Some well-known methods to generate a class of distributions are transformation of variables, use of conditional distributions, introducing of additional parameters or the use of Copulas. A Contour-Transformation technique is based on geometric approach and is used to generate a new class of densities. Contour-Transformation technique is different from the above methods in the sense that it may not be possible to obtain the density function obtained by a CT of a given density by using any of the above well-known methods.
In this paper we generate densities based on the size of Contours of a continuous density function f and study some of their distributional properties. As an Illustration contour densities of Exponential, Pareto distribution, Cauchy distribution have been discussed.
Unimodal, Symmetric, contour density, Chi square test for goodness of fit
