[Octave-patch-tracker] [patch #10051] [octave forge] (statistics) Add "squaredeuclidean" distance to function pdist

[anonymous]

Follow-up Comment #2, patch #10051 (project octave):

No, it is a bit confusing, but that is the way it is. The point is that the
"sqEuclidean" metric of silhouette is not the "seuclidean" metric of pdist,
but the "squaredeuclidean" metric, which was missing. The squared Euclidean
distance is fast to compute, but it does not satisfy the triangle inequality
and does not define a metric space, and that is probably why, so far, none
bothered enough to add it to pdist in Octave. I guess that the silhouette
method (which is an empirical method, after all) is usually better served by a
fast algorithm.

    _______________________________________________________

Reply to this item at:

  <https://savannah.gnu.org/patch/?10051>

_______________________________________________
  Message sent via Savannah
  https://savannah.gnu.org/