SDVThe Synthetic Data Vault

Single Table

Learn a tabular model to synthesize rows in a table

Multi Table

Learn a relational data model to synthesize multiple, related tables

Sequential

Learn a sequential or time series model to synthesize new events

The Synthetic Data Vault

What can you use synthetic data for?

Use a synthetic data in place of real data for added protection, or use it in addition to your real data as an enhancement.

Test Software

Expand Access

Pilot New Products

Augment Data

Plan Scenarios

View on Github

The SDV Ecosystem

Public, Source-Available Libraries

The SDV is an overall ecosystem for synthetic data models, benchmarks, and metrics. Explore publicly available libraries supporting the SDV. Each can be used as standalone packages for particular needs.

Copulas

Models & generates tabular data with classic statistical methods. Uses multivariate copulas.

Github

Docs

CTGAN

Models & generates tabular data with Deep Learning. Offers CTGAN and TVAE models.

Github

DeepEcho

Models & generates time series data with a mix of classic statistical models and Deep Learning.

Github

RDT

Discovers properties & transforms data for data science use. Reverses the transforms to reproduce realistic data.

Github

Docs

Synthetic Data Vault

Generates synthetic data across single table, relational, and time series data. Supports multiple models & evaluations.

Github

Docs

QUICKSTART

Try it out now!

Quickly discover SDV with just a few lines of code!

from sdv.datasets.demo import download_demo
from sdv.single_table import GaussianCopulaSynthesizer

real_data, metadata = download_demo(
  'single_table', 'fake_hotel_guests')

synthesizer = GaussianCopulaSynthesizer(metadata)
synthesizer.fit(real_data)

synthetic_data = synthesizer.sample(num_rows=10)

from sdv.datasets.demo import download_demo
from sdv.single_table import GaussianCopulaSynthesizer

real_data, metadata = download_demo(
  'single_table', 'fake_hotel_guests')

synthesizer = GaussianCopulaSynthesizer(metadata)
synthesizer.fit(real_data)

synthetic_data = synthesizer.sample(num_rows=10)

Install SDV

Join Us

Join Our Community

Chat with developers across the world. Stay up-to-date with the latest features, blogs, and news.

Twitter

Slack