check_mcts.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [],
   "source": [
    "import random"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [],
   "source": [
    "import model\r\n",
    "import selfplay\r\n",
    "import game\r\n",
    "import config"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [],
   "source": [
    "from visualise import vis"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [],
   "source": [
    "def get_probs_with_log(self, g: game.GameState):\r\n",
    "    \"runs N simulations and returns visit probabilities at root\"\r\n",
    "    root = selfplay.Node(g)\r\n",
    "    for i in range(config.num_simulate):\r\n",
    "        self.simulate(root)\r",
    "   # self.print_tree(root)\r\n",
    "    return [\r\n",
    "        root.children[i].visit/(root.visit-1) if i in root.children else 0.0\r\n",
    "        for i in range(config.num_actions)\r\n",
    "    ]\r\n",
    "\r\n",
    "def my_print_tree(self, n: selfplay.Node, depth=0, all=True):\r\n",
    "    if n.visit:\r\n",
    "        a = \">>\"\r\n",
    "    else:\r\n",
    "        a = \"--\"\r\n",
    "    if depth < 2 and (all or n.visit):\r\n",
    "        print(depth*\" \" + f\"boardosition: {n.state}, prob: {n.prob:0.2}, value: {n.value:0.2},\"\r\n",
    "                            f\" visit: {n.visit}, value_sum: {n.value_sum:0.2}\")\r\n",
    "        for ch in n.children.values():\r\n",
    "            self.print_tree(ch, depth+1)\r\n",
    "\r\n",
    "selfplay.MCTS.print_tree = my_print_tree\r\n",
    "selfplay.MCTS.get_probs = get_probs_with_log"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [],
   "source": [
    "net = model.Model()\n",
    "net.load()\r\n",
    "mcts = selfplay.MCTS(net)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [],
   "source": [
    "import chess"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {},
   "outputs": [],
   "source": [
    "g = game.GameState()\n",
    "g.board = chess.Board('3k3q/8/8/8/8/3K4/1b6/8 w - - 56 282')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {},
   "outputs": [],
   "source": [
    "p = mcts.get_probs(g)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "g=[[0 0 0 0 0 0 0 0 0]] probs=[0.05263157894736842, 0.10526315789473684, 0.2631578947368421, 0.05263157894736842, 0.05263157894736842, 0.15789473684210525, 0.05263157894736842, 0.05263157894736842, 0.21052631578947367] sum(probs)=0.9999999999999998\n",
      ". . .\n",
      ". . .\n",
      ". . .\n",
      "\n",
      "action=8\n",
      "g=[[0 0 0 0 0 0 0 0 1]] probs=[0.0, 0.05263157894736842, 0.10526315789473684, 0.10526315789473684, 0.0, 0.7368421052631579, 0.0, 0.0, 0.0] sum(probs)=1.0\n",
      ". . .\n",
      ". . .\n",
      ". . X\n",
      "\n",
      "action=5\n",
      "g=[[ 0  0  0  0  0 -1  0  0  1]] probs=[0.0, 0.0, 0.6842105263157895, 0.15789473684210525, 0.0, 0.0, 0.10526315789473684, 0.05263157894736842, 0.0] sum(probs)=1.0\n",
      ". . .\n",
      ". . O\n",
      ". . X\n",
      "\n",
      "action=2\n",
      "g=[[ 0  0  1  0  0 -1  0  0  1]] probs=[0.0, 0.0, 0.0, 0.3157894736842105, 0.0, 0.0, 0.5789473684210527, 0.10526315789473684, 0.0] sum(probs)=1.0\n",
      ". . X\n",
      ". . O\n",
      ". . X\n",
      "\n",
      "action=6\n",
      "g=[[ 0  0  1  0  0 -1 -1  0  1]] probs=[0.10526315789473684, 0.10526315789473684, 0.0, 0.5789473684210527, 0.10526315789473684, 0.0, 0.0, 0.10526315789473684, 0.0] sum(probs)=1.0\n",
      ". . X\n",
      ". . O\n",
      "O . X\n",
      "\n",
      "action=0\n",
      "g=[[ 1  0  1  0  0 -1 -1  0  1]] probs=[0.0, 0.3157894736842105, 0.0, 0.5263157894736842, 0.05263157894736842, 0.0, 0.0, 0.10526315789473684, 0.0] sum(probs)=0.9999999999999999\n",
      "X . X\n",
      ". . O\n",
      "O . X\n",
      "\n",
      "action=3\n",
      "g=[[ 1  0  1 -1  0 -1 -1  0  1]] probs=[0.0, 0.5263157894736842, 0.0, 0.0, 0.3684210526315789, 0.0, 0.0, 0.10526315789473684, 0.0] sum(probs)=0.9999999999999999\n",
      "X . X\n",
      "O . O\n",
      "O . X\n",
      "\n",
      "action=4\n",
      "g=[[ 1  0  1 -1  1 -1 -1  0  1]] g.terminated()=True\n",
      "X . X\n",
      "O X O\n",
      "O . X\n",
      "\n"
     ]
    }
   ],
   "source": [
    "g = game.GameState()\r\n",
    "history = []\r\n",
    "move_history = []\r\n",
    "\r\n",
    "while not g.terminated():\r\n",
    "    probs = mcts.get_probs(g)\r\n",
    "    print(f\"g={g.to_image()} {probs=} {sum(probs)=}\")\r\n",
    "    print(vis(str(g.to_image())))\r\n",
    "    history.append([g, probs, 0])\r\n",
    "    # choose action acc to probs\r\n",
    "    action = random.choices(list(range(config.num_actions)), probs)[0]\r\n",
    "    print(f\"{action=}\")\r\n",
    "    move_history.append(action)\r\n",
    "    g = g.next_state(action)\r\n",
    "\r\n",
    "print(f\"g={g.to_image()} {g.terminated()=}\")\r\n",
    "print(vis(str(g.to_image())))\r\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "# log game outcome\r\n",
    "winner = g.winner()\r\n",
    "# print(move_history, winner)d\r\n",
    "# fill in the value function\r\n",
    "if winner != 0:\r\n",
    "    for state in history:\r\n",
    "        g = state[0]\r\n",
    "        if g.player() == winner:\r\n",
    "            state[2] = -1\r\n",
    "        else:\r\n",
    "            state[2] = 1"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3 (ipykernel)",
   "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": "3.11.4"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}