Pairwise point-set matching using Wasserstein optimal transport. More...

#include <OpenMS/ANALYSIS/MAPMATCHING/WNetMatcher.h>

Classes
struct	MatchResult
	Result of a pairwise matching. More...

Public Types
enum class	DistanceMetric { L1 , L2 , LINF }
	Distance metric for comparing point positions. More...

Static Public Member Functions
static MatchResult	match (const std::vector< std::array< double, 2 > > &positions_a, const std::vector< double > &intensities_a, const std::vector< std::array< double, 2 > > &positions_b, const std::vector< double > &intensities_b, DistanceMetric metric=DistanceMetric::LINF, double max_distance=100.0, double trash_cost=100.0)
	Match two sets of 2D points using optimal transport.

static DistanceMetric	metricFromString (const std::string &s)
	Convert a string to a DistanceMetric enum value.

Detailed Description

Pairwise point-set matching using Wasserstein optimal transport.

Matches two sets of 2D points (with associated intensities) by solving a minimum-cost network flow problem. Returns 1-to-1 matched index pairs selected greedily by descending transport flow.

This provides a minimal, FeatureMap-independent interface to the WNetAlign algorithm. For feature-level grouping across multiple maps, use FeatureGroupingAlgorithmWNet instead.

Class Documentation

◆ OpenMS::WNetMatcher::MatchResult

struct OpenMS::WNetMatcher::MatchResult

Result of a pairwise matching.

Collaboration diagram for WNetMatcher::MatchResult:

[legend]

Class Members
double	cost = 0.0	Total transport cost.
vector< pair< Size, Size > >	matched_pairs	(index_a, index_b) of matched points

Member Enumeration Documentation

◆ DistanceMetric

enum class DistanceMetric

strong

Distance metric for comparing point positions.

Enumerator
L1
L2
LINF

Member Function Documentation

◆ match()

static MatchResult match	(	const std::vector< std::array< double, 2 > > &	positions_a,
		const std::vector< double > &	intensities_a,
		const std::vector< std::array< double, 2 > > &	positions_b,
		const std::vector< double > &	intensities_b,
		DistanceMetric	metric = `DistanceMetric::LINF`,
		double	max_distance = `100.0`,
		double	trash_cost = `100.0`
	)

static

Match two sets of 2D points using optimal transport.

Points are matched by minimizing the total transport cost between their positions, weighted by intensity. Unmatched points are sent to a "trash" node at the specified cost.

Parameters

[in]	positions_a	Positions of first point set (each {dim0, dim1})
[in]	intensities_a	Intensities/weights of first point set
[in]	positions_b	Positions of second point set (each {dim0, dim1})
[in]	intensities_b	Intensities/weights of second point set
[in]	metric	Distance metric (L1, L2, or LINF)
[in]	max_distance	Maximum distance to consider a match
[in]	trash_cost	Cost of leaving a point unmatched

Returns: MatchResult with paired indices and total cost

◆ metricFromString()

static DistanceMetric metricFromString ( const std::string & s )

static

Convert a string to a DistanceMetric enum value.

Recognizes "L1", "L2", and "LINF" (maps to DistanceMetric::L1/L2/LINF respectively). Any other input emits a warning via OPENMS_LOG_WARN and returns DistanceMetric::LINF as the fallback.

Classes

Public Types

Static Public Member Functions