Nektar++
 All Classes Namespaces Files Functions Variables Typedefs Enumerations Enumerator Friends Macros Pages
GlobalLinSysStaticCond.cpp
Go to the documentation of this file.
1 ///////////////////////////////////////////////////////////////////////////////
2 //
3 // File: GlobalLinSysStaticCond.cpp
4 //
5 // For more information, please see: http://www.nektar.info
6 //
7 // The MIT License
8 //
9 // Copyright (c) 2006 Division of Applied Mathematics, Brown University (USA),
10 // Department of Aeronautics, Imperial College London (UK), and Scientific
11 // Computing and Imaging Institute, University of Utah (USA).
12 //
13 // License for the specific language governing rights and limitations under
14 // Permission is hereby granted, free of charge, to any person obtaining a
15 // copy of this software and associated documentation files (the "Software"),
16 // to deal in the Software without restriction, including without limitation
17 // the rights to use, copy, modify, merge, publish, distribute, sublicense,
18 // and/or sell copies of the Software, and to permit persons to whom the
19 // Software is furnished to do so, subject to the following conditions:
20 //
21 // The above copyright notice and this permission notice shall be included
22 // in all copies or substantial portions of the Software.
23 //
24 // THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS
25 // OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
26 // FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL
27 // THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
28 // LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
29 // FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER
30 // DEALINGS IN THE SOFTWARE.
31 //
32 // Description: Implementation to linear solver using single-
33 // or multi-level static condensation
34 //
35 ///////////////////////////////////////////////////////////////////////////////
36 
43 
44 using namespace std;
45 
46 namespace Nektar
47 {
48  namespace MultiRegions
49  {
50  /**
51  * @class GlobalLinSysStaticCond
52  *
53  * Solves a linear system using single- or multi-level static
54  * condensation.
55  */
56 
57  /**
58  * For a matrix system of the form @f[
59  * \left[ \begin{array}{cc}
60  * \boldsymbol{A} & \boldsymbol{B}\\
61  * \boldsymbol{C} & \boldsymbol{D}
62  * \end{array} \right]
63  * \left[ \begin{array}{c} \boldsymbol{x_1}\\ \boldsymbol{x_2}
64  * \end{array}\right]
65  * = \left[ \begin{array}{c} \boldsymbol{y_1}\\ \boldsymbol{y_2}
66  * \end{array}\right],
67  * @f]
68  * where @f$\boldsymbol{D}@f$ and
69  * @f$(\boldsymbol{A-BD^{-1}C})@f$ are invertible, store and assemble
70  * a static condensation system, according to a given local to global
71  * mapping. #m_linSys is constructed by AssembleSchurComplement().
72  * @param mKey Associated matrix key.
73  * @param pLocMatSys LocalMatrixSystem
74  * @param locToGloMap Local to global mapping.
75  */
76  GlobalLinSysStaticCond::GlobalLinSysStaticCond(
77  const GlobalLinSysKey &pKey,
78  const boost::weak_ptr<ExpList> &pExpList,
79  const boost::shared_ptr<AssemblyMap> &pLocToGloMap)
80  : GlobalLinSys(pKey, pExpList, pLocToGloMap),
81  m_locToGloMap (pLocToGloMap)
82  {
83  }
84 
86  {
87  // Allocate memory for top-level structure
89 
90  // Construct this level
92  }
93 
94  /**
95  *
96  */
98  {
99 
100  }
101 
102 
103  /**
104  *
105  */
109  const AssemblyMapSharedPtr &pLocToGloMap,
110  const Array<OneD, const NekDouble> &dirForcing)
111  {
112  bool dirForcCalculated = (bool) dirForcing.num_elements();
113  bool atLastLevel = pLocToGloMap->AtLastLevel();
114  int scLevel = pLocToGloMap->GetStaticCondLevel();
115 
116  int nGlobDofs = pLocToGloMap->GetNumGlobalCoeffs();
117  int nGlobBndDofs = pLocToGloMap->GetNumGlobalBndCoeffs();
118  int nDirBndDofs = pLocToGloMap->GetNumGlobalDirBndCoeffs();
119  int nGlobHomBndDofs = nGlobBndDofs - nDirBndDofs;
120  int nLocBndDofs = pLocToGloMap->GetNumLocalBndCoeffs();
121  int nIntDofs = pLocToGloMap->GetNumGlobalCoeffs()
122  - nGlobBndDofs;
123 
124  Array<OneD, NekDouble> F = m_wsp + 2*nLocBndDofs;
126  if(nDirBndDofs && dirForcCalculated)
127  {
128  Vmath::Vsub(nGlobDofs,in.get(),1,dirForcing.get(),1,F.get(),1);
129  }
130  else
131  {
132  Vmath::Vcopy(nGlobDofs,in.get(),1,F.get(),1);
133  }
134 
135  NekVector<NekDouble> F_HomBnd(nGlobHomBndDofs,tmp=F+nDirBndDofs,
136  eWrapper);
137  NekVector<NekDouble> F_GlobBnd(nGlobBndDofs,F,eWrapper);
138  NekVector<NekDouble> F_LocBnd(nLocBndDofs,0.0);
139  NekVector<NekDouble> F_Int(nIntDofs,tmp=F+nGlobBndDofs,eWrapper);
140 
141  NekVector<NekDouble> V_GlobBnd(nGlobBndDofs,out,eWrapper);
142  NekVector<NekDouble> V_GlobHomBnd(nGlobHomBndDofs,
143  tmp=out+nDirBndDofs,
144  eWrapper);
145  NekVector<NekDouble> V_Int(nIntDofs,tmp=out+nGlobBndDofs,eWrapper);
146  NekVector<NekDouble> V_LocBnd(nLocBndDofs,m_wsp,eWrapper);
147 
148  NekVector<NekDouble> V_GlobHomBndTmp(nGlobHomBndDofs,0.0);
149 
150  // set up normalisation factor for right hand side on first SC level
151  DNekScalBlkMatSharedPtr sc = v_PreSolve(scLevel, F_GlobBnd);
152 
153  if(nGlobHomBndDofs)
154  {
155  // construct boundary forcing
156  if( nIntDofs && ((!dirForcCalculated) && (atLastLevel)) )
157  {
158  DNekScalBlkMat &BinvD = *m_BinvD;
159  DNekScalBlkMat &SchurCompl = *sc;
160 
161  // include dirichlet boundary forcing
162  pLocToGloMap->GlobalToLocalBnd(V_GlobBnd,V_LocBnd);
163  V_LocBnd = BinvD*F_Int + SchurCompl*V_LocBnd;
164  }
165  else if((!dirForcCalculated) && (atLastLevel))
166  {
167  // include dirichlet boundary forcing
168  DNekScalBlkMat &SchurCompl = *sc;
169  pLocToGloMap->GlobalToLocalBnd(V_GlobBnd,V_LocBnd);
170  V_LocBnd = SchurCompl*V_LocBnd;
171  }
172  else
173  {
174  DNekScalBlkMat &BinvD = *m_BinvD;
175  V_LocBnd = BinvD*F_Int;
176  }
177 
178  pLocToGloMap->AssembleBnd(V_LocBnd,V_GlobHomBndTmp,
179  nDirBndDofs);
180  F_HomBnd = F_HomBnd - V_GlobHomBndTmp;
181 
182  // Transform from original basis to low energy
183  v_BasisTransform(F, nDirBndDofs);
184 
185  // For parallel multi-level static condensation some
186  // processors may have different levels to others. This
187  // routine receives contributions to partition vertices from
188  // those lower levels, whilst not sending anything to the
189  // other partitions, and includes them in the modified right
190  // hand side vector.
191  int lcLevel = pLocToGloMap->GetLowestStaticCondLevel();
192  if(atLastLevel && scLevel < lcLevel)
193  {
194  // If this level is not the lowest level across all
195  // processes, we must do dummy communication for the
196  // remaining levels
197  Array<OneD, NekDouble> tmp(nGlobBndDofs);
198  for (int i = scLevel; i < lcLevel; ++i)
199  {
200  Vmath::Fill(nGlobBndDofs, 0.0, tmp, 1);
201  pLocToGloMap->UniversalAssembleBnd(tmp);
202  Vmath::Vcopy(nGlobHomBndDofs,
203  tmp.get()+nDirBndDofs, 1,
204  V_GlobHomBndTmp.GetPtr().get(), 1);
205  F_HomBnd = F_HomBnd - V_GlobHomBndTmp;
206  }
207  }
208 
209  // solve boundary system
210  if(atLastLevel)
211  {
212  Array<OneD, NekDouble> pert(nGlobBndDofs,0.0);
213  NekVector<NekDouble> Pert(nGlobBndDofs,pert,eWrapper);
214 
215  // Solve for difference from initial solution given inout;
217  nGlobBndDofs, F, pert, pLocToGloMap, nDirBndDofs);
218 
219  // Transform back to original basis
220  v_BasisInvTransform(pert);
221 
222  // Add back initial conditions onto difference
223  Vmath::Vadd(nGlobHomBndDofs,&out[nDirBndDofs],1,
224  &pert[nDirBndDofs],1,&out[nDirBndDofs],1);
225  }
226  else
227  {
228  m_recursiveSchurCompl->Solve(F,
229  V_GlobBnd.GetPtr(),
230  pLocToGloMap->GetNextLevelLocalToGlobalMap());
231  }
232  }
233 
234  // solve interior system
235  if(nIntDofs)
236  {
237  DNekScalBlkMat &invD = *m_invD;
238 
239  if(nGlobHomBndDofs || nDirBndDofs)
240  {
241  DNekScalBlkMat &C = *m_C;
242 
243  if(dirForcCalculated && nDirBndDofs)
244  {
245  pLocToGloMap->GlobalToLocalBnd(V_GlobHomBnd,V_LocBnd,
246  nDirBndDofs);
247  }
248  else
249  {
250  pLocToGloMap->GlobalToLocalBnd(V_GlobBnd,V_LocBnd);
251  }
252  F_Int = F_Int - C*V_LocBnd;
253  }
254 
255  V_Int = invD*F_Int;
256  }
257  }
258 
259 
260  /**
261  * If at the last level of recursion (or the only level in the case of
262  * single-level static condensation), assemble the Schur complement.
263  * For other levels, in the case of multi-level static condensation,
264  * the next level of the condensed system is computed.
265  * @param pLocToGloMap Local to global mapping.
266  */
268  const boost::shared_ptr<AssemblyMap>& pLocToGloMap)
269  {
270  int nLocalBnd = m_locToGloMap->GetNumLocalBndCoeffs();
271  int nGlobal = m_locToGloMap->GetNumGlobalCoeffs();
272  m_wsp = Array<OneD, NekDouble>(2*nLocalBnd + nGlobal, 0.0);
273 
274  if (pLocToGloMap->AtLastLevel())
275  {
276  v_AssembleSchurComplement(pLocToGloMap);
277  }
278  else
279  {
281  pLocToGloMap->GetNextLevelLocalToGlobalMap());
282  }
283  }
284 
286  {
287  return m_schurCompl->GetNumberOfBlockRows();
288  }
289 
290  /**
291  * For the first level in multi-level static condensation, or the only
292  * level in the case of single-level static condensation, allocate the
293  * condensed matrices and populate them with the local matrices
294  * retrieved from the expansion list.
295  * @param
296  */
298  const boost::shared_ptr<AssemblyMap>& pLocToGloMap)
299  {
300  int n;
301  int n_exp = m_expList.lock()->GetNumElmts();
302 
303  const Array<OneD,const unsigned int>& nbdry_size
304  = pLocToGloMap->GetNumLocalBndCoeffsPerPatch();
305  const Array<OneD,const unsigned int>& nint_size
306  = pLocToGloMap->GetNumLocalIntCoeffsPerPatch();
307 
308  // Setup Block Matrix systems
309  MatrixStorage blkmatStorage = eDIAGONAL;
311  ::AllocateSharedPtr(nbdry_size, nbdry_size, blkmatStorage);
313  ::AllocateSharedPtr(nbdry_size, nint_size , blkmatStorage);
315  ::AllocateSharedPtr(nint_size , nbdry_size, blkmatStorage);
317  ::AllocateSharedPtr(nint_size , nint_size , blkmatStorage);
318 
319  for(n = 0; n < n_exp; ++n)
320  {
321  if (m_linSysKey.GetMatrixType() ==
323  {
324  DNekScalMatSharedPtr loc_mat
326  m_expList.lock()->GetOffset_Elmt_Id(n));
327  m_schurCompl->SetBlock(n,n,loc_mat);
328  }
329  else
330  {
331  DNekScalBlkMatSharedPtr loc_schur
333  m_expList.lock()->GetOffset_Elmt_Id(n));
335  m_schurCompl->SetBlock(n, n, t = loc_schur->GetBlock(0,0));
336  m_BinvD ->SetBlock(n, n, t = loc_schur->GetBlock(0,1));
337  m_C ->SetBlock(n, n, t = loc_schur->GetBlock(1,0));
338  m_invD ->SetBlock(n, n, t = loc_schur->GetBlock(1,1));
339  }
340  }
341  }
342 
343  /**
344  *
345  */
347  const AssemblyMapSharedPtr& pLocToGloMap)
348  {
349  int i,j,n,cnt;
350  DNekScalBlkMatSharedPtr blkMatrices[4];
351 
352  // Create temporary matrices within an inner-local scope to ensure
353  // any references to the intermediate storage is lost before
354  // the recursive step, rather than at the end of the routine.
355  // This allows the schur complement matrix from this level to be
356  // disposed of in the next level after use without being retained
357  // due to lingering shared pointers.
358  {
359 
360  const Array<OneD,const unsigned int>& nBndDofsPerPatch
361  = pLocToGloMap->GetNumLocalBndCoeffsPerPatch();
362  const Array<OneD,const unsigned int>& nIntDofsPerPatch
363  = pLocToGloMap->GetNumLocalIntCoeffsPerPatch();
364 
365  // STEP 1:
366  // Based upon the schur complement of the the current level we
367  // will substructure this matrix in the form
368  // -- --
369  // | A B |
370  // | C D |
371  // -- --
372  // All matrices A,B,C and D are (diagonal) blockmatrices.
373  // However, as a start we will use an array of DNekMatrices as
374  // it is too hard to change the individual entries of a
375  // DNekScalBlkMatSharedPtr.
376 
377  // In addition, we will also try to ensure that the memory of
378  // the blockmatrices will be contiguous. This will probably
379  // enhance the efficiency
380  // - Calculate the total number of entries in the blockmatrices
381  int nPatches = pLocToGloMap->GetNumPatches();
382  int nEntriesA = 0; int nEntriesB = 0;
383  int nEntriesC = 0; int nEntriesD = 0;
384 
385  for(i = 0; i < nPatches; i++)
386  {
387  nEntriesA += nBndDofsPerPatch[i]*nBndDofsPerPatch[i];
388  nEntriesB += nBndDofsPerPatch[i]*nIntDofsPerPatch[i];
389  nEntriesC += nIntDofsPerPatch[i]*nBndDofsPerPatch[i];
390  nEntriesD += nIntDofsPerPatch[i]*nIntDofsPerPatch[i];
391  }
392 
393  // Now create the DNekMatrices and link them to the memory
394  // allocated above
395  Array<OneD, DNekMatSharedPtr> substructuredMat[4]
396  = {Array<OneD, DNekMatSharedPtr>(nPatches), //Matrix A
397  Array<OneD, DNekMatSharedPtr>(nPatches), //Matrix B
398  Array<OneD, DNekMatSharedPtr>(nPatches), //Matrix C
399  Array<OneD, DNekMatSharedPtr>(nPatches)}; //Matrix D
400 
401  // Initialise storage for the matrices. We do this separately
402  // for each matrix so the matrices may be independently
403  // deallocated when no longer required.
404  Array<OneD, NekDouble> storageA(nEntriesA,0.0);
405  Array<OneD, NekDouble> storageB(nEntriesB,0.0);
406  Array<OneD, NekDouble> storageC(nEntriesC,0.0);
407  Array<OneD, NekDouble> storageD(nEntriesD,0.0);
408 
409  Array<OneD, NekDouble> tmparray;
410  PointerWrapper wType = eWrapper;
411  int cntA = 0;
412  int cntB = 0;
413  int cntC = 0;
414  int cntD = 0;
415 
416  for(i = 0; i < nPatches; i++)
417  {
418  // Matrix A
419  tmparray = storageA+cntA;
420  substructuredMat[0][i] = MemoryManager<DNekMat>
421  ::AllocateSharedPtr(nBndDofsPerPatch[i],
422  nBndDofsPerPatch[i],
423  tmparray, wType);
424  // Matrix B
425  tmparray = storageB+cntB;
426  substructuredMat[1][i] = MemoryManager<DNekMat>
427  ::AllocateSharedPtr(nBndDofsPerPatch[i],
428  nIntDofsPerPatch[i],
429  tmparray, wType);
430  // Matrix C
431  tmparray = storageC+cntC;
432  substructuredMat[2][i] = MemoryManager<DNekMat>
433  ::AllocateSharedPtr(nIntDofsPerPatch[i],
434  nBndDofsPerPatch[i],
435  tmparray, wType);
436  // Matrix D
437  tmparray = storageD+cntD;
438  substructuredMat[3][i] = MemoryManager<DNekMat>
439  ::AllocateSharedPtr(nIntDofsPerPatch[i],
440  nIntDofsPerPatch[i],
441  tmparray, wType);
442 
443  cntA += nBndDofsPerPatch[i] * nBndDofsPerPatch[i];
444  cntB += nBndDofsPerPatch[i] * nIntDofsPerPatch[i];
445  cntC += nIntDofsPerPatch[i] * nBndDofsPerPatch[i];
446  cntD += nIntDofsPerPatch[i] * nIntDofsPerPatch[i];
447  }
448 
449  // Then, project SchurComplement onto
450  // the substructured matrices of the next level
452  DNekScalMatSharedPtr schurComplSubMat;
453  int schurComplSubMatnRows;
454  Array<OneD, const int> patchId, dofId;
457  NekDouble scale;
458 
459  for(n = cnt = 0; n < SchurCompl->GetNumberOfBlockRows(); ++n)
460  {
461  schurComplSubMat = SchurCompl->GetBlock(n,n);
462  schurComplSubMatnRows = schurComplSubMat->GetRows();
463 
464  scale = SchurCompl->GetBlock(n,n)->Scale();
465  Array<OneD, NekDouble> schurSubMat
466  = SchurCompl->GetBlock(n,n)->GetOwnedMatrix()->GetPtr();
467 
468  patchId = pLocToGloMap->GetPatchMapFromPrevLevel()
469  ->GetPatchId() + cnt;
470  dofId = pLocToGloMap->GetPatchMapFromPrevLevel()
471  ->GetDofId() + cnt;
472  isBndDof = pLocToGloMap->GetPatchMapFromPrevLevel()
473  ->IsBndDof() + cnt;
474  sign = pLocToGloMap->GetPatchMapFromPrevLevel()
475  ->GetSign() + cnt;
476 
477  // Set up Matrix;
478  for(i = 0; i < schurComplSubMatnRows; ++i)
479  {
480  int pId = patchId[i];
481  Array<OneD, NekDouble> subMat0
482  = substructuredMat[0][pId]->GetPtr();
483  Array<OneD, NekDouble> subMat1
484  = substructuredMat[1][patchId[i]]->GetPtr();
485  Array<OneD, NekDouble> subMat2
486  = substructuredMat[2][patchId[i]]->GetPtr();
487  Array<OneD, NekDouble> subMat3
488  = substructuredMat[3][patchId[i]]->GetPtr();
489  int subMat0rows = substructuredMat[0][pId]->GetRows();
490  int subMat1rows = substructuredMat[1][pId]->GetRows();
491  int subMat2rows = substructuredMat[2][pId]->GetRows();
492  int subMat3rows = substructuredMat[3][pId]->GetRows();
493 
494  if(isBndDof[i])
495  {
496  for(j = 0; j < schurComplSubMatnRows; ++j)
497  {
498  ASSERTL0(patchId[i]==patchId[j],
499  "These values should be equal");
500 
501  if(isBndDof[j])
502  {
503  subMat0[dofId[i]+dofId[j]*subMat0rows] +=
504  sign[i]*sign[j]*(
505  scale*schurSubMat[
506  i+j*schurComplSubMatnRows]);
507  }
508  else
509  {
510  subMat1[dofId[i]+dofId[j]*subMat1rows] +=
511  sign[i]*sign[j]*(
512  scale*schurSubMat[
513  i+j*schurComplSubMatnRows]);
514  }
515  }
516  }
517  else
518  {
519  for(j = 0; j < schurComplSubMatnRows; ++j)
520  {
521  ASSERTL0(patchId[i]==patchId[j],
522  "These values should be equal");
523 
524  if(isBndDof[j])
525  {
526  subMat2[dofId[i]+dofId[j]*subMat2rows] +=
527  sign[i]*sign[j]*(
528  scale*schurSubMat[
529  i+j*schurComplSubMatnRows]);
530  }
531  else
532  {
533  subMat3[dofId[i]+dofId[j]*subMat3rows] +=
534  sign[i]*sign[j]*(
535  scale*schurSubMat[
536  i+j*schurComplSubMatnRows]);
537  }
538  }
539  }
540  }
541  cnt += schurComplSubMatnRows;
542  }
543 
544  // STEP 2: condense the system
545  // This can be done elementally (i.e. patch per patch)
546  for(i = 0; i < nPatches; i++)
547  {
548  if(nIntDofsPerPatch[i])
549  {
550  Array<OneD, NekDouble> subMat0
551  = substructuredMat[0][i]->GetPtr();
552  Array<OneD, NekDouble> subMat1
553  = substructuredMat[1][i]->GetPtr();
554  Array<OneD, NekDouble> subMat2
555  = substructuredMat[2][i]->GetPtr();
556  int subMat0rows = substructuredMat[0][i]->GetRows();
557  int subMat1rows = substructuredMat[1][i]->GetRows();
558  int subMat2rows = substructuredMat[2][i]->GetRows();
559  int subMat2cols = substructuredMat[2][i]->GetColumns();
560 
561  // 1. D -> InvD
562  substructuredMat[3][i]->Invert();
563  // 2. B -> BInvD
564  (*substructuredMat[1][i]) = (*substructuredMat[1][i])*
565  (*substructuredMat[3][i]);
566  // 3. A -> A - BInvD*C (= schurcomplement)
567  // (*substructuredMat[0][i]) =(*substructuredMat[0][i])-
568  // (*substructuredMat[1][i])*(*substructuredMat[2][i]);
569  // Note: faster to use blas directly
570  Blas::Dgemm('N','N', subMat1rows, subMat2cols,
571  subMat2rows, -1.0, &subMat1[0], subMat1rows,
572  &subMat2[0], subMat2rows, 1.0, &subMat0[0],
573  subMat0rows);
574  }
575  }
576 
577  // STEP 3: fill the block matrices. However, do note that we
578  // first have to convert them to a DNekScalMat in order to be
579  // compatible with the first level of static condensation
580  const Array<OneD,const unsigned int>& nbdry_size
581  = pLocToGloMap->GetNumLocalBndCoeffsPerPatch();
582  const Array<OneD,const unsigned int>& nint_size
583  = pLocToGloMap->GetNumLocalIntCoeffsPerPatch();
584  MatrixStorage blkmatStorage = eDIAGONAL;
585 
586  blkMatrices[0] = MemoryManager<DNekScalBlkMat>
587  ::AllocateSharedPtr(nbdry_size, nbdry_size, blkmatStorage);
588  blkMatrices[1] = MemoryManager<DNekScalBlkMat>
589  ::AllocateSharedPtr(nbdry_size, nint_size , blkmatStorage);
590  blkMatrices[2] = MemoryManager<DNekScalBlkMat>
591  ::AllocateSharedPtr(nint_size , nbdry_size, blkmatStorage);
592  blkMatrices[3] = MemoryManager<DNekScalBlkMat>
593  ::AllocateSharedPtr(nint_size , nint_size , blkmatStorage);
594 
595  DNekScalMatSharedPtr tmpscalmat;
596  for(i = 0; i < nPatches; i++)
597  {
598  for(j = 0; j < 4; j++)
599  {
600  tmpscalmat= MemoryManager<DNekScalMat>
601  ::AllocateSharedPtr(1.0,substructuredMat[j][i]);
602  blkMatrices[j]->SetBlock(i,i,tmpscalmat);
603  }
604  }
605  }
606 
607  // We've finished with the Schur complement matrix passed to this
608  // level, so return the memory to the system. The Schur complement
609  // matrix need only be retained at the last level. Save the other
610  // matrices at this level though.
611  m_schurCompl.reset();
612 
614  m_linSysKey, m_expList, blkMatrices[0], blkMatrices[1],
615  blkMatrices[2], blkMatrices[3], pLocToGloMap);
616  }
617  }
618 }
Array< OneD, NekDouble > m_wsp
Workspace array for matrix multiplication.
#define ASSERTL0(condition, msg)
Definition: ErrorUtil.hpp:188
virtual void v_Solve(const Array< OneD, const NekDouble > &in, Array< OneD, NekDouble > &out, const AssemblyMapSharedPtr &locToGloMap, const Array< OneD, const NekDouble > &dirForcing=NullNekDouble1DArray)
Solve the linear system for given input and output vectors using a specified local to global map...
virtual GlobalLinSysStaticCondSharedPtr v_Recurse(const GlobalLinSysKey &mkey, const boost::weak_ptr< ExpList > &pExpList, const DNekScalBlkMatSharedPtr pSchurCompl, const DNekScalBlkMatSharedPtr pBinvD, const DNekScalBlkMatSharedPtr pC, const DNekScalBlkMatSharedPtr pInvD, const boost::shared_ptr< AssemblyMap > &locToGloMap)=0
static boost::shared_ptr< DataType > AllocateSharedPtr()
Allocate a shared pointer from the memory pool.
#define sign(a, b)
return the sign(b)*a
Definition: Polylib.cpp:22
boost::shared_ptr< AssemblyMap > AssemblyMapSharedPtr
Definition: AssemblyMap.h:53
DNekScalBlkMatSharedPtr m_schurCompl
Block Schur complement matrix.
GlobalLinSysStaticCondSharedPtr m_recursiveSchurCompl
Schur complement for Direct Static Condensation.
DNekScalBlkMatSharedPtr m_C
Block matrix.
void Fill(int n, const T alpha, T *x, const int incx)
Fill a vector with a constant value.
Definition: Vmath.cpp:46
void SolveLinearSystem(const int pNumRows, const Array< OneD, const NekDouble > &pInput, Array< OneD, NekDouble > &pOutput, const AssemblyMapSharedPtr &locToGloMap, const int pNumDir=0)
Solve the linear system for given input and output vectors.
Definition: GlobalLinSys.h:201
virtual void v_BasisTransform(Array< OneD, NekDouble > &pInOut, int offset)
STL namespace.
virtual void v_Initialise(const boost::shared_ptr< AssemblyMap > &locToGloMap)
Initialise this object.
void Initialise(const boost::shared_ptr< AssemblyMap > &pLocToGloMap)
Definition: GlobalLinSys.h:216
virtual DNekScalBlkMatSharedPtr v_GetStaticCondBlock(unsigned int n)
Retrieves a the static condensation block matrices from n-th expansion using the matrix key provided ...
DNekScalBlkMatSharedPtr m_invD
Block matrix.
boost::shared_ptr< DNekScalMat > DNekScalMatSharedPtr
virtual void v_AssembleSchurComplement(boost::shared_ptr< AssemblyMap > pLoctoGloMap)
virtual DNekScalMatSharedPtr v_GetBlock(unsigned int n)
Retrieves the block matrix from n-th expansion using the matrix key provided by the m_linSysKey...
virtual DNekScalBlkMatSharedPtr v_PreSolve(int scLevel, NekVector< NekDouble > &F_GlobBnd)
boost::shared_ptr< DNekScalBlkMat > DNekScalBlkMatSharedPtr
Definition: NekTypeDefs.hpp:74
double NekDouble
void SetupTopLevel(const boost::shared_ptr< AssemblyMap > &locToGloMap)
Set up the storage for the Schur complement or the top level of the multi-level Schur complement...
Describe a linear system.
PointerWrapper
Specifies if the pointer passed to a NekMatrix or NekVector is copied into an internal representation...
StdRegions::MatrixType GetMatrixType() const
Return the matrix type.
const GlobalLinSysKey m_linSysKey
Key associated with this linear system.
Definition: GlobalLinSys.h:127
void Vsub(int n, const T *x, const int incx, const T *y, const int incy, T *z, const int incz)
Subtract vector z = x-y.
Definition: Vmath.cpp:329
virtual int v_GetNumBlocks()
Get the number of blocks in this system.
A global linear system.
Definition: GlobalLinSys.h:74
boost::shared_ptr< AssemblyMap > m_locToGloMap
Local to global map.
void Vcopy(int n, const T *x, const int incx, T *y, const int incy)
Definition: Vmath.cpp:1047
virtual void v_BasisInvTransform(Array< OneD, NekDouble > &pInOut)
DNekScalBlkMatSharedPtr m_BinvD
Block matrix.
Array< OneD, DataType > & GetPtr()
Definition: NekVector.cpp:230
void ConstructNextLevelCondensedSystem(const boost::shared_ptr< AssemblyMap > &locToGloMap)
void Vadd(int n, const T *x, const int incx, const T *y, const int incy, T *z, const int incz)
Add vector z = x+y.
Definition: Vmath.cpp:285
const boost::weak_ptr< ExpList > m_expList
Local Matrix System.
Definition: GlobalLinSys.h:129