PredictionDemo.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<H1> Demo notebook </H1>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "\"\"\"\n",
    "IMPORTS\n",
    "\"\"\"\n",
    "from genetic_algorithm.population import *\n",
    "from genetic_algorithm.ga_operations import *\n",
    "from genetic_algorithm.algorithm import *\n",
    "import numpy as np\n",
    "import matplotlib.pyplot as plt\n",
    "%matplotlib notebook"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<b>Algorithm parameters</b>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "\"\"\"\n",
    "ALGORITHM PARAMETERS\n",
    "\"\"\"\n",
    "\n",
    "SIZE = 1 # number of variables\n",
    "MAX_DEPTH = 20 #maximum depth of a tree \n",
    "FUNCTIONS = {1: ['sin','cos','e','ln','tg','tanh','abs'], 2:['+', '-', '*', '/']} #unary and binary functions\n",
    "TERMINAL_SET = ['x'+str(i) for i in range(SIZE)] #create parameters name\n",
    "DEPTH = 4 #initial depth of trees\n",
    "POP_SIZE = 1000 #number of chromosomes in population\n",
    "NUM_FOR_SELECTION = 1000//3 #number of chromosomes to be chosen for selection\n",
    "NUMBER_ITERS = 1000 #number of iterations\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<b>Our first predicted function</b>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Best function: ['*', 'x0', 'x0']\n",
      "Best fitness: 133.1334000000001\n",
      "Best function: ['+', 'x0', '+', 'x0', '*', 'x0', 'x0']\n",
      "Best fitness: 2.3274238654394036e-29\n"
     ]
    }
   ],
   "source": [
    "\n",
    "\"\"\"\n",
    "DEFINE A FUNCTION TO PREDICT\n",
    "\"\"\"\n",
    "def f(x):\n",
    "    return x**2 + 2*x\n",
    "\n",
    "X = [[x] for x in np.arange(0, 10, 0.01)] #function inputs\n",
    "y = [[f(x[0])] for x in X] #function outputs\n",
    "\n",
    "pop = Population(POP_SIZE, NUM_FOR_SELECTION, FUNCTIONS, TERMINAL_SET, 6, MAX_DEPTH) #create the population\n",
    "alg = Algorithm(pop, NUMBER_ITERS, X, y, epoch_feedback=500)  #create the algorithm\n",
    "best = alg.train() #train the alrogithm\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<b>Graphs are overlaping since we exactly approximated the function</b>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "['+', 'x0', '+', 'x0', '*', 'x0', 'x0']\n",
      "2.3274238654394036e-29\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": [
    "print(best.gen) #print best function\n",
    "print(best.fitness)\n",
    "\n",
    "y_pred = [[best.evaluate_arg(x)] for x in X] #predictions of our best calculated function\n",
    " \n",
    "plt.plot(X, y, color='r', dashes=[6, 2], label='Expected')  #plot original function \n",
    "plt.plot(X, y_pred, color='g', dashes=[6, 3], label='Predicted') #plot calculated function\n",
    "plt.legend()\n",
    "plt.show() #show the plot"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Best function: ['tg', 'tanh', '*', 'e', 'tanh', 'cos', 'x0', '*', '*', '+', 'x0', 'x0', '*', 'x0', 'x0', '/', 'tanh', 'x0', 'sin', 'x0']\n",
      "Best fitness: 2.4939762383752933\n",
      "Best function: ['tg', 'tanh', '+', '/', 'sin', 'x0', 'sin', 'x0', '*', '+', 'sin', 'x0', 'abs', '*', '-', 'x0', 'abs', 'x0', '*', '/', 'sin', '*', '-', 'x0', 'abs', 'tg', 'x0', '*', '/', 'sin', 'x0', 'tg', 'x0', 'tanh', 'x0', 'tg', 'x0', 'tanh', '+', '/', 'tg', 'x0', 'tg', 'x0', 'tanh', 'x0', '+', '/', '+', '*', 'sin', 'x0', 'x0', 'tanh', '+', '/', '+', '/', 'sin', 'x0', 'x0', 'tanh', 'x0', 'x0', 'x0', 'x0', 'tanh', 'x0']\n",
      "Best fitness: 0.4621097514586274\n"
     ]
    }
   ],
   "source": [
    "\"\"\"\n",
    "DEFINE A FUNCTION TO PREDICT\n",
    "\"\"\"\n",
    "def f(x):\n",
    "    return np.exp(np.sin(x)) + np.sin(x) / x \n",
    "\n",
    "X = [[x] for x in np.arange(0.1, 30, 0.01)] #function inputs\n",
    "y = [[f(x[0])] for x in X] #function outputs\n",
    "\n",
    "alg = Algorithm(pop, NUMBER_ITERS, X, y, epoch_feedback=500)  #create the algorithm\n",
    "best = alg.train() #train the alrogithm\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<b>As we can see the predicted function has the same tendancy as the original function, a good approximation for only 1000 iterations</b>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<b></b>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "['e', 'tanh', '+', '/', 'sin', 'x0', 'x0', '*', '+', 'sin', 'x0', 'abs', '*', '-', 'x0', 'abs', 'x0', '*', '/', 'sin', '*', '-', 'x0', 'abs', 'tg', 'x0', '*', '/', 'sin', 'x0', 'tg', 'x0', 'tanh', 'x0', 'tg', 'x0', 'tanh', '+', '/', 'tg', 'x0', 'tg', 'x0', 'tanh', 'x0', '+', '/', '+', '/', 'sin', 'x0', 'x0', 'tanh', '+', '/', '+', '/', 'sin', 'x0', 'x0', 'tanh', 'x0', 'x0', 'tanh', 'x0', 'x0', 'tanh', 'x0']\n",
      "0.055763381492797935\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": [
    "print(best.gen) #print best function\n",
    "print(best.fitness)\n",
    "y_pred = [[best.evaluate_arg(x)] for x in X] #predictions of our best calculated function\n",
    " \n",
    "plt.plot(X, y, color='r', dashes=[6, 2], label='Expected')  #plot original function \n",
    "plt.plot(X, y_pred, color='g', dashes=[6, 3], label='Predicted') #plot calculated function\n",
    "plt.show() #show the plot"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python [conda env:AI_tenosr] *",
   "language": "python",
   "name": "conda-env-AI_tenosr-py"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.6.8"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}