Model_tesla_SARIMA_and_NLP.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Imports"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {},
   "outputs": [],
   "source": [
    "# The basics\n",
    "import pandas as pd\n",
    "import ast\n",
    "import numpy as np\n",
    "import matplotlib.pyplot as plt\n",
    "import matplotlib.ticker as mtick\n",
    "import time as t\n",
    "\n",
    "# Preprocessing\n",
    "from sklearn.preprocessing import MultiLabelBinarizer\n",
    "\n",
    "# Natural Language Processing\n",
    "from nltk.stem import WordNetLemmatizer\n",
    "from nltk.tokenize import RegexpTokenizer\n",
    "from nltk.corpus import stopwords\n",
    "from sklearn.feature_extraction.text import CountVectorizer\n",
    "\n",
    "# Models\n",
    "from sklearn.model_selection import train_test_split\n",
    "import statsmodels.api as sm\n",
    "from sklearn.linear_model import LinearRegression\n",
    "from xgboost import XGBRegressor\n",
    "from xgboost import XGBClassifier\n",
    "from sklearn.model_selection import GridSearchCV\n",
    "\n",
    "# Model evaluation\n",
    "from sklearn.model_selection import cross_val_score, cross_val_predict\n",
    "from sklearn import metrics\n",
    "from sklearn.metrics import confusion_matrix, accuracy_score, precision_score, roc_auc_score"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Build in News and Tweets"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "# NLP processing from the nltk package\n",
    "# Instantiate tokenizer to break text into lists of words\n",
    "#tokenizer = RegexpTokenizer(r'\\w+')\n",
    "tokenizer = RegexpTokenizer(r'[a-zA-Z0-9_@-]+')\n",
    "# Instantiate lemmatizer to turn plural words into singular for better comparison\n",
    "lemmatizer = WordNetLemmatizer()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Elon Musk's Tweets"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "elon = pd.read_csv('twitter/elonmusk_tweets.csv')\n",
    "\n",
    "# A bit of preprocessing and creating the datetime\n",
    "elon.text = elon.text.str.lstrip('b')\n",
    "elon.created_at = pd.to_datetime(elon.created_at)\n",
    "elon['date'] = elon.created_at.apply(lambda x: x.date())\n",
    "\n",
    "# First let's turn each title into a list of words\n",
    "elon['nlp_text'] = elon.text.apply(lambda x: tokenizer.tokenize(x.lower()))\n",
    "# Then we'll use the lemmatizer to change plurals to singulars to group them\n",
    "elon.nlp_text = elon.nlp_text.apply(lambda x: [lemmatizer.lemmatize(i) for i in x])\n",
    "# And rejoin the lists of words to one string for the count vectorizer\n",
    "elon.nlp_text = elon.nlp_text.apply(lambda x: ' '.join(x))\n",
    "\n",
    "# Recombine and create database with datetime index and nlp text ready to go\n",
    "elon.nlp_text = elon.nlp_text + ' '\n",
    "elon = elon.groupby('date')['nlp_text'].sum()\n",
    "elon = pd.DataFrame(elon)\n",
    "elon.columns = ['elon_tweet']"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Tesla company tweets"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "tesla = pd.read_csv('twitter/Tesla_tweets.csv')\n",
    "\n",
    "# A bit of preprocessing and creating the datetime\n",
    "tesla.text = tesla.text.str.lstrip('b')\n",
    "tesla.created_at = pd.to_datetime(tesla.created_at)\n",
    "tesla['date'] = tesla.created_at.apply(lambda x: x.date())\n",
    "\n",
    "# First let's turn each title into a list of words\n",
    "tesla['nlp_text'] = tesla.text.apply(lambda x: tokenizer.tokenize(x.lower()))\n",
    "# Then we'll use the lemmatizer to change plurals to singulars to group them\n",
    "tesla.nlp_text = tesla.nlp_text.apply(lambda x: [lemmatizer.lemmatize(i) for i in x])\n",
    "# And rejoin the lists of words to one string for the count vectorizer\n",
    "tesla.nlp_text = tesla.nlp_text.apply(lambda x: ' '.join(x))\n",
    "\n",
    "# Recombine and create database with datetime index and nlp text ready to go\n",
    "tesla.nlp_text = tesla.nlp_text + ' '\n",
    "tesla = tesla.groupby('date')['nlp_text'].sum()\n",
    "tesla = pd.DataFrame(tesla)\n",
    "tesla.columns = ['tesla_tweet']"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Techmeme headlines"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [],
   "source": [
    "techmeme = pd.read_csv('web_scraping/techmeme.csv', index_col=0)\n",
    "\n",
    "# A bit of preprocessing and creating the datetime\n",
    "techmeme['date_time'] = pd.to_datetime(techmeme.date)\n",
    "techmeme.date = techmeme.date_time.apply(lambda x: x.date())\n",
    "# Turn list into actual list\n",
    "techmeme.extra_sources = techmeme.extra_sources.apply(lambda x: ast.literal_eval(x))\n",
    "\n",
    "techmeme['sources'] = techmeme.extra_sources.copy()\n",
    "# Ben helped me with this lovely for loop to combine source columns\n",
    "for i, list_ in enumerate(techmeme.sources):\n",
    "    list_.append(techmeme.original_source[i])\n",
    "\n",
    "# Prepare titles to be combined    \n",
    "techmeme.original_title = techmeme.original_title.apply(lambda x: x.replace('\\n', ''))\n",
    "techmeme.extra_titles = techmeme.extra_titles.apply(lambda x: x.lstrip('[').rstrip(']'))\n",
    "techmeme['titles'] = techmeme.original_title + ' ' + techmeme.extra_titles\n",
    "\n",
    "# First let's turn each title into a list of words\n",
    "techmeme['nlp_text'] = techmeme.titles.apply(lambda x: tokenizer.tokenize(x.lower()))\n",
    "# Then we'll use the lemmatizer to change plurals to singulars to group them\n",
    "techmeme.nlp_text = techmeme.nlp_text.apply(lambda x: [lemmatizer.lemmatize(i) for i in x])\n",
    "# And rejoin the lists of words to one string for the count vectorizer\n",
    "techmeme.nlp_text = techmeme.nlp_text.apply(lambda x: ' '.join(x))\n",
    "\n",
    "# Rename columns\n",
    "techmeme.columns = ['date', 'original_source', 'original_title', 'extra_sources',\n",
    "       'extra_titles', 'date_time', 'news_sources', 'titles', 'news_text']\n",
    "\n",
    "# Combine entries by data and create new dataframe with datetime index\n",
    "techmeme.news_text = techmeme.news_text + ' '\n",
    "techmeme = techmeme.groupby('date')['news_sources', 'news_text'].sum()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Prep Evaluation Functions"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "# My go=to for classification evaluation\n",
    "def eval_sklearn_model(y_true, predictions, model=None, X=None):\n",
    "    \"\"\"This function takes the true values for y and the predictions made by the model and prints out the confusion matrix along with Accuracy, Precision, and, if model and X provided, Roc_Auc Scores.\"\"\"\n",
    "    cnf_matrix = confusion_matrix(y_true, predictions)\n",
    "\n",
    "    print('True Negative: ', cnf_matrix[0, 0], '| False Positive: ', cnf_matrix[0, 1])\n",
    "    print('False Negative: ', cnf_matrix[1, 0], '| True Positive: ', cnf_matrix[1, 1], '\\n')\n",
    "\n",
    "    sensitivity = cnf_matrix[1, 1]/ (cnf_matrix[1, 0] + cnf_matrix[1, 1])\n",
    "    specificity = cnf_matrix[0, 0]/ (cnf_matrix[0, 1] + cnf_matrix[0, 0])\n",
    "\n",
    "    print('Sensitivity (TP/ TP + FN): ', sensitivity)\n",
    "    print('Specificity (TN/ TN + FP): ', specificity, '\\n')\n",
    "\n",
    "    print('Accuracy: ', accuracy_score(y_true, predictions, normalize=True))\n",
    "    print('Precision: ', precision_score(y_true, predictions))\n",
    "    if model != None:\n",
    "        print('Roc-Auc: ', roc_auc_score(y_true, [x[1] for x in model.predict_proba(X)]))\n",
    "    else:\n",
    "        pass\n",
    "    print('\\n')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "# This is used to calculate the profit that would be expected if a model has been used to tell when to buy or sell\n",
    "def profit_calculator(dataframe, predict_col, early_stop=None):  \n",
    "    money_counter = 0\n",
    "    transactions = 0\n",
    "    own = False\n",
    "    last_buy = 0\n",
    "    current_price = 0\n",
    "    for num, i in enumerate(dataframe.iloc[:early_stop].iterrows()):\n",
    "        if i[1][predict_col] == 1 and own == False:\n",
    "            money_counter -= i[1]['next_day_open']\n",
    "            own = True\n",
    "            transactions += 1\n",
    "            last_buy = i[1]['next_day_open']\n",
    "        elif i[1][predict_col] == 0 and own == True:\n",
    "            money_counter += i[1]['next_day_open']\n",
    "            own = False\n",
    "            transactions += 1\n",
    "        else:\n",
    "            pass\n",
    "        current_price=i[1]['next_day_open']\n",
    "    print('Own?', own)\n",
    "    print('Last Buying price: $', last_buy)\n",
    "    print('Current price: $', current_price)\n",
    "    print('Cash? $', money_counter)\n",
    "    if own == True:\n",
    "        print('Profit: $', current_price + money_counter)\n",
    "    else:\n",
    "        print('Profit: $', money_counter)\n",
    "    print('Number of Transactins:', transactions)\n",
    "    print('Cost of transactions: $', transactions*5)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "metadata": {},
   "outputs": [],
   "source": [
    "def plot_profit(dataframe, predict_col, early_stop=None):  \n",
    "    X = []\n",
    "    y = []\n",
    "    money_counter = 0\n",
    "    transactions = 0\n",
    "    own = False\n",
    "    last_buy = 0\n",
    "    current_price = 0\n",
    "    for num, i in enumerate(dataframe.iloc[:early_stop].iterrows()):\n",
    "        if i[1][predict_col] == 1 and own == False:\n",
    "            money_counter -= i[1]['next_day_open']\n",
    "            own = True\n",
    "            transactions += 1\n",
    "            last_buy = i[1]['next_day_open']\n",
    "            X.append(i[0])\n",
    "            y.append(money_counter + i[1]['next_day_open'])\n",
    "        elif i[1][predict_col] == 0 and own == True:\n",
    "            money_counter += i[1]['next_day_open']\n",
    "            own = False\n",
    "            transactions += 1\n",
    "            X.append(i[0])\n",
    "            y.append(money_counter)\n",
    "        else:\n",
    "            pass\n",
    "        current_price=i[1]['next_day_open']\n",
    "    fig, ax = plt.subplots()\n",
    "    ax=pd.DataFrame(data=y, index=X)[0].plot(figsize=(12,8), title= 'Profit over Time', fontsize=14, ax=ax)\n",
    "    plt.title('Profit over Time', fontsize=20)\n",
    "    plt.xlabel('Date', fontsize=16)\n",
    "    plt.ylabel('Stock Price', fontsize=16)\n",
    "    fmt = '${x:,.0f}'\n",
    "    tick = mtick.StrMethodFormatter(fmt)\n",
    "    ax.yaxis.set_major_formatter(tick) \n",
    "    plt.yticks()\n",
    "    plt.xticks(rotation=25)\n",
    "    plt.show()\n",
    "    return X, y"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# First Models"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Combine them onto stock data\n",
    "\n",
    "In merging, I noticed that I was not retaining the same amounts of information.  I realized that since I do not have stock information on weekends, I have nothing to link tweets or news that happens over the weekend.  I should push this to help predict Monday prices, but I am not yet sure how to do that."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 339,
   "metadata": {},
   "outputs": [],
   "source": [
    "stock = pd.read_csv('TESLAHistoricalQuotes.csv')\n",
    "stock = stock.drop(0,0)\n",
    "stock.date = pd.to_datetime(stock.date)\n",
    "stock.date = stock.date.apply(lambda x: x.date())\n",
    "stock = stock.sort_values('date')\n",
    "stock.volume = stock.volume.astype(float)\n",
    "stock.set_index(stock.date, inplace=True)\n",
    "stock = stock.drop('date', 1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 340,
   "metadata": {},
   "outputs": [],
   "source": [
    "stock = pd.merge(stock, techmeme, how='left', left_index=True, right_index=True)\n",
    "stock = pd.merge(stock, tesla, how='left', left_index=True, right_index=True)\n",
    "stock = pd.merge(stock, elon, how='left', left_index=True, right_index=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 341,
   "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>close</th>\n",
       "      <th>volume</th>\n",
       "      <th>open</th>\n",
       "      <th>high</th>\n",
       "      <th>low</th>\n",
       "      <th>news_sources</th>\n",
       "      <th>news_text</th>\n",
       "      <th>tesla_tweet</th>\n",
       "      <th>elon_tweet</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>date</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>2017-11-16</th>\n",
       "      <td>312.50</td>\n",
       "      <td>5794409.0</td>\n",
       "      <td>313.99</td>\n",
       "      <td>318.14</td>\n",
       "      <td>311.30</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2017-11-17</th>\n",
       "      <td>315.05</td>\n",
       "      <td>13720830.0</td>\n",
       "      <td>325.67</td>\n",
       "      <td>326.67</td>\n",
       "      <td>313.15</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>going into plaid http t co ouu9s7cuwi bamf htt...</td>\n",
       "      <td>tesla semi http t co 7vlz7f46ji 0 to 100 km h ...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2017-11-20</th>\n",
       "      <td>308.74</td>\n",
       "      <td>8241821.0</td>\n",
       "      <td>313.79</td>\n",
       "      <td>315.50</td>\n",
       "      <td>304.75</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>@schesnais good point</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2017-11-21</th>\n",
       "      <td>317.81</td>\n",
       "      <td>7256381.0</td>\n",
       "      <td>310.86</td>\n",
       "      <td>318.23</td>\n",
       "      <td>308.71</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>initial hat offering going great w over 300k i...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2017-11-22</th>\n",
       "      <td>312.60</td>\n",
       "      <td>4913283.0</td>\n",
       "      <td>316.77</td>\n",
       "      <td>317.42</td>\n",
       "      <td>311.84</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "             close      volume    open    high     low news_sources news_text  \\\n",
       "date                                                                            \n",
       "2017-11-16  312.50   5794409.0  313.99  318.14  311.30          NaN       NaN   \n",
       "2017-11-17  315.05  13720830.0  325.67  326.67  313.15          NaN       NaN   \n",
       "2017-11-20  308.74   8241821.0  313.79  315.50  304.75          NaN       NaN   \n",
       "2017-11-21  317.81   7256381.0  310.86  318.23  308.71          NaN       NaN   \n",
       "2017-11-22  312.60   4913283.0  316.77  317.42  311.84          NaN       NaN   \n",
       "\n",
       "                                                  tesla_tweet  \\\n",
       "date                                                            \n",
       "2017-11-16                                                NaN   \n",
       "2017-11-17  going into plaid http t co ouu9s7cuwi bamf htt...   \n",
       "2017-11-20                                                NaN   \n",
       "2017-11-21                                                NaN   \n",
       "2017-11-22                                                NaN   \n",
       "\n",
       "                                                   elon_tweet  \n",
       "date                                                           \n",
       "2017-11-16                                                NaN  \n",
       "2017-11-17  tesla semi http t co 7vlz7f46ji 0 to 100 km h ...  \n",
       "2017-11-20                             @schesnais good point   \n",
       "2017-11-21  initial hat offering going great w over 300k i...  \n",
       "2017-11-22                                                NaN  "
      ]
     },
     "execution_count": 341,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "stock.tail()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Note on earliest NLP relevant info\n",
    "1. elon earliest tweet 2013-03-23\n",
    "2. tesla earliest tweet 2014-02-18\n",
    "3. techmeme earliest story 2006-12-07\n",
    "\n",
    "```stock.iloc[915]``` is 2014-02-18"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Run ARIMA Model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 342,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<table class=\"simpletable\">\n",
       "<caption>Statespace Model Results</caption>\n",
       "<tr>\n",
       "  <th>Dep. Variable:</th>                <td>close</td>             <th>  No. Observations:  </th>   <td>1866</td>   \n",
       "</tr>\n",
       "<tr>\n",
       "  <th>Model:</th>           <td>SARIMAX(1, 1, 3)x(0, 1, 1, 80)</td> <th>  Log Likelihood     </th> <td>-5456.155</td>\n",
       "</tr>\n",
       "<tr>\n",
       "  <th>Date:</th>                   <td>Wed, 20 Dec 2017</td>        <th>  AIC                </th> <td>10924.310</td>\n",
       "</tr>\n",
       "<tr>\n",
       "  <th>Time:</th>                       <td>12:30:04</td>            <th>  BIC                </th> <td>10957.499</td>\n",
       "</tr>\n",
       "<tr>\n",
       "  <th>Sample:</th>                    <td>06-29-2010</td>           <th>  HQIC               </th> <td>10936.539</td>\n",
       "</tr>\n",
       "<tr>\n",
       "  <th></th>                          <td>- 11-22-2017</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>ar.L1</th>    <td>    0.2551</td> <td>    0.425</td> <td>    0.601</td> <td> 0.548</td> <td>   -0.577</td> <td>    1.087</td>\n",
       "</tr>\n",
       "<tr>\n",
       "  <th>ma.L1</th>    <td>   -0.2205</td> <td>    0.425</td> <td>   -0.518</td> <td> 0.604</td> <td>   -1.054</td> <td>    0.613</td>\n",
       "</tr>\n",
       "<tr>\n",
       "  <th>ma.L2</th>    <td>    0.0068</td> <td>    0.027</td> <td>    0.253</td> <td> 0.800</td> <td>   -0.046</td> <td>    0.059</td>\n",
       "</tr>\n",
       "<tr>\n",
       "  <th>ma.L3</th>    <td>   -0.0481</td> <td>    0.018</td> <td>   -2.700</td> <td> 0.007</td> <td>   -0.083</td> <td>   -0.013</td>\n",
       "</tr>\n",
       "<tr>\n",
       "  <th>ma.S.L80</th> <td>   -0.9995</td> <td>    2.558</td> <td>   -0.391</td> <td> 0.696</td> <td>   -6.013</td> <td>    4.014</td>\n",
       "</tr>\n",
       "<tr>\n",
       "  <th>sigma2</th>   <td>   22.9819</td> <td>   58.554</td> <td>    0.392</td> <td> 0.695</td> <td>  -91.782</td> <td>  137.746</td>\n",
       "</tr>\n",
       "</table>\n",
       "<table class=\"simpletable\">\n",
       "<tr>\n",
       "  <th>Ljung-Box (Q):</th>          <td>33.16</td> <th>  Jarque-Bera (JB):  </th> <td>1844.42</td>\n",
       "</tr>\n",
       "<tr>\n",
       "  <th>Prob(Q):</th>                <td>0.77</td>  <th>  Prob(JB):          </th>  <td>0.00</td>  \n",
       "</tr>\n",
       "<tr>\n",
       "  <th>Heteroskedasticity (H):</th> <td>40.33</td> <th>  Skew:              </th>  <td>-0.18</td> \n",
       "</tr>\n",
       "<tr>\n",
       "  <th>Prob(H) (two-sided):</th>    <td>0.00</td>  <th>  Kurtosis:          </th>  <td>7.97</td>  \n",
       "</tr>\n",
       "</table>"
      ],
      "text/plain": [
       "<class 'statsmodels.iolib.summary.Summary'>\n",
       "\"\"\"\n",
       "                                 Statespace Model Results                                 \n",
       "==========================================================================================\n",
       "Dep. Variable:                              close   No. Observations:                 1866\n",
       "Model:             SARIMAX(1, 1, 3)x(0, 1, 1, 80)   Log Likelihood               -5456.155\n",
       "Date:                            Wed, 20 Dec 2017   AIC                          10924.310\n",
       "Time:                                    12:30:04   BIC                          10957.499\n",
       "Sample:                                06-29-2010   HQIC                         10936.539\n",
       "                                     - 11-22-2017                                         \n",
       "Covariance Type:                              opg                                         \n",
       "==============================================================================\n",
       "                 coef    std err          z      P>|z|      [0.025      0.975]\n",
       "------------------------------------------------------------------------------\n",
       "ar.L1          0.2551      0.425      0.601      0.548      -0.577       1.087\n",
       "ma.L1         -0.2205      0.425     -0.518      0.604      -1.054       0.613\n",
       "ma.L2          0.0068      0.027      0.253      0.800      -0.046       0.059\n",
       "ma.L3         -0.0481      0.018     -2.700      0.007      -0.083      -0.013\n",
       "ma.S.L80      -0.9995      2.558     -0.391      0.696      -6.013       4.014\n",
       "sigma2        22.9819     58.554      0.392      0.695     -91.782     137.746\n",
       "===================================================================================\n",
       "Ljung-Box (Q):                       33.16   Jarque-Bera (JB):              1844.42\n",
       "Prob(Q):                              0.77   Prob(JB):                         0.00\n",
       "Heteroskedasticity (H):              40.33   Skew:                            -0.18\n",
       "Prob(H) (two-sided):                  0.00   Kurtosis:                         7.97\n",
       "===================================================================================\n",
       "\n",
       "Warnings:\n",
       "[1] Covariance matrix calculated using the outer product of gradients (complex-step).\n",
       "\"\"\""
      ]
     },
     "execution_count": 342,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "mod = sm.tsa.statespace.SARIMAX(stock.close, trend='n', order=(1,1,3), seasonal_order=(0,1,1,80))\n",
    "results = mod.fit()\n",
    "results.summary()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 343,
   "metadata": {},
   "outputs": [],
   "source": [
    "stock['forecast'] = results.predict(dynamic=False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 344,
   "metadata": {},
   "outputs": [],
   "source": [
    "stock['next_day_open'] = stock.open.shift(-1) "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 345,
   "metadata": {},
   "outputs": [],
   "source": [
    "stock['target'] = stock.close.shift(-1)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Use CountVectorizer"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 346,
   "metadata": {},
   "outputs": [],
   "source": [
    "temp_df = stock[stock.news_text.notnull()]\n",
    "cvec = CountVectorizer(stop_words=stopwords.words('english'), ngram_range=(1,2), max_features=1000)\n",
    "cvec.fit(temp_df.news_text.dropna())\n",
    "cvec_df  = pd.DataFrame(cvec.transform(temp_df.news_text.dropna()).todense(), columns=cvec.get_feature_names())\n",
    "cvec_df.index = temp_df.index\n",
    "cvec_df.columns += \" _news_\"\n",
    "stock = stock.join(cvec_df).drop('news_text',1)\n",
    "\n",
    "temp_df = stock[stock.tesla_tweet.notnull()]\n",
    "cvec.fit(temp_df.tesla_tweet.dropna())\n",
    "cvec_df  = pd.DataFrame(cvec.transform(temp_df.tesla_tweet.dropna()).todense(), columns=cvec.get_feature_names())\n",
    "cvec_df.index = temp_df.index\n",
    "cvec_df.columns += \" _tesla_\"\n",
    "stock = stock.join(cvec_df).drop('tesla_tweet',1)\n",
    "\n",
    "temp_df = stock[stock.elon_tweet.notnull()]\n",
    "cvec.fit(temp_df.elon_tweet.dropna())\n",
    "cvec_df  = pd.DataFrame(cvec.transform(temp_df.elon_tweet.dropna()).todense(), columns=cvec.get_feature_names())\n",
    "cvec_df.index = temp_df.index\n",
    "cvec_df.columns += \" _elon_\"\n",
    "stock = stock.join(cvec_df).drop('elon_tweet',1)\n",
    "\n",
    "del temp_df"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 347,
   "metadata": {},
   "outputs": [],
   "source": [
    "stock.news_sources = stock.news_sources.fillna('')\n",
    "mlb = MultiLabelBinarizer()\n",
    "stock = stock.join(pd.DataFrame(mlb.fit_transform(stock.news_sources), columns=mlb.classes_, index=stock.index))\n",
    "stock = stock.drop('news_sources', 1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 348,
   "metadata": {},
   "outputs": [],
   "source": [
    "stock = stock.fillna(0)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 349,
   "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>close</th>\n",
       "      <th>volume</th>\n",
       "      <th>open</th>\n",
       "      <th>high</th>\n",
       "      <th>low</th>\n",
       "      <th>forecast</th>\n",
       "      <th>next_day_open</th>\n",
       "      <th>target</th>\n",
       "      <th>000 _news_</th>\n",
       "      <th>10 _news_</th>\n",
       "      <th>...</th>\n",
       "      <th>digitaltrends.com</th>\n",
       "      <th>documentcloud.org</th>\n",
       "      <th>everythingiCafe</th>\n",
       "      <th>iDownloadBlog.com</th>\n",
       "      <th>iDownloadBlog.com iDownloadBlog.com</th>\n",
       "      <th>iLounge</th>\n",
       "      <th>iMore</th>\n",
       "      <th>iPhone Hacks</th>\n",
       "      <th>iPhone in Canada Blog</th>\n",
       "      <th>lists.swift.org Mailing Lists</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>date</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>2010-06-29</th>\n",
       "      <td>23.89</td>\n",
       "      <td>18751150.0</td>\n",
       "      <td>19.00</td>\n",
       "      <td>25.0000</td>\n",
       "      <td>17.54</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>25.79</td>\n",
       "      <td>23.83</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2010-06-30</th>\n",
       "      <td>23.83</td>\n",
       "      <td>17165210.0</td>\n",
       "      <td>25.79</td>\n",
       "      <td>30.4192</td>\n",
       "      <td>23.30</td>\n",
       "      <td>23.890019</td>\n",
       "      <td>25.00</td>\n",
       "      <td>21.96</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2010-07-01</th>\n",
       "      <td>21.96</td>\n",
       "      <td>8216789.0</td>\n",
       "      <td>25.00</td>\n",
       "      <td>25.9200</td>\n",
       "      <td>20.27</td>\n",
       "      <td>23.830008</td>\n",
       "      <td>23.00</td>\n",
       "      <td>19.20</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2010-07-02</th>\n",
       "      <td>19.20</td>\n",
       "      <td>5135795.0</td>\n",
       "      <td>23.00</td>\n",
       "      <td>23.1000</td>\n",
       "      <td>18.71</td>\n",
       "      <td>21.959973</td>\n",
       "      <td>20.00</td>\n",
       "      <td>16.11</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2010-07-06</th>\n",
       "      <td>16.11</td>\n",
       "      <td>6858092.0</td>\n",
       "      <td>20.00</td>\n",
       "      <td>20.0000</td>\n",
       "      <td>15.83</td>\n",
       "      <td>19.199988</td>\n",
       "      <td>16.40</td>\n",
       "      <td>15.80</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 3306 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "            close      volume   open     high    low   forecast  \\\n",
       "date                                                              \n",
       "2010-06-29  23.89  18751150.0  19.00  25.0000  17.54   0.000000   \n",
       "2010-06-30  23.83  17165210.0  25.79  30.4192  23.30  23.890019   \n",
       "2010-07-01  21.96   8216789.0  25.00  25.9200  20.27  23.830008   \n",
       "2010-07-02  19.20   5135795.0  23.00  23.1000  18.71  21.959973   \n",
       "2010-07-06  16.11   6858092.0  20.00  20.0000  15.83  19.199988   \n",
       "\n",
       "            next_day_open  target  000 _news_  10 _news_  \\\n",
       "date                                                       \n",
       "2010-06-29          25.79   23.83         0.0        0.0   \n",
       "2010-06-30          25.00   21.96         0.0        0.0   \n",
       "2010-07-01          23.00   19.20         0.0        0.0   \n",
       "2010-07-02          20.00   16.11         0.0        0.0   \n",
       "2010-07-06          16.40   15.80         0.0        0.0   \n",
       "\n",
       "                        ...                digitaltrends.com  \\\n",
       "date                    ...                                    \n",
       "2010-06-29              ...                                0   \n",
       "2010-06-30              ...                                0   \n",
       "2010-07-01              ...                                0   \n",
       "2010-07-02              ...                                0   \n",
       "2010-07-06              ...                                0   \n",
       "\n",
       "            documentcloud.org  everythingiCafe  iDownloadBlog.com  \\\n",
       "date                                                                \n",
       "2010-06-29                  0                0                  0   \n",
       "2010-06-30                  0                0                  0   \n",
       "2010-07-01                  0                0                  0   \n",
       "2010-07-02                  0                0                  0   \n",
       "2010-07-06                  0                0                  0   \n",
       "\n",
       "            iDownloadBlog.com iDownloadBlog.com  iLounge  iMore  iPhone Hacks  \\\n",
       "date                                                                            \n",
       "2010-06-29                                    0        0      0             0   \n",
       "2010-06-30                                    0        0      0             0   \n",
       "2010-07-01                                    0        0      0             0   \n",
       "2010-07-02                                    0        0      0             0   \n",
       "2010-07-06                                    0        0      0             0   \n",
       "\n",
       "            iPhone in Canada Blog  lists.swift.org Mailing Lists  \n",
       "date                                                              \n",
       "2010-06-29                      0                              0  \n",
       "2010-06-30                      0                              0  \n",
       "2010-07-01                      0                              0  \n",
       "2010-07-02                      0                              0  \n",
       "2010-07-06                      0                              0  \n",
       "\n",
       "[5 rows x 3306 columns]"
      ]
     },
     "execution_count": 349,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "stock.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 351,
   "metadata": {},
   "outputs": [],
   "source": [
    "X_train, X_test, y_train, y_test = train_test_split(stock.drop(['target', 'close', 'open'], 1), stock['target'], test_size=0.3, random_state=42)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 353,
   "metadata": {},
   "outputs": [],
   "source": [
    "lr = LinearRegression()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 354,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "LinearRegression(copy_X=True, fit_intercept=True, n_jobs=1, normalize=False)"
      ]
     },
     "execution_count": 354,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "lr.fit(X_train, y_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 356,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Cross-validated scores: [ -2.13697039e+07  -7.38906389e+07  -1.90998464e+07  -1.94298254e+07\n",
      "  -1.98249349e+08  -1.82890011e+06]\n",
      "Average:  -55644710.655\n"
     ]
    }
   ],
   "source": [
    "scores = cross_val_score(lr, X_train, y_train, cv=6)\n",
    "print(\"Cross-validated scores:\", scores)\n",
    "print(\"Average: \", scores.mean())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 357,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Score: -1177536.34718\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAakAAAEKCAYAAACopKobAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x11f860d68>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "predictions = lr.predict(X_test)\n",
    "plt.scatter(y_test, predictions)\n",
    "plt.xlabel(\"True Values\")\n",
    "plt.ylabel(\"Predictions\")\n",
    "print(\"Score:\", lr.score(X_test, y_test))\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 359,
   "metadata": {},
   "outputs": [],
   "source": [
    "xgb = XGBRegressor(gamma=0, learning_rate=0.02, max_depth=3, n_estimators=100)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 360,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "XGBRegressor(base_score=0.5, colsample_bylevel=1, colsample_bytree=1, gamma=0,\n",
       "       learning_rate=0.02, max_delta_step=0, max_depth=3,\n",
       "       min_child_weight=1, missing=None, n_estimators=100, nthread=-1,\n",
       "       objective='reg:linear', reg_alpha=0, reg_lambda=1,\n",
       "       scale_pos_weight=1, seed=0, silent=True, subsample=1)"
      ]
     },
     "execution_count": 360,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "xgb.fit(X_train, y_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 361,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Score: 0.94280846105\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYgAAAEKCAYAAAAIO8L1AAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nKicIMzNL5QRhZmapnCDMzCzV/wfuaHT967aGDAAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x11f83aba8>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "predictions = xgb.predict(X_test)\n",
    "plt.scatter(y_test, predictions)\n",
    "plt.xlabel(\"True Values\")\n",
    "plt.ylabel(\"Predictions\")\n",
    "print(\"Score:\", xgb.score(X_test, y_test))\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 363,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "R2 Accuracy: 0.94280846105\n"
     ]
    }
   ],
   "source": [
    "accuracy = metrics.r2_score(y_test, predictions)\n",
    "print(\"R2 Accuracy:\", accuracy)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 364,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "MSE: 668.243820278\n"
     ]
    }
   ],
   "source": [
    "print(\"MSE:\", metrics.mean_squared_error(y_test, predictions))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Damn!  But, how much is the NLP adding?"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 365,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Score: 0.991110951772\n",
      "MSE: 102.352151481\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYsAAAEKCAYAAADjDHn2AAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\npGQhIiIlKVmIiEhJShYiIlLS/wf/L+gpZapRqAAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x11f83aac8>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.scatter(stock.target, stock.forecast)\n",
    "plt.xlabel(\"True Values\")\n",
    "plt.ylabel(\"Predictions\")\n",
    "print(\"Score:\", metrics.r2_score(stock.target, stock.forecast))\n",
    "print(\"MSE:\", metrics.mean_squared_error(stock.target, stock.forecast))\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Hmmmm. I'm guessing my one outlier is just the last observation since it is probably a zero (using .shift() and there being nothing there).  So my SARIMAX model is accurate, but I sort of knew that already."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 366,
   "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>target</th>\n",
       "      <th>forecast</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>date</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>2017-11-22</th>\n",
       "      <td>0.00</td>\n",
       "      <td>317.326469</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2010-07-06</th>\n",
       "      <td>15.80</td>\n",
       "      <td>19.199988</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2010-07-02</th>\n",
       "      <td>16.11</td>\n",
       "      <td>21.959973</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2010-07-09</th>\n",
       "      <td>17.05</td>\n",
       "      <td>17.460010</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2010-07-08</th>\n",
       "      <td>17.40</td>\n",
       "      <td>15.800004</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "            target    forecast\n",
       "date                          \n",
       "2017-11-22    0.00  317.326469\n",
       "2010-07-06   15.80   19.199988\n",
       "2010-07-02   16.11   21.959973\n",
       "2010-07-09   17.05   17.460010\n",
       "2010-07-08   17.40   15.800004"
      ]
     },
     "execution_count": 366,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "stock[['target', 'forecast']].sort_values('target').head()\n",
    "# YUP, last entry"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 367,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Score: 0.995793097015\n",
      "MSE: 48.414491673\n"
     ]
    }
   ],
   "source": [
    "print(\"Score:\", metrics.r2_score(stock.iloc[:-1].target, stock.iloc[:-1].forecast))\n",
    "print(\"MSE:\", metrics.mean_squared_error(stock.iloc[:-1].target, stock.iloc[:-1].forecast))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "And if I predicted the entirety with XGB:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 368,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "MSE: 650.015530157\n"
     ]
    }
   ],
   "source": [
    "print(\"MSE:\", metrics.mean_squared_error(stock.target, xgb.predict(stock.drop(['target', 'close', 'open'], 1))))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Let's look at change"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 69,
   "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>close</th>\n",
       "      <th>volume</th>\n",
       "      <th>open</th>\n",
       "      <th>high</th>\n",
       "      <th>low</th>\n",
       "      <th>forecast</th>\n",
       "      <th>next_day_open</th>\n",
       "      <th>target</th>\n",
       "      <th>000 _news_</th>\n",
       "      <th>10 _news_</th>\n",
       "      <th>...</th>\n",
       "      <th>documentcloud.org</th>\n",
       "      <th>everythingiCafe</th>\n",
       "      <th>iDownloadBlog.com</th>\n",
       "      <th>iDownloadBlog.com iDownloadBlog.com</th>\n",
       "      <th>iLounge</th>\n",
       "      <th>iMore</th>\n",
       "      <th>iPhone Hacks</th>\n",
       "      <th>iPhone in Canada Blog</th>\n",
       "      <th>lists.swift.org Mailing Lists</th>\n",
       "      <th>predict_grow</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>date</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>2010-06-29</th>\n",
       "      <td>23.89</td>\n",
       "      <td>18751150.0</td>\n",
       "      <td>19.00</td>\n",
       "      <td>25.0000</td>\n",
       "      <td>17.54</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>25.79</td>\n",
       "      <td>23.83</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2010-06-30</th>\n",
       "      <td>23.83</td>\n",
       "      <td>17165210.0</td>\n",
       "      <td>25.79</td>\n",
       "      <td>30.4192</td>\n",
       "      <td>23.30</td>\n",
       "      <td>23.890019</td>\n",
       "      <td>25.00</td>\n",
       "      <td>21.96</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2010-07-01</th>\n",
       "      <td>21.96</td>\n",
       "      <td>8216789.0</td>\n",
       "      <td>25.00</td>\n",
       "      <td>25.9200</td>\n",
       "      <td>20.27</td>\n",
       "      <td>23.830008</td>\n",
       "      <td>23.00</td>\n",
       "      <td>19.20</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2010-07-02</th>\n",
       "      <td>19.20</td>\n",
       "      <td>5135795.0</td>\n",
       "      <td>23.00</td>\n",
       "      <td>23.1000</td>\n",
       "      <td>18.71</td>\n",
       "      <td>21.959973</td>\n",
       "      <td>20.00</td>\n",
       "      <td>16.11</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2010-07-06</th>\n",
       "      <td>16.11</td>\n",
       "      <td>6858092.0</td>\n",
       "      <td>20.00</td>\n",
       "      <td>20.0000</td>\n",
       "      <td>15.83</td>\n",
       "      <td>19.199988</td>\n",
       "      <td>16.40</td>\n",
       "      <td>15.80</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 3307 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "            close      volume   open     high    low   forecast  \\\n",
       "date                                                              \n",
       "2010-06-29  23.89  18751150.0  19.00  25.0000  17.54   0.000000   \n",
       "2010-06-30  23.83  17165210.0  25.79  30.4192  23.30  23.890019   \n",
       "2010-07-01  21.96   8216789.0  25.00  25.9200  20.27  23.830008   \n",
       "2010-07-02  19.20   5135795.0  23.00  23.1000  18.71  21.959973   \n",
       "2010-07-06  16.11   6858092.0  20.00  20.0000  15.83  19.199988   \n",
       "\n",
       "            next_day_open  target  000 _news_  10 _news_      ...       \\\n",
       "date                                                          ...        \n",
       "2010-06-29          25.79   23.83         0.0        0.0      ...        \n",
       "2010-06-30          25.00   21.96         0.0        0.0      ...        \n",
       "2010-07-01          23.00   19.20         0.0        0.0      ...        \n",
       "2010-07-02          20.00   16.11         0.0        0.0      ...        \n",
       "2010-07-06          16.40   15.80         0.0        0.0      ...        \n",
       "\n",
       "            documentcloud.org  everythingiCafe  iDownloadBlog.com  \\\n",
       "date                                                                \n",
       "2010-06-29                  0                0                  0   \n",
       "2010-06-30                  0                0                  0   \n",
       "2010-07-01                  0                0                  0   \n",
       "2010-07-02                  0                0                  0   \n",
       "2010-07-06                  0                0                  0   \n",
       "\n",
       "            iDownloadBlog.com iDownloadBlog.com  iLounge  iMore  iPhone Hacks  \\\n",
       "date                                                                            \n",
       "2010-06-29                                    0        0      0             0   \n",
       "2010-06-30                                    0        0      0             0   \n",
       "2010-07-01                                    0        0      0             0   \n",
       "2010-07-02                                    0        0      0             0   \n",
       "2010-07-06                                    0        0      0             0   \n",
       "\n",
       "            iPhone in Canada Blog  lists.swift.org Mailing Lists  predict_grow  \n",
       "date                                                                            \n",
       "2010-06-29                      0                              0             0  \n",
       "2010-06-30                      0                              0             0  \n",
       "2010-07-01                      0                              0             1  \n",
       "2010-07-02                      0                              0             1  \n",
       "2010-07-06                      0                              0             1  \n",
       "\n",
       "[5 rows x 3307 columns]"
      ]
     },
     "execution_count": 69,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "stock.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 369,
   "metadata": {},
   "outputs": [],
   "source": [
    "stock['predict_grow'] = stock[['forecast', 'next_day_open']].apply(lambda x: 1 if x[0] - x[1] >= 0 else 0, axis=1)\n",
    "stock['true_grow'] = stock[['target', 'next_day_open']].apply(lambda x: 1 if x[0] - x[1] >= 0 else 0, axis=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 370,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "True Negative:  514 | False Positive:  443\n",
      "False Negative:  472 | True Positive:  437 \n",
      "\n",
      "Sensitivity (TP/ TP + FN):  0.480748074807\n",
      "Specificity (TN/ TN + FP):  0.537095088819 \n",
      "\n",
      "Accuracy:  0.509646302251\n",
      "Precision:  0.496590909091\n",
      "\n",
      "\n"
     ]
    }
   ],
   "source": [
    "eval_sklearn_model(stock.true_grow, stock.predict_grow)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Shiiiiiiiiiit\n",
    "\n",
    "Well that's the SARIMAX model on it's own... we can pray..."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 73,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Index(['close', 'volume', 'open', 'high', 'low', 'forecast', 'next_day_open',\n",
       "       'target', '000 _news_', '10 _news_',\n",
       "       ...\n",
       "       'everythingiCafe', 'iDownloadBlog.com',\n",
       "       'iDownloadBlog.com iDownloadBlog.com', 'iLounge', 'iMore',\n",
       "       'iPhone Hacks', 'iPhone in Canada Blog',\n",
       "       'lists.swift.org Mailing Lists', 'predict_grow', 'true_grow'],\n",
       "      dtype='object', length=3308)"
      ]
     },
     "execution_count": 73,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "stock.columns"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 371,
   "metadata": {},
   "outputs": [],
   "source": [
    "X = stock.drop(['target', 'true_grow', 'predict_grow'], 1)\n",
    "y = stock.true_grow"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 372,
   "metadata": {},
   "outputs": [],
   "source": [
    "X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.3, random_state=42)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 373,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0    957\n",
       "1    909\n",
       "Name: true_grow, dtype: int64"
      ]
     },
     "execution_count": 373,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "y.value_counts()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 81,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "XGBClassifier(base_score=0.5, colsample_bylevel=1, colsample_bytree=1,\n",
       "       gamma=0, learning_rate=0.1, max_delta_step=0, max_depth=3,\n",
       "       min_child_weight=1, missing=None, n_estimators=100, nthread=-1,\n",
       "       objective='binary:logistic', reg_alpha=0, reg_lambda=1,\n",
       "       scale_pos_weight=1, seed=0, silent=True, subsample=1)"
      ]
     },
     "execution_count": 81,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "xgb = XGBClassifier()\n",
    "xgb.fit(X_train, y_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 82,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "True Negative:  98 | False Positive:  196\n",
      "False Negative:  90 | True Positive:  176 \n",
      "\n",
      "Sensitivity (TP/ TP + FN):  0.661654135338\n",
      "Specificity (TN/ TN + FP):  0.333333333333 \n",
      "\n",
      "Accuracy:  0.489285714286\n",
      "Precision:  0.47311827957\n",
      "Roc-Auc:  0.494878778579\n",
      "\n",
      "\n"
     ]
    }
   ],
   "source": [
    "test_predictions = xgb.predict(X_test)\n",
    "eval_sklearn_model(y_test, test_predictions, model=xgb, X=X_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 84,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fitting 3 folds for each of 60 candidates, totalling 180 fits\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "[Parallel(n_jobs=1)]: Done 180 out of 180 | elapsed: 56.4min finished\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Decision Tree Classifier Score: 0.541 \n",
      "\n",
      "Elapsed Time: 3.41e+03  seconds \n",
      "\n",
      "XGBClassifier(base_score=0.5, colsample_bylevel=1, colsample_bytree=1,\n",
      "       gamma=0.1, learning_rate=0.1, max_delta_step=0, max_depth=5,\n",
      "       min_child_weight=3, missing=None, n_estimators=100, nthread=-1,\n",
      "       objective='binary:logistic', reg_alpha=0, reg_lambda=1,\n",
      "       scale_pos_weight=1, seed=0, silent=True, subsample=1) \n",
      "\n",
      "Best Hyperparameters we tested for \n",
      " {'params': [('gamma', 0.1), ('max_depth', 5), ('min_child_weight', 3)], 'score': 0.49469627234891383}\n"
     ]
    }
   ],
   "source": [
    "start_time = t.time()\n",
    "\n",
    "param_grid = {'max_depth':range(3,10,2),\n",
    "              'min_child_weight':range(1,6,2),\n",
    "              'gamma':[i/10.0 for i in range(0,5)]\n",
    "             }\n",
    "\n",
    "dt = XGBClassifier()\n",
    "\n",
    "grid = GridSearchCV(dt, param_grid, cv=3, scoring='roc_auc', verbose=1)\n",
    "\n",
    "grid.fit(X_train, y_train)\n",
    "\n",
    "best_results = {'params': list(grid.best_params_.items()), 'score': grid.best_score_}\n",
    "\n",
    "best_dt = grid.best_estimator_\n",
    "\n",
    "best_dt = best_dt.fit(X_train, y_train)\n",
    "\n",
    "score = best_dt.score(X_test, y_test)\n",
    "\n",
    "#print('Number of Models Run: ', len(n_estimators) * len(max_features) * len(min_samples_leaf) * 3)\n",
    "print(\"{} Score: {:0.3}\".format('Decision Tree Classifier', score.mean().round(3)), '\\n')\n",
    "print('Elapsed Time: {:0.3}'.format( t.time() - start_time), ' seconds', '\\n')\n",
    "print(grid.best_estimator_, '\\n')\n",
    "print('Best Hyperparameters we tested for', '\\n', best_results)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 85,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "True Negative:  148 | False Positive:  146\n",
      "False Negative:  111 | True Positive:  155 \n",
      "\n",
      "Sensitivity (TP/ TP + FN):  0.582706766917\n",
      "Specificity (TN/ TN + FP):  0.503401360544 \n",
      "\n",
      "Accuracy:  0.541071428571\n",
      "Precision:  0.514950166113\n",
      "Roc-Auc:  0.545515574651\n",
      "\n",
      "\n"
     ]
    }
   ],
   "source": [
    "test_predictions = best_dt.predict(X_test)\n",
    "eval_sklearn_model(y_test, test_predictions, model=best_dt, X=X_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 87,
   "metadata": {},
   "outputs": [],
   "source": [
    "import pickle"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 89,
   "metadata": {},
   "outputs": [],
   "source": [
    "pickle.dump(best_dt, open('xgb_model.pkl', 'wb'))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## So how did we do?\n",
    "\n",
    "```stock.iloc[1640]``` is 2017-01-03, the day I bought Tesla stocks"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 131,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "102.01999999999998"
      ]
     },
     "execution_count": 131,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Buy and hold strategy\n",
    "stock.iloc[-2]['next_day_open'] - stock.iloc[1640]['next_day_open']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 382,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Own? False\n",
      "Last Buying price: $ 310.86\n",
      "Current price: $ 316.77\n",
      "Cash? $ 54.59\n",
      "Profit: $ 54.59\n",
      "Number of Transactins: 102\n",
      "Cost of transactions: $ 510\n"
     ]
    }
   ],
   "source": [
    "# SARIMAX Model predictions\n",
    "profit_calculator(stock.iloc[1640:], 'predict_grow', -1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 374,
   "metadata": {},
   "outputs": [],
   "source": [
    "best_dt = pickle.load(open('xgb_model.pkl', 'rb'))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "stock['model_predict'] = best_dt.predict(stock.drop(['target', 'true_grow', 'predict_grow'], 1))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 383,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Own? False\n",
      "Last Buying price: $ 310.86\n",
      "Current price: $ 316.77\n",
      "Cash? $ 314.105\n",
      "Profit: $ 314.105\n",
      "Number of Transactins: 86\n",
      "Cost of transactions: $ 430\n"
     ]
    }
   ],
   "source": [
    "profit_calculator(stock.iloc[1640:], 'model_predict',-1)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "LIES!  This is has to be because I overfit and did not properly train/test split.  My model knows future news."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 384,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Own? False\n",
      "Last Buying price: $ 310.86\n",
      "Current price: $ 316.77\n",
      "Cash? $ 620.08\n",
      "Profit: $ 620.08\n",
      "Number of Transactins: 118\n",
      "Cost of transactions: $ 590\n"
     ]
    }
   ],
   "source": [
    "# True Possible returns\n",
    "profit_calculator(stock.iloc[1640:], 'true_grow',-1)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "My model did perform positively, so that's something over the entire time period.  I did a different sort of train/test split which unfortunately won't work in the real world.  Currently it is buying even if change is 0 or up and selling if it is below 0.  With no transaction cost, this might make sense."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Rebuild the Model from 2014 with proper train/test split\n",
    "\n",
    "### Note on earliest NLP relevant info\n",
    "1. elon earliest tweet 2013-03-23\n",
    "2. tesla earliest tweet 2014-02-18\n",
    "3. techmeme earliest story 2006-12-07\n",
    "\n",
    "```stock.iloc[915]``` is 2014-02-18"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [],
   "source": [
    "stock = pd.read_csv('TESLAHistoricalQuotes.csv')\n",
    "stock = stock.drop(0,0)\n",
    "stock.date = pd.to_datetime(stock.date)\n",
    "stock.date = stock.date.apply(lambda x: x.date())\n",
    "stock = stock.sort_values('date')\n",
    "stock.volume = stock.volume.astype(float)\n",
    "stock.set_index(stock.date, inplace=True)\n",
    "stock = stock.drop('date', 1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [],
   "source": [
    "stock = pd.merge(stock, techmeme, how='left', left_index=True, right_index=True)\n",
    "stock = pd.merge(stock, tesla, how='left', left_index=True, right_index=True)\n",
    "stock = pd.merge(stock, elon, how='left', left_index=True, right_index=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [],
   "source": [
    "# News sources binarize\n",
    "mlb = MultiLabelBinarizer()\n",
    "\n",
    "stock.news_sources = stock.news_sources.fillna('')\n",
    "stock = stock.join(pd.DataFrame(mlb.fit_transform(stock.news_sources), columns=mlb.classes_, index=stock.index))\n",
    "stock = stock.drop('news_sources', 1)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### SARIMAX\n",
    "\n",
    "The time series model requires yesterday's data to make a prediction.  We will always have that information, so the model will essentially need to be retrained daily to incorporate it.  (There may be a way to feed the existing model new information, but I am not sure.). It therefore makes sense to create the predictions before the train test split as each prediction should only be dependent on data from prior days."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [],
   "source": [
    "model_df = stock.iloc[915:].copy()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/anaconda/lib/python3.6/site-packages/statsmodels/base/model.py:496: ConvergenceWarning: Maximum Likelihood optimization failed to converge. Check mle_retvals\n",
      "  \"Check mle_retvals\", ConvergenceWarning)\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<table class=\"simpletable\">\n",
       "<caption>Statespace Model Results</caption>\n",
       "<tr>\n",
       "  <th>Dep. Variable:</th>                <td>close</td>             <th>  No. Observations:  </th>    <td>951</td>   \n",
       "</tr>\n",
       "<tr>\n",
       "  <th>Model:</th>           <td>SARIMAX(1, 1, 3)x(0, 1, 1, 80)</td> <th>  Log Likelihood     </th> <td>-2907.625</td>\n",
       "</tr>\n",
       "<tr>\n",
       "  <th>Date:</th>                   <td>Thu, 21 Dec 2017</td>        <th>  AIC                </th> <td>5827.249</td> \n",
       "</tr>\n",
       "<tr>\n",
       "  <th>Time:</th>                       <td>10:40:54</td>            <th>  BIC                </th> <td>5856.394</td> \n",
       "</tr>\n",
       "<tr>\n",
       "  <th>Sample:</th>                    <td>02-18-2014</td>           <th>  HQIC               </th> <td>5838.353</td> \n",
       "</tr>\n",
       "<tr>\n",
       "  <th></th>                          <td>- 11-22-2017</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>ar.L1</th>    <td>    0.4877</td> <td>    0.792</td> <td>    0.616</td> <td> 0.538</td> <td>   -1.064</td> <td>    2.040</td>\n",
       "</tr>\n",
       "<tr>\n",
       "  <th>ma.L1</th>    <td>   -0.4423</td> <td>    0.794</td> <td>   -0.557</td> <td> 0.577</td> <td>   -1.998</td> <td>    1.113</td>\n",
       "</tr>\n",
       "<tr>\n",
       "  <th>ma.L2</th>    <td>    0.0137</td> <td>    0.055</td> <td>    0.251</td> <td> 0.802</td> <td>   -0.093</td> <td>    0.121</td>\n",
       "</tr>\n",
       "<tr>\n",
       "  <th>ma.L3</th>    <td>   -0.0490</td> <td>    0.035</td> <td>   -1.386</td> <td> 0.166</td> <td>   -0.118</td> <td>    0.020</td>\n",
       "</tr>\n",
       "<tr>\n",
       "  <th>ma.S.L80</th> <td>   -0.9991</td> <td>    6.688</td> <td>   -0.149</td> <td> 0.881</td> <td>  -14.107</td> <td>   12.109</td>\n",
       "</tr>\n",
       "<tr>\n",
       "  <th>sigma2</th>   <td>   37.3230</td> <td>  248.806</td> <td>    0.150</td> <td> 0.881</td> <td> -450.328</td> <td>  524.974</td>\n",
       "</tr>\n",
       "</table>\n",
       "<table class=\"simpletable\">\n",
       "<tr>\n",
       "  <th>Ljung-Box (Q):</th>          <td>22.88</td> <th>  Jarque-Bera (JB):  </th> <td>109.70</td>\n",
       "</tr>\n",
       "<tr>\n",
       "  <th>Prob(Q):</th>                <td>0.99</td>  <th>  Prob(JB):          </th>  <td>0.00</td> \n",
       "</tr>\n",
       "<tr>\n",
       "  <th>Heteroskedasticity (H):</th> <td>1.06</td>  <th>  Skew:              </th>  <td>-0.23</td>\n",
       "</tr>\n",
       "<tr>\n",
       "  <th>Prob(H) (two-sided):</th>    <td>0.65</td>  <th>  Kurtosis:          </th>  <td>4.68</td> \n",
       "</tr>\n",
       "</table>"
      ],
      "text/plain": [
       "<class 'statsmodels.iolib.summary.Summary'>\n",
       "\"\"\"\n",
       "                                 Statespace Model Results                                 \n",
       "==========================================================================================\n",
       "Dep. Variable:                              close   No. Observations:                  951\n",
       "Model:             SARIMAX(1, 1, 3)x(0, 1, 1, 80)   Log Likelihood               -2907.625\n",
       "Date:                            Thu, 21 Dec 2017   AIC                           5827.249\n",
       "Time:                                    10:40:54   BIC                           5856.394\n",
       "Sample:                                02-18-2014   HQIC                          5838.353\n",
       "                                     - 11-22-2017                                         \n",
       "Covariance Type:                              opg                                         \n",
       "==============================================================================\n",
       "                 coef    std err          z      P>|z|      [0.025      0.975]\n",
       "------------------------------------------------------------------------------\n",
       "ar.L1          0.4877      0.792      0.616      0.538      -1.064       2.040\n",
       "ma.L1         -0.4423      0.794     -0.557      0.577      -1.998       1.113\n",
       "ma.L2          0.0137      0.055      0.251      0.802      -0.093       0.121\n",
       "ma.L3         -0.0490      0.035     -1.386      0.166      -0.118       0.020\n",
       "ma.S.L80      -0.9991      6.688     -0.149      0.881     -14.107      12.109\n",
       "sigma2        37.3230    248.806      0.150      0.881    -450.328     524.974\n",
       "===================================================================================\n",
       "Ljung-Box (Q):                       22.88   Jarque-Bera (JB):               109.70\n",
       "Prob(Q):                              0.99   Prob(JB):                         0.00\n",
       "Heteroskedasticity (H):               1.06   Skew:                            -0.23\n",
       "Prob(H) (two-sided):                  0.65   Kurtosis:                         4.68\n",
       "===================================================================================\n",
       "\n",
       "Warnings:\n",
       "[1] Covariance matrix calculated using the outer product of gradients (complex-step).\n",
       "\"\"\""
      ]
     },
     "execution_count": 13,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "mod = sm.tsa.statespace.SARIMAX(model_df.close, trend='n', order=(1,1,3), seasonal_order=(0,1,1,80))\n",
    "results = mod.fit()\n",
    "results.summary()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [],
   "source": [
    "model_df['forecast'] = results.predict(dynamic=False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [],
   "source": [
    "model_df['target'] = model_df.close.shift(-1)\n",
    "model_df['next_day_open'] = model_df.open.shift(-1) \n",
    "model_df['true_grow'] = model_df[['target', 'next_day_open']].apply(lambda x: 1 if x[0] - x[1] >= 0 else 0, axis=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [],
   "source": [
    "model_df.news_text = model_df.news_text.fillna('')\n",
    "model_df.tesla_tweet = model_df.tesla_tweet.fillna('')\n",
    "model_df.elon_tweet = model_df.elon_tweet.fillna('')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Train/test split\n",
    "\n",
    "Split on ```model_df.iloc[725]``` as it is 2017-01-03"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Drop day one as no forecast can be made and last day as there is no \"next_day_open\"\n",
    "train_df = model_df.iloc[1:725]\n",
    "test_df = model_df.iloc[725:-1]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [],
   "source": [
    "cvec = CountVectorizer(stop_words=stopwords.words('english'), ngram_range=(1,2), max_features=2000)\n",
    "\n",
    "# Fit news\n",
    "cvec.fit(train_df.news_text)\n",
    "\n",
    "# Transform train\n",
    "cvec_df  = pd.DataFrame(cvec.transform(train_df.news_text.dropna()).todense(), columns=cvec.get_feature_names())\n",
    "cvec_df.index = train_df.index\n",
    "cvec_df.columns += \" _news_\"\n",
    "train_df = train_df.join(cvec_df).drop('news_text',1)\n",
    "\n",
    "# Transform test\n",
    "cvec_df  = pd.DataFrame(cvec.transform(test_df.news_text.dropna()).todense(), columns=cvec.get_feature_names())\n",
    "cvec_df.index = test_df.index\n",
    "cvec_df.columns += \" _news_\"\n",
    "test_df = test_df.join(cvec_df).drop('news_text',1)\n",
    "\n",
    "# Fit tesla tweet\n",
    "cvec.fit(train_df.tesla_tweet)\n",
    "\n",
    "# Transform train\n",
    "cvec_df  = pd.DataFrame(cvec.transform(train_df.tesla_tweet.dropna()).todense(), columns=cvec.get_feature_names())\n",
    "cvec_df.index = train_df.index\n",
    "cvec_df.columns += \" _tesla_\"\n",
    "train_df = train_df.join(cvec_df).drop('tesla_tweet',1)\n",
    "\n",
    "# Transform test\n",
    "cvec_df  = pd.DataFrame(cvec.transform(test_df.tesla_tweet.dropna()).todense(), columns=cvec.get_feature_names())\n",
    "cvec_df.index = test_df.index\n",
    "cvec_df.columns += \" _tesla_\"\n",
    "test_df = test_df.join(cvec_df).drop('tesla_tweet',1)\n",
    "\n",
    "# Fit elon tweets\n",
    "cvec.fit(train_df.elon_tweet)\n",
    "\n",
    "# Transform train\n",
    "cvec_df  = pd.DataFrame(cvec.transform(train_df.elon_tweet.dropna()).todense(), columns=cvec.get_feature_names())\n",
    "cvec_df.index = train_df.index\n",
    "cvec_df.columns += \" _elon_\"\n",
    "train_df = train_df.join(cvec_df).drop('elon_tweet',1)\n",
    "\n",
    "# Transform test\n",
    "cvec_df  = pd.DataFrame(cvec.transform(test_df.elon_tweet.dropna()).todense(), columns=cvec.get_feature_names())\n",
    "cvec_df.index = test_df.index\n",
    "cvec_df.columns += \" _elon_\"\n",
    "test_df = test_df.join(cvec_df).drop('elon_tweet',1)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Alright, here we go!"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "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>close</th>\n",
       "      <th>volume</th>\n",
       "      <th>open</th>\n",
       "      <th>high</th>\n",
       "      <th>low</th>\n",
       "      <th>9to5Google</th>\n",
       "      <th>9to5Mac</th>\n",
       "      <th>@a_riley17</th>\n",
       "      <th>@aatilley</th>\n",
       "      <th>@adavies47</th>\n",
       "      <th>...</th>\n",
       "      <th>yet _elon_</th>\n",
       "      <th>yo _elon_</th>\n",
       "      <th>york _elon_</th>\n",
       "      <th>york http _elon_</th>\n",
       "      <th>youtube _elon_</th>\n",
       "      <th>yup _elon_</th>\n",
       "      <th>yup rt _elon_</th>\n",
       "      <th>zero _elon_</th>\n",
       "      <th>zqhjhtkuaj _elon_</th>\n",
       "      <th>zqhjhtkuaj http _elon_</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>date</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>2014-02-19</th>\n",
       "      <td>193.64</td>\n",
       "      <td>16038700.0</td>\n",
       "      <td>203.70</td>\n",
       "      <td>203.70</td>\n",
       "      <td>193.41</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2014-02-20</th>\n",
       "      <td>209.97</td>\n",
       "      <td>18000940.0</td>\n",
       "      <td>215.01</td>\n",
       "      <td>215.21</td>\n",
       "      <td>206.27</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2014-02-21</th>\n",
       "      <td>209.60</td>\n",
       "      <td>7817352.0</td>\n",
       "      <td>211.64</td>\n",
       "      <td>213.98</td>\n",
       "      <td>209.19</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2014-02-24</th>\n",
       "      <td>217.65</td>\n",
       "      <td>8286273.0</td>\n",
       "      <td>208.76</td>\n",
       "      <td>218.36</td>\n",
       "      <td>208.32</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2014-02-25</th>\n",
       "      <td>248.00</td>\n",
       "      <td>32648710.0</td>\n",
       "      <td>230.00</td>\n",
       "      <td>259.20</td>\n",
       "      <td>228.45</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 6307 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "             close      volume    open    high     low  9to5Google  9to5Mac  \\\n",
       "date                                                                          \n",
       "2014-02-19  193.64  16038700.0  203.70  203.70  193.41           0        1   \n",
       "2014-02-20  209.97  18000940.0  215.01  215.21  206.27           0        0   \n",
       "2014-02-21  209.60   7817352.0  211.64  213.98  209.19           0        0   \n",
       "2014-02-24  217.65   8286273.0  208.76  218.36  208.32           0        0   \n",
       "2014-02-25  248.00  32648710.0  230.00  259.20  228.45           0        0   \n",
       "\n",
       "            @a_riley17  @aatilley  @adavies47           ...            \\\n",
       "date                                                    ...             \n",
       "2014-02-19           0          0           0           ...             \n",
       "2014-02-20           0          0           0           ...             \n",
       "2014-02-21           0          0           0           ...             \n",
       "2014-02-24           0          0           0           ...             \n",
       "2014-02-25           0          0           0           ...             \n",
       "\n",
       "            yet _elon_  yo _elon_  york _elon_  york http _elon_  \\\n",
       "date                                                               \n",
       "2014-02-19           0          0            0                 0   \n",
       "2014-02-20           0          0            0                 0   \n",
       "2014-02-21           0          0            0                 0   \n",
       "2014-02-24           0          0            0                 0   \n",
       "2014-02-25           0          0            0                 0   \n",
       "\n",
       "            youtube _elon_  yup _elon_  yup rt _elon_  zero _elon_  \\\n",
       "date                                                                 \n",
       "2014-02-19               0           0              0            0   \n",
       "2014-02-20               0           0              0            0   \n",
       "2014-02-21               0           0              0            0   \n",
       "2014-02-24               0           0              0            0   \n",
       "2014-02-25               0           0              0            0   \n",
       "\n",
       "            zqhjhtkuaj _elon_  zqhjhtkuaj http _elon_  \n",
       "date                                                   \n",
       "2014-02-19                  0                       0  \n",
       "2014-02-20                  0                       0  \n",
       "2014-02-21                  0                       0  \n",
       "2014-02-24                  0                       0  \n",
       "2014-02-25                  0                       0  \n",
       "\n",
       "[5 rows x 6307 columns]"
      ]
     },
     "execution_count": 19,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "train_df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [],
   "source": [
    "X_train = train_df.drop(['target', 'true_grow'], 1)\n",
    "y_train = train_df.true_grow"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [],
   "source": [
    "X_test = test_df.drop(['target', 'true_grow'], 1)\n",
    "y_test = test_df.true_grow"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [],
   "source": [
    "xgb = XGBClassifier(base_score=0.5, colsample_bylevel=1, colsample_bytree=1,\n",
    "       gamma=0.1, learning_rate=0.1, max_delta_step=0, max_depth=5,\n",
    "       min_child_weight=3, missing=None, n_estimators=100, nthread=-1,\n",
    "       objective='binary:logistic', reg_alpha=0, reg_lambda=1,\n",
    "       scale_pos_weight=1, seed=0, silent=True, subsample=1) "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {
    "scrolled": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "True Negative:  102 | False Positive:  14\n",
      "False Negative:  82 | True Positive:  27 \n",
      "\n",
      "Sensitivity (TP/ TP + FN):  0.247706422018\n",
      "Specificity (TN/ TN + FP):  0.879310344828 \n",
      "\n",
      "Accuracy:  0.573333333333\n",
      "Precision:  0.658536585366\n",
      "Roc-Auc:  0.577942106928\n",
      "\n",
      "\n"
     ]
    }
   ],
   "source": [
    "xgb.fit(X_train, y_train)\n",
    "test_predictions = xgb.predict(X_test)\n",
    "eval_sklearn_model(y_test, test_predictions, model=xgb, X=X_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [],
   "source": [
    "test_df['model_predict'] = test_predictions"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Own? False\n",
      "Last Buying price: $ 299.5\n",
      "Current price: $ 316.77\n",
      "Cash? $ 63.9\n",
      "Profit: $ 63.9\n",
      "Number of Transactins: 54\n",
      "Cost of transactions: $ 270\n"
     ]
    }
   ],
   "source": [
    "profit_calculator(test_df, 'model_predict')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Own? False\n",
      "Last Buying price: $ 310.86\n",
      "Current price: $ 316.77\n",
      "Cash? $ 620.08\n",
      "Profit: $ 620.08\n",
      "Number of Transactins: 118\n",
      "Cost of transactions: $ 590\n"
     ]
    }
   ],
   "source": [
    "profit_calculator(test_df, 'true_grow')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAuQAAAIVCAYAAACZaurZAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nRU5ErkJggg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x123ddd748>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "X, y = plot_profit(test_df, 'model_predict', early_stop=None)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fitting 3 folds for each of 180 candidates, totalling 540 fits\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "[Parallel(n_jobs=1)]: Done 540 out of 540 | elapsed: 686.3min finished\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Decision Tree Classifier Score: 0.547 \n",
      "\n",
      "Elapsed Time: 4.12e+04  seconds \n",
      "\n",
      "XGBClassifier(base_score=0.5, colsample_bylevel=1, colsample_bytree=1,\n",
      "       gamma=0.4, learning_rate=0.1, max_delta_step=0, max_depth=3,\n",
      "       min_child_weight=1, missing=None, n_estimators=100, nthread=-1,\n",
      "       objective='binary:logistic', reg_alpha=0, reg_lambda=1,\n",
      "       scale_pos_weight=1, seed=0, silent=True, subsample=1) \n",
      "\n",
      "Best Hyperparameters we tested for \n",
      " {'params': [('gamma', 0.4), ('max_depth', 3), ('min_child_weight', 1), ('n_estimators', 100)], 'score': 0.51886353139604169}\n"
     ]
    }
   ],
   "source": [
    "# gridsearch xgboost\n",
    "start_time = t.time()\n",
    "\n",
    "param_grid = {'max_depth':range(3,10,2),\n",
    "              'min_child_weight':range(1,6,2),\n",
    "              'gamma':[i/10.0 for i in range(0,5)],\n",
    "              'n_estimators':[100,500,1000]\n",
    "             }\n",
    "\n",
    "dt = XGBClassifier()\n",
    "\n",
    "grid = GridSearchCV(dt, param_grid, cv=3, scoring='roc_auc', verbose=1)\n",
    "\n",
    "grid.fit(X_train, y_train)\n",
    "\n",
    "best_results = {'params': list(grid.best_params_.items()), 'score': grid.best_score_}\n",
    "\n",
    "best_dt = grid.best_estimator_\n",
    "\n",
    "best_dt = best_dt.fit(X_train, y_train)\n",
    "\n",
    "score = best_dt.score(X_test, y_test)\n",
    "\n",
    "#print('Number of Models Run: ', len(n_estimators) * len(max_features) * len(min_samples_leaf) * 3)\n",
    "print(\"{} Score: {:0.3}\".format('Decision Tree Classifier', score.mean().round(3)), '\\n')\n",
    "print('Elapsed Time: {:0.3}'.format( t.time() - start_time), ' seconds', '\\n')\n",
    "print(grid.best_estimator_, '\\n')\n",
    "print('Best Hyperparameters we tested for', '\\n', best_results)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "XGBClassifier(base_score=0.5, colsample_bylevel=1, colsample_bytree=1,\n",
    "       gamma=0.4, learning_rate=0.1, max_delta_step=0, max_depth=3,\n",
    "       min_child_weight=1, missing=None, n_estimators=100, nthread=-1,\n",
    "       objective='binary:logistic', reg_alpha=0, reg_lambda=1,\n",
    "       scale_pos_weight=1, seed=0, silent=True, subsample=1) \n",
    "\n",
    "Best Hyperparameters we tested for \n",
    " {'params': [('gamma', 0.4), ('max_depth', 3), ('min_child_weight', 1), ('n_estimators', 100)], 'score': 0.51886353139604169}"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# OH CRAP\n",
    "\n",
    "I need to rebuild the model EVERY DAY!"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Rebuild the Model from 2014 with proper train/test split\n",
    "\n",
    "### Note on earliest NLP relevant info\n",
    "1. elon earliest tweet 2013-03-23\n",
    "2. tesla earliest tweet 2014-02-18\n",
    "3. techmeme earliest story 2006-12-07\n",
    "\n",
    "```stock.iloc[915]``` is 2014-02-18"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [],
   "source": [
    "stock = pd.read_csv('TESLAHistoricalQuotes.csv')\n",
    "stock = stock.drop(0,0)\n",
    "stock.date = pd.to_datetime(stock.date)\n",
    "stock.date = stock.date.apply(lambda x: x.date())\n",
    "stock = stock.sort_values('date')\n",
    "stock.volume = stock.volume.astype(float)\n",
    "stock.set_index(stock.date, inplace=True)\n",
    "stock = stock.drop('date', 1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [],
   "source": [
    "stock = pd.merge(stock, techmeme, how='left', left_index=True, right_index=True)\n",
    "stock = pd.merge(stock, tesla, how='left', left_index=True, right_index=True)\n",
    "stock = pd.merge(stock, elon, how='left', left_index=True, right_index=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [],
   "source": [
    "# News sources binarize\n",
    "mlb = MultiLabelBinarizer()\n",
    "\n",
    "stock.news_sources = stock.news_sources.fillna('')\n",
    "stock = stock.join(pd.DataFrame(mlb.fit_transform(stock.news_sources), columns=mlb.classes_, index=stock.index))\n",
    "stock = stock.drop('news_sources', 1)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### SARIMAX\n",
    "\n",
    "The time series model requires yesterday's data to make a prediction.  We will always have that information, so the model will essentially need to be retrained daily to incorporate it.  (There may be a way to feed the existing model new information, but I am not sure.). It therefore makes sense to create the predictions before the train test split as each prediction should only be dependent on data from prior days."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {},
   "outputs": [],
   "source": [
    "model_df = stock.iloc[915:].copy()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/anaconda/lib/python3.6/site-packages/statsmodels/base/model.py:496: ConvergenceWarning: Maximum Likelihood optimization failed to converge. Check mle_retvals\n",
      "  \"Check mle_retvals\", ConvergenceWarning)\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<table class=\"simpletable\">\n",
       "<caption>Statespace Model Results</caption>\n",
       "<tr>\n",
       "  <th>Dep. Variable:</th>                <td>close</td>             <th>  No. Observations:  </th>    <td>951</td>   \n",
       "</tr>\n",
       "<tr>\n",
       "  <th>Model:</th>           <td>SARIMAX(1, 1, 3)x(0, 1, 1, 80)</td> <th>  Log Likelihood     </th> <td>-2907.625</td>\n",
       "</tr>\n",
       "<tr>\n",
       "  <th>Date:</th>                   <td>Wed, 20 Dec 2017</td>        <th>  AIC                </th> <td>5827.249</td> \n",
       "</tr>\n",
       "<tr>\n",
       "  <th>Time:</th>                       <td>21:16:05</td>            <th>  BIC                </th> <td>5856.394</td> \n",
       "</tr>\n",
       "<tr>\n",
       "  <th>Sample:</th>                    <td>02-18-2014</td>           <th>  HQIC               </th> <td>5838.353</td> \n",
       "</tr>\n",
       "<tr>\n",
       "  <th></th>                          <td>- 11-22-2017</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>ar.L1</th>    <td>    0.4877</td> <td>    0.792</td> <td>    0.616</td> <td> 0.538</td> <td>   -1.064</td> <td>    2.040</td>\n",
       "</tr>\n",
       "<tr>\n",
       "  <th>ma.L1</th>    <td>   -0.4423</td> <td>    0.794</td> <td>   -0.557</td> <td> 0.577</td> <td>   -1.998</td> <td>    1.113</td>\n",
       "</tr>\n",
       "<tr>\n",
       "  <th>ma.L2</th>    <td>    0.0137</td> <td>    0.055</td> <td>    0.251</td> <td> 0.802</td> <td>   -0.093</td> <td>    0.121</td>\n",
       "</tr>\n",
       "<tr>\n",
       "  <th>ma.L3</th>    <td>   -0.0490</td> <td>    0.035</td> <td>   -1.386</td> <td> 0.166</td> <td>   -0.118</td> <td>    0.020</td>\n",
       "</tr>\n",
       "<tr>\n",
       "  <th>ma.S.L80</th> <td>   -0.9991</td> <td>    6.688</td> <td>   -0.149</td> <td> 0.881</td> <td>  -14.107</td> <td>   12.109</td>\n",
       "</tr>\n",
       "<tr>\n",
       "  <th>sigma2</th>   <td>   37.3230</td> <td>  248.806</td> <td>    0.150</td> <td> 0.881</td> <td> -450.328</td> <td>  524.974</td>\n",
       "</tr>\n",
       "</table>\n",
       "<table class=\"simpletable\">\n",
       "<tr>\n",
       "  <th>Ljung-Box (Q):</th>          <td>22.88</td> <th>  Jarque-Bera (JB):  </th> <td>109.70</td>\n",
       "</tr>\n",
       "<tr>\n",
       "  <th>Prob(Q):</th>                <td>0.99</td>  <th>  Prob(JB):          </th>  <td>0.00</td> \n",
       "</tr>\n",
       "<tr>\n",
       "  <th>Heteroskedasticity (H):</th> <td>1.06</td>  <th>  Skew:              </th>  <td>-0.23</td>\n",
       "</tr>\n",
       "<tr>\n",
       "  <th>Prob(H) (two-sided):</th>    <td>0.65</td>  <th>  Kurtosis:          </th>  <td>4.68</td> \n",
       "</tr>\n",
       "</table>"
      ],
      "text/plain": [
       "<class 'statsmodels.iolib.summary.Summary'>\n",
       "\"\"\"\n",
       "                                 Statespace Model Results                                 \n",
       "==========================================================================================\n",
       "Dep. Variable:                              close   No. Observations:                  951\n",
       "Model:             SARIMAX(1, 1, 3)x(0, 1, 1, 80)   Log Likelihood               -2907.625\n",
       "Date:                            Wed, 20 Dec 2017   AIC                           5827.249\n",
       "Time:                                    21:16:05   BIC                           5856.394\n",
       "Sample:                                02-18-2014   HQIC                          5838.353\n",
       "                                     - 11-22-2017                                         \n",
       "Covariance Type:                              opg                                         \n",
       "==============================================================================\n",
       "                 coef    std err          z      P>|z|      [0.025      0.975]\n",
       "------------------------------------------------------------------------------\n",
       "ar.L1          0.4877      0.792      0.616      0.538      -1.064       2.040\n",
       "ma.L1         -0.4423      0.794     -0.557      0.577      -1.998       1.113\n",
       "ma.L2          0.0137      0.055      0.251      0.802      -0.093       0.121\n",
       "ma.L3         -0.0490      0.035     -1.386      0.166      -0.118       0.020\n",
       "ma.S.L80      -0.9991      6.688     -0.149      0.881     -14.107      12.109\n",
       "sigma2        37.3230    248.806      0.150      0.881    -450.328     524.974\n",
       "===================================================================================\n",
       "Ljung-Box (Q):                       22.88   Jarque-Bera (JB):               109.70\n",
       "Prob(Q):                              0.99   Prob(JB):                         0.00\n",
       "Heteroskedasticity (H):               1.06   Skew:                            -0.23\n",
       "Prob(H) (two-sided):                  0.65   Kurtosis:                         4.68\n",
       "===================================================================================\n",
       "\n",
       "Warnings:\n",
       "[1] Covariance matrix calculated using the outer product of gradients (complex-step).\n",
       "\"\"\""
      ]
     },
     "execution_count": 26,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "mod = sm.tsa.statespace.SARIMAX(model_df.close, trend='n', order=(1,1,3), seasonal_order=(0,1,1,80))\n",
    "results = mod.fit()\n",
    "results.summary()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {},
   "outputs": [],
   "source": [
    "model_df['forecast'] = results.predict(dynamic=False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {},
   "outputs": [],
   "source": [
    "model_df['target'] = model_df.close.shift(-1)\n",
    "model_df['next_day_open'] = model_df.open.shift(-1) \n",
    "model_df['true_grow'] = model_df[['target', 'next_day_open']].apply(lambda x: 1 if x[0] - x[1] >= 0 else 0, axis=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "metadata": {},
   "outputs": [],
   "source": [
    "model_df.news_text = model_df.news_text.fillna('')\n",
    "model_df.tesla_tweet = model_df.tesla_tweet.fillna('')\n",
    "model_df.elon_tweet = model_df.elon_tweet.fillna('')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## WE REBUILD HERE"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 103,
   "metadata": {},
   "outputs": [],
   "source": [
    "# We update this each time\n",
    "real_test_df = model_df.iloc[725:-1].copy()\n",
    "real_test_df['model_predict'] = np.NaN"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Train/test split\n",
    "\n",
    "Split on ```model_df.iloc[725]``` as it is 2017-01-03"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 93,
   "metadata": {},
   "outputs": [],
   "source": [
    "def count_vect_our_dfs(training_dataframe, testing_dataframe):\n",
    "    cvec = CountVectorizer(stop_words=stopwords.words('english'), ngram_range=(1,2), max_features=2000)\n",
    "\n",
    "    # Fit news\n",
    "    cvec.fit(training_dataframe.news_text)\n",
    "\n",
    "    # Transform train\n",
    "    cvec_df  = pd.DataFrame(cvec.transform(training_dataframe.news_text.dropna()).todense(), columns=cvec.get_feature_names())\n",
    "    cvec_df.index = training_dataframe.index\n",
    "    cvec_df.columns += \" _news_\"\n",
    "    training_dataframe = training_dataframe.join(cvec_df).drop('news_text',1)\n",
    "\n",
    "    # Transform test\n",
    "    cvec_df  = pd.DataFrame(cvec.transform(testing_dataframe.news_text.dropna()).todense(), columns=cvec.get_feature_names())\n",
    "    cvec_df.index = testing_dataframe.index\n",
    "    cvec_df.columns += \" _news_\"\n",
    "    testing_dataframe = testing_dataframe.join(cvec_df).drop('news_text',1)\n",
    "\n",
    "    # Fit tesla tweet\n",
    "    cvec.fit(training_dataframe.tesla_tweet)\n",
    "\n",
    "    # Transform train\n",
    "    cvec_df  = pd.DataFrame(cvec.transform(training_dataframe.tesla_tweet.dropna()).todense(), columns=cvec.get_feature_names())\n",
    "    cvec_df.index = training_dataframe.index\n",
    "    cvec_df.columns += \" _tesla_\"\n",
    "    training_dataframe = training_dataframe.join(cvec_df).drop('tesla_tweet',1)\n",
    "\n",
    "    # Transform test\n",
    "    cvec_df  = pd.DataFrame(cvec.transform(testing_dataframe.tesla_tweet.dropna()).todense(), columns=cvec.get_feature_names())\n",
    "    cvec_df.index = testing_dataframe.index\n",
    "    cvec_df.columns += \" _tesla_\"\n",
    "    testing_dataframe = testing_dataframe.join(cvec_df).drop('tesla_tweet',1)\n",
    "\n",
    "    # Fit elon tweets\n",
    "    cvec.fit(training_dataframe.elon_tweet)\n",
    "\n",
    "    # Transform train\n",
    "    cvec_df  = pd.DataFrame(cvec.transform(training_dataframe.elon_tweet.dropna()).todense(), columns=cvec.get_feature_names())\n",
    "    cvec_df.index = training_dataframe.index\n",
    "    cvec_df.columns += \" _elon_\"\n",
    "    training_dataframe = training_dataframe.join(cvec_df).drop('elon_tweet',1)\n",
    "\n",
    "    # Transform test\n",
    "    cvec_df  = pd.DataFrame(cvec.transform(testing_dataframe.elon_tweet.dropna()).todense(), columns=cvec.get_feature_names())\n",
    "    cvec_df.index = testing_dataframe.index\n",
    "    cvec_df.columns += \" _elon_\"\n",
    "    testing_dataframe = testing_dataframe.join(cvec_df).drop('elon_tweet',1)\n",
    "    \n",
    "    return training_dataframe, testing_dataframe"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Alright, here we go!"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 89,
   "metadata": {},
   "outputs": [],
   "source": [
    "xgb = XGBClassifier(base_score=0.5, colsample_bylevel=1, colsample_bytree=1,\n",
    "       gamma=0.1, learning_rate=0.1, max_delta_step=0, max_depth=5,\n",
    "       min_child_weight=3, missing=None, n_estimators=100, nthread=-1,\n",
    "       objective='binary:logistic', reg_alpha=0, reg_lambda=1,\n",
    "       scale_pos_weight=1, seed=0, silent=True, subsample=1) "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 104,
   "metadata": {},
   "outputs": [],
   "source": [
    "for num, i in enumerate(range(725, 950)):\n",
    "    train_df = model_df.iloc[1:i]\n",
    "    test_df = model_df.iloc[i:i+1]\n",
    "    train_cvec_df, test_cvec_df = count_vect_our_dfs(train_df, test_df)\n",
    "    train_df = train_cvec_df\n",
    "    test_df = test_cvec_df\n",
    "    X_train = train_df.drop(['target', 'true_grow'], 1)\n",
    "    y_train = train_df.true_grow\n",
    "    X_test = test_df.drop(['target', 'true_grow'], 1)\n",
    "    xgb.fit(X_train, y_train)\n",
    "    real_test_df.iat[num, -1] = xgb.predict(X_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 105,
   "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>close</th>\n",
       "      <th>volume</th>\n",
       "      <th>open</th>\n",
       "      <th>high</th>\n",
       "      <th>low</th>\n",
       "      <th>news_text</th>\n",
       "      <th>tesla_tweet</th>\n",
       "      <th>elon_tweet</th>\n",
       "      <th>9to5Google</th>\n",
       "      <th>9to5Mac</th>\n",
       "      <th>...</th>\n",
       "      <th>iLounge</th>\n",
       "      <th>iMore</th>\n",
       "      <th>iPhone Hacks</th>\n",
       "      <th>iPhone in Canada Blog</th>\n",
       "      <th>lists.swift.org Mailing Lists</th>\n",
       "      <th>forecast</th>\n",
       "      <th>target</th>\n",
       "      <th>next_day_open</th>\n",
       "      <th>true_grow</th>\n",
       "      <th>model_predict</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>date</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>2017-01-03</th>\n",
       "      <td>216.99</td>\n",
       "      <td>5901494.0</td>\n",
       "      <td>214.86</td>\n",
       "      <td>220.3300</td>\n",
       "      <td>210.9600</td>\n",
       "      <td></td>\n",
       "      <td></td>\n",
       "      <td></td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>214.345894</td>\n",
       "      <td>226.99</td>\n",
       "      <td>214.75</td>\n",
       "      <td>1</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2017-01-04</th>\n",
       "      <td>226.99</td>\n",
       "      <td>11204560.0</td>\n",
       "      <td>214.75</td>\n",
       "      <td>228.0000</td>\n",
       "      <td>214.3100</td>\n",
       "      <td></td>\n",
       "      <td>battery cell production begin at the gigafacto...</td>\n",
       "      <td></td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>216.666821</td>\n",
       "      <td>226.75</td>\n",
       "      <td>226.42</td>\n",
       "      <td>1</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2017-01-05</th>\n",
       "      <td>226.75</td>\n",
       "      <td>5908927.0</td>\n",
       "      <td>226.42</td>\n",
       "      <td>227.4800</td>\n",
       "      <td>221.9500</td>\n",
       "      <td></td>\n",
       "      <td></td>\n",
       "      <td>hold-down firing of @spacex falcon 9 at vanden...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>229.154491</td>\n",
       "      <td>229.01</td>\n",
       "      <td>226.93</td>\n",
       "      <td>1</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2017-01-06</th>\n",
       "      <td>229.01</td>\n",
       "      <td>5524153.0</td>\n",
       "      <td>226.93</td>\n",
       "      <td>230.3100</td>\n",
       "      <td>225.4500</td>\n",
       "      <td></td>\n",
       "      <td></td>\n",
       "      <td>@pensasoda http t co 6ies7xbubs via @wired</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>225.668192</td>\n",
       "      <td>231.28</td>\n",
       "      <td>228.97</td>\n",
       "      <td>1</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2017-01-09</th>\n",
       "      <td>231.28</td>\n",
       "      <td>3972716.0</td>\n",
       "      <td>228.97</td>\n",
       "      <td>231.9200</td>\n",
       "      <td>228.0000</td>\n",
       "      <td></td>\n",
       "      <td></td>\n",
       "      <td></td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>227.879001</td>\n",
       "      <td>229.87</td>\n",
       "      <td>232.00</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2017-01-10</th>\n",
       "      <td>229.87</td>\n",
       "      <td>3658981.0</td>\n",
       "      <td>232.00</td>\n",
       "      <td>232.0000</td>\n",
       "      <td>226.8900</td>\n",
       "      <td>chris lattner join tesla a vp of autopilot sof...</td>\n",
       "      <td>welcome chris lattner http t co aaig6rhhdf</td>\n",
       "      <td>rt @teslamotors welcome chris lattner http t c...</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>232.658214</td>\n",
       "      <td>229.73</td>\n",
       "      <td>229.07</td>\n",
       "      <td>1</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2017-01-11</th>\n",
       "      <td>229.73</td>\n",
       "      <td>3649910.0</td>\n",
       "      <td>229.07</td>\n",
       "      <td>229.9800</td>\n",
       "      <td>226.6800</td>\n",
       "      <td></td>\n",
       "      <td></td>\n",
       "      <td>@redletterdave the least cool character it s e...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>231.287654</td>\n",
       "      <td>229.59</td>\n",
       "      <td>229.06</td>\n",
       "      <td>1</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2017-01-12</th>\n",
       "      <td>229.59</td>\n",
       "      <td>3787547.0</td>\n",
       "      <td>229.06</td>\n",
       "      <td>230.7000</td>\n",
       "      <td>225.5800</td>\n",
       "      <td></td>\n",
       "      <td></td>\n",
       "      <td>@brandonr182 @sama have you tried steam engine...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>229.386931</td>\n",
       "      <td>237.75</td>\n",
       "      <td>230.00</td>\n",
       "      <td>1</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2017-01-13</th>\n",
       "      <td>237.75</td>\n",
       "      <td>6085610.0</td>\n",
       "      <td>230.00</td>\n",
       "      <td>237.8500</td>\n",
       "      <td>229.5900</td>\n",
       "      <td></td>\n",
       "      <td>building the supercharger network for the futu...</td>\n",
       "      <td>rt @spacex tracking ground wind for tomorrow s...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>227.790621</td>\n",
       "      <td>235.58</td>\n",
       "      <td>236.70</td>\n",
       "      <td>0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2017-01-17</th>\n",
       "      <td>235.58</td>\n",
       "      <td>4614731.0</td>\n",
       "      <td>236.70</td>\n",
       "      <td>239.9600</td>\n",
       "      <td>234.3700</td>\n",
       "      <td></td>\n",
       "      <td></td>\n",
       "      <td></td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>235.412726</td>\n",
       "      <td>238.36</td>\n",
       "      <td>236.65</td>\n",
       "      <td>1</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2017-01-18</th>\n",
       "      <td>238.36</td>\n",
       "      <td>3767272.0</td>\n",
       "      <td>236.65</td>\n",
       "      <td>239.7100</td>\n",
       "      <td>235.5800</td>\n",
       "      <td></td>\n",
       "      <td>rt @nasagiss 2016 wa the hottest year on recor...</td>\n",
       "      <td>rt @spacex updated launch and landing photo xe...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>233.357445</td>\n",
       "      <td>243.76</td>\n",
       "      <td>247.25</td>\n",
       "      <td>0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2017-01-19</th>\n",
       "      <td>243.76</td>\n",
       "      <td>7717244.0</td>\n",
       "      <td>247.25</td>\n",
       "      <td>248.6800</td>\n",
       "      <td>240.7500</td>\n",
       "      <td>nhtsa close probe into june 2016 tesla crash c...</td>\n",
       "      <td>from exotic island to brewery stadium amp scho...</td>\n",
       "      <td>@ukpjd similar to spacex the t is like a cross...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>238.216219</td>\n",
       "      <td>244.73</td>\n",
       "      <td>245.46</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2017-01-20</th>\n",
       "      <td>244.73</td>\n",
       "      <td>4199720.0</td>\n",
       "      <td>245.46</td>\n",
       "      <td>246.0000</td>\n",
       "      <td>243.0100</td>\n",
       "      <td></td>\n",
       "      <td></td>\n",
       "      <td></td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>241.626987</td>\n",
       "      <td>248.92</td>\n",
       "      <td>245.85</td>\n",
       "      <td>1</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2017-01-23</th>\n",
       "      <td>248.92</td>\n",
       "      <td>6262097.0</td>\n",
       "      <td>245.85</td>\n",
       "      <td>250.8899</td>\n",
       "      <td>245.5000</td>\n",
       "      <td></td>\n",
       "      <td></td>\n",
       "      <td>@stephenstohn yes safety should improve signif...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>244.079093</td>\n",
       "      <td>254.61</td>\n",
       "      <td>250.00</td>\n",
       "      <td>1</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2017-01-24</th>\n",
       "      <td>254.61</td>\n",
       "      <td>4958144.0</td>\n",
       "      <td>250.00</td>\n",
       "      <td>254.8000</td>\n",
       "      <td>249.6500</td>\n",
       "      <td></td>\n",
       "      <td>@itsangiechua look like the fam had fun @fshor...</td>\n",
       "      <td>@redletterdave ok fine soldier 76 my dark secr...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>249.905162</td>\n",
       "      <td>254.47</td>\n",
       "      <td>257.31</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2017-01-25</th>\n",
       "      <td>254.47</td>\n",
       "      <td>5145301.0</td>\n",
       "      <td>257.31</td>\n",
       "      <td>258.4600</td>\n",
       "      <td>251.8000</td>\n",
       "      <td></td>\n",
       "      <td></td>\n",
       "      <td>@theskinsfactory earthquake tend to have the b...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>256.502826</td>\n",
       "      <td>252.51</td>\n",
       "      <td>254.29</td>\n",
       "      <td>0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2017-01-26</th>\n",
       "      <td>252.51</td>\n",
       "      <td>3143717.0</td>\n",
       "      <td>254.29</td>\n",
       "      <td>255.7400</td>\n",
       "      <td>250.7500</td>\n",
       "      <td></td>\n",
       "      <td></td>\n",
       "      <td>rt @gizmodo a brief chat with elon musk about ...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>255.914770</td>\n",
       "      <td>252.95</td>\n",
       "      <td>251.38</td>\n",
       "      <td>1</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2017-01-27</th>\n",
       "      <td>252.95</td>\n",
       "      <td>3161774.0</td>\n",
       "      <td>251.38</td>\n",
       "      <td>253.0000</td>\n",
       "      <td>248.5200</td>\n",
       "      <td>tesla file suit against former autopilot direc...</td>\n",
       "      <td>honored to receive best car award for the thir...</td>\n",
       "      <td>hyperloop pod race happening this weekend at s...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>249.529318</td>\n",
       "      <td>250.63</td>\n",
       "      <td>252.53</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2017-01-30</th>\n",
       "      <td>250.63</td>\n",
       "      <td>3798638.0</td>\n",
       "      <td>252.53</td>\n",
       "      <td>255.2899</td>\n",
       "      <td>247.1000</td>\n",
       "      <td></td>\n",
       "      <td>meet the utility grid of the future tesla s po...</td>\n",
       "      <td>@reesandersen @flixrisk yes i believe that is ...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>252.093966</td>\n",
       "      <td>251.93</td>\n",
       "      <td>249.24</td>\n",
       "      <td>1</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2017-01-31</th>\n",
       "      <td>251.93</td>\n",
       "      <td>4112013.0</td>\n",
       "      <td>249.24</td>\n",
       "      <td>255.8900</td>\n",
       "      <td>247.7000</td>\n",
       "      <td></td>\n",
       "      <td></td>\n",
       "      <td>@sarahfkessler @qz headline amp article is mis...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>248.962630</td>\n",
       "      <td>249.24</td>\n",
       "      <td>253.05</td>\n",
       "      <td>0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2017-02-01</th>\n",
       "      <td>249.24</td>\n",
       "      <td>3953105.0</td>\n",
       "      <td>253.05</td>\n",
       "      <td>253.2000</td>\n",
       "      <td>249.0500</td>\n",
       "      <td>tesla ha been testing four autonomous car on c...</td>\n",
       "      <td></td>\n",
       "      <td></td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>250.975258</td>\n",
       "      <td>251.55</td>\n",
       "      <td>248.34</td>\n",
       "      <td>1</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2017-02-02</th>\n",
       "      <td>251.55</td>\n",
       "      <td>2498799.0</td>\n",
       "      <td>248.34</td>\n",
       "      <td>252.4200</td>\n",
       "      <td>247.7100</td>\n",
       "      <td></td>\n",
       "      <td></td>\n",
       "      <td>@lotte__world yes rt @wired the ai that beat t...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>252.409321</td>\n",
       "      <td>251.33</td>\n",
       "      <td>251.91</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2017-02-03</th>\n",
       "      <td>251.33</td>\n",
       "      <td>2185230.0</td>\n",
       "      <td>251.91</td>\n",
       "      <td>252.1790</td>\n",
       "      <td>249.6800</td>\n",
       "      <td></td>\n",
       "      <td>lithuania know the way to our heart a gigafact...</td>\n",
       "      <td>regarding the meeting at the white house http ...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>250.657184</td>\n",
       "      <td>257.77</td>\n",
       "      <td>251.00</td>\n",
       "      <td>1</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2017-02-06</th>\n",
       "      <td>257.77</td>\n",
       "      <td>3557600.0</td>\n",
       "      <td>251.00</td>\n",
       "      <td>257.8200</td>\n",
       "      <td>250.6300</td>\n",
       "      <td></td>\n",
       "      <td></td>\n",
       "      <td></td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>251.023598</td>\n",
       "      <td>257.48</td>\n",
       "      <td>258.19</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2017-02-07</th>\n",
       "      <td>257.48</td>\n",
       "      <td>4244063.0</td>\n",
       "      <td>258.19</td>\n",
       "      <td>260.0000</td>\n",
       "      <td>256.4200</td>\n",
       "      <td></td>\n",
       "      <td>rt @motortrend 0-60 mph is 2 28 second the @te...</td>\n",
       "      <td>@batlag yes @goel_ishan hoping for summer this...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>258.304469</td>\n",
       "      <td>262.08</td>\n",
       "      <td>257.35</td>\n",
       "      <td>1</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2017-02-08</th>\n",
       "      <td>262.08</td>\n",
       "      <td>3912428.0</td>\n",
       "      <td>257.35</td>\n",
       "      <td>263.3600</td>\n",
       "      <td>256.2000</td>\n",
       "      <td></td>\n",
       "      <td></td>\n",
       "      <td>@sundalj should all be uploaded by next month ...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>258.865779</td>\n",
       "      <td>269.20</td>\n",
       "      <td>266.25</td>\n",
       "      <td>1</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2017-02-09</th>\n",
       "      <td>269.20</td>\n",
       "      <td>7812600.0</td>\n",
       "      <td>266.25</td>\n",
       "      <td>271.1800</td>\n",
       "      <td>266.1500</td>\n",
       "      <td></td>\n",
       "      <td>supercharger availability is now displayed on ...</td>\n",
       "      <td>rt @teslamotors supercharger availability is n...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>262.016872</td>\n",
       "      <td>269.23</td>\n",
       "      <td>269.79</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2017-02-10</th>\n",
       "      <td>269.23</td>\n",
       "      <td>3618336.0</td>\n",
       "      <td>269.79</td>\n",
       "      <td>270.9500</td>\n",
       "      <td>266.1100</td>\n",
       "      <td></td>\n",
       "      <td></td>\n",
       "      <td>falcon 9 rocket now vertical at cape canaveral...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>267.395948</td>\n",
       "      <td>280.60</td>\n",
       "      <td>270.74</td>\n",
       "      <td>1</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2017-02-13</th>\n",
       "      <td>280.60</td>\n",
       "      <td>7023072.0</td>\n",
       "      <td>270.74</td>\n",
       "      <td>280.7899</td>\n",
       "      <td>270.5100</td>\n",
       "      <td></td>\n",
       "      <td></td>\n",
       "      <td></td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>271.352715</td>\n",
       "      <td>280.98</td>\n",
       "      <td>279.03</td>\n",
       "      <td>1</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2017-02-14</th>\n",
       "      <td>280.98</td>\n",
       "      <td>7341450.0</td>\n",
       "      <td>279.03</td>\n",
       "      <td>287.3900</td>\n",
       "      <td>278.6100</td>\n",
       "      <td></td>\n",
       "      <td></td>\n",
       "      <td></td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>280.909746</td>\n",
       "      <td>279.76</td>\n",
       "      <td>280.00</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</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",
       "      <td>...</td>\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",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2017-10-11</th>\n",
       "      <td>354.60</td>\n",
       "      <td>4494733.0</td>\n",
       "      <td>353.89</td>\n",
       "      <td>357.6000</td>\n",
       "      <td>351.1500</td>\n",
       "      <td></td>\n",
       "      <td>@trevorroberts9 we can help please dm u with y...</td>\n",
       "      <td>rt @spacex webcast of falcon 9 launch now live...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>353.586551</td>\n",
       "      <td>355.68</td>\n",
       "      <td>352.95</td>\n",
       "      <td>1</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2017-10-12</th>\n",
       "      <td>355.68</td>\n",
       "      <td>4073552.0</td>\n",
       "      <td>352.95</td>\n",
       "      <td>359.7800</td>\n",
       "      <td>352.6400</td>\n",
       "      <td></td>\n",
       "      <td></td>\n",
       "      <td>@_andyoneal @patriciapaje that s just to win o...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>358.030641</td>\n",
       "      <td>355.57</td>\n",
       "      <td>356.98</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2017-10-13</th>\n",
       "      <td>355.57</td>\n",
       "      <td>3538807.0</td>\n",
       "      <td>356.98</td>\n",
       "      <td>358.4900</td>\n",
       "      <td>353.6800</td>\n",
       "      <td></td>\n",
       "      <td></td>\n",
       "      <td></td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>353.445050</td>\n",
       "      <td>350.60</td>\n",
       "      <td>353.76</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2017-10-16</th>\n",
       "      <td>350.60</td>\n",
       "      <td>5373090.0</td>\n",
       "      <td>353.76</td>\n",
       "      <td>354.4800</td>\n",
       "      <td>347.1600</td>\n",
       "      <td></td>\n",
       "      <td></td>\n",
       "      <td>amazing sim of neutron star colliding http t c...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>358.925896</td>\n",
       "      <td>355.75</td>\n",
       "      <td>350.91</td>\n",
       "      <td>1</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2017-10-17</th>\n",
       "      <td>355.75</td>\n",
       "      <td>3290290.0</td>\n",
       "      <td>350.91</td>\n",
       "      <td>356.2200</td>\n",
       "      <td>350.0700</td>\n",
       "      <td></td>\n",
       "      <td>@sostre we can help please dm u with your cont...</td>\n",
       "      <td></td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>352.114824</td>\n",
       "      <td>359.65</td>\n",
       "      <td>355.97</td>\n",
       "      <td>1</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2017-10-18</th>\n",
       "      <td>359.65</td>\n",
       "      <td>4936024.0</td>\n",
       "      <td>355.97</td>\n",
       "      <td>363.0000</td>\n",
       "      <td>354.1300</td>\n",
       "      <td></td>\n",
       "      <td>@gadgetvillain can you send u over a dm with y...</td>\n",
       "      <td>rt @openai building advanced ai and only then ...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>354.809866</td>\n",
       "      <td>351.81</td>\n",
       "      <td>355.56</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2017-10-19</th>\n",
       "      <td>351.81</td>\n",
       "      <td>5055715.0</td>\n",
       "      <td>355.56</td>\n",
       "      <td>357.1465</td>\n",
       "      <td>348.2000</td>\n",
       "      <td></td>\n",
       "      <td>rt @businessinsider i took my entire family pl...</td>\n",
       "      <td>congratulation to coulter tilton of missouri b...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>358.948179</td>\n",
       "      <td>345.10</td>\n",
       "      <td>352.69</td>\n",
       "      <td>0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2017-10-20</th>\n",
       "      <td>345.10</td>\n",
       "      <td>4925657.0</td>\n",
       "      <td>352.69</td>\n",
       "      <td>354.5500</td>\n",
       "      <td>344.3400</td>\n",
       "      <td></td>\n",
       "      <td></td>\n",
       "      <td>@apta_info @baltimoresun not ready to do a pro...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>350.795980</td>\n",
       "      <td>337.02</td>\n",
       "      <td>349.88</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2017-10-23</th>\n",
       "      <td>337.02</td>\n",
       "      <td>5739598.0</td>\n",
       "      <td>349.88</td>\n",
       "      <td>349.9500</td>\n",
       "      <td>336.2500</td>\n",
       "      <td></td>\n",
       "      <td></td>\n",
       "      <td></td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>344.160377</td>\n",
       "      <td>337.34</td>\n",
       "      <td>338.80</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2017-10-24</th>\n",
       "      <td>337.34</td>\n",
       "      <td>4485692.0</td>\n",
       "      <td>338.80</td>\n",
       "      <td>342.8000</td>\n",
       "      <td>336.1600</td>\n",
       "      <td></td>\n",
       "      <td>hospital del ni xc3 xb1o is first of many sola...</td>\n",
       "      <td>rt @tesla hospital del ni xc3 xb1o is first of...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>337.297104</td>\n",
       "      <td>325.84</td>\n",
       "      <td>336.70</td>\n",
       "      <td>0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2017-10-25</th>\n",
       "      <td>325.84</td>\n",
       "      <td>8583619.0</td>\n",
       "      <td>336.70</td>\n",
       "      <td>337.5000</td>\n",
       "      <td>323.5600</td>\n",
       "      <td></td>\n",
       "      <td></td>\n",
       "      <td></td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>333.651923</td>\n",
       "      <td>326.17</td>\n",
       "      <td>327.78</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2017-10-26</th>\n",
       "      <td>326.17</td>\n",
       "      <td>5010087.0</td>\n",
       "      <td>327.78</td>\n",
       "      <td>330.2300</td>\n",
       "      <td>323.2000</td>\n",
       "      <td></td>\n",
       "      <td>@evil_igz @solarcity yes can you send u a dm w...</td>\n",
       "      <td>btw just want to express a word of appreciatio...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>324.541821</td>\n",
       "      <td>320.87</td>\n",
       "      <td>319.75</td>\n",
       "      <td>1</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2017-10-27</th>\n",
       "      <td>320.87</td>\n",
       "      <td>6968365.0</td>\n",
       "      <td>319.75</td>\n",
       "      <td>324.5900</td>\n",
       "      <td>316.6600</td>\n",
       "      <td></td>\n",
       "      <td></td>\n",
       "      <td></td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>324.564084</td>\n",
       "      <td>320.08</td>\n",
       "      <td>319.18</td>\n",
       "      <td>1</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2017-10-30</th>\n",
       "      <td>320.08</td>\n",
       "      <td>4247619.0</td>\n",
       "      <td>319.18</td>\n",
       "      <td>323.7800</td>\n",
       "      <td>317.2500</td>\n",
       "      <td></td>\n",
       "      <td>atmospheric co2 level hit an 800 000-year high...</td>\n",
       "      <td>rt @spacex falcon 9 and koreasat-5a went verti...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>324.219697</td>\n",
       "      <td>331.53</td>\n",
       "      <td>320.23</td>\n",
       "      <td>1</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2017-10-31</th>\n",
       "      <td>331.53</td>\n",
       "      <td>5650062.0</td>\n",
       "      <td>320.23</td>\n",
       "      <td>331.9500</td>\n",
       "      <td>320.1800</td>\n",
       "      <td></td>\n",
       "      <td></td>\n",
       "      <td></td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>319.830144</td>\n",
       "      <td>321.08</td>\n",
       "      <td>332.25</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2017-11-01</th>\n",
       "      <td>321.08</td>\n",
       "      <td>8242487.0</td>\n",
       "      <td>332.25</td>\n",
       "      <td>332.6089</td>\n",
       "      <td>320.2600</td>\n",
       "      <td></td>\n",
       "      <td></td>\n",
       "      <td></td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>335.547149</td>\n",
       "      <td>299.26</td>\n",
       "      <td>300.13</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2017-11-02</th>\n",
       "      <td>299.26</td>\n",
       "      <td>19771280.0</td>\n",
       "      <td>300.13</td>\n",
       "      <td>308.6900</td>\n",
       "      <td>292.6300</td>\n",
       "      <td></td>\n",
       "      <td>@drod0118 please send u a dm with your contact...</td>\n",
       "      <td></td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>321.775186</td>\n",
       "      <td>306.09</td>\n",
       "      <td>299.50</td>\n",
       "      <td>1</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2017-11-03</th>\n",
       "      <td>306.09</td>\n",
       "      <td>8885933.0</td>\n",
       "      <td>299.50</td>\n",
       "      <td>306.2500</td>\n",
       "      <td>295.1300</td>\n",
       "      <td></td>\n",
       "      <td></td>\n",
       "      <td></td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>299.548415</td>\n",
       "      <td>302.78</td>\n",
       "      <td>307.00</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2017-11-06</th>\n",
       "      <td>302.78</td>\n",
       "      <td>6480689.0</td>\n",
       "      <td>307.00</td>\n",
       "      <td>307.5000</td>\n",
       "      <td>299.0100</td>\n",
       "      <td>source chinese internet company tencent is dev...</td>\n",
       "      <td></td>\n",
       "      <td></td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>305.910012</td>\n",
       "      <td>306.05</td>\n",
       "      <td>301.02</td>\n",
       "      <td>1</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2017-11-07</th>\n",
       "      <td>306.05</td>\n",
       "      <td>5284844.0</td>\n",
       "      <td>301.02</td>\n",
       "      <td>306.5000</td>\n",
       "      <td>300.0300</td>\n",
       "      <td></td>\n",
       "      <td>rt @wsj what make l a traffic slightly le aggr...</td>\n",
       "      <td>@bob_richards one rocket slightly toasted</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>301.842300</td>\n",
       "      <td>304.39</td>\n",
       "      <td>305.50</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2017-11-08</th>\n",
       "      <td>304.39</td>\n",
       "      <td>4725355.0</td>\n",
       "      <td>305.50</td>\n",
       "      <td>306.8900</td>\n",
       "      <td>301.3001</td>\n",
       "      <td></td>\n",
       "      <td></td>\n",
       "      <td>atat xc3 xbcrk an xc4 xb1tkabir http t co al3w...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>306.634741</td>\n",
       "      <td>302.99</td>\n",
       "      <td>302.50</td>\n",
       "      <td>1</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2017-11-09</th>\n",
       "      <td>302.99</td>\n",
       "      <td>5440173.0</td>\n",
       "      <td>302.50</td>\n",
       "      <td>304.4600</td>\n",
       "      <td>296.3000</td>\n",
       "      <td></td>\n",
       "      <td>@lilamaria04 can you send u a dm we can get th...</td>\n",
       "      <td>if one day my word are against science choose ...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>303.907229</td>\n",
       "      <td>302.99</td>\n",
       "      <td>302.50</td>\n",
       "      <td>1</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2017-11-10</th>\n",
       "      <td>302.99</td>\n",
       "      <td>4621080.0</td>\n",
       "      <td>302.50</td>\n",
       "      <td>308.3600</td>\n",
       "      <td>301.8500</td>\n",
       "      <td></td>\n",
       "      <td></td>\n",
       "      <td>@newscientist sure</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>303.088822</td>\n",
       "      <td>315.40</td>\n",
       "      <td>300.13</td>\n",
       "      <td>1</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2017-11-13</th>\n",
       "      <td>315.40</td>\n",
       "      <td>7565242.0</td>\n",
       "      <td>300.13</td>\n",
       "      <td>316.8000</td>\n",
       "      <td>299.1100</td>\n",
       "      <td>steve jurvetson is leaving draper fisher jurve...</td>\n",
       "      <td></td>\n",
       "      <td>@newscientist say scientist xe2 x80 x94 among ...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>303.366486</td>\n",
       "      <td>308.70</td>\n",
       "      <td>315.00</td>\n",
       "      <td>0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2017-11-14</th>\n",
       "      <td>308.70</td>\n",
       "      <td>5661904.0</td>\n",
       "      <td>315.00</td>\n",
       "      <td>316.3500</td>\n",
       "      <td>306.9000</td>\n",
       "      <td></td>\n",
       "      <td>mit researcher rebut misleading @ft article xe...</td>\n",
       "      <td></td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>317.474216</td>\n",
       "      <td>311.30</td>\n",
       "      <td>306.01</td>\n",
       "      <td>1</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2017-11-15</th>\n",
       "      <td>311.30</td>\n",
       "      <td>5951835.0</td>\n",
       "      <td>306.01</td>\n",
       "      <td>312.4900</td>\n",
       "      <td>301.5000</td>\n",
       "      <td></td>\n",
       "      <td>tesla semi unveil 8pm pt tomorrow xe2 x80 x94 ...</td>\n",
       "      <td>sorry date is tomorrow night at 8pm this pic i...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>308.246135</td>\n",
       "      <td>312.50</td>\n",
       "      <td>313.99</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2017-11-16</th>\n",
       "      <td>312.50</td>\n",
       "      <td>5794409.0</td>\n",
       "      <td>313.99</td>\n",
       "      <td>318.1400</td>\n",
       "      <td>311.3000</td>\n",
       "      <td></td>\n",
       "      <td></td>\n",
       "      <td></td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>311.798508</td>\n",
       "      <td>315.05</td>\n",
       "      <td>325.67</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2017-11-17</th>\n",
       "      <td>315.05</td>\n",
       "      <td>13720830.0</td>\n",
       "      <td>325.67</td>\n",
       "      <td>326.6700</td>\n",
       "      <td>313.1500</td>\n",
       "      <td></td>\n",
       "      <td>going into plaid http t co ouu9s7cuwi bamf htt...</td>\n",
       "      <td>tesla semi http t co 7vlz7f46ji 0 to 100 km h ...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>314.420964</td>\n",
       "      <td>308.74</td>\n",
       "      <td>313.79</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2017-11-20</th>\n",
       "      <td>308.74</td>\n",
       "      <td>8241821.0</td>\n",
       "      <td>313.79</td>\n",
       "      <td>315.5000</td>\n",
       "      <td>304.7500</td>\n",
       "      <td></td>\n",
       "      <td></td>\n",
       "      <td>@schesnais good point</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>315.999310</td>\n",
       "      <td>317.81</td>\n",
       "      <td>310.86</td>\n",
       "      <td>1</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2017-11-21</th>\n",
       "      <td>317.81</td>\n",
       "      <td>7256381.0</td>\n",
       "      <td>310.86</td>\n",
       "      <td>318.2300</td>\n",
       "      <td>308.7100</td>\n",
       "      <td></td>\n",
       "      <td></td>\n",
       "      <td>initial hat offering going great w over 300k i...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>306.138106</td>\n",
       "      <td>312.60</td>\n",
       "      <td>316.77</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>225 rows × 311 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "             close      volume    open      high       low  \\\n",
       "date                                                         \n",
       "2017-01-03  216.99   5901494.0  214.86  220.3300  210.9600   \n",
       "2017-01-04  226.99  11204560.0  214.75  228.0000  214.3100   \n",
       "2017-01-05  226.75   5908927.0  226.42  227.4800  221.9500   \n",
       "2017-01-06  229.01   5524153.0  226.93  230.3100  225.4500   \n",
       "2017-01-09  231.28   3972716.0  228.97  231.9200  228.0000   \n",
       "2017-01-10  229.87   3658981.0  232.00  232.0000  226.8900   \n",
       "2017-01-11  229.73   3649910.0  229.07  229.9800  226.6800   \n",
       "2017-01-12  229.59   3787547.0  229.06  230.7000  225.5800   \n",
       "2017-01-13  237.75   6085610.0  230.00  237.8500  229.5900   \n",
       "2017-01-17  235.58   4614731.0  236.70  239.9600  234.3700   \n",
       "2017-01-18  238.36   3767272.0  236.65  239.7100  235.5800   \n",
       "2017-01-19  243.76   7717244.0  247.25  248.6800  240.7500   \n",
       "2017-01-20  244.73   4199720.0  245.46  246.0000  243.0100   \n",
       "2017-01-23  248.92   6262097.0  245.85  250.8899  245.5000   \n",
       "2017-01-24  254.61   4958144.0  250.00  254.8000  249.6500   \n",
       "2017-01-25  254.47   5145301.0  257.31  258.4600  251.8000   \n",
       "2017-01-26  252.51   3143717.0  254.29  255.7400  250.7500   \n",
       "2017-01-27  252.95   3161774.0  251.38  253.0000  248.5200   \n",
       "2017-01-30  250.63   3798638.0  252.53  255.2899  247.1000   \n",
       "2017-01-31  251.93   4112013.0  249.24  255.8900  247.7000   \n",
       "2017-02-01  249.24   3953105.0  253.05  253.2000  249.0500   \n",
       "2017-02-02  251.55   2498799.0  248.34  252.4200  247.7100   \n",
       "2017-02-03  251.33   2185230.0  251.91  252.1790  249.6800   \n",
       "2017-02-06  257.77   3557600.0  251.00  257.8200  250.6300   \n",
       "2017-02-07  257.48   4244063.0  258.19  260.0000  256.4200   \n",
       "2017-02-08  262.08   3912428.0  257.35  263.3600  256.2000   \n",
       "2017-02-09  269.20   7812600.0  266.25  271.1800  266.1500   \n",
       "2017-02-10  269.23   3618336.0  269.79  270.9500  266.1100   \n",
       "2017-02-13  280.60   7023072.0  270.74  280.7899  270.5100   \n",
       "2017-02-14  280.98   7341450.0  279.03  287.3900  278.6100   \n",
       "...            ...         ...     ...       ...       ...   \n",
       "2017-10-11  354.60   4494733.0  353.89  357.6000  351.1500   \n",
       "2017-10-12  355.68   4073552.0  352.95  359.7800  352.6400   \n",
       "2017-10-13  355.57   3538807.0  356.98  358.4900  353.6800   \n",
       "2017-10-16  350.60   5373090.0  353.76  354.4800  347.1600   \n",
       "2017-10-17  355.75   3290290.0  350.91  356.2200  350.0700   \n",
       "2017-10-18  359.65   4936024.0  355.97  363.0000  354.1300   \n",
       "2017-10-19  351.81   5055715.0  355.56  357.1465  348.2000   \n",
       "2017-10-20  345.10   4925657.0  352.69  354.5500  344.3400   \n",
       "2017-10-23  337.02   5739598.0  349.88  349.9500  336.2500   \n",
       "2017-10-24  337.34   4485692.0  338.80  342.8000  336.1600   \n",
       "2017-10-25  325.84   8583619.0  336.70  337.5000  323.5600   \n",
       "2017-10-26  326.17   5010087.0  327.78  330.2300  323.2000   \n",
       "2017-10-27  320.87   6968365.0  319.75  324.5900  316.6600   \n",
       "2017-10-30  320.08   4247619.0  319.18  323.7800  317.2500   \n",
       "2017-10-31  331.53   5650062.0  320.23  331.9500  320.1800   \n",
       "2017-11-01  321.08   8242487.0  332.25  332.6089  320.2600   \n",
       "2017-11-02  299.26  19771280.0  300.13  308.6900  292.6300   \n",
       "2017-11-03  306.09   8885933.0  299.50  306.2500  295.1300   \n",
       "2017-11-06  302.78   6480689.0  307.00  307.5000  299.0100   \n",
       "2017-11-07  306.05   5284844.0  301.02  306.5000  300.0300   \n",
       "2017-11-08  304.39   4725355.0  305.50  306.8900  301.3001   \n",
       "2017-11-09  302.99   5440173.0  302.50  304.4600  296.3000   \n",
       "2017-11-10  302.99   4621080.0  302.50  308.3600  301.8500   \n",
       "2017-11-13  315.40   7565242.0  300.13  316.8000  299.1100   \n",
       "2017-11-14  308.70   5661904.0  315.00  316.3500  306.9000   \n",
       "2017-11-15  311.30   5951835.0  306.01  312.4900  301.5000   \n",
       "2017-11-16  312.50   5794409.0  313.99  318.1400  311.3000   \n",
       "2017-11-17  315.05  13720830.0  325.67  326.6700  313.1500   \n",
       "2017-11-20  308.74   8241821.0  313.79  315.5000  304.7500   \n",
       "2017-11-21  317.81   7256381.0  310.86  318.2300  308.7100   \n",
       "\n",
       "                                                    news_text  \\\n",
       "date                                                            \n",
       "2017-01-03                                                      \n",
       "2017-01-04                                                      \n",
       "2017-01-05                                                      \n",
       "2017-01-06                                                      \n",
       "2017-01-09                                                      \n",
       "2017-01-10  chris lattner join tesla a vp of autopilot sof...   \n",
       "2017-01-11                                                      \n",
       "2017-01-12                                                      \n",
       "2017-01-13                                                      \n",
       "2017-01-17                                                      \n",
       "2017-01-18                                                      \n",
       "2017-01-19  nhtsa close probe into june 2016 tesla crash c...   \n",
       "2017-01-20                                                      \n",
       "2017-01-23                                                      \n",
       "2017-01-24                                                      \n",
       "2017-01-25                                                      \n",
       "2017-01-26                                                      \n",
       "2017-01-27  tesla file suit against former autopilot direc...   \n",
       "2017-01-30                                                      \n",
       "2017-01-31                                                      \n",
       "2017-02-01  tesla ha been testing four autonomous car on c...   \n",
       "2017-02-02                                                      \n",
       "2017-02-03                                                      \n",
       "2017-02-06                                                      \n",
       "2017-02-07                                                      \n",
       "2017-02-08                                                      \n",
       "2017-02-09                                                      \n",
       "2017-02-10                                                      \n",
       "2017-02-13                                                      \n",
       "2017-02-14                                                      \n",
       "...                                                       ...   \n",
       "2017-10-11                                                      \n",
       "2017-10-12                                                      \n",
       "2017-10-13                                                      \n",
       "2017-10-16                                                      \n",
       "2017-10-17                                                      \n",
       "2017-10-18                                                      \n",
       "2017-10-19                                                      \n",
       "2017-10-20                                                      \n",
       "2017-10-23                                                      \n",
       "2017-10-24                                                      \n",
       "2017-10-25                                                      \n",
       "2017-10-26                                                      \n",
       "2017-10-27                                                      \n",
       "2017-10-30                                                      \n",
       "2017-10-31                                                      \n",
       "2017-11-01                                                      \n",
       "2017-11-02                                                      \n",
       "2017-11-03                                                      \n",
       "2017-11-06  source chinese internet company tencent is dev...   \n",
       "2017-11-07                                                      \n",
       "2017-11-08                                                      \n",
       "2017-11-09                                                      \n",
       "2017-11-10                                                      \n",
       "2017-11-13  steve jurvetson is leaving draper fisher jurve...   \n",
       "2017-11-14                                                      \n",
       "2017-11-15                                                      \n",
       "2017-11-16                                                      \n",
       "2017-11-17                                                      \n",
       "2017-11-20                                                      \n",
       "2017-11-21                                                      \n",
       "\n",
       "                                                  tesla_tweet  \\\n",
       "date                                                            \n",
       "2017-01-03                                                      \n",
       "2017-01-04  battery cell production begin at the gigafacto...   \n",
       "2017-01-05                                                      \n",
       "2017-01-06                                                      \n",
       "2017-01-09                                                      \n",
       "2017-01-10        welcome chris lattner http t co aaig6rhhdf    \n",
       "2017-01-11                                                      \n",
       "2017-01-12                                                      \n",
       "2017-01-13  building the supercharger network for the futu...   \n",
       "2017-01-17                                                      \n",
       "2017-01-18  rt @nasagiss 2016 wa the hottest year on recor...   \n",
       "2017-01-19  from exotic island to brewery stadium amp scho...   \n",
       "2017-01-20                                                      \n",
       "2017-01-23                                                      \n",
       "2017-01-24  @itsangiechua look like the fam had fun @fshor...   \n",
       "2017-01-25                                                      \n",
       "2017-01-26                                                      \n",
       "2017-01-27  honored to receive best car award for the thir...   \n",
       "2017-01-30  meet the utility grid of the future tesla s po...   \n",
       "2017-01-31                                                      \n",
       "2017-02-01                                                      \n",
       "2017-02-02                                                      \n",
       "2017-02-03  lithuania know the way to our heart a gigafact...   \n",
       "2017-02-06                                                      \n",
       "2017-02-07  rt @motortrend 0-60 mph is 2 28 second the @te...   \n",
       "2017-02-08                                                      \n",
       "2017-02-09  supercharger availability is now displayed on ...   \n",
       "2017-02-10                                                      \n",
       "2017-02-13                                                      \n",
       "2017-02-14                                                      \n",
       "...                                                       ...   \n",
       "2017-10-11  @trevorroberts9 we can help please dm u with y...   \n",
       "2017-10-12                                                      \n",
       "2017-10-13                                                      \n",
       "2017-10-16                                                      \n",
       "2017-10-17  @sostre we can help please dm u with your cont...   \n",
       "2017-10-18  @gadgetvillain can you send u over a dm with y...   \n",
       "2017-10-19  rt @businessinsider i took my entire family pl...   \n",
       "2017-10-20                                                      \n",
       "2017-10-23                                                      \n",
       "2017-10-24  hospital del ni xc3 xb1o is first of many sola...   \n",
       "2017-10-25                                                      \n",
       "2017-10-26  @evil_igz @solarcity yes can you send u a dm w...   \n",
       "2017-10-27                                                      \n",
       "2017-10-30  atmospheric co2 level hit an 800 000-year high...   \n",
       "2017-10-31                                                      \n",
       "2017-11-01                                                      \n",
       "2017-11-02  @drod0118 please send u a dm with your contact...   \n",
       "2017-11-03                                                      \n",
       "2017-11-06                                                      \n",
       "2017-11-07  rt @wsj what make l a traffic slightly le aggr...   \n",
       "2017-11-08                                                      \n",
       "2017-11-09  @lilamaria04 can you send u a dm we can get th...   \n",
       "2017-11-10                                                      \n",
       "2017-11-13                                                      \n",
       "2017-11-14  mit researcher rebut misleading @ft article xe...   \n",
       "2017-11-15  tesla semi unveil 8pm pt tomorrow xe2 x80 x94 ...   \n",
       "2017-11-16                                                      \n",
       "2017-11-17  going into plaid http t co ouu9s7cuwi bamf htt...   \n",
       "2017-11-20                                                      \n",
       "2017-11-21                                                      \n",
       "\n",
       "                                                   elon_tweet  9to5Google  \\\n",
       "date                                                                        \n",
       "2017-01-03                                                              0   \n",
       "2017-01-04                                                              0   \n",
       "2017-01-05  hold-down firing of @spacex falcon 9 at vanden...           0   \n",
       "2017-01-06        @pensasoda http t co 6ies7xbubs via @wired            0   \n",
       "2017-01-09                                                              0   \n",
       "2017-01-10  rt @teslamotors welcome chris lattner http t c...           0   \n",
       "2017-01-11  @redletterdave the least cool character it s e...           0   \n",
       "2017-01-12  @brandonr182 @sama have you tried steam engine...           0   \n",
       "2017-01-13  rt @spacex tracking ground wind for tomorrow s...           0   \n",
       "2017-01-17                                                              0   \n",
       "2017-01-18  rt @spacex updated launch and landing photo xe...           0   \n",
       "2017-01-19  @ukpjd similar to spacex the t is like a cross...           0   \n",
       "2017-01-20                                                              0   \n",
       "2017-01-23  @stephenstohn yes safety should improve signif...           0   \n",
       "2017-01-24  @redletterdave ok fine soldier 76 my dark secr...           0   \n",
       "2017-01-25  @theskinsfactory earthquake tend to have the b...           0   \n",
       "2017-01-26  rt @gizmodo a brief chat with elon musk about ...           0   \n",
       "2017-01-27  hyperloop pod race happening this weekend at s...           0   \n",
       "2017-01-30  @reesandersen @flixrisk yes i believe that is ...           0   \n",
       "2017-01-31  @sarahfkessler @qz headline amp article is mis...           0   \n",
       "2017-02-01                                                              0   \n",
       "2017-02-02  @lotte__world yes rt @wired the ai that beat t...           0   \n",
       "2017-02-03  regarding the meeting at the white house http ...           0   \n",
       "2017-02-06                                                              0   \n",
       "2017-02-07  @batlag yes @goel_ishan hoping for summer this...           0   \n",
       "2017-02-08  @sundalj should all be uploaded by next month ...           0   \n",
       "2017-02-09  rt @teslamotors supercharger availability is n...           0   \n",
       "2017-02-10  falcon 9 rocket now vertical at cape canaveral...           0   \n",
       "2017-02-13                                                              0   \n",
       "2017-02-14                                                              0   \n",
       "...                                                       ...         ...   \n",
       "2017-10-11  rt @spacex webcast of falcon 9 launch now live...           0   \n",
       "2017-10-12  @_andyoneal @patriciapaje that s just to win o...           0   \n",
       "2017-10-13                                                              0   \n",
       "2017-10-16  amazing sim of neutron star colliding http t c...           0   \n",
       "2017-10-17                                                              0   \n",
       "2017-10-18  rt @openai building advanced ai and only then ...           0   \n",
       "2017-10-19  congratulation to coulter tilton of missouri b...           0   \n",
       "2017-10-20  @apta_info @baltimoresun not ready to do a pro...           0   \n",
       "2017-10-23                                                              0   \n",
       "2017-10-24  rt @tesla hospital del ni xc3 xb1o is first of...           0   \n",
       "2017-10-25                                                              0   \n",
       "2017-10-26  btw just want to express a word of appreciatio...           0   \n",
       "2017-10-27                                                              0   \n",
       "2017-10-30  rt @spacex falcon 9 and koreasat-5a went verti...           0   \n",
       "2017-10-31                                                              0   \n",
       "2017-11-01                                                              0   \n",
       "2017-11-02                                                              0   \n",
       "2017-11-03                                                              0   \n",
       "2017-11-06                                                              0   \n",
       "2017-11-07         @bob_richards one rocket slightly toasted            0   \n",
       "2017-11-08  atat xc3 xbcrk an xc4 xb1tkabir http t co al3w...           0   \n",
       "2017-11-09  if one day my word are against science choose ...           0   \n",
       "2017-11-10                                @newscientist sure            0   \n",
       "2017-11-13  @newscientist say scientist xe2 x80 x94 among ...           0   \n",
       "2017-11-14                                                              0   \n",
       "2017-11-15  sorry date is tomorrow night at 8pm this pic i...           0   \n",
       "2017-11-16                                                              0   \n",
       "2017-11-17  tesla semi http t co 7vlz7f46ji 0 to 100 km h ...           0   \n",
       "2017-11-20                             @schesnais good point            0   \n",
       "2017-11-21  initial hat offering going great w over 300k i...           0   \n",
       "\n",
       "            9to5Mac      ...        iLounge  iMore  iPhone Hacks  \\\n",
       "date                     ...                                       \n",
       "2017-01-03        0      ...              0      0             0   \n",
       "2017-01-04        0      ...              0      0             0   \n",
       "2017-01-05        0      ...              0      0             0   \n",
       "2017-01-06        0      ...              0      0             0   \n",
       "2017-01-09        0      ...              0      0             0   \n",
       "2017-01-10        1      ...              0      1             1   \n",
       "2017-01-11        0      ...              0      0             0   \n",
       "2017-01-12        0      ...              0      0             0   \n",
       "2017-01-13        0      ...              0      0             0   \n",
       "2017-01-17        0      ...              0      0             0   \n",
       "2017-01-18        0      ...              0      0             0   \n",
       "2017-01-19        0      ...              0      0             0   \n",
       "2017-01-20        0      ...              0      0             0   \n",
       "2017-01-23        0      ...              0      0             0   \n",
       "2017-01-24        0      ...              0      0             0   \n",
       "2017-01-25        0      ...              0      0             0   \n",
       "2017-01-26        0      ...              0      0             0   \n",
       "2017-01-27        0      ...              0      0             0   \n",
       "2017-01-30        0      ...              0      0             0   \n",
       "2017-01-31        0      ...              0      0             0   \n",
       "2017-02-01        0      ...              0      0             0   \n",
       "2017-02-02        0      ...              0      0             0   \n",
       "2017-02-03        0      ...              0      0             0   \n",
       "2017-02-06        0      ...              0      0             0   \n",
       "2017-02-07        0      ...              0      0             0   \n",
       "2017-02-08        0      ...              0      0             0   \n",
       "2017-02-09        0      ...              0      0             0   \n",
       "2017-02-10        0      ...              0      0             0   \n",
       "2017-02-13        0      ...              0      0             0   \n",
       "2017-02-14        0      ...              0      0             0   \n",
       "...             ...      ...            ...    ...           ...   \n",
       "2017-10-11        0      ...              0      0             0   \n",
       "2017-10-12        0      ...              0      0             0   \n",
       "2017-10-13        0      ...              0      0             0   \n",
       "2017-10-16        0      ...              0      0             0   \n",
       "2017-10-17        0      ...              0      0             0   \n",
       "2017-10-18        0      ...              0      0             0   \n",
       "2017-10-19        0      ...              0      0             0   \n",
       "2017-10-20        0      ...              0      0             0   \n",
       "2017-10-23        0      ...              0      0             0   \n",
       "2017-10-24        0      ...              0      0             0   \n",
       "2017-10-25        0      ...              0      0             0   \n",
       "2017-10-26        0      ...              0      0             0   \n",
       "2017-10-27        0      ...              0      0             0   \n",
       "2017-10-30        0      ...              0      0             0   \n",
       "2017-10-31        0      ...              0      0             0   \n",
       "2017-11-01        0      ...              0      0             0   \n",
       "2017-11-02        0      ...              0      0             0   \n",
       "2017-11-03        0      ...              0      0             0   \n",
       "2017-11-06        0      ...              0      0             0   \n",
       "2017-11-07        0      ...              0      0             0   \n",
       "2017-11-08        0      ...              0      0             0   \n",
       "2017-11-09        0      ...              0      0             0   \n",
       "2017-11-10        0      ...              0      0             0   \n",
       "2017-11-13        0      ...              0      0             0   \n",
       "2017-11-14        0      ...              0      0             0   \n",
       "2017-11-15        0      ...              0      0             0   \n",
       "2017-11-16        0      ...              0      0             0   \n",
       "2017-11-17        0      ...              0      0             0   \n",
       "2017-11-20        0      ...              0      0             0   \n",
       "2017-11-21        0      ...              0      0             0   \n",
       "\n",
       "            iPhone in Canada Blog  lists.swift.org Mailing Lists    forecast  \\\n",
       "date                                                                           \n",
       "2017-01-03                      0                              0  214.345894   \n",
       "2017-01-04                      0                              0  216.666821   \n",
       "2017-01-05                      0                              0  229.154491   \n",
       "2017-01-06                      0                              0  225.668192   \n",
       "2017-01-09                      0                              0  227.879001   \n",
       "2017-01-10                      1                              1  232.658214   \n",
       "2017-01-11                      0                              0  231.287654   \n",
       "2017-01-12                      0                              0  229.386931   \n",
       "2017-01-13                      0                              0  227.790621   \n",
       "2017-01-17                      0                              0  235.412726   \n",
       "2017-01-18                      0                              0  233.357445   \n",
       "2017-01-19                      1                              0  238.216219   \n",
       "2017-01-20                      0                              0  241.626987   \n",
       "2017-01-23                      0                              0  244.079093   \n",
       "2017-01-24                      0                              0  249.905162   \n",
       "2017-01-25                      0                              0  256.502826   \n",
       "2017-01-26                      0                              0  255.914770   \n",
       "2017-01-27                      0                              0  249.529318   \n",
       "2017-01-30                      0                              0  252.093966   \n",
       "2017-01-31                      0                              0  248.962630   \n",
       "2017-02-01                      0                              0  250.975258   \n",
       "2017-02-02                      0                              0  252.409321   \n",
       "2017-02-03                      0                              0  250.657184   \n",
       "2017-02-06                      0                              0  251.023598   \n",
       "2017-02-07                      0                              0  258.304469   \n",
       "2017-02-08                      0                              0  258.865779   \n",
       "2017-02-09                      0                              0  262.016872   \n",
       "2017-02-10                      0                              0  267.395948   \n",
       "2017-02-13                      0                              0  271.352715   \n",
       "2017-02-14                      0                              0  280.909746   \n",
       "...                           ...                            ...         ...   \n",
       "2017-10-11                      0                              0  353.586551   \n",
       "2017-10-12                      0                              0  358.030641   \n",
       "2017-10-13                      0                              0  353.445050   \n",
       "2017-10-16                      0                              0  358.925896   \n",
       "2017-10-17                      0                              0  352.114824   \n",
       "2017-10-18                      0                              0  354.809866   \n",
       "2017-10-19                      0                              0  358.948179   \n",
       "2017-10-20                      0                              0  350.795980   \n",
       "2017-10-23                      0                              0  344.160377   \n",
       "2017-10-24                      0                              0  337.297104   \n",
       "2017-10-25                      0                              0  333.651923   \n",
       "2017-10-26                      0                              0  324.541821   \n",
       "2017-10-27                      0                              0  324.564084   \n",
       "2017-10-30                      0                              0  324.219697   \n",
       "2017-10-31                      0                              0  319.830144   \n",
       "2017-11-01                      0                              0  335.547149   \n",
       "2017-11-02                      0                              0  321.775186   \n",
       "2017-11-03                      0                              0  299.548415   \n",
       "2017-11-06                      0                              0  305.910012   \n",
       "2017-11-07                      0                              0  301.842300   \n",
       "2017-11-08                      0                              0  306.634741   \n",
       "2017-11-09                      0                              0  303.907229   \n",
       "2017-11-10                      0                              0  303.088822   \n",
       "2017-11-13                      0                              0  303.366486   \n",
       "2017-11-14                      0                              0  317.474216   \n",
       "2017-11-15                      0                              0  308.246135   \n",
       "2017-11-16                      0                              0  311.798508   \n",
       "2017-11-17                      0                              0  314.420964   \n",
       "2017-11-20                      0                              0  315.999310   \n",
       "2017-11-21                      0                              0  306.138106   \n",
       "\n",
       "            target  next_day_open  true_grow  model_predict  \n",
       "date                                                         \n",
       "2017-01-03  226.99         214.75          1            0.0  \n",
       "2017-01-04  226.75         226.42          1            1.0  \n",
       "2017-01-05  229.01         226.93          1            1.0  \n",
       "2017-01-06  231.28         228.97          1            1.0  \n",
       "2017-01-09  229.87         232.00          0            0.0  \n",
       "2017-01-10  229.73         229.07          1            0.0  \n",
       "2017-01-11  229.59         229.06          1            1.0  \n",
       "2017-01-12  237.75         230.00          1            1.0  \n",
       "2017-01-13  235.58         236.70          0            1.0  \n",
       "2017-01-17  238.36         236.65          1            1.0  \n",
       "2017-01-18  243.76         247.25          0            1.0  \n",
       "2017-01-19  244.73         245.46          0            0.0  \n",
       "2017-01-20  248.92         245.85          1            0.0  \n",
       "2017-01-23  254.61         250.00          1            1.0  \n",
       "2017-01-24  254.47         257.31          0            0.0  \n",
       "2017-01-25  252.51         254.29          0            1.0  \n",
       "2017-01-26  252.95         251.38          1            1.0  \n",
       "2017-01-27  250.63         252.53          0            0.0  \n",
       "2017-01-30  251.93         249.24          1            0.0  \n",
       "2017-01-31  249.24         253.05          0            1.0  \n",
       "2017-02-01  251.55         248.34          1            1.0  \n",
       "2017-02-02  251.33         251.91          0            0.0  \n",
       "2017-02-03  257.77         251.00          1            1.0  \n",
       "2017-02-06  257.48         258.19          0            0.0  \n",
       "2017-02-07  262.08         257.35          1            0.0  \n",
       "2017-02-08  269.20         266.25          1            0.0  \n",
       "2017-02-09  269.23         269.79          0            0.0  \n",
       "2017-02-10  280.60         270.74          1            0.0  \n",
       "2017-02-13  280.98         279.03          1            0.0  \n",
       "2017-02-14  279.76         280.00          0            0.0  \n",
       "...            ...            ...        ...            ...  \n",
       "2017-10-11  355.68         352.95          1            0.0  \n",
       "2017-10-12  355.57         356.98          0            0.0  \n",
       "2017-10-13  350.60         353.76          0            0.0  \n",
       "2017-10-16  355.75         350.91          1            0.0  \n",
       "2017-10-17  359.65         355.97          1            0.0  \n",
       "2017-10-18  351.81         355.56          0            0.0  \n",
       "2017-10-19  345.10         352.69          0            1.0  \n",
       "2017-10-20  337.02         349.88          0            0.0  \n",
       "2017-10-23  337.34         338.80          0            0.0  \n",
       "2017-10-24  325.84         336.70          0            1.0  \n",
       "2017-10-25  326.17         327.78          0            0.0  \n",
       "2017-10-26  320.87         319.75          1            0.0  \n",
       "2017-10-27  320.08         319.18          1            0.0  \n",
       "2017-10-30  331.53         320.23          1            0.0  \n",
       "2017-10-31  321.08         332.25          0            0.0  \n",
       "2017-11-01  299.26         300.13          0            0.0  \n",
       "2017-11-02  306.09         299.50          1            1.0  \n",
       "2017-11-03  302.78         307.00          0            0.0  \n",
       "2017-11-06  306.05         301.02          1            0.0  \n",
       "2017-11-07  304.39         305.50          0            0.0  \n",
       "2017-11-08  302.99         302.50          1            1.0  \n",
       "2017-11-09  302.99         302.50          1            1.0  \n",
       "2017-11-10  315.40         300.13          1            1.0  \n",
       "2017-11-13  308.70         315.00          0            1.0  \n",
       "2017-11-14  311.30         306.01          1            0.0  \n",
       "2017-11-15  312.50         313.99          0            0.0  \n",
       "2017-11-16  315.05         325.67          0            0.0  \n",
       "2017-11-17  308.74         313.79          0            0.0  \n",
       "2017-11-20  317.81         310.86          1            0.0  \n",
       "2017-11-21  312.60         316.77          0            0.0  \n",
       "\n",
       "[225 rows x 311 columns]"
      ]
     },
     "execution_count": 105,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "real_test_df"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 111,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "True Negative:  77 | False Positive:  39\n",
      "False Negative:  60 | True Positive:  49 \n",
      "\n",
      "Sensitivity (TP/ TP + FN):  0.449541284404\n",
      "Specificity (TN/ TN + FP):  0.663793103448 \n",
      "\n",
      "Accuracy:  0.56\n",
      "Precision:  0.556818181818\n",
      "\n",
      "\n"
     ]
    }
   ],
   "source": [
    "eval_sklearn_model(real_test_df.true_grow, real_test_df.model_predict)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 112,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Model Profit:\n",
      "Own? False\n",
      "Last Buying price: $ 302.5\n",
      "Current price: $ 316.77\n",
      "Cash? $ 74.01000000000005\n",
      "Profit: $ 74.01000000000005\n",
      "Number of Transactins: 104\n",
      "Cost of transactions: $ 520\n",
      "\n",
      "Best Case Profit:\n",
      "Own? False\n",
      "Last Buying price: $ 310.86\n",
      "Current price: $ 316.77\n",
      "Cash? $ 620.0799999999999\n",
      "Profit: $ 620.0799999999999\n",
      "Number of Transactins: 118\n",
      "Cost of transactions: $ 590\n"
     ]
    }
   ],
   "source": [
    "print('Model Profit:')\n",
    "profit_calculator(real_test_df, 'model_predict')\n",
    "print('\\nBest Case Profit:')\n",
    "profit_calculator(real_test_df, 'true_grow')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.6.3"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}