Polish new algorithms

[janosg]

Polish and align a few things that slipped through the reviews for new algorithms.

• Improve error message if iminuit is not installed
• Set budget for PSO higher to reduce random failures
• Test stochastic algorithms separately from deterministic algorithms in test_many_algorithms. Stochastic algorithms would be all algorithms that have a seed option. They should be tested with a fixed seed to avoid spurious failures.

[codecov]

Codecov Report

Attention: Patch coverage is 71.42857% with 2 lines in your changes missing coverage. Please review.

Files with missing lines	Patch %	Lines
src/optimagic/optimizers/iminuit_migrad.py	66.66%	2 Missing ⚠️

Files with missing lines	Coverage Δ
src/optimagic/optimizers/nevergrad_optimizers.py	97.67% <100.00%> (+2.32%)	⬆️
src/optimagic/optimizers/iminuit_migrad.py	92.30% <66.66%> (-4.84%)	⬇️

🚀 New features to boost your workflow:

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

[timmens]

I've implemented point 3 by splitting all test functions into a stochastic and deterministic one. Alternatively, we could also define algo_options conditionally inside the test function depending on whether the algorithm is stochastic or not.

[janosg]

I've implemented point 3 by splitting all test functions into a stochastic and deterministic one. Alternatively, we could also define algo_options conditionally inside the test function depending on whether the algorithm is stochastic or not.

Thinking about that: We could just have called every algorithm with a seed. Optimagic would discard it automatically where it doesn't apply 🤦

On the other hand: Separating the stochastic algorithms will allow us to use something like flaky to re-run stochastic tests which is something we definitely don't want to do for deterministic algorithms.

So I would say we keep the separation.