CADETProcess.optimization.OptimizationProblem.create_initial_values

OptimizationProblem.create_initial_values(n_samples: int | None = 1, seed: int | None = None, burn_in: int | None = 100000, include_dependent_variables: bool | None = True) → ndarray

Create initial value within parameter space.

Uses hopsy (Highly Optimized toolbox for Polytope Sampling) to retrieve uniformly distributed samples from the parameter space.

Parameters:

n_samplesint, optional

Number of initial values to be drawn. The default is 1.

seedint, optional

Seed to initialize random numbers.

burn_inint, optional

Number of samples that are created to ensure uniform sampling. The actual initial values are then drawn from this set. The default is 100000.

include_dependent_variablesbool, optional

If True, include dependent variables in population. The default is True.

Returns:

valuesnp.ndarray

Initial values for starting the optimization.

Raises:

CADETProcessError

If not enough individuals fulfilling linear constraints are found.