Workflow.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "np.random.seed(1337)\n",
    "import pandas as pd\n",
    "import keras\n",
    "from keras.models import Sequential, Model\n",
    "from keras import initializers\n",
    "from keras.layers import Dense, Conv2D, MaxPooling2D, Flatten, Dropout\n",
    "import matplotlib.pylab as plt\n",
    "%matplotlib inline\n",
    "from sklearn.model_selection import cross_val_score\n",
    "from sklearn.model_selection import KFold\n",
    "from sklearn.preprocessing import StandardScaler\n",
    "from sklearn.pipeline import Pipeline\n",
    "from sklearn import preprocessing\n",
    "from keras.preprocessing.image import ImageDataGenerator\n",
    "from keras import optimizers\n",
    "from sklearn.model_selection import train_test_split\n",
    "import cv2\n",
    "import os\n",
    "import random\n",
    "from glob import glob\n",
    "import seaborn as sns\n",
    "from sklearn.utils import class_weight\n",
    "from keras import regularizers\n",
    "from imblearn.over_sampling import RandomOverSampler\n",
    "from sklearn.externals import joblib\n",
    "from keras import backend as k \n",
    "from keras.callbacks import ModelCheckpoint, LearningRateScheduler, TensorBoard, EarlyStopping\n",
    "from keras import applications\n",
    "import pickle\n",
    "from keras.models import model_from_json"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Setup (Data Preparation)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Load npz file containing image arrays\n",
    "x_npz = np.load(\"x_images_arrays.npz\")\n",
    "x = x_npz['arr_0']\n",
    "# Load disease labels\n",
    "y_npz = np.load(\"y_disease_labels.npz\")\n",
    "y = y_npz['arr_0']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "X_train, X_test, y_train, y_test = train_test_split(x,y, test_size=0.3, stratify=y)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(3168, 128, 128, 3)\n",
      "(1359, 128, 128, 3)\n",
      "(3168,)\n",
      "(1359,)\n"
     ]
    }
   ],
   "source": [
    "print(X_train.shape)\n",
    "print(X_test.shape)\n",
    "print(y_train.shape)\n",
    "print(y_test.shape)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "X_train_reshape = X_train.reshape(X_train.shape[0], -1).astype('float32')\n",
    "X_test_reshape = X_test.reshape(X_test.shape[0], -1).astype('float32')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "#one-hot encode y_train & y_test\n",
    "y_trainHot = keras.utils.to_categorical(y_train, 8)\n",
    "y_testHot = keras.utils.to_categorical(y_test, 8)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [],
   "source": [
    "epochs=5\n",
    "# optimizer = optimizers.Adam(lr=0.00003, decay=0.0001)\n",
    "optimizer = optimizers.SGD(lr=0.003, decay=0.0001, momentum=0.9)\n",
    "# optimizer= adam\n",
    "batch_size=100\n",
    "loss='categorical_crossentropy'\n",
    "activation='relu'\n",
    "out_activation='softmax'"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [],
   "source": [
    "tbcallback = keras.callbacks.TensorBoard(log_dir='./logs', histogram_freq=0, batch_size=32, \n",
    "                                         write_graph=True, write_images=True)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Balanced Class Weights"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "{0: 0.1858984888304862, 1: 7.859375, 2: 1.125, 3: 4.963815789473684, 4: 2.7875615763546797, 5: 3.9296875, 6: 2.947265625, 7: 2.2191176470588236}\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>0</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0.185898</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>7.859375</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>1.125000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>4.963816</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>2.787562</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>3.929688</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>2.947266</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>2.219118</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "          0\n",
       "0  0.185898\n",
       "1  7.859375\n",
       "2  1.125000\n",
       "3  4.963816\n",
       "4  2.787562\n",
       "5  3.929688\n",
       "6  2.947266\n",
       "7  2.219118"
      ]
     },
     "execution_count": 32,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "class_weights = class_weight.compute_class_weight('balanced', np.unique(y), y)\n",
    "class_weights = dict(enumerate(class_weights))\n",
    "print(class_weights)\n",
    "class_weights_df = pd.DataFrame.from_dict(class_weights, orient='index')\n",
    "class_weights_df"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train on 3168 samples, validate on 1359 samples\n",
      "Epoch 1/25\n",
      "3168/3168 [==============================] - 90s 28ms/step - loss: 14.0614 - acc: 0.6319 - val_loss: 5.2778 - val_acc: 0.6726\n",
      "Epoch 2/25\n",
      "3168/3168 [==============================] - 88s 28ms/step - loss: 14.1339 - acc: 0.6253 - val_loss: 5.2778 - val_acc: 0.6726\n",
      "Epoch 3/25\n",
      "3168/3168 [==============================] - 83s 26ms/step - loss: 14.0948 - acc: 0.6288 - val_loss: 5.2778 - val_acc: 0.6726\n",
      "Epoch 4/25\n",
      "3168/3168 [==============================] - 77s 24ms/step - loss: 14.1063 - acc: 0.6297 - val_loss: 5.2778 - val_acc: 0.6726\n",
      "Epoch 5/25\n",
      "3168/3168 [==============================] - 74s 24ms/step - loss: 14.1605 - acc: 0.6228 - val_loss: 5.2778 - val_acc: 0.6726\n",
      "Epoch 6/25\n",
      "3168/3168 [==============================] - 80s 25ms/step - loss: 14.0911 - acc: 0.6316 - val_loss: 5.2778 - val_acc: 0.6726\n",
      "Epoch 7/25\n",
      "3168/3168 [==============================] - 77s 24ms/step - loss: 14.1224 - acc: 0.6244 - val_loss: 5.2778 - val_acc: 0.6726\n",
      "Epoch 8/25\n",
      "3168/3168 [==============================] - 78s 25ms/step - loss: 14.0939 - acc: 0.6323 - val_loss: 5.2778 - val_acc: 0.6726\n",
      "Epoch 9/25\n",
      "3168/3168 [==============================] - 80s 25ms/step - loss: 14.1407 - acc: 0.6294 - val_loss: 5.2778 - val_acc: 0.6726\n",
      "Epoch 10/25\n",
      "3168/3168 [==============================] - 77s 24ms/step - loss: 14.0462 - acc: 0.6338 - val_loss: 5.2778 - val_acc: 0.6726\n",
      "Epoch 11/25\n",
      "3168/3168 [==============================] - 76s 24ms/step - loss: 14.1377 - acc: 0.6256 - val_loss: 5.2778 - val_acc: 0.6726\n",
      "Epoch 12/25\n",
      "3168/3168 [==============================] - 82s 26ms/step - loss: 14.1309 - acc: 0.6231 - val_loss: 5.2778 - val_acc: 0.6726\n",
      "Epoch 13/25\n",
      "3168/3168 [==============================] - 73s 23ms/step - loss: 14.1281 - acc: 0.6241 - val_loss: 5.2778 - val_acc: 0.6726\n",
      "Epoch 14/25\n",
      "3168/3168 [==============================] - 69s 22ms/step - loss: 14.0996 - acc: 0.6304 - val_loss: 5.2778 - val_acc: 0.6726\n",
      "Epoch 15/25\n",
      "3168/3168 [==============================] - 69s 22ms/step - loss: 14.1397 - acc: 0.6282 - val_loss: 5.2778 - val_acc: 0.6726\n",
      "Epoch 16/25\n",
      "3168/3168 [==============================] - 68s 22ms/step - loss: 14.1594 - acc: 0.6184 - val_loss: 5.2778 - val_acc: 0.6726\n",
      "Epoch 17/25\n",
      "3168/3168 [==============================] - 69s 22ms/step - loss: 14.0614 - acc: 0.6319 - val_loss: 5.2778 - val_acc: 0.6726\n",
      "Epoch 18/25\n",
      "3168/3168 [==============================] - 69s 22ms/step - loss: 14.0833 - acc: 0.6294 - val_loss: 5.2778 - val_acc: 0.6726\n",
      "Epoch 19/25\n",
      "3168/3168 [==============================] - 69s 22ms/step - loss: 14.1234 - acc: 0.6256 - val_loss: 5.2778 - val_acc: 0.6726\n",
      "Epoch 20/25\n",
      "3168/3168 [==============================] - 76s 24ms/step - loss: 14.0823 - acc: 0.6266 - val_loss: 5.2778 - val_acc: 0.6726\n",
      "Epoch 21/25\n",
      "3168/3168 [==============================] - 71s 22ms/step - loss: 14.1176 - acc: 0.6259 - val_loss: 5.2778 - val_acc: 0.6726\n",
      "Epoch 22/25\n",
      "3168/3168 [==============================] - 72s 23ms/step - loss: 14.1081 - acc: 0.6259 - val_loss: 5.2778 - val_acc: 0.6726\n",
      "Epoch 23/25\n",
      "3168/3168 [==============================] - 68s 21ms/step - loss: 14.0967 - acc: 0.6297 - val_loss: 5.2778 - val_acc: 0.6726\n",
      "Epoch 24/25\n",
      "3168/3168 [==============================] - 78s 25ms/step - loss: 14.1042 - acc: 0.6256 - val_loss: 5.2778 - val_acc: 0.6726\n",
      "Epoch 25/25\n",
      "3168/3168 [==============================] - 72s 23ms/step - loss: 14.1311 - acc: 0.6278 - val_loss: 5.2778 - val_acc: 0.6726\n"
     ]
    }
   ],
   "source": [
    "model = Sequential()\n",
    "model.add(Dense(150, input_dim=49152, activation=activation))\n",
    "model.add(Dropout(0.3))\n",
    "model.add(Dense(125, activation=activation))\n",
    "model.add(Dropout(0.3))\n",
    "model.add(Dense(100, activation=activation))\n",
    "model.add(Dropout(0.3))\n",
    "model.add(Dense(75, activation=activation))\n",
    "model.add(Dropout(0.3))\n",
    "model.add(Dense(50, activation=activation))\n",
    "model.add(Dropout(0.3))\n",
    "model.add(Dense(8, activation=out_activation))\n",
    "model.compile(loss=loss, optimizer=optimizer, metrics=['accuracy']) \n",
    "model.load_weights(\"my_model_weights.h5\")\n",
    "history = model.fit(X_train_reshape, y_trainHot, epochs=25, batch_size=batch_size, \n",
    "                    class_weight = class_weights, verbose=1, \n",
    "                    validation_data=[X_test_reshape, y_testHot], callbacks=[tbcallback])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "3168/3168 [==============================] - 31s 10ms/step\n",
      "1359/1359 [==============================] - 12s 9ms/step\n",
      "[5.281118400169142, 0.6723484848484849]\n",
      "[5.277816405447202, 0.6725533483131919]\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "results_train = model.evaluate(X_train_reshape, y_trainHot)\n",
    "results_test = model.evaluate(X_test_reshape, y_testHot)\n",
    "print(results_train)\n",
    "print(results_test)\n",
    "val_dict = history.history\n",
    "\n",
    "plt.clf()\n",
    "\n",
    "loss_values = val_dict['loss']\n",
    "val_loss_values = val_dict['val_loss']\n",
    "\n",
    "epochs = range(1, len(loss_values) + 1)\n",
    "plt.plot(epochs, loss_values, 'g', label='Training loss')\n",
    "plt.plot(epochs, val_loss_values, 'g.', label='Validation loss')\n",
    "\n",
    "plt.title('Training & validation loss')\n",
    "plt.xlabel('Epochs')\n",
    "plt.ylabel('Loss')\n",
    "plt.legend()\n",
    "plt.show()\n",
    "\n",
    "plt.clf()\n",
    "\n",
    "acc_values = val_dict['acc'] \n",
    "val_acc_values = val_dict['val_acc']\n",
    "\n",
    "plt.plot(epochs, acc_values, 'r', label='Training acc')\n",
    "plt.plot(epochs, val_acc_values, 'r.', label='Validation acc')\n",
    "plt.title('Training & validation accuracy')\n",
    "plt.xlabel('Epochs')\n",
    "plt.ylabel('Loss')\n",
    "plt.legend()\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Saved model to disk\n"
     ]
    }
   ],
   "source": [
    "# serialize model to JSON\n",
    "model_json = model.to_json()\n",
    "with open(\"balanced_mlp_model.json\", \"w\") as json_file:\n",
    "    json_file.write(model_json)\n",
    "# serialize weights to HDF5\n",
    "model.save_weights(\"balanced_mlp_model.h5\")\n",
    "print(\"Saved model to disk\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Loaded model from disk\n"
     ]
    }
   ],
   "source": [
    "# load json and create model\n",
    "json_file = open('balanced_mlp_model.json', 'r')\n",
    "loaded_model_json = json_file.read()\n",
    "json_file.close()\n",
    "loaded_model = model_from_json(loaded_model_json)\n",
    "# load weights into new model\n",
    "loaded_model.load_weights(\"balanced_mlp_model.h5\")\n",
    "print(\"Loaded model from disk\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1359/1359 [==============================] - 13s 9ms/step\n",
      "acc: 67.26%\n"
     ]
    }
   ],
   "source": [
    "# evaluate loaded model on test data\n",
    "loaded_model.compile(loss='categorical_crossentropy', optimizer=optimizer, metrics=['accuracy'])\n",
    "score = loaded_model.evaluate(X_test_reshape, y_testHot)\n",
    "print(\"%s: %.2f%%\" % (loaded_model.metrics_names[1], score[1]*100))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "3168/3168 [==============================] - 30s 10ms/step\n",
      "1359/1359 [==============================] - 13s 10ms/step\n",
      "[5.281118400169142, 0.6723484848484849]\n",
      "[5.277816405447202, 0.6725533483131919]\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "results_train = loaded_model.evaluate(X_train_reshape, y_trainHot)\n",
    "results_test = loaded_model.evaluate(X_test_reshape, y_testHot)\n",
    "print(results_train)\n",
    "print(results_test)\n",
    "val_dict = history.history\n",
    "\n",
    "plt.clf()\n",
    "\n",
    "loss_values = val_dict['loss']\n",
    "val_loss_values = val_dict['val_loss']\n",
    "\n",
    "epochs = range(1, len(loss_values) + 1)\n",
    "plt.plot(epochs, loss_values, 'g', label='Training loss')\n",
    "plt.plot(epochs, val_loss_values, 'g.', label='Validation loss')\n",
    "\n",
    "plt.title('Training & validation loss')\n",
    "plt.xlabel('Epochs')\n",
    "plt.ylabel('Loss')\n",
    "plt.legend()\n",
    "plt.show()\n",
    "\n",
    "plt.clf()\n",
    "\n",
    "acc_values = val_dict['acc'] \n",
    "val_acc_values = val_dict['val_acc']\n",
    "\n",
    "plt.plot(epochs, acc_values, 'r', label='Training acc')\n",
    "plt.plot(epochs, val_acc_values, 'r.', label='Validation acc')\n",
    "plt.title('Training & validation accuracy')\n",
    "plt.xlabel('Epochs')\n",
    "plt.ylabel('Loss')\n",
    "plt.legend()\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 68,
   "metadata": {},
   "outputs": [
    {
     "ename": "TypeError",
     "evalue": "can't pickle _thread.RLock objects",
     "output_type": "error",
     "traceback": [
      "\u001b[0;31m---------------------------------------------------------------------------\u001b[0m",
      "\u001b[0;31mTypeError\u001b[0m                                 Traceback (most recent call last)",
      "\u001b[0;32m<ipython-input-68-b41f9304db01>\u001b[0m in \u001b[0;36m<module>\u001b[0;34m()\u001b[0m\n\u001b[1;32m      1\u001b[0m \u001b[0mmodel\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0msave_weights\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m'my_model_weights.h5'\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m----> 2\u001b[0;31m \u001b[0mjoblib\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mdump\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mmodel\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m'mlp_balance.pickle'\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m",
      "\u001b[0;32m/anaconda3/lib/python3.6/site-packages/sklearn/externals/joblib/numpy_pickle.py\u001b[0m in \u001b[0;36mdump\u001b[0;34m(value, filename, compress, protocol, cache_size)\u001b[0m\n\u001b[1;32m    482\u001b[0m     \u001b[0;32melif\u001b[0m \u001b[0mis_filename\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    483\u001b[0m         \u001b[0;32mwith\u001b[0m \u001b[0mopen\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mfilename\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m'wb'\u001b[0m\u001b[0;34m)\u001b[0m \u001b[0;32mas\u001b[0m \u001b[0mf\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 484\u001b[0;31m             \u001b[0mNumpyPickler\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mf\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mprotocol\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0mprotocol\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mdump\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mvalue\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m    485\u001b[0m     \u001b[0;32melse\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    486\u001b[0m         \u001b[0mNumpyPickler\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mfilename\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mprotocol\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0mprotocol\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mdump\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mvalue\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m/anaconda3/lib/python3.6/pickle.py\u001b[0m in \u001b[0;36mdump\u001b[0;34m(self, obj)\u001b[0m\n\u001b[1;32m    407\u001b[0m         \u001b[0;32mif\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mproto\u001b[0m \u001b[0;34m>=\u001b[0m \u001b[0;36m4\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    408\u001b[0m             \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mframer\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mstart_framing\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 409\u001b[0;31m         \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0msave\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mobj\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m    410\u001b[0m         \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mwrite\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mSTOP\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    411\u001b[0m         \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mframer\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mend_framing\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m/anaconda3/lib/python3.6/site-packages/sklearn/externals/joblib/numpy_pickle.py\u001b[0m in \u001b[0;36msave\u001b[0;34m(self, obj)\u001b[0m\n\u001b[1;32m    279\u001b[0m             \u001b[0;32mreturn\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    280\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 281\u001b[0;31m         \u001b[0;32mreturn\u001b[0m \u001b[0mPickler\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0msave\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mself\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mobj\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m    282\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    283\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m/anaconda3/lib/python3.6/pickle.py\u001b[0m in \u001b[0;36msave\u001b[0;34m(self, obj, save_persistent_id)\u001b[0m\n\u001b[1;32m    519\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    520\u001b[0m         \u001b[0;31m# Save the reduce() output and finally memoize the object\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 521\u001b[0;31m         \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0msave_reduce\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mobj\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0mobj\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m*\u001b[0m\u001b[0mrv\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m    522\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    523\u001b[0m     \u001b[0;32mdef\u001b[0m \u001b[0mpersistent_id\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mself\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mobj\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m/anaconda3/lib/python3.6/pickle.py\u001b[0m in \u001b[0;36msave_reduce\u001b[0;34m(self, func, args, state, listitems, dictitems, obj)\u001b[0m\n\u001b[1;32m    632\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    633\u001b[0m         \u001b[0;32mif\u001b[0m \u001b[0mstate\u001b[0m \u001b[0;32mis\u001b[0m \u001b[0;32mnot\u001b[0m \u001b[0;32mNone\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 634\u001b[0;31m             \u001b[0msave\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mstate\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m    635\u001b[0m             \u001b[0mwrite\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mBUILD\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    636\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m/anaconda3/lib/python3.6/site-packages/sklearn/externals/joblib/numpy_pickle.py\u001b[0m in \u001b[0;36msave\u001b[0;34m(self, obj)\u001b[0m\n\u001b[1;32m    279\u001b[0m             \u001b[0;32mreturn\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    280\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 281\u001b[0;31m         \u001b[0;32mreturn\u001b[0m \u001b[0mPickler\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0msave\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mself\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mobj\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m    282\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    283\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m/anaconda3/lib/python3.6/pickle.py\u001b[0m in \u001b[0;36msave\u001b[0;34m(self, obj, save_persistent_id)\u001b[0m\n\u001b[1;32m    474\u001b[0m         \u001b[0mf\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mdispatch\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mget\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mt\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    475\u001b[0m         \u001b[0;32mif\u001b[0m \u001b[0mf\u001b[0m \u001b[0;32mis\u001b[0m \u001b[0;32mnot\u001b[0m \u001b[0;32mNone\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 476\u001b[0;31m             \u001b[0mf\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mself\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mobj\u001b[0m\u001b[0;34m)\u001b[0m \u001b[0;31m# Call unbound method with explicit self\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m    477\u001b[0m             \u001b[0;32mreturn\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    478\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m/anaconda3/lib/python3.6/pickle.py\u001b[0m in \u001b[0;36msave_dict\u001b[0;34m(self, obj)\u001b[0m\n\u001b[1;32m    819\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    820\u001b[0m         \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mmemoize\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mobj\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 821\u001b[0;31m         \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_batch_setitems\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mobj\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mitems\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m    822\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    823\u001b[0m     \u001b[0mdispatch\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0mdict\u001b[0m\u001b[0;34m]\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0msave_dict\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m/anaconda3/lib/python3.6/pickle.py\u001b[0m in \u001b[0;36m_batch_setitems\u001b[0;34m(self, items)\u001b[0m\n\u001b[1;32m    845\u001b[0m                 \u001b[0;32mfor\u001b[0m \u001b[0mk\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mv\u001b[0m \u001b[0;32min\u001b[0m \u001b[0mtmp\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    846\u001b[0m                     \u001b[0msave\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mk\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 847\u001b[0;31m                     \u001b[0msave\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mv\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m    848\u001b[0m                 \u001b[0mwrite\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mSETITEMS\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    849\u001b[0m             \u001b[0;32melif\u001b[0m \u001b[0mn\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m/anaconda3/lib/python3.6/site-packages/sklearn/externals/joblib/numpy_pickle.py\u001b[0m in \u001b[0;36msave\u001b[0;34m(self, obj)\u001b[0m\n\u001b[1;32m    279\u001b[0m             \u001b[0;32mreturn\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    280\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 281\u001b[0;31m         \u001b[0;32mreturn\u001b[0m \u001b[0mPickler\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0msave\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mself\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mobj\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m    282\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    283\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m/anaconda3/lib/python3.6/pickle.py\u001b[0m in \u001b[0;36msave\u001b[0;34m(self, obj, save_persistent_id)\u001b[0m\n\u001b[1;32m    519\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    520\u001b[0m         \u001b[0;31m# Save the reduce() output and finally memoize the object\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 521\u001b[0;31m         \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0msave_reduce\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mobj\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0mobj\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m*\u001b[0m\u001b[0mrv\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m    522\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    523\u001b[0m     \u001b[0;32mdef\u001b[0m \u001b[0mpersistent_id\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mself\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mobj\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m/anaconda3/lib/python3.6/pickle.py\u001b[0m in \u001b[0;36msave_reduce\u001b[0;34m(self, func, args, state, listitems, dictitems, obj)\u001b[0m\n\u001b[1;32m    632\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    633\u001b[0m         \u001b[0;32mif\u001b[0m \u001b[0mstate\u001b[0m \u001b[0;32mis\u001b[0m \u001b[0;32mnot\u001b[0m \u001b[0;32mNone\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 634\u001b[0;31m             \u001b[0msave\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mstate\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m    635\u001b[0m             \u001b[0mwrite\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mBUILD\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    636\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m/anaconda3/lib/python3.6/site-packages/sklearn/externals/joblib/numpy_pickle.py\u001b[0m in \u001b[0;36msave\u001b[0;34m(self, obj)\u001b[0m\n\u001b[1;32m    279\u001b[0m             \u001b[0;32mreturn\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    280\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 281\u001b[0;31m         \u001b[0;32mreturn\u001b[0m \u001b[0mPickler\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0msave\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mself\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mobj\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m    282\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    283\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m/anaconda3/lib/python3.6/pickle.py\u001b[0m in \u001b[0;36msave\u001b[0;34m(self, obj, save_persistent_id)\u001b[0m\n\u001b[1;32m    474\u001b[0m         \u001b[0mf\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mdispatch\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mget\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mt\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    475\u001b[0m         \u001b[0;32mif\u001b[0m \u001b[0mf\u001b[0m \u001b[0;32mis\u001b[0m \u001b[0;32mnot\u001b[0m \u001b[0;32mNone\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 476\u001b[0;31m             \u001b[0mf\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mself\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mobj\u001b[0m\u001b[0;34m)\u001b[0m \u001b[0;31m# Call unbound method with explicit self\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m    477\u001b[0m             \u001b[0;32mreturn\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    478\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m/anaconda3/lib/python3.6/pickle.py\u001b[0m in \u001b[0;36msave_dict\u001b[0;34m(self, obj)\u001b[0m\n\u001b[1;32m    819\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    820\u001b[0m         \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mmemoize\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mobj\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 821\u001b[0;31m         \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_batch_setitems\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mobj\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mitems\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m    822\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    823\u001b[0m     \u001b[0mdispatch\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0mdict\u001b[0m\u001b[0;34m]\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0msave_dict\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m/anaconda3/lib/python3.6/pickle.py\u001b[0m in \u001b[0;36m_batch_setitems\u001b[0;34m(self, items)\u001b[0m\n\u001b[1;32m    845\u001b[0m                 \u001b[0;32mfor\u001b[0m \u001b[0mk\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mv\u001b[0m \u001b[0;32min\u001b[0m \u001b[0mtmp\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    846\u001b[0m                     \u001b[0msave\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mk\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 847\u001b[0;31m                     \u001b[0msave\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mv\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m    848\u001b[0m                 \u001b[0mwrite\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mSETITEMS\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    849\u001b[0m             \u001b[0;32melif\u001b[0m \u001b[0mn\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m/anaconda3/lib/python3.6/site-packages/sklearn/externals/joblib/numpy_pickle.py\u001b[0m in \u001b[0;36msave\u001b[0;34m(self, obj)\u001b[0m\n\u001b[1;32m    279\u001b[0m             \u001b[0;32mreturn\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    280\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 281\u001b[0;31m         \u001b[0;32mreturn\u001b[0m \u001b[0mPickler\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0msave\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mself\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mobj\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m    282\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    283\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m/anaconda3/lib/python3.6/pickle.py\u001b[0m in \u001b[0;36msave\u001b[0;34m(self, obj, save_persistent_id)\u001b[0m\n\u001b[1;32m    474\u001b[0m         \u001b[0mf\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mdispatch\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mget\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mt\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    475\u001b[0m         \u001b[0;32mif\u001b[0m \u001b[0mf\u001b[0m \u001b[0;32mis\u001b[0m \u001b[0;32mnot\u001b[0m \u001b[0;32mNone\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 476\u001b[0;31m             \u001b[0mf\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mself\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mobj\u001b[0m\u001b[0;34m)\u001b[0m \u001b[0;31m# Call unbound method with explicit self\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m    477\u001b[0m             \u001b[0;32mreturn\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    478\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m/anaconda3/lib/python3.6/pickle.py\u001b[0m in \u001b[0;36msave_list\u001b[0;34m(self, obj)\u001b[0m\n\u001b[1;32m    779\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    780\u001b[0m         \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mmemoize\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mobj\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 781\u001b[0;31m         \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_batch_appends\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mobj\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m    782\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    783\u001b[0m     \u001b[0mdispatch\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0mlist\u001b[0m\u001b[0;34m]\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0msave_list\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m/anaconda3/lib/python3.6/pickle.py\u001b[0m in \u001b[0;36m_batch_appends\u001b[0;34m(self, items)\u001b[0m\n\u001b[1;32m    803\u001b[0m                 \u001b[0mwrite\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mMARK\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    804\u001b[0m                 \u001b[0;32mfor\u001b[0m \u001b[0mx\u001b[0m \u001b[0;32min\u001b[0m \u001b[0mtmp\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 805\u001b[0;31m                     \u001b[0msave\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mx\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m    806\u001b[0m                 \u001b[0mwrite\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mAPPENDS\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    807\u001b[0m             \u001b[0;32melif\u001b[0m \u001b[0mn\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m/anaconda3/lib/python3.6/site-packages/sklearn/externals/joblib/numpy_pickle.py\u001b[0m in \u001b[0;36msave\u001b[0;34m(self, obj)\u001b[0m\n\u001b[1;32m    279\u001b[0m             \u001b[0;32mreturn\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    280\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 281\u001b[0;31m         \u001b[0;32mreturn\u001b[0m \u001b[0mPickler\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0msave\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mself\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mobj\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m    282\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    283\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m/anaconda3/lib/python3.6/pickle.py\u001b[0m in \u001b[0;36msave\u001b[0;34m(self, obj, save_persistent_id)\u001b[0m\n\u001b[1;32m    519\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    520\u001b[0m         \u001b[0;31m# Save the reduce() output and finally memoize the object\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 521\u001b[0;31m         \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0msave_reduce\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mobj\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0mobj\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m*\u001b[0m\u001b[0mrv\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m    522\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    523\u001b[0m     \u001b[0;32mdef\u001b[0m \u001b[0mpersistent_id\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mself\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mobj\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m/anaconda3/lib/python3.6/pickle.py\u001b[0m in \u001b[0;36msave_reduce\u001b[0;34m(self, func, args, state, listitems, dictitems, obj)\u001b[0m\n\u001b[1;32m    632\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    633\u001b[0m         \u001b[0;32mif\u001b[0m \u001b[0mstate\u001b[0m \u001b[0;32mis\u001b[0m \u001b[0;32mnot\u001b[0m \u001b[0;32mNone\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 634\u001b[0;31m             \u001b[0msave\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mstate\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m    635\u001b[0m             \u001b[0mwrite\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mBUILD\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    636\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m/anaconda3/lib/python3.6/site-packages/sklearn/externals/joblib/numpy_pickle.py\u001b[0m in \u001b[0;36msave\u001b[0;34m(self, obj)\u001b[0m\n\u001b[1;32m    279\u001b[0m             \u001b[0;32mreturn\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    280\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 281\u001b[0;31m         \u001b[0;32mreturn\u001b[0m \u001b[0mPickler\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0msave\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mself\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mobj\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m    282\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    283\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m/anaconda3/lib/python3.6/pickle.py\u001b[0m in \u001b[0;36msave\u001b[0;34m(self, obj, save_persistent_id)\u001b[0m\n\u001b[1;32m    474\u001b[0m         \u001b[0mf\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mdispatch\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mget\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mt\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    475\u001b[0m         \u001b[0;32mif\u001b[0m \u001b[0mf\u001b[0m \u001b[0;32mis\u001b[0m \u001b[0;32mnot\u001b[0m \u001b[0;32mNone\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 476\u001b[0;31m             \u001b[0mf\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mself\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mobj\u001b[0m\u001b[0;34m)\u001b[0m \u001b[0;31m# Call unbound method with explicit self\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m    477\u001b[0m             \u001b[0;32mreturn\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    478\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m/anaconda3/lib/python3.6/pickle.py\u001b[0m in \u001b[0;36msave_dict\u001b[0;34m(self, obj)\u001b[0m\n\u001b[1;32m    819\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    820\u001b[0m         \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mmemoize\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mobj\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 821\u001b[0;31m         \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_batch_setitems\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mobj\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mitems\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m    822\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    823\u001b[0m     \u001b[0mdispatch\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0mdict\u001b[0m\u001b[0;34m]\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0msave_dict\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m/anaconda3/lib/python3.6/pickle.py\u001b[0m in \u001b[0;36m_batch_setitems\u001b[0;34m(self, items)\u001b[0m\n\u001b[1;32m    845\u001b[0m                 \u001b[0;32mfor\u001b[0m \u001b[0mk\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mv\u001b[0m \u001b[0;32min\u001b[0m \u001b[0mtmp\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    846\u001b[0m                     \u001b[0msave\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mk\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 847\u001b[0;31m                     \u001b[0msave\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mv\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m    848\u001b[0m                 \u001b[0mwrite\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mSETITEMS\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    849\u001b[0m             \u001b[0;32melif\u001b[0m \u001b[0mn\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m/anaconda3/lib/python3.6/site-packages/sklearn/externals/joblib/numpy_pickle.py\u001b[0m in \u001b[0;36msave\u001b[0;34m(self, obj)\u001b[0m\n\u001b[1;32m    279\u001b[0m             \u001b[0;32mreturn\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    280\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 281\u001b[0;31m         \u001b[0;32mreturn\u001b[0m \u001b[0mPickler\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0msave\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mself\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mobj\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m    282\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    283\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m/anaconda3/lib/python3.6/pickle.py\u001b[0m in \u001b[0;36msave\u001b[0;34m(self, obj, save_persistent_id)\u001b[0m\n\u001b[1;32m    519\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    520\u001b[0m         \u001b[0;31m# Save the reduce() output and finally memoize the object\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 521\u001b[0;31m         \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0msave_reduce\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mobj\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0mobj\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m*\u001b[0m\u001b[0mrv\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m    522\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    523\u001b[0m     \u001b[0;32mdef\u001b[0m \u001b[0mpersistent_id\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mself\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mobj\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m/anaconda3/lib/python3.6/pickle.py\u001b[0m in \u001b[0;36msave_reduce\u001b[0;34m(self, func, args, state, listitems, dictitems, obj)\u001b[0m\n\u001b[1;32m    632\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    633\u001b[0m         \u001b[0;32mif\u001b[0m \u001b[0mstate\u001b[0m \u001b[0;32mis\u001b[0m \u001b[0;32mnot\u001b[0m \u001b[0;32mNone\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 634\u001b[0;31m             \u001b[0msave\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mstate\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m    635\u001b[0m             \u001b[0mwrite\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mBUILD\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    636\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m/anaconda3/lib/python3.6/site-packages/sklearn/externals/joblib/numpy_pickle.py\u001b[0m in \u001b[0;36msave\u001b[0;34m(self, obj)\u001b[0m\n\u001b[1;32m    279\u001b[0m             \u001b[0;32mreturn\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    280\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 281\u001b[0;31m         \u001b[0;32mreturn\u001b[0m \u001b[0mPickler\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0msave\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mself\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mobj\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m    282\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    283\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m/anaconda3/lib/python3.6/pickle.py\u001b[0m in \u001b[0;36msave\u001b[0;34m(self, obj, save_persistent_id)\u001b[0m\n\u001b[1;32m    474\u001b[0m         \u001b[0mf\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mdispatch\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mget\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mt\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    475\u001b[0m         \u001b[0;32mif\u001b[0m \u001b[0mf\u001b[0m \u001b[0;32mis\u001b[0m \u001b[0;32mnot\u001b[0m \u001b[0;32mNone\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 476\u001b[0;31m             \u001b[0mf\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mself\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mobj\u001b[0m\u001b[0;34m)\u001b[0m \u001b[0;31m# Call unbound method with explicit self\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m    477\u001b[0m             \u001b[0;32mreturn\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    478\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m/anaconda3/lib/python3.6/pickle.py\u001b[0m in \u001b[0;36msave_dict\u001b[0;34m(self, obj)\u001b[0m\n\u001b[1;32m    819\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    820\u001b[0m         \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mmemoize\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mobj\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 821\u001b[0;31m         \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_batch_setitems\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mobj\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mitems\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m    822\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    823\u001b[0m     \u001b[0mdispatch\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0mdict\u001b[0m\u001b[0;34m]\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0msave_dict\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m/anaconda3/lib/python3.6/pickle.py\u001b[0m in \u001b[0;36m_batch_setitems\u001b[0;34m(self, items)\u001b[0m\n\u001b[1;32m    845\u001b[0m                 \u001b[0;32mfor\u001b[0m \u001b[0mk\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mv\u001b[0m \u001b[0;32min\u001b[0m \u001b[0mtmp\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    846\u001b[0m                     \u001b[0msave\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mk\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 847\u001b[0;31m                     \u001b[0msave\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mv\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m    848\u001b[0m                 \u001b[0mwrite\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mSETITEMS\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    849\u001b[0m             \u001b[0;32melif\u001b[0m \u001b[0mn\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m/anaconda3/lib/python3.6/site-packages/sklearn/externals/joblib/numpy_pickle.py\u001b[0m in \u001b[0;36msave\u001b[0;34m(self, obj)\u001b[0m\n\u001b[1;32m    279\u001b[0m             \u001b[0;32mreturn\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    280\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 281\u001b[0;31m         \u001b[0;32mreturn\u001b[0m \u001b[0mPickler\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0msave\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mself\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mobj\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m    282\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    283\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m/anaconda3/lib/python3.6/pickle.py\u001b[0m in \u001b[0;36msave\u001b[0;34m(self, obj, save_persistent_id)\u001b[0m\n\u001b[1;32m    519\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    520\u001b[0m         \u001b[0;31m# Save the reduce() output and finally memoize the object\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 521\u001b[0;31m         \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0msave_reduce\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mobj\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0mobj\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m*\u001b[0m\u001b[0mrv\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m    522\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    523\u001b[0m     \u001b[0;32mdef\u001b[0m \u001b[0mpersistent_id\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mself\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mobj\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m/anaconda3/lib/python3.6/pickle.py\u001b[0m in \u001b[0;36msave_reduce\u001b[0;34m(self, func, args, state, listitems, dictitems, obj)\u001b[0m\n\u001b[1;32m    632\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    633\u001b[0m         \u001b[0;32mif\u001b[0m \u001b[0mstate\u001b[0m \u001b[0;32mis\u001b[0m \u001b[0;32mnot\u001b[0m \u001b[0;32mNone\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 634\u001b[0;31m             \u001b[0msave\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mstate\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m    635\u001b[0m             \u001b[0mwrite\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mBUILD\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    636\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m/anaconda3/lib/python3.6/site-packages/sklearn/externals/joblib/numpy_pickle.py\u001b[0m in \u001b[0;36msave\u001b[0;34m(self, obj)\u001b[0m\n\u001b[1;32m    279\u001b[0m             \u001b[0;32mreturn\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    280\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 281\u001b[0;31m         \u001b[0;32mreturn\u001b[0m \u001b[0mPickler\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0msave\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mself\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mobj\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m    282\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    283\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m/anaconda3/lib/python3.6/pickle.py\u001b[0m in \u001b[0;36msave\u001b[0;34m(self, obj, save_persistent_id)\u001b[0m\n\u001b[1;32m    474\u001b[0m         \u001b[0mf\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mdispatch\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mget\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mt\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    475\u001b[0m         \u001b[0;32mif\u001b[0m \u001b[0mf\u001b[0m \u001b[0;32mis\u001b[0m \u001b[0;32mnot\u001b[0m \u001b[0;32mNone\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 476\u001b[0;31m             \u001b[0mf\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mself\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mobj\u001b[0m\u001b[0;34m)\u001b[0m \u001b[0;31m# Call unbound method with explicit self\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m    477\u001b[0m             \u001b[0;32mreturn\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    478\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m/anaconda3/lib/python3.6/pickle.py\u001b[0m in \u001b[0;36msave_dict\u001b[0;34m(self, obj)\u001b[0m\n\u001b[1;32m    819\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    820\u001b[0m         \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mmemoize\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mobj\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 821\u001b[0;31m         \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_batch_setitems\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mobj\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mitems\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m    822\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    823\u001b[0m     \u001b[0mdispatch\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0mdict\u001b[0m\u001b[0;34m]\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0msave_dict\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m/anaconda3/lib/python3.6/pickle.py\u001b[0m in \u001b[0;36m_batch_setitems\u001b[0;34m(self, items)\u001b[0m\n\u001b[1;32m    845\u001b[0m                 \u001b[0;32mfor\u001b[0m \u001b[0mk\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mv\u001b[0m \u001b[0;32min\u001b[0m \u001b[0mtmp\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    846\u001b[0m                     \u001b[0msave\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mk\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 847\u001b[0;31m                     \u001b[0msave\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mv\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m    848\u001b[0m                 \u001b[0mwrite\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mSETITEMS\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    849\u001b[0m             \u001b[0;32melif\u001b[0m \u001b[0mn\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m/anaconda3/lib/python3.6/site-packages/sklearn/externals/joblib/numpy_pickle.py\u001b[0m in \u001b[0;36msave\u001b[0;34m(self, obj)\u001b[0m\n\u001b[1;32m    279\u001b[0m             \u001b[0;32mreturn\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    280\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 281\u001b[0;31m         \u001b[0;32mreturn\u001b[0m \u001b[0mPickler\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0msave\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mself\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mobj\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m    282\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    283\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m/anaconda3/lib/python3.6/pickle.py\u001b[0m in \u001b[0;36msave\u001b[0;34m(self, obj, save_persistent_id)\u001b[0m\n\u001b[1;32m    494\u001b[0m             \u001b[0mreduce\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mgetattr\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mobj\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m\"__reduce_ex__\"\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;32mNone\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    495\u001b[0m             \u001b[0;32mif\u001b[0m \u001b[0mreduce\u001b[0m \u001b[0;32mis\u001b[0m \u001b[0;32mnot\u001b[0m \u001b[0;32mNone\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 496\u001b[0;31m                 \u001b[0mrv\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mreduce\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mproto\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m    497\u001b[0m             \u001b[0;32melse\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    498\u001b[0m                 \u001b[0mreduce\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mgetattr\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mobj\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m\"__reduce__\"\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;32mNone\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;31mTypeError\u001b[0m: can't pickle _thread.RLock objects"
     ]
    }
   ],
   "source": [
    "model.save_weights('my_model_weights.h5')\n",
    "joblib.dump(model, 'mlp_balance.pickle')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Random Oversampling"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [],
   "source": [
    "ros = RandomOverSampler(ratio='auto')\n",
    "X_trainRos, y_trainRos = ros.fit_sample(X_train_reshape, y_train)\n",
    "X_testRos, y_testRos = ros.fit_sample(X_test_reshape, y_test)\n",
    "\n",
    "y_trainRosHot = keras.utils.to_categorical(y_trainRos, num_classes = 8)\n",
    "y_testRosHot = keras.utils.to_categorical(y_testRos, num_classes = 8)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "X_train:  (3168, 128, 128, 3)\n",
      "X_trainFlat:  (3168, 49152)\n",
      "X_trainRos Shape:  (17040, 49152)\n",
      "X_testRos Shape:  (7312, 49152)\n",
      "Y_trainRosHot Shape:  (17040, 8)\n",
      "Y_testRosHot Shape:  (7312, 8)\n"
     ]
    }
   ],
   "source": [
    "print(\"X_train: \", X_train.shape)\n",
    "print(\"X_trainFlat: \", X_train_reshape.shape)\n",
    "print(\"X_trainRos Shape: \",X_trainRos.shape)\n",
    "print(\"X_testRos Shape: \",X_testRos.shape)\n",
    "print(\"Y_trainRosHot Shape: \",y_trainRosHot.shape)\n",
    "print(\"Y_testRosHot Shape: \",y_testRosHot.shape)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train on 17040 samples, validate on 7312 samples\n",
      "Epoch 1/15\n",
      "17040/17040 [==============================] - 297s 17ms/step - loss: 13.8273 - acc: 0.1256 - val_loss: 14.1033 - val_acc: 0.1250\n",
      "Epoch 2/15\n",
      "17040/17040 [==============================] - 284s 17ms/step - loss: 13.8537 - acc: 0.1205 - val_loss: 14.1033 - val_acc: 0.1250\n",
      "Epoch 3/15\n",
      "17040/17040 [==============================] - 287s 17ms/step - loss: 13.6912 - acc: 0.1259 - val_loss: 14.1033 - val_acc: 0.1250\n",
      "Epoch 4/15\n",
      "17040/17040 [==============================] - 290s 17ms/step - loss: 13.5288 - acc: 0.1258 - val_loss: 14.1033 - val_acc: 0.1250\n",
      "Epoch 5/15\n",
      "17040/17040 [==============================] - 292s 17ms/step - loss: 13.3521 - acc: 0.1279 - val_loss: 14.1033 - val_acc: 0.1250\n",
      "Epoch 6/15\n",
      "17040/17040 [==============================] - 293s 17ms/step - loss: 12.3657 - acc: 0.1249 - val_loss: 14.1033 - val_acc: 0.1250\n",
      "Epoch 7/15\n",
      "17040/17040 [==============================] - 272s 16ms/step - loss: 12.2570 - acc: 0.1258 - val_loss: 14.1033 - val_acc: 0.1250\n",
      "Epoch 8/15\n",
      "17040/17040 [==============================] - 276s 16ms/step - loss: 12.2251 - acc: 0.1282 - val_loss: 14.1033 - val_acc: 0.1250\n",
      "Epoch 9/15\n",
      "17040/17040 [==============================] - 261s 15ms/step - loss: 12.2680 - acc: 0.1240 - val_loss: 14.1033 - val_acc: 0.1250\n",
      "Epoch 10/15\n",
      "17040/17040 [==============================] - 262s 15ms/step - loss: 12.2450 - acc: 0.1246 - val_loss: 14.1033 - val_acc: 0.1250\n",
      "Epoch 11/15\n",
      "17040/17040 [==============================] - 258s 15ms/step - loss: 12.2693 - acc: 0.1236 - val_loss: 14.1033 - val_acc: 0.1250\n",
      "Epoch 12/15\n",
      "17040/17040 [==============================] - 283s 17ms/step - loss: 12.1734 - acc: 0.1264 - val_loss: 14.1033 - val_acc: 0.1250\n",
      "Epoch 13/15\n",
      "17040/17040 [==============================] - 276s 16ms/step - loss: 12.2408 - acc: 0.1249 - val_loss: 14.1033 - val_acc: 0.1250\n",
      "Epoch 14/15\n",
      "17040/17040 [==============================] - 272s 16ms/step - loss: 12.0620 - acc: 0.1248 - val_loss: 14.1033 - val_acc: 0.1250\n",
      "Epoch 15/15\n",
      "17040/17040 [==============================] - 268s 16ms/step - loss: 11.9705 - acc: 0.1271 - val_loss: 14.1033 - val_acc: 0.1250\n"
     ]
    }
   ],
   "source": [
    "model = Sequential()\n",
    "model.add(Dense(100, input_dim=49152, activation=activation))\n",
    "model.add(Dropout(0.3))\n",
    "model.add(Dense(25, activation=activation))\n",
    "model.add(Dropout(0.3))\n",
    "model.add(Dense(25, activation=activation))\n",
    "model.add(Dropout(0.3))\n",
    "model.add(Dense(25, activation=activation))\n",
    "model.add(Dropout(0.3))\n",
    "model.add(Dense(75, activation=activation))\n",
    "model.add(Dropout(0.3))\n",
    "model.add(Dense(10, activation=activation))\n",
    "model.add(Dropout(0.3))\n",
    "model.add(Dense(8, activation=out_activation))\n",
    "model.compile(loss=loss, optimizer=optimizer, metrics=['accuracy']) \n",
    "# model.load_weights(\"my_model_weights_ros.h5\")\n",
    "history = model.fit(X_trainRos, y_trainRosHot, epochs=15, batch_size=batch_size, verbose=1, \n",
    "                    validation_data=(X_testRos, y_testRosHot), callbacks=[tbcallback])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Saved model to disk\n"
     ]
    }
   ],
   "source": [
    "# serialize model to JSON\n",
    "model_json = model.to_json()\n",
    "with open(\"ros_mlp_model.json\", \"w\") as json_file:\n",
    "    json_file.write(model_json)\n",
    "# serialize weights to HDF5\n",
    "model.save_weights(\"ros_mlp_model.h5\")\n",
    "print(\"Saved model to disk\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "3168/3168 [==============================] - 26s 8ms/step\n",
      "1359/1359 [==============================] - 10s 8ms/step\n",
      "[5.281118400169142, 0.6723484848484849]\n",
      "[5.277816405447202, 0.6725533483131919]\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "results_train = model.evaluate(X_train_reshape, y_trainHot)\n",
    "results_test = model.evaluate(X_test_reshape, y_testHot)\n",
    "print(results_train)\n",
    "print(results_test)\n",
    "val_dict = history.history\n",
    "\n",
    "plt.clf()\n",
    "\n",
    "loss_values = val_dict['loss']\n",
    "val_loss_values = val_dict['val_loss']\n",
    "\n",
    "epochs = range(1, len(loss_values) + 1)\n",
    "plt.plot(epochs, loss_values, 'g', label='Training loss')\n",
    "plt.plot(epochs, val_loss_values, 'g.', label='Validation loss')\n",
    "\n",
    "plt.title('Training & validation loss')\n",
    "plt.xlabel('Epochs')\n",
    "plt.ylabel('Loss')\n",
    "plt.legend()\n",
    "plt.show()\n",
    "\n",
    "plt.clf()\n",
    "\n",
    "acc_values = val_dict['acc'] \n",
    "val_acc_values = val_dict['val_acc']\n",
    "\n",
    "plt.plot(epochs, acc_values, 'r', label='Training acc')\n",
    "plt.plot(epochs, val_acc_values, 'r.', label='Validation acc')\n",
    "plt.title('Training & validation accuracy')\n",
    "plt.xlabel('Epochs')\n",
    "plt.ylabel('Loss')\n",
    "plt.legend()\n",
    "plt.show()"
   ]
  },
  {
   "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": "3.6.5"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}