Shell script: Check if data in columns `X` from two CSV files are matched - bash

Question

For example I have
a.txt:
1 21 34
1 22 21
2 32 76
2 12 76
...
b.txt:
1 99 73
1 32 27
2 55 76
2 76 12
...
Expected output:
$ ./some_script 1 a.txt b.txt
0 # matched
# compare data in #1 column of a.txt to data in #1 column of b.txt
# data: a.txt b.txt
# 1 1
# 1 1
# 2 2
# 2 2
$ ./some_script 2 a.txt b.txt
1 # not matched
$ ./some_script 3 a.txt b.txt
1 # not matched
where parameters 1, 2, and 3 are column numbers.
Let's say, the some_script just did comparison between data in the same column from files a.txt and b.txt.
I need some program written in either bash, sed, or awk (or another possible programs) to do this job.

Answer 1

I would use a combination of paste and awk to achieve that
#!/bin/bash
[ -z "$1" -o -z "$2" -o -z "$3" ] && echo "Not enough arguments" && exit 1
[ ! -f "$2" -o ! -f "$3" ] && echo "input file(s) don't exist" && exit 1
awk -v var="$1" '$var!=$(NF/2+var){flag=1;exit}
END{print flag;}' <(paste "$2" "$3")
Save the file as, say, compare.sh, make it an executable and then run it like
./compare.sh 3 a.txt b.txt

Answer 2

[ "$(cut -d' ' -f1 a.txt)" = "$(cut -d' ' -f1 b.txt)" ]; echo $?
Explanation:
[ "string1" = "string2" ] - The test command. If the string1 equals to the string2, it returns 0, else 1. See man test for another information.
cut -d' ' -f1 a.txt - cut the first column from the file a.txt.
-d' ' - set the field delimiter to the space.
-f1 - select only the field number 1. You can use a variable, instead of the number 1 in this case, like the num=1; [ "$(cut -d' ' -f$num a.txt)" = "$(cut -d' ' -f$num b.txt)" ]; echo $?.
echo $? - print the exit status of the last executed program.

Answer 3

Simple one line solution with bash and awk
#!/bin/bash
[ "$(awk -F' ' "{print \$$1}" "$2")" == "$(awk -F' ' "{print \$$1}" "$3")" ] && echo 0 || echo 1
Output
./script 1 a.txt b.txt
0
./script 2 a.txt b.txt
1
./script 3 a.txt b.txt
1

Answer 4

Here's a bash version using custom file descriptors and arrays:
#!/bin/bash
exec 3< "$2"
exec 4< "$3"
while read -ru3 -a a && read -ru4 -a b; do
[ "${a[$(($1 - 1))]}" != "${b[$(($1 - 1))]}" ] && exit 1
done
exit 0