CODEWITHSUNDEEP
awksed
KLM117
KLM117

Reputation: 467

Subtracting from values ending with specific digits?

I have a .bed (.tsv) file that looks like this:

chr1    0   100000
chr1    100000  200000
chr1    200000  300000
chr1    300000  425234

I want to perform the operation -1 from only values in column 3 that end in "000", using sed or awk so that the output looks like:

chr1    0   99999
chr1    100000  199999
chr1    200000  299999
chr1    300000  425234

Embarassingly enough, the best I've come up with is:


awk {sub(/000$/,"999",$3); print $1,$2,$3}' oldfile > newfile

which simply substituites the last 3 digits for 999, rather than actually subtracting. Any help is appreciated is always!

Upvotes: 0

Views: 49

Answers (2)

Daweo
Daweo

Reputation: 36390

I would use GNU AWK for this as follows, let file.txt content be

chr1    0   100000
chr1    100000  200000
chr1    200000  300000
chr1    300000  425234

then

awk 'BEGIN{OFS="\t"}($3%1000==0){$3-=1}{print}' file.txt

output

chr1    0   99999
chr1    100000  199999
chr1    200000  299999
chr1    300000  425234

Explanation: Use tab (\t) as output field separator (OFS). If remainder from diving $3 by 1000 is zero (i.e. $3 is multiply of 1000) then subtract 1 from $3, for each line print.

(tested in gawk 4.2.1)

Upvotes: 0

tripleee
tripleee

Reputation: 189377

Awk can easily perform arithmetic, too.

awk 'BEGIN{FS=OFS="\t"} $3 ~ /000$/ {$3 -= 1}1' oldfile > newfile

This is assuming all the lines in your file always have three fields and that you want to print all the lines.

sed has no idea about even the simplest arithmetic so it's not particularly suitable for this.

Upvotes: 5

Related Questions