> Do you think that this averaged I/sig I could also be transferred to a
> averaged CC1/2 giving unobserved spots a CC1/2 of 0.
> --> <CC1/2>_all = shell_completeness * <CC1/2>_measured?

Actually yes it did occur to me (just after I hit 'send') that that
would be just as valid (or invalid depending on ones POV!) a tactic as
the <I / sigma(I)> one I suggested.  However since the ink is barely
dry on the K & D paper, it seemed somewhat premature to suggest
modifying their statistic! - and in any case it's still not clear to
me what is the consensus cut-off value for CC(1/2).

Cheers

-- Ian

Reply via email to