720 Final Synthetic Control.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Synthetic Control for the Purchase of Twitter\n",
    "\n",
    "`Objective`: Use the Synthetic Control method to effectively demonstrate the impact of Elon Musk purchasing Twitter on overall sentiment towards the company and on Twitter's stock price."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "# import libraries\n",
    "import pandas as pd\n",
    "import numpy as np\n",
    "import csv\n",
    "import snscrape.modules.twitter as sntwitter\n",
    "from vaderSentiment.vaderSentiment import SentimentIntensityAnalyzer\n",
    "from datetime import datetime, timedelta\n",
    "import datetime as dt\n",
    "import seaborn as sns\n",
    "import matplotlib.pyplot as plt\n",
    "from matplotlib import style\n",
    "style.use('ggplot')\n",
    "pd.set_option('display.max_colwidth', None)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Loading Tweets\n",
    "- Tweets with 'Twitter' keyword will represent Real Twitter Sentiment\n",
    "- Tweets with other keywords (Facebook, Snapchat, etc.) will combine to form Synthetic Twitter Sentiment"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Elon Musk buys twitter 04/25/2022 at 3pm\n",
    "# Start and end date are 1 day apart because we want a certain number of tweets per day\n",
    "start_date = '2022-04-18'\n",
    "end_date = '2022-04-19'"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 84,
   "metadata": {},
   "outputs": [],
   "source": [
    "# https://medium.datadriveninvestor.com/how-to-build-a-twitter-scraping-app-with-python-b3fc069a19c6\n",
    "\n",
    "# Initialize keywords\n",
    "keywords = ['Twitter', 'Facebook', 'Instagram', 'LinkedIn', 'Snapchat', 'Reddit']\n",
    "maxTweets = 2000\n",
    "\n",
    "# Iterate over all days one at a time\n",
    "for i in range(12):\n",
    "    start_date = dt.datetime.strftime(dt.datetime.strptime(start_date, '%Y-%m-%d') + timedelta(days=1), '%Y-%m-%d')\n",
    "    end_date = dt.datetime.strftime(dt.datetime.strptime(end_date, '%Y-%m-%d') + timedelta(days=1), '%Y-%m-%d')\n",
    "    \n",
    "    # Iterate over chosen keywords\n",
    "    for keyword in keywords:\n",
    "        # Open/create a file to append data to\n",
    "        csvFile = open(keyword + '_' + start_date + '.csv', 'a', newline='', encoding='utf8')\n",
    "        csvFile.truncate(0)\n",
    "        csvWriter = csv.writer(csvFile)\n",
    "\n",
    "        # Initialize scraper object\n",
    "        scraper = sntwitter.TwitterSearchScraper(keyword + ' lang:en since:' +  start_date + ' until:' + end_date)\n",
    "\n",
    "        # Iterate over tweet info and write to csv\n",
    "        for i, tweet in enumerate(scraper.get_items()):\n",
    "            if i >= maxTweets :\n",
    "                break\n",
    "            csvWriter.writerow([tweet.id, tweet.date, tweet.user, tweet.content, tweet.likeCount,\n",
    "                                tweet.retweetCount, tweet.quoteCount, tweet.replyCount])\n",
    "        csvFile.close()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 84,
   "metadata": {},
   "outputs": [],
   "source": [
    "# List of dates and keywords needed for import\n",
    "keywords = ['Twitter', 'Facebook', 'Instagram', 'LinkedIn', 'Snapchat', 'Reddit']\n",
    "dates = ['2022-04-19', '2022-04-20', '2022-04-21', '2022-04-22', '2022-04-23', '2022-04-24', '2022-04-25', '2022-04-26', \n",
    "        '2022-04-27', '2022-04-28', '2022-04-29', '2022-04-30']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 85,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(144000, 9)\n"
     ]
    },
    {
     "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>id</th>\n",
       "      <th>date</th>\n",
       "      <th>user</th>\n",
       "      <th>content</th>\n",
       "      <th>likes</th>\n",
       "      <th>retweets</th>\n",
       "      <th>quotes</th>\n",
       "      <th>replies</th>\n",
       "      <th>company</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1516567316531748866</td>\n",
       "      <td>2022-04-19 23:59:59+00:00</td>\n",
       "      <td>https://twitter.com/Jazzboneplyr1</td>\n",
       "      <td>@Alex343 @eb454 @paulkrugman @Twitter we were speaking about harassment of people wearing masks</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>Twitter</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1516567316292644864</td>\n",
       "      <td>2022-04-19 23:59:59+00:00</td>\n",
       "      <td>https://twitter.com/SandaraMary</td>\n",
       "      <td>@mhdksafa @Twitter Support your views!</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>Twitter</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>1516567316204560389</td>\n",
       "      <td>2022-04-19 23:59:59+00:00</td>\n",
       "      <td>https://twitter.com/ppotagonist</td>\n",
       "      <td>Incredibly important tweet for those of you who thrive on tearing apart who you identify as “twitter main character of the day” \\n\\nStop 🛑 https://t.co/2Awn7NPBsk</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>Twitter</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>1516567315957133317</td>\n",
       "      <td>2022-04-19 23:59:59+00:00</td>\n",
       "      <td>https://twitter.com/RayTesla_</td>\n",
       "      <td>@Brian07li @sloppytopgiver @defaultgunner @raptalksk @AnfieldAn The whole white ppl in hip hop thing is such a bs thing &amp;amp; it ONLY exists on Twitter. Hip hop is inclusive. Everybody is welcomed. Just respect the craft</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>Twitter</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>1516567311817318406</td>\n",
       "      <td>2022-04-19 23:59:58+00:00</td>\n",
       "      <td>https://twitter.com/Kasperinos</td>\n",
       "      <td>what’s up twitter</td>\n",
       "      <td>17</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>3</td>\n",
       "      <td>Twitter</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                    id                       date  \\\n",
       "0  1516567316531748866  2022-04-19 23:59:59+00:00   \n",
       "1  1516567316292644864  2022-04-19 23:59:59+00:00   \n",
       "2  1516567316204560389  2022-04-19 23:59:59+00:00   \n",
       "3  1516567315957133317  2022-04-19 23:59:59+00:00   \n",
       "4  1516567311817318406  2022-04-19 23:59:58+00:00   \n",
       "\n",
       "                                user  \\\n",
       "0  https://twitter.com/Jazzboneplyr1   \n",
       "1    https://twitter.com/SandaraMary   \n",
       "2    https://twitter.com/ppotagonist   \n",
       "3      https://twitter.com/RayTesla_   \n",
       "4     https://twitter.com/Kasperinos   \n",
       "\n",
       "                                                                                                                                                                                                                        content  \\\n",
       "0                                                                                                                               @Alex343 @eb454 @paulkrugman @Twitter we were speaking about harassment of people wearing masks   \n",
       "1                                                                                                                                                                                        @mhdksafa @Twitter Support your views!   \n",
       "2                                                            Incredibly important tweet for those of you who thrive on tearing apart who you identify as “twitter main character of the day” \\n\\nStop 🛑 https://t.co/2Awn7NPBsk   \n",
       "3  @Brian07li @sloppytopgiver @defaultgunner @raptalksk @AnfieldAn The whole white ppl in hip hop thing is such a bs thing &amp; it ONLY exists on Twitter. Hip hop is inclusive. Everybody is welcomed. Just respect the craft   \n",
       "4                                                                                                                                                                                                             what’s up twitter   \n",
       "\n",
       "   likes  retweets  quotes  replies  company  \n",
       "0      1         0       0        2  Twitter  \n",
       "1      0         0       0        0  Twitter  \n",
       "2      0         0       0        1  Twitter  \n",
       "3      0         0       0        1  Twitter  \n",
       "4     17         0       0        3  Twitter  "
      ]
     },
     "execution_count": 85,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Load all CSVs into one df \n",
    "df = pd.DataFrame()\n",
    "for keyword in keywords:\n",
    "    for date in dates:\n",
    "        df1 = pd.read_csv(keyword + '_' + date + '.csv', header=None).assign(company=keyword)\n",
    "        df = pd.concat([df, df1])\n",
    "df.reset_index(drop=True, inplace=True)\n",
    "df.columns = ['id', 'date', 'user', 'content', 'likes', 'retweets', 'quotes', 'replies', 'company']\n",
    "print(df.shape)\n",
    "df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 86,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Twitter      24000\n",
       "Facebook     24000\n",
       "Instagram    24000\n",
       "LinkedIn     24000\n",
       "Snapchat     24000\n",
       "Reddit       24000\n",
       "Name: company, dtype: int64"
      ]
     },
     "execution_count": 86,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Check number of tweets per company\n",
    "df.company.value_counts()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Calculating Sentiment Scores using VaderSentiment\n",
    "- These scores represent the attitude of the tweet\n",
    "- The Compound Scores will be used as the values for Synthetic Control"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 87,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[{'neg': 0.226, 'neu': 0.774, 'pos': 0.0, 'compound': -0.5423},\n",
       " {'neg': 0.0, 'neu': 0.572, 'pos': 0.428, 'compound': 0.4574},\n",
       " {'neg': 0.149, 'neu': 0.78, 'pos': 0.071, 'compound': -0.3197},\n",
       " {'neg': 0.0, 'neu': 0.857, 'pos': 0.143, 'compound': 0.6705},\n",
       " {'neg': 0.0, 'neu': 1.0, 'pos': 0.0, 'compound': 0.0}]"
      ]
     },
     "execution_count": 87,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Calculate sentiment scores and create list of dictionaries\n",
    "analyser = SentimentIntensityAnalyzer()\n",
    "scores = [analyser.polarity_scores(tweet) for tweet in df['content']]\n",
    "scores[0:5]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 88,
   "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>id</th>\n",
       "      <th>date</th>\n",
       "      <th>user</th>\n",
       "      <th>content</th>\n",
       "      <th>likes</th>\n",
       "      <th>retweets</th>\n",
       "      <th>quotes</th>\n",
       "      <th>replies</th>\n",
       "      <th>company</th>\n",
       "      <th>neg</th>\n",
       "      <th>neu</th>\n",
       "      <th>pos</th>\n",
       "      <th>compound</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1516567316531748866</td>\n",
       "      <td>2022-04-19 23:59:59+00:00</td>\n",
       "      <td>https://twitter.com/Jazzboneplyr1</td>\n",
       "      <td>@Alex343 @eb454 @paulkrugman @Twitter we were speaking about harassment of people wearing masks</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>Twitter</td>\n",
       "      <td>0.226</td>\n",
       "      <td>0.774</td>\n",
       "      <td>0.000</td>\n",
       "      <td>-0.5423</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1516567316292644864</td>\n",
       "      <td>2022-04-19 23:59:59+00:00</td>\n",
       "      <td>https://twitter.com/SandaraMary</td>\n",
       "      <td>@mhdksafa @Twitter Support your views!</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>Twitter</td>\n",
       "      <td>0.000</td>\n",
       "      <td>0.572</td>\n",
       "      <td>0.428</td>\n",
       "      <td>0.4574</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>1516567316204560389</td>\n",
       "      <td>2022-04-19 23:59:59+00:00</td>\n",
       "      <td>https://twitter.com/ppotagonist</td>\n",
       "      <td>Incredibly important tweet for those of you who thrive on tearing apart who you identify as “twitter main character of the day” \\n\\nStop 🛑 https://t.co/2Awn7NPBsk</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>Twitter</td>\n",
       "      <td>0.149</td>\n",
       "      <td>0.780</td>\n",
       "      <td>0.071</td>\n",
       "      <td>-0.3197</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>1516567315957133317</td>\n",
       "      <td>2022-04-19 23:59:59+00:00</td>\n",
       "      <td>https://twitter.com/RayTesla_</td>\n",
       "      <td>@Brian07li @sloppytopgiver @defaultgunner @raptalksk @AnfieldAn The whole white ppl in hip hop thing is such a bs thing &amp;amp; it ONLY exists on Twitter. Hip hop is inclusive. Everybody is welcomed. Just respect the craft</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>Twitter</td>\n",
       "      <td>0.000</td>\n",
       "      <td>0.857</td>\n",
       "      <td>0.143</td>\n",
       "      <td>0.6705</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>1516567311817318406</td>\n",
       "      <td>2022-04-19 23:59:58+00:00</td>\n",
       "      <td>https://twitter.com/Kasperinos</td>\n",
       "      <td>what’s up twitter</td>\n",
       "      <td>17</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>3</td>\n",
       "      <td>Twitter</td>\n",
       "      <td>0.000</td>\n",
       "      <td>1.000</td>\n",
       "      <td>0.000</td>\n",
       "      <td>0.0000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                    id                       date  \\\n",
       "0  1516567316531748866  2022-04-19 23:59:59+00:00   \n",
       "1  1516567316292644864  2022-04-19 23:59:59+00:00   \n",
       "2  1516567316204560389  2022-04-19 23:59:59+00:00   \n",
       "3  1516567315957133317  2022-04-19 23:59:59+00:00   \n",
       "4  1516567311817318406  2022-04-19 23:59:58+00:00   \n",
       "\n",
       "                                user  \\\n",
       "0  https://twitter.com/Jazzboneplyr1   \n",
       "1    https://twitter.com/SandaraMary   \n",
       "2    https://twitter.com/ppotagonist   \n",
       "3      https://twitter.com/RayTesla_   \n",
       "4     https://twitter.com/Kasperinos   \n",
       "\n",
       "                                                                                                                                                                                                                        content  \\\n",
       "0                                                                                                                               @Alex343 @eb454 @paulkrugman @Twitter we were speaking about harassment of people wearing masks   \n",
       "1                                                                                                                                                                                        @mhdksafa @Twitter Support your views!   \n",
       "2                                                            Incredibly important tweet for those of you who thrive on tearing apart who you identify as “twitter main character of the day” \\n\\nStop 🛑 https://t.co/2Awn7NPBsk   \n",
       "3  @Brian07li @sloppytopgiver @defaultgunner @raptalksk @AnfieldAn The whole white ppl in hip hop thing is such a bs thing &amp; it ONLY exists on Twitter. Hip hop is inclusive. Everybody is welcomed. Just respect the craft   \n",
       "4                                                                                                                                                                                                             what’s up twitter   \n",
       "\n",
       "   likes  retweets  quotes  replies  company    neg    neu    pos  compound  \n",
       "0      1         0       0        2  Twitter  0.226  0.774  0.000   -0.5423  \n",
       "1      0         0       0        0  Twitter  0.000  0.572  0.428    0.4574  \n",
       "2      0         0       0        1  Twitter  0.149  0.780  0.071   -0.3197  \n",
       "3      0         0       0        1  Twitter  0.000  0.857  0.143    0.6705  \n",
       "4     17         0       0        3  Twitter  0.000  1.000  0.000    0.0000  "
      ]
     },
     "execution_count": 88,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Add sentiment scores for each tweet to the df\n",
    "df = pd.concat([df, pd.DataFrame(scores)], axis=1)\n",
    "df.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Structuring Data for \"By-Hand\" Synthetic Control\n",
    "- Creating Synthetic Twitter Sentiment by simply taking the average Compound Sentiment Scores of all the other companies combined\n",
    "- This is a simpler version of Synthetic Control to use as a baseline"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 89,
   "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>id</th>\n",
       "      <th>date</th>\n",
       "      <th>user</th>\n",
       "      <th>content</th>\n",
       "      <th>likes</th>\n",
       "      <th>retweets</th>\n",
       "      <th>quotes</th>\n",
       "      <th>replies</th>\n",
       "      <th>company</th>\n",
       "      <th>neg</th>\n",
       "      <th>neu</th>\n",
       "      <th>pos</th>\n",
       "      <th>compound</th>\n",
       "      <th>Date</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1516567316531748866</td>\n",
       "      <td>2022-04-19 23:59:59+00:00</td>\n",
       "      <td>https://twitter.com/Jazzboneplyr1</td>\n",
       "      <td>@Alex343 @eb454 @paulkrugman @Twitter we were speaking about harassment of people wearing masks</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>Twitter</td>\n",
       "      <td>0.226</td>\n",
       "      <td>0.774</td>\n",
       "      <td>0.000</td>\n",
       "      <td>-0.5423</td>\n",
       "      <td>2022-04-19</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1516567316292644864</td>\n",
       "      <td>2022-04-19 23:59:59+00:00</td>\n",
       "      <td>https://twitter.com/SandaraMary</td>\n",
       "      <td>@mhdksafa @Twitter Support your views!</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>Twitter</td>\n",
       "      <td>0.000</td>\n",
       "      <td>0.572</td>\n",
       "      <td>0.428</td>\n",
       "      <td>0.4574</td>\n",
       "      <td>2022-04-19</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>1516567316204560389</td>\n",
       "      <td>2022-04-19 23:59:59+00:00</td>\n",
       "      <td>https://twitter.com/ppotagonist</td>\n",
       "      <td>Incredibly important tweet for those of you who thrive on tearing apart who you identify as “twitter main character of the day” \\n\\nStop 🛑 https://t.co/2Awn7NPBsk</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>Twitter</td>\n",
       "      <td>0.149</td>\n",
       "      <td>0.780</td>\n",
       "      <td>0.071</td>\n",
       "      <td>-0.3197</td>\n",
       "      <td>2022-04-19</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>1516567315957133317</td>\n",
       "      <td>2022-04-19 23:59:59+00:00</td>\n",
       "      <td>https://twitter.com/RayTesla_</td>\n",
       "      <td>@Brian07li @sloppytopgiver @defaultgunner @raptalksk @AnfieldAn The whole white ppl in hip hop thing is such a bs thing &amp;amp; it ONLY exists on Twitter. Hip hop is inclusive. Everybody is welcomed. Just respect the craft</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>Twitter</td>\n",
       "      <td>0.000</td>\n",
       "      <td>0.857</td>\n",
       "      <td>0.143</td>\n",
       "      <td>0.6705</td>\n",
       "      <td>2022-04-19</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>1516567311817318406</td>\n",
       "      <td>2022-04-19 23:59:58+00:00</td>\n",
       "      <td>https://twitter.com/Kasperinos</td>\n",
       "      <td>what’s up twitter</td>\n",
       "      <td>17</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>3</td>\n",
       "      <td>Twitter</td>\n",
       "      <td>0.000</td>\n",
       "      <td>1.000</td>\n",
       "      <td>0.000</td>\n",
       "      <td>0.0000</td>\n",
       "      <td>2022-04-19</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                    id                       date  \\\n",
       "0  1516567316531748866  2022-04-19 23:59:59+00:00   \n",
       "1  1516567316292644864  2022-04-19 23:59:59+00:00   \n",
       "2  1516567316204560389  2022-04-19 23:59:59+00:00   \n",
       "3  1516567315957133317  2022-04-19 23:59:59+00:00   \n",
       "4  1516567311817318406  2022-04-19 23:59:58+00:00   \n",
       "\n",
       "                                user  \\\n",
       "0  https://twitter.com/Jazzboneplyr1   \n",
       "1    https://twitter.com/SandaraMary   \n",
       "2    https://twitter.com/ppotagonist   \n",
       "3      https://twitter.com/RayTesla_   \n",
       "4     https://twitter.com/Kasperinos   \n",
       "\n",
       "                                                                                                                                                                                                                        content  \\\n",
       "0                                                                                                                               @Alex343 @eb454 @paulkrugman @Twitter we were speaking about harassment of people wearing masks   \n",
       "1                                                                                                                                                                                        @mhdksafa @Twitter Support your views!   \n",
       "2                                                            Incredibly important tweet for those of you who thrive on tearing apart who you identify as “twitter main character of the day” \\n\\nStop 🛑 https://t.co/2Awn7NPBsk   \n",
       "3  @Brian07li @sloppytopgiver @defaultgunner @raptalksk @AnfieldAn The whole white ppl in hip hop thing is such a bs thing &amp; it ONLY exists on Twitter. Hip hop is inclusive. Everybody is welcomed. Just respect the craft   \n",
       "4                                                                                                                                                                                                             what’s up twitter   \n",
       "\n",
       "   likes  retweets  quotes  replies  company    neg    neu    pos  compound  \\\n",
       "0      1         0       0        2  Twitter  0.226  0.774  0.000   -0.5423   \n",
       "1      0         0       0        0  Twitter  0.000  0.572  0.428    0.4574   \n",
       "2      0         0       0        1  Twitter  0.149  0.780  0.071   -0.3197   \n",
       "3      0         0       0        1  Twitter  0.000  0.857  0.143    0.6705   \n",
       "4     17         0       0        3  Twitter  0.000  1.000  0.000    0.0000   \n",
       "\n",
       "         Date  \n",
       "0  2022-04-19  \n",
       "1  2022-04-19  \n",
       "2  2022-04-19  \n",
       "3  2022-04-19  \n",
       "4  2022-04-19  "
      ]
     },
     "execution_count": 89,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Strip date column into correct string format\n",
    "df['Date'] = [date[0:10] for date in df['date']]\n",
    "df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 90,
   "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>neg</th>\n",
       "      <th>pos</th>\n",
       "      <th>neu</th>\n",
       "      <th>compound</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Date</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>2022-04-19</th>\n",
       "      <td>0.077192</td>\n",
       "      <td>0.116656</td>\n",
       "      <td>0.806155</td>\n",
       "      <td>0.124487</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2022-04-20</th>\n",
       "      <td>0.078219</td>\n",
       "      <td>0.118613</td>\n",
       "      <td>0.803159</td>\n",
       "      <td>0.115027</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2022-04-21</th>\n",
       "      <td>0.080174</td>\n",
       "      <td>0.115527</td>\n",
       "      <td>0.804297</td>\n",
       "      <td>0.110883</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2022-04-22</th>\n",
       "      <td>0.081254</td>\n",
       "      <td>0.114841</td>\n",
       "      <td>0.803901</td>\n",
       "      <td>0.105457</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2022-04-23</th>\n",
       "      <td>0.086941</td>\n",
       "      <td>0.121575</td>\n",
       "      <td>0.791491</td>\n",
       "      <td>0.099592</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2022-04-24</th>\n",
       "      <td>0.081282</td>\n",
       "      <td>0.114417</td>\n",
       "      <td>0.804308</td>\n",
       "      <td>0.094618</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2022-04-25</th>\n",
       "      <td>0.068813</td>\n",
       "      <td>0.119156</td>\n",
       "      <td>0.812038</td>\n",
       "      <td>0.140242</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2022-04-26</th>\n",
       "      <td>0.075836</td>\n",
       "      <td>0.116013</td>\n",
       "      <td>0.808158</td>\n",
       "      <td>0.117697</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2022-04-27</th>\n",
       "      <td>0.076698</td>\n",
       "      <td>0.118579</td>\n",
       "      <td>0.804717</td>\n",
       "      <td>0.123595</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2022-04-28</th>\n",
       "      <td>0.074909</td>\n",
       "      <td>0.118820</td>\n",
       "      <td>0.806273</td>\n",
       "      <td>0.119560</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2022-04-29</th>\n",
       "      <td>0.080138</td>\n",
       "      <td>0.115161</td>\n",
       "      <td>0.804706</td>\n",
       "      <td>0.097044</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2022-04-30</th>\n",
       "      <td>0.078179</td>\n",
       "      <td>0.118354</td>\n",
       "      <td>0.803474</td>\n",
       "      <td>0.119938</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                 neg       pos       neu  compound\n",
       "Date                                              \n",
       "2022-04-19  0.077192  0.116656  0.806155  0.124487\n",
       "2022-04-20  0.078219  0.118613  0.803159  0.115027\n",
       "2022-04-21  0.080174  0.115527  0.804297  0.110883\n",
       "2022-04-22  0.081254  0.114841  0.803901  0.105457\n",
       "2022-04-23  0.086941  0.121575  0.791491  0.099592\n",
       "2022-04-24  0.081282  0.114417  0.804308  0.094618\n",
       "2022-04-25  0.068813  0.119156  0.812038  0.140242\n",
       "2022-04-26  0.075836  0.116013  0.808158  0.117697\n",
       "2022-04-27  0.076698  0.118579  0.804717  0.123595\n",
       "2022-04-28  0.074909  0.118820  0.806273  0.119560\n",
       "2022-04-29  0.080138  0.115161  0.804706  0.097044\n",
       "2022-04-30  0.078179  0.118354  0.803474  0.119938"
      ]
     },
     "execution_count": 90,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Group by company and date and aggregate sentiment scores\n",
    "comp_date_scores = df.groupby(['company', 'Date'])[['neg', 'pos', 'neu', 'compound']].mean()\n",
    "# Segment just Twitter\n",
    "twitter = comp_date_scores.loc['Twitter']\n",
    "twitter"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 91,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Segment other companies\n",
    "others = comp_date_scores.loc[['Facebook', 'Instagram', 'LinkedIn', 'Snapchat', 'Reddit']]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 92,
   "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>neg</th>\n",
       "      <th>pos</th>\n",
       "      <th>neu</th>\n",
       "      <th>compound</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Date</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>2022-04-19</th>\n",
       "      <td>0.067979</td>\n",
       "      <td>0.103167</td>\n",
       "      <td>0.828855</td>\n",
       "      <td>0.114296</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2022-04-20</th>\n",
       "      <td>0.066546</td>\n",
       "      <td>0.121766</td>\n",
       "      <td>0.811692</td>\n",
       "      <td>0.118524</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2022-04-21</th>\n",
       "      <td>0.067966</td>\n",
       "      <td>0.102867</td>\n",
       "      <td>0.829169</td>\n",
       "      <td>0.106310</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2022-04-22</th>\n",
       "      <td>0.068818</td>\n",
       "      <td>0.104069</td>\n",
       "      <td>0.827115</td>\n",
       "      <td>0.110661</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2022-04-23</th>\n",
       "      <td>0.068436</td>\n",
       "      <td>0.103673</td>\n",
       "      <td>0.827889</td>\n",
       "      <td>0.106546</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2022-04-24</th>\n",
       "      <td>0.065002</td>\n",
       "      <td>0.106860</td>\n",
       "      <td>0.828143</td>\n",
       "      <td>0.124492</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2022-04-25</th>\n",
       "      <td>0.066829</td>\n",
       "      <td>0.103137</td>\n",
       "      <td>0.830031</td>\n",
       "      <td>0.105216</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2022-04-26</th>\n",
       "      <td>0.067345</td>\n",
       "      <td>0.106500</td>\n",
       "      <td>0.826155</td>\n",
       "      <td>0.119198</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2022-04-27</th>\n",
       "      <td>0.066682</td>\n",
       "      <td>0.101273</td>\n",
       "      <td>0.832049</td>\n",
       "      <td>0.114871</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2022-04-28</th>\n",
       "      <td>0.063645</td>\n",
       "      <td>0.100310</td>\n",
       "      <td>0.836041</td>\n",
       "      <td>0.117515</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2022-04-29</th>\n",
       "      <td>0.064897</td>\n",
       "      <td>0.100147</td>\n",
       "      <td>0.834958</td>\n",
       "      <td>0.110944</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2022-04-30</th>\n",
       "      <td>0.068343</td>\n",
       "      <td>0.101668</td>\n",
       "      <td>0.829985</td>\n",
       "      <td>0.107127</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                 neg       pos       neu  compound\n",
       "Date                                              \n",
       "2022-04-19  0.067979  0.103167  0.828855  0.114296\n",
       "2022-04-20  0.066546  0.121766  0.811692  0.118524\n",
       "2022-04-21  0.067966  0.102867  0.829169  0.106310\n",
       "2022-04-22  0.068818  0.104069  0.827115  0.110661\n",
       "2022-04-23  0.068436  0.103673  0.827889  0.106546\n",
       "2022-04-24  0.065002  0.106860  0.828143  0.124492\n",
       "2022-04-25  0.066829  0.103137  0.830031  0.105216\n",
       "2022-04-26  0.067345  0.106500  0.826155  0.119198\n",
       "2022-04-27  0.066682  0.101273  0.832049  0.114871\n",
       "2022-04-28  0.063645  0.100310  0.836041  0.117515\n",
       "2022-04-29  0.064897  0.100147  0.834958  0.110944\n",
       "2022-04-30  0.068343  0.101668  0.829985  0.107127"
      ]
     },
     "execution_count": 92,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Group others by date and aggregate sentiment scores\n",
    "others_combined = others.groupby('Date')[['neg', 'pos', 'neu', 'compound']].mean()\n",
    "others_combined"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 93,
   "metadata": {
    "scrolled": false
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 864x432 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# \"By-hand\" Synthetic Control using mean of Twitter vs mean of others combined\n",
    "plt.figure(figsize=(12,6))\n",
    "sns.lineplot(data=twitter, x='Date', y='compound', color='blue', label='Twitter')\n",
    "sns.lineplot(data=others_combined, x='Date', y='compound', color='green', label='Average of Others')\n",
    "plt.vlines(x='2022-04-25', ymin=.06, ymax=.20, linestyle=\":\", lw=2, label='Elon Musk Buys Twitter', color='black')\n",
    "plt.title('Twitter Compound Sentiment Score vs. Average of Others')\n",
    "plt.ylabel('Sentiment Score')\n",
    "plt.xlabel('Date')\n",
    "plt.xticks(rotation=70);\n",
    "plt.show();"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Not a great representation of the impact, but we can at least see the Synthetic and Real Twitter following each other somewhat closely pre-purchase.  The goal is to see the Real and Synthetic follow each other closely pre-purchase, and then diverge after the purchase."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Synthetic Control for Compound Sentiment Score (Round 1)\n",
    "- Synthetic Twitter created using Linear Regression to calculate weights for each company\n",
    "- The Synthetic sentiment score (combination of all others) will be plotted against the sentiment for only Twitter"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 94,
   "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>company</th>\n",
       "      <th>Date</th>\n",
       "      <th>neg</th>\n",
       "      <th>pos</th>\n",
       "      <th>neu</th>\n",
       "      <th>compound</th>\n",
       "      <th>Day</th>\n",
       "      <th>Twitter</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>Facebook</td>\n",
       "      <td>2022-04-19</td>\n",
       "      <td>0.074993</td>\n",
       "      <td>0.101807</td>\n",
       "      <td>0.823210</td>\n",
       "      <td>0.109483</td>\n",
       "      <td>19</td>\n",
       "      <td>False</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>Facebook</td>\n",
       "      <td>2022-04-20</td>\n",
       "      <td>0.071859</td>\n",
       "      <td>0.099841</td>\n",
       "      <td>0.828300</td>\n",
       "      <td>0.104920</td>\n",
       "      <td>20</td>\n",
       "      <td>False</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>Facebook</td>\n",
       "      <td>2022-04-21</td>\n",
       "      <td>0.070761</td>\n",
       "      <td>0.105739</td>\n",
       "      <td>0.823495</td>\n",
       "      <td>0.117222</td>\n",
       "      <td>21</td>\n",
       "      <td>False</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>Facebook</td>\n",
       "      <td>2022-04-22</td>\n",
       "      <td>0.069179</td>\n",
       "      <td>0.107949</td>\n",
       "      <td>0.822871</td>\n",
       "      <td>0.123297</td>\n",
       "      <td>22</td>\n",
       "      <td>False</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>Facebook</td>\n",
       "      <td>2022-04-23</td>\n",
       "      <td>0.068549</td>\n",
       "      <td>0.107415</td>\n",
       "      <td>0.824033</td>\n",
       "      <td>0.127343</td>\n",
       "      <td>23</td>\n",
       "      <td>False</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>67</th>\n",
       "      <td>Twitter</td>\n",
       "      <td>2022-04-26</td>\n",
       "      <td>0.075836</td>\n",
       "      <td>0.116013</td>\n",
       "      <td>0.808158</td>\n",
       "      <td>0.117697</td>\n",
       "      <td>26</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>68</th>\n",
       "      <td>Twitter</td>\n",
       "      <td>2022-04-27</td>\n",
       "      <td>0.076698</td>\n",
       "      <td>0.118579</td>\n",
       "      <td>0.804717</td>\n",
       "      <td>0.123595</td>\n",
       "      <td>27</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>69</th>\n",
       "      <td>Twitter</td>\n",
       "      <td>2022-04-28</td>\n",
       "      <td>0.074909</td>\n",
       "      <td>0.118820</td>\n",
       "      <td>0.806273</td>\n",
       "      <td>0.119560</td>\n",
       "      <td>28</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>70</th>\n",
       "      <td>Twitter</td>\n",
       "      <td>2022-04-29</td>\n",
       "      <td>0.080138</td>\n",
       "      <td>0.115161</td>\n",
       "      <td>0.804706</td>\n",
       "      <td>0.097044</td>\n",
       "      <td>29</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>71</th>\n",
       "      <td>Twitter</td>\n",
       "      <td>2022-04-30</td>\n",
       "      <td>0.078179</td>\n",
       "      <td>0.118354</td>\n",
       "      <td>0.803474</td>\n",
       "      <td>0.119938</td>\n",
       "      <td>30</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>72 rows × 8 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "     company        Date       neg       pos       neu  compound  Day  Twitter\n",
       "0   Facebook  2022-04-19  0.074993  0.101807  0.823210  0.109483   19    False\n",
       "1   Facebook  2022-04-20  0.071859  0.099841  0.828300  0.104920   20    False\n",
       "2   Facebook  2022-04-21  0.070761  0.105739  0.823495  0.117222   21    False\n",
       "3   Facebook  2022-04-22  0.069179  0.107949  0.822871  0.123297   22    False\n",
       "4   Facebook  2022-04-23  0.068549  0.107415  0.824033  0.127343   23    False\n",
       "..       ...         ...       ...       ...       ...       ...  ...      ...\n",
       "67   Twitter  2022-04-26  0.075836  0.116013  0.808158  0.117697   26     True\n",
       "68   Twitter  2022-04-27  0.076698  0.118579  0.804717  0.123595   27     True\n",
       "69   Twitter  2022-04-28  0.074909  0.118820  0.806273  0.119560   28     True\n",
       "70   Twitter  2022-04-29  0.080138  0.115161  0.804706  0.097044   29     True\n",
       "71   Twitter  2022-04-30  0.078179  0.118354  0.803474  0.119938   30     True\n",
       "\n",
       "[72 rows x 8 columns]"
      ]
     },
     "execution_count": 94,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# reset index and create new df\n",
    "df3 = comp_date_scores.reset_index()\n",
    "# Construct day column\n",
    "df3['Day'] = [date[8:10] for date in df3['Date']]\n",
    "df3['Day'] = df3['Day'].astype(int)\n",
    "# Add boolean column for Twitter or not\n",
    "df3['Twitter'] = np.where(df3['company'] == 'Twitter', True, False)\n",
    "df3"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 95,
   "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>company</th>\n",
       "      <th>Facebook</th>\n",
       "      <th>Instagram</th>\n",
       "      <th>LinkedIn</th>\n",
       "      <th>Reddit</th>\n",
       "      <th>Snapchat</th>\n",
       "      <th>Twitter</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th></th>\n",
       "      <th>Day</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 rowspan=\"6\" valign=\"top\">neg</th>\n",
       "      <th>19</th>\n",
       "      <td>0.074993</td>\n",
       "      <td>0.058905</td>\n",
       "      <td>0.053629</td>\n",
       "      <td>0.070246</td>\n",
       "      <td>0.082121</td>\n",
       "      <td>0.077192</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>20</th>\n",
       "      <td>0.071859</td>\n",
       "      <td>0.053891</td>\n",
       "      <td>0.044284</td>\n",
       "      <td>0.076901</td>\n",
       "      <td>0.085793</td>\n",
       "      <td>0.078219</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>21</th>\n",
       "      <td>0.070761</td>\n",
       "      <td>0.056804</td>\n",
       "      <td>0.048032</td>\n",
       "      <td>0.086847</td>\n",
       "      <td>0.077386</td>\n",
       "      <td>0.080174</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>22</th>\n",
       "      <td>0.069179</td>\n",
       "      <td>0.056787</td>\n",
       "      <td>0.046276</td>\n",
       "      <td>0.094155</td>\n",
       "      <td>0.077692</td>\n",
       "      <td>0.081254</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>23</th>\n",
       "      <td>0.068549</td>\n",
       "      <td>0.052056</td>\n",
       "      <td>0.047593</td>\n",
       "      <td>0.088407</td>\n",
       "      <td>0.085577</td>\n",
       "      <td>0.086941</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>24</th>\n",
       "      <td>0.068891</td>\n",
       "      <td>0.058554</td>\n",
       "      <td>0.041865</td>\n",
       "      <td>0.075534</td>\n",
       "      <td>0.080165</td>\n",
       "      <td>0.081282</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th rowspan=\"6\" valign=\"top\">pos</th>\n",
       "      <th>19</th>\n",
       "      <td>0.101807</td>\n",
       "      <td>0.105516</td>\n",
       "      <td>0.111208</td>\n",
       "      <td>0.093052</td>\n",
       "      <td>0.104254</td>\n",
       "      <td>0.116656</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>20</th>\n",
       "      <td>0.099841</td>\n",
       "      <td>0.095150</td>\n",
       "      <td>0.214449</td>\n",
       "      <td>0.098497</td>\n",
       "      <td>0.100893</td>\n",
       "      <td>0.118613</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>21</th>\n",
       "      <td>0.105739</td>\n",
       "      <td>0.098811</td>\n",
       "      <td>0.113702</td>\n",
       "      <td>0.093055</td>\n",
       "      <td>0.103027</td>\n",
       "      <td>0.115527</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>22</th>\n",
       "      <td>0.107949</td>\n",
       "      <td>0.099733</td>\n",
       "      <td>0.120908</td>\n",
       "      <td>0.094523</td>\n",
       "      <td>0.097229</td>\n",
       "      <td>0.114841</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>23</th>\n",
       "      <td>0.107415</td>\n",
       "      <td>0.103859</td>\n",
       "      <td>0.115346</td>\n",
       "      <td>0.089333</td>\n",
       "      <td>0.102415</td>\n",
       "      <td>0.121575</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>24</th>\n",
       "      <td>0.109761</td>\n",
       "      <td>0.102502</td>\n",
       "      <td>0.120067</td>\n",
       "      <td>0.098808</td>\n",
       "      <td>0.103161</td>\n",
       "      <td>0.114417</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th rowspan=\"6\" valign=\"top\">compound</th>\n",
       "      <th>19</th>\n",
       "      <td>0.109483</td>\n",
       "      <td>0.138275</td>\n",
       "      <td>0.178731</td>\n",
       "      <td>0.066992</td>\n",
       "      <td>0.078001</td>\n",
       "      <td>0.124487</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>20</th>\n",
       "      <td>0.104920</td>\n",
       "      <td>0.121838</td>\n",
       "      <td>0.235700</td>\n",
       "      <td>0.071814</td>\n",
       "      <td>0.058347</td>\n",
       "      <td>0.115027</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>21</th>\n",
       "      <td>0.117222</td>\n",
       "      <td>0.135306</td>\n",
       "      <td>0.184709</td>\n",
       "      <td>0.019024</td>\n",
       "      <td>0.075287</td>\n",
       "      <td>0.110883</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>22</th>\n",
       "      <td>0.123297</td>\n",
       "      <td>0.124281</td>\n",
       "      <td>0.222596</td>\n",
       "      <td>0.014303</td>\n",
       "      <td>0.068830</td>\n",
       "      <td>0.105457</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>23</th>\n",
       "      <td>0.127343</td>\n",
       "      <td>0.153063</td>\n",
       "      <td>0.208449</td>\n",
       "      <td>-0.008732</td>\n",
       "      <td>0.052609</td>\n",
       "      <td>0.099592</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>24</th>\n",
       "      <td>0.123130</td>\n",
       "      <td>0.143012</td>\n",
       "      <td>0.232405</td>\n",
       "      <td>0.055662</td>\n",
       "      <td>0.068251</td>\n",
       "      <td>0.094618</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th rowspan=\"6\" valign=\"top\">neu</th>\n",
       "      <th>19</th>\n",
       "      <td>0.823210</td>\n",
       "      <td>0.835569</td>\n",
       "      <td>0.835167</td>\n",
       "      <td>0.836699</td>\n",
       "      <td>0.813632</td>\n",
       "      <td>0.806155</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>20</th>\n",
       "      <td>0.828300</td>\n",
       "      <td>0.850957</td>\n",
       "      <td>0.741270</td>\n",
       "      <td>0.824620</td>\n",
       "      <td>0.813312</td>\n",
       "      <td>0.803159</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>21</th>\n",
       "      <td>0.823495</td>\n",
       "      <td>0.844376</td>\n",
       "      <td>0.838278</td>\n",
       "      <td>0.820102</td>\n",
       "      <td>0.819593</td>\n",
       "      <td>0.804297</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>22</th>\n",
       "      <td>0.822871</td>\n",
       "      <td>0.843482</td>\n",
       "      <td>0.832812</td>\n",
       "      <td>0.811322</td>\n",
       "      <td>0.825087</td>\n",
       "      <td>0.803901</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>23</th>\n",
       "      <td>0.824033</td>\n",
       "      <td>0.844088</td>\n",
       "      <td>0.837061</td>\n",
       "      <td>0.822261</td>\n",
       "      <td>0.811999</td>\n",
       "      <td>0.791491</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>24</th>\n",
       "      <td>0.821352</td>\n",
       "      <td>0.838954</td>\n",
       "      <td>0.838073</td>\n",
       "      <td>0.825658</td>\n",
       "      <td>0.816680</td>\n",
       "      <td>0.804308</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "company       Facebook  Instagram  LinkedIn    Reddit  Snapchat   Twitter\n",
       "         Day                                                             \n",
       "neg      19   0.074993   0.058905  0.053629  0.070246  0.082121  0.077192\n",
       "         20   0.071859   0.053891  0.044284  0.076901  0.085793  0.078219\n",
       "         21   0.070761   0.056804  0.048032  0.086847  0.077386  0.080174\n",
       "         22   0.069179   0.056787  0.046276  0.094155  0.077692  0.081254\n",
       "         23   0.068549   0.052056  0.047593  0.088407  0.085577  0.086941\n",
       "         24   0.068891   0.058554  0.041865  0.075534  0.080165  0.081282\n",
       "pos      19   0.101807   0.105516  0.111208  0.093052  0.104254  0.116656\n",
       "         20   0.099841   0.095150  0.214449  0.098497  0.100893  0.118613\n",
       "         21   0.105739   0.098811  0.113702  0.093055  0.103027  0.115527\n",
       "         22   0.107949   0.099733  0.120908  0.094523  0.097229  0.114841\n",
       "         23   0.107415   0.103859  0.115346  0.089333  0.102415  0.121575\n",
       "         24   0.109761   0.102502  0.120067  0.098808  0.103161  0.114417\n",
       "compound 19   0.109483   0.138275  0.178731  0.066992  0.078001  0.124487\n",
       "         20   0.104920   0.121838  0.235700  0.071814  0.058347  0.115027\n",
       "         21   0.117222   0.135306  0.184709  0.019024  0.075287  0.110883\n",
       "         22   0.123297   0.124281  0.222596  0.014303  0.068830  0.105457\n",
       "         23   0.127343   0.153063  0.208449 -0.008732  0.052609  0.099592\n",
       "         24   0.123130   0.143012  0.232405  0.055662  0.068251  0.094618\n",
       "neu      19   0.823210   0.835569  0.835167  0.836699  0.813632  0.806155\n",
       "         20   0.828300   0.850957  0.741270  0.824620  0.813312  0.803159\n",
       "         21   0.823495   0.844376  0.838278  0.820102  0.819593  0.804297\n",
       "         22   0.822871   0.843482  0.832812  0.811322  0.825087  0.803901\n",
       "         23   0.824033   0.844088  0.837061  0.822261  0.811999  0.791491\n",
       "         24   0.821352   0.838954  0.838073  0.825658  0.816680  0.804308"
      ]
     },
     "execution_count": 95,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Create transposed dataframe with column for each compnay\n",
    "features = ['neg', 'pos', 'compound', 'neu']\n",
    "\n",
    "inverted = (df3.query('Day < 25') # filter pre-intervention period\n",
    "            .pivot(index='company', columns='Day')[features] # make one column per day and one row per company\n",
    "            .T) # flip the table to have one column per company\n",
    "inverted"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 96,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Extract data for synthetic and real\n",
    "y = inverted['Twitter']\n",
    "X = inverted.drop(columns='Twitter')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 97,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([ 0.863, -0.358,  0.143,  0.026,  0.312])"
      ]
     },
     "execution_count": 97,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Run Linear regression to calculate weights of each company\n",
    "from sklearn.linear_model import LinearRegression\n",
    "weights_lr = LinearRegression(fit_intercept=False).fit(X, y).coef_\n",
    "weights_lr.round(3)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 98,
   "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>Company</th>\n",
       "      <th>Weight</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>Facebook</td>\n",
       "      <td>0.863</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>Instagram</td>\n",
       "      <td>-0.358</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>LinkedIn</td>\n",
       "      <td>0.143</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>Snapchat</td>\n",
       "      <td>0.026</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>Reddit</td>\n",
       "      <td>0.312</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "     Company  Weight\n",
       "0   Facebook   0.863\n",
       "1  Instagram  -0.358\n",
       "2   LinkedIn   0.143\n",
       "3   Snapchat   0.026\n",
       "4     Reddit   0.312"
      ]
     },
     "execution_count": 98,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Construct dataframe with weights\n",
    "keywords = ['Facebook', 'Instagram', 'LinkedIn', 'Snapchat', 'Reddit']\n",
    "weights = pd.DataFrame()\n",
    "weights['Company'] = keywords\n",
    "weights['Weight'] = weights_lr.round(3)\n",
    "weights"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 99,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Creating Synthetic Twitter for compound score by dotting weights with data\n",
    "twitter_synth_lr = (df3.query(\"company != 'Twitter'\").pivot(index='Day',\n",
    "                                                            columns='company')['compound'].values.dot(weights_lr))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 100,
   "metadata": {
    "scrolled": false
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 864x432 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Plot real vs. synthetic compound sentiment score\n",
    "plt.figure(figsize=(12,6))\n",
    "plt.plot(df3.query('Twitter')['Day'], df3.query('Twitter')['compound'], label='Twitter', color='blue')\n",
    "plt.plot(df3.query('Twitter')['Day'], twitter_synth_lr, label='Synthetic Control', color='red')\n",
    "plt.vlines(x=25, ymin=.06, ymax=.2, linestyle=':', lw=2, label='Elon Musk Buys Twitter', color='black')\n",
    "plt.ylabel('Sentiment Score')\n",
    "plt.xlabel('Day of Month')\n",
    "plt.title('Twitter Compound Sentiment Score vs. Synthetic Twitter (Round 1)')\n",
    "plt.legend();"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Still not a very good representation where we can see the impact of the purchase on sentiment. Need to recalculate weights."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Synthetic Control for Compound Sentiment Score (Round 2)\n",
    "- Weights calculated differently (without extrapolation) using functions below"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 101,
   "metadata": {},
   "outputs": [],
   "source": [
    "# From Causal Inference for The Brave and True (Matheus Facure Alves)\n",
    "from typing import List\n",
    "from operator import add\n",
    "from toolz import reduce, partial\n",
    "\n",
    "def loss_w(W, X, y) -> float:\n",
    "    return np.sqrt(np.mean((y - X.dot(W))**2))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 102,
   "metadata": {},
   "outputs": [],
   "source": [
    "# From Causal Inference for The Brave and True (Matheus Facure Alves)\n",
    "from scipy.optimize import fmin_slsqp\n",
    "\n",
    "def get_w(X, y):\n",
    "    w_start = [1/X.shape[1]]*X.shape[1]\n",
    "    weights = fmin_slsqp(partial(loss_w, X=X, y=y),\n",
    "                         np.array(w_start),\n",
    "                         f_eqcons=lambda x: np.sum(x) - 1,\n",
    "                         bounds=[(0.0, 1.0)]*len(w_start),\n",
    "                         disp=False)\n",
    "    return weights"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 103,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Sum: 1.0\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "array([0.2176, 0.    , 0.1798, 0.    , 0.6026])"
      ]
     },
     "execution_count": 103,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Recalculate weights using function above\n",
    "Twitter_weights = get_w(X, y)\n",
    "print(\"Sum:\", Twitter_weights.sum())\n",
    "np.round(Twitter_weights, 4)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 104,
   "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>Company</th>\n",
       "      <th>Weight</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>Facebook</td>\n",
       "      <td>0.2176</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>Instagram</td>\n",
       "      <td>0.0000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>LinkedIn</td>\n",
       "      <td>0.1798</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>Snapchat</td>\n",
       "      <td>0.0000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>Reddit</td>\n",
       "      <td>0.6026</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "     Company  Weight\n",
       "0   Facebook  0.2176\n",
       "1  Instagram  0.0000\n",
       "2   LinkedIn  0.1798\n",
       "3   Snapchat  0.0000\n",
       "4     Reddit  0.6026"
      ]
     },
     "execution_count": 104,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Add weights to new df\n",
    "keywords = ['Facebook', 'Instagram', 'LinkedIn', 'Snapchat', 'Reddit']\n",
    "weights = pd.DataFrame()\n",
    "weights['Company'] = keywords\n",
    "weights['Weight'] = np.round(Twitter_weights, 4)\n",
    "weights"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 105,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Creating second Synthetic Twitter for compound score by dotting weights with data\n",
    "twitter_synth = df3.query(\"company != 'Twitter'\").pivot(index='Day',\n",
    "                                                        columns='company')['compound'].values.dot(Twitter_weights)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 106,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAtoAAAGHCAYAAACOFvWHAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjMuMiwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy8vihELAAAACXBIWXMAAAsTAAALEwEAmpwYAACay0lEQVR4nOzdd1hT1xsH8O9NQgKEHZYi4kDrnrhQa2U46h6t1r2r1q1V3K17a90bZ62ts7XqT61WraOiVNwVVKwiQwh7Jzm/PyKRyAqYCe/neXg0N3e8uSe5eXPuGRxjjIEQQgghhBCiVTxDB0AIIYQQQkhpRIk2IYQQQgghOkCJNiGEEEIIITpAiTYhhBBCCCE6QIk2IYQQQgghOkCJNiGEEEIIITpAiTYxWnv27IFAIDB0GKQYKlWqhEWLFhk6jEJ999138PT0NHQYpBThOA4HDhzQyb7//PNPcByH169f62T/xWFs1+SUlBS4ubkhKCjI0KEUi7auQe3bt8fGjRu1EBHRJUq0iVZxHFfoX6VKlTTeV58+fRAREaF6fODAAXAcl2c9Pz8/DBkyRAvRFy4uLg7Tp0/HJ598AnNzczg7O+PTTz/Fvn37IJPJdH780uLXX39Fq1at4ODgALFYDE9PT/Tv3x9JSUlaPc5ff/0FjuMQHh6utnzatGm4efOmVo/1MUaMGIHPPvtMo3X1de5Mjb7OS3HKqiQEAgH27Nmjtszb2xuRkZEoX758ifZZmq/Jy5cvh5eXF5o0aaJalvu1WVpaolatWlizZo3OY9GmhIQETJo0CbVr14ZYLIarqyt69eqFJ0+eqK23cOFCzJ8/v8x//o0dJdpEqyIjI1V/J0+eBADcunVLtaw4NQ8WFhZwcXHRVaj5ysrKynf569ev0ahRIxw9ehTz5s1DcHAwrl27huHDh2PVqlV48OCBXuM0VRcvXkTPnj3h7++Pv/76C/fu3cOmTZtgY2ODzMxMvcRgZWUFR0dHvRxLmwx17rKzs2HM85oZw3tKl4RCIVxdXcHjlezrurRekzMyMrBlyxZ8/fXXeZ7buHEjIiMj8fDhQ0ycOBEzZszA9u3bdR2q1kRGRuLFixdYsGABgoODcerUKaSkpMDHxwfx8fGq9Zo2bQo3Nzfs27fPgNGSIjFCdOTq1asMAHvx4gVjjLGWLVuy2bNnq56fN28eA8DOnz+vWvbpp5+yb7/9ljHGWGBgIOPz+Ywxxi5dusQAqP0NHjyYDR48OM/yS5cuMcYYi4qKYoMHD2aOjo7MysqKeXt7s8uXL6uOlbPPU6dOsZYtWzKRSMQ2bNiQ72vp3Lkzc3FxYQkJCXmey8rKYikpKar/z5gxg5UvX56ZmZmxmjVrsoMHD6qtD4CtX7+effnll8zS0pK5u7uzX375hSUkJLB+/foxKysrVrlyZXbkyBHVNi9evGAA2L59+5iPjw8zNzdnlSpVYgcOHFDb95MnT9jnn3/OxGIxE4vFrHPnziw0NFT1fO5zmuPVq1dq5y3nvJw7d461bt2aWVhYsJo1a7KzZ8+qbXf37l3WokULJhKJWLVq1djhw4eZh4cHW7hwYb7nkDHGJk6cyBo3blzg8zlCQ0NZz549ma2tLbOzs2P+/v7s3r17eV7HX3/9xRo2bMgsLCyYl5cXu337ttr5yv3Xpk0bxhhj8+fPZ1WrVlXtK+fx4cOHmaenJ7OwsGDdunVjiYmJ7OjRo6x69erMysqK9erVK0/5Hzp0iNWvX5+JRCLm4eHBJk+erHovMMZYmzZt2PDhw9mCBQuYi4sLs7e3Z4MHD1atM3/+/DxxBgYGftS5CwsLY71792b29vbMwsKC1a1bl/3222+q53///XfWqFEjJhQKmZOTExszZoxazIMHD2a+vr5s/fr1zMPDg3Ecx5KTk4v8PH3o3LlzjMfjsf/++09t+U8//cREIhGLj49njDG2ePFiVrlyZSYUCpmjoyNr164dS0tLK/J1anpe5HI5q1y5Mlu8eLHa8pSUFGZtba063x9TVgDYpk2b2IABA5iVlRWrUKECW758udrxsrOz2fz581mlSpWYSCRitWrVYlu3blU97+HhkWf/jL3/PL569Uq1blFlXJDSdE0+fvw4s7S0ZNnZ2WrLAbD9+/erLWvUqBHr2bOn6rGm1+kP9+Pr68sGDx6seuzh4cHmzp3LJkyYwOzt7ZmzszObOnUqk8lkqnUyMjLY6NGjmY2NDbOzs2OjR49mAQEBatcgTcTGxjIA7Ndff1VbPm/ePNasWbNi7YvoFyXaRGc+vKjPnTuXNW/eXPV8q1atmJOTEwsICGCMMZaWlsaEQiE7c+YMY0z9op6Zmck2btzIALDIyEgWGRnJEhISWEJCAmvdujX78ssvVcszMzNZWloaq1mzJuvZsycLCgpioaGhbNGiRUwoFLJHjx4xxt5f1D/55BN28uRJ9vz5c7UvsxxxcXGMx+MVmkDmmDZtGnNwcGA///wz+/fff9nixYsZx3HswoULqnUAMBcXF7Znzx4WGhrKxowZwywsLFiHDh1YYGAgCw0NZePGjWOWlpYsNjaWMfY+cSxXrhw7cOAAe/LkCZs9ezbjOI4FBQWpzl/FihWZj48Pu337Nrt9+zb77LPPWNWqVVlmZmaec5qjoES7Xr167MyZM+zp06ds4MCBzNbWVpUcpaWlsfLly7OOHTuyu3fvsuvXrzMvLy9mYWFR6HlatmwZs7W1ZX///XeB60RFRTEXFxc2evRodu/ePfbkyRM2btw45uDgwGJiYlSvg+M41rp1a3blyhX2+PFj5u/vz6pUqcKys7OZTCZjJ0+eZADYrVu3WGRkJIuLi2OM5Z9oW1pass8//5yFhISwP//8kzk6OjJ/f3/V67ty5QpzdnZm06dPV20XGBjI7Ozs2L59+9izZ8/Y5cuXWd26ddmAAQNU67Rp04bZ2tqySZMmscePH7MzZ84wW1tbNm/ePMYYY8nJyaxfv36sRYsWqvdvQUmmJucuMjKSOTs7M19fX3b16lUWFhbGTpw4wX7//XfGGGMhISGMz+ezSZMmsUePHrHTp08zd3d3tZgHDx7MrK2tWffu3dk///zD7t27x5KSkor8PH1ILpczNzc3tmTJErXlnTp1Yl9++SVjjLGjR48ya2tr9uuvv7KXL1+yf/75h61du7ZYibYm52XJkiWsSpUqTKFQqJbt3LmT2drastTUVMbYx5UVAObs7My2b9/OwsLC2A8//MAAsIsXL6qd17p167L//e9/7Pnz5+ynn35itra2bOfOnYwxxmJiYhifz2fr1q1T7Z+xvIl2UWVcmNJyTWaMsUmTJrGWLVvmWZ47QVYoFOzChQvMwsKC9e3bV7WOptdpTRJtOzs7tnTpUvb06VP2008/MT6fz3bv3q0Wp5OTEztx4gR7/Pgxmzp1KrO2ti52ov3s2TMGgF2/fl1t+e+//874fD5LSkoq1v6I/lCiTXTmw4v6pUuXGJ/PZ4mJiSw1NZUJhUK2atUq1qRJE8aYsgbMzMxMVYP0YVK4f/9+VS1Pbh9e/HK2dXNzy1Pb0bZtWzZx4kRVPDm1xIX5+++/GQB29OjRQtfLeU2bNm1SW969e3fWtm1b1WMAqhgYU37BAmDjxo1TLZNKpQyAqpYqJ9GeM2eO2r5btGjB+vfvzxhTJg4WFhbs7du3quejoqKYubk527t3L2OseIl27tcbGRnJAKhqtXfs2MHEYjGTSqWqde7fv88AFJpop6amsi5dujAAzNXVlXXr1o2tW7dO9YOCMWXi+2ENjUKhYFWqVGFr165VvQ4A7M6dO6p1bty4wQCwJ0+eMMbyvv9y7//DRJvP56udt7FjxzIej6dK7BljbMKECWo1px4eHmzLli1q+758+TIDoDovbdq0YXXr1lVb5+uvv1ZLboYPH66qbS+MJuduzpw5zMXFRa2GOrcBAwaoPm85Tpw4wTiOY+Hh4YwxZUJoa2vLkpOTVeto8nnKz4wZM1jNmjVVj6Ojo5lAIGCnTp1ijDG2Zs0aVq1aNZaVlVXk6y+IJuclKiqKmZmZqdXUNm/enI0dO1b1+GPKCgAbP3682rJPPvlElbA+f/6ccRzHHj9+rLbO999/z+rXr696zOfz89zR+DDRLqqMC1NarsmMMdatWzfVD7bcADCRSMTEYjETCAQMALO0tGQ3b95kjBXvOq1Jot2lSxe1ddq3b69K6lNSUphIJGLbt29XW6dx48bFSrRlMhlr3749a9KkCZPL5WrPhYSEMADswYMHGu+P6Be10SZ606JFCwiFQly5cgVXr16Fh4cHBg0ahH/++QeJiYm4ePEimjZtCrFY/NHHCgoKQlRUFOzs7GBlZaX6u3r1KkJDQ9XWbdq0aaH7Yu/ap+bX6Se3sLAwZGVl4dNPP1Vb3qZNGzx8+FBtWf369VX/d3JyAp/PR7169VTL7O3tIRQKERMTo7ZdixYt1B63bNkSjx49AgA8fPgQtWrVUmt/7OLigk8++STP8TXRoEED1f9dXV3B5/MRHR0NAHj06BFq1qwJe3t71Tp16tSBra1tofu0tLTEr7/+ihcvXmDp0qUoX748li5dik8++QSPHz8GoCy7O3fuqJWbtbU1wsPD1cqO4zi18+jm5gYAqhiLw83NTe28ubq6wtXVFU5OTmrLcsrj7du3ePnyJaZMmaIWZ8eOHQEo3ws5cp/HnGOVJEZNzt2dO3fg7e1d4Gfo4cOH+b4/GWOq9xEA1KxZE1ZWVqrHxfk85TZ48GA8fvxY1Q740KFDkEgkaN++PQDgyy+/RHZ2Njw8PDBkyBDs378fycnJWj8vLi4u6NatG3bs2KE6Dzdv3sTIkSPV9vUxZVXYtrdv3wZjDF5eXmrnb8mSJYWev/wUVcbFYarXZABIT0+Hubl5vs8tXrwYd+/exaVLl9CyZUssWrQIzZo1A1C867QmCiv3Z8+eITMzE97e3mrrtGrVSuP9y+VyDBo0CE+fPsWxY8fytNXPOQfp6enFjp3oh/GM00NKPZFIBG9vb/zxxx8QCoXw8fGBk5MTatSogT///BMXL15UfQF/LIVCgZo1a+L48eN5nrO0tFR7XNSXSLVq1cDj8fDw4UP06NGjyGN/mJAzxvIsMzMzy7Pdh8s4joNCoSj0WDk/Ago69ofHz69DVXZ2dr77FgqFeZblxJPfayqOSpUqYciQIRgyZAgWL16M6tWrY8WKFQgMDIRCoYCvr2++w1blTuR5PB74fL7qcU48RZ2z/OR37gsrj5x/f/jhB7Rt2zbP/ipUqKD6/4fnUZNyLUxh5y5n/4Up6Pncyz/8TBTn85RbzZo14eXlhX379qFJkybYt28f+vXrpxoizs3NDU+ePMGlS5dw8eJFLFy4EDNmzMDff/8Nd3f3Ql/Hh4o6L6NHj8bnn3+Ot2/fYseOHWjSpEmeJOljyqqwbXP+vX79ep7zVZLP0cd89nIz1WsyoKygkEql+T7n4uICT09PeHp64vjx46hWrRoaNmyoNmJMUddpjuPyXF/zu1YWVu6aVtIUJCsrC1999RVCQkLw559/ql1XcuScg9yVAsS4UI020SsfHx9cvHgRFy9ehK+vr2rZ8ePHcefOHfj4+BS4bc4FTS6X51n+4TIvLy88f/4cNjY2qgtuzl9xh8lycHBAx44dsXHjRiQmJuZ5Pjs7G6mpqfD09IRIJMLly5fVnr9y5Qpq165drGMW5MNh6W7cuIGaNWsCAGrXro2HDx8iNjZW9Xx0dDSePn2qOr6zszPkcrlaLV1wcHCx46hduzYePXqEhIQE1bKHDx/me36KYm9vr1Zb7OXlhYcPH8LNzS1P2RXny6Sg94s2uLi4wN3dHf/++2+eGD09PQusaSsozpLG+OG5a9y4Ma5du4bU1NR8169du3ae9+fly5fBcRxq1apV4HE+5vM0aNAg/PTTTwgJCUFwcDAGDx6s9rxIJEKHDh2wYsUK3L9/H2lpaThx4oQGr75gH54XQHmdqVixIrZv3479+/fnqc3WREnLqnHjxgCA//77L8/5q1q1arH2X1QZF5cpXpMBoFGjRhrVQDs5OWHs2LGYOHEiGGMaX6ednZ3x5s0b1ePMzEy1uz6a8PT0hFAoxLVr19SWX79+vcht09LS0LVrVzx69AhXrlxBxYoV813v/v37cHJyKvB5YniUaBO98vHxwf3793H37l1VTaCPjw8OHjwIMzOzPE0jcqtcuTIA5Zi5b9++RUpKimr5nTt38OzZM8TGxiI7Oxv9+/dH5cqV0alTJ5w7dw7h4eH4+++/sXTp0hJ9iW/evBlmZmZo3LgxfvzxRzx69AhhYWE4cOAAvLy8EBoaCktLS0yYMAFz587FL7/8gtDQUCxZsgQnT57ErFmzin+y8rFr1y78+OOPePr0KebNm4cbN25g0qRJAIB+/frByckJffr0QXBwMO7cuYO+ffvCzc0Nffr0AaC8JWttbY2AgACEhobi7NmzWLBgQbHj6NevH6ytrTFgwACEhITg5s2bGDZsGCwsLArd7rvvvsO0adNw6dIlvHjxAvfv38e0adPw4MED1d2CcePGQS6Xo3v37rh69SrCw8Px119/Yfbs2Rp9QeXw8PAAj8fD6dOnERMTU6IfAYVZvHgx1q9fj0WLFuHBgwf4999/ceLEiXyHGytM5cqV8eTJE9WPpIKGpNPk3I0dOxYKhQLdunXDtWvX8OLFC5w6dQpnzpwBAHz77bcIDg7GlClT8OTJE5w9exbjx49H//79C/2i/pjP01dffYXExEQMGTIE9erVU2vus2vXLuzYsQMhISF4+fIlDh48iOTkZFXSf+vWLdSoUQO3bt0qcP+anBdAWas4atQoLFiwQFVTWFyaltWHPD09MWzYMIwcORL79+9HWFgYQkJCsHv3bixfvlxt/5cuXcKbN2/UfjDnVlQZF5epXpM7duyIFy9e4NWrV0WuO2HCBPz777/48ccfNb5O+/n5YevWrbhx4wYePHiAIUOGFDjUYEHEYjFGjx6NOXPm4Ndff8W///6L6dOn5xkP+0PJyclo3749/v33Xxw+fBg8Hg9RUVGIiorK00Tkzz//xOeff661uxxEBwzUNpyUAfl1RpPJZMzGxobVq1dPtSw+Pp7x+Xzm6+urtn1+HfcmTpzInJ2dGcdxqk4pz549Y61bt2ZisVitU19sbCwbPXq0agin8uXLs+7du7Pg4GDGWP7DZhUmJiaGTZkyhVWrVo2JRCLm5OTEPv30U7Z//35VB5+SDhuVXycokUjEduzYwRhTH96vTZs2quHkPuw09OTJE9axY0fV8H6dOnVSG96PMcZOnTrFatSowczNzZm3tzc7e/Zsvp0hPzwvH8YYHBzMmjdvzoRCIatSpQo7dOhQkcP7Xbx4kX355ZfMw8ODiUQiJpFImLe3d55hCsPDw1m/fv2Yo6MjEwqFrGLFiqx///7s+fPnjDHNOnUyxtjy5ctZ+fLlGY/HK3J4v9wWLlzIPDw81JYtXbqUubm5qS07fvw4a968ObOwsGDW1tasfv367Pvvv1c9nzNkXGH7jouLYx07dmQ2NjaFDu+n6bn7999/Wffu3ZmNjQ2zsLBg9erVUxuRIvfwfo6Ojmz06NH5Du/3oaI+T4Xp3r07A8BWrVqltvzo0aOsRYsWzM7OjllYWLDatWurRuFg7P17MXeZlvS8MMbY27dvmZmZGRs1alSe5z6mrPL7TH/YcU4mk7Hly5ezTz75hJmZmTGJRMI+/fRT9vPPP6vWOXPmDKtRowYTCoWFDu9XVBkXpLRdkz/77LM8wzbmVxaMMTZixAhWpUoVlpWVpdF1OjIyknXu3JlZW1uzChUqsM2bN+fbGfLD692HHWbT0tLYqFGjmI2NDbOxsWEjR44scni//IZOzPnLfX1ITk5mYrE4z0gkxLhwjBnxTASEEABAeHg4KleujKtXrxarIw0h5L1Hjx6hdu3auH37tqo5BzFdV69eRd++fVV3FMuaFStW4NKlSyW+m0H0g5qOEEIIKdUyMzPx/PlzzJw5E23atKEku5Ro3bo15s+fjxcvXhg6FIOwsLDAhg0bDB0GKQKNOkIIIaRUO3ToEIYNG4batWvjyJEjhg6HaNGoUaMMHYLBjB8/3tAhEA1Q0xFCCCGEEEJ0gJqOEEIIIYQQogOUaBNCCCGEEKIDlGgTQgghhBCiA6W6M2TuWZ3Ix3N0dCxwEgVi3KjsTBeVnWmicjNdVHamy1BlV9jsplSjTQghhBBCiA5Qok0IIYQQQogOUKJNCCGEEEKIDpTqNtqEEEIIKZsYY8jIyIBCoQDHcRpvFx0djczMTB1GRnRFl2XHGAOPx4O5uXmx3k+UaBNCCCGk1MnIyICZmRkEguKlOgKBAHw+X0dREV3SddnJZDJkZGTAwsJC422o6QghhBBCSh2FQlHsJJuQwggEAigUimJtQ4k2IYQQQkqd4tzeJ0RTxX1f0U89QgghhBAtkkql6NOnDwDg7du34PP5cHBwAAD8/vvvEAqFebbZt28fLCws8MUXX+Dw4cNo06YNXF1dAQA7duzAgAEDitVkgRgHSrQJIYQQQrTIwcEB58+fBwCsXr0aYrEYo0ePLnSbQYMGqf7/yy+/oEaNGqpEe+fOnejVq1exEm25XE5tzY2A3hLtu3fvIjAwEAqFAr6+vujevbva81evXsXJkycBAObm5hgxYgQqVaqk0baEEEIIIcZKoVCgQ4cOOHv2LB4+fIh27drh1q1bcHNzg7e3N/744w9s3rwZYrEYFSpUQEhICMaNGwdzc3P06dMH0dHR+OKLL2Bvb48jR47g8uXLWLVqFbKysuDh4YG1a9dCLBajWbNm6Nu3Ly5fvoyhQ4eiW7duhn7pZZ5eEm2FQoFdu3Zhzpw5kEgkmDlzJry8vFChQgXVOs7Ozvjuu+9gZWWFf/75B9u3b8eSJUs02pYQQgghpCDz5tng0SMzjdblOA6MsSLXq1UrGwsWJGm0Tx6Ph8zMTCQnJ+PWrVuoX78+/v77bzRt2hQSiUStprpz587Ys2cP5s6di/r16wNQNh355Zdf4ODgAKlUih9++AGHDx+GpaUlNm3ahO3bt2Py5MkAAJFIhBMnTmgUF9E9vSTaYWFhcHV1hYuLCwDA29sbQUFBasnyJ598ovp/tWrVEBcXp/G2hBBCCCHGzMvLC0FBQbh58ybGjx+PS5cugTGGZs2aFWs/d+7cwdOnT1W11dnZ2WjcuLHq+a5du2o1bvJx9JJoS6VSSCQS1WOJRILQ0NAC17948SIaNmxY7G0vXLiACxcuAACWLVsGR0dHbYRP3hEIBHROTRSVnekqrWUXHBwMAGjUqJGBI9GN0lpupiQ6Olo1vN+SJWk6OkrRaRSPxwOPx0OLFi0QFBSEiIgIdOrUCZs3bwafz4e/vz8EAoFqPYFAAI7jwOfzVfHnfszj8dCmTRts27Ytz7E4joO1tXWZHtZQ169dJBIV67Otl5LI7xZMQcOjPHjwAJcuXcKCBQuKva2fnx/8/PxUj2NjY0sSLimAo6MjnVMTRWVnukpr2VWsWBFA6b1Ol9ZyMyWZmZkl6gwoEAggk8m0FodCoYBCoUCTJk2wdOlSNG/eHAqFAnZ2drhw4QJmzJgBmUymWk8mk8HS0hKJiYmqOMRiMRITE2Fra4sGDRogICAAoaGhqFy5MtLT0/HmzRtUrVoVjDHI5XKtxm9KtF12+cnMzMzz2S5fvnyB6+tlHG2JRKJqCgIAcXFxsLe3z7Pey5cvsW3bNnz77bewtrYu1raEEEIIIcbK3d0dAFRNRZo0aQJbW1vY2dnlWffLL79EQEAA/P39kZ6ejv79+2PAgAHo3bs3JBIJ1q5di2+++QZ+fn7o0qULnj17ps+XQoqBY5q0+P9IcrkcEydOxLx58+Dg4ICZM2diwoQJqjcdoKzV+P777zFu3Di19tqabFuQN2/e6OT1lFVUQ2O6qOxMV2ktu9WrVwMApk6dauBIdKO0lpspSUtLg6WlZbG300etKNENfZRdfu+rwmq09dJ0hM/nY9iwYVi8eDEUCgXatm0Ld3d3nDt3DgDQrl07HDlyBCkpKdi5c6dqm2XLlhW4LSGEENO1Zs0aAKU30SaEEEBPNdqGQjXa2kU1NKaLys50ldayoxptomtUo132lNkabUIIISS30ppgE0JIbnrpDEkIIYQQQkhZQ4k2IYQQvbt37x7u3btn6DAIIUSnqOkIIYQQvevYsSMAICIiwsCREEKI7lCNNiGEEL2rW7cu6tata+gwCNGpH374AW3btoWfnx/8/f1VM6IW19mzZ/H06VPV4969eyMkJETj7V+9eoXjx4+rHoeEhGDu3Lkab5+dnY0lS5agZcuW8PHxQadOnXDx4kWNt8/tw9eiqeK+ZmNBNdqEEEL07uzZs4YOgRCdun37Ni5cuICzZ89CJBJBKpUiKyurRPs6e/Ys/Pz8UL169RJtn5No9+jRAwBQv3591K9fX+PtV65ciejoaFy8eBEikQhv377FjRs3ShRLYa9FJpOVuunjqUabEEIIIUTLYmJi4ODgAJFIBABwcHCAq6srrl69iuHDh6vWu3LlCkaMGAEAqFatGpYtWwY/Pz907twZb9++RVBQEM6fP49FixbB398f4eHhAIBTp06hU6dOaNWqFf7++28Aykn+Fi5ciM8//xx+fn7Yv38/AGDJkiW4desW/P39sX37dly/fh2DBg0CAKSmpmLy5Mnw9fWFn58ffv/9d7XXkZ6ejoMHD2LRokWq1+Lk5ISuXbsCAE6cOAFfX1/4+Phg8eLFqu00fS29e/fG0qVL0atXL+zcuRNXr15Fu3bt4OvriylTpiAzM1PbRaNXpetnAyGEEELIB2zmzYPZo0carctxHDSZYiS7Vi0kLVhQ4PNt2rTB2rVr0apVK7Ru3Rpdu3ZFixYt0KpVK8yePRtxcXGQSCQ4fPgwvvzySwDKMZobNWqEgIAALFq0CAcPHsSkSZPg7++vSlhzyGQy/P777/jjjz+wZs0aHD58GIcOHYK1tTVOnz6NzMxMdO/eHW3atMGsWbOwdetW7Nu3DwBw/fp11X7WrVsHa2tr/PHHHwCAhIQEtdfx4sULuLm5wdraOs9rjIqKwuLFi3H27FnY2triq6++wtmzZ9GhQ4divZakpCQcPXoUGRkZaNWqFQ4fPoyqVatiwoQJ2LdvH0aOHFlkeRgrqtEmhBCid40aNUKjRo0MHQYhOiMWi3H27FmsWLECEokEY8aMweHDh8FxHHr16oWjR48iMTERd+7cgY+PDwBAKBTC398fgLIfw+vXrwvc/+effw4AqFevnmq9y5cv48iRI/D390fnzp0RHx+PFy9eFBrn1atXMWTIENVjOzs7jV9jSEgIWrRoAYlEAoFAgJ49e+LmzZvFfi05tePPnj1DxYoVUbVqVQDAF198oaqtN1VUo00IIUTvoqOjDR0CKUMKq3n+kDZnF+Tz+fD29oa3tzdq1KiBX375BX369EGfPn0wZMgQiEQidO7cWdUuWSAQgOM41baFxSEUCvNdb9GiRfjss8/U1s1dg/0hxpjqmPmpXLkyIiIikJKSAisrqzzbFqQ4ryVnpsXSOFk51WgTQgjRuzt37uDOnTuGDoMQnQkLC8Pz589Vjx8+fIgKFSoAAFxdXeHi4oL169ermo0UxsrKCqmpqUWu16ZNG+zbtw/Z2dkAlDXEaWlphW7fpk0bBAYGqh5/2HTEwsICX331FebOnavqzBkdHY2jR4+iYcOGuHnzJqRSKeRyOU6cOIEWLVqU+LV4enri1atXqlr4o0ePonnz5kW+bmNGiTYhhBC9c3V1haurq6HDIERn0tLSMGnSJHz22Wfw8/NDaGgopk6dqnq+Z8+eKFeunEYjiXTr1g1btmxBu3btVJ0h89OvXz9Uq1YNHTp0gI+PD2bMmAGZTIaaNWuCz+fDz88P27dvV9tm4sSJSExMhI+PD/z8/PKt/Z4+fTokEgnatm0LHx8fDB8+HBKJBC4uLpg5cya++OIL+Pv7o27dumjfvn2JX4u5uTnWrFmDr7/+Gr6+vuDxeBg4cGCR58eYcaw01tO/8+bNG0OHUKo4OjoiNjbW0GGQEqCyM11UdqaJys3w0tLSVE0SikObTUcKM3v2bNSpUwdfffWVzo9VVuij7PJ7X5UvX77A9alGmxBCiN5Nnz4d06dPN3QYhBhEhw4d8PjxY/Ts2dPQoRAdo86QhBBC9O7gwYMAgBUrVhg4EkL0jyZsKjso0SaEEKJ3y5cvN3QIhBCic5RoE0II0bsBAwYYOgRCCNE5aqNNCCGEEEKIDlCiTQghRO/OnTuHc+fOGToMQgjRKUq0CSGE6N3QoUMxdOhQQ4dBiE65u7vD399f9bdx40YAQO/evRESEvLR+z98+DDc3Nxw9epV1bIzZ87Azc0Np06dKvb+rl+/jkGDBhW5To0aNeDv7w8/Pz/06dNHZ0NZDh8+HP7+/mjZsqXqmP7+/ggKCipwm5CQEMydO1cVa+51z549i6dPn+ok1oJQG21CCCF65+fnZ+gQCNE5c3NznD9/XqfHqFmzJk6ePInWrVsDAE6ePIlatWrp9JhNmzbFvn37AABLly7Fnj17MG3aNK0fZ9euXQCUCfPWrVtVxyxM/fr1Ub9+fQDAjRs3IBaL0aRJEwDKRNvPz0+jSYJyyGQyCAQlT5epRpsQQoje7d27F3v37jV0GIQY3IkTJ+Dr6wsfHx8sXrxYtbxatWpYtmwZ/Pz80LlzZ7x9+zbf7Zs2bYq7d+8iOzsbqampCA8PR+3atVXPN2vWDFKpFICytrd3794AlEloTg1xu3btkJKSorbfu3fvol27dnj58mWBsTPGkJKSAltbWwDA6tWrsXXrVtXzPj4+ePXqFVasWIGdO3eqli9btgy7du1CdHQ0evbsCX9/f/j4+ODvv/8u8nz5+voiMTERjDHUrl0bv/zyCwBg/PjxuHz5sqpW/tWrV9i/fz927NgBf39/3LhxA+fPn8eiRYvg7++P8PBwhIeHo3///ujQoQN69OiBsLAwAMCkSZPw3XffoXfv3mplUhKUaBNCCCGk1HNzc4Obm5vassGDB8PNzU2tv8C+ffvg5uamNqFSVFQU3Nzc0KhRo2IdMyMjQ63pyMmTJ9Wej4qKwuLFi/Hzzz/j3LlzuHv3rmqM7bS0NDRq1AgXLlxA8+bNVWPPf4jjOLRq1Qp//vkn/ve//6Fdu3YaxbZ161YsWbIE58+fx/Hjx2Fubq56LigoCAEBAQgMDISHh0eebW/dugV/f380adIEV69eRd++fQs91ldffaVKiBUKBX799Vf06NEDx48fR5s2bXD+/HmcP39e7QdCQby8vBAUFIR///0XHh4euHXrFgAgODgYXl5eqvXc3d0xcOBAjBw5EufPn0eLFi3g7++POXPm4Pz586hUqRKmT5+OhQsX4uzZs5g7dy5mzpyp2v758+c4fPgw5s+fX2RMhaGmI4QQQgghOlBU05GQkBC0aNECEokEANCzZ0/cvHkTHTp0gFAohL+/PwCgbt26au2wP9StWzfs3r0bSUlJmDdvHjZs2FBkbE2aNMH333+PHj16oGPHjqppxMPCwjBjxgz8+OOPcHV1zXfb3E1HNm3ahEWLFhU6Nr67uzvs7e3x4MEDvH37FrVr14aDgwMaNGiAqVOnQiaToX379qhTp06RcTdr1gx///03Xr9+jUGDBuHAgQOIjIyEvb09xGJxkdvnSE1NxZ07d/D111+rlmVlZan+37lzZ/D5fI33VxBKtAkhhOhdTs1iRESEgSMhZUV+77X8mi8NGjQI/fr1U1vm6uqqk/cqY6zA5wQCATiOAwDw+XzIZLIC123YsCGePHkCc3NzVK1aNc9+FAoFACAzM1O1fNy4cfD19cXFixfRpUsXHD58GADg7OyMzMxMPHjwoMBEO7d27dph5MiRqjhzjvXh8b766iv8/PPPiImJUdWAN2/eHEePHsUff/yBiRMnYvTo0fjiiy8KPV6zZs2wZ88eVKhQATNmzMCZM2fw+++/o2nTpkXGmptCoYCNjU2BP4QsLS2Ltb+CUNMRQgghhBADaNiwIW7evAmpVAq5XI4TJ06gRYsWJdrXzJkzERAQkGd5hQoVcO/ePQDA77//rloeHh6OmjVr4ptvvkH9+vVV7ZNtbGywb98+LFu2DNevXy/yuLdu3VI1L3F3d8f9+/cBAPfv38d///2nWq9jx464dOkSQkJC8NlnnwEAXr9+DUdHR/Tv3x99+/ZVbVsYNzc3SKVSvHjxAh4eHmjatCm2bt2KZs2a5VlXLBartT23srJCamoqAMDa2hru7u747bffACh/9Dx8+LDI4xcXJdqEEEL0LiIigmqzSan3YRvtJUuWqD3v4uKCmTNn4osvvoC/vz/q1q2L9u3bl+hYPj4+aNmyZZ7lU6ZMwbx589CjRw+1phA7d+6Ej48P/Pz8YG5ujrZt26qec3Jywt69ezF79mwEBwfn2WdOG20/Pz8cPXoU8+bNAwB8/vnnSEhIgL+/P/bt24cqVaqothEKhfD29kaXLl1UcVy/fh3t2rVDu3btcPr0aYwYMUKj19qwYUPVvps2bYqoqCjVyCK5+fv74+zZs/D398fff/+Nbt26YcuWLWjXrh3Cw8OxceNG/PTTT/Dz80Pbtm11MrY/xwq7b2Hi3rx5Y+gQShVHR0edjZVJdIvKznRR2ZkmKjfDS0tLK9Htf4FAUGgzDVIyCoUC7du3x7Zt29QScG3SR9nl977Kad+eH6rRJoQQQgghOvP06VO0bNkSrVq10lmSbayoMyQhhBC9Gzx4MID8O6MRQkqX6tWr48aNG4YOwyAo0SaEEKJ3Fy5cMHQIhBCic5RoE0II0bvAwEBDh0BKuVLcBY0YUHHfV5RoE0II0TtNZ68jpKR4PB5kMhkEAkp1iHbIZDLweMXr3qi3d9/du3cRGBgIhUIBX19fdO/eXe35iIgIbN68GS9evEDfvn3RtWtX1XOnTp3CxYsXwXEc3N3dMXbsWAiFQn2FTgghhBATY25ujoyMDGRmZqomftGESCRSm2iFmA5dlh1jDDweT22qek3oJdFWKBTYtWsX5syZA4lEgpkzZ8LLywsVKlRQrWNlZYWhQ4ciKChIbVupVIozZ85g7dq1EAqFWLNmDa5fv64a7JwQQojpOXDgAABgwIABBo6ElFYcx8HCwqLY29HQjKbLGMtOL8P7hYWFwdXVFS4uLhAIBPD29s6TUNva2sLT0zPfeeUVCgWysrIgl8uRlZUFe3t7fYRNCCFER2bMmIEZM2YYOgxCCNEpvdRoS6VSSCQS1WOJRILQ0FCNtnVwcECXLl0wZswYCIVC1K9fH/Xr19dVqIQQQvSgf//+hg6BEEJ0Ti+Jdn49NDVtL5WSkoKgoCBs2rQJlpaWWLNmDa5cuYJPP/00z7oXLlxQDRm1bNkyODo6flzgRI1AIKBzaqKo7ExXaS273bt3GzoEnSqt5VYWUNmZLmMsO70k2hKJBHFxcarHcXFxGjf/uH//PpydnWFjYwMAaNasGZ4+fZpvou3n5wc/Pz/VY2Nrp2PqjLHtE9EMlZ3porIzTVRupovKznQZquwMPgV71apVERkZiZiYGMhkMly/fh1eXl4abevo6IjQ0FBkZmaCMYb79+/Dzc1NxxETQgjRpaioKERFRRk6DEII0Sm91Gjz+XwMGzYMixcvhkKhQNu2beHu7o5z584BUI6nmpCQgICAAKSnp4PjOJw+fRpr1qxBtWrV0Lx5c8yYMQN8Ph+VKlVSq7UmhBBieho3bgxAObQrIYSUVhwrxVMnvXnzxtAhlCp0O810UdmZrtJado0aNQIABAcHGzgS3Sit5VYWUNmZLmNsOkLTJRFCCNG70ppgE0JIbnppo00IIYQQQkhZQ4k2IYQQQgghOkCJNiGEEL3r0KEDOnToYOgwCCFEp6iNNiGEEL27f/++oUMghBCdo0SbEEKI3p05c8bQIRBCiM5Rok0IIUTv6tWrZ+gQCCFE56iNNiGEEEIIITpAiTYhhBC9W716NVavXm3oMAghRKco0SaEEKJ3a9aswZo1awwdBiGE6BS10SaEEKJ3U6ZMMXQIhBCic5RoE0II0bupU6caOgRCCNE5ajpCCCGEEEKIDlCiTQghRO/u3buHe/fuGToMQgjRKWo6QgghRO86duwIAIiIiDBwJIQQojuUaBNCCNG7unXrGjoEQgjROUq0CSGE6N3Zs2cNHQIhhOgctdEmhBBCCCFEByjRJoQQQgghRAco0SaEEKJ3jRo1QqNGjQwdBiGE6BS10SaEEKJ30dHRhg6BEEJ0jhJtQgghenfnzh1Dh0AIITpHiTYhhBC9c3V1NXQIhBCic9RGmxBCCCGEEB2gRJsQQojeTZ8+HdOnTzd0GIQQolOUaBNCCNG7gwcP4uDBg4YOgxBCdIraaBNCCNG75cuXGzoEQgjROUq0CSGE6N2AAQMMHQIhhOgcNR0hhBBCCCFEByjRJoQQonfnzp3DuXPnDB0GIYToFDUdIYQQondDhw4FAERERBg4EkII0R1KtAkhhOidn5+foUMghBCdo0SbEEKI3u3du9fQIRBCiM5RG21CCCGEEEJ0QG812nfv3kVgYCAUCgV8fX3RvXt3tecjIiKwefNmvHjxAn379kXXrl1Vz6WmpmLr1q149eoVOI7DmDFjUL16dX2FTgghhBBCSLHpJdFWKBTYtWsX5syZA4lEgpkzZ8LLywsVKlRQrWNlZYWhQ4ciKCgoz/aBgYFo0KABpk6dCplMhszMTH2ETQghREfc3NwAUGdIQkjpppemI2FhYXB1dYWLiwsEAgG8vb3zJNS2trbw9PQEn89XW56WlobHjx/Dx8cHACAQCCAWi/URNiGEEEIIISWmlxptqVQKiUSieiyRSBAaGqrRtjExMbCxscHmzZvx8uVLVKlSBUOGDIG5uXmedS9cuIALFy4AAJYtWwZHR0ftvAACQPkjh86paaKyM12ltexK+53J0lpuZQGVnekyxrLTS6LNGMuzjOM4jbaVy+V48eIFhg0bhmrVqiEwMBAnTpxA375986zr5+enNmRUbGxsyYMmeTg6OtI5NVFUdqaLys40UbmZLio702WositfvnyBz+ml6YhEIkFcXJzqcVxcHOzt7TXeViKRoFq1agCA5s2b48WLFzqJkxBCCCGEEG3RS6JdtWpVREZGIiYmBjKZDNevX4eXl5dG29rZ2UEikeDNmzcAgPv376t1oiSEEGJ6Bg8ejMGDBxs6DEII0Sm9NB3h8/kYNmwYFi9eDIVCgbZt28Ld3R3nzp0DALRr1w4JCQkICAhAeno6OI7D6dOnsWbNGlhaWmLYsGFYv349ZDIZnJ2dMXbsWH2ETQghREdy+tMQQkhpxrH8GlCXEjm14EQ7qN2a6aKyM12ltexyV7SURqW13MoCKjvTZYxttGkKdkIIIXpXWhNsQgjJjaZgJ4QQQgghRAco0SaEEKJ3Bw4cwIEDBwwdBiGE6BQ1HSGEEKJ3M2bMAAAMGDDAwJEQQojuUKJNCCFE7/r372/oEAghROco0SaEEKJ3K1asMHQIhBCic9RGmxBCCCGEEB2gRJsQQojeRUVFISoqytBhEEKITlHTEUIIIXrXuHFjAEBERISBIyGEEN2hRJsQQojeubi4GDoEQgjROY0T7Xv37uHatWtITExEQEAAnj17hvT0dNSpU0eX8RFCCCmFgoODDR0CIYTonEZttM+cOYMdO3agXLlyePz4MQBAKBTip59+0mlwhBBCCCGEmCqNEu3Tp09j7ty56N69O3g85SZubm548+aNToMjhBBCCCHEVGmUaKenp8PR0VFtmUwmg0BATbwJIYQUX4cOHdChQwdDh0EIITqlUaJds2ZNnDhxQm3ZmTNnULt2bV3ERAghpJS7f/8+7t+/b+gwCCFEpzSqkh42bBiWL1+OP/74AxkZGZg4cSIsLS0xY8YMXcdHCCGkFDpz5oyhQyCEEJ0rMtFWKBSIiIjAggUL8N9//+Ht27eQSCTw9PRUtdcmhBBCiqNevXqGDoEQQnSuyEyZx+NhxYoVEAqF8PT0RIsWLVC9enVKsgkhhBBCCCmExm20nz59qutYCCGElBGrV6/G6tWrDR0GIYTolEZttJ2cnLB06VJ4eXlBIpGA4zjVc3369NFZcIQQQkqnNWvWAACmTp1q4EgIIUR3NEq0s7Ky0KRJEwCAVCrVaUCEEEJKvylTphg6BEII0TmNEu2xY8fqOg5CCCFlCNVkE0LKAo1nnImMjMS1a9cglUrh4OCAli1boly5crqMjRBCCCGEEJOlUWfI27dvIyAgABEREbCyssKbN28QEBCA27dv6zo+QgghpdC9e/dw7949Q4dBCCE6pVGN9qFDh/Dtt9+iTp06qmUPHz7E7t274eXlpbPgCCGElE4dO3YEAERERBg4EkII0R2NEm2pVIqaNWuqLatRowbi4uJ0EhQhhJDSrW7duoYOgRBCdE6jpiOVKlXCb7/9prbs1KlTqFSpki5iIoQQUsqdPXsWZ8+eNXQYhBCiUxrVaI8YMQLLly/HmTNnIJFIEBcXB5FIhOnTp+s6PkIIIYQQQkySRom2m5sb1q5di9DQUNWoI56enhAINB60hBBCCCGEkDJFo0w5PDwcVlZWqFGjhmpZbGwsUlJSqPkIIYSQYmvUqBEAIDg42MCREEKI7mjURnvDhg2Qy+Vqy2QyGTZu3KiToAghhJRu0dHRiI6ONnQYhBCiUxrVaMfGxsLFxUVtmaurK96+fauToAghhJRud+7cMXQIhBCicxrVaDs4OOD58+dqy54/fw57e3udBEUIIaR0c3V1haurq6HDIIQQndKoRrtTp05YuXIlunbtChcXF0RHR+O3335Dz549NT7Q3bt3ERgYCIVCAV9fX3Tv3l3t+YiICGzevBkvXrxA37590bVrV7XnFQoFAgIC4ODggICAAI2PSwghhBBCiCFolGj7+flBLBbj4sWLiIuLg0QiwaBBg9C8eXONDqJQKLBr1y7MmTMHEokEM2fOhJeXFypUqKBax8rKCkOHDkVQUFC++zh9+jTc3NyQnp6u0TEJIYQYr5zhYVesWGHgSAghRHc0Hp+vRYsWaNGiRYkOEhYWBldXV1U7b29vbwQFBakl2ra2trC1tc23B3pcXByCg4PRs2dPnDp1qkQxEEIIMR4HD/IBVMTSpQCfb+hoCCFENwpNtJ8/fw6BQICKFSsCAJKSkrBnzx68evUK1apVw6BBg2Bubl7kQaRSKSQSieqxRCJBaGioxkHu2bMHAwYMoNpsQggpBf77jw+BYBdkMgGGD8/Apk3xEIuZocMihBCtKzTR3rNnD3r37q1KtLdu3Yr4+Hj4+vri2rVrOHDgAEaMGFHkQRjLewHlOE6jAO/cuQNbW1tUqVIFDx8+LHTdCxcu4MKFCwCAZcuWwdHRUaNjEM0IBAI6pyaKys50lcaymzCBD6GQh1mzZFi0SIQ+fVxw7JgM5csbOjLtKY3lVlZQ2ZkuYyy7QhPtiIgI1KxZEwCQmpqKf/75B6tXr0b58uXh5eWFuXPnapRo50zbniMuLk7jEUv+/fdf3L59G//88w+ysrKQnp6O9evXY8KECXnW9fPzg5+fn+pxbGysRscgmnF0dKRzaqKo7ExXaSu7mzeFOHrUEdOmJeHrr1NQtaoIY8bYo2VLHvbtk6JmTZmhQ9SK0lZuZQmVnekyVNmVL6SWoNDh/eRyuWqa9dDQUNjZ2al25ujoiNTUVI0CqFq1KiIjIxETEwOZTIbr16/Dy8tLo2379euHrVu3YtOmTZg0aRLq1KmTb5JNCCHEuMnlwPz5NihfXgZPz5M4d+4c/Pwycfx4LBQKDt27O+LyZZGhwySEEK0ptEbb3d0dN27cgLe3N65du4a6deuqnpNKpbC0tNToIHw+H8OGDcPixYuhUCjQtm1buLu749y5cwCAdu3aISEhAQEBAUhPTwfHcTh9+jTWrFmj8TEIIYQYt19+scCDB0Js2hSP0aMHAVDeOa1TR4bffnuLQYMkGDjQAUuXJqJ//zQDR0sIIR+PY/k1oH7nyZMnWL58OQCAx+Nh4cKFqhrtU6dOITQ0FJMnT9ZPpCXw5s0bQ4dQqtDtNNNFZWe6SkvZJSdzaN3aGR4ecpw4EYshQwYDAPbu3au2zpgx9rh0yRzffJOMgIBk8DSaVs34lJZyK4uo7EyXMTYdKbRGu0aNGti8eTMiIyNRrlw5WFhYqJ5r1KgRvL29tRclIYSQUmvjRiu8fcvHnj1ScJx6gp3D2pphzx4pZs+2xaZN1vjvPwHWro1Hrq8eQggxKUWOo21hYYEqVarkWV5Y9k4IIYTkePmSj+3brdC7dxoaNMgudF2BAFi2LBGVK8uwcKEt3rzhIzBQColEoadoCSFEe0z0phwhhBBTsWiRDfh8hpkzkzRan+OA0aNTsW2bFA8fmqFLF0eEhdGsNoQQ00OJNiGEEJ25cUOI06ctMG5cClxd39dKu7m5wc3NrdBtO3fOwM8/xyIlhUO3bk64eVOo63AJIUSrKNEmhBCiE3I5MG+eLSpUkOHrr1NKtI/GjbPx22+xcHSU46uvJDh2jBpsE0JMh0aJ9vTp0/NdHhAQoNVgCCGElB6HD1vi0SMzzJ6dlKdDY0REBCIiIjTaj4eHHCdPxqJx4yyMH2+PtWutUPB4WYSQfGVkQHjzJujDo18aJdpRUVF5ljHGEB0drfWACCGEmL7kZA7Ll1ujadNMdOmS8dH7s7Nj+PHHOPTqlYZVq2wwebIdsrK0ECghZQCXlARJ//5w7NULNgsWULKtR4WOOrJx40YAgEwmU/0/x9u3b+Hu7q67yAghhJis9eutEBfHw/79SeA47exTKAR++CEBlSrJsHq1DSIi+NixQwo7O0oaCCkILzYWDv37w+zJE2T4+MBq+3YAQNK8edDah5MUqNBE28XFJd//cxyHTz75BC1atNBdZIQQQkxSeDgfO3da4Ysv0lGvXv7D+Q0enHfCGk1wHDBlSgo8POSYOtUO3bo5Yv9+KSpWlH903ISUNvyICEj69gXvzRtIAwOR2bYtbObNUybbHIekuXMp2daxQhPtL774AgBQrVo1NGjQQB/xEEIIMXELF9rAzIwhIKDg4fwuXLjwUcfo1Ssd5cvLMWKEA7p0cURgoBSNGhU+RjchZYkgLAySvn3BpaZC+tNPyGrSBACQ9K7piNW2bcpke84cSrZ1qMgJawCgQYMGePPmDcLDw5GRod7WzsfHRyeBEUIIMT1//SXE2bMWCAhIgotLwZPMBAYGfvSxWrTIwsmTsRg0yAFffOGIDRvi8fnnH98enBBTZ3bvHhz69wf4fMQeOQJZ7drvn+Q4JC1cCACw2roVACjZ1iGNEu1jx47h6NGj8PDwgEgkUnuOEm1CCCGAcji/776zhbu7DCNHFj6cX7t27bRyTE9PGX77LRZDhjhg1Ch7zJmThK+/TqWcgZRZwuvX4TB0KBR2dog7dAjyfGb3zkm2OcaUyTbHIWn2bEq2dUCjRPv06dNYsmQJPDw8dB0PIYQQE/Xjj5Z4/NgM27ZJYW6uv+NKJAr8/HMsJk2yx8KFtnj5UoCFCxMh0OgbjpDSQ3TuHBxGj4asYkXEHToERblyBa/McUhctEjZjGTLFjCOQ/KsWZRsa5lGlyGhUFjkDF6EEELKrqQkDitWWKN580x06lR0840DBw4AAAYMGKCV41tYAFu2xMPDQ4ZNm6zx+jUfW7bEw8qKRiQhZYPF0aOwmzwZ2XXrQrp/PxQODkVvxHFIXLwYAGC9eTMAULKtZRqNo92nTx/s3r0b8fHxUCgUan+EEELIunXWiI/n4fvvEzX6jp4xYwZmzJih1Rh4PGDWrGQsX56Ay5dF6NHDEW/e0ATIpPQT794N+wkTkNW8OeIOH9Ysyc7xLtlOHTQI1ps3w3rpUhpnW4s0qtHe/O5Xzh9//JHnucOHD2s3IkIIISbl+XM+du8Wo2/fNNSpI9Nom/79++ssngED0lChghxff22PLl2csHdvnMZxEWJSGIPV2rWwWb0a6R06IH7TJpSo3Vbumu1NmwCOQ3JAANVsa4FGifaHk9UQQgghORYutIFIxDB9erLG26xYsUKHEQGffZaJ48djMWiQBD17OmLLlnj4+mbq9JiE6JVCAZvvvoPVrl1I+/JLJKxciY/qmMDjKZNtxmD9Lu+jZPvjaXRPzcnJCU5OTpBIJBAIBKrHTk5Ouo6PEEKIEbtyRYhz5ywwYUIKnJ2NqzlhrVoynDr1FpUryzBkiAP27rU0dEiEaEd2NuwmTYLVrl1IGTkSCatXf1ySnYPHQ+KSJUgdMADWGzfCetkyakbykTQqldTUVOzcuRM3b96EQCDA/v37cfv2bYSFhaFv3766jpEQQogRksmA77+3hYeHDCNGFD6c34eioqIAAK6urroITcXVVYFjx+Iwdqw9Zs2yw8uXAsyZkwQeNd0mpiojA/ZjxsDi3DkkTZ+OlAkTtFvrzOMhcelSAFDWbHMckmfMoJrtEtLoUrNjxw5YWlpi8+bNELz7xVS9enVcv35dp8ERQggxXgcPWuLJEzPMmZOED6ZYKFLjxo3RuHFj3QT2AbGYYfduKYYOTcG2bVb4+mt7pKdT0kBMD5ecDMmAATA/fx4JixcjZeJE3STA75Lt1P79Yb1hA6yXL6ea7RLSqEb7/v372LZtmyrJBgAbGxskJibqLDBCCCHGKzGRw8qV1mjRIhMdOxZ/NkYXFxcdRFUwPh9YtCgJlSrJ8d13NvjiCwkCA6VwcjKu5i6EFIQXFweHAQNg9ugREjZsQHqPHjo+IA+Jy5YBAKw3bFDWbE+fTjXbxaRRom1paYnk5GTY29urlsXGxqo9JoQQUnasXWuNhAQevvtOs+H8PhQcHKz9oDQwYkQqKlSQ45tv7NCliyP275eiWjUakYQYN15EBCT9+kHw+jWku3Yh089PTwfOlWyvXw8AlGwXk0ZNR3x9fbF69Wo8ePAAjDE8ffoUmzZtgr+/v67jI4QQYmTCwvgIDBSjXz/Nh/MzJh06ZODYsThkZHDo1s0R164JDR0SIQXiP3sGxx49wI+ORtzBg/pLsnO8S7ZT+/WD9fr1sF65kpqRFINGiXa3bt3QokUL7Nq1C3K5HFu2bIGXlxc+//xzXcdHCCHEyCxcaAtz8+IN52ds6tfPxqlTsXB1laNfPwl+/tnC0CERkofgwQM49ugBLiMDsUeOIKt5c8MEwuMhcflyZbL9ww+wXrWKkm0NadR0hOM4dOrUCZ06ddJ1PIQQQozY5csiXLhgjjlzEuHoWPL2zR06dAAAnD17VluhFVuFCnKcOBGLUaMcMHmyPV6+FGDatGS6K06MgvDvv+EweDAUNjaIO3QI8qpVDRvQu2QbjMF63ToAQPK0adSMpAgaD7oYExOD//77DxkZ6p1eWrVqpfWgCCGEGB+ZDPjuOxtUqiTDsGGpH7Wv+/fvaymqj2Njw7B/fxwCAmyxbp01/vuPj1WrEoo9igoh2iS6cAEOX38NWYUKiPvxRyjc3AwdkhKPh8R3k01Zr1un7CA5bZphYzJyGiXax48fx5EjR+Du7g6h8H1bNo7jKNHO5fJlESwtGby8sugHHiGk1DlwwBJPn5ph927pRyeiZ86c0U5QWmBmBqxalYhKleRYtswGb97wsXOnFPb2dGuc6J/F8eOwmzQJ2bVqQXrgABQSiaFDUpeTbDMG67Vrlcn21KmGjspoaZRonzp1CsuXL0eFChV0HY9JW7HCGnfvCuHhIUPPnuno2TMNVarIDR0WIYR8tPh4DitX2qBly0y0a1f84fw+VK9ePS1EpT0cB4wfn4KKFWWYNMkeXbs6Yf/+OFSqRNdwoj+We/bAds4cZDVvDmlgIJi1taFDyh+Ph8SVK8ExBus1awCAku0CaNQZ0srKiqZb18Dhw3FYty4eFSvKsW6dFVq3dkHnzo7Ys8cSUilNQ0YIMV1r11ojKYkr8XB+pqJbtwwcPhyH+HgOXbo4IijIzNAhGY3sbODuXTPs2WOJY8csEBJihpSUUvxm0CfGYLVuHexmz0aGvz/i9u833iQ7B4+HhFWrkNanD6zXrIHVu4SbqOMYK7rb6D///IOrV6+iU6dOsLW1VXvO0dFRZ8F9rDdv3hjs2JGRPJw8aYEjRyzx+LEZBAKGtm0z0atXGvz9M2BubrDQSszR0RGxsbGGDoOUAJWd6TKGsgsLE8DX1wl9+6Zh+XLtTFS2evVqAMBUI60Fe/6cj0GDJHjzho916+LRtWvxavGNodw+VlwcD3fumOH2bSFu3xYiJESIjIy8ibWrqxxVq8pQtaoMnp7v/y1fXm6SU93rvewUCtgsWACrHTuQ1qsXEtasAQQad6EzPIUCdlOnwvLnn5E0dSpSpkwxWCiG+tyVL1++wOc0SrSDgoKwbds2JCfnHcrp8OHDHxedDhky0c7t0SMBjh61xIkTFoiK4sPaWoHOndPRq1c6mjXLMpkLUWn44iirqOxMlzGU3cCBDrh9W4i//oqBRKKdmRTd3nXuioiI0Mr+dEEq5TB8uANu3RJh1qwkjB2bonFtvjGUW3HI5cDTpwJVUn37thDh4cpkz8yMoW7dbDRunAUvryw0bJiNtDQOYWEChIUJ8OzZ+3+Tkt5/oZmbK1ClilyVeOck4VWqyCAWG2/7d72WnUwGu2+/heXPPyNl2DAkff89TCYpyE0uh920acpke9o0pEyebJAwjDHR1ugn086dO/HVV1+hZcuWap0hiWZq1ZKhVq0kzJqVhGvXhDh61BInT1rg0CEx3Nxk6NEjHb17p9PsZIQQo3PxoggXL5pj3rxErSXZADDFgLVemnJwYDh0KA5Tp9phyRIbvHzJx+LFiTArBa1JkpI4/PNPTlJthuBgIVJSlAmek5McXl5ZGDAgFV5e2ahbNyvfu7AffmcxBsTG8tQS77AwAe7fN8Pvv5tDoXj/K6VcOXmu2u9sVK0qR9Wq2ShfXlGqmyapyciA/TffwOLsWWVN8OTJpjtUHp+PhHdja9usWgUABku2jY1GNdojR47Etm3bwDOxX1nGUqOdn7Q0Dv/7nzmOHrXA5csiKBQc6tXLQs+e6ejePR1OTtr7QtMWU6uhIe9R2ZkuQ5Zddjbg5+cEuZzDxYsxKKv1LAoFsHKlNdavt0abNhnYujUeNjaFf3Ua02eOMeDFC76qpvrOHSH+/VcAxjjweAw1a8pUtdVeXlmoWFGu9XwvMxMIDxeo1YLn/CUnv88tLCwUak1Q3v/JYWGhn1pwfZQdl5ICh2HDILp2DYkLFiB1+HCdHk9v5HLYTZkCyyNHDFKzbYw12hol2r/++itkMhl69OgBroSfvrt37yIwMBAKhQK+vr7o3r272vMRERHYvHkzXrx4gb59+6Jr164AgNjYWGzatAkJCQngOA5+fn4az0hpzIl2bjExyvbcR49a4P59Ifh8hjZtMtGzZzo6dMjQ28WlKMb0xUGKh8rOdBmy7HbvFmPuXFsEBsahXbtMg8RgTA4ftsD06Xbw9JRh3z4p3NwKHpHEkOWWns4hJMQsV2JtBqmUDwCwtVWgceMsNGr0vhmIlZXhvmMYU34H5q4Fz/l79YoPxt7nHG5ueRNwT08ZXF21Wwuu67LjpFJIBg6E2f37SFizBum9e+vsWAaRO9n+9lukTJqkt0ObbKI9ZswYJCQkQCAQwMrKSu25LVu2FBmAQqHAxIkTMWfOHEgkEsycORMTJ05UGy4wMTERb9++RVBQEMRisSrRjo+PR3x8PKpUqYL09HQEBATg22+/1WioQVNJtHN7+lSAo0ctcOyYBd68EUAsVuDzzzPQq1cavL2zwOcbLjZK1kwXlZ3pMlTZSaUcWrd2Qd262Th0KE7rNZz37t0DYHzD/BXl6lUhRo50gKUlw549UtSrl53vevost4gInqqm+s4dIR48MINMpiwwT89sNG6craqt9vSUmUwT4PR09Vrw58/fJ+Opqe9fhFj8vha8SpX3yXjlyjJYWBT/uLosO15kJCT9+kHw8iWkW7cis107nRzH4ORy2E2eDMujR5E0fTpSJk7Uy2GNMdHWqI32+PHjPyqAsLAwuLq6wsXFBQDg7e2NoKAgtWTZ1tYWtra2CA4OVtvW3t4e9vb2AAALCwu4ublBKpWW2jG9q1eXYebMZMyYkYybN4U4dswCp05Z4JdfLOHqKkePHuno1SsNNWtSe25CiO6sWaPb4fw6duwIwLg7Q+andessnDwZi0GDHNCzpwSbN8frtbY/Kwt4+NBMlVjfvi1EZKSyBsbCQoEGDbIxZkwKvLyUtdYODsZxR7QkLCyAmjVleb7vGAOiot7Xguck4LduCXHsmKVqPY5jqFBBnicB9/SUwdlZ/23B+c+fQ/LVV+AlJCDuwAFkeXvrNwB94vORsHYtAMDm3UySuk62i642NgyNEu1atWp91EGkUikkuWY2kkgkCA0NLfZ+YmJi8OLFC3h6en5UPKaAxwO8vbPg7Z2FhQsTcf68OY4ds8SOHWJs2WKFmjWz0bt3Grp3T4erq/G15yaEmK6nTwXYt0+MgQPTUKOGbn7U161bVyf71YdPPpHht99iMXSoA4YPd8D33yd99JT0BYmN5b1LqM1w5476EHsVKsjQrFnmu9rqbNSsmW1So8KVFMcB5copUK5cFlq1ylJ7Lj2dw/PnfLVmKGFhAty8aYn09Pe14NbWCtUIKLlHRKlUSTfvd8HDh5D07w/I5Yj75Rdkm9idnBLJSbYZUybbHIeUCRN0cqjERA5jxthj8mQOTZro5BAlVmDTkWPHjqFnz54ACh/Cr0+fPkUe5MaNGwgJCcHo0aMBAFeuXEFYWBiGDRuWZ92ff/4Z5ubmqqYjOTIyMjB//nz07NkTzZo1y/c4Fy5cwIULFwAAy5YtQ1ZWVr7rmbK3b4FffuHh0CEebt3igeMY2rZl6N9fge7dFfigZY9WCQQCyGRUk26KqOxMl77LjjGgSxcBgoI4PHyYDSOeKsHg0tKAwYMF+PVXHsaNk2PFCrmqeV9Jyk0uBx494nDjBoe//+Zw4wYPz54pk2ozM4ZGjRiaN2do0UKBZs0YCrlbTT6gUAAREcDTpxz+/ZdT/RsayuHVq/dV2zweg58fsGdPNrQ18zl3/ToE3bsD1tbI/v13oEYN7ezYVMjl4I8YAf6PP0K2cCEU06drdfevXgFduwoQGsph926GL7/U/3ddYSPyFfjbNy4uLt//l4REIsmzv5zmIJqQyWRYvXo1WrduXWCSDQB+fn7w8/NTPS6NbVI5DvjyS+Xfs2d8HDumnKFr+HABxo1ToEOHDPTqlY7WrTO1XrNB7XxNF5Wd6dJ32V24IML58xJ8910igFTQ26ZwGzcCLi422LjRCk+fZmHTpgRYWjKNyi0piUNwsFBVY513iL0M9OuXhcaNs1C3bnaeIfaobIrHwgKoX1/5l1taWk4tuBkePxZg+3YrtGzJw/79cahcueAOr5oQXbwI+5EjIS9fHnE//QS5o2PZLLhly2CXkQHLuXORlJqKlI9skpzjwQMBBg+WIDUVOHAgDt272xhdG22NOkN+LLlcjokTJ2LevHlwcHDAzJkzMWHCBLi7u+dZ98MabcYYNm3aBCsrKwwZMqRYxzXFzpAlwRhw+7YQR44o23MnJPDg5CRH9+7KSXHq1MnWSls0StZMF5Wd6dJn2WVnAz4+zuA4hj/+eFsqxovWlz17LDF3ri3q1MnG3r1S1KrloFZujClnm8xpV53fEHs5HRa9vLLg7q79IfaIZv791wm9eil/8AQGxqNJk5LdHTc/eRL2EyYgu0YNSA8ehKKs3x6Sy2E3cSIsjx9H0syZSBk37qN29+efIowaZQ8bG4YDB+JQo4bMKDtDapRoDx06FIGBgXmWjxgxAjt37tQoiODgYOzduxcKhQJt27ZFz549ce7cOQBAu3btkJCQgICAAKSnp4PjOJibm2PNmjX477//MG/ePFSsWFE1tOBXX32FRo0aFXnMspJo55aZCVy8qByf+8IFc2Rnc6hePRu9eqWjR4/0QoejKgola6aLys506bPsduwQ47vvbLF3bxz8/HTbwS/nGv5hB3hTduGCCGPG2MPeXoGff2aIikosdIi9nLGrGzQw7BB7RJ2joyOCguIxcKAEb97wsXZtPLp1yyjWPiz374ftzJnIatoU0j17wGxsdBSticmdbM+ahZRvvinRbn76STnU5iefyLBvXxzKlVP2VTPZRHvQoEHYt2+f2jKZTIZRo0Zh9+7dHx+hjpTFRDu3+HgOp04px+cOChKB4xiaN89C795p+PzzjCInXPgQJWumi8rOdOmr7KRSHlq1ckbDhlk4cECq89pUU5iCvSRybmVHRb0fi9XTM1vVYbFxY9MaYq8syvnMSaUcRoxwwN9/izBjRhLGj08p+nPBGKw2boTNsmXI8PWFdNs2lGiMwdJMJlMm2ydOFDvZZgxYvdoaa9da47PPMrBtW7zaj1RjTLQLbcU7b948cByH7OxszJ8/X+25uLg4VK9eXTsREp2wt2cYODANAwem4eVLPo4ds8DRo5aYOtUes2cz+Psrx+f+7LNMukVMSBm3apU1UlI4zJ+fpJcmC3fu3NH9QQygTh0ZfvvtLf74wxHlyyeiUaMs2NtTbbUpcnBgOHQoDlOn2mH5chv89x8fS5cmFvx9yRhsFi2C1datSOvRQzniBn255iUQIOGHHwAANkuWKEcjGTu2yM2ysoBvv7XDkSOW6Ns3FcuWFVIWRqTQRNvHxweAchzstm3bqpZzHAdbW1vUqVNHt9ERrfHwkGPy5BRMmpSCf/4xw9Gjljh50hy//WYBiUSObt3S0bNnOho00E57bkKI6XjyRID9+y0xeHAqqlfXT499V1dXvRzHEMqXV2DyZAViY2k2TVMnEgEbNiTAw0OOdeus8eqVANu3S2Fr+8GPJ7kctjNmQHzoEFKHDEHiwoWg2xaFyEm2GYPN4sVgHIfUMWMKXD0picOoUQ64elWEadOSMGmSBncXjIRGTUciIiJUt/lMSVlvOlKU7Gzg0iURjh61xPnz5sjM5FCligy9eqWhZ890VKyo3p6bmh+YLio706XrsmMM+OorCe7fN8PVq9EmPcGJMaHPnOkqqOwOH1a2C65SRYb9+6WoUOHdd2RmJuzHjYPF6dNInjQJydOmwWSyQEOTyWA3YQIsT55E4pw5+Sbbb97wMGiQBKGhAqxcmYAvv0wvcHcm13Qkh5ubG0JCQhAeHo6MDPUOAZqMo02Mk5kZ0K5dJtq1y0RSEofff1e251650gYrV9qgadNM9OqVjs6d02FnR1++hJRG58+LcPWqCAsWJOo1yZ7+bizdFe9mjSt1aNz6UqdPH+WAAiNHOqBzZ0fs2SNFw2oJcBg+HKKrV5H43XdIHTnS0GGaFoEACevXg2MMtosWARyH1HdzrgDAo0cCDBwoQUoKh/374/Dpp6Y3P4pGNdq7du3CjRs3ULt2bYhEIrXnxmrQrsZQqEa7ZCIictpzWyA01AxCIYOfXwb69zeDu3scPDzkZWL2sdKEatdMly7LLitLOZwfn89w4YJ+h/MrbZ0hufh4iK5fh+jKFYj++gv8mBjEnDkDeRmYybi0KeozFxoqwMCBDpDFJOCuWwc4vbyLhFWrkP7ll3qMspSRyWA/fjwsfv0ViXPnInX0aFy5IsLIkfawsmLYvz8OtWoV/ePVZGu0r127hhUrVsCxrI8BWUa4uckxfnwKxo1Lwf37Zjh61AInTljg9Gk+ABeYmTFUrqycsrZatff/Vq0qg6Ul1XwTYioCA8V48UKA/fvj9N6paPny5fo9oLZlZEAYFATRX39BdPUqzO7dA8cYFFZWyGrRAvzYWNgsW4Z4DYfAJaajWjUZzux6AFHX/rB9HobDffai9RdtQY1FPoJAgPgNGwDGYLtwIe6GmGHg6TmoVk05fF/58gpDR1hiGiXa1tbWEIvFuo6FGBmOA+rVy0a9etmYOzcJr145ISgoBWFhAoSFCfDkiRn+9z9zyOXvLy9ubu+T7txJuESioCZrhBiRuDge1q2zho9PBnx89N9pb8CAAXo/5keRy2H28CFEV69CdPUqhEFB4DIywAQCZDVujOSpU5HZqhWyGzQAzMzgvG0bLBYsQEpQELKbNDF09ESL+OHh+GTEV+Dx4zC76QksO9wRg81TsWBBIt3t/RgCAaQbNiL2Xz7a/DoP66oI4Xd8EKytTbsCT6O3ROfOnbF+/Xr06NEDtra2as+5uLjoJDBiXAQCoEkThsqV1TshZGYCL18KEBqq/Hv2TPnv339bIj39fY9rOzvFu6Q7G56eMtWfu7scfP6HRyOE6NrKldZITVUO50fyx3/5UplYX7kC0bVr4CUkAACya9ZE6sCByGzdGlnNm4PlUxGlmDgR8i1bYLN4MeKOH6fOcaWE4PFjSPr1A7KzEffzz/imXn0kLE7B1q1WePWKjy1b4mnyoRLKzgZmzHDEkae/4GqFPvjmeQASD6UjddQoQ4f2UTRKtHNmf8xvBq/Dhw9rNyJiUkQioHp1WZ4hwRQKIDKSj7Aw9ST8/HlzHDr0PrM2N/+wGYoyEa9SRUZj/BOiI48eCXDwoCWGDk2Fp6dhOu3lnhnYWPCkUgj/+kvVHETw338AAHm5csho1w6ZrVsjs1UrKJydi96ZlRWSp0yB3cyZMD93Dhnt2+s4eqJrZrdvQzJoEJiFBeKOHYOsenXwAMydmwQPDxnmzLFFz56O2Lv3/UyFRDPJyRxGjbLHlSvmmDIlGRXHr0H6uEzYfv89AJh0sq1RZ0hTRZ0htUtbnQzi47l3zU/MVIn4s2cCvHzJB2PKWh+OY3B3l6tqvqtVy2mSkk3Dj5UAdYY0XdouO8aAPn0kePjQDH/9FW2wyVSMoTMkl54O4a1byqYgV69C+OABAEBhY4NMb29VYi2vWrXYNdKOjo6IjYyEs48PGI+Ht3/8AWpXYBry+8yJLl+G/fDhULi4IO6nnyB3d8+z3aVLInz9tT2srRn27YtD7do08owmIiN5GDhQOXzfihUJ6NPn3Z3z7GzYf/MNLH7/HYnz52uUbJtsZ8gcyilJpTQjJPko9vYMTZpko0mTbLXlGRnA8+cCVRtwZRJuhuvXRcjIeP8lJ5GoJ+A5/5YvL6f5AQgpwrlz5rh2TYTFixMMOmOhn5+f/g8ql8Ps3r337axv3waXlQVmZoYsLy8kTZ+OzNatkV2vnnaSYjMzJM2cCYeRI2F5+DDS+vf/+H0SvTP/7TfYjx8PWbVqiPvxRyicnPJdr23bTBw/HotBgyTo0cMRW7bEw9eXJi0qzOPHyuH7kpI47NsnRZs2uc6XmRniN21SdpD8/nvl0H8mOHyiRjXasbGx+OGHHxAeHg4A2L9/P27evIm7d+9idK7xDo0N1Whrl6F+KcrlyiEHQ0NzJ+DKJDwh4X1mbWGhUHXCzN0Zs3JlGT4YlbLMoRpt06XNssvMVA7nJxQynD//tvRXsDIG/osXysT6r7+U7awTEwEA2bVqKWusW7dGVrNmYJaWWj20qtwYg2O3buC/fo2Yv/7S+nGI9uX+zFkePAjbgABkNW4M6d69YB/0U8tPZCQPQ4Y44NEjMyxalIjBg9N0HbJJunpViJEjHSAWM+zdG4c6dQq4A5CdDfuxY2Fx+jQSv/8eqSNGFLhPk63R3r59Oxo2bIjvv/8ew4cPBwDUq1cP+/bt006EhBSCzwcqVpSjYkV5ntqBuDieKgHPaYISFCTE8ePvv8x4PIaKFeXvmp+od8bMM40uIaVYYKAY4eEC/PhjXKlNsnmxsRD99ReE72qtBe+apsjc3JD++efKxLplSyj0NVwtxyFpzhw49ugB8Y4dSJk4UT/HJR/NavNm2CxejAwfH8Rv3w6mYcehcuUUOHYsDmPH2mPWLDu8fCnAnDlJdMc1l19+scC0aXbw9FQO3+fmVkibdjMzxG/eDIwdC9v58wGg0GTb2Gh0qQ0LC0NAQAB4ud4llpaWSEujX2nEsCQSBSSSLDRvrj5bVFoah+fP+QgNVW8HfvmyCFlZ75uhODvLVbXfdepko379LNSoISu1SQgpu2JjlcP5+fpmqN+eNXFcWhqEN2+qmoOYPX4MAFDY2iKzZUukjBuHzNatIa9UyWAjf2Q1bYr0du1gtXkz0gYMgEIiMUgcREOMwXrJElhv2oS0bt2QsG4dIBQWaxdiMcPu3VLMn2+Dbdus8N9/fGzYkAALi7JducMY8MMPVli50gatWmVixw4pbGw0OCcfJtsch9R3Fb/GTqN0wtbWFlFRUWpV469fv6YJbIjRsrRkqFNHludWlEwG/PcfP09nzOPHLbBvn3KILnNzBerUkaF+/Sw0aKBMvitXpvbfxLStWGGN9HQO8+YlGjoUAB/RGVImg1lIiKo5iPD2bXDZ2WBCIbKaNEFSQICynXXdujCmsUOTZ86Eua8vrH74AUkLFhg6HFIQuRz8sWNhvXs3UgcOROLixSV+H/H5wKJFSahUSY7vvrPBF19IEBgohZNT2RyRJDsbmDXLFj/+KEbv3mlYuTKheL9fcpLtMWNgO2+eMtkeNkxn8WqLRol2ly5dsHz5cnTv3h0KhQJ//fUXjh8/ju7du+s4PEK0SyAAqlSRo0oVOdq1e1+rxxgQHs5HSIgQ//xjhpAQMxw8aIldu5TZtY2NAvXqZaNBg/fJd7lyNAkP0bHsbOWvw4/04IEAP/5oiREjUuHpKddCYHrEGATPnqmagoiuXwcvORmM45Bdpw5SRo5EVuvWyGrSRONb+4Ygq14daX37QrxvH1KHD4fcw8PQIZF82M6ZA/6+fUgePx7JM2Zo5S7IiBGpcHeX45tv7NC5syP275fmGRK3tEtJ4fD11/b4809zTJqUjGnTkkt2anPXbM+dCwBGn2xrPLzfrVu38Mcff+Dt27eQSCTw9/dH06ZNdR3fR6HOkNpV1jrUyWTA06cChIQIcfeuGe7eNcOTJ2aQyZRXB2dnOerXz1ar+TbWoQfLWtmZOl5MDKy2boXlvn3gpaeDmZtDIRaD5fpTWFmBWVqCWVmpP/fBY4XYClPnl8f9cDv8dCoD1uUtAXNzo55AhRcdrRrLWnT1KvhRUQAAWcWK7zswtmwJhYODgSMtWH6fOV5UFJxbtkRG+/ZI2LzZQJGRgghv3YJjjx6Qjx+P6IAAre8/JMQMQ4Y4ICODw/btUrRunVX0RqVAVBQPgwZJ8OSJAMuWJaJfPy00O87Kgv2YMbA4exYJixYhbehQAMbZGZLG0SYao2QNSE8HHj0yUyXfISFmCAszUz3v4SFTS77r1s2GWGz4jxiVnWngRUbCassWiA8eBLKykN6tG4R16yLj7VtwKSngUlPBpaaC9+5fLjUVXEoKeGlpyuc1rP1mfL4yEbe0VCbsBSTwykQ9/wSeWVlBkbOepeVHDYfHpaSot7P+918AgMLODpmtWqmSa1OqBS7oM2e9bBmsN2zA29OnkV2/vgEiI/nKzoZThw7gkpKgePAAsenpRW9TAq9f8zFokAOePftgzOhS6t9/BRgwwAGJiTxs2xaPtm212D8kKwv2o0fD4n//Q8LixUgbMsT0Eu3nz59DIBCgYsWKAICkpCTs2bMHr169QrVq1TBo0CCYm5trP2ItoURbuyhZy19SEod799ST74gIZdLB4zFUr66efNesmV3cfjUfhUtKgqRcOcSmpurvoKRY+BERsNq4EZY//QTI5Ujv3RvJ48ZBXqWK5p87xoCsrPdJeEoKZAmpmDHWDPZmyZg/5Q34GWng5UrYudRU5eN3iXqeBL4YyYbC3FyVgKsl8IXUuvMjIpTjWQcHg5PJwMzNkdm0KbLeJdbZtWvDVDtHFFRuXFISnL29IatVC3GHDxv1nYWyRLxtG2wXLIB01y5YDRig0++6pCRlM4orV8wxYUIypk8vYTMKI3ftmhAjRjjAwkI5gU+Bw/d9jA+Sbctp00wr0Z43bx569+6NevXqAQBWrFiB+Ph4tGnTBteuXYOHhwdGGPEQK5Roaxcl2pp7+5b3LukWIiRE2exEKlV2qBEKGWrVykb9+u/bfFetKvvoflu8uDgInj5V/oWFwezdv/yoKDAbG6T26YPUYcMgf/fDmRge/+VLZYL9yy8AgLQvv0TKuHFqZfQxn7tNm6ywZIkNDh2KxaefluA2tVyuTMI/rDnPXav+LlHPk8BrUOuuACCrX/99cxAvL2WzllKgsHIT79wJ2/nzEXfgADLbttVzZORDvDdv4PzZZ8hq3hzSvXvh6OSk8++63B0Du3dPw+rVCaXlrQ8AOHbMAlOm2KFKFRn275fCzU2HfUNyJduyjRsR06OH7o5VgBIn2sOHD8fWrVthZmaG1NRUjBgxAqtXr0b58uURGxuLuXPnYsuWLToJWhso0dYuSrRLjjHlLUNlW29l8n3vnhlSU5W1dWKxsrNl7uS7QgV53loOxsCLjIRZaCgEoaGqpFrw9Cn48fGq1RRiMWTVqqn+xC9egPfLL4BCgYz27ZE6YgSymjWj2jQD4T97BusNG2Bx7BjA5yOtXz+kjB0L+buROHIr6ecuJoaH1q2d4e2dicDA+KI30Idcte7N69ZFIoCHBpyCXZcKLbfMTDh/9hmYWIy3//ufUY2OUhbZf/01zC9cQMzFi5B7eOjtu44x5Y/hpUtt0LRpJnbtioeDg2mPSMIYsGGDFZYvt0GLFpnYtUuqn/kqsrJgN2UKzMaNw9saNXR/vA+UeMIauVwOwbt2d6GhobCzs1PtzNHREal0K5oQjXAc4O4uh7u7HF26ZABQznj57JlAVfN9964Zdu8WIyvLCjzI0cD2GTp4PEBzm4eoiccol/AEFuGh4KWkqParsLNDdvXqyPj8c2VSXb06sj09oShfXi2JNnd0xNupUyHeuxfi/fthceYMsurUQeqIEUjv2hVlfupMPRE8fQqr9ethcfIkIBQidehQpIwZA4Wrq9aPtWKFNTIzOcydm6T1fZcYxwEiERQiEb4LDDR0NIYjEiF5xgzYf/MNLI4dQ/oXXxg6ojJL9OefsDh1Cknffqv3PgAcB4wbl4KKFWWYNMkeXbo4Yv/+OFSpYmIjA70jkylr6Q8eFKNnzzSsWpWgv68WoRAJGzcqh502sgrBQhNtd3d33LhxA97e3rh27Rrq1q2rek4qlcKSppIlpMT4fKB65XTU4h5jkPgpBE6h4LmHQvEwFBb/PYNZYgZwT7nuG5TDTdTEf+JBSKtdHaIGnnBqUwWftLKFTdEzAgMAFOXKITkgACkTJ8Li2DGId+6E/aRJsFm8GKmDByNt4ED9zZZXxggeP4b1Dz/A/NQpMHNzpI4ahZSvv4bC2Vknx3vwQICffrLEqFGpRvul3a5dO0OHYFDpXbtCvG0brFeuRHqXLqWmyYxJyciA7ezZkFWujJQxYwwWRteuGShXLhbDhjmgSxcn7N4tRbNmpjUiSWoqh9Gj7XHxojnGj0/GjBmls915SRTadOTJkydYvnw5AIDH42HhwoWqGu1Tp04hNDQUkydP1k+kJUBNR7SLmo58hPR0CJ49e9/kI+fvxQu1NquyChUgq15d1eQju1o1JLlVR8hLR7U23y9fvv+NXLVqNho0yFYNMVi7dnae7+x8y44xiK5ehXjHDphfvAgmEiGtRw+kDh8OWa1aujwbZYbgwQNYr1sHizNnoLCyQuqQIUgdNapYMwMW93PHGNC7twRPnwrw118x+rltS/LQpNyEV6/CsW9fJM6di9TRo/UUGclhtXYtbFatQtyhQ8j89FPVckN914WH8zFwoASvX/OxZk0CevQwjRFJoqN5GDzYAY8emWHJkkQMGGC4WcNNbtQRAEhPT0dkZCTKlSsHi1yTAbx58wbm5uZwMOJxTCnR1i5KtIvGJSer2kznTqr5//0H7t1HjfF4kFeqhOzq1SHz9HyfWHt6KodJ04BUylN1ssxJvmNilO08BQKGGjVy2nsr23x7e9shIaHgsuOHhcFq925Y/PwzeOnpyqmrR4xApp+fyY76YEhm//wD63XrYH7hAhQ2NkgdPhwpw4eD2dsXe1/F/dydOmWOr792wLJlCRg40HBfeEU5cOAAAGDAgAEGjkQ3NC03h/79Ibx7F9HXroHZ2ek+MAIA4IeHw9nHBxnt2yP+g75mhvyui4/nMGKEA27eFOHbb5MwcWKKUdcMP30qwMCBDpBKedi6NR6+vlocvq8ETDLRNmWUaGsXJdrv8aTS950R3yXTZqGh4EdGqtZhQiFkVatC5umpnlRXrqz1NtGMAZGRPFVHy7t3hbh3zwxJSTmdLRnq189C48ZZ8PLKQqNG+U+uwyUkwPLQIYh374bgzRvIKlVC6vDhSPvySzArK63GXBqZBQUpE+w//4TCzg4pI0cidehQMFsN2/fkozifu4wM4LPPnGFlxfC//7016j52JZ6C3URoWm6Chw/h1L49UkePRtKcOXqIjIAxOAwcCGFQEGIuX87TR8LQ33WZmcC339rh6FFLfPllGpYvL+ZU5Xpy44YQw4c7QCRi2LdPirp1sw0dklEm2iWfYYCQ0o4x8KKjlbXT72qpVSN8xMWpVlNYWEBWrRoyvb3VOiTKPTw+ahKP4uA4oHx5BcqXz8Dnnys7WyoUwIsXymnlHz2ywbVrHDZvtoJcrqweqVo1G40bZ8PLS5mAV68uA8/ODqljxiB15EiYnz4Nq507YTt3LqxXrEDaV18hdehQGh4wH8IbN2C9di1E165B7uCApFmzkDp4sN5/nOzYYYVXrwQ4fDjWqJNsAOjfv7+hQzAKstq1kd6zJ8S7dys/X/mMPEO0y/zMGZhfuoTE777TSUfkjyUSAT/8kIBKlWRYvdoGERF87Nihp9E7NHT8uHL4Pg8P5fB97u7G2RfEGFCNNtGYoX/l64xcDv7r18okOqd2+l1SzUt6P2KDwtZW1W5aNXRe9eqQly9v9M0rcsouLY1DSIgZbt8W4s4dIW7fNkN8vDIjs7FRoGHDLFXy3bBhFmxsGMyCgyHetQsWp04phwfs0EE5PGDTpmV7eEDGILx6FdY//ADRzZuQOzkhZcwYpA0cqHETIE1o+rmLjuahVStnfPqpcpgwYljFuV7yX7+Gc+vWSO/WDQnr1uk2sDKOS02Fc5s2UNjZ4e3Zs/lWhhjTd92RIxaYNs0OlSrJsG+fFBUrGjahzT0kYYsWmdi5Uwo7O+NJI6lGmxBDkcnAj4iAIDwc/BcvIHj3xw8Ph+C//8Blv7/lJXdygszTE+k9eqgl1QpnZ5NPLC0tGVq0yEKLFsoe7Ywpa71zEu87d4RYu1YExjhwHMMnn8jQuHFbNP60JVr0D0fty7shPnAAFqdPI6tePeXwgF26wCjva+oKYxBdugTrdesgvHMHcldXJC5YgNR+/YBc/Vj0bflyG2RnG9lwfkQj8goVkDp0KMTbtyNl1CjqjKxD1mvWgB8ZCemWLXq74/gxevdOR/nycowY4YAuXRwRGChFo0aGaaIhkwFz5thi/34xunVLw9q1ehy+z4RRjTbRmDH9ys+XTKasmf4gmRaEh4P/6pVaMq2wsIC8UiXIKleGrHJl5f89PZHt6QlmxB18S6o4ZZeczOGff8xUifedO0JVW297ezla1E/EcNF++D3YDJuIp5A7O78fHrAYo2mYHMYgOn9emWCHhEDm5oaUb75BWp8+Oh2aTZOyu3fPDJ9/7ojRo1MxZ45pJNpRUVEAAFcjvHWvDcW9XnLx8XBp2RJZjRtDun+/DiMruwRPnsCpXTuk9emDxJUrC1zPGL/rwsIEGDTIAdHRfGzYEK9qIqgvqakcxoyxxx9/mOObb5IREJBslDdyjbFGmxJtLbKZOxeCZ8+gcHaG3MUFChcXyD/4vymPlWoUFx+ZDPxXryAID39fI51TO/3qlfr0zpaWeZPpypUhq1QJChcXk6+dLo6PKTuFQnmRV9Z6K5udhIWZAWDowDuHWZZr0Trlf5AJRIjr2AuYMAzyWjW1+wIMSaGA+ZkzsP7hB5g9fAhZxYpIGT8eab1766Umv6iyYwzo2VOC58+Vw/lZW5vGJZ06Q+ZltXkzbBYvRuzPPyOrZUsdRVZGMQZJz54QhIYi5sqVQitUjOK7Lh9xcTwMGeKAf/4xw5w5Sfj661S9fI3FxCiH73vwwAyLFydi0CDjHc3IGBNt479vYkqEQvCSkpTDucXEqCV9ORR2dsrE29lZmYS7uqon5jkJuQFvQRtcdnbeZDo8HILnz8F//Vo9mRaLlUPl1a6N9E6dlAl1TjJdCpp6GAMeD6heXYbq1WXo10+5LD6eQ3CwEHfueGPm7c+QFvwMI9I3Yshve2D524/4R9IWD/y+hvWXbVG3vsw0385yOcxPnVIm2P/+C1nlyohfuxbpPXoAZmaGjk7lt9/MceuWCCtWJJhMkg0ALi4uhg7B6KQMHQrx7t3KZPvUKaPv+2FKLH75BaJbt5CwapXJ3rWUSBT4+edYTJpkj4ULbREeLsCiRYk6bQETGqocvi82lofdu6Xw9zfs8H2miGq0dUWhAC8+HryoKPBjYsCLjgY/Olrt/7x3j3M3aVBtbmsLubNzvkm4wtVV+a+Li1Y7XRVFq78Us7PB/+8/ZQKd09QjJ7F+9Qqc/H2HD4VYrJZAqyXTTk6UTGtA17/yZTLgyRMBHl5NhdOJA/B5vB3l5a/xFNWwiTcet+v0R62mQtXwguXLK3QWy0eTyWBx8iSs1q+HWVgYsqtVQ8rEicqp6g0wlEdhZZeeDrRp4wxbW4azZ417OL+ypqSfOYvDh2E/ZQqkmzcjo1s3HURW9nDx8XBu0wbySpUQe+JEkT9gjLVGO4dCASxbZo1Nm6zh45OBLVviYWWl/VTu5k3l8H0CgXL4vvr1DT98X1GMsUabEm1DY0yZkOck4R8m5tHR4MXEKBPyzLy/JBXW1qqkO3cTlQ8TcyYWf3SoxX4DZ2WpJdO5m3rwX79WT6atrNST6UqVIK9SRZlMOzpSMv2R9H7xyc5G9uGzsNi+E87PbiOZb4tdGI518vF4iUooV06uGlbQy0s5m6XB+1NmZ8Pi2DFYr18PQXg4smvWRPLEicjo1MmgNYuFld0PP1hhxQob/PJLLLy9TWvK5tKuxJ85uRxO7duDS0tDzJ9/lq2OxjpiGxAAy4MH8fbsWchq1y5yfWNPtHMcPGiJmTNtUb26DPv2xWm1AuPkSXNMmmQPd3cZDhww/GgnmirTifbdu3cRGBgIhUIBX19fdO/eXe35iIgIbN68GS9evEDfvn3RtWtXjbctiEkk2ppiDFxCAvgxMe9rw3OS8Fy14/zoaHAZeTtJKMRi9eT7g8Q85//MyqrApDbfN3BOMp27Rjp3Mq14/8FXWFvn215aXrmyshMdJdM6Y8gvDrM7d94PD8gYntTojL32E3DgxaeIeKNsgmFuzlCvXpbauN5OTnqq9c7KguXPP8Nq40YIXr1CVp06SJk8GRnt2hnFrfuCyi4qiofWrZ3x2WeZ2LGDhvMzNh/zmRP98QckgwYhceFCpA4bpuXIyhazf/6BY5cuSB0+HEnff6/RNqaSaAPA5csijBplDysrhr1741CnTt4mq8XBGLB1qxiLFtmiadNM7N4thb296dTHltlEW6FQYOLEiZgzZw4kEglmzpyJiRMnokKFCqp1EhMT8fbtWwQFBUEsFqsSbU22LUipSrQ1xRi4pKQ8yXd+iTkvPT3P5goLi7xJ+Lt25Faurkh7+FA9mY6IUE+mbWwKTqYdHCiZNhBj+OLgvXkD8d69EB84AF5CArLq1UNE75H4Q9IbQSFWuH1biAcPzJCVpXyPeHjI0Ljx+1rvGjVk2m2LmJEBy59+gtWmTRC8eYOshg2RPGkSMn19jep9WlDZTZpkh5MnLfDnnzHw8DCN2qbcOnToAAA4e/asgSPRjY/6zDEGyRdfQPDvv4i5fh3M2lq7wZUVcjkcO3UCPyYGMZcva3wejeF6WRyPHytHJElI4GHz5vgSt6OWy4G5c22xd68YXbqkY926eJMbv8EYE229dIYMCwuDq6urqvOLt7c3goKC1JJlW1tb2NraIjg4uNjbklw4DszWFjJbW6B69YLXYwxcSkqhteNm9+9DFB0NXtr7Hsa2eDdxS+XKyPLygrx3b/Vk2t7eqJIUYjwU5csjeeZMpEycCIujRyHeuROV543HUJdF+HLwYKTtHYg0Swfcv/9+aMG//hLh2DFlPwRLSwUaNMguchr5onDp6bA8cABWW7aAHx2NzCZNkLhqFTI//dRk3rt375rhl18s8c03ySaZZAPA/fv3DR2C8eI4JM2ZA6dOnWC1ZQuSp083dEQmyXL/fgjv34d08+ZS/WOlZk0ZfvstFkOGOGDYMAcsWJCIoUOLNzJIWhqHsWPtcf68OcaMScGsWUnGcEOvVNBLoi2VSiHJNb6uRCJBaGio1re9cOECLly4AABYtmwZHB0dPyLqMsDJCahcucCnFe/+kJwMREZCkJ4OWYUKwLvyELz7o/HqjZ9AIDCuz8PkyWATJyL7wgXwN2yAzYoVsF6/Hop+/eA+bhw6dlS2o2RMjpcv5fj7bx5u3ODw999m2LxZqJpGvnp1hmbNFGjQgKFBA4Z69RhsbAo4ZkoKeNu3g792LbiYGCjatEH2vn3g2rSBNcfBWL+GPyw7xoBFiwRwcWH4/nsRrK1N8xN448YNADCu96UWffRnzs8P8t69YbVjB0STJwPlymkvuLIgKgpmK1ZA4eMDq2HDYFWMH9FGd73UgKMjcPkyMGgQw5w5doiOtsby5XKNOkhHRwNffSVAcDCHdetkGDNGCMC0Xn8OYyw7vSTa+bVO4TR80xdnWz8/P/j5+akem9KtH6Pn4PD+lgydV5NjtLdCGzUCAgMhCA1VtuP+8UeY7d6NzNatkTJiBDJ9fGBlxYOvL+Drq9wkZxp55RTyQpw5Y4b9+99fyipVkqF27WzUqZON2rWzUa+SFJVP74Z42zbw4+OR8emnSNm6FVnNmik3iIszwAvX3Idld/KkOW7ccMDq1fHIzExHPn2kTULFihUBlN7rtDY+c/xJk+B84gSyZ89G4ooVWoqsbLCbPBlm6el4+913kBfzM26010sNbN4MuLjYYMMGKzx9moWNGxNgaVnwnb+wMD4GDpQgJgbYtUuKdu0yTforvsw2HZFIJIjL9UaPi4uDvb29zrclhJgGWbVqSFy2DEkzZkD8448Q794NyeDBkFWujJQRI5D+xReqkXPym0Y+OpqHhw/N8OCB8u/hQzP89XsmJmATOmEdbBCPmw4dcK33NIh9G6KOUzYqKeQmd2s0PZ3DokU2qFMnC198kbePBSld5JUrI3XgQIj37UPqqFGQeXoaOiSTILx+HZbHjiF54kTIq1Y1dDh6xecDCxYkoVIlOebPt0Hv3hLs2SOFs3PezuW3bgkxdKgD+HyGI0fi0LCh8Q/fZ4r08jVTtWpVREZGIiYmBjKZDNevX4eXl5fOtyWEmBZmb4+Ub75B9M2bkG7eDIWdHexmz4aLlxdsFi4E//XrPNtwHODqqoCvbyYmTkzBzuXPcK/bNMRZVcQCzEdMjRaY6XcVw11/RcAJH4wZ44DWrV1Qo4YruneXYM4cG/z0kwUePBAYfe3w1q1ivHkjwIIFSSY/Zvbq1auxevVqQ4dh9FImTQIzN4f10qWGDsU0ZGXBdtYsyCpWRPL48YaOxmCGDUvFrl1SPH0qQOfOjnjyRL1e9bffzNG3rwQODgr89lssJdk6pLfh/YKDg7F3714oFAq0bdsWPXv2xLlz5wAA7dq1Q0JCAgICApCeng6O42Bubo41a9bA0tIy3201USZHHdEhU76dVtaZctmZ3bkDq507Yf777wBjyOjYESkjRyLby0ut8yIvLg7ibdsg3rMHvNRUpH/+OZInToSsTh3VOpmZwNOnZnj4UKCq/X70yAypqco6BzMzhmrVZKhT533Tk9q1sw0642JO2b15w8OnnzrD1zcT27aZ/nB+NAW75qzWroXNqlV4e+IEsps00co+SyurjRths3Qp4vbsQaa/f4n2YcrXyw/dv2+GwYMdkJbGYfv2eLRunYlt28RYuNAWXl5ZCAyMK1GncmNljE1HaMIaorHSdPEpa0pD2fEiIpTDAx48qBwesEEDpI4YgcymTWG1axcs9+0Dl5GB9K5dkTJhAmQ1ami0X4UCCA/nq5qc5DRBefv2fZVxpUoy1KqVrUrA69TJhouLfsb5zim78ePt8PvvFrh8OQbu7qY50khuObXZU6dONXAkuqHNzxyXlgbnli0h8/BA3PHjJjM6jr7xX7+GU5s2yPzsM8Tv2lXi/ZSG62VuERE8DB4sQWioAJ9+momLF83RqVM6fvghHhYWho5OuyjR1jNKtLWrtF18ypLSVHZcWhosjhyBeOdOmD17BgBgPB7Se/RQJthaascaHc1TJd85/4aHv7/96uQkV+t0Wbt2NipX1n67b0dHR5w7l4guXZwwfnwyAgKStXsAohPa/sxZ7t8Pu4AASHfvRkb79lrbb2liP2wYRFeu4O3ly5C/u2NSEqXpepkjOZnD6NH2+PNPc3z9dQrmzCmdw/dRoq1nlGhrV2m8+JQVpbLsFAqILl+GMDgYaT17Ql7IUJXakpzM4dEj9U6XT58KkJ2trGEUixWqmu/atZVNUKpXz4boI0bgk0gc4e0NvH7Nx9WrMbCyKrWX7FJF6585mQxOPj4Ax+HtH39Au7M3mT7RuXOQDB2KpNmzkTJ27Eftq1ReLwHIZEBYmAA1anzc7JHGzBgTbfqkEkJME4+HzLZtkdm2rd4OaW3N0KxZFpo1y1Ity8wEQkMFarXfP/9sqWr3LRAwVK/+fsjBOnWyUatWNmxsNEuYf/qJh+BgAdasiS9VSfa9e/cAAPXq1TNwJCZCIEDyzJlwGDEClocPI61/f0NHZDS49HTYzpuH7OrVkTJihKHDMVoCAUp1km2sKNEmhJCPIBIBderIUKeODIByyD2FAnj5kq9W8335sgi//GKp2s7DQ32875x237mb36alcZg9m4969UrfcH4dO3YEUHo7Q+pCRocOyPLygvXq1Ujv0QPM0rLojcoAq/XrIXj1CrFHjgBCoaHDIUQNJdqEEKJlPB5QubIclSvL0aVLhmp5TIx6u+8HD8xw+vT73kiOjnK1Nt/BwUJERHDYuLH0taesW7euoUMwPe+mZnfs3h3iHTuQMnGioSMyOH5YGKy2bEFar17IatHC0OEQkge10SYaK63t1soCKjvjlZzM4fHj94n3gwfq7b6/+EKOdeuiDRwlKS5dfubshw2D6No1xFy/DoVEopNjmATGIOnbF2b37iHmyhUonJy0slu6XpouaqNNCCFEjbU1Q9OmWWja9H2776ws4OlTAUJDzdCrlxgK/YwkSExE8syZMPfxgdUPPyBpwQJDh2Mw5r/+CtFffyFh8WKtJdmEaFspuxlJCCGmTyhUtvvu0SMdDg6GjoYYG1m1akj76iuI9+0D/+VLQ4djEFxSEmy/+w5Z9eohbeBAQ4dDSIEo0SaEEKJ3jRo1QqNGjQwdhslKnjIFjM+H9fLlhg7FIKxXrQLv7VskLl0K8PlFb0CIgVCiTQghRO+io6MRHU1tz0tK4eqK1FGjYHnyJMxCQgwdjl4JHjyAODAQaQMHIrtBA0OHQ0ihKNEmhBCid3fu3MGdO3cMHYZJSxk7FnIHB9gsXgyU3nEN1CkUsJs5Ewp7eyTNmGHoaAgpEiXahBBC9M7V1RWurq6GDsOkMWtrpEyaBNG1axD9+aehw9ELy59+gjA4GElz54LZ2Rk6HEKKRIk2IYQQYqJSBw6EzMNDWastlxs6HJ3iSaWwWbwYmc2aIb13b0OHQ4hGKNEmhBCid9OnT8f06dMNHYbpEwqRNGMGzB4/hsWxY4aORqesFy8Gl5KCxCVLoDaFKiFGjBJtQgghenfw4EEcPHjQ0GGUChlduiCrXj1Yr1wJZGQUvYEJEgYFQfzTT0gdORKyGjUMHQ4hGqMJawghhOjd8jI6LJ1O8HhImj0bjn36QLxnD1JHjzZ0RNolk8F25kzIy5VD8uTJho6GkGKhRJsQQojeDRgwwNAhlCpZrVoho21bWG/YgLS+fUtVR0Hx7t0we/wY0h07wMRiQ4dDSLFQ0xFCCCGkFEiaNQtcYiKsN240dChaw4uMhPWqVcjw8UFGx46GDoeQYqNEmxBCiN6dO3cO586dM3QYpYqsVi2k9+oF8e7d4EdEGDocrbD9/ntwcjkSFy2iDpDEJFGiTQghRO+GDh2KoUOHGjqMUif53Ugu1itXGjiSjye6fBkWv/2G5HHjIPfwMHQ4hJQIJdqEEEL0zs/PD35+foYOo9SRu7khdehQWBw5AsGjR4YOp+QyMmA7axZklSsjZcwYQ0dDSIlRok0IIUTv9u7di7179xo6jFIpedw4MBsb2CxdauhQSsxqyxYIwsORuHgxYG5u6HAIKTFKtAkhhJBShNnbI3n8eJhfvAjhtWuGDqfY+OHhsN6wAelduiCzTRtDh0PIR6FEmxBCCCllUocOhax8eeXU7AqFocPRHGOwnTsXTCBA4vz5ho6GkI9GiTYhhBC9c3Nzg5ubm6HDKL3MzZH87bcQhoTA/LffDB2NxszPnoX5xYtInjYNinLlDB0OIR+NEm1CCCGkFErv1QvZNWvCZvlyICvL0OEUiUtNhc28eciuWROpw4YZOhxCtIISbUIIIXoXERGBiFIy1rPR4vORNGsWBC9fQnzggKGjKZL12rUQvHmDxKVLAQFNXE1KB0q0CSGEkFIqs21bZHp7w2rtWnDJyYYOp0CCJ08g3rEDqX37IqtJE0OHQ4jWUKJNCCGElFYch6Q5c8CXSmG1ZYuho8kfY7CdNQvMygrJs2cbOhpCtIoSbUIIIXo3ePBgDB482NBhlAnZ9esjvWtXiLdvBy862tDh5GFx5AhEf/+NpFmzoHBwMHQ4hGgVJdqEEEL07sKFC7hw4YKhwygzkmbMACeTwXr1akOHooZLSIDNwoXIatgQaV99ZehwCNE66m1ACCFE7wIDAw0dQpkir1QJqQMHQrx3L1JHjYLM09PQIQEAbJYvBy8+HnE//gjwqO6PlD70riaEEKJ37dq1Q7t27QwdRpmSMmkSmIUFrI1kanazu3dhuX+/cnKdOnUMHQ4hOqG3Gu27d+8iMDAQCoUCvr6+6N69u9rzjDEEBgbin3/+gUgkwtixY1GlShUAwKlTp3Dx4kVwHAd3d3eMHTsWQqFQX6ETQgghJk8hkSBlzBjYrFyJ1KAgw47uIZfDduZMKJydkfztt4aLgxAd00uNtkKhwK5duzBr1iysXbsW165dw+vXr9XW+eeffxAVFYX169dj1KhR2LlzJwBAKpXizJkzWLZsGVavXg2FQoHr16/rI2xCCCE6cuDAARwwgbGdS5vUUaMgd3GBzaJFAGMGi8Ny/34I791D4vz5YNbWBouDEF3TS6IdFhYGV1dXuLi4QCAQwNvbG0FBQWrr3L59G59++ik4jkP16tWRmpqK+Ph4AMpEPSsrC3K5HFlZWbC3t9dH2IQQQnRkxowZmDFjhqHDKHOYpSWSp0yB8PZtmP/vfwaJgff2LWyWL0dmq1bI6NrVIDEQoi96aToilUohkUhUjyUSCUJDQ/Os4+joqLaOVCpF1apV0aVLF4wZMwZCoRD169dH/fr19RE2IYQQHenfv7+hQyiz0vr2hXjHDlgvXYoMPz+9z8Jos3AhuPR0JCxeDHCcXo9NiL7p5dPF8rk9xX3w4SponZSUFAQFBWHTpk2wtLTEmjVrcOXKFXz66ad51s89XNSyZcvUEnfy8QQCAZ1TE0VlZ7pKa9nt3r3b0CHolNGX25IlMPvySzifOgXFiBF6Oyx35QrMjh6FPCAA9s2b6+24xWH0ZUcKZIxlp5dEWyKRIC4uTvU4Li4uT/MPiUSC2NjYPOvcv38fzs7OsLGxAQA0a9YMT58+zTfR9vPzg5+fn+px7v2Rj+fo6Ejn1ERR2ZkuKjvTZPTl5u0NRy8v8L//HnHt2oFZWur+mFlZcPrmG8jc3fF2xAgwIz0/Rl92pECGKrvy5csX+Jxe2mhXrVoVkZGRiImJgUwmw/Xr1+Hl5aW2jpeXF65cuQLGGJ4+fQpLS0vY29vD0dERoaGhyMzMBGMM9+/fh5ubmz7CJoQQoiNRUVGIiooydBhlF8chcc4c8GNiIN6+XS+HtNq5E2ZPnyJx4UIwCwu9HJMQQ9NLjTafz8ewYcOwePFiKBQKtG3bFu7u7jh37hwA5XiqDRs2RHBwMCZMmAChUIixY8cCAKpVq4bmzZtjxowZ4PP5qFSpklqtNSGEENPTuHFjAEBERISBIym7sps0QXqHDrDasgVpAwdCkasvlbbxX7+G1Zo1SG/fHpn+/jo7DiHGhmP5NY4uJd68eWPoEEoVup1muqjsTFdpLbtGjRoBAIKDgw0ciW6YSrkJwsLg5OOD1MGDkbRwoc6OYz98OER//om3ly9DXqGCzo6jDaZSdiSvMtt0hBBCCMktODi41CbZpkTm6akchWT/fvDDw3VyDNH587A4exYpkycbfZJNiLZRok0IIYSUYclTp4IJBLBZvlzr++bS02E7dy6yPT2RMmqU1vdPiLGjRJsQQggpwxQuLkgdNQoWv/4Ks7t3tbpvqw0bIHj1ColLlgBCoVb3TYgpoESbEEKI3nXo0AEdOnQwdBjknZQxYyB3cIDN4sVam5qdHxam7GjZsyeyWrbUyj4JMTWUaBNCCNG7+/fv4/79+4YOg7zDrK2RMnkyRNevQ3TpkhZ2yGA3Zw6YSISkuXM/fn+EmChKtAkhhOjdmTNncObMGUOHQXJJHTAAskqVYLNkCSCXf9S+zH/9FaKrV5E0YwYUzs5aipAQ00OJNiGEEL2rV68e6tWrZ+gwSG5CIZKmT4fZ48ewOHq0xLvhkpNh+/33yKpbF2mDBmkxQEJMDyXahBBCCAEAZHTpgqz69WG9ciWQkVGifVivWgVeTAwSly4F+HwtR0iIaaFEmxBCiN6tXr0aq1evNnQY5EM8HpJmz4bgzRuIAwOLvbngwQOId+9G2oAByG7YUAcBEmJaKNEmhBCid2vWrMGaNWsMHQbJR1bLlsjw8YH1hg3g4uM131ChgN2sWVDY2yMpIEB3ARJiQijRJoQQondTpkzBlClTDB0GKUDSzJngkpJgvXGjxttYHj4M4Z07SJozB8zOTnfBEWJCBIYOgBBCSNkzdepUQ4dACiGrVQvpvXtDHBiI1GHDIHdzK3R9nlQKm0WLkNmsGdK/+EJPURJi/KhGmxBCCCF5JH/7LQDAesWKIte1XrIEXHKycgZIjtN1aISYDEq0CSGE6N29e/dw7949Q4dBCiF3c0PqsGGwOHoUgocPC1zPLCgI4kOHkDpyJGQ1augxQkKMHyXahBBC9K5jx47o2LGjocMgRUgeNw7M1hY2S5fmv4JMBruZM/H/9u49KMp6j+P4ey9cAhRYtEyIPCZUaqaIFzAwimzKmhxzTCdNcpimoJrpotBltJsVOuVkg9Kkboo2NVOmNqWHmI6o1UyKWhpZ0HhpwhssyM1Ly+75o+MeENak2n126/P6i93nYZ/P7HfAjz9+u9verx/N2nMv0oWKtoiI+N11113HddddZ3QM+R3umBiaH3mE8P/8h9Dt27scj7TbCfn+e06+8ALuyEgDEooENhVtERHxu82bN7N582ajY8hFaM3JwRkfT+8FC8Dl8txvPnKEXosWcTori9O3325gQpHApaItIiIi3oWH0zxnDqHffkv4xx977o5+4QVMTicnX3xRL4AU8UJFW0RERC7o1OTJ/HrttfQuKoKzZwnbupVLNm6k+eGHaf/Xv4yOJxKwVLRFRMTvUlJSSElJMTqGXCyL5bePZj90iMgVK4h++mmcAwbQkpdndDKRgKYPrBEREb87duyY0RGkh87ceCNnxo0j+qWXAKhfswbCww1OJRLYtKItIiJ+V1lZSWVlpdExpCdMJpqefRaAUxMnciYry+BAIoFPK9oiIuJ3/fr1MzqC/AG/DhvG8X//m/arrjI6ikhQUNEWERGRi+YcOtToCCJBQ1tHRETE7+bOncvcuXONjiEi4lMq2iIi4ndr165l7dq1RscQEfEpbR0RERG/KyoqMjqCiIjPqWiLiIjfzZgxw+gIIiI+p60jIiIiIiI+oKItIiJ+V1ZWRllZmdExRER8SltHRETE7+6//34AfvnlF4OTiIj4joq2iIj4XXZ2ttERRER8TkVbRET8btWqVUZHEBHxOe3RFhERERHxAb+taO/Zswe73Y7L5eLmm29m0qRJnY673W7sdju7d+8mLCyMvLw8Bg4cCEBrayslJSX8/PPPmEwmHnroIZKTk/0VXURERESkx/xStF0uFytWrODZZ58lLi6Op556itTUVBISEjzn7N69m6NHj7JkyRKqq6tZvnw5L7/8MgB2u53hw4fzxBNP4HQ6OXPmjD9ii4iIj8THxwN6MaSI/L35ZetITU0N/fr147LLLsNqtZKens6OHTs6nbNz504yMzMxmUwkJyfT2tpKQ0MDbW1tfP/999x0000AWK1WIiMj/RFbREREROQP88uKtsPhIC4uznM7Li6O6urqLuf06dOn0zkOhwOLxULv3r1ZunQphw4dYuDAgeTk5BAeHt7lOuXl5ZSXlwPw6quvdno8+fOsVque0yCl2QWvv+vs/u5/mfy7zu2fQLMLXoE4O78Ubbfb3eU+k8l0Uee0t7dz4MABZs+eTVJSEna7nfXr1zNt2rQu52dnZ3d6y6i6urq/IL2c06dPHz2nQUqzC16aXXDS3IKXZhe8jJpd//79vR7zy9aRuLg46uvrPbfr6+uJjY3tck7HJ+fcOXFxccTFxZGUlATA2LFjOXDggD9ii4iIiIj8YX4p2ldddRVHjhzh+PHjOJ1OvvzyS1JTUzudk5qaytatW3G73fz4449EREQQGxtLTEwMcXFx1NbWArB3795OL6IUEREREQlEftk6YrFYmD17NgsWLMDlcpGVlcUVV1xBWVkZABMmTGDEiBHs2rWLRx99lNDQUPLy8jzfP3v2bJYsWYLT6eTSSy/tdExEREREJBCZ3N1tjv6bOLcKLn8N7VsLXppd8NLsgpPmFrw0u+D1j92jLSIiIiLyT6OiLSIiIiLiAyraIiIiIiI+oKItIiIiIuIDKtoiIiIiIj6goi0iIiIi4gMq2iIiIiIiPqCiLSIiIiLiAyraIiIiIiI+8Lf+ZEgREREREaNoRVsuWmFhodER5A/S7IKXZhecNLfgpdkFr0CcnYq2iIiIiIgPqGiLiIiIiPiAirZctOzsbKMjyB+k2QUvzS44aW7BS7MLXoE4O70YUkRERETEB7SiLSIiIiLiA1ajA0hgqquro7i4mMbGRkwmE9nZ2dx+++20tLSwePFiTpw4Qd++fXnssceIiooyOq78j7e5lZaWUllZidVq5bLLLiMvL4/IyEij40oH3mZ3zsaNG1mzZg3Lly+nd+/eBiaV811odps2bWLz5s1YLBZSUlKYMWOGwWnlHG9zO3jwIG+//TZnz57FYrGQm5vLoEGDjI4rHZw9e5b58+fjdDppb29n7NixTJ06NSA7iraOSLcaGhpoaGhg4MCBnDp1isLCQubMmcOWLVuIiopi0qRJrF+/npaWFv3DEUC8za2+vp6hQ4disVhYs2YNgOYWYLzNLiEhgbq6Ot566y1++eUXXn31VRXtAONtdo2NjXz00UcUFhYSEhLCyZMniY6ONjqu/I+3ub3zzjtMnDiRESNGsGvXLjZu3Mhzzz1ndFzpwO12c+bMGcLDw3E6ncybN4+cnBy+/vrrgOso2joi3YqNjWXgwIEAXHLJJcTHx+NwONixYwfjx48HYPz48ezYscPImHIeb3O7/vrrsVgsACQnJ+NwOIyMKd3wNjuAVatWce+992IymYyMKF54m11ZWRl33XUXISEhACrZAcbb3EwmE6dOnQKgra2N2NhYI2NKN0wmE+Hh4QC0t7fT3t6OyWQKyI6irSPyu44fP86BAwcYNGgQJ0+e9PzSiY2NpampyeB04k3HuXX0+eefk56eblAquRgdZ7dz505sNhsDBgwwOpZchI6zKy0tZf/+/bz33nuEhIQwc+ZMbUEIUB3nNmvWLBYsWEBpaSkul4uXXnrJ6HjSDZfLRUFBAUePHuXWW28lKSkpIDuKVrTlgk6fPs1rr71GTk4OERERRseRi+RtbuvWrcNisZCRkWFgOrmQjrOzWCysW7eOe+65x+hYchHO/7lzuVy0tLSwYMECZs6cyeLFi9FuzcBz/tzKysqYNWsWy5YtY9asWZSUlBgdUbphNptZtGgRJSUl/PTTTxw+fNjoSN1S0RavnE4nr732GhkZGYwZMwb47U+fDQ0NwG/727RXNPB0NzeALVu2UFlZyaOPPqotCAHq/NkdO3aM48ePM2fOHPLz86mvr6egoIDGxkajo8p5uvu5s9lsjBkzBpPJxKBBgzCbzTQ3NxucVDrqbm4VFRWer9PS0qipqTEyovyOyMhIBg8ezJ49ewKyo6hoS7fcbjclJSXEx8dzxx13eO5PTU2loqIC+O2X0ahRo4yKKN3wNrc9e/awYcMGCgoKCAsLMzCheNPd7BITE1m+fDnFxcUUFxcTFxdHUVERMTExxoaVTrz93I0aNYp9+/YBUFtbi9PppFevXkbFlPN4m5vNZqOqqgqAffv20a9fP6MiihdNTU20trYCv70Dyd69e4mPjw/IjqJ3HZFu7d+/n3nz5pGYmOhZ/Zw+fTpJSUksXryYuro6+vTpw+OPP274W+fI/3mbm91ux+l0emaVlJTEAw88YGRUOY+32aWkpHjOyc/P55VXXgmIVRr5P2+zGzZsGEuXLuXQoUNYrVZmzpzJ0KFDDU4r53ibW0REBHa7HZfLRUhICLm5uZ4XTUpgOHToEMXFxbhcLtxuN2lpaUyZMoXm5uaA6ygq2iIiIiIiPqCtIyIiIiIiPqCiLSIiIiLiAyraIiIiIiI+oKItIiIiIuIDKtoiIiIiIj6goi0i8g9VW1vL3Llzue+++/j000+NjuORn5/Pt99+a3QMEZE/zWp0ABGRf5L8/HwaGxuxWCyYzWYSEhLIzMwkOzsbs9m/ax8bNmxg8ODBLFy4sNvjzz33HFVVVSxcuJABAwZ47l+4cCE7d+5k/vz5DBky5E9lOPdBPNOmTftTjyMiEohUtEVE/KygoIBhw4bR1tZGVVUVdrudmpoa8vLy/Jqjrq6O9PT0C55z+eWXs3XrVk/Rbm5uprq6Wh+aIyJyEVS0RUQMEhERQWpqKjExMTzzzDPccccdJCYmsmvXLt577z2OHTtGREQEWVlZTJ06FYBXXnmF4cOHc9ttt3ke58knn2Tq1KmMHj26yzV27tzJu+++i8PhYMCAAeTm5pKQkMDzzz9PVVUV+/fv55133qGoqIj+/ft3+f6MjAzKy8uZMWMGZrOZL774gtGjR1NZWek559dff2Xt2rV89dVXAKSlpXHvvfcSEhLCd999x5tvvsnEiRPZsGEDZrOZ6dOnk5WVRXl5Odu3bwfgk08+YciQIRQWFgJw8OBBVq9ezYkTJxg+fDj5+fmEhob+dU++iIgfaI+2iIjBBg0ahM1mY//+/QCEhYXx8MMPY7fbKSws5LPPPuPrr78GYPz48Wzbts3zvQcPHsThcHT6qPZzamtreeONN8jJyWH58uWMGDGCoqIinE4n8+fP59prr2X27NmUlpZ2W7IBYmNjSUhI4JtvvgGgoqKC8ePHdzpn3bp1VFdXs3DhQhYtWkRNTQ0ffvih53hjYyNtbW2UlJTw4IMPsmLFClpaWsjOzuaGG27grrvuorS01FOyAb766iuefvppiouLOXz4MFu2bPljT66IiIFUtEVEAoDNZqOlpQWAIUOGkJiYiNls5sorr2TcuHFUVVUBMGrUKI4cOcKRI0cA2Lp1K+np6VitXf9A+eWXXzJixAiGDRuG1Wrlzjvv5OzZs/zwww89ypaZmUlFRQW1tbW0tbWRnJzc6fj27du5++67iY6Opnfv3kyZMqXTfwYsFgtTpkzBarWSkpJCeHg4tbW1F7zmbbfdhs1mIyoqipEjR3Lw4MEeZRYRCQTaOiIiEgAcDgdRUVEAVFdX8+6773L48GGcTidOp5OxY8cCEBISQlpaGtu2bWPKlCl88cUXPPHEE90+ZkNDA3379vXcNpvN9OnTB4fD0aNsY8aMobS0lE2bNpGRkdFt9o7X6du3b6dr9OrVC4vF4rkdFhbG6dOnL3jNmJgYz9ehoaE9ziwiEgi0oi0iYrCamhocDgfXXHMNAEuWLGHkyJEsW7aMVatWccstt+B2uz3n33jjjWzbto19+/YRFhbWZYX5nNjYWE6cOOG57Xa7qaurw2az9ShfWFgYw4cPp6ysjMzMzC7HbTZbp+v05Bomk6lHWUREgomKtoiIQdra2qisrOSNN94gIyODxMREAE6dOkVUVBShoaHU1NR4XjB4TnJyMmazmdWrV3dbfM9JT09n9+7d7N27F6fTyccff0xISAhXX311j7NOnz6d559/nksvvbTLsXHjxrFu3Tqamppoamrigw8+6HbluzvR0dEcO3asx3lERIKBto6IiPhZUVERFosFk8lEQkICEydOZMKECZ7jubm5rF69mpUrVzJ48GDS0tJobW3t9BiZmZm8//77zJkzx+t1+vfvzyOPPMLKlSs97zpSUFDQ7X7u32Oz2byuUk+ePJm2tjaefPJJAMaOHcvkyZMv6nFvuukmXn/9dXJychg8eDBz587tcTYRkUBlcnf8e6SIiASFiooKysvLefHFF42OIiIiXmjriIhIkDlz5gxlZWVkZ2cbHUVERC5ARVtEJIjs2bOH3NxcoqOjueGGG4yOIyIiF6CtIyIiIiIiPqAVbRERERERH1DRFhERERHxARVtEREREREfUNEWEREREfEBFW0RERERER9Q0RYRERER8YH/AqHadopV8/x8AAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<Figure size 864x432 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Plot real vs. synthetic compound sentiment score\n",
    "plt.figure(figsize=(12,6))\n",
    "plt.plot(df3.query('Twitter')['Day'], df3.query('Twitter')['compound'], label='Twitter', color='blue')\n",
    "plt.plot(df3.query('Twitter')['Day'], twitter_synth, label='Synthetic Control', color='red')\n",
    "plt.vlines(x=25, ymin=.06, ymax=.2, linestyle=\":\", lw=2, label='Elon Musk Buys Twitter', color='black')\n",
    "plt.ylabel('Sentiment Score')\n",
    "plt.title('Twitter Compound Sentiment Score vs. Synthetic Twitter (Round 2)')\n",
    "plt.xlabel('Day of Month')\n",
    "plt.legend();"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Conclusions from Sentiment Score Synthetic Control\n",
    "- Overall, tweet sentiment scores from other companies did not provide us with a great Synthetic Twitter Sentiment.\n",
    "- Since sentiment is a calculated value itself, it is more difficult to make a Synthetic representation of it.\n",
    "- More control groups or more tweets may have helped create a more accurate representation.\n",
    "- The second round of Synthetic Control shows some divergence after the purchase, which is what we are looking for, but it does not allow us to clearly see the effect of the purchase on sentiment.\n",
    "- Now time to try stock price."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Synthetic Control for Stock Price\n",
    "- Not all companies used prior are publicly traded, so new companies had to be chosen to make up the Synthetic Twitter stock price.\n",
    "- The actual Twitter stock price will be plotted against the weighted combination of Facebook, Snapchat, and Pinterest."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 107,
   "metadata": {},
   "outputs": [],
   "source": [
    "tickers = {'Twitter':'TWTR', 'Facebook':'FB', 'Snapchat':'SNAP', 'Pinterest':'PINS'}"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 108,
   "metadata": {},
   "outputs": [],
   "source": [
    "# obtaining stock history from 4/19/22 - 4/29/22.\n",
    "import yfinance as yf\n",
    "\n",
    "stocks = pd.DataFrame()\n",
    "for company, ticker in tickers.items():\n",
    "    stock = yf.Ticker(ticker)\n",
    "\n",
    "    # get historical market data\n",
    "    hist = stock.history(start='2022-04-19', end='2022-04-30')\n",
    "    hist['Day'] = [19, 20, 21, 22, 25, 26, 27, 28, 29]\n",
    "    hist['company'] = company\n",
    "    stocks = pd.concat([stocks, hist])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 109,
   "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>Open</th>\n",
       "      <th>High</th>\n",
       "      <th>Low</th>\n",
       "      <th>Close</th>\n",
       "      <th>Volume</th>\n",
       "      <th>Dividends</th>\n",
       "      <th>Stock Splits</th>\n",
       "      <th>Day</th>\n",
       "      <th>company</th>\n",
       "      <th>Twitter</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",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>2022-04-19</th>\n",
       "      <td>47.299999</td>\n",
       "      <td>48.500000</td>\n",
       "      <td>45.000000</td>\n",
       "      <td>46.160000</td>\n",
       "      <td>125776000</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>19</td>\n",
       "      <td>Twitter</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2022-04-20</th>\n",
       "      <td>46.340000</td>\n",
       "      <td>47.110001</td>\n",
       "      <td>45.029999</td>\n",
       "      <td>46.720001</td>\n",
       "      <td>63360000</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>20</td>\n",
       "      <td>Twitter</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2022-04-21</th>\n",
       "      <td>45.509998</td>\n",
       "      <td>47.939999</td>\n",
       "      <td>45.500000</td>\n",
       "      <td>47.080002</td>\n",
       "      <td>70413600</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>21</td>\n",
       "      <td>Twitter</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2022-04-22</th>\n",
       "      <td>47.770000</td>\n",
       "      <td>49.730000</td>\n",
       "      <td>47.490002</td>\n",
       "      <td>48.930000</td>\n",
       "      <td>76504500</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>22</td>\n",
       "      <td>Twitter</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2022-04-25</th>\n",
       "      <td>51.020000</td>\n",
       "      <td>52.290001</td>\n",
       "      <td>50.240002</td>\n",
       "      <td>51.700001</td>\n",
       "      <td>176803600</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>25</td>\n",
       "      <td>Twitter</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2022-04-26</th>\n",
       "      <td>51.570000</td>\n",
       "      <td>51.619999</td>\n",
       "      <td>49.430000</td>\n",
       "      <td>49.680000</td>\n",
       "      <td>115020400</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>26</td>\n",
       "      <td>Twitter</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2022-04-27</th>\n",
       "      <td>49.060001</td>\n",
       "      <td>49.259998</td>\n",
       "      <td>47.820000</td>\n",
       "      <td>48.639999</td>\n",
       "      <td>59106900</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>27</td>\n",
       "      <td>Twitter</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2022-04-28</th>\n",
       "      <td>49.009998</td>\n",
       "      <td>49.919998</td>\n",
       "      <td>47.959999</td>\n",
       "      <td>49.110001</td>\n",
       "      <td>43816900</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>28</td>\n",
       "      <td>Twitter</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2022-04-29</th>\n",
       "      <td>49.450001</td>\n",
       "      <td>50.250000</td>\n",
       "      <td>48.740002</td>\n",
       "      <td>49.020000</td>\n",
       "      <td>38198200</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>29</td>\n",
       "      <td>Twitter</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2022-04-19</th>\n",
       "      <td>210.520004</td>\n",
       "      <td>218.429993</td>\n",
       "      <td>209.000000</td>\n",
       "      <td>217.309998</td>\n",
       "      <td>20089900</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>19</td>\n",
       "      <td>Facebook</td>\n",
       "      <td>False</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2022-04-20</th>\n",
       "      <td>213.610001</td>\n",
       "      <td>214.160004</td>\n",
       "      <td>198.919998</td>\n",
       "      <td>200.419998</td>\n",
       "      <td>43036500</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>20</td>\n",
       "      <td>Facebook</td>\n",
       "      <td>False</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2022-04-21</th>\n",
       "      <td>201.600006</td>\n",
       "      <td>202.970001</td>\n",
       "      <td>185.580002</td>\n",
       "      <td>188.070007</td>\n",
       "      <td>49954100</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>21</td>\n",
       "      <td>Facebook</td>\n",
       "      <td>False</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2022-04-22</th>\n",
       "      <td>190.369995</td>\n",
       "      <td>192.720001</td>\n",
       "      <td>183.270004</td>\n",
       "      <td>184.110001</td>\n",
       "      <td>36426800</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>22</td>\n",
       "      <td>Facebook</td>\n",
       "      <td>False</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2022-04-25</th>\n",
       "      <td>182.729996</td>\n",
       "      <td>187.179993</td>\n",
       "      <td>181.660004</td>\n",
       "      <td>186.990005</td>\n",
       "      <td>29281100</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>25</td>\n",
       "      <td>Facebook</td>\n",
       "      <td>False</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2022-04-26</th>\n",
       "      <td>186.630005</td>\n",
       "      <td>187.190002</td>\n",
       "      <td>179.500000</td>\n",
       "      <td>180.949997</td>\n",
       "      <td>31747400</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>26</td>\n",
       "      <td>Facebook</td>\n",
       "      <td>False</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2022-04-27</th>\n",
       "      <td>174.429993</td>\n",
       "      <td>181.210007</td>\n",
       "      <td>169.000000</td>\n",
       "      <td>174.949997</td>\n",
       "      <td>71062900</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>27</td>\n",
       "      <td>Facebook</td>\n",
       "      <td>False</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2022-04-28</th>\n",
       "      <td>202.919998</td>\n",
       "      <td>208.529999</td>\n",
       "      <td>192.899994</td>\n",
       "      <td>205.729996</td>\n",
       "      <td>100890600</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>28</td>\n",
       "      <td>Facebook</td>\n",
       "      <td>False</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2022-04-29</th>\n",
       "      <td>204.460007</td>\n",
       "      <td>212.479996</td>\n",
       "      <td>199.899994</td>\n",
       "      <td>200.470001</td>\n",
       "      <td>49137200</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>29</td>\n",
       "      <td>Facebook</td>\n",
       "      <td>False</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2022-04-19</th>\n",
       "      <td>32.369999</td>\n",
       "      <td>34.240002</td>\n",
       "      <td>32.320000</td>\n",
       "      <td>33.630001</td>\n",
       "      <td>29558200</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>19</td>\n",
       "      <td>Snapchat</td>\n",
       "      <td>False</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2022-04-20</th>\n",
       "      <td>32.930000</td>\n",
       "      <td>33.180000</td>\n",
       "      <td>30.700001</td>\n",
       "      <td>30.760000</td>\n",
       "      <td>38448300</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>20</td>\n",
       "      <td>Snapchat</td>\n",
       "      <td>False</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2022-04-21</th>\n",
       "      <td>31.365000</td>\n",
       "      <td>31.750000</td>\n",
       "      <td>28.754999</td>\n",
       "      <td>29.420000</td>\n",
       "      <td>69311500</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>21</td>\n",
       "      <td>Snapchat</td>\n",
       "      <td>False</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2022-04-22</th>\n",
       "      <td>28.670000</td>\n",
       "      <td>31.209999</td>\n",
       "      <td>28.559999</td>\n",
       "      <td>29.760000</td>\n",
       "      <td>77506100</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>22</td>\n",
       "      <td>Snapchat</td>\n",
       "      <td>False</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2022-04-25</th>\n",
       "      <td>29.280001</td>\n",
       "      <td>30.790001</td>\n",
       "      <td>28.910000</td>\n",
       "      <td>29.910000</td>\n",
       "      <td>42647800</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>25</td>\n",
       "      <td>Snapchat</td>\n",
       "      <td>False</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2022-04-26</th>\n",
       "      <td>29.520000</td>\n",
       "      <td>29.719999</td>\n",
       "      <td>28.150000</td>\n",
       "      <td>28.690001</td>\n",
       "      <td>34701800</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>26</td>\n",
       "      <td>Snapchat</td>\n",
       "      <td>False</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2022-04-27</th>\n",
       "      <td>27.969999</td>\n",
       "      <td>28.990000</td>\n",
       "      <td>26.559999</td>\n",
       "      <td>27.090000</td>\n",
       "      <td>40794900</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>27</td>\n",
       "      <td>Snapchat</td>\n",
       "      <td>False</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2022-04-28</th>\n",
       "      <td>27.910000</td>\n",
       "      <td>28.959999</td>\n",
       "      <td>26.440001</td>\n",
       "      <td>28.809999</td>\n",
       "      <td>44665200</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>28</td>\n",
       "      <td>Snapchat</td>\n",
       "      <td>False</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2022-04-29</th>\n",
       "      <td>29.070000</td>\n",
       "      <td>30.370001</td>\n",
       "      <td>28.400000</td>\n",
       "      <td>28.459999</td>\n",
       "      <td>39017400</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>29</td>\n",
       "      <td>Snapchat</td>\n",
       "      <td>False</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2022-04-19</th>\n",
       "      <td>21.620001</td>\n",
       "      <td>22.969999</td>\n",
       "      <td>21.360001</td>\n",
       "      <td>22.719999</td>\n",
       "      <td>8948400</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>19</td>\n",
       "      <td>Pinterest</td>\n",
       "      <td>False</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2022-04-20</th>\n",
       "      <td>22.370001</td>\n",
       "      <td>22.389999</td>\n",
       "      <td>21.030001</td>\n",
       "      <td>21.040001</td>\n",
       "      <td>14979700</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>20</td>\n",
       "      <td>Pinterest</td>\n",
       "      <td>False</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2022-04-21</th>\n",
       "      <td>21.150000</td>\n",
       "      <td>21.379999</td>\n",
       "      <td>19.870001</td>\n",
       "      <td>20.230000</td>\n",
       "      <td>16627300</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>21</td>\n",
       "      <td>Pinterest</td>\n",
       "      <td>False</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2022-04-22</th>\n",
       "      <td>19.910000</td>\n",
       "      <td>20.469999</td>\n",
       "      <td>19.469999</td>\n",
       "      <td>19.590000</td>\n",
       "      <td>14250800</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>22</td>\n",
       "      <td>Pinterest</td>\n",
       "      <td>False</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2022-04-25</th>\n",
       "      <td>19.469999</td>\n",
       "      <td>20.115000</td>\n",
       "      <td>19.299999</td>\n",
       "      <td>20.000000</td>\n",
       "      <td>12583000</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>25</td>\n",
       "      <td>Pinterest</td>\n",
       "      <td>False</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2022-04-26</th>\n",
       "      <td>19.860001</td>\n",
       "      <td>19.879999</td>\n",
       "      <td>18.910000</td>\n",
       "      <td>19.219999</td>\n",
       "      <td>13351700</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>26</td>\n",
       "      <td>Pinterest</td>\n",
       "      <td>False</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2022-04-27</th>\n",
       "      <td>18.870001</td>\n",
       "      <td>19.680000</td>\n",
       "      <td>18.320000</td>\n",
       "      <td>18.670000</td>\n",
       "      <td>27148800</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>27</td>\n",
       "      <td>Pinterest</td>\n",
       "      <td>False</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2022-04-28</th>\n",
       "      <td>20.200001</td>\n",
       "      <td>21.264999</td>\n",
       "      <td>19.049999</td>\n",
       "      <td>21.200001</td>\n",
       "      <td>30332400</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>28</td>\n",
       "      <td>Pinterest</td>\n",
       "      <td>False</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2022-04-29</th>\n",
       "      <td>21.280001</td>\n",
       "      <td>21.719999</td>\n",
       "      <td>20.490000</td>\n",
       "      <td>20.520000</td>\n",
       "      <td>14946400</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>29</td>\n",
       "      <td>Pinterest</td>\n",
       "      <td>False</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                  Open        High         Low       Close     Volume  \\\n",
       "Date                                                                    \n",
       "2022-04-19   47.299999   48.500000   45.000000   46.160000  125776000   \n",
       "2022-04-20   46.340000   47.110001   45.029999   46.720001   63360000   \n",
       "2022-04-21   45.509998   47.939999   45.500000   47.080002   70413600   \n",
       "2022-04-22   47.770000   49.730000   47.490002   48.930000   76504500   \n",
       "2022-04-25   51.020000   52.290001   50.240002   51.700001  176803600   \n",
       "2022-04-26   51.570000   51.619999   49.430000   49.680000  115020400   \n",
       "2022-04-27   49.060001   49.259998   47.820000   48.639999   59106900   \n",
       "2022-04-28   49.009998   49.919998   47.959999   49.110001   43816900   \n",
       "2022-04-29   49.450001   50.250000   48.740002   49.020000   38198200   \n",
       "2022-04-19  210.520004  218.429993  209.000000  217.309998   20089900   \n",
       "2022-04-20  213.610001  214.160004  198.919998  200.419998   43036500   \n",
       "2022-04-21  201.600006  202.970001  185.580002  188.070007   49954100   \n",
       "2022-04-22  190.369995  192.720001  183.270004  184.110001   36426800   \n",
       "2022-04-25  182.729996  187.179993  181.660004  186.990005   29281100   \n",
       "2022-04-26  186.630005  187.190002  179.500000  180.949997   31747400   \n",
       "2022-04-27  174.429993  181.210007  169.000000  174.949997   71062900   \n",
       "2022-04-28  202.919998  208.529999  192.899994  205.729996  100890600   \n",
       "2022-04-29  204.460007  212.479996  199.899994  200.470001   49137200   \n",
       "2022-04-19   32.369999   34.240002   32.320000   33.630001   29558200   \n",
       "2022-04-20   32.930000   33.180000   30.700001   30.760000   38448300   \n",
       "2022-04-21   31.365000   31.750000   28.754999   29.420000   69311500   \n",
       "2022-04-22   28.670000   31.209999   28.559999   29.760000   77506100   \n",
       "2022-04-25   29.280001   30.790001   28.910000   29.910000   42647800   \n",
       "2022-04-26   29.520000   29.719999   28.150000   28.690001   34701800   \n",
       "2022-04-27   27.969999   28.990000   26.559999   27.090000   40794900   \n",
       "2022-04-28   27.910000   28.959999   26.440001   28.809999   44665200   \n",
       "2022-04-29   29.070000   30.370001   28.400000   28.459999   39017400   \n",
       "2022-04-19   21.620001   22.969999   21.360001   22.719999    8948400   \n",
       "2022-04-20   22.370001   22.389999   21.030001   21.040001   14979700   \n",
       "2022-04-21   21.150000   21.379999   19.870001   20.230000   16627300   \n",
       "2022-04-22   19.910000   20.469999   19.469999   19.590000   14250800   \n",
       "2022-04-25   19.469999   20.115000   19.299999   20.000000   12583000   \n",
       "2022-04-26   19.860001   19.879999   18.910000   19.219999   13351700   \n",
       "2022-04-27   18.870001   19.680000   18.320000   18.670000   27148800   \n",
       "2022-04-28   20.200001   21.264999   19.049999   21.200001   30332400   \n",
       "2022-04-29   21.280001   21.719999   20.490000   20.520000   14946400   \n",
       "\n",
       "            Dividends  Stock Splits  Day    company  Twitter  \n",
       "Date                                                          \n",
       "2022-04-19          0             0   19    Twitter     True  \n",
       "2022-04-20          0             0   20    Twitter     True  \n",
       "2022-04-21          0             0   21    Twitter     True  \n",
       "2022-04-22          0             0   22    Twitter     True  \n",
       "2022-04-25          0             0   25    Twitter     True  \n",
       "2022-04-26          0             0   26    Twitter     True  \n",
       "2022-04-27          0             0   27    Twitter     True  \n",
       "2022-04-28          0             0   28    Twitter     True  \n",
       "2022-04-29          0             0   29    Twitter     True  \n",
       "2022-04-19          0             0   19   Facebook    False  \n",
       "2022-04-20          0             0   20   Facebook    False  \n",
       "2022-04-21          0             0   21   Facebook    False  \n",
       "2022-04-22          0             0   22   Facebook    False  \n",
       "2022-04-25          0             0   25   Facebook    False  \n",
       "2022-04-26          0             0   26   Facebook    False  \n",
       "2022-04-27          0             0   27   Facebook    False  \n",
       "2022-04-28          0             0   28   Facebook    False  \n",
       "2022-04-29          0             0   29   Facebook    False  \n",
       "2022-04-19          0             0   19   Snapchat    False  \n",
       "2022-04-20          0             0   20   Snapchat    False  \n",
       "2022-04-21          0             0   21   Snapchat    False  \n",
       "2022-04-22          0             0   22   Snapchat    False  \n",
       "2022-04-25          0             0   25   Snapchat    False  \n",
       "2022-04-26          0             0   26   Snapchat    False  \n",
       "2022-04-27          0             0   27   Snapchat    False  \n",
       "2022-04-28          0             0   28   Snapchat    False  \n",
       "2022-04-29          0             0   29   Snapchat    False  \n",
       "2022-04-19          0             0   19  Pinterest    False  \n",
       "2022-04-20          0             0   20  Pinterest    False  \n",
       "2022-04-21          0             0   21  Pinterest    False  \n",
       "2022-04-22          0             0   22  Pinterest    False  \n",
       "2022-04-25          0             0   25  Pinterest    False  \n",
       "2022-04-26          0             0   26  Pinterest    False  \n",
       "2022-04-27          0             0   27  Pinterest    False  \n",
       "2022-04-28          0             0   28  Pinterest    False  \n",
       "2022-04-29          0             0   29  Pinterest    False  "
      ]
     },
     "execution_count": 109,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Assigning a boolean column for Twitter or not\n",
    "stocks['Twitter'] = np.where(stocks['company'] == 'Twitter', True, False)\n",
    "stocks"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 110,
   "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>company</th>\n",
       "      <th>Facebook</th>\n",
       "      <th>Pinterest</th>\n",
       "      <th>Snapchat</th>\n",
       "      <th>Twitter</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th></th>\n",
       "      <th>Day</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th rowspan=\"4\" valign=\"top\">High</th>\n",
       "      <th>19</th>\n",
       "      <td>218.429993</td>\n",
       "      <td>22.969999</td>\n",
       "      <td>34.240002</td>\n",
       "      <td>48.500000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>20</th>\n",
       "      <td>214.160004</td>\n",
       "      <td>22.389999</td>\n",
       "      <td>33.180000</td>\n",
       "      <td>47.110001</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>21</th>\n",
       "      <td>202.970001</td>\n",
       "      <td>21.379999</td>\n",
       "      <td>31.750000</td>\n",
       "      <td>47.939999</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>22</th>\n",
       "      <td>192.720001</td>\n",
       "      <td>20.469999</td>\n",
       "      <td>31.209999</td>\n",
       "      <td>49.730000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th rowspan=\"4\" valign=\"top\">Low</th>\n",
       "      <th>19</th>\n",
       "      <td>209.000000</td>\n",
       "      <td>21.360001</td>\n",
       "      <td>32.320000</td>\n",
       "      <td>45.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>20</th>\n",
       "      <td>198.919998</td>\n",
       "      <td>21.030001</td>\n",
       "      <td>30.700001</td>\n",
       "      <td>45.029999</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>21</th>\n",
       "      <td>185.580002</td>\n",
       "      <td>19.870001</td>\n",
       "      <td>28.754999</td>\n",
       "      <td>45.500000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>22</th>\n",
       "      <td>183.270004</td>\n",
       "      <td>19.469999</td>\n",
       "      <td>28.559999</td>\n",
       "      <td>47.490002</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th rowspan=\"4\" valign=\"top\">Close</th>\n",
       "      <th>19</th>\n",
       "      <td>217.309998</td>\n",
       "      <td>22.719999</td>\n",
       "      <td>33.630001</td>\n",
       "      <td>46.160000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>20</th>\n",
       "      <td>200.419998</td>\n",
       "      <td>21.040001</td>\n",
       "      <td>30.760000</td>\n",
       "      <td>46.720001</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>21</th>\n",
       "      <td>188.070007</td>\n",
       "      <td>20.230000</td>\n",
       "      <td>29.420000</td>\n",
       "      <td>47.080002</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>22</th>\n",
       "      <td>184.110001</td>\n",
       "      <td>19.590000</td>\n",
       "      <td>29.760000</td>\n",
       "      <td>48.930000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "company      Facebook  Pinterest   Snapchat    Twitter\n",
       "      Day                                             \n",
       "High  19   218.429993  22.969999  34.240002  48.500000\n",
       "      20   214.160004  22.389999  33.180000  47.110001\n",
       "      21   202.970001  21.379999  31.750000  47.939999\n",
       "      22   192.720001  20.469999  31.209999  49.730000\n",
       "Low   19   209.000000  21.360001  32.320000  45.000000\n",
       "      20   198.919998  21.030001  30.700001  45.029999\n",
       "      21   185.580002  19.870001  28.754999  45.500000\n",
       "      22   183.270004  19.469999  28.559999  47.490002\n",
       "Close 19   217.309998  22.719999  33.630001  46.160000\n",
       "      20   200.419998  21.040001  30.760000  46.720001\n",
       "      21   188.070007  20.230000  29.420000  47.080002\n",
       "      22   184.110001  19.590000  29.760000  48.930000"
      ]
     },
     "execution_count": 110,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Create transposed dataframe with column for each compnay\n",
    "features = ['High', 'Low', 'Close']\n",
    "\n",
    "inverted = (stocks.query('Day < 25') # filter pre-intervention period\n",
    "            .pivot(index='company', columns='Day')[features] # make one column per day and one row per company\n",
    "            .T) # flip the table to have one column per company\n",
    "\n",
    "inverted"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 111,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Extract data for synthetic and real\n",
    "y = inverted['Twitter'].values\n",
    "X = inverted.drop(columns='Twitter').values"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 112,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([-1.041,  7.439,  3.152])"
      ]
     },
     "execution_count": 112,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Run Linear regression to calculate weights of each company\n",
    "from sklearn.linear_model import LinearRegression\n",
    "weights_lr = LinearRegression(fit_intercept=False).fit(X, y).coef_\n",
    "weights_lr.round(3)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 113,
   "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>Company</th>\n",
       "      <th>Weight</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>Facebook</td>\n",
       "      <td>-1.041</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>Pinterest</td>\n",
       "      <td>7.439</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>Snapchat</td>\n",
       "      <td>3.152</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "     Company  Weight\n",
       "0   Facebook  -1.041\n",
       "1  Pinterest   7.439\n",
       "2   Snapchat   3.152"
      ]
     },
     "execution_count": 113,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Construct dataframe with weights\n",
    "new_keywords = ['Facebook', 'Pinterest', 'Snapchat']\n",
    "weights = pd.DataFrame()\n",
    "weights['Company'] = new_keywords\n",
    "weights['Weight'] = weights_lr.round(3)\n",
    "weights"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 114,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Creating Synthetic Twitter for stock price by dotting weights with data\n",
    "twitter_synth_lr = (stocks.query(\"company != 'Twitter'\").pivot(index='Day',\n",
    "                                                            columns='company')['Close'].values.dot(weights_lr))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 115,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAtAAAAGHCAYAAACZNGVOAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjMuMiwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy8vihELAAAACXBIWXMAAAsTAAALEwEAmpwYAABvAklEQVR4nO3deVhUZfsH8O8s7PsAiuKelqmouYMLLkCammYuoShaZqXZov3E6lXrTXMp9U3TrDRcsNJyadUQTTQ1F1xTK01NBUE2QbaBmXl+f4yMjCyCwBw48/1cF5czZ85yz9wcvHm4z3MUQggBIiIiIiIqF6XUARARERER1SYsoImIiIiIKoAFNBERERFRBbCAJiIiIiKqABbQREREREQVwAKaiIiIiKgCWEATWbG1a9dCrVZLHYbkLP05KBQKREVFWex4cladn+XevXuhUChw/fr1atl/RfBcJapZWEAT1TIKhaLMryZNmpR7X6NGjUJ8fLzpeVRUFBQKRbH1goKCMH78+CqIvnS5ubmYNWsWWrRoAQcHB3h6eqJz585YtmyZaZ2JEyeid+/e1RpHeRUWV4Vfnp6e6NOnD/bv33/fbW/cuIHhw4dbIMrq9/3336NHjx7QaDRwcnJC8+bNMWbMGGRmZlbpcao792q1GmvXrjVbFhAQgBs3bqB+/foPtE+5nqtEBPDXWaJa5saNG6bHR44cwZAhQ3DkyBE0bNgQAKBSqcq9LwcHBzg4OFR5jGXJz8+Hra1tseUvvfQSfv31V3z00Udo164dMjMzceLECVy9etWi8VXU8ePHUa9ePSQmJuKtt97CgAED8Mcff5RYHBW+dx8fH8sHWg327NmDYcOGYdasWfjss89gZ2eHixcvYvv27dBqtVKHV2mVzZVcz1UiAiCIqNbav3+/ACAuX74shBCie/fu4u233za9Pnv2bAFA7Nq1y7SsV69e4v/+7/+EEEJERkYKlUolhBDi119/FQDMvsLDw0V4eHix5b/++qsQQojExEQRHh4uvLy8hLOzswgICBCxsbGmYxXu88cffxTdu3cXdnZ2Yvny5SW+Fzc3t1JfE0KIOXPmFIsjMjJSCCFEQkKCGDVqlHBzcxP29vYiMDBQHD161Gz7ixcviuHDhwsPDw/h4OAg/Pz8xA8//FDscxBCiNzcXPHUU0+JVq1aiWvXrpUYT+F7K/r69evXBQCxatUqIYQQAMRHH30kQkNDhaurqxg+fLhp+YYNG0zb3b59W7z66quiQYMGwtbWVjRu3FjMmzfP9Pr9Pud7RUdHC6VSKa5evWq2/OuvvxZ2dnYiPT1dCCHEvHnzRNOmTYWtra3w8vISISEhIicnp9T93uvVV18VHTt2LPV1vV4vmjZtavZehBAiKytLuLi4mPIXGBgonnvuOfHf//5X1K1bV3h4eIjw8HCRlZUlhCg79wDEihUrRFhYmHB2dhYNGjQQCxcuNDteQUGBmDNnjmjSpImws7MTrVq1MuVICCEaN25cbP9ClJzjsr6PyiKnc5WIhGABTVSL3fuf8qxZs0S3bt1Mr/fo0UN4e3uLmTNnCiGEyMnJEba2tmLHjh1CCPP/lLVarfj4448FAHHjxg1x48YNcevWLXHr1i3Rs2dPMXLkSNNyrVYrcnJyxKOPPiqGDRsmjh49Ki5cuCDmzp0rbG1txblz54QQd/9TfuSRR8R3330nLl26VGpB2rJlSzFw4ECRmppa4uu3b98Wo0ePFv7+/qY4cnJyhMFgEF26dBHt2rUT+/fvF6dPnxYjR44U7u7uIjk5WQghxI0bN0SdOnVEv379xP79+8XFixfF9u3bxU8//VTsc0hLSxM9evQQPXr0EGlpaaV+9iUVV6mpqQKAqfAAIDQajVi2bJm4ePGi+Ouvv0zLCwtog8EgAgMDRdOmTcW2bdvEP//8I2JjY8Vnn31mytn9Pud76fV64evrK95//32z5QMHDhQjR44UQgixZcsW4eLiIr7//nvx77//ihMnToilS5dWqIBesGCBcHNzE4cPHy51nffff180a9ZMGAwG07LVq1cLNzc3kZ2dLYQwFtBubm7itddeE+fPnxc7duwQbm5uYvbs2UKI0nNf+FnWqVNHfPbZZ+LixYvio48+EgDEnj17TMcLDw8Xfn5+4pdffhGXLl0SX3/9tXBzcxOrV68WQghx8+ZNoVKpxP/+9z/T/oUonuP7fR+VRU7nKhGxgCaq1e79T/nXX38VKpVKZGRkiOzsbGFrays+/PBD0blzZyGEcWTSxsbGNLJ378jrhg0bTKNvRfXr10+Eh4ebLYuMjBS+vr6ioKDAbHmfPn3Eq6++aooHgFi/fv1938tvv/0mGjVqJJRKpfDz8xPPP/+82L59u1nh9dxzz4nAwECz7WJiYgQAcfbsWdOyvLw84ePjI959910hhBD/+c9/RN26dU3v+16Fn8PVq1dFq1atxNChQ0Vubm6Z8d5bXGVmZoqJEycKtVotzpw5I4QwFnfPPvtssW2LFtCF8d87Yl40tvt9ziWJiIgQjz76qOl5UlKSUKvV4scffxRCCLFkyRLRokULkZ+fX+b7LEt2drYYPHiwACB8fHzEkCFDxP/+9z+RkpJiWicxMVHY2NiYjax269ZNTJ482fQ8MDBQ+Pn5me37hRdeMCswS8q9EMbPcurUqWbLHnnkEVMheunSJaFQKMT58+fN1nn33XdFu3btTM9VKpVpVLvQvTm+3/dRWeR0rhKRELyIkEhG/P39YWtri3379mH//v1o3Lgxxo0bhxMnTiAjIwN79uxBly5d4OTkVOljHT16FImJiXB3d4ezs7Ppa//+/bhw4YLZul26dLnv/rp3745//vkH+/fvR3h4OJKSkvD000/jySefhBCi1O3Onj0LT09PtGrVyrTMzs4OXbt2xdmzZwEAcXFxCAgIKPN9GwwG+Pv7o02bNvj2229hb29/35gB4JFHHoGzszPc3Nzwyy+/YP369WjTpo3p9fu997i4OHh4eKBTp04lvl6Rz7mo8PBwnD9/HkePHgUAfPXVV/D09MTjjz8OABg5ciQKCgrQuHFjjB8/Hhs2bMDt27fL9Z4LOTo64vvvv8fly5cxf/581K9fH/Pnz8cjjzyC8+fPAwDq1q2LIUOG4PPPPwdgzNfvv/+O559/3mxf7du3N3vu6+uLpKSkcsVR1rbHjh2DEAKdOnUy+/zef//9Mj+/kpTn+6i8avO5SkS8iJBIVuzs7BAQEIDdu3fD1tYWffv2hbe3N1q2bIm9e/diz549pgKqsgwGAx599FFs27at2GuOjo5mz8tbBKjVagQEBCAgIADTp09HVFQUxo4di3379iEwMLDU7UqajUAIYba8pHWKUiqVGDRoELZu3Ypz587Bz8+vXDH/8ssvqFevHjw8PKDRaIq9Xp73XlZsFfmci3r00UfRqVMnrF+/Hp07d8b69esxevRo01Rovr6++PPPP/Hrr79iz549eO+99xAREYHDhw+bLnIrryZNmmD8+PEYP3485s2bh4cffhiLFi1CZGQkAODFF1/EE088geTkZHz++efo3LlzsaL33ovVFAoFDAZDuY5f1raF/x48eLDY53W/74mSPMg2Jant5yqRteMINJHM9O3bF3v27MGePXvQr18/07Jt27YhLi4Offv2LXXbwkJEr9cXW37vsk6dOuHSpUtwdXVF8+bNzb4edNqvez366KMAgJs3b5YaR+vWrZGSkoJz586Zlmm1Whw5cgStW7cGAHTs2BEHDhxAdnZ2mcdbtWoVnnnmGfTp0wcnTpwoV4xNmjTBQw89VGLxXB4dO3ZEWloajh07VuLrlfmcx40bh6+//hqnTp3C8ePHER4ebva6nZ0d+vfvj0WLFuHMmTPIycnB9u3bH+h9FPLw8ICPj48pZ4Dx+69Ro0b47LPPsGHDhmKjz+VRUu7Lo2PHjgCAq1evFvv8HnrooQrtv7zfR+Ulp3OVyNqwgCaSmb59++LMmTM4efIk+vTpY1q2ceNG2NjYwN/fv9RtmzZtCsA4t29ycjKysrJMy+Pi4vDPP/8gJSUFBQUFGDNmDJo2bYqBAwciOjoaV65cweHDhzF//vwHKsICAwOxatUqHDt2DP/++y92796NyZMnw93d3fQ+mjZtij///BNnz55FSkoKtFot+vbtiy5dumD06NE4cOAA/vjjD4wbNw55eXl46aWXAACTJ0+GwWDAkCFDcODAAVy+fBk//vgjduzYUSyOZcuWITw8HP369Su1qK1Kffv2Rc+ePTFq1Ch89913uHz5Mg4cOIDVq1cDQKU+59DQUGRkZGD8+PFo27Yt2rVrZ3ptzZo1+Pzzz3Hq1Cn8+++/2LhxI27fvm1qhTly5AhatmyJI0eOlLr/d955B2+88QZ+/fVXXL58GWfOnMEbb7yBP/74A0899ZRpPYVCgUmTJuG///0v8vPzERoaWuHPqaTcl0fz5s3x7LPP4vnnn8eGDRtw8eJFnDp1Cl988QUWLlxotv9ff/0VCQkJSElJKXFfFfk+Ko/aeq4SETiNHVFtdu+FSUIIodPphKurq2jbtq1pWXp6ulCpVKJfv35m2997YZIQxqnJ6tSpIxQKhelipH/++Uf07NlTODk5mU2NlZKSIl588UVRv359YWNjI+rXry+GDh0qjh8/LoQoeaaK0syfP980E4GdnZ1o2LChGDNmjNnFgampqWLAgAHC1dW1zGnsevXqVeyivL/++ksMHTpUuLq6CgcHB9G2bdsSZ+Eo9Oabbwo3Nzdx6NChEuMtz3vDPdPVlbY8MzNTvPzyy8LHx0fY2NiIJk2aiPnz55tev9/nXJahQ4cKAOLDDz80W75lyxbh7+8v3N3dhYODg2jdurVpVoqi768w1yXZs2ePGDlypGjcuLGws7MTnp6eIiAgQERFRRVbNzk5WdjY2IhJkyYVe61wGrui3nvvPdG4cWPT89JyX9JnfO+FdDqdTixcuFA88sgjwsbGRnh6eopevXqJzZs3m9bZsWOHaNmypbC1tS1zGruyvo/KIqdzlYiEUAhRxtU5REREVeDcuXNo3bo1jh07ZmqrICKqrVhAExFRtdFqtYiPj8frr7+OjIwM7N27V+qQiIgqjT3QRERUbb766is0b94cly5dwqeffip1OEREVYIj0EREREREFcARaCIiIiKiCmABTURERERUASygiYiIiIgqoFbeyjshIUHqEKyCl5dXqTcUIPlgnq0D8yx/zLF1YJ4tq7S7dXIEmoiIiIioAlhAExERERFVAAtoIiIiIqIKqJU90ERERGSdhBDIy8uDwWCAQqGQOhyLS0pKglarlToMWRFCQKlUwt7evtzfUyygiYiIqNbIy8uDjY0N1GrrLGHUajVUKpXUYciOTqdDXl4eHBwcyrU+WziIiIio1jAYDFZbPFP1UavVMBgM5V6fBTQRERHVGtbYtkGWUZHvLf4KR0RERFROaWlpGDVqFAAgOTkZKpUKGo0GAPDTTz/B1ta22Dbr16+Hg4MDRowYgU2bNiEwMBA+Pj4AgM8//xxhYWHlbh2gmoEFNBEREVE5aTQa7Nq1CwCwePFiODk54cUXXyxzm3Hjxpkef/PNN2jZsqWpgF69ejWefvrpChXQer2efdASs1gBnZ2djVWrVuHatWtQKBR46aWXcPLkSezevRuurq4AgNDQUHTo0MFSIRERERFVisFgQP/+/bFz506cPXsWISEhOHLkCHx9fREQEIDdu3dj5cqVcHJyQoMGDXDq1Cm8/PLLsLe3x6hRo5CUlIQRI0bAw8MD3377LWJjY/Hhhx8iPz8fjRs3xtKlS+Hk5ISuXbvimWeewb59+zB+/HgMGTJE6rdu1SxWQEdGRqJ9+/aYPn06dDodtFotTp48iYEDB+LJJ5+0VBhEREQkE7Nnu+LcOZsq3WerVgX4738zy72+UqmEVqvF7du3ceTIEbRr1w6HDx9Gly5d4OnpaTayPGjQIKxduxazZs1Cu3btABhbOL755htoNBqkpaXho48+wqZNm+Do6IgVK1bgs88+w+uvvw4AsLOzww8//ACdTlel75kqziIFdE5ODs6fP48pU6YYD6pW8wpaIiIikoVOnTrh6NGj+P333zF16lT8+uuvEEKga9euFdpPXFwc/v77b9PockFBATp27Gh6nQOONYdFqtibN2/C1dUVK1euxL///otmzZph/PjxAIBffvkF+/btQ7NmzTBu3Dg4OzsX2z4mJgYxMTEAgAULFsDLy8sSYVs9tVrNz9oKMM/WwdryfPz4cQCwqrZAa8lxUlKSaRDu/fdzquko5SuPlEollEol/P39cfToUcTHx2PgwIFYuXIlVCoVgoODoVarTeup1WooFAqoVCrTeyj6XKlUIjAwEJ9++mmxYykUCri4uBij4yBktbCzsyv3OWSRDOj1ely+fBnPPvssWrRogcjISGzfvh39+/fH8OHDAQCbNm3C+vXrMXny5GLbBwUFISgoyPQ8JSXFEmFbPS8vL37WVoB5tg7WludGjRoBsK7/L6wlx1qttsZcQGcwGGAwGNC5c2fMnz8f3bp1g8FggLu7O2JiYhAREQGdTmdaT6fTwdHRERkZGaY2DCcnJ2RkZMDNzQ3t27fHzJkzceHCBTRt2hS5ublISEjAQw89BCEE9Ho9ALCFo5potdpi51D9+vVLXNci80B7enrC09MTLVq0AAB069YNly9fhru7u+m3sn79+uGff/6xRDhEREREVaZhw4YAYGrZ6Ny5M9zc3ODu7l5s3ZEjR2LmzJkIDg5Gbm4uxowZg7CwMAwfPhyenp5YunQppkyZgqCgIAwePJi1UQ2lEEIISxxo9uzZePHFF1G/fn1s3rwZWq0WgwYNgoeHBwDgxx9/xMWLF/Haa6/dd18JCQnVHC0B1jOaYe2YZ+tgbXlevHgxAGD69OkSR2I51pLjnJwcODo6Sh2GZNRqNUegq0lJ31uljUBbrInm2WefxbJly6DT6VCnTh1MnjwZkZGRuHLlChQKBby9vTFp0iRLhUNERDK2ZMkSANZVQBOR5VisgG7SpAkWLFhgtmzq1KmWOjwREVmRadOmSR0CEckYL+MkIiLZ4cgzEVUni1xESEREREQkFyygiYhIdk6fPo3Tp09LHQYRyRRbOIiISHYGDBgAAIiPj5c4EiKSI45AExGR7Pj5+cHPz0/qMEimPvroI/Tp0wdBQUEIDg423fmyonbu3Im///7b9Hz48OE4depUube/du0atm3bZnp+6tQpzJo1q9zbFxQU4P3330f37t3Rt29fDBw4EHv27Cn39kXd+17Kq6LvuabgCDQREcnOzp07pQ6BZOrYsWOIiYnBzp07YWdnh7S0NOTn5z/Qvnbu3ImgoCA8/PDDD7R9YQH91FNPAQDatWuHdu3alXv7Dz74AElJSdizZw/s7OyQnJyMQ4cOPVAsZb0XnU4nu9uPcwSaiIiIqJxu3rwJjUYDOzs7AIBGo4GPjw/279+P5557zrTevn37MHHiRABAixYtsGDBAgQFBWHQoEFITk7G0aNHsWvXLsydOxfBwcG4cuUKAOON5QYOHIgePXrg8OHDAAC9Xo/33nsPTzzxBHr37o0NGzYAAN5//30cOXIEwcHB+Oyzz3Dw4EGMGzcOAJCdnY3XX38d/fr1Q1BQEH766Sez95Gbm4uNGzdi7ty5pvfi7e2NJ598EgCwfft29OvXD3379sW8efNM25X3vQwfPhzz58/H008/jdWrV2P//v0ICQlBv379MG3aNGi12qpOjUXJ69cBIiIishqus2fD5ty5Kt1nQatWyPzvf0t9PTAwEEuXLkWPHj3Qs2dPPPnkk/D390ePHj3w9ttvIzU1FZ6enti0aRNGjhwJwHiHuw4dOmDmzJmYO3cuNm7ciNdeew3BwcGmQrSQTqfDTz/9hN27d2PJkiXYtGkTvvrqK7i4uODnn3+GXq/HoEGDEBgYiLfeegurVq3C+vXrAQAHDx407ed///sfXFxcsHv3bgDArVu3zN7H5cuX4evrCxcXl2LvMTExEfPmzcPOnTvh5uaG0NBQ7Ny5E/3796/Qe8nMzMSWLVuQl5eHHj16YNOmTXjooYfwyiuvYP369Xj++ecrnqAagiPQREQkOx06dECHDh2kDoNkyMnJCTt37sSiRYvg6emJl156CZs2bYJCocDTTz+NLVu2ICMjA3Fxcejbty8AwNbWFsHBwQCM/fnXr18vdf9PPPEEAKBt27am9WJjY/Htt98iODgYAwYMQHp6Oi5fvlxmnPv378f48eNNz93d3cv9Hk+dOgV/f394enpCrVZj2LBh+P333yv8XgpHs//55x80atQIDz30EABgxIgRptH12ooj0EREJDtJSUlSh0AWUNZIcXVSqVQICAhAQEAAWrZsiW+++QajRo3CqFGjMH78eNjZ2WHQoEGmvl+1Wg2FQmHaVqfTlbpvW1vbEtebO3cuevfuDbVabVpedMT5XkII0zFL0rRpU8THxyMrKwvOzs7Fti1NRd6Lo6PjffdXW3EEmoiIZCcuLg5xcXFSh0EydPHiRVy6dMn0/OzZs2jQoAEAwMfHB3Xr1sWyZctM7RtlcXZ2RnZ29n3XCwwMxPr161FQUADAOKKbk5NT5vaBgYGIjIw0Pb+3hcPBwQGhoaGYNWuW6SLIpKQkbNmyBY899hh+//13pKWlQa/XY/v27fD393/g99K8eXNcu3bNNGq+ZcsWdOvW7b7vuyZjAU1ERLLj4+MDHx8fqcMgGcrJycFrr72G3r17IygoCBcuXDC7dfywYcNQr169cs2sMWTIEHzyyScICQkxXURYktGjR6NFixbo378/evXqhYiICOh0Ojz66KNQqVQICgrCZ599ZrbNq6++ioyMDPTt2xdBQUEljlbPmDEDnp6e6NOnD/r27YvnnnsOnp6eqFu3Lt58802MGDECwcHB8PPzw+OPP/7A78Xe3h5LlizBCy+8gH79+kGpVGLs2LH3/XxqMoWohePqCQkJUodgFby8vJCSkiJ1GFTNmGfrwDzLn7XkOCcnx9QaUBO9/fbbaNOmDUJDQ6tl/0VbOKhqlfS9Vb9+/RLX5Qg0ERHJzowZMzBjxgypwyAr079/f5w/fx7Dhg2TOhSqZryIkIiIZGfjxo0AgEWLFkkcCVkT3sDHerCAJiIi2Vm4cKHUIRCRjLGAJiIi2QkLC5M6BCKSMfZAExERERFVAAtoIiKSnejoaERHR0sdBhHJFAtoIiKSnQkTJmDChAlSh0Ey1bBhQwQHB5u+Pv74YwDA8OHDcerUqUrvf9OmTfD19cX+/ftNy3bs2AFfX1/88MMPFd7fwYMHMW7cuPuu07JlSwQHByMoKAijRo2qtmkRn3vuOQQHB6N79+6mYwYHB+Po0aOlbnPq1CnMmjXLFGvRdXfu3Im///67WmItDXugiYhIdoKCgqQOgWTM3t4eu3btqtZjPProo/juu+/Qs2dPAMB3332HVq1aVesxu3TpgvXr1wMA5s+fj7Vr1+KNN96o8uOsWbMGgLEQXrVqlemYZWnXrh3atWsHADh06BCcnJzQuXNnAMYCOigoqFw3rymk0+lMt1p/EByBJiIi2Vm3bh3WrVsndRhkxbZv345+/fqhb9++mDdvnml5ixYtsGDBAgQFBWHQoEFITk4ucfsuXbrg5MmTKCgoQHZ2Nq5cuYLWrVubXu/atSvS0tIAGEdnhw8fDsBYXBaO6IaEhCArK8tsvydPnkRISAj+/fffUmMXQiArKwtubm4AgMWLF2PVqlWm1/v27Ytr165h0aJFWL16tWn5ggULsGbNGiQlJWHYsGEIDg5G3759cfjw4ft+Xv369UNGRgaEEGjdujW++eYbAMDUqVOxb98+0yj6tWvXsGHDBnz++ecIDg7GoUOHsGvXLsydOxfBwcG4cuUKrly5gjFjxqB///546qmncPHiRQDAa6+9hnfeeQfDhw83y8mDYAFNREREtZavry98fX3NloWHh8PX19esDz4qKgq+vr5mN9hJTEyEr68vOnToUKFj5uXlmbVwfPfdd2avJyYmYt68edi8eTOio6Nx8uRJ0xzROTk56NChA2JiYtCtWzfTnOX3UigU6NGjB/bu3YtffvkFISEh5Ypt1apVeP/997Fr1y5s27YN9vb2pteOHj2KmTNnIjIyEo0bNy627ZEjRxAcHIzOnTtj//79eOaZZ8o8VmhoqKnQNRgM+P777/HUU09h27ZtCAwMxK5du7Br1y6zwr80nTp1wtGjR/HXX3+hcePGOHLkCADg+PHj6Nixo2m9hg0bYuzYsXj++eexa9cu+Pv7Izg4GP/5z3+wa9cuNGnSBDNmzMB7772HnTt3YtasWXjzzTdN21+6dAmbNm3CnDlz7htTWdjCQURERFQB92vhOHXqFPz9/eHp6QkAGDZsGH7//Xf0798ftra2CA4OBgD4+fmZ9Tnfa8iQIfjiiy+QmZmJ2bNnY/ny5feNrXPnznj33Xfx1FNPYcCAAaZbUV+8eBERERH48ssv4ePjU+K2RVs4VqxYgblz55Y5p3rDhg3h4eGBP/74A8nJyWjdujU0Gg3at2+P6dOnQ6fT4fHHH0ebNm3uG3fXrl1x+PBhXL9+HePGjUNUVBRu3LgBDw8PODk53Xf7QtnZ2YiLi8MLL7xgWpafn296PGjQIKhUqnLvrzQsoImISHYKRyTj4+MljoSqW0k5Lql9JywsrNj84D4+PtXyPSKEKPU1tVoNhUIBAFCpVNDpdKWu+9hjj+HPP/+Evb09HnrooWL7MRgMAACtVmta/vLLL6Nfv37Ys2cPBg8ejE2bNgEA6tSpA61Wiz/++KPUArqokJAQPP/886Y4C4917/FCQ0OxefNm3Lx50zRi3a1bN2zZsgW7d+/Gq6++ihdffBEjRowo83hdu3bF2rVr0aBBA0RERGDHjh346aef0KVLl/vGWpTBYICrq2upv+A4OjpWaH+lYQsHERERURV67LHH8PvvvyMtLQ16vR7bt2+Hv7//A+3rzTffxMyZM4stb9CgAU6fPg0A+Omnn0zLr1y5gkcffRRTpkxBu3btTP2/rq6uWL9+PRYsWICDBw/e97hHjhwxtXk0bNgQZ86cAQCcOXMGV69eNa03YMAA/Prrrzh16hR69+4NALh+/Tq8vLwwZswYPPPMM6Zty+Lr64u0tDRcvnwZjRs3RpcuXbBq1Sp07dq12LpOTk5mvd3Ozs7Izs4GALi4uKBhw4am2UqEEDh79ux9j19RLKCJiEh24uPjOfpM1ebeHuj333/f7PW6devizTffxIgRIxAcHAw/Pz88/vjjD3Ssvn37onv37sWWT5s2DbNnz8ZTTz1l1pKwevVq9O3bF0FBQbC3t0efPn1Mr3l7e2PdunV4++23cfz48WL7LOyBDgoKwpYtWzB79mwAwBNPPIFbt24hODgY69evR7NmzUzb2NraIiAgAIMHDzbFcfDgQYSEhCAkJAQ///wzJk6cWK73+thjj5n23aVLFyQmJppm2igqODgYO3fuRHBwMA4fPowhQ4bgk08+QUhICK5cuYKPP/4YX3/9NYKCgtCnT59qmRNeIcr6O0MNlZCQIHUIVsHLy6va5oCkmoN5tg7Ms/xZS45zcnKq7M/wtZFarS6z7cPSDAYDHn/8cXz66admhXVtVNL3VmEP+b04Ak1EREREFfb333+je/fu6NGjR60vniuKFxESEZHshIeHAyj5YjIiqhoPP/wwDh06JHUYkmABTUREshMTEyN1CEQkYyygiYhIdiIjI6UOgapJLbx0i2qJinxvsYAmIiLZKe9d26j2USqV0Ol0UKtZwlDV0el0UCrLf2mgxb77srOzsWrVKly7dg0KhQIvvfQS6tevj6VLlyI5ORne3t54/fXX4ezsbKmQiIiIqJaxt7dHXl4etFqt6YYk1sTOzs7sRiZUeUIIKJVKs9ue34/FCujIyEizWztqtVps27YNfn5+GDp0KLZv347t27cXu0sQERFRRUVFRQEA/0+RIYVCAQcHB6nDkIy1TFdY01lkGrucnBycP38effv2BWCcw9DJyQlHjx5FYGAgACAwMBBHjx61RDhERCRzERERiIiIkDoMIpIpi4xA37x5E66urli5ciX+/fdfNGvWDOPHj0dGRgY8PDwAAB4eHsjMzLREOEREJHNjxoyROgQikjGLFNB6vR6XL1/Gs88+ixYtWiAyMhLbt28v9/YxMTGmKYkWLFgALy+vaoqUilKr1fysrQDzbB2sLc9ffPGF1CFYnLXl2FoxzzWDRQpoT09PeHp6okWLFgCAbt26Yfv27XBzc0N6ejo8PDyQnp4OV1fXErcPCgpCUFCQ6Tl7fyyDfVbWgXm2Dsyz/DHH1oF5tixJb+Xt7u4OT09PJCQkAADOnDmDBg0aoFOnToiNjQUAxMbGonPnzpYIh4iIZC4xMRGJiYlSh0FEMmWxWTieffZZLFu2DDqdDnXq1MHkyZMhhMDSpUuxZ88eeHl5Ydq0aZYKh4iIZKxjx44AgPj4eIkjISI5slgB3aRJEyxYsKDY8tmzZ1sqBCIishJ169aVOgQikjHexoeIiGTn+PHjUodARDJmkR5oIiIiIiK5YAFNRERERFQBLKCJiEh2+vfvj/79+0sdBhHJFHugiYhIds6cOSN1CEQkYyygiYhIdnbs2CF1CEQkYyygiYhIdtq2bSt1CEQkY+yBJiIiIiKqABbQREQkO4sXL8bixYulDoOIZIoFNBERyc6SJUuwZMkSqcMgIpliDzQREcnOtGnTpA6BiGSMBTQREcnO9OnTpQ6BiGSMLRxERERERBXAApqIiGTn9OnTOH36tNRhEJFMsYWDiIhkZ8CAAQCA+Ph4iSMhIjliAU1ERLLj5+cndQhEJGMsoImISHZ27twpdQhEJGPsgSYiIiIiqgAW0EREREREFcACmoiIZKdDhw7o0KGD1GEQkUyxB5qIiGQnKSlJ6hCISMZYQBMRkezExcVJHQIRyRgLaCIikh0fHx+pQyAiGWMPNBERERFRBbCAJiIi2ZkxYwZmzJghdRhEJFMsoImISHY2btyIjRs3Sh0GEckUe6CJiEh2Fi5cKHUIRCRjLKCJiEh2wsLCpA6BiGSMLRxERERERBXAApqIiGQnOjoa0dHRUodBRDLFFg4iIpKdCRMmAADi4+MljoSI5IgFNBERyU5QUJDUIRCRjLGAJiIi2Vm3bp3UIRCRjLEHmoiIiIioAiw2Aj1lyhTY29tDqVRCpVJhwYIF2Lx5M3bv3g1XV1cAQGhoKDp06GCpkIiIiIiIKsyiLRxz5swxFcuFBg4ciCeffNKSYRARkcz5+voC4EWERFQ92MJBRERERFQBCiGEsMSBpkyZAmdnZwBAcHAwgoKCsHnzZsTGxsLBwQHNmjXDuHHjTOsUFRMTg5iYGADAggULkJ+fb4mQrZ5arYZOp5M6DKpmzLN1YJ7ljzm2DsyzZdna2pa43GIFdFpaGjQaDTIyMjB37lxMmDAB9evXN7V0bNq0Cenp6Zg8efJ995WQkFDd4RIALy8vpKSkSB0GVTPm2Towz/LHHFsH5tmy6tevX+Jyi7VwaDQaAICbmxs6d+6Mixcvwt3dHUqlEkqlEv369cM///xjqXCIiIiIiB6IRQrovLw85Obmmh6fPn0ajRo1Qnp6ummdI0eOoGHDhpYIh4iIZC48PBzh4eFSh0FEMmWRWTgyMjLw4YcfAgD0ej169OiB9u3bY/ny5bhy5QoUCgW8vb0xadIkS4RDREQyV3jdDBFRdbBIAV23bl188MEHxZZPnTrVEocnIiIrExkZKXUIRCRjvJU3ERHJTkhIiNQhEJGMcR5oIiIiIqIKYAFNRESyExUVhaioKKnDICKZYgsHERHJTkREBAAgLCxM4kiISI5YQBMRkeyMGTNG6hCISMZYQBMRkewsWrRI6hCISMbYA01EREREVAEsoImISHYSExORmJgodRhEJFNs4SAiItnp2LEjACA+Pl7iSIhIjlhAExGR7NStW1fqEIhIxlhAExGR7Bw/flzqEIhIxtgDTURERERUASygiYiIiIgqgAU0ERHJTv/+/dG/f3+pwyAimWIPNBERyc6ZM2ekDoGIZIwFNBERyc6OHTukDoGIZIwFNBERyU7btm2lDoGIZIw90EREREREFcACmoiIZGfx4sVYvHix1GEQkUyxgCYiItlZsmQJlixZInUYRCRT7IEmIiLZmTZtmtQhEJGMlbuA1ul0uHDhAtLT0xEQEIC8vDwAgL29fbUFR0REdD9CAMnJSly7psK1a2pcvapCWtp/0apVAeLjtfD11UsdIhHJTLkK6KtXr2LhwoWwsbFBamoqAgICcO7cOcTGxuL111+v7hiJiMjKZWQocO2aClevGgvkwkLZWDSrkJdn3pHo5GRAdrYTAODhhwvQu7cWffpo0aWLFhz3IaLKKlcB/fnnn2PUqFHo1asXJkyYAABo1aoVPv3002oNjoiIrENubmGBrDIVyoX/Xr+uQkaGeYHs6mpAw4Z6NG+uQ+/eWjRqpEfDhro7/+px4cIpXL3qhISEtti71w5r1zrhs8+cYW9vQEBAPvr00aJ37zw0baqHQiHRmyaiWqtcBfT169fRs2dPs2X29vbIz8+vlqCIiEheCgqA+HiV2ejx9et3C+XkZJXZ+vb2Ag0aGAviTp3y0aiRDg0b6k2Fsru7KPN4TzwxAAAQHx+PF17IRk6OAocO2WLvXjv8+qs99uyxB+CGxo2NBXjv3nno3j0fTk5l75eICChnAe3t7Y1Lly7hoYceMi27ePEifHx8qi0wIiKqPQwGIDFRiWvX1Ka2irujyCrcuKGCwXB3qFelEvD1NY4WBwXlmRXHjRrp4e1tqNTIsJ+fn9lzR0eBfv206NdPCyAT//6rwt69dti71w7ffOOAdeucYGMj0KVLPvr0yUPv3lq0bKnj6DQRlUghhLjvr9txcXFYtWoVgoOD8cMPP2DYsGHYtWsXXnjhBbRr184ScZpJSEiw+DGtkZeXF1JSUqQOg6oZ82wdKptnIYD0dCWuXlUV60G+elWN+HgV8vPNq00fH2NBXFgcFx1F9vHRQ11D5oHSaoFjx+6OTp8/bwPAGH/v3sZiumdP7X1HvaXGc9k6MM+WVb9+/RKXl6uABoBLly5hz549SE5OhqenJ4KCgtCsWbMqDbK8WEBbBk9S68A8W4fy5DkrS1FicVw4opydbd6H7OGhN/UcFxbHxi8dGjTQ19qL9W7cUGLfPmMxvW+fHTIylFAqBR57rMA0Ot22bQFUqvvvy5J4LlsH5tmyKl1A1yQsoC2DJ6l1YJ6tg5eXF65fT8H160UL5KK9yCqkp5tXhE5OBlNbRWFxXLTNwtm51v33UWE6HXDypA327rXH3r12OHnSBkIo4OGhR2CgFr17axEYqEWdOgapQ+W5bCWYZ8uqVAH94YcfYuDAgXj00UdNy86fP4+ff/4Z06dPr7ooy4kFtGXwJLUOzLN86HTAjRuqItO73R1Jjo+3RUKCeYuFra2xD/neC/SM7RZ6eHhUrg9ZSh06dAAAHD9+vEr3m5ZWODpt7J9OSTH+0tGmTb5pqryOHfNhY1Olhy0XnsvWgXm2rNIK6HJ1oJ07d67YXZ0efvhhfPDBB5WPjIiIykUI4OZNZbHiuLDNIiFBBZ3ubsWrVArUq2cshoOCDKhTJ9usUK5b1wClsowD1mJJSUnVsl+NxoChQ3MxdGguDAbg3Dk1fv3VODq9apUzPv7YBc7OBvTsqb0zu4cWDRrwRi5EclOuAtrGxgZ5eXlwdHQ0LcvLy4OqpjWAERHVcrduKUosjo2tFmrk5ZkPCXt7G1srOnTIx5Ah5qPI9evrTSOhxlGrLAnekTTi4uKq/RhKJdCmjQ5t2mRh6tQsZGYqcODA3dHpHTscAAAtWty9kUvXrryRC5EclKuAbteuHT777DNMmjQJjo6OyMnJwZo1a9C+fftqDo+ISF5ycoreMMS8UL5+XYXMTPMhYTc3Axo21OHhh3Xo1097z8V6ejg4yL8P+UFIMc2qq6vAgAF5GDAgD0IAFy+qTcX0+vVO+PzzuzdyKZx7ulkz3siFqDYqVw90VlYWli9fjlOnTsHZ2RlZWVlo3749pk6dCicnp3IdaMqUKbC3t4dSqYRKpcKCBQuQlZWFpUuXIjk5Gd7e3nj99dfh7Ox8332xB9oy2GdlHZjnqpWfb7xhSGkzWRT2zBaytzeYzWTRoIH5lG9ublVTIDPP0srNvXsjl7177fDPP8Y/DTRsqDONTnfvrq3UhZnMsXVgni2rSmbhSE9PR2pqKry8vODu7l6hAKZMmYL58+fD1dXVtCwqKgrOzs4YOnQotm/fjqysLISFhd13XyygLYMnqXVgnitGr797w5B776Z39aoKiYnmNwxRqwUaNNDfKYyLX6zn5WWZC/WsLc8zZswAACxatEjiSEp29erdG7n89psdsrOVsLER6NTp7m3GW7Wq2I1crC3H1op5tqwKX0QohIDizplrMBin53Fzc4Obm5vZMmUlrkA5evQo3nnnHQBAYGAg3nnnnXIV0ERE1UUIIDVVWWqbRXy8CgUFd6sahUKgbl0DGjXSwd8/v9hMFj4++ho3X7A12LhxI4CaW0A3aqTHuHE5GDcuB/n55jdyef99V7z/vivq1i2cKi8PvXpp4eHBdh2imqLUAnr8+PFYt24dACA0NLTUHWzatKncB5s3bx4AIDg4GEFBQcjIyICHhwcAwMPDA5mZmeXeFxHRg7p9u+wbhuTkmA8MaDTGYtjPrwADB+aajSL7+uphZyfRG6FSLVy4UOoQys3WFggIyEdAQD7eeus2EhOViI21w9699oiOtsfmzY5QKgXat797I5d27WrejVyIrEmpLRwpKSnw8vICACQnJ5e6A29v73IdKC0tDRqNBhkZGZg7dy4mTJiARYsWYe3ataZ1JkyYgMjIyGLbxsTEICYmBgCwYMEC5Ofnl+uYVDlqtRo6nU7qMKiayTHPeXnAlSvAlSuKe76Af/9VIC3N/O/iLi4CTZoINGmCO/+aPy/HpRk1nhzzbA30euDYMQWio5WIjlbg6FEFhFBAoxEICjIgJMT4b716zLG1YJ4ty9bWtsTlpY5AFxbPBoMBK1aswNtvvw2bSswMr9FoABjbQDp37oyLFy/Czc0N6enp8PDwQHp6ull/dFFBQUEICgoyPWfvj2Wwz8o61MY863RAQoKqhLvpGUeRk5LMh+bs7AQaNNDdGUW+e7vpwlFkDw9Raq9pXp7xq7arjXkmo4ceAl56yfiVlqbA/v3GVo9ff7XD5s3G/8Zbty7A448LNGqUjWbNdGjaVAeNhi0fcsRz2bIe+EYqSqUSN2/eRGXu+J2XlwchBBwcHJCXl4fTp09j+PDh6NSpE2JjYzF06FDExsaic+fOD3wMIpIPg6HwhiHqIr3I5jcM0evvVrwqlUD9+sYL9Xr31pr1IDdsqEOdOvK9YQiVLDo6GgAQEhIicSRVS6MRGDIkD0OG5Jlu5FJ4m/Fly9TQ6TxM67q7G9C0qa7Il970uKpmdyGyVuWahWPPnj04f/48Ro4cCU9PT7PXynMRYVJSEj788EMAgF6vR48ePTBs2DDcvn0bS5cuNbWLTJs2jdPY1SD8Ldc6SJFnIcq+YUh8fPEbhtSpc3eqt3tnsqhXTy/JrZNrE2s7n319fQEA8fHxEkdiOa6uXjh+PB2XL6tNX5cuqXH5svGXTiHunlMajd6soG7aVIdmzYzPKzOVHlU/azuXpVapaexGjRpV6msVuYiwqrCAtgyepNahuvKck2O8UK+0G4bcvm3+y7e7u/GGIfcWx40a6eHrq4ODQ5WHaFWs7XwODw8HANPF8NagrBzn5QH//ltYWKuKFNdqJCaatzx5e+tNbSDmRTZv3FMTWNu5LLVKFdBVcRFhVWIBbRk8Sa3Dg+Y5Px+4fr3k4vjqVRVSU83/U3ZwML9hSGGhXNib7OrK/5irE89n+XvQHOfkKHDliqrIyPXdxzdvmp/HPj76O6PV5gV248Y63qLcQnguW9YD9UALIbB7925cvXoVzZo1Q+/evasjNiKqgYw3DFGZRpGvXzcvlBMTlWZ/EraxEfD1NRbH/fsXmEaRCwtlT0/L3DCEiCrG0VGgVSsdWrUqPrNDVpaxuP7nH7VZa8iOHfZIS7tbXCsUxusQireFGH9BLmUiAyoiNxfIyFAiI0OJzEwlbt1SFHmuwK1bxscGgwoFBR5QqQRUKtz5ElAqjY/V6ruPC5er1ebrlLZt+fYJKJXizj6Nj4tuU3Qd47bm65S2bW27TqXMAnrDhg347bff0LJlS3z55Ze4efMmRo4caanYiKia3b6twKVLCpw5Y1+kB9n4b0k3DKlXz1gMd++uNWuzaNhQBx8fA+elJZIZZ2eBNm10aNOmeHGdkaEocdT6hx8ccOvW3WpIqRRo2FBf4sWMDRvqob7vdAa1gxDGXzgyMozFb2am8k4BbCx+iz4vLIwLH2dmKqHVlj3C4OxsgKurAS4uShQU2MBgMA50GL8UpscGg/GxTnf3cdGLrmsqhaL04rtxYz1+/LFmjbqX+W176NAhvPPOO6hfvz6uX7+ORYsWsYAmqsVycxU4etQWBw7Y4sABO5w6ZXPnttPGaSa9vIwtFm3bFmDQoOI3DOEoEtUW1ngRoaW5uRlv7tK+fUGx19LSFPdcyGgsso8dc0RW1t3iWq02L66NrSHG576+lr+Lp14PswK36EiwsQC+OxKcmXl3PWOBrLjz87RkCoWAm5uAm5vhzpeAj48e7u53n7u6Gh+7uxvXc3U1PnZ1NZh+0XjQFg6DwVhUFy2yy1d8Gx8/+LaATqcwPb67jfk6er3izrZ3Xyvc1sPD8KAprTZlFtA5OTmm3o8GDRogKyvLIkHVNMqUFDhu3Ijs55+HcHSUOhyictNqgRMnbHHwoLFgjouzRUGBAmq1wGOP5eOVV7LQs6c9PDzS0bChHo6O7EMmosrTaAQ0mgJ07GheXAsBpKQoTQX13eJajUOHbJGbe7e4trUVaNTIWFDf7bk2ftWrV/rUlFotio3wFn9c8mtFi/uS2NiYF8AajcE0LWDR5Xcf333u7CwkbVNQKlFkEKToz3r+3H8Q9+2BLjoHtMFgKDYndN26das3whrAPjoarosWwWndOtx+4w3kjBoF/q2aaiKdDjhzxgYHDtjhwAFbHDlii7w8JRQKAT+/AkycmI3u3bXo0iUfTk7G89jLyw4pKbyrFckLR55rJoUC8PY2wNs7H126mL8mBJCUpCzWFnLpkhr799uZTW1pby/QpIkOPj56ZGebF8b3ToF5LwcHY1Hr7m4c4fX11aNVq7KL38JRYXv70m+6RNalzFk4ypq+rpC1TGNnc/Qo3N57D7ZxcSh45BFkvv02tH37Qs5nEq/0rfkMBuDPP9V3CmY7/P67rWl6uJYtC9C9uxbdu+eja1ct3N1LPtWZZ+vAPMufnHNsMAA3bqhw6ZLK7GLGpCQlnJ2FqQ3C1dW84C1siSj6vLa3osk5zzVRpaaxq2kkm8ZOCNj//DNc338f6itXoO3eHZmzZqHAz0+aeKoZT9KaRwjgn39UOHDADgcP2uHgQVvTlfBNmujuFMxaBATkw9u7fD1jzLN1kGWehQAKCqAoKADy86EoKIAiP984x6IQ0DdtWvsu7a8EWeaYimGeLeuBb+VNRSgUyBs4EHnBwXCKioLzkiXw7t8fOcOG4XZEBPQNGkgdIcnQ9esqHDhgi99+MxbNhTc9qFdPj759taZRZl9fvcSRkqwIYSpKTcVpQQGg1d59nJ8PxT3rlLSssKhVFBQYC94765ge37vdPduYrVv4uHB5Ga44OcFt7lzkDh2KWj/sSEQ1CkegK0GRmQnnFSvgvHo1IASyn30Wt6dOhXBzkzq0KsHfcqVx86YSBw/amWbK+Pdf4++5np56dO+ebxplbtJEXyUdRMyzBAwG86K0sHC8X3FaUADFnXXuV4jeu8wOQEF2donbFdu+MIYqJhQKwM4OwsYGwsYGsLWFsLW9+7jIv8LWFrjzb9HHxZbZ2EDY2d19fGf5m6+8gskA2gLQ1a+P7EmTkDN6NISTU5W/r5qC57J1YJ4tiy0c1UgZHw/XRYvgsGULhJsbbr/2GrLDw2v9iAdPUstIT1fg0CE700wZf/9tAwBwdTXA319rKpofeURXLS33ssuzwVC8EL13FLSkUdIiBWypxWkJBWyxUdI765TWWqAoKIBCV/UXbQqVyrwQvVNYFj5WOTqiQKk0L0RLKj5LKmTLKmpLKYRL2pclJ/yNjo4GhMBgtRrOK1fC7vffYXB3R/azzyJ7wgQYNBqLxWIpsjuXqUTMs2WxgLYA9R9/wG3uXNjt3w9d48bInDkTeYMH19oLDXmSVo+sLAUOH7Y1zZRx9qwNhFDAwcGAbt3yTT3MbdoUWGSylwrlWa8vXojeWxyWVIiWVLSWY3S1xD/bFylOzUZRC4+vr/pWFqFSmReE9xSnZsXjPevcOzJa1uhpqcXnvcWpjU2xkdz7fbNY+/lsc+wYnFeuhMMvv8Dg4ICc0aORPWmSrFrvrD3H1oJ5tiwW0JYiBOz27oXrvHmwOX8e+Y89hszZs5F/73w9tQBP0qqRmwvExdmaZso4edIGer0CtrYCHTsaC+YePfLRrl1+5f9oIQQUeXlQZGRAefv23X8zM6G881X4WHH7NpSZmbDV6aAr8qf9EvtOCwtdQ9VPZi/UavNRziJFY5nFaXn+tH+/EdXS2giKFqc2NrKYtpLns5H677+NhfS2bQCA3CFDkDV5MnQtW0ocWeUxx9aBebasShXQL730UonLbWxs4OnpiS5duiAkJAQqC/0nU6ML6EJ6PRy+/RauixZBlZiI3P79kfnmm9A3by51ZOXGk/TBFBQAJ0/a4rffbHHwoPHmJVqtAiqVQLt2BaYe5k6d8uHgcM/GOp2xwL1T3CpKKnoLi+I7jxX3rnuf9gChVsPg4gLh6gqDqyvUbm4oUCjMR0lLGhkt55/271ec3juSa02zJEjJ2s7nqKgoAEBYWFiJr6vi4+H02Wdw3LgRytxc5AUHI2vKFOR37mzJMKuUteXYWjHPllWpAvr777/H/v37MWDAAHh6eiIlJQW//PILunXrBmdnZ/z444/o0qVLqT+oqlqtKKDvUOTmwumzz+C8YgUUeXnICQvD7WnTYPDykjq0++JJWj56PXD2rA0O/GaDuH06/H00F3Z5mXDHLbRrnIIOzdPg1yAFzTzTYa81H/0tViTn5Nz3eAYnJwgXFxjc3Iz/3imERdF/73m96GvCwcGsrYh5tg7Wlufy3spbkZYGp3Xr4LRmDVTp6dB26YKsKVOg7dev1rXfWVuOrRXzbFmVKqCnTZuG//znP9AUuegiLS0Nc+fOxZIlS5CQkID33nsPn3zySdVFXIbaVEAXUqakwGXJEjhGRUE4OCBr8mRkT5pkLGZqKKs7SYWAMj299PaHO8sUt28jJ/42shOyoEvNhCrrNlwNt+CGDKhRdv+tsLExG/0t/Nfg6lqsKDZ77U5RLFxcqvxCLKvLs5WytjzPmDEDALBo0aJyra/IyYHjV1/B6dNPoY6PR0HLlsiaPBm5Tz5p/EtJLWBtObZWzLNlVWoe6PT0dNjb25sts7OzQ3p6OgCgXr16yM7OrmSI8mbw8kLG++8j69ln4Tp/vvHW4OvXI3PGDOQOHy6LHsvaSnXpEhy3bIHD1q1QX71a5ro5ahekGdyRbnDDLbgj394Xdg1dkNXIGQUtnOBY786IsIsLhJub2b8GV1fA3r7WjWoR1UblLZwLCUdHZD/3HLLHjYPDd9/BeeVKeLzyClwWLkT2iy8iJzS0Rg94EJFllauA7tixIz744AM89dRT8PT0RGpqKrZv346OHTsCAP7++294e3tXa6ByoW/eHOlr1iD78GG4vvcePKZNg/PnnyPzP/+BtndvqcOzGsq0NNh//z0cv/0WtidOQCiV0PboYZzeysMDKTp3nLriicN/eWLfKW/8fdMTmXCFtyfuXPRnnCnjoYZ3R5wNALKke0tEVBVsbJA7fDhyhw2D3e7dcF6xAm6zZsF56VLjFHjjx0N4eEgdJRFJrFwtHPn5+fjmm2/w+++/Iy0tDR4eHvD398fw4cNhZ2eHW7duQafTwctCfb21sYWjRELA/ocf4LpgAdT//ou8Xr2Q+fbb0LVpI3VkAOT1ZyIhgNx0LVQ7dsP1+2/hcWg3lHod0hq2xp+dRuJkqxFIUtXHpUtqHDhgh8uXjb9benjoERBw9+YlDz1UNTcvqUnklGcqnbXlOTExEQDg4+NT6X3ZHjkC5xUrYB8TA4OjI3LGjEHW88/DcKfPuqawthxbK+bZsjiNXU2m1cJp/Xq4/O9/UGRkIPfpp5E5Y4bkP5ylPEmFAHJzFcjKMn5lZyuRnV34WIGsLKXpcXa2spTlCuRkAX6ZBzE8Nwoj8A3ckYEE1MNGjMEGjMUZtDU7rrPz3bmYu3fX4tFHdbKfJII/jK2DteW5vBcRVoT6/HnjFHjffQcoFMgdNsw4BV6LFlV2jMqwthxbK+bZsipdQCckJODKlSvIy8szW963b9/KR1dBsiug71BkZMBl+XI4ffEFoFAga+JEZE2ZAuHqKkk8FTlJCwveBylys7KKFsdK03KDoXxDvfb2Bjg5CTg7izv/GtBS+TeeSN2IPglfo072v9DaOOFcy8H4s9NIpLbrCUcXJZycDHB2Ltzu7mO5F8z34g9j62Btee7QoQMA4Pjx41W+b9W1a8Yp8L78Esq8POQ+/jiypkxBwZ22RqlYW46tFfNsWZUqoLdu3YotW7agcePGsLOzM3ttzpw5VRNhBci1gC6kun4dLgsXwnHrVug1GmS9/jqyw8Kq7dbgQgC3byuQnq5EWtrdr4ICFyQl5dxT/BoL3pycu48rXvDeLVgdHY0Fb9Hi18lJFCmIzYvce5c7OQnTpBTKtDTYf/cdHLdsudvX3LMncp9+GnkDBkA4OlbL51fb8YexdWCeq54yNRVOkZFwioyE8tYtaP39kTV5MrR9+khysTBzbB2YZ8uqVAE9ceJEzJo1C40bN67ywB6E3AvoQjZnzsD1vfdgd+AAdE2aIPOtt5D3xBNl/mAuHAkuWgiX9lW0YNbpSt+nnZ15wXpvkVv4vLQit3C7wteqdEaovDzY79oFxy1bYPfrr1DodCho1Qo5Tz+N3KFDYaiC/ke54w9j68A8Vx9FdjYcN26E82efQXXjBgpatULWlCnIHTSoyqedLAtzbB2YZ8uqVAE9efJkLFu2DGoL/iAoi7UU0ACQlyug/3Ev6i19D87//oUbzbogOui/OOPiX6QYVpkVxXl5JRfDCoWAh4cBGk3xr8LlRV9/6CEPaLUpNW8KVIMBtkeOwGHLFjj8+COUmZnQ+/gg96mnkPP009A9+qjUEdYq/GFsHZhnC8jPh8O2bXBeuRI2Fy9C16gRsl54ATmjRqH4bUerHnNsHZhny6pUAR0bG4u//voLI0aMgJubm9lrSgkaRmtrAV1QgGJtEqWNCBd+5eQYP18VdBiPtfgvZqM+buBbPI35LvOQ7vVQqUXx3eJYD43GADc3UaHppmvaSaq6ePHufM3Xr8Pg6Ii8J55AztNPI797d86l/YBqWp6pelhbnvv37w8A2Llzp+UPbjDAftcuOC9fDtsTJ6D39DTOMR0eDuHuXm2HtbYcWyvm2bIqVUCPGjWq1Nc2bdr04FE9oJpQQOv1wK1b9y+Aiy7PzCz9lw1n55JHhO/98nK4jYd/+AR11q2AIj8f2ePGIev112EocpfIqlITTlJlaiocvvsODlu2wPbkSWNfc69exr7m/v3Z11wFakKeqfpZW56rYxaOChMCtr//DueVK2G/Zw8MTk7IGTsWWRMnwlCvXpUfztpybK2YZ8uqVAGdnJxc6mtS3EDF0gX0gQO2WL3ayaxVIiNDASFKbpVwcCi7TaKkx/dcm3lfyps34bJ4MRy/+grC0RFZL7+MrOeeq9I/E0p2kubm3u1r3rvX2NfcuvXdvua6dS0fk4zxh7F1sLY8nz59GgDQtm3b+6xpGeqzZ41T4H3/PaBSIWf4cGS9+CL0zZtX2TGsLcfWinm2LM4DXQm7dtlh0SLX+7RK6O+8LuDgYLmPVH3hAlznzYP9rl3Q16t399bgVdBaY9GT1GCA7eHDd/uab9+G3scHOcOGIffpp6Fr2dIycVgh/jC2DsxzzaD69184f/opHDdtArRa5A0YYJwCr337Su+bObYOzLNlVbiA/vTTT/HCCy8AAJYvXw5FKTM/vPzyy1UUYvnVhBaOmsb24EG4zp0L21OnUNCqFTJnzYK2V69K7dMSJ6n64kU4fPstHLZtM/Y1OzkZ+5qHDWNfs4Xwh7F1YJ5rFmVKCpzWrIHTunVQZmRA2707sqZMMf7cfsAp8Jhj68A8W1ZpBXSp02rUqVPH9LgqboVK1Ss/IAApP/4Ih++/h8uCBfAMDUVe797GW4O3aiV1eGaUKSl3+5pPnTL1Nd+OiGBfMxFVicWLFwMApk+fLnEkJTN4eeF2RASyJk82ToH3+efwHD0a+W3aIGvyZOQNGsQBBKIajC0ccqTVwikyEi7LlkGRmYnckSOR+X//V+GLVqr0t9zcXNhHR9/ta9br2ddcQ3A0wzpYW55rxEWEFaHVwnHrVjivXAn1pUvQNWmCrBdfRM6IEYC9fbl2YW05tlbMs2VVqgf6jz/+QJ06dVCnTh3cunULUVFRUCqVGD16NNyrcUqe0rCALh9Fejpcli2D09q1EEolsidNQtbkyRAuLuXavtInqcEA299/N/Y1//QT+5prKP4wtg7WlueaPgJdKr0e9r/8AucVK2B78iT03t7InjgR2ePGQbi6lrmpteXYWjHPllWpAvr111/H22+/DS8vL3z00UcAAFtbW2RmZiIiIqJqIy0HFtAVo7p6FS4LFsDxu++g9/TE7WnTkDNmDO53h5QHPUnVFy7c7WuOj7/b1/z008gPCOCfJWsY/jC2DsxzLSMEbA8cME6BFxsLg4sLsseORfbEiaX+xY45tg7Ms2VVuAe6qLS0NHh5eUGv1+PUqVNYuXIl1Gq16SLD8jIYDJg5cyY0Gg1mzpyJzZs3Y/fu3XC981t1aGgoOnToUKF90v3pGzXCrZUrkT1pElzfew/ub78N5zVrkPn228h7/PEHvmClKGVKChy2bzf2NZ8+bexrDgzE7TffNPY1W+AuXEREsqFQIL9HD6T16AGbM2fgvHIlnFetgvPq1cgZMcI4BV6zZlJHSWS1ylVAOzg44NatW7h27RoaNGgAe3t76HQ66HS6Ch3s559/hq+vL3Jzc03LBg4ciCeffLJiUdMDKWjfHqnffgu7XbvgOm8eNM89B22XLsicNQsFD/KLS2Ff87ffwi42Fgq9Hvlt2iBjzhxjX3ORC1GJiCypps0DXRkFfn5I/+QTqGbMgPOqVXD85hs4fvkl8gYONE6BJ4P3SFTblKuA7t+/P958803odDqMHz8eAPDnn3+aLtIoj9TUVBw/fhzDhg3Djz/++EDBUhVQKKANCUFy375w/OoruCxeDO/Bg5E7eDAyZ86EvkmTsrc3GGB76BActm6929dcrx6yXnzR2Nf8yCMWeRtERGUZMGAAgFp0EWE56Js2RcbChbg9fbppCjyHH3+EtmdP3J4yBRg6VOoQiaxGuWfhSEhIgFKpNE1pl5CQAJ1Oh0aNGpXrQIsXL8ZTTz2F3Nxc/PDDD6YWjtjYWDg4OKBZs2YYN24cnJ2dyxULVQ1FVhacV62C06pVUOh0yB43Drdfew1CozHrs1L//bfxYsCtW6FOSDD2NQ8caOxr9vdnX3Mtxn4662Btee7fvz8AYOfOnRJHUn0UmZlwioqC0+efQ3XzJgwdOiAjLAx5Tz7JtjkZs7ZzWWqVvhOhXq/HX3/9hbS0NGg0GjzyyCNQlbNoiouLw4kTJzBx4kScPXvWVEDfunXL1P+8adMmpKenY/LkycW2j4mJQUxMDABgwYIFyM/PL9dxqQISEqB67z0o164FXFygnzEDytGjYdiyBcqNG6E8cQJCpYIICoJhzBgYBg8GOF+zLKjV6gq3Y1HtwzzLWF4elFFRUC1bBsVff0G4uRl/Tk+cCNG6tdTRURXjuWxZtra2JS4vVwEdHx+PhQsXIj8/H56enkhNTYWNjQ0iIiLQoEGD+x78yy+/xL59+6BSqZCfn4/c3Fx06dIFr7zyimmdmzdvYuHChaaph8rCEejqo/7zT+OtwffsMS3L9/NDbuF8zd7eEkZH1YGjGdaBeZY/L09PZP74IxyjouDw889Q5Ocjv1MnZIeFIXfQIICj0rLAc9myKjUC/e677+Kxxx7D4MGDTbf0/v7773HixAnMmTOnQoEUHYFOT0+Hh4cHAODHH3/ExYsX8dprr913Hyygq5/tb7/B/e+/kdajB3QPPyx1OFSN+MPYOjDP8lc0x8rUVDh88w2coqKgvnwZBnd35Dz9NHLGjoWuRQuJI6XK4LlsWZWaxu7KlSuYNWuWqXgGjLNnbNu2rVJBRUVF4cqVK1AoFPD29sakSZMqtT+qOvk9esAwdCh0PEmJqBYqnBL1+PHjEkciDYOnJ7JffBHZL7wA24MHjb3S69fDec0aaLt2RU5YGHKfeKLcdzkkInPlKqA1Gg3OnTuHNm3amJadP3/eNHpcEa1bt0brOz1ZU6dOrfD2RERE95OUlCR1CDWDQoH87t2R3707lCkpcNy8GY4bN8Jj6lS4zZqFnBEjkBMWBl3z5lJHSlSrlKuADg0NxcKFC9GxY0fTnw6OHz/OApiIiGqkuLg4qUOocQxeXsiaPBlZL74I299+g9PGjXCKjITz559D6+9vHJUeMACws5M6VKIar0LT2B06dMjUt+zv719qX0h1Yw+0ZbDPyjowz9aBeZa/B8mxMjkZjps2wXHjRqivXoXewwO5I0cie8wY6B96qJoipcrguWxZlZ7GriZhAW0ZPEmtA/NsHZhn+atUjg0G2O3fD8eoKNj/8gsUej20AQHIDgtD3oABQClTeZHl8Vy2rApfRLh8+XKziwZL8/LLLz94VERERNVgxowZAIBFixZJHEktoVRCGxgIbWAglElJcPz6azh++SU0kydD7+mJnFGjkDN6NPRNm0odKVGNUOoI9DfffFOuHYwYMaJKAyoPjkBbBn/LtQ7Ms3Wwtjz7+voCkNetvO+nynOs18Nu3z7jqPSuXcZR6Z49kT1mDPIef5yj0hKxtnNZahUegZaiMCYiIqoKCxculDqE2k+lgrZPH2j79IEyMRGOX30Fx6++gubFF6H38kLOM88YR6UbN5Y6UiKLU5b14l9//YWoqKgSX9u4cSP+/vvvagmKiIioMsLCwhAWFiZ1GLJh8PFB1uuv4+ahQ0hdvx75HTrAeeVK1A0IgGb0aNj/9BNQUCB1mEQWU2YBvXXrVrRq1arE11q1aoWtW7dWS1BERERUA6lU0Pbrh/TISCQdPozM6dNh8/ff0EyahLpdusBlwQKorl2TOkqialdmAX3lyhW0b9++xNfatm2Ly5cvV0dMRERElRIdHY3o6Gipw5A1Q/36yJo2DUm//47UyEgUtG0L5xUrUMffH5qwMNjv3AnodFKHSVQtyryRSm5uLnQ6HWxLuFBAr9cjNze32gIjIiJ6UBMmTABgXRcRSkathjYkBNqQEKji4+/2Sj/3HPQ+PsZe6dBQ6Bs0kDpSoipT5gi0r68vTp06VeJrp06dMl3lTEREVJMEBQUhKChI6jCsjt7XF7ffeANJhw8j7YsvUNCqFZw/+gh1unWDZuxY2EVHc1SaZKHMEeiBAwfis88+g8FgQOfOnaFUKmEwGHD06FGsWbMG48aNs1ScRERE5bZu3TqpQ7BuajXyHn8ceY8/DtX163D88ks4fv01PCdMMI5Kjx6N7GeegYEDcVRL3fdOhD/++CM2b96MgoICuLq6IjMzE7a2thgxYgQGDRpkqTjNcB5oy+Bck9aBebYOzLP81fgcFxTAPiYGjlFRsIuNBRQKaPv2RXZYGLR9+wIqldQR1go1Ps8yU6lbeefk5ODvv/9GVlYWnJ2d8fDDD8PR0bHKgywvFtCWwZPUOjDP1oF5lr/alGPV1aumUWlVcjJ09esjZ/Ro5DzzDAz16kkdXo1Wm/IsB5UqoGsaFtCWwZPUOjDP1sHa8sw7EdYSBQWwj4423u1w3z4IlQp5QUHICQuDNjCQo9IlqJV5rsUqfCdCIiIiomplY4O8gQORN3AgVFeuGEelN22Cwy+/QNegAXJCQ42j0j4+UkdKZKbMWTiIiIhqo/j4eKsafZYDfZMmuP3WW0g6ehRpq1ZB36QJXD/4AHW7dIHHxImw27sXMBikDpMIAEegiYiIqCaxtUXe4MHIGzwYqkuX4PTll3DYtAkOO3ZA17Dh3V7pOnWkjpSsGEegiYiIqEbSN2uGzP/8B0nHjiFt5UroGzaE68KFqNu5Mzyefx62Bw5IHSJZKRbQREQkO+Hh4QgPD5c6DKoqdnbIGzIEqd98g6R9+5D93HOwPXQIXiNHwnHDBqmjIyvEApqIiGQnJiYGMTExUodB1UD/0EPInD0bSceOIa9fP7i9+Sbsf/hB6rDIyrAHmoiIZCcyMlLqEKi62dsj/dNPoQkNhcfUqUhzc4O2Vy+poyIrwRFoIiKSnZCQEISEhEgdBlUz4eCAtLVroWveHB7PPQebEyekDomsBAtoIiIiqrWEuztSN26EwcsLmrFjob5wQeqQyAqwgCYiItmJiopCVFSU1GGQhRjq1kXql18CajU8Q0Oh4hzgVM1YQBMRkexEREQgIiJC6jDIgvRNmyI1KgqKrCxoQkOhTE2VOiSSMRbQREQkO2PGjMGYMWOkDoMsTNemDdLWroU6Ph6asWOhyMqSOiSSKc7CQUREsrNo0SKpQyCJ5HfrhrRPPoFm4kRonnsOqevXA3Z2UodFMsMRaCIiIpIVbUgIbi1eDLvffoPHyy8Der3UIZHMsIAmIiLZSUxMRGJiotRhkIRyR4xAxuzZcPj5Z7i9+SYghNQhkYywhYOIiGSnY8eOAIB4zsZg1bJfeAHKtDS4fPwxDBoNbs+cKXVIJBMsoImISHbq1q0rdQhUQ9yeORPK9HS4LF8Og0aD7EmTpA6JZIAFNBERyc7x48elDoFqCoUCGfPnQ5meDrd334XBwwO5I0ZIHRXVcuyBJiIiInlTqZD+8cfQ9ugB9+nTYRcdLXVEVMtZtIA2GAyYMWMGFixYAADIysrCe++9h1deeQXvvfcesjhfIxEREVUHOzukrVmDgjZtoHnpJdgePix1RFSLWbSA/vnnn+Hr62t6vn37dvj5+WHZsmXw8/PD9u3bLRkOERHJVP/+/dG/f3+pw6AaRjg7I23DBuh8faEJD4f67FmpQ6JaymIFdGpqKo4fP45+/fqZlh09ehSBgYEAgMDAQBw9etRS4RARkYydOXMGZ86ckToMqoEMnp5I++orCGdneI4ZA9WVK1KHRLWQxS4iXLt2LcLCwpCbm2talpGRAQ8PDwCAh4cHMjMzS9w2JiYGMTExAIAFCxbAy8ur+gMmqNVqftZWgHm2DtaW50OHDgGAVb1na8txpXh5Qb9jB2z69UOdMWNQsHcvUK+e1FGVC/NcM1ikgI6Li4ObmxuaNWuGsw/w55KgoCAEBQWZnqekpFRleFQKLy8vftZWgHm2DtaW50aNGgGwrv8vrC3HlebtDZt16+A5ciQUAwYg5dtvIdzdpY7qvphny6pfv36Jyy1SQP/11184duwYTpw4gfz8fOTm5mLZsmVwc3NDeno6PDw8kJ6eDldXV0uEQ0RERISCxx5D+po10IwbB8348cbWDgcHqcOiWsAiPdCjR4/GqlWrsGLFCrz22mto06YNXnnlFXTq1AmxsbEAgNjYWHTu3NkS4RARkcwtXrwYixcvljoMqgW0vXohffly2B47Bo8XXgAKCqQOiWoBSeeBHjp0KE6fPo1XXnkFp0+fxtChQ6UMh4iIZGLJkiVYsmSJ1GFQLZE3eDAy3n8f9rt3w33aNMBgkDokquEsfifC1q1bo3Xr1gAAFxcXzJ4929IhEBGRzE2bNk3qEKiWyRk3Dsq0NLh+8AEMHh7IfPddQKGQOiyqoXgrbyIikp3p06dLHQLVQlmvvgplWhqc16yBwdMTWa++KnVIVEOxgCYiIiICAIUCme+8A2V6OlwXLYJBo0HO2LFSR0U1EAtoIiKSndOnTwMA2rZtK3EkVOsolbi1ZAmUGRlwe/NNGNzdkTd4sNRRUQ3DApqIiGRnwIABAID4+HiJI6FaycYG6Z9+Ck1oKDymTkWamxu0vXpJHRXVIJLOwkFERFQd/Pz84OfnJ3UYVIsJBwekrVsHXfPm8HjuOdicOCF1SFSDsIAmIiLZ2blzJ3bu3Cl1GFTLCTc3pG7cCIOXFzRjx0J94YLUIVENwQKaiIiIqBSGunWR+tVXgFoNz9BQqNgWRGABTURERFQmfZMmSN24EYrsbGhCQ6FMTZU6JJIYC2giIpKdDh06oEOHDlKHQTKia90aaWvXQh0fD83YsVBkZUkdEkmIBTQREclOUlISkpKSpA6DZCa/a1ekffIJbP74A5pnnwW0WqlDIomwgCYiItmJi4tDXFyc1GGQDGlDQnBr8WLYHTgAj5dfBvR6qUMiCbCAJiIi2fHx8YGPj4/UYZBM5Y4YgYw5c+Dw889we/NNQAipQyIL441UiIiIiCooe9IkKFNT4fLxxzBoNLg9c6bUIZEFsYAmIiLZmTFjBgBg0aJFEkdCcnZ75kwo09Phsnw5DBoNsidNkjokshAW0EREJDsbN24EwAKaqplCgYz586FMT4fbu+/C4OGB3BEjpI6KLIAFNBERyc7ChQulDoGshUqF9I8/hnLcOLhPnw6Dmxu0ISFSR0XVjBcREhGR7ISFhSEsLEzqMMha2Nkhbc0aFLRpA81LL8H28GGpI6JqxgKaiIiIqJKEszPSoqKg8/WFJjwc6rNnpQ6JqhELaCIikp3o6GhER0dLHQZZGYNGg7SvvoJwdobnmDFQXbkidUhUTVhAExGR7EyYMAETJkyQOgyyQnpfX6R+9RWg08EzNBRK3hFTllhAExGR7AQFBSEoKEjqMMhK6Vq0QNqGDVCmpMBzzBgobt2SOiSqYiygiYhIdtatW4d169ZJHQZZsYLHHkPamjVQX7wIzfjxUOTmSh0SVSEW0ERERETVIL9XL6QvXw7bY8fg8cILQEGB1CFRFWEBTURERFRN8gYPRsb8+bDfvRvu06YBBoPUIVEV4I1UiIhIdnx9fQEA8fHxEkdCBOSMHQtlWhpcFy2CwcMDme++CygUUodFlcACmoiIiKiaZb3yCpSpqXBeswYGT09kvfqq1CFRJbCAJiIi2eHIM9U4CgUy33kHyvR000h0zrhxUkdFD4gFNBEREZElKJW4tWQJlBkZcHvrLRg8PJA3eLDUUdED4EWERERERJZiY4P0Tz9FfufO8Jg6FXb79kkdET0AFtBERCQ74eHhCA8PlzoMohIJBwekrV0LXfPm8HjuOdicOCF1SFRBLKCJiEh2YmJiEBMTI3UYRKUSbm5I3bgRBm9veIaFQX3hgtQhUQWwgCYiItmJjIxEZGSk1GEQlclQty5Sv/wSwsYGnqGhUPHi11qDBTQREclOSEgIQkJCpA6D6L70TZogdeNGKLKzoQkNhTI1VeqQqBwsMgtHfn4+5syZA51OB71ej27dumHkyJHYvHkzdu/eDVdXVwBAaGgoOnToYImQiIiIiGoEXevWSFu7Fp6jR0MzdixSN2+GcHaWOiwqg0UKaBsbG8yZMwf29vbQ6XSYPXs22rdvDwAYOHAgnnzySUuEQUREViIqKgoAEBYWJnEkROWT37Ur0latgua556B59lmkbtgA2NlJHRaVwiItHAqFAvb29gAAvV4PvV4PBW9hSURE1SQiIgIRERFSh0FUIdrgYNxavBh2Bw7A4+WXAb1e6pCoFBa7kYrBYEBERAQSExPx+OOPo0WLFjhx4gR++eUX7Nu3D82aNcO4cePgzD9ZEBFRJY0ZM0bqEIgeSO6IEVCmp8Pt3XdhePNNZCxcCHDQscZRCCGEJQ+YnZ2NDz/8EBMmTICrq6up/3nTpk1IT0/H5MmTi21TdDqiBQsWID8/35IhWy21Wg2dTid1GFTNmGfrwDzLH3MsL6rZs6FauBD6iAjo//tf03Lm2bJsbW1LXG7xAhoAvvnmG9jZ2Zn1Pt+8eRMLFy7E4sWL77t9QkJCdYZHd3h5eSElJUXqMKiaMc/WgXmWP+ZYZoSAW0QEnDZuRMacOcieNAkA82xp9evXL3G5RXqgMzMzkZ2dDcA4I8eZM2fg6+uL9PR00zpHjhxBw4YNLREOERHJXGJiIhITE6UOg+jBKRTImD8fuQMHwu3dd+HwzTdSR0RFWKQHOj09HStWrIDBYIAQAv7+/ujYsSOWL1+OK1euQKFQwNvbG5Pu/HZFRERUGR07dgQAxPPGFFSbqVRIX74cyowMuE+fDoObGzB6tNRRESRq4agstnBYBv9MZB2YZ+tgbXkuvKfA8ePHJY7Ecqwtx9ZEkZUFz1GjYPPnn9Bt24bkNm0AJe+FZwmltXBYbBYOIiIiS7GmwpnkTzg7I23DBngOHQqbAQNQT62GwcsL+jp1YKhTx/ivt7f58zp1oPf2BhwcpA5fllhAExEREdVwBo0GqVu3wmvPHuReugTVzZtQJidDmZgIm9OnoUxJgcJgKL6dqyv03t5mhbahbl3zZXXqwODhwVHtCmABTURERFQLGLy8YJg8GbdLatXR66FMTYXy5k2okpON/968efff5GTYnjplLLrvTOxQlChpVPvOKDZHtYtjAU1ERLLTv39/AMDOnTsljoTIQlQq40hynTq43yzRiuzsu4V2UpJ5wZ2cDNWNGw8+ql1kdFvOo9osoImISHbOnDkjdQhENZZwcoK+aVPomzYte8Wio9qFxXXRUe2bN8s3ql23rnmP9r1tJLVwVJsFNBERyc6OHTukDoGo9qvsqHbhv+Ud1b7nYsjC4lrv64v8gIDqeY8PiAU0ERHJTtu2baUOgciqVGpUOynJbHTb9tQpKG/ehDInBwCga9IENw8csMC7KD8W0ERERERkGQ8wqq24U0jXJCygiYhIdhYvXgwAmD59usSRENGDKhzVronkeWkkERFZtSVLlmDJkiVSh0FEMsURaCIikp1p06ZJHQIRyRgLaCIikh22bhBRdWILBxERERFRBbCAJiIi2Tl9+jROnz4tdRhEJFNs4SAiItkZMGAAACA+Pl7iSIhIjlhAExGR7Pj5+UkdAhHJGAtoIiKSnZ07d0odAhHJGHugiYiIiIgqgAU0EREREVEFsIAmIiLZ6dChAzp06CB1GEQkU+yBJiIi2UlKSpI6BCKSMRbQREQkO3FxcVKHQEQyxgKaiIhkx8fHR+oQiEjG2ANNRERERFQBLKCJiEh2ZsyYgRkzZkgdBhHJFAtoIiKSnY0bN2Ljxo1Sh0FEMsUeaCIikp2FCxdKHQIRyRgLaCIikp2wsDCpQyAiGWMLBxERERFRBbCAJiIi2YmOjkZ0dLTUYRCRTLGFg4iIZGfChAkAgPj4eIkjISI5YgFNRESyExQUJHUIRCRjLKCJiEh21q1bJ3UIRCRj7IEmIiIiIqoAi4xA5+fnY86cOdDpdNDr9ejWrRtGjhyJrKwsLF26FMnJyfD29sbrr78OZ2dnS4RERERERPRALFJA29jYYM6cObC3t4dOp8Ps2bPRvn17HDlyBH5+fhg6dCi2b9+O7du3c+5OIiKqNF9fXwC8iJCIqodFWjgUCgXs7e0BAHq9Hnq9HgqFAkePHkVgYCAAIDAwEEePHrVEOERERERED8xiFxEaDAZEREQgMTERjz/+OFq0aIGMjAx4eHgAADw8PJCZmVnitjExMYiJiQEALFiwAF5eXpYK26qp1Wp+1laAebYO1pZnrVYrdQgWZ205tlbMc81gsQJaqVTigw8+QHZ2Nj788ENcvXq13NsGBQWZTUmUkpJSHSHSPby8vPhZWwHm2Towz/LHHFsH5tmy6tevX+Jyi8/C4eTkhFatWuHkyZNwc3NDeno6ACA9PR2urq6WDoeIiIiIqEIsUkBnZmYiOzsbgHFGjjNnzsDX1xedOnVCbGwsACA2NhadO3e2RDhERERERA/MIi0c6enpWLFiBQwGA4QQ8Pf3R8eOHfHwww9j6dKl2LNnD7y8vDBt2jRLhENERERE9MAUQgghdRAVlZCQIHUIVoF9VtaBebYOzLP8McfWgXm2rBrTA01EREREVJuxgCYiIiIiqgAW0EREREREFcACmoiIiIioAlhAExERERFVAAtoIiIiIqIKYAFNRERERFQBLKCJiIiIiCqABTQRERERUQXUyjsREhERERFJhSPQVKqZM2dKHQJZAPNsHZhn+WOOrQPzXDOwgCYiIiIiqgAW0EREREREFcACmkoVFBQkdQhkAcyzdWCe5Y85tg7Mc83AiwiJiIiIiCqAI9BERERERBWgljoAqhlSUlKwYsUK3Lp1CwqFAkFBQXjiiSeQlZWFpUuXIjk5Gd7e3nj99dfh7Owsdbj0AErL8YYNGxAXFwe1Wo26deti8uTJcHJykjpcekCl5bnQ999/j6ioKKxevRqurq4SRkqVUVaed+zYgZ07d0KlUqFDhw4ICwuTOFp6UKXl+cqVK/j888+Rn58PlUqFiRMnonnz5lKHa1VYQBMAQKVSYezYsWjWrBlyc3Mxc+ZMtG3bFnv37oWfnx+GDh2K7du3Y/v27fxhXEuVluO2bdti9OjRUKlUiIqKwrZt25jjWqy0PDdo0AApKSk4c+YMvLy8pA6TKqm0PN+6dQvHjh3Dhx9+CBsbG2RkZEgdKlVCaXmOiorC8OHD8dhjj+H48eOIiorCO++8I3W4VoUtHAQA8PDwQLNmzQAADg4O8PX1RVpaGo4ePYrAwEAAQGBgII4ePSplmFQJpeW4Xbt2UKlUAICHH34YaWlpUoZJlVRangFg3bp1GDNmDBQKhZQhUhUoLc/R0dEYMmQIbGxsAABubm5ShkmVVFqeFQoFcnNzAQA5OTnw8PCQMkyrxBFoKubmzZu4fPkymjdvjoyMDNOJ6eHhgczMTImjo6pQNMdF7dmzBwEBARJFRVWtaJ6PHTsGjUaDJk2aSB0WVbGied6wYQP+/PNPfP3117CxscHYsWP5p32ZKJrn8PBwzJs3Dxs2bIDBYMDcuXOlDs/qcASazOTl5WHx4sUYP348HB0dpQ6HqkFpOd66dStUKhV69uwpYXRUVYrmWaVSYevWrRg1apTUYVEVu/d8NhgMyMrKwrx58zB27FgsXboUnGyr9rs3z9HR0QgPD8cnn3yC8PBwrFq1SuoQrQ4LaDLR6XRYvHgxevbsia5duwIw/vkvPT0dAJCens6Ljmq5knIMAHv37kVcXBxeeeUV/nlfBu7Nc1JSEm7evIn/+7//w5QpU5CamoqIiAjcunVL6lCpEko6nzUaDbp27QqFQoHmzZtDqVTi9u3bEkdKlVFSnmNjY02P/f39cfHiRSlDtEosoAkAIITAqlWr4Ovri0GDBpmWd+rUCbGxsQCMJ2znzp2lCpEqqbQcnzx5Et999x0iIiJgZ2cnYYRUFUrKc6NGjbB69WqsWLECK1asgKenJxYuXAh3d3dpg6UHVtr53LlzZ/zxxx8AgISEBOh0Ori4uEgVJlVSaXnWaDQ4d+4cAOCPP/6Aj4+PVCFaLd5IhQAAf/75J2bPno1GjRqZRiBDQ0PRokULLF26FCkpKfDy8sK0adM4jV0tVVqOIyMjodPpTHlt0aIFJk2aJGWoVAml5blDhw6mdaZMmYL58+fzL0q1WGl5btu2LVauXIl///0XarUaY8eORZs2bSSOlh5UaXl2dHREZGQkDAYDbGxsMHHiRNPFhmQZLKCJiIiIiCqALRxERERERBXAApqIiIiIqAJYQBMRERERVQALaCIiIiKiCmABTURERERUASygiYhkJiEhATNmzMC4cePw888/Sx2OyZQpU3D69GmpwyAiqjS11AEQEcnBlClTcOvWLahUKiiVSjRo0AC9evVCUFAQlErLjlV89913aNWqFRYtWlTi6++88w7OnTuHRYsWoUmTJqblixYtwrFjxzBnzhy0bt26UjEU3rDlmWeeqdR+iIhqIhbQRERVJCIiAm3btkVOTg7OnTuHyMhIXLx4EZMnT7ZoHCkpKQgICChznXr16mHfvn2mAvr27du4cOECb65CRFQOLKCJiKqYo6MjOnXqBHd3d7z99tsYNGgQGjVqhOPHj+Prr79GUlISHB0d0adPH4wcORIAMH/+fLRv3x4DBgww7eeNN97AyJEj0aVLl2LHOHbsGL788kukpaWhSZMmmDhxIho0aIB3330X586dw59//om1a9di4cKFqF+/frHte/bsiZiYGISFhUGpVOLAgQPo0qUL4uLiTOsUFBRg48aNOHToEADA398fY8aMgY2NDc6ePYvly5dj4MCB+O6776BUKhEaGoo+ffogJiYGv/32GwDgp59+QuvWrTFz5kwAwJUrV7B+/XokJyejffv2mDJlCmxtbavuwycisgD2QBMRVZPmzZtDo9Hgzz//BADY2dnh5ZdfRmRkJGbOnIldu3bhyJEjAIDAwEDs37/ftO2VK1eQlpZmdgvuQgkJCfjoo48wfvx4rF69Go899hgWLlwInU6HOXPm4NFHH8Wzzz6LDRs2lFg8A4CHhwcaNGiAU6dOAQBiY2MRGBhots7WrVtx4cIFLFq0CB988AEuXryILVu2mF6/desWcnJysGrVKrz44otYs2YNsrKyEBQUhB49emDIkCHYsGGDqXgGgEOHDuGtt97CihUrcPXqVezdu/fBPlwiIgmxgCYiqkYajQZZWVkAgNatW6NRo0ZQKpVo3LgxunfvjnPnzgEAOnfujBs3buDGjRsAgH379iEgIABqdfE/FB48eBCPPfYY2rZtC7VajcGDByM/Px9//fVXhWLr1asXYmNjkZCQgJycHDz88MNmr//22294+umn4ebmBldXVwwfPtysyFepVBg+fDjUajU6dOgAe3t7JCQklHnMAQMGQKPRwNnZGR07dsSVK1cqFDMRUU3AFg4iomqUlpYGZ2dnAMCFCxfw5Zdf4urVq9DpdNDpdOjWrRsAwMbGBv7+/ti/fz+GDx+OAwcOYPr06SXuMz09Hd7e3qbnSqUSXl5eSEtLq1BsXbt2xYYNG7Bjxw707NmzxNiLHsfb29vsGC4uLlCpVKbndnZ2yMvLK/OY7u7upse2trYVjpmIqCbgCDQRUTW5ePEi0tLS0LJlSwDAsmXL0LFjR3zyySdYt24dgoODIYQwrd+7d2/s378ff/zxB+zs7IqNCBfy8PBAcnKy6bkQAikpKdBoNBWKz87ODu3bt0d0dDR69epV7HWNRmN2nIocQ6FQVCgWIqLahAU0EVEVy8nJQVxcHD766CP07NkTjRo1AgDk5ubC2dkZtra2uHjxoulCu0IPP/wwlEol1q9fX2JBWyggIAAnTpzAmTNnoNPp8MMPP8DGxgaPPPJIhWMNDQ3Fu+++izp16hR7rXv37ti6dSsyMzORmZmJb7/9tsSR6pK4ubkhKSmpwvEQEdUGbOEgIqoiCxcuhEqlgkKhQIMGDTBw4ECEhISYXp84cSLWr1+PL774Aq1atYK/vz+ys7PN9tGrVy9s2rQJ//d//1fqcerXr4+pU6fiiy++MM3CERERUWK/9P1oNJpSR5WHDRuGnJwcvPHGGwCAbt26YdiwYeXab9++fbFkyRKMHz8erVq1wowZMyocGxFRTaUQRf9+SEREkoqNjUVMTAzee+89qUMhIqJSsIWDiKiG0Gq1iI6ORlBQkNShEBFRGVhAExHVACdPnsTEiRPh5uaGHj16SB0OERGVgS0cREREREQVwBFoIiIiIqIKYAFNRERERFQBLKCJiIiIiCqABTQRERERUQWwgCYiIiIiqgAW0EREREREFfD/sBM8UqX4t+YAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<Figure size 864x432 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Plot real vs. synthetic stock price\n",
    "plt.figure(figsize=(12,6))\n",
    "plt.plot(stocks.query('Twitter')['Day'], stocks.query('Twitter')['Close'], label='Twitter', color='blue')\n",
    "plt.plot(stocks.query('Twitter')['Day'], twitter_synth_lr, label='Synthetic Control', color='red')\n",
    "plt.vlines(x=25, ymin=30, ymax=65, linestyle=':', lw=2, label='Elon Musk Buys Twitter', color='black')\n",
    "plt.ylabel('Closing Price')\n",
    "plt.xlabel('Day of Month')\n",
    "plt.title('Twitter Stock Price vs. Synthetic Twitter')\n",
    "plt.legend();\n",
    "# Best representation of the impact of Elon Musk purchase"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Here were can clearly see the impact of the purchase. The real Twitter and the Synthetic follow closely before the purchase, and the real stock stayed fairly level after purchase, while the Synthetic forecasts that it would have dropped significantly if Musk did not buy it. "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "On April 29, Twitter’s stock price was around 49 dollars, and the Synthetic Twitter forecasts it to be 33 dollars, showing a 33% difference."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "`Treatment Effect`: Elon Musk purchasing Twitter caused the stock price to increase 33% more than if he did not purchase it."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Conclusions\n",
    "\n",
    "- Applying Synthetic Control to Twitter’s Stock Price allowed us to see the treatment effect most clearly.\n",
    "- Creating a Synthetic Sentiment score was not very reliable to see the impact of the purchase because sentiment scores are already calculated values and can be misleading and volatile at times.\n",
    "- Overall, we see that Elon Musk purchasing Twitter had a positive impact on the stock price."
   ]
  }
 ],
 "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": "3.8.5"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}