- My Likelihood Agent.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Environment Setup"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {
    "pycharm": {
     "is_executing": true
    }
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/workspace/ML/reco-gym/recogym/envs/abstract.py:54: FutureWarning: Passing a negative integer is deprecated in version 1.0 and will not be supported in future version. Instead, use None to not limit the column width.\n",
      "  pd.set_option('display.max_colwidth', -1)\n"
     ]
    }
   ],
   "source": [
    "from copy import deepcopy\n",
    "from my_agent_utils import add_agent_id\n",
    "\n",
    "from my_entries.likehood_agent import LikelihoodAgent, CountFeatureProvider\n",
    "from recogym.evaluate_agent import plot_verify_agents, verify_agents\n",
    "%matplotlib inline\n",
    "import gym\n",
    "import matplotlib.pyplot as plt\n",
    "\n",
    "from recogym import env_1_args, Configuration\n",
    "\n",
    "# Set style for pretty plots\n",
    "plt.style.use('ggplot')\n",
    "\n",
    "products = 10\n",
    "\n",
    "offline_users = 1000\n",
    "online_users = 200\n",
    "\n",
    "# You can overwrite environment arguments here:\n",
    "\n",
    "env_1_args['random_seed'] = 42\n",
    "env_1_args['num_products']= products\n",
    "env_1_args['phi_var']=0.0\n",
    "env_1_args['number_of_flips']=1\n",
    "\n",
    "# env_1_args['sigma_mu_organic'] = 0.0\n",
    "# env_1_args['sigma_omega']=0\n",
    "# env_1_args['normalize_beta']=True\n",
    "\n",
    "# Initialize the gym for the first time by calling .make() and .init_gym()\n",
    "\n",
    "#---------------------------------------------\n",
    "%config InlineBackend.figure_format = 'retina'\n",
    "plt.rcParams['figure.figsize'] = [6, 3]\n",
    "\n",
    "env_1_args['number_of_flips'] = 0\n",
    "env_1_args['sigma_mu_organic'] = 0.0\n",
    "env_1_args['sigma_omega'] = 1\n",
    "env_1_args['random_seed'] = 42\n",
    "env_1_args['num_products'] = products\n",
    "env_1_args['K'] = 5\n",
    "env_1_args['number_of_flips'] = 5\n",
    "#---------------------------------------------\n",
    "\n",
    "env = gym.make('reco-gym-v1')\n",
    "env.init_gym(env_1_args)\n",
    "\n",
    "env.reset()\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "8. Define Likelihood Agent"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "pycharm": {
     "name": "#%%\n"
    }
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Organic Users: 0it [00:00, ?it/s]\n",
      "Users:   0%|          | 0/1000 [00:00<?, ?it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "# ------------------------------------------------------------------------------------\n",
      "#  GENERATE LOGS FOR AGENT None\n",
      "# ------------------------------------------------------------------------------------\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Users: 100%|██████████| 1000/1000 [00:13<00:00, 76.46it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "LikelihoodAgent train() num_products 10\n"
     ]
    }
   ],
   "source": [
    "from scipy.stats import beta\n",
    "\n",
    "config = Configuration(env_1_args)\n",
    "\n",
    "count_product_views_feature_provider = CountFeatureProvider(config)  # <----- CountFeatureProvider\n",
    "\n",
    "likelihood_agent = LikelihoodAgent(count_product_views_feature_provider)\n",
    "\n",
    "lh_log_data = deepcopy(env).generate_logs(offline_users)\n",
    "\n",
    "likelihood_agent.train(lh_log_data)\n",
    "\n",
    "bandits = lh_log_data[lh_log_data['z'] == 'bandit']\n",
    "successes = bandits[bandits['c'] == 1].shape[0]\n",
    "failures = bandits[bandits['c'] == 0].shape[0]\n",
    "\n",
    "lh_agent_res = add_agent_id(\n",
    "    'Likelihood',\n",
    "    beta.ppf(0.025, successes + 1, failures + 1),\n",
    "    beta.ppf(0.500, successes + 1, failures + 1),\n",
    "    beta.ppf(0.975, successes + 1, failures + 1))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {
    "pycharm": {
     "name": "#%%\n"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "102137"
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#ALL LEN\n",
    "\n",
    "len(lh_log_data)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {
    "pycharm": {
     "name": "#%%\n"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "17293"
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#BANDITS LEN\n",
    "len(bandits)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {
    "pycharm": {
     "name": "#%%\n"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "241"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#SUCCESS\n",
    "successes"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {
    "pycharm": {
     "name": "#%%\n"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "17052"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#FAILURES\n",
    "failures\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {
    "pycharm": {
     "name": "#%%\n"
    }
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Organic Users: 0it [00:00, ?it/s]\n",
      "Users:   1%|          | 9/1000 [00:00<00:11, 87.43it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "# ------------------------------------------------------------------------------------\n",
      "#  GENERATE LOGS FOR AGENT <my_entries.likehood_agent.LikelihoodAgent object at 0x134b16510>\n",
      "# ------------------------------------------------------------------------------------\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Users: 100%|██████████| 1000/1000 [00:15<00:00, 64.95it/s]\n"
     ]
    }
   ],
   "source": [
    "comb_result = verify_agents(\n",
    "    env,\n",
    "    number_of_users=online_users,\n",
    "    agents={\n",
    "        'Likelihood LogReg': likelihood_agent,\n",
    "    }\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {
    "pycharm": {
     "name": "#%%\n"
    }
   },
   "outputs": [
    {
     "ename": "AttributeError",
     "evalue": "'function' object has no attribute 'set_title'",
     "output_type": "error",
     "traceback": [
      "\u001B[0;31m---------------------------------------------------------------------------\u001B[0m",
      "\u001B[0;31mAttributeError\u001B[0m                            Traceback (most recent call last)",
      "\u001B[0;32m<ipython-input-8-3fa5afb8d3c2>\u001B[0m in \u001B[0;36m<module>\u001B[0;34m\u001B[0m\n\u001B[1;32m      1\u001B[0m \u001B[0mfig\u001B[0m \u001B[0;34m=\u001B[0m \u001B[0mplot_verify_agents\u001B[0m\u001B[0;34m(\u001B[0m\u001B[0mcomb_result\u001B[0m\u001B[0;34m)\u001B[0m\u001B[0;34m\u001B[0m\u001B[0;34m\u001B[0m\u001B[0m\n\u001B[1;32m      2\u001B[0m \u001B[0mplt\u001B[0m\u001B[0;34m.\u001B[0m\u001B[0mylabel\u001B[0m\u001B[0;34m(\u001B[0m\u001B[0;34m'CTR'\u001B[0m\u001B[0;34m)\u001B[0m\u001B[0;34m\u001B[0m\u001B[0;34m\u001B[0m\u001B[0m\n\u001B[0;32m----> 3\u001B[0;31m \u001B[0mplt\u001B[0m\u001B[0;34m.\u001B[0m\u001B[0msubplot\u001B[0m\u001B[0;34m.\u001B[0m\u001B[0mset_title\u001B[0m\u001B[0;34m(\u001B[0m\u001B[0;34m'Bandit 16%, Organic 84%'\u001B[0m\u001B[0;34m)\u001B[0m\u001B[0;34m\u001B[0m\u001B[0;34m\u001B[0m\u001B[0m\n\u001B[0m\u001B[1;32m      4\u001B[0m \u001B[0mplt\u001B[0m\u001B[0;34m.\u001B[0m\u001B[0mshow\u001B[0m\u001B[0;34m(\u001B[0m\u001B[0;34m)\u001B[0m\u001B[0;34m\u001B[0m\u001B[0;34m\u001B[0m\u001B[0m\n",
      "\u001B[0;31mAttributeError\u001B[0m: 'function' object has no attribute 'set_title'"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x216 with 1 Axes>"
      ]
     },
     "metadata": {
      "image/png": {
       "height": 292,
       "width": 400
      }
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig = plot_verify_agents(comb_result)\n",
    "plt.ylabel('CTR')\n",
    "plt.show()\n"
   ]
  }
 ],
 "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": "3.7.7"
  },
  "pycharm": {
   "stem_cell": {
    "cell_type": "raw",
    "source": [
     "\n"
    ],
    "metadata": {
     "collapsed": false
    }
   }
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}