Plots

`plot_fitness_evolution`(estimator[, metric, ...])	Plot one or more evolution metrics stored in `estimator.history`.
`plot_history`(estimator[, fields, source, ...])	Plot arbitrary history or logbook fields in an easier-to-read layout.
`plot_search_space`(estimator[, height, s, ...])	Plot the sampled search space used during the optimization.

sklearn_genetic.plots.plot_fitness_evolution(estimator, metric='fitness_best', metrics=None, *, kind='line', window=None, ax=None, title=None, palette=None)[source]

Plot one or more evolution metrics stored in estimator.history.

Parameters:

estimator: estimator object: A fitted estimator from GASearchCV or GAFeatureSelectionCV.
metric: str, default=”fitness_best”: Backward-compatible name for a single metric to plot.
metrics: list[str] | tuple[str] | None, default=None: Optional collection of history fields to plot together.
kind: {“line”, “bar”, “area”, “step”}, default=”line”: Plot style.
window: int | None, default=None: Optional rolling window applied before plotting.
ax: matplotlib.axes.Axes | None, default=None: Axis to draw on. A new axis is created if omitted.
title: str | None, default=None: Optional plot title.
palette: str | None, default=None: Optional seaborn palette name used for multiple series.

Returns:

matplotlib.axes.Axes: The axis used for the plot.

sklearn_genetic.plots.plot_history(estimator, fields=None, *, source='history', kind='line', rolling=None, subplots=None, figsize=None, title=None, palette=None)[source]

Plot arbitrary history or logbook fields in an easier-to-read layout.

Parameters:

estimator: estimator object: A fitted estimator with history or logbook data.
fields: list[str] | str | None, default=None: Explicit fields to plot. If omitted, numeric fields are selected automatically from the chosen source.
source: {“history”, “logbook”}, default=”history”: Data source to plot from.
kind: {“line”, “bar”, “area”, “step”}, default=”line”: Plot style for each field.
rolling: int | None, default=None: Optional rolling window applied to the plotted values.
subplots: bool | None, default=None: If True, plot one subplot per field. If False, overlay everything on one axis. If None, a readable default is chosen automatically.
figsize: tuple[float, float] | None, default=None: Optional figure size.
title: str | None, default=None: Optional figure title.
palette: str | None, default=None: Optional seaborn palette name.

Returns:

matplotlib.axes.Axes | numpy.ndarray[matplotlib.axes.Axes]: The created axis or axes.

sklearn_genetic.plots.plot_search_space(estimator, height=2, s=25, features=None, *, kind='pair', hue=None)[source]

Plot the sampled search space used during the optimization.

Parameters:

estimator: estimator object: A fitted estimator from GASearchCV.
height: float, default=2: Height of each facet for pair plots.
s: float, default=25: Marker size for scatter-based plots.
features: list[str] | None, default=None: Subset of fields to plot. If omitted, numeric parameter fields are used.
kind: {“pair”, “heatmap”}, default=”pair”: Plot style. pair shows pairwise relationships, while heatmap shows a correlation matrix.
hue: str | None, default=None: Optional column used to color the pair plot.

Returns:

seaborn.axisgrid.PairGrid | matplotlib.axes.Axes: Pair grid or heatmap axis depending on kind.