//fix phase of eigensystem and store phase of first entry of each eigenvector
void fix_phase(Eigen::MatrixXcd& V, Eigen::MatrixXcd& V_fix, std::vector<double>& phase) {
const int V3 = pars -> get_int("V3");
//helper variables:
//Number of eigenvectors
int n_ev;
//negative imaginary
std::complex<double> i_neg (0,-1);
//tmp factor and phase
std::complex<double> fac (1.,1.);
double tmp_phase = 0;
//get sizes right, resize if necessary
n_ev = V.cols();
if (phase.size() != n_ev) phase.resize(n_ev);
if (V_fix.cols() != n_ev) V_fix.resize(3*V3,n_ev);
//loop over all eigenvectors of system
for (int n = 0; n < n_ev; ++n) {
tmp_phase = std::arg(V(0,n));
phase.at(n) = tmp_phase;
fac = std::exp(i_neg*tmp_phase);
//Fix phase of eigenvector with negative polar angle of first entry
V_fix.col(n) = fac * V.col(n);
}
}
//TODO complex values as matrix entries too
//TODO support endomorphisms over Rn too
Expression EigenvectorsCommand::operator()(const QList< Analitza::Expression >& args)
{
Expression ret;
QStringList errors;
const Eigen::MatrixXcd eigeninfo = executeEigenSolver(args, true, errors);
if (!errors.isEmpty()) {
ret.addError(errors.first());
return ret;
}
const int neigenvectors = eigeninfo.rows();
QScopedPointer<Analitza::List> list(new Analitza::List);
for (int j = 0; j < neigenvectors; ++j) {
const Eigen::VectorXcd col = eigeninfo.col(j);
QScopedPointer<Analitza::Vector> eigenvector(new Analitza::Vector(neigenvectors));
for (int i = 0; i < neigenvectors; ++i) {
const std::complex<double> eigenvalue = col(i);
const double realpart = eigenvalue.real();
const double imagpart = eigenvalue.imag();
if (std::isnan(realpart) || std::isnan(imagpart)) {
ret.addError(QCoreApplication::tr("Returned eigenvalue is NaN", "NaN means Not a Number, is an invalid float number"));
return ret;
} else if (std::isinf(realpart) || std::isinf(imagpart)) {
ret.addError(QCoreApplication::tr("Returned eigenvalue is too big"));
return ret;
} else {
bool isonlyreal = true;
if (std::isnormal(imagpart)) {
isonlyreal = false;
}
Analitza::Cn * eigenvalueobj = 0;
if (isonlyreal) {
eigenvalueobj = new Analitza::Cn(realpart);
} else {
eigenvalueobj = new Analitza::Cn(realpart, imagpart);
}
eigenvector->appendBranch(eigenvalueobj);
}
}
list->appendBranch(eigenvector.take());
}
ret.setTree(list.take());
return ret;
}
// -----------------------------------------------------------------------------
// -----------------------------------------------------------------------------
void LapH::OperatorsForMesons::build_rvdaggerv(
const LapH::RandomVector& rnd_vec) {
// check if vdaggerv is already build
if(not is_vdaggerv_set){
std::cout << "\n\n\tCaution: vdaggerv is not set and rvdaggervr cannot be"
<< " computed\n\n" << std::endl;
exit(0);
}
clock_t t2 = clock();
std::cout << "\tbuild rvdaggerv:";
for(auto& rvdv_level1 : rvdaggerv)
for(auto& rvdv_level2 : rvdv_level1)
for(auto& rvdv_level3 : rvdv_level2)
rvdv_level3 = Eigen::MatrixXcd::Zero(4*dilE, nb_ev);
#pragma omp parallel for schedule(dynamic)
for(size_t t = 0; t < Lt; t++){
// rvdaggervr is calculated by multiplying vdaggerv with the same quantum
// numbers with random vectors from right and left.
for(const auto& op : operator_lookuptable.rvdaggerv_lookuptable){
Eigen::MatrixXcd vdv;
if(op.need_vdaggerv_daggering == false)
vdv = vdaggerv[op.id_vdaggerv][t];
else
vdv = vdaggerv[op.id_vdaggerv][t].adjoint();
size_t rid = 0;
for(const auto& rnd_id :
operator_lookuptable.ricQ1_lookup[op.id_ricQ_lookup].rnd_vec_ids){
for(size_t block = 0; block < 4; block++){
for(size_t vec_i = 0; vec_i < nb_ev; ++vec_i) {
size_t blk = block + vec_i * 4 + 4 * nb_ev * t;
rvdaggerv[op.id][t][rid].block(vec_i%dilE + dilE*block, 0, 1, nb_ev) +=
vdv.row(vec_i) * std::conj(rnd_vec(rnd_id, blk));
}}
rid++;
}
}}// time and operator loops end here
t2 = clock() - t2;
std::cout << std::setprecision(1) << "\t\tSUCCESS - " << std::fixed
<< ((float) t2)/CLOCKS_PER_SEC << " seconds" << std::endl;
}
//Reads in Eigenvectors from one Timeslice in binary format to V
void read_evectors_bin_ts(const char * path, const char* prefix, const int config_i, const int t,
const int nb_ev, Eigen::MatrixXcd& V) {
int V3 = pars -> get_int("V3");
//bool thorough = pars -> get_int("strict");
const int dim_row = 3 * V3;
//TODO: Change path getting to something keyword independent
//buffer for read in
std::complex<double>* eigen_vec = new std::complex<double>[dim_row];
//setting up file
char filename[200];
sprintf(filename, "%s/%s.%04d.%03d", path, prefix, config_i, t);
std::cout << "Reading file: " << filename << std::endl;
std::ifstream infile(filename, std::ifstream::binary);
for (int nev = 0; nev < nb_ev; ++nev) {
infile.read( reinterpret_cast<char*> (eigen_vec), 2*dim_row*sizeof(double));
V.col(nev) = Eigen::Map<Eigen::VectorXcd, 0 >(eigen_vec, dim_row);
eof_check(t,nev,nb_ev,infile.eof());
}
if(check_trace(V, nb_ev) != true){
std::cout << "Timeslice: " << t << ": Eigenvectors damaged, exiting" << std::endl;
exit(0);
}
//clean up
delete[] eigen_vec;
infile.close();
}
void write_eig_sys_bin(const char* prefix, const int config_i, const int t, const int nb_ev, Eigen::MatrixXcd& V) {
const int V3 = pars -> get_int("V3");
std::string path = pars -> get_path("res");
//set up filename
char file [200];
sprintf(file, "%s/%s.%04d.%03d", path.c_str(), prefix, config_i, t);
//sprintf(file, "%s.%04d.%03d", prefix, config_i, t);
if(check_trace(V, nb_ev) != true){
std::cout << "Timeslice: " << t << ": Eigenvectors damaged, abort writing" << std::endl;
exit(1);
}
std::cout << "Writing to file:" << file << std::endl;
std::ofstream outfile(file, std::ofstream::binary);
std::streamsize begin = outfile.tellp();
std::streamsize eigsys_bytes =2*3*V3*nb_ev*sizeof(double);
outfile.write(reinterpret_cast<char*> (V.data()), eigsys_bytes);
std::streamsize end = outfile.tellp();
if ( (end - begin)/eigsys_bytes != 1 ){
std::cout << "Timeslice: " << t << " Error: write incomplete, exiting" << std::endl;
std::cout << (end-begin) << " bytes instead of expected "<< eigsys_bytes << " bytes" << std::endl;
exit(1);
}
//std::cout << end - begin << " bytes written" << std::endl;
outfile.close();
}
int main(int argc, char ** argv) {
// initialize parameters /////////////////////////////////////////////////////
Parameters p;
#ifdef DEBUG
cout << "Memory usage of p: " << sizeof(p) << " bytes." << endl;
Eigen::MatrixXcd H = p.Ham();
cout << "size of Hamiltonian: " << H.rows() << "x" << H.cols() << endl;
cout << "Hamiltonian: " << H << endl;
cout << "Second element of Hamiltonian: " << p.Ham(0,1) << endl;
#endif
// propagate /////////////////////////////////////////////////////////////////
return 0;
}
// TODO: work on interface with eigenvector class
// transform matrix of eigenvectors with gauge array
Eigen::MatrixXcd GaugeField::trafo_ev(const Eigen::MatrixXcd &eig_sys) {
const ssize_t dim_row = eig_sys.rows();
const ssize_t dim_col = eig_sys.cols();
Eigen::MatrixXcd ret(dim_row, dim_col);
if (omega.shape()[0] == 0)
build_gauge_array(1);
// write_gauge_matrices("ev_trafo_log.bin",Omega);
for (ssize_t nev = 0; nev < dim_col; ++nev) {
for (ssize_t vol = 0; vol < dim_row; ++vol) {
int ind_c = vol % 3;
Eigen::Vector3cd tmp =
omega[0][ind_c].adjoint() * (eig_sys.col(nev)).segment(ind_c, 3);
(ret.col(nev)).segment(ind_c, 3) = tmp;
} // end loop nev
} // end loop vol
return ret;
}
// This returns a matrix that is the lower-triangular part (only column
// index <= row index) of the square of matrixToSquare.
Eigen::MatrixXcd SymmetricComplexMassMatrix::LowerTriangleOfSquareMatrix(
Eigen::MatrixXcd const& matrixToSquare ) const
{
Eigen::MatrixXcd valuesSquaredMatrix( numberOfRows,
numberOfRows );
for( size_t rowIndex( 0 );
rowIndex < numberOfRows;
++rowIndex )
{
for( size_t columnIndex( 0 );
columnIndex <= rowIndex;
++columnIndex )
{
valuesSquaredMatrix.coeffRef( rowIndex,
columnIndex ).real(0.0);
valuesSquaredMatrix.coeffRef( rowIndex,
columnIndex ).imag(0.0);
for( size_t sumIndex( 0 );
sumIndex < numberOfRows;
++sumIndex )
{
double temp = valuesSquaredMatrix.coeffRef( rowIndex,
columnIndex ).real()
+ ( ( matrixToSquare.coeff( sumIndex,
rowIndex ).real()
* matrixToSquare.coeff( sumIndex,
columnIndex ).real() )
+ ( matrixToSquare.coeff( sumIndex,
rowIndex ).imag()
* matrixToSquare.coeff( sumIndex,
columnIndex ).imag() ) );
valuesSquaredMatrix.coeffRef( rowIndex, columnIndex ).real(temp);
temp = valuesSquaredMatrix.coeffRef( rowIndex,
columnIndex ).imag()
+ ( ( matrixToSquare.coeff( sumIndex,
rowIndex ).real()
* matrixToSquare.coeff( sumIndex,
columnIndex ).imag() )
- ( matrixToSquare.coeff( sumIndex,
rowIndex ).imag()
* matrixToSquare.coeff( sumIndex,
columnIndex ).real() ) );
valuesSquaredMatrix.coeffRef( rowIndex,
columnIndex ).imag(temp);
// The Eigen routines don't bother looking at elements of
// valuesSquaredMatrix where columnIndex > rowIndex, so we don't even
// bother filling them with the conjugates of the transpose.
}
}
}
return valuesSquaredMatrix;
}
void BasicOperator::mult_dirac(const Eigen::MatrixXcd& matrix,
Eigen::MatrixXcd& reordered,
const size_t index) const {
const vec_pdg_Corr op_Corr = global_data->get_lookup_corr();
const size_t rows = matrix.rows();
const size_t cols = matrix.cols();
const size_t col = cols/4;
if( (rows != reordered.rows()) || (cols != reordered.cols()) ){
std::cout << "Error! In BasicOperator::mult_dirac: size of matrix and "
"reordered must be equal" << std::endl;
exit(0);
}
for(const auto& dirac : op_Corr[index].gamma){
if(dirac != 4){
for(size_t block = 0; block < 4; block++){
reordered.block(0, block * col, rows, col) =
gamma[dirac].value[block] *
matrix.block(0, gamma[dirac].row[block]*col, rows, col);
}
}
}
}
int main() {
const int dim_row = MAT_ENTRIES;
const int dim_col = 120;// number of eigenvectors
//Set up data structure
Eigen::MatrixXcd V;
Eigen::MatrixXcd V_dagger;
Eigen::MatrixXcd S;
V = Eigen::MatrixXcd::Zero( dim_row, dim_col);
V_dagger = Eigen::MatrixXcd::Zero( dim_row, dim_col);
S = Eigen::MatrixXcd::Zero( dim_row, dim_col);
read_eigenvectors_from_binary_ts("eigenvectors", 600, 0, V);
//read_eigenvectors_from_binary_ts("eigenvectors", 600, 0);
std::cout << "Calculating v v_dagger" << std::endl;
V_dagger = V.adjoint();
for(int i = 0; i < dim_row; ++i ){
for(int j = 0; j < dim_col; ++j) {
//std::complex<double> entry = V.row(i) * V_dagger.col(j);
if (std::abs(V(i,j)) > 1e-15)
std::cout << i << " " << j << " " << V(i,j) << std::endl;
}
}
}
static bool check_trace(const Eigen::MatrixXcd& V, const int nb_ev){
bool read_state = true;
Eigen::MatrixXcd VdV(nb_ev,nb_ev);
VdV = V.adjoint() * V;
double eps = 10e-10;
std::complex<double> trace (0.,0.), sum(0.,0.);
trace = VdV.trace();
sum = VdV.sum();
std::cout << trace.real() << std::endl;
if ( fabs( trace.real()) - nb_ev > eps ||
fabs(trace.imag()) > eps ||
fabs(sum.real()) - nb_ev > eps ||
fabs(sum.imag()) > eps)
read_state = false;
return read_state;
}
void CCorrelationFilters::build_otsdf(struct CDataStruct *img, struct CParamStruct *params, struct CFilterStruct *filt)
{
/*
* This function implements the correlation filter design proposed in the following publications.
*
* [1] Optimal trade-off synthetic discriminant function filters for arbitrary devices, B.V.K.Kumar, D.W.Carlson, A.Mahalanobis - Optics Letters, 1994.
*
* [2] Jason Thornton, "Matching deformed and occluded iris patterns: a probabilistic model based on discriminative cues." PhD thesis, Carnegie Mellon University, Pittsburgh, PA, USA, 2007.
*
* [3] Vishnu Naresh Boddeti, Jonathon M Smereka, and B. V. K. Vijaya Kumar, "A comparative evaluation of iris and ocular recognition methods on challenging ocular images." IJCB, 2011.
*
* [4] A. Mahalanobis, B.V.K. Kumar, D. Casasent, "Minimum average correlation energy filters," Applied Optics, 1987
*
* Notes: This filter design is good when the dimensionality of the data is greater than the training sample size. Setting the filter parameter params->alpha=0 results in the popular MACE filter. However, it is usually better to set alpha to a small number rather than setting it to 0. If you use MACE cite [4].
*/
filt->params = *params;
filt->filter.size_data = params->size_filt_freq;
filt->filter.size_data_freq = params->size_filt_freq;
filt->filter.num_elements_freq = img->num_elements_freq;
params->num_elements_freq = img->num_elements_freq;
filt->filter.data_freq = new complex<double>[img->num_elements_freq*img->num_channels];
Eigen::ArrayXcd filt_freq = Eigen::ArrayXcd::Zero(params->num_elements_freq*img->num_channels);
// If not set default to 1
if (params->wpos < 1) params->wpos = 1;
filt->params.wpos = params->wpos;
compute_psd_matrix(img, params);
Eigen::MatrixXcd Y = Eigen::MatrixXcd::Zero(img->num_elements_freq*img->num_channels,img->num_data);
Eigen::MatrixXcd u = Eigen::MatrixXcd::Zero(img->num_data,1);
Eigen::MatrixXcd temp = Eigen::MatrixXcd::Zero(img->num_data,img->num_data);
Eigen::MatrixXd tmp = Eigen::MatrixXd::Zero(img->num_data,img->num_data);
Eigen::Map<Eigen::MatrixXcd> X(img->data_freq,img->num_elements_freq*img->num_channels,img->num_data);
Eigen::ArrayXXcd temp1 = Eigen::ArrayXXcd::Zero(img->num_elements_freq,img->num_channels);
Eigen::ArrayXXcd temp2 = Eigen::ArrayXXcd::Zero(img->num_elements_freq,img->num_channels);
Eigen::Vector2i num_blocks_1, num_blocks_2;
num_blocks_1 << img->num_channels,img->num_channels;
num_blocks_2 << img->num_channels,1;
for (int k=0;k<img->num_data;k++){
temp2 = X.block(0,k,img->num_elements_freq*img->num_channels,1).array();
temp2.resize(img->num_elements_freq,img->num_channels);
fusion_matrix_multiply(temp1, img->Sinv, temp2, num_blocks_1, num_blocks_2);
temp1.resize(img->num_elements_freq*img->num_channels,1);
Y.block(0,k,img->num_elements_freq*img->num_channels,1) = temp1.matrix();
temp1.resize(img->num_elements_freq,img->num_channels);
if (img->labels[k] == 1)
{
u(k) = std::complex<double>(params->wpos,0);
}
else
{
u(k) = std::complex<double>(1,0);
}
}
temp = X.conjugate().transpose()*Y;
temp = temp.ldlt().solve(u);
filt_freq = Y*temp;
Y.resize(0,0);
Eigen::Map<Eigen::ArrayXcd>(filt->filter.data_freq,img->num_elements_freq*img->num_channels) = filt_freq;
filt->filter.num_data = 1;
filt->filter.num_channels = img->num_channels;
filt->filter.ptr_data.reserve(filt->filter.num_data);
filt->filter.ptr_data_freq.reserve(filt->filter.num_data);
ifft_data(&filt->filter);
fft_data(&filt->filter);
}
std::string cmplxMatrixString(const Eigen::MatrixXcd& m)
{
// Make a string from the bytes
return std::string((char *) m.data(), m.size() * sizeof(m(0, 0)));
}
inline int size_pn() const { return coef_iat_ipn.cols(); }
inline int size_at() const { return coef_iat_ipn.rows(); }
// -----------------------------------------------------------------------------
// -----------------------------------------------------------------------------
void LapH::OperatorsForMesons::build_rvdaggervr(
const LapH::RandomVector& rnd_vec) {
// check of vdaggerv is already build
if(not is_vdaggerv_set){
std::cout << "\n\n\tCaution: vdaggerv is not set and rvdaggervr cannot be"
<< " computed\n\n" << std::endl;
exit(0);
}
clock_t t2 = clock();
std::cout << "\tbuild rvdaggervr:";
for(auto& rvdvr_level1 : rvdaggervr)
for(auto& rvdvr_level2 : rvdvr_level1)
for(auto& rvdvr_level3 : rvdvr_level2)
rvdvr_level3 = Eigen::MatrixXcd::Zero(4*dilE, 4*dilE);
#pragma omp parallel for schedule(dynamic)
for(size_t t = 0; t < Lt; t++){
// rvdaggervr is calculated by multiplying vdaggerv with the same quantum
// numbers with random vectors from right and left.
for(const auto& op : operator_lookuptable.rvdaggervr_lookuptable){
Eigen::MatrixXcd vdv;
if(op.need_vdaggerv_daggering == false)
vdv = vdaggerv[op.id_vdaggerv][t];
else
vdv = vdaggerv[op.id_vdaggerv][t].adjoint();
size_t rid = 0;
int check = -1;
Eigen::MatrixXcd M; // Intermediate memory
for(const auto& rnd_id :
operator_lookuptable.ricQ2_lookup[op.id_ricQ_lookup].rnd_vec_ids){
if(check != rnd_id.first){ // this avoids recomputation
M = Eigen::MatrixXcd::Zero(nb_ev, 4*dilE);
for(size_t block = 0; block < 4; block++){
for(size_t vec_i = 0; vec_i < nb_ev; vec_i++) {
size_t blk = block + (vec_i + nb_ev * t) * 4;
M.block(0, vec_i%dilE + dilE*block, nb_ev, 1) +=
vdv.col(vec_i) * rnd_vec(rnd_id.first, blk);
}}
}
for(size_t block_x = 0; block_x < 4; block_x++){
for(size_t block_y = 0; block_y < 4; block_y++){
for(size_t vec_y = 0; vec_y < nb_ev; ++vec_y) {
size_t blk = block_y + (vec_y + nb_ev * t) * 4;
rvdaggervr[op.id][t][rid].block(
dilE*block_y + vec_y%dilE, dilE*block_x, 1, dilE) +=
M.block(vec_y, dilE*block_x, 1, dilE) *
std::conj(rnd_vec(rnd_id.second, blk));
}}}
check = rnd_id.first;
rid++;
}
}}// time and operator loops end here
t2 = clock() - t2;
std::cout << std::setprecision(1) << "\t\tSUCCESS - " << std::fixed
<< ((float) t2)/CLOCKS_PER_SEC << " seconds" << std::endl;
}
// -----------------------------------------------------------------------------
// -----------------------------------------------------------------------------
void LapH::Quarklines::build_Q2L(const Perambulator& peram,
const OperatorsForMesons& meson_operator,
const std::vector<QuarklineQ2Indices>& ql_lookup,
const std::vector<RandomIndexCombinationsQ2>& ric_lookup){
std::cout << "\tcomputing Q2L:";
clock_t time = clock();
#pragma omp parallel
{
Eigen::MatrixXcd M = Eigen::MatrixXcd::Zero(4 * dilE, 4 * nev);
for(size_t t1 = 0; t1 < Lt; t1++){
for(size_t t2 = 0; t2 < Lt/dilT; t2++){
for(size_t op = 0; op < ql_lookup.size(); op++){
size_t nb_rnd = ric_lookup[(ql_lookup[op]).
id_ric_lookup].rnd_vec_ids.size();
for(size_t rnd1 = 0; rnd1 < nb_rnd; rnd1++){
Q2L[t1][t2][op][rnd1].setZero();
}
}
}}
#pragma omp for schedule(dynamic)
for(size_t t1 = 0; t1 < Lt; t1++){
for(const auto& qll : ql_lookup){
size_t rnd_counter = 0;
int check = -1;
for(const auto& rnd_id : ric_lookup[qll.id_ric_lookup].rnd_vec_ids){
if(check != rnd_id.first){ // this avoids recomputation
for(size_t row = 0; row < 4; row++){
for(size_t col = 0; col < 4; col++){
if(!qll.need_vdaggerv_dag)
M.block(col*dilE, row*nev, dilE, nev) =
peram[rnd_id.first].block((t1*4 + row)*nev,
(t1/dilT*4 + col)*dilE,
nev, dilE).adjoint() *
meson_operator.return_vdaggerv(qll.id_vdaggerv, t1);
else
M.block(col*dilE, row*nev, dilE, nev) =
peram[rnd_id.first].block((t1*4 + row)*nev,
(t1/dilT*4 + col)*dilE,
nev, dilE).adjoint() *
meson_operator.return_vdaggerv(qll.id_vdaggerv, t1).adjoint();
// gamma_5 trick
if( ((row + col) == 3) || (abs(row - col) > 1) )
M.block(col*dilE, row*nev, dilE, nev) *= -1.;
}}
}
for(size_t t2 = 0; t2 < Lt/dilT; t2++){
Q2L[t1][t2][qll.id][rnd_counter].setZero(4*dilE, 4*dilE);
const size_t gamma_id = qll.gamma[0];
for(size_t block_dil = 0; block_dil < 4; block_dil++) {
const cmplx value = gamma[gamma_id].value[block_dil];
const size_t gamma_index = gamma[gamma_id].row[block_dil];
for(size_t row = 0; row < 4; row++){
for(size_t col = 0; col < 4; col++){
Q2L[t1][t2][qll.id][rnd_counter].
block(row*dilE, col*dilE, dilE, dilE) +=
value *
M.block(row*dilE, block_dil*nev, dilE, nev) *
peram[rnd_id.second].block(
(t1*4 + gamma_index)*nev,
(t2*4 + col)*dilE, nev, dilE);
}}
}
}
check = rnd_id.first;
rnd_counter++;
}
}}
} // pragma omp ends
time = clock() - time;
std::cout << "\t\t\tSUCCESS - " << ((float) time) / CLOCKS_PER_SEC
<< " seconds" << std::endl;
}
void CCorrelationFilters::build_mmcf(struct CDataStruct *img, struct CParamStruct *params, struct CFilterStruct *filt)
{
/*
* This function calls the correlation filter design proposed in the following publications.
*
* A. Rodriguez, Vishnu Naresh Boddeti, B.V.K. Vijaya Kumar and A. Mahalanobis, "Maximum Margin Correlation Filter: A New Approach for Localization and Classification", IEEE Transactions on Image Processing, 2012.
*
* Vishnu Naresh Boddeti, "Advances in Correlation Filters: Vector Features, Structured Prediction and Shape Alignment" PhD thesis, Carnegie Mellon University, Pittsburgh, PA, USA, 2012.
*
* Vishnu Naresh Boddeti and B.V.K. Vijaya Kumar, "Maximum Margin Vector Correlation Filters," Arxiv 1404.6031 (April 2014).
*
* Notes: This currently the best performing Correlation Filter design, especially when the training sample size is larger than the dimensionality of the data.
*/
filt->params = *params;
filt->filter.size_data = params->size_filt_freq;
filt->filter.size_data_freq = params->size_filt_freq;
filt->filter.num_elements_freq = img->num_elements_freq;
params->num_elements_freq = img->num_elements_freq;
filt->filter.data_freq = new complex<double>[img->num_elements_freq*img->num_channels];
Eigen::ArrayXcd filt_freq = Eigen::ArrayXcd::Zero(params->num_elements_freq*img->num_channels);
// If not set default to 1
if (params->wpos < 1) params->wpos = 1;
filt->params.wpos = params->wpos;
compute_psd_matrix(img, params);
Eigen::MatrixXcd Y = Eigen::MatrixXcd::Zero(img->num_elements_freq*img->num_channels,img->num_data);
Eigen::MatrixXcd u = Eigen::MatrixXcd::Zero(img->num_data,1);
Eigen::MatrixXd temp = Eigen::MatrixXd::Zero(img->num_data,img->num_data);
Eigen::Map<Eigen::MatrixXcd> X(img->data_freq,img->num_elements_freq*img->num_channels,img->num_data);
Eigen::ArrayXXcd temp1 = Eigen::ArrayXXcd::Zero(img->num_elements_freq,img->num_channels);
Eigen::ArrayXXcd temp2 = Eigen::ArrayXXcd::Zero(img->num_elements_freq,img->num_channels);
Eigen::Vector2i num_blocks_1, num_blocks_2;
num_blocks_1 << img->num_channels,img->num_channels;
num_blocks_2 << img->num_channels,1;
for (int k=0;k<img->num_data;k++){
temp2 = X.block(0,k,img->num_elements_freq*img->num_channels,1).array();
temp2.resize(img->num_elements_freq,img->num_channels);
fusion_matrix_multiply(temp1, img->Sinv, temp2, num_blocks_1, num_blocks_2);
temp1.resize(img->num_elements_freq*img->num_channels,1);
Y.block(0,k,img->num_elements_freq*img->num_channels,1) = temp1.matrix();
temp1.resize(img->num_elements_freq,img->num_channels);
if (img->labels[k] == 1)
{
u(k) = std::complex<double>(params->wpos,0);
}
else
{
u(k) = std::complex<double>(-1,0);
}
}
esvm::SVMClassifier libsvm;
libsvm.setC(params->C);
libsvm.setKernel(params->kernel_type);
libsvm.setWpos(params->wpos);
temp = (X.conjugate().transpose()*Y).real();
Eigen::Map<Eigen::MatrixXd> y(img->labels,img->num_data,1);
libsvm.train(temp, y);
temp.resize(0,0);
int nSV;
libsvm.getNSV(&nSV);
Eigen::VectorXi sv_indices = Eigen::VectorXi::Zero(nSV);
Eigen::VectorXd sv_coef = Eigen::VectorXd::Zero(nSV);
libsvm.getSI(sv_indices);
libsvm.getCoeff(sv_coef);
for (int k=0; k<nSV; k++) {
filt_freq += (Y.block(0,sv_indices[k]-1,img->num_elements_freq*img->num_channels,1)*sv_coef[k]).array();
}
Y.resize(0,0);
Eigen::Map<Eigen::ArrayXcd>(filt->filter.data_freq,img->num_elements_freq*img->num_channels) = filt_freq;
filt->filter.num_data = 1;
filt->filter.num_channels = img->num_channels;
filt->filter.ptr_data.reserve(filt->filter.num_data);
filt->filter.ptr_data_freq.reserve(filt->filter.num_data);
ifft_data(&filt->filter);
fft_data(&filt->filter);
}
// -----------------------------------------------------------------------------
// -----------------------------------------------------------------------------
void LapH::distillery::create_source(const size_t dil_t, const size_t dil_e,
std::complex<double>* source) {
if(dil_t >= param.dilution_size_so[0] ||
dil_e >= param.dilution_size_so[1] ){
std::cout << "dilution is out of bounds in \"create_source\"" <<
std::endl;
exit(0);
}
const size_t Lt = param.Lt;
const size_t Ls = param.Ls;
const size_t number_of_eigen_vec = param.nb_ev;
const size_t Vs = Ls*Ls*Ls;
const size_t dim_row = Vs*3;
int numprocs = 0, myid = 0;
MPI_Comm_size(MPI_COMM_WORLD, &numprocs);
MPI_Comm_rank(MPI_COMM_WORLD, &myid);
// setting source to zero
for(size_t dil_d = 0; dil_d < param.dilution_size_so[2]; ++dil_d)
for(size_t i = 0; i < 12*Vs*Lt/numprocs; ++i)
source[dil_d*12*Vs*Lt/numprocs + i] = {0.0, 0.0};
// indices of timeslices with non-zero entries
size_t nb_of_nonzero_t = Lt/param.dilution_size_so[0];
// TODO: think about smart pointer here!
size_t* t_index = new size_t[nb_of_nonzero_t];
create_dilution_lookup(nb_of_nonzero_t, param.dilution_size_so[0],
dil_t, param.dilution_type_so[0], t_index);
// indices of eigenvectors to be combined
size_t nb_of_ev_combined = number_of_eigen_vec/param.dilution_size_so[1];
size_t* ev_index = new size_t[nb_of_ev_combined];
create_dilution_lookup(nb_of_ev_combined, param.dilution_size_so[1],
dil_e, param.dilution_type_so[1], ev_index);
// indices of Dirac components to be combined
// TODO: This is needed only once - could be part of class
size_t nb_of_dirac_combined = 4/param.dilution_size_so[2];
size_t** d_index = new size_t*[param.dilution_size_so[2]];
for(size_t i = 0; i < param.dilution_size_so[2]; ++i)
d_index[i] = new size_t[nb_of_dirac_combined];
for(size_t dil_d = 0; dil_d < param.dilution_size_so[2]; ++dil_d)
create_dilution_lookup(nb_of_dirac_combined, param.dilution_size_so[2],
dil_d, param.dilution_type_so[2], d_index[dil_d]);
// creating the source
// running over nonzero timeslices
for(size_t t = 0; t < nb_of_nonzero_t; ++t){
// intermidiate memory
Eigen::MatrixXcd S = Eigen::MatrixXcd::Zero(dim_row, 4);
size_t time = t_index[t]; // helper index
if((int) (time / (Lt/numprocs)) == myid) {
size_t time_id = time % (Lt/numprocs); // helper index
// building source on one timeslice
for(size_t ev = 0; ev < nb_of_ev_combined; ++ev){
size_t ev_h = ev_index[ev] * 4; // helper index
for(size_t d = 0; d < 4; ++d){
S.col(d) += random_vector[time](ev_h+d) *
(V[time_id]).col(ev_index[ev]);
}
}
// copy the created source into output array
size_t t_h = time_id*Vs; // helper index
for(size_t x = 0; x < Vs; ++x){
size_t x_h = (t_h + x)*12; // helper index
size_t x_h2 = x*3; // helper index
for(size_t d2 = 0; d2 < param.dilution_size_so[2]; ++d2){
for(size_t d3 = 0; d3 < nb_of_dirac_combined; ++d3){
size_t d_h = x_h + d_index[d2][d3]*3; // helper index
for(size_t c = 0; c < 3; ++c){
source[d2*12*Vs*Lt/numprocs + d_h + c] =
S(x_h2 + c, d_index[d2][d3]);
}
}
}
}
}
} // end of loop over nonzero timeslices
MPI_Barrier(MPI_COMM_WORLD);
// freeing memory
delete[] t_index;
delete[] ev_index;
for(size_t i = 0; i < param.dilution_size_so[2]; ++i)
delete[] d_index[i];
delete[] d_index;
t_index = NULL;
ev_index = NULL;
d_index = NULL;
}
void BasicOperator::init_operator(const char dilution,
const LapH::VdaggerV& vdaggerv,
const LapH::Perambulator& peram){
const int Lt = global_data->get_Lt();
const size_t nb_ev = global_data->get_number_of_eigen_vec();
const std::vector<quark> quarks = global_data->get_quarks();
const size_t nb_rnd = quarks[0].number_of_rnd_vec;
const size_t dilE = quarks[0].number_of_dilution_E;
const int dilT = quarks[0].number_of_dilution_T;
const size_t Q2_size = 4 * dilE;
const vec_pdg_Corr op_Corr = global_data->get_lookup_corr();
const size_t nb_op = op_Corr.size();
std::cout << "\n" << std::endl;
clock_t time = clock();
#pragma omp parallel
{
Eigen::MatrixXcd M = Eigen::MatrixXcd::Zero(Q2_size, 4 * nb_ev);
#pragma omp for schedule(dynamic)
for(int t_0 = 0; t_0 < Lt; t_0++){
if(omp_get_thread_num() == 0)
std::cout << "\tcomputing double quarkline: "
<< std::setprecision(2) << (float) t_0/Lt*100 << "%\r"
<< std::flush;
for(const auto& op : op_Corr){
for(size_t rnd_i = 0; rnd_i < nb_rnd; ++rnd_i) {
for(int t = 0; t < Lt/dilT; t++){
// new momentum -> recalculate M[0]
// M only depends on momentum and displacement. first_vdv
// prevents repeated calculation for different gamma structures
if(op.first_vdv == true){
for(size_t col = 0; col < 4; ++col) {
for(size_t row = 0; row < 4; ++row) {
if(op.negative_momentum == false){
M.block(dilE * col, nb_ev * row, dilE, nb_ev) =
(peram(1, rnd_i).block(nb_ev * (4 * t_0 + row),
dilE * (4 * t + col),
nb_ev, dilE)).adjoint() *
vdaggerv.return_vdaggerv(op.id_vdv, t_0);
}
else {
M.block(dilE * col, nb_ev * row, dilE, nb_ev) =
(peram(1, rnd_i).block(nb_ev * (4 * t_0 + row),
dilE * (4 * t + col),
nb_ev, dilE)).adjoint() *
// TODO: calculate V^daggerV Omega from op.negative_momentum
// == false and multiply Omega from the left
// and then (V^daggerV Omega)^dagger * Omega
(vdaggerv.return_vdaggerv(op.id_vdv, t_0)).adjoint();
}
// gamma_5 trick. It changes the sign of the two upper right and two
// lower left blocks in dirac space
if( ((row + col) == 3) || (abs(row - col) > 1) )
M.block(dilE * col, row * nb_ev, dilE, nb_ev) *= -1.;
}}// loops over row and col end here
}//if over same gamma structure ends here
for(int ti = 0; ti < 3; ti++){
// getting the neighbour blocks
const int tend = (Lt/dilT+t + ti - 1)%(Lt/dilT);
for(size_t rnd_j = 0; rnd_j < nb_rnd; ++rnd_j) {
if(rnd_i != rnd_j){
//dilution of d-quark from left
for(size_t block_dil = 0; block_dil < 4; block_dil++){
cmplx value = 1.;
value_dirac(op.id, block_dil, value);
for(size_t col = 0; col < 4; col++){
for(size_t row = 0; row < 4; row++){
Q2[t_0][t][ti][op.id][rnd_i][rnd_j]
.block(row*dilE, col*dilE, dilE, dilE) += value *
M.block(row*dilE, block_dil* nb_ev, dilE, nb_ev) *
peram(0, rnd_j)
.block(4*nb_ev*t_0 + order_dirac(op.id, block_dil)*nb_ev,
Q2_size*tend + col*dilE, nb_ev, dilE);
}}}//dilution ends here
}}}// loops over rnd_j and ti block end here
}// loop over t ends here
}// loop over rnd_i ends here
}//loop operators
}// loops over t_0 ends here
}// pragma omp ends
std::cout << "\tcomputing double quarkline: 100.00%" << std::endl;
time = clock() - time;
std::cout << "\t\tSUCCESS - " << ((float) time) / CLOCKS_PER_SEC
<< " seconds" << std::endl;
}
