synthesizer: Fast, Robust, and High-Quality Synthetic Data Generation with a Tuneable Privacy-Utility Trade-Off

Synthesize numeric, categorical, mixed and time series data. Data circumstances including mixed (or zero-inflated) distributions and missing data patterns are reproduced in the synthetic data. A single parameter allows balancing between high-quality synthetic data that represents correlations of the original data and lower quality but more privacy safe synthetic data without correlations. Tuning can be done per variable or for the whole dataset.

Version:	0.5.0
Depends:	R (≥ 3.5.0)
Imports:	stats
Suggests:	tinytest, simplermarkdown
Published:	2025-07-10
DOI:	10.32614/CRAN.package.synthesizer
Author:	Mark van der Loo [aut, cre]
Maintainer:	Mark van der Loo <mark.vanderloo at gmail.com>
License:	EUPL version 1.1 | EUPL version 1.2 [expanded from: EUPL]
URL:	https://github.com/markvanderloo/synthesizer
NeedsCompilation:	no
Materials:	NEWS
CRAN checks:	synthesizer results

Documentation:

Reference manual:	synthesizer.pdf
Vignettes:	Introduction to synthesizer (source, R code)

Downloads:

Package source:	synthesizer_0.5.0.tar.gz
Windows binaries:	r-devel: synthesizer_0.4.0.zip, r-release: synthesizer_0.5.0.zip, r-oldrel: synthesizer_0.4.0.zip
macOS binaries:	r-release (arm64): synthesizer_0.5.0.tgz, r-oldrel (arm64): synthesizer_0.5.0.tgz, r-release (x86_64): synthesizer_0.5.0.tgz, r-oldrel (x86_64): synthesizer_0.5.0.tgz
Old sources:	synthesizer archive

Linking:

Please use the canonical form https://CRAN.R-project.org/package=synthesizer to link to this page.