Compute the Kuiper statistic to compare two samples.
The first set of data values.
The second set of data values.
The raw test statistic.
The probability of obtaining two samples this different from
the same distribution.
The fpp is quite approximate, especially for small samples.