BayesicFitting

Model Fitting and Evidence Calculation

View project on GitHub



class LinearModel( Model )Source

Anchestor of all linear models.

LinearModel implements the baseResult method needed in all linear models.

For Linear models it holds that

  f( x:p ) = ∑( pi * df( x )/dpi )

which means that only the partial derivatives to pi need to be given as basePartial. The baseResult follows directly from that one. It is implemented here.

Attributes

None of its own

Attributes from Model

     parameters, stdevs, npchain, _next, _head, _operation, xUnit, yUnit (relegated to model)

Attributes from FixedModel

     npmax, fixed, parlist, mlist

Attributes from BaseModel

     npbase, ndim, priors, posIndex, nonZero, tiny, deltaP, parNames

Author Do Kester

LinearModel( nparams, ndim=1, copy=None, **kwargs )

class for all linear models.

Parameters

  • nparams : int
         the number of parameters in this model
  • ndim : int
         the dimensionality of the inputs (default: 1)
  • copy : LinearModel
         model to be copied (default: None)
  • kwargs : dict
         Possibly includes keywords from
             Model : params
             FixedModel : fixed, names
             BaseModel : posIndex, nonZero

baseResult( xdata, params )
Returns the base result of linear models.

for linear models the result is the inner product of parameters and partial derivatives.

Parameters

  • xdata : array_like
         values at which to calculate the result
  • params : array_like
         values for the parameters.
Methods inherited from Model
Methods inherited from FixedModel
Methods inherited from BaseModel