
1290: is to be zero (and thus zeros it out before solving).
1292: Logically Collective on KSP
1294: Input Parameters:
1295: + ksp - iterative context obtained from KSPCreate()
1296: - flg - PETSC_TRUE indicates the guess is non-zero, PETSC_FALSE indicates the guess is zero
1298: Options database keys:
1299: . -ksp_initial_guess_nonzero : use nonzero initial guess; this takes an optional truth value (0/1/no/yes/true/false)
1301: Level: beginner
1303: Notes:
1304: If this is not called the X vector is zeroed in the call to KSPSolve().
1306: .keywords: set, initial guess, nonzero
1308: .seealso: KSPGetInitialGuessNonzero(), KSPSetGuessType(), KSPGuessType, KSP
1309: @*/
1310: PetscErrorCode KSPSetInitialGuessNonzero(KSP ksp,PetscBool flg)
1311: {
1315: ksp->guess_zero = (PetscBool) !(int)flg;
1316: return(0);
1317: }
1319: /*@
1320: KSPGetInitialGuessNonzero - Determines whether the KSP solver is using
1321: a zero initial guess.
1323: Not Collective
1325: Input Parameter:
1326: . ksp - iterative context obtained from KSPCreate()
1328: Output Parameter:
1329: . flag - PETSC_TRUE if guess is nonzero, else PETSC_FALSE
1331: Level: intermediate
1333: .keywords: set, initial guess, nonzero
1335: .seealso: KSPSetInitialGuessNonzero(), KSP
1336: @*/
1337: PetscErrorCode KSPGetInitialGuessNonzero(KSP ksp,PetscBool *flag)
1338: {
1342: if (ksp->guess_zero) *flag = PETSC_FALSE;
1343: else *flag = PETSC_TRUE;
1344: return(0);
1345: }
1347: /*@
1348: KSPSetErrorIfNotConverged - Causes KSPSolve() to generate an error if the solver has not converged.
1350: Logically Collective on KSP
1352: Input Parameters:
1353: + ksp - iterative context obtained from KSPCreate()
1354: - flg - PETSC_TRUE indicates you want the error generated
1356: Options database keys:
1357: . -ksp_error_if_not_converged : this takes an optional truth value (0/1/no/yes/true/false)
1359: Level: intermediate
1361: Notes:
1362: Normally PETSc continues if a linear solver fails to converge, you can call KSPGetConvergedReason() after a KSPSolve()
1363: to determine if it has converged.
1366: .seealso: KSPGetErrorIfNotConverged(), KSP
1367: @*/
1368: PetscErrorCode KSPSetErrorIfNotConverged(KSP ksp,PetscBool flg)
1369: {
1373: ksp->errorifnotconverged = flg;
1374: return(0);
1375: }
1377: /*@
1378: KSPGetErrorIfNotConverged - Will KSPSolve() generate an error if the solver does not converge?
1380: Not Collective
1382: Input Parameter:
1383: . ksp - iterative context obtained from KSPCreate()
1385: Output Parameter:
1386: . flag - PETSC_TRUE if it will generate an error, else PETSC_FALSE
1388: Level: intermediate
1390: .seealso: KSPSetErrorIfNotConverged(), KSP
1391: @*/
1392: PetscErrorCode KSPGetErrorIfNotConverged(KSP ksp,PetscBool *flag)
1393: {
1397: *flag = ksp->errorifnotconverged;
1398: return(0);
1399: }
1401: /*@
1402: KSPSetInitialGuessKnoll - Tells the iterative solver to use PCApply(pc,b,..) to compute the initial guess (The Knoll trick)
1404: Logically Collective on KSP
1406: Input Parameters:
1407: + ksp - iterative context obtained from KSPCreate()
1408: - flg - PETSC_TRUE or PETSC_FALSE
1410: Level: advanced
1412: Developer Note: the Knoll trick is not currently implemented using the KSPGuess class
1414: .keywords: set, initial guess, nonzero
1416: .seealso: KSPGetInitialGuessKnoll(), KSPSetInitialGuessNonzero(), KSPGetInitialGuessNonzero(), KSP
1417: @*/
1418: PetscErrorCode KSPSetInitialGuessKnoll(KSP ksp,PetscBool flg)
1419: {
1423: ksp->guess_knoll = flg;
1424: return(0);
1425: }
1427: /*@
1428: KSPGetInitialGuessKnoll - Determines whether the KSP solver is using the Knoll trick (using PCApply(pc,b,...) to compute
1429: the initial guess
1431: Not Collective
1433: Input Parameter:
1434: . ksp - iterative context obtained from KSPCreate()
1436: Output Parameter:
1437: . flag - PETSC_TRUE if using Knoll trick, else PETSC_FALSE
1439: Level: advanced
1441: .keywords: set, initial guess, nonzero
1443: .seealso: KSPSetInitialGuessKnoll(), KSPSetInitialGuessNonzero(), KSPGetInitialGuessNonzero(), KSP
1444: @*/
1445: PetscErrorCode KSPGetInitialGuessKnoll(KSP ksp,PetscBool *flag)
1446: {
1450: *flag = ksp->guess_knoll;
1451: return(0);
1452: }
1454: /*@
1455: KSPGetComputeSingularValues - Gets the flag indicating whether the extreme singular
1456: values will be calculated via a Lanczos or Arnoldi process as the linear
1457: system is solved.
1459: Not Collective
1461: Input Parameter:
1462: . ksp - iterative context obtained from KSPCreate()
1464: Output Parameter:
1465: . flg - PETSC_TRUE or PETSC_FALSE
1467: Options Database Key:
1468: . -ksp_monitor_singular_value - Activates KSPSetComputeSingularValues()
1470: Notes:
1471: Currently this option is not valid for all iterative methods.
1473: Many users may just want to use the monitoring routine
1474: KSPMonitorSingularValue() (which can be set with option -ksp_monitor_singular_value)
1475: to print the singular values at each iteration of the linear solve.
1477: Level: advanced
1479: .keywords: set, compute, singular values
1481: .seealso: KSPComputeExtremeSingularValues(), KSPMonitorSingularValue(), KSP
1482: @*/
1483: PetscErrorCode KSPGetComputeSingularValues(KSP ksp,PetscBool *flg)
1484: {
1488: *flg = ksp->calc_sings;
1489: return(0);
1490: }
1492: /*@
1493: KSPSetComputeSingularValues - Sets a flag so that the extreme singular
1494: values will be calculated via a Lanczos or Arnoldi process as the linear
1495: system is solved.
1497: Logically Collective on KSP
1499: Input Parameters:
1500: + ksp - iterative context obtained from KSPCreate()
1501: - flg - PETSC_TRUE or PETSC_FALSE
1503: Options Database Key:
1504: . -ksp_monitor_singular_value - Activates KSPSetComputeSingularValues()
1506: Notes:
1507: Currently this option is not valid for all iterative methods.
1509: Many users may just want to use the monitoring routine
1510: KSPMonitorSingularValue() (which can be set with option -ksp_monitor_singular_value)
1511: to print the singular values at each iteration of the linear solve.
1513: Level: advanced
1515: .keywords: set, compute, singular values
1517: .seealso: KSPComputeExtremeSingularValues(), KSPMonitorSingularValue(), KSP
1518: @*/
1519: PetscErrorCode KSPSetComputeSingularValues(KSP ksp,PetscBool flg)
1520: {
1524: ksp->calc_sings = flg;
1525: return(0);
1526: }
1528: /*@
1529: KSPGetComputeEigenvalues - Gets the flag indicating that the extreme eigenvalues
1530: values will be calculated via a Lanczos or Arnoldi process as the linear
1531: system is solved.
1533: Not Collective
1535: Input Parameter:
1536: . ksp - iterative context obtained from KSPCreate()
1538: Output Parameter:
1539: . flg - PETSC_TRUE or PETSC_FALSE
1541: Notes:
1542: Currently this option is not valid for all iterative methods.
1544: Level: advanced
1546: .keywords: set, compute, eigenvalues
1548: .seealso: KSPComputeEigenvalues(), KSPComputeEigenvaluesExplicitly(), KSP
1549: @*/
1550: PetscErrorCode KSPGetComputeEigenvalues(KSP ksp,PetscBool *flg)
1551: {
1555: *flg = ksp->calc_sings;
1556: return(0);
1557: }
1559: /*@
1560: KSPSetComputeEigenvalues - Sets a flag so that the extreme eigenvalues
1561: values will be calculated via a Lanczos or Arnoldi process as the linear
1562: system is solved.
1564: Logically Collective on KSP
1566: Input Parameters:
1567: + ksp - iterative context obtained from KSPCreate()
1568: - flg - PETSC_TRUE or PETSC_FALSE
1570: Notes:
1571: Currently this option is not valid for all iterative methods.
1573: Level: advanced
1575: .keywords: set, compute, eigenvalues
1577: .seealso: KSPComputeEigenvalues(), KSPComputeEigenvaluesExplicitly(), KSP
1578: @*/
1579: PetscErrorCode KSPSetComputeEigenvalues(KSP ksp,PetscBool flg)
1580: {
1584: ksp->calc_sings = flg;
1585: return(0);
1586: }
1588: /*@
1589: KSPSetComputeRitz - Sets a flag so that the Ritz or harmonic Ritz pairs
1590: will be calculated via a Lanczos or Arnoldi process as the linear
1591: system is solved.
1593: Logically Collective on KSP
1595: Input Parameters:
1596: + ksp - iterative context obtained from KSPCreate()
1597: - flg - PETSC_TRUE or PETSC_FALSE
1599: Notes:
1600: Currently this option is only valid for the GMRES method.
1602: Level: advanced
1604: .keywords: set, compute, ritz
1606: .seealso: KSPComputeRitz(), KSP
1607: @*/
1608: PetscErrorCode KSPSetComputeRitz(KSP ksp, PetscBool flg)
1609: {
1613: ksp->calc_ritz = flg;
1614: return(0);
1615: }
1617: /*@
1618: KSPGetRhs - Gets the right-hand-side vector for the linear system to
1619: be solved.
1621: Not Collective
1623: Input Parameter:
1624: . ksp - iterative context obtained from KSPCreate()
1626: Output Parameter:
1627: . r - right-hand-side vector
1629: Level: developer
1631: .keywords: get, right-hand-side, rhs
1633: .seealso: KSPGetSolution(), KSPSolve(), KSP
1634: @*/
1635: PetscErrorCode KSPGetRhs(KSP ksp,Vec *r)
1636: {
1640: *r = ksp->vec_rhs;
1641: return(0);
1642: }
1644: /*@
1645: KSPGetSolution - Gets the location of the solution for the
1646: linear system to be solved. Note that this may not be where the solution
1647: is stored during the iterative process; see KSPBuildSolution().
1649: Not Collective
1651: Input Parameters:
1652: . ksp - iterative context obtained from KSPCreate()
1654: Output Parameters:
1655: . v - solution vector
1657: Level: developer
1659: .keywords: get, solution
1661: .seealso: KSPGetRhs(), KSPBuildSolution(), KSPSolve(), KSP
1662: @*/
1663: PetscErrorCode KSPGetSolution(KSP ksp,Vec *v)
1664: {
1668: *v = ksp->vec_sol;
1669: return(0);
1670: }
1672: /*@
1673: KSPSetPC - Sets the preconditioner to be used to calculate the
1674: application of the preconditioner on a vector.
1676: Collective on KSP
1678: Input Parameters:
1679: + ksp - iterative context obtained from KSPCreate()
1680: - pc - the preconditioner object
1682: Notes:
1683: Use KSPGetPC() to retrieve the preconditioner context (for example,
1684: to free it at the end of the computations).
1686: Level: developer
1688: .keywords: set, precondition, Binv
1690: .seealso: KSPGetPC(), KSP
1691: @*/
1692: PetscErrorCode KSPSetPC(KSP ksp,PC pc)
1693: {
1700: PetscObjectReference((PetscObject)pc);
1701: PCDestroy(&ksp->pc);
1702: ksp->pc = pc;
1703: PetscLogObjectParent((PetscObject)ksp,(PetscObject)ksp->pc);
1704: return(0);
1705: }
1707: /*@
1708: KSPGetPC - Returns a pointer to the preconditioner context
1709: set with KSPSetPC().
1711: Not Collective
1713: Input Parameters:
1714: . ksp - iterative context obtained from KSPCreate()
1716: Output Parameter:
1717: . pc - preconditioner context
1719: Level: developer
1721: .keywords: get, preconditioner, Binv
1723: .seealso: KSPSetPC(), KSP
1724: @*/
1725: PetscErrorCode KSPGetPC(KSP ksp,PC *pc)
1726: {
1732: if (!ksp->pc) {
1733: PCCreate(PetscObjectComm((PetscObject)ksp),&ksp->pc);
1734: PetscObjectIncrementTabLevel((PetscObject)ksp->pc,(PetscObject)ksp,0);
1735: PetscLogObjectParent((PetscObject)ksp,(PetscObject)ksp->pc);
1736: PetscObjectSetOptions((PetscObject)ksp->pc,((PetscObject)ksp)->options);
1737: }
1738: *pc = ksp->pc;
1739: return(0);
1740: }
1742: /*@
1743: KSPMonitor - runs the user provided monitor routines, if they exist
1745: Collective on KSP
1747: Input Parameters:
1748: + ksp - iterative context obtained from KSPCreate()
1749: . it - iteration number
1750: - rnorm - relative norm of the residual
1752: Notes:
1753: This routine is called by the KSP implementations.
1754: It does not typically need to be called by the user.
1756: Level: developer
1758: .seealso: KSPMonitorSet()
1759: @*/
1760: PetscErrorCode KSPMonitor(KSP ksp,PetscInt it,PetscReal rnorm)
1761: {
1762: PetscInt i, n = ksp->numbermonitors;
1766: for (i=0; i<n; i++) {
1767: (*ksp->monitor[i])(ksp,it,rnorm,ksp->monitorcontext[i]);
1768: }
1769: return(0);
1770: }
1772: /*
1774: Checks if two monitors are identical; if they are then it destroys the new one
1775: */
1776: PetscErrorCode PetscMonitorCompare(PetscErrorCode (*nmon)(void),void *nmctx,PetscErrorCode (*nmdestroy)(void**),PetscErrorCode (*mon)(void),void *mctx,PetscErrorCode (*mdestroy)(void**),PetscBool *identical)
1777: {
1778: *identical = PETSC_FALSE;
1779: if (nmon == mon && nmdestroy == mdestroy) {
1780: if (nmctx == mctx) *identical = PETSC_TRUE;
1781: else if (nmdestroy == (PetscErrorCode (*)(void**)) PetscViewerAndFormatDestroy) {
1782: PetscViewerAndFormat *old = (PetscViewerAndFormat*)mctx, *newo = (PetscViewerAndFormat*)nmctx;
1783: if (old->viewer == newo->viewer && old->format == newo->format) *identical = PETSC_TRUE;
1784: }
1785: if (*identical) {
1786: if (mdestroy) {
1788: (*mdestroy)(&nmctx);
1789: }
1790: }
1791: }
1792: return(0);
1793: }
1795: /*@C
1796: KSPMonitorSet - Sets an ADDITIONAL function to be called at every iteration to monitor
1797: the residual/error etc.
1799: Logically Collective on KSP
1801: Input Parameters:
1802: + ksp - iterative context obtained from KSPCreate()
1803: . monitor - pointer to function (if this is NULL, it turns off monitoring
1804: . mctx - [optional] context for private data for the
1805: monitor routine (use NULL if no context is desired)
1806: - monitordestroy - [optional] routine that frees monitor context
1807: (may be NULL)
1809: Calling Sequence of monitor:
1810: $ monitor (KSP ksp, int it, PetscReal rnorm, void *mctx)
1812: + ksp - iterative context obtained from KSPCreate()
1813: . it - iteration number
1814: . rnorm - (estimated) 2-norm of (preconditioned) residual
1815: - mctx - optional monitoring context, as set by KSPMonitorSet()
1817: Options Database Keys:
1818: + -ksp_monitor - sets KSPMonitorDefault()
1819: . -ksp_monitor_true_residual - sets KSPMonitorTrueResidualNorm()
1820: . -ksp_monitor_max - sets KSPMonitorTrueResidualMaxNorm()
1821: . -ksp_monitor_lg_residualnorm - sets line graph monitor,
1822: uses KSPMonitorLGResidualNormCreate()
1823: . -ksp_monitor_lg_true_residualnorm - sets line graph monitor,
1824: uses KSPMonitorLGResidualNormCreate()
1825: . -ksp_monitor_singular_value - sets KSPMonitorSingularValue()
1826: - -ksp_monitor_cancel - cancels all monitors that have
1827: been hardwired into a code by
1828: calls to KSPMonitorSet(), but
1829: does not cancel those set via
1830: the options database.
1832: Notes:
1833: The default is to do nothing. To print the residual, or preconditioned
1834: residual if KSPSetNormType(ksp,KSP_NORM_PRECONDITIONED) was called, use
1835: KSPMonitorDefault() as the monitoring routine, with a ASCII viewer as the
1836: context.
1838: Several different monitoring routines may be set by calling
1839: KSPMonitorSet() multiple times; all will be called in the
1840: order in which they were set.
1842: Fortran Notes:
1843: Only a single monitor function can be set for each KSP object
1845: Level: beginner
1847: .keywords: set, monitor
1849: .seealso: KSPMonitorDefault(), KSPMonitorLGResidualNormCreate(), KSPMonitorCancel(), KSP
1850: @*/
1851: PetscErrorCode KSPMonitorSet(KSP ksp,PetscErrorCode (*monitor)(KSP,PetscInt,PetscReal,void*),void *mctx,PetscErrorCode (*monitordestroy)(void**))
1852: {
1853: PetscInt i;
1855: PetscBool identical;
1859: for (i=0; i<ksp->numbermonitors;i++) {
1860: PetscMonitorCompare((PetscErrorCode (*)(void))monitor,mctx,monitordestroy,(PetscErrorCode (*)(void))ksp->monitor[i],ksp->monitorcontext[i],ksp->monitordestroy[i],&identical);
1861: if (identical) return(0);
1862: }
1863: if (ksp->numbermonitors >= MAXKSPMONITORS) SETERRQ(PetscObjectComm((PetscObject)ksp),PETSC_ERR_ARG_OUTOFRANGE,"Too many KSP monitors set");
1864: ksp->monitor[ksp->numbermonitors] = monitor;
1865: ksp->monitordestroy[ksp->numbermonitors] = monitordestroy;
1866: ksp->monitorcontext[ksp->numbermonitors++] = (void*)mctx;
1867: return(0);
1868: }
1870: /*@
1871: KSPMonitorCancel - Clears all monitors for a KSP object.
1873: Logically Collective on KSP
1875: Input Parameters:
1876: . ksp - iterative context obtained from KSPCreate()
1878: Options Database Key:
1879: . -ksp_monitor_cancel - Cancels all monitors that have
1880: been hardwired into a code by calls to KSPMonitorSet(),
1881: but does not cancel those set via the options database.
1883: Level: intermediate
1885: .keywords: set, monitor
1887: .seealso: KSPMonitorDefault(), KSPMonitorLGResidualNormCreate(), KSPMonitorSet(), KSP
1888: @*/
1889: PetscErrorCode KSPMonitorCancel(KSP ksp)
1890: {
1892: PetscInt i;
1896: for (i=0; i<ksp->numbermonitors; i++) {
1897: if (ksp->monitordestroy[i]) {
1898: (*ksp->monitordestroy[i])(&ksp->monitorcontext[i]);
1899: }
1900: }
1901: ksp->numbermonitors = 0;
1902: return(0);
1903: }
1905: /*@C
1906: KSPGetMonitorContext - Gets the monitoring context, as set by
1907: KSPMonitorSet() for the FIRST monitor only.
1909: Not Collective
1911: Input Parameter:
1912: . ksp - iterative context obtained from KSPCreate()
1914: Output Parameter:
1915: . ctx - monitoring context
1917: Level: intermediate
1919: .keywords: get, monitor, context
1921: .seealso: KSPMonitorDefault(), KSPMonitorLGResidualNormCreate(), KSP
1922: @*/
1923: PetscErrorCode KSPGetMonitorContext(KSP ksp,void **ctx)
1924: {
1927: *ctx = (ksp->monitorcontext[0]);
1928: return(0);
1929: }
1931: /*@
1932: KSPSetResidualHistory - Sets the array used to hold the residual history.
1933: If set, this array will contain the residual norms computed at each
1934: iteration of the solver.
1936: Not Collective
1938: Input Parameters:
1939: + ksp - iterative context obtained from KSPCreate()
1940: . a - array to hold history
1941: . na - size of a
1942: - reset - PETSC_TRUE indicates the history counter is reset to zero
1943: for each new linear solve
1945: Level: advanced
1947: Notes:
1948: The array is NOT freed by PETSc so the user needs to keep track of
1949: it and destroy once the KSP object is destroyed.
1951: If 'a' is NULL then space is allocated for the history. If 'na' PETSC_DECIDE or PETSC_DEFAULT then a
1952: default array of length 10000 is allocated.
1954: .keywords: set, residual, history, norm
1956: .seealso: KSPGetResidualHistory(), KSP
1958: @*/
1959: PetscErrorCode KSPSetResidualHistory(KSP ksp,PetscReal a[],PetscInt na,PetscBool reset)
1960: {
1966: PetscFree(ksp->res_hist_alloc);
1967: if (na != PETSC_DECIDE && na != PETSC_DEFAULT && a) {
1968: ksp->res_hist = a;
1969: ksp->res_hist_max = na;
1970: } else {
1971: if (na != PETSC_DECIDE && na != PETSC_DEFAULT) ksp->res_hist_max = na;
1972: else ksp->res_hist_max = 10000; /* like default ksp->max_it */
1973: PetscCalloc1(ksp->res_hist_max,&ksp->res_hist_alloc);
1975: ksp->res_hist = ksp->res_hist_alloc;
1976: }
1977: ksp->res_hist_len = 0;
1978: ksp->res_hist_reset = reset;
1979: return(0);
1980: }
1982: /*@C
1983: KSPGetResidualHistory - Gets the array used to hold the residual history
1984: and the number of residuals it contains.
1986: Not Collective
1988: Input Parameter:
1989: . ksp - iterative context obtained from KSPCreate()
1991: Output Parameters:
1992: + a - pointer to array to hold history (or NULL)
1993: - na - number of used entries in a (or NULL)
1995: Level: advanced
1997: Notes:
1998: Can only be called after a KSPSetResidualHistory() otherwise a and na are set to zero
2000: The Fortran version of this routine has a calling sequence
2001: $ call KSPGetResidualHistory(KSP ksp, integer na, integer ierr)
2002: note that you have passed a Fortran array into KSPSetResidualHistory() and you need
2003: to access the residual values from this Fortran array you provided. Only the na (number of
2004: residual norms currently held) is set.
2006: .keywords: get, residual, history, norm
2008: .seealso: KSPGetResidualHistory(), KSP
2010: @*/
2011: PetscErrorCode KSPGetResidualHistory(KSP ksp,PetscReal *a[],PetscInt *na)
2012: {
2015: if (a) *a = ksp->res_hist;
2016: if (na) *na = ksp->res_hist_len;
2017: return(0);
2018: }
2020: /*@C
2021: KSPSetConvergenceTest - Sets the function to be used to determine
2022: convergence.
2024: Logically Collective on KSP
2026: Input Parameters:
2027: + ksp - iterative context obtained from KSPCreate()
2028: . converge - pointer to int function
2029: . cctx - context for private data for the convergence routine (may be null)
2030: - destroy - a routine for destroying the context (may be null)
2032: Calling sequence of converge:
2033: $ converge (KSP ksp, int it, PetscReal rnorm, KSPConvergedReason *reason,void *mctx)
2035: + ksp - iterative context obtained from KSPCreate()
2036: . it - iteration number
2037: . rnorm - (estimated) 2-norm of (preconditioned) residual
2038: . reason - the reason why it has converged or diverged
2039: - cctx - optional convergence context, as set by KSPSetConvergenceTest()
2042: Notes:
2043: Must be called after the KSP type has been set so put this after
2044: a call to KSPSetType(), or KSPSetFromOptions().
2046: The default convergence test, KSPConvergedDefault(), aborts if the
2047: residual grows to more than 10000 times the initial residual.
2049: The default is a combination of relative and absolute tolerances.
2050: The residual value that is tested may be an approximation; routines
2051: that need exact values should compute them.
2053: In the default PETSc convergence test, the precise values of reason
2054: are macros such as KSP_CONVERGED_RTOL, which are defined in petscksp.h.
2056: Level: advanced
2058: .keywords: set, convergence, test, context
2060: .seealso: KSPConvergedDefault(), KSPGetConvergenceContext(), KSPSetTolerances(), KSP, KSPGetConvergenceTest(), KSPGetAndClearConvergenceTest()
2061: @*/
2062: PetscErrorCode KSPSetConvergenceTest(KSP ksp,PetscErrorCode (*converge)(KSP,PetscInt,PetscReal,KSPConvergedReason*,void*),void *cctx,PetscErrorCode (*destroy)(void*))
2063: {
2068: if (ksp->convergeddestroy) {
2069: (*ksp->convergeddestroy)(ksp->cnvP);
2070: }
2071: ksp->converged = converge;
2072: ksp->convergeddestroy = destroy;
2073: ksp->cnvP = (void*)cctx;
2074: return(0);
2075: }
2077: /*@C
2078: KSPGetConvergenceTest - Gets the function to be used to determine
2079: convergence.
2081: Logically Collective on KSP
2083: Input Parameter:
2084: . ksp - iterative context obtained from KSPCreate()
2086: Output Parameter:
2087: + converge - pointer to convergence test function
2088: . cctx - context for private data for the convergence routine (may be null)
2089: - destroy - a routine for destroying the context (may be null)
2091: Calling sequence of converge:
2092: $ converge (KSP ksp, int it, PetscReal rnorm, KSPConvergedReason *reason,void *mctx)
2094: + ksp - iterative context obtained from KSPCreate()
2095: . it - iteration number
2096: . rnorm - (estimated) 2-norm of (preconditioned) residual
2097: . reason - the reason why it has converged or diverged
2098: - cctx - optional convergence context, as set by KSPSetConvergenceTest()
2100: Level: advanced
2102: .keywords: set, convergence, test, context
2104: .seealso: KSPConvergedDefault(), KSPGetConvergenceContext(), KSPSetTolerances(), KSP, KSPSetConvergenceTest(), KSPGetAndClearConvergenceTest()
2105: @*/
2106: PetscErrorCode KSPGetConvergenceTest(KSP ksp,PetscErrorCode (**converge)(KSP,PetscInt,PetscReal,KSPConvergedReason*,void*),void **cctx,PetscErrorCode (**destroy)(void*))
2107: {
2110: if (converge) *converge = ksp->converged;
2111: if (destroy) *destroy = ksp->convergeddestroy;
2112: if (cctx) *cctx = ksp->cnvP;
2113: return(0);
2114: }
2116: /*@C
2117: KSPGetAndClearConvergenceTest - Gets the function to be used to determine convergence. Removes the current test without calling destroy on the test context
2119: Logically Collective on KSP
2121: Input Parameter:
2122: . ksp - iterative context obtained from KSPCreate()
2124: Output Parameter:
2125: + converge - pointer to convergence test function
2126: . cctx - context for private data for the convergence routine
2127: - destroy - a routine for destroying the context
2129: Calling sequence of converge:
2130: $ converge (KSP ksp, int it, PetscReal rnorm, KSPConvergedReason *reason,void *mctx)
2132: + ksp - iterative context obtained from KSPCreate()
2133: . it - iteration number
2134: . rnorm - (estimated) 2-norm of (preconditioned) residual
2135: . reason - the reason why it has converged or diverged
2136: - cctx - optional convergence context, as set by KSPSetConvergenceTest()
2138: Level: advanced
2140: Notes: This is intended to be used to allow transfering the convergence test (and its context) to another testing object (for example another KSP) and then calling
2141: KSPSetConvergenceTest() on this original KSP. If you just called KSPGetConvergenceTest() followed by KSPSetConvergenceTest() the original context information
2142: would be destroyed and hence the transfered context would be invalid and trigger a crash on use
2144: .keywords: set, convergence, test, context
2146: .seealso: KSPConvergedDefault(), KSPGetConvergenceContext(), KSPSetTolerances(), KSP, KSPSetConvergenceTest(), KSPGetConvergenceTest()
2147: @*/
2148: PetscErrorCode KSPGetAndClearConvergenceTest(KSP ksp,PetscErrorCode (**converge)(KSP,PetscInt,PetscReal,KSPConvergedReason*,void*),void **cctx,PetscErrorCode (**destroy)(void*))
2149: {
2152: *converge = ksp->converged;
2153: *destroy = ksp->convergeddestroy;
2154: *cctx = ksp->cnvP;
2155: ksp->converged = NULL;
2156: ksp->cnvP = NULL;
2157: ksp->convergeddestroy = NULL;
2158: return(0);
2159: }
2161: /*@C
2162: KSPGetConvergenceContext - Gets the convergence context set with
2163: KSPSetConvergenceTest().
2165: Not Collective
2167: Input Parameter:
2168: . ksp - iterative context obtained from KSPCreate()
2170: Output Parameter:
2171: . ctx - monitoring context
2173: Level: advanced
2175: .keywords: get, convergence, test, context
2177: .seealso: KSPConvergedDefault(), KSPSetConvergenceTest(), KSP
2178: @*/
2179: PetscErrorCode KSPGetConvergenceContext(KSP ksp,void **ctx)
2180: {
2183: *ctx = ksp->cnvP;
2184: return(0);
2185: }
2187: /*@C
2188: KSPBuildSolution - Builds the approximate solution in a vector provided.
2189: This routine is NOT commonly needed (see KSPSolve()).
2191: Collective on KSP
2193: Input Parameter:
2194: . ctx - iterative context obtained from KSPCreate()
2196: Output Parameter:
2197: Provide exactly one of
2198: + v - location to stash solution.
2199: - V - the solution is returned in this location. This vector is created
2200: internally. This vector should NOT be destroyed by the user with
2201: VecDestroy().
2203: Notes:
2204: This routine can be used in one of two ways
2205: .vb
2206: KSPBuildSolution(ksp,NULL,&V);
2207: or
2208: KSPBuildSolution(ksp,v,NULL); or KSPBuildSolution(ksp,v,&v);
2209: .ve
2210: In the first case an internal vector is allocated to store the solution
2211: (the user cannot destroy this vector). In the second case the solution
2212: is generated in the vector that the user provides. Note that for certain
2213: methods, such as KSPCG, the second case requires a copy of the solution,
2214: while in the first case the call is essentially free since it simply
2215: returns the vector where the solution already is stored. For some methods
2216: like GMRES this is a reasonably expensive operation and should only be
2217: used in truly needed.
2219: Level: advanced
2221: .keywords: build, solution
2223: .seealso: KSPGetSolution(), KSPBuildResidual(), KSP
2224: @*/
2225: PetscErrorCode KSPBuildSolution(KSP ksp,Vec v,Vec *V)
2226: {
2231: if (!V && !v) SETERRQ(PetscObjectComm((PetscObject)ksp),PETSC_ERR_ARG_WRONG,"Must provide either v or V");
2232: if (!V) V = &v;
2233: (*ksp->ops->buildsolution)(ksp,v,V);
2234: return(0);
2235: }
2237: /*@C
2238: KSPBuildResidual - Builds the residual in a vector provided.
2240: Collective on KSP
2242: Input Parameter:
2243: . ksp - iterative context obtained from KSPCreate()
2245: Output Parameters:
2246: + v - optional location to stash residual. If v is not provided,
2247: then a location is generated.
2248: . t - work vector. If not provided then one is generated.
2249: - V - the residual
2251: Notes:
2252: Regardless of whether or not v is provided, the residual is
2253: returned in V.
2255: Level: advanced
2257: .keywords: KSP, build, residual
2259: .seealso: KSPBuildSolution()
2260: @*/
2261: PetscErrorCode KSPBuildResidual(KSP ksp,Vec t,Vec v,Vec *V)
2262: {
2264: PetscBool flag = PETSC_FALSE;
2265: Vec w = v,tt = t;
2269: if (!w) {
2270: VecDuplicate(ksp->vec_rhs,&w);
2271: PetscLogObjectParent((PetscObject)ksp,(PetscObject)w);
2272: }
2273: if (!tt) {
2274: VecDuplicate(ksp->vec_sol,&tt); flag = PETSC_TRUE;
2275: PetscLogObjectParent((PetscObject)ksp,(PetscObject)tt);
2276: }
2277: (*ksp->ops->buildresidual)(ksp,tt,w,V);
2278: if (flag) {VecDestroy(&tt);}
2279: return(0);
2280: }
2282: /*@
2283: KSPSetDiagonalScale - Tells KSP to symmetrically diagonally scale the system
2284: before solving. This actually CHANGES the matrix (and right hand side).
2286: Logically Collective on KSP
2288: Input Parameter:
2289: + ksp - the KSP context
2290: - scale - PETSC_TRUE or PETSC_FALSE
2292: Options Database Key:
2293: + -ksp_diagonal_scale -
2294: - -ksp_diagonal_scale_fix - scale the matrix back AFTER the solve
2297: Notes:
2298: Scales the matrix by D^(-1/2) A D^(-1/2) [D^(1/2) x ] = D^(-1/2) b
2299: where D_{ii} is 1/abs(A_{ii}) unless A_{ii} is zero and then it is 1.
2301: BE CAREFUL with this routine: it actually scales the matrix and right
2302: hand side that define the system. After the system is solved the matrix
2303: and right hand side remain scaled unless you use KSPSetDiagonalScaleFix()
2305: This should NOT be used within the SNES solves if you are using a line
2306: search.
2308: If you use this with the PCType Eisenstat preconditioner than you can
2309: use the PCEisenstatSetNoDiagonalScaling() option, or -pc_eisenstat_no_diagonal_scaling
2310: to save some unneeded, redundant flops.
2312: Level: intermediate
2314: .keywords: set, options, prefix, database
2316: .seealso: KSPGetDiagonalScale(), KSPSetDiagonalScaleFix(), KSP
2317: @*/
2318: PetscErrorCode KSPSetDiagonalScale(KSP ksp,PetscBool scale)
2319: {
2323: ksp->dscale = scale;
2324: return(0);
2325: }
2327: /*@
2328: KSPGetDiagonalScale - Checks if KSP solver scales the matrix and
2329: right hand side
2331: Not Collective
2333: Input Parameter:
2334: . ksp - the KSP context
2336: Output Parameter:
2337: . scale - PETSC_TRUE or PETSC_FALSE
2339: Notes:
2340: BE CAREFUL with this routine: it actually scales the matrix and right
2341: hand side that define the system. After the system is solved the matrix
2342: and right hand side remain scaled unless you use KSPSetDiagonalScaleFix()
2344: Level: intermediate
2346: .keywords: set, options, prefix, database
2348: .seealso: KSPSetDiagonalScale(), KSPSetDiagonalScaleFix(), KSP
2349: @*/
2350: PetscErrorCode KSPGetDiagonalScale(KSP ksp,PetscBool *scale)
2351: {
2355: *scale = ksp->dscale;
2356: return(0);
2357: }
2359: /*@
2360: KSPSetDiagonalScaleFix - Tells KSP to diagonally scale the system
2361: back after solving.
2363: Logically Collective on KSP
2365: Input Parameter:
2366: + ksp - the KSP context
2367: - fix - PETSC_TRUE to scale back after the system solve, PETSC_FALSE to not
2368: rescale (default)
2370: Notes:
2371: Must be called after KSPSetDiagonalScale()
2373: Using this will slow things down, because it rescales the matrix before and
2374: after each linear solve. This is intended mainly for testing to allow one
2375: to easily get back the original system to make sure the solution computed is
2376: accurate enough.
2378: Level: intermediate
2380: .keywords: set, options, prefix, database
2382: .seealso: KSPGetDiagonalScale(), KSPSetDiagonalScale(), KSPGetDiagonalScaleFix(), KSP
2383: @*/
2384: PetscErrorCode KSPSetDiagonalScaleFix(KSP ksp,PetscBool fix)
2385: {
2389: ksp->dscalefix = fix;
2390: return(0);
2391: }
2393: /*@
2394: KSPGetDiagonalScaleFix - Determines if KSP diagonally scales the system
2395: back after solving.
2397: Not Collective
2399: Input Parameter:
2400: . ksp - the KSP context
2402: Output Parameter:
2403: . fix - PETSC_TRUE to scale back after the system solve, PETSC_FALSE to not
2404: rescale (default)
2406: Notes:
2407: Must be called after KSPSetDiagonalScale()
2409: If PETSC_TRUE will slow things down, because it rescales the matrix before and
2410: after each linear solve. This is intended mainly for testing to allow one
2411: to easily get back the original system to make sure the solution computed is
2412: accurate enough.
2414: Level: intermediate
2416: .keywords: set, options, prefix, database
2418: .seealso: KSPGetDiagonalScale(), KSPSetDiagonalScale(), KSPSetDiagonalScaleFix(), KSP
2419: @*/
2420: PetscErrorCode KSPGetDiagonalScaleFix(KSP ksp,PetscBool *fix)
2421: {
2425: *fix = ksp->dscalefix;
2426: return(0);
2427: }
2429: /*@C
2430: KSPSetComputeOperators - set routine to compute the linear operators
2432: Logically Collective
2434: Input Arguments:
2435: + ksp - the KSP context
2436: . func - function to compute the operators
2437: - ctx - optional context
2439: Calling sequence of func:
2440: $ func(KSP ksp,Mat A,Mat B,void *ctx)
2442: + ksp - the KSP context
2443: . A - the linear operator
2444: . B - preconditioning matrix
2445: - ctx - optional user-provided context
2447: Notes:
2448: The user provided func() will be called automatically at the very next call to KSPSolve(). It will not be called at future KSPSolve() calls
2449: unless either KSPSetComputeOperators() or KSPSetOperators() is called before that KSPSolve() is called.
2451: To reuse the same preconditioner for the next KSPSolve() and not compute a new one based on the most recently computed matrix call KSPSetReusePreconditioner()
2453: Level: beginner
2455: .seealso: KSPSetOperators(), KSPSetComputeRHS(), DMKSPSetComputeOperators(), KSPSetComputeInitialGuess()
2456: @*/
2457: PetscErrorCode KSPSetComputeOperators(KSP ksp,PetscErrorCode (*func)(KSP,Mat,Mat,void*),void *ctx)
2458: {
2460: DM dm;
2464: KSPGetDM(ksp,&dm);
2465: DMKSPSetComputeOperators(dm,func,ctx);
2466: if (ksp->setupstage == KSP_SETUP_NEWRHS) ksp->setupstage = KSP_SETUP_NEWMATRIX;
2467: return(0);
2468: }
2470: /*@C
2471: KSPSetComputeRHS - set routine to compute the right hand side of the linear system
2473: Logically Collective
2475: Input Arguments:
2476: + ksp - the KSP context
2477: . func - function to compute the right hand side
2478: - ctx - optional context
2480: Calling sequence of func:
2481: $ func(KSP ksp,Vec b,void *ctx)
2483: + ksp - the KSP context
2484: . b - right hand side of linear system
2485: - ctx - optional user-provided context
2487: Notes:
2488: The routine you provide will be called EACH you call KSPSolve() to prepare the new right hand side for that solve
2490: Level: beginner
2492: .seealso: KSPSolve(), DMKSPSetComputeRHS(), KSPSetComputeOperators()
2493: @*/
2494: PetscErrorCode KSPSetComputeRHS(KSP ksp,PetscErrorCode (*func)(KSP,Vec,void*),void *ctx)
2495: {
2497: DM dm;
2501: KSPGetDM(ksp,&dm);
2502: DMKSPSetComputeRHS(dm,func,ctx);
2503: return(0);
2504: }
2506: /*@C
2507: KSPSetComputeInitialGuess - set routine to compute the initial guess of the linear system
2509: Logically Collective
2511: Input Arguments:
2512: + ksp - the KSP context
2513: . func - function to compute the initial guess
2514: - ctx - optional context
2516: Calling sequence of func:
2517: $ func(KSP ksp,Vec x,void *ctx)
2519: + ksp - the KSP context
2520: . x - solution vector
2521: - ctx - optional user-provided context
2523: Notes: This should only be used in conjunction with KSPSetComputeRHS(), KSPSetComputeOperators(), otherwise
2524: call KSPSetInitialGuessNonzero() and set the initial guess values in the solution vector passed to KSPSolve().
2526: Level: beginner
2528: .seealso: KSPSolve(), KSPSetComputeRHS(), KSPSetComputeOperators(), DMKSPSetComputeInitialGuess()
2529: @*/
2530: PetscErrorCode KSPSetComputeInitialGuess(KSP ksp,PetscErrorCode (*func)(KSP,Vec,void*),void *ctx)
2531: {
2533: DM dm;
2537: KSPGetDM(ksp,&dm);
2538: DMKSPSetComputeInitialGuess(dm,func,ctx);
2539: return(0);
2540: }