Add handling for failed simulations

[radka-j]

Closes #438

[codecov-commenter]

Codecov Report

Attention: Patch coverage is 95.12195% with 2 lines in your changes missing coverage. Please review.

Project coverage is 80.20%. Comparing base (275a06f) to head (395ee0b).
Report is 133 commits behind head on main.

Files with missing lines	Patch %	Lines
autoemulate/experimental/simulations/base.py	90.00%	2 Missing ⚠️

Additional details and impacted files

@@            Coverage Diff             @@
##             main     #643      +/-   ##
==========================================
+ Coverage   78.30%   80.20%   +1.89%     
==========================================
  Files         150      155       +5     
  Lines       10839    11132     +293     
==========================================
+ Hits         8488     8928     +440     
+ Misses       2351     2204     -147     

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.

[github-actions]

Coverage report

Click to see where and how coverage changed

File	Statements	Missing	Coverage	Coverage (new stmts)	Lines missing
autoemulate/experimental/calibration
history_matching.py
autoemulate/experimental/data
preprocessors.py
autoemulate/experimental/emulators
base.py
autoemulate/experimental/emulators/gaussian_process
__init__.py
autoemulate/experimental/learners
base.py
autoemulate/experimental/simulations
base.py					107, 177-181
autoemulate/experimental/transforms
base.py
autoemulate/simulations
naghavi_cardiac_ModularCirc.py
tests
test_compare.py
tests/experimental
test_experimental_base_simulator.py
test_experimental_bayesian_calibration.py
test_experimental_history_matching.py
test_learners.py
Project Total

This report was generated by python-coverage-comment-action

[review-notebook-app]

Check out this pull request on 

See visual diffs & provide feedback on Jupyter Notebooks.

---

Powered by ReviewNB

[radka-j]

• change forward_batch to return the successful x instead of failed indices.

[sgreenbury]

I think this looks great, thanks @radka-j! I had two comments:

1. I think it might be worth splitting forward_batch() into two methods: one for when it's expected for simulator to be infallible and one for fallible simulators so forward_batch() will always return the same number of results for inputs passed - I've pushed an update to this 395ee0b
2. We might consider extending the Simulator API to support simulators that can fail and always return a requested number of samples but this could be in a new issue (Extend/revise simulator API for batches #651)

I'll open a new issue to consider if the API for 1 would be preferable without two methods and also to consider if 2. would be useful but will merge this now for #632.