Four Peaks Problem.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import time\n",
    "import mlrose_hiive as mlrh\n",
    "import numpy as np\n",
    "import pandas as pd\n",
    "import matplotlib as mpl\n",
    "import matplotlib.pyplot as plt\n",
    "%matplotlib inline\n",
    "plt.rcParams['figure.figsize'] = [7, 7]\n",
    "\n",
    "random_state = 2020\n",
    "np.random.seed(1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "def plot_fitness_iteration(curve, title, x=\"Iteration\", y=\"Fitness\"):\n",
    "    plt.xlabel(x)\n",
    "    plt.ylabel(y)\n",
    "    plt.title(title)\n",
    "    length = len(curve)\n",
    "    plt.plot(range(length), curve, label=y, lw=2)\n",
    "    plt.legend(loc=\"best\")\n",
    "    plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Four Peaks Problem"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Problem definition\n",
    "length = 150\n",
    "t_pct = 0.9\n",
    "global eval_count\n",
    "eval_count = 0\n",
    "\n",
    "base_fit_func = mlrh.FourPeaks(t_pct=t_pct).evaluate\n",
    "\n",
    "def four_peaks_fit(state):\n",
    "    global eval_count\n",
    "    eval_count += 1\n",
    "\n",
    "    return base_fit_func(state)\n",
    "\n",
    "\n",
    "# Initialize custom fitness function object\n",
    "fitness = mlrh.CustomFitness(four_peaks_fit)\n",
    "\n",
    "prob = mlrh.DiscreteOpt(length=length,\n",
    "                        fitness_fn=fitness,\n",
    "                        maximize=True)\n",
    "experiment_name = \"four_peaks\"\n",
    "output_directory = \"four_peaks\""
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## MIMIC"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Running mimic\n",
      "*****************\n",
      "*** Run START ***\n",
      "*****************\n",
      "use_fast_mimic:[True], pop_size:[1000], keep_pct:[0.05], max_iters:[25]\n",
      "runner_name:[mimic], experiment_name:[four_peaks], attempt:[0], iteration:[0], done:[False], time:[0.05], fitness:[0.0000]\n",
      "\t[0 0 1 0 1 1 1 1 1 0 1 0 0 0 0 1 1 1 1 0 0 1 0 1 0 0 0 0 0 0 1 1 1 1 0 1 0// 1 1 0 0 0 0 1 1 0 0 0 0 0 1 0 1 1 0 1 0 1 0 0 1 1 0 0 1 1 0 0 0 0 1 0 0 0// 1 1 0 1 1 0 0 0 1 0 0 0 0 0 0 0 0 1 0 1 1 1 1 1...\n",
      "\n",
      "use_fast_mimic:[True], pop_size:[1000], keep_pct:[0.05], max_iters:[25]\n",
      "runner_name:[mimic], experiment_name:[four_peaks], attempt:[18], iteration:[25], done:[True], time:[12.80], fitness:[32.0000]\n",
      "\t[1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1.// 1. 1. 1. 1. 1. 1. 1. 1. 0. 1. 1. 1. 0. 1. 0. 1. 1. 0. 0. 1. 0. 0. 1. 0.// 0. 1. 1. 0. 1. 1. 0. 1. 1. 1. 0. 1. 1. 0. 0. 0. 0. ...\n",
      "\n",
      "***************\n",
      "*** Run END ***\n",
      "***************\n",
      "*****************\n",
      "*** Run START ***\n",
      "*****************\n",
      "use_fast_mimic:[True], pop_size:[1000], keep_pct:[0.1], max_iters:[25]\n",
      "runner_name:[mimic], experiment_name:[four_peaks], attempt:[0], iteration:[0], done:[False], time:[0.04], fitness:[0.0000]\n",
      "\t[0 0 1 0 1 1 1 1 1 0 1 0 0 0 0 1 1 1 1 0 0 1 0 1 0 0 0 0 0 0 1 1 1 1 0 1 0// 1 1 0 0 0 0 1 1 0 0 0 0 0 1 0 1 1 0 1 0 1 0 0 1 1 0 0 1 1 0 0 0 0 1 0 0 0// 1 1 0 1 1 0 0 0 1 0 0 0 0 0 0 0 0 1 0 1 1 1 1 1...\n",
      "\n",
      "use_fast_mimic:[True], pop_size:[1000], keep_pct:[0.1], max_iters:[25]\n",
      "runner_name:[mimic], experiment_name:[four_peaks], attempt:[6], iteration:[25], done:[True], time:[5.92], fitness:[63.0000]\n",
      "\t[1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1.// 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1.// 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 0. 1. ...\n",
      "\n",
      "***************\n",
      "*** Run END ***\n",
      "***************\n",
      "*****************\n",
      "*** Run START ***\n",
      "*****************\n",
      "use_fast_mimic:[True], pop_size:[1000], keep_pct:[0.15], max_iters:[25]\n",
      "runner_name:[mimic], experiment_name:[four_peaks], attempt:[0], iteration:[0], done:[False], time:[0.04], fitness:[0.0000]\n",
      "\t[0 0 1 0 1 1 1 1 1 0 1 0 0 0 0 1 1 1 1 0 0 1 0 1 0 0 0 0 0 0 1 1 1 1 0 1 0// 1 1 0 0 0 0 1 1 0 0 0 0 0 1 0 1 1 0 1 0 1 0 0 1 1 0 0 1 1 0 0 0 0 1 0 0 0// 1 1 0 1 1 0 0 0 1 0 0 0 0 0 0 0 0 1 0 1 1 1 1 1...\n",
      "\n",
      "use_fast_mimic:[True], pop_size:[1000], keep_pct:[0.15], max_iters:[25]\n",
      "runner_name:[mimic], experiment_name:[four_peaks], attempt:[2], iteration:[25], done:[True], time:[5.99], fitness:[56.0000]\n",
      "\t[1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1.// 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1.// 1. 1. 1. 1. 1. 1. 1. 1. 0. 1. 1. 0. 1. 0. 0. 0. 1. ...\n",
      "\n",
      "***************\n",
      "*** Run END ***\n",
      "***************\n",
      "*****************\n",
      "*** Run START ***\n",
      "*****************\n",
      "use_fast_mimic:[True], pop_size:[1000], keep_pct:[0.2], max_iters:[25]\n",
      "runner_name:[mimic], experiment_name:[four_peaks], attempt:[0], iteration:[0], done:[False], time:[0.05], fitness:[0.0000]\n",
      "\t[0 0 1 0 1 1 1 1 1 0 1 0 0 0 0 1 1 1 1 0 0 1 0 1 0 0 0 0 0 0 1 1 1 1 0 1 0// 1 1 0 0 0 0 1 1 0 0 0 0 0 1 0 1 1 0 1 0 1 0 0 1 1 0 0 1 1 0 0 0 0 1 0 0 0// 1 1 0 1 1 0 0 0 1 0 0 0 0 0 0 0 0 1 0 1 1 1 1 1...\n",
      "\n",
      "use_fast_mimic:[True], pop_size:[1000], keep_pct:[0.2], max_iters:[25]\n",
      "runner_name:[mimic], experiment_name:[four_peaks], attempt:[5], iteration:[25], done:[True], time:[8.53], fitness:[52.0000]\n",
      "\t[1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1.// 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1.// 1. 1. 1. 1. 0. 1. 0. 1. 1. 0. 1. 1. 0. 1. 1. 0. 0. ...\n",
      "\n",
      "***************\n",
      "*** Run END ***\n",
      "***************\n",
      "*****************\n",
      "*** Run START ***\n",
      "*****************\n",
      "use_fast_mimic:[True], pop_size:[2000], keep_pct:[0.05], max_iters:[25]\n",
      "runner_name:[mimic], experiment_name:[four_peaks], attempt:[0], iteration:[0], done:[False], time:[0.09], fitness:[0.0000]\n",
      "\t[0 0 1 0 1 1 1 1 1 0 1 0 0 0 0 1 1 1 1 0 0 1 0 1 0 0 0 0 0 0 1 1 1 1 0 1 0// 1 1 0 0 0 0 1 1 0 0 0 0 0 1 0 1 1 0 1 0 1 0 0 1 1 0 0 1 1 0 0 0 0 1 0 0 0// 1 1 0 1 1 0 0 0 1 0 0 0 0 0 0 0 0 1 0 1 1 1 1 1...\n",
      "\n",
      "use_fast_mimic:[True], pop_size:[2000], keep_pct:[0.05], max_iters:[25]\n",
      "runner_name:[mimic], experiment_name:[four_peaks], attempt:[1], iteration:[25], done:[True], time:[11.23], fitness:[79.0000]\n",
      "\t[1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1.// 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1.// 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. ...\n",
      "\n",
      "***************\n",
      "*** Run END ***\n",
      "***************\n",
      "*****************\n",
      "*** Run START ***\n",
      "*****************\n",
      "use_fast_mimic:[True], pop_size:[2000], keep_pct:[0.1], max_iters:[25]\n",
      "runner_name:[mimic], experiment_name:[four_peaks], attempt:[0], iteration:[0], done:[False], time:[0.09], fitness:[0.0000]\n",
      "\t[0 0 1 0 1 1 1 1 1 0 1 0 0 0 0 1 1 1 1 0 0 1 0 1 0 0 0 0 0 0 1 1 1 1 0 1 0// 1 1 0 0 0 0 1 1 0 0 0 0 0 1 0 1 1 0 1 0 1 0 0 1 1 0 0 1 1 0 0 0 0 1 0 0 0// 1 1 0 1 1 0 0 0 1 0 0 0 0 0 0 0 0 1 0 1 1 1 1 1...\n",
      "\n",
      "use_fast_mimic:[True], pop_size:[2000], keep_pct:[0.1], max_iters:[25]\n",
      "runner_name:[mimic], experiment_name:[four_peaks], attempt:[1], iteration:[25], done:[True], time:[7.81], fitness:[78.0000]\n",
      "\t[1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1.// 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1.// 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. ...\n",
      "\n",
      "***************\n",
      "*** Run END ***\n",
      "***************\n",
      "*****************\n",
      "*** Run START ***\n",
      "*****************\n",
      "use_fast_mimic:[True], pop_size:[2000], keep_pct:[0.15], max_iters:[25]\n",
      "runner_name:[mimic], experiment_name:[four_peaks], attempt:[0], iteration:[0], done:[False], time:[0.07], fitness:[0.0000]\n",
      "\t[0 0 1 0 1 1 1 1 1 0 1 0 0 0 0 1 1 1 1 0 0 1 0 1 0 0 0 0 0 0 1 1 1 1 0 1 0// 1 1 0 0 0 0 1 1 0 0 0 0 0 1 0 1 1 0 1 0 1 0 0 1 1 0 0 1 1 0 0 0 0 1 0 0 0// 1 1 0 1 1 0 0 0 1 0 0 0 0 0 0 0 0 1 0 1 1 1 1 1...\n",
      "\n",
      "use_fast_mimic:[True], pop_size:[2000], keep_pct:[0.15], max_iters:[25]\n",
      "runner_name:[mimic], experiment_name:[four_peaks], attempt:[1], iteration:[25], done:[True], time:[10.02], fitness:[65.0000]\n",
      "\t[1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1.// 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1.// 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. ...\n",
      "\n",
      "***************\n",
      "*** Run END ***\n",
      "***************\n",
      "*****************\n",
      "*** Run START ***\n",
      "*****************\n",
      "use_fast_mimic:[True], pop_size:[2000], keep_pct:[0.2], max_iters:[25]\n",
      "runner_name:[mimic], experiment_name:[four_peaks], attempt:[0], iteration:[0], done:[False], time:[0.09], fitness:[0.0000]\n",
      "\t[0 0 1 0 1 1 1 1 1 0 1 0 0 0 0 1 1 1 1 0 0 1 0 1 0 0 0 0 0 0 1 1 1 1 0 1 0// 1 1 0 0 0 0 1 1 0 0 0 0 0 1 0 1 1 0 1 0 1 0 0 1 1 0 0 1 1 0 0 0 0 1 0 0 0// 1 1 0 1 1 0 0 0 1 0 0 0 0 0 0 0 0 1 0 1 1 1 1 1...\n",
      "\n",
      "use_fast_mimic:[True], pop_size:[2000], keep_pct:[0.2], max_iters:[25]\n",
      "runner_name:[mimic], experiment_name:[four_peaks], attempt:[1], iteration:[25], done:[True], time:[12.24], fitness:[57.0000]\n",
      "\t[1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1.// 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1.// 1. 1. 1. 1. 1. 1. 1. 1. 1. 0. 0. 0. 1. 0. 1. 0. 0. ...\n",
      "\n",
      "***************\n",
      "*** Run END ***\n",
      "***************\n",
      "*****************\n",
      "*** Run START ***\n",
      "*****************\n",
      "use_fast_mimic:[True], pop_size:[3000], keep_pct:[0.05], max_iters:[25]\n",
      "runner_name:[mimic], experiment_name:[four_peaks], attempt:[0], iteration:[0], done:[False], time:[0.15], fitness:[0.0000]\n",
      "\t[0 0 1 0 1 1 1 1 1 0 1 0 0 0 0 1 1 1 1 0 0 1 0 1 0 0 0 0 0 0 1 1 1 1 0 1 0// 1 1 0 0 0 0 1 1 0 0 0 0 0 1 0 1 1 0 1 0 1 0 0 1 1 0 0 1 1 0 0 0 0 1 0 0 0// 1 1 0 1 1 0 0 0 1 0 0 0 0 0 0 0 0 1 0 1 1 1 1 1...\n",
      "\n",
      "use_fast_mimic:[True], pop_size:[3000], keep_pct:[0.05], max_iters:[25]\n",
      "runner_name:[mimic], experiment_name:[four_peaks], attempt:[3], iteration:[25], done:[True], time:[25.56], fitness:[66.0000]\n",
      "\t[1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1.// 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1.// 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. ...\n",
      "\n",
      "***************\n",
      "*** Run END ***\n",
      "***************\n",
      "*****************\n",
      "*** Run START ***\n",
      "*****************\n",
      "use_fast_mimic:[True], pop_size:[3000], keep_pct:[0.1], max_iters:[25]\n",
      "runner_name:[mimic], experiment_name:[four_peaks], attempt:[0], iteration:[0], done:[False], time:[0.14], fitness:[0.0000]\n",
      "\t[0 0 1 0 1 1 1 1 1 0 1 0 0 0 0 1 1 1 1 0 0 1 0 1 0 0 0 0 0 0 1 1 1 1 0 1 0// 1 1 0 0 0 0 1 1 0 0 0 0 0 1 0 1 1 0 1 0 1 0 0 1 1 0 0 1 1 0 0 0 0 1 0 0 0// 1 1 0 1 1 0 0 0 1 0 0 0 0 0 0 0 0 1 0 1 1 1 1 1...\n",
      "\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "use_fast_mimic:[True], pop_size:[3000], keep_pct:[0.1], max_iters:[25]\n",
      "runner_name:[mimic], experiment_name:[four_peaks], attempt:[1], iteration:[25], done:[True], time:[11.50], fitness:[82.0000]\n",
      "\t[1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1.// 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1.// 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. ...\n",
      "\n",
      "***************\n",
      "*** Run END ***\n",
      "***************\n",
      "*****************\n",
      "*** Run START ***\n",
      "*****************\n",
      "use_fast_mimic:[True], pop_size:[3000], keep_pct:[0.15], max_iters:[25]\n",
      "runner_name:[mimic], experiment_name:[four_peaks], attempt:[0], iteration:[0], done:[False], time:[0.13], fitness:[0.0000]\n",
      "\t[0 0 1 0 1 1 1 1 1 0 1 0 0 0 0 1 1 1 1 0 0 1 0 1 0 0 0 0 0 0 1 1 1 1 0 1 0// 1 1 0 0 0 0 1 1 0 0 0 0 0 1 0 1 1 0 1 0 1 0 0 1 1 0 0 1 1 0 0 0 0 1 0 0 0// 1 1 0 1 1 0 0 0 1 0 0 0 0 0 0 0 0 1 0 1 1 1 1 1...\n",
      "\n",
      "use_fast_mimic:[True], pop_size:[3000], keep_pct:[0.15], max_iters:[25]\n",
      "runner_name:[mimic], experiment_name:[four_peaks], attempt:[1], iteration:[25], done:[True], time:[16.16], fitness:[67.0000]\n",
      "\t[1. 1. 0. 1. 1. 0. 1. 1. 0. 1. 1. 1. 0. 0. 0. 0. 0. 0. 1. 1. 1. 1. 1. 0.// 1. 1. 1. 0. 1. 1. 0. 1. 0. 0. 0. 1. 1. 1. 1. 1. 1. 1. 1. 0. 0. 0. 1. 1.// 1. 0. 0. 0. 1. 1. 0. 1. 1. 1. 1. 0. 0. 0. 0. 1. 0. ...\n",
      "\n",
      "***************\n",
      "*** Run END ***\n",
      "***************\n",
      "*****************\n",
      "*** Run START ***\n",
      "*****************\n",
      "use_fast_mimic:[True], pop_size:[3000], keep_pct:[0.2], max_iters:[25]\n",
      "runner_name:[mimic], experiment_name:[four_peaks], attempt:[0], iteration:[0], done:[False], time:[0.15], fitness:[0.0000]\n",
      "\t[0 0 1 0 1 1 1 1 1 0 1 0 0 0 0 1 1 1 1 0 0 1 0 1 0 0 0 0 0 0 1 1 1 1 0 1 0// 1 1 0 0 0 0 1 1 0 0 0 0 0 1 0 1 1 0 1 0 1 0 0 1 1 0 0 1 1 0 0 0 0 1 0 0 0// 1 1 0 1 1 0 0 0 1 0 0 0 0 0 0 0 0 1 0 1 1 1 1 1...\n",
      "\n",
      "use_fast_mimic:[True], pop_size:[3000], keep_pct:[0.2], max_iters:[25]\n",
      "runner_name:[mimic], experiment_name:[four_peaks], attempt:[1], iteration:[25], done:[True], time:[17.44], fitness:[60.0000]\n",
      "\t[1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1.// 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1.// 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 0. 0. 1. 1. 1. ...\n",
      "\n",
      "***************\n",
      "*** Run END ***\n",
      "***************\n",
      "Run time: 145.34595933998935\n",
      "Saving: [four_peaks/four_peaks/mimic__four_peaks__run_stats_df.csv]\n",
      "Saving: [four_peaks/four_peaks/mimic__four_peaks__curves_df.csv]\n"
     ]
    }
   ],
   "source": [
    "mim = mlrh.MIMICRunner(problem=prob,\n",
    "                       experiment_name=experiment_name,\n",
    "                       output_directory=output_directory,\n",
    "                       seed=random_state,\n",
    "                       population_sizes=[1000, 2000, 3000],\n",
    "                       keep_percent_list=[0.05, 0.1, 0.15, 0.20],\n",
    "                       iteration_list=[25],\n",
    "                       max_attempts=100,\n",
    "                       use_fast_mimic=True)\n",
    "mim_stats, mim_curve = mim.run()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "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>Time</th>\n",
       "      <th>Fitness</th>\n",
       "      <th>Population Size</th>\n",
       "      <th>Keep Percent</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>19</th>\n",
       "      <td>11.504021</td>\n",
       "      <td>82.0</td>\n",
       "      <td>3000</td>\n",
       "      <td>0.10</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>11.225152</td>\n",
       "      <td>79.0</td>\n",
       "      <td>2000</td>\n",
       "      <td>0.05</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11</th>\n",
       "      <td>7.807094</td>\n",
       "      <td>78.0</td>\n",
       "      <td>2000</td>\n",
       "      <td>0.10</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>21</th>\n",
       "      <td>16.162705</td>\n",
       "      <td>67.0</td>\n",
       "      <td>3000</td>\n",
       "      <td>0.15</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>17</th>\n",
       "      <td>25.564504</td>\n",
       "      <td>66.0</td>\n",
       "      <td>3000</td>\n",
       "      <td>0.05</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>13</th>\n",
       "      <td>10.018241</td>\n",
       "      <td>65.0</td>\n",
       "      <td>2000</td>\n",
       "      <td>0.15</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>5.917637</td>\n",
       "      <td>63.0</td>\n",
       "      <td>1000</td>\n",
       "      <td>0.10</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>23</th>\n",
       "      <td>17.442336</td>\n",
       "      <td>60.0</td>\n",
       "      <td>3000</td>\n",
       "      <td>0.20</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>15</th>\n",
       "      <td>12.241212</td>\n",
       "      <td>57.0</td>\n",
       "      <td>2000</td>\n",
       "      <td>0.20</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>5.990759</td>\n",
       "      <td>56.0</td>\n",
       "      <td>1000</td>\n",
       "      <td>0.15</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "         Time  Fitness  Population Size  Keep Percent\n",
       "19  11.504021     82.0             3000          0.10\n",
       "9   11.225152     79.0             2000          0.05\n",
       "11   7.807094     78.0             2000          0.10\n",
       "21  16.162705     67.0             3000          0.15\n",
       "17  25.564504     66.0             3000          0.05\n",
       "13  10.018241     65.0             2000          0.15\n",
       "3    5.917637     63.0             1000          0.10\n",
       "23  17.442336     60.0             3000          0.20\n",
       "15  12.241212     57.0             2000          0.20\n",
       "5    5.990759     56.0             1000          0.15"
      ]
     },
     "execution_count": 14,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "show=10\n",
    "columns = ['Time', 'Fitness', 'Population Size', 'Keep Percent']\n",
    "pd.set_option('display.max_rows',show)\n",
    "df=pd.read_csv(\"./four_peaks/four_peaks/mimic__four_peaks__run_stats_df.csv\")\n",
    "df[columns].sort_values(by=['Fitness'], ascending=False).head(show)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "MIMIC - Total Function Evaluations: 78043\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 504x504 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "max_attempts = 100\n",
    "max_iters = 25\n",
    "keep_pct=0.1\n",
    "pop_size = 3000\n",
    "global eval_count\n",
    "eval_count = 0\n",
    "best_state, best_fitness, mimic_curve = mlrh.mimic(prob,\n",
    "                                             max_attempts=max_attempts,\n",
    "                                             max_iters=max_iters,\n",
    "                                             random_state=random_state,\n",
    "                                             pop_size=pop_size,\n",
    "                                             keep_pct=keep_pct,\n",
    "                                             curve=True)\n",
    "print(\"MIMIC - Total Function Evaluations:\", eval_count)\n",
    "plot_fitness_iteration(mimic_curve,\n",
    "                       \"Four Peaks - Mimic: Population Size: {}, Keep Percent: {}\".format(pop_size, keep_pct))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Genetic Algorithm"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Running ga\n",
      "*****************\n",
      "*** Run START ***\n",
      "*****************\n",
      "pop_size:[500], mutation_prob:[0.1], max_iters:[25]\n",
      "runner_name:[ga], experiment_name:[four_peaks], attempt:[0], iteration:[0], done:[False], time:[0.03], fitness:[0.0000]\n",
      "\t[0 0 1 0 1 1 1 1 1 0 1 0 0 0 0 1 1 1 1 0 0 1 0 1 0 0 0 0 0 0 1 1 1 1 0 1 0// 1 1 0 0 0 0 1 1 0 0 0 0 0 1 0 1 1 0 1 0 1 0 0 1 1 0 0 1 1 0 0 0 0 1 0 0 0// 1 1 0 1 1 0 0 0 1 0 0 0 0 0 0 0 0 1 0 1 1 1 1 1...\n",
      "\n",
      "pop_size:[500], mutation_prob:[0.1], max_iters:[25]\n",
      "runner_name:[ga], experiment_name:[four_peaks], attempt:[2], iteration:[25], done:[True], time:[2.41], fitness:[28.0000]\n",
      "\t[1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 0 1 1 0 1 0 1 1 1// 1 0 1 0 1 1 1 0 1 0 0 1 1 0 0 0 0 1 1 0 0 1 1 0 0 0 0 1 0 1 1 1 0 0 0 1 1// 0 1 1 0 0 0 0 1 0 0 1 0 1 0 1 0 0 0 0 1 0 1 0 1...\n",
      "\n",
      "***************\n",
      "*** Run END ***\n",
      "***************\n",
      "*****************\n",
      "*** Run START ***\n",
      "*****************\n",
      "pop_size:[500], mutation_prob:[0.25], max_iters:[25]\n",
      "runner_name:[ga], experiment_name:[four_peaks], attempt:[0], iteration:[0], done:[False], time:[0.02], fitness:[0.0000]\n",
      "\t[0 0 1 0 1 1 1 1 1 0 1 0 0 0 0 1 1 1 1 0 0 1 0 1 0 0 0 0 0 0 1 1 1 1 0 1 0// 1 1 0 0 0 0 1 1 0 0 0 0 0 1 0 1 1 0 1 0 1 0 0 1 1 0 0 1 1 0 0 0 0 1 0 0 0// 1 1 0 1 1 0 0 0 1 0 0 0 0 0 0 0 0 1 0 1 1 1 1 1...\n",
      "\n",
      "pop_size:[500], mutation_prob:[0.25], max_iters:[25]\n",
      "runner_name:[ga], experiment_name:[four_peaks], attempt:[1], iteration:[25], done:[True], time:[2.33], fitness:[32.0000]\n",
      "\t[1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 0 0 1 0 0// 0 1 0 1 0 1 1 0 0 1 0 1 0 0 1 1 0 1 0 0 0 1 1 1 0 1 0 0 1 1 1 1 0 1 0 0 1// 1 1 0 0 0 0 0 0 0 1 0 1 1 1 0 0 0 1 1 0 0 0 0 1...\n",
      "\n",
      "***************\n",
      "*** Run END ***\n",
      "***************\n",
      "*****************\n",
      "*** Run START ***\n",
      "*****************\n",
      "pop_size:[500], mutation_prob:[0.5], max_iters:[25]\n",
      "runner_name:[ga], experiment_name:[four_peaks], attempt:[0], iteration:[0], done:[False], time:[0.02], fitness:[0.0000]\n",
      "\t[0 0 1 0 1 1 1 1 1 0 1 0 0 0 0 1 1 1 1 0 0 1 0 1 0 0 0 0 0 0 1 1 1 1 0 1 0// 1 1 0 0 0 0 1 1 0 0 0 0 0 1 0 1 1 0 1 0 1 0 0 1 1 0 0 1 1 0 0 0 0 1 0 0 0// 1 1 0 1 1 0 0 0 1 0 0 0 0 0 0 0 0 1 0 1 1 1 1 1...\n",
      "\n",
      "pop_size:[500], mutation_prob:[0.5], max_iters:[25]\n",
      "runner_name:[ga], experiment_name:[four_peaks], attempt:[4], iteration:[25], done:[True], time:[2.30], fitness:[33.0000]\n",
      "\t[1 0 1 1 0 0 0 1 1 0 1 1 1 1 1 1 0 0 1 0 0 1 1 1 0 1 1 0 0 1 0 1 0 0 0 1 0// 0 1 0 1 1 0 0 0 1 1 0 1 1 1 1 0 0 1 1 1 1 0 1 1 1 1 1 1 0 0 0 0 0 1 1 1 0// 0 1 0 0 1 1 0 0 1 1 0 1 1 1 1 1 0 1 1 1 1 0 0 1...\n",
      "\n",
      "***************\n",
      "*** Run END ***\n",
      "***************\n",
      "*****************\n",
      "*** Run START ***\n",
      "*****************\n",
      "pop_size:[500], mutation_prob:[0.75], max_iters:[25]\n",
      "runner_name:[ga], experiment_name:[four_peaks], attempt:[0], iteration:[0], done:[False], time:[0.03], fitness:[0.0000]\n",
      "\t[0 0 1 0 1 1 1 1 1 0 1 0 0 0 0 1 1 1 1 0 0 1 0 1 0 0 0 0 0 0 1 1 1 1 0 1 0// 1 1 0 0 0 0 1 1 0 0 0 0 0 1 0 1 1 0 1 0 1 0 0 1 1 0 0 1 1 0 0 0 0 1 0 0 0// 1 1 0 1 1 0 0 0 1 0 0 0 0 0 0 0 0 1 0 1 1 1 1 1...\n",
      "\n",
      "pop_size:[500], mutation_prob:[0.75], max_iters:[25]\n",
      "runner_name:[ga], experiment_name:[four_peaks], attempt:[6], iteration:[25], done:[True], time:[2.21], fitness:[26.0000]\n",
      "\t[1 1 1 0 1 1 1 1 1 1 1 0 0 1 0 0 1 1 0 0 0 0 0 0 1 0 1 0 1 0 1 0 1 0 0 0 0// 0 1 0 0 0 0 1 1 1 0 0 1 1 1 1 0 0 1 0 1 1 0 0 1 1 0 0 0 1 1 1 1 1 1 1 1 0// 1 0 0 0 0 0 1 0 0 0 0 0 1 0 0 1 1 1 0 1 1 1 0 0...\n",
      "\n",
      "***************\n",
      "*** Run END ***\n",
      "***************\n",
      "*****************\n",
      "*** Run START ***\n",
      "*****************\n",
      "pop_size:[1000], mutation_prob:[0.1], max_iters:[25]\n",
      "runner_name:[ga], experiment_name:[four_peaks], attempt:[0], iteration:[0], done:[False], time:[0.05], fitness:[0.0000]\n",
      "\t[0 0 1 0 1 1 1 1 1 0 1 0 0 0 0 1 1 1 1 0 0 1 0 1 0 0 0 0 0 0 1 1 1 1 0 1 0// 1 1 0 0 0 0 1 1 0 0 0 0 0 1 0 1 1 0 1 0 1 0 0 1 1 0 0 1 1 0 0 0 0 1 0 0 0// 1 1 0 1 1 0 0 0 1 0 0 0 0 0 0 0 0 1 0 1 1 1 1 1...\n",
      "\n",
      "pop_size:[1000], mutation_prob:[0.1], max_iters:[25]\n",
      "runner_name:[ga], experiment_name:[four_peaks], attempt:[1], iteration:[25], done:[True], time:[4.75], fitness:[30.0000]\n",
      "\t[1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 0 1 0 0 0 0 0// 0 0 1 1 0 1 1 1 0 0 0 0 0 1 1 1 0 1 1 0 1 0 1 1 1 0 0 1 0 1 0 0 1 0 0 0 1// 0 0 1 0 1 0 0 1 0 0 0 1 1 0 1 0 0 1 1 1 0 1 1 0...\n",
      "\n",
      "***************\n",
      "*** Run END ***\n",
      "***************\n",
      "*****************\n",
      "*** Run START ***\n",
      "*****************\n",
      "pop_size:[1000], mutation_prob:[0.25], max_iters:[25]\n",
      "runner_name:[ga], experiment_name:[four_peaks], attempt:[0], iteration:[0], done:[False], time:[0.05], fitness:[0.0000]\n",
      "\t[0 0 1 0 1 1 1 1 1 0 1 0 0 0 0 1 1 1 1 0 0 1 0 1 0 0 0 0 0 0 1 1 1 1 0 1 0// 1 1 0 0 0 0 1 1 0 0 0 0 0 1 0 1 1 0 1 0 1 0 0 1 1 0 0 1 1 0 0 0 0 1 0 0 0// 1 1 0 1 1 0 0 0 1 0 0 0 0 0 0 0 0 1 0 1 1 1 1 1...\n",
      "\n",
      "pop_size:[1000], mutation_prob:[0.25], max_iters:[25]\n",
      "runner_name:[ga], experiment_name:[four_peaks], attempt:[2], iteration:[25], done:[True], time:[4.76], fitness:[34.0000]\n",
      "\t[1 0 1 1 1 1 0 0 1 0 1 0 1 1 0 1 0 0 1 0 1 1 1 1 1 0 0 0 1 0 1 1 0 1 0 1 1// 1 0 0 1 0 1 0 0 1 0 1 1 0 0 1 0 1 0 0 0 1 1 1 1 1 1 1 1 1 0 1 1 1 1 0 0 0// 0 1 1 0 0 1 0 1 0 0 1 1 0 1 1 0 0 1 1 0 0 0 1 1...\n",
      "\n",
      "***************\n",
      "*** Run END ***\n",
      "***************\n",
      "*****************\n",
      "*** Run START ***\n",
      "*****************\n",
      "pop_size:[1000], mutation_prob:[0.5], max_iters:[25]\n",
      "runner_name:[ga], experiment_name:[four_peaks], attempt:[0], iteration:[0], done:[False], time:[0.05], fitness:[0.0000]\n",
      "\t[0 0 1 0 1 1 1 1 1 0 1 0 0 0 0 1 1 1 1 0 0 1 0 1 0 0 0 0 0 0 1 1 1 1 0 1 0// 1 1 0 0 0 0 1 1 0 0 0 0 0 1 0 1 1 0 1 0 1 0 0 1 1 0 0 1 1 0 0 0 0 1 0 0 0// 1 1 0 1 1 0 0 0 1 0 0 0 0 0 0 0 0 1 0 1 1 1 1 1...\n",
      "\n",
      "pop_size:[1000], mutation_prob:[0.5], max_iters:[25]\n",
      "runner_name:[ga], experiment_name:[four_peaks], attempt:[5], iteration:[25], done:[True], time:[4.55], fitness:[33.0000]\n",
      "\t[1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 0 0 0 1// 0 0 0 1 1 0 0 1 1 1 1 0 1 1 0 0 0 0 0 1 1 1 1 1 0 0 1 0 0 0 0 1 0 1 1 0 1// 1 1 1 0 0 1 0 0 0 1 0 1 1 1 1 1 0 0 1 1 1 0 1 0...\n",
      "\n",
      "***************\n",
      "*** Run END ***\n",
      "***************\n",
      "*****************\n",
      "*** Run START ***\n",
      "*****************\n",
      "pop_size:[1000], mutation_prob:[0.75], max_iters:[25]\n",
      "runner_name:[ga], experiment_name:[four_peaks], attempt:[0], iteration:[0], done:[False], time:[0.05], fitness:[0.0000]\n",
      "\t[0 0 1 0 1 1 1 1 1 0 1 0 0 0 0 1 1 1 1 0 0 1 0 1 0 0 0 0 0 0 1 1 1 1 0 1 0// 1 1 0 0 0 0 1 1 0 0 0 0 0 1 0 1 1 0 1 0 1 0 0 1 1 0 0 1 1 0 0 0 0 1 0 0 0// 1 1 0 1 1 0 0 0 1 0 0 0 0 0 0 0 0 1 0 1 1 1 1 1...\n",
      "\n",
      "pop_size:[1000], mutation_prob:[0.75], max_iters:[25]\n",
      "runner_name:[ga], experiment_name:[four_peaks], attempt:[1], iteration:[25], done:[True], time:[4.44], fitness:[32.0000]\n",
      "\t[1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 0 0 0 0 1// 1 1 0 0 0 1 1 0 0 0 0 0 0 1 0 0 1 1 0 0 1 0 1 1 0 1 1 0 0 0 0 1 1 0 0 1 0// 0 1 0 1 1 1 1 1 0 1 1 0 1 1 1 1 1 1 1 1 1 0 1 0...\n",
      "\n",
      "***************\n",
      "*** Run END ***\n",
      "***************\n",
      "*****************\n",
      "*** Run START ***\n",
      "*****************\n",
      "pop_size:[2000], mutation_prob:[0.1], max_iters:[25]\n",
      "runner_name:[ga], experiment_name:[four_peaks], attempt:[0], iteration:[0], done:[False], time:[0.09], fitness:[0.0000]\n",
      "\t[0 0 1 0 1 1 1 1 1 0 1 0 0 0 0 1 1 1 1 0 0 1 0 1 0 0 0 0 0 0 1 1 1 1 0 1 0// 1 1 0 0 0 0 1 1 0 0 0 0 0 1 0 1 1 0 1 0 1 0 0 1 1 0 0 1 1 0 0 0 0 1 0 0 0// 1 1 0 1 1 0 0 0 1 0 0 0 0 0 0 0 0 1 0 1 1 1 1 1...\n",
      "\n",
      "pop_size:[2000], mutation_prob:[0.1], max_iters:[25]\n",
      "runner_name:[ga], experiment_name:[four_peaks], attempt:[3], iteration:[25], done:[True], time:[9.90], fitness:[36.0000]\n",
      "\t[1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 0// 1 1 1 0 0 1 0 1 0 1 0 0 1 1 0 1 0 1 1 0 0 0 0 0 0 0 1 0 1 0 0 0 1 1 1 0 0// 1 1 1 0 1 1 1 0 1 1 1 1 0 0 1 0 0 0 1 1 1 1 0 1...\n",
      "\n",
      "***************\n",
      "*** Run END ***\n",
      "***************\n",
      "*****************\n",
      "*** Run START ***\n",
      "*****************\n",
      "pop_size:[2000], mutation_prob:[0.25], max_iters:[25]\n",
      "runner_name:[ga], experiment_name:[four_peaks], attempt:[0], iteration:[0], done:[False], time:[0.09], fitness:[0.0000]\n",
      "\t[0 0 1 0 1 1 1 1 1 0 1 0 0 0 0 1 1 1 1 0 0 1 0 1 0 0 0 0 0 0 1 1 1 1 0 1 0// 1 1 0 0 0 0 1 1 0 0 0 0 0 1 0 1 1 0 1 0 1 0 0 1 1 0 0 1 1 0 0 0 0 1 0 0 0// 1 1 0 1 1 0 0 0 1 0 0 0 0 0 0 0 0 1 0 1 1 1 1 1...\n",
      "\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "pop_size:[2000], mutation_prob:[0.25], max_iters:[25]\n",
      "runner_name:[ga], experiment_name:[four_peaks], attempt:[5], iteration:[25], done:[True], time:[9.73], fitness:[38.0000]\n",
      "\t[0 1 0 1 1 1 0 1 0 0 1 1 1 1 1 0 0 0 1 0 0 1 0 1 0 1 0 1 0 0 1 1 0 0 0 1 1// 0 1 1 0 0 0 0 0 0 1 1 1 0 1 1 0 0 1 1 1 0 1 0 1 0 1 1 1 0 1 1 1 0 1 1 0 0// 0 0 0 1 0 1 1 0 1 1 0 1 0 1 0 1 1 0 0 0 1 0 0 0...\n",
      "\n",
      "***************\n",
      "*** Run END ***\n",
      "***************\n",
      "*****************\n",
      "*** Run START ***\n",
      "*****************\n",
      "pop_size:[2000], mutation_prob:[0.5], max_iters:[25]\n",
      "runner_name:[ga], experiment_name:[four_peaks], attempt:[0], iteration:[0], done:[False], time:[0.09], fitness:[0.0000]\n",
      "\t[0 0 1 0 1 1 1 1 1 0 1 0 0 0 0 1 1 1 1 0 0 1 0 1 0 0 0 0 0 0 1 1 1 1 0 1 0// 1 1 0 0 0 0 1 1 0 0 0 0 0 1 0 1 1 0 1 0 1 0 0 1 1 0 0 1 1 0 0 0 0 1 0 0 0// 1 1 0 1 1 0 0 0 1 0 0 0 0 0 0 0 0 1 0 1 1 1 1 1...\n",
      "\n",
      "pop_size:[2000], mutation_prob:[0.5], max_iters:[25]\n",
      "runner_name:[ga], experiment_name:[four_peaks], attempt:[1], iteration:[25], done:[True], time:[9.53], fitness:[37.0000]\n",
      "\t[1 1 1 0 0 0 0 0 1 0 0 1 1 0 1 0 0 1 1 1 0 1 0 1 0 1 1 1 0 0 1 1 0 0 1 1 0// 1 0 1 0 1 1 1 0 1 0 0 0 0 0 1 0 1 1 0 1 1 1 1 1 0 0 1 1 0 0 1 0 1 0 0 0 1// 0 0 0 1 1 1 0 1 0 1 0 1 0 0 1 1 0 0 0 1 1 1 1 1...\n",
      "\n",
      "***************\n",
      "*** Run END ***\n",
      "***************\n",
      "*****************\n",
      "*** Run START ***\n",
      "*****************\n",
      "pop_size:[2000], mutation_prob:[0.75], max_iters:[25]\n",
      "runner_name:[ga], experiment_name:[four_peaks], attempt:[0], iteration:[0], done:[False], time:[0.10], fitness:[0.0000]\n",
      "\t[0 0 1 0 1 1 1 1 1 0 1 0 0 0 0 1 1 1 1 0 0 1 0 1 0 0 0 0 0 0 1 1 1 1 0 1 0// 1 1 0 0 0 0 1 1 0 0 0 0 0 1 0 1 1 0 1 0 1 0 0 1 1 0 0 1 1 0 0 0 0 1 0 0 0// 1 1 0 1 1 0 0 0 1 0 0 0 0 0 0 0 0 1 0 1 1 1 1 1...\n",
      "\n",
      "pop_size:[2000], mutation_prob:[0.75], max_iters:[25]\n",
      "runner_name:[ga], experiment_name:[four_peaks], attempt:[3], iteration:[25], done:[True], time:[9.31], fitness:[36.0000]\n",
      "\t[1 0 1 1 1 1 1 0 1 0 0 1 0 0 0 1 1 1 1 0 0 1 0 0 0 1 1 0 0 1 0 0 0 0 1 0 1// 0 0 1 0 1 0 1 1 0 0 1 1 1 0 1 0 1 0 1 0 0 1 0 0 1 0 0 1 1 0 0 1 1 1 1 0 1// 0 0 1 0 0 1 0 0 1 0 0 0 0 1 0 1 0 1 0 1 1 0 0 0...\n",
      "\n",
      "***************\n",
      "*** Run END ***\n",
      "***************\n",
      "*****************\n",
      "*** Run START ***\n",
      "*****************\n",
      "pop_size:[3000], mutation_prob:[0.1], max_iters:[25]\n",
      "runner_name:[ga], experiment_name:[four_peaks], attempt:[0], iteration:[0], done:[False], time:[0.15], fitness:[0.0000]\n",
      "\t[0 0 1 0 1 1 1 1 1 0 1 0 0 0 0 1 1 1 1 0 0 1 0 1 0 0 0 0 0 0 1 1 1 1 0 1 0// 1 1 0 0 0 0 1 1 0 0 0 0 0 1 0 1 1 0 1 0 1 0 0 1 1 0 0 1 1 0 0 0 0 1 0 0 0// 1 1 0 1 1 0 0 0 1 0 0 0 0 0 0 0 0 1 0 1 1 1 1 1...\n",
      "\n",
      "pop_size:[3000], mutation_prob:[0.1], max_iters:[25]\n",
      "runner_name:[ga], experiment_name:[four_peaks], attempt:[3], iteration:[25], done:[True], time:[15.24], fitness:[40.0000]\n",
      "\t[1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1// 1 1 1 0 0 0 0 1 0 1 0 0 1 1 1 1 1 0 1 1 0 1 0 1 1 1 1 0 0 0 0 1 1 0 1 1 0// 1 1 1 1 1 1 1 1 0 0 0 0 0 1 1 1 0 1 1 0 0 0 1 1...\n",
      "\n",
      "***************\n",
      "*** Run END ***\n",
      "***************\n",
      "*****************\n",
      "*** Run START ***\n",
      "*****************\n",
      "pop_size:[3000], mutation_prob:[0.25], max_iters:[25]\n",
      "runner_name:[ga], experiment_name:[four_peaks], attempt:[0], iteration:[0], done:[False], time:[0.14], fitness:[0.0000]\n",
      "\t[0 0 1 0 1 1 1 1 1 0 1 0 0 0 0 1 1 1 1 0 0 1 0 1 0 0 0 0 0 0 1 1 1 1 0 1 0// 1 1 0 0 0 0 1 1 0 0 0 0 0 1 0 1 1 0 1 0 1 0 0 1 1 0 0 1 1 0 0 0 0 1 0 0 0// 1 1 0 1 1 0 0 0 1 0 0 0 0 0 0 0 0 1 0 1 1 1 1 1...\n",
      "\n",
      "pop_size:[3000], mutation_prob:[0.25], max_iters:[25]\n",
      "runner_name:[ga], experiment_name:[four_peaks], attempt:[5], iteration:[25], done:[True], time:[14.97], fitness:[32.0000]\n",
      "\t[1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 0 0 1 0 0// 1 0 1 1 1 1 0 1 0 1 1 1 0 1 1 1 0 1 0 1 1 1 0 0 1 1 1 1 1 1 1 0 0 1 0 1 0// 0 1 0 0 1 0 0 1 0 1 0 1 0 1 0 0 1 0 0 0 1 1 1 1...\n",
      "\n",
      "***************\n",
      "*** Run END ***\n",
      "***************\n",
      "*****************\n",
      "*** Run START ***\n",
      "*****************\n",
      "pop_size:[3000], mutation_prob:[0.5], max_iters:[25]\n",
      "runner_name:[ga], experiment_name:[four_peaks], attempt:[0], iteration:[0], done:[False], time:[0.14], fitness:[0.0000]\n",
      "\t[0 0 1 0 1 1 1 1 1 0 1 0 0 0 0 1 1 1 1 0 0 1 0 1 0 0 0 0 0 0 1 1 1 1 0 1 0// 1 1 0 0 0 0 1 1 0 0 0 0 0 1 0 1 1 0 1 0 1 0 0 1 1 0 0 1 1 0 0 0 0 1 0 0 0// 1 1 0 1 1 0 0 0 1 0 0 0 0 0 0 0 0 1 0 1 1 1 1 1...\n",
      "\n",
      "pop_size:[3000], mutation_prob:[0.5], max_iters:[25]\n",
      "runner_name:[ga], experiment_name:[four_peaks], attempt:[1], iteration:[25], done:[True], time:[14.60], fitness:[38.0000]\n",
      "\t[1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1// 1 0 0 0 0 0 0 0 1 1 0 1 0 1 1 1 1 0 1 0 0 1 1 0 1 1 0 0 1 0 1 0 0 1 1 1 0// 1 0 1 0 1 0 1 1 0 1 1 0 1 1 0 0 1 1 0 0 1 1 1 0...\n",
      "\n",
      "***************\n",
      "*** Run END ***\n",
      "***************\n",
      "*****************\n",
      "*** Run START ***\n",
      "*****************\n",
      "pop_size:[3000], mutation_prob:[0.75], max_iters:[25]\n",
      "runner_name:[ga], experiment_name:[four_peaks], attempt:[0], iteration:[0], done:[False], time:[0.15], fitness:[0.0000]\n",
      "\t[0 0 1 0 1 1 1 1 1 0 1 0 0 0 0 1 1 1 1 0 0 1 0 1 0 0 0 0 0 0 1 1 1 1 0 1 0// 1 1 0 0 0 0 1 1 0 0 0 0 0 1 0 1 1 0 1 0 1 0 0 1 1 0 0 1 1 0 0 0 0 1 0 0 0// 1 1 0 1 1 0 0 0 1 0 0 0 0 0 0 0 0 1 0 1 1 1 1 1...\n",
      "\n",
      "pop_size:[3000], mutation_prob:[0.75], max_iters:[25]\n",
      "runner_name:[ga], experiment_name:[four_peaks], attempt:[1], iteration:[25], done:[True], time:[14.37], fitness:[37.0000]\n",
      "\t[1 1 1 0 1 1 1 0 1 0 1 1 1 1 1 1 1 1 0 0 1 0 0 0 0 0 0 0 1 0 1 0 0 1 1 0 1// 0 0 0 1 0 1 1 1 0 0 0 0 0 1 0 1 0 1 1 1 1 0 0 0 1 1 1 1 0 0 0 1 0 1 1 0 1// 1 0 0 0 0 0 0 0 1 0 1 1 1 1 0 0 1 1 1 0 0 0 1 1...\n",
      "\n",
      "***************\n",
      "*** Run END ***\n",
      "***************\n",
      "Run time: 125.62498721998418\n",
      "Saving: [four_peaks/four_peaks/ga__four_peaks__run_stats_df.csv]\n",
      "Saving: [four_peaks/four_peaks/ga__four_peaks__curves_df.csv]\n"
     ]
    }
   ],
   "source": [
    "ga = mlrh.GARunner(problem=prob,\n",
    "                   experiment_name=experiment_name,\n",
    "                   output_directory=output_directory,\n",
    "                   seed=random_state,\n",
    "                   max_attempts=20,\n",
    "                   iteration_list=[25],\n",
    "                   population_sizes=[500, 1000, 2000, 3000],\n",
    "                   mutation_rates=[0.1, 0.25, 0.5, 0.75])\n",
    "ga_stats, ga_curve = ga.run()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {
    "scrolled": false
   },
   "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>Time</th>\n",
       "      <th>Fitness</th>\n",
       "      <th>Population Size</th>\n",
       "      <th>Mutation Rate</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>25</th>\n",
       "      <td>15.241749</td>\n",
       "      <td>40.0</td>\n",
       "      <td>3000</td>\n",
       "      <td>0.10</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>19</th>\n",
       "      <td>9.729882</td>\n",
       "      <td>38.0</td>\n",
       "      <td>2000</td>\n",
       "      <td>0.25</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>29</th>\n",
       "      <td>14.600447</td>\n",
       "      <td>38.0</td>\n",
       "      <td>3000</td>\n",
       "      <td>0.50</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>31</th>\n",
       "      <td>14.374018</td>\n",
       "      <td>37.0</td>\n",
       "      <td>3000</td>\n",
       "      <td>0.75</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>21</th>\n",
       "      <td>9.530168</td>\n",
       "      <td>37.0</td>\n",
       "      <td>2000</td>\n",
       "      <td>0.50</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>23</th>\n",
       "      <td>9.306833</td>\n",
       "      <td>36.0</td>\n",
       "      <td>2000</td>\n",
       "      <td>0.75</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>17</th>\n",
       "      <td>9.897134</td>\n",
       "      <td>36.0</td>\n",
       "      <td>2000</td>\n",
       "      <td>0.10</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11</th>\n",
       "      <td>4.758080</td>\n",
       "      <td>34.0</td>\n",
       "      <td>1000</td>\n",
       "      <td>0.25</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>2.295125</td>\n",
       "      <td>33.0</td>\n",
       "      <td>500</td>\n",
       "      <td>0.50</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>13</th>\n",
       "      <td>4.552805</td>\n",
       "      <td>33.0</td>\n",
       "      <td>1000</td>\n",
       "      <td>0.50</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "         Time  Fitness  Population Size  Mutation Rate\n",
       "25  15.241749     40.0             3000           0.10\n",
       "19   9.729882     38.0             2000           0.25\n",
       "29  14.600447     38.0             3000           0.50\n",
       "31  14.374018     37.0             3000           0.75\n",
       "21   9.530168     37.0             2000           0.50\n",
       "23   9.306833     36.0             2000           0.75\n",
       "17   9.897134     36.0             2000           0.10\n",
       "11   4.758080     34.0             1000           0.25\n",
       "5    2.295125     33.0              500           0.50\n",
       "13   4.552805     33.0             1000           0.50"
      ]
     },
     "execution_count": 17,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "show=10\n",
    "columns = ['Time', 'Fitness', 'Population Size', 'Mutation Rate']\n",
    "pd.set_option('display.max_rows',show)\n",
    "df=pd.read_csv(\"./four_peaks/four_peaks/ga__four_peaks__run_stats_df.csv\")\n",
    "df[columns].sort_values(by=['Fitness'], ascending=False).head(show)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Genetic Alg - Total Function Evaluations: 78038\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 504x504 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "max_attempts = 100\n",
    "max_iters = 25\n",
    "mutation_prob=0.1\n",
    "pop_size = 3000\n",
    "global eval_count\n",
    "eval_count = 0\n",
    "best_state, best_fitness, gen_curve = mlrh.genetic_alg(prob,\n",
    "                                                   max_attempts=max_attempts,\n",
    "                                                   max_iters=max_iters,\n",
    "                                                   random_state=random_state,\n",
    "                                                   pop_size=pop_size,\n",
    "                                                   mutation_prob=mutation_prob,\n",
    "                                                   curve=True)\n",
    "print(\"Genetic Alg - Total Function Evaluations:\", eval_count)\n",
    "plot_fitness_iteration(gen_curve,\n",
    "                       \"Four Peaks - Genetic Alg: mutation_prob: {}, pop_size: {}\".format(mutation_prob, pop_size))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Simulated Annealing"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Running sa\n",
      "*****************\n",
      "*** Run START ***\n",
      "*****************\n",
      "schedule:[0.05], max_iters:[2500]\n",
      "runner_name:[sa], experiment_name:[four_peaks], attempt:[0], iteration:[0], done:[False], time:[0.00], fitness:[0.0000]\n",
      "\t[0 0 1 0 1 1 1 1 1 0 1 0 0 0 0 1 1 1 1 0 0 1 0 1 0 0 0 0 0 0 1 1 1 1 0 1 0// 1 1 0 0 0 0 1 1 0 0 0 0 0 1 0 1 1 0 1 0 1 0 0 1 1 0 0 1 1 0 0 0 0 1 0 0 0// 1 1 0 1 1 0 0 0 1 0 0 0 0 0 0 0 0 1 0 1 1 1 1 1...\n",
      "\n",
      "schedule:[0.05], max_iters:[2500]\n",
      "runner_name:[sa], experiment_name:[four_peaks], attempt:[1], iteration:[2500], done:[True], time:[0.38], fitness:[28.0000]\n",
      "\t[1 0 0 1 1 1 0 0 1 0 1 0 1 1 1 1 0 1 1 1 0 0 1 0 0 1 1 0 0 0 1 0 1 0 0 0 0// 0 0 0 0 0 0 1 1 1 1 1 1 0 1 1 1 0 1 1 1 0 1 1 0 0 0 1 0 0 1 0 0 0 0 0 1 1// 0 0 1 1 0 1 1 0 1 1 1 0 0 0 0 1 1 0 0 0 0 1 1 1...\n",
      "\n",
      "***************\n",
      "*** Run END ***\n",
      "***************\n",
      "*****************\n",
      "*** Run START ***\n",
      "*****************\n",
      "schedule:[0.05], max_iters:[2500]\n",
      "runner_name:[sa], experiment_name:[four_peaks], attempt:[0], iteration:[0], done:[False], time:[0.00], fitness:[0.0000]\n",
      "\t[0 0 1 0 1 1 1 1 1 0 1 0 0 0 0 1 1 1 1 0 0 1 0 1 0 0 0 0 0 0 1 1 1 1 0 1 0// 1 1 0 0 0 0 1 1 0 0 0 0 0 1 0 1 1 0 1 0 1 0 0 1 1 0 0 1 1 0 0 0 0 1 0 0 0// 1 1 0 1 1 0 0 0 1 0 0 0 0 0 0 0 0 1 0 1 1 1 1 1...\n",
      "\n",
      "schedule:[0.05], max_iters:[2500]\n",
      "runner_name:[sa], experiment_name:[four_peaks], attempt:[1], iteration:[2500], done:[True], time:[0.39], fitness:[28.0000]\n",
      "\t[1 0 0 1 1 1 0 0 1 0 1 0 1 1 1 1 0 1 1 1 0 0 1 0 0 1 1 0 0 0 1 0 1 0 0 0 0// 0 0 0 0 0 0 1 1 1 1 1 1 0 1 1 1 0 1 1 1 0 1 1 0 0 0 1 0 0 1 0 0 0 0 0 1 1// 0 0 1 1 0 1 1 0 1 1 1 0 0 0 0 1 1 0 0 0 0 1 1 1...\n",
      "\n",
      "***************\n",
      "*** Run END ***\n",
      "***************\n",
      "*****************\n",
      "*** Run START ***\n",
      "*****************\n",
      "schedule:[0.05], max_iters:[2500]\n",
      "runner_name:[sa], experiment_name:[four_peaks], attempt:[0], iteration:[0], done:[False], time:[0.00], fitness:[0.0000]\n",
      "\t[0 0 1 0 1 1 1 1 1 0 1 0 0 0 0 1 1 1 1 0 0 1 0 1 0 0 0 0 0 0 1 1 1 1 0 1 0// 1 1 0 0 0 0 1 1 0 0 0 0 0 1 0 1 1 0 1 0 1 0 0 1 1 0 0 1 1 0 0 0 0 1 0 0 0// 1 1 0 1 1 0 0 0 1 0 0 0 0 0 0 0 0 1 0 1 1 1 1 1...\n",
      "\n",
      "schedule:[0.05], max_iters:[2500]\n",
      "runner_name:[sa], experiment_name:[four_peaks], attempt:[1], iteration:[2500], done:[True], time:[0.37], fitness:[28.0000]\n",
      "\t[1 0 0 1 1 1 0 0 1 0 1 0 1 1 1 1 0 1 1 1 0 0 1 0 0 1 1 0 0 0 1 0 1 0 0 0 0// 0 0 0 0 0 0 1 1 1 1 1 1 0 1 1 1 0 1 1 1 0 1 1 0 0 0 1 0 0 1 0 0 0 0 0 1 1// 0 0 1 1 0 1 1 0 1 1 1 0 0 0 0 1 1 0 0 0 0 1 1 1...\n",
      "\n",
      "***************\n",
      "*** Run END ***\n",
      "***************\n",
      "*****************\n",
      "*** Run START ***\n",
      "*****************\n",
      "schedule:[0.1], max_iters:[2500]\n",
      "runner_name:[sa], experiment_name:[four_peaks], attempt:[0], iteration:[0], done:[False], time:[0.00], fitness:[0.0000]\n",
      "\t[0 0 1 0 1 1 1 1 1 0 1 0 0 0 0 1 1 1 1 0 0 1 0 1 0 0 0 0 0 0 1 1 1 1 0 1 0// 1 1 0 0 0 0 1 1 0 0 0 0 0 1 0 1 1 0 1 0 1 0 0 1 1 0 0 1 1 0 0 0 0 1 0 0 0// 1 1 0 1 1 0 0 0 1 0 0 0 0 0 0 0 0 1 0 1 1 1 1 1...\n",
      "\n",
      "schedule:[0.1], max_iters:[2500]\n",
      "runner_name:[sa], experiment_name:[four_peaks], attempt:[1], iteration:[2500], done:[True], time:[0.37], fitness:[28.0000]\n",
      "\t[1 0 0 1 1 1 0 0 1 0 1 0 1 1 1 1 0 1 1 1 0 0 1 0 0 1 1 0 0 0 1 0 1 0 0 0 0// 0 0 0 0 0 0 1 1 1 1 1 1 0 1 1 1 0 1 1 1 0 1 1 0 0 0 1 0 0 1 0 0 0 0 0 1 1// 0 0 1 1 0 1 1 0 1 1 1 0 0 0 0 1 1 0 0 0 0 1 1 1...\n",
      "\n",
      "***************\n",
      "*** Run END ***\n",
      "***************\n",
      "*****************\n",
      "*** Run START ***\n",
      "*****************\n",
      "schedule:[0.1], max_iters:[2500]\n",
      "runner_name:[sa], experiment_name:[four_peaks], attempt:[0], iteration:[0], done:[False], time:[0.00], fitness:[0.0000]\n",
      "\t[0 0 1 0 1 1 1 1 1 0 1 0 0 0 0 1 1 1 1 0 0 1 0 1 0 0 0 0 0 0 1 1 1 1 0 1 0// 1 1 0 0 0 0 1 1 0 0 0 0 0 1 0 1 1 0 1 0 1 0 0 1 1 0 0 1 1 0 0 0 0 1 0 0 0// 1 1 0 1 1 0 0 0 1 0 0 0 0 0 0 0 0 1 0 1 1 1 1 1...\n",
      "\n",
      "schedule:[0.1], max_iters:[2500]\n",
      "runner_name:[sa], experiment_name:[four_peaks], attempt:[1], iteration:[2500], done:[True], time:[0.38], fitness:[28.0000]\n",
      "\t[1 0 0 1 1 1 0 0 1 0 1 0 1 1 1 1 0 1 1 1 0 0 1 0 0 1 1 0 0 0 1 0 1 0 0 0 0// 0 0 0 0 0 0 1 1 1 1 1 1 0 1 1 1 0 1 1 1 0 1 1 0 0 0 1 0 0 1 0 0 0 0 0 1 1// 0 0 1 1 0 1 1 0 1 1 1 0 0 0 0 1 1 0 0 0 0 1 1 1...\n",
      "\n",
      "***************\n",
      "*** Run END ***\n",
      "***************\n",
      "*****************\n",
      "*** Run START ***\n",
      "*****************\n",
      "schedule:[0.1], max_iters:[2500]\n",
      "runner_name:[sa], experiment_name:[four_peaks], attempt:[0], iteration:[0], done:[False], time:[0.00], fitness:[0.0000]\n",
      "\t[0 0 1 0 1 1 1 1 1 0 1 0 0 0 0 1 1 1 1 0 0 1 0 1 0 0 0 0 0 0 1 1 1 1 0 1 0// 1 1 0 0 0 0 1 1 0 0 0 0 0 1 0 1 1 0 1 0 1 0 0 1 1 0 0 1 1 0 0 0 0 1 0 0 0// 1 1 0 1 1 0 0 0 1 0 0 0 0 0 0 0 0 1 0 1 1 1 1 1...\n",
      "\n",
      "schedule:[0.1], max_iters:[2500]\n",
      "runner_name:[sa], experiment_name:[four_peaks], attempt:[1], iteration:[2500], done:[True], time:[0.36], fitness:[28.0000]\n",
      "\t[1 0 0 1 1 1 0 0 1 0 1 0 1 1 1 1 0 1 1 1 0 0 1 0 0 1 1 0 0 0 1 0 1 0 0 0 0// 0 0 0 0 0 0 1 1 1 1 1 1 0 1 1 1 0 1 1 1 0 1 1 0 0 0 1 0 0 1 0 0 0 0 0 1 1// 0 0 1 1 0 1 1 0 1 1 1 0 0 0 0 1 1 0 0 0 0 1 1 1...\n",
      "\n",
      "***************\n",
      "*** Run END ***\n",
      "***************\n",
      "*****************\n",
      "*** Run START ***\n",
      "*****************\n",
      "schedule:[0.5], max_iters:[2500]\n",
      "runner_name:[sa], experiment_name:[four_peaks], attempt:[0], iteration:[0], done:[False], time:[0.00], fitness:[0.0000]\n",
      "\t[0 0 1 0 1 1 1 1 1 0 1 0 0 0 0 1 1 1 1 0 0 1 0 1 0 0 0 0 0 0 1 1 1 1 0 1 0// 1 1 0 0 0 0 1 1 0 0 0 0 0 1 0 1 1 0 1 0 1 0 0 1 1 0 0 1 1 0 0 0 0 1 0 0 0// 1 1 0 1 1 0 0 0 1 0 0 0 0 0 0 0 0 1 0 1 1 1 1 1...\n",
      "\n",
      "schedule:[0.5], max_iters:[2500]\n",
      "runner_name:[sa], experiment_name:[four_peaks], attempt:[1], iteration:[2500], done:[True], time:[0.36], fitness:[28.0000]\n",
      "\t[1 0 0 1 1 1 0 0 1 0 1 0 1 1 1 1 0 1 1 1 0 0 1 0 0 1 1 0 0 0 1 0 1 0 0 0 0// 0 0 0 0 0 0 1 1 1 1 1 1 0 1 1 1 0 1 1 1 0 1 1 0 0 0 1 0 0 1 0 0 0 0 0 1 1// 0 0 1 1 0 1 1 0 1 1 1 0 0 0 0 1 1 0 0 0 0 1 1 1...\n",
      "\n",
      "***************\n",
      "*** Run END ***\n",
      "***************\n",
      "*****************\n",
      "*** Run START ***\n",
      "*****************\n",
      "schedule:[0.5], max_iters:[2500]\n",
      "runner_name:[sa], experiment_name:[four_peaks], attempt:[0], iteration:[0], done:[False], time:[0.00], fitness:[0.0000]\n",
      "\t[0 0 1 0 1 1 1 1 1 0 1 0 0 0 0 1 1 1 1 0 0 1 0 1 0 0 0 0 0 0 1 1 1 1 0 1 0// 1 1 0 0 0 0 1 1 0 0 0 0 0 1 0 1 1 0 1 0 1 0 0 1 1 0 0 1 1 0 0 0 0 1 0 0 0// 1 1 0 1 1 0 0 0 1 0 0 0 0 0 0 0 0 1 0 1 1 1 1 1...\n",
      "\n",
      "schedule:[0.5], max_iters:[2500]\n",
      "runner_name:[sa], experiment_name:[four_peaks], attempt:[1], iteration:[2500], done:[True], time:[0.39], fitness:[28.0000]\n",
      "\t[1 0 0 1 1 1 0 0 1 0 1 0 1 1 1 1 0 1 1 1 0 0 1 0 0 1 1 0 0 0 1 0 1 0 0 0 0// 0 0 0 0 0 0 1 1 1 1 1 1 0 1 1 1 0 1 1 1 0 1 1 0 0 0 1 0 0 1 0 0 0 0 0 1 1// 0 0 1 1 0 1 1 0 1 1 1 0 0 0 0 1 1 0 0 0 0 1 1 1...\n",
      "\n",
      "***************\n",
      "*** Run END ***\n",
      "***************\n",
      "*****************\n",
      "*** Run START ***\n",
      "*****************\n",
      "schedule:[0.5], max_iters:[2500]\n",
      "runner_name:[sa], experiment_name:[four_peaks], attempt:[0], iteration:[0], done:[False], time:[0.00], fitness:[0.0000]\n",
      "\t[0 0 1 0 1 1 1 1 1 0 1 0 0 0 0 1 1 1 1 0 0 1 0 1 0 0 0 0 0 0 1 1 1 1 0 1 0// 1 1 0 0 0 0 1 1 0 0 0 0 0 1 0 1 1 0 1 0 1 0 0 1 1 0 0 1 1 0 0 0 0 1 0 0 0// 1 1 0 1 1 0 0 0 1 0 0 0 0 0 0 0 0 1 0 1 1 1 1 1...\n",
      "\n",
      "schedule:[0.5], max_iters:[2500]\n",
      "runner_name:[sa], experiment_name:[four_peaks], attempt:[1], iteration:[2500], done:[True], time:[0.37], fitness:[28.0000]\n",
      "\t[1 0 0 1 1 1 0 0 0 0 1 0 1 1 1 1 0 1 1 1 0 0 1 0 0 1 1 0 0 0 1 0 1 0 0 0 0// 0 0 0 0 0 0 1 1 1 1 1 1 0 1 1 1 0 1 1 1 0 1 1 0 0 0 1 0 0 1 0 0 0 0 0 1 0// 0 0 1 1 0 1 1 0 1 1 1 0 0 0 0 1 1 0 0 0 0 1 1 1...\n",
      "\n",
      "***************\n",
      "*** Run END ***\n",
      "***************\n",
      "*****************\n",
      "*** Run START ***\n",
      "*****************\n",
      "schedule:[1], max_iters:[2500]\n",
      "runner_name:[sa], experiment_name:[four_peaks], attempt:[0], iteration:[0], done:[False], time:[0.00], fitness:[0.0000]\n",
      "\t[0 0 1 0 1 1 1 1 1 0 1 0 0 0 0 1 1 1 1 0 0 1 0 1 0 0 0 0 0 0 1 1 1 1 0 1 0// 1 1 0 0 0 0 1 1 0 0 0 0 0 1 0 1 1 0 1 0 1 0 0 1 1 0 0 1 1 0 0 0 0 1 0 0 0// 1 1 0 1 1 0 0 0 1 0 0 0 0 0 0 0 0 1 0 1 1 1 1 1...\n",
      "\n",
      "schedule:[1], max_iters:[2500]\n",
      "runner_name:[sa], experiment_name:[four_peaks], attempt:[1], iteration:[2500], done:[True], time:[0.37], fitness:[28.0000]\n",
      "\t[1 0 0 1 1 1 0 0 0 0 1 0 1 1 1 1 0 1 1 1 0 0 1 0 0 1 1 0 0 0 1 0 1 0 0 0 0// 0 0 0 0 0 0 1 1 1 1 1 1 0 1 1 1 0 1 1 1 0 1 1 0 0 0 1 0 0 1 0 0 0 0 0 1 0// 0 0 1 1 0 1 1 0 1 1 1 0 0 0 0 1 1 0 0 0 0 1 1 1...\n",
      "\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "***************\n",
      "*** Run END ***\n",
      "***************\n",
      "*****************\n",
      "*** Run START ***\n",
      "*****************\n",
      "schedule:[1], max_iters:[2500]\n",
      "runner_name:[sa], experiment_name:[four_peaks], attempt:[0], iteration:[0], done:[False], time:[0.00], fitness:[0.0000]\n",
      "\t[0 0 1 0 1 1 1 1 1 0 1 0 0 0 0 1 1 1 1 0 0 1 0 1 0 0 0 0 0 0 1 1 1 1 0 1 0// 1 1 0 0 0 0 1 1 0 0 0 0 0 1 0 1 1 0 1 0 1 0 0 1 1 0 0 1 1 0 0 0 0 1 0 0 0// 1 1 0 1 1 0 0 0 1 0 0 0 0 0 0 0 0 1 0 1 1 1 1 1...\n",
      "\n",
      "schedule:[1], max_iters:[2500]\n",
      "runner_name:[sa], experiment_name:[four_peaks], attempt:[1], iteration:[2500], done:[True], time:[0.38], fitness:[28.0000]\n",
      "\t[1 0 0 1 1 1 0 0 0 0 1 0 1 1 1 1 0 1 1 1 0 0 1 0 0 1 1 0 0 0 1 0 1 0 0 0 0// 0 0 0 0 0 0 1 1 1 1 1 1 0 1 1 1 0 1 1 1 0 1 1 0 0 0 1 0 0 1 0 0 0 0 0 1 0// 0 0 1 1 0 1 1 0 1 1 1 0 0 0 0 1 1 0 0 0 0 1 1 1...\n",
      "\n",
      "***************\n",
      "*** Run END ***\n",
      "***************\n",
      "*****************\n",
      "*** Run START ***\n",
      "*****************\n",
      "schedule:[1], max_iters:[2500]\n",
      "runner_name:[sa], experiment_name:[four_peaks], attempt:[0], iteration:[0], done:[False], time:[0.00], fitness:[0.0000]\n",
      "\t[0 0 1 0 1 1 1 1 1 0 1 0 0 0 0 1 1 1 1 0 0 1 0 1 0 0 0 0 0 0 1 1 1 1 0 1 0// 1 1 0 0 0 0 1 1 0 0 0 0 0 1 0 1 1 0 1 0 1 0 0 1 1 0 0 1 1 0 0 0 0 1 0 0 0// 1 1 0 1 1 0 0 0 1 0 0 0 0 0 0 0 0 1 0 1 1 1 1 1...\n",
      "\n",
      "schedule:[1], max_iters:[2500]\n",
      "runner_name:[sa], experiment_name:[four_peaks], attempt:[1], iteration:[2500], done:[True], time:[0.36], fitness:[28.0000]\n",
      "\t[1 0 0 1 1 1 0 0 0 0 1 0 1 1 1 1 0 1 1 1 0 0 1 0 0 1 1 0 0 1 1 0 1 0 1 0 0// 0 0 0 0 0 0 1 1 1 1 1 1 0 1 1 1 0 1 1 1 0 1 1 0 0 0 1 0 0 1 0 0 0 0 0 1 0// 0 0 1 1 0 1 1 0 1 1 1 0 0 0 0 1 1 0 0 1 0 1 1 1...\n",
      "\n",
      "***************\n",
      "*** Run END ***\n",
      "***************\n",
      "*****************\n",
      "*** Run START ***\n",
      "*****************\n",
      "schedule:[10], max_iters:[2500]\n",
      "runner_name:[sa], experiment_name:[four_peaks], attempt:[0], iteration:[0], done:[False], time:[0.00], fitness:[0.0000]\n",
      "\t[0 0 1 0 1 1 1 1 1 0 1 0 0 0 0 1 1 1 1 0 0 1 0 1 0 0 0 0 0 0 1 1 1 1 0 1 0// 1 1 0 0 0 0 1 1 0 0 0 0 0 1 0 1 1 0 1 0 1 0 0 1 1 0 0 1 1 0 0 0 0 1 0 0 0// 1 1 0 1 1 0 0 0 1 0 0 0 0 0 0 0 0 1 0 1 1 1 1 1...\n",
      "\n",
      "schedule:[10], max_iters:[2500]\n",
      "runner_name:[sa], experiment_name:[four_peaks], attempt:[1], iteration:[2500], done:[True], time:[0.39], fitness:[28.0000]\n",
      "\t[1 0 0 1 1 1 0 0 0 0 1 0 1 1 1 1 0 1 1 1 0 0 1 0 0 1 1 0 0 0 1 0 1 0 0 0 0// 0 0 0 0 0 0 1 1 1 1 1 1 0 1 1 1 0 1 1 1 0 1 1 0 0 0 1 0 0 1 0 0 0 0 0 1 0// 0 0 1 1 0 1 1 0 1 1 1 0 0 0 0 1 1 0 0 0 0 1 1 1...\n",
      "\n",
      "***************\n",
      "*** Run END ***\n",
      "***************\n",
      "*****************\n",
      "*** Run START ***\n",
      "*****************\n",
      "schedule:[10], max_iters:[2500]\n",
      "runner_name:[sa], experiment_name:[four_peaks], attempt:[0], iteration:[0], done:[False], time:[0.00], fitness:[0.0000]\n",
      "\t[0 0 1 0 1 1 1 1 1 0 1 0 0 0 0 1 1 1 1 0 0 1 0 1 0 0 0 0 0 0 1 1 1 1 0 1 0// 1 1 0 0 0 0 1 1 0 0 0 0 0 1 0 1 1 0 1 0 1 0 0 1 1 0 0 1 1 0 0 0 0 1 0 0 0// 1 1 0 1 1 0 0 0 1 0 0 0 0 0 0 0 0 1 0 1 1 1 1 1...\n",
      "\n",
      "schedule:[10], max_iters:[2500]\n",
      "runner_name:[sa], experiment_name:[four_peaks], attempt:[1], iteration:[2500], done:[True], time:[0.38], fitness:[23.0000]\n",
      "\t[1 0 0 1 1 1 0 0 0 0 1 0 1 1 1 1 0 1 1 1 0 0 1 0 0 1 1 0 0 0 1 0 1 0 0 0 0// 0 0 0 0 0 0 1 1 1 1 1 1 0 1 1 1 0 1 1 1 0 1 1 0 0 0 1 0 0 1 0 0 1 0 0 1 0// 0 0 1 1 0 1 1 0 1 1 1 0 0 0 0 1 1 0 0 0 0 1 1 1...\n",
      "\n",
      "***************\n",
      "*** Run END ***\n",
      "***************\n",
      "*****************\n",
      "*** Run START ***\n",
      "*****************\n",
      "schedule:[10], max_iters:[2500]\n",
      "runner_name:[sa], experiment_name:[four_peaks], attempt:[0], iteration:[0], done:[False], time:[0.00], fitness:[0.0000]\n",
      "\t[0 0 1 0 1 1 1 1 1 0 1 0 0 0 0 1 1 1 1 0 0 1 0 1 0 0 0 0 0 0 1 1 1 1 0 1 0// 1 1 0 0 0 0 1 1 0 0 0 0 0 1 0 1 1 0 1 0 1 0 0 1 1 0 0 1 1 0 0 0 0 1 0 0 0// 1 1 0 1 1 0 0 0 1 0 0 0 0 0 0 0 0 1 0 1 1 1 1 1...\n",
      "\n",
      "schedule:[10], max_iters:[2500]\n",
      "runner_name:[sa], experiment_name:[four_peaks], attempt:[1], iteration:[2500], done:[True], time:[0.34], fitness:[1.0000]\n",
      "\t[0 0 0 0 1 0 0 0 1 0 1 0 0 0 1 0 1 1 1 1 1 0 1 0 0 1 1 1 0 0 1 1 1 1 0 0 0// 0 1 0 1 1 0 1 0 0 0 0 1 0 1 1 0 0 1 0 0 0 1 1 0 0 0 0 0 1 1 1 0 0 0 0 1 0// 0 0 1 1 0 1 1 1 1 1 0 0 1 0 0 0 1 0 0 0 0 0 0 1...\n",
      "\n",
      "***************\n",
      "*** Run END ***\n",
      "***************\n",
      "*****************\n",
      "*** Run START ***\n",
      "*****************\n",
      "schedule:[20], max_iters:[2500]\n",
      "runner_name:[sa], experiment_name:[four_peaks], attempt:[0], iteration:[0], done:[False], time:[0.00], fitness:[0.0000]\n",
      "\t[0 0 1 0 1 1 1 1 1 0 1 0 0 0 0 1 1 1 1 0 0 1 0 1 0 0 0 0 0 0 1 1 1 1 0 1 0// 1 1 0 0 0 0 1 1 0 0 0 0 0 1 0 1 1 0 1 0 1 0 0 1 1 0 0 1 1 0 0 0 0 1 0 0 0// 1 1 0 1 1 0 0 0 1 0 0 0 0 0 0 0 0 1 0 1 1 1 1 1...\n",
      "\n",
      "schedule:[20], max_iters:[2500]\n",
      "runner_name:[sa], experiment_name:[four_peaks], attempt:[1], iteration:[2500], done:[True], time:[0.38], fitness:[23.0000]\n",
      "\t[1 0 0 1 1 1 0 0 0 0 1 0 1 1 1 1 0 1 1 1 0 0 1 0 0 1 1 0 0 0 1 0 1 0 0 0 0// 0 0 0 0 0 0 1 1 1 1 1 1 0 1 1 1 0 1 1 1 0 1 1 0 0 0 1 0 0 1 0 0 1 0 0 1 0// 0 0 1 1 0 1 1 0 1 1 1 0 0 0 0 1 1 0 0 0 0 1 1 1...\n",
      "\n",
      "***************\n",
      "*** Run END ***\n",
      "***************\n",
      "*****************\n",
      "*** Run START ***\n",
      "*****************\n",
      "schedule:[20], max_iters:[2500]\n",
      "runner_name:[sa], experiment_name:[four_peaks], attempt:[0], iteration:[0], done:[False], time:[0.00], fitness:[0.0000]\n",
      "\t[0 0 1 0 1 1 1 1 1 0 1 0 0 0 0 1 1 1 1 0 0 1 0 1 0 0 0 0 0 0 1 1 1 1 0 1 0// 1 1 0 0 0 0 1 1 0 0 0 0 0 1 0 1 1 0 1 0 1 0 0 1 1 0 0 1 1 0 0 0 0 1 0 0 0// 1 1 0 1 1 0 0 0 1 0 0 0 0 0 0 0 0 1 0 1 1 1 1 1...\n",
      "\n",
      "schedule:[20], max_iters:[2500]\n",
      "runner_name:[sa], experiment_name:[four_peaks], attempt:[1], iteration:[2500], done:[True], time:[0.36], fitness:[28.0000]\n",
      "\t[1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 0 0 1 0 1 0 0 1 0// 0 1 1 0 0 0 1 0 0 0 1 1 0 1 1 0 0 1 1 0 0 1 1 0 0 1 1 0 1 1 0 0 1 0 0 0 0// 1 0 1 1 1 1 1 1 1 1 1 0 0 0 1 0 1 0 0 1 0 0 1 1...\n",
      "\n",
      "***************\n",
      "*** Run END ***\n",
      "***************\n",
      "*****************\n",
      "*** Run START ***\n",
      "*****************\n",
      "schedule:[20], max_iters:[2500]\n",
      "runner_name:[sa], experiment_name:[four_peaks], attempt:[0], iteration:[0], done:[False], time:[0.00], fitness:[0.0000]\n",
      "\t[0 0 1 0 1 1 1 1 1 0 1 0 0 0 0 1 1 1 1 0 0 1 0 1 0 0 0 0 0 0 1 1 1 1 0 1 0// 1 1 0 0 0 0 1 1 0 0 0 0 0 1 0 1 1 0 1 0 1 0 0 1 1 0 0 1 1 0 0 0 0 1 0 0 0// 1 1 0 1 1 0 0 0 1 0 0 0 0 0 0 0 0 1 0 1 1 1 1 1...\n",
      "\n",
      "schedule:[20], max_iters:[2500]\n",
      "runner_name:[sa], experiment_name:[four_peaks], attempt:[1], iteration:[2500], done:[True], time:[0.35], fitness:[1.0000]\n",
      "\t[0 0 0 0 1 0 0 0 1 0 1 0 0 0 1 0 1 1 1 1 1 0 1 0 0 1 1 1 0 0 1 1 1 1 0 0 0// 0 1 0 1 1 0 1 0 0 0 0 1 0 1 1 0 0 1 0 0 0 1 1 0 0 0 0 0 1 1 1 0 0 0 0 1 0// 0 0 1 1 0 1 1 1 1 1 0 0 1 0 0 0 1 0 0 0 0 0 0 1...\n",
      "\n",
      "***************\n",
      "*** Run END ***\n",
      "***************\n",
      "*****************\n",
      "*** Run START ***\n",
      "*****************\n",
      "schedule:[25], max_iters:[2500]\n",
      "runner_name:[sa], experiment_name:[four_peaks], attempt:[0], iteration:[0], done:[False], time:[0.00], fitness:[0.0000]\n",
      "\t[0 0 1 0 1 1 1 1 1 0 1 0 0 0 0 1 1 1 1 0 0 1 0 1 0 0 0 0 0 0 1 1 1 1 0 1 0// 1 1 0 0 0 0 1 1 0 0 0 0 0 1 0 1 1 0 1 0 1 0 0 1 1 0 0 1 1 0 0 0 0 1 0 0 0// 1 1 0 1 1 0 0 0 1 0 0 0 0 0 0 0 0 1 0 1 1 1 1 1...\n",
      "\n",
      "schedule:[25], max_iters:[2500]\n",
      "runner_name:[sa], experiment_name:[four_peaks], attempt:[1], iteration:[2500], done:[True], time:[0.35], fitness:[23.0000]\n",
      "\t[1 0 0 1 1 1 0 0 0 0 1 0 1 1 1 1 0 1 1 1 0 0 1 0 0 1 1 0 0 0 1 0 1 0 0 0 0// 0 0 0 0 0 0 1 1 1 1 1 1 0 1 1 1 0 1 1 1 0 1 1 0 0 0 1 0 0 1 0 0 1 0 0 1 0// 0 0 1 1 0 1 1 0 1 1 1 0 0 0 0 1 1 0 0 0 0 1 1 1...\n",
      "\n",
      "***************\n",
      "*** Run END ***\n",
      "***************\n",
      "*****************\n",
      "*** Run START ***\n",
      "*****************\n",
      "schedule:[25], max_iters:[2500]\n",
      "runner_name:[sa], experiment_name:[four_peaks], attempt:[0], iteration:[0], done:[False], time:[0.00], fitness:[0.0000]\n",
      "\t[0 0 1 0 1 1 1 1 1 0 1 0 0 0 0 1 1 1 1 0 0 1 0 1 0 0 0 0 0 0 1 1 1 1 0 1 0// 1 1 0 0 0 0 1 1 0 0 0 0 0 1 0 1 1 0 1 0 1 0 0 1 1 0 0 1 1 0 0 0 0 1 0 0 0// 1 1 0 1 1 0 0 0 1 0 0 0 0 0 0 0 0 1 0 1 1 1 1 1...\n",
      "\n",
      "schedule:[25], max_iters:[2500]\n",
      "runner_name:[sa], experiment_name:[four_peaks], attempt:[1], iteration:[2500], done:[True], time:[0.38], fitness:[21.0000]\n",
      "\t[0 0 1 0 0 1 1 1 1 0 1 0 0 1 1 1 0 1 1 0 1 0 1 0 0 1 1 1 0 0 1 0 0 0 0 0 0// 0 1 0 0 0 0 1 0 0 0 1 1 0 1 1 0 0 1 0 0 0 1 1 0 0 1 1 1 1 1 0 0 1 0 1 1 0// 0 0 1 1 0 1 1 1 1 1 1 0 0 0 0 1 1 0 0 0 0 0 1 1...\n",
      "\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "***************\n",
      "*** Run END ***\n",
      "***************\n",
      "*****************\n",
      "*** Run START ***\n",
      "*****************\n",
      "schedule:[25], max_iters:[2500]\n",
      "runner_name:[sa], experiment_name:[four_peaks], attempt:[0], iteration:[0], done:[False], time:[0.00], fitness:[0.0000]\n",
      "\t[0 0 1 0 1 1 1 1 1 0 1 0 0 0 0 1 1 1 1 0 0 1 0 1 0 0 0 0 0 0 1 1 1 1 0 1 0// 1 1 0 0 0 0 1 1 0 0 0 0 0 1 0 1 1 0 1 0 1 0 0 1 1 0 0 1 1 0 0 0 0 1 0 0 0// 1 1 0 1 1 0 0 0 1 0 0 0 0 0 0 0 0 1 0 1 1 1 1 1...\n",
      "\n",
      "schedule:[25], max_iters:[2500]\n",
      "runner_name:[sa], experiment_name:[four_peaks], attempt:[1], iteration:[2500], done:[True], time:[0.34], fitness:[4.0000]\n",
      "\t[0 0 0 0 1 0 0 0 1 0 0 1 1 0 1 1 1 1 0 1 1 0 1 1 0 1 1 1 0 0 1 0 1 0 0 1 0// 0 1 1 1 0 0 0 0 0 1 1 1 0 1 1 1 0 1 0 1 1 1 1 1 0 1 0 0 0 1 1 0 1 0 1 1 0// 0 1 1 1 0 1 1 0 1 1 0 0 1 0 1 1 1 0 1 1 0 1 1 1...\n",
      "\n",
      "***************\n",
      "*** Run END ***\n",
      "***************\n",
      "Run time: 14.025780880008824\n",
      "Saving: [four_peaks/four_peaks/sa__four_peaks__run_stats_df.csv]\n",
      "Saving: [four_peaks/four_peaks/sa__four_peaks__curves_df.csv]\n"
     ]
    }
   ],
   "source": [
    "sa = mlrh.SARunner(problem=prob,\n",
    "                   experiment_name=experiment_name,\n",
    "                   output_directory=output_directory,\n",
    "                   seed=random_state,\n",
    "                   max_attempts=200,\n",
    "                   iteration_list=[2500],\n",
    "                   temperature_list=[0.05, 0.1, 0.5, 1, 10, 20, 25],\n",
    "                   decay_list=[mlrh.GeomDecay, mlrh.ExpDecay, mlrh.ArithDecay])\n",
    "sa_stats, sa_curve = sa.run()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "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>Time</th>\n",
       "      <th>Fitness</th>\n",
       "      <th>Temperature</th>\n",
       "      <th>schedule_type</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>21</th>\n",
       "      <td>0.382714</td>\n",
       "      <td>28.0</td>\n",
       "      <td>1.00</td>\n",
       "      <td>exponential</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>19</th>\n",
       "      <td>0.366579</td>\n",
       "      <td>28.0</td>\n",
       "      <td>1.00</td>\n",
       "      <td>geometric</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>0.378602</td>\n",
       "      <td>28.0</td>\n",
       "      <td>0.05</td>\n",
       "      <td>geometric</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>15</th>\n",
       "      <td>0.391623</td>\n",
       "      <td>28.0</td>\n",
       "      <td>0.50</td>\n",
       "      <td>exponential</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>23</th>\n",
       "      <td>0.360380</td>\n",
       "      <td>28.0</td>\n",
       "      <td>1.00</td>\n",
       "      <td>arithmetic</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>13</th>\n",
       "      <td>0.362886</td>\n",
       "      <td>28.0</td>\n",
       "      <td>0.50</td>\n",
       "      <td>geometric</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25</th>\n",
       "      <td>0.387314</td>\n",
       "      <td>28.0</td>\n",
       "      <td>10.00</td>\n",
       "      <td>geometric</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11</th>\n",
       "      <td>0.359405</td>\n",
       "      <td>28.0</td>\n",
       "      <td>0.10</td>\n",
       "      <td>arithmetic</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>0.381840</td>\n",
       "      <td>28.0</td>\n",
       "      <td>0.10</td>\n",
       "      <td>exponential</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>0.371453</td>\n",
       "      <td>28.0</td>\n",
       "      <td>0.10</td>\n",
       "      <td>geometric</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "        Time  Fitness  Temperature schedule_type\n",
       "21  0.382714     28.0         1.00   exponential\n",
       "19  0.366579     28.0         1.00     geometric\n",
       "1   0.378602     28.0         0.05     geometric\n",
       "15  0.391623     28.0         0.50   exponential\n",
       "23  0.360380     28.0         1.00    arithmetic\n",
       "13  0.362886     28.0         0.50     geometric\n",
       "25  0.387314     28.0        10.00     geometric\n",
       "11  0.359405     28.0         0.10    arithmetic\n",
       "9   0.381840     28.0         0.10   exponential\n",
       "7   0.371453     28.0         0.10     geometric"
      ]
     },
     "execution_count": 26,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "show=10\n",
    "columns = ['Time', 'Fitness', 'Temperature', 'schedule_type']\n",
    "pd.set_option('display.max_rows',show)\n",
    "df=pd.read_csv(\"./four_peaks/four_peaks/sa__four_peaks__run_stats_df.csv\")\n",
    "df[columns].sort_values(by=['Fitness'], ascending=False).head(show)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {
    "scrolled": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Simulated Annealing - Total Function Evaluations: 4797\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 504x504 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "max_attempts = 200\n",
    "max_iters = 2500\n",
    "init_temp = 1\n",
    "schedule = mlrh.ExpDecay(init_temp)\n",
    "global eval_count\n",
    "eval_count = 0\n",
    "best_state, best_fitness, sa_curve = mlrh.simulated_annealing(prob,\n",
    "                                                             max_attempts=max_attempts,\n",
    "                                                             max_iters=max_iters,\n",
    "                                                             random_state=random_state,\n",
    "                                                             schedule=schedule,\n",
    "                                                             curve=True)\n",
    "print(\"Simulated Annealing - Total Function Evaluations:\", eval_count)\n",
    "plot_fitness_iteration(sa_curve,\n",
    "                       \"Four Peaks - Simulated Annealing: schedule: {}, init_temp: {}\".format(schedule.__class__.__name__, init_temp))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Randomized Hill Climbing"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Running rhc\n",
      "*****************\n",
      "*** Run START ***\n",
      "*****************\n",
      "restarts:[10], max_iters:[2500], current_restart:[0]\n",
      "runner_name:[rhc], experiment_name:[four_peaks], attempt:[0], iteration:[0], done:[False], time:[0.00], fitness:[0.0000]\n",
      "\t[0 0 1 0 1 1 1 1 1 0 1 0 0 0 0 1 1 1 1 0 0 1 0 1 0 0 0 0 0 0 1 1 1 1 0 1 0// 1 1 0 0 0 0 1 1 0 0 0 0 0 1 0 1 1 0 1 0 1 0 0 1 1 0 0 1 1 0 0 0 0 1 0 0 0// 1 1 0 1 1 0 0 0 1 0 0 0 0 0 0 0 0 1 0 1 1 1 1 1...\n",
      "\n",
      "restarts:[10], max_iters:[2500], current_restart:[0]\n",
      "runner_name:[rhc], experiment_name:[four_peaks], attempt:[201], iteration:[467], done:[True], time:[0.04], fitness:[7.0000]\n",
      "\t[1 0 1 0 1 1 1 1 1 0 1 0 0 0 0 1 1 1 1 0 0 1 0 1 0 0 0 0 0 0 1 1 1 1 0 1 0// 1 1 0 0 0 0 1 1 0 0 0 0 0 1 0 1 1 0 1 0 1 0 0 1 1 0 0 1 1 0 0 0 0 1 0 0 0// 1 1 0 1 1 0 0 0 1 0 0 0 0 0 0 0 0 1 0 1 1 1 1 1...\n",
      "\n",
      "restarts:[10], max_iters:[2500], current_restart:[1]\n",
      "runner_name:[rhc], experiment_name:[four_peaks], attempt:[0], iteration:[0], done:[False], time:[0.06], fitness:[0.0000]\n",
      "\t[0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 1 0 1 0 0 1 0 0 1 1 0 1 1 0 1 1 0 1// 1 0 0 1 1 0 0 1 1 1 1 0 1 0 0 1 0 0 0 0 1 0 1 1 0 0 0 1 1 0 0 0 0 0 0 0 1// 1 1 0 1 0 1 1 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 1 0...\n",
      "\n",
      "restarts:[10], max_iters:[2500], current_restart:[1]\n",
      "runner_name:[rhc], experiment_name:[four_peaks], attempt:[201], iteration:[243], done:[True], time:[0.10], fitness:[3.0000]\n",
      "\t[1 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 1 0 1 0 0 1 0 0 1 1 0 1 1 0 1 1 0 1// 1 0 0 1 1 0 0 1 1 1 1 0 1 0 0 1 0 0 0 0 1 0 1 1 0 0 0 1 1 0 0 0 0 0 0 0 1// 1 1 0 1 0 1 1 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 1 0...\n",
      "\n",
      "restarts:[10], max_iters:[2500], current_restart:[2]\n",
      "runner_name:[rhc], experiment_name:[four_peaks], attempt:[0], iteration:[0], done:[False], time:[0.12], fitness:[1.0000]\n",
      "\t[1 0 0 1 1 1 0 1 0 1 1 0 1 0 0 1 1 0 1 0 1 1 0 0 0 1 0 0 0 0 0 0 1 1 1 0 1// 1 1 0 1 0 1 1 1 1 0 1 0 1 0 1 0 0 1 1 0 1 0 0 1 0 1 0 0 0 1 0 0 0 1 0 1 0// 1 0 0 0 1 0 0 0 1 0 0 1 0 1 0 0 0 0 0 0 1 1 1 0...\n",
      "\n",
      "restarts:[10], max_iters:[2500], current_restart:[2]\n",
      "runner_name:[rhc], experiment_name:[four_peaks], attempt:[201], iteration:[479], done:[True], time:[0.20], fitness:[13.0000]\n",
      "\t[1 1 1 1 1 1 1 1 1 1 1 1 1 0 0 1 1 0 1 0 1 1 0 0 0 1 0 0 0 0 0 0 1 1 1 0 1// 1 1 0 1 0 1 1 1 1 0 1 0 1 0 1 0 0 1 1 0 1 0 0 1 0 1 0 0 0 1 0 0 0 1 0 1 0// 1 0 0 0 1 0 0 0 1 0 0 1 0 1 0 0 0 0 0 0 1 1 1 0...\n",
      "\n",
      "restarts:[10], max_iters:[2500], current_restart:[3]\n",
      "runner_name:[rhc], experiment_name:[four_peaks], attempt:[0], iteration:[0], done:[False], time:[0.22], fitness:[0.0000]\n",
      "\t[0 1 1 1 0 0 0 0 1 0 0 0 0 1 1 1 1 1 0 1 1 0 1 1 1 0 1 0 0 1 1 0 1 1 1 1 1// 0 1 1 1 1 1 0 1 1 0 0 1 1 1 1 0 0 1 1 0 0 1 1 0 0 1 0 1 0 1 1 1 0 1 0 0 0// 0 1 0 1 0 0 0 0 0 1 0 1 0 1 0 1 0 1 0 1 1 1 1 0...\n",
      "\n",
      "restarts:[10], max_iters:[2500], current_restart:[3]\n",
      "runner_name:[rhc], experiment_name:[four_peaks], attempt:[201], iteration:[421], done:[True], time:[0.30], fitness:[6.0000]\n",
      "\t[1 1 1 1 1 1 0 0 1 0 0 0 0 1 1 1 1 1 0 1 1 0 1 1 1 0 1 0 0 1 1 0 1 1 1 1 1// 0 1 1 1 1 1 0 1 1 0 0 1 1 1 1 0 0 1 1 0 0 1 1 0 0 1 0 1 0 1 1 1 0 1 0 0 0// 0 1 0 1 0 0 0 0 0 1 0 1 0 1 0 1 0 1 0 1 1 1 1 0...\n",
      "\n",
      "restarts:[10], max_iters:[2500], current_restart:[4]\n",
      "runner_name:[rhc], experiment_name:[four_peaks], attempt:[0], iteration:[0], done:[False], time:[0.33], fitness:[0.0000]\n",
      "\t[0 1 1 1 0 1 1 1 0 0 0 1 0 1 0 1 0 0 1 0 0 0 1 1 0 0 1 1 1 0 1 1 0 1 0 1 1// 1 0 1 0 1 1 1 0 0 1 1 0 0 1 1 0 1 1 1 0 0 0 1 1 0 1 1 1 1 0 0 1 1 0 0 1 1// 0 1 1 1 1 0 1 1 0 0 1 0 1 1 1 0 0 0 0 1 0 0 1 0...\n",
      "\n",
      "restarts:[10], max_iters:[2500], current_restart:[4]\n",
      "runner_name:[rhc], experiment_name:[four_peaks], attempt:[201], iteration:[304], done:[True], time:[0.40], fitness:[8.0000]\n",
      "\t[1 1 1 1 1 1 1 1 0 0 0 1 0 1 0 1 0 0 1 0 0 0 1 1 0 0 1 1 1 0 1 1 0 1 0 1 1// 1 0 1 0 1 1 1 0 0 1 1 0 0 1 1 0 1 1 1 0 0 0 1 1 0 1 1 1 1 0 0 1 1 0 0 1 1// 0 1 1 1 1 0 1 1 0 0 1 0 1 1 1 0 0 0 0 1 0 0 1 0...\n",
      "\n",
      "restarts:[10], max_iters:[2500], current_restart:[5]\n",
      "runner_name:[rhc], experiment_name:[four_peaks], attempt:[0], iteration:[0], done:[False], time:[0.44], fitness:[5.0000]\n",
      "\t[0 0 1 1 1 1 0 0 0 0 0 1 1 1 1 0 0 0 0 1 1 1 1 1 1 1 0 0 1 0 0 0 0 0 0 0 0// 1 0 1 1 0 0 1 0 0 1 1 1 0 1 1 1 1 1 0 0 1 0 1 1 1 1 0 0 1 1 0 1 0 0 0 0 0// 0 0 0 0 1 0 0 0 0 1 0 0 0 1 1 0 1 1 0 1 0 1 1 1...\n",
      "\n",
      "restarts:[10], max_iters:[2500], current_restart:[5]\n",
      "runner_name:[rhc], experiment_name:[four_peaks], attempt:[201], iteration:[780], done:[True], time:[0.64], fitness:[11.0000]\n",
      "\t[0 0 1 1 1 1 0 0 0 0 0 1 1 1 1 0 0 0 0 1 1 1 1 1 1 1 0 0 1 0 0 0 0 0 0 0 0// 1 0 1 1 0 0 1 0 0 1 1 1 0 1 1 1 1 1 0 0 1 0 1 1 1 1 0 0 1 1 0 1 0 0 0 0 0// 0 0 0 0 1 0 0 0 0 1 0 0 0 1 1 0 1 1 0 1 0 1 1 1...\n",
      "\n",
      "restarts:[10], max_iters:[2500], current_restart:[6]\n",
      "runner_name:[rhc], experiment_name:[four_peaks], attempt:[0], iteration:[0], done:[False], time:[0.69], fitness:[5.0000]\n",
      "\t[1 1 0 0 0 0 0 0 0 0 0 0 1 0 1 0 1 1 1 0 0 0 1 1 0 1 0 1 1 0 1 0 1 0 1 1 1// 0 1 0 1 0 0 1 1 1 1 1 0 1 1 0 0 1 0 1 1 0 0 1 1 1 0 1 0 1 1 0 1 1 0 0 0 0// 0 0 0 1 1 1 1 1 0 0 0 1 0 0 1 0 1 1 0 1 1 1 1 1...\n",
      "\n",
      "restarts:[10], max_iters:[2500], current_restart:[6]\n",
      "runner_name:[rhc], experiment_name:[four_peaks], attempt:[201], iteration:[399], done:[True], time:[0.82], fitness:[13.0000]\n",
      "\t[1 1 0 0 0 0 0 0 0 0 0 0 1 0 1 0 1 1 1 0 0 0 1 1 0 1 0 1 1 0 1 0 1 0 1 1 1// 0 1 0 1 0 0 1 1 1 1 1 0 1 1 0 0 1 0 1 1 0 0 1 1 1 0 1 0 1 1 0 1 1 0 0 0 0// 0 0 0 1 1 1 1 1 0 0 0 1 0 0 1 0 1 1 0 1 1 1 1 1...\n",
      "\n",
      "restarts:[10], max_iters:[2500], current_restart:[7]\n",
      "runner_name:[rhc], experiment_name:[four_peaks], attempt:[0], iteration:[0], done:[False], time:[0.87], fitness:[6.0000]\n",
      "\t[0 0 0 0 1 1 0 0 1 1 1 1 0 0 0 1 0 1 1 1 1 1 0 0 1 1 1 1 1 1 0 0 0 0 0 0 1// 1 1 0 0 0 0 0 1 0 0 0 1 0 1 0 1 0 1 1 0 1 0 0 0 1 1 0 1 0 0 1 0 0 0 1 1 0// 1 0 1 0 1 0 1 1 0 1 0 1 1 0 0 1 1 0 1 0 0 0 0 1...\n",
      "\n",
      "restarts:[10], max_iters:[2500], current_restart:[7]\n",
      "runner_name:[rhc], experiment_name:[four_peaks], attempt:[201], iteration:[221], done:[True], time:[0.95], fitness:[7.0000]\n",
      "\t[0 0 0 0 1 1 0 0 1 1 1 1 0 0 0 1 0 1 1 1 1 1 0 0 1 1 1 1 1 1 0 0 0 0 0 0 1// 1 1 0 0 0 0 0 1 0 0 0 1 0 1 0 1 0 1 1 0 1 0 0 0 1 1 0 1 0 0 1 0 0 0 1 1 0// 1 0 1 0 1 0 1 1 0 1 0 1 1 0 0 1 1 0 1 0 0 0 0 1...\n",
      "\n",
      "restarts:[10], max_iters:[2500], current_restart:[8]\n",
      "runner_name:[rhc], experiment_name:[four_peaks], attempt:[0], iteration:[0], done:[False], time:[1.00], fitness:[0.0000]\n",
      "\t[0 0 1 1 0 0 1 1 0 0 1 1 1 0 0 0 1 0 1 0 1 0 1 1 0 0 1 0 1 1 0 0 1 0 1 0 0// 0 0 0 1 1 1 0 0 0 0 0 0 0 1 1 0 1 0 1 0 0 0 0 1 1 0 0 1 0 0 0 1 1 1 0 0 1// 0 1 0 1 1 1 0 1 0 0 0 0 1 1 0 1 0 1 0 1 1 1 1 1...\n",
      "\n",
      "restarts:[10], max_iters:[2500], current_restart:[8]\n",
      "runner_name:[rhc], experiment_name:[four_peaks], attempt:[201], iteration:[208], done:[True], time:[1.08], fitness:[1.0000]\n",
      "\t[1 0 1 1 0 0 1 1 0 0 1 1 1 0 0 0 1 0 1 0 1 0 1 1 0 0 1 0 1 1 0 0 1 0 1 0 0// 0 0 0 1 1 1 0 0 0 0 0 0 0 1 1 0 1 0 1 0 0 0 0 1 1 0 0 1 0 0 0 1 1 1 0 0 1// 0 1 0 1 1 1 0 1 0 0 0 0 1 1 0 1 0 1 0 1 1 1 1 1...\n",
      "\n",
      "restarts:[10], max_iters:[2500], current_restart:[9]\n",
      "runner_name:[rhc], experiment_name:[four_peaks], attempt:[0], iteration:[0], done:[False], time:[1.14], fitness:[1.0000]\n",
      "\t[1 0 1 1 1 0 0 0 1 1 1 0 0 1 0 1 0 0 0 1 0 1 0 0 0 0 1 1 0 1 1 0 1 0 0 1 0// 1 1 0 1 1 1 1 0 0 1 0 0 0 0 0 0 1 1 1 1 1 1 1 0 1 1 0 1 1 1 1 1 1 1 0 1 0// 1 1 1 0 0 0 1 1 0 0 1 1 0 1 0 1 0 1 1 1 1 1 0 0...\n",
      "\n",
      "restarts:[10], max_iters:[2500], current_restart:[9]\n",
      "runner_name:[rhc], experiment_name:[four_peaks], attempt:[201], iteration:[966], done:[True], time:[1.56], fitness:[23.0000]\n",
      "\t[1 0 1 1 1 0 0 0 1 1 1 0 0 1 0 1 0 0 0 1 0 1 0 0 0 0 1 1 0 1 1 0 1 0 0 1 0// 1 1 0 1 1 1 1 0 0 1 0 0 0 0 0 0 1 1 1 1 1 1 1 0 1 1 0 1 1 1 1 1 1 1 0 1 0// 1 1 1 0 0 0 1 1 0 0 1 1 0 1 0 1 0 1 1 1 1 1 0 0...\n",
      "\n",
      "restarts:[10], max_iters:[2500], current_restart:[10]\n",
      "runner_name:[rhc], experiment_name:[four_peaks], attempt:[0], iteration:[0], done:[False], time:[1.63], fitness:[0.0000]\n",
      "\t[0 1 1 1 1 0 0 0 1 1 1 0 1 0 0 0 1 0 0 0 0 0 0 1 0 1 0 1 1 0 0 0 1 0 1 1 0// 0 0 1 0 0 1 0 0 0 0 0 0 0 0 0 0 1 1 0 1 1 0 1 0 0 1 1 0 1 1 1 1 0 1 0 1 0// 1 0 0 0 1 0 0 1 1 1 1 0 1 0 1 0 1 1 1 1 0 0 0 1...\n",
      "\n",
      "restarts:[10], max_iters:[2500], current_restart:[10]\n",
      "runner_name:[rhc], experiment_name:[four_peaks], attempt:[201], iteration:[683], done:[True], time:[1.96], fitness:[14.0000]\n",
      "\t[1 1 1 1 1 1 1 1 1 1 1 1 1 1 0 0 1 0 0 0 0 0 0 1 0 1 0 1 1 0 0 0 1 0 1 1 0// 0 0 1 0 0 1 0 0 0 0 0 0 0 0 0 0 1 1 0 1 1 0 1 0 0 1 1 0 1 1 1 1 0 1 0 1 0// 1 0 0 0 1 0 0 1 1 1 1 0 1 0 1 0 1 1 1 1 0 0 0 1...\n",
      "\n",
      "***************\n",
      "*** Run END ***\n",
      "***************\n",
      "Run time: 2.0384462830261327\n",
      "Saving: [four_peaks/four_peaks/rhc__four_peaks__run_stats_df.csv]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Saving: [four_peaks/four_peaks/rhc__four_peaks__curves_df.csv]\n"
     ]
    }
   ],
   "source": [
    "rhc = mlrh.RHCRunner(problem=prob,\n",
    "                    experiment_name=experiment_name,\n",
    "                    output_directory=output_directory,\n",
    "                    seed=random_state,\n",
    "                    max_attempts=200,\n",
    "                    iteration_list=[2500],\n",
    "                    restart_list=[10])\n",
    "rhc_stats, rhc_curve = rhc.run()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "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>Time</th>\n",
       "      <th>Fitness</th>\n",
       "      <th>Restarts</th>\n",
       "      <th>current_restart</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>19</th>\n",
       "      <td>1.556132</td>\n",
       "      <td>23.0</td>\n",
       "      <td>10</td>\n",
       "      <td>9</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>21</th>\n",
       "      <td>1.960634</td>\n",
       "      <td>14.0</td>\n",
       "      <td>10</td>\n",
       "      <td>10</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>0.195303</td>\n",
       "      <td>13.0</td>\n",
       "      <td>10</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>13</th>\n",
       "      <td>0.815081</td>\n",
       "      <td>13.0</td>\n",
       "      <td>10</td>\n",
       "      <td>6</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11</th>\n",
       "      <td>0.639576</td>\n",
       "      <td>11.0</td>\n",
       "      <td>10</td>\n",
       "      <td>5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>0.398278</td>\n",
       "      <td>8.0</td>\n",
       "      <td>10</td>\n",
       "      <td>4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>15</th>\n",
       "      <td>0.947102</td>\n",
       "      <td>7.0</td>\n",
       "      <td>10</td>\n",
       "      <td>7</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>0.040760</td>\n",
       "      <td>7.0</td>\n",
       "      <td>10</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>0.302325</td>\n",
       "      <td>6.0</td>\n",
       "      <td>10</td>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>14</th>\n",
       "      <td>0.870834</td>\n",
       "      <td>6.0</td>\n",
       "      <td>10</td>\n",
       "      <td>7</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "        Time  Fitness  Restarts  current_restart\n",
       "19  1.556132     23.0        10                9\n",
       "21  1.960634     14.0        10               10\n",
       "5   0.195303     13.0        10                2\n",
       "13  0.815081     13.0        10                6\n",
       "11  0.639576     11.0        10                5\n",
       "9   0.398278      8.0        10                4\n",
       "15  0.947102      7.0        10                7\n",
       "1   0.040760      7.0        10                0\n",
       "7   0.302325      6.0        10                3\n",
       "14  0.870834      6.0        10                7"
      ]
     },
     "execution_count": 37,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "show=10\n",
    "columns = ['Time', 'Fitness', 'Restarts', 'current_restart']\n",
    "pd.set_option('display.max_rows',show)\n",
    "df=pd.read_csv(\"./four_peaks/four_peaks/rhc__four_peaks__run_stats_df.csv\")\n",
    "df[columns].sort_values(by=['Fitness'], ascending=False).head(show)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Randomized Hill Climbing - Total Function Evaluations: 5226\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 504x504 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "max_attempts = 200\n",
    "max_iters = 2500\n",
    "restarts = 10\n",
    "global eval_count\n",
    "eval_count = 0\n",
    "best_state, best_fitness, rhc_curve = mlrh.random_hill_climb(prob,\n",
    "                                                         max_attempts=max_attempts,\n",
    "                                                         max_iters=max_iters,\n",
    "                                                         random_state=random_state,\n",
    "                                                         restarts=restarts,\n",
    "                                                         curve=True)\n",
    "print(\"Randomized Hill Climbing - Total Function Evaluations:\", eval_count)\n",
    "plot_fitness_iteration(rhc_curve,\n",
    "                       \"Four Peaks - Randomized Hill Climbing: restarts: {}\".format(restarts))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Fitness vs Iterations"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 504x504 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "all_curves = {\"MIMIC\": mimic_curve, \"Genetic Alg\": gen_curve, \"Simulated Annealing\": sa_curve, \"Random Hill Climb\": rhc_curve}\n",
    "\n",
    "np.array([len(x) for x in all_curves]).max()\n",
    "\n",
    "plt.xlabel(\"Iterations\")\n",
    "plt.ylabel(\"Fitness\")\n",
    "plt.title(\"Four Peaks\")\n",
    "\n",
    "for name, curve in all_curves.items():\n",
    "    plt.plot(range(len(curve)), curve, label=name, lw=2)\n",
    "\n",
    "    plt.legend(loc=\"best\")\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Function Evaluations vs Input Size"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [],
   "source": [
    "def prob_gen(input=150):\n",
    "    length = input\n",
    "    t_pct = 0.9\n",
    "    global eval_count\n",
    "    eval_count = 0\n",
    "\n",
    "    base_fit_func = mlrh.FourPeaks(t_pct=t_pct).evaluate\n",
    "\n",
    "    def four_peaks_fit(state):\n",
    "        global eval_count\n",
    "        eval_count += 1\n",
    "\n",
    "        return base_fit_func(state)\n",
    "\n",
    "\n",
    "    # Initialize custom fitness function object\n",
    "    fitness = mlrh.CustomFitness(four_peaks_fit)\n",
    "\n",
    "    prob = mlrh.DiscreteOpt(length=length,\n",
    "                        fitness_fn=fitness,\n",
    "                        maximize=True)\n",
    "    \n",
    "    return prob"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [],
   "source": [
    "def eval_count_mimic(prob):\n",
    "    s = time.time()\n",
    "    max_attempts = 100\n",
    "    max_iters = 25\n",
    "    keep_pct=0.1\n",
    "    pop_size = 3000\n",
    "    global eval_count\n",
    "    eval_count = 0\n",
    "    best_state, best_fitness, mimic_curve = mlrh.mimic(prob,\n",
    "                                             max_attempts=max_attempts,\n",
    "                                             max_iters=max_iters,\n",
    "                                             random_state=random_state,\n",
    "                                             pop_size=pop_size,\n",
    "                                             keep_pct=keep_pct,\n",
    "                                             curve=True)\n",
    "    e = time.time()\n",
    "    t_diff = e-s\n",
    "    result = (eval_count, len(mimic_curve), t_diff)\n",
    "    return result"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [],
   "source": [
    "def eval_count_gen(prob):\n",
    "    s = time.time()\n",
    "    max_attempts = 100\n",
    "    max_iters = 25\n",
    "    mutation_prob=0.1\n",
    "    pop_size = 3000\n",
    "    global eval_count\n",
    "    eval_count = 0\n",
    "    best_state, best_fitness, gen_curve = mlrh.genetic_alg(prob,\n",
    "                                                   max_attempts=max_attempts,\n",
    "                                                   max_iters=max_iters,\n",
    "                                                   random_state=random_state,\n",
    "                                                   pop_size=pop_size,\n",
    "                                                   mutation_prob=mutation_prob,\n",
    "                                                   curve=True)\n",
    "    e = time.time()\n",
    "    t_diff = e-s\n",
    "    result = (eval_count, len(gen_curve), t_diff)\n",
    "    return result"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [],
   "source": [
    "def eval_count_sa(prob):\n",
    "    s = time.time()\n",
    "    max_attempts = 200\n",
    "    max_iters = 2500\n",
    "    init_temp = 1\n",
    "    schedule = mlrh.ExpDecay(init_temp)\n",
    "    global eval_count\n",
    "    eval_count = 0\n",
    "    best_state, best_fitness, sa_curve = mlrh.simulated_annealing(prob,\n",
    "                                                             max_attempts=max_attempts,\n",
    "                                                             max_iters=max_iters,\n",
    "                                                             random_state=random_state,\n",
    "                                                             schedule=schedule,\n",
    "                                                             curve=True)\n",
    "    e = time.time()\n",
    "    t_diff = e-s\n",
    "    result = (eval_count, len(sa_curve), t_diff)\n",
    "    return result"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [],
   "source": [
    "def eval_count_rhc(prob):\n",
    "    s = time.time()\n",
    "    max_attempts = 200\n",
    "    max_iters = 2500\n",
    "    restarts = 10\n",
    "    global eval_count\n",
    "    eval_count = 0\n",
    "    best_state, best_fitness, rhc_curve = mlrh.random_hill_climb(prob,\n",
    "                                                         max_attempts=max_attempts,\n",
    "                                                         max_iters=max_iters,\n",
    "                                                         random_state=random_state,\n",
    "                                                         restarts=restarts,\n",
    "                                                         curve=True)\n",
    "    e = time.time()\n",
    "    t_diff = e-s\n",
    "    result = (eval_count, len(rhc_curve), t_diff)\n",
    "    return result"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [],
   "source": [
    "np.random.seed(1)\n",
    "inputs = [10, 20, 30, 40, 50, 60, 70, 80]\n",
    "\n",
    "mimic_curve_eval = [eval_count_mimic(prob_gen(x)) for x in inputs]\n",
    "gen_curve_eval = [eval_count_gen(prob_gen(x)) for x in inputs]\n",
    "sa_curve_eval = [eval_count_sa(prob_gen(x)) for x in inputs]\n",
    "rhc_curve_eval = [eval_count_rhc(prob_gen(x)) for x in inputs]\n",
    "\n",
    "all_curves = {\"MIMIC\": mimic_curve_eval,\n",
    "              \"Genetic Alg\": gen_curve_eval,\n",
    "              \"Simulated Annealing\": sa_curve_eval,\n",
    "              \"Random Hill Climb\": rhc_curve_eval}\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 504x504 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 504x504 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 504x504 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "l = [(\"Inputs\", \"Function Evaluations\"),\n",
    "     (\"Inputs\", \"Iterations\"),\n",
    "     (\"Inputs\", \"Wall Clock Time (sec)\")]\n",
    "\n",
    "for i in range(3):\n",
    "    plt.xlabel(l[i][0])\n",
    "    plt.ylabel(l[i][1])\n",
    "    plt.title(\"Four Peaks Problem\")\n",
    "\n",
    "    for name, curve in all_curves.items():\n",
    "        plt.plot(inputs, [x[i] for x in curve], label=name, lw=2)\n",
    "\n",
    "        plt.legend(loc=\"best\")\n",
    "    plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 504x504 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.xlabel(\"Inputs\")\n",
    "plt.ylabel(\"Avg. Time (sec) / Iteration\")\n",
    "plt.title(\"Knapsack Problem\")\n",
    "\n",
    "for name, curve in all_curves.items():\n",
    "    avg_time_div_iterations = np.array([x[2] for x in curve]) / np.array([x[1] for x in curve])\n",
    "    plt.plot(inputs, avg_time_div_iterations, label=name, lw=2)\n",
    "\n",
    "    plt.legend(loc=\"best\")\n",
    "plt.show()"
   ]
  }
 ],
 "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": "3.7.5"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}