Stock Price Prediction 0.5.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/usr/local/miniconda3/envs/mimic3/lib/python3.6/site-packages/pandas_datareader/compat/__init__.py:7: FutureWarning: pandas.util.testing is deprecated. Use the functions in the public API at pandas.testing instead.\n",
      "  from pandas.util.testing import assert_frame_equal\n"
     ]
    }
   ],
   "source": [
    "import math\n",
    "import numpy as np\n",
    "import pandas as pd\n",
    "import pandas_datareader as web"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.preprocessing import MinMaxScaler\n",
    "from sklearn.model_selection import train_test_split"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "import seaborn as sns\n",
    "import matplotlib.pyplot as plt\n",
    "from pandas.plotting import register_matplotlib_converters\n",
    "register_matplotlib_converters()\n",
    "\n",
    "%matplotlib inline"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "import torch\n",
    "from torch import nn\n",
    "from torch.nn import functional as F\n",
    "from torch import optim\n",
    "from stochastic_rnn import StochasticLSTM\n",
    "from variational_module import MCDualMixin\n",
    "from regression_uncertainty_utils import aleatoric_uncertainty, epistemic_uncertainty, prediction_avg, uncertainty\n",
    "from train_utils import train_model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<torch._C.Generator at 0x7fe1d2f168d0>"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "np.random.seed(4321)\n",
    "torch.manual_seed(4321)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>High</th>\n",
       "      <th>Low</th>\n",
       "      <th>Open</th>\n",
       "      <th>Close</th>\n",
       "      <th>Volume</th>\n",
       "      <th>Adj Close</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Date</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>2012-01-03</th>\n",
       "      <td>58.928570</td>\n",
       "      <td>58.428570</td>\n",
       "      <td>58.485714</td>\n",
       "      <td>58.747143</td>\n",
       "      <td>75555200.0</td>\n",
       "      <td>50.857235</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2012-01-04</th>\n",
       "      <td>59.240002</td>\n",
       "      <td>58.468571</td>\n",
       "      <td>58.571430</td>\n",
       "      <td>59.062859</td>\n",
       "      <td>65005500.0</td>\n",
       "      <td>51.130558</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2012-01-05</th>\n",
       "      <td>59.792858</td>\n",
       "      <td>58.952858</td>\n",
       "      <td>59.278572</td>\n",
       "      <td>59.718571</td>\n",
       "      <td>67817400.0</td>\n",
       "      <td>51.698215</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2012-01-06</th>\n",
       "      <td>60.392857</td>\n",
       "      <td>59.888573</td>\n",
       "      <td>59.967144</td>\n",
       "      <td>60.342857</td>\n",
       "      <td>79573200.0</td>\n",
       "      <td>52.238651</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2012-01-09</th>\n",
       "      <td>61.107143</td>\n",
       "      <td>60.192856</td>\n",
       "      <td>60.785713</td>\n",
       "      <td>60.247143</td>\n",
       "      <td>98506096.0</td>\n",
       "      <td>52.155792</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                 High        Low       Open      Close      Volume  Adj Close\n",
       "Date                                                                         \n",
       "2012-01-03  58.928570  58.428570  58.485714  58.747143  75555200.0  50.857235\n",
       "2012-01-04  59.240002  58.468571  58.571430  59.062859  65005500.0  51.130558\n",
       "2012-01-05  59.792858  58.952858  59.278572  59.718571  67817400.0  51.698215\n",
       "2012-01-06  60.392857  59.888573  59.967144  60.342857  79573200.0  52.238651\n",
       "2012-01-09  61.107143  60.192856  60.785713  60.247143  98506096.0  52.155792"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df = web.DataReader('AAPL', data_source='yahoo', start='2012-01-01', end='2019-12-17').astype(np.float32)\n",
    "df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1152x576 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.figure(figsize=(16, 8))\n",
    "plt.title(\"Close Price History\")\n",
    "plt.plot(df[\"Close\"])\n",
    "plt.xlabel(\"Date\", fontsize=18)\n",
    "plt.ylabel(\"Close Price USD ($)\", fontsize=18)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "1603"
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data = df.filter([\"Close\"])\n",
    "dataset = data.values\n",
    "training_data_len = math.ceil(len(dataset) * .8)\n",
    "\n",
    "training_data_len"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[0.01316509],\n",
       "       [0.01457062],\n",
       "       [0.01748985],\n",
       "       ...,\n",
       "       [0.97658265],\n",
       "       [0.99755144],\n",
       "       [1.0000001 ]], dtype=float32)"
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "scaler = MinMaxScaler(feature_range=(0, 1))\n",
    "scaled_data = scaler.fit_transform(dataset)\n",
    "scaled_data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [],
   "source": [
    "STEP = 60"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [],
   "source": [
    "train_data = scaled_data[0:training_data_len, :]\n",
    "\n",
    "x_train = np.array([train_data[i-STEP:i, 0] for i in range(STEP, training_data_len)])\n",
    "y_train = np.array([train_data[i, 0] for i in range(STEP, training_data_len)])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [],
   "source": [
    "ENTRY_COUNT = len(y_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(1543, 60, 1)"
      ]
     },
     "execution_count": 13,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "x_train = x_train.reshape(ENTRY_COUNT, STEP, 1)\n",
    "x_train.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [],
   "source": [
    "x_train, x_val, y_train, y_val = train_test_split(x_train, y_train, test_size=0.3, random_state=42)\n",
    "train_dl = torch.utils.data.DataLoader([(x_train[i], y_train[i]) for i in range(len(y_train))], shuffle=True, batch_size=50)\n",
    "val_dl = torch.utils.data.DataLoader([(x_val[i], y_val[i]) for i in range(len(y_val))], shuffle=True, batch_size=50)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [],
   "source": [
    "class Net(nn.Module, MCDualMixin):\n",
    "    def __init__(self):\n",
    "        super(Net, self).__init__()\n",
    "        self.rnn = StochasticLSTM(1, 50, dropout=0.5, num_layers=1)\n",
    "        self.fc1 = nn.Linear(50, 25)\n",
    "        self.fc_mean = nn.Linear(25, 1)\n",
    "        self.fc_log_var = nn.Linear(25, 1)\n",
    "    \n",
    "    def regularizer(self):        \n",
    "        # Weight and bias regularizer\n",
    "        weight_sum, bias_sum, dropout_reg = self.rnn.regularizer()\n",
    "        \n",
    "        return weight_sum + bias_sum + dropout_reg\n",
    "    \n",
    "    def get_output_shape(self, x):\n",
    "        B = x.shape[1]\n",
    "        return (B,1), (B,1)\n",
    "\n",
    "    def forward(self, x):\n",
    "        result, _ = self.rnn(x)\n",
    "        result = result[-1,:]\n",
    "        result = torch.relu(self.fc1(result))\n",
    "        mean = self.fc_mean(result)\n",
    "        log_var = self.fc_log_var(result)\n",
    "        return mean, log_var"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [],
   "source": [
    "net = Net()\n",
    "def criterion(output, y):\n",
    "    mean, log_var = output\n",
    "    return 0.5 * torch.mean(torch.exp(-log_var) * (mean-y.view(-1,1))**2 + log_var)\n",
    "    #return torch.mean((mean - y.view(-1,1))**2)\n",
    "\n",
    "optimizer = optim.Adam(net.parameters(), lr=1e-3)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[1, 100.00%] train loss: -0.013724\n",
      "[1, 100.00%] validate loss: -0.119672\n",
      "[2, 100.00%] train loss: -0.350881\n",
      "[2, 100.00%] validate loss: -0.792926\n",
      "[3, 100.00%] train loss: -1.218622\n",
      "[3, 100.00%] validate loss: -1.420729\n",
      "[4, 100.00%] train loss: -1.423532\n",
      "[4, 100.00%] validate loss: -1.516659\n",
      "[5, 100.00%] train loss: -1.442718\n",
      "[5, 100.00%] validate loss: -1.529602\n",
      "[6, 100.00%] train loss: -1.441504\n",
      "[6, 100.00%] validate loss: -1.460087\n",
      "Retry 1/6\n",
      "[7, 100.00%] train loss: -1.476571\n",
      "[7, 100.00%] validate loss: -1.515593\n",
      "Retry 2/6\n",
      "[8, 100.00%] train loss: -1.481288\n",
      "[8, 100.00%] validate loss: -1.559829\n",
      "[9, 100.00%] train loss: -1.481756\n",
      "[9, 100.00%] validate loss: -1.531821\n",
      "Retry 1/6\n",
      "[10, 100.00%] train loss: -1.504717\n",
      "[10, 100.00%] validate loss: -1.615595\n",
      "[11, 100.00%] train loss: -1.512901\n",
      "[11, 100.00%] validate loss: -1.567260\n",
      "Retry 1/6\n",
      "[12, 100.00%] train loss: -1.514554\n",
      "[12, 100.00%] validate loss: -1.605565\n",
      "Retry 2/6\n",
      "[13, 100.00%] train loss: -1.539570\n",
      "[13, 100.00%] validate loss: -1.670401\n",
      "[14, 100.00%] train loss: -1.559449\n",
      "[14, 100.00%] validate loss: -1.666999\n",
      "Retry 1/6\n",
      "[15, 100.00%] train loss: -1.657957\n",
      "[15, 100.00%] validate loss: -1.760765\n",
      "[16, 100.00%] train loss: -1.717754\n",
      "[16, 100.00%] validate loss: -1.858662\n",
      "[17, 100.00%] train loss: -1.884723\n",
      "[17, 100.00%] validate loss: -2.055301\n",
      "[18, 100.00%] train loss: -2.084634\n",
      "[18, 100.00%] validate loss: -2.304468\n",
      "[19, 100.00%] train loss: -2.322118\n",
      "[19, 100.00%] validate loss: -2.638572\n",
      "[20, 100.00%] train loss: -2.510450\n",
      "[20, 100.00%] validate loss: -2.609157\n",
      "Retry 1/6\n",
      "[21, 100.00%] train loss: -2.674789\n",
      "[21, 100.00%] validate loss: -3.064799\n",
      "[22, 100.00%] train loss: -2.674001\n",
      "[22, 100.00%] validate loss: -3.079704\n",
      "[23, 100.00%] train loss: -2.891236\n",
      "[23, 100.00%] validate loss: -3.193616\n",
      "[24, 100.00%] train loss: -3.049439\n",
      "[24, 100.00%] validate loss: -3.444107\n",
      "[25, 100.00%] train loss: -3.233443\n",
      "[25, 100.00%] validate loss: -3.232377\n",
      "Retry 1/6\n",
      "[26, 100.00%] train loss: -3.222874\n",
      "[26, 100.00%] validate loss: -3.399259\n",
      "Retry 2/6\n",
      "[27, 100.00%] train loss: -3.398829\n",
      "[27, 100.00%] validate loss: -3.758904\n",
      "[28, 100.00%] train loss: -3.552268\n",
      "[28, 100.00%] validate loss: -3.790090\n",
      "[29, 100.00%] train loss: -3.524301\n",
      "[29, 100.00%] validate loss: -3.255094\n",
      "Retry 1/6\n",
      "[30, 100.00%] train loss: -3.540309\n",
      "[30, 100.00%] validate loss: -3.785956\n",
      "Retry 2/6\n",
      "[31, 100.00%] train loss: -3.638659\n",
      "[31, 100.00%] validate loss: -3.898989\n",
      "[32, 100.00%] train loss: -3.598556\n",
      "[32, 100.00%] validate loss: -3.943767\n",
      "[33, 100.00%] train loss: -3.529446\n",
      "[33, 100.00%] validate loss: -3.622451\n",
      "Retry 1/6\n",
      "[34, 100.00%] train loss: -3.718125\n",
      "[34, 100.00%] validate loss: -3.987788\n",
      "[35, 100.00%] train loss: -3.815814\n",
      "[35, 100.00%] validate loss: -4.038739\n",
      "[36, 100.00%] train loss: -3.690167\n",
      "[36, 100.00%] validate loss: -3.981146\n",
      "Retry 1/6\n",
      "[37, 100.00%] train loss: -3.756191\n",
      "[37, 100.00%] validate loss: -4.163666\n",
      "[38, 100.00%] train loss: -3.889328\n",
      "[38, 100.00%] validate loss: -4.128175\n",
      "Retry 1/6\n",
      "[39, 100.00%] train loss: -3.904732\n",
      "[39, 100.00%] validate loss: -3.915198\n",
      "Retry 2/6\n",
      "[40, 100.00%] train loss: -3.907220\n",
      "[40, 100.00%] validate loss: -4.058755\n",
      "Retry 3/6\n",
      "[41, 100.00%] train loss: -3.915555\n",
      "[41, 100.00%] validate loss: -3.560825\n",
      "Retry 4/6\n",
      "[42, 100.00%] train loss: -3.758583\n",
      "[42, 100.00%] validate loss: -3.890535\n",
      "Retry 5/6\n",
      "[43, 100.00%] train loss: -3.935635\n",
      "[43, 100.00%] validate loss: -4.172357\n",
      "[44, 100.00%] train loss: -3.952847\n",
      "[44, 100.00%] validate loss: -4.097226\n",
      "Retry 1/6\n",
      "[45, 100.00%] train loss: -3.694335\n",
      "[45, 100.00%] validate loss: -4.149095\n",
      "Retry 2/6\n",
      "[46, 100.00%] train loss: -3.971671\n",
      "[46, 100.00%] validate loss: -4.268177\n",
      "[47, 100.00%] train loss: -4.017874\n",
      "[47, 100.00%] validate loss: -4.229713\n",
      "Retry 1/6\n",
      "[48, 100.00%] train loss: -3.911349\n",
      "[48, 100.00%] validate loss: -4.046911\n",
      "Retry 2/6\n",
      "[49, 100.00%] train loss: -3.905191\n",
      "[49, 100.00%] validate loss: -4.246181\n",
      "Retry 3/6\n",
      "[50, 100.00%] train loss: -4.025958\n",
      "[50, 100.00%] validate loss: -4.270732\n",
      "[51, 100.00%] train loss: -4.028386\n",
      "[51, 100.00%] validate loss: -4.208905\n",
      "Retry 1/6\n",
      "[52, 100.00%] train loss: -3.891914\n",
      "[52, 100.00%] validate loss: -4.216770\n",
      "Retry 2/6\n",
      "[53, 100.00%] train loss: -4.029258\n",
      "[53, 100.00%] validate loss: -4.219796\n",
      "Retry 3/6\n",
      "[54, 100.00%] train loss: -4.005759\n",
      "[54, 100.00%] validate loss: -4.197976\n",
      "Retry 4/6\n",
      "[55, 100.00%] train loss: -3.941914\n",
      "[55, 100.00%] validate loss: -4.283027\n",
      "[56, 100.00%] train loss: -4.074443\n",
      "[56, 100.00%] validate loss: -4.332702\n",
      "[57, 100.00%] train loss: -3.924525\n",
      "[57, 100.00%] validate loss: -4.274918\n",
      "Retry 1/6\n",
      "[58, 100.00%] train loss: -4.061070\n",
      "[58, 100.00%] validate loss: -4.329870\n",
      "Retry 2/6\n",
      "[59, 100.00%] train loss: -4.109783\n",
      "[59, 100.00%] validate loss: -4.260747\n",
      "Retry 3/6\n",
      "[60, 100.00%] train loss: -3.981717\n",
      "[60, 100.00%] validate loss: -3.924793\n",
      "Retry 4/6\n",
      "[61, 100.00%] train loss: -3.892980\n",
      "[61, 100.00%] validate loss: -4.311236\n",
      "Retry 5/6\n",
      "[62, 100.00%] train loss: -4.046401\n",
      "[62, 100.00%] validate loss: -4.394182\n",
      "[63, 100.00%] train loss: -4.036377\n",
      "[63, 100.00%] validate loss: -4.399926\n",
      "[64, 100.00%] train loss: -4.084434\n",
      "[64, 100.00%] validate loss: -4.413023\n",
      "[65, 100.00%] train loss: -4.036327\n",
      "[65, 100.00%] validate loss: -4.422319\n",
      "[66, 100.00%] train loss: -4.130205\n",
      "[66, 100.00%] validate loss: -4.365760\n",
      "Retry 1/6\n",
      "[67, 100.00%] train loss: -4.090890\n",
      "[67, 100.00%] validate loss: -4.288909\n",
      "Retry 2/6\n",
      "[68, 100.00%] train loss: -4.083261\n",
      "[68, 100.00%] validate loss: -4.307276\n",
      "Retry 3/6\n",
      "[69, 100.00%] train loss: -4.165637\n",
      "[69, 100.00%] validate loss: -4.234450\n",
      "Retry 4/6\n",
      "[70, 100.00%] train loss: -4.099087\n",
      "[70, 100.00%] validate loss: -4.271874\n",
      "Retry 5/6\n",
      "[71, 100.00%] train loss: -3.940357\n",
      "[71, 100.00%] validate loss: -4.247865\n",
      "Retry 6/6\n",
      "CPU times: user 5min 20s, sys: 13 s, total: 5min 33s\n",
      "Wall time: 6min 20s\n"
     ]
    }
   ],
   "source": [
    "%%time\n",
    "net = train_model(net, train_dl, val_dl, criterion, optimizer, \"stock-0.5.pt\", 100, metrics=[], patience=6)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [],
   "source": [
    "test_data = scaled_data[training_data_len - 60:, :]\n",
    "x_test = torch.tensor([test_data[i-60:i, 0] for i in range(60, len(test_data))])\n",
    "y_test = dataset[training_data_len:,:]\n",
    "x_test = x_test.reshape(len(x_test), STEP, 1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "6.474945"
      ]
     },
     "execution_count": 19,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "net.eval()\n",
    "with torch.no_grad():\n",
    "    out_mean, out_log_var = net.sample(10, x_test.transpose(1, 0))\n",
    "    A = aleatoric_uncertainty(out_log_var)\n",
    "    E = epistemic_uncertainty(out_mean)\n",
    "    uncertainties = uncertainty(out_mean, out_log_var)\n",
    "    uncertainties = uncertainties.numpy()\n",
    "    predictions = prediction_avg(out_mean)\n",
    "    predictions = scaler.inverse_transform(predictions.numpy())\n",
    "\n",
    "rmse = np.sqrt(np.mean((predictions-y_test)**2))\n",
    "rmse"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1152x576 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "train = data[:training_data_len]\n",
    "valid = data[training_data_len:]\n",
    "valid.insert(1,\"Predictions\",predictions)\n",
    "valid.insert(2,\"Uncertainties\",uncertainties)\n",
    "\n",
    "plt.figure(figsize=(16, 8))\n",
    "plt.title(\"Model\")\n",
    "plt.xlabel(\"Date\", fontsize=18)\n",
    "plt.ylabel(\"Close Price USD ($)\", fontsize=18)\n",
    "plt.plot(train[\"Close\"])\n",
    "plt.plot(valid[[\"Close\", \"Predictions\"]])\n",
    "plt.legend([\"Train\", \"Val\", \"Predictions\"], loc=\"lower right\")\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "sns.scatterplot(x=np.abs(predictions-y_test).flatten(), y=uncertainties.flatten())\n",
    "plt.xlabel(\"Absolute error\")\n",
    "plt.ylabel(\"Uncertainty\")\n",
    "plt.title(\"Uncertainty against Error\")\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.6.10"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}