forked from pool/python-xarray
Dirk Mueller
dff17e69ab
- Update to 2023.8.0
* This release brings changes to minimum dependencies, allows
reading of datasets where a dimension name is associated with a
multidimensional variable (e.g. finite volume ocean model
output), and introduces a new xarray.Coordinates object.
## Announcements
* The xarray.Variable class is being refactored out to a new
project title 'namedarray'. See the design doc for more
details. Reach out to us on this discussion topic if you have
any thoughts.
## What's Changed
* Use variable name in all exceptions raised in as_variable by
@ZedThree in #7995
* Allow opening datasets with nD dimenson coordinate variables.
by @dcherian in #7989
* join together duplicate entries in the text repr by @keewis in
#7225
* Expose "Coordinates" as part of Xarray's public API by @benbovy
in #7368
* Update interpolate_na in dataset.py by @ilgast in #7974
* Add HDF5 Section to read/write docs page by @rwegener2 in #8012
* Add examples to docstrings by @harshitha1201 in #7937
* (chore) min versions bump by @jhamman in #8022
* Automatically chunk other in GroupBy binary ops. by @dcherian
in #7684
* change cumproduct to cumprod by @quantsnus in #8031
* Reduce pre-commit update frequency to monthly from weekly. by
@dcherian in #8033
* sort when encoding coordinates for deterministic outputs by
@itcarroll in #8034
* Zarr : Allow setting write_empty_chunks by @RKuttruff in #8016
* Count documentation by @Articoking in #8057
* unpin numpy by @keewis in #8061
- Release 2023.7.0
* This release brings improvements to the documentation on
wrapping numpy-like arrays, improved docstrings, and bug fixes.
## What's Changed
* Allow cubed arrays to be passed to flox groupby by @TomNicholas
in #7941
* Duck array documentation improvements by @TomNicholas in #7911
* Docstring examples by @harshitha1201 in #7881
* Chunked array docs by @TomNicholas in #7951
* ensure no forward slashes in names for HDF5-based backends by
@kmuehlbauer in #7953
* Move absolute path finder from open_mfdataset to own function
by @Illviljan in #7968
* Skip broken tests on Python 3.11 and Windows by @Illviljan in
#7972
* Examples added to docstrings by @harshitha1201 in #7936
* Fix typo in zarr.py by @johmathe in #7983
* Improve explanation in example "Working with Multidimensional
Coordinates" by @yvonnefroehlich in #7984
* Remove hue_style from plot1d docstring by @Illviljan in #7925
- Release 2023.06.0
* This release adds features to curvefit, improves the
performance of concatenation, and fixes various bugs.
## What's Changed
* Array API fixes for astype by @TomNicholas in #7847
* Add type hints to test_dtypes by @Illviljan in #7858
* adapt the pint + dask test to the newest version of pint by
@keewis in #7855
* Avoid explicit loop when updating OrderedSet by @Illviljan in
#7857
* Improve to_dask_dataframe performance by @Illviljan in #7844
* Changed duck typing exception to: (ImportError, AttributeError)
by @vhaasteren in #7874
* defer to numpy for the expected result by @keewis in #7875
* deprecate the cdms2 conversion methods by @keewis in #7876
* Implement multidimensional initial guess and bounds for
curvefit by @mgunyho in #7821
* Improve concat performance by @Illviljan in #7824
* Fix .groupby(multi index level) by @dcherian in #7830
* don't use CacheFileManager.__del__ on interpreter shutdown by
@keewis in #7880
* fix polyfit changing the original object by @malmans2 in #7900
* Fix flaky doctest for curvefit by @mgunyho in #7893
* Ensure dtype of reindex result matches dtype of the original
DataArray by @andersy005 in #7917
* Add errors option to curvefit by @mgunyho in #7891
* CF encoding should preserve vlen dtype for empty arrays by
@tomwhite in #7862
* use trusted publishers instead of a API token by @keewis in
#7899
* Add cfgrib,ipywidgets to doc env by @dcherian in #7888
* Fix regression with grouper object on an IndexVariable by
@mwtoews in #7920
* Fix check for chunk_store in zarr backend by @juntyr in #7923
* fixing rolling_window issue with cupy by @negin513 in #7938
- Add xarray-pr8139-pandas-fill_value.patch gh#pydata/xarray#8125,
gh#pydata/xarray#8139
OBS-URL: https://build.opensuse.org/request/show/1110376
OBS-URL: https://build.opensuse.org/package/show/devel:languages:python:numeric/python-xarray?expand=0&rev=82
39 lines
1.7 KiB
Diff
39 lines
1.7 KiB
Diff
From 3e48312c41cb26bbd4dc83ae286444f1469b7aa2 Mon Sep 17 00:00:00 2001
|
|
From: Maximilian Roos <m@maxroos.com>
|
|
Date: Fri, 1 Sep 2023 19:41:41 -0700
|
|
Subject: [PATCH] Fix pandas interpolate(fill_value=) error
|
|
|
|
Pandas no longer has a `fill_value` parameter for `interpolate`.
|
|
|
|
Weirdly I wasn't getting this locally, on pandas 2.1.0, only in CI on https://github.com/pydata/xarray/actions/runs/6054400455/job/16431747966?pr=8138.
|
|
|
|
Removing it passes locally, let's see whether this works in CI
|
|
---
|
|
xarray/tests/test_missing.py | 6 ++++--
|
|
1 file changed, 4 insertions(+), 2 deletions(-)
|
|
|
|
diff --git a/xarray/tests/test_missing.py b/xarray/tests/test_missing.py
|
|
index a6b6b1f80c..c303659116 100644
|
|
--- a/xarray/tests/test_missing.py
|
|
+++ b/xarray/tests/test_missing.py
|
|
@@ -104,7 +104,8 @@ def test_interpolate_pd_compat():
|
|
for dim in ["time", "x"]:
|
|
actual = da.interpolate_na(method=method, dim=dim, fill_value=np.nan)
|
|
expected = df.interpolate(
|
|
- method=method, axis=da.get_axis_num(dim), fill_value=(np.nan, np.nan)
|
|
+ method=method,
|
|
+ axis=da.get_axis_num(dim),
|
|
)
|
|
# Note, Pandas does some odd things with the left/right fill_value
|
|
# for the linear methods. This next line inforces the xarray
|
|
@@ -140,7 +141,8 @@ def test_interpolate_pd_compat_non_uniform_index():
|
|
method="linear", dim=dim, use_coordinate=True, fill_value=np.nan
|
|
)
|
|
expected = df.interpolate(
|
|
- method=method, axis=da.get_axis_num(dim), fill_value=np.nan
|
|
+ method=method,
|
|
+ axis=da.get_axis_num(dim),
|
|
)
|
|
|
|
# Note, Pandas does some odd things with the left/right fill_value
|