sha,message,author_date,committer_date,raw_author,raw_committer,repo,author,committer
685ba0671785eb402d4ed8aa390c7421ed80e5c6,"Marks slow, flaky, and failing tests (#1336)

* Adds marked tests as slow, flakey, and optionalci

Tests are marked using the following py.test options:

 * @slow which is run on default, not run if `--skip-slow`.
 * @optionalci is run on default, not run if `--skip-optional-ci`
 * @flakey which is not run on default, run if `--run-flakey`

Note, the `optionalci` category is needed in order to mark
tests that are environment dependent when run on continuous
integration (CI) hardware, e.g., tests that fail due to
variable loads on travis CI.

* Spelling: flakey -> flaky

* Test suite adjustments

* Typos

* Remove bottleneck from most CI builds

* xfail for cross engine read write netcdf4

* Try fix: Closes open file (for Windows error)

```
WindowsError: [Error 32] The process cannot access the file because it
is being used by another process
```

* Try fix: Closes open file (for Windows error)

```
WindowsError: [Error 32] The process cannot access the file because it
is being used by another process
```",2017-04-03T05:30:15Z,2017-04-03T05:30:15Z,b163753c1c9d5c0aca64dc350d604e668f1a8995,f10b21bed2846b879806f87039b77245b18e7671,13221727,4295853,1217238
56cec4613715fb77d86fe3b2bd88137adc97e74d,"Ensures drop=True case works with empty mask (#1342)

* Added test for where(mask, drop=True) [mask empty]

Ensures ds.where(mask, drop=True) works when the mask
is empty.

* Fix .get_index() for size 0 dimensions

This resolves issues with `where` with
`drop=True` with empty mask.",2017-04-02T22:43:52Z,2017-04-02T22:43:52Z,b163753c1c9d5c0aca64dc350d604e668f1a8995,f10b21bed2846b879806f87039b77245b18e7671,13221727,4295853,1217238
c0178b7b8b385dc65482099b1ff87ec81f7c51c2,"Ensures open_mfdataset attributes are retained (#1038)

Uses attributes from first file opened by
`open_mfdataset` to populate ds.attrs.",2017-03-31T03:11:07Z,2017-03-31T03:11:07Z,b163753c1c9d5c0aca64dc350d604e668f1a8995,f10b21bed2846b879806f87039b77245b18e7671,13221727,4295853,1217238
371d034372bc7522098a142a0debf93916c49102,"Fixes open_mfdataset too many open file error (#1198)

Includes testing to demonstrate an OSError associated
with opening too many files as encountered
using open_mfdataset.

Fixed for the following backends:
 * netCDF4 backend
 * scipy backend
 * pynio backend

Open/close operations on h5netcdf appear to have an
error associated with the h5netcdf library following
correspondence with @shoyer.
Thus, there are still challenges with h5netcdf;
hence, support for h5netcdf is currently disabled.

Note, by default `autoclose=False` for open_mfdataset so standard
behavior is unchanged unless `autoclose=True`.

This choice of default is to select standard xarray performance over
general removal of the OSError associated with opening too many files as
encountered using open_mfdataset.",2017-03-23T19:20:02Z,2017-03-23T19:20:02Z,b163753c1c9d5c0aca64dc350d604e668f1a8995,f10b21bed2846b879806f87039b77245b18e7671,13221727,4295853,1217238
9cf107b522188e206bd15f4620bc3b55ca30e616,"Adds cummulative operators to API  (#812)

* Adds nancumsum, nancumprod for numpy compatability

Needed until numpy v1.12, see
https://github.com/numpy/numpy/issues/7421

* Adds nancumsum, nancumprod to xarray functions",2016-10-03T21:05:32Z,2016-10-03T21:05:32Z,b163753c1c9d5c0aca64dc350d604e668f1a8995,f10b21bed2846b879806f87039b77245b18e7671,13221727,4295853,1217238
573541e30ee3136237642cc30f4004cd4af66281,Fixes documentation typo (#1032),2016-10-03T21:03:07Z,2016-10-03T21:03:07Z,b163753c1c9d5c0aca64dc350d604e668f1a8995,f10b21bed2846b879806f87039b77245b18e7671,13221727,4295853,1217238
c10a9df6522eca2c4d0942703f2b6fa21f9c8776,Fixes doc formating error (#1031),2016-10-03T16:15:19Z,2016-10-03T16:15:19Z,b163753c1c9d5c0aca64dc350d604e668f1a8995,f10b21bed2846b879806f87039b77245b18e7671,13221727,4295853,1217238
9f5a6ac3feb35feb7202aa00e3666c9310bab42c,Fixes doc typo (#845),2016-05-06T16:35:05Z,2016-05-06T16:35:05Z,b163753c1c9d5c0aca64dc350d604e668f1a8995,f10b21bed2846b879806f87039b77245b18e7671,13221727,4295853,1217238
28ee86afa683ce7c77e339d15d1220503df3d0ee,"Add drop=True option for where on Dataset and DataArray (#815)

By default ``where`` maintains the original size of the data.  For cases
where the selected data size is much smaller than the original data, use
of the option ``drop=True`` clips coordinate elements that are fully
masked.",2016-04-10T00:33:00Z,2016-04-10T00:33:00Z,b163753c1c9d5c0aca64dc350d604e668f1a8995,f10b21bed2846b879806f87039b77245b18e7671,13221727,4295853,1217238
759db6e17958b9e5800560cdf5cf46bf16ad5584,"Adds dask lock capability for backend writes

This fixes an error on an asynchronous write for `to_netcdf`
resulting in an `dask.async.RuntimeError: NetCDF: HDF error`

Resolves issue https://github.com/pydata/xarray/issues/793
following dask improvement at https://github.com/dask/dask/pull/1053
following advice of @shoyer.",2016-03-22T14:56:32Z,2016-03-22T22:12:54Z,0a7f3c7fc999b672b213b274bce2a60c16b84694,0a7f3c7fc999b672b213b274bce2a60c16b84694,13221727,4295853,4295853