plotting.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "import uproot\n",
    "import matplotlib.pyplot as plt\n",
    "import numpy as np\n",
    "import pandas as pd\n",
    "import utils\n",
    "import os"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "num_agents = 200\n",
    "num_iterations = 150\n",
    "# num_params = 61\n",
    "num_params = 25"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "default = utils.read_csv('checkpoint/default.csv')[0]\n",
    "default_params = default[:num_params]\n",
    "default_metrics = default[num_params:]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "metrics = [pd.read_csv('history/iteration' + str(i) + '.csv', header=None, usecols=[num_params, num_params + 1]).transpose().to_numpy()\n",
    "           for i in range(num_iterations)]\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "import matplotlib.animation as animation\n",
    "\n",
    "fig, ax = plt.subplots()\n",
    "\n",
    "def animate(i):\n",
    "    fig.clear()\n",
    "    ax = fig.add_subplot(111)\n",
    "    ax.set_xlim(0., 0.1)\n",
    "    ax.set_ylim(0.6, 0.8)\n",
    "    # s = ax.scatter(metrics[0][1], 1 - metrics[0][0], s=10)\n",
    "    s = ax.scatter(metrics[i][1], 1 - metrics[i][0], s=5, color='orchid', label='particles')\n",
    "    # s = ax.scatter(metrics[num_iterations - 1][1], 1 - metrics[num_iterations - 1][0], c='green', s=10)\n",
    "    s = ax.scatter([default_metrics[1]], [1 - default_metrics[0]], marker='s', color='blue', s=10, label='default')\n",
    "    # ax.set_xlabel(r'fakes $=\\frac{(N_{rec}-N_{ass})}{N_{rec}}$')\n",
    "    # ax.set_ylabel(r'eff $=\\frac{N_{ass}}{N_{sim}}$')\n",
    "    # ax.legend(loc='best')\n",
    "    ax.set_xlabel(\"Fake + duplicate rate\")\n",
    "    ax.set_ylabel(\"Efficiency\")\n",
    "    ax.legend()\n",
    "    ax.set_title(str(i))\n",
    "\n",
    "ani=animation.FuncAnimation(fig, animate, interval=200, frames=range(num_iterations))\n",
    "ani.save('checkpoint/metrics.gif', writer='pillow')\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [],
   "source": [
    "pareto_front = utils.read_csv('checkpoint/pareto_front.csv')\n",
    "pareto_front = pareto_front[pareto_front[:, num_params + 1].argsort()]\n",
    "pareto_x = [particle[num_params + 1] for particle in pareto_front]\n",
    "\n",
    "pareto_y = [1 - particle[num_params] for particle in pareto_front]\n",
    "# pareto_z = [particle[6] for particle in pareto_front]\n",
    "# pareto_t = [particle[5] + particle[6] for particle in pareto_front]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "598"
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(pareto_front)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 47,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "point1 = pareto_front[200]\n",
    "point2 = pareto_front[350]\n",
    "point3 = pareto_front[420]\n",
    "\n",
    "\n",
    "plt.scatter(pareto_x, pareto_y, s=5, color='turquoise', label='pareto front')\n",
    "\n",
    "plt.scatter(default_metrics[1], 1 - default_metrics[0], marker='s', color='blue', s=10, label='default')\n",
    "plt.scatter([point1[num_params + 1]], [1 - point1[num_params]], color='red', s=10, label='sample1')\n",
    "plt.scatter([point2[num_params + 1]], [1 - point2[num_params]], marker='^', color='black', s=10, label='sample2')\n",
    "plt.scatter([point3[num_params + 1]], [1 - point3[num_params]], marker='P', color='coral', s=15, label='sample3')\n",
    "plt.legend()\n",
    "\n",
    "# plt.scatter([point1[5] + point1[6], point2[5] + point2[6], point3[5] + point3[6]], [1 - point1[4], 1 - point2[4], 1 - point3[4]], color='red', s=8)\n",
    "\n",
    "# plt.scatter(point1[5], 1 - point1[4], color='red', s=8)\n",
    "plt.xlim(0.0, 0.1)\n",
    "plt.ylim(0.6, 0.8)\n",
    "plt.xlabel('Fake + duplicate rate')\n",
    "plt.ylabel('Efficiency')\n",
    "plt.title('Pareto Front')\n",
    "plt.savefig('checkpoint/pf.png')\n",
    "plt.show()\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 48,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[0.001810662567296451, 0.003087555630397035, 0.09246714653350764, 0.18152341708025987, 0.494168726360568, 14.138457372816534, 586.056241705347, 572.609333810121, 799.7658913490137, 529.236956888209, 781.2205619188171, 659.3251159558248, 685.3600247218071, 711.798504877761, 720.5207379590859, 607.1881073679643, 812.2225118513732, 754.319867935467, 746.914171448225, 566.801805795892, 642.8942235801952, 764.931566933547, 695.7574286025573, 685.6607185592361, 529.5465868725981]\n",
      "[0.002026680601574826, 0.003667961007414807, 0.13973166431929473, 0.2601286664785727, 0.6234191906795576, 12.748718895130501, 611.3509687106036, 599.0906311831084, 767.7894986875152, 554.1099289081602, 822.405233723454, 609.5225565238915, 654.9684509130346, 789.3707925441487, 781.3712354553714, 696.0467818105441, 767.6130509993511, 852.5787960669536, 758.7497788361125, 603.9374646454572, 681.0987713007432, 680.1745133327155, 627.7336421012832, 668.1327971090925, 664.1902624590334]\n",
      "[0.002270949818772689, 0.004548690015636908, 0.17495492817160524, 0.3026367397959566, 0.7312049631773163, 15.111468981530141, 701.9690695565533, 589.7385893770403, 782.9888558864833, 812.7232053168118, 691.9381445872277, 652.7402585639202, 691.4101496709853, 843.0526952411143, 912.6146104722875, 801.6729614334461, 751.7443955258144, 728.1079026643821, 713.1276878112122, 783.8495883443825, 783.1591843764348, 734.1050921181821, 601.1224345536785, 507.71003352463447, 754.106120229702]\n"
     ]
    }
   ],
   "source": [
    "print([point1[i] for i in range(num_params)])\n",
    "print([point2[i] for i in range(num_params)])\n",
    "print([point3[i] for i in range(num_params)])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 49,
   "metadata": {},
   "outputs": [],
   "source": [
    "selected_params = [default_params, point1[:num_params], point2[:num_params], point3[:num_params]]\n",
    "# selected_params"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 45,
   "metadata": {},
   "outputs": [],
   "source": [
    "utils.write_csv('checkpoint/selected_params.csv', selected_params)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [],
   "source": [
    "# print(point1)\n",
    "# print(point2)\n",
    "# print(point3)"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "cms",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.9.16"
  },
  "orig_nbformat": 4
 },
 "nbformat": 4,
 "nbformat_minor": 2
}