[cctbxbb] Bug in structure.expand_to_p1?

[Jan Marten Simons]

Am Montag 20 Juni 2011 19:16:16 schrieb Nat Echols:
> 2011/6/20 Jan Marten Simons <marten at xtal.rwth-aachen.de>
> 
> > It bothers me that there are x,y,z not in [0, 1[. Shouldn't the newly
> > expanded
> > cell only contain atoms within the unit cell?
> 
> I can't speak for Ralf, but FYI, this is exactly the opposite of what
> macromolecular crystallographers would expect - for us, the current
> behavior makes the most sense.

I see your point with mmcrystals where molecules might span the cell boundary 
quite often. But for inorganic crystallogrphy one usually just wants to have 
all atoms inside the asymmetric unit (even after expansion to P1 where the 
asymmetric unit is the unit cell).
Would any of you object if I add an optional parameter to expand_to_p1 to have 
all atoms inside the unit cell (defaulting to false of course)? I suppose the 
desired result can be achieved by calling 
.sites_mod_positive().sites_mod_short() on the result. The optional parameter 
would just be additional comfort and a little warning that by default 
expand_to_p1 will return scatterers with possible negative or bigger than 1 
fractional coordinates.

With regards,

 Dipl. Phys.
  Jan M. Simons
 
Institute of Crystallography
RWTH Aachen University