LogUniformPrior#

class gammapy.modeling.models.LogUniformPrior(**kwargs)[source]#

Bases: Prior

LogUniform Prior.

Equivalent to a uniform prior on the log of the parameter

Parameters:

minfloat, optional: Minimum value. Default is 1e-14.
maxfloat, optional: Maximum value. Default is 1e-10.

Attributes Summary

`covariance`
`default_parameters`
`frozen`	Frozen status of a model, True if all parameters are frozen.
`max`	Parameter of a `Prior`.
`min`	Parameter of a `Prior`.
`parameters`	Prior parameters as a `PriorParameters` object.
`parameters_unique_names`
`tag`
`type`
`weight`	Weight multiplied to the prior when evaluated.

Methods Summary

`__call__`(value)	Call evaluate method.
`copy`(**kwargs)	Deep copy.
`evaluate`(value, min, max)	Evaluate the likelihood penalization term (hence -2*).
`freeze`()	Freeze all parameters.
`from_dict`(data, **kwargs)	Get prior parameters from dictionary.
`from_parameters`(parameters, **kwargs)	Create model from parameter list.
`reassign`(datasets_names, new_datasets_names)	Reassign a model from one dataset to another.
`to_dict`([full_output])	Create dictionary for YAML serialisation.
`unfreeze`()	Restore parameters frozen status to default.

Attributes Documentation

covariance#

default_parameters = <gammapy.modeling.parameter.PriorParameters object>#

frozen#: Frozen status of a model, True if all parameters are frozen.

max#

Parameter of a Prior.

A prior is a probability density function of a model parameter and can take different forms, including but not limited to Gaussian distributions and uniform distributions. The prior includes information or knowledge about the dataset or the parameters of the fit.

Parameters:

namestr: Name.
valuefloat or Quantity: Value.
unitUnit or str, optional: Unit. Default is “”.

Examples

For a usage example see Priors tutorial.

min#

Parameter of a Prior.

A prior is a probability density function of a model parameter and can take different forms, including but not limited to Gaussian distributions and uniform distributions. The prior includes information or knowledge about the dataset or the parameters of the fit.

Parameters:

namestr: Name.
valuefloat or Quantity: Value.
unitUnit or str, optional: Unit. Default is “”.

Examples

For a usage example see Priors tutorial.

parameters#: Prior parameters as a PriorParameters object.

parameters_unique_names#

tag = ['LogUniformPrior']#

type#

weight#: Weight multiplied to the prior when evaluated.

Methods Documentation

__call__(value)#: Call evaluate method.

copy(**kwargs)#: Deep copy.

static evaluate(value, min, max)[source]#: Evaluate the likelihood penalization term (hence -2*). Note that this is currently a different scaling that the Uniform or Gaussian priors. With current implementation the TS of a source with/without LogUniform prior would be different… TBD

freeze()#: Freeze all parameters.

classmethod from_dict(data, **kwargs)#: Get prior parameters from dictionary.

classmethod from_parameters(parameters, **kwargs)#

Create model from parameter list.

Parameters:

parametersParameters: Parameters for init.

Returns:

modelModel: Model instance.

reassign(datasets_names, new_datasets_names)#

Reassign a model from one dataset to another.

Parameters:

datasets_namesstr or list: Name of the datasets where the model is currently defined.
new_datasets_namesstr or list: Name of the datasets where the model should be defined instead. If multiple names are given the two list must have the save length, as the reassignment is element-wise.

Returns:

modelModel: Reassigned model.

to_dict(full_output=False)#: Create dictionary for YAML serialisation.

unfreeze()#: Restore parameters frozen status to default.

LogUniformPrior#

This Page