Added Sympiler Wrapper #34
Conversation
| #ifdef POLYSOLVE_WITH_SYMPILER | ||
|
|
||
| #include <polysolve/LinearSolver.hpp> | ||
| #include <sympiler/parsy/cholesky_solver.h> |
There was a problem hiding this comment.
Let's move this header to the .cpp. This will require moving the definition of the constructor/destructor to the .cpp of this file as well.
There was a problem hiding this comment.
Is this possible given the member variables of type sym_lib::parsy::CSC and sym_lib::parsy::SolverSettings?
There was a problem hiding this comment.
I took a crack at it -- let me know how it looks
|
|
||
| std::unique_ptr<sym_lib::parsy::SolverSettings> m_solver; | ||
| std::unique_ptr<sym_lib::parsy::CSC> m_A_csc; | ||
| polysolve::StiffnessMatrix m_A_copy; |
There was a problem hiding this comment.
I am curious if we really need to store a copy of the matrix. Is this done in other solvers as well?
There was a problem hiding this comment.
Not sure about other solvers, but itseems like sympiler needs a SolverSettings object, which takes a pointer directly to the values and indices of a matrix. The pointers aren't const, so I don't think there's a guarantee that the solver object won't mess with the matrix data? That's why I've been doing the copy
| int solver_mode = 0; // 0 is normal solve, 1 is row/col addition | ||
| int factorize_status = -1; |
There was a problem hiding this comment.
I'd vote for using m_ prefix for member variables consistently.
No description provided.