Find most frequent line in file in bash

Question

Suppose I have a file similar to as follows:

Abigail 85
Kaylee 25
Kaylee 25
kaylee
Brooklyn
Kaylee 25
kaylee 25

I would like to find the most repeated line, the output must be just the line.

I've tried

sort list | uniq -c

but I need clean output, just the most repeated line (in this example Kaylee 25).

See also http://superuser.com/questions/383726/get-the-most-common-appearing-lines-from-file-in-linux — rogerdpack, Jan 23 '15 at 19:02

score 7 · Accepted Answer · answered Jun 04 '13 at 16:03

7

Kaizen ~

$ sort zlist | uniq -c | sort -r | head -1|  xargs | cut -d" " -f2-

Kaylee 25

does this help ?

answered Jun 04 '13 at 16:03

Nitin4873

16,804
1
13
15

I don't think the call to `xargs` is necessary here. – chepner Jun 04 '13 at 16:17
xargs .... its to supress the spaces , its plain and simple . you could also use translate if you liked !! – Nitin4873 Jun 04 '13 at 17:08
`xargs` will suppress valid consecutive spaces as well. You'll probably need to remove `xargs` and use some other method like `sed s/^\s\+[0-9]\+\s//'` to remove the count column instead of `cut` – Samveen Jun 04 '13 at 18:29
Use `sort --version-sort` instead of `sort -r`. For example, `--version-sort` will properly output "351, 51, 5", while `-r` will output "5, 351, 51". – fri Feb 24 '16 at 15:06
Note that this solution is wrong because the second `sort` doesn't have `-n` (pointed out by the [answer below](https://stackoverflow.com/a/26229009/)) – user202729 Aug 03 '19 at 09:04

score 4 · Answer 2 · edited Oct 07 '14 at 06:01

4

IMHO, none of these answers will sort the results correctly. The reason is that sort, without the -n, option will sort like this "1 10 11 2 3 4", etc., instead of "1 2 3 4 10 11 12". So, add -n like so:

sort zlist | uniq -c | sort -n -r | head -1

You can then, of course, pipe that to either xargs or sed as described earlier.

edited Oct 07 '14 at 06:01

Alex

1,549
2
16
41

answered Oct 07 '14 at 05:17

Sina Bahram

41
1

score 2 · Answer 3 · answered Jun 04 '13 at 16:25

2

awk -

awk '{a[$0]++; if(m<a[$0]){ m=a[$0];s[m]=$0}} END{print s[m]}' t.lis

answered Jun 04 '13 at 16:25

jim mcnamara

16,005
2
34
51

Interesting and different solution. – Samveen Jun 04 '13 at 18:31

score 0 · Answer 4 · answered Jun 04 '13 at 16:19

0

$ uniq -c list | sort -r | head -1 | awk '{$1=""}1'

Kaylee 25

Is this what you're looking for?

answered Jun 04 '13 at 16:19

mproffitt

2,409
18
24

Find most frequent line in file in bash

4 Answers4

Linked