view_results.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "import glob\n",
    "import numpy as np\n",
    "import seaborn as sns\n",
    "import matplotlib.pyplot as plt"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "result_folder = \"results/20250306_104959\"\n",
    "result_files = glob.glob(f\"{result_folder}/*/results.parquet\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "['results/20250306_104959/25451fc6/results.parquet',\n",
       " 'results/20250306_104959/27d4015f/results.parquet',\n",
       " 'results/20250306_104959/821791af/results.parquet',\n",
       " 'results/20250306_104959/c75e2504/results.parquet',\n",
       " 'results/20250306_104959/8c8dd530/results.parquet',\n",
       " 'results/20250306_104959/e44c611e/results.parquet',\n",
       " 'results/20250306_104959/7ef2372d/results.parquet',\n",
       " 'results/20250306_104959/c1657eea/results.parquet',\n",
       " 'results/20250306_104959/70a71cb9/results.parquet',\n",
       " 'results/20250306_104959/d2b3b2ef/results.parquet',\n",
       " 'results/20250306_104959/efc22d7e/results.parquet',\n",
       " 'results/20250306_104959/ebf1e7dc/results.parquet',\n",
       " 'results/20250306_104959/27e0c546/results.parquet',\n",
       " 'results/20250306_104959/065b90a9/results.parquet',\n",
       " 'results/20250306_104959/53ebaf7e/results.parquet',\n",
       " 'results/20250306_104959/a27407f0/results.parquet',\n",
       " 'results/20250306_104959/9941c0cd/results.parquet',\n",
       " 'results/20250306_104959/be61a03c/results.parquet',\n",
       " 'results/20250306_104959/42de13bf/results.parquet',\n",
       " 'results/20250306_104959/7ef75c99/results.parquet',\n",
       " 'results/20250306_104959/52c6e7f9/results.parquet',\n",
       " 'results/20250306_104959/84d800d0/results.parquet',\n",
       " 'results/20250306_104959/90922583/results.parquet',\n",
       " 'results/20250306_104959/8e9e3623/results.parquet',\n",
       " 'results/20250306_104959/b8aa985c/results.parquet',\n",
       " 'results/20250306_104959/6e0bec3a/results.parquet',\n",
       " 'results/20250306_104959/64313c2c/results.parquet',\n",
       " 'results/20250306_104959/d5692129/results.parquet',\n",
       " 'results/20250306_104959/e7faf00b/results.parquet',\n",
       " 'results/20250306_104959/1c7c20ad/results.parquet',\n",
       " 'results/20250306_104959/07261884/results.parquet',\n",
       " 'results/20250306_104959/7f61a4b7/results.parquet',\n",
       " 'results/20250306_104959/b5d35c87/results.parquet',\n",
       " 'results/20250306_104959/0f03b7f3/results.parquet',\n",
       " 'results/20250306_104959/a6bb9635/results.parquet',\n",
       " 'results/20250306_104959/0bbfa9e4/results.parquet',\n",
       " 'results/20250306_104959/8cd6ade5/results.parquet',\n",
       " 'results/20250306_104959/eaa0b939/results.parquet',\n",
       " 'results/20250306_104959/7dab4a43/results.parquet',\n",
       " 'results/20250306_104959/fd6d38ed/results.parquet',\n",
       " 'results/20250306_104959/1644be24/results.parquet',\n",
       " 'results/20250306_104959/e8a2245b/results.parquet',\n",
       " 'results/20250306_104959/6b680786/results.parquet',\n",
       " 'results/20250306_104959/43d8c4ed/results.parquet',\n",
       " 'results/20250306_104959/e9972084/results.parquet',\n",
       " 'results/20250306_104959/b2aff450/results.parquet',\n",
       " 'results/20250306_104959/c8ad07a9/results.parquet',\n",
       " 'results/20250306_104959/80534b7a/results.parquet',\n",
       " 'results/20250306_104959/39d824f3/results.parquet',\n",
       " 'results/20250306_104959/e7f61ada/results.parquet']"
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "result_files"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "run_dfs = [pd.read_parquet(f) for f in result_files]\n",
    "for run_id, df in enumerate(run_dfs):\n",
    "    df['run_id'] = run_id\n",
    "results_df = pd.concat(run_dfs, ignore_index=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 600x500 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "import seaborn as sns\n",
    "import matplotlib.pyplot as plt\n",
    "\n",
    "plt.figure(figsize=(6, 5))\n",
    "\n",
    "results_by_algo_and_horizon = results_df.groupby('algo')\n",
    "for m, (algo_name, algo_results) in enumerate(results_by_algo_and_horizon):\n",
    "    if algo_name in [\"CAP\"]:\n",
    "        algo_results['regret'] = algo_results['optimal_revenue'] - algo_results['expected_revenue']\n",
    "        total_regret = np.log(algo_results.groupby(['run_id', 'T'])['regret'].sum())\n",
    "        avg_total_regret = total_regret.groupby('T').mean().reset_index()\n",
    "        std_total_regret = total_regret.groupby('T').std().reset_index()\n",
    "        plt.errorbar(np.log(avg_total_regret['T']), avg_total_regret['regret'], yerr=std_total_regret['regret'], label=\"Empirical Regret\", capsize=5, marker=\"*\")\n",
    "\n",
    "x = np.linspace(4.8, 7.8, 100)\n",
    "y = 0.5 * x + 2.3\n",
    "plt.plot(x, y, color='red', linestyle='dashed', label=r\"Theoretical Regret = $\\Theta(T^{0.5})$\")\n",
    "\n",
    "plt.xlabel('log(T)')\n",
    "plt.ylabel('log(Regret)')\n",
    "plt.legend()\n",
    "plt.grid()\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": ".venv",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.12.3"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}