Changelog

Changelog#

Version 0.2.0#

Breaking Changes#

  • Dropped Python 3.8 support; now requires Python >=3.9 (#148)

  • Bumped torch dependency to >=2.8 (#148)

  • Replaced validate_data.py with pydantic-based validators (pydantic>=2.0 now required) (#148)

Bug Fixes#

  • Fixed IPCW device mismatch issue (#147)

Core Refactoring#

  • Refactored all loss modules (cox, weibull, momentum, survival) to use pydantic validators (#148)

  • Refactored all metric modules (auc, brier_score, cindex) to use new validators (#148)

  • Vectorized Cox partial likelihood and cumulative hazard computations (#148)

  • Added __init__.py with __all__ exports to all subpackages (#148)

Development Tooling#

  • Migrated from conda to uv as primary development tool (#148)

  • Added dependency-groups in pyproject.toml (dev, docs, publish) (#148)

  • Added ruff + mypy strict mode to pre-commit and CI (#148)

  • Modernized CI workflows to use uv + astral-sh/setup-uv@v5 (#148)

  • Upgraded workflow tests with better Python/PyTorch version coverage (#158)

Testing#

  • Added hypothesis-based property tests (#148)

  • Added shared test fixtures via conftest.py (#148)

  • Added pydantic validation tests (#148)

Documentation#

  • Added GOVERNANCE.md, CODEOWNERS, and release info for PyTorch Ecosystem compliance (#156)

  • Updated README with versioning, compatibility, and changelog links (#156)

  • Fixed URL typos and broken links (#150)

  • Added package overview docs and non-medical applications notebook (#148)

Version 0.1.6#

  • Fixed bugs (#118, #119, #137),

  • Added Full Disclaimer and Link to RST page for TorchSurv (#125)

  • Cox Proportional Hazards model:

    • Added Breslow’s estimator for the survival function (#127)

    • Added support for stratified models (#128)

  • Added survival model parameterized by log-hazard, with loss computed using the trapezoidal rule (#138)

  • Added support for time-dependent covariates (#138, #140)

Version 0.1.5#

  • Fixed bugs (#94), (#97)

  • Improved codebase to match Torch.jit.script (#67)

  • Added precommit (#105)

  • Fixed logo issue (#95)

Version 0.1.4#

  • JOSS review edits (#45, #59, #66, #68, #77, #78)

  • Torch.compile tests #81

  • Improved documentation (#73, #76)

  • Improved notebook #83

Version 0.1.3#

  • Tutorial dataset error on momentum.ipynb #50

  • Fix issue #48 - log_hazard returns torch.Inf

  • Fix warning with Spearman correlation #41

  • Added in-depth statistical background to link AUC to C-index #39

  • Created Conda Forge version #47

  • Updated CICD builds #53

Version 0.1.2#

  • Updated package documentation with publication links & badges (#9, #14, #16, #19, #21, #22, #24)

  • Fixed and documented package dependencies (#1)

Version 0.1.1#

  • Added metrics classes (AUC, Cindex, Brier score)

  • Added stats class with Kaplan Meier

  • Created Sphinx documentation

  • Created notebook examples

  • Added R benchmark comparison

Version 0.1.0#

  • Initial release of CoxPH and Weibull classes.