As discussed in
MAHOUT-165 at some point, Vector (and Matrix, but let's put that on a separate ticket) could do with a nice exposure of methods like the following:
this is good for generalized inner products and distances. Also nice:
Which generalizes norms and statistics (mean, median, stdDev) and things like that (number of positive values, or negative values, etc...).
These kind of thing exists in Colt, and we could just surface it up to the top.