[cctbxbb] Introducing arbitrary translations in symmetry operations

Ralf W. Grosse-Kunstleve rwgk at yahoo.com
Fri Jun 3 10:57:44 PDT 2011

Hi Emre,

This turns out to be a very long-standing oversight.
Could you try again with the next cctbx build?
Wait for build tag 2011_06_03_XXXX or higher.
It should give you 2*a,2*b,2*c.

The oversight was in cctbx/sgtbx/space_group_type.cpp, in the
cmp_change_of_basis_mx class. I was only evaluating "is one the
unit matrix but not the other". I had to add "is one a diagonal
matrix and not the other".


P.S.: If you want to try the new version straightaway, you can
use the command


Then manually build from sources as described here:


>From: Emre S. Tasci <emre.tasci at ehu.es>
>To: cctbx mailing list <cctbxbb at phenix-online.org>
>Sent: Friday, June 3, 2011 3:49 AM
>Subject: Re: [cctbxbb] Introducing arbitrary translations in symmetry operations
>Dear Ralf,
>Thank you very much for your answer. What we are actually trying to do is to refer to the tables of ITA (2006) 15.2.1.* where you have your space group & additional generators and voila! It's not the normalizers defined as a group but the group+normalizers defined as a new group.. 8)
>Here is yet another question-- I might be missing something obvious here so you got my apologies beforehand if it proves that I do:
>Take for instance SG #16, P222. I add the 3 translation operators
>plus the inversion:
>then I get  SG #47, Pmmm with (2*b,2*c,2*a) -- but why not (2*a,2*b,2*c)? As I said, I'm highly suspecting that I'm missing something very very obvious but at this moment I'm baffled.
>With my best regards,
>On 06/02/2011 07:32 AM, Ralf W. Grosse-Kunstleve wrote:
>>> But we couldn't find a way to introduce "x,y,z+t" while we are "expand_smx"ing the space group with these operators.
>> The space_group class only supports finite groups (and only in settings
>> that can be represented with integral rotation parts and rational
>> translation parts).
>> We have the class cctbx.sgtbx.search_symmetry, which multiplies the
>> discrete origin shifts into the space group and keeps track
>> of the continuous allowed origin shifts separately.
>> It would need new code to determine a full description of affine
>> normalizers. (Are they considered space groups?)
>-- =-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-
>Emre S. Tasci - http://www.emresururi.com
>Fisica de la Materia Condensada
>Facultad de Ciencia y Tecnologia
>Universidad del Pais Vasco
>Apartado 644
>48080 Bilbao / Spain
>cctbxbb mailing list
>cctbxbb at phenix-online.org
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://phenix-online.org/pipermail/cctbxbb/attachments/20110603/2657fe0e/attachment.htm>

More information about the cctbxbb mailing list