Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Maint/move newton update to function #121

Merged
merged 2 commits into from
Jul 2, 2024
Merged

Maint/move newton update to function #121

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
2 commits
Select commit Hold shift + click to select a range
24d1e37
MAINT: Moved the Newton solve into its own function
NateAM Jul 2, 2024
69364aa
TST: Added tests for the new solveNewtonUpdate function
NateAM Jul 2, 2024
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
77 changes: 44 additions & 33 deletions src/cpp/tardigrade_hydra.cpp
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -1420,68 +1420,79 @@ namespace tardigradeHydra{

}

void hydraBase::solveNonLinearProblem( ){
void hydraBase::solveNewtonUpdate( floatVector &deltaX_tr ){
/*!
* Solve the non-linear problem
* Solve the Newton update returning the trial value of the unknown vector
*
* \param &deltaX_tr: The trial change in the unknown vector
*/

// Form the initial unknown vector
TARDIGRADE_ERROR_TOOLS_CATCH( initializeUnknownVector( ) );

unsigned int rank;

floatVector deltaX( getNumUnknowns( ), 0 );
Eigen::Map< Eigen::Vector< floatType, -1 > > dx_map( deltaX_tr.data( ), getNumUnknowns( ) );

Eigen::Map< Eigen::Vector< floatType, -1 > > dx_map( deltaX.data( ), getUnknownVector( )->size( ) );
tardigradeVectorTools::solverType< floatType > linearSolver;

resetLSIteration( );
Eigen::Map< const Eigen::Matrix< floatType, -1, -1, Eigen::RowMajor > > J_map( getFlatJacobian( )->data( ), getNumUnknowns( ), getNumUnknowns( ) );

while( !checkConvergence( ) && checkIteration( ) ){
Eigen::Map< const Eigen::Vector< floatType, -1 > > R_map( getResidual( )->data( ), getNumUnknowns( ) );

floatVector X0 = *getUnknownVector( );
if ( *getUsePreconditioner( ) ){

tardigradeVectorTools::solverType< floatType > linearSolver;
if( *getPreconditionerIsDiagonal( ) ){

Eigen::Map< const Eigen::Matrix< floatType, -1, -1, Eigen::RowMajor > > J_map( getFlatJacobian( )->data( ), X0.size( ), X0.size( ) );
Eigen::Map< const Eigen::Vector< floatType, -1 > > p_map( getFlatPreconditioner( )->data( ), getNumUnknowns( ) );

Eigen::Map< const Eigen::Vector< floatType, -1 > > R_map( getResidual( )->data( ), X0.size( ) );
linearSolver = tardigradeVectorTools::solverType< floatType >( p_map.asDiagonal( ) * J_map );

if ( *getUsePreconditioner( ) ){
dx_map = -linearSolver.solve( p_map.asDiagonal( ) * R_map );

if( *getPreconditionerIsDiagonal( ) ){
}
else{

Eigen::Map< const Eigen::Vector< floatType, -1 > > p_map( getFlatPreconditioner( )->data( ), X0.size( ) );
Eigen::Map< const Eigen::Matrix< floatType, -1, -1 > > p_map( getFlatPreconditioner( )->data( ), getNumUnknowns( ), getNumUnknowns( ) );

linearSolver = tardigradeVectorTools::solverType< floatType >( p_map.asDiagonal( ) * J_map );
linearSolver = tardigradeVectorTools::solverType< floatType >( p_map * J_map );

dx_map = -linearSolver.solve( p_map.asDiagonal( ) * R_map );
dx_map = -linearSolver.solve( p_map * R_map );

}
else{
}

Eigen::Map< const Eigen::Matrix< floatType, -1, -1 > > p_map( getFlatPreconditioner( )->data( ), X0.size( ), X0.size( ) );
}
else{

linearSolver = tardigradeVectorTools::solverType< floatType >( p_map * J_map );
linearSolver = tardigradeVectorTools::solverType< floatType >( J_map );
dx_map = -linearSolver.solve( R_map );

dx_map = -linearSolver.solve( p_map * R_map );
}

}
rank = linearSolver.rank( );

}
else{
if ( rank != getResidual( )->size( ) ){

linearSolver = tardigradeVectorTools::solverType< floatType >( J_map );
dx_map = -linearSolver.solve( R_map );
TARDIGRADE_ERROR_TOOLS_CATCH( throw convergence_error( "The Jacobian is not full rank" ) );

}
}

rank = linearSolver.rank( );
}

if ( rank != getResidual( )->size( ) ){
void hydraBase::solveNonLinearProblem( ){
/*!
* Solve the non-linear problem
*/

// Form the initial unknown vector
TARDIGRADE_ERROR_TOOLS_CATCH( initializeUnknownVector( ) );

TARDIGRADE_ERROR_TOOLS_CATCH( throw convergence_error( "The Jacobian is not full rank" ) );
floatVector deltaX( getNumUnknowns( ), 0 );

}
resetLSIteration( );

while( !checkConvergence( ) && checkIteration( ) ){

floatVector X0 = *getUnknownVector( );

solveNewtonUpdate( deltaX );

updateUnknownVector( X0 + *getLambda( ) * deltaX );

Expand Down
2 changes: 2 additions & 0 deletions src/cpp/tardigrade_hydra.h
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -966,6 +966,8 @@ namespace tardigradeHydra{

void solveNonLinearProblem( );

void solveNewtonUpdate( floatVector &X_tr );

void setTolerance( const floatVector &tolerance );

void incrementIteration( ){ _iteration++; }
Expand Down
81 changes: 81 additions & 0 deletions src/cpp/tests/test_tardigrade_hydra.cpp
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -351,6 +351,12 @@ namespace tardigradeHydra{

}

static void solveNewtonUpdate( hydraBase &hydra, floatVector &deltaX ){

hydra.solveNewtonUpdate( deltaX );

}

static void solveNonLinearProblem( hydraBase &hydra ){

hydra.solveNonLinearProblem( );
Expand Down Expand Up @@ -3685,6 +3691,81 @@ BOOST_AUTO_TEST_CASE( test_hydraBase_decomposeUnknownVector ){

}

BOOST_AUTO_TEST_CASE( test_hydraBase_solveNewtonUpdate ){

class hydraBaseMock : public tardigradeHydra::hydraBase {

public:

using tardigradeHydra::hydraBase::hydraBase;

floatVector flatJacobian = { 1, 2, 3, 4, 5, 6, 7, 8, 2 };

floatVector residual = { 1, 2, 3 };

virtual const unsigned int getNumUnknowns( ) override{ return residual.size( ); }

protected:

virtual void initializeUnknownVector( ){

tardigradeHydra::unit_test::hydraBaseTester::set_residual( *this, residual );

tardigradeHydra::unit_test::hydraBaseTester::set_flatJacobian( *this, flatJacobian );

}


};

floatType time = 1.1;

floatType deltaTime = 2.2;

floatType temperature = 5.3;

floatType previousTemperature = 23.4;

floatVector deformationGradient = { 0.39293837, -0.42772133, -0.54629709,
0.10262954, 0.43893794, -0.15378708,
0.9615284 , 0.36965948, -0.0381362 };

floatVector previousDeformationGradient = { -0.21576496, -0.31364397, 0.45809941,
-0.12285551, -0.88064421, -0.20391149,
0.47599081, -0.63501654, -0.64909649 };

floatVector previousStateVariables = { 0.53155137, 0.53182759, 0.63440096, 0.84943179, 0.72445532,
0.61102351, 0.72244338, 0.32295891, 0.36178866, 0.22826323,
0.29371405, 0.63097612, 0.09210494, 0.43370117, 0.43086276,
0.4936851 , 0.42583029, 0.31226122, 0.42635131, 0.89338916,
0.94416002, 0.50183668, 0.62395295, 0.1156184 , 0.31728548,
0.41482621, 0.86630916, 0.25045537, 0.48303426, 0.98555979,
0.51948512, 0.61289453, 0.12062867, 0.8263408 , 0.60306013,
0.54506801, 0.34276383, 0.30412079 };

floatVector parameters = { 1, 2, 3, 4, 5 };

unsigned int numConfigurations = 4;

unsigned int numNonLinearSolveStateVariables = 5;

unsigned int dimension = 3;

hydraBaseMock hydra( time, deltaTime, temperature, previousTemperature, deformationGradient, previousDeformationGradient,
{ }, { },
previousStateVariables, parameters, numConfigurations, numNonLinearSolveStateVariables, dimension );

floatVector answer = { 1./3, -2./3, 0 };

floatVector result( 3, 0 );

tardigradeHydra::unit_test::hydraBaseTester::initializeUnknownVector( hydra );
tardigradeHydra::unit_test::hydraBaseTester::solveNewtonUpdate( hydra, result );

BOOST_CHECK( tardigradeVectorTools::fuzzyEquals( result, answer ) );

}

BOOST_AUTO_TEST_CASE( test_hydraBase_solveNonLinearProblem ){

class hydraBaseMock : public tardigradeHydra::hydraBase {
Expand Down
Loading