Pandas: resample categorical index data

Question

Assume some measurement data (in reality given about every minute) named logData :

import pandas as pd, numpy as np

idxData = pd.to_datetime(['08:00', '08:15', '08:30', '08:45', '09:00'])
logData = pd.DataFrame(np.array([1.0, 2.0, 3.0, 4.0, 5.0]), columns=['val'], index=idxData)
idxRng  = pd.interval_range(idxData[0], idxData[-1], freq='30min')
avgData = logData.groupby( pd.cut(logData.index, idxRng) ).mean()

The data is grouped into avgData eg looking like this:

                      val
(08:00:00, 08:30:00]  2.5
(08:30:00, 09:00:00]  4.5

This downsampled avgData should now (after performing some other calculations) be upsampled again, eg to a frequency of freq='10min' for further calculations. Since avgData.resample('10min') throws the following error, the question is how to resample categorical data ?

TypeError: Only valid with DatetimeIndex, TimedeltaIndex or PeriodIndex, but got an instance of 'CategoricalIndex'

Many thanks in advance!

Answer 1

为了使重新采样工作，您的索引需要具有 datetime64[ns] 数据类型 通过运行以下代码检查索引的数据类型。

avgData.index.dtype

Answer 2

It took my a little while to figure out how to meaningfully convert a categorical index, but index.categories.mid seems to work, allowing to resample the data via

avgData.set_index( pd.DatetimeIndex( avgData.index.categories.mid ), inplace=True)
avgData = avgData.resample('5min').interpolate(method='nearest')

which yields the expected result:

          val
08:15:00  2.5
08:20:00  2.5
08:25:00  2.5
08:30:00  2.5
08:35:00  4.5
08:40:00  4.5
08:45:00  4.5