PR G1-G2: replace mat_plot_2d.plot_array with _plot_array() bridge in…

[Jammy2211]

… all plotters

• Add autogalaxy/plot/plots/overlays.py with helpers to extract critical curves, caustics, and profile centres from Visuals2D as plain arrays
• Add _plot_array() and _plot_grid() bridge methods to the Plotter base class that route to the new direct-matplotlib plot_array()/plot_grid() functions
• Update all autogalaxy plotters to use self._plot_array() instead of self.mat_plot_2d.plot_array(), covering: LightProfilePlotter, BasisPlotter, MassPlotter, GalaxyPlotter, GalaxiesPlotter, AdaptPlotter, FitImagingPlotter, FitEllipsePlotter, FitEllipsePDFPlotter
• Remove redundant (and incorrect) cmap vmin/vmax mutation in FitImagingPlotter.figures_2d_of_galaxies before subtracted_image plot

https://claude.ai/code/session_0154SgR1ThCsfpbMZLnbS85k

[Copilot]

Pull request overview

This PR refactors Autogalaxy plotters to route common plotting calls through new Plotter helper methods, and introduces new utilities intended to convert Visuals2D overlay data into plain NumPy arrays for future plotting backends.

Changes:

• Add _plot_array() / _plot_grid() bridge methods to Plotter and update multiple plotters to call them.
• Add new autogalaxy.plot.plots.overlays helper functions to extract/normalize overlay data from Visuals2D.
• Remove some plot-specific colormap min/max manipulation in FitImagingPlotter and simplify AutoLabels usage in MassPlotter.

Reviewed changes

Copilot reviewed 11 out of 11 changed files in this pull request and generated 3 comments.

Show a summary per file

File	Description
autogalaxy/plot/abstract_plotters.py	Adds _plot_array / _plot_grid bridge methods used across plotters.
autogalaxy/profiles/plot/light_profile_plotters.py	Switches image plotting to the new _plot_array helper.
autogalaxy/profiles/plot/basis_plotters.py	Switches basis subplot image plotting to _plot_array.
autogalaxy/plot/mass_plotter.py	Routes array plots through _plot_array and removes cb_unit from AutoLabels.
autogalaxy/imaging/plot/fit_imaging_plotters.py	Uses _plot_array for galaxy image plots and removes NumPy-based vmin manipulation.
autogalaxy/galaxy/plot/galaxy_plotters.py	Switches galaxy image plotting to _plot_array.
autogalaxy/galaxy/plot/galaxies_plotters.py	Switches plane image/grid plotting to _plot_array / _plot_grid.
autogalaxy/galaxy/plot/adapt_plotters.py	Switches adapt image plotting to _plot_array.
autogalaxy/ellipse/plot/fit_ellipse_plotters.py	Switches ellipse plotting to _plot_array.
autogalaxy/plot/plots/overlays.py	Adds new overlay extraction/normalization helpers (currently with correctness concerns).
autogalaxy/plot/plots/init.py	Exposes the new overlay helpers via the package initializer.

---

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.

[Copilot]

Pull request overview

This PR refactors the PyAutoGalaxy plotting layer away from the previous Plotter/MatPlot/Visuals wrappers toward a set of direct matplotlib-backed plotting helper functions, and updates tests/config accordingly.

Changes:

• Introduces autogalaxy.plot.plot_utils (plot_array, plot_grid, _save_subplot, and overlay/curve helpers) and switches public plotting exports to these free functions.
• Replaces multiple Plotter-class implementations with function-based plotting modules (e.g. *_plots.py) for basis / galaxy / imaging / interferometer / quantity / ellipse plotting.
• Updates test suite to call the new plotting functions and adjusts matplotlib savefig monkeypatching and test visualize config.

Reviewed changes

Copilot reviewed 80 out of 83 changed files in this pull request and generated 5 comments.

Show a summary per file

File	Description
pyproject.toml	Changes autoarray dependency to a GitHub branch via VCS URL.
autogalaxy/plot/__init__.py	Replaces previous plotting API exports with free-function exports.
autogalaxy/plot/plot_utils.py	Adds new plotting/overlay helpers and subplot save utility.
autogalaxy/quantity/plot/fit_quantity_plots.py	Adds function-based subplot plotting for FitQuantity.
autogalaxy/quantity/model/plotter_interface.py	Updates plotter interface to call new fit_quantity_plots functions.
autogalaxy/profiles/plot/basis_plots.py	Adds function-based basis subplot plotting.
autogalaxy/galaxy/plot/*_plots.py	Adds function-based plotting for galaxies and adapt images.
autogalaxy/imaging/plot/fit_imaging_plots.py	Adds function-based subplot plotting for FitImaging.
autogalaxy/interferometer/plot/fit_interferometer_plots.py	Adds function-based subplot plotting for FitInterferometer.
autogalaxy/ellipse/plot/fit_ellipse_plots.py	Adds function-based plotting for ellipse fit and ellipse errors.
autogalaxy/ellipse/model/plotter_interface.py	Updates ellipse plotter interface to call new plotting functions.
test_autogalaxy/**	Updates tests to use the new plotting functions and adjusts expectations accordingly.
test_autogalaxy/conftest.py	Patches both pyplot.savefig and Figure.savefig for output interception.
test_autogalaxy/config/visualize.yaml	Simplifies/removes many legacy mat_wrap config sections.
.gitignore	Ignores additional generated test plot outputs.

---

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.

[Copilot]

PR description mentions adding autogalaxy/plot/plots/overlays.py helpers and adding _plot_array() / _plot_grid() bridge methods on a Plotter base class, but the current changeset appears to remove the Plotter-based API entirely and autogalaxy/plot now only exports free functions (and there is no plots/overlays.py in-tree). Please either update the PR description to match the implemented approach or add the missing module/bridge methods if they are still intended.

[Copilot]

autoarray is now pulled from a GitHub branch via a direct VCS URL. This can make installs non-reproducible (branch can move / disappear) and can break downstream packaging (e.g. conda, offline builds). Prefer a released version pin (or a commit SHA if VCS is unavoidable) and document why this override is required.

[Copilot]

disable_data_contours is accepted and passed through (e.g. by subplot_fit_ellipse / PlotterInterfaceEllipse.fit_ellipse), but it is not used inside _plot_data. This means the “no data contours” behavior is currently a no-op. Either implement the contour-disabling behavior (as previously done via the MatPlot contour toggle) or remove the parameter/call-sites to avoid misleading API.

[Copilot]

subplot_fit_ellipse passes ax=axes[1] into _plot_ellipse_residuals, but _plot_ellipse_residuals ignores the ax argument. Because fit_ellipse_plot_util.plot_ellipse_residuals(..., for_subplot=True) internally calls plt.subplot(1, 2, 2), this will draw onto a new axes / the current global figure state, which can lead to incorrect subplot composition. Consider updating plot_ellipse_residuals (or _plot_ellipse_residuals) to accept and use an explicit Axes instead of creating one via plt.subplot.

[Copilot]

_to_lines / _to_positions swallow all exceptions (except Exception: pass). This can silently drop overlays when inputs are malformed and make debugging very difficult. Consider catching only expected exceptions (e.g. TypeError, ValueError) and logging at least a debug warning with the offending type/value.

Suggested change

	except Exception:
	pass
	else:
	try:
	arr = np.array(item.array if hasattr(item, "array") else item)
	if arr.ndim == 2 and arr.shape[1] == 2 and len(arr) > 0:
	result.append(arr)
	except Exception:
	pass
	except (TypeError, ValueError):
	logger.debug(
	"_to_lines: failed to convert list element %r of type %s to array; skipping.",
	sub,
	type(sub),
	exc_info=True,
	)
	else:
	try:
	arr = np.array(item.array if hasattr(item, "array") else item)
	if arr.ndim == 2 and arr.shape[1] == 2 and len(arr) > 0:
	result.append(arr)
	except (TypeError, ValueError):
	logger.debug(
	"_to_lines: failed to convert item %r of type %s to array; skipping.",
	item,
	type(item),
	exc_info=True,
	)