cmc_curve.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "import matplotlib.pyplot as plt\n",
    "from scipy.interpolate import interp1d\n",
    "import re\n",
    "import ast\n",
    "\n",
    "%matplotlib inline"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAtkAAAEaCAYAAADE2UsZAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nZmZmZjlzkG1mZmZmlrP/A0iQt3jzXhxVAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x866ed30>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "import matplotlib.pyplot as plt\n",
    "import numpy as np\n",
    "\n",
    "f, axarr = plt.subplots(2, 3)\n",
    "x = range(1, 50)\n",
    "\n",
    "#------------------------------------------------------------------------------------\n",
    "file_scores = open('fusion_data/leaf_flower_top_50_scores.txt', 'r')\n",
    "for line in file_scores:\n",
    "    if line.split(':')[0] == 'max rule':\n",
    "        y_max = ast.literal_eval(line.split(':')[1].strip())\n",
    "    elif line.split(':')[0] == 'sum rule':\n",
    "        y_sum = ast.literal_eval(line.split(':')[1].strip())\n",
    "    elif line.split(':')[0] == 'prod rule':\n",
    "        y_prod = ast.literal_eval(line.split(':')[1].strip())\n",
    "    elif line.split(':')[0] == 'leaf':\n",
    "        y_or1 = ast.literal_eval(line.split(':')[1].strip())\n",
    "    elif line.split(':')[0] == 'flower':\n",
    "        y_or2 = ast.literal_eval(line.split(':')[1].strip())\n",
    "    elif line.split(':')[0] == 'rh rule':\n",
    "        y_rh = ast.literal_eval(line.split(':')[1].strip())\n",
    "    else:\n",
    "        y_svm = ast.literal_eval(line.split(':')[1].strip())\n",
    "file_scores.close()\n",
    "\n",
    "f_svm = interp1d(x, y_svm,  kind='cubic')\n",
    "f_max = interp1d(x, y_max,  kind='cubic')\n",
    "f_sum = interp1d(x, y_sum,  kind='cubic')\n",
    "f_prod = interp1d(x, y_prod,  kind='cubic')\n",
    "f_rh = interp1d(x, y_rh,  kind='cubic')\n",
    "f_or1 = interp1d(x, y_or1,  kind='cubic')\n",
    "f_or2 = interp1d(x, y_or2,  kind='cubic')\n",
    "\n",
    "l1, l2, l3, l4, l5, l6, l7 = axarr[0, 0].plot(x, f_prod(x), '-', x, f_sum(x), '-', x, f_rh(x), '-',\n",
    "                 x, f_svm(x), '-', x, f_max(x), '-', x, f_or1(x), '-', x, f_or2(x), '-')\n",
    "axarr[0, 0].set_title('leaf vs flower')\n",
    "axarr[0, 0].set_ylim([0.3,1])\n",
    "axarr[0, 0].set_xlim([0,20])\n",
    "\n",
    "#------------------------------------------------------------------------------------\n",
    "file_scores = open('fusion_data/leaf_entire_top_50_scores.txt', 'r')\n",
    "for line in file_scores:\n",
    "    if line.split(':')[0] == 'max rule':\n",
    "        y_max = ast.literal_eval(line.split(':')[1].strip())\n",
    "    elif line.split(':')[0] == 'sum rule':\n",
    "        y_sum = ast.literal_eval(line.split(':')[1].strip())\n",
    "    elif line.split(':')[0] == 'prod rule':\n",
    "        y_prod = ast.literal_eval(line.split(':')[1].strip())\n",
    "    elif line.split(':')[0] == 'leaf':\n",
    "        y_or1 = ast.literal_eval(line.split(':')[1].strip())\n",
    "    elif line.split(':')[0] == 'entire':\n",
    "        y_or2 = ast.literal_eval(line.split(':')[1].strip())\n",
    "    elif line.split(':')[0] == 'rh rule':\n",
    "        y_rh = ast.literal_eval(line.split(':')[1].strip())\n",
    "    else:\n",
    "        y_svm = ast.literal_eval(line.split(':')[1].strip())\n",
    "file_scores.close()\n",
    "\n",
    "f_svm = interp1d(x, y_svm,  kind='cubic')\n",
    "f_max = interp1d(x, y_max,  kind='cubic')\n",
    "f_sum = interp1d(x, y_sum,  kind='cubic')\n",
    "f_prod = interp1d(x, y_prod,  kind='cubic')\n",
    "f_rh = interp1d(x, y_rh,  kind='cubic')\n",
    "f_or1 = interp1d(x, y_or1,  kind='cubic')\n",
    "f_or2 = interp1d(x, y_or2,  kind='cubic')\n",
    "\n",
    "axarr[0, 1].plot(x, f_prod(x), '-', x, f_sum(x), '-', x, f_rh(x), '-',\n",
    "                 x, f_svm(x), '-', x, f_max(x), '-', x, f_or1(x), '-', x, f_or2(x), '-')\n",
    "axarr[0, 1].set_title('leaf vs entire')\n",
    "axarr[0, 1].set_ylim([0.3,1])\n",
    "axarr[0, 1].set_xlim([0,20])\n",
    "\n",
    "#------------------------------------------------------------------------------------\n",
    "file_scores = open('fusion_data/entire_flower_top_50_scores.txt', 'r')\n",
    "for line in file_scores:\n",
    "    if line.split(':')[0] == 'max rule':\n",
    "        y_max = ast.literal_eval(line.split(':')[1].strip())\n",
    "    elif line.split(':')[0] == 'sum rule':\n",
    "        y_sum = ast.literal_eval(line.split(':')[1].strip())\n",
    "    elif line.split(':')[0] == 'prod rule':\n",
    "        y_prod = ast.literal_eval(line.split(':')[1].strip())\n",
    "    elif line.split(':')[0] == 'entire':\n",
    "        y_or1 = ast.literal_eval(line.split(':')[1].strip())\n",
    "    elif line.split(':')[0] == 'flower':\n",
    "        y_or2 = ast.literal_eval(line.split(':')[1].strip())\n",
    "    elif line.split(':')[0] == 'rh rule':\n",
    "        y_rh = ast.literal_eval(line.split(':')[1].strip())\n",
    "    else:\n",
    "        y_svm = ast.literal_eval(line.split(':')[1].strip())\n",
    "file_scores.close()\n",
    "f_svm = interp1d(x, y_svm,  kind='cubic')\n",
    "f_max = interp1d(x, y_max,  kind='cubic')\n",
    "f_sum = interp1d(x, y_sum,  kind='cubic')\n",
    "f_prod = interp1d(x, y_prod,  kind='cubic')\n",
    "f_rh = interp1d(x, y_rh,  kind='cubic')\n",
    "f_or1 = interp1d(x, y_or1,  kind='cubic')\n",
    "f_or2 = interp1d(x, y_or2,  kind='cubic')\n",
    "axarr[1, 0].plot(x, f_prod(x), '-', x, f_sum(x), '-', x, f_rh(x), '-',\n",
    "                 x, f_svm(x), '-', x, f_max(x), '-', x, f_or1(x), '-', x, f_or2(x), '-')\n",
    "axarr[1, 0].set_title('entire vs flower')\n",
    "axarr[1, 0].set_ylim([0.3,1])\n",
    "axarr[1, 0].set_xlim([0,20])\n",
    "\n",
    "#------------------------------------------------------------------------------------\n",
    "file_scores = open('fusion_data/branch_leaf_top_50_scores.txt', 'r')\n",
    "for line in file_scores:\n",
    "    if line.split(':')[0] == 'max rule':\n",
    "        y_max = ast.literal_eval(line.split(':')[1].strip())\n",
    "    elif line.split(':')[0] == 'sum rule':\n",
    "        y_sum = ast.literal_eval(line.split(':')[1].strip())\n",
    "    elif line.split(':')[0] == 'prod rule':\n",
    "        y_prod = ast.literal_eval(line.split(':')[1].strip())\n",
    "    elif line.split(':')[0] == 'branch':\n",
    "        y_or1 = ast.literal_eval(line.split(':')[1].strip())\n",
    "    elif line.split(':')[0] == 'leaf':\n",
    "        y_or2 = ast.literal_eval(line.split(':')[1].strip())\n",
    "    elif line.split(':')[0] == 'rh rule':\n",
    "        y_rh = ast.literal_eval(line.split(':')[1].strip())\n",
    "    else:\n",
    "        y_svm = ast.literal_eval(line.split(':')[1].strip())\n",
    "file_scores.close()\n",
    "f_svm = interp1d(x, y_svm,  kind='cubic')\n",
    "f_max = interp1d(x, y_max,  kind='cubic')\n",
    "f_sum = interp1d(x, y_sum,  kind='cubic')\n",
    "f_prod = interp1d(x, y_prod,  kind='cubic')\n",
    "f_rh = interp1d(x, y_rh,  kind='cubic')\n",
    "f_or1 = interp1d(x, y_or1,  kind='cubic')\n",
    "f_or2 = interp1d(x, y_or2,  kind='cubic')\n",
    "axarr[1, 1].plot(x, f_prod(x), '-', x, f_sum(x), '-', x, f_rh(x), '-',\n",
    "                 x, f_svm(x), '-', x, f_max(x), '-', x, f_or1(x), '-', x, f_or2(x), '-')\n",
    "axarr[1, 1].set_title('branch vs leaf')\n",
    "axarr[1, 1].set_ylim([0.3,1])\n",
    "axarr[1, 1].set_xlim([0,20])\n",
    "\n",
    "#------------------------------------------------------------------------------------\n",
    "file_scores = open('fusion_data/branch_flower_top_50_scores.txt', 'r')\n",
    "for line in file_scores:\n",
    "    if line.split(':')[0] == 'max rule':\n",
    "        y_max = ast.literal_eval(line.split(':')[1].strip())\n",
    "    elif line.split(':')[0] == 'sum rule':\n",
    "        y_sum = ast.literal_eval(line.split(':')[1].strip())\n",
    "    elif line.split(':')[0] == 'prod rule':\n",
    "        y_prod = ast.literal_eval(line.split(':')[1].strip())\n",
    "    elif line.split(':')[0] == 'branch':\n",
    "        y_or1 = ast.literal_eval(line.split(':')[1].strip())\n",
    "    elif line.split(':')[0] == 'flower':\n",
    "        y_or2 = ast.literal_eval(line.split(':')[1].strip())\n",
    "    elif line.split(':')[0] == 'rh rule':\n",
    "        y_rh = ast.literal_eval(line.split(':')[1].strip())\n",
    "    else:\n",
    "        y_svm = ast.literal_eval(line.split(':')[1].strip())\n",
    "file_scores.close()\n",
    "f_svm = interp1d(x, y_svm,  kind='cubic')\n",
    "f_max = interp1d(x, y_max,  kind='cubic')\n",
    "f_sum = interp1d(x, y_sum,  kind='cubic')\n",
    "f_prod = interp1d(x, y_prod,  kind='cubic')\n",
    "f_rh = interp1d(x, y_rh,  kind='cubic')\n",
    "f_or1 = interp1d(x, y_or1,  kind='cubic')\n",
    "f_or2 = interp1d(x, y_or2,  kind='cubic')\n",
    "axarr[0, 2].plot(x, f_prod(x), '-', x, f_sum(x), '-', x, f_rh(x), '-',\n",
    "                 x, f_svm(x), '-', x, f_max(x), '-', x, f_or1(x), '-', x, f_or2(x), '-')\n",
    "axarr[0, 2].set_title('branch vs flower')\n",
    "axarr[0, 2].set_ylim([0.3,1])\n",
    "axarr[0, 2].set_xlim([0,20])\n",
    "\n",
    "#------------------------------------------------------------------------------------\n",
    "file_scores = open('fusion_data/branch_entire_top_50_scores.txt', 'r')\n",
    "for line in file_scores:\n",
    "    if line.split(':')[0] == 'max rule':\n",
    "        y_max = ast.literal_eval(line.split(':')[1].strip())\n",
    "    elif line.split(':')[0] == 'sum rule':\n",
    "        y_sum = ast.literal_eval(line.split(':')[1].strip())\n",
    "    elif line.split(':')[0] == 'prod rule':\n",
    "        y_prod = ast.literal_eval(line.split(':')[1].strip())\n",
    "    elif line.split(':')[0] == 'branch':\n",
    "        y_or1 = ast.literal_eval(line.split(':')[1].strip())\n",
    "    elif line.split(':')[0] == 'entire':\n",
    "        y_or2 = ast.literal_eval(line.split(':')[1].strip())\n",
    "    elif line.split(':')[0] == 'rh rule':\n",
    "        y_rh = ast.literal_eval(line.split(':')[1].strip())\n",
    "    else:\n",
    "        y_svm = ast.literal_eval(line.split(':')[1].strip())\n",
    "file_scores.close()\n",
    "f_svm = interp1d(x, y_svm,  kind='cubic')\n",
    "f_max = interp1d(x, y_max,  kind='cubic')\n",
    "f_sum = interp1d(x, y_sum,  kind='cubic')\n",
    "f_prod = interp1d(x, y_prod,  kind='cubic')\n",
    "f_rh = interp1d(x, y_rh,  kind='cubic')\n",
    "f_or1 = interp1d(x, y_or1,  kind='cubic')\n",
    "f_or2 = interp1d(x, y_or2,  kind='cubic')\n",
    "\n",
    "axarr[1, 2].plot(x, f_prod(x), '-', x, f_sum(x), '-', x, f_rh(x), '-',\n",
    "                 x, f_svm(x), '-', x, f_max(x), '-', x, f_or1(x), '-', x, f_or2(x), '-')\n",
    "axarr[1, 2].set_title('branch vs entire')\n",
    "axarr[1, 2].set_ylim([0.2,1])\n",
    "axarr[1, 2].set_xlim([0,20])\n",
    "axarr[1, 2].set_xticks(np.arange(0, 21, 5))\n",
    "\n",
    "# Fine-tune figure; hide x ticks for top plots and y ticks for right plots\n",
    "plt.setp([a.get_xticklabels() for a in axarr[0, :]], visible=False)\n",
    "plt.setp([a.get_yticklabels() for a in axarr[:, 1]], visible=False)\n",
    "plt.setp([a.get_yticklabels() for a in axarr[:, 2]], visible=False)\n",
    "f.legend((l1, l2, l3, l4, l5, l6, l7), ('Product rule', 'Sum rule','RHF',\n",
    "                                        'SVM', 'Max rule','Organ 1','Organ 2'), 'lower center',\n",
    "         ncol=7, frameon=False)\n",
    "\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 2",
   "language": "python",
   "name": "python2"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 2
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython2",
   "version": "2.7.11"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 0
}