notebooks/evaluation_rl_experiment.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "ef8e1119",
   "metadata": {},
   "outputs": [],
   "source": [
    "from typing import Tuple, Dict, Callable, Iterator, Union, Optional, List\n",
    "import os\n",
    "import sys\n",
    "import yaml \n",
    "\n",
    "import numpy as np\n",
    "import torch\n",
    "from torch import Tensor\n",
    "\n",
    "import gym\n",
    "\n",
    "# To import module code.\n",
    "module_path = os.path.abspath(os.path.join('..'))\n",
    "if module_path not in sys.path:\n",
    "    sys.path.append(module_path)\n",
    "from src.environment_api import EnvironmentObjective, manipulate_reward\n",
    "from src.policy_parameterizations import MLP, discretize\n",
    "from src.evaluate import (\n",
    "    offline_reward_evaluation,\n",
    "    postprocessing_interpolate_x,\n",
    "    plot_rewards_over_calls,\n",
    ")\n",
    "\n",
    "import matplotlib.pyplot as plt\n",
    "%config Completer.use_jedi = False"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "e85cc5d4",
   "metadata": {},
   "outputs": [],
   "source": [
    "plt.style.use('seaborn-whitegrid')\n",
    "tex_fonts = {\n",
    "    # Use LaTeX to write all text\n",
    "    \"text.usetex\": True,\n",
    "    \"font.family\": \"serif\",\n",
    "    # Use 10pt font in plots, to match 10pt font in document\n",
    "    \"axes.labelsize\": 9,\n",
    "    \"font.size\": 9,\n",
    "    # Make the legend/label fonts a little smaller\n",
    "    \"legend.fontsize\": 8,\n",
    "    \"xtick.labelsize\": 8,\n",
    "    \"ytick.labelsize\": 8\n",
    "}\n",
    "\n",
    "plt.rcParams.update(tex_fonts)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "ed00dbd0",
   "metadata": {},
   "outputs": [],
   "source": [
    "def plot_rewards_over_calls(\n",
    "    rewards_optimizer: List[torch.tensor],\n",
    "    names_optimizer: List[str],\n",
    "    title: str,\n",
    "    marker: List[str] = [\"o\", \">\"],\n",
    "    steps: int = 100,\n",
    "    markevery: int = 5,\n",
    "    figsize: Tuple[float] = (2.8, 1.7),\n",
    "    path_savefig: Optional[str] = None,\n",
    "):\n",
    "    \"\"\"Generate plot showing rewards over objective calls for multiple optimizer.\n",
    "\n",
    "    Args:\n",
    "        rewards_optimizer: List of torch tensors for every optimizer.\n",
    "        title: Plot title.\n",
    "        marker: Plot marker.\n",
    "        steps: Number which defines the x-th reward that should be plotted.\n",
    "        markevery: Number which defines the x-th reward which should be marked (after steps).\n",
    "        path_savefig: Path where to save the resulting figure.\n",
    "    \"\"\"\n",
    "    plt.figure(figsize=figsize)\n",
    "    for index_optimizer, rewards in enumerate(rewards_optimizer):\n",
    "        max_calls = rewards.shape[-1]\n",
    "        m = torch.mean(rewards, dim=0)[::steps]\n",
    "        std = torch.std(rewards, dim=0)[::steps]\n",
    "        plt.plot(\n",
    "            torch.linspace(0, max_calls, max_calls // steps),\n",
    "            m,\n",
    "            marker=marker[index_optimizer],\n",
    "            markevery=markevery,\n",
    "            markersize=3,\n",
    "            label=names_optimizer[index_optimizer],\n",
    "        )\n",
    "        plt.fill_between(\n",
    "            torch.linspace(0, max_calls, max_calls // steps),\n",
    "            m - std,\n",
    "            m + std,\n",
    "            alpha=0.2,\n",
    "        )\n",
    "    plt.xlabel(\"\\# of evaluations\")\n",
    "    plt.ylabel(\"Average Reward\")\n",
    "    plt.legend(loc=\"lower right\")\n",
    "    plt.title(title)\n",
    "    plt.xlim([0, max_calls])\n",
    "    if path_savefig:\n",
    "        plt.savefig(path_savefig, bbox_inches=\"tight\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "e3055daa",
   "metadata": {},
   "outputs": [],
   "source": [
    "def postprocess_data(configs: List[str], postprocess: bool = True):\n",
    "    \n",
    "    method_to_name = {'gibo': 'GIBO', 'rs': 'ARS', 'vbo': 'Vanilla BO'}\n",
    "    list_interpolated_rewards = []\n",
    "    list_names_optimizer = []\n",
    "    \n",
    "    for cfg_str in configs:\n",
    "\n",
    "        with open(cfg_str, 'r') as f:\n",
    "            cfg = yaml.load(f, Loader=yaml.Loader)\n",
    "\n",
    "        directory = '.' + cfg['out_dir']\n",
    "        \n",
    "        if postprocess:\n",
    "\n",
    "            print('Postprocess tracked parameters over optimization procedure.')\n",
    "\n",
    "            # Usecase 1: optimizing policy for a reinforcement learning environment.\n",
    "            mlp = MLP(*cfg['mlp']['layers'], add_bias=cfg['mlp']['add_bias'])\n",
    "            len_params = mlp.len_params\n",
    "\n",
    "            # In evaluation mode manipulation of state and reward is always None.\n",
    "            objective_env = EnvironmentObjective(\n",
    "                env=gym.make(cfg['environment_name']),\n",
    "                policy=mlp,\n",
    "                manipulate_state=None,\n",
    "                manipulate_reward=None,\n",
    "            )\n",
    "\n",
    "            # Load data.\n",
    "            print(f'Load data from {directory}.')\n",
    "            parameters = np.load(\n",
    "                os.path.join(directory, 'parameters.npy'), allow_pickle=True\n",
    "            ).item()\n",
    "            calls = np.load(\n",
    "                os.path.join(directory, 'calls.npy'), allow_pickle=True\n",
    "            ).item()\n",
    "\n",
    "            # Postprocess data (offline evaluation and interpolation).\n",
    "            print('Postprocess data: offline evaluation and interpolation.')\n",
    "            offline_rewards = offline_reward_evaluation(parameters, objective_env)\n",
    "            interpolated_rewards = postprocessing_interpolate_x(\n",
    "                offline_rewards, calls, max_calls=cfg['max_objective_calls']\n",
    "            )\n",
    "\n",
    "            # Save postprocessed data.\n",
    "            print(f'Save postprocessed data in {directory}.')\n",
    "            torch.save(\n",
    "                interpolated_rewards, os.path.join(directory, 'interpolated_rewards.pt')\n",
    "            )\n",
    "            torch.save(offline_rewards, os.path.join(directory, 'offline_rewards.pt'))\n",
    "\n",
    "        else:\n",
    "            interpolated_rewards = torch.load(\n",
    "                os.path.join(directory, 'interpolated_rewards.pt')\n",
    "            )\n",
    "\n",
    "        list_names_optimizer.append(method_to_name[cfg['method']])\n",
    "        list_interpolated_rewards.append(interpolated_rewards)\n",
    "        \n",
    "    return list_names_optimizer, list_interpolated_rewards"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "f401869b",
   "metadata": {},
   "outputs": [],
   "source": [
    "(list_names_optimizer, \n",
    " list_interpolated_rewards) = postprocess_data(configs=['../configs/rl_experiment/cartpole/rs_10runs.yaml',\n",
    "                                                        '../configs/rl_experiment/cartpole/gibo_10runs.yaml',\n",
    "                                                       ],\n",
    "                                               postprocess = False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "id": "8ad9c239",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 138.24x85.68 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.rcParams['lines.linewidth'] = 1.\n",
    "plot_rewards_over_calls(\n",
    "        rewards_optimizer=list_interpolated_rewards,\n",
    "        names_optimizer=list_names_optimizer,\n",
    "        title='Cartpole-v1',\n",
    "        marker=['o', '>'],\n",
    "        steps=10,\n",
    "        markevery=1,\n",
    "        figsize = (1.92, 1.19),\n",
    "        path_savefig=None #'../experiments/rl_experiments/cartpole/cartpole_v1_10runs.pdf',\n",
    "    )"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "id": "1bde3206",
   "metadata": {},
   "outputs": [],
   "source": [
    "(list_names_optimizer, \n",
    " list_interpolated_rewards) = postprocess_data(configs=['../configs/rl_experiment/swimmer/rs_10runs.yaml',\n",
    "                                                        '../configs/rl_experiment/swimmer/gibo_10runs.yaml',\n",
    "                                                       ],\n",
    "                                               postprocess = False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "id": "23cb3452",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 138.24x85.68 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.rcParams['lines.linewidth'] = 1.\n",
    "plot_rewards_over_calls(\n",
    "        rewards_optimizer=list_interpolated_rewards,\n",
    "        names_optimizer=list_names_optimizer,\n",
    "        title='Swimmer-v1',\n",
    "        marker=['o', '>'],\n",
    "        steps=50,\n",
    "        markevery=5,\n",
    "        figsize = (1.92, 1.19),\n",
    "        path_savefig=None #'../experiments/rl_experiments/swimmer/swimmer_v1_10runs.pdf',\n",
    "    )"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "id": "8cedde98",
   "metadata": {},
   "outputs": [],
   "source": [
    "(list_names_optimizer, \n",
    " list_interpolated_rewards) = postprocess_data(configs=['../configs/rl_experiment/hopper/rs_10runs.yaml',\n",
    "                                                        '../configs/rl_experiment/hopper/gibo_10runs.yaml',\n",
    "                                                       ],\n",
    "                                               postprocess = False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "id": "f9888de2",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 138.24x85.68 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.rcParams['lines.linewidth'] = 1.\n",
    "plot_rewards_over_calls(\n",
    "        rewards_optimizer=list_interpolated_rewards,\n",
    "        names_optimizer=list_names_optimizer,\n",
    "        title='Hopper-v1',\n",
    "        marker=['o', '>'],\n",
    "        steps=200,\n",
    "        markevery=5,\n",
    "        figsize = (1.92, 1.19),\n",
    "        path_savefig=None #'../experiments/rl_experiments/hopper/hopper_v1_10runs.pdf',\n",
    "    )"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "d2c5fff2",
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python [conda env:bga] *",
   "language": "python",
   "name": "conda-env-bga-py"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.8.5"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}