Development version

This is the latest (dev) documentation. It may contain unreleased features or breaking changes. For the stable release, use version 0.13.

Plotting Gallery

sklearn-genetic-opt ships three plot functions. This gallery shows each one with its main options.

Setup

import matplotlib.pyplot as plt
import pandas as pd
from sklearn.datasets import load_diabetes
from sklearn.model_selection import train_test_split
from sklearn.tree import DecisionTreeRegressor

from sklearn_genetic import GASearchCV
from sklearn_genetic.plots import plot_fitness_evolution, plot_history, plot_search_space
from sklearn_genetic.space import Categorical, Continuous, Integer

X, y = load_diabetes(return_X_y=True)
X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.33, random_state=42)

search = GASearchCV(
    DecisionTreeRegressor(random_state=42),
    cv=2,
    scoring="r2",
    population_size=8,
    generations=10,
    tournament_size=3,
    elitism=True,
    crossover_probability=0.9,
    mutation_probability=0.05,
    param_grid={
        "ccp_alpha": Continuous(0, 1),
        "criterion": Categorical(["squared_error", "absolute_error"]),
        "max_depth": Integer(2, 20),
        "min_samples_split": Integer(2, 30),
    },
    n_jobs=1,
)

search.fit(X_train, y_train)

Fitness Plots

plot_fitness_evolution shows how fitness changes across generations.

Default View

plot_fitness_evolution(search)
plt.show()

Multi-Metric With Smoothing

plot_fitness_evolution(
    search,
    metrics=["fitness_best", "fitness", "fitness_max"],
    window=2,
    kind="line",
    title="Fitness comparison with smoothing",
)
plt.show()

Parameter	Description
metrics	List of history fields to plot. Default: ["fitness_best"]
window	Rolling average window size. Default: 1 (no smoothing)
kind	"line" or "bar". Default: "line"
title	Chart title

History Plots

plot_history can plot any field from history or logbook. Use it to inspect fitness signals, diversity indicators, or optimizer-control events.

History Fields

plot_history(
    search,
    fields=["fitness_best", "fitness", "unique_individual_ratio", "genotype_diversity"],
    kind="line",
    subplots=True,
    title="Optimizer history overview",
)
plt.show()

Logbook Fields

plot_history(
    search,
    fields=["score", "fit_time", "score_time"],
    source="logbook",
    kind="area",
    title="Logbook fields from candidate evaluations",
)
plt.show()

Parameter	Description
fields	List of column names to plot
source	"history" (generation stats) or "logbook" (candidate evaluations)
kind	"line", "bar", or "area"
subplots	True to draw each field in its own subplot

Search-Space Plots

The search-space plots show how the optimizer explored the parameter space.

Pair Plot

Shows relationships between sampled parameters. Points are colored by a category column when hue is set.

plot_search_space(
    search,
    features=["ccp_alpha", "max_depth", "min_samples_split", "criterion"],
    hue="criterion",
    kind="pair",
)
plt.show()

Heatmap

Quick correlation view of sampled numeric parameters.

plot_search_space(
    search,
    features=["ccp_alpha", "max_depth", "min_samples_split"],
    kind="heatmap",
)
plt.show()

Parameter	Description
features	Parameter names to include. Omit to use all numeric parameters
hue	Categorical column to use for color coding (pair plot only)
kind	"pair" or "heatmap"

Reading History Directly

All plot functions read from search.history. You can also work with it directly as a DataFrame for custom reporting.

history = pd.DataFrame(search.history)

# Fields available in history
telemetry_columns = [
    "gen",
    "fitness_best",
    "fitness",
    "fitness_max",
    "unique_individual_ratio",
    "genotype_diversity",
    "mutation_probability",
    "selection_pressure",
    "random_immigrants",
    "duplicate_replacements",
    "local_refinements",
]

available = [c for c in telemetry_columns if c in history.columns]
print(history[available].tail())

When to Use Each Plot

Plot	Use when
plot_fitness_evolution	Quick fitness trend overview
plot_history	Inspecting telemetry: diversity, stagnation, control events
plot_search_space with kind="pair"	Understanding parameter interactions
plot_search_space with kind="heatmap"	Spotting correlations between numeric parameters

See Also

• Plots API — full parameter reference for all three functions
• Understanding Cross-Validation — what the generation log columns mean
• Advanced Optimizer Control — interpreting diversity and stagnation signals