Huge (ie unreasonable) GOFs on indexing, and potentially unreasonable matches

rowlesmr

I've been having a bash at some indexing recently, and I've run into some issues with GOFs of about 10000000000, and matches with unreasonable spacegroups.

The indexing range was made through the GUI using a peaks phase, and then converting to an indexing range.

Some screenshots:

This one I would deem a reasonable match, given the input data, and the method of peak picking to begin with, but the GOF is ridiculous.

Wth this one, I'd say that the GOF is reasonable, but the spacegroup chosen as the top match is anything but.

Any ideas?

johnsoevans

Dear Matthew, I like to wait a week before answering questions I don't know the answer to! :-)

As there are no other responses: are you just in the situation where you have a small number of reflections, and some eliminated as unindexed, such that you can always find a "perfect" cell if you allow 6 degrees of freedom plus a zero point? i.e. you're doing the equivalent of fitting a straight line to two observations?

rowlesmr

Ah. Bugger. I didn't realise that unindexed peaks did that.

Any idea re the second problem with indexed peaks being several degrees away from their associated data?

johnsoevans

Not really, is there a large zero? I haven't looked back at details of Alans algorithm and gof definition so this (and previous answer) could be wrong.....