Skip to content

1
1024程序员开源挑战赛

机器未来 / 1024程序员开源挑战赛
与 Fork 源项目一致

Fork自 GitCode / 1024程序员开源挑战赛（10.23-11.14）

通知 2
Star 0
Fork 0
1
1024程序员开源挑战赛
提交 33529055 编写于 作者: Z zhoushimin
浏览文件
操作

YaoLowCodeDemo

上级 3433b8a0
隐藏空白更改
内联 并排
Showing with 8 addition and 730 deletion
demo-widget @ 2489060e
Subproject commit 2489060e09630dff334e88b8e51e1efa1bfa1989
fund_tools.py 已删除 100644 → 0
浏览文件 @ 3433b8a0
from asyncio.windows_events import NULL
from cProfile import label
import akshare as ak
import pandas as pd
import numpy as np
from matplotlib import pyplot as plt
from matplotlib.pyplot import MultipleLocator
import matplotlib.dates as mdate
import requests
from bs4 import BeautifulSoup
import datetime
import time
import math
import os
import re
import random
from decimal import Decimal
dir_cumulative_net_value_trend = 'cumulative_net_value_trend'
dir_cumulative_return_trend = 'cumulative_return_trend'
# 基金行业常用常量定义
# 无风险年化收益率%
risk_free_annual_return_ratio = 0.0275
# 每年交易日天数
trading_days_per_year = 250
# user_agent列表
user_agent_list = [
'Mozilla/5.0 (Windows NT 6.1; WOW64) AppleWebKit/537.1 (KHTML, like Gecko) Chrome/21.0.1180.71 Safari/537.1 LBBROWSER',
'Mozilla/4.0 (compatible; MSIE 6.0; Windows NT 5.1; SV1; QQDownload 732; .NET4.0C; .NET4.0E)',
'Mozilla/5.0 (Windows NT 5.1) AppleWebKit/535.11 (KHTML, like Gecko) Chrome/17.0.963.84 Safari/535.11 SE 2.X MetaSr 1.0',
'Mozilla/5.0 (Windows NT 6.1; WOW64) AppleWebKit/537.36 (KHTML, like Gecko) Maxthon/4.4.3.4000 Chrome/30.0.1599.101 Safari/537.36',
'Mozilla/5.0 (Windows NT 6.1; WOW64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/38.0.2125.122 UBrowser/4.0.3214.0 Safari/537.36'
]
# referer列表
referer_list = [
'http://fund.eastmoney.com/110022.html',
'http://fund.eastmoney.com/110023.html',
'http://fund.eastmoney.com/110024.html',
'http://fund.eastmoney.com/110025.html'
]
# 获取一个随机user_agent和Referer
header = {'User-Agent': random.choice(user_agent_list),
'Referer': random.choice(referer_list)
}
# 获得字符串中的整数
def get_decimal(x):
return ''.join(re.findall(r"\d+\.?\d*", ''.join(x)))
# 获得指定后缀字符串中的整数
def get_decimal_suffix(x, suffix):
return get_decimal(re.findall(r"\d+\.?\d*"+suffix, ''.join(x)))
# 日志调试
def log(x):
file_handle = open('log.txt', mode='a', encoding='utf-8')
file_handle.write(x)
file_handle.close()
# 获得基金概要数据
def get_all_fund_outline():
if not os.path.exists('fund_em_fund_name_df.csv'):
df1 = ak.fund_name_em()
df1 = df1[['基金代码', '拼音缩写', '基金类型']]
df2 = ak.fund_open_fund_daily_em()
df = pd.merge(df1, df2, on='基金代码')
df.set_index(df['基金代码'])
df.to_csv('fund_em_fund_name_df.csv', encoding='utf_8_sig',index=None)
print(df)
else:
df = pd.read_csv('fund_em_fund_name_df.csv', dtype=object)
return df
#获得公墓基金每日净值
def get_all_open_fund_daily():
if not os.path.exists('open_fund_daily.csv'):
fund_em_fund_name_df = ak.fund_open_fund_daily_em()
fund_em_fund_name_df.set_index(fund_em_fund_name_df['基金代码'])
fund_em_fund_name_df.to_csv('fund_em_open_fund_daily.csv', encoding='utf_8_sig',index=None)
else:
fund_em_fund_name_df = pd.read_csv('fund_em_open_fund_daily.csv', dtype=object)
return fund_em_fund_name_df
# 根据关键字过滤基金列表
def query_fund_by_fundname_keyword(df, fundname_keyword):
# if not os.path.exists(fundname_keyword + '.csv'):
df_query = df[df['基金简称'].str.contains(fundname_keyword)]
df_query.set_index(df_query['基金代码'])
df_query.to_csv(fundname_keyword + '.csv', encoding='utf_8_sig',index=None);
# else:
# df_query = pd.read_csv(fundname_keyword + '.csv', dtype=object)
return df_query
# 获得分红数据
def get_fund_dividend_by_id(dir, fund_id):
filename = dir + '/%s' % (fund_id) + '-dividend.csv'
try:
if not os.path.exists(filename):
df = ak.fund_open_fund_info_em(fund=fund_id, indicator="分红送配详情")
print(df)
for i in np.arange(0, df.shape[0]):
df.iloc[i]['每份分红'] = get_decimal(df.iloc[i]['每份分红'])
df.to_csv(filename, encoding='utf_8_sig', index=None);
else:
df = pd.read_csv(filename, dtype=object)
return df
except:
return NULL
# 获得最近一次分红
def query_last_dividend_before_date(df, cur_date):
try:
dividend = '0.0'
# 查询最近一次的分红数据，满足条件反馈分红
for i in np.arange(0, df.shape[0]):
if str(cur_date) > df.iloc[df.shape[0] - 1 - i]['分红发放日']:
# print('Index:', df.shape[0] - 1 - i)
dividend = df.iloc[df.shape[0] - 1 - i]['每份分红']
# print("日期:%s"%cur_date, "分红%s"%dividend)
return dividend
except:
return '0.0'
# 获得指定基金历史数据
def get_fund_his_by_id(fund_id, dir):
filename = dir + '/%s'%(fund_id) + '.csv'
# print(filename)
try:
if not os.path.exists(filename):
df_total_net_value_trend = ak.fund_open_fund_info_em(fund=fund_id, indicator='累计净值走势')
df_unit_net_value_trend = ak.fund_open_fund_info_em(fund=fund_id, indicator="单位净值走势")
fund_em_info_df = pd.merge(df_total_net_value_trend, df_unit_net_value_trend, on='净值日期')
fund_em_info_df.set_index(fund_em_info_df['净值日期'])
if not isinstance(fund_em_info_df['日增长率'], float):
fund_em_info_df['日增长率'] = fund_em_info_df['日增长率'].astype(float)
df_dividend = get_fund_dividend_by_id(dir, fund_id)
# 复权列表
red_net_value_list = []
last_red_net_value = 0.0
# print(fund_em_info_df)
# print(f"fund_em_info_df.shape:{fund_em_info_df.shape}")
# 复权净值计算
for i in np.arange(0, fund_em_info_df.shape[0]):
dividend = float(query_last_dividend_before_date(df_dividend, fund_em_info_df.iloc[i]['净值日期']))
# print(dividend)
if dividend > 0.0:
try:
last_red_net_value = float(last_red_net_value * (1 + fund_em_info_df.iloc[i]['日增长率'] / 100))
red_net_value_list.append(last_red_net_value)
# print('date:',fund_em_info_df.iloc[i]['净值日期'], '单位净值:', fund_em_info_df.iloc[i]['单位净值'], \
# '日增长率:', fund_em_info_df.iloc[i]['日增长率'], '累计净值', fund_em_info_df.iloc[i]['累计净值'])
except:
print(fund_id, '日增长率数据异常:', fund_em_info_df.iloc[i]['日增长率'])
return ''
else:
try:
last_red_net_value = float(fund_em_info_df.iloc[i]['累计净值'])
red_net_value_list.append(last_red_net_value)
except:
print(fund_id, '累计净值数据异常:', fund_em_info_df.iloc[i]['累计净值'])
return ''
fund_em_info_df['复权净值'] = red_net_value_list
# print(fund_em_info_df)
fund_em_info_df.to_csv(filename, encoding='utf_8_sig',index=None)
else:
fund_em_info_df = pd.read_csv(filename, dtype=object)
fund_em_info_df.set_index(fund_em_info_df['净值日期'])
except Exception as e:
print(f"错误信息：{e}")
return fund_em_info_df
# 绘制基金净值走势
def draw_cumulative_net_value_trend(x, y, y_hushen300):
# plt.style.use('dark_background')
plt.rcParams["font.sans-serif"]=["SimHei"] #设置字体
plt.rcParams["axes.unicode_minus"]=False #该语句解决图像中的“-”负号的乱码问题
fig, ax = plt.subplots()
# fig.subplots_adjust(bottom=0.2)
# y_major_locator = MultipleLocator(0.1)
# # 设置X轴刻度为日期时间
# ax.xaxis.set_major_formatter(mdate.DateFormatter('%Y-%m-%d')) # 设置时间标签显示格式
# ax.yaxis.set_major_locator(y_major_locator)
# ax.xaxis_date()
# X轴刻度文字倾斜45度
# plt.xticks(rotation=45)
ax.set_title("基金走势图")
ax.set_xlabel("时间")
ax.set_ylabel("基金累计净值")
ax.plot(x, y, color='r', label="自选组合")
ax.plot(x, y_hushen300, color='b', label="沪深300-110020")
ax.legend()
plt.show()
# 计算基金成立时间
def calc_fund_established_time(df):
# print(df.dtypes)
# print(df.iloc[-1, 0], df.iloc[0, 0], df.iloc[-1, 0]-df.iloc[0, 0])
established_time = (df.iloc[-1, 0] - df.iloc[0, 0]).days/365
return established_time
# 计算基金最大回撤
def calc_fund_drawdown(fund_df, start='1970-01-01', end='2200-01-01'):
df = fund_df[['净值日期', '复权净值']].copy()
df['净值日期'] = pd.to_datetime(df['净值日期'])
st = df['净值日期'] >= start
et = df['净值日期'] <= end
res = st == et
df = df[res]
df['max2here'] = df['复权净值'].expanding().max()
df['dd2here'] = df['复权净值']/df['max2here']
end_date, remains = tuple(df.sort_values(by=['dd2here']).iloc[0][['净值日期', 'dd2here']])
# 计算最大回撤开始时间
start_date = df[df['净值日期']<=end_date].sort_values(by='复权净值', ascending=False).iloc[0]['净值日期']
drawdown = round((1-remains)*100, 2)
print('最大回撤 (%):', drawdown)
print('最大回撤开始时间:', start_date)
print('最大回撤结束时间:', end_date)
return drawdown, start_date, end_date
# 计算基金年华回报率
def cal_fund_annual_return(fund_df, start='1970-01-01', end='2200-01-01'):
df = fund_df[['净值日期', '复权净值']].copy()
# df['净值日期'] = pd.to_datetime(df['净值日期'])
#
# df['日收益'] = df['累计净值'] - df['累计净值'].shift(1)
# df['日收益率'] = (df['累计净值'] - df['累计净值'].shift(1))/df['累计净值'].shift(1)
st = df['净值日期'] >= start
et = df['净值日期'] <= end
res = st == et
df = df[res]
# df = df[df['净值日期'] >= '2020-01-09']
# 计算累计收益率
# df['累计收益率'] = df['日收益率'].expanding().sum()
total_return = round(((df.iloc[-1, 1] / df.iloc[0, 1]) - 1)*100, 2)
# df = df.fillna(0.0)
# print(df)
years = round(calc_fund_established_time(df), 2)
# print('分析周期:%f years'%years)
# 计算
# print("最新累计收益率:%f, "%df.iloc[-1, 4] + '初始累计收益率:%f'%df.iloc[0, 4])
if years >= 1:
annualized_returns = (df.iloc[-1, 1] / df.iloc[0, 1])**(1/years) - 1
else:
annualized_returns = (df.iloc[-1, 1] / df.iloc[0, 1]) - 1
annualized_returns = annualized_returns*100
print('annualized_returns:%f'%annualized_returns + '%')
return annualized_returns, total_return;
# 计算基金夏普比率
def cal_fund_sharpe_ratio(fund_df, withdrawal, start='1970-01-01', end='2200-01-01'):
df = fund_df[['净值日期', '日增长率']].copy()
st = df['净值日期'] >= start
et = df['净值日期'] <= end
res = st == et
df = df[res]
# print(df)
# 计算日收益率的均值
daily_return_ratio_average = np.mean(df['日增长率'].astype(float))*100
# print('daily_return_ratio_average:%f'%daily_return_ratio_average)
# 计算收益率的标准方差
return_ratio_std = df['日增长率'].astype(float).std()
# 计算无风险日收益率
daily_risk_free_return_ratio = (((1 + risk_free_annual_return_ratio)**(1/365)) - 1) * 100
# print('daily_risk_free_return_ratio:%f' % daily_risk_free_return_ratio)
# 计算夏普比率
try:
fund_sharpe_ratio = (daily_return_ratio_average - daily_risk_free_return_ratio) / return_ratio_std * math.sqrt(trading_days_per_year) / 100
except:
fund_sharpe_ratio = 0.0
# 计算卡玛比率
try:
fund_calmar_ratio = (daily_return_ratio_average - daily_risk_free_return_ratio) / withdrawal * 100
except:
fund_calmar_ratio = 0.0
print('sharpe_ratio:%f' % fund_sharpe_ratio + ', calmar_ratio:%f' % fund_calmar_ratio)
return fund_sharpe_ratio, fund_calmar_ratio
# 计算基金波动
def calc_fund_volatility(df_fund, start='1970-01-01', end='2200-01-01'):
# difflntotal = []
df = df_fund[['净值日期', '复权净值', '上一交易日复权净值']].copy()
# df['上一次累计净值'] = df['累计净值'].shift(1)
st = df['净值日期'] >= start
et = df['净值日期'] <= end
res = st == et
df = df[res]
# days = df.shape[0]
# print(days)
df['价格自然对数差'] = np.log(df['复权净值']) - np.log(df['上一交易日复权净值'])
volitality = np.std(df['价格自然对数差']) * 100 * math.sqrt(trading_days_per_year)
# 与前一种计算结果一致
# df['价格比率对数'] = np.log(df['累计净值'] / df['上一次累计净值'])
# volitality = np.std(df['价格比率对数'])* 100 * math.sqrt(trading_days_per_year)
print('volitality:%f'%volitality)
return volitality
# 计算基金kpi
def calc_fund_kpi(fund_id, df_cumulative_net_value_trend, start='1970-01-01', end='2200-01-01'):
dict_select = {
'code': fund_id, # 基金代码
'years': '', # 成立时间
'withdrawal': '', # 最大回撤
'annual_return': '', # 年化收益率
'total_return': '', # 累计收益率
'sharp': '', # 夏普比率
'calmar': '', # 卡玛比率
'volatility': '', # 波动率
}
df = df_cumulative_net_value_trend[['净值日期', '复权净值', '日增长率']].copy()
# 生成上一个交易日累计净值
df['上一交易日复权净值'] = df['复权净值'].shift(1)
# 数据格式转换
df['净值日期'] = pd.to_datetime(df['净值日期'])
df['复权净值'] = df['复权净值'].astype(float)
df['上一交易日复权净值'] = df['上一交易日复权净值'].astype(float)
# # 计算日收益和日收益率
# df['日收益'] = df['累计净值'] - df['上一交易日累计净值']
# df['日收益率'] = (df['累计净值'] - df['上一交易日累计净值'])/df['上一交易日累计净值']
# x = df['净值日期']
# y = df['累计净值']
#
# # 绘制基金累计净值走势图
# draw_cumulative_net_value_trend(x, y)
# 计算成立时间
years = calc_fund_established_time(df)
dict_select['years'] = round(years, 2)
print('基金[%s'%fund_id + ']成立时间:%d'%(int(years)) + '年%d'%((years-int(years))*365) + '天')
# 计算最大回撤
withdrawal, start_date, end_date = calc_fund_drawdown(df, start, end)
dict_select['withdrawal'] = round(withdrawal, 2)
# 计算年化收益
annual_return, total_return = cal_fund_annual_return(df, start, end)
dict_select['annual_return'] = round(annual_return, 2)
dict_select['total_return'] = round(total_return, 2)
# 计算夏普比率
sharpe_ratio, calmar_ratio = cal_fund_sharpe_ratio(df, withdrawal, start, end)
dict_select['sharp'] = round(sharpe_ratio, 2)
dict_select['calmar'] = round(calmar_ratio, 4)
# 计算波动率
volatility = calc_fund_volatility(df, start, end)
dict_select['volatility'] = round(volatility, 2)
print(dict_select)
return dict_select
# user_agent列表
user_agent_list = [
'Mozilla/5.0 (Windows NT 6.1; WOW64) AppleWebKit/537.1 (KHTML, like Gecko) Chrome/21.0.1180.71 Safari/537.1 LBBROWSER',
'Mozilla/4.0 (compatible; MSIE 6.0; Windows NT 5.1; SV1; QQDownload 732; .NET4.0C; .NET4.0E)',
'Mozilla/5.0 (Windows NT 5.1) AppleWebKit/535.11 (KHTML, like Gecko) Chrome/17.0.963.84 Safari/535.11 SE 2.X MetaSr 1.0',
'Mozilla/5.0 (Windows NT 6.1; WOW64) AppleWebKit/537.36 (KHTML, like Gecko) Maxthon/4.4.3.4000 Chrome/30.0.1599.101 Safari/537.36',
'Mozilla/5.0 (Windows NT 6.1; WOW64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/38.0.2125.122 UBrowser/4.0.3214.0 Safari/537.36'
]
# referer列表
referer_list = [
'http://fund.eastmoney.com/110022.html',
'http://fund.eastmoney.com/110023.html',
'http://fund.eastmoney.com/',
'http://fund.eastmoney.com/110025.html'
]
# 爬取网页基金数据
def get_url(baseUrl):
# 获取一个随机user_agent和Referer
headers = {'User-Agent': random.choice(user_agent_list), 'Referer': random.choice(referer_list)}
try:
resp = requests.get(baseUrl, headers=headers)
# print(resp.status_code)
if resp.status_code == 200:
# print(resp.content)
return resp.text
print("没有爬取到相应的内容")
return None
except:
print("没有爬取到相应的内容")
return None
# 获得基金费用
def get_fund_fee(code):
dict_select = {
'基金代码': code, # 基金代码
'scale': '', # 基金规模
'm_fee': '', # 管理费
'c_fee': '', # 托管费
'sale_fee': '', # 销售费用
'sub_fee': '', # 申购费<50W
# 'buy_fee_less_than_100': '', # 申购费<100W
# 'buy_fee_more_than_100': '', # 申购费>=100W
}
url = 'http://fundf10.eastmoney.com/jjfl_%s' % code + '.html'
html = get_url(url)
soup = BeautifulSoup(html, 'html.parser')
# log(soup.prettify())
target = ''
for co in soup.find_all(['span', 'td', 'span', 'label']):
# log('target%s' % target)
log('co.text:%s' % co.text)
if target != '':
# '''dict_select[target] = co.text.encode('utf-8')'''
dict_select[target] = co.text
if target == 'sale_fee':
break
target = ''
if co.text.find(u'资产规模') >= 0:
target = 'scale'
elif co.text == u'管理费率':
target = 'm_fee'
elif co.text == u'托管费率':
target = 'c_fee'
elif co.text == u'销售服务费率':
target = 'sale_fee'
# print(dict_select)
try:
dict_select['scale'] = float(get_decimal_suffix(dict_select['scale'], '亿元'))
if dict_select['scale'] == '':
dict_select['scale'] = 0.0
except:
dict_select['scale'] = 0.0
try:
dict_select['m_fee'] = float(get_decimal(dict_select['m_fee']))
if dict_select['m_fee'] == '':
dict_select['m_fee'] = 0.0
except:
dict_select['m_fee'] = 0.0
try:
dict_select['c_fee'] = float(get_decimal(dict_select['c_fee']))
if dict_select['c_fee'] == '':
dict_select['c_fee'] = 0.0
except:
dict_select['c_fee'] = 0.0
try:
dict_select['sale_fee'] = float(get_decimal(dict_select['sale_fee']))
if dict_select['sale_fee'] == '':
dict_select['sale_fee'] = 0.0
except:
dict_select['sale_fee'] = 0.0
print(dict_select)
return dict_select
# 获得基金费用列表
def get_fund_fee_list(df):
err_count = 0
fee_list = []
filename = 'fund_fee_list.csv'
filename_bak = 'fund_fee_list-bak.csv'
# print(filename)
if not os.path.exists(filename):
for i in np.arange(0, df.shape[0]):
# try:
print('基金编码[%s'%(df.iloc[i]['基金代码']) + ']正在读取费用...')
dict_item = get_fund_fee(df.iloc[i]['基金代码'])
dict_item['sub_fee'] = df.iloc[i]['手续费']
fee_list.append(dict_item)
# except:
# err_count = err_count + 1
# print('基金编码[%s' % (df.iloc[i]['基金代码']) + ']读取费用失败...')
time.sleep(1)
df_fee = pd.DataFrame(fee_list)
df_fee.to_csv(filename, encoding='utf_8_sig', index=None)
print('读取费率失败次数:%d' % err_count)
else:
df_fee = pd.read_csv(filename, dtype=object)
# for i in np.arange(0, df.shape[0]):
# try:
# print(df_fee.iloc[i]['基金代码'], df_fee.iloc[i]['sub_fee'])
# df_fee.iloc[i]['sub_fee'] = get_decimal(df_fee.iloc[i]['sub_fee'])
# except:
# df_fee.iloc[i]['sub_fee'] = 0.0
#
# df_fee.to_csv(filename_bak, encoding='utf_8_sig', index=None)
return df_fee
# 计算基金rank
def get_fund_rank(df_in, keywords, max_withdrawal, establish_year=3.0, start='1970-01-01', end='2200-01-01'):
fund_kpi_list = []
# 根据关键字检索
df_dist = query_fund_by_fundname_keyword(df_in, keywords)
df_dist = df_dist[df_dist['基金类型'].str.contains('场内') == False]
for i in np.arange(0, df_dist.shape[0]):
try:
print('基金编码[%s'%(df_dist.iloc[i]['基金代码']) + ']正在读取...')
# 爬取历史净值
df_cumulative_net_value_trend = get_fund_his_by_id(df_dist.iloc[i]['基金代码'], dir_cumulative_net_value_trend)
print('历史数据已加载...')
if (keywords != '债') or (keywords == '债' and df_cumulative_net_value_trend['日增长率'].astype(float).max(skipna=True) <= 2.0):
# 计算基金KPI
dict_kpi = calc_fund_kpi(df_dist.iloc[i]['基金代码'], df_cumulative_net_value_trend, start, end)
dict_kpi['code'] = df_dist.iloc[i]['基金代码']
dict_kpi['name'] = df_dist.iloc[i]['基金简称']
dict_kpi['type'] = df_dist.iloc[i]['基金类型']
print('KPI计算已完成...')
# 爬取费用
dict_kpi['scale'] = df_dist.iloc[i]['scale']
dict_kpi['m_fee'] = df_dist.iloc[i]['m_fee']
dict_kpi['c_fee'] = df_dist.iloc[i]['c_fee']
dict_kpi['sale_fee'] = df_dist.iloc[i]['sale_fee']
dict_kpi['sub_fee'] = df_dist.iloc[i]['sub_fee']
# print(dict_kpi)
fund_kpi_list.append(dict_kpi)
# time.sleep(1)
except Exception as e: # 未知异常的捕获
print(f"异常信息:{e}")
print('基金编码[%s'%(df_dist.iloc[i]['基金代码']) + ']读取失败')
# # time.sleep(1)
# print(fund_kpi_list)
df_kpi = pd.DataFrame(fund_kpi_list)
print(df_kpi)
# 删除异常数据行
# df_kpi.dropna(subset=['withdrawal'], inplace=True)
# 对数据筛选
df_kpi = df_kpi[df_kpi['years'].astype(float) >= establish_year]
df_kpi = df_kpi[df_kpi['withdrawal'].astype(float) <= max_withdrawal]
try:
df_kpi['total_fee'] = df_kpi['sub_fee'].astype(float) + df_kpi['m_fee'].astype(float) + df_kpi['c_fee'].astype(float) + df_kpi['sale_fee'].astype(float)
except Exception as e: # 未知异常的捕获
print(f"异常信息:{e}")
df_kpi['total_fee'] = 0
print('[%s'%keywords + ']计算累计费用失败，跳过!')
# 对数据排序, 'sharp', 'total_fee', 'annual_return'
df_kpi.sort_values(by=['sharp', 'withdrawal', 'calmar', 'total_fee'], ascending=[False, False, False, True], inplace=True)
# print(df_kpi.dtypes)
# df_kpi.to_csv('fund_kpi.csv')
df_kpi.to_excel('fund_rank-%s' % keywords + '.xlsx')
print(df_kpi)
return df_kpi
# 计算基金组合预测数据
def calc_fund_comb_kpi(fund_comb, start='1970-01-01', end='2200-01-01'):
df_comb = pd.DataFrame(columns=['净值日期', '复权净值', '复权净值临时', '日增长率'])
for i in np.arange(0, fund_comb.shape[0]):
# 爬取历史净值
df = get_fund_his_by_id(fund_comb.iloc[i]['code'], dir_cumulative_net_value_trend)
# df.dropna(axis=0, inplace=True)
# df.fillna(method='pad',axis=0, inplace=True)
# 过滤时间
st = df['净值日期'] >= start
et = df['净值日期'] <= end
res = st == et
df = df[res]
if df_comb.empty:
df_comb['净值日期'] = df['净值日期']
df_comb.set_index('净值日期', inplace=True)
df.set_index('净值日期', inplace=True)
# print(df)
print(fund_comb.iloc[i]['share'])
# if pd.isnull(df_comb.iloc[0]['复权净值']):
if df_comb['复权净值'].isnull().all():
df_comb['复权净值临时'] = 0.0
else:
df_comb.loc[:, '复权净值临时'] = df_comb.loc[:, '复权净值']
df_comb.loc[:, '复权净值'] = df.loc[:, '复权净值'].astype(float) * fund_comb.iloc[i]['share']
# print(type(df_comb.loc[:, '复权净值']), type(df_comb.loc[:, '复权净值临时']))
df_comb.loc[:, '复权净值'] = df_comb.loc[:, '复权净值'].astype(float).add(df_comb.loc[:, '复权净值临时'])
df_comb.fillna(method='pad', axis=0, inplace=True)
# 计算日增长率
df_comb['日增长率'] = (df_comb['复权净值'] - df_comb['复权净值'].shift(1))/df_comb['复权净值'].shift(1)
# print(df_comb)
return df_comb
if __name__ == "__main__":
pd.set_option('display.max_rows', 1000)
pd.set_option('display.max_columns', 10)
# get_fund_his_by_id('000051', dir_cumulative_net_value_trend)
# 获得公募基金列表
df = get_all_fund_outline()
# df = get_all_open_fund_daily()
df = df[df['申购状态'] == '开放申购']
# 加载费用
df_fee = get_fund_fee_list(df)
df_fee = df_fee[['基金代码', 'scale', 'm_fee', 'c_fee', 'sale_fee', 'sub_fee']].copy()
# get_fund_fee('000051')
df = pd.merge(df, df_fee, on='基金代码')
df.set_index(df['基金代码'])
# 分红测试代码
# df_dividend = get_fund_dividend_by_id('002898')
# df_dividend = query_last_dividend_before_date(dir_cumulative_net_value_trend, '002898', '2021-01-15')
# print(type(df_dividend))
# print(df_dividend)
df_kpi_csi300 = get_fund_rank(df, '沪深300', 60.0, 5, '2018-01-01', '2022-10-30')
df_kpi_csi500 = get_fund_rank(df, '中证500', 50.0, 5, '2018-01-01', '2022-10-30')
df_kpi_gem = get_fund_rank(df, '创业板', 50.0, 5, '2018-01-01', '2022-10-30')
df_kpi_gold = get_fund_rank(df, '黄金', 50.0, 5, '2018-01-01', '2022-10-30')
df_kpi_bond = get_fund_rank(df, '债', 30.0, 5, '2018-01-01', '2022-10-30')
df_kpi_sp500 = get_fund_rank(df, '标普500', 50.0, 5, '2018-01-01', '2022-10-30')
df_kpi_nasda = get_fund_rank(df, '纳斯达克', 50.0, 5, '2018-01-01', '2022-10-30')
#
# # 行业基金
# get_fund_rank(df, '新能源', 50.0, 3, '2018-01-01', '2021-02-18')
# get_fund_rank(df, '消费', 50.0, 3, '2018-01-01', '2021-02-18')
# get_fund_rank(df, '商品', 50.0, 3, '2018-01-01', '2021-02-18')
# get_fund_rank(df, '混合', 50.0, 3, '2018-01-01', '2021-03-28')
print(f"基金组合:{df_kpi_csi300.iloc[0]['code']}:0.25,\n \
{df_kpi_csi500.iloc[0]['code']}:0.15,\n \
{df_kpi_gem.iloc[0]['code']}: 0.10,\n \
{df_kpi_gold.iloc[0]['code']}: 0.10,\n \
{df_kpi_bond.iloc[0]['code']}: 0.20],\n \
{df_kpi_sp500.iloc[0]['code']}: 0.10],\n\
{df_kpi_nasda.iloc[0]['code']}: 0.10]")
# # 组合数据
# fund_comb = pd.DataFrame([['673100', 0.50], ['001879', 0.15], ['000187', 0.20], ['000218', 0.15]], columns=['code', 'share'])
fund_comb = pd.DataFrame([
[df_kpi_csi300.iloc[0]['code'], 0.25],
[df_kpi_csi500.iloc[0]['code'], 0.15],
[df_kpi_gem.iloc[0]['code'], 0.10],
[df_kpi_gold.iloc[0]['code'], 0.10],
[df_kpi_bond.iloc[0]['code'], 0.20],
[df_kpi_sp500.iloc[0]['code'], 0.10],
[df_kpi_nasda.iloc[0]['code'], 0.10]
],
columns=['code', 'share'])
df_comb = calc_fund_comb_kpi(fund_comb, '2017-06-02', '2022-10-30')
df_comb.reset_index(inplace=True)
x = df_comb['净值日期']
y = df_comb['复权净值']
fund_hs300 = pd.DataFrame([['000512', 1.00]], columns=['code', 'share'])
df_hs300 = calc_fund_comb_kpi(fund_hs300, '2017-06-02', '2022-10-30')
df_hs300.reset_index(inplace=True)
df_hs300.to_csv("参考沪深300.csv", encoding='utf_8_sig', index=None)
y_hs300 = df_hs300['复权净值']
# 绘制基金累计净值走势图
draw_cumulative_net_value_trend(x, y, y_hs300)
start.sh 0 → 100755
浏览文件 @ 33529055
cd demo-widget
# Create tables & set menu
yao migrate --reset
yao run flows.setmenu
yao start
Markdown is supported
0% 或 .
You are about to add 0 people to the discussion. Proceed with caution.
先完成此消息的编辑！
想要评论请 注册 或