moving every second row to a new column with awk

Question

Here is another noob question. I have a file like this

John
30
Mike
0.0786268
Tyson
0.114889
Gabriel
0.176072
Fiona
0.101895

I need to shift every second row to a new column so it should look like this

John   30
Mike   0.0786268
Tyson  0.114889
Gabriel 0.176072
Fiona   0.101895

I just now that awk 'NR%2==0' will print every second row, could somebody suggest me how to shift them to a new column as above. Any help is most appreciated.

Related, but not awk specific: http://stackoverflow.com/questions/8522851/concise-and-portable-join-on-the-unix-command-line — Michael J. Barber, Dec 28 '12 at 11:26

score 10 · Answer 1 · edited May 04 '18 at 17:30

10

awk '{printf "%s%s",$0,(NR%2?FS:RS)}' file

Explanation:

printf will not include a new line character at the end of line (unlike print)
(NR%2?FS:RS) decides new line (RS) or field seperator (FS) based on NR%2

If your colmns should be seperated by \t

awk '{printf "%s%s",$0,NR%2?"\t":RS}'

edited May 04 '18 at 17:30

Thamme Gowda

9,459
1
41
50

answered Dec 29 '12 at 22:19

Ed Morton

157,421
15
62
152

2

This is a most excellent and elegant illustration of AWK. No other solution comes close. – Greg A. Woods Dec 31 '12 at 21:32
1

This helped me. Thanks @ed-Morton – Hsehdar Jan 24 '18 at 07:50

score 8 · Answer 2 · answered Dec 28 '12 at 11:03

8

xargs could do it, the command line is really short:

xargs -n2 < file

answered Dec 28 '12 at 11:03

Kent

173,042
30
210
270

I know it's not awk, but I love this answer. Clever! – David M Nov 15 '18 at 00:31

score 1 · Answer 3 · answered Dec 28 '12 at 09:53

1

One way is:

awk '{ printf("%-10s ",$0); if(NR%2==0) printf("\n");}' file

See it

answered Dec 28 '12 at 09:53

codaddict

410,890
80
476
515

Thank you for the one liner, it really works fine. I have just started learning awk and if I could ask you a bit more, I guess printf("%-10s ",$0) will print every thing in a line, but I could not understand after this. Thanks again. – Amit Dec 28 '12 at 10:25
`printf` prints without a newline and `%s` is the format specification for a string. `%-10s` will pad with spaces to a length of 10, left justified, or truncate a longer string. I would actually `printf("%s ", $0)` instead, with a space after %s, alternatively a tab character, `\t`. – tripleee Dec 28 '12 at 14:12
1

printf is a builtin, not a function, so the () don't do what you think they do, the trailing semi-colon is superflous, and the way to print a newline is `print ""`. – Ed Morton Dec 29 '12 at 22:21

Vijay · Answer 4 · 2012-12-31T07:48:00.913

1

sed is better to do this than awk:

sed 'N;s/\n/ /g' yourfile

also in awk:

awk '{if(NR%2!=0){p=""}else{p="\n"};printf $0" "p}' your_file

also check another solution here

edited Dec 31 '12 at 07:48

answered Dec 28 '12 at 09:54

Vijay

59,537
86
209
308

2

sed is NOT better to do this than awk as that is NOT even close to how you would do it in awk. – Ed Morton Dec 29 '12 at 22:19

score 1 · Answer 5 · answered Dec 28 '12 at 10:36

I'd be tempted to use sed like this:

sed 'N;s/\n/ /' file

However, this won't nicely format your data. You could pipe it into column -t:

< file sed 'N;s/\n/ /' | column -t

But this seems unnecessary. It would be better to use paste like this:

paste - - < file

Results:

John    30
Mike    0.0786268
Tyson   0.114889
Gabriel 0.176072
Fiona   0.101895

score 0 · Answer 6 · answered Dec 28 '12 at 16:40

0

awk 'NR%2{s=$0; next} {print s, $0}' file

or using default action

awk 'NR%2{s=$0; next} {$0=s" "$0}1' file

answered Dec 28 '12 at 16:40

slitvinov

5,273
15
25

score 0 · Answer 7 · answered Dec 28 '12 at 17:15

0

paste - - < file > newfile

paste uses tab as a separator by default.

answered Dec 28 '12 at 17:15

glenn jackman

207,528
33
187
305

moving every second row to a new column with awk

7 Answers7