Logistic Regression Curve Plot — plotnine

A logistic regression visualization showing the characteristic S-shaped (sigmoid) probability curve for binary classification. The plot displays data points colored by their binary class, the fitted logistic curve representing predicted probabilities, confidence intervals around the curve, and an optional decision threshold line. This visualization is essential for understanding how a logistic model maps continuous input features to class probabilities.

Logistic Regression Curve Plot rendered with plotnine

Python source (plotnine)

""" anyplot.ai
logistic-regression: Logistic Regression Curve Plot
Library: plotnine 0.15.4 | Python 3.13.13
Quality: 94/100 | Updated: 2026-05-18
"""

import os

import numpy as np
import pandas as pd
import statsmodels.api as sm
from plotnine import (
    aes,
    element_line,
    element_rect,
    element_text,
    geom_hline,
    geom_line,
    geom_point,
    geom_ribbon,
    ggplot,
    labs,
    position_jitter,
    scale_color_manual,
    theme,
    theme_minimal,
)


# Theme tokens
THEME = os.getenv("ANYPLOT_THEME", "light")
PAGE_BG = "#FAF8F1" if THEME == "light" else "#1A1A17"
ELEVATED_BG = "#FFFDF6" if THEME == "light" else "#242420"
INK = "#1A1A17" if THEME == "light" else "#F0EFE8"
INK_SOFT = "#4A4A44" if THEME == "light" else "#B8B7B0"

# imprint semantic anchors
FAIL_COLOR = "#AE3030"  # imprint red - fail
PASS_COLOR = "#009E73"  # imprint green - pass
CURVE_COLOR = "#4467A3"  # imprint blue - regression curve

# Data - Exam score vs Pass/Fail outcome
np.random.seed(42)
n_samples = 150

# Generate exam scores with different distributions for pass/fail
scores_fail = np.random.normal(45, 12, 60)  # Lower scores tend to fail
scores_pass = np.random.normal(70, 10, 90)  # Higher scores tend to pass
scores = np.concatenate([scores_fail, scores_pass])
outcomes = np.concatenate([np.zeros(60), np.ones(90)])

# Add some noise to outcomes for realism
flip_indices = np.random.choice(n_samples, size=15, replace=False)
outcomes[flip_indices] = 1 - outcomes[flip_indices]

# Clip scores to reasonable range
scores = np.clip(scores, 20, 100)

# Fit logistic regression using statsmodels
X = sm.add_constant(scores)
model = sm.Logit(outcomes, X).fit(disp=0)

# Create smooth curve for predictions with confidence intervals
x_curve = np.linspace(20, 100, 200)
X_curve = sm.add_constant(x_curve)
predictions = model.get_prediction(X_curve)
y_pred = predictions.predicted
conf_int = predictions.conf_int(alpha=0.05)
y_lower = conf_int[:, 0]
y_upper = conf_int[:, 1]

# Create dataframes
df_points = pd.DataFrame(
    {"score": scores, "outcome": outcomes, "class": ["Fail" if o == 0 else "Pass" for o in outcomes]}
)

df_curve = pd.DataFrame({"score": x_curve, "probability": y_pred, "lower": y_lower, "upper": y_upper})

# Theme
anyplot_theme = theme(
    figure_size=(16, 9),
    plot_background=element_rect(fill=PAGE_BG, color=PAGE_BG),
    panel_background=element_rect(fill=PAGE_BG),
    panel_grid_major=element_line(color=INK, size=0.3, alpha=0.08),
    panel_grid_minor=element_line(color=INK, size=0.2, alpha=0.04),
    panel_border=element_rect(color=INK_SOFT, fill=None, size=0.5),
    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),
    plot_title=element_text(size=24, color=INK),
    legend_background=element_rect(fill=ELEVATED_BG, color=INK_SOFT),
    legend_text=element_text(size=16, color=INK_SOFT),
    legend_title=element_text(size=18, color=INK),
)

# Create plot
plot = (
    ggplot()
    # Confidence interval ribbon
    + geom_ribbon(data=df_curve, mapping=aes(x="score", ymin="lower", ymax="upper"), alpha=0.25, fill=CURVE_COLOR)
    # Fitted logistic curve
    + geom_line(data=df_curve, mapping=aes(x="score", y="probability"), color=CURVE_COLOR, size=2)
    # Decision threshold line at p=0.5
    + geom_hline(yintercept=0.5, linetype="dashed", color=INK_SOFT, size=1, alpha=0.6)
    # Data points with jitter
    + geom_point(
        data=df_points,
        mapping=aes(x="score", y="outcome", color="class"),
        size=4,
        alpha=0.6,
        position=position_jitter(width=0, height=0.03),
    )
    # Colors
    + scale_color_manual(values={"Fail": FAIL_COLOR, "Pass": PASS_COLOR})
    # Labels
    + labs(
        title="logistic-regression · python · plotnine · anyplot.ai",
        x="Exam Score (points)",
        y="Probability of Passing",
        color="Outcome",
    )
    + theme_minimal()
    + anyplot_theme
)

# Save
plot.save(f"plot-{THEME}.png", dpi=300)

Part of Logistic Regression Curve Plot on anyplot.ai.

Other implementations