At 03:48 06/09/2006, Bob Hanson wrote:
>Egon and Nico and others --
>
>I'm quite excited about this space group business and Jmol.

So am I.

>Probably
>some of these methods should be ported to the CDK, as they will have
>quite general interest.

Certainly. In fact I think this should be done as a general Blue 
Obelisk resource as it's so important. Other wise we run the risk of 
different variant implementations in Jmol, CDK, JUMBO, Openbabel, 
etc. It doesn't really alter the work to be done, simply how it is 
packaged. So if we have, say, BOSG (Blue Obelisk Space Groups) then 
it could comprise:
* definitive and tutorial material
* lookup tables (e.g. for HM-> Hall)
* Hall2SymOp generating code. This might have to be ported to C++ as 
well and possibly also FORTRAN for The solid state community
* other utility routines.

This has the advantage that the information is only held once so that 
bugs in lookup table are mended for everyone simultaneously



>The three classes are:
>
>   HallInfo  -- parses and inteprets Hall symbols,
>                generating the associated matrices
>
>   SymmetryOperation -- extends Matrix4f; parses and
>                generates "x,y,z" lists and matrices

Note that the use of x/12 is restricted to crystallography. In 
principle the matrix translations elements can be anything (although 
they won't generate a spacegroup). The spacegroup is generated by 
applying an infinite number of translation matrixes to the spacegroup 
type. I can see that Hall matrices could be used for lower 
dimensionality problems (e.g. surfaces and fibres) where some of the 
periodic translations are omitted. This is why CML uses real numbers 
in the 4*4 matrices

>   SpaceGroup -- overall container for a space group;
>                generates full operation list; looks up
>                symbol names; identifies unique axis;

This is formally a spacegroup type, I think

>                does the actual atom coordinate operations
>
The atom coordinate operations require knowledge of the structure and 
that can be held in many ways. It might apply when there is a clear 
externalized data structure like CIF or CML. But there is also the 
problem of how the periodic translations are actually applied  so I 
would omit this operation.





>Q: Do you think we should set it up so that the list of 600+ space
>groups is read by Jmol from an external source only if/when it needs it?

Yes. (I would prefer to see Jmol us a BO spacegroup routine for this purpose)

>Q: What's the proper way to do that -- would it be included in the
>distribution as a TXT file?

It can be held in CML iif this is useful. It can probably also be 
held in CIF with implicit semantics. I don't like TXT files if they 
can be avoided

>Q: Would Jmol read this automatically when it needed to? Or would we
>have some sort of DATA command that might read other data types as well.
>(Color schemes?) Or might we just set it up as a script file and use the
>mechanism already in place -- something like:

This is a general Jmol question - I would prefer to see as much 
generic functionality as possible normalised in BO resources as:
* it avoids multiple sources (e.g. we should only have one periodic 
table within all BO software. That is certainly my intention for 
JUMBO, though I haven't actually done it. It makes things like 
covalent radii much more normalised
* we use the same algorithms for bonding, bond orders, etc. So if 
atoms are bonded by default in Jmol then they should be in JUMBO and CDK
* we use the same colour schemes
etc.
etc.

Jmol has done massive work in this area and there is a lot in there 
that we can benefit from. As you know we are meeting in SF next week. 
Christoph will not be drinking beer...


>#_data spacegroup
>....
>....
>....
>#_end data
>
>so that it all can be in a script file?
>
>Syd, what do you know about the format of this sort of data? Do other
>programs read that data in a standard format? Perhaps XML?

There is a lot of similar stuff in solid state programs, especially 
GULP, CASTEP< SIESTA< DL_POLY, etc. There would be a great advantage 
to having symmetry normalised across them. I don't think it's difficult

P.




Peter Murray-Rust
Unilever Centre for Molecular Sciences Informatics
University of Cambridge,
Lensfield Road,  Cambridge CB2 1EW, UK
+44-1223-763069 


-------------------------------------------------------------------------
Using Tomcat but need to do more? Need to support web services, security?
Get stuff done quickly with pre-integrated technology to make your job easier
Download IBM WebSphere Application Server v.1.0.1 based on Apache Geronimo
http://sel.as-us.falkag.net/sel?cmd=lnk&kid=120709&bid=263057&dat=121642
_______________________________________________
Jmol-developers mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/jmol-developers

Reply via email to