On Thu, 16 Sept 2021 at 12:06, Gilles Sadowski <gillese...@gmail.com> wrote: <!-- SNIP --> > > I was seeing this as a "visual" validation test, so making it as > simple as possible (perhaps even with "interesting" values being > hard-coded).
I had thought of this as simple is best. However it should not be too hard to create an application that outputs the PDF, CDF, etc for any parameterisation. My idea would be to generate output over a range of parameterisations and then check them against a reference. For example the PDF and CDF plots for the parameterisations shown on wikipedia: https://en.wikipedia.org/wiki/Gamma_distribution https://en.wikipedia.org/wiki/L%C3%A9vy_distribution https://en.wikipedia.org/wiki/Poisson_distribution etc. Using just 1 set of hard coded values may not spot a problem. Suitable defaults can be coded into the program for each distribution. These can then be varied by command line arguments. Note that you cannot output the inverse CDF with the same input x range as the PDF, CDF and Survival functions since it is bounded to [0, 1]. This was one reason for parameterising the values. I am looking into making the application at the moment. I am not going to waste time over engineering this but I would like to have something that can test the distributions with a range of values so we can visually check them before the release. Alex --------------------------------------------------------------------- To unsubscribe, e-mail: dev-unsubscr...@commons.apache.org For additional commands, e-mail: dev-h...@commons.apache.org
