BDT_justXGB.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "from __future__ import division\n",
    "from __future__ import print_function\n",
    "\n",
    "import matplotlib\n",
    "\n",
    "import pandas\n",
    "\n",
    "import numpy as np\n",
    "import xgboost as xgb\n",
    "from sklearn.model_selection import train_test_split, StratifiedKFold\n",
    "from sklearn.metrics import accuracy_score\n",
    "\n",
    "import matplotlib.pyplot as plt\n",
    "\n",
    "from sklearn.model_selection import cross_val_score\n",
    "from sklearn.ensemble import RandomForestClassifier, ExtraTreesClassifier\n",
    "from sklearn.pipeline import Pipeline\n",
    "from sklearn import tree\n",
    "import pydotplus \n",
    "\n",
    "import scipy\n",
    "from scipy import sqrt\n",
    "\n",
    "import sys\n",
    "sys.path.append('./modules/')\n",
    "from Processors import *"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Skim size, Acceptance, xSec and luminosity"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "luminosity = 100\n",
    "\n",
    "signalFullSize = [50000, 50000, 49600, 50000, 50000, 50000, 100000, 100000, 100000, 99800]\n",
    "signalSkimSize = [403, 394, 390, 372, 479, 655, 1619, 1858, 2008, 2239]\n",
    "signalXSec = [2.4524985139, 3.2882620826, 3.8318006272, 4.5185294336, 4.9327584051,\n",
    "              5.6041453978, 5.7246882202, 5.6830461542, 5.5990314598, 5.3528318771]\n",
    "signalAcceptance = [a/b for a,b in zip(signalSkimSize, signalFullSize)]\n",
    "\n",
    "allSignalFullSize = sum(signalFullSize)\n",
    "allSignalSkimSize = sum(signalSkimSize)\n",
    "allSignalAcceptance = allSignalSkimSize/allSignalFullSize\n",
    "\n",
    "backgroundFullSize = [92925926, 47502020]\n",
    "backgroundSkimSize = [166157, 1896]\n",
    "backgroundXSec = [816, 61526.7]\n",
    "backgroundAcceptance = [a/b for a,b in zip(backgroundSkimSize, backgroundFullSize)]\n",
    "\n",
    "allBackgroundFullSize = sum(backgroundFullSize)\n",
    "allBackgroundSkimSize = sum(backgroundSkimSize)\n",
    "allBackGroundAcceptance = allBackgroundSkimSize/allBackgroundFullSize"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Calculate significância"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "def SignalSignificanceSelector(SignalSignificance):\n",
    "    temp = sorted(SignalSignificance)[0:min(len(SignalSignificance), 3)]\n",
    "    return sum(temp)/len(temp)\n",
    "\n",
    "def calculateFOM(Signals, Backgounds, signalIdx, backgroundIdx, minimum=False):\n",
    "    s_signal, s_background = [], 0\n",
    "    for i, signal in enumerate(Signals):\n",
    "        s_signal.append(len(signal) * signalXSec[signalIdx[i]] / signalFullSize[signalIdx[i]] * luminosity)\n",
    "    \n",
    "    for i, background in enumerate(Backgounds):\n",
    "        s_background += len(background) * backgroundXSec[backgroundIdx[i]] / backgroundFullSize[backgroundIdx[i]] * luminosity\n",
    "    \n",
    "    if s_background != 0:\n",
    "        return SignalSignificanceSelector(s_signal) / sqrt(s_background)\n",
    "    else:\n",
    "        return -1"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "def mergeDataFrame(data):\n",
    "    allData = data[0]\n",
    "    for _data in data[1:]:\n",
    "        allData = allData.append(_data, ignore_index=True)\n",
    "    return allData"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "403 events for Higgs with 260 GeV\n",
      "394 events for Higgs with 270 GeV\n",
      "390 events for Higgs with 280 GeV\n",
      "372 events for Higgs with 300 GeV\n",
      "479 events for Higgs with 320 GeV\n",
      "655 events for Higgs with 400 GeV\n",
      "1619 events for Higgs with 500 GeV\n",
      "1858 events for Higgs with 550 GeV\n",
      "2008 events for Higgs with 600 GeV\n",
      "2239 events for Higgs with 700 GeV\n",
      "166157 TT_Tune events\n",
      "1896 WJetsToLNu_Tune events\n",
      "\n",
      "10417 events of Higgs\n",
      "168053 background events\n"
     ]
    }
   ],
   "source": [
    "loc = './Data/'\n",
    "\n",
    "masses = [260, 270, 280, 300, 320, 400, 500, 550, 600, 700]\n",
    "backgroundLabels = ['TT', 'WJets']\n",
    "\n",
    "signalData = []\n",
    "backgroundData = []\n",
    "\n",
    "for mass in masses:\n",
    "    signalData.append(pandas.read_csv(loc + \"GluGluToRadionToHHTo2B2Tau_M-\" + \n",
    "                                      str(mass) + \"_narrow_13TeV-madgraph.csv\"))\n",
    "\n",
    "# same as 2 but only the first half of the events\n",
    "#backgroundData.append(pandas.read_csv(loc + \"TT_TuneCUETP8M1_13TeV-powheg-pythia8_1.csv\"))\n",
    "backgroundData.append(pandas.read_csv(loc + \"TT_TuneCUETP8M1_13TeV-powheg-pythia8_2.csv\"))\n",
    "backgroundData.append(pandas.read_csv(loc + \"WJetsToLNu_TuneCUETP8M1_13TeV-amcatnloFXFX-pythia8.csv\"))\n",
    "\n",
    "\n",
    "# drop columns without header\n",
    "for mass, signal_data in zip(masses, signalData):\n",
    "    signal_data.drop([x for x in signal_data.columns if 'Unnamed' in x], axis=1, inplace=True)\n",
    "    signal_data['gen_label'] = str(mass)\n",
    "\n",
    "for label, background_data in zip(backgroundLabels, backgroundData):\n",
    "    background_data.drop([x for x in background_data.columns if 'Unnamed' in x], axis=1, inplace=True)\n",
    "    background_data['gen_label'] = label\n",
    "\n",
    "allSignalData = mergeDataFrame(signalData)\n",
    "allBackgroundData = mergeDataFrame(backgroundData)\n",
    "\n",
    "for i in range(len(masses)):\n",
    "    print(\"{} events for Higgs with {} GeV\".format(len(signalData[i]), masses[i]))\n",
    "print(\"{} TT_Tune events\".format(len(backgroundData[0])))\n",
    "print(\"{} WJetsToLNu_Tune events\".format(len(backgroundData[1])), end='\\n\\n')\n",
    "print(\"{} events of Higgs\".format(len(allSignalData)))\n",
    "print(\"{} background events\".format(len(allBackgroundData)))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "dropMctMatch = True\n",
    "\n",
    "if dropMctMatch:\n",
    "    # drop signal that does not match thing\n",
    "    allSignalData = allSignalData[allSignalData['gen_mctMatch']]\n",
    "\n",
    "allData = allSignalData.append(allBackgroundData, ignore_index=True)\n",
    "allLabels = np.append(np.ones(len(allSignalData)),np.zeros(len(allBackgroundData)))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "particles = ['t_0', 't_1', 'b_0', 'b_1', 'h_tt', 'h_bb', 'diH']\n",
    "\n",
    "for p in particles:\n",
    "    moveToCartesian(allData, p) #Move pT, eta, and phi to p_x, p_y, and p_z\n",
    "    addEnergy(allData, p) #Calculate energy and absolute momentum\n",
    "    \n",
    "moveToCartesian(allData, 'mPT', False)  #Move Missing pT and phi to p_x and p_y\n",
    "addAbsMom(allData, 'mPT', False) #Calculate absolute missing transverse momentum\n",
    "addMT(allData, allData['t_1_pT'], allData['t_1_phi'], 't_1') #Calculate transverse mass of tau_mu"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## select train variables"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "genFeatures = [var for var in allData.columns if str.startswith(var, \"gen\")] #Generator features; not for training\n",
    "trainFeatures = [var for var in allData.columns if var not in genFeatures] #Reconstructed features; ok for training"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "pTEtaPhi = [var for var in trainFeatures for x in ['phi', 't_1_mass'] if x in var] # 'pT', 'eta', \n",
    "other = [var for var in trainFeatures for x in ['diH_kinFit_prob'] if x in var]\n",
    "trainFeatures = [var for var in trainFeatures if var not in pTEtaPhi + other]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "fsFeatures = [var for var in trainFeatures for p in particles + ['mPT'] if p in var]\n",
    "shapes = [var for var in trainFeatures for x in ['aplan', 'dShape', 'spher', 'upsilon'] if x in var]\n",
    "shapeFeatures = [var for var in trainFeatures if var in shapes]\n",
    "eventKinematicFeatures = ['centrality', 'eVis', 'hT', 'sT']\n",
    "jetFeatures = [var for var in trainFeatures if 'Jet' in var and 'Jets' not in var]\n",
    "multiplicityFeatures = ['nBJets', 'nJets', 'nPhotons', 'nTauJets']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Training on 64 features ['t_0_pT', 't_0_eta', 't_0_mass', 't_1_pT', 't_1_eta', 'b_0_pT', 'b_0_eta', 'b_0_mass', 'b_0_csv', 'b_1_pT', 'b_1_eta', 'b_1_mass', 'b_1_csv', 'mPT_pT', 'h_tt_pT', 'h_tt_eta', 'h_tt_mass', 'h_tt_svFit_mass', 'h_bb_pT', 'h_bb_eta', 'h_bb_mass', 'diH_pT', 'diH_eta', 'diH_mass', 'diH_kinFit_mass', 't_0_px', 't_0_py', 't_0_pz', 't_0_|p|', 't_0_E', 't_1_px', 't_1_py', 't_1_pz', 't_1_|p|', 't_1_E', 'b_0_px', 'b_0_py', 'b_0_pz', 'b_0_|p|', 'b_0_E', 'b_1_px', 'b_1_py', 'b_1_pz', 'b_1_|p|', 'b_1_E', 'h_tt_px', 'h_tt_py', 'h_tt_pz', 'h_tt_|p|', 'h_tt_E', 'h_bb_px', 'h_bb_py', 'h_bb_pz', 'h_bb_|p|', 'h_bb_E', 'diH_px', 'diH_py', 'diH_pz', 'diH_|p|', 'diH_E', 'mPT_px', 'mPT_py', 'mPT_|p|', 't_1_mT']\n"
     ]
    }
   ],
   "source": [
    "classTrainFeatures = fsFeatures\n",
    "print(\"Training on {} features {}\". format(len(classTrainFeatures), classTrainFeatures))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [],
   "source": [
    "removeFromTrain = [700]\n",
    "\n",
    "allDataTrainFeatures = allData\n",
    "\n",
    "for remove in removeFromTrain:\n",
    "    allLabels = allLabels[allDataTrainFeatures['gen_label'] != str(remove)]\n",
    "    allDataTrainFeatures = allDataTrainFeatures[allDataTrainFeatures['gen_label'] != str(remove)]\n",
    "\n",
    "X_train_full, X_test_full, y_train_full, y_test_full = train_test_split(allDataTrainFeatures, allLabels, test_size=0.2, random_state=2345)\n",
    "\n",
    "allDataTrainFeatures = allDataTrainFeatures[classTrainFeatures]\n",
    "\n",
    "X_train, X_test, y_train, y_test = train_test_split(allDataTrainFeatures, np.copy(allLabels), test_size=0.2, random_state=2345)\n",
    "for var in (X_train, X_test, X_train_full, X_test_full):\n",
    "    var.reset_index(inplace=True, drop=True)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Remove points who fail conversion to float32"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "def cleanData(X, y):\n",
    "    \"\"\"\n",
    "    Recives data and labels (X and y)\n",
    "    \n",
    "    Returns: X, y after removing points that would fail to convert to float32 \n",
    "    \"\"\"\n",
    "    temp = X[classTrainFeatures]\n",
    "    over  = (temp > np.finfo(np.float32).max)\n",
    "    under = (temp < np.finfo(np.float32).min)\n",
    "    selecting = pandas.Series(np.zeros(len(X)), dtype=np.bool)\n",
    "    \n",
    "    for label in over.columns:\n",
    "        if label != 'gen_label':\n",
    "            selecting = selecting | over[label] | under[label]\n",
    "    \n",
    "    y = y[np.logical_not(selecting)]#.astype('float32')\n",
    "    X = X[np.logical_not(selecting)].reset_index(drop=True)#.values.astype('float32')\n",
    "    \n",
    "    return X, y"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [],
   "source": [
    "X_train, y_train = cleanData(X_train, y_train)\n",
    "X_test , y_test  = cleanData(X_test , y_test )\n",
    "X_train_full, y_train_full = cleanData(X_train_full, y_train_full)\n",
    "X_test_full , y_test_full  = cleanData(X_test_full , y_test_full )"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "scale = len(allBackgroundData)/len(allSignalData)\n",
    "\n",
    "weights_train = np.ones(len(y_train))\n",
    "weights_test = np.ones(len(y_test))\n",
    "\n",
    "weights_train[y_train == 1] = scale\n",
    "weights_test[y_test == 1] = scale"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "CPU times: user 3min 34s, sys: 380 ms, total: 3min 34s\n",
      "Wall time: 33.1 s\n"
     ]
    }
   ],
   "source": [
    "%%time\n",
    "# fit model on training data\n",
    "model = xgb.XGBClassifier(random_state=1337, silent=False,\n",
    "                          scale_pos_weight=scale,\n",
    "                          objective='binary:logistic',\n",
    "                          n_estimators=100, eval_metric='auc',\n",
    "                          max_depth=4,\n",
    "                          n_jobs=8) #NOTE: number os threads to execute\n",
    "model.fit(X_train, y_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# make predictions for test data\n",
    "y_pred = model.predict(X_test)\n",
    "predictions = [round(value) for value in y_pred]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Accuracy           : 95.76591782097545%\n",
      "Accuracy signal    : 91.56118143459916%\n",
      "Accuracy background: 95.85475291950198%\n"
     ]
    }
   ],
   "source": [
    "# evaluate predictions\n",
    "accuracy = accuracy_score(y_test, predictions)\n",
    "print(\"Accuracy           : {:10}%\".format(accuracy * 100.0))\n",
    "\n",
    "accuracy = accuracy_score(y_test, predictions, sample_weight=y_test)\n",
    "print(\"Accuracy signal    : {:10}%\".format(accuracy * 100.0))\n",
    "\n",
    "accuracy = accuracy_score(y_test, predictions, sample_weight=np.logical_not(y_test))\n",
    "print(\"Accuracy background: {:10}%\".format(accuracy * 100.0))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0.964241706224\n",
      "0.940533493299\n"
     ]
    }
   ],
   "source": [
    "print(model.score(X_train, y_train, sample_weight=weights_train))\n",
    "print(model.score(X_test, y_test, sample_weight=weights_test))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAgMAAAHxCAYAAADwab/YAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nRU5ErkJggg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f52af0e3438>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "from sklearn import metrics\n",
    "\n",
    "plt.figure(figsize=[8, 8])\n",
    "plt.plot(*metrics.roc_curve(y_train, model.predict_proba(X_train)[:,1])[:2],\n",
    "         label=r'Dev, auc=${:.4f}$'.format(metrics.roc_auc_score(y_train, model.predict_proba(X_train)[:,1])),\n",
    "         color='b', linestyle='dashed')\n",
    "plt.plot(*metrics.roc_curve(y_test, model.predict_proba(X_test)[:,1])[:2], \n",
    "         label=r'Val, auc=${:.4f}$'.format(metrics.roc_auc_score(y_test, model.predict_proba(X_test)[:,1])),\n",
    "         color='b')\n",
    "plt.plot([0, 1], [0, 1], 'k--', label='No discrimination')\n",
    "plt.xlabel('Background acceptance', fontsize=24, color='black')\n",
    "plt.ylabel('Signal acceptance', fontsize=24, color='black')\n",
    "plt.legend(loc='best', fontsize=16)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "260 0.769141\n",
      "270 0.752057\n",
      "280 0.618748\n",
      "300 0.670057\n",
      "320 0.595731\n",
      "400 0.707749\n",
      "500 0.923004\n",
      "550 0.959421\n",
      "600 0.967992\n",
      "700 0.936069\n"
     ]
    }
   ],
   "source": [
    "backgroundTestData, signalTestData = [], []\n",
    "backgroundScore, signalScore = [], []\n",
    "\n",
    "for label in backgroundLabels:\n",
    "    selection = X_test_full['gen_label'] == label\n",
    "    var = X_test[selection]\n",
    "    \n",
    "    backgroundTestData.append(var)\n",
    "    backgroundScore.append(model.predict_proba(var[classTrainFeatures])[:,1])\n",
    "    \n",
    "trainMasses = masses[:]\n",
    "for mass in removeFromTrain: trainMasses.remove(mass)\n",
    "    \n",
    "for mass in trainMasses:\n",
    "    selection = X_test_full['gen_label'] == str(mass)\n",
    "    var = X_test[selection]\n",
    "    \n",
    "    signalTestData.append(var)\n",
    "    signalScore.append(model.predict_proba(var[classTrainFeatures])[:,1])\n",
    "    print(mass, np.average(model.predict_proba(var[classTrainFeatures])[:,1]))\n",
    "\n",
    "    \n",
    "for mass in removeFromTrain:\n",
    "    selection = allData['gen_label'] == str(mass)\n",
    "    var = allData[selection][classTrainFeatures]\n",
    "    \n",
    "    signalTestData.append(var)\n",
    "    signalScore.append(model.predict_proba(var[classTrainFeatures])[:,1])\n",
    "    print(mass, np.average(model.predict_proba(var[classTrainFeatures])[:,1]))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "def FOMAfterCut(p):\n",
    "    cutSignal, cutBackground = [], []\n",
    "    for data, scores in zip(signalTestData, signalScore):\n",
    "        cutSignal.append(data[(scores > p)])\n",
    "    for data, scores in zip(backgroundTestData, backgroundScore):\n",
    "        cutBackground.append(data[(scores > p)])\n",
    "\n",
    "    return calculateFOM(cutSignal, cutBackground, range(len(cutSignal)), range(len(cutBackground)))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {
    "scrolled": false
   },
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAX0AAAD8CAYAAACb4nSYAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nJn2llIogmvSVUiqC/H94LQ76NTd8mgAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f52ae4f7a90>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "x = np.linspace(0,.99,100)\n",
    "y = []\n",
    "for i in x:\n",
    "    #NOTE: since this is using only the test data this must be divided by sqrt(test_size)\n",
    "    y.append(FOMAfterCut(i)/sqrt(.2))\n",
    "\n",
    "plt.plot(x, y)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAgMAAAHxCAYAAADwab/YAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nilUDgQcAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f52ae4d3e48>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.figure(figsize=[8, 8])\n",
    "# only working for dropMctMatch = True\n",
    "\n",
    "for mass, signal in sorted(zip(trainMasses+removeFromTrain, signalTestData)):\n",
    "    # requires gen_mctMatch to be false for background\n",
    "    signalAndBackground = mergeDataFrame(backgroundTestData + [signal])\n",
    "    backgroundSize = sum([len(i) for i in backgroundTestData])\n",
    "    signalAndBackgroundLabel = np.append(np.zeros(backgroundSize, dtype=int), np.ones(len(signal), dtype=int))\n",
    "    \n",
    "    plt.plot(*metrics.roc_curve(signalAndBackgroundLabel, model.predict_proba(signalAndBackground)[:,1])[:2],\n",
    "             label=str(mass) + ' auc: {:.4f}'.format(\n",
    "             metrics.roc_auc_score(signalAndBackgroundLabel, model.predict_proba(signalAndBackground)[:,1])))\n",
    "\n",
    "plt.plot([0, 1], [0, 1], 'k--', label='No discrimination')\n",
    "plt.xlabel('Background acceptance', fontsize=24, color='black')\n",
    "plt.ylabel('Signal acceptance', fontsize=24, color='black')\n",
    "plt.legend(loc='best', fontsize=16)\n",
    "#plt.xlim(0.,0.4)\n",
    "#plt.ylim(0.6,1.)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "#from sklearn.model_selection import GridSearchCV\n",
    "#from sklearn.metrics import make_scorer\n",
    "#asd = GridSearchCV(model, {'max_depth': range(1,7)}, verbose=2, n_jobs=1, scoring=make_scorer(metrics.roc_auc_score))\n",
    "\n",
    "#asd.fit(X_train, y_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "#asd.cv_results_"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {
    "collapsed": true,
    "scrolled": true
   },
   "outputs": [],
   "source": [
    "#plt.plot(asd.cv_results_['param_max_depth'].data,\n",
    "#         asd.cv_results_['mean_test_score'])\n",
    "#plt.plot(asd.cv_results_['param_max_depth'].data,\n",
    "#         asd.cv_results_['mean_train_score'])\n",
    "#plt.show()"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "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": "3.5.2"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}