nprietveld
Dear All,
I would like to better understand the influence of weighting interatomic distance restraints on a refinement. Usually it is desiderable to reduce at minimum the restraint contribution to the refinement, just decreasing its weighting, at least until physically significant results are obtained. A point I would like to understand better is the "initial" value of a given weighting, and the "step" to be used to decrease its influence on the refinement. As an example, consider Distance_Restrain(Mg1 OH1 0 0 0 0, 2, 2.07710`_0.03415, 0.05,1). Should I start at the beginning of the refinement with an initial value for weighting of 5 or 555? If I wish to decrease its weighting, which "step" should I use? In conclusion, which is in your opinion the best way to evaluate the influence of "weighting" restraints? Is there any way to express it analitically?
Another interesting issue would be to allow the weighting value of a restraint to decrease during the refinement until say Rwp_dash decreases, stopping the process when a minimum is reached. Maybe somebody has got a way to do this?
Thanks for your attention.
Natale
johnsoevans
[in haste, sorry]
There's some discussion of this in the single crystal literature. I think there's something in both the shelx and crystals manuals. There's also the publication below (I've not gone back and looked at it):
Acta Cryst. (2002). B58, 835-840 [ doi:10.1107/S0108768102011370 ]