def createData(self, output, numDataPoints):
allFeatures = getMergedFeatures(
"C:/MissingWord/mergedFeatures" + str(self.sentenceLength) +
".txt", numDataPoints)
print("Loaded Features")
golds = loadGolds("C:/MissingWord/" + str(self.sentenceLength) +
"Gold.txt",
numGolds=len(allFeatures),
length=self.sentenceLength)
print(len(allFeatures))
sentences = []
for i in range(len(allFeatures)):
sentences.append(Sentence(golds[i], allFeatures[i]))
appendLabels(str(self.sentenceLength) + "rf15.txt", sentences)
appendLabels(str(self.sentenceLength) + "rf110.txt", sentences)
data = [(sentence.getAllFeatures(),
sentence.getGold().getRemovedIndex())
for sentence in sentences]
with open(output, "wb") as f:
pickle.dump(data, f)
return data
def makeData(self):
allFeatures = getMergedFeatures(
"C:/MissingWord/mergedFeatures" + str(self.sentenceLength) +
".txt", self.numData)
golds = loadGolds("C:/MissingWord/" + str(self.sentenceLength) +
"Gold.txt",
numGolds=len(allFeatures),
length=self.sentenceLength)
def createData(self, output, numDataPoints):
allFeatures = getMergedFeatures("C:/MissingWord/mergedFeatures"+str(self.sentenceLength)+".txt", numDataPoints)
print("Loaded Features")
golds = loadGolds("C:/MissingWord/"+str(self.sentenceLength)+"Gold.txt", numGolds = len(allFeatures), length = self.sentenceLength)
print(len(allFeatures))
sentences = []
for i in range(len(allFeatures)):
sentences.append(Sentence(golds[i], allFeatures[i]))
appendLabels(str(self.sentenceLength)+"rf15.txt", sentences)
appendLabels(str(self.sentenceLength)+"rf110.txt", sentences)
data = [(sentence.getAllFeatures(), sentence.getGold().getRemovedIndex()) for sentence in sentences]
with open(output, "wb") as f:
pickle.dump(data, f)
return data
testFeatures = [datum[0] for datum in data[cutoff:]]
testLabels = [datum[1] for datum in data[cutoff:]]
clf = RandomForestClassifier(n_estimators = 500, n_jobs = 7)
clf.fit(trainFeatures, trainLabels)
trainingPred = clf.predict(trainFeatures)
print(classification_report(trainLabels, trainingPred))
print(classification_report(testLabels, clf.predict(testFeatures)))
with open(modelSystem.nPosModelFile(self.sentenceLength), "wb") as f:
pickle.dump(clf, f)
def test(self, predictions, predictionsMade, golds):
print("Total Sentences "+str(len(predictions)))
toVerifyPrediction = []
toVerifyLabel = []
for predictionIndex in predictionsMade:
toVerifyPrediction.append(predictions[predictionIndex])
toVerifyLabel.append(golds[predictionIndex].getRemovedIndex())
print("Predicted "+str(len(toVerifyPrediction)))
print(classification_report(toVerifyPrediction, toVerifyLabel))
if __name__ == "__main__":
sentenceLength = 16
model = NPositionModel(sentenceLength)
predictions, predictionsMade = model.predict(bridge.loadMergedFeatures(sentenceLength, start = 100000, end = 110000))
model.test(predictions, predictionsMade, gold.loadGolds("C:/MissingWord/"+str(sentenceLength)+"Gold.txt", numGolds = 110000)[100000:110000])
def makeData(self):
allFeatures = getMergedFeatures("C:/MissingWord/mergedFeatures"+str(self.sentenceLength)+".txt", self.numData)
golds = loadGolds("C:/MissingWord/"+str(self.sentenceLength)+"Gold.txt", numGolds = len(allFeatures), length = self.sentenceLength)
print(classification_report(self.clf.predict(trainFeatures), trainLabels))
print(classification_report(self.clf.predict(testFeatures), testLabels))
def predict(self, features):
return self.clf.predict_proba(features)
def chunks(l, n):
for i in range(0, len(l), n):
yield l[i:i+n]
def export(self, filename):
if self.clf is not None:
with open(filename, "wb") as f:
pickle.dump(self.clf, f)
if __name__ == "__main__":
#output(15)
pms = PositionModelStructure(15, [5, 10])
'''
pms.trainIntermediateModels()
'''
#pms.trainPredictiveModel()
frameFeatures = bridge.loadMergedFeatures(15, 100000, 110000)
sentenceFeatures = modelSystem.rawToSentenceFeatures(frameFeatures)
pms.predict(sentenceFeatures)
pms.validatePrediction(gold.loadGolds("C:/MissingWord/"+str(15)+"Gold.txt", numGolds = 110000)[100000:110000])
for line in f:
chunks = parseSuperList(line.strip())
labelSet = []
for elem in chunks:
labelSet.append(round(float(parseArrayList(elem)[1])))
frameLabels.append(labelSet)
condensedFrameLabels = []
for label in frameLabels:
if 1 in label:
condensedFrameLabels.append(label.index(1))
else:
condensedFrameLabels.append(0)
golds = gold.loadGolds("C:/MissingWord/"+str(sentenceLength)+"gold.txt", numGolds = 100000)
goldLabels = [gold.getRemovedIndex() for gold in golds]
allPredictions = []
with open("C:/MissingWord/frame/15-10Combined.txt", "r") as f:
for line in f:
if len(line) > 0:
line = line.replace("array(", "")
line = line.replace(")", "")
line = line.strip()[1: -1]
predictionSet = []
for elem in line.split("], ["):
elem = elem.replace("]", "")
elem = elem.replace("[", "")
elems = [float(e) for e in elem.split(", ")]
if len(elems) == 1 or elems[0] > 0.5:
clf = RandomForestClassifier(n_estimators=500, n_jobs=7)
clf.fit(trainFeatures, trainLabels)
trainingPred = clf.predict(trainFeatures)
print(classification_report(trainLabels, trainingPred))
print(classification_report(testLabels, clf.predict(testFeatures)))
with open(modelSystem.nPosModelFile(self.sentenceLength), "wb") as f:
pickle.dump(clf, f)
def test(self, predictions, predictionsMade, golds):
print("Total Sentences " + str(len(predictions)))
toVerifyPrediction = []
toVerifyLabel = []
for predictionIndex in predictionsMade:
toVerifyPrediction.append(predictions[predictionIndex])
toVerifyLabel.append(golds[predictionIndex].getRemovedIndex())
print("Predicted " + str(len(toVerifyPrediction)))
print(classification_report(toVerifyPrediction, toVerifyLabel))
if __name__ == "__main__":
sentenceLength = 16
model = NPositionModel(sentenceLength)
predictions, predictionsMade = model.predict(
bridge.loadMergedFeatures(sentenceLength, start=100000, end=110000))
model.test(
predictions, predictionsMade,
gold.loadGolds("C:/MissingWord/" + str(sentenceLength) + "Gold.txt",
numGolds=110000)[100000:110000])