BayesicFitting

Model Fitting and Evidence Calculation

View project on GitHub



class ExponentialPrior( LaplacePrior )Source

Exponential prior distribution.

  Pr( x ) = exp( -x / scale )

By default scale = 1.

The domain is [0,+Inf]. In computational practice the domain is limited to about [0,36] scale units

Wrapper for LaplacePrior( center=0, scale=scale, limits=[0, hilim] )

Examples

pr = ExponentialPrior()                     # scale=1.0
pr = ExponentialPrior( scale=5.0 )          # scale=5

Attributes

  • scale : float
         scale of the exponential

Attributes from Prior

lowLimit, highLimit, deltaP, _lowDomain, _highDomain

Author: Do Kester.

ExponentialPrior( scale=1.0, hilimit=math.inf, prior=None )

Constructor.

Parameters

  • scale : float
         of the exponential
  • hilimit : float
         high limit
  • prior : ExponentialPrior
         prior to copy (with new scale if applicable)

copy( )

Copy the prior

shortName( )

Return a short name

Methods inherited from LaplacePrior
Methods inherited from Prior