def extend(request):
probabilityTable = bayesian.defaultTokensProbabilityTable()
requestResults = dict()
requestResults.update(csrf(request))
if "string" in request.POST:
testString = request.POST["string"].strip()
probability = bayesian.eventProbabilityFromStringAndTokensProbabilityTable(testString, probabilityTable)
requestResults["text"] = testString
requestResults["probability"] = probability
return render_to_response("amour/extend.html", requestResults)
def training(request):
requestResults = dict()
requestResults.update(csrf(request))
print request.POST
if "train1" in request.POST:
hitList = list()
misList = list()
counter = 1
trainLabel = "train%d" % counter
while trainLabel in request.POST:
if request.POST[trainLabel] == "pass":
pass
elif request.POST[trainLabel] == "hit":
text = request.POST["text%d" % counter]
hitList.append(text)
elif request.POST[trainLabel] == "mis":
text = request.POST["text%d" % counter]
misList.append(text)
counter = counter + 1
trainLabel = "train%d" % counter
bayesian.appendListToDefaultHitFileWithList(hitList)
bayesian.appendListToDefaultMisFileWithList(misList)
requestResults["message"] = "Training Completed"
print "Training Completed"
# client = APIClient(app_key=APP_KEY, app_secret=APP_SECRET, redirect_uri=CALLBACK_URL)
# access_token = request.COOKIES["access_token"]
# expires_in = request.COOKIES["expires_in"]
# client.set_access_token(access_token, expires_in)
#
# publicTimeline = client.get.statuses__public_timeline(count=10)
trainFileName = "trainFile.txt"
testStringList = bayesian.readListFromFileWithFileName(trainFileName)
statusesList = list()
probabilityTable = bayesian.defaultTokensProbabilityTable()
for item in testStringList:
state = dict()
state["text"] = item
state["probability"] = bayesian.eventProbabilityFromStringAndTokensProbabilityTable(item,
probabilityTable)
statusesList.append(state)
statusesList.sort(key=itemgetter("probability"), reverse=True)
requestResults["statusesList"] = statusesList
response = render_to_response("amour/training.html", requestResults)
return response
def analysis(request):
requestResults = dict()
requestResults.update(csrf(request))
realHitCounter = 0
realMisCounter = 0
postFlag = False
#print request.POST
if request.POST:
postFlag = True
counter = 1
trainLabel = "analysis%d" % counter
while trainLabel in request.POST:
if request.POST[trainLabel] == "hit":
realHitCounter += 1
else:
realMisCounter += 1
counter = counter + 1
trainLabel = "analysis%d" % counter
testFileName = "testFile.txt"
testStringList = bayesian.readListFromFileWithFileName(testFileName)
tokensProbabilityTable = bayesian.defaultTokensProbabilityTable()
hitCounter = 0
misCounter = 0
statusesList = list()
for item in testStringList:
probability = bayesian.eventProbabilityFromStringAndTokensProbabilityTable(item, tokensProbabilityTable)
if probability >= 0.9:
hitCounter += 1
else:
misCounter += 1
state = dict()
state["text"] = item
state["probability"] = probability
statusesList.append(state)
#statusesList.sort(key=itemgetter("probability"), reverse=True)
requestResults["totalCounter"] = hitCounter + misCounter
requestResults["postFlag"] = postFlag
requestResults["statusesList"] = statusesList
requestResults["hitCounter"] = hitCounter
requestResults["misCounter"] = misCounter
requestResults["realHitCounter"] = realHitCounter
requestResults["realMisCounter"] = realMisCounter
requestResults["error"] = float(math.fabs(realHitCounter - hitCounter) + math.fabs(realMisCounter - misCounter)) / (
hitCounter + misCounter)
return render_to_response("amour/analysis.html", requestResults)
def single(request):
probabilityTable = bayesian.defaultTokensProbabilityTable()
requestResults = dict()
requestResults.update(csrf(request))
if "string" in request.POST:
testString = request.POST["string"].strip()
probability = bayesian.eventProbabilityFromStringAndTokensProbabilityTable(testString, probabilityTable)
requestResults["text"] = testString
requestResults["probability"] = probability
elif "train" in request.POST:
text = request.POST["text"]
train = request.POST["train"]
tempList = [text]
if train == "hit":
bayesian.appendListToDefaultHitFileWithList(tempList)
elif train == "mis":
bayesian.appendListToDefaultMisFileWithList(tempList)
requestResults["message"] = "Training Completed"
response = render_to_response("amour/single.html", requestResults)
return response
