def maketeams(request):
"""
Actually adds the team that was requested by the addteams page
"""
#fh = open("log.txt")
#fh.write("I got to step 1 \n")
if auth(request):
parse.getData(request.POST['url']) # Sanitize your inputs
#fh.write("I got to step 2 \n")
t = Team.objects.get(link=request.POST['url'])
#fh.write("I got to step 3 \n")
return redirect('/admin/')
else:
return redirect('/login/')
def main():
decision_lists = {}
trainingFile = '/data/cs65/senseval3/train/EnglishLS.train'
trainData = getData(trainingFile)
testFile = '/data/cs65/senseval3/test/EnglishLS.test'
testData = getData(testFile)
k = 10
for word in trainData.keys():
decision_lists[word] = build_decision_list(trainData, word, k)
total = 0
correct = 0
correct_less = 0
correct_cutoff = 0
for word in testData.keys():
MFS = most_frequent_sense(trainData, word)
modified_declist = [x for x in decision_lists[word] if x[2] > 1.0]
for instance in testData[word].keys():
total += 1
instanceData = testData[word][instance]
classification = classify(instanceData, decision_lists[word], k, MFS)
less_rules = classify(instanceData, decision_lists[word][:100], k, MFS)
cutoff = classify(instanceData, modified_declist, k, MFS)
if classification in instanceData['answers']:
correct += 1
if less_rules in instanceData['answers']:
correct_less += 1
if cutoff in instanceData['answers']:
correct_cutoff += 1
accuracy = float(correct) / float(total)
accuracy_less = float(correct_less) / float(total)
accuracy_cutoff = float(correct_cutoff) / float(total)
print "Accurately classified %f of all words" % (accuracy)
print "%d correct of %d total" % (correct, total)
print "With 100 rules, accurately classified %f of all words" % (accuracy_less)
print "%d correct of %d total" % (correct_less, total)
print "With cutoff at 1.0, Accurately classified %f of all words" % (accuracy_cutoff)
print "%d correct of %d total" % (correct_cutoff, total)
"""
12. Got 0.593 accuracy, slightly better than the 0.571 accuracy
of the MFS baseline.
14. The modified classifications perform approximately the same; see
output
"""
# print decision_lists['organization.n'][:100]
"""
def classify(train, test): trainData, testData = getData(train, test); #print len(trainData[0]), len(testData[0]) #print len(trainData[0][0]), len(trainData[1]), len(testData[0][0]), len(testData[1]) #annClass = learnANN(trainData[0], trainData[1]); #annRes = annClass.predict(testData[0]); #annAcc = 0; bayesAcc = 0; svmAcc = 0; bayesClass = learnBayes(trainData[0], trainData[1]); bayesRes = bayesClass.predict(testData[0]); bayesRes = map(lambda x: 0 if x < 0.5 else 1, bayesRes) svmClass = learnSVM(trainData[0], trainData[1]); svmRes = svmClass.predict(testData[0]); svmRes = map(lambda x: 0 if x < 0.5 else 1, svmRes) for i in xrange(len(testData[1])): #if annRes[i] == testData[1][i]: # annAcc += 1 if bayesRes[i] == testData[1][i]: bayesAcc += 1 if svmRes[i] == testData[1][i]: svmAcc += 1 #print "ANN Accuracy:", annAcc/(len(testData[1])*1.0); print "Bayes Accuracy:", bayesAcc/(len(testData[1])*1.0); print "SVM Accuracy:", svmAcc/(len(testData[1])*1.0);
def updateDB():
data = getData('data.pickle')
with open('data.csv', 'w') as csvFile:
writer = csv.writer(csvFile)
writer.writerow(['Company', 'Keyword', 'Salience'])
for company in data:
for pair in data[company]:
writer.writerow([company, pair[0], pair[1]])
def buildBook(filename): if not os.access(filename, os.R_OK): print filename, 'is not exist' return basename = os.path.basename(filename) WORKDIR = basename try: data = parse.getData(filename) TITLE, AUTHOR, SECTIONS = parse.parse(data) except ParseError, e: print 'parse error', filename, e return
for rule in decList[key]:
if rule[1][1] in words:
return rule[1][0]
if rule[0] < 0:
break
tsenses = getSenses(testData, key)
tsenseFreq = map(lambda x: (x, tsenses.count(x)), set(tsenses))
mfs = freqSense(tsenseFreq)[0]
return mfs
if __name__=='__main__':
trainingFile = '/data/cs65/senseval3/train/EnglishLS.train'
data = getData(trainingFile)
testingFile = '/data/cs65/senseval3/test/EnglishLS.test'
testData = getData(testingFile)
k = 10
mfs = {} #values are tuples of mfs, count of mfs
decList = {}
for key in data.keys():
senses = list(set(getSenses(data, key)))
#below is list of frequencies
senseFreq = map(lambda x: (x, senses.count(x)), set(senses))
mfs[key] = freqSense(senseFreq)[0]
scores = []
counts = {}
for sense in senses:
senseInst = instanceSense(data, key, sense)
counts[sense] = countf(data, key, k, senseInst)
def doTest():
damage_data = parse.getData()
print("#### DAMAGE TEST ####")
if damage_data["jigglypuff"]["Jab1"] == 3.0:
print("jiggylypuff Jab1 -- PASS")
else:
print("jiggylypuff Jab1 -- FAIL")
if damage_data["robin"]["U-smash"] == 15.0:
print("robin U-smash -- PASS")
else:
print("robin U-smash -- FAIL")
if damage_data["pikachu"]["Nair"] == 8.5:
print("pikachu Nair -- PASS")
else:
print("pikachu Nair -- FAIL")
if damage_data["marth"]["Final Smash"] == 60.0:
print("marth Final Smash -- PASS")
else:
print("marth Final Smash -- FAIL")
if damage_data["metaknight"]["U-throw"] == 10.0:
print("metaknight U-throw -- PASS")
else:
print("metaknight U-throw -- FAIL")
if damage_data["falco"]["Reflector (ground)"] == 5.0:
print("falco Reflector (ground) -- PASS")
else:
print("falco Reflector (ground) -- FAIL")
if damage_data["villager"]["Timber (axe)"] == 14.0:
print("village Timber (axe) -- PASS")
else:
print("village Timber (axe) -- FAIL")
if damage_data["gamewatch"]["F-smash (normal)"] == 18.0:
print("gamewatch F-smash (normal) -- PASS")
else:
print("gamewatch F-smash (normal) -- FAIL")
if damage_data["wiifit"]["Header (head spike)"] == 15.0:
print("wiifit Header (head spike) -- PASS")
else:
print("wiifit Header (head spike) -- FAIL")
if damage_data["zelda"]["Phantom Strike (fully charged uppercut slash)"] == 12.0:
print("zelda Phantom Strike (fully charged uppercut slash) -- PASS")
else:
print("zelda Phantom Strike (fully charged uppercut slash) -- FAIL")
print("#### DAMAGE TEST END ####")
import sys
import config
import parse
import nn
if len(sys.argv) < 2:
sys.exit('Usage: %s directory-name' % sys.argv[0])
d = sys.argv[1]
#try:
translate = parse.buildTranslate(d)
data = parse.getData(d, "training", True)
data['translate'] = translate
nn = nn.neuralNetwork( data )
#print nn
#data = parse.getData(d, "test")
#for row in data['inputs']:
# print nn.predict( row )
#except Exception as error:
# print error
#! coding: UTF-8
import numpy as np
import parse as Parser
import kmeans as km
list = [
"北海道", "青森県", "岩手県", "宮城県", "秋田県", "山形県", "福島県", "茨城県", "栃木県", "群馬県",
"埼玉県", "千葉県", "東京", "神奈川県", "新潟県", "山梨県", "長野県", "富山県", "石川県", "福井県",
"岐阜県", "静岡県", "愛知県", "三重県", "滋賀県", "京都府", "大阪府", "兵庫県", "奈良県", "和歌山県",
"鳥取県", "島根県", "岡山県", "広島県", "山口県", "徳島県", "香川県", "愛媛県", "高知県", "福岡県",
"佐賀県", "長崎県", "熊本県", "大分県", "宮崎県", "鹿児島県", "沖縄"
]
print("クラスタ数を入力してください")
cluster = int(input())
result = km.kmeans(Parser.getData(), cluster)
for cluster_num in range(cluster):
cluster_list = []
for list_num in range(len(list)):
if result[list_num] == cluster_num:
cluster_list.append(list[list_num])
print("クラスタ" + str(cluster_num))
print(cluster_list)
print("--------------")
import sys
import config
import parse
import nn
if len(sys.argv) < 2:
sys.exit('Usage: %s directory-name' % sys.argv[0])
d = sys.argv[1]
#try:
translate = parse.buildTranslate(d)
data = parse.getData(d, "training", True)
data['translate'] = translate
nn = nn.neuralNetwork(data)
#print nn
#data = parse.getData(d, "test")
#for row in data['inputs']:
# print nn.predict( row )
#except Exception as error:
# print error
trainSenses = retrieveMostCommonSenses(trainData)
lexelts = testData.keys()
for lexelt in lexelts:
instances = testData[lexelt].keys()
for instance in instances:
senses = testData[lexelt][instance]["answers"]
if trainSenses[lexelt][0] in senses:
totalCorrect += 1
totalGuesses += 1
print float(totalCorrect)/totalGuesses * 100
if __name__=='__main__':
trainingFile = '/data/cs65/senseval3/train/EnglishLS.train'
trainData = getData(trainingFile)
testingFile = '/data/cs65/senseval3/test/EnglishLS.test'
testData = getData(testingFile)
# questions on training data
print "\nQuestion 1:\n"
question1(trainData)
print "\nQuestion 2:\n"
question2(trainData)
print "\nQuestion 3:\n"
question3(trainData)
print "\nQuestion 4:\n"
question4(trainData)
print "\nQuestion 5:\n"
question5(trainData)
print "\nQuestion 6:\n"
