Patterns in static

Apophenia

apop_waring.c File Reference

Data Structures

Functions

Variables


Detailed Description

The Waring distribution.


Function Documentation

double apop_rng_GHgB3 ( gsl_rng *  r,
double *  a 
)

RNG from a Generalized Hypergeometric type B3.

Devroye uses this as the base for many of his distribution-generators, including the Waring.


Variable Documentation

The Waring distribution Ignores the matrix structure of the input data, so send in a 1 x N, an N x 1, or an N x M.

apop_waring.estimate() is an MLE, so feed it appropriate apop_mle_settings.

$W(x,k, b,a) = (b-1) \gamma(b+a) \gamma(k+a) / [\gamma(a+1) \gamma(k+a+b)]$

$\ln W(x, b, a) = \ln(b-1) + \ln\gamma(b+a) + \ln\gamma(k+a) - \ln\gamma(a+1) - \ln\gamma(k+a+b)$

$dlnW/db = 1/(b-1) + \psi(b+a) - \psi(k+a+b)$

$dlnW/da = \psi(b+a) + \psi(k+a) - \psi(a+1) - \psi(k+a+b)$

To specify that you have frequency or ranking data, use

Apop_settings_add_group(your_model, apop_rank, NULL);
Todo:
This function needs better testing.

SourceForge.net Logo

Autogenerated by doxygen on 23 Nov 2009.