Logo ROOT   6.13/01
Reference Guide
List of all members | Public Types | Public Member Functions | Private Member Functions | Private Attributes | Friends | List of all members
TKDTreeBinning Class Reference

<- TKDTreeBinning - A class providing multidimensional binning ->

The class implements multidimensional binning by constructing a TKDTree inner structure from the data which is used as the bins. The bins are retrieved as two double*, one for the minimum bin edges, the other as the maximum bin edges. For one dimension one of these is enough to correctly define the bins. For the multidimensional case both minimum and maximum ones are necessary for the bins to be well defined. The bin edges of d-dimensional data is a d-tet of the bin's thresholds. For example if d=3 the minimum bin edges of bin b is of the form of the following array: {xbmin, ybmin, zbmin}. You also have the possibility to sort the bins by their density.

Details of usage can be found in $ROOTSYS/tutorials/math/kdTreeBinning.C and more information on the embedded TKDTree documentation.

Definition at line 29 of file TKDTreeBinning.h.

Public Types

enum  { kAdjustBinEdges = BIT(14) }
 

Public Member Functions

 TKDTreeBinning ()
 Default constructor (for I/O) More...
 
 TKDTreeBinning (UInt_t dataSize, UInt_t dataDim, Double_t *data, UInt_t nBins=100, bool adjustBinEdges=false)
 Class's constructor taking the size of the data points, dimension, a data array and the number of bins (default = 100). More...
 
 TKDTreeBinning (UInt_t dataSize, UInt_t dataDim, const std::vector< double > &data, UInt_t nBins=100, bool adjustBinEdges=false)
 Class's constructor taking the size of the data points, dimension, a data vector and the number of bins (default = 100). More...
 
 ~TKDTreeBinning ()
 Class's destructor. More...
 
void FillBinData (ROOT::Fit::BinData &data) const
 Fill the bin data set (class ROOT::Fit::BinData) with the result of the TKDTree binning. More...
 
UInt_t FindBin (const Double_t *point) const
 find the corresponding bin index given the coordinate of a point More...
 
const Double_t * GetBinCenter (UInt_t bin) const
 Returns the geometric center of of the bin. 'bin' is between 0 and fNBins - 1. More...
 
UInt_t GetBinContent (UInt_t bin) const
 Returns the number of points in bin. 'bin' is between 0 and fNBins - 1. More...
 
Double_t GetBinDensity (UInt_t bin) const
 Returns the density in bin. More...
 
std::pair< const Double_t *, const Double_t * > GetBinEdges (UInt_t bin) const
 Returns a pir with the bin's edges. 'bin' is between 0 and fNBins - 1. More...
 
UInt_t GetBinMaxDensity () const
 Return the bin with maximum density. More...
 
const Double_t * GetBinMaxEdges (UInt_t bin) const
 Returns the bin's maximum edges. 'bin' is between 0 and fNBins - 1. More...
 
UInt_t GetBinMinDensity () const
 Return the bin with minimum density. More...
 
const Double_t * GetBinMinEdges (UInt_t bin) const
 Returns the bin's minimum edges. 'bin' is between 0 and fNBins - 1. More...
 
std::pair< const Double_t *, const Double_t * > GetBinsEdges () const
 Returns a pair of an array with all bins minimum and maximum edges. More...
 
const Double_t * GetBinsMaxEdges () const
 Returns an array with all bins' maximum edges The edges are arranges as xmax_1,ymax_1, xmax_2,ymax_2,....xmax_{nbin},ymax_{nbin}. More...
 
const Double_t * GetBinsMinEdges () const
 Returns an array with all bins' minimum edges The edges are arranges as xmin_1,ymin_1, xmin_2,ymin_2,....xmin_{nbin},ymin_{nbin}. More...
 
Double_t GetBinVolume (UInt_t bin) const
 Returns the (hyper)volume of bin. 'bin' is between 0 and fNBins - 1. More...
 
const Double_t * GetBinWidth (UInt_t bin) const
 Returns a pointer to the vector of the bin widths. 'bin' is between 0 and fNBins - 1. More...
 
Double_t GetDataMax (UInt_t dim) const
 Returns the maximum of the data in the dim coordinate. 'dim' is between 0 and fDim - 1. More...
 
Double_t GetDataMin (UInt_t dim) const
 Returns the minimum of the data in the dim coordinate. 'dim' is between 0 and fDim - 1. More...
 
UInt_t GetDim () const
 Returns the number of dimensions. More...
 
const Double_t * GetDimData (UInt_t dim) const
 
UInt_t GetNBins () const
 Returns the number of bins. More...
 
const Double_t * GetOneDimBinEdges () const
 Returns a pointer to the vector of the bin edges for one dimensional binning only. More...
 
std::vector< std::vector< Double_t > > GetPointsInBin (UInt_t bin) const
 Return the corresponding point belonging to the bin i. More...
 
TKDTreeIDGetTree () const
 Returns the kD-Tree structure of the binning. More...
 
void SetNBins (UInt_t bins)
 Sets binning inner structure. More...
 
void SortBinsByDensity (Bool_t sortAsc=kTRUE)
 Sorts bins by their density. More...
 
const Double_t * SortOneDimBinEdges (Bool_t sortAsc=kTRUE)
 Sort the one-dimensional bin edges and retuns a pointer to them. More...
 

Private Member Functions

 TKDTreeBinning (TKDTreeBinning &bins)
 
TKDTreeBinning operator= (TKDTreeBinning &bins)
 Disallowed copy constructor. More...
 
void ReadjustMaxBinEdges (Double_t *binEdges)
 
void ReadjustMinBinEdges (Double_t *binEdges)
 
void SetBinMinMaxEdges (Double_t *binEdges)
 
void SetBinsContent ()
 
void SetBinsEdges ()
 
void SetCommonBinEdges (Double_t *binEdges)
 
void SetData (Double_t *data)
 Disallowed assign operator. More...
 
void SetData (const std::vector< double > &data)
 
void SetTreeData ()
 

Private Attributes

std::vector< Double_t > fBinMaxEdges
 The minimum values for the bins' edges for each dimension. More...
 
std::vector< Double_t > fBinMinEdges
 [fDataSize*fDim] The data from which a KDTree partition is computed for binning More...
 
std::vector< UInt_t > fBinsContent
 Flags if the bin edges are sorted densitywise (or by bin-edge for 1D) in ascending order. More...
 
std::vector< std::vector< std::pair< Bool_t, Bool_t > > > fCheckedBinEdges
 Minimum and maximum data values. More...
 
std::vector< std::map< Double_t, std::vector< UInt_t > > > fCommonBinEdges
 ! Auxiliary structure for readjusting the bin edges. Keeps the common bin boundaries More...
 
std::vector< Double_t > fData
 
TKDTreeIDfDataBins
 Index of the bins in the kd-tree (needed when bins are sorted) More...
 
UInt_t fDataSize
 The data dimension. More...
 
std::vector< std::pair< Double_t, Double_t > > fDataThresholds
 The data size. More...
 
UInt_t fDim
 The number of bins. More...
 
std::vector< UInt_t > fIndices
 The maximum values for the bins' edges for each dimension. More...
 
Bool_t fIsSorted
 
Bool_t fIsSortedAsc
 Flags if the bin edges are sorted densitywise (or by bin endges in case of 1-dim ) More...
 
UInt_t fNBins
 

Friends

struct CompareAsc
 ! Predicate for ascending sort More...
 
struct CompareDesc
 ! Predicate for descending sort More...
 

#include <TKDTreeBinning.h>

Inheritance diagram for TKDTreeBinning:
[legend]

Member Enumeration Documentation

◆ anonymous enum

anonymous enum
Enumerator
kAdjustBinEdges 

Definition at line 67 of file TKDTreeBinning.h.

Constructor & Destructor Documentation

◆ TKDTreeBinning() [1/4]

TKDTreeBinning::TKDTreeBinning ( TKDTreeBinning bins)
private

◆ TKDTreeBinning() [2/4]

TKDTreeBinning::TKDTreeBinning ( )

Default constructor (for I/O)

Definition at line 113 of file TKDTreeBinning.cxx.

◆ TKDTreeBinning() [3/4]

TKDTreeBinning::TKDTreeBinning ( UInt_t  dataSize,
UInt_t  dataDim,
Double_t *  data,
UInt_t  nBins = 100,
bool  adjustBinEdges = false 
)

Class's constructor taking the size of the data points, dimension, a data array and the number of bins (default = 100).

It is reccomended to have the number of bins as an exact divider of the data size. The data array must be organized with a stride=1 for the points and = N (the dataSize) for the dimension.

Thus data[] = x1,x2,x3,......xN, y1,y2,y3......yN, z1,z2,...........zN,....

Note that the passed dataSize is not the size of the array but is the number of points (N) The size of the array must be at least dataDim*dataSize

Definition at line 74 of file TKDTreeBinning.cxx.

◆ TKDTreeBinning() [4/4]

TKDTreeBinning::TKDTreeBinning ( UInt_t  dataSize,
UInt_t  dataDim,
const std::vector< double > &  data,
UInt_t  nBins = 100,
bool  adjustBinEdges = false 
)

Class's constructor taking the size of the data points, dimension, a data vector and the number of bins (default = 100).

It is reccomended to have the number of bins as an exact divider of the data size. The data array must be organized with a stride=1 for the points and = N (the dataSize) for the dimension.

Thus data[] = x1,x2,x3,......xN, y1,y2,y3......yN, z1,z2,...........zN,....

Note that the passed data vector may contains a larger size, in case extra coordinates are associated but not used in building the kdtree The size of thedata vector must be at least dataDim*dataSize

Definition at line 98 of file TKDTreeBinning.cxx.

◆ ~TKDTreeBinning()

TKDTreeBinning::~TKDTreeBinning ( )

Class's destructor.

Definition at line 123 of file TKDTreeBinning.cxx.

Member Function Documentation

◆ FillBinData()

void TKDTreeBinning::FillBinData ( ROOT::Fit::BinData data) const

Fill the bin data set (class ROOT::Fit::BinData) with the result of the TKDTree binning.

Definition at line 614 of file TKDTreeBinning.cxx.

◆ FindBin()

UInt_t TKDTreeBinning::FindBin ( const Double_t *  point) const

find the corresponding bin index given the coordinate of a point

Definition at line 624 of file TKDTreeBinning.cxx.

◆ GetBinCenter()

const Double_t * TKDTreeBinning::GetBinCenter ( UInt_t  bin) const

Returns the geometric center of of the bin. 'bin' is between 0 and fNBins - 1.

Definition at line 554 of file TKDTreeBinning.cxx.

◆ GetBinContent()

UInt_t TKDTreeBinning::GetBinContent ( UInt_t  bin) const

Returns the number of points in bin. 'bin' is between 0 and fNBins - 1.

Definition at line 429 of file TKDTreeBinning.cxx.

◆ GetBinDensity()

Double_t TKDTreeBinning::GetBinDensity ( UInt_t  bin) const

Returns the density in bin.

'bin' is between 0 and fNBins - 1 The density is the bin content/ bin volume

Definition at line 475 of file TKDTreeBinning.cxx.

◆ GetBinEdges()

std::pair< const Double_t *, const Double_t * > TKDTreeBinning::GetBinEdges ( UInt_t  bin) const

Returns a pir with the bin's edges. 'bin' is between 0 and fNBins - 1.

Definition at line 406 of file TKDTreeBinning.cxx.

◆ GetBinMaxDensity()

UInt_t TKDTreeBinning::GetBinMaxDensity ( ) const

Return the bin with maximum density.

Definition at line 584 of file TKDTreeBinning.cxx.

◆ GetBinMaxEdges()

const Double_t * TKDTreeBinning::GetBinMaxEdges ( UInt_t  bin) const

Returns the bin's maximum edges. 'bin' is between 0 and fNBins - 1.

Definition at line 393 of file TKDTreeBinning.cxx.

◆ GetBinMinDensity()

UInt_t TKDTreeBinning::GetBinMinDensity ( ) const

Return the bin with minimum density.

Definition at line 599 of file TKDTreeBinning.cxx.

◆ GetBinMinEdges()

const Double_t * TKDTreeBinning::GetBinMinEdges ( UInt_t  bin) const

Returns the bin's minimum edges. 'bin' is between 0 and fNBins - 1.

Definition at line 380 of file TKDTreeBinning.cxx.

◆ GetBinsEdges()

std::pair< const Double_t *, const Double_t * > TKDTreeBinning::GetBinsEdges ( ) const

Returns a pair of an array with all bins minimum and maximum edges.

Definition at line 370 of file TKDTreeBinning.cxx.

◆ GetBinsMaxEdges()

const Double_t * TKDTreeBinning::GetBinsMaxEdges ( ) const

Returns an array with all bins' maximum edges The edges are arranges as xmax_1,ymax_1, xmax_2,ymax_2,....xmax_{nbin},ymax_{nbin}.

Definition at line 360 of file TKDTreeBinning.cxx.

◆ GetBinsMinEdges()

const Double_t * TKDTreeBinning::GetBinsMinEdges ( ) const

Returns an array with all bins' minimum edges The edges are arranges as xmin_1,ymin_1, xmin_2,ymin_2,....xmin_{nbin},ymin_{nbin}.

Definition at line 350 of file TKDTreeBinning.cxx.

◆ GetBinVolume()

Double_t TKDTreeBinning::GetBinVolume ( UInt_t  bin) const

Returns the (hyper)volume of bin. 'bin' is between 0 and fNBins - 1.

Definition at line 488 of file TKDTreeBinning.cxx.

◆ GetBinWidth()

const Double_t * TKDTreeBinning::GetBinWidth ( UInt_t  bin) const

Returns a pointer to the vector of the bin widths. 'bin' is between 0 and fNBins - 1.

Definition at line 569 of file TKDTreeBinning.cxx.

◆ GetDataMax()

Double_t TKDTreeBinning::GetDataMax ( UInt_t  dim) const

Returns the maximum of the data in the dim coordinate. 'dim' is between 0 and fDim - 1.

Definition at line 465 of file TKDTreeBinning.cxx.

◆ GetDataMin()

Double_t TKDTreeBinning::GetDataMin ( UInt_t  dim) const

Returns the minimum of the data in the dim coordinate. 'dim' is between 0 and fDim - 1.

Definition at line 456 of file TKDTreeBinning.cxx.

◆ GetDim()

UInt_t TKDTreeBinning::GetDim ( ) const

Returns the number of dimensions.

Definition at line 424 of file TKDTreeBinning.cxx.

◆ GetDimData()

const Double_t * TKDTreeBinning::GetDimData ( UInt_t  dim) const

Definition at line 447 of file TKDTreeBinning.cxx.

◆ GetNBins()

UInt_t TKDTreeBinning::GetNBins ( ) const

Returns the number of bins.

Definition at line 419 of file TKDTreeBinning.cxx.

◆ GetOneDimBinEdges()

const double * TKDTreeBinning::GetOneDimBinEdges ( ) const

Returns a pointer to the vector of the bin edges for one dimensional binning only.

size of the vector is fNBins + 1 is the vector has been sorted in increasing bin edges N.B : if one does not call SortOneDimBinEdges the bins are not ordered

Definition at line 505 of file TKDTreeBinning.cxx.

◆ GetPointsInBin()

std::vector< std::vector< Double_t > > TKDTreeBinning::GetPointsInBin ( UInt_t  bin) const

Return the corresponding point belonging to the bin i.

Definition at line 639 of file TKDTreeBinning.cxx.

◆ GetTree()

TKDTreeID * TKDTreeBinning::GetTree ( ) const

Returns the kD-Tree structure of the binning.

Definition at line 439 of file TKDTreeBinning.cxx.

◆ operator=()

TKDTreeBinning TKDTreeBinning::operator= ( TKDTreeBinning bins)
private

Disallowed copy constructor.

◆ ReadjustMaxBinEdges()

void TKDTreeBinning::ReadjustMaxBinEdges ( Double_t *  binEdges)
private

Definition at line 329 of file TKDTreeBinning.cxx.

◆ ReadjustMinBinEdges()

void TKDTreeBinning::ReadjustMinBinEdges ( Double_t *  binEdges)
private

Definition at line 300 of file TKDTreeBinning.cxx.

◆ SetBinMinMaxEdges()

void TKDTreeBinning::SetBinMinMaxEdges ( Double_t *  binEdges)
private

Definition at line 267 of file TKDTreeBinning.cxx.

◆ SetBinsContent()

void TKDTreeBinning::SetBinsContent ( )
private

Definition at line 242 of file TKDTreeBinning.cxx.

◆ SetBinsEdges()

void TKDTreeBinning::SetBinsEdges ( )
private

Definition at line 251 of file TKDTreeBinning.cxx.

◆ SetCommonBinEdges()

void TKDTreeBinning::SetCommonBinEdges ( Double_t *  binEdges)
private

Definition at line 279 of file TKDTreeBinning.cxx.

◆ SetData() [1/2]

void TKDTreeBinning::SetData ( Double_t *  data)
private

Disallowed assign operator.

Definition at line 211 of file TKDTreeBinning.cxx.

◆ SetData() [2/2]

void TKDTreeBinning::SetData ( const std::vector< double > &  data)
private

Definition at line 224 of file TKDTreeBinning.cxx.

◆ SetNBins()

void TKDTreeBinning::SetNBins ( UInt_t  bins)

Sets binning inner structure.

Definition at line 129 of file TKDTreeBinning.cxx.

◆ SetTreeData()

void TKDTreeBinning::SetTreeData ( )
private

Definition at line 236 of file TKDTreeBinning.cxx.

◆ SortBinsByDensity()

void TKDTreeBinning::SortBinsByDensity ( Bool_t  sortAsc = kTRUE)

Sorts bins by their density.

Definition at line 159 of file TKDTreeBinning.cxx.

◆ SortOneDimBinEdges()

const Double_t * TKDTreeBinning::SortOneDimBinEdges ( Bool_t  sortAsc = kTRUE)

Sort the one-dimensional bin edges and retuns a pointer to them.

Definition at line 516 of file TKDTreeBinning.cxx.

Friends And Related Function Documentation

◆ CompareAsc

friend struct CompareAsc
friend

! Predicate for ascending sort

Definition at line 46 of file TKDTreeBinning.h.

◆ CompareDesc

friend struct CompareDesc
friend

! Predicate for descending sort

Definition at line 48 of file TKDTreeBinning.h.

Member Data Documentation

◆ fBinMaxEdges

std::vector<Double_t> TKDTreeBinning::fBinMaxEdges
private

The minimum values for the bins' edges for each dimension.

Definition at line 34 of file TKDTreeBinning.h.

◆ fBinMinEdges

std::vector<Double_t> TKDTreeBinning::fBinMinEdges
private

[fDataSize*fDim] The data from which a KDTree partition is computed for binning

Definition at line 33 of file TKDTreeBinning.h.

◆ fBinsContent

std::vector<UInt_t> TKDTreeBinning::fBinsContent
private

Flags if the bin edges are sorted densitywise (or by bin-edge for 1D) in ascending order.

Definition at line 45 of file TKDTreeBinning.h.

◆ fCheckedBinEdges

std::vector<std::vector<std::pair<Bool_t, Bool_t> > > TKDTreeBinning::fCheckedBinEdges
private

Minimum and maximum data values.

! Auxiliary structure for readjusting the bin edges. Flags if the bin edge was processed in the algorithm

Definition at line 41 of file TKDTreeBinning.h.

◆ fCommonBinEdges

std::vector<std::map<Double_t, std::vector<UInt_t> > > TKDTreeBinning::fCommonBinEdges
private

! Auxiliary structure for readjusting the bin edges. Keeps the common bin boundaries

Definition at line 42 of file TKDTreeBinning.h.

◆ fData

std::vector<Double_t> TKDTreeBinning::fData
private

Definition at line 32 of file TKDTreeBinning.h.

◆ fDataBins

TKDTreeID* TKDTreeBinning::fDataBins
private

Index of the bins in the kd-tree (needed when bins are sorted)

! The binning inner structure.

Definition at line 36 of file TKDTreeBinning.h.

◆ fDataSize

UInt_t TKDTreeBinning::fDataSize
private

The data dimension.

Definition at line 39 of file TKDTreeBinning.h.

◆ fDataThresholds

std::vector<std::pair<Double_t, Double_t> > TKDTreeBinning::fDataThresholds
private

The data size.

Definition at line 40 of file TKDTreeBinning.h.

◆ fDim

UInt_t TKDTreeBinning::fDim
private

The number of bins.

Definition at line 38 of file TKDTreeBinning.h.

◆ fIndices

std::vector<UInt_t> TKDTreeBinning::fIndices
private

The maximum values for the bins' edges for each dimension.

Definition at line 35 of file TKDTreeBinning.h.

◆ fIsSorted

Bool_t TKDTreeBinning::fIsSorted
private

Definition at line 43 of file TKDTreeBinning.h.

◆ fIsSortedAsc

Bool_t TKDTreeBinning::fIsSortedAsc
private

Flags if the bin edges are sorted densitywise (or by bin endges in case of 1-dim )

Definition at line 44 of file TKDTreeBinning.h.

◆ fNBins

UInt_t TKDTreeBinning::fNBins
private

Definition at line 37 of file TKDTreeBinning.h.

Libraries for TKDTreeBinning:
[legend]

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