SAMRAI::mesh::AsyncBergerRigoutsosNode< DIM >::CommonParams Struct Reference

Parameters shared among all dendogram nodes in an dendogram and collectively managed by those nodes. More...

#include <source/mesh/clustering/AsyncBergerRigoutsosNode.h>

List of all members.

Public Member Functions
	CommonParams (const tbox::Pointer< hier::PatchLevel< DIM > > level_, const int tag_data_index_, const int tag_val_, const hier::IntVector< DIM > min_box_, const double efficiency_tol_, const double combine_tol_, const tbox::SAMRAI_MPI::comm mpi_communicator_)
	Constructor sets all const members, which must be given in argument.
Static Public Member Functions
static int	getRank (const tbox::SAMRAI_MPI::comm &mpi_communicator_)
	Utility function for getting rank from an arbitrary communicator.
static int	getProcCount (const tbox::SAMRAI_MPI::comm &mpi_communicator_)
	Utility function for getting processor count from an arbitrary communicator.
Public Attributes
tbox::JobRelauncher	job_relauncher
hier::LayerNodeSet< DIM >	tag_node_set
	Layer of graph nodes coresponding to patch boxes on the tagged level.
hier::LayerNodeSet< DIM >	new_node_set
	Layer of graph nodes (boxes) generated by BR.
Connectivity	tag_cnect_new
	Connectivity from tag_node_set to new_node_set.
Connectivity	new_cnect_tag
	Connectivity from new_node_set to tag_node_set.
IntSet	edge_senders
	List of processes that will send neighbor data for locally owned boxes after the BR algorithm completes.
std::map< int, std::vector< int > >	edge_messages
	Outgoing messages to be sent to graph node owners describing new edges found by local process.
int	compute_edges
	Edge computation flag.
hier::IntVector< DIM >	max_gcw
	Ammount to grow a box when checking for overlap.
OwnerMode	owner_mode
	How to chose the group's owner.
bool	use_level_boxes
	Whether to use globally duplicated level boxes and globally duplicate new graph nodes.
const tbox::Pointer< hier::PatchLevel< DIM > >	level
const int	tag_data_index
const int	tag_val
const hier::IntVector< DIM >	min_box
const double	efficiency_tol
const double	combine_tol
Communication parameters
tbox::SAMRAI_MPI::comm	mpi_communicator
	MPI communicator used in all communications in the dendogram.
const int	rank
const int	nproc
int	tag_upper_bound
	Upperbound of valid tags.
int	available_mpi_tag
	Smallest unclaimed MPI tag in pool given to local process.
Performance monitors
tbox::Pointer< tbox::Timer >	t_cluster
tbox::Pointer< tbox::Timer >	t_continue_algorithm
tbox::Pointer< tbox::Timer >	t_compute_new_graph_edges
tbox::Pointer< tbox::Timer >	t_share_new_edges
tbox::Pointer< tbox::Timer >	t_share_new_edges_send
tbox::Pointer< tbox::Timer >	t_share_new_edges_recv
tbox::Pointer< tbox::Timer >	t_share_new_edges_unpack
tbox::Pointer< tbox::Timer >	t_reduce_histogram
tbox::Pointer< tbox::Timer >	t_bcast_acceptability
tbox::Pointer< tbox::Timer >	t_gather_grouping_criteria
tbox::Pointer< tbox::Timer >	t_bcast_child_groups
tbox::Pointer< tbox::Timer >	t_bcast_to_dropouts
Auxiliary data for analysis and debugging.
bool	log_node_history
	Whether to log major actions of dendogram node.
int	num_nodes_allocated
	Current number of dendogram nodes allocated.
int	max_nodes_allocated
	Highest number of dendogram nodes.
int	num_nodes_active
	Current number of dendogram nodes active.
int	max_nodes_active
	Highest number of dendogram nodes active.
int	num_nodes_owned
	Current number of dendogram nodes owned.
int	max_nodes_owned
	Highest number of dendogram nodes owned.
int	num_nodes_completed
	Current number of dendogram nodes completed.
int	max_generation
	Highest number of generation.
int	num_boxes_generated
	Current number of boxes generated.
int	num_conts_to_complete
	Number of continueAlgorithm calls for to complete nodes.
int	max_conts_to_complete
	Highest number of continueAlgorithm calls to complete nodes.

---

Detailed Description

template<int DIM>
struct SAMRAI::mesh::AsyncBergerRigoutsosNode< DIM >::CommonParams

Parameters shared among all dendogram nodes in an dendogram and collectively managed by those nodes.

In the implementation of the BR algorithm, some parameters are to be shared among all nodes in the dendogram, either for efficiency or coordinating the dendogram nodes. All such parameters are contained in a single CommonParams object.

---

Constructor & Destructor Documentation

template<int DIM>

SAMRAI::mesh::AsyncBergerRigoutsosNode< DIM >::CommonParams::CommonParams	(	const tbox::Pointer< hier::PatchLevel< DIM > >	level_,
		const int	tag_data_index_,
		const int	tag_val_,
		const hier::IntVector< DIM >	min_box_,
		const double	efficiency_tol_,
		const double	combine_tol_,
		const tbox::SAMRAI_MPI::comm	mpi_communicator_
	)

Constructor sets all const members, which must be given in argument.

---

Member Function Documentation

template<int DIM>

int SAMRAI::mesh::AsyncBergerRigoutsosNode< DIM >::CommonParams::getRank

(

const tbox::SAMRAI_MPI::comm &

mpi_communicator_

)

[static]

Utility function for getting rank from an arbitrary communicator.

template<int DIM>

int SAMRAI::mesh::AsyncBergerRigoutsosNode< DIM >::CommonParams::getProcCount

(

const tbox::SAMRAI_MPI::comm &

mpi_communicator_

)

[static]

Utility function for getting processor count from an arbitrary communicator.

---

Member Data Documentation

template<int DIM>

tbox::JobRelauncher SAMRAI::mesh::AsyncBergerRigoutsosNode< DIM >::CommonParams::job_relauncher

template<int DIM>

hier::LayerNodeSet<DIM> SAMRAI::mesh::AsyncBergerRigoutsosNode< DIM >::CommonParams::tag_node_set

Layer of graph nodes coresponding to patch boxes on the tagged level.

If edges are computed (see setComputeEdges()), the edges go between the graph nodes on the tagged level and the generated graph nodes.

template<int DIM>

hier::LayerNodeSet<DIM> SAMRAI::mesh::AsyncBergerRigoutsosNode< DIM >::CommonParams::new_node_set

Layer of graph nodes (boxes) generated by BR.

This is where we store the boxes resulting from the BR algorithm.

template<int DIM>

Connectivity SAMRAI::mesh::AsyncBergerRigoutsosNode< DIM >::CommonParams::tag_cnect_new

Connectivity from tag_node_set to new_node_set.

This is where we store the edges resulting from the BR algorithm. The edges are created locally for local nodes in tag_node_set.

template<int DIM>

Connectivity SAMRAI::mesh::AsyncBergerRigoutsosNode< DIM >::CommonParams::new_cnect_tag

Connectivity from new_node_set to tag_node_set.

The edges are created when the owners of nodes in tag_node_set share edge data with owners of nodes in new_node_set.

template<int DIM>

IntSet SAMRAI::mesh::AsyncBergerRigoutsosNode< DIM >::CommonParams::edge_senders

List of processes that will send neighbor data for locally owned boxes after the BR algorithm completes.

template<int DIM>

std::map<int,std::vector<int> > SAMRAI::mesh::AsyncBergerRigoutsosNode< DIM >::CommonParams::edge_messages

Outgoing messages to be sent to graph node owners describing new edges found by local process.

template<int DIM>

const tbox::Pointer<hier::PatchLevel<DIM> > SAMRAI::mesh::AsyncBergerRigoutsosNode< DIM >::CommonParams::level

template<int DIM>

const int SAMRAI::mesh::AsyncBergerRigoutsosNode< DIM >::CommonParams::tag_data_index

template<int DIM>

const int SAMRAI::mesh::AsyncBergerRigoutsosNode< DIM >::CommonParams::tag_val

template<int DIM>

const hier::IntVector<DIM> SAMRAI::mesh::AsyncBergerRigoutsosNode< DIM >::CommonParams::min_box

template<int DIM>

const double SAMRAI::mesh::AsyncBergerRigoutsosNode< DIM >::CommonParams::efficiency_tol

template<int DIM>

const double SAMRAI::mesh::AsyncBergerRigoutsosNode< DIM >::CommonParams::combine_tol

template<int DIM>

int SAMRAI::mesh::AsyncBergerRigoutsosNode< DIM >::CommonParams::compute_edges

Edge computation flag.

See setComputeEdges().

template<int DIM>

hier::IntVector<DIM> SAMRAI::mesh::AsyncBergerRigoutsosNode< DIM >::CommonParams::max_gcw

Ammount to grow a box when checking for overlap.

template<int DIM>

OwnerMode SAMRAI::mesh::AsyncBergerRigoutsosNode< DIM >::CommonParams::owner_mode

How to chose the group's owner.

template<int DIM>

bool SAMRAI::mesh::AsyncBergerRigoutsosNode< DIM >::CommonParams::use_level_boxes

Whether to use globally duplicated level boxes and globally duplicate new graph nodes.

template<int DIM>

tbox::SAMRAI_MPI::comm SAMRAI::mesh::AsyncBergerRigoutsosNode< DIM >::CommonParams::mpi_communicator

MPI communicator used in all communications in the dendogram.

Currently set to MPI_COMM_WORLD. Could eventually be set to a duplicate of MPI_COMM_WORLD if there is interference with other communications using MPI_COMM_WORLD.

template<int DIM>

const int SAMRAI::mesh::AsyncBergerRigoutsosNode< DIM >::CommonParams::rank

template<int DIM>

const int SAMRAI::mesh::AsyncBergerRigoutsosNode< DIM >::CommonParams::nproc

template<int DIM>

int SAMRAI::mesh::AsyncBergerRigoutsosNode< DIM >::CommonParams::tag_upper_bound

Upperbound of valid tags.

template<int DIM>

int SAMRAI::mesh::AsyncBergerRigoutsosNode< DIM >::CommonParams::available_mpi_tag

Smallest unclaimed MPI tag in pool given to local process.

template<int DIM>

tbox::Pointer<tbox::Timer> SAMRAI::mesh::AsyncBergerRigoutsosNode< DIM >::CommonParams::t_cluster

template<int DIM>

tbox::Pointer<tbox::Timer> SAMRAI::mesh::AsyncBergerRigoutsosNode< DIM >::CommonParams::t_continue_algorithm

template<int DIM>

tbox::Pointer<tbox::Timer> SAMRAI::mesh::AsyncBergerRigoutsosNode< DIM >::CommonParams::t_compute_new_graph_edges

template<int DIM>

tbox::Pointer<tbox::Timer> SAMRAI::mesh::AsyncBergerRigoutsosNode< DIM >::CommonParams::t_share_new_edges

template<int DIM>

tbox::Pointer<tbox::Timer> SAMRAI::mesh::AsyncBergerRigoutsosNode< DIM >::CommonParams::t_share_new_edges_send

template<int DIM>

tbox::Pointer<tbox::Timer> SAMRAI::mesh::AsyncBergerRigoutsosNode< DIM >::CommonParams::t_share_new_edges_recv

template<int DIM>

tbox::Pointer<tbox::Timer> SAMRAI::mesh::AsyncBergerRigoutsosNode< DIM >::CommonParams::t_share_new_edges_unpack

template<int DIM>

tbox::Pointer<tbox::Timer> SAMRAI::mesh::AsyncBergerRigoutsosNode< DIM >::CommonParams::t_reduce_histogram

template<int DIM>

tbox::Pointer<tbox::Timer> SAMRAI::mesh::AsyncBergerRigoutsosNode< DIM >::CommonParams::t_bcast_acceptability

template<int DIM>

tbox::Pointer<tbox::Timer> SAMRAI::mesh::AsyncBergerRigoutsosNode< DIM >::CommonParams::t_gather_grouping_criteria

template<int DIM>

tbox::Pointer<tbox::Timer> SAMRAI::mesh::AsyncBergerRigoutsosNode< DIM >::CommonParams::t_bcast_child_groups

template<int DIM>

tbox::Pointer<tbox::Timer> SAMRAI::mesh::AsyncBergerRigoutsosNode< DIM >::CommonParams::t_bcast_to_dropouts

template<int DIM>

bool SAMRAI::mesh::AsyncBergerRigoutsosNode< DIM >::CommonParams::log_node_history

Whether to log major actions of dendogram node.

template<int DIM>

int SAMRAI::mesh::AsyncBergerRigoutsosNode< DIM >::CommonParams::num_nodes_allocated

Current number of dendogram nodes allocated.

template<int DIM>

int SAMRAI::mesh::AsyncBergerRigoutsosNode< DIM >::CommonParams::max_nodes_allocated

Highest number of dendogram nodes.

template<int DIM>

int SAMRAI::mesh::AsyncBergerRigoutsosNode< DIM >::CommonParams::num_nodes_active

Current number of dendogram nodes active.

template<int DIM>

int SAMRAI::mesh::AsyncBergerRigoutsosNode< DIM >::CommonParams::max_nodes_active

Highest number of dendogram nodes active.

template<int DIM>

int SAMRAI::mesh::AsyncBergerRigoutsosNode< DIM >::CommonParams::num_nodes_owned

Current number of dendogram nodes owned.

template<int DIM>

int SAMRAI::mesh::AsyncBergerRigoutsosNode< DIM >::CommonParams::max_nodes_owned

Highest number of dendogram nodes owned.

template<int DIM>

int SAMRAI::mesh::AsyncBergerRigoutsosNode< DIM >::CommonParams::num_nodes_completed

Current number of dendogram nodes completed.

template<int DIM>

int SAMRAI::mesh::AsyncBergerRigoutsosNode< DIM >::CommonParams::max_generation

Highest number of generation.

template<int DIM>

int SAMRAI::mesh::AsyncBergerRigoutsosNode< DIM >::CommonParams::num_boxes_generated

Current number of boxes generated.

template<int DIM>

int SAMRAI::mesh::AsyncBergerRigoutsosNode< DIM >::CommonParams::num_conts_to_complete

Number of continueAlgorithm calls for to complete nodes.

template<int DIM>

int SAMRAI::mesh::AsyncBergerRigoutsosNode< DIM >::CommonParams::max_conts_to_complete

Highest number of continueAlgorithm calls to complete nodes.

---

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

• source/mesh/clustering/AsyncBergerRigoutsosNode.h
• source/mesh/clustering/AsyncBergerRigoutsosNode.C

---