Re: UnitCell for Windows
Peter Y. Zavalij ( (no email) )
Tue, 9 Jun 1998 09:51:54 -0400
>>Unlike most existing cell refinement programs, UnitCell does not require
>>initial estimates for cell constants;
>
>UnitCell does not require initial estimates for cell contants. Fine.
>But you need to give it reflection positions and hkl Miller indices.
>Right ? So that if you have Miller indices, you have somewhere initial
>estimates for cell constants ! Probably from a previous indexing...
>I have not found a situation where you could know the Miller
>indices and not an initial estimate for the cell constant (except the case
>you have lost your data :-). Could you give more details ?
>
Of course, you have to know either miller indices of cell parameters,
but many cell refinement programs required both miller indices, as well
as initial values for cell parameters. The later could be enough far from
real values but you have to give them in order to start unit cell refinement
BUT only when the NON-LINEAR algorithm is applied. LINEAR method
does not require initial values AND unit cell equations are LINEAR in
reciprocal space, moreover such thing as zero-shift, sample shift, etc.
could be presented with some approximation in linear form because of their
small value. That is, to my understanding, what authors mean.
However, two parts in the following sentence "UnitCell does not require
initial estimates for cell constants; it uses a non-linear least squares"
do not agree with each other.
Dr. Peter Y. Zavalij University Crystallographer
Materials Research Center, SUNY at Binghamton
Ph/Fax:(607)777-4623 E-mail:zavalij@binghamton.edu
http://imr.chem.binghamton.edu/zavalij/zavalij.html