from tools import local_file_util
file = [
line.split('\t')
for line in local_file_util.readFile('data/orgin_train_data.tsv')
]
s = set([line[4] for line in file])
# -*- coding: utf-8 -*
from tools import local_file_util
userComment_train = [
l[0].split(',') for l in [
line.split('\"') for line in local_file_util.readFile(
'bigdata/huangbaoche/huangbaoche_unzip/test/userComment_test.csv')
[1:]
]
]
orderHistory_train = [
line.split(',') for line in local_file_util.readFile(
'bigdata/huangbaoche/huangbaoche_unzip/test/orderHistory_test.csv')[1:]
]
userComment_train_dict = dict([[(line[0], line[1]), [line[2], line[3]]]
for line in userComment_train])
merge_res = []
for orderHistory_train_line in orderHistory_train:
add_line = []
userId_orderId = (orderHistory_train_line[0], orderHistory_train_line[1])
add_line = add_line + orderHistory_train_line
if userId_orderId in userComment_train_dict:
add_line = add_line + userComment_train_dict[userId_orderId]
else:
add_line = add_line + ['', '']
merge_res.append(add_line)
import xgboost as xgb
bst = xgb.Booster({'nthread': 4})
bst.load_model('xgb_model/xgb_v2.model')
import numpy as np
data = np.loadtxt('data/test_data.tsv', delimiter='\t')
test_x = data[:, 1:]
test_y = data[:, 0]
dtest = xgb.DMatrix(test_x, label=test_y)
eval = bst.predict(dtest)
from tools import local_file_util
file = [
line.split(',')[0] for line in local_file_util.readFile(
'bigdata/huangbaoche/huangbaoche_unzip/test/orderFuture_test.csv')[1:]
]
res = [line[0] + ',' + str(line[1]) for line in zip(file, list(eval))]
res.insert(0, 'userid,orderType')
local_file_util.writeFile('data/submit.csv', res)
# -*- coding: utf-8 -*
from tools import local_file_util
file = map(
lambda line: line.split(','),
local_file_util.readFile(
'bigdata/huangbaoche/huangbaoche_unzip/trainingset/action_train.csv')
[1:])
userId_actionTypeList_dict = {}
for line in file:
if line[0] in userId_actionTypeList_dict:
temp = userId_actionTypeList_dict[line[0]]
userId_actionTypeList_dict[line[0]] = temp + [line[2] + ':' + line[1]]
else:
userId_actionTypeList_dict[line[0]] = [line[2] + ':' + line[1]]
save_str = sorted(map(
lambda key: key + '\t' + str(userId_actionTypeList_dict[key].__len__(
)) + '\t' + '\t'.join(userId_actionTypeList_dict[key]),
userId_actionTypeList_dict),
key=lambda line: int(line.split('\t')[1]),
reverse=True)
local_file_util.writeFile('data/userId_actionTypeNum.tsv', save_str)
#userId actionNum(sort) time:actiontye time2:actiontype
# -*- coding: utf-8 -*
from tools import local_file_util
file = map(lambda line: line.split('\t'),
local_file_util.readFile('data/user_orderNum.tsv'))
orderNum_userId_dic = {}
for userId_orderNum in file:
if userId_orderNum[1] in orderNum_userId_dic:
temp = orderNum_userId_dic[userId_orderNum[1]] + [userId_orderNum[0]]
orderNum_userId_dic[userId_orderNum[1]] = temp
else:
orderNum_userId_dic[userId_orderNum[1]] = [userId_orderNum[0]]
orderNum_userIdNum = sorted(map(
lambda key: (key, orderNum_userId_dic[key].__len__()),
orderNum_userId_dic),
key=lambda tuple: tuple[1],
reverse=True)
save_str = map(lambda line: line[0] + '\t' + str(line[1]), orderNum_userIdNum)
local_file_util.writeFile('data/orderNum_userIdNum.tsv', save_str)
# -*- coding: utf-8 -*
#统计每个用户的订单数量
from tools import local_file_util
file = map(
lambda line: line.split(','),
local_file_util.readFile(
'bigdata/huangbaoche/huangbaoche_unzip/trainingset/orderHistory_train.csv'
)[1:])
userId_orderId_list = map(lambda line: (line[0], line[1]), file)
dic = {}
#userid: list(orderid1, oderid2)
for userId_orderId in userId_orderId_list:
if userId_orderId[0] in dic:
temp = dic[userId_orderId[0]] + [userId_orderId[1]]
dic[userId_orderId[0]] = temp
else:
dic[userId_orderId[0]] = [userId_orderId[1]]
userId_oderNum = sorted(map(lambda key: (key, dic[key].__len__()), dic),
key=lambda tuple: tuple[1],
reverse=True)
res_save_str = map(lambda line: line[0] + '\t' + str(line[1]), userId_oderNum)
local_file_util.writeFile('data/user_orderNum.tsv', res_save_str)
# -*- coding: utf-8 -*
#how to make test: 1: origin_train_line = ['-1', user_id] 2: trainningset to test 3:change save file
from tools import local_file_util
from itertools import groupby
import numpy as np
orderFuture_train = [
line.split(',') for line in local_file_util.readFile(
'bigdata/huangbaoche/huangbaoche_unzip/test/orderFuture_test.csv')[1:]
]
action_train = [
line.split(',') for line in local_file_util.readFile(
'bigdata/huangbaoche/huangbaoche_unzip/test/action_test.csv')[1:]
]
orderHistory_comment_train = [
line.split(',') for line in local_file_util.readFile(
'data/merge_orderHistory_userComment_test.csv')
]
userProfile_train = [
line.split(',') for line in local_file_util.readFile(
'bigdata/huangbaoche/huangbaoche_unzip/test/userProfile_test.csv')[1:]
]
continent_rate_dict = {
'大洋洲': 1.4 / 0.33,
'欧洲': 19.1 / 7.25,
'非洲': 2.2 / 8.69,
# -*- coding: utf-8 -*
#统计每个用户的订单数量
from tools import local_file_util
file =map(lambda line: line.split(','), local_file_util.readFile('bigdata/皇包车比赛/皇包车比赛数据-非压缩包/trainingset/orderHistory_train.csv')[1:])
userId_orderId_list = map(lambda line:(line[0], line[1]), file)
dic = {}
#userid: list(orderid1, oderid2)
for userId_orderId in userId_orderId_list:
if userId_orderId[0] in dic:
temp = dic[userId_orderId[0]]+ [userId_orderId[1]]
dic[userId_orderId[0]] = temp
else:
dic[userId_orderId[0]] = [userId_orderId[1]]
userId_oderNum = sorted(map(lambda key: (key, dic[key].__len__()), dic), key=lambda tuple: tuple[1], reverse=True)
res_save_str = map(lambda line: line[0] + '\t' + str(line[1]), userId_oderNum)
local_file_util.writeFile('data/user_orderNum.tsv', res_save_str)