Package weka.associations
Class RuleItem
java.lang.Object
weka.associations.RuleItem
- All Implemented Interfaces:
Serializable
,Comparable
,RevisionHandler
Class for storing an (class) association rule.
The premise and the consequence are stored each as separate item sets.
For every rule their expected predictive accuracy and the time of generation is stored.
These two measures allow to introduce a sort order for rules.
- Version:
- $Revision: 1.5 $
- Author:
- Stefan Mutter
- See Also:
-
Constructor Summary
Constructors -
Method Summary
Modifier and TypeMethodDescriptiondouble
accuracy()
Gets the expected predictive accuracy of a ruleint
compares two RuleItems and allows an ordering concerning expected predictive accuracy and time of generation Note: this class has a natural ordering that is inconsistent with equalsGets the consequence of a ruleboolean
returns whether two RuleItems are equalgenerateRuleItem
(ItemSet premise, ItemSet consequence, Instances instances, int genTime, int minRuleCount, double[] m_midPoints, Hashtable m_priors) Constructs a new RuleItem if the support of the given rule is above the support threshold.Returns the revision string.premise()
Gets the premise of a rule
-
Constructor Details
-
RuleItem
public RuleItem()Constructor for an empty RuleItem -
RuleItem
Constructor that generates a RuleItem out of a given one- Parameters:
toCopy
- RuleItem to copy
-
RuleItem
public RuleItem(ItemSet premise, ItemSet consequence, int genTime, int ruleSupport, double[] m_midPoints, Hashtable m_priors) Constructor- Parameters:
premise
- the premise of the future RuleItemconsequence
- the consequence of the future RuleItemgenTime
- the time of generation of the future RuleItemruleSupport
- support of the rulem_midPoints
- the mid poitns of the intervalsm_priors
- Hashtable containing the estimated prior probablilities
-
-
Method Details
-
generateRuleItem
public RuleItem generateRuleItem(ItemSet premise, ItemSet consequence, Instances instances, int genTime, int minRuleCount, double[] m_midPoints, Hashtable m_priors) Constructs a new RuleItem if the support of the given rule is above the support threshold.- Parameters:
premise
- the premiseconsequence
- the consequenceinstances
- the instancesgenTime
- the time of generation of the current premise and consequenceminRuleCount
- the support thresholdm_midPoints
- the mid points of the intervalsm_priors
- the estimated priori probabilities (in a hashtable)- Returns:
- a RuleItem if its support is above the threshold, null otherwise
-
compareTo
compares two RuleItems and allows an ordering concerning expected predictive accuracy and time of generation Note: this class has a natural ordering that is inconsistent with equals- Specified by:
compareTo
in interfaceComparable
- Parameters:
o
- RuleItem to compare- Returns:
- integer indicating the sort oder of the two RuleItems
-
equals
returns whether two RuleItems are equal -
accuracy
public double accuracy()Gets the expected predictive accuracy of a rule- Returns:
- the expected predictive accuracy of a rule stored as a RuleItem
-
premise
Gets the premise of a rule- Returns:
- the premise of a rule stored as a RuleItem
-
consequence
Gets the consequence of a rule- Returns:
- the consequence of a rule stored as a RuleItem
-
getRevision
Returns the revision string.- Specified by:
getRevision
in interfaceRevisionHandler
- Returns:
- the revision
-