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/1576#issuecomment-330271312,https://api.github.com/repos/pydata/xarray/issues/1576,330271312,MDEyOklzc3VlQ29tbWVudDMzMDI3MTMxMg==,1217238,2017-09-18T16:04:47Z,2017-09-18T16:04:47Z,MEMBER,"We currently decode anything with a `_FillValue` attribute to float, so that we can convert any values equal to the fill value to NaN. This ensure's that xarray's NaN skipping aggregations (e.g., `mean()`) work properly.

However, this isn't really a useful thing to do for a dataset like this where the values really represent enums/categories. It seems like the CF compliant way to indicate this is with the various [flag_*](http://cfconventions.org/Data/cf-conventions/cf-conventions-1.7/cf-conventions.html#flags) attributes. So we could look for those to indicate that we shouldn't fill-in fill values.

Eventually, we could possibly also use this for decoding into a true ""categorical"" dtype, but numpy doesn't have anything like that yet.","{""total_count"": 0, ""+1"": 0, ""-1"": 0, ""laugh"": 0, ""hooray"": 0, ""confused"": 0, ""heart"": 0, ""rocket"": 0, ""eyes"": 0}",,258500654