Set random seeds for Emulator models

[edwardchalstrey1]

Closes #465

TODO

• [x] Branches from #464 so merge that first
• ~~Ensure all emulators using the set_random_seed func~~ some models (LGBM and RandomForest) pass the random_seed int into the model params, so presumably you don't need to do both
• [ ] Ensure all emulators have tests for checking the predictions are consistent when the random_seed is set
• [x] Remove the seed being set separately in CNPDataset
  • Still doesn't result in test_fit_predict_deterministic_with_seed passing
• [ ] Consider refactoring so the seed is set at an even higher level (see comment below)
• [ ] Why does test_fit_predict_deterministic_with_seed always pass regardless of random_state for gaussian exact?
• [ ] May want to add a new issue to tackle later to ensure that tests pass for determinism on GPUs

Reviewers

@EdwinB12 if you have a better sense than me of how CNP/Gaussian processes work and how random seeds should affect the result would be great to get your review here. See my comments below, but my key questions are:

• Why does test_fit_predict_deterministic_with_seed fail for CNP?
  • There must be some randomness being introduced here still, but I can't see where
• Why does test_fit_predict_deterministic_with_seed always pass regardless of random_state for GP?

[github-actions[bot]]

Coverage report

Click to see where and how coverage changed

File	Statements	Missing	Coverage	Coverage (new stmts)	Lines missing
autoemulate/experimental
model_selection.py
autoemulate/experimental/emulators
base.py
lightgbm.py
random_forest.py
svm.py
autoemulate/experimental/emulators/gaussian_process
exact.py
autoemulate/experimental/emulators/neural_processes
conditional_neural_process.py
tests
test_compare.py
tests/experimental
test_experimental_base.py
test_experimental_gaussian_process_exact.py
test_experimental_lightgbm.py
test_experimental_random_forest.py
test_experimental_svm.py
Project Total

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

[codecov-commenter]

Codecov Report

All modified and coverable lines are covered by tests :white_check_mark:

Project coverage is 81.48%. Comparing base (71aba97) to head (fcc9357).

Additional details and impacted files

@@            Coverage Diff             @@
##             main     #479      +/-   ##
==========================================
+ Coverage   81.28%   81.48%   +0.20%     
==========================================
  Files         107      107              
  Lines        7420     7503      +83     
==========================================
+ Hits         6031     6114      +83     
  Misses       1389     1389              

:umbrella: View full report in Codecov by Sentry.
:loudspeaker: Have feedback on the report? Share it here.

:rocket: New features to boost your workflow:

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