feat(unitproc/bkt-decimate):Add Bash function to decimate stdin

[BK-2020-03.git] / unitproc / bkt-decimate

#!/bin/bash

decimate() {
    # Desc: Randomly remove 10% of stdin lines
    # Depends: GNU Coreutils 8.32 (shuf)
    # Version: 0.0.1

    # Read lines
    mapfile -t lines;

    # Calc lines to keep, lk
    lc="${#lines[@]}";
    lk="$((lc * 900 / 1000))";

    printf "%s\n" "${lines[@]}" | \
        nl -w1 -s' ' | \
        shuf | \
        head -n "$lk" | \
        sort -n -k1,1 | \
        cut -d' ' -f2- ;
}; # randomly eliminate 10% of lines

echo "$WARNING:This is a Bash function definition."