[sdpd] Re: New Dicvol04 tested with the indexing benchmarks

[Armel Le Bail <alb...@cristal.org>]

Hi,

The results of the DICVOL04 benchmark tests are availabe at :
http://www.cristal.org/uppw/benchmarks/index.html#RESULTS

The global note goes from -8 (DICVOL91) to -5 (DICVOL04).
If someone can obtain better results, please send me the
data file (.in) ready for the DICVOL04 calculations.

No change in default mode, some gains in manual mode
are obtained by using at line 5 :
0.05 0. -8 1 1      in cases B1 and C2

However, the tolerance for 8 impurity lines together with
the search for a zeropoint can lead to > 30 cells so that the
program stop. Before to use such a search for  impurity
and zeropoint, a first manual test would be preferable by
using a line 5 such as :
0.05 0. 0 0 1  (or if the zeropoint is known, use it of course).
(in the manual cases B2 and D2, this provides the solution
just as DICVOL91 provided it, whereas applying the line 5 as
0.05 0 -8 1 1 does not provide it).

Best wishes,

Armel

PS - these benchmarks being one limited series of 20 tests do not
pretend to provide any definitive classification. You may propose
other benchmarks.



------------------------ Yahoo! Groups Sponsor --------------------~--> 
$9.95 domain names from Yahoo!. Register anything.
http://us.click.yahoo.com/J8kdrA/y20IAA/yQLSAA/UIYolB/TM
--------------------------------------------------------------------~-> 

 
Yahoo! Groups Links

<*> To visit your group on the web, go to:
    http://groups.yahoo.com/group/sdpd/

<*> To unsubscribe from this group, send an email to:
    sdpd-unsubscribe...@yahoogroups.com

<*> Your use of Yahoo! Groups is subject to:
    http://docs.yahoo.com/info/terms/