Package csb :: Package statistics :: Module scalemixture :: Class ScaleMixture

Class ScaleMixture

         object --+        
                  |        
pdf.AbstractDensity --+    
                      |    
        pdf.BaseDensity --+
                          |
                         ScaleMixture

Robust probabilistic superposition and comparison of protein structures Martin Mechelke and Michael Habeck

Represenation of a distribution as a mixture of gaussians with a mean of zero and different inverse variances/scales. The number of scales equals the number of datapoints.

The underlying family is determined by a prior ScaleMixturePrior on the scales. Choosing a GammaPrior results in Stundent-t posterior, wheras a InvGammaPrior leads to a K-Distribution as posterior.

Nested Classes
Inherited from `pdf.AbstractDensity`: `__metaclass__`

Instance Methods

__init__(self, scales=array([ 1., 1.]), prior=None, d=3)
x.__init__(...) initializes x; see help(type(x)) for signature

source code

array

log_prob(self, x)
Evaluate the logarithm of the probability of observing values x. source code

random(self, shape=None)
Generate random samples from the probability distribution.

source code

Inherited from pdf.AbstractDensity: __call__, __getitem__, __setitem__, __str__, estimate, evaluate, get_params, set_params

Inherited from object: __delattr__, __format__, __getattribute__, __hash__, __new__, __reduce__, __reduce_ex__, __repr__, __setattr__, __sizeof__, __subclasshook__

Class Variables
	__abstractmethods__ = `frozenset([])`

Properties
	prior
	scales
Inherited from `pdf.AbstractDensity`: `estimator`, `parameters` Inherited from `object`: `__class__`

Method Details

init(self, scales=array([ 1., 1.]), prior=None, d=3)
(Constructor)

source code

x.__init__(...) initializes x; see help(type(x)) for signature

Overrides: object.__init__: (inherited documentation)

log_prob(self, x)

source code

Evaluate the logarithm of the probability of observing values x.

Parameters:

x - values

Returns: array

Overrides: pdf.AbstractDensity.log_prob

(inherited documentation)

random(self, shape=None)

source code

Generate random samples from the probability distribution.

Parameters:

size - number of values to sample

Overrides: pdf.AbstractDensity.random

(inherited documentation)

Property Details

prior

Get Method:: unreachable.prior(self)
Set Method:: unreachable.prior(self, value)

scales

Get Method:: unreachable.scales(self)
Set Method:: unreachable.scales(self, value)

Class ScaleMixture

__init__(self, scales=array([ 1., 1.]), prior=None, d=3) (Constructor)

log_prob(self, x)

random(self, shape=None)

prior

scales

init(self, scales=array([ 1., 1.]), prior=None, d=3)
(Constructor)