A<i>C++</i>Program for the Cramér-von Mises Two-Sample Test

As larger sets of high-throughput data in genomics and proteomics become more readily available, there is a growing need for fast algorithms designed to compute exact p values of distribution-free statistical tests. We present a program for computing the exact distribution of the two-sample Cramér-von Mises test statistic under the null hypothesis that the two samples are drawn from the same continuous distribution. The program makes it possible to handle substantially larger sample sizes than earlier proposed computational tools. The C++ source code for the program is published with this paper, and an R package is under development.