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/2969#issuecomment-494059784,https://api.github.com/repos/pydata/xarray/issues/2969,494059784,MDEyOklzc3VlQ29tbWVudDQ5NDA1OTc4NA==,8881170,2019-05-20T16:30:02Z,2019-05-20T16:30:02Z,CONTRIBUTOR,Thanks for the feedback and link to the other issue. I wasn't sure what to search to find other issues on this. The coordinate transformation seems like the most straightforward approach. ,"{""total_count"": 0, ""+1"": 0, ""-1"": 0, ""laugh"": 0, ""hooray"": 0, ""confused"": 0, ""heart"": 0, ""rocket"": 0, ""eyes"": 0}",,445175953
https://github.com/pydata/xarray/issues/2969#issuecomment-493249273,https://api.github.com/repos/pydata/xarray/issues/2969,493249273,MDEyOklzc3VlQ29tbWVudDQ5MzI0OTI3Mw==,2448579,2019-05-16T22:10:32Z,2019-05-16T22:10:32Z,MEMBER,"1. This has been reported before e.g. #1234 . I think we need a kwarg to control this behaviour. You can avoid this here by making `dateVar` a coordinate which seems logical in this case.
`ds = ds.rename({'dateVar': 'time'}).set_coords('time')`

2. This is because `where` fills in NaNs which is not representable by variables with int datatype. 
    So xarray casts the variable to float. I guess we could avoid this when `drop=True` but not when `drop=False`. You can fix this partially by setting `indexVar` to be a co-ordinate which again seems logical in this case.


```
ds = (ds.rename({'dateVar': 'time', 
                 'indexVar': 'N'})
      .set_coords(['time', 'N']))

print(ds)

ds.where(ds.N > 50, drop=True)
```

gives me

```
<xarray.Dataset>
Dimensions:   (N: 100, time: 120)
Coordinates:
  * N         (N) int64 0 1 2 3 4 5 6 7 8 9 10 ... 90 91 92 93 94 95 96 97 98 99
  * time      (time) datetime64[ns] 2005-02-01 2005-02-02 ... 2005-05-31
Data variables:
    floatVar  (time, N) float64 0.2382 0.9297 0.5391 ... 0.9527 0.6448 0.08718
    intVar    (time, N) int64 9 5 7 9 9 3 3 2 2 8 4 1 ... 8 9 1 8 1 8 9 5 7 4 4

<xarray.Dataset>
Dimensions:   (N: 49, time: 120)
Coordinates:
  * N         (N) int64 51 52 53 54 55 56 57 58 59 ... 92 93 94 95 96 97 98 99
  * time      (time) datetime64[ns] 2005-02-01 2005-02-02 ... 2005-05-31
Data variables:
    floatVar  (time, N) float64 0.5974 0.09831 0.3858 ... 0.9527 0.6448 0.08718
    intVar    (time, N) float64 3.0 8.0 7.0 5.0 4.0 4.0 ... 9.0 5.0 7.0 4.0 4.0
```

`intVar` has still been converted to float so you'll need to fix that manually.","{""total_count"": 0, ""+1"": 0, ""-1"": 0, ""laugh"": 0, ""hooray"": 0, ""confused"": 0, ""heart"": 0, ""rocket"": 0, ""eyes"": 0}",,445175953