Sig_back_separate_training_EarlyStop.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Using TensorFlow backend.\n"
     ]
    }
   ],
   "source": [
    "from pprint import pprint\n",
    "import math\n",
    "import numpy as np\n",
    "import h5py\n",
    "import matplotlib.pyplot as plt\n",
    "import scipy\n",
    "from scipy import ndimage\n",
    "import tensorflow as tf\n",
    "import keras\n",
    "from tensorflow.python.framework import ops\n",
    "#from cnn_utils import *\n",
    "import os\n",
    "#os.environ['CUDA_VISIBLE_DEVICES']='-1'\n",
    "import deepdish.io as io\n",
    "from keras.models import Sequential, Model\n",
    "from keras.layers import Input, Add, Dense, Dropout, LeakyReLU\n",
    "from keras.callbacks import ModelCheckpoint\n",
    "from keras.wrappers.scikit_learn import KerasRegressor\n",
    "from keras.utils import multi_gpu_model\n",
    "from keras.callbacks import EarlyStopping\n",
    "from keras.callbacks import ModelCheckpoint\n",
    "from keras.models import load_model\n",
    "from sklearn.model_selection import cross_val_score, cross_val_predict, train_test_split\n",
    "from sklearn.model_selection import KFold\n",
    "from sklearn.preprocessing import StandardScaler, MinMaxScaler\n",
    "from sklearn.pipeline import Pipeline\n",
    "from sklearn.metrics import r2_score, mean_squared_error, mean_absolute_error\n",
    "from sklearn.metrics import roc_curve, auc\n",
    "import random"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "data_sig_A = io.load(os.path.join('../allData/MEM-A-dbsigwSpectator.h5'))\n",
    "data_sig_B = io.load(os.path.join('../allData/MEM-B-dbsigwSpectator.h5'))\n",
    "data_back_A = io.load(os.path.join('../allData/MEM-A-dbbkgwSpectator.h5'))\n",
    "data_back_B = io.load(os.path.join('../allData/MEM-B-dbbkgwSpectator.h5'))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "X_Big_train = data_sig_A['X']\n",
    "target_BigSig_train = data_sig_A['signal_lh']\n",
    "y_BigSig_train = data_sig_A['y']\n",
    "target_BigBack_train = data_back_A['bkg_lh']\n",
    "y_BigBack_train = data_back_A['y']\n",
    "\n",
    "X_Big_test = data_sig_B['X']\n",
    "target_BigSig_test = data_sig_B['signal_lh']\n",
    "y_BigSig_test = data_sig_B['y']\n",
    "target_BigBack_test = data_back_B['bkg_lh']\n",
    "y_BigBack_test = data_back_B['y']\n",
    "\n",
    "softTarget_train = target_BigSig_train/(target_BigSig_train+(0.1*target_BigBack_train))\n",
    "softTarget_test = target_BigSig_test/(target_BigSig_test+(0.1*target_BigBack_test))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "#Defining Sequential Model\n",
    "\n",
    "#def sequential_model():\n",
    "   # seqModel = Sequential()\n",
    "   # seqModel.add(Dense(32, input_dim=30, kernel_initializer='glorot_normal', activation='relu'))\n",
    "   # seqModel.add(Dense(12, kernel_initializer='glorot_normal', activation='relu'))\n",
    "   # seqModel.add(Dense(1, kernel_initializer='glorot_normal',activation='sigmoid'))\n",
    "   # seqModel.compile(loss='mean_squared_error', optimizer='adam')\n",
    "   # return seqModel\n",
    "\n",
    "def sequential_model():\n",
    "    seqModel = Sequential()\n",
    "    seqModel.add(Dense(1024, input_dim=30, kernel_initializer='glorot_normal', activation='relu'))\n",
    "    seqModel.add(Dense(512, kernel_initializer='glorot_normal', activation='relu'))\n",
    "    seqModel.add(Dense(256, kernel_initializer='glorot_normal', activation='relu'))\n",
    "    seqModel.add(Dense(128, kernel_initializer='glorot_normal', activation='relu'))\n",
    "    seqModel.add(Dense(64, kernel_initializer='glorot_normal', activation='relu'))\n",
    "    seqModel.add(Dense(32, kernel_initializer='glorot_normal', activation='relu'))\n",
    "    seqModel.add(Dense(16, kernel_initializer='glorot_normal', activation='relu'))\n",
    "    seqModel.add(Dense(8, kernel_initializer='glorot_normal', activation='relu'))\n",
    "    seqModel.add(Dense(1, kernel_initializer='glorot_normal',activation='sigmoid'))\n",
    "    seqModel.compile(loss='binary_crossentropy', optimizer='adam')\n",
    "    return seqModel"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "es = EarlyStopping(monitor='val_loss', mode='min', verbose=1)\n",
    "#mc = ModelCheckpoint('best_model.h5', monitor='val_acc', mode='max', verbose=1, save_best_only=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "WARNING:tensorflow:From /home/debo/env5/local/lib/python2.7/site-packages/tensorflow/python/framework/op_def_library.py:263: colocate_with (from tensorflow.python.framework.ops) is deprecated and will be removed in a future version.\n",
      "Instructions for updating:\n",
      "Colocations handled automatically by placer.\n",
      "WARNING:tensorflow:From /home/debo/env5/local/lib/python2.7/site-packages/tensorflow/python/ops/math_ops.py:3066: to_int32 (from tensorflow.python.ops.math_ops) is deprecated and will be removed in a future version.\n",
      "Instructions for updating:\n",
      "Use tf.cast instead.\n",
      "Train on 3028985 samples, validate on 1009662 samples\n",
      "Epoch 1/100\n",
      "3028985/3028985 [==============================] - 115s 38us/step - loss: 0.0473 - val_loss: 0.0288\n",
      "Epoch 2/100\n",
      "3028985/3028985 [==============================] - 114s 38us/step - loss: 0.0215 - val_loss: 0.0187\n",
      "Epoch 3/100\n",
      "3028985/3028985 [==============================] - 114s 38us/step - loss: 0.0157 - val_loss: 0.0149\n",
      "Epoch 4/100\n",
      "3028985/3028985 [==============================] - 113s 37us/step - loss: 0.0128 - val_loss: 0.0142\n",
      "Epoch 5/100\n",
      "3028985/3028985 [==============================] - 113s 37us/step - loss: 0.0111 - val_loss: 0.0111\n",
      "Epoch 6/100\n",
      "3028985/3028985 [==============================] - 117s 39us/step - loss: 0.0097 - val_loss: 0.0106\n",
      "Epoch 7/100\n",
      "3028985/3028985 [==============================] - 120s 40us/step - loss: 0.0087 - val_loss: 0.0109\n",
      "Epoch 00007: early stopping\n"
     ]
    }
   ],
   "source": [
    "sc_X = StandardScaler()\n",
    "X_train = sc_X.fit_transform(X_Big_train)\n",
    "X_test = sc_X.fit_transform(X_Big_test)\n",
    "model = sequential_model()\n",
    "history = model.fit(X_train, y_BigSig_train, validation_data=(X_test, y_BigSig_test), epochs=100, batch_size=256, verbose=1, callbacks=[es])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1009662/1009662 [==============================] - 11s 11us/step\n"
     ]
    }
   ],
   "source": [
    "sig_predict = model.predict(X_test,batch_size=256, verbose=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/debo/env5/local/lib/python2.7/site-packages/matplotlib/axes/_axes.py:6571: UserWarning: The 'normed' kwarg is deprecated, and has been replaced by the 'density' kwarg.\n",
      "  warnings.warn(\"The 'normed' kwarg is deprecated, and has been \"\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "sigSoft= sig_predict[np.where(y_BigSig_test==1)]\n",
    "bkgSoft = sig_predict[np.where(y_BigSig_test==0)]\n",
    "_=plt.hist(sigSoft,bins=100, histtype='step', label='signal', normed = 1)\n",
    "                           \n",
    "_=plt.hist(bkgSoft,bins=100, histtype='step', color= 'red', label='background', normed = 1)\n",
    "\n",
    "_=plt.yscale('log')\n",
    "\n",
    "_=plt.legend(loc='upper center')\n",
    "\n",
    "_=plt.title('Signal_LH_to_(Sig_LH+0.1*Bkg_LH)')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/debo/env5/lib/python2.7/site-packages/ipykernel_launcher.py:4: RuntimeWarning: divide by zero encountered in divide\n",
      "  after removing the cwd from sys.path.\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "fpr, tpr, thresholds = roc_curve(y_BigSig_test,sig_predict)\n",
    "plt.figure()\n",
    "##plt.plot(fpr, tpr)\n",
    "plt.plot(tpr, 1/fpr)\n",
    "plt.xlim([0.75, 1])\n",
    "#plt.ylim([1, 10000])\n",
    "plt.ylabel('1/Bkg_eff')\n",
    "plt.xlabel('Sig_eff')\n",
    "plt.title('Receiver operating characteristic')\n",
    "plt.yscale('log')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# evaluate the model\n",
    "#_, train_acc = model.evaluate(X_train, y_BigSig_train, verbose=1)\n",
    "#_, test_acc = model.evaluate(X_test, y_BigSig_test, verbose=1)\n",
    "#print('Train: %.3f, Test: %.3f' % (train_acc, test_acc))\n",
    "# plot training history\n",
    "plt.plot(history.history['loss'], label='train')\n",
    "plt.plot(history.history['val_loss'], label='test')\n",
    "plt.legend()\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Shuffling columns (0, 4)\n",
      "Shuffling columns (8, 12)\n",
      "Shuffling columns (15, 19)\n",
      "Shuffling columns (23, 27)\n"
     ]
    }
   ],
   "source": [
    "d = io.load('../allData/MEM-A-dbsigwSpectator.h5')\n",
    "#ix = [random.sample(range(4), 4) for _ in xrange(d['X'].shape[0])]\n",
    "ix = np.argsort(d['X'][:, :4])[:, ::-1]\n",
    "\n",
    "def _pairwise(iterable):\n",
    "        '''s -> (s0, s1), (s2, s3), (s4, s5), ...'''\n",
    "        a = iter(iterable)\n",
    "        return izip(a, a)\n",
    "\n",
    "cols = [(0, 4), (8, 12), (15, 19), (23, 27)]\n",
    "for slc in cols:\n",
    "      print 'Shuffling columns {}'.format(slc)\n",
    "      d['X'][:, slice(*slc)] = np.array([row[sh] for row, sh in zip(d['X'][:, slice(*slc)], ix)])\n",
    "\n",
    "X_ordered = d['X']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Shuffling columns (0, 4)\n",
      "Shuffling columns (8, 12)\n",
      "Shuffling columns (15, 19)\n",
      "Shuffling columns (23, 27)\n"
     ]
    }
   ],
   "source": [
    "d1 = io.load('../allData/MEM-A-dbsigwSpectator.h5')\n",
    "ix = [random.sample(range(4), 4) for _ in xrange(d['X'].shape[0])]\n",
    "#ix = np.argsort(d['X'][:, :4])[:, ::-1]\n",
    "\n",
    "def _pairwise(iterable):\n",
    "        '''s -> (s0, s1), (s2, s3), (s4, s5), ...'''\n",
    "        a = iter(iterable)\n",
    "        return izip(a, a)\n",
    "\n",
    "cols = [(0, 4), (8, 12), (15, 19), (23, 27)]\n",
    "for slc in cols:\n",
    "      print 'Shuffling columns {}'.format(slc)\n",
    "      d1['X'][:, slice(*slc)] = np.array([row[sh] for row, sh in zip(d1['X'][:, slice(*slc)], ix)])\n",
    "\n",
    "X_shuffled = d1['X']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([ 1.55747299e+02,  1.82756454e+02,  1.37238022e+02,  2.49927765e+02,\n",
       "        1.58137192e+02,  1.13366745e+02,  5.22283478e+01,  1.53766127e+01,\n",
       "        2.09246492e+00, -7.29564130e-01, -6.98315203e-02, -3.07966858e-01,\n",
       "        1.67793083e+00,  1.32498848e+00, -1.72354221e-01,  1.33465338e+00,\n",
       "       -5.22415042e-01,  2.67757154e+00, -2.96595240e+00,  1.16517329e+00,\n",
       "        6.76925600e-01, -3.00133228e-01,  1.82927525e+00,  6.40781178e+02,\n",
       "        2.33637023e+02,  1.37653036e+02,  2.61915932e+02,  4.38136898e+02,\n",
       "        2.28317680e+02,  5.30060171e+01])"
      ]
     },
     "execution_count": 13,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X_Big_train[1,:]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([ 2.49927765e+02,  1.82756454e+02,  1.55747299e+02,  1.37238022e+02,\n",
       "        1.58137192e+02,  1.13366745e+02,  5.22283478e+01,  1.53766127e+01,\n",
       "       -3.07966858e-01, -7.29564130e-01,  2.09246492e+00, -6.98315203e-02,\n",
       "        1.67793083e+00,  1.32498848e+00, -1.72354221e-01, -2.96595240e+00,\n",
       "       -5.22415042e-01,  1.33465338e+00,  2.67757154e+00,  1.16517329e+00,\n",
       "        6.76925600e-01, -3.00133228e-01,  1.82927525e+00,  2.61915932e+02,\n",
       "        2.33637023e+02,  6.40781178e+02,  1.37653036e+02,  4.38136898e+02,\n",
       "        2.28317680e+02,  5.30060171e+01])"
      ]
     },
     "execution_count": 14,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X_ordered[1,:]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([ 1.37238022e+02,  1.55747299e+02,  2.49927765e+02,  1.82756454e+02,\n",
       "        1.58137192e+02,  1.13366745e+02,  5.22283478e+01,  1.53766127e+01,\n",
       "       -6.98315203e-02,  2.09246492e+00, -3.07966858e-01, -7.29564130e-01,\n",
       "        1.67793083e+00,  1.32498848e+00, -1.72354221e-01,  2.67757154e+00,\n",
       "        1.33465338e+00, -2.96595240e+00, -5.22415042e-01,  1.16517329e+00,\n",
       "        6.76925600e-01, -3.00133228e-01,  1.82927525e+00,  1.37653036e+02,\n",
       "        6.40781178e+02,  2.61915932e+02,  2.33637023e+02,  4.38136898e+02,\n",
       "        2.28317680e+02,  5.30060171e+01])"
      ]
     },
     "execution_count": 15,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X_shuffled[1,:]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train on 3028985 samples, validate on 1009662 samples\n",
      "Epoch 1/100\n",
      "3028985/3028985 [==============================] - 122s 40us/step - loss: 0.1140 - val_loss: 0.2039\n",
      "Epoch 2/100\n",
      "3028985/3028985 [==============================] - 122s 40us/step - loss: 0.0722 - val_loss: 0.2165\n",
      "Epoch 00002: early stopping\n"
     ]
    }
   ],
   "source": [
    "sc_X = StandardScaler()\n",
    "X_train1 = sc_X.fit_transform(X_ordered)\n",
    "X_test1 = sc_X.fit_transform(X_Big_test)\n",
    "model = sequential_model()\n",
    "history = model.fit(X_train1, y_BigSig_train, validation_data=(X_test1, y_BigSig_test), epochs=100, batch_size=256, verbose=1, callbacks=[es])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1009662/1009662 [==============================] - 11s 11us/step\n"
     ]
    }
   ],
   "source": [
    "sig_predict_ordered = model.predict(X_test1,batch_size=256, verbose=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "sigSoft= sig_predict_ordered[np.where(y_BigSig_test==1)]\n",
    "bkgSoft = sig_predict_ordered[np.where(y_BigSig_test==0)]\n",
    "_=plt.hist(sigSoft,bins=100, histtype='step', label='signal', normed = 1)\n",
    "                           \n",
    "_=plt.hist(bkgSoft,bins=100, histtype='step', color= 'red', label='background', normed = 1)\n",
    "\n",
    "_=plt.yscale('log')\n",
    "\n",
    "_=plt.legend(loc='upper center')\n",
    "\n",
    "_=plt.title('Signal_LH_to_(Sig_LH+0.1*Bkg_LH)')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/debo/env5/lib/python2.7/site-packages/ipykernel_launcher.py:4: RuntimeWarning: divide by zero encountered in divide\n",
      "  after removing the cwd from sys.path.\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "fpr_ordered, tpr_ordered, thresholds_ordered = roc_curve(y_BigSig_test,sig_predict_ordered)\n",
    "plt.figure()\n",
    "##plt.plot(fpr, tpr)\n",
    "plt.plot(tpr_ordered, 1/fpr_ordered)\n",
    "plt.xlim([0.75, 1])\n",
    "#plt.ylim([1, 10000])\n",
    "plt.ylabel('1/Bkg_eff')\n",
    "plt.xlabel('Sig_eff')\n",
    "plt.title('Receiver operating characteristic')\n",
    "plt.yscale('log')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.plot(history.history['loss'], label='train')\n",
    "plt.plot(history.history['val_loss'], label='test')\n",
    "plt.legend()\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train on 3028985 samples, validate on 1009662 samples\n",
      "Epoch 1/100\n",
      "3028985/3028985 [==============================] - 125s 41us/step - loss: 0.1447 - val_loss: 0.3204\n",
      "Epoch 2/100\n",
      "3028985/3028985 [==============================] - 125s 41us/step - loss: 0.1042 - val_loss: 0.3357\n",
      "Epoch 00002: early stopping\n"
     ]
    }
   ],
   "source": [
    "sc_X = StandardScaler()\n",
    "X_train2 = sc_X.fit_transform(X_shuffled)\n",
    "X_test2 = sc_X.fit_transform(X_Big_test)\n",
    "model = sequential_model()\n",
    "history = model.fit(X_train2, y_BigSig_train, validation_data=(X_test2, y_BigSig_test), epochs=100, batch_size=256, verbose=1, callbacks=[es])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1009662/1009662 [==============================] - 12s 11us/step\n"
     ]
    }
   ],
   "source": [
    "sig_predict_shuffled = model.predict(X_test2,batch_size=256, verbose=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "sigSoft= sig_predict_shuffled[np.where(y_BigSig_test==1)]\n",
    "bkgSoft = sig_predict_shuffled[np.where(y_BigSig_test==0)]\n",
    "_=plt.hist(sigSoft,bins=100, histtype='step', label='signal', normed = 1)\n",
    "                           \n",
    "_=plt.hist(bkgSoft,bins=100, histtype='step', color= 'red', label='background', normed = 1)\n",
    "\n",
    "_=plt.yscale('log')\n",
    "\n",
    "_=plt.legend(loc='upper center')\n",
    "\n",
    "_=plt.title('Signal_LH_to_(Sig_LH+0.1*Bkg_LH)')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/debo/env5/lib/python2.7/site-packages/ipykernel_launcher.py:4: RuntimeWarning: divide by zero encountered in divide\n",
      "  after removing the cwd from sys.path.\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "fpr_shuffled, tpr_shuffled, thresholds_shuffled = roc_curve(y_BigSig_test,sig_predict_shuffled)\n",
    "plt.figure()\n",
    "##plt.plot(fpr, tpr)\n",
    "plt.plot(tpr_shuffled, 1/fpr_shuffled)\n",
    "plt.xlim([0.75, 1])\n",
    "#plt.ylim([1, 10000])\n",
    "plt.ylabel('1/Bkg_eff')\n",
    "plt.xlabel('Sig_eff')\n",
    "plt.title('Receiver operating characteristic')\n",
    "plt.yscale('log')\n",
    "\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "sigSoft= softTarget_test[np.where(y_BigSig_test==1)]\n",
    "bkgSoft = softTarget_test[np.where(y_BigSig_test==0)]\n",
    "_=plt.hist(sigSoft,bins=100, histtype='step', label='signal', normed = 1)\n",
    "                           \n",
    "_=plt.hist(bkgSoft,bins=100, histtype='step', color= 'red', label='background', normed = 1)\n",
    "\n",
    "_=plt.yscale('log')\n",
    "\n",
    "_=plt.legend(loc='upper center')\n",
    "\n",
    "_=plt.title('Signal_LH_to_(Sig_LH+0.1*Bkg_LH)')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/debo/env5/lib/python2.7/site-packages/ipykernel_launcher.py:4: RuntimeWarning: divide by zero encountered in divide\n",
      "  after removing the cwd from sys.path.\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "fpr_MEM, tpr_MEM, thresholds_MEM = roc_curve(y_BigSig_test,softTarget_test)\n",
    "plt.figure()\n",
    "##plt.plot(fpr, tpr)\n",
    "plt.plot(tpr_MEM, 1/fpr_MEM)\n",
    "plt.xlim([0.75, 1])\n",
    "#plt.ylim([1, 10000])\n",
    "plt.ylabel('1/Bkg_eff')\n",
    "plt.xlabel('Sig_eff')\n",
    "plt.title('Receiver operating characteristic')\n",
    "plt.yscale('log')\n",
    "\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/debo/env5/lib/python2.7/site-packages/ipykernel_launcher.py:2: RuntimeWarning: divide by zero encountered in divide\n",
      "  \n",
      "/home/debo/env5/lib/python2.7/site-packages/ipykernel_launcher.py:3: RuntimeWarning: divide by zero encountered in divide\n",
      "  This is separate from the ipykernel package so we can avoid doing imports until\n",
      "/home/debo/env5/lib/python2.7/site-packages/ipykernel_launcher.py:4: RuntimeWarning: divide by zero encountered in divide\n",
      "  after removing the cwd from sys.path.\n",
      "/home/debo/env5/lib/python2.7/site-packages/ipykernel_launcher.py:5: RuntimeWarning: divide by zero encountered in divide\n",
      "  \"\"\"\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.figure()\n",
    "plt.plot(tpr_MEM, 1/fpr_MEM, label='MEM',color='b')\n",
    "plt.plot(tpr, 1/fpr, label='pt not ordered',color='g')\n",
    "plt.plot(tpr_ordered, 1/fpr_ordered, label='pt ordered',color='r')\n",
    "plt.plot(tpr_shuffled, 1/fpr_shuffled, label='pt shuffled',color='c')\n",
    "\n",
    "plt.xlim([0.75, 1])\n",
    "#plt.ylim([1, 10000])\n",
    "plt.ylabel('1/Bkg_eff')\n",
    "plt.xlabel('Sig_eff')\n",
    "plt.title('Receiver operating characteristic')\n",
    "plt.yscale('log')\n",
    "plt.legend(loc='upper right')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.plot(history.history['loss'], label='train')\n",
    "plt.plot(history.history['val_loss'], label='test')\n",
    "plt.legend()\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "env5",
   "language": "python",
   "name": "env5"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 2
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython2",
   "version": "2.7.12"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}