/[escript]/branches/ripleygmg_from_3668/ripley/src/Rectangle.cpp
ViewVC logotype

Annotation of /branches/ripleygmg_from_3668/ripley/src/Rectangle.cpp

Parent Directory Parent Directory | Revision Log Revision Log


Revision 3781 - (hide annotations)
Mon Jan 23 22:18:49 2012 UTC (7 years, 9 months ago) by caltinay
File size: 230612 byte(s)
Since setX() is not supported the constructors now take location and extent.
Python wrapper allows construction the old way with l0,l1 or with
tuples (x0,x1),(y0,y1).

1 caltinay 3691
2     /*******************************************************
3     *
4 caltinay 3764 * Copyright (c) 2003-2012 by University of Queensland
5 caltinay 3691 * Earth Systems Science Computational Center (ESSCC)
6     * http://www.uq.edu.au/esscc
7     *
8     * Primary Business: Queensland, Australia
9     * Licensed under the Open Software License version 3.0
10     * http://www.opensource.org/licenses/osl-3.0.php
11     *
12     *******************************************************/
13    
14     #include <ripley/Rectangle.h>
15     extern "C" {
16 caltinay 3748 #include "paso/SystemMatrix.h"
17 caltinay 3691 }
18    
19     #if USE_SILO
20     #include <silo.h>
21     #ifdef ESYS_MPI
22     #include <pmpio.h>
23     #endif
24     #endif
25    
26     #include <iomanip>
27    
28     using namespace std;
29    
30     namespace ripley {
31    
32 caltinay 3781 Rectangle::Rectangle(int n0, int n1, double x0, double y0, double x1,
33     double y1, int d0, int d1) :
34 caltinay 3691 RipleyDomain(2),
35     m_gNE0(n0),
36     m_gNE1(n1),
37 caltinay 3781 m_x0(x0),
38     m_y0(y0),
39     m_l0(x1-x0),
40     m_l1(y1-y0),
41 caltinay 3691 m_NX(d0),
42     m_NY(d1)
43     {
44     // ensure number of subdivisions is valid and nodes can be distributed
45     // among number of ranks
46     if (m_NX*m_NY != m_mpiInfo->size)
47     throw RipleyException("Invalid number of spatial subdivisions");
48    
49 caltinay 3752 if ((n0+1)%m_NX > 0 || (n1+1)%m_NY > 0)
50     throw RipleyException("Number of elements+1 must be separable into number of ranks in each dimension");
51 caltinay 3691
52 caltinay 3752 if ((m_NX > 1 && (n0+1)/m_NX<2) || (m_NY > 1 && (n1+1)/m_NY<2))
53     throw RipleyException("Too few elements for the number of ranks");
54    
55 caltinay 3764 // local number of elements (with and without overlap)
56     m_NE0 = m_ownNE0 = (m_NX>1 ? (n0+1)/m_NX : n0);
57 caltinay 3752 if (m_mpiInfo->rank%m_NX>0 && m_mpiInfo->rank%m_NX<m_NX-1)
58     m_NE0++;
59 caltinay 3764 else if (m_NX>1 && m_mpiInfo->rank%m_NX==m_NX-1)
60     m_ownNE0--;
61    
62     m_NE1 = m_ownNE1 = (m_NY>1 ? (n1+1)/m_NY : n1);
63 caltinay 3752 if (m_mpiInfo->rank/m_NX>0 && m_mpiInfo->rank/m_NX<m_NY-1)
64     m_NE1++;
65 caltinay 3764 else if (m_NY>1 && m_mpiInfo->rank/m_NX==m_NY-1)
66     m_ownNE1--;
67 caltinay 3752
68     // local number of nodes
69 caltinay 3691 m_N0 = m_NE0+1;
70     m_N1 = m_NE1+1;
71 caltinay 3752
72 caltinay 3691 // bottom-left node is at (offset0,offset1) in global mesh
73 caltinay 3752 m_offset0 = (n0+1)/m_NX*(m_mpiInfo->rank%m_NX);
74 caltinay 3753 if (m_offset0 > 0)
75 caltinay 3752 m_offset0--;
76     m_offset1 = (n1+1)/m_NY*(m_mpiInfo->rank/m_NX);
77 caltinay 3753 if (m_offset1 > 0)
78 caltinay 3752 m_offset1--;
79    
80 caltinay 3691 populateSampleIds();
81 caltinay 3756 createPattern();
82 caltinay 3691 }
83    
84     Rectangle::~Rectangle()
85     {
86     }
87    
88     string Rectangle::getDescription() const
89     {
90     return "ripley::Rectangle";
91     }
92    
93     bool Rectangle::operator==(const AbstractDomain& other) const
94     {
95 caltinay 3744 const Rectangle* o=dynamic_cast<const Rectangle*>(&other);
96     if (o) {
97     return (RipleyDomain::operator==(other) &&
98     m_gNE0==o->m_gNE0 && m_gNE1==o->m_gNE1
99 caltinay 3781 && m_x0==o->m_x0 && m_y0==o->m_y0
100 caltinay 3744 && m_l0==o->m_l0 && m_l1==o->m_l1
101     && m_NX==o->m_NX && m_NY==o->m_NY);
102     }
103 caltinay 3691
104     return false;
105     }
106    
107     void Rectangle::dump(const string& fileName) const
108     {
109     #if USE_SILO
110     string fn(fileName);
111     if (fileName.length() < 6 || fileName.compare(fileName.length()-5, 5, ".silo") != 0) {
112     fn+=".silo";
113     }
114    
115     const int NUM_SILO_FILES = 1;
116     const char* blockDirFmt = "/block%04d";
117     int driver=DB_HDF5;
118     DBfile* dbfile = NULL;
119    
120     #ifdef ESYS_MPI
121     PMPIO_baton_t* baton = NULL;
122     #endif
123    
124     if (m_mpiInfo->size > 1) {
125     #ifdef ESYS_MPI
126     baton = PMPIO_Init(NUM_SILO_FILES, PMPIO_WRITE, m_mpiInfo->comm,
127     0x1337, PMPIO_DefaultCreate, PMPIO_DefaultOpen,
128     PMPIO_DefaultClose, (void*)&driver);
129     // try the fallback driver in case of error
130     if (!baton && driver != DB_PDB) {
131     driver = DB_PDB;
132     baton = PMPIO_Init(NUM_SILO_FILES, PMPIO_WRITE, m_mpiInfo->comm,
133     0x1338, PMPIO_DefaultCreate, PMPIO_DefaultOpen,
134     PMPIO_DefaultClose, (void*)&driver);
135     }
136     if (baton) {
137 caltinay 3766 char siloPath[64];
138     snprintf(siloPath, 64, blockDirFmt, PMPIO_RankInGroup(baton, m_mpiInfo->rank));
139     dbfile = (DBfile*) PMPIO_WaitForBaton(baton, fn.c_str(), siloPath);
140 caltinay 3691 }
141     #endif
142     } else {
143     dbfile = DBCreate(fn.c_str(), DB_CLOBBER, DB_LOCAL,
144     getDescription().c_str(), driver);
145     // try the fallback driver in case of error
146     if (!dbfile && driver != DB_PDB) {
147     driver = DB_PDB;
148     dbfile = DBCreate(fn.c_str(), DB_CLOBBER, DB_LOCAL,
149     getDescription().c_str(), driver);
150     }
151 caltinay 3766 char siloPath[64];
152     snprintf(siloPath, 64, blockDirFmt, 0);
153     DBMkDir(dbfile, siloPath);
154     DBSetDir(dbfile, siloPath);
155 caltinay 3691 }
156    
157     if (!dbfile)
158     throw RipleyException("dump: Could not create Silo file");
159    
160     /*
161     if (driver==DB_HDF5) {
162     // gzip level 1 already provides good compression with minimal
163     // performance penalty. Some tests showed that gzip levels >3 performed
164     // rather badly on escript data both in terms of time and space
165     DBSetCompression("ERRMODE=FALLBACK METHOD=GZIP LEVEL=1");
166     }
167     */
168    
169     boost::scoped_ptr<double> x(new double[m_N0]);
170     boost::scoped_ptr<double> y(new double[m_N1]);
171     double* coords[2] = { x.get(), y.get() };
172 caltinay 3697 pair<double,double> xdx = getFirstCoordAndSpacing(0);
173     pair<double,double> ydy = getFirstCoordAndSpacing(1);
174 caltinay 3691 #pragma omp parallel
175     {
176 caltinay 3722 #pragma omp for nowait
177 caltinay 3691 for (dim_t i0 = 0; i0 < m_N0; i0++) {
178 caltinay 3697 coords[0][i0]=xdx.first+i0*xdx.second;
179 caltinay 3691 }
180 caltinay 3722 #pragma omp for nowait
181 caltinay 3691 for (dim_t i1 = 0; i1 < m_N1; i1++) {
182 caltinay 3697 coords[1][i1]=ydy.first+i1*ydy.second;
183 caltinay 3691 }
184     }
185 caltinay 3697 IndexVector dims = getNumNodesPerDim();
186    
187     // write mesh
188     DBPutQuadmesh(dbfile, "mesh", NULL, coords, &dims[0], 2, DB_DOUBLE,
189 caltinay 3691 DB_COLLINEAR, NULL);
190    
191 caltinay 3697 // write node ids
192     DBPutQuadvar1(dbfile, "nodeId", "mesh", (void*)&m_nodeId[0], &dims[0], 2,
193     NULL, 0, DB_INT, DB_NODECENT, NULL);
194    
195     // write element ids
196     dims = getNumElementsPerDim();
197     DBPutQuadvar1(dbfile, "elementId", "mesh", (void*)&m_elementId[0],
198     &dims[0], 2, NULL, 0, DB_INT, DB_ZONECENT, NULL);
199    
200     // rank 0 writes multimesh and multivar
201 caltinay 3691 if (m_mpiInfo->rank == 0) {
202     vector<string> tempstrings;
203 caltinay 3697 vector<char*> names;
204 caltinay 3691 for (dim_t i=0; i<m_mpiInfo->size; i++) {
205     stringstream path;
206     path << "/block" << setw(4) << setfill('0') << right << i << "/mesh";
207     tempstrings.push_back(path.str());
208 caltinay 3697 names.push_back((char*)tempstrings.back().c_str());
209 caltinay 3691 }
210 caltinay 3697 vector<int> types(m_mpiInfo->size, DB_QUAD_RECT);
211 caltinay 3691 DBSetDir(dbfile, "/");
212 caltinay 3697 DBPutMultimesh(dbfile, "multimesh", m_mpiInfo->size, &names[0],
213     &types[0], NULL);
214     tempstrings.clear();
215     names.clear();
216     for (dim_t i=0; i<m_mpiInfo->size; i++) {
217     stringstream path;
218     path << "/block" << setw(4) << setfill('0') << right << i << "/nodeId";
219     tempstrings.push_back(path.str());
220     names.push_back((char*)tempstrings.back().c_str());
221     }
222     types.assign(m_mpiInfo->size, DB_QUADVAR);
223     DBPutMultivar(dbfile, "nodeId", m_mpiInfo->size, &names[0],
224     &types[0], NULL);
225     tempstrings.clear();
226     names.clear();
227     for (dim_t i=0; i<m_mpiInfo->size; i++) {
228     stringstream path;
229     path << "/block" << setw(4) << setfill('0') << right << i << "/elementId";
230     tempstrings.push_back(path.str());
231     names.push_back((char*)tempstrings.back().c_str());
232     }
233     DBPutMultivar(dbfile, "elementId", m_mpiInfo->size, &names[0],
234     &types[0], NULL);
235 caltinay 3691 }
236    
237     if (m_mpiInfo->size > 1) {
238     #ifdef ESYS_MPI
239     PMPIO_HandOffBaton(baton, dbfile);
240     PMPIO_Finish(baton);
241     #endif
242     } else {
243     DBClose(dbfile);
244     }
245    
246     #else // USE_SILO
247     throw RipleyException("dump(): no Silo support");
248     #endif
249     }
250    
251 caltinay 3697 const int* Rectangle::borrowSampleReferenceIDs(int fsType) const
252 caltinay 3691 {
253 caltinay 3697 switch (fsType) {
254 caltinay 3691 case Nodes:
255 caltinay 3769 case ReducedNodes: // FIXME: reduced
256 caltinay 3691 return &m_nodeId[0];
257 caltinay 3750 case DegreesOfFreedom:
258 caltinay 3769 case ReducedDegreesOfFreedom: // FIXME: reduced
259 caltinay 3750 return &m_dofId[0];
260 caltinay 3691 case Elements:
261 caltinay 3733 case ReducedElements:
262 caltinay 3691 return &m_elementId[0];
263     case FaceElements:
264 caltinay 3733 case ReducedFaceElements:
265 caltinay 3691 return &m_faceId[0];
266     default:
267     break;
268     }
269    
270     stringstream msg;
271     msg << "borrowSampleReferenceIDs() not implemented for function space type "
272 caltinay 3701 << functionSpaceTypeAsString(fsType);
273 caltinay 3691 throw RipleyException(msg.str());
274     }
275    
276 caltinay 3757 bool Rectangle::ownSample(int fsType, index_t id) const
277 caltinay 3691 {
278 caltinay 3759 if (getMPISize()==1)
279     return true;
280    
281 caltinay 3757 switch (fsType) {
282     case Nodes:
283 caltinay 3769 case ReducedNodes: // FIXME: reduced
284 caltinay 3757 return (m_dofMap[id] < getNumDOF());
285     case DegreesOfFreedom:
286     case ReducedDegreesOfFreedom:
287     return true;
288     case Elements:
289     case ReducedElements:
290     // check ownership of element's bottom left node
291     return (m_dofMap[id%m_NE0+m_N0*(id/m_NE0)] < getNumDOF());
292     case FaceElements:
293     case ReducedFaceElements:
294 caltinay 3759 {
295 caltinay 3764 // determine which face the sample belongs to before
296 caltinay 3768 // checking ownership of corresponding element's first node
297 caltinay 3759 const IndexVector faces = getNumFacesPerBoundary();
298     dim_t n=0;
299     for (size_t i=0; i<faces.size(); i++) {
300     n+=faces[i];
301     if (id<n) {
302     index_t k;
303     if (i==1)
304 caltinay 3768 k=m_N0-2;
305 caltinay 3759 else if (i==3)
306 caltinay 3768 k=m_N0*(m_N1-2);
307 caltinay 3759 else
308     k=0;
309     // determine whether to move right or up
310     const index_t delta=(i/2==0 ? m_N0 : 1);
311     return (m_dofMap[k+(id-n+faces[i])*delta] < getNumDOF());
312     }
313     }
314     return false;
315     }
316 caltinay 3757 default:
317     break;
318 caltinay 3702 }
319 caltinay 3757
320     stringstream msg;
321     msg << "ownSample() not implemented for "
322     << functionSpaceTypeAsString(fsType);
323     throw RipleyException(msg.str());
324 caltinay 3691 }
325    
326 caltinay 3764 void Rectangle::setToNormal(escript::Data& out) const
327 caltinay 3691 {
328 caltinay 3764 if (out.getFunctionSpace().getTypeCode() == FaceElements) {
329     out.requireWrite();
330     #pragma omp parallel
331     {
332     if (m_faceOffset[0] > -1) {
333     #pragma omp for nowait
334     for (index_t k1 = 0; k1 < m_NE1; ++k1) {
335     double* o = out.getSampleDataRW(m_faceOffset[0]+k1);
336     // set vector at two quadrature points
337     *o++ = -1.;
338     *o++ = 0.;
339     *o++ = -1.;
340     *o = 0.;
341     }
342     }
343    
344     if (m_faceOffset[1] > -1) {
345     #pragma omp for nowait
346     for (index_t k1 = 0; k1 < m_NE1; ++k1) {
347     double* o = out.getSampleDataRW(m_faceOffset[1]+k1);
348     // set vector at two quadrature points
349     *o++ = 1.;
350     *o++ = 0.;
351     *o++ = 1.;
352     *o = 0.;
353     }
354     }
355    
356     if (m_faceOffset[2] > -1) {
357     #pragma omp for nowait
358     for (index_t k0 = 0; k0 < m_NE0; ++k0) {
359     double* o = out.getSampleDataRW(m_faceOffset[2]+k0);
360     // set vector at two quadrature points
361     *o++ = 0.;
362     *o++ = -1.;
363     *o++ = 0.;
364     *o = -1.;
365     }
366     }
367    
368     if (m_faceOffset[3] > -1) {
369     #pragma omp for nowait
370     for (index_t k0 = 0; k0 < m_NE0; ++k0) {
371     double* o = out.getSampleDataRW(m_faceOffset[3]+k0);
372     // set vector at two quadrature points
373     *o++ = 0.;
374     *o++ = 1.;
375     *o++ = 0.;
376     *o = 1.;
377     }
378     }
379     } // end of parallel section
380     } else if (out.getFunctionSpace().getTypeCode() == ReducedFaceElements) {
381     out.requireWrite();
382     #pragma omp parallel
383     {
384     if (m_faceOffset[0] > -1) {
385     #pragma omp for nowait
386     for (index_t k1 = 0; k1 < m_NE1; ++k1) {
387     double* o = out.getSampleDataRW(m_faceOffset[0]+k1);
388     *o++ = -1.;
389     *o = 0.;
390     }
391     }
392    
393     if (m_faceOffset[1] > -1) {
394     #pragma omp for nowait
395     for (index_t k1 = 0; k1 < m_NE1; ++k1) {
396     double* o = out.getSampleDataRW(m_faceOffset[1]+k1);
397     *o++ = 1.;
398     *o = 0.;
399     }
400     }
401    
402     if (m_faceOffset[2] > -1) {
403     #pragma omp for nowait
404     for (index_t k0 = 0; k0 < m_NE0; ++k0) {
405     double* o = out.getSampleDataRW(m_faceOffset[2]+k0);
406     *o++ = 0.;
407     *o = -1.;
408     }
409     }
410    
411     if (m_faceOffset[3] > -1) {
412     #pragma omp for nowait
413     for (index_t k0 = 0; k0 < m_NE0; ++k0) {
414     double* o = out.getSampleDataRW(m_faceOffset[3]+k0);
415     *o++ = 0.;
416     *o = 1.;
417     }
418     }
419     } // end of parallel section
420    
421     } else {
422     stringstream msg;
423     msg << "setToNormal() not implemented for "
424     << functionSpaceTypeAsString(out.getFunctionSpace().getTypeCode());
425     throw RipleyException(msg.str());
426     }
427     }
428    
429     void Rectangle::setToSize(escript::Data& out) const
430     {
431     if (out.getFunctionSpace().getTypeCode() == Elements
432     || out.getFunctionSpace().getTypeCode() == ReducedElements) {
433     out.requireWrite();
434     const dim_t numQuad=out.getNumDataPointsPerSample();
435     const double hSize=getFirstCoordAndSpacing(0).second;
436     const double vSize=getFirstCoordAndSpacing(1).second;
437     const double size=min(hSize,vSize);
438     #pragma omp parallel for
439     for (index_t k = 0; k < getNumElements(); ++k) {
440     double* o = out.getSampleDataRW(k);
441     fill(o, o+numQuad, size);
442     }
443     } else if (out.getFunctionSpace().getTypeCode() == FaceElements
444     || out.getFunctionSpace().getTypeCode() == ReducedFaceElements) {
445     out.requireWrite();
446     const dim_t numQuad=out.getNumDataPointsPerSample();
447     const double hSize=getFirstCoordAndSpacing(0).second;
448     const double vSize=getFirstCoordAndSpacing(1).second;
449     #pragma omp parallel
450     {
451     if (m_faceOffset[0] > -1) {
452     #pragma omp for nowait
453     for (index_t k1 = 0; k1 < m_NE1; ++k1) {
454     double* o = out.getSampleDataRW(m_faceOffset[0]+k1);
455     fill(o, o+numQuad, vSize);
456     }
457     }
458    
459     if (m_faceOffset[1] > -1) {
460     #pragma omp for nowait
461     for (index_t k1 = 0; k1 < m_NE1; ++k1) {
462     double* o = out.getSampleDataRW(m_faceOffset[1]+k1);
463     fill(o, o+numQuad, vSize);
464     }
465     }
466    
467     if (m_faceOffset[2] > -1) {
468     #pragma omp for nowait
469     for (index_t k0 = 0; k0 < m_NE0; ++k0) {
470     double* o = out.getSampleDataRW(m_faceOffset[2]+k0);
471     fill(o, o+numQuad, hSize);
472     }
473     }
474    
475     if (m_faceOffset[3] > -1) {
476     #pragma omp for nowait
477     for (index_t k0 = 0; k0 < m_NE0; ++k0) {
478     double* o = out.getSampleDataRW(m_faceOffset[3]+k0);
479     fill(o, o+numQuad, hSize);
480     }
481     }
482     } // end of parallel section
483    
484     } else {
485     stringstream msg;
486     msg << "setToSize() not implemented for "
487     << functionSpaceTypeAsString(out.getFunctionSpace().getTypeCode());
488     throw RipleyException(msg.str());
489     }
490     }
491    
492     Paso_SystemMatrixPattern* Rectangle::getPattern(bool reducedRowOrder,
493     bool reducedColOrder) const
494     {
495 caltinay 3776 /* FIXME: reduced
496 caltinay 3764 if (reducedRowOrder || reducedColOrder)
497     throw RipleyException("getPattern() not implemented for reduced order");
498 caltinay 3776 */
499 caltinay 3764 return m_pattern;
500     }
501    
502     void Rectangle::Print_Mesh_Info(const bool full) const
503     {
504     RipleyDomain::Print_Mesh_Info(full);
505     if (full) {
506     cout << " Id Coordinates" << endl;
507     cout.precision(15);
508     cout.setf(ios::scientific, ios::floatfield);
509     pair<double,double> xdx = getFirstCoordAndSpacing(0);
510     pair<double,double> ydy = getFirstCoordAndSpacing(1);
511     for (index_t i=0; i < getNumNodes(); i++) {
512     cout << " " << setw(5) << m_nodeId[i]
513     << " " << xdx.first+(i%m_N0)*xdx.second
514     << " " << ydy.first+(i/m_N0)*ydy.second << endl;
515     }
516     }
517     }
518    
519     IndexVector Rectangle::getNumNodesPerDim() const
520     {
521     IndexVector ret;
522     ret.push_back(m_N0);
523     ret.push_back(m_N1);
524     return ret;
525     }
526    
527     IndexVector Rectangle::getNumElementsPerDim() const
528     {
529     IndexVector ret;
530     ret.push_back(m_NE0);
531     ret.push_back(m_NE1);
532     return ret;
533     }
534    
535     IndexVector Rectangle::getNumFacesPerBoundary() const
536     {
537     IndexVector ret(4, 0);
538     //left
539     if (m_offset0==0)
540     ret[0]=m_NE1;
541     //right
542     if (m_mpiInfo->rank%m_NX==m_NX-1)
543     ret[1]=m_NE1;
544     //bottom
545     if (m_offset1==0)
546     ret[2]=m_NE0;
547     //top
548     if (m_mpiInfo->rank/m_NX==m_NY-1)
549     ret[3]=m_NE0;
550     return ret;
551     }
552    
553 caltinay 3766 IndexVector Rectangle::getNumSubdivisionsPerDim() const
554     {
555     IndexVector ret;
556     ret.push_back(m_NX);
557     ret.push_back(m_NY);
558     return ret;
559     }
560    
561 caltinay 3764 pair<double,double> Rectangle::getFirstCoordAndSpacing(dim_t dim) const
562     {
563     if (dim==0) {
564 caltinay 3781 return pair<double,double>(m_x0+(m_l0*m_offset0)/m_gNE0, m_l0/m_gNE0);
565 caltinay 3764 } else if (dim==1) {
566 caltinay 3781 return pair<double,double>(m_y0+(m_l1*m_offset1)/m_gNE1, m_l1/m_gNE1);
567 caltinay 3764 }
568     throw RipleyException("getFirstCoordAndSpacing(): invalid argument");
569     }
570    
571     //protected
572     dim_t Rectangle::getNumDOF() const
573     {
574     return (m_gNE0+1)/m_NX*(m_gNE1+1)/m_NY;
575     }
576    
577     //protected
578     dim_t Rectangle::getNumFaceElements() const
579     {
580     const IndexVector faces = getNumFacesPerBoundary();
581     dim_t n=0;
582     for (size_t i=0; i<faces.size(); i++)
583     n+=faces[i];
584     return n;
585     }
586    
587     //protected
588     void Rectangle::assembleCoordinates(escript::Data& arg) const
589     {
590     escriptDataC x = arg.getDataC();
591     int numDim = m_numDim;
592     if (!isDataPointShapeEqual(&x, 1, &numDim))
593     throw RipleyException("setToX: Invalid Data object shape");
594     if (!numSamplesEqual(&x, 1, getNumNodes()))
595     throw RipleyException("setToX: Illegal number of samples in Data object");
596    
597     pair<double,double> xdx = getFirstCoordAndSpacing(0);
598     pair<double,double> ydy = getFirstCoordAndSpacing(1);
599     arg.requireWrite();
600     #pragma omp parallel for
601     for (dim_t i1 = 0; i1 < m_N1; i1++) {
602     for (dim_t i0 = 0; i0 < m_N0; i0++) {
603     double* point = arg.getSampleDataRW(i0+m_N0*i1);
604     point[0] = xdx.first+i0*xdx.second;
605     point[1] = ydy.first+i1*ydy.second;
606     }
607     }
608     }
609    
610     //protected
611     void Rectangle::assembleGradient(escript::Data& out, escript::Data& in) const
612     {
613 caltinay 3701 const dim_t numComp = in.getDataPointSize();
614 caltinay 3702 const double h0 = m_l0/m_gNE0;
615     const double h1 = m_l1/m_gNE1;
616 caltinay 3724 const double cx0 = -1./h0;
617     const double cx1 = -.78867513459481288225/h0;
618     const double cx2 = -.5/h0;
619     const double cx3 = -.21132486540518711775/h0;
620     const double cx4 = .21132486540518711775/h0;
621     const double cx5 = .5/h0;
622     const double cx6 = .78867513459481288225/h0;
623     const double cx7 = 1./h0;
624     const double cy0 = -1./h1;
625     const double cy1 = -.78867513459481288225/h1;
626     const double cy2 = -.5/h1;
627     const double cy3 = -.21132486540518711775/h1;
628     const double cy4 = .21132486540518711775/h1;
629     const double cy5 = .5/h1;
630     const double cy6 = .78867513459481288225/h1;
631     const double cy7 = 1./h1;
632    
633 caltinay 3702 if (out.getFunctionSpace().getTypeCode() == Elements) {
634 caltinay 3760 out.requireWrite();
635 caltinay 3697 #pragma omp parallel for
636 caltinay 3707 for (index_t k1=0; k1 < m_NE1; ++k1) {
637     for (index_t k0=0; k0 < m_NE0; ++k0) {
638 caltinay 3769 const double* f_10 = in.getSampleDataRO(INDEX2(k0+1,k1, m_N0));
639     const double* f_11 = in.getSampleDataRO(INDEX2(k0+1,k1+1, m_N0));
640     const double* f_01 = in.getSampleDataRO(INDEX2(k0,k1+1, m_N0));
641     const double* f_00 = in.getSampleDataRO(INDEX2(k0,k1, m_N0));
642 caltinay 3702 double* o = out.getSampleDataRW(INDEX2(k0,k1,m_NE0));
643     for (index_t i=0; i < numComp; ++i) {
644 caltinay 3724 o[INDEX3(i,0,0,numComp,2)] = f_00[i]*cx1 + f_01[i]*cx3 + f_10[i]*cx6 + f_11[i]*cx4;
645     o[INDEX3(i,1,0,numComp,2)] = f_00[i]*cy1 + f_01[i]*cy6 + f_10[i]*cy3 + f_11[i]*cy4;
646     o[INDEX3(i,0,1,numComp,2)] = f_00[i]*cx1 + f_01[i]*cx3 + f_10[i]*cx6 + f_11[i]*cx4;
647     o[INDEX3(i,1,1,numComp,2)] = f_00[i]*cy3 + f_01[i]*cy4 + f_10[i]*cy1 + f_11[i]*cy6;
648     o[INDEX3(i,0,2,numComp,2)] = f_00[i]*cx3 + f_01[i]*cx1 + f_10[i]*cx4 + f_11[i]*cx6;
649     o[INDEX3(i,1,2,numComp,2)] = f_00[i]*cy1 + f_01[i]*cy6 + f_10[i]*cy3 + f_11[i]*cy4;
650     o[INDEX3(i,0,3,numComp,2)] = f_00[i]*cx3 + f_01[i]*cx1 + f_10[i]*cx4 + f_11[i]*cx6;
651     o[INDEX3(i,1,3,numComp,2)] = f_00[i]*cy3 + f_01[i]*cy4 + f_10[i]*cy1 + f_11[i]*cy6;
652 caltinay 3702 } /* end of component loop i */
653     } /* end of k0 loop */
654     } /* end of k1 loop */
655 caltinay 3711 } else if (out.getFunctionSpace().getTypeCode() == ReducedElements) {
656 caltinay 3760 out.requireWrite();
657 caltinay 3711 #pragma omp parallel for
658     for (index_t k1=0; k1 < m_NE1; ++k1) {
659     for (index_t k0=0; k0 < m_NE0; ++k0) {
660 caltinay 3769 const double* f_10 = in.getSampleDataRO(INDEX2(k0+1,k1, m_N0));
661     const double* f_11 = in.getSampleDataRO(INDEX2(k0+1,k1+1, m_N0));
662     const double* f_01 = in.getSampleDataRO(INDEX2(k0,k1+1, m_N0));
663     const double* f_00 = in.getSampleDataRO(INDEX2(k0,k1, m_N0));
664 caltinay 3711 double* o = out.getSampleDataRW(INDEX2(k0,k1,m_NE0));
665     for (index_t i=0; i < numComp; ++i) {
666 caltinay 3724 o[INDEX3(i,0,0,numComp,2)] = cx5*(f_10[i] + f_11[i]) + cx2*(f_00[i] + f_01[i]);
667     o[INDEX3(i,1,0,numComp,2)] = cy2*(f_00[i] + f_10[i]) + cy5*(f_01[i] + f_11[i]);
668 caltinay 3711 } /* end of component loop i */
669     } /* end of k0 loop */
670     } /* end of k1 loop */
671 caltinay 3707 } else if (out.getFunctionSpace().getTypeCode() == FaceElements) {
672 caltinay 3760 out.requireWrite();
673 caltinay 3722 #pragma omp parallel
674     {
675     if (m_faceOffset[0] > -1) {
676     #pragma omp for nowait
677     for (index_t k1=0; k1 < m_NE1; ++k1) {
678 caltinay 3769 const double* f_10 = in.getSampleDataRO(INDEX2(1,k1, m_N0));
679     const double* f_11 = in.getSampleDataRO(INDEX2(1,k1+1, m_N0));
680     const double* f_01 = in.getSampleDataRO(INDEX2(0,k1+1, m_N0));
681     const double* f_00 = in.getSampleDataRO(INDEX2(0,k1, m_N0));
682 caltinay 3722 double* o = out.getSampleDataRW(m_faceOffset[0]+k1);
683     for (index_t i=0; i < numComp; ++i) {
684 caltinay 3724 o[INDEX3(i,0,0,numComp,2)] = f_00[i]*cx1 + f_01[i]*cx3 + f_10[i]*cx6 + f_11[i]*cx4;
685     o[INDEX3(i,1,0,numComp,2)] = f_00[i]*cy0 + f_01[i]*cy7;
686     o[INDEX3(i,0,1,numComp,2)] = f_00[i]*cx3 + f_01[i]*cx1 + f_10[i]*cx4 + f_11[i]*cx6;
687     o[INDEX3(i,1,1,numComp,2)] = f_00[i]*cy0 + f_01[i]*cy7;
688 caltinay 3722 } /* end of component loop i */
689     } /* end of k1 loop */
690     } /* end of face 0 */
691     if (m_faceOffset[1] > -1) {
692     #pragma omp for nowait
693     for (index_t k1=0; k1 < m_NE1; ++k1) {
694 caltinay 3769 const double* f_10 = in.getSampleDataRO(INDEX2(m_N0-1,k1, m_N0));
695     const double* f_11 = in.getSampleDataRO(INDEX2(m_N0-1,k1+1, m_N0));
696     const double* f_01 = in.getSampleDataRO(INDEX2(m_N0-2,k1+1, m_N0));
697     const double* f_00 = in.getSampleDataRO(INDEX2(m_N0-2,k1, m_N0));
698 caltinay 3722 double* o = out.getSampleDataRW(m_faceOffset[1]+k1);
699     for (index_t i=0; i < numComp; ++i) {
700 caltinay 3724 o[INDEX3(i,0,0,numComp,2)] = f_00[i]*cx1 + f_01[i]*cx3 + f_10[i]*cx6 + f_11[i]*cx4;
701     o[INDEX3(i,1,0,numComp,2)] = f_10[i]*cy0 + f_11[i]*cy7;
702     o[INDEX3(i,0,1,numComp,2)] = f_00[i]*cx3 + f_01[i]*cx1 + f_10[i]*cx4 + f_11[i]*cx6;
703     o[INDEX3(i,1,1,numComp,2)] = f_10[i]*cy0 + f_11[i]*cy7;
704 caltinay 3722 } /* end of component loop i */
705     } /* end of k1 loop */
706     } /* end of face 1 */
707     if (m_faceOffset[2] > -1) {
708     #pragma omp for nowait
709     for (index_t k0=0; k0 < m_NE0; ++k0) {
710 caltinay 3769 const double* f_00 = in.getSampleDataRO(INDEX2(k0,0, m_N0));
711     const double* f_10 = in.getSampleDataRO(INDEX2(k0+1,0, m_N0));
712     const double* f_11 = in.getSampleDataRO(INDEX2(k0+1,1, m_N0));
713     const double* f_01 = in.getSampleDataRO(INDEX2(k0,1, m_N0));
714 caltinay 3722 double* o = out.getSampleDataRW(m_faceOffset[2]+k0);
715     for (index_t i=0; i < numComp; ++i) {
716 caltinay 3724 o[INDEX3(i,0,0,numComp,2)] = f_00[i]*cx0 + f_10[i]*cx7;
717     o[INDEX3(i,1,0,numComp,2)] = f_00[i]*cy1 + f_01[i]*cy6 + f_10[i]*cy3 + f_11[i]*cy4;
718     o[INDEX3(i,0,1,numComp,2)] = f_00[i]*cx0 + f_10[i]*cx7;
719     o[INDEX3(i,1,1,numComp,2)] = f_00[i]*cy3 + f_01[i]*cy4 + f_10[i]*cy1 + f_11[i]*cy6;
720 caltinay 3722 } /* end of component loop i */
721     } /* end of k0 loop */
722     } /* end of face 2 */
723     if (m_faceOffset[3] > -1) {
724     #pragma omp for nowait
725     for (index_t k0=0; k0 < m_NE0; ++k0) {
726 caltinay 3769 const double* f_11 = in.getSampleDataRO(INDEX2(k0+1,m_N1-1, m_N0));
727     const double* f_01 = in.getSampleDataRO(INDEX2(k0,m_N1-1, m_N0));
728     const double* f_10 = in.getSampleDataRO(INDEX2(k0+1,m_N1-2, m_N0));
729     const double* f_00 = in.getSampleDataRO(INDEX2(k0,m_N1-2, m_N0));
730 caltinay 3722 double* o = out.getSampleDataRW(m_faceOffset[3]+k0);
731     for (index_t i=0; i < numComp; ++i) {
732 caltinay 3724 o[INDEX3(i,0,0,numComp,2)] = f_01[i]*cx0 + f_11[i]*cx7;
733     o[INDEX3(i,1,0,numComp,2)] = f_00[i]*cy1 + f_01[i]*cy6 + f_10[i]*cy3 + f_11[i]*cy4;
734     o[INDEX3(i,0,1,numComp,2)] = f_01[i]*cx0 + f_11[i]*cx7;
735     o[INDEX3(i,1,1,numComp,2)] = f_00[i]*cy3 + f_01[i]*cy4 + f_10[i]*cy1 + f_11[i]*cy6;
736 caltinay 3722 } /* end of component loop i */
737     } /* end of k0 loop */
738     } /* end of face 3 */
739     } // end of parallel section
740 caltinay 3711 } else if (out.getFunctionSpace().getTypeCode() == ReducedFaceElements) {
741 caltinay 3760 out.requireWrite();
742 caltinay 3722 #pragma omp parallel
743     {
744     if (m_faceOffset[0] > -1) {
745     #pragma omp for nowait
746     for (index_t k1=0; k1 < m_NE1; ++k1) {
747 caltinay 3769 const double* f_10 = in.getSampleDataRO(INDEX2(1,k1, m_N0));
748     const double* f_11 = in.getSampleDataRO(INDEX2(1,k1+1, m_N0));
749     const double* f_01 = in.getSampleDataRO(INDEX2(0,k1+1, m_N0));
750     const double* f_00 = in.getSampleDataRO(INDEX2(0,k1, m_N0));
751 caltinay 3722 double* o = out.getSampleDataRW(m_faceOffset[0]+k1);
752     for (index_t i=0; i < numComp; ++i) {
753 caltinay 3724 o[INDEX3(i,0,0,numComp,2)] = cx5*(f_10[i] + f_11[i]) + cx2*(f_00[i] + f_01[i]);
754     o[INDEX3(i,1,0,numComp,2)] = f_00[i]*cy0 + f_01[i]*cy7;
755 caltinay 3722 } /* end of component loop i */
756     } /* end of k1 loop */
757     } /* end of face 0 */
758     if (m_faceOffset[1] > -1) {
759     #pragma omp for nowait
760     for (index_t k1=0; k1 < m_NE1; ++k1) {
761 caltinay 3769 const double* f_10 = in.getSampleDataRO(INDEX2(m_N0-1,k1, m_N0));
762     const double* f_11 = in.getSampleDataRO(INDEX2(m_N0-1,k1+1, m_N0));
763     const double* f_01 = in.getSampleDataRO(INDEX2(m_N0-2,k1+1, m_N0));
764     const double* f_00 = in.getSampleDataRO(INDEX2(m_N0-2,k1, m_N0));
765 caltinay 3722 double* o = out.getSampleDataRW(m_faceOffset[1]+k1);
766     for (index_t i=0; i < numComp; ++i) {
767 caltinay 3724 o[INDEX3(i,0,0,numComp,2)] = cx5*(f_10[i] + f_11[i]) + cx2*(f_00[i] + f_01[i]);
768     o[INDEX3(i,1,0,numComp,2)] = f_10[i]*cy0 + f_11[i]*cy7;
769 caltinay 3722 } /* end of component loop i */
770     } /* end of k1 loop */
771     } /* end of face 1 */
772     if (m_faceOffset[2] > -1) {
773     #pragma omp for nowait
774     for (index_t k0=0; k0 < m_NE0; ++k0) {
775 caltinay 3769 const double* f_00 = in.getSampleDataRO(INDEX2(k0,0, m_N0));
776     const double* f_10 = in.getSampleDataRO(INDEX2(k0+1,0, m_N0));
777     const double* f_11 = in.getSampleDataRO(INDEX2(k0+1,1, m_N0));
778     const double* f_01 = in.getSampleDataRO(INDEX2(k0,1, m_N0));
779 caltinay 3722 double* o = out.getSampleDataRW(m_faceOffset[2]+k0);
780     for (index_t i=0; i < numComp; ++i) {
781 caltinay 3724 o[INDEX3(i,0,0,numComp,2)] = f_00[i]*cx0 + f_10[i]*cx7;
782     o[INDEX3(i,1,0,numComp,2)] = cy2*(f_00[i] + f_10[i]) + cy5*(f_01[i] + f_11[i]);
783 caltinay 3722 } /* end of component loop i */
784     } /* end of k0 loop */
785     } /* end of face 2 */
786     if (m_faceOffset[3] > -1) {
787     #pragma omp for nowait
788     for (index_t k0=0; k0 < m_NE0; ++k0) {
789 caltinay 3769 const double* f_11 = in.getSampleDataRO(INDEX2(k0+1,m_N1-1, m_N0));
790     const double* f_01 = in.getSampleDataRO(INDEX2(k0,m_N1-1, m_N0));
791     const double* f_10 = in.getSampleDataRO(INDEX2(k0+1,m_N1-2, m_N0));
792     const double* f_00 = in.getSampleDataRO(INDEX2(k0,m_N1-2, m_N0));
793 caltinay 3722 double* o = out.getSampleDataRW(m_faceOffset[3]+k0);
794     for (index_t i=0; i < numComp; ++i) {
795 caltinay 3724 o[INDEX3(i,0,0,numComp,2)] = f_01[i]*cx0 + f_11[i]*cx7;
796     o[INDEX3(i,1,0,numComp,2)] = cy5*(f_01[i] + f_11[i]) + cy2*(f_00[i] + f_10[i]);
797 caltinay 3722 } /* end of component loop i */
798     } /* end of k0 loop */
799     } /* end of face 3 */
800     } // end of parallel section
801 caltinay 3702 }
802 caltinay 3701 }
803 caltinay 3697
804 caltinay 3764 //protected
805     void Rectangle::assembleIntegrate(vector<double>& integrals, escript::Data& arg) const
806 caltinay 3713 {
807 caltinay 3764 const dim_t numComp = arg.getDataPointSize();
808 caltinay 3713 const double h0 = m_l0/m_gNE0;
809     const double h1 = m_l1/m_gNE1;
810 caltinay 3764 const index_t left = (m_offset0==0 ? 0 : 1);
811     const index_t bottom = (m_offset1==0 ? 0 : 1);
812 caltinay 3713 if (arg.getFunctionSpace().getTypeCode() == Elements) {
813 caltinay 3764 const double w = h0*h1/4.;
814 caltinay 3713 #pragma omp parallel
815     {
816     vector<double> int_local(numComp, 0);
817 caltinay 3722 #pragma omp for nowait
818 caltinay 3764 for (index_t k1 = bottom; k1 < bottom+m_ownNE1; ++k1) {
819     for (index_t k0 = left; k0 < left+m_ownNE0; ++k0) {
820     const double* f = arg.getSampleDataRO(INDEX2(k0, k1, m_NE0));
821 caltinay 3713 for (index_t i=0; i < numComp; ++i) {
822 caltinay 3769 const double f0 = f[INDEX2(i,0,numComp)];
823     const double f1 = f[INDEX2(i,1,numComp)];
824     const double f2 = f[INDEX2(i,2,numComp)];
825     const double f3 = f[INDEX2(i,3,numComp)];
826 caltinay 3764 int_local[i]+=(f0+f1+f2+f3)*w;
827 caltinay 3713 } /* end of component loop i */
828     } /* end of k0 loop */
829     } /* end of k1 loop */
830    
831     #pragma omp critical
832     for (index_t i=0; i<numComp; i++)
833     integrals[i]+=int_local[i];
834 caltinay 3722 } // end of parallel section
835 caltinay 3713 } else if (arg.getFunctionSpace().getTypeCode() == ReducedElements) {
836 caltinay 3764 const double w = h0*h1;
837 caltinay 3713 #pragma omp parallel
838     {
839     vector<double> int_local(numComp, 0);
840 caltinay 3722 #pragma omp for nowait
841 caltinay 3764 for (index_t k1 = bottom; k1 < bottom+m_ownNE1; ++k1) {
842     for (index_t k0 = left; k0 < left+m_ownNE0; ++k0) {
843     const double* f = arg.getSampleDataRO(INDEX2(k0, k1, m_NE0));
844 caltinay 3713 for (index_t i=0; i < numComp; ++i) {
845 caltinay 3764 int_local[i]+=f[i]*w;
846 caltinay 3713 } /* end of component loop i */
847     } /* end of k0 loop */
848     } /* end of k1 loop */
849    
850     #pragma omp critical
851     for (index_t i=0; i<numComp; i++)
852     integrals[i]+=int_local[i];
853 caltinay 3722 } // end of parallel section
854 caltinay 3713 } else if (arg.getFunctionSpace().getTypeCode() == FaceElements) {
855 caltinay 3764 const double w0 = h0/2.;
856     const double w1 = h1/2.;
857 caltinay 3713 #pragma omp parallel
858     {
859     vector<double> int_local(numComp, 0);
860     if (m_faceOffset[0] > -1) {
861 caltinay 3722 #pragma omp for nowait
862 caltinay 3764 for (index_t k1 = bottom; k1 < bottom+m_ownNE1; ++k1) {
863     const double* f = arg.getSampleDataRO(m_faceOffset[0]+k1);
864 caltinay 3713 for (index_t i=0; i < numComp; ++i) {
865 caltinay 3769 const double f0 = f[INDEX2(i,0,numComp)];
866     const double f1 = f[INDEX2(i,1,numComp)];
867 caltinay 3764 int_local[i]+=(f0+f1)*w1;
868 caltinay 3713 } /* end of component loop i */
869     } /* end of k1 loop */
870     }
871    
872     if (m_faceOffset[1] > -1) {
873 caltinay 3722 #pragma omp for nowait
874 caltinay 3764 for (index_t k1 = bottom; k1 < bottom+m_ownNE1; ++k1) {
875     const double* f = arg.getSampleDataRO(m_faceOffset[1]+k1);
876 caltinay 3713 for (index_t i=0; i < numComp; ++i) {
877 caltinay 3769 const double f0 = f[INDEX2(i,0,numComp)];
878     const double f1 = f[INDEX2(i,1,numComp)];
879 caltinay 3764 int_local[i]+=(f0+f1)*w1;
880 caltinay 3713 } /* end of component loop i */
881     } /* end of k1 loop */
882     }
883    
884     if (m_faceOffset[2] > -1) {
885 caltinay 3722 #pragma omp for nowait
886 caltinay 3764 for (index_t k0 = left; k0 < left+m_ownNE0; ++k0) {
887     const double* f = arg.getSampleDataRO(m_faceOffset[2]+k0);
888 caltinay 3713 for (index_t i=0; i < numComp; ++i) {
889 caltinay 3769 const double f0 = f[INDEX2(i,0,numComp)];
890     const double f1 = f[INDEX2(i,1,numComp)];
891 caltinay 3764 int_local[i]+=(f0+f1)*w0;
892 caltinay 3713 } /* end of component loop i */
893     } /* end of k0 loop */
894     }
895    
896     if (m_faceOffset[3] > -1) {
897 caltinay 3722 #pragma omp for nowait
898 caltinay 3764 for (index_t k0 = left; k0 < left+m_ownNE0; ++k0) {
899     const double* f = arg.getSampleDataRO(m_faceOffset[3]+k0);
900 caltinay 3713 for (index_t i=0; i < numComp; ++i) {
901 caltinay 3769 const double f0 = f[INDEX2(i,0,numComp)];
902     const double f1 = f[INDEX2(i,1,numComp)];
903 caltinay 3764 int_local[i]+=(f0+f1)*w0;
904 caltinay 3713 } /* end of component loop i */
905     } /* end of k0 loop */
906     }
907    
908     #pragma omp critical
909     for (index_t i=0; i<numComp; i++)
910     integrals[i]+=int_local[i];
911 caltinay 3722 } // end of parallel section
912 caltinay 3713 } else if (arg.getFunctionSpace().getTypeCode() == ReducedFaceElements) {
913     #pragma omp parallel
914     {
915     vector<double> int_local(numComp, 0);
916     if (m_faceOffset[0] > -1) {
917 caltinay 3722 #pragma omp for nowait
918 caltinay 3764 for (index_t k1 = bottom; k1 < bottom+m_ownNE1; ++k1) {
919     const double* f = arg.getSampleDataRO(m_faceOffset[0]+k1);
920 caltinay 3713 for (index_t i=0; i < numComp; ++i) {
921     int_local[i]+=f[i]*h1;
922     } /* end of component loop i */
923     } /* end of k1 loop */
924     }
925    
926     if (m_faceOffset[1] > -1) {
927 caltinay 3722 #pragma omp for nowait
928 caltinay 3764 for (index_t k1 = bottom; k1 < bottom+m_ownNE1; ++k1) {
929     const double* f = arg.getSampleDataRO(m_faceOffset[1]+k1);
930 caltinay 3713 for (index_t i=0; i < numComp; ++i) {
931     int_local[i]+=f[i]*h1;
932     } /* end of component loop i */
933     } /* end of k1 loop */
934     }
935    
936     if (m_faceOffset[2] > -1) {
937 caltinay 3722 #pragma omp for nowait
938 caltinay 3764 for (index_t k0 = left; k0 < left+m_ownNE0; ++k0) {
939     const double* f = arg.getSampleDataRO(m_faceOffset[2]+k0);
940 caltinay 3713 for (index_t i=0; i < numComp; ++i) {
941     int_local[i]+=f[i]*h0;
942     } /* end of component loop i */
943     } /* end of k0 loop */
944     }
945    
946     if (m_faceOffset[3] > -1) {
947 caltinay 3722 #pragma omp for nowait
948 caltinay 3764 for (index_t k0 = left; k0 < left+m_ownNE0; ++k0) {
949     const double* f = arg.getSampleDataRO(m_faceOffset[3]+k0);
950 caltinay 3713 for (index_t i=0; i < numComp; ++i) {
951     int_local[i]+=f[i]*h0;
952     } /* end of component loop i */
953     } /* end of k0 loop */
954     }
955    
956     #pragma omp critical
957     for (index_t i=0; i<numComp; i++)
958     integrals[i]+=int_local[i];
959 caltinay 3722 } // end of parallel section
960 caltinay 3713 }
961     }
962    
963 caltinay 3691 //protected
964 caltinay 3756 dim_t Rectangle::insertNeighbourNodes(IndexVector& index, index_t node) const
965     {
966     const dim_t nDOF0 = (m_gNE0+1)/m_NX;
967     const dim_t nDOF1 = (m_gNE1+1)/m_NY;
968     const int x=node%nDOF0;
969     const int y=node/nDOF0;
970     dim_t num=0;
971     // loop through potential neighbours and add to index if positions are
972     // within bounds
973     for (int i1=-1; i1<2; i1++) {
974     for (int i0=-1; i0<2; i0++) {
975     // skip node itself
976     if (i0==0 && i1==0)
977     continue;
978     // location of neighbour node
979     const int nx=x+i0;
980     const int ny=y+i1;
981     if (nx>=0 && ny>=0 && nx<nDOF0 && ny<nDOF1) {
982     index.push_back(ny*nDOF0+nx);
983     num++;
984     }
985     }
986     }
987    
988     return num;
989     }
990    
991     //protected
992     void Rectangle::nodesToDOF(escript::Data& out, escript::Data& in) const
993     {
994     const dim_t numComp = in.getDataPointSize();
995     out.requireWrite();
996    
997     const index_t left = (m_offset0==0 ? 0 : 1);
998     const index_t bottom = (m_offset1==0 ? 0 : 1);
999     const dim_t nDOF0 = (m_gNE0+1)/m_NX;
1000     const dim_t nDOF1 = (m_gNE1+1)/m_NY;
1001     #pragma omp parallel for
1002     for (index_t i=0; i<nDOF1; i++) {
1003     for (index_t j=0; j<nDOF0; j++) {
1004     const index_t n=j+left+(i+bottom)*m_N0;
1005     const double* src=in.getSampleDataRO(n);
1006     copy(src, src+numComp, out.getSampleDataRW(j+i*nDOF0));
1007     }
1008     }
1009     }
1010    
1011     //protected
1012     void Rectangle::dofToNodes(escript::Data& out, escript::Data& in) const
1013     {
1014     const dim_t numComp = in.getDataPointSize();
1015     Paso_Coupler* coupler = Paso_Coupler_alloc(m_connector, numComp);
1016     in.requireWrite();
1017     Paso_Coupler_startCollect(coupler, in.getSampleDataRW(0));
1018    
1019     const dim_t numDOF = getNumDOF();
1020     out.requireWrite();
1021     const double* buffer = Paso_Coupler_finishCollect(coupler);
1022    
1023     #pragma omp parallel for
1024     for (index_t i=0; i<getNumNodes(); i++) {
1025     const double* src=(m_dofMap[i]<numDOF ?
1026     in.getSampleDataRO(m_dofMap[i])
1027     : &buffer[(m_dofMap[i]-numDOF)*numComp]);
1028     copy(src, src+numComp, out.getSampleDataRW(i));
1029     }
1030     }
1031    
1032 caltinay 3691 //private
1033     void Rectangle::populateSampleIds()
1034     {
1035 caltinay 3752 // identifiers are ordered from left to right, bottom to top globablly.
1036 caltinay 3697
1037     // build node distribution vector first.
1038     // rank i owns m_nodeDistribution[i+1]-nodeDistribution[i] nodes
1039     m_nodeDistribution.assign(m_mpiInfo->size+1, 0);
1040 caltinay 3752 const dim_t numDOF=getNumDOF();
1041     for (dim_t k=1; k<m_mpiInfo->size; k++) {
1042     m_nodeDistribution[k]=k*numDOF;
1043 caltinay 3697 }
1044     m_nodeDistribution[m_mpiInfo->size]=getNumDataPointsGlobal();
1045 caltinay 3691 m_nodeId.resize(getNumNodes());
1046 caltinay 3753 m_dofId.resize(numDOF);
1047     m_elementId.resize(getNumElements());
1048     m_faceId.resize(getNumFaceElements());
1049 caltinay 3697
1050 caltinay 3753 #pragma omp parallel
1051     {
1052     // nodes
1053     #pragma omp for nowait
1054     for (dim_t i1=0; i1<m_N1; i1++) {
1055     for (dim_t i0=0; i0<m_N0; i0++) {
1056     m_nodeId[i0+i1*m_N0] = (m_offset1+i1)*(m_gNE0+1)+m_offset0+i0;
1057     }
1058 caltinay 3697 }
1059    
1060 caltinay 3753 // degrees of freedom
1061     #pragma omp for nowait
1062     for (dim_t k=0; k<numDOF; k++)
1063     m_dofId[k] = m_nodeDistribution[m_mpiInfo->rank]+k;
1064 caltinay 3752
1065 caltinay 3753 // elements
1066     #pragma omp for nowait
1067 caltinay 3755 for (dim_t i1=0; i1<m_NE1; i1++) {
1068     for (dim_t i0=0; i0<m_NE0; i0++) {
1069     m_elementId[i0+i1*m_NE0]=(m_offset1+i1)*m_gNE0+m_offset0+i0;
1070     }
1071     }
1072 caltinay 3753
1073     // face elements
1074     #pragma omp for
1075     for (dim_t k=0; k<getNumFaceElements(); k++)
1076     m_faceId[k]=k;
1077     } // end parallel section
1078    
1079 caltinay 3735 m_nodeTags.assign(getNumNodes(), 0);
1080     updateTagsInUse(Nodes);
1081 caltinay 3697
1082 caltinay 3735 m_elementTags.assign(getNumElements(), 0);
1083     updateTagsInUse(Elements);
1084 caltinay 3697
1085 caltinay 3722 // generate face offset vector and set face tags
1086 caltinay 3704 const IndexVector facesPerEdge = getNumFacesPerBoundary();
1087 caltinay 3722 const index_t LEFT=1, RIGHT=2, BOTTOM=10, TOP=20;
1088     const index_t faceTag[] = { LEFT, RIGHT, BOTTOM, TOP };
1089 caltinay 3704 m_faceOffset.assign(facesPerEdge.size(), -1);
1090 caltinay 3722 m_faceTags.clear();
1091 caltinay 3704 index_t offset=0;
1092     for (size_t i=0; i<facesPerEdge.size(); i++) {
1093     if (facesPerEdge[i]>0) {
1094     m_faceOffset[i]=offset;
1095     offset+=facesPerEdge[i];
1096 caltinay 3722 m_faceTags.insert(m_faceTags.end(), facesPerEdge[i], faceTag[i]);
1097 caltinay 3704 }
1098     }
1099 caltinay 3722 setTagMap("left", LEFT);
1100     setTagMap("right", RIGHT);
1101     setTagMap("bottom", BOTTOM);
1102     setTagMap("top", TOP);
1103     updateTagsInUse(FaceElements);
1104 caltinay 3691 }
1105    
1106 caltinay 3699 //private
1107 caltinay 3756 void Rectangle::createPattern()
1108 caltinay 3699 {
1109 caltinay 3756 const dim_t nDOF0 = (m_gNE0+1)/m_NX;
1110     const dim_t nDOF1 = (m_gNE1+1)/m_NY;
1111     const index_t left = (m_offset0==0 ? 0 : 1);
1112     const index_t bottom = (m_offset1==0 ? 0 : 1);
1113    
1114     // populate node->DOF mapping with own degrees of freedom.
1115     // The rest is assigned in the loop further down
1116     m_dofMap.assign(getNumNodes(), 0);
1117     #pragma omp parallel for
1118 caltinay 3766 for (index_t i=bottom; i<bottom+nDOF1; i++) {
1119     for (index_t j=left; j<left+nDOF0; j++) {
1120 caltinay 3756 m_dofMap[i*m_N0+j]=(i-bottom)*nDOF0+j-left;
1121     }
1122     }
1123    
1124     // build list of shared components and neighbours by looping through
1125     // all potential neighbouring ranks and checking if positions are
1126 caltinay 3754 // within bounds
1127 caltinay 3756 const dim_t numDOF=nDOF0*nDOF1;
1128     vector<IndexVector> colIndices(numDOF); // for the couple blocks
1129     RankVector neighbour;
1130     IndexVector offsetInShared(1,0);
1131     IndexVector sendShared, recvShared;
1132     int numShared=0;
1133     const int x=m_mpiInfo->rank%m_NX;
1134     const int y=m_mpiInfo->rank/m_NX;
1135 caltinay 3754 for (int i1=-1; i1<2; i1++) {
1136     for (int i0=-1; i0<2; i0++) {
1137 caltinay 3756 // skip this rank
1138 caltinay 3754 if (i0==0 && i1==0)
1139     continue;
1140 caltinay 3756 // location of neighbour rank
1141 caltinay 3754 const int nx=x+i0;
1142     const int ny=y+i1;
1143 caltinay 3756 if (nx>=0 && ny>=0 && nx<m_NX && ny<m_NY) {
1144     neighbour.push_back(ny*m_NX+nx);
1145     if (i0==0) {
1146     // sharing top or bottom edge
1147     const int firstDOF=(i1==-1 ? 0 : numDOF-nDOF0);
1148     const int firstNode=(i1==-1 ? left : m_N0*(m_N1-1)+left);
1149     offsetInShared.push_back(offsetInShared.back()+nDOF0);
1150     for (dim_t i=0; i<nDOF0; i++, numShared++) {
1151     sendShared.push_back(firstDOF+i);
1152     recvShared.push_back(numDOF+numShared);
1153     if (i>0)
1154     colIndices[firstDOF+i-1].push_back(numShared);
1155     colIndices[firstDOF+i].push_back(numShared);
1156     if (i<nDOF0-1)
1157     colIndices[firstDOF+i+1].push_back(numShared);
1158     m_dofMap[firstNode+i]=numDOF+numShared;
1159     }
1160     } else if (i1==0) {
1161     // sharing left or right edge
1162     const int firstDOF=(i0==-1 ? 0 : nDOF0-1);
1163     const int firstNode=(i0==-1 ? bottom*m_N0 : (bottom+1)*m_N0-1);
1164     offsetInShared.push_back(offsetInShared.back()+nDOF1);
1165     for (dim_t i=0; i<nDOF1; i++, numShared++) {
1166     sendShared.push_back(firstDOF+i*nDOF0);
1167     recvShared.push_back(numDOF+numShared);
1168     if (i>0)
1169     colIndices[firstDOF+(i-1)*nDOF0].push_back(numShared);
1170     colIndices[firstDOF+i*nDOF0].push_back(numShared);
1171     if (i<nDOF1-1)
1172     colIndices[firstDOF+(i+1)*nDOF0].push_back(numShared);
1173     m_dofMap[firstNode+i*m_N0]=numDOF+numShared;
1174     }
1175     } else {
1176     // sharing a node
1177     const int dof=(i0+1)/2*(nDOF0-1)+(i1+1)/2*(numDOF-nDOF0);
1178     const int node=(i0+1)/2*(m_N0-1)+(i1+1)/2*m_N0*(m_N1-1);
1179     offsetInShared.push_back(offsetInShared.back()+1);
1180     sendShared.push_back(dof);
1181     recvShared.push_back(numDOF+numShared);
1182     colIndices[dof].push_back(numShared);
1183     m_dofMap[node]=numDOF+numShared;
1184     ++numShared;
1185     }
1186 caltinay 3754 }
1187 caltinay 3699 }
1188     }
1189 caltinay 3754
1190 caltinay 3756 // create connector
1191     Paso_SharedComponents *snd_shcomp = Paso_SharedComponents_alloc(
1192     numDOF, neighbour.size(), &neighbour[0], &sendShared[0],
1193     &offsetInShared[0], 1, 0, m_mpiInfo);
1194     Paso_SharedComponents *rcv_shcomp = Paso_SharedComponents_alloc(
1195     numDOF, neighbour.size(), &neighbour[0], &recvShared[0],
1196     &offsetInShared[0], 1, 0, m_mpiInfo);
1197     m_connector = Paso_Connector_alloc(snd_shcomp, rcv_shcomp);
1198     Paso_SharedComponents_free(snd_shcomp);
1199     Paso_SharedComponents_free(rcv_shcomp);
1200 caltinay 3754
1201 caltinay 3756 // create main and couple blocks
1202     Paso_Pattern *mainPattern = createMainPattern();
1203     Paso_Pattern *colPattern, *rowPattern;
1204     createCouplePatterns(colIndices, numShared, &colPattern, &rowPattern);
1205 caltinay 3754
1206 caltinay 3756 // allocate paso distribution
1207     Paso_Distribution* distribution = Paso_Distribution_alloc(m_mpiInfo,
1208     const_cast<index_t*>(&m_nodeDistribution[0]), 1, 0);
1209 caltinay 3755
1210 caltinay 3756 // finally create the system matrix
1211     m_pattern = Paso_SystemMatrixPattern_alloc(MATRIX_FORMAT_DEFAULT,
1212     distribution, distribution, mainPattern, colPattern, rowPattern,
1213     m_connector, m_connector);
1214 caltinay 3755
1215 caltinay 3756 Paso_Distribution_free(distribution);
1216 caltinay 3755
1217 caltinay 3756 // useful debug output
1218     /*
1219     cout << "--- rcv_shcomp ---" << endl;
1220     cout << "numDOF=" << numDOF << ", numNeighbors=" << neighbour.size() << endl;
1221     for (size_t i=0; i<neighbour.size(); i++) {
1222     cout << "neighbor[" << i << "]=" << neighbour[i]
1223     << " offsetInShared[" << i+1 << "]=" << offsetInShared[i+1] << endl;
1224 caltinay 3699 }
1225 caltinay 3756 for (size_t i=0; i<recvShared.size(); i++) {
1226     cout << "shared[" << i << "]=" << recvShared[i] << endl;
1227     }
1228     cout << "--- snd_shcomp ---" << endl;
1229     for (size_t i=0; i<sendShared.size(); i++) {
1230     cout << "shared[" << i << "]=" << sendShared[i] << endl;
1231     }
1232     cout << "--- dofMap ---" << endl;
1233     for (size_t i=0; i<m_dofMap.size(); i++) {
1234     cout << "m_dofMap[" << i << "]=" << m_dofMap[i] << endl;
1235     }
1236     cout << "--- colIndices ---" << endl;
1237     for (size_t i=0; i<colIndices.size(); i++) {
1238     cout << "colIndices[" << i << "].size()=" << colIndices[i].size() << endl;
1239     }
1240     */
1241 caltinay 3754
1242 caltinay 3756 /*
1243     cout << "--- main_pattern ---" << endl;
1244     cout << "M=" << mainPattern->numOutput << ", N=" << mainPattern->numInput << endl;
1245     for (size_t i=0; i<mainPattern->numOutput+1; i++) {
1246     cout << "ptr[" << i << "]=" << mainPattern->ptr[i] << endl;
1247     }
1248     for (size_t i=0; i<mainPattern->ptr[mainPattern->numOutput]; i++) {
1249     cout << "index[" << i << "]=" << mainPattern->index[i] << endl;
1250     }
1251     */
1252 caltinay 3754
1253 caltinay 3756 /*
1254     cout << "--- colCouple_pattern ---" << endl;
1255     cout << "M=" << colPattern->numOutput << ", N=" << colPattern->numInput << endl;
1256     for (size_t i=0; i<colPattern->numOutput+1; i++) {
1257     cout << "ptr[" << i << "]=" << colPattern->ptr[i] << endl;
1258     }
1259     for (size_t i=0; i<colPattern->ptr[colPattern->numOutput]; i++) {
1260     cout << "index[" << i << "]=" << colPattern->index[i] << endl;
1261     }
1262     */
1263 caltinay 3754
1264 caltinay 3756 /*
1265     cout << "--- rowCouple_pattern ---" << endl;
1266     cout << "M=" << rowPattern->numOutput << ", N=" << rowPattern->numInput << endl;
1267     for (size_t i=0; i<rowPattern->numOutput+1; i++) {
1268     cout << "ptr[" << i << "]=" << rowPattern->ptr[i] << endl;
1269 caltinay 3699 }
1270 caltinay 3756 for (size_t i=0; i<rowPattern->ptr[rowPattern->numOutput]; i++) {
1271     cout << "index[" << i << "]=" << rowPattern->index[i] << endl;
1272     }
1273     */
1274    
1275     Paso_Pattern_free(mainPattern);
1276     Paso_Pattern_free(colPattern);
1277     Paso_Pattern_free(rowPattern);
1278 caltinay 3699 }
1279    
1280 caltinay 3776 //private
1281     void Rectangle::addToMatrixAndRHS(Paso_SystemMatrix* S, escript::Data& F,
1282     const vector<double>& EM_S, const vector<double>& EM_F, bool addS,
1283     bool addF, index_t firstNode, dim_t nEq, dim_t nComp) const
1284     {
1285     IndexVector rowIndex;
1286     rowIndex.push_back(m_dofMap[firstNode]);
1287     rowIndex.push_back(m_dofMap[firstNode+1]);
1288     rowIndex.push_back(m_dofMap[firstNode+m_N0]);
1289     rowIndex.push_back(m_dofMap[firstNode+m_N0+1]);
1290     if (addF) {
1291     double *F_p=F.getSampleDataRW(0);
1292     for (index_t i=0; i<rowIndex.size(); i++) {
1293     if (rowIndex[i]<getNumDOF()) {
1294     for (index_t eq=0; eq<nEq; eq++) {
1295     F_p[INDEX2(eq, rowIndex[i], nEq)]+=EM_F[INDEX2(eq,i,nEq)];
1296     }
1297     }
1298     }
1299     }
1300     if (addS) {
1301     addToSystemMatrix(S, rowIndex, nEq, rowIndex, nComp, EM_S);
1302     }
1303     }
1304    
1305 caltinay 3702 //protected
1306 caltinay 3711 void Rectangle::interpolateNodesOnElements(escript::Data& out,
1307     escript::Data& in, bool reduced) const
1308 caltinay 3702 {
1309     const dim_t numComp = in.getDataPointSize();
1310 caltinay 3711 if (reduced) {
1311 caltinay 3760 out.requireWrite();
1312 caltinay 3724 const double c0 = .25;
1313 caltinay 3702 #pragma omp parallel for
1314 caltinay 3711 for (index_t k1=0; k1 < m_NE1; ++k1) {
1315     for (index_t k0=0; k0 < m_NE0; ++k0) {
1316 caltinay 3769 const double* f_11 = in.getSampleDataRO(INDEX2(k0+1,k1+1, m_N0));
1317     const double* f_10 = in.getSampleDataRO(INDEX2(k0+1,k1, m_N0));
1318     const double* f_00 = in.getSampleDataRO(INDEX2(k0,k1, m_N0));
1319     const double* f_01 = in.getSampleDataRO(INDEX2(k0,k1+1, m_N0));
1320 caltinay 3711 double* o = out.getSampleDataRW(INDEX2(k0,k1,m_NE0));
1321     for (index_t i=0; i < numComp; ++i) {
1322 caltinay 3724 o[INDEX2(i,numComp,0)] = c0*(f_00[i] + f_01[i] + f_10[i] + f_11[i]);
1323 caltinay 3711 } /* end of component loop i */
1324     } /* end of k0 loop */
1325     } /* end of k1 loop */
1326     } else {
1327 caltinay 3760 out.requireWrite();
1328 caltinay 3724 const double c0 = .16666666666666666667;
1329     const double c1 = .044658198738520451079;
1330     const double c2 = .62200846792814621559;
1331 caltinay 3711 #pragma omp parallel for
1332     for (index_t k1=0; k1 < m_NE1; ++k1) {
1333     for (index_t k0=0; k0 < m_NE0; ++k0) {
1334 caltinay 3769 const double* f_10 = in.getSampleDataRO(INDEX2(k0+1,k1, m_N0));
1335     const double* f_11 = in.getSampleDataRO(INDEX2(k0+1,k1+1, m_N0));
1336     const double* f_01 = in.getSampleDataRO(INDEX2(k0,k1+1, m_N0));
1337     const double* f_00 = in.getSampleDataRO(INDEX2(k0,k1, m_N0));
1338 caltinay 3711 double* o = out.getSampleDataRW(INDEX2(k0,k1,m_NE0));
1339     for (index_t i=0; i < numComp; ++i) {
1340 caltinay 3724 o[INDEX2(i,numComp,0)] = f_00[i]*c2 + f_11[i]*c1 + c0*(f_01[i] + f_10[i]);
1341     o[INDEX2(i,numComp,1)] = f_01[i]*c1 + f_10[i]*c2 + c0*(f_00[i] + f_11[i]);
1342     o[INDEX2(i,numComp,2)] = f_01[i]*c2 + f_10[i]*c1 + c0*(f_00[i] + f_11[i]);
1343     o[INDEX2(i,numComp,3)] = f_00[i]*c1 + f_11[i]*c2 + c0*(f_01[i] + f_10[i]);
1344 caltinay 3711 } /* end of component loop i */
1345     } /* end of k0 loop */
1346     } /* end of k1 loop */
1347     }
1348 caltinay 3702 }
1349    
1350     //protected
1351 caltinay 3711 void Rectangle::interpolateNodesOnFaces(escript::Data& out, escript::Data& in,
1352     bool reduced) const
1353 caltinay 3702 {
1354 caltinay 3704 const dim_t numComp = in.getDataPointSize();
1355 caltinay 3711 if (reduced) {
1356 caltinay 3760 out.requireWrite();
1357 caltinay 3724 const double c0 = .5;
1358     #pragma omp parallel
1359     {
1360     if (m_faceOffset[0] > -1) {
1361     #pragma omp for nowait
1362     for (index_t k1=0; k1 < m_NE1; ++k1) {
1363 caltinay 3769 const double* f_00 = in.getSampleDataRO(INDEX2(0,k1, m_N0));
1364     const double* f_01 = in.getSampleDataRO(INDEX2(0,k1+1, m_N0));
1365 caltinay 3724 double* o = out.getSampleDataRW(m_faceOffset[0]+k1);
1366     for (index_t i=0; i < numComp; ++i) {
1367     o[INDEX2(i,numComp,0)] = c0*(f_00[i] + f_01[i]);
1368     } /* end of component loop i */
1369     } /* end of k1 loop */
1370     } /* end of face 0 */
1371     if (m_faceOffset[1] > -1) {
1372     #pragma omp for nowait
1373     for (index_t k1=0; k1 < m_NE1; ++k1) {
1374 caltinay 3769 const double* f_10 = in.getSampleDataRO(INDEX2(m_N0-1,k1, m_N0));
1375     const double* f_11 = in.getSampleDataRO(INDEX2(m_N0-1,k1+1, m_N0));
1376 caltinay 3724 double* o = out.getSampleDataRW(m_faceOffset[1]+k1);
1377     for (index_t i=0; i < numComp; ++i) {
1378     o[INDEX2(i,numComp,0)] = c0*(f_10[i] + f_11[i]);
1379     } /* end of component loop i */
1380     } /* end of k1 loop */
1381     } /* end of face 1 */
1382     if (m_faceOffset[2] > -1) {
1383     #pragma omp for nowait
1384     for (index_t k0=0; k0 < m_NE0; ++k0) {
1385 caltinay 3769 const double* f_10 = in.getSampleDataRO(INDEX2(k0+1,0, m_N0));
1386     const double* f_00 = in.getSampleDataRO(INDEX2(k0,0, m_N0));
1387 caltinay 3724 double* o = out.getSampleDataRW(m_faceOffset[2]+k0);
1388     for (index_t i=0; i < numComp; ++i) {
1389     o[INDEX2(i,numComp,0)] = c0*(f_00[i] + f_10[i]);
1390     } /* end of component loop i */
1391     } /* end of k0 loop */
1392     } /* end of face 2 */
1393     if (m_faceOffset[3] > -1) {
1394     #pragma omp for nowait
1395     for (index_t k0=0; k0 < m_NE0; ++k0) {
1396 caltinay 3769 const double* f_01 = in.getSampleDataRO(INDEX2(k0,m_N1-1, m_N0));
1397     const double* f_11 = in.getSampleDataRO(INDEX2(k0+1,m_N1-1, m_N0));
1398 caltinay 3724 double* o = out.getSampleDataRW(m_faceOffset[3]+k0);
1399     for (index_t i=0; i < numComp; ++i) {
1400     o[INDEX2(i,numComp,0)] = c0*(f_01[i] + f_11[i]);
1401     } /* end of component loop i */
1402     } /* end of k0 loop */
1403     } /* end of face 3 */
1404     } // end of parallel section
1405 caltinay 3711 } else {
1406 caltinay 3760 out.requireWrite();
1407 caltinay 3724 const double c0 = 0.21132486540518711775;
1408     const double c1 = 0.78867513459481288225;
1409     #pragma omp parallel
1410     {
1411     if (m_faceOffset[0] > -1) {
1412     #pragma omp for nowait
1413     for (index_t k1=0; k1 < m_NE1; ++k1) {
1414 caltinay 3769 const double* f_01 = in.getSampleDataRO(INDEX2(0,k1+1, m_N0));
1415     const double* f_00 = in.getSampleDataRO(INDEX2(0,k1, m_N0));
1416 caltinay 3724 double* o = out.getSampleDataRW(m_faceOffset[0]+k1);
1417     for (index_t i=0; i < numComp; ++i) {
1418     o[INDEX2(i,numComp,0)] = f_00[i]*c1 + f_01[i]*c0;
1419     o[INDEX2(i,numComp,1)] = f_00[i]*c0 + f_01[i]*c1;
1420     } /* end of component loop i */
1421     } /* end of k1 loop */
1422     } /* end of face 0 */
1423     if (m_faceOffset[1] > -1) {
1424     #pragma omp for nowait
1425     for (index_t k1=0; k1 < m_NE1; ++k1) {
1426 caltinay 3769 const double* f_10 = in.getSampleDataRO(INDEX2(m_N0-1,k1, m_N0));
1427     const double* f_11 = in.getSampleDataRO(INDEX2(m_N0-1,k1+1, m_N0));
1428 caltinay 3724 double* o = out.getSampleDataRW(m_faceOffset[1]+k1);
1429     for (index_t i=0; i < numComp; ++i) {
1430     o[INDEX2(i,numComp,0)] = f_10[i]*c1 + f_11[i]*c0;
1431     o[INDEX2(i,numComp,1)] = f_10[i]*c0 + f_11[i]*c1;
1432     } /* end of component loop i */
1433     } /* end of k1 loop */
1434     } /* end of face 1 */
1435     if (m_faceOffset[2] > -1) {
1436     #pragma omp for nowait
1437     for (index_t k0=0; k0 < m_NE0; ++k0) {
1438 caltinay 3769 const double* f_10 = in.getSampleDataRO(INDEX2(k0+1,0, m_N0));
1439     const double* f_00 = in.getSampleDataRO(INDEX2(k0,0, m_N0));
1440 caltinay 3724 double* o = out.getSampleDataRW(m_faceOffset[2]+k0);
1441     for (index_t i=0; i < numComp; ++i) {
1442     o[INDEX2(i,numComp,0)] = f_00[i]*c1 + f_10[i]*c0;
1443     o[INDEX2(i,numComp,1)] = f_00[i]*c0 + f_10[i]*c1;
1444     } /* end of component loop i */
1445     } /* end of k0 loop */
1446     } /* end of face 2 */
1447     if (m_faceOffset[3] > -1) {
1448     #pragma omp for nowait
1449     for (index_t k0=0; k0 < m_NE0; ++k0) {
1450 caltinay 3769 const double* f_11 = in.getSampleDataRO(INDEX2(k0+1,m_N1-1, m_N0));
1451     const double* f_01 = in.getSampleDataRO(INDEX2(k0,m_N1-1, m_N0));
1452 caltinay 3724 double* o = out.getSampleDataRW(m_faceOffset[3]+k0);
1453     for (index_t i=0; i < numComp; ++i) {
1454     o[INDEX2(i,numComp,0)] = f_01[i]*c1 + f_11[i]*c0;
1455     o[INDEX2(i,numComp,1)] = f_01[i]*c0 + f_11[i]*c1;
1456     } /* end of component loop i */
1457     } /* end of k0 loop */
1458     } /* end of face 3 */
1459     } // end of parallel section
1460 caltinay 3711 }
1461 caltinay 3702 }
1462    
1463 caltinay 3748 //protected
1464     void Rectangle::assemblePDESingle(Paso_SystemMatrix* mat,
1465     escript::Data& rhs, const escript::Data& A, const escript::Data& B,
1466     const escript::Data& C, const escript::Data& D,
1467 caltinay 3769 const escript::Data& X, const escript::Data& Y) const
1468 caltinay 3748 {
1469     const double h0 = m_l0/m_gNE0;
1470     const double h1 = m_l1/m_gNE1;
1471     const double w0 = -0.1555021169820365539*h1/h0;
1472     const double w1 = 0.041666666666666666667;
1473 caltinay 3764 const double w2 = -0.15550211698203655390;
1474     const double w3 = 0.041666666666666666667*h0/h1;
1475     const double w4 = 0.15550211698203655390;
1476     const double w5 = -0.041666666666666666667;
1477     const double w6 = -0.01116454968463011277*h1/h0;
1478     const double w7 = 0.011164549684630112770;
1479     const double w8 = -0.011164549684630112770;
1480     const double w9 = -0.041666666666666666667*h1/h0;
1481 caltinay 3748 const double w10 = -0.041666666666666666667*h0/h1;
1482     const double w11 = 0.1555021169820365539*h1/h0;
1483     const double w12 = 0.1555021169820365539*h0/h1;
1484     const double w13 = 0.01116454968463011277*h0/h1;
1485     const double w14 = 0.01116454968463011277*h1/h0;
1486     const double w15 = 0.041666666666666666667*h1/h0;
1487     const double w16 = -0.01116454968463011277*h0/h1;
1488     const double w17 = -0.1555021169820365539*h0/h1;
1489     const double w18 = -0.33333333333333333333*h1/h0;
1490 caltinay 3769 const double w19 = 0.25;
1491     const double w20 = -0.25;
1492 caltinay 3748 const double w21 = 0.16666666666666666667*h0/h1;
1493     const double w22 = -0.16666666666666666667*h1/h0;
1494     const double w23 = -0.16666666666666666667*h0/h1;
1495     const double w24 = 0.33333333333333333333*h1/h0;
1496     const double w25 = 0.33333333333333333333*h0/h1;
1497     const double w26 = 0.16666666666666666667*h1/h0;
1498     const double w27 = -0.33333333333333333333*h0/h1;
1499     const double w28 = -0.032861463941450536761*h1;
1500     const double w29 = -0.032861463941450536761*h0;
1501     const double w30 = -0.12264065304058601714*h1;
1502     const double w31 = -0.0023593469594139828636*h1;
1503     const double w32 = -0.008805202725216129906*h0;
1504     const double w33 = -0.008805202725216129906*h1;
1505     const double w34 = 0.032861463941450536761*h1;
1506     const double w35 = 0.008805202725216129906*h1;
1507     const double w36 = 0.008805202725216129906*h0;
1508     const double w37 = 0.0023593469594139828636*h1;
1509     const double w38 = 0.12264065304058601714*h1;
1510     const double w39 = 0.032861463941450536761*h0;
1511     const double w40 = -0.12264065304058601714*h0;
1512     const double w41 = -0.0023593469594139828636*h0;
1513     const double w42 = 0.0023593469594139828636*h0;
1514     const double w43 = 0.12264065304058601714*h0;
1515     const double w44 = -0.16666666666666666667*h1;
1516     const double w45 = -0.083333333333333333333*h0;
1517     const double w46 = 0.083333333333333333333*h1;
1518     const double w47 = 0.16666666666666666667*h1;
1519     const double w48 = 0.083333333333333333333*h0;
1520     const double w49 = -0.16666666666666666667*h0;
1521     const double w50 = 0.16666666666666666667*h0;
1522     const double w51 = -0.083333333333333333333*h1;
1523     const double w52 = 0.025917019497006092316*h0*h1;
1524     const double w53 = 0.0018607582807716854616*h0*h1;
1525     const double w54 = 0.0069444444444444444444*h0*h1;
1526     const double w55 = 0.09672363354357992482*h0*h1;
1527     const double w56 = 0.00049858867864229740201*h0*h1;
1528     const double w57 = 0.055555555555555555556*h0*h1;
1529     const double w58 = 0.027777777777777777778*h0*h1;
1530     const double w59 = 0.11111111111111111111*h0*h1;
1531     const double w60 = -0.19716878364870322056*h1;
1532     const double w61 = -0.19716878364870322056*h0;
1533     const double w62 = -0.052831216351296779436*h0;
1534     const double w63 = -0.052831216351296779436*h1;
1535     const double w64 = 0.19716878364870322056*h1;
1536     const double w65 = 0.052831216351296779436*h1;
1537     const double w66 = 0.19716878364870322056*h0;
1538     const double w67 = 0.052831216351296779436*h0;
1539     const double w68 = -0.5*h1;
1540     const double w69 = -0.5*h0;
1541     const double w70 = 0.5*h1;
1542     const double w71 = 0.5*h0;
1543     const double w72 = 0.1555021169820365539*h0*h1;
1544     const double w73 = 0.041666666666666666667*h0*h1;
1545     const double w74 = 0.01116454968463011277*h0*h1;
1546     const double w75 = 0.25*h0*h1;
1547    
1548     rhs.requireWrite();
1549     #pragma omp parallel
1550     {
1551 caltinay 3754 for (index_t k1_0=0; k1_0<2; k1_0++) { // colouring
1552 caltinay 3748 #pragma omp for
1553     for (index_t k1=k1_0; k1<m_NE1; k1+=2) {
1554     for (index_t k0=0; k0<m_NE0; ++k0) {
1555     bool add_EM_S=false;
1556     bool add_EM_F=false;
1557     vector<double> EM_S(4*4, 0);
1558     vector<double> EM_F(4, 0);
1559     const index_t e = k0 + m_NE0*k1;
1560     ///////////////
1561     // process A //
1562     ///////////////
1563     if (!A.isEmpty()) {
1564     add_EM_S=true;
1565     const double* A_p=const_cast<escript::Data*>(&A)->getSampleDataRO(e);
1566     if (A.actsExpanded()) {
1567 caltinay 3769 const double A_00_0 = A_p[INDEX3(0,0,0,2,2)];
1568     const double A_10_0 = A_p[INDEX3(1,0,0,2,2)];
1569     const double A_01_0 = A_p[INDEX3(0,1,0,2,2)];
1570     const double A_11_0 = A_p[INDEX3(1,1,0,2,2)];
1571     const double A_00_1 = A_p[INDEX3(0,0,1,2,2)];
1572     const double A_10_1 = A_p[INDEX3(1,0,1,2,2)];
1573     const double A_01_1 = A_p[INDEX3(0,1,1,2,2)];
1574     const double A_11_1 = A_p[INDEX3(1,1,1,2,2)];
1575     const double A_00_2 = A_p[INDEX3(0,0,2,2,2)];
1576     const double A_10_2 = A_p[INDEX3(1,0,2,2,2)];
1577     const double A_01_2 = A_p[INDEX3(0,1,2,2,2)];
1578     const double A_11_2 = A_p[INDEX3(1,1,2,2,2)];
1579     const double A_00_3 = A_p[INDEX3(0,0,3,2,2)];
1580     const double A_10_3 = A_p[INDEX3(1,0,3,2,2)];
1581     const double A_01_3 = A_p[INDEX3(0,1,3,2,2)];
1582     const double A_11_3 = A_p[INDEX3(1,1,3,2,2)];
1583     const double tmp0_0 = A_01_0 + A_01_3;
1584     const double tmp1_0 = A_00_0 + A_00_1;
1585     const double tmp2_0 = A_11_0 + A_11_1 + A_11_2 + A_11_3;
1586     const double tmp3_0 = A_00_2 + A_00_3;
1587     const double tmp4_0 = A_10_1 + A_10_2;
1588     const double tmp5_0 = A_00_0 + A_00_1 + A_00_2 + A_00_3;
1589     const double tmp6_0 = A_01_3 + A_10_0;
1590     const double tmp7_0 = A_01_0 + A_10_3;
1591     const double tmp8_0 = A_01_1 + A_01_2 + A_10_1 + A_10_2;
1592     const double tmp9_0 = A_01_0 + A_10_0;
1593     const double tmp12_0 = A_11_0 + A_11_2;
1594     const double tmp10_0 = A_01_3 + A_10_3;
1595     const double tmp14_0 = A_01_0 + A_01_3 + A_10_0 + A_10_3;
1596     const double tmp13_0 = A_01_2 + A_10_1;
1597     const double tmp11_0 = A_11_1 + A_11_3;
1598     const double tmp18_0 = A_01_1 + A_10_1;
1599     const double tmp15_0 = A_01_1 + A_10_2;
1600     const double tmp16_0 = A_10_0 + A_10_3;
1601     const double tmp17_0 = A_01_1 + A_01_2;
1602     const double tmp19_0 = A_01_2 + A_10_2;
1603     const double tmp0_1 = A_10_3*w8;
1604     const double tmp1_1 = tmp0_0*w1;
1605     const double tmp2_1 = A_01_1*w4;
1606     const double tmp3_1 = tmp1_0*w0;
1607     const double tmp4_1 = A_01_2*w7;
1608     const double tmp5_1 = tmp2_0*w3;
1609     const double tmp6_1 = tmp3_0*w6;
1610     const double tmp7_1 = A_10_0*w2;
1611     const double tmp8_1 = tmp4_0*w5;
1612     const double tmp9_1 = tmp2_0*w10;
1613     const double tmp14_1 = A_10_0*w8;
1614     const double tmp23_1 = tmp3_0*w14;
1615     const double tmp35_1 = A_01_0*w8;
1616     const double tmp54_1 = tmp13_0*w8;
1617     const double tmp20_1 = tmp9_0*w4;
1618     const double tmp25_1 = tmp12_0*w12;
1619     const double tmp44_1 = tmp7_0*w7;
1620     const double tmp26_1 = tmp10_0*w4;
1621     const double tmp52_1 = tmp18_0*w8;
1622     const double tmp48_1 = A_10_1*w7;
1623     const double tmp46_1 = A_01_3*w8;
1624     const double tmp50_1 = A_01_0*w2;
1625     const double tmp56_1 = tmp19_0*w8;
1626     const double tmp19_1 = A_10_3*w2;
1627     const double tmp47_1 = A_10_2*w4;
1628     const double tmp16_1 = tmp3_0*w0;
1629     const double tmp18_1 = tmp1_0*w6;
1630     const double tmp31_1 = tmp11_0*w12;
1631     const double tmp55_1 = tmp15_0*w2;
1632     const double tmp39_1 = A_10_2*w7;
1633     const double tmp11_1 = tmp6_0*w7;
1634     const double tmp40_1 = tmp11_0*w17;
1635     const double tmp34_1 = tmp15_0*w8;
1636     const double tmp33_1 = tmp14_0*w5;
1637     const double tmp24_1 = tmp11_0*w13;
1638     const double tmp43_1 = tmp17_0*w5;
1639     const double tmp15_1 = A_01_2*w4;
1640     const double tmp53_1 = tmp19_0*w2;
1641     const double tmp27_1 = tmp3_0*w11;
1642     const double tmp32_1 = tmp13_0*w2;
1643     const double tmp10_1 = tmp5_0*w9;
1644     const double tmp37_1 = A_10_1*w4;
1645     const double tmp38_1 = tmp5_0*w15;
1646     const double tmp17_1 = A_01_1*w7;
1647     const double tmp12_1 = tmp7_0*w4;
1648     const double tmp22_1 = tmp10_0*w7;
1649     const double tmp57_1 = tmp18_0*w2;
1650     const double tmp28_1 = tmp9_0*w7;
1651     const double tmp29_1 = tmp1_0*w14;
1652     const double tmp51_1 = tmp11_0*w16;
1653     const double tmp42_1 = tmp12_0*w16;
1654     const double tmp49_1 = tmp12_0*w17;
1655     const double tmp21_1 = tmp1_0*w11;
1656     const double tmp45_1 = tmp6_0*w4;
1657     const double tmp13_1 = tmp8_0*w1;
1658     const double tmp36_1 = tmp16_0*w1;
1659     const double tmp41_1 = A_01_3*w2;
1660     const double tmp30_1 = tmp12_0*w13;
1661 caltinay 3748 EM_S[INDEX2(0,0,4)]+=tmp13_1 + tmp20_1 + tmp21_1 + tmp22_1 + tmp23_1 + tmp24_1 + tmp25_1;
1662 caltinay 3764 EM_S[INDEX2(1,0,4)]+=tmp36_1 + tmp37_1 + tmp39_1 + tmp3_1 + tmp43_1 + tmp46_1 + tmp50_1 + tmp5_1 + tmp6_1;
1663     EM_S[INDEX2(2,0,4)]+=tmp0_1 + tmp15_1 + tmp17_1 + tmp1_1 + tmp38_1 + tmp49_1 + tmp51_1 + tmp7_1 + tmp8_1;
1664 caltinay 3748 EM_S[INDEX2(3,0,4)]+=tmp10_1 + tmp32_1 + tmp33_1 + tmp34_1 + tmp9_1;
1665 caltinay 3764 EM_S[INDEX2(0,1,4)]+=tmp0_1 + tmp1_1 + tmp2_1 + tmp3_1 + tmp4_1 + tmp5_1 + tmp6_1 + tmp7_1 + tmp8_1;
1666     EM_S[INDEX2(1,1,4)]+=tmp21_1 + tmp23_1 + tmp30_1 + tmp31_1 + tmp33_1 + tmp56_1 + tmp57_1;
1667     EM_S[INDEX2(2,1,4)]+=tmp10_1 + tmp13_1 + tmp44_1 + tmp45_1 + tmp9_1;
1668 caltinay 3748 EM_S[INDEX2(3,1,4)]+=tmp35_1 + tmp36_1 + tmp37_1 + tmp38_1 + tmp39_1 + tmp40_1 + tmp41_1 + tmp42_1 + tmp43_1;
1669     EM_S[INDEX2(0,2,4)]+=tmp36_1 + tmp38_1 + tmp43_1 + tmp46_1 + tmp47_1 + tmp48_1 + tmp49_1 + tmp50_1 + tmp51_1;
1670 caltinay 3764 EM_S[INDEX2(1,2,4)]+=tmp10_1 + tmp11_1 + tmp12_1 + tmp13_1 + tmp9_1;
1671     EM_S[INDEX2(2,2,4)]+=tmp24_1 + tmp25_1 + tmp27_1 + tmp29_1 + tmp33_1 + tmp52_1 + tmp53_1;
1672     EM_S[INDEX2(3,2,4)]+=tmp14_1 + tmp15_1 + tmp16_1 + tmp17_1 + tmp18_1 + tmp19_1 + tmp1_1 + tmp5_1 + tmp8_1;
1673     EM_S[INDEX2(0,3,4)]+=tmp10_1 + tmp33_1 + tmp54_1 + tmp55_1 + tmp9_1;
1674 caltinay 3748 EM_S[INDEX2(1,3,4)]+=tmp14_1 + tmp19_1 + tmp1_1 + tmp2_1 + tmp38_1 + tmp40_1 + tmp42_1 + tmp4_1 + tmp8_1;
1675     EM_S[INDEX2(2,3,4)]+=tmp16_1 + tmp18_1 + tmp35_1 + tmp36_1 + tmp41_1 + tmp43_1 + tmp47_1 + tmp48_1 + tmp5_1;
1676 caltinay 3764 EM_S[INDEX2(3,3,4)]+=tmp13_1 + tmp26_1 + tmp27_1 + tmp28_1 + tmp29_1 + tmp30_1 + tmp31_1;
1677     } else { // constant data
1678 caltinay 3769 const double A_00 = A_p[INDEX2(0,0,2)];
1679     const double A_10 = A_p[INDEX2(1,0,2)];
1680     const double A_01 = A_p[INDEX2(0,1,2)];
1681     const double A_11 = A_p[INDEX2(1,1,2)];
1682     const double tmp0_0 = A_01 + A_10;
1683     const double tmp0_1 = A_00*w18;
1684     const double tmp1_1 = A_01*w19;
1685     const double tmp2_1 = A_10*w20;
1686     const double tmp3_1 = A_11*w21;
1687     const double tmp4_1 = A_00*w22;
1688     const double tmp5_1 = tmp0_0*w19;
1689     const double tmp6_1 = A_11*w23;
1690     const double tmp7_1 = A_11*w25;
1691     const double tmp8_1 = A_00*w24;
1692     const double tmp9_1 = tmp0_0*w20;
1693     const double tmp10_1 = A_01*w20;
1694     const double tmp11_1 = A_11*w27;
1695     const double tmp12_1 = A_00*w26;
1696     const double tmp13_1 = A_10*w19;
1697 caltinay 3748 EM_S[INDEX2(0,0,4)]+=tmp5_1 + tmp7_1 + tmp8_1;
1698 caltinay 3764 EM_S[INDEX2(1,0,4)]+=tmp0_1 + tmp10_1 + tmp13_1 + tmp3_1;
1699     EM_S[INDEX2(2,0,4)]+=tmp11_1 + tmp12_1 + tmp1_1 + tmp2_1;
1700 caltinay 3748 EM_S[INDEX2(3,0,4)]+=tmp4_1 + tmp6_1 + tmp9_1;
1701 caltinay 3764 EM_S[INDEX2(0,1,4)]+=tmp0_1 + tmp1_1 + tmp2_1 + tmp3_1;
1702     EM_S[INDEX2(1,1,4)]+=tmp7_1 + tmp8_1 + tmp9_1;
1703     EM_S[INDEX2(2,1,4)]+=tmp4_1 + tmp5_1 + tmp6_1;
1704 caltinay 3748 EM_S[INDEX2(3,1,4)]+=tmp10_1 + tmp11_1 + tmp12_1 + tmp13_1;
1705     EM_S[INDEX2(0,2,4)]+=tmp10_1 + tmp11_1 + tmp12_1 + tmp13_1;
1706 caltinay 3764 EM_S[INDEX2(1,2,4)]+=tmp4_1 + tmp5_1 + tmp6_1;
1707     EM_S[INDEX2(2,2,4)]+=tmp7_1 + tmp8_1 + tmp9_1;
1708     EM_S[INDEX2(3,2,4)]+=tmp0_1 + tmp1_1 + tmp2_1 + tmp3_1;
1709     EM_S[INDEX2(0,3,4)]+=tmp4_1 + tmp6_1 + tmp9_1;
1710 caltinay 3748 EM_S[INDEX2(1,3,4)]+=tmp11_1 + tmp12_1 + tmp1_1 + tmp2_1;
1711     EM_S[INDEX2(2,3,4)]+=tmp0_1 + tmp10_1 + tmp13_1 + tmp3_1;
1712 caltinay 3764 EM_S[INDEX2(3,3,4)]+=tmp5_1 + tmp7_1 + tmp8_1;
1713 caltinay 3748 }
1714     }
1715     ///////////////
1716     // process B //
1717     ///////////////
1718     if (!B.isEmpty()) {
1719     add_EM_S=true;
1720     const double* B_p=const_cast<escript::Data*>(&B)->getSampleDataRO(e);
1721     if (B.actsExpanded()) {
1722 caltinay 3769 const double B_0_0 = B_p[INDEX2(0,0,2)];
1723     const double B_1_0 = B_p[INDEX2(1,0,2)];
1724     const double B_0_1 = B_p[INDEX2(0,1,2)];
1725     const double B_1_1 = B_p[INDEX2(1,1,2)];
1726     const double B_0_2 = B_p[INDEX2(0,2,2)];
1727     const double B_1_2 = B_p[INDEX2(1,2,2)];
1728     const double B_0_3 = B_p[INDEX2(0,3,2)];
1729     const double B_1_3 = B_p[INDEX2(1,3,2)];
1730     const double tmp0_0 = B_1_0 + B_1_1;
1731     const double tmp1_0 = B_1_2 + B_1_3;
1732     const double tmp2_0 = B_0_1 + B_0_3;
1733     const double tmp3_0 = B_0_0 + B_0_2;
1734     const double tmp63_1 = B_1_1*w42;
1735     const double tmp79_1 = B_1_1*w40;
1736     const double tmp37_1 = tmp3_0*w35;
1737     const double tmp8_1 = tmp0_0*w32;
1738     const double tmp71_1 = B_0_1*w34;
1739     const double tmp19_1 = B_0_3*w31;
1740     const double tmp15_1 = B_0_3*w34;
1741     const double tmp9_1 = tmp3_0*w34;
1742     const double tmp35_1 = B_1_0*w36;
1743     const double tmp66_1 = B_0_3*w28;
1744     const double tmp28_1 = B_1_0*w42;
1745     const double tmp22_1 = B_1_0*w40;
1746     const double tmp16_1 = B_1_2*w29;
1747     const double tmp6_1 = tmp2_0*w35;
1748     const double tmp55_1 = B_1_3*w40;
1749     const double tmp50_1 = B_1_3*w42;
1750     const double tmp7_1 = tmp1_0*w29;
1751     const double tmp1_1 = tmp1_0*w32;
1752     const double tmp57_1 = B_0_3*w30;
1753     const double tmp18_1 = B_1_1*w32;
1754     const double tmp53_1 = B_1_0*w41;
1755     const double tmp61_1 = B_1_3*w36;
1756     const double tmp27_1 = B_0_3*w38;
1757     const double tmp64_1 = B_0_2*w30;
1758     const double tmp76_1 = B_0_1*w38;
1759     const double tmp39_1 = tmp2_0*w34;
1760     const double tmp62_1 = B_0_1*w31;
1761     const double tmp56_1 = B_0_0*w31;
1762     const double tmp49_1 = B_1_1*w36;
1763     const double tmp2_1 = B_0_2*w31;
1764     const double tmp23_1 = B_0_2*w33;
1765     const double tmp38_1 = B_1_1*w43;
1766     const double tmp74_1 = B_1_2*w41;
1767     const double tmp43_1 = B_1_1*w41;
1768     const double tmp58_1 = B_0_2*w28;
1769     const double tmp67_1 = B_0_0*w33;
1770     const double tmp33_1 = tmp0_0*w39;
1771     const double tmp4_1 = B_0_0*w28;
1772     const double tmp20_1 = B_0_0*w30;
1773     const double tmp13_1 = B_0_2*w38;
1774     const double tmp65_1 = B_1_2*w43;
1775     const double tmp0_1 = tmp0_0*w29;
1776     const double tmp41_1 = tmp3_0*w33;
1777     const double tmp73_1 = B_0_2*w37;
1778     const double tmp69_1 = B_0_0*w38;
1779     const double tmp48_1 = B_1_2*w39;
1780     const double tmp59_1 = B_0_1*w33;
1781     const double tmp17_1 = B_1_3*w41;
1782     const double tmp5_1 = B_0_3*w33;
1783     const double tmp3_1 = B_0_1*w30;
1784     const double tmp21_1 = B_0_1*w28;
1785     const double tmp42_1 = B_1_0*w29;
1786     const double tmp54_1 = B_1_2*w32;
1787     const double tmp60_1 = B_1_0*w39;
1788     const double tmp32_1 = tmp1_0*w36;
1789     const double tmp10_1 = B_0_1*w37;
1790     const double tmp14_1 = B_0_0*w35;
1791     const double tmp29_1 = B_0_1*w35;
1792     const double tmp26_1 = B_1_2*w36;
1793     const double tmp30_1 = B_1_3*w43;
1794     const double tmp70_1 = B_0_2*w35;
1795     const double tmp34_1 = B_1_3*w39;
1796     const double tmp51_1 = B_1_0*w43;
1797     const double tmp31_1 = B_0_2*w34;
1798     const double tmp45_1 = tmp3_0*w28;
1799     const double tmp11_1 = tmp1_0*w39;
1800     const double tmp52_1 = B_1_1*w29;
1801     const double tmp44_1 = B_1_3*w32;
1802     const double tmp25_1 = B_1_1*w39;
1803     const double tmp47_1 = tmp2_0*w33;
1804     const double tmp72_1 = B_1_3*w29;
1805     const double tmp40_1 = tmp2_0*w28;
1806     const double tmp46_1 = B_1_2*w40;
1807     const double tmp36_1 = B_1_2*w42;
1808     const double tmp24_1 = B_0_0*w37;
1809     const double tmp77_1 = B_0_3*w35;
1810     const double tmp68_1 = B_0_3*w37;
1811     const double tmp78_1 = B_0_0*w34;
1812     const double tmp12_1 = tmp0_0*w36;
1813     const double tmp75_1 = B_1_0*w32;
1814 caltinay 3748 EM_S[INDEX2(0,0,4)]+=tmp16_1 + tmp17_1 + tmp18_1 + tmp19_1 + tmp20_1 + tmp21_1 + tmp22_1 + tmp23_1;
1815 caltinay 3764 EM_S[INDEX2(1,0,4)]+=tmp0_1 + tmp1_1 + tmp68_1 + tmp69_1 + tmp70_1 + tmp71_1;
1816     EM_S[INDEX2(2,0,4)]+=tmp45_1 + tmp47_1 + tmp48_1 + tmp49_1 + tmp50_1 + tmp51_1;
1817 caltinay 3748 EM_S[INDEX2(3,0,4)]+=tmp32_1 + tmp33_1 + tmp6_1 + tmp9_1;
1818 caltinay 3764 EM_S[INDEX2(0,1,4)]+=tmp0_1 + tmp1_1 + tmp2_1 + tmp3_1 + tmp4_1 + tmp5_1;
1819     EM_S[INDEX2(1,1,4)]+=tmp72_1 + tmp73_1 + tmp74_1 + tmp75_1 + tmp76_1 + tmp77_1 + tmp78_1 + tmp79_1;
1820     EM_S[INDEX2(2,1,4)]+=tmp32_1 + tmp33_1 + tmp40_1 + tmp41_1;
1821 caltinay 3748 EM_S[INDEX2(3,1,4)]+=tmp34_1 + tmp35_1 + tmp36_1 + tmp37_1 + tmp38_1 + tmp39_1;
1822     EM_S[INDEX2(0,2,4)]+=tmp42_1 + tmp43_1 + tmp44_1 + tmp45_1 + tmp46_1 + tmp47_1;
1823 caltinay 3764 EM_S[INDEX2(1,2,4)]+=tmp6_1 + tmp7_1 + tmp8_1 + tmp9_1;
1824     EM_S[INDEX2(2,2,4)]+=tmp60_1 + tmp61_1 + tmp62_1 + tmp63_1 + tmp64_1 + tmp65_1 + tmp66_1 + tmp67_1;
1825     EM_S[INDEX2(3,2,4)]+=tmp10_1 + tmp11_1 + tmp12_1 + tmp13_1 + tmp14_1 + tmp15_1;
1826     EM_S[INDEX2(0,3,4)]+=tmp40_1 + tmp41_1 + tmp7_1 + tmp8_1;
1827 caltinay 3748 EM_S[INDEX2(1,3,4)]+=tmp37_1 + tmp39_1 + tmp52_1 + tmp53_1 + tmp54_1 + tmp55_1;
1828     EM_S[INDEX2(2,3,4)]+=tmp11_1 + tmp12_1 + tmp56_1 + tmp57_1 + tmp58_1 + tmp59_1;
1829 caltinay 3764 EM_S[INDEX2(3,3,4)]+=tmp24_1 + tmp25_1 + tmp26_1 + tmp27_1 + tmp28_1 + tmp29_1 + tmp30_1 + tmp31_1;
1830     } else { // constant data
1831 caltinay 3769 const double B_0 = B_p[0];
1832     const double B_1 = B_p[1];
1833     const double tmp0_1 = B_0*w44;
1834     const double tmp1_1 = B_1*w45;
1835     const double tmp2_1 = B_0*w46;
1836     const double tmp3_1 = B_0*w47;
1837     const double tmp4_1 = B_1*w48;
1838