Move Matplotlib to hard requirement #4156

tkoyama010 · 2023-03-18T17:14:42Z

Overview

resolves #3970

Details

None

akaszynski

Agreed. It's also a vtk requirement, so people already have to install this unless they ignore dependencies.

codecov · 2023-03-18T17:26:16Z

Codecov Report

Merging #4156 (7367d6b) into main (bf44573) will increase coverage by 0.00%.
The diff coverage is 94.44%.

@@           Coverage Diff           @@
##             main    #4156   +/-   ##
=======================================
  Coverage   95.58%   95.59%           
=======================================
  Files          95       95           
  Lines       20356    20333   -23     
=======================================
- Hits        19458    19438   -20     
+ Misses        898      895    -3

adeak

If we make matplotlib a hard dependency then we should at the same time tear out a bunch of logic related to the soft dependency status:

$ git grep matplotlib 'pyvista/*.py'
pyvista/core/filters/data_set.py:        Plot the variables of interest in 2D using matplotlib where the
pyvista/core/filters/data_set.py:            The string title of the matplotlib figure.
pyvista/core/filters/data_set.py:            Shows the matplotlib figure.
pyvista/core/filters/data_set.py:        # Ensure matplotlib is available
pyvista/core/filters/data_set.py:            import matplotlib.pyplot as plt
pyvista/core/filters/data_set.py:            raise ImportError('matplotlib must be available to use this filter.')
pyvista/core/filters/data_set.py:            The string title of the ``matplotlib`` figure.
pyvista/core/filters/data_set.py:            Shows the ``matplotlib`` figure when ``True``.
pyvista/core/filters/data_set.py:        # Ensure matplotlib is available
pyvista/core/filters/data_set.py:            import matplotlib.pyplot as plt
pyvista/core/filters/data_set.py:            raise ImportError('matplotlib must be installed to use this filter.')
pyvista/core/filters/data_set.py:        # create the matplotlib figure
pyvista/core/filters/data_set.py:            The string title of the `matplotlib` figure.
pyvista/core/filters/data_set.py:            Shows the matplotlib figure.
pyvista/core/filters/data_set.py:        # Ensure matplotlib is available
pyvista/core/filters/data_set.py:            import matplotlib.pyplot as plt
pyvista/core/filters/data_set.py:            raise ImportError('matplotlib must be installed to use this filter.')
pyvista/core/filters/data_set.py:        # create the matplotlib figure
pyvista/demos/demos.py:        import matplotlib  # noqa
pyvista/jupyter/pv_ipygany.py:        # attempt to matplotlib cmaps to ipygany
pyvista/plotting/charts.py:    """2D chart class similar to a ``matplotlib`` figure.
pyvista/plotting/charts.py:        the `matplotlib.pyplot.plot
pyvista/plotting/charts.py:        <https://matplotlib.org/stable/api/_as_gen/matplotlib.pyplot.plot.html>`_.
pyvista/plotting/charts.py:    """Create new chart from an existing matplotlib figure.
pyvista/plotting/charts.py:    figure : matplotlib.figure.Figure, optional
pyvista/plotting/charts.py:        The matplotlib figure to draw. If no figure is
pyvista/plotting/charts.py:    Plot streamlines of a vector field with varying colors (based on `this example <https://matplotlib.org/stable/gallery/images_contours_and_fields/plot_streamplot.html>`_).
pyvista/plotting/charts.py:    >>> import matplotlib.pyplot as plt
pyvista/plotting/charts.py:        "chart_name": "matplotlib chart",
pyvista/plotting/charts.py:            from matplotlib.backends.backend_agg import FigureCanvasAgg
pyvista/plotting/charts.py:            import matplotlib.figure  # noqa
pyvista/plotting/charts.py:            import matplotlib.pyplot as plt
pyvista/plotting/charts.py:            raise ImportError("ChartMPL requires matplotlib")
pyvista/plotting/charts.py:        # Close the underlying matplotlib figure when creating the sphinx gallery.
pyvista/plotting/charts.py:        # matplotlib figure (fetched by the 'matplotlib' scraper).
pyvista/plotting/charts.py:        """Retrieve the matplotlib figure associated with this chart.
pyvista/plotting/charts.py:        Create a matplotlib chart from an existing figure.
pyvista/plotting/charts.py:        >>> import matplotlib.pyplot as plt
pyvista/plotting/charts.py:        Plotter window is resized or the matplotlib figure is
pyvista/plotting/charts.py:        Create a matplotlib chart with title 'My Chart'.
pyvista/plotting/charts.py:        >>> import matplotlib.pyplot as plt
pyvista/plotting/charts.py:        Create a matplotlib chart with custom labels and show the legend.
pyvista/plotting/charts.py:        >>> import matplotlib.pyplot as plt
pyvista/plotting/colors.py:Used code from matplotlib.colors.  Thanks for your work.
pyvista/plotting/colors.py:# Following colors are copied from matplotlib.colors, synonyms (colors with a
pyvista/plotting/colors.py:matplotlib_default_colors = [
pyvista/plotting/colors.py:    """Fetch a colormap by name from matplotlib, colorcet, or cmocean."""
pyvista/plotting/colors.py:        if not has_module('matplotlib'):
pyvista/plotting/colors.py:                'The use of custom colormaps requires the installation of matplotlib.'
pyvista/plotting/colors.py:            from matplotlib import colormaps, colors
pyvista/plotting/colors.py:            import matplotlib
pyvista/core/filters/data_set.py:        Plot the variables of interest in 2D using matplotlib where the
pyvista/core/filters/data_set.py:            The string title of the matplotlib figure.
pyvista/core/filters/data_set.py:            Shows the matplotlib figure.
pyvista/core/filters/data_set.py:        # Ensure matplotlib is available
pyvista/core/filters/data_set.py:            import matplotlib.pyplot as plt
pyvista/core/filters/data_set.py:            raise ImportError('matplotlib must be available to use this filter.')
pyvista/core/filters/data_set.py:            The string title of the ``matplotlib`` figure.
pyvista/core/filters/data_set.py:            Shows the ``matplotlib`` figure when ``True``.
pyvista/core/filters/data_set.py:        # Ensure matplotlib is available
pyvista/core/filters/data_set.py:            import matplotlib.pyplot as plt
pyvista/core/filters/data_set.py:            raise ImportError('matplotlib must be installed to use this filter.')
pyvista/core/filters/data_set.py:        # create the matplotlib figure
pyvista/core/filters/data_set.py:            The string title of the `matplotlib` figure.
pyvista/core/filters/data_set.py:            Shows the matplotlib figure.
pyvista/core/filters/data_set.py:        # Ensure matplotlib is available
pyvista/core/filters/data_set.py:            import matplotlib.pyplot as plt
pyvista/core/filters/data_set.py:            raise ImportError('matplotlib must be installed to use this filter.')
pyvista/core/filters/data_set.py:        # create the matplotlib figure
pyvista/demos/demos.py:        import matplotlib  # noqa
pyvista/jupyter/pv_ipygany.py:        # attempt to matplotlib cmaps to ipygany
pyvista/plotting/charts.py:    """2D chart class similar to a ``matplotlib`` figure.
pyvista/plotting/charts.py:        the `matplotlib.pyplot.plot
pyvista/plotting/charts.py:        <https://matplotlib.org/stable/api/_as_gen/matplotlib.pyplot.plot.html>`_.
pyvista/plotting/charts.py:    """Create new chart from an existing matplotlib figure.
pyvista/plotting/charts.py:    figure : matplotlib.figure.Figure, optional
pyvista/plotting/charts.py:        The matplotlib figure to draw. If no figure is
pyvista/plotting/charts.py:    Plot streamlines of a vector field with varying colors (based on `this example <https://matplotlib.org/stable/gallery/images_contours_and_fields/plot_streamplot.html>`_).
pyvista/plotting/charts.py:    >>> import matplotlib.pyplot as plt
pyvista/plotting/charts.py:        "chart_name": "matplotlib chart",
pyvista/plotting/charts.py:            from matplotlib.backends.backend_agg import FigureCanvasAgg
pyvista/plotting/charts.py:            import matplotlib.figure  # noqa
pyvista/plotting/charts.py:            import matplotlib.pyplot as plt
pyvista/plotting/charts.py:            raise ImportError("ChartMPL requires matplotlib")
pyvista/plotting/charts.py:        # Close the underlying matplotlib figure when creating the sphinx gallery.
pyvista/plotting/charts.py:        # matplotlib figure (fetched by the 'matplotlib' scraper).
pyvista/plotting/charts.py:        """Retrieve the matplotlib figure associated with this chart.
pyvista/plotting/charts.py:        Create a matplotlib chart from an existing figure.
pyvista/plotting/charts.py:        >>> import matplotlib.pyplot as plt
pyvista/plotting/charts.py:        Plotter window is resized or the matplotlib figure is
pyvista/plotting/charts.py:        Create a matplotlib chart with title 'My Chart'.
pyvista/plotting/charts.py:        >>> import matplotlib.pyplot as plt
pyvista/plotting/charts.py:        Create a matplotlib chart with custom labels and show the legend.
pyvista/plotting/charts.py:        >>> import matplotlib.pyplot as plt
pyvista/plotting/colors.py:Used code from matplotlib.colors.  Thanks for your work.
pyvista/plotting/colors.py:# Following colors are copied from matplotlib.colors, synonyms (colors with a
pyvista/plotting/colors.py:matplotlib_default_colors = [
pyvista/plotting/colors.py:    """Fetch a colormap by name from matplotlib, colorcet, or cmocean."""
pyvista/plotting/colors.py:        if not has_module('matplotlib'):
pyvista/plotting/colors.py:                'The use of custom colormaps requires the installation of matplotlib.'
pyvista/plotting/colors.py:            from matplotlib import colormaps, colors
pyvista/plotting/colors.py:            import matplotlib
pyvista/plotting/colors.py:            if not scooby.meets_version(matplotlib.__version__, min_req):
pyvista/plotting/colors.py:                    f'matplotlib>={min_req}.'
pyvista/plotting/colors.py:        if not has_module('matplotlib'):
pyvista/plotting/colors.py:                'The use of custom colormaps requires the installation of matplotlib.'
pyvista/plotting/colors.py:        from matplotlib.colors import ListedColormap
pyvista/plotting/colors.py:    """Return the default color cycler (matches matplotlib's default)."""
pyvista/plotting/colors.py:        raise ImportError('cycler not installed. Please install matplotlib.')
pyvista/plotting/colors.py:    return cycler('color', matplotlib_default_colors)
pyvista/plotting/colors.py:        raise ImportError('cycler not installed. Please install matplotlib.')
pyvista/plotting/colors.py:def get_matplotlib_theme_cycler():
pyvista/plotting/colors.py:    """Return matplotlib's current theme's color cycler."""
pyvista/plotting/colors.py:        import matplotlib.pyplot as plt
pyvista/plotting/colors.py:        raise ImportError('Please install matplotlib.')
pyvista/plotting/colors.py:        raise ImportError('cycler not installed. Please install matplotlib.')
pyvista/plotting/colors.py:    * ``'default'`` - Use the default color cycler (matches matplotlib's default)
pyvista/plotting/colors.py:    * ``'matplotlib`` - Dynamically get matplotlib's current theme's color cycler.
pyvista/plotting/colors.py:        raise ImportError('cycler not installed. Please install matplotlib.')
pyvista/plotting/colors.py:        elif color_cycler == 'matplotlib':
pyvista/plotting/colors.py:            return get_matplotlib_theme_cycler()
pyvista/plotting/composite_mapper.py:        This uses ``matplotlib``'s color cycler.
pyvista/plotting/composite_mapper.py:        if has_module('matplotlib'):
pyvista/plotting/composite_mapper.py:            import matplotlib
pyvista/plotting/composite_mapper.py:            colors = cycle(matplotlib.rcParams['axes.prop_cycle'])
pyvista/plotting/composite_mapper.py:            warnings.warn('Please install matplotlib for color cycles.')
pyvista/plotting/composite_mapper.py:            if cmap is None:  # Set default map if matplotlib is available
pyvista/plotting/composite_mapper.py:                if has_module('matplotlib'):
pyvista/plotting/lookup_table.py:        Color map from ``matplotlib``, ``colorcet``, or ``cmocean``. Either
pyvista/plotting/lookup_table.py:        Apply ``matplotlib``'s ``'cividis'`` color map.
pyvista/plotting/lookup_table.py:        if not has_module('matplotlib'):  # pragma: no cover
pyvista/plotting/mapper.py:            # Set default map if matplotlib is available
pyvista/plotting/mapper.py:                if has_module('matplotlib'):
pyvista/plotting/mapper.py:                    # no requires matplotlib
pyvista/plotting/plotting.py:            If a string, this is the name of the ``matplotlib`` colormap to use
pyvista/plotting/plotting.py:            Color each block by a solid color using matplotlib's color cycler.
pyvista/plotting/plotting.py:            If a string, this is the name of the ``matplotlib`` colormap to use
pyvista/plotting/plotting.py:            each block by a solid color using matplotlib's color cycler.
pyvista/plotting/plotting.py:            If a string, this is the name of the ``matplotlib`` colormap to use
pyvista/plotting/plotting.py:                if not has_module('matplotlib'):
pyvista/plotting/plotting.py:                    raise ImportError('Please install matplotlib for color maps.')
pyvista/plotting/renderer.py:            * ``'default'`` - Use the default color cycler (matches matplotlib's default)
pyvista/plotting/renderer.py:            * ``'matplotlib`` - Dynamically get matplotlib's current theme's color cycler.
pyvista/plotting/renderer.py:        ``matplotlib``'s ``grid`` function.
pyvista/plotting/renderer.py:            list, or a matplotlib color string (e.g. ``'w'`` or ``'white'``
pyvista/plotting/renderers.py:            * ``'default'`` - Use the default color cycler (matches matplotlib's default)
pyvista/plotting/renderers.py:            * ``'matplotlib`` - Dynamically get matplotlib's current theme's color cycler.
pyvista/themes.py:                'Unable to set a default theme colormap without matplotlib. '
pyvista/themes.py:            * ``'default'`` - Use the default color cycler (matches matplotlib's default)
pyvista/themes.py:            * ``'matplotlib`` - Dynamically get matplotlib's current theme's color cycler.
pyvista/utilities/errors.py:            matplotlib : 3.6.0
pyvista/utilities/errors.py:            'matplotlib',
pyvista/utilities/misc.py:    """Return if a figure can be created with matplotlib."""
pyvista/utilities/misc.py:        import matplotlib.pyplot as plt
pyvista/utilities/sphinx_gallery.py:    'matplotlib' and 'mayavi'. Details on this implementation can be found in

A bunch of the above lines signal snippets where we go out of our way to check if matplotlib is installed. We should remove those rather then let them sit around as debt.

Here's a filter for the imports alone, but this misses some cases, e.g. cycler availability checking:

$ git grep 'from matplotlib\|import matplotlib' 'pyvista/*.py'
pyvista/core/filters/data_set.py:            import matplotlib.pyplot as plt
pyvista/core/filters/data_set.py:            import matplotlib.pyplot as plt
pyvista/core/filters/data_set.py:            import matplotlib.pyplot as plt
pyvista/demos/demos.py:        import matplotlib  # noqa
pyvista/plotting/charts.py:    >>> import matplotlib.pyplot as plt
pyvista/plotting/charts.py:            from matplotlib.backends.backend_agg import FigureCanvasAgg
pyvista/plotting/charts.py:            import matplotlib.figure  # noqa
pyvista/plotting/charts.py:            import matplotlib.pyplot as plt
pyvista/plotting/charts.py:        >>> import matplotlib.pyplot as plt
pyvista/plotting/charts.py:        >>> import matplotlib.pyplot as plt
pyvista/plotting/charts.py:        >>> import matplotlib.pyplot as plt
pyvista/plotting/colors.py:Used code from matplotlib.colors.  Thanks for your work.
pyvista/plotting/colors.py:# Following colors are copied from matplotlib.colors, synonyms (colors with a
pyvista/plotting/colors.py:    """Fetch a colormap by name from matplotlib, colorcet, or cmocean."""
pyvista/plotting/colors.py:            from matplotlib import colormaps, colors
pyvista/plotting/colors.py:            import matplotlib
pyvista/plotting/colors.py:        from matplotlib.colors import ListedColormap
pyvista/plotting/colors.py:        import matplotlib.pyplot as plt
pyvista/plotting/composite_mapper.py:            import matplotlib
pyvista/utilities/misc.py:        import matplotlib.pyplot as plt

pyvista/utilities/misc.py

adeak

Great job tearing out the legacy @tkoyama010 :) I have a few comments, and I think there's a few more things we could/should get rid of:

$ git grep -i matplotlib |grep -in 'require\|installed'
28:docker/requirements.txt:matplotlib
54:examples/02-plot/color_cycler.py:#    This requires matplotlib (or at least cycler) to be installed.
71:examples_trame/requirements.txt:matplotlib
133:pyvista/plotting/composite_mapper.py:            for when displaying ``scalars``. Requires Matplotlib to be
142:pyvista/plotting/mapper.py:            for when displaying ``scalars``. Requires Matplotlib to be
145:pyvista/plotting/plotting.py:            Matplotlib to be installed.  ``colormap`` is also an accepted alias
149:pyvista/plotting/plotting.py:            Matplotlib to be installed.  ``colormap`` is also an accepted alias
153:pyvista/plotting/plotting.py:            Matplotlib to be installed.  ``colormap`` is also an accepted alias
161:pyvista/themes.py:        displaying ``scalars``. Requires Matplotlib to be installed.
167:requirements.txt:matplotlib>=3.5.0,<3.7.2
179:tests/plotting/test_plotting.py:    reason='VTK and Matplotlib version incompatibility. For VTK<=9.2.2, MathText requires matplotlib<3.6',
180:tests/plotting/test_plotting.py:    For VTK<=9.2.2, this requires matplotlib<3.6
193:tests/test_filters.py:    """This test requires matplotlib."""
194:tests/test_filters.py:    """This test requires matplotlib."""
195:tests/test_filters.py:    """This test requires matplotlib."""

Most of these signals stale comments or remarks in docstrings, and stale requirements because matplotlib is now a direct dependency of pyvista.

pyvista/plotting/charts.py

pyvista/plotting/colors.py

adeak · 2023-03-19T12:49:54Z

pyvista/plotting/mapper.py

@@ -713,7 +712,6 @@ def set_scalars(
                    self.lookup_table.apply_cmap(cmap, n_colors)
                else:  # pragma: no cover
                    # should be impossible to get to this point as VTK package
-                    # no requires matplotlib


I think this was a typo, and the second part of this sentence was meant to be

# now requires matplotlib

In which case we should restore this line and fix the typo. Or reword the two-line comment to mention our own matplotlib dependency. Something like

# should be impossible to get to this point as matplotlib is a dependency

(give or take line lengths and wording).

Or perhaps the whole comment is meaningless now, I haven't checked the big picture. In this case I'd raise in the else, maybe an AssertionError to signal that this should never happen.

Turns out you can trigger this section of code even with matplotlib installed with:

import pyvista pyvista.global_theme.cmap = None mesh = pyvista.Sphere() mesh.plot(scalars=range(mesh.n_points))

This is because by setting pyvista.global_theme.cmap = None, now it's possible for cmap to be None within the mapper.

The easiest fix is to make sure that our theme must have a valid default color map. Since we now require matplotlib, we can do this and completely remove this section of code.

requirements.txt

akaszynski · 2023-03-20T14:09:49Z

@tkoyama010, I'm going to add support for matplotlib<3.5.0. If we require it, I'd like to have a lower bound for it.

banesullivan · 2023-03-20T14:38:31Z

I'd prefer not to bound the version of matplotlib if possible. I think there are really only two things where the version matters:

The recent changes to the colormap API -- very important and heavily used by us
The recent fixes in VTK to support MathText/LaTeX with Matplotlib 3.6

Instead, can we support the color mapping API across versions and have a helper utility to warn about incompatible VTK and MPL versions with MathText

akaszynski · 2023-03-20T14:43:23Z

Fully agree.

I'd prefer not to bound the version of matplotlib if possible. I think there are really only two things where the version matters:

The recent changes to the colormap API -- very important and heavily used by us

The recent fixes in VTK to support MathText/LaTeX with Matplotlib 3.6

Instead, can we support the color mapping API across versions and have a helper utility to warn about incompatible VTK and MPL versions with MathText

Fully agree. I'll deal with (1), can you deal with (2)?

banesullivan · 2023-03-20T14:48:00Z

Will do -- let me know when you're wrapped up, and I will improve the MathText support utils with:

An helper utility to check compatible versions -- will raise a dedicated error
An example for the gallery using MathText/LaTeX symbols and documentation on the supported version

pyvista/plotting/colors.py

tests/test_lookup_table.py

Co-authored-by: Andras Deak <adeak@users.noreply.github.com>

adeak · 2023-03-21T00:05:51Z

Phase 1 with older matplotlib compatibility looks good to me, thanks @akaszynski. There's one test failure with matplotlib 3.4:

FAILED tests/test_theme.py::test_cmap - TypeError: 'module' object is not subscriptable

This is because we get the wrong kind of error here:

pyvista/tests/test_theme.py

Lines 311 to 312 in bffe561

    
           with pytest.raises(KeyError, match='not a color map'): 
        
               default_theme.cmap = 'not a color map'

So this is a testing issue rather than a library issue, and we don't want to make the test suite work for older matplotlib (I assume).

On second thought, this might actually be an issue. Setting a non-existent cmap on a theme should probably be more informative:

>>> import pyvista as pv
>>> pv.global_theme.cmap = 'potato'
Traceback (most recent call last):
  File "<stdin>", line 1, in <module>
  File "/home/adeak/pyvista/pyvista/themes.py", line 2039, in cmap
    get_cmap_safe(cmap)  # for validation
  File "/home/adeak/pyvista/pyvista/plotting/colors.py", line 1221, in get_cmap_safe
    cmap = colormaps[cmap]
TypeError: 'module' object is not subscriptable

pyvista/plotting/colors.py

pyvista/plotting/composite_mapper.py

tkoyama010 · 2023-03-21T03:26:40Z

Sorry, I didn't realize the discussion was still going on and deleted >= 3.5.0. I will revert as the situation requires.

Co-authored-by: Andras Deak <adeak@users.noreply.github.com>

akaszynski · 2023-03-21T23:31:53Z

Resolved the KeyError issue in 69075a6 and now a ValueError is always raised.

I also added a lower limit in pyproject.toml of matplotlib>=3.0.1. This is necessary because matplotlib==3.0.0 has the following issue:

Successfully installed matplotlib-3.0.0
(py38) (py311) pyvista $ python -m pytest -nauto                                             Traceback (most recent call last):
  File "/usr/lib/python3.8/runpy.py", line 194, in _run_module_as_main
    return _run_code(code, main_globals, None,
  File "/usr/lib/python3.8/runpy.py", line 87, in _run_code
    exec(code, run_globals)
  File "/home/alex/python/py38/lib/python3.8/site-packages/pytest/__main__.py", line 5, in <module>
    raise SystemExit(pytest.console_main())
  File "/home/alex/python/py38/lib/python3.8/site-packages/_pytest/config/__init__.py", line 187, in console_main
    code = main()
  File "/home/alex/python/py38/lib/python3.8/site-packages/_pytest/config/__init__.py", line 145, in main
    config = _prepareconfig(args, plugins)
  File "/home/alex/python/py38/lib/python3.8/site-packages/_pytest/config/__init__.py", line 324, in _prepareconfig
    config = pluginmanager.hook.pytest_cmdline_parse(
  File "/home/alex/python/py38/lib/python3.8/site-packages/pluggy/_hooks.py", line 265, in __call__
    return self._hookexec(self.name, self.get_hookimpls(), kwargs, firstresult)
  File "/home/alex/python/py38/lib/python3.8/site-packages/pluggy/_manager.py", line 80, in _hookexec
    return self._inner_hookexec(hook_name, methods, kwargs, firstresult)
  File "/home/alex/python/py38/lib/python3.8/site-packages/pluggy/_callers.py", line 55, in _multicall
    gen.send(outcome)
  File "/home/alex/python/py38/lib/python3.8/site-packages/_pytest/helpconfig.py", line 102, in pytest_cmdline_parse
    config: Config = outcome.get_result()
  File "/home/alex/python/py38/lib/python3.8/site-packages/pluggy/_result.py", line 60, in get_result
    raise ex[1].with_traceback(ex[2])
  File "/home/alex/python/py38/lib/python3.8/site-packages/pluggy/_callers.py", line 39, in _multicall
    res = hook_impl.function(*args)
  File "/home/alex/python/py38/lib/python3.8/site-packages/_pytest/config/__init__.py", line 1016, in pytest_cmdline_parse
    self.parse(args)
  File "/home/alex/python/py38/lib/python3.8/site-packages/_pytest/config/__init__.py", line 1304, in parse
    self._preparse(args, addopts=addopts)
  File "/home/alex/python/py38/lib/python3.8/site-packages/_pytest/config/__init__.py", line 1187, in _preparse
    self.pluginmanager.load_setuptools_entrypoints("pytest11")
  File "/home/alex/python/py38/lib/python3.8/site-packages/pluggy/_manager.py", line 287, in load_setuptools_entrypoints
    plugin = ep.load()
  File "/usr/lib/python3.8/importlib/metadata.py", line 77, in load
    module = import_module(match.group('module'))
  File "/usr/lib/python3.8/importlib/__init__.py", line 127, in import_module
    return _bootstrap._gcd_import(name[level:], package, level)
  File "<frozen importlib._bootstrap>", line 1014, in _gcd_import
  File "<frozen importlib._bootstrap>", line 991, in _find_and_load
  File "<frozen importlib._bootstrap>", line 961, in _find_and_load_unlocked
  File "<frozen importlib._bootstrap>", line 219, in _call_with_frames_removed
  File "<frozen importlib._bootstrap>", line 1014, in _gcd_import
  File "<frozen importlib._bootstrap>", line 991, in _find_and_load
  File "<frozen importlib._bootstrap>", line 975, in _find_and_load_unlocked
  File "<frozen importlib._bootstrap>", line 671, in _load_unlocked
  File "/home/alex/python/py38/lib/python3.8/site-packages/_pytest/assertion/rewrite.py", line 168, in exec_module
    exec(co, module.__dict__)
  File "/home/alex/python/py38/lib/python3.8/site-packages/pytest_pyvista/__init__.py", line 5, in <module>
    from .pytest_pyvista import VerifyImageCache
  File "/home/alex/python/py38/lib/python3.8/site-packages/_pytest/assertion/rewrite.py", line 168, in exec_module
    exec(co, module.__dict__)
  File "/home/alex/python/py38/lib/python3.8/site-packages/pytest_pyvista/pytest_pyvista.py", line 8, in <module>
    import pyvista
  File "/home/alex/source/pyvista/pyvista/__init__.py", line 10, in <module>
    from pyvista.plotting import *
  File "/home/alex/source/pyvista/pyvista/plotting/__init__.py", line 4, in <module>
    from .charts import Chart, Chart2D, ChartMPL, ChartBox, ChartPie
  File "/home/alex/source/pyvista/pyvista/plotting/charts.py", line 10, in <module>
    from matplotlib.backends.backend_agg import FigureCanvasAgg
  File "/home/alex/python/py38/lib/python3.8/site-packages/matplotlib/__init__.py", line 1111, in <module>
    rcParamsOrig = RcParams(rcParams.copy())
  File "/home/alex/python/py38/lib/python3.8/site-packages/matplotlib/__init__.py", line 891, in __getitem__
    from matplotlib import pyplot as plt
  File "/home/alex/python/py38/lib/python3.8/site-packages/matplotlib/pyplot.py", line 32, in <module>
    import matplotlib.colorbar
  File "/home/alex/python/py38/lib/python3.8/site-packages/matplotlib/colorbar.py", line 40, in <module>
    import matplotlib._constrained_layout as constrained_layout
  File "/home/alex/python/py38/lib/python3.8/site-packages/matplotlib/_constrained_layout.py", line 52, in <module>
    from matplotlib.legend import Legend
  File "/home/alex/python/py38/lib/python3.8/site-packages/matplotlib/legend.py", line 43, in <module>
    from matplotlib.offsetbox import HPacker, VPacker, TextArea, DrawingArea
  File "/home/alex/python/py38/lib/python3.8/site-packages/matplotlib/offsetbox.py", line 33, in <module>
    from matplotlib.image import BboxImage
  File "/home/alex/python/py38/lib/python3.8/site-packages/matplotlib/image.py", line 19, in <module>
    from matplotlib.backend_bases import FigureCanvasBase
  File "/home/alex/python/py38/lib/python3.8/site-packages/matplotlib/backend_bases.py", line 46, in <module>
    from matplotlib import (
ImportError: cannot import name 'get_backend' from partially initialized module 'matplotlib' (most likely due to a circular import) (/home/alex/python/py38/lib/python3.8/site-packages/matplotlib/__init__.py)

Considering that matplotlib==3.0.1, I think it's good to say that > 4 years is quite good in regards to backwards compatibility.

akaszynski

Great work all. Nice to see net decreases of code without removing functionality.

There was one unresolved issue that I've opened issue #4174 for, but otherwise this is good to go.

tkoyama010 · 2023-03-22T05:43:01Z

Thank you all for reviews, comments, and support. I could not have done it without your help.

adeak

LGTM, thanks a lot @tkoyama010!

tkoyama010 added 2 commits March 19, 2023 02:08

Move Matplotlib to a hard requirement

1d65c80

Move Matplotlib to a hard requirement

1644b7b

github-actions bot added dependencies Pull requests that update a dependency file documentation Anything related to the documentation/website maintenance Low-impact maintenance activity labels Mar 18, 2023

akaszynski previously approved these changes Mar 18, 2023

View reviewed changes

Remove unused code

cf9517a

tkoyama010 dismissed akaszynski’s stale review via cf9517a March 18, 2023 17:27

tkoyama010 added 2 commits March 19, 2023 02:29

Fix COLORMAPS

e940d40

Fix test errors

4f3579d

adeak reviewed Mar 19, 2023

View reviewed changes

pyvista/utilities/misc.py Show resolved Hide resolved

tkoyama010 added 4 commits March 19, 2023 17:45

Merge branch 'main' into maint/move-matplotlib-to-hard-requirement

5e074e2

Fix matplotlib import

754787c

Fix document warning

53597a5

Minor revert

2791ee9

tkoyama010 requested review from adeak and akaszynski March 19, 2023 11:22

adeak reviewed Mar 19, 2023

View reviewed changes

tkoyama010 added 3 commits March 20, 2023 11:22

Merge branch 'main' into maint/move-matplotlib-to-hard-requirement

be5c1a8

Update charts.py

f9fc9df

Apply suggestions from code review

c09296f

allow matplotlib<3.5.0

9c0b367

akaszynski reviewed Mar 20, 2023

View reviewed changes

pyvista/plotting/colors.py Outdated Show resolved Hide resolved

Apply suggestions from code review

31c7f54

adeak reviewed Mar 20, 2023

View reviewed changes

pyvista/plotting/colors.py Outdated Show resolved Hide resolved

tests/test_lookup_table.py Show resolved Hide resolved

Update pyvista/plotting/colors.py

8abb6e6

Co-authored-by: Andras Deak <adeak@users.noreply.github.com>

adeak reviewed Mar 21, 2023

View reviewed changes

pyvista/plotting/colors.py Outdated Show resolved Hide resolved

Apply suggestions from code review

1a64e0f

tkoyama010 commented Mar 21, 2023

View reviewed changes

pyvista/plotting/composite_mapper.py Outdated Show resolved Hide resolved

Apply suggestions from code review

093da4a

tkoyama010 and others added 4 commits March 21, 2023 12:31

Merge branch 'main' into maint/move-matplotlib-to-hard-requirement

c813f4f

Additional fix

fec3984

Co-authored-by: Andras Deak <adeak@users.noreply.github.com>

Merge branch 'main' into maint/move-matplotlib-to-hard-requirement

b115fb2

raise consistent error

69075a6

akaszynski added 2 commits March 21, 2023 18:16

remove 'Youre not supposed to be here" code

af1d6eb

Merge branch 'main' into maint/move-matplotlib-to-hard-requirement

7367d6b

akaszynski mentioned this pull request Mar 22, 2023

Implement Matplotlib version check for latex and mathtex #4174

Closed

akaszynski approved these changes Mar 22, 2023

View reviewed changes

adeak approved these changes Mar 22, 2023

View reviewed changes

tkoyama010 merged commit dfe47b4 into main Mar 23, 2023

tkoyama010 deleted the maint/move-matplotlib-to-hard-requirement branch March 23, 2023 02:36

akaszynski mentioned this pull request Apr 30, 2023

Release 0.39.0 #4359

Merged

6 tasks

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Move Matplotlib to hard requirement #4156

Move Matplotlib to hard requirement #4156

tkoyama010 commented Mar 18, 2023

akaszynski left a comment

codecov bot commented Mar 18, 2023 •

edited

adeak left a comment •

edited

adeak left a comment

adeak Mar 19, 2023

akaszynski Mar 22, 2023

akaszynski commented Mar 20, 2023

banesullivan commented Mar 20, 2023

akaszynski commented Mar 20, 2023

banesullivan commented Mar 20, 2023

adeak commented Mar 21, 2023 •

edited

tkoyama010 commented Mar 21, 2023

akaszynski commented Mar 21, 2023

akaszynski left a comment

tkoyama010 commented Mar 22, 2023 •

edited

adeak left a comment

Move Matplotlib to hard requirement #4156

Move Matplotlib to hard requirement #4156

Conversation

tkoyama010 commented Mar 18, 2023

Overview

Details

akaszynski left a comment

Choose a reason for hiding this comment

codecov bot commented Mar 18, 2023 • edited

Codecov Report

adeak left a comment • edited

Choose a reason for hiding this comment

adeak left a comment

Choose a reason for hiding this comment

adeak Mar 19, 2023

Choose a reason for hiding this comment

akaszynski Mar 22, 2023

Choose a reason for hiding this comment

akaszynski commented Mar 20, 2023

banesullivan commented Mar 20, 2023

akaszynski commented Mar 20, 2023

banesullivan commented Mar 20, 2023

adeak commented Mar 21, 2023 • edited

tkoyama010 commented Mar 21, 2023

akaszynski commented Mar 21, 2023

akaszynski left a comment

Choose a reason for hiding this comment

tkoyama010 commented Mar 22, 2023 • edited

adeak left a comment

Choose a reason for hiding this comment

codecov bot commented Mar 18, 2023 •

edited

adeak left a comment •

edited

adeak commented Mar 21, 2023 •

edited

tkoyama010 commented Mar 22, 2023 •

edited