🚀v0.4.1 and v0.4.2

[jsnel]

pyglotaran 0.4.2 Release Notes

Code name: Continue "GTA Save Game"

The v0.4.2 is a minor patch for the v0.4.1 maintenance release which first established a result comparison framework (#760) to benchmark future releases against. This comparison framework is implemented as a github action workflow which can be run on any PR or release branch. The comparison involves running a set of example scripts hosted in the pyglotaran-examples repository using the code in the PR or release-branch and comparing the output against a 'gold standard' reference - initially generated using the v0.4 code base which was manually checked and validated against other sources. This framework guards pyglotaran against future regression, both in terms of performance and accuracy, going forward and gives users the confidence to migrate from one version to the next.

As part of the release v0.4.1 was tested against 'itself' (based on v0.4.0) which revealed some minor consistency bugs (#747 and #791) which were then simultaneously solved for the next release and 'backported' to this version. Further development revealed a minor bug in saving result files (not in the optimization process) which is patched in the v0.4.2 release (#935). In addition the v0.4.2 maintenance branch bring some internal improvement to the test_result_consistency #936, further validating the checking process.

As noted before in the v0.4.1 release, the output (estimates, plots, etc) of example scripts were validated by manual inspection and comparing against pre-existing analysis software packages:

• TIMP (the analysis core of Glotaran v1.5.1) as described in "TIMP: An R Package for Modeling Multi-way Spectroscopic Measurements" , DOI: 10.18637/jss.v018.i03
  • Glotaran v1.5.1 as described in the publication "Glotaran: A Java-Based Graphical User Interface for the R Package TIMP", DOI: 10.18637/jss.v049.i03
• paramGUI used for teaching global analysis to natural science students at the VU for many years.
• TIM as described in the reference work on "Global and target analysis of time-resolved spectra", DOI: 10.1016/j.bbabio.2004.04.011

✨ Features

• Integration test result validation (🧪🚇 Add integration test result validation (maintain/0.4.1) #760) (introduced in v0.4.1, updated in v0.4.2).

🩹 Bug fixes

New in v0.4.2

• 🩹🚧 Backport of bugfix 🩹 Bugfix Use normalized initial_concentrations in result creation for decay megacomplex #927 discovered in PR ✨ Add simple decay megacomplexes #860 related to initial_concentration normalization when saving results (🚀🩹 v0.4.2 release preparation and bugfix backport for result consistency checking purposes #935).

Part of v0.4.1

• Fix unintended saving of sub-optimal parameters (0ece818, backport from Feature: Full Models #747).
• Improve ordering in k_matrix involved_compartments function ([Backport] Improve ordering in k_matrix involved_compartments function #791).

🚧 Maintenance

New in v0.4.2:

• 🚇🚧 Updated 'gold standard' result comparison reference (old -> new).
• 🚇 Refine test_result_consistency (🚇Update test_result_consistency #936).

[jsnel]

Added v0.4.2 release notes to the discussion. Since v0.4.2 only brings a small bug fix and no new features, we can re-use this discussion thread. The bug fix was also forward ported to v0.5.1 discussion here.

Also 👋 bye 👋 2021, hello ✨2022✨.

Feels like a good 👌 year to 🚀 release 🚀 1.0 !
-- Joris ( @jsnel )