apparun.parameters#

Classes#

`EnumParam`	Impact model enum parameter.
`FloatParam`	Impact model float parameter.
`ImpactModelParam`	Impact model parameter.
`ImpactModelParams`

Module Contents#

class apparun.parameters.EnumParam#

Bases: ImpactModelParam

Impact model enum parameter. Contains a default, list of options and associated weights for monte carlo run. runs.

corresponds(symbol_name: str) → bool#

draw_to_distrib(samples: numpy.ndarray) → List[str]#

full_option_name(option: str) → str#: Return option with parameter name as a prefix. :param option: enum option. :return: option name with parameter name as a prefix.

look_up_table() → Dict[str, Dict[str, int]]#: Map every option to its one hot encoded vector, i.e. a dict with dummies names and associated value (1 if corresponding to current option, 0 otherwise). :return: a nested dict with all option and corresponding one hot encoded vector.

transform(values: str | List[str]) → Dict[str, float | numpy.array]#: Transform option, or list of options to be readily usable by ImpactModel. Value(s) one hot encoded, then mapped to a dict with parameter name as a key. :param values: :return: a dict mapping parameter name and transformed values.

property dummies_names#: Give dummies names of all possible options. A dummy name is the option with parameter name as a prefix. :return: a list containing dummies names for all possible options.

property options#

type: str = 'enum'#

weights: Dict[str, float]#

class apparun.parameters.FloatParam(**kwargs)#

Bases: ImpactModelParam

Impact model float parameter. Contains a default, min and max values, as well as a distribution for monte carlo runs. Supported distributions is “linear”.

corresponds(symbol_name: str) → bool#

draw_to_distrib(samples: numpy.ndarray) → List[float]#

transform(values: float | List[float]) → Dict[str, float | numpy.array]#: Transform float value, or list of float values to be readily usable by ImpactModel. Value(s) are mapped to a dict with parameter name as a key. :param values: :return: a dict mapping parameter name and transformed values.

update_bounds()#: Compute min and max attributes from default attribute if pm, or pm_perc attributes is not None.

update_default(new_value: float)#: Change default value of the parameter, and compute new min and max values if they are defined using pm or pm_perc. :param new_value: new default value

distrib: str | None = 'linear'#

max: float | None = None#

min: float | None = None#

pm: float | None = None#

pm_perc: float | None = None#

type: str = 'float'#

class apparun.parameters.ImpactModelParam#

Bases: pydantic.BaseModel

Impact model parameter.

corresponds(symbol_name: str) → bool#

draw_to_distrib(samples) → List[str] | List[float]#

static from_dict(impact_model_param: dict) → ImpactModelParam#: Convert dict to ImpactModelParam object. Subclass must be specified in ‘type’ field. :param impact_model_param: dict containing construction parameters of the param. :param frozen: should base model be frozen? Allows it to be hashable. :return: constructed param

name_is_valid() → bool#: Check if parameter’s name is considered as valid. Parameter name should only be composed of letters, numbers, underscores, and contain at least one letter. :return: if parameter’s name is valid

to_dict() → dict#: Convert self to dict. :return: self as a dict

update_default(new_value: float | str)#: Change default value of the parameter, and compute new bounds if relevant. :param new_value: new default value

default: str | float#

name: str#

class apparun.parameters.ImpactModelParams(**args)#

Bases: pydantic.BaseModel

draw_to_distrib(samples: numpy.ndarray) → Dict[str, List[float] | List[str]]#

find_corresponding_parameter(symbol_name: str, must_find_one: bool | None = True) → ImpactModelParam | List[ImpactModelParam]#

static from_list(parameters) → ImpactModelParams#

get_default_values(parameter_names: List) → Dict#

get_missing_parameter_names(parameters: List | Dict) → List#

get_parameter_by_name(parameter_name: str) → ImpactModelParam | None#

set_sobol_problem()#

sobol_draw(n) → numpy.ndarray#

to_list(sorted_by_name: bool | None = False) → List[Dict]#: Convert each parameter of parameters attribute to a dict, return them as a list, optionally sorted by parameter’s name :param sorted_by_name: sort parameters list by parameter’s name if True :return: list of parameter, themselves as dict

uniform_draw(n) → numpy.ndarray#

update_defaults(defaults: Dict[str, str | float])#: Change default value of the parameters contained in argument, and compute new bounds if relevant. :param defaults: dict containing parameter name as key and new default as value

property names#

parameters: List[ImpactModelParam]#

sobol_problem: Dict | None = None#