Taxi agent and Eight Puzzle agent.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Building a Problem Solving Agent"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "In this assignment, I have built an agent that solves problems by searching through a graph (state space) for the solution.<br>\n",
    "I have used four searching algorithms for this purpose.\n",
    "* Breadth First Search\n",
    "* Depth First Search\n",
    "* Greedy Best First Search\n",
    "* A* Search"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Importing required libraries\n",
    "import pandas as pd\n",
    "import numpy as np\n",
    "import networkx as nx\n",
    "import matplotlib.pyplot as plt\n",
    "%matplotlib inline"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "process = lambda column: column.str.replace(' ', '').str.capitalize()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Reading in the dataset containing cities and pairwise distance"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>City_1</th>\n",
       "      <th>City_2</th>\n",
       "      <th>travel_dist</th>\n",
       "      <th>straight_dist</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>Agartala</td>\n",
       "      <td>Aizawl</td>\n",
       "      <td>342</td>\n",
       "      <td>146</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>Aizawl</td>\n",
       "      <td>Imphal</td>\n",
       "      <td>400</td>\n",
       "      <td>173</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>Amaravathi</td>\n",
       "      <td>Bangalore</td>\n",
       "      <td>663</td>\n",
       "      <td>498</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>Amaravathi</td>\n",
       "      <td>Chennai</td>\n",
       "      <td>448</td>\n",
       "      <td>391</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>Amaravathi</td>\n",
       "      <td>Bhubaneswar</td>\n",
       "      <td>819</td>\n",
       "      <td>715</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "       City_1       City_2  travel_dist  straight_dist\n",
       "0    Agartala       Aizawl          342            146\n",
       "1      Aizawl       Imphal          400            173\n",
       "2  Amaravathi    Bangalore          663            498\n",
       "3  Amaravathi      Chennai          448            391\n",
       "4  Amaravathi  Bhubaneswar          819            715"
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data_df = pd.read_csv(\"Cities_Distances.csv\", header=None)\n",
    "data_df.columns = [\"City_1\", \"City_2\", \"travel_dist\", \"straight_dist\"]\n",
    "data_df.City_1 = process(data_df.City_1)\n",
    "data_df.City_2 = process(data_df.City_2)\n",
    "data_df.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Reading the dataset containing the heuristic distances from other cities to Bangalore"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "bangalore_heuristics = pd.read_csv('heuristics_bangalore.csv', header=0)\n",
    "bangalore_heuristics.columns = ['City_1', 'City_2', 'Distance']\n",
    "bangalore_heuristics.City_1 = process(bangalore_heuristics.City_1)\n",
    "bangalore_heuristics.City_2 = process(bangalore_heuristics.City_2)\n",
    "bangalore_heuristics.set_index(['City_1', 'City_2'], inplace=True)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Drawing the cities and the edges connecting them"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "c:\\users\\sravan vinakota\\appdata\\local\\programs\\python\\python36\\lib\\site-packages\\networkx\\drawing\\nx_pylab.py:611: MatplotlibDeprecationWarning: isinstance(..., numbers.Number)\n",
      "  if cb.is_numlike(alpha):\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 691.2x518.4 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "map_ = nx.Graph()\n",
    "\n",
    "for row in data_df.set_index(['City_1', 'City_2']).iterrows():\n",
    "    map_.add_edge(*row[0], **row[1])\n",
    "\n",
    "fig = plt.figure(figsize=(6.4 * 1.5, 4.8 * 1.5))\n",
    "nx.draw(map_, with_labels=True, pos=nx.kamada_kawai_layout(map_), node_color='blue')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "###### Making a class(problem solving agent) that contains the function for heuristics and other helper functions"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Class for the taxi agent"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "class Taxiagent:\n",
    "    def __init__(self, init_state, final_state, space, heuristics):\n",
    "#         Parameters:\n",
    "#             init_state: the initial state of the agent\n",
    "#             final_state: the goal state of the agent\n",
    "#             space: the space of the taxi agent\n",
    "#             heuristics: the heuristics graph or function\n",
    "#         Returns:\n",
    "#             the taxiagent object\n",
    "\n",
    "        self.init_state = init_state\n",
    "        self.final_state = final_state\n",
    "        self.space = space\n",
    "        self.heuristics = heuristics\n",
    "    \n",
    "    def heuristic_fn(self, n):\n",
    "        return self.heuristics['Distance'][(n, self.final_state)]\n",
    "    \n",
    "    def g_n(self, n1, n2):\n",
    "        for edge in self.space.edges(data=True):\n",
    "            if n1 in edge and n2 in edge:\n",
    "                return edge[-1]['straight_dist']\n",
    "    \n",
    "    def action_fn(self, current_state, with_heuristics=False):\n",
    "#         Parameters:\n",
    "#             current_state: the current_state of the problem\n",
    "#         Returns:\n",
    "#             the list of neighbors of the current state\n",
    "        \n",
    "        neighbors = list(nx.Graph.neighbors(self.space, current_state))\n",
    "#         print(f'Neighbors: {neighbors}')\n",
    "        if with_heuristics:\n",
    "            h_n = [self.heuristic_fn(n) for n in neighbors]\n",
    "            return list(zip(neighbors, h_n))\n",
    "        else:\n",
    "            return neighbors"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Class for the Eight Puzzle agent"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [],
   "source": [
    "class EightPuzzle:\n",
    "    def __init__(self, init_state, final_state, heuristics=None, size=3):\n",
    "#         Parameters:\n",
    "#             init_state: the initial state of the agent\n",
    "#             final_state: the goal state of the agent\n",
    "#             heuristics: the heuristics graph or function\n",
    "#         Returns:\n",
    "#             the taxiagent object\n",
    "        self.init_state = self.MTL(np.array(init_state))\n",
    "        self.final_state = self.MTL(np.array(final_state))\n",
    "        self.heuristics = heuristics\n",
    "        self.size = size\n",
    "\n",
    "    def heuristic_fn(self, n):\n",
    "        reshaped_final = self.LTM(self.final_state)\n",
    "        reshaped_curr = self.LTM(n)\n",
    "        \n",
    "        total = 0\n",
    "        for i in range(self.size ** 2):\n",
    "            curr_pos = (np.where(reshaped_curr == i)[0][0], np.where(reshaped_curr == i)[1][0])\n",
    "            final_pos = (np.where(reshaped_final == i)[0][0], np.where(reshaped_final == i)[1][0])\n",
    "            \n",
    "            total += abs(curr_pos[0] - final_pos[0]) + abs(curr_pos[1] - final_pos[1])\n",
    "        \n",
    "        return total\n",
    "    \n",
    "    def g_n(self, n1, n2):\n",
    "        return 1 # edge cost is assumed to be 1 in this case\n",
    "    \n",
    "    def action_fn(self, current_state, with_heuristics=False):\n",
    "#         Parameters:\n",
    "#             current_state: the current_state of the problem\n",
    "#         Returns:\n",
    "#             the list of neighbors of the current state\n",
    "        reshaped_curr = self.LTM(current_state)\n",
    "        pos_0 = (np.where(reshaped_curr == 0)[0][0], np.where(reshaped_curr == 0)[1][0])\n",
    "        \n",
    "        neighbors = []\n",
    "        # check position of empty and apply up, down, left, right\n",
    "        if pos_0[0] != 0: # going up\n",
    "            n = np.copy(reshaped_curr)\n",
    "            temp = n[pos_0[0] - 1][pos_0[1]]\n",
    "            n[pos_0[0] - 1][pos_0[1]] = 0\n",
    "            n[pos_0[0]][pos_0[1]] = temp\n",
    "            neighbors.append(self.MTL(n))\n",
    "        \n",
    "        if pos_0[0] != self.size - 1: # going down\n",
    "            n = np.copy(reshaped_curr)\n",
    "            temp = n[pos_0[0] + 1][pos_0[1]]\n",
    "            n[pos_0[0] + 1][pos_0[1]] = 0\n",
    "            n[pos_0[0]][pos_0[1]] = temp\n",
    "            neighbors.append(self.MTL(n))\n",
    "        \n",
    "        if pos_0[1] != 0: # going left\n",
    "            n = np.copy(reshaped_curr)\n",
    "            temp = n[pos_0[0]][pos_0[1] - 1]\n",
    "            n[pos_0[0]][pos_0[1] - 1] = 0\n",
    "            n[pos_0[0]][pos_0[1]] = temp\n",
    "            neighbors.append(self.MTL(n))\n",
    "        \n",
    "        if pos_0[1] != self.size - 1: # going right\n",
    "            n = np.copy(reshaped_curr)\n",
    "            temp = n[pos_0[0]][pos_0[1] + 1]\n",
    "            n[pos_0[0]][pos_0[1] + 1] = 0\n",
    "            n[pos_0[0]][pos_0[1]] = temp\n",
    "            neighbors.append(self.MTL(n))\n",
    "        \n",
    "        if with_heuristics:\n",
    "            h_s = [self.heuristic_fn(i) for i in neighbors]\n",
    "            return list(zip(neighbors, h_s))\n",
    "        else:\n",
    "            return neighbors\n",
    "    \n",
    "    def LTM(self, state):\n",
    "        return np.reshape(list(map(int, state)), (self.size, self.size))\n",
    "    \n",
    "    def MTL(self, state):\n",
    "        return ''.join(map(str, state.flatten()))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "##### The four search algortihms"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [],
   "source": [
    "def BFS(agent):\n",
    "#     Parameters:\n",
    "#         agent: the problem solving agent\n",
    "#     Returns:\n",
    "#         path: the sequence of actions\n",
    "    \n",
    "    parent = None # memory of parent\n",
    "    frontier = [(agent.init_state, parent)] # discovered but not yet explored\n",
    "    explored = [] # explored\n",
    "    found = False # flag to determine if path exists or not\n",
    "    \n",
    "    while frontier:\n",
    "        current, parent = frontier.pop(0)\n",
    "        explored.append((current, parent))\n",
    "        if current == agent.final_state:\n",
    "            found = True # set the flag to True if the goal state has been found\n",
    "            break\n",
    "        \n",
    "        # function to filter out neighbors that have already been explored\n",
    "        check_not_visited = lambda x: x not in list(map(lambda x: x[0], explored)) and x not in list(map(lambda x: x[0], frontier))\n",
    "        new_neighbors = agent.action_fn(current) # get the neighbors of the current state\n",
    "        non_visited_neighbors = list(filter(check_not_visited, new_neighbors)) # apply the check_not_visited function on the set of new neighbors\n",
    "        new_nodes = list(zip(non_visited_neighbors, [current] * len(new_neighbors))) # group the filtered out neighbors with their parent to append to the frontier\n",
    "        frontier.extend(new_nodes)\n",
    "    \n",
    "    if found:\n",
    "        path = []\n",
    "        state = agent.final_state\n",
    "        \n",
    "        while state: # loop through the explored set to get the path to the final state\n",
    "            path.append(state)\n",
    "            state = tuple(filter(lambda x: x[0] == state, explored))[0][1]\n",
    "        \n",
    "        return True, list(reversed(path))\n",
    "    return False, None"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [],
   "source": [
    "def DFS(agent):\n",
    "#     Parameters:\n",
    "#         agent: the problem solving agent\n",
    "#     Returns:\n",
    "#         path: the sequence of actions\n",
    "    \n",
    "    parent = None # memory of parent\n",
    "    frontier = [(agent.init_state, parent)] # discovered but not yet explored\n",
    "    explored = [] # explored\n",
    "    found = False # flag to determine if path exists or not\n",
    "    \n",
    "    while frontier:\n",
    "        current, parent = frontier.pop(-1)\n",
    "        explored.append((current, parent))\n",
    "        if current == agent.final_state:\n",
    "            found = True # set the flag to True if the goal state has been found\n",
    "            break\n",
    "        \n",
    "        # function to filter out neighbors that have already been explored\n",
    "        check_not_visited = lambda x: x not in list(map(lambda x: x[0], explored)) and x not in list(map(lambda x: x[0], frontier))\n",
    "        new_neighbors = agent.action_fn(current) # get the neighbors of the current state\n",
    "        non_visited_neighbors = list(filter(check_not_visited, new_neighbors)) # apply the check_not_visited function on the set of new neighbors\n",
    "        new_nodes = list(zip(non_visited_neighbors, [current] * len(new_neighbors))) # group the filtered out neighbors with their parent to append to the frontier\n",
    "        frontier.extend(new_nodes)\n",
    "    \n",
    "    if found:\n",
    "        path = []\n",
    "        state = agent.final_state\n",
    "        \n",
    "        while state: # loop through the explored set to get the path to the final state\n",
    "            path.append(state)\n",
    "            state = tuple(filter(lambda x: x[0] == state, explored))[0][1]\n",
    "        \n",
    "        return True, list(reversed(path))\n",
    "    return False, None"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [],
   "source": [
    "def GBFS(agent):\n",
    "#     Parameters:\n",
    "#         agent: the problem solving agent\n",
    "#     Returns:\n",
    "#         path: the sequence of actions\n",
    "    \n",
    "    initial_cost = agent.heuristic_fn(agent.init_state)\n",
    "    parent = None # memory of parent\n",
    "    frontier = [(agent.init_state, parent, initial_cost)] # discovered but not yet explored\n",
    "    explored = [] # explored\n",
    "    found = False # flag to determine if path found or not\n",
    "    \n",
    "    while frontier:\n",
    "        frontier = sorted(frontier, key=lambda x: x[-1])\n",
    "        current, parent, cost = frontier.pop(0)\n",
    "        explored.append((current, parent, cost))\n",
    "#         print(f'Explored: {explored}')\n",
    "        if current == agent.final_state:\n",
    "            found = True # set the flag to True if the goal state has been found\n",
    "            break\n",
    "        \n",
    "        # function to filter out neighbors that have already been explored\n",
    "        check_not_visited = lambda x: x[0] not in list(map(lambda x: x[0], explored)) and x[0] not in list(map(lambda x: x[0], frontier))\n",
    "        new_neighbors = agent.action_fn(current, True) # get the neighbors of the current state\n",
    "        non_visited_neighbors = list(filter(check_not_visited, new_neighbors)) # apply the check_not_visited function on the set of new neighbors\n",
    "        new_nodes = list(zip(map(lambda x: x[0], non_visited_neighbors), [current] * len(new_neighbors), map(lambda x: x[-1], non_visited_neighbors)))\n",
    "        # group the filtered out neighbors with their parent to append to the frontier\n",
    "#         print(f'New Nodes: {new_nodes}')\n",
    "        frontier.extend(new_nodes)\n",
    "    \n",
    "    if found:\n",
    "        path = []\n",
    "        state = agent.final_state\n",
    "        \n",
    "        while state: # loop through the explored set to get the path to the final state\n",
    "            path.append(state)\n",
    "            state = tuple(filter(lambda x: x[0] == state, explored))[0][1]\n",
    "        \n",
    "        return True, list(reversed(path))\n",
    "    return False, None"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [],
   "source": [
    "def A_star(agent):\n",
    "#     Parameters:\n",
    "#         agent: the problem solving agent\n",
    "#     Returns:\n",
    "#         path: the sequence of actions\n",
    "    \n",
    "    initial_cost = agent.heuristic_fn(agent.init_state)\n",
    "    parent = None # memory of parent\n",
    "    frontier = [(agent.init_state, parent, initial_cost)] # discovered but not yet explored\n",
    "    explored = [] # explored\n",
    "    found = False # flag to determine if path found or not\n",
    "    \n",
    "    while frontier:\n",
    "        frontier = sorted(frontier, key=lambda x: x[-1])\n",
    "        current, parent, cost = frontier.pop(0)\n",
    "        explored.append((current, parent, cost))\n",
    "#         print(f'Explored: {explored}')\n",
    "        if current == agent.final_state:\n",
    "            found = True # set the flag to True if the goal state has been found\n",
    "            break\n",
    "        \n",
    "        # function to filter out neighbors that have already been explored\n",
    "        check_not_visited = lambda x: x[0] not in list(map(lambda x: x[0], explored))\n",
    "        new_neighbors = agent.action_fn(current, True) # get the neighbors of the current state\n",
    "        non_visited_neighbors = list(filter(check_not_visited, new_neighbors)) # apply the check_not_visited function on the set of new neighbors\n",
    "        new_nodes = list(zip(map(lambda x: x[0], non_visited_neighbors), [current] * len(new_neighbors), map(lambda x: cost - agent.heuristic_fn(current) + agent.g_n(current, x[0]) + x[-1], non_visited_neighbors)))\n",
    "        # group the filtered out neighbors with their parent to append to the frontier\n",
    "#         print(f'New Nodes: {new_nodes}')\n",
    "        frontier.extend(new_nodes)\n",
    "    \n",
    "    if found:\n",
    "        path = []\n",
    "        state = agent.final_state\n",
    "        \n",
    "        while state:\n",
    "            path.append(state)\n",
    "            state = tuple(filter(lambda x: x[0] == state, explored))[0][1]\n",
    "        \n",
    "        return True, list(reversed(path))\n",
    "    return False, None"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Enter source city (Choose from the graph displayed above) (please mind case and spelling): Gangtok\n",
      "Goal city is taken to be Bangalore by default.\n",
      "Please select algorithm:\n",
      "1. BFS\n",
      "2. DFS\n",
      "3. GBFS\n",
      "4. A*\n",
      "\n",
      "(Input an integer):4\n",
      "A_star: Gangtok -> Kolkata -> Bhubaneswar -> Amaravathi -> Bangalore\n"
     ]
    }
   ],
   "source": [
    "algo_mapping = {\n",
    "    1: 'BFS',\n",
    "    2: 'DFS',\n",
    "    3: 'GBFS',\n",
    "    4: 'A_star'\n",
    "}\n",
    "# mapping from integer to algorithm\n",
    "\n",
    "source = input(\"Enter source city (Choose from the graph displayed above) (please mind case and spelling): \")\n",
    "print(\"Goal city is taken to be Bangalore by default.\")\n",
    "goal = 'Bangalore'# input(\"Enter destination city (please mind spelling): \") # Goal is now Bangalore by default\n",
    "algo = int(input(\"\"\"Please select algorithm:\n",
    "1. BFS\n",
    "2. DFS\n",
    "3. GBFS\n",
    "4. A*\\n\n",
    "(Input an integer):\"\"\"))\n",
    "\n",
    "taxi = Taxiagent(source.strip(), goal.strip(), map_, bangalore_heuristics) # making the taxi agent\n",
    "\n",
    "if algo in range(1, 5):\n",
    "    exec('success_status, path = ' + str(algo_mapping[algo]) + '(taxi)')\n",
    "else:\n",
    "    print(\"Check your input.\")\n",
    "\n",
    "try:\n",
    "    if success_status:\n",
    "        print(f\"{algo_mapping[algo]}: \", end=\"\") # printing the path\n",
    "        print(*path, sep=\" -> \")\n",
    "    else:\n",
    "        print(\"No path found.\")\n",
    "except:\n",
    "    pass"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {
    "scrolled": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Enter elements for initial state:\n",
      "Enter row 0: 0 1 2\n",
      "Enter row 1: 3 4 5\n",
      "Enter row 2: 6 7 8\n",
      "\n",
      "Enter elements for goal state:\n",
      "Enter row 0: 1 2 3\n",
      "Enter row 1: 4 5 6\n",
      "Enter row 2: 7 8 0\n",
      "Please select algorithm:\n",
      "1. BFS\n",
      "2. DFS\n",
      "3. GBFS\n",
      "4. A*\n",
      "\n",
      "(Input an integer):4\n",
      "A_star: \n",
      "[[0 1 2]\n",
      " [3 4 5]\n",
      " [6 7 8]]\n",
      "\n",
      "->[[1 0 2]\n",
      " [3 4 5]\n",
      " [6 7 8]]\n",
      "\n",
      "->[[1 4 2]\n",
      " [3 0 5]\n",
      " [6 7 8]]\n",
      "\n",
      "->[[1 4 2]\n",
      " [0 3 5]\n",
      " [6 7 8]]\n",
      "\n",
      "->[[1 4 2]\n",
      " [6 3 5]\n",
      " [0 7 8]]\n",
      "\n",
      "->[[1 4 2]\n",
      " [6 3 5]\n",
      " [7 0 8]]\n",
      "\n",
      "->[[1 4 2]\n",
      " [6 3 5]\n",
      " [7 8 0]]\n",
      "\n",
      "->[[1 4 2]\n",
      " [6 3 0]\n",
      " [7 8 5]]\n",
      "\n",
      "->[[1 4 2]\n",
      " [6 0 3]\n",
      " [7 8 5]]\n",
      "\n",
      "->[[1 4 2]\n",
      " [0 6 3]\n",
      " [7 8 5]]\n",
      "\n",
      "->[[1 4 2]\n",
      " [7 6 3]\n",
      " [0 8 5]]\n",
      "\n",
      "->[[1 4 2]\n",
      " [7 6 3]\n",
      " [8 0 5]]\n",
      "\n",
      "->[[1 4 2]\n",
      " [7 0 3]\n",
      " [8 6 5]]\n",
      "\n",
      "->[[1 0 2]\n",
      " [7 4 3]\n",
      " [8 6 5]]\n",
      "\n",
      "->[[1 2 0]\n",
      " [7 4 3]\n",
      " [8 6 5]]\n",
      "\n",
      "->[[1 2 3]\n",
      " [7 4 0]\n",
      " [8 6 5]]\n",
      "\n",
      "->[[1 2 3]\n",
      " [7 4 5]\n",
      " [8 6 0]]\n",
      "\n",
      "->[[1 2 3]\n",
      " [7 4 5]\n",
      " [8 0 6]]\n",
      "\n",
      "->[[1 2 3]\n",
      " [7 4 5]\n",
      " [0 8 6]]\n",
      "\n",
      "->[[1 2 3]\n",
      " [0 4 5]\n",
      " [7 8 6]]\n",
      "\n",
      "->[[1 2 3]\n",
      " [4 0 5]\n",
      " [7 8 6]]\n",
      "\n",
      "->[[1 2 3]\n",
      " [4 5 0]\n",
      " [7 8 6]]\n",
      "\n",
      "->[[1 2 3]\n",
      " [4 5 6]\n",
      " [7 8 0]]\n"
     ]
    }
   ],
   "source": [
    "algo_mapping = {\n",
    "    1: 'BFS',\n",
    "    2: 'DFS',\n",
    "    3: 'GBFS',\n",
    "    4: 'A_star'\n",
    "}# mapping from integer to algorithm\n",
    "\n",
    "N = 3\n",
    "\n",
    "init_state = []\n",
    "print(\"Enter elements for initial state:\")\n",
    "for row in range(N):\n",
    "    row_ = list(map(int, input(f'Enter row {row}: ').split()))\n",
    "    init_state.append(row_)\n",
    "\n",
    "final_state = []\n",
    "print(\"\\nEnter elements for goal state:\")\n",
    "for row in range(N):\n",
    "    row_ = list(map(int, input(f'Enter row {row}: ').split()))\n",
    "    final_state.append(row_)\n",
    "\n",
    "algo = int(input(\"\"\"Please select algorithm:\n",
    "1. BFS\n",
    "2. DFS\n",
    "3. GBFS\n",
    "4. A*\\n\n",
    "(Input an integer):\"\"\"))\n",
    "\n",
    "\n",
    "eight = EightPuzzle(init_state, final_state) # making the eight puzzle agent\n",
    "\n",
    "if algo in range(1, 5):\n",
    "    exec('success_status, path = ' + str(algo_mapping[algo]) + '(eight)')\n",
    "else:\n",
    "    print(\"Check your input.\")\n",
    "\n",
    "try:\n",
    "    if success_status:\n",
    "        print(f\"{algo_mapping[algo]}: \")\n",
    "        print(*[eight.LTM(state) for state in path], sep=\"\\n\\n->\") # printing the path\n",
    "    else:\n",
    "        print(\"No path found.\")\n",
    "except:\n",
    "    pass"
   ]
  }
 ],
 "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": "3.6.6"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}