✨✨✨ Megacomplex based models, full models (spectrotemporal), damped oscillations models (DOAS), 🧪 test result validation ✔️ #778
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
|
|
Codecov Report
@@ Coverage Diff @@
## main #778 +/- ##
=======================================
+ Coverage 81.7% 84.7% +2.9%
=======================================
Files 79 77 -2
Lines 4024 4343 +319
Branches 706 785 +79
=======================================
+ Hits 3291 3682 +391
+ Misses 606 521 -85
- Partials 127 140 +13
Continue to review full report at Codecov.
|
jsnel
added
Status: In Progress
This comment has been minimized.
This comment has been minimized.
…otaran.model (#734) * Changed calculate_matrix function to return xarrays * Added relation and constraint to basemodel * Added functions to apply relations and constraints * Removed LabelAndMatrix class * Added relations and constraints tests to base model * Added penalties to base model * Adapted models to changes * Address xarray deprecation warnings Fixed DeprecationWarning: Using a DataArray object to construct a variable is ambiguous, please extract the data using the .data property. This will raise a TypeError in 0.19.0. Co-authored-by: Jörn Weißenborn <[email protected]> Co-authored-by: Joris Snellenburg <[email protected]>
* 🔧 Allowed mypy to run on a subset of glotaran modules that will be checked are: - utils - plugin_system - deprecation * 🔧 Partially activated docstring QA tools and fixed issues modules that will be checked are: - utils - plugin_system - deprecation
This PR removes models like kinetic-image in favor of independent megacomplex model. Therefor, there are a lot of internal changes. Key Changes for users: * use `default-megacomplex` instead of `model-type` in yml spec * simulations can now be done by setting `global_megacomplex` in a dataset model * heterogenous dataset models can be analyzed in one optimization, e.g. a spectral and a temporal dataset * the user now needs to set `group=true` in a scheme to have a grouped analysis.
* Added benchmark for Problem class * removed print * ♻️ 'Refactored by Sourcery' * 🧹 Moved glotaran/analysis/test/test_relations.py to benchmark/pytest/analysis/test_problem.py Use 'pytest benchmark/pytest/' to run the benchmarks * Numerous performance tweaks * Don't weight data with ones if no weight supplied * Fic duplicate call to create result_dataset * Removed dead code * Switched back to pure numpy in problem * Fixed example * Cleanunp * Update glotaran/analysis/problem_ungrouped.py Co-authored-by: Sebastian Weigand <[email protected]> * Update glotaran/analysis/problem_ungrouped.py Co-authored-by: Sebastian Weigand <[email protected]> * Update glotaran/analysis/util.py Co-authored-by: Sebastian Weigand <[email protected]> Co-authored-by: Sourcery AI <> Co-authored-by: s-weigand <[email protected]>
* Added Full Model functionality (also known as spectrotemporal models) * Added finalizing functions for full models and changed spectral decay test to full model test * Added flake8-print to precommit * Fixed parameter history and set correct optizemet parameter prior to resultdata creation * Don't apply clp relations and constraints on full models * Added auto inference of grouping * Update glotaran/test/test_spectral_decay_full_model.py * Correction of cost calculation Co-authored-by: Joris Snellenburg <[email protected]> Co-authored-by: Sebastian Weigand <[email protected]>
* Small fix for baseline megacomplex * Save baseline as name 'baseline'
* Added SpectralShapeGaussian. * Replaced property energy_spectrum of SpectralMegacomplex with invert and axis_scale. * Made invert and axis_scale dataset properties. * Added guard and meaningful exception message for spectral skewness * Fixed scaling * Made SpectralShapeSkewedGaussian decendent of SpectralShapeGaussian and added fallback for skewness == 0. * Move sanatize.py to utils module and correct typo sanatize -> sanitize (incl rename to sanitize.py) * Move regex patterns to seperate module in utils * Add sanity_scientific_notation_conversion Convert scientific notation string (e.g. 1E7) to proper floats * Fixed spectral megacomplex test parameters and added test for inverted axis * Removed model_item._from_list * Added convenience in model_item.from_dict for automatically convert float or int typed properties which are parsed as strings. * Made amplitude of shape optional Co-authored-by: Joris Snellenburg <[email protected]>
Various fixes and improvements to the glotaran command line interface. * Changed CLI save plugin to folder * Added outputformat option to CLI * Added basic test for CLI * Rename CLI entrypoint to main and add more CLI tests * 👌 CLI use same default for non_negative_least_squares as scheme * 👌 CLI dedent pluginlist output * 🩹 CLI fixed result outformat accepting none supported formats Co-authored-by: Joris Snellenburg <[email protected]> Co-authored-by: s-weigand <[email protected]>
* ✨ Added 'deprecate_dict_entry' to deprecate dict keys and/or values * 🧹📚 Changed Warns OverDueDeprecation to Raises and added missing Raises OverDueDeprecation to deprecate_module_attribute and deprecate_submodule * 🩹🧪 Reimplemented check_deprecations and with deprecate_dict_entry and renamed it to 'model_spec_deprecations' This change also ensures that the deprecation warning is thrown when users calls 'load_model' so python will show the warning * ♻️🧪 Changed deprecation_warning_on_call_test_helper to test for file emitting the warning and return the WarningsRecorder * 📚 Added docs for deprecate_dict_entry * 🧹 Removed unused record variables * 🧪 Added test for deprecated 'spectral_constraints' * 👌 Changed DeprecationWarning to GlotaranApiDeprecationWarning (GlotaranApiDeprecationWarning is a subclass of UserWarning so it won't be filtered automatically by python) * 📚 Updated docs to reflect the usage of GlotaranApiDeprecationWarning
* Added DampedOscillationMegacomplex * Added damped-oscillation to setup.cfg plugins Co-authored-by: Joris Snellenburg <[email protected]>
* 🩹 🔧 Only run interrogate on glotaran folder * 👌 🚧 Raised interrogate threshold for 52% to 55% (current value 56.1%)
* 🚇 Only run base compare commit and last HEAD commit on PR benchmark * 👌 Update comment w/o deleting the last one (comment should have an 'edit' history)
…790) The deprecated equal_area_penalties would be called in lieu of the new clp_area_penalties function The pyglotaran-examples have in their model still equal_area_penalties which is correctly swapped with clp_area_penalties but them model.equal_area_penalties was still called in some places instead of model.clp_area_penalties
|
This pull request fixes 3 alerts when merging dd0ce80 into 482a53b - view on LGTM.com fixed alerts:
|
Avoiding the use of set to improve ordering Refactor at least guarantees better than random ordering using list(set(x)) TODO: further improve ordering to use initial_concentrations ordering
|
This pull request fixes 3 alerts when merging a6a5a51 into 482a53b - view on LGTM.com fixed alerts:
|
* Fix and re-enable IRF Dispersion Test Improve the tests in test_spectral_irf Added NoDispersion test Re-enable IrfDispersion tests on the CI (de-Volkswagen the tests) Improve assignment of center_dispersion to result dataset * Rename center_dispersion to irf_center_location in result The name irf_center_location in the results describes more clearly what the variable holds. Deprecated irf model item center_dispersion for center_dispersion_coefficients Deprecated irf model item width_dispersion for width_dispersion_coefficients
jsnel
changed the title
|
It's time to merge staging into main. Note to 📚 reviewers 📚: Please note that each individual PR merged into the staging branch had already been review separately, so there's no need to review this PR in its entirety. If you find issues create a new issues for it to be picked up after merging this PR into main, before preparing the v0.5.0 release. Note to 🚧 maintainers 🚧: This PR, because of it's size, will be merged (by 🚧 @jsnel 🚧 !!! ) as a rebase commit, rather than a squash and merge, in order to retain some of the historical value provided by the individual commits - and keep all the extensive commit messages. Note to supervisor (@ism200) : This is the big one you have been waiting for. |
This comment has been minimized.
This comment has been minimized.
4 tasks
joernweissenborn
approved these changes
Sep 16, 2021
|
🚀 Final score ... +7,897 −5,366 😱 Thanks for the tremendous effort @joernweissenborn and @s-weigand! |
This was referenced Sep 16, 2021
1 task
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
Megacomplex composition, spectrotempraol (full) models and fitting damped oscillations are now a thing!
This (rather massive) PR bring in some amazing changes - detailed below, consistency checking and significant speed improvements!
This PR makes that pyglotaran can finally be called 'feature complete' and sets the stage for a v0.5.0 release.
Change summary
Major new features
Other major changes
Minor changes:
Checklist
Closes issues
closes #777
(and many issues linked to in the above referenced PRs)