1
0
mirror of https://github.com/paboyle/Grid.git synced 2024-11-10 07:55:35 +00:00

Covariant laplacian and implicit integration

This commit is contained in:
Guido Cossu 2017-02-20 11:17:27 +00:00
parent bafb101e4f
commit 97a6b61551
9 changed files with 366 additions and 95 deletions

View File

@ -37,25 +37,19 @@ namespace Grid {
namespace QCD {
////////////////////////////////////////////////////////////////////
class RNGModuleParameters: Serializable {
struct RNGModuleParameters: Serializable {
GRID_SERIALIZABLE_CLASS_MEMBERS(RNGModuleParameters,
std::string, serial_seeds,
std::string, parallel_seeds,);
public:
std::vector<int> SerialSeed;
std::vector<int> ParallelSeed;
RNGModuleParameters(const std::vector<int> S = std::vector<int>(),
const std::vector<int> P = std::vector<int>())
: SerialSeed(S), ParallelSeed(P) {}
std::vector<int> getSerialSeeds(){return strToVec<int>(serial_seeds);}
std::vector<int> getParallelSeeds(){return strToVec<int>(parallel_seeds);}
RNGModuleParameters(): serial_seeds("1"), parallel_seeds("1"){}
template <class ReaderClass >
RNGModuleParameters(Reader<ReaderClass>& Reader){
read(Reader, "RandomNumberGenerator", *this);
SerialSeed = strToVec<int>(serial_seeds);
ParallelSeed = strToVec<int>(parallel_seeds);
}
};
@ -82,12 +76,14 @@ public:
GridParallelRNG& get_pRNG() { return *pRNG_.get(); }
void seed() {
if (Params_.SerialSeed.size() == 0 && Params_.ParallelSeed.size() == 0) {
std::cout << "Seeds not initialized" << std::endl;
auto SerialSeeds = Params_.getSerialSeeds();
auto ParallelSeeds = Params_.getParallelSeeds();
if (SerialSeeds.size() == 0 && ParallelSeeds.size() == 0) {
std::cout << GridLogError << "Seeds not initialized" << std::endl;
exit(1);
}
sRNG_.SeedFixedIntegers(Params_.SerialSeed);
pRNG_->SeedFixedIntegers(Params_.ParallelSeed);
sRNG_.SeedFixedIntegers(SerialSeeds);
pRNG_->SeedFixedIntegers(ParallelSeeds);
}
};

View File

@ -75,6 +75,8 @@ class HMCResourceManager {
bool have_RNG;
bool have_CheckPointer;
// NOTE: operator << is not overloaded for std::vector<string>
// so thsi function is necessary
void output_vector_string(const std::vector<std::string> &vs){
for (auto &i: vs)
std::cout << i << " ";
@ -85,13 +87,13 @@ class HMCResourceManager {
public:
HMCResourceManager() : have_RNG(false), have_CheckPointer(false) {}
template <class ReaderClass >
template <class ReaderClass, class vector_type = vComplex >
void initialize(ReaderClass &Read){
// assumes we are starting from the main node
// Geometry
GridModuleParameters GridPar(Read);
GridFourDimModule GridMod( GridPar) ;
GridFourDimModule<vector_type> GridMod( GridPar) ;
AddGrid("gauge", GridMod);
// Checkpointer
@ -100,9 +102,6 @@ class HMCResourceManager {
std::string cp_type;
read(Read,"name", cp_type);
std::cout << "Registered types " << std::endl;
// NOTE: operator << is not overloaded for std::vector<string>
// so it complains here
//std::cout << CPfactory.getBuilderList() << std::endl;
output_vector_string(CPfactory.getBuilderList());
@ -178,7 +177,7 @@ class HMCResourceManager {
// Add a named grid set, 4d shortcut
void AddFourDimGrid(std::string s) {
GridFourDimModule Mod;
GridFourDimModule<vComplex> Mod;
AddGrid(s, Mod);
}

View File

@ -31,41 +31,48 @@ directory
#define HMC_GRID_MODULES
namespace Grid {
namespace QCD {
// Resources
// Modules for grids
class GridModuleParameters: Serializable{
// Introduce another namespace HMCModules?
class GridModuleParameters: Serializable{
public:
GRID_SERIALIZABLE_CLASS_MEMBERS(GridModuleParameters,
std::string, lattice,
std::string, mpi);
std::string, mpi);
public:
// these namings are ugly
// also ugly the distinction between the serializable members
// and this
std::vector<int> lattice_v;
std::vector<int> mpi_v;
std::vector<int> getLattice(){return strToVec<int>(lattice);}
std::vector<int> getMpi() {return strToVec<int>(mpi);}
GridModuleParameters(const std::vector<int> l_ = std::vector<int>(),
const std::vector<int> mpi_ = std::vector<int>()):lattice_v(l_), mpi_v(mpi_){}
template <class ReaderClass>
GridModuleParameters(Reader<ReaderClass>& Reader) {
read(Reader, "LatticeGrid", *this);
lattice_v = strToVec<int>(lattice);
mpi_v = strToVec<int>(mpi);
if (mpi_v.size() != lattice_v.size()) {
std::cout << "Error in GridModuleParameters: lattice and mpi dimensions "
void check(){
if (getLattice().size() != getMpi().size()) {
std::cout << GridLogError
<< "Error in GridModuleParameters: lattice and mpi dimensions "
"do not match"
<< std::endl;
exit(1);
}
}
template <class ReaderClass>
GridModuleParameters(Reader<ReaderClass>& Reader) {
read(Reader, name, *this);
check();
}
// Save on file
template< class WriterClass>
void save(Writer<WriterClass>& Writer){
check();
write(Writer, name, *this);
}
private:
std::string name = "LatticeGrid";
};
// Lower level class
class GridModule {
public:
GridCartesian* get_full() {
@ -84,27 +91,33 @@ class GridModule {
};
// helpers
// FIXME define a class accepting also real vtypes
////////////////////////////////////
// Classes for the user
////////////////////////////////////
// Note: the space time grid must be out of the QCD namespace
template< class vector_type>
class GridFourDimModule : public GridModule {
public:
// add a function to create the module from a Reader
GridFourDimModule() {
using namespace QCD;
set_full(SpaceTimeGrid::makeFourDimGrid(
GridDefaultLatt(), GridDefaultSimd(4, vComplex::Nsimd()),
GridDefaultLatt(), GridDefaultSimd(4, vector_type::Nsimd()),
GridDefaultMpi()));
set_rb(SpaceTimeGrid::makeFourDimRedBlackGrid(grid_.get()));
}
template <class vector_type = vComplex>
GridFourDimModule(GridModuleParameters Params) {
if (Params.lattice_v.size() == 4) {
using namespace QCD;
Params.check();
std::vector<int> lattice_v = Params.getLattice();
std::vector<int> mpi_v = Params.getMpi();
if (lattice_v.size() == 4) {
set_full(SpaceTimeGrid::makeFourDimGrid(
Params.lattice_v, GridDefaultSimd(4, vector_type::Nsimd()),
Params.mpi_v));
lattice_v, GridDefaultSimd(4, vector_type::Nsimd()),
mpi_v));
set_rb(SpaceTimeGrid::makeFourDimRedBlackGrid(grid_.get()));
} else {
std::cout
std::cout << GridLogError
<< "Error in GridFourDimModule: lattice dimension different from 4"
<< std::endl;
exit(1);
@ -112,10 +125,9 @@ class GridFourDimModule : public GridModule {
}
};
typedef GridFourDimModule<vComplex> GridDefaultFourDimModule;
} // namespace QCD
} // namespace Grid
#endif // HMC_GRID_MODULES

View File

@ -131,6 +131,49 @@ class Integrator {
as[level].apply(update_P_hireps, Representations, Mom, U, ep);
}
void implicit_update_P(MomentaField& Mom, Field& U, int level, double ep) {
// Fundamental updates, include smearing
MomentaField Msum(Mom._grid);
Msum = zero;
for (int a = 0; a < as[level].actions.size(); ++a) {
// Compute the force
// We need to compute the derivative of the actions
// only once
Field force(U._grid);
conformable(U._grid, Mom._grid);
Field& Us = Smearer.get_U(as[level].actions.at(a)->is_smeared);
as[level].actions.at(a)->deriv(Us, force); // deriv should NOT include Ta
std::cout << GridLogIntegrator << "Smearing (on/off): " << as[level].actions.at(a)->is_smeared << std::endl;
if (as[level].actions.at(a)->is_smeared) Smearer.smeared_force(force);
force = FieldImplementation::projectForce(force); // Ta for gauge fields
Real force_abs = std::sqrt(norm2(force)/U._grid->gSites());
std::cout << GridLogIntegrator << "Force average: " << force_abs << std::endl;
Msum += force;
}
MomentaField NewMom = Mom;
MomentaField OldMom = Mom;
double threshold = 1e-6;
// Here run recursively
do{
MomentaField MomDer(Mom._grid);
OldMom = NewMom;
// Compute the derivative of the kinetic term
// with respect to the gauge field
// Laplacian.Mder(NewMom, MomDer);
// NewMom = Mom - ep*(MomDer + Msum);
} while (norm2(NewMom - OldMom) > threshold);
Mom = NewMom;
// update the auxiliary fields momenta
}
void update_U(Field& U, double ep) {
update_U(P, U, ep);

View File

@ -285,6 +285,74 @@ class ForceGradient : public Integrator<FieldImplementation, SmearingPolicy,
}
}
};
////////////////////////////////
// Riemannian Manifold HMC
// Girolami et al
////////////////////////////////
template <class FieldImplementation, class SmearingPolicy,
class RepresentationPolicy =
Representations<FundamentalRepresentation> >
class ImplicitLeapFrog : public Integrator<FieldImplementation, SmearingPolicy,
RepresentationPolicy> {
public:
typedef ImplicitLeapFrog<FieldImplementation, SmearingPolicy, RepresentationPolicy>
Algorithm;
INHERIT_FIELD_TYPES(FieldImplementation);
// Riemannian manifold metric operator
// Hermitian operator Fisher
std::string integrator_name(){return "ImplicitLeapFrog";}
ImplicitLeapFrog(GridBase* grid, IntegratorParameters Par,
ActionSet<Field, RepresentationPolicy>& Aset, SmearingPolicy& Sm)
: Integrator<FieldImplementation, SmearingPolicy, RepresentationPolicy>(
grid, Par, Aset, Sm){};
void step(Field& U, int level, int _first, int _last) {
int fl = this->as.size() - 1;
// level : current level
// fl : final level
// eps : current step size
// Get current level step size
RealD eps = this->Params.trajL/this->Params.MDsteps;
for (int l = 0; l <= level; ++l) eps /= this->as[l].multiplier;
int multiplier = this->as[level].multiplier;
for (int e = 0; e < multiplier; ++e) {
int first_step = _first && (e == 0);
int last_step = _last && (e == multiplier - 1);
if (first_step) { // initial half step
this->implicit_update_P(U, level, eps / 2.0);
}
if (level == fl) { // lowest level
this->update_U(U, eps);
} else { // recursive function call
this->step(U, level + 1, first_step, last_step);
}
int mm = last_step ? 1 : 2;
this->update_P(U, level, mm * eps / 2.0);
}
}
};
}
}

View File

@ -33,13 +33,27 @@ directory
namespace Grid {
namespace QCD {
////////////////////////////////////////////////////////////
// Laplacian operator L on adjoint fields
//
// phi: adjoint field
// L: D_mu^dag D_mu
//
// L phi(x) = Sum_mu [ U_mu(x)phi(x+mu)U_mu(x)^dag +
// U_mu(x-mu)^dag phi(x-mu)U_mu(x-mu)
// -2phi(x)]
//
// Operator designed to be encapsulated by
// an HermitianLinearOperator<.. , ..>
////////////////////////////////////////////////////////////
template <class Impl>
class LaplacianAdjointField {
public:
INHERIT_GIMPL_TYPES(Impl);
typedef SU<Nc>::LatticeAlgebraVector AVector;
LaplacianAdjointField(GridBase* grid) : U(Nd, grid){};
LaplacianAdjointField(GridBase* grid, const RealD k = 1.0) :
U(Nd, grid), kappa(k){};
void ImportGauge(const GaugeField& _U) {
for (int mu = 0; mu < Nd; mu++) {
@ -49,61 +63,48 @@ class LaplacianAdjointField {
void Mdiag(const GaugeLinkField& in, GaugeLinkField& out) { assert(0); }
void Mdir(const GaugeLinkField& in, GaugeLinkField& out, int dir, int disp) { assert(0); }
/*
// Operator with algebra vector inputs and outputs
void M2(const AVector& in, AVector& out) {
double kappa = 0.9;
//Reconstruct matrix
GaugeLinkField tmp(in._grid);
GaugeLinkField tmp2(in._grid);
GaugeLinkField sum(in._grid);
GaugeLinkField out_mat(in._grid);
GaugeLinkField in_mat(in._grid);
SU<Nc>::FundamentalLieAlgebraMatrix(in, in_mat);
sum = zero;
for (int mu = 0; mu < Nd; mu++) {
tmp = U[mu] * Cshift(in_mat, mu, +1) * adj(U[mu]);
tmp2 = adj(U[mu]) * in_mat * U[mu];
sum += tmp + Cshift(tmp2, mu, -1) - 2.0 * in_mat;
}
out_mat = (1.0 - kappa) * in_mat - kappa/(double(4*Nd)) * sum;
// Project
SU<Nc>::projectOnAlgebra(out, out_mat);
void Mdir(const GaugeLinkField& in, GaugeLinkField& out, int dir, int disp) {
assert(0);
}
*/
void M(const GaugeLinkField& in, GaugeLinkField& out) {
double kappa = 0.999;
//Reconstruct matrix
void M(const GaugeLinkField& in, GaugeLinkField& out) {
GaugeLinkField tmp(in._grid);
GaugeLinkField tmp2(in._grid);
GaugeLinkField sum(in._grid);
sum = zero;
for (int mu = 0; mu < Nd; mu++) {
tmp = U[mu] * Cshift(in, mu, +1) * adj(U[mu]);
tmp = U[mu] * Cshift(in, mu, +1) * adj(U[mu]);
tmp2 = adj(U[mu]) * in * U[mu];
sum += tmp + Cshift(tmp2, mu, -1) - 2.0 * in;
}
out = (1.0 - kappa) * in - kappa/(double(4*Nd)) * sum;
out = (1.0 - kappa) * in - kappa / (double(4 * Nd)) * sum;
}
void MDeriv(const GaugeLinkField& in, GaugeLinkField& out, bool dag){
RealD factor = - kappa / (double(4 * Nd))
if (!dag)
out = factor * Cshift(in, mu, +1) * adj(U[mu]) + adj(U[mu]) * in;
else
out = factor * U[mu] * Cshift(in, mu, +1) + in * U[mu];
}
private:
RealD kappa;
std::vector<GaugeLinkField> U;
};
// This is just a debug tests
// not meant to be used
template <class Impl>
class LaplacianAlgebraField {
public:
INHERIT_GIMPL_TYPES(Impl);
typedef SU<Nc>::LatticeAlgebraVector AVector;
LaplacianAlgebraField(GridBase* grid) : U(Nd, grid){};
LaplacianAlgebraField(GridBase* grid, const RealD k) :
U(Nd, grid), kappa(k){};
void ImportGauge(const GaugeField& _U) {
for (int mu = 0; mu < Nd; mu++) {
@ -117,28 +118,28 @@ class LaplacianAlgebraField {
// Operator with algebra vector inputs and outputs
void M(const AVector& in, AVector& out) {
double kappa = 0.999;
//Reconstruct matrix
GaugeLinkField tmp(in._grid);
GaugeLinkField tmp2(in._grid);
GaugeLinkField sum(in._grid);
GaugeLinkField out_mat(in._grid);
GaugeLinkField in_mat(in._grid);
// Reconstruct matrix
SU<Nc>::FundamentalLieAlgebraMatrix(in, in_mat);
sum = zero;
for (int mu = 0; mu < Nd; mu++) {
tmp = U[mu] * Cshift(in_mat, mu, +1) * adj(U[mu]);
tmp = U[mu] * Cshift(in_mat, mu, +1) * adj(U[mu]);
tmp2 = adj(U[mu]) * in_mat * U[mu];
sum += tmp + Cshift(tmp2, mu, -1) - 2.0 * in_mat;
}
out_mat = (1.0 - kappa) * in_mat - kappa/(double(4*Nd)) * sum;
out_mat = (1.0 - kappa) * in_mat - kappa / (double(4 * Nd)) * sum;
// Project
SU<Nc>::projectOnAlgebra(out, out_mat);
}
private:
RealD kappa;
std::vector<GaugeLinkField> U;
};

View File

@ -59,8 +59,8 @@ int main(int argc, char **argv) {
TheHMC.Resources.LoadBinaryCheckpointer(CPparams);
RNGModuleParameters RNGpar;
RNGpar.SerialSeed = {1,2,3,4,5};
RNGpar.ParallelSeed = {6,7,8,9,10};
RNGpar.serial_seeds = "1 2 3 4 5";
RNGpar.parallel_seeds = "6 7 8 9 10";
TheHMC.Resources.SetRNGSeeds(RNGpar);
// Construct observables

View File

@ -0,0 +1,147 @@
/*************************************************************************************
Grid physics library, www.github.com/paboyle/Grid
Source file: ./tests/Test_hmc_WilsonFermionGauge.cc
Copyright (C) 2015
Author: Peter Boyle <pabobyle@ph.ed.ac.uk>
Author: neo <cossu@post.kek.jp>
Author: Guido Cossu <guido.cossu@ed.ac.uk>
This program 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 2 of the License, or
(at your option) any later version.
This program 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 this program; if not, write to the Free Software Foundation, Inc.,
51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA.
See the full license in the file "LICENSE" in the top level distribution
directory
*************************************************************************************/
/* END LEGAL */
#include <Grid/Grid.h>
int main(int argc, char **argv) {
using namespace Grid;
using namespace Grid::QCD;
Grid_init(&argc, &argv);
int threads = GridThread::GetThreads();
// here make a routine to print all the relevant information on the run
std::cout << GridLogMessage << "Grid is setup to use " << threads << " threads" << std::endl;
// Typedefs to simplify notation
typedef GenericHMCRunner<MinimumNorm2> HMCWrapper; // Uses the default minimum norm
typedef WilsonImplR FermionImplPolicy;
typedef WilsonTMFermionR FermionAction;
typedef typename FermionAction::FermionField FermionField;
//::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::
HMCWrapper TheHMC;
// Grid from the command line
TheHMC.Resources.AddFourDimGrid("gauge");
// Possibile to create the module by hand
// hardcoding parameters or using a Reader
// Checkpointer definition
CheckpointerParameters CPparams;
CPparams.config_prefix = "ckpoint_lat";
CPparams.rng_prefix = "ckpoint_rng";
CPparams.saveInterval = 5;
CPparams.format = "IEEE64BIG";
TheHMC.Resources.LoadBinaryCheckpointer(CPparams);
RNGModuleParameters RNGpar;
RNGpar.SerialSeed = {1,2,3,4,5};
RNGpar.ParallelSeed = {6,7,8,9,10};
TheHMC.Resources.SetRNGSeeds(RNGpar);
// Construct observables
// here there is too much indirection
PlaquetteObsParameters PlPar;
PlPar.output_prefix = "Plaquette";
PlaquetteMod<HMCWrapper::ImplPolicy> PlaqModule(PlPar);
TheHMC.Resources.AddObservable(&PlaqModule);
//////////////////////////////////////////////
/////////////////////////////////////////////////////////////
// Collect actions, here use more encapsulation
// need wrappers of the fermionic classes
// that have a complex construction
// standard
RealD beta = 3.9 ;
SymanzikGaugeActionR Waction(beta);
auto GridPtr = TheHMC.Resources.GetCartesian();
auto GridRBPtr = TheHMC.Resources.GetRBCartesian();
// temporarily need a gauge field
LatticeGaugeField U(GridPtr);
Real mass = -0.89163;
Real mu = 0.01;
// Can we define an overloaded operator that does not need U and initialises
// it with zeroes?
FermionAction FermOp(U, *GridPtr, *GridRBPtr, mass, mu);
ConjugateGradient<FermionField> CG(1.0e-8, 2000);
TwoFlavourPseudoFermionAction<FermionImplPolicy> Nf2(FermOp, CG, CG);
// With modules
/*
TwoFlavourFmodule<FermionImplPolicy> TwoFMod(Reader);
*/
// Set smearing (true/false), default: false
Nf2.is_smeared = false;
// Collect actions
ActionLevel<HMCWrapper::Field> Level1(1);
Level1.push_back(&Nf2);
ActionLevel<HMCWrapper::Field> Level2(4);
Level2.push_back(&Waction);
TheHMC.TheAction.push_back(Level1);
TheHMC.TheAction.push_back(Level2);
/////////////////////////////////////////////////////////////
/*
double rho = 0.1; // smearing parameter
int Nsmear = 2; // number of smearing levels
Smear_Stout<HMCWrapper::ImplPolicy> Stout(rho);
SmearedConfiguration<HMCWrapper::ImplPolicy> SmearingPolicy(
UGrid, Nsmear, Stout);
*/
// HMC parameters are serialisable
TheHMC.Parameters.MD.MDsteps = 20;
TheHMC.Parameters.MD.trajL = 1.0;
TheHMC.ReadCommandLine(argc, argv); // these can be parameters from file
TheHMC.Run(); // no smearing
// TheHMC.Run(SmearingPolicy); // for smearing
Grid_finalize();
} // main

View File

@ -48,28 +48,33 @@ int main (int argc, char ** argv)
LatticeGaugeField Umu(&Grid);
SU<Nc>::HotConfiguration(pRNG,Umu);
double Kappa = 0.9999;
typedef SU<Nc>::LatticeAlgebraVector AVector;
// Source and result in the algebra
AVector src_vec(&Grid); random(pRNG, src_vec);
AVector result_vec(&Grid); result_vec = zero;
LatticeColourMatrix src(&Grid);
SU<Nc>::FundamentalLieAlgebraMatrix(src_vec, src);
LatticeColourMatrix result(&Grid); result=zero;
LaplacianAdjointField<PeriodicGimplR> Laplacian(&Grid);
LaplacianAdjointField<PeriodicGimplR> Laplacian(&Grid, Kappa);
Laplacian.ImportGauge(Umu);
HermitianLinearOperator<LaplacianAdjointField<PeriodicGimplR>,LatticeColourMatrix> HermOp(Laplacian);
ConjugateGradient<LatticeColourMatrix> CG(1.0e-8,10000);
std::cout << GridLogMessage << "Testing the Laplacian using the full matrix" <<std::endl;
CG(HermOp,src,result); // fastest
// Tests also the version using the algebra decomposition
LaplacianAlgebraField<PeriodicGimplR> LaplacianAlgebra(&Grid);
LaplacianAlgebraField<PeriodicGimplR> LaplacianAlgebra(&Grid, Kappa);
LaplacianAlgebra.ImportGauge(Umu);
HermitianLinearOperator<LaplacianAlgebraField<PeriodicGimplR>,AVector> HermOpAlg(LaplacianAlgebra);
ConjugateGradient<AVector> CG_Algebra(1.0e-8,10000);
std::cout << GridLogMessage << "Testing the Laplacian using the algebra vectors" <<std::endl;
CG_Algebra(HermOpAlg,src_vec,result_vec);
LatticeColourMatrix result2(&Grid);