Parameter¶

class ionworkspipeline.Parameter(name, initial_value=None, bounds=None, prior=None, normalize=None, check_bounds=None, check_initial_value=None, initial_guess_distribution=None, _fitting_scale=None, _is_transformed=None)¶

Parameter object for use in data fits, stores the initial value and bounds

Inherits from pybamm.InputParameter for compatibility with PyBaMM models. Like every PyBaMM symbol it is immutable once constructed: the with_* methods (e.g. with_bounds()) return a changed copy instead.

Parameters¶

namestr: The name of the parameter.
initial_valuefloat or int, optional: The initial value to be assigned to the parameter. If not provided and bounds are given, will be set to the mean of finite bounds. Defaults to 1 if no bounds are given.
boundstuple, optional: A tuple defining the lower and upper bounds for the parameter. Defaults to (-np.inf, np.inf).
prioriwp.stats.Distribution, optional: The prior distribution for the parameter.
normalizebool, optional: Whether to normalize the parameter by its initial value. Defaults to True.
check_boundsbool, optional: Whether to check the bounds of the parameter. Defaults to True.
check_initial_valuebool, optional: Whether to check the initial value of the parameter. Defaults to True.
initial_guess_distributioniwp.stats.Distribution, optional: The initial guess distribution for the parameter. Defaults to a uniform distribution between the bounds. The provided distribution must be either a iwp.stats.Uniform or iwp.stats.PointMass.

Extends: ConfigMixin

See also: Parameter — field-level documentation.

property base_parameter¶: Returns the base parameter

property bounds¶: Returns the corrected parameter bounds

get_bounds()¶: Returns the unmodified bounds

get_initial_guess_distribution() → Distribution | None¶: Returns the unmodified initial guess distribution for the parameter

get_initial_value()¶: Returns the unmodified initial value

property initial_guess_distribution: Distribution¶: Returns the initial guess distribution for the parameter

property initial_value¶: Returns the corrected initial value

property is_transformed¶

Returns whether the parameter represents a transformed variable.

Returns¶

bool: True if the parameter is a transformed variable (e.g., “Log10(Diffusion Coefficient)”) False if the parameter is in its original form (e.g., “Diffusion Coefficient”)

Examples¶

>>> from ionworkspipeline.data_fits.parameter import Parameter
>>> from ionworkspipeline.data_fits.transforms import Log10
>>> param = Parameter("Diffusion Coefficient", 1e-9, (1e-10, 1e-8))
>>> param.is_transformed
False
>>> transformed_param = Log10(param)
>>> transformed_param.is_transformed
True

property prior¶: Returns the prior

to_config() → dict¶

Convert the Parameter to a JSON-serializable configuration dictionary.

Returns a dictionary that can be used with parse_parameters() to reconstruct the parameter. The name field is excluded since it comes from the dictionary key in the parser.

Returns¶

dict: Configuration dictionary with all initialization parameters, excluding validation-only parameters (check_bounds, check_initial_value) and name (which comes from the parser’s dict key).

to_dict(name)¶

Convert the parameter to a dictionary for fitting.

This method is used by the DataFit class to process parameters for fitting. For a basic Parameter object, it simply returns a dictionary with one entry mapping the given name to this parameter object.

Parameters¶

namestr: The name to use for the parameter in the model. This may be different from the parameter’s internal name.

Returns¶

dict: A dictionary with a single entry mapping the given name to this parameter.

with_bounds(bounds, check_initial_value=True, check_bounds=True)¶

Return a copy of this parameter with different bounds.

Parameters¶

boundstuple: The new (lower, upper) bounds; either may be None for unbounded.
check_initial_valuebool, optional: Whether to check the initial value lies within the new bounds. Defaults to True.
check_boundsbool, optional: Whether to check the new bounds are increasing. Defaults to True.

Returns¶

Parameter: The new parameter. This one is unchanged.

with_fitting_scale(fitting_scale=None)¶

Return a copy of this parameter with a different fitting scale.

Parameters¶

fitting_scalefloat, optional: The scale factor the parameter is normalized by, used as its magnitude. None scales by the initial value. Ignored when normalize is False, where the scale is always 1.

Returns¶

Parameter: The new parameter. This one is unchanged.

with_initial_guess_distribution(initial_guess_distribution)¶

Return a copy of this parameter with a different initial guess distribution.

Parameters¶

initial_guess_distributioniwp.stats.Uniform, iwp.stats.PointMass or None: The new distribution, which must lie within the bounds.

Returns¶

Parameter: The new parameter. This one is unchanged.

with_initial_value(initial_value, check_initial_value=True)¶

Return a copy of this parameter with a different initial value.

Parameters¶

initial_valuefloat or None: The new initial value. None derives it from the bounds.
check_initial_valuebool, optional: Whether to check the new value lies within the bounds. Defaults to True.

Returns¶

Parameter: The new parameter. This one is unchanged.

with_normalize(normalize)¶

Return a copy of this parameter with normalization switched on or off.

Parameters¶

normalizebool: Whether to normalize the parameter by its fitting scale.

Returns¶

Parameter: The new parameter. This one is unchanged.

with_prior(prior)¶

Return a copy of this parameter with a different prior.

Parameters¶

prioriwp.priors.Prior, iwp.stats.Distribution, dict or None: The new prior, in any form the constructor accepts.

Returns¶

Parameter: The new parameter. This one is unchanged.