evaluation.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 78,
   "id": "86829276",
   "metadata": {},
   "outputs": [],
   "source": [
    "from keras.models import Sequential\n",
    "from keras.layers import Dense, Dropout\n",
    "from keras.models import load_model\n",
    "import tensorflow as tf\n",
    "from sklearn.tree import DecisionTreeRegressor\n",
    "from sklearn.model_selection import KFold\n",
    "from sklearn.model_selection import train_test_split\n",
    "from sklearn.metrics import mean_squared_error\n",
    "from sklearn.ensemble import RandomForestRegressor,GradientBoostingRegressor\n",
    "import matplotlib.pyplot as plt\n",
    "import numpy as np\n",
    "import pandas as pd\n",
    "\n",
    "%matplotlib notebook\n",
    "\n",
    "from src.model import emb_model as Emb_model\n",
    "from src.train import train_and_get_deep_learning_model"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "155f6e84",
   "metadata": {},
   "source": [
    "# Task 1\n",
    "# Initialize data and model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 58,
   "id": "c2b88b8e",
   "metadata": {},
   "outputs": [],
   "source": [
    "def load_data(path = \"./data/preprocessed_train_data.csv\", is_np=True):\n",
    "    train_data = pd.read_csv(path)\n",
    "    if not is_np:\n",
    "        return train_data\n",
    "    X = train_data.iloc[:,:-1]\n",
    "    X = np.array(X)\n",
    "    y = train_data[\"price\"]\n",
    "    y = np.array(y)\n",
    "    return X, y"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 59,
   "id": "998477c3",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(16589, 123) (16589,)\n",
      "(16589, 123) (16589,)\n",
      "(16589, 124)\n"
     ]
    }
   ],
   "source": [
    "X, y = load_data(\"./data/preprocessed_train_data.csv\")\n",
    "df = pd.read_csv(\"./data/preprocessed_train_data.csv\")\n",
    "print(X.shape, y.shape)\n",
    "print(df.shape)\n",
    "\n",
    "# same random_state as data split for training\n",
    "X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.2, random_state=2021)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 60,
   "id": "44caf18a",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(16589, 123) (16589,)\n",
      "(13271, 123) (13271,) (3318, 123) (3318,)\n",
      "<src.model.simple_widedeep_model object at 0x0000012E358B9208>\n",
      "Epoch 1/400\n",
      "166/166 [==============================] - 1s 3ms/step - loss: 17830258688.0000 - rmse: 133529.9844 - val_loss: 2571390720.0000 - val_rmse: 50708.8828\n",
      "Epoch 2/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 2528207872.0000 - rmse: 50281.2891 - val_loss: 1394348928.0000 - val_rmse: 37340.9805\n",
      "Epoch 3/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 1998107648.0000 - rmse: 44700.1992 - val_loss: 1281703296.0000 - val_rmse: 35800.8828\n",
      "Epoch 4/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 1853382016.0000 - rmse: 43050.9219 - val_loss: 1135808896.0000 - val_rmse: 33701.7656\n",
      "Epoch 5/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 1805754752.0000 - rmse: 42494.1719 - val_loss: 1075279104.0000 - val_rmse: 32791.4492\n",
      "Epoch 6/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 1688681472.0000 - rmse: 41093.5703 - val_loss: 1037702528.0000 - val_rmse: 32213.3906\n",
      "Epoch 7/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 1671269504.0000 - rmse: 40881.1641 - val_loss: 1112743680.0000 - val_rmse: 33357.8125\n",
      "Epoch 8/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 1604848640.0000 - rmse: 40060.5625 - val_loss: 915345024.0000 - val_rmse: 30254.6699\n",
      "Epoch 9/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 1546407808.0000 - rmse: 39324.3906 - val_loss: 896093824.0000 - val_rmse: 29934.8262\n",
      "Epoch 10/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 1577067648.0000 - rmse: 39712.3125 - val_loss: 883295040.0000 - val_rmse: 29720.2793\n",
      "Epoch 11/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 1508499584.0000 - rmse: 38839.4062 - val_loss: 828627136.0000 - val_rmse: 28785.8848\n",
      "Epoch 12/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 1497009792.0000 - rmse: 38691.2109 - val_loss: 818901952.0000 - val_rmse: 28616.4629\n",
      "Epoch 13/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 1439506816.0000 - rmse: 37940.8320 - val_loss: 793362560.0000 - val_rmse: 28166.6934\n",
      "Epoch 14/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 1331729536.0000 - rmse: 36492.8711 - val_loss: 760373248.0000 - val_rmse: 27574.8672\n",
      "Epoch 15/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 1298299520.0000 - rmse: 36031.9219 - val_loss: 760287104.0000 - val_rmse: 27573.3047\n",
      "Epoch 16/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 1290214272.0000 - rmse: 35919.5547 - val_loss: 722813376.0000 - val_rmse: 26885.1895\n",
      "Epoch 17/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 1278710656.0000 - rmse: 35759.0625 - val_loss: 790887680.0000 - val_rmse: 28122.7246\n",
      "Epoch 18/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 1261683968.0000 - rmse: 35520.1914 - val_loss: 701810112.0000 - val_rmse: 26491.6992\n",
      "Epoch 19/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 1199366016.0000 - rmse: 34631.8633 - val_loss: 696543680.0000 - val_rmse: 26392.1133\n",
      "Epoch 20/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 1201087104.0000 - rmse: 34656.7031 - val_loss: 631647232.0000 - val_rmse: 25132.5938\n",
      "Epoch 21/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 1118631680.0000 - rmse: 33445.9531 - val_loss: 637675520.0000 - val_rmse: 25252.2383\n",
      "Epoch 22/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 1062909440.0000 - rmse: 32602.2910 - val_loss: 570601792.0000 - val_rmse: 23887.2734\n",
      "Epoch 23/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 1043067200.0000 - rmse: 32296.5508 - val_loss: 563753088.0000 - val_rmse: 23743.4844\n",
      "Epoch 24/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 1031125248.0000 - rmse: 32111.1387 - val_loss: 549149376.0000 - val_rmse: 23433.9355\n",
      "Epoch 25/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 1024568384.0000 - rmse: 32008.8789 - val_loss: 534973888.0000 - val_rmse: 23129.5020\n",
      "Epoch 26/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 1010539712.0000 - rmse: 31788.9863 - val_loss: 543096384.0000 - val_rmse: 23304.4277\n",
      "Epoch 27/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 980836736.0000 - rmse: 31318.3125 - val_loss: 599129920.0000 - val_rmse: 24477.1309\n",
      "Epoch 28/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 1003632832.0000 - rmse: 31680.1641 - val_loss: 551128064.0000 - val_rmse: 23476.1172\n",
      "Epoch 29/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 959896960.0000 - rmse: 30982.2031 - val_loss: 517891296.0000 - val_rmse: 22757.2246\n",
      "Epoch 30/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 949743552.0000 - rmse: 30817.9102 - val_loss: 507917024.0000 - val_rmse: 22537.0137\n",
      "Epoch 31/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 948520576.0000 - rmse: 30798.0605 - val_loss: 497145248.0000 - val_rmse: 22296.7539\n",
      "Epoch 32/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 906088000.0000 - rmse: 30101.2949 - val_loss: 500469472.0000 - val_rmse: 22371.1758\n",
      "Epoch 33/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 872347776.0000 - rmse: 29535.5332 - val_loss: 482000480.0000 - val_rmse: 21954.5098\n",
      "Epoch 34/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 867573184.0000 - rmse: 29454.5957 - val_loss: 469859648.0000 - val_rmse: 21676.2461\n",
      "Epoch 35/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 884865472.0000 - rmse: 29746.6875 - val_loss: 570062208.0000 - val_rmse: 23875.9746\n",
      "Epoch 36/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 771353280.0000 - rmse: 27773.2480 - val_loss: 561329856.0000 - val_rmse: 23692.4004\n",
      "Epoch 37/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 817450944.0000 - rmse: 28591.0996 - val_loss: 479350688.0000 - val_rmse: 21894.0781\n",
      "Epoch 38/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 765140160.0000 - rmse: 27661.1680 - val_loss: 446899552.0000 - val_rmse: 21139.9980\n",
      "Epoch 39/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 831771584.0000 - rmse: 28840.4512 - val_loss: 611455168.0000 - val_rmse: 24727.6191\n",
      "Epoch 40/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 897452032.0000 - rmse: 29957.5039 - val_loss: 468957984.0000 - val_rmse: 21655.4375\n",
      "Epoch 41/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 757578624.0000 - rmse: 27524.1465 - val_loss: 457176384.0000 - val_rmse: 21381.6836\n",
      "Epoch 42/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 768196672.0000 - rmse: 27716.3613 - val_loss: 442752672.0000 - val_rmse: 21041.6895\n",
      "Epoch 43/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 810273600.0000 - rmse: 28465.3047 - val_loss: 462280000.0000 - val_rmse: 21500.6973\n",
      "Epoch 44/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 804123072.0000 - rmse: 28357.0645 - val_loss: 471245312.0000 - val_rmse: 21708.1855\n",
      "Epoch 45/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 681257408.0000 - rmse: 26100.9082 - val_loss: 448478432.0000 - val_rmse: 21177.3086\n",
      "Epoch 46/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 817473088.0000 - rmse: 28591.4863 - val_loss: 487294144.0000 - val_rmse: 22074.7402\n",
      "Epoch 47/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 819937664.0000 - rmse: 28634.5527 - val_loss: 447777536.0000 - val_rmse: 21160.7539\n",
      "Epoch 48/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 752197504.0000 - rmse: 27426.2188 - val_loss: 448570784.0000 - val_rmse: 21179.4902\n",
      "Epoch 49/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 756144320.0000 - rmse: 27498.0781 - val_loss: 480853600.0000 - val_rmse: 21928.3750\n",
      "Epoch 50/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 791295808.0000 - rmse: 28129.9805 - val_loss: 434726272.0000 - val_rmse: 20850.0898\n",
      "Epoch 51/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 742617792.0000 - rmse: 27251.0137 - val_loss: 458829344.0000 - val_rmse: 21420.3027\n",
      "Epoch 52/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 697108544.0000 - rmse: 26402.8125 - val_loss: 441212928.0000 - val_rmse: 21005.0684\n",
      "Epoch 53/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 672996480.0000 - rmse: 25942.1758 - val_loss: 412282304.0000 - val_rmse: 20304.7363\n",
      "Epoch 54/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 722922560.0000 - rmse: 26887.2188 - val_loss: 453473344.0000 - val_rmse: 21294.9141\n",
      "Epoch 55/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 615453312.0000 - rmse: 24808.3320 - val_loss: 420832160.0000 - val_rmse: 20514.1934\n",
      "Epoch 56/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 706319680.0000 - rmse: 26576.6758 - val_loss: 407715712.0000 - val_rmse: 20191.9707\n",
      "Epoch 57/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 745859264.0000 - rmse: 27310.4238 - val_loss: 465732640.0000 - val_rmse: 21580.8398\n",
      "Epoch 58/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 768682112.0000 - rmse: 27725.1172 - val_loss: 501146656.0000 - val_rmse: 22386.3047\n",
      "Epoch 59/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 740098368.0000 - rmse: 27204.7480 - val_loss: 435316960.0000 - val_rmse: 20864.2500\n",
      "Epoch 60/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 620657792.0000 - rmse: 24913.0039 - val_loss: 420332000.0000 - val_rmse: 20502.0000\n",
      "Epoch 61/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 651213824.0000 - rmse: 25518.8906 - val_loss: 498817632.0000 - val_rmse: 22334.2246\n",
      "Epoch 62/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 665563776.0000 - rmse: 25798.5234 - val_loss: 401958560.0000 - val_rmse: 20048.9043\n",
      "Epoch 63/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 705663936.0000 - rmse: 26564.3359 - val_loss: 393411200.0000 - val_rmse: 19834.5957\n",
      "Epoch 64/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 674469952.0000 - rmse: 25970.5586 - val_loss: 410201248.0000 - val_rmse: 20253.4258\n",
      "Epoch 65/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 615387264.0000 - rmse: 24807.0000 - val_loss: 386192576.0000 - val_rmse: 19651.7832\n",
      "Epoch 66/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 662588416.0000 - rmse: 25740.7930 - val_loss: 404827968.0000 - val_rmse: 20120.3379\n",
      "Epoch 67/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 631580864.0000 - rmse: 25131.2734 - val_loss: 489812448.0000 - val_rmse: 22131.7070\n",
      "Epoch 68/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 689456256.0000 - rmse: 26257.5000 - val_loss: 411431200.0000 - val_rmse: 20283.7676\n",
      "Epoch 69/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 665931712.0000 - rmse: 25805.6523 - val_loss: 378894016.0000 - val_rmse: 19465.1992\n",
      "Epoch 70/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 634492480.0000 - rmse: 25189.1348 - val_loss: 415549888.0000 - val_rmse: 20385.0410\n",
      "Epoch 71/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 686726144.0000 - rmse: 26205.4609 - val_loss: 380028928.0000 - val_rmse: 19494.3301\n",
      "Epoch 72/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 616178432.0000 - rmse: 24822.9414 - val_loss: 395074592.0000 - val_rmse: 19876.4824\n",
      "Epoch 73/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 645272768.0000 - rmse: 25402.2207 - val_loss: 417706368.0000 - val_rmse: 20437.8652\n",
      "Epoch 74/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 626503936.0000 - rmse: 25030.0605 - val_loss: 410030080.0000 - val_rmse: 20249.1992\n",
      "Epoch 75/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 620547712.0000 - rmse: 24910.7949 - val_loss: 489142976.0000 - val_rmse: 22116.5762\n",
      "Epoch 76/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 622396224.0000 - rmse: 24947.8711 - val_loss: 374121312.0000 - val_rmse: 19342.2148\n",
      "Epoch 77/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 617613184.0000 - rmse: 24851.8242 - val_loss: 388916288.0000 - val_rmse: 19720.9609\n",
      "Epoch 78/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 619621184.0000 - rmse: 24892.1914 - val_loss: 368444384.0000 - val_rmse: 19194.9043\n",
      "Epoch 79/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 624052608.0000 - rmse: 24981.0449 - val_loss: 442314720.0000 - val_rmse: 21031.2793\n",
      "Epoch 80/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 605589952.0000 - rmse: 24608.7383 - val_loss: 369877888.0000 - val_rmse: 19232.2090\n",
      "Epoch 81/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 614295360.0000 - rmse: 24784.9824 - val_loss: 505893984.0000 - val_rmse: 22492.0879\n",
      "Epoch 82/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 618754944.0000 - rmse: 24874.7852 - val_loss: 364959200.0000 - val_rmse: 19103.9062\n",
      "Epoch 83/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 651122176.0000 - rmse: 25517.0957 - val_loss: 371341088.0000 - val_rmse: 19270.2129\n",
      "Epoch 84/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 595191552.0000 - rmse: 24396.5488 - val_loss: 393902720.0000 - val_rmse: 19846.9824\n",
      "Epoch 85/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 574393088.0000 - rmse: 23966.5000 - val_loss: 377665824.0000 - val_rmse: 19433.6270\n",
      "Epoch 86/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 536862272.0000 - rmse: 23170.2891 - val_loss: 361256608.0000 - val_rmse: 19006.7520\n",
      "Epoch 87/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 575402368.0000 - rmse: 23987.5469 - val_loss: 376236192.0000 - val_rmse: 19396.8086\n",
      "Epoch 88/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 573352320.0000 - rmse: 23944.7773 - val_loss: 363503456.0000 - val_rmse: 19065.7656\n",
      "Epoch 89/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 543503808.0000 - rmse: 23313.1680 - val_loss: 409358976.0000 - val_rmse: 20232.6211\n",
      "Epoch 90/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 620337728.0000 - rmse: 24906.5801 - val_loss: 349693120.0000 - val_rmse: 18700.0840\n",
      "Epoch 91/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 593022208.0000 - rmse: 24352.0469 - val_loss: 390711552.0000 - val_rmse: 19766.4258\n",
      "Epoch 92/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 555820288.0000 - rmse: 23575.8418 - val_loss: 381055104.0000 - val_rmse: 19520.6328\n",
      "Epoch 93/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 592744704.0000 - rmse: 24346.3496 - val_loss: 450195552.0000 - val_rmse: 21217.8125\n",
      "Epoch 94/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 577584768.0000 - rmse: 24032.9941 - val_loss: 383377696.0000 - val_rmse: 19580.0332\n",
      "Epoch 95/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 561376000.0000 - rmse: 23693.3750 - val_loss: 354831200.0000 - val_rmse: 18836.9629\n",
      "Epoch 96/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 566345216.0000 - rmse: 23798.0078 - val_loss: 382538880.0000 - val_rmse: 19558.6016\n",
      "Epoch 97/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 512319008.0000 - rmse: 22634.4648 - val_loss: 346179936.0000 - val_rmse: 18605.9121\n",
      "Epoch 98/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 608708864.0000 - rmse: 24672.0254 - val_loss: 366402016.0000 - val_rmse: 19141.6309\n",
      "Epoch 99/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 603187776.0000 - rmse: 24559.8809 - val_loss: 514103488.0000 - val_rmse: 22673.8496\n",
      "Epoch 100/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 550755840.0000 - rmse: 23468.1875 - val_loss: 352888192.0000 - val_rmse: 18785.3184\n",
      "Epoch 101/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 483182560.0000 - rmse: 21981.4141 - val_loss: 393888960.0000 - val_rmse: 19846.6367\n",
      "Epoch 102/400\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "166/166 [==============================] - 0s 2ms/step - loss: 531358176.0000 - rmse: 23051.2070 - val_loss: 413447680.0000 - val_rmse: 20333.4121\n",
      "Epoch 103/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 516528384.0000 - rmse: 22727.2617 - val_loss: 522949184.0000 - val_rmse: 22868.0820\n",
      "Epoch 104/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 591684224.0000 - rmse: 24324.5605 - val_loss: 344719808.0000 - val_rmse: 18566.6309\n",
      "Epoch 105/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 544277184.0000 - rmse: 23329.7480 - val_loss: 412650432.0000 - val_rmse: 20313.7988\n",
      "Epoch 106/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 562472832.0000 - rmse: 23716.5098 - val_loss: 363928384.0000 - val_rmse: 19076.9062\n",
      "Epoch 107/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 526391168.0000 - rmse: 22943.2168 - val_loss: 390885408.0000 - val_rmse: 19770.8223\n",
      "Epoch 108/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 497088128.0000 - rmse: 22295.4727 - val_loss: 415995808.0000 - val_rmse: 20395.9746\n",
      "Epoch 109/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 516567840.0000 - rmse: 22728.1289 - val_loss: 351343616.0000 - val_rmse: 18744.1621\n",
      "Epoch 110/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 484734048.0000 - rmse: 22016.6758 - val_loss: 360811776.0000 - val_rmse: 18995.0469\n",
      "Epoch 111/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 509987712.0000 - rmse: 22582.9082 - val_loss: 359288640.0000 - val_rmse: 18954.9102\n",
      "Epoch 112/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 495575168.0000 - rmse: 22261.5176 - val_loss: 346516576.0000 - val_rmse: 18614.9551\n",
      "Epoch 113/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 543621888.0000 - rmse: 23315.7012 - val_loss: 430294528.0000 - val_rmse: 20743.5410\n",
      "Epoch 114/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 563184512.0000 - rmse: 23731.5098 - val_loss: 354123296.0000 - val_rmse: 18818.1641\n",
      "Epoch 115/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 552800960.0000 - rmse: 23511.7188 - val_loss: 345662848.0000 - val_rmse: 18592.0098\n",
      "Epoch 116/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 578351232.0000 - rmse: 24048.9336 - val_loss: 404960640.0000 - val_rmse: 20123.6348\n",
      "Epoch 117/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 511480384.0000 - rmse: 22615.9316 - val_loss: 370080576.0000 - val_rmse: 19237.4785\n",
      "Epoch 118/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 549119360.0000 - rmse: 23433.2969 - val_loss: 346345248.0000 - val_rmse: 18610.3535\n",
      "Epoch 119/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 520710720.0000 - rmse: 22819.0859 - val_loss: 390602400.0000 - val_rmse: 19763.6641\n",
      "Epoch 120/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 546681856.0000 - rmse: 23381.2285 - val_loss: 350256640.0000 - val_rmse: 18715.1445\n",
      "Epoch 121/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 532866784.0000 - rmse: 23083.9082 - val_loss: 374766208.0000 - val_rmse: 19358.8789\n",
      "Epoch 122/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 465847328.0000 - rmse: 21583.4961 - val_loss: 351960768.0000 - val_rmse: 18760.6172\n",
      "Epoch 123/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 481874048.0000 - rmse: 21951.6289 - val_loss: 345358912.0000 - val_rmse: 18583.8340\n",
      "Epoch 124/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 529389088.0000 - rmse: 23008.4570 - val_loss: 349060960.0000 - val_rmse: 18683.1738\n",
      "Epoch 125/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 529727232.0000 - rmse: 23015.8047 - val_loss: 333409856.0000 - val_rmse: 18259.5137\n",
      "Epoch 126/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 547329664.0000 - rmse: 23395.0781 - val_loss: 343403840.0000 - val_rmse: 18531.1582\n",
      "Epoch 127/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 505117056.0000 - rmse: 22474.8086 - val_loss: 383870752.0000 - val_rmse: 19592.6191\n",
      "Epoch 128/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 500449920.0000 - rmse: 22370.7383 - val_loss: 417721440.0000 - val_rmse: 20438.2344\n",
      "Epoch 129/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 504385312.0000 - rmse: 22458.5234 - val_loss: 395343328.0000 - val_rmse: 19883.2422\n",
      "Epoch 130/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 518674784.0000 - rmse: 22774.4336 - val_loss: 358228160.0000 - val_rmse: 18926.9160\n",
      "Epoch 131/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 481676128.0000 - rmse: 21947.1211 - val_loss: 367707584.0000 - val_rmse: 19175.7031\n",
      "Epoch 132/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 456937152.0000 - rmse: 21376.0879 - val_loss: 405349664.0000 - val_rmse: 20133.2969\n",
      "Epoch 133/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 474195488.0000 - rmse: 21776.0293 - val_loss: 338954944.0000 - val_rmse: 18410.7285\n",
      "Epoch 134/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 513786496.0000 - rmse: 22666.8594 - val_loss: 360672416.0000 - val_rmse: 18991.3770\n",
      "Epoch 135/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 523878336.0000 - rmse: 22888.3887 - val_loss: 425690720.0000 - val_rmse: 20632.2734\n",
      "Epoch 136/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 430207648.0000 - rmse: 20741.4473 - val_loss: 411007872.0000 - val_rmse: 20273.3281\n",
      "Epoch 137/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 498695072.0000 - rmse: 22331.4824 - val_loss: 463439456.0000 - val_rmse: 21527.6445\n",
      "Epoch 138/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 475162176.0000 - rmse: 21798.2148 - val_loss: 358415776.0000 - val_rmse: 18931.8711\n",
      "Epoch 139/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 509764928.0000 - rmse: 22577.9746 - val_loss: 357531136.0000 - val_rmse: 18908.4941\n",
      "Epoch 140/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 475674368.0000 - rmse: 21809.9609 - val_loss: 330206144.0000 - val_rmse: 18171.5742\n",
      "Epoch 141/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 446493056.0000 - rmse: 21130.3828 - val_loss: 328724416.0000 - val_rmse: 18130.7598\n",
      "Epoch 142/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 498389664.0000 - rmse: 22324.6426 - val_loss: 513715392.0000 - val_rmse: 22665.2910\n",
      "Epoch 143/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 456023456.0000 - rmse: 21354.7051 - val_loss: 339862624.0000 - val_rmse: 18435.3633\n",
      "Epoch 144/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 491733440.0000 - rmse: 22175.0625 - val_loss: 409238464.0000 - val_rmse: 20229.6426\n",
      "Epoch 145/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 440579872.0000 - rmse: 20989.9941 - val_loss: 436044352.0000 - val_rmse: 20881.6758\n",
      "Epoch 146/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 463852192.0000 - rmse: 21537.2285 - val_loss: 334838912.0000 - val_rmse: 18298.6035\n",
      "Epoch 147/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 491584896.0000 - rmse: 22171.7148 - val_loss: 355367232.0000 - val_rmse: 18851.1855\n",
      "Epoch 148/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 433480928.0000 - rmse: 20820.2051 - val_loss: 484797408.0000 - val_rmse: 22018.1152\n",
      "Epoch 149/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 469145952.0000 - rmse: 21659.7773 - val_loss: 393851392.0000 - val_rmse: 19845.6895\n",
      "Epoch 150/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 517097248.0000 - rmse: 22739.7715 - val_loss: 354249440.0000 - val_rmse: 18821.5156\n",
      "Epoch 151/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 425712224.0000 - rmse: 20632.7949 - val_loss: 326425184.0000 - val_rmse: 18067.2402\n",
      "Epoch 152/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 486314688.0000 - rmse: 22052.5430 - val_loss: 339792000.0000 - val_rmse: 18433.4473\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 153/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 408606112.0000 - rmse: 20214.0078 - val_loss: 334719776.0000 - val_rmse: 18295.3477\n",
      "Epoch 154/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 456789376.0000 - rmse: 21372.6309 - val_loss: 381417120.0000 - val_rmse: 19529.9023\n",
      "Epoch 155/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 444852352.0000 - rmse: 21091.5234 - val_loss: 374731840.0000 - val_rmse: 19357.9922\n",
      "Epoch 156/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 441107104.0000 - rmse: 21002.5508 - val_loss: 353701408.0000 - val_rmse: 18806.9512\n",
      "Epoch 157/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 432097216.0000 - rmse: 20786.9473 - val_loss: 334553440.0000 - val_rmse: 18290.8027\n",
      "Epoch 158/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 481845568.0000 - rmse: 21950.9805 - val_loss: 351915968.0000 - val_rmse: 18759.4238\n",
      "Epoch 159/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 463268544.0000 - rmse: 21523.6738 - val_loss: 360791520.0000 - val_rmse: 18994.5137\n",
      "Epoch 160/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 426782880.0000 - rmse: 20658.7246 - val_loss: 353641184.0000 - val_rmse: 18805.3496\n",
      "Epoch 161/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 425658656.0000 - rmse: 20631.4961 - val_loss: 447619968.0000 - val_rmse: 21157.0312\n",
      "Epoch 162/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 486159392.0000 - rmse: 22049.0234 - val_loss: 343796768.0000 - val_rmse: 18541.7578\n",
      "Epoch 163/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 488710144.0000 - rmse: 22106.7891 - val_loss: 368786112.0000 - val_rmse: 19203.8047\n",
      "Epoch 164/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 427550336.0000 - rmse: 20677.2910 - val_loss: 381033728.0000 - val_rmse: 19520.0859\n",
      "Epoch 165/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 402517952.0000 - rmse: 20062.8496 - val_loss: 346819424.0000 - val_rmse: 18623.0879\n",
      "Epoch 166/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 515011168.0000 - rmse: 22693.8574 - val_loss: 346144704.0000 - val_rmse: 18604.9648\n",
      "Epoch 167/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 411124096.0000 - rmse: 20276.1953 - val_loss: 334735360.0000 - val_rmse: 18295.7734\n",
      "Epoch 168/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 460730144.0000 - rmse: 21464.6250 - val_loss: 378631776.0000 - val_rmse: 19458.4629\n",
      "Epoch 169/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 460988704.0000 - rmse: 21470.6484 - val_loss: 506913728.0000 - val_rmse: 22514.7441\n",
      "Epoch 170/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 417392064.0000 - rmse: 20430.1758 - val_loss: 382813280.0000 - val_rmse: 19565.6152\n",
      "Epoch 171/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 405334048.0000 - rmse: 20132.9102 - val_loss: 354210656.0000 - val_rmse: 18820.4844\n",
      "Epoch 172/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 442937216.0000 - rmse: 21046.0742 - val_loss: 353855072.0000 - val_rmse: 18811.0352\n",
      "Epoch 173/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 398774048.0000 - rmse: 19969.3281 - val_loss: 333960512.0000 - val_rmse: 18274.5859\n",
      "Epoch 174/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 433458624.0000 - rmse: 20819.6699 - val_loss: 330001984.0000 - val_rmse: 18165.9570\n",
      "Epoch 175/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 456616032.0000 - rmse: 21368.5762 - val_loss: 381469440.0000 - val_rmse: 19531.2422\n",
      "Epoch 176/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 411764480.0000 - rmse: 20291.9805 - val_loss: 332993440.0000 - val_rmse: 18248.1074\n",
      "Epoch 177/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 387552960.0000 - rmse: 19686.3652 - val_loss: 362518528.0000 - val_rmse: 19039.9199\n",
      "Epoch 178/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 449410464.0000 - rmse: 21199.3027 - val_loss: 337268800.0000 - val_rmse: 18364.8789\n",
      "Epoch 179/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 425113664.0000 - rmse: 20618.2852 - val_loss: 363853152.0000 - val_rmse: 19074.9355\n",
      "Epoch 180/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 421597472.0000 - rmse: 20532.8398 - val_loss: 338176928.0000 - val_rmse: 18389.5879\n",
      "Epoch 181/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 394036672.0000 - rmse: 19850.3574 - val_loss: 348018624.0000 - val_rmse: 18655.2578\n",
      "Epoch 182/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 424219808.0000 - rmse: 20596.5977 - val_loss: 349550176.0000 - val_rmse: 18696.2617\n",
      "Epoch 183/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 417003456.0000 - rmse: 20420.6621 - val_loss: 379029280.0000 - val_rmse: 19468.6738\n",
      "Epoch 184/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 456321248.0000 - rmse: 21361.6777 - val_loss: 395084256.0000 - val_rmse: 19876.7266\n",
      "Epoch 185/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 445030176.0000 - rmse: 21095.7383 - val_loss: 353838624.0000 - val_rmse: 18810.5996\n",
      "Epoch 186/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 428612448.0000 - rmse: 20702.9570 - val_loss: 347158464.0000 - val_rmse: 18632.1875\n",
      "Epoch 187/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 408431328.0000 - rmse: 20209.6836 - val_loss: 370925536.0000 - val_rmse: 19259.4277\n",
      "Epoch 188/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 473618336.0000 - rmse: 21762.7734 - val_loss: 331571328.0000 - val_rmse: 18209.0996\n",
      "Epoch 189/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 382416896.0000 - rmse: 19555.4824 - val_loss: 401739008.0000 - val_rmse: 20043.4277\n",
      "Epoch 190/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 397659040.0000 - rmse: 19941.3906 - val_loss: 381083616.0000 - val_rmse: 19521.3633\n",
      "Epoch 191/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 409024672.0000 - rmse: 20224.3574 - val_loss: 344091392.0000 - val_rmse: 18549.7012\n",
      "Epoch 192/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 425719872.0000 - rmse: 20632.9805 - val_loss: 342137600.0000 - val_rmse: 18496.9609\n",
      "Epoch 193/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 379959936.0000 - rmse: 19492.5605 - val_loss: 328592512.0000 - val_rmse: 18127.1211\n",
      "Epoch 194/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 391503616.0000 - rmse: 19786.4512 - val_loss: 356361504.0000 - val_rmse: 18877.5391\n",
      "Epoch 195/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 394641184.0000 - rmse: 19865.5781 - val_loss: 365368128.0000 - val_rmse: 19114.6055\n",
      "Epoch 196/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 413177376.0000 - rmse: 20326.7656 - val_loss: 459750304.0000 - val_rmse: 21441.7891\n",
      "Epoch 197/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 447978272.0000 - rmse: 21165.4980 - val_loss: 334393536.0000 - val_rmse: 18286.4297\n",
      "Epoch 198/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 415174112.0000 - rmse: 20375.8223 - val_loss: 345417600.0000 - val_rmse: 18585.4141\n",
      "Epoch 199/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 415944256.0000 - rmse: 20394.7109 - val_loss: 361277600.0000 - val_rmse: 19007.3047\n",
      "Epoch 200/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 390859264.0000 - rmse: 19770.1602 - val_loss: 331910656.0000 - val_rmse: 18218.4160\n",
      "Epoch 201/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 371854656.0000 - rmse: 19283.5332 - val_loss: 338965792.0000 - val_rmse: 18411.0234\n",
      "Epoch 202/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 417338080.0000 - rmse: 20428.8535 - val_loss: 351615936.0000 - val_rmse: 18751.4258\n",
      "Epoch 203/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 390045408.0000 - rmse: 19749.5664 - val_loss: 334700768.0000 - val_rmse: 18294.8281\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 204/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 329685472.0000 - rmse: 18157.2422 - val_loss: 341001888.0000 - val_rmse: 18466.2363\n",
      "Epoch 205/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 369600736.0000 - rmse: 19225.0020 - val_loss: 334344064.0000 - val_rmse: 18285.0781\n",
      "Epoch 206/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 369610336.0000 - rmse: 19225.2520 - val_loss: 319799712.0000 - val_rmse: 17882.9453\n",
      "Epoch 207/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 355543744.0000 - rmse: 18855.8672 - val_loss: 343923680.0000 - val_rmse: 18545.1797\n",
      "Epoch 208/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 388432096.0000 - rmse: 19708.6816 - val_loss: 331427712.0000 - val_rmse: 18205.1562\n",
      "Epoch 209/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 349283936.0000 - rmse: 18689.1387 - val_loss: 429767456.0000 - val_rmse: 20730.8340\n",
      "Epoch 210/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 450916704.0000 - rmse: 21234.7988 - val_loss: 345078848.0000 - val_rmse: 18576.2988\n",
      "Epoch 211/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 377022080.0000 - rmse: 19417.0566 - val_loss: 340830432.0000 - val_rmse: 18461.5938\n",
      "Epoch 212/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 362293568.0000 - rmse: 19034.0117 - val_loss: 348715712.0000 - val_rmse: 18673.9316\n",
      "Epoch 213/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 387849824.0000 - rmse: 19693.9023 - val_loss: 341852032.0000 - val_rmse: 18489.2402\n",
      "Epoch 214/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 368836640.0000 - rmse: 19205.1211 - val_loss: 341446496.0000 - val_rmse: 18478.2715\n",
      "Epoch 215/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 402512000.0000 - rmse: 20062.7012 - val_loss: 333461056.0000 - val_rmse: 18260.9160\n",
      "Epoch 216/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 380481856.0000 - rmse: 19505.9434 - val_loss: 373721536.0000 - val_rmse: 19331.8789\n",
      "Epoch 217/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 374735616.0000 - rmse: 19358.0898 - val_loss: 385157408.0000 - val_rmse: 19625.4277\n",
      "Epoch 218/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 340369504.0000 - rmse: 18449.1055 - val_loss: 331797760.0000 - val_rmse: 18215.3164\n",
      "Epoch 219/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 370385504.0000 - rmse: 19245.4023 - val_loss: 467171744.0000 - val_rmse: 21614.1562\n",
      "Epoch 220/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 355692896.0000 - rmse: 18859.8223 - val_loss: 339918624.0000 - val_rmse: 18436.8828\n",
      "Epoch 221/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 338861920.0000 - rmse: 18408.2031 - val_loss: 348413248.0000 - val_rmse: 18665.8301\n",
      "Epoch 222/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 346450016.0000 - rmse: 18613.1680 - val_loss: 352216288.0000 - val_rmse: 18767.4258\n",
      "Epoch 223/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 381469728.0000 - rmse: 19531.2500 - val_loss: 403778336.0000 - val_rmse: 20094.2363\n",
      "Epoch 224/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 357102048.0000 - rmse: 18897.1445 - val_loss: 352615776.0000 - val_rmse: 18778.0664\n",
      "Epoch 225/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 423298752.0000 - rmse: 20574.2246 - val_loss: 360351872.0000 - val_rmse: 18982.9355\n",
      "Epoch 226/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 344318848.0000 - rmse: 18555.8301 - val_loss: 361138816.0000 - val_rmse: 19003.6523\n",
      "Epoch 227/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 356538720.0000 - rmse: 18882.2324 - val_loss: 362212512.0000 - val_rmse: 19031.8809\n",
      "Epoch 228/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 361718016.0000 - rmse: 19018.8867 - val_loss: 349519392.0000 - val_rmse: 18695.4375\n",
      "Epoch 229/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 340031392.0000 - rmse: 18439.9395 - val_loss: 406072064.0000 - val_rmse: 20151.2305\n",
      "Epoch 230/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 362470656.0000 - rmse: 19038.6621 - val_loss: 347550240.0000 - val_rmse: 18642.6992\n",
      "Epoch 231/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 368450784.0000 - rmse: 19195.0723 - val_loss: 354358176.0000 - val_rmse: 18824.4043\n",
      "Epoch 232/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 398286144.0000 - rmse: 19957.1074 - val_loss: 357621792.0000 - val_rmse: 18910.8906\n",
      "Epoch 233/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 338291520.0000 - rmse: 18392.7031 - val_loss: 352593440.0000 - val_rmse: 18777.4707\n",
      "Epoch 234/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 339253568.0000 - rmse: 18418.8379 - val_loss: 326952480.0000 - val_rmse: 18081.8281\n",
      "Epoch 235/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 344298304.0000 - rmse: 18555.2773 - val_loss: 339608256.0000 - val_rmse: 18428.4629\n",
      "Epoch 236/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 291311680.0000 - rmse: 17067.8555 - val_loss: 351606656.0000 - val_rmse: 18751.1777\n",
      "Epoch 237/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 347089600.0000 - rmse: 18630.3418 - val_loss: 353943232.0000 - val_rmse: 18813.3789\n",
      "Epoch 238/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 386494624.0000 - rmse: 19659.4668 - val_loss: 424144640.0000 - val_rmse: 20594.7715\n",
      "Epoch 239/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 361741856.0000 - rmse: 19019.5117 - val_loss: 356913120.0000 - val_rmse: 18892.1445\n",
      "Epoch 240/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 353500800.0000 - rmse: 18801.6172 - val_loss: 338764096.0000 - val_rmse: 18405.5449\n",
      "Epoch 241/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 332561632.0000 - rmse: 18236.2715 - val_loss: 342580384.0000 - val_rmse: 18508.9277\n",
      "Epoch 242/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 339213024.0000 - rmse: 18417.7363 - val_loss: 339872096.0000 - val_rmse: 18435.6211\n",
      "Epoch 243/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 321677216.0000 - rmse: 17935.3613 - val_loss: 439070720.0000 - val_rmse: 20954.0137\n",
      "Epoch 244/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 324359296.0000 - rmse: 18009.9785 - val_loss: 335166560.0000 - val_rmse: 18307.5547\n",
      "Epoch 245/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 368160864.0000 - rmse: 19187.5176 - val_loss: 518491840.0000 - val_rmse: 22770.4160\n",
      "Epoch 246/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 323997376.0000 - rmse: 17999.9277 - val_loss: 366895904.0000 - val_rmse: 19154.5273\n",
      "Epoch 247/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 333204544.0000 - rmse: 18253.8906 - val_loss: 340907520.0000 - val_rmse: 18463.6816\n",
      "Epoch 248/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 318413312.0000 - rmse: 17844.1387 - val_loss: 372236448.0000 - val_rmse: 19293.4297\n",
      "Epoch 249/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 347465984.0000 - rmse: 18640.4395 - val_loss: 331903168.0000 - val_rmse: 18218.2090\n",
      "Epoch 250/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 320454752.0000 - rmse: 17901.2500 - val_loss: 386735680.0000 - val_rmse: 19665.5957\n",
      "Epoch 251/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 317276896.0000 - rmse: 17812.2676 - val_loss: 335583552.0000 - val_rmse: 18318.9395\n",
      "Epoch 252/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 400797952.0000 - rmse: 20019.9395 - val_loss: 434424288.0000 - val_rmse: 20842.8477\n",
      "Epoch 253/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 372377120.0000 - rmse: 19297.0762 - val_loss: 360262784.0000 - val_rmse: 18980.5898\n",
      "Epoch 254/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 340517376.0000 - rmse: 18453.1133 - val_loss: 326790176.0000 - val_rmse: 18077.3379\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 255/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 328549664.0000 - rmse: 18125.9395 - val_loss: 355922944.0000 - val_rmse: 18865.9199\n",
      "Epoch 256/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 342429472.0000 - rmse: 18504.8496 - val_loss: 357666240.0000 - val_rmse: 18912.0664\n",
      "Epoch 257/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 335231840.0000 - rmse: 18309.3379 - val_loss: 427471712.0000 - val_rmse: 20675.3887\n",
      "Epoch 258/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 312136928.0000 - rmse: 17667.3965 - val_loss: 337344544.0000 - val_rmse: 18366.9414\n",
      "Epoch 259/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 343485920.0000 - rmse: 18533.3730 - val_loss: 334127680.0000 - val_rmse: 18279.1602\n",
      "Epoch 260/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 343898272.0000 - rmse: 18544.4941 - val_loss: 382001664.0000 - val_rmse: 19544.8633\n",
      "Epoch 261/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 322442752.0000 - rmse: 17956.6914 - val_loss: 337499200.0000 - val_rmse: 18371.1504\n",
      "Epoch 262/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 295033952.0000 - rmse: 17176.5527 - val_loss: 321211200.0000 - val_rmse: 17922.3652\n",
      "Epoch 263/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 322579264.0000 - rmse: 17960.4922 - val_loss: 323388608.0000 - val_rmse: 17983.0098\n",
      "Epoch 264/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 339144224.0000 - rmse: 18415.8691 - val_loss: 361879712.0000 - val_rmse: 19023.1367\n",
      "Epoch 265/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 340185664.0000 - rmse: 18444.1230 - val_loss: 342250048.0000 - val_rmse: 18500.0020\n",
      "Epoch 266/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 312870272.0000 - rmse: 17688.1387 - val_loss: 349754784.0000 - val_rmse: 18701.7324\n",
      "Epoch 267/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 283093376.0000 - rmse: 16825.3789 - val_loss: 384621632.0000 - val_rmse: 19611.7734\n",
      "Epoch 268/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 293205664.0000 - rmse: 17123.2500 - val_loss: 352164928.0000 - val_rmse: 18766.0586\n",
      "Epoch 269/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 331702304.0000 - rmse: 18212.6953 - val_loss: 382879520.0000 - val_rmse: 19567.3066\n",
      "Epoch 270/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 353298048.0000 - rmse: 18796.2246 - val_loss: 340193088.0000 - val_rmse: 18444.3242\n",
      "Epoch 271/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 324427904.0000 - rmse: 18011.8828 - val_loss: 351171200.0000 - val_rmse: 18739.5625\n",
      "Epoch 272/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 339425536.0000 - rmse: 18423.5059 - val_loss: 360524768.0000 - val_rmse: 18987.4902\n",
      "Epoch 273/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 296010816.0000 - rmse: 17204.9648 - val_loss: 345984256.0000 - val_rmse: 18600.6523\n",
      "Epoch 274/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 327358752.0000 - rmse: 18093.0586 - val_loss: 374531264.0000 - val_rmse: 19352.8105\n",
      "Epoch 275/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 292011296.0000 - rmse: 17088.3379 - val_loss: 352575648.0000 - val_rmse: 18776.9980\n",
      "Epoch 276/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 295936416.0000 - rmse: 17202.8027 - val_loss: 352546400.0000 - val_rmse: 18776.2188\n",
      "Epoch 277/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 335961664.0000 - rmse: 18329.2578 - val_loss: 391485152.0000 - val_rmse: 19785.9844\n",
      "Epoch 278/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 299262144.0000 - rmse: 17299.1953 - val_loss: 327280096.0000 - val_rmse: 18090.8848\n",
      "Epoch 279/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 338288064.0000 - rmse: 18392.6094 - val_loss: 375716512.0000 - val_rmse: 19383.4082\n",
      "Epoch 280/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 313923520.0000 - rmse: 17717.8867 - val_loss: 333364896.0000 - val_rmse: 18258.2832\n",
      "Epoch 281/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 329848256.0000 - rmse: 18161.7246 - val_loss: 338061504.0000 - val_rmse: 18386.4492\n",
      "Epoch 282/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 316951008.0000 - rmse: 17803.1172 - val_loss: 331569504.0000 - val_rmse: 18209.0508\n",
      "Epoch 283/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 290666912.0000 - rmse: 17048.9570 - val_loss: 325126176.0000 - val_rmse: 18031.2559\n",
      "Epoch 284/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 316552352.0000 - rmse: 17791.9180 - val_loss: 346969184.0000 - val_rmse: 18627.1094\n",
      "Epoch 285/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 296016544.0000 - rmse: 17205.1309 - val_loss: 337807552.0000 - val_rmse: 18379.5410\n",
      "Epoch 286/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 299637792.0000 - rmse: 17310.0488 - val_loss: 381888576.0000 - val_rmse: 19541.9688\n",
      "Epoch 287/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 297539904.0000 - rmse: 17249.3457 - val_loss: 349555488.0000 - val_rmse: 18696.4023\n",
      "Epoch 288/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 335712768.0000 - rmse: 18322.4668 - val_loss: 351794816.0000 - val_rmse: 18756.1934\n",
      "Epoch 289/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 275239968.0000 - rmse: 16590.3574 - val_loss: 325931648.0000 - val_rmse: 18053.5781\n",
      "Epoch 290/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 317688480.0000 - rmse: 17823.8184 - val_loss: 350307648.0000 - val_rmse: 18716.5078\n",
      "Epoch 291/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 272334656.0000 - rmse: 16502.5645 - val_loss: 369271904.0000 - val_rmse: 19216.4492\n",
      "Epoch 292/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 281891712.0000 - rmse: 16789.6309 - val_loss: 335594176.0000 - val_rmse: 18319.2305\n",
      "Epoch 293/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 273054112.0000 - rmse: 16524.3496 - val_loss: 411904288.0000 - val_rmse: 20295.4258\n",
      "Epoch 294/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 282531360.0000 - rmse: 16808.6699 - val_loss: 359641632.0000 - val_rmse: 18964.2207\n",
      "Epoch 295/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 286907680.0000 - rmse: 16938.3496 - val_loss: 375488896.0000 - val_rmse: 19377.5352\n",
      "Epoch 296/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 293429728.0000 - rmse: 17129.7910 - val_loss: 354571872.0000 - val_rmse: 18830.0781\n",
      "Epoch 297/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 295840736.0000 - rmse: 17200.0215 - val_loss: 350927712.0000 - val_rmse: 18733.0645\n",
      "Epoch 298/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 292028896.0000 - rmse: 17088.8535 - val_loss: 314033664.0000 - val_rmse: 17720.9941\n",
      "Epoch 299/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 291444832.0000 - rmse: 17071.7559 - val_loss: 378709280.0000 - val_rmse: 19460.4551\n",
      "Epoch 300/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 333731136.0000 - rmse: 18268.3105 - val_loss: 320064576.0000 - val_rmse: 17890.3496\n",
      "Epoch 301/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 287119552.0000 - rmse: 16944.6016 - val_loss: 328844608.0000 - val_rmse: 18134.0723\n",
      "Epoch 302/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 308052672.0000 - rmse: 17551.4297 - val_loss: 346853824.0000 - val_rmse: 18624.0117\n",
      "Epoch 303/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 315060096.0000 - rmse: 17749.9316 - val_loss: 375769056.0000 - val_rmse: 19384.7637\n",
      "Epoch 304/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 304549024.0000 - rmse: 17451.3340 - val_loss: 331494176.0000 - val_rmse: 18206.9824\n",
      "Epoch 305/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 269084800.0000 - rmse: 16403.8047 - val_loss: 353732128.0000 - val_rmse: 18807.7676\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 306/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 301304096.0000 - rmse: 17358.1133 - val_loss: 426145824.0000 - val_rmse: 20643.2988\n",
      "Epoch 307/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 292568384.0000 - rmse: 17104.6309 - val_loss: 423445984.0000 - val_rmse: 20577.8027\n",
      "Epoch 308/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 285936864.0000 - rmse: 16909.6680 - val_loss: 331964128.0000 - val_rmse: 18219.8828\n",
      "Epoch 309/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 270434464.0000 - rmse: 16444.8926 - val_loss: 323561856.0000 - val_rmse: 17987.8262\n",
      "Epoch 310/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 257425728.0000 - rmse: 16044.4922 - val_loss: 372028992.0000 - val_rmse: 19288.0527\n",
      "Epoch 311/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 278032896.0000 - rmse: 16674.3184 - val_loss: 327426016.0000 - val_rmse: 18094.9160\n",
      "Epoch 312/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 271962848.0000 - rmse: 16491.2969 - val_loss: 312866528.0000 - val_rmse: 17688.0332\n",
      "Epoch 313/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 297293120.0000 - rmse: 17242.1895 - val_loss: 317152288.0000 - val_rmse: 17808.7695\n",
      "Epoch 314/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 322760320.0000 - rmse: 17965.5312 - val_loss: 330223520.0000 - val_rmse: 18172.0527\n",
      "Epoch 315/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 265653168.0000 - rmse: 16298.8701 - val_loss: 302267968.0000 - val_rmse: 17385.8555\n",
      "Epoch 316/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 290521440.0000 - rmse: 17044.6895 - val_loss: 345765504.0000 - val_rmse: 18594.7715\n",
      "Epoch 317/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 274054176.0000 - rmse: 16554.5820 - val_loss: 313278464.0000 - val_rmse: 17699.6738\n",
      "Epoch 318/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 261453056.0000 - rmse: 16169.5098 - val_loss: 351242656.0000 - val_rmse: 18741.4688\n",
      "Epoch 319/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 287191456.0000 - rmse: 16946.7246 - val_loss: 336187680.0000 - val_rmse: 18335.4219\n",
      "Epoch 320/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 314113248.0000 - rmse: 17723.2402 - val_loss: 329045824.0000 - val_rmse: 18139.6211\n",
      "Epoch 321/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 290983200.0000 - rmse: 17058.2305 - val_loss: 363988864.0000 - val_rmse: 19078.4922\n",
      "Epoch 322/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 256923296.0000 - rmse: 16028.8271 - val_loss: 334867040.0000 - val_rmse: 18299.3730\n",
      "Epoch 323/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 321388128.0000 - rmse: 17927.3008 - val_loss: 369838624.0000 - val_rmse: 19231.1895\n",
      "Epoch 324/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 296515520.0000 - rmse: 17219.6270 - val_loss: 310410112.0000 - val_rmse: 17618.4590\n",
      "Epoch 325/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 255484048.0000 - rmse: 15983.8682 - val_loss: 327656000.0000 - val_rmse: 18101.2715\n",
      "Epoch 326/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 297935808.0000 - rmse: 17260.8164 - val_loss: 328212864.0000 - val_rmse: 18116.6465\n",
      "Epoch 327/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 314081440.0000 - rmse: 17722.3438 - val_loss: 330255168.0000 - val_rmse: 18172.9238\n",
      "Epoch 328/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 270961248.0000 - rmse: 16460.9004 - val_loss: 320726208.0000 - val_rmse: 17908.8301\n",
      "Epoch 329/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 290426656.0000 - rmse: 17041.9082 - val_loss: 311815744.0000 - val_rmse: 17658.3047\n",
      "Epoch 330/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 262011072.0000 - rmse: 16186.7559 - val_loss: 306168704.0000 - val_rmse: 17497.6777\n",
      "Epoch 331/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 291498784.0000 - rmse: 17073.3359 - val_loss: 360125920.0000 - val_rmse: 18976.9844\n",
      "Epoch 332/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 307317568.0000 - rmse: 17530.4746 - val_loss: 310131744.0000 - val_rmse: 17610.5586\n",
      "Epoch 333/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 292266624.0000 - rmse: 17095.8066 - val_loss: 312571744.0000 - val_rmse: 17679.6992\n",
      "Epoch 334/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 276729952.0000 - rmse: 16635.2031 - val_loss: 341972992.0000 - val_rmse: 18492.5117\n",
      "Epoch 335/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 274444288.0000 - rmse: 16566.3594 - val_loss: 341464960.0000 - val_rmse: 18478.7715\n",
      "Epoch 336/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 256005648.0000 - rmse: 16000.1768 - val_loss: 337892224.0000 - val_rmse: 18381.8457\n",
      "Epoch 337/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 294115776.0000 - rmse: 17149.8047 - val_loss: 341567328.0000 - val_rmse: 18481.5410\n",
      "Epoch 338/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 271152992.0000 - rmse: 16466.7246 - val_loss: 322093856.0000 - val_rmse: 17946.9727\n",
      "Epoch 339/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 268177008.0000 - rmse: 16376.1113 - val_loss: 313880224.0000 - val_rmse: 17716.6660\n",
      "Epoch 340/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 282745056.0000 - rmse: 16815.0254 - val_loss: 380377728.0000 - val_rmse: 19503.2754\n",
      "Epoch 341/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 251180256.0000 - rmse: 15848.6670 - val_loss: 348955296.0000 - val_rmse: 18680.3457\n",
      "Epoch 342/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 308269632.0000 - rmse: 17557.6094 - val_loss: 553651904.0000 - val_rmse: 23529.8086\n",
      "Epoch 343/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 293458464.0000 - rmse: 17130.6289 - val_loss: 358175136.0000 - val_rmse: 18925.5156\n",
      "Epoch 344/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 244265456.0000 - rmse: 15628.9941 - val_loss: 338933472.0000 - val_rmse: 18410.1465\n",
      "Epoch 345/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 273001952.0000 - rmse: 16522.7715 - val_loss: 331232032.0000 - val_rmse: 18199.7812\n",
      "Epoch 346/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 271572800.0000 - rmse: 16479.4668 - val_loss: 353386208.0000 - val_rmse: 18798.5684\n",
      "Epoch 347/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 285747936.0000 - rmse: 16904.0801 - val_loss: 333863136.0000 - val_rmse: 18271.9219\n",
      "Epoch 348/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 283750944.0000 - rmse: 16844.9082 - val_loss: 438355712.0000 - val_rmse: 20936.9453\n",
      "Epoch 349/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 277355968.0000 - rmse: 16654.0078 - val_loss: 317133728.0000 - val_rmse: 17808.2480\n",
      "Epoch 350/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 238423072.0000 - rmse: 15440.9541 - val_loss: 317761344.0000 - val_rmse: 17825.8613\n",
      "Epoch 351/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 247360176.0000 - rmse: 15727.6885 - val_loss: 315222240.0000 - val_rmse: 17754.5000\n",
      "Epoch 352/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 245817168.0000 - rmse: 15678.5576 - val_loss: 333301792.0000 - val_rmse: 18256.5547\n",
      "Epoch 353/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 245303840.0000 - rmse: 15662.1787 - val_loss: 313543808.0000 - val_rmse: 17707.1680\n",
      "Epoch 354/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 273094336.0000 - rmse: 16525.5664 - val_loss: 330762272.0000 - val_rmse: 18186.8711\n",
      "Epoch 355/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 252951952.0000 - rmse: 15904.4629 - val_loss: 366148512.0000 - val_rmse: 19135.0078\n",
      "Epoch 356/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 261065600.0000 - rmse: 16157.5244 - val_loss: 320306016.0000 - val_rmse: 17897.0957\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 357/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 306679744.0000 - rmse: 17512.2734 - val_loss: 324534848.0000 - val_rmse: 18014.8516\n",
      "Epoch 358/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 258961456.0000 - rmse: 16092.2793 - val_loss: 323100320.0000 - val_rmse: 17974.9922\n",
      "Epoch 359/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 279380000.0000 - rmse: 16714.6641 - val_loss: 313171584.0000 - val_rmse: 17696.6543\n",
      "Epoch 360/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 236638416.0000 - rmse: 15383.0557 - val_loss: 330832960.0000 - val_rmse: 18188.8145\n",
      "Epoch 361/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 287453088.0000 - rmse: 16954.4414 - val_loss: 338419840.0000 - val_rmse: 18396.1914\n",
      "Epoch 362/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 267002288.0000 - rmse: 16340.2051 - val_loss: 346165824.0000 - val_rmse: 18605.5312\n",
      "Epoch 363/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 223275520.0000 - rmse: 14942.4072 - val_loss: 306993920.0000 - val_rmse: 17521.2422\n",
      "Epoch 364/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 267259840.0000 - rmse: 16348.0840 - val_loss: 345409600.0000 - val_rmse: 18585.1992\n",
      "Epoch 365/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 243122912.0000 - rmse: 15592.3994 - val_loss: 305186400.0000 - val_rmse: 17469.5859\n",
      "Epoch 366/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 284627136.0000 - rmse: 16870.8965 - val_loss: 323868288.0000 - val_rmse: 17996.3418\n",
      "Epoch 367/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 241139312.0000 - rmse: 15528.6611 - val_loss: 393425312.0000 - val_rmse: 19834.9512\n",
      "Epoch 368/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 261173408.0000 - rmse: 16160.8604 - val_loss: 326617152.0000 - val_rmse: 18072.5527\n",
      "Epoch 369/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 242712448.0000 - rmse: 15579.2314 - val_loss: 337312608.0000 - val_rmse: 18366.0723\n",
      "Epoch 370/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 294667872.0000 - rmse: 17165.8926 - val_loss: 381865024.0000 - val_rmse: 19541.3672\n",
      "Epoch 371/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 259915600.0000 - rmse: 16121.8984 - val_loss: 328883904.0000 - val_rmse: 18135.1562\n",
      "Epoch 372/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 278552096.0000 - rmse: 16689.8809 - val_loss: 356227584.0000 - val_rmse: 18873.9922\n",
      "Epoch 373/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 237729264.0000 - rmse: 15418.4717 - val_loss: 372629984.0000 - val_rmse: 19303.6270\n",
      "Epoch 374/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 252166256.0000 - rmse: 15879.7432 - val_loss: 325365312.0000 - val_rmse: 18037.8848\n",
      "Epoch 375/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 227371296.0000 - rmse: 15078.8359 - val_loss: 313176352.0000 - val_rmse: 17696.7891\n",
      "Epoch 376/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 277599456.0000 - rmse: 16661.3164 - val_loss: 343191040.0000 - val_rmse: 18525.4160\n",
      "Epoch 377/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 309338848.0000 - rmse: 17588.0312 - val_loss: 307617184.0000 - val_rmse: 17539.0195\n",
      "Epoch 378/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 246046096.0000 - rmse: 15685.8564 - val_loss: 311931872.0000 - val_rmse: 17661.5938\n",
      "Epoch 379/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 282418592.0000 - rmse: 16805.3145 - val_loss: 331107648.0000 - val_rmse: 18196.3633\n",
      "Epoch 380/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 239551360.0000 - rmse: 15477.4473 - val_loss: 338480448.0000 - val_rmse: 18397.8379\n",
      "Epoch 381/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 224204896.0000 - rmse: 14973.4727 - val_loss: 307479200.0000 - val_rmse: 17535.0840\n",
      "Epoch 382/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 241761616.0000 - rmse: 15548.6855 - val_loss: 318345088.0000 - val_rmse: 17842.2285\n",
      "Epoch 383/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 222641968.0000 - rmse: 14921.1924 - val_loss: 334781856.0000 - val_rmse: 18297.0449\n",
      "Epoch 384/400\n",
      "166/166 [==============================] - ETA: 0s - loss: 232868640.0000 - rmse: 15260.03 - 0s 2ms/step - loss: 272152864.0000 - rmse: 16497.0566 - val_loss: 393238944.0000 - val_rmse: 19830.2539\n",
      "Epoch 385/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 242460384.0000 - rmse: 15571.1396 - val_loss: 308521120.0000 - val_rmse: 17564.7695\n",
      "Epoch 386/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 240041552.0000 - rmse: 15493.2744 - val_loss: 308569696.0000 - val_rmse: 17566.1523\n",
      "Epoch 387/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 247152992.0000 - rmse: 15721.1006 - val_loss: 337211136.0000 - val_rmse: 18363.3086\n",
      "Epoch 388/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 254122880.0000 - rmse: 15941.2324 - val_loss: 305447488.0000 - val_rmse: 17477.0566\n",
      "Epoch 389/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 236854144.0000 - rmse: 15390.0664 - val_loss: 346845600.0000 - val_rmse: 18623.7910\n",
      "Epoch 390/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 234473072.0000 - rmse: 15312.5137 - val_loss: 397108800.0000 - val_rmse: 19927.5898\n",
      "Epoch 391/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 241639952.0000 - rmse: 15544.7725 - val_loss: 339130816.0000 - val_rmse: 18415.5039\n",
      "Epoch 392/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 238419280.0000 - rmse: 15440.8320 - val_loss: 339064512.0000 - val_rmse: 18413.7051\n",
      "Epoch 393/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 289353088.0000 - rmse: 17010.3809 - val_loss: 318052704.0000 - val_rmse: 17834.0312\n",
      "Epoch 394/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 251567984.0000 - rmse: 15860.8945 - val_loss: 325300864.0000 - val_rmse: 18036.0996\n",
      "Epoch 395/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 210239520.0000 - rmse: 14499.6387 - val_loss: 332609632.0000 - val_rmse: 18237.5879\n",
      "Epoch 396/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 229168080.0000 - rmse: 15138.2988 - val_loss: 312536064.0000 - val_rmse: 17678.6895\n",
      "Epoch 397/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 248215680.0000 - rmse: 15754.8623 - val_loss: 319301120.0000 - val_rmse: 17868.9980\n",
      "Epoch 398/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 264352096.0000 - rmse: 16258.9082 - val_loss: 342615296.0000 - val_rmse: 18509.8711\n",
      "Epoch 399/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 241591136.0000 - rmse: 15543.2021 - val_loss: 322686624.0000 - val_rmse: 17963.4805\n",
      "Epoch 400/400\n",
      "166/166 [==============================] - 0s 2ms/step - loss: 237155792.0000 - rmse: 15399.8633 - val_loss: 333460672.0000 - val_rmse: 18260.9062\n",
      "104/104 [==============================] - 0s 1ms/step - loss: 481569792.0000 - rmse: 21944.6992\n",
      "[481569792.0, 21944.69921875]\n"
     ]
    },
    {
     "data": {
      "application/javascript": [
       "/* Put everything inside the global mpl namespace */\n",
       "/* global mpl */\n",
       "window.mpl = {};\n",
       "\n",
       "mpl.get_websocket_type = function () {\n",
       "    if (typeof WebSocket !== 'undefined') {\n",
       "        return WebSocket;\n",
       "    } else if (typeof MozWebSocket !== 'undefined') {\n",
       "        return MozWebSocket;\n",
       "    } else {\n",
       "        alert(\n",
       "            'Your browser does not have WebSocket support. ' +\n",
       "                'Please try Chrome, Safari or Firefox ≥ 6. ' +\n",
       "                'Firefox 4 and 5 are also supported but you ' +\n",
       "                'have to enable WebSockets in about:config.'\n",
       "        );\n",
       "    }\n",
       "};\n",
       "\n",
       "mpl.figure = function (figure_id, websocket, ondownload, parent_element) {\n",
       "    this.id = figure_id;\n",
       "\n",
       "    this.ws = websocket;\n",
       "\n",
       "    this.supports_binary = this.ws.binaryType !== undefined;\n",
       "\n",
       "    if (!this.supports_binary) {\n",
       "        var warnings = document.getElementById('mpl-warnings');\n",
       "        if (warnings) {\n",
       "            warnings.style.display = 'block';\n",
       "            warnings.textContent =\n",
       "                'This browser does not support binary websocket messages. ' +\n",
       "                'Performance may be slow.';\n",
       "        }\n",
       "    }\n",
       "\n",
       "    this.imageObj = new Image();\n",
       "\n",
       "    this.context = undefined;\n",
       "    this.message = undefined;\n",
       "    this.canvas = undefined;\n",
       "    this.rubberband_canvas = undefined;\n",
       "    this.rubberband_context = undefined;\n",
       "    this.format_dropdown = undefined;\n",
       "\n",
       "    this.image_mode = 'full';\n",
       "\n",
       "    this.root = document.createElement('div');\n",
       "    this.root.setAttribute('style', 'display: inline-block');\n",
       "    this._root_extra_style(this.root);\n",
       "\n",
       "    parent_element.appendChild(this.root);\n",
       "\n",
       "    this._init_header(this);\n",
       "    this._init_canvas(this);\n",
       "    this._init_toolbar(this);\n",
       "\n",
       "    var fig = this;\n",
       "\n",
       "    this.waiting = false;\n",
       "\n",
       "    this.ws.onopen = function () {\n",
       "        fig.send_message('supports_binary', { value: fig.supports_binary });\n",
       "        fig.send_message('send_image_mode', {});\n",
       "        if (fig.ratio !== 1) {\n",
       "            fig.send_message('set_dpi_ratio', { dpi_ratio: fig.ratio });\n",
       "        }\n",
       "        fig.send_message('refresh', {});\n",
       "    };\n",
       "\n",
       "    this.imageObj.onload = function () {\n",
       "        if (fig.image_mode === 'full') {\n",
       "            // Full images could contain transparency (where diff images\n",
       "            // almost always do), so we need to clear the canvas so that\n",
       "            // there is no ghosting.\n",
       "            fig.context.clearRect(0, 0, fig.canvas.width, fig.canvas.height);\n",
       "        }\n",
       "        fig.context.drawImage(fig.imageObj, 0, 0);\n",
       "    };\n",
       "\n",
       "    this.imageObj.onunload = function () {\n",
       "        fig.ws.close();\n",
       "    };\n",
       "\n",
       "    this.ws.onmessage = this._make_on_message_function(this);\n",
       "\n",
       "    this.ondownload = ondownload;\n",
       "};\n",
       "\n",
       "mpl.figure.prototype._init_header = function () {\n",
       "    var titlebar = document.createElement('div');\n",
       "    titlebar.classList =\n",
       "        'ui-dialog-titlebar ui-widget-header ui-corner-all ui-helper-clearfix';\n",
       "    var titletext = document.createElement('div');\n",
       "    titletext.classList = 'ui-dialog-title';\n",
       "    titletext.setAttribute(\n",
       "        'style',\n",
       "        'width: 100%; text-align: center; padding: 3px;'\n",
       "    );\n",
       "    titlebar.appendChild(titletext);\n",
       "    this.root.appendChild(titlebar);\n",
       "    this.header = titletext;\n",
       "};\n",
       "\n",
       "mpl.figure.prototype._canvas_extra_style = function (_canvas_div) {};\n",
       "\n",
       "mpl.figure.prototype._root_extra_style = function (_canvas_div) {};\n",
       "\n",
       "mpl.figure.prototype._init_canvas = function () {\n",
       "    var fig = this;\n",
       "\n",
       "    var canvas_div = (this.canvas_div = document.createElement('div'));\n",
       "    canvas_div.setAttribute(\n",
       "        'style',\n",
       "        'border: 1px solid #ddd;' +\n",
       "            'box-sizing: content-box;' +\n",
       "            'clear: both;' +\n",
       "            'min-height: 1px;' +\n",
       "            'min-width: 1px;' +\n",
       "            'outline: 0;' +\n",
       "            'overflow: hidden;' +\n",
       "            'position: relative;' +\n",
       "            'resize: both;'\n",
       "    );\n",
       "\n",
       "    function on_keyboard_event_closure(name) {\n",
       "        return function (event) {\n",
       "            return fig.key_event(event, name);\n",
       "        };\n",
       "    }\n",
       "\n",
       "    canvas_div.addEventListener(\n",
       "        'keydown',\n",
       "        on_keyboard_event_closure('key_press')\n",
       "    );\n",
       "    canvas_div.addEventListener(\n",
       "        'keyup',\n",
       "        on_keyboard_event_closure('key_release')\n",
       "    );\n",
       "\n",
       "    this._canvas_extra_style(canvas_div);\n",
       "    this.root.appendChild(canvas_div);\n",
       "\n",
       "    var canvas = (this.canvas = document.createElement('canvas'));\n",
       "    canvas.classList.add('mpl-canvas');\n",
       "    canvas.setAttribute('style', 'box-sizing: content-box;');\n",
       "\n",
       "    this.context = canvas.getContext('2d');\n",
       "\n",
       "    var backingStore =\n",
       "        this.context.backingStorePixelRatio ||\n",
       "        this.context.webkitBackingStorePixelRatio ||\n",
       "        this.context.mozBackingStorePixelRatio ||\n",
       "        this.context.msBackingStorePixelRatio ||\n",
       "        this.context.oBackingStorePixelRatio ||\n",
       "        this.context.backingStorePixelRatio ||\n",
       "        1;\n",
       "\n",
       "    this.ratio = (window.devicePixelRatio || 1) / backingStore;\n",
       "    if (this.ratio !== 1) {\n",
       "        fig.send_message('set_dpi_ratio', { dpi_ratio: this.ratio });\n",
       "    }\n",
       "\n",
       "    var rubberband_canvas = (this.rubberband_canvas = document.createElement(\n",
       "        'canvas'\n",
       "    ));\n",
       "    rubberband_canvas.setAttribute(\n",
       "        'style',\n",
       "        'box-sizing: content-box; position: absolute; left: 0; top: 0; z-index: 1;'\n",
       "    );\n",
       "\n",
       "    var resizeObserver = new ResizeObserver(function (entries) {\n",
       "        var nentries = entries.length;\n",
       "        for (var i = 0; i < nentries; i++) {\n",
       "            var entry = entries[i];\n",
       "            var width, height;\n",
       "            if (entry.contentBoxSize) {\n",
       "                if (entry.contentBoxSize instanceof Array) {\n",
       "                    // Chrome 84 implements new version of spec.\n",
       "                    width = entry.contentBoxSize[0].inlineSize;\n",
       "                    height = entry.contentBoxSize[0].blockSize;\n",
       "                } else {\n",
       "                    // Firefox implements old version of spec.\n",
       "                    width = entry.contentBoxSize.inlineSize;\n",
       "                    height = entry.contentBoxSize.blockSize;\n",
       "                }\n",
       "            } else {\n",
       "                // Chrome <84 implements even older version of spec.\n",
       "                width = entry.contentRect.width;\n",
       "                height = entry.contentRect.height;\n",
       "            }\n",
       "\n",
       "            // Keep the size of the canvas and rubber band canvas in sync with\n",
       "            // the canvas container.\n",
       "            if (entry.devicePixelContentBoxSize) {\n",
       "                // Chrome 84 implements new version of spec.\n",
       "                canvas.setAttribute(\n",
       "                    'width',\n",
       "                    entry.devicePixelContentBoxSize[0].inlineSize\n",
       "                );\n",
       "                canvas.setAttribute(\n",
       "                    'height',\n",
       "                    entry.devicePixelContentBoxSize[0].blockSize\n",
       "                );\n",
       "            } else {\n",
       "                canvas.setAttribute('width', width * fig.ratio);\n",
       "                canvas.setAttribute('height', height * fig.ratio);\n",
       "            }\n",
       "            canvas.setAttribute(\n",
       "                'style',\n",
       "                'width: ' + width + 'px; height: ' + height + 'px;'\n",
       "            );\n",
       "\n",
       "            rubberband_canvas.setAttribute('width', width);\n",
       "            rubberband_canvas.setAttribute('height', height);\n",
       "\n",
       "            // And update the size in Python. We ignore the initial 0/0 size\n",
       "            // that occurs as the element is placed into the DOM, which should\n",
       "            // otherwise not happen due to the minimum size styling.\n",
       "            if (width != 0 && height != 0) {\n",
       "                fig.request_resize(width, height);\n",
       "            }\n",
       "        }\n",
       "    });\n",
       "    resizeObserver.observe(canvas_div);\n",
       "\n",
       "    function on_mouse_event_closure(name) {\n",
       "        return function (event) {\n",
       "            return fig.mouse_event(event, name);\n",
       "        };\n",
       "    }\n",
       "\n",
       "    rubberband_canvas.addEventListener(\n",
       "        'mousedown',\n",
       "        on_mouse_event_closure('button_press')\n",
       "    );\n",
       "    rubberband_canvas.addEventListener(\n",
       "        'mouseup',\n",
       "        on_mouse_event_closure('button_release')\n",
       "    );\n",
       "    // Throttle sequential mouse events to 1 every 20ms.\n",
       "    rubberband_canvas.addEventListener(\n",
       "        'mousemove',\n",
       "        on_mouse_event_closure('motion_notify')\n",
       "    );\n",
       "\n",
       "    rubberband_canvas.addEventListener(\n",
       "        'mouseenter',\n",
       "        on_mouse_event_closure('figure_enter')\n",
       "    );\n",
       "    rubberband_canvas.addEventListener(\n",
       "        'mouseleave',\n",
       "        on_mouse_event_closure('figure_leave')\n",
       "    );\n",
       "\n",
       "    canvas_div.addEventListener('wheel', function (event) {\n",
       "        if (event.deltaY < 0) {\n",
       "            event.step = 1;\n",
       "        } else {\n",
       "            event.step = -1;\n",
       "        }\n",
       "        on_mouse_event_closure('scroll')(event);\n",
       "    });\n",
       "\n",
       "    canvas_div.appendChild(canvas);\n",
       "    canvas_div.appendChild(rubberband_canvas);\n",
       "\n",
       "    this.rubberband_context = rubberband_canvas.getContext('2d');\n",
       "    this.rubberband_context.strokeStyle = '#000000';\n",
       "\n",
       "    this._resize_canvas = function (width, height, forward) {\n",
       "        if (forward) {\n",
       "            canvas_div.style.width = width + 'px';\n",
       "            canvas_div.style.height = height + 'px';\n",
       "        }\n",
       "    };\n",
       "\n",
       "    // Disable right mouse context menu.\n",
       "    this.rubberband_canvas.addEventListener('contextmenu', function (_e) {\n",
       "        event.preventDefault();\n",
       "        return false;\n",
       "    });\n",
       "\n",
       "    function set_focus() {\n",
       "        canvas.focus();\n",
       "        canvas_div.focus();\n",
       "    }\n",
       "\n",
       "    window.setTimeout(set_focus, 100);\n",
       "};\n",
       "\n",
       "mpl.figure.prototype._init_toolbar = function () {\n",
       "    var fig = this;\n",
       "\n",
       "    var toolbar = document.createElement('div');\n",
       "    toolbar.classList = 'mpl-toolbar';\n",
       "    this.root.appendChild(toolbar);\n",
       "\n",
       "    function on_click_closure(name) {\n",
       "        return function (_event) {\n",
       "            return fig.toolbar_button_onclick(name);\n",
       "        };\n",
       "    }\n",
       "\n",
       "    function on_mouseover_closure(tooltip) {\n",
       "        return function (event) {\n",
       "            if (!event.currentTarget.disabled) {\n",
       "                return fig.toolbar_button_onmouseover(tooltip);\n",
       "            }\n",
       "        };\n",
       "    }\n",
       "\n",
       "    fig.buttons = {};\n",
       "    var buttonGroup = document.createElement('div');\n",
       "    buttonGroup.classList = 'mpl-button-group';\n",
       "    for (var toolbar_ind in mpl.toolbar_items) {\n",
       "        var name = mpl.toolbar_items[toolbar_ind][0];\n",
       "        var tooltip = mpl.toolbar_items[toolbar_ind][1];\n",
       "        var image = mpl.toolbar_items[toolbar_ind][2];\n",
       "        var method_name = mpl.toolbar_items[toolbar_ind][3];\n",
       "\n",
       "        if (!name) {\n",
       "            /* Instead of a spacer, we start a new button group. */\n",
       "            if (buttonGroup.hasChildNodes()) {\n",
       "                toolbar.appendChild(buttonGroup);\n",
       "            }\n",
       "            buttonGroup = document.createElement('div');\n",
       "            buttonGroup.classList = 'mpl-button-group';\n",
       "            continue;\n",
       "        }\n",
       "\n",
       "        var button = (fig.buttons[name] = document.createElement('button'));\n",
       "        button.classList = 'mpl-widget';\n",
       "        button.setAttribute('role', 'button');\n",
       "        button.setAttribute('aria-disabled', 'false');\n",
       "        button.addEventListener('click', on_click_closure(method_name));\n",
       "        button.addEventListener('mouseover', on_mouseover_closure(tooltip));\n",
       "\n",
       "        var icon_img = document.createElement('img');\n",
       "        icon_img.src = '_images/' + image + '.png';\n",
       "        icon_img.srcset = '_images/' + image + '_large.png 2x';\n",
       "        icon_img.alt = tooltip;\n",
       "        button.appendChild(icon_img);\n",
       "\n",
       "        buttonGroup.appendChild(button);\n",
       "    }\n",
       "\n",
       "    if (buttonGroup.hasChildNodes()) {\n",
       "        toolbar.appendChild(buttonGroup);\n",
       "    }\n",
       "\n",
       "    var fmt_picker = document.createElement('select');\n",
       "    fmt_picker.classList = 'mpl-widget';\n",
       "    toolbar.appendChild(fmt_picker);\n",
       "    this.format_dropdown = fmt_picker;\n",
       "\n",
       "    for (var ind in mpl.extensions) {\n",
       "        var fmt = mpl.extensions[ind];\n",
       "        var option = document.createElement('option');\n",
       "        option.selected = fmt === mpl.default_extension;\n",
       "        option.innerHTML = fmt;\n",
       "        fmt_picker.appendChild(option);\n",
       "    }\n",
       "\n",
       "    var status_bar = document.createElement('span');\n",
       "    status_bar.classList = 'mpl-message';\n",
       "    toolbar.appendChild(status_bar);\n",
       "    this.message = status_bar;\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.request_resize = function (x_pixels, y_pixels) {\n",
       "    // Request matplotlib to resize the figure. Matplotlib will then trigger a resize in the client,\n",
       "    // which will in turn request a refresh of the image.\n",
       "    this.send_message('resize', { width: x_pixels, height: y_pixels });\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.send_message = function (type, properties) {\n",
       "    properties['type'] = type;\n",
       "    properties['figure_id'] = this.id;\n",
       "    this.ws.send(JSON.stringify(properties));\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.send_draw_message = function () {\n",
       "    if (!this.waiting) {\n",
       "        this.waiting = true;\n",
       "        this.ws.send(JSON.stringify({ type: 'draw', figure_id: this.id }));\n",
       "    }\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.handle_save = function (fig, _msg) {\n",
       "    var format_dropdown = fig.format_dropdown;\n",
       "    var format = format_dropdown.options[format_dropdown.selectedIndex].value;\n",
       "    fig.ondownload(fig, format);\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.handle_resize = function (fig, msg) {\n",
       "    var size = msg['size'];\n",
       "    if (size[0] !== fig.canvas.width || size[1] !== fig.canvas.height) {\n",
       "        fig._resize_canvas(size[0], size[1], msg['forward']);\n",
       "        fig.send_message('refresh', {});\n",
       "    }\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.handle_rubberband = function (fig, msg) {\n",
       "    var x0 = msg['x0'] / fig.ratio;\n",
       "    var y0 = (fig.canvas.height - msg['y0']) / fig.ratio;\n",
       "    var x1 = msg['x1'] / fig.ratio;\n",
       "    var y1 = (fig.canvas.height - msg['y1']) / fig.ratio;\n",
       "    x0 = Math.floor(x0) + 0.5;\n",
       "    y0 = Math.floor(y0) + 0.5;\n",
       "    x1 = Math.floor(x1) + 0.5;\n",
       "    y1 = Math.floor(y1) + 0.5;\n",
       "    var min_x = Math.min(x0, x1);\n",
       "    var min_y = Math.min(y0, y1);\n",
       "    var width = Math.abs(x1 - x0);\n",
       "    var height = Math.abs(y1 - y0);\n",
       "\n",
       "    fig.rubberband_context.clearRect(\n",
       "        0,\n",
       "        0,\n",
       "        fig.canvas.width / fig.ratio,\n",
       "        fig.canvas.height / fig.ratio\n",
       "    );\n",
       "\n",
       "    fig.rubberband_context.strokeRect(min_x, min_y, width, height);\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.handle_figure_label = function (fig, msg) {\n",
       "    // Updates the figure title.\n",
       "    fig.header.textContent = msg['label'];\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.handle_cursor = function (fig, msg) {\n",
       "    var cursor = msg['cursor'];\n",
       "    switch (cursor) {\n",
       "        case 0:\n",
       "            cursor = 'pointer';\n",
       "            break;\n",
       "        case 1:\n",
       "            cursor = 'default';\n",
       "            break;\n",
       "        case 2:\n",
       "            cursor = 'crosshair';\n",
       "            break;\n",
       "        case 3:\n",
       "            cursor = 'move';\n",
       "            break;\n",
       "    }\n",
       "    fig.rubberband_canvas.style.cursor = cursor;\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.handle_message = function (fig, msg) {\n",
       "    fig.message.textContent = msg['message'];\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.handle_draw = function (fig, _msg) {\n",
       "    // Request the server to send over a new figure.\n",
       "    fig.send_draw_message();\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.handle_image_mode = function (fig, msg) {\n",
       "    fig.image_mode = msg['mode'];\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.handle_history_buttons = function (fig, msg) {\n",
       "    for (var key in msg) {\n",
       "        if (!(key in fig.buttons)) {\n",
       "            continue;\n",
       "        }\n",
       "        fig.buttons[key].disabled = !msg[key];\n",
       "        fig.buttons[key].setAttribute('aria-disabled', !msg[key]);\n",
       "    }\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.handle_navigate_mode = function (fig, msg) {\n",
       "    if (msg['mode'] === 'PAN') {\n",
       "        fig.buttons['Pan'].classList.add('active');\n",
       "        fig.buttons['Zoom'].classList.remove('active');\n",
       "    } else if (msg['mode'] === 'ZOOM') {\n",
       "        fig.buttons['Pan'].classList.remove('active');\n",
       "        fig.buttons['Zoom'].classList.add('active');\n",
       "    } else {\n",
       "        fig.buttons['Pan'].classList.remove('active');\n",
       "        fig.buttons['Zoom'].classList.remove('active');\n",
       "    }\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.updated_canvas_event = function () {\n",
       "    // Called whenever the canvas gets updated.\n",
       "    this.send_message('ack', {});\n",
       "};\n",
       "\n",
       "// A function to construct a web socket function for onmessage handling.\n",
       "// Called in the figure constructor.\n",
       "mpl.figure.prototype._make_on_message_function = function (fig) {\n",
       "    return function socket_on_message(evt) {\n",
       "        if (evt.data instanceof Blob) {\n",
       "            /* FIXME: We get \"Resource interpreted as Image but\n",
       "             * transferred with MIME type text/plain:\" errors on\n",
       "             * Chrome.  But how to set the MIME type?  It doesn't seem\n",
       "             * to be part of the websocket stream */\n",
       "            evt.data.type = 'image/png';\n",
       "\n",
       "            /* Free the memory for the previous frames */\n",
       "            if (fig.imageObj.src) {\n",
       "                (window.URL || window.webkitURL).revokeObjectURL(\n",
       "                    fig.imageObj.src\n",
       "                );\n",
       "            }\n",
       "\n",
       "            fig.imageObj.src = (window.URL || window.webkitURL).createObjectURL(\n",
       "                evt.data\n",
       "            );\n",
       "            fig.updated_canvas_event();\n",
       "            fig.waiting = false;\n",
       "            return;\n",
       "        } else if (\n",
       "            typeof evt.data === 'string' &&\n",
       "            evt.data.slice(0, 21) === 'data:image/png;base64'\n",
       "        ) {\n",
       "            fig.imageObj.src = evt.data;\n",
       "            fig.updated_canvas_event();\n",
       "            fig.waiting = false;\n",
       "            return;\n",
       "        }\n",
       "\n",
       "        var msg = JSON.parse(evt.data);\n",
       "        var msg_type = msg['type'];\n",
       "\n",
       "        // Call the  \"handle_{type}\" callback, which takes\n",
       "        // the figure and JSON message as its only arguments.\n",
       "        try {\n",
       "            var callback = fig['handle_' + msg_type];\n",
       "        } catch (e) {\n",
       "            console.log(\n",
       "                \"No handler for the '\" + msg_type + \"' message type: \",\n",
       "                msg\n",
       "            );\n",
       "            return;\n",
       "        }\n",
       "\n",
       "        if (callback) {\n",
       "            try {\n",
       "                // console.log(\"Handling '\" + msg_type + \"' message: \", msg);\n",
       "                callback(fig, msg);\n",
       "            } catch (e) {\n",
       "                console.log(\n",
       "                    \"Exception inside the 'handler_\" + msg_type + \"' callback:\",\n",
       "                    e,\n",
       "                    e.stack,\n",
       "                    msg\n",
       "                );\n",
       "            }\n",
       "        }\n",
       "    };\n",
       "};\n",
       "\n",
       "// from http://stackoverflow.com/questions/1114465/getting-mouse-location-in-canvas\n",
       "mpl.findpos = function (e) {\n",
       "    //this section is from http://www.quirksmode.org/js/events_properties.html\n",
       "    var targ;\n",
       "    if (!e) {\n",
       "        e = window.event;\n",
       "    }\n",
       "    if (e.target) {\n",
       "        targ = e.target;\n",
       "    } else if (e.srcElement) {\n",
       "        targ = e.srcElement;\n",
       "    }\n",
       "    if (targ.nodeType === 3) {\n",
       "        // defeat Safari bug\n",
       "        targ = targ.parentNode;\n",
       "    }\n",
       "\n",
       "    // pageX,Y are the mouse positions relative to the document\n",
       "    var boundingRect = targ.getBoundingClientRect();\n",
       "    var x = e.pageX - (boundingRect.left + document.body.scrollLeft);\n",
       "    var y = e.pageY - (boundingRect.top + document.body.scrollTop);\n",
       "\n",
       "    return { x: x, y: y };\n",
       "};\n",
       "\n",
       "/*\n",
       " * return a copy of an object with only non-object keys\n",
       " * we need this to avoid circular references\n",
       " * http://stackoverflow.com/a/24161582/3208463\n",
       " */\n",
       "function simpleKeys(original) {\n",
       "    return Object.keys(original).reduce(function (obj, key) {\n",
       "        if (typeof original[key] !== 'object') {\n",
       "            obj[key] = original[key];\n",
       "        }\n",
       "        return obj;\n",
       "    }, {});\n",
       "}\n",
       "\n",
       "mpl.figure.prototype.mouse_event = function (event, name) {\n",
       "    var canvas_pos = mpl.findpos(event);\n",
       "\n",
       "    if (name === 'button_press') {\n",
       "        this.canvas.focus();\n",
       "        this.canvas_div.focus();\n",
       "    }\n",
       "\n",
       "    var x = canvas_pos.x * this.ratio;\n",
       "    var y = canvas_pos.y * this.ratio;\n",
       "\n",
       "    this.send_message(name, {\n",
       "        x: x,\n",
       "        y: y,\n",
       "        button: event.button,\n",
       "        step: event.step,\n",
       "        guiEvent: simpleKeys(event),\n",
       "    });\n",
       "\n",
       "    /* This prevents the web browser from automatically changing to\n",
       "     * the text insertion cursor when the button is pressed.  We want\n",
       "     * to control all of the cursor setting manually through the\n",
       "     * 'cursor' event from matplotlib */\n",
       "    event.preventDefault();\n",
       "    return false;\n",
       "};\n",
       "\n",
       "mpl.figure.prototype._key_event_extra = function (_event, _name) {\n",
       "    // Handle any extra behaviour associated with a key event\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.key_event = function (event, name) {\n",
       "    // Prevent repeat events\n",
       "    if (name === 'key_press') {\n",
       "        if (event.which === this._key) {\n",
       "            return;\n",
       "        } else {\n",
       "            this._key = event.which;\n",
       "        }\n",
       "    }\n",
       "    if (name === 'key_release') {\n",
       "        this._key = null;\n",
       "    }\n",
       "\n",
       "    var value = '';\n",
       "    if (event.ctrlKey && event.which !== 17) {\n",
       "        value += 'ctrl+';\n",
       "    }\n",
       "    if (event.altKey && event.which !== 18) {\n",
       "        value += 'alt+';\n",
       "    }\n",
       "    if (event.shiftKey && event.which !== 16) {\n",
       "        value += 'shift+';\n",
       "    }\n",
       "\n",
       "    value += 'k';\n",
       "    value += event.which.toString();\n",
       "\n",
       "    this._key_event_extra(event, name);\n",
       "\n",
       "    this.send_message(name, { key: value, guiEvent: simpleKeys(event) });\n",
       "    return false;\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.toolbar_button_onclick = function (name) {\n",
       "    if (name === 'download') {\n",
       "        this.handle_save(this, null);\n",
       "    } else {\n",
       "        this.send_message('toolbar_button', { name: name });\n",
       "    }\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.toolbar_button_onmouseover = function (tooltip) {\n",
       "    this.message.textContent = tooltip;\n",
       "};\n",
       "mpl.toolbar_items = [[\"Home\", \"Reset original view\", \"fa fa-home icon-home\", \"home\"], [\"Back\", \"Back to previous view\", \"fa fa-arrow-left icon-arrow-left\", \"back\"], [\"Forward\", \"Forward to next view\", \"fa fa-arrow-right icon-arrow-right\", \"forward\"], [\"\", \"\", \"\", \"\"], [\"Pan\", \"Left button pans, Right button zooms\\nx/y fixes axis, CTRL fixes aspect\", \"fa fa-arrows icon-move\", \"pan\"], [\"Zoom\", \"Zoom to rectangle\\nx/y fixes axis, CTRL fixes aspect\", \"fa fa-square-o icon-check-empty\", \"zoom\"], [\"\", \"\", \"\", \"\"], [\"Download\", \"Download plot\", \"fa fa-floppy-o icon-save\", \"download\"]];\n",
       "\n",
       "mpl.extensions = [\"eps\", \"jpeg\", \"pdf\", \"png\", \"ps\", \"raw\", \"svg\", \"tif\"];\n",
       "\n",
       "mpl.default_extension = \"png\";/* global mpl */\n",
       "\n",
       "var comm_websocket_adapter = function (comm) {\n",
       "    // Create a \"websocket\"-like object which calls the given IPython comm\n",
       "    // object with the appropriate methods. Currently this is a non binary\n",
       "    // socket, so there is still some room for performance tuning.\n",
       "    var ws = {};\n",
       "\n",
       "    ws.close = function () {\n",
       "        comm.close();\n",
       "    };\n",
       "    ws.send = function (m) {\n",
       "        //console.log('sending', m);\n",
       "        comm.send(m);\n",
       "    };\n",
       "    // Register the callback with on_msg.\n",
       "    comm.on_msg(function (msg) {\n",
       "        //console.log('receiving', msg['content']['data'], msg);\n",
       "        // Pass the mpl event to the overridden (by mpl) onmessage function.\n",
       "        ws.onmessage(msg['content']['data']);\n",
       "    });\n",
       "    return ws;\n",
       "};\n",
       "\n",
       "mpl.mpl_figure_comm = function (comm, msg) {\n",
       "    // This is the function which gets called when the mpl process\n",
       "    // starts-up an IPython Comm through the \"matplotlib\" channel.\n",
       "\n",
       "    var id = msg.content.data.id;\n",
       "    // Get hold of the div created by the display call when the Comm\n",
       "    // socket was opened in Python.\n",
       "    var element = document.getElementById(id);\n",
       "    var ws_proxy = comm_websocket_adapter(comm);\n",
       "\n",
       "    function ondownload(figure, _format) {\n",
       "        window.open(figure.canvas.toDataURL());\n",
       "    }\n",
       "\n",
       "    var fig = new mpl.figure(id, ws_proxy, ondownload, element);\n",
       "\n",
       "    // Call onopen now - mpl needs it, as it is assuming we've passed it a real\n",
       "    // web socket which is closed, not our websocket->open comm proxy.\n",
       "    ws_proxy.onopen();\n",
       "\n",
       "    fig.parent_element = element;\n",
       "    fig.cell_info = mpl.find_output_cell(\"<div id='\" + id + \"'></div>\");\n",
       "    if (!fig.cell_info) {\n",
       "        console.error('Failed to find cell for figure', id, fig);\n",
       "        return;\n",
       "    }\n",
       "    fig.cell_info[0].output_area.element.one(\n",
       "        'cleared',\n",
       "        { fig: fig },\n",
       "        fig._remove_fig_handler\n",
       "    );\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.handle_close = function (fig, msg) {\n",
       "    var width = fig.canvas.width / fig.ratio;\n",
       "    fig.cell_info[0].output_area.element.off(\n",
       "        'cleared',\n",
       "        fig._remove_fig_handler\n",
       "    );\n",
       "\n",
       "    // Update the output cell to use the data from the current canvas.\n",
       "    fig.push_to_output();\n",
       "    var dataURL = fig.canvas.toDataURL();\n",
       "    // Re-enable the keyboard manager in IPython - without this line, in FF,\n",
       "    // the notebook keyboard shortcuts fail.\n",
       "    IPython.keyboard_manager.enable();\n",
       "    fig.parent_element.innerHTML =\n",
       "        '<img src=\"' + dataURL + '\" width=\"' + width + '\">';\n",
       "    fig.close_ws(fig, msg);\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.close_ws = function (fig, msg) {\n",
       "    fig.send_message('closing', msg);\n",
       "    // fig.ws.close()\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.push_to_output = function (_remove_interactive) {\n",
       "    // Turn the data on the canvas into data in the output cell.\n",
       "    var width = this.canvas.width / this.ratio;\n",
       "    var dataURL = this.canvas.toDataURL();\n",
       "    this.cell_info[1]['text/html'] =\n",
       "        '<img src=\"' + dataURL + '\" width=\"' + width + '\">';\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.updated_canvas_event = function () {\n",
       "    // Tell IPython that the notebook contents must change.\n",
       "    IPython.notebook.set_dirty(true);\n",
       "    this.send_message('ack', {});\n",
       "    var fig = this;\n",
       "    // Wait a second, then push the new image to the DOM so\n",
       "    // that it is saved nicely (might be nice to debounce this).\n",
       "    setTimeout(function () {\n",
       "        fig.push_to_output();\n",
       "    }, 1000);\n",
       "};\n",
       "\n",
       "mpl.figure.prototype._init_toolbar = function () {\n",
       "    var fig = this;\n",
       "\n",
       "    var toolbar = document.createElement('div');\n",
       "    toolbar.classList = 'btn-toolbar';\n",
       "    this.root.appendChild(toolbar);\n",
       "\n",
       "    function on_click_closure(name) {\n",
       "        return function (_event) {\n",
       "            return fig.toolbar_button_onclick(name);\n",
       "        };\n",
       "    }\n",
       "\n",
       "    function on_mouseover_closure(tooltip) {\n",
       "        return function (event) {\n",
       "            if (!event.currentTarget.disabled) {\n",
       "                return fig.toolbar_button_onmouseover(tooltip);\n",
       "            }\n",
       "        };\n",
       "    }\n",
       "\n",
       "    fig.buttons = {};\n",
       "    var buttonGroup = document.createElement('div');\n",
       "    buttonGroup.classList = 'btn-group';\n",
       "    var button;\n",
       "    for (var toolbar_ind in mpl.toolbar_items) {\n",
       "        var name = mpl.toolbar_items[toolbar_ind][0];\n",
       "        var tooltip = mpl.toolbar_items[toolbar_ind][1];\n",
       "        var image = mpl.toolbar_items[toolbar_ind][2];\n",
       "        var method_name = mpl.toolbar_items[toolbar_ind][3];\n",
       "\n",
       "        if (!name) {\n",
       "            /* Instead of a spacer, we start a new button group. */\n",
       "            if (buttonGroup.hasChildNodes()) {\n",
       "                toolbar.appendChild(buttonGroup);\n",
       "            }\n",
       "            buttonGroup = document.createElement('div');\n",
       "            buttonGroup.classList = 'btn-group';\n",
       "            continue;\n",
       "        }\n",
       "\n",
       "        button = fig.buttons[name] = document.createElement('button');\n",
       "        button.classList = 'btn btn-default';\n",
       "        button.href = '#';\n",
       "        button.title = name;\n",
       "        button.innerHTML = '<i class=\"fa ' + image + ' fa-lg\"></i>';\n",
       "        button.addEventListener('click', on_click_closure(method_name));\n",
       "        button.addEventListener('mouseover', on_mouseover_closure(tooltip));\n",
       "        buttonGroup.appendChild(button);\n",
       "    }\n",
       "\n",
       "    if (buttonGroup.hasChildNodes()) {\n",
       "        toolbar.appendChild(buttonGroup);\n",
       "    }\n",
       "\n",
       "    // Add the status bar.\n",
       "    var status_bar = document.createElement('span');\n",
       "    status_bar.classList = 'mpl-message pull-right';\n",
       "    toolbar.appendChild(status_bar);\n",
       "    this.message = status_bar;\n",
       "\n",
       "    // Add the close button to the window.\n",
       "    var buttongrp = document.createElement('div');\n",
       "    buttongrp.classList = 'btn-group inline pull-right';\n",
       "    button = document.createElement('button');\n",
       "    button.classList = 'btn btn-mini btn-primary';\n",
       "    button.href = '#';\n",
       "    button.title = 'Stop Interaction';\n",
       "    button.innerHTML = '<i class=\"fa fa-power-off icon-remove icon-large\"></i>';\n",
       "    button.addEventListener('click', function (_evt) {\n",
       "        fig.handle_close(fig, {});\n",
       "    });\n",
       "    button.addEventListener(\n",
       "        'mouseover',\n",
       "        on_mouseover_closure('Stop Interaction')\n",
       "    );\n",
       "    buttongrp.appendChild(button);\n",
       "    var titlebar = this.root.querySelector('.ui-dialog-titlebar');\n",
       "    titlebar.insertBefore(buttongrp, titlebar.firstChild);\n",
       "};\n",
       "\n",
       "mpl.figure.prototype._remove_fig_handler = function (event) {\n",
       "    var fig = event.data.fig;\n",
       "    fig.close_ws(fig, {});\n",
       "};\n",
       "\n",
       "mpl.figure.prototype._root_extra_style = function (el) {\n",
       "    el.style.boxSizing = 'content-box'; // override notebook setting of border-box.\n",
       "};\n",
       "\n",
       "mpl.figure.prototype._canvas_extra_style = function (el) {\n",
       "    // this is important to make the div 'focusable\n",
       "    el.setAttribute('tabindex', 0);\n",
       "    // reach out to IPython and tell the keyboard manager to turn it's self\n",
       "    // off when our div gets focus\n",
       "\n",
       "    // location in version 3\n",
       "    if (IPython.notebook.keyboard_manager) {\n",
       "        IPython.notebook.keyboard_manager.register_events(el);\n",
       "    } else {\n",
       "        // location in version 2\n",
       "        IPython.keyboard_manager.register_events(el);\n",
       "    }\n",
       "};\n",
       "\n",
       "mpl.figure.prototype._key_event_extra = function (event, _name) {\n",
       "    var manager = IPython.notebook.keyboard_manager;\n",
       "    if (!manager) {\n",
       "        manager = IPython.keyboard_manager;\n",
       "    }\n",
       "\n",
       "    // Check for shift+enter\n",
       "    if (event.shiftKey && event.which === 13) {\n",
       "        this.canvas_div.blur();\n",
       "        // select the cell after this one\n",
       "        var index = IPython.notebook.find_cell_index(this.cell_info[0]);\n",
       "        IPython.notebook.select(index + 1);\n",
       "    }\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.handle_save = function (fig, _msg) {\n",
       "    fig.ondownload(fig, null);\n",
       "};\n",
       "\n",
       "mpl.find_output_cell = function (html_output) {\n",
       "    // Return the cell and output element which can be found *uniquely* in the notebook.\n",
       "    // Note - this is a bit hacky, but it is done because the \"notebook_saving.Notebook\"\n",
       "    // IPython event is triggered only after the cells have been serialised, which for\n",
       "    // our purposes (turning an active figure into a static one), is too late.\n",
       "    var cells = IPython.notebook.get_cells();\n",
       "    var ncells = cells.length;\n",
       "    for (var i = 0; i < ncells; i++) {\n",
       "        var cell = cells[i];\n",
       "        if (cell.cell_type === 'code') {\n",
       "            for (var j = 0; j < cell.output_area.outputs.length; j++) {\n",
       "                var data = cell.output_area.outputs[j];\n",
       "                if (data.data) {\n",
       "                    // IPython >= 3 moved mimebundle to data attribute of output\n",
       "                    data = data.data;\n",
       "                }\n",
       "                if (data['text/html'] === html_output) {\n",
       "                    return [cell, data, j];\n",
       "                }\n",
       "            }\n",
       "        }\n",
       "    }\n",
       "};\n",
       "\n",
       "// Register the function which deals with the matplotlib target/channel.\n",
       "// The kernel may be null if the page has been refreshed.\n",
       "if (IPython.notebook.kernel !== null) {\n",
       "    IPython.notebook.kernel.comm_manager.register_target(\n",
       "        'matplotlib',\n",
       "        mpl.mpl_figure_comm\n",
       "    );\n",
       "}\n"
      ],
      "text/plain": [
       "<IPython.core.display.Javascript object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/html": [
       "<img src=\"data:image/png;base64,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\" width=\"640\">"
      ],
      "text/plain": [
       "<IPython.core.display.HTML object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Model: \"simple_widedeep_model_3\"\n",
      "_________________________________________________________________\n",
      "Layer (type)                 Output Shape              Param #   \n",
      "=================================================================\n",
      "dense_80 (Dense)             multiple                  860       \n",
      "_________________________________________________________________\n",
      "dense_81 (Dense)             multiple                  39        \n",
      "_________________________________________________________________\n",
      "dense_82 (Dense)             multiple                  4         \n",
      "_________________________________________________________________\n",
      "dense_83 (Dense)             multiple                  36        \n",
      "_________________________________________________________________\n",
      "dense_84 (Dense)             multiple                  156       \n",
      "_________________________________________________________________\n",
      "dense_85 (Dense)             multiple                  1240      \n",
      "_________________________________________________________________\n",
      "dense_86 (Dense)             multiple                  3840      \n",
      "_________________________________________________________________\n",
      "dropout_7 (Dropout)          multiple                  0         \n",
      "_________________________________________________________________\n",
      "dense_87 (Dense)             multiple                  4128      \n",
      "_________________________________________________________________\n",
      "dense_88 (Dense)             multiple                  528       \n",
      "_________________________________________________________________\n",
      "dense_89 (Dense)             multiple                  272       \n",
      "_________________________________________________________________\n",
      "dense_90 (Dense)             multiple                  108       \n",
      "_________________________________________________________________\n",
      "dense_91 (Dense)             multiple                  5         \n",
      "=================================================================\n",
      "Total params: 11,216\n",
      "Trainable params: 11,216\n",
      "Non-trainable params: 0\n",
      "_________________________________________________________________\n",
      "None\n"
     ]
    }
   ],
   "source": [
    "# train or load model\n",
    "\n",
    "# baseline model\n",
    "baseline_model = load_model(\"./models/baseline_model.h5\")\n",
    "\n",
    "# embedded model\n",
    "emb_model = Emb_model()\n",
    "emb_model.load_weights(\"./models/emb_model\")\n",
    "\n",
    "# random forest\n",
    "rf_regressor = RandomForestRegressor(n_estimators=200, max_depth=20, random_state=2021)\n",
    "# rf_regressor.fit(X_tree_train, y_tree_train)\n",
    "rf_regressor.fit(X_train, y_train)\n",
    "\n",
    "# gradient boost \n",
    "gb_regressor = GradientBoostingRegressor(n_estimators=400, learning_rate=0.15, random_state=2021)\n",
    "# gb_regressor.fit(X_tree_train, y_tree_train)\n",
    "gb_regressor.fit(X_train, y_train)\n",
    "\n",
    "# wide & deep model\n",
    "wnd_model = train_and_get_deep_learning_model(\"simple_widedeep\")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "016a061f",
   "metadata": {},
   "source": [
    "## RMSE in different test part"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 70,
   "id": "5fa881df",
   "metadata": {},
   "outputs": [],
   "source": [
    "def rmse_in_different_test_part():\n",
    "    # import data\n",
    "    kfold = KFold(n_splits=5)\n",
    "    bl_rmse = []\n",
    "    eb_rmse = []\n",
    "    rf_rmse = []\n",
    "    gb_rmse = []\n",
    "    wnd_rmse = []\n",
    "    ensemble_rmse = []\n",
    "\n",
    "\n",
    "    for _, test in kfold.split(X_test, y_test):\n",
    "        bl_predict = np.squeeze(baseline_model.predict(X_test[test]))\n",
    "        bl_rmse.append(mean_squared_error(y_test[test], bl_predict, squared=False))\n",
    "        \n",
    "        eb_predict = np.squeeze(emb_model.predict(X_test[test]))\n",
    "        eb_rmse.append(mean_squared_error(y_test[test], eb_predict, squared=False))\n",
    "\n",
    "        rf_predict = rf_regressor.predict(X_test[test])\n",
    "        rf_rmse.append(mean_squared_error(y_test[test], rf_predict, squared=False))\n",
    "\n",
    "        gb_predict = gb_regressor.predict(X_test[test])\n",
    "        gb_rmse.append(mean_squared_error(y_test[test], gb_predict, squared=False))\n",
    "        \n",
    "        wnd_predict = np.squeeze(wnd_model.predict(X_test[test]))\n",
    "        wnd_rmse.append(mean_squared_error(y_test[test], wnd_predict, squared=False))\n",
    "\n",
    "        ensemble_predict = 0.2 * gb_predict + 0.2 * rf_predict + 0.3 * bl_predict + 0.3 * eb_predict\n",
    "        ensemble_rmse.append(mean_squared_error(y_test[test], ensemble_predict, squared=False))\n",
    "\n",
    "    x = range(1, 6)\n",
    "    plt.figure(figsize=(9,6))\n",
    "    plt.plot(x, bl_rmse, c=\"r\", label=\"BaselineModel_RMSE\")\n",
    "    plt.plot(x, eb_rmse, c=\"pink\", label=\"EmbeddedModel_RMSE\")\n",
    "    plt.plot(x, rf_rmse, c=\"y\", label=\"RandomForest_RMSE\")\n",
    "    plt.plot(x, gb_rmse, c=\"b\", label=\"GradientBoosting_RMSE\")\n",
    "    plt.plot(x, wnd_rmse, c=\"grey\", label=\"Wide&Deep_RMSE\")\n",
    "    plt.plot(x, ensemble_rmse, c=\"black\", label=\"ensemble_RMSE\")\n",
    "    \n",
    "    plt.xlabel(\"Set No.\")\n",
    "    plt.ylabel(\"RMSE\")\n",
    "    plt.xticks(np.arange(1, 6, 1), [f\"test set {i}\" for i in range(1, 6)])\n",
    "    plt.legend()\n",
    "    plt.show()\n",
    "    plt.savefig(\"x1 RMSE in different part.png\")\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 71,
   "id": "26340ff7",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "application/javascript": [
       "/* Put everything inside the global mpl namespace */\n",
       "/* global mpl */\n",
       "window.mpl = {};\n",
       "\n",
       "mpl.get_websocket_type = function () {\n",
       "    if (typeof WebSocket !== 'undefined') {\n",
       "        return WebSocket;\n",
       "    } else if (typeof MozWebSocket !== 'undefined') {\n",
       "        return MozWebSocket;\n",
       "    } else {\n",
       "        alert(\n",
       "            'Your browser does not have WebSocket support. ' +\n",
       "                'Please try Chrome, Safari or Firefox ≥ 6. ' +\n",
       "                'Firefox 4 and 5 are also supported but you ' +\n",
       "                'have to enable WebSockets in about:config.'\n",
       "        );\n",
       "    }\n",
       "};\n",
       "\n",
       "mpl.figure = function (figure_id, websocket, ondownload, parent_element) {\n",
       "    this.id = figure_id;\n",
       "\n",
       "    this.ws = websocket;\n",
       "\n",
       "    this.supports_binary = this.ws.binaryType !== undefined;\n",
       "\n",
       "    if (!this.supports_binary) {\n",
       "        var warnings = document.getElementById('mpl-warnings');\n",
       "        if (warnings) {\n",
       "            warnings.style.display = 'block';\n",
       "            warnings.textContent =\n",
       "                'This browser does not support binary websocket messages. ' +\n",
       "                'Performance may be slow.';\n",
       "        }\n",
       "    }\n",
       "\n",
       "    this.imageObj = new Image();\n",
       "\n",
       "    this.context = undefined;\n",
       "    this.message = undefined;\n",
       "    this.canvas = undefined;\n",
       "    this.rubberband_canvas = undefined;\n",
       "    this.rubberband_context = undefined;\n",
       "    this.format_dropdown = undefined;\n",
       "\n",
       "    this.image_mode = 'full';\n",
       "\n",
       "    this.root = document.createElement('div');\n",
       "    this.root.setAttribute('style', 'display: inline-block');\n",
       "    this._root_extra_style(this.root);\n",
       "\n",
       "    parent_element.appendChild(this.root);\n",
       "\n",
       "    this._init_header(this);\n",
       "    this._init_canvas(this);\n",
       "    this._init_toolbar(this);\n",
       "\n",
       "    var fig = this;\n",
       "\n",
       "    this.waiting = false;\n",
       "\n",
       "    this.ws.onopen = function () {\n",
       "        fig.send_message('supports_binary', { value: fig.supports_binary });\n",
       "        fig.send_message('send_image_mode', {});\n",
       "        if (fig.ratio !== 1) {\n",
       "            fig.send_message('set_dpi_ratio', { dpi_ratio: fig.ratio });\n",
       "        }\n",
       "        fig.send_message('refresh', {});\n",
       "    };\n",
       "\n",
       "    this.imageObj.onload = function () {\n",
       "        if (fig.image_mode === 'full') {\n",
       "            // Full images could contain transparency (where diff images\n",
       "            // almost always do), so we need to clear the canvas so that\n",
       "            // there is no ghosting.\n",
       "            fig.context.clearRect(0, 0, fig.canvas.width, fig.canvas.height);\n",
       "        }\n",
       "        fig.context.drawImage(fig.imageObj, 0, 0);\n",
       "    };\n",
       "\n",
       "    this.imageObj.onunload = function () {\n",
       "        fig.ws.close();\n",
       "    };\n",
       "\n",
       "    this.ws.onmessage = this._make_on_message_function(this);\n",
       "\n",
       "    this.ondownload = ondownload;\n",
       "};\n",
       "\n",
       "mpl.figure.prototype._init_header = function () {\n",
       "    var titlebar = document.createElement('div');\n",
       "    titlebar.classList =\n",
       "        'ui-dialog-titlebar ui-widget-header ui-corner-all ui-helper-clearfix';\n",
       "    var titletext = document.createElement('div');\n",
       "    titletext.classList = 'ui-dialog-title';\n",
       "    titletext.setAttribute(\n",
       "        'style',\n",
       "        'width: 100%; text-align: center; padding: 3px;'\n",
       "    );\n",
       "    titlebar.appendChild(titletext);\n",
       "    this.root.appendChild(titlebar);\n",
       "    this.header = titletext;\n",
       "};\n",
       "\n",
       "mpl.figure.prototype._canvas_extra_style = function (_canvas_div) {};\n",
       "\n",
       "mpl.figure.prototype._root_extra_style = function (_canvas_div) {};\n",
       "\n",
       "mpl.figure.prototype._init_canvas = function () {\n",
       "    var fig = this;\n",
       "\n",
       "    var canvas_div = (this.canvas_div = document.createElement('div'));\n",
       "    canvas_div.setAttribute(\n",
       "        'style',\n",
       "        'border: 1px solid #ddd;' +\n",
       "            'box-sizing: content-box;' +\n",
       "            'clear: both;' +\n",
       "            'min-height: 1px;' +\n",
       "            'min-width: 1px;' +\n",
       "            'outline: 0;' +\n",
       "            'overflow: hidden;' +\n",
       "            'position: relative;' +\n",
       "            'resize: both;'\n",
       "    );\n",
       "\n",
       "    function on_keyboard_event_closure(name) {\n",
       "        return function (event) {\n",
       "            return fig.key_event(event, name);\n",
       "        };\n",
       "    }\n",
       "\n",
       "    canvas_div.addEventListener(\n",
       "        'keydown',\n",
       "        on_keyboard_event_closure('key_press')\n",
       "    );\n",
       "    canvas_div.addEventListener(\n",
       "        'keyup',\n",
       "        on_keyboard_event_closure('key_release')\n",
       "    );\n",
       "\n",
       "    this._canvas_extra_style(canvas_div);\n",
       "    this.root.appendChild(canvas_div);\n",
       "\n",
       "    var canvas = (this.canvas = document.createElement('canvas'));\n",
       "    canvas.classList.add('mpl-canvas');\n",
       "    canvas.setAttribute('style', 'box-sizing: content-box;');\n",
       "\n",
       "    this.context = canvas.getContext('2d');\n",
       "\n",
       "    var backingStore =\n",
       "        this.context.backingStorePixelRatio ||\n",
       "        this.context.webkitBackingStorePixelRatio ||\n",
       "        this.context.mozBackingStorePixelRatio ||\n",
       "        this.context.msBackingStorePixelRatio ||\n",
       "        this.context.oBackingStorePixelRatio ||\n",
       "        this.context.backingStorePixelRatio ||\n",
       "        1;\n",
       "\n",
       "    this.ratio = (window.devicePixelRatio || 1) / backingStore;\n",
       "    if (this.ratio !== 1) {\n",
       "        fig.send_message('set_dpi_ratio', { dpi_ratio: this.ratio });\n",
       "    }\n",
       "\n",
       "    var rubberband_canvas = (this.rubberband_canvas = document.createElement(\n",
       "        'canvas'\n",
       "    ));\n",
       "    rubberband_canvas.setAttribute(\n",
       "        'style',\n",
       "        'box-sizing: content-box; position: absolute; left: 0; top: 0; z-index: 1;'\n",
       "    );\n",
       "\n",
       "    var resizeObserver = new ResizeObserver(function (entries) {\n",
       "        var nentries = entries.length;\n",
       "        for (var i = 0; i < nentries; i++) {\n",
       "            var entry = entries[i];\n",
       "            var width, height;\n",
       "            if (entry.contentBoxSize) {\n",
       "                if (entry.contentBoxSize instanceof Array) {\n",
       "                    // Chrome 84 implements new version of spec.\n",
       "                    width = entry.contentBoxSize[0].inlineSize;\n",
       "                    height = entry.contentBoxSize[0].blockSize;\n",
       "                } else {\n",
       "                    // Firefox implements old version of spec.\n",
       "                    width = entry.contentBoxSize.inlineSize;\n",
       "                    height = entry.contentBoxSize.blockSize;\n",
       "                }\n",
       "            } else {\n",
       "                // Chrome <84 implements even older version of spec.\n",
       "                width = entry.contentRect.width;\n",
       "                height = entry.contentRect.height;\n",
       "            }\n",
       "\n",
       "            // Keep the size of the canvas and rubber band canvas in sync with\n",
       "            // the canvas container.\n",
       "            if (entry.devicePixelContentBoxSize) {\n",
       "                // Chrome 84 implements new version of spec.\n",
       "                canvas.setAttribute(\n",
       "                    'width',\n",
       "                    entry.devicePixelContentBoxSize[0].inlineSize\n",
       "                );\n",
       "                canvas.setAttribute(\n",
       "                    'height',\n",
       "                    entry.devicePixelContentBoxSize[0].blockSize\n",
       "                );\n",
       "            } else {\n",
       "                canvas.setAttribute('width', width * fig.ratio);\n",
       "                canvas.setAttribute('height', height * fig.ratio);\n",
       "            }\n",
       "            canvas.setAttribute(\n",
       "                'style',\n",
       "                'width: ' + width + 'px; height: ' + height + 'px;'\n",
       "            );\n",
       "\n",
       "            rubberband_canvas.setAttribute('width', width);\n",
       "            rubberband_canvas.setAttribute('height', height);\n",
       "\n",
       "            // And update the size in Python. We ignore the initial 0/0 size\n",
       "            // that occurs as the element is placed into the DOM, which should\n",
       "            // otherwise not happen due to the minimum size styling.\n",
       "            if (width != 0 && height != 0) {\n",
       "                fig.request_resize(width, height);\n",
       "            }\n",
       "        }\n",
       "    });\n",
       "    resizeObserver.observe(canvas_div);\n",
       "\n",
       "    function on_mouse_event_closure(name) {\n",
       "        return function (event) {\n",
       "            return fig.mouse_event(event, name);\n",
       "        };\n",
       "    }\n",
       "\n",
       "    rubberband_canvas.addEventListener(\n",
       "        'mousedown',\n",
       "        on_mouse_event_closure('button_press')\n",
       "    );\n",
       "    rubberband_canvas.addEventListener(\n",
       "        'mouseup',\n",
       "        on_mouse_event_closure('button_release')\n",
       "    );\n",
       "    // Throttle sequential mouse events to 1 every 20ms.\n",
       "    rubberband_canvas.addEventListener(\n",
       "        'mousemove',\n",
       "        on_mouse_event_closure('motion_notify')\n",
       "    );\n",
       "\n",
       "    rubberband_canvas.addEventListener(\n",
       "        'mouseenter',\n",
       "        on_mouse_event_closure('figure_enter')\n",
       "    );\n",
       "    rubberband_canvas.addEventListener(\n",
       "        'mouseleave',\n",
       "        on_mouse_event_closure('figure_leave')\n",
       "    );\n",
       "\n",
       "    canvas_div.addEventListener('wheel', function (event) {\n",
       "        if (event.deltaY < 0) {\n",
       "            event.step = 1;\n",
       "        } else {\n",
       "            event.step = -1;\n",
       "        }\n",
       "        on_mouse_event_closure('scroll')(event);\n",
       "    });\n",
       "\n",
       "    canvas_div.appendChild(canvas);\n",
       "    canvas_div.appendChild(rubberband_canvas);\n",
       "\n",
       "    this.rubberband_context = rubberband_canvas.getContext('2d');\n",
       "    this.rubberband_context.strokeStyle = '#000000';\n",
       "\n",
       "    this._resize_canvas = function (width, height, forward) {\n",
       "        if (forward) {\n",
       "            canvas_div.style.width = width + 'px';\n",
       "            canvas_div.style.height = height + 'px';\n",
       "        }\n",
       "    };\n",
       "\n",
       "    // Disable right mouse context menu.\n",
       "    this.rubberband_canvas.addEventListener('contextmenu', function (_e) {\n",
       "        event.preventDefault();\n",
       "        return false;\n",
       "    });\n",
       "\n",
       "    function set_focus() {\n",
       "        canvas.focus();\n",
       "        canvas_div.focus();\n",
       "    }\n",
       "\n",
       "    window.setTimeout(set_focus, 100);\n",
       "};\n",
       "\n",
       "mpl.figure.prototype._init_toolbar = function () {\n",
       "    var fig = this;\n",
       "\n",
       "    var toolbar = document.createElement('div');\n",
       "    toolbar.classList = 'mpl-toolbar';\n",
       "    this.root.appendChild(toolbar);\n",
       "\n",
       "    function on_click_closure(name) {\n",
       "        return function (_event) {\n",
       "            return fig.toolbar_button_onclick(name);\n",
       "        };\n",
       "    }\n",
       "\n",
       "    function on_mouseover_closure(tooltip) {\n",
       "        return function (event) {\n",
       "            if (!event.currentTarget.disabled) {\n",
       "                return fig.toolbar_button_onmouseover(tooltip);\n",
       "            }\n",
       "        };\n",
       "    }\n",
       "\n",
       "    fig.buttons = {};\n",
       "    var buttonGroup = document.createElement('div');\n",
       "    buttonGroup.classList = 'mpl-button-group';\n",
       "    for (var toolbar_ind in mpl.toolbar_items) {\n",
       "        var name = mpl.toolbar_items[toolbar_ind][0];\n",
       "        var tooltip = mpl.toolbar_items[toolbar_ind][1];\n",
       "        var image = mpl.toolbar_items[toolbar_ind][2];\n",
       "        var method_name = mpl.toolbar_items[toolbar_ind][3];\n",
       "\n",
       "        if (!name) {\n",
       "            /* Instead of a spacer, we start a new button group. */\n",
       "            if (buttonGroup.hasChildNodes()) {\n",
       "                toolbar.appendChild(buttonGroup);\n",
       "            }\n",
       "            buttonGroup = document.createElement('div');\n",
       "            buttonGroup.classList = 'mpl-button-group';\n",
       "            continue;\n",
       "        }\n",
       "\n",
       "        var button = (fig.buttons[name] = document.createElement('button'));\n",
       "        button.classList = 'mpl-widget';\n",
       "        button.setAttribute('role', 'button');\n",
       "        button.setAttribute('aria-disabled', 'false');\n",
       "        button.addEventListener('click', on_click_closure(method_name));\n",
       "        button.addEventListener('mouseover', on_mouseover_closure(tooltip));\n",
       "\n",
       "        var icon_img = document.createElement('img');\n",
       "        icon_img.src = '_images/' + image + '.png';\n",
       "        icon_img.srcset = '_images/' + image + '_large.png 2x';\n",
       "        icon_img.alt = tooltip;\n",
       "        button.appendChild(icon_img);\n",
       "\n",
       "        buttonGroup.appendChild(button);\n",
       "    }\n",
       "\n",
       "    if (buttonGroup.hasChildNodes()) {\n",
       "        toolbar.appendChild(buttonGroup);\n",
       "    }\n",
       "\n",
       "    var fmt_picker = document.createElement('select');\n",
       "    fmt_picker.classList = 'mpl-widget';\n",
       "    toolbar.appendChild(fmt_picker);\n",
       "    this.format_dropdown = fmt_picker;\n",
       "\n",
       "    for (var ind in mpl.extensions) {\n",
       "        var fmt = mpl.extensions[ind];\n",
       "        var option = document.createElement('option');\n",
       "        option.selected = fmt === mpl.default_extension;\n",
       "        option.innerHTML = fmt;\n",
       "        fmt_picker.appendChild(option);\n",
       "    }\n",
       "\n",
       "    var status_bar = document.createElement('span');\n",
       "    status_bar.classList = 'mpl-message';\n",
       "    toolbar.appendChild(status_bar);\n",
       "    this.message = status_bar;\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.request_resize = function (x_pixels, y_pixels) {\n",
       "    // Request matplotlib to resize the figure. Matplotlib will then trigger a resize in the client,\n",
       "    // which will in turn request a refresh of the image.\n",
       "    this.send_message('resize', { width: x_pixels, height: y_pixels });\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.send_message = function (type, properties) {\n",
       "    properties['type'] = type;\n",
       "    properties['figure_id'] = this.id;\n",
       "    this.ws.send(JSON.stringify(properties));\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.send_draw_message = function () {\n",
       "    if (!this.waiting) {\n",
       "        this.waiting = true;\n",
       "        this.ws.send(JSON.stringify({ type: 'draw', figure_id: this.id }));\n",
       "    }\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.handle_save = function (fig, _msg) {\n",
       "    var format_dropdown = fig.format_dropdown;\n",
       "    var format = format_dropdown.options[format_dropdown.selectedIndex].value;\n",
       "    fig.ondownload(fig, format);\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.handle_resize = function (fig, msg) {\n",
       "    var size = msg['size'];\n",
       "    if (size[0] !== fig.canvas.width || size[1] !== fig.canvas.height) {\n",
       "        fig._resize_canvas(size[0], size[1], msg['forward']);\n",
       "        fig.send_message('refresh', {});\n",
       "    }\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.handle_rubberband = function (fig, msg) {\n",
       "    var x0 = msg['x0'] / fig.ratio;\n",
       "    var y0 = (fig.canvas.height - msg['y0']) / fig.ratio;\n",
       "    var x1 = msg['x1'] / fig.ratio;\n",
       "    var y1 = (fig.canvas.height - msg['y1']) / fig.ratio;\n",
       "    x0 = Math.floor(x0) + 0.5;\n",
       "    y0 = Math.floor(y0) + 0.5;\n",
       "    x1 = Math.floor(x1) + 0.5;\n",
       "    y1 = Math.floor(y1) + 0.5;\n",
       "    var min_x = Math.min(x0, x1);\n",
       "    var min_y = Math.min(y0, y1);\n",
       "    var width = Math.abs(x1 - x0);\n",
       "    var height = Math.abs(y1 - y0);\n",
       "\n",
       "    fig.rubberband_context.clearRect(\n",
       "        0,\n",
       "        0,\n",
       "        fig.canvas.width / fig.ratio,\n",
       "        fig.canvas.height / fig.ratio\n",
       "    );\n",
       "\n",
       "    fig.rubberband_context.strokeRect(min_x, min_y, width, height);\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.handle_figure_label = function (fig, msg) {\n",
       "    // Updates the figure title.\n",
       "    fig.header.textContent = msg['label'];\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.handle_cursor = function (fig, msg) {\n",
       "    var cursor = msg['cursor'];\n",
       "    switch (cursor) {\n",
       "        case 0:\n",
       "            cursor = 'pointer';\n",
       "            break;\n",
       "        case 1:\n",
       "            cursor = 'default';\n",
       "            break;\n",
       "        case 2:\n",
       "            cursor = 'crosshair';\n",
       "            break;\n",
       "        case 3:\n",
       "            cursor = 'move';\n",
       "            break;\n",
       "    }\n",
       "    fig.rubberband_canvas.style.cursor = cursor;\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.handle_message = function (fig, msg) {\n",
       "    fig.message.textContent = msg['message'];\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.handle_draw = function (fig, _msg) {\n",
       "    // Request the server to send over a new figure.\n",
       "    fig.send_draw_message();\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.handle_image_mode = function (fig, msg) {\n",
       "    fig.image_mode = msg['mode'];\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.handle_history_buttons = function (fig, msg) {\n",
       "    for (var key in msg) {\n",
       "        if (!(key in fig.buttons)) {\n",
       "            continue;\n",
       "        }\n",
       "        fig.buttons[key].disabled = !msg[key];\n",
       "        fig.buttons[key].setAttribute('aria-disabled', !msg[key]);\n",
       "    }\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.handle_navigate_mode = function (fig, msg) {\n",
       "    if (msg['mode'] === 'PAN') {\n",
       "        fig.buttons['Pan'].classList.add('active');\n",
       "        fig.buttons['Zoom'].classList.remove('active');\n",
       "    } else if (msg['mode'] === 'ZOOM') {\n",
       "        fig.buttons['Pan'].classList.remove('active');\n",
       "        fig.buttons['Zoom'].classList.add('active');\n",
       "    } else {\n",
       "        fig.buttons['Pan'].classList.remove('active');\n",
       "        fig.buttons['Zoom'].classList.remove('active');\n",
       "    }\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.updated_canvas_event = function () {\n",
       "    // Called whenever the canvas gets updated.\n",
       "    this.send_message('ack', {});\n",
       "};\n",
       "\n",
       "// A function to construct a web socket function for onmessage handling.\n",
       "// Called in the figure constructor.\n",
       "mpl.figure.prototype._make_on_message_function = function (fig) {\n",
       "    return function socket_on_message(evt) {\n",
       "        if (evt.data instanceof Blob) {\n",
       "            /* FIXME: We get \"Resource interpreted as Image but\n",
       "             * transferred with MIME type text/plain:\" errors on\n",
       "             * Chrome.  But how to set the MIME type?  It doesn't seem\n",
       "             * to be part of the websocket stream */\n",
       "            evt.data.type = 'image/png';\n",
       "\n",
       "            /* Free the memory for the previous frames */\n",
       "            if (fig.imageObj.src) {\n",
       "                (window.URL || window.webkitURL).revokeObjectURL(\n",
       "                    fig.imageObj.src\n",
       "                );\n",
       "            }\n",
       "\n",
       "            fig.imageObj.src = (window.URL || window.webkitURL).createObjectURL(\n",
       "                evt.data\n",
       "            );\n",
       "            fig.updated_canvas_event();\n",
       "            fig.waiting = false;\n",
       "            return;\n",
       "        } else if (\n",
       "            typeof evt.data === 'string' &&\n",
       "            evt.data.slice(0, 21) === 'data:image/png;base64'\n",
       "        ) {\n",
       "            fig.imageObj.src = evt.data;\n",
       "            fig.updated_canvas_event();\n",
       "            fig.waiting = false;\n",
       "            return;\n",
       "        }\n",
       "\n",
       "        var msg = JSON.parse(evt.data);\n",
       "        var msg_type = msg['type'];\n",
       "\n",
       "        // Call the  \"handle_{type}\" callback, which takes\n",
       "        // the figure and JSON message as its only arguments.\n",
       "        try {\n",
       "            var callback = fig['handle_' + msg_type];\n",
       "        } catch (e) {\n",
       "            console.log(\n",
       "                \"No handler for the '\" + msg_type + \"' message type: \",\n",
       "                msg\n",
       "            );\n",
       "            return;\n",
       "        }\n",
       "\n",
       "        if (callback) {\n",
       "            try {\n",
       "                // console.log(\"Handling '\" + msg_type + \"' message: \", msg);\n",
       "                callback(fig, msg);\n",
       "            } catch (e) {\n",
       "                console.log(\n",
       "                    \"Exception inside the 'handler_\" + msg_type + \"' callback:\",\n",
       "                    e,\n",
       "                    e.stack,\n",
       "                    msg\n",
       "                );\n",
       "            }\n",
       "        }\n",
       "    };\n",
       "};\n",
       "\n",
       "// from http://stackoverflow.com/questions/1114465/getting-mouse-location-in-canvas\n",
       "mpl.findpos = function (e) {\n",
       "    //this section is from http://www.quirksmode.org/js/events_properties.html\n",
       "    var targ;\n",
       "    if (!e) {\n",
       "        e = window.event;\n",
       "    }\n",
       "    if (e.target) {\n",
       "        targ = e.target;\n",
       "    } else if (e.srcElement) {\n",
       "        targ = e.srcElement;\n",
       "    }\n",
       "    if (targ.nodeType === 3) {\n",
       "        // defeat Safari bug\n",
       "        targ = targ.parentNode;\n",
       "    }\n",
       "\n",
       "    // pageX,Y are the mouse positions relative to the document\n",
       "    var boundingRect = targ.getBoundingClientRect();\n",
       "    var x = e.pageX - (boundingRect.left + document.body.scrollLeft);\n",
       "    var y = e.pageY - (boundingRect.top + document.body.scrollTop);\n",
       "\n",
       "    return { x: x, y: y };\n",
       "};\n",
       "\n",
       "/*\n",
       " * return a copy of an object with only non-object keys\n",
       " * we need this to avoid circular references\n",
       " * http://stackoverflow.com/a/24161582/3208463\n",
       " */\n",
       "function simpleKeys(original) {\n",
       "    return Object.keys(original).reduce(function (obj, key) {\n",
       "        if (typeof original[key] !== 'object') {\n",
       "            obj[key] = original[key];\n",
       "        }\n",
       "        return obj;\n",
       "    }, {});\n",
       "}\n",
       "\n",
       "mpl.figure.prototype.mouse_event = function (event, name) {\n",
       "    var canvas_pos = mpl.findpos(event);\n",
       "\n",
       "    if (name === 'button_press') {\n",
       "        this.canvas.focus();\n",
       "        this.canvas_div.focus();\n",
       "    }\n",
       "\n",
       "    var x = canvas_pos.x * this.ratio;\n",
       "    var y = canvas_pos.y * this.ratio;\n",
       "\n",
       "    this.send_message(name, {\n",
       "        x: x,\n",
       "        y: y,\n",
       "        button: event.button,\n",
       "        step: event.step,\n",
       "        guiEvent: simpleKeys(event),\n",
       "    });\n",
       "\n",
       "    /* This prevents the web browser from automatically changing to\n",
       "     * the text insertion cursor when the button is pressed.  We want\n",
       "     * to control all of the cursor setting manually through the\n",
       "     * 'cursor' event from matplotlib */\n",
       "    event.preventDefault();\n",
       "    return false;\n",
       "};\n",
       "\n",
       "mpl.figure.prototype._key_event_extra = function (_event, _name) {\n",
       "    // Handle any extra behaviour associated with a key event\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.key_event = function (event, name) {\n",
       "    // Prevent repeat events\n",
       "    if (name === 'key_press') {\n",
       "        if (event.which === this._key) {\n",
       "            return;\n",
       "        } else {\n",
       "            this._key = event.which;\n",
       "        }\n",
       "    }\n",
       "    if (name === 'key_release') {\n",
       "        this._key = null;\n",
       "    }\n",
       "\n",
       "    var value = '';\n",
       "    if (event.ctrlKey && event.which !== 17) {\n",
       "        value += 'ctrl+';\n",
       "    }\n",
       "    if (event.altKey && event.which !== 18) {\n",
       "        value += 'alt+';\n",
       "    }\n",
       "    if (event.shiftKey && event.which !== 16) {\n",
       "        value += 'shift+';\n",
       "    }\n",
       "\n",
       "    value += 'k';\n",
       "    value += event.which.toString();\n",
       "\n",
       "    this._key_event_extra(event, name);\n",
       "\n",
       "    this.send_message(name, { key: value, guiEvent: simpleKeys(event) });\n",
       "    return false;\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.toolbar_button_onclick = function (name) {\n",
       "    if (name === 'download') {\n",
       "        this.handle_save(this, null);\n",
       "    } else {\n",
       "        this.send_message('toolbar_button', { name: name });\n",
       "    }\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.toolbar_button_onmouseover = function (tooltip) {\n",
       "    this.message.textContent = tooltip;\n",
       "};\n",
       "mpl.toolbar_items = [[\"Home\", \"Reset original view\", \"fa fa-home icon-home\", \"home\"], [\"Back\", \"Back to previous view\", \"fa fa-arrow-left icon-arrow-left\", \"back\"], [\"Forward\", \"Forward to next view\", \"fa fa-arrow-right icon-arrow-right\", \"forward\"], [\"\", \"\", \"\", \"\"], [\"Pan\", \"Left button pans, Right button zooms\\nx/y fixes axis, CTRL fixes aspect\", \"fa fa-arrows icon-move\", \"pan\"], [\"Zoom\", \"Zoom to rectangle\\nx/y fixes axis, CTRL fixes aspect\", \"fa fa-square-o icon-check-empty\", \"zoom\"], [\"\", \"\", \"\", \"\"], [\"Download\", \"Download plot\", \"fa fa-floppy-o icon-save\", \"download\"]];\n",
       "\n",
       "mpl.extensions = [\"eps\", \"jpeg\", \"pdf\", \"png\", \"ps\", \"raw\", \"svg\", \"tif\"];\n",
       "\n",
       "mpl.default_extension = \"png\";/* global mpl */\n",
       "\n",
       "var comm_websocket_adapter = function (comm) {\n",
       "    // Create a \"websocket\"-like object which calls the given IPython comm\n",
       "    // object with the appropriate methods. Currently this is a non binary\n",
       "    // socket, so there is still some room for performance tuning.\n",
       "    var ws = {};\n",
       "\n",
       "    ws.close = function () {\n",
       "        comm.close();\n",
       "    };\n",
       "    ws.send = function (m) {\n",
       "        //console.log('sending', m);\n",
       "        comm.send(m);\n",
       "    };\n",
       "    // Register the callback with on_msg.\n",
       "    comm.on_msg(function (msg) {\n",
       "        //console.log('receiving', msg['content']['data'], msg);\n",
       "        // Pass the mpl event to the overridden (by mpl) onmessage function.\n",
       "        ws.onmessage(msg['content']['data']);\n",
       "    });\n",
       "    return ws;\n",
       "};\n",
       "\n",
       "mpl.mpl_figure_comm = function (comm, msg) {\n",
       "    // This is the function which gets called when the mpl process\n",
       "    // starts-up an IPython Comm through the \"matplotlib\" channel.\n",
       "\n",
       "    var id = msg.content.data.id;\n",
       "    // Get hold of the div created by the display call when the Comm\n",
       "    // socket was opened in Python.\n",
       "    var element = document.getElementById(id);\n",
       "    var ws_proxy = comm_websocket_adapter(comm);\n",
       "\n",
       "    function ondownload(figure, _format) {\n",
       "        window.open(figure.canvas.toDataURL());\n",
       "    }\n",
       "\n",
       "    var fig = new mpl.figure(id, ws_proxy, ondownload, element);\n",
       "\n",
       "    // Call onopen now - mpl needs it, as it is assuming we've passed it a real\n",
       "    // web socket which is closed, not our websocket->open comm proxy.\n",
       "    ws_proxy.onopen();\n",
       "\n",
       "    fig.parent_element = element;\n",
       "    fig.cell_info = mpl.find_output_cell(\"<div id='\" + id + \"'></div>\");\n",
       "    if (!fig.cell_info) {\n",
       "        console.error('Failed to find cell for figure', id, fig);\n",
       "        return;\n",
       "    }\n",
       "    fig.cell_info[0].output_area.element.one(\n",
       "        'cleared',\n",
       "        { fig: fig },\n",
       "        fig._remove_fig_handler\n",
       "    );\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.handle_close = function (fig, msg) {\n",
       "    var width = fig.canvas.width / fig.ratio;\n",
       "    fig.cell_info[0].output_area.element.off(\n",
       "        'cleared',\n",
       "        fig._remove_fig_handler\n",
       "    );\n",
       "\n",
       "    // Update the output cell to use the data from the current canvas.\n",
       "    fig.push_to_output();\n",
       "    var dataURL = fig.canvas.toDataURL();\n",
       "    // Re-enable the keyboard manager in IPython - without this line, in FF,\n",
       "    // the notebook keyboard shortcuts fail.\n",
       "    IPython.keyboard_manager.enable();\n",
       "    fig.parent_element.innerHTML =\n",
       "        '<img src=\"' + dataURL + '\" width=\"' + width + '\">';\n",
       "    fig.close_ws(fig, msg);\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.close_ws = function (fig, msg) {\n",
       "    fig.send_message('closing', msg);\n",
       "    // fig.ws.close()\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.push_to_output = function (_remove_interactive) {\n",
       "    // Turn the data on the canvas into data in the output cell.\n",
       "    var width = this.canvas.width / this.ratio;\n",
       "    var dataURL = this.canvas.toDataURL();\n",
       "    this.cell_info[1]['text/html'] =\n",
       "        '<img src=\"' + dataURL + '\" width=\"' + width + '\">';\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.updated_canvas_event = function () {\n",
       "    // Tell IPython that the notebook contents must change.\n",
       "    IPython.notebook.set_dirty(true);\n",
       "    this.send_message('ack', {});\n",
       "    var fig = this;\n",
       "    // Wait a second, then push the new image to the DOM so\n",
       "    // that it is saved nicely (might be nice to debounce this).\n",
       "    setTimeout(function () {\n",
       "        fig.push_to_output();\n",
       "    }, 1000);\n",
       "};\n",
       "\n",
       "mpl.figure.prototype._init_toolbar = function () {\n",
       "    var fig = this;\n",
       "\n",
       "    var toolbar = document.createElement('div');\n",
       "    toolbar.classList = 'btn-toolbar';\n",
       "    this.root.appendChild(toolbar);\n",
       "\n",
       "    function on_click_closure(name) {\n",
       "        return function (_event) {\n",
       "            return fig.toolbar_button_onclick(name);\n",
       "        };\n",
       "    }\n",
       "\n",
       "    function on_mouseover_closure(tooltip) {\n",
       "        return function (event) {\n",
       "            if (!event.currentTarget.disabled) {\n",
       "                return fig.toolbar_button_onmouseover(tooltip);\n",
       "            }\n",
       "        };\n",
       "    }\n",
       "\n",
       "    fig.buttons = {};\n",
       "    var buttonGroup = document.createElement('div');\n",
       "    buttonGroup.classList = 'btn-group';\n",
       "    var button;\n",
       "    for (var toolbar_ind in mpl.toolbar_items) {\n",
       "        var name = mpl.toolbar_items[toolbar_ind][0];\n",
       "        var tooltip = mpl.toolbar_items[toolbar_ind][1];\n",
       "        var image = mpl.toolbar_items[toolbar_ind][2];\n",
       "        var method_name = mpl.toolbar_items[toolbar_ind][3];\n",
       "\n",
       "        if (!name) {\n",
       "            /* Instead of a spacer, we start a new button group. */\n",
       "            if (buttonGroup.hasChildNodes()) {\n",
       "                toolbar.appendChild(buttonGroup);\n",
       "            }\n",
       "            buttonGroup = document.createElement('div');\n",
       "            buttonGroup.classList = 'btn-group';\n",
       "            continue;\n",
       "        }\n",
       "\n",
       "        button = fig.buttons[name] = document.createElement('button');\n",
       "        button.classList = 'btn btn-default';\n",
       "        button.href = '#';\n",
       "        button.title = name;\n",
       "        button.innerHTML = '<i class=\"fa ' + image + ' fa-lg\"></i>';\n",
       "        button.addEventListener('click', on_click_closure(method_name));\n",
       "        button.addEventListener('mouseover', on_mouseover_closure(tooltip));\n",
       "        buttonGroup.appendChild(button);\n",
       "    }\n",
       "\n",
       "    if (buttonGroup.hasChildNodes()) {\n",
       "        toolbar.appendChild(buttonGroup);\n",
       "    }\n",
       "\n",
       "    // Add the status bar.\n",
       "    var status_bar = document.createElement('span');\n",
       "    status_bar.classList = 'mpl-message pull-right';\n",
       "    toolbar.appendChild(status_bar);\n",
       "    this.message = status_bar;\n",
       "\n",
       "    // Add the close button to the window.\n",
       "    var buttongrp = document.createElement('div');\n",
       "    buttongrp.classList = 'btn-group inline pull-right';\n",
       "    button = document.createElement('button');\n",
       "    button.classList = 'btn btn-mini btn-primary';\n",
       "    button.href = '#';\n",
       "    button.title = 'Stop Interaction';\n",
       "    button.innerHTML = '<i class=\"fa fa-power-off icon-remove icon-large\"></i>';\n",
       "    button.addEventListener('click', function (_evt) {\n",
       "        fig.handle_close(fig, {});\n",
       "    });\n",
       "    button.addEventListener(\n",
       "        'mouseover',\n",
       "        on_mouseover_closure('Stop Interaction')\n",
       "    );\n",
       "    buttongrp.appendChild(button);\n",
       "    var titlebar = this.root.querySelector('.ui-dialog-titlebar');\n",
       "    titlebar.insertBefore(buttongrp, titlebar.firstChild);\n",
       "};\n",
       "\n",
       "mpl.figure.prototype._remove_fig_handler = function (event) {\n",
       "    var fig = event.data.fig;\n",
       "    fig.close_ws(fig, {});\n",
       "};\n",
       "\n",
       "mpl.figure.prototype._root_extra_style = function (el) {\n",
       "    el.style.boxSizing = 'content-box'; // override notebook setting of border-box.\n",
       "};\n",
       "\n",
       "mpl.figure.prototype._canvas_extra_style = function (el) {\n",
       "    // this is important to make the div 'focusable\n",
       "    el.setAttribute('tabindex', 0);\n",
       "    // reach out to IPython and tell the keyboard manager to turn it's self\n",
       "    // off when our div gets focus\n",
       "\n",
       "    // location in version 3\n",
       "    if (IPython.notebook.keyboard_manager) {\n",
       "        IPython.notebook.keyboard_manager.register_events(el);\n",
       "    } else {\n",
       "        // location in version 2\n",
       "        IPython.keyboard_manager.register_events(el);\n",
       "    }\n",
       "};\n",
       "\n",
       "mpl.figure.prototype._key_event_extra = function (event, _name) {\n",
       "    var manager = IPython.notebook.keyboard_manager;\n",
       "    if (!manager) {\n",
       "        manager = IPython.keyboard_manager;\n",
       "    }\n",
       "\n",
       "    // Check for shift+enter\n",
       "    if (event.shiftKey && event.which === 13) {\n",
       "        this.canvas_div.blur();\n",
       "        // select the cell after this one\n",
       "        var index = IPython.notebook.find_cell_index(this.cell_info[0]);\n",
       "        IPython.notebook.select(index + 1);\n",
       "    }\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.handle_save = function (fig, _msg) {\n",
       "    fig.ondownload(fig, null);\n",
       "};\n",
       "\n",
       "mpl.find_output_cell = function (html_output) {\n",
       "    // Return the cell and output element which can be found *uniquely* in the notebook.\n",
       "    // Note - this is a bit hacky, but it is done because the \"notebook_saving.Notebook\"\n",
       "    // IPython event is triggered only after the cells have been serialised, which for\n",
       "    // our purposes (turning an active figure into a static one), is too late.\n",
       "    var cells = IPython.notebook.get_cells();\n",
       "    var ncells = cells.length;\n",
       "    for (var i = 0; i < ncells; i++) {\n",
       "        var cell = cells[i];\n",
       "        if (cell.cell_type === 'code') {\n",
       "            for (var j = 0; j < cell.output_area.outputs.length; j++) {\n",
       "                var data = cell.output_area.outputs[j];\n",
       "                if (data.data) {\n",
       "                    // IPython >= 3 moved mimebundle to data attribute of output\n",
       "                    data = data.data;\n",
       "                }\n",
       "                if (data['text/html'] === html_output) {\n",
       "                    return [cell, data, j];\n",
       "                }\n",
       "            }\n",
       "        }\n",
       "    }\n",
       "};\n",
       "\n",
       "// Register the function which deals with the matplotlib target/channel.\n",
       "// The kernel may be null if the page has been refreshed.\n",
       "if (IPython.notebook.kernel !== null) {\n",
       "    IPython.notebook.kernel.comm_manager.register_target(\n",
       "        'matplotlib',\n",
       "        mpl.mpl_figure_comm\n",
       "    );\n",
       "}\n"
      ],
      "text/plain": [
       "<IPython.core.display.Javascript object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/html": [
       "<img src=\"data:image/png;base64,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\" width=\"900\">"
      ],
      "text/plain": [
       "<IPython.core.display.HTML object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "rmse_in_different_test_part()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "d05dddc5",
   "metadata": {},
   "source": [
    "## Price-Quantile"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 72,
   "id": "c6314e0b",
   "metadata": {},
   "outputs": [],
   "source": [
    "def price_quantile():\n",
    "    df_y = [df.quantile(i*0.01).price for i in range(0, 101)]\n",
    "    df_x = [(i)*0.01 for i in range(len(df_y))]\n",
    "    quarter_y = [df.quantile(i*0.1).price for i in range(11)]\n",
    "    quarter_x = [(i)*0.1 for i in range(11)]\n",
    "\n",
    "    plt.figure(figsize=(9,6))\n",
    "#     plt.scatter(df_x, df_y)\n",
    "    plt.plot(df_x, df_y)\n",
    "    plt.scatter(quarter_x, quarter_y, c='r', label=\"Price with quantile{0.0,0.1,....1.0}\")\n",
    "    \n",
    "    plt.xlabel(\"Quantile\")\n",
    "    plt.ylabel(\"Price\")\n",
    "    plt.legend()\n",
    "    plt.show()\n",
    "    plt.savefig(\"x2 price of different quantile.png\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 73,
   "id": "e85926eb",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "application/javascript": [
       "/* Put everything inside the global mpl namespace */\n",
       "/* global mpl */\n",
       "window.mpl = {};\n",
       "\n",
       "mpl.get_websocket_type = function () {\n",
       "    if (typeof WebSocket !== 'undefined') {\n",
       "        return WebSocket;\n",
       "    } else if (typeof MozWebSocket !== 'undefined') {\n",
       "        return MozWebSocket;\n",
       "    } else {\n",
       "        alert(\n",
       "            'Your browser does not have WebSocket support. ' +\n",
       "                'Please try Chrome, Safari or Firefox ≥ 6. ' +\n",
       "                'Firefox 4 and 5 are also supported but you ' +\n",
       "                'have to enable WebSockets in about:config.'\n",
       "        );\n",
       "    }\n",
       "};\n",
       "\n",
       "mpl.figure = function (figure_id, websocket, ondownload, parent_element) {\n",
       "    this.id = figure_id;\n",
       "\n",
       "    this.ws = websocket;\n",
       "\n",
       "    this.supports_binary = this.ws.binaryType !== undefined;\n",
       "\n",
       "    if (!this.supports_binary) {\n",
       "        var warnings = document.getElementById('mpl-warnings');\n",
       "        if (warnings) {\n",
       "            warnings.style.display = 'block';\n",
       "            warnings.textContent =\n",
       "                'This browser does not support binary websocket messages. ' +\n",
       "                'Performance may be slow.';\n",
       "        }\n",
       "    }\n",
       "\n",
       "    this.imageObj = new Image();\n",
       "\n",
       "    this.context = undefined;\n",
       "    this.message = undefined;\n",
       "    this.canvas = undefined;\n",
       "    this.rubberband_canvas = undefined;\n",
       "    this.rubberband_context = undefined;\n",
       "    this.format_dropdown = undefined;\n",
       "\n",
       "    this.image_mode = 'full';\n",
       "\n",
       "    this.root = document.createElement('div');\n",
       "    this.root.setAttribute('style', 'display: inline-block');\n",
       "    this._root_extra_style(this.root);\n",
       "\n",
       "    parent_element.appendChild(this.root);\n",
       "\n",
       "    this._init_header(this);\n",
       "    this._init_canvas(this);\n",
       "    this._init_toolbar(this);\n",
       "\n",
       "    var fig = this;\n",
       "\n",
       "    this.waiting = false;\n",
       "\n",
       "    this.ws.onopen = function () {\n",
       "        fig.send_message('supports_binary', { value: fig.supports_binary });\n",
       "        fig.send_message('send_image_mode', {});\n",
       "        if (fig.ratio !== 1) {\n",
       "            fig.send_message('set_dpi_ratio', { dpi_ratio: fig.ratio });\n",
       "        }\n",
       "        fig.send_message('refresh', {});\n",
       "    };\n",
       "\n",
       "    this.imageObj.onload = function () {\n",
       "        if (fig.image_mode === 'full') {\n",
       "            // Full images could contain transparency (where diff images\n",
       "            // almost always do), so we need to clear the canvas so that\n",
       "            // there is no ghosting.\n",
       "            fig.context.clearRect(0, 0, fig.canvas.width, fig.canvas.height);\n",
       "        }\n",
       "        fig.context.drawImage(fig.imageObj, 0, 0);\n",
       "    };\n",
       "\n",
       "    this.imageObj.onunload = function () {\n",
       "        fig.ws.close();\n",
       "    };\n",
       "\n",
       "    this.ws.onmessage = this._make_on_message_function(this);\n",
       "\n",
       "    this.ondownload = ondownload;\n",
       "};\n",
       "\n",
       "mpl.figure.prototype._init_header = function () {\n",
       "    var titlebar = document.createElement('div');\n",
       "    titlebar.classList =\n",
       "        'ui-dialog-titlebar ui-widget-header ui-corner-all ui-helper-clearfix';\n",
       "    var titletext = document.createElement('div');\n",
       "    titletext.classList = 'ui-dialog-title';\n",
       "    titletext.setAttribute(\n",
       "        'style',\n",
       "        'width: 100%; text-align: center; padding: 3px;'\n",
       "    );\n",
       "    titlebar.appendChild(titletext);\n",
       "    this.root.appendChild(titlebar);\n",
       "    this.header = titletext;\n",
       "};\n",
       "\n",
       "mpl.figure.prototype._canvas_extra_style = function (_canvas_div) {};\n",
       "\n",
       "mpl.figure.prototype._root_extra_style = function (_canvas_div) {};\n",
       "\n",
       "mpl.figure.prototype._init_canvas = function () {\n",
       "    var fig = this;\n",
       "\n",
       "    var canvas_div = (this.canvas_div = document.createElement('div'));\n",
       "    canvas_div.setAttribute(\n",
       "        'style',\n",
       "        'border: 1px solid #ddd;' +\n",
       "            'box-sizing: content-box;' +\n",
       "            'clear: both;' +\n",
       "            'min-height: 1px;' +\n",
       "            'min-width: 1px;' +\n",
       "            'outline: 0;' +\n",
       "            'overflow: hidden;' +\n",
       "            'position: relative;' +\n",
       "            'resize: both;'\n",
       "    );\n",
       "\n",
       "    function on_keyboard_event_closure(name) {\n",
       "        return function (event) {\n",
       "            return fig.key_event(event, name);\n",
       "        };\n",
       "    }\n",
       "\n",
       "    canvas_div.addEventListener(\n",
       "        'keydown',\n",
       "        on_keyboard_event_closure('key_press')\n",
       "    );\n",
       "    canvas_div.addEventListener(\n",
       "        'keyup',\n",
       "        on_keyboard_event_closure('key_release')\n",
       "    );\n",
       "\n",
       "    this._canvas_extra_style(canvas_div);\n",
       "    this.root.appendChild(canvas_div);\n",
       "\n",
       "    var canvas = (this.canvas = document.createElement('canvas'));\n",
       "    canvas.classList.add('mpl-canvas');\n",
       "    canvas.setAttribute('style', 'box-sizing: content-box;');\n",
       "\n",
       "    this.context = canvas.getContext('2d');\n",
       "\n",
       "    var backingStore =\n",
       "        this.context.backingStorePixelRatio ||\n",
       "        this.context.webkitBackingStorePixelRatio ||\n",
       "        this.context.mozBackingStorePixelRatio ||\n",
       "        this.context.msBackingStorePixelRatio ||\n",
       "        this.context.oBackingStorePixelRatio ||\n",
       "        this.context.backingStorePixelRatio ||\n",
       "        1;\n",
       "\n",
       "    this.ratio = (window.devicePixelRatio || 1) / backingStore;\n",
       "    if (this.ratio !== 1) {\n",
       "        fig.send_message('set_dpi_ratio', { dpi_ratio: this.ratio });\n",
       "    }\n",
       "\n",
       "    var rubberband_canvas = (this.rubberband_canvas = document.createElement(\n",
       "        'canvas'\n",
       "    ));\n",
       "    rubberband_canvas.setAttribute(\n",
       "        'style',\n",
       "        'box-sizing: content-box; position: absolute; left: 0; top: 0; z-index: 1;'\n",
       "    );\n",
       "\n",
       "    var resizeObserver = new ResizeObserver(function (entries) {\n",
       "        var nentries = entries.length;\n",
       "        for (var i = 0; i < nentries; i++) {\n",
       "            var entry = entries[i];\n",
       "            var width, height;\n",
       "            if (entry.contentBoxSize) {\n",
       "                if (entry.contentBoxSize instanceof Array) {\n",
       "                    // Chrome 84 implements new version of spec.\n",
       "                    width = entry.contentBoxSize[0].inlineSize;\n",
       "                    height = entry.contentBoxSize[0].blockSize;\n",
       "                } else {\n",
       "                    // Firefox implements old version of spec.\n",
       "                    width = entry.contentBoxSize.inlineSize;\n",
       "                    height = entry.contentBoxSize.blockSize;\n",
       "                }\n",
       "            } else {\n",
       "                // Chrome <84 implements even older version of spec.\n",
       "                width = entry.contentRect.width;\n",
       "                height = entry.contentRect.height;\n",
       "            }\n",
       "\n",
       "            // Keep the size of the canvas and rubber band canvas in sync with\n",
       "            // the canvas container.\n",
       "            if (entry.devicePixelContentBoxSize) {\n",
       "                // Chrome 84 implements new version of spec.\n",
       "                canvas.setAttribute(\n",
       "                    'width',\n",
       "                    entry.devicePixelContentBoxSize[0].inlineSize\n",
       "                );\n",
       "                canvas.setAttribute(\n",
       "                    'height',\n",
       "                    entry.devicePixelContentBoxSize[0].blockSize\n",
       "                );\n",
       "            } else {\n",
       "                canvas.setAttribute('width', width * fig.ratio);\n",
       "                canvas.setAttribute('height', height * fig.ratio);\n",
       "            }\n",
       "            canvas.setAttribute(\n",
       "                'style',\n",
       "                'width: ' + width + 'px; height: ' + height + 'px;'\n",
       "            );\n",
       "\n",
       "            rubberband_canvas.setAttribute('width', width);\n",
       "            rubberband_canvas.setAttribute('height', height);\n",
       "\n",
       "            // And update the size in Python. We ignore the initial 0/0 size\n",
       "            // that occurs as the element is placed into the DOM, which should\n",
       "            // otherwise not happen due to the minimum size styling.\n",
       "            if (width != 0 && height != 0) {\n",
       "                fig.request_resize(width, height);\n",
       "            }\n",
       "        }\n",
       "    });\n",
       "    resizeObserver.observe(canvas_div);\n",
       "\n",
       "    function on_mouse_event_closure(name) {\n",
       "        return function (event) {\n",
       "            return fig.mouse_event(event, name);\n",
       "        };\n",
       "    }\n",
       "\n",
       "    rubberband_canvas.addEventListener(\n",
       "        'mousedown',\n",
       "        on_mouse_event_closure('button_press')\n",
       "    );\n",
       "    rubberband_canvas.addEventListener(\n",
       "        'mouseup',\n",
       "        on_mouse_event_closure('button_release')\n",
       "    );\n",
       "    // Throttle sequential mouse events to 1 every 20ms.\n",
       "    rubberband_canvas.addEventListener(\n",
       "        'mousemove',\n",
       "        on_mouse_event_closure('motion_notify')\n",
       "    );\n",
       "\n",
       "    rubberband_canvas.addEventListener(\n",
       "        'mouseenter',\n",
       "        on_mouse_event_closure('figure_enter')\n",
       "    );\n",
       "    rubberband_canvas.addEventListener(\n",
       "        'mouseleave',\n",
       "        on_mouse_event_closure('figure_leave')\n",
       "    );\n",
       "\n",
       "    canvas_div.addEventListener('wheel', function (event) {\n",
       "        if (event.deltaY < 0) {\n",
       "            event.step = 1;\n",
       "        } else {\n",
       "            event.step = -1;\n",
       "        }\n",
       "        on_mouse_event_closure('scroll')(event);\n",
       "    });\n",
       "\n",
       "    canvas_div.appendChild(canvas);\n",
       "    canvas_div.appendChild(rubberband_canvas);\n",
       "\n",
       "    this.rubberband_context = rubberband_canvas.getContext('2d');\n",
       "    this.rubberband_context.strokeStyle = '#000000';\n",
       "\n",
       "    this._resize_canvas = function (width, height, forward) {\n",
       "        if (forward) {\n",
       "            canvas_div.style.width = width + 'px';\n",
       "            canvas_div.style.height = height + 'px';\n",
       "        }\n",
       "    };\n",
       "\n",
       "    // Disable right mouse context menu.\n",
       "    this.rubberband_canvas.addEventListener('contextmenu', function (_e) {\n",
       "        event.preventDefault();\n",
       "        return false;\n",
       "    });\n",
       "\n",
       "    function set_focus() {\n",
       "        canvas.focus();\n",
       "        canvas_div.focus();\n",
       "    }\n",
       "\n",
       "    window.setTimeout(set_focus, 100);\n",
       "};\n",
       "\n",
       "mpl.figure.prototype._init_toolbar = function () {\n",
       "    var fig = this;\n",
       "\n",
       "    var toolbar = document.createElement('div');\n",
       "    toolbar.classList = 'mpl-toolbar';\n",
       "    this.root.appendChild(toolbar);\n",
       "\n",
       "    function on_click_closure(name) {\n",
       "        return function (_event) {\n",
       "            return fig.toolbar_button_onclick(name);\n",
       "        };\n",
       "    }\n",
       "\n",
       "    function on_mouseover_closure(tooltip) {\n",
       "        return function (event) {\n",
       "            if (!event.currentTarget.disabled) {\n",
       "                return fig.toolbar_button_onmouseover(tooltip);\n",
       "            }\n",
       "        };\n",
       "    }\n",
       "\n",
       "    fig.buttons = {};\n",
       "    var buttonGroup = document.createElement('div');\n",
       "    buttonGroup.classList = 'mpl-button-group';\n",
       "    for (var toolbar_ind in mpl.toolbar_items) {\n",
       "        var name = mpl.toolbar_items[toolbar_ind][0];\n",
       "        var tooltip = mpl.toolbar_items[toolbar_ind][1];\n",
       "        var image = mpl.toolbar_items[toolbar_ind][2];\n",
       "        var method_name = mpl.toolbar_items[toolbar_ind][3];\n",
       "\n",
       "        if (!name) {\n",
       "            /* Instead of a spacer, we start a new button group. */\n",
       "            if (buttonGroup.hasChildNodes()) {\n",
       "                toolbar.appendChild(buttonGroup);\n",
       "            }\n",
       "            buttonGroup = document.createElement('div');\n",
       "            buttonGroup.classList = 'mpl-button-group';\n",
       "            continue;\n",
       "        }\n",
       "\n",
       "        var button = (fig.buttons[name] = document.createElement('button'));\n",
       "        button.classList = 'mpl-widget';\n",
       "        button.setAttribute('role', 'button');\n",
       "        button.setAttribute('aria-disabled', 'false');\n",
       "        button.addEventListener('click', on_click_closure(method_name));\n",
       "        button.addEventListener('mouseover', on_mouseover_closure(tooltip));\n",
       "\n",
       "        var icon_img = document.createElement('img');\n",
       "        icon_img.src = '_images/' + image + '.png';\n",
       "        icon_img.srcset = '_images/' + image + '_large.png 2x';\n",
       "        icon_img.alt = tooltip;\n",
       "        button.appendChild(icon_img);\n",
       "\n",
       "        buttonGroup.appendChild(button);\n",
       "    }\n",
       "\n",
       "    if (buttonGroup.hasChildNodes()) {\n",
       "        toolbar.appendChild(buttonGroup);\n",
       "    }\n",
       "\n",
       "    var fmt_picker = document.createElement('select');\n",
       "    fmt_picker.classList = 'mpl-widget';\n",
       "    toolbar.appendChild(fmt_picker);\n",
       "    this.format_dropdown = fmt_picker;\n",
       "\n",
       "    for (var ind in mpl.extensions) {\n",
       "        var fmt = mpl.extensions[ind];\n",
       "        var option = document.createElement('option');\n",
       "        option.selected = fmt === mpl.default_extension;\n",
       "        option.innerHTML = fmt;\n",
       "        fmt_picker.appendChild(option);\n",
       "    }\n",
       "\n",
       "    var status_bar = document.createElement('span');\n",
       "    status_bar.classList = 'mpl-message';\n",
       "    toolbar.appendChild(status_bar);\n",
       "    this.message = status_bar;\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.request_resize = function (x_pixels, y_pixels) {\n",
       "    // Request matplotlib to resize the figure. Matplotlib will then trigger a resize in the client,\n",
       "    // which will in turn request a refresh of the image.\n",
       "    this.send_message('resize', { width: x_pixels, height: y_pixels });\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.send_message = function (type, properties) {\n",
       "    properties['type'] = type;\n",
       "    properties['figure_id'] = this.id;\n",
       "    this.ws.send(JSON.stringify(properties));\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.send_draw_message = function () {\n",
       "    if (!this.waiting) {\n",
       "        this.waiting = true;\n",
       "        this.ws.send(JSON.stringify({ type: 'draw', figure_id: this.id }));\n",
       "    }\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.handle_save = function (fig, _msg) {\n",
       "    var format_dropdown = fig.format_dropdown;\n",
       "    var format = format_dropdown.options[format_dropdown.selectedIndex].value;\n",
       "    fig.ondownload(fig, format);\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.handle_resize = function (fig, msg) {\n",
       "    var size = msg['size'];\n",
       "    if (size[0] !== fig.canvas.width || size[1] !== fig.canvas.height) {\n",
       "        fig._resize_canvas(size[0], size[1], msg['forward']);\n",
       "        fig.send_message('refresh', {});\n",
       "    }\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.handle_rubberband = function (fig, msg) {\n",
       "    var x0 = msg['x0'] / fig.ratio;\n",
       "    var y0 = (fig.canvas.height - msg['y0']) / fig.ratio;\n",
       "    var x1 = msg['x1'] / fig.ratio;\n",
       "    var y1 = (fig.canvas.height - msg['y1']) / fig.ratio;\n",
       "    x0 = Math.floor(x0) + 0.5;\n",
       "    y0 = Math.floor(y0) + 0.5;\n",
       "    x1 = Math.floor(x1) + 0.5;\n",
       "    y1 = Math.floor(y1) + 0.5;\n",
       "    var min_x = Math.min(x0, x1);\n",
       "    var min_y = Math.min(y0, y1);\n",
       "    var width = Math.abs(x1 - x0);\n",
       "    var height = Math.abs(y1 - y0);\n",
       "\n",
       "    fig.rubberband_context.clearRect(\n",
       "        0,\n",
       "        0,\n",
       "        fig.canvas.width / fig.ratio,\n",
       "        fig.canvas.height / fig.ratio\n",
       "    );\n",
       "\n",
       "    fig.rubberband_context.strokeRect(min_x, min_y, width, height);\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.handle_figure_label = function (fig, msg) {\n",
       "    // Updates the figure title.\n",
       "    fig.header.textContent = msg['label'];\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.handle_cursor = function (fig, msg) {\n",
       "    var cursor = msg['cursor'];\n",
       "    switch (cursor) {\n",
       "        case 0:\n",
       "            cursor = 'pointer';\n",
       "            break;\n",
       "        case 1:\n",
       "            cursor = 'default';\n",
       "            break;\n",
       "        case 2:\n",
       "            cursor = 'crosshair';\n",
       "            break;\n",
       "        case 3:\n",
       "            cursor = 'move';\n",
       "            break;\n",
       "    }\n",
       "    fig.rubberband_canvas.style.cursor = cursor;\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.handle_message = function (fig, msg) {\n",
       "    fig.message.textContent = msg['message'];\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.handle_draw = function (fig, _msg) {\n",
       "    // Request the server to send over a new figure.\n",
       "    fig.send_draw_message();\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.handle_image_mode = function (fig, msg) {\n",
       "    fig.image_mode = msg['mode'];\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.handle_history_buttons = function (fig, msg) {\n",
       "    for (var key in msg) {\n",
       "        if (!(key in fig.buttons)) {\n",
       "            continue;\n",
       "        }\n",
       "        fig.buttons[key].disabled = !msg[key];\n",
       "        fig.buttons[key].setAttribute('aria-disabled', !msg[key]);\n",
       "    }\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.handle_navigate_mode = function (fig, msg) {\n",
       "    if (msg['mode'] === 'PAN') {\n",
       "        fig.buttons['Pan'].classList.add('active');\n",
       "        fig.buttons['Zoom'].classList.remove('active');\n",
       "    } else if (msg['mode'] === 'ZOOM') {\n",
       "        fig.buttons['Pan'].classList.remove('active');\n",
       "        fig.buttons['Zoom'].classList.add('active');\n",
       "    } else {\n",
       "        fig.buttons['Pan'].classList.remove('active');\n",
       "        fig.buttons['Zoom'].classList.remove('active');\n",
       "    }\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.updated_canvas_event = function () {\n",
       "    // Called whenever the canvas gets updated.\n",
       "    this.send_message('ack', {});\n",
       "};\n",
       "\n",
       "// A function to construct a web socket function for onmessage handling.\n",
       "// Called in the figure constructor.\n",
       "mpl.figure.prototype._make_on_message_function = function (fig) {\n",
       "    return function socket_on_message(evt) {\n",
       "        if (evt.data instanceof Blob) {\n",
       "            /* FIXME: We get \"Resource interpreted as Image but\n",
       "             * transferred with MIME type text/plain:\" errors on\n",
       "             * Chrome.  But how to set the MIME type?  It doesn't seem\n",
       "             * to be part of the websocket stream */\n",
       "            evt.data.type = 'image/png';\n",
       "\n",
       "            /* Free the memory for the previous frames */\n",
       "            if (fig.imageObj.src) {\n",
       "                (window.URL || window.webkitURL).revokeObjectURL(\n",
       "                    fig.imageObj.src\n",
       "                );\n",
       "            }\n",
       "\n",
       "            fig.imageObj.src = (window.URL || window.webkitURL).createObjectURL(\n",
       "                evt.data\n",
       "            );\n",
       "            fig.updated_canvas_event();\n",
       "            fig.waiting = false;\n",
       "            return;\n",
       "        } else if (\n",
       "            typeof evt.data === 'string' &&\n",
       "            evt.data.slice(0, 21) === 'data:image/png;base64'\n",
       "        ) {\n",
       "            fig.imageObj.src = evt.data;\n",
       "            fig.updated_canvas_event();\n",
       "            fig.waiting = false;\n",
       "            return;\n",
       "        }\n",
       "\n",
       "        var msg = JSON.parse(evt.data);\n",
       "        var msg_type = msg['type'];\n",
       "\n",
       "        // Call the  \"handle_{type}\" callback, which takes\n",
       "        // the figure and JSON message as its only arguments.\n",
       "        try {\n",
       "            var callback = fig['handle_' + msg_type];\n",
       "        } catch (e) {\n",
       "            console.log(\n",
       "                \"No handler for the '\" + msg_type + \"' message type: \",\n",
       "                msg\n",
       "            );\n",
       "            return;\n",
       "        }\n",
       "\n",
       "        if (callback) {\n",
       "            try {\n",
       "                // console.log(\"Handling '\" + msg_type + \"' message: \", msg);\n",
       "                callback(fig, msg);\n",
       "            } catch (e) {\n",
       "                console.log(\n",
       "                    \"Exception inside the 'handler_\" + msg_type + \"' callback:\",\n",
       "                    e,\n",
       "                    e.stack,\n",
       "                    msg\n",
       "                );\n",
       "            }\n",
       "        }\n",
       "    };\n",
       "};\n",
       "\n",
       "// from http://stackoverflow.com/questions/1114465/getting-mouse-location-in-canvas\n",
       "mpl.findpos = function (e) {\n",
       "    //this section is from http://www.quirksmode.org/js/events_properties.html\n",
       "    var targ;\n",
       "    if (!e) {\n",
       "        e = window.event;\n",
       "    }\n",
       "    if (e.target) {\n",
       "        targ = e.target;\n",
       "    } else if (e.srcElement) {\n",
       "        targ = e.srcElement;\n",
       "    }\n",
       "    if (targ.nodeType === 3) {\n",
       "        // defeat Safari bug\n",
       "        targ = targ.parentNode;\n",
       "    }\n",
       "\n",
       "    // pageX,Y are the mouse positions relative to the document\n",
       "    var boundingRect = targ.getBoundingClientRect();\n",
       "    var x = e.pageX - (boundingRect.left + document.body.scrollLeft);\n",
       "    var y = e.pageY - (boundingRect.top + document.body.scrollTop);\n",
       "\n",
       "    return { x: x, y: y };\n",
       "};\n",
       "\n",
       "/*\n",
       " * return a copy of an object with only non-object keys\n",
       " * we need this to avoid circular references\n",
       " * http://stackoverflow.com/a/24161582/3208463\n",
       " */\n",
       "function simpleKeys(original) {\n",
       "    return Object.keys(original).reduce(function (obj, key) {\n",
       "        if (typeof original[key] !== 'object') {\n",
       "            obj[key] = original[key];\n",
       "        }\n",
       "        return obj;\n",
       "    }, {});\n",
       "}\n",
       "\n",
       "mpl.figure.prototype.mouse_event = function (event, name) {\n",
       "    var canvas_pos = mpl.findpos(event);\n",
       "\n",
       "    if (name === 'button_press') {\n",
       "        this.canvas.focus();\n",
       "        this.canvas_div.focus();\n",
       "    }\n",
       "\n",
       "    var x = canvas_pos.x * this.ratio;\n",
       "    var y = canvas_pos.y * this.ratio;\n",
       "\n",
       "    this.send_message(name, {\n",
       "        x: x,\n",
       "        y: y,\n",
       "        button: event.button,\n",
       "        step: event.step,\n",
       "        guiEvent: simpleKeys(event),\n",
       "    });\n",
       "\n",
       "    /* This prevents the web browser from automatically changing to\n",
       "     * the text insertion cursor when the button is pressed.  We want\n",
       "     * to control all of the cursor setting manually through the\n",
       "     * 'cursor' event from matplotlib */\n",
       "    event.preventDefault();\n",
       "    return false;\n",
       "};\n",
       "\n",
       "mpl.figure.prototype._key_event_extra = function (_event, _name) {\n",
       "    // Handle any extra behaviour associated with a key event\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.key_event = function (event, name) {\n",
       "    // Prevent repeat events\n",
       "    if (name === 'key_press') {\n",
       "        if (event.which === this._key) {\n",
       "            return;\n",
       "        } else {\n",
       "            this._key = event.which;\n",
       "        }\n",
       "    }\n",
       "    if (name === 'key_release') {\n",
       "        this._key = null;\n",
       "    }\n",
       "\n",
       "    var value = '';\n",
       "    if (event.ctrlKey && event.which !== 17) {\n",
       "        value += 'ctrl+';\n",
       "    }\n",
       "    if (event.altKey && event.which !== 18) {\n",
       "        value += 'alt+';\n",
       "    }\n",
       "    if (event.shiftKey && event.which !== 16) {\n",
       "        value += 'shift+';\n",
       "    }\n",
       "\n",
       "    value += 'k';\n",
       "    value += event.which.toString();\n",
       "\n",
       "    this._key_event_extra(event, name);\n",
       "\n",
       "    this.send_message(name, { key: value, guiEvent: simpleKeys(event) });\n",
       "    return false;\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.toolbar_button_onclick = function (name) {\n",
       "    if (name === 'download') {\n",
       "        this.handle_save(this, null);\n",
       "    } else {\n",
       "        this.send_message('toolbar_button', { name: name });\n",
       "    }\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.toolbar_button_onmouseover = function (tooltip) {\n",
       "    this.message.textContent = tooltip;\n",
       "};\n",
       "mpl.toolbar_items = [[\"Home\", \"Reset original view\", \"fa fa-home icon-home\", \"home\"], [\"Back\", \"Back to previous view\", \"fa fa-arrow-left icon-arrow-left\", \"back\"], [\"Forward\", \"Forward to next view\", \"fa fa-arrow-right icon-arrow-right\", \"forward\"], [\"\", \"\", \"\", \"\"], [\"Pan\", \"Left button pans, Right button zooms\\nx/y fixes axis, CTRL fixes aspect\", \"fa fa-arrows icon-move\", \"pan\"], [\"Zoom\", \"Zoom to rectangle\\nx/y fixes axis, CTRL fixes aspect\", \"fa fa-square-o icon-check-empty\", \"zoom\"], [\"\", \"\", \"\", \"\"], [\"Download\", \"Download plot\", \"fa fa-floppy-o icon-save\", \"download\"]];\n",
       "\n",
       "mpl.extensions = [\"eps\", \"jpeg\", \"pdf\", \"png\", \"ps\", \"raw\", \"svg\", \"tif\"];\n",
       "\n",
       "mpl.default_extension = \"png\";/* global mpl */\n",
       "\n",
       "var comm_websocket_adapter = function (comm) {\n",
       "    // Create a \"websocket\"-like object which calls the given IPython comm\n",
       "    // object with the appropriate methods. Currently this is a non binary\n",
       "    // socket, so there is still some room for performance tuning.\n",
       "    var ws = {};\n",
       "\n",
       "    ws.close = function () {\n",
       "        comm.close();\n",
       "    };\n",
       "    ws.send = function (m) {\n",
       "        //console.log('sending', m);\n",
       "        comm.send(m);\n",
       "    };\n",
       "    // Register the callback with on_msg.\n",
       "    comm.on_msg(function (msg) {\n",
       "        //console.log('receiving', msg['content']['data'], msg);\n",
       "        // Pass the mpl event to the overridden (by mpl) onmessage function.\n",
       "        ws.onmessage(msg['content']['data']);\n",
       "    });\n",
       "    return ws;\n",
       "};\n",
       "\n",
       "mpl.mpl_figure_comm = function (comm, msg) {\n",
       "    // This is the function which gets called when the mpl process\n",
       "    // starts-up an IPython Comm through the \"matplotlib\" channel.\n",
       "\n",
       "    var id = msg.content.data.id;\n",
       "    // Get hold of the div created by the display call when the Comm\n",
       "    // socket was opened in Python.\n",
       "    var element = document.getElementById(id);\n",
       "    var ws_proxy = comm_websocket_adapter(comm);\n",
       "\n",
       "    function ondownload(figure, _format) {\n",
       "        window.open(figure.canvas.toDataURL());\n",
       "    }\n",
       "\n",
       "    var fig = new mpl.figure(id, ws_proxy, ondownload, element);\n",
       "\n",
       "    // Call onopen now - mpl needs it, as it is assuming we've passed it a real\n",
       "    // web socket which is closed, not our websocket->open comm proxy.\n",
       "    ws_proxy.onopen();\n",
       "\n",
       "    fig.parent_element = element;\n",
       "    fig.cell_info = mpl.find_output_cell(\"<div id='\" + id + \"'></div>\");\n",
       "    if (!fig.cell_info) {\n",
       "        console.error('Failed to find cell for figure', id, fig);\n",
       "        return;\n",
       "    }\n",
       "    fig.cell_info[0].output_area.element.one(\n",
       "        'cleared',\n",
       "        { fig: fig },\n",
       "        fig._remove_fig_handler\n",
       "    );\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.handle_close = function (fig, msg) {\n",
       "    var width = fig.canvas.width / fig.ratio;\n",
       "    fig.cell_info[0].output_area.element.off(\n",
       "        'cleared',\n",
       "        fig._remove_fig_handler\n",
       "    );\n",
       "\n",
       "    // Update the output cell to use the data from the current canvas.\n",
       "    fig.push_to_output();\n",
       "    var dataURL = fig.canvas.toDataURL();\n",
       "    // Re-enable the keyboard manager in IPython - without this line, in FF,\n",
       "    // the notebook keyboard shortcuts fail.\n",
       "    IPython.keyboard_manager.enable();\n",
       "    fig.parent_element.innerHTML =\n",
       "        '<img src=\"' + dataURL + '\" width=\"' + width + '\">';\n",
       "    fig.close_ws(fig, msg);\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.close_ws = function (fig, msg) {\n",
       "    fig.send_message('closing', msg);\n",
       "    // fig.ws.close()\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.push_to_output = function (_remove_interactive) {\n",
       "    // Turn the data on the canvas into data in the output cell.\n",
       "    var width = this.canvas.width / this.ratio;\n",
       "    var dataURL = this.canvas.toDataURL();\n",
       "    this.cell_info[1]['text/html'] =\n",
       "        '<img src=\"' + dataURL + '\" width=\"' + width + '\">';\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.updated_canvas_event = function () {\n",
       "    // Tell IPython that the notebook contents must change.\n",
       "    IPython.notebook.set_dirty(true);\n",
       "    this.send_message('ack', {});\n",
       "    var fig = this;\n",
       "    // Wait a second, then push the new image to the DOM so\n",
       "    // that it is saved nicely (might be nice to debounce this).\n",
       "    setTimeout(function () {\n",
       "        fig.push_to_output();\n",
       "    }, 1000);\n",
       "};\n",
       "\n",
       "mpl.figure.prototype._init_toolbar = function () {\n",
       "    var fig = this;\n",
       "\n",
       "    var toolbar = document.createElement('div');\n",
       "    toolbar.classList = 'btn-toolbar';\n",
       "    this.root.appendChild(toolbar);\n",
       "\n",
       "    function on_click_closure(name) {\n",
       "        return function (_event) {\n",
       "            return fig.toolbar_button_onclick(name);\n",
       "        };\n",
       "    }\n",
       "\n",
       "    function on_mouseover_closure(tooltip) {\n",
       "        return function (event) {\n",
       "            if (!event.currentTarget.disabled) {\n",
       "                return fig.toolbar_button_onmouseover(tooltip);\n",
       "            }\n",
       "        };\n",
       "    }\n",
       "\n",
       "    fig.buttons = {};\n",
       "    var buttonGroup = document.createElement('div');\n",
       "    buttonGroup.classList = 'btn-group';\n",
       "    var button;\n",
       "    for (var toolbar_ind in mpl.toolbar_items) {\n",
       "        var name = mpl.toolbar_items[toolbar_ind][0];\n",
       "        var tooltip = mpl.toolbar_items[toolbar_ind][1];\n",
       "        var image = mpl.toolbar_items[toolbar_ind][2];\n",
       "        var method_name = mpl.toolbar_items[toolbar_ind][3];\n",
       "\n",
       "        if (!name) {\n",
       "            /* Instead of a spacer, we start a new button group. */\n",
       "            if (buttonGroup.hasChildNodes()) {\n",
       "                toolbar.appendChild(buttonGroup);\n",
       "            }\n",
       "            buttonGroup = document.createElement('div');\n",
       "            buttonGroup.classList = 'btn-group';\n",
       "            continue;\n",
       "        }\n",
       "\n",
       "        button = fig.buttons[name] = document.createElement('button');\n",
       "        button.classList = 'btn btn-default';\n",
       "        button.href = '#';\n",
       "        button.title = name;\n",
       "        button.innerHTML = '<i class=\"fa ' + image + ' fa-lg\"></i>';\n",
       "        button.addEventListener('click', on_click_closure(method_name));\n",
       "        button.addEventListener('mouseover', on_mouseover_closure(tooltip));\n",
       "        buttonGroup.appendChild(button);\n",
       "    }\n",
       "\n",
       "    if (buttonGroup.hasChildNodes()) {\n",
       "        toolbar.appendChild(buttonGroup);\n",
       "    }\n",
       "\n",
       "    // Add the status bar.\n",
       "    var status_bar = document.createElement('span');\n",
       "    status_bar.classList = 'mpl-message pull-right';\n",
       "    toolbar.appendChild(status_bar);\n",
       "    this.message = status_bar;\n",
       "\n",
       "    // Add the close button to the window.\n",
       "    var buttongrp = document.createElement('div');\n",
       "    buttongrp.classList = 'btn-group inline pull-right';\n",
       "    button = document.createElement('button');\n",
       "    button.classList = 'btn btn-mini btn-primary';\n",
       "    button.href = '#';\n",
       "    button.title = 'Stop Interaction';\n",
       "    button.innerHTML = '<i class=\"fa fa-power-off icon-remove icon-large\"></i>';\n",
       "    button.addEventListener('click', function (_evt) {\n",
       "        fig.handle_close(fig, {});\n",
       "    });\n",
       "    button.addEventListener(\n",
       "        'mouseover',\n",
       "        on_mouseover_closure('Stop Interaction')\n",
       "    );\n",
       "    buttongrp.appendChild(button);\n",
       "    var titlebar = this.root.querySelector('.ui-dialog-titlebar');\n",
       "    titlebar.insertBefore(buttongrp, titlebar.firstChild);\n",
       "};\n",
       "\n",
       "mpl.figure.prototype._remove_fig_handler = function (event) {\n",
       "    var fig = event.data.fig;\n",
       "    fig.close_ws(fig, {});\n",
       "};\n",
       "\n",
       "mpl.figure.prototype._root_extra_style = function (el) {\n",
       "    el.style.boxSizing = 'content-box'; // override notebook setting of border-box.\n",
       "};\n",
       "\n",
       "mpl.figure.prototype._canvas_extra_style = function (el) {\n",
       "    // this is important to make the div 'focusable\n",
       "    el.setAttribute('tabindex', 0);\n",
       "    // reach out to IPython and tell the keyboard manager to turn it's self\n",
       "    // off when our div gets focus\n",
       "\n",
       "    // location in version 3\n",
       "    if (IPython.notebook.keyboard_manager) {\n",
       "        IPython.notebook.keyboard_manager.register_events(el);\n",
       "    } else {\n",
       "        // location in version 2\n",
       "        IPython.keyboard_manager.register_events(el);\n",
       "    }\n",
       "};\n",
       "\n",
       "mpl.figure.prototype._key_event_extra = function (event, _name) {\n",
       "    var manager = IPython.notebook.keyboard_manager;\n",
       "    if (!manager) {\n",
       "        manager = IPython.keyboard_manager;\n",
       "    }\n",
       "\n",
       "    // Check for shift+enter\n",
       "    if (event.shiftKey && event.which === 13) {\n",
       "        this.canvas_div.blur();\n",
       "        // select the cell after this one\n",
       "        var index = IPython.notebook.find_cell_index(this.cell_info[0]);\n",
       "        IPython.notebook.select(index + 1);\n",
       "    }\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.handle_save = function (fig, _msg) {\n",
       "    fig.ondownload(fig, null);\n",
       "};\n",
       "\n",
       "mpl.find_output_cell = function (html_output) {\n",
       "    // Return the cell and output element which can be found *uniquely* in the notebook.\n",
       "    // Note - this is a bit hacky, but it is done because the \"notebook_saving.Notebook\"\n",
       "    // IPython event is triggered only after the cells have been serialised, which for\n",
       "    // our purposes (turning an active figure into a static one), is too late.\n",
       "    var cells = IPython.notebook.get_cells();\n",
       "    var ncells = cells.length;\n",
       "    for (var i = 0; i < ncells; i++) {\n",
       "        var cell = cells[i];\n",
       "        if (cell.cell_type === 'code') {\n",
       "            for (var j = 0; j < cell.output_area.outputs.length; j++) {\n",
       "                var data = cell.output_area.outputs[j];\n",
       "                if (data.data) {\n",
       "                    // IPython >= 3 moved mimebundle to data attribute of output\n",
       "                    data = data.data;\n",
       "                }\n",
       "                if (data['text/html'] === html_output) {\n",
       "                    return [cell, data, j];\n",
       "                }\n",
       "            }\n",
       "        }\n",
       "    }\n",
       "};\n",
       "\n",
       "// Register the function which deals with the matplotlib target/channel.\n",
       "// The kernel may be null if the page has been refreshed.\n",
       "if (IPython.notebook.kernel !== null) {\n",
       "    IPython.notebook.kernel.comm_manager.register_target(\n",
       "        'matplotlib',\n",
       "        mpl.mpl_figure_comm\n",
       "    );\n",
       "}\n"
      ],
      "text/plain": [
       "<IPython.core.display.Javascript object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/html": [
       "<img src=\"data:image/png;base64,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\" width=\"900\">"
      ],
      "text/plain": [
       "<IPython.core.display.HTML object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "price_quantile()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "07baaac1",
   "metadata": {},
   "source": [
    "## RMSE in different price range"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 67,
   "id": "c1760f25",
   "metadata": {},
   "outputs": [],
   "source": [
    "# compute rmse in given price range\n",
    "def rmse_in_one_price_range(left_quantile=0, right_quantile=1):\n",
    "    xy_test = np.column_stack((X_test, y_test))\n",
    "    xy_test_price = xy_test[df.quantile(left_quantile).price <= xy_test[:,-1]]\n",
    "    xy_test_price = xy_test_price[xy_test_price[:,-1] < df.quantile(right_quantile).price]\n",
    "    \n",
    "#     xy_tree_test = np.column_stack((X_tree_test, y_tree_test))\n",
    "#     xy_tree_test_price = xy_tree_test[df.quantile(left_quantile).price <= xy_tree_test[:,-1]]\n",
    "#     xy_tree_test_price = xy_tree_test_price[xy_tree_test_price[:,-1] < df.quantile(right_quantile).price]\n",
    "    \n",
    "    def compute_rmse(X, y):\n",
    "        bl_predict = np.squeeze(baseline_model.predict(X))\n",
    "        bl_rmse = mean_squared_error(y, bl_predict, squared=False)\n",
    "\n",
    "        eb_predict = np.squeeze(emb_model.predict(X))\n",
    "        eb_rmse = mean_squared_error(y, eb_predict, squared=False)\n",
    "\n",
    "        rf_predict = rf_regressor.predict(X)\n",
    "        rf_rmse = mean_squared_error(y, rf_predict, squared=False)\n",
    "\n",
    "        gb_predict = gb_regressor.predict(X)\n",
    "        gb_rmse = mean_squared_error(y, gb_predict, squared=False)\n",
    "        \n",
    "        wnd_predict = np.squeeze(wnd_model.predict(X))\n",
    "        wnd_rmse = mean_squared_error(y, wnd_predict, squared=False)\n",
    "\n",
    "        ensemble_predict = 0.2 * gb_predict + 0.2 * rf_predict + 0.3 * bl_predict + 0.3 * eb_predict\n",
    "        ensemble_rmse = mean_squared_error(y, ensemble_predict, squared=False)\n",
    "        \n",
    "        return bl_rmse, eb_rmse, rf_rmse, gb_rmse, wnd_rmse, ensemble_rmse\n",
    "    \n",
    "    return compute_rmse(xy_test_price[:,:-1], xy_test_price[:,-1])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 74,
   "id": "e3c11d19",
   "metadata": {},
   "outputs": [],
   "source": [
    "def rmse_in_different_price_range():\n",
    "    bl_rmse_lst = []\n",
    "    eb_rmse_lst = []\n",
    "    rf_rmse_lst = []\n",
    "    gb_rmse_lst = []\n",
    "    wnd_rmse_lst = []\n",
    "    ensemble_rmse_lst = []\n",
    "    for i in range(10):\n",
    "        bl_rmse, eb_rmse, rf_rmse, gb_rmse, wnd_rmse, ensemble_rmse = rmse_in_one_price_range(i*0.1, (i+1)*0.1)\n",
    "        bl_rmse_lst.append(bl_rmse)\n",
    "        eb_rmse_lst.append(eb_rmse)\n",
    "        rf_rmse_lst.append(rf_rmse)\n",
    "        gb_rmse_lst.append(gb_rmse)\n",
    "        wnd_rmse_lst.append(wnd_rmse)\n",
    "        ensemble_rmse_lst.append(ensemble_rmse)\n",
    "    x = [i*0.1 for i in range(10)]\n",
    "    \n",
    "    \n",
    "    plt.figure(figsize=(9,6))\n",
    "    plt.plot(x, bl_rmse_lst, c=\"r\", label=\"BaselineModel_RMSE\")\n",
    "    plt.plot(x, bl_rmse_lst, c=\"pink\", label=\"EmbeddedModel_RMSE\")\n",
    "    plt.plot(x, rf_rmse_lst, c=\"y\", label=\"RandomForest_RMSE\")\n",
    "    plt.plot(x, gb_rmse_lst, c=\"b\", label=\"GradientBoosting_RMSE\")\n",
    "    plt.plot(x,wnd_rmse_lst, c=\"grey\", label=\"Wide&Deep_RMSE\")\n",
    "    plt.plot(x, ensemble_rmse_lst, c=\"black\", label=\"ensemble_RMSE\")\n",
    "    \n",
    "    plt.xlabel(\"Price Range(in quantile)\")\n",
    "    plt.ylabel(\"RMSE\")\n",
    "    \n",
    "    plt.xticks(x, [f\"[{round(i, 1)}, {round(i+0.1, 1)})\" for i in x])\n",
    "    plt.legend()\n",
    "    plt.show()\n",
    "    plt.savefig(\"x3 RMSE in different price ranges.png\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 75,
   "id": "c80f45cf",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "application/javascript": [
       "/* Put everything inside the global mpl namespace */\n",
       "/* global mpl */\n",
       "window.mpl = {};\n",
       "\n",
       "mpl.get_websocket_type = function () {\n",
       "    if (typeof WebSocket !== 'undefined') {\n",
       "        return WebSocket;\n",
       "    } else if (typeof MozWebSocket !== 'undefined') {\n",
       "        return MozWebSocket;\n",
       "    } else {\n",
       "        alert(\n",
       "            'Your browser does not have WebSocket support. ' +\n",
       "                'Please try Chrome, Safari or Firefox ≥ 6. ' +\n",
       "                'Firefox 4 and 5 are also supported but you ' +\n",
       "                'have to enable WebSockets in about:config.'\n",
       "        );\n",
       "    }\n",
       "};\n",
       "\n",
       "mpl.figure = function (figure_id, websocket, ondownload, parent_element) {\n",
       "    this.id = figure_id;\n",
       "\n",
       "    this.ws = websocket;\n",
       "\n",
       "    this.supports_binary = this.ws.binaryType !== undefined;\n",
       "\n",
       "    if (!this.supports_binary) {\n",
       "        var warnings = document.getElementById('mpl-warnings');\n",
       "        if (warnings) {\n",
       "            warnings.style.display = 'block';\n",
       "            warnings.textContent =\n",
       "                'This browser does not support binary websocket messages. ' +\n",
       "                'Performance may be slow.';\n",
       "        }\n",
       "    }\n",
       "\n",
       "    this.imageObj = new Image();\n",
       "\n",
       "    this.context = undefined;\n",
       "    this.message = undefined;\n",
       "    this.canvas = undefined;\n",
       "    this.rubberband_canvas = undefined;\n",
       "    this.rubberband_context = undefined;\n",
       "    this.format_dropdown = undefined;\n",
       "\n",
       "    this.image_mode = 'full';\n",
       "\n",
       "    this.root = document.createElement('div');\n",
       "    this.root.setAttribute('style', 'display: inline-block');\n",
       "    this._root_extra_style(this.root);\n",
       "\n",
       "    parent_element.appendChild(this.root);\n",
       "\n",
       "    this._init_header(this);\n",
       "    this._init_canvas(this);\n",
       "    this._init_toolbar(this);\n",
       "\n",
       "    var fig = this;\n",
       "\n",
       "    this.waiting = false;\n",
       "\n",
       "    this.ws.onopen = function () {\n",
       "        fig.send_message('supports_binary', { value: fig.supports_binary });\n",
       "        fig.send_message('send_image_mode', {});\n",
       "        if (fig.ratio !== 1) {\n",
       "            fig.send_message('set_dpi_ratio', { dpi_ratio: fig.ratio });\n",
       "        }\n",
       "        fig.send_message('refresh', {});\n",
       "    };\n",
       "\n",
       "    this.imageObj.onload = function () {\n",
       "        if (fig.image_mode === 'full') {\n",
       "            // Full images could contain transparency (where diff images\n",
       "            // almost always do), so we need to clear the canvas so that\n",
       "            // there is no ghosting.\n",
       "            fig.context.clearRect(0, 0, fig.canvas.width, fig.canvas.height);\n",
       "        }\n",
       "        fig.context.drawImage(fig.imageObj, 0, 0);\n",
       "    };\n",
       "\n",
       "    this.imageObj.onunload = function () {\n",
       "        fig.ws.close();\n",
       "    };\n",
       "\n",
       "    this.ws.onmessage = this._make_on_message_function(this);\n",
       "\n",
       "    this.ondownload = ondownload;\n",
       "};\n",
       "\n",
       "mpl.figure.prototype._init_header = function () {\n",
       "    var titlebar = document.createElement('div');\n",
       "    titlebar.classList =\n",
       "        'ui-dialog-titlebar ui-widget-header ui-corner-all ui-helper-clearfix';\n",
       "    var titletext = document.createElement('div');\n",
       "    titletext.classList = 'ui-dialog-title';\n",
       "    titletext.setAttribute(\n",
       "        'style',\n",
       "        'width: 100%; text-align: center; padding: 3px;'\n",
       "    );\n",
       "    titlebar.appendChild(titletext);\n",
       "    this.root.appendChild(titlebar);\n",
       "    this.header = titletext;\n",
       "};\n",
       "\n",
       "mpl.figure.prototype._canvas_extra_style = function (_canvas_div) {};\n",
       "\n",
       "mpl.figure.prototype._root_extra_style = function (_canvas_div) {};\n",
       "\n",
       "mpl.figure.prototype._init_canvas = function () {\n",
       "    var fig = this;\n",
       "\n",
       "    var canvas_div = (this.canvas_div = document.createElement('div'));\n",
       "    canvas_div.setAttribute(\n",
       "        'style',\n",
       "        'border: 1px solid #ddd;' +\n",
       "            'box-sizing: content-box;' +\n",
       "            'clear: both;' +\n",
       "            'min-height: 1px;' +\n",
       "            'min-width: 1px;' +\n",
       "            'outline: 0;' +\n",
       "            'overflow: hidden;' +\n",
       "            'position: relative;' +\n",
       "            'resize: both;'\n",
       "    );\n",
       "\n",
       "    function on_keyboard_event_closure(name) {\n",
       "        return function (event) {\n",
       "            return fig.key_event(event, name);\n",
       "        };\n",
       "    }\n",
       "\n",
       "    canvas_div.addEventListener(\n",
       "        'keydown',\n",
       "        on_keyboard_event_closure('key_press')\n",
       "    );\n",
       "    canvas_div.addEventListener(\n",
       "        'keyup',\n",
       "        on_keyboard_event_closure('key_release')\n",
       "    );\n",
       "\n",
       "    this._canvas_extra_style(canvas_div);\n",
       "    this.root.appendChild(canvas_div);\n",
       "\n",
       "    var canvas = (this.canvas = document.createElement('canvas'));\n",
       "    canvas.classList.add('mpl-canvas');\n",
       "    canvas.setAttribute('style', 'box-sizing: content-box;');\n",
       "\n",
       "    this.context = canvas.getContext('2d');\n",
       "\n",
       "    var backingStore =\n",
       "        this.context.backingStorePixelRatio ||\n",
       "        this.context.webkitBackingStorePixelRatio ||\n",
       "        this.context.mozBackingStorePixelRatio ||\n",
       "        this.context.msBackingStorePixelRatio ||\n",
       "        this.context.oBackingStorePixelRatio ||\n",
       "        this.context.backingStorePixelRatio ||\n",
       "        1;\n",
       "\n",
       "    this.ratio = (window.devicePixelRatio || 1) / backingStore;\n",
       "    if (this.ratio !== 1) {\n",
       "        fig.send_message('set_dpi_ratio', { dpi_ratio: this.ratio });\n",
       "    }\n",
       "\n",
       "    var rubberband_canvas = (this.rubberband_canvas = document.createElement(\n",
       "        'canvas'\n",
       "    ));\n",
       "    rubberband_canvas.setAttribute(\n",
       "        'style',\n",
       "        'box-sizing: content-box; position: absolute; left: 0; top: 0; z-index: 1;'\n",
       "    );\n",
       "\n",
       "    var resizeObserver = new ResizeObserver(function (entries) {\n",
       "        var nentries = entries.length;\n",
       "        for (var i = 0; i < nentries; i++) {\n",
       "            var entry = entries[i];\n",
       "            var width, height;\n",
       "            if (entry.contentBoxSize) {\n",
       "                if (entry.contentBoxSize instanceof Array) {\n",
       "                    // Chrome 84 implements new version of spec.\n",
       "                    width = entry.contentBoxSize[0].inlineSize;\n",
       "                    height = entry.contentBoxSize[0].blockSize;\n",
       "                } else {\n",
       "                    // Firefox implements old version of spec.\n",
       "                    width = entry.contentBoxSize.inlineSize;\n",
       "                    height = entry.contentBoxSize.blockSize;\n",
       "                }\n",
       "            } else {\n",
       "                // Chrome <84 implements even older version of spec.\n",
       "                width = entry.contentRect.width;\n",
       "                height = entry.contentRect.height;\n",
       "            }\n",
       "\n",
       "            // Keep the size of the canvas and rubber band canvas in sync with\n",
       "            // the canvas container.\n",
       "            if (entry.devicePixelContentBoxSize) {\n",
       "                // Chrome 84 implements new version of spec.\n",
       "                canvas.setAttribute(\n",
       "                    'width',\n",
       "                    entry.devicePixelContentBoxSize[0].inlineSize\n",
       "                );\n",
       "                canvas.setAttribute(\n",
       "                    'height',\n",
       "                    entry.devicePixelContentBoxSize[0].blockSize\n",
       "                );\n",
       "            } else {\n",
       "                canvas.setAttribute('width', width * fig.ratio);\n",
       "                canvas.setAttribute('height', height * fig.ratio);\n",
       "            }\n",
       "            canvas.setAttribute(\n",
       "                'style',\n",
       "                'width: ' + width + 'px; height: ' + height + 'px;'\n",
       "            );\n",
       "\n",
       "            rubberband_canvas.setAttribute('width', width);\n",
       "            rubberband_canvas.setAttribute('height', height);\n",
       "\n",
       "            // And update the size in Python. We ignore the initial 0/0 size\n",
       "            // that occurs as the element is placed into the DOM, which should\n",
       "            // otherwise not happen due to the minimum size styling.\n",
       "            if (width != 0 && height != 0) {\n",
       "                fig.request_resize(width, height);\n",
       "            }\n",
       "        }\n",
       "    });\n",
       "    resizeObserver.observe(canvas_div);\n",
       "\n",
       "    function on_mouse_event_closure(name) {\n",
       "        return function (event) {\n",
       "            return fig.mouse_event(event, name);\n",
       "        };\n",
       "    }\n",
       "\n",
       "    rubberband_canvas.addEventListener(\n",
       "        'mousedown',\n",
       "        on_mouse_event_closure('button_press')\n",
       "    );\n",
       "    rubberband_canvas.addEventListener(\n",
       "        'mouseup',\n",
       "        on_mouse_event_closure('button_release')\n",
       "    );\n",
       "    // Throttle sequential mouse events to 1 every 20ms.\n",
       "    rubberband_canvas.addEventListener(\n",
       "        'mousemove',\n",
       "        on_mouse_event_closure('motion_notify')\n",
       "    );\n",
       "\n",
       "    rubberband_canvas.addEventListener(\n",
       "        'mouseenter',\n",
       "        on_mouse_event_closure('figure_enter')\n",
       "    );\n",
       "    rubberband_canvas.addEventListener(\n",
       "        'mouseleave',\n",
       "        on_mouse_event_closure('figure_leave')\n",
       "    );\n",
       "\n",
       "    canvas_div.addEventListener('wheel', function (event) {\n",
       "        if (event.deltaY < 0) {\n",
       "            event.step = 1;\n",
       "        } else {\n",
       "            event.step = -1;\n",
       "        }\n",
       "        on_mouse_event_closure('scroll')(event);\n",
       "    });\n",
       "\n",
       "    canvas_div.appendChild(canvas);\n",
       "    canvas_div.appendChild(rubberband_canvas);\n",
       "\n",
       "    this.rubberband_context = rubberband_canvas.getContext('2d');\n",
       "    this.rubberband_context.strokeStyle = '#000000';\n",
       "\n",
       "    this._resize_canvas = function (width, height, forward) {\n",
       "        if (forward) {\n",
       "            canvas_div.style.width = width + 'px';\n",
       "            canvas_div.style.height = height + 'px';\n",
       "        }\n",
       "    };\n",
       "\n",
       "    // Disable right mouse context menu.\n",
       "    this.rubberband_canvas.addEventListener('contextmenu', function (_e) {\n",
       "        event.preventDefault();\n",
       "        return false;\n",
       "    });\n",
       "\n",
       "    function set_focus() {\n",
       "        canvas.focus();\n",
       "        canvas_div.focus();\n",
       "    }\n",
       "\n",
       "    window.setTimeout(set_focus, 100);\n",
       "};\n",
       "\n",
       "mpl.figure.prototype._init_toolbar = function () {\n",
       "    var fig = this;\n",
       "\n",
       "    var toolbar = document.createElement('div');\n",
       "    toolbar.classList = 'mpl-toolbar';\n",
       "    this.root.appendChild(toolbar);\n",
       "\n",
       "    function on_click_closure(name) {\n",
       "        return function (_event) {\n",
       "            return fig.toolbar_button_onclick(name);\n",
       "        };\n",
       "    }\n",
       "\n",
       "    function on_mouseover_closure(tooltip) {\n",
       "        return function (event) {\n",
       "            if (!event.currentTarget.disabled) {\n",
       "                return fig.toolbar_button_onmouseover(tooltip);\n",
       "            }\n",
       "        };\n",
       "    }\n",
       "\n",
       "    fig.buttons = {};\n",
       "    var buttonGroup = document.createElement('div');\n",
       "    buttonGroup.classList = 'mpl-button-group';\n",
       "    for (var toolbar_ind in mpl.toolbar_items) {\n",
       "        var name = mpl.toolbar_items[toolbar_ind][0];\n",
       "        var tooltip = mpl.toolbar_items[toolbar_ind][1];\n",
       "        var image = mpl.toolbar_items[toolbar_ind][2];\n",
       "        var method_name = mpl.toolbar_items[toolbar_ind][3];\n",
       "\n",
       "        if (!name) {\n",
       "            /* Instead of a spacer, we start a new button group. */\n",
       "            if (buttonGroup.hasChildNodes()) {\n",
       "                toolbar.appendChild(buttonGroup);\n",
       "            }\n",
       "            buttonGroup = document.createElement('div');\n",
       "            buttonGroup.classList = 'mpl-button-group';\n",
       "            continue;\n",
       "        }\n",
       "\n",
       "        var button = (fig.buttons[name] = document.createElement('button'));\n",
       "        button.classList = 'mpl-widget';\n",
       "        button.setAttribute('role', 'button');\n",
       "        button.setAttribute('aria-disabled', 'false');\n",
       "        button.addEventListener('click', on_click_closure(method_name));\n",
       "        button.addEventListener('mouseover', on_mouseover_closure(tooltip));\n",
       "\n",
       "        var icon_img = document.createElement('img');\n",
       "        icon_img.src = '_images/' + image + '.png';\n",
       "        icon_img.srcset = '_images/' + image + '_large.png 2x';\n",
       "        icon_img.alt = tooltip;\n",
       "        button.appendChild(icon_img);\n",
       "\n",
       "        buttonGroup.appendChild(button);\n",
       "    }\n",
       "\n",
       "    if (buttonGroup.hasChildNodes()) {\n",
       "        toolbar.appendChild(buttonGroup);\n",
       "    }\n",
       "\n",
       "    var fmt_picker = document.createElement('select');\n",
       "    fmt_picker.classList = 'mpl-widget';\n",
       "    toolbar.appendChild(fmt_picker);\n",
       "    this.format_dropdown = fmt_picker;\n",
       "\n",
       "    for (var ind in mpl.extensions) {\n",
       "        var fmt = mpl.extensions[ind];\n",
       "        var option = document.createElement('option');\n",
       "        option.selected = fmt === mpl.default_extension;\n",
       "        option.innerHTML = fmt;\n",
       "        fmt_picker.appendChild(option);\n",
       "    }\n",
       "\n",
       "    var status_bar = document.createElement('span');\n",
       "    status_bar.classList = 'mpl-message';\n",
       "    toolbar.appendChild(status_bar);\n",
       "    this.message = status_bar;\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.request_resize = function (x_pixels, y_pixels) {\n",
       "    // Request matplotlib to resize the figure. Matplotlib will then trigger a resize in the client,\n",
       "    // which will in turn request a refresh of the image.\n",
       "    this.send_message('resize', { width: x_pixels, height: y_pixels });\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.send_message = function (type, properties) {\n",
       "    properties['type'] = type;\n",
       "    properties['figure_id'] = this.id;\n",
       "    this.ws.send(JSON.stringify(properties));\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.send_draw_message = function () {\n",
       "    if (!this.waiting) {\n",
       "        this.waiting = true;\n",
       "        this.ws.send(JSON.stringify({ type: 'draw', figure_id: this.id }));\n",
       "    }\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.handle_save = function (fig, _msg) {\n",
       "    var format_dropdown = fig.format_dropdown;\n",
       "    var format = format_dropdown.options[format_dropdown.selectedIndex].value;\n",
       "    fig.ondownload(fig, format);\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.handle_resize = function (fig, msg) {\n",
       "    var size = msg['size'];\n",
       "    if (size[0] !== fig.canvas.width || size[1] !== fig.canvas.height) {\n",
       "        fig._resize_canvas(size[0], size[1], msg['forward']);\n",
       "        fig.send_message('refresh', {});\n",
       "    }\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.handle_rubberband = function (fig, msg) {\n",
       "    var x0 = msg['x0'] / fig.ratio;\n",
       "    var y0 = (fig.canvas.height - msg['y0']) / fig.ratio;\n",
       "    var x1 = msg['x1'] / fig.ratio;\n",
       "    var y1 = (fig.canvas.height - msg['y1']) / fig.ratio;\n",
       "    x0 = Math.floor(x0) + 0.5;\n",
       "    y0 = Math.floor(y0) + 0.5;\n",
       "    x1 = Math.floor(x1) + 0.5;\n",
       "    y1 = Math.floor(y1) + 0.5;\n",
       "    var min_x = Math.min(x0, x1);\n",
       "    var min_y = Math.min(y0, y1);\n",
       "    var width = Math.abs(x1 - x0);\n",
       "    var height = Math.abs(y1 - y0);\n",
       "\n",
       "    fig.rubberband_context.clearRect(\n",
       "        0,\n",
       "        0,\n",
       "        fig.canvas.width / fig.ratio,\n",
       "        fig.canvas.height / fig.ratio\n",
       "    );\n",
       "\n",
       "    fig.rubberband_context.strokeRect(min_x, min_y, width, height);\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.handle_figure_label = function (fig, msg) {\n",
       "    // Updates the figure title.\n",
       "    fig.header.textContent = msg['label'];\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.handle_cursor = function (fig, msg) {\n",
       "    var cursor = msg['cursor'];\n",
       "    switch (cursor) {\n",
       "        case 0:\n",
       "            cursor = 'pointer';\n",
       "            break;\n",
       "        case 1:\n",
       "            cursor = 'default';\n",
       "            break;\n",
       "        case 2:\n",
       "            cursor = 'crosshair';\n",
       "            break;\n",
       "        case 3:\n",
       "            cursor = 'move';\n",
       "            break;\n",
       "    }\n",
       "    fig.rubberband_canvas.style.cursor = cursor;\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.handle_message = function (fig, msg) {\n",
       "    fig.message.textContent = msg['message'];\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.handle_draw = function (fig, _msg) {\n",
       "    // Request the server to send over a new figure.\n",
       "    fig.send_draw_message();\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.handle_image_mode = function (fig, msg) {\n",
       "    fig.image_mode = msg['mode'];\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.handle_history_buttons = function (fig, msg) {\n",
       "    for (var key in msg) {\n",
       "        if (!(key in fig.buttons)) {\n",
       "            continue;\n",
       "        }\n",
       "        fig.buttons[key].disabled = !msg[key];\n",
       "        fig.buttons[key].setAttribute('aria-disabled', !msg[key]);\n",
       "    }\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.handle_navigate_mode = function (fig, msg) {\n",
       "    if (msg['mode'] === 'PAN') {\n",
       "        fig.buttons['Pan'].classList.add('active');\n",
       "        fig.buttons['Zoom'].classList.remove('active');\n",
       "    } else if (msg['mode'] === 'ZOOM') {\n",
       "        fig.buttons['Pan'].classList.remove('active');\n",
       "        fig.buttons['Zoom'].classList.add('active');\n",
       "    } else {\n",
       "        fig.buttons['Pan'].classList.remove('active');\n",
       "        fig.buttons['Zoom'].classList.remove('active');\n",
       "    }\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.updated_canvas_event = function () {\n",
       "    // Called whenever the canvas gets updated.\n",
       "    this.send_message('ack', {});\n",
       "};\n",
       "\n",
       "// A function to construct a web socket function for onmessage handling.\n",
       "// Called in the figure constructor.\n",
       "mpl.figure.prototype._make_on_message_function = function (fig) {\n",
       "    return function socket_on_message(evt) {\n",
       "        if (evt.data instanceof Blob) {\n",
       "            /* FIXME: We get \"Resource interpreted as Image but\n",
       "             * transferred with MIME type text/plain:\" errors on\n",
       "             * Chrome.  But how to set the MIME type?  It doesn't seem\n",
       "             * to be part of the websocket stream */\n",
       "            evt.data.type = 'image/png';\n",
       "\n",
       "            /* Free the memory for the previous frames */\n",
       "            if (fig.imageObj.src) {\n",
       "                (window.URL || window.webkitURL).revokeObjectURL(\n",
       "                    fig.imageObj.src\n",
       "                );\n",
       "            }\n",
       "\n",
       "            fig.imageObj.src = (window.URL || window.webkitURL).createObjectURL(\n",
       "                evt.data\n",
       "            );\n",
       "            fig.updated_canvas_event();\n",
       "            fig.waiting = false;\n",
       "            return;\n",
       "        } else if (\n",
       "            typeof evt.data === 'string' &&\n",
       "            evt.data.slice(0, 21) === 'data:image/png;base64'\n",
       "        ) {\n",
       "            fig.imageObj.src = evt.data;\n",
       "            fig.updated_canvas_event();\n",
       "            fig.waiting = false;\n",
       "            return;\n",
       "        }\n",
       "\n",
       "        var msg = JSON.parse(evt.data);\n",
       "        var msg_type = msg['type'];\n",
       "\n",
       "        // Call the  \"handle_{type}\" callback, which takes\n",
       "        // the figure and JSON message as its only arguments.\n",
       "        try {\n",
       "            var callback = fig['handle_' + msg_type];\n",
       "        } catch (e) {\n",
       "            console.log(\n",
       "                \"No handler for the '\" + msg_type + \"' message type: \",\n",
       "                msg\n",
       "            );\n",
       "            return;\n",
       "        }\n",
       "\n",
       "        if (callback) {\n",
       "            try {\n",
       "                // console.log(\"Handling '\" + msg_type + \"' message: \", msg);\n",
       "                callback(fig, msg);\n",
       "            } catch (e) {\n",
       "                console.log(\n",
       "                    \"Exception inside the 'handler_\" + msg_type + \"' callback:\",\n",
       "                    e,\n",
       "                    e.stack,\n",
       "                    msg\n",
       "                );\n",
       "            }\n",
       "        }\n",
       "    };\n",
       "};\n",
       "\n",
       "// from http://stackoverflow.com/questions/1114465/getting-mouse-location-in-canvas\n",
       "mpl.findpos = function (e) {\n",
       "    //this section is from http://www.quirksmode.org/js/events_properties.html\n",
       "    var targ;\n",
       "    if (!e) {\n",
       "        e = window.event;\n",
       "    }\n",
       "    if (e.target) {\n",
       "        targ = e.target;\n",
       "    } else if (e.srcElement) {\n",
       "        targ = e.srcElement;\n",
       "    }\n",
       "    if (targ.nodeType === 3) {\n",
       "        // defeat Safari bug\n",
       "        targ = targ.parentNode;\n",
       "    }\n",
       "\n",
       "    // pageX,Y are the mouse positions relative to the document\n",
       "    var boundingRect = targ.getBoundingClientRect();\n",
       "    var x = e.pageX - (boundingRect.left + document.body.scrollLeft);\n",
       "    var y = e.pageY - (boundingRect.top + document.body.scrollTop);\n",
       "\n",
       "    return { x: x, y: y };\n",
       "};\n",
       "\n",
       "/*\n",
       " * return a copy of an object with only non-object keys\n",
       " * we need this to avoid circular references\n",
       " * http://stackoverflow.com/a/24161582/3208463\n",
       " */\n",
       "function simpleKeys(original) {\n",
       "    return Object.keys(original).reduce(function (obj, key) {\n",
       "        if (typeof original[key] !== 'object') {\n",
       "            obj[key] = original[key];\n",
       "        }\n",
       "        return obj;\n",
       "    }, {});\n",
       "}\n",
       "\n",
       "mpl.figure.prototype.mouse_event = function (event, name) {\n",
       "    var canvas_pos = mpl.findpos(event);\n",
       "\n",
       "    if (name === 'button_press') {\n",
       "        this.canvas.focus();\n",
       "        this.canvas_div.focus();\n",
       "    }\n",
       "\n",
       "    var x = canvas_pos.x * this.ratio;\n",
       "    var y = canvas_pos.y * this.ratio;\n",
       "\n",
       "    this.send_message(name, {\n",
       "        x: x,\n",
       "        y: y,\n",
       "        button: event.button,\n",
       "        step: event.step,\n",
       "        guiEvent: simpleKeys(event),\n",
       "    });\n",
       "\n",
       "    /* This prevents the web browser from automatically changing to\n",
       "     * the text insertion cursor when the button is pressed.  We want\n",
       "     * to control all of the cursor setting manually through the\n",
       "     * 'cursor' event from matplotlib */\n",
       "    event.preventDefault();\n",
       "    return false;\n",
       "};\n",
       "\n",
       "mpl.figure.prototype._key_event_extra = function (_event, _name) {\n",
       "    // Handle any extra behaviour associated with a key event\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.key_event = function (event, name) {\n",
       "    // Prevent repeat events\n",
       "    if (name === 'key_press') {\n",
       "        if (event.which === this._key) {\n",
       "            return;\n",
       "        } else {\n",
       "            this._key = event.which;\n",
       "        }\n",
       "    }\n",
       "    if (name === 'key_release') {\n",
       "        this._key = null;\n",
       "    }\n",
       "\n",
       "    var value = '';\n",
       "    if (event.ctrlKey && event.which !== 17) {\n",
       "        value += 'ctrl+';\n",
       "    }\n",
       "    if (event.altKey && event.which !== 18) {\n",
       "        value += 'alt+';\n",
       "    }\n",
       "    if (event.shiftKey && event.which !== 16) {\n",
       "        value += 'shift+';\n",
       "    }\n",
       "\n",
       "    value += 'k';\n",
       "    value += event.which.toString();\n",
       "\n",
       "    this._key_event_extra(event, name);\n",
       "\n",
       "    this.send_message(name, { key: value, guiEvent: simpleKeys(event) });\n",
       "    return false;\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.toolbar_button_onclick = function (name) {\n",
       "    if (name === 'download') {\n",
       "        this.handle_save(this, null);\n",
       "    } else {\n",
       "        this.send_message('toolbar_button', { name: name });\n",
       "    }\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.toolbar_button_onmouseover = function (tooltip) {\n",
       "    this.message.textContent = tooltip;\n",
       "};\n",
       "mpl.toolbar_items = [[\"Home\", \"Reset original view\", \"fa fa-home icon-home\", \"home\"], [\"Back\", \"Back to previous view\", \"fa fa-arrow-left icon-arrow-left\", \"back\"], [\"Forward\", \"Forward to next view\", \"fa fa-arrow-right icon-arrow-right\", \"forward\"], [\"\", \"\", \"\", \"\"], [\"Pan\", \"Left button pans, Right button zooms\\nx/y fixes axis, CTRL fixes aspect\", \"fa fa-arrows icon-move\", \"pan\"], [\"Zoom\", \"Zoom to rectangle\\nx/y fixes axis, CTRL fixes aspect\", \"fa fa-square-o icon-check-empty\", \"zoom\"], [\"\", \"\", \"\", \"\"], [\"Download\", \"Download plot\", \"fa fa-floppy-o icon-save\", \"download\"]];\n",
       "\n",
       "mpl.extensions = [\"eps\", \"jpeg\", \"pdf\", \"png\", \"ps\", \"raw\", \"svg\", \"tif\"];\n",
       "\n",
       "mpl.default_extension = \"png\";/* global mpl */\n",
       "\n",
       "var comm_websocket_adapter = function (comm) {\n",
       "    // Create a \"websocket\"-like object which calls the given IPython comm\n",
       "    // object with the appropriate methods. Currently this is a non binary\n",
       "    // socket, so there is still some room for performance tuning.\n",
       "    var ws = {};\n",
       "\n",
       "    ws.close = function () {\n",
       "        comm.close();\n",
       "    };\n",
       "    ws.send = function (m) {\n",
       "        //console.log('sending', m);\n",
       "        comm.send(m);\n",
       "    };\n",
       "    // Register the callback with on_msg.\n",
       "    comm.on_msg(function (msg) {\n",
       "        //console.log('receiving', msg['content']['data'], msg);\n",
       "        // Pass the mpl event to the overridden (by mpl) onmessage function.\n",
       "        ws.onmessage(msg['content']['data']);\n",
       "    });\n",
       "    return ws;\n",
       "};\n",
       "\n",
       "mpl.mpl_figure_comm = function (comm, msg) {\n",
       "    // This is the function which gets called when the mpl process\n",
       "    // starts-up an IPython Comm through the \"matplotlib\" channel.\n",
       "\n",
       "    var id = msg.content.data.id;\n",
       "    // Get hold of the div created by the display call when the Comm\n",
       "    // socket was opened in Python.\n",
       "    var element = document.getElementById(id);\n",
       "    var ws_proxy = comm_websocket_adapter(comm);\n",
       "\n",
       "    function ondownload(figure, _format) {\n",
       "        window.open(figure.canvas.toDataURL());\n",
       "    }\n",
       "\n",
       "    var fig = new mpl.figure(id, ws_proxy, ondownload, element);\n",
       "\n",
       "    // Call onopen now - mpl needs it, as it is assuming we've passed it a real\n",
       "    // web socket which is closed, not our websocket->open comm proxy.\n",
       "    ws_proxy.onopen();\n",
       "\n",
       "    fig.parent_element = element;\n",
       "    fig.cell_info = mpl.find_output_cell(\"<div id='\" + id + \"'></div>\");\n",
       "    if (!fig.cell_info) {\n",
       "        console.error('Failed to find cell for figure', id, fig);\n",
       "        return;\n",
       "    }\n",
       "    fig.cell_info[0].output_area.element.one(\n",
       "        'cleared',\n",
       "        { fig: fig },\n",
       "        fig._remove_fig_handler\n",
       "    );\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.handle_close = function (fig, msg) {\n",
       "    var width = fig.canvas.width / fig.ratio;\n",
       "    fig.cell_info[0].output_area.element.off(\n",
       "        'cleared',\n",
       "        fig._remove_fig_handler\n",
       "    );\n",
       "\n",
       "    // Update the output cell to use the data from the current canvas.\n",
       "    fig.push_to_output();\n",
       "    var dataURL = fig.canvas.toDataURL();\n",
       "    // Re-enable the keyboard manager in IPython - without this line, in FF,\n",
       "    // the notebook keyboard shortcuts fail.\n",
       "    IPython.keyboard_manager.enable();\n",
       "    fig.parent_element.innerHTML =\n",
       "        '<img src=\"' + dataURL + '\" width=\"' + width + '\">';\n",
       "    fig.close_ws(fig, msg);\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.close_ws = function (fig, msg) {\n",
       "    fig.send_message('closing', msg);\n",
       "    // fig.ws.close()\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.push_to_output = function (_remove_interactive) {\n",
       "    // Turn the data on the canvas into data in the output cell.\n",
       "    var width = this.canvas.width / this.ratio;\n",
       "    var dataURL = this.canvas.toDataURL();\n",
       "    this.cell_info[1]['text/html'] =\n",
       "        '<img src=\"' + dataURL + '\" width=\"' + width + '\">';\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.updated_canvas_event = function () {\n",
       "    // Tell IPython that the notebook contents must change.\n",
       "    IPython.notebook.set_dirty(true);\n",
       "    this.send_message('ack', {});\n",
       "    var fig = this;\n",
       "    // Wait a second, then push the new image to the DOM so\n",
       "    // that it is saved nicely (might be nice to debounce this).\n",
       "    setTimeout(function () {\n",
       "        fig.push_to_output();\n",
       "    }, 1000);\n",
       "};\n",
       "\n",
       "mpl.figure.prototype._init_toolbar = function () {\n",
       "    var fig = this;\n",
       "\n",
       "    var toolbar = document.createElement('div');\n",
       "    toolbar.classList = 'btn-toolbar';\n",
       "    this.root.appendChild(toolbar);\n",
       "\n",
       "    function on_click_closure(name) {\n",
       "        return function (_event) {\n",
       "            return fig.toolbar_button_onclick(name);\n",
       "        };\n",
       "    }\n",
       "\n",
       "    function on_mouseover_closure(tooltip) {\n",
       "        return function (event) {\n",
       "            if (!event.currentTarget.disabled) {\n",
       "                return fig.toolbar_button_onmouseover(tooltip);\n",
       "            }\n",
       "        };\n",
       "    }\n",
       "\n",
       "    fig.buttons = {};\n",
       "    var buttonGroup = document.createElement('div');\n",
       "    buttonGroup.classList = 'btn-group';\n",
       "    var button;\n",
       "    for (var toolbar_ind in mpl.toolbar_items) {\n",
       "        var name = mpl.toolbar_items[toolbar_ind][0];\n",
       "        var tooltip = mpl.toolbar_items[toolbar_ind][1];\n",
       "        var image = mpl.toolbar_items[toolbar_ind][2];\n",
       "        var method_name = mpl.toolbar_items[toolbar_ind][3];\n",
       "\n",
       "        if (!name) {\n",
       "            /* Instead of a spacer, we start a new button group. */\n",
       "            if (buttonGroup.hasChildNodes()) {\n",
       "                toolbar.appendChild(buttonGroup);\n",
       "            }\n",
       "            buttonGroup = document.createElement('div');\n",
       "            buttonGroup.classList = 'btn-group';\n",
       "            continue;\n",
       "        }\n",
       "\n",
       "        button = fig.buttons[name] = document.createElement('button');\n",
       "        button.classList = 'btn btn-default';\n",
       "        button.href = '#';\n",
       "        button.title = name;\n",
       "        button.innerHTML = '<i class=\"fa ' + image + ' fa-lg\"></i>';\n",
       "        button.addEventListener('click', on_click_closure(method_name));\n",
       "        button.addEventListener('mouseover', on_mouseover_closure(tooltip));\n",
       "        buttonGroup.appendChild(button);\n",
       "    }\n",
       "\n",
       "    if (buttonGroup.hasChildNodes()) {\n",
       "        toolbar.appendChild(buttonGroup);\n",
       "    }\n",
       "\n",
       "    // Add the status bar.\n",
       "    var status_bar = document.createElement('span');\n",
       "    status_bar.classList = 'mpl-message pull-right';\n",
       "    toolbar.appendChild(status_bar);\n",
       "    this.message = status_bar;\n",
       "\n",
       "    // Add the close button to the window.\n",
       "    var buttongrp = document.createElement('div');\n",
       "    buttongrp.classList = 'btn-group inline pull-right';\n",
       "    button = document.createElement('button');\n",
       "    button.classList = 'btn btn-mini btn-primary';\n",
       "    button.href = '#';\n",
       "    button.title = 'Stop Interaction';\n",
       "    button.innerHTML = '<i class=\"fa fa-power-off icon-remove icon-large\"></i>';\n",
       "    button.addEventListener('click', function (_evt) {\n",
       "        fig.handle_close(fig, {});\n",
       "    });\n",
       "    button.addEventListener(\n",
       "        'mouseover',\n",
       "        on_mouseover_closure('Stop Interaction')\n",
       "    );\n",
       "    buttongrp.appendChild(button);\n",
       "    var titlebar = this.root.querySelector('.ui-dialog-titlebar');\n",
       "    titlebar.insertBefore(buttongrp, titlebar.firstChild);\n",
       "};\n",
       "\n",
       "mpl.figure.prototype._remove_fig_handler = function (event) {\n",
       "    var fig = event.data.fig;\n",
       "    fig.close_ws(fig, {});\n",
       "};\n",
       "\n",
       "mpl.figure.prototype._root_extra_style = function (el) {\n",
       "    el.style.boxSizing = 'content-box'; // override notebook setting of border-box.\n",
       "};\n",
       "\n",
       "mpl.figure.prototype._canvas_extra_style = function (el) {\n",
       "    // this is important to make the div 'focusable\n",
       "    el.setAttribute('tabindex', 0);\n",
       "    // reach out to IPython and tell the keyboard manager to turn it's self\n",
       "    // off when our div gets focus\n",
       "\n",
       "    // location in version 3\n",
       "    if (IPython.notebook.keyboard_manager) {\n",
       "        IPython.notebook.keyboard_manager.register_events(el);\n",
       "    } else {\n",
       "        // location in version 2\n",
       "        IPython.keyboard_manager.register_events(el);\n",
       "    }\n",
       "};\n",
       "\n",
       "mpl.figure.prototype._key_event_extra = function (event, _name) {\n",
       "    var manager = IPython.notebook.keyboard_manager;\n",
       "    if (!manager) {\n",
       "        manager = IPython.keyboard_manager;\n",
       "    }\n",
       "\n",
       "    // Check for shift+enter\n",
       "    if (event.shiftKey && event.which === 13) {\n",
       "        this.canvas_div.blur();\n",
       "        // select the cell after this one\n",
       "        var index = IPython.notebook.find_cell_index(this.cell_info[0]);\n",
       "        IPython.notebook.select(index + 1);\n",
       "    }\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.handle_save = function (fig, _msg) {\n",
       "    fig.ondownload(fig, null);\n",
       "};\n",
       "\n",
       "mpl.find_output_cell = function (html_output) {\n",
       "    // Return the cell and output element which can be found *uniquely* in the notebook.\n",
       "    // Note - this is a bit hacky, but it is done because the \"notebook_saving.Notebook\"\n",
       "    // IPython event is triggered only after the cells have been serialised, which for\n",
       "    // our purposes (turning an active figure into a static one), is too late.\n",
       "    var cells = IPython.notebook.get_cells();\n",
       "    var ncells = cells.length;\n",
       "    for (var i = 0; i < ncells; i++) {\n",
       "        var cell = cells[i];\n",
       "        if (cell.cell_type === 'code') {\n",
       "            for (var j = 0; j < cell.output_area.outputs.length; j++) {\n",
       "                var data = cell.output_area.outputs[j];\n",
       "                if (data.data) {\n",
       "                    // IPython >= 3 moved mimebundle to data attribute of output\n",
       "                    data = data.data;\n",
       "                }\n",
       "                if (data['text/html'] === html_output) {\n",
       "                    return [cell, data, j];\n",
       "                }\n",
       "            }\n",
       "        }\n",
       "    }\n",
       "};\n",
       "\n",
       "// Register the function which deals with the matplotlib target/channel.\n",
       "// The kernel may be null if the page has been refreshed.\n",
       "if (IPython.notebook.kernel !== null) {\n",
       "    IPython.notebook.kernel.comm_manager.register_target(\n",
       "        'matplotlib',\n",
       "        mpl.mpl_figure_comm\n",
       "    );\n",
       "}\n"
      ],
      "text/plain": [
       "<IPython.core.display.Javascript object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/html": [
       "<img src=\"data:image/png;base64,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\" width=\"900\">"
      ],
      "text/plain": [
       "<IPython.core.display.HTML object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "rmse_in_different_price_range()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "c0932193",
   "metadata": {},
   "source": [
    "# Task 3"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 76,
   "id": "02b7ec15",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(4463, 53) (4463,) (1116, 53) (1116,)\n"
     ]
    }
   ],
   "source": [
    "def load_data(path = \"./data/preprocessed_data_task3.csv\"):\n",
    "\n",
    "    train_data = pd.read_csv(path)\n",
    "\n",
    "    X = train_data.iloc[:,0:-1]\n",
    "    X = np.array(X)\n",
    "    y = train_data[\"pred\"]\n",
    "    y = np.array(y)\n",
    "\n",
    "    x_train, x_test, y_train, y_test = train_test_split(X, y, test_size=0.2, random_state=2021)\n",
    "    print(x_train.shape, y_train.shape, x_test.shape, y_test.shape)\n",
    "    return x_train, y_train, x_test, y_test\n",
    "\n",
    "x_train, y_train, x_test, y_test = load_data()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 79,
   "id": "5c5cda14",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "WARNING:tensorflow:Unresolved object in checkpoint: (root).optimizer\n",
      "WARNING:tensorflow:Unresolved object in checkpoint: (root).optimizer.iter\n",
      "WARNING:tensorflow:Unresolved object in checkpoint: (root).optimizer.beta_1\n",
      "WARNING:tensorflow:Unresolved object in checkpoint: (root).optimizer.beta_2\n",
      "WARNING:tensorflow:Unresolved object in checkpoint: (root).optimizer.decay\n",
      "WARNING:tensorflow:Unresolved object in checkpoint: (root).optimizer's state 'm' for (root).emb1.kernel\n",
      "WARNING:tensorflow:Unresolved object in checkpoint: (root).optimizer's state 'm' for (root).emb1.bias\n",
      "WARNING:tensorflow:Unresolved object in checkpoint: (root).optimizer's state 'm' for (root).emb2.kernel\n",
      "WARNING:tensorflow:Unresolved object in checkpoint: (root).optimizer's state 'm' for (root).emb2.bias\n",
      "WARNING:tensorflow:Unresolved object in checkpoint: (root).optimizer's state 'm' for (root).emb3.kernel\n",
      "WARNING:tensorflow:Unresolved object in checkpoint: (root).optimizer's state 'm' for (root).emb3.bias\n",
      "WARNING:tensorflow:Unresolved object in checkpoint: (root).optimizer's state 'm' for (root).emb4.kernel\n",
      "WARNING:tensorflow:Unresolved object in checkpoint: (root).optimizer's state 'm' for (root).emb4.bias\n",
      "WARNING:tensorflow:Unresolved object in checkpoint: (root).optimizer's state 'm' for (root).num_emb.kernel\n",
      "WARNING:tensorflow:Unresolved object in checkpoint: (root).optimizer's state 'm' for (root).num_emb.bias\n",
      "WARNING:tensorflow:Unresolved object in checkpoint: (root).optimizer's state 'm' for (root).linear1.kernel\n",
      "WARNING:tensorflow:Unresolved object in checkpoint: (root).optimizer's state 'm' for (root).linear1.bias\n",
      "WARNING:tensorflow:Unresolved object in checkpoint: (root).optimizer's state 'm' for (root).linear2.kernel\n",
      "WARNING:tensorflow:Unresolved object in checkpoint: (root).optimizer's state 'm' for (root).linear2.bias\n",
      "WARNING:tensorflow:Unresolved object in checkpoint: (root).optimizer's state 'm' for (root).linear3.kernel\n",
      "WARNING:tensorflow:Unresolved object in checkpoint: (root).optimizer's state 'm' for (root).linear3.bias\n",
      "WARNING:tensorflow:Unresolved object in checkpoint: (root).optimizer's state 'm' for (root).linear4.kernel\n",
      "WARNING:tensorflow:Unresolved object in checkpoint: (root).optimizer's state 'm' for (root).linear4.bias\n",
      "WARNING:tensorflow:Unresolved object in checkpoint: (root).optimizer's state 'm' for (root).linear5.kernel\n",
      "WARNING:tensorflow:Unresolved object in checkpoint: (root).optimizer's state 'm' for (root).linear5.bias\n",
      "WARNING:tensorflow:Unresolved object in checkpoint: (root).optimizer's state 'm' for (root).linear6.kernel\n",
      "WARNING:tensorflow:Unresolved object in checkpoint: (root).optimizer's state 'm' for (root).linear6.bias\n",
      "WARNING:tensorflow:Unresolved object in checkpoint: (root).optimizer's state 'v' for (root).emb1.kernel\n",
      "WARNING:tensorflow:Unresolved object in checkpoint: (root).optimizer's state 'v' for (root).emb1.bias\n",
      "WARNING:tensorflow:Unresolved object in checkpoint: (root).optimizer's state 'v' for (root).emb2.kernel\n",
      "WARNING:tensorflow:Unresolved object in checkpoint: (root).optimizer's state 'v' for (root).emb2.bias\n",
      "WARNING:tensorflow:Unresolved object in checkpoint: (root).optimizer's state 'v' for (root).emb3.kernel\n",
      "WARNING:tensorflow:Unresolved object in checkpoint: (root).optimizer's state 'v' for (root).emb3.bias\n",
      "WARNING:tensorflow:Unresolved object in checkpoint: (root).optimizer's state 'v' for (root).emb4.kernel\n",
      "WARNING:tensorflow:Unresolved object in checkpoint: (root).optimizer's state 'v' for (root).emb4.bias\n",
      "WARNING:tensorflow:Unresolved object in checkpoint: (root).optimizer's state 'v' for (root).num_emb.kernel\n",
      "WARNING:tensorflow:Unresolved object in checkpoint: (root).optimizer's state 'v' for (root).num_emb.bias\n",
      "WARNING:tensorflow:Unresolved object in checkpoint: (root).optimizer's state 'v' for (root).linear1.kernel\n",
      "WARNING:tensorflow:Unresolved object in checkpoint: (root).optimizer's state 'v' for (root).linear1.bias\n",
      "WARNING:tensorflow:Unresolved object in checkpoint: (root).optimizer's state 'v' for (root).linear2.kernel\n",
      "WARNING:tensorflow:Unresolved object in checkpoint: (root).optimizer's state 'v' for (root).linear2.bias\n",
      "WARNING:tensorflow:Unresolved object in checkpoint: (root).optimizer's state 'v' for (root).linear3.kernel\n",
      "WARNING:tensorflow:Unresolved object in checkpoint: (root).optimizer's state 'v' for (root).linear3.bias\n",
      "WARNING:tensorflow:Unresolved object in checkpoint: (root).optimizer's state 'v' for (root).linear4.kernel\n",
      "WARNING:tensorflow:Unresolved object in checkpoint: (root).optimizer's state 'v' for (root).linear4.bias\n",
      "WARNING:tensorflow:Unresolved object in checkpoint: (root).optimizer's state 'v' for (root).linear5.kernel\n",
      "WARNING:tensorflow:Unresolved object in checkpoint: (root).optimizer's state 'v' for (root).linear5.bias\n",
      "WARNING:tensorflow:Unresolved object in checkpoint: (root).optimizer's state 'v' for (root).linear6.kernel\n",
      "WARNING:tensorflow:Unresolved object in checkpoint: (root).optimizer's state 'v' for (root).linear6.bias\n",
      "WARNING:tensorflow:A checkpoint was restored (e.g. tf.train.Checkpoint.restore or tf.keras.Model.load_weights) but not all checkpointed values were used. See above for specific issues. Use expect_partial() on the load status object, e.g. tf.train.Checkpoint.restore(...).expect_partial(), to silence these warnings, or use assert_consumed() to make the check explicit. See https://www.tensorflow.org/guide/checkpoint#loading_mechanics for details.\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "GradientBoostingRegressor(learning_rate=0.15, max_features='sqrt',\n",
       "                          n_estimators=300, random_state=2021)"
      ]
     },
     "execution_count": 79,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# train\n",
    "\n",
    "dt = DecisionTreeRegressor(max_depth=100)\n",
    "dt.fit(x_train, y_train)\n",
    "\n",
    "rf = RandomForestRegressor(n_estimators=50, max_depth=10, random_state=2021)\n",
    "rf.fit(x_train, y_train)\n",
    "\n",
    "gb = GradientBoostingRegressor(n_estimators=300, learning_rate=0.15, max_features='sqrt', random_state=2021)\n",
    "gb.fit(x_train, y_train)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 80,
   "id": "aae07f0b",
   "metadata": {},
   "outputs": [],
   "source": [
    "dt_predict = dt.predict(x_test)\n",
    "dt_rmse = mean_squared_error(y_test, dt_predict, squared=False)\n",
    "\n",
    "rf_predict = rf.predict(x_test)\n",
    "rf_rmse = mean_squared_error(y_test, rf_predict, squared=False)\n",
    "\n",
    "gb_predict = gb.predict(x_test)\n",
    "gb_rmse = mean_squared_error(y_test, gb_predict, squared=False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 81,
   "id": "d9d12033",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0.2064706984295075 0.15199629539484402 0.12198192319786115\n"
     ]
    }
   ],
   "source": [
    "print(dt_rmse, rf_rmse, gb_rmse)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 88,
   "id": "42e7d69b",
   "metadata": {},
   "outputs": [],
   "source": [
    "def rmse_in_different_test_part():\n",
    "    kfold = KFold(n_splits=5)\n",
    "    dt_rmse = []\n",
    "    rf_rmse = []\n",
    "    gb_rmse = []\n",
    "    for _, test in kfold.split(x_test, y_test):\n",
    "        dt_predict = dt.predict(x_test[test])\n",
    "        dt_rmse.append(mean_squared_error(y_test[test], dt_predict, squared=False))\n",
    "\n",
    "        rf_predict = rf.predict(x_test[test])\n",
    "        rf_rmse.append(mean_squared_error(y_test[test], rf_predict, squared=False))\n",
    "\n",
    "        gb_predict = gb.predict(x_test[test])\n",
    "        gb_rmse.append(mean_squared_error(y_test[test], gb_predict, squared=False))\n",
    "\n",
    "    x = range(1, 6)\n",
    "    plt.figure(figsize=(9,6))\n",
    "    plt.plot(x, dt_rmse, c=\"r\", label=\"DecisionTree_RMSE\")\n",
    "    plt.plot(x, rf_rmse, c=\"pink\", label=\"RandomForest_RMSE\")\n",
    "    plt.plot(x, gb_rmse, c=\"y\", label=\"GradientBoost_RMSE\")\n",
    "\n",
    "    plt.xlabel(\"Set No.\")\n",
    "    plt.ylabel(\"RMSE\")\n",
    "    plt.xticks(np.arange(1, 6, 1), [f\"test set {i}\" for i in range(1, 6)])\n",
    "    plt.legend()\n",
    "    plt.show()\n",
    "    plt.savefig(\"x4 RMSE in different part.png\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 89,
   "id": "033e3884",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "application/javascript": [
       "/* Put everything inside the global mpl namespace */\n",
       "/* global mpl */\n",
       "window.mpl = {};\n",
       "\n",
       "mpl.get_websocket_type = function () {\n",
       "    if (typeof WebSocket !== 'undefined') {\n",
       "        return WebSocket;\n",
       "    } else if (typeof MozWebSocket !== 'undefined') {\n",
       "        return MozWebSocket;\n",
       "    } else {\n",
       "        alert(\n",
       "            'Your browser does not have WebSocket support. ' +\n",
       "                'Please try Chrome, Safari or Firefox ≥ 6. ' +\n",
       "                'Firefox 4 and 5 are also supported but you ' +\n",
       "                'have to enable WebSockets in about:config.'\n",
       "        );\n",
       "    }\n",
       "};\n",
       "\n",
       "mpl.figure = function (figure_id, websocket, ondownload, parent_element) {\n",
       "    this.id = figure_id;\n",
       "\n",
       "    this.ws = websocket;\n",
       "\n",
       "    this.supports_binary = this.ws.binaryType !== undefined;\n",
       "\n",
       "    if (!this.supports_binary) {\n",
       "        var warnings = document.getElementById('mpl-warnings');\n",
       "        if (warnings) {\n",
       "            warnings.style.display = 'block';\n",
       "            warnings.textContent =\n",
       "                'This browser does not support binary websocket messages. ' +\n",
       "                'Performance may be slow.';\n",
       "        }\n",
       "    }\n",
       "\n",
       "    this.imageObj = new Image();\n",
       "\n",
       "    this.context = undefined;\n",
       "    this.message = undefined;\n",
       "    this.canvas = undefined;\n",
       "    this.rubberband_canvas = undefined;\n",
       "    this.rubberband_context = undefined;\n",
       "    this.format_dropdown = undefined;\n",
       "\n",
       "    this.image_mode = 'full';\n",
       "\n",
       "    this.root = document.createElement('div');\n",
       "    this.root.setAttribute('style', 'display: inline-block');\n",
       "    this._root_extra_style(this.root);\n",
       "\n",
       "    parent_element.appendChild(this.root);\n",
       "\n",
       "    this._init_header(this);\n",
       "    this._init_canvas(this);\n",
       "    this._init_toolbar(this);\n",
       "\n",
       "    var fig = this;\n",
       "\n",
       "    this.waiting = false;\n",
       "\n",
       "    this.ws.onopen = function () {\n",
       "        fig.send_message('supports_binary', { value: fig.supports_binary });\n",
       "        fig.send_message('send_image_mode', {});\n",
       "        if (fig.ratio !== 1) {\n",
       "            fig.send_message('set_dpi_ratio', { dpi_ratio: fig.ratio });\n",
       "        }\n",
       "        fig.send_message('refresh', {});\n",
       "    };\n",
       "\n",
       "    this.imageObj.onload = function () {\n",
       "        if (fig.image_mode === 'full') {\n",
       "            // Full images could contain transparency (where diff images\n",
       "            // almost always do), so we need to clear the canvas so that\n",
       "            // there is no ghosting.\n",
       "            fig.context.clearRect(0, 0, fig.canvas.width, fig.canvas.height);\n",
       "        }\n",
       "        fig.context.drawImage(fig.imageObj, 0, 0);\n",
       "    };\n",
       "\n",
       "    this.imageObj.onunload = function () {\n",
       "        fig.ws.close();\n",
       "    };\n",
       "\n",
       "    this.ws.onmessage = this._make_on_message_function(this);\n",
       "\n",
       "    this.ondownload = ondownload;\n",
       "};\n",
       "\n",
       "mpl.figure.prototype._init_header = function () {\n",
       "    var titlebar = document.createElement('div');\n",
       "    titlebar.classList =\n",
       "        'ui-dialog-titlebar ui-widget-header ui-corner-all ui-helper-clearfix';\n",
       "    var titletext = document.createElement('div');\n",
       "    titletext.classList = 'ui-dialog-title';\n",
       "    titletext.setAttribute(\n",
       "        'style',\n",
       "        'width: 100%; text-align: center; padding: 3px;'\n",
       "    );\n",
       "    titlebar.appendChild(titletext);\n",
       "    this.root.appendChild(titlebar);\n",
       "    this.header = titletext;\n",
       "};\n",
       "\n",
       "mpl.figure.prototype._canvas_extra_style = function (_canvas_div) {};\n",
       "\n",
       "mpl.figure.prototype._root_extra_style = function (_canvas_div) {};\n",
       "\n",
       "mpl.figure.prototype._init_canvas = function () {\n",
       "    var fig = this;\n",
       "\n",
       "    var canvas_div = (this.canvas_div = document.createElement('div'));\n",
       "    canvas_div.setAttribute(\n",
       "        'style',\n",
       "        'border: 1px solid #ddd;' +\n",
       "            'box-sizing: content-box;' +\n",
       "            'clear: both;' +\n",
       "            'min-height: 1px;' +\n",
       "            'min-width: 1px;' +\n",
       "            'outline: 0;' +\n",
       "            'overflow: hidden;' +\n",
       "            'position: relative;' +\n",
       "            'resize: both;'\n",
       "    );\n",
       "\n",
       "    function on_keyboard_event_closure(name) {\n",
       "        return function (event) {\n",
       "            return fig.key_event(event, name);\n",
       "        };\n",
       "    }\n",
       "\n",
       "    canvas_div.addEventListener(\n",
       "        'keydown',\n",
       "        on_keyboard_event_closure('key_press')\n",
       "    );\n",
       "    canvas_div.addEventListener(\n",
       "        'keyup',\n",
       "        on_keyboard_event_closure('key_release')\n",
       "    );\n",
       "\n",
       "    this._canvas_extra_style(canvas_div);\n",
       "    this.root.appendChild(canvas_div);\n",
       "\n",
       "    var canvas = (this.canvas = document.createElement('canvas'));\n",
       "    canvas.classList.add('mpl-canvas');\n",
       "    canvas.setAttribute('style', 'box-sizing: content-box;');\n",
       "\n",
       "    this.context = canvas.getContext('2d');\n",
       "\n",
       "    var backingStore =\n",
       "        this.context.backingStorePixelRatio ||\n",
       "        this.context.webkitBackingStorePixelRatio ||\n",
       "        this.context.mozBackingStorePixelRatio ||\n",
       "        this.context.msBackingStorePixelRatio ||\n",
       "        this.context.oBackingStorePixelRatio ||\n",
       "        this.context.backingStorePixelRatio ||\n",
       "        1;\n",
       "\n",
       "    this.ratio = (window.devicePixelRatio || 1) / backingStore;\n",
       "    if (this.ratio !== 1) {\n",
       "        fig.send_message('set_dpi_ratio', { dpi_ratio: this.ratio });\n",
       "    }\n",
       "\n",
       "    var rubberband_canvas = (this.rubberband_canvas = document.createElement(\n",
       "        'canvas'\n",
       "    ));\n",
       "    rubberband_canvas.setAttribute(\n",
       "        'style',\n",
       "        'box-sizing: content-box; position: absolute; left: 0; top: 0; z-index: 1;'\n",
       "    );\n",
       "\n",
       "    var resizeObserver = new ResizeObserver(function (entries) {\n",
       "        var nentries = entries.length;\n",
       "        for (var i = 0; i < nentries; i++) {\n",
       "            var entry = entries[i];\n",
       "            var width, height;\n",
       "            if (entry.contentBoxSize) {\n",
       "                if (entry.contentBoxSize instanceof Array) {\n",
       "                    // Chrome 84 implements new version of spec.\n",
       "                    width = entry.contentBoxSize[0].inlineSize;\n",
       "                    height = entry.contentBoxSize[0].blockSize;\n",
       "                } else {\n",
       "                    // Firefox implements old version of spec.\n",
       "                    width = entry.contentBoxSize.inlineSize;\n",
       "                    height = entry.contentBoxSize.blockSize;\n",
       "                }\n",
       "            } else {\n",
       "                // Chrome <84 implements even older version of spec.\n",
       "                width = entry.contentRect.width;\n",
       "                height = entry.contentRect.height;\n",
       "            }\n",
       "\n",
       "            // Keep the size of the canvas and rubber band canvas in sync with\n",
       "            // the canvas container.\n",
       "            if (entry.devicePixelContentBoxSize) {\n",
       "                // Chrome 84 implements new version of spec.\n",
       "                canvas.setAttribute(\n",
       "                    'width',\n",
       "                    entry.devicePixelContentBoxSize[0].inlineSize\n",
       "                );\n",
       "                canvas.setAttribute(\n",
       "                    'height',\n",
       "                    entry.devicePixelContentBoxSize[0].blockSize\n",
       "                );\n",
       "            } else {\n",
       "                canvas.setAttribute('width', width * fig.ratio);\n",
       "                canvas.setAttribute('height', height * fig.ratio);\n",
       "            }\n",
       "            canvas.setAttribute(\n",
       "                'style',\n",
       "                'width: ' + width + 'px; height: ' + height + 'px;'\n",
       "            );\n",
       "\n",
       "            rubberband_canvas.setAttribute('width', width);\n",
       "            rubberband_canvas.setAttribute('height', height);\n",
       "\n",
       "            // And update the size in Python. We ignore the initial 0/0 size\n",
       "            // that occurs as the element is placed into the DOM, which should\n",
       "            // otherwise not happen due to the minimum size styling.\n",
       "            if (width != 0 && height != 0) {\n",
       "                fig.request_resize(width, height);\n",
       "            }\n",
       "        }\n",
       "    });\n",
       "    resizeObserver.observe(canvas_div);\n",
       "\n",
       "    function on_mouse_event_closure(name) {\n",
       "        return function (event) {\n",
       "            return fig.mouse_event(event, name);\n",
       "        };\n",
       "    }\n",
       "\n",
       "    rubberband_canvas.addEventListener(\n",
       "        'mousedown',\n",
       "        on_mouse_event_closure('button_press')\n",
       "    );\n",
       "    rubberband_canvas.addEventListener(\n",
       "        'mouseup',\n",
       "        on_mouse_event_closure('button_release')\n",
       "    );\n",
       "    // Throttle sequential mouse events to 1 every 20ms.\n",
       "    rubberband_canvas.addEventListener(\n",
       "        'mousemove',\n",
       "        on_mouse_event_closure('motion_notify')\n",
       "    );\n",
       "\n",
       "    rubberband_canvas.addEventListener(\n",
       "        'mouseenter',\n",
       "        on_mouse_event_closure('figure_enter')\n",
       "    );\n",
       "    rubberband_canvas.addEventListener(\n",
       "        'mouseleave',\n",
       "        on_mouse_event_closure('figure_leave')\n",
       "    );\n",
       "\n",
       "    canvas_div.addEventListener('wheel', function (event) {\n",
       "        if (event.deltaY < 0) {\n",
       "            event.step = 1;\n",
       "        } else {\n",
       "            event.step = -1;\n",
       "        }\n",
       "        on_mouse_event_closure('scroll')(event);\n",
       "    });\n",
       "\n",
       "    canvas_div.appendChild(canvas);\n",
       "    canvas_div.appendChild(rubberband_canvas);\n",
       "\n",
       "    this.rubberband_context = rubberband_canvas.getContext('2d');\n",
       "    this.rubberband_context.strokeStyle = '#000000';\n",
       "\n",
       "    this._resize_canvas = function (width, height, forward) {\n",
       "        if (forward) {\n",
       "            canvas_div.style.width = width + 'px';\n",
       "            canvas_div.style.height = height + 'px';\n",
       "        }\n",
       "    };\n",
       "\n",
       "    // Disable right mouse context menu.\n",
       "    this.rubberband_canvas.addEventListener('contextmenu', function (_e) {\n",
       "        event.preventDefault();\n",
       "        return false;\n",
       "    });\n",
       "\n",
       "    function set_focus() {\n",
       "        canvas.focus();\n",
       "        canvas_div.focus();\n",
       "    }\n",
       "\n",
       "    window.setTimeout(set_focus, 100);\n",
       "};\n",
       "\n",
       "mpl.figure.prototype._init_toolbar = function () {\n",
       "    var fig = this;\n",
       "\n",
       "    var toolbar = document.createElement('div');\n",
       "    toolbar.classList = 'mpl-toolbar';\n",
       "    this.root.appendChild(toolbar);\n",
       "\n",
       "    function on_click_closure(name) {\n",
       "        return function (_event) {\n",
       "            return fig.toolbar_button_onclick(name);\n",
       "        };\n",
       "    }\n",
       "\n",
       "    function on_mouseover_closure(tooltip) {\n",
       "        return function (event) {\n",
       "            if (!event.currentTarget.disabled) {\n",
       "                return fig.toolbar_button_onmouseover(tooltip);\n",
       "            }\n",
       "        };\n",
       "    }\n",
       "\n",
       "    fig.buttons = {};\n",
       "    var buttonGroup = document.createElement('div');\n",
       "    buttonGroup.classList = 'mpl-button-group';\n",
       "    for (var toolbar_ind in mpl.toolbar_items) {\n",
       "        var name = mpl.toolbar_items[toolbar_ind][0];\n",
       "        var tooltip = mpl.toolbar_items[toolbar_ind][1];\n",
       "        var image = mpl.toolbar_items[toolbar_ind][2];\n",
       "        var method_name = mpl.toolbar_items[toolbar_ind][3];\n",
       "\n",
       "        if (!name) {\n",
       "            /* Instead of a spacer, we start a new button group. */\n",
       "            if (buttonGroup.hasChildNodes()) {\n",
       "                toolbar.appendChild(buttonGroup);\n",
       "            }\n",
       "            buttonGroup = document.createElement('div');\n",
       "            buttonGroup.classList = 'mpl-button-group';\n",
       "            continue;\n",
       "        }\n",
       "\n",
       "        var button = (fig.buttons[name] = document.createElement('button'));\n",
       "        button.classList = 'mpl-widget';\n",
       "        button.setAttribute('role', 'button');\n",
       "        button.setAttribute('aria-disabled', 'false');\n",
       "        button.addEventListener('click', on_click_closure(method_name));\n",
       "        button.addEventListener('mouseover', on_mouseover_closure(tooltip));\n",
       "\n",
       "        var icon_img = document.createElement('img');\n",
       "        icon_img.src = '_images/' + image + '.png';\n",
       "        icon_img.srcset = '_images/' + image + '_large.png 2x';\n",
       "        icon_img.alt = tooltip;\n",
       "        button.appendChild(icon_img);\n",
       "\n",
       "        buttonGroup.appendChild(button);\n",
       "    }\n",
       "\n",
       "    if (buttonGroup.hasChildNodes()) {\n",
       "        toolbar.appendChild(buttonGroup);\n",
       "    }\n",
       "\n",
       "    var fmt_picker = document.createElement('select');\n",
       "    fmt_picker.classList = 'mpl-widget';\n",
       "    toolbar.appendChild(fmt_picker);\n",
       "    this.format_dropdown = fmt_picker;\n",
       "\n",
       "    for (var ind in mpl.extensions) {\n",
       "        var fmt = mpl.extensions[ind];\n",
       "        var option = document.createElement('option');\n",
       "        option.selected = fmt === mpl.default_extension;\n",
       "        option.innerHTML = fmt;\n",
       "        fmt_picker.appendChild(option);\n",
       "    }\n",
       "\n",
       "    var status_bar = document.createElement('span');\n",
       "    status_bar.classList = 'mpl-message';\n",
       "    toolbar.appendChild(status_bar);\n",
       "    this.message = status_bar;\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.request_resize = function (x_pixels, y_pixels) {\n",
       "    // Request matplotlib to resize the figure. Matplotlib will then trigger a resize in the client,\n",
       "    // which will in turn request a refresh of the image.\n",
       "    this.send_message('resize', { width: x_pixels, height: y_pixels });\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.send_message = function (type, properties) {\n",
       "    properties['type'] = type;\n",
       "    properties['figure_id'] = this.id;\n",
       "    this.ws.send(JSON.stringify(properties));\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.send_draw_message = function () {\n",
       "    if (!this.waiting) {\n",
       "        this.waiting = true;\n",
       "        this.ws.send(JSON.stringify({ type: 'draw', figure_id: this.id }));\n",
       "    }\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.handle_save = function (fig, _msg) {\n",
       "    var format_dropdown = fig.format_dropdown;\n",
       "    var format = format_dropdown.options[format_dropdown.selectedIndex].value;\n",
       "    fig.ondownload(fig, format);\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.handle_resize = function (fig, msg) {\n",
       "    var size = msg['size'];\n",
       "    if (size[0] !== fig.canvas.width || size[1] !== fig.canvas.height) {\n",
       "        fig._resize_canvas(size[0], size[1], msg['forward']);\n",
       "        fig.send_message('refresh', {});\n",
       "    }\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.handle_rubberband = function (fig, msg) {\n",
       "    var x0 = msg['x0'] / fig.ratio;\n",
       "    var y0 = (fig.canvas.height - msg['y0']) / fig.ratio;\n",
       "    var x1 = msg['x1'] / fig.ratio;\n",
       "    var y1 = (fig.canvas.height - msg['y1']) / fig.ratio;\n",
       "    x0 = Math.floor(x0) + 0.5;\n",
       "    y0 = Math.floor(y0) + 0.5;\n",
       "    x1 = Math.floor(x1) + 0.5;\n",
       "    y1 = Math.floor(y1) + 0.5;\n",
       "    var min_x = Math.min(x0, x1);\n",
       "    var min_y = Math.min(y0, y1);\n",
       "    var width = Math.abs(x1 - x0);\n",
       "    var height = Math.abs(y1 - y0);\n",
       "\n",
       "    fig.rubberband_context.clearRect(\n",
       "        0,\n",
       "        0,\n",
       "        fig.canvas.width / fig.ratio,\n",
       "        fig.canvas.height / fig.ratio\n",
       "    );\n",
       "\n",
       "    fig.rubberband_context.strokeRect(min_x, min_y, width, height);\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.handle_figure_label = function (fig, msg) {\n",
       "    // Updates the figure title.\n",
       "    fig.header.textContent = msg['label'];\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.handle_cursor = function (fig, msg) {\n",
       "    var cursor = msg['cursor'];\n",
       "    switch (cursor) {\n",
       "        case 0:\n",
       "            cursor = 'pointer';\n",
       "            break;\n",
       "        case 1:\n",
       "            cursor = 'default';\n",
       "            break;\n",
       "        case 2:\n",
       "            cursor = 'crosshair';\n",
       "            break;\n",
       "        case 3:\n",
       "            cursor = 'move';\n",
       "            break;\n",
       "    }\n",
       "    fig.rubberband_canvas.style.cursor = cursor;\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.handle_message = function (fig, msg) {\n",
       "    fig.message.textContent = msg['message'];\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.handle_draw = function (fig, _msg) {\n",
       "    // Request the server to send over a new figure.\n",
       "    fig.send_draw_message();\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.handle_image_mode = function (fig, msg) {\n",
       "    fig.image_mode = msg['mode'];\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.handle_history_buttons = function (fig, msg) {\n",
       "    for (var key in msg) {\n",
       "        if (!(key in fig.buttons)) {\n",
       "            continue;\n",
       "        }\n",
       "        fig.buttons[key].disabled = !msg[key];\n",
       "        fig.buttons[key].setAttribute('aria-disabled', !msg[key]);\n",
       "    }\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.handle_navigate_mode = function (fig, msg) {\n",
       "    if (msg['mode'] === 'PAN') {\n",
       "        fig.buttons['Pan'].classList.add('active');\n",
       "        fig.buttons['Zoom'].classList.remove('active');\n",
       "    } else if (msg['mode'] === 'ZOOM') {\n",
       "        fig.buttons['Pan'].classList.remove('active');\n",
       "        fig.buttons['Zoom'].classList.add('active');\n",
       "    } else {\n",
       "        fig.buttons['Pan'].classList.remove('active');\n",
       "        fig.buttons['Zoom'].classList.remove('active');\n",
       "    }\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.updated_canvas_event = function () {\n",
       "    // Called whenever the canvas gets updated.\n",
       "    this.send_message('ack', {});\n",
       "};\n",
       "\n",
       "// A function to construct a web socket function for onmessage handling.\n",
       "// Called in the figure constructor.\n",
       "mpl.figure.prototype._make_on_message_function = function (fig) {\n",
       "    return function socket_on_message(evt) {\n",
       "        if (evt.data instanceof Blob) {\n",
       "            /* FIXME: We get \"Resource interpreted as Image but\n",
       "             * transferred with MIME type text/plain:\" errors on\n",
       "             * Chrome.  But how to set the MIME type?  It doesn't seem\n",
       "             * to be part of the websocket stream */\n",
       "            evt.data.type = 'image/png';\n",
       "\n",
       "            /* Free the memory for the previous frames */\n",
       "            if (fig.imageObj.src) {\n",
       "                (window.URL || window.webkitURL).revokeObjectURL(\n",
       "                    fig.imageObj.src\n",
       "                );\n",
       "            }\n",
       "\n",
       "            fig.imageObj.src = (window.URL || window.webkitURL).createObjectURL(\n",
       "                evt.data\n",
       "            );\n",
       "            fig.updated_canvas_event();\n",
       "            fig.waiting = false;\n",
       "            return;\n",
       "        } else if (\n",
       "            typeof evt.data === 'string' &&\n",
       "            evt.data.slice(0, 21) === 'data:image/png;base64'\n",
       "        ) {\n",
       "            fig.imageObj.src = evt.data;\n",
       "            fig.updated_canvas_event();\n",
       "            fig.waiting = false;\n",
       "            return;\n",
       "        }\n",
       "\n",
       "        var msg = JSON.parse(evt.data);\n",
       "        var msg_type = msg['type'];\n",
       "\n",
       "        // Call the  \"handle_{type}\" callback, which takes\n",
       "        // the figure and JSON message as its only arguments.\n",
       "        try {\n",
       "            var callback = fig['handle_' + msg_type];\n",
       "        } catch (e) {\n",
       "            console.log(\n",
       "                \"No handler for the '\" + msg_type + \"' message type: \",\n",
       "                msg\n",
       "            );\n",
       "            return;\n",
       "        }\n",
       "\n",
       "        if (callback) {\n",
       "            try {\n",
       "                // console.log(\"Handling '\" + msg_type + \"' message: \", msg);\n",
       "                callback(fig, msg);\n",
       "            } catch (e) {\n",
       "                console.log(\n",
       "                    \"Exception inside the 'handler_\" + msg_type + \"' callback:\",\n",
       "                    e,\n",
       "                    e.stack,\n",
       "                    msg\n",
       "                );\n",
       "            }\n",
       "        }\n",
       "    };\n",
       "};\n",
       "\n",
       "// from http://stackoverflow.com/questions/1114465/getting-mouse-location-in-canvas\n",
       "mpl.findpos = function (e) {\n",
       "    //this section is from http://www.quirksmode.org/js/events_properties.html\n",
       "    var targ;\n",
       "    if (!e) {\n",
       "        e = window.event;\n",
       "    }\n",
       "    if (e.target) {\n",
       "        targ = e.target;\n",
       "    } else if (e.srcElement) {\n",
       "        targ = e.srcElement;\n",
       "    }\n",
       "    if (targ.nodeType === 3) {\n",
       "        // defeat Safari bug\n",
       "        targ = targ.parentNode;\n",
       "    }\n",
       "\n",
       "    // pageX,Y are the mouse positions relative to the document\n",
       "    var boundingRect = targ.getBoundingClientRect();\n",
       "    var x = e.pageX - (boundingRect.left + document.body.scrollLeft);\n",
       "    var y = e.pageY - (boundingRect.top + document.body.scrollTop);\n",
       "\n",
       "    return { x: x, y: y };\n",
       "};\n",
       "\n",
       "/*\n",
       " * return a copy of an object with only non-object keys\n",
       " * we need this to avoid circular references\n",
       " * http://stackoverflow.com/a/24161582/3208463\n",
       " */\n",
       "function simpleKeys(original) {\n",
       "    return Object.keys(original).reduce(function (obj, key) {\n",
       "        if (typeof original[key] !== 'object') {\n",
       "            obj[key] = original[key];\n",
       "        }\n",
       "        return obj;\n",
       "    }, {});\n",
       "}\n",
       "\n",
       "mpl.figure.prototype.mouse_event = function (event, name) {\n",
       "    var canvas_pos = mpl.findpos(event);\n",
       "\n",
       "    if (name === 'button_press') {\n",
       "        this.canvas.focus();\n",
       "        this.canvas_div.focus();\n",
       "    }\n",
       "\n",
       "    var x = canvas_pos.x * this.ratio;\n",
       "    var y = canvas_pos.y * this.ratio;\n",
       "\n",
       "    this.send_message(name, {\n",
       "        x: x,\n",
       "        y: y,\n",
       "        button: event.button,\n",
       "        step: event.step,\n",
       "        guiEvent: simpleKeys(event),\n",
       "    });\n",
       "\n",
       "    /* This prevents the web browser from automatically changing to\n",
       "     * the text insertion cursor when the button is pressed.  We want\n",
       "     * to control all of the cursor setting manually through the\n",
       "     * 'cursor' event from matplotlib */\n",
       "    event.preventDefault();\n",
       "    return false;\n",
       "};\n",
       "\n",
       "mpl.figure.prototype._key_event_extra = function (_event, _name) {\n",
       "    // Handle any extra behaviour associated with a key event\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.key_event = function (event, name) {\n",
       "    // Prevent repeat events\n",
       "    if (name === 'key_press') {\n",
       "        if (event.which === this._key) {\n",
       "            return;\n",
       "        } else {\n",
       "            this._key = event.which;\n",
       "        }\n",
       "    }\n",
       "    if (name === 'key_release') {\n",
       "        this._key = null;\n",
       "    }\n",
       "\n",
       "    var value = '';\n",
       "    if (event.ctrlKey && event.which !== 17) {\n",
       "        value += 'ctrl+';\n",
       "    }\n",
       "    if (event.altKey && event.which !== 18) {\n",
       "        value += 'alt+';\n",
       "    }\n",
       "    if (event.shiftKey && event.which !== 16) {\n",
       "        value += 'shift+';\n",
       "    }\n",
       "\n",
       "    value += 'k';\n",
       "    value += event.which.toString();\n",
       "\n",
       "    this._key_event_extra(event, name);\n",
       "\n",
       "    this.send_message(name, { key: value, guiEvent: simpleKeys(event) });\n",
       "    return false;\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.toolbar_button_onclick = function (name) {\n",
       "    if (name === 'download') {\n",
       "        this.handle_save(this, null);\n",
       "    } else {\n",
       "        this.send_message('toolbar_button', { name: name });\n",
       "    }\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.toolbar_button_onmouseover = function (tooltip) {\n",
       "    this.message.textContent = tooltip;\n",
       "};\n",
       "mpl.toolbar_items = [[\"Home\", \"Reset original view\", \"fa fa-home icon-home\", \"home\"], [\"Back\", \"Back to previous view\", \"fa fa-arrow-left icon-arrow-left\", \"back\"], [\"Forward\", \"Forward to next view\", \"fa fa-arrow-right icon-arrow-right\", \"forward\"], [\"\", \"\", \"\", \"\"], [\"Pan\", \"Left button pans, Right button zooms\\nx/y fixes axis, CTRL fixes aspect\", \"fa fa-arrows icon-move\", \"pan\"], [\"Zoom\", \"Zoom to rectangle\\nx/y fixes axis, CTRL fixes aspect\", \"fa fa-square-o icon-check-empty\", \"zoom\"], [\"\", \"\", \"\", \"\"], [\"Download\", \"Download plot\", \"fa fa-floppy-o icon-save\", \"download\"]];\n",
       "\n",
       "mpl.extensions = [\"eps\", \"jpeg\", \"pdf\", \"png\", \"ps\", \"raw\", \"svg\", \"tif\"];\n",
       "\n",
       "mpl.default_extension = \"png\";/* global mpl */\n",
       "\n",
       "var comm_websocket_adapter = function (comm) {\n",
       "    // Create a \"websocket\"-like object which calls the given IPython comm\n",
       "    // object with the appropriate methods. Currently this is a non binary\n",
       "    // socket, so there is still some room for performance tuning.\n",
       "    var ws = {};\n",
       "\n",
       "    ws.close = function () {\n",
       "        comm.close();\n",
       "    };\n",
       "    ws.send = function (m) {\n",
       "        //console.log('sending', m);\n",
       "        comm.send(m);\n",
       "    };\n",
       "    // Register the callback with on_msg.\n",
       "    comm.on_msg(function (msg) {\n",
       "        //console.log('receiving', msg['content']['data'], msg);\n",
       "        // Pass the mpl event to the overridden (by mpl) onmessage function.\n",
       "        ws.onmessage(msg['content']['data']);\n",
       "    });\n",
       "    return ws;\n",
       "};\n",
       "\n",
       "mpl.mpl_figure_comm = function (comm, msg) {\n",
       "    // This is the function which gets called when the mpl process\n",
       "    // starts-up an IPython Comm through the \"matplotlib\" channel.\n",
       "\n",
       "    var id = msg.content.data.id;\n",
       "    // Get hold of the div created by the display call when the Comm\n",
       "    // socket was opened in Python.\n",
       "    var element = document.getElementById(id);\n",
       "    var ws_proxy = comm_websocket_adapter(comm);\n",
       "\n",
       "    function ondownload(figure, _format) {\n",
       "        window.open(figure.canvas.toDataURL());\n",
       "    }\n",
       "\n",
       "    var fig = new mpl.figure(id, ws_proxy, ondownload, element);\n",
       "\n",
       "    // Call onopen now - mpl needs it, as it is assuming we've passed it a real\n",
       "    // web socket which is closed, not our websocket->open comm proxy.\n",
       "    ws_proxy.onopen();\n",
       "\n",
       "    fig.parent_element = element;\n",
       "    fig.cell_info = mpl.find_output_cell(\"<div id='\" + id + \"'></div>\");\n",
       "    if (!fig.cell_info) {\n",
       "        console.error('Failed to find cell for figure', id, fig);\n",
       "        return;\n",
       "    }\n",
       "    fig.cell_info[0].output_area.element.one(\n",
       "        'cleared',\n",
       "        { fig: fig },\n",
       "        fig._remove_fig_handler\n",
       "    );\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.handle_close = function (fig, msg) {\n",
       "    var width = fig.canvas.width / fig.ratio;\n",
       "    fig.cell_info[0].output_area.element.off(\n",
       "        'cleared',\n",
       "        fig._remove_fig_handler\n",
       "    );\n",
       "\n",
       "    // Update the output cell to use the data from the current canvas.\n",
       "    fig.push_to_output();\n",
       "    var dataURL = fig.canvas.toDataURL();\n",
       "    // Re-enable the keyboard manager in IPython - without this line, in FF,\n",
       "    // the notebook keyboard shortcuts fail.\n",
       "    IPython.keyboard_manager.enable();\n",
       "    fig.parent_element.innerHTML =\n",
       "        '<img src=\"' + dataURL + '\" width=\"' + width + '\">';\n",
       "    fig.close_ws(fig, msg);\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.close_ws = function (fig, msg) {\n",
       "    fig.send_message('closing', msg);\n",
       "    // fig.ws.close()\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.push_to_output = function (_remove_interactive) {\n",
       "    // Turn the data on the canvas into data in the output cell.\n",
       "    var width = this.canvas.width / this.ratio;\n",
       "    var dataURL = this.canvas.toDataURL();\n",
       "    this.cell_info[1]['text/html'] =\n",
       "        '<img src=\"' + dataURL + '\" width=\"' + width + '\">';\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.updated_canvas_event = function () {\n",
       "    // Tell IPython that the notebook contents must change.\n",
       "    IPython.notebook.set_dirty(true);\n",
       "    this.send_message('ack', {});\n",
       "    var fig = this;\n",
       "    // Wait a second, then push the new image to the DOM so\n",
       "    // that it is saved nicely (might be nice to debounce this).\n",
       "    setTimeout(function () {\n",
       "        fig.push_to_output();\n",
       "    }, 1000);\n",
       "};\n",
       "\n",
       "mpl.figure.prototype._init_toolbar = function () {\n",
       "    var fig = this;\n",
       "\n",
       "    var toolbar = document.createElement('div');\n",
       "    toolbar.classList = 'btn-toolbar';\n",
       "    this.root.appendChild(toolbar);\n",
       "\n",
       "    function on_click_closure(name) {\n",
       "        return function (_event) {\n",
       "            return fig.toolbar_button_onclick(name);\n",
       "        };\n",
       "    }\n",
       "\n",
       "    function on_mouseover_closure(tooltip) {\n",
       "        return function (event) {\n",
       "            if (!event.currentTarget.disabled) {\n",
       "                return fig.toolbar_button_onmouseover(tooltip);\n",
       "            }\n",
       "        };\n",
       "    }\n",
       "\n",
       "    fig.buttons = {};\n",
       "    var buttonGroup = document.createElement('div');\n",
       "    buttonGroup.classList = 'btn-group';\n",
       "    var button;\n",
       "    for (var toolbar_ind in mpl.toolbar_items) {\n",
       "        var name = mpl.toolbar_items[toolbar_ind][0];\n",
       "        var tooltip = mpl.toolbar_items[toolbar_ind][1];\n",
       "        var image = mpl.toolbar_items[toolbar_ind][2];\n",
       "        var method_name = mpl.toolbar_items[toolbar_ind][3];\n",
       "\n",
       "        if (!name) {\n",
       "            /* Instead of a spacer, we start a new button group. */\n",
       "            if (buttonGroup.hasChildNodes()) {\n",
       "                toolbar.appendChild(buttonGroup);\n",
       "            }\n",
       "            buttonGroup = document.createElement('div');\n",
       "            buttonGroup.classList = 'btn-group';\n",
       "            continue;\n",
       "        }\n",
       "\n",
       "        button = fig.buttons[name] = document.createElement('button');\n",
       "        button.classList = 'btn btn-default';\n",
       "        button.href = '#';\n",
       "        button.title = name;\n",
       "        button.innerHTML = '<i class=\"fa ' + image + ' fa-lg\"></i>';\n",
       "        button.addEventListener('click', on_click_closure(method_name));\n",
       "        button.addEventListener('mouseover', on_mouseover_closure(tooltip));\n",
       "        buttonGroup.appendChild(button);\n",
       "    }\n",
       "\n",
       "    if (buttonGroup.hasChildNodes()) {\n",
       "        toolbar.appendChild(buttonGroup);\n",
       "    }\n",
       "\n",
       "    // Add the status bar.\n",
       "    var status_bar = document.createElement('span');\n",
       "    status_bar.classList = 'mpl-message pull-right';\n",
       "    toolbar.appendChild(status_bar);\n",
       "    this.message = status_bar;\n",
       "\n",
       "    // Add the close button to the window.\n",
       "    var buttongrp = document.createElement('div');\n",
       "    buttongrp.classList = 'btn-group inline pull-right';\n",
       "    button = document.createElement('button');\n",
       "    button.classList = 'btn btn-mini btn-primary';\n",
       "    button.href = '#';\n",
       "    button.title = 'Stop Interaction';\n",
       "    button.innerHTML = '<i class=\"fa fa-power-off icon-remove icon-large\"></i>';\n",
       "    button.addEventListener('click', function (_evt) {\n",
       "        fig.handle_close(fig, {});\n",
       "    });\n",
       "    button.addEventListener(\n",
       "        'mouseover',\n",
       "        on_mouseover_closure('Stop Interaction')\n",
       "    );\n",
       "    buttongrp.appendChild(button);\n",
       "    var titlebar = this.root.querySelector('.ui-dialog-titlebar');\n",
       "    titlebar.insertBefore(buttongrp, titlebar.firstChild);\n",
       "};\n",
       "\n",
       "mpl.figure.prototype._remove_fig_handler = function (event) {\n",
       "    var fig = event.data.fig;\n",
       "    fig.close_ws(fig, {});\n",
       "};\n",
       "\n",
       "mpl.figure.prototype._root_extra_style = function (el) {\n",
       "    el.style.boxSizing = 'content-box'; // override notebook setting of border-box.\n",
       "};\n",
       "\n",
       "mpl.figure.prototype._canvas_extra_style = function (el) {\n",
       "    // this is important to make the div 'focusable\n",
       "    el.setAttribute('tabindex', 0);\n",
       "    // reach out to IPython and tell the keyboard manager to turn it's self\n",
       "    // off when our div gets focus\n",
       "\n",
       "    // location in version 3\n",
       "    if (IPython.notebook.keyboard_manager) {\n",
       "        IPython.notebook.keyboard_manager.register_events(el);\n",
       "    } else {\n",
       "        // location in version 2\n",
       "        IPython.keyboard_manager.register_events(el);\n",
       "    }\n",
       "};\n",
       "\n",
       "mpl.figure.prototype._key_event_extra = function (event, _name) {\n",
       "    var manager = IPython.notebook.keyboard_manager;\n",
       "    if (!manager) {\n",
       "        manager = IPython.keyboard_manager;\n",
       "    }\n",
       "\n",
       "    // Check for shift+enter\n",
       "    if (event.shiftKey && event.which === 13) {\n",
       "        this.canvas_div.blur();\n",
       "        // select the cell after this one\n",
       "        var index = IPython.notebook.find_cell_index(this.cell_info[0]);\n",
       "        IPython.notebook.select(index + 1);\n",
       "    }\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.handle_save = function (fig, _msg) {\n",
       "    fig.ondownload(fig, null);\n",
       "};\n",
       "\n",
       "mpl.find_output_cell = function (html_output) {\n",
       "    // Return the cell and output element which can be found *uniquely* in the notebook.\n",
       "    // Note - this is a bit hacky, but it is done because the \"notebook_saving.Notebook\"\n",
       "    // IPython event is triggered only after the cells have been serialised, which for\n",
       "    // our purposes (turning an active figure into a static one), is too late.\n",
       "    var cells = IPython.notebook.get_cells();\n",
       "    var ncells = cells.length;\n",
       "    for (var i = 0; i < ncells; i++) {\n",
       "        var cell = cells[i];\n",
       "        if (cell.cell_type === 'code') {\n",
       "            for (var j = 0; j < cell.output_area.outputs.length; j++) {\n",
       "                var data = cell.output_area.outputs[j];\n",
       "                if (data.data) {\n",
       "                    // IPython >= 3 moved mimebundle to data attribute of output\n",
       "                    data = data.data;\n",
       "                }\n",
       "                if (data['text/html'] === html_output) {\n",
       "                    return [cell, data, j];\n",
       "                }\n",
       "            }\n",
       "        }\n",
       "    }\n",
       "};\n",
       "\n",
       "// Register the function which deals with the matplotlib target/channel.\n",
       "// The kernel may be null if the page has been refreshed.\n",
       "if (IPython.notebook.kernel !== null) {\n",
       "    IPython.notebook.kernel.comm_manager.register_target(\n",
       "        'matplotlib',\n",
       "        mpl.mpl_figure_comm\n",
       "    );\n",
       "}\n"
      ],
      "text/plain": [
       "<IPython.core.display.Javascript object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/html": [
       "<img src=\"data:image/png;base64,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\" width=\"900\">"
      ],
      "text/plain": [
       "<IPython.core.display.HTML object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "rmse_in_different_test_part()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "a7274450",
   "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": "3.6.8"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}