pokemon-color-search/nearest.py

#!/usr/bin/env python3
import csv
import math
import random
from argparse import ArgumentParser

from convert import rgb_to_cieluv


def norm(r, g, b):
    return math.sqrt(r * r + g * g + b * b)


parser = ArgumentParser()
parser.add_argument(
    "color", nargs="?", default=None, help="Target color, randomized if not provided"
)
parser.add_argument(
    "-n", "--number", default=1, type=int, help="Number of Pokemon to find"
)
parser.add_argument(
    "-c", "--closeness", default=2, type=int, help="Closeness coefficient"
)
parser.add_argument("-d", "--database", default="database.csv", help="Database file")
parser.add_argument("-x", "--exclude-x", action="store_true", help="Exclude X")
parser.add_argument("-z", "--normalize", action="store_true", help="Normalize q and Y")
parser.add_argument(
    "-l",
    "--convert-luv",
    action="store_true",
    help="Convert input color to CIELUV before calculation",
)
parser.add_argument("-v", "--verbose", action="store_true", help="Print raw scores")
args = parser.parse_args()

if args.number <= 0:
    raise ValueError("Must request a number greater than 0")

if args.color is not None:
    cleaned_color = args.color.strip("#")
    if len(cleaned_color) != 6:
        raise ValueError("Color must be a 6 digit hex")
    color = (
        int(cleaned_color[0:2], base=16),
        int(cleaned_color[2:4], base=16),
        int(cleaned_color[4:6], base=16),
    )
else:
    color = tuple(int(random.random() * 255) for _ in range(3))
    print(f"Generated random color: #{''.join(hex(c)[2:] for c in color)} / {color}")

if args.convert_luv:
    color = rgb_to_cieluv(*color)

yfactor = args.closeness
if args.normalize:
    yfactor /= norm(*color)

results = []
with open(args.database) as infile:
    for name, x, *y in csv.reader(infile, delimiter=",", quotechar="'"):
        xval = 0 if args.exclude_x else float(x)
        yvec = [float(y_c) for y_c in y]
        yval = sum(y_c * c for y_c, c in zip(yvec, color))
        if args.normalize:
            yval /= norm(*yvec)
        score = xval - yfactor * yval
        results.append((score, name))

if args.number > 1:
    for i, (score, name) in enumerate(sorted(results)[: args.number]):
        print(f"{i+1}:\t{name}")
        if args.verbose:
            print(f"\t\t\t{score=}")
else:
    best_score, best_name = min(results)
    print(f"Best match: {best_name}")
    if args.verbose:
        print(f"{score=}")