Actual source code: epsimpl.h
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: */
11: #pragma once
13: #include <slepceps.h>
14: #include <slepc/private/bvimpl.h>
16: /* SUBMANSEC = EPS */
18: SLEPC_EXTERN PetscBool EPSRegisterAllCalled;
19: SLEPC_EXTERN PetscBool EPSMonitorRegisterAllCalled;
20: SLEPC_EXTERN PetscErrorCode EPSRegisterAll(void);
21: SLEPC_EXTERN PetscErrorCode EPSMonitorRegisterAll(void);
22: SLEPC_EXTERN PetscLogEvent EPS_SetUp,EPS_Solve,EPS_CISS_SVD;
24: typedef struct _EPSOps *EPSOps;
26: struct _EPSOps {
27: PetscErrorCode (*solve)(EPS);
28: PetscErrorCode (*setup)(EPS);
29: PetscErrorCode (*setupsort)(EPS);
30: PetscErrorCode (*setfromoptions)(EPS,PetscOptionItems);
31: PetscErrorCode (*publishoptions)(EPS);
32: PetscErrorCode (*destroy)(EPS);
33: PetscErrorCode (*reset)(EPS);
34: PetscErrorCode (*view)(EPS,PetscViewer);
35: PetscErrorCode (*backtransform)(EPS);
36: PetscErrorCode (*computevectors)(EPS);
37: PetscErrorCode (*setdefaultst)(EPS);
38: PetscErrorCode (*setdstype)(EPS);
39: };
41: /*
42: Maximum number of monitors you can run with a single EPS
43: */
44: #define MAXEPSMONITORS 5
46: /*
47: The solution process goes through several states
48: */
49: typedef enum { EPS_STATE_INITIAL,
50: EPS_STATE_SETUP,
51: EPS_STATE_SOLVED,
52: EPS_STATE_EIGENVECTORS } EPSStateType;
54: /*
55: To classify the different solvers into categories
56: */
57: typedef enum { EPS_CATEGORY_KRYLOV, /* Krylov solver: relies on STApply and STBackTransform (same as OTHER) */
58: EPS_CATEGORY_PRECOND, /* Preconditioned solver: uses ST only to manage preconditioner */
59: EPS_CATEGORY_CONTOUR, /* Contour integral: ST used to solve linear systems at integration points */
60: EPS_CATEGORY_OTHER } EPSSolverType;
62: /*
63: To check for unsupported features at EPSSetUp_XXX()
64: */
65: typedef enum { EPS_FEATURE_BALANCE=1, /* balancing */
66: EPS_FEATURE_ARBITRARY=2, /* arbitrary selection of eigepairs */
67: EPS_FEATURE_REGION=4, /* nontrivial region for filtering */
68: EPS_FEATURE_EXTRACTION=8, /* extraction technique different from Ritz */
69: EPS_FEATURE_CONVERGENCE=16, /* convergence test selected by user */
70: EPS_FEATURE_STOPPING=32, /* stopping test */
71: EPS_FEATURE_THRESHOLD=64, /* threshold stopping test */
72: EPS_FEATURE_TWOSIDED=128 /* two-sided variant */
73: } EPSFeatureType;
75: /*
76: Defines the EPS data structure
77: */
78: struct _p_EPS {
79: PETSCHEADER(struct _EPSOps);
80: /*------------------------- User parameters ---------------------------*/
81: PetscInt max_it; /* maximum number of iterations */
82: PetscInt nev; /* number of eigenvalues to compute */
83: PetscInt ncv; /* number of basis vectors */
84: PetscInt mpd; /* maximum dimension of projected problem */
85: PetscInt nini,ninil; /* number of initial vectors (negative means not copied yet) */
86: PetscInt nds; /* number of basis vectors of deflation space */
87: PetscScalar target; /* target value */
88: PetscReal tol; /* tolerance */
89: PetscReal thres; /* threshold */
90: PetscBool threlative; /* threshold is relative */
91: EPSConv conv; /* convergence test */
92: EPSStop stop; /* stopping test */
93: EPSWhich which; /* which part of the spectrum to be sought */
94: PetscReal inta,intb; /* interval [a,b] for spectrum slicing */
95: EPSProblemType problem_type; /* which kind of problem to be solved */
96: EPSExtraction extraction; /* which kind of extraction to be applied */
97: EPSBalance balance; /* the balancing method */
98: PetscInt balance_its; /* number of iterations of the balancing method */
99: PetscReal balance_cutoff; /* cutoff value for balancing */
100: PetscBool trueres; /* whether the true residual norm must be computed */
101: PetscBool trackall; /* whether all the residuals must be computed */
102: PetscBool purify; /* whether eigenvectors need to be purified */
103: PetscBool twosided; /* whether to compute left eigenvectors (two-sided solver) */
105: /*-------------- User-provided functions and contexts -----------------*/
106: EPSConvergenceTestFn *converged;
107: EPSConvergenceTestFn *convergeduser;
108: PetscCtxDestroyFn *convergeddestroy;
109: EPSStoppingTestFn *stopping;
110: EPSStoppingTestFn *stoppinguser;
111: PetscCtxDestroyFn *stoppingdestroy;
112: SlepcArbitrarySelectionFn *arbitrary;
113: void *convergedctx;
114: void *stoppingctx;
115: void *arbitraryctx;
116: EPSMonitorFn *monitor[MAXEPSMONITORS];
117: PetscCtxDestroyFn *monitordestroy[MAXEPSMONITORS];
118: void *monitorcontext[MAXEPSMONITORS];
119: PetscInt numbermonitors;
121: /*----------------- Child objects and working data -------------------*/
122: ST st; /* spectral transformation object */
123: DS ds; /* direct solver object */
124: BV V; /* set of basis vectors and computed eigenvectors */
125: BV W; /* left basis vectors (if left eigenvectors requested) */
126: RG rg; /* optional region for filtering */
127: SlepcSC sc; /* sorting criterion data */
128: Vec D; /* diagonal matrix for balancing */
129: Vec *IS,*ISL; /* references to user-provided initial spaces */
130: Vec *defl; /* references to user-provided deflation space */
131: PetscScalar *eigr,*eigi; /* real and imaginary parts of eigenvalues */
132: PetscReal *errest; /* error estimates */
133: PetscScalar *rr,*ri; /* values computed by user's arbitrary selection function */
134: PetscInt *perm; /* permutation for eigenvalue ordering */
135: PetscInt nwork; /* number of work vectors */
136: Vec *work; /* work vectors */
137: void *data; /* placeholder for solver-specific stuff */
139: /* ----------------------- Status variables --------------------------*/
140: EPSStateType state; /* initial -> setup -> solved -> eigenvectors */
141: EPSSolverType categ; /* solver category */
142: PetscInt nconv; /* number of converged eigenvalues */
143: PetscInt its; /* number of iterations so far computed */
144: PetscInt n,nloc; /* problem dimensions (global, local) */
145: PetscReal nrma,nrmb; /* computed matrix norms */
146: PetscBool useds; /* whether the solver uses the DS object or not */
147: PetscBool isgeneralized;
148: PetscBool ispositive;
149: PetscBool ishermitian;
150: PetscBool isstructured;
151: EPSConvergedReason reason;
152: };
154: /*
155: Macros to test valid EPS arguments
156: */
157: #if !defined(PETSC_USE_DEBUG)
159: #define EPSCheckSolved(h,arg) do {(void)(h);} while (0)
161: #else
163: #define EPSCheckSolved(h,arg) \
164: do { \
165: PetscCheck((h)->state>=EPS_STATE_SOLVED,PetscObjectComm((PetscObject)(h)),PETSC_ERR_ARG_WRONGSTATE,"Must call EPSSolve() first: Parameter #%d",arg); \
166: } while (0)
168: #endif
170: /*
171: Macros to check settings at EPSSetUp()
172: */
174: /* EPSCheckHermitianDefinite: the problem is HEP or GHEP */
175: #define EPSCheckHermitianDefiniteCondition(eps,condition,msg) \
176: do { \
177: if (condition) { \
178: PetscCheck((eps)->ishermitian,PetscObjectComm((PetscObject)(eps)),PETSC_ERR_SUP,"The solver '%s'%s cannot be used for non-%s problems",((PetscObject)(eps))->type_name,(msg),SLEPC_STRING_HERMITIAN); \
179: PetscCheck(!(eps)->isgeneralized || (eps)->ispositive,PetscObjectComm((PetscObject)(eps)),PETSC_ERR_SUP,"The solver '%s'%s requires that the problem is %s-definite",((PetscObject)(eps))->type_name,(msg),SLEPC_STRING_HERMITIAN); \
180: } \
181: } while (0)
182: #define EPSCheckHermitianDefinite(eps) EPSCheckHermitianDefiniteCondition(eps,PETSC_TRUE,"")
184: /* EPSCheckHermitian: the problem is HEP, GHEP, or GHIEP */
185: #define EPSCheckHermitianCondition(eps,condition,msg) \
186: do { \
187: if (condition) { \
188: PetscCheck((eps)->ishermitian,PetscObjectComm((PetscObject)(eps)),PETSC_ERR_SUP,"The solver '%s'%s cannot be used for non-%s problems",((PetscObject)(eps))->type_name,(msg),SLEPC_STRING_HERMITIAN); \
189: } \
190: } while (0)
191: #define EPSCheckHermitian(eps) EPSCheckHermitianCondition(eps,PETSC_TRUE,"")
193: /* EPSCheckDefinite: the problem is not GHIEP */
194: #define EPSCheckDefiniteCondition(eps,condition,msg) \
195: do { \
196: if (condition) { \
197: PetscCheck(!(eps)->isgeneralized || !(eps)->ishermitian || (eps)->ispositive,PetscObjectComm((PetscObject)(eps)),PETSC_ERR_SUP,"The solver '%s'%s cannot be used for %s-indefinite problems",((PetscObject)(eps))->type_name,(msg),SLEPC_STRING_HERMITIAN); \
198: } \
199: } while (0)
200: #define EPSCheckDefinite(eps) EPSCheckDefiniteCondition(eps,PETSC_TRUE,"")
202: /* EPSCheckStandard: the problem is HEP or NHEP */
203: #define EPSCheckStandardCondition(eps,condition,msg) \
204: do { \
205: if (condition) { \
206: PetscCheck(!(eps)->isgeneralized,PetscObjectComm((PetscObject)(eps)),PETSC_ERR_SUP,"The solver '%s'%s cannot be used for generalized problems",((PetscObject)(eps))->type_name,(msg)); \
207: } \
208: } while (0)
209: #define EPSCheckStandard(eps) EPSCheckStandardCondition(eps,PETSC_TRUE,"")
211: /* EPSCheckNotStructured: the problem is not structured */
212: #define EPSCheckNotStructuredCondition(eps,condition,msg) \
213: do { \
214: if (condition) { \
215: PetscCheck(!(eps)->isstructured,PetscObjectComm((PetscObject)(eps)),PETSC_ERR_SUP,"The solver '%s'%s does not provide support for structured eigenproblems",((PetscObject)(eps))->type_name,(msg)); \
216: } \
217: } while (0)
218: #define EPSCheckNotStructured(eps) EPSCheckNotStructuredCondition(eps,PETSC_TRUE,"")
220: /* EPSCheckSinvert: shift-and-invert ST */
221: #define EPSCheckSinvertCondition(eps,condition,msg) \
222: do { \
223: if (condition) { \
224: PetscBool __flg; \
225: PetscCall(PetscObjectTypeCompare((PetscObject)(eps)->st,STSINVERT,&__flg)); \
226: PetscCheck(__flg,PetscObjectComm((PetscObject)(eps)),PETSC_ERR_SUP,"The solver '%s'%s requires a shift-and-invert spectral transform",((PetscObject)(eps))->type_name,(msg)); \
227: } \
228: } while (0)
229: #define EPSCheckSinvert(eps) EPSCheckSinvertCondition(eps,PETSC_TRUE,"")
231: /* EPSCheckSinvertCayley: shift-and-invert or Cayley ST */
232: #define EPSCheckSinvertCayleyCondition(eps,condition,msg) \
233: do { \
234: if (condition) { \
235: PetscBool __flg; \
236: PetscCall(PetscObjectTypeCompareAny((PetscObject)(eps)->st,&__flg,STSINVERT,STCAYLEY,"")); \
237: PetscCheck(__flg,PetscObjectComm((PetscObject)(eps)),PETSC_ERR_SUP,"The solver '%s'%s requires shift-and-invert or Cayley transform",((PetscObject)(eps))->type_name,(msg)); \
238: } \
239: } while (0)
240: #define EPSCheckSinvertCayley(eps) EPSCheckSinvertCayleyCondition(eps,PETSC_TRUE,"")
242: /* Check for unsupported features */
243: #define EPSCheckUnsupportedCondition(eps,mask,condition,msg) \
244: do { \
245: if (condition) { \
246: PetscCheck(!((mask) & EPS_FEATURE_BALANCE) || (eps)->balance==EPS_BALANCE_NONE,PetscObjectComm((PetscObject)(eps)),PETSC_ERR_SUP,"The solver '%s'%s does not support balancing",((PetscObject)(eps))->type_name,(msg)); \
247: PetscCheck(!((mask) & EPS_FEATURE_ARBITRARY) || !(eps)->arbitrary,PetscObjectComm((PetscObject)(eps)),PETSC_ERR_SUP,"The solver '%s'%s does not support arbitrary selection of eigenpairs",((PetscObject)(eps))->type_name,(msg)); \
248: if ((mask) & EPS_FEATURE_REGION) { \
249: PetscBool __istrivial; \
250: PetscCall(RGIsTrivial((eps)->rg,&__istrivial)); \
251: PetscCheck(__istrivial,PetscObjectComm((PetscObject)(eps)),PETSC_ERR_SUP,"The solver '%s'%s does not support region filtering",((PetscObject)(eps))->type_name,(msg)); \
252: } \
253: PetscCheck(!((mask) & EPS_FEATURE_EXTRACTION) || (eps)->extraction==EPS_RITZ,PetscObjectComm((PetscObject)(eps)),PETSC_ERR_SUP,"The solver '%s'%s only supports Ritz extraction",((PetscObject)(eps))->type_name,(msg)); \
254: PetscCheck(!((mask) & EPS_FEATURE_CONVERGENCE) || (eps)->converged==EPSConvergedRelative,PetscObjectComm((PetscObject)(eps)),PETSC_ERR_SUP,"The solver '%s'%s only supports the default convergence test",((PetscObject)(eps))->type_name,(msg)); \
255: PetscCheck(!((mask) & EPS_FEATURE_STOPPING) || (eps)->stopping==EPSStoppingBasic,PetscObjectComm((PetscObject)(eps)),PETSC_ERR_SUP,"The solver '%s'%s only supports the default stopping test",((PetscObject)(eps))->type_name,(msg)); \
256: PetscCheck(!((mask) & EPS_FEATURE_THRESHOLD) || (eps)->stopping!=EPSStoppingThreshold,PetscObjectComm((PetscObject)(eps)),PETSC_ERR_SUP,"The solver '%s'%s does not support the threshold stopping test",((PetscObject)(eps))->type_name,(msg)); \
257: PetscCheck(!((mask) & EPS_FEATURE_TWOSIDED) || !(eps)->twosided,PetscObjectComm((PetscObject)(eps)),PETSC_ERR_SUP,"The solver '%s'%s cannot compute left eigenvectors (no two-sided variant)",((PetscObject)(eps))->type_name,(msg)); \
258: } \
259: } while (0)
260: #define EPSCheckUnsupported(eps,mask) EPSCheckUnsupportedCondition(eps,mask,PETSC_TRUE,"")
262: /* Check for ignored features */
263: #define EPSCheckIgnoredCondition(eps,mask,condition,msg) \
264: do { \
265: if (condition) { \
266: if (((mask) & EPS_FEATURE_BALANCE) && (eps)->balance!=EPS_BALANCE_NONE) PetscCall(PetscInfo((eps),"The solver '%s'%s ignores the balancing settings\n",((PetscObject)(eps))->type_name,(msg))); \
267: if (((mask) & EPS_FEATURE_ARBITRARY) && (eps)->arbitrary) PetscCall(PetscInfo((eps),"The solver '%s'%s ignores the settings for arbitrary selection of eigenpairs\n",((PetscObject)(eps))->type_name,(msg))); \
268: if ((mask) & EPS_FEATURE_REGION) { \
269: PetscBool __istrivial; \
270: PetscCall(RGIsTrivial((eps)->rg,&__istrivial)); \
271: if (!__istrivial) PetscCall(PetscInfo((eps),"The solver '%s'%s ignores the specified region\n",((PetscObject)(eps))->type_name,(msg))); \
272: } \
273: if (((mask) & EPS_FEATURE_EXTRACTION) && (eps)->extraction!=EPS_RITZ) PetscCall(PetscInfo((eps),"The solver '%s'%s ignores the extraction settings\n",((PetscObject)(eps))->type_name,(msg))); \
274: if (((mask) & EPS_FEATURE_CONVERGENCE) && (eps)->converged!=EPSConvergedRelative) PetscCall(PetscInfo((eps),"The solver '%s'%s ignores the convergence test settings\n",((PetscObject)(eps))->type_name,(msg))); \
275: if (((mask) & EPS_FEATURE_STOPPING) && (eps)->stopping!=EPSStoppingBasic) PetscCall(PetscInfo((eps),"The solver '%s'%s ignores the stopping test settings\n",((PetscObject)(eps))->type_name,(msg))); \
276: if (((mask) & EPS_FEATURE_TWOSIDED) && (eps)->twosided) PetscCall(PetscInfo((eps),"The solver '%s'%s ignores the two-sided flag\n",((PetscObject)(eps))->type_name,(msg))); \
277: } \
278: } while (0)
279: #define EPSCheckIgnored(eps,mask) EPSCheckIgnoredCondition(eps,mask,PETSC_TRUE,"")
281: /*
282: EPSSetCtxThreshold - Fills EPSStoppingCtx with data needed for the threshold stopping test
283: */
284: #define EPSSetCtxThreshold(eps,eigr,eigi,k) \
285: do { \
286: if (eps->stop==EPS_STOP_THRESHOLD && k) { \
287: PetscScalar __kr=eigr[k-1],__ki=eigi[k-1],__kr0=eigr[0],__ki0=eigi[0]; \
288: PetscCall(STBackTransform(eps->st,1,&__kr,&__ki)); \
289: PetscCall(STBackTransform(eps->st,1,&__kr0,&__ki0)); \
290: if (eps->which==EPS_LARGEST_MAGNITUDE || eps->which==EPS_SMALLEST_MAGNITUDE) { \
291: ((EPSStoppingCtx)eps->stoppingctx)->firstev = SlepcAbsEigenvalue(__kr0,__ki0); \
292: ((EPSStoppingCtx)eps->stoppingctx)->lastev = SlepcAbsEigenvalue(__kr,__ki); \
293: } else { \
294: ((EPSStoppingCtx)eps->stoppingctx)->firstev = PetscRealPart(__kr0); \
295: ((EPSStoppingCtx)eps->stoppingctx)->lastev = PetscRealPart(__kr); \
296: } \
297: } \
298: } while (0)
300: /*
301: EPS_SetInnerProduct - set B matrix for inner product if appropriate.
302: */
303: static inline PetscErrorCode EPS_SetInnerProduct(EPS eps)
304: {
305: Mat B;
307: PetscFunctionBegin;
308: if (!eps->V) PetscCall(EPSGetBV(eps,&eps->V));
309: if (eps->ispositive || (eps->isgeneralized && eps->ishermitian)) {
310: PetscCall(STGetBilinearForm(eps->st,&B));
311: PetscCall(BVSetMatrix(eps->V,B,PetscNot(eps->ispositive)));
312: if (eps->twosided) PetscCall(BVSetMatrix(eps->W,B,PetscNot(eps->ispositive)));
313: PetscCall(MatDestroy(&B));
314: } else PetscCall(BVSetMatrix(eps->V,NULL,PETSC_FALSE));
315: PetscFunctionReturn(PETSC_SUCCESS);
316: }
318: /*
319: EPS_Purify - purify the first k vectors in the V basis
320: */
321: static inline PetscErrorCode EPS_Purify(EPS eps,PetscInt k)
322: {
323: PetscInt i;
324: Vec v,z;
326: PetscFunctionBegin;
327: PetscCall(BVCreateVec(eps->V,&v));
328: for (i=0;i<k;i++) {
329: PetscCall(BVCopyVec(eps->V,i,v));
330: PetscCall(BVGetColumn(eps->V,i,&z));
331: PetscCall(STApply(eps->st,v,z));
332: PetscCall(BVRestoreColumn(eps->V,i,&z));
333: }
334: PetscCall(VecDestroy(&v));
335: PetscFunctionReturn(PETSC_SUCCESS);
336: }
338: /*
339: EPS_KSPSetOperators - Sets the KSP matrices, see also ST_KSPSetOperators()
340: */
341: static inline PetscErrorCode EPS_KSPSetOperators(KSP ksp,Mat A,Mat B)
342: {
343: const char *prefix;
345: PetscFunctionBegin;
346: PetscCall(KSPSetOperators(ksp,A,B));
347: PetscCall(MatGetOptionsPrefix(B,&prefix));
348: if (!prefix) {
349: /* set Mat prefix to be the same as KSP to enable setting command-line options (e.g. MUMPS)
350: only applies if the Mat has no user-defined prefix */
351: PetscCall(KSPGetOptionsPrefix(ksp,&prefix));
352: PetscCall(MatSetOptionsPrefix(B,prefix));
353: }
354: PetscFunctionReturn(PETSC_SUCCESS);
355: }
357: /*
358: EPS_GetActualConverged - Gets the actual value of nconv; in special cases the
359: number of available eigenvalues is larger than the computed ones
360: */
361: static inline PetscErrorCode EPS_GetActualConverged(EPS eps,PetscInt *nconv)
362: {
363: PetscFunctionBegin;
364: *nconv = eps->nconv;
365: if (eps->isstructured) {
366: if (eps->problem_type == EPS_BSE && (eps->which == EPS_SMALLEST_MAGNITUDE || eps->which == EPS_LARGEST_MAGNITUDE || eps->which == EPS_TARGET_MAGNITUDE)) *nconv *= 2;
367: if (eps->problem_type == EPS_HAMILT && (eps->which == EPS_SMALLEST_MAGNITUDE || eps->which == EPS_LARGEST_MAGNITUDE || eps->which == EPS_TARGET_MAGNITUDE)) *nconv *= 2;
368: }
369: PetscFunctionReturn(PETSC_SUCCESS);
370: }
372: /*
373: EPS_GetEigenvector - Gets the i-th eigenvector taking into account the case
374: where i exceeds the number of computed vectors (structure-preserving solver).
375: The argument V should be eps->V for right eigenvectors, eps->W for left ones.
376: */
377: static inline PetscErrorCode EPS_GetEigenvector(EPS eps,BV V,PetscInt i,Vec Vr,Vec Vi)
378: {
379: PetscInt k;
380: Vec v0,v1,w,w0,w1;
381: Mat H;
382: IS is[2];
383: #if !defined(PETSC_USE_COMPLEX)
384: PetscInt k0,k1,k2,iquad;
385: PetscReal nrm,nrmr=0.0,nrmi=0.0,sgn;
386: #endif
388: PetscFunctionBegin;
389: if (!eps->isstructured) {
390: k = eps->perm[i];
391: PetscCall(BV_GetEigenvector(V,k,eps->eigi[k],Vr,Vi));
392: } else {
393: if (eps->problem_type == EPS_BSE && (eps->which == EPS_SMALLEST_MAGNITUDE || eps->which == EPS_LARGEST_MAGNITUDE || eps->which == EPS_TARGET_MAGNITUDE)) {
394: /* BSE problem, even index is +lambda, odd index is -lambda */
395: k = eps->perm[i/2];
396: if (i%2) {
397: /* eigenvector of -lambda is J*conj(X) where J=[0 I; I 0] and x is eigenvector of lambda */
398: PetscCall(VecDuplicate(Vr?Vr:Vi,&w));
399: PetscCall(STGetMatrix(eps->st,0,&H));
400: PetscCall(MatNestGetISs(H,is,NULL));
401: if (Vr) {
402: PetscCall(BV_GetEigenvector(V,k,eps->eigi[k],w,NULL));
403: PetscCall(VecConjugate(w));
404: PetscCall(VecGetSubVector(w,is[0],&w0));
405: PetscCall(VecGetSubVector(w,is[1],&w1));
406: PetscCall(VecGetSubVector(Vr,is[0],&v0));
407: PetscCall(VecGetSubVector(Vr,is[1],&v1));
408: PetscCall(VecCopy(w1,v0));
409: PetscCall(VecCopy(w0,v1));
410: PetscCall(VecRestoreSubVector(w,is[0],&w0));
411: PetscCall(VecRestoreSubVector(w,is[1],&w1));
412: PetscCall(VecRestoreSubVector(Vr,is[0],&v0));
413: PetscCall(VecRestoreSubVector(Vr,is[1],&v1));
414: }
415: #if !defined(PETSC_USE_COMPLEX)
416: if (Vi) {
417: PetscCall(BV_GetEigenvector(V,k,eps->eigi[k],NULL,w));
418: PetscCall(VecScale(w,-1.0));
419: PetscCall(VecGetSubVector(w,is[0],&w0));
420: PetscCall(VecGetSubVector(w,is[1],&w1));
421: PetscCall(VecGetSubVector(Vi,is[0],&v0));
422: PetscCall(VecGetSubVector(Vi,is[1],&v1));
423: PetscCall(VecCopy(w1,v0));
424: PetscCall(VecCopy(w0,v1));
425: PetscCall(VecRestoreSubVector(w,is[0],&w0));
426: PetscCall(VecRestoreSubVector(w,is[1],&w1));
427: PetscCall(VecRestoreSubVector(Vi,is[0],&v0));
428: PetscCall(VecRestoreSubVector(Vi,is[1],&v1));
429: }
430: #endif
431: PetscCall(VecDestroy(&w));
432: } else {
433: PetscCall(BV_GetEigenvector(V,k,eps->eigi[k],Vr,Vi));
434: }
435: } else if (eps->problem_type == EPS_HAMILT) {
436: k = eps->perm[i/2];
437: #if !defined(PETSC_USE_COMPLEX)
438: if (eps->eigi[k]==0.0) { /* real eigenvalue */
439: if (Vr) {
440: PetscCall(BVCopyVec(V,k+eps->ncv/2+1,Vr));
441: PetscCall(BVGetColumn(V,k,&w));
442: PetscCall(VecAXPY(Vr,(i%2)?-eps->eigr[k]:eps->eigr[k],w));
443: PetscCall(BVRestoreColumn(V,k,&w));
444: PetscCall(VecNorm(Vr,NORM_2,&nrmr));
445: }
446: if (Vi) PetscCall(VecZeroEntries(Vi));
447: nrm = nrmr;
448: } else if (eps->eigr[k]==0.0 ) { /* purely imaginary eigenvalue */
449: if (Vr) {
450: PetscCall(BVCopyVec(V,k+eps->ncv/2+1,Vr));
451: PetscCall(VecNorm(Vr,NORM_2,&nrmr));
452: }
453: if (Vi) {
454: PetscCall(BVCopyVec(V,k,Vi));
455: PetscCall(VecScale(Vi,(i%2)?-eps->eigi[k]:eps->eigi[k]));
456: PetscCall(VecNorm(Vi,NORM_2,&nrmi));
457: }
458: nrm = SlepcAbs(nrmr,nrmi);
459: } else { /* quadruple eigenvalue (-conj(lambda),-lambda,lambda,conj(lambda)) */
460: iquad = i%2; /* index within the 4 values */
461: if (i>=2) {
462: k2 = eps->perm[(i-2)/2];
463: if (eps->eigr[k]==eps->eigr[k2] && eps->eigi[k]==-eps->eigi[k2]) iquad += 2;
464: }
465: k0 = (iquad<2)? k: k2;
466: k1 = k0+1;
467: /* Vr+Vi*i obtained as eig*u+v where u=ur+ui*i is stored in cols k0 (ur) and k1 (ui) and
468: v=vr+vi*i is in cols shifted by ncv/2+1.
469: For lambda=eigr+eigi*i:
470: Vr+Vi*i = (eigr+eigi*i)(ur+ui*i) + vr+vi*i
471: Vr+Vi*i = eigr*(ur+ui*i) - eigi*ui+eigi*ur*i + vr+vi*i
472: Vr+Vi*i = eigr*ur-eigi*ui+vr + (eigi*ur+eigr*ui+vi)*i
473: For -conj(lambda): eigr, ui and vi have the signs changed
474: For -lambda: eigr and eigi have the signs changed
475: For conj(lambda): eigi, ui and vi have the signs changed */
476: if (Vr) {
477: sgn = (iquad<2)? -1.0: 1.0;
478: PetscCall(BVCopyVec(V,k0,Vr)); /* ur */
479: PetscCall(VecScale(Vr,sgn*eps->eigr[k0]));
480: PetscCall(BVGetColumn(V,k1,&w)); /* ui */
481: PetscCall(VecAXPY(Vr,-sgn*eps->eigi[k0],w));
482: PetscCall(BVRestoreColumn(V,k1,&w));
483: PetscCall(BVGetColumn(V,k0+eps->ncv/2+1,&w)); /* vr */
484: PetscCall(VecAXPY(Vr,1.0,w));
485: PetscCall(BVRestoreColumn(V,k0+eps->ncv/2+1,&w));
486: PetscCall(VecNorm(Vr,NORM_2,&nrmr));
487: }
488: if (Vi) {
489: sgn = (iquad%2)? -1.0: 1.0;
490: PetscCall(BVCopyVec(V,k0,Vi)); /* ur */
491: PetscCall(VecScale(Vi,sgn*eps->eigi[k0]));
492: PetscCall(BVGetColumn(V,k1,&w)); /* ui */
493: PetscCall(VecAXPY(Vi,sgn*eps->eigr[k0],w));
494: PetscCall(BVRestoreColumn(V,k1,&w));
495: PetscCall(BVGetColumn(V,k1+eps->ncv/2+1,&w)); /* vi */
496: sgn = (iquad%3)? 1.0: -1.0;
497: PetscCall(VecAXPY(Vi,sgn,w));
498: PetscCall(BVRestoreColumn(V,k1+eps->ncv/2+1,&w));
499: PetscCall(VecNorm(Vi,NORM_2,&nrmi));
500: }
501: nrm = SlepcAbs(nrmr,nrmi);
502: }
503: if (Vr) PetscCall(VecScale(Vr,1.0/nrm));
504: if (Vi) PetscCall(VecScale(Vi,1.0/nrm));
505: #endif
506: } else SETERRQ(PetscObjectComm((PetscObject)eps),PETSC_ERR_LIB,"Inconsistent state");
507: }
508: PetscFunctionReturn(PETSC_SUCCESS);
509: }
511: SLEPC_INTERN PetscErrorCode EPSSetWhichEigenpairs_Default(EPS);
512: SLEPC_INTERN PetscErrorCode EPSSetDimensions_Default(EPS,PetscInt*,PetscInt*,PetscInt*);
513: SLEPC_INTERN PetscErrorCode EPSBackTransform_Default(EPS);
514: SLEPC_INTERN PetscErrorCode EPSComputeVectors(EPS);
515: SLEPC_INTERN PetscErrorCode EPSComputeVectors_Hermitian(EPS);
516: SLEPC_INTERN PetscErrorCode EPSComputeVectors_Schur(EPS);
517: SLEPC_INTERN PetscErrorCode EPSComputeVectors_Indefinite(EPS);
518: SLEPC_INTERN PetscErrorCode EPSComputeVectors_Twosided(EPS);
519: SLEPC_INTERN PetscErrorCode EPSComputeVectors_Slice(EPS);
520: SLEPC_INTERN PetscErrorCode EPSComputeResidualNorm_Private(EPS,PetscBool,PetscScalar,PetscScalar,Vec,Vec,Vec*,PetscReal*);
521: SLEPC_INTERN PetscErrorCode EPSComputeRitzVector(EPS,PetscScalar*,PetscScalar*,BV,Vec,Vec);
522: SLEPC_INTERN PetscErrorCode EPSGetStartVector(EPS,PetscInt,PetscBool*);
523: SLEPC_INTERN PetscErrorCode EPSGetLeftStartVector(EPS,PetscInt,PetscBool*);
524: SLEPC_INTERN PetscErrorCode MatEstimateSpectralRange_EPS(Mat,PetscReal*,PetscReal*);
526: /* Private functions of the solver implementations */
528: SLEPC_INTERN PetscErrorCode EPSDelayedArnoldi(EPS,PetscScalar*,PetscInt,PetscInt,PetscInt*,PetscReal*,PetscBool*);
529: SLEPC_INTERN PetscErrorCode EPSDelayedArnoldi1(EPS,PetscScalar*,PetscInt,PetscInt,PetscInt*,PetscReal*,PetscBool*);
530: SLEPC_INTERN PetscErrorCode EPSKrylovConvergence(EPS,PetscBool,PetscInt,PetscInt,PetscReal,PetscReal,PetscReal,PetscInt*);
531: SLEPC_INTERN PetscErrorCode EPSPseudoLanczos(EPS,PetscReal*,PetscReal*,PetscReal*,PetscInt,PetscInt*,PetscBool*,PetscBool*,PetscReal*,Vec);
532: SLEPC_INTERN PetscErrorCode EPSBuildBalance_Krylov(EPS);
533: SLEPC_INTERN PetscErrorCode EPSSetDefaultST(EPS);
534: SLEPC_INTERN PetscErrorCode EPSSetDefaultST_Precond(EPS);
535: SLEPC_INTERN PetscErrorCode EPSSetDefaultST_GMRES(EPS);
536: SLEPC_INTERN PetscErrorCode EPSSetDefaultST_NoFactor(EPS);
537: SLEPC_INTERN PetscErrorCode EPSSetUpSort_Basic(EPS);
538: SLEPC_INTERN PetscErrorCode EPSSetUpSort_Default(EPS);