313_GeneticAlgorithm_Camouflage.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "provenance": [],
      "authorship_tag": "ABX9TyNpsfDJ4tIKwvKvvLaYi44H",
      "include_colab_link": true
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    },
    "language_info": {
      "name": "python"
    }
  },
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "view-in-github",
        "colab_type": "text"
      },
      "source": [
        "<a href=\"https://colab.research.google.com/github/bnsreenu/python_for_microscopists/blob/master/313_GeneticAlgorithm_Camouflage.ipynb\" target=\"_parent\"><img src=\"https://colab.research.google.com/assets/colab-badge.svg\" alt=\"Open In Colab\"/></a>"
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "https://youtu.be/shReZko1xns"
      ],
      "metadata": {
        "id": "Dz3iCQ3f1XWM"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "\n",
        "\n",
        "#Understanding the evolutionary part of the Genetic Algorithms <br>\n",
        "Genetic Algorithms (GAs) are inspired by the process of natural selection and evolution found in nature. GAs are a type of optimization technique that imitates the process of evolution and natural selection. They operate by creating a population of candidate solutions to a problem, and then repeatedly iterating through the following steps: selection, crossover, and mutation.\n",
        "<p>\n",
        "Read my blog on this topic here:\n",
        "https://www.linkedin.com/pulse/genetic-algorithms-optimization-problems-sreenivas-bhattiprolu"
      ],
      "metadata": {
        "id": "q9Qge_0eNXfP"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "The first part of the code below implements a genetic algorithm simulation to find the best RGB values that match a target RGB value. It involves the following 3 steps:\n",
        "<br>\n",
        "1. calc_fitness: calculates the mean absolute difference between an input RGB chromosome and the target RGB values.\n",
        "2.evolve_population: evolves a population of chromosomes by selecting the best chromosomes (elites), mating them to create offspring, and applying random mutations to the offspring.\n",
        "3. simulate: runs the genetic algorithm simulation by initializing a population of random RGB values, evolving the population over a specified number of generations, and returning the best chromosome found.\n",
        "\n",
        "The target RGB values are passed as input to the simulate function, and the simulation runs for 50 generations with a population size of 100, mutation rate of 0.01, and 10% of the population being selected as elites. The RGB values and fitness of the best individual in each generation are printed. The final output is the best chromosome found, which is the individual with the minimum fitness value (the closest match to the target RGB)."
      ],
      "metadata": {
        "id": "jVZJ4bCKLYgH"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "import random\n",
        "import numpy as np\n",
        "from matplotlib import pyplot as plt"
      ],
      "metadata": {
        "id": "RgEkEOtJoxcZ"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "#A quick summary of our fitness function - mean absolute difference between two arrays\n",
        "#Calculate absolute value element wise and then mean.\n",
        "print(np.mean(np.abs(np.array([50, 50, 10]) - np.array([255, 0, 0]))))"
      ],
      "metadata": {
        "id": "l-JEEPL1QuJG",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "98471cee-288d-4262-cdde-3427a0e71c35"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "88.33333333333333\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "Define the necessary helper functions.\n"
      ],
      "metadata": {
        "id": "K3OYGTvjRn84"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "def calc_fitness(chromosome, target_rgb):\n",
        "    \"\"\"\n",
        "    Calculate the fitness of a given chromosome.\n",
        "\n",
        "    Parameters:\n",
        "        chromosome (np.ndarray): An array representing the chromosome.\n",
        "        target_rgb (np.ndarray): The target RGB values.\n",
        "\n",
        "    Returns:\n",
        "        float: The mean absolute difference between the chromosome and target RGB values.\n",
        "    \"\"\"\n",
        "    return np.mean(np.abs(chromosome - target_rgb))"
      ],
      "metadata": {
        "id": "HjeH3jsIoxXN"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "def evolve_population(population, target_rgb, mutation_rate, elite_ratio):\n",
        "    \"\"\"\n",
        "    Evolve a population of chromosomes.\n",
        "\n",
        "    Parameters:\n",
        "        population (List[np.ndarray]): A list of arrays representing the chromosomes in the population.\n",
        "        target_rgb (np.ndarray): The target RGB values.\n",
        "        mutation_rate (float): The probability of a chromosome mutating.\n",
        "        elite_ratio (float): The ratio of chromosomes to be selected as elites.\n",
        "\n",
        "    Returns:\n",
        "        List[np.ndarray]: The evolved population of chromosomes.\n",
        "    \"\"\"\n",
        "    # Calculate fitness values for each chromosome in the population\n",
        "    fitness_values = [calc_fitness(c, target_rgb) for c in population]\n",
        "\n",
        "    # Select the chromosomes with the best fitness values as elites\n",
        "    elite_index = np.argsort(fitness_values)[:int(len(population) * elite_ratio)]\n",
        "    elites = [population[i] for i in elite_index]\n",
        "\n",
        "    # Create a new population by mating the elites\n",
        "    new_population = []\n",
        "    while len(new_population) < len(population) - len(elites):\n",
        "        # Select two parents randomly from the elites\n",
        "        parent1, parent2 = random.choices(elites, k=2)\n",
        "\n",
        "        # Create a child chromosome by combining the genes of the two parents\n",
        "        child = np.zeros(parent1.shape)\n",
        "        for i in range(len(parent1)):\n",
        "            if random.random() < 0.5:\n",
        "                child[i] = parent1[i]\n",
        "            else:\n",
        "                child[i] = parent2[i]\n",
        "\n",
        "            # Mutate the child chromosome with a probability determined by mutation_rate\n",
        "            if random.random() < mutation_rate:\n",
        "                child[i] = np.random.randint(0, 256)\n",
        "\n",
        "        # Add the child chromosome to the new population\n",
        "        new_population.append(child)\n",
        "\n",
        "    # Return the evolved population\n",
        "    return elites + new_population"
      ],
      "metadata": {
        "id": "Ko605RpzoxOP"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "def simulate(size, target_rgb, generations, mutation_rate, elite_ratio):\n",
        "    \"\"\"\n",
        "    Simulate the evolution of a population of chromosomes.\n",
        "\n",
        "    Parameters:\n",
        "        size (int): The size of the population.\n",
        "        target_rgb (np.ndarray): The target RGB values.\n",
        "        generations (int): The number of generations to evolve the population.\n",
        "        mutation_rate (float): The probability of a chromosome mutating.\n",
        "        elite_ratio (float): The ratio of chromosomes to be selected as elites.\n",
        "\n",
        "    Returns:\n",
        "        np.ndarray: The best chromosome found.\n",
        "    \"\"\"\n",
        "    # Initialize the population with random RGB values\n",
        "    population = [np.random.randint(0, 256, size=(3,)) for _ in range(size)]\n",
        "\n",
        "    # Evolve the population over the specified number of generations\n",
        "    for i in range(generations):\n",
        "        # Evolve the population by selecting elites, mutating, and recombining\n",
        "        population = evolve_population(population, target_rgb, mutation_rate, elite_ratio)\n",
        "\n",
        "        # Get the individual with the best fitness (closest to target RGB)\n",
        "        best_fit = min(population, key=lambda c: calc_fitness(c, target_rgb))\n",
        "\n",
        "        # Print the generation number and the RGB values and fitness of the best individual\n",
        "        print(f'Generation {i}: Best fitness {calc_fitness(best_fit, target_rgb)}, RGB values {best_fit}')\n",
        "\n",
        "    # Return the individual with the minimum fitness value\n",
        "    return min(population, key=lambda c: calc_fitness(c, target_rgb))"
      ],
      "metadata": {
        "id": "bMMhjkAfo5gn"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "target_rgb = np.array([255, 255, 0])\n",
        "best_camouflage = simulate(100, target_rgb, 50, 0.01, 0.1)\n",
        "\n",
        "print(\"Target is %s and the best campuflage is %s .\" % (target_rgb, best_camouflage))"
      ],
      "metadata": {
        "id": "iT8KjnAoHWoA",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "a59d28ea-76d3-40c5-e228-ceef16aa5b4e"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Generation 0: Best fitness 11.0, RGB values [237. 251.  11.]\n",
            "Generation 1: Best fitness 10.333333333333334, RGB values [239. 251.  11.]\n",
            "Generation 2: Best fitness 10.333333333333334, RGB values [239. 251.  11.]\n",
            "Generation 3: Best fitness 7.333333333333333, RGB values [248. 251.  11.]\n",
            "Generation 4: Best fitness 7.333333333333333, RGB values [248. 251.  11.]\n",
            "Generation 5: Best fitness 7.333333333333333, RGB values [248. 251.  11.]\n",
            "Generation 6: Best fitness 7.333333333333333, RGB values [248. 251.  11.]\n",
            "Generation 7: Best fitness 5.333333333333333, RGB values [254. 251.  11.]\n",
            "Generation 8: Best fitness 5.333333333333333, RGB values [254. 251.  11.]\n",
            "Generation 9: Best fitness 5.333333333333333, RGB values [254. 251.  11.]\n",
            "Generation 10: Best fitness 5.333333333333333, RGB values [254. 251.  11.]\n",
            "Generation 11: Best fitness 5.333333333333333, RGB values [254. 251.  11.]\n",
            "Generation 12: Best fitness 5.333333333333333, RGB values [254. 251.  11.]\n",
            "Generation 13: Best fitness 5.333333333333333, RGB values [254. 251.  11.]\n",
            "Generation 14: Best fitness 5.333333333333333, RGB values [254. 251.  11.]\n",
            "Generation 15: Best fitness 5.333333333333333, RGB values [254. 251.  11.]\n",
            "Generation 16: Best fitness 5.333333333333333, RGB values [254. 251.  11.]\n",
            "Generation 17: Best fitness 5.333333333333333, RGB values [254. 251.  11.]\n",
            "Generation 18: Best fitness 5.333333333333333, RGB values [254. 251.  11.]\n",
            "Generation 19: Best fitness 5.333333333333333, RGB values [254. 251.  11.]\n",
            "Generation 20: Best fitness 4.0, RGB values [254. 255.  11.]\n",
            "Generation 21: Best fitness 4.0, RGB values [254. 255.  11.]\n",
            "Generation 22: Best fitness 3.6666666666666665, RGB values [255. 255.  11.]\n",
            "Generation 23: Best fitness 3.6666666666666665, RGB values [255. 255.  11.]\n",
            "Generation 24: Best fitness 3.6666666666666665, RGB values [255. 255.  11.]\n",
            "Generation 25: Best fitness 3.6666666666666665, RGB values [255. 255.  11.]\n",
            "Generation 26: Best fitness 3.6666666666666665, RGB values [255. 255.  11.]\n",
            "Generation 27: Best fitness 3.6666666666666665, RGB values [255. 255.  11.]\n",
            "Generation 28: Best fitness 3.6666666666666665, RGB values [255. 255.  11.]\n",
            "Generation 29: Best fitness 3.6666666666666665, RGB values [255. 255.  11.]\n",
            "Generation 30: Best fitness 3.6666666666666665, RGB values [255. 255.  11.]\n",
            "Generation 31: Best fitness 3.6666666666666665, RGB values [255. 255.  11.]\n",
            "Generation 32: Best fitness 3.6666666666666665, RGB values [255. 255.  11.]\n",
            "Generation 33: Best fitness 3.6666666666666665, RGB values [255. 255.  11.]\n",
            "Generation 34: Best fitness 3.6666666666666665, RGB values [255. 255.  11.]\n",
            "Generation 35: Best fitness 3.6666666666666665, RGB values [255. 255.  11.]\n",
            "Generation 36: Best fitness 3.6666666666666665, RGB values [255. 255.  11.]\n",
            "Generation 37: Best fitness 3.6666666666666665, RGB values [255. 255.  11.]\n",
            "Generation 38: Best fitness 3.6666666666666665, RGB values [255. 255.  11.]\n",
            "Generation 39: Best fitness 3.6666666666666665, RGB values [255. 255.  11.]\n",
            "Generation 40: Best fitness 3.6666666666666665, RGB values [255. 255.  11.]\n",
            "Generation 41: Best fitness 3.6666666666666665, RGB values [255. 255.  11.]\n",
            "Generation 42: Best fitness 3.6666666666666665, RGB values [255. 255.  11.]\n",
            "Generation 43: Best fitness 3.6666666666666665, RGB values [255. 255.  11.]\n",
            "Generation 44: Best fitness 3.6666666666666665, RGB values [255. 255.  11.]\n",
            "Generation 45: Best fitness 3.6666666666666665, RGB values [255. 255.  11.]\n",
            "Generation 46: Best fitness 3.6666666666666665, RGB values [255. 255.  11.]\n",
            "Generation 47: Best fitness 3.6666666666666665, RGB values [255. 255.  11.]\n",
            "Generation 48: Best fitness 3.6666666666666665, RGB values [255. 255.  11.]\n",
            "Generation 49: Best fitness 3.6666666666666665, RGB values [255. 255.  11.]\n",
            "Target is [255 255   0] and the best campuflage is [255. 255.  11.] .\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# Create a figure and axis to plot the RGB values\n",
        "fig, ax = plt.subplots(1, 2, figsize=(8, 4))\n",
        "\n",
        "# Plot each RGB value in its own axis\n",
        "ax[0].imshow(np.reshape(target_rgb/255, (1, 1, 3)))\n",
        "ax[1].imshow(np.reshape(best_camouflage/255, (1, 1, 3)))\n",
        "\n",
        "# Remove the axis labels and grid\n",
        "ax[0].axis(\"off\")\n",
        "ax[1].axis(\"off\")\n",
        "\n",
        "#Add titles\n",
        "ax[0].set_title(\"Target Color\")\n",
        "ax[1].set_title(\"Camouflage Color\")\n",
        "# Show the plot\n",
        "plt.show()"
      ],
      "metadata": {
        "id": "a9Tt4QfAPOYs",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 341
        },
        "outputId": "2b808922-b347-430d-f20e-8b6cb3a3ad3e"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 800x400 with 2 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "#Now, let us extend the above example to a 2D image <br>\n",
        "The following code is similar to the previous code in terms of the basic structure and the purpose of the code, which is to simulate an evolutionary process except here we are working with 2-dimensional images rather than a single RGB value. This helps us visualize the evolution process.  \n",
        "<p>\n",
        "We will pick a specific color as our target color, for example, yellow with RGB = [222, 165, 33] and watch the random pixels evolve into this color.\n"
      ],
      "metadata": {
        "id": "Za-zEvmtEPhX"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "import random\n",
        "import numpy as np\n",
        "import matplotlib.pyplot as plt"
      ],
      "metadata": {
        "id": "faabk3nmpVz9"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "def calc_fitness(chromosome, target_grid):\n",
        "    \"\"\"\n",
        "    Calculates the fitness of a chromosome by comparing it to the target grid.\n",
        "\n",
        "    Parameters:\n",
        "        chromosome (np.array): The chromosome to calculate the fitness for.\n",
        "        target_grid (np.array): The target grid to compare the chromosome to.\n",
        "\n",
        "    Returns:\n",
        "        float: The fitness of the chromosome.\n",
        "    \"\"\"\n",
        "    return np.mean(np.abs(chromosome - target_grid))"
      ],
      "metadata": {
        "id": "pEI9cVJEpVsl"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "\n",
        "def evolve_population(population, target_grid, mutation_rate, elite_ratio):\n",
        "    \"\"\"\n",
        "    Evolves the population by selecting the best chromosomes (elites), breeding them, and mutating the offspring.\n",
        "\n",
        "    Parameters:\n",
        "        population (list of np.array): The current population of chromosomes.\n",
        "        target_grid (np.array): The target grid to compare the chromosomes to.\n",
        "        mutation_rate (float): The rate of mutation for the offspring.\n",
        "        elite_ratio (float): The ratio of the population that should be considered as elites.\n",
        "\n",
        "    Returns:\n",
        "        list of np.array: The evolved population of chromosomes.\n",
        "    \"\"\"\n",
        "    # Calculate fitness values for each chromosome in the population\n",
        "    fitness_values = [calc_fitness(c, target_grid) for c in population]\n",
        "\n",
        "    # Determine the indices of the best chromosomes (elites)\n",
        "    elite_index = np.argsort(fitness_values)[:int(len(population) * elite_ratio)]\n",
        "\n",
        "    # Select the best chromosomes (elites) from the population\n",
        "    elites = [population[i] for i in elite_index]\n",
        "\n",
        "    # Initialize the new population\n",
        "    new_population = []\n",
        "\n",
        "    # Create new offspring by breeding the elites\n",
        "    while len(new_population) < len(population) - len(elites):\n",
        "        # Select two parent chromosomes randomly from the elites\n",
        "        parent1, parent2 = random.choices(elites, k=2)\n",
        "\n",
        "        # Initialize the child chromosome\n",
        "        child = np.zeros(parent1.shape)\n",
        "\n",
        "        # Loop through the elements of the parent chromosomes\n",
        "        for i in range(len(parent1)):\n",
        "            for j in range(len(parent1[i])):\n",
        "                for k in range(len(parent1[i][j])):\n",
        "                    # Randomly inherit the element from either parent\n",
        "                    if random.random() < 0.5:\n",
        "                        child[i][j][k] = parent1[i][j][k]\n",
        "                    else:\n",
        "                        child[i][j][k] = parent2[i][j][k]\n",
        "\n",
        "                    # Mutate the element with a certain probability\n",
        "                    if random.random() < mutation_rate:\n",
        "                        child[i][j][k] = np.random.randint(0, 256)\n",
        "\n",
        "        # Add the child chromosome to the new population\n",
        "        new_population.append(child)\n",
        "\n",
        "    # Combine the elites and the offspring to form the evolved population\n",
        "    return elites + new_population\n"
      ],
      "metadata": {
        "id": "853wthbWpVkg"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "\n",
        "images =[]\n",
        "def simulate(size, target_grid, generations, mutation_rate, elite_ratio):\n",
        "    \"\"\"\n",
        "    This function simulates the evolutionary process to generate new images.\n",
        "\n",
        "    Parameters:\n",
        "    size (int) : The size of the population.\n",
        "    target_grid (numpy array) : The target image for the evolutionary process.\n",
        "    generations (int) : The number of generations to run the simulation for.\n",
        "    mutation_rate (float) : The rate of mutation of the images in the population.\n",
        "    elite_ratio (float) : The ratio of the population that will be considered as elites.\n",
        "\n",
        "    Returns:\n",
        "    (tuple) : A tuple of the best image generated by the simulation and a list of all images generated in the simulation.\n",
        "    \"\"\"\n",
        "    # The size of each image in the population\n",
        "    grid_size = (16, 16, 3)\n",
        "    # initialize the population with random values\n",
        "    population = [np.random.randint(0, 256, size=grid_size) for _ in range(size)]\n",
        "\n",
        "\n",
        "    for i in range(generations):\n",
        "      # evolve the population for each generation\n",
        "        population = evolve_population(population, target_grid, mutation_rate, elite_ratio)\n",
        "         # find the chromosome with the best fitness value\n",
        "        best_fit = min(population, key=lambda c: calc_fitness(c, target_grid))\n",
        "        # print the generation number and best fitness value\n",
        "        print(f'Generation {i}: Best fitness {calc_fitness(best_fit, target_grid)}')\n",
        "\n",
        "        # store the generated image\n",
        "        images.append(best_fit.astype(np.uint8))\n",
        "\n",
        "        # commented out code for displaying or saving images\n",
        "        #fig, ax = plt.subplots()\n",
        "        #ax.imshow(best_fit.astype(np.uint8))\n",
        "        #plt.title(f'Generation {i}')\n",
        "        #fig.tight_layout()\n",
        "        #plt.show()\n",
        "        # filename = f'camo_images/yellow/gen_{i}.png'\n",
        "        # plt.savefig(filename)\n",
        "\n",
        "    # Return the best fit image and the list of all images\n",
        "    return min(population, key=lambda c: calc_fitness(c, target_grid)), images"
      ],
      "metadata": {
        "id": "_64Cxmx1phBi"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "target_grid = np.zeros((16, 16, 3), dtype=np.uint8)\n",
        "#Yellow color from African Savannah [222, 165, 33]\n",
        "# Green grass [0, 154, 23]\n",
        "target_grid[:] = [222, 165, 33]\n",
        "\n",
        "pop_size = 100\n",
        "num_generations = 50\n",
        "mut_rate = 0.01\n",
        "elite_ratio = 0.1\n",
        "best_camouflage, images = simulate(pop_size, target_grid, num_generations, mut_rate, elite_ratio)\n",
        "\n"
      ],
      "metadata": {
        "id": "mGfWR_azERVc",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "f16ba74b-6607-4f83-cebd-e35560971fd9"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Generation 0: Best fitness 81.09375\n",
            "Generation 1: Best fitness 79.03515625\n",
            "Generation 2: Best fitness 76.32291666666667\n",
            "Generation 3: Best fitness 73.9453125\n",
            "Generation 4: Best fitness 72.34244791666667\n",
            "Generation 5: Best fitness 70.33984375\n",
            "Generation 6: Best fitness 69.05208333333333\n",
            "Generation 7: Best fitness 67.80989583333333\n",
            "Generation 8: Best fitness 66.32291666666667\n",
            "Generation 9: Best fitness 65.07421875\n",
            "Generation 10: Best fitness 63.608072916666664\n",
            "Generation 11: Best fitness 62.8828125\n",
            "Generation 12: Best fitness 62.338541666666664\n",
            "Generation 13: Best fitness 61.1953125\n",
            "Generation 14: Best fitness 60.399739583333336\n",
            "Generation 15: Best fitness 59.60546875\n",
            "Generation 16: Best fitness 59.057291666666664\n",
            "Generation 17: Best fitness 57.91796875\n",
            "Generation 18: Best fitness 57.016927083333336\n",
            "Generation 19: Best fitness 56.661458333333336\n",
            "Generation 20: Best fitness 55.3359375\n",
            "Generation 21: Best fitness 54.890625\n",
            "Generation 22: Best fitness 54.19921875\n",
            "Generation 23: Best fitness 53.78125\n",
            "Generation 24: Best fitness 53.26171875\n",
            "Generation 25: Best fitness 52.384114583333336\n",
            "Generation 26: Best fitness 51.169270833333336\n",
            "Generation 27: Best fitness 50.73828125\n",
            "Generation 28: Best fitness 50.6875\n",
            "Generation 29: Best fitness 50.11328125\n",
            "Generation 30: Best fitness 49.697916666666664\n",
            "Generation 31: Best fitness 49.2265625\n",
            "Generation 32: Best fitness 48.6328125\n",
            "Generation 33: Best fitness 48.359375\n",
            "Generation 34: Best fitness 47.89453125\n",
            "Generation 35: Best fitness 47.6015625\n",
            "Generation 36: Best fitness 47.188802083333336\n",
            "Generation 37: Best fitness 46.93359375\n",
            "Generation 38: Best fitness 46.180989583333336\n",
            "Generation 39: Best fitness 46.0625\n",
            "Generation 40: Best fitness 45.513020833333336\n",
            "Generation 41: Best fitness 45.135416666666664\n",
            "Generation 42: Best fitness 44.8203125\n",
            "Generation 43: Best fitness 44.494791666666664\n",
            "Generation 44: Best fitness 44.33203125\n",
            "Generation 45: Best fitness 43.927083333333336\n",
            "Generation 46: Best fitness 43.434895833333336\n",
            "Generation 47: Best fitness 43.174479166666664\n",
            "Generation 48: Best fitness 43.072916666666664\n",
            "Generation 49: Best fitness 42.631510416666664\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "Let us plot images in a grid to visualize the evolution"
      ],
      "metadata": {
        "id": "D9tB_btxDGLD"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "import numpy as np\n",
        "import matplotlib.pyplot as plt\n",
        "\n",
        "## Function to display a grid of images\n",
        "def display_images(images, num_images=50, rows=10, cols=5):\n",
        "    fig, axs = plt.subplots(rows, cols, figsize=(20, 20))\n",
        "    for i, ax in enumerate(axs.ravel()):\n",
        "        if i < num_images:\n",
        "            ax.imshow(images[i], cmap='gray')\n",
        "            ax.axis('off')\n",
        "        else:\n",
        "            fig.delaxes(ax)\n",
        "    plt.tight_layout()\n",
        "    plt.show()\n",
        "\n",
        "# Plot images:\n",
        "\n",
        "display_images(images, rows=5, cols=10)\n"
      ],
      "metadata": {
        "id": "ZVkbE6Sq6Ual",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 815
        },
        "outputId": "7cf71a79-d9bb-4490-b0c8-5761ec090cf2"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 2000x2000 with 50 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        }
      ]
    }
  ]
}