FROC2.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Using TensorFlow backend.\n"
     ]
    }
   ],
   "source": [
    "import cv2\n",
    "import numpy as np\n",
    "import os \n",
    "import dicom\n",
    "import data\n",
    "import copy\n",
    "from keras.models import Sequential\n",
    "from keras.layers import Dense, Dropout, Activation, Flatten\n",
    "from keras.layers import Convolution2D, MaxPooling2D\n",
    "from keras.utils import np_utils\n",
    "from scipy import ndimage\n",
    "from keras.optimizers import SGD\n",
    "from keras.utils import np_utils\n",
    "import matplotlib.pyplot as plt\n",
    "from skimage.morphology import remove_small_objects\n",
    "from scipy import ndimage\n",
    "from sklearn import metrics,metrics\n",
    "from scipy.ndimage.measurements import label\n",
    "%matplotlib inline"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "def getPara(predict, true, threshold):\n",
    "    (TP, FP, TN, FN, class_lable) = perf_measure(true, predict, threshold)\n",
    "    if((TP + FN)== 0):\n",
    "        TPR = 0\n",
    "    else:\n",
    "        TPR = np.float(TP)/(TP + FN)\n",
    "    \n",
    "    class_lable = class_lable.astype(bool).reshape(264,132)\n",
    "    true = predict.astype(bool).reshape(264,132)\n",
    "    \n",
    "    predict2 = remove_small_objects(class_lable, 64,in_place=False)\n",
    "    true2 = remove_small_objects(true, 64,in_place=False)\n",
    "    labeled_array1, num_features1 = label(predict2)\n",
    "    labeled_array2, num_features2 = label(true2)\n",
    "    FP_num = num_features1 - num_features2\n",
    "    return TPR, FP_num"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "def perf_measure(y_actual, predict, threshold):\n",
    "    TP = 0\n",
    "    FP = 0\n",
    "    TN = 0\n",
    "    FN = 0\n",
    "    predict = transfer_prob(predict, threshold)\n",
    "    for i in range(len(predict)): \n",
    "        if y_actual[i]==predict[i]==1:\n",
    "           TP += 1\n",
    "    for i in range(len(predict)): \n",
    "        if y_actual[i]==1 and y_actual[i]!=predict[i]:\n",
    "           FP += 1\n",
    "    for i in range(len(predict)): \n",
    "        if y_actual[i]==predict[i]==0:\n",
    "           TN += 1\n",
    "    for i in range(len(predict)): \n",
    "        if y_actual[i]==0 and y_actual[i]!=predict[i]:\n",
    "           FN += 1\n",
    "\n",
    "    return(TP, FP, TN, FN, predict)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "def transfer_prob(y_score, threshold):\n",
    "    y_result = []\n",
    "    for i in range(len(y_score)):\n",
    "        if y_score[i] >= threshold:\n",
    "            y_result.append(1)\n",
    "        else:\n",
    "            y_result.append(0)\n",
    "    return np.asarray(y_result)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYMAAAEZCAYAAAB1mUk3AAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nhQAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x11404fb50>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "\n",
    "TPR_list= np.load('./resolution/TPR_list_reso.npy') \n",
    "FP_num_list = np.load('./resolution/FP_num_list_reso.npy')\n",
    "\n",
    "#TPR_list1= np.load('TPR_list_pure_good.npy') \n",
    "#FP_num_list1 = np.load('FP_num_list_pure_good.npy')\n",
    "\n",
    "#TPR_list2= np.load('TPR_list_160.npy') \n",
    "#FP_num_list2 = np.load('FP_num_list_160.npy')\n",
    "\n",
    "# false_positive_rate, true_positive_rate, threshol = metrics.roc_curve(y_true.reshape(y_true.shape[0]*y_true.shape[1]), y_score.reshape(y_score.shape[0]*y_score.shape[1]))\n",
    "# thresholds = []\n",
    "# count = 1\n",
    "# for i in range(threshol.shape[0]):\n",
    "#     if(threshol[i] > 0.01 and count%100 == 0):\n",
    "#         thresholds.append(threshol[i])\n",
    "#     count += 1\n",
    "# thresholds = np.asarray(thresholds)\n",
    "plt.gca().set_color_cycle(['red', 'green', 'blue'])\n",
    "plt.title('Free response Receiver Operating Characteristic')\n",
    "plt.plot((FP_num_list), (TPR_list), marker = 'o')\n",
    "# plt.plot((FP_num_list1), (TPR_list1), marker = 'o')\n",
    "# plt.plot((FP_num_list2), (TPR_list2), marker = 'o')\n",
    "\n",
    "plt.ylim([ 0.3,0.85])\n",
    "plt.xlim([-0.1,6])\n",
    "plt.legend(loc='lower right')\n",
    "plt.ylabel('True Positive Rate')\n",
    "plt.xlabel('False Negative Numbers')\n",
    "plt.legend(['Filter size: 160', 'Filter size: 64', 'Filter size: 160'], loc='lower right')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "plt.imshow(y_score[18].reshape(264,132),'gray')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "getPara(y_score[18], y_true[18], 0.43136)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "getPara(y_score[18], y_true[18], 0.40587)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "    (TP, FP, TN, FN, class_lable) = perf_measure(y_true[18], y_score[18], 0.43136)\n",
    "    class_lable = class_lable.astype(bool).reshape(264,132)\n",
    "    true =  y_true[18].astype(bool).reshape(264,132)\n",
    "    \n",
    "    predict2 = remove_small_objects(class_lable, 64,in_place=False)\n",
    "    true2 = remove_small_objects(true, 64,in_place=False)\n",
    "    labeled_array1, num_features1 = label(predict2)\n",
    "    labeled_array2, num_features2 = label(true2)\n",
    "    FP_num = num_features1 - num_features2"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "num_features1"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "plt.imshow(predict2,'gray')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "    (TP, FP, TN, FN, class_lable) = perf_measure(y_true[18], y_score[18], 0.40587)\n",
    "    class_lable = class_lable.astype(bool).reshape(264,132)\n",
    "    true =  y_true[18].astype(bool).reshape(264,132)\n",
    "    \n",
    "    predict2 = remove_small_objects(class_lable, 64,in_place=False)\n",
    "    true2 = remove_small_objects(true, 64,in_place=False)\n",
    "    labeled_array1, num_features1 = label(predict2)\n",
    "    labeled_array2, num_features2 = label(true2)\n",
    "    FP_num = num_features1 - num_features2"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "num_features1"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 106,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "y_score = np.load('./resolution/predicted_prob.npy')\n",
    "y_true = np.load('./resolution/answer_image.npy')\n",
    "reso = np.load('./resolution/resolution.npy')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 107,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "y_score = y_score[0:y_score.shape[0],1]\n",
    "y_true = y_true[0:y_true.shape[0],1]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 108,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "scores = []\n",
    "trues = []\n",
    "next_start = 0\n",
    "for i in range(len(reso)):\n",
    "    ysize = (reso[i][0] - 56)\n",
    "    xsize = (reso[i][0]*0.5 - 28)\n",
    "    scores.append(y_score[next_start: next_start + np.int(ysize * xsize )])\n",
    "    trues.append(y_true[next_start: next_start + np.int(ysize * xsize )])\n",
    "    next_start = np.int( next_start + ysize * xsize )\n",
    "    "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 111,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(456.0, 228.0)\n",
      "(456.0, 228.0)\n",
      "(456.0, 228.0)\n",
      "(456.0, 228.0)\n",
      "(456.0, 228.0)\n",
      "(456.0, 228.0)\n",
      "(328.0, 164.0)\n",
      "(328.0, 164.0)\n",
      "(328.0, 164.0)\n",
      "(456.0, 228.0)\n",
      "(456.0, 228.0)\n",
      "(456.0, 228.0)\n",
      "(456.0, 228.0)\n",
      "(456.0, 228.0)\n",
      "(584.0, 292.0)\n",
      "(584.0, 292.0)\n",
      "(584.0, 292.0)\n",
      "(584.0, 292.0)\n",
      "(584.0, 292.0)\n",
      "(328.0, 164.0)\n",
      "(328.0, 164.0)\n",
      "(328.0, 164.0)\n"
     ]
    }
   ],
   "source": [
    "for i in range(len(reso)):\n",
    "    ysize = (reso[i][0] - 56)\n",
    "    xsize = (reso[i][0]*0.5 - 28)\n",
    "    print(ysize,xsize)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 112,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "import cv2\n",
    "import numpy as np\n",
    "import os\n",
    "import dicom\n",
    "import data\n",
    "import copy\n",
    "from keras.models import Sequential\n",
    "from keras.layers import Dense, Dropout, Activation, Flatten\n",
    "from keras.layers import Convolution2D, MaxPooling2D\n",
    "from keras.utils import np_utils\n",
    "from scipy import ndimage\n",
    "from keras.optimizers import SGD\n",
    "from keras.utils import np_utils\n",
    "from skimage.morphology import remove_small_objects\n",
    "from scipy import ndimage\n",
    "from sklearn import metrics, metrics\n",
    "from scipy.ndimage.measurements import label\n",
    "\n",
    "\n",
    "def getPara(predict, true, threshold, resolution):\n",
    "    (TP, FP, TN, FN, class_lable) = perf_measure(true, predict, threshold)\n",
    "    if((TP + FN) == 0):\n",
    "        TPR = 0\n",
    "    else:\n",
    "        TPR = np.float(TP) / (TP + FN)\n",
    "\n",
    "    class_lable = class_lable.astype(\n",
    "        bool).reshape(resolution[0], resolution[1])\n",
    "    true = true.astype(bool).reshape(resolution[0], resolution[1])\n",
    "\n",
    "    predict2 = remove_small_objects(class_lable, 160, in_place=False)\n",
    "    labeled_array1, num_features1 = label(predict2)\n",
    "    labeled_array2, num_features2 = label(true)\n",
    "    FP_num = num_features1 - num_features2\n",
    "    if FP_num < 0:\n",
    "        FP_num = 0\n",
    "    return TPR, FP_num\n",
    "\n",
    "\n",
    "def perf_measure(y_actual, predict, threshold):\n",
    "    TP = 0\n",
    "    FP = 0\n",
    "    TN = 0\n",
    "    FN = 0\n",
    "    predict = transfer_prob(predict, threshold)\n",
    "    for i in range(len(predict)):\n",
    "        if y_actual[i] == predict[i] == 1:\n",
    "            TP += 1\n",
    "    for i in range(len(predict)):\n",
    "        if y_actual[i] == 0 and y_actual[i] != predict[i]:\n",
    "            FP += 1\n",
    "    for i in range(len(predict)):\n",
    "        if y_actual[i] == predict[i] == 0:\n",
    "            TN += 1\n",
    "    for i in range(len(predict)):\n",
    "        if y_actual[i] == 1 and y_actual[i] != predict[i]:\n",
    "            FN += 1\n",
    "\n",
    "    return(TP, FP, TN, FN, predict)\n",
    "\n",
    "\n",
    "def transfer_prob(y_score, threshold):\n",
    "    y_result = []\n",
    "    for i in range(len(y_score)):\n",
    "        if y_score[i] >= threshold:\n",
    "            y_result.append(1)\n",
    "        else:\n",
    "            y_result.append(0)\n",
    "    return np.asarray(y_result)\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 194,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "y_score = np.load('./resolution/predicted_prob.npy')\n",
    "y_true = np.load('./resolution/answer_image.npy')\n",
    "reso = np.load('./resolution/resolution.npy')\n",
    "\n",
    "y_score = y_score[0:y_score.shape[0], 1]\n",
    "y_true = y_true[0:y_true.shape[0], 1]\n",
    "\n",
    "scores = []\n",
    "trues = []\n",
    "next_start = 0\n",
    "for i in range(len(reso)):\n",
    "    ysize = (reso[i][0] - 56)\n",
    "    xsize = (reso[i][0] * 0.5 - 28)\n",
    "    reso[i][0] = ysize\n",
    "    reso[i][1] = xsize\n",
    "    scores.append(y_score[next_start: next_start + np.int(ysize * xsize)])\n",
    "    trues.append(y_true[next_start: next_start + np.int(ysize * xsize)])\n",
    "    next_start = np.int(next_start + ysize * xsize)\n",
    "    #print(next_start,ysize,xsize)\n",
    "\n",
    "                                                                                                                                                                                             \n",
    "y_score = np.asarray(scores)\n",
    "y_true = np.asarray(trues)\n",
    "#y_true = y_true.astype(np.int)\n",
    "\n",
    "set_size = 22\n",
    "\n",
    "count = 1\n",
    "\n",
    "\n",
    "thresholds = []\n",
    "tmp = 0\n",
    "for m in range(1, 10, 1):\n",
    "    tmp += 1\n",
    "    # print(m/np.float(100))\n",
    "    thresholds.append(m / np.float(100))\n",
    "for i in range(1, 10, 1):\n",
    "    thresholds.append(i / np.float(10))\n",
    "for m in range(90, 100, 1):\n",
    "    thresholds.append(m / np.float(100))\n",
    "tmp = 0\n",
    "for m in range(900, 1000, 1):\n",
    "    tmp += 1\n",
    "    if(tmp % 10 == 0):\n",
    "        thresholds.append(m / np.float(1000))\n",
    "thresholds = sorted(thresholds, reverse=True)\n",
    "thresholds = np.asarray(thresholds)\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 198,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/Users/fangzhiyuan/anaconda/lib/python2.7/site-packages/ipykernel/__main__.py:28: DeprecationWarning: using a non-integer number instead of an integer will result in an error in the future\n",
      "/Users/fangzhiyuan/anaconda/lib/python2.7/site-packages/ipykernel/__main__.py:29: DeprecationWarning: using a non-integer number instead of an integer will result in an error in the future\n"
     ]
    },
    {
     "ename": "KeyboardInterrupt",
     "evalue": "",
     "output_type": "error",
     "traceback": [
      "\u001b[0;31m---------------------------------------------------------------------------\u001b[0m",
      "\u001b[0;31mKeyboardInterrupt\u001b[0m                         Traceback (most recent call last)",
      "\u001b[0;32m<ipython-input-198-710c5c3fdcb3>\u001b[0m in \u001b[0;36m<module>\u001b[0;34m()\u001b[0m\n\u001b[1;32m      9\u001b[0m     \u001b[0;32mfor\u001b[0m \u001b[0mi\u001b[0m \u001b[0;32min\u001b[0m \u001b[0mrange\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mset_size\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m     10\u001b[0m         \u001b[0;32mif\u001b[0m \u001b[0mi\u001b[0m \u001b[0;32mnot\u001b[0m \u001b[0;32min\u001b[0m \u001b[0mdelete\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m---> 11\u001b[0;31m             \u001b[0mTPR\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mFP_num\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mgetPara\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0my_score\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0mi\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0my_true\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0mi\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mastype\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mnp\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mint\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m,\u001b[0m  \u001b[0mthresholds\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0mt\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mreso\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0mi\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m",
      "\u001b[0;32m<ipython-input-112-a2e3f65d010b>\u001b[0m in \u001b[0;36mgetPara\u001b[0;34m(predict, true, threshold, resolution)\u001b[0m\n\u001b[1;32m     19\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m     20\u001b[0m \u001b[0;32mdef\u001b[0m \u001b[0mgetPara\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mpredict\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mtrue\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mthreshold\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mresolution\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m---> 21\u001b[0;31m     \u001b[0;34m(\u001b[0m\u001b[0mTP\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mFP\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mTN\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mFN\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mclass_lable\u001b[0m\u001b[0;34m)\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mperf_measure\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mtrue\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mpredict\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mthreshold\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m     22\u001b[0m     \u001b[0;32mif\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mTP\u001b[0m \u001b[0;34m+\u001b[0m \u001b[0mFN\u001b[0m\u001b[0;34m)\u001b[0m \u001b[0;34m==\u001b[0m \u001b[0;36m0\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m     23\u001b[0m         \u001b[0mTPR\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0;36m0\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m<ipython-input-112-a2e3f65d010b>\u001b[0m in \u001b[0;36mperf_measure\u001b[0;34m(y_actual, predict, threshold)\u001b[0m\n\u001b[1;32m     45\u001b[0m     \u001b[0mpredict\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mtransfer_prob\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mpredict\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mthreshold\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m     46\u001b[0m     \u001b[0;32mfor\u001b[0m \u001b[0mi\u001b[0m \u001b[0;32min\u001b[0m \u001b[0mrange\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mlen\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mpredict\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m---> 47\u001b[0;31m         \u001b[0;32mif\u001b[0m \u001b[0my_actual\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0mi\u001b[0m\u001b[0;34m]\u001b[0m \u001b[0;34m==\u001b[0m \u001b[0mpredict\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0mi\u001b[0m\u001b[0;34m]\u001b[0m \u001b[0;34m==\u001b[0m \u001b[0;36m1\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m     48\u001b[0m             \u001b[0mTP\u001b[0m \u001b[0;34m+=\u001b[0m \u001b[0;36m1\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m     49\u001b[0m     \u001b[0;32mfor\u001b[0m \u001b[0mi\u001b[0m \u001b[0;32min\u001b[0m \u001b[0mrange\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mlen\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mpredict\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;31mKeyboardInterrupt\u001b[0m: "
     ]
    }
   ],
   "source": [
    "TPR_list = []\n",
    "FP_num_list = []\n",
    "delete = [6, 7, 12, 19, 21]\n",
    "\n",
    "#delete = [6,7,8,9,10,11,12,13,19,20,21]\n",
    "for t in range(1, thresholds.size):\n",
    "    tpr_sum = 0\n",
    "    fp_sum = 0\n",
    "    for i in range(set_size):\n",
    "        if i not in delete:\n",
    "            TPR, FP_num = getPara(y_score[i], y_true[i].astype(np.int),  thresholds[t], reso[i])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 199,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/Users/fangzhiyuan/anaconda/lib/python2.7/site-packages/ipykernel/__main__.py:28: DeprecationWarning: using a non-integer number instead of an integer will result in an error in the future\n",
      "/Users/fangzhiyuan/anaconda/lib/python2.7/site-packages/ipykernel/__main__.py:29: DeprecationWarning: using a non-integer number instead of an integer will result in an error in the future\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "(0.5129474485910129, 6)"
      ]
     },
     "execution_count": 199,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "getPara(y_score[i], y_true[i].astype(np.int),  thresholds[t], reso[i])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 170,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "predict = y_score[i]\n",
    "true = y_true[i]\n",
    "threshold = 0.5"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 174,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "    TP = 0\n",
    "    FP = 0\n",
    "    TN = 0\n",
    "    FN = 0\n",
    "    predict = transfer_prob(predict, threshold)\n",
    "    y_actual = true"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 175,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "ename": "IndexError",
     "evalue": "invalid index to scalar variable.",
     "output_type": "error",
     "traceback": [
      "\u001b[0;31m---------------------------------------------------------------------------\u001b[0m",
      "\u001b[0;31mIndexError\u001b[0m                                Traceback (most recent call last)",
      "\u001b[0;32m<ipython-input-175-b11be20127c3>\u001b[0m in \u001b[0;36m<module>\u001b[0;34m()\u001b[0m\n\u001b[1;32m      1\u001b[0m \u001b[0;32mfor\u001b[0m \u001b[0mi\u001b[0m \u001b[0;32min\u001b[0m \u001b[0mrange\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mlen\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mpredict\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m----> 2\u001b[0;31m     \u001b[0;32mif\u001b[0m \u001b[0my_actual\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0mi\u001b[0m\u001b[0;34m]\u001b[0m \u001b[0;34m==\u001b[0m \u001b[0mpredict\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0mi\u001b[0m\u001b[0;34m]\u001b[0m \u001b[0;34m==\u001b[0m \u001b[0;36m1\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m      3\u001b[0m         \u001b[0mTP\u001b[0m \u001b[0;34m+=\u001b[0m \u001b[0;36m1\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m      4\u001b[0m \u001b[0;32mfor\u001b[0m \u001b[0mi\u001b[0m \u001b[0;32min\u001b[0m \u001b[0mrange\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mlen\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mpredict\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m      5\u001b[0m     \u001b[0;32mif\u001b[0m \u001b[0my_actual\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0mi\u001b[0m\u001b[0;34m]\u001b[0m \u001b[0;34m==\u001b[0m \u001b[0;36m0\u001b[0m \u001b[0;32mand\u001b[0m \u001b[0my_actual\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0mi\u001b[0m\u001b[0;34m]\u001b[0m \u001b[0;34m!=\u001b[0m \u001b[0mpredict\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0mi\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;31mIndexError\u001b[0m: invalid index to scalar variable."
     ]
    }
   ],
   "source": [
    "    for i in range(len(predict)):\n",
    "        if y_actual[i] == predict[i] == 1:\n",
    "            TP += 1\n",
    "    for i in range(len(predict)):\n",
    "        if y_actual[i] == 0 and y_actual[i] != predict[i]:\n",
    "            FP += 1\n",
    "    for i in range(len(predict)):\n",
    "        if y_actual[i] == predict[i] == 0:\n",
    "            TN += 1\n",
    "    for i in range(len(predict)):\n",
    "        if y_actual[i] == 1 and y_actual[i] != predict[i]:\n",
    "            FN += 1"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 176,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "ename": "IndexError",
     "evalue": "invalid index to scalar variable.",
     "output_type": "error",
     "traceback": [
      "\u001b[0;31m---------------------------------------------------------------------------\u001b[0m",
      "\u001b[0;31mIndexError\u001b[0m                                Traceback (most recent call last)",
      "\u001b[0;32m<ipython-input-176-3e14171ce079>\u001b[0m in \u001b[0;36m<module>\u001b[0;34m()\u001b[0m\n\u001b[1;32m      1\u001b[0m \u001b[0;32mfor\u001b[0m \u001b[0mi\u001b[0m \u001b[0;32min\u001b[0m \u001b[0mrange\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mlen\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mpredict\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m----> 2\u001b[0;31m     \u001b[0;32mif\u001b[0m \u001b[0my_actual\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0mi\u001b[0m\u001b[0;34m]\u001b[0m \u001b[0;34m==\u001b[0m \u001b[0mpredict\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0mi\u001b[0m\u001b[0;34m]\u001b[0m \u001b[0;34m==\u001b[0m \u001b[0;36m1\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m      3\u001b[0m         \u001b[0mTP\u001b[0m \u001b[0;34m+=\u001b[0m \u001b[0;36m1\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;31mIndexError\u001b[0m: invalid index to scalar variable."
     ]
    }
   ],
   "source": [
    "    for i in range(len(predict)):\n",
    "        if y_actual[i] == predict[i] == 1:\n",
    "            TP += 1"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 196,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([0, 0, 0, ..., 0, 0, 0])"
      ]
     },
     "execution_count": 196,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "y_true[i].astype(np.int)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 200,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[  4.56000000e+02,   2.28000000e+02,   5.85900000e-01,\n",
       "          5.85900000e-01],\n",
       "       [  4.56000000e+02,   2.28000000e+02,   5.85900000e-01,\n",
       "          5.85900000e-01],\n",
       "       [  4.56000000e+02,   2.28000000e+02,   5.85900000e-01,\n",
       "          5.85900000e-01],\n",
       "       [  4.56000000e+02,   2.28000000e+02,   5.85900000e-01,\n",
       "          5.85900000e-01],\n",
       "       [  4.56000000e+02,   2.28000000e+02,   5.85900000e-01,\n",
       "          5.85900000e-01],\n",
       "       [  4.56000000e+02,   2.28000000e+02,   5.85900000e-01,\n",
       "          5.85900000e-01],\n",
       "       [  3.28000000e+02,   1.64000000e+02,   7.29166687e-01,\n",
       "          7.29166687e-01],\n",
       "       [  3.28000000e+02,   1.64000000e+02,   7.29166687e-01,\n",
       "          7.29166687e-01],\n",
       "       [  3.28000000e+02,   1.64000000e+02,   7.29166687e-01,\n",
       "          7.29166687e-01],\n",
       "       [  4.56000000e+02,   2.28000000e+02,   5.46900000e-01,\n",
       "          5.46900000e-01],\n",
       "       [  4.56000000e+02,   2.28000000e+02,   5.46900000e-01,\n",
       "          5.46900000e-01],\n",
       "       [  4.56000000e+02,   2.28000000e+02,   5.46900000e-01,\n",
       "          5.46900000e-01],\n",
       "       [  4.56000000e+02,   2.28000000e+02,   5.46900000e-01,\n",
       "          5.46900000e-01],\n",
       "       [  4.56000000e+02,   2.28000000e+02,   5.46900000e-01,\n",
       "          5.46900000e-01],\n",
       "       [  5.84000000e+02,   2.92000000e+02,   5.62500000e-01,\n",
       "          5.62500000e-01],\n",
       "       [  5.84000000e+02,   2.92000000e+02,   5.62500000e-01,\n",
       "          5.62500000e-01],\n",
       "       [  5.84000000e+02,   2.92000000e+02,   5.62500000e-01,\n",
       "          5.62500000e-01],\n",
       "       [  5.84000000e+02,   2.92000000e+02,   5.62500000e-01,\n",
       "          5.62500000e-01],\n",
       "       [  5.84000000e+02,   2.92000000e+02,   5.62500000e-01,\n",
       "          5.62500000e-01],\n",
       "       [  3.28000000e+02,   1.64000000e+02,   7.29166687e-01,\n",
       "          7.29166687e-01],\n",
       "       [  3.28000000e+02,   1.64000000e+02,   7.29166687e-01,\n",
       "          7.29166687e-01],\n",
       "       [  3.28000000e+02,   1.64000000e+02,   7.29166687e-01,\n",
       "          7.29166687e-01]])"
      ]
     },
     "execution_count": 200,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "reso"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 201,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "ename": "IndexError",
     "evalue": "tuple index out of range",
     "output_type": "error",
     "traceback": [
      "\u001b[0;31m---------------------------------------------------------------------------\u001b[0m",
      "\u001b[0;31mIndexError\u001b[0m                                Traceback (most recent call last)",
      "\u001b[0;32m<ipython-input-201-e8803c15ce9d>\u001b[0m in \u001b[0;36m<module>\u001b[0;34m()\u001b[0m\n\u001b[1;32m      7\u001b[0m \u001b[0;31m#FP_num_list2 = np.load('FP_num_list1.npy')\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m      8\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m----> 9\u001b[0;31m \u001b[0mfalse_positive_rate\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mtrue_positive_rate\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mthreshol\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mmetrics\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mroc_curve\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0my_true\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mreshape\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0my_true\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mshape\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0;36m0\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m*\u001b[0m\u001b[0my_true\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mshape\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0;36m1\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0my_score\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mreshape\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0my_score\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mshape\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0;36m0\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m*\u001b[0m\u001b[0my_score\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mshape\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0;36m1\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m     10\u001b[0m \u001b[0mthresholds\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0;34m[\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m     11\u001b[0m \u001b[0mcount\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0;36m1\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;31mIndexError\u001b[0m: tuple index out of range"
     ]
    }
   ],
   "source": [
    "\n",
    "\n",
    "TPR_list1= np.load('TPR_list_160.npy') \n",
    "FP_num_list1 = np.load('FP_num_list_160.npy')\n",
    "\n",
    "#TPR_list2= np.load('TPR_list1.npy') \n",
    "#FP_num_list2 = np.load('FP_num_list1.npy')\n",
    "\n",
    "false_positive_rate, true_positive_rate, threshol = metrics.roc_curve(y_true.reshape(y_true.shape[0]*y_true.shape[1]), y_score.reshape(y_score.shape[0]*y_score.shape[1]))\n",
    "thresholds = []\n",
    "count = 1\n",
    "for i in range(threshol.shape[0]):\n",
    "    if(threshol[i] > 0.01 and count%100 == 0):\n",
    "        thresholds.append(threshol[i])\n",
    "    count += 1\n",
    "thresholds = np.asarray(thresholds)\n",
    "   "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "\n",
    "TPR_list= np.load('./resolution/TPR_list_reso.npy') \n",
    "FP_num_list = np.load('./resolution/FP_num_list_reso.npy')\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([ 0.40961644,  0.41331948,  0.43437496,  0.43639786,  0.44750569,\n",
       "        0.44814098,  0.45687315,  0.45767279,  0.46589002,  0.46626547,\n",
       "        0.47305448,  0.47305448,  0.47910395,  0.47944498,  0.48343548,\n",
       "        0.48405351,  0.48764436,  0.48844036,  0.49338583,  0.49338583,\n",
       "        0.5295392 ,  0.56103594,  0.59098634,  0.62394333,  0.65527835,\n",
       "        0.6847334 ,  0.71333   ,  0.74796118,  0.75104789,  0.75621759,\n",
       "        0.76063337,  0.76641144,  0.77209585,  0.77744104,  0.78603683,\n",
       "        0.79783071,  0.81357874])"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "TPR_list"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([ 0.    ,  0.    ,  0.    ,  0.    ,  0.0625,  0.0625,  0.0625,\n",
       "        0.0625,  0.125 ,  0.125 ,  0.25  ,  0.25  ,  0.25  ,  0.25  ,\n",
       "        0.25  ,  0.25  ,  0.25  ,  0.25  ,  0.25  ,  0.25  ,  0.375 ,\n",
       "        0.5   ,  0.75  ,  1.125 ,  1.375 ,  1.5   ,  1.8125,  2.5625,\n",
       "        2.6875,  2.875 ,  3.1875,  3.375 ,  3.5625,  3.6875,  4.3125,\n",
       "        4.9375,  5.375 ])"
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "FP_num_list"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "TPR_list= np.load('./resolution/TPR_list_reso.npy') \n",
    "FP_num_list = np.load('./resolution/FP_num_list_reso.npy')\n",
    "\n",
    "TPR_list1= np.load('./resolution/TPR_list_64.npy') \n",
    "FP_num_list1 = np.load('./resolution/FP_num_list_64.npy')\n",
    "\n",
    "TPR_list2= np.load('./resolution/TPR_list_128.npy') \n",
    "FP_num_list2 = np.load('./resolution/FP_num_list_128.npy')\n",
    "\n",
    "plt.gca().set_color_cycle(['red', 'green', 'blue'])\n",
    "plt.title('Free response Receiver Operating Characteristic')\n",
    "plt.plot((FP_num_list), (TPR_list), marker = 'o')\n",
    "plt.plot((FP_num_list1), (TPR_list1), marker = 'o')\n",
    "plt.plot((FP_num_list2), (TPR_list2), marker = 'o')\n",
    "plt.ylim([ 0.3,0.85])\n",
    "plt.xlim([-0.1,4])\n",
    "plt.legend(loc='lower right')\n",
    "plt.ylabel('True Positive Rate')\n",
    "plt.xlabel('False Negative Numbers')\n",
    "plt.legend(['Filter size: 160', 'Filter size: 64', 'Filter size: 128'], loc='lower right')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYgAAAEZCAYAAACNebLAAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x114075550>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "\n",
    "TPR_list= np.load('./resolution/TPR_list_600.npy') \n",
    "FP_num_list = np.load('./resolution/FP_num_list_600.npy')\n",
    "TPR_list1= np.load('./resolution/TPR_list_500.npy')\n",
    "FP_num_list1 = np.load('./resolution/FP_num_list_500.npy')\n",
    "\n",
    "TPR_list2= np.load('./resolution/TPR_list_490.npy') \n",
    "FP_num_list2 = np.load('./resolution/FP_num_list_490.npy')\n",
    "\n",
    "TPR_list3= np.load('./resolution/TPR_list_400.npy') \n",
    "FP_num_list3 = np.load('./resolution/FP_num_list_400.npy')\n",
    "\n",
    "TPR_list4= np.load('./resolution/TPR_list_noreso.npy') \n",
    "FP_num_list4 = np.load('./resolution/FP_num_list_noreso.npy')\n",
    "\n",
    "TPR_list5= np.load('./resolution/TPR_list_special.npy') \n",
    "FP_num_list5 = np.load('./resolution/FP_num_list_spec.npy')\n",
    "\n",
    "plt.gca().set_color_cycle(['red', 'green', 'blue','orange','purple','yellow'])\n",
    "plt.title('Free response Receiver Operating Characteristic')\n",
    "plt.plot((FP_num_list), (TPR_list), marker = 'o')\n",
    "plt.plot((FP_num_list1), (TPR_list1), marker = 'o') \n",
    "plt.plot((FP_num_list2), (TPR_list2), marker = 'o') \n",
    "plt.plot((FP_num_list3), (TPR_list3), marker = 'o') \n",
    "plt.plot((FP_num_list4), (TPR_list4), marker = 'o') \n",
    "\n",
    "plt.ylim([ 0.3,0.9]) \n",
    "plt.xlim([-0.1,3])\n",
    "plt.grid()\n",
    "plt.legend(loc='lower right') \n",
    "plt.ylabel('True Positive Rate')\n",
    "plt.xlabel('False Negative Numbers')\n",
    "plt.legend([ 'Filter size: 600', 'Filter size: 500', 'Filter size: un reso_500','Filter size: 400','unstable resolution'], loc='lower right')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYgAAAEZCAYAAACNebLAAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nAABJRU5ErkJggg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x1143c1fd0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "\n",
    "TPR_list4= np.load('./resolution/TPR_list_special.npy') \n",
    "FP_num_list4 = np.load('./resolution/FP_num_list_special.npy')\n",
    "\n",
    "plt.gca().set_color_cycle(['red', 'green', 'blue','orange','purple'])\n",
    "plt.title('Free response Receiver Operating Characteristic')\n",
    "\n",
    "plt.plot((FP_num_list4), (TPR_list4), marker = 'o') \n",
    "\n",
    "plt.ylim([ 0.3,0.95]) \n",
    "plt.xlim([-0.1,3])\n",
    "plt.grid()\n",
    "plt.legend(loc='lower right') \n",
    "plt.ylabel('True Positive Rate')\n",
    "plt.xlabel('False Negative Numbers')\n",
    "plt.legend([ 'Filter size: 600', 'Filter size: 500', 'Filter size: un reso_500','Filter size: 400','unstable resolution'], loc='lower right')\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
}