opm-simulators/tests/test_autodiffmatrix.cpp
Arne Morten Kvarving 02b4644612 workaround eigen 3.3.2 issue
eigen returns a matrix with all coefficients stored from the operator/
this makes the comparison fail as x has no nonzeros
2017-10-06 12:48:43 +02:00

432 lines
10 KiB
C++

/*
Copyright 2014 SINTEF ICT, Applied Mathematics.
This file is part of the Open Porous Media project (OPM).
OPM is free software: you can redistribute it and/or modify
it under the terms of the GNU General Public License as published by
the Free Software Foundation, either version 3 of the License, or
(at your option) any later version.
OPM is distributed in the hope that it will be useful,
but WITHOUT ANY WARRANTY; without even the implied warranty of
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
GNU General Public License for more details.
You should have received a copy of the GNU General Public License
along with OPM. If not, see <http://www.gnu.org/licenses/>.
*/
#include <config.h>
#if HAVE_DYNAMIC_BOOST_TEST
#define BOOST_TEST_DYN_LINK
#endif
#define BOOST_TEST_MODULE AutoDiffMatrixTest
#include <opm/autodiff/AutoDiffMatrix.hpp>
#include <opm/autodiff/AutoDiffHelpers.hpp>
#include <boost/test/unit_test.hpp>
typedef Eigen::SparseMatrix<double> Sp;
typedef Opm::AutoDiffMatrix Mat;
using namespace Opm;
bool
operator ==(const Eigen::SparseMatrix<double>& A,
const Eigen::SparseMatrix<double>& B)
{
// Two SparseMatrices are equal if
// 0) They have the same ordering (enforced by equal types)
// 1) They have the same outer and inner dimensions
// 2) They have the same number of non-zero elements
// 3) They have the same sparsity structure
// 4) The non-zero elements are equal
// 1) Outer and inner dimensions
bool eq = (A.outerSize() == B.outerSize());
eq = eq && (A.innerSize() == B.innerSize());
// 2) Equal number of non-zero elements
eq = eq && (A.nonZeros() == B.nonZeros());
for (typename Eigen::SparseMatrix<double>::Index
k0 = 0, kend = A.outerSize(); eq && (k0 < kend); ++k0) {
for (typename Eigen::SparseMatrix<double>::InnerIterator
iA(A, k0), iB(B, k0); eq && (iA && iB); ++iA, ++iB) {
// 3) Sparsity structure
eq = (iA.row() == iB.row()) && (iA.col() == iB.col());
// 4) Equal non-zero elements
eq = eq && (iA.value() == iB.value());
}
}
return eq;
// Note: Investigate implementing this operator as
// return A.cwiseNotEqual(B).count() == 0;
}
BOOST_AUTO_TEST_CASE(Initialization)
{
// Setup.
Mat z = Mat(3, 3);
Mat i = Mat::createIdentity(3);
Eigen::Array<double, Eigen::Dynamic, 1> d1(3);
d1 << 0.2, 1.2, 13.4;
Mat d = Mat(d1.matrix().asDiagonal());
Eigen::Matrix<double, Eigen::Dynamic, Eigen::Dynamic> s1(3,2);
s1 <<
1.0, 0.0, 2.0,
0.0, 1.0, 0.0;
Sp s2(s1.sparseView());
Mat s = Mat(s2);
}
BOOST_AUTO_TEST_CASE(EigenConversion)
{
// Setup.
Mat z = Mat(3, 3);
Mat i = Mat::createIdentity(3);
Eigen::Array<double, Eigen::Dynamic, 1> d1(3);
d1 << 0.2, 1.2, 13.4;
Mat d = Mat(d1.matrix().asDiagonal());
Eigen::Matrix<double, Eigen::Dynamic, Eigen::Dynamic> s1(3,2);
s1 <<
1.0, 0.0, 2.0,
0.0, 1.0, 0.0;
Sp s2(s1.sparseView());
Mat s = Mat(s2);
// Convert to Eigen::SparseMatrix
Sp x;
z.toSparse(x);
Sp z1(3,3);
BOOST_CHECK(x == z1);
i.toSparse(x);
Sp i1(Eigen::Matrix<double, Eigen::Dynamic, Eigen::Dynamic>::Identity(3,3).sparseView());
BOOST_CHECK(x == i1);
d.toSparse(x);
Sp d2 = spdiag(d1);
BOOST_CHECK(x == d2);
s.toSparse(x);
BOOST_CHECK(x == s2);
}
BOOST_AUTO_TEST_CASE(AdditionOps)
{
// Setup.
Mat z = Mat(3, 3);
Sp zs(3,3);
Mat i = Mat::createIdentity(3);
Sp is(Eigen::Matrix<double, Eigen::Dynamic, Eigen::Dynamic>::Identity(3,3).sparseView());
Eigen::Array<double, Eigen::Dynamic, 1> d1(3);
d1 << 0.2, 1.2, 13.4;
Mat d = Mat(d1.matrix().asDiagonal());
Sp ds = spdiag(d1);
Eigen::Matrix<double, Eigen::Dynamic, Eigen::Dynamic> s1(3,3);
s1 <<
1.0, 0.0, 2.0,
0.0, 1.0, 0.0,
0.0, 0.0, 2.0;
Sp ss(s1.sparseView());
Mat s = Mat(ss);
// Convert to Eigen::SparseMatrix
Sp x;
z.toSparse(x);
BOOST_CHECK(x == zs);
i.toSparse(x);
BOOST_CHECK(x == is);
d.toSparse(x);
BOOST_CHECK(x == ds);
s.toSparse(x);
BOOST_CHECK(x == ss);
// Adding zero.
auto zpz = z + z;
zpz.toSparse(x);
BOOST_CHECK(x == zs);
auto ipz = i + z;
ipz.toSparse(x);
BOOST_CHECK(x == is);
auto dpz = d + z;
dpz.toSparse(x);
BOOST_CHECK(x == ds);
auto spz = s + z;
spz.toSparse(x);
BOOST_CHECK(x == ss);
}
BOOST_AUTO_TEST_CASE(MultOps)
{
// Setup.
Mat z = Mat(3, 3);
Sp zs(3,3);
Mat i = Mat::createIdentity(3);
Sp is(Eigen::Matrix<double, Eigen::Dynamic, Eigen::Dynamic>::Identity(3,3).sparseView());
Eigen::Array<double, Eigen::Dynamic, 1> d1(3);
d1 << 0.2, 1.2, 13.4;
Mat d = Mat(d1.matrix().asDiagonal());
Sp ds = spdiag(d1);
Eigen::Matrix<double, Eigen::Dynamic, Eigen::Dynamic> s1(3,3);
s1 <<
1.0, 0.0, 2.0,
0.0, 1.0, 0.0,
0.0, 0.0, 2.0;
Sp ss(s1.sparseView());
Mat s = Mat(ss);
// Convert to Eigen::SparseMatrix
Sp x;
z.toSparse(x);
BOOST_CHECK(x == zs);
i.toSparse(x);
BOOST_CHECK(x == is);
d.toSparse(x);
BOOST_CHECK(x == ds);
s.toSparse(x);
BOOST_CHECK(x == ss);
//Multiply by zero matrix
auto ztz = z * z;
ztz.toSparse(x);
BOOST_CHECK(x == zs*zs);
auto itz = i * z;
itz.toSparse(x);
BOOST_CHECK(x == is*zs);
auto dtz = d * z;
dtz.toSparse(x);
BOOST_CHECK(x == ds*zs);
auto stz = s * z;
stz.toSparse(x);
BOOST_CHECK(x == ss*zs);
//Multiply by identity matrix
auto zti = z * i;
zti.toSparse(x);
BOOST_CHECK(x == zs*is);
auto iti = i * i;
iti.toSparse(x);
BOOST_CHECK(x == is*is);
auto dti = d * i;
dti.toSparse(x);
BOOST_CHECK(x == ds*is);
auto sti = s * i;
sti.toSparse(x);
BOOST_CHECK(x == ss*is);
// Multiply by diagonal matrix.
auto ztd = z * d;
ztd.toSparse(x);
BOOST_CHECK(x == zs*ds);
auto itd = i * d;
itd.toSparse(x);
BOOST_CHECK(x == is*ds);
auto dtd = d * d;
dtd.toSparse(x);
BOOST_CHECK(x == ds*ds);
auto std = s * d;
std.toSparse(x);
BOOST_CHECK(x == ss*ds);
// Multiply by sparse matrix.
auto zts = z * s;
zts.toSparse(x);
BOOST_CHECK(x == zs*ss);
auto its = i * s;
its.toSparse(x);
BOOST_CHECK(x == is*ss);
auto dts = d * s;
dts.toSparse(x);
BOOST_CHECK(x == ds*ss);
auto sts = s * s;
sts.toSparse(x);
BOOST_CHECK(x == ss*ss);
}
BOOST_AUTO_TEST_CASE(MultOpsDouble)
{
// Setup.
Mat z = Mat(3, 3);
Sp zs(3,3);
Mat i = Mat::createIdentity(3);
Sp is(Eigen::Matrix<double, Eigen::Dynamic, Eigen::Dynamic>::Identity(3,3).sparseView());
Eigen::Array<double, Eigen::Dynamic, 1> d1(3);
d1 << 0.2, 1.2, 13.4;
Mat d = Mat(d1.matrix().asDiagonal());
Sp ds = spdiag(d1);
Eigen::Matrix<double, Eigen::Dynamic, Eigen::Dynamic> s1(3,3);
s1 <<
1.0, 0.0, 2.0,
0.0, 1.0, 0.0,
0.0, 0.0, 2.0;
Sp ss(s1.sparseView());
Mat s = Mat(ss);
static double factor = 5.3;
Sp x;
auto zd = z*factor;
zd.toSparse(x);
BOOST_CHECK(x == zs*factor);
auto id = i*factor;
id.toSparse(x);
BOOST_CHECK(x == is*factor);
auto dd = d*factor;
dd.toSparse(x);
BOOST_CHECK(x == ds*factor);
auto sd = s*factor;
sd.toSparse(x);
BOOST_CHECK(x == ss*factor);
}
BOOST_AUTO_TEST_CASE(DivOpsDouble)
{
// Setup.
Mat z = Mat(3, 3);
Sp zs(3,3);
Mat i = Mat::createIdentity(3);
Sp is(Eigen::Matrix<double, Eigen::Dynamic, Eigen::Dynamic>::Identity(3,3).sparseView());
Eigen::Array<double, Eigen::Dynamic, 1> d1(3);
d1 << 0.2, 1.2, 13.4;
Mat d = Mat(d1.matrix().asDiagonal());
Sp ds = spdiag(d1);
Eigen::Matrix<double, Eigen::Dynamic, Eigen::Dynamic> s1(3,3);
s1 <<
1.0, 0.0, 2.0,
0.0, 1.0, 0.0,
0.0, 0.0, 2.0;
Sp ss(s1.sparseView());
Mat s = Mat(ss);
static double factor = 5.3;
Sp x;
auto zd = z/factor;
zd.toSparse(x);
Sp tmp = zs/factor;
tmp.prune(1e-16);
BOOST_CHECK(x == tmp);
auto id = i/factor;
id.toSparse(x);
tmp = is/factor;
tmp.prune(1e-16);
BOOST_CHECK(x == tmp);
auto dd = d/factor;
dd.toSparse(x);
tmp = ds/factor;
tmp.prune(1e-16);
BOOST_CHECK(x == tmp);
auto sd = s/factor;
sd.toSparse(x);
tmp = ss/factor;
tmp.prune(1e-16);
BOOST_CHECK(x == tmp);
}
BOOST_AUTO_TEST_CASE(MultVectorXd)
{
Mat z = Mat(3, 3);
Sp zs(3,3);
Mat i = Mat::createIdentity(3);
Sp is(Eigen::Matrix<double, Eigen::Dynamic, Eigen::Dynamic>::Identity(3,3).sparseView());
Eigen::Array<double, Eigen::Dynamic, 1> d1(3);
d1 << 0.2, 1.2, 13.4;
Mat d = Mat(d1.matrix().asDiagonal());
Sp ds = spdiag(d1);
Eigen::Matrix<double, Eigen::Dynamic, Eigen::Dynamic> s1(3,3);
s1 <<
1.0, 0.0, 2.0,
0.0, 1.0, 0.0,
0.0, 0.0, 2.0;
Sp ss(s1.sparseView());
Mat s = Mat(ss);
Eigen::VectorXd vec(3);
vec << 1.0, 2.0, 3.0;
Sp x;
Eigen::VectorXd zd = z*vec;
BOOST_CHECK(zd == zs*vec);
Eigen::VectorXd id = i*vec;
BOOST_CHECK(id == is*vec);
Eigen::VectorXd dd = d*vec;
BOOST_CHECK(dd == ds*vec);
Eigen::VectorXd sd = s*vec;
BOOST_CHECK(sd == ss*vec);
}
BOOST_AUTO_TEST_CASE(Coeff)
{
Eigen::Matrix<double, Eigen::Dynamic, Eigen::Dynamic> s1(3,2);
s1 <<
1.0, 0.0, 2.0,
0.0, 1.0, 0.0;
Sp s2(s1.sparseView());
Mat s = Mat(s2);
for (int row=0; row<s1.rows(); ++row) {
for (int col=0; col<s1.cols(); ++col) {
double a = s.coeff(row, col);
double b = s1(row, col);
BOOST_CHECK_EQUAL(a, b);
}
}
}
BOOST_AUTO_TEST_CASE(nonZeros)
{
Mat z = Mat(3, 3);
Mat i = Mat::createIdentity(3);
Eigen::Array<double, Eigen::Dynamic, 1> d1(3);
d1 << 0.2, 1.2, 13.4;
Mat d = Mat(d1.matrix().asDiagonal());
Sp ds = spdiag(d1);
Eigen::Matrix<double, Eigen::Dynamic, Eigen::Dynamic> s1(3,3);
s1 <<
1.0, 0.0, 2.0,
0.0, 1.0, 0.0,
0.0, 0.0, 2.0;
Sp ss(s1.sparseView());
Mat s = Mat(ss);
BOOST_CHECK_EQUAL(z.nonZeros(), 0);
BOOST_CHECK_EQUAL(i.nonZeros(), 3);
BOOST_CHECK_EQUAL(d.nonZeros(), 3);
BOOST_CHECK_EQUAL(s.nonZeros(), 4);
}