(I hope you all can read my english)
I have to disturb another time Mr Thioulouse for the module Distances, but
also
I would like to make known some performances and some little bugs of ADE-4
for Win95/NT to all
the users.
The release of Distances preceding the last one performs Principal
Coordinate Analysis in half of the time the last one does.
In particular, on a Pentium 200mhz with 128Mb RAM, it takes, on a
2787x2787 (!) distance matrix, 12,25 hours and, on a 1969x1969 one, about
8 hours.
A 819x819 matrix could be diagonalized in about 10 minutes on a Pentium
(133mhz)
with more than 16MB RAM.
Two questions:
1) Why are the computation time so different (the new version of
distance
ask for more RAM also, under WinNT) ? Why are they worst ?
2) One can select the distance type (Euclidean, Chi2, Jaccard) in the
option 'Distances|Canonical
Distance' no more, why ?
I think also that if you ask for a column distance matrix and THEN for a
row distance matrix, from the same file and with the same file as output,
you
will always obtain a column distance matrix unless you restart the module.
Thank you for attention.
Federico Spinazzi
This archive was generated by hypermail 2b30 : Sat Feb 10 2001 - 10:21:36 MET