Talk:Generalised hyperbolic distribution

From Wikipedia, the free encyclopedia
Jump to: navigation, search

The Density Function[edit]

The density function at the right hand table must be wrong, it appears to diverge to infinity because of the exponential term, and I cannot see any other term to suppress it. 192.38.66.116 (talk) 08:33, 15 December 2008 (UTC)

The Bessel function goes to zero and acts as the suppression. —Preceding unsigned comment added by 80.70.52.5 (talk) 15:41, 14 July 2010 (UTC)

For clarity, I suggest that the GH should be mentioned or defined explicitly as GH(lambda, alpha, beta, delta, mu), in agreement with the parameter order in the "related distributions" section; and the "parameters" box on the right should list them in the same order.Funem? (talk) 15:37, 17 August 2011 (UTC)

The configuration for obtaining a t-distribution has alpha set to zero, but if alpha is zero, the PDF returns zero. — Preceding unsigned comment added by 130.95.207.120 (talk) 05:59, 20 April 2012 (UTC)

Not really. K(0) approaches infinity, while the 1/0 term in the denominator approaches infinity as well, so L'Hôpital's Rule applies.

I suggest expanding this article into details. Most importantly, we should add the normal mean variance mixture representation to make it much more intuitive, and derive the pdf from there. Also it is important to add on different parameterizations as well as how to switch to each other, and the extension to multivariate case.[1] — Preceding unsigned comment added by 98.229.29.30 (talk) 14:40, 3 February 2013 (UTC)

  1. ^ http://cran.r-project.org/web/packages/ghyp/vignettes/Generalized_Hyperbolic_Distribution.pdf