Pr/bo only

[ju-he]

Description

This PR adds a full Bayesian Optimization workflow to Propulate and integrates it into the public API, docs, tutorials, and tests.

Summary of changes:

• Adds new propagator BayesianOptimizer in propulate/propagators/bayesopt.py.
• Implements GP-based BO on CPU (SingleCPUFitter, scikit-learn backend) with robust GP fitting (_robust_lbfgs) and diagnostics.
• Adds acquisition functions and factory support:
  • EI, PI, UCB (implemented in minimization form as mu - kappa * sigma)
  • create_acquisition, rank-stretch scaling across MPI ranks
  • optional annealing and mid-run acquisition switching
• Adds mixed search-space support (float, integer, ordinal integer, categorical via one-hot + projection).
• Adds sparse subsampling for GP training (top_m elites + diversity sampling) and safer handling of non-finite losses.
• Adds BO initial design strategies (sobol default, lhs, random) and hyperparameter optimization scheduling.
• Exposes BO components in propulate/propagators/__init__.py and adds LCB alias export.
• Updates propulate/population.py:
  • normalizes numpy scalar types to canonical Python types
  • fixes categorical slice reset in Individual.__setitem__
• Adds comprehensive BO test suite in tests/test_bayesian_optimizer.py.
• Adds docs and usage/tutorial integration:
  • docs/bayesian_optimizer.rst
  • docs/tut_bayesopt.rst
  • references from docs/usage.rst, docs/tut_propulator.rst, docs/algos_explained.rst
  • adds docs/autoapi/index.rst
• Adds runnable example: tutorials/bayesian_optimizer_example.py.

Motivation/context:

• Enables sample-efficient optimization for expensive black-box objectives.
• Extends Propulate beyond population-based heuristics with a GP/acquisition-based optimizer.
• Supports heterogeneous search spaces needed in practical HPO/NAS settings.

Dependencies required for this change:

• Added runtime deps: scikit-learn, scipy

Validation notes:

• All tests pass

Type of change

• New feature (non-breaking change which adds functionality)
• This change requires a documentation update

Checklist:

• My code follows the style guidelines of this project
• I have performed a self-review of my code
• I have commented my code, particularly in hard-to-understand areas
• I have made corresponding changes to the documentation
• My changes generate no new warnings
• I have added tests that prove my fix is effective or that my feature works
• New and existing tests pass locally with my changes
• Any dependent changes have been merged and published in downstream modules