A partial dependence plot (PDP) showing the marginal effect of one feature on the predicted outcome of a machine learning model. The plot displays how predictions change as a feature varies across its range, while averaging over the effects of all other features. This visualization is essential for understanding the relationship between individual features and model predictions in interpretable machine learning.
""" anyplot.ai
pdp-basic: Partial Dependence Plot
Library: pygal 3.1.0 | Python 3.13.13
Quality: 91/100 | Updated: 2026-05-15
"""
import os
import numpy as np
import pygal
from pygal.style import Style
from sklearn.datasets import make_regression
from sklearn.ensemble import GradientBoostingRegressor
from sklearn.inspection import partial_dependence
# Theme tokens
THEME = os.getenv("ANYPLOT_THEME", "light")
PAGE_BG = "#FAF8F1" if THEME == "light" else "#1A1A17"
INK = "#1A1A17" if THEME == "light" else "#F0EFE8"
INK_MUTED = "#6B6A63" if THEME == "light" else "#A8A79F"
IMPRINT = ("#009E73", "#C475FD", "#4467A3", "#BD8233", "#AE3030", "#2ABCCD", "#954477")
# Generate data and train model
np.random.seed(42)
X, y = make_regression(n_samples=500, n_features=5, noise=10, random_state=42)
model = GradientBoostingRegressor(n_estimators=50, max_depth=3, random_state=42)
model.fit(X, y)
# Calculate partial dependence for feature 0
feature_idx = 0
pdp_result = partial_dependence(model, X, features=[feature_idx], kind="average", grid_resolution=100)
feature_values = pdp_result["grid_values"][0]
pd_values = pdp_result["average"][0]
# Calculate confidence interval using individual predictions
pdp_individual = partial_dependence(model, X, features=[feature_idx], kind="individual", grid_resolution=100)
individual_preds = pdp_individual["individual"][0]
pd_std = np.std(individual_preds, axis=0)
ci_lower = pd_values - 1.96 * pd_std / np.sqrt(len(X))
ci_upper = pd_values + 1.96 * pd_std / np.sqrt(len(X))
# Custom style for theme-adaptive rendering
custom_style = Style(
background=PAGE_BG,
plot_background=PAGE_BG,
foreground=INK,
foreground_strong=INK,
foreground_subtle=INK_MUTED,
colors=IMPRINT,
title_font_size=28,
label_font_size=22,
major_label_font_size=18,
legend_font_size=16,
value_font_size=14,
stroke_width=3,
)
# Create XY chart for PDP line plot
chart = pygal.XY(
width=4800,
height=2700,
style=custom_style,
title="pdp-basic · pygal · anyplot.ai",
x_title="Feature Value (standardized)",
y_title="Partial Dependence",
show_legend=True,
legend_at_bottom=True,
show_dots=False,
stroke_style={"width": 4},
show_x_guides=True,
show_y_guides=True,
dots_size=6,
margin_bottom=150,
)
# Create XY data points for main PDP line
pdp_points = [(float(x), float(y)) for x, y in zip(feature_values, pd_values, strict=True)]
# Create confidence interval points
ci_upper_points = [(float(x), float(y)) for x, y in zip(feature_values, ci_upper, strict=True)]
ci_lower_points = [(float(x), float(y)) for x, y in zip(feature_values, ci_lower, strict=True)]
# Add data series
chart.add("Partial Dependence", pdp_points, stroke_style={"width": 5})
chart.add("95% CI Upper", ci_upper_points, stroke_style={"width": 2, "dasharray": "8,4"})
chart.add("95% CI Lower", ci_lower_points, stroke_style={"width": 2, "dasharray": "8,4"})
# Add rug plot showing training data distribution
rug_indices = np.random.choice(len(X), size=min(40, len(X)), replace=False)
rug_x_values = X[rug_indices, feature_idx]
y_min = float(np.min(pd_values) - 0.15 * (np.max(pd_values) - np.min(pd_values)))
rug_points = [(float(x), y_min) for x in sorted(rug_x_values)]
chart.add("Training Data (rug)", rug_points, stroke=False, dots_size=5)
# Save as PNG and HTML
chart.render_to_png(f"plot-{THEME}.png")
with open(f"plot-{THEME}.html", "wb") as f:
f.write(chart.render())
Part of Partial Dependence Plot on anyplot.ai.