Unix，按数字排序

Question

I want to sort my file by decreasing number of the second column, but not changing the place of the header. Currently it is like this in a csv file that has two columns like this:

Person     Age

Sarah      15

Robert     23

Emma       31

Bob        9

I want it to be like this:

Person     Age

Emma       31

Robert     23

Sarah      15

Bob        9

Please help a beginner in Unix.

Answer 1

Read http://man7.org/linux/man-pages/man1/sort.1.html to learn about the sort command on UNIX/Linux. That's how I and everyone else learned.

$ sort --key 2 --reverse --numeric-sort myfile > mysortedfile

The only problem is that this sorts the line with "Person Age" as if it had the numeric value of zero, and moves it to the last line of the file. You then have to move that line after sorting the file.

Answer 2

This will do the trick except remove blank lines.

sed -n '1 ! p' sortme.txt | sort --key 2 --reverse --numeric-sort | sed '1 i\\Person\\tAge' > sorted.txt

~Pat

Answer 3

To solve the "header in front" problem as outlined in Bills answer, you could use awk in addition to sort :

awk 'NR==1 {print ; next } { print | "sort -n -r -k2"  }' yourfile > sorted_file

• the condition NR==1 applies only for the first line: it is printed
• all other line are piped to the sort command (you could do the sorting in awk itself, but that would be a bit longer )