Unify properties

[jo-mueller]

Hi @zoccoler ,

here comes the result of my works in synchronizing functionality between the scatter and the histogram. The key changes are the _create_plot() function that creates the plot. I added new functions _draw_selection_on_plot and color_indices_to_rgba which, as the name suggests, handle the overlay drawing and make it easier to retrieve the color_indices as rgba values without the need to access private methods.

I also replaced some code by functions I found elsewhere. For the statistical histogram, for instance, there's a scipy function that essentially does that :)

PS: Tests haven't been updated yet (WIP) and will likely fail.

Long copilot description

This pull request introduces a new abstract base class, Artist, in the src/biaplotter/artists_base.py file. The class is designed to serve as a foundation for creating customizable plotting components in the BiAPlotter framework. It provides a structured interface for managing data, colormaps, and visibility, while also emitting signals for changes in data and color indices.

New Abstract Base Class for Plotting (Artist)

• Class Definition: Added the Artist class, which inherits from ABC and defines an interface for plotting components. It includes parameters for axes (ax), data (data), colormaps (overlay_colormap), and color indices (color_indices).
• Signals: Introduced data_changed_signal and color_indices_changed_signal to notify changes in the data and color_indices properties, respectively.
• Abstract Methods: Defined three abstract methods—_create_plot, _draw_selection_on_plot, and color_indices_to_rgba—to enforce implementation in derived classes.

Plot Management and Customization

• Dynamic Plot Updates: Added methods like _update_axes_limits, _modify_plot, and _remove_artists to dynamically adjust axes limits, modify plots, and remove plot elements.

[jo-mueller]

@zoccoler I think this is good to go. I fixed the tests and updated the documentation notebooks 👍

[zoccoler]

Hi @jo-mueller ,

I approved the changes and added back the functionality of cmin (now a property), which can set a minimal counts value for the histogram (values below that will be turned to np.nan and, therefore, become transparent). The default value is 0 as before, so nothing should change with this value.

Minor updates are:

• updating notebooks with these changes (cmin property and reset method)
• documentation
• pinning minimal numpy version (due to using np.nanpercentile now)
• bumping biaplotter version to 0.2.0

If you agree, please merge and I am also in favor of making the new release right after that.

[jo-mueller]

@zoccoler Awe-some! Will merge and release and then be able to merge a bunch of stuff right after that for the clusters-plotter 👍