issue_comments: 148276168

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/issues/624#issuecomment-148276168 https://api.github.com/repos/pydata/xarray/issues/624 148276168 MDEyOklzc3VlQ29tbWVudDE0ODI3NjE2OA== 1217238 2015-10-15T04:12:01Z 2015-10-15T04:12:01Z MEMBER

I agree that ds.roll(time=100) seems like the more consistent syntax.

shift seems like the natural companion method to roll. It should work the same way, but insert NaNs instead of other of the arrays. No need to add shift in this PR, of course, but it's a natural follow up.

Does it makes sense to roll in integer or coordinate space? e.g., should we be able to write ds.roll(longitude=180) regardless of the actual spatial resolution? I'm thinking it's probably better to stick with integers, simply because coordinates are not necessarily evenly spaced.

 
{
    "total_count": 0,
    "+1": 0,
    "-1": 0,
    "laugh": 0,
    "hooray": 0,
    "confused": 0,
    "heart": 0,
    "rocket": 0,
    "eyes": 0
}
   111471076