Mark test_binstats_quantile as xfail in the GMT Legacy Tests workflow

[weiji14]

Description of proposed changes

The test_binstats_quantile test (introduced in #3012) is failing on the legacy CI tests. Trying to see if it's due to xarray/pandas/numpy or using old version of GMT 6.4 Edit: seems to be that GMT 6.4 gives different results to GMT 6.5+, most likely due to changes in GenericMappingTools/gmt#8243

Traceback from https://github.com/GenericMappingTools/pygmt/actions/runs/16895808645/job/47865247702#step:6:1098

___________________________ test_binstats_quantile ____________________________
    def test_binstats_quantile():
        """
        Test binstats quantile statistic functionality.
        """
        temp_grid = binstats(
            data="@capitals.gmt",
            spacing=5,
            statistic="quantile",
            quantile_value=75,
            search_radius="1000k",
            aspatial="2=population",
            region="g",
        )
        assert temp_grid.dims == ("y", "x")
        assert temp_grid.gmt.gtype is GridType.CARTESIAN
        assert temp_grid.gmt.registration is GridRegistration.GRIDLINE
        npt.assert_allclose(temp_grid.max(), 15047685)
>       npt.assert_allclose(temp_grid.min(), 53)
..\pygmt\tests\test_binstats.py:71: 
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _
args = (<function assert_allclose.<locals>.compare at 0x000001D761A02E80>, array(0., dtype=float32), array(53))
kwds = {'equal_nan': True, 'err_msg': '', 'header': 'Not equal to tolerance rtol=1e-07, atol=0', 'verbose': True}
    @wraps(func)
    def inner(*args, **kwds):
        with self._recreate_cm():
>           return func(*args, **kwds)
E           AssertionError: 
E           Not equal to tolerance rtol=1e-07, atol=0
E           
E           Mismatched elements: 1 / 1 (100%)
E           Max absolute difference: 53.
E           Max relative difference: 1.
E            x: array(0., dtype=float32)
E            y: array(53)
C:\hostedtoolcache\windows\Python\3.11.9\x64\Lib\contextlib.py:81: AssertionError

Output of pygmt.show_versions() from CI:

PyGMT information:
  version: v0.17.0.dev39+g354b50a0f
System information:
  python: 3.11.9 | packaged by conda-forge | (main, Apr 19 2024, 18:36:13) [GCC 12.3.0]
  executable: /home/runner/micromamba/envs/pygmt/bin/python
  machine: Linux-6.8.0-1031-azure-x86_64-with-glibc2.35
Dependency information:
  numpy: 1.26.4
  pandas: 2.3.1
  xarray: 2025.7.1
  packaging: 24.2
  contextily: 1.5.2
  geopandas: 1.0.1
  IPython: 9.4.0
  pyarrow: 16.1.0
  rioxarray: 0.19.0
  gdal: 3.8.5
  ghostscript: 9.56.1
GMT library information:
  version: 6.4.0
  padding: 2
  share dir: /home/runner/micromamba/envs/pygmt/share/gmt
  plugin dir: /home/runner/micromamba/envs/pygmt/lib/gmt/plugins
  library path: /home/runner/micromamba/envs/pygmt/lib/libgmt.so
  cores: 4
  grid layout: rows
  image layout: 
  binary version: 6.4.0

Patches #3012

Preview:

Reminders

• Run make format and make check to make sure the code follows the style guide.
• Add tests for new features or tests that would have caught the bug that you're fixing.
• Add new public functions/methods/classes to doc/api/index.rst.
• Write detailed docstrings for all functions/methods.
• If wrapping a new module, open a 'Wrap new GMT module' issue and submit reasonably-sized PRs.
• If adding new functionality, add an example to docstrings or tutorials.

Slash Commands

You can write slash commands (/command) in the first line of a comment to perform
specific operations. Supported slash command is:

• /format: automatically format and lint the code

[weiji14]

This PR probably won't be needed if #4017 goes ahead, depends on whether we can get #4018 to work before the next release.

[weiji14]

Unsure which values are the correct ones. Do we want to check in upstream GMT on what might have changed binstat from GMT 6.4 to 6.5?

[seisman]

Likely related to GenericMappingTools/gmt#8243.

[weiji14]

Thanks for the reminder, forgot about that thread! Have changed to an xfail in commit 1dc2e42