CAD-USD_PricePart_2_Implementing_Forecast_Models.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "id": "ec5932c2",
   "metadata": {},
   "source": [
    "# Forecasting the Price & Trend of CAD-USD Exchange based on historical Data "
   ]
  },
  {
   "cell_type": "markdown",
   "id": "7177a377",
   "metadata": {},
   "source": [
    "In the previous notebook (\"CAD-USD Price - Part 1 - Analysis\"), I cleaned & analysed historic data (Nov 21, 2017 to Nov 21, 2022) on CAD-USD price. The analysis involved ETS Decomposition as well as Holt-Winters methods to identify if there is any Trend or seasonality in the data.\n",
    "\n",
    "In this section, I will use the same dataset to predict future prices using a variety of Forecasting Models. At the end, the performance of each of the models will be evaluated. "
   ]
  },
  {
   "cell_type": "markdown",
   "id": "d54e1e92",
   "metadata": {},
   "source": [
    "### Goal\n",
    "\n",
    "Since I am dealing with exchange prices, this can be very tricky to predict as it's dependent on several other factors apart from time. \n",
    "So I will attempt the forecast the price & the future trend "
   ]
  },
  {
   "cell_type": "markdown",
   "id": "f023dfb8",
   "metadata": {},
   "source": [
    "### Importing Libraries"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "158ccfd2",
   "metadata": {},
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "import pandas as pd\n",
    "from matplotlib import pyplot as plt\n",
    "from matplotlib import dates \n",
    "from statsmodels.tsa.filters.hp_filter import hpfilter\n",
    "from statsmodels.tsa.seasonal import seasonal_decompose\n",
    "from statsmodels.tsa.holtwinters import ExponentialSmoothing\n",
    "from sklearn.metrics import mean_squared_error, mean_absolute_error"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "65496c14",
   "metadata": {},
   "source": [
    "### Dataset"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "45b6e920",
   "metadata": {},
   "source": [
    "I will use the data on daily CAD/USD conversion rates from November 21, 2017 - November 21, 2022 from Yahoo finance\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 660,
   "id": "6edf6a38",
   "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>Date</th>\n",
       "      <th>Open</th>\n",
       "      <th>High</th>\n",
       "      <th>Low</th>\n",
       "      <th>Close</th>\n",
       "      <th>Adj Close</th>\n",
       "      <th>Volume</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>2017-11-22</td>\n",
       "      <td>0.782963</td>\n",
       "      <td>0.786411</td>\n",
       "      <td>0.782791</td>\n",
       "      <td>0.782840</td>\n",
       "      <td>0.782840</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>2017-11-23</td>\n",
       "      <td>0.786999</td>\n",
       "      <td>0.789204</td>\n",
       "      <td>0.785960</td>\n",
       "      <td>0.787092</td>\n",
       "      <td>0.787092</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>2017-11-24</td>\n",
       "      <td>0.786596</td>\n",
       "      <td>0.787836</td>\n",
       "      <td>0.784560</td>\n",
       "      <td>0.786621</td>\n",
       "      <td>0.786621</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>2017-11-27</td>\n",
       "      <td>0.786559</td>\n",
       "      <td>0.788706</td>\n",
       "      <td>0.784252</td>\n",
       "      <td>0.786535</td>\n",
       "      <td>0.786535</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>2017-11-28</td>\n",
       "      <td>0.783816</td>\n",
       "      <td>0.784191</td>\n",
       "      <td>0.779818</td>\n",
       "      <td>0.783797</td>\n",
       "      <td>0.783797</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "         Date      Open      High       Low     Close  Adj Close  Volume\n",
       "0  2017-11-22  0.782963  0.786411  0.782791  0.782840   0.782840     0.0\n",
       "1  2017-11-23  0.786999  0.789204  0.785960  0.787092   0.787092     0.0\n",
       "2  2017-11-24  0.786596  0.787836  0.784560  0.786621   0.786621     0.0\n",
       "3  2017-11-27  0.786559  0.788706  0.784252  0.786535   0.786535     0.0\n",
       "4  2017-11-28  0.783816  0.784191  0.779818  0.783797   0.783797     0.0"
      ]
     },
     "execution_count": 660,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data = pd.read_csv(r'C:\\Users\\olalawal\\Downloads\\CADUSD.csv')\n",
    "data.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "c71cf2d4",
   "metadata": {},
   "source": [
    "### Data Cleaning"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "ff8034fe",
   "metadata": {},
   "source": [
    "From the analysis in Part 1, we already discovered there were some unwanted columns & NAN entries, so we will remove them"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 661,
   "id": "56fb249f",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "DatetimeIndex(['2017-11-22', '2017-11-23', '2017-11-24', '2017-11-27',\n",
       "               '2017-11-28', '2017-11-29', '2017-11-30', '2017-12-01',\n",
       "               '2017-12-04', '2017-12-05',\n",
       "               ...\n",
       "               '2022-11-09', '2022-11-10', '2022-11-11', '2022-11-14',\n",
       "               '2022-11-15', '2022-11-16', '2022-11-17', '2022-11-18',\n",
       "               '2022-11-21', '2022-11-22'],\n",
       "              dtype='datetime64[ns]', name='Date', length=1305, freq='B')"
      ]
     },
     "execution_count": 661,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data['Date'] = pd.to_datetime(data['Date'])\n",
    "\n",
    "data.set_index('Date', drop=True, inplace=True)\n",
    "# We will use a business day frequency since we only have prices from Monday to Friday\n",
    "data.index.freq='B'\n",
    "data.index"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 662,
   "id": "ac4e574c",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(1304, 4)"
      ]
     },
     "execution_count": 662,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data.drop(labels=['Adj Close', 'Volume'], axis = 1, inplace=True)\n",
    "data.dropna(axis =0, how = 'any', inplace=True)\n",
    "data.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 663,
   "id": "24b6a9cd",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "DatetimeIndex(['2017-11-22', '2017-11-23', '2017-11-24', '2017-11-27',\n",
       "               '2017-11-28', '2017-11-29', '2017-11-30', '2017-12-01',\n",
       "               '2017-12-04', '2017-12-05',\n",
       "               ...\n",
       "               '2022-11-09', '2022-11-10', '2022-11-11', '2022-11-14',\n",
       "               '2022-11-15', '2022-11-16', '2022-11-17', '2022-11-18',\n",
       "               '2022-11-21', '2022-11-22'],\n",
       "              dtype='datetime64[ns]', name='Date', length=1304, freq=None)"
      ]
     },
     "execution_count": 663,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data.index"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 664,
   "id": "c9fe9ac4",
   "metadata": {},
   "outputs": [],
   "source": [
    "#The freq has been reverted to None. Dont know how to deal with this \n",
    "data.index=data.index.to_period('B')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 665,
   "id": "f67656b3",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "PeriodIndex(['2017-11-22', '2017-11-23', '2017-11-24', '2017-11-27',\n",
       "             '2017-11-28', '2017-11-29', '2017-11-30', '2017-12-01',\n",
       "             '2017-12-04', '2017-12-05',\n",
       "             ...\n",
       "             '2022-11-09', '2022-11-10', '2022-11-11', '2022-11-14',\n",
       "             '2022-11-15', '2022-11-16', '2022-11-17', '2022-11-18',\n",
       "             '2022-11-21', '2022-11-22'],\n",
       "            dtype='period[B]', name='Date', length=1304)"
      ]
     },
     "execution_count": 665,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data.index"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "07f18e2c",
   "metadata": {},
   "source": [
    "### Train/Test Split"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "93c9f28e",
   "metadata": {},
   "source": [
    "I want the model to be able to forecast prices within 6 months into the future, so my test set should have data for at least 6 months after the train set.\n",
    "so for the train set: I will select approximately the last 9 months in the dataset (20 business days in a month for 9 months is 180)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 666,
   "id": "144d39fd",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<BusinessDay>"
      ]
     },
     "execution_count": 666,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "train_set = data[:1124]\n",
    "test_set = data[1124:]\n",
    "test_set.index.freq "
   ]
  },
  {
   "cell_type": "markdown",
   "id": "b7700157",
   "metadata": {},
   "source": [
    "### Implementing different Models"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "6fbf405f",
   "metadata": {},
   "source": [
    "### Model 1 - Holt-Winters Triple Exponential Smoothing "
   ]
  },
  {
   "cell_type": "markdown",
   "id": "cb221864",
   "metadata": {},
   "source": [
    "The Holt-Winters Triple ES can be used to model data that has both trends & seasonality\n",
    "\n",
    "From the analysis in part 1, I saw that the data has trends and a little bit of seasonality (although, not a lot)\n",
    "\n",
    "After carefully analysing the seasonality plot from the ETS decomposition, I'll say the data has some monthly seasonality\n",
    "\n",
    "However I will use 3 different values for seasonal periods:\n",
    "- 5 business days for weekly cycle\n",
    "- 20 business days in a monthly cycle\n",
    "- 260 for year"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 667,
   "id": "2674fba1",
   "metadata": {},
   "outputs": [],
   "source": [
    "#Using a seasonal period of 260 (5 business days in a week & 52 weeks in a year)\n",
    "\n",
    "\n",
    "fitted_model_weekly = ExponentialSmoothing(train_set['Close'], trend ='add', seasonal = 'add', seasonal_periods=5).fit()\n",
    "fitted_model_monthly = ExponentialSmoothing(train_set['Close'], trend ='add', seasonal = 'add', seasonal_periods=20).fit()\n",
    "fitted_model_annual = ExponentialSmoothing(train_set['Close'], trend ='add', seasonal = 'add', seasonal_periods=260).fit()\n",
    "predicted_prices_week = fitted_model_weekly.forecast(180) #Using 180 because thats the amount of data points that we want to forecast (just like the data set )\n",
    "predicted_prices_month = fitted_model_monthly.forecast(180)\n",
    "predicted_prices_annual = fitted_model_annual.forecast(180)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "e4e16bd2",
   "metadata": {},
   "source": [
    "#### Evaluating Model 1 - Holt-Winters Triple Exponential Smoothing "
   ]
  },
  {
   "cell_type": "markdown",
   "id": "3bcd646d",
   "metadata": {},
   "source": [
    "Now let's see how the model performed by plotting the result & calculating the RMSE for each model "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 668,
   "id": "5cac6026",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<AxesSubplot:xlabel='Date'>"
      ]
     },
     "execution_count": 668,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 864x360 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "train_set['Close'].plot(legend = True, label = 'TRAIN SET', figsize = (12,5))\n",
    "test_set['Close'].plot(legend = True, label = 'TEST SET')\n",
    "predicted_prices_week.plot(legend = True, label = 'Predicted Prices - Weekly Seasonality')\n",
    "predicted_prices_month.plot(legend = True, label = 'Predicted Prices - Monthly Seasonality')\n",
    "predicted_prices_annual.plot(legend = True, label = 'Predicted Prices - Annual Seasonality')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 669,
   "id": "72035e74",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<AxesSubplot:xlabel='Date'>"
      ]
     },
     "execution_count": 669,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "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": [
    "test_set['Close'].plot(legend = True, label = 'TEST SET')\n",
    "predicted_prices_week.plot(legend = True, label = 'Predicted Prices - Weekly Seasonality')\n",
    "predicted_prices_month.plot(legend = True, label = 'Predicted Prices - Monthly Seasonality')\n",
    "predicted_prices_annual.plot(legend = True, label = 'Predicted Prices - Annual Seasonality')"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "42329422",
   "metadata": {},
   "source": [
    "### Evaluate Model 1- Holt-Winters ET model"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "ab5e0159",
   "metadata": {},
   "source": [
    "### Calculate RMSE "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 670,
   "id": "469c22fe",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "The RMSE for the Holt-Winters ET model with weekly seasonality is 0.024502088350959086\n",
      "The RMSE for the Holt-Winters ET model with monthly seasonality is 0.02410175027437978\n",
      "The RMSE for the Holt-Winters ET model with weekly seasonality is 0.03380904834072822\n"
     ]
    }
   ],
   "source": [
    "#The rmse for the weekly seasnonality model \n",
    "rmse_weekly = mean_squared_error(test_set['Close'],predicted_prices_week, squared=False )\n",
    "\n",
    "#The rmse for the monthly seasonality model \n",
    "rmse_monthly = mean_squared_error(test_set['Close'],predicted_prices_month, squared=False )\n",
    "\n",
    "#The rmse for the annual seasonality model \n",
    "rmse_annual = mean_squared_error(test_set['Close'],predicted_prices_annual, squared=False )\n",
    "\n",
    "print(f'The RMSE for the Holt-Winters ET model with weekly seasonality is {rmse_weekly}')\n",
    "print(f'The RMSE for the Holt-Winters ET model with monthly seasonality is {rmse_monthly}')\n",
    "print(f'The RMSE for the Holt-Winters ET model with weekly seasonality is {rmse_annual}')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 671,
   "id": "1adf26e0",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "count    180.000000\n",
       "mean       0.768410\n",
       "std        0.021806\n",
       "min        0.721761\n",
       "25%        0.754429\n",
       "50%        0.773500\n",
       "75%        0.782154\n",
       "max        0.801134\n",
       "Name: Close, dtype: float64"
      ]
     },
     "execution_count": 671,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "test_set['Close'].describe()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "b02e540a",
   "metadata": {},
   "source": [
    "### How did the Models perform?"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "5fc7f557",
   "metadata": {},
   "source": [
    "**Which seasonality period gave the best performance?**\n",
    "   - The model with the monthly seasonality period gave the best performance, This is not surprising as the ETS decomposition already gave an idea that there is a bit of monthly seasonality in the data"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "9e65ea16",
   "metadata": {},
   "source": [
    "**Is the RMSE good or bad?**\n",
    "  - The relative RMSE (RMSE/mean of the dataset) = 0.031 (about 3%)\n",
    "  - The RMSE is greater than the SD of the dataset\n",
    "  \n",
    " My conclusion is that the RMSE is not so good "
   ]
  },
  {
   "cell_type": "markdown",
   "id": "429302d9",
   "metadata": {},
   "source": [
    "**Was the model able to predict the future trend? i.e prices stable, increasing or decreasing**\n",
    "   - The three models were not able to correctly predict the future trends. Infact two of the models predicted the prices being relatively stable for the period and we can see that was not the case "
   ]
  },
  {
   "cell_type": "markdown",
   "id": "e5412959",
   "metadata": {},
   "source": [
    "### Conclusion & Recommendation\n",
    "\n",
    "This Holt-Winters ET model didnt correctly predict the price or the future trend. This could be because this kind of data (CAD/USD exhange prices) depends on a lot more factors rather than the time stamp.\n",
    "I'll try other models to see if they will give better performance"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "1448e0d4",
   "metadata": {},
   "source": [
    "#### Model 2 - Auto Regression Model"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "462947a2",
   "metadata": {},
   "source": [
    "AR - Just like regression but the parameters is the time series data with lags. \n",
    "So parameter one can be the time series itself with one lag backwards.\n",
    "The order of the AR model is the no of lags used. So in the example above, the order is 1.\n",
    "The more order you use, the better performance but it will reach a stage when performance starts to decline. So this is why you need to select the optimal order to use."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 672,
   "id": "1f94ebc9",
   "metadata": {},
   "outputs": [],
   "source": [
    "from statsmodels.tsa.ar_model import AutoReg,ARResults"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 673,
   "id": "7956ce34",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<AxesSubplot:xlabel='Date'>"
      ]
     },
     "execution_count": 673,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "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": [
    "data['Close'].plot()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "42013c99",
   "metadata": {},
   "source": [
    "Seeing the way my data is right now, AR model will most likely not perform well since there isnt a clear linearity. But let me find out!"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 674,
   "id": "b3842a18",
   "metadata": {},
   "outputs": [],
   "source": [
    "import warnings\n",
    "warnings.filterwarnings('ignore')"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "518a78f7",
   "metadata": {},
   "source": [
    "### Training a 1st order AR model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 675,
   "id": "c437174c",
   "metadata": {},
   "outputs": [],
   "source": [
    "AR_Order1 = AutoReg(train_set['Close'], lags=1).fit()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 676,
   "id": "4c75b048",
   "metadata": {},
   "outputs": [],
   "source": [
    "start = len(train_set['Close']) + 1"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 677,
   "id": "ab87dd11",
   "metadata": {},
   "outputs": [],
   "source": [
    "end = len(train_set['Close']) + len(test_set['Close'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 678,
   "id": "d0e77f86",
   "metadata": {},
   "outputs": [],
   "source": [
    "predictions1 = AR_Order1.predict(start = start, end = end).rename('AR(1) Predictions')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 679,
   "id": "a795f65c",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "2022-03-16    0.779753\n",
       "2022-03-17    0.779649\n",
       "2022-03-18    0.779545\n",
       "2022-03-21    0.779443\n",
       "2022-03-22    0.779342\n",
       "                ...   \n",
       "2022-11-16    0.770777\n",
       "2022-11-17    0.770756\n",
       "2022-11-18    0.770737\n",
       "2022-11-21    0.770717\n",
       "2022-11-22    0.770697\n",
       "Freq: B, Name: AR(1) Predictions, Length: 180, dtype: float64"
      ]
     },
     "execution_count": 679,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "predictions1\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 680,
   "id": "31355063",
   "metadata": {},
   "outputs": [],
   "source": [
    "AR_Order2 = AutoReg(train_set['Close'], lags=2).fit()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 681,
   "id": "8cc548e5",
   "metadata": {},
   "outputs": [],
   "source": [
    "predictions2 = AR_Order2.predict(start = start, end = end).rename('AR(2) Predictions')"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "f6b74855",
   "metadata": {},
   "source": [
    "### Evaluating Model 2 - AutoRegression Model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 682,
   "id": "aa9ca27b",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<AxesSubplot:xlabel='Date'>"
      ]
     },
     "execution_count": 682,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "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": [
    "test_set['Close'].plot(legend = True)\n",
    "predictions1.plot(legend=True)\n",
    "predictions2.plot(legend=True)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "a5d1f324",
   "metadata": {},
   "source": [
    "### Calculate the RMSE"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 683,
   "id": "2c1b71c5",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "RMSE for first order AR model is 0.0204744407499538\n",
      "RMSE for second order AR model is 0.020361060586043315\n"
     ]
    }
   ],
   "source": [
    "rmse_lag1 = mean_squared_error(test_set['Close'],predictions1, squared=False)\n",
    "rmse_lag2 = mean_squared_error(test_set['Close'],predictions2, squared=False)\n",
    "print(f'RMSE for first order AR model is {rmse_lag1}')\n",
    "print(f'RMSE for second order AR model is {rmse_lag2}')"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "504286a0",
   "metadata": {},
   "source": [
    "### How did the AR Models perform?"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "0ccf89d1",
   "metadata": {},
   "source": [
    "**Which lag gave the best performance?**\n",
    "   - The second order AR model gave a slightly lower RMSE than the other model."
   ]
  },
  {
   "cell_type": "markdown",
   "id": "b0443b54",
   "metadata": {},
   "source": [
    "**How do I select the best lag (order) to use for the AR model?**\n",
    "  - Lets have statsmodels tell us the best order to use "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 684,
   "id": "d991192a",
   "metadata": {},
   "outputs": [],
   "source": [
    "from statsmodels.tsa.ar_model import ar_select_order"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 685,
   "id": "bfcdbd25",
   "metadata": {},
   "outputs": [],
   "source": [
    "select_order = ar_select_order(data['Close'], maxlag = 20, ic = 'aic')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 686,
   "id": "07fc9cde",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[1, 2, 3]"
      ]
     },
     "execution_count": 686,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "select_order.ar_lags"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "a5051bfd",
   "metadata": {},
   "source": [
    "Statsmodel says a lag of 1 is the optimal lag, even though lag 2 gave less rmse. This is probably because statsmodel uses aic to judge the best lag and this penalises more complex models. "
   ]
  },
  {
   "cell_type": "markdown",
   "id": "4435e46d",
   "metadata": {},
   "source": [
    "**Was the model able to predict the future trend?**"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "263eac6a",
   "metadata": {},
   "source": [
    "- In my opinion, This AR model did not give a good performance. \n",
    "- Even though the RMSE is lower than that of the Holt-Winters model, this AR model tries to plot a best fit line for the stock prices which is not appropriate for this kind of dataset because it doesnt have a linear relationship with time. \n",
    "- Hence, the model was not able to accurately forecast the future prices or trends\n",
    "- Let's try more models"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "37e5d4e7",
   "metadata": {},
   "source": [
    "### Model 3 - AutoRegressive Integrated Moving Average (ARIMA)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "01a99bca",
   "metadata": {},
   "source": [
    "From Analysis, I also confirmed that the data is not stationary, So I will need to transform it to be stationary in order to evaluate it & what type of parameters to use for ARIMA."
   ]
  },
  {
   "cell_type": "markdown",
   "id": "1e6dae48",
   "metadata": {},
   "source": [
    "The Differencing technique can be used to transform the dataset to stationary. You may have to difference multiple times before finally reaching stationarity."
   ]
  },
  {
   "cell_type": "markdown",
   "id": "6ba2055c",
   "metadata": {},
   "source": [
    "Pyramid arima is a library that can be used to do a grid search for p,d,q and it will also tell you whether to use a seasonal arima model (by giving P,D,Q).\n",
    "It uses the AIC metric to decide on the optimal p,d,q for you."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 687,
   "id": "1bcaa748",
   "metadata": {},
   "outputs": [],
   "source": [
    "from pmdarima import auto_arima"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 688,
   "id": "15d66825",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Performing stepwise search to minimize aic\n",
      " ARIMA(0,1,0)(0,0,0)[0] intercept   : AIC=-11031.002, Time=0.30 sec\n",
      " ARIMA(1,1,0)(0,0,0)[0] intercept   : AIC=-11029.464, Time=0.31 sec\n",
      " ARIMA(0,1,1)(0,0,0)[0] intercept   : AIC=-11029.551, Time=0.31 sec\n",
      " ARIMA(0,1,0)(0,0,0)[0]             : AIC=-11032.924, Time=0.09 sec\n",
      " ARIMA(1,1,1)(0,0,0)[0] intercept   : AIC=-11028.173, Time=0.33 sec\n",
      "\n",
      "Best model:  ARIMA(0,1,0)(0,0,0)[0]          \n",
      "Total fit time: 1.398 seconds\n"
     ]
    }
   ],
   "source": [
    "grid_search = auto_arima(data['Close'], start_p = 0, start_q = 0, max_q = 4, max_p = 6, trace = True, seasonal = True) #you can use the entire data\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 689,
   "id": "031b4b90",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<table class=\"simpletable\">\n",
       "<caption>SARIMAX Results</caption>\n",
       "<tr>\n",
       "  <th>Dep. Variable:</th>           <td>y</td>        <th>  No. Observations:  </th>    <td>1304</td>   \n",
       "</tr>\n",
       "<tr>\n",
       "  <th>Model:</th>           <td>SARIMAX(0, 1, 0)</td> <th>  Log Likelihood     </th>  <td>5517.462</td> \n",
       "</tr>\n",
       "<tr>\n",
       "  <th>Date:</th>            <td>Wed, 30 Nov 2022</td> <th>  AIC                </th> <td>-11032.924</td>\n",
       "</tr>\n",
       "<tr>\n",
       "  <th>Time:</th>                <td>16:28:06</td>     <th>  BIC                </th> <td>-11027.752</td>\n",
       "</tr>\n",
       "<tr>\n",
       "  <th>Sample:</th>             <td>11-22-2017</td>    <th>  HQIC               </th> <td>-11030.984</td>\n",
       "</tr>\n",
       "<tr>\n",
       "  <th></th>                   <td>- 11-22-2022</td>   <th>                     </th>      <td> </td>    \n",
       "</tr>\n",
       "<tr>\n",
       "  <th>Covariance Type:</th>        <td>opg</td>       <th>                     </th>      <td> </td>    \n",
       "</tr>\n",
       "</table>\n",
       "<table class=\"simpletable\">\n",
       "<tr>\n",
       "     <td></td>       <th>coef</th>     <th>std err</th>      <th>z</th>      <th>P>|z|</th>  <th>[0.025</th>    <th>0.975]</th>  \n",
       "</tr>\n",
       "<tr>\n",
       "  <th>sigma2</th> <td> 1.228e-05</td> <td> 2.71e-07</td> <td>   45.350</td> <td> 0.000</td> <td> 1.17e-05</td> <td> 1.28e-05</td>\n",
       "</tr>\n",
       "</table>\n",
       "<table class=\"simpletable\">\n",
       "<tr>\n",
       "  <th>Ljung-Box (L1) (Q):</th>     <td>0.46</td> <th>  Jarque-Bera (JB):  </th> <td>1006.89</td>\n",
       "</tr>\n",
       "<tr>\n",
       "  <th>Prob(Q):</th>                <td>0.50</td> <th>  Prob(JB):          </th>  <td>0.00</td>  \n",
       "</tr>\n",
       "<tr>\n",
       "  <th>Heteroskedasticity (H):</th> <td>1.25</td> <th>  Skew:              </th>  <td>0.05</td>  \n",
       "</tr>\n",
       "<tr>\n",
       "  <th>Prob(H) (two-sided):</th>    <td>0.02</td> <th>  Kurtosis:          </th>  <td>7.31</td>  \n",
       "</tr>\n",
       "</table><br/><br/>Warnings:<br/>[1] Covariance matrix calculated using the outer product of gradients (complex-step)."
      ],
      "text/plain": [
       "<class 'statsmodels.iolib.summary.Summary'>\n",
       "\"\"\"\n",
       "                               SARIMAX Results                                \n",
       "==============================================================================\n",
       "Dep. Variable:                      y   No. Observations:                 1304\n",
       "Model:               SARIMAX(0, 1, 0)   Log Likelihood                5517.462\n",
       "Date:                Wed, 30 Nov 2022   AIC                         -11032.924\n",
       "Time:                        16:28:06   BIC                         -11027.752\n",
       "Sample:                    11-22-2017   HQIC                        -11030.984\n",
       "                         - 11-22-2022                                         \n",
       "Covariance Type:                  opg                                         \n",
       "==============================================================================\n",
       "                 coef    std err          z      P>|z|      [0.025      0.975]\n",
       "------------------------------------------------------------------------------\n",
       "sigma2      1.228e-05   2.71e-07     45.350      0.000    1.17e-05    1.28e-05\n",
       "===================================================================================\n",
       "Ljung-Box (L1) (Q):                   0.46   Jarque-Bera (JB):              1006.89\n",
       "Prob(Q):                              0.50   Prob(JB):                         0.00\n",
       "Heteroskedasticity (H):               1.25   Skew:                             0.05\n",
       "Prob(H) (two-sided):                  0.02   Kurtosis:                         7.31\n",
       "===================================================================================\n",
       "\n",
       "Warnings:\n",
       "[1] Covariance matrix calculated using the outer product of gradients (complex-step).\n",
       "\"\"\""
      ]
     },
     "execution_count": 689,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "grid_search.summary()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "cdf86015",
   "metadata": {},
   "source": [
    "Pyramid Arima suggests:\n",
    "- P = 0\n",
    "- d = 1 - The data should be differenced 1 time\n",
    "- q = 0\n",
    "\n",
    "This is a random walk model (find out more)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 690,
   "id": "0f9471c4",
   "metadata": {},
   "outputs": [],
   "source": [
    "from statsmodels.tsa.arima.model import ARIMA, ARIMAResults"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 691,
   "id": "a9de716a",
   "metadata": {},
   "outputs": [],
   "source": [
    "arima_mod = ARIMA(train_set['Close'], order = (0,1,0)).fit()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 692,
   "id": "5c855d1e",
   "metadata": {},
   "outputs": [],
   "source": [
    "predictions_arima = arima_mod.predict(start=start, end = end)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 693,
   "id": "20b7b60c",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "2022-03-16    0.779964\n",
       "2022-03-17    0.779964\n",
       "2022-03-18    0.779964\n",
       "2022-03-21    0.779964\n",
       "2022-03-22    0.779964\n",
       "                ...   \n",
       "2022-11-16    0.779964\n",
       "2022-11-17    0.779964\n",
       "2022-11-18    0.779964\n",
       "2022-11-21    0.779964\n",
       "2022-11-22    0.779964\n",
       "Freq: B, Name: predicted_mean, Length: 180, dtype: float64"
      ]
     },
     "execution_count": 693,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "preditions_arima"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "cd2f4312",
   "metadata": {},
   "source": [
    "### Evaluating Model 2 - ARIMA Model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 694,
   "id": "3337b351",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<AxesSubplot:xlabel='Date'>"
      ]
     },
     "execution_count": 694,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "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": [
    "test_set['Close'].plot(legend = True)\n",
    "preditions_arima.plot(legend=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 562,
   "id": "c2dd7658",
   "metadata": {},
   "outputs": [],
   "source": [
    "from statsmodels.tools.eval_measures import rmse"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 563,
   "id": "57c0147d",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.024624180260693164"
      ]
     },
     "execution_count": 563,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "rmse(preditions_arima,test_set['Close'])"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "6d75b1fe",
   "metadata": {},
   "source": [
    "The ARIMA model tries to use the average of the past to predict the future stock prices. This is not such a good performance."
   ]
  },
  {
   "cell_type": "markdown",
   "id": "ea5a4a7c",
   "metadata": {},
   "source": [
    "### Model 4 - Recurrent Neural Network"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "4d397243",
   "metadata": {},
   "source": [
    "### Step 1 - Scale the Dataset"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "79549894",
   "metadata": {},
   "source": [
    "When working with neural networks, you need to scale the data so it all falls between 0 & 1"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 695,
   "id": "6a76ff0d",
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.preprocessing import MinMaxScaler"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 696,
   "id": "80d70ab2",
   "metadata": {},
   "outputs": [],
   "source": [
    "scaler = MinMaxScaler()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 697,
   "id": "a7d6e99b",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "MinMaxScaler()"
      ]
     },
     "execution_count": 697,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "scaler.fit(train_set[['Close']]) #Only fit on the training data\n",
    "#This will find the max value in the training data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 698,
   "id": "cb34dfea",
   "metadata": {},
   "outputs": [],
   "source": [
    "scaled_train = scaler.transform(train_set[['Close']])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 699,
   "id": "407668d5",
   "metadata": {},
   "outputs": [],
   "source": [
    "scaled_test = scaler.transform(test_set[['Close']])"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "e02ee50a",
   "metadata": {},
   "source": [
    "### Step 2 - Transofrm the data to this format; Input = [X1, x2, x3, ... Xn] Output = [Y]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 700,
   "id": "5024275e",
   "metadata": {},
   "outputs": [],
   "source": [
    "#Use a time series generator for your dataset, to transform it to the way you want \n",
    "from tensorflow.keras.preprocessing.sequence import TimeseriesGenerator"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 701,
   "id": "845c4053",
   "metadata": {},
   "outputs": [],
   "source": [
    "train_generator = TimeseriesGenerator(scaled_train,scaled_train, length = 20, batch_size=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 702,
   "id": "310d9b93",
   "metadata": {},
   "outputs": [],
   "source": [
    "X,y = train_generator[0]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 703,
   "id": "69dae554",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(array([[[0.66045459],\n",
       "         [0.69061875],\n",
       "         [0.68727742],\n",
       "         [0.68666733],\n",
       "         [0.66724365],\n",
       "         [0.64281863],\n",
       "         [0.62245853],\n",
       "         [0.60988068],\n",
       "         [0.68486542],\n",
       "         [0.70294831],\n",
       "         [0.69545693],\n",
       "         [0.65485024],\n",
       "         [0.62417531],\n",
       "         [0.62331692],\n",
       "         [0.62631773],\n",
       "         [0.61989047],\n",
       "         [0.6422582 ],\n",
       "         [0.65155858],\n",
       "         [0.61714505],\n",
       "         [0.62121707]]]),\n",
       " array([[0.61552049]]))"
      ]
     },
     "execution_count": 703,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X,y"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 704,
   "id": "0e12926c",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(array([[[0.69061875],\n",
       "         [0.68727742],\n",
       "         [0.68666733],\n",
       "         [0.66724365],\n",
       "         [0.64281863],\n",
       "         [0.62245853],\n",
       "         [0.60988068],\n",
       "         [0.68486542],\n",
       "         [0.70294831],\n",
       "         [0.69545693],\n",
       "         [0.65485024],\n",
       "         [0.62417531],\n",
       "         [0.62331692],\n",
       "         [0.62631773],\n",
       "         [0.61989047],\n",
       "         [0.6422582 ],\n",
       "         [0.65155858],\n",
       "         [0.61714505],\n",
       "         [0.62121707],\n",
       "         [0.61552049]]]),\n",
       " array([[0.63776763]]))"
      ]
     },
     "execution_count": 704,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "train_generator[1]"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "184bb3b1",
   "metadata": {},
   "source": [
    "So I have generated a sequence in our dataset\n",
    "The first sequence takes the first 20 prices as input and has the 21st price as the output\n",
    "If we look at the second sequence, the output from the last sequence is now the 20th value of the input array and this sequence of data will continue till the end."
   ]
  },
  {
   "cell_type": "markdown",
   "id": "752c5830",
   "metadata": {},
   "source": [
    "I have selected 20 as the initial timestep because the data has a monthly seasonality (20 business days in a month)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 729,
   "id": "363bf414",
   "metadata": {},
   "outputs": [],
   "source": [
    "from tensorflow.keras.models import Sequential\n",
    "from tensorflow.keras.layers import Dense\n",
    "from tensorflow.keras.layers import LSTM\n",
    "from tensorflow.keras.layers import Dropout\n",
    "from keras.callbacks import EarlyStopping"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "c357926b",
   "metadata": {},
   "source": [
    "### Step 3 - Create the NN"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 745,
   "id": "97516434",
   "metadata": {},
   "outputs": [],
   "source": [
    "model = Sequential()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "3342dcc8",
   "metadata": {},
   "source": [
    "Add the LSTM layer.\n",
    "- No of units - 150 # I have chosen this value randomly.This can be changed to try to improve the performance of the model\n",
    "- Activation fxn - 'relu' # I have selected the relu fxn has it usually gives the best performance\n",
    "- input_shape - (20,1) - Since I already transofrmed the data to a sequence with each row being an array of 20 inputs and another array of one output, I need to tell the LSTM layer to expect this as well.\n",
    "- I'm telling the model to use a sequence 20 previous prices to predict the next price.\n",
    "- Ideally you should use the seasonality of this data. so if you have monthly data that has annual seasonality you can use 12, 1\n",
    "- I have selected 20 as the initial timestep because the data has a monthly seasonality (20 business days in a month)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 746,
   "id": "4a2737c7",
   "metadata": {},
   "outputs": [],
   "source": [
    "model.add(LSTM(100, activation='relu', input_shape=(20, 1)))"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "6c8cfc5c",
   "metadata": {},
   "source": [
    "Add the dense Layer for the output"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 747,
   "id": "a7fa9d58",
   "metadata": {},
   "outputs": [],
   "source": [
    "model.add(Dense(1)) #output prediction"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 748,
   "id": "f4424745",
   "metadata": {},
   "outputs": [],
   "source": [
    "#Add dropout to prevent overfitting"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 749,
   "id": "f40c152c",
   "metadata": {},
   "outputs": [],
   "source": [
    "#model.add(Dropout(0.1))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 750,
   "id": "5cc1b1a5",
   "metadata": {},
   "outputs": [],
   "source": [
    "model.compile(optimizer = 'adam', loss = 'mse')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 751,
   "id": "a931b290",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Model: \"sequential_31\"\n",
      "_________________________________________________________________\n",
      " Layer (type)                Output Shape              Param #   \n",
      "=================================================================\n",
      " lstm_37 (LSTM)              (None, 100)               40800     \n",
      "                                                                 \n",
      " dense_29 (Dense)            (None, 1)                 101       \n",
      "                                                                 \n",
      "=================================================================\n",
      "Total params: 40,901\n",
      "Trainable params: 40,901\n",
      "Non-trainable params: 0\n",
      "_________________________________________________________________\n"
     ]
    }
   ],
   "source": [
    "model.summary()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "858613d1",
   "metadata": {},
   "source": [
    "### Step 4 - Fit the Model on the sequenced training set"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "cde8c97a",
   "metadata": {},
   "source": [
    "Now we will fit the model on the training generator data (not the original training set)\n",
    "This is because we are using LSTM and our model has been told to expect data with input shape (20,1)\n",
    "\n",
    "Since I cant use the validation split for LSTM, I will use early stopping to prevent the model from overfitting"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 752,
   "id": "2207717a",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/30\n",
      "1104/1104 [==============================] - 9s 6ms/step - loss: 0.0053\n",
      "Epoch 2/30\n",
      "1104/1104 [==============================] - 7s 6ms/step - loss: 0.0017\n",
      "Epoch 3/30\n",
      "1104/1104 [==============================] - 7s 6ms/step - loss: 0.0013\n",
      "Epoch 4/30\n",
      "1104/1104 [==============================] - 7s 6ms/step - loss: 0.0011\n",
      "Epoch 5/30\n",
      "1104/1104 [==============================] - 7s 6ms/step - loss: 0.0011\n",
      "Epoch 6/30\n",
      "1104/1104 [==============================] - 7s 6ms/step - loss: 8.4070e-04\n",
      "Epoch 7/30\n",
      "1104/1104 [==============================] - 7s 6ms/step - loss: 8.9973e-04\n",
      "Epoch 8/30\n",
      "1104/1104 [==============================] - 7s 6ms/step - loss: 8.1007e-04\n",
      "Epoch 9/30\n",
      "1104/1104 [==============================] - 7s 6ms/step - loss: 8.2999e-04\n",
      "Epoch 10/30\n",
      "1104/1104 [==============================] - 7s 6ms/step - loss: 8.0178e-04\n",
      "Epoch 11/30\n",
      "1104/1104 [==============================] - 7s 6ms/step - loss: 7.7183e-04\n",
      "Epoch 12/30\n",
      "1104/1104 [==============================] - 7s 6ms/step - loss: 7.8547e-04\n",
      "Epoch 13/30\n",
      "1104/1104 [==============================] - 7s 6ms/step - loss: 7.4440e-04\n",
      "Epoch 14/30\n",
      "1104/1104 [==============================] - 7s 7ms/step - loss: 7.7629e-04\n",
      "Epoch 15/30\n",
      "1104/1104 [==============================] - 7s 6ms/step - loss: 7.5060e-04\n",
      "Epoch 16/30\n",
      "1104/1104 [==============================] - 7s 6ms/step - loss: 7.1000e-04\n",
      "Epoch 17/30\n",
      "1104/1104 [==============================] - 7s 6ms/step - loss: 7.3196e-04\n",
      "Epoch 18/30\n",
      "1104/1104 [==============================] - 7s 6ms/step - loss: 7.2627e-04\n",
      "Epoch 19/30\n",
      "1104/1104 [==============================] - 7s 6ms/step - loss: 7.1918e-04\n",
      "Epoch 20/30\n",
      "1104/1104 [==============================] - 7s 6ms/step - loss: 7.0960e-04\n",
      "Epoch 21/30\n",
      "1104/1104 [==============================] - 7s 6ms/step - loss: 6.9790e-04\n",
      "Epoch 22/30\n",
      "1104/1104 [==============================] - 7s 6ms/step - loss: 7.1157e-04\n",
      "Epoch 23/30\n",
      "1104/1104 [==============================] - 7s 6ms/step - loss: 7.0059e-04\n",
      "Epoch 24/30\n",
      "1104/1104 [==============================] - 7s 6ms/step - loss: 7.0673e-04\n",
      "Epoch 25/30\n",
      "1104/1104 [==============================] - 7s 6ms/step - loss: 6.7852e-04\n",
      "Epoch 26/30\n",
      "1104/1104 [==============================] - 7s 6ms/step - loss: 6.7242e-04\n",
      "Epoch 27/30\n",
      "1104/1104 [==============================] - 7s 6ms/step - loss: 6.6942e-04\n",
      "Epoch 28/30\n",
      "1104/1104 [==============================] - 7s 6ms/step - loss: 6.7417e-04\n",
      "Epoch 29/30\n",
      "1104/1104 [==============================] - 7s 6ms/step - loss: 6.7732e-04\n",
      "Epoch 30/30\n",
      "1104/1104 [==============================] - 7s 6ms/step - loss: 6.5554e-04\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "<keras.callbacks.History at 0x2a121ba3550>"
      ]
     },
     "execution_count": 752,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "model.fit(train_generator,epochs = 30)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "36153267",
   "metadata": {},
   "source": [
    "I can see that the loss was already converging around 0.0007, which means that the training has been effective and the no of epochs selected was okay.\n",
    "If the loss didnt converge, then it means the no of epochs wasnt enough"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 753,
   "id": "852e21fc",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "dict_keys(['loss'])"
      ]
     },
     "execution_count": 753,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "model.history.history.keys()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 754,
   "id": "d5f2a5e6",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[<matplotlib.lines.Line2D at 0x2a12793b760>]"
      ]
     },
     "execution_count": 754,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "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": [
    "plt.plot(range(len(model.history.history['loss'])),model.history.history['loss'] )"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "c0bb9f6b",
   "metadata": {},
   "source": [
    "### Step 5 - Evaluate on the Test Data"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "1db97460",
   "metadata": {},
   "source": [
    "Evaluate on test data"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "27e4a660",
   "metadata": {},
   "source": [
    "so currently this network will see 20 history prices and predict the 21st price \n",
    "so we need the last 20 points of the training data to predict the first point of the test data"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "d85f9419",
   "metadata": {},
   "source": [
    "Let's test with the last 20 points in the train set and see what the model will predict"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 755,
   "id": "c639e0f7",
   "metadata": {},
   "outputs": [],
   "source": [
    "first_eval_batch = scaled_train[-20:]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 756,
   "id": "460e5b84",
   "metadata": {},
   "outputs": [],
   "source": [
    "#reshape to what the network expects\n",
    "first_eval_batch = first_eval_batch.reshape((1,20,1))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 757,
   "id": "b3a52432",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(1, 20, 1)"
      ]
     },
     "execution_count": 757,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "first_eval_batch.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 758,
   "id": "a53f9ee7",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1/1 [==============================] - 0s 188ms/step\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "array([[0.630296]], dtype=float32)"
      ]
     },
     "execution_count": 758,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "model.predict(first_eval_batch)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "cffaf479",
   "metadata": {},
   "source": [
    "Now lets predict for the entire test set"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "76530c82",
   "metadata": {},
   "source": [
    "### Forecast Using the Model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 759,
   "id": "d3616657",
   "metadata": {},
   "outputs": [],
   "source": [
    "# This will hold the predictions as we go along\n",
    "test_predictions = []\n",
    "\n",
    "#create the first evaluation batch with the last 20 inputs from the train set\n",
    "#This will be used to predict the first value in the test set\n",
    "first_eval_batch = scaled_train[-20:]\n",
    "\n",
    "#reshape to what the RNN wants (same format as output of time series generator)\n",
    "current_batch = first_eval_batch.reshape((1,20,1))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 760,
   "id": "9358e502",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1/1 [==============================] - 0s 24ms/step\n",
      "1/1 [==============================] - 0s 23ms/step\n",
      "1/1 [==============================] - 0s 22ms/step\n",
      "1/1 [==============================] - 0s 35ms/step\n",
      "1/1 [==============================] - 0s 23ms/step\n",
      "1/1 [==============================] - 0s 23ms/step\n",
      "1/1 [==============================] - 0s 21ms/step\n",
      "1/1 [==============================] - 0s 37ms/step\n",
      "1/1 [==============================] - 0s 23ms/step\n",
      "1/1 [==============================] - 0s 24ms/step\n",
      "1/1 [==============================] - 0s 21ms/step\n",
      "1/1 [==============================] - 0s 22ms/step\n",
      "1/1 [==============================] - 0s 20ms/step\n",
      "1/1 [==============================] - 0s 21ms/step\n",
      "1/1 [==============================] - 0s 21ms/step\n",
      "1/1 [==============================] - 0s 20ms/step\n",
      "1/1 [==============================] - 0s 22ms/step\n",
      "1/1 [==============================] - 0s 21ms/step\n",
      "1/1 [==============================] - 0s 21ms/step\n",
      "1/1 [==============================] - 0s 20ms/step\n",
      "1/1 [==============================] - 0s 20ms/step\n",
      "1/1 [==============================] - 0s 21ms/step\n",
      "1/1 [==============================] - 0s 20ms/step\n",
      "1/1 [==============================] - 0s 24ms/step\n",
      "1/1 [==============================] - 0s 20ms/step\n",
      "1/1 [==============================] - 0s 19ms/step\n",
      "1/1 [==============================] - 0s 21ms/step\n",
      "1/1 [==============================] - 0s 21ms/step\n",
      "1/1 [==============================] - 0s 20ms/step\n",
      "1/1 [==============================] - 0s 21ms/step\n",
      "1/1 [==============================] - 0s 20ms/step\n",
      "1/1 [==============================] - 0s 20ms/step\n",
      "1/1 [==============================] - 0s 25ms/step\n",
      "1/1 [==============================] - 0s 23ms/step\n",
      "1/1 [==============================] - 0s 21ms/step\n",
      "1/1 [==============================] - 0s 23ms/step\n",
      "1/1 [==============================] - 0s 21ms/step\n",
      "1/1 [==============================] - 0s 22ms/step\n",
      "1/1 [==============================] - 0s 26ms/step\n",
      "1/1 [==============================] - 0s 22ms/step\n",
      "1/1 [==============================] - 0s 27ms/step\n",
      "1/1 [==============================] - 0s 22ms/step\n",
      "1/1 [==============================] - 0s 21ms/step\n",
      "1/1 [==============================] - 0s 23ms/step\n",
      "1/1 [==============================] - 0s 24ms/step\n",
      "1/1 [==============================] - 0s 23ms/step\n",
      "1/1 [==============================] - 0s 21ms/step\n",
      "1/1 [==============================] - 0s 21ms/step\n",
      "1/1 [==============================] - 0s 23ms/step\n",
      "1/1 [==============================] - 0s 23ms/step\n",
      "1/1 [==============================] - 0s 24ms/step\n",
      "1/1 [==============================] - 0s 22ms/step\n",
      "1/1 [==============================] - 0s 22ms/step\n",
      "1/1 [==============================] - 0s 20ms/step\n",
      "1/1 [==============================] - 0s 24ms/step\n",
      "1/1 [==============================] - 0s 25ms/step\n",
      "1/1 [==============================] - 0s 20ms/step\n",
      "1/1 [==============================] - 0s 23ms/step\n",
      "1/1 [==============================] - 0s 22ms/step\n",
      "1/1 [==============================] - 0s 22ms/step\n",
      "1/1 [==============================] - 0s 23ms/step\n",
      "1/1 [==============================] - 0s 21ms/step\n",
      "1/1 [==============================] - 0s 19ms/step\n",
      "1/1 [==============================] - 0s 21ms/step\n",
      "1/1 [==============================] - 0s 27ms/step\n",
      "1/1 [==============================] - 0s 25ms/step\n",
      "1/1 [==============================] - 0s 23ms/step\n",
      "1/1 [==============================] - 0s 23ms/step\n",
      "1/1 [==============================] - 0s 28ms/step\n",
      "1/1 [==============================] - 0s 25ms/step\n",
      "1/1 [==============================] - 0s 25ms/step\n",
      "1/1 [==============================] - 0s 23ms/step\n",
      "1/1 [==============================] - 0s 24ms/step\n",
      "1/1 [==============================] - 0s 24ms/step\n",
      "1/1 [==============================] - 0s 24ms/step\n",
      "1/1 [==============================] - 0s 25ms/step\n",
      "1/1 [==============================] - 0s 24ms/step\n",
      "1/1 [==============================] - 0s 26ms/step\n",
      "1/1 [==============================] - 0s 26ms/step\n",
      "1/1 [==============================] - 0s 24ms/step\n",
      "1/1 [==============================] - 0s 27ms/step\n",
      "1/1 [==============================] - 0s 24ms/step\n",
      "1/1 [==============================] - 0s 24ms/step\n",
      "1/1 [==============================] - 0s 26ms/step\n",
      "1/1 [==============================] - 0s 27ms/step\n",
      "1/1 [==============================] - 0s 25ms/step\n",
      "1/1 [==============================] - 0s 24ms/step\n",
      "1/1 [==============================] - 0s 26ms/step\n",
      "1/1 [==============================] - 0s 25ms/step\n",
      "1/1 [==============================] - 0s 25ms/step\n",
      "1/1 [==============================] - 0s 26ms/step\n",
      "1/1 [==============================] - 0s 26ms/step\n",
      "1/1 [==============================] - 0s 25ms/step\n",
      "1/1 [==============================] - 0s 27ms/step\n",
      "1/1 [==============================] - 0s 27ms/step\n",
      "1/1 [==============================] - 0s 30ms/step\n",
      "1/1 [==============================] - 0s 27ms/step\n",
      "1/1 [==============================] - 0s 27ms/step\n",
      "1/1 [==============================] - 0s 21ms/step\n",
      "1/1 [==============================] - 0s 22ms/step\n",
      "1/1 [==============================] - 0s 21ms/step\n",
      "1/1 [==============================] - 0s 20ms/step\n",
      "1/1 [==============================] - 0s 20ms/step\n",
      "1/1 [==============================] - 0s 21ms/step\n",
      "1/1 [==============================] - 0s 21ms/step\n",
      "1/1 [==============================] - 0s 22ms/step\n",
      "1/1 [==============================] - 0s 26ms/step\n",
      "1/1 [==============================] - 0s 24ms/step\n",
      "1/1 [==============================] - 0s 22ms/step\n",
      "1/1 [==============================] - 0s 25ms/step\n",
      "1/1 [==============================] - 0s 24ms/step\n",
      "1/1 [==============================] - 0s 26ms/step\n",
      "1/1 [==============================] - 0s 23ms/step\n",
      "1/1 [==============================] - 0s 26ms/step\n",
      "1/1 [==============================] - 0s 25ms/step\n",
      "1/1 [==============================] - 0s 25ms/step\n",
      "1/1 [==============================] - 0s 27ms/step\n",
      "1/1 [==============================] - 0s 24ms/step\n",
      "1/1 [==============================] - 0s 22ms/step\n",
      "1/1 [==============================] - 0s 21ms/step\n",
      "1/1 [==============================] - 0s 23ms/step\n",
      "1/1 [==============================] - 0s 20ms/step\n",
      "1/1 [==============================] - 0s 29ms/step\n",
      "1/1 [==============================] - 0s 22ms/step\n",
      "1/1 [==============================] - 0s 21ms/step\n",
      "1/1 [==============================] - 0s 21ms/step\n",
      "1/1 [==============================] - 0s 21ms/step\n",
      "1/1 [==============================] - 0s 21ms/step\n",
      "1/1 [==============================] - 0s 21ms/step\n",
      "1/1 [==============================] - 0s 21ms/step\n",
      "1/1 [==============================] - 0s 21ms/step\n",
      "1/1 [==============================] - 0s 20ms/step\n",
      "1/1 [==============================] - 0s 21ms/step\n",
      "1/1 [==============================] - 0s 20ms/step\n",
      "1/1 [==============================] - 0s 21ms/step\n",
      "1/1 [==============================] - 0s 23ms/step\n",
      "1/1 [==============================] - 0s 21ms/step\n",
      "1/1 [==============================] - 0s 21ms/step\n",
      "1/1 [==============================] - 0s 21ms/step\n",
      "1/1 [==============================] - 0s 22ms/step\n",
      "1/1 [==============================] - 0s 21ms/step\n",
      "1/1 [==============================] - 0s 21ms/step\n",
      "1/1 [==============================] - 0s 20ms/step\n",
      "1/1 [==============================] - 0s 20ms/step\n",
      "1/1 [==============================] - 0s 21ms/step\n",
      "1/1 [==============================] - 0s 21ms/step\n",
      "1/1 [==============================] - 0s 21ms/step\n",
      "1/1 [==============================] - 0s 23ms/step\n",
      "1/1 [==============================] - 0s 22ms/step\n",
      "1/1 [==============================] - 0s 21ms/step\n",
      "1/1 [==============================] - 0s 21ms/step\n",
      "1/1 [==============================] - 0s 21ms/step\n",
      "1/1 [==============================] - 0s 22ms/step\n",
      "1/1 [==============================] - 0s 20ms/step\n",
      "1/1 [==============================] - 0s 21ms/step\n",
      "1/1 [==============================] - 0s 22ms/step\n",
      "1/1 [==============================] - 0s 21ms/step\n",
      "1/1 [==============================] - 0s 26ms/step\n",
      "1/1 [==============================] - 0s 25ms/step\n",
      "1/1 [==============================] - 0s 21ms/step\n",
      "1/1 [==============================] - 0s 21ms/step\n",
      "1/1 [==============================] - 0s 20ms/step\n",
      "1/1 [==============================] - 0s 19ms/step\n",
      "1/1 [==============================] - 0s 22ms/step\n",
      "1/1 [==============================] - 0s 21ms/step\n",
      "1/1 [==============================] - 0s 22ms/step\n",
      "1/1 [==============================] - 0s 20ms/step\n",
      "1/1 [==============================] - 0s 21ms/step\n",
      "1/1 [==============================] - 0s 20ms/step\n",
      "1/1 [==============================] - 0s 22ms/step\n",
      "1/1 [==============================] - 0s 21ms/step\n",
      "1/1 [==============================] - 0s 22ms/step\n",
      "1/1 [==============================] - 0s 22ms/step\n",
      "1/1 [==============================] - 0s 23ms/step\n",
      "1/1 [==============================] - 0s 20ms/step\n",
      "1/1 [==============================] - 0s 21ms/step\n",
      "1/1 [==============================] - 0s 22ms/step\n",
      "1/1 [==============================] - 0s 22ms/step\n",
      "1/1 [==============================] - 0s 22ms/step\n",
      "1/1 [==============================] - 0s 22ms/step\n"
     ]
    }
   ],
   "source": [
    "for i in range(len(test_set['Close'])):\n",
    "    \n",
    "    #Get the prediction one time step ahead of the past 20 points\n",
    "    current_pred = model.predict(current_batch)[0]\n",
    "    \n",
    "    #store the prediction\n",
    "    test_predictions.append(current_pred)\n",
    "    \n",
    "    #Update the current batch to include the prediction\n",
    "    #you need to drop the first value from the current batch and add the prediction to the end\n",
    "    #the indexing below is because the shape is 1,20,1.\n",
    "    #x =1 , y = 20, z = 1\n",
    "    #so we are grabbing the full sample ( x= 1), only from the 2nd up to the last of the y dimension\n",
    "    # and then add the new prediction\n",
    "    current_batch = np.append(current_batch[:,1:,:],[[current_pred]], axis = 1)\n",
    "    "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 761,
   "id": "bddeb758",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[array([0.630296], dtype=float32),\n",
       " array([0.62816674], dtype=float32),\n",
       " array([0.6217693], dtype=float32),\n",
       " array([0.6152964], dtype=float32),\n",
       " array([0.60892576], dtype=float32),\n",
       " array([0.6026868], dtype=float32),\n",
       " array([0.59656405], dtype=float32),\n",
       " array([0.5905422], dtype=float32),\n",
       " array([0.584611], dtype=float32),\n",
       " array([0.57876396], dtype=float32),\n",
       " array([0.5729971], dtype=float32),\n",
       " array([0.56730866], dtype=float32),\n",
       " array([0.5616974], dtype=float32),\n",
       " array([0.5561633], dtype=float32),\n",
       " array([0.55070704], dtype=float32),\n",
       " array([0.54532886], dtype=float32),\n",
       " array([0.5400289], dtype=float32),\n",
       " array([0.5348065], dtype=float32),\n",
       " array([0.5296616], dtype=float32),\n",
       " array([0.5245945], dtype=float32),\n",
       " array([0.51960576], dtype=float32),\n",
       " array([0.5146953], dtype=float32),\n",
       " array([0.50986284], dtype=float32),\n",
       " array([0.50510836], dtype=float32),\n",
       " array([0.5004316], dtype=float32),\n",
       " array([0.4958325], dtype=float32),\n",
       " array([0.49131072], dtype=float32),\n",
       " array([0.48686606], dtype=float32),\n",
       " array([0.4824983], dtype=float32),\n",
       " array([0.478207], dtype=float32),\n",
       " array([0.4739917], dtype=float32),\n",
       " array([0.46985215], dtype=float32),\n",
       " array([0.46578777], dtype=float32),\n",
       " array([0.4617982], dtype=float32),\n",
       " array([0.45788276], dtype=float32),\n",
       " array([0.45404106], dtype=float32),\n",
       " array([0.45027244], dtype=float32),\n",
       " array([0.44657636], dtype=float32),\n",
       " array([0.44295216], dtype=float32),\n",
       " array([0.43939906], dtype=float32),\n",
       " array([0.43591648], dtype=float32),\n",
       " array([0.43250388], dtype=float32),\n",
       " array([0.42916036], dtype=float32),\n",
       " array([0.42588514], dtype=float32),\n",
       " array([0.42267758], dtype=float32),\n",
       " array([0.4195369], dtype=float32),\n",
       " array([0.41646218], dtype=float32),\n",
       " array([0.41345263], dtype=float32),\n",
       " array([0.41050738], dtype=float32),\n",
       " array([0.40762568], dtype=float32),\n",
       " array([0.4048066], dtype=float32),\n",
       " array([0.40204936], dtype=float32),\n",
       " array([0.39935303], dtype=float32),\n",
       " array([0.39671677], dtype=float32),\n",
       " array([0.3941396], dtype=float32),\n",
       " array([0.39162064], dtype=float32),\n",
       " array([0.38915896], dtype=float32),\n",
       " array([0.38675368], dtype=float32),\n",
       " array([0.38440394], dtype=float32),\n",
       " array([0.3821088], dtype=float32),\n",
       " array([0.37986723], dtype=float32),\n",
       " array([0.37767845], dtype=float32),\n",
       " array([0.3755415], dtype=float32),\n",
       " array([0.37345535], dtype=float32),\n",
       " array([0.3714192], dtype=float32),\n",
       " array([0.36943212], dtype=float32),\n",
       " array([0.36749315], dtype=float32),\n",
       " array([0.36560145], dtype=float32),\n",
       " array([0.36375606], dtype=float32),\n",
       " array([0.36195612], dtype=float32),\n",
       " array([0.3602007], dtype=float32),\n",
       " array([0.35848892], dtype=float32),\n",
       " array([0.35681987], dtype=float32),\n",
       " array([0.35519272], dtype=float32),\n",
       " array([0.35360664], dtype=float32),\n",
       " array([0.3520607], dtype=float32),\n",
       " array([0.3505541], dtype=float32),\n",
       " array([0.349086], dtype=float32),\n",
       " array([0.34765553], dtype=float32),\n",
       " array([0.34626192], dtype=float32),\n",
       " array([0.3449043], dtype=float32),\n",
       " array([0.3434923], dtype=float32),\n",
       " array([0.34205034], dtype=float32),\n",
       " array([0.3405909], dtype=float32),\n",
       " array([0.33911827], dtype=float32),\n",
       " array([0.3376357], dtype=float32),\n",
       " array([0.3361454], dtype=float32),\n",
       " array([0.3346495], dtype=float32),\n",
       " array([0.33315015], dtype=float32),\n",
       " array([0.331622], dtype=float32),\n",
       " array([0.3300309], dtype=float32),\n",
       " array([0.32836422], dtype=float32),\n",
       " array([0.32661295], dtype=float32),\n",
       " array([0.32477155], dtype=float32),\n",
       " array([0.32283664], dtype=float32),\n",
       " array([0.32080656], dtype=float32),\n",
       " array([0.31868047], dtype=float32),\n",
       " array([0.31645656], dtype=float32),\n",
       " array([0.31413263], dtype=float32),\n",
       " array([0.31170666], dtype=float32),\n",
       " array([0.30917645], dtype=float32),\n",
       " array([0.3065387], dtype=float32),\n",
       " array([0.30379033], dtype=float32),\n",
       " array([0.3009275], dtype=float32),\n",
       " array([0.2979458], dtype=float32),\n",
       " array([0.2948409], dtype=float32),\n",
       " array([0.2916078], dtype=float32),\n",
       " array([0.288241], dtype=float32),\n",
       " array([0.28473464], dtype=float32),\n",
       " array([0.2810826], dtype=float32),\n",
       " array([0.27727824], dtype=float32),\n",
       " array([0.2733146], dtype=float32),\n",
       " array([0.2691843], dtype=float32),\n",
       " array([0.2648793], dtype=float32),\n",
       " array([0.26039112], dtype=float32),\n",
       " array([0.25571087], dtype=float32),\n",
       " array([0.25082916], dtype=float32),\n",
       " array([0.24573588], dtype=float32),\n",
       " array([0.24042034], dtype=float32),\n",
       " array([0.23484223], dtype=float32),\n",
       " array([0.22895525], dtype=float32),\n",
       " array([0.22274703], dtype=float32),\n",
       " array([0.2161492], dtype=float32),\n",
       " array([0.20900768], dtype=float32),\n",
       " array([0.20146072], dtype=float32),\n",
       " array([0.19392361], dtype=float32),\n",
       " array([0.18659818], dtype=float32),\n",
       " array([0.17968974], dtype=float32),\n",
       " array([0.17353243], dtype=float32),\n",
       " array([0.16841991], dtype=float32),\n",
       " array([0.164263], dtype=float32),\n",
       " array([0.16088384], dtype=float32),\n",
       " array([0.15822239], dtype=float32),\n",
       " array([0.15617791], dtype=float32),\n",
       " array([0.15458699], dtype=float32),\n",
       " array([0.15329295], dtype=float32),\n",
       " array([0.15217452], dtype=float32),\n",
       " array([0.1511531], dtype=float32),\n",
       " array([0.15018553], dtype=float32),\n",
       " array([0.14925732], dtype=float32),\n",
       " array([0.14836805], dtype=float32),\n",
       " array([0.14752509], dtype=float32),\n",
       " array([0.14673479], dtype=float32),\n",
       " array([0.14600271], dtype=float32),\n",
       " array([0.14533095], dtype=float32),\n",
       " array([0.14471832], dtype=float32),\n",
       " array([0.14416242], dtype=float32),\n",
       " array([0.14365914], dtype=float32),\n",
       " array([0.14320272], dtype=float32),\n",
       " array([0.14278765], dtype=float32),\n",
       " array([0.14240915], dtype=float32),\n",
       " array([0.14206305], dtype=float32),\n",
       " array([0.14174592], dtype=float32),\n",
       " array([0.14145485], dtype=float32),\n",
       " array([0.14118737], dtype=float32),\n",
       " array([0.14094159], dtype=float32),\n",
       " array([0.14071582], dtype=float32),\n",
       " array([0.1405086], dtype=float32),\n",
       " array([0.1403186], dtype=float32),\n",
       " array([0.14014451], dtype=float32),\n",
       " array([0.13998513], dtype=float32),\n",
       " array([0.13983925], dtype=float32),\n",
       " array([0.13970575], dtype=float32),\n",
       " array([0.13958357], dtype=float32),\n",
       " array([0.13947172], dtype=float32),\n",
       " array([0.13936932], dtype=float32),\n",
       " array([0.13927555], dtype=float32),\n",
       " array([0.13918965], dtype=float32),\n",
       " array([0.13911097], dtype=float32),\n",
       " array([0.13903889], dtype=float32),\n",
       " array([0.13897286], dtype=float32),\n",
       " array([0.13891238], dtype=float32),\n",
       " array([0.13885698], dtype=float32),\n",
       " array([0.13880624], dtype=float32),\n",
       " array([0.13875979], dtype=float32),\n",
       " array([0.13871725], dtype=float32),\n",
       " array([0.1386783], dtype=float32),\n",
       " array([0.13864262], dtype=float32),\n",
       " array([0.13860993], dtype=float32),\n",
       " array([0.13858001], dtype=float32)]"
      ]
     },
     "execution_count": 761,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "test_predictions"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "0d499707",
   "metadata": {},
   "source": [
    "Note that the predictions are still scaled"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "d5f05e79",
   "metadata": {},
   "source": [
    "### Step 5 - Transform predictions back to original data "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 762,
   "id": "b1ade42b",
   "metadata": {},
   "outputs": [],
   "source": [
    "True_predictions = scaler.inverse_transform(test_predictions)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 763,
   "id": "9902929b",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[0.77858878],\n",
       "       [0.77828864],\n",
       "       [0.77738685],\n",
       "       [0.77647441],\n",
       "       [0.77557639],\n",
       "       [0.77469694],\n",
       "       [0.77383386],\n",
       "       [0.77298501],\n",
       "       [0.77214894],\n",
       "       [0.77132473],\n",
       "       [0.77051182],\n",
       "       [0.76970996],\n",
       "       [0.76891899],\n",
       "       [0.76813889],\n",
       "       [0.76736977],\n",
       "       [0.76661165],\n",
       "       [0.76586456],\n",
       "       [0.76512839],\n",
       "       [0.76440316],\n",
       "       [0.76368889],\n",
       "       [0.76298567],\n",
       "       [0.76229348],\n",
       "       [0.76161229],\n",
       "       [0.76094208],\n",
       "       [0.76028284],\n",
       "       [0.75963454],\n",
       "       [0.75899714],\n",
       "       [0.75837061],\n",
       "       [0.75775492],\n",
       "       [0.75715001],\n",
       "       [0.75655582],\n",
       "       [0.7559723 ],\n",
       "       [0.75539938],\n",
       "       [0.754837  ],\n",
       "       [0.75428507],\n",
       "       [0.75374354],\n",
       "       [0.7532123 ],\n",
       "       [0.7526913 ],\n",
       "       [0.75218042],\n",
       "       [0.75167957],\n",
       "       [0.75118866],\n",
       "       [0.75070761],\n",
       "       [0.7502363 ],\n",
       "       [0.74977462],\n",
       "       [0.74932248],\n",
       "       [0.74887976],\n",
       "       [0.74844634],\n",
       "       [0.74802211],\n",
       "       [0.74760694],\n",
       "       [0.74720073],\n",
       "       [0.74680335],\n",
       "       [0.74641468],\n",
       "       [0.7460346 ],\n",
       "       [0.74566299],\n",
       "       [0.7452997 ],\n",
       "       [0.74494463],\n",
       "       [0.74459763],\n",
       "       [0.74425857],\n",
       "       [0.74392735],\n",
       "       [0.74360382],\n",
       "       [0.74328784],\n",
       "       [0.74297931],\n",
       "       [0.74267808],\n",
       "       [0.74238401],\n",
       "       [0.74209699],\n",
       "       [0.74181689],\n",
       "       [0.74154357],\n",
       "       [0.74127691],\n",
       "       [0.74101678],\n",
       "       [0.74076306],\n",
       "       [0.74051561],\n",
       "       [0.74027431],\n",
       "       [0.74003904],\n",
       "       [0.73980968],\n",
       "       [0.7395861 ],\n",
       "       [0.73936818],\n",
       "       [0.73915581],\n",
       "       [0.73894886],\n",
       "       [0.73874722],\n",
       "       [0.73855077],\n",
       "       [0.7383594 ],\n",
       "       [0.73816036],\n",
       "       [0.7379571 ],\n",
       "       [0.73775137],\n",
       "       [0.73754379],\n",
       "       [0.7373348 ],\n",
       "       [0.73712473],\n",
       "       [0.73691386],\n",
       "       [0.73670251],\n",
       "       [0.7364871 ],\n",
       "       [0.73626281],\n",
       "       [0.73602788],\n",
       "       [0.73578101],\n",
       "       [0.73552145],\n",
       "       [0.7352487 ],\n",
       "       [0.73496253],\n",
       "       [0.73466284],\n",
       "       [0.73434935],\n",
       "       [0.73402176],\n",
       "       [0.73367979],\n",
       "       [0.73332313],\n",
       "       [0.73295131],\n",
       "       [0.73256389],\n",
       "       [0.73216034],\n",
       "       [0.73174004],\n",
       "       [0.73130236],\n",
       "       [0.73084662],\n",
       "       [0.73037203],\n",
       "       [0.72987776],\n",
       "       [0.72936297],\n",
       "       [0.7288267 ],\n",
       "       [0.72826797],\n",
       "       [0.72768576],\n",
       "       [0.72707891],\n",
       "       [0.72644625],\n",
       "       [0.72578652],\n",
       "       [0.72509838],\n",
       "       [0.72438042],\n",
       "       [0.72363113],\n",
       "       [0.72284483],\n",
       "       [0.72201499],\n",
       "       [0.72113987],\n",
       "       [0.72020982],\n",
       "       [0.71920314],\n",
       "       [0.71813931],\n",
       "       [0.71707686],\n",
       "       [0.71604425],\n",
       "       [0.71507042],\n",
       "       [0.71420248],\n",
       "       [0.71348181],\n",
       "       [0.71289584],\n",
       "       [0.71241951],\n",
       "       [0.71204434],\n",
       "       [0.71175615],\n",
       "       [0.71153189],\n",
       "       [0.71134948],\n",
       "       [0.71119182],\n",
       "       [0.71104784],\n",
       "       [0.71091145],\n",
       "       [0.71078061],\n",
       "       [0.71065526],\n",
       "       [0.71053643],\n",
       "       [0.71042503],\n",
       "       [0.71032183],\n",
       "       [0.71022714],\n",
       "       [0.71014078],\n",
       "       [0.71006242],\n",
       "       [0.70999148],\n",
       "       [0.70992714],\n",
       "       [0.70986863],\n",
       "       [0.70981528],\n",
       "       [0.70976649],\n",
       "       [0.70972179],\n",
       "       [0.70968076],\n",
       "       [0.70964305],\n",
       "       [0.70960841],\n",
       "       [0.70957658],\n",
       "       [0.70954737],\n",
       "       [0.70952059],\n",
       "       [0.70949605],\n",
       "       [0.70947358],\n",
       "       [0.70945302],\n",
       "       [0.7094342 ],\n",
       "       [0.70941698],\n",
       "       [0.70940121],\n",
       "       [0.70938678],\n",
       "       [0.70937356],\n",
       "       [0.70936145],\n",
       "       [0.70935036],\n",
       "       [0.7093402 ],\n",
       "       [0.70933089],\n",
       "       [0.70932237],\n",
       "       [0.70931456],\n",
       "       [0.70930741],\n",
       "       [0.70930086],\n",
       "       [0.70929486],\n",
       "       [0.70928937],\n",
       "       [0.70928434],\n",
       "       [0.70927973],\n",
       "       [0.70927552]])"
      ]
     },
     "execution_count": 763,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "True_predictions"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "0f28092f",
   "metadata": {},
   "source": [
    "Now I have the original values"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 764,
   "id": "760d1085",
   "metadata": {},
   "outputs": [],
   "source": [
    "original_test = test_set[['Close']]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 765,
   "id": "3e4ccf70",
   "metadata": {},
   "outputs": [],
   "source": [
    "original_test['predictions'] = True_predictions"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "bc653021",
   "metadata": {},
   "source": [
    "### Step 6 - Evaluate RNN Performance"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 767,
   "id": "4ae6bc87",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<AxesSubplot:xlabel='Date'>"
      ]
     },
     "execution_count": 767,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "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": [
    "original_test.plot(legend=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 768,
   "id": "bc01bd4f",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.03507562465172962"
      ]
     },
     "execution_count": 768,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "mean_squared_error(original_test['Close'],original_test['predictions'],squared = False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "94103d48",
   "metadata": {},
   "outputs": [],
   "source": [
    "### Step 7 - Tuning Model - Improving Performance"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "649045fc",
   "metadata": {},
   "source": [
    "So far, this RNN has given the highest RMSE. Why did it give such poor performance compared to the rest?\n",
    "I will continue tweaking to see if there will be improvements \n",
    "so far I have changed sequence length from 5 to 20 and this caused the rmse to drop but still on (0.03963)\n",
    "Try different timesteps & plot the results "
   ]
  },
  {
   "cell_type": "markdown",
   "id": "95c2f032",
   "metadata": {},
   "source": [
    "1. I'll try different timesteps values & compare performances"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 636,
   "id": "a4dd4d70",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/25\n",
      "1104/1104 [==============================] - 9s 7ms/step - loss: 0.0068\n",
      "Epoch 2/25\n",
      "1104/1104 [==============================] - 7s 6ms/step - loss: 0.0018\n",
      "Epoch 3/25\n",
      "1104/1104 [==============================] - 7s 7ms/step - loss: 0.0013\n",
      "Epoch 4/25\n",
      "1104/1104 [==============================] - 7s 7ms/step - loss: 0.0011\n",
      "Epoch 5/25\n",
      "1104/1104 [==============================] - 7s 7ms/step - loss: 0.0010\n",
      "Epoch 6/25\n",
      "1104/1104 [==============================] - 7s 7ms/step - loss: 9.3508e-04\n",
      "Epoch 7/25\n",
      "1104/1104 [==============================] - 7s 7ms/step - loss: 9.0222e-04\n",
      "Epoch 8/25\n",
      "1104/1104 [==============================] - 7s 7ms/step - loss: 8.1627e-04\n",
      "Epoch 9/25\n",
      "1104/1104 [==============================] - 7s 6ms/step - loss: 9.0808e-04\n",
      "Epoch 10/25\n",
      "1104/1104 [==============================] - 7s 7ms/step - loss: 8.4012e-04\n",
      "Epoch 11/25\n",
      "1104/1104 [==============================] - 7s 7ms/step - loss: 7.8424e-04\n",
      "Epoch 12/25\n",
      "1104/1104 [==============================] - 7s 6ms/step - loss: 7.3269e-04\n",
      "Epoch 13/25\n",
      "1104/1104 [==============================] - 7s 6ms/step - loss: 7.9770e-04\n",
      "Epoch 14/25\n",
      "1104/1104 [==============================] - 7s 6ms/step - loss: 7.3843e-04\n",
      "Epoch 15/25\n",
      "1104/1104 [==============================] - 7s 7ms/step - loss: 8.4881e-04\n",
      "Epoch 16/25\n",
      "1104/1104 [==============================] - 7s 7ms/step - loss: 7.1602e-04\n",
      "Epoch 17/25\n",
      "1104/1104 [==============================] - 7s 6ms/step - loss: 7.7065e-04\n",
      "Epoch 18/25\n",
      "1104/1104 [==============================] - 7s 7ms/step - loss: 7.2248e-04\n",
      "Epoch 19/25\n",
      "1104/1104 [==============================] - 7s 6ms/step - loss: 7.1454e-04\n",
      "Epoch 20/25\n",
      "1104/1104 [==============================] - 7s 6ms/step - loss: 7.2414e-04\n",
      "Epoch 21/25\n",
      "1104/1104 [==============================] - 7s 6ms/step - loss: 7.3582e-04\n",
      "Epoch 22/25\n",
      "1104/1104 [==============================] - 7s 6ms/step - loss: 6.7618e-04\n",
      "Epoch 23/25\n",
      "1104/1104 [==============================] - 7s 6ms/step - loss: 7.2188e-04\n",
      "Epoch 24/25\n",
      "1104/1104 [==============================] - 7s 6ms/step - loss: 7.1915e-04\n",
      "Epoch 25/25\n",
      "1104/1104 [==============================] - 7s 6ms/step - loss: 6.9249e-04\n",
      "1/1 [==============================] - 0s 187ms/step\n",
      "1/1 [==============================] - 0s 22ms/step\n",
      "1/1 [==============================] - 0s 20ms/step\n",
      "1/1 [==============================] - 0s 22ms/step\n",
      "1/1 [==============================] - 0s 38ms/step\n",
      "1/1 [==============================] - 0s 21ms/step\n",
      "1/1 [==============================] - 0s 19ms/step\n",
      "1/1 [==============================] - 0s 21ms/step\n",
      "1/1 [==============================] - 0s 21ms/step\n",
      "1/1 [==============================] - 0s 20ms/step\n",
      "1/1 [==============================] - 0s 20ms/step\n",
      "1/1 [==============================] - 0s 21ms/step\n",
      "1/1 [==============================] - 0s 21ms/step\n",
      "1/1 [==============================] - 0s 21ms/step\n",
      "1/1 [==============================] - 0s 21ms/step\n",
      "1/1 [==============================] - 0s 19ms/step\n",
      "1/1 [==============================] - 0s 21ms/step\n",
      "1/1 [==============================] - 0s 21ms/step\n",
      "1/1 [==============================] - 0s 20ms/step\n",
      "1/1 [==============================] - 0s 20ms/step\n",
      "1/1 [==============================] - 0s 20ms/step\n",
      "1/1 [==============================] - 0s 21ms/step\n",
      "1/1 [==============================] - 0s 28ms/step\n",
      "1/1 [==============================] - 0s 20ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 21ms/step\n",
      "1/1 [==============================] - 0s 21ms/step\n",
      "1/1 [==============================] - 0s 20ms/step\n",
      "1/1 [==============================] - 0s 19ms/step\n",
      "1/1 [==============================] - 0s 24ms/step\n",
      "1/1 [==============================] - 0s 20ms/step\n",
      "1/1 [==============================] - 0s 21ms/step\n",
      "1/1 [==============================] - 0s 22ms/step\n",
      "1/1 [==============================] - 0s 22ms/step\n",
      "1/1 [==============================] - 0s 21ms/step\n",
      "1/1 [==============================] - 0s 21ms/step\n",
      "1/1 [==============================] - 0s 23ms/step\n",
      "1/1 [==============================] - 0s 22ms/step\n",
      "1/1 [==============================] - 0s 23ms/step\n",
      "1/1 [==============================] - 0s 26ms/step\n",
      "1/1 [==============================] - 0s 21ms/step\n",
      "1/1 [==============================] - 0s 19ms/step\n",
      "1/1 [==============================] - 0s 21ms/step\n",
      "1/1 [==============================] - 0s 20ms/step\n",
      "1/1 [==============================] - 0s 20ms/step\n",
      "1/1 [==============================] - 0s 20ms/step\n",
      "1/1 [==============================] - 0s 20ms/step\n",
      "1/1 [==============================] - 0s 21ms/step\n",
      "1/1 [==============================] - 0s 32ms/step\n",
      "1/1 [==============================] - 0s 22ms/step\n",
      "1/1 [==============================] - 0s 22ms/step\n",
      "1/1 [==============================] - 0s 21ms/step\n",
      "1/1 [==============================] - 0s 22ms/step\n",
      "1/1 [==============================] - 0s 20ms/step\n",
      "1/1 [==============================] - 0s 20ms/step\n",
      "1/1 [==============================] - 0s 20ms/step\n",
      "1/1 [==============================] - 0s 21ms/step\n",
      "1/1 [==============================] - 0s 20ms/step\n",
      "1/1 [==============================] - 0s 20ms/step\n",
      "1/1 [==============================] - 0s 21ms/step\n",
      "1/1 [==============================] - 0s 21ms/step\n",
      "1/1 [==============================] - 0s 20ms/step\n",
      "1/1 [==============================] - 0s 20ms/step\n",
      "1/1 [==============================] - 0s 21ms/step\n",
      "1/1 [==============================] - 0s 21ms/step\n",
      "1/1 [==============================] - 0s 21ms/step\n",
      "1/1 [==============================] - 0s 20ms/step\n",
      "1/1 [==============================] - 0s 21ms/step\n",
      "1/1 [==============================] - 0s 43ms/step\n",
      "1/1 [==============================] - 0s 20ms/step\n",
      "1/1 [==============================] - 0s 20ms/step\n",
      "1/1 [==============================] - 0s 22ms/step\n",
      "1/1 [==============================] - 0s 21ms/step\n",
      "1/1 [==============================] - 0s 21ms/step\n",
      "1/1 [==============================] - 0s 22ms/step\n",
      "1/1 [==============================] - 0s 21ms/step\n",
      "1/1 [==============================] - 0s 30ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 20ms/step\n",
      "1/1 [==============================] - 0s 20ms/step\n",
      "1/1 [==============================] - 0s 19ms/step\n",
      "1/1 [==============================] - 0s 22ms/step\n",
      "1/1 [==============================] - 0s 22ms/step\n",
      "1/1 [==============================] - 0s 20ms/step\n",
      "1/1 [==============================] - 0s 19ms/step\n",
      "1/1 [==============================] - 0s 21ms/step\n",
      "1/1 [==============================] - 0s 22ms/step\n",
      "1/1 [==============================] - 0s 24ms/step\n",
      "1/1 [==============================] - 0s 24ms/step\n",
      "1/1 [==============================] - 0s 20ms/step\n",
      "1/1 [==============================] - 0s 22ms/step\n",
      "1/1 [==============================] - 0s 20ms/step\n",
      "1/1 [==============================] - 0s 21ms/step\n",
      "1/1 [==============================] - 0s 21ms/step\n",
      "1/1 [==============================] - 0s 21ms/step\n",
      "1/1 [==============================] - 0s 20ms/step\n",
      "1/1 [==============================] - 0s 21ms/step\n",
      "1/1 [==============================] - 0s 20ms/step\n",
      "1/1 [==============================] - 0s 22ms/step\n",
      "1/1 [==============================] - 0s 20ms/step\n",
      "1/1 [==============================] - 0s 20ms/step\n",
      "1/1 [==============================] - 0s 19ms/step\n",
      "1/1 [==============================] - 0s 23ms/step\n",
      "1/1 [==============================] - 0s 20ms/step\n",
      "1/1 [==============================] - 0s 21ms/step\n",
      "1/1 [==============================] - 0s 20ms/step\n",
      "1/1 [==============================] - 0s 20ms/step\n",
      "1/1 [==============================] - 0s 19ms/step\n",
      "1/1 [==============================] - 0s 20ms/step\n",
      "1/1 [==============================] - 0s 21ms/step\n",
      "1/1 [==============================] - 0s 20ms/step\n",
      "1/1 [==============================] - 0s 37ms/step\n",
      "1/1 [==============================] - 0s 21ms/step\n",
      "1/1 [==============================] - 0s 19ms/step\n",
      "1/1 [==============================] - 0s 20ms/step\n",
      "1/1 [==============================] - 0s 20ms/step\n",
      "1/1 [==============================] - 0s 20ms/step\n",
      "1/1 [==============================] - 0s 19ms/step\n",
      "1/1 [==============================] - 0s 20ms/step\n",
      "1/1 [==============================] - 0s 21ms/step\n",
      "1/1 [==============================] - 0s 21ms/step\n",
      "1/1 [==============================] - 0s 20ms/step\n",
      "1/1 [==============================] - 0s 21ms/step\n",
      "1/1 [==============================] - 0s 20ms/step\n",
      "1/1 [==============================] - 0s 20ms/step\n",
      "1/1 [==============================] - 0s 20ms/step\n",
      "1/1 [==============================] - 0s 21ms/step\n",
      "1/1 [==============================] - 0s 19ms/step\n",
      "1/1 [==============================] - 0s 20ms/step\n",
      "1/1 [==============================] - 0s 22ms/step\n",
      "1/1 [==============================] - 0s 20ms/step\n",
      "1/1 [==============================] - 0s 21ms/step\n",
      "1/1 [==============================] - 0s 21ms/step\n",
      "1/1 [==============================] - 0s 20ms/step\n",
      "1/1 [==============================] - 0s 21ms/step\n",
      "1/1 [==============================] - 0s 21ms/step\n",
      "1/1 [==============================] - 0s 20ms/step\n",
      "1/1 [==============================] - 0s 22ms/step\n",
      "1/1 [==============================] - 0s 21ms/step\n",
      "1/1 [==============================] - 0s 21ms/step\n",
      "1/1 [==============================] - 0s 22ms/step\n",
      "1/1 [==============================] - 0s 20ms/step\n",
      "1/1 [==============================] - 0s 19ms/step\n",
      "1/1 [==============================] - 0s 22ms/step\n",
      "1/1 [==============================] - 0s 29ms/step\n",
      "1/1 [==============================] - 0s 20ms/step\n",
      "1/1 [==============================] - 0s 20ms/step\n",
      "1/1 [==============================] - 0s 20ms/step\n",
      "1/1 [==============================] - 0s 20ms/step\n",
      "1/1 [==============================] - 0s 20ms/step\n",
      "1/1 [==============================] - 0s 21ms/step\n",
      "1/1 [==============================] - 0s 20ms/step\n",
      "1/1 [==============================] - 0s 21ms/step\n",
      "1/1 [==============================] - 0s 21ms/step\n",
      "1/1 [==============================] - 0s 21ms/step\n",
      "1/1 [==============================] - 0s 20ms/step\n",
      "1/1 [==============================] - 0s 19ms/step\n",
      "1/1 [==============================] - 0s 20ms/step\n",
      "1/1 [==============================] - 0s 20ms/step\n",
      "1/1 [==============================] - 0s 28ms/step\n",
      "1/1 [==============================] - 0s 22ms/step\n",
      "1/1 [==============================] - 0s 21ms/step\n",
      "1/1 [==============================] - 0s 20ms/step\n",
      "1/1 [==============================] - 0s 23ms/step\n",
      "1/1 [==============================] - 0s 22ms/step\n",
      "1/1 [==============================] - 0s 21ms/step\n",
      "1/1 [==============================] - 0s 19ms/step\n",
      "1/1 [==============================] - 0s 21ms/step\n",
      "1/1 [==============================] - 0s 19ms/step\n",
      "1/1 [==============================] - 0s 27ms/step\n",
      "1/1 [==============================] - 0s 21ms/step\n",
      "1/1 [==============================] - 0s 20ms/step\n",
      "1/1 [==============================] - 0s 20ms/step\n",
      "1/1 [==============================] - 0s 20ms/step\n",
      "1/1 [==============================] - 0s 21ms/step\n",
      "1/1 [==============================] - 0s 21ms/step\n",
      "1/1 [==============================] - 0s 21ms/step\n",
      "1/1 [==============================] - 0s 19ms/step\n",
      "1/1 [==============================] - 0s 21ms/step\n",
      "1/1 [==============================] - 0s 20ms/step\n",
      "Epoch 1/25\n",
      "1094/1094 [==============================] - 11s 9ms/step - loss: 0.0066\n",
      "Epoch 2/25\n",
      "1094/1094 [==============================] - 10s 9ms/step - loss: 0.0015\n",
      "Epoch 3/25\n",
      "1094/1094 [==============================] - 9s 9ms/step - loss: 0.0012\n",
      "Epoch 4/25\n",
      "1094/1094 [==============================] - 9s 9ms/step - loss: 0.0010\n",
      "Epoch 5/25\n",
      "1094/1094 [==============================] - 9s 9ms/step - loss: 0.0010\n",
      "Epoch 6/25\n",
      "1094/1094 [==============================] - 10s 9ms/step - loss: 9.5011e-04\n",
      "Epoch 7/25\n",
      "1094/1094 [==============================] - 10s 9ms/step - loss: 8.6027e-04\n",
      "Epoch 8/25\n",
      "1094/1094 [==============================] - 9s 9ms/step - loss: 8.0179e-04\n",
      "Epoch 9/25\n",
      "1094/1094 [==============================] - 10s 9ms/step - loss: 8.3174e-04\n",
      "Epoch 10/25\n",
      "1094/1094 [==============================] - 10s 9ms/step - loss: 7.9962e-04\n",
      "Epoch 11/25\n",
      "1094/1094 [==============================] - 10s 9ms/step - loss: 7.8199e-04\n",
      "Epoch 12/25\n",
      "1094/1094 [==============================] - 11s 10ms/step - loss: 7.6153e-04\n",
      "Epoch 13/25\n",
      "1094/1094 [==============================] - 12s 11ms/step - loss: 7.7167e-04\n",
      "Epoch 14/25\n",
      "1094/1094 [==============================] - 10s 9ms/step - loss: 7.9012e-04\n",
      "Epoch 15/25\n",
      "1094/1094 [==============================] - 12s 11ms/step - loss: 7.4685e-04\n",
      "Epoch 16/25\n",
      "1094/1094 [==============================] - 11s 10ms/step - loss: 7.7385e-04\n",
      "Epoch 17/25\n",
      "1094/1094 [==============================] - 11s 10ms/step - loss: 7.3359e-04\n",
      "Epoch 18/25\n",
      "1094/1094 [==============================] - 11s 10ms/step - loss: 7.4456e-04\n",
      "Epoch 19/25\n",
      "1094/1094 [==============================] - 12s 11ms/step - loss: 7.2304e-04\n",
      "Epoch 20/25\n",
      "1094/1094 [==============================] - 12s 11ms/step - loss: 7.1221e-04\n",
      "Epoch 21/25\n",
      "1094/1094 [==============================] - 12s 11ms/step - loss: 7.4391e-04\n",
      "Epoch 22/25\n",
      "1094/1094 [==============================] - 11s 10ms/step - loss: 6.9825e-04\n",
      "Epoch 23/25\n",
      "1094/1094 [==============================] - 11s 10ms/step - loss: 7.0305e-04\n",
      "Epoch 24/25\n",
      "1094/1094 [==============================] - 12s 11ms/step - loss: 7.1200e-04\n",
      "Epoch 25/25\n",
      "1094/1094 [==============================] - 11s 10ms/step - loss: 7.0985e-04\n",
      "1/1 [==============================] - 0s 232ms/step\n",
      "1/1 [==============================] - 0s 30ms/step\n",
      "1/1 [==============================] - 0s 30ms/step\n",
      "1/1 [==============================] - 0s 26ms/step\n",
      "1/1 [==============================] - 0s 26ms/step\n",
      "1/1 [==============================] - 0s 25ms/step\n",
      "1/1 [==============================] - 0s 26ms/step\n",
      "1/1 [==============================] - 0s 27ms/step\n",
      "1/1 [==============================] - 0s 26ms/step\n",
      "1/1 [==============================] - 0s 28ms/step\n",
      "1/1 [==============================] - 0s 30ms/step\n",
      "1/1 [==============================] - 0s 32ms/step\n",
      "1/1 [==============================] - 0s 30ms/step\n",
      "1/1 [==============================] - 0s 32ms/step\n",
      "1/1 [==============================] - 0s 32ms/step\n",
      "1/1 [==============================] - 0s 30ms/step\n",
      "1/1 [==============================] - 0s 32ms/step\n",
      "1/1 [==============================] - 0s 32ms/step\n",
      "1/1 [==============================] - 0s 32ms/step\n",
      "1/1 [==============================] - 0s 30ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 23ms/step\n",
      "1/1 [==============================] - 0s 24ms/step\n",
      "1/1 [==============================] - 0s 29ms/step\n",
      "1/1 [==============================] - 0s 28ms/step\n",
      "1/1 [==============================] - 0s 26ms/step\n",
      "1/1 [==============================] - 0s 25ms/step\n",
      "1/1 [==============================] - 0s 26ms/step\n",
      "1/1 [==============================] - 0s 26ms/step\n",
      "1/1 [==============================] - 0s 26ms/step\n",
      "1/1 [==============================] - 0s 26ms/step\n",
      "1/1 [==============================] - 0s 27ms/step\n",
      "1/1 [==============================] - 0s 28ms/step\n",
      "1/1 [==============================] - 0s 26ms/step\n",
      "1/1 [==============================] - 0s 27ms/step\n",
      "1/1 [==============================] - 0s 28ms/step\n",
      "1/1 [==============================] - 0s 23ms/step\n",
      "1/1 [==============================] - 0s 30ms/step\n",
      "1/1 [==============================] - 0s 28ms/step\n",
      "1/1 [==============================] - 0s 25ms/step\n",
      "1/1 [==============================] - 0s 26ms/step\n",
      "1/1 [==============================] - 0s 25ms/step\n",
      "1/1 [==============================] - 0s 28ms/step\n",
      "1/1 [==============================] - 0s 25ms/step\n",
      "1/1 [==============================] - 0s 27ms/step\n",
      "1/1 [==============================] - 0s 27ms/step\n",
      "1/1 [==============================] - 0s 26ms/step\n",
      "1/1 [==============================] - 0s 25ms/step\n",
      "1/1 [==============================] - 0s 28ms/step\n",
      "1/1 [==============================] - 0s 24ms/step\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1/1 [==============================] - 0s 27ms/step\n",
      "1/1 [==============================] - 0s 23ms/step\n",
      "1/1 [==============================] - 0s 25ms/step\n",
      "1/1 [==============================] - 0s 29ms/step\n",
      "1/1 [==============================] - 0s 25ms/step\n",
      "1/1 [==============================] - 0s 28ms/step\n",
      "1/1 [==============================] - 0s 25ms/step\n",
      "1/1 [==============================] - 0s 26ms/step\n",
      "1/1 [==============================] - 0s 22ms/step\n",
      "1/1 [==============================] - 0s 27ms/step\n",
      "1/1 [==============================] - 0s 28ms/step\n",
      "1/1 [==============================] - 0s 30ms/step\n",
      "1/1 [==============================] - 0s 25ms/step\n",
      "1/1 [==============================] - 0s 27ms/step\n",
      "1/1 [==============================] - 0s 28ms/step\n",
      "1/1 [==============================] - 0s 24ms/step\n",
      "1/1 [==============================] - 0s 26ms/step\n",
      "1/1 [==============================] - 0s 26ms/step\n",
      "1/1 [==============================] - 0s 28ms/step\n",
      "1/1 [==============================] - 0s 26ms/step\n",
      "1/1 [==============================] - 0s 28ms/step\n",
      "1/1 [==============================] - 0s 28ms/step\n",
      "1/1 [==============================] - 0s 24ms/step\n",
      "1/1 [==============================] - 0s 27ms/step\n",
      "1/1 [==============================] - 0s 25ms/step\n",
      "1/1 [==============================] - 0s 27ms/step\n",
      "1/1 [==============================] - 0s 26ms/step\n",
      "1/1 [==============================] - 0s 23ms/step\n",
      "1/1 [==============================] - 0s 27ms/step\n",
      "1/1 [==============================] - 0s 29ms/step\n",
      "1/1 [==============================] - 0s 30ms/step\n",
      "1/1 [==============================] - 0s 24ms/step\n",
      "1/1 [==============================] - 0s 26ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 23ms/step\n",
      "1/1 [==============================] - 0s 28ms/step\n",
      "1/1 [==============================] - 0s 25ms/step\n",
      "1/1 [==============================] - 0s 24ms/step\n",
      "1/1 [==============================] - 0s 26ms/step\n",
      "1/1 [==============================] - 0s 25ms/step\n",
      "1/1 [==============================] - 0s 25ms/step\n",
      "1/1 [==============================] - 0s 22ms/step\n",
      "1/1 [==============================] - 0s 30ms/step\n",
      "1/1 [==============================] - 0s 26ms/step\n",
      "1/1 [==============================] - 0s 25ms/step\n",
      "1/1 [==============================] - 0s 27ms/step\n",
      "1/1 [==============================] - 0s 24ms/step\n",
      "1/1 [==============================] - 0s 30ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 29ms/step\n",
      "1/1 [==============================] - 0s 26ms/step\n",
      "1/1 [==============================] - 0s 30ms/step\n",
      "1/1 [==============================] - 0s 35ms/step\n",
      "1/1 [==============================] - 0s 28ms/step\n",
      "1/1 [==============================] - 0s 28ms/step\n",
      "1/1 [==============================] - 0s 30ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 26ms/step\n",
      "1/1 [==============================] - 0s 30ms/step\n",
      "1/1 [==============================] - 0s 24ms/step\n",
      "1/1 [==============================] - 0s 25ms/step\n",
      "1/1 [==============================] - 0s 26ms/step\n",
      "1/1 [==============================] - 0s 22ms/step\n",
      "1/1 [==============================] - 0s 26ms/step\n",
      "1/1 [==============================] - 0s 26ms/step\n",
      "1/1 [==============================] - 0s 24ms/step\n",
      "1/1 [==============================] - 0s 28ms/step\n",
      "1/1 [==============================] - 0s 26ms/step\n",
      "1/1 [==============================] - 0s 26ms/step\n",
      "1/1 [==============================] - 0s 24ms/step\n",
      "1/1 [==============================] - 0s 25ms/step\n",
      "1/1 [==============================] - 0s 25ms/step\n",
      "1/1 [==============================] - 0s 22ms/step\n",
      "1/1 [==============================] - 0s 22ms/step\n",
      "1/1 [==============================] - 0s 25ms/step\n",
      "1/1 [==============================] - 0s 25ms/step\n",
      "1/1 [==============================] - 0s 28ms/step\n",
      "1/1 [==============================] - 0s 26ms/step\n",
      "1/1 [==============================] - 0s 23ms/step\n",
      "1/1 [==============================] - 0s 26ms/step\n",
      "1/1 [==============================] - 0s 25ms/step\n",
      "1/1 [==============================] - 0s 23ms/step\n",
      "1/1 [==============================] - 0s 27ms/step\n",
      "1/1 [==============================] - 0s 26ms/step\n",
      "1/1 [==============================] - 0s 23ms/step\n",
      "1/1 [==============================] - 0s 26ms/step\n",
      "1/1 [==============================] - 0s 26ms/step\n",
      "1/1 [==============================] - 0s 27ms/step\n",
      "1/1 [==============================] - 0s 24ms/step\n",
      "1/1 [==============================] - 0s 24ms/step\n",
      "1/1 [==============================] - 0s 27ms/step\n",
      "1/1 [==============================] - 0s 26ms/step\n",
      "1/1 [==============================] - 0s 24ms/step\n",
      "1/1 [==============================] - 0s 27ms/step\n",
      "1/1 [==============================] - 0s 23ms/step\n",
      "1/1 [==============================] - 0s 24ms/step\n",
      "1/1 [==============================] - 0s 25ms/step\n",
      "1/1 [==============================] - 0s 24ms/step\n",
      "1/1 [==============================] - 0s 23ms/step\n",
      "1/1 [==============================] - 0s 23ms/step\n",
      "1/1 [==============================] - 0s 28ms/step\n",
      "1/1 [==============================] - 0s 23ms/step\n",
      "1/1 [==============================] - 0s 30ms/step\n",
      "1/1 [==============================] - 0s 27ms/step\n",
      "1/1 [==============================] - 0s 24ms/step\n",
      "1/1 [==============================] - 0s 24ms/step\n",
      "1/1 [==============================] - 0s 25ms/step\n",
      "1/1 [==============================] - 0s 25ms/step\n",
      "1/1 [==============================] - 0s 23ms/step\n",
      "1/1 [==============================] - 0s 23ms/step\n",
      "1/1 [==============================] - 0s 27ms/step\n",
      "1/1 [==============================] - 0s 26ms/step\n",
      "1/1 [==============================] - 0s 26ms/step\n",
      "1/1 [==============================] - 0s 25ms/step\n",
      "1/1 [==============================] - 0s 23ms/step\n",
      "1/1 [==============================] - 0s 27ms/step\n",
      "1/1 [==============================] - 0s 26ms/step\n",
      "1/1 [==============================] - 0s 23ms/step\n",
      "1/1 [==============================] - 0s 23ms/step\n",
      "1/1 [==============================] - 0s 27ms/step\n",
      "1/1 [==============================] - 0s 24ms/step\n",
      "1/1 [==============================] - 0s 27ms/step\n",
      "1/1 [==============================] - 0s 26ms/step\n",
      "1/1 [==============================] - 0s 25ms/step\n",
      "1/1 [==============================] - 0s 28ms/step\n",
      "1/1 [==============================] - 0s 25ms/step\n",
      "1/1 [==============================] - 0s 36ms/step\n",
      "1/1 [==============================] - 0s 24ms/step\n",
      "1/1 [==============================] - 0s 25ms/step\n",
      "Epoch 1/25\n",
      "1084/1084 [==============================] - 18s 14ms/step - loss: 0.0101\n",
      "Epoch 2/25\n",
      "1084/1084 [==============================] - 14s 13ms/step - loss: 0.0020\n",
      "Epoch 3/25\n",
      "1084/1084 [==============================] - 15s 14ms/step - loss: 0.0015\n",
      "Epoch 4/25\n",
      "1084/1084 [==============================] - 14s 13ms/step - loss: 0.0013\n",
      "Epoch 5/25\n",
      "1084/1084 [==============================] - 13s 12ms/step - loss: 0.0011\n",
      "Epoch 6/25\n",
      "1084/1084 [==============================] - 13s 12ms/step - loss: 9.4848e-04\n",
      "Epoch 7/25\n",
      "1084/1084 [==============================] - 13s 12ms/step - loss: 9.0584e-04\n",
      "Epoch 8/25\n",
      "1084/1084 [==============================] - 14s 13ms/step - loss: 8.6227e-04\n",
      "Epoch 9/25\n",
      "1084/1084 [==============================] - 13s 12ms/step - loss: 8.2601e-04\n",
      "Epoch 10/25\n",
      "1084/1084 [==============================] - 13s 12ms/step - loss: 7.7634e-04\n",
      "Epoch 11/25\n",
      "1084/1084 [==============================] - 13s 12ms/step - loss: 7.9830e-04\n",
      "Epoch 12/25\n",
      "1084/1084 [==============================] - 13s 12ms/step - loss: 7.5966e-04\n",
      "Epoch 13/25\n",
      "1084/1084 [==============================] - 13s 12ms/step - loss: 7.7634e-04\n",
      "Epoch 14/25\n",
      "1084/1084 [==============================] - 13s 12ms/step - loss: 7.1719e-04\n",
      "Epoch 15/25\n",
      "1084/1084 [==============================] - 13s 12ms/step - loss: 7.6480e-04\n",
      "Epoch 16/25\n",
      "1084/1084 [==============================] - 13s 12ms/step - loss: 7.9215e-04\n",
      "Epoch 17/25\n",
      "1084/1084 [==============================] - 12s 11ms/step - loss: 7.3779e-04\n",
      "Epoch 18/25\n",
      "1084/1084 [==============================] - 13s 12ms/step - loss: 7.3509e-04\n",
      "Epoch 19/25\n",
      "1084/1084 [==============================] - 13s 12ms/step - loss: 7.3724e-04\n",
      "Epoch 20/25\n",
      "1084/1084 [==============================] - 13s 12ms/step - loss: 7.2794e-04\n",
      "Epoch 21/25\n",
      "1084/1084 [==============================] - 13s 12ms/step - loss: 7.4737e-04\n",
      "Epoch 22/25\n",
      "1084/1084 [==============================] - 14s 13ms/step - loss: 6.9129e-04\n",
      "Epoch 23/25\n",
      "1084/1084 [==============================] - 13s 12ms/step - loss: 6.6932e-04\n",
      "Epoch 24/25\n",
      "1084/1084 [==============================] - 13s 12ms/step - loss: 6.9522e-04\n",
      "Epoch 25/25\n",
      "1084/1084 [==============================] - 12s 11ms/step - loss: 7.0039e-04\n",
      "1/1 [==============================] - 0s 191ms/step\n",
      "1/1 [==============================] - 0s 25ms/step\n",
      "1/1 [==============================] - 0s 24ms/step\n",
      "1/1 [==============================] - 0s 38ms/step\n",
      "1/1 [==============================] - 0s 23ms/step\n",
      "1/1 [==============================] - 0s 22ms/step\n",
      "1/1 [==============================] - 0s 25ms/step\n",
      "1/1 [==============================] - 0s 24ms/step\n",
      "1/1 [==============================] - 0s 23ms/step\n",
      "1/1 [==============================] - 0s 25ms/step\n",
      "1/1 [==============================] - 0s 23ms/step\n",
      "1/1 [==============================] - 0s 24ms/step\n",
      "1/1 [==============================] - 0s 25ms/step\n",
      "1/1 [==============================] - 0s 22ms/step\n",
      "1/1 [==============================] - 0s 23ms/step\n",
      "1/1 [==============================] - 0s 24ms/step\n",
      "1/1 [==============================] - 0s 25ms/step\n",
      "1/1 [==============================] - 0s 23ms/step\n",
      "1/1 [==============================] - 0s 23ms/step\n",
      "1/1 [==============================] - 0s 23ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 24ms/step\n",
      "1/1 [==============================] - 0s 23ms/step\n",
      "1/1 [==============================] - 0s 23ms/step\n",
      "1/1 [==============================] - 0s 22ms/step\n",
      "1/1 [==============================] - 0s 24ms/step\n",
      "1/1 [==============================] - 0s 23ms/step\n",
      "1/1 [==============================] - 0s 22ms/step\n",
      "1/1 [==============================] - 0s 22ms/step\n",
      "1/1 [==============================] - 0s 22ms/step\n",
      "1/1 [==============================] - 0s 23ms/step\n",
      "1/1 [==============================] - 0s 23ms/step\n",
      "1/1 [==============================] - 0s 23ms/step\n",
      "1/1 [==============================] - 0s 23ms/step\n",
      "1/1 [==============================] - 0s 26ms/step\n",
      "1/1 [==============================] - 0s 26ms/step\n",
      "1/1 [==============================] - 0s 30ms/step\n",
      "1/1 [==============================] - 0s 28ms/step\n",
      "1/1 [==============================] - 0s 28ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 24ms/step\n",
      "1/1 [==============================] - 0s 24ms/step\n",
      "1/1 [==============================] - 0s 23ms/step\n",
      "1/1 [==============================] - 0s 25ms/step\n",
      "1/1 [==============================] - 0s 24ms/step\n",
      "1/1 [==============================] - 0s 24ms/step\n",
      "1/1 [==============================] - 0s 27ms/step\n",
      "1/1 [==============================] - 0s 24ms/step\n",
      "1/1 [==============================] - 0s 25ms/step\n",
      "1/1 [==============================] - 0s 23ms/step\n",
      "1/1 [==============================] - 0s 29ms/step\n",
      "1/1 [==============================] - 0s 24ms/step\n",
      "1/1 [==============================] - 0s 23ms/step\n",
      "1/1 [==============================] - 0s 24ms/step\n",
      "1/1 [==============================] - 0s 24ms/step\n",
      "1/1 [==============================] - 0s 27ms/step\n",
      "1/1 [==============================] - 0s 24ms/step\n",
      "1/1 [==============================] - 0s 21ms/step\n",
      "1/1 [==============================] - 0s 24ms/step\n",
      "1/1 [==============================] - 0s 26ms/step\n",
      "1/1 [==============================] - 0s 23ms/step\n",
      "1/1 [==============================] - 0s 24ms/step\n",
      "1/1 [==============================] - 0s 27ms/step\n",
      "1/1 [==============================] - 0s 26ms/step\n",
      "1/1 [==============================] - 0s 23ms/step\n",
      "1/1 [==============================] - 0s 22ms/step\n",
      "1/1 [==============================] - 0s 22ms/step\n",
      "1/1 [==============================] - 0s 24ms/step\n",
      "1/1 [==============================] - 0s 24ms/step\n",
      "1/1 [==============================] - 0s 23ms/step\n",
      "1/1 [==============================] - 0s 23ms/step\n",
      "1/1 [==============================] - 0s 23ms/step\n",
      "1/1 [==============================] - 0s 22ms/step\n",
      "1/1 [==============================] - 0s 24ms/step\n",
      "1/1 [==============================] - 0s 24ms/step\n",
      "1/1 [==============================] - 0s 24ms/step\n",
      "1/1 [==============================] - 0s 25ms/step\n",
      "1/1 [==============================] - 0s 26ms/step\n",
      "1/1 [==============================] - 0s 25ms/step\n",
      "1/1 [==============================] - 0s 24ms/step\n",
      "1/1 [==============================] - 0s 23ms/step\n",
      "1/1 [==============================] - 0s 25ms/step\n",
      "1/1 [==============================] - 0s 27ms/step\n",
      "1/1 [==============================] - 0s 25ms/step\n",
      "1/1 [==============================] - 0s 24ms/step\n",
      "1/1 [==============================] - 0s 23ms/step\n",
      "1/1 [==============================] - 0s 24ms/step\n",
      "1/1 [==============================] - 0s 23ms/step\n",
      "1/1 [==============================] - 0s 25ms/step\n",
      "1/1 [==============================] - 0s 22ms/step\n",
      "1/1 [==============================] - 0s 25ms/step\n",
      "1/1 [==============================] - 0s 24ms/step\n",
      "1/1 [==============================] - 0s 24ms/step\n",
      "1/1 [==============================] - 0s 24ms/step\n",
      "1/1 [==============================] - 0s 23ms/step\n",
      "1/1 [==============================] - 0s 24ms/step\n",
      "1/1 [==============================] - 0s 23ms/step\n",
      "1/1 [==============================] - 0s 24ms/step\n",
      "1/1 [==============================] - 0s 24ms/step\n",
      "1/1 [==============================] - 0s 23ms/step\n",
      "1/1 [==============================] - 0s 25ms/step\n",
      "1/1 [==============================] - 0s 25ms/step\n",
      "1/1 [==============================] - 0s 24ms/step\n",
      "1/1 [==============================] - 0s 23ms/step\n",
      "1/1 [==============================] - 0s 28ms/step\n",
      "1/1 [==============================] - 0s 28ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 30ms/step\n",
      "1/1 [==============================] - 0s 29ms/step\n",
      "1/1 [==============================] - 0s 27ms/step\n",
      "1/1 [==============================] - 0s 25ms/step\n",
      "1/1 [==============================] - 0s 23ms/step\n",
      "1/1 [==============================] - 0s 22ms/step\n",
      "1/1 [==============================] - 0s 24ms/step\n",
      "1/1 [==============================] - 0s 24ms/step\n",
      "1/1 [==============================] - 0s 28ms/step\n",
      "1/1 [==============================] - 0s 27ms/step\n",
      "1/1 [==============================] - 0s 26ms/step\n",
      "1/1 [==============================] - 0s 25ms/step\n",
      "1/1 [==============================] - 0s 25ms/step\n",
      "1/1 [==============================] - 0s 25ms/step\n",
      "1/1 [==============================] - 0s 24ms/step\n",
      "1/1 [==============================] - 0s 32ms/step\n",
      "1/1 [==============================] - 0s 29ms/step\n",
      "1/1 [==============================] - 0s 23ms/step\n",
      "1/1 [==============================] - 0s 28ms/step\n",
      "1/1 [==============================] - 0s 27ms/step\n",
      "1/1 [==============================] - 0s 29ms/step\n",
      "1/1 [==============================] - 0s 25ms/step\n",
      "1/1 [==============================] - 0s 28ms/step\n",
      "1/1 [==============================] - 0s 22ms/step\n",
      "1/1 [==============================] - 0s 23ms/step\n",
      "1/1 [==============================] - 0s 32ms/step\n",
      "1/1 [==============================] - 0s 26ms/step\n",
      "1/1 [==============================] - 0s 24ms/step\n",
      "1/1 [==============================] - 0s 23ms/step\n",
      "1/1 [==============================] - 0s 24ms/step\n",
      "1/1 [==============================] - 0s 25ms/step\n",
      "1/1 [==============================] - 0s 25ms/step\n",
      "1/1 [==============================] - 0s 23ms/step\n",
      "1/1 [==============================] - 0s 22ms/step\n",
      "1/1 [==============================] - 0s 23ms/step\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1/1 [==============================] - 0s 24ms/step\n",
      "1/1 [==============================] - 0s 24ms/step\n",
      "1/1 [==============================] - 0s 22ms/step\n",
      "1/1 [==============================] - 0s 23ms/step\n",
      "1/1 [==============================] - 0s 24ms/step\n",
      "1/1 [==============================] - 0s 24ms/step\n",
      "1/1 [==============================] - 0s 25ms/step\n",
      "1/1 [==============================] - 0s 25ms/step\n",
      "1/1 [==============================] - 0s 25ms/step\n",
      "1/1 [==============================] - 0s 24ms/step\n",
      "1/1 [==============================] - 0s 23ms/step\n",
      "1/1 [==============================] - 0s 24ms/step\n",
      "1/1 [==============================] - 0s 23ms/step\n",
      "1/1 [==============================] - 0s 24ms/step\n",
      "1/1 [==============================] - 0s 23ms/step\n",
      "1/1 [==============================] - 0s 24ms/step\n",
      "1/1 [==============================] - 0s 25ms/step\n",
      "1/1 [==============================] - 0s 23ms/step\n",
      "1/1 [==============================] - 0s 23ms/step\n",
      "1/1 [==============================] - 0s 21ms/step\n",
      "1/1 [==============================] - 0s 23ms/step\n",
      "1/1 [==============================] - 0s 22ms/step\n",
      "1/1 [==============================] - 0s 22ms/step\n",
      "1/1 [==============================] - 0s 25ms/step\n",
      "1/1 [==============================] - 0s 24ms/step\n",
      "1/1 [==============================] - 0s 25ms/step\n",
      "1/1 [==============================] - 0s 24ms/step\n",
      "1/1 [==============================] - 0s 23ms/step\n",
      "1/1 [==============================] - 0s 24ms/step\n",
      "1/1 [==============================] - 0s 23ms/step\n",
      "1/1 [==============================] - 0s 23ms/step\n",
      "1/1 [==============================] - 0s 23ms/step\n",
      "1/1 [==============================] - 0s 25ms/step\n",
      "1/1 [==============================] - 0s 25ms/step\n",
      "1/1 [==============================] - 0s 22ms/step\n",
      "1/1 [==============================] - 0s 25ms/step\n",
      "1/1 [==============================] - 0s 33ms/step\n",
      "1/1 [==============================] - 0s 23ms/step\n",
      "Epoch 1/25\n",
      "1074/1074 [==============================] - 18s 15ms/step - loss: 0.0068\n",
      "Epoch 2/25\n",
      "1074/1074 [==============================] - 16s 15ms/step - loss: 0.0017\n",
      "Epoch 3/25\n",
      "1074/1074 [==============================] - 17s 15ms/step - loss: 0.0013\n",
      "Epoch 4/25\n",
      "1074/1074 [==============================] - 16s 15ms/step - loss: 0.0012\n",
      "Epoch 5/25\n",
      "1074/1074 [==============================] - 15s 14ms/step - loss: 0.0010\n",
      "Epoch 6/25\n",
      "1074/1074 [==============================] - 15s 14ms/step - loss: 9.4837e-04\n",
      "Epoch 7/25\n",
      "1074/1074 [==============================] - 16s 15ms/step - loss: 8.7465e-04\n",
      "Epoch 8/25\n",
      "1074/1074 [==============================] - 16s 15ms/step - loss: 8.3762e-04\n",
      "Epoch 9/25\n",
      "1074/1074 [==============================] - 16s 15ms/step - loss: 8.2851e-04\n",
      "Epoch 10/25\n",
      "1074/1074 [==============================] - 16s 15ms/step - loss: 7.6305e-04\n",
      "Epoch 11/25\n",
      "1074/1074 [==============================] - 15s 14ms/step - loss: 7.4449e-04\n",
      "Epoch 12/25\n",
      "1074/1074 [==============================] - 15s 14ms/step - loss: 7.9077e-04\n",
      "Epoch 13/25\n",
      "1074/1074 [==============================] - 16s 15ms/step - loss: 7.6420e-04\n",
      "Epoch 14/25\n",
      "1074/1074 [==============================] - 15s 14ms/step - loss: 7.6746e-04\n",
      "Epoch 15/25\n",
      "1074/1074 [==============================] - 15s 14ms/step - loss: 7.8215e-04\n",
      "Epoch 16/25\n",
      "1074/1074 [==============================] - 16s 15ms/step - loss: 7.2598e-04\n",
      "Epoch 17/25\n",
      "1074/1074 [==============================] - 15s 14ms/step - loss: 7.4948e-04\n",
      "Epoch 18/25\n",
      "1074/1074 [==============================] - 15s 14ms/step - loss: 7.3759e-04\n",
      "Epoch 19/25\n",
      "1074/1074 [==============================] - 16s 15ms/step - loss: 6.7897e-04\n",
      "Epoch 20/25\n",
      "1074/1074 [==============================] - 16s 15ms/step - loss: 7.2599e-04\n",
      "Epoch 21/25\n",
      "1074/1074 [==============================] - 15s 14ms/step - loss: 7.4053e-04\n",
      "Epoch 22/25\n",
      "1074/1074 [==============================] - 16s 15ms/step - loss: 6.8550e-04\n",
      "Epoch 23/25\n",
      "1074/1074 [==============================] - 15s 14ms/step - loss: 7.3021e-04\n",
      "Epoch 24/25\n",
      "1074/1074 [==============================] - 16s 15ms/step - loss: 6.8209e-04\n",
      "Epoch 25/25\n",
      "1074/1074 [==============================] - 17s 16ms/step - loss: 7.1403e-04\n",
      "1/1 [==============================] - 0s 215ms/step\n",
      "1/1 [==============================] - 0s 25ms/step\n",
      "1/1 [==============================] - 0s 26ms/step\n",
      "1/1 [==============================] - 0s 25ms/step\n",
      "1/1 [==============================] - 0s 23ms/step\n",
      "1/1 [==============================] - 0s 29ms/step\n",
      "1/1 [==============================] - 0s 24ms/step\n",
      "1/1 [==============================] - 0s 23ms/step\n",
      "1/1 [==============================] - 0s 26ms/step\n",
      "1/1 [==============================] - 0s 25ms/step\n",
      "1/1 [==============================] - 0s 23ms/step\n",
      "1/1 [==============================] - 0s 25ms/step\n",
      "1/1 [==============================] - 0s 24ms/step\n",
      "1/1 [==============================] - 0s 23ms/step\n",
      "1/1 [==============================] - 0s 24ms/step\n",
      "1/1 [==============================] - 0s 29ms/step\n",
      "1/1 [==============================] - 0s 25ms/step\n",
      "1/1 [==============================] - 0s 25ms/step\n",
      "1/1 [==============================] - 0s 22ms/step\n",
      "1/1 [==============================] - 0s 22ms/step\n",
      "1/1 [==============================] - 0s 23ms/step\n",
      "1/1 [==============================] - 0s 24ms/step\n",
      "1/1 [==============================] - 0s 23ms/step\n",
      "1/1 [==============================] - 0s 25ms/step\n",
      "1/1 [==============================] - 0s 25ms/step\n",
      "1/1 [==============================] - 0s 25ms/step\n",
      "1/1 [==============================] - 0s 24ms/step\n",
      "1/1 [==============================] - 0s 27ms/step\n",
      "1/1 [==============================] - 0s 25ms/step\n",
      "1/1 [==============================] - 0s 22ms/step\n",
      "1/1 [==============================] - 0s 23ms/step\n",
      "1/1 [==============================] - 0s 24ms/step\n",
      "1/1 [==============================] - 0s 22ms/step\n",
      "1/1 [==============================] - 0s 23ms/step\n",
      "1/1 [==============================] - 0s 26ms/step\n",
      "1/1 [==============================] - 0s 35ms/step\n",
      "1/1 [==============================] - 0s 26ms/step\n",
      "1/1 [==============================] - 0s 24ms/step\n",
      "1/1 [==============================] - 0s 30ms/step\n",
      "1/1 [==============================] - 0s 29ms/step\n",
      "1/1 [==============================] - 0s 29ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 28ms/step\n",
      "1/1 [==============================] - 0s 28ms/step\n",
      "1/1 [==============================] - 0s 32ms/step\n",
      "1/1 [==============================] - 0s 29ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 32ms/step\n",
      "1/1 [==============================] - 0s 30ms/step\n",
      "1/1 [==============================] - 0s 30ms/step\n",
      "1/1 [==============================] - 0s 35ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 28ms/step\n",
      "1/1 [==============================] - 0s 43ms/step\n",
      "1/1 [==============================] - 0s 29ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 29ms/step\n",
      "1/1 [==============================] - 0s 29ms/step\n",
      "1/1 [==============================] - 0s 32ms/step\n",
      "1/1 [==============================] - 0s 30ms/step\n",
      "1/1 [==============================] - 0s 30ms/step\n",
      "1/1 [==============================] - 0s 30ms/step\n",
      "1/1 [==============================] - 0s 35ms/step\n",
      "1/1 [==============================] - 0s 28ms/step\n",
      "1/1 [==============================] - 0s 35ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 24ms/step\n",
      "1/1 [==============================] - 0s 26ms/step\n",
      "1/1 [==============================] - 0s 24ms/step\n",
      "1/1 [==============================] - 0s 24ms/step\n",
      "1/1 [==============================] - 0s 23ms/step\n",
      "1/1 [==============================] - 0s 22ms/step\n",
      "1/1 [==============================] - 0s 25ms/step\n",
      "1/1 [==============================] - 0s 27ms/step\n",
      "1/1 [==============================] - 0s 29ms/step\n",
      "1/1 [==============================] - 0s 24ms/step\n",
      "1/1 [==============================] - 0s 25ms/step\n",
      "1/1 [==============================] - 0s 25ms/step\n",
      "1/1 [==============================] - 0s 26ms/step\n",
      "1/1 [==============================] - 0s 26ms/step\n",
      "1/1 [==============================] - 0s 28ms/step\n",
      "1/1 [==============================] - 0s 23ms/step\n",
      "1/1 [==============================] - 0s 26ms/step\n",
      "1/1 [==============================] - 0s 29ms/step\n",
      "1/1 [==============================] - 0s 26ms/step\n",
      "1/1 [==============================] - 0s 24ms/step\n",
      "1/1 [==============================] - 0s 26ms/step\n",
      "1/1 [==============================] - 0s 28ms/step\n",
      "1/1 [==============================] - 0s 25ms/step\n",
      "1/1 [==============================] - 0s 24ms/step\n",
      "1/1 [==============================] - 0s 24ms/step\n",
      "1/1 [==============================] - 0s 23ms/step\n",
      "1/1 [==============================] - 0s 24ms/step\n",
      "1/1 [==============================] - 0s 24ms/step\n",
      "1/1 [==============================] - 0s 25ms/step\n",
      "1/1 [==============================] - 0s 23ms/step\n",
      "1/1 [==============================] - 0s 23ms/step\n",
      "1/1 [==============================] - 0s 24ms/step\n",
      "1/1 [==============================] - 0s 25ms/step\n",
      "1/1 [==============================] - 0s 23ms/step\n",
      "1/1 [==============================] - 0s 61ms/step\n",
      "1/1 [==============================] - 0s 33ms/step\n",
      "1/1 [==============================] - 0s 30ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 30ms/step\n",
      "1/1 [==============================] - 0s 27ms/step\n",
      "1/1 [==============================] - 0s 30ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 28ms/step\n",
      "1/1 [==============================] - 0s 29ms/step\n",
      "1/1 [==============================] - 0s 26ms/step\n",
      "1/1 [==============================] - 0s 32ms/step\n",
      "1/1 [==============================] - 0s 28ms/step\n",
      "1/1 [==============================] - 0s 26ms/step\n",
      "1/1 [==============================] - 0s 27ms/step\n",
      "1/1 [==============================] - 0s 27ms/step\n",
      "1/1 [==============================] - 0s 32ms/step\n",
      "1/1 [==============================] - 0s 30ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 30ms/step\n",
      "1/1 [==============================] - 0s 30ms/step\n",
      "1/1 [==============================] - 0s 33ms/step\n",
      "1/1 [==============================] - 0s 35ms/step\n",
      "1/1 [==============================] - 0s 33ms/step\n",
      "1/1 [==============================] - 0s 29ms/step\n",
      "1/1 [==============================] - 0s 32ms/step\n",
      "1/1 [==============================] - 0s 32ms/step\n",
      "1/1 [==============================] - 0s 40ms/step\n",
      "1/1 [==============================] - 0s 28ms/step\n",
      "1/1 [==============================] - 0s 34ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 34ms/step\n",
      "1/1 [==============================] - 0s 34ms/step\n",
      "1/1 [==============================] - 0s 29ms/step\n",
      "1/1 [==============================] - 0s 29ms/step\n",
      "1/1 [==============================] - 0s 38ms/step\n",
      "1/1 [==============================] - 0s 37ms/step\n",
      "1/1 [==============================] - 0s 27ms/step\n",
      "1/1 [==============================] - 0s 28ms/step\n",
      "1/1 [==============================] - 0s 36ms/step\n",
      "1/1 [==============================] - 0s 29ms/step\n",
      "1/1 [==============================] - 0s 24ms/step\n",
      "1/1 [==============================] - 0s 30ms/step\n",
      "1/1 [==============================] - 0s 30ms/step\n",
      "1/1 [==============================] - 0s 25ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 37ms/step\n",
      "1/1 [==============================] - 0s 27ms/step\n",
      "1/1 [==============================] - 0s 27ms/step\n",
      "1/1 [==============================] - 0s 25ms/step\n",
      "1/1 [==============================] - 0s 27ms/step\n",
      "1/1 [==============================] - 0s 27ms/step\n",
      "1/1 [==============================] - 0s 32ms/step\n",
      "1/1 [==============================] - 0s 28ms/step\n",
      "1/1 [==============================] - 0s 27ms/step\n",
      "1/1 [==============================] - 0s 27ms/step\n",
      "1/1 [==============================] - 0s 26ms/step\n",
      "1/1 [==============================] - 0s 30ms/step\n",
      "1/1 [==============================] - 0s 32ms/step\n",
      "1/1 [==============================] - 0s 24ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 27ms/step\n",
      "1/1 [==============================] - 0s 34ms/step\n",
      "1/1 [==============================] - 0s 30ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 30ms/step\n",
      "1/1 [==============================] - 0s 33ms/step\n",
      "1/1 [==============================] - 0s 32ms/step\n",
      "1/1 [==============================] - 0s 26ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 26ms/step\n",
      "1/1 [==============================] - 0s 23ms/step\n",
      "1/1 [==============================] - 0s 29ms/step\n",
      "1/1 [==============================] - 0s 27ms/step\n",
      "1/1 [==============================] - 0s 24ms/step\n",
      "1/1 [==============================] - 0s 28ms/step\n",
      "1/1 [==============================] - 0s 25ms/step\n",
      "1/1 [==============================] - 0s 28ms/step\n",
      "1/1 [==============================] - 0s 28ms/step\n",
      "1/1 [==============================] - 0s 25ms/step\n",
      "Epoch 1/25\n",
      "1064/1064 [==============================] - 21s 18ms/step - loss: 0.0067\n",
      "Epoch 2/25\n",
      "1064/1064 [==============================] - 17s 16ms/step - loss: 0.0018\n",
      "Epoch 3/25\n",
      "1064/1064 [==============================] - 17s 16ms/step - loss: 0.0014\n",
      "Epoch 4/25\n",
      "1064/1064 [==============================] - 17s 16ms/step - loss: 0.0012\n",
      "Epoch 5/25\n",
      "1064/1064 [==============================] - 17s 16ms/step - loss: 0.0011\n",
      "Epoch 6/25\n",
      "1064/1064 [==============================] - 18s 17ms/step - loss: 9.0337e-04\n",
      "Epoch 7/25\n",
      "1064/1064 [==============================] - 17s 16ms/step - loss: 9.1092e-04\n",
      "Epoch 8/25\n",
      "1064/1064 [==============================] - 18s 17ms/step - loss: 9.1820e-04\n",
      "Epoch 9/25\n",
      "1064/1064 [==============================] - 18s 17ms/step - loss: 8.4455e-04\n",
      "Epoch 10/25\n",
      "1064/1064 [==============================] - 18s 17ms/step - loss: 7.4183e-04\n",
      "Epoch 11/25\n",
      "1064/1064 [==============================] - 18s 17ms/step - loss: 7.7801e-04\n",
      "Epoch 12/25\n",
      "1064/1064 [==============================] - 18s 17ms/step - loss: 7.1130e-04\n",
      "Epoch 13/25\n",
      "1064/1064 [==============================] - 17s 16ms/step - loss: 7.3468e-04\n",
      "Epoch 14/25\n",
      "1064/1064 [==============================] - 17s 16ms/step - loss: 7.3675e-04\n",
      "Epoch 15/25\n",
      "1064/1064 [==============================] - 17s 16ms/step - loss: 7.5766e-04\n",
      "Epoch 16/25\n",
      "1064/1064 [==============================] - 17s 16ms/step - loss: 7.2822e-04\n",
      "Epoch 17/25\n",
      "1064/1064 [==============================] - 17s 16ms/step - loss: 7.3595e-04\n",
      "Epoch 18/25\n",
      "1064/1064 [==============================] - 25s 24ms/step - loss: 7.3552e-04\n",
      "Epoch 19/25\n",
      "1064/1064 [==============================] - 30s 28ms/step - loss: 7.3621e-04\n",
      "Epoch 20/25\n",
      "1064/1064 [==============================] - 20s 19ms/step - loss: 6.9466e-04\n",
      "Epoch 21/25\n",
      "1064/1064 [==============================] - 20s 19ms/step - loss: 7.2850e-04\n",
      "Epoch 22/25\n",
      "1064/1064 [==============================] - 20s 19ms/step - loss: 6.6244e-04\n",
      "Epoch 23/25\n",
      "1064/1064 [==============================] - 21s 19ms/step - loss: 7.2724e-04\n",
      "Epoch 24/25\n",
      "1064/1064 [==============================] - 25s 24ms/step - loss: 7.0473e-04\n",
      "Epoch 25/25\n",
      "1064/1064 [==============================] - 21s 19ms/step - loss: 7.4343e-04\n",
      "1/1 [==============================] - 0s 316ms/step\n",
      "1/1 [==============================] - 0s 44ms/step\n",
      "1/1 [==============================] - 0s 25ms/step\n",
      "1/1 [==============================] - 0s 58ms/step\n",
      "1/1 [==============================] - 0s 61ms/step\n",
      "1/1 [==============================] - 0s 78ms/step\n",
      "1/1 [==============================] - 0s 78ms/step\n",
      "1/1 [==============================] - 0s 78ms/step\n",
      "1/1 [==============================] - 0s 64ms/step\n",
      "1/1 [==============================] - 0s 52ms/step\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1/1 [==============================] - 0s 67ms/step\n",
      "1/1 [==============================] - 0s 82ms/step\n",
      "1/1 [==============================] - 0s 71ms/step\n",
      "1/1 [==============================] - 0s 87ms/step\n",
      "1/1 [==============================] - 0s 78ms/step\n",
      "1/1 [==============================] - 0s 71ms/step\n",
      "1/1 [==============================] - 0s 56ms/step\n",
      "1/1 [==============================] - 0s 42ms/step\n",
      "1/1 [==============================] - 0s 71ms/step\n",
      "1/1 [==============================] - 0s 45ms/step\n",
      "1/1 [==============================] - 0s 77ms/step\n",
      "1/1 [==============================] - 0s 81ms/step\n",
      "1/1 [==============================] - 0s 82ms/step\n",
      "1/1 [==============================] - 0s 34ms/step\n",
      "1/1 [==============================] - 0s 37ms/step\n",
      "1/1 [==============================] - 0s 33ms/step\n",
      "1/1 [==============================] - 0s 24ms/step\n",
      "1/1 [==============================] - 0s 26ms/step\n",
      "1/1 [==============================] - 0s 54ms/step\n",
      "1/1 [==============================] - 0s 46ms/step\n",
      "1/1 [==============================] - 0s 72ms/step\n",
      "1/1 [==============================] - 0s 94ms/step\n",
      "1/1 [==============================] - 0s 30ms/step\n",
      "1/1 [==============================] - 0s 53ms/step\n",
      "1/1 [==============================] - 0s 80ms/step\n",
      "1/1 [==============================] - 0s 81ms/step\n",
      "1/1 [==============================] - 0s 61ms/step\n",
      "1/1 [==============================] - 0s 41ms/step\n",
      "1/1 [==============================] - 0s 26ms/step\n",
      "1/1 [==============================] - 0s 63ms/step\n",
      "1/1 [==============================] - 0s 60ms/step\n",
      "1/1 [==============================] - 0s 42ms/step\n",
      "1/1 [==============================] - 0s 27ms/step\n",
      "1/1 [==============================] - 0s 28ms/step\n",
      "1/1 [==============================] - 0s 22ms/step\n",
      "1/1 [==============================] - 0s 67ms/step\n",
      "1/1 [==============================] - 0s 69ms/step\n",
      "1/1 [==============================] - 0s 56ms/step\n",
      "1/1 [==============================] - 0s 27ms/step\n",
      "1/1 [==============================] - 0s 47ms/step\n",
      "1/1 [==============================] - 0s 64ms/step\n",
      "1/1 [==============================] - 0s 92ms/step\n",
      "1/1 [==============================] - 0s 56ms/step\n",
      "1/1 [==============================] - 0s 88ms/step\n",
      "1/1 [==============================] - 0s 28ms/step\n",
      "1/1 [==============================] - 0s 28ms/step\n",
      "1/1 [==============================] - 0s 26ms/step\n",
      "1/1 [==============================] - 0s 27ms/step\n",
      "1/1 [==============================] - 0s 27ms/step\n",
      "1/1 [==============================] - 0s 26ms/step\n",
      "1/1 [==============================] - 0s 22ms/step\n",
      "1/1 [==============================] - 0s 54ms/step\n",
      "1/1 [==============================] - 0s 87ms/step\n",
      "1/1 [==============================] - 0s 33ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 28ms/step\n",
      "1/1 [==============================] - 0s 28ms/step\n",
      "1/1 [==============================] - 0s 26ms/step\n",
      "1/1 [==============================] - 0s 29ms/step\n",
      "1/1 [==============================] - 0s 52ms/step\n",
      "1/1 [==============================] - 0s 82ms/step\n",
      "1/1 [==============================] - 0s 69ms/step\n",
      "1/1 [==============================] - 0s 70ms/step\n",
      "1/1 [==============================] - 0s 54ms/step\n",
      "1/1 [==============================] - 0s 65ms/step\n",
      "1/1 [==============================] - 0s 72ms/step\n",
      "1/1 [==============================] - 0s 52ms/step\n",
      "1/1 [==============================] - 0s 74ms/step\n",
      "1/1 [==============================] - 0s 84ms/step\n",
      "1/1 [==============================] - 0s 52ms/step\n",
      "1/1 [==============================] - 0s 49ms/step\n",
      "1/1 [==============================] - 0s 29ms/step\n",
      "1/1 [==============================] - 0s 75ms/step\n",
      "1/1 [==============================] - 0s 81ms/step\n",
      "1/1 [==============================] - 0s 28ms/step\n",
      "1/1 [==============================] - 0s 28ms/step\n",
      "1/1 [==============================] - 0s 80ms/step\n",
      "1/1 [==============================] - 0s 82ms/step\n",
      "1/1 [==============================] - 0s 27ms/step\n",
      "1/1 [==============================] - 0s 80ms/step\n",
      "1/1 [==============================] - 0s 37ms/step\n",
      "1/1 [==============================] - 0s 52ms/step\n",
      "1/1 [==============================] - 0s 83ms/step\n",
      "1/1 [==============================] - 0s 30ms/step\n",
      "1/1 [==============================] - 0s 26ms/step\n",
      "1/1 [==============================] - 0s 41ms/step\n",
      "1/1 [==============================] - 0s 36ms/step\n",
      "1/1 [==============================] - 0s 25ms/step\n",
      "1/1 [==============================] - 0s 26ms/step\n",
      "1/1 [==============================] - 0s 22ms/step\n",
      "1/1 [==============================] - 0s 28ms/step\n",
      "1/1 [==============================] - 0s 25ms/step\n",
      "1/1 [==============================] - 0s 51ms/step\n",
      "1/1 [==============================] - 0s 66ms/step\n",
      "1/1 [==============================] - 0s 67ms/step\n",
      "1/1 [==============================] - 0s 44ms/step\n",
      "1/1 [==============================] - 0s 24ms/step\n",
      "1/1 [==============================] - 0s 42ms/step\n",
      "1/1 [==============================] - 0s 73ms/step\n",
      "1/1 [==============================] - 0s 85ms/step\n",
      "1/1 [==============================] - 0s 25ms/step\n",
      "1/1 [==============================] - 0s 54ms/step\n",
      "1/1 [==============================] - 0s 41ms/step\n",
      "1/1 [==============================] - 0s 24ms/step\n",
      "1/1 [==============================] - 0s 25ms/step\n",
      "1/1 [==============================] - 0s 48ms/step\n",
      "1/1 [==============================] - 0s 42ms/step\n",
      "1/1 [==============================] - 0s 56ms/step\n",
      "1/1 [==============================] - 0s 35ms/step\n",
      "1/1 [==============================] - 0s 83ms/step\n",
      "1/1 [==============================] - 0s 67ms/step\n",
      "1/1 [==============================] - 0s 81ms/step\n",
      "1/1 [==============================] - 0s 71ms/step\n",
      "1/1 [==============================] - 0s 26ms/step\n",
      "1/1 [==============================] - 0s 39ms/step\n",
      "1/1 [==============================] - 0s 72ms/step\n",
      "1/1 [==============================] - 0s 82ms/step\n",
      "1/1 [==============================] - 0s 39ms/step\n",
      "1/1 [==============================] - 0s 26ms/step\n",
      "1/1 [==============================] - 0s 26ms/step\n",
      "1/1 [==============================] - 0s 26ms/step\n",
      "1/1 [==============================] - 0s 63ms/step\n",
      "1/1 [==============================] - 0s 82ms/step\n",
      "1/1 [==============================] - 0s 84ms/step\n",
      "1/1 [==============================] - 0s 89ms/step\n",
      "1/1 [==============================] - 0s 53ms/step\n",
      "1/1 [==============================] - 0s 67ms/step\n",
      "1/1 [==============================] - 0s 61ms/step\n",
      "1/1 [==============================] - 0s 83ms/step\n",
      "1/1 [==============================] - 0s 77ms/step\n",
      "1/1 [==============================] - 0s 71ms/step\n",
      "1/1 [==============================] - 0s 41ms/step\n",
      "1/1 [==============================] - 0s 24ms/step\n",
      "1/1 [==============================] - 0s 78ms/step\n",
      "1/1 [==============================] - 0s 61ms/step\n",
      "1/1 [==============================] - 0s 51ms/step\n",
      "1/1 [==============================] - 0s 90ms/step\n",
      "1/1 [==============================] - 0s 24ms/step\n",
      "1/1 [==============================] - 0s 27ms/step\n",
      "1/1 [==============================] - 0s 25ms/step\n",
      "1/1 [==============================] - 0s 24ms/step\n",
      "1/1 [==============================] - 0s 42ms/step\n",
      "1/1 [==============================] - 0s 45ms/step\n",
      "1/1 [==============================] - 0s 37ms/step\n",
      "1/1 [==============================] - 0s 50ms/step\n",
      "1/1 [==============================] - 0s 87ms/step\n",
      "1/1 [==============================] - 0s 28ms/step\n",
      "1/1 [==============================] - 0s 80ms/step\n",
      "1/1 [==============================] - 0s 26ms/step\n",
      "1/1 [==============================] - 0s 70ms/step\n",
      "1/1 [==============================] - 0s 56ms/step\n",
      "1/1 [==============================] - 0s 82ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 54ms/step\n",
      "1/1 [==============================] - 0s 63ms/step\n",
      "1/1 [==============================] - 0s 77ms/step\n",
      "1/1 [==============================] - 0s 27ms/step\n",
      "1/1 [==============================] - 0s 25ms/step\n",
      "1/1 [==============================] - 0s 53ms/step\n",
      "1/1 [==============================] - 0s 78ms/step\n",
      "1/1 [==============================] - 0s 62ms/step\n",
      "1/1 [==============================] - 0s 25ms/step\n",
      "1/1 [==============================] - 0s 58ms/step\n",
      "1/1 [==============================] - 0s 24ms/step\n",
      "1/1 [==============================] - 0s 56ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 54ms/step\n",
      "1/1 [==============================] - 0s 56ms/step\n",
      "1/1 [==============================] - 0s 72ms/step\n",
      "1/1 [==============================] - 0s 56ms/step\n",
      "Timestep 20 has an rmse of 0.017536682244544254.\n",
      "Timestep 30 has an rmse of 0.018036320018463226.\n",
      "Timestep 40 has an rmse of 0.019453517364745744.\n",
      "Timestep 50 has an rmse of 0.027204263953219645.\n",
      "Timestep 60 has an rmse of 113371312727710.56.\n"
     ]
    },
    {
     "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": [
    "timesteps = [20,30,40,50,60]\n",
    "rmse = []\n",
    "\n",
    "for step in timesteps:\n",
    "    train_generator = TimeseriesGenerator(scaled_train,scaled_train, length = step, batch_size=1)\n",
    "    model2 = Sequential()\n",
    "    model2.add(LSTM(150, activation='relu', input_shape=(step, 1)))\n",
    "    model2.add(Dense(1)) #output prediction\n",
    "    model2.compile(optimizer = 'adam', loss = 'mse')\n",
    "    model2.fit(train_generator,epochs = 25)\n",
    "    model2.history.history.keys()\n",
    "    plt.plot(range(len(model2.history.history['loss'])),model2.history.history['loss'], label = f\"Loss plot for Timestep:{step}\")\n",
    "    \n",
    "    first_eval_batch2 = scaled_train[-step:]\n",
    "    current_batch2 = first_eval_batch2.reshape((1,step,1))\n",
    "    \n",
    "    test_predictions2 = []\n",
    "    \n",
    "    for i in range(len(test_set['Close'])):\n",
    "    \n",
    "        current_pred2 = model2.predict(current_batch2)[0]\n",
    "    \n",
    "        #store the prediction\n",
    "        test_predictions2.append(current_pred2)\n",
    "\n",
    "        current_batch2 = np.append(current_batch2[:,1:,:],[[current_pred2]], axis = 1)\n",
    "    \n",
    "    predictions_true = scaler.inverse_transform(test_predictions2)\n",
    "    \n",
    "    original_test = test_set[['Close']]\n",
    "    \n",
    "    original_test['predictions'] = predictions_true\n",
    "    \n",
    "    \n",
    "    rms = mean_squared_error(original_test['Close'],original_test['predictions'],squared = False)\n",
    "    rmse.append(rms)\n",
    "    \n",
    "for i in range (len(rmse)):\n",
    "    print(f'Timestep {timesteps[i]} has an rmse of {rmse[i]}.')"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "ad9cd594",
   "metadata": {},
   "source": [
    "From the results above, we see that time steps of 20 gave the best results and the rmse is lower than that of the previous models. But even with the rmse, the model is still not able to predict the trends in the prices. Even if it cant get the exact price, its tsill not able to correctly predict whether there is an increase or decrease"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "a3604376",
   "metadata": {},
   "source": [
    "3. I'll add another LSTM layer\n",
    "Lets add an LSTM layer to see how ir performs(you can also add dropout if you fear there is overfitting)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 769,
   "id": "941a70b7",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/30\n",
      "1104/1104 [==============================] - 16s 12ms/step - loss: 0.0059\n",
      "Epoch 2/30\n",
      "1104/1104 [==============================] - 13s 12ms/step - loss: 0.0023\n",
      "Epoch 3/30\n",
      "1104/1104 [==============================] - 13s 12ms/step - loss: 0.0017\n",
      "Epoch 4/30\n",
      "1104/1104 [==============================] - 13s 12ms/step - loss: 0.0014\n",
      "Epoch 5/30\n",
      "1104/1104 [==============================] - 13s 12ms/step - loss: 0.0012\n",
      "Epoch 6/30\n",
      "1104/1104 [==============================] - 14s 12ms/step - loss: 0.0011\n",
      "Epoch 7/30\n",
      "1104/1104 [==============================] - 13s 12ms/step - loss: 9.8755e-04\n",
      "Epoch 8/30\n",
      "1104/1104 [==============================] - 13s 12ms/step - loss: 9.7189e-04\n",
      "Epoch 9/30\n",
      "1104/1104 [==============================] - 13s 12ms/step - loss: 9.9939e-04\n",
      "Epoch 10/30\n",
      "1104/1104 [==============================] - 14s 12ms/step - loss: 9.6275e-04\n",
      "Epoch 11/30\n",
      "1104/1104 [==============================] - 13s 12ms/step - loss: 9.2973e-04\n",
      "Epoch 12/30\n",
      "1104/1104 [==============================] - 13s 12ms/step - loss: 9.2520e-04\n",
      "Epoch 13/30\n",
      "1104/1104 [==============================] - 13s 12ms/step - loss: 8.2667e-04\n",
      "Epoch 14/30\n",
      "1104/1104 [==============================] - 13s 12ms/step - loss: 7.9371e-04\n",
      "Epoch 15/30\n",
      "1104/1104 [==============================] - 13s 12ms/step - loss: 8.0511e-04\n",
      "Epoch 16/30\n",
      "1104/1104 [==============================] - 13s 12ms/step - loss: 8.4500e-04\n",
      "Epoch 17/30\n",
      "1104/1104 [==============================] - 13s 11ms/step - loss: 7.9054e-04\n",
      "Epoch 18/30\n",
      "1104/1104 [==============================] - 13s 12ms/step - loss: 8.4894e-04\n",
      "Epoch 19/30\n",
      "1104/1104 [==============================] - 13s 12ms/step - loss: 7.8206e-04\n",
      "Epoch 20/30\n",
      "1104/1104 [==============================] - 13s 12ms/step - loss: 7.8127e-04\n",
      "Epoch 21/30\n",
      "1104/1104 [==============================] - 13s 12ms/step - loss: 8.0698e-04\n",
      "Epoch 22/30\n",
      "1104/1104 [==============================] - 13s 12ms/step - loss: 7.9750e-04\n",
      "Epoch 23/30\n",
      "1104/1104 [==============================] - 13s 12ms/step - loss: 7.3020e-04\n",
      "Epoch 24/30\n",
      "1104/1104 [==============================] - 13s 12ms/step - loss: 7.6968e-04\n",
      "Epoch 25/30\n",
      "1104/1104 [==============================] - 13s 12ms/step - loss: 8.0628e-04\n",
      "Epoch 26/30\n",
      "1104/1104 [==============================] - 13s 12ms/step - loss: 7.5521e-04\n",
      "Epoch 27/30\n",
      "1104/1104 [==============================] - 13s 12ms/step - loss: 7.8172e-04\n",
      "Epoch 28/30\n",
      "1104/1104 [==============================] - 13s 12ms/step - loss: 7.4064e-04\n",
      "Epoch 29/30\n",
      "1104/1104 [==============================] - 13s 12ms/step - loss: 7.5290e-04\n",
      "Epoch 30/30\n",
      "1104/1104 [==============================] - 13s 12ms/step - loss: 7.2182e-04\n",
      "1/1 [==============================] - 0s 22ms/step\n",
      "1/1 [==============================] - 0s 20ms/step\n",
      "1/1 [==============================] - 0s 21ms/step\n",
      "1/1 [==============================] - 0s 22ms/step\n",
      "1/1 [==============================] - 0s 21ms/step\n",
      "1/1 [==============================] - 0s 24ms/step\n",
      "1/1 [==============================] - 0s 22ms/step\n",
      "1/1 [==============================] - 0s 22ms/step\n",
      "1/1 [==============================] - 0s 24ms/step\n",
      "1/1 [==============================] - 0s 22ms/step\n",
      "1/1 [==============================] - 0s 25ms/step\n",
      "1/1 [==============================] - 0s 20ms/step\n",
      "1/1 [==============================] - 0s 22ms/step\n",
      "1/1 [==============================] - 0s 20ms/step\n",
      "1/1 [==============================] - 0s 22ms/step\n",
      "1/1 [==============================] - 0s 23ms/step\n",
      "1/1 [==============================] - 0s 21ms/step\n",
      "1/1 [==============================] - 0s 21ms/step\n",
      "1/1 [==============================] - 0s 21ms/step\n",
      "1/1 [==============================] - 0s 21ms/step\n",
      "1/1 [==============================] - 0s 21ms/step\n",
      "1/1 [==============================] - 0s 21ms/step\n",
      "1/1 [==============================] - 0s 22ms/step\n",
      "1/1 [==============================] - 0s 22ms/step\n",
      "1/1 [==============================] - 0s 22ms/step\n",
      "1/1 [==============================] - 0s 20ms/step\n",
      "1/1 [==============================] - 0s 20ms/step\n",
      "1/1 [==============================] - 0s 28ms/step\n",
      "1/1 [==============================] - 0s 20ms/step\n",
      "1/1 [==============================] - 0s 22ms/step\n",
      "1/1 [==============================] - 0s 21ms/step\n",
      "1/1 [==============================] - 0s 21ms/step\n",
      "1/1 [==============================] - 0s 21ms/step\n",
      "1/1 [==============================] - 0s 23ms/step\n",
      "1/1 [==============================] - 0s 22ms/step\n",
      "1/1 [==============================] - 0s 21ms/step\n",
      "1/1 [==============================] - 0s 23ms/step\n",
      "1/1 [==============================] - 0s 21ms/step\n",
      "1/1 [==============================] - 0s 20ms/step\n",
      "1/1 [==============================] - 0s 20ms/step\n",
      "1/1 [==============================] - 0s 21ms/step\n",
      "1/1 [==============================] - 0s 23ms/step\n",
      "1/1 [==============================] - 0s 21ms/step\n",
      "1/1 [==============================] - 0s 22ms/step\n",
      "1/1 [==============================] - 0s 21ms/step\n",
      "1/1 [==============================] - 0s 21ms/step\n",
      "1/1 [==============================] - 0s 21ms/step\n",
      "1/1 [==============================] - 0s 21ms/step\n",
      "1/1 [==============================] - 0s 21ms/step\n",
      "1/1 [==============================] - 0s 24ms/step\n",
      "1/1 [==============================] - 0s 21ms/step\n",
      "1/1 [==============================] - 0s 22ms/step\n",
      "1/1 [==============================] - 0s 22ms/step\n",
      "1/1 [==============================] - 0s 22ms/step\n",
      "1/1 [==============================] - 0s 22ms/step\n",
      "1/1 [==============================] - 0s 20ms/step\n",
      "1/1 [==============================] - 0s 20ms/step\n",
      "1/1 [==============================] - 0s 21ms/step\n",
      "1/1 [==============================] - 0s 22ms/step\n",
      "1/1 [==============================] - 0s 20ms/step\n",
      "1/1 [==============================] - 0s 21ms/step\n",
      "1/1 [==============================] - 0s 22ms/step\n",
      "1/1 [==============================] - 0s 20ms/step\n",
      "1/1 [==============================] - 0s 20ms/step\n",
      "1/1 [==============================] - 0s 22ms/step\n",
      "1/1 [==============================] - 0s 21ms/step\n",
      "1/1 [==============================] - 0s 23ms/step\n",
      "1/1 [==============================] - 0s 20ms/step\n",
      "1/1 [==============================] - 0s 22ms/step\n",
      "1/1 [==============================] - 0s 24ms/step\n",
      "1/1 [==============================] - 0s 22ms/step\n",
      "1/1 [==============================] - 0s 21ms/step\n",
      "1/1 [==============================] - 0s 21ms/step\n",
      "1/1 [==============================] - 0s 24ms/step\n",
      "1/1 [==============================] - 0s 20ms/step\n",
      "1/1 [==============================] - 0s 23ms/step\n",
      "1/1 [==============================] - 0s 21ms/step\n",
      "1/1 [==============================] - 0s 21ms/step\n",
      "1/1 [==============================] - 0s 22ms/step\n",
      "1/1 [==============================] - 0s 21ms/step\n",
      "1/1 [==============================] - 0s 20ms/step\n",
      "1/1 [==============================] - 0s 22ms/step\n",
      "1/1 [==============================] - 0s 22ms/step\n",
      "1/1 [==============================] - 0s 23ms/step\n",
      "1/1 [==============================] - 0s 23ms/step\n",
      "1/1 [==============================] - 0s 22ms/step\n",
      "1/1 [==============================] - 0s 22ms/step\n",
      "1/1 [==============================] - 0s 22ms/step\n",
      "1/1 [==============================] - 0s 21ms/step\n",
      "1/1 [==============================] - 0s 19ms/step\n",
      "1/1 [==============================] - 0s 21ms/step\n",
      "1/1 [==============================] - 0s 20ms/step\n",
      "1/1 [==============================] - 0s 21ms/step\n",
      "1/1 [==============================] - 0s 20ms/step\n",
      "1/1 [==============================] - 0s 22ms/step\n",
      "1/1 [==============================] - 0s 21ms/step\n",
      "1/1 [==============================] - 0s 23ms/step\n",
      "1/1 [==============================] - 0s 25ms/step\n",
      "1/1 [==============================] - 0s 22ms/step\n",
      "1/1 [==============================] - 0s 22ms/step\n",
      "1/1 [==============================] - 0s 22ms/step\n",
      "1/1 [==============================] - 0s 21ms/step\n",
      "1/1 [==============================] - 0s 22ms/step\n",
      "1/1 [==============================] - 0s 21ms/step\n",
      "1/1 [==============================] - 0s 21ms/step\n",
      "1/1 [==============================] - 0s 21ms/step\n",
      "1/1 [==============================] - 0s 22ms/step\n",
      "1/1 [==============================] - 0s 22ms/step\n",
      "1/1 [==============================] - 0s 20ms/step\n",
      "1/1 [==============================] - 0s 22ms/step\n",
      "1/1 [==============================] - 0s 22ms/step\n",
      "1/1 [==============================] - 0s 22ms/step\n",
      "1/1 [==============================] - 0s 21ms/step\n",
      "1/1 [==============================] - 0s 25ms/step\n",
      "1/1 [==============================] - 0s 20ms/step\n",
      "1/1 [==============================] - 0s 21ms/step\n",
      "1/1 [==============================] - 0s 22ms/step\n",
      "1/1 [==============================] - 0s 22ms/step\n",
      "1/1 [==============================] - 0s 21ms/step\n",
      "1/1 [==============================] - 0s 21ms/step\n",
      "1/1 [==============================] - 0s 21ms/step\n",
      "1/1 [==============================] - 0s 20ms/step\n",
      "1/1 [==============================] - 0s 20ms/step\n",
      "1/1 [==============================] - 0s 21ms/step\n",
      "1/1 [==============================] - 0s 22ms/step\n",
      "1/1 [==============================] - 0s 21ms/step\n",
      "1/1 [==============================] - 0s 22ms/step\n",
      "1/1 [==============================] - 0s 20ms/step\n",
      "1/1 [==============================] - 0s 21ms/step\n",
      "1/1 [==============================] - 0s 23ms/step\n",
      "1/1 [==============================] - 0s 20ms/step\n",
      "1/1 [==============================] - 0s 21ms/step\n",
      "1/1 [==============================] - 0s 22ms/step\n",
      "1/1 [==============================] - 0s 22ms/step\n",
      "1/1 [==============================] - 0s 21ms/step\n",
      "1/1 [==============================] - 0s 22ms/step\n",
      "1/1 [==============================] - 0s 22ms/step\n",
      "1/1 [==============================] - 0s 21ms/step\n",
      "1/1 [==============================] - 0s 21ms/step\n",
      "1/1 [==============================] - 0s 22ms/step\n",
      "1/1 [==============================] - 0s 22ms/step\n",
      "1/1 [==============================] - 0s 21ms/step\n",
      "1/1 [==============================] - 0s 21ms/step\n",
      "1/1 [==============================] - 0s 22ms/step\n",
      "1/1 [==============================] - 0s 22ms/step\n",
      "1/1 [==============================] - 0s 24ms/step\n",
      "1/1 [==============================] - 0s 22ms/step\n",
      "1/1 [==============================] - 0s 20ms/step\n",
      "1/1 [==============================] - 0s 22ms/step\n",
      "1/1 [==============================] - 0s 23ms/step\n",
      "1/1 [==============================] - 0s 21ms/step\n",
      "1/1 [==============================] - 0s 21ms/step\n",
      "1/1 [==============================] - 0s 21ms/step\n",
      "1/1 [==============================] - 0s 21ms/step\n",
      "1/1 [==============================] - 0s 21ms/step\n",
      "1/1 [==============================] - 0s 22ms/step\n",
      "1/1 [==============================] - 0s 21ms/step\n",
      "1/1 [==============================] - 0s 24ms/step\n",
      "1/1 [==============================] - 0s 19ms/step\n",
      "1/1 [==============================] - 0s 24ms/step\n",
      "1/1 [==============================] - 0s 21ms/step\n",
      "1/1 [==============================] - 0s 23ms/step\n",
      "1/1 [==============================] - 0s 22ms/step\n",
      "1/1 [==============================] - 0s 21ms/step\n",
      "1/1 [==============================] - 0s 21ms/step\n",
      "1/1 [==============================] - 0s 22ms/step\n",
      "1/1 [==============================] - 0s 36ms/step\n",
      "1/1 [==============================] - 0s 21ms/step\n",
      "1/1 [==============================] - 0s 22ms/step\n",
      "1/1 [==============================] - 0s 21ms/step\n",
      "1/1 [==============================] - 0s 20ms/step\n",
      "1/1 [==============================] - 0s 20ms/step\n",
      "1/1 [==============================] - 0s 21ms/step\n",
      "1/1 [==============================] - 0s 21ms/step\n",
      "1/1 [==============================] - 0s 21ms/step\n",
      "1/1 [==============================] - 0s 20ms/step\n",
      "1/1 [==============================] - 0s 30ms/step\n",
      "1/1 [==============================] - 0s 21ms/step\n",
      "1/1 [==============================] - 0s 22ms/step\n",
      "1/1 [==============================] - 0s 21ms/step\n"
     ]
    },
    {
     "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": [
    "train_gen = TimeseriesGenerator(scaled_train,scaled_train, length = 20, batch_size=1)\n",
    "stacked_model = Sequential()\n",
    "stacked_model.add(LSTM(150, activation='relu', input_shape=(20, 1), return_sequences=True))\n",
    "stacked_model.add(LSTM(150, activation='relu'))\n",
    "stacked_model.add(Dense(1)) #output prediction\n",
    "stacked_model.compile(optimizer = 'adam', loss = 'mse')\n",
    "stacked_model.fit(train_gen,epochs = 30)\n",
    "stacked_model.history.history.keys()\n",
    "plt.plot(range(len(stacked_model.history.history['loss'])),stacked_model.history.history['loss'])\n",
    "\n",
    "# This will hold the predictions as we go along\n",
    "test_predictions = []\n",
    "\n",
    "first_eval_batch = scaled_train[-20:]\n",
    "\n",
    "current_batch = first_eval_batch.reshape((1,20,1))\n",
    "\n",
    "for i in range(len(test_set['Close'])):\n",
    "    \n",
    "    current_pred = model.predict(current_batch)[0]\n",
    "\n",
    "    test_predictions.append(current_pred)\n",
    "\n",
    "    current_batch = np.append(current_batch[:,1:,:],[[current_pred]], axis = 1)\n",
    "    \n",
    "True_predictions = scaler.inverse_transform(test_predictions)\n",
    "    \n",
    "original_test = test_set[['Close']]\n",
    "\n",
    "original_test['predictions'] = True_predictions"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 770,
   "id": "06a3d5fd",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<AxesSubplot:xlabel='Date'>"
      ]
     },
     "execution_count": 770,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "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": [
    "original_test.plot(legend=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 771,
   "id": "9dfd1bd9",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.03507562465172962"
      ]
     },
     "execution_count": 771,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "mean_squared_error(original_test['Close'],original_test['predictions'],squared = False)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "98f8c1f2",
   "metadata": {},
   "source": [
    "**Even if it doesnt get the values, was it able to get the trend?**\n",
    "   - This model has a higher RMSE than all the other models and also doesn't forecast the future trends properly"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "7e186bf0",
   "metadata": {},
   "source": [
    "### Model 5 - Facebook Prophet"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "3f473cef",
   "metadata": {},
   "outputs": [],
   "source": [
    "print('head')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "c2dbce48",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Collecting prophetNote: you may need to restart the kernel to use updated packages.\n",
      "  Downloading prophet-1.1.1-cp39-cp39-win_amd64.whl (12.1 MB)\n",
      "Requirement already satisfied: matplotlib>=2.0.0 in c:\\users\\olalawal\\anaconda3\\lib\\site-packages (from prophet) (3.5.1)\n",
      "Requirement already satisfied: tqdm>=4.36.1 in c:\\users\\olalawal\\anaconda3\\lib\\site-packages (from prophet) (4.64.0)\n",
      "Collecting cmdstanpy>=1.0.4\n",
      "  Downloading cmdstanpy-1.0.8-py3-none-any.whl (81 kB)\n",
      "\n",
      "Requirement already satisfied: python-dateutil>=2.8.0 in c:\\users\\olalawal\\anaconda3\\lib\\site-packages (from prophet) (2.8.2)\n",
      "Requirement already satisfied: pandas>=1.0.4 in c:\\users\\olalawal\\anaconda3\\lib\\site-packages (from prophet) (1.3.5)\n",
      "Requirement already satisfied: wheel>=0.37.0 in c:\\users\\olalawal\\anaconda3\\lib\\site-packages (from prophet) (0.37.1)\n",
      "Requirement already satisfied: numpy>=1.15.4 in c:\\users\\olalawal\\anaconda3\\lib\\site-packages (from prophet) (1.21.5)\n",
      "Collecting setuptools-git>=1.2\n",
      "  Downloading setuptools_git-1.2-py2.py3-none-any.whl (10 kB)\n",
      "Collecting convertdate>=2.1.2\n",
      "  Downloading convertdate-2.4.0-py3-none-any.whl (47 kB)\n",
      "Requirement already satisfied: setuptools>=42 in c:\\users\\olalawal\\anaconda3\\lib\\site-packages (from prophet) (61.2.0)\n",
      "Collecting holidays>=0.14.2\n",
      "  Downloading holidays-0.17.2-py3-none-any.whl (187 kB)\n",
      "Collecting LunarCalendar>=0.0.9\n",
      "  Downloading LunarCalendar-0.0.9-py2.py3-none-any.whl (18 kB)\n",
      "Collecting pymeeus<=1,>=0.3.13\n",
      "  Downloading PyMeeus-0.5.11.tar.gz (5.4 MB)\n",
      "Collecting hijri-converter\n",
      "  Downloading hijri_converter-2.2.4-py3-none-any.whl (14 kB)\n",
      "Collecting korean-lunar-calendar\n",
      "  Downloading korean_lunar_calendar-0.3.1-py3-none-any.whl (9.0 kB)\n",
      "Requirement already satisfied: pytz in c:\\users\\olalawal\\anaconda3\\lib\\site-packages (from LunarCalendar>=0.0.9->prophet) (2021.3)\n",
      "Collecting ephem>=3.7.5.3\n",
      "  Downloading ephem-4.1.3-cp39-cp39-win_amd64.whl (1.4 MB)\n",
      "Requirement already satisfied: fonttools>=4.22.0 in c:\\users\\olalawal\\anaconda3\\lib\\site-packages (from matplotlib>=2.0.0->prophet) (4.25.0)\n",
      "Requirement already satisfied: kiwisolver>=1.0.1 in c:\\users\\olalawal\\anaconda3\\lib\\site-packages (from matplotlib>=2.0.0->prophet) (1.3.2)\n",
      "Requirement already satisfied: cycler>=0.10 in c:\\users\\olalawal\\anaconda3\\lib\\site-packages (from matplotlib>=2.0.0->prophet) (0.11.0)\n",
      "Requirement already satisfied: pillow>=6.2.0 in c:\\users\\olalawal\\anaconda3\\lib\\site-packages (from matplotlib>=2.0.0->prophet) (9.1.0)\n",
      "Requirement already satisfied: pyparsing>=2.2.1 in c:\\users\\olalawal\\anaconda3\\lib\\site-packages (from matplotlib>=2.0.0->prophet) (3.0.4)\n",
      "Requirement already satisfied: packaging>=20.0 in c:\\users\\olalawal\\anaconda3\\lib\\site-packages (from matplotlib>=2.0.0->prophet) (21.3)\n",
      "Requirement already satisfied: six>=1.5 in c:\\users\\olalawal\\anaconda3\\lib\\site-packages (from python-dateutil>=2.8.0->prophet) (1.16.0)\n",
      "Requirement already satisfied: colorama in c:\\users\\olalawal\\anaconda3\\lib\\site-packages (from tqdm>=4.36.1->prophet) (0.4.4)\n",
      "Building wheels for collected packages: pymeeus\n",
      "  Building wheel for pymeeus (setup.py): started\n",
      "  Building wheel for pymeeus (setup.py): finished with status 'done'\n",
      "  Created wheel for pymeeus: filename=PyMeeus-0.5.11-py3-none-any.whl size=730985 sha256=0a65211df33a18f826c9dd05cf9eecc35be0c612c1d2a0440ca8c8d7c80bf5ef\n",
      "  Stored in directory: c:\\users\\olalawal\\appdata\\local\\pip\\cache\\wheels\\33\\3d\\82\\4579e9cca41ff991140b2e050bc6df3a38292f26e4fa06b15d\n",
      "Successfully built pymeeus\n",
      "Installing collected packages: pymeeus, korean-lunar-calendar, hijri-converter, ephem, convertdate, setuptools-git, LunarCalendar, holidays, cmdstanpy, prophet\n",
      "Successfully installed LunarCalendar-0.0.9 cmdstanpy-1.0.8 convertdate-2.4.0 ephem-4.1.3 hijri-converter-2.2.4 holidays-0.17.2 korean-lunar-calendar-0.3.1 prophet-1.1.1 pymeeus-0.5.11 setuptools-git-1.2\n"
     ]
    }
   ],
   "source": [
    "pip install prophet"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "301b7c32",
   "metadata": {},
   "outputs": [],
   "source": [
    "from prophet import Prophet"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "9680fe82",
   "metadata": {},
   "source": [
    "Prophet expects the data to be in a particular format.\n",
    "- Date column needs to be called 'ds'\n",
    "- Date col needs to be a pandas datetime object\n",
    "- The target variable col needs to be called 'y'\n",
    "- prophet expects daily data tho it can accept monthly\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "d9dea746",
   "metadata": {},
   "source": [
    "I'll clean the data to fit the format expected by Prophet"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "id": "da4f5c95",
   "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>Date</th>\n",
       "      <th>Close</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>2017-11-22</td>\n",
       "      <td>0.782840</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>2017-11-23</td>\n",
       "      <td>0.787092</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>2017-11-24</td>\n",
       "      <td>0.786621</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>2017-11-27</td>\n",
       "      <td>0.786535</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>2017-11-28</td>\n",
       "      <td>0.783797</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1300</th>\n",
       "      <td>2022-11-16</td>\n",
       "      <td>0.753381</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1301</th>\n",
       "      <td>2022-11-17</td>\n",
       "      <td>0.749878</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1302</th>\n",
       "      <td>2022-11-18</td>\n",
       "      <td>0.750672</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1303</th>\n",
       "      <td>2022-11-21</td>\n",
       "      <td>0.747557</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1304</th>\n",
       "      <td>2022-11-22</td>\n",
       "      <td>0.747641</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>1305 rows × 2 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "            Date     Close\n",
       "0     2017-11-22  0.782840\n",
       "1     2017-11-23  0.787092\n",
       "2     2017-11-24  0.786621\n",
       "3     2017-11-27  0.786535\n",
       "4     2017-11-28  0.783797\n",
       "...          ...       ...\n",
       "1300  2022-11-16  0.753381\n",
       "1301  2022-11-17  0.749878\n",
       "1302  2022-11-18  0.750672\n",
       "1303  2022-11-21  0.747557\n",
       "1304  2022-11-22  0.747641\n",
       "\n",
       "[1305 rows x 2 columns]"
      ]
     },
     "execution_count": 18,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "prices = pd.read_csv(r'C:\\Users\\olalawal\\Downloads\\CADUSD.csv', usecols = ['Date', 'Close'] )\n",
    "prices"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "dc50ad91",
   "metadata": {},
   "source": [
    "Change the column names"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "id": "c28fd224",
   "metadata": {},
   "outputs": [],
   "source": [
    "prices.columns = ['ds','y']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "id": "20c9fcdd",
   "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>ds</th>\n",
       "      <th>y</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>2017-11-22</td>\n",
       "      <td>0.782840</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>2017-11-23</td>\n",
       "      <td>0.787092</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>2017-11-24</td>\n",
       "      <td>0.786621</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>2017-11-27</td>\n",
       "      <td>0.786535</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>2017-11-28</td>\n",
       "      <td>0.783797</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1300</th>\n",
       "      <td>2022-11-16</td>\n",
       "      <td>0.753381</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1301</th>\n",
       "      <td>2022-11-17</td>\n",
       "      <td>0.749878</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1302</th>\n",
       "      <td>2022-11-18</td>\n",
       "      <td>0.750672</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1303</th>\n",
       "      <td>2022-11-21</td>\n",
       "      <td>0.747557</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1304</th>\n",
       "      <td>2022-11-22</td>\n",
       "      <td>0.747641</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>1305 rows × 2 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "              ds         y\n",
       "0     2017-11-22  0.782840\n",
       "1     2017-11-23  0.787092\n",
       "2     2017-11-24  0.786621\n",
       "3     2017-11-27  0.786535\n",
       "4     2017-11-28  0.783797\n",
       "...          ...       ...\n",
       "1300  2022-11-16  0.753381\n",
       "1301  2022-11-17  0.749878\n",
       "1302  2022-11-18  0.750672\n",
       "1303  2022-11-21  0.747557\n",
       "1304  2022-11-22  0.747641\n",
       "\n",
       "[1305 rows x 2 columns]"
      ]
     },
     "execution_count": 20,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "prices"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "id": "de4155a5",
   "metadata": {},
   "outputs": [],
   "source": [
    "prices['ds'] = pd.to_datetime(prices['ds'])"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "8bca8174",
   "metadata": {},
   "source": [
    "### Train/Test Split"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "dd91856f",
   "metadata": {},
   "source": [
    "I'll split the data same way I split for the other the models.\n",
    "Test set will have the last 9 months of data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 64,
   "id": "bf9d148c",
   "metadata": {},
   "outputs": [],
   "source": [
    "train = prices[:1125]\n",
    "test = prices[1125:]"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "dc26bdcc",
   "metadata": {},
   "source": [
    "Create the model\n",
    "\n",
    "- Prophet automatically checks if there is any seasonality & uses what it finds but you can override it"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 68,
   "id": "e0c3d5fa",
   "metadata": {},
   "outputs": [],
   "source": [
    "model = Prophet()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "0205332f",
   "metadata": {},
   "source": [
    "Fit to the training data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 69,
   "id": "1fea61bb",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "23:10:08 - cmdstanpy - INFO - Chain [1] start processing\n",
      "23:10:09 - cmdstanpy - INFO - Chain [1] done processing\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "<prophet.forecaster.Prophet at 0x16fe1775a90>"
      ]
     },
     "execution_count": 69,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "model.fit(train)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "12c09e85",
   "metadata": {},
   "source": [
    "Make the future dataframe \n",
    "- This will hold both the current dates (in the train set) + the dates for the next periods that you specify\n",
    "- Since I want to predict 9 months (180 business days) into the future, I will use a period of 180"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 70,
   "id": "a8508251",
   "metadata": {},
   "outputs": [],
   "source": [
    "future = model.make_future_dataframe(periods = 180, freq='B')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 71,
   "id": "102eb664",
   "metadata": {},
   "outputs": [],
   "source": [
    "forecast = model.predict(future)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 72,
   "id": "baf70965",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<AxesSubplot:xlabel='ds'>"
      ]
     },
     "execution_count": 72,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 864x576 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "ax = forecast.plot(x='ds', y = 'yhat', label='Predictions', legend=True, figsize=(12,8))\n",
    "test.plot(x='ds', y='y', label ='True Test data', legend=True, ax=ax, xlim = ('2022-03-15','2022-11-22'))"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "059f3b21",
   "metadata": {},
   "source": [
    "### Calculate RMSE "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 73,
   "id": "cca14724",
   "metadata": {},
   "outputs": [],
   "source": [
    "from statsmodels.tools.eval_measures import rmse"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 74,
   "id": "3ac87a56",
   "metadata": {},
   "outputs": [],
   "source": [
    "predictions = forecast.iloc[-180:]['yhat']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 75,
   "id": "fed37d1f",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "1125    0.770164\n",
       "1126    0.769607\n",
       "1127    0.768783\n",
       "1128    0.767233\n",
       "1129    0.767053\n",
       "          ...   \n",
       "1300    0.744590\n",
       "1301    0.744521\n",
       "1302    0.744139\n",
       "1303    0.743583\n",
       "1304    0.743607\n",
       "Name: yhat, Length: 180, dtype: float64"
      ]
     },
     "execution_count": 75,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "predictions"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 76,
   "id": "030a80d9",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.01863211618564846"
      ]
     },
     "execution_count": 76,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "rmse(test['y'], predictions)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "cbff3f94",
   "metadata": {},
   "source": [
    "**How did the Prophet Model Perform?**"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "84883c1e",
   "metadata": {},
   "source": [
    "- The prophet model gave the lowest RMSE I've seen so far.\n",
    "- The plot also shows that the model is able to do a fair job of forecasting the future trend.\n",
    "- It was able to forecast the downward trend around May, upward trend around june and even the drop seen around the end of november"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "3bfeeaf5",
   "metadata": {},
   "source": [
    "#### Cross Validation"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "2fc58712",
   "metadata": {},
   "source": [
    "To confirm the performance, I'll do a bit of cross validation"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 77,
   "id": "d4794b67",
   "metadata": {},
   "outputs": [],
   "source": [
    "#Perform cross validation on several sections"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 78,
   "id": "37c85d2b",
   "metadata": {},
   "outputs": [],
   "source": [
    "from prophet.diagnostics import cross_validation, performance_metrics\n",
    "from prophet.plot import plot_cross_validation_metric"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 90,
   "id": "7d3d6ffd",
   "metadata": {},
   "outputs": [],
   "source": [
    "#initial training period - 2 years \n",
    "\n",
    "initial = 260 * 3\n",
    "initial = str(initial) + 'days'\n",
    "\n",
    "#Period length for cross val\n",
    "period = 60\n",
    "period = str(period) + 'days'\n",
    "\n",
    "#horizon of prediction for each fold \n",
    "horizon = 180 #6 months\n",
    "horizon = str(horizon) + 'days'"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 91,
   "id": "fb819811",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "0814503465144541a2822a7a65daed82",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/11 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "23:27:50 - cmdstanpy - INFO - Chain [1] start processing\n",
      "23:27:51 - cmdstanpy - INFO - Chain [1] done processing\n",
      "23:27:51 - cmdstanpy - INFO - Chain [1] start processing\n",
      "23:27:51 - cmdstanpy - INFO - Chain [1] done processing\n",
      "23:27:52 - cmdstanpy - INFO - Chain [1] start processing\n",
      "23:27:52 - cmdstanpy - INFO - Chain [1] done processing\n",
      "23:27:53 - cmdstanpy - INFO - Chain [1] start processing\n",
      "23:27:53 - cmdstanpy - INFO - Chain [1] done processing\n",
      "23:27:54 - cmdstanpy - INFO - Chain [1] start processing\n",
      "23:27:54 - cmdstanpy - INFO - Chain [1] done processing\n",
      "23:27:55 - cmdstanpy - INFO - Chain [1] start processing\n",
      "23:27:55 - cmdstanpy - INFO - Chain [1] done processing\n",
      "23:27:56 - cmdstanpy - INFO - Chain [1] start processing\n",
      "23:27:57 - cmdstanpy - INFO - Chain [1] done processing\n",
      "23:27:57 - cmdstanpy - INFO - Chain [1] start processing\n",
      "23:27:57 - cmdstanpy - INFO - Chain [1] done processing\n",
      "23:27:58 - cmdstanpy - INFO - Chain [1] start processing\n",
      "23:27:59 - cmdstanpy - INFO - Chain [1] done processing\n",
      "23:27:59 - cmdstanpy - INFO - Chain [1] start processing\n",
      "23:28:00 - cmdstanpy - INFO - Chain [1] done processing\n",
      "23:28:00 - cmdstanpy - INFO - Chain [1] start processing\n",
      "23:28:01 - cmdstanpy - INFO - Chain [1] done processing\n"
     ]
    }
   ],
   "source": [
    "prices_cv = cross_validation(model, initial = initial , period=period, horizon = horizon)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 92,
   "id": "9e740bf6",
   "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>horizon</th>\n",
       "      <th>mse</th>\n",
       "      <th>rmse</th>\n",
       "      <th>mae</th>\n",
       "      <th>mape</th>\n",
       "      <th>mdape</th>\n",
       "      <th>smape</th>\n",
       "      <th>coverage</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>18 days</td>\n",
       "      <td>0.000137</td>\n",
       "      <td>0.011698</td>\n",
       "      <td>0.009280</td>\n",
       "      <td>0.012109</td>\n",
       "      <td>0.009632</td>\n",
       "      <td>0.012120</td>\n",
       "      <td>0.460993</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>19 days</td>\n",
       "      <td>0.000140</td>\n",
       "      <td>0.011847</td>\n",
       "      <td>0.009472</td>\n",
       "      <td>0.012365</td>\n",
       "      <td>0.010017</td>\n",
       "      <td>0.012386</td>\n",
       "      <td>0.443262</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>20 days</td>\n",
       "      <td>0.000139</td>\n",
       "      <td>0.011771</td>\n",
       "      <td>0.009430</td>\n",
       "      <td>0.012318</td>\n",
       "      <td>0.010085</td>\n",
       "      <td>0.012349</td>\n",
       "      <td>0.444782</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>21 days</td>\n",
       "      <td>0.000141</td>\n",
       "      <td>0.011865</td>\n",
       "      <td>0.009533</td>\n",
       "      <td>0.012453</td>\n",
       "      <td>0.010268</td>\n",
       "      <td>0.012495</td>\n",
       "      <td>0.443262</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>22 days</td>\n",
       "      <td>0.000144</td>\n",
       "      <td>0.012018</td>\n",
       "      <td>0.009711</td>\n",
       "      <td>0.012683</td>\n",
       "      <td>0.010465</td>\n",
       "      <td>0.012730</td>\n",
       "      <td>0.429472</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>158</th>\n",
       "      <td>176 days</td>\n",
       "      <td>0.003409</td>\n",
       "      <td>0.058391</td>\n",
       "      <td>0.046928</td>\n",
       "      <td>0.060060</td>\n",
       "      <td>0.063896</td>\n",
       "      <td>0.061749</td>\n",
       "      <td>0.725768</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>159</th>\n",
       "      <td>177 days</td>\n",
       "      <td>0.003486</td>\n",
       "      <td>0.059045</td>\n",
       "      <td>0.047575</td>\n",
       "      <td>0.060882</td>\n",
       "      <td>0.064931</td>\n",
       "      <td>0.062590</td>\n",
       "      <td>0.716312</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>160</th>\n",
       "      <td>178 days</td>\n",
       "      <td>0.003438</td>\n",
       "      <td>0.058631</td>\n",
       "      <td>0.047381</td>\n",
       "      <td>0.060588</td>\n",
       "      <td>0.064931</td>\n",
       "      <td>0.062244</td>\n",
       "      <td>0.720745</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>161</th>\n",
       "      <td>179 days</td>\n",
       "      <td>0.003358</td>\n",
       "      <td>0.057947</td>\n",
       "      <td>0.046997</td>\n",
       "      <td>0.060080</td>\n",
       "      <td>0.065636</td>\n",
       "      <td>0.061653</td>\n",
       "      <td>0.723404</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>162</th>\n",
       "      <td>180 days</td>\n",
       "      <td>0.003343</td>\n",
       "      <td>0.057822</td>\n",
       "      <td>0.047085</td>\n",
       "      <td>0.060212</td>\n",
       "      <td>0.065191</td>\n",
       "      <td>0.061748</td>\n",
       "      <td>0.726064</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>163 rows × 8 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "     horizon       mse      rmse       mae      mape     mdape     smape  \\\n",
       "0    18 days  0.000137  0.011698  0.009280  0.012109  0.009632  0.012120   \n",
       "1    19 days  0.000140  0.011847  0.009472  0.012365  0.010017  0.012386   \n",
       "2    20 days  0.000139  0.011771  0.009430  0.012318  0.010085  0.012349   \n",
       "3    21 days  0.000141  0.011865  0.009533  0.012453  0.010268  0.012495   \n",
       "4    22 days  0.000144  0.012018  0.009711  0.012683  0.010465  0.012730   \n",
       "..       ...       ...       ...       ...       ...       ...       ...   \n",
       "158 176 days  0.003409  0.058391  0.046928  0.060060  0.063896  0.061749   \n",
       "159 177 days  0.003486  0.059045  0.047575  0.060882  0.064931  0.062590   \n",
       "160 178 days  0.003438  0.058631  0.047381  0.060588  0.064931  0.062244   \n",
       "161 179 days  0.003358  0.057947  0.046997  0.060080  0.065636  0.061653   \n",
       "162 180 days  0.003343  0.057822  0.047085  0.060212  0.065191  0.061748   \n",
       "\n",
       "     coverage  \n",
       "0    0.460993  \n",
       "1    0.443262  \n",
       "2    0.444782  \n",
       "3    0.443262  \n",
       "4    0.429472  \n",
       "..        ...  \n",
       "158  0.725768  \n",
       "159  0.716312  \n",
       "160  0.720745  \n",
       "161  0.723404  \n",
       "162  0.726064  \n",
       "\n",
       "[163 rows x 8 columns]"
      ]
     },
     "execution_count": 92,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "performance_metrics(prices_cv)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 93,
   "id": "fe43868b",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Users\\olalawal\\Anaconda3\\lib\\site-packages\\prophet\\plot.py:544: FutureWarning: casting timedelta64[ns] values to int64 with .astype(...) is deprecated and will raise in a future version. Use .view(...) instead.\n",
      "  x_plt = df_none['horizon'].astype('timedelta64[ns]').astype(np.int64) / float(dt_conversions[i])\n",
      "C:\\Users\\olalawal\\Anaconda3\\lib\\site-packages\\prophet\\plot.py:545: FutureWarning: casting timedelta64[ns] values to int64 with .astype(...) is deprecated and will raise in a future version. Use .view(...) instead.\n",
      "  x_plt_h = df_h['horizon'].astype('timedelta64[ns]').astype(np.int64) / float(dt_conversions[i])\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x432 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plot_cross_validation_metric(prices_cv, metric='rmse');"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "474c1c27",
   "metadata": {},
   "source": [
    "Even though, the RMSE from the cross validation is higher. I believe this is because less data is been used to train the model"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "1ecc0a3c",
   "metadata": {},
   "source": [
    "## Final Recommendation"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "d68b8593",
   "metadata": {},
   "source": [
    "After implementing different models and comparing their performance on forecasting future trends & prices;\n",
    "- Most of the models didnt perform well and this might be because of the sort of data we are dealing with.\n",
    "- Can we really predict future exchange price based on just past prices. What happends to the external factors that actually affect this price?\n",
    "\n",
    "Therefore, I recommend the following:\n",
    "1. Do more research and obtain data that can be more suitable for this project\n",
    "2. If the business can't obtain more data, The Prophet model can be deployed but bearing in mind its limitations "
   ]
  },
  {
   "cell_type": "markdown",
   "id": "2ccd5a99",
   "metadata": {},
   "source": [
    "## Next Steps"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "4706764a",
   "metadata": {},
   "source": [
    "1. Deploy the Prophet Model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "444cb7a2",
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3 (ipykernel)",
   "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": "3.9.13"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}