Actual source code: nepopts.c
1: /*
2: - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
3: SLEPc - Scalable Library for Eigenvalue Problem Computations
4: Copyright (c) 2002-, Universitat Politecnica de Valencia, Spain
6: This file is part of SLEPc.
7: SLEPc is distributed under a 2-clause BSD license (see LICENSE).
8: - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
9: */
10: /*
11: NEP routines related to options that can be set via the command-line
12: or procedurally
13: */
15: #include <slepc/private/nepimpl.h>
16: #include <petscdraw.h>
18: /*@C
19: NEPMonitorSetFromOptions - Sets a monitor function and viewer appropriate for the type
20: indicated by the user.
22: Collective
24: Input Parameters:
25: + nep - the nonlinear eigensolver context
26: . opt - the command line option for this monitor
27: . name - the monitor type one is seeking
28: . ctx - an optional user context for the monitor, or `NULL`
29: - trackall - whether this monitor tracks all eigenvalues or not
31: Level: developer
33: .seealso: [](ch:nep), `NEPMonitorSet()`, `NEPSetTrackAll()`
34: @*/
35: PetscErrorCode NEPMonitorSetFromOptions(NEP nep,const char opt[],const char name[],PetscCtx ctx,PetscBool trackall)
36: {
37: PetscErrorCode (*mfunc)(NEP,PetscInt,PetscInt,PetscScalar*,PetscScalar*,PetscReal*,PetscInt,void*);
38: PetscErrorCode (*cfunc)(PetscViewer,PetscViewerFormat,void*,PetscViewerAndFormat**);
39: PetscErrorCode (*dfunc)(PetscViewerAndFormat**);
40: PetscViewerAndFormat *vf;
41: PetscViewer viewer;
42: PetscViewerFormat format;
43: PetscViewerType vtype;
44: char key[PETSC_MAX_PATH_LEN];
45: PetscBool flg;
47: PetscFunctionBegin;
48: PetscCall(PetscOptionsCreateViewer(PetscObjectComm((PetscObject)nep),((PetscObject)nep)->options,((PetscObject)nep)->prefix,opt,&viewer,&format,&flg));
49: if (!flg) PetscFunctionReturn(PETSC_SUCCESS);
51: PetscCall(PetscViewerGetType(viewer,&vtype));
52: PetscCall(SlepcMonitorMakeKey_Internal(name,vtype,format,key));
53: PetscCall(PetscFunctionListFind(NEPMonitorList,key,&mfunc));
54: PetscCheck(mfunc,PetscObjectComm((PetscObject)nep),PETSC_ERR_SUP,"Specified viewer and format not supported");
55: PetscCall(PetscFunctionListFind(NEPMonitorCreateList,key,&cfunc));
56: PetscCall(PetscFunctionListFind(NEPMonitorDestroyList,key,&dfunc));
57: if (!cfunc) cfunc = PetscViewerAndFormatCreate_Internal;
58: if (!dfunc) dfunc = PetscViewerAndFormatDestroy;
60: PetscCall((*cfunc)(viewer,format,ctx,&vf));
61: PetscCall(PetscViewerDestroy(&viewer));
62: PetscCall(NEPMonitorSet(nep,mfunc,vf,(PetscCtxDestroyFn*)dfunc));
63: if (trackall) PetscCall(NEPSetTrackAll(nep,PETSC_TRUE));
64: PetscFunctionReturn(PETSC_SUCCESS);
65: }
67: /*@
68: NEPSetFromOptions - Sets `NEP` options from the options database.
69: This routine must be called before `NEPSetUp()` if the user is to be
70: allowed to configure the solver.
72: Collective
74: Input Parameter:
75: . nep - the nonlinear eigensolver context
77: Note:
78: To see all options, run your program with the `-help` option.
80: Level: beginner
82: .seealso: [](ch:nep), `NEPSetOptionsPrefix()`
83: @*/
84: PetscErrorCode NEPSetFromOptions(NEP nep)
85: {
86: char type[256];
87: PetscBool set,flg,flg1,flg2,flg3,flg4,flg5,bval;
88: PetscReal r;
89: PetscScalar s;
90: PetscInt i,j,k;
91: NEPRefine refine;
92: NEPRefineScheme scheme;
94: PetscFunctionBegin;
96: PetscCall(NEPRegisterAll());
97: PetscObjectOptionsBegin((PetscObject)nep);
98: PetscCall(PetscOptionsFList("-nep_type","Nonlinear eigensolver method","NEPSetType",NEPList,(char*)(((PetscObject)nep)->type_name?((PetscObject)nep)->type_name:NEPRII),type,sizeof(type),&flg));
99: if (flg) PetscCall(NEPSetType(nep,type));
100: else if (!((PetscObject)nep)->type_name) PetscCall(NEPSetType(nep,NEPRII));
102: PetscCall(PetscOptionsBoolGroupBegin("-nep_general","General nonlinear eigenvalue problem","NEPSetProblemType",&flg));
103: if (flg) PetscCall(NEPSetProblemType(nep,NEP_GENERAL));
104: PetscCall(PetscOptionsBoolGroupEnd("-nep_rational","Rational eigenvalue problem","NEPSetProblemType",&flg));
105: if (flg) PetscCall(NEPSetProblemType(nep,NEP_RATIONAL));
107: refine = nep->refine;
108: PetscCall(PetscOptionsEnum("-nep_refine","Iterative refinement method","NEPSetRefine",NEPRefineTypes,(PetscEnum)refine,(PetscEnum*)&refine,&flg1));
109: i = nep->npart;
110: PetscCall(PetscOptionsInt("-nep_refine_partitions","Number of partitions of the communicator for iterative refinement","NEPSetRefine",nep->npart,&i,&flg2));
111: r = nep->rtol;
112: PetscCall(PetscOptionsReal("-nep_refine_tol","Tolerance for iterative refinement","NEPSetRefine",nep->rtol==(PetscReal)PETSC_DETERMINE?SLEPC_DEFAULT_TOL/1000:nep->rtol,&r,&flg3));
113: j = nep->rits;
114: PetscCall(PetscOptionsInt("-nep_refine_its","Maximum number of iterations for iterative refinement","NEPSetRefine",nep->rits,&j,&flg4));
115: scheme = nep->scheme;
116: PetscCall(PetscOptionsEnum("-nep_refine_scheme","Scheme used for linear systems within iterative refinement","NEPSetRefine",NEPRefineSchemes,(PetscEnum)scheme,(PetscEnum*)&scheme,&flg5));
117: if (flg1 || flg2 || flg3 || flg4 || flg5) PetscCall(NEPSetRefine(nep,refine,i,r,j,scheme));
119: i = nep->max_it;
120: PetscCall(PetscOptionsInt("-nep_max_it","Maximum number of iterations","NEPSetTolerances",nep->max_it,&i,&flg1));
121: r = nep->tol;
122: PetscCall(PetscOptionsReal("-nep_tol","Tolerance","NEPSetTolerances",SlepcDefaultTol(nep->tol),&r,&flg2));
123: if (flg1 || flg2) PetscCall(NEPSetTolerances(nep,r,i));
125: PetscCall(PetscOptionsBoolGroupBegin("-nep_conv_rel","Relative error convergence test","NEPSetConvergenceTest",&flg));
126: if (flg) PetscCall(NEPSetConvergenceTest(nep,NEP_CONV_REL));
127: PetscCall(PetscOptionsBoolGroup("-nep_conv_norm","Convergence test relative to the matrix norms","NEPSetConvergenceTest",&flg));
128: if (flg) PetscCall(NEPSetConvergenceTest(nep,NEP_CONV_NORM));
129: PetscCall(PetscOptionsBoolGroup("-nep_conv_abs","Absolute error convergence test","NEPSetConvergenceTest",&flg));
130: if (flg) PetscCall(NEPSetConvergenceTest(nep,NEP_CONV_ABS));
131: PetscCall(PetscOptionsBoolGroupEnd("-nep_conv_user","User-defined convergence test","NEPSetConvergenceTest",&flg));
132: if (flg) PetscCall(NEPSetConvergenceTest(nep,NEP_CONV_USER));
134: PetscCall(PetscOptionsBoolGroupBegin("-nep_stop_basic","Stop iteration if all eigenvalues converged or max_it reached","NEPSetStoppingTest",&flg));
135: if (flg) PetscCall(NEPSetStoppingTest(nep,NEP_STOP_BASIC));
136: PetscCall(PetscOptionsBoolGroupEnd("-nep_stop_user","User-defined stopping test","NEPSetStoppingTest",&flg));
137: if (flg) PetscCall(NEPSetStoppingTest(nep,NEP_STOP_USER));
139: i = nep->nev;
140: PetscCall(PetscOptionsInt("-nep_nev","Number of eigenvalues to compute","NEPSetDimensions",nep->nev,&i,&flg1));
141: j = nep->ncv;
142: PetscCall(PetscOptionsInt("-nep_ncv","Number of basis vectors","NEPSetDimensions",nep->ncv,&j,&flg2));
143: k = nep->mpd;
144: PetscCall(PetscOptionsInt("-nep_mpd","Maximum dimension of projected problem","NEPSetDimensions",nep->mpd,&k,&flg3));
145: if (flg1 || flg2 || flg3) PetscCall(NEPSetDimensions(nep,i,j,k));
147: PetscCall(PetscOptionsBoolGroupBegin("-nep_largest_magnitude","Compute largest eigenvalues in magnitude","NEPSetWhichEigenpairs",&flg));
148: if (flg) PetscCall(NEPSetWhichEigenpairs(nep,NEP_LARGEST_MAGNITUDE));
149: PetscCall(PetscOptionsBoolGroup("-nep_smallest_magnitude","Compute smallest eigenvalues in magnitude","NEPSetWhichEigenpairs",&flg));
150: if (flg) PetscCall(NEPSetWhichEigenpairs(nep,NEP_SMALLEST_MAGNITUDE));
151: PetscCall(PetscOptionsBoolGroup("-nep_largest_real","Compute eigenvalues with largest real parts","NEPSetWhichEigenpairs",&flg));
152: if (flg) PetscCall(NEPSetWhichEigenpairs(nep,NEP_LARGEST_REAL));
153: PetscCall(PetscOptionsBoolGroup("-nep_smallest_real","Compute eigenvalues with smallest real parts","NEPSetWhichEigenpairs",&flg));
154: if (flg) PetscCall(NEPSetWhichEigenpairs(nep,NEP_SMALLEST_REAL));
155: PetscCall(PetscOptionsBoolGroup("-nep_largest_imaginary","Compute eigenvalues with largest imaginary parts","NEPSetWhichEigenpairs",&flg));
156: if (flg) PetscCall(NEPSetWhichEigenpairs(nep,NEP_LARGEST_IMAGINARY));
157: PetscCall(PetscOptionsBoolGroup("-nep_smallest_imaginary","Compute eigenvalues with smallest imaginary parts","NEPSetWhichEigenpairs",&flg));
158: if (flg) PetscCall(NEPSetWhichEigenpairs(nep,NEP_SMALLEST_IMAGINARY));
159: PetscCall(PetscOptionsBoolGroup("-nep_target_magnitude","Compute eigenvalues closest to target","NEPSetWhichEigenpairs",&flg));
160: if (flg) PetscCall(NEPSetWhichEigenpairs(nep,NEP_TARGET_MAGNITUDE));
161: PetscCall(PetscOptionsBoolGroup("-nep_target_real","Compute eigenvalues with real parts closest to target","NEPSetWhichEigenpairs",&flg));
162: if (flg) PetscCall(NEPSetWhichEigenpairs(nep,NEP_TARGET_REAL));
163: PetscCall(PetscOptionsBoolGroup("-nep_target_imaginary","Compute eigenvalues with imaginary parts closest to target","NEPSetWhichEigenpairs",&flg));
164: if (flg) PetscCall(NEPSetWhichEigenpairs(nep,NEP_TARGET_IMAGINARY));
165: PetscCall(PetscOptionsBoolGroup("-nep_all","Compute all eigenvalues in a region","NEPSetWhichEigenpairs",&flg));
166: if (flg) PetscCall(NEPSetWhichEigenpairs(nep,NEP_ALL));
167: PetscCall(PetscOptionsBoolGroupEnd("-nep_which_user","Select the user-defined selection criterion","NEPSetWhichEigenpairs",&flg));
168: if (flg) PetscCall(NEPSetWhichEigenpairs(nep,NEP_WHICH_USER));
170: PetscCall(PetscOptionsScalar("-nep_target","Value of the target","NEPSetTarget",nep->target,&s,&flg));
171: if (flg) {
172: if (nep->which!=NEP_TARGET_REAL && nep->which!=NEP_TARGET_IMAGINARY) PetscCall(NEPSetWhichEigenpairs(nep,NEP_TARGET_MAGNITUDE));
173: PetscCall(NEPSetTarget(nep,s));
174: }
176: PetscCall(PetscOptionsBool("-nep_two_sided","Use two-sided variant (to compute left eigenvectors)","NEPSetTwoSided",nep->twosided,&bval,&flg));
177: if (flg) PetscCall(NEPSetTwoSided(nep,bval));
179: /* -----------------------------------------------------------------------*/
180: /*
181: Cancels all monitors hardwired into code before call to NEPSetFromOptions()
182: */
183: PetscCall(PetscOptionsBool("-nep_monitor_cancel","Remove any hardwired monitor routines","NEPMonitorCancel",PETSC_FALSE,&flg,&set));
184: if (set && flg) PetscCall(NEPMonitorCancel(nep));
185: PetscCall(NEPMonitorSetFromOptions(nep,"-nep_monitor","first_approximation",NULL,PETSC_FALSE));
186: PetscCall(NEPMonitorSetFromOptions(nep,"-nep_monitor_all","all_approximations",NULL,PETSC_TRUE));
187: PetscCall(NEPMonitorSetFromOptions(nep,"-nep_monitor_conv","convergence_history",NULL,PETSC_FALSE));
189: /* -----------------------------------------------------------------------*/
190: PetscCall(PetscOptionsName("-nep_view","Print detailed information on solver used","NEPView",&set));
191: PetscCall(PetscOptionsName("-nep_view_vectors","View computed eigenvectors","NEPVectorsView",&set));
192: PetscCall(PetscOptionsName("-nep_view_values","View computed eigenvalues","NEPValuesView",&set));
193: PetscCall(PetscOptionsName("-nep_converged_reason","Print reason for convergence, and number of iterations","NEPConvergedReasonView",&set));
194: PetscCall(PetscOptionsName("-nep_error_absolute","Print absolute errors of each eigenpair","NEPErrorView",&set));
195: PetscCall(PetscOptionsName("-nep_error_relative","Print relative errors of each eigenpair","NEPErrorView",&set));
197: PetscTryTypeMethod(nep,setfromoptions,PetscOptionsObject);
198: PetscCall(PetscObjectProcessOptionsHandlers((PetscObject)nep,PetscOptionsObject));
199: PetscOptionsEnd();
201: if (!nep->V) PetscCall(NEPGetBV(nep,&nep->V));
202: PetscCall(BVSetFromOptions(nep->V));
203: if (!nep->rg) PetscCall(NEPGetRG(nep,&nep->rg));
204: PetscCall(RGSetFromOptions(nep->rg));
205: if (nep->useds) {
206: if (!nep->ds) PetscCall(NEPGetDS(nep,&nep->ds));
207: PetscCall(NEPSetDSType(nep));
208: PetscCall(DSSetFromOptions(nep->ds));
209: }
210: if (!nep->refineksp) PetscCall(NEPRefineGetKSP(nep,&nep->refineksp));
211: PetscCall(KSPSetFromOptions(nep->refineksp));
212: if (nep->fui==NEP_USER_INTERFACE_SPLIT) for (i=0;i<nep->nt;i++) PetscCall(FNSetFromOptions(nep->f[i]));
213: PetscFunctionReturn(PETSC_SUCCESS);
214: }
216: /*@
217: NEPGetTolerances - Gets the tolerance and maximum iteration count used
218: by the `NEP` convergence tests.
220: Not Collective
222: Input Parameter:
223: . nep - the nonlinear eigensolver context
225: Output Parameters:
226: + tol - the convergence tolerance
227: - maxits - maximum number of iterations
229: Notes:
230: The user can specify `NULL` for any parameter that is not needed.
232: Level: intermediate
234: .seealso: [](ch:nep), `NEPSetTolerances()`
235: @*/
236: PetscErrorCode NEPGetTolerances(NEP nep,PetscReal *tol,PetscInt *maxits)
237: {
238: PetscFunctionBegin;
240: if (tol) *tol = nep->tol;
241: if (maxits) *maxits = nep->max_it;
242: PetscFunctionReturn(PETSC_SUCCESS);
243: }
245: /*@
246: NEPSetTolerances - Sets the tolerance and maximum iteration count used
247: by the `NEP` convergence tests.
249: Logically Collective
251: Input Parameters:
252: + nep - the nonlinear eigensolver context
253: . tol - the convergence tolerance
254: - maxits - maximum number of iterations to use
256: Options Database Keys:
257: + -nep_tol tol - sets the convergence tolerance
258: - -nep_max_it maxits - sets the maximum number of iterations allowed
260: Note:
261: Use `PETSC_CURRENT` to retain the current value of any of the parameters.
262: Use `PETSC_DETERMINE` for either argument to assign a default value computed
263: internally (may be different in each solver).
264: For `maxits` use `PETSC_UNLIMITED` to indicate there is no upper bound on this value.
266: Level: intermediate
268: .seealso: [](ch:nep), `NEPGetTolerances()`
269: @*/
270: PetscErrorCode NEPSetTolerances(NEP nep,PetscReal tol,PetscInt maxits)
271: {
272: PetscFunctionBegin;
276: if (tol == (PetscReal)PETSC_DETERMINE) {
277: nep->tol = PETSC_DETERMINE;
278: nep->state = NEP_STATE_INITIAL;
279: } else if (tol != (PetscReal)PETSC_CURRENT) {
280: PetscCheck(tol>0.0,PetscObjectComm((PetscObject)nep),PETSC_ERR_ARG_OUTOFRANGE,"Illegal value of tol. Must be > 0");
281: nep->tol = tol;
282: }
283: if (maxits == PETSC_DETERMINE) {
284: nep->max_it = PETSC_DETERMINE;
285: nep->state = NEP_STATE_INITIAL;
286: } else if (maxits == PETSC_UNLIMITED) {
287: nep->max_it = PETSC_INT_MAX;
288: } else if (maxits != PETSC_CURRENT) {
289: PetscCheck(maxits>0,PetscObjectComm((PetscObject)nep),PETSC_ERR_ARG_OUTOFRANGE,"Illegal value of maxits. Must be > 0");
290: nep->max_it = maxits;
291: }
292: PetscFunctionReturn(PETSC_SUCCESS);
293: }
295: /*@
296: NEPGetDimensions - Gets the number of eigenvalues to compute
297: and the dimension of the subspace.
299: Not Collective
301: Input Parameter:
302: . nep - the nonlinear eigensolver context
304: Output Parameters:
305: + nev - number of eigenvalues to compute
306: . ncv - the maximum dimension of the subspace to be used by the solver
307: - mpd - the maximum dimension allowed for the projected problem
309: Note:
310: The user can specify `NULL` for any parameter that is not needed.
312: Level: intermediate
314: .seealso: [](ch:nep), `NEPSetDimensions()`
315: @*/
316: PetscErrorCode NEPGetDimensions(NEP nep,PetscInt *nev,PetscInt *ncv,PetscInt *mpd)
317: {
318: PetscFunctionBegin;
320: if (nev) *nev = nep->nev;
321: if (ncv) *ncv = nep->ncv;
322: if (mpd) *mpd = nep->mpd;
323: PetscFunctionReturn(PETSC_SUCCESS);
324: }
326: /*@
327: NEPSetDimensions - Sets the number of eigenvalues to compute
328: and the dimension of the subspace.
330: Logically Collective
332: Input Parameters:
333: + nep - the nonlinear eigensolver context
334: . nev - number of eigenvalues to compute
335: . ncv - the maximum dimension of the subspace to be used by the solver
336: - mpd - the maximum dimension allowed for the projected problem
338: Options Database Keys:
339: + -nep_nev nev - sets the number of eigenvalues
340: . -nep_ncv ncv - sets the dimension of the subspace
341: - -nep_mpd mpd - sets the maximum projected dimension
343: Notes:
344: Use `PETSC_DETERMINE` for `ncv` and `mpd` to assign a reasonably good value, which is
345: dependent on the solution method. For any of the arguments, use `PETSC_CURRENT`
346: to preserve the current value.
348: The parameters `ncv` and `mpd` are intimately related, so that the user is advised
349: to set one of them at most. Normal usage is\:
351: 1. in cases where `nev` is small, the user sets `ncv` (a reasonable default is `2*nev`).
352: 2. in cases where `nev` is large, the user sets `mpd`.
354: The value of `ncv` should always be between `nev` and `(nev+mpd)`, typically
355: `ncv=nev+mpd`. If `nev` is not too large, `mpd=nev` is a reasonable choice, otherwise
356: a smaller value should be used.
358: Level: intermediate
360: .seealso: [](ch:nep), `NEPGetDimensions()`
361: @*/
362: PetscErrorCode NEPSetDimensions(NEP nep,PetscInt nev,PetscInt ncv,PetscInt mpd)
363: {
364: PetscFunctionBegin;
369: if (nev != PETSC_CURRENT) {
370: PetscCheck(nev>0,PetscObjectComm((PetscObject)nep),PETSC_ERR_ARG_OUTOFRANGE,"Illegal value of nev. Must be > 0");
371: nep->nev = nev;
372: }
373: if (ncv == PETSC_DETERMINE) {
374: nep->ncv = PETSC_DETERMINE;
375: } else if (ncv != PETSC_CURRENT) {
376: PetscCheck(ncv>0,PetscObjectComm((PetscObject)nep),PETSC_ERR_ARG_OUTOFRANGE,"Illegal value of ncv. Must be > 0");
377: nep->ncv = ncv;
378: }
379: if (mpd == PETSC_DETERMINE) {
380: nep->mpd = PETSC_DETERMINE;
381: } else if (mpd != PETSC_CURRENT) {
382: PetscCheck(mpd>0,PetscObjectComm((PetscObject)nep),PETSC_ERR_ARG_OUTOFRANGE,"Illegal value of mpd. Must be > 0");
383: nep->mpd = mpd;
384: }
385: nep->state = NEP_STATE_INITIAL;
386: PetscFunctionReturn(PETSC_SUCCESS);
387: }
389: /*@
390: NEPSetWhichEigenpairs - Specifies which portion of the spectrum is
391: to be sought.
393: Logically Collective
395: Input Parameters:
396: + nep - the nonlinear eigensolver context
397: - which - the portion of the spectrum to be sought, see `NEPWhich` for possible values
399: Options Database Keys:
400: + -nep_largest_magnitude - sets largest eigenvalues in magnitude
401: . -nep_smallest_magnitude - sets smallest eigenvalues in magnitude
402: . -nep_largest_real - sets largest real parts
403: . -nep_smallest_real - sets smallest real parts
404: . -nep_largest_imaginary - sets largest imaginary parts
405: . -nep_smallest_imaginary - sets smallest imaginary parts
406: . -nep_target_magnitude - sets eigenvalues closest to target
407: . -nep_target_real - sets real parts closest to target
408: . -nep_target_imaginary - sets imaginary parts closest to target
409: . -nep_all - sets all eigenvalues in a region
410: - -nep_which_user - select the user-defined selection criterion
412: Notes:
413: Not all eigensolvers implemented in `NEP` account for all the possible values
414: of `which`. Also, some values make sense only for certain types of
415: problems. If SLEPc is compiled for real numbers `NEP_LARGEST_IMAGINARY`
416: and `NEP_SMALLEST_IMAGINARY` use the absolute value of the imaginary part
417: for eigenvalue selection.
419: The target is a scalar value provided with `NEPSetTarget()`.
421: The criterion `NEP_TARGET_IMAGINARY` is available only in case PETSc and
422: SLEPc have been built with complex scalars.
424: `NEP_ALL` is intended for use in the context of the `NEPCISS` solver for
425: computing all eigenvalues in a region.
427: Level: intermediate
429: .seealso: [](ch:nep), `NEPGetWhichEigenpairs()`, `NEPSetTarget()`, `NEPSetDimensions()`, `NEPSetEigenvalueComparison()`, `NEPWhich`
430: @*/
431: PetscErrorCode NEPSetWhichEigenpairs(NEP nep,NEPWhich which)
432: {
433: PetscFunctionBegin;
436: switch (which) {
437: case NEP_LARGEST_MAGNITUDE:
438: case NEP_SMALLEST_MAGNITUDE:
439: case NEP_LARGEST_REAL:
440: case NEP_SMALLEST_REAL:
441: case NEP_LARGEST_IMAGINARY:
442: case NEP_SMALLEST_IMAGINARY:
443: case NEP_TARGET_MAGNITUDE:
444: case NEP_TARGET_REAL:
445: #if defined(PETSC_USE_COMPLEX)
446: case NEP_TARGET_IMAGINARY:
447: #endif
448: case NEP_ALL:
449: case NEP_WHICH_USER:
450: if (nep->which != which) {
451: nep->state = NEP_STATE_INITIAL;
452: nep->which = which;
453: }
454: break;
455: #if !defined(PETSC_USE_COMPLEX)
456: case NEP_TARGET_IMAGINARY:
457: SETERRQ(PetscObjectComm((PetscObject)nep),PETSC_ERR_SUP,"NEP_TARGET_IMAGINARY can be used only with complex scalars");
458: #endif
459: default:
460: SETERRQ(PetscObjectComm((PetscObject)nep),PETSC_ERR_ARG_OUTOFRANGE,"Invalid 'which' value");
461: }
462: PetscFunctionReturn(PETSC_SUCCESS);
463: }
465: /*@
466: NEPGetWhichEigenpairs - Returns which portion of the spectrum is to be
467: sought.
469: Not Collective
471: Input Parameter:
472: . nep - the nonlinear eigensolver context
474: Output Parameter:
475: . which - the portion of the spectrum to be sought
477: Level: intermediate
479: .seealso: [](ch:nep), `NEPSetWhichEigenpairs()`, `NEPWhich`
480: @*/
481: PetscErrorCode NEPGetWhichEigenpairs(NEP nep,NEPWhich *which)
482: {
483: PetscFunctionBegin;
485: PetscAssertPointer(which,2);
486: *which = nep->which;
487: PetscFunctionReturn(PETSC_SUCCESS);
488: }
490: /*@C
491: NEPSetEigenvalueComparison - Specifies the eigenvalue comparison function
492: when `NEPSetWhichEigenpairs()` is set to `NEP_WHICH_USER`.
494: Logically Collective
496: Input Parameters:
497: + nep - the nonlinear eigensolver context
498: . comp - a pointer to the comparison function, see `SlepcEigenvalueComparisonFn` for the calling sequence
499: - ctx - a context pointer (the last parameter to the comparison function)
501: Level: advanced
503: .seealso: [](ch:nep), `NEPSetWhichEigenpairs()`, `NEPWhich`
504: @*/
505: PetscErrorCode NEPSetEigenvalueComparison(NEP nep,SlepcEigenvalueComparisonFn *comp,PetscCtx ctx)
506: {
507: PetscFunctionBegin;
509: nep->sc->comparison = comp;
510: nep->sc->comparisonctx = ctx;
511: nep->which = NEP_WHICH_USER;
512: PetscFunctionReturn(PETSC_SUCCESS);
513: }
515: /*@
516: NEPSetProblemType - Specifies the type of the nonlinear eigenvalue problem.
518: Logically Collective
520: Input Parameters:
521: + nep - the nonlinear eigensolver context
522: - type - a known type of nonlinear eigenvalue problem
524: Options Database Keys:
525: + -nep_general - general problem with no particular structure
526: - -nep_rational - a rational eigenvalue problem defined in split form with all $f_i$ rational
528: Notes:
529: See `NEPProblemType` for possible problem types.
531: This function is used to provide a hint to the `NEP` solver to exploit certain
532: properties of the nonlinear eigenproblem. This hint may be used or not,
533: depending on the solver. By default, no particular structure is assumed.
535: Level: intermediate
537: .seealso: [](ch:nep), `NEPSetType()`, `NEPGetProblemType()`, `NEPProblemType`
538: @*/
539: PetscErrorCode NEPSetProblemType(NEP nep,NEPProblemType type)
540: {
541: PetscFunctionBegin;
544: PetscCheck(type==NEP_GENERAL || type==NEP_RATIONAL,PetscObjectComm((PetscObject)nep),PETSC_ERR_ARG_WRONG,"Unknown eigenvalue problem type");
545: if (type != nep->problem_type) {
546: nep->problem_type = type;
547: nep->state = NEP_STATE_INITIAL;
548: }
549: PetscFunctionReturn(PETSC_SUCCESS);
550: }
552: /*@
553: NEPGetProblemType - Gets the problem type from the `NEP` object.
555: Not Collective
557: Input Parameter:
558: . nep - the nonlinear eigensolver context
560: Output Parameter:
561: . type - the problem type
563: Level: intermediate
565: .seealso: [](ch:nep), `NEPSetProblemType()`, `NEPProblemType`
566: @*/
567: PetscErrorCode NEPGetProblemType(NEP nep,NEPProblemType *type)
568: {
569: PetscFunctionBegin;
571: PetscAssertPointer(type,2);
572: *type = nep->problem_type;
573: PetscFunctionReturn(PETSC_SUCCESS);
574: }
576: /*@
577: NEPSetTwoSided - Sets the solver to use a two-sided variant so that left
578: eigenvectors are also computed.
580: Logically Collective
582: Input Parameters:
583: + nep - the nonlinear eigensolver context
584: - twosided - whether the two-sided variant is to be used or not
586: Options Database Key:
587: . -nep_two_sided (true|false) - toggles the twosided flag
589: Notes:
590: If the user sets `twosided`=`PETSC_TRUE` then the solver uses a variant of
591: the algorithm that computes both right and left eigenvectors. This is
592: usually much more costly. This option is not available in all solvers,
593: see table [](#tab:solversn).
595: When using two-sided solvers, the problem matrices must have both the
596: `MATOP_MULT` and `MATOP_MULT_TRANSPOSE` operations defined.
598: Level: advanced
600: .seealso: [](ch:nep), `NEPGetTwoSided()`, `NEPGetLeftEigenvector()`
601: @*/
602: PetscErrorCode NEPSetTwoSided(NEP nep,PetscBool twosided)
603: {
604: PetscFunctionBegin;
607: if (twosided!=nep->twosided) {
608: nep->twosided = twosided;
609: nep->state = NEP_STATE_INITIAL;
610: }
611: PetscFunctionReturn(PETSC_SUCCESS);
612: }
614: /*@
615: NEPGetTwoSided - Returns the flag indicating whether a two-sided variant
616: of the algorithm is being used or not.
618: Not Collective
620: Input Parameter:
621: . nep - the nonlinear eigensolver context
623: Output Parameter:
624: . twosided - the returned flag
626: Level: advanced
628: .seealso: [](ch:nep), `NEPSetTwoSided()`
629: @*/
630: PetscErrorCode NEPGetTwoSided(NEP nep,PetscBool *twosided)
631: {
632: PetscFunctionBegin;
634: PetscAssertPointer(twosided,2);
635: *twosided = nep->twosided;
636: PetscFunctionReturn(PETSC_SUCCESS);
637: }
639: /*@C
640: NEPSetConvergenceTestFunction - Sets a function to compute the error estimate
641: used in the convergence test.
643: Logically Collective
645: Input Parameters:
646: + nep - the nonlinear eigensolver context
647: . conv - convergence test function, see `NEPConvergenceTestFn` for the calling sequence
648: . ctx - context for private data for the convergence routine (may be `NULL`)
649: - destroy - a routine for destroying the context (may be `NULL`), see `PetscCtxDestroyFn`
650: for the calling sequence
652: Notes:
653: When this is called with a user-defined function, then the convergence
654: criterion is set to `NEP_CONV_USER`, see `NEPSetConvergenceTest()`.
656: If the error estimate returned by the convergence test function is less than
657: the tolerance, then the eigenvalue is accepted as converged.
659: Level: advanced
661: .seealso: [](ch:nep), `NEPSetConvergenceTest()`, `NEPSetTolerances()`
662: @*/
663: PetscErrorCode NEPSetConvergenceTestFunction(NEP nep,NEPConvergenceTestFn *conv,PetscCtx ctx,PetscCtxDestroyFn *destroy)
664: {
665: PetscFunctionBegin;
667: if (nep->convergeddestroy) PetscCall((*nep->convergeddestroy)(&nep->convergedctx));
668: nep->convergeduser = conv;
669: nep->convergeddestroy = destroy;
670: nep->convergedctx = ctx;
671: if (conv == NEPConvergedRelative) nep->conv = NEP_CONV_REL;
672: else if (conv == NEPConvergedNorm) nep->conv = NEP_CONV_NORM;
673: else if (conv == NEPConvergedAbsolute) nep->conv = NEP_CONV_ABS;
674: else {
675: nep->conv = NEP_CONV_USER;
676: nep->converged = nep->convergeduser;
677: }
678: PetscFunctionReturn(PETSC_SUCCESS);
679: }
681: /*@
682: NEPSetConvergenceTest - Specifies how to compute the error estimate
683: used in the convergence test.
685: Logically Collective
687: Input Parameters:
688: + nep - the nonlinear eigensolver context
689: - conv - the type of convergence test, see `NEPConv` for possible values
691: Options Database Keys:
692: + -nep_conv_abs - sets the absolute convergence test
693: . -nep_conv_rel - sets the convergence test relative to the eigenvalue
694: . -nep_conv_norm - sets the convergence test relative to the matrix norms
695: - -nep_conv_user - selects the user-defined convergence test
697: Level: intermediate
699: .seealso: [](ch:nep), `NEPGetConvergenceTest()`, `NEPSetConvergenceTestFunction()`, `NEPSetStoppingTest()`, `NEPConv`
700: @*/
701: PetscErrorCode NEPSetConvergenceTest(NEP nep,NEPConv conv)
702: {
703: PetscFunctionBegin;
706: switch (conv) {
707: case NEP_CONV_ABS: nep->converged = NEPConvergedAbsolute; break;
708: case NEP_CONV_REL: nep->converged = NEPConvergedRelative; break;
709: case NEP_CONV_NORM: nep->converged = NEPConvergedNorm; break;
710: case NEP_CONV_USER:
711: PetscCheck(nep->convergeduser,PetscObjectComm((PetscObject)nep),PETSC_ERR_ORDER,"Must call NEPSetConvergenceTestFunction() first");
712: nep->converged = nep->convergeduser;
713: break;
714: default:
715: SETERRQ(PetscObjectComm((PetscObject)nep),PETSC_ERR_ARG_OUTOFRANGE,"Invalid 'conv' value");
716: }
717: nep->conv = conv;
718: PetscFunctionReturn(PETSC_SUCCESS);
719: }
721: /*@
722: NEPGetConvergenceTest - Gets the method used to compute the error estimate
723: used in the convergence test.
725: Not Collective
727: Input Parameter:
728: . nep - the nonlinear eigensolver context
730: Output Parameter:
731: . conv - the type of convergence test
733: Level: intermediate
735: .seealso: [](ch:nep), `NEPSetConvergenceTest()`, `NEPConv`
736: @*/
737: PetscErrorCode NEPGetConvergenceTest(NEP nep,NEPConv *conv)
738: {
739: PetscFunctionBegin;
741: PetscAssertPointer(conv,2);
742: *conv = nep->conv;
743: PetscFunctionReturn(PETSC_SUCCESS);
744: }
746: /*@C
747: NEPSetStoppingTestFunction - Sets a function to decide when to stop the outer
748: iteration of the eigensolver.
750: Logically Collective
752: Input Parameters:
753: + nep - the nonlinear eigensolver context
754: . stop - the stopping test function, see `NEPStoppingTestFn` for the calling sequence
755: . ctx - context for private data for the stopping routine (may be `NULL`)
756: - destroy - a routine for destroying the context (may be `NULL`), see `PetscCtxDestroyFn`
757: for the calling sequence
759: Note:
760: When implementing a function for this, normal usage is to first call the
761: default routine `NEPStoppingBasic()` and then set `reason` to `NEP_CONVERGED_USER`
762: if some user-defined conditions have been met. To let the eigensolver continue
763: iterating, the result must be left as `NEP_CONVERGED_ITERATING`.
765: Level: advanced
767: .seealso: [](ch:nep), `NEPSetStoppingTest()`, `NEPStoppingBasic()`
768: @*/
769: PetscErrorCode NEPSetStoppingTestFunction(NEP nep,NEPStoppingTestFn *stop,PetscCtx ctx,PetscCtxDestroyFn *destroy)
770: {
771: PetscFunctionBegin;
773: if (nep->stoppingdestroy) PetscCall((*nep->stoppingdestroy)(&nep->stoppingctx));
774: nep->stoppinguser = stop;
775: nep->stoppingdestroy = destroy;
776: nep->stoppingctx = ctx;
777: if (stop == NEPStoppingBasic) nep->stop = NEP_STOP_BASIC;
778: else {
779: nep->stop = NEP_STOP_USER;
780: nep->stopping = nep->stoppinguser;
781: }
782: PetscFunctionReturn(PETSC_SUCCESS);
783: }
785: /*@
786: NEPSetStoppingTest - Specifies how to decide the termination of the outer
787: loop of the eigensolver.
789: Logically Collective
791: Input Parameters:
792: + nep - the nonlinear eigensolver context
793: - stop - the type of stopping test, see `NEPStop`
795: Options Database Keys:
796: + -nep_stop_basic - sets the default stopping test
797: - -nep_stop_user - selects the user-defined stopping test
799: Level: advanced
801: .seealso: [](ch:nep), `NEPGetStoppingTest()`, `NEPSetStoppingTestFunction()`, `NEPSetConvergenceTest()`, `NEPStop`
802: @*/
803: PetscErrorCode NEPSetStoppingTest(NEP nep,NEPStop stop)
804: {
805: PetscFunctionBegin;
808: switch (stop) {
809: case NEP_STOP_BASIC: nep->stopping = NEPStoppingBasic; break;
810: case NEP_STOP_USER:
811: PetscCheck(nep->stoppinguser,PetscObjectComm((PetscObject)nep),PETSC_ERR_ORDER,"Must call NEPSetStoppingTestFunction() first");
812: nep->stopping = nep->stoppinguser;
813: break;
814: default:
815: SETERRQ(PetscObjectComm((PetscObject)nep),PETSC_ERR_ARG_OUTOFRANGE,"Invalid 'stop' value");
816: }
817: nep->stop = stop;
818: PetscFunctionReturn(PETSC_SUCCESS);
819: }
821: /*@
822: NEPGetStoppingTest - Gets the method used to decide the termination of the outer
823: loop of the eigensolver.
825: Not Collective
827: Input Parameter:
828: . nep - the nonlinear eigensolver context
830: Output Parameter:
831: . stop - the type of stopping test
833: Level: advanced
835: .seealso: [](ch:nep), `NEPSetStoppingTest()`, `NEPStop`
836: @*/
837: PetscErrorCode NEPGetStoppingTest(NEP nep,NEPStop *stop)
838: {
839: PetscFunctionBegin;
841: PetscAssertPointer(stop,2);
842: *stop = nep->stop;
843: PetscFunctionReturn(PETSC_SUCCESS);
844: }
846: /*@
847: NEPSetTrackAll - Specifies if the solver must compute the residual of all
848: approximate eigenpairs or not.
850: Logically Collective
852: Input Parameters:
853: + nep - the nonlinear eigensolver context
854: - trackall - whether compute all residuals or not
856: Notes:
857: If the user sets `trackall`=`PETSC_TRUE` then the solver explicitly computes
858: the residual for each eigenpair approximation. Computing the residual is
859: usually an expensive operation and solvers commonly compute the associated
860: residual to the first unconverged eigenpair.
862: The option `-nep_monitor_all` automatically activates this option.
864: Level: developer
866: .seealso: [](ch:nep), `NEPGetTrackAll()`
867: @*/
868: PetscErrorCode NEPSetTrackAll(NEP nep,PetscBool trackall)
869: {
870: PetscFunctionBegin;
873: nep->trackall = trackall;
874: PetscFunctionReturn(PETSC_SUCCESS);
875: }
877: /*@
878: NEPGetTrackAll - Returns the flag indicating whether all residual norms must
879: be computed or not.
881: Not Collective
883: Input Parameter:
884: . nep - the nonlinear eigensolver context
886: Output Parameter:
887: . trackall - the returned flag
889: Level: developer
891: .seealso: [](ch:nep), `NEPSetTrackAll()`
892: @*/
893: PetscErrorCode NEPGetTrackAll(NEP nep,PetscBool *trackall)
894: {
895: PetscFunctionBegin;
897: PetscAssertPointer(trackall,2);
898: *trackall = nep->trackall;
899: PetscFunctionReturn(PETSC_SUCCESS);
900: }
902: /*@
903: NEPSetRefine - Specifies the refinement type (and options) to be used
904: after the solve.
906: Logically Collective
908: Input Parameters:
909: + nep - the nonlinear eigensolver context
910: . refine - refinement type, see `NEPRefine` for possible values
911: . npart - number of partitions of the communicator
912: . tol - the convergence tolerance
913: . its - maximum number of refinement iterations
914: - scheme - which scheme to be used for solving the involved linear systems, see `NEPRefineScheme`
915: for possible values
917: Options Database Keys:
918: + -nep_refine (none|simple|multiple) - set the refinement type
919: . -nep_refine_partitions npart - set the number of partitions
920: . -nep_refine_tol tol - set the tolerance
921: . -nep_refine_its its - set the number of iterations
922: - -nep_refine_scheme (schur|mbe|explicit) - set the scheme for the linear solves
924: Notes:
925: This function configures the parameters of Newton iterative refinement,
926: see section [](#sec:refine) for a discussion of the different strategies
927: in the context of polynomial eigenproblems.
929: By default, iterative refinement is disabled, since it may be very
930: costly. There are two possible refinement strategies, simple and multiple.
931: The simple approach performs iterative refinement on each of the
932: converged eigenpairs individually, whereas the multiple strategy works
933: with the invariant pair as a whole, refining all eigenpairs simultaneously.
934: The latter may be required for the case of multiple eigenvalues.
936: In some cases, especially when using direct solvers within the
937: iterative refinement method, it may be helpful for improved scalability
938: to split the communicator in several partitions. The `npart` parameter
939: indicates how many partitions to use (defaults to 1).
941: The `tol` and `its` parameters specify the stopping criterion. In the simple
942: method, refinement continues until the residual of each eigenpair is
943: below the tolerance (`tol` defaults to the `NEP` tolerance, but may be set to a
944: different value). In contrast, the multiple method simply performs its
945: refinement iterations (just one by default).
947: The `scheme` argument is used to change the way in which linear systems are
948: solved. Possible choices are explicit, mixed block elimination (MBE),
949: and Schur complement.
951: Use `PETSC_CURRENT` to retain the current value of `npart`, `tol` or `its`. Use
952: `PETSC_DETERMINE` to assign a default value.
954: Level: intermediate
956: .seealso: [](ch:nep), [](#sec:refine), `NEPGetRefine()`
957: @*/
958: PetscErrorCode NEPSetRefine(NEP nep,NEPRefine refine,PetscInt npart,PetscReal tol,PetscInt its,NEPRefineScheme scheme)
959: {
960: PetscMPIInt size;
962: PetscFunctionBegin;
969: nep->refine = refine;
970: if (refine) { /* process parameters only if not REFINE_NONE */
971: if (npart!=nep->npart) {
972: PetscCall(PetscSubcommDestroy(&nep->refinesubc));
973: PetscCall(KSPDestroy(&nep->refineksp));
974: }
975: if (npart == PETSC_DETERMINE) {
976: nep->npart = 1;
977: } else if (npart != PETSC_CURRENT) {
978: PetscCallMPI(MPI_Comm_size(PetscObjectComm((PetscObject)nep),&size));
979: PetscCheck(npart>0 && npart<=size,PetscObjectComm((PetscObject)nep),PETSC_ERR_ARG_OUTOFRANGE,"Illegal value of npart");
980: nep->npart = npart;
981: }
982: if (tol == (PetscReal)PETSC_DETERMINE) {
983: nep->rtol = PETSC_DETERMINE;
984: } else if (tol != (PetscReal)PETSC_CURRENT) {
985: PetscCheck(tol>0.0,PetscObjectComm((PetscObject)nep),PETSC_ERR_ARG_OUTOFRANGE,"Illegal value of tol. Must be > 0");
986: nep->rtol = tol;
987: }
988: if (its==PETSC_DETERMINE) {
989: nep->rits = PETSC_DETERMINE;
990: } else if (its != PETSC_CURRENT) {
991: PetscCheck(its>=0,PetscObjectComm((PetscObject)nep),PETSC_ERR_ARG_OUTOFRANGE,"Illegal value of its. Must be >= 0");
992: nep->rits = its;
993: }
994: nep->scheme = scheme;
995: }
996: nep->state = NEP_STATE_INITIAL;
997: PetscFunctionReturn(PETSC_SUCCESS);
998: }
1000: /*@
1001: NEPGetRefine - Gets the refinement strategy used by the `NEP` object, and the
1002: associated parameters.
1004: Not Collective
1006: Input Parameter:
1007: . nep - the nonlinear eigensolver context
1009: Output Parameters:
1010: + refine - refinement type
1011: . npart - number of partitions of the communicator
1012: . tol - the convergence tolerance
1013: . its - maximum number of refinement iterations
1014: - scheme - the scheme used for solving linear systems
1016: Level: intermediate
1018: Note:
1019: The user can specify `NULL` for any parameter that is not needed.
1021: .seealso: [](ch:nep), `NEPSetRefine()`
1022: @*/
1023: PetscErrorCode NEPGetRefine(NEP nep,NEPRefine *refine,PetscInt *npart,PetscReal *tol,PetscInt *its,NEPRefineScheme *scheme)
1024: {
1025: PetscFunctionBegin;
1027: if (refine) *refine = nep->refine;
1028: if (npart) *npart = nep->npart;
1029: if (tol) *tol = nep->rtol;
1030: if (its) *its = nep->rits;
1031: if (scheme) *scheme = nep->scheme;
1032: PetscFunctionReturn(PETSC_SUCCESS);
1033: }
1035: /*@
1036: NEPSetOptionsPrefix - Sets the prefix used for searching for all
1037: `NEP` options in the database.
1039: Logically Collective
1041: Input Parameters:
1042: + nep - the nonlinear eigensolver context
1043: - prefix - the prefix string to prepend to all `NEP` option requests
1045: Notes:
1046: A hyphen (-) must NOT be given at the beginning of the prefix name.
1047: The first character of all runtime options is AUTOMATICALLY the
1048: hyphen.
1050: For example, to distinguish between the runtime options for two
1051: different `NEP` contexts, one could call
1052: .vb
1053: NEPSetOptionsPrefix(nep1,"neig1_")
1054: NEPSetOptionsPrefix(nep2,"neig2_")
1055: .ve
1057: Level: advanced
1059: .seealso: [](ch:nep), `NEPAppendOptionsPrefix()`, `NEPGetOptionsPrefix()`
1060: @*/
1061: PetscErrorCode NEPSetOptionsPrefix(NEP nep,const char prefix[])
1062: {
1063: PetscFunctionBegin;
1065: if (!nep->V) PetscCall(NEPGetBV(nep,&nep->V));
1066: PetscCall(BVSetOptionsPrefix(nep->V,prefix));
1067: if (!nep->ds) PetscCall(NEPGetDS(nep,&nep->ds));
1068: PetscCall(DSSetOptionsPrefix(nep->ds,prefix));
1069: if (!nep->rg) PetscCall(NEPGetRG(nep,&nep->rg));
1070: PetscCall(RGSetOptionsPrefix(nep->rg,prefix));
1071: PetscCall(PetscObjectSetOptionsPrefix((PetscObject)nep,prefix));
1072: PetscFunctionReturn(PETSC_SUCCESS);
1073: }
1075: /*@
1076: NEPAppendOptionsPrefix - Appends to the prefix used for searching for all
1077: `NEP` options in the database.
1079: Logically Collective
1081: Input Parameters:
1082: + nep - the nonlinear eigensolver context
1083: - prefix - the prefix string to prepend to all `NEP` option requests
1085: Notes:
1086: A hyphen (-) must NOT be given at the beginning of the prefix name.
1087: The first character of all runtime options is AUTOMATICALLY the hyphen.
1089: Level: advanced
1091: .seealso: [](ch:nep), `NEPSetOptionsPrefix()`, `NEPGetOptionsPrefix()`
1092: @*/
1093: PetscErrorCode NEPAppendOptionsPrefix(NEP nep,const char prefix[])
1094: {
1095: PetscFunctionBegin;
1097: if (!nep->V) PetscCall(NEPGetBV(nep,&nep->V));
1098: PetscCall(BVAppendOptionsPrefix(nep->V,prefix));
1099: if (!nep->ds) PetscCall(NEPGetDS(nep,&nep->ds));
1100: PetscCall(DSAppendOptionsPrefix(nep->ds,prefix));
1101: if (!nep->rg) PetscCall(NEPGetRG(nep,&nep->rg));
1102: PetscCall(RGAppendOptionsPrefix(nep->rg,prefix));
1103: PetscCall(PetscObjectAppendOptionsPrefix((PetscObject)nep,prefix));
1104: PetscFunctionReturn(PETSC_SUCCESS);
1105: }
1107: /*@
1108: NEPGetOptionsPrefix - Gets the prefix used for searching for all
1109: `NEP` options in the database.
1111: Not Collective
1113: Input Parameter:
1114: . nep - the nonlinear eigensolver context
1116: Output Parameter:
1117: . prefix - pointer to the prefix string used is returned
1119: Level: advanced
1121: .seealso: [](ch:nep), `NEPSetOptionsPrefix()`, `NEPAppendOptionsPrefix()`
1122: @*/
1123: PetscErrorCode NEPGetOptionsPrefix(NEP nep,const char *prefix[])
1124: {
1125: PetscFunctionBegin;
1127: PetscAssertPointer(prefix,2);
1128: PetscCall(PetscObjectGetOptionsPrefix((PetscObject)nep,prefix));
1129: PetscFunctionReturn(PETSC_SUCCESS);
1130: }