Feature Importance Bar Chart — Bokeh

A horizontal bar chart displaying feature importances from machine learning models, with features sorted by importance value and bars colored by a gradient to emphasize relative contribution. This visualization is essential for model interpretability, helping data scientists and stakeholders understand which features drive model predictions. The horizontal orientation allows for readable feature names of varying lengths.

Feature Importance Bar Chart rendered with Bokeh

Python source (Bokeh)

""" anyplot.ai
bar-feature-importance: Feature Importance Bar Chart
Library: bokeh 3.9.0 | Python 3.13.13
Quality: 72/100 | Updated: 2026-05-10
"""

from bokeh.io import export_png, save
from bokeh.models import ColumnDataSource, LabelSet, LinearColorMapper
from bokeh.palettes import Blues9
from bokeh.plotting import figure
from bokeh.resources import CDN


# Data: Feature importances from a classification model
features = [
    "Income",
    "Credit Score",
    "Age",
    "Employment Years",
    "Debt Ratio",
    "Num Accounts",
    "Loan Amount",
    "Education Level",
    "Num Inquiries",
    "Home Ownership",
    "Payment History",
    "Account Balance",
]
importances = [0.185, 0.162, 0.124, 0.098, 0.089, 0.076, 0.068, 0.058, 0.052, 0.041, 0.032, 0.015]

# Sort by importance (highest at top when plotted)
sorted_pairs = sorted(zip(features, importances, strict=True), key=lambda x: x[1])
features_sorted = [p[0] for p in sorted_pairs]
importances_sorted = [p[1] for p in sorted_pairs]

# Create data source
source = ColumnDataSource(
    data={
        "features": features_sorted,
        "importances": importances_sorted,
        "labels": [f"{imp:.3f}" for imp in importances_sorted],
        "label_x": [imp + 0.008 for imp in importances_sorted],  # offset for label placement
    }
)

# Color mapper for gradient effect (light to dark based on importance)
color_mapper = LinearColorMapper(palette=list(reversed(Blues9)), low=min(importances), high=max(importances))

# Create figure with categorical y-axis
p = figure(
    width=4800,
    height=2700,
    y_range=features_sorted,
    x_range=(0, max(importances) * 1.15),
    title="bar-feature-importance · bokeh · pyplots.ai",
    x_axis_label="Importance Score",
    toolbar_location=None,
)

# Draw horizontal bars with color gradient
p.hbar(
    y="features",
    right="importances",
    height=0.7,
    source=source,
    fill_color={"field": "importances", "transform": color_mapper},
    line_color="#306998",
    line_width=2,
)

# Add value labels at end of bars
labels = LabelSet(
    x="label_x",
    y="features",
    text="labels",
    source=source,
    text_font_size="18pt",
    text_color="#306998",
    text_baseline="middle",
)
p.add_layout(labels)

# Styling for large canvas
p.title.text_font_size = "32pt"
p.title.text_color = "#306998"
p.xaxis.axis_label_text_font_size = "24pt"
p.xaxis.major_label_text_font_size = "18pt"
p.yaxis.major_label_text_font_size = "18pt"

# Grid styling (subtle)
p.xgrid.grid_line_color = "#cccccc"
p.xgrid.grid_line_alpha = 0.3
p.xgrid.grid_line_dash = "dashed"
p.ygrid.grid_line_color = None

# Axis styling
p.xaxis.axis_line_color = "#666666"
p.yaxis.axis_line_color = "#666666"
p.xaxis.major_tick_line_color = "#666666"
p.yaxis.major_tick_line_color = "#666666"

# Background
p.background_fill_color = "#fafafa"
p.border_fill_color = "white"
p.outline_line_color = None

# Save outputs
export_png(p, filename="plot.png")
save(p, filename="plot.html", resources=CDN, title="bar-feature-importance · bokeh · pyplots.ai")

Part of Feature Importance Bar Chart on anyplot.ai.

Feature Importance Bar Chart — Bokeh

Python source (Bokeh)

Other implementations