CNN.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Convolutional Autoencoder in Keras for EEG\n",
    "\n",
    "Autoencoder is a neural network which consists of two parts: \n",
    "* encoder,\n",
    "* decoder, \n",
    "\n",
    "both are neural networks. \n",
    "\n",
    "In the work we create contractive autoencoder. It means that the layer between encoder and decoder has less neurons than an input layer. It is aimed to reduce input data diminution space and delete the noise with valuable information extraction. \n",
    "\n",
    "Decoder transforms reduced representation to the target representation.\n",
    "\n",
    "__Representation of EEG data__"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Using TensorFlow backend.\n"
     ]
    }
   ],
   "source": [
    "from keras.models import Model\n",
    "from keras.layers import Dense, Input\n",
    "import numpy as np\n",
    "import os\n",
    "from sklearn.preprocessing import StandardScaler\n",
    "from sklearn.model_selection import train_test_split\n",
    "import h5py\n",
    "from sklearn.metrics import mean_squared_error"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## CNN AE\n",
    "\n",
    "The encoder will consist in a stack of Conv1D and MaxPooling1D layers (max pooling being used for spatial down-sampling), while the decoder will consist in a stack of Conv1D and UpSampling1D layers.\n",
    "\n",
    "In the proccess different hyperparameters were tuned, such as *number of filters*, *filter lengths*, *length of batch* and others.\n",
    "### Model\n",
    "* Input — (channels_num = 61, window_size = 500, 1)\n",
    "\n",
    "#### Encoder\n",
    "\n",
    "    * Conv1D(nb_filters = 64, filter_length = 5) + ReLU ---- (channels_num = 61, window_size = 500, 64)\n",
    "    * MaxPooling1D ----------------------------------------- (channels_num = 61, window_size = 250, 64)\n",
    "    * Conv1D(nb_filters = 32, filter_length = 5) + ReLU ---- (channels_num = 61, window_size = 250, 32)\n",
    "    * MaxPooling1D ----------------------------------------- (channels_num = 61, window_size = 125, 32)\n",
    "    * Dense + ReLU ----------------------------------------- (channels_num = 61, window_size = 125, 1)\n",
    "\n",
    "#### Decoder\n",
    "\n",
    "    * Conv1D(nb_filters = 32, filter_length = 5) + ReLU ---- (channels_num = 61, window_size = 125, 32)\n",
    "    * UpSampling1D ----------------------------------------- (channels_num = 61, window_size = 250, 32)\n",
    "    * Conv1D(nb_filters = 64, filter_length = 5) + ReLU ---- (channels_num = 61, window_size = 250, 64)\n",
    "    * UpSampling1D ----------------------------------------- (channels_num = 61, window_size = 500, 64)\n",
    "    * Conv1D(nb_filters = 64, filter_length = 5) + Sigmoid - (channels_num = 61, window_size = 500, 1)\n",
    "\n",
    "\n",
    "* Output — (channels_num = 61, window_size = 500, 1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "from keras.layers import Input, Conv1D, Dense, MaxPooling1D, Dropout, UpSampling1D\n",
    "from keras.models import Model\n",
    "from keras.models import Sequential\n",
    "from keras import backend as K\n",
    "\n",
    "\n",
    "def create_ae(dim = 1):\n",
    "\n",
    "    channels_num = 58\n",
    "    nb_filter_1 = 64\n",
    "    nb_filter_2 = 32\n",
    "    filter_length = 5\n",
    "\n",
    "    window_size = 500\n",
    "    encoding_dim = dim\n",
    "\n",
    "    input_ = Input(shape=(window_size, 1))\n",
    "\n",
    "    encoder = Sequential((\n",
    "        Conv1D(nb_filter=nb_filter_1, filter_length=filter_length, activation='relu', padding='same', input_shape=(window_size, 1)),\n",
    "        Dropout(0.4),\n",
    "        MaxPooling1D(),\n",
    "        Conv1D(nb_filter=nb_filter_2, filter_length=filter_length, activation='relu', padding='same'),\n",
    "        MaxPooling1D(),\n",
    "        Dense(encoding_dim, activation='relu')\n",
    "    ))\n",
    "    print(encoder.summary())\n",
    "    decoder = Sequential((\n",
    "        Conv1D(nb_filter=nb_filter_2, filter_length=filter_length, activation='relu', padding='same', input_shape=encoder.output_shape[-2:]),\n",
    "        UpSampling1D(),\n",
    "        Conv1D(nb_filter=nb_filter_1, filter_length=filter_length, padding='same', activation='relu'),\n",
    "        UpSampling1D(),\n",
    "        Conv1D(nb_filter=1, filter_length=filter_length, padding='same', activation='sigmoid')\n",
    "    ))\n",
    "    print(decoder.summary())\n",
    "    autoencoder = Model(input_, decoder(encoder(input_)), name=\"autoencoder\")\n",
    "    autoencoder.compile(loss='mse', optimizer='adam', metrics=['mae']) # .compile(optimizer='adadelta', loss='binary_crossentropy')\n",
    "\n",
    "    print(autoencoder.summary())\n",
    "    return autoencoder"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Batch Generator\n",
    "\n",
    "To fit our model we need to define batch_generator, so that on every step we could choose random 0.5 s interval from the dataset.\n",
    "\n",
    "__ForThreadSafety:__"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "import threading\n",
    "class threadsafe_iter:\n",
    "    def __init__(self, it):\n",
    "        self.it = it\n",
    "        self.lock = threading.Lock()\n",
    "    def __iter__(self):\n",
    "        return self\n",
    "    def __next__(self):\n",
    "        with self.lock:\n",
    "            return next(self.it)\n",
    "\n",
    "def threadsafe_generator(f):\n",
    "    def g(*a, **kw):\n",
    "        return threadsafe_iter(f(*a, **kw))\n",
    "    return g"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "__And now to the generator.__ \n",
    "\n",
    "First we will separate one file for validation test. \n",
    "Then define *batch\\_length (window\\_size)* for our data.\n",
    "\n",
    "At the each step, the file and batch from the file will be chosen randomly. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "32\n",
      "303000\n"
     ]
    }
   ],
   "source": [
    "train_eeg_dir = \"../data/train/\"\n",
    "files_ = [x for x in os.listdir(train_eeg_dir) \n",
    "                 if x[-3:] == \".h5\"]\n",
    "print(len(files_))\n",
    "file_test_ind = np.random.choice(np.arange(len(files_))) * 0 + 0\n",
    "file_test_ = files_[file_test_ind]\n",
    "files_.pop(file_test_ind)\n",
    "\n",
    "\n",
    "batch_length = 0.5 # 0.5 s of data segment\n",
    "random_rate = 0.5\n",
    "\n",
    "h5_file = h5py.File(train_eeg_dir + file_test_, 'r')\n",
    "a_group_key = list(h5_file.keys())[0]\n",
    "raw = np.array(h5_file[a_group_key])\n",
    "if (raw.shape[1] % 500 != 0):\n",
    "    add = 500 - (raw.shape[1] % 500)\n",
    "    raw = np.append(raw, np.zeros((58, add)), axis = 1)\n",
    "\n",
    "iters = []\n",
    "print(raw.shape[1])\n",
    "for i in range(int(raw.shape[1] / 500)):\n",
    "    start, stop = i * 500, (i + 1) * 500\n",
    "    iters.append((start, stop))\n",
    "x_tests = [raw[:, start:stop] for start, stop in iters]\n",
    "\n",
    "@threadsafe_generator\n",
    "def generate_batch():\n",
    "    while True:\n",
    "        cur_files = [[] for _ in range(len(files_))]\n",
    "        batches = [[] for _ in range(len(files_))]\n",
    "        files_count = len(files_)\n",
    "        while files_count > 0:\n",
    "            file_ind = np.random.choice(np.arange(len(files_)))\n",
    "            if cur_files[file_ind] is None: continue\n",
    "            if len(cur_files[file_ind]) == 0:\n",
    "                h5_file = h5py.File(train_eeg_dir + file_test_, 'r')\n",
    "                a_group_key = list(h5_file.keys())[0]\n",
    "                raw = np.array(h5_file[a_group_key])\n",
    "                if (raw.shape[1] % 500 != 0):\n",
    "                    add = 500 - (raw.shape[1] % 500)\n",
    "                    raw = np.append(raw, np.zeros((58, add)), axis = 1)\n",
    "                cur_files[file_ind] = raw\n",
    "                batches[file_ind] = np.arange(int(raw.shape[1] / 500))\n",
    "            begin = np.random.choice(np.arange(len(batches[file_ind])))\n",
    "            \n",
    "            start, stop = batches[file_ind][begin] * 500, (batches[file_ind][begin] + 1) * 500\n",
    "            data = cur_files[file_ind][:, start:stop]\n",
    "            data += random_rate * np.random.normal(loc=0.0, scale=1.0, size=data.shape)\n",
    "            yield (np.expand_dims(data, axis=2).astype('float32')/255., np.expand_dims(data, axis=2).astype('float32')/255.) # add noise later\n",
    "            \n",
    "            batches[file_ind] = np.delete(batches[file_ind], begin)\n",
    "            if len(batches[file_ind]) == 0:\n",
    "                cur_files[file_ind] = None\n",
    "                files_count -= 1"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Let's fit our Model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/opt/conda/anaconda3/lib/python3.6/site-packages/ipykernel_launcher.py:20: UserWarning: Update your `Conv1D` call to the Keras 2 API: `Conv1D(activation=\"relu\", padding=\"same\", input_shape=(500, 1), filters=64, kernel_size=5)`\n",
      "/opt/conda/anaconda3/lib/python3.6/site-packages/ipykernel_launcher.py:23: UserWarning: Update your `Conv1D` call to the Keras 2 API: `Conv1D(activation=\"relu\", padding=\"same\", filters=32, kernel_size=5)`\n",
      "/opt/conda/anaconda3/lib/python3.6/site-packages/ipykernel_launcher.py:29: UserWarning: Update your `Conv1D` call to the Keras 2 API: `Conv1D(activation=\"relu\", padding=\"same\", input_shape=(125, 3), filters=32, kernel_size=5)`\n",
      "/opt/conda/anaconda3/lib/python3.6/site-packages/ipykernel_launcher.py:31: UserWarning: Update your `Conv1D` call to the Keras 2 API: `Conv1D(padding=\"same\", activation=\"relu\", filters=64, kernel_size=5)`\n",
      "/opt/conda/anaconda3/lib/python3.6/site-packages/ipykernel_launcher.py:33: UserWarning: Update your `Conv1D` call to the Keras 2 API: `Conv1D(padding=\"same\", activation=\"sigmoid\", filters=1, kernel_size=5)`\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "_________________________________________________________________\n",
      "Layer (type)                 Output Shape              Param #   \n",
      "=================================================================\n",
      "conv1d_1 (Conv1D)            (None, 500, 64)           384       \n",
      "_________________________________________________________________\n",
      "dropout_1 (Dropout)          (None, 500, 64)           0         \n",
      "_________________________________________________________________\n",
      "max_pooling1d_1 (MaxPooling1 (None, 250, 64)           0         \n",
      "_________________________________________________________________\n",
      "conv1d_2 (Conv1D)            (None, 250, 32)           10272     \n",
      "_________________________________________________________________\n",
      "max_pooling1d_2 (MaxPooling1 (None, 125, 32)           0         \n",
      "_________________________________________________________________\n",
      "dense_1 (Dense)              (None, 125, 3)            99        \n",
      "=================================================================\n",
      "Total params: 10,755\n",
      "Trainable params: 10,755\n",
      "Non-trainable params: 0\n",
      "_________________________________________________________________\n",
      "None\n",
      "_________________________________________________________________\n",
      "Layer (type)                 Output Shape              Param #   \n",
      "=================================================================\n",
      "conv1d_3 (Conv1D)            (None, 125, 32)           512       \n",
      "_________________________________________________________________\n",
      "up_sampling1d_1 (UpSampling1 (None, 250, 32)           0         \n",
      "_________________________________________________________________\n",
      "conv1d_4 (Conv1D)            (None, 250, 64)           10304     \n",
      "_________________________________________________________________\n",
      "up_sampling1d_2 (UpSampling1 (None, 500, 64)           0         \n",
      "_________________________________________________________________\n",
      "conv1d_5 (Conv1D)            (None, 500, 1)            321       \n",
      "=================================================================\n",
      "Total params: 11,137\n",
      "Trainable params: 11,137\n",
      "Non-trainable params: 0\n",
      "_________________________________________________________________\n",
      "None\n",
      "_________________________________________________________________\n",
      "Layer (type)                 Output Shape              Param #   \n",
      "=================================================================\n",
      "input_1 (InputLayer)         (None, 500, 1)            0         \n",
      "_________________________________________________________________\n",
      "sequential_1 (Sequential)    (None, 125, 3)            10755     \n",
      "_________________________________________________________________\n",
      "sequential_2 (Sequential)    (None, 500, 1)            11137     \n",
      "=================================================================\n",
      "Total params: 21,892\n",
      "Trainable params: 21,892\n",
      "Non-trainable params: 0\n",
      "_________________________________________________________________\n",
      "None\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/opt/conda/anaconda3/lib/python3.6/site-packages/ipykernel_launcher.py:8: UserWarning: The semantics of the Keras 2 argument `steps_per_epoch` is not the same as the Keras 1 argument `samples_per_epoch`. `steps_per_epoch` is the number of batches to draw from the generator at each epoch. Basically steps_per_epoch = samples_per_epoch/batch_size. Similarly `nb_val_samples`->`validation_steps` and `val_samples`->`steps` arguments have changed. Update your method calls accordingly.\n",
      "  \n",
      "/opt/conda/anaconda3/lib/python3.6/site-packages/ipykernel_launcher.py:8: UserWarning: Update your `fit_generator` call to the Keras 2 API: `fit_generator(<__main__...., validation_data=(array([[[..., steps_per_epoch=18600, epochs=24)`\n",
      "  \n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/24\n",
      "18600/18600 [==============================] - 569s - loss: 2.1587e-04 - mean_absolute_error: 0.0021 - val_loss: 5.3453e-10 - val_mean_absolute_error: 1.5707e-06\n",
      "Epoch 2/24\n",
      "18600/18600 [==============================] - 483s - loss: 3.8445e-06 - mean_absolute_error: 0.0016 - val_loss: 2.6814e-11 - val_mean_absolute_error: 3.5252e-07\n",
      "Epoch 3/24\n",
      "18600/18600 [==============================] - 483s - loss: 3.8447e-06 - mean_absolute_error: 0.0016 - val_loss: 1.4525e-11 - val_mean_absolute_error: 2.6223e-07\n",
      "Epoch 4/24\n",
      "18600/18600 [==============================] - 483s - loss: 3.8446e-06 - mean_absolute_error: 0.0016 - val_loss: 7.2435e-12 - val_mean_absolute_error: 1.8865e-07\n",
      "Epoch 5/24\n",
      "18600/18600 [==============================] - ETA: 0s - loss: 3.8443e-06 - mean_absolute_error: 0.001 - 481s - loss: 3.8443e-06 - mean_absolute_error: 0.0016 - val_loss: 5.9922e-12 - val_mean_absolute_error: 1.7266e-07\n",
      "Epoch 6/24\n",
      "18600/18600 [==============================] - 485s - loss: 3.8448e-06 - mean_absolute_error: 0.0016 - val_loss: 3.9418e-12 - val_mean_absolute_error: 1.4233e-07\n",
      "Epoch 7/24\n",
      "18600/18600 [==============================] - 486s - loss: 3.8444e-06 - mean_absolute_error: 0.0016 - val_loss: 2.9382e-12 - val_mean_absolute_error: 1.2440e-07\n",
      "Epoch 8/24\n",
      "18600/18600 [==============================] - 483s - loss: 3.8451e-06 - mean_absolute_error: 0.0016 - val_loss: 2.7533e-12 - val_mean_absolute_error: 1.2088e-07\n",
      "Epoch 9/24\n",
      "18600/18600 [==============================] - 484s - loss: 3.8449e-06 - mean_absolute_error: 0.0016 - val_loss: 2.3025e-12 - val_mean_absolute_error: 1.1162e-07\n",
      "Epoch 10/24\n",
      "18600/18600 [==============================] - 484s - loss: 3.8446e-06 - mean_absolute_error: 0.0016 - val_loss: 2.2543e-12 - val_mean_absolute_error: 1.1062e-07\n",
      "Epoch 11/24\n",
      "18600/18600 [==============================] - 473s - loss: 3.8446e-06 - mean_absolute_error: 0.0016 - val_loss: 2.2321e-12 - val_mean_absolute_error: 1.1011e-07\n",
      "Epoch 12/24\n",
      "18600/18600 [==============================] - 487s - loss: 3.8447e-06 - mean_absolute_error: 0.0016 - val_loss: 2.0490e-12 - val_mean_absolute_error: 1.0608e-07\n",
      "Epoch 13/24\n",
      "18600/18600 [==============================] - 485s - loss: 3.8444e-06 - mean_absolute_error: 0.0016 - val_loss: 1.6962e-12 - val_mean_absolute_error: 9.7741e-08\n",
      "Epoch 14/24\n",
      "18600/18600 [==============================] - 483s - loss: 3.8445e-06 - mean_absolute_error: 0.0016 - val_loss: 1.5565e-12 - val_mean_absolute_error: 9.4133e-08\n",
      "Epoch 15/24\n",
      "18600/18600 [==============================] - 485s - loss: 3.8449e-06 - mean_absolute_error: 0.0016 - val_loss: 1.4565e-12 - val_mean_absolute_error: 9.1459e-08\n",
      "Epoch 16/24\n",
      "18600/18600 [==============================] - 485s - loss: 3.8448e-06 - mean_absolute_error: 0.0016 - val_loss: 1.4442e-12 - val_mean_absolute_error: 9.1087e-08\n",
      "Epoch 17/24\n",
      "18600/18600 [==============================] - 486s - loss: 3.8444e-06 - mean_absolute_error: 0.0016 - val_loss: 1.4238e-12 - val_mean_absolute_error: 9.0552e-08\n",
      "Epoch 18/24\n",
      "18600/18600 [==============================] - 484s - loss: 3.8449e-06 - mean_absolute_error: 0.0016 - val_loss: 1.2659e-12 - val_mean_absolute_error: 8.6072e-08\n",
      "Epoch 19/24\n",
      "18600/18600 [==============================] - 483s - loss: 3.8448e-06 - mean_absolute_error: 0.0016 - val_loss: 1.1665e-12 - val_mean_absolute_error: 8.3144e-08\n",
      "Epoch 20/24\n",
      "18600/18600 [==============================] - 487s - loss: 3.8448e-06 - mean_absolute_error: 0.0016 - val_loss: 1.2088e-12 - val_mean_absolute_error: 8.4333e-08\n",
      "Epoch 21/24\n",
      "18600/18600 [==============================] - 483s - loss: 3.8448e-06 - mean_absolute_error: 0.0016 - val_loss: 1.2561e-12 - val_mean_absolute_error: 8.5682e-08\n",
      "Epoch 22/24\n",
      "18600/18600 [==============================] - 482s - loss: 3.8452e-06 - mean_absolute_error: 0.0016 - val_loss: 1.1466e-12 - val_mean_absolute_error: 8.2437e-08\n",
      "Epoch 23/24\n",
      "18600/18600 [==============================] - 485s - loss: 3.8448e-06 - mean_absolute_error: 0.0016 - val_loss: 1.1218e-12 - val_mean_absolute_error: 8.1668e-08\n",
      "Epoch 24/24\n",
      "18600/18600 [==============================] - 483s - loss: 3.8444e-06 - mean_absolute_error: 0.0016 - val_loss: 1.0008e-12 - val_mean_absolute_error: 7.7862e-08\n"
     ]
    }
   ],
   "source": [
    "dims = [3]\n",
    "name = \"Girls_CNN_model_\"\n",
    "for dim in dims:\n",
    "    autoencoder = create_ae(dim)\n",
    "    history = autoencoder.fit_generator(generate_batch(),\n",
    "        samples_per_epoch=600 * len(files_), nb_epoch=24,\n",
    "        validation_data=(np.expand_dims(x_tests[123], axis=2).astype('float32')/255.,\n",
    "        np.expand_dims(x_tests[123], axis=2).astype('float32')/255.))\n",
    "    autoencoder.save(name + str(dim) + \".h5\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(7495.8731155524001, 0.00037771457204455639)"
      ]
     },
     "execution_count": 20,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "loss = np.sum([(autoencoder.predict(np.expand_dims(x, axis=2).astype('float32')/255.)[:, :, 0] * 255. - x)**2 for x in x_tests[0 : -1]])\n",
    "div = np.sum([np.sum(x**2) for x in x_tests])\n",
    "loss / div, div"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "dict_keys(['val_loss', 'val_mean_absolute_error', 'loss', 'mean_absolute_error'])\n"
     ]
    }
   ],
   "source": [
    "print(history.history.keys())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAA4QAAAFNCAYAAABYNqFuAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nrkJggg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f364e9ec080>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "import matplotlib.pyplot as plt\n",
    "plt.figure(figsize=(15, 5))\n",
    "plt.subplot(1, 2, 1)\n",
    "plt.plot(history.history['loss'])\n",
    "plt.title('model loss')\n",
    "plt.ylabel('loss')\n",
    "plt.yscale('log')\n",
    "plt.xlabel('epoch')\n",
    "\n",
    "plt.subplot(1, 2, 2)\n",
    "plt.plot(history.history['val_loss'])\n",
    "plt.title('model validation loss')\n",
    "plt.ylabel('validation loss')\n",
    "plt.xlabel('epoch')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAA5UAAAFNCAYAAACQWku4AAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n/X9TmSlajj2i6gAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f35f8131f98>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.figure(figsize=(15, 5))\n",
    "plt.subplot(1, 2, 1)\n",
    "plt.plot(history.history['mean_absolute_error'])\n",
    "plt.title('model MAE')\n",
    "plt.ylabel('error')\n",
    "plt.xlabel('epoch')\n",
    "\n",
    "plt.subplot(1, 2, 2)\n",
    "plt.plot(history.history['val_mean_absolute_error'])\n",
    "plt.title('model validation MAE')\n",
    "plt.ylabel('validation error')\n",
    "plt.xlabel('epoch')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "import types\n",
    "import tempfile\n",
    "import keras.models\n",
    "import pickle\n",
    "import mne as mn\n",
    "from sklearn.metrics import mean_squared_error"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "def make_keras_picklable():\n",
    "    def __getstate__(self):\n",
    "        model_str = \"\"\n",
    "        with tempfile.NamedTemporaryFile(suffix='.hdf5', delete=True) as fd:\n",
    "            keras.models.save_model(self, fd.name, overwrite=True)\n",
    "            model_str = fd.read()\n",
    "        d = { 'model_str': model_str }\n",
    "        return d\n",
    "\n",
    "    def __setstate__(self, state):\n",
    "        with tempfile.NamedTemporaryFile(suffix='.hdf5', delete=True) as fd:\n",
    "            fd.write(state['model_str'])\n",
    "            fd.flush()\n",
    "            model = keras.models.load_model(fd.name)\n",
    "        self.__dict__ = model.__dict__\n",
    "\n",
    "\n",
    "    cls = keras.models.Model\n",
    "    cls.__getstate__ = __getstate__\n",
    "    cls.__setstate__ = __setstate__"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "make_keras_picklable()\n",
    "pickle.dump(encoder, open(\"encoder.p\", \"wb\"))\n",
    "pickle.dump(decoder, open(\"decoder.p\", \"wb\"))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/opt/conda/anaconda3/lib/python3.6/site-packages/keras/models.py:245: UserWarning: No training configuration found in save file: the model was *not* compiled. Compile it manually.\n",
      "  warnings.warn('No training configuration found in save file: '\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1.953125\n"
     ]
    }
   ],
   "source": [
    "load_encoder = pickle.load(open(\"encoder.p\", \"rb\"))\n",
    "load_decoder = pickle.load(open(\"decoder.p\", \"rb\"))\n",
    "x_reduce = load_encoder.predict(np.expand_dims(x_tests[321], axis=2).astype('float32')/255.)\n",
    "alpha = x_reduce.shape[1] / 64\n",
    "x_pred = load_decoder.predict(x_reduce)\n",
    "mse = mean_squared_error(x_tests[321].astype('float32')/255.,x_tests[321].astype('float32')/255.)\n",
    "score = (1 + mse) * alpha\n",
    "\n",
    "print(score)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(61, 500, 1)\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAA50AAAFNCAYAAACDjQtMAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nXFb5+gAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f35f15577f0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[[  3.10140094e-05   1.01841977e-06   9.11373643e-09 ...,   7.73356224e-09\n",
      "    7.01472118e-07   3.18319217e-05]\n",
      " [  3.10140094e-05   1.01841977e-06   9.11373643e-09 ...,   7.73356224e-09\n",
      "    7.01472118e-07   3.18319217e-05]\n",
      " [  3.10140094e-05   1.01841977e-06   9.11373643e-09 ...,   7.73356224e-09\n",
      "    7.01472118e-07   3.18319217e-05]\n",
      " ..., \n",
      " [  3.10140094e-05   1.01841977e-06   9.11373643e-09 ...,   7.73356224e-09\n",
      "    7.01472118e-07   3.18319217e-05]\n",
      " [  3.10140094e-05   1.01841977e-06   9.11373643e-09 ...,   7.73356224e-09\n",
      "    7.01472118e-07   3.18319217e-05]\n",
      " [  3.10140094e-05   1.01841977e-06   9.11373643e-09 ...,   7.73356224e-09\n",
      "    7.01472118e-07   3.18319217e-05]]\n"
     ]
    }
   ],
   "source": [
    "start, stop = raw.time_as_index([100, 100.5])  # 100 s to 120 s data segment\n",
    "_, times = raw[:, start:stop]\n",
    "\n",
    "plt.figure(figsize=(15, 5))\n",
    "plt.subplot(1,2,1)\n",
    "plt.plot(times, (x_tests[321].astype('float32')/255.).T)\n",
    "plt.xlabel('time (s)')\n",
    "plt.ylabel('EEG data (T)')\n",
    "plt.title('Before AE')\n",
    "\n",
    "print(x_pred.shape)\n",
    "plt.subplot(1,2,2)\n",
    "plt.plot(times, np.squeeze(x_pred, axis=2).T)\n",
    "plt.xlabel('time (s)')\n",
    "plt.ylabel('EEG data (T)')\n",
    "plt.title('After AE')\n",
    "plt.show()\n",
    "\n",
    "print(np.squeeze(x_pred, axis=2))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "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": "3.6.2"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}