results-analysis.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"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "df_server_gai = pd.read_csv('experiment_logs/gaikube/server_stats_msgke_20240630_115907.csv')\n",
    "df_server_vpa = pd.read_csv('experiment_logs/vpa_gke/server_stats_msgke_20240630_144821.csv')\n",
    "df_server_ms = pd.read_csv('experiment_logs/ms_gke/server_stats_msgke_20240630_160859.csv')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>num_moves</th>\n",
       "      <th>num_scalings</th>\n",
       "      <th>cost</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>4.5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>4.5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>4.5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>4.5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>4.5</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   num_moves  num_scalings  cost\n",
       "0          0             0   4.5\n",
       "1          2             0   4.5\n",
       "2          2             0   4.5\n",
       "3          1             0   4.5\n",
       "4          1             0   4.5"
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_server_ms.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 142,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 600x400 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.figure(figsize=(6,4))\n",
    "plt.plot(df_server_gai.index, df_server_gai['cost'], label='GAIKube')\n",
    "plt.plot(df_server_vpa.index, df_server_vpa['cost'], label='SVG')\n",
    "plt.plot(df_server_ms.index, df_server_ms['cost'], label='MVG')\n",
    "\n",
    "#plt.title('Cost')\n",
    "plt.xlabel('Iterations')\n",
    "plt.ylabel('Cost')\n",
    "plt.legend()\n",
    "plt.grid(True)\n",
    "plt.ylim([0,10])\n",
    "plt.savefig('cost.pdf', format='pdf')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 141,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 400x300 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.figure(figsize=(4,3))\n",
    "plt.plot(df_server_gai.index, df_server_gai['num_scalings'], label='GAIKube')\n",
    "plt.plot(df_server_ms.index, df_server_ms['num_scalings'], label='MVG')\n",
    "plt.plot(df_server_vpa.index, df_server_vpa['num_scalings'], label='SVG')\n",
    "#plt.title('num_scalings')\n",
    "plt.xlabel('Number of Requests')\n",
    "plt.ylabel('Number of Scalings')\n",
    "plt.legend()\n",
    "plt.grid(True)\n",
    "plt.ylim([0,5])\n",
    "plt.savefig('cost.pdf', format='pdf')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 140,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAgsAAAF4CAYAAADaAbo8AAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjcuMiwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy8pXeV/AAAACXBIWXMAAA9hAAAPYQGoP6dpAABcnklEQVR4nO3deXhTZdo/8O/J2qZJulFaCt3YFVmURYEZFFkcmHFDxcENAX1nBF8Ffo7v4AaMM8K4g/vC4jiDG6PO6LhVZFERhlVQEYW2tCylpVvapEmznN8fISGlbZq0Jz0nyfdzXb2uNjk9udOTNPd5znM/tyCKoggiIiKiNqjkDoCIiIiUjckCERERBcVkgYiIiIJiskBERERBMVkgIiKioJgsEBERUVBMFoiIiCgoJgtEREQUFJMFIiIiCorJAhEREQUla7KwZMkSCILQ7GvgwIFyhkRERERn0cgdwKBBg/D555/7f9ZoZA+JiIiIAsj+yazRaJCVlSV3GERERNQG2ZOFn3/+GdnZ2UhISMDo0aOxbNky5Obmtrqtw+GAw+Hw/+zxeFBdXY309HQIgtBVIRMREUU9URRRX1+P7OxsqFTBZyUIcrao/vjjj9HQ0IABAwbgxIkTWLp0KY4dO4bvvvsOJpOpxfZLlizB0qVLZYiUiIgoNpWVlaFXr15Bt5E1WThbbW0t8vLy8OSTT2LOnDkt7j97ZKGurg65ubkoLi5uNbnoCKfTiY0bN2L8+PHQarWS7JM6hsdCGXgclIHHQTli5VjU19ejoKAAtbW1SE5ODrqt7JchAqWkpKB///44dOhQq/fr9Xro9foWt6elpcFsNksSg9PphMFgQHp6elS/CGIBj4Uy8DgoA4+DcsTKsfDFHsplfEWts9DQ0IDDhw+jR48ecodCREREp8maLNxzzz3YvHkzSkpKsHXrVlx99dVQq9WYMWOGnGERERFRAFkvQxw9ehQzZsxAVVUVMjIy8Itf/ALbtm1DRkaGnGERERFRAFmThTfffFPOhyciog5yu91wOp1yhyELp9MJjUYDu90Ot9stdzhtUqvV0Gg0kiwtoKgJjkREpHwNDQ04evQoFFRM16VEUURWVhbKysoUv8aPwWBAjx49oNPpOrUfJgtERBQyt9uNo0ePwmAwICMjQ/EflpHg8XjQ0NAAo9HY7mJGchFFEU1NTaisrERxcTH69evXqViZLBARUcicTidEUURGRgYSExPlDkcWHo8HTU1NSEhIUGyyAACJiYnQarU4cuSIP96OUu6zJCIixYrHEYVoJFUyw2SBiIiIgmKyQEREREExWSAiIlKAJUuWYNiwYXKH0SomC0REFDfKy8tx9913o2/fvkhISEBmZibGjh2LF154ATabrdm2y5Ytg1qtxmOPPdZiP+vWrUNaWpr/57Vr1yIlJaXZNgcOHEBOTg6uu+46NDU1ReT5dBUmC0REFBeKiopw/vnn47PPPsMjjzyCPXv24JtvvsG9996LDz/8EJ9//nmz7VevXo17770Xq1evDvuxduzYgV/+8pf41a9+hbfeeqvT6xzIjckCERF1mCiKsDW5ZPkKd1GouXPnQqPRYOfOnZg+fTrOOecc9O7dG1deeSX+85//4PLLL/dvu3nzZjQ2NuJPf/oTLBYLtm7dGvLjfPHFF7j00ksxZ84cvPLKK1CpVK2OPLz//vutVpW89NJLyMnJgcFgwPTp01FXV9fs/ldffRXnnHMOEhISMHDgQDz//PNh/R06gussEBFRhzU63Tj3oU9leewf/nQZDLrQPsaqqqr8IwpJSUmtbhP4wb1q1SrMmDEDWq0WM2bMwKpVqzBmzJh2H+e9997DDTfcgCVLluD//u//QnsiAQ4dOoS3334bH3zwASwWC+bMmYO5c+fiH//4BwDgH//4Bx566CE8++yzOP/887Fnzx7cfvvtSEpKwsyZM8N+vFBxZIGIiGLeoUOHIIoiBgwY0Oz2bt26wWg0wmg0+j/cLRYL1q9fj5tuugkAcNNNN+Htt99GQ0ND0MdoaGjAddddhz/84Q8dShQAwG63429/+xuGDRuGcePG4ZlnnsGbb76J8vJyAMDixYvxxBNPYNq0aSgoKMC0adOwYMECvPTSSx16vFBxZIGIiDosUavGD3+6TLbH7qz//ve/8Hg8uPHGG+FwOAAAb7zxBvr06YOhQ4cCAIYNG4a8vDy89dZbmDNnTtvxJCbiF7/4BV555RXMmDED55xzTtjx5ObmomfPnv6fR48eDY/Hg4MHD8JkMuHw4cOYM2cObr/9dv82LpcLycnJYT9WOJgsEBFRhwmCEPKlADn17dsXgiDg4MGDzW7v3bs3ADRbunrVqlX4/vvvodGceV4ejwerV68Omiyo1Wq8//77mDZtGsaPH4+NGzf6EwaVStVijkW4XTt9IxuvvPIKLrzwwhaPHUm8DEFERDEvPT0dkyZNwrPPPgur1drmdvv378fOnTuxadMm7N271/+1adMmfPPNN/jxxx+DPo5er8e7776LkSNHYvz48fjhhx8AABkZGaivr2/22Hv37m3x+6WlpTh+/Lj/523btkGlUmHAgAHIzMxEdnY2ioqK0Ldv32ZfBQUFYf5FwqP8dJCIiEgCzz//PMaOHYsRI0ZgyZIlGDJkCFQqFXbs2IEff/wRw4cPx6pVqzBq1CiMGzeuxe+PHDkSq1atwl//+tegj6PX6/HPf/4T1113HcaPH48vvvgCF154IQwGA+677z7cdddd2L59O9auXdvidxMSEjBz5kw8/vjjsFgsuOuuuzB9+nRkZWUBAJYuXYq77roLycnJ+NWvfgWHw4GdO3eipqYGCxculOTv1BqOLBARUVzo06cP9uzZg4kTJ2LRokUYOnQoRowYgWeeeQb33HMPFi9ejL///e+45pprWv39a665Bn/7299Cunyg0+mwfv16jBkzBuPHj8fx48fx97//HR999BEGDx6MN954A0uWLGnxe3379sW0adMwdepUTJ48GUOGDGlWGnnbbbfh1VdfxZo1azB48GBcfPHFWLt2bcRHFgQx3EJVBbFYLEhOTkZdXR3MZrMk+3Q6nfjoo48wdepUaLVaSfZJHcNjoQw8DsqglONgt9tRXFyMgoKCTrU8jmYejwcWiwVms1nRLaqB4McrnM9QZT9LIiIikh2TBSIiIgqKyQIREREFxWSBiIiIgmKyQEREREExWSAiIqKgmCwQERFRUEwWiIiIKCgmC0RERBQUkwUiIiIKiskCERHFhcrKStxxxx3Izc2FXq9HVlYWLrvsMmzevBndunXD8uXLW/29hx9+GJmZmf6eEE1NTVi5ciVGjBiBpKQkJCcnY+jQoXjggQeadYyMJUwWiIgoLlxzzTXYs2cPXnvtNfz000/497//jUsuuQR1dXW46aabsGbNmha/I4oi1q5di1tuuQVarRYOhwOXXXYZnnzyScycORNbtmzB/v37sXLlSpw6dQrPPPOMDM8s8tiimoiIOk4UAadNnsfWGgBBCGnT2tpafPnll9i0aRMuvvhiAEBeXh5GjRoFACgoKMCKFSvw1Vdf4Re/+IX/9zZv3oyioiLMmTMHAPDUU0/hq6++wsaNG/GLX/zC30gqNzcXF198MaK4N2NQTBaIiKjjnDbgkWx5Hvu+44AuKaRNjUYjjEYj3n//fVx00UXQ6/XN7h88eDBGjhyJ1atXN0sW1qxZgzFjxmDgwIEAgDfeeAMTJ07EkCFDWn0cIcTkJdrwMgQREcU8jUaDtWvX4rXXXkNKSgrGjh2L++67D/v27fNvM2fOHLzzzjtoaGgAANTX12P9+vWYPXu2f5uffvoJAwYMaLbvq6++2p+MjBkzpmueUBfjyAIREXWc1uA9w5frscNwzTXX4Ne//jW+/PJLbNu2DR9//DEeffRRvPrqq7j11lsxY8YMLFiwAG+//TZmz56Nt956CyqVCtdff33Q/T7//POwWq1YuXIltmzZ0plnpFgcWSAioo4TBO+lADm+OjDkn5CQgEmTJuHBBx/E1q1bceutt2Lx4sUAALPZjGuvvdY/0XHNmjWYPn06jEaj//f79euHgwcPNttnjx490LdvX6SlpXXiD6lsTBaIiChunXvuubBarf6f58yZg6+++goffvghtm7d6p/Y6DNjxgx8/vnnzS5fxANehiAiophXVVWF6667DrNnz8aQIUNgMpmwc+dOPProo7jyyiv9240bNw59+/bFLbfcgoEDB7aYg7BgwQL85z//wZVXXonFixdj3LhxSE1NxU8//YSPP/4YarW6q59al2CyQEREMc9oNOLCCy/EU089hcOHD8PpdCInJwe333477rvvPv92giBg9uzZuO+++7Bo0aIW+0lISEBhYSH++te/Yu3atbjvvvvg8XhQUFCAKVOmYMGCBV35tLoMkwUiIop5er0ey5Ytw7Jly9rddtGiRa0mCoH7mj9/Ph566CH/OguxLj6eJREREXUYkwUiIiIKiskCERERBcVkgYiIiIJiskBERERBMVkgIiKioJgsEBERUVBMFoiIiCgoJgtEREQUFJMFIiIiCorJAhERxbxbb70VgiDg97//fYv75s2bB0EQcOutt+Lyyy/Hr371q1b38eWXX0IQhGYdJ//5z3/i0ksvRWpqKhITEzFgwADMnj0be/bsidhzkQOTBSIiigs5OTl488030djY6L/Nbrdj3bp1yM3NBeBtUV1YWIijR4+2+P01a9ZgxIgRGDJkCADgj3/8I66//noMGzYM//73v3Hw4EGsW7cOvXv3DtpbIhopppHU8uXLsWjRItx99914+umn5Q6HiIhCIIoiGl2N7W8YAYmaRAiCEPL2F1xwAQ4fPox3330XN954IwDg3XffRW5uLgoKCgAAv/nNb5CRkYG1a9figQce8P9uQ0MD3nnnHTz22GMAgB07duCxxx7DihUrcNddd/m3y83NxfDhwyGKohRPUTEUkSzs2LEDL730kj9bIyKi6NDoasSF6y6U5bG337AdBq0hrN+ZPXs21qxZ408WVq9ejVmzZmHTpk0AAI1Gg1tuuQVr167F/fff709G3nnnHbjdbsyYMQOA9/KD0WjE3LlzW32ccJKYaCB7stDQ0IAbb7wRr7zyCv785z8H3dbhcMDhcPh/tlgsAACn0wmn0ylJPL79SLU/6jgeC2XgcVAGpRwHp9MJURTh8Xj8X3IJ5/FFUYQoirjhhhuwaNEiFBcXAwC+/vprrFu3Dhs3bvQ/r1tvvRWPPfYYNm7ciEsuuQSA9xLEtGnTYDKZIIoiDh06hN69e0OlUvljeOqpp7B48WL/Y5aVlSE5OVnaJx0mj8cDURThdDqhVqub3RfOa0n2ZGHevHn49a9/jYkTJ7abLCxbtgxLly5tcftnn30GgyG87LI9hYWFku6POo7HQhl4HJRB7uOg0WiQlZWFhoYGNDU1QRRFfPbrz2SJxWlzwiJYQtvW6YTL5YJer8fkyZPx8ssvQxRFTJ48GTqdDi6XC06nExaLBdnZ2Rg1ahRefvllXHDBBSgqKsKXX36JDz74wH+SCgBut7vZz9deey3Gjx+PnTt34ne/+x0sFovsIwxNTU1obGzEli1b4HK5mt1ns9lC3o+sycKbb76J3bt3Y8eOHSFtv2jRIixcuND/s8ViQU5ODiZPngyz2SxJTE6nE4WFhZg0aRK0Wq0k+6SO4bFQBh4HZVDKcbDb7SgrK4PRaERCQgIAIBnynj2HQqvVQqPRwGw24/bbb/fPM3jmmWdgNpuh0Wig1Wr9nyW333477r77brz00ktYv349+vTpgylTpkAQBIiiiD59+mD79u1ITEz0Hw+z2YycnBzU1tYCAEwmk2SfTR1lt9uRmJiIcePG+Y+XT2Ci0x7ZkoWysjLcfffdKCwsbPEE2qLX66HX61vcrtVqJX/zRGKf1DE8FsrA46AMch8Ht9sNQRCgUqmgUkVPQZ0gCP64p06dit///vcQBAFTpkyBSqVqdj8A/Pa3v8WCBQvw5ptv4vXXX8cdd9zhH8b3eDy45ppr8PLLL+PFF1/E3Xff3eyxfPtQwt/I99xae92E8zqSLVnYtWsXKioqcMEFF/hvc7vd2LJlC5599lk4HI4W11eIiIg6S61W48CBA/7vW2M0GnH99ddj0aJFsFgsuPXWW5vdP2rUKCxcuBD/7//9Pxw5cgTTpk1DTk4OTpw4gVWrVjVLPGKBbM9kwoQJ2L9/P/bu3ev/GjFiBG688Ubs3buXiQIREUWM2Wxu9xLBnDlzUFNTg8suuwzZ2dkt7n/sscewbt067NmzB7/5zW/Qr18/XHfddfB4PPjmm29kvwQhJdlGFkwmE84777xmtyUlJSE9Pb3F7URERJ2xdu3aoPe///77LW4bPXp0u+slTJ8+HdOnT+9EZNEhdsZIiIiIKCJkL50M5FsUg4iIiJSDIwtEREQUFJMFIiIiCorJAhERhS3WGiXFKqmOE5MFIiIKma+svampSeZIKBS+JZ07u5CXoiY4EhGRsmk0GhgMBlRWVkKr1cbUwkOh8ng8aGpqgt1uV+zzF0URNpsNFRUVSElJ6fTaRUwWiIgoZIIgoEePHiguLsaRI0fkDkcWoiiisbERiYmJsjeKak9KSgqysrI6vR8mC0REFBadTod+/frF7aUIp9OJLVu2YNy4cYrul6LVaiVbDZnJAhERhU2lUoXcBDDWqNVquFwuJCQkKDpZkJIyL7YQERGRYjBZICIioqCYLBAREVFQTBaIiIgoKCYLREREFBSTBSIiIgqKyQIREREFxWSBiIiIgmKyQEREREExWSAiIqKgmCwQERFRUEwWiIiIKCgmC0RERBQUkwUiIiIKiskCERERBcVkgYiIiIJiskBERERBMVkgIiKioJgsEBERUVBhJwulpaUQRbHF7aIoorS0VJKgiIiISDnCThYKCgpQWVnZ4vbq6moUFBRIEhQREREpR9jJgiiKEAShxe0NDQ1ISEiQJCgiIiJSDk2oGy5cuBAAIAgCHnzwQRgMBv99brcb27dvx7BhwyQPkIiIiOQVcrKwZ88eAN6Rhf3790On0/nv0+l0GDp0KO655x7pIyQiIiJZhZwsbNy4EQAwa9YsrFixAmazOWJBERERkXKEnCz4rFmzJhJxEBERkUKFnSxYrVYsX74cGzZsQEVFBTweT7P7i4qKJAuOiIiI5Bd2snDbbbdh8+bNuPnmm9GjR49WKyOIiIgodoSdLHz88cf4z3/+g7Fjx0YiHiIiIlKYsNdZSE1NRVpaWiRiISIiIgUKO1l4+OGH8dBDD8Fms0UiHiIiIlKYsC9DPPHEEzh8+DAyMzORn58PrVbb7P7du3dLFhwRERHJL+xk4aqrropAGERERKRUYScLixcvjkQcREREpFBhz1kgIiKi+BL2yIJKpQq6toLb7e5UQERERKQsYScL7733XrOfnU4n9uzZg9deew1Lly6VLDAiIiJShrCThSuvvLLFbddeey0GDRqEt956C3PmzJEkMCIiIlIGyeYsXHTRRdiwYYNUuyMiIiKFkCRZaGxsxMqVK9GzZ08pdkdEREQKEvZliNTU1GYTHEVRRH19PQwGA/7+979LGhwRERHJL+xk4emnn272s0qlQkZGBi688EKkpqZKFRcREREpRNjJwsyZMyV78BdeeAEvvPACSkpKAACDBg3CQw89hClTpkj2GERERNQ5YScLAFBbW4tVq1bhwIEDALwf8rNnz0ZycnJY++nVqxeWL1+Ofv36QRRFvPbaa7jyyiuxZ88eDBo0qCOhERERkcTCnuC4c+dO9OnTB0899RSqq6tRXV2NJ598En369Am7idTll1+OqVOnol+/fujfvz/+8pe/wGg0Ytu2beGGRURERBES9sjCggULcMUVV+CVV16BRuP9dZfLhdtuuw3z58/Hli1bOhSI2+3GO++8A6vVitGjR7e6jcPhgMPh8P9ssVgAeBeGcjqdHXrcs/n2I9X+qON4LJSBx0EZeByUI1aORTjxC6IoiuHsPDExEXv27MHAgQOb3f7DDz9gxIgRsNls4ewO+/fvx+jRo2G322E0GrFu3TpMnTq11W2XLFnS6iqR69atg8FgCOtxiYiI4pnNZsMNN9yAuro6mM3moNuGnSxkZmbi9ddfx+TJk5vd/umnn+KWW27ByZMnwwq2qakJpaWlqKurw/r16/Hqq69i8+bNOPfcc1ts29rIQk5ODk6dOtXuEw2V0+lEYWEhJk2aBK1WK8k+qWN4LJSBx0EZeByUI1aOhcViQbdu3UJKFsK+DHH99ddjzpw5ePzxxzFmzBgAwNdff40//OEPmDFjRtjB6nQ69O3bFwAwfPhw7NixAytWrMBLL73UYlu9Xg+9Xt/idq1WK/kBi8Q+qWN4LJSBx0EZeByUI9qPRTixh50sPP744xAEAbfccgtcLpf/Ae+44w4sX7483N214PF4mo0eEBERkbzCThZ0Oh1WrFiBZcuW4fDhwwCAPn36dGjOwKJFizBlyhTk5uaivr4e69atw6ZNm/Dpp5+GvS8iIiKKjJCTBbfbje+//x79+vVDYmIiDAYDBg8eDMDbG2Lfvn0477zzoFKFXo1ZUVGBW265BSdOnEBycjKGDBmCTz/9FJMmTQr/mRAREVFEhJwsvP7663j22Wexffv2FvdptVrMnj0b8+fPx0033RTyg69atSrkbYmIiEgeIQ8DrFq1Cvfccw/UanWL+zQaDe699168/PLLkgZHRERE8gs5WTh48CAuuuiiNu8fOXKkf/lnIiIiih0hJwtWq9W/YmJr6uvrw16QiYiIiJQv5GShX79+2Lp1a5v3f/XVV+jXr58kQREREZFyhJws3HDDDXjggQewb9++Fvd9++23eOihh3DDDTdIGhwRERHJL+RqiAULFuDjjz/G8OHDMXHiRH9viB9//BGff/45xo4diwULFkQsUCIiIpJHyMmCVqvFZ599hqeeegrr1q3Dli1bIIqiv7X0/Pnzo3rZSyIiImpdWCs4arVa3Hvvvbj33nsjFQ8REREpTOjLLRIREVFcYrJAREREQTFZICIioqCYLBAREVFQnU4W3G439u7di5qaGiniISIiIoUJO1mYP3++v1uk2+3GxRdfjAsuuAA5OTnYtGmT1PERERGRzMJOFtavX4+hQ4cCAD744AMUFxfjxx9/xIIFC3D//fdLHiARERHJK+xk4dSpU8jKygIAfPTRR7juuuvQv39/zJ49G/v375c8QCIiIpJX2MlCZmYmfvjhB7jdbnzyySeYNGkSAMBms0GtVkseIBEREckrrBUcAWDWrFmYPn06evToAUEQMHHiRADA9u3b/f0iiIiIKHaEnSwsWbIE5513HsrKynDddddBr9cDANRqNf74xz9KHiARERHJK+xkAQCuvfbaFrfNnDmz08EQERGR8nQoWdiwYQM2bNiAiooKeDyeZvetXr1aksCIiIhIGcJOFpYuXYo//elPGDFihH/eAhEREcWusJOFF198EWvXrsXNN98ciXiIiIhIYcIunWxqasKYMWMiEQsREREpUNjJwm233YZ169ZFIhYiIiJSoLAvQ9jtdrz88sv4/PPPMWTIEGi12mb3P/nkk5IFR0RERPILO1nYt28fhg0bBgD47rvvmt3HyY5ERESxJ+xkYePGjZGIg4iIiBQq7DkLgY4ePYqjR49KFQsREREpUNjJgsfjwZ/+9CckJycjLy8PeXl5SElJwcMPP9xigSYiIiKKfmFfhrj//vuxatUqLF++HGPHjgUAfPXVV1iyZAnsdjv+8pe/SB4kERERySfsZOG1117Dq6++iiuuuMJ/25AhQ9CzZ0/MnTuXyQIREVGMCfsyRHV1dautqAcOHIjq6mpJgiIiIiLlCDtZGDp0KJ599tkWtz/77LMYOnSoJEERERGRcoR9GeLRRx/Fr3/9a3z++ecYPXo0AOCbb75BWVkZPvroI8kDJCIiInmFPbJw8cUX46effsLVV1+N2tpa1NbWYtq0aTh48CB++ctfRiJGIiIiklHYIwsAkJ2dzYmMREREcSKkZGHfvn0477zzoFKpsG/fvqDbDhkyRJLAiIiISBlCShaGDRuG8vJydO/eHcOGDYMgCBBFscV2giDA7XZLHiQRERHJJ6Rkobi4GBkZGf7viYiIKH6ElCzk5eX5vz9y5AjGjBkDjab5r7pcLmzdurXZtkRERBT9wq6GGD9+fKuLL9XV1WH8+PGSBEVERETKEXayIIoiBEFocXtVVRWSkpIkCYqIiIiUI+TSyWnTpgHwTmK89dZbodfr/fe53W7s27cPY8aMkT5CIiIiklXIyUJycjIA78iCyWRCYmKi/z6dToeLLroIt99+u/QREhERkaxCThbWrFkDAMjPz8c999zDSw5ERERxIuwVHBcvXhyJOIiIiEihOrTc8/r16/H222+jtLQUTU1Nze7bvXu3JIERERGRMoRdDbFy5UrMmjULmZmZ2LNnD0aNGoX09HQUFRVhypQpkYiRiIiIZBR2svD888/j5ZdfxjPPPAOdTod7770XhYWFuOuuu1BXVxeJGImIiEhGYScLpaWl/hLJxMRE1NfXAwBuvvlmvPHGG9JGR0RERLILO1nIysryr+CYm5uLbdu2AfD2jGituVQwy5Ytw8iRI2EymdC9e3dcddVVOHjwYLghERERUQSFnSxceuml+Pe//w0AmDVrFhYsWIBJkybh+uuvx9VXXx3WvjZv3ox58+Zh27ZtKCwshNPpxOTJk2G1WsMNi4iIiCIk7GqIl19+GR6PBwAwb948pKenY+vWrbjiiivwu9/9Lqx9ffLJJ81+Xrt2Lbp3745du3Zh3Lhx4YbWpd7ctwX//unzkLYdnzcWt4+8LMIRye/j/SeQbNBiTJ9ucodC1CEfFX2EQ7WH2t1OJagwtWAqeqf07oKovA7WHMTnjZ+j6NsiqFXqoNsO6jYIE3IndFFkFA/CShZcLhceeeQRzJ49G7169QIA/Pa3v8Vvf/tbSYLxTZBMS0tr9X6HwwGHw+H/2WKxAACcTiecTqckMfj2097+HtlxP0RNbUj73Pfdh7junLFIClgiO9ZUWZswb91uJGrV2HnfeGjUYQ9atRDqsaDIipfjUFpfiv/78v9C3n5X+S68PPHlCEbU3ENbH8LPjp+x6ftN7W6rElT47OrPkJbQ+v9S6pxYeU+EE78ghjnRwGg04rvvvkN+fn64cQXl8XhwxRVXoLa2Fl999VWr2yxZsgRLly5tcfu6detgMBgkjSeYJo8bS+uWQBBEpNlHQR0k56rQ7YCgcuIG7V04N6l7l8XY1Y5agcf2ef8O9w9zoXtiO79ApDDfN32PN2xvwCyYMUg7qM3t7KIde5x7kCAk4H7z/a021pOaW3Rjad1SeODBSN1IaIL8z/nW+S1sog2zkmahj7ZPxGOj6GWz2XDDDTegrq4OZrM56LZhX4aYMGECNm/eLHmyMG/ePHz33XdtJgoAsGjRIixcuND/s8ViQU5ODiZPntzuEw2V0+lEYWEhJk2aBK1W2+o2h6pOQPjUm2N9ePNKJGh1be7votevQhNKIWSbMXXsVEliVKKvDlUB+3YBAHLOHYEJ53Q+MQrlWFDkxctxOP7dcWAf8Mu8X+LhMQ+3uZ3dZcfYt8fCLtpx0aUXIT0xPeKxFdcVw/MfD3TQ4ZkrvGXrbbl709348viXyDwnE1P7x+7/HDnFynvCNzofirCThSlTpuCPf/wj9u/fj+HDh7foEXHFFVeEu0vceeed+PDDD7Flyxb/5Y3W6PX6Zt0ufbRareQHLNg+T1hrvd+4DTAZgvfISNfm4ISnFD/VFkX1i6o9Fofb/31xtV3S5xqJ40vhi/XjcKThCACgb1rfoM9Tq9Wip7EnjjYcRZmtDFnmrIjHVmYtAwB0U3eDTqcLGl/f1L748viXONJwJKaPlxJE+3sinNjDThbmzp0LAHjyySdb3CcIAtxud4vb2yKKIv73f/8X7733HjZt2oSCgoJww5FFWV0lAEAtGtvdNsdUgBN1X6OsviTCUcmr2npm2e9DFQ0yRkLUMUW1RQCAguT2/w/1TumNow1HUVRbhJFZIyMdGorqvLFlqDLa3dYXv+93iKQQ9iw0j8fT5lc4iQLgvfTw97//HevWrYPJZEJ5eTnKy8vR2NgYblhd6mRDFQBAJ5ja3fbcbn0BANXOsojGJLeawGShkskCRReP6EGJpQQA0Du5/QoH3zZd9YHsTxbU7ScLvgqN4triiMZE8aXzU9Y74YUXXkBdXR0uueQS9OjRw//11ltvyRlWu07avMlCgrr9eRIjew4EADiEcrjCTKaiSVVAsnC4oiHsBbqI5FRuLUejqxEalQY5ppx2t5crWeiuan8ukC+2isYK1DfVRzQuih9hX4ZYuXJlq7cLgoCEhAT07dsX48aNg1odvA4YQNR+oFQ31gAAjJrkdrcd1asfRFENQeXEt+UlGN4zNmcn19jOJAsNDhdOWhzISk6QMSKi0Pk+jPNMedCo2v+32JVD/R7Rg+I67yhBKCMLJp0JGYkZqGysRHFdMYZkDIl0iBQHwk4WnnrqKVRWVsJmsyE1NRUAUFNTA4PBAKPRiIqKCvTu3RsbN25ETk77GXo0qnXUAgDMupR2t03Q6qB1Z8ClKcf2owdiNlkInLMAeOctMFmgaOGbrxDqIku+ZKHCVoGGpgYYde3PX+qok9aT3lEPQYM0VWjrJvRO7o3KxkoU1RUxWSBJhH0Z4pFHHsHIkSPx888/o6qqClVVVfjpp59w4YUXYsWKFSgtLUVWVhYWLFgQiXgVod5ZCwBITUgNaftkTU8AwPeV7a8MF618yYJJ780/D3PeAkWRYov3zD3fnB/S9sn6ZKQneEsmfXMdIsU3qtDL1Atqof0RWwDIT85v9rtEnRV2svDAAw/gqaeeQp8+Z86Q+/bti8cffxyLFi1Cr1698Oijj+Lrr7+WNFAlsbm8tandEkPL8rOT8gEAJZbYfeNWW70rgV2Q502gWBFB0STckYXAbSN9KcK3/wJz6NViXT2ngmJf2MnCiRMn4HK5WtzucrlQXl4OAMjOzva3ro5FdtGbLGQaQ1uMZUCaN7GqtMdmRYTHI/rnLIwq8CZQTBYomvjOwEOphPDxfyDXdlGyEEJJp4+/IoIjCySRsJOF8ePH43e/+x327Nnjv23Pnj244447cOmllwIA9u/fHzVrJnSES/QmQtmm0BomXdBjAACgEccjFpOc6u0uuD3eyar+ZIGXIShK1NhrUOPwTloO9TIE0HWTHDszslBWX4Ymd1M7WxO1L+xkYdWqVUhLS8Pw4cP9KyqOGDECaWlpWLVqFQBv/4gnnnhC8mCVwOPxwKPyfhDmJLc/MxkAxuSd4/1GbUVR9clIhSabKqu3uZdRr8HALO/aE5X1DtQ1RneTFYoPvg/j7KRsGLSh95jxfSBH+uzdt/9wRhYyEjNg1BrhET04YjkSqdAojoRdDZGVlYXCwkL8+OOP+OmnnwAAAwYMwIABA/zbjB8/XroIFabGboWg8l6GyU8Jrf9BusEEwZUGUVONraU/oHdaZiRD7HK+SxBpSTqYErTIMieg3GLHoYoGDM8LbRIokVz8Z+4p4Y2GBp69O91OaNXSL/tba69Ftb3aG5+5ACUoCen3BEFA7+Te2HdqH4rqitAvtZ/ksVF86fCiTAMHDsQVV1yBK664olmiEOtKaioAAKJHg3RD6OVSJnU2AODbkz9FJC45+SY3piZ5m9v07e79uxzmvAWKAv7JjWHMVwCA7obuSNImwS26I3b27ktkeiT1QKImvFauXPaZpBTSyMLChQvx8MMPIykpqVnXx9a01jMilvj6QqjEJKhUoedamQm5sDi+Q1EMLsFaffoyRJrBe2bVt7sRXx06xfJJigq+sslwhvkB79l7gbkA31V9h2JLMfqm9pU+tg5cgvDx/Q4nOZIUQkoW9uzZA6fT6f++LV3R111ux+tPAQA0Yvt9IQL1Se2Nn8uB8sbYu37oG1lIS/J2BO2T4e3EyYoIiga+HgrhjiwA3qqD76q+845O5Ekd2ZlRgQ7F1kVzKig+hJQsbNy4sdXv45GviZReFV6yMDSzPz4pB+rdsVcR4R9ZSPKOLPQ5fRmCFRGkdDanDcet3vdkRz6QIz3U35GySR9f+WRJXQk8ogcqQdZWQBTl+OoJU6XNO9nIEEJfiEBjcs8FAIiaalTZYmsNirNHFnxzFsqqbbA7Y7d5FkU/3+qLqfrUkFdkDRTps/eOrP/g09PYE1qVFna3HccbYu8khbpWyNUQs2fPDmm71atXdziYaFBj99Zjm7QpYf1e77RMwJ0EqK3YeuQALj9nVASik8fZIwsZRj3MCRpY7C4Un7LinB7td+ckkkNnztyB5smC1Gfvja5G/4d8OCtL+mhUGuSZ83Co9hCK6orQy9RLstgo/oT8yl67di02btyI2tpa1NTUtPkV6yyn+0Kk6FPC/t1EeCsidp84KGFE8qu2NR9ZEATBP7rAeQukZB1Z5jlQL1Mv/9n7CesJKUNDSV0JRIhI0acgLSG0peXPxnkLJJWQRxbuuOMOvPHGGyguLsasWbNw0003IS2tYy/gaNbgrAOADr15MxJyUOr8GQerD0sdlqxqrL51Fs7UmfftbsTu0lomC6RonRnmB846e68tQk9jT8li68zkRp/eKb2BIyyfpM4LeWThueeew4kTJ3Dvvffigw8+QE5ODqZPn45PP/0UoihGMkZFaXR7+0JkGMJPFvJM+QCA49bYqojwdZxMNej8t/nXWuAkR1KwzpQm+kSqRFGS2MwsnyRphHWBTa/XY8aMGSgsLMQPP/yAQYMGYe7cucjPz0dDQ3x8KDSd7gvRI8QmUoHO6+5dRa3OdVTSmOTkcLnR4PCuaJl++jIEAPTJ4GUIUjaXx4Uj9d7EvTNn75GqiOjsfAqgeWfMeDqpI+l1eDaOSqWCIAgQRRFud/zMeHcL3mShpzm0JlKBLuzl7RHhVFfC7oyN5i41pysh1CoBpoQzV7V8IwtFp6z+JlNESlJWXwaXx4VETSKykrI6vJ9IzQvo7CUSwNsYS4CAOkedf9looo4IK1lwOBx44403MGnSJPTv3x/79+/Hs88+i9LSUhiNoS99HK2aXC6IqkYAQG6IfSECDc3Kh+jRQhDc+O/Rn6UOTxaBlyBUqjOLcvVKNUCnUaHJ5cHRGptc4RG1yXfmnm/O71QVg79VtYQjCy6Py1/W2dHJlwCQoElAttE7sZrzFqgzQn6HzJ07Fz169MDy5cvxm9/8BmVlZXjnnXcwderUsJY9jmZldacgCN6z5JyU8C9DaNRq6EXvGcyOYz9KGptczjSRat5ER60S0LsbV3Ik5fKfuXfiwxgA8pO9Z++1jlrJzt6P1h/1j3r0SOrRqX2xIoKkEHI1xIsvvojc3Fz07t0bmzdvxubNm1vd7t1335UsOKUpPd0XAu5EGLT64Bu3IU2bg3JPGX44dUjCyORTZT3TcfJsfbsb8WN5PQ5VNGDCObHVaZOiX0cbSJ0tUZOIbGM2jjUcQ1FtEdKyOl8lJtWoB+B9fl8e+5IjC9QpIScLt9xyS1z0fgjmmMXbF0IldvySS44pH+V1W1FWHxtZfk07yQLAkQVSJilKE30Kkgu8yUJdEUZkjej0/qSY3Ojjn+RYy2SBOi7kZGHt2rURDCM6nDjdREonhNcXItDAtD7YUQdUO2OjIqKqlbJJH19FBMsnSWlEUZSkNNGnILkAXx37SrKhfqljA8501yTqiPiYbCCRitN9IRLUHV++eETPgQAAu1AOj8cjSVxy8o0spLczssCyLVKSk7aTsLlsUAtq5JpyO70/qecFSFEJ4ePbR7m1HDYnJxtTxzBZCENVozdZMIbZRCrQhb0GQBRVEFQOfFdRKlVosvFXQ7SSLBR0S4JKACx2FyobHF0dGlGbfMP8OaYcaNXadrZun5QVEaIoSnqJJFmf7F9xlpMcqaOYLISh1l4LADDrUjq8jyS9Hhq3d42GbWUHJIhKXtVB5iwkaNXISTMA4LwFUhYpz9wD93PCeqLTZ+8VtgpYnVaoBTXyzHlShBeR8k6KL0wWwlB/uolUR1rZBjJrvN3f9ldE/1oLZ0onWyYLANDXN2+ByQIpSGcbSJ0tJeFMs6fOzg2QetQDYLJAncdkIQzW030huiV2rjSqp8F7tlAcA2/cYKWTACsiSJmkHOb38S/73MmqAykrIXxYEUGdxWQhDA6PN1nITOpcstA/rQ8AoNJR1umY5CSKYtDSSQDo40sWWBFBChKJZEGqSY5SXyIBIte/guIHk4UwOP1NpMLvCxFoaFZ/AIBNPNHpmORksbvgOt33obXSSSCgfLLC2mVxEQUT2CchPzlfsv1K1X1SyrJJH1/icbT+KJwep2T7pfjBZCEMHpX37DgnOfy+EIHG5p3r/UZdj7Laqs6GJRvfqEKSTo0ErbrVbXyXIcotdtTb+U+K5Of7MM40ZCJJmyTZfqWaFxCJUY9MQyYMGgNcogtlluge0SR5MFkIUY2tAYLK+2GXl9q5kYVMYzIEdwoA4OvS7zsbmmyqgpRN+iQnapFh8i6NfbiSowskv0h8GAfur9RS2uGzd0uTBacavYu/STmyIAgCL0VQpzBZCFFJrbcvhOhRo3tSx9dZ8DEK3k5we8sPdnpfcgm2IFMgX0UEJzmSEkhdCeGTlZSFRE2i9+y9vmNn777Yuhu6w6iTtpMvKyKoM5gshOjo6SZSKtEoSZfN7oneVeMORfHs5Gpb+yMLACsiSFkiNbIQePZeXNuxeQuRmNzo46+IYLJAHcBkIURHT/eF0KDjfSEC9Tn9xi23HZFkf3IItiBTICYLpCSRKE306ezZe6QSGUC60k6KT0wWQlTR4J2IqO9EE6lAQ7p7KyLq3ccl2Z8c/GWTbVRC+PiSBTaUIrnZXXYcb/C+5yJy9q7gZMG3zxJLCTxi9Peloa7FZCFEFTZvspDYiSZSgS7KOQcA4FZXo7YxOif+hTLBEThTPllabYPD5Y54XERtOWI5AhEizDqzf8VFKXW2fDISZZM+vUy9oBE0aHQ14qT1pOT7p9jGZCFENaf7Qpi0nZ/cCAD90rMAdyIEQcTW0h8l2WdXC3WCY6ZZD6NeA7dHxJEqdr0j+QSeuQuCIPn+AxdmCrfTqsPtwLGGY979SDz5EgC0Ki1yzd65Upy3QOFishCiuqYaAECyvnN9IXxUKhUS4K2I2H0iOisiQh1ZEAThzEqOnLdAMvInCxH4MAaAHHMONIIGNpcNJ23hnb2X1HkvD5h0JqQnpEckPlZEUEcxWQiR1VkHAOiWKE2yAADd9DkAgINVhyTbZ1fyNZFqb2QBYPkkKYO/bDICcwIA79l7jjmn2WOFKrASIhKjHgCXfaaOY7IQosbTfSG6GaS7zplv9r5xj1pLJNtnV6oOcWQBYEUEKUMkKyF8Onr2HsnJjT5sKEUdxWQhRA6Pty9EVpJ0w4ODMvoBAGqdxyTbZ1dpcnlQb3cBCHFkgckCyczlceGIxVuqHNEPZCUnCxI1u6L4w2QhRG7B+yHXMzlDsn1e2MtbEeFUnYTd2STZfrtC7elLECoBMCdo293elywUnWqAxxPexC8iKRxrOAanxwm9Wo9sY3bEHqejQ/2Rnk8BAPnmfABAjaMGNfaaiD0OxR4mCyFwud0QVd7yxlwJk4WhWfkQPRoIKjd2H4+uYUH/5EaDDipV+9dXc1IToVOrYHd6cKy2MdLhEbXgO5vON+dDJUTuX19Hzt7dHjeO1HlHPQrMkbtEYtAa0COpBwCOLlB4mCyE4GhdFQTBezaclypdsqDTaKATswAA/z0WXeWTNSGu3uijUauQ380AADjExZlIBl0xzA+cGVmotlejzlEX0u8cbziOJk8TdCpdREc9AFZEUMcwWQhBSW2F9xt3AgxavaT7TtX2AgD8cOpnSfcbaaGWTQbyr+TIeQskA9+kvoKUyJ25A96z96wk70lAqB/Ivu3yk/OhVrXe7l0qrIigjmCyEIJjp/tCqERplnoO1CspHwBwxBJdQ4LhlE36sHyS5BTJJk1n85+9h1h10FWjHgAbSlHHMFkIQXm9d6lnrSBty1gAGJjeFwBQ1XRU8n1HUjhlkz5cmInkIopi134ghznUL0dsHe2MSfGJyUIITlpP94VQSdMXItCI7AEAAIdwAh5P9DR3qQ5xqedA/vLJyoawl8Il6ozKxko0OBugElTIM+dF/PHCHer3r/8Q4UskwJlk4bj1OGxOLr9OoWGyEIKqRm+JUZJGmr4QgS7KHQBRFACVHQcqo2e9heqAaohQ9e5mhCAAtTanf84DUVfwfRjnmHKgU4f+mu2ocCoiRFH0n+V3xchCakIqUk8vW19iKYn441FskDVZ2LJlCy6//HJkZ2dDEAS8//77cobTplqHN1kw61Ik37dJnwi1uxsAYHvZAcn3HynVYVZDAECiTo2eKYkAOMmRupa/m2MEyxID+UYWjjccR6MreKlwlb0K9c76Lhv1ADrfHZPij6zJgtVqxdChQ/Hcc8/JGUa76ptqAXgz8kgwa3oCAPZVRk9FREeSBaD5pQiirtJVlRA+aQlpSNYnQ4ToXzWyLb7Yehp7Qq+WttqqLayIoHDJmixMmTIFf/7zn3H11VfLGUa7rG5vrXS6hE2kAvUweNvGFkfReu0dThZYEUEy6MpKCMDbaTXUioiunNzow2WfKVwauQMIh8PhgMPh8P9ssXibOzmdTjidTkkew7efwP3ZPRZADWQkpkr2OIH6mAtwwAYctm/A4NUXBd9YVENVcyVUtmGSPHZ3kx6rZw5Hd1PbZzSP7nwUnxz5pNlttiwnkrKA323RIoQFHP0cLg+S+rnxTgWwfnX724si8MDry0J/AAmoVQJMCe2/NSbkTMD9o+7vgojk4/Q4cecXd+L7uu/x+PrHgcg0Q4w4S5P3f0VuUm5E3sOtyTflY0/FHizeuhjL/tv2a9jusgMA8kx5QWNr7X9TR+UavScoG0s34pdv/jLotmpBjdvPux3T+0/v9OPKwe6yY+7GuSHNz9AIGvzP4P/Btf2uDbpdKMfiw+IPsXLvSrg8rrDibc+/r/g3jFppKvPCeS1FVbKwbNkyLF26tMXtn332GQwGg6SPVVhY6P++yVMPqIFTxcfxUdVHkj4OAGTY1BA9aggqNwBru9s3JW5D46lBkjx2jc2JZ9Z/gdGZrVcnuEQX3qp7CyKa3y9ovJ8blg7MU1SF8aqT47PJDaDW0e5m+Oehf2JQ5SDohMhPmJPLUddR7GjYAQCwNUX3zHmjYMTh7YdRJpR1yePpmryvC7vbDrvb3u72QqmAj060//8l8H9TR9k8NiQgAXbRjlpHbbvbr9m9BsZD0peOd4XDzsPYa90b8vZrdq+B4efQPk+CHYtX61/FKfepkB83VJ999hkShARJ9mWzhf6ejqpkYdGiRVi4cKH/Z4vFgpycHEyePBlmszRljU6nE4WFhZg0aRK0Wm+DpPv//icIAH41Zjwuyu0vyeOc7fLay1BcfTLoNkcaDmPld0uRmV6L568a0+nHfOWrEry35zgMWb0xdcqAVrc5VHsI4kcijFoj1k5eCwA4XteIOa/tRoJWjffuaGckpBV1NidqG9vPaF1uF77duxdDhw2DRt01L9WnNxzCj+X1uGdyP0wY2L3N7W77/DbUOmoxYPQAnJN2TpfEJocPij4AtgE56hw8OvFRaDXtNw1TqixDFgxaaU8qgpmKqZhpnQmbq/1/yCatCRmG4EvJt/a/qTMua7oMlY2VQbepbKzEHV/cgWpUY/KvJkMTTqavEG/99BawExiVOQr3jri3ze1O2k5i3sZ5qEENfjXlV0H7h7R3LERRxGP/fAxwA0/88glJJ67mmfIkW+XTNzofiqg68nq9Hnp9y+FyrVYryZuntX3WNlohqLynz30yekj+OD79MrLRLyP4mvB1jt5Y+d1SVDkqkdddhyRtUqcec0S+Be/tOY6iKlubz6vUWgrAu+rbgG7ehMJmq4Gn6RhSDYn+2yLB6XTC+fNxTOo3LGJ/97NtOaDBD6WlaGjICPrceif3xu6K3Si1lmJI5pAuiU0ORxq8k/Oy1dkYkD6gy45DrMhJyZF8n1L9v0vTpiEtKS3oNv3EfkhQJ8DutqPCUdFl1RpSOlLvfQ0PyhgU9D3dx9MHWpUWdrcdlY5K9DL1anffbR2LqsYq1DXVQYCAcbnjkKCRZiRAauG8jrjOQjtKar2ZtyiqkWVMkTWWZH0y0hPSAUgzMSmUyYb+xWICSs46OrkxGoQ6AdO/ZG4UTUrtCN/xz1BL10CNoodKUCE/OR9A9L7WQ53cqlFp/MlQZ6tEfL+fbcxWbKIQLlmThYaGBuzduxd79+4FABQXF2Pv3r0oLS2VM6xmyk4nC4InCSqV/LmVlCVPvuWXj9U2orHJ3eo2/sViUs680WI5WegTYmmnL3mK9dnkvufXTdVN5khILv7XepT1r/Hxr7GR3H7ZrFTrT4TzmNFC1k+/nTt34vzzz8f5558PAFi4cCHOP/98PPTQQ3KG1czxem+yoBGVMblHypKn9CQdUgxaiCJQdKr1D0ffP4jArDyWkwXfOhClVTY43W0vv+1LnmI5WWhyN+FovbdnSXd12/M3KLb51qaIxpGF+qZ6VDR6uwaHUpoq1f/Xri7V7Qqyzlm45JJLFN8j4KS1GgCgj0BfiI6QcvhbEAT0zTBi55EaHKpowKDs5stZe0QPSupKvI8bmCzYYjdZyDInwKjXoMHhwpEqK/p2b73TqO/vcaT+CFweV1RO/GpPqaUUbtGNJE0STIL0HVcpOkTzmgy+mDMSM2DStf8aDrcBWFvkWDsj0uQfV1e4Sps3WTColZEsSL3ymu9MurXll483HIfdbYdOpUNPY0//7TUxPLIgCAL6ZHgnjgabt5CVlIVETSJcHhfK6rumFK+r+eerJBdAEKJ0gQXqtMAPUKWf3J0t3A/twPbdnXmu/sdNYbIQN6rt3mTBqJW+iVRH+F70ZfVlcLo7vzhLsOWXfS/4vOTmpTodaSIVTUJppa0SVMg35wOI3SVzW5vcSvEnz5wHlaBCg7Oh3VJLpQlMeEORb86HAAF1jjr///5w2Zw2lFvLAXBkIa7UnV6wJFkXmaWew5VpyESSNglu0Y3S+s5PBA32wdjWdbdYnrMABCRQIVZEROPwbCjC/UdLsUmn1iHH5C0BjbbEuLUJ2sEkaBKQbfSWsHf0ufr+H/j6g8QKJgvtaHB6+0KkRagvRLgEQfCf6UnxxvWVChafssJ11oS+tj4sYj5ZyAitIiLUtf+jVVd3aiTl8l/+jLLXekfmDnR2jkYszlcAmCy0q9HjXeGquyH44iVdSco3bs+URCRoVXC6RZRWN19pzrf/eBtZ6OOfx2GFx9P2dctY7twXOLnVd7mF4lc0trRucjfhaIO3miec0bHOPtdYLJsEmCy0q+l0spBlTJc5kjP8w98S1D2rVAJ6dzv94Vh5pi+FKIqtZshOtwcWu7cxSqwmC3lpBmjVAhqdbpywtL2mf+AZSLRN/GrPCesJ2N12aFXaZpNbKT5FY0XEEcsReEQPjFojMhJDX1SssxURHFmIUy7BOxTd06ycFeykHhJs7Rp9tb0aliYLBAjNlnitOV02qRKA5MTYXPpXo1YhP739iohcUy7Ugho2lw0nbcH7ekQb32srz5wXk2WhFB6pSgq7UuCHdjjVPIEVEZ193FjCZCEIl9sNUeUdms9JVk6y4HsRllhK4BHbXjgoVK0lC74XfE9jz2bLldZYvRUYKQYd1OH0po4yoUxy1Kq1UTvxqz2c3EiBfK+DysZK1DfVyxxNaDr6Gvb9fy23lsPmDK/TqtPjRJnFW0odS2WTAJOFoI7XV0MQvB/GuSnKWe42x5QDjUqDRlejv0SnM1orn/RXQpz1gq+yens3pxpic1TBJ+SKiCgcng1FLK5ARx1n0pnQPdG7ime0JMbhVkL4JOuTkZbgnaMW7vu6zFIGl+iCQWNApiEzrN9VOiYLQRypOV1T7EmASZ8obzABNCoN8kzSNDwBmi/M5Lv23laNvW9kIT2pZffPWBJssapAsdpQKlaHUqnjom3Z5868hjt62SWWFzJjshDEUYs3WVB5OtcKOhKk/JDKSzdArRLQ4HDhpMXRbL9nZ+XVvpGFpNgeWegTbvlklJxthaLZ5NYYG0qljvOPokVBQymP6EGJpQSAPMlCLCbZTBaCOFFfBQDQCspY6jmQlKsH6jVq5KYZAJwZdm/rRV99emQhLcZHFnqfXvK52trkLxVtTSyWT1bbq1HnqGsxuZXim7+ksFb5ycLxhuNwuB3QqrT+RZbC0dHyyVgtmwSYLAR10upNFhJUymuiI/Xqgb4z6cOVDbA6rf7Z/We/6Gv8TaRie2TBoNOgZ4r30tPhIKMLvr+P7wM2FvheU9nGbCRqlHP5jeQVTaNo/qXqO1jNw5GFlpgsBFF1em3wJI3yluyUemJd4IQ+32I86QnpLZYrrfIvyBTbIwtAaJMck7RJ/olMsTLJkZUQ1Brf/5yjDUfhcDtkjia4zk7Q9Z2MlVnK4PSE1oPHI3rOjCykxN57h8lCELWn+0KYdSmyxtEa32WIGkcNauw1nd5f4AdjsOvVZzpOxvbIAhB+RUQ0nHGFgpUQ1Jpuid1g0prgET04YjkidzhBdXbOTaYhEwaNAS7R5S+FbM9J60k0uhqhETT+kupYwmQhCEtTLQAgRa+MvhCBDFoDspM61/AkUGD5ZLBug1Ux3nEyULgNpaJllnh7YnkolTpOEIQzFREKT4zbWqo+VIIghD0fybddrjkXWlXsnUwxWQjC5vJeg85IVE5fiEBSvnH7nJ7QV1nvwMHqwwCCjyzEeukkEL8jC6yEoLZEw7oibS1VH65w39exnmQzWQjCfrovREaSMpMFKbsemhK0yDJ7V2r8+XSycPY1a1EU/ZUBsV46CZzpPnmsthG2Jleb28VSRYTNafMv9MVuk3Q2f7Kg4IqItpaqD1e4k8hjuRICYLIQlFP0nlFmm5SzemMgqTvB9emeBMCFk43eTm1nZ8jWJjeaTrexjoeRhdQknb9ZVlFAk62z+Y7D8YbjsLvabjwVDXyvpbSENKQkpMgbDClONCTGvtiyjdnNlqoPly9ZDndkgclCHHKrvMlCr2RlJguSV0RkGKHSVUOEp9ksfx/fJYgErQqJOrUkj6l0fQNKStuSnpAOs84MEaLiJ361J9b/4VHnBPalcXvcMkfTOqkm6Pou8xbXFYfUg6etJfJjBZOFNljsNggqb3lQXooy1/j2vRmOW4+H3fCkNX27G6HSVwDwZtVnL1daFUfzFXz6hDBvQRCEmJm3wEoICqansSd0Kh0cbgeOW4/LHU6rpJo7kGPKgUbw9uA5aQ3eVbbWXovq06X2sXr5jslCG0rrvAsyiaIK2aYUeYNpQ2pCKlJPV2r4ljbtjD7djVDpvMlCsMmN8TBfwSfsiogoTxZifZIWdY5apUZesncegFInOba1VH24tCotcs253n2287723d8jqQcMWkOnHlepmCy0oazO+6EpeJKgUin3zyTlNcTAkYVexpYTg+KpbNIn7IqIKC+fZLJA7VF6RYSvd4UUr+FQRwzj4X2j3E9BmR1v8A4paUTlLfUcSMoa/wyjHtoEb/Msg9ByPfUzZZPxlyyUVFnhcrd93TIaJn61x+lx+hegidXrrtR5Sr7k1qyaR4J5N6G+r+Nhrg+ThTacbDgFANCpjDJHEpyUWb4IEYLOmyy47Rkt7vePLMRRstDDnACDTg2nW8SR6rbnhfiOwxHLEbg8bZdZKlmZpQwu0YVETWKLya1EPkoeRfP9H2xtqfqOCLV8MtYnNwJMFtp0qtG7hLJBrby+EIGkPKMtt5ZDFJogimrUWlp22ozHkQWVSvB3oAx2KaJHUg/o1Xo4PU4cazjWVeFJKvDs6OzJrUQ+gf9zRFGUOZrmpD7DD7U83b/GQoxObgSYLLSp+nS/BaNW2cmCL8svrS8NueFJW3wveE9TOooqG1vcX22Lv5EFILTySbVK7e/XodRrue1hJQSFIs+cBwECLE0WVNmr5A6nGannDvg+/Kvt1ai117a6TaOrEccbvJUhHFmIQ3W+vhA65fWFCJSVlIVETSJcHheO1h/t1L58bzSPo3urZ9HVcTiyAMTPss/xMEmLOi9Bk4Cexp4AlJcYS1UJ4WPQGtAjqYd33228r0vqSiBCRIo+BWkJylztVwpMFtpgdXqXek5NUHayoBJU/jPazn5I+ZOFpgwUnbLC7Wk+xFgTh9UQwJlk4XA7yYK/V4cCr+WGgskChUqpzdMiMdGwvZOAeHnfMFlog83tbSLVzaDsZAEIfw3ztvje+CpXJppcHhytaT6hr8rfnjpOk4VKa9BrtEovKQvGI3rOXHdNid3rriQN/2vdopzXutPj9I+uSvnB3d68sHiohACYLLSpSawHAPQwtqwKUBqpZif7PiyyDd41FgKH3V1uD+oavXMi4i1ZyEtPgkYloMHhQrml7d4PgWcgSpv41Z6T1pNodDVCI2iQY8qROxxSOCVWRPiqeVpbqr4z2jsZi5e5PkwW2uASTjeRMqfLHEn7pLhWXmOvQY3DO6lzQFofAM2ThRqbN1EQBCAlzi5DaNUq5KV7V2ULNm8hz5wHlaBCg7MBlY2VXRWeJHyvnVxzLrSq+FmhkzpGieuK+M/wW1mqvjPaGzGMh7JJgMlCqzweEaLK22UwL7m7zNG0L7C8p6NntP5ObUnZ6N/dmyAFzv6vOV0JkZKohVoVf2V1fTLan+SoU+vQy9gLgLL+iYYiXoZSSRq+18lJ20lYnW13ZO1KkXoNB3aVbXQ1rxJzeVz+pfZj/b3DZKEVNW47BMG7Wl9eqvIvQ+SacqEW1LC5bDhpC97wpC2B16tbm/1fHYcLMgU6M28htIqIaJu3EC9DqSSNZH0y0hO8JxVKea37JxpKfIaflpCGFH0KRIgoqStpdt+xhmNwebwLmfmqJmIVk4VWVDV5M2XRo4dJnyhzNO3TqrX+68wdPaMNHMILTBZ8IxXxWjbpE2r5ZLRWRHBkgcKltOZpvvdcJF7DbV3q9U3wzDfnQyXE9sdpbD+7Dqp2e6sA1B5lL/UcqLNntIFZeUG3JKgEwGJ3obLB26a7Ok7LJn3OJAvBh1yjfmQhxq+7knSU9Fr3iB7/5YBIjI61NUfDPyIbB0k2k4VW1Lq8yYJWiKJkoZN1z8W1Z4ahE7Rq5KQ1n9BXHadlkz6+OQunGhyos7W9UmY0LsxUa69Ftd3bOC2Wl6slafk/QBUwiuav5lFFppqnrcRIyg6XSsdkoRUWj/fsMUHVsj+CUnXmQ8rmtOG49Xiz/fiXOGayAABI0muQnZwAADhUWd/mdr5/oJWNlahvans7JfG9Znok9YBBa5A5GooWSkqMfTHkmfKgUWkk339b5ZP+0Yw4GJFjstCKBo93ZMGgUXZfiECdeeP6XvCp+lT/ipVnX6OP92QBAPqEMG/BpDOhe6K3gkYJ/0RDES8r0JG0fK+XsvoyON2d60vTWZGa3Ojje64llhJ/V1lRFONqYjCThVbYTicLZl2KvIGEIT85H4C34Umdoy6s321tclufjDOrFgJnSifjOlkIoXwSUNbwbCg4uZE6oruhO5K0SXCLbpTWl8oai+817Fv6Xmqt9eCpF+thdVmhFtTINeVG5HGVhMlCKxrh/YBM0afIG0gYAlctC3fCUWuT284+i4730kmg+bLPwfjXvVDQUrjBxNMkLZKOIAj+OS5yj6JJ3UDqbK314KlwVwAAckw50KpjfyEzJgutcJxOFrolRlcHsY5eimitF7vvg7HcYke93Rn3pZNAGN0nfdc3a6MrWYiHoVSSllR9aTorkpUQPmdXRFR6KpvdHuuYLLTCKXgvQ3Q3KH+p50AdrYhoLStPTtQiw6QH4D2TjvfSSeBMslBWY4Pd6W5zOyVN/GpPo6sRxxtOT26Ng0laJC0lLPscWM0TqcsQQMuKiEp3ZbPbYx2ThVa4Ty/13MMUZclCBz6kXB4XjtQfafb7Pr6KiP1Ha+FweVe0jOc5C+lJOqQYtBBFoCjIpQjf3/Fow1E43I6uCq9DSupKIEJEij4FaQnRNZJG8lNCQ6nApeojWc1z9iiKb2QhXpJsJgut8PWF6GVW/lLPgTqS5ZfVl/mXK81Kymp2n+9MenuxN2vXa1Qw6NQSRRt9BEHwJ1CHgiz73C2xG0xaEzyiB0csR7oqvA5hJQR1RmCVgEf0yBKDf4JuhFurn91VliMLca7eYYeg9p4NRkNfiEC+F+3xhuOwu9pupRwocBbx2cuV+pKFHSXeZCEtSSdpN7doFMq8BUEQziz7rPBLEayEoM7oZeoFjUqDRlcjyq3lssTQVQmvrweP1WlFUV0RGkTv/4B4ee8wWThLWZ03WxRFFbJN0TUsm5aQBrPO7G14cnrCT3uCLfPrKxU8afEmT/F8CcKnz1mLVbXFN1lU6ZMcWQlBnaFRaZBnygMgX2LcVQlvYA+ejUc3AgC6J3rLR+MBk4WzlNadAgAIHgM06ugachcEIez12lurhPDxnUX7MFkIo/ukQmaJt4eVENRZnV1qvrMCl6qPNF9C8kXZF81+jgdMFs5yot6bLGhEk8yRdEy4neCC1SdnmvUw6s8sncpk4UyyUHTKCrdHbHO7aKiIcHlccbVcLUWGnOuKNLoaWyxVH0m+x/ix5kcA8dVLhcnCWU5avdfndUKUJgthzE4WRTFoIxRBEPyLMwHxXTbp0zMlEQlaFZpcHpRV29rcLnDil9vTdpmlnI7WH/VPbu2R1EPucChKyVkRUVJXAqD5UvWRdHZSzZGFOFZp8yYLCeroTBbCqYg4aTsJqzP4cqW+2f8ARxYAQKUS0Ltb+5Mcexp7QqfSweF2+M98lCbY5FaiUMnZqrqrJ+iefVLFkYUu9txzzyE/Px8JCQm48MIL8d///le2WKrtNQAAkyZFthg6w/diPmI54m940hbfGy3YcqWB8xaYLHj5KyKCzFtQq9TIS/ZO/FLqvAVWQpAU8pPzIUBAjaMGNaf/f3aVSDeQOtvZ7xUmC13orbfewsKFC7F48WLs3r0bQ4cOxWWXXYaKigpZ4qlrqgUAmHXR03EyULYxG3q1Hk6PE8cajgXdNpTJbUwWWgp52WcFLFgTDCc3khQSNYnINmYD6Po5Ol39Gg7swZMoJMbVQmbSN/4O05NPPonbb78ds2bNAgC8+OKL+M9//oPVq1fjj3/8Y9cGI4qwOb2XIbrpjUBT8IZBSqQCkG/KxcHan/Hvn/6Jc1L7t7nt1qNfAgAKjL3afK59UwQkwrtmQzedq2v/Jk4n1G6H9zFF5TRq6Z+qQiLs+LG0HIV7jW1uJzamAAA+/OkLVJ2S/a3Wwrby3QCA+hoDCvcebnM7l8uNHyod0H1bDI0muiqEYomSj4NJyARwDK/ueBubjAe67HF3ndgHAKg5lRD0NSylZFUWTuIkktzp2LCvpMuPxSXn5UMrw/EXRFFse0p3hDU1NcFgMGD9+vW46qqr/LfPnDkTtbW1+Ne//tVse4fDAYfjzPK5dXV1yM3NRXFxMUwmCeYYNFmx4M3x2Jmox9LKKlxma+z8PmXwYHoaCo2hL3u6+FQVplij87kqWaEhEQ9mKH/J8HXHytHbFfySFVEwK1KT8YZZvnle7x09gR7urplI/GRqCt42G3F5vRX3V3ftZRcAqP7dtzCZzJLsq76+HgUFBaitrUVycjuj6aKMjh07JgIQt27d2uz2P/zhD+KoUaNabL948WIRAL/4xS9+8Ytf/JLoq6ysrN3Pa+WNjQaxaNEiLFy40P+zx+NBdXU10tPTJVuG2GKxICcnB2VlZTCbpcneqGN4LJSBx0EZeByUI1aOhSiKqK+vR3Z2drvbyposdOvWDWq1GidPnmx2+8mTJ5GVldVie71eD71e3+y2lJSUiMRmNpuj+kUQS3gslIHHQRl4HJQjFo5Fu5cfTpO1GkKn02H48OHYsGGD/zaPx4MNGzZg9OjRMkZGREREPrJfhli4cCFmzpyJESNGYNSoUXj66adhtVr91RFEREQkL9mTheuvvx6VlZV46KGHUF5ejmHDhuGTTz5BZmamLPHo9XosXry4xeUO6no8FsrA46AMPA7KEY/HQtbSSSIiIlI+2VdwJCIiImVjskBERERBMVkgIiKioJgsEBERUVBMFs6ipHbZ8WDLli24/PLLkZ2dDUEQ8P777ze7XxRFPPTQQ+jRowcSExMxceJE/Pzzz/IEG8OWLVuGkSNHwmQyoXv37rjqqqtw8ODBZtvY7XbMmzcP6enpMBqNuOaaa1osqEad98ILL2DIkCH+BX9Gjx6Njz/+2H8/j0PXW758OQRBwPz58/23xdtxYLIQQGntsuOB1WrF0KFD8dxzz7V6/6OPPoqVK1fixRdfxPbt25GUlITLLrsMdru9iyONbZs3b8a8efOwbds2FBYWwul0YvLkybBaz3QZXbBgAT744AO888472Lx5M44fP45p06bJGHVs6tWrF5YvX45du3Zh586duPTSS3HllVfi+++/B8Dj0NV27NiBl156CUOGDGl2e9wdB0k6QsWIUaNGifPmzfP/7Ha7xezsbHHZsmUyRhU/AIjvvfee/2ePxyNmZWWJjz32mP+22tpaUa/Xi2+88YYMEcaPiooKEYC4efNmURS9f3etViu+8847/m0OHDggAhC/+eYbucKMG6mpqeKrr77K49DF6uvrxX79+omFhYXixRdfLN59992iKMbn+4EjC6c1NTVh165dmDhxov82lUqFiRMn4ptvvpExsvhVXFyM8vLyZsckOTkZF154IY9JhNXV1QEA0tLSAAC7du2C0+lsdiwGDhyI3NxcHosIcrvdePPNN2G1WjF69Ggehy42b948/PrXv2729wbi8/0g+wqOSnHq1Cm43e4WK0dmZmbixx9/lCmq+FZeXg4ArR4T330kPY/Hg/nz52Ps2LE477zzAHiPhU6na9G4jcciMvbv34/Ro0fDbrfDaDTivffew7nnnou9e/fyOHSRN998E7t378aOHTta3BeP7wcmC0TUzLx58/Ddd9/hq6++kjuUuDVgwADs3bsXdXV1WL9+PWbOnInNmzfLHVbcKCsrw913343CwkIkJCTIHY4i8DLEaeG2y6bI8/3deUy6zp133okPP/wQGzduRK9evfy3Z2VloampCbW1tc2257GIDJ1Oh759+2L48OFYtmwZhg4dihUrVvA4dJFdu3ahoqICF1xwATQaDTQaDTZv3oyVK1dCo9EgMzMz7o4Dk4XT2C5beQoKCpCVldXsmFgsFmzfvp3HRGKiKOLOO+/Ee++9hy+++AIFBQXN7h8+fDi0Wm2zY3Hw4EGUlpbyWHQBj8cDh8PB49BFJkyYgP3792Pv3r3+rxEjRuDGG2/0fx9vx4GXIQKwXXbXa2howKFDh/w/FxcXY+/evUhLS0Nubi7mz5+PP//5z+jXrx8KCgrw4IMPIjs7G1dddZV8QcegefPmYd26dfjXv/4Fk8nkv+6anJyMxMREJCcnY86cOVi4cCHS0tJgNpvxv//7vxg9ejQuuugimaOPLYsWLcKUKVOQm5uL+vp6rFu3Dps2bcKnn37K49BFTCaTf76OT1JSEtLT0/23x91xkLscQ2meeeYZMTc3V9TpdOKoUaPEbdu2yR1STNu4caMIoMXXzJkzRVH0lk8++OCDYmZmpqjX68UJEyaIBw8elDfoGNTaMQAgrlmzxr9NY2OjOHfuXDE1NVU0GAzi1VdfLZ44cUK+oGPU7Nmzxby8PFGn04kZGRnihAkTxM8++8x/P4+DPAJLJ0Ux/o4DW1QTERFRUJyzQEREREExWSAiIqKgmCwQERFRUEwWiIiIKCgmC0RERBQUkwUiIiIKiskCERERBcVkgYiIiIJiskBEXS4/Px9PP/203GEQUYiYLBDFuFtvvdXfS+OSSy7B/Pnzu+yx165di5SUlBa379ixA//zP//TZXEQUeewkRQRha2pqQk6na7Dv5+RkSFhNEQUaRxZIIoTt956KzZv3owVK1ZAEAQIgoCSkhIAwHfffYcpU6bAaDQiMzMTN998M06dOuX/3UsuuQR33nkn5s+fj27duuGyyy4DADz55JMYPHgwkpKSkJOTg7lz56KhoQEAsGnTJsyaNQt1dXX+x1uyZAmAlpchSktLceWVV8JoNMJsNmP69Ok4efKk//4lS5Zg2LBheP3115Gfn4/k5GT89re/RX19vX+b9evXY/DgwUhMTER6ejomTpwIq9Uaob8mUXxhskAUJ1asWIHRo0fj9ttvx4kTJ3DixAnk5OSgtrYWl156Kc4//3zs3LkTn3zyCU6ePInp06c3+/3XXnsNOp0OX3/9NV588UUAgEqlwsqVK/H999/jtddewxdffIF7770XADBmzBg8/fTTMJvN/se75557WsTl8Xhw5ZVXorq6Gps3b0ZhYSGKiopw/fXXN9vu8OHDeP/99/Hhhx/iww8/xObNm7F8+XIAwIkTJzBjxgzMnj0bBw4cwKZNmzBt2jSwTx6RNHgZgihOJCcnQ6fTwWAwICsry3/7s88+i/PPPx+PPPKI/7bVq1cjJycHP/30E/r37w8A6NevHx599NFm+wyc/5Cfn48///nP+P3vf4/nn38eOp0OycnJEASh2eOdbcOGDdi/fz+Ki4uRk5MDAPjb3/6GQYMGYceOHRg5ciQAb1Kxdu1amEwmAMDNN9+MDRs24C9/+QtOnDgBl8uFadOmIS8vDwAwePDgTvy1iCgQRxaI4ty3336LjRs3wmg0+r8GDhwIwHs27zN8+PAWv/v5559jwoQJ6NmzJ0wmE26++WZUVVXBZrOF/PgHDhxATk6OP1EAgHPPPRcpKSk4cOCA/7b8/Hx/ogAAPXr0QEVFBQBg6NChmDBhAgYPHozrrrsOr7zyCmpqakL/IxBRUEwWiOJcQ0MDLr/8cuzdu7fZ188//4xx48b5t0tKSmr2eyUlJfjNb36DIUOG4J///Cd27dqF5557DoB3AqTUtFpts58FQYDH4wEAqNVqFBYW4uOPP8a5556LZ555BgMGDEBxcbHkcRDFIyYLRHFEp9PB7XY3u+2CCy7A999/j/z8fPTt27fZ19kJQqBdu3bB4/HgiSeewEUXXYT+/fvj+PHj7T7e2c455xyUlZWhrKzMf9sPP/yA2tpanHvuuSE/N0EQMHbsWCxduhR79uyBTqfDe++9F/LvE1HbmCwQxZH8/Hxs374dJSUlOHXqFDweD+bNm4fq6mrMmDEDO3bswOHDh/Hpp59i1qxZQT/o+/btC6fTiWeeeQZFRUV4/fXX/RMfAx+voaEBGzZswKlTp1q9PDFx4kQMHjwYN954I3bv3o3//ve/uOWWW3DxxRdjxIgRIT2v7du345FHHsHOnTtRWlqKd999F5WVlTjnnHPC+wMRUauYLBDFkXvuuQdqtRrnnnsuMjIyUFpaiuzsbHz99ddwu92YPHkyBg8ejPnz5yMlJQUqVdv/IoYOHYonn3wSf/3rX3HeeefhH//4B5YtW9ZsmzFjxuD3v/89rr/+emRkZLSYIAl4RwT+9a9/ITU1FePGjcPEiRPRu3dvvPXWWyE/L7PZjC1btmDq1Kno378/HnjgATzxxBOYMmVK6H8cImqTILK2iIiIiILgyAIREREFxWSBiIiIgmKyQEREREExWSAiIqKgmCwQERFRUEwWiIiIKCgmC0RERBQUkwUiIiIKiskCERERBcVkgYiIiIJiskBERERB/X/QDDgKWmVh4AAAAABJRU5ErkJggg==",
      "text/plain": [
       "<Figure size 600x400 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.figure(figsize=(6,4))\n",
    "plt.plot(df_server_gai.index, df_server_gai['num_moves'], label='GAIKube')\n",
    "plt.plot(df_server_vpa.index, df_server_vpa['num_moves'], label='SVG')\n",
    "plt.plot(df_server_ms.index, df_server_ms['num_moves'], label='MVG')\n",
    "\n",
    "#plt.title('Cost')\n",
    "plt.xlabel('Iterations')\n",
    "plt.ylabel('Migrations Count')\n",
    "plt.legend()\n",
    "plt.grid(True)\n",
    "plt.ylim([0,5])\n",
    "plt.savefig('migrations.pdf', format='pdf')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 59,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.3409090909090909"
      ]
     },
     "execution_count": 59,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "sum(df_server_gai['num_moves'])/df_server_gai.shape[0]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 60,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.0"
      ]
     },
     "execution_count": 60,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "sum(df_server_vpa['num_moves'])/df_server_vpa.shape[0]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 63,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.8409090909090909"
      ]
     },
     "execution_count": 63,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "sum(df_server_ms['num_moves'])/df_server_ms.shape[0]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 67,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "5.965909090909091"
      ]
     },
     "execution_count": 67,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "sum(df_server_gai['cost'])/df_server_gai.shape[0]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 66,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "4.5"
      ]
     },
     "execution_count": 66,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "sum(df_server_vpa['cost'])/df_server_vpa.shape[0]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 65,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "4.5"
      ]
     },
     "execution_count": 65,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "sum(df_server_ms['cost'])/df_server_ms.shape[0]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Nodes Stats ##"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [],
   "source": [
    "df_pred = pd.read_csv('/home/babarali/Extended_ProKube/datasets/bitbrains/times_prediction.csv')\n",
    "df_pred = df_pred[['ds', 'unique_id', 'timesfm']]\n",
    "df_pred = df_pred.iloc[1554:1683]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [],
   "source": [
    "df_nodes_utils_gai = pd.read_csv('experiment_logs/gaikube/nodes_utils_stats_msgke_20240630_115907.csv')\n",
    "df_nodes_utils_vpa = pd.read_csv('experiment_logs/vpa_gke/nodes_utils_stats_msgke_20240630_144821.csv')\n",
    "df_nodes_utils_ms = pd.read_csv('experiment_logs/ms_gke/nodes_utils_stats_msgke_20240630_160859.csv')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [],
   "source": [
    "df_nodes_utils_gai['cpu_usage_percent'] = (df_nodes_utils_gai['cpu_req']/ df_nodes_utils_gai['cpu_alloc'])*100\n",
    "df_nodes_utils_gai['ram_usage_percent'] = (df_nodes_utils_gai['ram_req']/ df_nodes_utils_gai['ram_alloc'])*100\n",
    "\n",
    "df_nodes_utils_vpa['cpu_usage_percent'] = (df_nodes_utils_vpa['cpu_req']/ df_nodes_utils_vpa['cpu_alloc'])*100\n",
    "df_nodes_utils_vpa['ram_usage_percent'] = (df_nodes_utils_vpa['ram_req']/ df_nodes_utils_vpa['ram_alloc'])*100\n",
    "\n",
    "df_nodes_utils_ms['cpu_usage_percent'] = (df_nodes_utils_ms['cpu_req']/ df_nodes_utils_ms['cpu_alloc'])*100\n",
    "df_nodes_utils_ms['ram_usage_percent'] = (df_nodes_utils_ms['ram_req']/ df_nodes_utils_ms['ram_alloc'])*100\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>mid</th>\n",
       "      <th>cpu_cap</th>\n",
       "      <th>ram_cap</th>\n",
       "      <th>cpu_alloc</th>\n",
       "      <th>ram_alloc</th>\n",
       "      <th>cpu_req</th>\n",
       "      <th>ram_req</th>\n",
       "      <th>cpu_remain</th>\n",
       "      <th>ram_remain</th>\n",
       "      <th>cpu_usage_percent</th>\n",
       "      <th>ram_usage_percent</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>4</td>\n",
       "      <td>1930</td>\n",
       "      <td>2873</td>\n",
       "      <td>1759</td>\n",
       "      <td>2591</td>\n",
       "      <td>171</td>\n",
       "      <td>282</td>\n",
       "      <td>91.139896</td>\n",
       "      <td>90.184476</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1</td>\n",
       "      <td>4</td>\n",
       "      <td>8</td>\n",
       "      <td>3920</td>\n",
       "      <td>6176</td>\n",
       "      <td>3572</td>\n",
       "      <td>2836</td>\n",
       "      <td>348</td>\n",
       "      <td>3340</td>\n",
       "      <td>91.122449</td>\n",
       "      <td>45.919689</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>2</td>\n",
       "      <td>6</td>\n",
       "      <td>8</td>\n",
       "      <td>5915</td>\n",
       "      <td>6175</td>\n",
       "      <td>568</td>\n",
       "      <td>887</td>\n",
       "      <td>5347</td>\n",
       "      <td>5288</td>\n",
       "      <td>9.602705</td>\n",
       "      <td>14.364372</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>4</td>\n",
       "      <td>1930</td>\n",
       "      <td>2873</td>\n",
       "      <td>1759</td>\n",
       "      <td>2591</td>\n",
       "      <td>171</td>\n",
       "      <td>282</td>\n",
       "      <td>91.139896</td>\n",
       "      <td>90.184476</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>1</td>\n",
       "      <td>4</td>\n",
       "      <td>8</td>\n",
       "      <td>3920</td>\n",
       "      <td>6176</td>\n",
       "      <td>3572</td>\n",
       "      <td>2836</td>\n",
       "      <td>348</td>\n",
       "      <td>3340</td>\n",
       "      <td>91.122449</td>\n",
       "      <td>45.919689</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>127</th>\n",
       "      <td>1</td>\n",
       "      <td>4</td>\n",
       "      <td>8</td>\n",
       "      <td>3920</td>\n",
       "      <td>6176</td>\n",
       "      <td>3572</td>\n",
       "      <td>2836</td>\n",
       "      <td>348</td>\n",
       "      <td>3340</td>\n",
       "      <td>91.122449</td>\n",
       "      <td>45.919689</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>128</th>\n",
       "      <td>2</td>\n",
       "      <td>6</td>\n",
       "      <td>8</td>\n",
       "      <td>5915</td>\n",
       "      <td>6175</td>\n",
       "      <td>568</td>\n",
       "      <td>887</td>\n",
       "      <td>5347</td>\n",
       "      <td>5288</td>\n",
       "      <td>9.602705</td>\n",
       "      <td>14.364372</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>129</th>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>4</td>\n",
       "      <td>1930</td>\n",
       "      <td>2873</td>\n",
       "      <td>1759</td>\n",
       "      <td>2591</td>\n",
       "      <td>171</td>\n",
       "      <td>282</td>\n",
       "      <td>91.139896</td>\n",
       "      <td>90.184476</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>130</th>\n",
       "      <td>1</td>\n",
       "      <td>4</td>\n",
       "      <td>8</td>\n",
       "      <td>3920</td>\n",
       "      <td>6176</td>\n",
       "      <td>3572</td>\n",
       "      <td>2836</td>\n",
       "      <td>348</td>\n",
       "      <td>3340</td>\n",
       "      <td>91.122449</td>\n",
       "      <td>45.919689</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>131</th>\n",
       "      <td>2</td>\n",
       "      <td>6</td>\n",
       "      <td>8</td>\n",
       "      <td>5915</td>\n",
       "      <td>6175</td>\n",
       "      <td>568</td>\n",
       "      <td>887</td>\n",
       "      <td>5347</td>\n",
       "      <td>5288</td>\n",
       "      <td>9.602705</td>\n",
       "      <td>14.364372</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>132 rows × 11 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "     mid  cpu_cap  ram_cap  cpu_alloc  ram_alloc  cpu_req  ram_req  \\\n",
       "0      0        2        4       1930       2873     1759     2591   \n",
       "1      1        4        8       3920       6176     3572     2836   \n",
       "2      2        6        8       5915       6175      568      887   \n",
       "3      0        2        4       1930       2873     1759     2591   \n",
       "4      1        4        8       3920       6176     3572     2836   \n",
       "..   ...      ...      ...        ...        ...      ...      ...   \n",
       "127    1        4        8       3920       6176     3572     2836   \n",
       "128    2        6        8       5915       6175      568      887   \n",
       "129    0        2        4       1930       2873     1759     2591   \n",
       "130    1        4        8       3920       6176     3572     2836   \n",
       "131    2        6        8       5915       6175      568      887   \n",
       "\n",
       "     cpu_remain  ram_remain  cpu_usage_percent  ram_usage_percent  \n",
       "0           171         282          91.139896          90.184476  \n",
       "1           348        3340          91.122449          45.919689  \n",
       "2          5347        5288           9.602705          14.364372  \n",
       "3           171         282          91.139896          90.184476  \n",
       "4           348        3340          91.122449          45.919689  \n",
       "..          ...         ...                ...                ...  \n",
       "127         348        3340          91.122449          45.919689  \n",
       "128        5347        5288           9.602705          14.364372  \n",
       "129         171         282          91.139896          90.184476  \n",
       "130         348        3340          91.122449          45.919689  \n",
       "131        5347        5288           9.602705          14.364372  \n",
       "\n",
       "[132 rows x 11 columns]"
      ]
     },
     "execution_count": 21,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_nodes_utils_vpa"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [],
   "source": [
    "unique_mids = df_nodes_utils_gai['mid'].unique()\n",
    "\n",
    "# Create a dictionary to store DataFrames for each unique 'mid'\n",
    "df_dict_gai = {mid: df_nodes_utils_gai[df_nodes_utils_gai['mid'] == mid].reset_index(drop=True) for mid in unique_mids}\n",
    "df_dict_vpa = {mid: df_nodes_utils_vpa[df_nodes_utils_vpa['mid'] == mid].reset_index(drop=True) for mid in unique_mids}\n",
    "df_dict_ms = {mid: df_nodes_utils_ms[df_nodes_utils_ms['mid'] == mid].reset_index(drop=True) for mid in unique_mids}"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>mid</th>\n",
       "      <th>cpu_cap</th>\n",
       "      <th>ram_cap</th>\n",
       "      <th>cpu_alloc</th>\n",
       "      <th>ram_alloc</th>\n",
       "      <th>cpu_req</th>\n",
       "      <th>ram_req</th>\n",
       "      <th>cpu_remain</th>\n",
       "      <th>ram_remain</th>\n",
       "      <th>cpu_usage_percent</th>\n",
       "      <th>ram_usage_percent</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>4</td>\n",
       "      <td>1930</td>\n",
       "      <td>2873</td>\n",
       "      <td>259</td>\n",
       "      <td>591</td>\n",
       "      <td>1671</td>\n",
       "      <td>2282</td>\n",
       "      <td>13.419689</td>\n",
       "      <td>20.570832</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>4</td>\n",
       "      <td>1930</td>\n",
       "      <td>2873</td>\n",
       "      <td>259</td>\n",
       "      <td>591</td>\n",
       "      <td>1671</td>\n",
       "      <td>2282</td>\n",
       "      <td>13.419689</td>\n",
       "      <td>20.570832</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>4</td>\n",
       "      <td>1930</td>\n",
       "      <td>2873</td>\n",
       "      <td>259</td>\n",
       "      <td>591</td>\n",
       "      <td>1671</td>\n",
       "      <td>2282</td>\n",
       "      <td>13.419689</td>\n",
       "      <td>20.570832</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>4</td>\n",
       "      <td>1930</td>\n",
       "      <td>2873</td>\n",
       "      <td>259</td>\n",
       "      <td>591</td>\n",
       "      <td>1671</td>\n",
       "      <td>2282</td>\n",
       "      <td>13.419689</td>\n",
       "      <td>20.570832</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>4</td>\n",
       "      <td>1930</td>\n",
       "      <td>2873</td>\n",
       "      <td>259</td>\n",
       "      <td>591</td>\n",
       "      <td>1671</td>\n",
       "      <td>2282</td>\n",
       "      <td>13.419689</td>\n",
       "      <td>20.570832</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   mid  cpu_cap  ram_cap  cpu_alloc  ram_alloc  cpu_req  ram_req  cpu_remain  \\\n",
       "0    0        2        4       1930       2873      259      591        1671   \n",
       "1    0        2        4       1930       2873      259      591        1671   \n",
       "2    0        2        4       1930       2873      259      591        1671   \n",
       "3    0        2        4       1930       2873      259      591        1671   \n",
       "4    0        2        4       1930       2873      259      591        1671   \n",
       "\n",
       "   ram_remain  cpu_usage_percent  ram_usage_percent  \n",
       "0        2282          13.419689          20.570832  \n",
       "1        2282          13.419689          20.570832  \n",
       "2        2282          13.419689          20.570832  \n",
       "3        2282          13.419689          20.570832  \n",
       "4        2282          13.419689          20.570832  "
      ]
     },
     "execution_count": 23,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_dict_gai[0].head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>ds</th>\n",
       "      <th>unique_id</th>\n",
       "      <th>timesfm</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>2013-09-11 09:50:00</td>\n",
       "      <td>2</td>\n",
       "      <td>70.743904</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>2013-09-11 09:55:00</td>\n",
       "      <td>2</td>\n",
       "      <td>69.392440</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>2013-09-11 10:00:00</td>\n",
       "      <td>2</td>\n",
       "      <td>68.575750</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>2013-09-11 10:05:00</td>\n",
       "      <td>2</td>\n",
       "      <td>67.589910</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>2013-09-11 10:10:00</td>\n",
       "      <td>2</td>\n",
       "      <td>73.271800</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>2013-09-11 10:15:00</td>\n",
       "      <td>2</td>\n",
       "      <td>73.402440</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>2013-09-11 10:20:00</td>\n",
       "      <td>2</td>\n",
       "      <td>72.894646</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>2013-09-11 10:25:00</td>\n",
       "      <td>2</td>\n",
       "      <td>72.071625</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>2013-09-11 10:30:00</td>\n",
       "      <td>2</td>\n",
       "      <td>71.410770</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>2013-09-11 10:35:00</td>\n",
       "      <td>2</td>\n",
       "      <td>70.649100</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10</th>\n",
       "      <td>2013-09-11 10:40:00</td>\n",
       "      <td>2</td>\n",
       "      <td>78.828940</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11</th>\n",
       "      <td>2013-09-11 10:45:00</td>\n",
       "      <td>2</td>\n",
       "      <td>78.516920</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12</th>\n",
       "      <td>2013-09-11 10:50:00</td>\n",
       "      <td>2</td>\n",
       "      <td>77.966300</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>13</th>\n",
       "      <td>2013-09-11 10:55:00</td>\n",
       "      <td>2</td>\n",
       "      <td>76.407036</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>14</th>\n",
       "      <td>2013-09-11 11:00:00</td>\n",
       "      <td>2</td>\n",
       "      <td>75.341810</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>15</th>\n",
       "      <td>2013-09-11 11:05:00</td>\n",
       "      <td>2</td>\n",
       "      <td>73.591980</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>16</th>\n",
       "      <td>2013-09-11 11:10:00</td>\n",
       "      <td>2</td>\n",
       "      <td>80.882770</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>17</th>\n",
       "      <td>2013-09-11 11:15:00</td>\n",
       "      <td>2</td>\n",
       "      <td>80.079780</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>18</th>\n",
       "      <td>2013-09-11 11:20:00</td>\n",
       "      <td>2</td>\n",
       "      <td>78.402840</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>19</th>\n",
       "      <td>2013-09-11 11:25:00</td>\n",
       "      <td>2</td>\n",
       "      <td>76.518524</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>20</th>\n",
       "      <td>2013-09-11 11:30:00</td>\n",
       "      <td>2</td>\n",
       "      <td>74.350100</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>21</th>\n",
       "      <td>2013-09-11 11:35:00</td>\n",
       "      <td>2</td>\n",
       "      <td>71.840260</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>22</th>\n",
       "      <td>2013-09-11 11:40:00</td>\n",
       "      <td>2</td>\n",
       "      <td>76.612060</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>23</th>\n",
       "      <td>2013-09-11 11:45:00</td>\n",
       "      <td>2</td>\n",
       "      <td>75.853000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>24</th>\n",
       "      <td>2013-09-11 11:50:00</td>\n",
       "      <td>2</td>\n",
       "      <td>75.373510</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25</th>\n",
       "      <td>2013-09-11 11:55:00</td>\n",
       "      <td>2</td>\n",
       "      <td>74.937910</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>26</th>\n",
       "      <td>2013-09-11 12:00:00</td>\n",
       "      <td>2</td>\n",
       "      <td>73.896140</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>27</th>\n",
       "      <td>2013-09-11 12:05:00</td>\n",
       "      <td>2</td>\n",
       "      <td>73.164030</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>28</th>\n",
       "      <td>2013-09-11 12:10:00</td>\n",
       "      <td>2</td>\n",
       "      <td>83.375980</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>29</th>\n",
       "      <td>2013-09-11 12:15:00</td>\n",
       "      <td>2</td>\n",
       "      <td>83.112946</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>30</th>\n",
       "      <td>2013-09-11 12:20:00</td>\n",
       "      <td>2</td>\n",
       "      <td>83.415245</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>31</th>\n",
       "      <td>2013-09-11 12:25:00</td>\n",
       "      <td>2</td>\n",
       "      <td>83.370470</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>32</th>\n",
       "      <td>2013-09-11 12:30:00</td>\n",
       "      <td>2</td>\n",
       "      <td>83.004010</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>33</th>\n",
       "      <td>2013-09-11 12:35:00</td>\n",
       "      <td>2</td>\n",
       "      <td>83.519320</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>34</th>\n",
       "      <td>2013-09-11 12:40:00</td>\n",
       "      <td>2</td>\n",
       "      <td>73.549480</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>35</th>\n",
       "      <td>2013-09-11 12:45:00</td>\n",
       "      <td>2</td>\n",
       "      <td>72.589200</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>36</th>\n",
       "      <td>2013-09-11 12:50:00</td>\n",
       "      <td>2</td>\n",
       "      <td>71.549810</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>37</th>\n",
       "      <td>2013-09-11 12:55:00</td>\n",
       "      <td>2</td>\n",
       "      <td>70.132900</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>38</th>\n",
       "      <td>2013-09-11 13:00:00</td>\n",
       "      <td>2</td>\n",
       "      <td>68.370250</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>39</th>\n",
       "      <td>2013-09-11 13:05:00</td>\n",
       "      <td>2</td>\n",
       "      <td>67.543015</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>40</th>\n",
       "      <td>2013-09-11 13:10:00</td>\n",
       "      <td>2</td>\n",
       "      <td>69.278824</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>41</th>\n",
       "      <td>2013-09-11 13:15:00</td>\n",
       "      <td>2</td>\n",
       "      <td>68.632484</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>42</th>\n",
       "      <td>2013-09-11 13:20:00</td>\n",
       "      <td>2</td>\n",
       "      <td>67.225390</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                     ds  unique_id    timesfm\n",
       "0   2013-09-11 09:50:00          2  70.743904\n",
       "1   2013-09-11 09:55:00          2  69.392440\n",
       "2   2013-09-11 10:00:00          2  68.575750\n",
       "3   2013-09-11 10:05:00          2  67.589910\n",
       "4   2013-09-11 10:10:00          2  73.271800\n",
       "5   2013-09-11 10:15:00          2  73.402440\n",
       "6   2013-09-11 10:20:00          2  72.894646\n",
       "7   2013-09-11 10:25:00          2  72.071625\n",
       "8   2013-09-11 10:30:00          2  71.410770\n",
       "9   2013-09-11 10:35:00          2  70.649100\n",
       "10  2013-09-11 10:40:00          2  78.828940\n",
       "11  2013-09-11 10:45:00          2  78.516920\n",
       "12  2013-09-11 10:50:00          2  77.966300\n",
       "13  2013-09-11 10:55:00          2  76.407036\n",
       "14  2013-09-11 11:00:00          2  75.341810\n",
       "15  2013-09-11 11:05:00          2  73.591980\n",
       "16  2013-09-11 11:10:00          2  80.882770\n",
       "17  2013-09-11 11:15:00          2  80.079780\n",
       "18  2013-09-11 11:20:00          2  78.402840\n",
       "19  2013-09-11 11:25:00          2  76.518524\n",
       "20  2013-09-11 11:30:00          2  74.350100\n",
       "21  2013-09-11 11:35:00          2  71.840260\n",
       "22  2013-09-11 11:40:00          2  76.612060\n",
       "23  2013-09-11 11:45:00          2  75.853000\n",
       "24  2013-09-11 11:50:00          2  75.373510\n",
       "25  2013-09-11 11:55:00          2  74.937910\n",
       "26  2013-09-11 12:00:00          2  73.896140\n",
       "27  2013-09-11 12:05:00          2  73.164030\n",
       "28  2013-09-11 12:10:00          2  83.375980\n",
       "29  2013-09-11 12:15:00          2  83.112946\n",
       "30  2013-09-11 12:20:00          2  83.415245\n",
       "31  2013-09-11 12:25:00          2  83.370470\n",
       "32  2013-09-11 12:30:00          2  83.004010\n",
       "33  2013-09-11 12:35:00          2  83.519320\n",
       "34  2013-09-11 12:40:00          2  73.549480\n",
       "35  2013-09-11 12:45:00          2  72.589200\n",
       "36  2013-09-11 12:50:00          2  71.549810\n",
       "37  2013-09-11 12:55:00          2  70.132900\n",
       "38  2013-09-11 13:00:00          2  68.370250\n",
       "39  2013-09-11 13:05:00          2  67.543015\n",
       "40  2013-09-11 13:10:00          2  69.278824\n",
       "41  2013-09-11 13:15:00          2  68.632484\n",
       "42  2013-09-11 13:20:00          2  67.225390"
      ]
     },
     "execution_count": 24,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_pred2 = df_pred[df_pred['unique_id']==2].reset_index(drop=True)\n",
    "df_pred4 = df_pred[df_pred['unique_id']==4].reset_index(drop=True)\n",
    "df_pred6 = df_pred[df_pred['unique_id']==6].reset_index(drop=True)\n",
    "\n",
    "df_pred2"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 139,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 700x600 with 3 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig, (ax1, ax2, ax3) = plt.subplots(ncols=1, nrows=3, figsize=(7,6), sharex=True)\n",
    "\n",
    "ax1.plot(df_pred2.index, df_pred2['timesfm'], label='Predicted CPU')\n",
    "ax1.plot(df_dict_gai[0].index, df_dict_gai[0]['cpu_usage_percent'], label='GAIKube')\n",
    "ax1.plot(df_dict_vpa[0].index, df_dict_vpa[0]['cpu_usage_percent'], label='SVG')\n",
    "ax1.plot(df_dict_ms[0].index, df_dict_ms[0]['cpu_usage_percent'], label='MVG')\n",
    "ax1.set_title('2 Core')\n",
    "#ax1.set_xlabel('Number of Requests')\n",
    "ax1.set_ylabel('Requested CPU [%]')\n",
    "ax1.legend(loc='lower right')\n",
    "ax1.grid(True)\n",
    "ax1.set_ylim([0,100])\n",
    "\n",
    "\n",
    "ax2.plot(df_pred4.index, df_pred4['timesfm'], label='Predicted CPU')\n",
    "ax2.plot(df_dict_gai[1].index, df_dict_gai[1]['cpu_usage_percent'], label='GAIKube')\n",
    "ax2.plot(df_dict_vpa[1].index, df_dict_vpa[1]['cpu_usage_percent'], label='SVG')\n",
    "ax2.plot(df_dict_ms[1].index, df_dict_ms[1]['cpu_usage_percent'], label='MVG')\n",
    "ax2.set_title('4 Core')\n",
    "#ax2.set_xlabel('Number of Requests')\n",
    "ax2.set_ylabel('Requested CPU [%]')\n",
    "ax2.legend(loc='lower right')\n",
    "ax2.grid(True)\n",
    "ax2.set_ylim([0,100])\n",
    "\n",
    "ax3.plot(df_pred6.index, df_pred6['timesfm'], label='Predicted CPU')\n",
    "ax3.plot(df_dict_gai[2].index, df_dict_gai[2]['cpu_usage_percent'], label='GAIKube')\n",
    "ax3.plot(df_dict_vpa[2].index, df_dict_vpa[2]['cpu_usage_percent'], label='SVG')\n",
    "ax3.plot(df_dict_ms[2].index, df_dict_ms[2]['cpu_usage_percent'], label='MVG')\n",
    "ax3.set_title('6 Core')\n",
    "ax3.set_xlabel('Iterations')\n",
    "ax3.set_ylabel('Requested CPU [%]')\n",
    "ax3.legend(loc='lower right')\n",
    "#ax2.legend(loc='upper right')\n",
    "ax3.grid(True)\n",
    "ax3.set_ylim([0,100])\n",
    "\n",
    "#plt.setp(ax3.get_xticklabels(), rotation=30, ha='right')\n",
    "plt.savefig('cpu_usage.pdf', format='pdf')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 157,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>mid</th>\n",
       "      <th>cpu_cap</th>\n",
       "      <th>ram_cap</th>\n",
       "      <th>cpu_alloc</th>\n",
       "      <th>ram_alloc</th>\n",
       "      <th>cpu_req</th>\n",
       "      <th>ram_req</th>\n",
       "      <th>cpu_remain</th>\n",
       "      <th>ram_remain</th>\n",
       "      <th>cpu_usage_percent</th>\n",
       "      <th>ram_usage_percent</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1</td>\n",
       "      <td>4</td>\n",
       "      <td>8</td>\n",
       "      <td>3920</td>\n",
       "      <td>6176</td>\n",
       "      <td>2572</td>\n",
       "      <td>2836</td>\n",
       "      <td>1348</td>\n",
       "      <td>3340</td>\n",
       "      <td>65.612245</td>\n",
       "      <td>45.919689</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>1</td>\n",
       "      <td>4</td>\n",
       "      <td>8</td>\n",
       "      <td>3920</td>\n",
       "      <td>6176</td>\n",
       "      <td>2572</td>\n",
       "      <td>2836</td>\n",
       "      <td>1348</td>\n",
       "      <td>3340</td>\n",
       "      <td>65.612245</td>\n",
       "      <td>45.919689</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>1</td>\n",
       "      <td>4</td>\n",
       "      <td>8</td>\n",
       "      <td>3920</td>\n",
       "      <td>6176</td>\n",
       "      <td>2572</td>\n",
       "      <td>2836</td>\n",
       "      <td>1348</td>\n",
       "      <td>3340</td>\n",
       "      <td>65.612245</td>\n",
       "      <td>45.919689</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10</th>\n",
       "      <td>1</td>\n",
       "      <td>4</td>\n",
       "      <td>8</td>\n",
       "      <td>3920</td>\n",
       "      <td>6176</td>\n",
       "      <td>2572</td>\n",
       "      <td>2836</td>\n",
       "      <td>1348</td>\n",
       "      <td>3340</td>\n",
       "      <td>65.612245</td>\n",
       "      <td>45.919689</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>13</th>\n",
       "      <td>1</td>\n",
       "      <td>4</td>\n",
       "      <td>8</td>\n",
       "      <td>3920</td>\n",
       "      <td>6176</td>\n",
       "      <td>2572</td>\n",
       "      <td>2836</td>\n",
       "      <td>1348</td>\n",
       "      <td>3340</td>\n",
       "      <td>65.612245</td>\n",
       "      <td>45.919689</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>16</th>\n",
       "      <td>1</td>\n",
       "      <td>4</td>\n",
       "      <td>8</td>\n",
       "      <td>3920</td>\n",
       "      <td>6176</td>\n",
       "      <td>2572</td>\n",
       "      <td>2836</td>\n",
       "      <td>1348</td>\n",
       "      <td>3340</td>\n",
       "      <td>65.612245</td>\n",
       "      <td>45.919689</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>19</th>\n",
       "      <td>1</td>\n",
       "      <td>4</td>\n",
       "      <td>8</td>\n",
       "      <td>3920</td>\n",
       "      <td>6176</td>\n",
       "      <td>2572</td>\n",
       "      <td>2836</td>\n",
       "      <td>1348</td>\n",
       "      <td>3340</td>\n",
       "      <td>65.612245</td>\n",
       "      <td>45.919689</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>22</th>\n",
       "      <td>1</td>\n",
       "      <td>4</td>\n",
       "      <td>8</td>\n",
       "      <td>3920</td>\n",
       "      <td>6176</td>\n",
       "      <td>2572</td>\n",
       "      <td>2836</td>\n",
       "      <td>1348</td>\n",
       "      <td>3340</td>\n",
       "      <td>65.612245</td>\n",
       "      <td>45.919689</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25</th>\n",
       "      <td>1</td>\n",
       "      <td>4</td>\n",
       "      <td>8</td>\n",
       "      <td>3920</td>\n",
       "      <td>6176</td>\n",
       "      <td>2572</td>\n",
       "      <td>2836</td>\n",
       "      <td>1348</td>\n",
       "      <td>3340</td>\n",
       "      <td>65.612245</td>\n",
       "      <td>45.919689</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>28</th>\n",
       "      <td>1</td>\n",
       "      <td>4</td>\n",
       "      <td>8</td>\n",
       "      <td>3920</td>\n",
       "      <td>6176</td>\n",
       "      <td>2572</td>\n",
       "      <td>2836</td>\n",
       "      <td>1348</td>\n",
       "      <td>3340</td>\n",
       "      <td>65.612245</td>\n",
       "      <td>45.919689</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>31</th>\n",
       "      <td>1</td>\n",
       "      <td>4</td>\n",
       "      <td>8</td>\n",
       "      <td>3920</td>\n",
       "      <td>6176</td>\n",
       "      <td>2572</td>\n",
       "      <td>2836</td>\n",
       "      <td>1348</td>\n",
       "      <td>3340</td>\n",
       "      <td>65.612245</td>\n",
       "      <td>45.919689</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>34</th>\n",
       "      <td>1</td>\n",
       "      <td>4</td>\n",
       "      <td>8</td>\n",
       "      <td>3920</td>\n",
       "      <td>6176</td>\n",
       "      <td>2572</td>\n",
       "      <td>2836</td>\n",
       "      <td>1348</td>\n",
       "      <td>3340</td>\n",
       "      <td>65.612245</td>\n",
       "      <td>45.919689</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>37</th>\n",
       "      <td>1</td>\n",
       "      <td>4</td>\n",
       "      <td>8</td>\n",
       "      <td>3920</td>\n",
       "      <td>6176</td>\n",
       "      <td>2572</td>\n",
       "      <td>2836</td>\n",
       "      <td>1348</td>\n",
       "      <td>3340</td>\n",
       "      <td>65.612245</td>\n",
       "      <td>45.919689</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>40</th>\n",
       "      <td>1</td>\n",
       "      <td>4</td>\n",
       "      <td>8</td>\n",
       "      <td>3920</td>\n",
       "      <td>6176</td>\n",
       "      <td>2572</td>\n",
       "      <td>2836</td>\n",
       "      <td>1348</td>\n",
       "      <td>3340</td>\n",
       "      <td>65.612245</td>\n",
       "      <td>45.919689</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>43</th>\n",
       "      <td>1</td>\n",
       "      <td>4</td>\n",
       "      <td>8</td>\n",
       "      <td>3920</td>\n",
       "      <td>6176</td>\n",
       "      <td>2572</td>\n",
       "      <td>2836</td>\n",
       "      <td>1348</td>\n",
       "      <td>3340</td>\n",
       "      <td>65.612245</td>\n",
       "      <td>45.919689</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>46</th>\n",
       "      <td>1</td>\n",
       "      <td>4</td>\n",
       "      <td>8</td>\n",
       "      <td>3920</td>\n",
       "      <td>6176</td>\n",
       "      <td>2572</td>\n",
       "      <td>2836</td>\n",
       "      <td>1348</td>\n",
       "      <td>3340</td>\n",
       "      <td>65.612245</td>\n",
       "      <td>45.919689</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>49</th>\n",
       "      <td>1</td>\n",
       "      <td>4</td>\n",
       "      <td>8</td>\n",
       "      <td>3920</td>\n",
       "      <td>6176</td>\n",
       "      <td>2572</td>\n",
       "      <td>2836</td>\n",
       "      <td>1348</td>\n",
       "      <td>3340</td>\n",
       "      <td>65.612245</td>\n",
       "      <td>45.919689</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>52</th>\n",
       "      <td>1</td>\n",
       "      <td>4</td>\n",
       "      <td>8</td>\n",
       "      <td>3920</td>\n",
       "      <td>6176</td>\n",
       "      <td>2572</td>\n",
       "      <td>2836</td>\n",
       "      <td>1348</td>\n",
       "      <td>3340</td>\n",
       "      <td>65.612245</td>\n",
       "      <td>45.919689</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>55</th>\n",
       "      <td>1</td>\n",
       "      <td>4</td>\n",
       "      <td>8</td>\n",
       "      <td>3920</td>\n",
       "      <td>6176</td>\n",
       "      <td>2572</td>\n",
       "      <td>2836</td>\n",
       "      <td>1348</td>\n",
       "      <td>3340</td>\n",
       "      <td>65.612245</td>\n",
       "      <td>45.919689</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>58</th>\n",
       "      <td>1</td>\n",
       "      <td>4</td>\n",
       "      <td>8</td>\n",
       "      <td>3920</td>\n",
       "      <td>6176</td>\n",
       "      <td>2572</td>\n",
       "      <td>2836</td>\n",
       "      <td>1348</td>\n",
       "      <td>3340</td>\n",
       "      <td>65.612245</td>\n",
       "      <td>45.919689</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>61</th>\n",
       "      <td>1</td>\n",
       "      <td>4</td>\n",
       "      <td>8</td>\n",
       "      <td>3920</td>\n",
       "      <td>6176</td>\n",
       "      <td>2572</td>\n",
       "      <td>2836</td>\n",
       "      <td>1348</td>\n",
       "      <td>3340</td>\n",
       "      <td>65.612245</td>\n",
       "      <td>45.919689</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>64</th>\n",
       "      <td>1</td>\n",
       "      <td>4</td>\n",
       "      <td>8</td>\n",
       "      <td>3920</td>\n",
       "      <td>6176</td>\n",
       "      <td>2572</td>\n",
       "      <td>2836</td>\n",
       "      <td>1348</td>\n",
       "      <td>3340</td>\n",
       "      <td>65.612245</td>\n",
       "      <td>45.919689</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>67</th>\n",
       "      <td>1</td>\n",
       "      <td>4</td>\n",
       "      <td>8</td>\n",
       "      <td>3920</td>\n",
       "      <td>6176</td>\n",
       "      <td>2572</td>\n",
       "      <td>2836</td>\n",
       "      <td>1348</td>\n",
       "      <td>3340</td>\n",
       "      <td>65.612245</td>\n",
       "      <td>45.919689</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>70</th>\n",
       "      <td>1</td>\n",
       "      <td>4</td>\n",
       "      <td>8</td>\n",
       "      <td>3920</td>\n",
       "      <td>6176</td>\n",
       "      <td>2572</td>\n",
       "      <td>2836</td>\n",
       "      <td>1348</td>\n",
       "      <td>3340</td>\n",
       "      <td>65.612245</td>\n",
       "      <td>45.919689</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>73</th>\n",
       "      <td>1</td>\n",
       "      <td>4</td>\n",
       "      <td>8</td>\n",
       "      <td>3920</td>\n",
       "      <td>6176</td>\n",
       "      <td>2572</td>\n",
       "      <td>2836</td>\n",
       "      <td>1348</td>\n",
       "      <td>3340</td>\n",
       "      <td>65.612245</td>\n",
       "      <td>45.919689</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>76</th>\n",
       "      <td>1</td>\n",
       "      <td>4</td>\n",
       "      <td>8</td>\n",
       "      <td>3920</td>\n",
       "      <td>6176</td>\n",
       "      <td>2572</td>\n",
       "      <td>2836</td>\n",
       "      <td>1348</td>\n",
       "      <td>3340</td>\n",
       "      <td>65.612245</td>\n",
       "      <td>45.919689</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>79</th>\n",
       "      <td>1</td>\n",
       "      <td>4</td>\n",
       "      <td>8</td>\n",
       "      <td>3920</td>\n",
       "      <td>6176</td>\n",
       "      <td>2572</td>\n",
       "      <td>2836</td>\n",
       "      <td>1348</td>\n",
       "      <td>3340</td>\n",
       "      <td>65.612245</td>\n",
       "      <td>45.919689</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>82</th>\n",
       "      <td>1</td>\n",
       "      <td>4</td>\n",
       "      <td>8</td>\n",
       "      <td>3920</td>\n",
       "      <td>6176</td>\n",
       "      <td>2572</td>\n",
       "      <td>2836</td>\n",
       "      <td>1348</td>\n",
       "      <td>3340</td>\n",
       "      <td>65.612245</td>\n",
       "      <td>45.919689</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>85</th>\n",
       "      <td>1</td>\n",
       "      <td>4</td>\n",
       "      <td>8</td>\n",
       "      <td>3920</td>\n",
       "      <td>6176</td>\n",
       "      <td>2572</td>\n",
       "      <td>2836</td>\n",
       "      <td>1348</td>\n",
       "      <td>3340</td>\n",
       "      <td>65.612245</td>\n",
       "      <td>45.919689</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>88</th>\n",
       "      <td>1</td>\n",
       "      <td>4</td>\n",
       "      <td>8</td>\n",
       "      <td>3920</td>\n",
       "      <td>6176</td>\n",
       "      <td>2572</td>\n",
       "      <td>2836</td>\n",
       "      <td>1348</td>\n",
       "      <td>3340</td>\n",
       "      <td>65.612245</td>\n",
       "      <td>45.919689</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>91</th>\n",
       "      <td>1</td>\n",
       "      <td>4</td>\n",
       "      <td>8</td>\n",
       "      <td>3920</td>\n",
       "      <td>6176</td>\n",
       "      <td>2572</td>\n",
       "      <td>2836</td>\n",
       "      <td>1348</td>\n",
       "      <td>3340</td>\n",
       "      <td>65.612245</td>\n",
       "      <td>45.919689</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>94</th>\n",
       "      <td>1</td>\n",
       "      <td>4</td>\n",
       "      <td>8</td>\n",
       "      <td>3920</td>\n",
       "      <td>6176</td>\n",
       "      <td>2572</td>\n",
       "      <td>2836</td>\n",
       "      <td>1348</td>\n",
       "      <td>3340</td>\n",
       "      <td>65.612245</td>\n",
       "      <td>45.919689</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>97</th>\n",
       "      <td>1</td>\n",
       "      <td>4</td>\n",
       "      <td>8</td>\n",
       "      <td>3920</td>\n",
       "      <td>6176</td>\n",
       "      <td>2572</td>\n",
       "      <td>2836</td>\n",
       "      <td>1348</td>\n",
       "      <td>3340</td>\n",
       "      <td>65.612245</td>\n",
       "      <td>45.919689</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>100</th>\n",
       "      <td>1</td>\n",
       "      <td>4</td>\n",
       "      <td>8</td>\n",
       "      <td>3920</td>\n",
       "      <td>6176</td>\n",
       "      <td>2572</td>\n",
       "      <td>2836</td>\n",
       "      <td>1348</td>\n",
       "      <td>3340</td>\n",
       "      <td>65.612245</td>\n",
       "      <td>45.919689</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>103</th>\n",
       "      <td>1</td>\n",
       "      <td>4</td>\n",
       "      <td>8</td>\n",
       "      <td>3920</td>\n",
       "      <td>6176</td>\n",
       "      <td>2572</td>\n",
       "      <td>2836</td>\n",
       "      <td>1348</td>\n",
       "      <td>3340</td>\n",
       "      <td>65.612245</td>\n",
       "      <td>45.919689</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>106</th>\n",
       "      <td>1</td>\n",
       "      <td>4</td>\n",
       "      <td>8</td>\n",
       "      <td>3920</td>\n",
       "      <td>6176</td>\n",
       "      <td>2572</td>\n",
       "      <td>2836</td>\n",
       "      <td>1348</td>\n",
       "      <td>3340</td>\n",
       "      <td>65.612245</td>\n",
       "      <td>45.919689</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>109</th>\n",
       "      <td>1</td>\n",
       "      <td>4</td>\n",
       "      <td>8</td>\n",
       "      <td>3920</td>\n",
       "      <td>6176</td>\n",
       "      <td>2572</td>\n",
       "      <td>2836</td>\n",
       "      <td>1348</td>\n",
       "      <td>3340</td>\n",
       "      <td>65.612245</td>\n",
       "      <td>45.919689</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>112</th>\n",
       "      <td>1</td>\n",
       "      <td>4</td>\n",
       "      <td>8</td>\n",
       "      <td>3920</td>\n",
       "      <td>6176</td>\n",
       "      <td>2572</td>\n",
       "      <td>2836</td>\n",
       "      <td>1348</td>\n",
       "      <td>3340</td>\n",
       "      <td>65.612245</td>\n",
       "      <td>45.919689</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>115</th>\n",
       "      <td>1</td>\n",
       "      <td>4</td>\n",
       "      <td>8</td>\n",
       "      <td>3920</td>\n",
       "      <td>6176</td>\n",
       "      <td>2572</td>\n",
       "      <td>2836</td>\n",
       "      <td>1348</td>\n",
       "      <td>3340</td>\n",
       "      <td>65.612245</td>\n",
       "      <td>45.919689</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>118</th>\n",
       "      <td>1</td>\n",
       "      <td>4</td>\n",
       "      <td>8</td>\n",
       "      <td>3920</td>\n",
       "      <td>6176</td>\n",
       "      <td>2572</td>\n",
       "      <td>2836</td>\n",
       "      <td>1348</td>\n",
       "      <td>3340</td>\n",
       "      <td>65.612245</td>\n",
       "      <td>45.919689</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>121</th>\n",
       "      <td>1</td>\n",
       "      <td>4</td>\n",
       "      <td>8</td>\n",
       "      <td>3920</td>\n",
       "      <td>6176</td>\n",
       "      <td>2572</td>\n",
       "      <td>2836</td>\n",
       "      <td>1348</td>\n",
       "      <td>3340</td>\n",
       "      <td>65.612245</td>\n",
       "      <td>45.919689</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>124</th>\n",
       "      <td>1</td>\n",
       "      <td>4</td>\n",
       "      <td>8</td>\n",
       "      <td>3920</td>\n",
       "      <td>6176</td>\n",
       "      <td>2572</td>\n",
       "      <td>2836</td>\n",
       "      <td>1348</td>\n",
       "      <td>3340</td>\n",
       "      <td>65.612245</td>\n",
       "      <td>45.919689</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>127</th>\n",
       "      <td>1</td>\n",
       "      <td>4</td>\n",
       "      <td>8</td>\n",
       "      <td>3920</td>\n",
       "      <td>6176</td>\n",
       "      <td>2572</td>\n",
       "      <td>2836</td>\n",
       "      <td>1348</td>\n",
       "      <td>3340</td>\n",
       "      <td>65.612245</td>\n",
       "      <td>45.919689</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>130</th>\n",
       "      <td>1</td>\n",
       "      <td>4</td>\n",
       "      <td>8</td>\n",
       "      <td>3920</td>\n",
       "      <td>6176</td>\n",
       "      <td>2572</td>\n",
       "      <td>2836</td>\n",
       "      <td>1348</td>\n",
       "      <td>3340</td>\n",
       "      <td>65.612245</td>\n",
       "      <td>45.919689</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "     mid  cpu_cap  ram_cap  cpu_alloc  ram_alloc  cpu_req  ram_req  \\\n",
       "1      1        4        8       3920       6176     2572     2836   \n",
       "4      1        4        8       3920       6176     2572     2836   \n",
       "7      1        4        8       3920       6176     2572     2836   \n",
       "10     1        4        8       3920       6176     2572     2836   \n",
       "13     1        4        8       3920       6176     2572     2836   \n",
       "16     1        4        8       3920       6176     2572     2836   \n",
       "19     1        4        8       3920       6176     2572     2836   \n",
       "22     1        4        8       3920       6176     2572     2836   \n",
       "25     1        4        8       3920       6176     2572     2836   \n",
       "28     1        4        8       3920       6176     2572     2836   \n",
       "31     1        4        8       3920       6176     2572     2836   \n",
       "34     1        4        8       3920       6176     2572     2836   \n",
       "37     1        4        8       3920       6176     2572     2836   \n",
       "40     1        4        8       3920       6176     2572     2836   \n",
       "43     1        4        8       3920       6176     2572     2836   \n",
       "46     1        4        8       3920       6176     2572     2836   \n",
       "49     1        4        8       3920       6176     2572     2836   \n",
       "52     1        4        8       3920       6176     2572     2836   \n",
       "55     1        4        8       3920       6176     2572     2836   \n",
       "58     1        4        8       3920       6176     2572     2836   \n",
       "61     1        4        8       3920       6176     2572     2836   \n",
       "64     1        4        8       3920       6176     2572     2836   \n",
       "67     1        4        8       3920       6176     2572     2836   \n",
       "70     1        4        8       3920       6176     2572     2836   \n",
       "73     1        4        8       3920       6176     2572     2836   \n",
       "76     1        4        8       3920       6176     2572     2836   \n",
       "79     1        4        8       3920       6176     2572     2836   \n",
       "82     1        4        8       3920       6176     2572     2836   \n",
       "85     1        4        8       3920       6176     2572     2836   \n",
       "88     1        4        8       3920       6176     2572     2836   \n",
       "91     1        4        8       3920       6176     2572     2836   \n",
       "94     1        4        8       3920       6176     2572     2836   \n",
       "97     1        4        8       3920       6176     2572     2836   \n",
       "100    1        4        8       3920       6176     2572     2836   \n",
       "103    1        4        8       3920       6176     2572     2836   \n",
       "106    1        4        8       3920       6176     2572     2836   \n",
       "109    1        4        8       3920       6176     2572     2836   \n",
       "112    1        4        8       3920       6176     2572     2836   \n",
       "115    1        4        8       3920       6176     2572     2836   \n",
       "118    1        4        8       3920       6176     2572     2836   \n",
       "121    1        4        8       3920       6176     2572     2836   \n",
       "124    1        4        8       3920       6176     2572     2836   \n",
       "127    1        4        8       3920       6176     2572     2836   \n",
       "130    1        4        8       3920       6176     2572     2836   \n",
       "\n",
       "     cpu_remain  ram_remain  cpu_usage_percent  ram_usage_percent  \n",
       "1          1348        3340          65.612245          45.919689  \n",
       "4          1348        3340          65.612245          45.919689  \n",
       "7          1348        3340          65.612245          45.919689  \n",
       "10         1348        3340          65.612245          45.919689  \n",
       "13         1348        3340          65.612245          45.919689  \n",
       "16         1348        3340          65.612245          45.919689  \n",
       "19         1348        3340          65.612245          45.919689  \n",
       "22         1348        3340          65.612245          45.919689  \n",
       "25         1348        3340          65.612245          45.919689  \n",
       "28         1348        3340          65.612245          45.919689  \n",
       "31         1348        3340          65.612245          45.919689  \n",
       "34         1348        3340          65.612245          45.919689  \n",
       "37         1348        3340          65.612245          45.919689  \n",
       "40         1348        3340          65.612245          45.919689  \n",
       "43         1348        3340          65.612245          45.919689  \n",
       "46         1348        3340          65.612245          45.919689  \n",
       "49         1348        3340          65.612245          45.919689  \n",
       "52         1348        3340          65.612245          45.919689  \n",
       "55         1348        3340          65.612245          45.919689  \n",
       "58         1348        3340          65.612245          45.919689  \n",
       "61         1348        3340          65.612245          45.919689  \n",
       "64         1348        3340          65.612245          45.919689  \n",
       "67         1348        3340          65.612245          45.919689  \n",
       "70         1348        3340          65.612245          45.919689  \n",
       "73         1348        3340          65.612245          45.919689  \n",
       "76         1348        3340          65.612245          45.919689  \n",
       "79         1348        3340          65.612245          45.919689  \n",
       "82         1348        3340          65.612245          45.919689  \n",
       "85         1348        3340          65.612245          45.919689  \n",
       "88         1348        3340          65.612245          45.919689  \n",
       "91         1348        3340          65.612245          45.919689  \n",
       "94         1348        3340          65.612245          45.919689  \n",
       "97         1348        3340          65.612245          45.919689  \n",
       "100        1348        3340          65.612245          45.919689  \n",
       "103        1348        3340          65.612245          45.919689  \n",
       "106        1348        3340          65.612245          45.919689  \n",
       "109        1348        3340          65.612245          45.919689  \n",
       "112        1348        3340          65.612245          45.919689  \n",
       "115        1348        3340          65.612245          45.919689  \n",
       "118        1348        3340          65.612245          45.919689  \n",
       "121        1348        3340          65.612245          45.919689  \n",
       "124        1348        3340          65.612245          45.919689  \n",
       "127        1348        3340          65.612245          45.919689  \n",
       "130        1348        3340          65.612245          45.919689  "
      ]
     },
     "execution_count": 157,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_dict_gai[1]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 68,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>mid</th>\n",
       "      <th>cpu_cap</th>\n",
       "      <th>ram_cap</th>\n",
       "      <th>cpu_alloc</th>\n",
       "      <th>ram_alloc</th>\n",
       "      <th>cpu_req</th>\n",
       "      <th>ram_req</th>\n",
       "      <th>cpu_remain</th>\n",
       "      <th>ram_remain</th>\n",
       "      <th>cpu_usage_percent</th>\n",
       "      <th>ram_usage_percent</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>2</td>\n",
       "      <td>6</td>\n",
       "      <td>8</td>\n",
       "      <td>5915</td>\n",
       "      <td>6175</td>\n",
       "      <td>567</td>\n",
       "      <td>871</td>\n",
       "      <td>5348</td>\n",
       "      <td>5304</td>\n",
       "      <td>9.585799</td>\n",
       "      <td>14.105263</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>2</td>\n",
       "      <td>6</td>\n",
       "      <td>8</td>\n",
       "      <td>5915</td>\n",
       "      <td>6175</td>\n",
       "      <td>567</td>\n",
       "      <td>871</td>\n",
       "      <td>5348</td>\n",
       "      <td>5304</td>\n",
       "      <td>9.585799</td>\n",
       "      <td>14.105263</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>2</td>\n",
       "      <td>6</td>\n",
       "      <td>8</td>\n",
       "      <td>5915</td>\n",
       "      <td>6175</td>\n",
       "      <td>567</td>\n",
       "      <td>871</td>\n",
       "      <td>5348</td>\n",
       "      <td>5304</td>\n",
       "      <td>9.585799</td>\n",
       "      <td>14.105263</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11</th>\n",
       "      <td>2</td>\n",
       "      <td>6</td>\n",
       "      <td>8</td>\n",
       "      <td>5915</td>\n",
       "      <td>6175</td>\n",
       "      <td>567</td>\n",
       "      <td>871</td>\n",
       "      <td>5348</td>\n",
       "      <td>5304</td>\n",
       "      <td>9.585799</td>\n",
       "      <td>14.105263</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>14</th>\n",
       "      <td>2</td>\n",
       "      <td>6</td>\n",
       "      <td>8</td>\n",
       "      <td>5915</td>\n",
       "      <td>6175</td>\n",
       "      <td>567</td>\n",
       "      <td>871</td>\n",
       "      <td>5348</td>\n",
       "      <td>5304</td>\n",
       "      <td>9.585799</td>\n",
       "      <td>14.105263</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>17</th>\n",
       "      <td>2</td>\n",
       "      <td>6</td>\n",
       "      <td>8</td>\n",
       "      <td>5915</td>\n",
       "      <td>6175</td>\n",
       "      <td>567</td>\n",
       "      <td>871</td>\n",
       "      <td>5348</td>\n",
       "      <td>5304</td>\n",
       "      <td>9.585799</td>\n",
       "      <td>14.105263</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>20</th>\n",
       "      <td>2</td>\n",
       "      <td>6</td>\n",
       "      <td>8</td>\n",
       "      <td>5915</td>\n",
       "      <td>6175</td>\n",
       "      <td>567</td>\n",
       "      <td>871</td>\n",
       "      <td>5348</td>\n",
       "      <td>5304</td>\n",
       "      <td>9.585799</td>\n",
       "      <td>14.105263</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>23</th>\n",
       "      <td>2</td>\n",
       "      <td>6</td>\n",
       "      <td>8</td>\n",
       "      <td>5915</td>\n",
       "      <td>6175</td>\n",
       "      <td>567</td>\n",
       "      <td>871</td>\n",
       "      <td>5348</td>\n",
       "      <td>5304</td>\n",
       "      <td>9.585799</td>\n",
       "      <td>14.105263</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>26</th>\n",
       "      <td>2</td>\n",
       "      <td>6</td>\n",
       "      <td>8</td>\n",
       "      <td>5915</td>\n",
       "      <td>6175</td>\n",
       "      <td>567</td>\n",
       "      <td>871</td>\n",
       "      <td>5348</td>\n",
       "      <td>5304</td>\n",
       "      <td>9.585799</td>\n",
       "      <td>14.105263</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>29</th>\n",
       "      <td>2</td>\n",
       "      <td>6</td>\n",
       "      <td>8</td>\n",
       "      <td>5915</td>\n",
       "      <td>6175</td>\n",
       "      <td>567</td>\n",
       "      <td>871</td>\n",
       "      <td>5348</td>\n",
       "      <td>5304</td>\n",
       "      <td>9.585799</td>\n",
       "      <td>14.105263</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>32</th>\n",
       "      <td>2</td>\n",
       "      <td>6</td>\n",
       "      <td>8</td>\n",
       "      <td>5915</td>\n",
       "      <td>6175</td>\n",
       "      <td>567</td>\n",
       "      <td>871</td>\n",
       "      <td>5348</td>\n",
       "      <td>5304</td>\n",
       "      <td>9.585799</td>\n",
       "      <td>14.105263</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>35</th>\n",
       "      <td>2</td>\n",
       "      <td>6</td>\n",
       "      <td>8</td>\n",
       "      <td>5915</td>\n",
       "      <td>6175</td>\n",
       "      <td>567</td>\n",
       "      <td>871</td>\n",
       "      <td>5348</td>\n",
       "      <td>5304</td>\n",
       "      <td>9.585799</td>\n",
       "      <td>14.105263</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "    mid  cpu_cap  ram_cap  cpu_alloc  ram_alloc  cpu_req  ram_req  cpu_remain  \\\n",
       "2     2        6        8       5915       6175      567      871        5348   \n",
       "5     2        6        8       5915       6175      567      871        5348   \n",
       "8     2        6        8       5915       6175      567      871        5348   \n",
       "11    2        6        8       5915       6175      567      871        5348   \n",
       "14    2        6        8       5915       6175      567      871        5348   \n",
       "17    2        6        8       5915       6175      567      871        5348   \n",
       "20    2        6        8       5915       6175      567      871        5348   \n",
       "23    2        6        8       5915       6175      567      871        5348   \n",
       "26    2        6        8       5915       6175      567      871        5348   \n",
       "29    2        6        8       5915       6175      567      871        5348   \n",
       "32    2        6        8       5915       6175      567      871        5348   \n",
       "35    2        6        8       5915       6175      567      871        5348   \n",
       "\n",
       "    ram_remain  cpu_usage_percent  ram_usage_percent  \n",
       "2         5304           9.585799          14.105263  \n",
       "5         5304           9.585799          14.105263  \n",
       "8         5304           9.585799          14.105263  \n",
       "11        5304           9.585799          14.105263  \n",
       "14        5304           9.585799          14.105263  \n",
       "17        5304           9.585799          14.105263  \n",
       "20        5304           9.585799          14.105263  \n",
       "23        5304           9.585799          14.105263  \n",
       "26        5304           9.585799          14.105263  \n",
       "29        5304           9.585799          14.105263  \n",
       "32        5304           9.585799          14.105263  \n",
       "35        5304           9.585799          14.105263  "
      ]
     },
     "execution_count": 68,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_dict_ms[2]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Client SLA Stats\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "((1748, 9), (1692, 9), (1744, 9))"
      ]
     },
     "execution_count": 35,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "model_accuracies = {\n",
    "    'yolov5n': 0.457,\n",
    "    'yolov5s': 0.568,\n",
    "    'yolov5m': 0.641\n",
    "}\n",
    "df_client_gai = pd.read_csv('experiment_logs/gaikube/clients_stats_msgke_20240630_115907.csv')\n",
    "df_client_vpa = pd.read_csv('experiment_logs/vpa_gke/clients_stats_msgke_20240630_144821.csv')\n",
    "df_client_ms = pd.read_csv('experiment_logs/ms_gke/clients_stats_msgke_20240630_160859.csv')\n",
    "df_client_gai.shape, df_client_vpa.shape, df_client_ms.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "metadata": {},
   "outputs": [],
   "source": [
    "df_client_gai= df_client_gai[['cid', 'model_name', 'processing_delay (ms)', 'core']]\n",
    "df_client_vpa = df_client_vpa[['cid', 'model_name', 'processing_delay (ms)', 'core']]\n",
    "df_client_ms = df_client_ms[['cid', 'model_name', 'processing_delay (ms)', 'core']]\n",
    "\n",
    "df_client_gai['model_accuracy'] = df_client_gai['model_name'].map(model_accuracies)\n",
    "df_client_vpa['model_accuracy'] = df_client_vpa['model_name'].map(model_accuracies)\n",
    "df_client_ms['model_accuracy'] = df_client_ms['model_name'].map(model_accuracies)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 241,
   "metadata": {},
   "outputs": [],
   "source": [
    "# total_traffic = df_client_gai.groupby('core').size()\n",
    "# traffic_by_core_model = df_client_gai.groupby(['core', 'model_name']).size().unstack(fill_value=0)\n",
    "# average_accuracies = {}\n",
    "# for core in traffic_by_core_model.index:\n",
    "#     total_traffic_core = total_traffic[core]\n",
    "#     weighted_sum = 0\n",
    "#     for model in traffic_by_core_model.columns:\n",
    "#         traffic_by_model = traffic_by_core_model.loc[core, model]\n",
    "#         if traffic_by_model > 0:\n",
    "#             model_accuracy = model_accuracies[model]\n",
    "#             weighted_sum += (traffic_by_model / total_traffic_core) * model_accuracy\n",
    "#     average_accuracies[core] = weighted_sum * 100  # Convert to percentage\n",
    "# average_accuracies"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([ 500, 1000, 2000])"
      ]
     },
     "execution_count": 37,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_client_gai['core'].unique()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array(['yolov5s', 'yolov5m', 'yolov5n'], dtype=object)"
      ]
     },
     "execution_count": 38,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_client_gai[df_client_gai['core']==1000]['model_name'].unique()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "core\n",
       "500     218\n",
       "1000     46\n",
       "2000      3\n",
       "dtype: int64"
      ]
     },
     "execution_count": 39,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_client_ms[df_client_ms['processing_delay (ms)'] > 700].groupby('core').size()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "core\n",
       "1000    872\n",
       "500     436\n",
       "2000    436\n",
       "Name: count, dtype: int64"
      ]
     },
     "execution_count": 40,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_client_ms['core'].value_counts()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>cid</th>\n",
       "      <th>model_name</th>\n",
       "      <th>processing_delay (ms)</th>\n",
       "      <th>core</th>\n",
       "      <th>model_accuracy</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1</td>\n",
       "      <td>yolov5s</td>\n",
       "      <td>427.858591</td>\n",
       "      <td>1000</td>\n",
       "      <td>0.568</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>1</td>\n",
       "      <td>yolov5s</td>\n",
       "      <td>678.848028</td>\n",
       "      <td>1000</td>\n",
       "      <td>0.568</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>1</td>\n",
       "      <td>yolov5s</td>\n",
       "      <td>451.465607</td>\n",
       "      <td>1000</td>\n",
       "      <td>0.568</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>13</th>\n",
       "      <td>1</td>\n",
       "      <td>yolov5s</td>\n",
       "      <td>667.403936</td>\n",
       "      <td>1000</td>\n",
       "      <td>0.568</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>17</th>\n",
       "      <td>1</td>\n",
       "      <td>yolov5s</td>\n",
       "      <td>399.151087</td>\n",
       "      <td>1000</td>\n",
       "      <td>0.568</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1725</th>\n",
       "      <td>1</td>\n",
       "      <td>yolov5n</td>\n",
       "      <td>325.529575</td>\n",
       "      <td>1000</td>\n",
       "      <td>0.457</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1729</th>\n",
       "      <td>1</td>\n",
       "      <td>yolov5n</td>\n",
       "      <td>593.114853</td>\n",
       "      <td>1000</td>\n",
       "      <td>0.457</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1733</th>\n",
       "      <td>1</td>\n",
       "      <td>yolov5n</td>\n",
       "      <td>298.827648</td>\n",
       "      <td>1000</td>\n",
       "      <td>0.457</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1737</th>\n",
       "      <td>1</td>\n",
       "      <td>yolov5n</td>\n",
       "      <td>598.983765</td>\n",
       "      <td>1000</td>\n",
       "      <td>0.457</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1741</th>\n",
       "      <td>1</td>\n",
       "      <td>yolov5n</td>\n",
       "      <td>319.641829</td>\n",
       "      <td>1000</td>\n",
       "      <td>0.457</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>436 rows × 5 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "      cid model_name  processing_delay (ms)  core  model_accuracy\n",
       "1       1    yolov5s             427.858591  1000           0.568\n",
       "5       1    yolov5s             678.848028  1000           0.568\n",
       "9       1    yolov5s             451.465607  1000           0.568\n",
       "13      1    yolov5s             667.403936  1000           0.568\n",
       "17      1    yolov5s             399.151087  1000           0.568\n",
       "...   ...        ...                    ...   ...             ...\n",
       "1725    1    yolov5n             325.529575  1000           0.457\n",
       "1729    1    yolov5n             593.114853  1000           0.457\n",
       "1733    1    yolov5n             298.827648  1000           0.457\n",
       "1737    1    yolov5n             598.983765  1000           0.457\n",
       "1741    1    yolov5n             319.641829  1000           0.457\n",
       "\n",
       "[436 rows x 5 columns]"
      ]
     },
     "execution_count": 41,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_client_ms[df_client_ms['cid']==1]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>cid</th>\n",
       "      <th>model_name</th>\n",
       "      <th>processing_delay (ms)</th>\n",
       "      <th>core</th>\n",
       "      <th>model_accuracy</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0</td>\n",
       "      <td>yolov5n</td>\n",
       "      <td>634.610653</td>\n",
       "      <td>500</td>\n",
       "      <td>0.457</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1</td>\n",
       "      <td>yolov5s</td>\n",
       "      <td>550.533772</td>\n",
       "      <td>1000</td>\n",
       "      <td>0.568</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>2</td>\n",
       "      <td>yolov5s</td>\n",
       "      <td>487.960815</td>\n",
       "      <td>1000</td>\n",
       "      <td>0.568</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>3</td>\n",
       "      <td>yolov5m</td>\n",
       "      <td>554.638147</td>\n",
       "      <td>2000</td>\n",
       "      <td>0.641</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>0</td>\n",
       "      <td>yolov5n</td>\n",
       "      <td>913.888931</td>\n",
       "      <td>500</td>\n",
       "      <td>0.457</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1743</th>\n",
       "      <td>3</td>\n",
       "      <td>yolov5m</td>\n",
       "      <td>676.305532</td>\n",
       "      <td>2000</td>\n",
       "      <td>0.641</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1744</th>\n",
       "      <td>0</td>\n",
       "      <td>yolov5m</td>\n",
       "      <td>664.591551</td>\n",
       "      <td>2000</td>\n",
       "      <td>0.641</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1745</th>\n",
       "      <td>1</td>\n",
       "      <td>yolov5s</td>\n",
       "      <td>567.901611</td>\n",
       "      <td>1000</td>\n",
       "      <td>0.568</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1746</th>\n",
       "      <td>2</td>\n",
       "      <td>yolov5s</td>\n",
       "      <td>581.035376</td>\n",
       "      <td>1000</td>\n",
       "      <td>0.568</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1747</th>\n",
       "      <td>3</td>\n",
       "      <td>yolov5m</td>\n",
       "      <td>531.518459</td>\n",
       "      <td>2000</td>\n",
       "      <td>0.641</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>1748 rows × 5 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "      cid model_name  processing_delay (ms)  core  model_accuracy\n",
       "0       0    yolov5n             634.610653   500           0.457\n",
       "1       1    yolov5s             550.533772  1000           0.568\n",
       "2       2    yolov5s             487.960815  1000           0.568\n",
       "3       3    yolov5m             554.638147  2000           0.641\n",
       "4       0    yolov5n             913.888931   500           0.457\n",
       "...   ...        ...                    ...   ...             ...\n",
       "1743    3    yolov5m             676.305532  2000           0.641\n",
       "1744    0    yolov5m             664.591551  2000           0.641\n",
       "1745    1    yolov5s             567.901611  1000           0.568\n",
       "1746    2    yolov5s             581.035376  1000           0.568\n",
       "1747    3    yolov5m             531.518459  2000           0.641\n",
       "\n",
       "[1748 rows x 5 columns]"
      ]
     },
     "execution_count": 42,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_client_gai"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>cid</th>\n",
       "      <th>model_name</th>\n",
       "      <th>processing_delay (ms)</th>\n",
       "      <th>core</th>\n",
       "      <th>model_accuracy</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0</td>\n",
       "      <td>yolov5n</td>\n",
       "      <td>472.021580</td>\n",
       "      <td>500</td>\n",
       "      <td>0.457</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1</td>\n",
       "      <td>yolov5s</td>\n",
       "      <td>466.292620</td>\n",
       "      <td>1000</td>\n",
       "      <td>0.568</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>2</td>\n",
       "      <td>yolov5s</td>\n",
       "      <td>465.268612</td>\n",
       "      <td>1000</td>\n",
       "      <td>0.568</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>3</td>\n",
       "      <td>yolov5m</td>\n",
       "      <td>488.454103</td>\n",
       "      <td>2000</td>\n",
       "      <td>0.641</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>0</td>\n",
       "      <td>yolov5n</td>\n",
       "      <td>926.285982</td>\n",
       "      <td>500</td>\n",
       "      <td>0.457</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1687</th>\n",
       "      <td>3</td>\n",
       "      <td>yolov5m</td>\n",
       "      <td>632.038355</td>\n",
       "      <td>2000</td>\n",
       "      <td>0.641</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1688</th>\n",
       "      <td>0</td>\n",
       "      <td>yolov5n</td>\n",
       "      <td>863.586664</td>\n",
       "      <td>500</td>\n",
       "      <td>0.457</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1689</th>\n",
       "      <td>1</td>\n",
       "      <td>yolov5s</td>\n",
       "      <td>413.468599</td>\n",
       "      <td>1000</td>\n",
       "      <td>0.568</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1690</th>\n",
       "      <td>2</td>\n",
       "      <td>yolov5s</td>\n",
       "      <td>356.239796</td>\n",
       "      <td>1000</td>\n",
       "      <td>0.568</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1691</th>\n",
       "      <td>3</td>\n",
       "      <td>yolov5m</td>\n",
       "      <td>451.718330</td>\n",
       "      <td>2000</td>\n",
       "      <td>0.641</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>1692 rows × 5 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "      cid model_name  processing_delay (ms)  core  model_accuracy\n",
       "0       0    yolov5n             472.021580   500           0.457\n",
       "1       1    yolov5s             466.292620  1000           0.568\n",
       "2       2    yolov5s             465.268612  1000           0.568\n",
       "3       3    yolov5m             488.454103  2000           0.641\n",
       "4       0    yolov5n             926.285982   500           0.457\n",
       "...   ...        ...                    ...   ...             ...\n",
       "1687    3    yolov5m             632.038355  2000           0.641\n",
       "1688    0    yolov5n             863.586664   500           0.457\n",
       "1689    1    yolov5s             413.468599  1000           0.568\n",
       "1690    2    yolov5s             356.239796  1000           0.568\n",
       "1691    3    yolov5m             451.718330  2000           0.641\n",
       "\n",
       "[1692 rows x 5 columns]"
      ]
     },
     "execution_count": 43,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_client_vpa"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 72,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>cid</th>\n",
       "      <th>SLA_violation_percentage</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0</td>\n",
       "      <td>1.372998</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1</td>\n",
       "      <td>2.974828</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>2</td>\n",
       "      <td>8.237986</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>3</td>\n",
       "      <td>1.144165</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   cid  SLA_violation_percentage\n",
       "0    0                  1.372998\n",
       "1    1                  2.974828\n",
       "2    2                  8.237986\n",
       "3    3                  1.144165"
      ]
     },
     "execution_count": 72,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "def violation_calculator(df, threshold):\n",
    "    violations = df[df['processing_delay (ms)'] > threshold]\n",
    "    return (violations.shape[0]/df.shape[0])*100\n",
    "\n",
    "# latency_bound = 700\n",
    "# vr = violation_calculator(df_client_vpa,latency_bound)\n",
    "\n",
    "def per_cid_violation_calculator(df, threshold):\n",
    "    result = df.groupby('cid').apply(lambda x: violation_calculator(x, threshold)).reset_index(name='SLA_violation_percentage')\n",
    "    return result\n",
    "\n",
    "latency_bound = 700\n",
    "vr_gai= per_cid_violation_calculator(df_client_gai, latency_bound)\n",
    "vr_vpa = per_cid_violation_calculator(df_client_vpa, latency_bound)\n",
    "vr_ms = per_cid_violation_calculator(df_client_ms, latency_bound)\n",
    "vr_gai"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 73,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>cid</th>\n",
       "      <th>SLA_violation_percentage</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0</td>\n",
       "      <td>49.881797</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1</td>\n",
       "      <td>8.747045</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>2</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>3</td>\n",
       "      <td>0.472813</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   cid  SLA_violation_percentage\n",
       "0    0                 49.881797\n",
       "1    1                  8.747045\n",
       "2    2                  0.000000\n",
       "3    3                  0.472813"
      ]
     },
     "execution_count": 73,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "vr_vpa"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 74,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>cid</th>\n",
       "      <th>SLA_violation_percentage</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0</td>\n",
       "      <td>50.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1</td>\n",
       "      <td>3.899083</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>2</td>\n",
       "      <td>6.651376</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>3</td>\n",
       "      <td>0.688073</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   cid  SLA_violation_percentage\n",
       "0    0                 50.000000\n",
       "1    1                  3.899083\n",
       "2    2                  6.651376\n",
       "3    3                  0.688073"
      ]
     },
     "execution_count": 74,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "vr_ms"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 76,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>cid</th>\n",
       "      <th>gai_violations</th>\n",
       "      <th>vpa_violations</th>\n",
       "      <th>ms_violations</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0</td>\n",
       "      <td>1.372998</td>\n",
       "      <td>49.881797</td>\n",
       "      <td>50.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1</td>\n",
       "      <td>2.974828</td>\n",
       "      <td>8.747045</td>\n",
       "      <td>3.899083</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>2</td>\n",
       "      <td>8.237986</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>6.651376</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>3</td>\n",
       "      <td>1.144165</td>\n",
       "      <td>0.472813</td>\n",
       "      <td>0.688073</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   cid  gai_violations  vpa_violations  ms_violations\n",
       "0    0        1.372998       49.881797      50.000000\n",
       "1    1        2.974828        8.747045       3.899083\n",
       "2    2        8.237986        0.000000       6.651376\n",
       "3    3        1.144165        0.472813       0.688073"
      ]
     },
     "execution_count": 76,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "vr_gai['vpa_violations'] = vr_vpa['SLA_violation_percentage']\n",
    "vr_gai['ms_violations'] = vr_ms['SLA_violation_percentage']\n",
    "vr_gai.rename(columns={\"SLA_violation_percentage\": \"gai_violations\"}, inplace=True)\n",
    "vr_gai"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 138,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 600x400 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig, ax = plt.subplots(figsize=(6, 4))\n",
    "\n",
    "# Set the bar width\n",
    "bar_width = 0.25\n",
    "\n",
    "# Set positions for the bars\n",
    "r1 = range(len(vr_gai['cid']))\n",
    "r2 = [x + bar_width for x in r1]\n",
    "r3 = [x + bar_width for x in r2]\n",
    "\n",
    "# Create the bars\n",
    "ax.bar(r1, vr_gai['gai_violations'], width=bar_width, label='GAIKube')\n",
    "ax.bar(r2, vr_gai['vpa_violations'], width=bar_width, label='SVG')\n",
    "ax.bar(r3, vr_gai['ms_violations'], width=bar_width, label='MVG')\n",
    "\n",
    "# Add labels and title\n",
    "ax.set_xlabel('Container Number')\n",
    "ax.set_ylabel('Violation [%]')\n",
    "#ax.set_title('Violations per CID')\n",
    "ax.set_xticks([r + bar_width for r in range(len(vr_gai['cid']))])\n",
    "ax.set_xticklabels(vr_gai['cid'])\n",
    "\n",
    "ax.grid()\n",
    "ax.legend()\n",
    "plt.ylim([0,100])\n",
    "plt.savefig('per_contaier_violations.pdf', format='pdf')\n",
    "# Show plot\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 47,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "3.4324942791762014"
      ]
     },
     "execution_count": 47,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "violation_calculator(df_client_gai, latency_bound)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 48,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "14.775413711583923"
      ]
     },
     "execution_count": 48,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "violation_calculator(df_client_vpa, latency_bound)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 49,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "15.309633027522937"
      ]
     },
     "execution_count": 49,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "violation_calculator(df_client_ms, latency_bound)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Accuracy Calculator"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 98,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>cid</th>\n",
       "      <th>model_name</th>\n",
       "      <th>processing_delay (ms)</th>\n",
       "      <th>core</th>\n",
       "      <th>model_accuracy</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0</td>\n",
       "      <td>yolov5n</td>\n",
       "      <td>472.021580</td>\n",
       "      <td>500</td>\n",
       "      <td>0.457</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1</td>\n",
       "      <td>yolov5s</td>\n",
       "      <td>466.292620</td>\n",
       "      <td>1000</td>\n",
       "      <td>0.568</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>2</td>\n",
       "      <td>yolov5s</td>\n",
       "      <td>465.268612</td>\n",
       "      <td>1000</td>\n",
       "      <td>0.568</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>3</td>\n",
       "      <td>yolov5m</td>\n",
       "      <td>488.454103</td>\n",
       "      <td>2000</td>\n",
       "      <td>0.641</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>0</td>\n",
       "      <td>yolov5n</td>\n",
       "      <td>926.285982</td>\n",
       "      <td>500</td>\n",
       "      <td>0.457</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>1</td>\n",
       "      <td>yolov5s</td>\n",
       "      <td>674.696922</td>\n",
       "      <td>1000</td>\n",
       "      <td>0.568</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>2</td>\n",
       "      <td>yolov5s</td>\n",
       "      <td>601.025820</td>\n",
       "      <td>1000</td>\n",
       "      <td>0.568</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>3</td>\n",
       "      <td>yolov5m</td>\n",
       "      <td>628.035069</td>\n",
       "      <td>2000</td>\n",
       "      <td>0.641</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>0</td>\n",
       "      <td>yolov5n</td>\n",
       "      <td>328.474998</td>\n",
       "      <td>500</td>\n",
       "      <td>0.457</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>1</td>\n",
       "      <td>yolov5s</td>\n",
       "      <td>496.792078</td>\n",
       "      <td>1000</td>\n",
       "      <td>0.568</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10</th>\n",
       "      <td>2</td>\n",
       "      <td>yolov5s</td>\n",
       "      <td>368.584633</td>\n",
       "      <td>1000</td>\n",
       "      <td>0.568</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11</th>\n",
       "      <td>3</td>\n",
       "      <td>yolov5m</td>\n",
       "      <td>529.941797</td>\n",
       "      <td>2000</td>\n",
       "      <td>0.641</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "    cid model_name  processing_delay (ms)  core  model_accuracy\n",
       "0     0    yolov5n             472.021580   500           0.457\n",
       "1     1    yolov5s             466.292620  1000           0.568\n",
       "2     2    yolov5s             465.268612  1000           0.568\n",
       "3     3    yolov5m             488.454103  2000           0.641\n",
       "4     0    yolov5n             926.285982   500           0.457\n",
       "5     1    yolov5s             674.696922  1000           0.568\n",
       "6     2    yolov5s             601.025820  1000           0.568\n",
       "7     3    yolov5m             628.035069  2000           0.641\n",
       "8     0    yolov5n             328.474998   500           0.457\n",
       "9     1    yolov5s             496.792078  1000           0.568\n",
       "10    2    yolov5s             368.584633  1000           0.568\n",
       "11    3    yolov5m             529.941797  2000           0.641"
      ]
     },
     "execution_count": 98,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_client_vpa.head(12)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 101,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>cid</th>\n",
       "      <th>model_accuracy</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0</td>\n",
       "      <td>0.638895</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1</td>\n",
       "      <td>0.560105</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>2</td>\n",
       "      <td>0.568773</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>3</td>\n",
       "      <td>0.641000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   cid  model_accuracy\n",
       "0    0        0.638895\n",
       "1    1        0.560105\n",
       "2    2        0.568773\n",
       "3    3        0.641000"
      ]
     },
     "execution_count": 101,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "acc_cid_gai = df_client_gai.groupby('cid')['model_accuracy'].mean().reset_index()\n",
    "acc_cid_gai"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 100,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>cid</th>\n",
       "      <th>vpa_accuracy</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0</td>\n",
       "      <td>0.457</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1</td>\n",
       "      <td>0.568</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>2</td>\n",
       "      <td>0.568</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>3</td>\n",
       "      <td>0.641</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   cid  vpa_accuracy\n",
       "0    0         0.457\n",
       "1    1         0.568\n",
       "2    2         0.568\n",
       "3    3         0.641"
      ]
     },
     "execution_count": 100,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "acc_cid_vpa = df_client_vpa.groupby('cid')['model_accuracy'].mean().reset_index()\n",
    "acc_cid_vpa.columns = ['cid', 'vpa_accuracy']\n",
    "acc_cid_vpa"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 97,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>cid</th>\n",
       "      <th>ms_accuracy</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0</td>\n",
       "      <td>0.457000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1</td>\n",
       "      <td>0.546106</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>2</td>\n",
       "      <td>0.543961</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>3</td>\n",
       "      <td>0.641000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   cid  ms_accuracy\n",
       "0    0     0.457000\n",
       "1    1     0.546106\n",
       "2    2     0.543961\n",
       "3    3     0.641000"
      ]
     },
     "execution_count": 97,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "acc_cid_ms = df_client_ms.groupby('cid')['model_accuracy'].mean().reset_index()\n",
    "acc_cid_ms.columns = ['cid', 'ms_accuracy']\n",
    "acc_cid_ms"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 334,
   "metadata": {},
   "outputs": [],
   "source": [
    "traffic_by_core_model = df_client_ms.groupby(['cid', 'model_name']).size().unstack(fill_value=0)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 335,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th>model_name</th>\n",
       "      <th>yolov5m</th>\n",
       "      <th>yolov5n</th>\n",
       "      <th>yolov5s</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>cid</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0</td>\n",
       "      <td>436</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>0</td>\n",
       "      <td>86</td>\n",
       "      <td>350</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>10</td>\n",
       "      <td>101</td>\n",
       "      <td>325</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>436</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "model_name  yolov5m  yolov5n  yolov5s\n",
       "cid                                  \n",
       "0                 0      436        0\n",
       "1                 0       86      350\n",
       "2                10      101      325\n",
       "3               436        0        0"
      ]
     },
     "execution_count": 335,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "traffic_by_core_model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 104,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>cid</th>\n",
       "      <th>gai_acc</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0</td>\n",
       "      <td>63.889474</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1</td>\n",
       "      <td>56.010526</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>2</td>\n",
       "      <td>56.877346</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>3</td>\n",
       "      <td>64.100000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   cid    gai_acc\n",
       "0    0  63.889474\n",
       "1    1  56.010526\n",
       "2    2  56.877346\n",
       "3    3  64.100000"
      ]
     },
     "execution_count": 104,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "total_traffic = df_client_gai.groupby('cid').size()\n",
    "traffic_by_core_model = df_client_gai.groupby(['cid', 'model_name']).size().unstack(fill_value=0)\n",
    "average_accuracies = {}\n",
    "for cid in traffic_by_core_model.index:\n",
    "    total_traffic_core = total_traffic[cid]\n",
    "    weighted_sum = 0\n",
    "    for model in traffic_by_core_model.columns:\n",
    "        traffic_by_model = traffic_by_core_model.loc[cid, model]\n",
    "        if traffic_by_model > 0:\n",
    "            model_accuracy = model_accuracies[model]\n",
    "            weighted_sum += (traffic_by_model / total_traffic_core) * model_accuracy\n",
    "    average_accuracies[cid] = weighted_sum * 100  # Convert to percentage\n",
    "\n",
    "gai_acc = pd.DataFrame(list(average_accuracies.items()), columns=['cid', 'gai_acc'])\n",
    "\n",
    "gai_acc"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 105,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>cid</th>\n",
       "      <th>vpa_acc</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0</td>\n",
       "      <td>45.7</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1</td>\n",
       "      <td>56.8</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>2</td>\n",
       "      <td>56.8</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>3</td>\n",
       "      <td>64.1</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   cid  vpa_acc\n",
       "0    0     45.7\n",
       "1    1     56.8\n",
       "2    2     56.8\n",
       "3    3     64.1"
      ]
     },
     "execution_count": 105,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Calculate average accuracy per cid\n",
    "total_traffic = df_client_vpa.groupby('cid').size()\n",
    "traffic_by_core_model = df_client_vpa.groupby(['cid', 'model_name']).size().unstack(fill_value=0)\n",
    "average_accuracies = {}\n",
    "for cid in traffic_by_core_model.index:\n",
    "    total_traffic_cid = traffic_by_core_model.loc[cid].sum()\n",
    "    weighted_sum = 0\n",
    "    for model in traffic_by_core_model.columns:\n",
    "        traffic_by_model = traffic_by_core_model.loc[cid, model]\n",
    "        if traffic_by_model > 0:\n",
    "            model_accuracy = model_accuracies[model]\n",
    "            weighted_sum += (traffic_by_model / total_traffic_cid) * model_accuracy\n",
    "    average_accuracies[cid] = weighted_sum * 100  # Convert to percentage\n",
    "\n",
    "# Convert the result to a dataframe\n",
    "vpa_acc = pd.DataFrame(list(average_accuracies.items()), columns=['cid', 'vpa_acc'])\n",
    "\n",
    "vpa_acc"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 106,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>cid</th>\n",
       "      <th>ms_acc</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0</td>\n",
       "      <td>45.700000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1</td>\n",
       "      <td>54.610550</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>2</td>\n",
       "      <td>54.396101</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>3</td>\n",
       "      <td>64.100000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   cid     ms_acc\n",
       "0    0  45.700000\n",
       "1    1  54.610550\n",
       "2    2  54.396101\n",
       "3    3  64.100000"
      ]
     },
     "execution_count": 106,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Calculate average accuracy per cid\n",
    "total_traffic = df_client_ms.groupby('cid').size()\n",
    "traffic_by_core_model = df_client_ms.groupby(['cid', 'model_name']).size().unstack(fill_value=0)\n",
    "average_accuracies = {}\n",
    "for cid in traffic_by_core_model.index:\n",
    "    total_traffic_cid = traffic_by_core_model.loc[cid].sum()\n",
    "    weighted_sum = 0\n",
    "    for model in traffic_by_core_model.columns:\n",
    "        traffic_by_model = traffic_by_core_model.loc[cid, model]\n",
    "        if traffic_by_model > 0:\n",
    "            model_accuracy = model_accuracies[model]\n",
    "            weighted_sum += (traffic_by_model / total_traffic_cid) * model_accuracy\n",
    "    average_accuracies[cid] = weighted_sum * 100  # Convert to percentage\n",
    "\n",
    "# Convert the result to a dataframe\n",
    "ms_acc = pd.DataFrame(list(average_accuracies.items()), columns=['cid', 'ms_acc'])\n",
    "ms_acc"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 108,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>cid</th>\n",
       "      <th>gai_acc</th>\n",
       "      <th>vpa_acc</th>\n",
       "      <th>ms_acc</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0</td>\n",
       "      <td>63.889474</td>\n",
       "      <td>45.7</td>\n",
       "      <td>45.700000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1</td>\n",
       "      <td>56.010526</td>\n",
       "      <td>56.8</td>\n",
       "      <td>54.610550</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>2</td>\n",
       "      <td>56.877346</td>\n",
       "      <td>56.8</td>\n",
       "      <td>54.396101</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>3</td>\n",
       "      <td>64.100000</td>\n",
       "      <td>64.1</td>\n",
       "      <td>64.100000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   cid    gai_acc  vpa_acc     ms_acc\n",
       "0    0  63.889474     45.7  45.700000\n",
       "1    1  56.010526     56.8  54.610550\n",
       "2    2  56.877346     56.8  54.396101\n",
       "3    3  64.100000     64.1  64.100000"
      ]
     },
     "execution_count": 108,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "gai_acc['vpa_acc'] = vpa_acc['vpa_acc']\n",
    "gai_acc['ms_acc'] = ms_acc['ms_acc']\n",
    "gai_acc"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 114,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(45.7, 64.1)"
      ]
     },
     "execution_count": 114,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "min(model_accuracies.values())*100, max(model_accuracies.values())*100"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 126,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "60.21933638443936"
      ]
     },
     "execution_count": 126,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "sum(gai_acc['gai_acc'])/gai_acc.shape[0]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 127,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "55.85"
      ]
     },
     "execution_count": 127,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "sum(gai_acc['vpa_acc'])/gai_acc.shape[0]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 128,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "54.7016628440367"
      ]
     },
     "execution_count": 128,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "sum(gai_acc['ms_acc'])/gai_acc.shape[0]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 125,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(4, 4)"
      ]
     },
     "execution_count": 125,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "gai_acc.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 137,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 600x400 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig, ax = plt.subplots(figsize=(6, 4))\n",
    "\n",
    "# Set the bar width\n",
    "bar_width = 0.25\n",
    "\n",
    "# Set positions for the bars\n",
    "r1 = range(len(gai_acc['cid']))\n",
    "r2 = [x + bar_width for x in r1]\n",
    "r3 = [x + bar_width for x in r2]\n",
    "\n",
    "# Create the bars\n",
    "ax.bar(r1, gai_acc['gai_acc'],  width=bar_width, label='GAIKube')\n",
    "ax.bar(r2, gai_acc['vpa_acc'],  width=bar_width, label='SVG')\n",
    "ax.bar(r3, gai_acc['ms_acc'],  width=bar_width, label='MVG')\n",
    "\n",
    "# Add labels and title\n",
    "ax.set_xlabel('Container Number')\n",
    "ax.set_ylabel('Yolo5 Accuracy (%)')\n",
    "#ax.set_title('Model Accuracies per CID')\n",
    "ax.set_xticks([r + bar_width for r in range(len(gai_acc['cid']))])\n",
    "ax.set_xticklabels(gai_acc['cid'])\n",
    "\n",
    "# Add legend\n",
    "ax.legend()\n",
    "\n",
    "# Add horizontal lines\n",
    "# Add horizontal lines\n",
    "min_line = 45.7\n",
    "max_line = 64.1\n",
    "ax.axhline(y=min_line, color='black', linestyle='--', linewidth=1)\n",
    "ax.axhline(y=max_line, color='black', linestyle='--', linewidth=1)\n",
    "\n",
    "# Add labels to the horizontal lines\n",
    "ax.text(len(gai_acc['cid']) - 0.5, min_line, 'Min', verticalalignment='bottom', horizontalalignment='right', color='black', fontsize=10, fontweight='bold')\n",
    "ax.text(len(gai_acc['cid']) - 0.5, max_line, 'Max', verticalalignment='bottom', horizontalalignment='right', color='black', fontsize=10, fontweight='bold')\n",
    "\n",
    "\n",
    "#plt.grid()\n",
    "plt.ylim([0,100])\n",
    "plt.savefig('yolo_mean_acc.pdf', format='pdf')\n",
    "# Show plot\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "bitbrain",
   "language": "python",
   "name": "bitbrain"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.11.4"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}