我有一個帶有 column 的時間序列資料框volt。我在這個資料幀上應用了滾動視窗操作，視窗為 24 小時。滾動視窗的代碼是
telemetry['datetime'] = pd.to_datetime(telemetry['datetime'])
a = telemetry.set_index('datetime').groupby('machineID').rolling(window=1, freq='24h').mean()['volt']
滾動平均值后資料框看起來像這樣
machineID datetime
1 2015-01-01 06:00:00 151.919999
2015-01-01 07:00:00 174.522001
2015-01-01 08:00:00 146.912822
2015-01-01 09:00:00 179.530561
2015-01-01 10:00:00 180.544277
...
1000 2016-01-01 02:00:00 174.076779
2016-01-01 03:00:00 130.151378
2016-01-01 04:00:00 140.423844
2016-01-01 05:00:00 160.007424
2016-01-01 06:00:00 164.590354
Name: volt, Length: 8761000, dtype: float64
最后一列是伏特的平均值。
現在我想對datetime列上的 12H 頻率應用重采樣。
a = a.resample('12H', on='datetime')
但是得到這個錯誤
---------------------------------------------------------------------------
KeyError Traceback (most recent call last)
<ipython-input-114-d1e27c8b07ce> in <module>
----> 1 a = a.resample('12h', on='datetime')
/anaconda/envs/azureml_py36/lib/python3.6/site-packages/pandas/core/generic.py in resample(self, rule, how, axis, fill_method, closed, label, convention, kind, loffset, limit, base, on, level)
8447 base=base,
8448 key=on,
-> 8449 level=level,
8450 )
8451 return _maybe_process_deprecations(
/anaconda/envs/azureml_py36/lib/python3.6/site-packages/pandas/core/resample.py in resample(obj, kind, **kwds)
1304 """
1305 tg = TimeGrouper(**kwds)
-> 1306 return tg._get_resampler(obj, kind=kind)
1307
1308
/anaconda/envs/azureml_py36/lib/python3.6/site-packages/pandas/core/resample.py in _get_resampler(self, obj, kind)
1428
1429 """
-> 1430 self._set_grouper(obj)
1431
1432 ax = self.ax
/anaconda/envs/azureml_py36/lib/python3.6/site-packages/pandas/core/groupby/grouper.py in _set_grouper(self, obj, sort)
171 else:
172 if key not in obj._info_axis:
--> 173 raise KeyError("The grouper name {0} is not found".format(key))
174 ax = Index(obj[key], name=key)
175
KeyError: 'The grouper name datetime is not found'
我不知道如何在滾動視窗平均值后應用重采樣。
uj5u.com熱心網友回復：
使用level替代on和補充聚合函式：
a = a.resample('12H', level='datetime').mean()
編輯：
a = (telemetry.set_index('datetime')
.groupby('machineID')
.rolling(window=1, freq='24h')['volt']
.mean())
b = a.groupby(['machineID', pd.Grouper(freq='12H', level='datetime')]).mean()
轉載請註明出處，本文鏈接：https://www.uj5u.com/shujuku/382529.html