WTI Crude Oil Price Prediction.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# WTI Crude Oil Price Prediction - ConocoPhillips Energy Analytics\n",
    "\n",
    "### In this notebook, we will be cleaning data obtained both from ConocoPhillips and the EIA in the hopes of putting it through a model that will predict the price of WTI crude oil. The notebook is split into multiple parts: data cleaning, models, and analysis. \n",
    "\n",
    "## Part I - Data Cleaning"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "For the data, we will use prices since 1986, as well as other metrics that would logically influence crude oil price."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 810,
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "import matplotlib.pyplot as plt\n",
    "import numpy as np"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Import the data and turn missing values into values that Python can interpret."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 677,
   "metadata": {},
   "outputs": [],
   "source": [
    "prices = pd.read_csv(\"WTI_price.csv\", na_values = (\"NaN\"))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Here we're converting the date into an index by which we can sort/arrange the data."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 678,
   "metadata": {},
   "outputs": [],
   "source": [
    "prices.index = pd.to_datetime(prices['Date'])\n",
    "prices.drop('Date', axis = 1, inplace = True)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Let's plot the data and see what the prices look like."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "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": [
    "#Included as a suggestion from Jupyter.\n",
    "from pandas.plotting import register_matplotlib_converters\n",
    "register_matplotlib_converters()\n",
    "\n",
    "plt.figure(figsize = (12,8))\n",
    "plt.plot(prices)\n",
    "plt.title(\"WTI Crude Oil Prices\")\n",
    "plt.xlabel(\"Year\")\n",
    "plt.ylabel(\"Price in USD\")\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Part 2 - Using an LSTM Network"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The first thing we have to do is establish a baseline. To do this, we will run a Long Short-Term Memory neural network on the prices data because LSTM networks have historically worked well in time-series forecasting. We will gather a measure of error from this, and if our model that uses more predictors can beat the error, then we will use the latter model.\n",
    "\n",
    "What about the fact that the LSTM is training on far more data than the multivariate model? To answer this, we will run our multivariate model sans the consumption data in order to have two models that have been run on the same timeframe. If this model has an error that is close to that of the LSTM, we will likely go with the multivariate model, since it's far easier to interpret, and we can add consumption data later on to hopefully reduce the error further."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Let's start working on the LSTM. We have to import a function that will turn our prices into values that fall within a range of (0,1) since that's what the LSTM network interprets."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 94,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Using TensorFlow backend.\n"
     ]
    }
   ],
   "source": [
    "from sklearn.preprocessing import MinMaxScaler\n",
    "\n",
    "import keras\n",
    "from keras.models import Sequential\n",
    "from keras.layers import Dense\n",
    "from keras.layers import LSTM\n",
    "from keras.layers import Dropout"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Now, we have to split the data into a training set and test set. For the Energy Analytics course, we have been tasked with training up until the end of December 2018 and testing on 2019 data. To do that, let's find which index corresponds with the last trading day in December."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 287,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(8319, 1)"
      ]
     },
     "execution_count": 287,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "prices['1986-01-01':'2019-01-01'].shape"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Looks like we'll be training the neural network on 8319 data points. Now, we have to split the data into lists of prices that include the previous 60 trading days. This is because the model will use the 60 previous days to get trained, then use the 61st trading day as a response. Let's scale the prices then split the data."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 288,
   "metadata": {},
   "outputs": [],
   "source": [
    "scaler = MinMaxScaler(feature_range = (0, 1))\n",
    "prices_scaled = scaler.fit_transform(prices)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 680,
   "metadata": {},
   "outputs": [],
   "source": [
    "prices = np.array(prices)\n",
    "prices_training = []\n",
    "prices_response = []\n",
    "for i in range(60, 8319):\n",
    "    prices_training.append(prices_scaled[i-60:i, 0])\n",
    "    prices_response.append(prices_scaled[i, 0])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We have to convert our new lists into numpy arrays, then turn the training set into a 3-dimensional array so that the LSTM network can interpret it (in this case, the third dimension will be 1 because there is only one feature being used for training)."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 346,
   "metadata": {},
   "outputs": [],
   "source": [
    "prices_training, prices_response = np.array(prices_training), np.array(prices_response)\n",
    "prices_training = np.reshape(prices_training, (prices_training.shape[0], prices_training.shape[1], 1))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "For the LSTM network, we will add 4 layers with 50 nodes each, and add dropout regularization to each to prevent overfitting. Then, we'll finish up with a dense layer that converts the values of the final 50 nodes into a single value that can be used for price prediction."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 138,
   "metadata": {},
   "outputs": [],
   "source": [
    "model = Sequential()\n",
    "model.add(LSTM(50, return_sequences = True, input_shape = (prices_training.shape[1], 1)))\n",
    "model.add(Dropout(.2))\n",
    "model.add(LSTM(units=50, return_sequences=True))\n",
    "model.add(Dropout(0.2))\n",
    "model.add(LSTM(units=50, return_sequences=True))\n",
    "model.add(Dropout(0.2))\n",
    "model.add(LSTM(units=50))\n",
    "model.add(Dropout(0.2))\n",
    "model.add(Dense(units = 1))\n",
    "model.compile(optimizer = 'adam', loss = 'mean_squared_error')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 139,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "WARNING:tensorflow:From /Users/andrew/anaconda3/lib/python3.7/site-packages/tensorflow/python/ops/math_grad.py:1250: add_dispatch_support.<locals>.wrapper (from tensorflow.python.ops.array_ops) is deprecated and will be removed in a future version.\n",
      "Instructions for updating:\n",
      "Use tf.where in 2.0, which has the same broadcast rule as np.where\n",
      "WARNING:tensorflow:From /Users/andrew/anaconda3/lib/python3.7/site-packages/keras/backend/tensorflow_backend.py:1033: The name tf.assign_add is deprecated. Please use tf.compat.v1.assign_add instead.\n",
      "\n",
      "Epoch 1/10\n",
      "8259/8259 [==============================] - 39s 5ms/step - loss: 0.0041\n",
      "Epoch 2/10\n",
      "8259/8259 [==============================] - 37s 4ms/step - loss: 0.0016\n",
      "Epoch 3/10\n",
      "8259/8259 [==============================] - 36s 4ms/step - loss: 0.0013\n",
      "Epoch 4/10\n",
      "8259/8259 [==============================] - 37s 4ms/step - loss: 0.0012\n",
      "Epoch 5/10\n",
      "8259/8259 [==============================] - 36s 4ms/step - loss: 0.0010\n",
      "Epoch 6/10\n",
      "8259/8259 [==============================] - 36s 4ms/step - loss: 0.0010\n",
      "Epoch 7/10\n",
      "8259/8259 [==============================] - 38s 5ms/step - loss: 9.6356e-04\n",
      "Epoch 8/10\n",
      "8259/8259 [==============================] - 37s 4ms/step - loss: 8.4464e-04\n",
      "Epoch 9/10\n",
      "8259/8259 [==============================] - 38s 5ms/step - loss: 8.3654e-04\n",
      "Epoch 10/10\n",
      "8259/8259 [==============================] - 37s 4ms/step - loss: 7.8421e-04\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "<keras.callbacks.History at 0x1a389c8a20>"
      ]
     },
     "execution_count": 139,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "model.fit(prices_training, prices_response, epochs = 10, batch_size = 32)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "After running the model through 10 epochs, we can predict the held out data (namely indices 8319 and beyond from the original prices dataset), and then compare them with the actual data to calculate an error. To do this, we have to execute the same idea on the test set that we did on the training set by splitting up the data into the 60 previous price points and labelling them with the 61st price point, which will be predicted by the model."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 408,
   "metadata": {},
   "outputs": [],
   "source": [
    "prices_test = []\n",
    "for i in range(8319, 8476):\n",
    "    prices_test.append(prices_scaled[i-60:i, 0])\n",
    "\n",
    "prices_test = np.array(prices_test)\n",
    "prices_test = np.reshape(prices_test, (prices_test.shape[0], prices_test.shape[1], 1))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 411,
   "metadata": {},
   "outputs": [],
   "source": [
    "predictions = model.predict(prices_test)\n",
    "predictions = scaler.inverse_transform(predictions)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Now that we have our predictions, let's plot them on top of the real data to get a visual understanding of how well the neural network performed."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 412,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 936x648 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.figure(figsize = (13,9))\n",
    "plt.plot(prices[8319:8476], color = 'blue', label = 'Actual WTI Prices')\n",
    "plt.plot(predictions, color = 'red', label = 'Predicted WTI Prices')\n",
    "plt.title('WTI Crude Oil Price Prediction')\n",
    "plt.xlabel('Index')\n",
    "plt.ylabel('Price')\n",
    "plt.legend()\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Looks great! Now that we have a visual representation, we can calculate the mean squared error to get an understanding of how close the the predictions were to reality."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 146,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "38.89"
      ]
     },
     "execution_count": 146,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "residuals = (predictions - prices[8319:8474])\n",
    "mse = round(np.mean(residuals ** 2), 2)\n",
    "mse"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The above model looks good, but there's probably something that's been bothering you if you've paid attention to HOW we've actually predicted the data: this prediction isn't actually useful in the real world because it relies on the previous 60 prices to make predictions, and in reality, we don't have access to these datapoints in long-term forecasting. So, if we're going to use a LSTM network, we have to build one that outputs an n x 1 vector that can predict a series of days worth of data. In this case, we're hoping the model can capture patterns over weeks, months, years, and even decades and extrapolate that to rest of 2019 and all of 2020.\n",
    "\n",
    "After doing some research, there are some methods that can predict multistep time series that involve convolutional neural networks that work in conjunction with the LSTM, but for now, let's stick with using an LSTM network. The idea here is that we will train the model with 152 days of training data for each iteration (rather than 60), and then run a loop of predictions that cuts out the first training point and inserts the most recent prediction. This loop will be run 152 times, and the last prediction will be based completely on prior predicted data rather than data that we supposedly are not supposed to know. This cuts out the need for using test data for predictions, and will show us if the neural network is actually able to predict future data."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 338,
   "metadata": {},
   "outputs": [],
   "source": [
    "#Repeating the same process \n",
    "\n",
    "prices_training2 = []\n",
    "prices_response2 = []\n",
    "for i in range(157, 8319):\n",
    "    prices_training2.append(prices_scaled[i-157:i, 0])\n",
    "    prices_response2.append(prices_scaled[i, 0])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 348,
   "metadata": {},
   "outputs": [],
   "source": [
    "prices_training2, prices_response2 = np.array(prices_training2), np.array(prices_response2)\n",
    "prices_training2 = np.reshape(prices_training, (prices_training.shape[0], 1, prices_training.shape[1]))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Let's create a validation set for the model to work with."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 356,
   "metadata": {},
   "outputs": [],
   "source": [
    "prices_xval = []\n",
    "prices_yval = []\n",
    "for i in range(8317, 8474):\n",
    "    prices_xval.append(prices_scaled[i-60:i, 0])\n",
    "    prices_yval.append(prices_scaled[i, 0])\n",
    "\n",
    "prices_xval, prices_yval = np.array(prices_xval), np.array(prices_yval)\n",
    "prices_xval = np.reshape(prices_xval, (prices_xval.shape[0], 1, prices_xval.shape[1]))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 359,
   "metadata": {},
   "outputs": [],
   "source": [
    "#New model with one hidden layer and 100 nodes. For this model, the features are the 60 previous price points,\n",
    "#rather than having one feature with 60 steps.\n",
    "\n",
    "model2 = Sequential()\n",
    "model2.add(LSTM(100, input_shape = (prices_training2.shape[1], prices_training2.shape[2])))\n",
    "model2.add(Dropout(.2))\n",
    "model2.add(Dense(units = 1))\n",
    "model2.compile(optimizer = 'adam', loss = 'mean_squared_error')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 361,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train on 8259 samples, validate on 157 samples\n",
      "Epoch 1/10\n",
      "8259/8259 [==============================] - 1s 81us/step - loss: 6.6732e-04 - val_loss: 2.5507e-04\n",
      "Epoch 2/10\n",
      "8259/8259 [==============================] - 1s 74us/step - loss: 5.5734e-04 - val_loss: 6.1048e-04\n",
      "Epoch 3/10\n",
      "8259/8259 [==============================] - 1s 73us/step - loss: 4.8663e-04 - val_loss: 3.3615e-04\n",
      "Epoch 4/10\n",
      "8259/8259 [==============================] - 1s 73us/step - loss: 4.8589e-04 - val_loss: 4.4073e-04\n",
      "Epoch 5/10\n",
      "8259/8259 [==============================] - 1s 73us/step - loss: 4.2833e-04 - val_loss: 1.8652e-04\n",
      "Epoch 6/10\n",
      "8259/8259 [==============================] - 1s 73us/step - loss: 4.5636e-04 - val_loss: 1.6479e-04\n",
      "Epoch 7/10\n",
      "8259/8259 [==============================] - 1s 73us/step - loss: 4.2601e-04 - val_loss: 4.2958e-04\n",
      "Epoch 8/10\n",
      "8259/8259 [==============================] - 1s 73us/step - loss: 4.0925e-04 - val_loss: 1.8046e-04\n",
      "Epoch 9/10\n",
      "8259/8259 [==============================] - 1s 72us/step - loss: 4.7451e-04 - val_loss: 3.2671e-04\n",
      "Epoch 10/10\n",
      "8259/8259 [==============================] - 1s 75us/step - loss: 3.9221e-04 - val_loss: 2.6916e-04\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "<keras.callbacks.History at 0x1a480a9ac8>"
      ]
     },
     "execution_count": 361,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "model2.fit(prices_training2, prices_response, epochs = 10, batch_size = 32, validation_data = (prices_xval, prices_yval))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Here, we're predicting the first \"unknown\" data point, appending it to the end of the new set of features, and deleting the oldest until we have a set of data that is filled completely with new, predicted data."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 443,
   "metadata": {},
   "outputs": [],
   "source": [
    "prices_test3 = [prices_scaled[8260:8320]]\n",
    "predicted_prices = []\n",
    "predicted_prices = np.array(predicted_prices)\n",
    "                 \n",
    "for i in range(0, 157):\n",
    "    prices_test3 = np.array(prices_test3)\n",
    "    prices_test3 = np.reshape(prices_test3, (prices_test3.shape[0], 1, prices_test3.shape[1]))\n",
    "    new_prediction = model2.predict(prices_test3)\n",
    "    prices_test3 = np.append(prices_test3, new_prediction)\n",
    "    predicted_prices = np.append(predicted_prices, new_prediction)\n",
    "    prices_test3 = np.delete(prices_test3, 0)\n",
    "    prices_test3 = np.reshape(prices_test3, (1, prices_test3.shape[0]))\n",
    "\n",
    "#Transpose the columns into rows to be plotted, then reverse the scaling.\n",
    "predicted_prices = np.reshape(predicted_prices, (-1,1))\n",
    "predicted_prices = scaler.inverse_transform(predicted_prices)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Here we're just running a similar simulation to the very first one we did, but with the new model."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 403,
   "metadata": {},
   "outputs": [],
   "source": [
    "prices_test_model2 = []\n",
    "for i in range(8319, 8476):\n",
    "    prices_test_model2.append(prices_scaled[i-60:i, 0])\n",
    "\n",
    "prices_test_model2 = np.array(prices_test_model2)\n",
    "prices_test_model2 = np.reshape(prices_test_model2, (prices_test_model2.shape[0], 1, prices_test_model2.shape[1]))\n",
    "\n",
    "predictions2 = model2.predict(prices_test_model2)\n",
    "predictions2 = scaler.inverse_transform(predictions2)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Now that we have all of our predictions, let's plot all of them on one graph and compare them."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 449,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 936x648 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.figure(figsize = (13,9))\n",
    "plt.plot(prices[8319:8476], color = 'blue', label = 'Actual WTI Prices')\n",
    "plt.plot(predicted_prices, color = 'orange', label = 'Predicted WTI Prices for Model 3')\n",
    "plt.plot(predictions2, color = 'red', label = 'Predicted WTI Prices for Model 2')\n",
    "plt.plot(predictions, color = 'green', label = 'Predicted WTI Prices for Model 1')\n",
    "plt.title('WTI Crude Oil Price Prediction')\n",
    "plt.xlabel('Index')\n",
    "plt.ylabel('Price')\n",
    "plt.legend()\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Here we can see that training the model using different numbers of nodes and layers, as well as changing the length of features or how the features are defined in the first place can change our prediction. Clearly, it's difficult to predict a long series of time steps (such as with price prediction), or else everybody would be rich. However, we can clearly that see that there's some shape to the curve, so the model clearly learned something about how the prices act in some way. \n",
    "\n",
    "With future tweaking, it's likely that the prediction could get closer to the truth, but that likely takes months, and even years, to accomplish, and the purpose of this notebook is just to explore methods of prediction."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Part 3 - Using Random Forest and Facebook Prophet"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "To switch gears, let's train a random forest model on the dataset that includes all four previously described predictors. We'll set the random_state to 1 so that we can get the same results during different iterations.\n",
    "\n",
    "Something to mention about this model is that the accuracy can only be computed based on pre-existing data, so for time-series data, it is not a good method for prediction. However, the purpose of this sprint is to get familiar with data analysis and using computational methods on oil data, so using random forest in this case will be okay. \n",
    "\n",
    "What we can do, though, is use another form of time series prediction to gather future data for each predictor using Facebook's Prophet time series modeling tool, and inputting imports, exports, and consumption to get future predictions."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Nice! Our model can predict a price given a vector of four predictors with about 88% accuracy. Now the issue is modeling future metrics (imports, exports, consumptions, and stockpiles) using support vector regression."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 466,
   "metadata": {},
   "outputs": [],
   "source": [
    "from fbprophet import Prophet"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 652,
   "metadata": {},
   "outputs": [],
   "source": [
    "#Importing and cleaning data, then choosing the selected features that are macroeconomically relevant.\n",
    "\n",
    "imports = pd.read_csv(\"Imports.csv\", na_values = (\"NaN\"))\n",
    "exports = pd.read_csv(\"Exports.csv\", na_values = (\"NaN\"))\n",
    "consumption = pd.read_csv(\"Consumption.csv\", na_values = (\"NaN\"))\n",
    "\n",
    "imports = pd.DataFrame((imports['Date'], imports['Weekly U.S. Imports of Crude Oil  (Thousand Barrels per Day)'])).transpose()\n",
    "exports = pd.DataFrame((exports['Date'], exports['Weekly U.S. Exports of Crude Oil  (Thousand Barrels per Day)'])).transpose()\n",
    "consumption = pd.DataFrame((consumption['Date'], consumption['Weekly U.S. Product Supplied of Petroleum Products  (Thousand Barrels per Day)'])).transpose()\n",
    "\n",
    "imports.dropna(inplace = True)\n",
    "exports.dropna(inplace = True)\n",
    "consumption.dropna(inplace = True)\n",
    "\n",
    "#Changing the names of the columns for Prophet to interpret.\n",
    "imports.rename(columns = {'Date':'ds', 'Weekly U.S. Imports of Crude Oil  (Thousand Barrels per Day)':'y'}, inplace = True)\n",
    "exports.rename(columns = {'Date':'ds', 'Weekly U.S. Exports of Crude Oil  (Thousand Barrels per Day)':'y'}, inplace = True)\n",
    "consumption.rename(columns = {'Date':'ds', 'Weekly  U.S. Product Supplied of Petroleum Products (Thousand Barrels per Day)':'y'}, inplace = True)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Let's also include our prices in our dataframe, and concatenate everthing together."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 807,
   "metadata": {},
   "outputs": [],
   "source": [
    "#Concatenation\n",
    "metrics = pd.DataFrame(data = \n",
    "            (\n",
    "                \n",
    "            imports[\"Weekly U.S. Imports of Crude Oil  (Thousand Barrels per Day)\"],\n",
    "            exports['Weekly U.S. Exports of Crude Oil  (Thousand Barrels per Day)'],\n",
    "            consumption['Weekly U.S. Product Supplied of Petroleum Products  (Thousand Barrels per Day)'])\n",
    "                       \n",
    "            ).transpose()\n",
    "\n",
    "#Create a prices dataframe and append it to metrics.\n",
    "prices_df = pd.read_csv('WTI_price.csv')\n",
    "prices_df.loc[prices_df['Date'] == 'Jan 02, 2019']\n",
    "prices_df.index = pd.to_datetime(prices_df['Date'])\n",
    "prices_df.drop('Date', axis = 1, inplace = True)\n",
    "metrics['prices'] = prices_df\n",
    "metrics.dropna(inplace = True)\n",
    "\n",
    "#Create a numpy array for plotting.\n",
    "prices_plot = np.array(metrics['prices']['2019-01-01':])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 588,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "INFO:fbprophet:Disabling weekly seasonality. Run prophet with weekly_seasonality=True to override this.\n",
      "INFO:fbprophet:Disabling daily seasonality. Run prophet with daily_seasonality=True to override this.\n",
      "INFO:fbprophet:Disabling weekly seasonality. Run prophet with weekly_seasonality=True to override this.\n",
      "INFO:fbprophet:Disabling daily seasonality. Run prophet with daily_seasonality=True to override this.\n",
      "INFO:fbprophet:Disabling weekly seasonality. Run prophet with weekly_seasonality=True to override this.\n",
      "INFO:fbprophet:Disabling daily seasonality. Run prophet with daily_seasonality=True to override this.\n"
     ]
    }
   ],
   "source": [
    "#Building the models for each, and creating dataframes that have future dates as the index for the model to feed into.\n",
    "#Final line for each section is the prediction itself.\n",
    "m_imports = Prophet()\n",
    "m_imports.fit(imports)\n",
    "future_imports = m_imports.make_future_dataframe(periods = 500)\n",
    "forecast_imports = m_imports.predict(future_imports)\n",
    "\n",
    "m_exports = Prophet()\n",
    "m_exports.fit(exports)\n",
    "future_exports = m_exports.make_future_dataframe(periods = 500)\n",
    "forecast_exports = m_exports.predict(future_exports)\n",
    "\n",
    "m_consumption = Prophet()\n",
    "m_consumption.fit(exports)\n",
    "future_consumption  = m_consumption.make_future_dataframe(periods = 500)\n",
    "forecast_consumption = m_consumption.predict(future_consumption)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Let's figure out where to slice our data to use for the test predictions."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 574,
   "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>trend</th>\n",
       "      <th>yhat_lower</th>\n",
       "      <th>yhat_upper</th>\n",
       "      <th>trend_lower</th>\n",
       "      <th>trend_upper</th>\n",
       "      <th>additive_terms</th>\n",
       "      <th>additive_terms_lower</th>\n",
       "      <th>additive_terms_upper</th>\n",
       "      <th>yearly</th>\n",
       "      <th>yearly_lower</th>\n",
       "      <th>yearly_upper</th>\n",
       "      <th>multiplicative_terms</th>\n",
       "      <th>multiplicative_terms_lower</th>\n",
       "      <th>multiplicative_terms_upper</th>\n",
       "      <th>yhat</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>1560</th>\n",
       "      <td>2019-08-31</td>\n",
       "      <td>7461.747131</td>\n",
       "      <td>6935.302076</td>\n",
       "      <td>8429.959077</td>\n",
       "      <td>7461.747131</td>\n",
       "      <td>7461.747131</td>\n",
       "      <td>202.160454</td>\n",
       "      <td>202.160454</td>\n",
       "      <td>202.160454</td>\n",
       "      <td>202.160454</td>\n",
       "      <td>202.160454</td>\n",
       "      <td>202.160454</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>7663.907585</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "             ds        trend   yhat_lower   yhat_upper  trend_lower  \\\n",
       "1560 2019-08-31  7461.747131  6935.302076  8429.959077  7461.747131   \n",
       "\n",
       "      trend_upper  additive_terms  additive_terms_lower  additive_terms_upper  \\\n",
       "1560  7461.747131      202.160454            202.160454            202.160454   \n",
       "\n",
       "          yearly  yearly_lower  yearly_upper  multiplicative_terms  \\\n",
       "1560  202.160454    202.160454    202.160454                   0.0   \n",
       "\n",
       "      multiplicative_terms_lower  multiplicative_terms_upper         yhat  \n",
       "1560                         0.0                         0.0  7663.907585  "
      ]
     },
     "execution_count": 574,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "forecast_imports.loc[forecast_imports['ds'] == '2019-08-31']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 579,
   "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>trend</th>\n",
       "      <th>yhat_lower</th>\n",
       "      <th>yhat_upper</th>\n",
       "      <th>trend_lower</th>\n",
       "      <th>trend_upper</th>\n",
       "      <th>additive_terms</th>\n",
       "      <th>additive_terms_lower</th>\n",
       "      <th>additive_terms_upper</th>\n",
       "      <th>yearly</th>\n",
       "      <th>yearly_lower</th>\n",
       "      <th>yearly_upper</th>\n",
       "      <th>multiplicative_terms</th>\n",
       "      <th>multiplicative_terms_lower</th>\n",
       "      <th>multiplicative_terms_upper</th>\n",
       "      <th>yhat</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>1503</th>\n",
       "      <td>2019-08-31</td>\n",
       "      <td>2300.795666</td>\n",
       "      <td>1993.173946</td>\n",
       "      <td>2534.168587</td>\n",
       "      <td>2300.795666</td>\n",
       "      <td>2300.795666</td>\n",
       "      <td>-46.154994</td>\n",
       "      <td>-46.154994</td>\n",
       "      <td>-46.154994</td>\n",
       "      <td>-46.154994</td>\n",
       "      <td>-46.154994</td>\n",
       "      <td>-46.154994</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>2254.640671</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "             ds        trend   yhat_lower   yhat_upper  trend_lower  \\\n",
       "1503 2019-08-31  2300.795666  1993.173946  2534.168587  2300.795666   \n",
       "\n",
       "      trend_upper  additive_terms  additive_terms_lower  additive_terms_upper  \\\n",
       "1503  2300.795666      -46.154994            -46.154994            -46.154994   \n",
       "\n",
       "         yearly  yearly_lower  yearly_upper  multiplicative_terms  \\\n",
       "1503 -46.154994    -46.154994    -46.154994                   0.0   \n",
       "\n",
       "      multiplicative_terms_lower  multiplicative_terms_upper         yhat  \n",
       "1503                         0.0                         0.0  2254.640671  "
      ]
     },
     "execution_count": 579,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "forecast_exports.loc[forecast_exports['ds'] == '2019-08-31']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 580,
   "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>trend</th>\n",
       "      <th>yhat_lower</th>\n",
       "      <th>yhat_upper</th>\n",
       "      <th>trend_lower</th>\n",
       "      <th>trend_upper</th>\n",
       "      <th>additive_terms</th>\n",
       "      <th>additive_terms_lower</th>\n",
       "      <th>additive_terms_upper</th>\n",
       "      <th>yearly</th>\n",
       "      <th>yearly_lower</th>\n",
       "      <th>yearly_upper</th>\n",
       "      <th>multiplicative_terms</th>\n",
       "      <th>multiplicative_terms_lower</th>\n",
       "      <th>multiplicative_terms_upper</th>\n",
       "      <th>yhat</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>1503</th>\n",
       "      <td>2019-08-31</td>\n",
       "      <td>2300.795666</td>\n",
       "      <td>1973.838858</td>\n",
       "      <td>2528.354688</td>\n",
       "      <td>2300.795666</td>\n",
       "      <td>2300.795666</td>\n",
       "      <td>-46.154994</td>\n",
       "      <td>-46.154994</td>\n",
       "      <td>-46.154994</td>\n",
       "      <td>-46.154994</td>\n",
       "      <td>-46.154994</td>\n",
       "      <td>-46.154994</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>2254.640671</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "             ds        trend   yhat_lower   yhat_upper  trend_lower  \\\n",
       "1503 2019-08-31  2300.795666  1973.838858  2528.354688  2300.795666   \n",
       "\n",
       "      trend_upper  additive_terms  additive_terms_lower  additive_terms_upper  \\\n",
       "1503  2300.795666      -46.154994            -46.154994            -46.154994   \n",
       "\n",
       "         yearly  yearly_lower  yearly_upper  multiplicative_terms  \\\n",
       "1503 -46.154994    -46.154994    -46.154994                   0.0   \n",
       "\n",
       "      multiplicative_terms_lower  multiplicative_terms_upper         yhat  \n",
       "1503                         0.0                         0.0  2254.640671  "
      ]
     },
     "execution_count": 580,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "forecast_consumption.loc[forecast_consumption['ds'] == '2019-08-31']"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Now that we have our data, let's concatenate everything together for the random forest to interpret."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 809,
   "metadata": {},
   "outputs": [],
   "source": [
    "#Create a new dataframe containing just test data.\n",
    "imports_predictions = forecast_imports.iloc[1560:]['yhat']\n",
    "exports_predictions = forecast_exports.iloc[1503:]['yhat']\n",
    "consumption_predictions = forecast_consumption.iloc[1503:]['yhat']\n",
    "\n",
    "future_predictions = pd.DataFrame((imports_predictions, exports_predictions, consumption_predictions)).transpose()\n",
    "future_predictions.dropna(inplace = True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 665,
   "metadata": {},
   "outputs": [],
   "source": [
    "#Split up the data into training and test sets.\n",
    "train_metrics, test_metrics, train_prices, test_prices = train_test_split(\n",
    "     metrics[:'2019-01-01'].drop('prices', axis = 1), metrics['prices'][:'2019-01-01'], test_size = .25, random_state = 1)\n",
    "rf = RandomForestRegressor(n_estimators = 2000, random_state = 1)\n",
    "rf.fit(train_metrics, train_prices)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 783,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "86.98"
      ]
     },
     "execution_count": 783,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#Calculate RSS and MAPE error metrics.\n",
    "\n",
    "predictions_rf = rf.predict(test_metrics)\n",
    "abs_errors = abs(predictions_rf - test_prices)\n",
    "mape = 100 * (abs_errors / predictions_rf)\n",
    "accuracy = round(100 - np.mean(mape), 2)\n",
    "accuracy"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Our random forest model has an accuracy of about 87%. This isn't fantastic, but again we're just trying to familiarize ourselves with the methodologies."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 784,
   "metadata": {},
   "outputs": [],
   "source": [
    "future_predictions_rf = rf.predict(future_predictions)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Let's plot our 2020 predictions, as well as our test set predictions."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 794,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Text(0.5, 1.0, 'WTI Crude Oil Price Prediction through 2020')"
      ]
     },
     "execution_count": 794,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 936x648 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.figure(figsize = (13,9))\n",
    "plt.plot(future_predictions_rf)\n",
    "plt.xlabel('Days since September 01, 2019')\n",
    "plt.ylabel('Price')\n",
    "plt.title('WTI Crude Oil Price Prediction through 2020')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Now, let's predict the test data and compare to the actual data."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 786,
   "metadata": {},
   "outputs": [],
   "source": [
    "test_predictions_rf = rf.predict(metrics['2019-01-01':].drop('prices', axis = 1))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 793,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.legend.Legend at 0x1a4f1beeb8>"
      ]
     },
     "execution_count": 793,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 936x648 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.figure(figsize = (13,9))\n",
    "plt.plot(test_predictions_rf, label = 'Predicted')\n",
    "plt.xlabel('Weeks since Jan 01, 2019')\n",
    "plt.ylabel('Price in USD')\n",
    "plt.plot(prices_plot, label = 'Actual')\n",
    "plt.legend()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Let's calculate MSE and the average price."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 796,
   "metadata": {},
   "outputs": [],
   "source": [
    "test_error = np.mean((test_predictions_rf - prices_plot) ** 2)\n",
    "average_price = np.mean(test_predictions_rf)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 797,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "31.771192724691904"
      ]
     },
     "execution_count": 797,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "test_error"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 798,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "58.23082080645136"
      ]
     },
     "execution_count": 798,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "average_price"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "So, we have a test error (MSE) of 31.77, and average price of $58.23 for 2020. Obviously, this has performed better than the LSTM, but there is some indication that the data could be overfitting somewhat due to the apparent randomness seen in the plot of test predictions vs. actual prices."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.7.3"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}