def calculateWeightage():
global totalSample
global jsonData
global Featureset
csv_filename = "WebcredNormalized.csv"
f = open(csv_filename, 'r')
data = f.readlines()
pipe = Pipeline()
# get json data
jsonData = pipe.converttojson(data)
totalSets = 10
# sets of possible weightages
weightage = []
totalSample = int(
subprocess.check_output(['wc', '-l', csv_filename]).split(' ')[0]) - 1
filterKeys = ['url', 'wot', 'cookie', 'redirected']
FeaturesName = list((set(jsonData[0].keys()) - set(filterKeys)))
Featureset = []
alexaScoreSet = []
for i in range(totalSets):
count = 0
# select sample sets
while True:
sample = getjsonData()
featurevalue, alexaScore, wotScore = getFeatureValue([sample],
FeaturesName)
if checksimiliarData(featurevalue[0]):
Featureset.append(featurevalue[0])
alexaScoreSet.append(alexaScore)
count += 1
if count == len(FeaturesName) - 1:
# sum of all weightage== 1, misc. genre == 0.1
temp = []
for j in range(len(FeaturesName)):
temp.append(1)
Featureset.append(temp)
alexaScoreSet.append([0.9])
break
# get weightage of individual feature
weightage.append(getWeightage(Featureset, alexaScoreSet))
print 'getting', i, 'set of weightages'
finalWeightage = np.mean(weightage, axis=0).tolist()
total = 0
for i in finalWeightage:
total += i
print total
print finalWeightage
print weightage, alexaSimilarityScore
elif action == 'bn':
global totalSample
global jsonData
global Featureset
Featureset = []
alexaScoreSet = []
wotScoreSet = []
csv_filename = "WebcredNormalized.csv"
f = open(csv_filename, 'r')
data = f.readlines()
pipe = Pipeline()
# get json data
jsonData = pipe.converttojson(data)
totalSample = int(
subprocess.check_output(['wc', '-l', csv_filename
]).split(' ')[0]) - 1
filterKeys = ['url', 'wot', 'cookie', 'redirected']
FeaturesName = list((set(jsonData[0].keys()) - set(filterKeys)))
count = 0
tried = 0
# building matrix wiht 1000 samples
while True:
tried += 1
try:
sample = getjsonData()
featurevalue, alexaScore, wotScore = getFeatureValue(
[sample], FeaturesName)
if checksimiliarData(featurevalue[0]):
