When you do your crosstable read, the first 17 columns in your read are duplicated in your final table. The number of duplicates depends on the number of cross-months that you have, in your case 13 I think.
To show the correct value you need to select only one "Months", for example Dec-10.
But I don't think you want to duplicate the LMAT-values at all so you need to split this into two tables, one containg non-duplicated LMAT and one containg the LEUData. To make this work you need to make a key between the tables which might be a bit difficult.
Another option is to use Set Analysis on the sum( LMAT'11)/1000, like
i have used the expression wch you have give, the total came correct but the individual values are wrong in the table, as for the other method i am trying to break the table in parts, hope its works
