html_url,issue_url,id,node_id,user,created_at,updated_at,author_association,body,reactions,performed_via_github_app,issue
https://github.com/pydata/xarray/pull/134#issuecomment-43548881,https://api.github.com/repos/pydata/xarray/issues/134,43548881,MDEyOklzc3VlQ29tbWVudDQzNTQ4ODgx,1217238,2014-05-19T20:00:48Z,2014-05-19T20:00:48Z,MEMBER,"The reason I cast all datetime64 to datetime64[ns] is because pandas will not let you make an Index of datetime64 objects with anything other than ns precision. If you try to make it an Index with dtype=object you'll actually get an array of datetime.datetime objects:

```
>>> pd.Index(pd.date_range('2000-01-01', periods=5).values.astype('datetime64[us]'), dtype='object').values
array([datetime.datetime(2000, 1, 1, 0, 0),
       datetime.datetime(2000, 1, 2, 0, 0),
       datetime.datetime(2000, 1, 3, 0, 0),
       datetime.datetime(2000, 1, 4, 0, 0),
       datetime.datetime(2000, 1, 5, 0, 0)], dtype=object)
```

But I do agree this is not terribly consistent nor fully thought through. And it should certainly be well-documented.
","{""total_count"": 0, ""+1"": 0, ""-1"": 0, ""laugh"": 0, ""hooray"": 0, ""confused"": 0, ""heart"": 0, ""rocket"": 0, ""eyes"": 0}",,33772168
https://github.com/pydata/xarray/pull/134#issuecomment-43545468,https://api.github.com/repos/pydata/xarray/issues/134,43545468,MDEyOklzc3VlQ29tbWVudDQzNTQ1NDY4,514053,2014-05-19T19:27:26Z,2014-05-19T19:27:26Z,CONTRIBUTOR,"Also worth considering: how should datetime64[us] datetimes be handled?  Currently they get cast to [ns] which, since datetimes do not, could get confusing.
","{""total_count"": 0, ""+1"": 0, ""-1"": 0, ""laugh"": 0, ""hooray"": 0, ""confused"": 0, ""heart"": 0, ""rocket"": 0, ""eyes"": 0}",,33772168
https://github.com/pydata/xarray/pull/134#issuecomment-43538733,https://api.github.com/repos/pydata/xarray/issues/134,43538733,MDEyOklzc3VlQ29tbWVudDQzNTM4NzMz,1217238,2014-05-19T18:23:53Z,2014-05-19T18:23:53Z,MEMBER,"I agree, we should either ensure datetime64[ns] or ensure that operations on datetime objects preserve dtype. The latest commit should verify this.

My main reason for not doing the former is that I thought it would be nice (in theory) to support using plain datetime objects if datetime64[ns] does not have a long enough time range for some users. But for most users, I suspect they would indeed rather have datetime64[ns].
","{""total_count"": 0, ""+1"": 0, ""-1"": 0, ""laugh"": 0, ""hooray"": 0, ""confused"": 0, ""heart"": 0, ""rocket"": 0, ""eyes"": 0}",,33772168
https://github.com/pydata/xarray/pull/134#issuecomment-43535578,https://api.github.com/repos/pydata/xarray/issues/134,43535578,MDEyOklzc3VlQ29tbWVudDQzNTM1NTc4,514053,2014-05-19T17:55:00Z,2014-05-19T17:55:00Z,CONTRIBUTOR,"Yeah all fixed.

In #125 I went the route of forcing datetimes to be datetime64[ns].  This is probably part of a broader conversation, but doing so might save some future headaches.  Of course ... it would also restrict us to nanosecond precision.  Basically I feel like we should either force datetimes to be datetime64[ns] or make sure that operations on datetime objects preserve their type.

Probably worth getting this in and picking that conversation back up if needed.  In which case could you add tests which make sure variables with datetime objects are still datetime objects after concatenation?  If those start getting cast to datetime[ns] it'll start get confusing for users.
","{""total_count"": 0, ""+1"": 0, ""-1"": 0, ""laugh"": 0, ""hooray"": 0, ""confused"": 0, ""heart"": 0, ""rocket"": 0, ""eyes"": 0}",,33772168
https://github.com/pydata/xarray/pull/134#issuecomment-43531956,https://api.github.com/repos/pydata/xarray/issues/134,43531956,MDEyOklzc3VlQ29tbWVudDQzNTMxOTU2,1217238,2014-05-19T17:23:02Z,2014-05-19T17:23:02Z,MEMBER,"@akleeman I have a work-around up and ready for testing.
","{""total_count"": 0, ""+1"": 0, ""-1"": 0, ""laugh"": 0, ""hooray"": 0, ""confused"": 0, ""heart"": 0, ""rocket"": 0, ""eyes"": 0}",,33772168
https://github.com/pydata/xarray/pull/134#issuecomment-43478250,https://api.github.com/repos/pydata/xarray/issues/134,43478250,MDEyOklzc3VlQ29tbWVudDQzNDc4MjUw,1217238,2014-05-19T08:42:55Z,2014-05-19T08:45:03Z,MEMBER,"The later is a (newly exposed) bug in `pandas.isnull`. I'm waiting on Travis before submitting the PR upstream but here is the fix: https://github.com/shoyer/pandas/compare/isnull-0d-object-array

Looks like we'll need to add a work-around for now...
","{""total_count"": 0, ""+1"": 0, ""-1"": 0, ""laugh"": 0, ""hooray"": 0, ""confused"": 0, ""heart"": 0, ""rocket"": 0, ""eyes"": 0}",,33772168
https://github.com/pydata/xarray/pull/134#issuecomment-43476062,https://api.github.com/repos/pydata/xarray/issues/134,43476062,MDEyOklzc3VlQ29tbWVudDQzNDc2MDYy,1217238,2014-05-19T08:15:00Z,2014-05-19T08:15:00Z,MEMBER,"With regards to datetime.datetime being converted into integers, the issue is that pandas currently does dtype inference when indexing an Index (https://github.com/pydata/pandas/issues/6370). Fortunately the next version of pandas (0.14), due out in a few weeks, stops doing this.

I had made some attempts to fix this previously but it was untested and only sort of worked. I think this latest commit should fix that up.

Let me check on that second issue...
","{""total_count"": 0, ""+1"": 0, ""-1"": 0, ""laugh"": 0, ""hooray"": 0, ""confused"": 0, ""heart"": 0, ""rocket"": 0, ""eyes"": 0}",,33772168
https://github.com/pydata/xarray/pull/134#issuecomment-43473367,https://api.github.com/repos/pydata/xarray/issues/134,43473367,MDEyOklzc3VlQ29tbWVudDQzNDczMzY3,514053,2014-05-19T07:35:53Z,2014-05-19T07:35:53Z,CONTRIBUTOR,"The reorganization does make things cleaner, but the behavior changed relative to #125.  In particular, while this patch fixes concatenation with datetime64 times it doesn't work with datetimes:

```
In [2]: dates = [datetime.datetime(2011, 1, i + 1) for i in range(10)]
In [3]: ds = xray.Dataset({'time': ('time', dates)})
In [4]: xray.Dataset.concat([ds.indexed(time=slice(0, 4)), ds.indexed(time=slice(4, 8))], 'time')['time'].values
Out[4]: 
array([1293840000000000000L, 1293926400000000000L, 1294012800000000000L,
       1294099200000000000L, 1294185600000000000L, 1294272000000000000L,
       1294358400000000000L, 1294444800000000000L], dtype=object)
```

and not sure if this was broken before or not ...

```
In [5]: xray.Dataset.concat([x for _, x in ds.groupby('time')], 'time')['time'].values
ValueError: cannot slice a 0-d array
```
","{""total_count"": 0, ""+1"": 0, ""-1"": 0, ""laugh"": 0, ""hooray"": 0, ""confused"": 0, ""heart"": 0, ""rocket"": 0, ""eyes"": 0}",,33772168