|
ErrorDistribution = class ErrorDistribution(builtins.object) |
| |
ErrorDistribution(fixed=None, constrain=None, copy=None)
ErrorDistribution defines general methods for a error distribution.
Error distributions are used to calculate the likelihoods.
Author Do Kester.
Attributes
----------
hyperpar : HyperParameter
hyperparameter for the error distribution
deltaP : float
delta for calculating numerical derivatives
ncalls : int
number of calls to the logLikelihood
nparts : int
number of calls to the partial of the logLikelihood
sumweight : float
sum over the weights or ndata
ndata : int
number of points in data
hypar : [float]
list of values for the hyperparameters
nphypar : int
number of hyper parameters in this error distribution
constrain : None or callable
None: Use logLikelihood as is
callable: logL = func( logL, problem, allpars, lowLhood )
returning a (modified) value of the logLikelihood.
|
| |
Constructor:
- ErrorDistribution( fixed=None, constrain=None, copy=None )
- Constructor.
Parameters
----------
fixed : dictionary of {int:float}
int list if parameters to fix permanently. Default None.
float list of values for the fixed parameters.
constrain : None or callable
function as: func( logL, problem, allpars )
returning a (modified) value of the logLikelihood.
copy : ErrorDistribution
distribution to be copied.
Raise
-----
ValueError when constrain is not a callable method.
Methods defined here:
- acceptWeight()
- True if the distribution accepts weights.
- copy()
- Return copy of this.
- domain2Unit( dval, ks )
- Return value in [0,1] for the selected parameter.
Parameters
----------
dval : float
hyper parameter value in domain
ks : int
selecting index
- getChisq( problem, allpars=None )
- Return chisq
*** Gaussian approximation ***
Sum over the (weighted) squared residuals
Parameters
----------
problem : Problem
to be solved
allpars : array_like
None take parameters from problem.model
list of all parameters in the problem
- getGaussianScale( problem, allpars=None )
- Return the noise scale.
*** Gaussian approximation ***
Parameters
----------
problem : Problem
to be solved
allpars : array_like
None take parameters from problem.model
list of all parameters in the problem
- getResiduals( problem, allpars=None )
- Return residuals: ydata - model.result
Parameters
----------
problem : Problem
to be solved
allpars : array_like
None take parameters from problem.model
list of all parameters in the problem
- hyparname( k )
- Return name of the hyperparameter
Parameters
----------
k : int
index of the hyperparameter
- isBound()
- True when all priors of its (hyper)parameters are bound
- keepFixed( fixed=None )
- Keeps (hyper)parameters fixed at the provided values.
1. Repeated calls start from scratch.<br>
2. Reset with keepFixed( fixed=None )
Parameters
----------
fixed : dictionary of {int:float}
int list if parameters to fix permanently. Default None.
float list of values for the fixed parameters.
- logCLhood( problem, allpars )
- Return the constrained log( likelihood ).
Parameters
----------
problem : Problem
to be solved
allpars : array_like
parameters of the problem
- logLhood( problem, allpars )
- Return the log( likelihood ).
Parameters
----------
problem : Problem
to be solved
allpars : array_like
parameters of the problem
- numPartialLogL( problem, allpars, fitIndex )
- Return d log( likelihood ) / dp, numerically calculated.
Parameters
----------
problem : Problem
to be solved
allpars : array_like
parameters of the problem
fitIndex : array_like
indices of parameters to be fitted
- partialLogL( problem, allpars, fitIndex )
- Return the partial derivative of log( likelihood ) to the parameters.
Parameters
----------
problem : Problem
to be solved
allpars : array_like
parameters of the problem
fitIndex : array_like
indices of parameters to be fitted
- partialLogL_alt( problem, allpars, fitIndex )
- Return the partial derivative of log( likelihood ) to the parameters.
Alternative calculation.
Parameters
----------
problem : Problem
to be solved
allpars : array_like
parameters of the problem
fitIndex : array_like
indices of parameters to be fitted
- setLimits( limits )
- Set limits on the hyper parameter(s).
Parameters
----------
limits : [low,high]
low : float or array_like
low limits
high : float or array_like
high limits
- setPriors( priors )
- Set priors on the hyper parameter(s).
Parameters
----------
priors : (list of) Prior
prior distribution for the hyperparameters
- setResult()
- toSigma( scale )
- Return sigma, the squareroot of the variance.
Parameter
--------
scale : float
the scale of this distribution.
Return default value : scale
- unit2Domain( uval, ks )
- Return domain value for the selected parameter.
Parameters
----------
uval : float
unit value of hyper parameter
ks : int
selecting index
- updateLogL( problem, allpars, parval=None )
- Return a update of the log( likelihood ) given a change in a few parameter.
This method provides the opportunity to optimize the logL calculation.
Providing this one, automatically provides the previous one.
For now it just refers to logLikelihood() itself.
Parameters
----------
problem : Problem
to be solved
param : array_like
parameters of the model
parval : dict of {int : float}
int index of a parameter
float (old) value of the parameter
|
|