Pandas：将带有 DatetimeIndex 的列表字典转换为 DataFrame

Question

I need some help from pandas experts :) I need to efficiently create DataFrame from dictionary that contains list of two DatetimeIndexes as values.

Here is short example of such dictionary:

import pandas as pd

d = {
  871:[pd.date_range('20180131', '20180331', freq='M', normalize=True),
       pd.date_range('20180228', '20180430', freq='M', normalize=True)],
  872:[pd.date_range('20180228', '20180331', freq='M', normalize=True),
       pd.date_range('20180331', '20180430', freq='M', normalize=True)]}

d
Out[3]: 
{871: [DatetimeIndex(['2018-01-31', '2018-02-28', '2018-03-31'], dtype='datetime64[ns]', freq='M'),
  DatetimeIndex(['2018-02-28', '2018-03-31', '2018-04-30'], dtype='datetime64[ns]', freq='M')],
 872: [DatetimeIndex(['2018-02-28', '2018-03-31'], dtype='datetime64[ns]', freq='M'),
  DatetimeIndex(['2018-03-31', '2018-04-30'], dtype='datetime64[ns]', freq='M')]}

What I want to get is a DataFrame where dict keys used as indexes and two pd.Timestamp columns that correspond to those two DatetimeIndexes from above.

Here is how it should look like:

result_df
Out[6]: 
         Start        End
871 2018-01-31 2018-02-28
871 2018-02-28 2018-03-31
871 2018-03-31 2018-04-30
872 2018-02-28 2018-03-31
872 2018-03-31 2018-04-30

PS Actual task is to split large DataFrame with date ranges, such as below, to single months.

df_original
Out[19]: 
           Start        End
Index                      
871   2018-01-31 2018-02-28
872   2018-02-28 2018-04-30

Large is several hundred thousand rows.

Answer 1

Use:

d1 = {k: pd.DataFrame(list(zip(*v)), columns=['Start','End']) for k, v in d.items()}
df = pd.concat(d1).reset_index(level=1, drop=True)
print (df)
         Start        End
871 2018-01-31 2018-02-28
871 2018-02-28 2018-03-31
871 2018-03-31 2018-04-30
872 2018-02-28 2018-03-31
872 2018-03-31 2018-04-30