Navigation

Global	Documentation
Home	Manual
Readme	Glossary
Notes	Design
Examples	Classes
	Trouble
	Style

class HyperParameter( object )	Source

Values and priors for the parameter(s) of an ErrorDistribution.

Hyperparameters are not directly related to the model, they are parameters of the error distribution.

Information about the scale of the noise is stored in a derived class, noiseScale.

The full use of priors is reserved for Bayesian calculations as in NestedSampler

Attributes

• hypar : float
       the value of the hyperparameter. Default: 1.0
• stdev : float
       the standard deviation of the hyperparameter. Default: None
• prior : Prior
       the prior for the hyperparameter.
• isFixed : boolean
       keep the hyperparameter fixed at the value given by hypar.
       default: True

HyperParameter( hypar=1, isFixed=True, prior=None, limits=None, copy=None )

Constructor.

Parameters

• hypar : float
       value of the hyperparameter
• isFixed : bool
       True: Consider the hyperparameter as fixed
       False: Optimize the parameter too (when relevant)
               It might need a prior and/or limits to be set
• prior : None or Prior
       None : no prior is set if no limits are given else JeffreysPrior
       prior probability on the hyperparameter
• limits : None or list of 2 floats [lo,hi]
       low limit and high limit on hypar.
• copy : HyperParameter
       HyperParameter to copy

copy( )

Return a copy.

checkPrior( )

Raises

ValueError when no prior has been set.

setLimits( limits )

Set the limits on the scale within the prior.

Parameters

• limits : list of 2 float
       the [low,high] limits.

getLimits( )

Return the limits on the scale.

isBound( )

Return true is the itergral over the prior is bound.

domain2Unit( dval )

Return a value in [0,1] given a value within the valid domain of a parameter for the prior distribution.

Parameters

• dval : float
       value within the domain of a parameter

unit2Domain( uval )

Return a value within the valid domain of the parameter given a value between [0,1] for the prior distribution.

Parameters

• uval : float
       value within [0,1]

partialDomain2Unit( dval )

Return a the derivate of the domain2Unit function to dval.

Parameters

• dval : float
       value within the domain of a parameter