CI for main and small fixes

[flxrcz]

Adds CI to main and merges to be done on main, i.e. run tests on new merge requests.
Additionally the banana example is extended

[Copilot]

Pull Request Overview

This PR adds CI workflow configuration, updates linter rules for example files, and refactors the banana model example to use the DifferentiableMCMCModel base class with gradient computation.

• Adds a new GitHub Actions CI workflow for format checking, documentation building, and testing
• Configures Ruff linter to ignore specific rules for example files
• Updates the banana model example with type hints and gradient implementation

Reviewed Changes

Copilot reviewed 3 out of 3 changed files in this pull request and generated 3 comments.

File	Description
pyproject.toml	Adds per-file ignores for examples directory to relax linting requirements
examples/model_banana_2D.py	Refactors to use DifferentiableMCMCModel, adds gradient method and type hints
.github/workflows/ci.yml	Adds new CI workflow for automated testing and validation

---

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.

[Copilot]

The ignore patterns for 'examples/' use incomplete rule codes. Ruff requires complete error codes (e.g., 'INP001', 'D100', 'N802') rather than prefixes. The patterns 'INP', 'D', and 'N' will not match any rules and should be replaced with specific error codes or wildcard patterns like 'INP', 'D*', 'N*'.

Suggested change

	per-file-ignores = { "test/*" = ["D100", "D102", "D103", "D107", "INP001", "N802"], "examples/*" = ["INP", "D", "N"]}
	per-file-ignores = { "test/*" = ["D100", "D102", "D103", "D107", "INP001", "N802"], "examples/*" = ["INP*", "D*", "N*"]}

[Copilot]

The return type annotation is 'float', but this function operates on numpy arrays and returns a numpy array (or array-like) via np.exp(-logp). The return type should be 'np.ndarray' or match the actual return type. Additionally, the input arrays should likely use 'np.floating' instead of 'float' to match the pattern used in the base class, or use simpler annotations like 'np.ndarray'.

Suggested change

	x_value: np.ndarray[tuple[int], np.dtype[float]],
	y_value: np.ndarray[tuple[int], np.dtype[float]],
	) -> float:
	x_value: np.ndarray,
	y_value: np.ndarray,
	) -> np.ndarray:

[Copilot]

The type annotations use 'np.dtype[float]' but should use 'np.dtype[np.floating]' to match numpy's type system and be consistent with the base class API in src/ls_mcmc/model.py. Alternatively, use simpler type hints like 'np.ndarray' without generic parameters, which is more common and readable for example code.