pm::IncidenceMatrix< symmetric > Class Template Reference

0/1 incidence matrix. More...

Inherits pm::IncidenceMatrix_base< symmetric >, and pm::GenericIncidenceMatrix< pm::IncidenceMatrix< symmetric > >.

List of all members.

Public Member Functions
	IncidenceMatrix ()
	Create an empty IncidenceMatrix.
	IncidenceMatrix (int r, int c)
	Create an empty IncidenceMatrix with r rows and c columns initialized with zeroes.
template<typename Iterator>
	IncidenceMatrix (int r, int c, Iterator src)
	Create an IncidenceMatrix IncidenceMatrix with r rows and c columns and initialize it from a data sequence.
template<typename Container>
	IncidenceMatrix (const Container &src, typename enable_if< nothing *, isomorphic_to_container_of< Container, Set< int > >::value &&!symmetric::value >::type=0)
	number of columns not known in advance
template<typename Container>
	IncidenceMatrix (const Container &src, typename enable_if< int, isomorphic_to_container_of< Container, Set< int > >::value &&!symmetric::value >::type c)
	number of columns given explicitly
IncidenceMatrix &	operator= (const IncidenceMatrix &M)
	Assignement operator.
void	swap (IncidenceMatrix &M)
	Swap the contents with that of another matrix in an efficient way.
void	resize (int m, int n)
	Extend or truncate to new dimensions (m rows, n columns). Surviving elements keep their values, new elements are implicitly false.
void	clear ()
	Clear contents.
void	clear (int r, int c)
	Clear contents.
reference	operator() (int i, int j)
	Entry at row i column j.
const_reference	operator() (int i, int j) const
	Entry at row i column j (const).
bool	exists (int i, int j) const
	Returns the entry at position (i,j).
void	squeeze ()
	Delete empty rows and columns, renumber the rest and reduce the dimensions.
void	squeeze_rows ()
	Delete empty rows, renumber the rest and reduce the dimensions.
void	squeeze_cols ()
	Delete empty columns, renumber the rest and reduce the dimensions.
template<typename Iterator>
void	permute_rows (Iterator perm)
	Permute the rows according to the given permutation.
template<typename Iterator>
void	permute_cols (Iterator perm)
	Permute the columns according to the given permutation.
template<typename Iterator>
void	permute_inv_rows (Iterator inv_perm)
	Permute the rows according to the inverse of the given permutation.
template<typename Iterator>
void	permute_inv_cols (Iterator inv_perm)
	Permute the columns according to the inverse of the given permutation.
Protected Member Functions
template<typename Iterator>
void	_init (Iterator src, True)
	initialize from a dense boolean sequence in row order
template<typename Iterator>
void	_init (Iterator src, False)
	initialize rowwise from a sequence of sets

---

Detailed Description

template<typename symmetric>
class pm::IncidenceMatrix< symmetric >

0/1 incidence matrix.

The only persistent class from the incidence matrix family. The implementation is based on a two-dimensional grid of balanced binary search (AVL) trees, the same as for

See also:

SparseMatrix. The whole internal data structure is attached to a smart pointer with

{reference counting}.

A symmetric incidence matrix is a square matrix whose elements `(i,j)` and `(j,i)` are always equal. Internally it is stored in a triangular form, avoiding redundant elements, but appears as a full square.

---

Constructor & Destructor Documentation

template<typename symmetric>

template<typename Iterator>

pm::IncidenceMatrix< symmetric >::IncidenceMatrix	(	int	r,
		int	c,
		Iterator	src
	)			[inline]

Create an IncidenceMatrix IncidenceMatrix with r rows and c columns and initialize it from a data sequence.

src should iterate either over r c boolean values, corresponding to the elements in the row order (the column index changes first,) or over r sets with integer elements (or convertible to integer), which are assigned to the matrix rows.

In the symmetric case the redundant elements must be present in the input sequence; their values are ignored.

Parameters:

	r	the number of rows
	c	the number of columns
	src	an iterator

---

Member Function Documentation

template<typename symmetric>

void pm::IncidenceMatrix< symmetric >::resize	(	int	m,
		int	n
	)			[inline]

Extend or truncate to new dimensions (m rows, n columns). Surviving elements keep their values, new elements are implicitly false.

IncidenceMatrix deploys an adaptive reallocation strategy similar to std::vector, reserving additional stock memory by every reallocation. If you repeatedly increase the matrix dimensions by one, the amortized reallocation costs will be proportional to the logarithm of the final dimension.

A special case, looking at the first glance like a "no operation": { M.resize(M.rows(), M.cols()) }, gets rid of this extra allocated storage.

---

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

• include/core/polymake/IncidenceMatrix.h