DeepNeuro_gym.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Using TensorFlow backend.\n"
     ]
    }
   ],
   "source": [
    "import os\n",
    "import numpy as np\n",
    "import gym\n",
    "import tensorflow as tf\n",
    "from time import time\n",
    "from GA.genetic_algorithm import GeneticNetworks\n",
    "\n",
    "from utils.atari_environment import AtariEnvironment\n",
    "from utils.continuous_environments import Environment\n",
    "\n",
    "from keras.backend.tensorflow_backend import set_session\n",
    "from utils.networks import get_session"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "gym.logger.set_level(40)\n",
    "os.environ['TF_CPP_MIN_LOG_LEVEL'] = '3'"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "tensorboard_logs = \"GA/tensorboard_gym\""
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "!mkdir GA/tensorboard_gym\n",
    "!mkdir saved_models"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Clear any logs from previous runs\n",
    "import shutil\n",
    "shutil.rmtree(tensorboard_logs)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "WARNING:tensorflow:From /home/jovyan/work/utils/networks.py:9: The name tf.ConfigProto is deprecated. Please use tf.compat.v1.ConfigProto instead.\n",
      "\n",
      "WARNING:tensorflow:From /home/jovyan/work/utils/networks.py:11: The name tf.Session is deprecated. Please use tf.compat.v1.Session instead.\n",
      "\n"
     ]
    }
   ],
   "source": [
    "set_session(get_session())\n",
    "summary_writer = tf.summary.FileWriter(tensorboard_logs)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "observation_space  (6,)\n",
      "action_space  Discrete(3)\n",
      "3 6 Acrobot-v1\n"
     ]
    }
   ],
   "source": [
    "env_name = 'Acrobot-v1'\n",
    "\n",
    "args = {\n",
    "    'gym_env': gym.make(env_name),\n",
    "    'action_repeat': 4\n",
    "}\n",
    "env = Environment(**args)\n",
    "print('observation_space ', env.get_state_size())\n",
    "print('action_space ', gym.make(env_name).action_space)\n",
    "\n",
    "action_dim = gym.make(env_name).action_space.n\n",
    "space_dim = env.get_state_size()[0]\n",
    "\n",
    "print(action_dim, space_dim, env_name)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "--------------------------------------------------\n",
      "Logging params: \n",
      "population_size 64\n",
      "generations 200\n",
      "mutation_variance 0.1\n",
      "survival_ratio 0.1\n",
      "both_parent_percentage 0.8\n",
      "one_parent_percentage 0.1\n",
      "episodes 15\n",
      "--------------------------------------------------\n"
     ]
    }
   ],
   "source": [
    "genetic_pop = GeneticNetworks(architecture=(space_dim * args['action_repeat'],16,action_dim),\n",
    "                                population_size=64, \n",
    "                                generations=200,\n",
    "                                episodes=15, \n",
    "                                mutation_variance=0.1,\n",
    "                                render_env=False,\n",
    "                                verbose=True)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Train"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Generation:   0%|          | 0/200 [00:00<?, ? gen/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "WARNING:tensorflow:From /home/jovyan/work/utils/networks.py:16: The name tf.Summary is deprecated. Please use tf.compat.v1.Summary instead.\n",
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Generation:10| Highest Reward:-78.33333333333333| Average Reward:-90.37916666666666| std Reward: 9.091570233219098| Stagnation: 11| Population size: 64:   5%|▌         | 10/200 [03:38<48:22, 15.28s/ gen]  "
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "CPU times: user 847 ms, sys: 405 ms, total: 1.25 s\n",
      "Wall time: 3min 38s\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "[{'max': -85.13333333333334,\n",
       "  'mean': -431.37395833333335,\n",
       "  'std': 130.73999091559313},\n",
       " {'max': -96.06666666666666,\n",
       "  'mean': -355.6177083333333,\n",
       "  'std': 156.6647292620399},\n",
       " {'max': -82.53333333333333,\n",
       "  'mean': -431.52083333333337,\n",
       "  'std': 141.69451288825155},\n",
       " {'max': -84.33333333333333,\n",
       "  'mean': -390.0864583333333,\n",
       "  'std': 150.21309570799352},\n",
       " {'max': -106.8, 'mean': -425.1645833333333, 'std': 134.22817066925734},\n",
       " {'max': -82.6, 'mean': -216.8875, 'std': 155.31468185366623},\n",
       " {'max': -85.8, 'mean': -408.703125, 'std': 150.4872004199506},\n",
       " {'max': -75.2, 'mean': -111.39791666666667, 'std': 39.63793246505381},\n",
       " {'max': -80.46666666666667,\n",
       "  'mean': -161.07083333333333,\n",
       "  'std': 113.65295669700909},\n",
       " {'max': -78.33333333333333,\n",
       "  'mean': -90.37916666666666,\n",
       "  'std': 9.091570233219098}]"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "%%time\n",
    "genetic_pop.fit(env, summary_writer, num_cpus=4, env_args=args)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "          max        mean         std\n",
      "0  -85.133333 -431.373958  130.739991\n",
      "1  -96.066667 -355.617708  156.664729\n",
      "2  -82.533333 -431.520833  141.694513\n",
      "3  -84.333333 -390.086458  150.213096\n",
      "4 -106.800000 -425.164583  134.228171\n",
      "5  -82.600000 -216.887500  155.314682\n",
      "6  -85.800000 -408.703125  150.487200\n",
      "7  -75.200000 -111.397917   39.637932\n",
      "8  -80.466667 -161.070833  113.652957\n",
      "9  -78.333333  -90.379167    9.091570\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "import matplotlib.pyplot as plt\n",
    "import pandas as pd\n",
    "\n",
    "df = pd.DataFrame(genetic_pop.global_info)\n",
    "print(df)\n",
    "\n",
    "fig = plt.figure(facecolor='white')\n",
    "ax = fig.add_subplot(111)\n",
    "ax.plot(df['max'], label='max')\n",
    "plt.plot(df['mean'], label='mean')\n",
    "plt.legend()\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Evaluate"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "-100.33333333333333"
      ]
     },
     "execution_count": 16,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Lets observe our best network\n",
    "genetic_pop.best_network.evaluate(env, episodes=3, render_env=False, record=False)"
   ]
  },
  {
   "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": "3.7.6"
  },
  "widgets": {
   "application/vnd.jupyter.widget-state+json": {
    "state": {},
    "version_major": 2,
    "version_minor": 0
   }
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}