Files
solver/GNN/OndselSolver/SparseMatrix.h
forbes-0023 98051ba0c9 feat: add Phase 1 constraint solver addon, move prior content to GNN/
- Move existing OndselSolver, GNN ML layer, and tooling into GNN/
  directory for integration in later phases
- Add Create addon scaffold: package.xml, Init.py
- Add expression DAG with eval, symbolic diff, simplification
- Add parameter table with fixed/free variable tracking
- Add quaternion rotation as polynomial Expr trees
- Add RigidBody entity (7 DOF: position + unit quaternion)
- Add constraint classes: Coincident, DistancePointPoint, Fixed
- Add Newton-Raphson solver with symbolic Jacobian + numpy lstsq
- Add pre-solve passes: substitution + single-equation
- Add DOF counting via Jacobian SVD rank
- Add KindredSolver IKCSolver bridge for kcsolve integration
- Add 82 unit tests covering all modules

Registers as 'kindred' solver via kcsolve.register_solver() when
loaded by Create's addon_loader.
2026-02-20 20:35:47 -06:00

252 lines
7.3 KiB
C++

/***************************************************************************
* Copyright (c) 2023 Ondsel, Inc. *
* *
* This file is part of OndselSolver. *
* *
* See LICENSE file for details about copyright. *
***************************************************************************/
#pragma once
#include <sstream>
#include "RowTypeMatrix.h"
#include "SparseRow.h"
#include "DiagonalMatrix.h"
#include "FullMatrix.h"
namespace MbD {
template<typename T>
class SparseMatrix;
using SpMatDsptr = std::shared_ptr<SparseMatrix<double>>;
template<typename T>
using SpMatsptr = std::shared_ptr<SparseMatrix<T>>;
class GESpMatParPvPrecise;
template<typename T>
class SparseMatrix : public RowTypeMatrix<SpRowsptr<T>>
{
public:
SparseMatrix(size_t m) : RowTypeMatrix<SpRowsptr<T>>(m)
{
}
SparseMatrix(size_t m, size_t n) {
for (size_t i = 0; i < m; i++)
{
auto row = std::make_shared<SparseRow<T>>(n);
this->push_back(row);
}
}
SparseMatrix(std::initializer_list<std::initializer_list<std::initializer_list<double>>> list2D) {
for (auto& rowList : list2D)
{
auto row = std::make_shared<SparseRow<T>>(rowList);
this->push_back(row);
}
}
void atiput(size_t i, SpRowsptr<T> spRow);
void atijplusDiagonalMatrix(size_t i, size_t j, DiagMatDsptr diagMat);
void atijminusDiagonalMatrix(size_t i, size_t j, DiagMatDsptr diagMat);
double sumOfSquares() override;
void zeroSelf() override;
void atijplusFullRow(size_t i, size_t j, FRowsptr<T> fullRow);
void atijplusFullColumn(size_t i, size_t j, FColsptr<T> fullCol);
void atijplusFullMatrix(size_t i, size_t j, FMatDsptr fullMat);
void atijminusFullMatrix(size_t i, size_t j, FMatDsptr fullMat);
void atijplusTransposeFullMatrix(size_t i, size_t j, FMatDsptr fullMat);
void atijplusFullMatrixtimes(size_t i, size_t j, FMatDsptr fullMat, T factor);
void atijminusFullColumn(size_t i, size_t j, FColDsptr fullCol);
void atijminusTransposeFullMatrix(size_t i, size_t j, FMatDsptr fullMat);
void atijplusNumber(size_t i, size_t j, double value);
void atijminusNumber(size_t i, size_t j, double value);
void atijput(size_t i, size_t j, T value);
double maxMagnitude() override;
FColsptr<T> timesFullColumn(FColsptr<T> fullCol);
SpMatsptr<T> plusSparseMatrix(SpMatsptr<T> spMat);
std::shared_ptr<SparseMatrix<T>> clonesptr();
void magnifySelf(T factor);
std::ostream& printOn(std::ostream& s) const override;
};
template<typename T>
inline void SparseMatrix<T>::atiput(size_t i, SpRowsptr<T> spRow)
{
this->at(i) = spRow;
}
template<typename T>
inline void SparseMatrix<T>::atijplusDiagonalMatrix(size_t i, size_t j, DiagMatDsptr diagMat)
{
auto n = diagMat->nrow();
for (size_t ii = 0; ii < n; ii++)
{
this->atijplusNumber(i + ii, j + ii, diagMat->at(ii));
}
}
template<>
inline void SparseMatrix<double>::atijminusDiagonalMatrix(size_t i1, size_t j1, DiagMatDsptr diagMat)
{
auto n = diagMat->nrow();
for (size_t ii = 0; ii < n; ii++)
{
this->atijminusNumber(i1 + ii, j1 + ii, diagMat->at(ii));
}
}
template<typename T>
inline double SparseMatrix<T>::sumOfSquares()
{
double sum = 0.0;
for (size_t i = 0; i < this->size(); i++)
{
sum += this->at(i)->sumOfSquares();
}
return sum;
}
template<>
inline void SparseMatrix<double>::zeroSelf()
{
for (size_t i = 0; i < this->size(); i++) {
this->at(i)->zeroSelf();
}
}
template<typename T>
inline void SparseMatrix<T>::atijplusFullRow(size_t i, size_t j, FRowsptr<T> fullRow)
{
this->at(i)->atiplusFullRow(j, fullRow);
}
template<typename T>
inline void SparseMatrix<T>::atijplusFullColumn(size_t i, size_t j, FColsptr<T> fullCol)
{
for (size_t ii = 0; ii < fullCol->size(); ii++)
{
this->atijplusNumber(i + ii, j, fullCol->at(ii));
}
}
template<typename T>
inline void SparseMatrix<T>::atijminusFullColumn(size_t i, size_t j, FColDsptr fullCol)
{
for (size_t ii = 0; ii < fullCol->size(); ii++)
{
this->atijminusNumber(i + ii, j, fullCol->at(ii));
}
}
template<typename T>
inline void SparseMatrix<T>::atijplusFullMatrix(size_t i, size_t j, FMatDsptr fullMat)
{
for (size_t ii = 0; ii < fullMat->nrow(); ii++)
{
this->at(i + ii)->atiplusFullRow(j, fullMat->at(ii));
}
}
template<typename T>
inline void SparseMatrix<T>::atijminusFullMatrix(size_t i, size_t j, FMatDsptr fullMat)
{
for (size_t ii = 0; ii < fullMat->nrow(); ii++)
{
this->at(i + ii)->atiminusFullRow(j, fullMat->at(ii));
}
}
template<typename T>
inline void SparseMatrix<T>::atijplusTransposeFullMatrix(size_t i, size_t j, FMatDsptr fullMat)
{
for (size_t ii = 0; ii < fullMat->nrow(); ii++)
{
this->atijplusFullColumn(i, j + ii, fullMat->at(ii)->transpose());
}
}
template<typename T>
inline void SparseMatrix<T>::atijminusTransposeFullMatrix(size_t i, size_t j, FMatDsptr fullMat)
{
for (size_t ii = 0; ii < fullMat->nrow(); ii++)
{
this->atijminusFullColumn(i, j + ii, fullMat->at(ii)->transpose());
}
}
template<typename T>
inline void SparseMatrix<T>::atijplusFullMatrixtimes(size_t i, size_t j, FMatDsptr fullMat, T factor)
{
for (size_t ii = 0; ii < fullMat->nrow(); ii++)
{
this->at(i + ii)->atiplusFullRowtimes(j, fullMat->at(ii), factor);
}
}
template<typename T>
inline void SparseMatrix<T>::atijplusNumber(size_t i, size_t j, double value)
{
this->at(i)->atiplusNumber(j, value);
}
template<typename T>
inline void SparseMatrix<T>::atijminusNumber(size_t i, size_t j, double value)
{
this->at(i)->atiminusNumber(j, value);
}
template<typename T>
inline void SparseMatrix<T>::atijput(size_t i, size_t j, T value)
{
this->at(i)->atiput(j, value);
}
template<typename T>
inline double SparseMatrix<T>::maxMagnitude()
{
double max = 0.0;
for (size_t i = 0; i < this->size(); i++)
{
double element = this->at(i)->maxMagnitude();
if (max < element) max = element;
}
return max;
}
template<typename T>
inline std::ostream& SparseMatrix<T>::printOn(std::ostream& s) const
{
s << "SpMat[" << std::endl;
for (size_t i = 0; i < this->size(); i++)
{
s << *(this->at(i)) << std::endl;
}
s << "]" << std::endl;
return s;
}
template<typename T>
inline FColsptr<T> SparseMatrix<T>::timesFullColumn(FColsptr<T> fullCol)
{
//"a*b = a(i,j)b(j) sum j."
auto nrow = this->nrow();
auto answer = std::make_shared<FullColumn<T>>(nrow);
for (size_t i = 0; i < nrow; i++)
{
answer->at(i) = this->at(i)->timesFullColumn(fullCol);
}
return answer;
}
template<typename T>
inline SpMatsptr<T> SparseMatrix<T>::plusSparseMatrix(SpMatsptr<T> spMat)
{
//"a + b."
//"Assume all checking of validity of this operation has been done."
//"Just evaluate quickly."
auto answer = clonesptr();
for (size_t i = 0; i < answer->size(); i++)
{
answer->atiput(i, answer->at(i)->plusSparseRow(spMat->at(i)));
}
return answer;
}
template<typename T>
inline std::shared_ptr<SparseMatrix<T>> SparseMatrix<T>::clonesptr()
{
return std::make_shared<SparseMatrix<T>>(*this);
}
template<typename T>
inline void SparseMatrix<T>::magnifySelf(T factor)
{
for (size_t i = 0; i < this->size(); i++) {
this->at(i)->magnifySelf(factor);
}
}
}