Question

我试图对市场数据进行一些统计分析，并设法将交易平台的tick数据转换为python数据框，格式如下所示。

我需要进行一项测试，看看当天会话（09：30-16：15）多长时间等于或超过会议剩余时间（16:15 - 09:30）打印的高点或低点。有人可以帮我这个吗？

export default React.createClass({

    change: function(){
        console.log(this.props.classes);
    },

    render: function(){

        return(
            <div>

             <select onChange = {this.change}>
             {
                Object.keys(this.props.classes).map((value, key) =>{

                return <option key = {key}>{value}</option>
                }
             )}
              </select>
            </div>

        )

    }

});

期望的输出：

print (df.iloc[85:95])
                              Price  bidVol  askVol
2015-12-28 08:39:12.000000  2041.00       2       0
2015-12-28 08:39:12.001000  2041.00       1       0
2015-12-28 08:47:19.000000  2041.25       0       2
2015-12-28 09:18:00.999999  2040.25       0       5
2015-12-28 09:33:30.000000  2039.75       0       1
2015-12-28 09:35:58.000000  2039.75       0       5
2015-12-28 09:44:35.000000  2039.75       1       0
2015-12-28 09:58:39.999999  2039.00       0       1
2015-12-28 09:58:55.000000  2039.25       1       0
2015-12-28 10:02:59.000000  2038.50       1       0

我试试：

high 2015-12-28  True or False
low  2015-12-28  True or False

All data

Answer 1

您可以尝试：

import pandas as pd

#for all data remove iloc[85:95]
df = pd.read_hdf('tubus.h5').iloc[85:95]
print (df)
                              Price  bidVol  askVol
2015-12-28 08:39:12.000000  2041.00       2       0
2015-12-28 08:39:12.001000  2041.00       1       0
2015-12-28 08:47:19.000000  2041.25       0       2
2015-12-28 09:18:00.999999  2040.25       0       5
2015-12-28 09:33:30.000000  2039.75       0       1
2015-12-28 09:35:58.000000  2039.75       0       5
2015-12-28 09:44:35.000000  2039.75       1       0
2015-12-28 09:58:39.999999  2039.00       0       1
2015-12-28 09:58:55.000000  2039.25       1       0
2015-12-28 10:02:59.000000  2038.50       1       0

daystart = '9:30'
dayend = '16:14:59'

#filter day data, include start and end 
day_session = df.between_time(daystart,dayend, include_start=True, include_end=True)
print (day_session.head())
                              Price  bidVol  askVol
2015-12-28 09:33:30.000000  2039.75       0       1
2015-12-28 09:35:58.000000  2039.75       0       5
2015-12-28 09:44:35.000000  2039.75       1       0
2015-12-28 09:58:39.999999  2039.00       0       1
2015-12-28 09:58:55.000000  2039.25       1       0

#filter night data, exclude start and end
night_session = df.between_time(dayend,daystart, include_start=False, include_end=False)
print (night_session.head())
                              Price  bidVol  askVol
2015-12-28 08:39:12.000000  2041.00       2       0
2015-12-28 08:39:12.001000  2041.00       1       0
2015-12-28 08:47:19.000000  2041.25       0       2
2015-12-28 09:18:00.999999  2040.25       0       5

#resample only column Price
#version above 0.18.0
df_day = day_session.Price.resample('D').ohlc()
df_night = night_session.Price.resample('D').ohlc()

#version bellow 0.18.0
#df_day = day_session.Price.resample('D', how='ohlc')
#df_night = night_session.Price.resample('D', how='ohlc')


#day data are longer as nightly, so you need reindex them
df_day = df_day.reindex(df_night.index)

print (df_day)
               open     high     low   close
2015-12-28  2039.75  2039.75  2038.5  2038.5

print (df_night)
              open     high      low    close
2015-12-28  2041.0  2041.25  2040.25  2040.25

#compare columns high and low
print (df_day.high >=  df_night.high)
2015-12-28    False
Freq: D, Name: high, dtype: bool

print (df_day.low <= df_night.low)
2015-12-28    True
Freq: D, Name: low, dtype: bool

刻度数据的时间序列分析

1 个答案: