coreneuron::scopmath::sparse Namespace Reference

Functions
void	delete_item (Item *item)
void	linkitem (Item *item, Item *ii)
void	insert (SparseObj *so, Item *item)
void	increase_order (SparseObj *so, unsigned row)
template<enabled_code code_to_enable = enabled_code::all>
Elm *	getelm (SparseObj *so, unsigned row, unsigned col, Elm *new_elem)
	Return pointer to (row, col) element maintaining order in rows. More...
List *	newlist ()
	The following routines support the concept of a list. More...
void	freelist (List *list)
void	check_assert (SparseObj *so)
void	re_link (SparseObj *so, unsigned i)
void	free_elm (SparseObj *so)
void	init_minorder (SparseObj *so)
void	reduce_order (SparseObj *so, unsigned row)
void	get_next_pivot (SparseObj *so, unsigned i)
void	initeqn (SparseObj *so, unsigned maxeqn)
void	spar_minorder (SparseObj *so)
	Minimum ordering algorithm to determine the order that the matrix should be solved. More...
void	init_coef_list (SparseObj *so, int _iml)
void	subrow (SparseObj *so, Elm *pivot, Elm *rowsub, int _iml)
void	bksub (SparseObj *so, int _iml)
int	matsol (SparseObj *so, int _iml)
template<typename SPFUN >
void	create_coef_list (SparseObj *so, int n, SPFUN fun, _threadargsproto_)
template<enabled_code code_to_enable = enabled_code::all>
double *	thread_getelm (SparseObj *so, int row, int col, int _iml)

Function Documentation

bksub()

void coreneuron::scopmath::sparse::bksub ( SparseObj *  so, int  _iml  )

inline

Definition at line 453 of file sparse_thread.hpp.

check_assert()

void coreneuron::scopmath::sparse::check_assert ( SparseObj *  so )

inline

Definition at line 205 of file sparse_thread.hpp.

create_coef_list()

template<typename SPFUN >

void coreneuron::scopmath::sparse::create_coef_list	(	SparseObj *	so,
		int	n,
		SPFUN	fun,
		_threadargsproto_
	)

Definition at line 487 of file sparse_thread.hpp.

delete_item()

void coreneuron::scopmath::sparse::delete_item ( Item *  item )

inline

Definition at line 20 of file sparse_thread.hpp.

free_elm()

void coreneuron::scopmath::sparse::free_elm ( SparseObj *  so )

inline

Definition at line 294 of file sparse_thread.hpp.

freelist()

void coreneuron::scopmath::sparse::freelist ( List *  list )

inline

Definition at line 196 of file sparse_thread.hpp.

get_next_pivot()

void coreneuron::scopmath::sparse::get_next_pivot ( SparseObj *  so, unsigned  i  )

inline

Definition at line 345 of file sparse_thread.hpp.

getelm()

template<enabled_code code_to_enable = enabled_code::all>

Elm* coreneuron::scopmath::sparse::getelm	(	SparseObj *	so,
		unsigned	row,
		unsigned	col,
		Elm *	new_elem
	)

Return pointer to (row, col) element maintaining order in rows.

See check_assert in minorder for info about how this matrix is supposed to look. If new_elem is nonzero and an element would otherwise be created, new is used instead. This is because linking an element is highly nontrivial. The biggest difference is that elements are no longer removed and this saves much time allocating and freeing during the solve phase.

Definition at line 71 of file sparse_thread.hpp.

increase_order()

void coreneuron::scopmath::sparse::increase_order ( SparseObj *  so, unsigned  row  )

inline

Definition at line 51 of file sparse_thread.hpp.

init_coef_list()

void coreneuron::scopmath::sparse::init_coef_list ( SparseObj *  so, int  _iml  )

inline

Definition at line 426 of file sparse_thread.hpp.

init_minorder()

void coreneuron::scopmath::sparse::init_minorder ( SparseObj *  so )

inline

Definition at line 302 of file sparse_thread.hpp.

initeqn()

void coreneuron::scopmath::sparse::initeqn ( SparseObj *  so, unsigned  maxeqn  )

inline

Definition at line 379 of file sparse_thread.hpp.

insert()

void coreneuron::scopmath::sparse::insert ( SparseObj *  so, Item *  item  )

inline

Definition at line 35 of file sparse_thread.hpp.

linkitem()

void coreneuron::scopmath::sparse::linkitem ( Item *  item, Item *  ii  )

inline

Definition at line 28 of file sparse_thread.hpp.

matsol()

int coreneuron::scopmath::sparse::matsol ( SparseObj *  so, int  _iml  )

inline

Definition at line 466 of file sparse_thread.hpp.

newlist()

List* coreneuron::scopmath::sparse::newlist ( )

inline

The following routines support the concept of a list.

Modified from modl. The list is a doubly linked list. A special item with element 0 is always at the tail of the list and is denoted as the List pointer itself. list->next point to the first item in the list and list->prev points to the last item in the list. i.e. the list is circular. Note that in an empty list next and prev points to itself.

It is intended that this implementation be hidden from the user via the following function calls.

Definition at line 188 of file sparse_thread.hpp.

re_link()

void coreneuron::scopmath::sparse::re_link ( SparseObj *  so, unsigned  i  )

inline

Definition at line 233 of file sparse_thread.hpp.

reduce_order()

void coreneuron::scopmath::sparse::reduce_order ( SparseObj *  so, unsigned  row  )

inline

Definition at line 334 of file sparse_thread.hpp.

spar_minorder()

void coreneuron::scopmath::sparse::spar_minorder ( SparseObj *  so )

inline

Minimum ordering algorithm to determine the order that the matrix should be solved.

Also make sure all needed elements are present. This does not mess up the matrix.

Definition at line 416 of file sparse_thread.hpp.

subrow()

void coreneuron::scopmath::sparse::subrow ( SparseObj *  so, Elm *  pivot, Elm *  rowsub, int  _iml  )

inline

Definition at line 440 of file sparse_thread.hpp.

thread_getelm()

template<enabled_code code_to_enable = enabled_code::all>

double* coreneuron::scopmath::sparse::thread_getelm	(	SparseObj *	so,
		int	row,
		int	col,
		int	_iml
	)

Definition at line 505 of file sparse_thread.hpp.