Organic vs Likelihood.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {
    "pycharm": {}
   },
   "source": [
    "# _RecoGym_: Organic vs Likelihood\n",
    "\n",
    "# Please note this is a slow notebook to run\n",
    "\n",
    "The _`Agent`_ that uses _Organic_ _`Event`_ counter shows quite good results in predicting a click.\n",
    "In this notebook you shall find how that _`Agent`_ performs when the amount of _`Products`_ flips in the _`Environment`_ is big enough to make such kind of predictions hard.\n",
    "\n",
    "**Note:** the evaluation of the data provided in that notebook takes some time."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "pycharm": {}
   },
   "source": [
    "Below you shall find the code that initialises the _`Environment`_."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {
    "pycharm": {}
   },
   "outputs": [],
   "source": [
    "import gym\n",
    "import matplotlib.pyplot as plt\n",
    "\n",
    "%matplotlib notebook\n",
    "%config InlineBackend.figure_format = 'retina'\n",
    "plt.rcParams['figure.figsize'] = [6, 3]\n",
    "\n",
    "from recogym import build_agent_init, env_1_args, gather_agent_stats, plot_agent_stats\n",
    "\n",
    "RandomSeed = 42\n",
    "\n",
    "TrainingDataSamples = (100, 500, 1000, 2000, 3000, 5000, 8000, 10000, 13000, 14000, 15000)\n",
    "TestingDataSamples = 15000\n",
    "StatEpochs = 10\n",
    "StatEpochsNewRandomSeed = True\n",
    "\n",
    "std_env_args = {\n",
    "    **env_1_args,\n",
    "    'random_seed': RandomSeed,\n",
    "}\n",
    "\n",
    "env = gym.make('reco-gym-v1')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "pycharm": {}
   },
   "source": [
    "## _With_ Product Flips\n",
    "\n",
    "In this case, there is a maximum of _`Products`_ flips in the _`Environment`_ setup."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "pycharm": {}
   },
   "source": [
    "The code that imports _`Agents`_."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "pycharm": {}
   },
   "outputs": [],
   "source": [
    "from recogym.agents import OrganicUserEventCounterAgent, organic_user_count_args\n",
    "from recogym.agents import LogregPolyAgent, logreg_poly_args"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {
    "pycharm": {}
   },
   "outputs": [],
   "source": [
    "agent_inits = {\n",
    "    **build_agent_init('User Organic', OrganicUserEventCounterAgent, organic_user_count_args),\n",
    "    **build_agent_init('LogReg Poly', LogregPolyAgent, logreg_poly_args),\n",
    "}"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "pycharm": {}
   },
   "source": [
    "Gathering statistics..."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {
    "collapsed": true,
    "pycharm": {}
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Agent: User Organic\n",
      "Start: Num of Offline Users: 100\n",
      "Start: Agent Training #0\n",
      "Start: Agent Training #4\n",
      "Start: Agent Training #1\n",
      "Start: Agent Training #3\n",
      "Start: Agent Training #2\n",
      "Start: Agent Training #9\n",
      "Start: Agent Training #6\n",
      "Start: Agent Training #5\n",
      "Start: Agent Training #8\n",
      "Start: Agent Training #7\n",
      "Start: Agent Testing #4\n",
      "Start: Agent Testing #2\n",
      "Start: Agent Testing #8\n",
      "Start: Agent Testing #3\n",
      "Start: Agent Testing #5\n",
      "Start: Agent Testing #0\n",
      "Start: Agent Testing #7\n",
      "Start: Agent Testing #9\n",
      "Start: Agent Testing #1\n",
      "Start: Agent Testing #6\n",
      "End: Agent Testing #3 (123.39248251914978s)\n",
      "End: Agent Testing #8 (124.11361193656921s)\n",
      "End: Agent Testing #1 (124.84213304519653s)\n",
      "End: Agent Testing #5 (125.17546486854553s)\n",
      "End: Agent Testing #6 (125.45348310470581s)\n",
      "End: Agent Testing #0 (125.68971347808838s)\n",
      "End: Agent Testing #7 (126.19205284118652s)\n",
      "End: Agent Testing #4 (127.16778755187988s)\n",
      "End: Agent Testing #2 (127.79104852676392s)\n",
      "End: Agent Testing #9 (128.10331797599792s)\n",
      "End: Num of Offline Users: 100 (128.33882689476013s)\n",
      "Start: Num of Offline Users: 500\n",
      "Start: Agent Training #0\n",
      "Start: Agent Training #1\n",
      "Start: Agent Training #6\n",
      "Start: Agent Training #5\n",
      "Start: Agent Training #3\n",
      "Start: Agent Training #7\n",
      "Start: Agent Training #2\n",
      "Start: Agent Training #9\n",
      "Start: Agent Training #4\n",
      "Start: Agent Training #8\n",
      "Start: Agent Testing #5\n",
      "Start: Agent Testing #2\n",
      "Start: Agent Testing #8\n",
      "Start: Agent Testing #3\n",
      "Start: Agent Testing #4\n",
      "Start: Agent Testing #9\n",
      "Start: Agent Testing #6\n",
      "Start: Agent Testing #1\n",
      "Start: Agent Testing #0\n",
      "Start: Agent Testing #7\n",
      "End: Agent Testing #3 (123.8380184173584s)\n",
      "End: Agent Testing #5 (123.8584930896759s)\n",
      "End: Agent Testing #8 (123.98458862304688s)\n",
      "End: Agent Testing #1 (125.77044534683228s)\n",
      "End: Agent Testing #4 (126.28837060928345s)\n",
      "End: Agent Testing #6 (126.75386047363281s)\n",
      "End: Agent Testing #0 (126.98685359954834s)\n",
      "End: Agent Testing #2 (127.96317744255066s)\n",
      "End: Agent Testing #7 (128.11734318733215s)\n",
      "End: Agent Testing #9 (128.57282376289368s)\n",
      "End: Num of Offline Users: 500 (128.79154753684998s)\n",
      "Start: Num of Offline Users: 1000\n",
      "Start: Agent Training #0\n",
      "Start: Agent Training #4\n",
      "Start: Agent Training #2\n",
      "Start: Agent Training #1\n",
      "Start: Agent Training #3\n",
      "Start: Agent Training #5\n",
      "Start: Agent Training #6\n",
      "Start: Agent Training #7\n",
      "Start: Agent Training #8\n",
      "Start: Agent Training #9\n",
      "Start: Agent Testing #5\n",
      "Start: Agent Testing #2\n",
      "Start: Agent Testing #3\n",
      "Start: Agent Testing #6\n",
      "Start: Agent Testing #4\n",
      "Start: Agent Testing #0\n",
      "Start: Agent Testing #1\n",
      "Start: Agent Testing #7\n",
      "Start: Agent Testing #9\n",
      "Start: Agent Testing #8\n",
      "End: Agent Testing #0 (128.8860900402069s)\n",
      "End: Agent Testing #4 (130.20413947105408s)\n",
      "End: Agent Testing #5 (130.41836881637573s)\n",
      "End: Agent Testing #3 (130.50418639183044s)\n",
      "End: Agent Testing #1 (130.62047600746155s)\n",
      "End: Agent Testing #2 (130.6842339038849s)\n",
      "End: Agent Testing #8 (131.18756937980652s)\n",
      "End: Agent Testing #6 (131.64665842056274s)\n",
      "End: Agent Testing #9 (131.87730073928833s)\n",
      "End: Agent Testing #7 (132.85054421424866s)\n",
      "End: Num of Offline Users: 1000 (133.0790991783142s)\n",
      "Start: Num of Offline Users: 2000\n",
      "Start: Agent Training #1\n",
      "Start: Agent Training #0\n",
      "Start: Agent Training #3\n",
      "Start: Agent Training #2\n",
      "Start: Agent Training #5\n",
      "Start: Agent Training #4\n",
      "Start: Agent Training #6\n",
      "Start: Agent Training #7\n",
      "Start: Agent Training #8\n",
      "Start: Agent Training #9\n",
      "Start: Agent Testing #2\n",
      "Start: Agent Testing #8\n",
      "Start: Agent Testing #0\n",
      "Start: Agent Testing #1\n",
      "Start: Agent Testing #3\n",
      "Start: Agent Testing #9\n",
      "Start: Agent Testing #5\n",
      "Start: Agent Testing #6\n",
      "Start: Agent Testing #7\n",
      "Start: Agent Testing #4\n",
      "End: Agent Testing #1 (136.18636012077332s)\n",
      "End: Agent Testing #9 (136.23517632484436s)\n",
      "End: Agent Testing #2 (136.6247866153717s)\n",
      "End: Agent Testing #5 (136.6869022846222s)\n",
      "End: Agent Testing #0 (136.75905394554138s)\n",
      "End: Agent Testing #3 (137.1209523677826s)\n",
      "End: Agent Testing #7 (137.37060165405273s)\n",
      "End: Agent Testing #6 (137.45069456100464s)\n",
      "End: Agent Testing #4 (137.7482681274414s)\n",
      "End: Agent Testing #8 (139.1800615787506s)\n",
      "End: Num of Offline Users: 2000 (139.42576575279236s)\n",
      "Start: Num of Offline Users: 3000\n",
      "Start: Agent Training #2\n",
      "Start: Agent Training #0\n",
      "Start: Agent Training #1\n",
      "Start: Agent Training #4\n",
      "Start: Agent Training #3\n",
      "Start: Agent Training #5\n",
      "Start: Agent Training #6\n",
      "Start: Agent Training #7\n",
      "Start: Agent Training #8\n",
      "Start: Agent Training #9\n",
      "Start: Agent Testing #2\n",
      "Start: Agent Testing #1\n",
      "Start: Agent Testing #8\n",
      "Start: Agent Testing #3\n",
      "Start: Agent Testing #0\n",
      "Start: Agent Testing #4\n",
      "Start: Agent Testing #5\n",
      "Start: Agent Testing #6\n",
      "Start: Agent Testing #7\n",
      "Start: Agent Testing #9\n",
      "End: Agent Testing #3 (142.20453596115112s)\n",
      "End: Agent Testing #4 (142.64829683303833s)\n",
      "End: Agent Testing #1 (143.28881978988647s)\n",
      "End: Agent Testing #8 (144.0823974609375s)\n",
      "End: Agent Testing #0 (144.23245573043823s)\n",
      "End: Agent Testing #2 (145.05368280410767s)\n",
      "End: Agent Testing #7 (145.19709086418152s)\n",
      "End: Agent Testing #6 (145.2858476638794s)\n",
      "End: Agent Testing #9 (145.80567383766174s)\n",
      "End: Agent Testing #5 (145.84813404083252s)\n",
      "End: Num of Offline Users: 3000 (146.1327257156372s)\n",
      "Start: Num of Offline Users: 5000\n",
      "Start: Agent Training #0\n",
      "Start: Agent Training #3\n",
      "Start: Agent Training #5\n",
      "Start: Agent Training #6\n",
      "Start: Agent Training #2\n",
      "Start: Agent Training #1\n",
      "Start: Agent Training #4\n",
      "Start: Agent Training #7\n",
      "Start: Agent Training #8\n",
      "Start: Agent Training #9\n",
      "Start: Agent Testing #1\n",
      "Start: Agent Testing #2\n",
      "Start: Agent Testing #5\n",
      "Start: Agent Testing #0\n",
      "Start: Agent Testing #3\n",
      "Start: Agent Testing #8\n",
      "Start: Agent Testing #9\n",
      "Start: Agent Testing #7\n",
      "Start: Agent Testing #4\n",
      "Start: Agent Testing #6\n",
      "End: Agent Testing #8 (159.62501120567322s)\n",
      "End: Agent Testing #4 (161.1475546360016s)\n",
      "End: Agent Testing #5 (161.35995054244995s)\n",
      "End: Agent Testing #1 (161.40881276130676s)\n",
      "End: Agent Testing #9 (162.07729840278625s)\n",
      "End: Agent Testing #0 (162.09308099746704s)\n",
      "End: Agent Testing #3 (162.12963104248047s)\n",
      "End: Agent Testing #6 (162.7776505947113s)\n",
      "End: Agent Testing #7 (163.10224652290344s)\n",
      "End: Agent Testing #2 (163.8222999572754s)\n",
      "End: Num of Offline Users: 5000 (164.17798614501953s)\n",
      "Start: Num of Offline Users: 8000\n",
      "Start: Agent Training #1\n",
      "Start: Agent Training #2\n",
      "Start: Agent Training #0\n",
      "Start: Agent Training #4\n",
      "Start: Agent Training #7\n",
      "Start: Agent Training #3\n",
      "Start: Agent Training #5\n",
      "Start: Agent Training #6\n",
      "Start: Agent Training #8\n",
      "Start: Agent Training #9\n",
      "Start: Agent Testing #1\n",
      "Start: Agent Testing #8\n",
      "Start: Agent Testing #7\n",
      "Start: Agent Testing #4\n",
      "Start: Agent Testing #9\n",
      "Start: Agent Testing #0\n",
      "Start: Agent Testing #2\n",
      "Start: Agent Testing #5\n",
      "Start: Agent Testing #6\n",
      "Start: Agent Testing #3\n",
      "End: Agent Testing #8 (179.05088806152344s)\n",
      "End: Agent Testing #9 (180.56373143196106s)\n",
      "End: Agent Testing #2 (181.96650338172913s)\n",
      "End: Agent Testing #1 (182.61831998825073s)\n",
      "End: Agent Testing #4 (182.62855672836304s)\n",
      "End: Agent Testing #7 (183.99360728263855s)\n",
      "End: Agent Testing #6 (184.26527571678162s)\n",
      "End: Agent Testing #5 (184.60087251663208s)\n",
      "End: Agent Testing #3 (184.69710540771484s)\n",
      "End: Agent Testing #0 (184.87769651412964s)\n",
      "End: Num of Offline Users: 8000 (185.47872185707092s)\n",
      "Start: Num of Offline Users: 10000\n",
      "Start: Agent Training #0\n",
      "Start: Agent Training #1\n",
      "Start: Agent Training #4\n",
      "Start: Agent Training #2\n",
      "Start: Agent Training #3\n",
      "Start: Agent Training #5\n",
      "Start: Agent Training #7\n",
      "Start: Agent Training #6\n",
      "Start: Agent Training #9\n",
      "Start: Agent Training #8\n",
      "Start: Agent Testing #4\n",
      "Start: Agent Testing #9\n",
      "Start: Agent Testing #5\n",
      "Start: Agent Testing #2\n",
      "Start: Agent Testing #8\n",
      "Start: Agent Testing #0\n",
      "Start: Agent Testing #3\n",
      "Start: Agent Testing #6\n",
      "Start: Agent Testing #1\n",
      "Start: Agent Testing #7\n",
      "End: Agent Testing #9 (195.36398839950562s)\n",
      "End: Agent Testing #8 (196.06080794334412s)\n",
      "End: Agent Testing #3 (196.44597625732422s)\n",
      "End: Agent Testing #2 (196.91143894195557s)\n",
      "End: Agent Testing #5 (197.92338585853577s)\n",
      "End: Agent Testing #6 (199.02118515968323s)\n",
      "End: Agent Testing #0 (199.40210604667664s)\n",
      "End: Agent Testing #7 (200.64196944236755s)\n",
      "End: Agent Testing #1 (201.11883544921875s)\n",
      "End: Agent Testing #4 (201.65825986862183s)\n",
      "End: Num of Offline Users: 10000 (202.1901273727417s)\n",
      "Start: Num of Offline Users: 13000\n",
      "Start: Agent Training #0\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Start: Agent Training #2\n",
      "Start: Agent Training #1\n",
      "Start: Agent Training #3\n",
      "Start: Agent Training #5\n",
      "Start: Agent Training #4\n",
      "Start: Agent Training #6\n",
      "Start: Agent Training #7\n",
      "Start: Agent Training #8\n",
      "Start: Agent Training #9\n",
      "Start: Agent Testing #5\n",
      "Start: Agent Testing #8\n",
      "Start: Agent Testing #2\n",
      "Start: Agent Testing #4\n",
      "Start: Agent Testing #7\n",
      "Start: Agent Testing #6\n",
      "Start: Agent Testing #3\n",
      "Start: Agent Testing #1\n",
      "Start: Agent Testing #9\n",
      "Start: Agent Testing #0\n",
      "End: Agent Testing #5 (219.03441309928894s)\n",
      "End: Agent Testing #2 (219.1699550151825s)\n",
      "End: Agent Testing #7 (220.1464171409607s)\n",
      "End: Agent Testing #9 (220.52815294265747s)\n",
      "End: Agent Testing #8 (221.17253613471985s)\n",
      "End: Agent Testing #1 (221.95933318138123s)\n",
      "End: Agent Testing #3 (221.97778248786926s)\n",
      "End: Agent Testing #0 (222.3791115283966s)\n",
      "End: Agent Testing #6 (223.48102927207947s)\n",
      "End: Agent Testing #4 (248.54747343063354s)\n",
      "End: Num of Offline Users: 13000 (249.1576554775238s)\n",
      "Start: Num of Offline Users: 14000\n",
      "Start: Agent Training #0\n",
      "Start: Agent Training #2\n",
      "Start: Agent Training #4\n",
      "Start: Agent Training #3\n",
      "Start: Agent Training #6\n",
      "Start: Agent Training #5\n",
      "Start: Agent Training #1\n",
      "Start: Agent Training #7\n",
      "Start: Agent Training #8\n",
      "Start: Agent Training #9\n",
      "Start: Agent Testing #8\n",
      "Start: Agent Testing #2\n",
      "Start: Agent Testing #4\n",
      "Start: Agent Testing #9\n",
      "Start: Agent Testing #5\n",
      "Start: Agent Testing #3\n",
      "Start: Agent Testing #6\n",
      "Start: Agent Testing #0\n",
      "Start: Agent Testing #7\n",
      "Start: Agent Testing #1\n",
      "End: Agent Testing #8 (226.67042875289917s)\n",
      "End: Agent Testing #2 (229.81895351409912s)\n",
      "End: Agent Testing #6 (230.70438957214355s)\n",
      "End: Agent Testing #7 (230.87767624855042s)\n",
      "End: Agent Testing #9 (231.96205615997314s)\n",
      "End: Agent Testing #5 (232.34873509407043s)\n",
      "End: Agent Testing #4 (232.5915403366089s)\n",
      "End: Agent Testing #1 (233.0155189037323s)\n",
      "End: Agent Testing #3 (234.0117859840393s)\n",
      "End: Agent Testing #0 (235.7787799835205s)\n",
      "End: Num of Offline Users: 14000 (236.4021759033203s)\n",
      "Start: Num of Offline Users: 15000\n",
      "Start: Agent Training #0\n",
      "Start: Agent Training #3\n",
      "Start: Agent Training #2\n",
      "Start: Agent Training #1\n",
      "Start: Agent Training #6\n",
      "Start: Agent Training #7\n",
      "Start: Agent Training #4\n",
      "Start: Agent Training #8\n",
      "Start: Agent Training #9\n",
      "Start: Agent Training #5\n",
      "Start: Agent Testing #8\n",
      "Start: Agent Testing #4\n",
      "Start: Agent Testing #2\n",
      "Start: Agent Testing #9\n",
      "Start: Agent Testing #6\n",
      "Start: Agent Testing #5\n",
      "Start: Agent Testing #7\n",
      "Start: Agent Testing #1\n",
      "Start: Agent Testing #3\n",
      "Start: Agent Testing #0\n",
      "End: Agent Testing #2 (233.6702311038971s)\n",
      "End: Agent Testing #8 (235.06134009361267s)\n",
      "End: Agent Testing #4 (235.46665859222412s)\n",
      "End: Agent Testing #5 (236.0094735622406s)\n",
      "End: Agent Testing #7 (236.31099796295166s)\n",
      "End: Agent Testing #9 (236.52324223518372s)\n",
      "End: Agent Testing #0 (236.54944777488708s)\n",
      "End: Agent Testing #1 (236.6013844013214s)\n",
      "End: Agent Testing #6 (236.73209285736084s)\n",
      "End: Agent Testing #3 (239.92768502235413s)\n",
      "End: Num of Offline Users: 15000 (240.5820505619049s)\n",
      "Agent: LogReg Poly\n",
      "Start: Num of Offline Users: 100\n",
      "Start: Agent Training #1\n",
      "Start: Agent Training #3\n",
      "Start: Agent Training #0\n",
      "Start: Agent Training #8\n",
      "Start: Agent Training #9\n",
      "Start: Agent Training #6\n",
      "Start: Agent Training #4\n",
      "Start: Agent Training #2\n",
      "Start: Agent Training #5\n",
      "Start: Agent Training #7\n",
      "Start: Agent Testing #4\n",
      "Start: Agent Testing #2\n",
      "Start: Agent Testing #8\n",
      "Start: Agent Testing #3\n",
      "Start: Agent Testing #9\n",
      "Start: Agent Testing #5\n",
      "Start: Agent Testing #0\n",
      "Start: Agent Testing #7\n",
      "Start: Agent Testing #1\n",
      "Start: Agent Testing #6\n",
      "End: Agent Testing #2 (1882.031091451645s)\n",
      "End: Agent Testing #3 (1891.061695098877s)\n",
      "End: Agent Testing #8 (1897.865297794342s)\n",
      "End: Agent Testing #0 (1899.4052293300629s)\n",
      "End: Agent Testing #6 (1900.1340644359589s)\n",
      "End: Agent Testing #7 (1910.9622151851654s)\n",
      "End: Agent Testing #9 (1924.3815376758575s)\n",
      "End: Agent Testing #1 (1925.7388424873352s)\n",
      "End: Agent Testing #5 (1951.08443069458s)\n",
      "End: Agent Testing #4 (2121.1789252758026s)\n",
      "End: Num of Offline Users: 100 (2121.3673825263977s)\n",
      "Start: Num of Offline Users: 500\n",
      "Start: Agent Training #0\n",
      "Start: Agent Training #1\n",
      "Start: Agent Training #2\n",
      "Start: Agent Training #5\n",
      "Start: Agent Training #8\n",
      "Start: Agent Training #3\n",
      "Start: Agent Training #4\n",
      "Start: Agent Training #7\n",
      "Start: Agent Training #6\n",
      "Start: Agent Training #9\n",
      "Start: Agent Testing #2\n",
      "Start: Agent Testing #5\n",
      "Start: Agent Testing #8\n",
      "Start: Agent Testing #4\n",
      "Start: Agent Testing #9\n",
      "Start: Agent Testing #6\n",
      "Start: Agent Testing #1\n",
      "Start: Agent Testing #3\n",
      "Start: Agent Testing #0\n",
      "Start: Agent Testing #7\n",
      "End: Agent Testing #8 (1952.4591872692108s)\n",
      "End: Agent Testing #9 (1991.674465417862s)\n",
      "End: Agent Testing #5 (1996.0949895381927s)\n",
      "End: Agent Testing #3 (2002.3544654846191s)\n",
      "End: Agent Testing #4 (2005.1301791667938s)\n",
      "End: Agent Testing #7 (2008.0394084453583s)\n",
      "End: Agent Testing #1 (2013.6898565292358s)\n",
      "End: Agent Testing #6 (2018.0395152568817s)\n",
      "End: Agent Testing #0 (2021.9099016189575s)\n",
      "End: Agent Testing #2 (2026.5971326828003s)\n",
      "End: Num of Offline Users: 500 (2026.7800390720367s)\n",
      "Start: Num of Offline Users: 1000\n",
      "Start: Agent Training #9\n",
      "Start: Agent Training #7\n",
      "Start: Agent Training #2\n",
      "Start: Agent Training #4\n",
      "Start: Agent Training #8\n",
      "Start: Agent Training #3\n",
      "Start: Agent Training #5\n",
      "Start: Agent Training #1\n",
      "Start: Agent Training #6\n",
      "Start: Agent Training #0\n",
      "Start: Agent Testing #5\n",
      "Start: Agent Testing #2\n",
      "Start: Agent Testing #3\n",
      "Start: Agent Testing #0\n",
      "Start: Agent Testing #6\n",
      "Start: Agent Testing #9\n",
      "Start: Agent Testing #8\n",
      "Start: Agent Testing #1\n",
      "Start: Agent Testing #4\n",
      "Start: Agent Testing #7\n",
      "End: Agent Testing #8 (2014.3710491657257s)\n",
      "End: Agent Testing #7 (2035.2643628120422s)\n",
      "End: Agent Testing #0 (2039.2826232910156s)\n",
      "End: Agent Testing #6 (2039.5954439640045s)\n",
      "End: Agent Testing #2 (2046.4049036502838s)\n",
      "End: Agent Testing #3 (2053.476474761963s)\n",
      "End: Agent Testing #4 (2056.638435125351s)\n",
      "End: Agent Testing #5 (2106.183948993683s)\n",
      "End: Agent Testing #9 (2108.375521659851s)\n",
      "End: Agent Testing #1 (2371.455108165741s)\n",
      "End: Num of Offline Users: 1000 (2371.7118785381317s)\n",
      "Start: Num of Offline Users: 2000\n",
      "Start: Agent Training #2\n",
      "Start: Agent Training #0\n",
      "Start: Agent Training #1\n",
      "Start: Agent Training #3\n",
      "Start: Agent Training #4\n",
      "Start: Agent Training #6\n",
      "Start: Agent Training #9\n",
      "Start: Agent Training #8\n",
      "Start: Agent Training #5\n",
      "Start: Agent Training #7\n",
      "Start: Agent Testing #2\n",
      "Start: Agent Testing #0\n",
      "Start: Agent Testing #3\n",
      "Start: Agent Testing #8\n",
      "Start: Agent Testing #5\n",
      "Start: Agent Testing #4\n",
      "Start: Agent Testing #6\n",
      "Start: Agent Testing #1\n",
      "Start: Agent Testing #9\n",
      "Start: Agent Testing #7\n",
      "End: Agent Testing #8 (2142.999589920044s)\n",
      "End: Agent Testing #7 (2160.908604621887s)\n",
      "End: Agent Testing #6 (2163.6594507694244s)\n",
      "End: Agent Testing #1 (2176.5968775749207s)\n",
      "End: Agent Testing #4 (2178.7974886894226s)\n",
      "End: Agent Testing #0 (2187.7216141223907s)\n",
      "End: Agent Testing #2 (2193.109820127487s)\n",
      "End: Agent Testing #5 (2213.068551301956s)\n",
      "End: Agent Testing #9 (2229.143986225128s)\n",
      "End: Agent Testing #3 (2239.0544481277466s)\n",
      "End: Num of Offline Users: 2000 (2239.32678771019s)\n",
      "Start: Num of Offline Users: 3000\n",
      "Start: Agent Training #4\n",
      "Start: Agent Training #0\n",
      "Start: Agent Training #8\n",
      "Start: Agent Training #2\n",
      "Start: Agent Training #9\n",
      "Start: Agent Training #7\n",
      "Start: Agent Training #1\n",
      "Start: Agent Training #3\n",
      "Start: Agent Training #6\n",
      "Start: Agent Training #5\n",
      "Start: Agent Testing #8\n",
      "Start: Agent Testing #0\n",
      "Start: Agent Testing #3\n",
      "Start: Agent Testing #2\n",
      "Start: Agent Testing #1\n",
      "Start: Agent Testing #7\n",
      "Start: Agent Testing #6\n",
      "Start: Agent Testing #9\n",
      "Start: Agent Testing #5\n",
      "Start: Agent Testing #4\n",
      "End: Agent Testing #3 (2192.8923659324646s)\n",
      "End: Agent Testing #2 (2193.9980370998383s)\n",
      "End: Agent Testing #1 (2207.4448137283325s)\n",
      "End: Agent Testing #7 (2221.1669597625732s)\n",
      "End: Agent Testing #8 (2222.7275054454803s)\n",
      "End: Agent Testing #4 (2249.461811065674s)\n",
      "End: Agent Testing #0 (2267.6760606765747s)\n",
      "End: Agent Testing #6 (2318.0835196971893s)\n",
      "End: Agent Testing #9 (2322.3488075733185s)\n",
      "End: Agent Testing #5 (2371.4642162323s)\n",
      "End: Num of Offline Users: 3000 (2371.7111036777496s)\n",
      "Start: Num of Offline Users: 5000\n",
      "Start: Agent Training #0\n",
      "Start: Agent Training #5\n",
      "Start: Agent Training #6\n",
      "Start: Agent Training #1\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Start: Agent Training #4\n",
      "Start: Agent Training #8\n",
      "Start: Agent Training #2\n",
      "Start: Agent Training #7\n",
      "Start: Agent Training #3\n",
      "Start: Agent Training #9\n",
      "Start: Agent Testing #3\n",
      "Start: Agent Testing #0\n",
      "Start: Agent Testing #2\n",
      "Start: Agent Testing #8\n",
      "Start: Agent Testing #4\n",
      "Start: Agent Testing #5\n",
      "Start: Agent Testing #1\n",
      "Start: Agent Testing #9\n",
      "Start: Agent Testing #7\n",
      "Start: Agent Testing #6\n",
      "End: Agent Testing #7 (2390.1742782592773s)\n",
      "End: Agent Testing #4 (2395.6259303092957s)\n",
      "End: Agent Testing #8 (2404.0798659324646s)\n",
      "End: Agent Testing #9 (2412.2140130996704s)\n",
      "End: Agent Testing #3 (2415.4738945961s)\n",
      "End: Agent Testing #6 (2421.3527805805206s)\n",
      "End: Agent Testing #0 (2463.6384150981903s)\n",
      "End: Agent Testing #2 (2469.19864320755s)\n",
      "End: Agent Testing #5 (2504.283156633377s)\n",
      "End: Agent Testing #1 (2511.0701310634613s)\n",
      "End: Num of Offline Users: 5000 (2511.4058351516724s)\n",
      "Start: Num of Offline Users: 8000\n",
      "Start: Agent Training #3\n",
      "Start: Agent Training #1\n",
      "Start: Agent Training #8\n",
      "Start: Agent Training #2\n",
      "Start: Agent Training #5\n",
      "Start: Agent Training #6\n",
      "Start: Agent Training #7\n",
      "Start: Agent Training #0\n",
      "Start: Agent Training #9\n",
      "Start: Agent Training #4\n",
      "Start: Agent Testing #9\n",
      "Start: Agent Testing #2\n",
      "Start: Agent Testing #3\n",
      "Start: Agent Testing #1\n",
      "Start: Agent Testing #8\n",
      "Start: Agent Testing #4\n",
      "Start: Agent Testing #5\n",
      "Start: Agent Testing #7\n",
      "Start: Agent Testing #0\n",
      "Start: Agent Testing #6\n",
      "End: Agent Testing #9 (2427.7156994342804s)\n",
      "End: Agent Testing #7 (2448.7035105228424s)\n",
      "End: Agent Testing #0 (2475.477529525757s)\n",
      "End: Agent Testing #2 (2493.9938366413116s)\n",
      "End: Agent Testing #1 (2495.3289070129395s)\n",
      "End: Agent Testing #5 (2512.365116596222s)\n",
      "End: Agent Testing #4 (2524.632965326309s)\n",
      "End: Agent Testing #6 (2535.5862288475037s)\n",
      "End: Agent Testing #3 (2560.790731906891s)\n",
      "End: Agent Testing #8 (2660.1423332691193s)\n",
      "End: Num of Offline Users: 8000 (2660.6299636363983s)\n",
      "Start: Num of Offline Users: 10000\n",
      "Start: Agent Training #4\n",
      "Start: Agent Training #9\n",
      "Start: Agent Training #3\n",
      "Start: Agent Training #6\n",
      "Start: Agent Training #1\n",
      "Start: Agent Training #5\n",
      "Start: Agent Training #2\n",
      "Start: Agent Training #7\n",
      "Start: Agent Training #0\n",
      "Start: Agent Training #8\n",
      "Start: Agent Testing #1\n",
      "Start: Agent Testing #0\n",
      "Start: Agent Testing #5\n",
      "Start: Agent Testing #4\n",
      "Start: Agent Testing #2\n",
      "Start: Agent Testing #3\n",
      "Start: Agent Testing #8\n",
      "Start: Agent Testing #9\n",
      "Start: Agent Testing #7\n",
      "Start: Agent Testing #6\n",
      "End: Agent Testing #1 (2281.002685070038s)\n",
      "End: Agent Testing #0 (2296.4338417053223s)\n",
      "End: Agent Testing #5 (2307.619724035263s)\n",
      "End: Agent Testing #3 (2311.9892926216125s)\n",
      "End: Agent Testing #2 (2321.4908154010773s)\n",
      "End: Agent Testing #7 (2337.9414904117584s)\n",
      "End: Agent Testing #8 (2343.647889614105s)\n",
      "End: Agent Testing #9 (2343.73930311203s)\n",
      "End: Agent Testing #6 (2386.6923081874847s)\n",
      "End: Agent Testing #4 (2454.5798723697662s)\n",
      "End: Num of Offline Users: 10000 (2455.1426272392273s)\n",
      "Start: Num of Offline Users: 13000\n",
      "Start: Agent Training #1\n",
      "Start: Agent Training #9\n",
      "Start: Agent Training #0\n",
      "Start: Agent Training #2\n",
      "Start: Agent Training #6\n",
      "Start: Agent Training #8\n",
      "Start: Agent Training #3\n",
      "Start: Agent Training #7\n",
      "Start: Agent Training #5\n",
      "Start: Agent Training #4\n",
      "Start: Agent Testing #8\n",
      "Start: Agent Testing #2\n",
      "Start: Agent Testing #5\n",
      "Start: Agent Testing #0\n",
      "Start: Agent Testing #1\n",
      "Start: Agent Testing #9\n",
      "Start: Agent Testing #3\n",
      "Start: Agent Testing #6\n",
      "Start: Agent Testing #7\n",
      "Start: Agent Testing #4\n",
      "End: Agent Testing #2 (2381.8222353458405s)\n",
      "End: Agent Testing #1 (2393.305884361267s)\n",
      "End: Agent Testing #9 (2400.7532567977905s)\n",
      "End: Agent Testing #5 (2403.5055859088898s)\n",
      "End: Agent Testing #0 (2407.5946519374847s)\n",
      "End: Agent Testing #8 (2409.1234908103943s)\n",
      "End: Agent Testing #7 (2410.0158867836s)\n",
      "End: Agent Testing #4 (2437.497500181198s)\n",
      "End: Agent Testing #3 (2479.2976536750793s)\n",
      "End: Agent Testing #6 (2536.5243649482727s)\n",
      "End: Num of Offline Users: 13000 (2537.1661150455475s)\n",
      "Start: Num of Offline Users: 14000\n",
      "Start: Agent Training #4\n",
      "Start: Agent Training #2\n",
      "Start: Agent Training #3\n",
      "Start: Agent Training #1\n",
      "Start: Agent Training #6\n",
      "Start: Agent Training #7\n",
      "Start: Agent Training #5\n",
      "Start: Agent Training #0\n",
      "Start: Agent Training #8\n",
      "Start: Agent Training #9\n",
      "Start: Agent Testing #8\n",
      "Start: Agent Testing #4\n",
      "Start: Agent Testing #5\n",
      "Start: Agent Testing #2\n",
      "Start: Agent Testing #3\n",
      "Start: Agent Testing #9\n",
      "Start: Agent Testing #7\n",
      "Start: Agent Testing #0\n",
      "Start: Agent Testing #6\n",
      "Start: Agent Testing #1\n",
      "End: Agent Testing #5 (2409.016356945038s)\n",
      "End: Agent Testing #6 (2410.3050808906555s)\n",
      "End: Agent Testing #8 (2420.1634714603424s)\n",
      "End: Agent Testing #7 (2423.79527425766s)\n",
      "End: Agent Testing #0 (2444.445494890213s)\n",
      "End: Agent Testing #4 (2451.8483278751373s)\n",
      "End: Agent Testing #3 (2453.7571926116943s)\n",
      "End: Agent Testing #1 (2474.422869682312s)\n",
      "End: Agent Testing #2 (2488.09029006958s)\n",
      "End: Agent Testing #9 (2603.9986975193024s)\n",
      "End: Num of Offline Users: 14000 (2604.6487367153168s)\n",
      "Start: Num of Offline Users: 15000\n",
      "Start: Agent Training #0\n",
      "Start: Agent Training #9\n",
      "Start: Agent Training #6\n",
      "Start: Agent Training #5\n",
      "Start: Agent Training #1\n",
      "Start: Agent Training #7\n",
      "Start: Agent Training #4\n",
      "Start: Agent Training #8\n",
      "Start: Agent Training #2\n",
      "Start: Agent Training #3\n",
      "Start: Agent Testing #8\n",
      "Start: Agent Testing #4\n",
      "Start: Agent Testing #2\n",
      "Start: Agent Testing #1\n",
      "Start: Agent Testing #7\n",
      "Start: Agent Testing #9\n",
      "Start: Agent Testing #6\n",
      "Start: Agent Testing #3\n",
      "Start: Agent Testing #5\n",
      "Start: Agent Testing #0\n",
      "End: Agent Testing #2 (2438.3086552619934s)\n",
      "End: Agent Testing #0 (2458.8984081745148s)\n",
      "End: Agent Testing #1 (2466.939116001129s)\n",
      "End: Agent Testing #6 (2469.3080229759216s)\n",
      "End: Agent Testing #4 (2473.0328390598297s)\n",
      "End: Agent Testing #7 (2492.803640604019s)\n",
      "End: Agent Testing #9 (2514.1078536510468s)\n",
      "End: Agent Testing #3 (2525.9673380851746s)\n",
      "End: Agent Testing #8 (2538.895731449127s)\n",
      "End: Agent Testing #5 (2570.4917330741882s)\n",
      "End: Num of Offline Users: 15000 (2571.213179588318s)\n"
     ]
    }
   ],
   "source": [
    "agent_stats01 = gather_agent_stats(\n",
    "    env,\n",
    "    std_env_args,\n",
    "    {\n",
    "        'num_products': 10,\n",
    "        'number_of_flips': 5,\n",
    "    },\n",
    "    agent_inits,\n",
    "    TrainingDataSamples,\n",
    "    TestingDataSamples,\n",
    "    StatEpochs,\n",
    "    StatEpochsNewRandomSeed\n",
    ")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "pycharm": {}
   },
   "source": [
    "Plotting results..."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "pycharm": {},
    "scrolled": false
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1152x576 with 1 Axes>"
      ]
     },
     "metadata": {
      "image/png": {
       "height": 481,
       "width": 958
      },
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plot_agent_stats(agent_stats01)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "pycharm": {}
   },
   "source": [
    "# _Without_ Product Flips\n",
    "\n",
    "This time, without any _`Product`_ flips."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true,
    "pycharm": {}
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Agent: User Organic\n",
      "Start: Num of Offline Users: 100\n",
      "Start: Agent Training #1\n",
      "Start: Agent Training #0\n",
      "Start: Agent Training #6\n",
      "Start: Agent Training #4\n",
      "Start: Agent Training #2\n",
      "Start: Agent Training #5\n",
      "Start: Agent Training #7\n",
      "Start: Agent Training #8\n",
      "Start: Agent Training #9\n",
      "Start: Agent Training #3\n",
      "Start: Agent Testing #4\n",
      "Start: Agent Testing #8\n",
      "Start: Agent Testing #9\n",
      "Start: Agent Testing #3\n",
      "Start: Agent Testing #2\n",
      "Start: Agent Testing #5\n",
      "Start: Agent Testing #0\n",
      "Start: Agent Testing #7\n",
      "Start: Agent Testing #1\n",
      "Start: Agent Testing #6\n",
      "End: Agent Testing #8 (120.56862950325012s)\n",
      "End: Agent Testing #0 (120.7930679321289s)\n",
      "End: Agent Testing #2 (121.80716681480408s)\n",
      "End: Agent Testing #1 (121.96314716339111s)\n",
      "End: Agent Testing #6 (122.17742705345154s)\n",
      "End: Agent Testing #5 (123.80630397796631s)\n",
      "End: Agent Testing #3 (123.99241280555725s)\n",
      "End: Agent Testing #4 (124.81978368759155s)\n",
      "End: Agent Testing #9 (126.49068665504456s)\n",
      "End: Agent Testing #7 (127.25687646865845s)\n",
      "End: Num of Offline Users: 100 (127.48194742202759s)\n",
      "Start: Num of Offline Users: 500\n",
      "Start: Agent Training #0\n",
      "Start: Agent Training #2\n",
      "Start: Agent Training #3\n",
      "Start: Agent Training #4\n",
      "Start: Agent Training #5\n",
      "Start: Agent Training #9\n",
      "Start: Agent Training #8\n",
      "Start: Agent Training #6\n",
      "Start: Agent Training #7\n",
      "Start: Agent Training #1\n",
      "Start: Agent Testing #2\n",
      "Start: Agent Testing #3\n",
      "Start: Agent Testing #8\n",
      "Start: Agent Testing #6\n",
      "Start: Agent Testing #1\n",
      "Start: Agent Testing #4\n",
      "Start: Agent Testing #5\n",
      "Start: Agent Testing #7\n",
      "Start: Agent Testing #9\n",
      "Start: Agent Testing #0\n",
      "End: Agent Testing #4 (125.93728590011597s)\n",
      "End: Agent Testing #5 (126.06220245361328s)\n",
      "End: Agent Testing #7 (126.14840483665466s)\n",
      "End: Agent Testing #1 (126.65745282173157s)\n",
      "End: Agent Testing #6 (127.32791900634766s)\n",
      "End: Agent Testing #3 (127.92899560928345s)\n",
      "End: Agent Testing #2 (128.21523094177246s)\n",
      "End: Agent Testing #8 (128.21734309196472s)\n",
      "End: Agent Testing #9 (130.21062016487122s)\n",
      "End: Agent Testing #0 (130.47900938987732s)\n",
      "End: Num of Offline Users: 500 (130.69138479232788s)\n",
      "Start: Num of Offline Users: 1000\n",
      "Start: Agent Training #0\n",
      "Start: Agent Training #2\n",
      "Start: Agent Training #1\n",
      "Start: Agent Training #3\n",
      "Start: Agent Training #4\n",
      "Start: Agent Training #6\n",
      "Start: Agent Training #5\n",
      "Start: Agent Training #7\n",
      "Start: Agent Training #8\n",
      "Start: Agent Training #9\n",
      "Start: Agent Testing #3\n",
      "Start: Agent Testing #5\n",
      "Start: Agent Testing #2\n",
      "Start: Agent Testing #1\n",
      "Start: Agent Testing #6\n",
      "Start: Agent Testing #0\n",
      "Start: Agent Testing #8\n",
      "Start: Agent Testing #9\n",
      "Start: Agent Testing #7\n",
      "Start: Agent Testing #4\n",
      "End: Agent Testing #2 (127.52003359794617s)\n",
      "End: Agent Testing #3 (128.3723111152649s)\n",
      "End: Agent Testing #5 (128.92526078224182s)\n",
      "End: Agent Testing #0 (129.6938076019287s)\n",
      "End: Agent Testing #9 (129.8172106742859s)\n",
      "End: Agent Testing #6 (130.67796754837036s)\n",
      "End: Agent Testing #1 (130.800639629364s)\n",
      "End: Agent Testing #4 (130.8219027519226s)\n",
      "End: Agent Testing #7 (130.8873462677002s)\n",
      "End: Agent Testing #8 (131.2753872871399s)\n",
      "End: Num of Offline Users: 1000 (131.51917171478271s)\n",
      "Start: Num of Offline Users: 2000\n",
      "Start: Agent Training #1\n",
      "Start: Agent Training #2\n",
      "Start: Agent Training #0\n",
      "Start: Agent Training #3\n",
      "Start: Agent Training #5\n",
      "Start: Agent Training #6\n",
      "Start: Agent Training #4\n",
      "Start: Agent Training #8\n",
      "Start: Agent Training #9\n",
      "Start: Agent Training #7\n",
      "Start: Agent Testing #2\n",
      "Start: Agent Testing #5\n",
      "Start: Agent Testing #0\n",
      "Start: Agent Testing #3\n",
      "Start: Agent Testing #8\n",
      "Start: Agent Testing #1\n",
      "Start: Agent Testing #6\n",
      "Start: Agent Testing #4\n",
      "Start: Agent Testing #9\n",
      "Start: Agent Testing #7\n",
      "End: Agent Testing #2 (134.1877522468567s)\n",
      "End: Agent Testing #3 (135.36970019340515s)\n",
      "End: Agent Testing #8 (135.604159116745s)\n",
      "End: Agent Testing #9 (136.21337699890137s)\n",
      "End: Agent Testing #0 (136.64241242408752s)\n",
      "End: Agent Testing #1 (136.75382947921753s)\n",
      "End: Agent Testing #7 (136.86580061912537s)\n",
      "End: Agent Testing #4 (137.2482852935791s)\n",
      "End: Agent Testing #6 (137.53022813796997s)\n",
      "End: Agent Testing #5 (138.19259214401245s)\n",
      "End: Num of Offline Users: 2000 (138.48669052124023s)\n",
      "Start: Num of Offline Users: 3000\n",
      "Start: Agent Training #2\n",
      "Start: Agent Training #1\n",
      "Start: Agent Training #0\n",
      "Start: Agent Training #3\n",
      "Start: Agent Training #4\n",
      "Start: Agent Training #5\n",
      "Start: Agent Training #9\n",
      "Start: Agent Training #8\n",
      "Start: Agent Training #6\n",
      "Start: Agent Training #7\n",
      "Start: Agent Testing #3\n",
      "Start: Agent Testing #2\n",
      "Start: Agent Testing #8\n",
      "Start: Agent Testing #0\n",
      "Start: Agent Testing #7\n",
      "Start: Agent Testing #1\n",
      "Start: Agent Testing #5\n",
      "Start: Agent Testing #9\n",
      "Start: Agent Testing #4\n",
      "Start: Agent Testing #6\n",
      "End: Agent Testing #0 (143.9300467967987s)\n",
      "End: Agent Testing #7 (144.79538893699646s)\n",
      "End: Agent Testing #2 (145.1324863433838s)\n",
      "End: Agent Testing #3 (146.05884909629822s)\n",
      "End: Agent Testing #9 (146.7557430267334s)\n",
      "End: Agent Testing #8 (147.30620861053467s)\n",
      "End: Agent Testing #5 (147.78487348556519s)\n",
      "End: Agent Testing #1 (148.25524377822876s)\n",
      "End: Agent Testing #4 (148.37989258766174s)\n",
      "End: Agent Testing #6 (149.2246265411377s)\n",
      "End: Num of Offline Users: 3000 (149.57694244384766s)\n",
      "Start: Num of Offline Users: 5000\n",
      "Start: Agent Training #0\n",
      "Start: Agent Training #2\n",
      "Start: Agent Training #1\n",
      "Start: Agent Training #3\n",
      "Start: Agent Training #4\n",
      "Start: Agent Training #5\n",
      "Start: Agent Training #6\n",
      "Start: Agent Training #8\n",
      "Start: Agent Training #7\n",
      "Start: Agent Training #9\n",
      "Start: Agent Testing #3\n",
      "Start: Agent Testing #8\n",
      "Start: Agent Testing #5\n",
      "Start: Agent Testing #0\n",
      "Start: Agent Testing #1\n",
      "Start: Agent Testing #2\n",
      "Start: Agent Testing #4\n",
      "Start: Agent Testing #7\n",
      "Start: Agent Testing #9\n",
      "Start: Agent Testing #6\n",
      "End: Agent Testing #2 (157.13761234283447s)\n",
      "End: Agent Testing #8 (157.78914093971252s)\n",
      "End: Agent Testing #4 (157.91233444213867s)\n",
      "End: Agent Testing #3 (158.10847854614258s)\n",
      "End: Agent Testing #1 (158.30453872680664s)\n",
      "End: Agent Testing #7 (158.36190128326416s)\n",
      "End: Agent Testing #5 (158.7505202293396s)\n",
      "End: Agent Testing #9 (159.14403867721558s)\n",
      "End: Agent Testing #0 (160.6490330696106s)\n",
      "End: Agent Testing #6 (161.06949520111084s)\n",
      "End: Num of Offline Users: 5000 (161.3822376728058s)\n",
      "Start: Num of Offline Users: 8000\n",
      "Start: Agent Training #0\n",
      "Start: Agent Training #1\n",
      "Start: Agent Training #2\n",
      "Start: Agent Training #3\n",
      "Start: Agent Training #5\n",
      "Start: Agent Training #6\n",
      "Start: Agent Training #4\n",
      "Start: Agent Training #7\n",
      "Start: Agent Training #8\n",
      "Start: Agent Training #9\n",
      "Start: Agent Testing #8\n",
      "Start: Agent Testing #6\n",
      "Start: Agent Testing #1\n",
      "Start: Agent Testing #2\n",
      "Start: Agent Testing #0\n",
      "Start: Agent Testing #3\n",
      "Start: Agent Testing #7\n",
      "Start: Agent Testing #5\n",
      "Start: Agent Testing #4\n",
      "Start: Agent Testing #9\n",
      "End: Agent Testing #9 (184.8623354434967s)\n",
      "End: Agent Testing #0 (186.24745321273804s)\n",
      "End: Agent Testing #7 (186.563711643219s)\n",
      "End: Agent Testing #8 (186.720356464386s)\n",
      "End: Agent Testing #3 (186.7642993927002s)\n",
      "End: Agent Testing #1 (188.02420377731323s)\n",
      "End: Agent Testing #4 (188.08228659629822s)\n",
      "End: Agent Testing #6 (188.19826316833496s)\n",
      "End: Agent Testing #2 (188.61295986175537s)\n",
      "End: Agent Testing #5 (189.73710656166077s)\n",
      "End: Num of Offline Users: 8000 (190.2125060558319s)\n",
      "Start: Num of Offline Users: 10000\n",
      "Start: Agent Training #2\n",
      "Start: Agent Training #5\n",
      "Start: Agent Training #1\n",
      "Start: Agent Training #6\n",
      "Start: Agent Training #7\n",
      "Start: Agent Training #0\n",
      "Start: Agent Training #4\n",
      "Start: Agent Training #3\n",
      "Start: Agent Training #8\n",
      "Start: Agent Training #9\n",
      "Start: Agent Testing #8\n",
      "Start: Agent Testing #3\n",
      "Start: Agent Testing #5\n",
      "Start: Agent Testing #9\n",
      "Start: Agent Testing #2\n",
      "Start: Agent Testing #1\n",
      "Start: Agent Testing #0\n",
      "Start: Agent Testing #6\n",
      "Start: Agent Testing #4\n",
      "Start: Agent Testing #7\n",
      "End: Agent Testing #8 (197.71193051338196s)\n",
      "End: Agent Testing #9 (198.25532937049866s)\n",
      "End: Agent Testing #2 (199.49465322494507s)\n",
      "End: Agent Testing #5 (199.53775024414062s)\n",
      "End: Agent Testing #6 (199.8428659439087s)\n",
      "End: Agent Testing #0 (200.48362803459167s)\n",
      "End: Agent Testing #7 (200.9443974494934s)\n",
      "End: Agent Testing #1 (201.21114540100098s)\n",
      "End: Agent Testing #4 (201.6709339618683s)\n",
      "End: Agent Testing #3 (203.96210384368896s)\n",
      "End: Num of Offline Users: 10000 (204.48380374908447s)\n",
      "Start: Num of Offline Users: 13000\n",
      "Start: Agent Training #0\n",
      "Start: Agent Training #3\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Start: Agent Training #1\n",
      "Start: Agent Training #2\n",
      "Start: Agent Training #6\n",
      "Start: Agent Training #5\n",
      "Start: Agent Training #9\n",
      "Start: Agent Training #7\n",
      "Start: Agent Training #8\n",
      "Start: Agent Training #4\n",
      "Start: Agent Testing #8\n",
      "Start: Agent Testing #4\n",
      "Start: Agent Testing #1\n",
      "Start: Agent Testing #5\n",
      "Start: Agent Testing #3\n",
      "Start: Agent Testing #9\n",
      "Start: Agent Testing #2\n",
      "Start: Agent Testing #7\n",
      "Start: Agent Testing #0\n",
      "Start: Agent Testing #6\n",
      "End: Agent Testing #8 (216.92610597610474s)\n",
      "End: Agent Testing #9 (216.97651863098145s)\n",
      "End: Agent Testing #4 (219.69095301628113s)\n",
      "End: Agent Testing #7 (220.10654497146606s)\n",
      "End: Agent Testing #0 (220.26104044914246s)\n",
      "End: Agent Testing #6 (221.02877235412598s)\n",
      "End: Agent Testing #3 (221.32490801811218s)\n",
      "End: Agent Testing #5 (221.4810860157013s)\n",
      "End: Agent Testing #1 (221.69982028007507s)\n",
      "End: Agent Testing #2 (222.46597337722778s)\n",
      "End: Num of Offline Users: 13000 (223.1223394870758s)\n",
      "Start: Num of Offline Users: 14000\n",
      "Start: Agent Training #1\n",
      "Start: Agent Training #0\n",
      "Start: Agent Training #2\n",
      "Start: Agent Training #4\n",
      "Start: Agent Training #3\n",
      "Start: Agent Training #8\n",
      "Start: Agent Training #5\n",
      "Start: Agent Training #7\n",
      "Start: Agent Training #9\n",
      "Start: Agent Training #6\n",
      "Start: Agent Testing #4\n",
      "Start: Agent Testing #8\n",
      "Start: Agent Testing #1\n",
      "Start: Agent Testing #9\n",
      "Start: Agent Testing #5\n",
      "Start: Agent Testing #7\n",
      "Start: Agent Testing #2\n",
      "Start: Agent Testing #3\n",
      "Start: Agent Testing #0\n",
      "Start: Agent Testing #6\n",
      "End: Agent Testing #9 (232.01247191429138s)\n",
      "End: Agent Testing #8 (232.40218043327332s)\n",
      "End: Agent Testing #4 (232.58170747756958s)\n",
      "End: Agent Testing #2 (232.98685312271118s)\n",
      "End: Agent Testing #1 (233.19117140769958s)\n",
      "End: Agent Testing #6 (233.19172716140747s)\n",
      "End: Agent Testing #5 (233.6593918800354s)\n",
      "End: Agent Testing #0 (233.9578890800476s)\n",
      "End: Agent Testing #3 (235.5008864402771s)\n",
      "End: Agent Testing #7 (235.98341989517212s)\n",
      "End: Num of Offline Users: 14000 (236.64559650421143s)\n",
      "Start: Num of Offline Users: 15000\n",
      "Start: Agent Training #0\n",
      "Start: Agent Training #1\n",
      "Start: Agent Training #9\n",
      "Start: Agent Training #7\n",
      "Start: Agent Training #6\n",
      "Start: Agent Training #5\n",
      "Start: Agent Training #3\n",
      "Start: Agent Training #2\n",
      "Start: Agent Training #4\n",
      "Start: Agent Training #8\n",
      "Start: Agent Testing #8\n",
      "Start: Agent Testing #4\n",
      "Start: Agent Testing #7\n",
      "Start: Agent Testing #0\n",
      "Start: Agent Testing #9\n",
      "Start: Agent Testing #1\n",
      "Start: Agent Testing #2\n",
      "Start: Agent Testing #5\n",
      "Start: Agent Testing #3\n",
      "Start: Agent Testing #6\n",
      "End: Agent Testing #8 (230.5016450881958s)\n",
      "End: Agent Testing #7 (232.94115471839905s)\n",
      "End: Agent Testing #2 (233.3243489265442s)\n",
      "End: Agent Testing #9 (234.0310456752777s)\n",
      "End: Agent Testing #5 (234.9141252040863s)\n",
      "End: Agent Testing #4 (235.22772789001465s)\n",
      "End: Agent Testing #0 (235.30178093910217s)\n",
      "End: Agent Testing #6 (236.426598072052s)\n",
      "End: Agent Testing #1 (236.75095319747925s)\n",
      "End: Agent Testing #3 (237.2825210094452s)\n",
      "End: Num of Offline Users: 15000 (238.0306544303894s)\n",
      "Agent: LogReg Poly\n",
      "Start: Num of Offline Users: 100\n",
      "Start: Agent Training #2\n",
      "Start: Agent Training #4\n",
      "Start: Agent Training #1\n",
      "Start: Agent Training #5\n",
      "Start: Agent Training #8\n",
      "Start: Agent Training #3\n",
      "Start: Agent Training #0\n",
      "Start: Agent Training #6\n",
      "Start: Agent Training #7\n",
      "Start: Agent Training #9\n",
      "Start: Agent Testing #4\n",
      "Start: Agent Testing #8\n",
      "Start: Agent Testing #2\n",
      "Start: Agent Testing #5\n",
      "Start: Agent Testing #3\n",
      "Start: Agent Testing #0\n",
      "Start: Agent Testing #9\n",
      "Start: Agent Testing #7\n",
      "Start: Agent Testing #1\n",
      "Start: Agent Testing #6\n",
      "End: Agent Testing #2 (1905.7672550678253s)\n",
      "End: Agent Testing #3 (1908.675678730011s)\n",
      "End: Agent Testing #9 (1911.5746819972992s)\n",
      "End: Agent Testing #4 (1911.9238619804382s)\n",
      "End: Agent Testing #7 (1919.1231608390808s)\n",
      "End: Agent Testing #5 (1920.191602230072s)\n",
      "End: Agent Testing #6 (1925.483838558197s)\n",
      "End: Agent Testing #8 (1936.3428599834442s)\n",
      "End: Agent Testing #0 (1953.9585921764374s)\n",
      "End: Agent Testing #1 (1992.4614322185516s)\n",
      "End: Num of Offline Users: 100 (1992.7045271396637s)\n",
      "Start: Num of Offline Users: 500\n",
      "Start: Agent Training #7\n",
      "Start: Agent Training #3\n",
      "Start: Agent Training #6\n",
      "Start: Agent Training #4\n",
      "Start: Agent Training #5\n",
      "Start: Agent Training #0\n",
      "Start: Agent Training #1\n",
      "Start: Agent Training #8\n",
      "Start: Agent Training #2\n",
      "Start: Agent Training #9\n",
      "Start: Agent Testing #4\n",
      "Start: Agent Testing #5\n",
      "Start: Agent Testing #8\n",
      "Start: Agent Testing #3\n",
      "Start: Agent Testing #9\n",
      "Start: Agent Testing #1\n",
      "Start: Agent Testing #2\n",
      "Start: Agent Testing #6\n",
      "Start: Agent Testing #0\n",
      "Start: Agent Testing #7\n",
      "End: Agent Testing #8 (1956.799228668213s)\n",
      "End: Agent Testing #3 (1965.3080990314484s)\n",
      "End: Agent Testing #5 (1966.058173418045s)\n",
      "End: Agent Testing #0 (1972.120177745819s)\n",
      "End: Agent Testing #4 (1982.010862827301s)\n",
      "End: Agent Testing #9 (1986.5687255859375s)\n",
      "End: Agent Testing #7 (1999.3858819007874s)\n",
      "End: Agent Testing #6 (2014.9334433078766s)\n",
      "End: Agent Testing #1 (2096.4889085292816s)\n",
      "End: Agent Testing #2 (2206.393836736679s)\n",
      "End: Num of Offline Users: 500 (2206.6139237880707s)\n",
      "Start: Num of Offline Users: 1000\n",
      "Start: Agent Training #0\n",
      "Start: Agent Training #1\n",
      "Start: Agent Training #9\n",
      "Start: Agent Training #3\n",
      "Start: Agent Training #6\n",
      "Start: Agent Training #2\n",
      "Start: Agent Training #7\n",
      "Start: Agent Training #8\n",
      "Start: Agent Training #4\n",
      "Start: Agent Training #5\n",
      "Start: Agent Testing #3\n",
      "Start: Agent Testing #4\n",
      "Start: Agent Testing #8\n",
      "Start: Agent Testing #1\n",
      "Start: Agent Testing #9\n",
      "Start: Agent Testing #5\n",
      "Start: Agent Testing #0\n",
      "Start: Agent Testing #6\n",
      "Start: Agent Testing #2\n",
      "Start: Agent Testing #7\n"
     ]
    }
   ],
   "source": [
    "agent_stats02 = gather_agent_stats(\n",
    "    env,\n",
    "    std_env_args,\n",
    "    {\n",
    "        'num_products': 10,\n",
    "        'number_of_flips': 0,\n",
    "    },\n",
    "    agent_inits,\n",
    "    TrainingDataSamples,\n",
    "    TestingDataSamples,\n",
    "    StatEpochs,\n",
    "    StatEpochsNewRandomSeed\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {
    "pycharm": {},
    "scrolled": false
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1152x576 with 1 Axes>"
      ]
     },
     "metadata": {
      "image/png": {
       "height": 481,
       "width": 964
      },
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plot_agent_stats(agent_stats02)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "pycharm": {}
   },
   "source": [
    "# Resolution"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "pycharm": {}
   },
   "source": [
    "_Logistic Regression_ with a polynomial feature set\n",
    "* _**does**_ outperform _User Organic Counter_ _**when**_ there are some _`Product`_ flips in the _`Environment`_, i.e. the probability of the click has _**no**_ direct relation to the amount of viewed _`Products`_;\n",
    "* and, it has almost the same _Click Through Rate_ _**when**_ there is no _`Product`_ flips."
   ]
  }
 ],
 "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": "3.6.3"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}