Darwin  1.10(beta)
 All Classes Namespaces Files Functions Variables Typedefs Enumerations Enumerator Friends Groups Pages
Public Member Functions | List of all members
drwnPRCurve Class Reference

Precision-recall curve. More...

Inheritance diagram for drwnPRCurve:

Public Member Functions

 drwnPRCurve ()
 default constructor
 drwnPRCurve (const drwnClassificationResults &c)
 copy constructor
vector< pair< double, double > > getCurve () const
 return a list of points defining the precision-recall curve
void writeCurve (const char *filename) const
 write the precision-recall curve to a space-delimited file
double averagePrecision (unsigned numPoints=11) const
 extract the average precision (area under the curve)
- Public Member Functions inherited from drwnClassificationResults
 drwnClassificationResults ()
 default constructor
 drwnClassificationResults (const drwnClassificationResults &c)
 copy constructor
int numPositives () const
 return the number os positive samples accumulated
int numNegatives () const
 return the number of negative samples accumulated
int numSamples () const
 return the total number (positive and negative) of samples accumulated
int numThresholds () const
 return the number of unique classification scores
int numMisses () const
 return the number of positive samples that have not been scored
double getPosWeight () const
 return the relative weight of a positive sample to a negative sample
void setPosWeight (double w)
 set the relative weight of a positive sample to a negative sample
void normalize ()
 this will change the weight of the positive examples such that overall positive and negative examples will have the same weight
void clear ()
 clear the accumulated scores
bool write (const char *filename) const
 write the accumulated scores to file
bool read (const char *filename)
 read accumulated scores from file
void accumulate (const drwnClassificationResults &c)
 accumulate results from another drwnClassificationResults object
void accumulate (const drwnClassifierDataset &dataset, drwnClassifier const *classifier, int positiveClassId=1)
 Accumulate results from a classifier run on a dataset. The positiveClass parameter indicates the positive class label for multi-class classifiers.
void accumulatePositives (double score, int count=1)
 accumulate a single positive example
void accumulatePositives (const vector< double > &scores)
 accumulate multiple positive examples
void accumulateNegatives (double score, int count=1)
 accumulate a single negative example
void accumulateNegatives (const vector< double > &scores)
 accumulate multiple negative examples
void accumulateMisses (int count=1)
 accumulate unscored positive examples (misses)

Additional Inherited Members

- Static Public Attributes inherited from drwnClassificationResults
static bool INCLUDE_MISSES = false
 true if some positive samples are never scored
- Protected Attributes inherited from drwnClassificationResults
map< double, pair< int, int > > _scoredResults
 number of positives (first) and negatives (second) grouped by score
int _numPositiveSamples
 must be greater than sum(_scoredResults.first)
int _numNegativeSamples
 must be must be equal to sum(_scoredResults.second)
double _posWeight
 weight of positive-to-negative count

Detailed Description

Precision-recall curve.

See Also
drwnClassificationResults, drwnML Tutorial

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