def __init__(self,
evaluator=orange.RuleEvaluator_Entropy(),
beamWidth=5,
alpha=1.0,
**kwds):
self.__dict__.update(kwds)
self.ruleFinder = orange.RuleBeamFinder()
self.ruleFinder.ruleFilter = orange.RuleBeamFilter_Width(
width=beamWidth)
self.ruleFinder.evaluator = evaluator
self.ruleFinder.validator = orange.RuleValidator_LRS(alpha=alpha)
def __init__(self,
evaluator=orange.RuleEvaluator_Laplace(),
beamWidth=5,
alpha=1.0,
**kwds):
self.__dict__.update(kwds)
self.ruleFinder = orange.RuleBeamFinder()
self.ruleFinder.ruleFilter = orange.RuleBeamFilter_Width(
width=beamWidth)
self.ruleFinder.evaluator = evaluator
self.ruleFinder.validator = orange.RuleValidator_LRS(alpha=alpha)
self.ruleFinder.ruleStoppingValidator = orange.RuleValidator_LRS(
alpha=1.0)
self.ruleStopping = RuleStopping_apriori()
self.dataStopping = orange.RuleDataStoppingCriteria_NoPositives()
def setLearner(self):
if hasattr(self, "btnApply"):
self.btnApply.setFocus()
#progress bar
self.progressBarInit()
#learner / specific handling in case of EVC learning (completely different type of class)
if self.useMaxRuleLength:
maxRuleLength = self.MaxRuleLength
else:
maxRuleLength = -1
if self.QualityButton == 2:
self.learner = orngCN2.CN2EVCUnorderedLearner(
width=self.BeamWidth,
rule_sig=self.Alpha,
att_sig=self.stepAlpha,
min_coverage=self.MinCoverage,
max_rule_complexity=maxRuleLength)
if self.preprocessor:
self.learner = self.preprocessor.wrapLearner(self.learner)
self.learner.name = self.name
# self.learner.progressCallback=CN2ProgressBar(self)
self.send("Learner", self.learner)
else:
self.learner = orngCN2.CN2UnorderedLearner()
self.learner.name = self.name
# self.learner.progressCallback=CN2ProgressBar(self)
# self.send("Learner",self.learner)
ruleFinder = orange.RuleBeamFinder()
if self.QualityButton == 0:
ruleFinder.evaluator = orange.RuleEvaluator_Laplace()
elif self.QualityButton == 1:
ruleFinder.evaluator = orngCN2.mEstimate(self.m)
elif self.QualityButton == 3:
ruleFinder.evaluator = orngCN2.WRACCEvaluator()
ruleFinder.ruleStoppingValidator = orange.RuleValidator_LRS(
alpha=self.stepAlpha,
min_coverage=self.MinCoverage,
max_rule_complexity=maxRuleLength)
ruleFinder.validator = orange.RuleValidator_LRS(
alpha=self.Alpha,
min_coverage=self.MinCoverage,
max_rule_complexity=maxRuleLength)
ruleFinder.ruleFilter = orange.RuleBeamFilter_Width(
width=self.BeamWidth)
self.learner.ruleFinder = ruleFinder
if self.CoveringButton == 0:
self.learner.coverAndRemove = orange.RuleCovererAndRemover_Default(
)
elif self.CoveringButton == 1:
self.learner.coverAndRemove = orngCN2.CovererAndRemover_multWeights(
mult=self.Weight)
if self.preprocessor:
self.learner = self.preprocessor.wrapLearner(self.learner)
self.learner.name = self.name
self.send("Learner", self.learner)
self.classifier = None
self.error()
if self.data:
oldDomain = orange.Domain(self.data.domain)
learnData = orange.ExampleTable(oldDomain, self.data)
self.learner.progressCallback = CN2ProgressBar(self)
self.classifier = self.learner(learnData)
self.learner.progressCallback = None
self.classifier.name = self.name
for r in self.classifier.rules:
r.examples = orange.ExampleTable(oldDomain, r.examples)
self.classifier.examples = orange.ExampleTable(
oldDomain, self.classifier.examples)
self.classifier.setattr("data", self.classifier.examples)
self.error("")
## except orange.KernelException, (errValue):
## self.classifier=None
## self.error(errValue)
## except Exception:
## self.classifier=None
## if not self.data.domain.classVar:
## self.error("Classless domain.")
## elif self.data.domain.classVar.varType == orange.VarTypes.Continuous:
## self.error("CN2 can learn only from discrete class.")
## else:
## self.error("Unknown error")
self.send("Classifier", self.classifier)
self.send("Unordered CN2 Classifier", self.classifier)
self.progressBarFinished()
import orange
import orngCN2
data = orange.ExampleTable("titanic.tab")
# create learner
learner = orngCN2.CN2Learner()
cl = learner(data)
for r in cl.rules:
print orngCN2.ruleToString(r)
print "*****"
learner = orngCN2.CN2UnorderedLearner()
learner.ruleFinder = orange.RuleBeamFinder()
learner.ruleFinder.evaluator = orngCN2.mEstimate(m=50)
cl = learner(data)
for r in cl.rules:
print orngCN2.ruleToString(r)
print "****"
learner = orngCN2.CN2SDUnorderedLearner()
learner.ruleFinder.ruleStoppingValidator = orange.RuleValidator_LRS(
alpha=0.01, min_coverage=10, max_rule_complexity=2)
learner.ruleFinder.ruleFilter = orange.RuleBeamFilter_Width(width=50)
cl = learner(data)
for r in cl.rules:
def __init__(self, k):
self.k = k
self.counter = orange.newmetaid()
self.weightID = orange.newmetaid()
self.rbf = orange.RuleBeamFinder()
self.rbf.evaluator = RuleEvaluator_WRAcc()
def __init__(self,
m=2,
opt_reduction=2,
min_improved=1,
min_improved_perc=0.0,
nsampling=100,
width=5,
rule_sig=0.05,
att_sig=0.5,
max_rule_complexity=5,
min_coverage=5,
add_sub_rules=True,
min_cl_sig=0.5,
min_beta=0.0,
set_prefix_rules=False,
alternative_learner=None,
**kwds):
"""
Parameters:
EVC related:
m ... m-estimate to be corrected with EVC (default 2)
opt_reduction ... types of EVC correction; 0=no correction, 1=pessimistic, 2=normal (default 2)
nsampling ... number of samples in estimating extreme value distribution (for EVC) (default 100)
Probabilistic covering:
min_improved ... minimal number of examples improved in probabilistic covering (default 1)
min_improved_perc ... minimal percentage of covered examples that need to be improved (default 0.0)
Beam search:
width ... beam width (default 5)
Rule Validation:
rule_sig ... minimal rule significance (default 1.0)
att_sig ... minimal attribute significance in rule (default 1.0)
max_rule_complexity ... maximum number of conditions in rule (default 5)
min_coverage ... minimal number of covered examples (default 5)
Classifier (LCR) related:
add_sub_rules ... add sub rules ? (default True)
min_cl_sig ... minimal significance of beta in classifier (default 0.5)
min_beta ... minimal beta value (default 0.0)
set_prefix_rules ... should ordered prefix rules be added? (default False)
alternative_learner ... use rule-learner as a correction method for other machine learning methods. (default None)
"""
# argument ID which is passed to abcn2 learner
# two learners, learner with arguments and learners without
self.ruleFinder = orange.RuleBeamFinder()
self.ruleFilter = orange.RuleBeamFilter_Width(width=width)
evdGet = orange.EVDistGetter_Standard()
self.ruleFinder.evaluator = orange.RuleEvaluator_mEVC(
m=m,
evDistGetter=evdGet,
min_improved=min_improved,
min_improved_perc=min_improved_perc)
self.ruleFinder.evaluator.returnExpectedProb = True
self.ruleFinder.evaluator.optimismReduction = opt_reduction
self.ruleFinder.evaluator.ruleAlpha = rule_sig
self.ruleFinder.evaluator.attributeAlpha = att_sig
self.ruleFinder.ruleStoppingValidator = orange.RuleValidator_LRS(
alpha=1.0,
min_quality=0.,
max_rule_complexity=max_rule_complexity - 1)
self.ruleFinder.evaluator.validator = orange.RuleValidator_LRS(
alpha=1.0, min_quality=0., min_coverage=min_coverage)
self.coverAndRemove = CovererAndRemover_Prob()
self.ruleStopping = None
self.dataStopping = orange.RuleDataStoppingCriteria_NoPositives()
self.N = nsampling
self.add_sub_rules = add_sub_rules
self.min_cl_sig = min_cl_sig
self.min_beta = min_beta
self.set_prefix_rules = set_prefix_rules
self.alternativeLearner = alternative_learner
self.__dict__.update(kwds)
