public class DIBD extends Discretizer
This class implements the DIBD
classOfInstances, cutPoints, iClassIndex, realAttributes, realValues
Constructor and Description |
---|
DIBD()
Constructor of the class
|
Modifier and Type | Method and Description |
---|---|
int |
computeIntervalNI(int class_,
int start,
int end)
It computes the number of instances with attribute value in the interval [start, end] and class class_
|
int |
computeNI(int class_,
int value,
int option)
It computes the number of instances of class class_ (or all classes) and attribute value <= or > than
value, according to option
|
protected java.util.Vector |
discretizeAttribute(int attribute,
int[] values,
int begin,
int end)
Returns a vector with the discretized values
|
double |
Ecom(int start,
int end)
It computes the compound distributional index needed to compute the compound decrement of a cutpoint
|
double |
Ed(int start,
int end)
It computes the decision distributional index needed to compute the compound distributional index (Ecom)
|
double |
Eleft(int value)
It computes the Left decision distributional index needed to compute the entropy of a cutpoint
|
double |
Eright(int value)
It computes the Right decision distributional index needed to compute the entropy of a cutpoint
|
double |
Ev(int start,
int end)
It computes the value distributional index needed to compute the compound distributional index (Ecom)
|
applyDiscretization, buildCutPoints, discretize, getCutPoint, getNumIntervals, sortValues
protected java.util.Vector discretizeAttribute(int attribute, int[] values, int begin, int end)
Returns a vector with the discretized values
discretizeAttribute
in class Discretizer
attribute
- number of the attributevalues
- vector of indexes of the instances sorted from lowest to highestbegin
- index of the instance with the lowest value of attributeend
- index of the instance with the lowest value of attributepublic int computeNI(int class_, int value, int option)
It computes the number of instances of class class_ (or all classes) and attribute value <= or > than value, according to option
class_
- class of instances computedvalue
- value to compareoption
- is equal to one tag: LEFT (for <=) or RIGHT (for >)public int computeIntervalNI(int class_, int start, int end)
It computes the number of instances with attribute value in the interval [start, end] and class class_
class_
- class of instances computedstart
- begin of the intervalend
- end of the intervalpublic double Eleft(int value)
It computes the Left decision distributional index needed to compute the entropy of a cutpoint
value
- index of the cutpoint studiedpublic double Eright(int value)
It computes the Right decision distributional index needed to compute the entropy of a cutpoint
value
- index of the cutpoint studiedpublic double Ecom(int start, int end)
It computes the compound distributional index needed to compute the compound decrement of a cutpoint
start
- begin of the intervalend
- end of the intervalpublic double Ed(int start, int end)
It computes the decision distributional index needed to compute the compound distributional index (Ecom)
start
- begin of the intervalend
- end of the intervalpublic double Ev(int start, int end)
It computes the value distributional index needed to compute the compound distributional index (Ecom)
start
- begin of the intervalend
- end of the interval