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: plotnine 0.15.4 | Python 3.13.13
Quality: 94/100 | Updated: 2026-05-15
"""
import os
import numpy as np
import pandas as pd
from plotnine import (
aes,
element_blank,
element_line,
element_rect,
element_text,
geom_line,
geom_ribbon,
geom_segment,
ggplot,
labs,
theme,
theme_minimal,
)
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_SOFT = "#4A4A44" if THEME == "light" else "#B8B7B0"
BRAND = "#009E73" # Okabe-Ito position 1
ACCENT = "#C475FD" # Okabe-Ito position 2 for rug
# Data - Train a model and compute partial dependence
np.random.seed(42)
# Generate synthetic data for a regression problem
X, y = make_regression(n_samples=500, n_features=5, noise=10, random_state=42)
feature_names = ["Energy Consumption", "Room Size", "Occupancy Rate", "Ventilation", "Age"]
# Train a gradient boosting model
model = GradientBoostingRegressor(n_estimators=100, max_depth=3, random_state=42)
model.fit(X, y)
# Compute partial dependence for Room Size (feature index 1)
feature_idx = 1
# Get partial dependence values
pd_results = partial_dependence(model, X, features=[feature_idx], kind="average", grid_resolution=80)
grid_actual = pd_results["grid_values"][0]
# Compute ICE curves for confidence interval estimation
pd_individual = partial_dependence(model, X, features=[feature_idx], kind="individual", grid_resolution=80)
ice_values = pd_individual["individual"][0]
# Calculate confidence interval (mean ± 1.96 * std for 95% CI)
pd_mean = ice_values.mean(axis=0)
pd_std = ice_values.std(axis=0)
ci_lower = pd_mean - 1.96 * pd_std
ci_upper = pd_mean + 1.96 * pd_std
# Create DataFrame for plotting
df = pd.DataFrame(
{"feature_value": grid_actual, "partial_dependence": pd_mean, "ci_lower": ci_lower, "ci_upper": ci_upper}
)
# Rug data - sample of training data positioned at the axis baseline
y_min = df["partial_dependence"].min()
y_max = df["partial_dependence"].max()
rug_height = (y_max - y_min) * 0.03
rug_sample = pd.DataFrame({"x": X[:100, feature_idx], "y": y_min - rug_height, "yend": y_min})
# Plot
plot = (
ggplot(df, aes(x="feature_value", y="partial_dependence"))
+ geom_ribbon(aes(ymin="ci_lower", ymax="ci_upper"), alpha=0.15, fill=BRAND, color=BRAND, size=0.5)
+ geom_line(color=BRAND, size=2)
+ geom_segment(data=rug_sample, mapping=aes(x="x", xend="x", y="y", yend="yend"), color=ACCENT, alpha=0.6, size=0.8)
+ labs(
title="pdp-basic · plotnine · anyplot.ai",
x="Room Size (standardized)",
y="Partial Dependence (avg. prediction)",
)
+ theme_minimal()
+ theme(
figure_size=(16, 9),
plot_background=element_rect(fill=PAGE_BG, color=PAGE_BG),
panel_background=element_rect(fill=PAGE_BG, color=PAGE_BG),
panel_grid_major=element_line(color=INK_SOFT, size=0.3, alpha=0.10),
panel_grid_minor=element_blank(),
panel_border=element_rect(color=INK_SOFT, fill=None, size=0.5),
plot_title=element_text(size=24, weight="bold", color=INK),
axis_title=element_text(size=20, color=INK),
axis_text=element_text(size=16, color=INK_SOFT),
axis_line=element_line(color=INK_SOFT, size=0.5),
)
)
# Save
plot.save(f"plot-{THEME}.png", dpi=300, verbose=False)
Part of Partial Dependence Plot on anyplot.ai.