[sdpd] Multi-decomposition

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

Hi,

Review of the latest J. Appl. Cryst. issue (volume 37, part 2,
April 2004)

Thanks to Altomare et al. continuing to adapt the Le Bail
decomposition method to the application of direct methods
in the EXPO software.

In Altomare et al. J. Appl. Cryst. 34 (2001) 704-709,
"random partitions of the cluster intensity are generated
and then submitted to cycles of a Le Bail algorithm to
make them compatible with the experimental pattern".

In Altomare et al. J. Appl. Cryst. 36 (2003) 906-913,
"described a decomposition technique based on systematic
variation of the n intensities Ii of each cluster, under the condition
that the cluster intensity I is a constant...indeed, each of
the 1500 decomposition process aims at providing a suitable input
to the tangent formula..."

Now, in Altomare et al. J. Appl. Cryst. 37 (2004) 204-209,
"error -correcting codes are used to explore the set of
decomposition trials obtained by the partitions of various
clusters of overlapping reflections". decomposition trials
are reduced to 27 instead of 1500.

So this is a way to present multiple intensity distributions,
compatible with the powder pattern (rather than to apply
the rough equipartition), to the direct method.

The question is : is there any other application than
the direct methods which would need these multi-trial-
decompositions ?

If not, why to call that process "a procedure for the
decomposition of a powder diffraction pattern" ?

I would instead call that process "multi-decomposition
for direct methods applications", or something similar.

Only my 2 cents.

Armel

PS - If someone wish to comment the two papers
"High-throughput powder diffraction. etc" parts I and II,
please do so.



 
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/