mixpp: bdm::mDirich Class Reference

bdm::mDirich Class Reference

Random Walk on Dirichlet. More...

#include <exp_family.h>

Inheritance diagram for bdm::mDirich:

bdm::pdf_internal< EPDF > bdm::pdf bdm::root List of all members.

Public Member Functions

void condition (const vec &val)
void from_setting (const Setting &set)
void to_setting (Setting &set) const
void validate ()

Protected Attributes

double k
 constant $ k $ of the random walk
vec & _beta
 cache of beta_i
vec betac
 stabilizing coefficient $ \beta_c $

Detailed Description

Random Walk on Dirichlet.

Using simple assignment

\[ \beta = rvc / k + \beta_c \]

hence, mean value = rvc, variance = (k+1)*mean*mean;

The greater k is, the greater is the variance of the random walk;

$ \beta_c $ is used as regularizing element to avoid corner cases, i.e. when one element of rvc is zero. By default is it set to 0.1;


Member Function Documentation

void bdm::mDirich::from_setting ( const Setting &  set  )  [virtual]

Create object from the following structure

    class = 'mDirich';
    k = 1;                      % multiplicative constant k
    --- optional ---
    beta0 = [...];              % initial values of beta
    betac = [...];              % initial values of beta stabilizing coefficients
    --- inherited fields ---
    bdm::pdf::from_setting
fulfilling form

\[ f(rv|rvc) = Di(rvc*k) \]

If the optional fields are not given, they will be filled as follows:

    beta0 = [1,1,1,...];
    betac = 0.1 * [1,1,1,...];

Reimplemented from bdm::pdf.


The documentation for this class was generated from the following files:
Generated on 2 Dec 2013 for mixpp by  doxygen 1.4.7