1 // Copyright (C) 2007-2012 CEA/DEN, EDF R&D
3 // This library is free software; you can redistribute it and/or
4 // modify it under the terms of the GNU Lesser General Public
5 // License as published by the Free Software Foundation; either
6 // version 2.1 of the License.
8 // This library is distributed in the hope that it will be useful,
9 // but WITHOUT ANY WARRANTY; without even the implied warranty of
10 // MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
11 // Lesser General Public License for more details.
13 // You should have received a copy of the GNU Lesser General Public
14 // License along with this library; if not, write to the Free Software
15 // Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA
17 // See http://www.salome-platform.org/ or email : webmaster.salome@opencascade.com
20 #include "OverlapDEC.hxx"
21 #include "CommInterface.hxx"
22 #include "ParaFIELD.hxx"
23 #include "MPIProcessorGroup.hxx"
24 #include "OverlapElementLocator.hxx"
25 #include "OverlapInterpolationMatrix.hxx"
27 \defgroup overlapdec OverlapDEC
28 The \c OverlapDEC enables the \ref InterpKerRemapGlobal "conservative remapping" of fields between two parallel codes. This remapping is based on the computation of intersection volumes on a \b same \b processor \b group. On this processor group are defined two field-templates called A and B. The computation is possible for 3D meshes, 2D meshes, 3D-surface meshes, 1D meshes and 2D-curve meshes. Dimensions must be similar for the distribution templates A and B.
29 The main difference with \ref interpkerneldec is that this \ref dec manages 2 field templates on each processor of the processor group (A and B) called source and target.
30 Furthermore all processors in processor group cooperates in global interpolation matrix computation. In this respect \ref InterpKernelIDEC is a specialization of \c OverlapDEC.
32 \section ParaMEDMEMOverlapDECAlgorithmDescription Algorithm Description
34 Let's consider the following use case that is ran in ParaMEDMEMTest_OverlapDEC.cxx to describes the different steps of the computation. The processor group contains 3 processors.
35 \anchor ParaMEDMEMOverlapDECImgTest1
36 \image html OverlapDEC1.png "Example showing the use case in order to explain the different steps."
38 \subsection ParaMEDMEMOverlapDECAlgoStep1 Step 1 : Bounding box exchange and global interaction between procs computation.
40 In order to reduce as much as possible the amount of communications between distant processors, every processor computes a bounding box for A and B. Then a AllToAll communication is performed so that
41 every processor can compute the \b global interactions between processor.
42 This computation leads every processor to compute the same global TODO list expressed as a list of pair. A pair (x,y) means that proc \b x fieldtemplate A can interact with fieltemplate B of proc \b y because the two bounding boxes interact.
43 In the \ref ParaMEDMEMOverlapDECImgTest1 "example above" this computation leads to the following a \b global TODO list :
45 \b (0,0),(0,1),(1,0),(1,2),(2,0),(2,1),(2,2)
47 Here the pair (0,2) does not appear because the bounding box of fieldtemplateA of proc#2 does not intersect that of fieldtemplate B on proc#0.
49 Stage performed by ParaMEDMEM::OverlapElementLocator::computeBoundingBoxes.
51 \subsection ParaMEDMEMOverlapDECAlgoStep2 Step 2 : Computation of local TODO list
53 Starting from the global interaction previously computed in \ref ParaMEDMEMOverlapDECAlgoStep1 "Step 1", each proc computes the TODO list per proc.
54 The following rules is chosen : a pair (x,y) can be treated by either proc #x or proc #y, in order to reduce the amount of data transfert among
55 processors. The algorithm chosen for load balancing is the following : Each processor has an empty \b local TODO list at the beginning. Then for each pair (k,m) in
56 \b global TODO list, if proc#k has less temporary local list than proc#m pair, (k,m) is added to temparary local TODO list of proc#k.
57 If proc#m has less temporary local TODO list than proc#k pair, (k,m) is added to temporary local TODO list of proc#m.
58 If proc#k and proc#m have the same amount of temporary local TODO list pair, (k,m) is added to temporary local TODO list of proc#k.
60 In the \ref ParaMEDMEMOverlapDECImgTest1 "example above" this computation leads to the following local TODO list :
63 - proc#1 : (0,1),(1,0)
64 - proc#2 : (1,2),(2,0),(2,1),(2,2)
66 The algorithm described here is not perfect for this use case, we hope to enhance it soon.
68 At this stage each proc knows precisely its \b local TODO list (with regard to interpolation). The \b local TODO list of other procs than local
69 is kept for future computations.
71 \subsection ParaMEDMEMOverlapDECAlgoStep3 Step 3 : Matrix echange between procs
73 Knowing the \b local TODO list, the aim now is to exchange field-templates between procs. Each proc computes knowing TODO list per
74 proc computed in \ref ParaMEDMEMOverlapDECAlgoStep2 "Step 2" the exchange TODO list :
76 In the \ref ParaMEDMEMOverlapDECImgTest1 "example above" the exchange TODO list gives the following results :
78 Sending TODO list per proc :
80 - proc #0 : Send fieldtemplate A to Proc#1, Send fieldtemplate B to Proc#1, Send fieldtemplate B to Proc#2
81 - Proc #1 : Send fieldtemplate A to Proc#2, Send fieldtemplate B to Proc#2
84 Receiving TODO list per proc :
86 - proc #0 : No receiving
87 - proc #1 : receiving fieldtemplate A from Proc#0, receiving fieldtemplate B from Proc#0
88 - proc #2 : receiving fieldtemplate B from Proc#0, receiving fieldtemplate A from Proc#1, receiving fieldtemplate B from Proc#1
90 To avoid as much as possible large volumes of transfers between procs, only relevant parts of meshes are sent. In order for proc#k to send fieldtemplate A to fieldtemplate B
91 of proc #m., proc#k computes the part of mesh A contained in the boundingbox B of proc#m. It implies that the corresponding cellIds or nodeIds of the
92 corresponding part are sent to proc #m too.
94 Let's consider the couple (k,m) in the TODO list. This couple is treated by either k or m as seen in \ref ParaMEDMEMOverlapDECAlgoStep2 "here in Step2".
96 As will be dealt in Step 6, for final matrix-vector computations, the resulting matrix of the couple (k,m) whereever it is computed (proc #k or proc #m)
97 will be stored in \b proc#m.
99 - If proc #k is in charge (performs the matrix computation) for this couple (k,m), target ids (cells or nodes) of the mesh in proc #m are renumbered, because proc #m has seelected a sub mesh of the target mesh to avoid large amounts of data to transfer. In this case as proc #m is ultimately in charge of the matrix, proc #k must keep preciously the
100 source ids needed to be sent to proc#m. No problem will appear for matrix assembling in proc m for source ids because no restriction was done.
101 Concerning source ids to be sent for the matrix-vector computation, proc k will know precisely which source ids field values to send to proc #m.
102 This is embodied by OverlapMapping::keepTracksOfTargetIds in proc m.
104 - If proc #m is in charge (performs matrix computation) for this couple (k,m), source ids (cells or nodes) of the mesh in proc #k are renumbered, because proc #k has selected a sub mesh of the source mesh to avoid large amounts of data to transfer. In this case as proc #k is ultimately in charge of the matrix, proc #m receives the source ids
105 from remote proc #k, and thus the matrix is directly correct, no need for renumbering as in \ref ParaMEDMEMOverlapDECAlgoStep5 "Step 5". However proc #k must
106 keep track of the ids sent to proc #m for te matrix-vector computation.
107 This is incarnated by OverlapMapping::keepTracksOfSourceIds in proc k.
109 This step is performed in ParaMEDMEM::OverlapElementLocator::exchangeMeshes method.
111 \subsection ParaMEDMEMOverlapDECAlgoStep4 Step 4 : Computation of the interpolation matrix
113 After mesh exchange in \ref ParaMEDMEMOverlapDECAlgoStep3 "Step3" each processor has all the required information to treat its \b local TODO list computed in
114 \ref ParaMEDMEMOverlapDECAlgoStep2 "Step2". This step is potentially CPU costly, which is why the \b local TODO list per proc is expected to
115 be as well balanced as possible.
117 The interpolation is performed as \ref ParaMEDMEM::MEDCouplingRemapper "Remapper" does.
119 This operation is performed by OverlapInterpolationMatrix::addContribution method.
121 \subsection ParaMEDMEMOverlapDECAlgoStep5 Step 5 : Global matrix construction.
123 After having performed the TODO list at the end of \ref ParaMEDMEMOverlapDECAlgoStep4 "Step4" we need to assemble the final matrix.
125 The final aim is to have a distributed matrix \f$ M_k \f$ on each proc#k. In order to reduce data exchange during the matrix product process,
126 \f$ M_k \f$ is built using sizeof(Proc group) \c std::vector< \c std::map<int,double> \c >.
128 For a proc#k, it is necessary to fetch info of all matrices built in \ref ParaMEDMEMOverlapDECAlgoStep4 "Step4" where the first element in pair (i,j)
131 After this step, the matrix repartition is the following after a call to ParaMEDMEM::OverlapMapping::prepare :
133 - proc#0 : (0,0),(1,0),(2,0)
134 - proc#1 : (0,1),(2,1)
135 - proc#2 : (1,2),(2,2)
137 Tuple (2,1) computed on proc 2 is stored in proc 1 after execution of the function "prepare". This is an example of item 0 in \ref ParaMEDMEMOverlapDECAlgoStep2 "Step2".
138 Tuple (0,1) computed on proc 1 is stored in proc 1 too. This is an example of item 1 in \ref ParaMEDMEMOverlapDECAlgoStep2 "Step2".
140 In the end ParaMEDMEM::OverlapMapping::_proc_ids_to_send_vector_st will contain :
146 In the end ParaMEDMEM::OverlapMapping::_proc_ids_to_recv_vector_st will contain :
152 The method in charge to perform this is : ParaMEDMEM::OverlapMapping::prepare.
156 OverlapDEC::OverlapDEC(const std::set<int>& procIds, const MPI_Comm& world_comm):_own_group(true),_interpolation_matrix(0),
157 _source_field(0),_own_source_field(false),
158 _target_field(0),_own_target_field(false)
160 ParaMEDMEM::CommInterface comm;
161 int *ranks_world=new int[procIds.size()]; // ranks of sources and targets in world_comm
162 std::copy(procIds.begin(),procIds.end(),ranks_world);
163 MPI_Group group,world_group;
164 comm.commGroup(world_comm,&world_group);
165 comm.groupIncl(world_group,procIds.size(),ranks_world,&group);
166 delete [] ranks_world;
168 comm.commCreate(world_comm,group,&theComm);
169 comm.groupFree(&group);
170 if(theComm==MPI_COMM_NULL)
175 std::set<int> idsUnion;
176 for(std::size_t i=0;i<procIds.size();i++)
178 _group=new MPIProcessorGroup(comm,idsUnion,theComm);
181 OverlapDEC::~OverlapDEC()
185 if(_own_source_field)
186 delete _source_field;
187 if(_own_target_field)
188 delete _target_field;
189 delete _interpolation_matrix;
192 void OverlapDEC::sendRecvData(bool way)
200 void OverlapDEC::sendData()
202 _interpolation_matrix->multiply();
205 void OverlapDEC::recvData()
207 throw INTERP_KERNEL::Exception("Not implemented yet !!!!");
208 //_interpolation_matrix->transposeMultiply();
211 void OverlapDEC::synchronize()
215 delete _interpolation_matrix;
216 _interpolation_matrix=new OverlapInterpolationMatrix(_source_field,_target_field,*_group,*this,*this);
217 OverlapElementLocator locator(_source_field,_target_field,*_group);
218 locator.copyOptions(*this);
219 locator.exchangeMeshes(*_interpolation_matrix);
220 std::vector< std::pair<int,int> > jobs=locator.getToDoList();
221 std::string srcMeth=locator.getSourceMethod();
222 std::string trgMeth=locator.getTargetMethod();
223 for(std::vector< std::pair<int,int> >::const_iterator it=jobs.begin();it!=jobs.end();it++)
225 const MEDCouplingPointSet *src=locator.getSourceMesh((*it).first);
226 const DataArrayInt *srcIds=locator.getSourceIds((*it).first);
227 const MEDCouplingPointSet *trg=locator.getTargetMesh((*it).second);
228 const DataArrayInt *trgIds=locator.getTargetIds((*it).second);
229 _interpolation_matrix->addContribution(src,srcIds,srcMeth,(*it).first,trg,trgIds,trgMeth,(*it).second);
231 _interpolation_matrix->prepare(locator.getProcsInInteraction());
232 _interpolation_matrix->computeDeno();
235 void OverlapDEC::attachSourceLocalField(ParaFIELD *field, bool ownPt)
239 if(_own_source_field)
240 delete _source_field;
242 _own_source_field=ownPt;
245 void OverlapDEC::attachTargetLocalField(ParaFIELD *field, bool ownPt)
249 if(_own_target_field)
250 delete _target_field;
252 _own_target_field=ownPt;
255 bool OverlapDEC::isInGroup() const
259 return _group->containsMyRank();