Answers/Code/lstm_stock_predictor_fng.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# LSTM Stock Predictor Using Fear and Greed Index\n",
    "\n",
    "In this notebook, you will build and train a custom LSTM RNN that uses a 10 day window of Bitcoin fear and greed index values to predict the 11th day closing price. \n",
    "\n",
    "You will need to:\n",
    "\n",
    "1. Prepare the data for training and testing\n",
    "2. Build and train a custom LSTM RNN\n",
    "3. Evaluate the performance of the model"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Data Preparation\n",
    "\n",
    "In this section, you will need to prepare the training and testing data for the model. The model will use a rolling 10 day window to predict the 11th day closing price.\n",
    "\n",
    "You will need to:\n",
    "1. Use the `window_data` function to generate the X and y values for the model.\n",
    "2. Split the data into 70% training and 30% testing\n",
    "3. Apply the MinMaxScaler to the X and y values\n",
    "4. Reshape the X_train and X_test data for the model. Note: The required input format for the LSTM is:\n",
    "\n",
    "```python\n",
    "reshape((X_train.shape[0], X_train.shape[1], 1))\n",
    "```"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 615,
   "metadata": {},
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "import pandas as pd\n",
    "import hvplot.pandas\n",
    "\n",
    "%matplotlib inline"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 616,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Set the random seed for reproducibility\n",
    "# Note: This is for the homework solution, but it is good practice to comment this out and run multiple experiments to evaluate your model\n",
    "from numpy.random import seed\n",
    "seed(1)\n",
    "from tensorflow import random\n",
    "random.set_seed(2)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 617,
   "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>fng_value</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>date</th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>2019-07-29</th>\n",
       "      <td>19</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2019-07-28</th>\n",
       "      <td>16</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2019-07-27</th>\n",
       "      <td>47</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2019-07-26</th>\n",
       "      <td>24</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2019-07-25</th>\n",
       "      <td>42</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "             fng_value\n",
       "date                  \n",
       "2019-07-29          19\n",
       "2019-07-28          16\n",
       "2019-07-27          47\n",
       "2019-07-26          24\n",
       "2019-07-25          42"
      ]
     },
     "execution_count": 617,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Load the fear and greed sentiment data for Bitcoin\n",
    "df = pd.read_csv('../Data/btc_sentiment.csv', index_col=\"date\", infer_datetime_format=True, parse_dates=True)\n",
    "df = df.drop(columns=\"fng_classification\")\n",
    "df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 618,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Date\n",
       "2019-07-25    9882.429688\n",
       "2019-07-26    9847.450195\n",
       "2019-07-27    9478.320313\n",
       "2019-07-28    9531.769531\n",
       "2019-07-29    9529.889648\n",
       "Name: Close, dtype: float64"
      ]
     },
     "execution_count": 618,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Load the historical closing prices for Bitcoin\n",
    "df2 = pd.read_csv('../Data/btc_historic.csv', index_col=\"Date\", infer_datetime_format=True, parse_dates=True)['Close']\n",
    "df2 = df2.sort_index()\n",
    "df2.tail()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 619,
   "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>fng_value</th>\n",
       "      <th>Close</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>2019-07-25</th>\n",
       "      <td>42</td>\n",
       "      <td>9882.429688</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2019-07-26</th>\n",
       "      <td>24</td>\n",
       "      <td>9847.450195</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2019-07-27</th>\n",
       "      <td>47</td>\n",
       "      <td>9478.320313</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2019-07-28</th>\n",
       "      <td>16</td>\n",
       "      <td>9531.769531</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2019-07-29</th>\n",
       "      <td>19</td>\n",
       "      <td>9529.889648</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "             fng_value        Close\n",
       "2019-07-25          42  9882.429688\n",
       "2019-07-26          24  9847.450195\n",
       "2019-07-27          47  9478.320313\n",
       "2019-07-28          16  9531.769531\n",
       "2019-07-29          19  9529.889648"
      ]
     },
     "execution_count": 619,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Join the data into a single DataFrame\n",
    "df = df.join(df2, how=\"inner\")\n",
    "df.tail()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 620,
   "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>fng_value</th>\n",
       "      <th>Close</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>2018-02-01</th>\n",
       "      <td>30</td>\n",
       "      <td>9114.719727</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2018-02-02</th>\n",
       "      <td>15</td>\n",
       "      <td>8870.820313</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2018-02-03</th>\n",
       "      <td>40</td>\n",
       "      <td>9251.269531</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2018-02-04</th>\n",
       "      <td>24</td>\n",
       "      <td>8218.049805</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2018-02-05</th>\n",
       "      <td>11</td>\n",
       "      <td>6937.080078</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "             fng_value        Close\n",
       "2018-02-01          30  9114.719727\n",
       "2018-02-02          15  8870.820313\n",
       "2018-02-03          40  9251.269531\n",
       "2018-02-04          24  8218.049805\n",
       "2018-02-05          11  6937.080078"
      ]
     },
     "execution_count": 620,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 621,
   "metadata": {},
   "outputs": [],
   "source": [
    "# This function accepts the column number for the features (X) and the target (y)\n",
    "# It chunks the data up with a rolling window of Xt-n to predict Xt\n",
    "# It returns a numpy array of X any y\n",
    "def window_data(df, window, feature_col_number, target_col_number):\n",
    "    X = []\n",
    "    y = []\n",
    "    for i in range(len(df) - window - 1):\n",
    "        features = df.iloc[i:(i + window), feature_col_number]\n",
    "        target = df.iloc[(i + window), target_col_number]\n",
    "        X.append(features)\n",
    "        y.append(target)\n",
    "    return np.array(X), np.array(y).reshape(-1, 1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 622,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Predict Closing Prices using a 10 day window of fear and greed index values and a target of the 11th day closing price\n",
    "# Try a window size anywhere from 1 to 10 and see how the model performance changes\n",
    "window_size = 1\n",
    "#window_size = 2\n",
    "#window_size = 3\n",
    "#window_size = 4\n",
    "#window_size = 5\n",
    "#window_size = 6\n",
    "#window_size = 7\n",
    "#window_size = 8\n",
    "#window_size = 9\n",
    "#window_size = 10\n",
    "\n",
    "# Column index 1 is the `Close` column\n",
    "feature_column = 0\n",
    "target_column = 1\n",
    "X, y = window_data(df, window_size, feature_column, target_column)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 623,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Use 70% of the data for training and the remainder for testing\n",
    "split = int(0.7 * len(X))\n",
    "X_train = X[: split - 1]\n",
    "X_test = X[split:]\n",
    "y_train = y[: split - 1]\n",
    "y_test = y[split:]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 624,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Use MinMaxScaler to scale the data between 0 and 1. \n",
    "from sklearn.preprocessing import MinMaxScaler\n",
    "\n",
    "scaler = MinMaxScaler()\n",
    "scaler.fit(X)\n",
    "X_train = scaler.transform(X_train)\n",
    "X_test = scaler.transform(X_test)\n",
    "scaler.fit(y)\n",
    "y_train = scaler.transform(y_train)\n",
    "y_test = scaler.transform(y_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 625,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "X_train sample values:\n",
      "[[[0.25287356]]] \n",
      "\n",
      "X_test sample values:\n",
      "[[[0.40229885]]]\n"
     ]
    }
   ],
   "source": [
    "# Reshape the features for the model\n",
    "X_train = X_train.reshape((X_train.shape[0], X_train.shape[1], 1))\n",
    "X_test = X_test.reshape((X_test.shape[0], X_test.shape[1], 1))\n",
    "print (f\"X_train sample values:\\n{X_train[:1]} \\n\")\n",
    "print (f\"X_test sample values:\\n{X_test[:1]}\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "---"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Build and Train the LSTM RNN\n",
    "\n",
    "In this section, you will design a custom LSTM RNN and fit (train) it using the training data.\n",
    "\n",
    "You will need to:\n",
    "1. Define the model architecture\n",
    "2. Compile the model\n",
    "3. Fit the model to the training data\n",
    "\n",
    "### Hints:\n",
    "You will want to use the same model architecture and random seed for both notebooks. This is necessary to accurately compare the performance of the FNG model vs the closing price model. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 626,
   "metadata": {},
   "outputs": [],
   "source": [
    "from tensorflow.keras.models import Sequential\n",
    "from tensorflow.keras.layers import LSTM, Dense, Dropout"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 627,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Build the LSTM model. \n",
    "# The return sequences need to be set to True if you are adding additional LSTM layers, but \n",
    "# You don't have to do this for the final layer. \n",
    "model = Sequential()\n",
    "\n",
    "# Initial model setup\n",
    "number_units = 30\n",
    "dropout_fraction = 0.2\n",
    "\n",
    "# Layer 1\n",
    "model.add(LSTM(\n",
    "    units=number_units,\n",
    "    return_sequences=True,\n",
    "    input_shape=(X_train.shape[1], 1))\n",
    "    )\n",
    "model.add(Dropout(dropout_fraction))\n",
    "\n",
    "# Layer 2\n",
    "#model.add(LSTM(units=number_units, activation='sigmoid', return_sequences=True))\n",
    "model.add(LSTM(units=number_units, activation='tanh', return_sequences=True))\n",
    "model.add(Dropout(dropout_fraction))\n",
    "\n",
    "# Layer 3\n",
    "model.add(LSTM(units=number_units))\n",
    "model.add(Dropout(dropout_fraction))\n",
    "\n",
    "# Output layer\n",
    "model.add(Dense(1))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 628,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Compile the model\n",
    "model.compile(optimizer=\"adam\", loss=\"mean_squared_error\") "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 629,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Model: \"sequential_26\"\n",
      "_________________________________________________________________\n",
      "Layer (type)                 Output Shape              Param #   \n",
      "=================================================================\n",
      "lstm_78 (LSTM)               (None, 1, 30)             3840      \n",
      "_________________________________________________________________\n",
      "dropout_78 (Dropout)         (None, 1, 30)             0         \n",
      "_________________________________________________________________\n",
      "lstm_79 (LSTM)               (None, 1, 30)             7320      \n",
      "_________________________________________________________________\n",
      "dropout_79 (Dropout)         (None, 1, 30)             0         \n",
      "_________________________________________________________________\n",
      "lstm_80 (LSTM)               (None, 30)                7320      \n",
      "_________________________________________________________________\n",
      "dropout_80 (Dropout)         (None, 30)                0         \n",
      "_________________________________________________________________\n",
      "dense_26 (Dense)             (None, 1)                 31        \n",
      "=================================================================\n",
      "Total params: 18,511\n",
      "Trainable params: 18,511\n",
      "Non-trainable params: 0\n",
      "_________________________________________________________________\n"
     ]
    }
   ],
   "source": [
    "# Summarize the model\n",
    "model.summary() "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 630,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Creating validation data sets\n",
    "from sklearn.model_selection import train_test_split\n",
    "X_train, X_val, y_train, y_val = train_test_split(X_train, y_train, train_size=0.7, shuffle=False, random_state=2) "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 631,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train on 263 samples, validate on 114 samples\n",
      "Epoch 1/20\n",
      "263/263 [==============================] - 5s 19ms/sample - loss: 0.2154 - val_loss: 0.0187\n",
      "Epoch 2/20\n",
      "263/263 [==============================] - 0s 191us/sample - loss: 0.1921 - val_loss: 0.0147\n",
      "Epoch 3/20\n",
      "263/263 [==============================] - 0s 208us/sample - loss: 0.1662 - val_loss: 0.0121\n",
      "Epoch 4/20\n",
      "263/263 [==============================] - 0s 198us/sample - loss: 0.1372 - val_loss: 0.0127\n",
      "Epoch 5/20\n",
      "263/263 [==============================] - 0s 205us/sample - loss: 0.1032 - val_loss: 0.0201\n",
      "Epoch 6/20\n",
      "263/263 [==============================] - 0s 205us/sample - loss: 0.0695 - val_loss: 0.0407\n",
      "Epoch 7/20\n",
      "263/263 [==============================] - 0s 209us/sample - loss: 0.0392 - val_loss: 0.0811\n",
      "Epoch 8/20\n",
      "263/263 [==============================] - 0s 216us/sample - loss: 0.0220 - val_loss: 0.1272\n",
      "Epoch 9/20\n",
      "263/263 [==============================] - 0s 197us/sample - loss: 0.0177 - val_loss: 0.1408\n",
      "Epoch 10/20\n",
      "263/263 [==============================] - 0s 199us/sample - loss: 0.0166 - val_loss: 0.1258\n",
      "Epoch 11/20\n",
      "263/263 [==============================] - 0s 206us/sample - loss: 0.0153 - val_loss: 0.1134\n",
      "Epoch 12/20\n",
      "263/263 [==============================] - 0s 199us/sample - loss: 0.0157 - val_loss: 0.1106\n",
      "Epoch 13/20\n",
      "263/263 [==============================] - 0s 192us/sample - loss: 0.0155 - val_loss: 0.1134\n",
      "Epoch 14/20\n",
      "263/263 [==============================] - 0s 191us/sample - loss: 0.0161 - val_loss: 0.1177\n",
      "Epoch 15/20\n",
      "263/263 [==============================] - 0s 196us/sample - loss: 0.0160 - val_loss: 0.1187\n",
      "Epoch 16/20\n",
      "263/263 [==============================] - 0s 191us/sample - loss: 0.0165 - val_loss: 0.1173\n",
      "Epoch 17/20\n",
      "263/263 [==============================] - 0s 237us/sample - loss: 0.0140 - val_loss: 0.1144\n",
      "Epoch 18/20\n",
      "263/263 [==============================] - 0s 191us/sample - loss: 0.0140 - val_loss: 0.1136\n",
      "Epoch 19/20\n",
      "263/263 [==============================] - 0s 187us/sample - loss: 0.0165 - val_loss: 0.1151\n",
      "Epoch 20/20\n",
      "263/263 [==============================] - 0s 186us/sample - loss: 0.0151 - val_loss: 0.1152\n"
     ]
    }
   ],
   "source": [
    "# Train the model\n",
    "# Use at least 10 epochs\n",
    "# Do not shuffle the data\n",
    "# Experiement with the batch size, but a smaller batch size is recommended\n",
    "training_history = model.fit(\n",
    "    X_train, \n",
    "    y_train, \n",
    "    validation_data=(X_val, y_val),\n",
    "    epochs=20, \n",
    "    shuffle=False, \n",
    "    batch_size=30, \n",
    "    verbose=1\n",
    "    )"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 632,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train on 263 samples, validate on 114 samples\n",
      "Epoch 1/50\n",
      "263/263 [==============================] - 0s 544us/sample - loss: 0.0168 - val_loss: 0.0922\n",
      "Epoch 2/50\n",
      "263/263 [==============================] - 0s 505us/sample - loss: 0.0195 - val_loss: 0.0970\n",
      "Epoch 3/50\n",
      "263/263 [==============================] - 0s 459us/sample - loss: 0.0169 - val_loss: 0.1011\n",
      "Epoch 4/50\n",
      "263/263 [==============================] - 0s 455us/sample - loss: 0.0188 - val_loss: 0.0942\n",
      "Epoch 5/50\n",
      "263/263 [==============================] - 0s 452us/sample - loss: 0.0186 - val_loss: 0.0941\n",
      "Epoch 6/50\n",
      "263/263 [==============================] - 0s 496us/sample - loss: 0.0185 - val_loss: 0.0945\n",
      "Epoch 7/50\n",
      "263/263 [==============================] - 0s 517us/sample - loss: 0.0186 - val_loss: 0.0948\n",
      "Epoch 8/50\n",
      "263/263 [==============================] - 0s 561us/sample - loss: 0.0194 - val_loss: 0.0998\n",
      "Epoch 9/50\n",
      "263/263 [==============================] - 0s 490us/sample - loss: 0.0173 - val_loss: 0.0916\n",
      "Epoch 10/50\n",
      "263/263 [==============================] - 0s 491us/sample - loss: 0.0177 - val_loss: 0.0962\n",
      "Epoch 11/50\n",
      "263/263 [==============================] - 0s 482us/sample - loss: 0.0153 - val_loss: 0.0988\n",
      "Epoch 12/50\n",
      "263/263 [==============================] - 0s 474us/sample - loss: 0.0151 - val_loss: 0.0933\n",
      "Epoch 13/50\n",
      "263/263 [==============================] - 0s 518us/sample - loss: 0.0162 - val_loss: 0.0939\n",
      "Epoch 14/50\n",
      "263/263 [==============================] - 0s 495us/sample - loss: 0.0165 - val_loss: 0.0900\n",
      "Epoch 15/50\n",
      "263/263 [==============================] - 0s 476us/sample - loss: 0.0156 - val_loss: 0.1004\n",
      "Epoch 16/50\n",
      "263/263 [==============================] - 0s 519us/sample - loss: 0.0166 - val_loss: 0.0969\n",
      "Epoch 17/50\n",
      "263/263 [==============================] - 0s 533us/sample - loss: 0.0163 - val_loss: 0.0970\n",
      "Epoch 18/50\n",
      "263/263 [==============================] - 0s 593us/sample - loss: 0.0153 - val_loss: 0.0943\n",
      "Epoch 19/50\n",
      "263/263 [==============================] - 0s 513us/sample - loss: 0.0171 - val_loss: 0.0927\n",
      "Epoch 20/50\n",
      "263/263 [==============================] - 0s 473us/sample - loss: 0.0156 - val_loss: 0.0937\n",
      "Epoch 21/50\n",
      "263/263 [==============================] - 0s 516us/sample - loss: 0.0153 - val_loss: 0.0965\n",
      "Epoch 22/50\n",
      "263/263 [==============================] - 0s 503us/sample - loss: 0.0146 - val_loss: 0.0903\n",
      "Epoch 23/50\n",
      "263/263 [==============================] - 0s 578us/sample - loss: 0.0151 - val_loss: 0.1004\n",
      "Epoch 24/50\n",
      "263/263 [==============================] - 0s 524us/sample - loss: 0.0141 - val_loss: 0.0945\n",
      "Epoch 25/50\n",
      "263/263 [==============================] - 0s 481us/sample - loss: 0.0141 - val_loss: 0.0983\n",
      "Epoch 26/50\n",
      "263/263 [==============================] - 0s 516us/sample - loss: 0.0141 - val_loss: 0.0933\n",
      "Epoch 27/50\n",
      "263/263 [==============================] - 0s 523us/sample - loss: 0.0159 - val_loss: 0.0952\n",
      "Epoch 28/50\n",
      "263/263 [==============================] - 0s 531us/sample - loss: 0.0144 - val_loss: 0.0981\n",
      "Epoch 29/50\n",
      "263/263 [==============================] - 0s 606us/sample - loss: 0.0149 - val_loss: 0.0951\n",
      "Epoch 30/50\n",
      "263/263 [==============================] - 0s 494us/sample - loss: 0.0147 - val_loss: 0.0930\n",
      "Epoch 31/50\n",
      "263/263 [==============================] - 0s 487us/sample - loss: 0.0161 - val_loss: 0.0905\n",
      "Epoch 32/50\n",
      "263/263 [==============================] - 0s 451us/sample - loss: 0.0152 - val_loss: 0.0977\n",
      "Epoch 33/50\n",
      "263/263 [==============================] - 0s 474us/sample - loss: 0.0144 - val_loss: 0.0920\n",
      "Epoch 34/50\n",
      "263/263 [==============================] - 0s 484us/sample - loss: 0.0149 - val_loss: 0.0910\n",
      "Epoch 35/50\n",
      "263/263 [==============================] - 0s 486us/sample - loss: 0.0138 - val_loss: 0.0976\n",
      "Epoch 36/50\n",
      "263/263 [==============================] - 0s 463us/sample - loss: 0.0143 - val_loss: 0.0926\n",
      "Epoch 37/50\n",
      "263/263 [==============================] - 0s 490us/sample - loss: 0.0149 - val_loss: 0.0931\n",
      "Epoch 38/50\n",
      "263/263 [==============================] - 0s 510us/sample - loss: 0.0161 - val_loss: 0.0914\n",
      "Epoch 39/50\n",
      "263/263 [==============================] - 0s 468us/sample - loss: 0.0135 - val_loss: 0.0955\n",
      "Epoch 40/50\n",
      "263/263 [==============================] - 0s 458us/sample - loss: 0.0144 - val_loss: 0.0985\n",
      "Epoch 41/50\n",
      "263/263 [==============================] - 0s 461us/sample - loss: 0.0136 - val_loss: 0.0926\n",
      "Epoch 42/50\n",
      "263/263 [==============================] - 0s 468us/sample - loss: 0.0145 - val_loss: 0.0969\n",
      "Epoch 43/50\n",
      "263/263 [==============================] - 0s 496us/sample - loss: 0.0144 - val_loss: 0.0964\n",
      "Epoch 44/50\n",
      "263/263 [==============================] - 0s 466us/sample - loss: 0.0138 - val_loss: 0.0943\n",
      "Epoch 45/50\n",
      "263/263 [==============================] - 0s 488us/sample - loss: 0.0136 - val_loss: 0.0937\n",
      "Epoch 46/50\n",
      "263/263 [==============================] - 0s 478us/sample - loss: 0.0143 - val_loss: 0.0977\n",
      "Epoch 47/50\n",
      "263/263 [==============================] - 0s 488us/sample - loss: 0.0145 - val_loss: 0.0981\n",
      "Epoch 48/50\n",
      "263/263 [==============================] - 0s 466us/sample - loss: 0.0145 - val_loss: 0.0965\n",
      "Epoch 49/50\n",
      "263/263 [==============================] - 0s 502us/sample - loss: 0.0147 - val_loss: 0.0949\n",
      "Epoch 50/50\n",
      "263/263 [==============================] - 0s 470us/sample - loss: 0.0134 - val_loss: 0.0954\n"
     ]
    }
   ],
   "source": [
    "training_history = model.fit(\n",
    "    X_train, \n",
    "    y_train, \n",
    "    validation_data=(X_val, y_val),\n",
    "    epochs=50, \n",
    "    shuffle=False, \n",
    "    batch_size=10, \n",
    "    verbose=1) "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "---"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Model Performance\n",
    "\n",
    "In this section, you will evaluate the model using the test data. \n",
    "\n",
    "You will need to:\n",
    "1. Evaluate the model using the `X_test` and `y_test` data.\n",
    "2. Use the X_test data to make predictions\n",
    "3. Create a DataFrame of Real (y_test) vs predicted values. \n",
    "4. Plot the Real vs predicted values as a line chart\n",
    "\n",
    "### Hints\n",
    "Remember to apply the `inverse_transform` function to the predicted and y_test values to recover the actual closing prices."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 633,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      " Evaluate on test data for 1-day rolling bitcoin closing prices:\n",
      "\n",
      " Based on FNG Index, with unshuffled valuation, test loss on LSTM RNN is 0.1152.\n"
     ]
    }
   ],
   "source": [
    "# Evaluate the model on the test data using `evaluate`\n",
    "print(f'\\n Evaluate on test data for {window_size}-day rolling bitcoin closing prices:')\n",
    "results = model.evaluate(X_test, y_test, verbose=0)\n",
    "print(f'\\n Based on FNG Index, with unshuffled valuation, test loss on LSTM RNN is {results:,.4f}.')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 634,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Make some predictions\n",
    "predicted = model.predict(X_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 635,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Recover the original prices instead of the scaled version\n",
    "predicted_prices = scaler.inverse_transform(predicted)\n",
    "real_prices = scaler.inverse_transform(y_test.reshape(-1, 1))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 636,
   "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>Real</th>\n",
       "      <th>Predicted</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>3670.919922</td>\n",
       "      <td>7951.530762</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>3670.919922</td>\n",
       "      <td>7828.386719</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>3912.570068</td>\n",
       "      <td>7648.010254</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>3924.239990</td>\n",
       "      <td>9310.462891</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>3974.050049</td>\n",
       "      <td>9458.948242</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "          Real    Predicted\n",
       "0  3670.919922  7951.530762\n",
       "1  3670.919922  7828.386719\n",
       "2  3912.570068  7648.010254\n",
       "3  3924.239990  9310.462891\n",
       "4  3974.050049  9458.948242"
      ]
     },
     "execution_count": 636,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Create a DataFrame of Real and Predicted values\n",
    "btc_fng = pd.DataFrame({\n",
    "    \"Real\": real_prices.ravel(),\n",
    "    \"Predicted\": predicted_prices.ravel()\n",
    "})\n",
    "btc_fng.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 637,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<Figure size 432x288 with 0 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "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": [
    "# Plot the real vs predicted values as a line chartimport matplotlib\n",
    "import matplotlib.pyplot as plt\n",
    "import matplotlib\n",
    "import matplotlib.pyplot as plt\n",
    "plt.figure()\n",
    "btc_fng.plot(title=f\"{window_size}-Day Rolling Unshuffled Real Vs. FNG Predicted Bitcoin Prices\")\n",
    "plt.savefig('../Images/btc_fng_unshuffled_1day.png')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 638,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<Figure size 432x288 with 0 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "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": [
    "# Plotting loss\n",
    "epochs = 50\n",
    "loss_df = pd.DataFrame(\n",
    "    {\n",
    "        \"Epoch\": range(1, epochs + 1),\n",
    "        \"Train\": training_history.history[\"loss\"],\n",
    "        \"Val\": training_history.history[\"val_loss\"],\n",
    "    }\n",
    ")\n",
    "loss_df.set_index(\"Epoch\", inplace=True)\n",
    "\n",
    "plt.figure()\n",
    "loss_df.plot(title=f\"Loss on LSTM {window_size}-Day Unshuffled Model for FNG Predicted Bitcoin Closing Prices\")\n",
    "plt.savefig('../Images/btc_fng_loss_unshuffled_1day.png')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 639,
   "metadata": {},
   "outputs": [],
   "source": [
    "import h5py\n",
    "\n",
    "model.save('../Models/btc_fng_unshuffled_1day.h5')  # creates a HDF5 file 'my_model.h5'\n",
    "del model  # deletes the existing model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 640,
   "metadata": {},
   "outputs": [],
   "source": [
    "from keras.models import load_model\n",
    "import tensorflow as tf\n",
    "from tensorflow import keras"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 641,
   "metadata": {},
   "outputs": [],
   "source": [
    "# returns a compiled model\n",
    "# identical to the previous one\n",
    "model = tf.keras.models.load_model('../Models/btc_fng_unshuffled_1day.h5')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### One-Day Window\n",
    "_**Shuffled Valuation**_"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### _**Data Preparation**_"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 569,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Predict Closing Prices using a 10 day window of fear and greed index values and a target of the 11th day closing price\n",
    "# Try a window size anywhere from 1 to 10 and see how the model performance changes\n",
    "window_size = 1\n",
    "\n",
    "# Column index 1 is the `Close` column\n",
    "feature_column = 0\n",
    "target_column = 1\n",
    "X, y = window_data(df, window_size, feature_column, target_column)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 570,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Use 70% of the data for training and the remainder for testing\n",
    "split = int(0.7 * len(X))\n",
    "X_train = X[: split - 1]\n",
    "X_test = X[split:]\n",
    "y_train = y[: split - 1]\n",
    "y_test = y[split:]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 571,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Use MinMaxScaler to scale the data between 0 and 1. \n",
    "from sklearn.preprocessing import MinMaxScaler\n",
    "\n",
    "scaler = MinMaxScaler()\n",
    "scaler.fit(X)\n",
    "X_train = scaler.transform(X_train)\n",
    "X_test = scaler.transform(X_test)\n",
    "scaler.fit(y)\n",
    "y_train = scaler.transform(y_train)\n",
    "y_test = scaler.transform(y_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 572,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "X_train sample values:\n",
      "[[[0.25287356]]] \n",
      "\n",
      "X_test sample values:\n",
      "[[[0.40229885]]]\n"
     ]
    }
   ],
   "source": [
    "# Reshape the features for the model\n",
    "X_train = X_train.reshape((X_train.shape[0], X_train.shape[1], 1))\n",
    "X_test = X_test.reshape((X_test.shape[0], X_test.shape[1], 1))\n",
    "print (f\"X_train sample values:\\n{X_train[:1]} \\n\")\n",
    "print (f\"X_test sample values:\\n{X_test[:1]}\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### _**Model Training**_"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 573,
   "metadata": {},
   "outputs": [],
   "source": [
    "from tensorflow.keras.models import Sequential\n",
    "from tensorflow.keras.layers import LSTM, Dense, Dropout"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 574,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Build the LSTM model. \n",
    "# The return sequences need to be set to True if you are adding additional LSTM layers, but \n",
    "# You don't have to do this for the final layer. \n",
    "# Define the LSTM RNN model.\n",
    "\n",
    "model = Sequential()\n",
    "\n",
    "# Initial model setup\n",
    "number_units = 30\n",
    "dropout_fraction = 0.2\n",
    "\n",
    "# Layer 1\n",
    "model.add(LSTM(\n",
    "    units=number_units,\n",
    "    return_sequences=True,\n",
    "    input_shape=(X_train.shape[1], 1))\n",
    "    )\n",
    "model.add(Dropout(dropout_fraction))\n",
    "\n",
    "# Layer 2\n",
    "model.add(LSTM(units=number_units, activation='tanh', return_sequences=True))\n",
    "model.add(Dropout(dropout_fraction))\n",
    "\n",
    "# Layer 3\n",
    "model.add(LSTM(units=number_units))\n",
    "model.add(Dropout(dropout_fraction))\n",
    "\n",
    "# Output layer\n",
    "model.add(Dense(1))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 575,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Compile the model\n",
    "model.compile(optimizer=\"adam\", \n",
    "              loss=\"mean_squared_error\") "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 576,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Model: \"sequential_24\"\n",
      "_________________________________________________________________\n",
      "Layer (type)                 Output Shape              Param #   \n",
      "=================================================================\n",
      "lstm_72 (LSTM)               (None, 1, 30)             3840      \n",
      "_________________________________________________________________\n",
      "dropout_72 (Dropout)         (None, 1, 30)             0         \n",
      "_________________________________________________________________\n",
      "lstm_73 (LSTM)               (None, 1, 30)             7320      \n",
      "_________________________________________________________________\n",
      "dropout_73 (Dropout)         (None, 1, 30)             0         \n",
      "_________________________________________________________________\n",
      "lstm_74 (LSTM)               (None, 30)                7320      \n",
      "_________________________________________________________________\n",
      "dropout_74 (Dropout)         (None, 30)                0         \n",
      "_________________________________________________________________\n",
      "dense_24 (Dense)             (None, 1)                 31        \n",
      "=================================================================\n",
      "Total params: 18,511\n",
      "Trainable params: 18,511\n",
      "Non-trainable params: 0\n",
      "_________________________________________________________________\n"
     ]
    }
   ],
   "source": [
    "# Summarize the model\n",
    "model.summary() "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 577,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Creating validation data sets\n",
    "from sklearn.model_selection import train_test_split\n",
    "X_train, X_val, y_train, y_val = train_test_split(X_train, y_train, train_size=0.7, random_state=2) "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 578,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train on 263 samples, validate on 114 samples\n",
      "Epoch 1/20\n",
      "263/263 [==============================] - 6s 22ms/sample - loss: 0.1512 - val_loss: 0.1487\n",
      "Epoch 2/20\n",
      "263/263 [==============================] - 0s 202us/sample - loss: 0.1300 - val_loss: 0.1260\n",
      "Epoch 3/20\n",
      "263/263 [==============================] - 0s 213us/sample - loss: 0.1074 - val_loss: 0.1018\n",
      "Epoch 4/20\n",
      "263/263 [==============================] - 0s 216us/sample - loss: 0.0846 - val_loss: 0.0768\n",
      "Epoch 5/20\n",
      "263/263 [==============================] - 0s 198us/sample - loss: 0.0613 - val_loss: 0.0540\n",
      "Epoch 6/20\n",
      "263/263 [==============================] - 0s 206us/sample - loss: 0.0449 - val_loss: 0.0395\n",
      "Epoch 7/20\n",
      "263/263 [==============================] - 0s 249us/sample - loss: 0.0379 - val_loss: 0.0362\n",
      "Epoch 8/20\n",
      "263/263 [==============================] - 0s 195us/sample - loss: 0.0387 - val_loss: 0.0359\n",
      "Epoch 9/20\n",
      "263/263 [==============================] - 0s 197us/sample - loss: 0.0406 - val_loss: 0.0361\n",
      "Epoch 10/20\n",
      "263/263 [==============================] - 0s 208us/sample - loss: 0.0384 - val_loss: 0.0364\n",
      "Epoch 11/20\n",
      "263/263 [==============================] - 0s 198us/sample - loss: 0.0383 - val_loss: 0.0360\n",
      "Epoch 12/20\n",
      "263/263 [==============================] - 0s 204us/sample - loss: 0.0383 - val_loss: 0.0351\n",
      "Epoch 13/20\n",
      "263/263 [==============================] - 0s 205us/sample - loss: 0.0375 - val_loss: 0.0345\n",
      "Epoch 14/20\n",
      "263/263 [==============================] - 0s 187us/sample - loss: 0.0364 - val_loss: 0.0339\n",
      "Epoch 15/20\n",
      "263/263 [==============================] - 0s 208us/sample - loss: 0.0377 - val_loss: 0.0336\n",
      "Epoch 16/20\n",
      "263/263 [==============================] - 0s 192us/sample - loss: 0.0367 - val_loss: 0.0331\n",
      "Epoch 17/20\n",
      "263/263 [==============================] - 0s 200us/sample - loss: 0.0360 - val_loss: 0.0328\n",
      "Epoch 18/20\n",
      "263/263 [==============================] - 0s 201us/sample - loss: 0.0357 - val_loss: 0.0323\n",
      "Epoch 19/20\n",
      "263/263 [==============================] - 0s 199us/sample - loss: 0.0354 - val_loss: 0.0313\n",
      "Epoch 20/20\n",
      "263/263 [==============================] - 0s 203us/sample - loss: 0.0364 - val_loss: 0.0310\n"
     ]
    }
   ],
   "source": [
    "# Train the model\n",
    "# Use at least 10 epochs\n",
    "# Do not shuffle the data\n",
    "# Experiement with the batch size, but a smaller batch size is recommended\n",
    "training_history = model.fit(\n",
    "    X_train, \n",
    "    y_train, \n",
    "    validation_data=(X_val, y_val),\n",
    "    epochs=20, \n",
    "    shuffle=False, \n",
    "    batch_size=30, \n",
    "    verbose=1\n",
    "  ) "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 579,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train on 263 samples, validate on 114 samples\n",
      "Epoch 1/50\n",
      "263/263 [==============================] - 0s 534us/sample - loss: 0.0331 - val_loss: 0.0297\n",
      "Epoch 2/50\n",
      "263/263 [==============================] - 0s 541us/sample - loss: 0.0369 - val_loss: 0.0300\n",
      "Epoch 3/50\n",
      "263/263 [==============================] - 0s 492us/sample - loss: 0.0332 - val_loss: 0.0286\n",
      "Epoch 4/50\n",
      "263/263 [==============================] - 0s 487us/sample - loss: 0.0376 - val_loss: 0.0298\n",
      "Epoch 5/50\n",
      "263/263 [==============================] - 0s 503us/sample - loss: 0.0353 - val_loss: 0.0291\n",
      "Epoch 6/50\n",
      "263/263 [==============================] - 0s 472us/sample - loss: 0.0361 - val_loss: 0.0288\n",
      "Epoch 7/50\n",
      "263/263 [==============================] - 0s 479us/sample - loss: 0.0362 - val_loss: 0.0286\n",
      "Epoch 8/50\n",
      "263/263 [==============================] - 0s 489us/sample - loss: 0.0339 - val_loss: 0.0284\n",
      "Epoch 9/50\n",
      "263/263 [==============================] - 0s 495us/sample - loss: 0.0352 - val_loss: 0.0285\n",
      "Epoch 10/50\n",
      "263/263 [==============================] - 0s 508us/sample - loss: 0.0377 - val_loss: 0.0287\n",
      "Epoch 11/50\n",
      "263/263 [==============================] - 0s 547us/sample - loss: 0.0352 - val_loss: 0.0282\n",
      "Epoch 12/50\n",
      "263/263 [==============================] - 0s 534us/sample - loss: 0.0344 - val_loss: 0.0289\n",
      "Epoch 13/50\n",
      "263/263 [==============================] - 0s 486us/sample - loss: 0.0353 - val_loss: 0.0283\n",
      "Epoch 14/50\n",
      "263/263 [==============================] - 0s 624us/sample - loss: 0.0347 - val_loss: 0.0290\n",
      "Epoch 15/50\n",
      "263/263 [==============================] - 0s 528us/sample - loss: 0.0353 - val_loss: 0.0283\n",
      "Epoch 16/50\n",
      "263/263 [==============================] - 0s 570us/sample - loss: 0.0349 - val_loss: 0.0281\n",
      "Epoch 17/50\n",
      "263/263 [==============================] - 0s 520us/sample - loss: 0.0355 - val_loss: 0.0284\n",
      "Epoch 18/50\n",
      "263/263 [==============================] - 0s 508us/sample - loss: 0.0348 - val_loss: 0.0283\n",
      "Epoch 19/50\n",
      "263/263 [==============================] - 0s 529us/sample - loss: 0.0336 - val_loss: 0.0280\n",
      "Epoch 20/50\n",
      "263/263 [==============================] - 0s 523us/sample - loss: 0.0344 - val_loss: 0.0282\n",
      "Epoch 21/50\n",
      "263/263 [==============================] - 0s 480us/sample - loss: 0.0345 - val_loss: 0.0286\n",
      "Epoch 22/50\n",
      "263/263 [==============================] - 0s 482us/sample - loss: 0.0330 - val_loss: 0.0282\n",
      "Epoch 23/50\n",
      "263/263 [==============================] - 0s 456us/sample - loss: 0.0321 - val_loss: 0.0276\n",
      "Epoch 24/50\n",
      "263/263 [==============================] - 0s 444us/sample - loss: 0.0344 - val_loss: 0.0287\n",
      "Epoch 25/50\n",
      "263/263 [==============================] - 0s 836us/sample - loss: 0.0343 - val_loss: 0.0276\n",
      "Epoch 26/50\n",
      "263/263 [==============================] - 0s 522us/sample - loss: 0.0332 - val_loss: 0.0286\n",
      "Epoch 27/50\n",
      "263/263 [==============================] - 0s 482us/sample - loss: 0.0325 - val_loss: 0.0279\n",
      "Epoch 28/50\n",
      "263/263 [==============================] - 0s 455us/sample - loss: 0.0342 - val_loss: 0.0287\n",
      "Epoch 29/50\n",
      "263/263 [==============================] - 0s 476us/sample - loss: 0.0330 - val_loss: 0.0278\n",
      "Epoch 30/50\n",
      "263/263 [==============================] - 0s 519us/sample - loss: 0.0343 - val_loss: 0.0281\n",
      "Epoch 31/50\n",
      "263/263 [==============================] - 0s 474us/sample - loss: 0.0339 - val_loss: 0.0285\n",
      "Epoch 32/50\n",
      "263/263 [==============================] - 0s 481us/sample - loss: 0.0354 - val_loss: 0.0287\n",
      "Epoch 33/50\n",
      "263/263 [==============================] - 0s 541us/sample - loss: 0.0326 - val_loss: 0.0280\n",
      "Epoch 34/50\n",
      "263/263 [==============================] - 0s 520us/sample - loss: 0.0341 - val_loss: 0.0282\n",
      "Epoch 35/50\n",
      "263/263 [==============================] - 0s 496us/sample - loss: 0.0337 - val_loss: 0.0286\n",
      "Epoch 36/50\n",
      "263/263 [==============================] - 0s 578us/sample - loss: 0.0333 - val_loss: 0.0284\n",
      "Epoch 37/50\n",
      "263/263 [==============================] - 0s 523us/sample - loss: 0.0323 - val_loss: 0.0284\n",
      "Epoch 38/50\n",
      "263/263 [==============================] - 0s 482us/sample - loss: 0.0338 - val_loss: 0.0281\n",
      "Epoch 39/50\n",
      "263/263 [==============================] - 0s 475us/sample - loss: 0.0347 - val_loss: 0.0285\n",
      "Epoch 40/50\n",
      "263/263 [==============================] - 0s 450us/sample - loss: 0.0343 - val_loss: 0.0283\n",
      "Epoch 41/50\n",
      "263/263 [==============================] - 0s 475us/sample - loss: 0.0338 - val_loss: 0.0287\n",
      "Epoch 42/50\n",
      "263/263 [==============================] - 0s 492us/sample - loss: 0.0343 - val_loss: 0.0277\n",
      "Epoch 43/50\n",
      "263/263 [==============================] - 0s 484us/sample - loss: 0.0343 - val_loss: 0.0288\n",
      "Epoch 44/50\n",
      "263/263 [==============================] - 0s 487us/sample - loss: 0.0357 - val_loss: 0.0282\n",
      "Epoch 45/50\n",
      "263/263 [==============================] - 0s 539us/sample - loss: 0.0336 - val_loss: 0.0287\n",
      "Epoch 46/50\n",
      "263/263 [==============================] - 0s 488us/sample - loss: 0.0332 - val_loss: 0.0279\n",
      "Epoch 47/50\n",
      "263/263 [==============================] - 0s 491us/sample - loss: 0.0334 - val_loss: 0.0281\n",
      "Epoch 48/50\n",
      "263/263 [==============================] - 0s 491us/sample - loss: 0.0334 - val_loss: 0.0282\n",
      "Epoch 49/50\n",
      "263/263 [==============================] - 0s 503us/sample - loss: 0.0342 - val_loss: 0.0283\n",
      "Epoch 50/50\n",
      "263/263 [==============================] - 0s 523us/sample - loss: 0.0349 - val_loss: 0.0283\n"
     ]
    }
   ],
   "source": [
    "training_history = model.fit(\n",
    "    X_train, \n",
    "    y_train, \n",
    "    validation_data=(X_val, y_val),\n",
    "    epochs=50, \n",
    "    shuffle=False, \n",
    "    batch_size=10, \n",
    "    verbose=1) "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### _**Model Performance**_"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 580,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      " Evaluate on test data for rolling 1-day bitcoin closing prices:\n",
      "\n",
      " Based on FNG Index, test loss on LSTM RNN is 0.1019.\n"
     ]
    }
   ],
   "source": [
    "# Evaluate the model on the test data using `evaluate`\n",
    "print(f'\\n Evaluate on test data for rolling {window_size}-day bitcoin closing prices:')\n",
    "results = model.evaluate(X_test, y_test, verbose=0)\n",
    "print(f'\\n Based on FNG Index, test loss on LSTM RNN is {results:,.4f}.')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 581,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Make some predictions\n",
    "# Make predictions using the testing data X_test\n",
    "predicted = model.predict(X_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 582,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Recover the original prices instead of the scaled version\n",
    "predicted_prices = scaler.inverse_transform(predicted)\n",
    "real_prices = scaler.inverse_transform(y_test.reshape(-1, 1))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 583,
   "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>Real</th>\n",
       "      <th>Predicted</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>3670.919922</td>\n",
       "      <td>7237.300293</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>3670.919922</td>\n",
       "      <td>7088.113770</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>3912.570068</td>\n",
       "      <td>6873.434570</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>3924.239990</td>\n",
       "      <td>9024.020508</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>3974.050049</td>\n",
       "      <td>9234.712891</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "          Real    Predicted\n",
       "0  3670.919922  7237.300293\n",
       "1  3670.919922  7088.113770\n",
       "2  3912.570068  6873.434570\n",
       "3  3924.239990  9024.020508\n",
       "4  3974.050049  9234.712891"
      ]
     },
     "execution_count": 583,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Create a DataFrame of Real and Predicted values\n",
    "btc_fng = pd.DataFrame({\n",
    "    \"Real\": real_prices.ravel(),\n",
    "    \"Predicted\": predicted_prices.ravel()\n",
    "})\n",
    "btc_fng.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 584,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<Figure size 432x288 with 0 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "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": [
    "# Plot the real vs predicted values as a line chart\n",
    "import matplotlib.pyplot as plt\n",
    "import matplotlib\n",
    "import matplotlib.pyplot as plt\n",
    "plt.figure()\n",
    "btc_fng.plot(title=f\"{window_size}-Day Rolling Real vs. FNG Predicted Bitcoin Prices\")\n",
    "plt.savefig('../Images/btc_fng_1day.png')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 585,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<Figure size 432x288 with 0 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "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": [
    "# Plotting loss\n",
    "epochs = 50\n",
    "loss_df = pd.DataFrame(\n",
    "    {\n",
    "        \"Epoch\": range(1, epochs + 1),\n",
    "        \"Train\": training_history.history[\"loss\"],\n",
    "        \"Val\": training_history.history[\"val_loss\"],\n",
    "    }\n",
    ")\n",
    "loss_df.set_index(\"Epoch\", inplace=True)\n",
    "\n",
    "plt.figure()\n",
    "loss_df.plot(title=f\"Loss on LSTM {window_size}-Day Model for FNG Predicted Bitcoin Closing Prices\")\n",
    "plt.savefig('../Images/btc_fng_loss_1day.png')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### _**Save the Model**_"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 586,
   "metadata": {},
   "outputs": [],
   "source": [
    "import h5py\n",
    "\n",
    "model.save('../Models/btc_fng_1day.h5')  # creates a HDF5 file 'my_model.h5'\n",
    "del model  # deletes the existing model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 587,
   "metadata": {},
   "outputs": [],
   "source": [
    "from keras.models import load_model\n",
    "import tensorflow as tf\n",
    "from tensorflow import keras"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 588,
   "metadata": {},
   "outputs": [],
   "source": [
    "# returns a compiled model\n",
    "# identical to the previous one\n",
    "model = tf.keras.models.load_model('../Models/btc_fng_1day.h5')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### One-Day Window\n",
    "_**Default Valuation**_"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 658,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Predict Closing Prices using a 10 day window of fear and greed index values and a target of the 11th day closing price\n",
    "# Try a window size anywhere from 1 to 10 and see how the model performance changes\n",
    "window_size = 1\n",
    "\n",
    "# Column index 1 is the `Close` column\n",
    "feature_column = 0\n",
    "target_column = 1\n",
    "X, y = window_data(df, window_size, feature_column, target_column)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 659,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Use 70% of the data for training and the remainder for testing\n",
    "split = int(0.7 * len(X))\n",
    "X_train = X[: split - 1]\n",
    "X_test = X[split:]\n",
    "y_train = y[: split - 1]\n",
    "y_test = y[split:]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 660,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Use MinMaxScaler to scale the data between 0 and 1. \n",
    "from sklearn.preprocessing import MinMaxScaler\n",
    "\n",
    "scaler = MinMaxScaler()\n",
    "scaler.fit(X)\n",
    "X_train = scaler.transform(X_train)\n",
    "X_test = scaler.transform(X_test)\n",
    "scaler.fit(y)\n",
    "y_train = scaler.transform(y_train)\n",
    "y_test = scaler.transform(y_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 661,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "X_train sample values:\n",
      "[[[0.25287356]]] \n",
      "\n",
      "X_test sample values:\n",
      "[[[0.40229885]]]\n"
     ]
    }
   ],
   "source": [
    "# Reshape the features for the model\n",
    "X_train = X_train.reshape((X_train.shape[0], X_train.shape[1], 1))\n",
    "X_test = X_test.reshape((X_test.shape[0], X_test.shape[1], 1))\n",
    "print (f\"X_train sample values:\\n{X_train[:1]} \\n\")\n",
    "print (f\"X_test sample values:\\n{X_test[:1]}\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 662,
   "metadata": {},
   "outputs": [],
   "source": [
    "from tensorflow.keras.models import Sequential\n",
    "from tensorflow.keras.layers import LSTM, Dense, Dropout"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 663,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Build the LSTM model. \n",
    "# The return sequences need to be set to True if you are adding additional LSTM layers, but \n",
    "# You don't have to do this for the final layer. \n",
    "# Define the LSTM RNN model.\n",
    "\n",
    "model = Sequential()\n",
    "\n",
    "# Initial model setup\n",
    "number_units = 30\n",
    "dropout_fraction = 0.2\n",
    "\n",
    "# Layer 1\n",
    "model.add(LSTM(\n",
    "    units=number_units,\n",
    "    return_sequences=True,\n",
    "    input_shape=(X_train.shape[1], 1))\n",
    "    )\n",
    "model.add(Dropout(dropout_fraction))\n",
    "\n",
    "# Layer 2\n",
    "model.add(LSTM(units=number_units, activation='tanh', return_sequences=True))\n",
    "model.add(Dropout(dropout_fraction))\n",
    "\n",
    "# Layer 3\n",
    "model.add(LSTM(units=number_units))\n",
    "model.add(Dropout(dropout_fraction))\n",
    "\n",
    "# Output layer\n",
    "model.add(Dense(1))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 664,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Compile the model\n",
    "model.compile(optimizer=\"adam\", \n",
    "              loss=\"mean_squared_error\") "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 665,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Model: \"sequential_28\"\n",
      "_________________________________________________________________\n",
      "Layer (type)                 Output Shape              Param #   \n",
      "=================================================================\n",
      "lstm_84 (LSTM)               (None, 1, 30)             3840      \n",
      "_________________________________________________________________\n",
      "dropout_84 (Dropout)         (None, 1, 30)             0         \n",
      "_________________________________________________________________\n",
      "lstm_85 (LSTM)               (None, 1, 30)             7320      \n",
      "_________________________________________________________________\n",
      "dropout_85 (Dropout)         (None, 1, 30)             0         \n",
      "_________________________________________________________________\n",
      "lstm_86 (LSTM)               (None, 30)                7320      \n",
      "_________________________________________________________________\n",
      "dropout_86 (Dropout)         (None, 30)                0         \n",
      "_________________________________________________________________\n",
      "dense_28 (Dense)             (None, 1)                 31        \n",
      "=================================================================\n",
      "Total params: 18,511\n",
      "Trainable params: 18,511\n",
      "Non-trainable params: 0\n",
      "_________________________________________________________________\n"
     ]
    }
   ],
   "source": [
    "# Summarize the model\n",
    "model.summary() "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 666,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train on 377 samples\n",
      "Epoch 1/20\n",
      "377/377 [==============================] - 3s 9ms/sample - loss: 0.1557\n",
      "Epoch 2/20\n",
      "377/377 [==============================] - 0s 176us/sample - loss: 0.1327\n",
      "Epoch 3/20\n",
      "377/377 [==============================] - 0s 167us/sample - loss: 0.1086\n",
      "Epoch 4/20\n",
      "377/377 [==============================] - 0s 176us/sample - loss: 0.0843\n",
      "Epoch 5/20\n",
      "377/377 [==============================] - 0s 175us/sample - loss: 0.0633\n",
      "Epoch 6/20\n",
      "377/377 [==============================] - 0s 175us/sample - loss: 0.0487\n",
      "Epoch 7/20\n",
      "377/377 [==============================] - 0s 173us/sample - loss: 0.0442\n",
      "Epoch 8/20\n",
      "377/377 [==============================] - 0s 197us/sample - loss: 0.0403\n",
      "Epoch 9/20\n",
      "377/377 [==============================] - 0s 155us/sample - loss: 0.0417\n",
      "Epoch 10/20\n",
      "377/377 [==============================] - 0s 164us/sample - loss: 0.0388\n",
      "Epoch 11/20\n",
      "377/377 [==============================] - 0s 168us/sample - loss: 0.0403\n",
      "Epoch 12/20\n",
      "377/377 [==============================] - 0s 151us/sample - loss: 0.0390\n",
      "Epoch 13/20\n",
      "377/377 [==============================] - 0s 154us/sample - loss: 0.0385\n",
      "Epoch 14/20\n",
      "377/377 [==============================] - 0s 147us/sample - loss: 0.0387\n",
      "Epoch 15/20\n",
      "377/377 [==============================] - 0s 149us/sample - loss: 0.0395\n",
      "Epoch 16/20\n",
      "377/377 [==============================] - 0s 143us/sample - loss: 0.0383\n",
      "Epoch 17/20\n",
      "377/377 [==============================] - 0s 144us/sample - loss: 0.0380\n",
      "Epoch 18/20\n",
      "377/377 [==============================] - 0s 145us/sample - loss: 0.0377\n",
      "Epoch 19/20\n",
      "377/377 [==============================] - 0s 146us/sample - loss: 0.0382\n",
      "Epoch 20/20\n",
      "377/377 [==============================] - 0s 146us/sample - loss: 0.0381\n"
     ]
    }
   ],
   "source": [
    "# Train the model\n",
    "# Use at least 10 epochs\n",
    "# Do not shuffle the data\n",
    "# Experiement with the batch size, but a smaller batch size is recommended\n",
    "training_history = model.fit(\n",
    "    X_train, \n",
    "    y_train, \n",
    "    #validation_data=(X_val, y_val),\n",
    "    epochs=20, \n",
    "    shuffle=False, \n",
    "    batch_size=30, \n",
    "    verbose=1\n",
    "  ) "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 667,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train on 377 samples\n",
      "Epoch 1/50\n",
      "377/377 [==============================] - 0s 441us/sample - loss: 0.0402\n",
      "Epoch 2/50\n",
      "377/377 [==============================] - 0s 414us/sample - loss: 0.0451\n",
      "Epoch 3/50\n",
      "377/377 [==============================] - 0s 417us/sample - loss: 0.0420\n",
      "Epoch 4/50\n",
      "377/377 [==============================] - 0s 400us/sample - loss: 0.0430\n",
      "Epoch 5/50\n",
      "377/377 [==============================] - 0s 412us/sample - loss: 0.0435\n",
      "Epoch 6/50\n",
      "377/377 [==============================] - 0s 402us/sample - loss: 0.0426\n",
      "Epoch 7/50\n",
      "377/377 [==============================] - 0s 403us/sample - loss: 0.0429\n",
      "Epoch 8/50\n",
      "377/377 [==============================] - 0s 403us/sample - loss: 0.0426\n",
      "Epoch 9/50\n",
      "377/377 [==============================] - 0s 412us/sample - loss: 0.0431\n",
      "Epoch 10/50\n",
      "377/377 [==============================] - 0s 397us/sample - loss: 0.0396\n",
      "Epoch 11/50\n",
      "377/377 [==============================] - 0s 392us/sample - loss: 0.0400\n",
      "Epoch 12/50\n",
      "377/377 [==============================] - 0s 394us/sample - loss: 0.0396\n",
      "Epoch 13/50\n",
      "377/377 [==============================] - 0s 406us/sample - loss: 0.0395\n",
      "Epoch 14/50\n",
      "377/377 [==============================] - 0s 397us/sample - loss: 0.0415\n",
      "Epoch 15/50\n",
      "377/377 [==============================] - 0s 411us/sample - loss: 0.0390\n",
      "Epoch 16/50\n",
      "377/377 [==============================] - 0s 404us/sample - loss: 0.0390\n",
      "Epoch 17/50\n",
      "377/377 [==============================] - 0s 396us/sample - loss: 0.0400\n",
      "Epoch 18/50\n",
      "377/377 [==============================] - 0s 405us/sample - loss: 0.0407\n",
      "Epoch 19/50\n",
      "377/377 [==============================] - 0s 405us/sample - loss: 0.0404\n",
      "Epoch 20/50\n",
      "377/377 [==============================] - 0s 419us/sample - loss: 0.0386\n",
      "Epoch 21/50\n",
      "377/377 [==============================] - 0s 402us/sample - loss: 0.0390\n",
      "Epoch 22/50\n",
      "377/377 [==============================] - 0s 426us/sample - loss: 0.0398\n",
      "Epoch 23/50\n",
      "377/377 [==============================] - 0s 451us/sample - loss: 0.0381\n",
      "Epoch 24/50\n",
      "377/377 [==============================] - 0s 388us/sample - loss: 0.0391\n",
      "Epoch 25/50\n",
      "377/377 [==============================] - 0s 386us/sample - loss: 0.0394\n",
      "Epoch 26/50\n",
      "377/377 [==============================] - 0s 390us/sample - loss: 0.0389\n",
      "Epoch 27/50\n",
      "377/377 [==============================] - 0s 396us/sample - loss: 0.0383\n",
      "Epoch 28/50\n",
      "377/377 [==============================] - 0s 445us/sample - loss: 0.0380\n",
      "Epoch 29/50\n",
      "377/377 [==============================] - 0s 398us/sample - loss: 0.0388\n",
      "Epoch 30/50\n",
      "377/377 [==============================] - 0s 398us/sample - loss: 0.0388\n",
      "Epoch 31/50\n",
      "377/377 [==============================] - 0s 401us/sample - loss: 0.0388\n",
      "Epoch 32/50\n",
      "377/377 [==============================] - 0s 392us/sample - loss: 0.0374\n",
      "Epoch 33/50\n",
      "377/377 [==============================] - 0s 402us/sample - loss: 0.0376\n",
      "Epoch 34/50\n",
      "377/377 [==============================] - 0s 438us/sample - loss: 0.0386\n",
      "Epoch 35/50\n",
      "377/377 [==============================] - 0s 407us/sample - loss: 0.0378\n",
      "Epoch 36/50\n",
      "377/377 [==============================] - 0s 392us/sample - loss: 0.0392\n",
      "Epoch 37/50\n",
      "377/377 [==============================] - 0s 377us/sample - loss: 0.0369\n",
      "Epoch 38/50\n",
      "377/377 [==============================] - 0s 382us/sample - loss: 0.0365\n",
      "Epoch 39/50\n",
      "377/377 [==============================] - 0s 397us/sample - loss: 0.0379\n",
      "Epoch 40/50\n",
      "377/377 [==============================] - 0s 389us/sample - loss: 0.0373\n",
      "Epoch 41/50\n",
      "377/377 [==============================] - 0s 398us/sample - loss: 0.0377\n",
      "Epoch 42/50\n",
      "377/377 [==============================] - 0s 403us/sample - loss: 0.0371\n",
      "Epoch 43/50\n",
      "377/377 [==============================] - 0s 385us/sample - loss: 0.0368\n",
      "Epoch 44/50\n",
      "377/377 [==============================] - 0s 387us/sample - loss: 0.0372\n",
      "Epoch 45/50\n",
      "377/377 [==============================] - 0s 408us/sample - loss: 0.0381\n",
      "Epoch 46/50\n",
      "377/377 [==============================] - 0s 395us/sample - loss: 0.0377\n",
      "Epoch 47/50\n",
      "377/377 [==============================] - 0s 507us/sample - loss: 0.0374\n",
      "Epoch 48/50\n",
      "377/377 [==============================] - 0s 404us/sample - loss: 0.0377\n",
      "Epoch 49/50\n",
      "377/377 [==============================] - 0s 431us/sample - loss: 0.0373\n",
      "Epoch 50/50\n",
      "377/377 [==============================] - 0s 410us/sample - loss: 0.0369\n"
     ]
    }
   ],
   "source": [
    "training_history = model.fit(\n",
    "    X_train, \n",
    "    y_train, \n",
    "#    validation_data=(X_val, y_val),\n",
    "    epochs=50, \n",
    "    shuffle=False, \n",
    "    batch_size=10, \n",
    "    verbose=1) "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 668,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      " Evaluate on test data for rolling 1-day bitcoin closing prices:\n",
      "\n",
      " Based on FNG Index, without evaluation, test loss on LSTM RNN is 0.0904.\n"
     ]
    }
   ],
   "source": [
    "# Evaluate the model on the test data using `evaluate`\n",
    "print(f'\\n Evaluate on test data for rolling {window_size}-day bitcoin closing prices:')\n",
    "results = model.evaluate(X_test, y_test, verbose=0)\n",
    "print(f'\\n Based on FNG Index, without evaluation, test loss on LSTM RNN is {results:,.4f}.')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 669,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Make some predictions\n",
    "# Make predictions using the testing data X_test\n",
    "predicted = model.predict(X_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 670,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Recover the original prices instead of the scaled version\n",
    "predicted_prices = scaler.inverse_transform(predicted)\n",
    "real_prices = scaler.inverse_transform(y_test.reshape(-1, 1))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 671,
   "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>Real</th>\n",
       "      <th>Predicted</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>3670.919922</td>\n",
       "      <td>7085.916992</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>3670.919922</td>\n",
       "      <td>6946.062988</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>3912.570068</td>\n",
       "      <td>6743.067871</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>3924.239990</td>\n",
       "      <td>8687.127930</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>3974.050049</td>\n",
       "      <td>8867.420898</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "          Real    Predicted\n",
       "0  3670.919922  7085.916992\n",
       "1  3670.919922  6946.062988\n",
       "2  3912.570068  6743.067871\n",
       "3  3924.239990  8687.127930\n",
       "4  3974.050049  8867.420898"
      ]
     },
     "execution_count": 671,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Create a DataFrame of Real and Predicted values\n",
    "btc_fng = pd.DataFrame({\n",
    "    \"Real\": real_prices.ravel(),\n",
    "    \"Predicted\": predicted_prices.ravel()\n",
    "})\n",
    "btc_fng.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 672,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<Figure size 432x288 with 0 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "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": [
    "# Plot the real vs predicted values as a line chart\n",
    "import matplotlib.pyplot as plt\n",
    "import matplotlib\n",
    "import matplotlib.pyplot as plt\n",
    "plt.figure()\n",
    "btc_fng.plot(title=f\"{window_size}-Day Rolling Real vs. FNG Originally Predicted Bitcoin Prices\")\n",
    "plt.savefig('../Images/btc_fng_original_1day.png')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 675,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<Figure size 432x288 with 0 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "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": [
    "# Plotting loss\n",
    "epochs = 50\n",
    "loss_df = pd.DataFrame(\n",
    "    {\n",
    "        \"Epoch\": range(1, epochs + 1),\n",
    "        \"Train\": training_history.history[\"loss\"],\n",
    "#        \"Val\": training_history.history[\"val_loss\"],\n",
    "    }\n",
    ")\n",
    "loss_df.set_index(\"Epoch\", inplace=True)\n",
    "\n",
    "plt.figure()\n",
    "loss_df.plot(title=f\"Loss on LSTM {window_size}-Day Model for FNG Originally Predicted Bitcoin Closing Prices\")\n",
    "plt.savefig('../Images/btc_fng_loss_original_1day.png')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "file_extension": ".py",
  "kernelspec": {
   "display_name": "mlenv",
   "language": "python",
   "name": "mlenv"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.7.6"
  },
  "mimetype": "text/x-python",
  "name": "python",
  "npconvert_exporter": "python",
  "pygments_lexer": "ipython3",
  "version": 3
 },
 "nbformat": 4,
 "nbformat_minor": 4
}