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/issues/808#issuecomment-269569458,https://api.github.com/repos/pydata/xarray/issues/808,269569458,MDEyOklzc3VlQ29tbWVudDI2OTU2OTQ1OA==,2443309,2016-12-29T01:44:43Z,2016-12-29T01:44:43Z,MEMBER,see #1190 for a belated fix for this issue.,"{""total_count"": 0, ""+1"": 0, ""-1"": 0, ""laugh"": 0, ""hooray"": 0, ""confused"": 0, ""heart"": 0, ""rocket"": 0, ""eyes"": 0}",,144575316
https://github.com/pydata/xarray/issues/808#issuecomment-203468065,https://api.github.com/repos/pydata/xarray/issues/808,203468065,MDEyOklzc3VlQ29tbWVudDIwMzQ2ODA2NQ==,2443309,2016-03-30T14:46:49Z,2016-03-30T14:46:49Z,MEMBER,"Setuptools has the concept of `extras_require` but the application doesn't really work with our optional import of `bottleneck` ([ref](http://pythonhosted.org/setuptools/setuptools.html#declaring-extras-optional-features-with-their-own-dependencies)). I think what may work here is to check the `bottleneck` version on import and raise a warning if the version is not sufficient for all the functionality we use it for.

So [here](https://github.com/pydata/xarray/blob/master/xarray/core/ops.py#L20) may look something like:

``` Python
try:
    import bottleneck as bn
    has_bottleneck = True
    if bn.__version__ < 1.0:
        warnings.warn('insufficient bottleneck version...')
except ImportError:
    # use numpy methods instead
    bn = np
    has_bottleneck = False
```

Alternatively, we could raise an error or fall back to use the numpy methods.
","{""total_count"": 0, ""+1"": 0, ""-1"": 0, ""laugh"": 0, ""hooray"": 0, ""confused"": 0, ""heart"": 0, ""rocket"": 0, ""eyes"": 0}",,144575316