Some explanations:

Col B serves to assign an arbitrary unique "suffix" to the data in col A.
This suffix assignment is necessary to distinguish between all multiple
occurences of the source numbers which you have in col A. Eg there could be
multiple occurences of 10, -10, 11, -11, etc all the way down in col A.

Col C then checks for the match of the original source numbers in col A
concatenated with the corresponding suffixes in col B against the full table
array composed. Where it matches (ie cancels out), the expression returns a
null string. Where it doesn't match, it returns an arbitrary row number as a
flag. This flag can then be used in the CF formula to format col A for all
the non-cancellations, ie all the unique source numbers in col A which do
not cancel out

The flag could also be used as well in another col E, to "float up" all the
non-cancellations from col A.

Easiest to see what's happening by referring
to this small working sample (as posted earlier):
http://cjoint.com/?kbafj66Yrv
NettOffPosnNeg.xls

--
Max
Singapore
http://savefile.com/projects/236895
xdemechanik
---