# sdv-dev/CTGAN

Conditional GAN for generating synthetic tabular data.

Repository: https://github.com/sdv-dev/CTGAN
Canonical: https://ross.abutalabs.com/products/ctgan
Language: Python
License: NOASSERTION
License Family: other
Topics: synthetic-data, generative-adversarial-network, tabular-data, data-generation, synthetic-data-generation
Last push: 2026-08-10T15:01:33+00:00

## Health v2 (maintenance only)
Score: 80/100 (v2, computed 2026-09-03T02:20:16.233290+00:00)
- activity 97, release rhythm 46, longevity 100
- inputs: {"age_days": 2551, "days_push": 23, "days_rel": 201, "gap_med": 108.0, "n_releases_24m": 5}
- flags: no_license
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 1562, forks 332 (observed 2026-08-28T04:05:03.944390+00:00)

## What it is
CTGAN is a Python library of deep learning based synthetic data generators for single-table tabular data, implementing the CTGAN conditional GAN and TVAE models. It learns from real datasets and generates high-fidelity synthetic rows for testing, privacy, and ML workflows.

## Use cases
- generate synthetic tabular data from a csv
- create fake training data with a gan
- augment a small dataset with synthetic rows
- generate privacy-safe test data for a database table
- train a model on synthetic tabular data
- balance imbalanced tabular datasets with synthetic samples

## When to choose
- you need synthetic single-table data with realistic column distributions and correlations
- you want a Python API for CTGAN or TVAE models
- you need to generate data for testing, demos, or ML augmentation

## When to avoid
- you need synthetic data for relational or multi-table schemas - use the full SDV library instead
- you need image, text, or time-series generation
- you require production-grade stability - the library is labeled Pre-Alpha

## Facets
- artifact type: library
- maturity: active
- function: machine-learning, deep-learning, data-generation
- domain: machine-learning, data-science, artificial-intelligence
- platform: python
- tags: synthetic-data, gan, tabular-data, tvae, generative-models

## Member repositories
- sdv-dev/CTGAN (main) score 80

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:05:03.944390+00:00.
- Health v2: computed from the inputs above; adoption is never an input.
- Inferred fields (summary, facets, guidance): AI-extracted, prompt v1, taxonomy v1, on 2026-08-30T04:00:19.101533+00:00, confidence not recorded.
  - readme: https://github.com/sdv-dev/CTGAN (fetched 2026-08-28T04:05:03.944390+00:00, sha 56d2a72c6943)
  - registry_pypi: https://pypi.org/pypi/ctgan/json (fetched 2026-08-29T11:29:18.829958+00:00, sha d009e5ce8f31)
- Data as of 2026-08-30T08:39:29.467469+00:00.