08-12-2014, 09:55 AM

The Prime seems to have build in PDF, CDF and ICDF functions for the following standard textbook distributions: Normal, T, Chi Sq., F, Binomial, and Poisson.

While this is great, would including a few other distributions be considered? There are many useful theoretical distributions out there, but I would think that the following standard four would add a lot of power to Prime stat's capabilities:

continuous: lognormal, exponential

discrete: hypergeometric, negative-binomial

Also, I would think that adding some easy to use graphical displays for those distributions would be of great benefit (to all users, but especially to students). Inference/Confidence Interval option already have such a display, but only for Normal and T PDF, using calculated standard error for their SD parameter.

This is pushing it, but if random numbers could be generated from each of the available distributions as well (not only Normal), that would be positively awesome.

Cheers!

While this is great, would including a few other distributions be considered? There are many useful theoretical distributions out there, but I would think that the following standard four would add a lot of power to Prime stat's capabilities:

continuous: lognormal, exponential

discrete: hypergeometric, negative-binomial

Also, I would think that adding some easy to use graphical displays for those distributions would be of great benefit (to all users, but especially to students). Inference/Confidence Interval option already have such a display, but only for Normal and T PDF, using calculated standard error for their SD parameter.

This is pushing it, but if random numbers could be generated from each of the available distributions as well (not only Normal), that would be positively awesome.

Cheers!