def trainUsingNaiveBayes(Data, delim):
#print "In train Func"
NB = NaiveBayes()
##instances = []
NB.totalInstances = len(Data)
count = 1
for sample in Data:
instance = dataInstance(sample[1],sample[0],delim)
NB.trainOnline(instance, count)
count += 1
##instances.append(instance)
#instances.append(dataInstance(sample[1],sample[0]))
#print len(instances[-1].feats)
#print len(instances)
#dummy = raw_input()
#instance.feats = ""
print
#print "Made instances"
#NB.train(instances)
NB.smooth()
return NB
def predictGender(NB, Data, outFile, delim):
status = 0
outFile = open(outFile,'w')
correct = 0
for sample in Data:
instance = dataInstance(sample[1],sample[0],delim)
#print len(instance.feats)
#dummy = raw_input()
outputLabel = NB.classify(instance)
if outputLabel == instance.classLabel:
correct += 1
outFile.write(sample[0]+"\t"+sample[1]+ "\t" + outputLabel +"\n")
status += 1
try:
dummy = 1/(status%1000)
except:
print status,
print
outFile.write(str(round(correct*100.0/len(Data),2)))
outFile.close()
return round(correct*100.0/len(Data),2)