public class ChiSquareSplitCrit
extends weka.classifiers.trees.j48.SplitCriterion
Modifier and Type | Field and Description |
---|---|
private static long |
serialVersionUID
for serialization
|
Constructor and Description |
---|
ChiSquareSplitCrit() |
Modifier and Type | Method and Description |
---|---|
double |
BFfactor(int c,
int r)
Compute factor for Bonferroni correction.
|
double |
bonferroniFactor(int c,
int r,
boolean hasMissing,
boolean ordered)
Compute factor for Bonferroni correction for Nominal (Free) and Ordinal (Monotonic) attributes
taking into account whether there is missing values or not (Floating).
|
double |
chiVal(weka.classifiers.trees.j48.Distribution bags)
Computes chi-squared statistic for given distribution.
|
double |
chiVal(double[][] matrix)
Computes chi-squared statistic for given matrix.
|
double |
combinations(int n,
int k)
Compute the binomial coefficient or the number of ways/combinations to choose k elements,
disregarding their order, from a set of n elements.
|
java.lang.String |
getRevision()
Returns the revision string.
|
double |
splitCritValue(weka.classifiers.trees.j48.Distribution bags)
Computes chi-squared probability for given distribution.
|
double |
splitCritValue(double ch,
double dg)
Computes chi-squared probability for given ch statistic and dg degrees of freedom.
|
private static final long serialVersionUID
public java.lang.String getRevision()
weka.core.RevisionHandler
public double chiVal(weka.classifiers.trees.j48.Distribution bags)
public double chiVal(double[][] matrix)
public double splitCritValue(weka.classifiers.trees.j48.Distribution bags)
splitCritValue
in class weka.classifiers.trees.j48.SplitCriterion
public double splitCritValue(double ch, double dg)
public double BFfactor(int c, int r)
public double combinations(int n, int k)
n
- number of elements of the setk
- elements to choosepublic double bonferroniFactor(int c, int r, boolean hasMissing, boolean ordered)
c
- Number of known categories or values that the attribute can take
(included missing value, if exists)r
- Number of groups of values in which the original values have been combinedhasMissing
- Indicates if missing values are presentordered
- Indicates if the attribute's categories lie on an ordinal scale