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/1217#issuecomment-273687248,https://api.github.com/repos/pydata/xarray/issues/1217,273687248,MDEyOklzc3VlQ29tbWVudDI3MzY4NzI0OA==,1217238,2017-01-19T05:42:25Z,2017-01-19T05:43:22Z,MEMBER,"For reference, it may be helpful to try your example on a smaller dataset:
```
ds = xr.Dataset()
ds['data1'] = xr.DataArray(np.arange(100), coords={'t1': np.linspace(0, 1, 100)})
ds['data1b'] = xr.DataArray(np.arange(100, 200), coords={'t1': np.linspace(0, 1, 100)})
ds['data2'] = xr.DataArray(np.arange(200, 500), coords={'t2': np.linspace(0, 1, 300)})
ds['data2b'] = xr.DataArray(np.arange(600, 900), coords={'t2': np.linspace(0, 1, 300)})
```
This is what the data looks like:
```
>>> ds
<xarray.Dataset>
Dimensions:  (t1: 100, t2: 300)
Coordinates:
  * t1       (t1) float64 0.0 0.0101 0.0202 0.0303 0.0404 0.05051 0.06061 ...
  * t2       (t2) float64 0.0 0.003344 0.006689 0.01003 0.01338 0.01672 ...
Data variables:
    data1    (t1) int64 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 ...
    data1b   (t1) int64 100 101 102 103 104 105 106 107 108 109 110 111 112 ...
    data2    (t2) int64 200 201 202 203 204 205 206 207 208 209 210 211 212 ...
    data2b   (t2) int64 600 601 602 603 604 605 606 607 608 609 610 611 612 ...
```
and here's what happens with `where`:
```
>>> ds.where(ds.data1 < 50, drop=True)
<xarray.Dataset>
Dimensions:  (t1: 50, t2: 300)
Coordinates:
  * t1       (t1) float64 0.0 0.0101 0.0202 0.0303 0.0404 0.05051 0.06061 ...
  * t2       (t2) float64 0.0 0.003344 0.006689 0.01003 0.01338 0.01672 ...
Data variables:
    data1    (t1) float64 0.0 1.0 2.0 3.0 4.0 5.0 6.0 7.0 8.0 9.0 10.0 11.0 ...
    data1b   (t1) float64 100.0 101.0 102.0 103.0 104.0 105.0 106.0 107.0 ...
    data2    (t2, t1) float64 200.0 200.0 200.0 200.0 200.0 200.0 200.0 ...
    data2b   (t2, t1) float64 600.0 600.0 600.0 600.0 600.0 600.0 600.0 ...
```

I suspect this probably isn't really doing what you want, unless you really want two-dimensional versions of `data2` and `data2b`. It probably makes more sense to subset out the relevant variables first.

Broadcasting producing gigantic arrays without any warning is really a NumPy issue, e.g., try`np.where(True, np.zeros((1000000, 1)), np.ones((1, 1000000)))` (this will probably crash your computer!).","{""total_count"": 0, ""+1"": 0, ""-1"": 0, ""laugh"": 0, ""hooray"": 0, ""confused"": 0, ""heart"": 0, ""rocket"": 0, ""eyes"": 0}",,201617371