def get_data(ID):
data = crawler.fetch_item(ID)
if data['condition']== u'new':
condition = 1.0
else: condition = 0.0
data = [data['title'], float(data['init_price'])**(0.5), float(data['seller_point']), condition]
return data
def predict_KN(ID):
testee = crawler.fetch_item(ID)
maker = search_maker(int(testee['category_id']))
if maker == 1: return 'Error'
result = dic_KN[maker].predict(ID)
cu = result[1].sort('current_price')
return (result[0], cu.ix[cu.index[2], 'current_price'], cu.ix[cu.index[7], 'current_price'])
def predict(self,ID,threhold = 0.0,rate=2):
testee = fetch_item(ID)
try:
maker = self.categoryID[int(testee['category_id'])]
return self.Auction_model[maker].predict(testee['title'],testee['description'],threhold = threhold,rate=rate)
except KeyError as e:
print e
raise NameError('No category_id found')
def predict(self, ID, threhold=0.0, rate=2):
testee = fetch_item(ID)
if self.categoryID[int(testee['category_id'])] not in self.makers:
raise NameError('No category_id found')
else:
return self.Auction.predict(testee['title'],
testee['description'],
threhold=threhold,
rate=rate)
def predict_KN(ID):
testee = crawler.fetch_item(ID)
maker = search_maker(int(testee['category_id']))
if maker == 1: return 'Error'
result = dic_KN[maker].predict(ID)
cu = result[1].sort('current_price')
return (result[0], cu.ix[cu.index[2],
'current_price'], cu.ix[cu.index[7],
'current_price'])
def predict(self, ID, threhold=0.0, rate=2):
testee = fetch_item(ID)
try:
maker = self.categoryID[int(testee['category_id'])]
return self.Auction_model[maker].predict(testee['title'],
testee['description'],
threhold=threhold,
rate=rate)
except KeyError as e:
print e
raise NameError('No category_id found')
def get_data(ID):
data = crawler.fetch_item(ID)
if data['condition'] == u'new':
condition = 1.0
else:
condition = 0.0
data = [
data['title'],
float(data['init_price'])**(0.5),
float(data['seller_point']), condition
]
return data
def predict(self,ID,threhold = 0.0,rate=1):
testee = crawler.fetch_item(ID)
maker = self.search_maker(int(testee['category_id']))
if maker == None:
raise NameError('No category_id found')
title_similarity = self.setence_to_similarity(testee['title'],self.model_dict[maker].title_lda)
description_similarity = self.setence_to_similarity(testee['title'],self.model_dict[maker].description_lda)
sim = [(n,(rate+1)*s1*s2/(s1+rate*s2)) for ((n,s1),s2) in zip(enumerate(title_similarity),description_similarity) if (s1 !=0 and s2 != 0)]
print sim
p_list = sorted([self.model_dict[maker].price[n] for (n,x) in sim if x > threhold ])
l=len(p_list)
if l == 0:
return (None,[])
else:
return (p_list[l/2],p_list[l/4],p_list[l*3/4])
def predict_LR2(ID):
testee = crawler.fetch_item(ID)
maker = search_maker(int(testee['category_id']))
if maker == 1: return 'Error'
return dic_LR2[maker].predict(ID)
def predict(self, ID, threhold=0.0, rate=2):
testee = fetch_item(ID)
if self.categoryID[int(testee["category_id"])] not in self.makers:
raise NameError("No category_id found")
else:
return self.Auction.predict(testee["title"], testee["description"], threhold=threhold, rate=rate)
def predict_LR2(ID):
testee = crawler.fetch_item(ID)
maker = search_maker(int(testee['category_id']))
if maker == 1: return 'Error'
return dic_LR2[maker].predict(ID)
def predict(self,ID,threhold = 0.0,rate=2):
testee = fetch_item(ID)
if self.categoryID[int(testee['category_id'])] not in self.makers:
raise NameError('No category_id found')
else:
return self.Auction.predict(testee['title'],testee['description'],threhold = threhold,rate=rate)
'2084307168',
'2084193602',
'2084307178',
'2084193594',
'2084307171',
'2084307189',
'2084193603',
'2084307177',
'2084193585',
'2084307180',
'2084042163',
'2084307169',
'2084048237',
'2084307192',
'2084307167',
'2084307187',
'2084307172'
]
random.shuffle(ctgs)
for c in ctgs:
print 'start fething items in category %s... ' % c
for i in range(3):
ids = crawler.fetch_item_list(c, i)
for j in ids:
result = crawler.fetch_item(j)
archiver.add_item(result)
print 'page', i, 'count', len(ids)
print 'end'
