I have a file with 1000+ columns of data. I need to sum each row (not column). How can I do this with an awk one-liner?
Thank you
Example file:
1 1 1 1 1 1 1 1 1 1
2 2 2 2 2 2 2 2 2 2
3 3 3 3 3 3 3 3 3 3
The output should be:
10
20
30
Please give a sample of your input file, otherwise we've no idea how to write any awk, one line or otherwise. And a clearer description of your requirement would help.
You have to sum each row how? Column 1 of row 1 with column 1 of row 2, or....?
Thanks.
I'm feelin' lucky, so here's a shot in the dark:
awk '{for(i=1;i<=NF;i++) t+=$i; print t; t=0}'
Shrinking it a bit further at clarity's expense:
awk '{for(i=t=0;i<NF;) t+=$++i; $0=t}1'
If these are of no use, oh well .. it's the thought that counts 
Alister
I saw your post, and thought... I'll give him a second to add code tags... and, of course you did 
Nice job with such little info to go on
Thanks for the swift response.
-EDIT- did not see the edited part ^^
Assuming 'data' is yout datafile and numbers are separated with a comma:
awk '{ for(i=1; i<=NF;i++) j+=$i; print j; j=0 }' data
seems to do the job.
unix.com$ cat data
1 2 3 4 5
6 7 8 9 10
11 12 13 14 15
unix.com$ awk '{ for(i=1; i<=NF;i++) j+=$i; print j; j=0 }' data
15
40
65
0
Hi, scottn:
Heheheh. When I saw that I'd forgotten the code tags, and saw your post just above, I remembered how many times I've seen your name in edits and figured I'd better get on that ASAP
Cheers,
Alister
Via bash
# while read line;do echo $((${line// /+}));done < infile
10
20
30
Another one, since it probably doesn't have to be awk after all
tr \ '+' < infile | bc
Thanks for the multiple options.
If blank lines should be handled and summed to zero:
tr \ '+' < infile | sed 's/^$/0/' | tr \ '+' | bc
Alister
One in perl,
perl -lane 'foreach (@F) { $s+=$_; } print $s;' file
You forgot to zero s for each line read.
Alister
Yes..It was a mistake...Thanks for pointing it out.
perl -lane '$s=0; foreach (@F) { $s+=$_; } print $s;' file