simple_es_example.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Simple example of using evolution strategies inside `es.py` for toy deterministic fitness function."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "import matplotlib.pyplot as plt\n",
    "import cma\n",
    "from es import SimpleGA, CMAES, PEPG, OpenES"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "# from https://github.com/CMA-ES/pycma/blob/master/cma/fitness_functions.py\n",
    "def rastrigin(x):\n",
    "    \"\"\"Rastrigin test objective function, shifted by 10. units away from origin\"\"\"\n",
    "    x = np.copy(x)\n",
    "    x -= 10.0\n",
    "    if not np.isscalar(x[0]):\n",
    "        N = len(x[0])\n",
    "        return -np.array([10 * N + sum(xi**2 - 10 * np.cos(2 * np.pi * xi)) for xi in x])\n",
    "    N = len(x)\n",
    "    return -(10 * N + sum(x**2 - 10 * np.cos(2 * np.pi * x)))\n",
    "\n",
    "fit_func = rastrigin"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "NPARAMS = 100        # make this a 100-dimensinal problem.\n",
    "NPOPULATION = 101    # use population size of 101.\n",
    "MAX_ITERATION = 4000 # run each solver for 5000 generations."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "# defines a function to use solver to solve fit_func\n",
    "def test_solver(solver):\n",
    "    history = []\n",
    "    for j in range(MAX_ITERATION):\n",
    "        solutions = solver.ask()\n",
    "        fitness_list = np.zeros(solver.popsize)\n",
    "        for i in range(solver.popsize):\n",
    "            fitness_list[i] = fit_func(solutions[i])\n",
    "        solver.tell(fitness_list)\n",
    "        result = solver.result() # first element is the best solution, second element is the best fitness\n",
    "        history.append(result[1])\n",
    "        if (j+1) % 100 == 0:\n",
    "            print(\"fitness at iteration\", (j+1), result[1])\n",
    "    print(\"local optimum discovered by solver:\\n\", result[0])\n",
    "    print(\"fitness score at this local optimum:\", result[1])\n",
    "    return history"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "This is F(0):\n",
      "-10000.0\n"
     ]
    }
   ],
   "source": [
    "x = np.zeros(NPARAMS) # 100-dimensional problem\n",
    "print(\"This is F(0):\")\n",
    "print(rastrigin(x))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "-0.0\n",
      "global optimum point:\n",
      " [ 10.  10.  10.  10.  10.  10.  10.  10.  10.  10.  10.  10.  10.  10.  10.\n",
      "  10.  10.  10.  10.  10.  10.  10.  10.  10.  10.  10.  10.  10.  10.  10.\n",
      "  10.  10.  10.  10.  10.  10.  10.  10.  10.  10.  10.  10.  10.  10.  10.\n",
      "  10.  10.  10.  10.  10.  10.  10.  10.  10.  10.  10.  10.  10.  10.  10.\n",
      "  10.  10.  10.  10.  10.  10.  10.  10.  10.  10.  10.  10.  10.  10.  10.\n",
      "  10.  10.  10.  10.  10.  10.  10.  10.  10.  10.  10.  10.  10.  10.  10.\n",
      "  10.  10.  10.  10.  10.  10.  10.  10.  10.  10.]\n"
     ]
    }
   ],
   "source": [
    "x = np.ones(NPARAMS)*10. # 100-dimensional problem\n",
    "print(rastrigin(x))\n",
    "print(\"global optimum point:\\n\", x)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [],
   "source": [
    "# defines genetic algorithm solver\n",
    "ga = SimpleGA(NPARAMS,                # number of model parameters\n",
    "               sigma_init=0.5,        # initial standard deviation\n",
    "               popsize=NPOPULATION,   # population size\n",
    "               elite_ratio=0.1,       # percentage of the elites\n",
    "               forget_best=False,     # forget the historical best elites\n",
    "               weight_decay=0.00,     # weight decay coefficient\n",
    "              )"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "fitness at iteration 100 -1135.25076943\n",
      "fitness at iteration 200 -968.111229243\n",
      "fitness at iteration 300 -955.78797871\n",
      "fitness at iteration 400 -920.327139724\n",
      "fitness at iteration 500 -920.327139724\n",
      "fitness at iteration 600 -910.553812306\n",
      "fitness at iteration 700 -889.68806569\n",
      "fitness at iteration 800 -852.255322754\n",
      "fitness at iteration 900 -852.255322754\n",
      "fitness at iteration 1000 -797.614253941\n",
      "fitness at iteration 1100 -721.587177388\n",
      "fitness at iteration 1200 -654.698421656\n",
      "fitness at iteration 1300 -626.756546326\n",
      "fitness at iteration 1400 -588.103750242\n",
      "fitness at iteration 1500 -515.337515896\n",
      "fitness at iteration 1600 -476.77271097\n",
      "fitness at iteration 1700 -405.8241054\n",
      "fitness at iteration 1800 -376.054801064\n",
      "fitness at iteration 1900 -351.268683967\n",
      "fitness at iteration 2000 -327.414122797\n",
      "fitness at iteration 2100 -293.677266891\n",
      "fitness at iteration 2200 -282.148910382\n",
      "fitness at iteration 2300 -251.378116918\n",
      "fitness at iteration 2400 -244.029768007\n",
      "fitness at iteration 2500 -217.547069954\n",
      "fitness at iteration 2600 -211.001197639\n",
      "fitness at iteration 2700 -198.57375004\n",
      "fitness at iteration 2800 -193.870712966\n",
      "fitness at iteration 2900 -189.966366147\n",
      "fitness at iteration 3000 -184.559791821\n",
      "fitness at iteration 3100 -178.434189963\n",
      "fitness at iteration 3200 -177.121079173\n",
      "fitness at iteration 3300 -172.77955777\n",
      "fitness at iteration 3400 -169.092147595\n",
      "fitness at iteration 3500 -168.508687514\n",
      "fitness at iteration 3600 -166.91151829\n",
      "fitness at iteration 3700 -164.701194779\n",
      "fitness at iteration 3800 -164.278514256\n",
      "fitness at iteration 3900 -163.792434205\n",
      "fitness at iteration 4000 -162.775219097\n",
      "local optimum discovered by solver:\n",
      " [ 10.97672509   9.99450072  10.01229443  10.98084311  11.99732787\n",
      "   9.99165997   9.01100864   8.00388958  10.00066367  11.9825391\n",
      "   9.98526607   9.97856286  10.98530549   9.98037197  10.0034753\n",
      "   9.00147727  10.00885316  10.96998115   8.01378946   8.01954088\n",
      "  11.01152498   8.99183882  10.00503022  10.99640128   8.03010486\n",
      "   9.98853839  10.01487612   9.99790571   9.98735097  10.98771468\n",
      "   9.00320488  11.97548268   8.00068976   8.98203378   8.00878006\n",
      "   9.01794576   8.01503657  10.99911461   9.02026417  11.00321313\n",
      "   8.99948125  11.01421686  10.04155664   8.98371565  10.01549771\n",
      "  10.00357877  12.9858397    8.01141082   8.98751366   7.0015745\n",
      "  10.00014372   8.99374037  11.0195508    9.9998437    8.00542152\n",
      "  10.01256026   9.98510566  10.01246651   9.99142455  10.01258364\n",
      "  10.98714778   8.01499142   9.99883081  10.96514086  11.99265527\n",
      "   8.99812563   9.01776565   8.99779376  10.00613626   9.98654185\n",
      "   8.98783937   9.00291737  10.01511231  10.976475     7.02339543\n",
      "   9.03420956  12.97164085   9.96914034   8.00342471  10.98319757\n",
      "  10.00786304   8.01192057   9.00796048   8.03361722  10.02784622\n",
      "   9.0214906    9.00041672  10.99610603  10.0017115   10.99841305\n",
      "   7.99971194  10.98709663  10.02068818   8.01339088   9.98087992\n",
      "   8.01767784   8.99116741   8.00887198  10.99415526   9.99633568]\n",
      "fitness score at this local optimum: -162.775219097\n"
     ]
    }
   ],
   "source": [
    "ga_history = test_solver(ga)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(50_w,101)-aCMA-ES (mu_w=27.2,w_1=8%) in dimension 100 (seed=234125, Tue Sep  4 18:07:01 2018)\n"
     ]
    }
   ],
   "source": [
    "# defines CMA-ES algorithm solver\n",
    "cmaes = CMAES(NPARAMS,\n",
    "              popsize=NPOPULATION,\n",
    "              weight_decay=0.0,\n",
    "              sigma_init = 0.5\n",
    "          )"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "fitness at iteration 100 -992.8697680170322\n",
      "fitness at iteration 200 -855.8365206423846\n",
      "fitness at iteration 300 -783.3718976444816\n",
      "fitness at iteration 400 -313.025161688519\n",
      "fitness at iteration 500 -42.050337808791596\n",
      "fitness at iteration 600 -41.78846503773707\n",
      "fitness at iteration 700 -41.788275430231806\n",
      "fitness at iteration 800 -41.78827536015649\n",
      "fitness at iteration 900 -41.788275360099306\n",
      "fitness at iteration 1000 -41.788275360099306\n",
      "fitness at iteration 1100 -41.788275360099306\n",
      "fitness at iteration 1200 -41.788275360099306\n",
      "fitness at iteration 1300 -41.788275360099306\n",
      "fitness at iteration 1400 -41.788275360099306\n",
      "fitness at iteration 1500 -41.788275360099306\n",
      "fitness at iteration 1600 -41.788275360099306\n",
      "fitness at iteration 1700 -41.788275360099306\n",
      "fitness at iteration 1800 -41.788275360099306\n",
      "fitness at iteration 1900 -41.788275360099306\n",
      "fitness at iteration 2000 -41.788275360099306\n",
      "fitness at iteration 2100 -41.788275360099306\n",
      "fitness at iteration 2200 -41.788275360099306\n",
      "fitness at iteration 2300 -41.788275360099306\n",
      "fitness at iteration 2400 -41.788275360099306\n",
      "fitness at iteration 2500 -41.788275360099306\n",
      "fitness at iteration 2600 -41.788275360099306\n",
      "fitness at iteration 2700 -41.788275360099306\n",
      "fitness at iteration 2800 -41.788275360099306\n",
      "fitness at iteration 2900 -41.788275360099306\n",
      "fitness at iteration 3000 -41.788275360099306\n",
      "fitness at iteration 3100 -41.788275360099306\n",
      "fitness at iteration 3200 -41.788275360099306\n",
      "fitness at iteration 3300 -41.788275360099306\n",
      "fitness at iteration 3400 -41.788275360099306\n",
      "fitness at iteration 3500 -41.788275360099306\n",
      "fitness at iteration 3600 -41.788275360099306\n",
      "fitness at iteration 3700 -41.788275360099306\n",
      "fitness at iteration 3800 -41.788275360099306\n",
      "fitness at iteration 3900 -41.788275360099306\n",
      "fitness at iteration 4000 -41.788275360099306\n",
      "local optimum discovered by solver:\n",
      " [ 10.           9.00504136  10.           9.00504136   9.00504137\n",
      "   9.00504136  10.          10.99495864  10.           9.00504136\n",
      "  10.00000001  10.99495865   9.00504136  10.          10.          10.          10.\n",
      "   9.00504136  10.99495863  10.          10.99495863  10.          10.00000001\n",
      "  10.00000001   8.01008777  10.           9.00504137  10.           9.99999999\n",
      "  10.          10.00000001   9.99999999  10.           9.00504137  10.\n",
      "   9.00504136   9.00504136  10.00000001  10.99495863  10.99495864\n",
      "   9.00504136  10.          10.00000001   9.99999999   9.00504137\n",
      "   9.99999999  10.          10.           9.00504136  10.99495864\n",
      "  10.99495863  10.           9.99999999  10.          10.          10.\n",
      "   9.00504136  10.99495864  10.00000001  10.           9.00504137  10.\n",
      "  10.99495864  10.99495864  10.00000001  10.99495863   9.00504135  10.\n",
      "   9.99999999  10.          10.          10.99495864  10.99495865\n",
      "   9.00504136  10.          10.           9.99999999  10.           9.99999999\n",
      "  10.00000001  10.00000001  10.           9.00504136   9.99999999\n",
      "  10.99495863  10.          10.99495863  10.99495864  10.00000001\n",
      "   9.00504136  10.00000001  10.          10.          10.           9.00504136\n",
      "  10.          10.           9.99999999  10.          10.00000002]\n",
      "fitness score at this local optimum: -41.788275360099306\n"
     ]
    }
   ],
   "source": [
    "cma_history = test_solver(cmaes)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [],
   "source": [
    "# defines PEPG (NES) solver\n",
    "pepg = PEPG(NPARAMS,                         # number of model parameters\n",
    "            sigma_init=0.5,                  # initial standard deviation\n",
    "            learning_rate=0.1,               # learning rate for standard deviation\n",
    "            learning_rate_decay=1.0,       # don't anneal the learning rate\n",
    "            popsize=NPOPULATION,             # population size\n",
    "            average_baseline=False,          # set baseline to average of batch\n",
    "            weight_decay=0.00,            # weight decay coefficient\n",
    "            rank_fitness=False,           # use rank rather than fitness numbers\n",
    "            forget_best=False)            # don't keep the historical best solution)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "fitness at iteration 100 -3634.61752301\n",
      "fitness at iteration 200 -1341.36432923\n",
      "fitness at iteration 300 -984.891975596\n",
      "fitness at iteration 400 -820.140411991\n",
      "fitness at iteration 500 -630.913844661\n",
      "fitness at iteration 600 -462.167795841\n",
      "fitness at iteration 700 -403.452034725\n",
      "fitness at iteration 800 -319.551110879\n",
      "fitness at iteration 900 -219.408990937\n",
      "fitness at iteration 1000 -207.067627133\n",
      "fitness at iteration 1100 -201.696733769\n",
      "fitness at iteration 1200 -165.567121035\n",
      "fitness at iteration 1300 -152.353599613\n",
      "fitness at iteration 1400 -144.409940202\n",
      "fitness at iteration 1500 -140.03668276\n",
      "fitness at iteration 1600 -136.413788473\n",
      "fitness at iteration 1700 -134.440263492\n",
      "fitness at iteration 1800 -133.462306314\n",
      "fitness at iteration 1900 -131.932555482\n",
      "fitness at iteration 2000 -131.619768809\n",
      "fitness at iteration 2100 -130.921828364\n",
      "fitness at iteration 2200 -130.921828364\n",
      "fitness at iteration 2300 -129.964253508\n",
      "fitness at iteration 2400 -129.683166947\n",
      "fitness at iteration 2500 -129.424069963\n",
      "fitness at iteration 2600 -129.251329282\n",
      "fitness at iteration 2700 -129.251329282\n",
      "fitness at iteration 2800 -129.077562363\n",
      "fitness at iteration 2900 -128.85565274\n",
      "fitness at iteration 3000 -128.82250865\n",
      "fitness at iteration 3100 -128.82250865\n",
      "fitness at iteration 3200 -128.82250865\n",
      "fitness at iteration 3300 -128.82250865\n",
      "fitness at iteration 3400 -128.82250865\n",
      "fitness at iteration 3500 -128.354664319\n",
      "fitness at iteration 3600 -128.059813352\n",
      "fitness at iteration 3700 -127.970780694\n",
      "fitness at iteration 3800 -127.95519121\n",
      "fitness at iteration 3900 -127.936529492\n",
      "fitness at iteration 4000 -127.903155474\n",
      "local optimum discovered by solver:\n",
      " [ 10.01961473  10.00384561   8.99603969  10.00376039  10.00202244\n",
      "  10.98965524  10.99056633  11.98453385  10.99028118   9.99710707\n",
      "   9.00665572  10.99628287   9.9987487    9.99418439  10.01378318\n",
      "  10.99918179  10.99739287  11.9763445    9.98129826   9.01064205\n",
      "   9.99918261  10.9905115   10.99716731  10.99215052   8.95031247\n",
      "  10.00513128   8.01056788  10.00526991  10.99336915  11.99245762\n",
      "   8.99664358  10.99589626  10.99752655  12.98180671   9.99886669\n",
      "  11.98638826  11.0040743    9.00522993  11.98876187  11.98827291\n",
      "  11.00065362   9.00298452   9.99843922  11.98387404   9.01123762\n",
      "  10.00400764  10.99387755  10.99074414   9.99254525   9.99552831\n",
      "  10.98626598   9.9959352   10.99966687  10.98732211  11.01700644\n",
      "  11.98498094   8.01701287   8.99417894  10.00038439   9.99978239\n",
      "  11.00482054  10.99051431  10.00028136   8.01588162  10.9897487\n",
      "  10.00112807   8.98334035  12.99419065   9.01039287   9.99622875\n",
      "  10.00279622  11.98657915  10.99825347  10.00117812   9.9985847\n",
      "  11.99224988   9.99847225  10.00560234   9.99973099  10.00844029\n",
      "  10.00553205   9.99582849   9.99451903  10.99497116  10.99004273\n",
      "  11.99128624  10.98286678  10.99461444  10.01648052  10.98055134\n",
      "  11.99104532   9.99960975  10.00290433   9.99892735  12.00398291\n",
      "  10.00305775  10.99785973  11.98693747   9.99910614   9.98826357]\n",
      "fitness score at this local optimum: -127.903155474\n"
     ]
    }
   ],
   "source": [
    "pepg_history = test_solver(pepg)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [],
   "source": [
    "# defines OpenAI's ES algorithm solver. Note that we needed to anneal the sigma parameter\n",
    "oes = OpenES(NPARAMS,                  # number of model parameters\n",
    "            sigma_init=0.5,            # initial standard deviation\n",
    "            sigma_decay=0.999,         # don't anneal standard deviation\n",
    "            learning_rate=0.1,         # learning rate for standard deviation\n",
    "            learning_rate_decay = 1.0, # annealing the learning rate\n",
    "            popsize=NPOPULATION,       # population size\n",
    "            antithetic=False,          # whether to use antithetic sampling\n",
    "            weight_decay=0.00,         # weight decay coefficient\n",
    "            rank_fitness=False,        # use rank rather than fitness numbers\n",
    "            forget_best=False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "fitness at iteration 100 -2547.72753072\n",
      "fitness at iteration 200 -837.308947128\n",
      "fitness at iteration 300 -795.150652289\n",
      "fitness at iteration 400 -795.150652289\n",
      "fitness at iteration 500 -795.150652289\n",
      "fitness at iteration 600 -725.37344417\n",
      "fitness at iteration 700 -651.22994325\n",
      "fitness at iteration 800 -615.048148284\n",
      "fitness at iteration 900 -555.995472926\n",
      "fitness at iteration 1000 -494.296291472\n",
      "fitness at iteration 1100 -427.11601359\n",
      "fitness at iteration 1200 -385.325115707\n",
      "fitness at iteration 1300 -363.263045154\n",
      "fitness at iteration 1400 -296.737702306\n",
      "fitness at iteration 1500 -296.737702306\n",
      "fitness at iteration 1600 -296.737702306\n",
      "fitness at iteration 1700 -268.211850956\n",
      "fitness at iteration 1800 -228.436063113\n",
      "fitness at iteration 1900 -228.436063113\n",
      "fitness at iteration 2000 -198.152382653\n",
      "fitness at iteration 2100 -188.20569125\n",
      "fitness at iteration 2200 -177.675502216\n",
      "fitness at iteration 2300 -177.675502216\n",
      "fitness at iteration 2400 -175.776811507\n",
      "fitness at iteration 2500 -163.618514907\n",
      "fitness at iteration 2600 -157.053733385\n",
      "fitness at iteration 2700 -143.880760633\n",
      "fitness at iteration 2800 -143.721770211\n",
      "fitness at iteration 2900 -143.721770211\n",
      "fitness at iteration 3000 -136.063215595\n",
      "fitness at iteration 3100 -132.834987971\n",
      "fitness at iteration 3200 -132.834987971\n",
      "fitness at iteration 3300 -132.834987971\n",
      "fitness at iteration 3400 -132.834987971\n",
      "fitness at iteration 3500 -132.834987971\n",
      "fitness at iteration 3600 -132.834987971\n",
      "fitness at iteration 3700 -132.834987971\n",
      "fitness at iteration 3800 -132.834987971\n",
      "fitness at iteration 3900 -119.073762315\n",
      "fitness at iteration 4000 -119.073762315\n",
      "local optimum discovered by solver:\n",
      " [  9.96222399  10.03581485   9.98721106  10.9794934   11.03020663\n",
      "  10.96076394  11.00446454  11.00247883  10.98392224  10.94027612\n",
      "  10.0504529   10.97830386   9.0312129    9.92128733   8.92664915\n",
      "  10.00189549   8.95754956  10.99974551   9.97280674   9.98013623\n",
      "  10.94491159   8.94378391  10.07967037  10.97622273  10.97014864\n",
      "   8.98547047  10.00366378  11.03164668  12.03111177  10.95637449\n",
      "  10.98361494  10.9748333   11.09157689  10.04065297  11.00006433\n",
      "  10.05198133  10.99228515   9.97397095   9.11027833  12.96501191\n",
      "   9.99265986   9.00709144  10.0509037   10.01431034   9.940433\n",
      "  10.97644353  10.01497257  11.92533512   9.99708173   9.12054999\n",
      "  10.992724    11.00459189  10.01703682  12.02695887   8.00249103\n",
      "  10.00167295  11.03613397  10.9779862   10.00749282  10.98325682\n",
      "  11.05239174  10.00549102  11.9927345    9.00513705  10.00676037\n",
      "   9.97190821  10.98698358   9.00707313   9.9436845   10.99244854\n",
      "  10.00464128  10.97691325   9.98609334  10.04023315  10.01552519\n",
      "  10.96793756  13.0317341    8.96804311   9.96510546  10.00518361\n",
      "  10.93546217  11.00502035  10.06336554   9.9907779   10.05001362\n",
      "  10.98710824   9.01820945   8.9904683    9.05672193   9.99382182\n",
      "  11.06678637  10.9756182    9.94234177  10.95908423   9.97904453\n",
      "   9.9734558    9.94889348  10.97144241  10.96124909   9.0287699 ]\n",
      "fitness score at this local optimum: -119.073762315\n"
     ]
    }
   ],
   "source": [
    "oes_history = test_solver(oes)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [],
   "source": [
    "# defines OpenAI's ES algorithm solver. Note that we needed to anneal the sigma parameter\n",
    "# this version turns on antithetic sampling. It doesn't really help, and sometimes hurts performance.\n",
    "oes_antithetic = OpenES(NPARAMS,            # number of model parameters\n",
    "                 sigma_init=0.5,            # initial standard deviation\n",
    "                 sigma_decay=0.999,         # don't anneal standard deviation\n",
    "                 learning_rate=0.1,         # learning rate for standard deviation\n",
    "                 learning_rate_decay=1.0,   # annealing the learning rate\n",
    "                 popsize=NPOPULATION+1,     # population size\n",
    "                 antithetic=True,           # whether to use antithetic sampling\n",
    "                 weight_decay=0.00,         # weight decay coefficient\n",
    "                 rank_fitness=False,        # use rank rather than fitness numbers\n",
    "                 forget_best=False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "fitness at iteration 100 -3249.4558748\n",
      "fitness at iteration 200 -905.683279874\n",
      "fitness at iteration 300 -782.282701518\n",
      "fitness at iteration 400 -782.282701518\n",
      "fitness at iteration 500 -782.282701518\n",
      "fitness at iteration 600 -752.944311414\n",
      "fitness at iteration 700 -681.00298853\n",
      "fitness at iteration 800 -629.418122969\n",
      "fitness at iteration 900 -571.037612384\n",
      "fitness at iteration 1000 -506.852114247\n",
      "fitness at iteration 1100 -440.331013572\n",
      "fitness at iteration 1200 -380.178757215\n",
      "fitness at iteration 1300 -335.373179856\n",
      "fitness at iteration 1400 -335.373179856\n",
      "fitness at iteration 1500 -319.827206653\n",
      "fitness at iteration 1600 -300.108061502\n",
      "fitness at iteration 1700 -290.415921385\n",
      "fitness at iteration 1800 -235.871192792\n",
      "fitness at iteration 1900 -207.680078014\n",
      "fitness at iteration 2000 -207.680078014\n",
      "fitness at iteration 2100 -207.680078014\n",
      "fitness at iteration 2200 -207.680078014\n",
      "fitness at iteration 2300 -207.680078014\n",
      "fitness at iteration 2400 -207.680078014\n",
      "fitness at iteration 2500 -206.826183977\n",
      "fitness at iteration 2600 -194.751620165\n",
      "fitness at iteration 2700 -194.751620165\n",
      "fitness at iteration 2800 -194.751620165\n",
      "fitness at iteration 2900 -194.751620165\n",
      "fitness at iteration 3000 -184.163308157\n",
      "fitness at iteration 3100 -184.163308157\n",
      "fitness at iteration 3200 -184.163308157\n",
      "fitness at iteration 3300 -183.48751021\n",
      "fitness at iteration 3400 -174.345761059\n",
      "fitness at iteration 3500 -174.345761059\n",
      "fitness at iteration 3600 -174.345761059\n",
      "fitness at iteration 3700 -174.345761059\n",
      "fitness at iteration 3800 -174.345761059\n",
      "fitness at iteration 3900 -174.345761059\n",
      "fitness at iteration 4000 -174.345761059\n",
      "local optimum discovered by solver:\n",
      " [ 11.11888794  11.11738517  11.05755268  10.97955645  10.01730772\n",
      "   8.99037223  10.05631989  11.93023509  10.01683307  10.90354934\n",
      "   9.04082188   8.94886975   9.10047183   9.97122681  10.11512472\n",
      "   9.91322231   9.93481562  10.09862652   9.89323582  11.04935905\n",
      "   9.96734987  11.01444488  10.98531918  10.06688134  11.14169528\n",
      "  10.02138664   9.97270597  10.99687272  10.06114293   9.99064074\n",
      "  10.03533774  10.96545774  10.07952058  11.02385502  10.94115267\n",
      "  11.04676738  12.05077682  11.01796634  11.95468163   9.94293649\n",
      "   9.07971396   9.87962282   9.93819201  10.00743849  11.06174766\n",
      "  10.86197627  10.12220057  10.00278321   9.94279514   9.95592564\n",
      "  10.06193083  11.94087051  11.06637086  10.91863424  10.92866464\n",
      "  10.00300299  12.02358086   8.92083608  11.03192709   9.96456953\n",
      "  11.94505205  11.04076089   8.93660681  11.03771843  10.9731474\n",
      "  11.98934571  10.91490246  11.04677918   8.96246687   9.99570488\n",
      "  10.96014829  10.06468044   8.95612409  10.02019367  10.97810255\n",
      "  11.97961555  10.87702715  12.04049814  10.01191171   9.96356814\n",
      "   9.97247573   9.06371765  11.07953805  10.86567807  11.0254814\n",
      "  10.94385121  10.99931823  11.01254205   9.09950367  12.02094477\n",
      "   8.97796135  11.85216809  10.99209484  10.02231263  11.09378876\n",
      "   9.04402783   9.98980739  10.96810385   9.98329216  11.92838249]\n",
      "fitness score at this local optimum: -174.345761059\n"
     ]
    }
   ],
   "source": [
    "oes_antithetic_history = test_solver(oes_antithetic)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x10f210ba8>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Create a new figure of size 8x6 points, using 100 dots per inch\n",
    "best_history = [0] * MAX_ITERATION\n",
    "plt.figure(figsize=(16,8), dpi=150)\n",
    "\n",
    "optimum_line, = plt.plot(best_history, color=\"black\", linewidth=0.5, linestyle=\"-.\", label='Global Optimum')\n",
    "ga_line, = plt.plot(ga_history, color=\"green\", linewidth=1.0, linestyle=\"-\", label='GA')\n",
    "oes_line, = plt.plot(oes_history, color=\"orange\", linewidth=1.0, linestyle=\"-\", label='OpenAI-ES')\n",
    "pepg_line, = plt.plot(pepg_history, color=\"blue\", linewidth=1.0, linestyle=\"-\", label='PEPG / NES')\n",
    "cma_line, = plt.plot(cma_history, color=\"red\", linewidth=1.0, linestyle=\"-\", label='CMA-ES')\n",
    "\n",
    "plt.legend(handles=[optimum_line, ga_line, cma_line, pepg_line, oes_line], loc=4)\n",
    "\n",
    "# Set x limits\n",
    "plt.xlim(0,2500)\n",
    "\n",
    "plt.xlabel('generation')\n",
    "plt.ylabel('fitness')\n",
    "\n",
    "# plt.savefig(\"./rastrigin_10d.svg\")\n",
    "\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.6.5"
  },
  "toc": {
   "base_numbering": 1,
   "nav_menu": {},
   "number_sections": true,
   "sideBar": true,
   "skip_h1_title": false,
   "title_cell": "Table of Contents",
   "title_sidebar": "Contents",
   "toc_cell": false,
   "toc_position": {},
   "toc_section_display": true,
   "toc_window_display": false
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}