深300指数 PB&PE 处于历史最低10% 时买入持有2年的年化收益率

import pandas as pd

ii = index_indicator('000300.XSHG', end_date=20170828)

ii.head()

ii.describe()

# s_pb: Series类型, trade_date作为索引，pb 作为值,累积历史数据用来统计
s_pb = pd.Series()
# s_pe: Series类型, trade_date作为索引，pe_ttm 作为值,累积历史数据用来统计
s_pe = pd.Series()
# selected_pb_date: 数组类型， 存放用 pb洗出来的 trade_date 时间戳
selected_pb_date = []
#  selected_pe_date: 数组类型， 存放用 pe洗出来的 trade_date 时间戳
selected_pe_date = []

# pandas.Series.quantile: pandas API 返回给定分位数
# 文档地址： http://pandas.pydata.org/pandas-docs/stable/generated/pandas.Series.quantile.html

for index, row in ii.iterrows():
    if s_pb.any() and row['pb'] < s_pb.quantile(0.1):
        selected_pb_date.append(row['trade_date'])
    if s_pe.any() and row['pe_ttm'] < s_pe.quantile(0.1):
        selected_pe_date.append(row['trade_date'])
    s_pb = s_pb.append(pd.Series({row['trade_date']: row['pb']}), ignore_index=False)
    s_pe = s_pe.append(pd.Series({row['trade_date']: row['pe_ttm']}), ignore_index=False)

selected_pb_date

len(selected_pb_date)

selected_pe_date

len(selected_pe_date)

# pb
ii_pb = ii.loc[ii['trade_date'].isin(selected_pb_date)];ii_pb

ii_pb.count()

# pe
ii_pe = ii.loc[ii['trade_date'].isin(selected_pe_date)];ii_pe

ii_pe.count()

get_price('000300.XSHG', start_date='2005-01-01', end_date='2016-02-12').describe()

'''

闰年2月有29天

pandas.Timestamp 相关

文档地址: [Timestamp](http://pandas.pydata.org/pandas-docs/stable/generated/pandas.Timestamp.html#pandas.Timestamp)

'''

def leap_month_day(date, modify_year_number = 0, modify_day_number = 0):
    if date.is_leap_year and date.is_month_end and date.month == 2 :
        nday = date.day - 1
    else:
        nday = date.day
    if modify_day_number != 0 and (nday + modify_day_number) > date.days_in_month:
        nmonth = date.month + 1
        nday = nday + modify_day_number - date.days_in_month
    return pd.Timestamp(date.year + modify_year_number, date.month, nday + modify_day_number)

# 利用 成交量低于2年平均值 这个条件筛选

# final_pb_date: 存放  pb 相关的 日期
final_pb_date = []
for date in selected_pb_date:
    volumes = get_price('000300.XSHG', start_date=leap_month_day(date, -2), end_date=date, fields='volume')
    if volumes[-1] < volumes.mean():
        final_pb_date.append(date)

#print(final_pb_date)

# rates_in_pb： 保存 pb 条件下得到的利润率结果
rates_in_pb = []

for day in final_pb_date:
    price_then = get_price('000300.XSHG', start_date=day, end_date=day, fields='close').values[-1]

    price_late = get_price('000300.XSHG', start_date=leap_month_day(day, 2), end_date=leap_month_day(day, 2), fields='close')
    day_time = 1
    while(price_late.empty):
        price_late = get_price('000300.XSHG', start_date=leap_month_day(day, 2), end_date=leap_month_day(day, 2, day_time), fields='close')
        day_time += 1
    price_late = price_late.values[-1]

    #print('price_then:', price_then)
    #print('price_late:', price_late)

    rates_in_pb.append(price_late / price_then - 1)

#print('rates_in_pb:', rates_in_pb)

s_pb_rates = pd.Series(rates_in_pb)
s_pb_rates

s_pb_rates.describe()

final_pe_date = []
for date in selected_pe_date:
    if date.is_leap_year and date.is_month_end and date.month == 2 :
        nday = date.day - 1
    else:
        nday = date.day
    volumes = get_price('000300.XSHG', start_date=pd.Timestamp(date.year - 2, date.month, nday), end_date=date, fields='volume')
    if volumes[-1] < volumes.mean():
        final_pe_date.append(date)
#print(len(final_pe_date))

rates_in_pe = []

for day in final_pe_date:
    price_then = get_price('000300.XSHG', start_date=day, end_date=day, fields='close').values[-1]

    price_late = get_price('000300.XSHG', start_date=leap_month_day(day, 2), end_date=leap_month_day(day, 2), fields='close')
    day_time = 1
    while(price_late.empty):
        price_late = get_price('000300.XSHG', start_date=leap_month_day(day, 2), end_date=leap_month_day(day, 2, day_time), fields='close')
        day_time += 1
    price_late = price_late.values[-1]

    #print('price_then:', price_then)
    #print('price_late:', price_late)
    rates_in_pe.append(price_late / price_then - 1)

#print('rates_in_pb:', rates_in_pb)

s_pe_rates = pd.Series(rates_in_pe)
s_pe_rates

s_pe_rates.describe()