Ackley/Akley.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 81,
   "id": "01d68de3-adb6-402a-932a-40f1f87c5a9c",
   "metadata": {},
   "outputs": [],
   "source": [
    "import random\n",
    "import numpy as np\n",
    "import math\n",
    "import matplotlib.pyplot as plt"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 82,
   "id": "4951dfa4-8a79-4e59-bd09-4983598af488",
   "metadata": {},
   "outputs": [],
   "source": [
    "sx = 100"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 83,
   "id": "0b687aa1-a409-4241-915c-265e656a0cb7",
   "metadata": {},
   "outputs": [],
   "source": [
    "class point:\n",
    "    \n",
    "    def __init__(self, x , y):\n",
    "        self.x = x\n",
    "        self.y = y\n",
    "        self.h = fitness_value(x, y)\n",
    "        self.xbin = to_binary(x)\n",
    "        self.ybin = to_binary(y)\n",
    "        self.x_dec_place = to_binary_NoneSigin(x)\n",
    "        self.y_dec_place = to_binary_NoneSigin(y)\n",
    "        \n",
    "    "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 84,
   "id": "57ac7f40-07f0-4640-a8a1-5d85edd466c6",
   "metadata": {},
   "outputs": [],
   "source": [
    "def print_population(populations):\n",
    "    population = []\n",
    "    population = populations\n",
    "    for i in range(len(population)):\n",
    "        print(i+1,\":\\n\")\n",
    "        \n",
    "        print(\"x: \",population[i].x)\n",
    "        print(\"y: \",population[i].y)\n",
    "        print(\"h: \",population[i].h,end='\\n\\n')\n",
    "        \n",
    "#         print(\"x_binary: \",population[i].xbin,\"\\n\")\n",
    "#         print(\"y_binary: \",population[i].ybin,\"\\n\")\n",
    "#         print(population[i].y_dec_place)\n",
    "        "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 85,
   "id": "49e33646-6f6d-467f-9328-f407d5d11dcd",
   "metadata": {},
   "outputs": [],
   "source": [
    "def fitness_value(x, y):\n",
    "\n",
    "    value = -20.0 * np.exp(-0.2 * np.sqrt(0.5 * (x**2 + y**2))) - np.exp(0.5 * (np.cos(2 * np.pi * x) + np.cos(2 * np.pi * y))) + np.e + 20\n",
    "    \n",
    "    return value"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 86,
   "id": "9a9248ed-b853-46e3-a931-b06bc73f11e9",
   "metadata": {},
   "outputs": [],
   "source": [
    "def to_binary(num):\n",
    "    x = num\n",
    "    if x < 0 :\n",
    "        x = x * -1\n",
    "        \n",
    "    x_bin = \"{0:b}\".format(int(x))\n",
    "    x_str = ['0','0','0','0','0']\n",
    "    \n",
    "    j = 0\n",
    "    for i in range(len(x_str) - len(x_bin) , 5):\n",
    "        x_str[i] = x_bin[j]\n",
    "        j +=1\n",
    "        \n",
    "    if num >= 0 :\n",
    "        x_str.insert(0,'0')\n",
    "    else :\n",
    "        x_str.insert(0,'1')\n",
    "        \n",
    "    return x_str"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 87,
   "id": "d6605309-3d7a-46ba-ba89-934db59adf79",
   "metadata": {},
   "outputs": [],
   "source": [
    "\n",
    "def to_binary_NoneSigin(x): \n",
    "    \n",
    "    x = round(x,4)\n",
    "    x2 = str(x)    \n",
    "\n",
    "    temp = ''\n",
    "    for i in range(len(x2)):\n",
    "        if x2[i] == '.':\n",
    "            index = i\n",
    "            break\n",
    "\n",
    "    for i in range(index + 1 , len(x2)): \n",
    "        temp = temp + x2[i]\n",
    "        \n",
    "    while len(temp) < 4:\n",
    "        temp = '0' + str(temp)\n",
    "    \n",
    "    x_bin = \"{0:b}\".format(int(temp))\n",
    "    \n",
    "    x_str = ['0','0','0','0','0','0','0','0','0','0','0','0','0','0']\n",
    "    \n",
    "    j = 0\n",
    "    for i in range(len(x_str) - len(x_bin) , 14):\n",
    "        x_str[i] = x_bin[j]\n",
    "        j +=1\n",
    "        \n",
    "    return x_str\n",
    "    \n",
    "    \n",
    "    "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 88,
   "id": "1c33b22e",
   "metadata": {},
   "outputs": [],
   "source": [
    "def to_decimal(xbin, x_dec_place) :\n",
    "    \n",
    "    x = to_decimal_z(xbin)\n",
    "    x += to_decimal_place(x_dec_place)\n",
    "\n",
    "    if xbin[0] == '1':\n",
    "        x *= -1\n",
    "        \n",
    "    return x"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 89,
   "id": "252a6673",
   "metadata": {},
   "outputs": [],
   "source": [
    "def to_decimal_z(xbin) :\n",
    "    \n",
    "    x=''\n",
    "    for i in range(1,len(xbin)):\n",
    "        x = x[:len(xbin)] + xbin[i] + x[len(xbin):]\n",
    "     \n",
    "    x = int(x, 2)\n",
    "    \n",
    "    return x"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 90,
   "id": "3a1a322b",
   "metadata": {},
   "outputs": [],
   "source": [
    "def to_decimal_place(x_dec_place) :\n",
    "\n",
    "    x=''\n",
    "    for i in range(0,len(x_dec_place)):\n",
    "        x = x[:len(x_dec_place)] + x_dec_place[i] + x[len(x_dec_place):]\n",
    "    \n",
    "    x = int(x, 2)\n",
    "\n",
    "    return x * (10 ** -4)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 91,
   "id": "6958cf4d-cdee-41e8-a040-304cc9da45cd",
   "metadata": {},
   "outputs": [],
   "source": [
    "def first_population(size):\n",
    "    popu = []\n",
    "    for i in range(size):\n",
    "        x = random.uniform(-30 , 30)\n",
    "        y = random.uniform(-30 , 30)\n",
    "        popu.append(point(x,y))\n",
    "    \n",
    "    return popu"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 92,
   "id": "9dfd337d-275b-4556-aabf-56d84d10e27b",
   "metadata": {},
   "outputs": [],
   "source": [
    "def select_parents(population,size):\n",
    "    \n",
    "    All_Parent = []\n",
    "    sort_population = []\n",
    "    weights = []\n",
    "    sort_population = sorted(population, key=lambda p: p.h, reverse=True)\n",
    "    \n",
    "    for i in range(len(sort_population)):\n",
    "        weights.append(i+1)\n",
    "    \n",
    "    All_Parent = random.choices(sort_population, weights = weights , k = size)\n",
    "        \n",
    "    return All_Parent"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 93,
   "id": "ba109579",
   "metadata": {},
   "outputs": [],
   "source": [
    "def select_parents_computation(population,size):\n",
    "    \n",
    "    All_Parent = []\n",
    "    sort_population = []\n",
    "    \n",
    "    \n",
    "    for i in range(0,size) :\n",
    "        temp_population = population.copy()\n",
    "        random_population = []\n",
    "        for j in range(0,len(population)//5) :\n",
    "            index = random.randint(0,len(population)-1-j)\n",
    "            random_population.append(temp_population[index])\n",
    "            temp_population.pop(index)\n",
    "            \n",
    "        sort_population = sorted(random_population, key=lambda p: p.h, reverse=False)\n",
    "        All_Parent.append(sort_population[0])\n",
    "    \n",
    "        \n",
    "    return All_Parent"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 94,
   "id": "299bb210",
   "metadata": {},
   "outputs": [],
   "source": [
    "def cross_one_point_z(parent1, parent2) :\n",
    "    index = random.choice([1,2,3,4,5])\n",
    "    child = []\n",
    "\n",
    "    for i in range(0,len(parent1)) :\n",
    "        if i < index :\n",
    "            child.append(parent1[i])\n",
    "        else :\n",
    "            child.append(parent2[i])\n",
    "    \n",
    "    return child"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 95,
   "id": "edabdfde",
   "metadata": {},
   "outputs": [],
   "source": [
    "def cross_one_point_palce(parent1, parent2) :            \n",
    "    index = random.choice([0,1,2,3,4,5,6,7,8,9,10,11,12,13])\n",
    "    child = []\n",
    "\n",
    "    for i in range(0,len(parent1)) :\n",
    "        if i < index :\n",
    "            child.append(parent1[i])\n",
    "        else :\n",
    "            child.append(parent2[i])\n",
    "\n",
    "    return child"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 96,
   "id": "11ecd918",
   "metadata": {},
   "outputs": [],
   "source": [
    "def cross_uniform_z(parent1, parent2) :\n",
    "    child = []\n",
    "\n",
    "    for i in range(0,len(parent1)) :\n",
    "        if random.choice(['1', '2']) == '1' :\n",
    "            child.append(parent1[i])\n",
    "        else :\n",
    "            child.append(parent2[i])\n",
    "    \n",
    "    return child"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 97,
   "id": "c4fa26e7",
   "metadata": {},
   "outputs": [],
   "source": [
    "def cross_uniform_palce(parent1, parent2) :\n",
    "    child = []\n",
    "\n",
    "    for i in range(0,len(parent1)) :\n",
    "        if random.choice(['1', '2']) == '1' :\n",
    "            child.append(parent1[i])\n",
    "        else :\n",
    "            child.append(parent2[i])\n",
    "    \n",
    "    return child"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 98,
   "id": "0e0f3e6b",
   "metadata": {},
   "outputs": [],
   "source": [
    "def cross_over(parents) : \n",
    "    childeren = []\n",
    "    for i in range(0,len(parents)-1) :\n",
    "\n",
    "#        xbin_child1 = cross_one_point_z(parents[i].xbin,parents[i+1].xbin)\n",
    "        xbin_child1 = cross_uniform_z(parents[i].xbin,parents[i+1].xbin)\n",
    "        xbin_child2 = xbin_child1[::-1]\n",
    "#        ybin_child1 = cross_one_point_z(parents[i].ybin,parents[i+1].ybin)\n",
    "        ybin_child1 = cross_uniform_z(parents[i].ybin,parents[i+1].ybin)\n",
    "        ybin_child2 = ybin_child1[::-1]\n",
    "        \n",
    "#        x_dec_place_child1 = cross_one_point_palce(parents[i].x_dec_place,parents[i+1].x_dec_place)\n",
    "        x_dec_place_child1 = cross_uniform_palce(parents[i].x_dec_place,parents[i+1].x_dec_place)\n",
    "        x_dec_place_child2 = x_dec_place_child1[::-1]\n",
    "#        y_dec_place_child1 = cross_one_point_palce(parents[i].y_dec_place,parents[i+1].y_dec_place)\n",
    "        y_dec_place_child1 = cross_uniform_palce(parents[i].y_dec_place,parents[i+1].y_dec_place)\n",
    "        y_dec_place_child2 = y_dec_place_child1[::-1]\n",
    "        \n",
    "        x_child1 = to_decimal(xbin_child1,x_dec_place_child1)\n",
    "        y_child1 = to_decimal(ybin_child1,y_dec_place_child1)\n",
    "        x_child2 = to_decimal(xbin_child2,x_dec_place_child2)\n",
    "        y_child2 = to_decimal(ybin_child2,y_dec_place_child2)\n",
    "        \n",
    "        childeren.append(point(x_child1,y_child1))\n",
    "        childeren.append(point(x_child2,y_child2))\n",
    "        \n",
    "        i += 1\n",
    "        \n",
    "        \n",
    "    return childeren"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 99,
   "id": "ea1bd2fd",
   "metadata": {},
   "outputs": [],
   "source": [
    "def mutate(childeren):\n",
    "    mutatant = childeren\n",
    "    \n",
    "    for i in range(0,len(mutatant)) :\n",
    "        index = random.randint(0,len(mutatant[i].xbin)-1)\n",
    "        if random.choice(['0', '1']) == '1' :\n",
    "            if random.random() < 0.2:\n",
    "                mutatant[i].xbin[index] = random.choice(['0', '1'])\n",
    "        else :\n",
    "            if random.random() < 0.2:\n",
    "                mutatant[i].ybin[index] = random.choice(['0', '1'])\n",
    "    \n",
    "    \n",
    "    for i in range(0,len(mutatant)) :\n",
    "        index = random.randint(0,len(mutatant[i].x_dec_place)-1)\n",
    "        if random.choice(['0', '1']) == '1' :\n",
    "            if random.random() < 0.2:\n",
    "                mutatant[i].x_dec_place[index] = random.choice(['0', '1'])\n",
    "        else :\n",
    "            if random.random() < 0.2:\n",
    "                mutatant[i].y_dec_place[index] = random.choice(['0', '1'])\n",
    "        \n",
    "    return mutatant"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 100,
   "id": "aac9059d",
   "metadata": {},
   "outputs": [],
   "source": [
    "def replacement(population, parents, childeren) :\n",
    "    population_copy = population.copy()\n",
    "    parents_copy = parents.copy()\n",
    "    childeren_copy = childeren.copy()\n",
    "    \n",
    "    newlistChild = []\n",
    "    newlistParent = []\n",
    "    newlistChild = sorted(childeren_copy, key=lambda ch: ch.h, reverse=False)      #STEP 6\n",
    "    newlistParent = sorted(parents_copy, key=lambda p: p.h, reverse=True)\n",
    "\n",
    "    for i in range(0,int((len(parents_copy)*30/100)+0.5)) :\n",
    "        newlistParent[i] = newlistChild[i]\n",
    "    \n",
    "    newpopulation = newlistParent\n",
    "    sort_pop = sorted(population_copy, key=lambda p: p.h, reverse=True)\n",
    "    \n",
    "    for i in range(0,len(newpopulation)) :\n",
    "        sort_pop[i] = newpopulation[i]\n",
    "        \n",
    "    return sort_pop"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 101,
   "id": "15f28c42",
   "metadata": {},
   "outputs": [],
   "source": [
    "def GetBest(population) :\n",
    "    pop = sorted(population, key=lambda ch: ch.h, reverse=False)\n",
    "    print(\"x of best point in population: \",pop[0].x)\n",
    "    print(\"y of best point in population: \",pop[0].y)\n",
    "    print(\"fitness of best point in population: \",pop[0].h,end='\\n\\n')\n",
    "    return pop[0].h"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 105,
   "id": "7565d57c-e067-4ca4-a391-c9822d790104",
   "metadata": {},
   "outputs": [],
   "source": [
    "def main():\n",
    "    draw_fitness = []    #for draw result\n",
    "    cnt=0                # for count number of repeat best in generations\n",
    "    Best=100             # maximum of ackley\n",
    "    gen=0                # for count number of generations\n",
    "    target=0             # minimum of ackley\n",
    "    \n",
    "    firstpopulation = first_population(sx)   #STEP 1\n",
    "    current_population = firstpopulation    #STEP 2\n",
    "    \n",
    "    while True :\n",
    "        if gen > 1000 :\n",
    "            break\n",
    "            \n",
    "        if cnt == 800 :\n",
    "            break\n",
    "        \n",
    "        parents = select_parents(current_population,sx//2)   #STEP 3\n",
    "#        parents = select_parents_computation(current_population,sx//2)   #STEP 3\n",
    "#        print_population(current_population)\n",
    "\n",
    "        childeren = cross_over(parents)      #STEP 4\n",
    "\n",
    "        childeren = mutate(childeren)        #STEP 5\n",
    "        \n",
    "        next_population = replacement(current_population, parents, childeren)  #STEP 6        \n",
    "\n",
    "        BestNew = GetBest(current_population)      #STEP 7\n",
    "        \n",
    "        if Best == BestNew :\n",
    "            cnt += 1\n",
    "        elif Best > BestNew :\n",
    "            cnt=0\n",
    "            Best = BestNew\n",
    "        else :\n",
    "            cnt = 0\n",
    "        \n",
    "        print('gen :', gen)\n",
    "        print(\"Best (in gen \",gen,\") :\",Best)\n",
    "        print(\"********************\")\n",
    "        draw_fitness.append(Best)\n",
    "        if Best == target :\n",
    "            break\n",
    "            \n",
    "        current_population = next_population\n",
    "        gen += 1\n",
    "    hist = draw_fitness\n",
    "    plt.plot(hist)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 109,
   "id": "78d0b7d8-27da-4a5f-9725-a00cb27f4ecd",
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "x of best point in population:  0.9863671282114836\n",
      "y of best point in population:  -5.81294244388344\n",
      "fitness of best point in population:  12.03529647150577\n",
      "\n",
      "gen : 0\n",
      "Best (in gen  0 ) : 12.03529647150577\n",
      "********************\n",
      "x of best point in population:  -3.3236\n",
      "y of best point in population:  1.8826\n",
      "fitness of best point in population:  9.907275585399821\n",
      "\n",
      "gen : 1\n",
      "Best (in gen  1 ) : 9.907275585399821\n",
      "********************\n",
      "x of best point in population:  -0.8738\n",
      "y of best point in population:  -2.335\n",
      "fitness of best point in population:  7.559730340118854\n",
      "\n",
      "gen : 2\n",
      "Best (in gen  2 ) : 7.559730340118854\n",
      "********************\n",
      "x of best point in population:  1.0656\n",
      "y of best point in population:  -0.0993\n",
      "fitness of best point in population:  3.1549228312879407\n",
      "\n",
      "gen : 3\n",
      "Best (in gen  3 ) : 3.1549228312879407\n",
      "********************\n",
      "x of best point in population:  0.17200000000000001\n",
      "y of best point in population:  -0.0993\n",
      "fitness of best point in population:  1.3735484359229382\n",
      "\n",
      "gen : 4\n",
      "Best (in gen  4 ) : 1.3735484359229382\n",
      "********************\n",
      "x of best point in population:  -0.0176\n",
      "y of best point in population:  -0.0037\n",
      "fitness of best point in population:  0.059459166617973125\n",
      "\n",
      "gen : 5\n",
      "Best (in gen  5 ) : 0.059459166617973125\n",
      "********************\n",
      "x of best point in population:  -0.0176\n",
      "y of best point in population:  -0.0037\n",
      "fitness of best point in population:  0.059459166617973125\n",
      "\n",
      "gen : 6\n",
      "Best (in gen  6 ) : 0.059459166617973125\n",
      "********************\n",
      "x of best point in population:  -0.0176\n",
      "y of best point in population:  -0.0037\n",
      "fitness of best point in population:  0.059459166617973125\n",
      "\n",
      "gen : 7\n",
      "Best (in gen  7 ) : 0.059459166617973125\n",
      "********************\n",
      "x of best point in population:  -0.0054\n",
      "y of best point in population:  -0.0008\n",
      "fitness of best point in population:  0.0162335414824355\n",
      "\n",
      "gen : 8\n",
      "Best (in gen  8 ) : 0.0162335414824355\n",
      "********************\n",
      "x of best point in population:  -0.0008\n",
      "y of best point in population:  -0.0009000000000000001\n",
      "fitness of best point in population:  0.003444488048362615\n",
      "\n",
      "gen : 9\n",
      "Best (in gen  9 ) : 0.003444488048362615\n",
      "********************\n",
      "x of best point in population:  -0.0008\n",
      "y of best point in population:  -0.0009000000000000001\n",
      "fitness of best point in population:  0.003444488048362615\n",
      "\n",
      "gen : 10\n",
      "Best (in gen  10 ) : 0.003444488048362615\n",
      "********************\n",
      "x of best point in population:  -0.0008\n",
      "y of best point in population:  -0.0009000000000000001\n",
      "fitness of best point in population:  0.003444488048362615\n",
      "\n",
      "gen : 11\n",
      "Best (in gen  11 ) : 0.003444488048362615\n",
      "********************\n",
      "x of best point in population:  -0.0008\n",
      "y of best point in population:  -0.0009000000000000001\n",
      "fitness of best point in population:  0.003444488048362615\n",
      "\n",
      "gen : 12\n",
      "Best (in gen  12 ) : 0.003444488048362615\n",
      "********************\n",
      "x of best point in population:  -0.0001\n",
      "y of best point in population:  -0.0001\n",
      "fitness of best point in population:  0.00040053256728356246\n",
      "\n",
      "gen : 13\n",
      "Best (in gen  13 ) : 0.00040053256728356246\n",
      "********************\n",
      "x of best point in population:  -0.0\n",
      "y of best point in population:  -0.0\n",
      "fitness of best point in population:  0.0\n",
      "\n",
      "gen : 14\n",
      "Best (in gen  14 ) : 0.0\n",
      "********************\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "main()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "62bd5daf",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "84a20dfa-b32d-4da2-b701-099a9a2eb597",
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "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": "3.9.6"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}