preCICE v3.2.0
|
#include <Eigen/Core>
#include <algorithm>
#include <boost/test/data/test_case.hpp>
#include <cmath>
#include <map>
#include <memory>
#include <utility>
#include <vector>
#include "acceleration/Acceleration.hpp"
#include "acceleration/BaseQNAcceleration.hpp"
#include "acceleration/IQNILSAcceleration.hpp"
#include "acceleration/IQNIMVJAcceleration.hpp"
#include "acceleration/SharedPointer.hpp"
#include "acceleration/impl/ConstantPreconditioner.hpp"
#include "acceleration/impl/QRFactorization.hpp"
#include "acceleration/impl/ResidualSumPreconditioner.hpp"
#include "acceleration/impl/SharedPointer.hpp"
#include "acceleration/test/helper.hpp"
#include "cplscheme/Constants.hpp"
#include "cplscheme/CouplingData.hpp"
#include "cplscheme/SharedPointer.hpp"
#include "mesh/Mesh.hpp"
#include "mesh/SharedPointer.hpp"
#include "testing/Meshes.hpp"
#include "testing/TestContext.hpp"
#include "testing/Testing.hpp"
#include "utils/EigenHelperFunctions.hpp"
Go to the source code of this file.
Typedefs | |
using | DataMap = std::map<int, PtrCouplingData> |
Functions | |
PRECICE_TEST_SETUP (""_on(4_ranks).setupIntraComm()) | |
Test that runs on 4 processors. | |
cplscheme::PtrCouplingData | makeCouplingData (mesh::PtrData data, mesh::PtrMesh mesh, bool exchangeSubsteps) |
using DataMap = std::map<int, PtrCouplingData> |
Definition at line 40 of file AccelerationIntraCommTest.cpp.
|
inline |
Definition at line 10 of file helper.hpp.
PRECICE_TEST_SETUP | ( | "" | _on4_ranks).setupIntraComm( | ) |
Test that runs on 4 processors.
Tests a non-contigous owner distributed at the outMesh.
Using meshes of different sizes, outMesh is smaller then inMesh.
Using meshes of different sizes, inMesh is smaller then outMesh.
Using a more heterogeneous distribution of vertices and owner.
Test with a homogeneous distribution of mesh among ranks.
same as 2DV4, but strictly linear input values, converges and gives correct results
Some ranks are empty, does not converge.
Test with a very heterogeneous distributed and non-continuous ownership.
Using a more heterogeneous distributon of vertices and owner.
processor with 4 vertices
processor with 4 vertices
processor with no vertices
processor with 2 vertices
processor with 4 vertices
processor with 4 vertices
processor with no vertices
processor with 2 vertices
processor with no vertices
processor with 11 vertices
processor with 11 vertices
processor with no vertices
Definition at line 45 of file AccelerationIntraCommTest.cpp.