issue_comments: 284367479

This data as json

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/1289#issuecomment-284367479 https://api.github.com/repos/pydata/xarray/issues/1289 284367479 MDEyOklzc3VlQ29tbWVudDI4NDM2NzQ3OQ== 6815844 2017-03-06T11:08:09Z 2017-03-06T11:08:09Z MEMBER

we try to do it when possible because it leads to more predictable results.

OK. I totally agree with this. (Today I experienced a bug caused by unexpectedly aligned data_vars in other project. I realized predictable behavior is really important :) )

I think this PR is ready for review. If anyone has comments about the following behavior, I'm happy to update the code. + Only data_vars depending on the given rolling-dim are reduced. The others remain unchanged but are still stored in the reduced dataset (not dropped). + If rolling-dim is not included in Dataset.dims, this class raises KeyErrror. + The order of data_vars remain unchanged even after reduce(). + All the attrs are dropped.

 
{
    "total_count": 0,
    "+1": 0,
    "-1": 0,
    "laugh": 0,
    "hooray": 0,
    "confused": 0,
    "heart": 0,
    "rocket": 0,
    "eyes": 0
}
   211323643