Setup of pipeline integration testing up to modeling

[HealthyPear]

Requirements

• Merge PR Small improvements to modeling script #114
• Update documentation

Description

This PR refactors integration tests in a single module called test_pipeline.py where,

• the whole pipeline will be tested on standard settings starting from diffuse test files stored in the ccin2p3 dataserver
• integration tests follow the pipeline workflow
• they produce their own output files and are skipped if a previous test fails to produce the expected output
• current tests cover training and modeling
• all tests are performed on a CTAN and CTAS Prod3b baseline analysis using all available cameras
• all tests share a temporary directory where all produced files have unique suffixes defined by their test case

[codecov]

Codecov Report

Merging #116 (bf85742) into master (e87557c) will increase coverage by 7.69%.
The diff coverage is 100.00%.

@@            Coverage Diff             @@
##           master     #116      +/-   ##
==========================================
+ Coverage   38.61%   46.30%   +7.69%     
==========================================
  Files          22       22              
  Lines        1950     1978      +28     
==========================================
+ Hits          753      916     +163     
+ Misses       1197     1062     -135     

Impacted Files	Coverage Δ
protopipe/scripts/tests/test_pipeline.py	100.00% <100.00%> (ø)
protopipe/pipeline/utils.py	54.54% <0.00%> (+3.63%)	⬆️
protopipe/pipeline/event_preparer.py	76.62% <0.00%> (+7.27%)	⬆️
protopipe/mva/utils.py	30.43% <0.00%> (+18.11%)	⬆️
protopipe/mva/train_model.py	60.78% <0.00%> (+45.09%)	⬆️
protopipe/scripts/build_model.py	61.29% <0.00%> (+50.00%)	⬆️

---

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 7632ef2...bf85742. Read the comment docs.

[HealthyPear]

@kosack did you notice PR #114 ? I can't merge this one before because I am using some of the new flag options in the model script