back to list

Kees val

🔗Gene Ward Smith <gwsmith@svpal.org>

9/5/2005 11:39:48 PM

Another idea we can try with this kees metric business is to refine
the semistandard val. If v is a val, and if u has its p prime
coordinate equal to the coordiate for v over log2(p), then taking the
val with the minimum maximum for |u[i]-u[j]| gives a more restrictive
condition than the semistandard condition; it still does not suffice
to enforce a unique result in all cases.

🔗Paul Erlich <perlich@aya.yale.edu>

9/6/2005 1:29:46 PM

--- In tuning-math@yahoogroups.com, "Gene Ward Smith" <gwsmith@s...>
wrote:
> Another idea we can try with this kees metric business is to refine
> the semistandard val. If v is a val, and if u has its p prime
> coordinate equal to the coordiate for v over log2(p), then taking the
> val with the minimum maximum for |u[i]-u[j]| gives a more restrictive
> condition than the semistandard condition; it still does not suffice
> to enforce a unique result in all cases.

You lost me.