Genetic_Algorithm.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "from utils import *\n",
    "import random\n",
    "import seaborn as sns\n",
    "import matplotlib.pyplot as plt\n",
    "import numpy as np"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "322\n"
     ]
    }
   ],
   "source": [
    "with open(\"defects.csv\") as f:\n",
    "    lines = f.readlines()\n",
    "    defects = [line.strip().split(\",\") for line in lines[1:]]\n",
    "\n",
    "# Defects never occur at precise integer coordinates, so a defect will never be shared by two tiles.\n",
    "# We floor each defect coordinate to get the tile it belongs to.\n",
    "# We then count the number of defects in each tile.\n",
    "defect_counts = {}\n",
    "for x, y in defects:\n",
    "    x = int(float(x))\n",
    "    if x not in defect_counts:\n",
    "        defect_counts[x] = {\"a\": 0, \"b\": 0, \"c\": 0}\n",
    "\n",
    "    defect_counts[x][y] += 1\n",
    "\n",
    "print(len(defect_counts))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "dough_length = 500\n",
    "\n",
    "biscuits = {\n",
    "    0: {\"length\": 4, \"value\": 6, \"defects\": {\"a\": 4, \"b\": 2, \"c\": 3}, \"id\": 0},\n",
    "    1: {\"length\": 8, \"value\": 12, \"defects\": {\"a\": 5, \"b\": 4, \"c\": 4}, \"id\": 1},\n",
    "    2: {\"length\": 2, \"value\": 1, \"defects\": {\"a\": 1, \"b\": 2, \"c\": 1}, \"id\": 2},\n",
    "    3: {\"length\": 5, \"value\": 8, \"defects\": {\"a\": 2, \"b\": 3, \"c\": 2}, \"id\": 3},\n",
    "}"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [],
   "source": [
    "class Individual:\n",
    "    \"\"\"\n",
    "    Represents an individual in the genetic algorithm for biscuit production optimization.\n",
    "    The chromosome is a list of <dough_length> biscuit types for each position. -1 represents a missing biscuit.\n",
    "    Biscuit types are represented by integers 0, 1, 2, 3.\n",
    "    e.g. [-1, -1, 2, -1, -1] represents a solution with a biscuit of type 2 at position 2.\n",
    "    \"\"\"\n",
    "\n",
    "    def __init__(self, chromosome=None):\n",
    "        \"\"\"\n",
    "        Initializes a new individual.\n",
    "\n",
    "        Args:\n",
    "            chromosome (list): The chromosome representing the biscuit types for each gene. If not provided, a random chromosome will be generated.\n",
    "\n",
    "        Returns:\n",
    "            None\n",
    "        \"\"\"\n",
    "        if chromosome is None:\n",
    "            self.initialize_chromosome()\n",
    "        else:\n",
    "            self.chromosome = chromosome\n",
    "        self.fitness = self.calculate_fitness()\n",
    "\n",
    "    @staticmethod\n",
    "    def convert_chromosome_to_solution(chromosome):\n",
    "        \"\"\"\n",
    "        Convert a chromosome to a solution (they have different formats).\n",
    "\n",
    "        Args:\n",
    "            chromosome (list): The chromosome to convert.\n",
    "\n",
    "        Returns:\n",
    "            list: The converted solution.\n",
    "        \"\"\"\n",
    "        solution = []\n",
    "        for position, biscuit_type in enumerate(chromosome):\n",
    "            if biscuit_type in [0, 1, 2, 3]:\n",
    "                solution.append((position, biscuit_type))\n",
    "        return solution\n",
    "\n",
    "    @staticmethod\n",
    "    def convert_solution_to_chromosome(solution):\n",
    "        \"\"\"\n",
    "        Convert a solution to a chromosome (they have different formats).\n",
    "\n",
    "        Args:\n",
    "            solution (list): The solution to convert.\n",
    "\n",
    "        Returns:\n",
    "            list: The converted chromosome.\n",
    "        \"\"\"\n",
    "        chromosome = [-1] * dough_length\n",
    "        for position, biscuit_type in solution:\n",
    "            chromosome[position] = biscuit_type\n",
    "        return chromosome\n",
    "\n",
    "    def initialize_chromosome(self):\n",
    "        \"\"\"\n",
    "        Initializes the chromosome by randomly selecting biscuit types for each gene.\n",
    "\n",
    "        Returns:\n",
    "            list: The initialized chromosome.\n",
    "        \"\"\"\n",
    "        self.chromosome = [random.choice(list(biscuits.keys()) + [-1]) for _ in range(dough_length)]\n",
    "        self.clean_chromosome()\n",
    "        return self.chromosome\n",
    "\n",
    "    def clean_chromosome(self):\n",
    "        \"\"\"\n",
    "        Cleans the chromosome by removing invalid biscuits and updating the fitness.\n",
    "\n",
    "        Returns:\n",
    "            list: The cleaned chromosome.\n",
    "        \"\"\"\n",
    "        self.chromosome = Individual.convert_solution_to_chromosome(check_solution(Individual.convert_chromosome_to_solution(self.chromosome))[1])\n",
    "        self.fitness = self.calculate_fitness()\n",
    "        return self.chromosome\n",
    "    \n",
    "    def calculate_fitness(self):\n",
    "        \"\"\"\n",
    "        Calculates the fitness of the individual based on the biscuit in the chromosome.\n",
    "\n",
    "        Returns:\n",
    "            int: The fitness value.\n",
    "        \"\"\"\n",
    "        return sum([biscuits[biscuit_type]['value'] for biscuit_type in self.chromosome if biscuit_type != -1])\n",
    "\n",
    "    def mutate(self, mutation_rate):\n",
    "        \"\"\"\n",
    "        Mutates the individual's chromosome by randomly changing some genes.\n",
    "\n",
    "        Args:\n",
    "            mutation_rate (float): The probability of a gene being mutated.\n",
    "\n",
    "        Returns:\n",
    "            list: The mutated chromosome.\n",
    "        \"\"\"\n",
    "        chromosome_array = np.array(self.chromosome)\n",
    "        mutation_indices = np.random.rand(len(chromosome_array)) < mutation_rate\n",
    "        mutation_values = list(biscuits.keys()) + [-1]\n",
    "        chromosome_array[mutation_indices] = np.random.choice(mutation_values, size=np.sum(mutation_indices))\n",
    "        self.chromosome = chromosome_array.tolist()\n",
    "\n",
    "        return self.chromosome\n",
    "\n",
    "class GeneticAlgorithm:\n",
    "    \"\"\"\n",
    "    A class representing a genetic algorithm.\n",
    "    \"\"\"\n",
    "\n",
    "    def __init__(self, pop_size, mutation_rate, selection_rate):\n",
    "        self.pop_size = pop_size\n",
    "        self.mutation_rate = mutation_rate\n",
    "        self.population = None\n",
    "        self.selection_rate = selection_rate\n",
    "\n",
    "        self.fitness_history = []\n",
    "\n",
    "    def run(self, generations):\n",
    "        \"\"\"\n",
    "        Runs the genetic algorithm for the specified number of generations.\n",
    "\n",
    "        Args:\n",
    "        - generations (int): The number of generations to run.\n",
    "\n",
    "        Returns:\n",
    "        - best_solution (Individual): The best solution found.\n",
    "        \"\"\"\n",
    "        pop_size = self.pop_size\n",
    "        mutation_rate = self.mutation_rate\n",
    "\n",
    "        if self.population is None:\n",
    "            self.population = self.initialize_population(pop_size)\n",
    "        \n",
    "        self.fitness_history.append([individual.fitness for individual in self.population])\n",
    "\n",
    "        for i in range(generations):\n",
    "            new_population = []\n",
    "            population = self.tournament_selection(self.population, int(self.pop_size * self.selection_rate))\n",
    "            while len(new_population) < pop_size:\n",
    "                parent1, parent2 = random.sample(population, 2)\n",
    "                child1, child2 = self.crossover(parent1, parent2)\n",
    "                child1.mutate(mutation_rate)\n",
    "                child2.mutate(mutation_rate)\n",
    "                child1.clean_chromosome()\n",
    "                child2.clean_chromosome()\n",
    "                new_population.extend([child1, child2])\n",
    "            \n",
    "            self.population = new_population\n",
    "            best_fitness = max(ind.fitness for ind in self.population)\n",
    "            print(f\"Generation {i}, best fitness = {best_fitness}\")\n",
    "            self.fitness_history.append([individual.fitness for individual in self.population])\n",
    "\n",
    "        best_solution = max(self.population, key=lambda x: x.fitness)\n",
    "        return best_solution\n",
    "\n",
    "\n",
    "    # Initialize population\n",
    "    def initialize_population(self, pop_size):\n",
    "        \"\"\"\n",
    "        Initializes the population with random individuals.\n",
    "\n",
    "        Args:\n",
    "        - pop_size (int): The size of the population.\n",
    "\n",
    "        Returns:\n",
    "        - population (list): The initialized population.\n",
    "        \"\"\"\n",
    "        return [Individual() for _ in range(pop_size)]\n",
    "\n",
    "    # Selection\n",
    "    def tournament_selection(self, population, tournament_size):\n",
    "        \"\"\"\n",
    "        Perform tournament selection on a population.\n",
    "\n",
    "        Args:\n",
    "        population (list): A list of individuals in the population.\n",
    "        tournament_size (int): The number of individuals in each tournament.\n",
    "\n",
    "        Returns:\n",
    "        list: A list of selected individuals for crossover.\n",
    "        \"\"\"\n",
    "\n",
    "        # Always select the best individual\n",
    "        best_individual = max(population, key=lambda individual: individual.fitness)\n",
    "        selected_individuals = [best_individual]\n",
    "\n",
    "        for _ in range(len(population) // 2):\n",
    "            # Select random individuals for the tournament\n",
    "            tournament = random.sample(population, tournament_size)\n",
    "            \n",
    "            # Select the best individual from the tournament\n",
    "            best_individual = max(tournament, key=lambda individual: individual.fitness)\n",
    "            selected_individuals.append(best_individual)\n",
    "\n",
    "        return selected_individuals\n",
    "\n",
    "    # Crossover\n",
    "    def crossover(self, parent1, parent2):\n",
    "        \"\"\"\n",
    "        Performs crossover between two parents.\n",
    "\n",
    "        Args:\n",
    "        - parent1 (Individual): The first parent.\n",
    "        - parent2 (Individual): The second parent.\n",
    "\n",
    "        Returns:\n",
    "        - child1 (Individual): The first child.\n",
    "        - child2 (Individual): The second child.\n",
    "        \"\"\"\n",
    "        crossover_point = random.randint(1, dough_length - 1)\n",
    "        child1 = Individual(parent1.chromosome[:crossover_point] + parent2.chromosome[crossover_point:])\n",
    "        child2 = Individual(parent2.chromosome[:crossover_point] + parent1.chromosome[crossover_point:])\n",
    "        return child1, child2\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Generation 0, best fitness = 642\n",
      "Generation 1, best fitness = 658\n",
      "Generation 2, best fitness = 654\n",
      "Generation 3, best fitness = 665\n",
      "Generation 4, best fitness = 671\n",
      "Generation 5, best fitness = 672\n",
      "Generation 6, best fitness = 678\n",
      "Generation 7, best fitness = 685\n",
      "Generation 8, best fitness = 686\n",
      "Generation 9, best fitness = 691\n",
      "Generation 10, best fitness = 692\n",
      "Generation 11, best fitness = 693\n",
      "Generation 12, best fitness = 695\n",
      "Generation 13, best fitness = 698\n",
      "Generation 14, best fitness = 701\n",
      "Generation 15, best fitness = 703\n",
      "Generation 16, best fitness = 705\n",
      "Generation 17, best fitness = 706\n",
      "Generation 18, best fitness = 708\n",
      "Generation 19, best fitness = 706\n",
      "Generation 20, best fitness = 709\n",
      "Generation 21, best fitness = 708\n",
      "Generation 22, best fitness = 710\n",
      "Generation 23, best fitness = 710\n",
      "Generation 24, best fitness = 710\n",
      "Generation 25, best fitness = 710\n",
      "Generation 26, best fitness = 710\n",
      "Generation 27, best fitness = 710\n",
      "Generation 28, best fitness = 711\n",
      "Generation 29, best fitness = 711\n",
      "Generation 30, best fitness = 712\n",
      "Generation 31, best fitness = 712\n",
      "Generation 32, best fitness = 711\n",
      "Generation 33, best fitness = 711\n",
      "Generation 34, best fitness = 712\n",
      "Generation 35, best fitness = 713\n",
      "Generation 36, best fitness = 713\n",
      "Generation 37, best fitness = 713\n",
      "Generation 38, best fitness = 713\n",
      "Generation 39, best fitness = 713\n",
      "Generation 40, best fitness = 713\n",
      "Generation 41, best fitness = 713\n",
      "Generation 42, best fitness = 713\n",
      "Generation 43, best fitness = 714\n",
      "Generation 44, best fitness = 713\n",
      "Generation 45, best fitness = 714\n",
      "Generation 46, best fitness = 715\n",
      "Generation 47, best fitness = 716\n",
      "Generation 48, best fitness = 716\n",
      "Generation 49, best fitness = 717\n",
      "Generation 50, best fitness = 719\n",
      "Generation 51, best fitness = 719\n",
      "Generation 52, best fitness = 719\n",
      "Generation 53, best fitness = 719\n",
      "Generation 54, best fitness = 719\n",
      "Generation 55, best fitness = 719\n",
      "Generation 56, best fitness = 719\n",
      "Generation 57, best fitness = 719\n",
      "Generation 58, best fitness = 719\n",
      "Generation 59, best fitness = 719\n",
      "Generation 60, best fitness = 721\n",
      "Generation 61, best fitness = 721\n",
      "Generation 62, best fitness = 723\n",
      "Generation 63, best fitness = 723\n",
      "Generation 64, best fitness = 724\n",
      "Generation 65, best fitness = 724\n",
      "Generation 66, best fitness = 724\n",
      "Generation 67, best fitness = 724\n",
      "Generation 68, best fitness = 724\n",
      "Generation 69, best fitness = 724\n",
      "Generation 70, best fitness = 724\n",
      "Generation 71, best fitness = 724\n",
      "Generation 72, best fitness = 724\n",
      "Generation 73, best fitness = 724\n",
      "Generation 74, best fitness = 724\n",
      "Generation 75, best fitness = 724\n",
      "Generation 76, best fitness = 724\n",
      "Generation 77, best fitness = 725\n",
      "Generation 78, best fitness = 725\n",
      "Generation 79, best fitness = 724\n",
      "Generation 80, best fitness = 723\n",
      "Generation 81, best fitness = 724\n",
      "Generation 82, best fitness = 724\n",
      "Generation 83, best fitness = 725\n",
      "Generation 84, best fitness = 725\n",
      "Generation 85, best fitness = 725\n",
      "Generation 86, best fitness = 725\n",
      "Generation 87, best fitness = 725\n",
      "Generation 88, best fitness = 725\n",
      "Generation 89, best fitness = 725\n",
      "Generation 90, best fitness = 725\n",
      "Generation 91, best fitness = 725\n",
      "Generation 92, best fitness = 725\n",
      "Generation 93, best fitness = 725\n",
      "Generation 94, best fitness = 726\n",
      "Generation 95, best fitness = 726\n",
      "Generation 96, best fitness = 726\n",
      "Generation 97, best fitness = 726\n",
      "Generation 98, best fitness = 726\n",
      "Generation 99, best fitness = 726\n",
      "Generation 100, best fitness = 726\n",
      "Generation 101, best fitness = 726\n",
      "Generation 102, best fitness = 726\n",
      "Generation 103, best fitness = 728\n",
      "Generation 104, best fitness = 728\n",
      "Generation 105, best fitness = 728\n",
      "Generation 106, best fitness = 728\n",
      "Generation 107, best fitness = 728\n",
      "Generation 108, best fitness = 727\n",
      "Generation 109, best fitness = 728\n",
      "Generation 110, best fitness = 728\n",
      "Generation 111, best fitness = 728\n",
      "Generation 112, best fitness = 728\n",
      "Generation 113, best fitness = 728\n",
      "Generation 114, best fitness = 728\n",
      "Generation 115, best fitness = 728\n",
      "Generation 116, best fitness = 728\n",
      "Generation 117, best fitness = 728\n",
      "Generation 118, best fitness = 729\n",
      "Generation 119, best fitness = 729\n",
      "Generation 120, best fitness = 729\n",
      "Generation 121, best fitness = 729\n",
      "Generation 122, best fitness = 729\n",
      "Generation 123, best fitness = 729\n",
      "Generation 124, best fitness = 729\n",
      "Generation 125, best fitness = 729\n",
      "Generation 126, best fitness = 729\n",
      "Generation 127, best fitness = 729\n",
      "Generation 128, best fitness = 729\n",
      "Generation 129, best fitness = 730\n",
      "Generation 130, best fitness = 729\n",
      "Generation 131, best fitness = 729\n",
      "Generation 132, best fitness = 729\n",
      "Generation 133, best fitness = 729\n",
      "Generation 134, best fitness = 729\n",
      "Generation 135, best fitness = 729\n",
      "Generation 136, best fitness = 729\n",
      "Generation 137, best fitness = 729\n",
      "Generation 138, best fitness = 729\n",
      "Generation 139, best fitness = 729\n",
      "Generation 140, best fitness = 731\n",
      "Generation 141, best fitness = 729\n",
      "Generation 142, best fitness = 731\n",
      "Generation 143, best fitness = 731\n",
      "Generation 144, best fitness = 731\n",
      "Generation 145, best fitness = 731\n",
      "Generation 146, best fitness = 731\n",
      "Generation 147, best fitness = 731\n",
      "Generation 148, best fitness = 731\n",
      "Generation 149, best fitness = 731\n",
      "Generation 150, best fitness = 731\n",
      "Generation 151, best fitness = 731\n",
      "Generation 152, best fitness = 731\n",
      "Generation 153, best fitness = 731\n",
      "Generation 154, best fitness = 731\n",
      "Generation 155, best fitness = 731\n",
      "Generation 156, best fitness = 731\n",
      "Generation 157, best fitness = 731\n",
      "Generation 158, best fitness = 731\n",
      "Generation 159, best fitness = 731\n",
      "Generation 160, best fitness = 731\n",
      "Generation 161, best fitness = 731\n",
      "Generation 162, best fitness = 731\n",
      "Generation 163, best fitness = 731\n",
      "Generation 164, best fitness = 731\n",
      "Generation 165, best fitness = 731\n",
      "Generation 166, best fitness = 732\n",
      "Generation 167, best fitness = 732\n",
      "Generation 168, best fitness = 732\n",
      "Generation 169, best fitness = 732\n",
      "Generation 170, best fitness = 732\n",
      "Generation 171, best fitness = 732\n",
      "Generation 172, best fitness = 732\n",
      "Generation 173, best fitness = 732\n",
      "Generation 174, best fitness = 732\n",
      "Generation 175, best fitness = 734\n",
      "Generation 176, best fitness = 732\n",
      "Generation 177, best fitness = 732\n",
      "Generation 178, best fitness = 733\n",
      "Generation 179, best fitness = 732\n",
      "Generation 180, best fitness = 732\n",
      "Generation 181, best fitness = 734\n",
      "Generation 182, best fitness = 734\n",
      "Generation 183, best fitness = 734\n",
      "Generation 184, best fitness = 734\n",
      "Generation 185, best fitness = 734\n",
      "Generation 186, best fitness = 734\n",
      "Generation 187, best fitness = 734\n",
      "Generation 188, best fitness = 734\n",
      "Generation 189, best fitness = 734\n",
      "Generation 190, best fitness = 734\n",
      "Generation 191, best fitness = 734\n",
      "Generation 192, best fitness = 734\n",
      "Generation 193, best fitness = 734\n",
      "Generation 194, best fitness = 734\n",
      "Generation 195, best fitness = 734\n",
      "Generation 196, best fitness = 734\n",
      "Generation 197, best fitness = 734\n",
      "Generation 198, best fitness = 734\n",
      "Generation 199, best fitness = 734\n",
      "Generation 200, best fitness = 734\n",
      "Generation 201, best fitness = 734\n",
      "Generation 202, best fitness = 734\n",
      "Generation 203, best fitness = 734\n",
      "Generation 204, best fitness = 734\n",
      "Generation 205, best fitness = 734\n",
      "Generation 206, best fitness = 734\n",
      "Generation 207, best fitness = 734\n",
      "Generation 208, best fitness = 734\n",
      "Generation 209, best fitness = 734\n",
      "Generation 210, best fitness = 734\n",
      "Generation 211, best fitness = 734\n",
      "Generation 212, best fitness = 734\n",
      "Generation 213, best fitness = 734\n",
      "Generation 214, best fitness = 734\n",
      "Generation 215, best fitness = 734\n",
      "Generation 216, best fitness = 734\n",
      "Generation 217, best fitness = 734\n",
      "Generation 218, best fitness = 734\n",
      "Generation 219, best fitness = 734\n",
      "Generation 220, best fitness = 734\n",
      "Generation 221, best fitness = 738\n",
      "Generation 222, best fitness = 736\n",
      "Generation 223, best fitness = 736\n",
      "Generation 224, best fitness = 738\n",
      "Generation 225, best fitness = 738\n",
      "Generation 226, best fitness = 738\n",
      "Generation 227, best fitness = 738\n",
      "Generation 228, best fitness = 738\n",
      "Generation 229, best fitness = 738\n",
      "Generation 230, best fitness = 738\n",
      "Generation 231, best fitness = 738\n",
      "Generation 232, best fitness = 738\n",
      "Generation 233, best fitness = 738\n",
      "Generation 234, best fitness = 738\n",
      "Generation 235, best fitness = 738\n",
      "Generation 236, best fitness = 738\n",
      "Generation 237, best fitness = 738\n",
      "Generation 238, best fitness = 738\n",
      "Generation 239, best fitness = 738\n",
      "Generation 240, best fitness = 739\n",
      "Generation 241, best fitness = 739\n",
      "Generation 242, best fitness = 739\n",
      "Generation 243, best fitness = 739\n",
      "Generation 244, best fitness = 740\n",
      "Generation 245, best fitness = 740\n",
      "Generation 246, best fitness = 740\n",
      "Generation 247, best fitness = 740\n",
      "Generation 248, best fitness = 740\n",
      "Generation 249, best fitness = 740\n",
      "(740, [(0, 1), (8, 1), (16, 3), (21, 1), (29, 3), (34, 2), (36, 0), (40, 0), (45, 3), (50, 0), (54, 1), (62, 3), (67, 1), (75, 1), (83, 0), (87, 3), (92, 0), (98, 0), (102, 3), (107, 0), (111, 2), (114, 0), (118, 0), (122, 0), (126, 3), (131, 1), (139, 0), (144, 3), (149, 1), (158, 1), (167, 3), (172, 0), (176, 1), (185, 2), (188, 1), (196, 1), (204, 1), (212, 1), (220, 1), (228, 1), (236, 1), (244, 1), (252, 1), (260, 3), (265, 3), (270, 3), (275, 0), (279, 1), (287, 3), (294, 3), (299, 0), (303, 0), (307, 3), (312, 1), (320, 3), (325, 1), (333, 3), (338, 3), (343, 3), (348, 3), (353, 0), (357, 1), (365, 0), (369, 0), (373, 1), (381, 2), (383, 3), (388, 3), (393, 3), (398, 0), (402, 3), (407, 1), (415, 3), (420, 1), (428, 1), (436, 1), (444, 0), (448, 3), (453, 1), (461, 3), (466, 3), (471, 1), (479, 1), (487, 1), (495, 3)], True)\n"
     ]
    }
   ],
   "source": [
    "# Genetic Algorithm Parameters\n",
    "POP_SIZE = 1000\n",
    "MUTATION_RATE = 0.05\n",
    "GENERATIONS = 250\n",
    "SELECTION_RATE = 0.05\n",
    "\n",
    "genetic_algorithm = GeneticAlgorithm(POP_SIZE, MUTATION_RATE, SELECTION_RATE)\n",
    "result = genetic_algorithm.run(GENERATIONS)\n",
    "\n",
    "print(check_solution(Individual.convert_chromosome_to_solution(result.chromosome)))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 576x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Plot fitness history (mean and best fitness of each generation)\n",
    "\n",
    "sns.set_style(\"darkgrid\")\n",
    "sns.set_palette(\"colorblind\")\n",
    "sns.set_context(\"paper\")\n",
    "\n",
    "fig, ax = plt.subplots(figsize=(8, 4))\n",
    "ax.plot(np.mean(genetic_algorithm.fitness_history, axis=1), label=\"Mean fitness\")\n",
    "ax.plot(np.max(genetic_algorithm.fitness_history, axis=1), label=\"Best fitness\")\n",
    "ax.set_xlabel(\"Generation\")\n",
    "ax.set_ylabel(\"Fitness\")\n",
    "ax.legend()\n",
    "plt.tight_layout()\n",
    "plt.savefig(\"fitness_history_ga.png\", dpi=300)\n",
    "plt.show()\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(740, [(0, 1), (8, 1), (16, 3), (21, 1), (29, 3), (34, 2), (36, 0), (40, 0), (45, 3), (50, 0), (54, 1), (62, 3), (67, 1), (75, 1), (83, 0), (87, 3), (92, 0), (98, 0), (102, 3), (107, 0), (111, 2), (114, 0), (118, 0), (122, 0), (126, 3), (131, 1), (139, 0), (144, 3), (149, 1), (158, 1), (167, 3), (172, 0), (176, 1), (185, 2), (188, 1), (196, 1), (204, 1), (212, 1), (220, 1), (228, 1), (236, 1), (244, 1), (252, 1), (260, 3), (265, 3), (270, 3), (275, 0), (279, 1), (287, 3), (294, 3), (299, 0), (303, 0), (307, 3), (312, 1), (320, 3), (325, 1), (333, 3), (338, 3), (343, 3), (348, 3), (353, 0), (357, 1), (365, 0), (369, 0), (373, 1), (381, 2), (383, 3), (388, 3), (393, 3), (398, 0), (402, 3), (407, 1), (415, 3), (420, 1), (428, 1), (436, 1), (444, 0), (448, 3), (453, 1), (461, 3), (466, 3), (471, 1), (479, 1), (487, 1), (495, 3)], True)\n"
     ]
    }
   ],
   "source": [
    "from utils import check_solution\n",
    "# Build and check solution\n",
    "solution = Individual.convert_chromosome_to_solution(result.chromosome)\n",
    "print(check_solution(solution))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [],
   "source": [
    "from utils import save_solution\n",
    "# Save solution\n",
    "save_solution(solution, \"solution_genetic_algorithm.txt\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "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": "3.9.7"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}