The Kolmogorov distribution has distribution function

For a sample of size n from a continuous distribution function F(x),

K(x) is the limiting distribution of √n sup

For the two-sample problem with a sample of size m from F(x) and a sample of size n from G(x),

K(x) is also the limiting distribution of √((m+n)/mn) sup

For example, K(1) = 0.73. The computation is accurate to .000005.