Parameters Modules for PySkOptimize

class pyskoptimize.params.BaseParamModel(*, name: str)

The base abstract class for all scikit-learn compatible parameters

Variables

name – The name of the parameter

abstract to_param()

The abstract class to get the parameter space with the current distribution :return: The skopt parameter

class pyskoptimize.params.CategoricalParamModel(*, name: str, categories: List)

The class to handle categorical parameter values

Variables

• name – The name of the parameter

• categories – The list of categories we are going to use

to_param()

This converts the param to what skopt needs

Returns

The categorical

class pyskoptimize.params.DefaultBooleanParamModel(*, name: str, valueBool: bool)

The class for the default parameter that is a boolean

to_param() → bool

This will create the default parameter for the boolean value

Returns

class pyskoptimize.params.DefaultCollectionParamModel(*, name: str, valueCollection: Tuple)

The class for the default parameter that is an iterable

to_param() → Tuple

This will create the default parameter for the iterable value

Returns

class pyskoptimize.params.DefaultFloatParamModel(*, name: str, valueFloat: float)

The class for the default parameter that is a float

to_param() → float

This will create the default parameter for the float value

Returns

class pyskoptimize.params.DefaultIntegerParamModel(*, name: str, valueInt: int)

The class for the default parameter that is an integer

to_param() → int

This will create the default parameter for the integer value

Returns

class pyskoptimize.params.DefaultStringParamModel(*, name: str, valueStr: str)

The class for the default parameter that is an string

to_param() → str

This will create the default parameter for the string value

Returns

class pyskoptimize.params.HasDefaultParameterSpace

Whether an object has a default parameter space

abstract get_default_parameter_space(name: str) → Dict

The abstract method to create the parameter search :param name: :return:

class pyskoptimize.params.HasParameterSpace

This is the trait for creating the parameter space

abstract get_parameter_space(name: str) → Dict

The abstract method to create the parameter search :param name: :return:

class pyskoptimize.params.NormallyDistributedParamModel(*, name: str, log_scale: bool = False, mu: Union[float, int], sigma: Union[float, int])

The class for normally (or log-normally) distributed parameters

Variables

• name – The name of the parameter

• mu – The mean

• sigma – The variance

• log_scale – A boolean if we are using the log scale

to_param() → Tuple

This converts the param to what skopt needs

Returns

The skopt parameter

class pyskoptimize.params.NumericParamModel(*, name: str, log_scale: bool = False)

An abstract base class for all purely numeric parameters

Variables

• name – The name of the parameter

• log_scale – A boolean if we are using the log scale

class pyskoptimize.params.UniformlyDistributedIntegerParamModel(*, name: str, lowInt: int, highInt: int)

This is for the uniform integer distribution

to_param()

This converts the param to what skopt needs

Returns

The integer parameter

class pyskoptimize.params.UniformlyDistributedParamModel(*, name: str, log_scale: bool = False, low: Union[float, int], high: Union[float, int])

The class for uniformly (or log-uniformly) distributed parameters

Variables

• name – The name of the parameter

• low – The lowest value

• high – The highest value

• log_scale – A boolean if we are using the log scale

to_param()

This converts the param to what skopt needs

Returns

The skopt parameter