Implementation:DistrictDataLabs Yellowbrick VisualPipeline

Knowledge Sources	DistrictDataLabs_Yellowbrick
Domains	Model_Selection, Visualization
Last Updated	2026-02-08 05:00 GMT

Overview

Extended scikit-learn Pipeline that integrates Yellowbrick visualizers into the pipeline workflow, enabling automatic rendering of all visual steps.

Description

The VisualPipeline subclasses scikit-learn's Pipeline and adds the ability to detect and render Yellowbrick Visualizer steps. The show method iterates over all visualizer steps and renders each one, optionally saving figures to disk. The visual_steps property provides a read-only dictionary of visualizer steps in the pipeline.

Usage

Import VisualPipeline when building machine learning pipelines that include Yellowbrick visualizers alongside sklearn transformers and estimators.

Code Reference

Source Location

• Repository: DistrictDataLabs_Yellowbrick
• File: yellowbrick/pipeline.py
• Lines: 1-112

Signature

class VisualPipeline(Pipeline):
    @property
    def visual_steps(self):
        """Read-only dict of visualizer steps."""

    def show(self, outdir=None, ext=".pdf", **kwargs):
        """Renders all visualizer steps."""

    def fit_transform_show(self, X, y=None, outpath=None, **kwargs):
        """Fit, transform, and show all visualizations."""

Import

from yellowbrick.pipeline import VisualPipeline

I/O Contract

Inputs

Name	Type	Required	Description
steps	list of (name, transform)	Yes	Pipeline steps (sklearn convention)
outdir	str	No	Directory to save figures
ext	str	No	Figure file extension (default: ".pdf")

Outputs

Name	Type	Description
visual_steps	dict	Dictionary of name: Visualizer pairs
show()	None	Renders all visualizers

Usage Examples

from sklearn.preprocessing import StandardScaler
from sklearn.svm import SVC
from yellowbrick.pipeline import VisualPipeline
from yellowbrick.features import RadViz

pipe = VisualPipeline([
    ("scaler", StandardScaler()),
    ("radviz", RadViz()),
    ("svc", SVC()),
])

pipe.fit(X, y)
pipe.show(outdir="./figures")

Related Pages

• Environment:DistrictDataLabs_Yellowbrick_Python_Scikit_Learn_Environment