Computing the Kolmogorov-Smirnov Distribution when the Underlying cdf is Purely Discrete, Mixed or Continuous

Dimitrova, D. S., Kaishev, V. K. & Tan, S. (2017). Computing the Kolmogorov-Smirnov Distribution when the Underlying cdf is Purely Discrete, Mixed or Continuous. .

[img]
Preview
Text - Draft Version
Download (633kB) | Preview

Abstract

The distribution of the Kolmogorov-Smirnov (K-S) test statistic has been widely studied under the assumption that the underlying theoretical cdf, F(x), is continuous. However, there are many real-life applications in which fitting discrete or mixed distributions is required. Nevertheless, due to inherent difficulties, the distribution of the K-S statistic when F(x) has jump discontinuities has been studied to a much lesser extent and no exact and efficient computational methods have been proposed in the literature. In this paper, we provide a fast and accurate method to compute the (complementary) cdf of the K-S statistic when F(x) is discontinuous, and thus obtain exact p values of the K-S test. Our approach is to express the complementary cdf through the rectangle probability for uniform order statistics, and to compute it using Fast Fourier Transform(FFT). Secondly, we provide a C++ and an R implementation of the proposed method, which fills in the existing gap in statistical software. We give also a useful extension of the Schmid’s asymptotic formula for the distribution of the K-S statistic, relaxing his requirement for F(x) to be increasing between jumps and thus allowing for any general mixed or purely discrete F(x). The numerical performance of the proposed FFT-based method, implemented both in C++ and in the R package KSgeneral, is illustrated when F(x) is mixed, purely discrete, and continuous. The performance of the general asymptotic formula is also studied.

Item Type: Monograph (Working Paper)
Uncontrolled Keywords: Kolmogorov-Smirnov test statistic; discontinuous (discrete or mixed) distribution; Fast Fourier Transform; double boundary non-crossing; rectangle probability for uniform order statistics
Divisions: Cass Business School > Faculty of Actuarial Science & Insurance
URI: http://openaccess.city.ac.uk/id/eprint/18541

Actions (login required)

View Item View Item

Downloads

Downloads per month over past year

View more statistics