Parameters

Parameters#

class BaseParameter(*args, **kwargs)[source]#

Bases: Variable[V]

property prior: BasePDF | None#

Returns the prior distribution associated with this parameter.

Returns:: Prior distribution, or None if no prior is set.
Return type:: BasePDF | None

scale(slope: ArrayLike = 1.0, offset: ArrayLike = 0.0) → Modifier[source]#

Creates a linear modifier driven by this parameter.

Parameters:

slope – Multiplicative factor applied to the histogram.
offset – Additive shift applied to the histogram.

Returns:

Modifier representing the linear effect.

Return type:

Modifier

class NormalParameter(*args, **kwargs)[source]#

Bases: Parameter[V]

Parameter whose default prior is the standard normal distribution.

Provides convenience methods for log-normal scaling and template morphing.

property prior: BasePDF#

Returns the standard normal prior distribution for this parameter.

Returns:: Standard normal distribution.
Return type:: BasePDF

scale_log_asymmetric(up: ArrayLike, down: ArrayLike) → Modifier[source]#

Creates an asymmetric log-normal modifier for this parameter.

Parameters:

up – Scaling factor applied to upward deviations.
down – Scaling factor applied to downward deviations.

Returns:

Modifier representing the asymmetric exponential effect.

Return type:

Modifier

scale_log_symmetric(kappa: ArrayLike) → Modifier[source]#

Creates a symmetric log-normal modifier for this parameter.

Parameters:: kappa – scaling factor
Returns:: Modifier representing the symmetric exponential effect.
Return type:: Modifier

morphing(up_template: H, down_template: H) → Modifier[source]#

Creates a vertical template morphing modifier for this parameter.

Parameters:

up_template – Template used for upward variations.
down_template – Template used for downward variations.

Returns:

Modifier modelling the morphing effect.

Return type:

Modifier

class Parameter(*args, **kwargs)[source]#

Bases: BaseParameter[V]

Generic parameter with optional bounds, priors, and metadata.

value#: Current parameter value (mutable via .get_value()).

name#: Optional human-readable identifier.

lower#: Optional lower bound enforced via transformations.

upper#: Optional upper bound enforced via transformations.

prior#: Optional prior distribution.

frozen#: Whether the parameter participates in optimisation.

transform#: Optional transformation applied during unwrap/wrap.

tags#: Additional metadata tags.

Examples

>>> import evermore as evm
>>> theta = evm.Parameter(value=1.0, lower=0.0, upper=2.0)
>>> theta.get_value()
Array(1., dtype=float32)

Parameters

Contents

Parameters#