Scoring functions used by MRMScoring. More...

Classes
struct	pair_hash
	Simple hash function for Scoring::pos2D. More...

struct	XCorrArrayType

Typedefs
Type defs and helper structures
typedef std::pair< unsigned int, unsigned int >	pos2D

typedef std::pair< int, double >	XCorrEntry
	Cross Correlation array contains (lag,correlation) pairs.

Functions
Helper functions
OPENSWATHALGO_DLLAPI double	NormalizedManhattanDist (std::vector< double > &x, std::vector< double > &y)
	Calculate the normalized Manhattan distance between two vectors.

OPENSWATHALGO_DLLAPI double	RootMeanSquareDeviation (const std::vector< double > &x, const std::vector< double > &y)
	Calculate the RMSD (root mean square deviation) between two vectors.

OPENSWATHALGO_DLLAPI double	SpectralAngle (const std::vector< double > &x, const std::vector< double > &y)
	Calculate the spectral angle (acosine of the normalized dotproduct) between two vectors.

OPENSWATHALGO_DLLAPI XCorrArrayType	normalizedCrossCorrelation (std::vector< double > &data1, std::vector< double > &data2, const int maxdelay, const int lag)

OPENSWATHALGO_DLLAPI XCorrArrayType	normalizedCrossCorrelationPost (std::vector< double > &normalized_data1, std::vector< double > &normalized_data2, const int maxdelay, const int lag)
	Calculate crosscorrelation on std::vector data that is already normalized.

OPENSWATHALGO_DLLAPI XCorrArrayType	calculateCrossCorrelation (const std::vector< double > &data1, const std::vector< double > &data2, const int maxdelay, const int lag)
	Calculate crosscorrelation on std::vector data without normalization.

OPENSWATHALGO_DLLAPI XCorrArrayType::const_iterator	xcorrArrayGetMaxPeak (const XCorrArrayType &array)
	Find best peak in an cross-correlation (highest apex)

OPENSWATHALGO_DLLAPI void	standardize_data (std::vector< double > &data)
	Standardize a vector (subtract mean, divide by standard deviation)

OPENSWATHALGO_DLLAPI void	normalize_sum (std::vector< double > &x)
	Divide each element of x by the sum of the vector.

OPENSWATHALGO_DLLAPI unsigned int	computeAndAppendRank (const std::vector< double > &v, std::vector< unsigned int > &ranks)

OPENSWATHALGO_DLLAPI std::vector< unsigned int >	computeRankVector (const std::vector< std::vector< double > > &intensity, std::vector< std::vector< unsigned int > > &ranks)

OPENSWATHALGO_DLLAPI double	rankedMutualInformation (std::vector< unsigned int > &ranked_data1, std::vector< unsigned int > &ranked_data2, const unsigned int max_rank1, const unsigned int max_rank2)

Detailed Description

Scoring functions used by MRMScoring.

Many helper functions to calculate cross-correlations between data

Typedef Documentation

◆ pos2D

typedef std::pair<unsigned int, unsigned int> pos2D

◆ XCorrEntry

typedef std::pair<int, double> XCorrEntry

Cross Correlation array contains (lag,correlation) pairs.

Function Documentation

◆ calculateCrossCorrelation()

OPENSWATHALGO_DLLAPI XCorrArrayType calculateCrossCorrelation	(	const std::vector< double > &	data1,
		const std::vector< double > &	data2,
		const int	maxdelay,
		const int	lag
	)

Calculate crosscorrelation on std::vector data without normalization.

◆ computeAndAppendRank()

OPENSWATHALGO_DLLAPI unsigned int computeAndAppendRank	(	const std::vector< double > &	v,
		std::vector< unsigned int > &	ranks
	)

Referenced by MRMTransitionGroupPicker::pickFragmentChromatograms().

◆ computeRankVector()

OPENSWATHALGO_DLLAPI std::vector< unsigned int > computeRankVector	(	const std::vector< std::vector< double > > &	intensity,
		std::vector< std::vector< unsigned int > > &	ranks
	)

◆ normalize_sum()

OPENSWATHALGO_DLLAPI void normalize_sum ( std::vector< double > & x )

Divide each element of x by the sum of the vector.

Parameters

[in,out] x Vector to normalize in-place; unchanged if sum is zero

◆ normalizedCrossCorrelation()

OPENSWATHALGO_DLLAPI XCorrArrayType normalizedCrossCorrelation	(	std::vector< double > &	data1,
		std::vector< double > &	data2,
		const int	maxdelay,
		const int	lag
	)

Calculate crosscorrelation on std::vector data (which is first normalized) NOTE: this replaces calcxcorr

Referenced by MRMTransitionGroupPicker::computeQuality_().

◆ normalizedCrossCorrelationPost()

OPENSWATHALGO_DLLAPI XCorrArrayType normalizedCrossCorrelationPost	(	std::vector< double > &	normalized_data1,
		std::vector< double > &	normalized_data2,
		const int	maxdelay,
		const int	lag
	)

Calculate crosscorrelation on std::vector data that is already normalized.

◆ NormalizedManhattanDist()

OPENSWATHALGO_DLLAPI double NormalizedManhattanDist	(	std::vector< double > &	x,
		std::vector< double > &	y
	)

Calculate the normalized Manhattan distance between two vectors.

Equivalent to the function "delta_ratio_sum" from mQuest to calculate similarity between library intensity and experimental ones.

The delta_ratio_sum is calculated as follows:

\[ d = \sqrt{\frac{1}{N} \sum_{i=0}^N |\frac{x_i}{\mu_x} - \frac{y_i}{\mu_y}|) } \]

Parameters

[in,out]	x	First intensity vector; normalized in-place via normalize_sum
[in,out]	y	Second intensity vector; normalized in-place via normalize_sum

◆ rankedMutualInformation()

OPENSWATHALGO_DLLAPI double rankedMutualInformation	(	std::vector< unsigned int > &	ranked_data1,
		std::vector< unsigned int > &	ranked_data2,
		const unsigned int	max_rank1,
		const unsigned int	max_rank2
	)

Referenced by MRMTransitionGroupPicker::pickFragmentChromatograms().

◆ RootMeanSquareDeviation()

OPENSWATHALGO_DLLAPI double RootMeanSquareDeviation	(	const std::vector< double > &	x,
		const std::vector< double > &	y
	)

Calculate the RMSD (root mean square deviation) between two vectors.

The RMSD is calculated as follows:

\[ RMSD = \sqrt{\frac{1}{N} \sum_{i=0}^N (x_i - y_i)^2 } \]

Parameters

[in]	x	First data vector
[in]	y	Second data vector (must have same size as x)

◆ SpectralAngle()

OPENSWATHALGO_DLLAPI double SpectralAngle	(	const std::vector< double > &	x,
		const std::vector< double > &	y
	)

Calculate the spectral angle (acosine of the normalized dotproduct) between two vectors.

The spectral angle is calculated as follows:

\[ \theta = acos \left( \frac{\sum_{i=0}^N (x_i * y_i))}{\sqrt{\sum_{i=0}^N (x_i * x_i) \sum_{i=0}^N (y_i * y_i)} } \right) \]

Parameters

[in]	x	First intensity vector
[in]	y	Second intensity vector (must have same size as x)

◆ standardize_data()

OPENSWATHALGO_DLLAPI void standardize_data ( std::vector< double > & data )

Standardize a vector (subtract mean, divide by standard deviation)

◆ xcorrArrayGetMaxPeak()

OPENSWATHALGO_DLLAPI XCorrArrayType::const_iterator xcorrArrayGetMaxPeak ( const XCorrArrayType & array )

Find best peak in an cross-correlation (highest apex)

Referenced by MRMTransitionGroupPicker::computeQuality_().

Classes

Typedefs

Functions

Detailed Description

Typedef Documentation

◆ pos2D

◆ XCorrEntry

Function Documentation

◆ calculateCrossCorrelation()

◆ computeAndAppendRank()

◆ computeRankVector()

◆ normalize_sum()

◆ normalizedCrossCorrelation()

◆ normalizedCrossCorrelationPost()

◆ NormalizedManhattanDist()

◆ rankedMutualInformation()

◆ RootMeanSquareDeviation()

◆ SpectralAngle()

◆ standardize_data()

◆ xcorrArrayGetMaxPeak()