libMesh
Public Member Functions | Static Public Member Functions | Protected Member Functions | Protected Attributes | Static Protected Attributes | List of all members
libMesh::Partitioner Class Referenceabstract

The Partitioner class provides a uniform interface for partitioning algorithms. More...

#include <partitioner.h>

Inheritance diagram for libMesh::Partitioner:
[legend]

Public Member Functions

 Partitioner ()
 Constructor. More...
 
 Partitioner (const Partitioner &)=default
 Copy/move ctor, copy/move assignment operator, and destructor are all explicitly defaulted for this class. More...
 
 Partitioner (Partitioner &&)=default
 
Partitioneroperator= (const Partitioner &)=default
 
Partitioneroperator= (Partitioner &&)=default
 
virtual ~Partitioner ()=default
 
virtual std::unique_ptr< Partitionerclone () const =0
 
virtual void partition (MeshBase &mesh, const unsigned int n)
 Partitions the MeshBase into n parts by setting processor_id() on Nodes and Elems. More...
 
virtual void partition (MeshBase &mesh)
 Partitions the MeshBase into mesh.n_processors() by setting processor_id() on Nodes and Elems. More...
 
virtual void partition_range (MeshBase &, MeshBase::element_iterator, MeshBase::element_iterator, const unsigned int)
 Partitions elements in the range (it, end) into n parts. More...
 
void repartition (MeshBase &mesh, const unsigned int n)
 Repartitions the MeshBase into n parts. More...
 
void repartition (MeshBase &mesh)
 Repartitions the MeshBase into mesh.n_processors() parts. More...
 
virtual void attach_weights (ErrorVector *)
 Attach weights that can be used for partitioning. More...
 

Static Public Member Functions

static void partition_unpartitioned_elements (MeshBase &mesh)
 These functions assign processor IDs to newly-created elements (in parallel) which are currently assigned to processor 0. More...
 
static void partition_unpartitioned_elements (MeshBase &mesh, const unsigned int n)
 
static void set_parent_processor_ids (MeshBase &mesh)
 This function is called after partitioning to set the processor IDs for the inactive parent elements. More...
 
static void set_node_processor_ids (MeshBase &mesh)
 This function is called after partitioning to set the processor IDs for the nodes. More...
 
static void processor_pairs_to_interface_nodes (MeshBase &mesh, std::map< std::pair< processor_id_type, processor_id_type >, std::set< dof_id_type >> &processor_pair_to_nodes)
 On the partitioning interface, a surface is shared by two and only two processors. More...
 
static void set_interface_node_processor_ids_linear (MeshBase &mesh)
 Nodes on the partitioning interface is linearly assigned to each pair of processors. More...
 
static void set_interface_node_processor_ids_BFS (MeshBase &mesh)
 Nodes on the partitioning interface is clustered into two groups BFS (Breadth First Search)scheme for per pair of processors. More...
 
static void set_interface_node_processor_ids_petscpartitioner (MeshBase &mesh)
 Nodes on the partitioning interface is partitioned into two groups using a PETSc partitioner for each pair of processors. More...
 

Protected Member Functions

void single_partition (MeshBase &mesh)
 Trivially "partitions" the mesh for one processor. More...
 
void single_partition_range (MeshBase::element_iterator it, MeshBase::element_iterator end)
 Slightly generalized version of single_partition which acts on a range of elements defined by the pair of iterators (it, end). More...
 
virtual void _do_partition (MeshBase &mesh, const unsigned int n)=0
 This is the actual partitioning method which must be overridden in derived classes. More...
 
virtual void _do_repartition (MeshBase &mesh, const unsigned int n)
 This is the actual re-partitioning method which can be overridden in derived classes. More...
 
virtual void _find_global_index_by_pid_map (const MeshBase &mesh)
 Construct contiguous global indices for the current partitioning. More...
 
virtual void build_graph (const MeshBase &mesh)
 Build a dual graph for partitioner. More...
 
void assign_partitioning (const MeshBase &mesh, const std::vector< dof_id_type > &parts)
 Assign the computed partitioning to the mesh. More...
 

Protected Attributes

ErrorVector_weights
 The weights that might be used for partitioning. More...
 
std::unordered_map< dof_id_type, dof_id_type_global_index_by_pid_map
 Maps active element ids into a contiguous range, as needed by parallel partitioner. More...
 
std::vector< dof_id_type_n_active_elem_on_proc
 The number of active elements on each processor. More...
 
std::vector< std::vector< dof_id_type > > _dual_graph
 A dual graph corresponds to the mesh, and it is typically used in paritioner. More...
 
std::vector< Elem * > _local_id_to_elem
 

Static Protected Attributes

static const dof_id_type communication_blocksize
 The blocksize to use when doing blocked parallel communication. More...
 

Detailed Description

The Partitioner class provides a uniform interface for partitioning algorithms.

It takes a reference to a MeshBase object as input, which it will partition into a number of subdomains.

Author
Benjamin S. Kirk
Date
2003 Base class for all concrete Partitioner instantiations.

Definition at line 50 of file partitioner.h.

Constructor & Destructor Documentation

◆ Partitioner() [1/3]

libMesh::Partitioner::Partitioner ( )

Constructor.

Definition at line 57 of file partitioner.h.

57 : _weights(nullptr) {}
ErrorVector * _weights
The weights that might be used for partitioning.
Definition: partitioner.h:267

◆ Partitioner() [2/3]

libMesh::Partitioner::Partitioner ( const Partitioner )
default

Copy/move ctor, copy/move assignment operator, and destructor are all explicitly defaulted for this class.

◆ Partitioner() [3/3]

libMesh::Partitioner::Partitioner ( Partitioner &&  )
default

◆ ~Partitioner()

virtual libMesh::Partitioner::~Partitioner ( )
virtualdefault

Member Function Documentation

◆ _do_partition()

virtual void libMesh::Partitioner::_do_partition ( MeshBase mesh,
const unsigned int  n 
)
protectedpure virtual

This is the actual partitioning method which must be overridden in derived classes.

It is called via the public partition() method above by the user.

Implemented in libMesh::SubdomainPartitioner, libMesh::CentroidPartitioner, libMesh::ParmetisPartitioner, libMesh::SFCPartitioner, libMesh::MappedSubdomainPartitioner, libMesh::MetisPartitioner, libMesh::LinearPartitioner, libMesh::HilbertSFCPartitioner, and libMesh::MortonSFCPartitioner.

Referenced by _do_repartition().

◆ _do_repartition()

virtual void libMesh::Partitioner::_do_repartition ( MeshBase mesh,
const unsigned int  n 
)
protectedvirtual

This is the actual re-partitioning method which can be overridden in derived classes.

Note
The default behavior is to simply call the partition function.

Reimplemented in libMesh::ParmetisPartitioner.

Definition at line 237 of file partitioner.h.

References _do_partition().

238  { this->_do_partition (mesh, n); }
virtual void _do_partition(MeshBase &mesh, const unsigned int n)=0
This is the actual partitioning method which must be overridden in derived classes.

◆ _find_global_index_by_pid_map()

virtual void libMesh::Partitioner::_find_global_index_by_pid_map ( const MeshBase mesh)
protectedvirtual

Construct contiguous global indices for the current partitioning.

The global indices are ordered part-by-part

◆ assign_partitioning()

void libMesh::Partitioner::assign_partitioning ( const MeshBase mesh,
const std::vector< dof_id_type > &  parts 
)
protected

Assign the computed partitioning to the mesh.

◆ attach_weights()

virtual void libMesh::Partitioner::attach_weights ( ErrorVector )
virtual

Attach weights that can be used for partitioning.

This ErrorVector should be exactly the same on every processor and should have mesh->max_elem_id() entries.

Reimplemented in libMesh::MetisPartitioner.

Definition at line 203 of file partitioner.h.

203 { libmesh_not_implemented(); }

◆ build_graph()

virtual void libMesh::Partitioner::build_graph ( const MeshBase mesh)
protectedvirtual

Build a dual graph for partitioner.

Reimplemented in libMesh::ParmetisPartitioner.

◆ clone()

virtual std::unique_ptr<Partitioner> libMesh::Partitioner::clone ( ) const
pure virtual
Returns
A copy of this partitioner wrapped in a smart pointer.

This is used when copying meshes, and must be overridden in the derived classes.

Implemented in libMesh::SubdomainPartitioner, libMesh::CentroidPartitioner, libMesh::ParmetisPartitioner, libMesh::SFCPartitioner, libMesh::HilbertSFCPartitioner, libMesh::MortonSFCPartitioner, libMesh::MappedSubdomainPartitioner, libMesh::LinearPartitioner, and libMesh::MetisPartitioner.

◆ operator=() [1/2]

Partitioner& libMesh::Partitioner::operator= ( const Partitioner )
default

◆ operator=() [2/2]

Partitioner& libMesh::Partitioner::operator= ( Partitioner &&  )
default

◆ partition() [1/2]

virtual void libMesh::Partitioner::partition ( MeshBase mesh,
const unsigned int  n 
)
virtual

Partitions the MeshBase into n parts by setting processor_id() on Nodes and Elems.

Note
If you are implementing a new type of Partitioner, you most likely do not want to override the partition() function, see instead the protected virtual _do_partition() method below. The partition() function is responsible for doing a lot of libmesh-internals-specific setup and finalization before and after the _do_partition() function is called. The only responsibility of the _do_partition() function, on the other hand, is to set the processor IDs of the elements according to a specific partitioning algorithm. See, e.g. MetisPartitioner for an example.

◆ partition() [2/2]

virtual void libMesh::Partitioner::partition ( MeshBase mesh)
virtual

Partitions the MeshBase into mesh.n_processors() by setting processor_id() on Nodes and Elems.

Note
If you are implementing a new type of Partitioner, you most likely do not want to override the partition() function, see instead the protected virtual _do_partition() method below. The partition() function is responsible for doing a lot of libmesh-internals-specific setup and finalization before and after the _do_partition() function is called. The only responsibility of the _do_partition() function, on the other hand, is to set the processor IDs of the elements according to a specific partitioning algorithm. See, e.g. MetisPartitioner for an example.

◆ partition_range()

virtual void libMesh::Partitioner::partition_range ( MeshBase ,
MeshBase::element_iterator  ,
MeshBase::element_iterator  ,
const unsigned int   
)
virtual

Partitions elements in the range (it, end) into n parts.

The mesh from which the iterators are created must also be passed in, since it is a parallel object and has other useful information in it.

Although partition_range() is part of the public Partitioner interface, it should not generally be called by applications. Its main purpose is to support the SubdomainPartitioner, which uses it internally to individually partition ranges of elements before combining them into the final partitioning. Most of the time, the protected _do_partition() function is implemented in terms of partition_range() by passing a range which includes all the elements of the Mesh.

Reimplemented in libMesh::CentroidPartitioner, libMesh::SFCPartitioner, libMesh::MappedSubdomainPartitioner, libMesh::LinearPartitioner, and libMesh::MetisPartitioner.

Definition at line 127 of file partitioner.h.

131  { libmesh_not_implemented(); }

◆ partition_unpartitioned_elements() [1/2]

static void libMesh::Partitioner::partition_unpartitioned_elements ( MeshBase mesh)
static

These functions assign processor IDs to newly-created elements (in parallel) which are currently assigned to processor 0.

◆ partition_unpartitioned_elements() [2/2]

static void libMesh::Partitioner::partition_unpartitioned_elements ( MeshBase mesh,
const unsigned int  n 
)
static

◆ processor_pairs_to_interface_nodes()

static void libMesh::Partitioner::processor_pairs_to_interface_nodes ( MeshBase mesh,
std::map< std::pair< processor_id_type, processor_id_type >, std::set< dof_id_type >> &  processor_pair_to_nodes 
)
static

On the partitioning interface, a surface is shared by two and only two processors.

Try to find which pair of processors corresponds to which surfaces, and store their nodes.

◆ repartition() [1/2]

void libMesh::Partitioner::repartition ( MeshBase mesh,
const unsigned int  n 
)

Repartitions the MeshBase into n parts.

(Some partitioning algorithms can repartition more efficiently than computing a new partitioning from scratch.) The default behavior is to simply call this->partition(mesh,n).

◆ repartition() [2/2]

void libMesh::Partitioner::repartition ( MeshBase mesh)

Repartitions the MeshBase into mesh.n_processors() parts.

This is required since some partitioning algorithms can repartition more efficiently than computing a new partitioning from scratch.

◆ set_interface_node_processor_ids_BFS()

static void libMesh::Partitioner::set_interface_node_processor_ids_BFS ( MeshBase mesh)
static

Nodes on the partitioning interface is clustered into two groups BFS (Breadth First Search)scheme for per pair of processors.

◆ set_interface_node_processor_ids_linear()

static void libMesh::Partitioner::set_interface_node_processor_ids_linear ( MeshBase mesh)
static

Nodes on the partitioning interface is linearly assigned to each pair of processors.

◆ set_interface_node_processor_ids_petscpartitioner()

static void libMesh::Partitioner::set_interface_node_processor_ids_petscpartitioner ( MeshBase mesh)
static

Nodes on the partitioning interface is partitioned into two groups using a PETSc partitioner for each pair of processors.

◆ set_node_processor_ids()

static void libMesh::Partitioner::set_node_processor_ids ( MeshBase mesh)
static

This function is called after partitioning to set the processor IDs for the nodes.

By definition, a Node's processor ID is the minimum processor ID for all of the elements which share the node.

◆ set_parent_processor_ids()

static void libMesh::Partitioner::set_parent_processor_ids ( MeshBase mesh)
static

This function is called after partitioning to set the processor IDs for the inactive parent elements.

A parent's processor ID is the same as its first child.

◆ single_partition()

void libMesh::Partitioner::single_partition ( MeshBase mesh)
protected

Trivially "partitions" the mesh for one processor.

Simply loops through the elements and assigns all of them to processor 0. Is is provided as a separate function so that derived classes may use it without reimplementing it.

◆ single_partition_range()

void libMesh::Partitioner::single_partition_range ( MeshBase::element_iterator  it,
MeshBase::element_iterator  end 
)
protected

Slightly generalized version of single_partition which acts on a range of elements defined by the pair of iterators (it, end).

Member Data Documentation

◆ _dual_graph

std::vector<std::vector<dof_id_type> > libMesh::Partitioner::_dual_graph
protected

A dual graph corresponds to the mesh, and it is typically used in paritioner.

A vertex represents an element, and its neighbors are the element neighbors.

Definition at line 288 of file partitioner.h.

◆ _global_index_by_pid_map

std::unordered_map<dof_id_type, dof_id_type> libMesh::Partitioner::_global_index_by_pid_map
protected

Maps active element ids into a contiguous range, as needed by parallel partitioner.

Definition at line 272 of file partitioner.h.

◆ _local_id_to_elem

std::vector<Elem *> libMesh::Partitioner::_local_id_to_elem
protected

Definition at line 291 of file partitioner.h.

◆ _n_active_elem_on_proc

std::vector<dof_id_type> libMesh::Partitioner::_n_active_elem_on_proc
protected

The number of active elements on each processor.

Note
ParMETIS requires that each processor have some active elements; it will abort if any processor passes a nullptr _part array.

Definition at line 281 of file partitioner.h.

◆ _weights

ErrorVector* libMesh::Partitioner::_weights
protected

The weights that might be used for partitioning.

Definition at line 267 of file partitioner.h.

Referenced by libMesh::MetisPartitioner::attach_weights().

◆ communication_blocksize

const dof_id_type libMesh::Partitioner::communication_blocksize
staticprotected

The blocksize to use when doing blocked parallel communication.

This limits the maximum vector size which can be used in a single communication step.

Definition at line 244 of file partitioner.h.


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