decisioncanvas 1.0.1

Easy decision boundary visualization for classifiers

decisioncanvas

A Python library for visualizing machine learning classifier decision boundaries with automatic dimensionality reduction and standardization.

Overview

decisioncanvas provides a simple, reliable way to create publication-quality visualizations of how classification models separate data. The library automatically handles high-dimensional data through PCA dimensionality reduction and ensures consistent scaling through built-in standardization.

Key Features

• Simple API: Single function call for complete visualization
• High-dimensional support: Automatic PCA with explained variance reporting
• Scikit-learn compatibility: Works with any scikit-learn classifier
• Multi-class support: Handles binary and multi-class classification
• Built-in preprocessing: Automatic feature standardization
• Customizable visualization: Configurable grid resolution, colors, and styling
• Educational focus: Ideal for learning and teaching machine learning concepts

Installation

pip install decisioncanvas

Quick Start

from decisioncanvas import plot_decision_boundary
from sklearn.datasets import load_iris
from sklearn.linear_model import LogisticRegression

# Load data and create model
X, y = load_iris(return_X_y=True)
model = LogisticRegression(max_iter=300)

# Visualize decision boundary
plot_decision_boundary(model, X, y)

Usage Examples

Basic Classification

from decisioncanvas import plot_decision_boundary
from sklearn.datasets import load_iris
from sklearn.model_selection import train_test_split
from sklearn.linear_model import LogisticRegression

# Prepare data
data = load_iris()
X_train, X_test, y_train, y_test = train_test_split(
    data.data, data.target, test_size=0.3, random_state=42
)

# Train model
model = LogisticRegression(max_iter=200)
model.fit(X_train, y_train)

# Visualize with custom labels
plot_decision_boundary(
    model,
    X_train,
    y_train,
    class_names=data.target_names,
    title='Iris Classification Decision Boundary'
)

Advanced Customization

plot_decision_boundary(
    model,
    X,
    y,
    grid_resolution=400,
    alpha=0.4,
    padding=1.5,
    title='Custom Decision Boundary Visualization',
    figsize=(10, 8)
)

API Reference

plot_decision_boundary

plot_decision_boundary(
    model, X, y,
    class_names=None,
    standardize=True,
    pca_components=2,
    grid_resolution=300,
    padding=1.0,
    cmap_light=None,
    cmap_bold=None,
    alpha=0.3,
    title="Decision Boundary",
    fit_model=True,
    figsize=(8, 6),
    random_state=None
)

Parameters

Parameter	Type	Description	Default
model	estimator	Scikit-learn compatible classifier	Required
X	array-like	Feature matrix (n_samples, n_features)	Required
y	array-like	Target labels (n_samples,)	Required
class_names	list or dict	Custom class labels for legend	None
standardize	bool	Apply feature standardization	True
pca_components	int	Number of PCA components for dimensionality reduction	2
grid_resolution	int	Resolution of decision boundary grid	300
padding	float	Padding around data points in plot	1.0
cmap_light	colormap	Colormap for decision regions	None (auto)
cmap_bold	list	Colors for data points	None (auto)
alpha	float	Transparency of decision regions	0.3
title	str	Plot title	"Decision Boundary"
fit_model	bool	Whether to fit the model on provided data	True
figsize	tuple	Figure size (width, height)	(8, 6)
random_state	int	Random state for PCA reproducibility	None

Use Cases

• Model evaluation: Visualize how different classifiers separate your data
• Educational purposes: Demonstrate classification concepts in tutorials
• Research presentations: Create publication-ready decision boundary plots
• Debugging: Identify potential overfitting or data separation issues

Requirements

• Python >= 3.7
• numpy >= 1.18.0
• scikit-learn >= 0.24.0
• matplotlib >= 3.2.0

Contributing

Contributions are welcome! Please feel free to submit issues, feature requests, or pull requests.

1. Fork the repository
2. Create a feature branch
3. Make your changes
4. Add tests if applicable
5. Submit a pull request

License

This project is licensed under the MIT License - see the LICENSE file for details.

Authors

• Krunal Wankhade - [e-mail]
• Parimal Kalpande - [e-mail]

Links

• PyPI: https://pypi.org/project/decisioncanvas/
• GitHub: https://github.com/KrunalWankhade9021/Decesion-Canvas