Brief awk tutorial

cols.txt

abc	1	2	3
def	4	5	6
ga	7	9	10
hij	1	5	99

tutorial.md

Example input

$ cat cols.txt
abc	1	2	3
def	4	5	6
ga	7	9	10
hij	1	5	99

The columns are separated by tabs, but any whitespace will do for awk.

Matching lines

Remember that "numbers" are just character strings!

Regular expressions

An "a" in the first column:

$ awk '$1 ~ /a/' cols.txt
abc	1	2	3
ga	7	9	10

An "a" at the start of the first column:

$ awk '$1 ~ /^a/' cols.txt
abc	1	2	3

A "9" in the fourth column:

$ awk '$4 ~ /9/' cols.txt
hij	1	5	99

A "9" in any column:

$ awk '/9/' cols.txt
ga	7	9	10
hij	1	5	99

A "99" in any column:

$ awk '/99/' cols.txt
hij	1	5	99

Exact string matches

Second column must be exactly "1":

$ awk '$2 == "1"' cols.txt
abc	1	2	3
hij	1	5	99

Third column must be exactly "5":

$ awk '$3 == "5"' cols.txt
def	4	5	6
hij	1	5	99

Fourth column must be exactly "9":

$ awk '$4 == "9"' cols.txt

Fourth column must be exactly "99":

$ awk '$4 == "99"' cols.txt
hij	1	5	99

Extracting output

Where the match occurs, print the entire line:

$ awk '$2 == "1" { print $0 }' cols.txt
abc	1	2	3
hij	1	5	99

... print the fourth column:

$ awk '$2 == "1" { print $4 }' cols.txt
3
99

... print the third and fourth columns:

$ awk '$2 == "1" { print $3 $4 }' cols.txt
23
599

... separated by a space:

$ awk '$2 == "1" { print $3, $4 }' cols.txt
2 3
5 99

... separated by a tab:

$ awk '$2 == "1" { print $3 "\t" $4 }' cols.txt
2	3
5	99

Arithmetic

Digit strings can also be treated as numbers by awk.

Operations

Add values:

$ awk '{ print 2 + 2 }' cols.txt
4
4
4
4

... once:

$ awk 'BEGIN { print 2 + 2 }' cols.txt
4

Add columns:

$ awk '{ print $2 + $3 + $4 }' cols.txt
6
15
26
105

Calculate means, even if they are non-integer:

$ awk '{ print ($2 + $3 + $4) / 3 }' cols.txt
2
5
8.66667
35

Running sums

Sum the values in the fourth column:

$ awk '{ sum += $4 } END { print sum }' cols.txt
118

... if the second column matches:

$ awk '$2 == "1" { sum += $4 } END { print sum }' cols.txt
102

... and also print the values used:

$ awk '$2 == "1" { print $4; sum += $4 } END { print sum }' cols.txt
3
99
102

... and also print the rows used:

$ awk '$2 == "1" { print; sum += $4 } END { print sum }' cols.txt
abc	1	2	3
hij	1	5	99
102

... and use an initial value:

$ awk 'BEGIN { sum = 1000 } $2 == "1" { print; sum += $4 } END { print sum }' cols.txt
abc	1	2	3
hij	1	5	99
1102