Implementation of different sampling strategies.

Definition at line 32 of file RSampler.hxx.

Public Member Functions
	RSampler (const RFlat2DMatrix &dataset, std::vector< std::vector< std::size_t > > &rows, const std::string &sampleType, float sampleRatio, bool replacement=false, bool shuffle=true, std::size_t setSeed=0)

	~RSampler ()

void	GatherSamples (RFlat2DMatrix &ShuffledTensor, const std::vector< std::size_t > &ResampledRows, std::size_t numResampled, const std::vector< std::size_t > &OtherRows)
	Fill the sampled entries: the resampled entries followed by all the other ones, in shuffled order.

std::size_t	GetNumEntries ()

void	RandomOversampler (RFlat2DMatrix &ShuffledTensor)
	Oversample entries randomly from the minority dataset.

void	RandomUndersampler (RFlat2DMatrix &ShuffledTensor)
	Undersample entries randomly from the majority dataset.

void	Sampler (RFlat2DMatrix &SampledTensor)
	Collection of sampling types.

void	SampleWithoutReplacement (std::size_t n_samples, std::size_t max)
	Add indices without replacement to fSamples.

void	SampleWithReplacement (std::size_t n_samples, std::size_t max)
	Add indices with replacement to fSamples.

void	SetupRandomOversampler ()
	Calculate fNumEntries and major/minor variables for the random oversampler.

void	SetupRandomUndersampler ()
	Calculate fNumEntries and major/minor variables for the random undersampler.

void	SetupSampler ()
	Calculate fNumEntries and major/minor variables.

Private Attributes
const RFlat2DMatrix &	fDataset

std::size_t	fMajor

std::size_t	fMinor

std::size_t	fNumEntries

std::size_t	fNumMajor

std::size_t	fNumMinor

std::size_t	fNumResampledMajor

std::size_t	fNumResampledMinor

bool	fReplacement

std::vector< std::vector< std::size_t > > &	fRows

float	fSampleRatio

std::vector< std::size_t >	fSamples

std::string	fSampleType

std::size_t	fSetSeed

bool	fShuffle

std::unique_ptr< RFlat2DMatrixOperators >	fTensorOperators

#include </github/home/ROOT-CI/src/tree/ml/src/RSampler.hxx>

Constructor & Destructor Documentation

◆ RSampler()

ROOT::Experimental::Internal::ML::RSampler::RSampler	(	const RFlat2DMatrix &	dataset,
		std::vector< std::vector< std::size_t > > &	rows,
		const std::string &	sampleType,
		float	sampleRatio,
		bool	replacement = false,
		bool	shuffle = true,
		std::size_t	setSeed = 0 )

Definition at line 15 of file RSampler.cxx.

◆ ~RSampler()

ROOT::Experimental::Internal::ML::RSampler::~RSampler ( )

default

Member Function Documentation

◆ GatherSamples()

void ROOT::Experimental::Internal::ML::RSampler::GatherSamples	(	RFlat2DMatrix &	ShuffledTensor,
		const std::vector< std::size_t > &	ResampledRows,
		std::size_t	numResampled,
		const std::vector< std::size_t > &	OtherRows )

Fill the sampled entries: the resampled entries followed by all the other ones, in shuffled order.

Parameters

[in]	ShuffledTensor	Tensor with all the sampled entries
[in]	ResampledRows	Rows of the class the entries in fSamples are taken from
[in]	numResampled	Number of resampled entries
[in]	OtherRows	Rows of the class whose entries are all taken

Definition at line 131 of file RSampler.cxx.

◆ GetNumEntries()

std::size_t ROOT::Experimental::Internal::ML::RSampler::GetNumEntries ( )

inline

Definition at line 80 of file RSampler.hxx.

◆ RandomOversampler()

void ROOT::Experimental::Internal::ML::RSampler::RandomOversampler ( RFlat2DMatrix & ShuffledTensor )

Oversample entries randomly from the minority dataset.

Parameters

[in] SampledTensor Tensor with all the sampled entries

Definition at line 118 of file RSampler.cxx.

◆ RandomUndersampler()

void ROOT::Experimental::Internal::ML::RSampler::RandomUndersampler ( RFlat2DMatrix & ShuffledTensor )

Undersample entries randomly from the majority dataset.

Parameters

[in] SampledTensor Tensor with all the sampled entries

Definition at line 102 of file RSampler.cxx.

◆ Sampler()

void ROOT::Experimental::Internal::ML::RSampler::Sampler ( RFlat2DMatrix & SampledTensor )

Collection of sampling types.

Parameters

[in] SampledTensor Tensor with all the sampled entries

Definition at line 46 of file RSampler.cxx.

◆ SampleWithoutReplacement()

void ROOT::Experimental::Internal::ML::RSampler::SampleWithoutReplacement	(	std::size_t	n_samples,
		std::size_t	max )

Add indices without replacement to fSamples.

Parameters

[in]	n_samples	Number of indices to sample
[in]	max	Max index of the sample distribution

Definition at line 184 of file RSampler.cxx.

◆ SampleWithReplacement()

void ROOT::Experimental::Internal::ML::RSampler::SampleWithReplacement	(	std::size_t	n_samples,
		std::size_t	max )

Add indices with replacement to fSamples.

Parameters

[in]	n_samples	Number of indices to sample
[in]	max	Max index of the sample distribution

Definition at line 151 of file RSampler.cxx.

◆ SetupRandomOversampler()

void ROOT::Experimental::Internal::ML::RSampler::SetupRandomOversampler ( )

Calculate fNumEntries and major/minor variables for the random oversampler.

Definition at line 83 of file RSampler.cxx.

◆ SetupRandomUndersampler()

void ROOT::Experimental::Internal::ML::RSampler::SetupRandomUndersampler ( )

Calculate fNumEntries and major/minor variables for the random undersampler.

Definition at line 57 of file RSampler.cxx.

◆ SetupSampler()

void ROOT::Experimental::Internal::ML::RSampler::SetupSampler ( )

Calculate fNumEntries and major/minor variables.

Definition at line 34 of file RSampler.cxx.

Member Data Documentation

◆ fDataset

const RFlat2DMatrix& ROOT::Experimental::Internal::ML::RSampler::fDataset

private

Definition at line 34 of file RSampler.hxx.

◆ fMajor

std::size_t ROOT::Experimental::Internal::ML::RSampler::fMajor

private

Definition at line 44 of file RSampler.hxx.

◆ fMinor

std::size_t ROOT::Experimental::Internal::ML::RSampler::fMinor

private

Definition at line 45 of file RSampler.hxx.

◆ fNumEntries

std::size_t ROOT::Experimental::Internal::ML::RSampler::fNumEntries

private

Definition at line 42 of file RSampler.hxx.

◆ fNumMajor

std::size_t ROOT::Experimental::Internal::ML::RSampler::fNumMajor

private

Definition at line 46 of file RSampler.hxx.

◆ fNumMinor

std::size_t ROOT::Experimental::Internal::ML::RSampler::fNumMinor

private

Definition at line 47 of file RSampler.hxx.

◆ fNumResampledMajor

std::size_t ROOT::Experimental::Internal::ML::RSampler::fNumResampledMajor

private

Definition at line 48 of file RSampler.hxx.

◆ fNumResampledMinor

std::size_t ROOT::Experimental::Internal::ML::RSampler::fNumResampledMinor

private

Definition at line 49 of file RSampler.hxx.

◆ fReplacement

bool ROOT::Experimental::Internal::ML::RSampler::fReplacement

private

Definition at line 39 of file RSampler.hxx.

◆ fRows

std::vector<std::vector<std::size_t> >& ROOT::Experimental::Internal::ML::RSampler::fRows

private

Definition at line 36 of file RSampler.hxx.

◆ fSampleRatio

float ROOT::Experimental::Internal::ML::RSampler::fSampleRatio

private

Definition at line 38 of file RSampler.hxx.

◆ fSamples

std::vector<std::size_t> ROOT::Experimental::Internal::ML::RSampler::fSamples

private

Definition at line 51 of file RSampler.hxx.

◆ fSampleType

std::string ROOT::Experimental::Internal::ML::RSampler::fSampleType

private

Definition at line 37 of file RSampler.hxx.

◆ fSetSeed

std::size_t ROOT::Experimental::Internal::ML::RSampler::fSetSeed

private

Definition at line 41 of file RSampler.hxx.

◆ fShuffle

bool ROOT::Experimental::Internal::ML::RSampler::fShuffle

private

Definition at line 40 of file RSampler.hxx.

◆ fTensorOperators

std::unique_ptr<RFlat2DMatrixOperators> ROOT::Experimental::Internal::ML::RSampler::fTensorOperators

private

Definition at line 53 of file RSampler.hxx.

tree/ml/src/RSampler.hxx

tree/ml/src/RSampler.cxx

Public Member Functions

Private Attributes

Constructor & Destructor Documentation

◆ RSampler()

◆ ~RSampler()

Member Function Documentation

◆ GatherSamples()

◆ GetNumEntries()

◆ RandomOversampler()

◆ RandomUndersampler()

◆ Sampler()

◆ SampleWithoutReplacement()

◆ SampleWithReplacement()

◆ SetupRandomOversampler()

◆ SetupRandomUndersampler()

◆ SetupSampler()

Member Data Documentation

◆ fDataset

◆ fMajor

◆ fMinor

◆ fNumEntries

◆ fNumMajor

◆ fNumMinor

◆ fNumResampledMajor

◆ fNumResampledMinor

◆ fReplacement

◆ fRows

◆ fSampleRatio

◆ fSamples

◆ fSampleType

◆ fSetSeed

◆ fShuffle

◆ fTensorOperators