nb_model_all.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# All Data Models\n",
    "\n",
    "1. Train the PCA models\n",
    "\n",
    "        This results in pickled PCA models.\n",
    "        \n",
    "2. Train autoencoder models (as replacement for PCA)\n",
    "\n",
    "        This results in saved autoencoder models.\n",
    "        \n",
    "3. Train a k-means cluster that sits on top of PCA & autoencoder\n",
    "\n",
    "        This results in pickled k-means models."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## LOAD"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "task_name = 'all_data'"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "model_path = './models/{0}/'.format(task_name)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "import os\n",
    "import numpy as np\n",
    "import pandas as pd\n",
    "import warnings\n",
    "import pickle\n",
    "\n",
    "from tqdm import tqdm\n",
    "import matplotlib.pyplot as plt"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "save_path = './data/tmp_analysis/{0}'.format(task_name)\n",
    "save_train_path = save_path + '_train.csv'\n",
    "save_test_path = save_path + '_test.csv'\n",
    "save_valid_path = save_path + '_valid.csv'\n",
    "\n",
    "\n",
    "Train = pd.read_csv(save_train_path)\n",
    "Test = pd.read_csv(save_test_path)\n",
    "Valid = pd.read_csv(save_valid_path)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "X = np.asarray(Train.to_numpy()[:,3:]).astype('float32')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "del Train"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [],
   "source": [
    "Xtest = np.asarray(Test.to_numpy()[:,3:]).astype('float32')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [],
   "source": [
    "del Test"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [],
   "source": [
    "Xvalid = np.asarray(Valid.to_numpy()[:,3:]).astype('float32')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [],
   "source": [
    "del Valid"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## PCA Fit\n",
    "I found 32 components to be roughly the inflection curve.\n",
    "I am also creating an 18  component variant just in case we want a direct comparison with action units"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.decomposition import PCA"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[3.02898280e+06 1.48192985e+06 8.50408697e+04 3.55503803e+04\n",
      " 1.42465207e+04 1.00788078e+04 8.03081786e+03 2.93665604e+03\n",
      " 1.74589976e+03 1.00896830e+03 7.85833858e+02 6.95647183e+02\n",
      " 5.76001481e+02 4.85647584e+02 3.70484387e+02 3.13074567e+02\n",
      " 2.23914548e+02 1.90445931e+02 1.61881174e+02 1.55147901e+02\n",
      " 1.42459612e+02 1.08798510e+02 9.41354586e+01 7.94734513e+01\n",
      " 7.24062832e+01 6.16170017e+01 5.54279534e+01 4.92837016e+01\n",
      " 4.52762860e+01 3.86788726e+01 3.17843279e+01 2.72607169e+01]\n"
     ]
    }
   ],
   "source": [
    "pca_n_32 = PCA(n_components=32)\n",
    "face_pca_n_32 = pca_n_32.fit(X)\n",
    "\n",
    "print(face_pca_n_32.explained_variance_)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [],
   "source": [
    "pickle.dump(face_pca_n_32,open(model_path+'pca_n_32.p','wb'))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[3.02898391e+06 1.48192875e+06 8.50409329e+04 3.55503917e+04\n",
      " 1.42465537e+04 1.00788114e+04 8.03082720e+03 2.93666053e+03\n",
      " 1.74590084e+03 1.00896758e+03 7.85834127e+02 6.95647677e+02\n",
      " 5.76001167e+02 4.85647674e+02 3.70484566e+02 3.13074477e+02\n",
      " 2.23914212e+02 1.90445819e+02]\n"
     ]
    }
   ],
   "source": [
    "pca_n_18 = PCA(n_components=18)\n",
    "face_pca_n_18 = pca_n_18.fit(X)\n",
    "\n",
    "print(face_pca_n_18.explained_variance_)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [],
   "source": [
    "pickle.dump(face_pca_n_18,open(model_path+'pca_n_18.p','wb'))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Autoencoder\n",
    "Right now just doing a single layer \n",
    "Perhaps we can try a two layer encoder / two layer decoder next?"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "32 Wide encoder"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [],
   "source": [
    "import keras\n",
    "from keras import layers"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(712824, 136)\n"
     ]
    }
   ],
   "source": [
    "print(X.shape)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [],
   "source": [
    "input_dim = X.shape[1]\n",
    "encoding_dim = 32\n",
    "\n",
    "input_layer = keras.Input(shape=(input_dim,))\n",
    "encoded_layer = layers.Dense(encoding_dim, activation='relu')(input_layer)\n",
    "decoded_layer = layers.Dense(input_dim, activation='relu')(encoded_layer)\n",
    "\n",
    "autoencoder_32 = keras.Model(input_layer,decoded_layer)\n",
    "\n",
    "encoder_32 = keras.Model(input_layer, encoded_layer)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [],
   "source": [
    "autoencoder_32.compile(optimizer='adam',\n",
    "                    loss='binary_crossentropy')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Model: \"model\"\n",
      "_________________________________________________________________\n",
      "Layer (type)                 Output Shape              Param #   \n",
      "=================================================================\n",
      "input_1 (InputLayer)         [(None, 136)]             0         \n",
      "_________________________________________________________________\n",
      "dense (Dense)                (None, 32)                4384      \n",
      "_________________________________________________________________\n",
      "dense_1 (Dense)              (None, 136)               4488      \n",
      "=================================================================\n",
      "Total params: 8,872\n",
      "Trainable params: 8,872\n",
      "Non-trainable params: 0\n",
      "_________________________________________________________________\n"
     ]
    }
   ],
   "source": [
    "autoencoder_32.summary()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/2\n",
      "11138/11138 [==============================] - 27s 2ms/step - loss: -6594.5979 - val_loss: -7894.4731\n",
      "Epoch 2/2\n",
      "11138/11138 [==============================] - 26s 2ms/step - loss: -7895.8183 - val_loss: -7895.5244\n"
     ]
    }
   ],
   "source": [
    "history = autoencoder_32.fit(X, X, epochs=2, batch_size=64, verbose=1, validation_data=(Xtest, Xtest))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYUAAAD4CAYAAAAD6PrjAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjMuMywgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy/Il7ecAAAACXBIWXMAAAsTAAALEwEAmpwYAAAwV0lEQVR4nO3dd3gVZfbA8e9JJ6GmICUIUVCa0iJiXcFCRCUoiGIBXRVXZXdd14JrZV1/a1vbWgFdQVFELKCilBXLqoAB6UUCIgQUQqgJkJDk/P6YSbhJbkjgtpTzeZ55mPvOO/eeTEJOZt6Z94iqYowxxgCEhToAY4wxNYclBWOMMaUsKRhjjCllScEYY0wpSwrGGGNKRYQ6AF8lJiZqu3btQh2GMcbUKgsXLtyuqknl22t9UmjXrh0ZGRmhDsMYY2oVEfnFW7tdPjLGGFPKkoIxxphSlhSMMcaUqvVjCsYYc6QOHjxIVlYWBw4cCHUoARcTE0NycjKRkZHV6m9JwRhT72RlZdGoUSPatWuHiIQ6nIBRVXJycsjKyiIlJaVa+9S4y0cikiYia0QkU0RGhzoeY0zdc+DAARISEup0QgAQERISEo7ojKhGJQURCQdeBC4EOgPDRKRzaKMyxtRFdT0hlDjSr7NGJQWgN5CpqutVtQCYDKQH4oMWvP8sS/87ORBvbYwxtVZNSwqtgU0er7PctjJEZKSIZIhIRnZ29hF/yMGCfOJXTqTd13ewYe3yo4/WGGOOwq5du3jppZeOeL8BAwawa9cu/wfkoaYlhWpR1bGqmqqqqUlJFZ7SrlJkVDSNhr8NAvnvXMuuPXsCEKUxxnhXWVIoLCw87H4zZsygadOmAYrKUdOSwmagjcfrZLfN745p25Ft5z3PicXrWfTqSAqLigPxMcYYU8Ho0aNZt24d3bt355RTTuGss85i4MCBdO7sDKEOGjSIXr160aVLF8aOHVu6X7t27di+fTsbNmygU6dO3HTTTXTp0oULLriA/fv3+yW2mnZL6g9ABxFJwUkGVwJXBerD2p85hFUbvqdf5ng+mvgUg66/O1AfZYypocZ8vIKVW/x7taBzq8Y8dEmXSrc/9thjLF++nMWLF/Pll19y0UUXsXz58tLbRl9//XXi4+PZv38/p5xyCoMHDyYhIaHMe6xdu5Z33nmHcePGMXToUN5//32uueYan2OvUWcKqloIjAJmAquAKaq6IpCf2WnY42xo1Iu0DU/w2ZzZgfwoY4zxqnfv3mWeI3j++efp1q0bffr0YdOmTaxdu7bCPikpKXTv3h2AXr16sWHDBr/EUtPOFFDVGcCMoH1geARtbnqHPc+dRqdvRpGR/DmpHav3kIcxpvY73F/0wRIXF1e6/uWXXzJnzhy+//57YmNjOeecc7w+ZxAdHV26Hh4e7rfLRzXqTCFUwhsfQ+SVE0mWbPZMHsmmnLxQh2SMqcMaNWrE3r17vW7bvXs3zZo1IzY2ltWrVzNv3rygxmZJwdWww5nsPvMB+rGAz8fdR27+4e8CMMaYo5WQkMAZZ5xB165dueuuu8psS0tLo7CwkE6dOjF69Gj69OkT1NhEVYP6gf6Wmpqqfiuyo0r2f66k2S+zeKbVv/jrTdcTFlY/nno0pj5ZtWoVnTp1CnUYQePt6xWRhaqaWr6vnSl4EiHpqnHkxbVhxJYxvDrju1BHZIwxQWVJobyYxjQe/jZNw/bTY8Ff+eTHjaGOyBhjgsaSghfSoithA5+jT9gqfv3wbyzL2h3qkIwxJigsKVQioscw9ncbwU1hHzPxjZfYtqfuF+MwxhhLCofR4JIn2Z90Mg8cfJ4H3/iEAweLQh2SMcYElCWFw4mIpsFVbxETFcmftv+dh6ZmUNvv1jLGmMOxpFCVZm2Junw8ncN+oeeK/2PcN+tDHZExppY72qmzAZ599ln27dvn54gOsaRQHSdcgJ51J1dEfMm6mS8zd/W2UEdkjKnFanJSqHFzH9VU0vdvFG36gUc2vMHV77SnzW1X0755o1CHZYyphTynzj7//PNp3rw5U6ZMIT8/n0svvZQxY8aQl5fH0KFDycrKoqioiAceeICtW7eyZcsW+vbtS2JiInPnzvV7bJYUqissnPDLX4eXz+KZ3Ke5+Y3WTBrVn6axUaGOzBjji89Gw2/L/PueLU6CCx+rdLPn1NmzZs1i6tSpLFiwAFVl4MCBfP3112RnZ9OqVSs+/fRTwJkTqUmTJjz99NPMnTuXxMRE/8bssstHRyIukfArJtBacvhL7jOMmpTBQSvOY4zxwaxZs5g1axY9evSgZ8+erF69mrVr13LSSScxe/Zs7rnnHr755huaNGkSlHjsTOFItemN9H+U8z6/h4wNE/jHJ40Zk9411FEZY47WYf6iDwZV5d577+Xmm2+usG3RokXMmDGD+++/n3PPPZcHH3ww4PHYmcLROPVm6HIpd0dOYc38z3h7vk2FYYypPs+ps/v378/rr79Obm4uAJs3b2bbtm1s2bKF2NhYrrnmGu666y4WLVpUYd9A8CkpiMjlIrJCRIpFJLXctntFJFNE1ohIf4/2NLctU0RGe7SniMh8t/1dEam5F+tFYOC/kcT2vBLzIi9M+5r563NCHZUxppbwnDp79uzZXHXVVZx22mmcdNJJDBkyhL1797Js2TJ69+5N9+7dGTNmDPfffz8AI0eOJC0tjb59+wYkNp+mzhaRTkAx8Cpwp6pmuO2dgXeA3kArYA5wgrvbT8D5QBZOTeZhqrpSRKYAH6jqZBF5BViiqi9XFYNfp84+UttWoeP6say4HTfqg7w/6ne0iY8NTSzGmGqzqbMDNHW2qq5S1TVeNqUDk1U1X1V/BjJxEkRvIFNV16tqATAZSBcRAfoBU939JwCDfIktKJp3Qi55npOLVnJb8VvcNDHDivMYY2q1QI0ptAY2ebzOctsqa08AdqlqYbl2r0RkpIhkiEhGdna2XwM/YidfDqfcxAg+ISX7C/7y7mKKi20qDGNM7VRlUhCROSKy3MuSHowAvVHVsaqaqqqpSUlJoQrjkP6PQutePNdgHGtXLebp2T+FOiJjTBXqyzxmR/p1VnlLqqqedxRxbAbaeLxOdtuopD0HaCoiEe7Zgmf/mi8iGi6fQOSrZ/NOkxfpOzeeE1o0YmC3VqGOzBjjRUxMDDk5OSQkJOBcva6bVJWcnBxiYmKqvU+gnlOYDrwtIk/jDDR3ABYAAnQQkRScX/pXAlepqorIXGAIzjjDCGBagGILjKZtkMHjaPHWEF5pOomb34uhXUIsJyc3DXVkxphykpOTycrKIuSXn4MgJiaG5OTkavf3KSmIyKXAv4Ek4FMRWayq/VV1hXs30UqgELhNVYvcfUYBM4Fw4HVVXeG+3T3AZBH5B/Aj8JovsYVE+/OQ393DOV89xvUxHRg5MZrpo86geePqZ2ljTOBFRkaSkpIS6jBqJJ9uSa0JQnpLqjfFRTBpCMU//4+hhWMoPKYbk0f2ISYyPNSRGWNMqYDckmq8CAuHy8YT1rA5Exu+wM+bsvjbB8vqzaCWMaZ2s6QQCHEJMHQisQe28VGriXz44ybGfm3FeYwxNZ8lhUBJ7gVp/yRlx/94rvVcHvt8NV+s3hrqqIwx5rAsKQTSKTdC1yFcsuM/DEtcz5/eWczarYGbyMoYY3xlSSGQROCS55CEDjxS+AxtInZx48QMdu0rCHVkxhjjlSWFQItuCFe8SXjhAd5LeJXsXbnc9vYiK85jjKmRLCkEQ9KJkP5vGm5byLQTZvJtZg7/+GRlqKMyxpgKLCkES9fBcOof6PDzm/yry3omfP+LFecxxtQ4lhSC6fxHILk3l216jCuPO8CD05Yzz4rzGGNqEEsKwRQRBZe/gURE82jBE5wYH8Ytby1k0459oY7MGGMASwrB16Q1DB5P+PY1vNt6CkXFxdw4wYrzGGNqBksKoXB8P+j7NxqueZ8PTllNZnauFecxxtQIlhRC5aw7of35tF/0KM+eWcTslVutOI8xJuQsKYRKWBhcNhYatuDiNaP5fY/GvDA3k+lLtoQ6MmNMPWZJIZRi42HoBCR3G/cXPMupbZty13tLWJq1K9SRGWPqKUsKoda6J6Q9Rti6Ofyn/VckNoxm5MSFbNtzINSRGWPqIZ+Sgog8KSKrRWSpiHwoIk09tt0rIpkiskZE+nu0p7ltmSIy2qM9RUTmu+3vikiUL7HVKqm/h5OvIPbbJ3jn3H3sOXCQm95cyIGDRaGOzBhTz/h6pjAb6KqqJwM/AfcCiEhnnPrLXYA04CURCReRcOBF4EKgMzDM7QvwOPCMqrYHdgI3+Bhb7SECFz8DSR05du6feOniY1iyaRf3WnEeY0yQ+ZQUVHWWqpbcYD8PKKkOnQ5MVtV8Vf0ZyAR6u0umqq5X1QJgMpAuIgL0A6a6+08ABvkSW60TFQdXvAmFBZyz5E7uPDeFD3/cbMV5jDFB5c8xhd8Dn7nrrYFNHtuy3LbK2hOAXR4JpqTdKxEZKSIZIpKRnZ3tp/BrgMQOkP4CbM7gtoP/4aKTW1pxHmNMUFWZFERkjogs97Kke/S5DygEJgUy2BKqOlZVU1U1NSkpKRgfGTxdBkGf25AFY3mmUyZdWjW24jzGmKCpMimo6nmq2tXLMg1ARK4DLgau1kMXwDcDbTzeJtltq6w9B2gqIhHl2uun88dAmz5Ezbid1y9qTExkODdOzGBnnhXnMcYElq93H6UBdwMDVdVzVrfpwJUiEi0iKUAHYAHwA9DBvdMoCmcwerqbTOYCQ9z9RwDTfImtVguPhMv/A1GxNJ9xE+OHdeTXXQesOI8xJuB8HVN4AWgEzBaRxSLyCoCqrgCmACuBz4HbVLXIHTMYBcwEVgFT3L4A9wB3iEgmzhjDaz7GVrs1bgWDX4OctXT/8SH+79KufLcuh0esOI8xJoAiqu5SOff20cq2PQo86qV9BjDDS/t6nLuTTInjfgd974MvHmFIm1NZc9aZjPvmZ05s0YirT20b6uiMMXWQPdFc0515B5yQBjP/xuiTcjnnxCQemrbCivMYYwLCkkJNFxYGl74CjVsSPvV6/p1+LG0TYq04jzEmICwp1AYNmsHQNyEvm0af3sL4a3tSrFhxHmOM31lSqC1adYcBT8C6L0hZ8SIvXtXTivMYY/zOkkJt0nMEdLsKvnqcM1nMAxd1YvbKrfxr9ppQR2aMqSMsKdQmInDRv+CYLvDBjYzoHMaw3m14ce46pi2uv8/6GWP8x5JCbRMVC0MnQnER8t51jBnQgd7t4rl76lIrzmOM8Zklhdoo4XgY9BJsWUTUf+/n5Wt6WnEeY4xfWFKorTpdAqf/EX4YT8L66YwfkWrFeYwxPrOkUJud+zAcezp8/Cc6hW/m6aHdrTiPMcYnlhRqs/AId+K8hvDutaR1iOOv55/Ahz9u5lUrzmOMOQqWFGq7Ri1gyOuwYx1M/yOj+h7PxSe35PHPV/PfVVacxxhzZCwp1AUpZ8G5D8KKD5EFr/LkkG50adWYP09ezE9WnMcYcwQsKdQVZ9wOJw6AWffT4LcMxl6b6hTnmWDFeYwx1WdJoa4QgUEvQ5NkeO86WkXkMnZ4L37bbcV5jDHVZ0mhLmnQ1Jk4b/8OeP8GeiY35p+XnWTFeYwx1eZrOc5HRGSpW3Vtloi0cttFRJ4XkUx3e0+PfUaIyFp3GeHR3ktElrn7PC8i4kts9VbLk2HAU/DzV/DlPxncK5mRZx/HxO9/YdL8X0IdnTGmhvP1TOFJVT1ZVbsDnwAPuu0X4tRl7gCMBF4GEJF44CHgVJwqaw+JSDN3n5eBmzz2S/Mxtvqr57XQ4xr4+kn4aSb3pHWkr1uc5/t1VpzHGFM5n5KCqu7xeBkHlDwxlQ5MVMc8oKmItAT6A7NVdYeq7gRmA2nutsaqOk+dp64mAoN8ia3eG/AUtDgJPhhJ+O6NPDesB20TYrl10kI25lhxHmOMdz6PKYjIoyKyCbiaQ2cKrYFNHt2y3LbDtWd5aa/sM0eKSIaIZGRnZ/v6JdRNkQ2cifNUYcpwGocXMX7EKRQr3DTRivMYY7yrMimIyBwRWe5lSQdQ1ftUtQ0wCRgV6IDdzxyrqqmqmpqUlBSMj6yd4o+DS1+GXxfD56NJSYwrLc5z+2QrzmOMqajKpKCq56lqVy/LtHJdJwGD3fXNQBuPbclu2+Hak720G191vMh5hmHhf2DJZM7skMgDF3ViziorzmOMqcjXu486eLxMB1a769OB4e5dSH2A3ar6KzATuEBEmrkDzBcAM91te0Skj3vX0XCgfNIxR6vfA9DuLPj4dti6ghGnt7PiPMYYr3wdU3jMvZS0FOcX/J/d9hnAeiATGAfcCqCqO4BHgB/c5e9uG26f8e4+64DPfIzNlAiPgMGvQUwTePdaJH8vYwZ2pXeKU5xnyaZdoY7QGFNDSG2fYjk1NVUzMjJCHUbt8Mt38MbF0HEADH2TnLwCBr7wLYXFxUwfdSbHNI4JdYTGmCARkYWqmlq+3Z5ork/ang7nPQyrPobvXyShYTTjR6Sy90AhI604jzEGSwr1z+l/hI4Xw+wH4Zfv6dSycWlxntHvL7XiPMbUc5YU6hsRp75zs7bw3nWQu420ri346/kn8NHiLVacx5h6zpJCfRTTxJk478BumPp7KCpkVL/2VpzHGGNJod5q0RUufho2fANzH0VEeHJIN7q2amLFeYypxywp1Gfdr4KeI+B/T8Oaz2gQFc7Y4b1oEGXFeYyprywp1HcXPgEtu8EHN8OOn2nZpAGvXusU57l1khXnMaa+saRQ30XGOBPnCTBlOBw8QM9jm/HPy07i+/U5/P1jK85jTH1iScFAs3Zw6Vj4bSl8dhdAaXGeN+f9wlvzrDiPMfWFJQXjODENzvorLJoIP04CKC3O8/B0K85jTH1hScEc0vc+SDkbPr0DfltGeJjw3LAetEuMs+I8xtQTlhTMIWHhMPh1aNAM3r0W9u+icUwk44enWnEeY+oJSwqmrIZJcPkbsHsTTLsNVGmXGMdLV1txHmPqA0sKpqJj+8D5f4fVn8B3zwNwRvtDxXmemmXFeYypqywpGO/63Aqd02HOGNjwLUBpcZ6XvrTiPMbUVZYUjHciMPAFiE+BqdfD3t8QESvOY0wd55ekICJ/FREVkUT3tYjI8yKSKSJLRaSnR98RIrLWXUZ4tPcSkWXuPs+7ZTlNKMU0dibOy99bOnFeVEQYL1/dk6RG0Yx8M4Otew6EOkpjjB/5nBREpA1OKc6NHs0XAh3cZSTwsts3HngIOBXoDTzk1mrG7XOTx35pvsZm/OCYznDxs/DLt/DF3wFIaBjNuOFWnMeYusgfZwrPAHcDnrekpAMT1TEPaCoiLYH+wGxV3aGqO4HZQJq7rbGqzlOnystEYJAfYjP+0O0KSP09fPscrPoEgE4tG/PMFVacx5i6xqekICLpwGZVXVJuU2tgk8frLLftcO1ZXtor+9yRIpIhIhnZ2dk+fAWm2tIeg1Y94KNbIGcdAP27tODOC5ziPK98ZcV5jKkLqkwKIjJHRJZ7WdKBvwEPBj7MslR1rKqmqmpqUlJSsD++foqIhssngITBlBFwcD8At/V1ivM8MXM1c1ZacR5jarsqk4KqnqeqXcsvwHogBVgiIhuAZGCRiLQANgNtPN4m2W07XHuyl3ZTkzRrC4PHw9bl8OmdAOWK8/xoxXmMqeWO+vKRqi5T1eaq2k5V2+Fc8umpqr8B04Hh7l1IfYDdqvorMBO4QESauQPMFwAz3W17RKSPe9fRcGCaj1+bCYQO58PZd8Hit5zJ86C0OE9sdIQV5zGmlgvUcwozcM4kMoFxwK0AqroDeAT4wV3+7rbh9hnv7rMO+CxAsRlfnTMajuvrnC386gwnlRbn2WPFeYypzaS23zWSmpqqGRkZoQ6j/snbDq+eDeGRMPJLZxI94INFWdwxZQnX9mnLI4O6hjZGY0ylRGShqqaWb7cnms3RiUt0J87Lgg9vgWLnzOCynsncbMV5jKm1LCmYo9emN1zwKPz0GXz7bGnz3Vacx5hay5KC8c2pN0OXS+GLR+DnrwHKFOe5xYrzGFOrWFIwvhGBgf+GhPbO/Eh7fgUoLc6jCjdO/IG9Bw6GOFBjTHVYUjC+i27kTJxXsM+ZUbXISQAlxXnWZefxl3etOI8xtYElBeMfzTvCwOdh4/cw5+HS5jPaJ/LgxZ2Zs2qbFecxphaICHUApg45aQhsnAffv+AMQndOB2D4aW1Z/dteXvpyHSe2aER690qntTLGhJidKRj/6v8otO4FH90G2zMB3OI8Xaw4jzG1gCUF418lE+eFR8KU4c44A5QpznPTRCvOY0xNZUnB+F/TNjB4HGxbCZ/eAe5T8yXFeXLzCxk5McOK8xhTA1lSMIHR/jxnjqQl78DCN0qbS4vzZO224jzG1ECWFEzgnH03HH8ufHY3bF5U2mzFeYypuSwpmMAJC4PLxkFcc6cwz74dpZtu69ueS7q1suI8xtQwlhRMYMUlwNCJsPdX+PDm0onzRIQnBp9sxXmMqWEsKZjAS+4Faf+EtbPgf/8qbbbiPMbUPJYUTHCcciN0HQJz/w/Wf1na7Fmc55ZJC604jzEh5lNSEJGHRWSziCx2lwEe2+4VkUwRWSMi/T3a09y2TBEZ7dGeIiLz3fZ3RSTKl9hMDSMClzwHiSfA1Btgz5bSTT2PbcZjl53EvPU7GPPxihAGaYzxx5nCM6ra3V1mAIhIZ+BKoAuQBrwkIuEiEg68CFwIdAaGuX0BHnffqz2wE7jBD7GZmiS6oTNxXuEBeO+60onz4FBxnrfmbeRNK85jTMgE6vJROjBZVfNV9Wecusu93SVTVderagEwGUgXEQH6AVPd/ScAgwIUmwmlpBOcqbY3zYfZD5bZdHdaR/p1bM4YK85jTMj4IymMEpGlIvK6iDRz21oDmzz6ZLltlbUnALtUtbBcu1ciMlJEMkQkIzs72w9fggmqrpfBqX+AeS/B8g9Km8PDhOeu7G7FeYwJoSqTgojMEZHlXpZ04GXgeKA78Cvwr8O9l7+o6lhVTVXV1KSkpGB8pPG38x+B5N4w/Y+Q/VNpcyMrzmNMSFWZFFT1PFXt6mWZpqpbVbVIVYuBcTiXhwA2A2083ibZbausPQdoKiIR5dpNXRURBZe/4UygN2U4FOSVbipfnKfIivMYEzS+3n3U0uPlpcByd306cKWIRItICtABWAD8AHRw7zSKwhmMnq7OBDhzgSHu/iOAab7EZmqBJq1h8GuQvRo+vr104jyw4jzGhIqvYwpPiMgyEVkK9AX+AqCqK4ApwErgc+A294yiEBgFzARWAVPcvgD3AHeISCbOGMNrPsZmaoPj+0Lf+2DZFMgo+y0fflpbhvU+lpe/XMdHP9qJozHBILV9lsrU1FTNyMgIdRjGF8XF8M4VzkNt13/uPAHtKigs5prX5rN40y7eu/k0urVpGrIwjalLRGShqqaWb7cnmk3ohYXBpa9CwxbwXtmJ86Iiwnjlml40t+I8xgSFJQVTM8TGw9AJkLsVPripdOI8gPi4KMaPSCXPivMYE3CWFEzN0bonpD0GmXPg6yfLbOrY4lBxnnusOI8xAWNJwdQsqb+Hk6+AL/8Jmf8ts+kCtzjPtMVbePmrdSEK0Ji6zZKCqVlE4OJnoHkneP9G2J1VZnNJcZ4nZ66x4jzGBIAlBVPzRMU5E+cVHXQqthUeqrNQvjjPmt+sOI8x/mRJwdRMie0h/QXYnAGz7iuzqUFUOOOGpzrFeSb+YMV5jPEjSwqm5uoyCPrcBgvGwrKpZTa1aBLD2Gt7sXVPvhXnMcaPLCmYmu38MdCmD0z/E2xbXWZTDyvOY4zfWVIwNVt4JFz+H4iKdSbOy88ts9mK8xjjX5YUTM3XuJUzcV7OWvj4T2UmzoNDxXkenr6C79ZtD1GQxtQNlhRM7XDc76Df/bD8fVgwrsymkuI8KYlx3DppEb/k5FXyJsaYqlhSMLXHGX+BE9Jg5t9g0w9lNpUpzjMhw4rzGHOULCmY2iMsDC59xbmc9N4IyCt7qahdYhwvX92T9dutOI8xR8uSgqldGjSDoROdhPD+jVBcdnK809sn8tAlVpzHmKNlScHUPq26w4AnYP1c+OrxCpuv7dOWq0614jzGHA2fk4KI/FFEVovIChF5wqP9XhHJFJE1ItLfoz3NbcsUkdEe7SkiMt9tf9ct12mMdz1HQLer4KsnYO2cMptEhIcv6ULvlHjufn8pizftCk2MxtRCvtZo7gukA91UtQvwlNveGaf+chcgDXhJRMJFJBx4EbgQ6AwMc/sCPA48o6rtgZ3ADb7EZuo4EbjoX3BMF/jgRti1scxmz+I8Iydm8NtuK85jTHX4eqZwC/CYquYDqOo2tz0dmKyq+ar6M5AJ9HaXTFVdr6oFwGQgXUQE6AeUzGUwARjkY2ymrouKdcYXiovcifPyy2wuU5znTSvOY0x1+JoUTgDOci/7fCUip7jtrYFNHv2y3LbK2hOAXapaWK7dKxEZKSIZIpKRnZ3t45dgarWE42HQS7BlEXx+b4XNJcV5lm224jzGVEeVSUFE5ojIci9LOhABxAN9gLuAKe5f/QGlqmNVNVVVU5OSkgL9caam63QJnP5HyHgNlk6psNkpznOiFecxphoiquqgqudVtk1EbgE+UOfPrwUiUgwkApuBNh5dk902KmnPAZqKSIR7tuDZ35iqnfswZC2Ej/8MLU5yivR4uPWc41n9216enLmGE5o34rzOx4QmTmNqOF8vH30E9AUQkROAKGA7MB24UkSiRSQF6AAsAH4AOrh3GkXhDEZPd5PKXGCI+74jgGk+xmbqk/AId+K8hvDutZBftviOFecxpnp8TQqvA8eJyHKcQeMR6lgBTAFWAp8Dt6lqkXsWMAqYCawCprh9Ae4B7hCRTJwxhtd8jM3UN41aOIlhx3qYNqrCxHnli/PssOI8xlQgtX3gLTU1VTMyMkIdhqlJ/vcszHkI0h6DPrdU2Pzjxp1cMXYePY9typs3nEpkuD3DaeofEVmoqqnl2+1/g6l7zvgznHgRzLofNs6rsLnHsc14fLBTnOfh6VacxxhPlhRM3SPi3KbapA28dx3kVrxt+dIeydz8u+OYNN+K8xjjyZKCqZsaNHUebNu/E96/ocLEeQB397fiPMaUZ0nB1F0tT4YBT8HPX8Hc/6uw2YrzGFORJQVTt/W8FnpcC988BT/NrLDZivMYU5YlBVP3DXjSeaDtg5Gws+L4gWdxntsnW3EeU79ZUjB1X2QDGPqm89zClOFwsOKMqSXFef67ehtPzrTiPKb+sqRg6of4FKeU56+L4fN7vHYpKc7zylfr+PDHrODGZ0wNYUnB1B8dB8AZt8PCN2DxOxU2iwhjBnbh1JR47nl/mRXnMfWSJQVTv/R7ANqdBZ/8BbZWfHAtMjyMl604j6nHLCmY+iU8Aga/BjFNnInzDuyu0MWK85j6zJKCqX8aHeNMnLdzA0y7rcLEeVC2OM/dU604j6k/LCmY+qnt6XD+GFj1MXz/otcuJcV5pi/ZwktfWnEeUz9YUjD112mjnKptsx+EX77z2uXWc45nYLdWPDVrDbNXbg1ygMYEnyUFU3+JQPqL0KwtvHc97K34S19EeGLIyZzUugm3W3EeUw9YUjD1W0wT58G2A7udifOKCit2iQxn7LWpxFlxHlMP+JQURORdEVnsLhtEZLHHtntFJFNE1ohIf4/2NLctU0RGe7SniMh8t/1dt1ynMYHXoitc/DRs+Abm/sN7lyYxvHptL7buyefWSQs5WFQc5CCNCQ6fkoKqXqGq3VW1O/A+8AGAiHTGqb/cBUgDXhKRcBEJB14ELgQ6A8PcvgCPA8+oantgJ3CDL7EZc0S6XwW9roP/PQOrZ3jtYsV5TH3gl8tHIiLAUKDkMdF0YLKq5qvqz0Am0NtdMlV1vaoW4NR1Tnf37wdMdfefAAzyR2zGVFva49CyG3z4B9jxs9cuZYrzfL8huPEZEwT+GlM4C9iqqmvd162BTR7bs9y2ytoTgF2qWliu3SsRGSkiGSKSkZ1dsaqWMUclMsYpzCMCU66Fg/u9distzvPxSr7LtOI8pm6pMimIyBwRWe5lSffoNoxDZwkBp6pjVTVVVVOTkpKC9bGmPmjWDi4bC78tgxl3ee1SUpznuMQ4bn3bivOYuqXKpKCq56lqVy/LNAARiQAuA9712G0z0MbjdbLbVll7DtDUfS/PdmOC74T+cNZf4cc34ce3vHZpFBPJ+BGpgBXnMXWLPy4fnQesVlXPuYanA1eKSLSIpAAdgAXAD0AH906jKJzB6OnqzCEwFxji7j8CmOaH2Iw5On3vg5Sz4dO/wq9LvXZpmxDHS1dZcR5Tt/gjKVxJuUtHqroCmAKsBD4HblPVInfMYBQwE1gFTHH7AtwD3CEimThjDK/5ITZjjk5YOAx+HRo0cwrz7N/ltZsV5zF1jdT2ib5SU1M1IyMj1GGYumrjPHjjIjghDa54yxmELkdVue+j5bw9fyPPXNGNS3skhyBQY46MiCxU1dTy7fZEszGHc2wfOP8RWP0JfPe81y7li/P8uHFnkIM0xn8sKRhTlT63QOdBMOdh2PA/r11KivMc0ziam99caMV5TK1lScGYqojAwH9D/HHuxHm/ee0WHxfF+OGnWHEeU6vZmIIx1bV1JYw/F1r1gOHTnSpuXsxeuZWRb2bQuWVjOrZoTELDKBLiokhoGO3+e2g9JjI8yF+EMY7KxhQsKRhzJJa8Cx+OhNP/BBc8Umm3dxZs5O35G9mRV0B2bj4Fhd4n0IuLCncSREniiIsukzScdqctPi6KyHA7uTf+UVlS8P6njjHGu25XwKb5zqBzm1Oh08Veuw3rfSzDeh8LOHcn5RUUkZObT05eATm5BWXX8/LJyS1g864DLM3azY68AgoreeahSYPICsmi9CzETRyJbkJpGhtFeFjFu6WMORxLCsYcqbR/wpYf4aNboHknSDj+sN1FhIbRETSMjqBtQlyVb6+q7NlfyHY3WVRIIHlO27rsXBZsKGDnvgJvZaYJE2gWeyiBxDeMItFNIE7y8DgjiYumcYMIxMstt6Z+sctHxhyNXRvh1bOhcWu4YTZExYYslKJiZee+smcdObn57MgrYLubQHJyC5zXufnsOVCxkBBAZLgQHxdFfFy0kzDKJ5DSxOKclcRGhVsSqcXs8pEx/tT0WLhsHEy6HGbc6ZT1DNEvyPAwIbFhNIkNo4FGVfYvKCxm5z4nQZRJJHkeySS3gA05eeTkFrCvwPtdVDGRYWUuYZUmk4bOeoJHAom3QfVaw5KCMUerw/lw9l3w9RPO+EKvEaGOqFqiIsI4pnEMxzSOqVb//QVFpYmj5GwjJ89jPdcZTF/z21625xVUOqjeMDqiNEEklEsgpWci7llJMxtUDxlLCsb44pzRkPWDM812y27QqnuoI/K7BlHhJEfFktys6ktkqkpufmHp2UaOlwSyI6+ArJ37WJK1ix15BZVOJNg0NtJJEh5nGwkNnQRSNrFE07RBJGE2qO4XNqZgjK/ytjvjC2ERcPNXziR6plqKi5U9Bw6WJpDKxkFKEsvhBtVLEkW8e3dWyV1Y8Q0PJZCSxNI4xgbVbUzBmECJS4TLJ8B/LoQPb4Er34Ywu/RRHWFhQtNY5/bZ9s0bVtm/sKiYnfsOkpOXz47cQwnE88xkR14BK7bsYXtuPnsPM6hePoGUrpdPLA2jiI2qP78q689XakwgtTkF+j8Kn90N3z7jFOkxfhcRHkZSo2iSGkVXq39+YRE78w56nG04ZyDlz0x+3p7HjryqB9U9zza8JRDn7q3aPahuScEYf+k90nmw7Yt/QPIpTpEeE1LREeG0aBJOiybVG1TfV1BYetkqJy/fTR4eycR9Qn31b3vJyS2goMj7oHqj6Aj3slXFcZAyDx42jCI+NoqIGjSobknBGH8RgUued+o7T/093Pw1NG4V6qjMEYiNiiA2PoI28dUbVN+bX8iO3EMJZId7OWu7e4vvjrx8Nu3Yx48bd7Fz3+EH1SvMj1V6ZuKelbjrgR5U9ykpiEh34BUgBigEblXVBeKM4DwHDAD2Adep6iJ3nxHA/e5b/ENVJ7jtvYA3gAbADODPWttHwU39E90Qhr4J4/rBC6dAZIAeagvYIGkA3rc2xQrVjleAxu7SrqrOsaCxgqpSpEpxsVKsUKxKUcn6fqU4z92uTlt5+4EDQJgIYQKxdy6hYVzVYzFHwtczhSeAMar6mYgMcF+fA1yIU5e5A3Aq8DJwqojEAw8BqYACC0VkuqrudPvcBMzHSQppwGc+xmdM8DXvCFdPgWVTA/DmAfo7KWB/f1m8JW8rOEt1LxQVq5JfWER+YTH5hcUcOOiuHyx22g8Wc1pUlN9D9TUpKE6iBGgCbHHX04GJ7l/680SkqYi0xEkYs1V1B4CIzAbSRORLoLGqznPbJwKDsKRgaqt2ZzqLMUcpDOeySYMgf66vSeF2YKaIPIXzNZzutrcGNnn0y3LbDtee5aXdKxEZCYwEOPbYY336AowxxhxSZVIQkTlACy+b7gPOBf6iqu+LyFDgNeA8/4ZYkaqOBcaC8/BaoD/PGGPqiyqTgqpW+kvevczzZ/fle8B4d30z0Maja7LbthnnEpJn+5due7KX/sYYY4LI15tjtwC/c9f7AWvd9enAcHH0AXar6q/ATOACEWkmIs2AC4CZ7rY9ItLHvXNpODDNx9iMMcYcIV/HFG4CnhORCJw7pUa67TNwbkfNxLkl9XoAVd0hIo8AP7j9/l4y6AzcyqFbUj/DBpmNMSbobEI8Y4yphyqbEK/mPFttjDEm5CwpGGOMKVXrLx+JSDbwy1Hunghs92M4/mJxHRmL68hYXEemrsbVVlWTyjfW+qTgCxHJ8HZNLdQsriNjcR0Zi+vI1Le47PKRMcaYUpYUjDHGlKrvSWFsqAOohMV1ZCyuI2NxHZl6FVe9HlMwxhhTVn0/UzDGGOPBkoIxxphSdTYpiEiaiKwRkUwRGe1le7SIvOtuny8i7Ty23eu2rxGR/kGM6Q4RWSkiS0XkvyLS1mNbkYgsdpfp/orpCGK7TkSyPWK40WPbCBFZ6y4jghzXMx4x/SQiuzy2BeSYicjrIrJNRJZXsl1E5Hk35qUi0tNjWyCPVVVxXe3Gs0xEvhORbh7bNrjti0XEr/PGVCOuc0Rkt8f36kGPbYf9/gc4rrs8Ylru/jzFu9sCebzaiMhc93fBChH5s5c+gfsZU9U6twDhwDrgOCAKWAJ0LtfnVuAVd/1K4F13vbPbPxpIcd8nPEgx9QVi3fVbSmJyX+eG+HhdB7zgZd94YL37bzN3vVmw4irX/4/A64E+ZsDZQE9geSXbB+BM6ChAH2B+oI9VNeM6veTzcErmzvfYtgFIDNHxOgf4xNfvv7/jKtf3EuCLIB2vlkBPd70R8JOX/48B+xmrq2cKvYFMVV2vqgXAZJwSoZ7SgQnu+lTgXBERt32yquar6s84M732DkZMqjpXVfe5L+dRtsZEIFXneFWmP26JVXVqbc/Gqa8diriGAe/46bMrpapfAzsO06W0HK06JWZLytEG8lhVGZeqfud+LgTx56sax6syvvxc+juuoPxsAajqr6q6yF3fC6yiYiXKgP2M1dWkUFnZT699VLUQ2A0kVHPfQMXk6QbKTh8eIyIZIjJPRAb5IZ6jiW2we6o6VURKiigF6ngd0Xu7l9pSgC88mgN5zA7nSMvRhkL5ny8FZonIQnHK3QbbaSKyREQ+E5EubluNOF4iEovzi/V9j+agHC9xLmv3AOaX2xSwnzFf6ymYABCRa4BUDhUwAmeeks0ichzwhYgsU9V1QQzrY+AdVc0XkZtxzrL6BfHzq3IlMFVVizzaQn3MaiQR6YuTFM70aD7TPVbNgdkistr9SzoYFuF8r3JFZADwEdAhSJ9dHZcA3+qh2i8QhOMlIg1xEtHtqrrHn+99OHX1TKGycqBe+4hTJKgJkFPNfQMVEyJyHk7964Gqml/Srqqb3X/X45Qw7eGHmKodm6rmeMQzHuhV3X0DGZeHKyl3eh/gY3Y4hytHG6hjVS0icjLO9y9dVXNK2j2O1TbgQ/xzybRaVHWPqua66zOASBFJpAYcL9fhfrYCcrxEJBInIUxS1Q+8dAncz1ggBkpCveCcAa3HuZxQMkDVpVyf2yg70DzFXe9C2YHm9fhnoLk6MfXAGVjrUK69GRDtrifilD3154BbdWJr6bF+KTBPDw1s/ezG2Mxdjw9WXG6/jjgDfxLEY9aOygdOL6LsIOCCQB+rasZ1LM4Y2enl2uOARh7r3wFpQYyrRcn3DueX60b32FXr+x+ouNztTXDGHeKCdbzcr30i8Oxh+gTsZ8xvB7emLTij8z/h/JK9z237O85f4AAxwHvuf5IFwHEe+97n7rcGuDCIMc0BtgKL3WW62346sMz9T7EMuCEEx+ufwAo3hrlAR499f+8ex0zg+mDG5b5+GHis3H4BO2Y4fzX+ChzEuWZ7A/AH4A/udgFedGNeBqQG6VhVFdd4YKfHz1eG236ce5yWuN/j+4Ic1yiPn615eCQtb9//YMXl9rkO58YTz/0CfbzOxBmzWOrxvRoQrJ8xm+bCGGNMqbo6pmCMMeYoWFIwxhhTypKCMcaYUpYUjDHGlLKkYIwxppQlBWOMMaUsKRhjjCn1/5aayxHf4jqfAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig, ax = plt.subplots()\n",
    "\n",
    "plt.plot([0]+history.history['loss'], label='train')\n",
    "plt.plot([0]+history.history['val_loss'], label='test')\n",
    "plt.legend()\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [],
   "source": [
    "predictions = autoencoder_32.predict(Xvalid)\n",
    "# TODO I should probably do something with the validation set"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [],
   "source": [
    "autoencoder_32.save(model_path + 'autoencoder_32.h5', save_format='h5')\n",
    "encoder_32.save(model_path + 'encoder_32.h5', save_format='h5')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "18 wide encoder"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {},
   "outputs": [],
   "source": [
    "import keras\n",
    "from keras import layers"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {},
   "outputs": [],
   "source": [
    "input_dim = X.shape[1]\n",
    "encoding_dim = 18\n",
    "\n",
    "input_layer = keras.Input(shape=(input_dim,))\n",
    "encoded_layer = layers.Dense(encoding_dim, activation='relu')(input_layer)\n",
    "decoded_layer = layers.Dense(input_dim, activation='relu')(encoded_layer)\n",
    "\n",
    "autoencoder_18 = keras.Model(input_layer,decoded_layer)\n",
    "\n",
    "encoder_18 = keras.Model(input_layer, encoded_layer)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {},
   "outputs": [],
   "source": [
    "autoencoder_18.compile(optimizer='adam',\n",
    "                    loss='binary_crossentropy')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Model: \"model_2\"\n",
      "_________________________________________________________________\n",
      "Layer (type)                 Output Shape              Param #   \n",
      "=================================================================\n",
      "input_2 (InputLayer)         [(None, 136)]             0         \n",
      "_________________________________________________________________\n",
      "dense_2 (Dense)              (None, 18)                2466      \n",
      "_________________________________________________________________\n",
      "dense_3 (Dense)              (None, 136)               2584      \n",
      "=================================================================\n",
      "Total params: 5,050\n",
      "Trainable params: 5,050\n",
      "Non-trainable params: 0\n",
      "_________________________________________________________________\n"
     ]
    }
   ],
   "source": [
    "autoencoder_18.summary()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {},
   "outputs": [],
   "source": [
    "history = autoencoder_18.fit(X, X, epochs=2, batch_size=64, verbose=0, validation_data=(Xtest, Xtest))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig, ax = plt.subplots()\n",
    "\n",
    "plt.plot([0]+history.history['loss'], label='train')\n",
    "plt.plot([0]+history.history['val_loss'], label='test')\n",
    "plt.legend()\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {},
   "outputs": [],
   "source": [
    "predictions = autoencoder_18.predict(Xvalid)\n",
    "# TODO I should probably do something with the validation set"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {},
   "outputs": [],
   "source": [
    "autoencoder_18.save(model_path + 'autoencoder_18.h5', save_format='h5')\n",
    "encoder_18.save(model_path + 'encoder_18.h5', save_format='h5')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## K-means\n",
    "Not sure what this is going to do \n",
    "(my guess is it is going to cluster similar participants)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.cluster import KMeans"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {},
   "outputs": [],
   "source": [
    "def visualize_labels(y):\n",
    "    idx = np.arange(len(y))\n",
    "\n",
    "    fig, (ax1,ax2) = plt.subplots(nrows=2, sharex=True)\n",
    "\n",
    "    ax1.imshow(y[np.newaxis,:], cmap=\"plasma\", aspect=\"auto\")\n",
    "    ax1.set_yticks([])\n",
    "\n",
    "    ax2.scatter(idx,y)\n",
    "\n",
    "    plt.tight_layout()\n",
    "    plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "metadata": {},
   "outputs": [],
   "source": [
    "face_pca_n_32 = pickle.load(open(model_path+'pca_n_32.p','rb'))\n",
    "\n",
    "X_pca = face_pca_n_32.transform(X)\n",
    "kmeans_5 = KMeans(n_clusters=5, random_state=0).fit(X_pca)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "y = kmeans_5.predict(face_pca_n_32.transform(Xtest))\n",
    "visualize_labels(y)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "y = kmeans_5.predict(face_pca_n_32.transform(Xvalid))\n",
    "visualize_labels(y)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "metadata": {},
   "outputs": [],
   "source": [
    "pickle.dump(kmeans_5,open(model_path+'kmeans_5_from_pca.p','wb'))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "   "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "WARNING:tensorflow:No training configuration found in the save file, so the model was *not* compiled. Compile it manually.\n"
     ]
    }
   ],
   "source": [
    "encoder_32 = keras.models.load_model(model_path + 'encoder_32.h5')\n",
    "\n",
    "X_auto = encoder_32.predict(X)\n",
    "kmeans_5 = KMeans(n_clusters=5, random_state=0).fit(X_auto)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "y = kmeans_5.predict(encoder_32.predict(Xtest))\n",
    "visualize_labels(y)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "y = kmeans_5.predict(encoder_32.predict(Xvalid))\n",
    "visualize_labels(y)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "metadata": {},
   "outputs": [],
   "source": [
    "pickle.dump(kmeans_5,open(model_path+'kmeans_5_from_auto.p','wb'))"
   ]
  }
 ],
 "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": "3.6.13"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}