add utility functions

williamgilpin · Feb 21, 2020 · 9521ac5 · 9521ac5
1 parent 9c075fe
commit 9521ac5
Show file tree

Hide file tree

Showing 3 changed files with 122 additions and 38 deletions.
diff --git a/demos.ipynb b/demos.ipynb
@@ -9,7 +9,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 3,
+   "execution_count": 1,
    "metadata": {},
    "outputs": [
     {
@@ -56,16 +56,16 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 6,
+   "execution_count": 2,
    "metadata": {},
    "outputs": [
     {
      "data": {
       "text/plain": [
-       "[<matplotlib.lines.Line2D at 0x1c4a8e7940>]"
+       "[<matplotlib.lines.Line2D at 0x1c3ff211d0>]"
       ]
      },
-     "execution_count": 6,
+     "execution_count": 2,
      "metadata": {},
      "output_type": "execute_result"
     },
@@ -117,7 +117,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 7,
+   "execution_count": 3,
    "metadata": {},
    "outputs": [],
    "source": [
@@ -147,22 +147,22 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 8,
+   "execution_count": 4,
    "metadata": {},
    "outputs": [
     {
      "name": "stdout",
      "output_type": "stream",
      "text": [
-      "Model: \"sequential_2\"\n",
+      "Model: \"sequential\"\n",
       "_________________________________________________________________\n",
       "Layer (type)                 Output Shape              Param #   \n",
       "=================================================================\n",
-      "gaussian_noise_2 (GaussianNo (None, 10, 1)             0         \n",
+      "gaussian_noise (GaussianNois (None, 10, 1)             0         \n",
       "_________________________________________________________________\n",
-      "lstm_2 (LSTM)                (None, 10)                480       \n",
+      "lstm (LSTM)                  (None, 10)                480       \n",
       "_________________________________________________________________\n",
-      "batch_normalization_2 (Batch (None, 10)                40        \n",
+      "batch_normalization (BatchNo (None, 10)                40        \n",
       "=================================================================\n",
       "Total params: 520\n",
       "Trainable params: 500\n",
@@ -205,29 +205,9 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 9,
+   "execution_count": 6,
    "metadata": {},
-   "outputs": [
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "0.07771948337554932\n"
-     ]
-    },
-    {
-     "data": {
-      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAXcAAAD8CAYAAACMwORRAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADl0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uIDIuMi40LCBodHRwOi8vbWF0cGxvdGxpYi5vcmcv7US4rQAAIABJREFUeJzt3Xd4HOW59/HvvbvaVW+WZNmSreKGK27YNJsOxgQ7QAJ2QgkhQE5CSA5pkOSQQN40SHJCCAkhHIcSaiAkBgwOxYTqIhv3XiRblmXJ6rLq7j7vH88Ky7KazaqMfH+uS9fujmZnbs2ufvPMM02MMSillBpYXH1dgFJKqfDTcFdKqQFIw10ppQYgDXellBqANNyVUmoA0nBXSqkBSMNdKaUGIA13pZQagDTclVJqAPL01YxTUlJMdnZ2X81eKaUcafXq1YeMMaldjddn4Z6dnU1eXl5fzV4ppRxJRAq6M552yyil1ACk4a6UUgOQhrtSSg1AGu5KKTUAabgrpdQApOGulFIDkIa7UkoNQI4L91X55fx66Tb8gWBfl6KUUv1Wt8JdROaIyDYR2Skid3YwztUisllENonI0+Et84i1eyv5w7KdNPo13JVSqiNdnqEqIm7gIeAioBBYJSKLjTGbW40zCrgLOMsYUyEiaT1VsNdj10eN/iAxvp6ai1JKOVt3Wu4zgJ3GmN3GmCbgWWB+m3FuBh4yxlQAGGNKwlvmEb5QuDdpy10ppTrUnXDPAPa1el0YGtbaaGC0iHwgIstFZE57ExKRW0QkT0TySktLT6jgIy33wAm9XymlTgbdCXdpZ5hp89oDjALOBRYCj4pI4jFvMuYRY8x0Y8z01NQuL2rWLp/HDWjLXSmlOtOdcC8EhrV6nQkUtTPOv4wxzcaYPcA2bNiHXes+d6WUUu3rTrivAkaJSI6IeIEFwOI24/wTOA9ARFKw3TS7w1loC5+Gu1JKdanLcDfG+IHbgKXAFuB5Y8wmEblXROaFRlsKlInIZmAZ8F1jTFlPFKx97kop1bVu3azDGLMEWNJm2N2tnhvgjtBPj9KjZZRSqmuOO0NV+9yVUqprjgv3lqNlNNyVUqpjDgx37ZZRSqmuODbcdYeqUkp1zHHh7tWWu1JKdclx4a597kop1TXHhbu23JVSqmuOC3e3S/C4RPvclVKqE44Ld7Ctd225K6VUxxwZ7j6PS/vclVKqE44Md225K6VU5xwZ7j6PW1vuSinVCUeGu7bclVKqc44Md9vnrkfLKKVURxwZ7l7doaqUUp1yZLjr0TJKKdU5R4a71+PWPnellOqEI8NdW+5KKdU5R4a7PVpGd6gqpVRHHBnu2nJXSqnOOTbctc9dKaU65tBw1zNUlVKqM44Mdz1DVSmlOufIcNczVJVSqnOODHev20XQgD+grXellGqPI8PdF2HL1n53pZRqnyPD3evW+6gqpVRnuhXuIjJHRLaJyE4RubOd339JREpFZG3o5yvhL/UIX4Qb0Ja7Ukp1xNPVCCLiBh4CLgIKgVUistgYs7nNqM8ZY27rgRqPoS13pZTqXHda7jOAncaY3caYJuBZYH7PltW5I33uesSMUkq1pzvhngHsa/W6MDSsratEZL2IvCAiw9qbkIjcIiJ5IpJXWlp6AuVaLS137ZZRSqn2dSfcpZ1hps3rl4FsY8wk4E3g8fYmZIx5xBgz3RgzPTU19fgqbUX73JVSqnPdCfdCoHVLPBMoaj2CMabMGNMYevkXYFp4ymuf9rkrpVTnuhPuq4BRIpIjIl5gAbC49QgiMqTVy3nAlvCVeCztc1dKqc51ebSMMcYvIrcBSwE3sMgYs0lE7gXyjDGLgdtFZB7gB8qBL/VgzdpyV0qpLnQZ7gDGmCXAkjbD7m71/C7grvCW1rFIPUNVKaU65dAzVO0OVW25K6VU+xwZ7nptGaWU6pwjw/1In7vuUFVKqfY4Mty15a6UUp1zZLjr0TJKKdU5R4a7x+3C7RJtuSulVAccGe5gW+9NeicmpZRql2PDPTLCRV2Tv6/LUEqpfsmx4Z4U7aWyrrmvy1BKqX7JseGeHOOl/HBTX5ehlFL9kmPDPUnDXSmlOuTYcE+O9lJRp+GulFLtcWy4J8V4qTjcjDFt7xuilFLKseE+KMZLUyDI4Sa9BIFSSrXl2HBPivECUF6rXTNKKdWWY8M9OSYCgHLtd1dKqWM4NtyTom3LvUKPmFFKqWM4NtyTW7plNNyVUuoYGu5KKTUAOTbcY30eItyife5KKdUOx4a7iJAU7dU+d6WUaodjwx30+jJKKdURR4d7UrSGu1JKtcfR4Z4c69U+d6WUaoezw1373JVSql2ODvekGC+V9c0EgnrxMKWUas3R4Z4cHYExUKldM0opdZRuhbuIzBGRbSKyU0Tu7GS8z4mIEZHp4SuxY+kJkQAcqGrojdkppZRjdBnuIuIGHgIuBcYBC0VkXDvjxQG3AyvCXWRHhiZGAVBUWd9bs1RKKUfoTst9BrDTGLPbGNMEPAvMb2e8nwL3Ab3WjM4Ihft+DXellDpKd8I9A9jX6nVhaNgnRGQKMMwY80oYa+tScowXn8elLXellGqjO+Eu7Qz75PAUEXEB/wt8u8sJidwiInkikldaWtr9KjueHhmJUdpyV0qpNroT7oXAsFavM4GiVq/jgAnAOyKSD5wOLG5vp6ox5hFjzHRjzPTU1NQTr7qVjKQo9lfqDlWllGqtO+G+ChglIjki4gUWAItbfmmMqTLGpBhjso0x2cByYJ4xJq9HKm5jaEKUdssopVQbXYa7McYP3AYsBbYAzxtjNonIvSIyr6cL7EpGUhSlNY00NOuNspVSqoWnOyMZY5YAS9oMu7uDcc/99GV1X8vhkMVVDWSnxPTmrJVSqt9y3hmq25fCizdDoBmAoYn2RCbdqaqUUkc4L9wPbYcNz0OzDfPMxGhAw10ppVpzXrh7bEsdfyMAgxN8iMD+Cg13pZRq4cBw99lHvz380edxkxrro1DDXSmlPuHAcD+65Q6QnRJDQdnhPipIKaX6HweG+9Etd4ARqTHsPqThrpRSLRwY7se23HNSYig/3KTXdVdKqRAHhvuxLffclFgAbb0rpVSIA8O9peXeKtxT7clLu0s13JVSChwZ7i0t9yPdMsOSo/G4hN2ltX1UlFJK9S8ODPdjW+4RbhfDk6PZo90ySikFODLcj225g+2a0W4ZpZSyHBjux7bcAXJTY9lTdphA0LTzJqWUOrk4ONyPbrnnpMTQ5A/qZQiUUgpHhvuxh0ICjB0SD8Cmoqrerkgppfod54W7u/0+97FD4ohwC+v3a7grpZQDw90DLs8xLXefx82Y9DjWF1b2UWFKKdV/OC/cwfa7t2m5A0zKTGR9YRXG6E5VpdTJzaHh7jum5Q4wKSOBmgY/+WV1fVCUUkr1Hw4N9/Zb7hMzEwC0a0YpddJzaLi333IfPTgOn8fF+kLdqaqUOrk5NNwj2w33CLeLCRkJrNlb0QdFKaVU/+HQcPe12y0DMCMnmQ2FVdQ3BXq5KKWU6j8cGu7tt9zBhrs/aPhYW+9KqZOYQ8O945b7tKwkXAIr9pT3clFKKdV/ODTcO265x0dGMG5oPCv2lPVyUUop1X84NNw7brkDzMgexMd7K2n0a7+7Uurk1K1wF5E5IrJNRHaKyJ3t/P6rIrJBRNaKyPsiMi78pbbSScsdYGZuMo3+IKvztd9dKXVy6jLcRcQNPARcCowDFrYT3k8bYyYaYyYD9wG/DXulrXXRcp81KoVor5vF64p6tAyllOqvutNynwHsNMbsNsY0Ac8C81uPYIypbvUyBujZi7t00XKP9nq4ZHw6SzYc0K4ZpdRJqTvhngHsa/W6MDTsKCLydRHZhW253x6e8jrQRcsdYP7koVQ3+HlnW2mPlqKUUv1Rd8Jd2hl2TMvcGPOQMWYE8H3gR+1OSOQWEckTkbzS0k8Rui0t906u/nj2yBQGxXi1a0YpdVLqTrgXAsNavc4EOkvMZ4HPtvcLY8wjxpjpxpjpqamp3a+yLY8PMBBo7ngUt4tzxqSyfFeZXgJYKXXS6U64rwJGiUiOiHiBBcDi1iOIyKhWLy8DdoSvxHZ0cJPstqZnJVN2uIk9hw73aDlKKdXfeLoawRjjF5HbgKWAG1hkjNkkIvcCecaYxcBtInIh0AxUADf0ZNGfhHugqdPRpmcnAbC6oILc1NgeLUkppfqTLsMdwBizBFjSZtjdrZ5/M8x1da6bLfeRqbHER3pYXVDB56cP63RcpZQaSBx6hmpLuHd+xIzLJUzLSiKvQE9mUkqdXBwa7j772EXLHWB6djI7S2qprOu8C0cppQYSh4Z797plAKZn2X73D3fphcSUUicPh4Z7S8u9824ZsJcATo3z8a+1+3u4KKWU6j8cGu7db7l73C4unzSUZVtLqarr+Lh4pZQaSBwa7t1vuQNcMSWDpkCQJRsP9GBRSinVfzg03LvfcgeYkBHPiNQYXlxd2INFKaVU/+HQcD++lruIsHDGcPIKKthQWNWDhSmlVP/g0HA/vpY7wDWnDSPW5+HR93f3UFFKKdV/ODTcj6/lDhAXGcGC04bxyvoDFFXW91BhSinVPzg03I+/5Q7wpbOycQk8tGxnDxSllFL9h0PD/fhb7gCZSdEsnDGc51btI1+vFKmUGsCcGe4uN7gijrvlDnDb+SOJcLv47Rvbe6AwpZTqH5wZ7hC6G9PxtdwB0uIiWThjOK9tPEB1g57UpJQamJwb7pHxUH9iV3u8bNIQmgOGZVtLwlyUUkr1D84N97h0qCk+obdOGZZIWpyP1zee2PuVUqq/c3C4D4GaE7ucgMslXDI+nXe2lVLfFAhzYUop1fccHO7pJxzuAHMmpFPfHOBt7ZpRSg1Azg73+gpoPv4jZgBm5iQzPDmah/+zC2NMmItTSqm+5eBwH2Ifa0+s39zjdnHbeSPZsL+KZdu09a6UGlgcHO7p9vEEd6oCXDE1g2HJUTzw5g5tvSulBhQHh/tQ+/gp+t0j3C5unT2CdYVVehNtpdSA4uBw//Qtd4CrpmaSEBXBovf3hKEopZTqH5wb7lFJ4PZ9qpY7QJTXzRdmDmfppmL2ldeFqTillOpbzg13Edt6r/70t867/owsXCI88NaOMBSmlFJ9z7nhDp/qRKbWhiREcfPsXF5YXciK3WVhKEwppfqWw8P9xC9B0Nbt548iMymKH/1zI03+YFimqZRSfcXZ4R4/NGzhHuV1c+/88ewoqdVb8SmlHK9b4S4ic0Rkm4jsFJE72/n9HSKyWUTWi8hbIpIV/lLbEZcOTTXQWBOWyZ1/ymDmjE/n92/t0J2rSilH6zLcRcQNPARcCowDForIuDajfQxMN8ZMAl4A7gt3oe1KHG4fy8N3GOOP543DJcIvX9satmkqpVRv607LfQaw0xiz2xjTBDwLzG89gjFmmTGmpam7HMgMb5kdSAutY0q2hG2SQxKi+MqsXF7dcID1hZVhm65SSvWm7oR7BrCv1evC0LCO3AS81t4vROQWEckTkbzS0tLuV9mR5Fxwe6Fk86efVis3z8ohOcbLfa9vC+t0lVKqt3Qn3KWdYe1eiEVErgWmA/e393tjzCPGmOnGmOmpqandr7Ij7ghIGR3WljtAXGQEXz9vJO/vPMT7Ow6FddpKKdUbuhPuhcCwVq8zgaK2I4nIhcAPgXnGmOO/uemJShsb9nAH+OLM4WQkRvGr17fqRcWUUo7TnXBfBYwSkRwR8QILgMWtRxCRKcCfscHeu9fPTT0FqvZCQ3VYJxsZ4ea/LxrNhv1VLNmgt+NTSjlLl+FujPEDtwFLgS3A88aYTSJyr4jMC412PxAL/F1E1orI4g4mF34tO1VLw98/fsWUDEakxvCHZTu19a6UchRPd0YyxiwBlrQZdner5xeGua7uSxtrH0s2w7DTwjppt0u49ZwRfO+F9by34xCzR4dhP4FSSvUCZ5+hCpCYBRHRYT9ipsX8yUMZHO/jkXf1rFWllHM4P9xdLsiYBvnv98jkfR43Xz4rR4+cUUo5ivPDHWDkhXBwY1gu/9ueG87MJiclhh+8tIH6pkCPzEMppcJp4IQ7wM43e2TykRFufn7FRPaW1/GHZXrNd6VU/zcwwn3weHtP1Z1v9NgszhgxiHmnDmXR+/mU1fbeYfxKKXUiBka4i8DIC2DXOxDw99hsbr9gFI3+AI+8pztXlVL928AId4Axc6GxCtY+1WOzGJkWy7xTh/LEhwUcrG7osfkopdSnNYDC/VIYfia8cTfUhuGiZB341oWjMRi+/+J6PbFJKdVvDZxwF4HP/C80HYZ//6jHZpOdEsMP5o7lnW2lPLNyX9dvUEqpPjBwwh0g7RQ463ZY/yzs/k+PzebamVnMyE7m92/toDmg91tVSvU/AyvcAWZ/F5Ky4dU7wnb7vbZcLuGr5+ZSXN3Aaxv1omJKqf5n4IV7RJTtninfDX+eDcUbemQ2545OIyclhkXvh+8Wf0opFS4DL9wBRpwPN7wCzfXw/PUQDH/XicslfOnMbNbuq+SjXWVhn75SSn0aAzPcAbLPgkt+blvw21/vkVlcc9owMhKjuOflTfi1710p1Y8M3HAHGDsPEobB8j/2yOQjI9z8YO5YthbX8NSKvT0yD6WUOhEDO9zdHphxC+S/B9uX9sgs5k5M5+yRKdz7ymb+saawR+ahlFLHa2CHO8D0GyF9Ijx3LWx5OeyTFxEevm4aM3OSueP5dSzZ0DNXplRKqeMx8MPdFwfXL4bBE2zAv3gz7HgDGqrCNotYn4e/3ngak4cl8v0X1lNQdjhs01ZKqRMx8MMdIDoZbnwNZn8PNr0ET30O/nQW1FeEbRY+j5s/fGEKLpfwtafW0NCs131XSvWdkyPcASIi4fwfwvd2wdVPQnURvP6DsM4iMyma33z+VDYVVfP/Xu2Z2/4ppVR3nDzh3iIyAcbNg7P/G9Y9Df/8OhSuDtvkLxw3mFtm5/K35Xt5eV1R2KarlFLH4+QL9xbnfA+mfcl20zx6Abz1U9uary2B/avh8ImfmPTdS8YwLSuJu/6xgT2HtP9dKdX7pK8uWzt9+nSTl5fXJ/M+SmMNvHYnrP3b0cPFBRnTYfTFMOkaSBx+XJMtqqxn7u/fIz4ygp9fMZGzR6WEsWil1MlKRFYbY6Z3Od5JH+4t8j+AQ9sg0AzxQ6F4I+xYCkUfgysCRl0MlXshdTSc90MYNKLLSebll/Ptv6+joKyOK6dm8JN544mPjOiFP0YpNVBpuIdL5T547zf28MmUkbBvFTQfhrRxMPOrMO2GTt/e0BzgoWU7+eM7u4iP9HDZpCF89ZwRZCZF99IfoJQaSDTce0rNQfj4Sdi2xPbNz/01zLi5y7et21fJX97bzRubDxLldfPzKyZy3pg0orzuXihaKTVQaLj3tEAzPHcdbH8NRlwAQ0613TZZZ8L4K+yx9e3IP3SYr/5tNVuLa3C7hLhID5lJUfzumsmMTIvr5T9CKeU0YQ13EZkDPAC4gUeNMb9s8/vZwO+AScACY8wLXU3T8eEO4G+ED38PKx6B+nKIHgS1ByEixrbmz7wdYgYd87aG5gDv7zjE+sJKquqbeXVDMc2BIBeNG0yTP8jM3GQuGZ/OoBgvr28sJinGy+m5R6ZzoKqe9PhIRKQ3/1qlVD8QtnAXETewHbgIKARWAQuNMZtbjZMNxAPfARafNOHeIuAHEwC3F4rXw4cPwoYXICIaRp5vW/kjzodTF0Jk/DFv31dex9efXkNJdSMAxdUNeD0uclNi2Fpcgwh85+IxfO3cEbz08X7ueH4dcyem88urJukOWqVOMuEM9zOAnxhjLgm9vgvAGPOLdsZ9DHjlpAv39pRug3fvt0fbAJTttI8xaTDx8zDr2+226o0xbD9YyxMf5bMqv5zrz8hm5Z5yFq8r4vxT0lixu4yUOB+FFfUkRXu58axsbjo7h8gI23dfWdfE/sp6xg9N6KU/VCnVm7ob7p5uTCsD2NfqdSEw8wSLugW4BWD48OM7btxxUsfAVY8eeV2YB7uWwcGNsOJPsOovkD0LPJHgcsNF90ByLiLCmPQ4fnbFRPu+gJ8vzshkUmYCv3htKzFeN89cN4aDTVH875s7uH/pNjYUVvHQF6ciwJcfW8X6wiqeueV0Tstuv99fKTXwdSfc2+vYPaG9sMaYR4BHwLbcT2QajpU53f6AbdWvfhx2vQXihqpCeHiWPVGqugiGTrHdOHFD4M0fI1HJfOVzizg99yziC95g6MPjGDrpGp744q/5v1Wp/PSVzXz/xfWMTItlzd5K4iI9fO2pNbx829mkJ0QeU0qTP0hxVQPDB0Wzt6yOv7y3mxvOzNIdukoNINot0x9U7oOld9kdtHHptpVfEtqlkToWDpfY+8FedC+8+2s7/HCJDf9Z3+a+QzP507sFGAPnZvn4n9kJfObZMmJ8Hq45LZPXNhRT3dBMQlQEkzITWb67jANVDTywYDKLPshn3b5KItzCnAlDmDwsketOz8LrOfbKFE3+IIGgCdvhm69vLMbtEi4aNzgs01PqZBDOPncPdofqBcB+7A7VLxhjNrUz7mNouIdH+W7bwh95IRwutdehL3jftvRvfhv8DfDG3bBvBUz4HDvP/g0vfbyf2wtux1eylr1zn+DL78ays6SWmTnJjEyL5WB1A2v3VTIiNZYGf5B1+yoB+On88WzcX817O0opqmrgK2fn8J1LxvD21hImZiSwq7SW3/x7O1sOVBM0hjHp8Zx/SipzJw5h3JD4Do/aKa5qoLi6gVMzE44ZZ9m2Em56bBU+j5v/fO9c0uIiMcawKr+CcUPjifV1Z6Py0zHGEDTgdulRR8o5wn0o5FzsoY5uYJEx5mcici+QZ4xZLCKnAS8BSUADUGyMGd/ZNDXcj1MwYO8FG5kIU6+zw4yBD34Hb/4ERl9qL4nw0R8gJhX8jTQtfJ6yhIkMKX4HTNCeVTtoBFTvp7ZgLZ9/O57LB+3na/4nYdqNMOEq7n55M098VMDw5Gj2ltd9MvvclBjmTEjH4xJW5pezck85QQM5KTFMy0oi2utmc1E1Z45M4cKxadz78mbyCuz18ueMT8ftFt7ZWsJlk4aQHh/JXz/IJy3eR35ZHdfOHM498yfwzMq93PWPDSRERfBf547g5lm5xwRvoz/AtuIa/EFDZmIUafHHdju1VdvoR4CYVisMYwx3PL+OvIJyHr9xBrmpsd3+KLYcqGZQjPeYeW8rriFoDGOHHHtE1PHaW1ZHUyBITkpMj618Vuwu4x9r9nPP/PGf7JBX/Z+exHQy+fBBWPZzaK6DMZfBpb+Cv14K1fshMQsq9hwZNz4DaorBBDAzbkW2vvLJa3wJBFNP4caqr7C5IZkfXz6OgrI6fB4X150+DN+BNVBbDOkTKYsYytLNJSzdVMzmA9XUNvgZkRbDxv3VAAyK8fKVWbkEgkEeeGsHEW4X545J5a0tJTQHgkzPTuaRkct5d1cV3y6YwbcuHM0f3t7JxIwE4iI9vLW1hBk5yUzKSKC20U/QGLYdrGVLUTVNgSAAHpdwz/zxXDwunVfWF/Hwf3aRHOPjjotGc87oVOqa/LywupDfv7WDxGgvf//qGQwOBfI/P97Pt55bS4RbiIuMINbnIdrr5pHrpjN80LGXhjDGUN8c4LlV+/jpK5vJGhTD4tvOIi50KOo/P97P915YT8AYbj9/FMkxETT6gwxJiGJIYiSj0mI/Gbet5kCQCPeRbrA/vbOLX72+FYAZOck88eUZPLtyL4NifVx+6lDKau0hs4NifSf8lcnLL+f6RSupawrwwILJzJ+cccLTaq3RH8Dn0RVFT9JwP9nUldtLIoyZa8+ObaiCN++BA2vh9K/ZFvv+NbD7HUjOsZc0Xvs3e1G0m/5tVwAFH8H65wikjiMw6zt437sPBo+zx+tv/pddWbRIHQunLoBJV0P8UIwxiDG8u7OM93ce4tZZOQxqLISaIvbFTyXGF0Gy6zDVrjhMEBJ2vAgv3YoRF99IeJBXipNIiXLx1sUlJEy4hBe2+7ln8SaaAkHio2wo5qTEMGV4IpMzE4mMcPP4R/m8s630k5JmZCdzsKaBgrI6IiNcNAcMgaDhrJGDWLu3kiGJUcw7dSgHqxv419oixqTH8csrJ/Kjf24kPiqCVfnl+Dwu/vjFqUzLSqbicBO//vc2Fq8torbJT8u/ysycZPIKKrho7GCuPT2L5/L28fK6ImbmJJMc4+W1jcXHfDzxkR5+dsVEJg9LZHVBBUs3FTMtK4n0hEh++NJGzhwxiN9ePZmiqnou/d17nD0qhWlZSdy/dBsZiVHsr6wH4NbZuTyXZw9ee3DhFGaNSj1qPtUNzazcXc60rCQiPC4efHsH44bEc9nEIXjcLowxPLViL//v1c0MSYiisTnAiLRYnrxppv0MRfAHgvxneymjB8exo6SGV9Yf4JsXjGJ4cjT/2V7K1uIaUmJ9XDklA1errYqfLN7EK+uLeO2bs0mNO/EVz6dVUtMQWll7KCg7jMftIiMx6rin0xwIUn646ZMGQYtA0HCwuoGhHUzTHwhi4KgVdjhpuKvOBQO2tT94PEy48sjwdc/CS7fa5/GZ9laEwWbb9z/+iiMriQ0vwL7lgNijfLwxcGi77frJOhO2LoGqvXY6gyfaaZRuhVM+AwnDYPVf7VFBJVtg6GR2zfkbKasfIGHF/bbrac4vCEy8BpfLdaS/3hh7BnDlPjBBAnFDeXpbkIamAKflJHNqeiTNxsWyHeWs2F1OlNfFRePSmZwcYGV+OV97KZ9DtU3E+jycnjuI+4d/RFL5erulE53M1uJqbli0koPVjeSmxFBYUU/AGD47OYOMpCiivW6GJERyeY7w6KpD/PxNG7KRES5umZXLbedkE+FxseVgPUkxEURHeCiqqqewop6Hlu1kbWgfB0BSdAQVdc2A7fLaU3aYYUnRNAeC1DUFePMb00lNTOT3y3bx2ze2c8dFo1mxp4wPdpYxZrA9qmnbwRoGx/sYHB+Jz+OivjnAjoO1NPqDDIrxkhgdwa5Sez+B9PhIZo1KYe2+SnaU1DJrVAq/nZfNf955g++uSeKW2bn8Y81+/rBwCm9vLeHP7+4+6uuSkxLDadlJPJ8ZLUQUAAAPgElEQVRX+MmwmTnJlB1uorq+mbkTh/DYh/kALJwxjF9cOYm8/HJ+9+YOZo9O4ZbZ9iqqr64/wKIP9jA9O4lTMxOpbwqw5UA1zYEgI9JimTMhnbS4SOqa/Dz63h62HKhmTHocgaChpsFPhFu4evowMpKiuPflzczISebKqZls3F9FdUMzWw7UcN/rW8lJieFnV0zkS4tW4g8afnDZWM4/JY1Yr4egMSTFeENfKcP/vb+HzUXVfP/SUz4J8tpGP1/+6yo+3lfBgwunMmdCOgCHahv5xtMfs3xPGX/90mmcOyaNuiY/awoqGRzvY1Csj2sfXcG+8jouGj+YyycN5YwRgz7p9jLGsLW4hsHxkSSHajheGu7qxBgDS74LDZVw2W/BHWFXBL52+qTLdtmbnZRshsZaSBkFu962O4JHnA+nzAW3z+4XiIi2ob/mSQg02evkf+Z3sPFFeO17kHse7HkXRl8Chw9B4UpIGWNvcN5cby/DfHAT1LS6u5W44TP/CyPOg9WPwfKH7SUgLrgbxn7GnkOw5WVYfBt44+Dmt2mOTsUlgnvHUnjmGjudhGF2P0VkAocvf5jH19WyOr+CkYNjuXJKJmPS444sm/XPwcvfhJRR7Lr8H5Q0uBmRGkNa8354ZoHd0X35A+CLt+cvpE+ChiqacfPKthqaA8bupxiexIe7ythaXM31Z2Tz7vZSHvswHxH4n7QPGL3mp/aeAqfdTOXse0isy6feRPDqPi9zJ6ZjDDy5vIAdB2spO9xIY3OQyAgXWYNiOGPEIB5atpO95XU8uHAK9U0BXlhdyEe7y8hNjeX607O4YlQErqeuhIMbubLxJ6wxo4mKcOMSONwU4MqpGZyamUhCVASD4yO5YdFKmgJB/uvcEfzXuSNYvLaIX722lZGDYwkEDesLq5iWlcTEjAQe/yifKcMSWbO3Ep/HRaM/yO0XjKK6vpnHPswnIzGK4uoGAkGbPT6Piwi3i9pGG96jB8dRVFlPRV0zmUlRFFbU4xK7z6SxOYjP4yI3LfaTAwJOSY9ja3HNJ1+LGTnJfLy3guaAISXWx+jBsXy46+ib70zPSuKS8ensPlTLMyv3IQKxXg+pcT6ag0ECAcPBmkZyUmLYc+gwDyyYbFckf/yQ0ppGUuN81DcFuGRCOn/P20dzwCACg2J81DQ0c/H4dN7ZVkJNgx+A5BgvaXE+quqbOVDVwL3zx3P9Gdkn9C+q4a76hjE2vD0dbJY3he5M5Y2xj8GADf8Pfg9RiXDruzaIN78EK/5sAzoiGqoLIXmEXUEkZYPLY3cw73zzyLTHfdaucA5uCJ0c5oGmWkifaIenjbPX/Dm4EfIes91Tl/wc3vgf8MbaQ1CTsmHBU3YLpXwPrHkcdrxpj1iqK7NbIOmToHgDnHKZvaTEgbWw6v8AY2/jWJF/pCZPpA18X4I9qW30xXBgPax9yv4tKWPsyiL3XDjn+3br58+zIfM0iBtsV36nfAa2L7XL9PwfQf779vmF90DisGOXf+VegnFDaQoKkYUfQFSSXQYtW0CHdsDTV9t9LcC6xAt5b+zdzJ+cwVV/+pD0GBcvnFOKd/SF9jMp380HFYkcOtxk++ZDdykz0cmICIGgYemmYmb5P8K363XO334VkZGRfG5aJl+YOZw7nlvLm1tKEIErp2TysysmUN8U4GBNAxFuF8OTo/G4hN2HDvPU8r3sPlRLcoyXhTOGc1p2Mg3NAbxuFy6XUFRZz41/XcWOkhp+/flT2XKgmsXrivjyWTlMzLRnZZ+RO4iX1x/gN//exgPXnMqkjESW7yknv6yOuia/XdmtKaSgzB4w8KUzs7nujCwefGsH/qDB7RIq6pq5/vQsTh8xiBv/upLVBRUMT47mYHUjT988k7hID5c/+AHNgSCfnz6Mi8cPZtWecl7dcICfzp/A7NGpNPoDfLDzEBsKqympaeBgdSM+j4tzRqdy3ilpJ9x1peGunKXpsA36dq6906FAMyz/kz0SaNw8SM6109i9zJ4NHGi2oTbpatj+OrzwZQj6beiPuhguve/ocNzzLjy9wF6vP26o3UoQlz2TOHEYRKfYFcLka2H5Q/ZQVLDj5My2Wzqxg2HTP+y4zXV2hRGXbkO6eL1dcTXVYM8NNHbrw+0Ffz0MnWpPYgv64WvL7VbIizfZ6Y28yF6cbv9qO7y53v7dSdlHVpTRKVBZYLu/hkyGpCy7rwRs19mpC+2Kd9Uiu0W28Bm78tr0T7juJdiymOoZ/01U3p+IeP9+ewE8X6ztCptxC4ydZ7fqSrfYfTVTvmi3UCKiIOcc+NtV4K/HzP0NnHYT0lAJqx/HnzmDje5x5KbGEF+bD2uegIypMOx0O/2Dm+zK0xtjl7U7tOP5wHq7sht1sV0RVeyByATqUidRVBNkZPN2+/dHJ9vPWlx2BfzWPZB+Kky5Fh67zA6/+nFIyPzkozbGUFXfTFMgSFqwzE4/c9rR36+CD2H7UurOuIObnt7C8j1l/PnaaVwcvQO2vsL28d8iMibh6B3wwYC97MjhQzDzVrs1G2Ya7kq1VV9hf6KS7E97aopt+JRshswZMPbyY1vHLSr3Qd0hewRSbFrn826qs5edOFxmw37yF2H9szasL/4ZbP6n7bJKHAZnfQuyzrDv8zfZfRtZZ9tg3vU25J5jw2P5H+1O7qY6wNhh3li7FbAydKXSc39wZAtg9zs2lLPPgnkP2sDfuwIWXXykzuFnQtEaG9YxKXal642xWxpgt56mXm+3Tj5+0gZn0G9XNDFptv6KAnt/4pWPQGM1eKLgmiftCmnxN2yXX0fSxsPpX7WXz37vt/YoLnHZ6bcYPBGGz4RVj9oV3fgrbXdZMGC7whrtEVsMnQJFa+2WnzsChkwKfRaH7Ypu6vV2BfLMAvs5jp4DaWNtV2Bilq21uQ4GT6Dp6qcpNikMX/cbWxcGRl1iTyzc8He7fN1eu8wKPrANiKDfrnwyZ9gtsUEj7DLY865tcGR2mc/t0nBX6mRWV25bw6ljjgyrLbHh723V0jQGFl1iu3lGXmi3RiKi4bY8SMg4Ms4HD9gVyQU/PrL/panOvu/Qdnt+xZTrbYj+5Tz7+7HzbDfYq9+xt7AESBkNC5+1K9HSLfaorrRxdgV5aDu88WPbBQe2m+20r9iut6Rsu/O/fDcs/aEN4ynX2S62oo/tSjhuqF2hnfVN+PeP7MrsnO/DhKvgnV9A1X7bNeVvtF1pLZKy7X2Slz9swzxod3QzaKR9/6vfsd1T4+bZw46nXGdXAkt/YMcTl12h+pvsdC/4sT34YP1zsPcjKFxlt35aeKJg7v1Hzlc5ThruSqnuMeZIf3zeIrtzeezlJz69HW9C/BAbxgC1pbDlXzZEh5959MqlLX+jvdaS22u7Udo7+7m21F5lNesM21pvrD52S6yx1ob7mEtta741Y+x1nQpX25XTlGtti7tlOdQU2y2kEefbraz9q+GJz9r5TPw8XPkXO97Kv9h6J37Ojtcy7bY1G2O72yoL7Ipg6JSO90l1g4a7UkqFy/7VsHkxnHsXRHR9VnRPCuclf5VS6uSWMc3+OEjPnEKllFKqT2m4K6XUAKThrpRSA5CGu1JKDUAa7kopNQBpuCul1ACk4a6UUgOQhrtSSg1AfXaGqoiUAgUn+PYU4FAYywmn/lqb1nV8tK7j119rG2h1ZRljUrsaqc/C/dMQkbzunH7bF/prbVrX8dG6jl9/re1krUu7ZZRSagDScFdKqQHIqeH+SF8X0In+WpvWdXy0ruPXX2s7KetyZJ+7Ukqpzjm15a6UUqoTjgt3EZkjIttEZKeI3NmHdQwTkWUiskVENonIN0PDfyIi+0Vkbehnbh/Uli8iG0LzzwsNSxaRN0RkR+ixg5uI9lhNY1otk7UiUi0i3+qr5SUii0SkREQ2thrW7jIS6/eh79x6EZnay3XdLyJbQ/N+SUQSQ8OzRaS+1bJ7uJfr6vCzE5G7Qstrm4hc0lN1dVLbc63qyheRtaHhvbLMOsmH3vuOGWMc8wO4gV1ALuAF1gHj+qiWIcDU0PM4YDswDvgJ8J0+Xk75QEqbYfcBd4ae3wn8qo8/x2Igq6+WFzAbmAps7GoZAXOB1wABTgdW9HJdFwOe0PNftaoru/V4fbC82v3sQv8H6wAfkBP6n3X3Zm1tfv8b4O7eXGad5EOvfcec1nKfAew0xuw2xjQBzwLz+6IQY8wBY8ya0PMaYAuQ0Re1dNN84PHQ88eBz/ZhLRcAu4wxJ3oS26dmjHkXKG8zuKNlNB94wljLgUQRGdJbdRlj/m2M8YdeLgcye2Lex1tXJ+YDzxpjGo0xe4Cd2P/dXq9NRAS4Gnimp+bfQU0d5UOvfcecFu4ZwL5WrwvpB4EqItnAFGBFaNBtoU2rRb3d/RFigH+LyGoRuSU0bLAx5gDYLx6Q1gd1tVjA0f9sfb28WnS0jPrT9+7L2BZeixwR+VhE/iMis/qgnvY+u/60vGYBB40xO1oN69Vl1iYfeu075rRwb+dW6PTp4T4iEgu8CHzLGFMN/AkYAUwGDmA3CXvbWcaYqcClwNdFZHYf1NAuEfEC84C/hwb1h+XVlX7xvRORHwJ+4KnQoAPAcGPMFOAO4GkRie/Fkjr67PrF8gpZyNENiV5dZu3kQ4ejtjPsUy0zp4V7ITCs1etMoKiPakFEIrAf3FPGmH8AGGMOGmMCxpgg8Bd6cHO0I8aYotBjCfBSqIaDLZt5oceS3q4r5FJgjTHmYKjGPl9erXS0jPr8eyciNwCfAb5oQp20oW6PstDz1di+7dG9VVMnn12fLy8AEfEAVwLPtQzrzWXWXj7Qi98xp4X7KmCUiOSEWoALgMV9UUioL+//gC3GmN+2Gt66n+wKYGPb9/ZwXTEiEtfyHLszbiN2Od0QGu0G4F+9WVcrR7Wk+np5tdHRMloMXB86ouF0oKpl07o3iMgc4PvAPGNMXavhqSLiDj3PBUYBu3uxro4+u8XAAhHxiUhOqK6VvVVXKxcCW40xhS0DemuZdZQP9OZ3rKf3Gof7B7tXeTt2jfvDPqzjbOxm03pgbehnLvAksCE0fDEwpJfrysUeqbAO2NSyjIBBwFvAjtBjch8ss2igDEhoNaxPlhd2BXMAaMa2mm7qaBlhN5kfCn3nNgDTe7mundj+2Jbv2cOhca8KfcbrgDXA5b1cV4efHfDD0PLaBlza259laPhjwFfbjNsry6yTfOi175ieoaqUUgOQ07pllFJKdYOGu1JKDUAa7kopNQBpuCul1ACk4a6UUgOQhrtSSg1AGu5KKTUAabgrpdQA9P8BgdJY5ZWl2SgAAAAASUVORK5CYII=\n",
-      "text/plain": [
-       "<Figure size 432x288 with 1 Axes>"
-      ]
-     },
-     "metadata": {
-      "needs_background": "light"
-     },
-     "output_type": "display_data"
-    }
-   ],
+   "outputs": [],
    "source": [
     "train_history = autoencoder.fit(x=tf.convert_to_tensor(X_train), \n",
     "                                y=tf.convert_to_tensor(X_train),\n",