Zoltan2 Namespace Reference

Created by mbenlioglu on Aug 31, 2020. More...

Namespaces
	Details
	Impl

Classes
class	AlgTwoGhostLayer
class	AlgDistance1TwoGhostLayer
class	AlgDistance1
class	AlgDistance2
class	AlgPartialDistance2
class	AlgSerialGreedy
class	AlgBlockMapping
class	AlgDefaultMapping
struct	GMWM_triplet
class	AlgAMD
class	AlgMetis
class	AlgNatural
class	AlgND
class	AlgRandom
class	AlgRCM
class	AlgSortedDegree
class	Matcher
	An implementation of the Matcher interface that operates on Epetra matrices and Graphs. More...
class	SortPairs
class	AlgBlock
class	uMultiSortItem
	Class for sorting items with multiple values. First sorting with respect to val[0], then val[1] then ... val[count-1]. The last tie breaking is done with index values. Used for task mapping partitioning where the points on a cut line needs to be distributed consistently. More...
struct	uSortItem
	Sort items for quick sort function. More...
struct	uSignedSortItem
struct	Zoltan2_AlgMJ_TrackCallsCounter
class	AlgMJ
	Multi Jagged coordinate partitioning algorithm. More...
struct	Zoltan2_MJArrayType
struct	ArrayCombinationReducer
struct	ReduceWeightsFunctor
struct	ArrayReducer
struct	ReduceArrayFunctor
class	Zoltan2_AlgMJ
	Multi Jagged coordinate partitioning algorithm. More...
class	AlgParMA
class	AlgParMETIS
class	AlgPuLP
class	AlgQuotient
class	AlgSarma
class	AlgPTScotch
class	coordinateModelPartBox
	coordinateModelPartBox Class, represents the boundaries of the box which is a result of a geometric partitioning algorithm. More...
class	GridHash
	GridHash Class, Hashing Class for part boxes. More...
class	AlgMatrix
class	GNO_LNO_PAIR
class	KmeansHeap
	KmeansHeap Class, max heap, but holds the minimum values. More...
class	KMeansCluster
class	KMeansAlgorithm
	KMeansAlgorithm Class that performs clustering of the coordinates, and returns the closest set of coordinates. More...
class	CommunicationModel
	CommunicationModel Base Class that performs mapping between the coordinate partitioning result. More...
class	CoordinateCommunicationModel
	CoordinateModelInput Class that performs mapping between the coordinate partitioning result and mpi ranks base on the coordinate results and mpi physical coordinates. More...
class	CoordinateTaskMapper
class	AlgZoltan
class	AlgForTestingOnly
class	Algorithm
	Algorithm defines the base class for all algorithms. More...
struct	TPL_Traits
struct	TPL_Traits< first_t, first_t >
struct	TPL_Traits< ZOLTAN_ID_PTR, second_t >
struct	TPL_Traits< first_t, ZOLTAN_ID_PTR >
class	Zoltan2_Directory_Node
class	Zoltan2_DD_Update_Msg
class	Zoltan2_DD_Find_Msg
class	Zoltan2_DD_Remove_Msg
class	Zoltan2_Directory
	Zoltan2_Directory is an abstract base class. More...
class	Zoltan2_Directory_Simple
class	Zoltan2_Directory_Vector
class	Zoltan2_Directory_Plan
class	Zoltan2_Directory_Comm
class	DebugManager
	DebugManager contains the methods that perform output of debug and status messages. More...
class	Environment
	The user parameters, debug, timing and memory profiling output objects, and error checking methods. More...
class	NotImplemented
	Exception thrown when a called base-class method is not implemented. More...
class	IntegerRangeListValidator
	A ParameterList validator for integer range lists. More...
class	IntegerRangeListValidatorXMLConverter
	XML conversion code for IntegerRangeListValidator. More...
class	Machine
	MachineClass Base class for representing machine coordinates, networks, etc. More...
class	MachineDragonflyRCA
	A Dragonfly (e.g. Cori, Trinity, & Theta) Machine Class for task mapping. More...
class	MachineDragonflyRCAForTesting
	A Dragonfly (e.g. Cori, Trinity, Theta) Machine Class for testing only. A more realistic machine should be used for task mapping (i.e. see Zoltan2_MachineDragonflyRCA.hpp). More...
class	MachineForTesting
	A Machine Class for testing only A more realistic machine should be used for task mapping. More...
class	MachineRepresentation
	MachineRepresentation Class Base class for representing machine coordinates, networks, etc. More...
class	DefaultMachine
	A Default MachineRepresentation Class. More...
class	MachineTorusRCA
	An RCA Machine class on Torus Networks. More...
class	MachineTorusRCAForTesting
	An RCA Machine Class (Torus Networks) for testing only A more realistic machine should be used for task mapping. More...
class	MachineTorusTopoMgr
	A TopoManager Machine Class on Torus Networks. More...
class	MachineTorusBGQTest
	A TopoManager Machine Class (Torus Networks) for testing only A more realistic machine should be used for task mapping. More...
class	MetricOutputManager
	MetricOutputManager handles output of profiling messages. More...
class	TimerManager
class	BaseAdapterRoot
	BaseAdapter defines methods required by all Adapters. More...
class	BaseAdapter
class	AdapterWithCoords
class	VectorAdapter
	VectorAdapter defines the interface for vector input. More...
class	AdapterWithCoordsWrapper
class	APFMeshAdapter
class	BasicIdentifierAdapter
	This class represents a collection of global Identifiers and their associated weights, if any. More...
class	BasicKokkosIdentifierAdapter
	This class represents a collection of global Identifiers and their associated weights, if any. More...
class	BasicVectorAdapter
	BasicVectorAdapter represents a vector (plus optional weights) supplied by the user as pointers to strided arrays. More...
class	GraphAdapter
	GraphAdapter defines the interface for graph-based user data. More...
class	IdentifierAdapter
	IdentifierAdapter defines the interface for identifiers. More...
class	BasicUserTypes
	A simple class that can be the User template argument for an InputAdapter. More...
struct	InputTraits
	The traits required of User input classes or structures. More...
class	MatrixAdapter
	MatrixAdapter defines the adapter interface for matrices. More...
class	MeshAdapter
	MeshAdapter defines the interface for mesh input. More...
class	PamgenMeshAdapter
	This class represents a mesh. More...
class	TpetraCrsGraphAdapter
	Provides access for Zoltan2 to Tpetra::CrsGraph data. More...
class	TpetraCrsMatrixAdapter
	Provides access for Zoltan2 to Tpetra::CrsMatrix data. More...
class	TpetraRowGraphAdapter
	Provides access for Zoltan2 to Tpetra::RowGraph data. More...
class	TpetraRowMatrixAdapter
	Provides access for Zoltan2 to Tpetra::RowMatrix data. More...
class	XpetraCrsGraphAdapter
	Provides access for Zoltan2 to Xpetra::CrsGraph data. More...
class	XpetraCrsMatrixAdapter
	Provides access for Zoltan2 to Xpetra::CrsMatrix data. More...
class	XpetraMultiVectorAdapter
	An adapter for Xpetra::MultiVector. More...
struct	XpetraTraits
	Defines the traits required for Tpetra, Eptra and Xpetra objects. More...
class	CommGraphModel
	CommGraphModel defines the interface required for communication graph. More...
class	CoordinateModel
	This class provides geometric coordinates with optional weights to the Zoltan2 algorithm. More...
class	GraphModel
	GraphModel defines the interface required for graph models. More...
class	HyperGraphModel
	HyperGraphModel defines the interface required for hyper graph models. More...
class	IdentifierModel
	IdentifierModel defines the interface for all identifier models. More...
class	Model
	The base class for all model classes. More...
class	ColoringProblem
	ColoringProblem sets up coloring problems for the user. More...
class	ColoringSolution
	The class containing coloring solution. More...
class	MappingProblem
	MappingProblem enables mapping of a partition (either computed or input) to MPI ranks. More...
class	MappingSolution
	PartitionMapping maps a solution or an input distribution to ranks. More...
class	MatrixPartitioningProblem
	MatrixPartitioningProblem sets up partitioning problems for the user. More...
class	MatrixPartitioningSolution
	A PartitioningSolution is a solution to a partitioning problem. More...
class	OrderingProblem
	OrderingProblem sets up ordering problems for the user. More...
class	OrderingSolution
	The class containing ordering solutions. More...
class	LocalOrderingSolution
class	GlobalOrderingSolution
class	PartitioningProblem
	PartitioningProblem sets up partitioning problems for the user. More...
class	PartitioningSolution
	A PartitioningSolution is a solution to a partitioning problem. More...
class	PartitionMapping
	PartitionMapping maps a solution or an input distribution to ranks. More...
class	ProblemRoot
	ProblemRoot allows ptr storage and safe dynamic_cast of all. More...
class	Problem
	Problem base class from which other classes (PartitioningProblem, ColoringProblem, OrderingProblem, MatchingProblem, etc.) derive. More...
class	Solution
	Just a placeholder for now. More...
class	TpetraCrsColorer
class	TpetraCrsColorer< Tpetra::BlockCrsMatrix< SC, LO, GO, NO > >
class	ZoltanCrsColorer
class	Zoltan2CrsColorer
class	Zoltan2CrsColorer< Tpetra::BlockCrsMatrix< SC, LO, GO, NO > >
class	BaseClassMetrics
class	perProcessorComponentMetrics
class	EvaluateBaseClassRoot
	A base class for EvaluatePartition, EvaluateOrdering, ... More...
class	EvaluateBaseClass
class	EvaluateMapping
	A class that computes and returns quality metrics. More...
class	EvaluateOrdering
	A class that computes and returns quality metrics. base class for the local and global ordering versions. More...
class	EvaluateLocalOrdering
class	EvaluateGlobalOrdering
class	EvaluatePartition
	A class that computes and returns quality metrics. More...
class	GraphMetrics
class	ImbalanceMetrics
class	StridedData
	The StridedData class manages lists of weights or coordinates. More...
class	Sphynx
class	SphynxProblem
struct	type_name
struct	gid_set_t
struct	lid_set_t
class	IDs
class	Single_GID
class	Multiple_GID
class	Single_User
class	Vector_User
class	Single_User_Single_GID
class	Single_User_Multiple_GID
class	Vector_User_Single_GID
class	Vector_User_Multiple_GID
class	TestManager

Typedefs
typedef ::Tpetra::Details::DefaultTypes::local_ordinal_type	default_lno_t
typedef ::Tpetra::Details::DefaultTypes::global_ordinal_type	default_gno_t
typedef size_t	default_offset_t
typedef ::Tpetra::Details::DefaultTypes::scalar_type	default_scalar_t
typedef ::Tpetra::Details::DefaultTypes::node_type	default_node_t
typedef int	default_part_t
typedef std::bitset < NUM_MODEL_FLAGS >	modelFlag_t
typedef Tpetra::global_size_t	global_size_t

Enumerations
enum	blockParams {   block_balanceCount, block_balanceWeight, block_minTotalWeight, block_minMaximumWeight,   block_balanceTotalMaximum, NUM_BLOCK_PARAMS }
	The boolean parameters of interest to the Block algorithm. More...
enum	RangeType { RANGE_INCLUDES_ALL, RANGE_IS_EMPTY, RANGE_IS_LISTED, NUM_RANGE_TYPES }
	Codes that indicate how to interpret the Array<int> representing the user's integer range list. More...
enum	AssertionLevel {   NO_ASSERTIONS, BASIC_ASSERTION, COMPLEX_ASSERTION, DEBUG_MODE_ASSERTION,   NUM_ASSERTION_LEVELS }
	Level of error checking or assertions desired. More...
enum	MessageOutputLevel {   NO_STATUS, BASIC_STATUS, DETAILED_STATUS, VERBOSE_DETAILED_STATUS,   NUM_STATUS_OUTPUT_LEVELS }
	The amount of debugging or status output to print. More...
enum	TimerType {   NO_TIMERS, MACRO_TIMERS, MICRO_TIMERS, BOTH_TIMERS,   TEST_TIMERS, NUM_TIMING_OPTIONS }
	The type of timers which should be active. More...
enum	OSType { COUT_STREAM, CERR_STREAM, NULL_STREAM, NUM_OUTPUT_STREAMS }
	Output stream types. More...
enum	multiCriteriaNorm { normMinimizeTotalWeight, normBalanceTotalMaximum, normMinimizeMaximumWeight, normNumNorms }
	Enumerator used in code for multicriteria norm choice. More...
enum	BaseAdapterType {   InvalidAdapterType = 0, IdentifierAdapterType, VectorAdapterType, MatrixAdapterType,   GraphAdapterType, MeshAdapterType }
	An enum to identify general types of adapters. More...
enum	GraphEntityType { GRAPH_VERTEX, GRAPH_EDGE }
	Enumerated entity type for graphs: Vertices or Edges. More...
enum	MatrixEntityType { MATRIX_ROW, MATRIX_COLUMN, MATRIX_NONZERO }
enum	MeshEntityType { MESH_VERTEX, MESH_EDGE, MESH_FACE, MESH_REGION }
	Enumerate entity types for meshes: Regions, Faces, Edges, or Vertices. More...
enum	EntityTopologyType {   POINT, LINE_SEGMENT, POLYGON, TRIANGLE,   QUADRILATERAL, POLYHEDRON, TETRAHEDRON, HEXAHEDRON,   PRISM, PYRAMID }
	Enumerate entity topology types for meshes: points,lines,polygons,triangles,quadrilaterals, polyhedrons, tetrahedrons, hexhedrons, prisms, or pyramids. More...
enum	CentricView { HYPEREDGE_CENTRIC, VERTEX_CENTRIC }
	Enumerate the views for the pins: HYPEREDGE_CENTRIC: pins are the global ids of the vertices as seen by the hyperedges VERTEX_CENTRIC: pins are the global ids of the hyperedges as seen by the vertices. More...
enum	ModelType {   HypergraphModelType, GraphModelType, CoordinateModelType, IdentifierModelType,   MAX_NUM_MODEL_TYPES }
	An identifier for the general type of model. More...
enum	ModelFlags {   GENERATE_CONSECUTIVE_IDS, BUILD_LOCAL_GRAPH, SYMMETRIZE_INPUT_TRANSPOSE, SYMMETRIZE_INPUT_BIPARTITE,   VERTICES_ARE_MATRIX_ROWS, VERTICES_ARE_MATRIX_COLUMNS, VERTICES_ARE_MATRIX_NONZEROS, VERTICES_ARE_MESH_NODES,   VERTICES_ARE_MESH_ELEMENTS, REMOVE_SELF_EDGES, BUILD_SUBSET_GRAPH, NUM_MODEL_FLAGS }
	Flags are set by a Problem to tell a Model what transformations it may need to do on the user's input. More...
enum	TestMode { Replace = 0, Add, Aggregate, TestMode_Max }

Functions
template<typename vtx_t , typename wgt_t >
static bool	compare_triplets (GMWM_triplet< vtx_t, wgt_t > a, GMWM_triplet< vtx_t, wgt_t > b)
template<typename vtx_t , typename wgt_t >
vtx_t	GreedyMWM (int *idx, vtx_t *adj, wgt_t *wgt, vtx_t tnVtx, vtx_t *match)
template<typename Adapter >
int	AlgSpectral (const RCP< GraphModel< Adapter > > &model, const RCP< LocalOrderingSolution< typename Adapter::lno_t > > &solution, const RCP< Teuchos::ParameterList > &pl, const RCP< const Teuchos::Comm< int > > &comm)
template<class IT , class WT >
void	uqsort (IT n, uSortItem< IT, WT > *arr)
	Quick sort function. Sorts the arr of uSortItems, with respect to increasing vals. DOCWORK: Document input params. More...
template<class IT , class WT , class SIGN >
void	uqSignsort (IT n, uSignedSortItem< IT, WT, SIGN > *arr)
	Quick sort function. Sorts the arr of uSignedSortItems, with respect to increasing vals. More...
static void	getScotchParameters (Teuchos::ParameterList &pl)
template<typename it >
it	z2Fact (it x)
template<typename IT >
void	ithPermutation (const IT n, IT i, IT *perm)
template<typename part_t >
void	getGridCommunicationGraph (part_t taskCount, part_t *&task_comm_xadj, part_t *&task_comm_adj, std::vector< int > grid_dims)
template<typename Adapter , typename scalar_t , typename part_t >
void	getSolutionCenterCoordinates (const Environment *envConst, const Teuchos::Comm< int > *comm, const Zoltan2::CoordinateModel< typename Adapter::base_adapter_t > *coords, const part_t *parts, int coordDim, part_t ntasks, scalar_t **partCenters)
template<typename Adapter , typename scalar_t , typename part_t >
void	getCoarsenedPartGraph (const Environment *envConst, const Teuchos::Comm< int > *comm, const Zoltan2::GraphModel< typename Adapter::base_adapter_t > *graph, part_t np, const part_t *parts, ArrayRCP< part_t > &g_part_xadj, ArrayRCP< part_t > &g_part_adj, ArrayRCP< scalar_t > &g_part_ew)
template<typename T >
void	fillContinousArray (T *arr, size_t arrSize, T *val)
	fillContinousArray function More...
template<typename part_t , typename pcoord_t , typename tcoord_t >
void	coordinateTaskMapperInterface (RCP< const Teuchos::Comm< int > > problemComm, int proc_dim, int num_processors, pcoord_t **machine_coords, int task_dim, part_t num_tasks, tcoord_t **task_coords, part_t *task_comm_xadj, part_t *task_comm_adj, pcoord_t *task_communication_edge_weight_, part_t *proc_to_task_xadj, part_t *proc_to_task_adj, int recursion_depth, Kokkos::View< part_t *, Kokkos::HostSpace > part_no_array, const part_t *machine_dimensions, int num_ranks_per_node=1, bool divide_to_prime_first=false)
	Constructor The interface function that calls CoordinateTaskMapper which will also perform the mapping operation. The result mapping can be obtained by -proc_to_task_xadj: Holds the beginning and end indices of tasks on proc_to_task_adj that is assigned to a processor. The tasks assigned to processor i are between proc_to_task_xadj[i] and proc_to_task_xadj[i + 1] on proc_to_task_adj. More...
template<typename proc_coord_t , typename v_lno_t >
void	visualize_mapping (int myRank, const int machine_coord_dim, const int num_ranks, proc_coord_t **machine_coords, const v_lno_t num_tasks, const v_lno_t *task_communication_xadj, const v_lno_t *task_communication_adj, const int *task_to_rank)
template<typename Adapter >
static int	zoltanNumObj (void *data, int *ierr)
template<typename Adapter >
static void	zoltanObjList (void *data, int nGidEnt, int nLidEnt, ZOLTAN_ID_PTR gids, ZOLTAN_ID_PTR lids, int wdim, float *wgts, int *ierr)
template<typename Adapter >
static void	zoltanParts (void *data, int, int nLidEnt, int nObj, ZOLTAN_ID_PTR, ZOLTAN_ID_PTR lids, int *parts, int *ierr)
template<typename Adapter >
static int	zoltanNumGeom (void *data, int *ierr)
template<typename Adapter >
static void	zoltanGeom (void *data, int, int nLidEnt, int nObj, ZOLTAN_ID_PTR, ZOLTAN_ID_PTR lids, int nDim, double *coords, int *ierr)
template<typename Adapter >
static void	zoltanHGSizeCS_withGraphAdapter (void *data, int *nLists, int *nPins, int *format, int *ierr)
template<typename Adapter >
static void	zoltanHGCS_withGraphAdapter (void *data, int nGidEnt, int nLists, int, int, ZOLTAN_ID_PTR listIds, int *listIdx, ZOLTAN_ID_PTR pinIds, int *ierr)
template<typename Adapter >
static void	zoltanHGSizeEdgeWts_withGraphAdapter (void *data, int *nEdges, int *ierr)
template<typename Adapter >
static void	zoltanHGEdgeWts_withGraphAdapter (void *data, int nGidEnt, int nLidEnt, int nEdges, int, ZOLTAN_ID_PTR edgeGids, ZOLTAN_ID_PTR edgeLids, float *edgeWgts, int *ierr)
template<typename Adapter >
static void	zoltanHGSizeCS_withMatrixAdapter (void *data, int *nLists, int *nPins, int *format, int *ierr)
template<typename Adapter >
static void	zoltanHGCS_withMatrixAdapter (void *data, int nGidEnt, int nLists, int nPins, int format, ZOLTAN_ID_PTR listIds, int *listIdx, ZOLTAN_ID_PTR pinIds, int *ierr)
template<typename Adapter >
static void	zoltanHGSizeCS_withMeshAdapter (void *data, int *nLists, int *nPins, int *format, int *ierr)
template<typename Adapter >
static void	zoltanHGCS_withMeshAdapter (void *data, int nGidEnt, int nLists, int nPins, int format, ZOLTAN_ID_PTR listIds, int *listIdx, ZOLTAN_ID_PTR pinIds, int *ierr)
void	makeDebugManager (int rank, bool iPrint, int level, std::string fname, int ost, Teuchos::RCP< DebugManager > &mgr)
	Create an output manager for debugging or status information. More...
template<typename metric_t >
void	makeMetricOutputManager (int rank, bool iPrint, std::string fname, int ost, Teuchos::RCP< MetricOutputManager< metric_t > > &mgr, std::string units, int fieldWidth, RCP< std::ofstream > &fptr)
	Create an output manager for a metric value. More...
template<typename Integral >
bool	validIntegralRangeList (const Teuchos::Array< Integral > &vals)
	A helper function that indicates whether an array is a valid integer range list. More...
template<typename Integral >
bool	allValuesAreInRangeList (const Teuchos::Array< Integral > &vals)
	A helper function that determines if all values are in the list. More...
template<typename Integral >
bool	allValuesAreInRangeList (const Teuchos::ParameterEntry &e)
	A helper function that determines if all values are in the list. More...
template<typename Integral >
bool	noValuesAreInRangeList (const Teuchos::Array< Integral > &vals)
	A helper function that determines if no values are in the list. More...
template<typename Integral >
bool	noValuesAreInRangeList (const Teuchos::ParameterEntry &e)
	A helper function that determines if no values are in the list. More...
template<typename Integral >
bool	IsInRangeList (const Integral val, const Teuchos::Array< Integral > &valList, bool sorted=true)
	A helper function that determines if a value is in the list. More...
template<typename Integral >
bool	IsInRangeList (const Integral val, const Teuchos::ParameterEntry &e)
	A helper function that determines if a value is in the list. More...
template<typename Integral >
Teuchos::ArrayView< Integral >	getList (Teuchos::Array< Integral > &irl)
	A helper function that returns a view of the list. More...
template<typename Integral >
void	printIntegralRangeList (std::ostream &os, Teuchos::Array< Integral > &irl)
	A helper function that prints the meaning of an encoded integer range list. More...
void	createAllParameters (Teuchos::ParameterList &pList)
	Create a list of all Zoltan2 parameters and validators. More...
static void	setValidatorsInList (const Teuchos::ParameterList &plSome, const Teuchos::ParameterList &plAll, Teuchos::ParameterList &plVal)
	Create a parameter list that can validate a specific list of parameters. More...
void	createValidatorList (const Teuchos::ParameterList &plIn, Teuchos::ParameterList &plOut)
	Create a list by adding validators to the users parameter list. More...
void	printListDocumentation (const Teuchos::ParameterList &pl, std::ostream &os, std::string listNames)
template<typename User >
RCP< Tpetra::CrsMatrix< int, typename MeshAdapter< User > ::lno_t, typename MeshAdapter < User >::gno_t, typename MeshAdapter< User >::node_t > >	get2ndAdjsMatFromAdjs (const Teuchos::RCP< const MeshAdapter< User > > &ia, const RCP< const Comm< int > > comm, Zoltan2::MeshEntityType sourcetarget, Zoltan2::MeshEntityType through)
template<typename User >
void	get2ndAdjsViewFromAdjs (const Teuchos::RCP< const MeshAdapter< User > > &ia, const RCP< const Comm< int > > comm, Zoltan2::MeshEntityType sourcetarget, Zoltan2::MeshEntityType through, Teuchos::ArrayRCP< typename MeshAdapter< User >::offset_t > &offsets, Teuchos::ArrayRCP< typename MeshAdapter< User >::gno_t > &adjacencyIds)
template<typename SolutionAdapter , typename DataAdapter >
size_t	getImportList (const PartitioningSolution< SolutionAdapter > &solution, const DataAdapter *const data, ArrayRCP< typename DataAdapter::gno_t > &imports)
	From a PartitioningSolution, get a list of IDs to be imported. Assumes part numbers in PartitioningSolution are to be used as rank numbers. More...
void	AlltoAllCount (const Comm< int > &comm, const Environment &env, const ArrayView< const int > &sendCount, const ArrayView< int > &recvCount)
	Each process sends a value to every process, an all-to-all. More...
template<>
void	AlltoAllv (const Comm< int > &comm, const Environment &env, const ArrayView< const std::string > &sendBuf, const ArrayView< const int > &sendCount, ArrayRCP< std::string > &recvBuf, const ArrayView< int > &recvCount)
template<>
void	AlltoAllv (const Comm< int > &comm, const Environment &env, const ArrayView< const unsigned short > &sendBuf, const ArrayView< const int > &sendCount, ArrayRCP< unsigned short > &recvBuf, const ArrayView< int > &recvCount)
template<>
void	AlltoAllv (const Comm< int > &comm, const Environment &env, const ArrayView< const unsigned char > &sendBuf, const ArrayView< const int > &sendCount, ArrayRCP< unsigned char > &recvBuf, const ArrayView< int > &recvCount)
template<typename T >
void	AlltoAllv (const Comm< int > &comm, const Environment &env, const ArrayView< const T > &sendBuf, const ArrayView< const int > &sendCount, ArrayRCP< T > &recvBuf, const ArrayView< int > &recvCount)
	AlltoAllv sends/receives data to/from all processes. More...
template<typename gno_t >
size_t	findUniqueGidsCommon (size_t num_keys, int num_gid, ZOLTAN_ID_PTR ddkeys, char *ddnewgids, MPI_Comm mpicomm)
template<typename lno_t , typename gno_t >
size_t	findUniqueGids (Tpetra::MultiVector< gno_t, lno_t, gno_t > &keys, Tpetra::Vector< gno_t, lno_t, gno_t > &gids)
template<typename key_t , typename gno_t >
size_t	findUniqueGids (std::vector< key_t > &keys, std::vector< gno_t > &gids, const Teuchos::Comm< int > &comm)
template<typename Adapter , typename MachineRep = MachineRepresentation<typename Adapter::scalar_t, typename Adapter::part_t>>
void	globalWeightedByPart (const RCP< const Environment > &env, const RCP< const Comm< int > > &comm, const RCP< const GraphModel< typename Adapter::base_adapter_t > > &graph, const ArrayView< const typename Adapter::part_t > &parts, typename Adapter::part_t &numParts, ArrayRCP< RCP< BaseClassMetrics< typename Adapter::scalar_t > > > &metrics, ArrayRCP< typename Adapter::scalar_t > &globalSums, bool bMessages=true, const RCP< const MachineRep > machine=Teuchos::null)
	Given the local partitioning, compute the global weighted cuts in each part. More...
template<typename scalar_t , typename part_t >
void	printGraphMetricsHeader (std::ostream &os, part_t targetNumParts, part_t numParts)
	Print out header info for graph metrics. More...
template<typename scalar_t , typename part_t >
void	printGraphMetrics (std::ostream &os, part_t targetNumParts, part_t numParts, const ArrayView< RCP< BaseClassMetrics< scalar_t > > > &infoList)
	Print out list of graph metrics. More...
template<typename scalar_t , typename part_t >
void	printGraphMetrics (std::ostream &os, part_t targetNumParts, part_t numParts, RCP< BaseClassMetrics< scalar_t >> metricValue)
	Print out header and a single graph metric. More...
template<typename scalar_t , typename lno_t , typename part_t >
void	globalSumsByPart (const RCP< const Environment > &env, const RCP< const Comm< int > > &comm, const ArrayView< const part_t > &part, int vwgtDim, const ArrayView< StridedData< lno_t, scalar_t > > &vwgts, multiCriteriaNorm mcNorm, part_t targetNumParts, part_t &numExistingParts, part_t &numNonemptyParts, ArrayRCP< RCP< BaseClassMetrics< scalar_t > > > &metrics, ArrayRCP< scalar_t > &globalSums)
	Given the local partitioning, compute the global sums in each part. More...
template<typename Adapter >
void	imbalanceMetrics (const RCP< const Environment > &env, const RCP< const Comm< int > > &comm, multiCriteriaNorm mcNorm, const Adapter *ia, const PartitioningSolution< Adapter > *solution, const ArrayView< const typename Adapter::part_t > &partArray, const RCP< const GraphModel< typename Adapter::base_adapter_t > > &graphModel, typename Adapter::part_t &numExistingParts, typename Adapter::part_t &numNonemptyParts, ArrayRCP< RCP< BaseClassMetrics< typename Adapter::scalar_t > > > &metrics)
	Compute imbalance metrics for a distribution. More...
template<typename scalar_t , typename part_t >
void	printImbalanceMetricsHeader (std::ostream &os, part_t targetNumParts, part_t numExistingParts, part_t numNonemptyParts)
	Print out header info for imbalance metrics. More...
template<typename scalar_t , typename part_t >
void	printImbalanceMetrics (std::ostream &os, part_t targetNumParts, part_t numExistingParts, part_t numNonemptyParts, const ArrayView< RCP< BaseClassMetrics< scalar_t > > > &infoList)
	Print out list of imbalance metrics. More...
template<typename scalar_t , typename part_t >
void	printImbalanceMetrics (std::ostream &os, part_t targetNumParts, part_t numExistingParts, part_t numNonemptyParts, RCP< BaseClassMetrics< scalar_t >> metricValue)
	Print out header and a single imbalance metric. More...
void	addNumberToFileName (int number, std::string fname, std::string &newf)
	Helper method to add number to a file name. More...
template<typename metric_t , typename scalar_t >
RCP< metric_t >	addNewMetric (const RCP< const Environment > &env, ArrayRCP< RCP< BaseClassMetrics< scalar_t > > > &metrics)
template<typename scalar_t >
void	getStridedStats (const ArrayView< scalar_t > &v, int stride, int offset, scalar_t &min, scalar_t &max, scalar_t &sum)
	Find min, max and sum of metric values. More...
template<typename scalar_t >
void	getStridedStats (const ArrayView< scalar_t > &v, int stride, int offset, scalar_t &max, scalar_t &sum)
	Find max and sum of graph metric values. More...
template<typename scalar_t , typename lno_t , typename part_t >
void	normedPartWeights (const RCP< const Environment > &env, part_t numberOfParts, const ArrayView< const part_t > &parts, const ArrayView< StridedData< lno_t, scalar_t > > &vwgts, multiCriteriaNorm mcNorm, scalar_t *weights)
	Compute the total weight in each part on this process. More...
template<typename scalar_t , typename part_t >
void	computeImbalances (part_t numExistingParts, part_t targetNumParts, const scalar_t *psizes, scalar_t sumVals, const scalar_t *vals, scalar_t &min, scalar_t &max, scalar_t &avg)
	Compute the imbalance. More...
template<typename scalar_t , typename part_t >
void	computeImbalances (part_t numExistingParts, part_t targetNumParts, int numSizes, ArrayView< ArrayRCP< scalar_t > > psizes, scalar_t sumVals, const scalar_t *vals, scalar_t &min, scalar_t &max, scalar_t &avg)
	Compute the imbalance in the case of multiple part sizes. More...
template<typename scalar_t >
scalar_t	normedWeight (ArrayView< scalar_t > weights, multiCriteriaNorm norm)
	Compute the norm of the vector of weights. More...
template<typename lno_t , typename scalar_t >
scalar_t	normedWeight (ArrayView< StridedData< lno_t, scalar_t > > weights, lno_t idx, multiCriteriaNorm norm)
	Compute the norm of the vector of weights stored as StridedData. More...
template<typename scalar_t , typename T >
static void	getInputArrayHelper (ArrayRCP< const T > &target, const ArrayRCP< const scalar_t > &src)
template<typename scalar_t >
static void	getInputArrayHelper (ArrayRCP< const scalar_t > &target, const ArrayRCP< const scalar_t > &src)
long	getProcessKilobytes ()
template<typename scalar_t >
bool	outsideRegion (scalar_t val, scalar_t mark, double epsilon)
static void	AssertCondition (bool condition, const std::string &message, const char *file=__FILE__, int line=__LINE__)
std::string	Zoltan2_Version ()
static void	getSphynxValidParameters (ParameterList &pl)
	Set up validators specific to this algorithm. More...
static void	setSphynxValidatorsInList (const Teuchos::ParameterList &plSome, const Teuchos::ParameterList &plAll, Teuchos::ParameterList &plVal)
std::string	Sphynx_Version ()
template<typename key_t , typename gno_t >
size_t	findUniqueGids (const std::vector< key_t > &keys, std::vector< gno_t > &gids, Teuchos::RCP< const Teuchos::Comm< int > > &comm)
	DECL_TYPE_NAME (int)
	DECL_TYPE_NAME (long long)
void	checkNUnique (std::string &name, size_t nUniqueGids, size_t nExpected)
template<typename gno_t >
void	checkMaxGid (std::string &name, std::vector< gno_t > &gids, gno_t maxExpected, Teuchos::RCP< const Teuchos::Comm< int > > &comm)
template<typename gno_t >
void	checkMinGid (std::string &name, std::vector< gno_t > &gids, gno_t minExpected, Teuchos::RCP< const Teuchos::Comm< int > > &comm)
template<typename gno_t >
void	checkNLocallyUnique (std::string &name, std::vector< gno_t > &gids, size_t nExpected)
template<typename gno_t >
void	test1 (Teuchos::RCP< const Teuchos::Comm< int > > &comm)
template<typename gno_t >
void	test2 (Teuchos::RCP< const Teuchos::Comm< int > > &comm)
template<typename gno_t >
void	test3 (Teuchos::RCP< const Teuchos::Comm< int > > &comm)
template<typename gno_t >
void	test4 (Teuchos::RCP< const Teuchos::Comm< int > > &comm)
void	print_proc_safe (const std::string &message, Teuchos::RCP< const Teuchos::Comm< int > > comm, bool only_rank0=true)
int	runDirectoryTests (int narg, char **arg)

Variables
static const std::string	fail = "FAIL "
static const std::string	pass = " "

Detailed Description

Created by mbenlioglu on Aug 31, 2020.

CommGraphModel creates a graph representing the communication topology of

the MPI ranks for a given XpetraCrsGraphAdapter object. If there are n MPI ranks in the given communicator, then this model contains n vertices so that each vertex represents an MPI rank. If rank i sends a message to rank j (during the mat-vec on the matrix corresponding to the given adapter), then there is a directed edge from vertex i to vertex j in the graph. The size of the edge is the number of nonzeros that cause that message. The weight of vertex i is the number of nonzeros currently residing at rank i.

Since the above mentioned graph is too small, we migrate it into a subset of ranks, which we call activeRanks. nActiveRanks_ denotes the number of active ranks and is computed as n/threshold_.

Typedef Documentation

typedef ::Tpetra::Details::DefaultTypes::local_ordinal_type Zoltan2::default_lno_t

Definition at line 37 of file Zoltan2_InputTraits.hpp.

typedef ::Tpetra::Details::DefaultTypes::global_ordinal_type Zoltan2::default_gno_t

Definition at line 40 of file Zoltan2_InputTraits.hpp.

typedef size_t Zoltan2::default_offset_t

Definition at line 43 of file Zoltan2_InputTraits.hpp.

typedef ::Tpetra::Details::DefaultTypes::scalar_type Zoltan2::default_scalar_t

Definition at line 46 of file Zoltan2_InputTraits.hpp.

typedef ::Tpetra::Details::DefaultTypes::node_type Zoltan2::default_node_t

Definition at line 49 of file Zoltan2_InputTraits.hpp.

typedef int Zoltan2::default_part_t

Definition at line 52 of file Zoltan2_InputTraits.hpp.

typedef std::bitset<NUM_MODEL_FLAGS> Zoltan2::modelFlag_t

Definition at line 57 of file Zoltan2_Model.hpp.

typedef Tpetra::global_size_t Zoltan2::global_size_t

Definition at line 83 of file Zoltan2_Standards.hpp.

Enumeration Type Documentation

enum Zoltan2::blockParams

The boolean parameters of interest to the Block algorithm.

Enumerator
block_balanceCount	objective = balance_object_count
block_balanceWeight	objective = balance_object_weight
block_minTotalWeight	objective = mc_minimize_total_weight
block_minMaximumWeight	objective = mc_minimize_maximum_weight
block_balanceTotalMaximum	objective = mc_balance_total_maximum
NUM_BLOCK_PARAMS

Definition at line 29 of file Zoltan2_AlgBlock.hpp.

enum Zoltan2::RangeType

Codes that indicate how to interpret the Array<int> representing the user's integer range list.

Enumerator
RANGE_INCLUDES_ALL	all values were listed
RANGE_IS_EMPTY	no values were listed
RANGE_IS_LISTED	the listed values are in the Array<int>
NUM_RANGE_TYPES

Definition at line 40 of file Zoltan2_IntegerRangeList.hpp.

enum Zoltan2::AssertionLevel

Level of error checking or assertions desired.

Each assertion in the code must have a level. Tests for logic errors should always be level DEBUG_MODE_ASSERTION. Quick tests are BASIC, longer tests for input errors are COMPLEX.

The user sets the assertion level with the parameter error_check_level.

If compiled with Z2_OMIT_ALL_ERROR_CHECKING, error checks don't happen.

Enumerator
NO_ASSERTIONS	no assertion checks will be done
BASIC_ASSERTION	fast typical checks for valid arguments
COMPLEX_ASSERTION	more involved, like validate a graph
DEBUG_MODE_ASSERTION	checks for logic errors
NUM_ASSERTION_LEVELS

Definition at line 51 of file Zoltan2_Parameters.hpp.

enum Zoltan2::MessageOutputLevel

The amount of debugging or status output to print.

Each debug/status message must have an output level. The user specfies the level desired with the debug_level parameter.

If Zoltan2 is compiled with Z2_OMIT_ALL_STATUS_MESSAGES, no messages will be processed.

Enumerator
NO_STATUS	don't display status/debug messages
BASIC_STATUS	the status at each high level step
DETAILED_STATUS	sub-steps, each method's entry and exit
VERBOSE_DETAILED_STATUS	include more detail about sub-steps
NUM_STATUS_OUTPUT_LEVELS

Definition at line 68 of file Zoltan2_Parameters.hpp.

enum Zoltan2::TimerType

The type of timers which should be active.

If timing is requested, timer_type can specify MACRO timing or MICRO timing. For example, a MACRO timer would time an algorithm. A MICRO timer would time steps in the algorithm. You can also ask for BOTH, but be aware that your MACRO timer is also timing the MICRO timers as well as the algorithm.

If Zoltan2 is compiled with Z2_OMIT_ALL_PROFILING timing messages are ignored.

Enumerator
NO_TIMERS	No timing data will be collected (the default).
MACRO_TIMERS	Time an algorithm (or other entity) as a whole.
MICRO_TIMERS	Time the substeps of an entity.
BOTH_TIMERS	Run both MACRO and MICRO timers.
TEST_TIMERS	Timers added while testing, removed later.
NUM_TIMING_OPTIONS

Definition at line 88 of file Zoltan2_Parameters.hpp.

enum Zoltan2::OSType

Output stream types.

Enumerator
COUT_STREAM	std::cout
CERR_STREAM	std::cerr
NULL_STREAM	/dev/null: do actions but don't output results
NUM_OUTPUT_STREAMS

Definition at line 100 of file Zoltan2_Parameters.hpp.

enum Zoltan2::multiCriteriaNorm

Enumerator used in code for multicriteria norm choice.

Enumerator
normMinimizeTotalWeight	1-norm = Manhattan norm
normBalanceTotalMaximum	2-norm = sqrt of sum of squares
normMinimizeMaximumWeight	inf-norm = maximum norm
normNumNorms

Definition at line 109 of file Zoltan2_Parameters.hpp.

enum Zoltan2::BaseAdapterType

An enum to identify general types of adapters.

Enumerator
InvalidAdapterType	unused value
IdentifierAdapterType	identifier data, just a list of IDs
VectorAdapterType	vector data
MatrixAdapterType	matrix data
GraphAdapterType	graph data
MeshAdapterType	mesh data

Definition at line 32 of file Zoltan2_Adapter.hpp.

enum Zoltan2::GraphEntityType

Enumerated entity type for graphs: Vertices or Edges.

Enumerator
GRAPH_VERTEX
GRAPH_EDGE

Definition at line 24 of file Zoltan2_GraphAdapter.hpp.

enum Zoltan2::MatrixEntityType

Enumerator
MATRIX_ROW
MATRIX_COLUMN
MATRIX_NONZERO

Definition at line 22 of file Zoltan2_MatrixAdapter.hpp.

enum Zoltan2::MeshEntityType

Enumerate entity types for meshes: Regions, Faces, Edges, or Vertices.

Enumerator
MESH_VERTEX
MESH_EDGE
MESH_FACE
MESH_REGION

Definition at line 27 of file Zoltan2_MeshAdapter.hpp.

enum Zoltan2::EntityTopologyType

Enumerate entity topology types for meshes: points,lines,polygons,triangles,quadrilaterals, polyhedrons, tetrahedrons, hexhedrons, prisms, or pyramids.

Enumerator
POINT
LINE_SEGMENT
POLYGON
TRIANGLE
QUADRILATERAL
POLYHEDRON
TETRAHEDRON
HEXAHEDRON
PRISM
PYRAMID

Definition at line 39 of file Zoltan2_MeshAdapter.hpp.

enum Zoltan2::CentricView

Enumerate the views for the pins: HYPEREDGE_CENTRIC: pins are the global ids of the vertices as seen by the hyperedges VERTEX_CENTRIC: pins are the global ids of the hyperedges as seen by the vertices.

Enumerator
HYPEREDGE_CENTRIC
VERTEX_CENTRIC

Definition at line 38 of file Zoltan2_HyperGraphModel.hpp.

enum Zoltan2::ModelType

An identifier for the general type of model.

Enumerator
HypergraphModelType
GraphModelType
CoordinateModelType
IdentifierModelType
MAX_NUM_MODEL_TYPES

Definition at line 24 of file Zoltan2_Model.hpp.

enum Zoltan2::ModelFlags

Flags are set by a Problem to tell a Model what transformations it may need to do on the user's input.

Enumerator
GENERATE_CONSECUTIVE_IDS	algorithm requires consecutive ids
BUILD_LOCAL_GRAPH	model represents graph within only one rank
SYMMETRIZE_INPUT_TRANSPOSE	model must symmetrize input
SYMMETRIZE_INPUT_BIPARTITE	model must symmetrize input
VERTICES_ARE_MATRIX_ROWS	use matrix rows as graph vertices
VERTICES_ARE_MATRIX_COLUMNS	use columns as graph vertices
VERTICES_ARE_MATRIX_NONZEROS	use nonzeros as graph vertices
VERTICES_ARE_MESH_NODES	use mesh nodes as vertices
VERTICES_ARE_MESH_ELEMENTS	use mesh elements as vertices
REMOVE_SELF_EDGES	algorithm requires no self edges
BUILD_SUBSET_GRAPH	ignore invalid neighbors
NUM_MODEL_FLAGS

Definition at line 38 of file Zoltan2_Model.hpp.

enum Zoltan2::TestMode

Enumerator
Replace
Add
Aggregate
TestMode_Max

Definition at line 21 of file directoryTest_Impl.hpp.

Function Documentation

template<typename vtx_t , typename wgt_t >

static bool Zoltan2::compare_triplets ( GMWM_triplet< vtx_t, wgt_t >  a, GMWM_triplet< vtx_t, wgt_t >  b  )

static

Definition at line 41 of file Zoltan2_GreedyMWM.hpp.

template<typename vtx_t , typename wgt_t >

vtx_t Zoltan2::GreedyMWM	(	int *	idx,
		vtx_t *	adj,
		wgt_t *	wgt,
		vtx_t	tnVtx,
		vtx_t *	match
	)

Definition at line 49 of file Zoltan2_GreedyMWM.hpp.

template<typename Adapter >

int Zoltan2::AlgSpectral	(	const RCP< GraphModel< Adapter > > &	model,
		const RCP< LocalOrderingSolution< typename Adapter::lno_t > > &	solution,
		const RCP< Teuchos::ParameterList > &	pl,
		const RCP< const Teuchos::Comm< int > > &	comm
	)

Definition at line 27 of file Zoltan2_AlgSpectral.hpp.

template<class IT , class WT >

void Zoltan2::uqsort	(	IT	n,
		uSortItem< IT, WT > *	arr
	)

Quick sort function. Sorts the arr of uSortItems, with respect to increasing vals. DOCWORK: Document input params.

Definition at line 159 of file Zoltan2_AlgMultiJagged.hpp.

template<class IT , class WT , class SIGN >

void Zoltan2::uqSignsort	(	IT	n,
		uSignedSortItem< IT, WT, SIGN > *	arr
	)

Quick sort function. Sorts the arr of uSignedSortItems, with respect to increasing vals.

Definition at line 268 of file Zoltan2_AlgMultiJagged.hpp.

static void Zoltan2::getScotchParameters ( Teuchos::ParameterList &  pl )

inlinestatic

Definition at line 29 of file Zoltan2_AlgScotch.hpp.

template<typename it >

it Zoltan2::z2Fact ( it  x )

inline

Definition at line 84 of file Zoltan2_TaskMapping.hpp.

template<typename IT >

void Zoltan2::ithPermutation	(	const IT	n,
		IT	i,
		IT *	perm
	)

Definition at line 97 of file Zoltan2_TaskMapping.hpp.

template<typename part_t >

void Zoltan2::getGridCommunicationGraph	(	part_t	taskCount,
		part_t *&	task_comm_xadj,
		part_t *&	task_comm_adj,
		std::vector< int >	grid_dims
	)

Definition at line 126 of file Zoltan2_TaskMapping.hpp.

template<typename Adapter , typename scalar_t , typename part_t >

void Zoltan2::getSolutionCenterCoordinates	(	const Environment *	envConst,
		const Teuchos::Comm< int > *	comm,
		const Zoltan2::CoordinateModel< typename Adapter::base_adapter_t > *	coords,
		const part_t *	parts,
		int	coordDim,
		part_t	ntasks,
		scalar_t **	partCenters
	)

Definition at line 160 of file Zoltan2_TaskMapping.hpp.

template<typename Adapter , typename scalar_t , typename part_t >

void Zoltan2::getCoarsenedPartGraph	(	const Environment *	envConst,
		const Teuchos::Comm< int > *	comm,
		const Zoltan2::GraphModel< typename Adapter::base_adapter_t > *	graph,
		part_t	np,
		const part_t *	parts,
		ArrayRCP< part_t > &	g_part_xadj,
		ArrayRCP< part_t > &	g_part_adj,
		ArrayRCP< scalar_t > &	g_part_ew
	)

Definition at line 247 of file Zoltan2_TaskMapping.hpp.

template<typename T >

void Zoltan2::fillContinousArray	(	T *	arr,
		size_t	arrSize,
		T *	val
	)

fillContinousArray function

Parameters

arr	array to be filled in with values.
arrSize	the size of the array.
val	the pointer to the value to be filled. if given NULL, the filling performs arr[i] = i.

Definition at line 908 of file Zoltan2_TaskMapping.hpp.

template<typename part_t , typename pcoord_t , typename tcoord_t >

void Zoltan2::coordinateTaskMapperInterface	(	RCP< const Teuchos::Comm< int > >	problemComm,
		int	proc_dim,
		int	num_processors,
		pcoord_t **	machine_coords,
		int	task_dim,
		part_t	num_tasks,
		tcoord_t **	task_coords,
		part_t *	task_comm_xadj,
		part_t *	task_comm_adj,
		pcoord_t *	task_communication_edge_weight_,
		part_t *	proc_to_task_xadj,
		part_t *	proc_to_task_adj,
		int	recursion_depth,
		Kokkos::View< part_t *, Kokkos::HostSpace >	part_no_array,
		const part_t *	machine_dimensions,
		int	num_ranks_per_node = 1,
		bool	divide_to_prime_first = false
	)

Constructor The interface function that calls CoordinateTaskMapper which will also perform the mapping operation. The result mapping can be obtained by -proc_to_task_xadj: Holds the beginning and end indices of tasks on proc_to_task_adj that is assigned to a processor. The tasks assigned to processor i are between proc_to_task_xadj[i] and proc_to_task_xadj[i + 1] on proc_to_task_adj.

-proc_to_task_adj: holds the task adj array.

-task_comm_xadj, task_comm_adj, task_communication_edge_weight_ can be provided NULL. In this case all processors will calculate the same mapping.

-If task_comm_xadj, task_comm_adj and provided, algorithm will perform rotations, and processors will calculate different mappings, and best one will be reduced.

-If task_communication_edge_weight_ is provided with task_comm_xadj, task_comm_adj, this will be used when cost is calculated.

-recursion_depth is a mandatory argument. In the case part_no_array is not null, this parameter should represent the length of part_no_array. If part_no_array is given as NULL, then this will give the recursion depth for the algorithm, Maximum number is ceil(log_2(min(num_processors, num_tasks))), and providing a higher number will be equivalant to this. Partitioning algorithm will work as RCB when maximum number is given, which performs the best mapping results.

-part_no_array: The best results are obtained when this parameter is given as NULL. But if this is provided, partitioning will use this array for partitioning each dimension to the given numbers. The multiplication of these numbers should be equal to min(num_processors, num_tasks).

-machine_dimensions: This can be NULL, but if provided the algorithm will perform shift of the machine coords so that the largest gap is treated as wrap-around link.

Parameters

problemComm	is the communication object.
proc_dim	dimensions of the processor coordinates.
num_processors	is the number of processors
machine_coords	is the coordinates of the processors.
task_dim	is the dimension of the tasks.
num_tasks	is the number of tasks.
task_coords	is the coordinates of the tasks.
task_comm_xadj	is the task communication graphs xadj array. (task i's adjacency is between task_comm_xadj[i] and task_comm_xadj[i + 1])
task_comm_adj	is task communication graphs adj array.
task_communication_edge_weight_	is the weight of the communication in task graph.
proc_to_task_xadj	is is the output for tasks showing which proc has the which parts. (proc-i will own the tasks from proc_to_task_xadj[i] to proc_to_task_xadj[i + 1])
proc_to_task_adj	is the ouput list of tasks pointed by proc_to_task_xadj
recursion_depth	is the recursion depth that will be applied to partitioning. If part_no_array is provided, then it is the length of this array.
part_no_array	if part_no_array is provided, partitioning algorithm will be forced to use this array for partitioning. However, the multiplication of each entries in this array should be equal to min(num_processors, num_tasks).
*machine_dimensions,:	the dimensions of the machine network. For example for hopper 17x8x24 This can be NULL, but if provided the algorithm will perform shift of the machine coords so that the largest gap is treated as wrap-around link.

Definition at line 3277 of file Zoltan2_TaskMapping.hpp.

template<typename proc_coord_t , typename v_lno_t >

void Zoltan2::visualize_mapping ( int  myRank, const int  machine_coord_dim, const int  num_ranks, proc_coord_t **  machine_coords, const v_lno_t  num_tasks, const v_lno_t *  task_communication_xadj, const v_lno_t *  task_communication_adj, const int *  task_to_rank  )

inline

Definition at line 3357 of file Zoltan2_TaskMapping.hpp.

template<typename Adapter >

static int Zoltan2::zoltanNumObj ( void *  data, int *  ierr  )

static

Definition at line 41 of file Zoltan2_AlgZoltanCallbacks.hpp.

template<typename Adapter >

static void Zoltan2::zoltanObjList ( void *  data, int  nGidEnt, int  nLidEnt, ZOLTAN_ID_PTR  gids, ZOLTAN_ID_PTR  lids, int  wdim, float *  wgts, int *  ierr  )

static

Definition at line 50 of file Zoltan2_AlgZoltanCallbacks.hpp.

template<typename Adapter >

static void Zoltan2::zoltanParts ( void *  data, int  , int  nLidEnt, int  nObj, ZOLTAN_ID_PTR  , ZOLTAN_ID_PTR  lids, int *  parts, int *  ierr  )

static

Definition at line 93 of file Zoltan2_AlgZoltanCallbacks.hpp.

template<typename Adapter >

static int Zoltan2::zoltanNumGeom ( void *  data, int *  ierr  )

static

Definition at line 113 of file Zoltan2_AlgZoltanCallbacks.hpp.

template<typename Adapter >

static void Zoltan2::zoltanGeom ( void *  data, int  , int  nLidEnt, int  nObj, ZOLTAN_ID_PTR  , ZOLTAN_ID_PTR  lids, int  nDim, double *  coords, int *  ierr  )

static

Definition at line 123 of file Zoltan2_AlgZoltanCallbacks.hpp.

template<typename Adapter >

static void Zoltan2::zoltanHGSizeCS_withGraphAdapter ( void *  data, int *  nLists, int *  nPins, int *  format, int *  ierr  )

static

Definition at line 153 of file Zoltan2_AlgZoltanCallbacks.hpp.

template<typename Adapter >

static void Zoltan2::zoltanHGCS_withGraphAdapter ( void *  data, int  nGidEnt, int  nLists, int  , int  , ZOLTAN_ID_PTR  listIds, int *  listIdx, ZOLTAN_ID_PTR  pinIds, int *  ierr  )

static

Definition at line 171 of file Zoltan2_AlgZoltanCallbacks.hpp.

template<typename Adapter >

static void Zoltan2::zoltanHGSizeEdgeWts_withGraphAdapter ( void *  data, int *  nEdges, int *  ierr  )

static

Definition at line 224 of file Zoltan2_AlgZoltanCallbacks.hpp.

template<typename Adapter >

static void Zoltan2::zoltanHGEdgeWts_withGraphAdapter ( void *  data, int  nGidEnt, int  nLidEnt, int  nEdges, int  , ZOLTAN_ID_PTR  edgeGids, ZOLTAN_ID_PTR  edgeLids, float *  edgeWgts, int *  ierr  )

static

Definition at line 242 of file Zoltan2_AlgZoltanCallbacks.hpp.

template<typename Adapter >

static void Zoltan2::zoltanHGSizeCS_withMatrixAdapter ( void *  data, int *  nLists, int *  nPins, int *  format, int *  ierr  )

static

Definition at line 305 of file Zoltan2_AlgZoltanCallbacks.hpp.

template<typename Adapter >

static void Zoltan2::zoltanHGCS_withMatrixAdapter ( void *  data, int  nGidEnt, int  nLists, int  nPins, int  format, ZOLTAN_ID_PTR  listIds, int *  listIdx, ZOLTAN_ID_PTR  pinIds, int *  ierr  )

static

Definition at line 346 of file Zoltan2_AlgZoltanCallbacks.hpp.

template<typename Adapter >

static void Zoltan2::zoltanHGSizeCS_withMeshAdapter ( void *  data, int *  nLists, int *  nPins, int *  format, int *  ierr  )

static

Definition at line 417 of file Zoltan2_AlgZoltanCallbacks.hpp.

template<typename Adapter >

static void Zoltan2::zoltanHGCS_withMeshAdapter ( void *  data, int  nGidEnt, int  nLists, int  nPins, int  format, ZOLTAN_ID_PTR  listIds, int *  listIdx, ZOLTAN_ID_PTR  pinIds, int *  ierr  )

static

Definition at line 454 of file Zoltan2_AlgZoltanCallbacks.hpp.

void Zoltan2::makeDebugManager	(	int	rank,
		bool	iPrint,
		int	level,
		std::string	fname,
		int	ost,
		Teuchos::RCP< DebugManager > &	mgr
	)

Create an output manager for debugging or status information.

Parameters

rank	the MPI rank of the calling process in the application
iPrint	true if this process should output information
fname	name of file to which output is to be appended, or or Z2_UNSET_STRING
ost	output stream type
mgr	on return, a pointer to the created output manager

Definition at line 42 of file Zoltan2_Environment.cpp.

template<typename metric_t >

void Zoltan2::makeMetricOutputManager	(	int	rank,
		bool	iPrint,
		std::string	fname,
		int	ost,
		Teuchos::RCP< MetricOutputManager< metric_t > > &	mgr,
		std::string	units,
		int	fieldWidth,
		RCP< std::ofstream > &	fptr
	)

Create an output manager for a metric value.

Parameters

rank	the MPI rank of the calling process in the application
iPrint	true if this process should print metric information
fname	name of file to which output is to be appended, or or Z2_UNSET_STRING
ost	output stream type
mgr	on return, a pointer to the created output manager
fptr	on return, an RCP to an ofstream if one was created.

The template parameter is the data type of the entity being measured.

Definition at line 620 of file Zoltan2_Environment.hpp.

template<typename Integral >

bool Zoltan2::validIntegralRangeList

(

const Teuchos::Array< Integral > &

vals

)

A helper function that indicates whether an array is a valid integer range list.

Parameters

vals	An array that may encode an integer range list.

Returns

true if the array encodes such a list, false otherwise.

Definition at line 60 of file Zoltan2_IntegerRangeList.hpp.

template<typename Integral >

bool Zoltan2::allValuesAreInRangeList

(

const Teuchos::Array< Integral > &

vals

)

A helper function that determines if all values are in the list.

Parameters

vals	An array encoding an integer range list.

Returns

true if the array encoding implies all values, false otherwise.

If the array is not a valid encoding of an integer range list, a std::runtime_error will be thrown.

If the user's parameter value was "all", then after validation the parameter value will be an array of size one containing a code that indicates all values are included, and this function will return true.

Definition at line 90 of file Zoltan2_IntegerRangeList.hpp.

template<typename Integral >

bool Zoltan2::allValuesAreInRangeList

(

const Teuchos::ParameterEntry &

e

)

A helper function that determines if all values are in the list.

Parameters

e	A parameter entry

Returns

true if the entry value is an array encoding all values, false otherwise.

If the entry value is not a valid encoding of an integer range list, a std::runtime_error will be thrown.

If the user's parameter value was "all", then after validation the parameter value will be an array of size one containing a code that indicates all values are included, and this function will return true.

Definition at line 112 of file Zoltan2_IntegerRangeList.hpp.

template<typename Integral >

bool Zoltan2::noValuesAreInRangeList

(

const Teuchos::Array< Integral > &

vals

)

A helper function that determines if no values are in the list.

Parameters

vals	An array encoding an integer range list.

Returns

true if the array encoding implies no values, false otherwise.

If the array is not a valid encoding of an integer range list, a std::runtime_error will be thrown.

If the user's parameter value was empty, then after validation the parameter value will be an array of size one containing a code that indicates no values are included, and this function will return true.

Definition at line 137 of file Zoltan2_IntegerRangeList.hpp.

template<typename Integral >

bool Zoltan2::noValuesAreInRangeList

(

const Teuchos::ParameterEntry &

e

)

A helper function that determines if no values are in the list.

Parameters

e	A parameter entry

Returns

true if the entry value is an array encoding no values, false otherwise.

If the entry value is not a valid encoding of an integer range list, a std::runtime_error will be thrown.

If the user's parameter value was empty, then after validation the parameter value will be an array of size one containing a code that indicates no values are included, and this function will return true.

Definition at line 159 of file Zoltan2_IntegerRangeList.hpp.

template<typename Integral >

bool Zoltan2::IsInRangeList	(	const Integral	val,
		const Teuchos::Array< Integral > &	valList,
		bool	sorted = true
	)

A helper function that determines if a value is in the list.

Parameters

val	A value that could be in the list.
vals	An array encoding an integer range list.
sorted	Set to false if the integer range list is not sorted

Returns

true if the encoding of vals implies val is in the list, false otherwise.

If vals is not a valid encoding of an integer range list, a std::runtime_error will be thrown.

Definition at line 184 of file Zoltan2_IntegerRangeList.hpp.

template<typename Integral >

bool Zoltan2::IsInRangeList	(	const Integral	val,
		const Teuchos::ParameterEntry &	e
	)

A helper function that determines if a value is in the list.

Parameters

val	A value that could be in the list.
e	A parameter entry

Returns

true if the entry value implies val is in the list, false otherwise.

If the entry value is not a valid encoding of an integer range list, a std::runtime_error will be thrown.

Definition at line 219 of file Zoltan2_IntegerRangeList.hpp.

template<typename Integral >

Teuchos::ArrayView<Integral> Zoltan2::getList

(

Teuchos::Array< Integral > &

irl

)

A helper function that returns a view of the list.

Parameters

irl	The value of an integer range list parameter after it has been validated

Returns

If the user's parameter value did not imply "all" or "none", then a view of the array of integers implied by the value is returned. Otherwise an empty array is returned.

Definition at line 261 of file Zoltan2_IntegerRangeList.hpp.

template<typename Integral >

void Zoltan2::printIntegralRangeList	(	std::ostream &	os,
		Teuchos::Array< Integral > &	irl
	)

A helper function that prints the meaning of an encoded integer range list.

Parameters

os	An output stream to which the list will be printed.
val	A value of an integer range list parameter after it has been validated.

Definition at line 281 of file Zoltan2_IntegerRangeList.hpp.

void Zoltan2::createAllParameters

(

Teuchos::ParameterList &

pList

)

Create a list of all Zoltan2 parameters and validators.

Parameters

pList

on return, pList is the parameter list that was created.

This is the validating parameter list that can be used to process the user's parameter list.

Definition at line 35 of file Zoltan2_Parameters.cpp.

static void Zoltan2::setValidatorsInList ( const Teuchos::ParameterList &  plSome, const Teuchos::ParameterList &  plAll, Teuchos::ParameterList &  plVal  )

static

Create a parameter list that can validate a specific list of parameters.

Parameters

plSome	the user's parameter list
plAll	a parameter list with all Zoltan2 parameters and their validators
plVal	on return is a parameter list with all of the user's parameters, but with validators.

Environment::commitParameters() calls validateParametersAndSetDefaults() on the user's parameter list rather than validateParameters() because we want the validators' validateAndModify() methods to be called rather than the validate() methods. But unfortunately, validateAndModify() in addition to modifying the parameter if necessary also sets it to a default if the parameter does not appear in the user's parameter list.

We want the user's parameters to be modified, but we do not want unset parameters to appear in the validated user's parameter list. To achieve this, we create a validating list that contains only the parameters that appear in the user's parameter list.

Definition at line 84 of file Zoltan2_Parameters.cpp.

void Zoltan2::createValidatorList	(	const Teuchos::ParameterList &	plIn,
		Teuchos::ParameterList &	plOut
	)

Create a list by adding validators to the users parameter list.

Parameters

plIn	the user's parameter list
plOut	a new parameter list which is the user's list with our validators added.

Definition at line 115 of file Zoltan2_Parameters.cpp.

void Zoltan2::printListDocumentation	(	const Teuchos::ParameterList &	pl,
		std::ostream &	os,
		std::string	listNames
	)

Definition at line 131 of file Zoltan2_Parameters.cpp.

template<typename User >

RCP<Tpetra::CrsMatrix<int, typename MeshAdapter<User>::lno_t, typename MeshAdapter<User>::gno_t, typename MeshAdapter<User>::node_t> > Zoltan2::get2ndAdjsMatFromAdjs	(	const Teuchos::RCP< const MeshAdapter< User > > &	ia,
		const RCP< const Comm< int > >	comm,
		Zoltan2::MeshEntityType	sourcetarget,
		Zoltan2::MeshEntityType	through
	)

Definition at line 28 of file Zoltan2_ModelHelpers.hpp.

template<typename User >

void Zoltan2::get2ndAdjsViewFromAdjs	(	const Teuchos::RCP< const MeshAdapter< User > > &	ia,
		const RCP< const Comm< int > >	comm,
		Zoltan2::MeshEntityType	sourcetarget,
		Zoltan2::MeshEntityType	through,
		Teuchos::ArrayRCP< typename MeshAdapter< User >::offset_t > &	offsets,
		Teuchos::ArrayRCP< typename MeshAdapter< User >::gno_t > &	adjacencyIds
	)

Definition at line 161 of file Zoltan2_ModelHelpers.hpp.

template<typename SolutionAdapter , typename DataAdapter >

size_t Zoltan2::getImportList	(	const PartitioningSolution< SolutionAdapter > &	solution,
		const DataAdapter *const	data,
		ArrayRCP< typename DataAdapter::gno_t > &	imports
	)

From a PartitioningSolution, get a list of IDs to be imported. Assumes part numbers in PartitioningSolution are to be used as rank numbers.

The list returned by solution.getPartListView() is an export list, detailing to which part each ID should move. This method provides a new list, listing the IDs to be imported to this processor.

Parameters

imports

on return is the list of global Ids assigned to this process under the Solution.

TODO: Add a version that takes a Mapping solution as well for k!=p.

Definition at line 44 of file Zoltan2_PartitioningHelpers.hpp.

void Zoltan2::AlltoAllCount	(	const Comm< int > &	comm,
		const Environment &	env,
		const ArrayView< const int > &	sendCount,
		const ArrayView< int > &	recvCount
	)

Each process sends a value to every process, an all-to-all.

Parameters

comm	The communicator for the process group involved
env	The environment, required for error messages
sendCount	The number to send to process p is in sendCount[p].
recvCount	On return, The number received from process p will be in recvCount[p].

Note: If Teuchos::Comm adds an AlltoAll method, we should use it instead of this function. TODO

Definition at line 36 of file Zoltan2_AlltoAll.cpp.

template<>

void Zoltan2::AlltoAllv	(	const Comm< int > &	comm,
		const Environment &	env,
		const ArrayView< const std::string > &	sendBuf,
		const ArrayView< const int > &	sendCount,
		ArrayRCP< std::string > &	recvBuf,
		const ArrayView< int > &	recvCount
	)

Definition at line 95 of file Zoltan2_AlltoAll.cpp.

template<>

void Zoltan2::AlltoAllv	(	const Comm< int > &	comm,
		const Environment &	env,
		const ArrayView< const unsigned short > &	sendBuf,
		const ArrayView< const int > &	sendCount,
		ArrayRCP< unsigned short > &	recvBuf,
		const ArrayView< int > &	recvCount
	)

Definition at line 213 of file Zoltan2_AlltoAll.cpp.

template<>

void Zoltan2::AlltoAllv	(	const Comm< int > &	comm,
		const Environment &	env,
		const ArrayView< const unsigned char > &	sendBuf,
		const ArrayView< const int > &	sendCount,
		ArrayRCP< unsigned char > &	recvBuf,
		const ArrayView< int > &	recvCount
	)

Definition at line 235 of file Zoltan2_AlltoAll.cpp.

template<typename T >

void Zoltan2::AlltoAllv	(	const Comm< int > &	comm,
		const Environment &	env,
		const ArrayView< const T > &	sendBuf,
		const ArrayView< const int > &	sendCount,
		ArrayRCP< T > &	recvBuf,
		const ArrayView< int > &	recvCount
	)

AlltoAllv sends/receives data to/from all processes.

Parameters

comm	The communicator for the process group involved
env	The environment, required for error messages
sendBuf	The data to be sent, in destination process rank order
sendCount	The number of Ts to send to process p is in sendCount[p].
recvBuf	On return, recvBuf has been allocated and contains the packets sent to this process by others.
recvCount	On return, The number of Ts received from process p will be in recvCount[p].

The data type T must be a type for which either Zoltan2::IdentifierTraits are defined or Teuchos::DirectSerializationTraits is defined.

AlltoAll uses only point-to-point messages. This is to avoid the MPI limitation of integer offsets and counters in collective operations. In other words, sendBuf.size() can exceed a value that fits into 32 bits. However each point to point message size must fit in an int.

It also avoids non-scalable MPI data structures that are associated with collective operations.

In addition it can be used for Zoltan2 global ID data types that are not serializable by Teuchos point-to-point messages.

Definition at line 57 of file Zoltan2_AlltoAll.hpp.

template<typename gno_t >

size_t Zoltan2::findUniqueGidsCommon	(	size_t	num_keys,
		int	num_gid,
		ZOLTAN_ID_PTR	ddkeys,
		char *	ddnewgids,
		MPI_Comm	mpicomm
	)

Definition at line 33 of file Zoltan2_findUniqueGids.hpp.

template<typename lno_t , typename gno_t >

size_t Zoltan2::findUniqueGids	(	Tpetra::MultiVector< gno_t, lno_t, gno_t > &	keys,
		Tpetra::Vector< gno_t, lno_t, gno_t > &	gids
	)

Definition at line 89 of file Zoltan2_findUniqueGids.hpp.

template<typename key_t , typename gno_t >

size_t Zoltan2::findUniqueGids	(	std::vector< key_t > &	keys,
		std::vector< gno_t > &	gids,
		const Teuchos::Comm< int > &	comm
	)

Definition at line 160 of file Zoltan2_findUniqueGids.hpp.

template<typename Adapter , typename MachineRep = MachineRepresentation<typename Adapter::scalar_t, typename Adapter::part_t>>

void Zoltan2::globalWeightedByPart	(	const RCP< const Environment > &	env,
		const RCP< const Comm< int > > &	comm,
		const RCP< const GraphModel< typename Adapter::base_adapter_t > > &	graph,
		const ArrayView< const typename Adapter::part_t > &	parts,
		typename Adapter::part_t &	numParts,
		ArrayRCP< RCP< BaseClassMetrics< typename Adapter::scalar_t > > > &	metrics,
		ArrayRCP< typename Adapter::scalar_t > &	globalSums,
		bool	bMessages = true,
		const RCP< const MachineRep >	machine = Teuchos::null
	)

Given the local partitioning, compute the global weighted cuts in each part.

Parameters

env	Environment for error handling
comm	communicator
graph	Graph model
part	part[i] is the part ID for local object i
numParts	on return this is the global number of parts.
metrics	on return points to a list of named GraphMetricValues cuts that each contains the global max and sum over parts of the item being measured. The list may contain "edge cuts", or "weight 0", "weight 1" and so on in that order. If uniform weights were given, then only "edge cuts" appears. If one set of non-uniform weights were given, then "weight 0" appear. Finally, if multiple weights were given, we have the individual weights "weight 0", "weight 1", and so on.
globalSums	If weights are uniform, the globalSums is the numParts totals of global number of cuts in each part. Suppose the number of weights is W. If W is 1, then on return this is an array of length numParts . The numParts entries are the total weight in each part. If W is greater than one, then the length of this array is W*numParts . The entries are the sum of the individual weights in each part, by weight index by part number. The array is allocated here.

globalWeightedByPart() must be called by all processes in comm.

Definition at line 58 of file Zoltan2_GraphMetricsUtility.hpp.

template<typename scalar_t , typename part_t >

void Zoltan2::printGraphMetricsHeader	(	std::ostream &	os,
		part_t	targetNumParts,
		part_t	numParts
	)

Print out header info for graph metrics.

Definition at line 496 of file Zoltan2_GraphMetricsUtility.hpp.

template<typename scalar_t , typename part_t >

void Zoltan2::printGraphMetrics	(	std::ostream &	os,
		part_t	targetNumParts,
		part_t	numParts,
		const ArrayView< RCP< BaseClassMetrics< scalar_t > > > &	infoList
	)

Print out list of graph metrics.

Definition at line 511 of file Zoltan2_GraphMetricsUtility.hpp.

template<typename scalar_t , typename part_t >

void Zoltan2::printGraphMetrics	(	std::ostream &	os,
		part_t	targetNumParts,
		part_t	numParts,
		RCP< BaseClassMetrics< scalar_t >>	metricValue
	)

Print out header and a single graph metric.

Definition at line 529 of file Zoltan2_GraphMetricsUtility.hpp.

template<typename scalar_t , typename lno_t , typename part_t >

void Zoltan2::globalSumsByPart	(	const RCP< const Environment > &	env,
		const RCP< const Comm< int > > &	comm,
		const ArrayView< const part_t > &	part,
		int	vwgtDim,
		const ArrayView< StridedData< lno_t, scalar_t > > &	vwgts,
		multiCriteriaNorm	mcNorm,
		part_t	targetNumParts,
		part_t &	numExistingParts,
		part_t &	numNonemptyParts,
		ArrayRCP< RCP< BaseClassMetrics< scalar_t > > > &	metrics,
		ArrayRCP< scalar_t > &	globalSums
	)

Given the local partitioning, compute the global sums in each part.

Parameters

env	Environment for error handling
comm	communicator
part	part[i] is the part ID for local object i
vwgts	vwgts[w] is the StridedData object representing weight index w. The number of weights (which must be at least one TODO WHY?) is taken to be vwgts.size().
mcNorm	the multiCriteria norm to be used if the number of weights is greater than one.
targetNumParts	input: number of requested parts
numExistingParts	on return this is the maximum part ID + 1.
numNonemptyParts	on return this is the number of those parts that are non-empty.
metrics	on return points to a list of named MetricValues objects that each contains the global min, max and avg over parts of the item being measured. The list may contain "object count", "normed weight", "weight 0", "weight 1" and so on in that order. If uniform weights were given, then only "object count" appears. If one set of non-uniform weights were given, then "object count" and "weight 0" appear. Finally, if multiple weights were given, we have "object count", then "normed weight", then the individual weights "weight 0", "weight 1", and so on.
globalSums	If weights are uniform, the globalSums is the numExistingParts totals of global number of objects in each part. Suppose the number of weights is W. If W is 1, then on return this is an array of length 2*numExistingParts . The first numExistingParts entries are the count of objects in each part and the second is the total weight in each part. If W is greater than one, then the length of this array is (2+W)*numExistingParts . The first numExistingParts entries are the count of objects in each part. The next numExistingParts entries are the sum of the normed weights in each part. The final entries are the sum of the individual weights in each part, by weight index by part number. The array is allocated here.

() must be called by all processes in comm. The imbalance metrics are not yet set in the MetricValues objects, because they require part size information.

Definition at line 65 of file Zoltan2_ImbalanceMetricsUtility.hpp.

template<typename Adapter >

void Zoltan2::imbalanceMetrics	(	const RCP< const Environment > &	env,
		const RCP< const Comm< int > > &	comm,
		multiCriteriaNorm	mcNorm,
		const Adapter *	ia,
		const PartitioningSolution< Adapter > *	solution,
		const ArrayView< const typename Adapter::part_t > &	partArray,
		const RCP< const GraphModel< typename Adapter::base_adapter_t > > &	graphModel,
		typename Adapter::part_t &	numExistingParts,
		typename Adapter::part_t &	numNonemptyParts,
		ArrayRCP< RCP< BaseClassMetrics< typename Adapter::scalar_t > > > &	metrics
	)

Compute imbalance metrics for a distribution.

Parameters

env	The problem environment.
comm	The problem communicator.
ia	the InputAdapter object which corresponds to the Solution.
solution	the PartitioningSolution to be evaluated.
mcNorm	is the multicriteria norm to use if the number of weights is greater than one. See the multiCriteriaNorm enumerator for mcNorm values.
graphModel	the graph model.
numExistingParts	on return is the max Part ID + 1.
numNonemptyParts	on return is the global number of parts to which objects are assigned.
metrics	on return points to a list of named MetricValues objects that each contains the global min, max and avg over parts and also imbalance measures of the item being measured. The list may contain "object count", "normed weight", "weight 0", "weight 1" and so on in that order. If uniform weights were given, then only "object count" appears. If one set of non-uniform weights were given, then "object count" and "weight 0" appear. Finally, if multiple weights were given, we have "object count", then "normed weight", then the individual weights "weight 0", "weight 1", and so on.

objectMetrics() must be called by all processes in comm. See the metricOffset enumerator in the MetricValues class for the interpretation of the metric quantities.

Todo:

check that part sizes sum to one if we're doing COMPLEX_ASSERTION

Definition at line 314 of file Zoltan2_ImbalanceMetricsUtility.hpp.

template<typename scalar_t , typename part_t >

void Zoltan2::printImbalanceMetricsHeader	(	std::ostream &	os,
		part_t	targetNumParts,
		part_t	numExistingParts,
		part_t	numNonemptyParts
	)

Print out header info for imbalance metrics.

Definition at line 506 of file Zoltan2_ImbalanceMetricsUtility.hpp.

template<typename scalar_t , typename part_t >

void Zoltan2::printImbalanceMetrics	(	std::ostream &	os,
		part_t	targetNumParts,
		part_t	numExistingParts,
		part_t	numNonemptyParts,
		const ArrayView< RCP< BaseClassMetrics< scalar_t > > > &	infoList
	)

Print out list of imbalance metrics.

Definition at line 526 of file Zoltan2_ImbalanceMetricsUtility.hpp.

template<typename scalar_t , typename part_t >

void Zoltan2::printImbalanceMetrics	(	std::ostream &	os,
		part_t	targetNumParts,
		part_t	numExistingParts,
		part_t	numNonemptyParts,
		RCP< BaseClassMetrics< scalar_t >>	metricValue
	)

Print out header and a single imbalance metric.

Definition at line 547 of file Zoltan2_ImbalanceMetricsUtility.hpp.

void Zoltan2::addNumberToFileName	(	int	number,
		std::string	fname,
		std::string &	newf
	)

Helper method to add number to a file name.

Parameters

number	the number (such as process rank) to add
fname	the file name to modify
newf	on return newf is fname with the rank added to the name.

If fname has no dot in it, then the rank is added to the end of the name. Otherwise the rank is added before the first dot in fname.

Definition at line 20 of file Zoltan2_IO.cpp.

template<typename metric_t , typename scalar_t >

RCP<metric_t> Zoltan2::addNewMetric	(	const RCP< const Environment > &	env,
		ArrayRCP< RCP< BaseClassMetrics< scalar_t > > > &	metrics
	)

Definition at line 29 of file Zoltan2_MetricUtility.hpp.

template<typename scalar_t >

void Zoltan2::getStridedStats	(	const ArrayView< scalar_t > &	v,
		int	stride,
		int	offset,
		scalar_t &	min,
		scalar_t &	max,
		scalar_t &	sum
	)

Find min, max and sum of metric values.

Parameters

v	a list of values
stride	the value such that v[offset + stride*i] will be included in the calculation for all possible i.
offset	the offset at which calculation will begin.
min	on return, min will hold the minimum of the values.
max	on return, max will hold the maximum of the values.
sum	on return, sum will hold the sum of the values.

Definition at line 54 of file Zoltan2_MetricUtility.hpp.

template<typename scalar_t >

void Zoltan2::getStridedStats	(	const ArrayView< scalar_t > &	v,
		int	stride,
		int	offset,
		scalar_t &	max,
		scalar_t &	sum
	)

Find max and sum of graph metric values.

Parameters

v	a list of values
stride	the value such that v[offset + stride*i] will be included in the calculation for all possible i.
offset	the offset at which calculation will begin.
max	on return, max will hold the maximum of the values.
sum	on return, sum will hold the sum of the values.

Definition at line 76 of file Zoltan2_MetricUtility.hpp.

template<typename scalar_t , typename lno_t , typename part_t >

void Zoltan2::normedPartWeights	(	const RCP< const Environment > &	env,
		part_t	numberOfParts,
		const ArrayView< const part_t > &	parts,
		const ArrayView< StridedData< lno_t, scalar_t > > &	vwgts,
		multiCriteriaNorm	mcNorm,
		scalar_t *	weights
	)

Compute the total weight in each part on this process.

Parameters

env	the Environment for error messages
numberOfParts	the number of Parts with respect to which weights should be computed.
parts	the part Id for each object, which may range from 0 to one less than numberOfParts
vwgts	vwgts[w] is the StridedData object representing weight index w. vwgts[w][i] is the w'th weight for object i.
mcNorm	the multiCriteria norm, to be used if the number of weights is greater than one.
weights	on return, weights[p] is the total weight for part p. weights is allocated by the caller

Todo:

• Zoltan_norm() in Zoltan may scale the weight. Do we ever need this?

< 1-norm = Manhattan norm

< 2-norm = sqrt of sum of squares

< inf-norm = maximum norm

Definition at line 107 of file Zoltan2_MetricUtility.hpp.

template<typename scalar_t , typename part_t >

void Zoltan2::computeImbalances	(	part_t	numExistingParts,
		part_t	targetNumParts,
		const scalar_t *	psizes,
		scalar_t	sumVals,
		const scalar_t *	vals,
		scalar_t &	min,
		scalar_t &	max,
		scalar_t &	avg
	)

Compute the imbalance.

Parameters

numExistingParts	the max Part ID + 1, which is the length of the vals array.
targetNumParts	the number of parts desired, which is the length of the psizes array if it is defined.
psizes	if part sizes are not uniform then psizes[p] is the part size for part p, for p ranging from zero to one less than targetNumParts. Part sizes must sum to one. If part sizes are uniform, psizes should be NULL.
sumVals	is the sum of the values in the vals list.
vals	vals[p] is the amount in part p, for p ranging from zero to one less than numExistingParts.
min	on return, min will be the minimum (best) imbalance of all the parts.
max	on return, max will be the maximum imbalance of all the parts.
avg	on return avg will be the average imbalance across the parts.

Imbalance is a value between zero and one. If target is the desired amount in part p and actual is the actual amount in part p, then the imbalance is:

abs(target - actual) / target

If the part is supposed to be empty (target is zero), then no imbalance is computed for that part. If actual for that part is non-zero, then other parts are too small and the imbalance will be found in those other parts.

Definition at line 204 of file Zoltan2_MetricUtility.hpp.

template<typename scalar_t , typename part_t >

void Zoltan2::computeImbalances	(	part_t	numExistingParts,
		part_t	targetNumParts,
		int	numSizes,
		ArrayView< ArrayRCP< scalar_t > >	psizes,
		scalar_t	sumVals,
		const scalar_t *	vals,
		scalar_t &	min,
		scalar_t &	max,
		scalar_t &	avg
	)

Compute the imbalance in the case of multiple part sizes.

Parameters

numExistingParts	the max Part ID + 1, which is the length of the vals array.
targetNumParts	the number of parts desired, which is the length of the psizes array if it is defined.
numSizes	the number of part size arrays
psizes	is an array of numSizes pointers to part size arrays. If the part sizes for index w are uniform, then psizes[w] should be NULL. Otherwise it should point to targetNumParts sizes, and the sizes for each index should sum to one.
sumVals	is the sum of the values in the vals list.
vals	vals[p] is the amount in part p, for p ranging from zero to one less than numExistingParts.
min	on return, min will be the minimum (best) imbalance of all the parts.
max	on return, max will be the maximum imbalance of all the parts.
avg	on return avg will be the average imbalance across the parts.

Imbalance is a value between zero and one. If target is the desired amount in part p and actual is the actual amount in part p, then the imbalance is:

abs(target - actual) / target

If the part is supposed to be empty (target is zero), then no imbalance is computed for that part. If actual for that part is non-zero, then other parts are too small and the imbalance will be found in those other parts.

Definition at line 302 of file Zoltan2_MetricUtility.hpp.

template<typename scalar_t >

scalar_t Zoltan2::normedWeight	(	ArrayView< scalar_t >	weights,
		multiCriteriaNorm	norm
	)

Compute the norm of the vector of weights.

< 1-norm = Manhattan norm

< 2-norm = sqrt of sum of squares

< inf-norm = maximum norm

Definition at line 420 of file Zoltan2_MetricUtility.hpp.

template<typename lno_t , typename scalar_t >

scalar_t Zoltan2::normedWeight	(	ArrayView< StridedData< lno_t, scalar_t > >	weights,
		lno_t	idx,
		multiCriteriaNorm	norm
	)

Compute the norm of the vector of weights stored as StridedData.

Definition at line 469 of file Zoltan2_MetricUtility.hpp.

template<typename scalar_t , typename T >

static void Zoltan2::getInputArrayHelper ( ArrayRCP< const T > &  target, const ArrayRCP< const scalar_t > &  src  )

static

Definition at line 133 of file Zoltan2_StridedData.hpp.

template<typename scalar_t >

static void Zoltan2::getInputArrayHelper ( ArrayRCP< const scalar_t > &  target, const ArrayRCP< const scalar_t > &  src  )

static

Definition at line 156 of file Zoltan2_StridedData.hpp.

long Zoltan2::getProcessKilobytes

(

)

Definition at line 34 of file Zoltan2_Util.cpp.

template<typename scalar_t >

bool Zoltan2::outsideRegion ( scalar_t  val, scalar_t  mark, double  epsilon  )

inline

Definition at line 28 of file Zoltan2_Util.hpp.

static void Zoltan2::AssertCondition ( bool  condition, const std::string &  message, const char *  file = __FILE__, int  line = __LINE__  )

inlinestatic

Definition at line 33 of file Zoltan2_Util.hpp.

std::string Zoltan2::Zoltan2_Version

(

)

Definition at line 24 of file Zoltan2_Version.cpp.

static void Zoltan2::getSphynxValidParameters ( ParameterList &  pl )

static

Set up validators specific to this algorithm.

Definition at line 35 of file Zoltan2_SphynxProblem.hpp.

static void Zoltan2::setSphynxValidatorsInList ( const Teuchos::ParameterList &  plSome, const Teuchos::ParameterList &  plAll, Teuchos::ParameterList &  plVal  )

static

Definition at line 72 of file Zoltan2_SphynxProblem.hpp.

std::string Zoltan2::Sphynx_Version

(

)

Definition at line 15 of file Zoltan2_SphynxVersion.cpp.

template<typename key_t , typename gno_t >

size_t Zoltan2::findUniqueGids	(	const std::vector< key_t > &	keys,
		std::vector< gno_t > &	gids,
		Teuchos::RCP< const Teuchos::Comm< int > > &	comm
	)

Definition at line 30 of file directoryTest_findUniqueGids.cpp.

Zoltan2::DECL_TYPE_NAME

(

int

)

Zoltan2::DECL_TYPE_NAME

(

long

long

)

void Zoltan2::checkNUnique	(	std::string &	name,
		size_t	nUniqueGids,
		size_t	nExpected
	)

Definition at line 94 of file directoryTest_findUniqueGids.cpp.

template<typename gno_t >

void Zoltan2::checkMaxGid	(	std::string &	name,
		std::vector< gno_t > &	gids,
		gno_t	maxExpected,
		Teuchos::RCP< const Teuchos::Comm< int > > &	comm
	)

Definition at line 104 of file directoryTest_findUniqueGids.cpp.

template<typename gno_t >

void Zoltan2::checkMinGid	(	std::string &	name,
		std::vector< gno_t > &	gids,
		gno_t	minExpected,
		Teuchos::RCP< const Teuchos::Comm< int > > &	comm
	)

Definition at line 126 of file directoryTest_findUniqueGids.cpp.

template<typename gno_t >

void Zoltan2::checkNLocallyUnique	(	std::string &	name,
		std::vector< gno_t > &	gids,
		size_t	nExpected
	)

Definition at line 148 of file directoryTest_findUniqueGids.cpp.

template<typename gno_t >

void Zoltan2::test1

(

Teuchos::RCP< const Teuchos::Comm< int > > &

comm

)

Definition at line 175 of file directoryTest_findUniqueGids.cpp.

template<typename gno_t >

void Zoltan2::test2

(

Teuchos::RCP< const Teuchos::Comm< int > > &

comm

)

Definition at line 220 of file directoryTest_findUniqueGids.cpp.

template<typename gno_t >

void Zoltan2::test3

(

Teuchos::RCP< const Teuchos::Comm< int > > &

comm

)

Definition at line 272 of file directoryTest_findUniqueGids.cpp.

template<typename gno_t >

void Zoltan2::test4

(

Teuchos::RCP< const Teuchos::Comm< int > > &

comm

)

Definition at line 336 of file directoryTest_findUniqueGids.cpp.

void Zoltan2::print_proc_safe	(	const std::string &	message,
		Teuchos::RCP< const Teuchos::Comm< int > >	comm,
		bool	only_rank0 = true
	)

Definition at line 48 of file directoryTest_Impl.hpp.

int Zoltan2::runDirectoryTests	(	int	narg,
		char **	arg
	)

Definition at line 1355 of file directoryTest_Impl.hpp.

Variable Documentation

const std::string Zoltan2::fail = "FAIL "

static

Definition at line 90 of file directoryTest_findUniqueGids.cpp.

const std::string Zoltan2::pass = " "

static

Definition at line 91 of file directoryTest_findUniqueGids.cpp.