Sparse vector algorithms

Sparse vector algorithms. More...

Collaboration diagram for Sparse vector algorithms:

Data Structures
class	bm::sparse_vector_scanner< SV, S_FACTOR >
	algorithms for sparse_vector scan/search More...
class	bm::set2set_11_transform< SV >
	Integer set to set transformation (functional image in groups theory) https://en.wikipedia.org/wiki/Image_(mathematics). More...

Functions
template<typename SV>
void	bm::dynamic_range_clip_high (SV &svect, unsigned high_bit)
	Clip dynamic range for signal higher than specified.
template<typename SV>
void	bm::dynamic_range_clip_low (SV &svect, unsigned low_bit)
	Clip dynamic range for signal lower than specified (boost).
template<typename SV>
bool	bm::sparse_vector_find_first_mismatch (const SV &sv1, const SV &sv2, typename SV::size_type &midx, bm::null_support null_proc=bm::use_null)
	Find first mismatch (element which is different) between two sparse vectors (uses linear scan in bit-vector planes).
template<typename SV1, typename SV2>
void	bm::sparse_vector_find_mismatch (typename SV1::bvector_type &bv, const SV1 &sv1, const SV2 &sv2, bm::null_support null_proc)
	Find mismatch vector, indicating positions of mismatch between two sparse vectors (uses linear scan in bit-vector planes).

Detailed Description

Sparse vector algorithms.

Function Documentation

dynamic_range_clip_high()

template<typename SV>

void bm::dynamic_range_clip_high	(	SV &	svect,
		unsigned	high_bit )

Clip dynamic range for signal higher than specified.

Parameters

svect	- sparse vector to do clipping
high_bit	- max bit (inclusive) allowed for this signal vector

See also

dynamic_range_clip_low

Definition at line 67 of file bmsparsevec_algo.h.

References BM_ASSERT.

dynamic_range_clip_low()

template<typename SV>

void bm::dynamic_range_clip_low	(	SV &	svect,
		unsigned	low_bit )

Clip dynamic range for signal lower than specified (boost).

Parameters

svect	- sparse vector to do clipping
low_bit	- low bit (inclusive) allowed for this signal vector

See also

dynamic_range_clip_high

Definition at line 107 of file bmsparsevec_algo.h.

References BM_ASSERT.

sparse_vector_find_first_mismatch()

template<typename SV>

bool bm::sparse_vector_find_first_mismatch	(	const SV &	sv1,
		const SV &	sv2,
		typename SV::size_type &	midx,
		bm::null_support	null_proc = bm::use_null )

Find first mismatch (element which is different) between two sparse vectors (uses linear scan in bit-vector planes).

Function works with both NULL and NOT NULL vectors NULL means unassigned (uncertainty), so first mismatch NULL is a mismatch even if values in vectors can be formally the same (i.e. 0)

Parameters

sv1	- vector 1
sv2	- vector 2
midx	- mismatch index
null_proc	- defines if we want to include (not) NULL vector into comparison (bm::use_null) or not. By default search takes NULL vector into account

Returns

true if mismatch found

See also

sparse_vector_find_mismatch

Examples

strsvsample06.cpp, svsample09.cpp, xsample06.cpp, and xsample07.cpp.

Definition at line 170 of file bmsparsevec_algo.h.

References BM_ASSERT, id_max, bm::conditional< b >::test(), and use_null.

Referenced by compare_sv(), main(), main(), and test_mismatch_search().

sparse_vector_find_mismatch()

template<typename SV1, typename SV2>

void bm::sparse_vector_find_mismatch	(	typename SV1::bvector_type &	bv,
		const SV1 &	sv1,
		const SV2 &	sv2,
		bm::null_support	null_proc )

Find mismatch vector, indicating positions of mismatch between two sparse vectors (uses linear scan in bit-vector planes).

Function works with both NULL and NOT NULL vectors

Parameters

bv	- [out] - bit-ector with mismatch positions indicated as 1s
sv1	- vector 1
sv2	- vector 2
null_proc	- rules of processing for (not) NULL plane bm::no_null - NULLs from both vectors are treated as uncertainty and NOT included into final result bm::use_null - difference in NULLs accounted into the result

See also

sparse_vector_find_first_mismatch

Definition at line 348 of file bmsparsevec_algo.h.

References bm::bvector< Alloc >::bit_or(), bm::bvector< Alloc >::bit_xor(), BM_ASSERT, bm::bvector< Alloc >::invert(), no_null, bm::bvector< Alloc >::opt_none, bm::bvector< Alloc >::resize(), bm::conditional< b >::test(), use_null, and xor_swap().