YData
Open-source synthetic data generation for tabular and time-series data.
About
YData Synthetic (ydata-synthetic) is an open-source Python package for generating synthetic tabular and time-series data with generative models. It bundles GAN-family generators including Vanilla GAN, CGAN, WGAN, WGAN-GP, DRAGAN, CTGAN, and TimeGAN, plus a Streamlit-based guided UI that walks users from data loading to synthetic sample visualization.
The project is positioned as an educational on-ramp to generative AI for data: data augmentation, bias mitigation, dataset balancing, and privacy-preserving data sharing. The maintainers recommend ydata-sdk for production-scale quality and performance, keeping ydata-synthetic for exploratory and educational use.
What it can do
- GAN-based synthetic tabular data generation (CTGAN, WGAN-GP, and more)
- Time-series synthesis with TimeGAN and DoppelGANger
- Guided Streamlit UI for generation workflows
- Dataset balancing and bias mitigation tooling
- Evaluation metrics for synthetic data quality
Setup
pip install ydata-synthetic, then follow the example notebooks. For production scale, the maintainers point to ydata-sdk and YData Fabric.
Prerequisites
Python with TensorFlow 2.0; pandas familiarity.
Use cases
Works with
Trust score — 7/10
Scored 7/10 (metrics tier, metrics-v1). Evidence: Verified via vendor site/docs/GitHub and reputable coverage, wave 107; Trending score 68/100 from researcher assessment.
Sources
Every fact above was checked against these sources. Where a claim comes from the provider and couldn't be independently verified, the text says so.