Question

我从互联网上提取了一些数据，基本上是一整年的2列小时数据：

france.GetData(base_scenario, utils.enumerate_periods(start,end,'H','CET'))

输出

2015-12-31 23:00:00+00:00    23.86
2016-01-01 00:00:00+00:00    22.39
2016-01-01 01:00:00+00:00    20.59
2016-01-01 02:00:00+00:00    16.81
2016-01-01 03:00:00+00:00    17.41
2016-01-01 04:00:00+00:00    17.02
2016-01-01 05:00:00+00:00    15.86...

我想基本上添加两个列'peak'小时和'off off'小时缩放列。因此，如果当天的时间在0800和1800之间，则峰值列中将有1，如果超出这些时间，则在非峰值列中将有1。

有人可以解释一下如何做到这一点。

非常感谢

Answer 1

我认为如果不是DatetimeIndex，您可以使用to_datetime，然后使用between_time到列peak并测试notnull - 如果NaN获取False，如果某个值获得True。然后布尔值转换为int（False - ＆gt; 0和True - ＆gt; 1）astype，最后一列peak获取peak-off（感谢Quickbeam2k1）：

df = pd.DataFrame({'col': {'2016-01-01 01:00:00+00:00': 20.59, '2016-01-01 07:00:00+00:00': 15.86, '2016-01-01 10:00:00+00:00': 15.86, '2016-01-01 09:00:00+00:00': 15.86, '2016-01-01 02:00:00+00:00': 16.81, '2016-01-01 03:00:00+00:00': 17.41, '2016-01-01 05:00:00+00:00': 15.86, '2016-01-01 04:00:00+00:00': 17.02, '2016-01-01 08:00:00+00:00': 15.86, '2015-12-31 23:00:00+00:00': 23.86, '2016-01-01 18:00:00+00:00': 15.86, '2016-01-01 06:00:00+00:00': 15.86, '2016-01-01 00:00:00+00:00': 22.39}})
print (df)
                             col
2015-12-31 23:00:00+00:00  23.86
2016-01-01 00:00:00+00:00  22.39
2016-01-01 01:00:00+00:00  20.59
2016-01-01 02:00:00+00:00  16.81
2016-01-01 03:00:00+00:00  17.41
2016-01-01 04:00:00+00:00  17.02
2016-01-01 05:00:00+00:00  15.86
2016-01-01 06:00:00+00:00  15.86
2016-01-01 07:00:00+00:00  15.86
2016-01-01 08:00:00+00:00  15.86
2016-01-01 09:00:00+00:00  15.86
2016-01-01 10:00:00+00:00  15.86
2016-01-01 18:00:00+00:00  15.86

print (df.index)
Index(['2015-12-31 23:00:00+00:00', '2016-01-01 00:00:00+00:00',
       '2016-01-01 01:00:00+00:00', '2016-01-01 02:00:00+00:00',
       '2016-01-01 03:00:00+00:00', '2016-01-01 04:00:00+00:00',
       '2016-01-01 05:00:00+00:00', '2016-01-01 06:00:00+00:00',
       '2016-01-01 07:00:00+00:00', '2016-01-01 08:00:00+00:00',
       '2016-01-01 09:00:00+00:00', '2016-01-01 10:00:00+00:00',
       '2016-01-01 18:00:00+00:00'],
      dtype='object')

df.index = pd.to_datetime(df.index) 

print (df.index)
DatetimeIndex(['2015-12-31 23:00:00', '2016-01-01 00:00:00',
               '2016-01-01 01:00:00', '2016-01-01 02:00:00',
               '2016-01-01 03:00:00', '2016-01-01 04:00:00',
               '2016-01-01 05:00:00', '2016-01-01 06:00:00',
               '2016-01-01 07:00:00', '2016-01-01 08:00:00',
               '2016-01-01 09:00:00', '2016-01-01 10:00:00',
               '2016-01-01 18:00:00'],
              dtype='datetime64[ns]', freq=None)

df['peak'] = df.between_time('08:00', '18:00')
df['peak'] = df['peak'].notnull().astype(int)
df['peak-off'] = -df['peak'] + 1
print (df)
                       col  peak  peak-off
2015-12-31 23:00:00  23.86     0         1
2016-01-01 00:00:00  22.39     0         1
2016-01-01 01:00:00  20.59     0         1
2016-01-01 02:00:00  16.81     0         1
2016-01-01 03:00:00  17.41     0         1
2016-01-01 04:00:00  17.02     0         1
2016-01-01 05:00:00  15.86     0         1
2016-01-01 06:00:00  15.86     0         1
2016-01-01 07:00:00  15.86     0         1
2016-01-01 08:00:00  15.86     1         0
2016-01-01 09:00:00  15.86     1         0
2016-01-01 10:00:00  15.86     1         0
2016-01-01 18:00:00  15.86     1         0

另一个解决方案是首先按条件获取boolean掩码，然后将其转换为int，以反转掩码使用~：

h1 = pd.datetime.strptime('08:00:00', '%H:%M:%S').time()
h2 = pd.datetime.strptime('18:00:00', '%H:%M:%S').time()
times = df.index.time
mask = (times >= h1) & (times <= h2)
df['peak'] = mask.astype(int)
df['peak-off'] = (~mask).astype(int)
print (df)
                       col  peak  peak-off
2015-12-31 23:00:00  23.86     0         1
2016-01-01 00:00:00  22.39     0         1
2016-01-01 01:00:00  20.59     0         1
2016-01-01 02:00:00  16.81     0         1
2016-01-01 03:00:00  17.41     0         1
2016-01-01 04:00:00  17.02     0         1
2016-01-01 05:00:00  15.86     0         1
2016-01-01 06:00:00  15.86     0         1
2016-01-01 07:00:00  15.86     0         1
2016-01-01 08:00:00  15.86     1         0
2016-01-01 09:00:00  15.86     1         0
2016-01-01 10:00:00  15.86     1         0
2016-01-01 18:00:00  15.86     1         0

如果只有小时数据解决方案可以更简单 - 使用DatetimeIndex.hour作为掩码：

df.index = pd.to_datetime(df.index) 
print (df.index)
h = df.index.hour
mask = (h >= 8) & (h <= 18)
df['peak'] = mask.astype(int)
df['peak-off'] = (~mask).astype(int)
print (df)
                       col  peak  peak-off
2015-12-31 23:00:00  23.86     0         1
2016-01-01 00:00:00  22.39     0         1
2016-01-01 01:00:00  20.59     0         1
2016-01-01 02:00:00  16.81     0         1
2016-01-01 03:00:00  17.41     0         1
2016-01-01 04:00:00  17.02     0         1
2016-01-01 05:00:00  15.86     0         1
2016-01-01 06:00:00  15.86     0         1
2016-01-01 07:00:00  15.86     0         1
2016-01-01 08:00:00  15.86     1         0
2016-01-01 09:00:00  15.86     1         0
2016-01-01 10:00:00  15.86     1         0
2016-01-01 18:00:00  15.86     1         0

根据条件时间序列添加列pandas系列

1 个答案: