本文内容
最近看《写给程序员的数据挖掘指南》，研究推荐算法，书中的测试数据集是 Book-Crossing Dataset 提供的亚马逊用户对书籍评分的真实数据。推荐大家看本书，写得不错，立刻就能对推荐算法上手，甚至应用到你的项目中。
Book-Crossing Dataset 提供两种格式的数据集：CVS 格式和 SQL dump，问题是：
如果你有 UE 打开 cvs 文件，有乱码。无论如何转换编码，都不行~因为，这个文件是亚马逊通过程序持久化后，再导出来的。你还会发现，文件中有 html 标记，另外，关于用户名，书名等等信息，基本都是德文的（看域名就知道了）~
虽然，作者提供了加载测试数据集的 python 代码，不过不能导入到 MySQL 数据库中，其中，作者只是简单地按分号来分割字段内容（虽然推荐算法并不需要全部字段），可数据集中包含类似“ऩ”或“\“”这样的字符，不可能导入到 MySQL 数据库中~
你也许会问，作者都不导入到数据库，你为什么要导？因为，作者提供的推荐算法属于内存模型，也就是一次性把数据加载到内存，但之前，总还是要持久化吧~
因此，只能改造一下作者的 Python 代码~
Github Demo
改造后测试数据集
Python
# -*- coding: utf-8 -*-
import mysql.connector
import codecs
import string
import os
import sys
import ConfigParser
from collections import OrderedDict
import re
class MysqlPythonFacotry(object):
"""
Python Class for connecting with MySQL server.
"""
__instance = None
__host = None
__user = None
__password = None
__database = None
__session = None
__connection = None
def __init__(self, host='localhost', user='root', password='', database=''):
self.__host = host
self.__user = user
self.__password = password
self.__database = database
## End def __init__
def open(self):
try:
cnx = mysql.connector.connect(host=self.__host,\
user= self.__user,\
password= self.__password,\
database= self.__database)
self.__connection = cnx
self.__session = cnx.cursor()
except mysql.connector.Error as e:
print('connect fails!{}'.format(e))
## End def open
def close(self):
self.__session.close()
self.__connection.close()
## End def close
def select(self, table, where=None, *args, **kwargs):
result = None
query = 'SELECT '
keys = args
values = tuple(kwargs.values())
l = len(keys) - 1
for i, key in enumerate(keys):
query += "`" + key + "`"
if i <; l:
query += ","
## End for keys
query += 'FROM %s' % table
if where:
query += " WHERE %s" % where
## End if where
self.__session.execute(query, values)
number_rows = self.__session.rowcount
number_columns = len(self.__session.description)
result = self.__session.fetchall()
return result
## End def select
def update(self, table, where=None, *args, **kwargs):
try:
query = "UPDATE %s SET " % table
keys = kwargs.keys()
values = tuple(kwargs.values()) + tuple(args)
l = len(keys) - 1
for i, key in enumerate(keys):
query += "`" + key + "` = %s"
if i <; l:
query += ","
## End if i less than 1
## End for keys
query += " WHERE %s" % where
self.__session.execute(query, values)
self.__connection.commit()
# Obtain rows affected
update_rows = self.__session.rowcount
except mysql.connector.Error as e:
print(e.value)
return update_rows
## End function update
def insert(self, table, *args, **kwargs):
values = None
query = "INSERT INTO %s " % table
if kwargs:
keys = kwargs.keys()
values = tuple(kwargs.values())
query += "(" + ",".join(["`%s`"] * len(keys)) % tuple(keys) + ") VALUES (" + ",".join(["%s"] * len(values)) + ")"
elif args:
values = args
query += " VALUES(" + ",".join(["%s"] * len(values)) + ")"
self.__session.execute(query, values)
self.__connection.commit()
cnt = self.__session.rowcount
return cnt
## End def insert
def delete(self, table, where=None, *args):
query = "DELETE FROM %s" % table
if where:
query += ' WHERE %s' % where
values = tuple(args)
self.__session.execute(query, values)
self.__connection.commit()
delete_rows = self.__session.rowcount
return delete_rows
## End def delete
def select_advanced(self, sql, *args):
od = OrderedDict(args)
query = sql
values = tuple(od.values())
self.__session.execute(query, values)
number_rows = self.__session.rowcount
number_columns = len(self.__session.description)
result = self.__session.fetchall()
return result
## End def select_advanced
## End class
class ErrorMyProgram(Exception):
"""
My Exception Error Class
"""
def __init__(self, value):
self.value = value
##End def __init__
def __str__(self):
return repr(self.value)
##End def __str__
## End class ErrorMyProgram
class LoadAppConf(object):
"""
Load app.conf Config File Class
"""
__configFileName = "app.conf"
def __init__(self):
config = ConfigParser.ConfigParser()
config.read(self.__configFileName)
self.biz_db_host = config.get("biz_db","host")
self.biz_db_user = config.get("biz_db","user")
self.biz_db_password = config.get("biz_db","password")
self.biz_db_database = config.get("biz_db","database")
## End def __init__
## End class LoadAppConf
class Biz_Base(object):
"""
biz base class
"""
def __init__(self, db):
self.db = db
## End def __init__
## End class Biz_Base
class Biz_bx_book_ratings(Biz_Base):
"""
bx_book_ratings table
"""
__tableName = "bx_book_ratings"
def __init__(self, db):
Biz_Base.__init__(self, db)
## End def __init__
def insert(self, userid, isbn, bookrating):
cnt = self.db.insert(self.__tableName,\
userid = userid, \
isbn = isbn,\
bookrating = bookrating)
return cnt >; 0
## End def insert
## End class Biz_bx_book_ratings
class Biz_bx_books(Biz_Base):
"""
bx_books table
"""
__tableName = "bx_books"
def __init__(self, db):
Biz_Base.__init__(self, db)
## End def __init__
def insert(self, isbn, booktitle, bookauthor, yearofpublication, publisher, imageurls, imageurlm, imageurll):
cnt = self.db.insert(self.__tableName,\
isbn = isbn, \
booktitle = booktitle, \
bookauthor = bookauthor,\
yearofpublication = yearofpublication, \
publisher = publisher, \
imageurls = imageurls, \
imageurlm = imageurlm, \
imageurll = imageurll)
return cnt >; 0
## End def insert
## End class Biz_bx_books
class Biz_bx_users(Biz_Base):
"""
bx_users table
"""
__tableName = "bx_users"
def __init__(self, db):
Biz_Base.__init__(self, db)
## End def __init__
def insert(self, userid, location, age):
cnt = self.db.insert(self.__tableName,\
userid = userid, \
location = location,\
age = age)
return cnt >; 0
## End def insert
## End class Biz_bx_users
def regx(l):
"""
split line by regex
"""
p = re.compile(r'"[^"]*"')
return p.findall(l)
## End def regx
class LoadDataset(object):
"""
bx_books table
"""
__loadConf = None
__users = None
__books = None
__book_ratings = None
__bizDb = None
def __init__(self):
self.__loadConf = LoadAppConf()
self.__bizDb = MysqlPythonFacotry(self.__loadConf.biz_db_host,\
self.__loadConf.biz_db_user, \
self.__loadConf.biz_db_password,\
self.__loadConf.biz_db_database)
self.__users = Biz_bx_users(self.__bizDb)
self.__books = Biz_bx_books(self.__bizDb)
self.__book_ratings = Biz_bx_book_ratings(self.__bizDb)
self.__bizDb.open()
## End def __init__
def toDB(self, path=''):
"""
loads the BX book dataset. Path is where the BX files are
located
"""
self.data = {}
i = 0
j = 0
try:
#
# First load book ratings into self.data
#
f = codecs.open(path + "BX-Book-Ratings.csv", 'r', 'utf8')
for line in f:
i += 1
j += 1
print(j)
print(line)
#separate line into fields
fields = line.split(';')
user = fields[0].strip('"')
book = fields[1].strip('"')
rating = int(fields[2].strip().strip('"'))
self.__book_ratings.insert(user, book, rating)
f.close()
#
# Now load books into self.productid2name
# Books contains isbn, title, and author among other fields
#
j = 0
f = codecs.open(path + "BX-Books.csv", 'r', 'utf8')
for line in f:
i += 1
j += 1
print(j)
print(line)
#separate line into fields
fields = regx(line)
isbn = fields[0].strip('"')
title = fields[1].strip('"')
author = fields[2].strip().strip('"')
yearOfPublication = fields[3].strip().strip('"')
publisher = fields[4].strip().strip('"')
imageUrlS = fields[5].strip().strip('"')
imageUrlM = fields[6].strip().strip('"')
imageUrlL = fields[7].strip().strip('"')
self.__books.insert(isbn, title, author, yearOfPublication, publisher, imageUrlS, imageUrlM, imageUrlL)
f.close()
#
# Now load user info into both self.userid2name and
# self.username2id
#
j = 0
f = codecs.open(path + "BX-Users.csv", 'r', 'utf8')
for line in f:
i += 1
j += 1
print(j)
print(line)
#separate line into fields
fields = regx(line)
userid = fields[0].strip('"')
location = fields[1].strip('"')
if len(fields) >; 2:
age = fields[2].strip().strip('"')
else:
age = None
if age != None:
value = location + ' (age: ' + age + ')'
else:
value = location
if age == None:
age =0
self.__users.insert(userid, location, age)
f.close()
except ErrorMyProgram as e:
print(e.value)
finally:
self.__bizDb.close()
print(i)
## End def toDB
## End class LoadData
