test.csv

uniqcol.py

#!/usr/bin/env python

import sys
import re

eol = re.compile('\r?\n$')

def build_header_table(cols):
    return [[0 for col in cols] for col in cols]

def get_row(line):
    line = eol.sub('', line)
    return line.split("\t")

def populate_checklist(table):
    xs = []
    for j, t2 in enumerate(table):
        for k, _ in enumerate(t2):
            if j != k:
                xs.append([j,k])
    return xs


def main(handle):
    first = 1
    header = None
    table = None
    checklist = None
    removed = []

    for line in handle:
        row = get_row(line)
        if first:
            header = row
            table = build_header_table(header)
            checklist = populate_checklist(table)
            first = 0
            continue

        for j, k in checklist:
            if row[j] != row[k]:
                checklist.remove([j,k])
                checklist.remove([k,j])

    for j, k in checklist:
        checklist.remove([k,j])
        print("{}\t{}".format(j+1, k+1))

if __name__ == "__main__":
    handle = open(sys.argv[1], "rb") if len(sys.argv) >= 2 else sys.stdin
    main(handle)

./uniqcol.py <(csv2tsv test.csv)
2	4
2	5
4	5

Need to merge these pairs now to find common subsets 🤔