Hi, I have a data file that contains the following data. Q: How do I replace the dashes "-" (using awk gsub) with "Nan" with replacing the minus sign in front of the digits)? Thanks, 0,11:58:15:028,4095,0,-61.4,1,2,-77,-78,38,-80,-81,38,-,-,-,-,-,- 0,11:58:15:128,4095,0,-61.6,1,2,-76,-78,38,-80,-81,37,-,-,-,-,-,- 0,11:58:15:228,4095,0,-61.6,1,2,-76,-78,38,-80,-81,37,-,-,-,-,-,- 0,11:58:15:328,4095,0,-61.6,1,2,-76,-78,37,-80,-81,37,-,-,-,-,-,- 0,11:58:15:428,4095,0,-61.5,1,2,-76,-77,38,-80,-81,37,-,-,-,-,-,-

opps, I mean ... without replacing the minus sign in front of the digits)?

I am using Solaris, so I'm using nawk. One way is to look at each field and only change fields consisting of "-": #!/bin/ksh nawk ' BEGIN { FS=","; OFS="," } { for(i=1; i<=NF; i++) gsub("^[-]$", "Nan", $i) print $0 } ' datafile.txt

Thanks nails. I was hoping to avoid checking each field...

It's been a verrrry long time since I did anything with awk, but wouldn't modifying nails script to: gsub("^[,-,]$", ",Nan,", $i) give you what you need? I don't remember if you have to escape the commas or not. MIKE http://www.skeptic.com/