Okay, I think I am almost done. Just to let you guys know, I have found a lot of things on my own... I would like to give data totals for a given week from Sunday to Saturday. Most of the questions I asked I had somewhat of a start on before I asked (not necerally anywhere close to the end answer). I am totally clueless on this one. 03/16/2003 10355.2 03/17/2003 10770.9 03/18/2003 9259.4 03/19/2003 8131.2 03/20/2003 8914.6 03/21/2003 10473.9 03/22/2003 12896.2 03/23/2003 11312.6 03/24/2003 11207.6 03/25/2003 9841.2 03/26/2003 8470.0 03/27/2003 10481.5 03/28/2003 12226.4 03/29/2003 15764.8 This is what I want... Week ending 03/22/2003 70801.4 week ending 03/29/2003 79304.1

Robert: Totally brute force: #!/usr/bin/ksh awk ' BEGIN { total=0 } { total+=$2 if ($1 == "03/22/2003" || $1 == "03/29/2003") { printf("Week ending %s is: %.2f\n", $1, total) total=0 } } ' data.file # end awk script Regards, Nails

hai, i read ur question. very interesting programming question. but i can't understand fully. wad my doubt was whether v have to generalise the code, or i have to find sum from 16th to 22nd and 23rd to 29th. plz explain me the question breifly. byee kannanvijay

Yes, I am looking for a sum of the 2nd column. The script nails gave me works, but it gives a sum for the first one of everything before it. Is there a way to tell it to go back only 7 days? Is there a way to have it automatically figure out the dates? Thanks for what you have given me so far. Robert