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/2063#issuecomment-384741776,https://api.github.com/repos/pydata/xarray/issues/2063,384741776,MDEyOklzc3VlQ29tbWVudDM4NDc0MTc3Ng==,1217238,2018-04-26T18:23:11Z,2018-04-26T18:23:11Z,MEMBER,"OK, so it looks like a dask.array issue. (I got runtime warnings, too, but I think those can be safely ignored.)","{""total_count"": 0, ""+1"": 0, ""-1"": 0, ""laugh"": 0, ""hooray"": 0, ""confused"": 0, ""heart"": 0, ""rocket"": 0, ""eyes"": 0}",,314670220
https://github.com/pydata/xarray/issues/2063#issuecomment-384707323,https://api.github.com/repos/pydata/xarray/issues/2063,384707323,MDEyOklzc3VlQ29tbWVudDM4NDcwNzMyMw==,1217238,2018-04-26T16:35:24Z,2018-04-26T18:21:40Z,MEMBER,"I think the following script should reproduce the issue and identify which library is the source of the issue: 
```
import dask.array as da
import numpy as np
import pandas as pd
import bottleneck

rng = np.random.RandomState(0)
x = np.concatenate([rng.randn(16, 8).astype(np.float32), np.nan * np.zeros((4, 8)).astype(np.float32)])

print('numpy', np.nanvar(x))
print('dask', da.nanvar(da.from_array(x, 4)).compute())
print('pandas', pd.Series(x.ravel()).var(ddof=0))
print('bottleneck', bottleneck.nanvar(x))
```

On my laptop, this prints:
```
numpy 1.0792575
dask 1.0792574
pandas 1.0792568922042847
bottleneck 1.0792568922042847
```","{""total_count"": 0, ""+1"": 0, ""-1"": 0, ""laugh"": 0, ""hooray"": 0, ""confused"": 0, ""heart"": 0, ""rocket"": 0, ""eyes"": 0}",,314670220
https://github.com/pydata/xarray/issues/2063#issuecomment-381651113,https://api.github.com/repos/pydata/xarray/issues/2063,381651113,MDEyOklzc3VlQ29tbWVudDM4MTY1MTExMw==,1217238,2018-04-16T15:44:05Z,2018-04-16T15:44:05Z,MEMBER,"Thanks for the report. This is likely an upstream bug but I can help direct this to the appropriate place.

Can you please share the output of `xarray.show_versions()`? At least, which versions (if any) of bottleneck, numpy and pandas do you have installed?","{""total_count"": 0, ""+1"": 0, ""-1"": 0, ""laugh"": 0, ""hooray"": 0, ""confused"": 0, ""heart"": 0, ""rocket"": 0, ""eyes"": 0}",,314670220