REAP

Peter writes:

The second global stat is the difference between the means for the data sets (effectively, the distances used in a clustering algorithm). This could be presented to the user as a matrix with the off diagonal elements on one side of the matrix being the difference/variance. The rows and columns of this matrix are data sets. If you include the diagonal, that could be the means/variances for each data set alone; i.e., the simple table mentioned above. No need to repeat the other side of the off diagonal elements since they simply repeat - would present too many numbers.