matplotlib / matplotlib/matplotlib

[MNT]: Confusing edgecolor behavior for Patch and Collection

Abierto
#29,261 12 comentarios 0 reacciones 0 asignados Ver en GitHub

Nadie ha tomado este issue todavía.

Maintenance
Lenguaje dominante
Python
Estrellas
23.2k
Forks
8.5k
Merge medio
1 d 6 h
PR fusionados (30 d)
66

Descripción

Noted while reviewing documentation on the current behavoir in #29252.

Summary

Patches and Collections do not draw edges by default. This was introduced in #5596 for matplotlib 2.0. It seems that this was first achieved by a linewidth 0, but it this was changed to using an edgecolor of "none" while maintaining a finite linewidth.

As a consequence rcParams["patch.edgecolor"] does not have an effect by default. There are, however two cases when rcParams["patch.edgecolor"] becomes relevant:

  1. If facecolor is "none", Patch and Collection take rcParams["patch.edgecolor"] as edgecolor.
    I assume, this is done to prevent the artist being completely invisible (no facecolor and no edgecolor).
  2. If rcParams["patch.force_edgecolor"] is set to true, rcParams["patch.edgecolor"] is used instead of "none" as the

This is overall very confusing.

Ping @efiring, who introduced rcParams["patch.force_edgecolor"] in #6904. Can you comment on the background and whether the proposal below is a reasonable alternative?

Proposed fix

Can we switch to the following logic: patch.edgecolor is generally used as edgecolor and now defaults to "none" (to achieve the same behavior as before). To handle case 1. (prevent invisible artists) introduce a new rcParams["patch.edgecolor_fallback"] that has a special color only for that case.

Needed changes to default rcParams

matplotlibrc before:

#patch.edgecolor: black
#patch.force_edgecolor: False

matplotlibrc after

#patch.edgecolor: "none"            # default edgecolor for Patch and Collection
#patch.edgecolor_fallback: "black"  # edgecolor to be used if facecolor is "none" to prevent completely invisible artists
Needed changes to styles

Existing styles would migrate

  • If patch.force_edgecolor was False:
    from:
    patch.edgecolor: COLOR
    #patch.force_edgecolor: False
    
    to
    #patch.edgecolor: "none"
    patch.edgecolor_fallback:  COLOR
    
  • If patch.force_edgecolor was True:
    from:
    patch.edgecolor: COLOR
    patch.force_edgecolor: True
    
    to
    patch.edgecolor: COLOR
    patch.edgecolor_fallback:  COLOR
    
Transition path
  • patch.force_edgecolor gets deprecated and raises a warning when used on rcParams.
  • styles that don't use any of patch.edgecolor or patch.force_edgecolor do not have to do anything.
  • styles that have set patch.edgecolor but not patch.force_edgecolor are a very edge case. They either have overlooked that the default force_edgecolor=False makes the setting ineffective, or they have intentionally only configured the somewhat arcane edge color fallback for non-filled artists. In the former case, they get "auto-fixed" through the new behavior. I'd argue that we can affort to not runtime-warn on the latter case (but if really wanted, one could add a check at least for loading stylesheets).

Guía de contribución

Abrir la guía de contribución

Primeros pasos

  1. Lee el issue completo y luego la guía de contribución del proyecto.
  2. Comenta en el issue que vas a ocuparte — evita que dos personas hagan lo mismo.
  3. Haz un fork del repositorio y trabaja en una rama.
  4. Abre un pull request que haga referencia al número del issue.

Línea de trabajo

No se nombra ningún archivo fuente ni ninguna prueba. Empieza trazando el manejo de edgecolor de Patch y Collection y las configuraciones de rcParams, matplotlibrc y style descritas en la propuesta; el trabajo está terminado cuando los valores predeterminados propuestos, el comportamiento de fallback, la deprecación de force_edgecolor y las migraciones de style existentes se hayan validado de forma coherente.

Escrito por el modelo de indexación a partir del texto del issue.

Evaluación

Stack tecnológico
python
Área
data-visualization
Tipo de issue
Nueva funcionalidad
Dificultad
5/5
Tiempo estimado
Más de una semana
Estado de actividad
Estancado
Claridad
Bastante claro
Aptitud para principiantes
35/100

Recibe los nuevos issues en tu correo

Un resumen breve de issues de GitHub para principiantes.