Chi-Square Goodness of Fit Test for Bivariate Distributions in Polygonal Area

Abstract

An adaptation of the chi-square goodness test for the bivariate data observed in an arbitrary area is given in this study. An arbitrary area defined in a two-dimensional plane is defined by the polygonal area approach. The inside of polygons whose dominant points are determined are triangulated with the grid points of equal distance. The Delaunay algorithm was preferred in the triangulation process. Triangular areas are used to determine the classes in the polygonal area and calculate the observed and expected frequencies in these classes for the Chi-square goodness of fit test.

Keywords: Bivariate distributions, chi-square goodness of fit test, polygonal distributions.