huffman.ipynb

markov.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [],
   "source": [
    "import re\n",
    "from collections import defaultdict\n",
    "\n",
    "# https://github.com/CGLemon/chinese-chess-PGN\n",
    "with open(\"WXF-41743games.pgns\", encoding=\"utf8\") as f:\n",
    "    data = f.read()\n",
    "\n",
    "cc = defaultdict(lambda:defaultdict(int))\n",
    "\n",
    "for d in data.split(\"\\n\\n\"):\n",
    "    bd = [[None for j in range(10)] for i in range(9)]\n",
    "    for i, p in enumerate([(0, 0), (1, 0), (2, 0), (3, 0), (4, 0), (5, 0), (6, 0), (7, 0), (8, 0), (1, 2), (7, 2), (0, 3), (2, 3), (4, 3), (6, 3), (8, 3)]):\n",
    "        bd[p[0]][p[1]] = i\n",
    "    last_m = (-1, 0, 0)\n",
    "    for z in re.findall(r\"\\d+\\. (..)-(..) (..)-(..)\", d):\n",
    "        p1, p2, q1, q2 = tuple(map(lambda t: (ord(t[0]) - ord(\"A\"), int(t[1])), z))\n",
    "        assert bd[p1[0]][p1[1]] != None\n",
    "        c = bd[p1[0]][p1[1]]\n",
    "        m = (c, (p2[0] - p1[0] + 9) % 9, (p2[1] - p1[1] + 10) % 10)\n",
    "        bd[p2[0]][p2[1]] = bd[p1[0]][p1[1]]\n",
    "        bd[p1[0]][p1[1]] = None\n",
    "        cc[last_m][m] += 1\n",
    "        last_m = m\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "6.019194684029047\n"
     ]
    }
   ],
   "source": [
    "\n",
    "ts = 0\n",
    "tm = 0\n",
    "for m, cnt in cc.items():\n",
    "    freq = set((cnt[k], k) for k in cnt)\n",
    "    total_moves = sum(cnt.values())\n",
    "    total_bits = 0\n",
    "    while len(freq) > 1:\n",
    "        x = min(freq, key=lambda t: t[0])\n",
    "        freq.remove(x)\n",
    "        y = min(freq, key=lambda t: t[0])\n",
    "        freq.remove(y)\n",
    "        freq.add((x[0] + y[0], (x[1], y[1])))\n",
    "        total_bits += x[0] + y[0]\n",
    "    ts += total_bits\n",
    "    tm += total_moves\n",
    "print(ts / tm)\n"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.11.2"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}