Evaluating the Fit for a Pareto/Bradford Distribution<meta http-equiv="Content-Type" content="text/html; charset=UTF-8" />JMP has features for evaluating the fit of a continuous variable to a distribution. In the JMP Pro 9.0.3 user interface, these functions can be accessed via the "Analyze, Distribution" menu. There is a nice selection of pre-loaded distribution definitions that can be used for the evaluation. I would like to evaluate fit versus a Pareto / Bradford Distribution, which is not in the list. Is it possible for me to add the equation for this distribution so that I can use it for the evaluation of fit?
Test System: SAS JMP Pro 9.0.3 32-bit, Microsoft Windows 7 Professional Service Pack 1 64-bit.
Re: Evaluating the Fit for a Pareto/Bradford DistributionI am not aware of a way to add items to the list.
But according to the information in the link you provided, Y = ln(X/xm) is exponentially distributed with rate alpha, where X is pareto distributed with parameter alpha and xm.
Thus, with a good estimate of xm, the minimum treshold, you can apply the tranformation using a separate column formula and fit a simple exponential distribution to the transformed data.
