[sdpd] Powder Indexing Benchmarks

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

Hello,

As a conclusion of the Unindexed Powder Pattern of
the Week (UPPW) series, the Powder Indexing Benchmarks
are proposed :
            http://www.cristal.org/uppw/benchmarks/
There is a series of 10 tests based on the various bethanechol
chloride powder diffraction data. The tests are to be performed
2 times (in default mode, and in manual mode). Provisional
results are given for ITO13, TREOR90, DICVOL91 and McMaille.

Please (users or developers) send contributions for those lacking
indexing programs (N-TREOR, X-Cell, SVD-Index, AUTOX,
CRYSFIRE, and individual programs inside of the CRYSFIRE
suite, etc). Sorry that these benchmarks cannot be used
for testing indexing programs needing the raw powder data
(EFLECH/INDEX, GAIN).

These benchmarks are made for giving elements to newcomers
in the field in order to judge by themselves and select indexing
programs adapted to their data quality. Generally, all indexing
programs will succeed with excellent data : no impurity line,
negligible zeroshift and high accuracy. Differences between
indexing programs may become apparent in presence of
inaccurate data (synchrotron not always available, diffractometer
not always well aligned) and/or with impure samples (chemists
are doing their best, but...).

The results at that Web page are not frozen. If improvements
are obtained, for any indexing program, which would change a
failure to one of the 2x10 tests into a success, then please send
the input and output corresponding files. The scores will be
updated, thanks.

Moreover, the benchmarks may force developers to improve
their programs in order to offer more efficiency to the users
(obtaining a better score as a bonus).

Best wishes with these benchmarks, appearing timely with
a sudden multiplication of the available programs. Are recent
programs at least as efficient as the old classic ones ? The
Powder Indexing Benchmarks try to answer.

Armel



 
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/