/usr/include/trilinos/Ifpack_AdditiveSchwarz.h is in libtrilinos-ifpack-dev 12.12.1-5.
This file is owned by root:root, with mode 0o644.
The actual contents of the file can be viewed below.
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 409 410 411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 432 433 434 435 436 437 438 439 440 441 442 443 444 445 446 447 448 449 450 451 452 453 454 455 456 457 458 459 460 461 462 463 464 465 466 467 468 469 470 471 472 473 474 475 476 477 478 479 480 481 482 483 484 485 486 487 488 489 490 491 492 493 494 495 496 497 498 499 500 501 502 503 504 505 506 507 508 509 510 511 512 513 514 515 516 517 518 519 520 521 522 523 524 525 526 527 528 529 530 531 532 533 534 535 536 537 538 539 540 541 542 543 544 545 546 547 548 549 550 551 552 553 554 555 556 557 558 559 560 561 562 563 564 565 566 567 568 569 570 571 572 573 574 575 576 577 578 579 580 581 582 583 584 585 586 587 588 589 590 591 592 593 594 595 596 597 598 599 600 601 602 603 604 605 606 607 608 609 610 611 612 613 614 615 616 617 618 619 620 621 622 623 624 625 626 627 628 629 630 631 632 633 634 635 636 637 638 639 640 641 642 643 644 645 646 647 648 649 650 651 652 653 654 655 656 657 658 659 660 661 662 663 664 665 666 667 668 669 670 671 672 673 674 675 676 677 678 679 680 681 682 683 684 685 686 687 688 689 690 691 692 693 694 695 696 697 698 699 700 701 702 703 704 705 706 707 708 709 710 711 712 713 714 715 716 717 718 719 720 721 722 723 724 725 726 727 728 729 730 731 732 733 734 735 736 737 738 739 740 741 742 743 744 745 746 747 748 749 750 751 752 753 754 755 756 757 758 759 760 761 762 763 764 765 766 767 768 769 770 771 772 773 774 775 776 777 778 779 780 781 782 783 784 785 786 787 788 789 790 791 792 793 794 795 796 797 798 799 800 801 802 803 804 805 806 807 808 809 810 811 812 813 814 815 816 817 818 819 820 821 822 823 824 825 826 827 828 829 830 831 832 833 834 835 836 837 838 839 840 841 842 843 844 845 846 847 848 849 850 851 852 853 854 855 856 857 858 859 860 861 862 863 864 865 866 867 868 869 870 871 872 873 874 875 876 877 878 879 880 881 882 883 884 885 886 887 888 889 890 891 892 893 894 895 896 897 898 899 900 901 902 903 904 905 906 907 908 909 910 911 912 913 914 915 916 917 918 919 920 921 922 923 924 925 926 927 928 929 930 931 932 933 934 935 936 937 938 939 940 941 942 943 944 945 946 947 948 949 950 951 952 953 954 955 956 957 958 959 960 961 962 963 964 965 966 967 968 969 970 971 972 973 974 975 976 977 978 979 980 981 982 983 984 985 986 987 988 989 990 991 992 993 994 995 996 997 998 999 1000 1001 1002 1003 1004 1005 1006 1007 1008 1009 1010 1011 1012 1013 1014 1015 1016 1017 1018 1019 1020 1021 1022 1023 1024 1025 1026 1027 1028 1029 1030 1031 1032 1033 1034 1035 1036 1037 1038 1039 1040 1041 1042 1043 1044 1045 1046 1047 1048 1049 1050 1051 1052 1053 1054 1055 1056 1057 1058 1059 1060 1061 1062 1063 1064 1065 1066 1067 1068 1069 1070 1071 1072 1073 1074 1075 1076 1077 1078 1079 1080 1081 1082 1083 1084 1085 1086 1087 1088 1089 1090 1091 1092 1093 1094 1095 1096 1097 1098 1099 1100 1101 1102 1103 1104 1105 1106 1107 1108 1109 1110 1111 1112 1113 1114 1115 1116 1117 1118 1119 1120 1121 1122 1123 1124 1125 1126 1127 1128 1129 1130 1131 1132 1133 1134 1135 1136 1137 1138 1139 1140 1141 1142 1143 1144 1145 1146 1147 1148 1149 1150 1151 1152 1153 1154 1155 1156 1157 1158 1159 1160 1161 1162 1163 1164 1165 1166 1167 1168 1169 1170 1171 1172 1173 1174 1175 1176 1177 1178 1179 1180 1181 1182 1183 1184 1185 1186 1187 1188 1189 1190 1191 1192 1193 1194 1195 1196 1197 1198 1199 1200 1201 1202 1203 1204 1205 1206 1207 1208 1209 1210 1211 1212 1213 1214 1215 1216 1217 1218 1219 1220 1221 1222 1223 1224 1225 1226 1227 1228 1229 1230 1231 1232 1233 1234 1235 | /*
//@HEADER
// ***********************************************************************
//
// Ifpack: Object-Oriented Algebraic Preconditioner Package
// Copyright (2002) Sandia Corporation
//
// Under terms of Contract DE-AC04-94AL85000, there is a non-exclusive
// license for use of this work by or on behalf of the U.S. Government.
//
// Redistribution and use in source and binary forms, with or without
// modification, are permitted provided that the following conditions are
// met:
//
// 1. Redistributions of source code must retain the above copyright
// notice, this list of conditions and the following disclaimer.
//
// 2. Redistributions in binary form must reproduce the above copyright
// notice, this list of conditions and the following disclaimer in the
// documentation and/or other materials provided with the distribution.
//
// 3. Neither the name of the Corporation nor the names of the
// contributors may be used to endorse or promote products derived from
// this software without specific prior written permission.
//
// THIS SOFTWARE IS PROVIDED BY SANDIA CORPORATION "AS IS" AND ANY
// EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
// IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR
// PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL SANDIA CORPORATION OR THE
// CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL,
// EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO,
// PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR
// PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF
// LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING
// NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS
// SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
//
// Questions? Contact Michael A. Heroux (maherou@sandia.gov)
//
// ***********************************************************************
//@HEADER
*/
#ifndef IFPACK_ADDITIVESCHWARZ_H
#define IFPACK_ADDITIVESCHWARZ_H
#include "Ifpack_ConfigDefs.h"
#include "Ifpack_Preconditioner.h"
#include "Ifpack_ConfigDefs.h"
#include "Ifpack_Preconditioner.h"
#include "Ifpack_Reordering.h"
#include "Ifpack_RCMReordering.h"
#include "Ifpack_METISReordering.h"
#include "Ifpack_LocalFilter.h"
#include "Ifpack_NodeFilter.h"
#include "Ifpack_SingletonFilter.h"
#include "Ifpack_ReorderFilter.h"
#include "Ifpack_Utils.h"
#include "Ifpack_OverlappingRowMatrix.h"
#include "Epetra_CombineMode.h"
#include "Epetra_MultiVector.h"
#include "Epetra_Map.h"
#include "Epetra_Comm.h"
#include "Epetra_Time.h"
#include "Epetra_LinearProblem.h"
#include "Epetra_RowMatrix.h"
#include "Epetra_CrsMatrix.h"
#include "Teuchos_ParameterList.hpp"
#include "Teuchos_RefCountPtr.hpp"
#ifdef HAVE_IFPACK_PARALLEL_SUBDOMAIN_SOLVERS
#include "Ifpack_SubdomainFilter.h"
#include "EpetraExt_Reindex_CrsMatrix.h"
#include "EpetraExt_Reindex_MultiVector.h"
#endif
#ifdef IFPACK_NODE_AWARE_CODE
#include "EpetraExt_OperatorOut.h"
#include "EpetraExt_RowMatrixOut.h"
#include "EpetraExt_BlockMapOut.h"
#endif
#ifdef HAVE_IFPACK_AMESOS
#include "Ifpack_AMDReordering.h"
#endif
//! Ifpack_AdditiveSchwarz: a class to define Additive Schwarz preconditioners of Epetra_RowMatrix's.
/*!
Class Ifpack_AdditiveSchwarz enables the construction of Additive
Schwarz (one-level overlapping domain decomposition) preconditioners,
for a given Epetra_RowMatrix.
Ifpack_AdditiveSchwarz is derived from Ifpack_Preconditioner,
itself derived from Epetra_Operator. An application of
the Additive Schwarz preconditioner can be obtained
by calling method ApplyInverse().
One-level overlapping domain decomposition preconditioners use
local solvers, of Dirichlet type. This means that the inverse of
the local matrix (with minimal or wider overlap) is applied to
the residual to be preconditioned.
The preconditioner can be written as:
\f[
P_{AS}^{-1} = \sum_{i=1}^M P_i A_i^{-1} R_i ,
\f]
where \f$M\f$ is the number of subdomains (that is, the number of
processors in
the computation), \f$R_i\f$ is an operator that restricts the global
vector to the vector lying on subdomain \f$i\f$, \f$P_i\f$ is the
prolongator operator, and
\f[
A_i = R_i A P_i.
\f]
The construction of Schwarz preconditioners is mainly composed by
two steps:
- definition of the restriction and prolongation operator
\f$R_i\f$ and \f$R_i^T\f$. If minimal overlap is chosen, their
implementation is trivial, \f$R_i\f$ will return all the local
components. For wider overlaps, instead, Epetra_Import and
Epetra_Export will be used to import/export data. The user
must provide both the matrix to be preconditioned (which is suppose
to have minimal-overlap) and the matrix with wider overlap.
- definition of a technique to apply the inverse of \f$A_i\f$.
To solve on each subdomain, the user can adopt any class, derived
from Ifpack_Preconditioner. This can be easily accomplished, as
Ifpack_AdditiveSchwarz is templated with the solver for each subdomain.
The local matrix \f$A_i\f$ can be filtered, to eliminate singletons, and
reordered. At the present time, RCM and METIS can be used to reorder the
local matrix.
The complete list of supported parameters is reported in page \ref ifp_params.
\author Marzio Sala, SNL 9214.
\date Last modified on 22-Jan-05.
*/
template<typename T>
class Ifpack_AdditiveSchwarz : public virtual Ifpack_Preconditioner {
public:
//@{ \name Constructors/Destructors
//! Ifpack_AdditiveSchwarz constructor with given Epetra_RowMatrix.
/*! Creates an Ifpack_AdditiveSchwarz preconditioner with overlap.
* To use minimal-overlap, OverlappingMatrix is omitted
* (as defaulted to 0).
*
* \param
* Matrix - (In) Pointer to matrix to be preconditioned
*
* \param
* OverlappingMatrix - (In) Pointer to the matrix extended with the
* desired level of overlap.
*/
Ifpack_AdditiveSchwarz(Epetra_RowMatrix* Matrix_in,
int OverlapLevel_in = 0);
//! Destructor
virtual ~Ifpack_AdditiveSchwarz() {};
//@}
//@{ \name Attribute set methods.
//! If set true, transpose of this operator will be applied (not implemented).
/*! This flag allows the transpose of the given operator to be used
* implicitly.
\param
UseTranspose_in - (In) If true, multiply by the transpose of operator,
otherwise just use operator.
\return Integer error code, set to 0 if successful. Set to -1 if this implementation does not support transpose.
*/
virtual int SetUseTranspose(bool UseTranspose_in);
//@}
//@{ \name Mathematical functions.
//! Applies the matrix to X, returns the result in Y.
/*!
\param
X - (In) A Epetra_MultiVector of dimension NumVectors
to multiply with matrix.
\param
Y -(Out) A Epetra_MultiVector of dimension NumVectors
containing the result.
\return Integer error code, set to 0 if successful.
*/
virtual int Apply(const Epetra_MultiVector& X, Epetra_MultiVector& Y) const;
//! Applies the preconditioner to X, returns the result in Y.
/*!
\param
X - (In) A Epetra_MultiVector of dimension NumVectors to be preconditioned.
\param
Y -(Out) A Epetra_MultiVector of dimension NumVectors containing result.
\return Integer error code, set to 0 if successful.
\warning In order to work with AztecOO, any implementation of this method
must support the case where X and Y are the same object.
*/
virtual int ApplyInverse(const Epetra_MultiVector& X, Epetra_MultiVector& Y) const;
//! Returns the infinity norm of the global matrix (not implemented)
virtual double NormInf() const;
//@}
//@{ \name Attribute access functions
//! Returns a character string describing the operator
virtual const char * Label() const;
//! Returns the current UseTranspose setting.
virtual bool UseTranspose() const;
//! Returns true if the \e this object can provide an approximate Inf-norm, false otherwise.
virtual bool HasNormInf() const;
//! Returns a pointer to the Epetra_Comm communicator associated with this operator.
virtual const Epetra_Comm & Comm() const;
//! Returns the Epetra_Map object associated with the domain of this operator.
virtual const Epetra_Map & OperatorDomainMap() const;
//! Returns the Epetra_Map object associated with the range of this operator.
virtual const Epetra_Map & OperatorRangeMap() const;
//@}
//! Returns \c true if the preconditioner has been successfully initialized.
virtual bool IsInitialized() const
{
return(IsInitialized_);
}
//! Returns \c true if the preconditioner has been successfully computed.
virtual bool IsComputed() const
{
return(IsComputed_);
}
//! Sets the parameters.
/*! Sets the parameter for the additive Schwarz preconditioner,
* as well as for all the preconditioners that may need to
* be defined on each subblock.
* Parameters accepted by List are:
* - \c "schwarz: combine mode" : It must be an Epetra_CombineMode.
* Default: \c Zero.
* It Can be assume of the following values:
* - Add: Components on the receiving processor will be added together;
* - Zero: Off-processor components will be ignored;
* - Insert: Off-processor components will be inserted into locations on
* receiving processor replacing existing values.
* - Average: Off-processor components will be averaged with existing;
* - AbsMax: Magnitudes of Off-processor components will be
* maxed with magnitudes of existing components on the receiving
* processor.
* - \c "schwarz: compute condest" : if \c true, \c Compute() will
* estimate the condition number of the preconditioner.
* Default: \c true.
*/
virtual int SetParameters(Teuchos::ParameterList& List);
// @}
// @{ Query methods
//! Initialized the preconditioner.
virtual int Initialize();
//! Computes the preconditioner.
virtual int Compute();
//! Computes the estimated condition number and returns its value.
virtual double Condest(const Ifpack_CondestType CT = Ifpack_Cheap,
const int MaxIters = 1550,
const double Tol = 1e-9,
Epetra_RowMatrix* Matrix_in = 0);
//! Returns the estimated condition number, or -1.0 if not computed.
virtual double Condest() const
{
return(Condest_);
}
//! Returns a refernence to the internally stored matrix.
virtual const Epetra_RowMatrix& Matrix() const
{
return(*Matrix_);
}
//! Returns \c true is an overlapping matrix is present.
virtual bool IsOverlapping() const
{
return(IsOverlapping_);
}
//! Prints major information about this preconditioner.
virtual std::ostream& Print(std::ostream&) const;
virtual const T* Inverse() const
{
return(&*Inverse_);
}
//! Returns the number of calls to Initialize().
virtual int NumInitialize() const
{
return(NumInitialize_);
}
//! Returns the number of calls to Compute().
virtual int NumCompute() const
{
return(NumCompute_);
}
//! Returns the number of calls to ApplyInverse().
virtual int NumApplyInverse() const
{
return(NumApplyInverse_);
}
//! Returns the time spent in Initialize().
virtual double InitializeTime() const
{
return(InitializeTime_);
}
//! Returns the time spent in Compute().
virtual double ComputeTime() const
{
return(ComputeTime_);
}
//! Returns the time spent in ApplyInverse().
virtual double ApplyInverseTime() const
{
return(ApplyInverseTime_);
}
//! Returns the number of flops in the initialization phase.
virtual double InitializeFlops() const
{
return(InitializeFlops_);
}
virtual double ComputeFlops() const
{
return(ComputeFlops_);
}
virtual double ApplyInverseFlops() const
{
return(ApplyInverseFlops_);
}
//! Returns the level of overlap.
virtual int OverlapLevel() const
{
return(OverlapLevel_);
}
//! Returns a reference to the internally stored list.
virtual const Teuchos::ParameterList& List() const
{
return(List_);
}
protected:
// @}
// @{ Internal merhods.
//! Copy constructor (should never be used)
Ifpack_AdditiveSchwarz(const Ifpack_AdditiveSchwarz& RHS)
{ }
//! Sets up the localized matrix and the singleton filter.
int Setup();
// @}
// @{ Internal data.
//! Pointers to the matrix to be preconditioned.
Teuchos::RefCountPtr<const Epetra_RowMatrix> Matrix_;
//! Pointers to the overlapping matrix.
Teuchos::RefCountPtr<Ifpack_OverlappingRowMatrix> OverlappingMatrix_;
//! Localized version of Matrix_ or OverlappingMatrix_.
/*
//TODO if we choose to expose the node aware code, i.e., no ifdefs,
//TODO then we should switch to this definition.
Teuchos::RefCountPtr<Epetra_RowMatrix> LocalizedMatrix_;
*/
#ifdef HAVE_IFPACK_PARALLEL_SUBDOMAIN_SOLVERS
//! The subdomain matrix (either LocalFilter or SubdomainFilter)
Teuchos::RCP<Epetra_RowMatrix> LocalizedMatrix_;
//! The subdomain matrix in Epetra_CrsMatrix format
Teuchos::RCP<Epetra_CrsMatrix> SubdomainCrsMatrix_;
//! The reindexed version of SubdomainCrsMatrix_
Teuchos::RCP<Epetra_CrsMatrix> ReindexedCrsMatrix_;
// The following data members are needed when doing ApplyInverse
// with an Ifpack_SubdomainFilter local matrix
Teuchos::RCP<Epetra_Map> tempMap_;
Teuchos::RCP<Epetra_Map> tempDomainMap_;
Teuchos::RCP<Epetra_Map> tempRangeMap_;
Teuchos::RCP<EpetraExt::CrsMatrix_Reindex> SubdomainMatrixReindexer_;
Teuchos::RCP<EpetraExt::MultiVector_Reindex> DomainVectorReindexer_;
Teuchos::RCP<EpetraExt::MultiVector_Reindex> RangeVectorReindexer_;
mutable Teuchos::RCP<Epetra_MultiVector> tempX_;
mutable Teuchos::RCP<Epetra_MultiVector> tempY_;
#else
# ifdef IFPACK_NODE_AWARE_CODE
Teuchos::RefCountPtr<Ifpack_NodeFilter> LocalizedMatrix_;
# else
Teuchos::RefCountPtr<Ifpack_LocalFilter> LocalizedMatrix_;
# endif
#endif
#ifdef HAVE_IFPACK_PARALLEL_SUBDOMAIN_SOLVERS
// Some data members used for determining the subdomain id (color)
//! The MPI rank of the current process
int MpiRank_;
//! The total number of MPI processes in the simulation
int NumMpiProcs_;
//! The number of MPI processes per subdomain
int NumMpiProcsPerSubdomain_;
//! The number of subdomains
int NumSubdomains_;
//! The unique ID associated with the subdomain
int SubdomainId_;
#endif
//! Contains the label of \c this object.
std::string Label_;
//! If true, the preconditioner has been successfully initialized.
bool IsInitialized_;
//! If true, the preconditioner has been successfully computed.
bool IsComputed_;
//! If \c true, solve with the transpose (not supported by all solvers).
bool UseTranspose_;
//! If true, overlapping is used
bool IsOverlapping_;
//! Level of overlap among the processors.
int OverlapLevel_;
//! Stores a copy of the list given in SetParameters()
Teuchos::ParameterList List_;
//! Combine mode for off-process elements (only if overlap is used)
Epetra_CombineMode CombineMode_;
//! Contains the estimated condition number.
double Condest_;
//! If \c true, compute the condition number estimate each time Compute() is called.
bool ComputeCondest_;
//! If \c true, reorder the local matrix.
bool UseReordering_;
//! Type of reordering of the local matrix.
std::string ReorderingType_;
//! Pointer to a reordering object.
Teuchos::RefCountPtr<Ifpack_Reordering> Reordering_;
//! Pointer to the reorderd matrix.
Teuchos::RefCountPtr<Ifpack_ReorderFilter> ReorderedLocalizedMatrix_;
//! Filter for singletons.
bool FilterSingletons_;
//! filtering object.
Teuchos::RefCountPtr<Ifpack_SingletonFilter> SingletonFilter_;
//! Contains the number of successful calls to Initialize().
int NumInitialize_;
//! Contains the number of successful call to Compute().
int NumCompute_;
//! Contains the number of successful call to ApplyInverse().
mutable int NumApplyInverse_;
//! Contains the time for all successful calls to Initialize().
double InitializeTime_;
//! Contains the time for all successful calls to Compute().
double ComputeTime_;
//! Contains the time for all successful calls to ApplyInverse().
mutable double ApplyInverseTime_;
//! Contains the number of flops for Initialize().
double InitializeFlops_;
//! Contains the number of flops for Compute().
double ComputeFlops_;
//! Contain sthe number of flops for ApplyInverse().
mutable double ApplyInverseFlops_;
//! Object used for timing purposes.
Teuchos::RefCountPtr<Epetra_Time> Time_;
//! Pointer to the local solver.
Teuchos::RefCountPtr<T> Inverse_;
//! Vectors used in overlap solve.
#ifdef HAVE_IFPACK_PARALLEL_SUBDOMAIN_SOLVERS
mutable Teuchos::RefCountPtr<Epetra_MultiVector> OverlappingX;
mutable Teuchos::RefCountPtr<Epetra_MultiVector> OverlappingY;
#endif
# ifdef IFPACK_NODE_AWARE_CODE
mutable Teuchos::RefCountPtr<Epetra_MultiVector> OverlappingX;
mutable Teuchos::RefCountPtr<Epetra_MultiVector> OverlappingY;
#endif
}; // class Ifpack_AdditiveSchwarz<T>
//==============================================================================
template<typename T>
Ifpack_AdditiveSchwarz<T>::
Ifpack_AdditiveSchwarz(Epetra_RowMatrix* Matrix_in,
int OverlapLevel_in) :
#ifdef HAVE_IFPACK_PARALLEL_SUBDOMAIN_SOLVERS
MpiRank_(0),
NumMpiProcs_(1),
NumMpiProcsPerSubdomain_(1),
NumSubdomains_(1),
SubdomainId_(0),
#endif
IsInitialized_(false),
IsComputed_(false),
UseTranspose_(false),
IsOverlapping_(false),
OverlapLevel_(OverlapLevel_in),
CombineMode_(Zero),
Condest_(-1.0),
ComputeCondest_(true),
UseReordering_(false),
ReorderingType_("none"),
FilterSingletons_(false),
NumInitialize_(0),
NumCompute_(0),
NumApplyInverse_(0),
InitializeTime_(0.0),
ComputeTime_(0.0),
ApplyInverseTime_(0.0),
InitializeFlops_(0.0),
ComputeFlops_(0.0),
ApplyInverseFlops_(0.0)
{
// Construct a reference-counted pointer with the input matrix, don't manage the memory.
Matrix_ = Teuchos::rcp( Matrix_in, false );
if (Matrix_->Comm().NumProc() == 1)
OverlapLevel_ = 0;
if ((OverlapLevel_ != 0) && (Matrix_->Comm().NumProc() > 1))
IsOverlapping_ = true;
// Sets parameters to default values
Teuchos::ParameterList List_in;
SetParameters(List_in);
}
#ifdef IFPACK_NODE_AWARE_CODE
extern int ML_NODE_ID;
#endif
//==============================================================================
template<typename T>
int Ifpack_AdditiveSchwarz<T>::Setup()
{
using std::cerr;
using std::endl;
Epetra_RowMatrix* MatrixPtr;
#ifndef HAVE_IFPACK_PARALLEL_SUBDOMAIN_SOLVERS
# ifdef IFPACK_NODE_AWARE_CODE
/*
sleep(3);
if (Comm().MyPID() == 0) cout << "Printing out ovArowmap" << endl;
Comm().Barrier();
EpetraExt::BlockMapToMatrixMarketFile("ovArowmap",OverlappingMatrix_->RowMatrixRowMap());
if (Comm().MyPID() == 0) cout << "Printing out ovAcolmap" << endl;
Comm().Barrier();
EpetraExt::BlockMapToMatrixMarketFile("ovAcolmap",OverlappingMatrix_->RowMatrixColMap());
Comm().Barrier();
*/
/*
EpetraExt::RowMatrixToMatlabFile("ovA",*OverlappingMatrix_);
fprintf(stderr,"p %d n %d matrix file done\n",Comm().MyPID(),ML_NODE_ID);
Comm().Barrier();
*/
int nodeID;
try{ nodeID = List_.get("ML node id",0);}
catch(...){fprintf(stderr,"%s","Ifpack_AdditiveSchwarz<T>::Setup(): no parameter \"ML node id\"\n\n");
std::cout << List_ << std::endl;}
# endif
#endif
try{
if (OverlappingMatrix_ != Teuchos::null)
{
#ifdef HAVE_IFPACK_PARALLEL_SUBDOMAIN_SOLVERS
if (NumMpiProcsPerSubdomain_ > 1) {
LocalizedMatrix_ = Teuchos::rcp(new Ifpack_SubdomainFilter(OverlappingMatrix_, SubdomainId_));
} else {
LocalizedMatrix_ = Teuchos::rcp(new Ifpack_LocalFilter(OverlappingMatrix_));
}
#else
# ifdef IFPACK_NODE_AWARE_CODE
Ifpack_NodeFilter *tt = new Ifpack_NodeFilter(OverlappingMatrix_,nodeID); //FIXME
LocalizedMatrix_ = Teuchos::rcp(tt);
//LocalizedMatrix_ = Teuchos::rcp( new Ifpack_LocalFilter(OverlappingMatrix_) );
# else
LocalizedMatrix_ = Teuchos::rcp( new Ifpack_LocalFilter(OverlappingMatrix_) );
# endif
#endif
}
else
{
#ifdef HAVE_IFPACK_PARALLEL_SUBDOMAIN_SOLVERS
if (NumMpiProcsPerSubdomain_ > 1) {
LocalizedMatrix_ = Teuchos::rcp(new Ifpack_SubdomainFilter(Matrix_, SubdomainId_));
} else {
LocalizedMatrix_ = Teuchos::rcp(new Ifpack_LocalFilter(Matrix_));
}
#else
# ifdef IFPACK_NODE_AWARE_CODE
Ifpack_NodeFilter *tt = new Ifpack_NodeFilter(Matrix_,nodeID); //FIXME
LocalizedMatrix_ = Teuchos::rcp(tt);
//LocalizedMatrix_ = Teuchos::rcp( new Ifpack_LocalFilter(Matrix_) );
# else
LocalizedMatrix_ = Teuchos::rcp( new Ifpack_LocalFilter(Matrix_) );
# endif
#endif
}
}
catch(...) {
fprintf(stderr,"%s","AdditiveSchwarz Setup: problem creating local filter matrix.\n");
}
if (LocalizedMatrix_ == Teuchos::null)
IFPACK_CHK_ERR(-5);
// users may want to skip singleton check
if (FilterSingletons_) {
SingletonFilter_ = Teuchos::rcp( new Ifpack_SingletonFilter(LocalizedMatrix_) );
MatrixPtr = &*SingletonFilter_;
}
else
MatrixPtr = &*LocalizedMatrix_;
if (UseReordering_) {
// create reordering and compute it
if (ReorderingType_ == "rcm")
Reordering_ = Teuchos::rcp( new Ifpack_RCMReordering() );
else if (ReorderingType_ == "metis")
Reordering_ = Teuchos::rcp( new Ifpack_METISReordering() );
#ifdef HAVE_IFPACK_AMESOS
else if (ReorderingType_ == "amd" )
Reordering_ = Teuchos::rcp( new Ifpack_AMDReordering() );
#endif
else {
cerr << "reordering type not correct (" << ReorderingType_ << ")" << endl;
exit(EXIT_FAILURE);
}
if (Reordering_ == Teuchos::null) IFPACK_CHK_ERR(-5);
IFPACK_CHK_ERR(Reordering_->SetParameters(List_));
IFPACK_CHK_ERR(Reordering_->Compute(*MatrixPtr));
// now create reordered localized matrix
ReorderedLocalizedMatrix_ =
Teuchos::rcp( new Ifpack_ReorderFilter(Teuchos::rcp( MatrixPtr, false ), Reordering_) );
if (ReorderedLocalizedMatrix_ == Teuchos::null) IFPACK_CHK_ERR(-5);
MatrixPtr = &*ReorderedLocalizedMatrix_;
}
#ifdef HAVE_IFPACK_PARALLEL_SUBDOMAIN_SOLVERS
// The subdomain matrix needs to be reindexed by Amesos so we need to make a CrsMatrix
// and then reindex it with EpetraExt.
// The reindexing is done here because this feature is only implemented in Amesos_Klu,
// and it is desired to have reindexing with other direct solvers in the Amesos package
SubdomainCrsMatrix_.reset(new Epetra_CrsMatrix(Copy, MatrixPtr->RowMatrixRowMap(), -1));
Teuchos::RCP<Epetra_Import> tempImporter = Teuchos::rcp(new Epetra_Import(SubdomainCrsMatrix_->Map(), MatrixPtr->Map()));
SubdomainCrsMatrix_->Import(*MatrixPtr, *tempImporter, Insert);
SubdomainCrsMatrix_->FillComplete();
if (NumMpiProcsPerSubdomain_ > 1) {
tempMap_.reset(new Epetra_Map(SubdomainCrsMatrix_->RowMap().NumGlobalElements(),
SubdomainCrsMatrix_->RowMap().NumMyElements(),
0, SubdomainCrsMatrix_->Comm()));
tempRangeMap_.reset(new Epetra_Map(SubdomainCrsMatrix_->OperatorRangeMap().NumGlobalElements(),
SubdomainCrsMatrix_->OperatorRangeMap().NumMyElements(),
0, SubdomainCrsMatrix_->Comm()));
tempDomainMap_.reset(new Epetra_Map(SubdomainCrsMatrix_->OperatorDomainMap().NumGlobalElements(),
SubdomainCrsMatrix_->OperatorDomainMap().NumMyElements(),
0, SubdomainCrsMatrix_->Comm()));
SubdomainMatrixReindexer_.reset(new EpetraExt::CrsMatrix_Reindex(*tempMap_));
DomainVectorReindexer_.reset(new EpetraExt::MultiVector_Reindex(*tempDomainMap_));
RangeVectorReindexer_.reset(new EpetraExt::MultiVector_Reindex(*tempRangeMap_));
ReindexedCrsMatrix_.reset(&((*SubdomainMatrixReindexer_)(*SubdomainCrsMatrix_)), false);
MatrixPtr = &*ReindexedCrsMatrix_;
Inverse_ = Teuchos::rcp( new T(&*ReindexedCrsMatrix_) );
} else {
Inverse_ = Teuchos::rcp( new T(&*SubdomainCrsMatrix_) );
}
#else
Inverse_ = Teuchos::rcp( new T(MatrixPtr) );
#endif
if (Inverse_ == Teuchos::null)
IFPACK_CHK_ERR(-5);
return(0);
}
//==============================================================================
template<typename T>
int Ifpack_AdditiveSchwarz<T>::SetParameters(Teuchos::ParameterList& List_in)
{
#ifdef HAVE_IFPACK_PARALLEL_SUBDOMAIN_SOLVERS
MpiRank_ = Matrix_->Comm().MyPID();
NumMpiProcs_ = Matrix_->Comm().NumProc();
NumMpiProcsPerSubdomain_ = List_in.get("subdomain: number-of-processors", 1);
NumSubdomains_ = NumMpiProcs_ / NumMpiProcsPerSubdomain_;
SubdomainId_ = MpiRank_ / NumMpiProcsPerSubdomain_;
if (NumSubdomains_ == 1) {
IsOverlapping_ = false;
}
#endif
// compute the condition number each time Compute() is invoked.
ComputeCondest_ = List_in.get("schwarz: compute condest", ComputeCondest_);
// combine mode
if( Teuchos::ParameterEntry *combineModeEntry = List_in.getEntryPtr("schwarz: combine mode") )
{
if( typeid(std::string) == combineModeEntry->getAny().type() )
{
std::string mode = List_in.get("schwarz: combine mode", "Add");
if (mode == "Add")
CombineMode_ = Add;
else if (mode == "Zero")
CombineMode_ = Zero;
else if (mode == "Insert")
CombineMode_ = Insert;
else if (mode == "InsertAdd")
CombineMode_ = InsertAdd;
else if (mode == "Average")
CombineMode_ = Average;
else if (mode == "AbsMax")
CombineMode_ = AbsMax;
else
{
TEUCHOS_TEST_FOR_EXCEPTION(
true,std::logic_error
,"Error, The (Epetra) combine mode of \""<<mode<<"\" is not valid! Only the values"
" \"Add\", \"Zero\", \"Insert\", \"InsertAdd\", \"Average\", and \"AbsMax\" are accepted!"
);
}
}
else if ( typeid(Epetra_CombineMode) == combineModeEntry->getAny().type() )
{
CombineMode_ = Teuchos::any_cast<Epetra_CombineMode>(combineModeEntry->getAny());
}
else
{
// Throw exception with good error message!
Teuchos::getParameter<std::string>(List_in,"schwarz: combine mode");
}
}
else
{
// Make the default be a std::string to be consistent with the valid parameters!
List_in.get("schwarz: combine mode","Zero");
}
// type of reordering
ReorderingType_ = List_in.get("schwarz: reordering type", ReorderingType_);
if (ReorderingType_ == "none")
UseReordering_ = false;
else
UseReordering_ = true;
// if true, filter singletons. NOTE: the filtered matrix can still have
// singletons! A simple example: upper triangular matrix, if I remove
// the lower node, I still get a matrix with a singleton! However, filter
// singletons should help for PDE problems with Dirichlet BCs.
FilterSingletons_ = List_in.get("schwarz: filter singletons", FilterSingletons_);
// This copy may be needed by Amesos or other preconditioners.
List_ = List_in;
return(0);
}
//==============================================================================
template<typename T>
int Ifpack_AdditiveSchwarz<T>::Initialize()
{
IsInitialized_ = false;
IsComputed_ = false; // values required
Condest_ = -1.0; // zero-out condest
if (Time_ == Teuchos::null)
Time_ = Teuchos::rcp( new Epetra_Time(Comm()) );
Time_->ResetStartTime();
// compute the overlapping matrix if necessary
if (IsOverlapping_) {
#ifdef HAVE_IFPACK_PARALLEL_SUBDOMAIN_SOLVERS
if (NumMpiProcsPerSubdomain_ > 1) {
OverlappingMatrix_ = Teuchos::rcp( new Ifpack_OverlappingRowMatrix(Matrix_, OverlapLevel_, SubdomainId_) );
} else {
OverlappingMatrix_ = Teuchos::rcp( new Ifpack_OverlappingRowMatrix(Matrix_, OverlapLevel_));
}
#else
# ifdef IFPACK_NODE_AWARE_CODE
int myNodeID;
try{ myNodeID = List_.get("ML node id",-1);}
catch(...){fprintf(stderr,"pid %d: no such entry (returned %d)\n",Comm().MyPID(),myNodeID);}
/*
cout << "pid " << Comm().MyPID()
<< ": calling Ifpack_OverlappingRowMatrix with myNodeID = "
<< myNodeID << ", OverlapLevel_ = " << OverlapLevel_ << endl;
*/
OverlappingMatrix_ = Teuchos::rcp( new Ifpack_OverlappingRowMatrix(Matrix_, OverlapLevel_, myNodeID) );
# else
OverlappingMatrix_ = Teuchos::rcp( new Ifpack_OverlappingRowMatrix(Matrix_, OverlapLevel_) );
# endif
#endif
if (OverlappingMatrix_ == Teuchos::null) {
IFPACK_CHK_ERR(-5);
}
}
# ifdef IFPACK_NODE_AWARE_CODE
/*
sleep(1);
Comm().Barrier();
*/
# endif
IFPACK_CHK_ERR(Setup());
# ifdef IFPACK_NODE_AWARE_CODE
/*
sleep(1);
Comm().Barrier();
*/
#endif
if (Inverse_ == Teuchos::null)
IFPACK_CHK_ERR(-5);
if (LocalizedMatrix_ == Teuchos::null)
IFPACK_CHK_ERR(-5);
IFPACK_CHK_ERR(Inverse_->SetUseTranspose(UseTranspose()));
IFPACK_CHK_ERR(Inverse_->SetParameters(List_));
IFPACK_CHK_ERR(Inverse_->Initialize());
// Label is for Aztec-like solvers
Label_ = "Ifpack_AdditiveSchwarz, ";
if (UseTranspose())
Label_ += ", transp";
Label_ += ", ov = " + Ifpack_toString(OverlapLevel_)
+ ", local solver = \n\t\t***** `" + std::string(Inverse_->Label()) + "'";
IsInitialized_ = true;
++NumInitialize_;
InitializeTime_ += Time_->ElapsedTime();
#ifdef IFPACK_FLOPCOUNTERS
// count flops by summing up all the InitializeFlops() in each
// Inverse. Each Inverse() can only give its flops -- it acts on one
// process only
double partial = Inverse_->InitializeFlops();
double total;
Comm().SumAll(&partial, &total, 1);
InitializeFlops_ += total;
#endif
return(0);
}
//==============================================================================
template<typename T>
int Ifpack_AdditiveSchwarz<T>::Compute()
{
if (IsInitialized() == false)
IFPACK_CHK_ERR(Initialize());
Time_->ResetStartTime();
IsComputed_ = false;
Condest_ = -1.0;
IFPACK_CHK_ERR(Inverse_->Compute());
IsComputed_ = true; // need this here for Condest(Ifpack_Cheap)
++NumCompute_;
ComputeTime_ += Time_->ElapsedTime();
#ifdef IFPACK_FLOPCOUNTERS
// sum up flops
double partial = Inverse_->ComputeFlops();
double total;
Comm().SumAll(&partial, &total, 1);
ComputeFlops_ += total;
#endif
// reset the Label
std::string R = "";
if (UseReordering_)
R = ReorderingType_ + " reord, ";
if (ComputeCondest_)
Condest(Ifpack_Cheap);
// add Condest() to label
Label_ = "Ifpack_AdditiveSchwarz, ov = " + Ifpack_toString(OverlapLevel_)
+ ", local solver = \n\t\t***** `" + std::string(Inverse_->Label()) + "'"
+ "\n\t\t***** " + R + "Condition number estimate = "
+ Ifpack_toString(Condest_);
return(0);
}
//==============================================================================
template<typename T>
int Ifpack_AdditiveSchwarz<T>::SetUseTranspose(bool UseTranspose_in)
{
// store the flag -- it will be set in Initialize() if Inverse_ does not
// exist.
UseTranspose_ = UseTranspose_in;
// If Inverse_ exists, pass it right now.
if (Inverse_!=Teuchos::null)
IFPACK_CHK_ERR(Inverse_->SetUseTranspose(UseTranspose_in));
return(0);
}
//==============================================================================
template<typename T>
int Ifpack_AdditiveSchwarz<T>::
Apply(const Epetra_MultiVector& X, Epetra_MultiVector& Y) const
{
IFPACK_CHK_ERR(Matrix_->Apply(X,Y));
return(0);
}
//==============================================================================
template<typename T>
double Ifpack_AdditiveSchwarz<T>::NormInf() const
{
return(-1.0);
}
//==============================================================================
template<typename T>
const char * Ifpack_AdditiveSchwarz<T>::Label() const
{
return(Label_.c_str());
}
//==============================================================================
template<typename T>
bool Ifpack_AdditiveSchwarz<T>::UseTranspose() const
{
return(UseTranspose_);
}
//==============================================================================
template<typename T>
bool Ifpack_AdditiveSchwarz<T>::HasNormInf() const
{
return(false);
}
//==============================================================================
template<typename T>
const Epetra_Comm & Ifpack_AdditiveSchwarz<T>::Comm() const
{
return(Matrix_->Comm());
}
//==============================================================================
template<typename T>
const Epetra_Map & Ifpack_AdditiveSchwarz<T>::OperatorDomainMap() const
{
return(Matrix_->OperatorDomainMap());
}
//==============================================================================
template<typename T>
const Epetra_Map & Ifpack_AdditiveSchwarz<T>::OperatorRangeMap() const
{
return(Matrix_->OperatorRangeMap());
}
//==============================================================================
template<typename T>
int Ifpack_AdditiveSchwarz<T>::
ApplyInverse(const Epetra_MultiVector& X, Epetra_MultiVector& Y) const
{
// compute the preconditioner is not done by the user
if (!IsComputed())
IFPACK_CHK_ERR(-3);
int NumVectors = X.NumVectors();
if (NumVectors != Y.NumVectors())
IFPACK_CHK_ERR(-2); // wrong input
Time_->ResetStartTime();
Teuchos::RefCountPtr<Epetra_MultiVector> OverlappingX;
Teuchos::RefCountPtr<Epetra_MultiVector> OverlappingY;
Teuchos::RefCountPtr<Epetra_MultiVector> Xtmp;
// for flop count, see bottom of this function
#ifdef IFPACK_FLOPCOUNTERS
double pre_partial = Inverse_->ApplyInverseFlops();
double pre_total;
Comm().SumAll(&pre_partial, &pre_total, 1);
#endif
// process overlap, may need to create vectors and import data
if (IsOverlapping()) {
#ifdef HAVE_IFPACK_PARALLEL_SUBDOMAIN_SOLVERS
if (OverlappingX == Teuchos::null) {
OverlappingX = Teuchos::rcp( new Epetra_MultiVector(OverlappingMatrix_->RowMatrixRowMap(), X.NumVectors()) );
if (OverlappingX == Teuchos::null) IFPACK_CHK_ERR(-5);
} else assert(OverlappingX->NumVectors() == X.NumVectors());
if (OverlappingY == Teuchos::null) {
OverlappingY = Teuchos::rcp( new Epetra_MultiVector(OverlappingMatrix_->RowMatrixRowMap(), Y.NumVectors()) );
if (OverlappingY == Teuchos::null) IFPACK_CHK_ERR(-5);
} else assert(OverlappingY->NumVectors() == Y.NumVectors());
#else
# ifdef IFPACK_NODE_AWARE_CODE
if (OverlappingX == Teuchos::null) {
OverlappingX = Teuchos::rcp( new Epetra_MultiVector(OverlappingMatrix_->RowMatrixRowMap(),
X.NumVectors()) );
if (OverlappingX == Teuchos::null) IFPACK_CHK_ERR(-5);
} else assert(OverlappingX->NumVectors() == X.NumVectors());
if (OverlappingY == Teuchos::null) {
OverlappingY = Teuchos::rcp( new Epetra_MultiVector(OverlappingMatrix_->RowMatrixRowMap(),
Y.NumVectors()) );
if (OverlappingY == Teuchos::null) IFPACK_CHK_ERR(-5);
} else assert(OverlappingY->NumVectors() == Y.NumVectors());
#else
OverlappingX = Teuchos::rcp( new Epetra_MultiVector(OverlappingMatrix_->RowMatrixRowMap(),
X.NumVectors()) );
OverlappingY = Teuchos::rcp( new Epetra_MultiVector(OverlappingMatrix_->RowMatrixRowMap(),
Y.NumVectors()) );
if (OverlappingY == Teuchos::null) IFPACK_CHK_ERR(-5);
# endif
#endif
OverlappingY->PutScalar(0.0);
OverlappingX->PutScalar(0.0);
IFPACK_CHK_ERR(OverlappingMatrix_->ImportMultiVector(X,*OverlappingX,Insert));
// FIXME: this will not work with overlapping and non-zero starting
// solutions. The same for other cases below.
// IFPACK_CHK_ERR(OverlappingMatrix_->ImportMultiVector(Y,*OverlappingY,Insert));
}
else {
Xtmp = Teuchos::rcp( new Epetra_MultiVector(X) );
OverlappingX = Xtmp;
OverlappingY = Teuchos::rcp( &Y, false );
}
if (FilterSingletons_) {
// process singleton filter
Epetra_MultiVector ReducedX(SingletonFilter_->Map(),NumVectors);
Epetra_MultiVector ReducedY(SingletonFilter_->Map(),NumVectors);
IFPACK_CHK_ERR(SingletonFilter_->SolveSingletons(*OverlappingX,*OverlappingY));
IFPACK_CHK_ERR(SingletonFilter_->CreateReducedRHS(*OverlappingY,*OverlappingX,ReducedX));
// process reordering
if (!UseReordering_) {
IFPACK_CHK_ERR(Inverse_->ApplyInverse(ReducedX,ReducedY));
}
else {
Epetra_MultiVector ReorderedX(ReducedX);
Epetra_MultiVector ReorderedY(ReducedY);
IFPACK_CHK_ERR(Reordering_->P(ReducedX,ReorderedX));
IFPACK_CHK_ERR(Inverse_->ApplyInverse(ReorderedX,ReorderedY));
IFPACK_CHK_ERR(Reordering_->Pinv(ReorderedY,ReducedY));
}
// finish up with singletons
IFPACK_CHK_ERR(SingletonFilter_->UpdateLHS(ReducedY,*OverlappingY));
}
else {
// process reordering
if (!UseReordering_) {
#ifdef HAVE_IFPACK_PARALLEL_SUBDOMAIN_SOLVERS
if (NumMpiProcsPerSubdomain_ > 1) {
tempX_.reset(&((*RangeVectorReindexer_)(*OverlappingX)), false);
tempY_.reset(&((*DomainVectorReindexer_)(*OverlappingY)), false);
IFPACK_CHK_ERR(Inverse_->ApplyInverse(*tempX_,*tempY_));
} else {
IFPACK_CHK_ERR(Inverse_->ApplyInverse(*OverlappingX, *OverlappingY));
}
#else
IFPACK_CHK_ERR(Inverse_->ApplyInverse(*OverlappingX,*OverlappingY));
#endif
}
else {
Epetra_MultiVector ReorderedX(*OverlappingX);
Epetra_MultiVector ReorderedY(*OverlappingY);
IFPACK_CHK_ERR(Reordering_->P(*OverlappingX,ReorderedX));
IFPACK_CHK_ERR(Inverse_->ApplyInverse(ReorderedX,ReorderedY));
IFPACK_CHK_ERR(Reordering_->Pinv(ReorderedY,*OverlappingY));
}
}
if (IsOverlapping()) {
IFPACK_CHK_ERR(OverlappingMatrix_->ExportMultiVector(*OverlappingY,Y,
CombineMode_));
}
#ifdef IFPACK_FLOPCOUNTERS
// add flops. Note the we only have to add the newly counted
// flops -- and each Inverse returns the cumulative sum
double partial = Inverse_->ApplyInverseFlops();
double total;
Comm().SumAll(&partial, &total, 1);
ApplyInverseFlops_ += total - pre_total;
#endif
// FIXME: right now I am skipping the overlap and singletons
++NumApplyInverse_;
ApplyInverseTime_ += Time_->ElapsedTime();
return(0);
}
//==============================================================================
template<typename T>
std::ostream& Ifpack_AdditiveSchwarz<T>::
Print(std::ostream& os) const
{
using std::endl;
#ifdef IFPACK_FLOPCOUNTERS
double IF = InitializeFlops();
double CF = ComputeFlops();
double AF = ApplyInverseFlops();
double IFT = 0.0, CFT = 0.0, AFT = 0.0;
if (InitializeTime() != 0.0)
IFT = IF / InitializeTime();
if (ComputeTime() != 0.0)
CFT = CF / ComputeTime();
if (ApplyInverseTime() != 0.0)
AFT = AF / ApplyInverseTime();
#endif
if (Matrix().Comm().MyPID())
return(os);
os << endl;
os << "================================================================================" << endl;
os << "Ifpack_AdditiveSchwarz, overlap level = " << OverlapLevel_ << endl;
if (CombineMode_ == Insert)
os << "Combine mode = Insert" << endl;
else if (CombineMode_ == Add)
os << "Combine mode = Add" << endl;
else if (CombineMode_ == Zero)
os << "Combine mode = Zero" << endl;
else if (CombineMode_ == Average)
os << "Combine mode = Average" << endl;
else if (CombineMode_ == AbsMax)
os << "Combine mode = AbsMax" << endl;
os << "Condition number estimate = " << Condest_ << endl;
os << "Global number of rows = " << Matrix_->NumGlobalRows64() << endl;
#ifdef HAVE_IFPACK_PARALLEL_SUBDOMAIN_SOLVERS
os << endl;
os << "================================================================================" << endl;
os << "Subcommunicator stats" << endl;
os << "Number of MPI processes in simulation: " << NumMpiProcs_ << endl;
os << "Number of subdomains: " << NumSubdomains_ << endl;
os << "Number of MPI processes per subdomain: " << NumMpiProcsPerSubdomain_ << endl;
#endif
os << endl;
os << "Phase # calls Total Time (s) Total MFlops MFlops/s" << endl;
os << "----- ------- -------------- ------------ --------" << endl;
os << "Initialize() " << std::setw(5) << NumInitialize()
<< " " << std::setw(15) << InitializeTime()
#ifdef IFPACK_FLOPCOUNTERS
<< " " << std::setw(15) << 1.0e-6 * IF
<< " " << std::setw(15) << 1.0e-6 * IFT
#endif
<< endl;
os << "Compute() " << std::setw(5) << NumCompute()
<< " " << std::setw(15) << ComputeTime()
#ifdef IFPACK_FLOPCOUNTERS
<< " " << std::setw(15) << 1.0e-6 * CF
<< " " << std::setw(15) << 1.0e-6 * CFT
#endif
<< endl;
os << "ApplyInverse() " << std::setw(5) << NumApplyInverse()
<< " " << std::setw(15) << ApplyInverseTime()
#ifdef IFPACK_FLOPCOUNTERS
<< " " << std::setw(15) << 1.0e-6 * AF
<< " " << std::setw(15) << 1.0e-6 * AFT
#endif
<< endl;
os << "================================================================================" << endl;
os << endl;
return(os);
}
#include "Ifpack_Condest.h"
//==============================================================================
template<typename T>
double Ifpack_AdditiveSchwarz<T>::
Condest(const Ifpack_CondestType CT, const int MaxIters,
const double Tol, Epetra_RowMatrix* Matrix_in)
{
if (!IsComputed()) // cannot compute right now
return(-1.0);
Condest_ = Ifpack_Condest(*this, CT, MaxIters, Tol, Matrix_in);
return(Condest_);
}
#endif // IFPACK_ADDITIVESCHWARZ_H
|