dolfin::ParallelRefinement Class Reference

Data structure and methods for refining meshes in parallel. More...

#include <ParallelRefinement.h>

Public Member Functions
	ParallelRefinement (const Mesh &mesh)
	Constructor.
	~ParallelRefinement ()
	Destructor.
const Mesh &	mesh () const
	Original mesh associated with this refinement.
bool	is_marked (std::size_t edge_index) const
void	mark (std::size_t edge_index)
void	mark_all ()
	Mark all edges in mesh.
void	mark (const MeshFunction< bool > &refinement_marker)
void	mark (const MeshEntity &cell)
std::vector< std::size_t >	marked_edge_list (const MeshEntity &cell) const
void	update_logical_edgefunction ()
	Transfer marked edges between processes.
void	create_new_vertices ()
std::shared_ptr< const std::map< std::size_t, std::size_t > >	edge_to_new_vertex () const
void	new_cell (const Cell &cell)
void	new_cell (std::size_t i0, std::size_t i1, std::size_t i2, std::size_t i3)
void	new_cell (std::size_t i0, std::size_t i1, std::size_t i2)
void	new_cells (const std::vector< std::size_t > &idx)
void	partition (Mesh &new_mesh, bool redistribute) const
void	build_local (Mesh &new_mesh) const

Detailed Description

Data structure and methods for refining meshes in parallel.

ParallelRefinement encapsulates two main features: a distributed MeshFunction defined over the mesh edes, which can be updated across processes, and storage for local mesh data, which can be used to construct the new Mesh

Member Function Documentation

build_local()

void ParallelRefinement::build_local

(

Mesh &

new_mesh

)

const

Build local mesh from internal data when not running in parallel

Parameters

new_mesh

(Mesh)

create_new_vertices()

void ParallelRefinement::create_new_vertices

(

)

Add new vertex for each marked edge, and create new_vertex_coordinates and global_edge->new_vertex mapping. Communicate new vertices with MPI to all affected processes.

edge_to_new_vertex()

std::shared_ptr< const std::map< std::size_t, std::size_t > > ParallelRefinement::edge_to_new_vertex

(

)

const

Mapping of old edge (to be removed) to new global vertex number. Useful for forming new topology

is_marked()

bool ParallelRefinement::is_marked

(

std::size_t

edge_index

)

const

Return marked status of edge

Parameters

edge_index

(std::size_t)

mark() [1/3]

void ParallelRefinement::mark

(

std::size_t

edge_index

)

Mark edge by index

Parameters

edge_index

(std::size_t) Index of edge to mark

mark() [2/3]

void ParallelRefinement::mark

(

const MeshFunction< bool > &

refinement_marker

)

Mark all edges incident on entities indicated by refinement marker

Parameters

refinement_marker

(const MeshFunction<bool>)

mark() [3/3]

void ParallelRefinement::mark

(

const MeshEntity &

cell

)

Mark all incident edges of an entity

Parameters

cell	(MeshEntity)

marked_edge_list()

std::vector< std::size_t > ParallelRefinement::marked_edge_list

(

const MeshEntity &

cell

)

const

Return list of marked edges incident on this MeshEntity - usually a cell

Parameters

cell	(const MeshEntity)

new_cell() [1/3]

void ParallelRefinement::new_cell

(

const Cell &

cell

)

Add a new cell to the list in 3D or 2D

Parameters

cell	(const Cell)

new_cell() [2/3]

void ParallelRefinement::new_cell	(	std::size_t	i0,
		std::size_t	i1,
		std::size_t	i2,
		std::size_t	i3
	)

Add a new cell with vertex indices

Parameters

i0	(std::size_t)
i1	(std::size_t)
i2	(std::size_t)
i3	(std::size_t)

new_cell() [3/3]

void ParallelRefinement::new_cell	(	std::size_t	i0,
		std::size_t	i1,
		std::size_t	i2
	)

Add a new cell with vertex indices

Parameters

i0	(std::size_t)
i1	(std::size_t)
i2	(std::size_t)

new_cells()

void ParallelRefinement::new_cells

(

const std::vector< std::size_t > &

idx

)

Add new cells with vertex indices

Parameters

idx	(const std::vector<std::size_t>)

partition()

void ParallelRefinement::partition	(	Mesh &	new_mesh,
		bool	redistribute
	)		const

Use vertex and topology data to partition new mesh across processes

Parameters

new_mesh	(Mesh)
redistribute	(bool)

---

The documentation for this class was generated from the following files:

• /home/fenics/shared/dolfin/dolfin/refinement/ParallelRefinement.h
• /home/fenics/shared/dolfin/dolfin/refinement/ParallelRefinement.cpp