reset_index + set_index with drop=False

[cancan101]

Is there a simple way to make the Index in a DataFrame into columns in a while still keeping the index?

Currently I do the following and am wondering if this is the best I can do:

df.reset_index().set_index(df.index.names, drop=False)

[jreback]

In [16]: df = DataFrame(np.random.randn(5,2),columns=['A','B'],index=date_range('20130101',periods=5))

In [17]: df
Out[17]: 
                   A         B
2013-01-01 -0.075832 -0.571500
2013-01-02 -0.484581 -0.208359
2013-01-03 -0.579949  2.218587
2013-01-04 -0.636561 -0.741129
2013-01-05  0.629130  0.693871

[5 rows x 2 columns]

In [18]: df['the_index'] = df.index

In [19]: df
Out[19]: 
                   A         B  the_index
2013-01-01 -0.075832 -0.571500 2013-01-01
2013-01-02 -0.484581 -0.208359 2013-01-02
2013-01-03 -0.579949  2.218587 2013-01-03
2013-01-04 -0.636561 -0.741129 2013-01-04
2013-01-05  0.629130  0.693871 2013-01-05

[5 rows x 3 columns]

[cancan101]

@jreback What about multi indexes?

[cancan101]

Also strictly speaking the answer you proposed would be:

df2=df.copy()
df2['the_index'] = df2.index

[jreback]

@cancan101 you said Index. !

[cancan101]

That is true. The reason I used names was for a general purpose solution.

All that being said, what is the best bet for a multi index?

Also df.reset_index().set_index(df.index.names, drop=False) isn't quite right if the name if the Index is None

[jreback]

In [55]: df2 = df.reset_index()

In [56]: df2.index = df.index

In [57]: df2
Out[57]: 
             level_0    level_1         A         B
a 2013-01-01       a 2013-01-01 -0.592502  0.851388
  2013-01-02       a 2013-01-02 -1.358005  0.118389
  2013-01-03       a 2013-01-03 -1.835907  0.284208
b 2013-01-01       b 2013-01-01 -0.009812  0.421922
  2013-01-02       b 2013-01-02  0.662348  0.312630
  2013-01-03       b 2013-01-03  1.368136  0.859343

[6 rows x 4 columns]