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/1992#issuecomment-592913599,https://api.github.com/repos/pydata/xarray/issues/1992,592913599,MDEyOklzc3VlQ29tbWVudDU5MjkxMzU5OQ==,5635139,2020-02-29T07:22:00Z,2020-02-29T07:22:00Z,MEMBER,We made some decent progress with numbagg!,"{""total_count"": 0, ""+1"": 0, ""-1"": 0, ""laugh"": 0, ""hooray"": 0, ""confused"": 0, ""heart"": 0, ""rocket"": 0, ""eyes"": 0}",,305663416
https://github.com/pydata/xarray/issues/1992#issuecomment-592909477,https://api.github.com/repos/pydata/xarray/issues/1992,592909477,MDEyOklzc3VlQ29tbWVudDU5MjkwOTQ3Nw==,26384082,2020-02-29T06:43:03Z,2020-02-29T06:43:03Z,NONE,"In order to maintain a list of currently relevant issues, we mark issues as stale after a period of inactivity

If this issue remains relevant, please comment here or remove the `stale` label; otherwise it will be marked as closed automatically
","{""total_count"": 0, ""+1"": 0, ""-1"": 0, ""laugh"": 0, ""hooray"": 0, ""confused"": 0, ""heart"": 0, ""rocket"": 0, ""eyes"": 0}",,305663416
https://github.com/pydata/xarray/issues/1992#issuecomment-377669882,https://api.github.com/repos/pydata/xarray/issues/1992,377669882,MDEyOklzc3VlQ29tbWVudDM3NzY2OTg4Mg==,5635139,2018-03-31T06:19:33Z,2018-03-31T06:19:33Z,MEMBER,First attempt at https://github.com/shoyer/numbagg/pull/4,"{""total_count"": 0, ""+1"": 0, ""-1"": 0, ""laugh"": 0, ""hooray"": 0, ""confused"": 0, ""heart"": 0, ""rocket"": 0, ""eyes"": 0}",,305663416
https://github.com/pydata/xarray/issues/1992#issuecomment-373511743,https://api.github.com/repos/pydata/xarray/issues/1992,373511743,MDEyOklzc3VlQ29tbWVudDM3MzUxMTc0Mw==,1217238,2018-03-15T20:26:51Z,2018-03-15T20:26:51Z,MEMBER,"I started numbagg with the intent of using it for xarray, but never got around to finishing it up. I think it would be perfectly fine to add as an optional dependency. Mostly we are using bottleneck instead because it's already out there and is slightly more complete, but certainly writing new aggregation functions with numbagg rather than bottleneck would be much easier.

One thing I was also excited about doing with numbagg was writing fast grouped aggregation functions, e.g., to make xarray's groupby().mean() comparable in performance to pandas. I had some promising preliminary results, but now I can't find the notebook where I wrote them!

Out of curiosity, I ran the benchmark suite from the numbagg README again. Both Numba and bottleneck have been significantly improved over the past three years, but Numba now beats bottleneck across the board:
```
%timeit numbagg.nanmean(x)
1000 loops, best of 3: 1.92 ms per loop
%timeit numbagg.nanmean(x, axis=0)
100 loops, best of 3: 2.75 ms per loop
%timeit numbagg.nanmean(x, axis=1)
1000 loops, best of 3: 1.86 ms per loop

%timeit bottleneck.nanmean(x)
100 loops, best of 3: 2.75 ms per loop
%timeit bottleneck.nanmean(x, axis=0)
100 loops, best of 3: 3.47 ms per loop
%timeit bottleneck.nanmean(x, axis=1)
100 loops, best of 3: 2.72 ms per loop
```","{""total_count"": 1, ""+1"": 1, ""-1"": 0, ""laugh"": 0, ""hooray"": 0, ""confused"": 0, ""heart"": 0, ""rocket"": 0, ""eyes"": 0}",,305663416