deepbridge-synthetic 2.0.0

Privacy-Preserving Synthetic Data Generation - Standalone library

DeepBridge Synthetic

Privacy-Preserving Synthetic Data Generation

Standalone Package - No Dependencies on DeepBridge!

This package can be used independently for synthetic data generation. It was extracted from DeepBridge v1.x to provide a focused, standalone solution. See Migration Guide if migrating from v1.x.

Installation

pip install deepbridge-synthetic

Quick Start

from deepbridge_synthetic import Synthesize

# Generate synthetic data
synthesizer = Synthesize(
    data=original_df,
    method='gaussian_copula'
)

synthetic_df = synthesizer.generate(n_samples=10000)

Features

• Gaussian Copula: Statistical modeling for synthetic data
• Privacy-Preserving: Generate data while protecting privacy
• Quality Metrics: Evaluate synthetic data quality
• Distributed Computing: Uses Dask for large datasets
• Multiple Methods: Various generation algorithms

Documentation

Full documentation: https://deepbridge.readthedocs.io/en/latest/synthetic/

Related Projects

• deepbridge - Model Validation Toolkit
• deepbridge-distillation - Model Distillation

License

MIT License - see LICENSE