How to sort numbers in two columns in Ascending Order

Question

[Kumar]

I have a large file with un-sorted order numbers in column 2 and 3. I need to sort these numbers in Ascending Order. Please see example file below and suggest.

File:

NODE_55_length_30858_cov_27.421 19951 19901
NODE_55_length_30858_cov_27.421 19930 19900
NODE_55_length_30858_cov_27.421 17578 17613
NODE_55_length_30858_cov_27.421 16544 16578
NODE_55_length_30858_cov_27.421 19982 19932

OUTPUT:

NODE_55_length_30858_cov_27.421 19901 19951
NODE_55_length_30858_cov_27.421 19900 19930
NODE_55_length_30858_cov_27.421 17578 17613
NODE_55_length_30858_cov_27.421 16544 16578
NODE_55_length_30858_cov_27.421 19932 19982

[Mensur Dlakic]

With thanks to two members who answered, I'd like to hear opinions about the original question. It strikes me as a column-sorting question which is outside of the scope.

Does the fact the poster is interested in sorting the names of assembly contigs make it within the scope? I see this pop up occasionally where non-bioinformatics questions are answered because they are tangentially related to bioinformatics topics, while other questions are quickly shut down.

[Carambakaracho]

in my practical bioinformatic work (as customer service) this is a very frequent task. This allows at least two conclusions

1. Some bioinformatics specialists in customer service jobs are overpaid column sorting experts
2. Complex column sorting is a non trivial task for beginners.

For narcissistic reasons I'd largely favor the second conclusion, but I admit complex sorting was way less trivial when I needed a Schwartzian transform (years ago in Perl) for things that python for example can do naturally.

Answer 1

[JC]

Using perl-one-liner:

$ perl -lane '($F[1]>$F[2]) ? print "$F[0]\t$F[2]\t$F[1]" : print $_' < in
NODE_55_length_30858_cov_27.421 19901   19951
NODE_55_length_30858_cov_27.421 19900   19930
NODE_55_length_30858_cov_27.421 17578   17613
NODE_55_length_30858_cov_27.421 16544   16578
NODE_55_length_30858_cov_27.421 19932   19982

Answer 2

[shawn.w.foley]

You can do something like that with awk.

awk 'BEGIN{OFS="\t";FS="\t"} {if ($2 < $3) print $0; else print $1,$3,$2}' file.txt

This defines the field seperator (FS) and output field separator (OFS) as tabs, then states that if column 2 < column 3 print the line, else print column 1, column 3, column 2.

Input:

NODE_55_length_30858_cov_27.421 19951   19901
NODE_55_length_30858_cov_27.421 19930   19900
NODE_55_length_30858_cov_27.421 17578   17613
NODE_55_length_30858_cov_27.421 16544   16578
NODE_55_length_30858_cov_27.421 19982   19932

Output:

NODE_55_length_30858_cov_27.421 19901   19951
NODE_55_length_30858_cov_27.421 19900   19930
NODE_55_length_30858_cov_27.421 17578   17613
NODE_55_length_30858_cov_27.421 16544   16578
NODE_55_length_30858_cov_27.421 19932   19982

[Kumar]

Thanks for helping..