Inheritance diagram for Sundance::PartitionedRectangleMesher:

Public Member Functions
	PartitionedRectangleMesher (double ax, double bx, int nx, int npx, double ay, double by, int ny, int npy, const MeshType &meshType, int verbosity=0, const MPIComm &comm=MPIComm::world())
	PartitionedRectangleMesher (double ax, double bx, int nx, double ay, double by, int ny, const MeshType &meshType, int verbosity=0, const MPIComm &comm=MPIComm::world())
	PartitionedRectangleMesher (const ParameterList &params)
virtual	~PartitionedRectangleMesher ()
virtual std::string	description () const
virtual RCP< MeshSourceBase >	getRcp ()
Static Public Member Functions
static void	balanceXY (int n, int npx, int npy)
Protected Member Functions
virtual Mesh	fillMesh () const
Private Attributes
double	ax_
double	bx_
int	nx_
int	npx_
double	ay_
double	by_
int	ny_
int	npy_

Detailed Description

PartitionedRectangleMesher meshes the rectangle $\left[a_x, b_x\right] \otimes \left[a_y, b_y\right]$ with $n_x \otimes n_y$ elements per processor. The rectangle is partitioned among processors, with $np_x$ equal sized subdomains in the $x$ direction and $np_y$ in the $y$ direction.

Definition at line 63 of file SundancePartitionedRectangleMesher.hpp.

Constructor & Destructor Documentation

Sundance::PartitionedRectangleMesher::PartitionedRectangleMesher	(	double	ax,
		double	bx,
		int	nx,
		int	npx,
		double	ay,
		double	by,
		int	ny,
		int	npy,
		const MeshType &	meshType,
		int	verbosity = `0`,
		const MPIComm &	comm = `MPIComm::world()`
	)		`[inline]`

Set up meshing of the rectangle $\left[a_x, b_x\right] \otimes \left[a_y, b_y\right]$ with $n_x \otimes n_y$ elements per processor. The rectangle is partitioned among processors, with $np_x$ equal sized subdomains in the $x$ direction and $np_y$ in the $y$ direction.

Definition at line 75 of file SundancePartitionedRectangleMesher.hpp.

Sundance::PartitionedRectangleMesher::PartitionedRectangleMesher	(	double	ax,
		double	bx,
		int	nx,
		double	ay,
		double	by,
		int	ny,
		const MeshType &	meshType,
		int	verbosity = `0`,
		const MPIComm &	comm = `MPIComm::world()`
	)		`[inline]`

Set up meshing of the rectangle $\left[a_x, b_x\right] \otimes \left[a_y, b_y\right]$ with $n_x \otimes n_y$ elements per processor. The balance() function is used to choose npx and npy.

Definition at line 91 of file SundancePartitionedRectangleMesher.hpp.

References balanceXY(), Sundance::MeshSourceBase::comm(), Playa::MPIComm::getNProc(), npx_, and npy_.

PartitionedRectangleMesher::PartitionedRectangleMesher ( const ParameterList & params )

Create a rectangle mesher from a ParameterList

Definition at line 52 of file SundancePartitionedRectangleMesher.cpp.

References npx_, and npy_.

virtual Sundance::PartitionedRectangleMesher::~PartitionedRectangleMesher ( ) [inline, virtual]

Definition at line 109 of file SundancePartitionedRectangleMesher.hpp.