eval_and_analyze.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "provenance": [],
      "authorship_tag": "ABX9TyMJjPaI29YCfyYewxCsG6Ps",
      "include_colab_link": true
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    },
    "language_info": {
      "name": "python"
    }
  },
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "view-in-github",
        "colab_type": "text"
      },
      "source": [
        "<a href=\"https://colab.research.google.com/github/nathanwispinski/meta-rl/blob/main/eval_and_analyze.ipynb\" target=\"_parent\"><img src=\"https://colab.research.google.com/assets/colab-badge.svg\" alt=\"Open In Colab\"/></a>"
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "# eval_and_analyze.ipynb\n",
        "\n",
        "This is a Google Colab notebook to demo model evaluation and analysis of a recurrent neural network in a two-armed bandit task using reinforcement learning.\n",
        "\n",
        "You can load your own `*.pickle` file generated from training, or use the provided `trained_agent.pickle` file.\n",
        "\n",
        "For more details, see the GitHub repository (https://github.com/nathanwispinski/meta-rl)."
      ],
      "metadata": {
        "id": "krddZGO8DjYD"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "# Colab setup"
      ],
      "metadata": {
        "id": "O7cU8rD-D6PI"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "#@title Clone GitHub repository.\n",
        "!git clone https://github.com/nathanwispinski/meta-rl"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "cellView": "form",
        "id": "pGkgcQwGD8gi",
        "outputId": "94557213-d944-4891-e7cf-88e8c7630932"
      },
      "execution_count": 1,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Cloning into 'meta-rl'...\n",
            "remote: Enumerating objects: 57, done.\u001b[K\n",
            "remote: Counting objects: 100% (57/57), done.\u001b[K\n",
            "remote: Compressing objects: 100% (44/44), done.\u001b[K\n",
            "remote: Total 57 (delta 22), reused 38 (delta 12), pack-reused 0\u001b[K\n",
            "Unpacking objects: 100% (57/57), 148.95 KiB | 556.00 KiB/s, done.\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "#@title Change working directory to cloned repository (i.e., /content/meta-rl/).\n",
        "%cd meta-rl"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "cellView": "form",
        "id": "N5bjP_1zD2Zk",
        "outputId": "f3595e75-3a1e-42c7-877c-e8febe45bda2"
      },
      "execution_count": 2,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "/content/meta-rl\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# @title Install dependencies from `requirements.txt`.\n",
        "!pip install -r requirements.txt"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 1000
        },
        "cellView": "form",
        "id": "FHd8LbgVEAXD",
        "outputId": "05bd89b4-eb43-436d-90dc-974fdf2cd27d"
      },
      "execution_count": 3,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Looking in indexes: https://pypi.org/simple, https://us-python.pkg.dev/colab-wheels/public/simple/\n",
            "Collecting absl_py==1.3.0\n",
            "  Downloading absl_py-1.3.0-py3-none-any.whl (124 kB)\n",
            "\u001b[2K     \u001b[90m━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[32m124.6/124.6 KB\u001b[0m \u001b[31m1.5 MB/s\u001b[0m eta \u001b[36m0:00:00\u001b[0m\n",
            "\u001b[?25hCollecting chex==0.1.5\n",
            "  Downloading chex-0.1.5-py3-none-any.whl (85 kB)\n",
            "\u001b[2K     \u001b[90m━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[32m85.3/85.3 KB\u001b[0m \u001b[31m4.9 MB/s\u001b[0m eta \u001b[36m0:00:00\u001b[0m\n",
            "\u001b[?25hCollecting dm_haiku==0.0.9\n",
            "  Downloading dm_haiku-0.0.9-py3-none-any.whl (352 kB)\n",
            "\u001b[2K     \u001b[90m━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[32m352.1/352.1 KB\u001b[0m \u001b[31m5.3 MB/s\u001b[0m eta \u001b[36m0:00:00\u001b[0m\n",
            "\u001b[?25hCollecting jax==0.3.25\n",
            "  Downloading jax-0.3.25.tar.gz (1.1 MB)\n",
            "\u001b[2K     \u001b[90m━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[32m1.1/1.1 MB\u001b[0m \u001b[31m16.4 MB/s\u001b[0m eta \u001b[36m0:00:00\u001b[0m\n",
            "\u001b[?25h  Preparing metadata (setup.py) ... \u001b[?25l\u001b[?25hdone\n",
            "Collecting jaxlib==0.3.25\n",
            "  Downloading jaxlib-0.3.25-cp38-cp38-manylinux2014_x86_64.whl (71.2 MB)\n",
            "\u001b[2K     \u001b[90m━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[32m71.2/71.2 MB\u001b[0m \u001b[31m9.0 MB/s\u001b[0m eta \u001b[36m0:00:00\u001b[0m\n",
            "\u001b[?25hCollecting matplotlib==3.6.2\n",
            "  Downloading matplotlib-3.6.2-cp38-cp38-manylinux_2_12_x86_64.manylinux2010_x86_64.whl (9.4 MB)\n",
            "\u001b[2K     \u001b[90m━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[32m9.4/9.4 MB\u001b[0m \u001b[31m13.7 MB/s\u001b[0m eta \u001b[36m0:00:00\u001b[0m\n",
            "\u001b[?25hCollecting ml_collections==0.1.1\n",
            "  Downloading ml_collections-0.1.1.tar.gz (77 kB)\n",
            "\u001b[2K     \u001b[90m━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[32m77.9/77.9 KB\u001b[0m \u001b[31m5.2 MB/s\u001b[0m eta \u001b[36m0:00:00\u001b[0m\n",
            "\u001b[?25h  Preparing metadata (setup.py) ... \u001b[?25l\u001b[?25hdone\n",
            "Requirement already satisfied: numpy==1.22.4 in /usr/local/lib/python3.8/dist-packages (from -r requirements.txt (line 8)) (1.22.4)\n",
            "Collecting optax==0.1.4\n",
            "  Downloading optax-0.1.4-py3-none-any.whl (154 kB)\n",
            "\u001b[2K     \u001b[90m━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[32m154.9/154.9 KB\u001b[0m \u001b[31m2.0 MB/s\u001b[0m eta \u001b[36m0:00:00\u001b[0m\n",
            "\u001b[?25hCollecting rlax==0.1.4\n",
            "  Downloading rlax-0.1.4-py3-none-any.whl (128 kB)\n",
            "\u001b[2K     \u001b[90m━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[32m128.7/128.7 KB\u001b[0m \u001b[31m6.6 MB/s\u001b[0m eta \u001b[36m0:00:00\u001b[0m\n",
            "\u001b[?25hCollecting scipy==1.9.3\n",
            "  Downloading scipy-1.9.3-cp38-cp38-manylinux_2_17_x86_64.manylinux2014_x86_64.whl (33.8 MB)\n",
            "\u001b[2K     \u001b[90m━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[32m33.8/33.8 MB\u001b[0m \u001b[31m7.4 MB/s\u001b[0m eta \u001b[36m0:00:00\u001b[0m\n",
            "\u001b[?25hRequirement already satisfied: dm-tree>=0.1.5 in /usr/local/lib/python3.8/dist-packages (from chex==0.1.5->-r requirements.txt (line 2)) (0.1.8)\n",
            "Requirement already satisfied: toolz>=0.9.0 in /usr/local/lib/python3.8/dist-packages (from chex==0.1.5->-r requirements.txt (line 2)) (0.12.0)\n",
            "Requirement already satisfied: tabulate>=0.8.9 in /usr/local/lib/python3.8/dist-packages (from dm_haiku==0.0.9->-r requirements.txt (line 3)) (0.8.10)\n",
            "Collecting jmp>=0.0.2\n",
            "  Downloading jmp-0.0.4-py3-none-any.whl (18 kB)\n",
            "Requirement already satisfied: opt_einsum in /usr/local/lib/python3.8/dist-packages (from jax==0.3.25->-r requirements.txt (line 4)) (3.3.0)\n",
            "Requirement already satisfied: typing_extensions in /usr/local/lib/python3.8/dist-packages (from jax==0.3.25->-r requirements.txt (line 4)) (4.5.0)\n",
            "Requirement already satisfied: pillow>=6.2.0 in /usr/local/lib/python3.8/dist-packages (from matplotlib==3.6.2->-r requirements.txt (line 6)) (8.4.0)\n",
            "Requirement already satisfied: cycler>=0.10 in /usr/local/lib/python3.8/dist-packages (from matplotlib==3.6.2->-r requirements.txt (line 6)) (0.11.0)\n",
            "Requirement already satisfied: python-dateutil>=2.7 in /usr/local/lib/python3.8/dist-packages (from matplotlib==3.6.2->-r requirements.txt (line 6)) (2.8.2)\n",
            "Requirement already satisfied: packaging>=20.0 in /usr/local/lib/python3.8/dist-packages (from matplotlib==3.6.2->-r requirements.txt (line 6)) (23.0)\n",
            "Requirement already satisfied: kiwisolver>=1.0.1 in /usr/local/lib/python3.8/dist-packages (from matplotlib==3.6.2->-r requirements.txt (line 6)) (1.4.4)\n",
            "Requirement already satisfied: fonttools>=4.22.0 in /usr/local/lib/python3.8/dist-packages (from matplotlib==3.6.2->-r requirements.txt (line 6)) (4.38.0)\n",
            "Requirement already satisfied: pyparsing>=2.2.1 in /usr/local/lib/python3.8/dist-packages (from matplotlib==3.6.2->-r requirements.txt (line 6)) (3.0.9)\n",
            "Collecting contourpy>=1.0.1\n",
            "  Downloading contourpy-1.0.7-cp38-cp38-manylinux_2_17_x86_64.manylinux2014_x86_64.whl (300 kB)\n",
            "\u001b[2K     \u001b[90m━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[32m300.0/300.0 KB\u001b[0m \u001b[31m5.0 MB/s\u001b[0m eta \u001b[36m0:00:00\u001b[0m\n",
            "\u001b[?25hRequirement already satisfied: PyYAML in /usr/local/lib/python3.8/dist-packages (from ml_collections==0.1.1->-r requirements.txt (line 7)) (6.0)\n",
            "Requirement already satisfied: six in /usr/local/lib/python3.8/dist-packages (from ml_collections==0.1.1->-r requirements.txt (line 7)) (1.15.0)\n",
            "Requirement already satisfied: contextlib2 in /usr/local/lib/python3.8/dist-packages (from ml_collections==0.1.1->-r requirements.txt (line 7)) (0.5.5)\n",
            "Collecting distrax>=0.0.2\n",
            "  Downloading distrax-0.1.3-py3-none-any.whl (317 kB)\n",
            "\u001b[2K     \u001b[90m━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[32m318.0/318.0 KB\u001b[0m \u001b[31m5.3 MB/s\u001b[0m eta \u001b[36m0:00:00\u001b[0m\n",
            "\u001b[?25hCollecting dm-env\n",
            "  Downloading dm_env-1.6-py3-none-any.whl (26 kB)\n",
            "Requirement already satisfied: tensorflow-probability>=0.15.0 in /usr/local/lib/python3.8/dist-packages (from distrax>=0.0.2->rlax==0.1.4->-r requirements.txt (line 10)) (0.19.0)\n",
            "Requirement already satisfied: cloudpickle>=1.3 in /usr/local/lib/python3.8/dist-packages (from tensorflow-probability>=0.15.0->distrax>=0.0.2->rlax==0.1.4->-r requirements.txt (line 10)) (2.2.1)\n",
            "Requirement already satisfied: decorator in /usr/local/lib/python3.8/dist-packages (from tensorflow-probability>=0.15.0->distrax>=0.0.2->rlax==0.1.4->-r requirements.txt (line 10)) (4.4.2)\n",
            "Requirement already satisfied: gast>=0.3.2 in /usr/local/lib/python3.8/dist-packages (from tensorflow-probability>=0.15.0->distrax>=0.0.2->rlax==0.1.4->-r requirements.txt (line 10)) (0.4.0)\n",
            "Building wheels for collected packages: jax, ml_collections\n",
            "  Building wheel for jax (setup.py) ... \u001b[?25l\u001b[?25hdone\n",
            "  Created wheel for jax: filename=jax-0.3.25-py3-none-any.whl size=1308510 sha256=e00885a564a877034ca4197abcc85b29769dc867c4c7a1f9135e4f04ec723f9b\n",
            "  Stored in directory: /root/.cache/pip/wheels/ac/cc/1f/327114e0ec22ee90b4db16c01e1fc843883e18501098a7bbec\n",
            "  Building wheel for ml_collections (setup.py) ... \u001b[?25l\u001b[?25hdone\n",
            "  Created wheel for ml_collections: filename=ml_collections-0.1.1-py3-none-any.whl size=94524 sha256=941f6fdcab2e8428f47f67b4453782c2d15c506b31d3ce15844f9eea500b2977\n",
            "  Stored in directory: /root/.cache/pip/wheels/6d/9f/a9/9e8309035a5bf09ed9086bbca8c9b74cb6413d3eb203e2bc8c\n",
            "Successfully built jax ml_collections\n",
            "Installing collected packages: scipy, jmp, contourpy, absl_py, ml_collections, matplotlib, jaxlib, jax, dm_haiku, dm-env, chex, optax, distrax, rlax\n",
            "  Attempting uninstall: scipy\n",
            "    Found existing installation: scipy 1.7.3\n",
            "    Uninstalling scipy-1.7.3:\n",
            "      Successfully uninstalled scipy-1.7.3\n",
            "  Attempting uninstall: absl_py\n",
            "    Found existing installation: absl-py 1.4.0\n",
            "    Uninstalling absl-py-1.4.0:\n",
            "      Successfully uninstalled absl-py-1.4.0\n",
            "  Attempting uninstall: matplotlib\n",
            "    Found existing installation: matplotlib 3.5.3\n",
            "    Uninstalling matplotlib-3.5.3:\n",
            "      Successfully uninstalled matplotlib-3.5.3\n",
            "  Attempting uninstall: jaxlib\n",
            "    Found existing installation: jaxlib 0.4.4+cuda11.cudnn82\n",
            "    Uninstalling jaxlib-0.4.4+cuda11.cudnn82:\n",
            "      Successfully uninstalled jaxlib-0.4.4+cuda11.cudnn82\n",
            "  Attempting uninstall: jax\n",
            "    Found existing installation: jax 0.4.4\n",
            "    Uninstalling jax-0.4.4:\n",
            "      Successfully uninstalled jax-0.4.4\n",
            "Successfully installed absl_py-1.3.0 chex-0.1.5 contourpy-1.0.7 distrax-0.1.3 dm-env-1.6 dm_haiku-0.0.9 jax-0.3.25 jaxlib-0.3.25 jmp-0.0.4 matplotlib-3.6.2 ml_collections-0.1.1 optax-0.1.4 rlax-0.1.4 scipy-1.9.3\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "application/vnd.colab-display-data+json": {
              "pip_warning": {
                "packages": [
                  "matplotlib",
                  "mpl_toolkits"
                ]
              }
            }
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "# Import dependencies"
      ],
      "metadata": {
        "id": "F7OsfWxvEBxf"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "#@title Import dependencies after install.\n",
        "import pickle\n",
        "import numpy as np\n",
        "import matplotlib\n",
        "import matplotlib.pyplot as plt\n",
        "import matplotlib.patches as mpatches\n",
        "\n",
        "import modules.envs as envs\n",
        "import modules.agents as agents"
      ],
      "metadata": {
        "cellView": "form",
        "id": "Gg37yqjjED9-"
      },
      "execution_count": 4,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "source": [
        "# Load trained agent"
      ],
      "metadata": {
        "id": "6Sy06wEdEJnK"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "#@title Load trained agent (your own file, or the provided `trained_agent.pickle` file).\n",
        "path = \"./\" #@param {type:\"string\"}\n",
        "filename = \"trained_agent.pickle\" #@param {type:\"string\"}"
      ],
      "metadata": {
        "cellView": "form",
        "id": "FVsAq1bUEFOt"
      },
      "execution_count": 5,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "#@title Load trained agent with pickle.\n",
        "with open(path + filename, 'rb') as fp:\n",
        "    training_results = pickle.load(fp)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "cellView": "form",
        "id": "0UKJH4q8EY28",
        "outputId": "368b2644-f159-4f36-a964-c1b5a86498a7"
      },
      "execution_count": 6,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "WARNING:jax._src.lib.xla_bridge:No GPU/TPU found, falling back to CPU. (Set TF_CPP_MIN_LOG_LEVEL=0 and rerun for more info.)\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "#@title Unpack pickle file.\n",
        "params = training_results[0]['params']\n",
        "training_config = training_results[0]['config']\n",
        "agent_config = training_config['agent']\n"
      ],
      "metadata": {
        "cellView": "form",
        "id": "JewwbHcgEqDF"
      },
      "execution_count": 7,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "source": [
        "# Evaluation"
      ],
      "metadata": {
        "id": "yq64NhM3Et3c"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "#@title Initialize evaluation environment.\n",
        "env = envs.BanditEnv(\n",
        "    steps_per_episode=100,\n",
        "    num_arms=2,\n",
        "    reward_structure=\"correlated\",\n",
        "    )\n",
        "observation = env.reset()"
      ],
      "metadata": {
        "cellView": "form",
        "id": "WAV5KlsNEu89"
      },
      "execution_count": 8,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "#@title Initialize agent.\n",
        "agent = agents.create_agent(\n",
        "    agent_config=agent_config,\n",
        "    observation=observation,\n",
        "    num_actions=env.num_actions,\n",
        ")"
      ],
      "metadata": {
        "cellView": "form",
        "id": "AYlckKiKE5EL"
      },
      "execution_count": 9,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "#@title Setup initial LSTM memory state.\n",
        "initial_lstm_state = agent.get_initial_lstm_state()\n",
        "lstm_state = initial_lstm_state"
      ],
      "metadata": {
        "cellView": "form",
        "id": "_TtvzDmHE8E9"
      },
      "execution_count": 10,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "#@title # Overwrite agent paramters with saved training parameters.\n",
        "agent.params = params"
      ],
      "metadata": {
        "cellView": "form",
        "id": "b0Lpus7IFAje"
      },
      "execution_count": 11,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "#@title Setup evaluation episodes to test the agent on.\n",
        "\n",
        "# Make evaluation episodes\n",
        "arm1_probs = np.arange(101)/100\n",
        "arm2_probs = 1. - arm1_probs\n",
        "\n",
        "eval_episodes = np.arange(len(arm1_probs))\n",
        "\n",
        "# Initialize data to save\n",
        "eval_actions = np.zeros((len(arm1_probs), env._steps_per_episode))\n",
        "eval_actions[:] = np.nan"
      ],
      "metadata": {
        "cellView": "form",
        "id": "F5JqrjGOFEyG"
      },
      "execution_count": 12,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "#@title Evaluation loop.\n",
        "for (arm1_prob, arm2_prob, ep) in zip(arm1_probs, arm2_probs, eval_episodes):\n",
        "    \n",
        "    step = 0\n",
        "    done = False\n",
        "    lstm_state = initial_lstm_state\n",
        "    observation = env.reset(arm_probs=np.array([arm1_prob, arm2_prob]))\n",
        "    \n",
        "    while not done:\n",
        "\n",
        "        action, pi_out, v_out, new_lstm_state, _ = agent.get_action(observation, lstm_state)\n",
        "        next_observation, reward, done, info = env.step(action)\n",
        "\n",
        "        # Save data\n",
        "        eval_actions[ep, step] = action\n",
        "\n",
        "        observation = next_observation\n",
        "        lstm_state = new_lstm_state\n",
        "        step += 1\n",
        "\n",
        "print(\"Done evaluation.\")"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "cellView": "form",
        "id": "-ggLxtaOFNXP",
        "outputId": "ec30fa6e-2016-4dc1-829f-c8780a5f0c8a"
      },
      "execution_count": 13,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Done evaluation.\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "# Analysis"
      ],
      "metadata": {
        "id": "xs8QWVkmFSJ2"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "#@title Figure parameters.\n",
        "tick_fontsize = 16\n",
        "label_fontsize = 20"
      ],
      "metadata": {
        "cellView": "form",
        "id": "Bw6XAfFEFTFp"
      },
      "execution_count": 14,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "#@title Setup y tick labels to convey arm win probabilities.\n",
        "ytick_vals = [0, eval_episodes[int(len(eval_episodes)/2)], eval_episodes[-1]]\n",
        "ytick_labels = []\n",
        "for i in ytick_vals:\n",
        "    cur_probs = str(int(arm1_probs[i]*100)) + \", \" + str(int(arm2_probs[i]*100))\n",
        "    ytick_labels.append(cur_probs)"
      ],
      "metadata": {
        "cellView": "form",
        "id": "oGYI4jVxFYps"
      },
      "execution_count": 15,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "#@title Setup custom legend.\n",
        "cmap_name = 'tab20c'\n",
        "cmap = matplotlib.cm.get_cmap(cmap_name)\n",
        "zero_color = cmap(0.)\n",
        "one_color = cmap(1.)\n",
        "patch1 = mpatches.Patch(color=zero_color, label='Arm 1')\n",
        "patch2 = mpatches.Patch(color=one_color, label='Arm 2')"
      ],
      "metadata": {
        "cellView": "form",
        "id": "VMtKTmqiFcm8"
      },
      "execution_count": 16,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "#@title Define plotting function.\n",
        "\n",
        "def plot_action_results(data):\n",
        "    fig = plt.figure(figsize=(8, 6))\n",
        "    ax = plt.subplot()\n",
        "    plt.imshow(data, cmap='tab20c', origin='lower')\n",
        "    plt.yticks(\n",
        "        ticks=ytick_vals,\n",
        "        labels=ytick_labels,\n",
        "        fontsize=tick_fontsize,\n",
        "        )\n",
        "    plt.xticks(\n",
        "        ticks=[0, env._steps_per_episode],\n",
        "        labels=[0, env._steps_per_episode],\n",
        "        fontsize=tick_fontsize,\n",
        "        )\n",
        "    plt.xlabel(\"Trials\", fontsize=label_fontsize)\n",
        "    plt.ylabel(\"Arm win probabilities \\n (Arm 1 %, Arm 2 %)\", fontsize=label_fontsize)\n",
        "    ax.spines[['right', 'top']].set_visible(False)\n",
        "    plt.legend(handles=[patch1, patch2],\n",
        "              loc='center left',\n",
        "              bbox_to_anchor=(1, 0.5),\n",
        "              frameon=False,\n",
        "              fontsize=label_fontsize,\n",
        "              borderaxespad=0.,\n",
        "              )\n",
        "    plt.tight_layout()\n",
        "    plt.show()"
      ],
      "metadata": {
        "cellView": "form",
        "id": "4e7MoKqCHJ2x"
      },
      "execution_count": 17,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "#@title Plot action results from our trained reinforcement learning model.\n",
        "plot_action_results(eval_actions)\n"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 441
        },
        "cellView": "form",
        "id": "tuNHnoCBFgBM",
        "outputId": "70fef2a7-2c3c-46ec-d8de-d9cc131bc5b0"
      },
      "execution_count": 18,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 576x432 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "# Other models\n",
        "\n",
        "We can also plot a completely random agent, and an agent with insider knowledge about the arm win probabilities, for completeness."
      ],
      "metadata": {
        "id": "IwWwHrCsHkPc"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "#@title Plot action results from a random model.\n",
        "\n",
        "random_actions = np.random.randint(low=0, high=2, size=np.shape(eval_actions))\n",
        "\n",
        "plot_action_results(random_actions)\n"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 441
        },
        "cellView": "form",
        "id": "OssoysiqFx5P",
        "outputId": "e5a9a666-8d06-442a-d0f9-2f031a70a2ff"
      },
      "execution_count": 19,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 576x432 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "#@title Plot action results from a model with insider knowledge.\n",
        "\n",
        "# Initialize data\n",
        "insider_actions = np.zeros((len(arm1_probs), env._steps_per_episode))\n",
        "insider_actions[:] = np.nan\n",
        "\n",
        "for (arm1_prob, arm2_prob, ep) in zip(arm1_probs, arm2_probs, eval_episodes):\n",
        "    if arm1_prob > arm2_prob:\n",
        "        actions = np.zeros(env._steps_per_episode)\n",
        "    elif arm1_prob == arm2_prob:\n",
        "        actions = np.random.randint(low=0, high=2, size=env._steps_per_episode)\n",
        "    else:\n",
        "        actions = np.ones(env._steps_per_episode)\n",
        "    insider_actions[ep, :] = actions\n",
        "\n",
        "plot_action_results(insider_actions)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 441
        },
        "cellView": "form",
        "id": "K-Ybz0ITHyHU",
        "outputId": "65f79538-37b6-46e8-eedf-c172575249a7"
      },
      "execution_count": 20,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 576x432 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ]
    }
  ]
}