Model_03-SGD-k=3_3e-4.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "#Mandatory imports\n",
    "import numpy as np\n",
    "import pandas as pd\n",
    "from sklearn.model_selection import train_test_split\n",
    "from sklearn.metrics import log_loss\n",
    "from sklearn.model_selection import StratifiedKFold, StratifiedShuffleSplit\n",
    "from os.path import join as opj\n",
    "from matplotlib import pyplot as plt\n",
    "from mpl_toolkits.mplot3d import Axes3D\n",
    "import pylab\n",
    "plt.rcParams['figure.figsize'] = 10, 10\n",
    "%matplotlib inline"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "train = pd.read_json(\"input/train.json\")\n",
    "target_train=train['is_iceberg']\n",
    "test = pd.read_json(\"input/test.json\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "target_train=train['is_iceberg']\n",
    "test['inc_angle']=pd.to_numeric(test['inc_angle'], errors='coerce')\n",
    "train['inc_angle']=pd.to_numeric(train['inc_angle'], errors='coerce')#We have only 133 NAs.\n",
    "train['inc_angle']=train['inc_angle'].fillna(method='pad')\n",
    "X_angle=train['inc_angle']\n",
    "test['inc_angle']=pd.to_numeric(test['inc_angle'], errors='coerce')\n",
    "X_test_angle=test['inc_angle']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "#Generate the training data\n",
    "X_band_1=np.array([np.array(band).astype(np.float32).reshape(75, 75) for band in train[\"band_1\"]])\n",
    "X_band_2=np.array([np.array(band).astype(np.float32).reshape(75, 75) for band in train[\"band_2\"]])\n",
    "X_band_3=(X_band_1+X_band_2)/2\n",
    "#X_band_3=np.array([np.full((75, 75), angel).astype(np.float32) for angel in train[\"inc_angle\"]])\n",
    "X_train = np.concatenate([X_band_1[:, :, :, np.newaxis]\n",
    "                          , X_band_2[:, :, :, np.newaxis]\n",
    "                         , X_band_3[:, :, :, np.newaxis]], axis=-1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "X_band_test_1=np.array([np.array(band).astype(np.float32).reshape(75, 75) for band in test[\"band_1\"]])\n",
    "X_band_test_2=np.array([np.array(band).astype(np.float32).reshape(75, 75) for band in test[\"band_2\"]])\n",
    "X_band_test_3=(X_band_test_1+X_band_test_2)/2\n",
    "#X_band_test_3=np.array([np.full((75, 75), angel).astype(np.float32) for angel in test[\"inc_angle\"]])\n",
    "X_test = np.concatenate([X_band_test_1[:, :, :, np.newaxis]\n",
    "                          , X_band_test_2[:, :, :, np.newaxis]\n",
    "                         , X_band_test_3[:, :, :, np.newaxis]], axis=-1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Using TensorFlow backend.\n"
     ]
    }
   ],
   "source": [
    "#Import Keras.\n",
    "from matplotlib import pyplot\n",
    "from keras.optimizers import RMSprop\n",
    "# from keras.preprocessing.image import ImageDataGenerator\n",
    "from keras.models import Sequential\n",
    "from keras.layers import Conv2D, MaxPooling2D, Dense, Dropout, Input, Flatten, Activation\n",
    "from keras.layers import GlobalMaxPooling2D\n",
    "from keras.layers.normalization import BatchNormalization\n",
    "from keras.layers.merge import Concatenate\n",
    "from keras.models import Model\n",
    "from keras import initializers\n",
    "from keras.optimizers import Adam\n",
    "from keras.optimizers import rmsprop\n",
    "from keras.optimizers import SGD\n",
    "from keras.optimizers import Nadam\n",
    "from keras.optimizers import Adadelta\n",
    "from keras.layers.advanced_activations import LeakyReLU, PReLU\n",
    "from keras.callbacks import ModelCheckpoint, Callback, EarlyStopping\n",
    "\n",
    "from keras.datasets import cifar10\n",
    "from keras.applications.inception_v3 import InceptionV3\n",
    "from keras.applications.vgg16 import VGG16\n",
    "from keras.applications.xception import Xception\n",
    "from keras.applications.mobilenet import MobileNet\n",
    "from keras.applications.vgg19 import VGG19\n",
    "from keras.layers import Concatenate, Dense, LSTM, Input, concatenate\n",
    "from keras.preprocessing import image\n",
    "from keras.applications.vgg16 import preprocess_input"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [],
   "source": [
    "#Data Aug for multi-input\n",
    "from keras.preprocessing.image import ImageDataGenerator\n",
    "batch_size=64\n",
    "# Define the image transformations here\n",
    "gen = ImageDataGenerator(horizontal_flip = True,\n",
    "                         vertical_flip = True,\n",
    "                         width_shift_range = 0.,\n",
    "                         height_shift_range = 0.,\n",
    "                         channel_shift_range=0,\n",
    "                         zoom_range = 0.2,\n",
    "                         rotation_range = 10)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Here is the function that merges our two generators\n",
    "# We use the exact same generator with the same random seed for both the y and angle arrays\n",
    "def gen_flow_for_two_inputs(X1, X2, y):\n",
    "    genX1 = gen.flow(X1,y,  batch_size=batch_size,seed=55)\n",
    "    genX2 = gen.flow(X1,X2, batch_size=batch_size,seed=55)\n",
    "    while True:\n",
    "            X1i = genX1.next()\n",
    "            X2i = genX2.next()\n",
    "            #Assert arrays are equal - this was for peace of mind, but slows down training\n",
    "            #np.testing.assert_array_equal(X1i[0],X2i[0])\n",
    "            yield [X1i[0], X2i[1]], X1i[1]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Finally create generator\n",
    "def get_callbacks(filepath, patience=2):\n",
    "    es = EarlyStopping('val_loss', patience=10, mode=\"min\")\n",
    "    msave = ModelCheckpoint(filepath, save_best_only=True)\n",
    "    return [es, msave]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [],
   "source": [
    "def getVggAngleModel():\n",
    "    input_2 = Input(shape=[1], name=\"angle\")\n",
    "    angle_layer = Dense(1, )(input_2)\n",
    "    base_model = VGG16(weights='imagenet', include_top=False, \n",
    "                 input_shape=X_train.shape[1:], classes=1)\n",
    "    x = base_model.get_layer('block5_pool').output\n",
    "    \n",
    "\n",
    "    x = GlobalMaxPooling2D()(x)\n",
    "    merge_one = concatenate([x, angle_layer])\n",
    "    merge_one = Dense(512, activation='relu', name='fc2')(merge_one)\n",
    "    merge_one = Dropout(0.3)(merge_one)\n",
    "    merge_one = Dense(512, activation='relu', name='fc3')(merge_one)\n",
    "    merge_one = Dropout(0.3)(merge_one)\n",
    "    \n",
    "    predictions = Dense(1, activation='sigmoid')(merge_one)\n",
    "    \n",
    "    model = Model(inputs=[base_model.input, input_2], outputs=predictions)\n",
    "    \n",
    "#     adadelta= Adadelta(lr=1.0, rho=0.95, epsilon=None, decay=0.0)\n",
    "#     model.compile(loss='binary_crossentropy',\n",
    "#                   optimizer=adadelta,\n",
    "#                   metrics=['accuracy'])\n",
    "    \n",
    "    sgd = SGD(lr=3e-4, decay=1e-6, momentum=0.9, nesterov=True)\n",
    "    model.compile(loss='binary_crossentropy',\n",
    "                  optimizer=sgd,\n",
    "                  metrics=['accuracy'])\n",
    "    return model\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {
    "scrolled": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "===================FOLD= 0\n",
      "Epoch 1/50\n",
      "24/24 [==============================] - 19s 777ms/step - loss: 0.7456 - acc: 0.5755 - val_loss: 0.4876 - val_acc: 0.7346\n",
      "Epoch 2/50\n",
      "24/24 [==============================] - 13s 543ms/step - loss: 0.5228 - acc: 0.7395 - val_loss: 0.3300 - val_acc: 0.8467\n",
      "Epoch 3/50\n",
      "24/24 [==============================] - 13s 542ms/step - loss: 0.4043 - acc: 0.8125 - val_loss: 0.2974 - val_acc: 0.8449\n",
      "Epoch 4/50\n",
      "24/24 [==============================] - 13s 550ms/step - loss: 0.3690 - acc: 0.8358 - val_loss: 0.2770 - val_acc: 0.8729\n",
      "Epoch 5/50\n",
      "24/24 [==============================] - 13s 548ms/step - loss: 0.3256 - acc: 0.8566 - val_loss: 0.2599 - val_acc: 0.8710\n",
      "Epoch 6/50\n",
      "24/24 [==============================] - 13s 555ms/step - loss: 0.3398 - acc: 0.8547 - val_loss: 0.2529 - val_acc: 0.8916\n",
      "Epoch 7/50\n",
      "24/24 [==============================] - 13s 555ms/step - loss: 0.3036 - acc: 0.8622 - val_loss: 0.2419 - val_acc: 0.8935\n",
      "Epoch 8/50\n",
      "24/24 [==============================] - 13s 553ms/step - loss: 0.2830 - acc: 0.8751 - val_loss: 0.2385 - val_acc: 0.8916\n",
      "Epoch 9/50\n",
      "24/24 [==============================] - 13s 558ms/step - loss: 0.2592 - acc: 0.8826 - val_loss: 0.2327 - val_acc: 0.9009\n",
      "Epoch 10/50\n",
      "24/24 [==============================] - 13s 531ms/step - loss: 0.2701 - acc: 0.8823 - val_loss: 0.2507 - val_acc: 0.8822\n",
      "Epoch 11/50\n",
      "24/24 [==============================] - 13s 556ms/step - loss: 0.2554 - acc: 0.8847 - val_loss: 0.2235 - val_acc: 0.8972\n",
      "Epoch 12/50\n",
      "24/24 [==============================] - 13s 556ms/step - loss: 0.2544 - acc: 0.8850 - val_loss: 0.2186 - val_acc: 0.9065\n",
      "Epoch 13/50\n",
      "24/24 [==============================] - 13s 551ms/step - loss: 0.2380 - acc: 0.8901 - val_loss: 0.2118 - val_acc: 0.9103\n",
      "Epoch 14/50\n",
      "24/24 [==============================] - 13s 557ms/step - loss: 0.2306 - acc: 0.9136 - val_loss: 0.2074 - val_acc: 0.9178\n",
      "Epoch 15/50\n",
      "24/24 [==============================] - 13s 530ms/step - loss: 0.2292 - acc: 0.9023 - val_loss: 0.2243 - val_acc: 0.9009\n",
      "Epoch 16/50\n",
      "24/24 [==============================] - 13s 537ms/step - loss: 0.2179 - acc: 0.9076 - val_loss: 0.2281 - val_acc: 0.9047\n",
      "Epoch 17/50\n",
      "24/24 [==============================] - 14s 565ms/step - loss: 0.2218 - acc: 0.9073 - val_loss: 0.2138 - val_acc: 0.9121\n",
      "Epoch 18/50\n",
      "24/24 [==============================] - 13s 555ms/step - loss: 0.2214 - acc: 0.8973 - val_loss: 0.1955 - val_acc: 0.9252\n",
      "Epoch 19/50\n",
      "24/24 [==============================] - 13s 536ms/step - loss: 0.2050 - acc: 0.9100 - val_loss: 0.2092 - val_acc: 0.9065\n",
      "Epoch 20/50\n",
      "24/24 [==============================] - 13s 550ms/step - loss: 0.2120 - acc: 0.9141 - val_loss: 0.1936 - val_acc: 0.9196\n",
      "Epoch 21/50\n",
      "24/24 [==============================] - 13s 536ms/step - loss: 0.1711 - acc: 0.9261 - val_loss: 0.2074 - val_acc: 0.9215\n",
      "Epoch 22/50\n",
      "24/24 [==============================] - 13s 532ms/step - loss: 0.1848 - acc: 0.9228 - val_loss: 0.2057 - val_acc: 0.9196\n",
      "Epoch 23/50\n",
      "24/24 [==============================] - 13s 535ms/step - loss: 0.1909 - acc: 0.9240 - val_loss: 0.2136 - val_acc: 0.9140\n",
      "Epoch 24/50\n",
      "24/24 [==============================] - 13s 536ms/step - loss: 0.1747 - acc: 0.9274 - val_loss: 0.2161 - val_acc: 0.9121\n",
      "Epoch 25/50\n",
      "24/24 [==============================] - 13s 530ms/step - loss: 0.1931 - acc: 0.9196 - val_loss: 0.2238 - val_acc: 0.9234\n",
      "Epoch 26/50\n",
      "24/24 [==============================] - 13s 535ms/step - loss: 0.1780 - acc: 0.9223 - val_loss: 0.2109 - val_acc: 0.9252\n",
      "Epoch 27/50\n",
      "24/24 [==============================] - 13s 530ms/step - loss: 0.1613 - acc: 0.9355 - val_loss: 0.2065 - val_acc: 0.9178\n",
      "Epoch 28/50\n",
      "24/24 [==============================] - 13s 535ms/step - loss: 0.1605 - acc: 0.9315 - val_loss: 0.2127 - val_acc: 0.9234\n",
      "Epoch 29/50\n",
      "24/24 [==============================] - 13s 536ms/step - loss: 0.1830 - acc: 0.9218 - val_loss: 0.2082 - val_acc: 0.9196\n",
      "Epoch 30/50\n",
      "24/24 [==============================] - 13s 554ms/step - loss: 0.1640 - acc: 0.9310 - val_loss: 0.1919 - val_acc: 0.9271\n",
      "Epoch 31/50\n",
      "24/24 [==============================] - 13s 534ms/step - loss: 0.1618 - acc: 0.9261 - val_loss: 0.1993 - val_acc: 0.9196\n",
      "Epoch 32/50\n",
      "24/24 [==============================] - 13s 531ms/step - loss: 0.1646 - acc: 0.9329 - val_loss: 0.2224 - val_acc: 0.9234\n",
      "Epoch 33/50\n",
      "24/24 [==============================] - 13s 535ms/step - loss: 0.1610 - acc: 0.9296 - val_loss: 0.2100 - val_acc: 0.9346\n",
      "Epoch 34/50\n",
      "24/24 [==============================] - 13s 535ms/step - loss: 0.1509 - acc: 0.9411 - val_loss: 0.2186 - val_acc: 0.9215\n",
      "Epoch 35/50\n",
      "24/24 [==============================] - 13s 535ms/step - loss: 0.1516 - acc: 0.9316 - val_loss: 0.2015 - val_acc: 0.9252\n",
      "Epoch 36/50\n",
      "24/24 [==============================] - 13s 535ms/step - loss: 0.1529 - acc: 0.9441 - val_loss: 0.2333 - val_acc: 0.9084\n",
      "Epoch 37/50\n",
      "24/24 [==============================] - 13s 530ms/step - loss: 0.1592 - acc: 0.9398 - val_loss: 0.2021 - val_acc: 0.9215\n",
      "Epoch 38/50\n",
      "24/24 [==============================] - 13s 536ms/step - loss: 0.1354 - acc: 0.9432 - val_loss: 0.2286 - val_acc: 0.9215\n",
      "Epoch 39/50\n",
      "24/24 [==============================] - 13s 530ms/step - loss: 0.1569 - acc: 0.9338 - val_loss: 0.2285 - val_acc: 0.9196\n",
      "Epoch 40/50\n",
      "24/24 [==============================] - 13s 536ms/step - loss: 0.1354 - acc: 0.9448 - val_loss: 0.2479 - val_acc: 0.9215\n",
      "Train loss: 0.106742089588\n",
      "Train accuracy: 0.957904584559\n",
      "Test loss: 0.191861295032\n",
      "Test accuracy: 0.927102805521\n",
      "\n",
      "===================FOLD= 1\n",
      "Epoch 1/50\n",
      "24/24 [==============================] - 14s 581ms/step - loss: 0.9238 - acc: 0.5517 - val_loss: 0.5406 - val_acc: 0.7121\n",
      "Epoch 2/50\n",
      "24/24 [==============================] - 13s 554ms/step - loss: 0.6088 - acc: 0.6883 - val_loss: 0.3454 - val_acc: 0.8262\n",
      "Epoch 3/50\n",
      "24/24 [==============================] - 13s 549ms/step - loss: 0.4405 - acc: 0.7935 - val_loss: 0.2829 - val_acc: 0.8542\n",
      "Epoch 4/50\n",
      "24/24 [==============================] - 13s 554ms/step - loss: 0.3912 - acc: 0.8129 - val_loss: 0.2460 - val_acc: 0.8991\n",
      "Epoch 5/50\n",
      "24/24 [==============================] - 13s 550ms/step - loss: 0.3362 - acc: 0.8540 - val_loss: 0.2199 - val_acc: 0.9047\n",
      "Epoch 6/50\n",
      "24/24 [==============================] - 13s 536ms/step - loss: 0.2925 - acc: 0.8655 - val_loss: 0.2226 - val_acc: 0.9159\n",
      "Epoch 7/50\n",
      "24/24 [==============================] - 13s 535ms/step - loss: 0.3344 - acc: 0.8435 - val_loss: 0.2331 - val_acc: 0.9178\n",
      "Epoch 8/50\n",
      "24/24 [==============================] - 13s 530ms/step - loss: 0.2825 - acc: 0.8664 - val_loss: 0.2298 - val_acc: 0.9159\n",
      "Epoch 9/50\n",
      "24/24 [==============================] - 13s 535ms/step - loss: 0.2549 - acc: 0.8762 - val_loss: 0.2280 - val_acc: 0.9159\n",
      "Epoch 10/50\n",
      "24/24 [==============================] - 13s 531ms/step - loss: 0.2782 - acc: 0.8778 - val_loss: 0.2297 - val_acc: 0.9196\n",
      "Epoch 11/50\n",
      "24/24 [==============================] - 13s 536ms/step - loss: 0.2747 - acc: 0.8736 - val_loss: 0.2267 - val_acc: 0.9196\n",
      "Epoch 12/50\n",
      "24/24 [==============================] - 13s 554ms/step - loss: 0.2581 - acc: 0.8851 - val_loss: 0.2140 - val_acc: 0.9271\n",
      "Epoch 13/50\n",
      "24/24 [==============================] - 13s 531ms/step - loss: 0.2266 - acc: 0.8973 - val_loss: 0.2641 - val_acc: 0.8972\n",
      "Epoch 14/50\n",
      "24/24 [==============================] - 13s 554ms/step - loss: 0.2520 - acc: 0.8895 - val_loss: 0.2074 - val_acc: 0.9234\n",
      "Epoch 15/50\n",
      "24/24 [==============================] - 13s 550ms/step - loss: 0.2670 - acc: 0.8883 - val_loss: 0.1979 - val_acc: 0.9346\n",
      "Epoch 16/50\n",
      "24/24 [==============================] - 13s 535ms/step - loss: 0.2097 - acc: 0.9100 - val_loss: 0.2031 - val_acc: 0.9290\n",
      "Epoch 17/50\n",
      "24/24 [==============================] - 13s 536ms/step - loss: 0.2220 - acc: 0.8988 - val_loss: 0.2110 - val_acc: 0.9271\n",
      "Epoch 18/50\n",
      "24/24 [==============================] - 13s 536ms/step - loss: 0.2118 - acc: 0.9147 - val_loss: 0.1993 - val_acc: 0.9308\n",
      "Epoch 19/50\n",
      "24/24 [==============================] - 13s 555ms/step - loss: 0.2208 - acc: 0.9114 - val_loss: 0.1922 - val_acc: 0.9290\n",
      "Epoch 20/50\n",
      "24/24 [==============================] - 13s 531ms/step - loss: 0.2081 - acc: 0.9102 - val_loss: 0.1980 - val_acc: 0.9271\n",
      "Epoch 21/50\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "24/24 [==============================] - 13s 536ms/step - loss: 0.2121 - acc: 0.9162 - val_loss: 0.2002 - val_acc: 0.9364\n",
      "Epoch 22/50\n",
      "24/24 [==============================] - 13s 530ms/step - loss: 0.2115 - acc: 0.9116 - val_loss: 0.2211 - val_acc: 0.9215\n",
      "Epoch 23/50\n",
      "24/24 [==============================] - 13s 536ms/step - loss: 0.1993 - acc: 0.9118 - val_loss: 0.2082 - val_acc: 0.9252\n",
      "Epoch 24/50\n",
      "24/24 [==============================] - 13s 534ms/step - loss: 0.1750 - acc: 0.9223 - val_loss: 0.1954 - val_acc: 0.9308\n",
      "Epoch 25/50\n",
      "24/24 [==============================] - 13s 531ms/step - loss: 0.1912 - acc: 0.9190 - val_loss: 0.1925 - val_acc: 0.9271\n",
      "Epoch 26/50\n",
      "24/24 [==============================] - 13s 534ms/step - loss: 0.1831 - acc: 0.9214 - val_loss: 0.2465 - val_acc: 0.9215\n",
      "Epoch 27/50\n",
      "24/24 [==============================] - 13s 530ms/step - loss: 0.2070 - acc: 0.9099 - val_loss: 0.1936 - val_acc: 0.9327\n",
      "Epoch 28/50\n",
      "24/24 [==============================] - 13s 535ms/step - loss: 0.1700 - acc: 0.9268 - val_loss: 0.2149 - val_acc: 0.9327\n",
      "Epoch 29/50\n",
      "24/24 [==============================] - 13s 535ms/step - loss: 0.1736 - acc: 0.9265 - val_loss: 0.2282 - val_acc: 0.9215\n",
      "Train loss: 0.145599535573\n",
      "Train accuracy: 0.935453695823\n",
      "Test loss: 0.192168537228\n",
      "Test accuracy: 0.928971963508\n",
      "\n",
      "===================FOLD= 2\n",
      "Epoch 1/50\n",
      "24/24 [==============================] - 16s 684ms/step - loss: 0.7796 - acc: 0.5372 - val_loss: 0.5543 - val_acc: 0.7566\n",
      "Epoch 2/50\n",
      "24/24 [==============================] - 13s 553ms/step - loss: 0.5134 - acc: 0.7185 - val_loss: 0.4262 - val_acc: 0.7584\n",
      "Epoch 3/50\n",
      "24/24 [==============================] - 13s 549ms/step - loss: 0.4004 - acc: 0.8095 - val_loss: 0.3379 - val_acc: 0.8427\n",
      "Epoch 4/50\n",
      "24/24 [==============================] - 13s 555ms/step - loss: 0.3386 - acc: 0.8425 - val_loss: 0.3351 - val_acc: 0.8408\n",
      "Epoch 5/50\n",
      "24/24 [==============================] - 13s 549ms/step - loss: 0.3128 - acc: 0.8560 - val_loss: 0.3075 - val_acc: 0.8652\n",
      "Epoch 6/50\n",
      "24/24 [==============================] - 13s 536ms/step - loss: 0.2887 - acc: 0.8795 - val_loss: 0.3300 - val_acc: 0.8521\n",
      "Epoch 7/50\n",
      "24/24 [==============================] - 13s 537ms/step - loss: 0.2687 - acc: 0.8844 - val_loss: 0.3326 - val_acc: 0.8596\n",
      "Epoch 8/50\n",
      "24/24 [==============================] - 13s 550ms/step - loss: 0.2400 - acc: 0.8945 - val_loss: 0.3060 - val_acc: 0.8689\n",
      "Epoch 9/50\n",
      "24/24 [==============================] - 13s 554ms/step - loss: 0.2528 - acc: 0.8885 - val_loss: 0.2915 - val_acc: 0.8708\n",
      "Epoch 10/50\n",
      "24/24 [==============================] - 13s 549ms/step - loss: 0.2299 - acc: 0.8943 - val_loss: 0.2825 - val_acc: 0.8764\n",
      "Epoch 11/50\n",
      "24/24 [==============================] - 13s 534ms/step - loss: 0.2218 - acc: 0.9043 - val_loss: 0.2870 - val_acc: 0.8727\n",
      "Epoch 12/50\n",
      "24/24 [==============================] - 13s 535ms/step - loss: 0.2155 - acc: 0.9073 - val_loss: 0.2993 - val_acc: 0.8689\n",
      "Epoch 13/50\n",
      "24/24 [==============================] - 13s 532ms/step - loss: 0.2250 - acc: 0.9073 - val_loss: 0.2892 - val_acc: 0.8727\n",
      "Epoch 14/50\n",
      "24/24 [==============================] - 13s 535ms/step - loss: 0.1911 - acc: 0.9281 - val_loss: 0.2848 - val_acc: 0.8727\n",
      "Epoch 15/50\n",
      "24/24 [==============================] - 13s 533ms/step - loss: 0.1937 - acc: 0.9178 - val_loss: 0.2852 - val_acc: 0.8689\n",
      "Epoch 16/50\n",
      "24/24 [==============================] - 13s 555ms/step - loss: 0.1932 - acc: 0.9183 - val_loss: 0.2657 - val_acc: 0.8839\n",
      "Epoch 17/50\n",
      "24/24 [==============================] - 13s 556ms/step - loss: 0.1788 - acc: 0.9273 - val_loss: 0.2999 - val_acc: 0.8764\n",
      "Epoch 18/50\n",
      "24/24 [==============================] - 13s 535ms/step - loss: 0.1884 - acc: 0.9284 - val_loss: 0.2740 - val_acc: 0.8858\n",
      "Epoch 19/50\n",
      "24/24 [==============================] - 13s 536ms/step - loss: 0.1688 - acc: 0.9391 - val_loss: 0.2859 - val_acc: 0.8783\n",
      "Epoch 20/50\n",
      "24/24 [==============================] - 13s 531ms/step - loss: 0.1779 - acc: 0.9335 - val_loss: 0.2694 - val_acc: 0.8914\n",
      "Epoch 21/50\n",
      "24/24 [==============================] - 13s 535ms/step - loss: 0.1903 - acc: 0.9233 - val_loss: 0.2952 - val_acc: 0.8858\n",
      "Epoch 22/50\n",
      "24/24 [==============================] - 13s 531ms/step - loss: 0.1630 - acc: 0.9360 - val_loss: 0.2949 - val_acc: 0.8652\n",
      "Epoch 23/50\n",
      "24/24 [==============================] - 13s 554ms/step - loss: 0.1618 - acc: 0.9402 - val_loss: 0.2615 - val_acc: 0.8820\n",
      "Epoch 24/50\n",
      "24/24 [==============================] - 13s 535ms/step - loss: 0.1750 - acc: 0.9399 - val_loss: 0.2644 - val_acc: 0.8914\n",
      "Epoch 25/50\n",
      "24/24 [==============================] - 13s 548ms/step - loss: 0.1776 - acc: 0.9285 - val_loss: 0.2554 - val_acc: 0.8951\n",
      "Epoch 26/50\n",
      "24/24 [==============================] - 13s 534ms/step - loss: 0.1617 - acc: 0.9392 - val_loss: 0.2589 - val_acc: 0.8951\n",
      "Epoch 27/50\n",
      "24/24 [==============================] - 13s 530ms/step - loss: 0.1504 - acc: 0.9429 - val_loss: 0.2718 - val_acc: 0.8876\n",
      "Epoch 28/50\n",
      "24/24 [==============================] - 13s 554ms/step - loss: 0.1818 - acc: 0.9271 - val_loss: 0.2473 - val_acc: 0.8933\n",
      "Epoch 29/50\n",
      "24/24 [==============================] - 13s 535ms/step - loss: 0.1405 - acc: 0.9444 - val_loss: 0.2652 - val_acc: 0.8876\n",
      "Epoch 30/50\n",
      "24/24 [==============================] - 13s 530ms/step - loss: 0.1455 - acc: 0.9423 - val_loss: 0.2620 - val_acc: 0.8951\n",
      "Epoch 31/50\n",
      "24/24 [==============================] - 13s 535ms/step - loss: 0.1303 - acc: 0.9525 - val_loss: 0.2768 - val_acc: 0.8914\n",
      "Epoch 32/50\n",
      "24/24 [==============================] - 13s 531ms/step - loss: 0.1492 - acc: 0.9449 - val_loss: 0.2948 - val_acc: 0.8933\n",
      "Epoch 33/50\n",
      "24/24 [==============================] - 13s 534ms/step - loss: 0.1410 - acc: 0.9511 - val_loss: 0.2630 - val_acc: 0.9026\n",
      "Epoch 34/50\n",
      "24/24 [==============================] - 13s 534ms/step - loss: 0.1440 - acc: 0.9482 - val_loss: 0.2896 - val_acc: 0.8858\n",
      "Epoch 35/50\n",
      "24/24 [==============================] - 13s 535ms/step - loss: 0.1357 - acc: 0.9500 - val_loss: 0.2641 - val_acc: 0.8970\n",
      "Epoch 36/50\n",
      "24/24 [==============================] - 13s 535ms/step - loss: 0.1320 - acc: 0.9525 - val_loss: 0.2937 - val_acc: 0.8951\n",
      "Epoch 37/50\n",
      "24/24 [==============================] - 13s 531ms/step - loss: 0.1404 - acc: 0.9531 - val_loss: 0.2839 - val_acc: 0.8839\n",
      "Epoch 38/50\n",
      "24/24 [==============================] - 13s 534ms/step - loss: 0.1356 - acc: 0.9454 - val_loss: 0.2810 - val_acc: 0.8933\n",
      "Train loss: 0.124804835286\n",
      "Train accuracy: 0.959813084112\n",
      "Test loss: 0.247270283181\n",
      "Test accuracy: 0.893258427413\n",
      "\n",
      " Train Log Loss Validation=  0.137365193607\n",
      " Test Log Loss Validation=  0.210410405777\n"
     ]
    }
   ],
   "source": [
    "#def myAngleCV(X_train, X_angle, X_test):\n",
    "K=3\n",
    "folds = list(StratifiedKFold(n_splits=K, shuffle=True, random_state=16).split(X_train, target_train))\n",
    "y_test_pred_log = 0\n",
    "y_train_pred_log=0\n",
    "y_valid_pred_log = 0.0*target_train\n",
    "for j, (train_idx, test_idx) in enumerate(folds):\n",
    "    print('\\n===================FOLD=',j)\n",
    "    X_train_cv = X_train[train_idx]\n",
    "    y_train_cv = target_train[train_idx]\n",
    "    X_holdout = X_train[test_idx]\n",
    "    Y_holdout= target_train[test_idx]\n",
    "    \n",
    "    #Angle\n",
    "    X_angle_cv=X_angle[train_idx]\n",
    "    X_angle_hold=X_angle[test_idx]\n",
    "    #define file path and get callbacks\n",
    "    file_path = \"%s_aug_model_weights.hdf5\"%j\n",
    "    callbacks = get_callbacks(filepath=file_path, patience=5)\n",
    "    gen_flow = gen_flow_for_two_inputs(X_train_cv, X_angle_cv, y_train_cv)\n",
    "    galaxyModel= getVggAngleModel()\n",
    "    histr = galaxyModel.fit_generator(\n",
    "            gen_flow,\n",
    "            steps_per_epoch=24,\n",
    "            epochs=50,\n",
    "            shuffle=True,\n",
    "            verbose=1,\n",
    "            validation_data=([X_holdout,X_angle_hold], Y_holdout),\n",
    "            callbacks=callbacks)\n",
    "     #Getting the Best Model\n",
    "    galaxyModel.load_weights(filepath=file_path)\n",
    "    #Getting Training Score\n",
    "    score = galaxyModel.evaluate([X_train_cv,X_angle_cv], y_train_cv, verbose=0)\n",
    "    print('Train loss:', score[0])\n",
    "    print('Train accuracy:', score[1])\n",
    "    #Getting Test Score\n",
    "    score = galaxyModel.evaluate([X_holdout,X_angle_hold], Y_holdout, verbose=0)\n",
    "    print('Test loss:', score[0])\n",
    "    print('Test accuracy:', score[1])\n",
    "\n",
    "    #Getting validation Score.\n",
    "    pred_valid=galaxyModel.predict([X_holdout,X_angle_hold])\n",
    "    y_valid_pred_log[test_idx] = pred_valid.reshape(pred_valid.shape[0])\n",
    "\n",
    "    #Getting Test Scores\n",
    "    temp_test=galaxyModel.predict([X_test, X_test_angle])\n",
    "    y_test_pred_log+=temp_test.reshape(temp_test.shape[0])\n",
    "\n",
    "    #Getting Train Scores\n",
    "    temp_train=galaxyModel.predict([X_train, X_angle])\n",
    "    y_train_pred_log+=temp_train.reshape(temp_train.shape[0])\n",
    "\n",
    "y_test_pred_log=y_test_pred_log/K\n",
    "y_train_pred_log=y_train_pred_log/K\n",
    "\n",
    "print('\\n Train Log Loss Validation= ',log_loss(target_train, y_train_pred_log))\n",
    "print(' Test Log Loss Validation= ',log_loss(target_train, y_valid_pred_log))\n",
    "#return y_test_pred_log"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [],
   "source": [
    "#preds=myAngleCV(X_train, X_angle, X_test)\n",
    "preds = y_test_pred_log"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Submission to kaggle\n",
    "submission = pd.DataFrame()\n",
    "submission['id']=test['id']\n",
    "submission['is_iceberg']=preds\n",
    "submission.to_csv('sub_SGD_3folds_3e-4.csv', index=False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "dict_keys(['acc', 'val_acc', 'loss', 'val_loss'])\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYUAAAEWCAYAAACJ0YulAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz\nAAALEgAACxIB0t1+/AAAIABJREFUeJzt3Xd8leXZwPHflb0TSFhJWLJkKTKddSs4UGu1DhxtFW1r\n1db6qq1atfWtb2ut1Tpa6164EZUqoKIiyt5hiowkBEIgk+xc7x/3STiEkBwgJ+ck5/p+PvnkjOec\nc50H8lzPc4/rFlXFGGOMAQgLdADGGGOChyUFY4wxDSwpGGOMaWBJwRhjTANLCsYYYxpYUjDGGNPA\nkoIJKSLygoj8ycdtN4nIGf6OyZhgYknBGGNMA0sKxrRDIhIR6BhMx2RJwQQdT7PN7SKyXETKRORZ\nEekmIv8VkRIRmSUinby2nygiq0SkUERmi8hgr+eOEZHFnte9AcQ0+qzzRGSp57VzReQoH2M8V0SW\niEixiGwVkfsaPX+i5/0KPc9f63k8VkT+JiKbRaRIROZ4HjtFRLKb2A9neG7fJyJvi8grIlIMXCsi\nY0XkG89nbBORf4pIlNfrh4rITBHZJSLbReR3ItJdRPaISKrXdiNFJF9EIn357qZjs6RggtXFwJnA\nQOB84L/A74AuuP+3NwOIyEDgdeBWz3PTgQ9EJMpzgJwKvAx0Bt7yvC+e1x4DPAfcAKQC/wKmiUi0\nD/GVAVcDKcC5wM9F5ELP+/b2xPu4J6YRwFLP6x4GRgHHe2L6H6DOx31yAfC25zNfBWqBXwNpwHHA\n6cAvPDEkArOAj4F0oD/wqarmAbOBS73e9ypgiqpW+xiH6cAsKZhg9biqblfVHOArYJ6qLlHVCuA9\n4BjPdj8GPlLVmZ6D2sNALO6geywQCTyqqtWq+jawwOszJgP/UtV5qlqrqi8ClZ7XNUtVZ6vqClWt\nU9XluMR0sufpK4BZqvq653MLVHWpiIQBPwVuUdUcz2fOVdVKH/fJN6o61fOZ5aq6SFW/VdUaVd2E\nS2r1MZwH5Knq31S1QlVLVHWe57kXgUkAIhIOXI5LnMZYUjBBa7vX7fIm7id4bqcDm+ufUNU6YCuQ\n4XkuR/et+rjZ63Zv4DZP80uhiBQCPT2va5aIjBORzz3NLkXAjbgzdjzv8V0TL0vDNV819ZwvtjaK\nYaCIfCgieZ4mpf/1IQaA94EhItIXdzVWpKrzDzEm08FYUjDtXS7u4A6AiAjugJgDbAMyPI/V6+V1\neyvwoKqmeP3EqerrPnzua8A0oKeqJgNPA/WfsxXo18RrdgIVB3iuDIjz+h7huKYnb41LGj8FrAEG\nqGoSrnnNO4Yjmgrcc7X1Ju5q4SrsKsF4saRg2rs3gXNF5HRPR+ltuCagucA3QA1ws4hEisgPgbFe\nr30GuNFz1i8iEu/pQE704XMTgV2qWiEiY3FNRvVeBc4QkUtFJEJEUkVkhOcq5jngERFJF5FwETnO\n04exDojxfH4kcDfQUt9GIlAMlIrIkcDPvZ77EOghIreKSLSIJIrIOK/nXwKuBSZiScF4saRg2jVV\nXYs7430cdyZ+PnC+qlapahXwQ9zBbxeu/+Fdr9cuBK4H/gnsBjZ4tvXFL4AHRKQEuBeXnOrfdwtw\nDi5B7cJ1Mh/tefq3wApc38Yu4P+AMFUt8rznf3BXOWXAPqORmvBbXDIqwSW4N7xiKME1DZ0P5AHr\ngVO9nv8a18G9WFW9m9RMiBNbZMeY0CQinwGvqep/Ah2LCR6WFIwJQSIyBpiJ6xMpCXQ8JnhY85Ex\nIUZEXsTNYbjVEoJpzK4UjDHGNLArBWOMMQ3aXVGttLQ07dOnT6DDMMaYdmXRokU7VbXx3Jf9tLuk\n0KdPHxYuXBjoMIwxpl0REZ+GHlvzkTHGmAaWFIwxxjSwpGCMMaZBu+tTaEp1dTXZ2dlUVFQEOhS/\niomJITMzk8hIWwvFGOMfHSIpZGdnk5iYSJ8+fdi3IGbHoaoUFBSQnZ1N3759Ax2OMaaD6hDNRxUV\nFaSmpnbYhAAgIqSmpnb4qyFjTGB1iKQAdOiEUC8UvqMxJrA6RPORMca0N3uqasjZXU5OofvZU1nL\nBSPS6ZoUE9C4LCm0gsLCQl577TV+8YtfHNTrzjnnHF577TVSUlL8FJkxHZuq8sW6fHaWVjFhWHfi\now/+kFZXp3xfUEb3pJhDen1TauuUnaWV5BSWs62wgm1F7sCf60kAObvL2b2ner/XPTxjLVeO682N\nJx8RsORgSaEVFBYW8uSTT+6XFGpqaoiIOPAunj59ur9DM6bDWpNXzB8/zOLrDQUA3D9tFRePymTS\nsb3p3zWhhVdD9u49vLMoh7cXb2XrrnJEoG9qPIN7JDEkPYkhnt9dE6P3abqtq1MKyqrIK6ogt6i8\n4Xf9wT+3sILtxRXU1O1bbDQ2MpyMTrFkpMRyVGYKGSmxZHrup6fEUllTx5Ofb+DFbzbx6rzNXDGu\nFz8/uV+bJ4d2VyV19OjR2rjMxerVqxk8eHCAIoLLLruM999/n0GDBhEZGUlMTAydOnVizZo1rFu3\njgsvvJCtW7dSUVHBLbfcwuTJk4G9JTtKS0uZMGECJ554InPnziUjI4P333+f2NjY/T4r0N/VmEDb\nWVrJIzPXMWX+FpJiI/n1GQMZmp7EK99uZvqKPKpq6zihfypXHduHMwZ3JSJ8b9dpRXUtH6/M461F\nW5n7XQGqcEL/VM4Z3oOC0iqycovJ2lbMll17Gl6TGh/FkT0Sqa5VthWVs72okqraun1iigoPo3ty\nDD2SY0hPiaVHcgw9UmJJT46hR3Is6SkxJMdG+tQvuLmgjH9+toF3l+QQESatlhxEZJGqjm5xu46W\nFO7/YBVZucWt+plD0pP4w/lDD/j8pk2bOO+881i5ciWzZ8/m3HPPZeXKlQ1DR3ft2kXnzp0pLy9n\nzJgxfPHFF6Smpu6TFPr378/ChQsZMWIEl156KRMnTmTSpEn7fZYlBROqqmrqeHHuJh77dD3l1bVc\ndVxvbjl9AClxUQ3b7Cyt5I0FW3n1283kFlWQnhzDFeN6MbJ3Jz5cvo0PluZSUllDZqdYLhnVk4tH\nZZDZKW6/zyquqGbNthKycovI2lbM2rwSoiPC6ZESQ/fkGNKTPQf+5Fh6pMTQOS6KsLDWHQjS2snB\n16RgzUd+MHbs2H3mEjz22GO89957AGzdupX169eTmpq6z2v69u3LiBEjABg1ahSbNm1qs3iNORQl\nFdXM27iL6MgwThrQYvHNJuUUlvPHD7IA6JESQ0ZKbMOZdUZKLGkJ0YjAzKzt/O/01Wwq2MOpg7rw\n+3OHNNlElJYQzS9P7c8NPziCT9fs4OVvNvPwjHUAxESGcc6wHvxodCbH9k1t9iCeFBPJ2L6dGdu3\n8yF9r9bQOzWev15yNDed1p9/fraBl77ZTO/OcVx7gn/nKXW4pNDcGX1biY+Pb7g9e/ZsZs2axTff\nfENcXBynnHJKk3MNoqOjG26Hh4dTXl7eJrGajqO6to7y6loqqmrd7+o6wgT6d01oleHMVTV1LNmy\nm6837GTOhp0syy6i1tNufvNp/fn1mQMP6nM27Szjyv/Mo6i8mu7JMXy5Pp89VbX7bBMZLqTERZFf\nUkn/rgm88JMxnDKoa4vvHREextlDu3P20O58l1/Kmm0l/GBgGokx7a8agHdy6NYG/QsdLikEQmJi\nIiUlTa9qWFRURKdOnYiLi2PNmjV8++23bRyd8Yfl2YXMytrOmL6dGdc3laiItpnyU1JRzZIthSzc\ntIuFm3ezfkcp5Z4kUFvXdFPwGYO78qcLh9M9+eAPKJsLypiZtZ05G3Yyb+MuyqtrCRM4KjOFn5/c\nj+P7pzJ1SQ6PfbaB/NIq/nThMMJ9aEZZm1fCpGfnUVunTJl8LMMyklFVistr3IidIjdSJ7eogryi\nCkb2SuHysb326R/wVb8uCfTr0nLHc7DrnRrf8katwJJCK0hNTeWEE05g2LBhxMbG0q1bt4bnxo8f\nz9NPP83gwYMZNGgQxx57bAAjNYeraE81f52xhlfnbaG+Oy4xJoJTB3XlrKHdOHlgl1Y9G80rqmDB\npl0NSWD1tmLqFMLE9XWdNqgr8dERxEaFERsZToznJzYynNiocL7fWcbjn63nzL9/wT3nDuGS0Zk+\nnc3vKK7g77PW8+bCrdTWKf26xHPp6ExO6J/GuCNSSY7d+x2POyKVtIRonpz9HbvLqnj0shHERIYf\n8L2XZxdy9XPziY4I47XJxzKgWyLgJmcmx0WSHBfJkPSkw9955pB0uI7mji6UvmswUVXeWZzDn6ev\nZveeKq45vg+/OKU/y7YWMiMrj1mrd7CrrIrIcOH4fmmcOaQbZw7pdsiX+zuKK/j1m0sbhlvGRoYz\nsncKo3t3ZnSfThzTqxMJPo6p37SzjDveWc6873dx0oA0/vzD4U12roK7EvnXFxt5ds731NTVceW4\n3lz/gyPISNl/JFxjz835ngc+zGJc3848c81okppIjvO/38VPX1hASlwkr113LL1Sm47DtL6QHX3U\n0YXSdw0Wa/KKuWfqShZs2s3IXin88cJhDE1P3meb2jpl0ebdzMzKY0bWdjYX7EEErhjbizsmHNnk\nAfJA5qzfya1vLKGsspabTuvPSQPSGNwjichDaDqpV1envDp/Cw9NXw3AHROOZNK43g2drVU1dbw6\nbzOPf7aBXWVVnH90Or89a+BBN1m8vzSH295cxoBuibz40zF0TdybFL9cl8/klxeSkRLLq9cde0jN\nWebQWVLooELpuwZaaWUNj85cx/NzN5EUE8FdEwbzo1GZLQ49VFXWbS/l9flbeOmbTXRJjOaBC4Zx\n9tDuzb6utk75x6frefyz9fTvksCTV45saFppLdm793DXuyv4av1OxvbpzEMXD2dFThEPz1jL1l3l\nHN8vlTsnHMlRmYc+y3722h38/JXFdEmM5uWfjaV3ajwfr8zj5teX0L9rAi/9bCxpCdEtv5FpVZYU\nOqhQ+q5tqaqmjtzCcrJ3l5O9ew9bd+/h7UXZ7Cip5LIxvfifswfRKT6q5TdqZNnWQu54Zzlr8koY\nP7Q7918wtMkmpR0lFdw6ZSlzvyvg4pGZ/PHCocRF+afLT1V5a1E2f/owi+KKGgAG90jizglH8oMB\naa0yUmnJlt389IUFhIcJVx/Xh398up6jMpN54dqxJMe1vxFAHYElhQ4qFL7rIzPX8f7SHK47sS+X\njulJdMSBOy0PRUV1LW8t3MriLYVk795D9u5y8oor8P5TCA8TjspM5t7zhnBMr06H9XnVtXU889VG\nHp21nuiIMO6aMJjLxvRsuOKYu2EnN09ZSmllNQ9cMIxLR/c8rM/z1fbiCp6d8z1Hdk/kwhEZrT75\nasOOUq55bj45he4K5JmrR7dabSFz8CwpdFAd/bs+8+VGHpy+mvTkGHKLKshIieVXp/Xn4lGZh9Wm\nDu7g/ObCrTz+6QbyiivokRxDz85x9OwUR2anWM9PHD07x9I9KeaQhj825/udZfzu3RV8s7GAsX07\n878XDeOj5Xn849N19E2L58krRzGoe+s2FwVaXlEF/125jcvH9mp2RJLxP0sKHVRH/q5vLdzK7W8v\n59yjevDYZcfw9Yad/G3mOpZtLaRX5zhuOX0AFx6T4dM4eG+1dcr7S3N4dNZ6tuzaw8heKfz27EEc\n3y/NT9/kwFSVtxZm86eP9jbdXDginQcvGm5n0cavrMxFGzrU0tkAjz76KJMnTyYuLrSH5s3M2s6d\n767gpAFpPHLp0YSHCT8Y2IWTBqTx2ZodPDJzHbe9tYwnZm/g1jMGct7wHi02d9TVKR+vyuORmevY\nsKOUIT2SeO7a0Zw6qGvAFiwSES4d05NTjuzCP2at5+ieKVwyyre5A8a0BbtSaAXeBfEOVn1RvLQ0\n385aA/1d/eHbjQVc/dx8BvdI4rXrxjV5xqyqfLIqj7/PXM/a7SX0TYunb1o8CdERJMZEkBgTSWJM\nBEme23WqPPf196zMKaZfl3h+c+YgJgzr3urt5sa0F3al0IbuvPNOvvvuO0aMGMGZZ55J165defPN\nN6msrOSiiy7i/vvvp6ysjEsvvZTs7Gxqa2u555572L59O7m5uZx66qmkpaXx+eefB/qrtLmVOUVc\n/+JCenaK5flrxxywCUVEGD+sB2cN6c5HK7bx5sKt7Cip4Lv8GkoqaiipqKa6dt8TnJ6dY/nbJUcf\nUpOTMaGq4yWF/94JeSta9z27D4cJDx3w6YceeoiVK1eydOlSZsyYwdtvv838+fNRVSZOnMiXX35J\nfn4+6enpfPTRR4CriZScnMwjjzzC559/7vOVQkeyaWcZ1z4/n8SYCF7+2Tg6+zDkMyxMOP/odM4/\nOn2fx1WVypq6hgSxp6qWgd0S26wmkTEdRcdLCgE2Y8YMZsyYwTHHHANAaWkp69ev56STTuK2227j\njjvu4LzzzuOkk04KcKSBtb24oqEg2kuTx5HuQxmF5ohIQ92fLok2McqYQ9XxkkIzZ/RtQVW56667\nuOGGG/Z7bvHixUyfPp27776b008/nXvvvTcAEQZe0Z5qrn52PrvLqnjt+mN9WjrRGNM2Ol5SCADv\n0tlnn30299xzD1deeSUJCQnk5OQQGRlJTU0NnTt3ZtKkSaSkpPCf//xnn9e29+aj6to61m8vJWtb\nMau3FVNQWklpZS1llTWUVdVQWllDaUUNZZU17KmuJTIsjOd/Moajex56OQVjTOuzpNAKvEtnT5gw\ngSuuuILjjjsOgISEBF555RU2bNjA7bffTlhYGJGRkTz11FMATJ48mfHjx5Oent5uOppLK2tYva2Y\nrNxiVuUWsSq3mPXbSxvWrY2JDKNLYjQJ0ZEkRIfTOT6Knp3jSIiKICEmgvjoCH4wII3RfQK3qpUx\npml+HZIqIuOBfwDhwH9U9aFGz/cGngO6ALuASaqa3dx7BuOQ1LYUyO9aWVPLE59/x9Ozv2tIAJ3j\noxiansSQ9CSG9EhiaHoyfdPibbSPaV/KCiA+teXt2rGAD0kVkXDgCeBMIBtYICLTVDXLa7OHgZdU\n9UUROQ34M3CVv2Iyh27hpl3c8c5yvssv44IR6Uw8Op2h6cl0S4q2iVem/cpbCZ/eD+tnwA9uh1N/\nDyH+/9mfzUdjgQ2quhFARKYAFwDeSWEI8BvP7c+BqX6MxxyCkopq/vrJWl7+djPpybE+r5FrTFAr\n3AKf/y8smwIxyXDEqfDlX6GmEs58IKQTgz+TQgaw1et+NjCu0TbLgB/impguAhJFJFVVC7w3EpHJ\nwGSAXr16Nflhqtrhz1jbevb5p6u3c/fUleQVV3Dt8X347VmDrD6Pad/27IKv/gbz/w0InHAznPhr\niE6G/94Ocx+D2ioY/5DviUHVXWnsXA9JPSApA5LSIaE7RBx8ufVAC/Rf+G+Bf4rItcCXQA5Q23gj\nVf038G9wfQqNn4+JiaGgoIDU1NQOmxhUlYKCAmJiDrxaVVF5NU9/8R0RYcKPRmUe8kLfO0sruf+D\nLD5Ylsugbok8eeXIwy4fbYKIKhR8B9kLYNtSqCprfvtuQ2HEFe6Muq0VZcOSV90BOnM0pI+E2EMY\nsVa1B+Y9BXMehapS931OuQuSM/duc87DEB4N3z7hrhjOfQTCWpj8WLwNProN1n7UxJMC8V1cgkjK\ngB5HwXE3QXRwD8H2Z1LIAbwLw2d6Hmugqrm4KwVEJAG4WFULD/aDMjMzyc7OJj8//zDCDX4xMTFk\nZmY2+dyMVXncPXUlO0srAXj8sw2M69uZS0f35JzhPYiNar5scWllDfO/L+Cr9Tt5b0kOeypr+c2Z\nA7nx5H42K7i927MLshdCzkLP70VQ4fkzi4xv/mCvtbDkZfjsT+5AOvYGSOvv2+fu3ACr33cVBvqc\nBIPPhwQfmh5VYcu3MO9pWP0BaF39E+5X2kDIGO2SROZo6DoUwiOgrhZKd0BJLhQ3+tn0FZRsg0Hn\nwOn3QtcmBmuIwNkPurP7OX93VwwTH4ewJv52VGHxSzDjHqj1NDkdcxWU5Hl9/jYoznG3d29yiWPp\nq3D+P6Dfab7tw3o5i2DmH+C0e6BX4waX1uW30UciEgGsA07HJYMFwBWqusprmzRgl6rWiciDQK2q\nNjujq6nRR6FsZ2kl901bxYfLtzG4RxJ/ufgouiZF8/aibN5auJVNBXtIiI7g/KPTuXR0JiN6piAi\n1NTWsTyniDnrdzJn/U4Wb9lNTZ0SHRHGif3TuHPCka2+FGRI2TofPn8QUvvD4InQ+wR34GortdXu\nYP7Nk1Cw3j0mYdB1CGSM8hxQx7gDbFMHPW+5S2Dev2DlO+5A2f9MGHejO7B5n0mrwvZVsHqaO5jv\n8HQfJnSD0u2AQK/jYMhElyCSG53g1FS6z5j3NGxb5pLVyGtg7PXuds5iT2Jb5H6XeU4CI2IhrrM7\nIGujhoawSNekkzYQTroNeh/f8r5ThS/+D2b/GYZfAhc+ve+/3a6NMO1ml2h6nwgTH4PUfi2/7+a5\nMO1XULABRlwJZ/3Jxd2cgu/g0wcgayrEpcH5j7p9dwiCYj0FETkHeBQ3JPU5VX1QRB4AFqrqNBH5\nEW7EkeKaj36pqpXNvaclBUdVmbo0h/s/yGJPZS03n96fG07ut89CNKrK/O938ebCbKav2EZ5dS0D\nuyXQOzWebzcWUFJRgwgMTU/ixP6uTPWo3p1sMZTDoQrf/BNm3QdxqVBRDDXl7vagc2DIBdD3ZP+1\nNau6g/KnD7iDT+YYOPJcd2adfszhNV2U7oCFz8PCZ91BPnUAjLsBehwNaz5yn7trIyDu4Dt4Igw+\nzzWd7MiCrGlum/pkkT7SJYi+J8Pa/8Ki592BvsuR7n2P+jFEHaAJVBUKN++98ikv9LTne5pqEj1t\n+3GpLTcBHchXj7iRSYMnwsXPuuT57ZPw2YMQHumuDkZec3DvX13hEs7X/3Cxnfuw+z/RWMl2+PIv\nsOgF16R1/K/g+Jsg+tBP1IIiKfiDJQXIKSzn9++tYPbafEb2SuEvPzqK/l2b/89SUlHNh8tdddGC\n0iqO75fKiQPSOL5fmk+F6EJOTSXUVBxcO3r5bpj6C1g73Z3NXfAEhEXAhlnugLjuE6gqcZ2ag8a7\ng82As1ovQWya45oYcha6A+sZ98HA8a0/kqamyp25fvsU5C52j4VFuCaiIRPhyPOabyaqb1Za/YG7\nCgFAXKzH3uiSRLD0DX7zBHzyO+h/hmuGy10MAyfAeY+4BHSoti2H938Jecvd/5VzHobE7lBZAnMf\nh7n/dM1So66Fk+/wrdmtBZYUOhhVJaewnE9WbeeRGWupU/if8YO4+rg+NlGsKXkrIKU3xCQd3Otq\na2DpKzD7IXfWOuQCGPdz6Dmm+dflLIK3rnXtyGf9yZ3pNj6wVVfAxtnubHnNR65dv/+ZcPmUw2ta\n8h5rn5gOp/4Ojr68bZqrshe69vJ+p7XcFNKUwi0umfU6Fjof0erhtYr5z8D037rmm3P+AkN/2DpJ\nq7YGvnnc/V+LiHZ9EsumwJ6dMPQi13/gS7OUjywptHPVtXWsyi1m0ebdLNq8i0Wbd7O92LWsndg/\njT//cDg9O4f2am1NUnXjz7/8C0QluLbbcTe0/MelCms+hFn3uzb4zLGuuWXZ61BZ7Nrhx90IQy7c\n98xe1bW3z7jbNVlc8gJkjmo5ztpqWPAf+PhO974T/u/gv2vpDndlsOx1l/xO/I37rpGHV3HWNCF3\niTvJOJTE15KdG+CDm2Hz1+5q68z73f+3VmZJoR0qrqjmmS83Mu/7XSzPLqSi2o26yEiJZVTvTozu\n04mRvToxND2pww69PSw1lfD+TbDiTRh+qTubW/ku1FW7Zpr6ztHG+27zXHdwzZ7vOiRP/4NrhxeB\nylJ30J33L5csErrB6J/B6J+4s7v3b3Jn/gMnwIVPHvxB45Pfuz6Icx52Haq+KsqGF8+HohyXCE78\ntX8OWKZt1NVBaZ47sfDT37YlhXZGVZn88iI+Xb2d4ZkpjOrViVG93U/35APPTTAee3bBG5Pc2dZp\n97iRJiJuREp952hZvjvoj7sBjroMira6K4N1/3V/jKfc5a4smmp2qauDjZ/Bt0/DhpkQHgWxnd17\nnnm/G39+KH/MdbUw5QpYPxMmve3bUMXdm1xCKC+ESe9Az7EH/7km5FhSaGee//p77v8gi3vOG8LP\nTuzb9gGowvaVrvNv+yroNswNW8wYFfxnoLs2wquXuPbpC5+C4T/af5uaSlj1nusc3bYUohKhusz9\nPvFWdxUR5WNz3M4Nbkbs9pXuquJwx41XlsBz46FwK1w3E7oMOvC2Bd/BixPdBKyr3oOMkYf32SZk\nWFJoR1ZkF/HDp77m5IFdeObq0W3XNKTqxn6vft+Njtn9vRvLntLbnY3WTxbq3M+TIEa79vJuw4Nn\n+v6WeTDlcjfB6bLXofdxzW+v6uYQLHrBjeg44ZbgSHqFW+CZ011iuu6zpit25q91CaGuGq6a6mbI\nGuMjSwrtRElFNec9Poeqmjqm33wSnfw9PLSmypU3qJ9gVJzjhhP2PdkNjTvyPEjo4s5ec5fsHQee\nvdC1eYIbN93j6L1XEpljIKXXgZtPqivc0DvvWbVlLcw+j4p3Hb2ZY9xnZIzav7zBynfhvRshOQOu\nfLtVR2oERPZCeP4c912vnur6LOptX+USgoTBNdOanpFrTDMsKbQDqsrNU5YyfcU2pkw+ljGtsehM\nab6bHFSc2/R0/7J8QN2Bvf/pbqz8oPEQ20JtI1WXQLIX7E0UuUvdxCxwNV7qryTSR0LZzr0JIG+F\nO7sFSMp02yT3bL4Nfs8u9xn5a2m4YkkdsLe0QWk+fPEQ9DwWLnut49TCX/E2vPMzOPoK13Et4vbz\nyxdCRAxc8wGkDQh0lKYdCvh6CqZlbyzYygfLcrn97EGHnxAaqj8+4ya91ItJ8VRt7OGaGxLToeuR\nbjLOwcyOFHFlCZIz3RhqcMMqt6/aW3oge4HrtK0XGe/avI/75d7mp6QeB/e9Kor2LW+wYZYbDQQw\n7GK44EmI7EAd8cN/5GYiz/4zdBnohii+8kOITnJXCME6lt90GHal0JZK8iB/DRxxCmvzSpj4zzmM\n6dOZl35vXFiDAAAX70lEQVQ6lrBDnYBWtcfVipnzqBtPP+IKVx4gOdONqPG187S1lO92dWvi0lwT\nR0t1dQ5WfXmDkjw3l+BQSxgEM1V45zpY+bZLrAld3BVCStNl443xhV0pBJuaSnjlYti+kpqjruC2\n7y8iMSaSR3589KElhNoaV3Fx9p9d9ceBE1z1x25DWj/2gxHbCY44xX/vLwKd+rifjkrElcgoznFX\ngFdPPbySCsYcBEsKbWXW/W4I47CLCVv+Og/XfUXZxGfpmniQTR+qrkTCp/fDznWuI/biZ6HPCf6J\n2wRGZAxcOx3Q1r/aMqYZlhTawoZP3cIdY65navqveXdxf56O+xdxn1wEUf+Aoy5p+T1qa1yhtbmP\nubb71AHw41fcaCGb3dwxdcSmMRP0LCn4W9lOmPpz6HIkG0fewe+fWsSQXqcQ9eOr4L3r4d3rYPMc\nGP9/TXeY7tnl6uLPf8bNwE3u5RbpGDGpbevzG2NCgh1V/EnVLapRvpuFJz3Ldc8sISoijMcuP4aI\n5FjXefjZH11t9ZxFcMmLe8fa71jt6u0sm+KGffY5ya0bO2iCNScYY/zGkoI/LXoe1k7ni76/4dqp\nxQzqlshTk0bRI9lTxbJ+oY5ex8PUG+FfJ8MPbnPllTfOduPSh1/iavV0Hx7Ib2KMCRGWFPwlfy36\n8e9YGTOKa1eP5KJjMnjwouFNr5U8aDzc8BW8/RO3YldiuhtJNPLajjMpyxjTLlhS8IeaSsqn/ISK\nmkhuKP8ZD1x4FJPG9Wq+plFKTzfaZEcWdBvqriKMMaaNWVJoZarKmlf/h8EFq/hDxJ08+bNzGdEz\npeUXgisylz7CvwEaY0wzLCm0oorqWl565QUmb36BWfHncecvbrP1j40x7YolBV+pwge3wJ4CN7s0\nscfemkJJGWhid379wlz+kPNHCuL6curN/yY82hKCMaZ9saTgq8LNsPhFlww2feUKtXkR4FGNJCJM\nCb96GkTHByZOY4w5DJYUfJW7xP2+/HVX57+qDIq3QXEOtYU5vPDJXFLrdnLexdfY4ifGmHbLkoKv\ncha7dXm7DnX3o+IhrT+k9eftBVv4Y1EyT105kogjD7I0tDHGBBErruKr3CVu3eJGy1CWV9XyyMx1\nHNMrhfHDugcoOGOMaR2WFHxRV+dWv2pikfTnvv6e7cWV3DVhcNutrWyMMX5iScEXBRugqsQtM+ll\nV1kVT8/+jjMGd2Ns3yBY/N0YYw6TJQVf5C52v9OP2efhxz9bT1lVDXeMHxSAoIwxpvVZUvBFzmK3\nLGKXvQf/LQV7eOXbzfx4TE8GdDuItY6NMSaIWVLwRe5i6HH0PiWr/zpjLeFhwq1nDAxgYMYY07os\nKbSkthryVuzTdLQ8u5APluVy3YlH0C3pIJfTNMaYIGZJoSU7VkNNRcPII1Xlz9PX0Dk+ihtOPiLA\nwRljTOuypNCSRp3Ms9fl883GAm4+rT+JMVbe2hjTsVhSaEnOYohJhs5HUFunPDR9Db1T47hiXO9A\nR2aMMa3OkkJLcpe4qwQR3l2czdrtJdx+9iCiImzXGWM6HjuyNae6wq2Elj6Sujrl7zPXcXRmMucO\nt/pGxpiOyZJCc/JWQF0NZIxk86495BZVcPnYFpbVNMaYdsySQnPqy2Wnj2RVrls/YVhGcgADMsYY\n//JrUhCR8SKyVkQ2iMidTTzfS0Q+F5ElIrJcRM7xZzwHLXcxxHeFpHSycouJCBMGdEsIdFTGGOM3\nfksKIhIOPAFMAIYAl4vIkEab3Q28qarHAJcBT/ornkOSs9jNTxBhVW4x/bsmEB0R3vLrjDGmnfLn\nlcJYYIOqblTVKmAKcEGjbRRI8txOBnL9GM/BqSyBnesaKqOuyi1maLo1HRljOjZ/JoUMYKvX/WzP\nY97uAyaJSDYwHfhVU28kIpNFZKGILMzPz/dHrPvbtgxQyBjJjpIKdpZWMjQ9qcWXGWNMexbojubL\ngRdUNRM4B3hZRPaLSVX/raqjVXV0ly5d2iaynL0zmVflFgMwxJKCMaaD82dSyAF6et3P9Dzm7WfA\nmwCq+g0QA6T5MSbf5S6G5F4Qn0aWJQVjTIjwZ1JYAAwQkb4iEoXrSJ7WaJstwOkAIjIYlxTaqH2o\nBTmLIcPVO8rKLaZX5ziSrNaRMaaD81tSUNUa4CbgE2A1bpTRKhF5QEQmeja7DbheRJYBrwPXqqr6\nKyaf7dkFhZu9OpmLGNLDrhKMMR1fhD/fXFWn4zqQvR+71+t2FnCCP2M4JF6VUUsqqtlUsIeLR2YG\nNiZjjGkDPl0piMi7InJuU53AHVJO/UzmEazJKwFgaIZdKRhjOj5fD/JPAlcA60XkIRHp2CvV5y6G\n1AEQk8yqHFfeYkgPm6NgjOn4fEoKqjpLVa8ERgKbgFkiMldEfiIiHa/3NXdJw0prq3KLSY2PoltS\ndICDMsYY//O5OUhEUoFrgeuAJcA/cElipl8iC5TibVCyrWGltaxtxQxJT7LKqMaYkOBrn8J7wFdA\nHHC+qk5U1TdU9VdAx6oQ19DJPJKqmjrWbS+x+QnGmJDh6+ijx1T186aeUNXRrRhP4OUsBgmH7sNZ\nv6OE6lq1mkfGmJDha/PREBFJqb8jIp1E5Bd+iimwcpdA18EQFdcwk9lqHhljQoWvSeF6VS2sv6Oq\nu4Hr/RNSAKm65iNPf8Kq3GJiI8Ppkxof4MCMMaZt+JoUwsWrp9WzVkKUf0IKoN2boHx3w8ijrNxi\nBvdIJDzMOpmNMaHB16TwMfCGiJwuIqfjSlJ87L+wAsSrk7muTsnaZmsoGGNCi68dzXcANwA/99yf\nCfzHLxEFUu4SCI+CrkPYunsPpZU1NvLIGBNSfEoKqloHPOX56bhylkD34RARxarcAsA6mY0xocXX\neQoDRORtEckSkY31P/4Ork3V1cK2pQ2VUbNyiwkPEwZ2SwxwYMYY03Z8bT56HvgD8HfgVOAnBH7V\ntoOn6jqSi3PdT0nu3tuFm6Gq1Ku8RRH9uyQQExke4KCNMabt+JoUYlX1UxERVd0M3Ccii4B7W3ph\n0Pj2KZh1H9RUNHpCIKEbJKXD0B/CgLMBNxz1xP7BsQicMca0FV+TQqWnbPZ6EbkJt6xm+ypv0W0o\njLkOkjJcAqj/SegG4fvW9MsvqWRHSaV1MhtjQo6vSeEWXN2jm4E/4pqQrvFXUH7R9wfuxwdZ22xN\nZmNMaGoxKXgmqv1YVX8LlOL6Ezq0VbluDYWhtoaCMSbEtNhZrKq1wIltEEvQWJVbTGanWJLjOt5S\nEcYY0xxfm4+WiMg04C2grP5BVX3XL1EF2OrcYob0sKYjY0zo8TUpxAAFwGlejynQ4ZJCWWUN3xeU\nccGIjECHYowxbc7XGc0dvh+h3uptxajaTGZjTGjyKSmIyPO4K4N9qOpPWz2iAKsfeTQ0w5KCMSb0\n+Np89KHX7RjgIiC39cMJvFU5xXSKi6R7UkygQzHGmDbna/PRO973ReR1YI5fIgqwVduKGJqejNfy\nEcYYEzIOtX7RAKBrawYSDKpr61iXV2r9CcaYkOVrn0IJ+/Yp5OHWWOhQNuwopaq2zmYyG2NClq/N\nRyFRP3pVrqeT2ZKCMSZE+bqewkUikux1P0VELvRfWIGRlVtMTGQYfdPaV60/Y4xpLb72KfxBVYvq\n76hqIW59hQ5lVW4RR3ZPIjzMOpmNMaHJ16TQ1Ha+DmdtF1SVrG3F1nRkjAlpviaFhSLyiIj08/w8\nAizyZ2BtLXt3OSUVNQxNt8qoxpjQ5WtS+BVQBbwBTAEqgF/6K6hA2FTg6vz172r9CcaY0OXr6KMy\n4E4/xxJQO4orAeiaGB3gSIwxJnB8HX00U0RSvO53EpFP/BdW28svdUmhiyUFY0wI87X5KM0z4ggA\nVd1NB5vRnF9SSXxUOPHRHar/3BhjDoqvSaFORHrV3xGRPjRRNbU9yy+ptKsEY0zI8/W0+PfAHBH5\nAhDgJGCy36IKgB0lFZYUjDEhz6crBVX9GBgNrAVeB24Dylt6nYiMF5G1IrJBRPbrqBaRv4vIUs/P\nOhEpbOp92oJdKRhjjO8F8a4DbgEygaXAscA37Ls8Z+PXhANPAGcC2cACEZmmqln126jqr722/xVw\nzCF8h1aRX1LJSQO6BOrjjTEmKPjap3ALMAbYrKqn4g7eLZ3VjwU2qOpGVa3CzW+4oJntL8ddhbS5\niupaiitq7ErBGBPyfE0KFapaASAi0aq6BhjUwmsygK1e97M9j+1HRHoDfYHPfIynVeWXeIajJlhS\nMMaENl87mrM98xSmAjNFZDewuRXjuAx4W1Vrm3pSRCbj6dju1atXU5scloY5CkmWFIwxoc3XGc0X\neW7eJyKfA8nAxy28LAfo6XU/0/NYUy6jmbIZqvpv4N8Ao0ePbvWhsHalYIwxzkHP1FLVL3zcdAEw\nQET64pLBZcAVjTcSkSOBTriO64DYUWIlLowxBg59jeYWqWoNcBPwCbAaeFNVV4nIAyIy0WvTy4Ap\nqhqwyXD5JZWIQOf4qECFYIwxQcGvNR1UdTowvdFj9za6f58/Y/BFfkklqfHRRIT7LUcaY0y7YEdB\nbOKaMcbUs6QA5FuJC2OMASwpAJ4rBRt5ZIwxlhRUlfzSSrraHAVjjLGkUFReTXWt2pWCMcZgSaFh\njoL1KRhjjCWFvbOZLSkYY4wlhXybzWyMMQ0sKdiVgjHGNAj5pLCjpIKYyDASov06udsYY9qFkE8K\n9bOZRSTQoRhjTMBZUiitpGtiTKDDMMaYoBDySWFHsc1mNsaYeiGfFPJLrRieMcbUC+mkUFlTS+Ge\naksKxhjjEdJJoaC0CrA5CsYYUy+kk4KVuDDGmH2FdFKwiWvGGLMvSwpgQ1KNMcbDkgKQmhAV4EiM\nMSY4hHRS2FFSQef4KCLDQ3o3GGNMg5A+GtoynMYYs6/QTgq2DKcxxuwjtJOCXSkYY8w+QjYpqCo7\nSqzEhTHGeAvZpFBcUUNVTZ0lBWOM8RKyScEmrhljzP5CNinsKKkALCkYY4y3kE0Ke2czW1Iwxph6\nIZ8UuiRYiQtjjKkXukmhtJKoiDCSYiMCHYoxxgSN0E0KnmU4RSTQoRhjTNAI3aRgy3AaY8x+Qjcp\n2MQ1Y4zZT0gnBRt5ZIwx+wrJpFBdW0dBWZVdKRhjTCMhmRQKSqsAm7hmjDGNhWRSsGU4jTGmaaGZ\nFEqtxIUxxjTFr0lBRMaLyFoR2SAidx5gm0tFJEtEVonIa/6Mp96OYiuGZ4wxTfHbdF4RCQeeAM4E\nsoEFIjJNVbO8thkA3AWcoKq7RaSrv+LxVt98lJYQ1RYfZ4wx7YY/rxTGAhtUdaOqVgFTgAsabXM9\n8ISq7gZQ1R1+jKdBfmklKXGRREeEt8XHGWNMu+HPpJABbPW6n+15zNtAYKCIfC0i34rI+KbeSEQm\ni8hCEVmYn59/2IHZMpzGGNO0QHc0RwADgFOAy4FnRCSl8Uaq+m9VHa2qo7t06XLYH2rLcBpjTNP8\nmRRygJ5e9zM9j3nLBqaparWqfg+swyUJv7ISF8YY0zR/JoUFwAAR6SsiUcBlwLRG20zFXSUgImm4\n5qSNfowJVbUSF8YYcwB+SwqqWgPcBHwCrAbeVNVVIvKAiEz0bPYJUCAiWcDnwO2qWuCvmABKK2so\nr661KwVjjGmCX1eYUdXpwPRGj93rdVuB33h+2kTDimuWFIwxZj+B7mhuc7YMpzHGHFjoJYVST92j\nJLtSMMaYxkIuKTSUuLB5CsYYs5+QSwr5pZVEhgvJsZGBDsUYY4JO6CWFkkrSEqIJC5NAh2KMMUEn\nJJOCzVEwxpimhVxSsBIXxhhzYCGXFKzEhTHGHFhIJYXaOmVXmVVINcaYAwmppFBQVkmdQpckm7hm\njDFNCamkYHMUjDGmeSGVFOpnM1ufgjHGNC20koKn7pENSTXGmKaFZFKwKwVjjGlayCWFxJgIYiLD\nAx2KMcYEpZBLCnaVYIwxBxZyScH6E4wx5sBCKinsKKmgS6LNUTDGmAMJqaSQX2KzmY0xpjkhkxTK\nKmsoq6q1PgVjjGlGyCSFnaU2R8EYY1oSMklhh81RMMaYFoVMUrCJa8YY0zJLCsYYYxqETFLokRzD\nWUO60TkuKtChGGNM0IoIdABt5ayh3TlraPdAh2GMMUEtZK4UjDHGtMySgjHGmAaWFIwxxjSwpGCM\nMaaBJQVjjDENLCkYY4xpYEnBGGNMA0sKxhhjGoiqBjqGgyIi+cDmQ3x5GrCzFcPxh/YQI7SPOC3G\n1mExto5Ax9hbVbu0tFG7SwqHQ0QWquroQMfRnPYQI7SPOC3G1mExto72ECNY85ExxhgvlhSMMcY0\nCLWk8O9AB+CD9hAjtI84LcbWYTG2jvYQY2j1KRhjjGleqF0pGGOMaYYlBWOMMQ1CJimIyHgRWSsi\nG0TkzkDH0xQR2SQiK0RkqYgsDHQ8ACLynIjsEJGVXo91FpGZIrLe87tTEMZ4n4jkePblUhE5J8Ax\n9hSRz0UkS0RWicgtnseDZl82E2Ow7csYEZkvIss8cd7vebyviMzz/I2/ISIBW2axmRhfEJHvvfbl\niEDFeCAh0acgIuHAOuBMIBtYAFyuqlkBDawREdkEjFbVoJmEIyI/AEqBl1R1mOexvwC7VPUhT4Lt\npKp3BFmM9wGlqvpwoOLyJiI9gB6qulhEEoFFwIXAtQTJvmwmxksJrn0pQLyqlopIJDAHuAX4DfCu\nqk4RkaeBZar6VJDFeCPwoaq+HYi4fBEqVwpjgQ2qulFVq4ApwAUBjqldUNUvgV2NHr4AeNFz+0Xc\ngSNgDhBjUFHVbaq62HO7BFgNZBBE+7KZGIOKOqWeu5GeHwVOA+oPtoHelweKMeiFSlLIALZ63c8m\nCP+z4/7TzBCRRSIyOdDBNKObqm7z3M4DugUymGbcJCLLPc1LAW3i8iYifYBjgHkE6b5sFCME2b4U\nkXARWQrsAGYC3wGFqlrj2STgf+ONY1TV+n35oGdf/l1EogMYYpNCJSm0Fyeq6khgAvBLT7NIUFPX\n/hiMZ0BPAf2AEcA24G+BDccRkQTgHeBWVS32fi5Y9mUTMQbdvlTVWlUdAWTiWgKODHBI+2kco4gM\nA+7CxToG6AwErNn1QEIlKeQAPb3uZ3oeCyqqmuP5vQN4D/efPRht97Q/17dD7whwPPtR1e2eP8o6\n4BmCYF962pbfAV5V1Xc9DwfVvmwqxmDcl/VUtRD4HDgOSBGRCM9TQfM37hXjeE8TnapqJfA8QbQv\n64VKUlgADPCMTogCLgOmBTimfYhIvKdzDxGJB84CVjb/qoCZBlzjuX0N8H4AY2lS/YHW4yICvC89\nHY/PAqtV9RGvp4JmXx4oxiDcl11EJMVzOxY3gGQ17sD7I89mgd6XTcW4xusEQHB9HkH3Nx4So48A\nPMPoHgXCgedU9cEAh7QPETkCd3UAEAG8FgwxisjrwCm4sr/bgT8AU4E3gV64MuaXqmrAOnoPEOMp\nuOYOBTYBN3i13bc5ETkR+ApYAdR5Hv4drs0+KPZlMzFeTnDty6NwHcnhuBPbN1X1Ac/f0BRcs8wS\nYJLnjDyYYvwM6AIIsBS40atDOiiETFIwxhjTslBpPjLGGOMDSwrGGGMaWFIwxhjTwJKCMcaYBpYU\njDHGNLCkYEwbEpFTROTDQMdhzIFYUjDGGNPAkoIxTRCRSZ56+EtF5F+e4malniJmq0TkUxHp4tl2\nhIh86yly9l59wTgR6S8iszw19ReLSD/P2yeIyNsiskZEXvXMbjUmKFhSMKYRERkM/Bg4wVPQrBa4\nEogHFqrqUOAL3MxpgJeAO1T1KNxs4PrHXwWeUNWjgeNxxeTAVR+9FRgCHAGc4PcvZYyPIlrexJiQ\nczowCljgOYmPxRWqqwPe8GzzCvCuiCQDKar6hefxF4G3PHWsMlT1PQBVrQDwvN98Vc323F8K9MEt\nwmJMwFlSMGZ/Aryoqnft86DIPY22O9QaMd71eGqxv0MTRKz5yJj9fQr8SES6QsM6yr1xfy/1VTiv\nAOaoahGwW0RO8jx+FfCFZ+WybBG50PMe0SIS16bfwphDYGcoxjSiqlkicjduFbwwoBr4JVCGWyzl\nblxz0o89L7kGeNpz0N8I/MTz+FXAv0TkAc97XNKGX8OYQ2JVUo3xkYiUqmpCoOMwxp+s+cgYY0wD\nu1IwxhjTwK4UjDHGNLCkYIwxpoElBWOMMQ0sKRhjjGlgScEYY0yD/we4GLV16MqFZAAAAABJRU5E\nrkJggg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f20433c5f98>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYUAAAEWCAYAAACJ0YulAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nGGOMqWJJwRhjTJX/D2ri9Mds4l5xAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f203fd80630>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# list all data in history\n",
    "print(histr.history.keys())\n",
    "# summarize history for accuracy\n",
    "plt.plot(histr.history['acc'])\n",
    "plt.plot(histr.history['val_acc'])\n",
    "plt.title('model accuracy')\n",
    "plt.ylabel('accuracy')\n",
    "plt.xlabel('epoch')\n",
    "plt.legend(['train', 'test'], loc='upper left')\n",
    "plt.show()\n",
    "# summarize history for loss\n",
    "plt.plot(histr.history['loss'])\n",
    "plt.plot(histr.history['val_loss'])\n",
    "plt.title('model loss')\n",
    "plt.ylabel('loss')\n",
    "plt.xlabel('epoch')\n",
    "plt.legend(['train', 'test'], loc='upper left')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "---"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [],
   "source": [
    "import matplotlib.image as mpimg\n",
    "\n",
    "min1 = np.min(X_holdout)\n",
    "max1 = np.max(X_holdout)\n",
    "diff1 = max1 - min1\n",
    "my_sum = lambda x: (x - min1)/diff1\n",
    "\n",
    "X_holdout2 = my_sum(X_holdout)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(534, 75, 75, 3)"
      ]
     },
     "execution_count": 16,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X_holdout2.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "9       0\n",
       "13      1\n",
       "14      0\n",
       "19      1\n",
       "26      1\n",
       "27      0\n",
       "28      1\n",
       "36      1\n",
       "38      1\n",
       "42      0\n",
       "43      0\n",
       "44      0\n",
       "52      0\n",
       "63      0\n",
       "73      1\n",
       "76      0\n",
       "77      1\n",
       "79      0\n",
       "80      0\n",
       "81      1\n",
       "83      1\n",
       "86      1\n",
       "89      0\n",
       "92      0\n",
       "93      0\n",
       "95      0\n",
       "96      1\n",
       "97      1\n",
       "100     0\n",
       "101     0\n",
       "       ..\n",
       "1533    0\n",
       "1537    0\n",
       "1539    0\n",
       "1541    0\n",
       "1542    0\n",
       "1543    0\n",
       "1545    0\n",
       "1548    0\n",
       "1549    0\n",
       "1552    0\n",
       "1553    0\n",
       "1559    0\n",
       "1561    0\n",
       "1566    0\n",
       "1567    0\n",
       "1569    0\n",
       "1576    0\n",
       "1577    0\n",
       "1579    0\n",
       "1585    0\n",
       "1587    0\n",
       "1588    0\n",
       "1590    0\n",
       "1591    0\n",
       "1596    0\n",
       "1597    0\n",
       "1599    0\n",
       "1601    0\n",
       "1602    0\n",
       "1603    0\n",
       "Name: is_iceberg, Length: 534, dtype: int64"
      ]
     },
     "execution_count": 17,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "Y_holdout"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.text.Text at 0x7f203fce2f28>"
      ]
     },
     "execution_count": 18,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAUUAAAEICAYAAADIsubvAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nDIIgEP4Cx4klPm4oTFIAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f20433c5c88>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.imshow(X_holdout2[0], cmap='gray', interpolation='nearest');\n",
    "tl = \"Actual label : \" +str(Y_holdout[9])+ \",\"+\" iceberg_probability : \"+str(pred_valid[0])\n",
    "plt.title(tl)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.text.Text at 0x7f203fc49160>"
      ]
     },
     "execution_count": 19,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAUUAAAEICAYAAADIsubvAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f203fda32b0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.imshow(X_holdout2[2], cmap='gray', interpolation='nearest');\n",
    "tl = \"Actual label : \" +str(Y_holdout[14])+ \",\"+\" iceberg_probability : \"+str(pred_valid[2])\n",
    "plt.title(tl)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(534, 1)"
      ]
     },
     "execution_count": 20,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "pred_valid.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(534,)"
      ]
     },
     "execution_count": 21,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "Y_holdout.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.text.Text at 0x7f203fc27748>"
      ]
     },
     "execution_count": 22,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAUUAAAEICAYAAADIsubvAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nI/ik6DiOI/wJ2bvmJU2ZhQ8AAAAASUVORK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f203fc444a8>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.imshow(X_holdout2[7], cmap='gray', interpolation='nearest');\n",
    "tl = \"Actual label : \" +str(Y_holdout[28])+ \",\"+\" iceberg_probability : \"+str(pred_valid[7])\n",
    "plt.title(tl)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.text.Text at 0x7f203fe474e0>"
      ]
     },
     "execution_count": 23,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAUUAAAEICAYAAADIsubvAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nli9mAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f203fbdea90>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.imshow(X_holdout2[532], cmap='gray', interpolation='nearest');\n",
    "tl = \"Actual label : \" +str(Y_holdout[1602])+ \",\"+\" iceberg_probability : \"+str(pred_valid[532])\n",
    "plt.title(tl)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.text.Text at 0x7f203fdc3be0>"
      ]
     },
     "execution_count": 24,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAUUAAAEICAYAAADIsubvAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f203fe4b160>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.imshow(X_holdout2[15], cmap='gray', interpolation='nearest');\n",
    "tl = \"Actual label : \" +str(Y_holdout[73])+ \",\"+\" iceberg_probability : \"+str(pred_valid[15])\n",
    "plt.title(tl)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "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
}