Scrabble Art Generator

In my humble opinion, art is incredibly personal and displaying a purchased canvas of "A Starry Night" - although beautiful - totally misses the mark. With this in mind, I acquired a used copy of Scrabble for some board game themed wall art! My goal: scrabbly connecting a bunch of adjectives to my name.

Instead of manually finding a hodgepodge of properly connected adjectives, I decided to automate this search. In a nutshell, this super naive algorithm boils down to a 2 dimensional regex solver that randomly samples adjectives from Wordnet.


import argparse
import random
import re
import numpy as np
from nltk.corpus import wordnet as wn

class Scrabbler():
    def __init__(self, my_name, board_size): = my_name
        self.board_size = board_size

        self.board = np.empty([board_size, board_size], dtype=object)
        self.board[:] = r'.?'

        self.used_words = []
        self.look_at_rows = False


    def read_clean_word_list(self):
        all_words = set()
        for adjective in wn.all_synsets(wn.ADJ):
            # Remove Roman Numerals
            if 'more than' in adjective.definition():
            for lemma in adjective.lemma_names():
                if lemma.isalpha() and len(lemma) < self.board_size:

        self.all_words = list(all_words) + []

    def place_name_on_board(self):
        starting_loc = (self.board_size - len( // 2
        end_loc = starting_loc + len(
        center_row = self.board_size // 2
        self.board[center_row][starting_loc:end_loc] = list(

    def try_to_place(line, word):
        if any(characters.isalpha() for characters in line):
            smushed_line = f'^{"".join(line)}$'
            smushed_line_regrex = re.compile(smushed_line)

                num_characters_in_line = sum(
                    characters.isalpha() for characters in line
                word_length = len(word)

                for position in range(0, len(line) - word_length):
                    new_line = [' '] * len(line)
                    new_line[position:position+word_length] = list(word)

                    total_in_common = sum(
                        1 if a == b else 0 for (a, b) in zip(line, new_line)

                    if total_in_common == num_characters_in_line:
                        revert_back_to_regex = [
                            '.?' if c == ' ' else c for c in new_line
                        return revert_back_to_regex

    def placeWord(self, board, word):
        self.temp_board = None
        copied_board = np.copy(board)

        for index, line in enumerate(board):
            letters_in_common = set(line).intersection(set(word))
            if len(letters_in_common) > 0:
                new_word = self.try_to_place(line, word)
                if new_word:
                    copied_board[index] = new_word
                    self.temp_board = copied_board

    def check_for_valid_lines(self):
        if self.temp_board is None:
            return False

        for temp_board_orientation in [self.temp_board, self.temp_board.T]:
            for row in temp_board_orientation:
                words = row.sum().replace('.?', ' ').split()
                for word in words:
                    if len(word) > 1:
                        if not word in self.all_words:
                            return False

        return True

    def add_valid_word_to_board(self):

        for word in self.all_words:
            if word in self.used_words:

            if self.look_at_rows:
                self.placeWord(self.board, word)
                self.placeWord(self.board.T, word)

            is_valid = self.check_for_valid_lines()
            if is_valid:
                if not self.look_at_rows:
                    self.temp_board = self.temp_board.T
                self.board[:, :] = self.temp_board
                return word

    def print_board(self):
        for row in self.board:
            clean_row = "".join(row).replace(".?", " ")

    def run(self):
        while True:
            valid_word = self.add_valid_word_to_board()
            print('.', end="", flush=True)
            if not valid_word:

            self.look_at_rows = ~self.look_at_rows


def check_wordnet_exists():
    except LookupError:
        print('Downloading wordnet, is this okay? (y/n)')
        if input() == 'y':
            import nltk

if __name__ == '__main__':
    parser = argparse.ArgumentParser()
    parser.add_argument('--name', nargs='*', help='Sets your name')
    parser.add_argument('--size', type=int, default=15,
                        help='Sets the size of the board')
    parser.add_argument('--iter', type=int, default=1,
                        help='Sets the number of boards generated')
    args = parser.parse_args()

    my_name = "".join(
    board_size = args.size
    number_of_iterations = args.iter

    if not my_name:
        print("Missing `--name <My Name>` argument")
    elif len(my_name) > board_size:
        board_size = len(my_name)


    print('Running Scrabbler\n')
    for _ in range(number_of_iterations):
        scrabbler = Scrabbler(my_name, board_size)


Recent Posts

Data StarFeld: A LSTM network beautifully smooshes the scripts from Star Trek: The Next Generation and Seinfeld together.

Plancky Why Does the Sun Shine: How does a bunch of hydrogen atoms minding their own business turn into a humongous light spewing fireball?

Random Purple Violet Bear Documentary: Art Clokey inspired my most rebellious teenage outburst.


Hi, I'm Chris and I'm a Data Scientist living in Canada with my wonderful wife and occasionally wonderful Shih Tzu, Bubba. I'm always on the lookout for cool new experiences, so feel free to send a message!

Professional Interests: Data Science, Sustainability

Personal Interests: Gaming, Biking, Epic Fantasy

Primary Disinterests: Pursuing a Second PhD, Multi-Origin Non-Esoteric Coffee

Cookie Interests: Chocolate Chip, Crackle, Gingersnap