Refactor model.from_dict to parse megacomplex_type from dict and add simple_generator for testing

[jsnel]

One of the issues with current tests is that they are hard to read because of large in-line dicts with minimal differences between them, occasionally resulting in the introduction of errors (such as defining a parallel model as a sequential model, illustrated by this commit fixing that error: dce7f0b)

To fix this we need:

• a simple model + parameters generator
• Model.from_dict to pickup its type from dict instead of only through passing in kwargs

This PR introduces those changes, and then illustrates the use of the simple generator by using it to refactor a single class (ThreeComponentParallel) used in tests.

This PR opens up the possibility for further refactoring tests for improved readability, to come in a future PR (see #815).

Co-authored-by: Sebastian Weigand [email protected]

Change summary

• ✨🧪 Added plugin registry monkeypatch context managers for testing
• ✨🧪 Added new simple_generator for generating model and parameters using a minimal specification (for testing)
• ♻️ Refactored ThreeComponentParallel in test_decay_megacomplex.py as an example of the new features

Checklist

• ✔️ Passing the tests (mandatory for all PR's)
• 👌 Closes issue (mandatory for ✨ feature and 🩹 bug fix PR's)
• 🧪 Adds new tests for the feature (mandatory for ✨ feature and 🩹 bug fix PR's)
• 📚 Adds documentation of the feature

Closes issues

closes #816

[github-actions]

👈 Launch a binder notebook on branch jsnel/pyglotaran/refactor/megacomplex_types

[codecov]

Codecov Report

Merging #807 (d6a31e6) into staging (0f36985) will increase coverage by 0.2%.
The diff coverage is 91.2%.

@@            Coverage Diff            @@
##           staging    #807     +/-   ##
=========================================
+ Coverage     84.5%   84.7%   +0.2%     
=========================================
  Files           75      77      +2     
  Lines         4201    4343    +142     
  Branches       757     785     +28     
=========================================
+ Hits          3553    3682    +129     
- Misses         515     521      +6     
- Partials       133     140      +7     

Impacted Files	Coverage Δ
glotaran/testing/model_generators.py	88.6% <88.6%> (ø)
glotaran/model/model.py	85.7% <88.8%> (+0.1%)	⬆️
glotaran/builtin/io/yml/yml.py	79.6% <100.0%> (-0.8%)	⬇️
glotaran/testing/plugin_system.py	100.0% <100.0%> (ø)

---

Continue to review full report at Codecov.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update 0f36985...d6a31e6. Read the comment docs.

[jsnel]

This PR opens up the possibility for further refactoring the tests, for instance the other tests in test_decay_megacomplex.py could be refacorted along the same lines:

• OneComponentOneChannel
• ThreeComponentSequential
  And perhaps:
• OneComponentOneChannelGaussianIrf
  (this may requires some further changes to the simple generators)

This could be an easy issue for a new contributor (@s-weigand could you create a new issue for this with some details). This could then be picked up after merging this PR into staging (@joernweissenborn).

[s-weigand]

Accidentally the pre-commit check for darglint and pydocstyle were "VWed" out for the testing module 😱
I fixed it in 8b4d3bf, sorry the extra noise.

[s-weigand]

Just some minor nitpicky things. 😅

Also, could we rename simple_generator.py into something like model_generators.py, if we have the need to create more we can make it a sub_package?
SimpleGenerator sounds like we will need more complex ones.

[sonarcloud]

Kudos, SonarCloud Quality Gate passed!   

0 Bugs
0 Vulnerabilities
0 Security Hotspots
11 Code Smells

No Coverage information
0.6% Duplication

[github-actions]

Benchmark is done. Checkout the benchmark result page.
Benchmark differences below 5% might be due to CI noise.

Benchmark diff

Parametrized benchmark signatures:

BenchmarkOptimize.time_optimize(index_dependent, grouped, weight)

All benchmarks:

       before           after         ratio
     [dc00e6da]       [d6a31e66]
     <v0.4.0>                   
-      49.8±0.6ms       36.5±0.4ms     0.73  BenchmarkOptimize.time_optimize(False, False, False)
-       278±0.5ms         43.5±1ms     0.16  BenchmarkOptimize.time_optimize(False, False, True)
-      73.0±0.6ms       61.1±0.4ms     0.84  BenchmarkOptimize.time_optimize(False, True, False)
         75.8±1ms       66.4±0.9ms    ~0.87  BenchmarkOptimize.time_optimize(False, True, True)
       49.7±0.2ms       47.3±0.5ms     0.95  BenchmarkOptimize.time_optimize(True, False, False)
-         278±1ms         56.6±2ms     0.20  BenchmarkOptimize.time_optimize(True, False, True)
       72.9±0.6ms         72.9±1ms     1.00  BenchmarkOptimize.time_optimize(True, True, False)
         77.1±4ms        79.8±50ms     1.04  BenchmarkOptimize.time_optimize(True, True, True)
             180M             180M     1.00  IntegrationTwoDatasets.peakmem_create_result
             198M             197M     1.00  IntegrationTwoDatasets.peakmem_optimize
-         232±3ms          190±3ms     0.82  IntegrationTwoDatasets.time_create_result
       5.02±0.08s       1.84±0.09s    ~0.37  IntegrationTwoDatasets.time_optimize

Implemented requested changes: add docstrings to semi-private methods, rename simple_generator to model_generators

[sourcery-ai]

Sourcery Code Quality Report

✅  Merging this PR will increase code quality in the affected files by 0.28%.

Quality metrics	Before	After	Change
Complexity	7.24 ⭐	7.15 ⭐	-0.09 👍
Method Length	101.71 🙂	96.29 🙂	-5.42 👍
Working memory	11.28 😞	11.42 😞	0.14 👎
Quality	52.46% 🙂	52.74% 🙂	0.28% 👍

Other metrics	Before	After	Change
Lines	206	194	-12

Changed files	Quality Before	Quality After	Quality Change
glotaran/builtin/io/yml/yml.py	52.46% 🙂	52.74% 🙂	0.28% 👍

Here are some functions in these files that still need a tune-up:

File	Function	Complexity	Length	Working Memory	Quality	Recommendation
glotaran/builtin/io/yml/yml.py	YmlProjectIo.load_scheme	13 🙂	257 ⛔	15 😞	34.68% 😞	Try splitting into smaller methods. Extract out complex expressions
glotaran/builtin/io/yml/yml.py	YmlProjectIo.save_result	3 ⭐	214 ⛔	11 😞	52.52% 🙂	Try splitting into smaller methods. Extract out complex expressions

Legend and Explanation

The emojis denote the absolute quality of the code:

• ⭐ excellent
• 🙂 good
• 😞 poor
• ⛔ very poor

The 👍 and 👎 indicate whether the quality has improved or gotten worse with this pull request.

---

Please see our documentation here for details on how these metrics are calculated.

We are actively working on this report - lots more documentation and extra metrics to come!

Help us improve this quality report!