PEER NEES NEESit

• HOME
• USER
• DEVELOPER
• PROJECTS
• SUPPORT
• SITE MAP

• Dev Doc
• API
• Source
• Download
• Builds
• Bug Reports
• Message Board

DistributedBandSPDLinSOE.cpp Go to the documentation of this file. /* ****************************************************************** ** ** OpenSees - Open System for Earthquake Engineering Simulation ** ** Pacific Earthquake Engineering Research Center ** ** ** ** ** ** (C) Copyright 1999, The Regents of the University of California ** ** All Rights Reserved. ** ** ** ** Commercial use of this program without express permission of the ** ** University of California, Berkeley, is strictly prohibited. See ** ** file 'COPYRIGHT' in main directory for information on usage and ** ** redistribution, and for a DISCLAIMER OF ALL WARRANTIES. ** ** ** ** Developed by: ** ** Frank McKenna (fmckenna@ce.berkeley.edu) ** ** Gregory L. Fenves (fenves@ce.berkeley.edu) ** ** Filip C. Filippou (filippou@ce.berkeley.edu) ** ** ** ** ****************************************************************** */ // $Revision: 1.1 $ // $Date: 2005/12/06 22:04:33 $ // $Source: /usr/local/cvs/OpenSees/SRC/system_of_eqn/linearSOE/bandSPD/DistributedBandSPDLinSOE.cpp,v $ // Written: fmk // Revision: A // // Description: This file contains the implementation for BandSPDLinSOE #include <DistributedBandSPDLinSOE.h> #include <BandSPDLinSolver.h> #include <Matrix.h> #include <Graph.h> #include <Vertex.h> #include <VertexIter.h> #include <f2c.h> #include <math.h> #include <Channel.h> #include <FEM_ObjectBroker.h> DistributedBandSPDLinSOE::DistributedBandSPDLinSOE(BandSPDLinSolver &theSolvr) :BandSPDLinSOE(theSolvr, LinSOE_TAGS_DistributedBandSPDLinSOE), processID(0), numChannels(0), theChannels(0), localCol(0), workArea(0), sizeWork(0), myVectB(0), myB(0) { theSolvr.setLinearSOE(*this); } DistributedBandSPDLinSOE::~DistributedBandSPDLinSOE() { if (theChannels != 0) delete [] theChannels; if (localCol != 0) for (int i=0; i<numChannels; i++) if (localCol[i] != 0) delete localCol[i]; delete [] localCol; if (workArea != 0) delete [] workArea; if (myVectB != 0) delete myVectB; if (myB != 0) delete [] myB; } int DistributedBandSPDLinSOE::setSize(Graph &theGraph) { int result = 0; int oldSize = size; half_band = 0; int maxNumSubVertex = 0; // if subprocess, collect graph, send it off, // vector back containing size of system, etc. if (processID != 0) { Channel *theChannel = theChannels[0]; theGraph.sendSelf(0, *theChannel); static ID data(2); theChannel->recvID(0, 0, data); size = data(0); half_band = data(1); ID *subMap = new ID(theGraph.getNumVertex()); localCol[0] = subMap; Vertex *vertex; VertexIter &theSubVertices = theGraph.getVertices(); int cnt = 0; while((vertex = theSubVertices()) != 0) (*subMap)(cnt++) = vertex->getTag(); theChannel->sendID(0, 0, *subMap); } // if main domain, collect graphs from all subdomains, // merge into 1, number this one, send to subdomains the // id containing dof tags & start id's. else { // from each distributed soe recv it's graph // and merge them into master graph FEM_ObjectBroker theBroker; for (int j=0; j<numChannels; j++) { Channel *theChannel = theChannels[j]; Graph theSubGraph; theSubGraph.recvSelf(0, *theChannel, theBroker); theGraph.merge(theSubGraph); int numSubVertex = theSubGraph.getNumVertex(); ID *subMap = new ID(numSubVertex); localCol[j] = subMap; if (numSubVertex > maxNumSubVertex) maxNumSubVertex = numSubVertex; } size = theGraph.getNumVertex(); // // determine the number of superdiagonals and subdiagonals // Vertex *vertexPtr; VertexIter &theVertices = theGraph.getVertices(); while ((vertexPtr = theVertices()) != 0) { int vertexNum = vertexPtr->getTag(); const ID &theAdjacency = vertexPtr->getAdjacency(); for (int i=0; i<theAdjacency.Size(); i++) { int otherNum = theAdjacency(i); int diff = vertexNum-otherNum; if (half_band < diff) half_band = diff; } } half_band += 1; // include the diagonal static ID data(2); data(0) = size; data(1) = half_band; // to each distributed soe send the size data // and merge them into master graph for (int j=0; j<numChannels; j++) { Channel *theChannel = theChannels[j]; theChannel->sendID(0, 0, data); ID *subMap = localCol[j]; theChannel->recvID(0, 0, *subMap); } } int numCols = theGraph.getNumVertex(); if (processID != 0) { ID &globalMap = *(localCol[0]); ID *localMap = new ID(size); localMap->Zero(); for (int k=0; k< globalMap.Size(); k++) (*localMap)(globalMap(k)) = k; delete localCol[0]; localCol[0] = localMap; } int newSize; if (processID != 0) newSize = numCols * half_band; else newSize = size * half_band; // newSize = size * half_band; if (newSize != Asize) { // we have to get another space for A if (processID == 0) { if (workArea != 0) delete [] workArea; workArea = new double [newSize]; sizeWork = newSize; } if (A != 0) delete [] A; A = new double[newSize]; if (A == 0) { opserr << "WARNING DistributedBandSPDLinSOE::DistributedBandSPDLinSOE :"; opserr << " ran out of memory for A (size, half_band) ("; opserr << size <<", " << half_band << "\n"; Asize = 0; size = 0; half_band = 0; result= -1; } else Asize = newSize; } // zero the matrix for (int i=0; i<Asize; i++) A[i] = 0; factored = false; if (size > Bsize) { // we have to get space for the vectors // delete the old if (B != 0) delete [] B; if (X != 0) delete [] X; if (myB != 0) delete [] myB; // create the new B = new double[size]; X = new double[size]; myB = new double[size]; if (B == 0 || X == 0 || myB == 0) { opserr << "WARNING DistributedBandSPDLinSOE::DistributedBandSPDLinSOE :"; opserr << " ran out of memory for vectors (size) ("; opserr << size << ") \n"; Bsize = 0; size = 0; half_band = 0; result = -1; } else Bsize = size; } // zero the vectors for (int j=0; j<size; j++) { B[j] = 0; X[j] = 0; myB[j] = 0; } // get new Vector objects if size has changes if (oldSize != size) { if (vectX != 0) delete vectX; if (vectB != 0) delete vectB; if (myVectB != 0) delete myVectB; vectX = new Vector(X,size); vectB = new Vector(B,size); vectX = new Vector(X,size); myVectB = new Vector(myB, size); } // invoke setSize() on the Solver if (processID == 0) { LinearSOESolver *theSolvr = this->getSolver(); int solverOK = theSolvr->setSize(); if (solverOK < 0) { opserr << "WARNING:DistributedBandSPDLinSOE::setSize :"; opserr << " solver failed setSize()\n"; return solverOK; } } return result; } int DistributedBandSPDLinSOE::addA(const Matrix &m, const ID &id, double fact) { // check for a quick return if (fact == 0.0) return 0; // check that m and id are of similar size int idSize = id.Size(); if (idSize != m.noRows() && idSize != m.noCols()) { opserr << "BandSPDLinSOE::addA() - Matrix and ID not of similar sizes\n"; return -1; } ID *theMap =0; if (numChannels > 0) theMap = localCol[0]; if (fact == 1.0) { // do not need to multiply for (int i=0; i<idSize; i++) { int col = id(i); if (col < size && col >= 0) { // double *coliiPtr = A +(col+1)*half_band -1; double *coliiPtr; if (processID == 0) coliiPtr = A + (col+1) * half_band -1; else coliiPtr = A + ((*theMap)(col)+1)*half_band-1; int minColRow = col - half_band + 1; for (int j=0; j<idSize; j++) { int row = id(j); if (row <size && row >= 0 && row <= col && row >= minColRow) { // only add upper double *APtr = coliiPtr + (row-col); *APtr += m(j,i); } } // for j } } // for i } else { for (int i=0; i<idSize; i++) { int col = id(i); if (col < size && col >= 0) { // double *coliiPtr = A +(col+1)*half_band -1; double *coliiPtr; if (processID == 0) coliiPtr = A + (col+1) * half_band -1; else coliiPtr = A + ((*theMap)(col)+1)*half_band-1; int minColRow = col - half_band +1; for (int j=0; j<idSize; j++) { int row = id(j); if (row <size && row >= 0 && row <= col && row >= minColRow) { // only add upper double *APtr = coliiPtr + (row-col); *APtr += m(j,i)*fact; } } // for j } } // for i } return 0; return 0; } int DistributedBandSPDLinSOE::solve(void) { static ID result(1); // // if subprocess send B and A and receive back result X, B & result // if (processID != 0) { Channel *theChannel = theChannels[0]; // send B theChannel->sendVector(0, 0, *myVectB); // send A in packets placed in vector X if (factored == false) { Vector vectA(A, Asize); theChannel->sendVector(0, 0, vectA); } // receive X,B and result theChannel->recvVector(0, 0, *vectX); theChannel->recvVector(0, 0, *vectB); theChannel->recvID(0, 0, result); factored = true; } // // if main process, recv B & A from all, solve and send back X, B & result // else { *vectB = *myVectB; // receive X and A contribution from subprocess & add them in for (int j=0; j<numChannels; j++) { // get X & add Channel *theChannel = theChannels[j]; theChannel->recvVector(0, 0, *vectX); *vectB += *vectX; // get A & add using local map if (factored == false) { const ID &localMap = *(localCol[j]); int localSize = localMap.Size() * half_band; Vector vectA(workArea, localSize); theChannel->recvVector(0, 0, vectA); int loc = 0; for (int i=0; i<localMap.Size(); i++) { int pos = localMap(i)*half_band; for (int k=0; k<half_band; k++) A[pos++] += workArea[loc++]; } } } // solve result(0) = this->LinearSOE::solve(); // send results back for (int j=0; j<numChannels; j++) { Channel *theChannel = theChannels[j]; theChannel->sendVector(0, 0, *vectX); theChannel->sendVector(0, 0, *vectB); theChannel->sendID(0, 0, result); } } return result(0); } int DistributedBandSPDLinSOE::addB(const Vector &v, const ID &id, double fact) { // check for a quick return if (fact == 0.0) return 0; // check that m and id are of similar size int idSize = id.Size(); if (idSize != v.Size() ) { opserr << "BandSPDLinSOE::addB() - Vector and ID not of similar sizes\n"; return -1; } if (fact == 1.0) { // do not need to multiply if fact == 1.0 for (int i=0; i<idSize; i++) { int pos = id(i); if (pos <size && pos >= 0) myB[pos] += v(i); } } else if (fact == -1.0) { for (int i=0; i<idSize; i++) { int pos = id(i); if (pos <size && pos >= 0) myB[pos] -= v(i); } } else { for (int i=0; i<idSize; i++) { int pos = id(i); if (pos <size && pos >= 0) myB[pos] += v(i) * fact; } } return 0; } int DistributedBandSPDLinSOE::setB(const Vector &v, double fact) { // check for a quick return if (fact == 0.0) return 0; if (v.Size() != size) { opserr << "WARNING DistributedBandGenLinSOE::setB() -"; opserr << " incomptable sizes " << size << " and " << v.Size() << endln; return -1; } if (fact == 1.0) { // do not need to multiply if fact == 1.0 for (int i=0; i<size; i++) { myB[i] = v(i); } } else if (fact == -1.0) { for (int i=0; i<size; i++) { myB[i] = -v(i); } } else { for (int i=0; i<size; i++) { myB[i] = v(i) * fact; } } return 0; } void DistributedBandSPDLinSOE::zeroB(void) { double *Bptr = myB; for (int i=0; i<size; i++) *Bptr++ = 0; } const Vector & DistributedBandSPDLinSOE::getB(void) { if (processID != 0) { Channel *theChannel = theChannels[0]; // send B & recv merged B theChannel->sendVector(0, 0, *myVectB); theChannel->recvVector(0, 0, *vectB); } // // if main process, recv B & A from all, solve and send back X, B & result // else { // receive X and A contribution from subprocess & add them in *vectB = *myVectB; Vector remoteB(workArea, size); for (int j=0; j<numChannels; j++) { Channel *theChannel = theChannels[j]; theChannel->recvVector(0, 0, remoteB); *vectB += remoteB; } // send results back for (int j=0; j<numChannels; j++) { Channel *theChannel = theChannels[j]; theChannel->sendVector(0, 0, *vectB); } } return *vectB; } int DistributedBandSPDLinSOE::sendSelf(int commitTag, Channel &theChannel) { int sendID =0; // if P0 check if already sent. If already sent use old processID; if not allocate a new process // id for remote part of object, enlarge channel * to hold a channel * for this remote object. // if not P0, send current processID if (processID == 0) { // check if already using this object bool found = false; for (int i=0; i<numChannels; i++) if (theChannels[i] == &theChannel) { sendID = i+1; found = true; } // if new object, enlarge Channel pointers to hold new channel * & allocate new ID if (found == false) { int nextNumChannels = numChannels + 1; Channel **nextChannels = new Channel *[nextNumChannels]; if (nextNumChannels == 0) { opserr << "DistributedBandSPDLinSOE::sendSelf() - failed to allocate channel array of size: " << nextNumChannels << endln; return -1; } for (int i=0; i<numChannels; i++) nextChannels[i] = theChannels[i]; nextChannels[numChannels] = &theChannel; numChannels = nextNumChannels; if (theChannels != 0) delete [] theChannels; theChannels = nextChannels; if (localCol != 0) delete [] localCol; localCol = new ID *[numChannels]; if (localCol == 0) { opserr << "DistributedBandSPDLinSOE::sendSelf() - failed to allocate id array of size: " << nextNumChannels << endln; return -1; } for (int i=0; i<numChannels; i++) localCol[i] = 0; // allocate new processID for remote object sendID = numChannels; } } else sendID = processID; // send remotes processID ID idData(1); idData(0) = sendID; int res = theChannel.sendID(0, commitTag, idData); if (res < 0) { opserr <<"WARNING DistributedBandSPDLinSOE::sendSelf() - failed to send data\n"; return -1; } return 0; return 0; } int DistributedBandSPDLinSOE::recvSelf(int commitTag, Channel &theChannel, FEM_ObjectBroker &theBroker) { ID idData(1); int res = theChannel.recvID(0, commitTag, idData); if (res < 0) { opserr <<"WARNING DistributedBandSPDLinSOE::recvSelf() - failed to send data\n"; return -1; } processID = idData(0); numChannels = 1; theChannels = new Channel *[1]; theChannels[0] = &theChannel; localCol = new ID *[numChannels]; for (int i=0; i<numChannels; i++) localCol[i] = 0; return 0; } int DistributedBandSPDLinSOE::setProcessID(int dTag) { processID = dTag; return 0; } int DistributedBandSPDLinSOE::setChannels(int nChannels, Channel **theC) { numChannels = nChannels; if (theChannels != 0) delete [] theChannels; theChannels = new Channel *[numChannels]; for (int i=0; i<numChannels; i++) theChannels[i] = theC[i]; localCol = new ID *[nChannels]; for (int i=0; i<numChannels; i++) localCol[i] = 0; return 0; } Generated on Mon Oct 23 15:05:28 2006 for OpenSees by  1.5.0

---

opensees-support @ berkeley.edu

Supported by the National Science Foundation

©2006, UC Regents