Changelog¶
All notable changes to circumplex will be documented in this page and in the
repository CHANGELOG.md.
Unreleased¶
Massive update
This is the largest update to circumplex so far and forms the basis of the upcoming 0.3 release. A major goal of this work has been to target feature parity and a familiar API for users of the fantastic R circumplex package. The implementation is regression-tested against the R package to ensure equivalent results, and the Python package has been substantially reorganized, expanded, and hardened for public use.
Added¶
- A modular package layout with dedicated
analysis,core,data,instruments,utils, andvisualizationmodules. - Mean-based and correlation-based structural summary workflows with bootstrap-backed inference and richer plotting helpers.
- Built-in instrument definitions, instrument registration utilities, and tidying helpers for ipsatization, scoring, and normative standardization.
- Expanded API reference pages, tutorial content, examples, and regression fixtures to improve discoverability and validate parity with the reference R package.
- Added regression coverage against the R implementation to ensure the Python package produces equivalent results for supported structural summary workflows.
Changed¶
- Reoriented the project around feature parity and a similar user experience to the R circumplex package, so core workflows feel familiar to users moving between the two implementations.
- Refactored the public package around a clearer, more maintainable architecture for analysis, data loading, scoring, and plotting.
- Updated project infrastructure around
uv,tox,zensical, and GitHub Actions to give the project a cleaner release, documentation, and CI story. - Refreshed the README, documentation landing page, and documentation navigation to reflect the new package structure and upcoming release scope.
Fixed¶
- Aligned package metadata, licensing, docs configuration, and release workflows so published artifacts and hosted documentation follow the same project configuration.