Lung_Cancer_Analysis.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "import numpy as np # linear algebra\n",
    "import pandas as pd # data processing, CSV file I/O (e.g. pd.read_csv)\n",
    "import matplotlib.pyplot as plt\n",
    "import seaborn as sns\n",
    "import os\n",
    "import glob\n",
    "%matplotlib inline\n",
    "\n",
    "p = sns.color_palette()\n",
    "\n",
    "os.listdir('./')\n",
    "\n",
    "seed = 42"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Patient '00cba091fa4ad62cc3200a657aeb957e' has 134 scans\n",
      "Patient '0a099f2549429d29b32f349e95fb2244' has 128 scans\n",
      "Patient '0a0c32c9e08cc2ea76a71649de56be6d' has 133 scans\n",
      "Patient '0a38e7597ca26f9374f8ea2770ba870d' has 110 scans\n",
      "Patient '0acbebb8d463b4b9ca88cf38431aac69' has 203 scans\n",
      "Patient '0b20184e0cd497028bdd155d9fb42dc9' has 196 scans\n",
      "Patient '0bd0e3056cbf23a1cb7f0f0b18446068' has 280 scans\n",
      "Patient '0c0de3749d4fe175b7a5098b060982a1' has 123 scans\n",
      "Patient '0c37613214faddf8701ca41e6d43f56e' has 164 scans\n",
      "Patient '0c59313f52304e25d5a7dcf9877633b1' has 244 scans\n",
      "Patient '0c60f4b87afcb3e2dfa65abbbf3ef2f9' has 136 scans\n",
      "Patient '0c98fcb55e3f36d0c2b6507f62f4c5f1' has 180 scans\n",
      "Patient '0c9d8314f9c69840e25febabb1229fa4' has 221 scans\n",
      "Patient '0ca943d821204ceb089510f836a367fd' has 147 scans\n",
      "Patient '0d06d764d3c07572074d468b4cff954f' has 435 scans\n",
      "Patient '0d19f1c627df49eb223771c28548350e' has 183 scans\n",
      "Patient '0d2fcf787026fece4e57be167d079383' has 126 scans\n",
      "Patient '0d941a3ad6c889ac451caf89c46cb92a' has 177 scans\n",
      "Patient '0ddeb08e9c97227853422bd71a2a695e' has 171 scans\n",
      "Patient '0de72529c30fe642bc60dcb75c87f6bd' has 113 scans\n",
      "----\n",
      "Total patients 21 Total DCM files 3604\n"
     ]
    }
   ],
   "source": [
    "for d in os.listdir('./sample_images'):\n",
    "    if '.DS_Store' not in d:\n",
    "        print(\"Patient '{}' has {} scans\".format(d, len(os.listdir('./sample_images/' + d))))\n",
    "print('----')\n",
    "print('Total patients {} Total DCM files {}'.format(len(os.listdir('./sample_images')), \n",
    "                                                      len(glob.glob('./sample_images/*/*.dcm'))))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.text.Text at 0x116cee410>"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYEAAAEZCAYAAABxbJkKAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nJAAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x116bf7ad0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "patient_sizes = [len(os.listdir('./sample_images/' + d)) for d in os.listdir('./sample_images') if '.DS_Store' not in d]\n",
    "plt.hist(patient_sizes, color=p[2])\n",
    "plt.ylabel('Number of patients')\n",
    "plt.xlabel('DICOM files')\n",
    "plt.title('Histogram of DICOM count per patient')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Training Set"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>id</th>\n",
       "      <th>cancer</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0015ceb851d7251b8f399e39779d1e7d</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>0030a160d58723ff36d73f41b170ec21</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>003f41c78e6acfa92430a057ac0b306e</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>006b96310a37b36cccb2ab48d10b49a3</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>008464bb8521d09a42985dd8add3d0d2</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                 id  cancer\n",
       "0  0015ceb851d7251b8f399e39779d1e7d       1\n",
       "1  0030a160d58723ff36d73f41b170ec21       0\n",
       "2  003f41c78e6acfa92430a057ac0b306e       0\n",
       "3  006b96310a37b36cccb2ab48d10b49a3       1\n",
       "4  008464bb8521d09a42985dd8add3d0d2       1"
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_train = pd.read_csv('./stage1_labels.csv')\n",
    "df_train.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Number of training patients: 1397\n",
      "Cancer rate: 25.91%\n"
     ]
    }
   ],
   "source": [
    "print('Number of training patients: {}'.format(len(df_train)))\n",
    "print('Cancer rate: {:.4}%'.format(df_train.cancer.mean()*100))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Naive Submission\n",
    "Since the evaluation metric used in this competition is LogLoss and not something like AUC, this means that we can often gain an improvement just by aligning the probabilities of our sample submission to that of the training set.\n",
    "Before I try making a naive submission, I will calculate what the score of this submission would be on the training set to get a comparison."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Training logloss is 0.572141489479\n"
     ]
    }
   ],
   "source": [
    "from sklearn.metrics import log_loss\n",
    "logloss = log_loss(df_train.cancer, np.zeros_like(df_train.cancer) + df_train.cancer.mean())\n",
    "print('Training logloss is {}'.format(logloss))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "# sample = pd.read_csv('./stage1_sample_submission.csv')\n",
    "# sample['cancer'] = df_train.cancer.mean()\n",
    "# sample.to_csv('naive_submission.csv', index=False)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Looking at images"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "import dicom"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Filename: ./0a67f9edb4915467ac16a565955898d3.dcm\n"
     ]
    }
   ],
   "source": [
    "dcm = './0a67f9edb4915467ac16a565955898d3.dcm'\n",
    "print('Filename: {}'.format(dcm))\n",
    "dcm = dicom.read_file(dcm)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(0008, 0005) Specific Character Set              CS: 'ISO_IR 100'\n",
      "(0008, 0016) SOP Class UID                       UI: CT Image Storage\n",
      "(0008, 0018) SOP Instance UID                    UI: 1.2.840.113654.2.55.240087524148038410985780799448670801102\n",
      "(0008, 0060) Modality                            CS: 'CT'\n",
      "(0008, 103e) Series Description                  LO: 'Axial'\n",
      "(0010, 0010) Patient's Name                      PN: '0a0c32c9e08cc2ea76a71649de56be6d'\n",
      "(0010, 0020) Patient ID                          LO: '0a0c32c9e08cc2ea76a71649de56be6d'\n",
      "(0010, 0030) Patient's Birth Date                DA: '19000101'\n",
      "(0018, 0060) KVP                                 DS: ''\n",
      "(0020, 000d) Study Instance UID                  UI: 2.25.60037070027156423276159501017920151735078954137544798194660\n",
      "(0020, 000e) Series Instance UID                 UI: 2.25.58703274222857573910779974742342423982066946347485459782406\n",
      "(0020, 0011) Series Number                       IS: '1'\n",
      "(0020, 0012) Acquisition Number                  IS: '1'\n",
      "(0020, 0013) Instance Number                     IS: '45'\n",
      "(0020, 0020) Patient Orientation                 CS: ''\n",
      "(0020, 0032) Image Position (Patient)            DS: ['-160.100006', '-142.500000', '-93.410004']\n",
      "(0020, 0037) Image Orientation (Patient)         DS: ['1', '0', '0', '0', '1', '0']\n",
      "(0020, 0052) Frame of Reference UID              UI: 2.25.10816424791949633522058525471061381137400269917518537485573\n",
      "(0020, 1040) Position Reference Indicator        LO: 'SN'\n",
      "(0020, 1041) Slice Location                      DS: '-93.410004'\n",
      "(0028, 0002) Samples per Pixel                   US: 1\n",
      "(0028, 0004) Photometric Interpretation          CS: 'MONOCHROME2'\n",
      "(0028, 0006) Planar Configuration                US: 0\n",
      "(0028, 0010) Rows                                US: 512\n",
      "(0028, 0011) Columns                             US: 512\n",
      "(0028, 0030) Pixel Spacing                       DS: ['0.664062', '0.664062']\n",
      "(0028, 0100) Bits Allocated                      US: 16\n",
      "(0028, 0101) Bits Stored                         US: 12\n",
      "(0028, 0102) High Bit                            US: 11\n",
      "(0028, 0103) Pixel Representation                US: 0\n",
      "(0028, 0120) Pixel Padding Value                 US: 0\n",
      "(0028, 1050) Window Center                       DS: '40'\n",
      "(0028, 1051) Window Width                        DS: '400'\n",
      "(0028, 1052) Rescale Intercept                   DS: '-1024'\n",
      "(0028, 1053) Rescale Slope                       DS: '1'\n",
      "(0028, 1054) Rescale Type                        LO: 'HU'\n",
      "(7fe0, 0010) Pixel Data                          OW: Array of 524288 bytes\n"
     ]
    }
   ],
   "source": [
    "print dcm"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "-93.410004\n",
      "[[0 0 0 ..., 0 0 0]\n",
      " [0 0 0 ..., 0 0 0]\n",
      " [0 0 0 ..., 0 0 0]\n",
      " ..., \n",
      " [0 0 0 ..., 0 0 0]\n",
      " [0 0 0 ..., 0 0 0]\n",
      " [0 0 0 ..., 0 0 0]]\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAQgAAAEBCAYAAACE+Ts3AAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x104005650>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAQgAAAEBCAYAAACE+Ts3AAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nb3hPXbOlVgAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x119e87f90>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "img = dcm.pixel_array\n",
    "print dcm.SliceLocation  \n",
    "img[img == -2000] = 0\n",
    "print img\n",
    "\n",
    "plt.axis('off')\n",
    "plt.imshow(img)\n",
    "plt.show()\n",
    "\n",
    "plt.axis('off')\n",
    "plt.imshow(-img) # Invert colors with -\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "                                 id  cancer\n",
      "0  0015ceb851d7251b8f399e39779d1e7d       1\n",
      "1  0030a160d58723ff36d73f41b170ec21       0\n",
      "2  003f41c78e6acfa92430a057ac0b306e       0\n",
      "3  006b96310a37b36cccb2ab48d10b49a3       1\n",
      "4  008464bb8521d09a42985dd8add3d0d2       1\n"
     ]
    }
   ],
   "source": [
    "labels = pd.read_csv('stage1_labels.csv')\n",
    "print labels.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "# import csv\n",
    "# features = open('features.csv', 'wb')\n",
    "# wr = csv.writer(features)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "metadata": {
    "collapsed": false,
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "starting with new patient 00cba091fa4ad62cc3200a657aeb957e\n",
      "starting with new patient 0a099f2549429d29b32f349e95fb2244\n",
      "starting with new patient 0a0c32c9e08cc2ea76a71649de56be6d\n",
      "starting with new patient 0a38e7597ca26f9374f8ea2770ba870d\n",
      "starting with new patient 0acbebb8d463b4b9ca88cf38431aac69\n",
      "starting with new patient 0bd0e3056cbf23a1cb7f0f0b18446068\n",
      "starting with new patient 0c0de3749d4fe175b7a5098b060982a1\n",
      "starting with new patient 0c37613214faddf8701ca41e6d43f56e\n",
      "starting with new patient 0c59313f52304e25d5a7dcf9877633b1\n",
      "starting with new patient 0c60f4b87afcb3e2dfa65abbbf3ef2f9\n",
      "starting with new patient 0c98fcb55e3f36d0c2b6507f62f4c5f1\n",
      "starting with new patient 0c9d8314f9c69840e25febabb1229fa4\n",
      "starting with new patient 0ca943d821204ceb089510f836a367fd\n",
      "starting with new patient 0d06d764d3c07572074d468b4cff954f\n",
      "starting with new patient 0d19f1c627df49eb223771c28548350e\n",
      "starting with new patient 0d2fcf787026fece4e57be167d079383\n",
      "starting with new patient 0d941a3ad6c889ac451caf89c46cb92a\n",
      "starting with new patient 0ddeb08e9c97227853422bd71a2a695e\n",
      "starting with new patient 0de72529c30fe642bc60dcb75c87f6bd\n"
     ]
    }
   ],
   "source": [
    "a = []\n",
    "b = []\n",
    "ids = []\n",
    "imgs = []\n",
    "for i in os.listdir('./sample_images/'):\n",
    "    if '.DS_Store' not in i and '0b20184e0cd497028bdd155d9fb42dc9' not in i:\n",
    "        print \"starting with new patient\", i\n",
    "        label = int(labels[labels['id'] == i]['cancer'])\n",
    "#         print label\n",
    "        for j in os.listdir('./sample_images/'+i):\n",
    "            ids.append(i)\n",
    "            imgs.append(j)\n",
    "            b.append(label)\n",
    "            dcm = dicom.read_file('./sample_images/'+i+'/'+j)\n",
    "            img = dcm.pixel_array\n",
    "            img[img == -2000] = 0\n",
    "            img = np.sum(img, axis=1) ## sum of each row\n",
    "#             print img.shape\n",
    "            img = list(img.flatten())\n",
    "            loca = float(dcm.SliceLocation)\n",
    "            img.extend([loca])\n",
    "            a.append(img)\n",
    "\n",
    "#             break\n",
    "#         break\n",
    "df = pd.DataFrame()\n",
    "df['id'] = ids\n",
    "df['image'] = imgs\n",
    "df['label'] = b\n",
    "df['feat'] = a\n",
    "\n",
    "X = a\n",
    "Y = b "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "from sklearn.ensemble import RandomForestClassifier\n",
    "from sklearn.cross_validation import train_test_split\n",
    "from sklearn.metrics import accuracy_score\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "3408\n"
     ]
    }
   ],
   "source": [
    "# df = pd.read_csv('./features.csv')\n",
    "# X = np.arrray(df)\n",
    "print len(df)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 89,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "20\n"
     ]
    }
   ],
   "source": [
    "# X_train, X_test, y_train, y_test = train_test_split(X, Y, test_size=0.50, random_state=seed)\n",
    "\n",
    "patients = [i for i in os.listdir('./sample_images/') if '.DS_Store' not in i]\n",
    "print len(patients)\n",
    "patients_train = np.random.choice(patients, 15)\n",
    "\n",
    "X_train = []\n",
    "y_train = []\n",
    "X_test = []\n",
    "y_test = []\n",
    "\n",
    "for i in range(len(df)):\n",
    "    if df['id'][i] in patients_train:\n",
    "        X_train.append(np.array(df['feat'][i]))\n",
    "        y_train.append(df['label'][i])\n",
    "    else:\n",
    "        X_test.append(np.array(df['feat'][i]))\n",
    "        y_test.append(df['label'][i])\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 90,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "accuracy score 0.573326248672\n",
      "[0 0 0 ..., 0 0 1] [0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0] [1476, 1960, 2412, 2828, 3179, 3491, 3766, 3970, 4392, 4758, 4996, 5336, 5888, 6225, 6299, 6499, 6732, 6756, 7103, 7226, 7359, 7762, 8451, 9558, 11337, 13576, 15971, 17492, 17903, 17981, 17886, 17530, 17072, 16850, 16691, 16442, 22592, 45643, 82157, 111770, 125957, 133831, 141942, 149692, 157179, 164585, 170994, 177265, 184051, 189431, 193653, 198114, 202819, 208077, 212463, 215678, 219179, 223759, 227684, 230481, 233441, 236972, 240964, 244762, 247934, 250822, 254041, 257047, 260058, 263551, 266879, 269698, 272268, 275815, 279319, 283609, 288992, 293800, 299069, 305154, 310011, 313345, 316575, 320060, 322341, 325257, 331049, 336284, 340144, 344891, 348727, 351081, 355498, 360567, 365598, 371780, 377553, 380726, 382680, 385152, 388664, 390896, 393574, 397516, 399752, 402834, 404787, 406220, 408674, 409122, 410860, 414798, 416894, 416394, 418166, 422630, 422466, 421328, 424116, 427572, 427850, 423778, 420958, 421556, 421044, 420532, 420812, 420482, 422040, 424276, 423494, 422684, 425072, 429282, 432676, 434380, 438512, 444420, 449360, 452048, 452738, 454094, 456364, 456022, 455478, 456174, 457090, 457494, 457032, 457290, 458818, 459622, 457316, 454012, 451430, 448914, 448664, 449862, 449364, 449414, 450012, 450096, 451234, 453156, 457960, 466834, 472506, 472320, 471926, 472616, 470686, 465182, 461104, 459346, 458014, 456498, 453368, 451964, 452192, 451566, 451114, 453634, 454160, 453514, 452714, 452794, 457870, 460438, 460676, 461756, 461682, 461846, 464662, 467772, 468156, 468224, 469974, 473266, 475786, 477244, 478014, 479972, 483920, 485772, 486646, 489926, 492246, 494744, 497192, 498178, 499452, 500556, 502550, 505070, 507952, 511850, 513582, 513164, 513828, 516682, 518892, 518842, 518374, 517018, 520090, 522296, 519166, 516036, 518712, 521906, 519586, 516502, 516736, 520868, 520418, 517906, 515710, 514178, 515672, 515952, 516644, 517870, 516816, 517038, 519488, 520132, 520730, 521680, 521062, 519926, 522558, 527126, 524138, 520480, 522580, 520542, 518121, 518069, 515469, 516061, 518945, 517387, 516729, 520976, 519668, 516358, 516139, 515251, 516880, 519437, 519801, 519666, 521904, 519788, 518814, 520246, 518972, 517800, 516056, 516694, 519128, 519456, 519544, 515200, 512382, 515668, 517256, 517180, 517120, 516002, 515840, 517308, 516866, 516570, 517896, 518366, 514530, 512486, 512358, 509864, 505766, 499178, 492172, 486696, 482334, 478820, 475808, 472834, 469234, 468852, 469232, 464542, 459902, 459356, 459450, 458360, 459964, 463644, 466808, 466904, 466588, 467826, 468420, 470412, 474570, 476676, 476976, 481600, 485486, 488210, 493074, 491404, 484722, 482998, 483922, 482220, 477888, 473956, 471128, 468178, 464708, 462000, 465640, 469336, 467038, 465214, 464110, 461100, 461000, 464146, 465222, 461830, 458840, 455184, 450088, 447992, 447538, 446500, 443022, 440490, 440046, 438560, 436908, 435944, 434442, 432686, 431976, 430076, 427064, 427078, 427428, 423376, 421530, 424970, 424732, 418822, 412042, 407320, 405212, 404206, 402348, 397884, 391886, 387652, 383796, 379072, 375266, 371192, 366752, 363446, 359206, 354976, 352486, 350380, 345500, 336998, 328910, 321944, 314578, 307814, 302444, 297860, 293352, 289100, 281926, 271774, 263434, 256126, 247436, 237418, 224118, 204396, 176416, 138084, 100178, 82214, 80640, 81192, 83330, 90576, 104086, 119496, 125804, 106290, 69561, 43833, 32387, 28633, 28391, 28336, 27661, 27107, 27154, 27429, 27129, 26708, 26906, 26953, 26835, 26800, 26635, 26508, 26636, 26903, 27675, 28803, 30594, 32574, 35956, 40283, 45365, 50465, 56100, 61205, 68788, 76949, 85478, 94780, 102952, 112157, 120460, 128068, 136861, 148308, 165363, 194414, 223060, 225878, 204043, 170704, 124475, 71162, 34237, 22954, 21830, 21387, 20707, 20054, 19393, 19239, 18875, 18335, 18032, 17622, 17518, 17256, 16701, 16240, 16152, 15981, 15671, 15494, 14985, 14765, 14347, 14169, 13834, 13646, 12949, 12467, 11783, 11167, 10715, 10109, 9451, 8937, 8242, 7324, 6659, 5936, 4774, 3376, -316.200012]\n"
     ]
    }
   ],
   "source": [
    "model = RandomForestClassifier(random_state=seed)\n",
    "model.fit(X_train, y_train)\n",
    "predicted = model.predict(X_test)\n",
    "predicted_prob = model.predict_proba(X_test)\n",
    "# print 'log loss', log_loss(y_test, predicted)\n",
    "print 'accuracy score', accuracy_score(y_test, predicted)\n",
    "\n",
    "print predicted, y_test, X[0]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 72,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "# sample = pd.read_csv('./stage1_sample_submission.csv')\n",
    "# test_ids = sample['id']\n",
    "# print test_ids\n",
    "# print np.sum(predicted_prob)\n",
    "# predicted = model.predict(X_test)\n",
    "# sample['cancer'] = \n",
    "# sample.to_csv('submission2.csv', index=False)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>id</th>\n",
       "      <th>image</th>\n",
       "      <th>label</th>\n",
       "      <th>feat</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>00cba091fa4ad62cc3200a657aeb957e</td>\n",
       "      <td>034673134cbef5ea15ff9e0c8090500a.dcm</td>\n",
       "      <td>0</td>\n",
       "      <td>[1476, 1960, 2412, 2828, 3179, 3491, 3766, 397...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>00cba091fa4ad62cc3200a657aeb957e</td>\n",
       "      <td>0484f5a7f55eb7b6743cadaffcce586d.dcm</td>\n",
       "      <td>0</td>\n",
       "      <td>[1208, 1630, 2088, 2464, 2663, 3042, 3307, 355...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>00cba091fa4ad62cc3200a657aeb957e</td>\n",
       "      <td>053a0460fb45227bd8e0e7b514a71e8e.dcm</td>\n",
       "      <td>0</td>\n",
       "      <td>[1318, 1700, 2176, 2641, 2930, 3107, 3308, 366...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>00cba091fa4ad62cc3200a657aeb957e</td>\n",
       "      <td>0678ae61b7c56dfaf35477f6d01a6ffc.dcm</td>\n",
       "      <td>0</td>\n",
       "      <td>[1426, 1894, 2428, 2789, 3061, 3444, 3835, 397...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>00cba091fa4ad62cc3200a657aeb957e</td>\n",
       "      <td>06ba0baeb0f26fb7aed5249eebbb77a9.dcm</td>\n",
       "      <td>0</td>\n",
       "      <td>[1376, 2032, 2398, 2750, 3256, 4247, 5684, 736...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>00cba091fa4ad62cc3200a657aeb957e</td>\n",
       "      <td>0791f624c2f313bfe490ac940947977e.dcm</td>\n",
       "      <td>0</td>\n",
       "      <td>[1292, 1842, 2274, 2495, 2887, 3180, 3346, 347...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>00cba091fa4ad62cc3200a657aeb957e</td>\n",
       "      <td>08eb3a007acc90a45b6e1b92fb479adc.dcm</td>\n",
       "      <td>0</td>\n",
       "      <td>[1260, 1786, 2004, 2411, 2659, 2836, 3144, 338...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>00cba091fa4ad62cc3200a657aeb957e</td>\n",
       "      <td>090173a7ed57a872b62fd7c0b1d24be5.dcm</td>\n",
       "      <td>0</td>\n",
       "      <td>[1200, 1710, 1908, 2231, 2570, 2875, 3129, 332...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>00cba091fa4ad62cc3200a657aeb957e</td>\n",
       "      <td>0a291d1b12b86213d813e3796f14b329.dcm</td>\n",
       "      <td>0</td>\n",
       "      <td>[1194, 1836, 2036, 2313, 2623, 2952, 3135, 331...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>00cba091fa4ad62cc3200a657aeb957e</td>\n",
       "      <td>0af07f64d14d6b0451036cc742a96fbf.dcm</td>\n",
       "      <td>0</td>\n",
       "      <td>[1130, 1550, 1974, 2153, 2566, 2883, 2988, 325...</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                 id                                 image  \\\n",
       "0  00cba091fa4ad62cc3200a657aeb957e  034673134cbef5ea15ff9e0c8090500a.dcm   \n",
       "1  00cba091fa4ad62cc3200a657aeb957e  0484f5a7f55eb7b6743cadaffcce586d.dcm   \n",
       "2  00cba091fa4ad62cc3200a657aeb957e  053a0460fb45227bd8e0e7b514a71e8e.dcm   \n",
       "3  00cba091fa4ad62cc3200a657aeb957e  0678ae61b7c56dfaf35477f6d01a6ffc.dcm   \n",
       "4  00cba091fa4ad62cc3200a657aeb957e  06ba0baeb0f26fb7aed5249eebbb77a9.dcm   \n",
       "5  00cba091fa4ad62cc3200a657aeb957e  0791f624c2f313bfe490ac940947977e.dcm   \n",
       "6  00cba091fa4ad62cc3200a657aeb957e  08eb3a007acc90a45b6e1b92fb479adc.dcm   \n",
       "7  00cba091fa4ad62cc3200a657aeb957e  090173a7ed57a872b62fd7c0b1d24be5.dcm   \n",
       "8  00cba091fa4ad62cc3200a657aeb957e  0a291d1b12b86213d813e3796f14b329.dcm   \n",
       "9  00cba091fa4ad62cc3200a657aeb957e  0af07f64d14d6b0451036cc742a96fbf.dcm   \n",
       "\n",
       "   label                                               feat  \n",
       "0      0  [1476, 1960, 2412, 2828, 3179, 3491, 3766, 397...  \n",
       "1      0  [1208, 1630, 2088, 2464, 2663, 3042, 3307, 355...  \n",
       "2      0  [1318, 1700, 2176, 2641, 2930, 3107, 3308, 366...  \n",
       "3      0  [1426, 1894, 2428, 2789, 3061, 3444, 3835, 397...  \n",
       "4      0  [1376, 2032, 2398, 2750, 3256, 4247, 5684, 736...  \n",
       "5      0  [1292, 1842, 2274, 2495, 2887, 3180, 3346, 347...  \n",
       "6      0  [1260, 1786, 2004, 2411, 2659, 2836, 3144, 338...  \n",
       "7      0  [1200, 1710, 1908, 2231, 2570, 2875, 3129, 332...  \n",
       "8      0  [1194, 1836, 2036, 2313, 2623, 2952, 3135, 331...  \n",
       "9      0  [1130, 1550, 1974, 2153, 2566, 2883, 2988, 325...  "
      ]
     },
     "execution_count": 40,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.head(10)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "anaconda-cloud": {},
  "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": "2.7.10"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 0
}