Parameter Space (CADETProcess.parameter_space)

ParameterSpace defines a feasible input domain and writes parameter values into evaluation objects. It is the foundation OptimizationProblem builds on, but is also directly usable on its own for design-of-experiments studies, surrogate training data, and sensitivity screenings.

Space

ParameterSpace()	Container for parameters, evaluation objects, constraints, and dependencies.
TransformedSpace(space)	Optimizer-facing view of a ParameterSpace in normalized coordinates.

Parameters

ParameterBase(name)	Base class for an optimizable parameter.
RangedParameter(name, parameter_type, lb, ...)	Scalar parameter bounded by a lower and upper limit.
ChoiceParameter(name, valid_values)	Parameter constrained to a finite set of allowed values.

Sampling

SamplerBase()	Abstract sampler strategy.
HopsySampler([pool_size])	Uniform polytope sampler using hopsy (Hit-and-Run MCMC).
LatinHypercubeSampler()	Latin Hypercube sampler via scipy.stats.qmc.
SobolSampler()	Sobol sequence sampler via scipy.stats.qmc.
chebyshev_center(space)	Compute the Chebyshev center of the independent-variable polytope.

Constraints and Dependencies

LinearConstraint(parameters[, lhs, b])	Linear inequality constraint: lhs · x <= b.
LinearEqualityConstraint(parameters[, lhs, b])	Linear equality constraint: lhs · x = b.
ParameterDependency(dependent_parameter, ...)	Declare that one parameter depends on others via a transform.

Mappers

ParameterMapperBase(evaluation_objects)	Base class: broadcasts a write over a list of evaluation objects.
DotPathMapper(evaluation_objects, path)	Write a value to the leaf referenced by a dot-separated path.
IndexedMapper(evaluation_objects, path[, index])	Read-patch-write mapper for individual array elements or slices.
CallableMapper(evaluation_objects, fn)	Write a value by delegating to a user-supplied function.
parse_path(path)	Split a path string into typed segments.