An optimized algorithm to determine the values of the exact cumulative distribution function of some discrete statistics (Q1966370)

Let \(a=(a_1,\dots,a_N)\) be a finite sequence of real numbers. The authors propose an algorithmic procedure which allows to calculate the number of combinations formed of \(k\) elements \((m\leq k\leq M)\) of distinct order among the sequence \(a\), whose sum is less than or equal to \(X\). As a consequence, the exact distributions of a certain rank statistics for a finite samples can be calculated as well as exact distributions in so-called models of association indices.