plot_data.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "import numpy as np\n",
    "import matplotlib.pyplot as plt\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [],
   "source": [
    "df = pd.read_csv(\"high.csv\")\n",
    "models = [\"GRUNet\", \"Neg\", \"Pack\", \"Sig\", \"Seq\", \"FCNet\" ]\n",
    "metrics = [\"L1\", \"MSE\", \"RMSE\"]\n",
    "pat = '|'.join(metrics)\n",
    "s = df['Nets'].str.extract('('+ pat + ')', expand=False)\n",
    "groups = df.groupby(s)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<Axes: title={'center': 'R2 score for high volume type'}>"
      ]
     },
     "execution_count": 40,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig, ax = plt.subplots()\n",
    "labels = df[\"Nets\"].str.replace(\"\\r\\n Trained with \", \"_\")\n",
    "r2_scores = df[\"Mean\\r\\n R2\\r\\n Score\"]\n",
    "bar_width = 0.2\n",
    "x = np.arange(len(models))*1.5\n",
    "\n",
    "\n",
    "for kk, (name, group) in enumerate(groups):\n",
    "    r2_scores = group[\"Mean\\r\\n R2\\r\\n Score\"].values\n",
    "    idx = np.argmax(r2_scores)\n",
    "    bar = ax.bar(-bar_width + x + kk*bar_width, r2_scores, width=bar_width)\n",
    "    # ax.bar_label(bar, fontsize=8, padding=4)\n",
    "#colors = [\"brown\" if val == max(r2_scores) else \"bisque\" if val == sorted(r2_scores, reverse=True)[2] else \"cornflowerblue\" for val in r2_scores]\n",
    "\n",
    "#bar = ax.bar(labels, r2_scores, align=\"edge\", width=0.3, color=colors)\n",
    "#ax.bar_label(bar, fontsize=8, padding=4)\n",
    "#plt.setp(ax.get_xticklabels(), fontsize=10, rotation='vertical')\n",
    "plt.xticks(x,models)\n",
    "ax.set_title(\"evaluation with R2 score for high volume type\")\n",
    "plt.legend(metrics)\n",
    "ax"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "4"
      ]
     },
     "execution_count": 27,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "|"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "base",
   "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": "3.10.9"
  },
  "orig_nbformat": 4
 },
 "nbformat": 4,
 "nbformat_minor": 2
}