CODEWITHSUNDEEP
bashshell
user2519184
user2519184

Reputation: 67

count numbers in file using BASH script

i have file with the following data records:

0.005
0.027
0.352
1.005
0.117
0.752

I'm trying to read the file using bash line by line and count the numbers into 3 groups:

if [($num < 0.2)] then min++
if [($num > 0.2) && ($num < 1)] then med++
if [($num > 1)] then max++

At the end I need to print out the 3 variable to file which count all the records. Any idea how to read the file in bash and manipulating it using float in bash?

Upvotes: 0

Views: 373

Answers (2)

tshiono
tshiono

Reputation: 22012

If you want to do it with bash by all means, possible workaround will be to multiply the numbers by 1,000 (or whatever appropriate) and treat them as integers. How about:

while read -r x; do
    num=$((10#$(printf "%.3f" "$x" | tr -d ".")))
#   echo "$num"
    (( num < 200 )) && (( min++ ))
    (( num > 200 && num < 1000 )) && (( med++ ))
    (( num > 1000 )) && (( max++ ))
done < "data.txt"
printf "min=%d med=%d max=%d\n" "$min" "$med" "$max"

As @dibery comments, you need to decide how to handle exact 0.2 and 1.

Upvotes: 0

dibery
dibery

Reputation: 3460

Unfortunately, bash doesn't like floating numbers very much; however, you may use other text processing tools like awk:

awk '{if($0 < 0.2)min++; else if($0 < 1)med++; else max++;} END {print min, med, max;}' file.txt

Also please note that your algorithm doesn't handle numbers exactly at 0.2 and 1.

Upvotes: 1

Related Questions