def _makeInstance(self, i, instances=None): inst = Instance(len(i)) if instances is not None: inst.setDataset(instances) for (attName, value) in i.iteritems(): if attName in self.numericAttributes: value = Double(value) else: value = String(value) attr = self.attName2Obj[attName] #print self.attName2Domain #print "attName, value", attName, value inst.setValue(attr, value) return inst
def build_instances(state,dataset):
class_attributes = ["Sunny", "Fog", "Rain", "Snow", "Hail", "Thunder", "Tornado"]
header = ["state","lat", "lon", "day","temp","dewp","weather"]
#build attributes based on the header and types
attributes = []
for h in header[:-1]:
attributes.append(Attribute(h))
#add the classification attribute
classification_vector = FastVector(len(class_attributes))
for c in class_attributes:
classification_vector.addElement(c)
attributes.append(Attribute("toClassify", classification_vector))
fvWekaAttributes = FastVector(len(dataset[0]))
for a in attributes:
fvWekaAttributes.addElement(a)
training_set = Instances("C4.5Set", fvWekaAttributes, len(dataset))
training_set.setClassIndex(len(header)-1)
for d in dataset:
inst = Instance(len(d))
for i in range(len(d)-1):
try:
inst.setValue(fvWekaAttributes.elementAt(i), float(d[i]))
except:
pass
#print "failed on", i, d[i], d[i].__class__
inst.setValue(fvWekaAttributes.elementAt(len(d)-1), d[-1])
training_set.add(inst)
j48 = J48()
j48.buildClassifier(training_set)
return state,parse_tree(str(j48))
def _makeInstance(self, i, instances=None):
inst = Instance(len(i))
if instances is not None:
inst.setDataset(instances)
for (attName, value) in i.iteritems():
if attName in self.numericAttributes: value = Double(value)
else: value = String(value)
attr = self.attName2Obj[attName]
#print self.attName2Domain
#print "attName, value", attName, value
inst.setValue(attr, value)
return inst
class JeroR(Classifier, JythonSerializableObject):
"""
JeroR is a Jython implementation of the Weka classifier ZeroR
'author' -- FracPete (fracpete at waikato dot ac dot nz)
'version' -- $Revision$
"""
# the documentation can be generated with HappyDoc:
# http://happydoc.sourceforge.net/
# Example command:
# happydoc --title Weka -d ./doc ./src
# the chosen class value
__ClassValue = Instance.missingValue()
# the class attribute
__Class = None
# the counts for each class label
__Counts = None
def listOptions(self):
"""
Returns an enumeration describing the available options.
Return:
an enumeration of all the available options.
"""
return Classifier.listOptions(self)
def setOptions(self, options):
"""
Parses a given list of options.
Parameter(s):
'options' -- the list of options as an array of strings
"""
Classifier.setOptions(self, options)
return
def getOptions(self):
"""
Gets the current settings of the Classifier as string array.
Return:
an array of strings suitable for passing to setOptions
"""
return Classifier.getOptions(self)
def getCapabilities(self):
"""
returns the capabilities of this classifier
Return:
the capabilities of this classifier
"""
result = Classifier.getCapabilities(self)
# attributes
result.enable(Capability.NOMINAL_ATTRIBUTES)
result.enable(Capability.NUMERIC_ATTRIBUTES)
result.enable(Capability.DATE_ATTRIBUTES)
result.enable(Capability.STRING_ATTRIBUTES)
result.enable(Capability.RELATIONAL_ATTRIBUTES)
result.enable(Capability.MISSING_VALUES)
# class
result.enable(Capability.NOMINAL_CLASS)
result.enable(Capability.NUMERIC_CLASS)
result.enable(Capability.DATE_CLASS)
result.enable(Capability.MISSING_CLASS_VALUES)
# instances
result.setMinimumNumberInstances(0)
return result
def buildClassifier(self, instances):
"""
builds the ZeroR classifier with the given data
Parameter(s):
'instances' -- the data to build the classifier from
"""
self.getCapabilities().testWithFail(instances)
# remove instances with missing class
instances = Instances(instances)
instances.deleteWithMissingClass()
sumOfWeights = 0
self.__Class = instances.classAttribute()
self.__ClassValue = 0
self.__Counts = None
if (instances.classAttribute().isNumeric()):
self.__Counts = None
elif (instances.classAttribute().isNominal()):
self.__Counts = jarray.zeros(instances.numClasses(), 'd')
for i in range(len(self.__Counts)):
self.__Counts[i] = 1
sumOfWeights = instances.numClasses()
enu = instances.enumerateInstances()
while (enu.hasMoreElements()):
instance = enu.nextElement()
if (not instance.classIsMissing()):
if (instances.classAttribute().isNominal()):
self.__Counts[int(
instance.classValue())] += instance.weight()
else:
self.__ClassValue += instance.weight(
) * instance.classValue()
sumOfWeights += instance.weight()
if (instances.classAttribute().isNumeric()):
if (Utils.gr(sumOfWeights, 0)):
self.__ClassValue /= sumOfWeights
else:
self.__ClassValue = Utils.maxIndex(self.__Counts)
Utils.normalize(self.__Counts, sumOfWeights)
return
def classifyInstance(self, instance):
"""
returns the prediction for the given instance
Parameter(s):
'instance' -- the instance to predict the class value for
Return:
the prediction for the given instance
"""
return self.__ClassValue
def distributionForInstance(self, instance):
"""
returns the class distribution for the given instance
Parameter(s):
'instance' -- the instance to calculate the class distribution for
Return:
the class distribution for the given instance
"""
result = None
if (self.__Counts == None):
result = jarray.zeros(1, 'd')
result[0] = self.__ClassValue
else:
result = self.__Counts[:]
return result
def toString(self):
"""
Prints a string representation of the classifier
Return:
string representation of the classifier
"""
if (self.__Class == None):
return "JeroR: No model built yet."
if (self.__Counts == None):
return "JeroR predicts class value: " + str(self.__ClassValue)
else:
return "JeroR predicts class value: " + str(
self.__Class.value(int(self.__ClassValue)))
def processThread(self, tempCursor, postChain, data, dataNum, j48, linreg): if len(postChain) < 2: return # return if there are no replies threadSoFar = [] self.threadWrapper_add(threadSoFar, postChain[0], postChain[1]) preprocessCorpus([item[3] for item in postChain]) dbHandler.insert(tempCursor, postChain[0][0], postChain[1][0]) for i in xrange(2, len(postChain)): child = [postChain[i][0], []] hiLinreg = [[-1, -1.0], [-1, -1.0]] # [postid, linregValue] posJ48 = [-1, 0] # [postid, sum of yields] for j in xrange(0, i): self.threadWrapper_add(threadSoFar, postChain[j], postChain[i]) time, cos, quoted, newestparent, dist, pos, lastown, diaLen, diaDist, prevsparent, firstpost = self.computeFeatures(threadSoFar) threadSoFar = threadSoFar[:-1] postPair = Instance(data.numAttributes()) postPair.setValue(0, time) postPair.setValue(1, cos) postPair.setValue(2, quoted) postPair.setValue(3, newestparent) postPair.setValue(4, dist) postPair.setValue(5, pos) postPair.setValue(6, data.attribute(6).indexOfValue(lastown)) postPair.setValue(7, diaLen) postPair.setValue(8, diaDist) postPair.setValue(9, prevsparent) postPair.setValue(10, firstpost) postPair.setDataset(dataNum) linregValue = linreg.classifyInstance(postPair) if linregValue > hiLinreg[0][1]: hiLinreg[1] = hiLinreg[0] hiLinreg[0] = [postChain[j][0], linregValue] else: if linregValue > hiLinreg[1][1]: hiLinreg[1] = [postChain[j][0], linregValue] postPair.setDataset(data) j48Class = j48.classifyInstance(postPair) if j48Class == 1.0: posJ48 = [postChain[j][0], posJ48[1]+1] if child[1] == []: if posJ48[1] == 0: child[1] = [hiLinreg[1][0]] else: if posJ48[1] > 1: child[1] = [hiLinreg[0][0]] else: child[1] = [posJ48[0]] self.threadWrapper_add(threadSoFar, [item for item in postChain if item[0] == child[1][0]][0], postChain[i]) dbHandler.insert(tempCursor, child[1][0], child[0])