随机森林预测财务报表是否舞弊
爬取财务报表
“”"
e.g: http://data.eastmoney.com/bbsj/201806/lrb.html
“”"
import requests
import re
from multiprocessing import Pool
import json
import csv
import pandas as pd
import os
import time
设置文件保存在c盘eastmoney文件夹下
file_path = ‘C:\eastmoney’
if not os.path.exists(file_path):
os.mkdir(file_path)
os.chdir(file_path)
1 设置表格爬取时期
def set_table():
print(’*’ * 80)
print(’\t\t\t\t东方财富网报表下载’)
print(‘作者:高级农民工 2018.10.10’)
print(’--------------’)
# 1 设置财务报表获取时期
year = int(float(input('请输入要查询的年份(四位数2016-2019):\n')))
# int表示取整,里面加float是因为输入的是str,直接int会报错,float则不会
# https://stackoverflow.com/questions/1841565/valueerror-invalid-literal-for-int-with-base-10
while (year < 2016 or year > 2019):
year = int(float(input('年份数值输入错误,请重新输入:\n')))
quarter = int(float(input('请输入小写数字季度(1:1季报,2-年中报,3:3季报,4-年报):\n')))
while (quarter < 1 or quarter > 4):
quarter = int(float(input('季度数值输入错误,请重新输入:\n')))
# 转换为所需的quarter 两种方法,2表示两位数,0表示不满2位用0补充,
# http://www.runoob.com/python/att-string-format.html
quarter = '{:02d}'.format(quarter * 3)
# quarter = '%02d' %(int(month)*3)
# 确定季度所对应的最后一天是30还是31号
if (quarter == '06') or (quarter == '09'):
day = 30
else:
day = 31
date = '{}-{}-{}' .format(year, quarter, day)
# print('date:', date) # 测试日期 ok
# 2 设置财务报表种类
tables = int(
input('请输入查询的报表种类对应的数字(1-业绩报表;2-业绩快报表:3-业绩预告表;4-预约披露时间表;5-资产负债表;6-利润表;7-现金流量表): \n'))
dict_tables = {1: '业绩报表', 2: '业绩快报表', 3: '业绩预告表',
4: '预约披露时间表', 5: '资产负债表', 6: '利润表', 7: '现金流量表'}
dict = {1: 'YJBB', 2: 'YJKB', 3: 'YJYG',
4: 'YYPL', 5: 'ZCFZB', 6: 'LRB', 7: 'XJLLB'}
category = dict[tables]
# js请求参数里的type,第1-4个表的前缀是'YJBB20_',后3个表是'CWBB_'
# 设置se