Talk:Gompertz distribution

Incorrect PDF plot

The plots of the PDF are not correct. At x=0, the value of the PDF is f(0, eta, b) = eta*b. Only the red curve looks correct. All the others have the wrong value at x=0. — Preceding unsigned comment added by T113355 (talk • contribs) 06:04, 24 February 2023 (UTC)[reply]

Doubt

The Gompertz cumulative distribution function contains the term $e^{bx}$ . When the data series contains x-values over 100 this term causes difficulty as it becomes extremely large. On Excel I obtain the value $1.042$ x $10^{304}$ using bx = 700, while for bx = 800 there is a warning because Ecxel can not handle the outcome any more as it becomes too large. How is that? Asitgoes (talk) 10:29, 2 June 2020 (UTC)[reply]

You may want to look at the wiki page on numerical analysis for info on that. Or an intro to floating-point arithmetic. This article will not help. Closed Limelike Curves (talk) 01:59, 15 April 2022 (UTC)[reply]

Merge with Gumbel Distribution

The Gompertz distribution is just a different parametrization of a Gumbel distribution, and as a result the two articles should be merged. Closed Limelike Curves (talk) 01:56, 15 April 2022 (UTC)[reply]

Why was this version chosen?

The Gompertz distribution is historically defined to have a hazard function $h(x)=k\,exp(\alpha \,x)$ . This is how this distribution was obtained. See, for example [Garg, Rao, Redmond 1970]https://www.jstor.org/stable/pdf/2346545.pdf Using such a representation would make it easier to give an explanation for the motivation of this distribution. Mltam (talk) 22:36, 15 December 2022 (UTC)[reply]