TAlgCTBuilder Class Reference

Base class for cluster tree construction algorithms based on graph partitioning with graph defined by a sparse matrix.

#include <TAlgCTBuilder.hh>

Inheritance diagram for TAlgCTBuilder:

Public Member Functions
	TAlgCTBuilder (TAlgPartStrat *part_strat, const uint n_min=CFG::Cluster::nmin, const uint min_leaf_lvl=0)
	construct cluster tree builder with given partition strategy and tree parameters
virtual	~TAlgCTBuilder ()
	dtor
void	set_high_deg_fac (const uint fac)
	activate/deactivate (if fac = 0) high degree node separation
void	set_use_edge_weights (const bool use_edge_weights, const bool sym_edge_weights)
virtual TClusterTree *	build (const TSparseMatrix *S, const idx_t idx_ofs=0) const
	build cluster tree using connectivity in sparse matrix S
virtual TCluster *	divide (const TGraph &graph, const uint lvl, TPermutation &perm, const idx_t idx_ofs, const uint n_min, const TSparseMatrix *S, const uint max_lvl) const
	divide graph graph and build corresponding cluster tree
virtual void	partition (const TGraph &graph, TNodeSet &left, TNodeSet &right) const
	compute graph bi-partitioning of graph and store result in left and right

Protected Member Functions
virtual void	scc_partition (const TGraph &graph, TNodeSet &left, TNodeSet &right) const
	same as
virtual TCluster *	build_leaf (const TGraph &graph, const idx_t idx_ofs, TPermutation &perm) const
	build leaf node for indices in graph
virtual uint	adjust_n_min (const TSparseMatrix *S) const
	adjust n_min based on sparse matrix if default value of 0 was given in constructor
virtual void	check_flow (const TGraph &graph, TNodeSet &left, TNodeSet &right, const TSparseMatrix *S) const
	analyze connections between sub graphs left and right and swap if necessary

Member Function Documentation

virtual void scc_partition ( const TGraph &  graph, TNodeSet &  left, TNodeSet &  right  ) const

protectedvirtual

See Also

partition, but first check graph for connected components

void set_use_edge_weights	(	const bool	use_edge_weights,
		const bool	sym_edge_weights
	)

activate/deactivate the use of edge weights for the graph bi-partitioning and whether they should be handled symmetric