Brian Smith wrote:
> sensible default is to use r-6 sum averaging over all possible distance
hmm, 'r-6 sum averaging' it shouldn't get any more strange then this.
but seriously, I have trouble imagining a scenario where r-6 averaging
is useful when it comes to treating multiple possible assignments. The
sum of intensities of the underlying assignment possibilities is what
is considered, so one should the r-6 sum.
Only if you do ensemble averaging or time averaging it starts to make
sense to look at the r-6 average. (and for time averaging you
usually need the r-3 average)
So in the normal case where people want to make a single structure model
that fits the data as good as possible, it's always the r-6 sum that you
need (which should therefore be the default, imho).
or am I missing something?
Eiso
|