Graph Fitting by Compartmentalization Method, Calculation of In-Set Differentiation, Propensity Score and Permutation Statistical Test

Abstract

By using points of sufficient density on the coordinate plane, any desired degree of graph can be drawn by dividing the coordinate plane into equal parts. Furthermore, whether a numerical set is divided within itself and whether the quantities formed by two numerical sets are separate from each other can be determined using averages without making any assumptions. Besides, robust p values in permutation statistical test can be obtained within large datasets through correction factor.