patch both xarray and erddapy for pandas 2.0 #426

ocefpaf · 2023-04-04T15:29:45Z

Checklist

Ran python show_diff.py and posted the output as part of the PR.
Modifications won't affect packages built in the future.

noarch::erddapy-0.6.0-py_0.tar.bz2
-    "pandas >=0.20.3",
+    "pandas >=0.20.3,<2",
noarch::erddapy-0.7.0-py_0.tar.bz2
-    "pandas >=0.20.3",
+    "pandas >=0.20.3,<2",
noarch::erddapy-0.7.1-py_0.tar.bz2
-    "pandas >=0.20.3",
+    "pandas >=0.20.3,<2",
noarch::erddapy-0.7.2-py_0.tar.bz2
-    "pandas >=0.20.3",
+    "pandas >=0.20.3,<2",
noarch::erddapy-0.8.0-pyhd8ed1ab_0.tar.bz2
-    "pandas >=0.20.3",
+    "pandas >=0.20.3,<2",
noarch::erddapy-0.9.0-pyhd8ed1ab_0.tar.bz2
-    "pandas >=0.20.3",
+    "pandas >=0.20.3,<2",
noarch::erddapy-1.0.0-pyhd8ed1ab_0.tar.bz2
-    "pandas >=0.20.3",
+    "pandas >=0.20.3,<2",
noarch::erddapy-1.1.0-pyhd8ed1ab_0.tar.bz2
-    "pandas >=0.20.3",
+    "pandas >=0.20.3,<2",
noarch::erddapy-1.1.1-pyhd8ed1ab_0.tar.bz2
-    "pandas >=0.20.3",
+    "pandas >=0.20.3,<2",
noarch::erddapy-1.2.0-pyhd8ed1ab_0.tar.bz2
-    "pandas >=0.20.3",
+    "pandas >=0.20.3,<2",
noarch::erddapy-1.2.1-pyhd8ed1ab_0.tar.bz2
-    "pandas >=0.20.3",
+    "pandas >=0.20.3,<2",
noarch::xarray-2022.10.0-pyhd8ed1ab_0.tar.bz2
-    "pandas >=1.3",
+    "pandas >=1.3,<2",
noarch::xarray-2022.11.0-pyhd8ed1ab_0.tar.bz2
-    "pandas >=1.3",
+    "pandas >=1.3,<2",
noarch::pmgr-2.1.2-pyhd8ed1ab_0.conda
-    "mysqlclient =1.3.12|>=2.0.3",
+    "mysqlclient ==1.3.12.*|>=2.0.3",
noarch::xarray-2022.12.0-pyhd8ed1ab_0.conda
-    "pandas >=1.3",
+    "pandas >=1.3,<2",
noarch::xarray-2023.1.0-pyhd8ed1ab_0.conda
-    "pandas >=1.3",
+    "pandas >=1.3,<2",

conda-forge-webservices · 2023-04-04T15:29:51Z

Hi! This is the friendly automated conda-forge-linting service.

I just wanted to let you know that I linted all conda-recipes in your PR (recipe) and found it was in an excellent condition.

ocefpaf · 2023-04-04T15:32:07Z

@conda-forge/core this is ready for review! I'm not sure what is causing the diff in

noarch::pmgr-2.1.2-pyhd8ed1ab_0.conda
-    "mysqlclient =1.3.12|>=2.0.3",
+    "mysqlclient ==1.3.12.*|>=2.0.3",

Looks like some library used in the diff got updated and is "fixing/updating" the comparison there.

recipe/gen_patch_json.py

ocefpaf · 2023-04-04T16:18:42Z

Latest diff:

noarch::erddapy-0.6.0-py_0.tar.bz2
-    "pandas >=0.20.3",
+    "pandas >=0.20.3,<2",
noarch::erddapy-0.7.0-py_0.tar.bz2
-    "pandas >=0.20.3",
+    "pandas >=0.20.3,<2",
noarch::erddapy-0.7.1-py_0.tar.bz2
-    "pandas >=0.20.3",
+    "pandas >=0.20.3,<2",
noarch::erddapy-0.7.2-py_0.tar.bz2
-    "pandas >=0.20.3",
+    "pandas >=0.20.3,<2",
noarch::erddapy-0.8.0-pyhd8ed1ab_0.tar.bz2
-    "pandas >=0.20.3",
+    "pandas >=0.20.3,<2",
noarch::erddapy-0.9.0-pyhd8ed1ab_0.tar.bz2
-    "pandas >=0.20.3",
+    "pandas >=0.20.3,<2",
noarch::erddapy-1.0.0-pyhd8ed1ab_0.tar.bz2
-    "pandas >=0.20.3",
+    "pandas >=0.20.3,<2",
noarch::erddapy-1.1.0-pyhd8ed1ab_0.tar.bz2
-    "pandas >=0.20.3",
+    "pandas >=0.20.3,<2",
noarch::erddapy-1.1.1-pyhd8ed1ab_0.tar.bz2
-    "pandas >=0.20.3",
+    "pandas >=0.20.3,<2",
noarch::erddapy-1.2.0-pyhd8ed1ab_0.tar.bz2
-    "pandas >=0.20.3",
+    "pandas >=0.20.3,<2",
noarch::erddapy-1.2.1-pyhd8ed1ab_0.tar.bz2
-    "pandas >=0.20.3",
+    "pandas >=0.20.3,<2",
noarch::xarray-2022.10.0-pyhd8ed1ab_0.tar.bz2
-    "pandas >=1.3",
+    "pandas >=1.3,<2",
noarch::xarray-2022.11.0-pyhd8ed1ab_0.tar.bz2
-    "pandas >=1.3",
+    "pandas >=1.3,<2",
noarch::pmgr-2.1.2-pyhd8ed1ab_0.conda
-    "mysqlclient =1.3.12|>=2.0.3",
+    "mysqlclient ==1.3.12.*|>=2.0.3",
noarch::xarray-2022.12.0-pyhd8ed1ab_0.conda
-    "pandas >=1.3",
+    "pandas >=1.3,<2",
noarch::xarray-2023.1.0-pyhd8ed1ab_0.conda
-    "pandas >=1.3",
+    "pandas >=1.3,<2",
noarch::xarray-2023.2.0-pyhd8ed1ab_0.conda
-    "pandas >=1.4",
+    "pandas >=1.4,<2",

dcherian · 2023-04-04T16:20:13Z

LGTM. How do we decide how far back to go?

ocefpaf · 2023-04-04T16:24:49Z

LGTM. How do we decide how far back to go?

This one will patch all that is less-equal 2023.1.0 and pandas is pinned with pandas >=1.3 .

        if record_name == "xarray" and packaging.version.Version(record["version"]) <= packaging.version.Version("2023.1.0"):
            _replace_pin("pandas >=1.3", "pandas >=1.3,<2", deps, record)

That is patching xarray-2022.10.0-pyhd8ed1ab_0, xarray-2022.11.0-pyhd8ed1ab_0, xarray-2022.12.0-pyhd8ed1ab_0, and 2023.1.0-pyhd8ed1ab_0. If we want to patch more we need to modify the pandas version we are matching. IMO, not worth it.

This one only patches xarray 2023.2.0.

        if record_name == "xarray" and packaging.version.Version(record["version"]) == packaging.version.Version("2023.2.0"):
            _replace_pin("pandas >=1.4", "pandas >=1.4,<2", deps, record)

beckermr · 2023-04-04T17:47:58Z

Usually, upper bounds need to exclude prereleases. There are examples in the patches code of what those look like, usually ending in a0.

ocefpaf · 2023-04-04T17:53:21Z

Usually, upper bounds need to exclude prereleases. There are examples in the patches code of what those look like, usually ending in a0.

@dcherian you don't publish xarray pre-releases, right? Is that something we need to worry about here?

beckermr · 2023-04-04T17:54:13Z

I mean the pandas upper bound at <2, not xarray.

dcherian · 2023-04-04T17:58:49Z

you don't publish xarray pre-releases, right?

Not for the versions discussed here. The last one was 2022.06.0 IIRC

ocefpaf · 2023-04-04T18:07:55Z

I mean the pandas upper bound at <2, not xarray.

I really hate the treatment pre/dev releases got in PyPI and what that forces us to do :-/

Latest diff:

noarch::erddapy-0.6.0-py_0.tar.bz2
-    "pandas >=0.20.3",
+    "pandas >=0.20.3,<2a0",
noarch::erddapy-0.7.0-py_0.tar.bz2
-    "pandas >=0.20.3",
+    "pandas >=0.20.3,<2a0",
noarch::erddapy-0.7.1-py_0.tar.bz2
-    "pandas >=0.20.3",
+    "pandas >=0.20.3,<2a0",
noarch::erddapy-0.7.2-py_0.tar.bz2
-    "pandas >=0.20.3",
+    "pandas >=0.20.3,<2a0",
noarch::erddapy-0.8.0-pyhd8ed1ab_0.tar.bz2
-    "pandas >=0.20.3",
+    "pandas >=0.20.3,<2a0",
noarch::erddapy-0.9.0-pyhd8ed1ab_0.tar.bz2
-    "pandas >=0.20.3",
+    "pandas >=0.20.3,<2a0",
noarch::erddapy-1.0.0-pyhd8ed1ab_0.tar.bz2
-    "pandas >=0.20.3",
+    "pandas >=0.20.3,<2a0",
noarch::erddapy-1.1.0-pyhd8ed1ab_0.tar.bz2
-    "pandas >=0.20.3",
+    "pandas >=0.20.3,<2a0",
noarch::erddapy-1.1.1-pyhd8ed1ab_0.tar.bz2
-    "pandas >=0.20.3",
+    "pandas >=0.20.3,<2a0",
noarch::erddapy-1.2.0-pyhd8ed1ab_0.tar.bz2
-    "pandas >=0.20.3",
+    "pandas >=0.20.3,<2a0",
noarch::erddapy-1.2.1-pyhd8ed1ab_0.tar.bz2
-    "pandas >=0.20.3",
+    "pandas >=0.20.3,<2a0",
noarch::xarray-2022.10.0-pyhd8ed1ab_0.tar.bz2
-    "pandas >=1.3",
+    "pandas >=1.3,<2a0",
noarch::xarray-2022.11.0-pyhd8ed1ab_0.tar.bz2
-    "pandas >=1.3",
+    "pandas >=1.3,<2a0",
noarch::pmgr-2.1.2-pyhd8ed1ab_0.conda
-    "mysqlclient =1.3.12|>=2.0.3",
+    "mysqlclient ==1.3.12.*|>=2.0.3",
noarch::xarray-2022.12.0-pyhd8ed1ab_0.conda
-    "pandas >=1.3",
+    "pandas >=1.3,<2a0",
noarch::xarray-2023.1.0-pyhd8ed1ab_0.conda
-    "pandas >=1.3",
+    "pandas >=1.3,<2a0",
noarch::xarray-2023.2.0-pyhd8ed1ab_0.conda
-    "pandas >=1.4",
+    "pandas >=1.4,<2a0",

jaimergp

LGTM. Maybe we can add timestamp checks too but for now let's fix the issue.

- a patch is on its way at ioos/erddapy#300 and conda-forge/conda-forge-repodata-patches-feedstock#426

dcherian · 2023-04-05T15:31:03Z

Thanks all!

patch both xarray and erddapy for pandas 2.0

16482d2

ocefpaf requested a review from a team as a code owner April 4, 2023 15:29

This was referenced Apr 4, 2023

bad conda solve with pandas 2 pydata/xarray#7716

Closed

Impossible to import erddapy with pandas 2.0 ioos/erddapy#299

Closed

use packaging instead of pkg_resources

b157022

dcherian reviewed Apr 4, 2023

View reviewed changes

recipe/gen_patch_json.py Outdated Show resolved Hide resolved

adjust for 2023.2.0

67ba097

use a0 pandas upper limit

56ec5a3

jaimergp approved these changes Apr 5, 2023

View reviewed changes

jaimergp merged commit 2f1df76 into conda-forge:main Apr 5, 2023

gmaze added a commit to euroargodev/argopy that referenced this pull request Apr 5, 2023

Drop support for Pandas < 2.0 in CI tests

37ecd50

- a patch is on its way at ioos/erddapy#300 and conda-forge/conda-forge-repodata-patches-feedstock#426

gmaze added a commit to euroargodev/argopy that referenced this pull request Apr 5, 2023

Drop support for Pandas 2.0 in CI tests [skip-ci]

3738d7d

- a patch is on its way at ioos/erddapy#300 and conda-forge/conda-forge-repodata-patches-feedstock#426

ocefpaf deleted the patch_xarray_erddapy_4_pandas_2 branch April 5, 2023 09:17

jrbourbeau mentioned this pull request Apr 5, 2023

Add xarray back to Python 3.11 once there is pandas 2.0 support dask/dask#10152

Closed

weiji14 mentioned this pull request Apr 6, 2023

Use np.asarray to convert a 1-D array to datetime type in array_to_datetime GenericMappingTools/pygmt#2481

Merged

7 tasks

seisman mentioned this pull request Sep 27, 2024

SPEC 0: Set minimum supported versions to pandas>=2.0 and xarray>=2023.04 GenericMappingTools/pygmt#3460

Merged

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

patch both xarray and erddapy for pandas 2.0 #426

patch both xarray and erddapy for pandas 2.0 #426

ocefpaf commented Apr 4, 2023

conda-forge-webservices bot commented Apr 4, 2023

ocefpaf commented Apr 4, 2023

ocefpaf commented Apr 4, 2023

dcherian commented Apr 4, 2023

ocefpaf commented Apr 4, 2023

beckermr commented Apr 4, 2023

ocefpaf commented Apr 4, 2023

beckermr commented Apr 4, 2023

dcherian commented Apr 4, 2023

ocefpaf commented Apr 4, 2023 •

edited

Loading

jaimergp left a comment

dcherian commented Apr 5, 2023

patch both xarray and erddapy for pandas 2.0 #426

patch both xarray and erddapy for pandas 2.0 #426

Conversation

ocefpaf commented Apr 4, 2023

conda-forge-webservices bot commented Apr 4, 2023

ocefpaf commented Apr 4, 2023

ocefpaf commented Apr 4, 2023

dcherian commented Apr 4, 2023

ocefpaf commented Apr 4, 2023

beckermr commented Apr 4, 2023

ocefpaf commented Apr 4, 2023

beckermr commented Apr 4, 2023

dcherian commented Apr 4, 2023

ocefpaf commented Apr 4, 2023 • edited Loading

jaimergp left a comment

Choose a reason for hiding this comment

dcherian commented Apr 5, 2023

ocefpaf commented Apr 4, 2023 •

edited

Loading