python - Q : Python CSV, 能否根据日期时间列中的不同日期自动将 CSV 中的数据行发送到数据框

Question

Sample CSV File可以在这里查看

数据描述:

我有大量数据，其中有名为 Time_Stamp 的日期时间列 [例如2017/6/21 01:09:30。每条记录以秒为单位创建行...因此 2017/6/21 01:09:30、2017/6/21 01:09:31 等等向前。也有不同的日期和不同的时间，例如 2017/6/22 03:38:43。

我想要得到什么:

• 将 Time_Stamp 的日期为 2017/6/21 ..(time).. 的行获取到 Dataframe 中，然后按其他日期(例如2017/6/22 ..(时间)..)也放入单独的数据帧中。

我的Python代码:

import csv
from datetime import date,time,datetime
import pandas as pd


Data = pd.read_csv('MainD2.csv', parse_dates=['Time_Stamp'], infer_datetime_format=True)
Data['Date'] = Data.Time_Stamp.dt.date 
Data['Time'] = Data.Time_Stamp.dt.time 

Data['Time_Stamp'] = pd.to_datetime(Data['Time_Stamp'])
df = pd.DataFrame(Data)

下面是我通过手动获取时间范围和日期尝试的代码: - 我希望这是自动化的，但我不知道该怎么做 -

countIC = 0
countIC2 = 0
countIC3 = 0
def getMask(start,end):
    mask = (df['Time_Stamp'] > start) & (df['Time_Stamp'] <= end)
    return mask

start = '2017-06-21 01:09:21'
end = '2017-06-21 01:18:54'
timerange1 = Data.loc[mask]

InputCurrent =timerange1.AC_Input_Current.tolist()

for i in InputCurrent:
    if float(i) >= 4.9:
        countIC += 1
        a = (countIC/60)
        a1 = (a/60)
    else:
        countIC2 += 1
        b = (countIC2/60)
        b1 = (b/60)
countIC3 = (a+b)
print ('Start: ', start, ' - ', 'End:', end ,' \n')
print ('%.2f' %a,'[in Mins]',' or ','%.2f' %a1,'[Hrs]',' of work') 
print ('%.2f' %b,'[in Mins]',' or ','%.2f' %b1,'[Hrs]',' of no work done')
print ('Total duration recorded: ','%.2f' %countIC3,'[Mins]')

countICC = 0
countIC22 = 0
countIC33 = 0

def getMask(start,end):
    mask1 = (df['Time_Stamp'] > start) & (df['Time_Stamp'] <= end)
    return mask1

start = '2017-06-21 19:54:20'
end = '2017-06-21 23:15:04'
timerange2 = Data.loc[mask1]

InputCurrent =timerange2.AC_Input_Current.tolist()

for i in InputCurrent:
    if float(i) >= 4.9:
        countICC += 1
        a = (countICC/60)
        a1 = (a/60)
    else:
        countIC22 += 1
        b = (countIC22/60)
        b1 = (b/60)
countIC33 = (a+b)
print ('Start: ', start, ' - ', 'End:', end ,' \n')
print ('%.2f' %a,'[in Mins]',' or ','%.2f' %a1,'[Hrs]',' of work') 
print ('%.2f' %b,'[in Mins]',' or ','%.2f' %b1,'[Hrs]',' of no work done')
print ('Total duration recorded: ','%.2f' %countIC33,'[Mins]')

上面代码的输出:

Start:  2017-06-21 01:09:21  -  End: 2017-06-21 01:18:54  

3.10 [in Mins]  or  0.05 [Hrs]  of work done
6.23 [in Mins]  or  0.10 [Hrs]  of No work done
Total Duration recorded:  9.33 [Mins]

Start:  2017-06-21 19:54:20  -  End: 2017-06-21 23:15:04  

125.00 [in Mins]  or  2.08 [Hrs]  of work done
472.38 [in Mins]  or  7.87 [Hrs]  of No work done
Total Duration recorded:  597.38 [Mins]

Answer 1

首先:你的代码非常像复制粘贴，你定义了函数 getMask 两次，你只需要执行一次。与您的输出字符串相同。

而且你计算工作/非工作时间的方法不是很好。在每次迭代中将最终值除以 60，您也可以在最后进行此计算。

但是，pandas 拥有非常强大的工具来选择、索引和分析数据。尝试下面的代码:

    import csv
    from datetime import date,time,datetime 
    import pandas as pd

    def print_output(df):
        worktime = len(df.loc[df['AC_Input_Current'] >= 4.9])
        noworktime = len(df) - worktime

        print ('Start: ', df.index.min(), ' - ', 'End:', df.index.max())
        print ('%.2f' %(worktime/60),'[in Mins]',' or ','%.2f' %(worktime/3600),'[Hrs]',' of work') 
        print ('%.2f' %(noworktime/60),'[in Mins]',' or ','%.2f' %(noworktime/3600),'[Hrs]',' of no work done')
        print ('Total duration recorded: ','%.2f' %((noworktime + worktime)/60),'[Mins]\n')

    dateparse = lambda x: pd.datetime.strptime(x, '%d/%m/%Y %H:%M:%S')

    Data = pd.read_csv('MainD2.csv', parse_dates=['Time_Stamp'], date_parser = dateparse, index_col='Time_Stamp')

    print_output(Data['2017-06-23'].between_time('08:00', '23:15:04'))

    Data.resample('D').apply(lambda x: print_output(x))

print_output() 打印所选数据的报告。通过日期和 df. Between_time() 方法选择数据。请注意，数据框现在将日期时间作为索引，与您的示例不同。

如果您想制作每日报告，您可以使用df.resample()方法并应用您自己的函数，在本例中为print_output()

这会打印:第一份报告来自于选定时间的手动 print_output()，下面是每日报告。

    Start:  2017-06-23 08:43:31  -  End: 2017-06-23 23:10:15
    33.02 [in Mins]  or  0.55 [Hrs]  of work
    210.40 [in Mins]  or  3.51 [Hrs]  of no work done
    Total duration recorded:  243.42 [Mins]

    Start:  2017-06-21 01:09:21  -  End: 2017-06-21 23:15:04
    43.75 [in Mins]  or  0.73 [Hrs]  of work
    161.63 [in Mins]  or  2.69 [Hrs]  of no work done
    Total duration recorded:  205.38 [Mins]

    Start:  2017-06-22 00:38:54  -  End: 2017-06-22 05:47:29
    30.25 [in Mins]  or  0.50 [Hrs]  of work
    271.40 [in Mins]  or  4.52 [Hrs]  of no work done
    Total duration recorded:  301.65 [Mins]

    Start:  2017-06-23 00:39:50  -  End: 2017-06-23 23:10:15
    44.23 [in Mins]  or  0.74 [Hrs]  of work
    263.35 [in Mins]  or  4.39 [Hrs]  of no work done
    Total duration recorded:  307.58 [Mins]

    Start:  2017-06-24 00:38:10  -  End: 2017-06-24 13:33:49
    77.83 [in Mins]  or  1.30 [Hrs]  of work
    120.00 [in Mins]  or  2.00 [Hrs]  of no work done
    Total duration recorded:  197.83 [Mins]

    Start:  2017-06-25 09:54:25  -  End: 2017-06-25 20:16:05
    75.62 [in Mins]  or  1.26 [Hrs]  of work
    131.28 [in Mins]  or  2.19 [Hrs]  of no work done
    Total duration recorded:  206.90 [Mins]

    Start:  2017-06-26 00:51:32  -  End: 2017-06-26 01:05:55
    9.18 [in Mins]  or  0.15 [Hrs]  of work
    4.92 [in Mins]  or  0.08 [Hrs]  of no work done
    Total duration recorded:  14.10 [Mins]

编辑、绘制数据:

首先，我对matplotlib不太熟悉，所以请参阅文档。这只是一个非常基本的介绍。

import matplotlib.pyplot as plt

然后您可以在 print_output() 函数的末尾添加:

df.AC_Input_Current.plot(kind='line')
plt.show()