UNDUPL (CCP4: Supported Program)


undupl - remove duplicates from Madnes data, after Absurd


undupl hklin in.mtz hklout out.mtz
[optional input]


This program removes duplicate observations from Madnes data, after the data have been processed by Absurd and sorted. The problem which UNDUPL addresses arises in Madnes if the integration of a series of images is restarted. It is normal to reprocess some of the same images, to avoid losing the reflections which were unfinished and the end of the first run. Some reflections will then be measured in both runs, but they are NOT independent observations, since they come from the same images, and they need to be removed from the dataset.

Reflections are considered to be redundant if they have the same H, K, L, M/ISYM (symmetry number), and have less than a specified difference in the other columns. Limits for checking to override the defaults may be given as as the first line of command input, but the defaults will normally suffice.


Line 1:
(optional) limits (free-format) on columns 6 onwards (usually use default, which may be set with /)
Columns which may be checked are (with default limits): I (-1), SIGI (-1), IERROR (-1), TIME (-1), XDET (3), YDET (3), ROT (-1).
A negative number means no check, i.e. the defaults check only the detector positions XDET, YDET
Line 2:
(optional) filename for list of redundant reflections set to * to output inline


HKLIN: Input mtz file

HKLOUT: Output mtz file


Phil Evans, MRC-LMB Cambridge.


Unix examples:
# Use defaults (normal case)
undupl hklin bsr2_sort hklout bsr2_sortu </dev/null

#  Send duplicates to file
undupl hklin bsr2_sort hklout bsr2_sortu << EOF