[sdpd] Re: checkcell , version from 19.04.2000

[L.M.D.Cranswick...@dl.ac.uk]

Query posted on the Rietveld list:
> I have a version of the program CHEKCELL from
> 19.04.2000, but we some problems, in my copy
> this program run have a problem.
> The message is :
> runtime error 255 at 00002FF2

I have been testing out the latest pre-release version of 
Chekcell - mainly to help get rid of some bugs out related 
to Array errors (of which the following seems to be) and problems 
doing a "Best Solution" involving Rhombohedral cells.  
(This is available on the web below and will be E-mailed to 
anyone on request.  As mentioned below a new bug-fix 
update will be on the web soon)

-----

Hype on why you should closely follow getting the 
latest releases of Chekcell (which are presently
occuring quite often) follows:

When combined with XFIT/Winfit/Winplotr peak profiling 
(including Fundamental Parameters mode in XFIT for 
laboratory Bragg Brentano data) and the Crysfire indexing 
suite by Robin Shirley, Chekcell can give some quite 
impressive results.

An almost finished tutorial demonstrating this (including
example datafiles and output) is at:
   
  http://www.ccp14.ac.uk/tutorial/lmgp/chekcell_large_hex.htm


The question posed by this tutorial is what cell would
you choose from the following results provided via 
the Crysfire suite:

Top solution FOM (Figure of Merit) of 163.09 - Monoclinic
 18.6457 2.5543 4.4247 90.000 94.534 90.000
Vol=163.09 (V/V1=1)
   
or a few 1000 other trial cells indexing 20 out of
20 peaks including the not so obvious(?) 166th best 
solution: hexagonal with an FOM of 76.6
 5.1089 5.1089 55.7630 90.000 90.000 120.000
Vol=1260.443  (V/V1=6.00)

Using the "Best Solution" option automatically crunching 
through more than 1000 trial cells, Chekcell suggests
the Hexagonal Cell is the "Estimated Best Solution"
based on R3, R-3, R32, R3M, R-3M giving 37 Indexed 
to 93 Calculated peaks.  Where as the monoclinic
cell based on P2, PM and P2/M gives 37 Indexed to
142 Calculated peaks. 
  
TEM was originally used on the example data provided
with this tutorial and shows the hexagonal cell is
correct.  (knowing the answer in advance of course
makes this tutorial potentially quite contrived, it
can provide "food for thought")

===================

There are other examples of Chekcell doing similar work
on "unsolvable" cells - but I do not have permission yet
to based tutorial examples around these examples.
The resulting structure (in press), would not make
sense if you tried to use the best monoclinic cell.

===================

There will also be at least one update in the very 
near future - tonight or more likely tomorrow night.
(will post out when this happens)

The version on the website is quite happy with
Hexagonal/Rhombohedral cells but may have some
minor "array" problems on some systems that are
still being worked out (they seem to affect NT
machines more)

  http://www.ccp14.ac.uk/ccp/web-mirrors/lmgp-laugier-bochu/
  http://ccp14.sims.nrc.ca/ccp/web-mirrors/lmgp-laugier-bochu/
  ftp://ftp.minerals.csiro.au/pub/xtallography/ccp14/ccp/web-mirrors/lmgp-laugier-bochu/

Summary is to check for updates often - and it is recommended to
try the new version of Chekcell out with old data in "Best 
Solution" mode.  (old versions of Chekcell and other LMGP suite
software is also on the above sites in case a newer version
causes more problems to you then a previous one)

If you do strike problems, please feel free to Email them 
my way - including PC type, OS (Win95/98/ or NT) - amount of RAM.

Lachlan.

PS:  The soon to be released update has a major speed improvement
in performing the "Best Solution" option.

-- 
Lachlan M. D. Cranswick

Collaborative Computational Project No 14 (CCP14)
    for Single Crystal and Powder Diffraction
Daresbury Laboratory, Warrington, WA4 4AD U.K
Tel: +44-1925-603703  Fax: +44-1925-603124
E-mail: l.cranswick...@dl.ac.uk  Ext: 3703  Room C14
                           http://www.ccp14.ac.uk