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/637#issuecomment-268204399,https://api.github.com/repos/pydata/xarray/issues/637,268204399,MDEyOklzc3VlQ29tbWVudDI2ODIwNDM5OQ==,10050469,2016-12-20T10:08:35Z,2016-12-20T10:08:35Z,MEMBER,closed via https://github.com/pydata/xarray/pull/1168,"{""total_count"": 0, ""+1"": 0, ""-1"": 0, ""laugh"": 0, ""hooray"": 0, ""confused"": 0, ""heart"": 0, ""rocket"": 0, ""eyes"": 0}",,112928260
https://github.com/pydata/xarray/pull/637#issuecomment-233201657,https://api.github.com/repos/pydata/xarray/issues/637,233201657,MDEyOklzc3VlQ29tbWVudDIzMzIwMTY1Nw==,1217238,2016-07-17T20:17:49Z,2016-07-17T20:17:49Z,MEMBER,"@robintw The challenge is that we want to handle `aspect` consistently even if we're doing a multi-faceted plot. Right now, we use the opposite convention (to matplotlib) for multi-faceted plots. So either:
1. We change the existing behavior for muilt-panel plots (and become inconsistent with Seaborn).
2. We change the existing behavior of `aspect` for single panel plots (and become inconsistent with matplotlib).

I would lean toward (2), currently implemented in this PR given that I think `aspect` is used more often for muilt-panel plots, but I'm OK with either option. If we pick (1), then we'll need to adjust both this PR and the existing behavior for multi-panel plots.
","{""total_count"": 0, ""+1"": 0, ""-1"": 0, ""laugh"": 0, ""hooray"": 0, ""confused"": 0, ""heart"": 0, ""rocket"": 0, ""eyes"": 0}",,112928260
https://github.com/pydata/xarray/pull/637#issuecomment-152744407,https://api.github.com/repos/pydata/xarray/issues/637,152744407,MDEyOklzc3VlQ29tbWVudDE1Mjc0NDQwNw==,2443309,2015-10-31T15:43:07Z,2015-10-31T15:43:07Z,MEMBER,"Bummer.   I'll just say I think we need to stay with the `matplotlib` convention. We require `matplotlib` as a dependency for plotting, not `seaborn`.  We are also passing the `aspect` argument to a `matplotlib` function/class.  I don't think it makes sense to say the `aspect` argument is passed to the `matplotlib` figure creator... after we revere itss order. 
","{""total_count"": 0, ""+1"": 0, ""-1"": 0, ""laugh"": 0, ""hooray"": 0, ""confused"": 0, ""heart"": 0, ""rocket"": 0, ""eyes"": 0}",,112928260
https://github.com/pydata/xarray/pull/637#issuecomment-152676248,https://api.github.com/repos/pydata/xarray/issues/637,152676248,MDEyOklzc3VlQ29tbWVudDE1MjY3NjI0OA==,1217238,2015-10-31T00:00:21Z,2015-10-31T00:00:35Z,MEMBER,"Ugh -- the `aspect` argument in matplotlib (e.g., on `imshow`) uses the opposite convention from Seaborn (https://github.com/mwaskom/seaborn/issues/746). So we'll have to break with at least one of those libraries to make this consistent.
","{""total_count"": 0, ""+1"": 0, ""-1"": 0, ""laugh"": 0, ""hooray"": 0, ""confused"": 0, ""heart"": 0, ""rocket"": 0, ""eyes"": 0}",,112928260