pandas - 重新采样 TimeSeries 时无法导入名称 'NaT'

Question

我有几年的数据，例如(作为 pandas TimeSeries):

1997-04-02     0.0
1997-04-03     0.0
1997-04-04     2.0
1997-04-05     2.0
1997-04-06     0.0
1997-04-07     0.0
1997-04-08     0.0

日期是 'pandas.core.indexes.datetimes.DatetimeIndex' 类型的索引

df.resample('Y').sum() 

给出错误“无法导入名称‘NaT’”。

错误的来源是什么？我尝试使用格式为 '%Y-%m-%d %H:%M:%S' 的 to_datetime，但没有帮助。

异常全文:

    ---------------------------------------------------------------------------
ImportError                               Traceback (most recent call last)
<ipython-input-342-df3b79446059> in <module>
     12     globals()['df_'+col]=nm #.resample('M').sum()
     13 df_dataKlin27417c=df_dataKlin27417.loc['1997-04-02':'1997-05-06']
---> 14 rex=df_dataKlin27417.resample('M').sum()
     15 #rex
     16 df_allst

~/anaconda3/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)
   7102         Show which entries in a DataFrame are not NA.
   7103 
-> 7104         >>> df = pd.DataFrame({'age': [5, 6, np.NaN],
   7105         ...                    'born': [pd.NaT, pd.Timestamp('1939-05-27'),
   7106         ...                             pd.Timestamp('1940-04-25')],

~/anaconda3/lib/python3.6/site-packages/pandas/core/resample.py in <module>
      7 
      8 from pandas._libs import lib
----> 9 from pandas._libs.tslibs import NaT, Timestamp
     10 from pandas._libs.tslibs.frequencies import is_subperiod, is_superperiod
     11 from pandas._libs.tslibs.period import IncompatibleFrequency

ImportError: cannot import name 'NaT'

Answer 1

我也有这个问题，环境是Linux Centos，但它适用于我的 Macbook 系统。我不知道它是否只发生在 Linux 上。