bump min deps for 0.15 #3713

dcherian · 2020-01-21T16:39:07Z

Closes Use numpy's normalize_axis_index, if available #1679
Passes black . && mypy . && flake8
Fully documented, including whats-new.rst for all changes and api.rst for new API

ci/requirements/doc.yml

crusaderky · 2020-01-21T16:46:48Z

ci/requirements/py36-min-all-deps.yml

@@ -40,7 +40,7 @@ dependencies:
  - pytest-cov
  - pytest-env
  - rasterio=1.0
-  - scipy=1.0  # Policy allows for 1.2, but scipy>=1.1 breaks numpy=1.14
+  - scipy=1.2  # Policy allows for 1.3


Why not use 1.3 then?

I was getting segmentation faults locally. I'll bump once I fix the tests.

crusaderky · 2020-01-21T16:48:53Z

doc/whats-new.rst

+  - scipy 1.3
+  - bottleneck 1.3
+  - h5py 2.10
+  - rasterio 1.1


py36-min-all-deps.yml should have matching changes for bottleneck, h5py, rasterio

Yeah I'm doing this incrementally since the env wasn't getting created locally.

crusaderky · 2020-01-21T16:49:04Z

doc/whats-new.rst

+  - bottleneck 1.3
+  - h5py 2.10
+  - rasterio 1.1
+  - pseudonetcdf


what about pseudonetcdf?

xarray/core/dataset.py

xarray/core/nputils.py

dcherian · 2020-01-21T17:03:15Z

@keewis: looks like pint doesn't have nanquantile? https://dev.azure.com/xarray/xarray/_build/results?buildId=1989&view=logs&jobId=78b48a04-306f-5a15-9ac3-dd2fdb28db5e&j=acc1347f-8235-55b0-95c7-0e2189e61659&t=7284452c-b1fa-561e-b473-09e477cf0d79

jthielen · 2020-01-21T17:08:39Z

looks like pint doesn't have nanquantile?

@dcherian Hope you don't mind me jumping in, but that is correct at the moment (see hgrecco/pint#981). It would be easy to add to Pint, though.

keewis · 2020-01-21T17:13:05Z

as @jthielen said, it does not since I never realized we used nanquantile (did we change that recently?). For now, I'd mark these tests as xfail(reason="not implemented by pint") since none of the pint-support PRs are merged yet. Can be a new PR, though.

This reverts commit d078cf1.

pep8speaks · 2020-01-21T17:13:35Z

Hello @dcherian! Thanks for updating this PR. We checked the lines you've touched for PEP 8 issues, and found:

There are currently no PEP 8 issues detected in this Pull Request. Cheers! 🍻

Comment last updated at 2020-01-24 17:47:42 UTC

dcherian · 2020-01-21T17:15:22Z

as @jthielen said, it does not since I never realized we used nanquantile (did we change that recently?).

I changed it in this PR. But I guess I could just revert that bit

keewis · 2020-01-21T17:17:22Z

I think it's fine if you mark the failing tests as xfail

crusaderky · 2020-01-22T12:51:48Z

Related: #3714

* 'master' of github.com:pydata/xarray: setuptools-scm and one-liner setup.py (pydata#3714)

* 'master' of github.com:pydata/xarray: apply_ufunc: Add meta kwarg + bump dask to 2.2 (pydata#3660)

dcherian · 2020-01-22T15:55:19Z

This output is confusing to me. It hasn't been 6 months since bottleneck 1.3 was released. Why is that the "policy version". Similarly for zarr 2.4

Package       Required             Policy               Status
------------- -------------------- -------------------- ------
python        3.6     (2016-12-23) 3.6     (2016-12-23) =
boto3         1.9     (2018-09-10) 1.10    (2019-10-22) <
bottleneck    1.2     (2017-09-18) 1.3     (2019-11-13) <
cartopy       0.17    (2018-11-17) 0.17    (2018-11-17) =
cdms2         3.1     (2018-11-02) 3.1     (2018-11-02) =
cfgrib        0.9     (2019-02-25) 0.9     (2019-02-25) =
cftime        1.0     (2018-04-27) 1.0     (2018-04-27) =
dask          1.2     (2019-04-13) 2.2     (2019-08-01) <
distributed   1.27    (2019-04-12) 2.2     (2019-08-01) <
h5netcdf      0.7     (2019-02-27) 0.7     (2019-02-27) =
h5py          2.9     (2018-12-20) 2.10    (2019-09-10) <
hdf5          1.10    (2017-09-18) 1.10    (2017-09-18) =
iris          2.2     (2018-10-12) 2.3     (2019-12-21) <
lxml          4.4     (2019-07-27) 4.4     (2019-07-27) =
matplotlib    3.1     (2019-05-18) 3.1     (2019-05-18) =
nc-time-axis  1.2     (2019-01-25) 1.2     (2019-01-25) =
netcdf4       1.4     (2018-05-08) 1.4     (2018-05-08) =
numba         0.44    (2019-06-18) 0.46    (2019-10-25) <
numpy         1.15    (2018-07-27) 1.15    (2018-07-27) =
pandas        0.25    (2019-07-18) 0.25    (2019-07-18) =
pseudonetcdf  3.0     (2019-07-18) 3.1     (2019-10-18) <
pydap         3.2     (2018-10-16) 3.2     (2018-10-16) =
pynio         1.5     (2017-12-01) 1.5     (2017-12-01) =
rasterio      1.0     (2018-12-21) 1.1     (2019-10-15) <
scipy         1.3     (2019-05-28) 1.3     (2019-05-28) =
seaborn       0.9     (2018-07-17) 0.9     (2018-07-17) =
toolz         0.10    (2019-07-12) 0.10    (2019-07-12) =
zarr          2.3     (2019-03-25) 2.4     (2020-01-11) <

crusaderky · 2020-01-23T18:40:29Z

Because NEP29 states support for all versions in the latest X months, not the most recent version released more than X months ago.

dcherian · 2020-01-24T17:47:10Z

bumping the following seems a little excessive to me.

bottleneck    1.2     (2017-09-18) 1.3     (2019-11-13) <
iris          2.2     (2018-10-12) 2.3     (2019-12-21) <
numba         0.44    (2019-06-18) 0.46    (2019-10-25) <
pseudonetcdf  3.0     (2019-07-18) 3.1     (2019-10-18) <
rasterio      1.0     (2018-12-21) 1.1     (2019-10-15) <
zarr          2.3     (2019-03-25) 2.4     (2020-01-11) <

Is anyone opposed to merging this PR in its current form and bumping up minimum versions at the next major release?

crusaderky · 2020-01-24T20:28:25Z

I have no problems with supporting versions older than those mandated by the policy as long as there isn't any major benefit in dropping them...

max-sixty · 2020-01-24T20:48:46Z

For sure no harm in keeping older ones for this release. But to the extent we can have a policy we agree on, then it's not a judgement call each release. Should we change the policy?

crusaderky · 2020-01-24T21:39:46Z

We could increase the support window for seldom-updated packages (read: not dask) to 1 year.

dcherian · 2020-01-24T21:45:43Z

the most recent version released more than X months ago.

To me, this seems most sensible.

EDIT: crusaderky's 1 year proposal sounds good too

* 'master' of github.com:pydata/xarray: (27 commits) bump min deps for 0.15 (pydata#3713) setuptools-scm and isort tweaks (pydata#3720) Allow binned coordinates on 1D plots y-axis. (pydata#3685) apply_ufunc: Add meta kwarg + bump dask to 2.2 (pydata#3660) setuptools-scm and one-liner setup.py (pydata#3714) Feature/align in dot (pydata#3699) ENH: enable `H5NetCDFStore` to work with already open h5netcdf.File a… (pydata#3618) One-off isort run (pydata#3705) hardcoded xarray.__all__ (pydata#3703) Bump mypy to v0.761 (pydata#3704) remove DataArray and Dataset constructor deprecations for 0.15 (pydata#3560) Tests for variables with units (pydata#3654) Add an example notebook using apply_ufunc to vectorize 1D functions (pydata#3629) Use encoding['dtype'] over data.dtype when possible within CFMaskCoder.encode (pydata#3652) allow passing any iterable to drop when dropping variables (pydata#3693) Typo on DataSet/DataArray.to_dict documentation (pydata#3692) Fix mypy type checking tests failure in ds.merge (pydata#3690) Explicitly convert result of pd.to_datetime to a timezone-naive type (pydata#3688) ds.merge(da) bugfix (pydata#3677) fix docstring for combine_first: returns a Dataset (pydata#3683) ...

* master: Add support for CFTimeIndex in get_clean_interp_index (pydata#3631) sel with categorical index (pydata#3670) bump min deps for 0.15 (pydata#3713) setuptools-scm and isort tweaks (pydata#3720) Allow binned coordinates on 1D plots y-axis. (pydata#3685) apply_ufunc: Add meta kwarg + bump dask to 2.2 (pydata#3660) setuptools-scm and one-liner setup.py (pydata#3714) Feature/align in dot (pydata#3699) ENH: enable `H5NetCDFStore` to work with already open h5netcdf.File a… (pydata#3618) One-off isort run (pydata#3705) hardcoded xarray.__all__ (pydata#3703) Bump mypy to v0.761 (pydata#3704) remove DataArray and Dataset constructor deprecations for 0.15 (pydata#3560) Tests for variables with units (pydata#3654) Add an example notebook using apply_ufunc to vectorize 1D functions (pydata#3629) Use encoding['dtype'] over data.dtype when possible within CFMaskCoder.encode (pydata#3652)

dcherian added 3 commits January 21, 2020 09:12

Bump numpy to 1.15, pandas to 0.25, scipy to 1.2

05c9443

Remove some backcompat code

3540398

bump pandas in doc.yml

d078cf1

dcherian mentioned this pull request Jan 21, 2020

release 0.15.0? #3702

Closed

11 tasks

crusaderky requested changes Jan 21, 2020

View reviewed changes

Fix repr test

fe0ae22

review cleanups.

e5653be

scipy to 1.3

f5d4fbe

Revert "bump pandas in doc.yml"

41e87d9

This reverts commit d078cf1.

bugfix.

fdc7830

fix tests.

d0d9c14

dcherian added 4 commits January 22, 2020 08:41

Merge branch 'master' of github.com:pydata/xarray into min-deps-15

b03f6ab

* 'master' of github.com:pydata/xarray: setuptools-scm and one-liner setup.py (pydata#3714)

update setup.cfg

c3e42a7

Merge branch 'master' of github.com:pydata/xarray into min-deps-15

4746492

* 'master' of github.com:pydata/xarray: apply_ufunc: Add meta kwarg + bump dask to 2.2 (pydata#3660)

Update whats-new

6e4d6eb

dcherian closed this Jan 24, 2020

dcherian reopened this Jan 24, 2020

crusaderky approved these changes Jan 24, 2020

View reviewed changes

dcherian merged commit 9c72866 into pydata:master Jan 24, 2020

dcherian deleted the min-deps-15 branch January 24, 2020 21:47

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

bump min deps for 0.15 #3713

bump min deps for 0.15 #3713

dcherian commented Jan 21, 2020 •

edited

Loading

crusaderky Jan 21, 2020

dcherian Jan 21, 2020

crusaderky Jan 21, 2020

dcherian Jan 21, 2020

crusaderky Jan 21, 2020

dcherian commented Jan 21, 2020

jthielen commented Jan 21, 2020

keewis commented Jan 21, 2020

pep8speaks commented Jan 21, 2020 •

edited

Loading

dcherian commented Jan 21, 2020

keewis commented Jan 21, 2020

crusaderky commented Jan 22, 2020

dcherian commented Jan 22, 2020

crusaderky commented Jan 23, 2020

dcherian commented Jan 24, 2020

crusaderky commented Jan 24, 2020

max-sixty commented Jan 24, 2020

crusaderky commented Jan 24, 2020

dcherian commented Jan 24, 2020 •

edited

Loading

bump min deps for 0.15 #3713

bump min deps for 0.15 #3713

Conversation

dcherian commented Jan 21, 2020 • edited Loading

crusaderky Jan 21, 2020

Choose a reason for hiding this comment

dcherian Jan 21, 2020

Choose a reason for hiding this comment

crusaderky Jan 21, 2020

Choose a reason for hiding this comment

dcherian Jan 21, 2020

Choose a reason for hiding this comment

crusaderky Jan 21, 2020

Choose a reason for hiding this comment

dcherian commented Jan 21, 2020

jthielen commented Jan 21, 2020

keewis commented Jan 21, 2020

pep8speaks commented Jan 21, 2020 • edited Loading

Comment last updated at 2020-01-24 17:47:42 UTC

dcherian commented Jan 21, 2020

keewis commented Jan 21, 2020

crusaderky commented Jan 22, 2020

dcherian commented Jan 22, 2020

crusaderky commented Jan 23, 2020

dcherian commented Jan 24, 2020

crusaderky commented Jan 24, 2020

max-sixty commented Jan 24, 2020

crusaderky commented Jan 24, 2020

dcherian commented Jan 24, 2020 • edited Loading

dcherian commented Jan 21, 2020 •

edited

Loading

pep8speaks commented Jan 21, 2020 •

edited

Loading

dcherian commented Jan 24, 2020 •

edited

Loading