Selection

pylit.core.selection.heuristic(config, prep, has_sigma)

Perform parameter selection using a heuristic approach.

Parameters:

• config (Configuration) – Configuration object.

• prep (Preparation) – Preparation object.

• has_sigma (bool) – Whether the model uses kernel widths as parameters.

Return type:

Tuple[ndarray, ndarray]

Returns:

The heuristic kernel centers and, if applicable, the kernel widths.

pylit.core.selection.simulated_annealing(config, prep, has_sigma, max_iter=1000)

Perform parameter selection using simulated annealing.

This method optimizes kernel centers and widths by iteratively proposing new parameter values and accepting them based on a Metropolis–Hastings acceptance criterion.

Parameters:

• config (Configuration) – Configuration object.

• prep (Preparation) – Preparation object.

• has_sigma (bool) – Whether the model uses kernel widths as parameters.

• max_iter (int) – Maximum number of iterations performed during the optimization procedure.

Return type:

ndarray

Returns:

The optimized kernel centers and, if available, the optimized kernel widths.