OpenMS
ROCCurve Class Reference

ROCCurves show the trade-off in sensitivity and specificity for binary classifiers using different cutoff values. More...

#include <OpenMS/ML/ROCCURVE/ROCCurve.h>

Collaboration diagram for ROCCurve:
[legend]

Classes

class  simsortdec
 predicate for sort() More...
 

Public Member Functions

 ROCCurve ()
 default constructor More...
 
 ROCCurve (const std::vector< std::pair< double, bool >> &pairs)
 constructor with value, class pairs More...
 
virtual ~ROCCurve ()=default
 destructor More...
 
 ROCCurve (const ROCCurve &source)
 copy constructor More...
 
ROCCurveoperator= (const ROCCurve &source)
 assignment operator More...
 
void insertPair (double score, bool clas)
 insert score, type pair More...
 
double AUC ()
 returns Area Under Curve More...
 
double rocN (Size N)
 returns ROC-N score (e.g. ROC-50). Returns -1 if not enough false positives were found More...
 
std::vector< std::pair< double, double > > curve (UInt resolution=10)
 some points in the ROC Curve More...
 
double cutoffPos (double fraction=0.95)
 
double cutoffNeg (double fraction=0.95)
 

Private Member Functions

void sort ()
 sorts data and caches if sorted More...
 
void count ()
 counts global pos and neg More...
 
double trapezoidal_area (double x1, double x2, double y1, double y2)
 

Private Attributes

std::vector< std::pair< double, bool > > score_clas_pairs_
 
UInt pos_ {}
 
UInt neg_ {}
 
bool sorted_ {}
 

Detailed Description

ROCCurves show the trade-off in sensitivity and specificity for binary classifiers using different cutoff values.

[This class is buggy and usage is discouraged!]

Constructor & Destructor Documentation

◆ ROCCurve() [1/3]

ROCCurve ( )

default constructor

◆ ROCCurve() [2/3]

ROCCurve ( const std::vector< std::pair< double, bool >> &  pairs)
explicit

constructor with value, class pairs

◆ ~ROCCurve()

virtual ~ROCCurve ( )
virtualdefault

destructor

◆ ROCCurve() [3/3]

ROCCurve ( const ROCCurve source)

copy constructor

Member Function Documentation

◆ AUC()

double AUC ( )

returns Area Under Curve

◆ count()

void count ( )
inlineprivate

counts global pos and neg

◆ curve()

std::vector<std::pair<double, double> > curve ( UInt  resolution = 10)

some points in the ROC Curve

◆ cutoffNeg()

double cutoffNeg ( double  fraction = 0.95)

returns the score at which you would need to set a cutoff to have fraction positives returns -1 if there are not enough positives

◆ cutoffPos()

double cutoffPos ( double  fraction = 0.95)

returns the score at which you would need to set a cutoff to have fraction positives returns -1 if there are not enough positives

◆ insertPair()

void insertPair ( double  score,
bool  clas 
)

insert score, type pair

◆ operator=()

ROCCurve& operator= ( const ROCCurve source)

assignment operator

◆ rocN()

double rocN ( Size  N)

returns ROC-N score (e.g. ROC-50). Returns -1 if not enough false positives were found

◆ sort()

void sort ( )
inlineprivate

sorts data and caches if sorted

◆ trapezoidal_area()

double trapezoidal_area ( double  x1,
double  x2,
double  y1,
double  y2 
)
inlineprivate

calculates area with trapezoidal rule

Parameters
x1,x2,y1,y2

Member Data Documentation

◆ neg_

UInt neg_ {}
private

◆ pos_

UInt pos_ {}
private

◆ score_clas_pairs_

std::vector<std::pair<double, bool> > score_clas_pairs_
private

◆ sorted_

bool sorted_ {}
private