Version 2.1.0

New Features

• Multiple table support. Metasyn now supports the ability to fit and synthesize multiple related tables at once.
• Multiple bug fixes

What's Changed

• Give more useful errors for Fitters and Distributions by @qubixes in #412
• add hospital and druguse demo dataframes to documentation
• Remove some sources of None by @qubixes in #414
• Update OS version to Ubuntu 24.04 for readthedocs by @qubixes in #424
• Full api reference by @MSVermet in #423
• Remove exception for documentation in GenAI use by @qubixes in #426
• Fix truncated normal by @qubixes in #428
• Improve the default discrete normal distributions by @qubixes in #425
• Improve the performance of read/write stat by @qubixes in #427
• Multiple table/dataframe synthesis by @qubixes in #415
• Bump version to 2.1.0 by @qubixes in #429

New Contributors

• @MSVermet made their first contribution in #423

Full Changelog: v2.0.0...v2.1.0

Version 2.0.0

New features

• 🗒️Metasyn now supports reading and writing files: generate synthetic data from different formats (e.g., .dta, .sav, .csv) and write it in those same formats, in the exact same way
• 🕦 Metasyn now gives nice progress bars when it is fitting or synthesizing data
• 🗄️Metasyn now supports larger-than-memory data through chunked reading: fit and synthesize big data!
• 💡We overhauled the whole back-end to separate the concept of a Distribution from the concept of a Fitter (which learns the parameters of a distribution). This makes building plugins easier!
• ⬇️ Many other small improvements and bug fixes, see below.

What's Changed

• rebase develop on main by @vankesteren in #345
• update dockerfile for dependency management by @vankesteren in #349
• Fix numpy dependency, move to stable, add 3.13 ci by @vankesteren in #348
• Make the vartype determination easier by @qubixes in #350
• Set the default prop missing for data free to 0 by @qubixes in #346
• Improve error messaging for variable fitting by @qubixes in #347
• Split varspec param by @qubixes in #351
• Create a new option to output the schema to a file by @qubixes in #352
• Allow the synthesized output to be seeded by @qubixes in #353
• Coverage by @qubixes in #354
• Add a parameter to skip checks for empty series by @qubixes in #356
• Add rudimentary file handling support. by @qubixes in #355
• Add excel reader and add conversion options by @qubixes in #365
• Add Stata reader by @qubixes in #368
• Add option to set encoding by @qubixes in #372
• Check the filename before synthesizing by @qubixes in #373
• Add nested progress bar to synthesize() by @vankesteren in #367
• Convert file reader to interface by @qubixes in #374
• Add ability to steer reading of files in configfile by @qubixes in #375
• Add distribution module to exports for type completion by @vankesteren in #379
• Rework privacy a little bit by @qubixes in #378
• Partially rework structure by @qubixes in #381
• Add variable type argument for default distribution by @qubixes in #389
• Fix typing for unique key distribution by @qubixes in #390
• Check whether we can remove pandas by @qubixes in #393
• Add citation section to readme by @vankesteren in #391
• Modernize the dockerfile by @vankesteren in #397
• Improve error message when a plugin fails to load by @qubixes in #398
• Add feature to use chunking for pyreadstat by @qubixes in #399
• Update documentation for version 2.0 by @vankesteren in #403
• Improved error messages for fitter and distribution by @qubixes in #400
• Rename distribution registries to plugins by @qubixes in #405
• Improve the provenance information in the GMF by @qubixes in #406
• Bump version to 2.0.0 by @vankesteren in #407
• Release 2.0.0 by @vankesteren in #408
• Set specific version of setuptools for pypi by @vankesteren in #410

Full Changelog: v1.1.0...v2.0.0

Version 1.1.0

This release improves documentation and the developer experience, adds new built-in datasets, adds TOML export functionality (in addition to JSON), and improves warnings

What's Changed

• Move polars to version 1, fix test errors, rerun examples by @vankesteren in #299
• Set information criterion for NA distribution high by @qubixes in #298
• Convert int columns in spaceship data to ints by @vankesteren in #300
• Fix with branch switching by @qubixes in #301
• Adding the paper to the docs folder for submission to JOSS by @vankesteren in #302
• Add JOSS submission status badge to readme by @vankesteren in #305
• Shorten JOSS paper by @vankesteren in #306
• Fix broken link and link to README by @qubixes in #308
• One line for readthedocs and tutorials next steps by @qubixes in #310
• Add survey demo dataset by @vankesteren in #313
• Add gif to readme by @vankesteren in #304
• Update README.md by @qubixes in #318
• Simplify CI & development experience by @vankesteren in #316
• Improved demo datasets by @qubixes in #325
• Add new dataset and make regex more used by @qubixes in #330
• Change the representation of MetaVar/MetaFrame by @qubixes in #331
• Add export functionality to TOML in addition to JSON by @qubixes in #327
• Improve unique warning by @qubixes in #332
• Improve the toml comments by @qubixes in #333
• Change utctimestamp -> timestamp by @qubixes in #336
• Improve test documentation by @qubixes in #335
• Run core ci on push by @vankesteren in #341
• Simplified paper description and added relevant examples by @vankesteren in #337
• Documentation update 1.1 by @qubixes in #340
• Remove stray print statement and add comment. by @qubixes in #343

Full Changelog: v1.0.1...v1.1.0

Version 1.0.3

This release fixes a bug with categorical data due to a change in polars, and it updates some minor things in the readme and the documentation.

What's Changed

• Add JOSS submission status badge to readme by @vankesteren in #305
• Shorten JOSS paper by @vankesteren in #306
• Fix broken link and link to README by @qubixes in #308
• One line for readthedocs and tutorials next steps by @qubixes in #310
• Add survey demo dataset by @vankesteren in #313
• Add gif to readme by @vankesteren in #304
• Update README.md by @qubixes in #318
• Release 1.0.3 by @vankesteren in #320

Full Changelog: v1.0.2...v1.0.3

Version 1.0.2

This is a minor bugfix release before submission to the Journal of Open Source Software

What's Changed

• Move polars to version 1, fix test errors, rerun examples by @vankesteren in #299
• Set information criterion for NA distribution high by @qubixes in #298
• Convert int columns in spaceship data to ints by @vankesteren in #300
• Fix with branch switching by @qubixes in #301
• Adding the paper to the docs folder for submission to JOSS by @vankesteren in #302
• Release 1.0.2 by @vankesteren in #303

Full Changelog: v1.0.1...v1.0.2

Version 1.0.1

This release mainly fixes compatibility with the new numpy and polars versions.

What's Changed

• Fix defaults so plugins don't need manual fitkwarg by @qubixes in #296

Full Changelog: v1.0.0...v1.0.1

Version 1.0.0

The first stable release of metasyn!

Highlights

• Lot of documentation updates, thanks mainly to @Samuwhale.
• More descriptive error handling.
• Add a way to find plugins that are needed using a plugin registry.

What's Changed

• Plugin registry by @qubixes in #286
• Add direct links to readme images for them to work in pypi by @Samuwhale in #288
• Better errors by @Samuwhale in #283
• Code of conduct by @vankesteren in #289
• Update repo status to active by @vankesteren in #290
• Remodel API docs with inspiration from Scipy by @qubixes in #292
• Review of the documentation by @qubixes in #293
• Ignore Mypy error by @qubixes in #294
• Docs improvements based on doc review by @Samuwhale in #295

Full Changelog: v0.8.0...v1.0.0

Version 0.8.0

What's Changed

• Remove notebook checking for osx by @qubixes in #267
• Further update readme by @vankesteren in #264
• fix basic example by @Samuwhale in #268
• Add Zenodo DOI to readme by @vankesteren in #271
• Update polars version by @vankesteren in #273
• Revarspec by @qubixes in #275
• Fixes an issue with distspec and privacy by @qubixes in #276
• Add creation settings by @qubixes in #272
• Fix leftovers of the VarConfig era by @qubixes in #277
• Fix issue of privacy when user provides distrib by @qubixes in #280

Full Changelog: v0.7.1...v0.8.0

v0.7.1

A small release with mainly bug fixes and improvements to the command line interface.

What's Changed

• Add Python 3.12 as possible option for install by @qubixes in #248
• Fix categorical type annotation by @vankesteren in #254
• Improve readme by @Samuwhale in #247
• Fix to_dict method of BasePrivacy by @qubixes in #259
• Demo datasets by @vankesteren in #256
• Improved cli by @qubixes in #262
• Fix boolean types to be supported as categorical by @qubixes in #260
• Remove general toml by @qubixes in #261

Full Changelog: v0.7.0...v0.7.1

v0.7.0

We are preparing to release version 1.0 soon. In the mean time, this version (0.7.0) is a preparation for that. This version is not considered a super stable version, so you can also wait until 1.0 for that when the interface should stabilize.

What's Changed

• Added CLI documentation by @Samuwhale in #183
• Added section on the name change in the FAQ by @Samuwhale in #181
• Progress bar by @qubixes in #188
• Added documentation on printing and previewing MetaFrames / GMFs by @Samuwhale in #195
• Add create-meta subcommand by @qubixes in #192
• Improved CLI docs + added create-meta documentation by @Samuwhale in #200
• Cleaned and improved examples by @Samuwhale in #193
• Fix deprecation warning by @qubixes in #201
• Ruff + Tox integration by @Samuwhale in #199
• Python req updated from 3.7 to 3.8 on docs by @Samuwhale in #205
• Updated version number on the docs by @Samuwhale in #207
• Make small changes to the documentation by @qubixes in #209
• Improve CLI tests by @qubixes in #202
• Fix for new version of lingua by @qubixes in #210
• Allow automatic changes with strings by @qubixes in #211
• Change from AIC to BIC by @qubixes in #212
• Discrete normal by @vankesteren in #216
• Change AIC to BIC by @qubixes in #217
• Change it so that dt distribution inherit from base by @qubixes in #221
• Constant distributions by @vankesteren in #219
• Add docker version badge to readme by @vankesteren in #225
• Improved API and Developer guide by @Samuwhale in #220
• Improved module/package docstrings by @Samuwhale in #226
• Add CI action to push releases to dockerhub by @vankesteren in #231
• Refactor metasyn and make CLI more feature complete by @qubixes in #227
• Update LICENSE by @Samuwhale in #236
• Homogenize distribution parameters by @vankesteren in #239
• Remove legacy code and explicit pandas dependency by @qubixes in #242
• Add links to plugin templates to docs by @Samuwhale in #235
• Change consecutive parameter in unique_key to bool by @qubixes in #244
• Fix a small cli doc error by @vankesteren in #246
• Harmonize is_unique to unique by @qubixes in #245
• Improved docstrings at a class level by @Samuwhale in #234

Full Changelog: v0.6.0...v0.7.0