bash help

[Liy]

Hi there,

I cant figure it out how to solve a problem in unix, so a little help would be appritiated.

So, I want to delete a line (or swap the two column) only if the column3 is greater then column4. Its a tab delimited bed file. The expected output would be the same as the original file, but swapped coordinates if start is greater then stop. Maybe awk, sed?

x1 I 139154 139256 y1 +
x2 I 166268 166340 y2 +
x3 I 181135 181248 y3 +
x4 I 182597 182516 y4 +

Thanks a lot!
L

[rhinoceros]

awk '$4>$3{print}' file > output

Only prints lines where column4 > column3

awk '$3>$4{print $1" "$2" "$4" "$3" "$5}' file > output

Only print lines where column3 > column4, but print them as "column1<space>column2<space>column4<space>column3<space>column5"

[blakeoft]

awk '$4>$3{print}' file > output

Only prints lines where column4 > column3

awk '$3>$4{print $1" "$2" "$4" "$3" "$5}' file > output

Only print lines where column3 > column4, but print them as "column1<space>column2<space>column4<space>column3<space>column5"

I suggest doing both of these except using ">>" instead of ">" when sending to the output file and then sorting the output file if you need to.

[GenoMax]

Since Liy wants to keep the file otherwise unchanged, building on rhinoceros's code.

BEGIN   { RS = "\n"}{
        if ($4 > $3) {
        print $0
        }
        else    {
         #OFS = "\t" (uncomment if you need the output to be tab separated, your example above seems to have spaces)
        print $1, $2, $4, $3, $5, $6
        }
}

$ cat file | awk -f code_above.awk > save_to_a_new_file

[blakeoft]

You could also try:

awk '{if ($3>$4) {print $1,$2,$4,$3,$5,$6} else {print}}' file.bed > new_file.bed

[Liy]

Thank you very much, all of you!