使用bash比较txt列中的浮动数据

Question

I have thousands of files containg a column of non integers data. I would like to build a script that iterates through all the folders, checks if the values in the file are equal/different from 0.000000.

The column of data are like the following one:

-0.572650
-0.002281
-0.080300
-0.803684
-0.498033
-0.206755
-0.182819
-0.222126
-0.116538
 0.720913
 0.707343
 1.134845
-0.462461
-0.382780
-0.378505
-0.436588
-0.464654
-0.207534
-0.231861
-0.155786
-0.196779
 0.977137
 1.024751
 1.037147

I have tried something like:

#!/bin/bash


a=$(awk '{print $1}' tmp.txt)


if (( $(echo "$a == 0.000000" |bc -l) ));
 then
 echo "MULLIKEN MISSING"
elif (( $(echo "$a != 0.000000" |bc -l) )); 
 then
echo "OK"
fi

but it is does not work in manner I would like to.

I would like the script to be able to detect if the data in column are all equal to 0.000000 or not.

Answer 1

I would stick just with awk: this will be quite efficient as you only need to spawn one process to handle the whole file, as opposed to calling bc for every number.

if awk '$1 == 0 {exit 0} END {exit 1}' file; then
    echo "has a zero"
else
    echo "no zeroes"
fi

---

EDIT: I have a logic error: the awk END block is executed when the exit command is called, so the above will always exit 1 . Use this instead:

if awk 'BEGIN {rc = 1} $1 == 0 {rc = 0; exit} END {exit rc}' file; then
    echo "has a zero"
else
    echo "no zeroes"
fi

or, exit 1 when a zero is found. This simplifies the awk command:

if awk '$1 == 0 {exit 1}' file; then
    echo "no zeroes"
else
    echo "has a zero"
fi

---

If you want to stick with plain bash, you can read each number from the file, and remove the dot so that you only have an integer (no floating point in bash)

zero=false
while read -r value; do
    if (( 10#${value//./} == 0 )); then
        zero=true
        break
    fi
done < file

$zero && echo "has a zero" || echo "no zeroes"

The 10# is to force base-10 interpretation, so that bash does not puke on invalid octal numbers like 080000.

Answer 2

Hi this code is quite similar to yours and it should work.

#!/bin/bash

a=$(awk '{print $1}' tmp.txt)

for i in ${a[@]}; do
    result=$( echo $i == 0.0 | bc -l)
    if [[ $result == "0" ]];
    then
        echo "OK"
    else
        echo "NOT OK"
    fi
done

Note that reading with AWK in this way, then using variable as array is not the safest method.