I have a pandas.DataFrame df1, indexed with a pandas.DateRange object.
If I have a d1 and d2, as datetimes, why does df[d1:d2]not work, and how can I obtain this slice?
Add a comment
|
2 Answers
This works:
In [25]: df.ix[d1:d2]
Out[25]:
A B C D
2000-01-10 1.149815 0.686696 -1.230991 -1.610557
2000-01-11 -1.296118 -0.172950 -0.603887 0.383690
2000-01-12 -1.034574 -0.523238 0.626968 0.471755
2000-01-13 -0.193280 1.857499 -0.046383 0.849935
2000-01-14 -1.043492 -0.820525 0.868685 -0.773050
2000-01-17 -1.622019 -0.363992 1.207590 0.577290
cf. http://pandas.pydata.org/pandas-docs/stable/indexing.html#advanced-indexing-with-labels
On first principles df[d1:d2] should work as it does for Series:
In [27]: df['A'][d1:d2]
Out[27]:
2000-01-10 1.149815
2000-01-11 -1.296118
2000-01-12 -1.034574
2000-01-13 -0.193280
2000-01-14 -1.043492
2000-01-17 -1.622019
Name: A
Creating an issue here: https://github.com/pydata/pandas/issues/946
Comments
Try the truncate method:
df.truncate(before=d1, after=d2)
It won't modify your original df and will return a truncated one.
From docs:
Function truncate a sorted DataFrame / Series before and/or after
some particular dates.
Parameters
----------
before : date
Truncate before date
after : date
Truncate after date
Returns
-------
truncated : type of caller
5 Comments
Paragon
Can you link to a source for this? I'm on pandas.pydata.org/pandas-docs/stable/… and I haven't found the truncate function.
saroele
It works, thanks. Is there a reason why the more convenient
df[d1:d2] doesnt work?saroele
@Paragon: here's the link to the description of truncate in the current documentation (v0.7.2): pandas.pydata.org/pandas-docs/stable/generated/…
Wes McKinney
@saroele no reason, just an API oversight. See my answer below-- if someone would contribute some docs about truncate that would be helpful.
eumiro
@saroele - I believe
df[d1:d2] would imply, that both d1 and d2 are contained in the index (just like in lists). before/after doesn't sound that restrictive.