Value_Function_Approximation_Compendio.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "provenance": [],
      "collapsed_sections": [
        "o11bM0eqbvYC",
        "csd6dOJNHEQt",
        "K6I7WJwc1vMw",
        "cZjc5fAz8ab_"
      ]
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    },
    "language_info": {
      "name": "python"
    }
  },
  "cells": [
    {
      "cell_type": "markdown",
      "source": [
        "#Value Function Approximation - Compendio\n",
        "\n",
        "Mario Fiorino"
      ],
      "metadata": {
        "id": "UUKWWcG9Jcg_"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "**Limiti dei metodi RL tabulari (look-up table):**\n",
        "\n",
        "i requisiti di memoria per archiviare in una tabelle spazi di stato/azione molto grandi, oggi sembrano irrealizzabili; inoltre, visitare adeguatamente ciascuna coppia stato-azione richiede un tempo enorme\n",
        "\n",
        "\n",
        "**Idea:**\n",
        "\n",
        "superare tali limitazioni con  tecniche di approssimazione di funzioni. In pratica : si utilizza una funzione (ad esempio le reti neural) che approssima la value-fuction di ogni stato."
      ],
      "metadata": {
        "id": "jBPWH_VoJp2N"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "## Nozioni principali\n",
        "\n",
        "**Modellizazzione lineare**\n",
        "\n",
        "Idea è di approssimare la function value $v(s;\\mathbf{w})$ , oppure la value-action $q$, con una funzione lineare parametrizzata dai i pesi $\\mathbf{w}$:\n",
        "\n",
        "$v(s;\\mathbf{w}) =  \\mathbf{w} ^\\intercal \\mathbf{x}(s) =  \\sum_i ^d w_i x_i(s)  $\n",
        "\n",
        "\n",
        "Dove il vettore colonna di dimensione $d$:\n",
        "\n",
        "$\\mathbf{x}(s) = (x_1(s), x_2(s),...,x_d(s) )^\\intercal $\n",
        "\n",
        "è detto **feature vector**; questo mappa ogni stato $s$ in $\\mathbb{R}^d$. Precisamente i $d$ componenti, dette features, sono del tipo $ x_i : s → \\mathbb{R}$.\n",
        "\n",
        "In genere, il numero $d$ delle features è molto inferiore alla dimensione dello spazio degli stati.\n",
        "\n",
        "Si assume che le features siano funzioni note.\n",
        "\n",
        "I parametri $\\mathbf{w}$ sono le \"low-dimensional embeddings\" che desideriamo apprendere.\n",
        "\n",
        "Per maggiori info sulle \"Feature Construction for Linear Methods\" vedere il testo di Sutton & Barto, 2018, PDF 231\n",
        "\n",
        "\n",
        "Maggiori info sul Linear Q-learning:\n",
        "\n",
        "https://gibberblot.github.io/rl-notes/single-agent/function-approximation.html\n",
        "\n",
        "Ref:\n",
        "\n",
        "https://users.ece.cmu.edu/~yuejiec/ece18813B_notes/lecture16-linear-function-approximation.pdf\n"
      ],
      "metadata": {
        "id": "SoNKGmPILe95"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "**Modellizzazione non lineare**\n",
        "\n",
        "Questo tipo di modellizzazione utilizza funzioni non lineari per approssimare la funzione $v(s;\\mathbf{w})$, in particolare usa reti neurali artificiali addestrate con la backpropagation. Ha guadagnato grande popolarità negli ultimi anni sotto il nome di **deep reinforcement learning**."
      ],
      "metadata": {
        "id": "3ilNWSgAVFv1"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "**Gradient Descent Algorithm**\n",
        "\n",
        "Definita $J(\\mathbf{w})$, funzione differenziabile rispetti il vettore che ne definisce i parametri $\\mathbf{w}$\n",
        "\n",
        "Per trovare il minimo locale di $J(\\mathbf{w})$ aggiustimo iterativamente i valori di $\\mathbf{w}$ in direzione del gradiente di $J$:\n",
        "\n",
        "$\\mathbf{w}_{t+1} \\doteq \\mathbf{w}_{t} - α ∇_{\\mathbf{w}}J(\\mathbf{w}_{t}) $\n",
        "\n",
        "Sia il learning rate:  $ α > 0 $\n"
      ],
      "metadata": {
        "id": "ZX1jue7cVPQ5"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "##Incremental update\n",
        "\n",
        "\n",
        "Premesse:\n",
        "\n",
        "La derivata rispetto $w$ di una funzione del tipo $(1-w)^2$ è la seguente : $\\frac{d (1-w)^2}{dw} =  \\frac{d (-w)}{dw} \\cdot 2 \\cdot (1-w)  = -  2 \\cdot (1-w) $\n",
        "\n",
        "$\\;$\n",
        "\n",
        "Definizioni:\n",
        "\n",
        "1.\n",
        "\n",
        "True value function:\n",
        "\n",
        "$v_\\pi(s)$ è un ideale value-fuction che si ottiene seguendo una certa policy $\\pi$.  Di fatto $v_\\pi(s)$ è una funzione valore sconosciuta (se lo fosse, non avremo bisogno di queste tecniche di Value Function Approximation) , si usa solo con finalità didattiche. $v_\\pi(s)$ è funge da nostro target.\n",
        "\n",
        "2.\n",
        "\n",
        "Approximate value function\n",
        "\n",
        "$v̂(s;\\mathbf{w})$ è il modello che ci interessa apprendere, in funzione degli stati e dei pesi.\n",
        "\n",
        "3.\n",
        "\n",
        "Objective function (o loss fuction):\n",
        "\n",
        "$J(s; \\mathbf{w}) = \\frac{1}{2} \\; \\mathbb{E}_{s \\sim \\pi} [ (v_\\pi(s) - v̂(s;\\mathbf{w}))^2 ]$\n",
        "\n",
        "\n",
        "\n",
        "che quantifica la discrepanza tra le previsioni del modello e i dati del target. Nota l'unico ruolo del fattore $\\frac{1}{2}$  è quello di semplificare la notazione matematica, rendendola più concisa dopo l'operazione di derivazione che seguirà. La presenza o meno di tale fattore non modifica i nostri intenti.\n",
        "\n",
        "\n",
        "\n",
        "$\\;$\n",
        "\n",
        "Scopo:\n",
        "\n",
        "al fine di trovare un minimo locale di $J(s; \\mathbf{w})$ applichiamo il Gradient Descent Algorithm:\n",
        "\n",
        "$∇_{\\mathbf{w}} J(s; \\mathbf{w}) = - [ (v_\\pi(s) - v̂(s;\\mathbf{w}))] \\cdot ∇_{\\mathbf{w}} v̂(s;\\mathbf{w}) $\n",
        "\n",
        "che inserito nella formula di update dei parametri, la trasforma :\n",
        "\n",
        "$\\mathbf{w}_{t+1} \\doteq \\mathbf{w}_{t} + α [ (v_\\pi(s_t) - v̂(s_t;\\mathbf{w_t}))] \\cdot ∇_{\\mathbf{w_t}} v̂(s_t;\\mathbf{w_t}) $\n",
        "\n",
        "\n",
        "$ ∇_{\\mathbf{w_t}} v̂(s_t;\\mathbf{w_t}) = ( \\frac{\\partial v̂(s_t;\\mathbf{w_t}) }{\\partial w_1},\\frac{\\partial v̂(s_t; \\mathbf{w_t}) }{\\partial w_2} ,...,\\frac{\\partial v̂(s_t; \\mathbf{w_t}) }{\\partial w_d})^\\intercal $ è il vettore gradiente rispetto $\\mathbf{w_t}$. In pratica un vettore colonna delle derivate parziali rispetto alle singole componenti.\n",
        "\n",
        "\n",
        "Si noti.\n",
        "\n",
        "Nel caso della modellizazzione lineare la formula si riduce a:\n",
        "\n",
        "$\\mathbf{w}_{t+1} \\doteq \\mathbf{w}_{t} + α [ (v_\\pi(s_t) - v̂(s_t;\\mathbf{w_t}))] \\cdot \\mathbf{x}(s_t) $\n",
        "\n",
        "$\\;$\n",
        "\n",
        "Caso concreto:\n",
        "\n",
        "In pratica non possiamo eseguire l'aggiornamento esatto dei pesi visto sopra perché $v_\\pi(s)$ è  una idealizzazione. Però possiamo usare una qualche approssimazione di questo target, ottenuta dai dati a disposizione; indichiamo tale approssimazione con $U(s_t)$.\n",
        "\n",
        "L' update rule diventa:\n",
        "\n",
        "$\\mathbf{w}_{t+1} \\doteq \\mathbf{w}_{t} + α [ (U(s_t)  - v̂(s_t;\\mathbf{w_t}))] \\cdot ∇_{\\mathbf{w_t}} v̂(s_t;\\mathbf{w_t}) $\n",
        "\n",
        "$\\;$\n",
        "\n",
        "Come concretizzare $U(s_t)$ ?\n",
        "\n",
        "\n",
        "1.\n",
        "\n",
        "Monte Carlo target - stimare unbiased :\n",
        "$U(s_t) = G(s_t)$.\n",
        "\n",
        "Ovvero il return totale $G(s_t)$ che si ottiene partendo dalla stato $s_t$ , e segundo la policy $\\pi$\n",
        "\n",
        "Per il Gradient Monte Carlo Algorithm vedi Sutton & Barto, 2018,  PDF 224.\n",
        "\n",
        "2.\n",
        "\n",
        "Bootstrapping targets - stimantori*biased. In particolare nel caso TD(0):\n",
        "\n",
        "$U(s_t) = R_{t+1} + γ \\cdot v̂(s_{t+1}, \\mathbf{w}_{t})$\n",
        "\n",
        "I metodi TD(n) non sono veri e propri \"gradient methods\", ma sono detti semi-gradient TD(n). In tali metodi di bootstrapping (incluso DP), il target $U(s_t)$, non è utilizzato esplicitamente nel calcolo del gradiente, viene infatti pensato come una costante e non in funzione di $\\mathbf{w}$. Vedi : Sutton & Barto, 2018, PDF 224. L'algoritmo richiede questa impostazione per garantire una ragionevole convergenza; ed i motivi precisi per cui questo sia necessario non sono ancora del tutto chiari.\n",
        "\n",
        "Update rule TD(0):\n",
        "\n",
        "$\\mathbf{w}_{t+1} \\doteq \\mathbf{w}_{t} + α [ R_{t+1} + γ \\cdot v̂(s_{t+1}, \\mathbf{w}_{t}) - v̂(s_t;\\mathbf{w_t})] \\cdot ∇_{\\mathbf{w_t}} v̂(s_t;\\mathbf{w_t}) $\n",
        "\n",
        "\n",
        "\n",
        "\n",
        "\n",
        "\n",
        "\n",
        "\n",
        "\n",
        "\n",
        "\n",
        "\n",
        "\n",
        "\n",
        "\n",
        "\n",
        "\n"
      ],
      "metadata": {
        "id": "03L-P7QnZjCv"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "##The deadly triad\n",
        "\n",
        "\n",
        "Sutton & Barto, 2018,PDF 286\n",
        "\n",
        "L'integrazione dei seguenti tre elementi all'interno di un singolo algoritmo può determinare l'insorgere di fenomeni di instabilità e divergenza:\n",
        "\n",
        "- Function approximation:\n",
        "un modo potente e scalabile per trattare spazi di stati molto più grande della memoria e delle risorse computazionali disponibili\n",
        "\n",
        "- Bootstrapping: \"learning value estimates from other value estimates\", vedi la dynamic\n",
        "programming ed il  Temporal-Difference method\n",
        "\n",
        "- Off-policy training: avere una behavior policy che genera i dati, diversa dalla target policy ovvero, dalla politica ottimale che l'agente desidera imparare. Vedi Q-learning.\n",
        "\n",
        "\n",
        "$\\;$\n",
        "\n",
        "Sutton scrive:\n",
        "\n",
        "*If any two elements of the deadly triad are present, but not all three, then instability can be avoided.*\n",
        "\n",
        "\n"
      ],
      "metadata": {
        "id": "ecfBa20KMfRF"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "# Coding: Solve gym's Cartpole problem DQN vs. Gradient Monte Carlo control and more ..."
      ],
      "metadata": {
        "id": "t_HW5QInQQbk"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "##Moduli"
      ],
      "metadata": {
        "id": "o11bM0eqbvYC"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "import sys\n",
        "IN_COLAB = \"google.colab\" in sys.modules\n",
        "\n",
        "if IN_COLAB:\n",
        "    !apt install python-opengl\n",
        "    !apt install ffmpeg\n",
        "    !apt install xvfb\n",
        "    !pip install PyVirtualDisplay==3.0\n",
        "    !pip install gymnasium==0.28.1\n",
        "    from pyvirtualdisplay import Display\n",
        "\n",
        "    # Start virtual display\n",
        "    dis = Display(visible=0, size=(400, 400))\n",
        "    dis.start()\n",
        "\n",
        "import os\n",
        "from typing import Dict, List, Tuple\n",
        "import warnings\n",
        "warnings.filterwarnings('ignore')\n",
        "\n",
        "import gymnasium as gym\n",
        "import matplotlib.pyplot as plt\n",
        "import numpy as np\n",
        "\n",
        "import torch\n",
        "import torch.nn as nn\n",
        "import torch.nn.functional as F\n",
        "import torch.optim as optim\n",
        "\n",
        "from IPython.display import clear_output"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "LSvE0B3GbxYW",
        "outputId": "96b42776-e241-46ec-8f55-d5e5559cea71"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Reading package lists... Done\n",
            "Building dependency tree... Done\n",
            "Reading state information... Done\n",
            "\u001b[1;31mE: \u001b[0mUnable to locate package python-opengl\u001b[0m\n",
            "Reading package lists... Done\n",
            "Building dependency tree... Done\n",
            "Reading state information... Done\n",
            "ffmpeg is already the newest version (7:4.4.2-0ubuntu0.22.04.1).\n",
            "0 upgraded, 0 newly installed, 0 to remove and 45 not upgraded.\n",
            "Reading package lists... Done\n",
            "Building dependency tree... Done\n",
            "Reading state information... Done\n",
            "The following additional packages will be installed:\n",
            "  libfontenc1 libxfont2 libxkbfile1 x11-xkb-utils xfonts-base xfonts-encodings xfonts-utils\n",
            "  xserver-common\n",
            "The following NEW packages will be installed:\n",
            "  libfontenc1 libxfont2 libxkbfile1 x11-xkb-utils xfonts-base xfonts-encodings xfonts-utils\n",
            "  xserver-common xvfb\n",
            "0 upgraded, 9 newly installed, 0 to remove and 45 not upgraded.\n",
            "Need to get 7,813 kB of archives.\n",
            "After this operation, 11.9 MB of additional disk space will be used.\n",
            "Get:1 http://archive.ubuntu.com/ubuntu jammy/main amd64 libfontenc1 amd64 1:1.1.4-1build3 [14.7 kB]\n",
            "Get:2 http://archive.ubuntu.com/ubuntu jammy/main amd64 libxfont2 amd64 1:2.0.5-1build1 [94.5 kB]\n",
            "Get:3 http://archive.ubuntu.com/ubuntu jammy/main amd64 libxkbfile1 amd64 1:1.1.0-1build3 [71.8 kB]\n",
            "Get:4 http://archive.ubuntu.com/ubuntu jammy/main amd64 x11-xkb-utils amd64 7.7+5build4 [172 kB]\n",
            "Get:5 http://archive.ubuntu.com/ubuntu jammy/main amd64 xfonts-encodings all 1:1.0.5-0ubuntu2 [578 kB]\n",
            "Get:6 http://archive.ubuntu.com/ubuntu jammy/main amd64 xfonts-utils amd64 1:7.7+6build2 [94.6 kB]\n",
            "Get:7 http://archive.ubuntu.com/ubuntu jammy/main amd64 xfonts-base all 1:1.0.5 [5,896 kB]\n",
            "Get:8 http://archive.ubuntu.com/ubuntu jammy-updates/main amd64 xserver-common all 2:21.1.4-2ubuntu1.7~22.04.10 [28.5 kB]\n",
            "Get:9 http://archive.ubuntu.com/ubuntu jammy-updates/universe amd64 xvfb amd64 2:21.1.4-2ubuntu1.7~22.04.10 [863 kB]\n",
            "Fetched 7,813 kB in 0s (33.6 MB/s)\n",
            "Selecting previously unselected package libfontenc1:amd64.\n",
            "(Reading database ... 121918 files and directories currently installed.)\n",
            "Preparing to unpack .../0-libfontenc1_1%3a1.1.4-1build3_amd64.deb ...\n",
            "Unpacking libfontenc1:amd64 (1:1.1.4-1build3) ...\n",
            "Selecting previously unselected package libxfont2:amd64.\n",
            "Preparing to unpack .../1-libxfont2_1%3a2.0.5-1build1_amd64.deb ...\n",
            "Unpacking libxfont2:amd64 (1:2.0.5-1build1) ...\n",
            "Selecting previously unselected package libxkbfile1:amd64.\n",
            "Preparing to unpack .../2-libxkbfile1_1%3a1.1.0-1build3_amd64.deb ...\n",
            "Unpacking libxkbfile1:amd64 (1:1.1.0-1build3) ...\n",
            "Selecting previously unselected package x11-xkb-utils.\n",
            "Preparing to unpack .../3-x11-xkb-utils_7.7+5build4_amd64.deb ...\n",
            "Unpacking x11-xkb-utils (7.7+5build4) ...\n",
            "Selecting previously unselected package xfonts-encodings.\n",
            "Preparing to unpack .../4-xfonts-encodings_1%3a1.0.5-0ubuntu2_all.deb ...\n",
            "Unpacking xfonts-encodings (1:1.0.5-0ubuntu2) ...\n",
            "Selecting previously unselected package xfonts-utils.\n",
            "Preparing to unpack .../5-xfonts-utils_1%3a7.7+6build2_amd64.deb ...\n",
            "Unpacking xfonts-utils (1:7.7+6build2) ...\n",
            "Selecting previously unselected package xfonts-base.\n",
            "Preparing to unpack .../6-xfonts-base_1%3a1.0.5_all.deb ...\n",
            "Unpacking xfonts-base (1:1.0.5) ...\n",
            "Selecting previously unselected package xserver-common.\n",
            "Preparing to unpack .../7-xserver-common_2%3a21.1.4-2ubuntu1.7~22.04.10_all.deb ...\n",
            "Unpacking xserver-common (2:21.1.4-2ubuntu1.7~22.04.10) ...\n",
            "Selecting previously unselected package xvfb.\n",
            "Preparing to unpack .../8-xvfb_2%3a21.1.4-2ubuntu1.7~22.04.10_amd64.deb ...\n",
            "Unpacking xvfb (2:21.1.4-2ubuntu1.7~22.04.10) ...\n",
            "Setting up libfontenc1:amd64 (1:1.1.4-1build3) ...\n",
            "Setting up xfonts-encodings (1:1.0.5-0ubuntu2) ...\n",
            "Setting up libxkbfile1:amd64 (1:1.1.0-1build3) ...\n",
            "Setting up libxfont2:amd64 (1:2.0.5-1build1) ...\n",
            "Setting up x11-xkb-utils (7.7+5build4) ...\n",
            "Setting up xfonts-utils (1:7.7+6build2) ...\n",
            "Setting up xfonts-base (1:1.0.5) ...\n",
            "Setting up xserver-common (2:21.1.4-2ubuntu1.7~22.04.10) ...\n",
            "Setting up xvfb (2:21.1.4-2ubuntu1.7~22.04.10) ...\n",
            "Processing triggers for man-db (2.10.2-1) ...\n",
            "Processing triggers for fontconfig (2.13.1-4.2ubuntu5) ...\n",
            "Processing triggers for libc-bin (2.35-0ubuntu3.4) ...\n",
            "/sbin/ldconfig.real: /usr/local/lib/libtbb.so.12 is not a symbolic link\n",
            "\n",
            "/sbin/ldconfig.real: /usr/local/lib/libtbbmalloc.so.2 is not a symbolic link\n",
            "\n",
            "/sbin/ldconfig.real: /usr/local/lib/libtbbbind.so.3 is not a symbolic link\n",
            "\n",
            "/sbin/ldconfig.real: /usr/local/lib/libtbbmalloc_proxy.so.2 is not a symbolic link\n",
            "\n",
            "/sbin/ldconfig.real: /usr/local/lib/libtbbbind_2_5.so.3 is not a symbolic link\n",
            "\n",
            "/sbin/ldconfig.real: /usr/local/lib/libtbbbind_2_0.so.3 is not a symbolic link\n",
            "\n",
            "Collecting PyVirtualDisplay==3.0\n",
            "  Downloading PyVirtualDisplay-3.0-py3-none-any.whl (15 kB)\n",
            "Installing collected packages: PyVirtualDisplay\n",
            "Successfully installed PyVirtualDisplay-3.0\n",
            "Collecting gymnasium==0.28.1\n",
            "  Downloading gymnasium-0.28.1-py3-none-any.whl (925 kB)\n",
            "\u001b[2K     \u001b[90m━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[32m925.5/925.5 kB\u001b[0m \u001b[31m15.2 MB/s\u001b[0m eta \u001b[36m0:00:00\u001b[0m\n",
            "\u001b[?25hRequirement already satisfied: numpy>=1.21.0 in /usr/local/lib/python3.10/dist-packages (from gymnasium==0.28.1) (1.25.2)\n",
            "Collecting jax-jumpy>=1.0.0 (from gymnasium==0.28.1)\n",
            "  Downloading jax_jumpy-1.0.0-py3-none-any.whl (20 kB)\n",
            "Requirement already satisfied: cloudpickle>=1.2.0 in /usr/local/lib/python3.10/dist-packages (from gymnasium==0.28.1) (2.2.1)\n",
            "Requirement already satisfied: typing-extensions>=4.3.0 in /usr/local/lib/python3.10/dist-packages (from gymnasium==0.28.1) (4.11.0)\n",
            "Collecting farama-notifications>=0.0.1 (from gymnasium==0.28.1)\n",
            "  Downloading Farama_Notifications-0.0.4-py3-none-any.whl (2.5 kB)\n",
            "Installing collected packages: farama-notifications, jax-jumpy, gymnasium\n",
            "Successfully installed farama-notifications-0.0.4 gymnasium-0.28.1 jax-jumpy-1.0.0\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "## DQN"
      ],
      "metadata": {
        "id": "csd6dOJNHEQt"
      }
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "XRFp4Ra8JJ26"
      },
      "outputs": [],
      "source": [
        "#codice ispirato:\n",
        "#https://colab.research.google.com/github/Curt-Park/rainbow-is-all-you-need/blob/master/01.dqn.ipynb"
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "### Network\n",
        "\n"
      ],
      "metadata": {
        "id": "sAq5m-iFfEem"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "class Network(nn.Module):\n",
        "    def __init__(self, in_dim: int, out_dim: int):\n",
        "      # in:  state\n",
        "      # out: 2 x q_value_state_action\n",
        "        super(Network, self).__init__()\n",
        "\n",
        "        self.layers = nn.Sequential(\n",
        "            nn.Linear(in_dim, 128),\n",
        "            nn.ReLU(),\n",
        "            nn.Linear(128, 128),\n",
        "            nn.ReLU(),\n",
        "            nn.Linear(128, out_dim)\n",
        "        )\n",
        "\n",
        "    def forward(self, x: torch.Tensor) -> torch.Tensor:\n",
        "        return self.layers(x)"
      ],
      "metadata": {
        "id": "MEC0Z5FmfFPH"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "source": [
        "###Experience Replay Memory\n",
        "\n",
        "The Experience Replay Memory data structure, indicated in this notebook with $U(D)$, allows to reduce the probability that adjacent transitions ($s, a, r, s'$) have a correlation.\n",
        "\n",
        "This structure allows the agent to sample experiences in a uniform random way, thus ensuring an independent distribution of transitions."
      ],
      "metadata": {
        "id": "ZjeI-K4Wb5nI"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "class ReplayBuffer:\n",
        "    \"\"\"A simple numpy replay buffer.\"\"\"\n",
        "\n",
        "    def __init__(self, obs_dim: int, size: int, batch_size: int = 32):\n",
        "        self.obs_buf = np.zeros([size, obs_dim], dtype=np.float32)\n",
        "        self.next_obs_buf = np.zeros([size, obs_dim], dtype=np.float32)\n",
        "        self.acts_buf = np.zeros([size], dtype=np.float32)\n",
        "        self.rews_buf = np.zeros([size], dtype=np.float32)\n",
        "        self.done_buf = np.zeros(size, dtype=np.float32)\n",
        "        self.max_size, self.batch_size = size, batch_size\n",
        "        self.ptr, self.size, = 0, 0\n",
        "\n",
        "    def store(\n",
        "        self,\n",
        "        obs: np.ndarray,\n",
        "        act: np.ndarray,\n",
        "        rew: float,\n",
        "        next_obs: np.ndarray,\n",
        "        done: bool,\n",
        "    ):\n",
        "        self.obs_buf[self.ptr] = obs\n",
        "        self.next_obs_buf[self.ptr] = next_obs\n",
        "        self.acts_buf[self.ptr] = act\n",
        "        self.rews_buf[self.ptr] = rew\n",
        "        self.done_buf[self.ptr] = done\n",
        "        self.ptr = (self.ptr + 1) % self.max_size\n",
        "        self.size = min(self.size + 1, self.max_size)\n",
        "\n",
        "    def sample_batch(self) -> Dict[str, np.ndarray]:\n",
        "        idxs = np.random.choice(self.size, size=self.batch_size, replace=False)\n",
        "        return dict(obs=self.obs_buf[idxs],\n",
        "                    next_obs=self.next_obs_buf[idxs],\n",
        "                    acts=self.acts_buf[idxs],\n",
        "                    rews=self.rews_buf[idxs],\n",
        "                    done=self.done_buf[idxs])\n",
        "\n",
        "    def __len__(self) -> int:\n",
        "        return self.size\n",
        "\n",
        " #Note:\n",
        " # numpy.ndarray. numpy.ndarray is even faster than list due to the fact that it\n",
        " #is a homogeneous array of fixed-size items, so you can get the benefits of locality of reference.\n",
        " # Whereas list is an array of pointers to objects, even when all of them are of the same type."
      ],
      "metadata": {
        "id": "KHyWwhyOejd3"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "source": [
        "### DQN Agent\n",
        "\n",
        "Compendio della classe DQNAgent .\n",
        "\n",
        "| Method           | Descrizione                                          |\n",
        "| ---              | ---                                                  |\n",
        "|select_action     | select an action from the input state.               |\n",
        "|step              | take an action and return the response of the env.   |\n",
        "|compute_dqn_loss  | return dqn loss.                                     |\n",
        "|update_model      | update the model by gradient descent.                |\n",
        "|target_hard_update| hard update from the local model to the target model.|\n",
        "|train             | train the agent during num_frames.                   |\n",
        "|test              | test the agent (1 episode).                          |\n",
        "|plot              | plot the training progresses.                        |\n"
      ],
      "metadata": {
        "id": "kJEcEwDQfjH2"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "**Fixed Q-target**\n",
        "\n",
        "\n",
        "DQN uses an iterative update that adjusts the action-values ($Q$) towards target values that are only periodically updated, thereby reducing correlations with the target; if not, it is easily divergy because the target continuously moves. The Q-learning update at iteration $i$ uses the following loss function:\n",
        "\n",
        "$$\n",
        "L_i(\\theta_i) = \\mathbb{E}_{(s,a,r,s') \\sim U(D)} \\big[ \\big( r + \\gamma \\max_{a'} Q(s',a';\\theta_i^-) - Q(s, a; \\theta_i) \\big)^2 \\big]\n",
        "$$\n",
        "\n",
        "in which $\\gamma$ is the discount factor determining the agent’s horizon, $\\theta_i$ are the parameters of the Q-network at iteration $i$ and $\\theta_i^-$ are the network parameters used to compute the target at iteration $i$. The target network parameters $\\theta_i^-$ are only updated with the Q-network parameters ($\\theta_i$) every C steps and are held fixed between individual updates.\n",
        "\n",
        "**Gradient clipping**\n",
        "\n",
        "The authors also found it helpful to clip the error term from the update $r + \\gamma \\max_{a'} Q(s', a'; \\theta_i^-) - Q(s,a,;\\theta_i)$ to be between -1 and 1. Because the absolute value loss function $|x|$ has a derivative of -1 for all negative values of x and a derivative of 1 for all positive values of x, clipping the squared error to be between -1 and 1 corresponds to using an absolute value loss function for errors outside of the (-1,1) interval. This form of error clipping further improved the stability of the algorithm."
      ],
      "metadata": {
        "id": "r7VjAjUrhbwO"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "class DQNAgent:\n",
        "    \"\"\"DQN Agent interacting with environment.\n",
        "\n",
        "    Attribute:\n",
        "        env (gym.Env): openAI Gym environment\n",
        "        memory (ReplayBuffer): replay memory to store transitions\n",
        "        batch_size (int): batch size for sampling\n",
        "        epsilon (float): parameter for epsilon greedy policy\n",
        "        epsilon_decay (float): step size to decrease epsilon\n",
        "        max_epsilon (float): max value of epsilon\n",
        "        min_epsilon (float): min value of epsilon\n",
        "        target_update (int): period for target model's hard update\n",
        "        gamma (float): discount factor\n",
        "        dqn (Network): model to train and select actions\n",
        "        dqn_target (Network): target model to update\n",
        "        optimizer (torch.optim): optimizer for training dqn\n",
        "        transition (list): transition information including\n",
        "                           state, action, reward, next_state, done\n",
        "    \"\"\"\n",
        "\n",
        "    def __init__(\n",
        "        self,\n",
        "        env: gym.Env,\n",
        "        memory_size: int,\n",
        "        batch_size: int,\n",
        "        target_update: int,\n",
        "        epsilon_decay: float,\n",
        "        max_epsilon: float = 1.0,\n",
        "        min_epsilon: float = 0.1,\n",
        "        gamma: float = 0.95,\n",
        "    ):\n",
        "        \"\"\"Initialization.\n",
        "\n",
        "        Args:\n",
        "            env (gym.Env): openAI Gym environment\n",
        "            memory_size (int): length of memory\n",
        "            batch_size (int): batch size for sampling\n",
        "            target_update (int): period for target model's hard update\n",
        "            epsilon_decay (float): step size to decrease epsilon\n",
        "            lr (float): learning rate\n",
        "            max_epsilon (float): max value of epsilon\n",
        "            min_epsilon (float): min value of epsilon\n",
        "            gamma (float): discount factor\n",
        "        \"\"\"\n",
        "        obs_dim = env.observation_space.shape[0]\n",
        "        action_dim = env.action_space.n\n",
        "\n",
        "        self.env = env\n",
        "        self.memory = ReplayBuffer(obs_dim, memory_size, batch_size)\n",
        "        self.batch_size = batch_size\n",
        "        self.epsilon = max_epsilon\n",
        "        self.epsilon_decay = epsilon_decay\n",
        "        self.max_epsilon = max_epsilon\n",
        "        self.min_epsilon = min_epsilon\n",
        "        self.target_update = target_update\n",
        "        self.gamma = gamma\n",
        "\n",
        "        # device: cpu / gpu\n",
        "        self.device = torch.device(\n",
        "            \"cuda\" if torch.cuda.is_available() else \"cpu\"\n",
        "        )\n",
        "\n",
        "        # networks: dqn, dqn_target\n",
        "        self.dqn = Network(obs_dim, action_dim).to(self.device)\n",
        "        self.dqn_target = Network(obs_dim, action_dim).to(self.device)\n",
        "        self.dqn_target.load_state_dict(self.dqn.state_dict())\n",
        "        #https://pytorch.org/tutorials/recipes/recipes/what_is_state_dict.html\n",
        "        self.dqn_target.eval()\n",
        "        # The model modality to make predictions\n",
        "        # .eval() is equivalent with self.train(False)\n",
        "        #https://pytorch.org/docs/stable/generated/torch.nn.Module.html\n",
        "        #https://www.youtube.com/watch?v=GzjRE3MUx6Q\n",
        "\n",
        "        # optimizer\n",
        "        self.optimizer = optim.Adam(self.dqn.parameters())\n",
        "\n",
        "        # transition to store in memory\n",
        "        self.transition = list()\n",
        "\n",
        "        # mode: train / test\n",
        "        self.is_test = False\n",
        "\n",
        "    def select_action(self, state: np.ndarray) -> np.ndarray:\n",
        "        \"\"\"Select an action from the input state.\"\"\"\n",
        "        # epsilon greedy policy\n",
        "        if self.epsilon > np.random.random():\n",
        "            selected_action = self.env.action_space.sample()\n",
        "        else:\n",
        "            selected_action = self.dqn(\n",
        "                torch.FloatTensor(state).to(self.device)\n",
        "            ).argmax()\n",
        "            selected_action = selected_action.detach().cpu().numpy()\n",
        "\n",
        "        if not self.is_test:\n",
        "            self.transition = [state, selected_action]\n",
        "\n",
        "        return selected_action\n",
        "\n",
        "    def step(self, action: np.ndarray) -> Tuple[np.ndarray, np.float64, bool]:\n",
        "        \"\"\"Take an action and return the response of the env.\"\"\"\n",
        "        next_state, reward, terminated, truncated, _ = self.env.step(action)\n",
        "        done = terminated or truncated\n",
        "\n",
        "        if not self.is_test:\n",
        "            self.transition += [reward, next_state, done]\n",
        "            self.memory.store(*self.transition)\n",
        "\n",
        "        return next_state, reward, done\n",
        "\n",
        "    def update_model(self) -> torch.Tensor:\n",
        "        \"\"\"Update the model by gradient descent.\"\"\"\n",
        "        samples = self.memory.sample_batch()\n",
        "\n",
        "        loss = self._compute_dqn_loss(samples)\n",
        "\n",
        "        self.optimizer.zero_grad()\n",
        "        loss.backward()\n",
        "        self.optimizer.step()\n",
        "\n",
        "        return loss.item()\n",
        "\n",
        "    def train(self, num_frames: int, plotting_interval: int = 200):\n",
        "        \"\"\"Train the agent.\"\"\"\n",
        "        self.is_test = False\n",
        "\n",
        "        state, _ = self.env.reset()\n",
        "        update_cnt = 0\n",
        "        epsilons = []\n",
        "        losses = []\n",
        "        scores = []\n",
        "        deb_list = []\n",
        "        score = 0\n",
        "        ep=1\n",
        "\n",
        "\n",
        "        for frame_idx in range(1, num_frames + 1):\n",
        "            action = self.select_action(state)\n",
        "            next_state, reward, done = self.step(action)\n",
        "\n",
        "            state = next_state\n",
        "            score += reward\n",
        "\n",
        "            # if episode ends\n",
        "            if done:\n",
        "                state, _ = self.env.reset()\n",
        "                scores.append(score)\n",
        "                # Debug\n",
        "                if ep % 10 == 0:\n",
        "                  deb_list.append((ep, frame_idx, score))\n",
        "                ep+=1\n",
        "                score = 0\n",
        "\n",
        "            # if training is ready\n",
        "            if len(self.memory) >= self.batch_size:\n",
        "                loss = self.update_model()\n",
        "                losses.append(loss)\n",
        "                update_cnt += 1\n",
        "\n",
        "                # linearly decrease epsilon\n",
        "                self.epsilon = max(\n",
        "                    self.min_epsilon, self.epsilon - (\n",
        "                        self.max_epsilon - self.min_epsilon\n",
        "                    ) * self.epsilon_decay\n",
        "                )\n",
        "                epsilons.append(self.epsilon)\n",
        "\n",
        "                # if hard update is needed\n",
        "                if update_cnt % self.target_update == 0:\n",
        "                    self._target_hard_update()\n",
        "\n",
        "            # plotting\n",
        "            if frame_idx % plotting_interval == 0:\n",
        "                self._plot(frame_idx, scores, losses, epsilons)\n",
        "\n",
        "        self.env.close()\n",
        "\n",
        "        print('\\n...End of training...\\n')\n",
        "        print('Resume')\n",
        "        print('Episode - Tot frame -  Score Episode')\n",
        "        for ideb in deb_list:\n",
        "          print(ideb)\n",
        "\n",
        "    def test(self, video_folder: str) -> None:\n",
        "        \"\"\"Test the agent.\"\"\"\n",
        "        self.is_test = True\n",
        "\n",
        "        # for recording a video\n",
        "        naive_env = self.env\n",
        "        self.env = gym.wrappers.RecordVideo(self.env, video_folder=video_folder)\n",
        "\n",
        "        state, _ = self.env.reset()\n",
        "        done = False\n",
        "        score = 0\n",
        "\n",
        "        while not done:\n",
        "            action = self.select_action(state)\n",
        "            next_state, reward, done = self.step(action)\n",
        "\n",
        "            state = next_state\n",
        "            score += reward\n",
        "\n",
        "        print(\"score: \", score)\n",
        "        self.env.close()\n",
        "\n",
        "        # reset\n",
        "        self.env = naive_env\n",
        "\n",
        "    def _compute_dqn_loss(self, samples: Dict[str, np.ndarray]) -> torch.Tensor:\n",
        "        \"\"\"Return dqn loss.\"\"\"\n",
        "        device = self.device  # for shortening the following lines\n",
        "        state = torch.FloatTensor(samples[\"obs\"]).to(device)\n",
        "        next_state = torch.FloatTensor(samples[\"next_obs\"]).to(device)\n",
        "        action = torch.LongTensor(samples[\"acts\"].reshape(-1, 1)).to(device)\n",
        "        reward = torch.FloatTensor(samples[\"rews\"].reshape(-1, 1)).to(device)\n",
        "        done = torch.FloatTensor(samples[\"done\"].reshape(-1, 1)).to(device)\n",
        "\n",
        "        # G_t   = r + gamma * v(s_{t+1})  if state != Terminal\n",
        "        #       = r                       otherwise\n",
        "        curr_q_value = self.dqn(state).gather(1, action)\n",
        "        next_q_value = self.dqn_target(\n",
        "            next_state\n",
        "        ).max(dim=1, keepdim=True)[0].detach()\n",
        "        mask = 1 - done\n",
        "        #if done == True then mask == 0 -> target = reward\n",
        "        target = (reward + self.gamma * next_q_value * mask).to(self.device)\n",
        "\n",
        "        # calculate dqn loss\n",
        "        loss =   F.smooth_l1_loss(curr_q_value, target)\n",
        "        #Alternative\n",
        "        #loss_f = nn.MSELoss()\n",
        "        #loss = loss_f(curr_q_value, target)\n",
        "\n",
        "        return loss\n",
        "\n",
        "    def _target_hard_update(self):\n",
        "        \"\"\"Hard update: target <- local.\"\"\"\n",
        "        self.dqn_target.load_state_dict(self.dqn.state_dict())\n",
        "\n",
        "    def _plot(\n",
        "        self,\n",
        "        frame_idx: int,\n",
        "        scores: List[float],\n",
        "        losses: List[float],\n",
        "        epsilons: List[float],\n",
        "    ):\n",
        "        \"\"\"Plot the training progresses.\"\"\"\n",
        "        clear_output(True)\n",
        "        plt.figure(figsize=(20, 5))\n",
        "        plt.subplot(131)\n",
        "        plt.title('frame %s. score: %s' % (frame_idx, np.mean(scores[-10:])))\n",
        "        plt.plot(scores)\n",
        "        plt.subplot(132)\n",
        "        plt.title('loss')\n",
        "        plt.plot(losses)\n",
        "        plt.subplot(133)\n",
        "        plt.title('epsilons')\n",
        "        plt.plot(epsilons)\n",
        "        plt.show()\n",
        "\n"
      ],
      "metadata": {
        "id": "4nh60iF0fMyD"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "source": [
        "### Parametri ed inizializzazioni DQN"
      ],
      "metadata": {
        "id": "heZz3WlanSle"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "# environment\n",
        "env = gym.make(\"CartPole-v1\", max_episode_steps=300, render_mode=\"rgb_array\")\n",
        "\n",
        "\n",
        "# parameters\n",
        "num_frames = 12_000\n",
        "memory_size = 1000\n",
        "batch_size = 32\n",
        "target_update = 300\n",
        "epsilon_decay = 1 / 4_000\n",
        "\n",
        "agent = DQNAgent(env, memory_size, batch_size, target_update, epsilon_decay)"
      ],
      "metadata": {
        "id": "WWDwrrWtfUkt"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "source": [
        "### Train DQN"
      ],
      "metadata": {
        "id": "jlUCp-UxpPSV"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "agent.train(num_frames)\n",
        "#Commenti:\n",
        "#https://wandb.ai/peacekurella/Hitchiker-s-Guide-to-the-Galaxy-of-Reinforcement-Learning/reports/A-Brief-Overview-of-DQN-training--Vmlldzo1NTk5Mjk#:~:text=One%20interesting%20phenomenon%20to%20watch,and%20the%20policy%20network%20synchronize.\n"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 588
        },
        "id": "yolIlBUjpHW0",
        "outputId": "5114ea0d-e029-4e2b-d4b2-2364f38806a1"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 2000x500 with 3 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "\n",
            "...End of training...\n",
            "\n",
            "Resume\n",
            "Episode - Tot frame -  Score Episode\n",
            "(10, 202, 19.0)\n",
            "(20, 397, 32.0)\n",
            "(30, 600, 24.0)\n",
            "(40, 819, 31.0)\n",
            "(50, 1038, 15.0)\n",
            "(60, 1331, 37.0)\n",
            "(70, 1931, 177.0)\n",
            "(80, 2990, 195.0)\n",
            "(90, 4622, 195.0)\n",
            "(100, 6905, 201.0)\n",
            "(110, 9612, 62.0)\n",
            "(120, 11944, 300.0)\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "#### Test DQN"
      ],
      "metadata": {
        "id": "FONrCLRnz-R2"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "video_folder=\"videos/dqn\"\n",
        "agent.test(video_folder=video_folder)\n"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "VdEpdEWkzy_2",
        "outputId": "5e11c969-f8e7-4cea-f8c0-3e04b614eaa2"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Moviepy - Building video /content/videos/dqn/rl-video-episode-0.mp4.\n",
            "Moviepy - Writing video /content/videos/dqn/rl-video-episode-0.mp4\n",
            "\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": []
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Moviepy - Done !\n",
            "Moviepy - video ready /content/videos/dqn/rl-video-episode-0.mp4\n",
            "score:  300.0\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "import base64\n",
        "import glob\n",
        "import io\n",
        "import os\n",
        "\n",
        "from IPython.display import HTML, display\n",
        "\n",
        "\n",
        "def ipython_show_video(path: str) -> None:\n",
        "    \"\"\"Show a video at `path` within IPython Notebook.\"\"\"\n",
        "    if not os.path.isfile(path):\n",
        "        raise NameError(\"Cannot access: {}\".format(path))\n",
        "\n",
        "    video = io.open(path, \"r+b\").read()\n",
        "    encoded = base64.b64encode(video)\n",
        "\n",
        "    display(HTML(\n",
        "        data=\"\"\"\n",
        "        <video width=\"320\" height=\"240\" alt=\"test\" controls>\n",
        "        <source src=\"data:video/mp4;base64,{0}\" type=\"video/mp4\"/>\n",
        "        </video>\n",
        "        \"\"\".format(encoded.decode(\"ascii\"))\n",
        "    ))\n",
        "\n",
        "\n",
        "def show_latest_video(video_folder: str) -> str:\n",
        "    \"\"\"Show the most recently recorded video from video folder.\"\"\"\n",
        "    list_of_files = glob.glob(os.path.join(video_folder, \"*.mp4\"))\n",
        "    latest_file = max(list_of_files, key=os.path.getctime)\n",
        "    ipython_show_video(latest_file)\n",
        "    return latest_file\n",
        "\n",
        "\n",
        "latest_file = show_latest_video(video_folder=video_folder)\n",
        "print(\"Played:\", latest_file)"
      ],
      "metadata": {
        "id": "YB4h6Zsqz9Sm"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "source": [
        "##Gradient Monte Carlo target control\n",
        "\n",
        "In this section the Gradient Monte Carlo target control algorithm is implemented, using a network of smaller sizes than the one used in the DQN section (and with a greater number of episodes for training)."
      ],
      "metadata": {
        "id": "vT6mBtsViAZ_"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "class QNetwork(nn.Module):\n",
        "    def __init__(self, input_dim, hidden_dim, output_dim):\n",
        "        super(QNetwork, self).__init__()\n",
        "        self.fc1 = nn.Linear(input_dim, hidden_dim)\n",
        "        self.relu = nn.ReLU()\n",
        "        self.fc2 = nn.Linear(hidden_dim, output_dim)\n",
        "\n",
        "    def forward(self, x):\n",
        "        x = self.fc1(x)\n",
        "        x = self.relu(x)\n",
        "        x = self.fc2(x)\n",
        "        return x\n",
        "\n",
        "def epsilon_greedy_policy(q_values, epsilon):\n",
        "    if np.random.rand() < epsilon:\n",
        "        return np.random.randint(len(q_values))\n",
        "    else:\n",
        "        return np.argmax(q_values)\n",
        "\n",
        "def gradient_monte_carlo_control(env, q_net, optimizer, num_episodes, gamma=1.0, epsilon=0.1):\n",
        "\n",
        "    q_net.train()\n",
        "\n",
        "    scores = []\n",
        "    losses = []\n",
        "\n",
        "    for episode in range(1,num_episodes+1):\n",
        "        state,_ = env.reset()\n",
        "        #In the recent versions of OpenAI gym, env.reset() returns a tuple of (observation, info)\n",
        "        # and not just observation\n",
        "\n",
        "        rewards = []\n",
        "        states = []\n",
        "        actions = []\n",
        "\n",
        "        done = False\n",
        "        #trunc = False\n",
        "        # \"trunc\"  is used to avoid peaks in scores during the training, example score = 100_000.\n",
        "        # Setting the score limit to \"max_episode_steps\".\n",
        "\n",
        "        #while done== False and trunc == False:\n",
        "        while done== False:\n",
        "            state_tensor = torch.FloatTensor(state).unsqueeze(0)\n",
        "            q_values = q_net(state_tensor).detach().numpy().squeeze()\n",
        "            action = epsilon_greedy_policy(q_values, epsilon)\n",
        "\n",
        "            states.append(state)\n",
        "            actions.append(action)\n",
        "\n",
        "            next_state, reward, done, trunc, _ = env.step(action)\n",
        "            rewards.append(reward)\n",
        "            state = next_state\n",
        "\n",
        "        # Calculate returns\n",
        "        returns = []\n",
        "        G = 0\n",
        "        for reward in reversed(rewards):\n",
        "            G = reward + gamma * G\n",
        "            returns.insert(0, G)\n",
        "\n",
        "        # Update Q-network\n",
        "        loss_mean = 0\n",
        "        optimizer.zero_grad()\n",
        "        for state, action, G in zip(states, actions, returns):\n",
        "            state_tensor = torch.FloatTensor(state).unsqueeze(0)\n",
        "            G_tensor = torch.FloatTensor([G])\n",
        "            q_values = q_net(state_tensor)\n",
        "            q_value = q_values[0, action]\n",
        "            loss = abs(q_value - G_tensor)\n",
        "            loss_mean += loss\n",
        "            loss.backward()\n",
        "            optimizer.step()\n",
        "\n",
        "        losses.append(loss_mean.detach().numpy()/len(states))\n",
        "        scores.append(returns[0])\n",
        "\n",
        "        if (episode % 50) == 0:\n",
        "            #print(f\"Episode {episode + 1}/{num_episodes}, Loss: {loss.item()}, Score : {returns[0]}\")\n",
        "            clear_output(True)\n",
        "            plt.figure(figsize=(12, 5))\n",
        "            plt.subplot(131)\n",
        "            plt.title('Episode %s. - Score %s' % (episode,scores[episode-1]))\n",
        "            plt.plot(scores)\n",
        "            plt.subplot(132)\n",
        "            plt.title('Episode %s. - Loss mean %s' % (episode,losses[episode-1]))\n",
        "            plt.plot(losses)\n",
        "            plt.show()\n",
        "\n",
        "env = gym.make('CartPole-v1', max_episode_steps=300)\n",
        "input_dim = env.observation_space.shape[0]\n",
        "hidden_dim = 128\n",
        "output_dim = env.action_space.n\n",
        "q_net = QNetwork(input_dim, hidden_dim, output_dim)\n",
        "optimizer = optim.Adam(q_net.parameters(), lr=0.001)\n",
        "\n",
        "num_episodes = 500\n",
        "gradient_monte_carlo_control(env, q_net, optimizer, num_episodes)\n"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 391
        },
        "id": "PssJlpWwip--",
        "outputId": "7e0ddcb3-3cc4-48e8-e51b-0a3e3d3a1465"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 1000x400 with 2 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "def evaluate_policy(env, q_net, num_episodes):\n",
        "    q_net.eval()\n",
        "    total_rewards = []\n",
        "\n",
        "    for episode in range(num_episodes):\n",
        "        state,_ = env.reset()\n",
        "        total_reward = 0\n",
        "        done = False\n",
        "\n",
        "        while not done:\n",
        "            state_tensor = torch.FloatTensor(state).unsqueeze(0)\n",
        "            q_values = q_net(state_tensor).detach().numpy().squeeze()\n",
        "            action = np.argmax(q_values)\n",
        "\n",
        "            next_state, reward, done, _, _ = env.step(action)\n",
        "            total_reward += reward\n",
        "            state = next_state\n",
        "\n",
        "        total_rewards.append(total_reward)\n",
        "\n",
        "    average_reward = np.mean(total_rewards)\n",
        "    return average_reward\n",
        "\n",
        "# Evaluate the trained policy\n",
        "evaluation_episodes = 100\n",
        "average_reward = evaluate_policy(env, q_net, evaluation_episodes)\n",
        "print(f\"Average Reward over {evaluation_episodes} episodes: {average_reward}\")"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "AQovhy2nnGO8",
        "outputId": "a1bd916c-ffde-4356-b5a0-036a42cd9281"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Average Reward over 100 episodes: 575.99\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "#### Test MC"
      ],
      "metadata": {
        "id": "K6I7WJwc1vMw"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "video_folder=\"videos/dqn\"\n",
        "\n",
        "def campione_test(video_folder: str) -> None:\n",
        "        env = gym.make(\"CartPole-v1\",  render_mode=\"rgb_array\")\n",
        "        # for recording a video\n",
        "        naive_env = env\n",
        "        env = gym.wrappers.RecordVideo(env, video_folder=video_folder)\n",
        "\n",
        "        state, _ = env.reset()\n",
        "        done = False\n",
        "        total_reward = 0\n",
        "\n",
        "        while not done:\n",
        "            state_tensor = torch.FloatTensor(state).unsqueeze(0)\n",
        "            q_values = q_net(state_tensor).detach().numpy().squeeze()\n",
        "            action = np.argmax(q_values)\n",
        "\n",
        "            next_state, reward, done, _, _ = env.step(action)\n",
        "            total_reward += reward\n",
        "            state = next_state\n",
        "\n",
        "        print(\"Score: \", total_reward)\n",
        "\n",
        "\n",
        "        env.close()\n",
        "\n",
        "        # reset\n",
        "        env = naive_env\n",
        "\n",
        "campione_test(video_folder=video_folder)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "177692c7-e4ad-4d71-e082-a50392838f3b",
        "id": "UVXRzvah1vMx"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Moviepy - Building video /content/videos/dqn/rl-video-episode-0.mp4.\n",
            "Moviepy - Writing video /content/videos/dqn/rl-video-episode-0.mp4\n",
            "\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": []
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Moviepy - Done !\n",
            "Moviepy - video ready /content/videos/dqn/rl-video-episode-0.mp4\n",
            "Score:  30228.0\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "import base64\n",
        "import glob\n",
        "import io\n",
        "import os\n",
        "\n",
        "from IPython.display import HTML, display\n",
        "\n",
        "env = gym.make(\"CartPole-v1\", max_episode_steps=300, render_mode=\"rgb_array\")\n",
        "\n",
        "def ipython_show_video(path: str) -> None:\n",
        "    \"\"\"Show a video at `path` within IPython Notebook.\"\"\"\n",
        "    if not os.path.isfile(path):\n",
        "        raise NameError(\"Cannot access: {}\".format(path))\n",
        "\n",
        "    video = io.open(path, \"r+b\").read()\n",
        "    encoded = base64.b64encode(video)\n",
        "\n",
        "    display(HTML(\n",
        "        data=\"\"\"\n",
        "        <video width=\"320\" height=\"240\" alt=\"test\" controls>\n",
        "        <source src=\"data:video/mp4;base64,{0}\" type=\"video/mp4\"/>\n",
        "        </video>\n",
        "        \"\"\".format(encoded.decode(\"ascii\"))\n",
        "    ))\n",
        "\n",
        "\n",
        "def show_latest_video(video_folder: str) -> str:\n",
        "    \"\"\"Show the most recently recorded video from video folder.\"\"\"\n",
        "    list_of_files = glob.glob(os.path.join(video_folder, \"*.mp4\"))\n",
        "    latest_file = max(list_of_files, key=os.path.getctime)\n",
        "    ipython_show_video(latest_file)\n",
        "    return latest_file\n",
        "\n",
        "\n",
        "latest_file = show_latest_video(video_folder=video_folder)\n",
        "print(\"Played:\", latest_file)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 279
        },
        "id": "HDHnb3sP1vM0",
        "outputId": "18c90152-b9c5-4326-fc38-5b9e12c13d5f"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<IPython.core.display.HTML object>"
            ],
            "text/html": [
              "\n",
              "        <video width=\"320\" height=\"240\" alt=\"test\" controls>\n",
              "        <source src=\"data:video/mp4;base64,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\" type=\"video/mp4\"/>\n",
              "        </video>\n",
              "        "
            ]
          },
          "metadata": {}
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Played: videos/dqn/rl-video-episode-0.mp4\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "## DQN with imitative mode\n",
        "\n",
        "In this section, the network implemented with the Gradient Monte Carlo target control algorithm helps, in imitation mode every 5 steps, the DQN network (Note, for this training, a smaller number of episodes was used compared to classical DQN training)"
      ],
      "metadata": {
        "id": "cZjc5fAz8ab_"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "### DQN Agent with imitative mode\n",
        "\n"
      ],
      "metadata": {
        "id": "bfOh_cSu8acH"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "class DQNAgent:\n",
        "    \"\"\"DQN Agent interacting with environment.\n",
        "\n",
        "    Attribute:\n",
        "        env (gym.Env): openAI Gym environment\n",
        "        memory (ReplayBuffer): replay memory to store transitions\n",
        "        batch_size (int): batch size for sampling\n",
        "        epsilon (float): parameter for epsilon greedy policy\n",
        "        epsilon_decay (float): step size to decrease epsilon\n",
        "        max_epsilon (float): max value of epsilon\n",
        "        min_epsilon (float): min value of epsilon\n",
        "        target_update (int): period for target model's hard update\n",
        "        gamma (float): discount factor\n",
        "        dqn (Network): model to train and select actions\n",
        "        dqn_target (Network): target model to update\n",
        "        optimizer (torch.optim): optimizer for training dqn\n",
        "        transition (list): transition information including\n",
        "                           state, action, reward, next_state, done\n",
        "    \"\"\"\n",
        "\n",
        "    def __init__(\n",
        "        self,\n",
        "        env: gym.Env,\n",
        "        memory_size: int,\n",
        "        batch_size: int,\n",
        "        target_update: int,\n",
        "        epsilon_decay: float,\n",
        "        max_epsilon: float = 1.0,\n",
        "        min_epsilon: float = 0.1,\n",
        "        gamma: float = 0.95,\n",
        "    ):\n",
        "        \"\"\"Initialization.\n",
        "\n",
        "        Args:\n",
        "            env (gym.Env): openAI Gym environment\n",
        "            memory_size (int): length of memory\n",
        "            batch_size (int): batch size for sampling\n",
        "            target_update (int): period for target model's hard update\n",
        "            epsilon_decay (float): step size to decrease epsilon\n",
        "            lr (float): learning rate\n",
        "            max_epsilon (float): max value of epsilon\n",
        "            min_epsilon (float): min value of epsilon\n",
        "            gamma (float): discount factor\n",
        "        \"\"\"\n",
        "        obs_dim = env.observation_space.shape[0]\n",
        "        action_dim = env.action_space.n\n",
        "\n",
        "        self.env = env\n",
        "        self.memory = ReplayBuffer(obs_dim, memory_size, batch_size)\n",
        "        self.batch_size = batch_size\n",
        "        self.epsilon = max_epsilon\n",
        "        self.epsilon_decay = epsilon_decay\n",
        "        self.max_epsilon = max_epsilon\n",
        "        self.min_epsilon = min_epsilon\n",
        "        self.target_update = target_update\n",
        "        self.gamma = gamma\n",
        "\n",
        "        # device: cpu / gpu\n",
        "        self.device = torch.device(\n",
        "            \"cuda\" if torch.cuda.is_available() else \"cpu\"\n",
        "        )\n",
        "\n",
        "        # networks: dqn, dqn_target\n",
        "        self.dqn = Network(obs_dim, action_dim).to(self.device)\n",
        "        self.dqn_target = Network(obs_dim, action_dim).to(self.device)\n",
        "        self.dqn_target.load_state_dict(self.dqn.state_dict())\n",
        "        self.dqn_target.eval()\n",
        "\n",
        "\n",
        "        # optimizer\n",
        "        self.optimizer = optim.Adam(self.dqn.parameters())\n",
        "\n",
        "        # transition to store in memory\n",
        "        self.transition = list()\n",
        "\n",
        "        # mode: train / test\n",
        "        self.is_test = False\n",
        "\n",
        "    def select_action(self, state, frame_idx = num_frames) :\n",
        "\n",
        "        #imitative mode\n",
        "        if frame_idx % 5 == 0: # < (num_frames/2):\n",
        "              state_tensor = torch.FloatTensor(state).unsqueeze(0)\n",
        "              q_values = q_net(state_tensor).detach().numpy().squeeze()\n",
        "              selected_action = np.argmax(q_values)\n",
        "\n",
        "        else:\n",
        "        # epsilon greedy policy\n",
        "          if self.epsilon > np.random.random():\n",
        "              selected_action = self.env.action_space.sample()\n",
        "          else:\n",
        "            selected_action = self.dqn(\n",
        "                torch.FloatTensor(state).to(self.device)\n",
        "            ).argmax()\n",
        "            selected_action = selected_action.detach().cpu().numpy()\n",
        "\n",
        "        if not self.is_test:\n",
        "            self.transition = [state, selected_action]\n",
        "\n",
        "        return selected_action\n",
        "\n",
        "    def step(self, action: np.ndarray) -> Tuple[np.ndarray, np.float64, bool]:\n",
        "        \"\"\"Take an action and return the response of the env.\"\"\"\n",
        "        next_state, reward, terminated, truncated, _ = self.env.step(action)\n",
        "        done = terminated or truncated\n",
        "\n",
        "        if not self.is_test:\n",
        "            self.transition += [reward, next_state, done]\n",
        "            self.memory.store(*self.transition)\n",
        "\n",
        "        return next_state, reward, done\n",
        "\n",
        "    def update_model(self) -> torch.Tensor:\n",
        "        \"\"\"Update the model by gradient descent.\"\"\"\n",
        "        samples = self.memory.sample_batch()\n",
        "\n",
        "        loss = self._compute_dqn_loss(samples)\n",
        "\n",
        "        self.optimizer.zero_grad()\n",
        "        loss.backward()\n",
        "        self.optimizer.step()\n",
        "\n",
        "        return loss.item()\n",
        "\n",
        "    def train(self, num_frames: int, plotting_interval: int = 200):\n",
        "        \"\"\"Train the agent.\"\"\"\n",
        "        self.is_test = False\n",
        "\n",
        "        state, _ = self.env.reset()\n",
        "        update_cnt = 0\n",
        "        epsilons = []\n",
        "        losses = []\n",
        "        scores = []\n",
        "        deb_list = []\n",
        "        score = 0\n",
        "        ep=1\n",
        "\n",
        "\n",
        "        for frame_idx in range(1, num_frames + 1):\n",
        "\n",
        "            action = self.select_action(state, frame_idx)\n",
        "\n",
        "            next_state, reward, done = self.step(action)\n",
        "\n",
        "            state = next_state\n",
        "            score += reward\n",
        "\n",
        "            # if episode ends\n",
        "            if done:\n",
        "                state, _ = self.env.reset()\n",
        "                scores.append(score)\n",
        "                # Debug\n",
        "                if ep % 10 == 0:\n",
        "                  deb_list.append((ep, frame_idx, score))\n",
        "                ep+=1\n",
        "                score = 0\n",
        "\n",
        "            # if training is ready\n",
        "            if len(self.memory) >= self.batch_size:\n",
        "                loss = self.update_model()\n",
        "                losses.append(loss)\n",
        "                update_cnt += 1\n",
        "\n",
        "                # linearly decrease epsilon\n",
        "                self.epsilon = max(\n",
        "                    self.min_epsilon, self.epsilon - (\n",
        "                        self.max_epsilon - self.min_epsilon\n",
        "                    ) * self.epsilon_decay\n",
        "                )\n",
        "                epsilons.append(self.epsilon)\n",
        "\n",
        "                # if hard update is needed\n",
        "                if update_cnt % self.target_update == 0:\n",
        "                    self._target_hard_update()\n",
        "\n",
        "            # plotting\n",
        "            if frame_idx % plotting_interval == 0:\n",
        "                self._plot(frame_idx, scores, losses, epsilons)\n",
        "\n",
        "        self.env.close()\n",
        "\n",
        "        print('\\n...End of training...\\n')\n",
        "\n",
        "\n",
        "    def test(self, video_folder: str) -> None:\n",
        "        \"\"\"Test the agent.\"\"\"\n",
        "        self.is_test = True\n",
        "\n",
        "        # for recording a video\n",
        "        naive_env = self.env\n",
        "        self.env = gym.wrappers.RecordVideo(self.env, video_folder=video_folder)\n",
        "\n",
        "        state, _ = self.env.reset()\n",
        "        done = False\n",
        "        score = 0\n",
        "\n",
        "        while not done:\n",
        "            action = self.select_action(state)\n",
        "            next_state, reward, done = self.step(action)\n",
        "\n",
        "            state = next_state\n",
        "            score += reward\n",
        "\n",
        "        print(\"score: \", score)\n",
        "        self.env.close()\n",
        "\n",
        "        # reset\n",
        "        self.env = naive_env\n",
        "\n",
        "    def test_avr_100(self):\n",
        "        \"\"\"Test the agent.\"\"\"\n",
        "        self.is_test = True\n",
        "\n",
        "        total_rewards = []\n",
        "\n",
        "        for e in range(1,100+1):\n",
        "          state, _ = self.env.reset()\n",
        "          done = False\n",
        "          score = 0\n",
        "\n",
        "          while not done:\n",
        "             action = self.select_action(state)\n",
        "             next_state, reward, done = self.step(action)\n",
        "\n",
        "             state = next_state\n",
        "             score += reward\n",
        "\n",
        "          total_rewards.append(score)\n",
        "\n",
        "        average_reward = np.mean(total_rewards)\n",
        "        self.env.close()\n",
        "        return average_reward\n",
        "\n",
        "\n",
        "\n",
        "    def _compute_dqn_loss(self, samples: Dict[str, np.ndarray]) -> torch.Tensor:\n",
        "        \"\"\"Return dqn loss.\"\"\"\n",
        "        device = self.device  # for shortening the following lines\n",
        "        state = torch.FloatTensor(samples[\"obs\"]).to(device)\n",
        "        next_state = torch.FloatTensor(samples[\"next_obs\"]).to(device)\n",
        "        action = torch.LongTensor(samples[\"acts\"].reshape(-1, 1)).to(device)\n",
        "        reward = torch.FloatTensor(samples[\"rews\"].reshape(-1, 1)).to(device)\n",
        "        done = torch.FloatTensor(samples[\"done\"].reshape(-1, 1)).to(device)\n",
        "\n",
        "        # G_t   = r + gamma * v(s_{t+1})  if state != Terminal\n",
        "        #       = r                       otherwise\n",
        "        curr_q_value = self.dqn(state).gather(1, action)\n",
        "        next_q_value = self.dqn_target(\n",
        "            next_state\n",
        "        ).max(dim=1, keepdim=True)[0].detach()\n",
        "        mask = 1 - done\n",
        "        #if done == True then mask == 0 -> target = reward\n",
        "        target = (reward + self.gamma * next_q_value * mask).to(self.device)\n",
        "\n",
        "        # calculate dqn loss\n",
        "        loss =   F.smooth_l1_loss(curr_q_value, target)\n",
        "\n",
        "\n",
        "        return loss\n",
        "\n",
        "    def _target_hard_update(self):\n",
        "        \"\"\"Hard update: target <- local.\"\"\"\n",
        "        self.dqn_target.load_state_dict(self.dqn.state_dict())\n",
        "\n",
        "    def _plot(\n",
        "        self,\n",
        "        frame_idx: int,\n",
        "        scores: List[float],\n",
        "        losses: List[float],\n",
        "        epsilons: List[float],\n",
        "    ):\n",
        "        \"\"\"Plot the training progresses.\"\"\"\n",
        "        clear_output(True)\n",
        "        plt.figure(figsize=(20, 5))\n",
        "        plt.subplot(131)\n",
        "        plt.title('frame %s. score: %s' % (frame_idx, np.mean(scores[-10:])))\n",
        "        plt.plot(scores)\n",
        "        plt.subplot(132)\n",
        "        plt.title('loss')\n",
        "        plt.plot(losses)\n",
        "        plt.subplot(133)\n",
        "        plt.title('epsilons')\n",
        "        plt.plot(epsilons)\n",
        "        plt.show()\n",
        "\n"
      ],
      "metadata": {
        "id": "0FKZea7R8acI"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "source": [
        "### Parametri ed inizializzazioni"
      ],
      "metadata": {
        "id": "9FIxEVew8acK"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "# environment\n",
        "env = gym.make(\"CartPole-v1\", max_episode_steps=300, render_mode=\"rgb_array\")\n",
        "\n",
        "\n",
        "# parameters\n",
        "num_frames = 6_000\n",
        "memory_size = 600\n",
        "batch_size = 32\n",
        "target_update = 300\n",
        "epsilon_decay = 1 / 3_000\n",
        "\n",
        "agent = DQNAgent(env, memory_size, batch_size, target_update, epsilon_decay)"
      ],
      "metadata": {
        "id": "SKMX-4Ni8acK"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "source": [
        "### Train"
      ],
      "metadata": {
        "id": "BxK7NM9A8acK"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "agent.train(num_frames)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 253
        },
        "outputId": "b38d045b-c247-41ff-c623-d57f3c4ac41c",
        "id": "F_gX7Kqh8acL"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 2000x500 with 3 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "\n",
            "...End of training...\n",
            "\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "#### Test"
      ],
      "metadata": {
        "id": "PtT-jjGUDyfH"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "video_folder=\"videos/dqn\"\n",
        "agent.test(video_folder=video_folder)\n"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "228b7e9a-3cd0-4a3d-d4dc-ea4dffb1424a",
        "id": "k_NlHfNvDyfJ"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Moviepy - Building video /content/videos/dqn/rl-video-episode-0.mp4.\n",
            "Moviepy - Writing video /content/videos/dqn/rl-video-episode-0.mp4\n",
            "\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": []
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Moviepy - Done !\n",
            "Moviepy - video ready /content/videos/dqn/rl-video-episode-0.mp4\n",
            "score:  300.0\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "r = agent.test_avr_100()\n",
        "print(\"Average score over 100 episodes:\",r)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "wNzaZcoCQt7h",
        "outputId": "acbd6909-dd5b-44d6-977c-2447a7bb06f2"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Average score over 100 episodes: 231.48\n"
          ]
        }
      ]
    }
  ]
}