That leaves the question of how to label these statistics in a consistent,
clear and concise way: suggestions?
Phil Evans
Here is a suggestion. :)
FWIW, dtscaleaverage labels the mean I/sigmaI of the individual input
reflections as "I/sig unavg". That is, before any averaging is
performed.
After scaling and adjustment of sigmas, the I/sigmaI for
that unique hkl is calculated based on the weighted average of the
input observations that contribute to that unique hkl.
Then the mean I/sigmaI for these unique reflections as they appear
in the output reflection file is listed as well and labeled "I/sig avg".
That is, after any averaging is performed.
Here is the text the output file describing these two columns:
==========
I/sig unavg is the mean I/sig for the unaveraged reflections
in the input file.
I/sig avg is the mean I/sig for the unique reflections
in the output file.
==========
I think these columns are consistent with the columns that appear in SCALA
output. Rather than "I/sig unavg" or "I/sig avg", one could have
"I/sig in" and "I/sig out".
My issue is that errors, uncertainty and noise in the input reflections
are treated as if they are normally distributed and random without a hint
of systematic or erratic errors. This is wishful thinking in most cases.
Jim
