Sarah-Analysis.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Named Entity Recognition Analysis\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 114,
   "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>tweet_id</th>\n",
       "      <th>Compound_Score</th>\n",
       "      <th>Positive_Score</th>\n",
       "      <th>Negative_Score</th>\n",
       "      <th>Neutral_Score</th>\n",
       "      <th>external_author_id</th>\n",
       "      <th>author</th>\n",
       "      <th>content</th>\n",
       "      <th>region</th>\n",
       "      <th>language</th>\n",
       "      <th>publish_date</th>\n",
       "      <th>following</th>\n",
       "      <th>followers</th>\n",
       "      <th>updates</th>\n",
       "      <th>post_type</th>\n",
       "      <th>account_type</th>\n",
       "      <th>retweet</th>\n",
       "      <th>account_category</th>\n",
       "      <th>new_june_2018</th>\n",
       "      <th>alt_external_id</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>914580356430536707</td>\n",
       "      <td>0.0000</td>\n",
       "      <td>0.000</td>\n",
       "      <td>0.000</td>\n",
       "      <td>1.000</td>\n",
       "      <td>906000000000000000</td>\n",
       "      <td>10_GOP</td>\n",
       "      <td>\"We have a sitting Democrat US Senator on tria...</td>\n",
       "      <td>Unknown</td>\n",
       "      <td>English</td>\n",
       "      <td>2017-10-01 19:58:00</td>\n",
       "      <td>1052</td>\n",
       "      <td>9636</td>\n",
       "      <td>253</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Right</td>\n",
       "      <td>0</td>\n",
       "      <td>RightTroll</td>\n",
       "      <td>0</td>\n",
       "      <td>905874659358453760</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>914621840496189440</td>\n",
       "      <td>0.0000</td>\n",
       "      <td>0.000</td>\n",
       "      <td>0.000</td>\n",
       "      <td>1.000</td>\n",
       "      <td>906000000000000000</td>\n",
       "      <td>10_GOP</td>\n",
       "      <td>Marshawn Lynch arrives to game in anti-Trump s...</td>\n",
       "      <td>Unknown</td>\n",
       "      <td>English</td>\n",
       "      <td>2017-10-01 22:43:00</td>\n",
       "      <td>1054</td>\n",
       "      <td>9637</td>\n",
       "      <td>254</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Right</td>\n",
       "      <td>0</td>\n",
       "      <td>RightTroll</td>\n",
       "      <td>0</td>\n",
       "      <td>905874659358453760</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>914623490375979008</td>\n",
       "      <td>-0.1531</td>\n",
       "      <td>0.126</td>\n",
       "      <td>0.198</td>\n",
       "      <td>0.676</td>\n",
       "      <td>906000000000000000</td>\n",
       "      <td>10_GOP</td>\n",
       "      <td>Daughter of fallen Navy Sailor delivers powerf...</td>\n",
       "      <td>Unknown</td>\n",
       "      <td>English</td>\n",
       "      <td>2017-10-01 22:50:00</td>\n",
       "      <td>1054</td>\n",
       "      <td>9637</td>\n",
       "      <td>255</td>\n",
       "      <td>RETWEET</td>\n",
       "      <td>Right</td>\n",
       "      <td>1</td>\n",
       "      <td>RightTroll</td>\n",
       "      <td>0</td>\n",
       "      <td>905874659358453760</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>914639143690555392</td>\n",
       "      <td>0.0000</td>\n",
       "      <td>0.000</td>\n",
       "      <td>0.000</td>\n",
       "      <td>1.000</td>\n",
       "      <td>906000000000000000</td>\n",
       "      <td>10_GOP</td>\n",
       "      <td>JUST IN: President Trump dedicates Presidents ...</td>\n",
       "      <td>Unknown</td>\n",
       "      <td>English</td>\n",
       "      <td>2017-10-01 23:52:00</td>\n",
       "      <td>1062</td>\n",
       "      <td>9642</td>\n",
       "      <td>256</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Right</td>\n",
       "      <td>0</td>\n",
       "      <td>RightTroll</td>\n",
       "      <td>0</td>\n",
       "      <td>905874659358453760</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>914312219952861184</td>\n",
       "      <td>0.6399</td>\n",
       "      <td>0.413</td>\n",
       "      <td>0.000</td>\n",
       "      <td>0.587</td>\n",
       "      <td>906000000000000000</td>\n",
       "      <td>10_GOP</td>\n",
       "      <td>19,000 RESPECTING our National Anthem! #StandF...</td>\n",
       "      <td>Unknown</td>\n",
       "      <td>English</td>\n",
       "      <td>2017-10-01 02:13:00</td>\n",
       "      <td>1050</td>\n",
       "      <td>9645</td>\n",
       "      <td>246</td>\n",
       "      <td>RETWEET</td>\n",
       "      <td>Right</td>\n",
       "      <td>1</td>\n",
       "      <td>RightTroll</td>\n",
       "      <td>0</td>\n",
       "      <td>905874659358453760</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "             tweet_id  Compound_Score  Positive_Score  Negative_Score  \\\n",
       "0  914580356430536707          0.0000           0.000           0.000   \n",
       "1  914621840496189440          0.0000           0.000           0.000   \n",
       "2  914623490375979008         -0.1531           0.126           0.198   \n",
       "3  914639143690555392          0.0000           0.000           0.000   \n",
       "4  914312219952861184          0.6399           0.413           0.000   \n",
       "\n",
       "   Neutral_Score  external_author_id  author  \\\n",
       "0          1.000  906000000000000000  10_GOP   \n",
       "1          1.000  906000000000000000  10_GOP   \n",
       "2          0.676  906000000000000000  10_GOP   \n",
       "3          1.000  906000000000000000  10_GOP   \n",
       "4          0.587  906000000000000000  10_GOP   \n",
       "\n",
       "                                             content   region language  \\\n",
       "0  \"We have a sitting Democrat US Senator on tria...  Unknown  English   \n",
       "1  Marshawn Lynch arrives to game in anti-Trump s...  Unknown  English   \n",
       "2  Daughter of fallen Navy Sailor delivers powerf...  Unknown  English   \n",
       "3  JUST IN: President Trump dedicates Presidents ...  Unknown  English   \n",
       "4  19,000 RESPECTING our National Anthem! #StandF...  Unknown  English   \n",
       "\n",
       "          publish_date  following  followers  updates post_type account_type  \\\n",
       "0  2017-10-01 19:58:00       1052       9636      253       NaN        Right   \n",
       "1  2017-10-01 22:43:00       1054       9637      254       NaN        Right   \n",
       "2  2017-10-01 22:50:00       1054       9637      255   RETWEET        Right   \n",
       "3  2017-10-01 23:52:00       1062       9642      256       NaN        Right   \n",
       "4  2017-10-01 02:13:00       1050       9645      246   RETWEET        Right   \n",
       "\n",
       "   retweet account_category  new_june_2018     alt_external_id  \n",
       "0        0       RightTroll              0  905874659358453760  \n",
       "1        0       RightTroll              0  905874659358453760  \n",
       "2        1       RightTroll              0  905874659358453760  \n",
       "3        0       RightTroll              0  905874659358453760  \n",
       "4        1       RightTroll              0  905874659358453760  "
      ]
     },
     "execution_count": 114,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "%matplotlib inline\n",
    "# Dependencies and Setup\n",
    "import matplotlib.pyplot as plt\n",
    "import pandas as pd\n",
    "import numpy as np\n",
    "import spacy\n",
    "from spacy import displacy\n",
    "from spacymoji import Emoji\n",
    "\n",
    "# import data and read to df\n",
    "csv = \"../Output/all_data.csv\"\n",
    "df = pd.read_csv(csv, low_memory=False)\n",
    "\n",
    "df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 124,
   "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>Compound_Score</th>\n",
       "      <th>Positive_Score</th>\n",
       "      <th>Negative_Score</th>\n",
       "      <th>Neutral_Score</th>\n",
       "      <th>updates</th>\n",
       "      <th>following</th>\n",
       "      <th>followers</th>\n",
       "      <th>account_type</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>author</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>10_GOP</th>\n",
       "      <td>0.078718</td>\n",
       "      <td>0.131511</td>\n",
       "      <td>0.089804</td>\n",
       "      <td>0.778685</td>\n",
       "      <td>176.500000</td>\n",
       "      <td>985.018817</td>\n",
       "      <td>7044.306452</td>\n",
       "      <td>Right</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1D_NICOLE_</th>\n",
       "      <td>0.177359</td>\n",
       "      <td>0.193545</td>\n",
       "      <td>0.069000</td>\n",
       "      <td>0.737455</td>\n",
       "      <td>373.500000</td>\n",
       "      <td>58.500000</td>\n",
       "      <td>51.477273</td>\n",
       "      <td>Koch</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1ERIK_LEE</th>\n",
       "      <td>-0.549850</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.214000</td>\n",
       "      <td>0.786000</td>\n",
       "      <td>333.000000</td>\n",
       "      <td>239.000000</td>\n",
       "      <td>74.000000</td>\n",
       "      <td>Right</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2NDHALFONION</th>\n",
       "      <td>0.439200</td>\n",
       "      <td>0.186000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.814000</td>\n",
       "      <td>10.000000</td>\n",
       "      <td>22.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>Right</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4EVER_SUSAN</th>\n",
       "      <td>0.042470</td>\n",
       "      <td>0.133794</td>\n",
       "      <td>0.104810</td>\n",
       "      <td>0.761413</td>\n",
       "      <td>37.047619</td>\n",
       "      <td>75.650794</td>\n",
       "      <td>57.777778</td>\n",
       "      <td>Right</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "              Compound_Score  Positive_Score  Negative_Score  Neutral_Score  \\\n",
       "author                                                                        \n",
       "10_GOP              0.078718        0.131511        0.089804       0.778685   \n",
       "1D_NICOLE_          0.177359        0.193545        0.069000       0.737455   \n",
       "1ERIK_LEE          -0.549850        0.000000        0.214000       0.786000   \n",
       "2NDHALFONION        0.439200        0.186000        0.000000       0.814000   \n",
       "4EVER_SUSAN         0.042470        0.133794        0.104810       0.761413   \n",
       "\n",
       "                 updates   following    followers account_type  \n",
       "author                                                          \n",
       "10_GOP        176.500000  985.018817  7044.306452        Right  \n",
       "1D_NICOLE_    373.500000   58.500000    51.477273         Koch  \n",
       "1ERIK_LEE     333.000000  239.000000    74.000000        Right  \n",
       "2NDHALFONION   10.000000   22.000000     1.000000        Right  \n",
       "4EVER_SUSAN    37.047619   75.650794    57.777778        Right  "
      ]
     },
     "execution_count": 124,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# group the dataframe by account because I'm curious\n",
    "\n",
    "account_df = df[[\"author\", \"Compound_Score\", \"Positive_Score\", \"Negative_Score\", \"Neutral_Score\", \"updates\", \"following\", \"followers\"]]\n",
    "account_type_ref = df[[\"author\", \"account_type\"]].drop_duplicates()\n",
    "account_df = account_df.groupby(df[\"author\"]).mean()\n",
    "account_df = account_df.merge(account_type_ref, left_index=True, right_on='author').set_index('author')\n",
    "\n",
    "account_df.head(5)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 56,
   "metadata": {},
   "outputs": [],
   "source": [
    "# import and create a default dictionary to store words as it finds them from function below\n",
    "from collections import defaultdict\n",
    "word_counts = defaultdict(int) #template for simpler function code\n",
    "\n",
    "# Load the spaCy model\n",
    "nlp = spacy.load('en_core_web_sm')\n",
    "\n",
    "# make separate dfs for left and right trolls\n",
    "right_troll_df = df[df[\"account_category\"] == \"RightTroll\"]\n",
    "left_troll_df = df[df[\"account_category\"] == \"LeftTroll\"]\n",
    "\n",
    "# function to run NER \n",
    "def getEntities(tweet):\n",
    "    output = nlp(tweet)\n",
    "    ents = output.ents\n",
    "    return_list = []\n",
    "    for ent in ents:\n",
    "        return_list.append(ent.text.lower())\n",
    "        word_counts[ent.text.lower()] += 1\n",
    "    return return_list\n",
    "\n",
    "# run on a random sample of Left and Right Trolls\n",
    "Right = right_troll_df['content'].sample(n=25000).apply(getEntities)\n",
    "Right_Counts = word_counts #dict for right measurements\n",
    "del word_counts \n",
    "word_counts = defaultdict(int)\n",
    "Left = left_troll_df['content'].sample(n=25000).apply(getEntities)\n",
    "Left_Counts = word_counts #dict for left measurements\n",
    "del word_counts"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 57,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1116 - rt\n",
      "801 -  \n",
      "757 - #\n",
      "652 - obama\n",
      "594 - hillary\n",
      "507 - trump\n",
      "421 - maga\n",
      "402 - america\n",
      "382 - cnn\n",
      "291 - us\n",
      "263 - one\n",
      "241 - american\n",
      "239 - gop\n",
      "238 - russia\n",
      "218 - democrats\n",
      "205 - clinton\n",
      "203 - u.s.\n",
      "201 - today\n",
      "194 - fbi\n",
      "187 - 🇺\n",
      "184 - 2\n",
      "183 - americans\n",
      "182 - donald trump\n",
      "180 - first\n",
      "177 - fakenews\n",
      "167 - democrat\n",
      "158 - muslim\n",
      "157 - congress\n",
      "157 - dems\n",
      "157 - hillary clinton\n",
      "145 - lol\n",
      "144 - pjnet\n",
      "137 - antifa\n",
      "133 - nfl\n",
      "131 - 4\n",
      "127 - iran\n",
      "127 - dem\n",
      "126 - north korea\n",
      "126 - charlottesville\n",
      "115 - dnc\n",
      "115 - muslims\n",
      "114 - two\n",
      "112 - ️\n",
      "108 - california\n",
      "107 - republicans\n",
      "105 - #mar\n",
      "102 - msm\n",
      "101 - texas\n",
      "95 - islam\n",
      "95 - israel\n"
     ]
    }
   ],
   "source": [
    "#get the value from the key-value dictionary\n",
    "def getSecond(x):\n",
    "    return x[1]\n",
    "\n",
    "#Get Top Entities from the Right Trolls df and sort by most instances to least\n",
    "top = 50\n",
    "sorted_entities = sorted(Right_Counts.items(), key=getSecond, reverse=True)\n",
    "for ent in sorted_entities[0:top]:\n",
    "    print(\"{} - {}\".format(ent[1], ent[0]))\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 59,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "733 -  \n",
      "621 - #\n",
      "481 - blacklivesmatter\n",
      "326 - trump\n",
      "314 - today\n",
      "299 - one\n",
      "269 - first\n",
      "231 - 2\n",
      "216 - america\n",
      "215 - us\n",
      "187 - rt\n",
      "166 - ️\n",
      "158 - obama\n",
      "137 - chicago\n",
      "136 - 1\n",
      "135 - 4\n",
      "128 - 2016\n",
      "122 - donald trump\n",
      "122 - gop\n",
      "117 - 3\n",
      "115 - @youtube\n",
      "112 - russia\n",
      "107 - 5\n",
      "105 - tonight\n",
      "105 - trump's\n",
      "105 - american\n",
      "103 - two\n",
      "95 - black\n",
      "90 - 2017\n",
      "90 - u.s.\n",
      "82 - 10\n",
      "75 - americans\n",
      "74 - blm\n",
      "72 - clinton\n",
      "68 - staywoke\n",
      "64 - hillary\n",
      "64 - republicans\n",
      "64 - blackhistorymonth\n",
      "63 - dagr8fm\n",
      "60 - 6\n",
      "60 - texas\n",
      "60 - god\n",
      "58 - @realdonaldtrump\n",
      "57 - tomorrow\n",
      "57 - muslim\n",
      "56 - rapstationradio\n",
      "56 - baltimore\n",
      "55 - \\|\n",
      "55 - fbi\n",
      "53 - @josephjett\n"
     ]
    }
   ],
   "source": [
    "#Get Top Entities from the Left Trolls df and sort by most instances to least\n",
    "top = 50\n",
    "sorted_entities = sorted(Left_Counts.items(), key=getSecond, reverse=True)\n",
    "for ent in sorted_entities[0:top]:\n",
    "    print(\"{} - {}\".format(ent[1], ent[0]))\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Most common words \n",
    "Here are some of the most common words between left and right trolls. The search_terms list below will include different spellings of these words or words that are very closely associated, like if \"trump\" was a most used word, \"donald\" would mean the same thing and be added to this list. I will search all of these words with a function to find content that \"contains\" certain words so that I can see a comprehensive overview of how many tweets mentioned the subjects, and not just the most used word of the subjects below.\n",
    "\n",
    "1. Trump\n",
    "2. BlackLivesMatter\n",
    "3. Hillary\n",
    "4. Wall\n",
    "5. Obama\n",
    "6. America\n",
    "7. Republican\n",
    "8. Democrat\n",
    "9. MAGA\n",
    "10. Islam\n",
    "11. Fake news\n",
    "12. Russia\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# make a df for the search terms so you can store the frequency that they appear in a search \n",
    "    # and the avg sentiment score that the right and the left associates with them\n",
    "common_words_scored = [\"Subject\", \"Right_Frequency\", \"Left_Frequency\", \"Right_Sentiment\", \"Left_Sentiment\"]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 93,
   "metadata": {},
   "outputs": [],
   "source": [
    "search_terms = {\"Trump\": [\"Trump\", \"Donald\", \"DJT\", \"trump's\", \"realdonaldtrump\"],\n",
    "                \"BlackLivesMatter\": [\"BlackLivesMatter\", \"blacklives\", \"black\", \"blm\", \"get woke\", \"blackhistorymonth\", \"african american\"],\n",
    "                \"Hillary\": [\"Hillary\", \"Clinton\"],\n",
    "                \"Wall\": [\"Wall\", \"build\", \"border\"],\n",
    "                \"Obama\": [\"Obama\", \"obamacare\", \"barack\"],\n",
    "                \"America\": [\"America\", \"USA\", \"united states\", \"merca\", \"u.s.\", \"American\"],\n",
    "                \"Republican\": [\"Republican\", \"gop\", \"conservative\", \"right wing\"],\n",
    "                \"Democrat\": [\"Democrat\", \"dem\", \"libs\", \"liberal\", \"left wing\", \"dnc\"],\n",
    "                \"Maga\": [\"Maga\", \"make america great again\"],\n",
    "                \"Islam\": [\"Islam\", \"muslim\", \"iran\", \"iraq\", \"israel\", \"middle east\",\"isis\"],\n",
    "                \"Fakenews\": [\"Fake news\",\"fakenews\",\"fox\",\"cnn\",\"msnbc\",\"WaPo\",\"Washington Post\",\"NYT\",\"New York Times\",\"HuffPost\",\"Huffington\",\"Yahoo\"],\n",
    "                \"Russia\": [\"Russia\"]\n",
    "               }"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 202,
   "metadata": {
    "scrolled": true
   },
   "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>Topic</th>\n",
       "      <th>Right_Frequency</th>\n",
       "      <th>Left_Frequency</th>\n",
       "      <th>Right_Sentiment</th>\n",
       "      <th>Left_Sentiment</th>\n",
       "      <th>Right_Sentiment_Std</th>\n",
       "      <th>Left_Sentiment_Std</th>\n",
       "      <th>Right_Updates</th>\n",
       "      <th>Left_Updates</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>Trump</td>\n",
       "      <td>139455</td>\n",
       "      <td>28178</td>\n",
       "      <td>-0.035907</td>\n",
       "      <td>-0.069399</td>\n",
       "      <td>0.456759</td>\n",
       "      <td>0.456759</td>\n",
       "      <td>10070.070740</td>\n",
       "      <td>3576.116509</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>BlackLivesMatter</td>\n",
       "      <td>14513</td>\n",
       "      <td>53130</td>\n",
       "      <td>-0.145995</td>\n",
       "      <td>-0.018735</td>\n",
       "      <td>0.454990</td>\n",
       "      <td>0.454990</td>\n",
       "      <td>8034.139392</td>\n",
       "      <td>3078.636345</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>Hillary</td>\n",
       "      <td>52230</td>\n",
       "      <td>5311</td>\n",
       "      <td>-0.085268</td>\n",
       "      <td>-0.013561</td>\n",
       "      <td>0.422539</td>\n",
       "      <td>0.422539</td>\n",
       "      <td>9004.150029</td>\n",
       "      <td>2584.819808</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>Wall</td>\n",
       "      <td>10403</td>\n",
       "      <td>3785</td>\n",
       "      <td>-0.049571</td>\n",
       "      <td>0.063941</td>\n",
       "      <td>0.430326</td>\n",
       "      <td>0.430326</td>\n",
       "      <td>12549.104681</td>\n",
       "      <td>3792.928666</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>Obama</td>\n",
       "      <td>45185</td>\n",
       "      <td>5938</td>\n",
       "      <td>-0.107213</td>\n",
       "      <td>-0.016021</td>\n",
       "      <td>0.434051</td>\n",
       "      <td>0.434051</td>\n",
       "      <td>7752.373819</td>\n",
       "      <td>3443.387504</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>America</td>\n",
       "      <td>135688</td>\n",
       "      <td>56162</td>\n",
       "      <td>-0.041184</td>\n",
       "      <td>0.004665</td>\n",
       "      <td>0.461271</td>\n",
       "      <td>0.461271</td>\n",
       "      <td>10244.696281</td>\n",
       "      <td>3618.212368</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>Republican</td>\n",
       "      <td>28177</td>\n",
       "      <td>6593</td>\n",
       "      <td>-0.034941</td>\n",
       "      <td>-0.055502</td>\n",
       "      <td>0.433535</td>\n",
       "      <td>0.433535</td>\n",
       "      <td>8166.365546</td>\n",
       "      <td>3034.659487</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>Democrat</td>\n",
       "      <td>59551</td>\n",
       "      <td>8762</td>\n",
       "      <td>-0.159778</td>\n",
       "      <td>-0.040120</td>\n",
       "      <td>0.446818</td>\n",
       "      <td>0.446818</td>\n",
       "      <td>9717.473796</td>\n",
       "      <td>3416.699612</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>Maga</td>\n",
       "      <td>18707</td>\n",
       "      <td>2317</td>\n",
       "      <td>0.045751</td>\n",
       "      <td>0.126126</td>\n",
       "      <td>0.453496</td>\n",
       "      <td>0.453496</td>\n",
       "      <td>18903.959641</td>\n",
       "      <td>3899.510574</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>Islam</td>\n",
       "      <td>42693</td>\n",
       "      <td>6939</td>\n",
       "      <td>-0.186188</td>\n",
       "      <td>-0.184455</td>\n",
       "      <td>0.447204</td>\n",
       "      <td>0.447204</td>\n",
       "      <td>6979.531492</td>\n",
       "      <td>3090.720421</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10</th>\n",
       "      <td>Fakenews</td>\n",
       "      <td>35193</td>\n",
       "      <td>7394</td>\n",
       "      <td>-0.123897</td>\n",
       "      <td>-0.031851</td>\n",
       "      <td>0.431529</td>\n",
       "      <td>0.431529</td>\n",
       "      <td>15550.683261</td>\n",
       "      <td>3521.203814</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11</th>\n",
       "      <td>Russia</td>\n",
       "      <td>11948</td>\n",
       "      <td>3546</td>\n",
       "      <td>-0.163501</td>\n",
       "      <td>-0.071404</td>\n",
       "      <td>0.420719</td>\n",
       "      <td>0.420719</td>\n",
       "      <td>13048.254017</td>\n",
       "      <td>4725.677383</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "               Topic  Right_Frequency  Left_Frequency  Right_Sentiment  \\\n",
       "0              Trump           139455           28178        -0.035907   \n",
       "1   BlackLivesMatter            14513           53130        -0.145995   \n",
       "2            Hillary            52230            5311        -0.085268   \n",
       "3               Wall            10403            3785        -0.049571   \n",
       "4              Obama            45185            5938        -0.107213   \n",
       "5            America           135688           56162        -0.041184   \n",
       "6         Republican            28177            6593        -0.034941   \n",
       "7           Democrat            59551            8762        -0.159778   \n",
       "8               Maga            18707            2317         0.045751   \n",
       "9              Islam            42693            6939        -0.186188   \n",
       "10          Fakenews            35193            7394        -0.123897   \n",
       "11            Russia            11948            3546        -0.163501   \n",
       "\n",
       "    Left_Sentiment  Right_Sentiment_Std  Left_Sentiment_Std  Right_Updates  \\\n",
       "0        -0.069399             0.456759            0.456759   10070.070740   \n",
       "1        -0.018735             0.454990            0.454990    8034.139392   \n",
       "2        -0.013561             0.422539            0.422539    9004.150029   \n",
       "3         0.063941             0.430326            0.430326   12549.104681   \n",
       "4        -0.016021             0.434051            0.434051    7752.373819   \n",
       "5         0.004665             0.461271            0.461271   10244.696281   \n",
       "6        -0.055502             0.433535            0.433535    8166.365546   \n",
       "7        -0.040120             0.446818            0.446818    9717.473796   \n",
       "8         0.126126             0.453496            0.453496   18903.959641   \n",
       "9        -0.184455             0.447204            0.447204    6979.531492   \n",
       "10       -0.031851             0.431529            0.431529   15550.683261   \n",
       "11       -0.071404             0.420719            0.420719   13048.254017   \n",
       "\n",
       "    Left_Updates  \n",
       "0    3576.116509  \n",
       "1    3078.636345  \n",
       "2    2584.819808  \n",
       "3    3792.928666  \n",
       "4    3443.387504  \n",
       "5    3618.212368  \n",
       "6    3034.659487  \n",
       "7    3416.699612  \n",
       "8    3899.510574  \n",
       "9    3090.720421  \n",
       "10   3521.203814  \n",
       "11   4725.677383  "
      ]
     },
     "execution_count": 202,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "final_df = defaultdict(list)\n",
    "right_troll_df[\"content\"] = right_troll_df['content'].str.lower()\n",
    "left_troll_df[\"content\"] = left_troll_df['content'].str.lower()\n",
    "\n",
    "# functions to find the frequency of common words and their avg sentiment from left and right account\n",
    "\n",
    "def calculateRightFrequency(termsList):\n",
    "    related_rows = right_troll_df[right_troll_df['content'].str.contains(\"|\".join(termsList))]\n",
    "    return len(related_rows)\n",
    "\n",
    "def calculateLeftFrequency(termsList):\n",
    "    related_rows = left_troll_df[left_troll_df['content'].str.contains(\"|\".join(termsList))]\n",
    "    return len(related_rows)\n",
    "\n",
    "def calculateRightSentiment(termsList):\n",
    "    related_rows = right_troll_df[right_troll_df['content'].str.contains(\"|\".join(termsList))]\n",
    "    return pd.to_numeric(related_rows[\"Compound_Score\"]).mean()\n",
    "\n",
    "def calculateLeftSentiment(termsList):\n",
    "    related_rows = left_troll_df[left_troll_df['content'].str.contains(\"|\".join(termsList))]\n",
    "    return pd.to_numeric(related_rows[\"Compound_Score\"]).mean()\n",
    "\n",
    "def calculateRightStdDev(termsList):\n",
    "    related_rows = right_troll_df[right_troll_df['content'].str.contains(\"|\".join(termsList))]\n",
    "    return pd.to_numeric(related_rows[\"Compound_Score\"]).std()\n",
    "\n",
    "def calculateLeftStdDev(termsList):\n",
    "    related_rows = left_troll_df[left_troll_df['content'].str.contains(\"|\".join(termsList))]\n",
    "    return pd.to_numeric(related_rows[\"Compound_Score\"]).std()\n",
    "\n",
    "def calculateRightUpdates(termsList):\n",
    "    related_rows = right_troll_df[right_troll_df['content'].str.contains(\"|\".join(termsList))]\n",
    "    return pd.to_numeric(related_rows[\"updates\"]).mean()\n",
    "\n",
    "def calculateLeftUpdates(termsList):\n",
    "    related_rows = left_troll_df[left_troll_df['content'].str.contains(\"|\".join(termsList))]\n",
    "    return pd.to_numeric(related_rows[\"updates\"]).mean()\n",
    "\n",
    "for term in search_terms:\n",
    "    search_terms[term] = [ t.lower() for t in search_terms[term] ]\n",
    "    final_df['Topic'].append(term)\n",
    "    final_df[\"Right_Frequency\"].append(calculateRightFrequency(search_terms[term]))\n",
    "    final_df[\"Left_Frequency\"].append(calculateLeftFrequency(search_terms[term]))\n",
    "    final_df[\"Right_Sentiment\"].append(calculateRightSentiment(search_terms[term]))\n",
    "    final_df[\"Left_Sentiment\"].append(calculateLeftSentiment(search_terms[term]))\n",
    "    final_df[\"Right_Sentiment_Std\"].append(calculateRightStdDev(search_terms[term]))\n",
    "    final_df[\"Left_Sentiment_Std\"].append(calculateRightStdDev(search_terms[term]))\n",
    "    final_df[\"Right_Updates\"].append(calculateRightUpdates(search_terms[term]))\n",
    "    final_df[\"Left_Updates\"].append(calculateLeftUpdates(search_terms[term]))\n",
    "\n",
    "final_df = pd.DataFrame(final_df)\n",
    "final_df.head(12)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 234,
   "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>Topic</th>\n",
       "      <th>Right_Frequency</th>\n",
       "      <th>Left_Frequency</th>\n",
       "      <th>Right_Sentiment</th>\n",
       "      <th>Left_Sentiment</th>\n",
       "      <th>Right_Sentiment_Std</th>\n",
       "      <th>Left_Sentiment_Std</th>\n",
       "      <th>Right_Updates</th>\n",
       "      <th>Left_Updates</th>\n",
       "      <th>Cumulative_Appearance</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>Trump</td>\n",
       "      <td>139455</td>\n",
       "      <td>28178</td>\n",
       "      <td>-0.035907</td>\n",
       "      <td>-0.069399</td>\n",
       "      <td>0.456759</td>\n",
       "      <td>0.456759</td>\n",
       "      <td>10070.070740</td>\n",
       "      <td>3576.116509</td>\n",
       "      <td>167633</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>BlackLivesMatter</td>\n",
       "      <td>14513</td>\n",
       "      <td>53130</td>\n",
       "      <td>-0.145995</td>\n",
       "      <td>-0.018735</td>\n",
       "      <td>0.454990</td>\n",
       "      <td>0.454990</td>\n",
       "      <td>8034.139392</td>\n",
       "      <td>3078.636345</td>\n",
       "      <td>67643</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>Hillary</td>\n",
       "      <td>52230</td>\n",
       "      <td>5311</td>\n",
       "      <td>-0.085268</td>\n",
       "      <td>-0.013561</td>\n",
       "      <td>0.422539</td>\n",
       "      <td>0.422539</td>\n",
       "      <td>9004.150029</td>\n",
       "      <td>2584.819808</td>\n",
       "      <td>57541</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>Wall</td>\n",
       "      <td>10403</td>\n",
       "      <td>3785</td>\n",
       "      <td>-0.049571</td>\n",
       "      <td>0.063941</td>\n",
       "      <td>0.430326</td>\n",
       "      <td>0.430326</td>\n",
       "      <td>12549.104681</td>\n",
       "      <td>3792.928666</td>\n",
       "      <td>14188</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>Obama</td>\n",
       "      <td>45185</td>\n",
       "      <td>5938</td>\n",
       "      <td>-0.107213</td>\n",
       "      <td>-0.016021</td>\n",
       "      <td>0.434051</td>\n",
       "      <td>0.434051</td>\n",
       "      <td>7752.373819</td>\n",
       "      <td>3443.387504</td>\n",
       "      <td>51123</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "              Topic  Right_Frequency  Left_Frequency  Right_Sentiment  \\\n",
       "0             Trump           139455           28178        -0.035907   \n",
       "1  BlackLivesMatter            14513           53130        -0.145995   \n",
       "2           Hillary            52230            5311        -0.085268   \n",
       "3              Wall            10403            3785        -0.049571   \n",
       "4             Obama            45185            5938        -0.107213   \n",
       "\n",
       "   Left_Sentiment  Right_Sentiment_Std  Left_Sentiment_Std  Right_Updates  \\\n",
       "0       -0.069399             0.456759            0.456759   10070.070740   \n",
       "1       -0.018735             0.454990            0.454990    8034.139392   \n",
       "2       -0.013561             0.422539            0.422539    9004.150029   \n",
       "3        0.063941             0.430326            0.430326   12549.104681   \n",
       "4       -0.016021             0.434051            0.434051    7752.373819   \n",
       "\n",
       "   Left_Updates  Cumulative_Appearance  \n",
       "0   3576.116509                 167633  \n",
       "1   3078.636345                  67643  \n",
       "2   2584.819808                  57541  \n",
       "3   3792.928666                  14188  \n",
       "4   3443.387504                  51123  "
      ]
     },
     "execution_count": 234,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#find and sort by the highest cumulative frequency topics appeared\n",
    "final_df[\"Cumulative_Appearance\"] = final_df[\"Right_Frequency\"] + final_df[\"Left_Frequency\"]\n",
    "final_df.sort_values(\"Cumulative_Appearance\", ascending=False)\n",
    "final_df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 292,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[\"this eminem freestyle proves that celebs still haven't learned that nobody cares what they have to say. sad!  #hiphopawards\",\n",
       " 'eminem is a trump supporter?!  😮 https://t.co/jncuj7sqnu',\n",
       " 'cringe video: eminem goes on anti-trump rant on bet awards https://t.co/z2abknmgob',\n",
       " 'barf! kaepernick gives eminem’s anti-trump rant a ‘thumbs up’ https://t.co/7ethvo4ygv',\n",
       " 'oops! liberals who love eminem have conveniently forgotten his sexist lyrics https://t.co/ccmbqsxcjj']"
      ]
     },
     "execution_count": 292,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#find example tweets for presentation\n",
    "list(right_troll_df[right_troll_df[\"content\"].str.contains(\"eminem\")][\"content\"].head())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 96,
   "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>content</th>\n",
       "      <th>Neutral_Score</th>\n",
       "      <th>Negative_Score</th>\n",
       "      <th>Positive_Score</th>\n",
       "      <th>Compound_Score</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>35975</th>\n",
       "      <td>rt richardstiller4: polnewsforever ten_gop tha...</td>\n",
       "      <td>0.49</td>\n",
       "      <td>0.307</td>\n",
       "      <td>0.203</td>\n",
       "      <td>-0.3595</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>43986</th>\n",
       "      <td>new post: thanks obama! illegal alien satantis...</td>\n",
       "      <td>0.519</td>\n",
       "      <td>0.34600000000000003</td>\n",
       "      <td>0.135</td>\n",
       "      <td>-0.7712</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>80656</th>\n",
       "      <td>it's cheap and easy to say, \"thanks obama!\" th...</td>\n",
       "      <td>0.757</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.243</td>\n",
       "      <td>0.7263</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>93460</th>\n",
       "      <td>does it seem like everyone in the world wants ...</td>\n",
       "      <td>0.693</td>\n",
       "      <td>0.129</td>\n",
       "      <td>0.17800000000000002</td>\n",
       "      <td>0.128</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>165041</th>\n",
       "      <td>isis celebrates victory in ramadi ➠ thanks oba...</td>\n",
       "      <td>0.488</td>\n",
       "      <td>0.19</td>\n",
       "      <td>0.322</td>\n",
       "      <td>0.4019</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                                  content Neutral_Score  \\\n",
       "35975   rt richardstiller4: polnewsforever ten_gop tha...          0.49   \n",
       "43986   new post: thanks obama! illegal alien satantis...         0.519   \n",
       "80656   it's cheap and easy to say, \"thanks obama!\" th...         0.757   \n",
       "93460   does it seem like everyone in the world wants ...         0.693   \n",
       "165041  isis celebrates victory in ramadi ➠ thanks oba...         0.488   \n",
       "\n",
       "             Negative_Score       Positive_Score Compound_Score  \n",
       "35975                 0.307                0.203        -0.3595  \n",
       "43986   0.34600000000000003                0.135        -0.7712  \n",
       "80656                   0.0                0.243         0.7263  \n",
       "93460                 0.129  0.17800000000000002          0.128  \n",
       "165041                 0.19                0.322         0.4019  "
      ]
     },
     "execution_count": 96,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#look at tweets to see how the sentiment scores did on sarcasm\n",
    "(right_troll_df[right_troll_df[\"content\"].str.contains(\"thanks obama\")][[\"content\",\"Neutral_Score\",\"Negative_Score\",\"Positive_Score\",\"Compound_Score\"]]).head(5)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 238,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "<Figure size 432x288 with 0 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "#find a better weay to plot. start again\n",
    "ind = 1\n",
    "\n",
    "# get the min sentiment and max sentiment in order to plot the Text later on the midpoint\n",
    "def getXTextLocation(locations):\n",
    "    lower = min(locations)\n",
    "    higher = max(locations)\n",
    "    return ((higher - lower)/2)+lower\n",
    "\n",
    "fig,ax = plt.subplots(1)\n",
    "\n",
    "for row in final_df.iterrows():\n",
    "    row_data = row[1]\n",
    "    ax.plot([row_data['Right_Sentiment'], row_data['Left_Sentiment']], [ind, ind], c='black')\n",
    "    ax.scatter([row_data['Right_Sentiment'], row_data['Left_Sentiment']], [ind, ind], s=[row_data['Right_Frequency']/500, row_data['Left_Frequency']/500], c=['red', 'blue'])\n",
    "    ax.text(getXTextLocation([row_data['Right_Sentiment'], row_data['Left_Sentiment']]), ind+.15, row_data['Topic'])\n",
    "    ind += 1\n",
    "plt.xlim(-1, 1) \n",
    "ax.get_yaxis().set_visible(False)\n",
    "plt.title(\"Sentiment and Frequency of Most Used Words\")\n",
    "plt.xlabel(\"Compound VADER Score\")\n",
    "plt.show()\n",
    "plt.savefig(\"Images/sentiment_leftandright_1.png\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 295,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 576x504 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "<Figure size 432x288 with 0 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "#narrow in for better viewing of data\n",
    "index = 1\n",
    "\n",
    "# get the min sentiment and max sentiment in order to plot the Text on the midpoints\n",
    "def getXTextLocation(locations):\n",
    "    lower = min(locations)\n",
    "    higher = max(locations)\n",
    "    return ((higher - lower)/2)+lower # midpoint\n",
    "\n",
    "# make one figure\n",
    "fig,ax = plt.subplots(1, figsize=(8, 7))\n",
    "\n",
    "# loop through rows to make both a line graph and scatter plot, then call the function to add the text labels\n",
    "for row in final_df.iterrows():\n",
    "    row_data = row[1] #find value\n",
    "    ax.scatter([row_data['Right_Sentiment'], row_data['Left_Sentiment']], [index, index], s=[row_data['Right_Frequency']/300, row_data['Left_Frequency']/300], c=['red', 'blue'], zorder=1)\n",
    "    ax.plot([row_data['Right_Sentiment'], row_data['Left_Sentiment']], [index, index], c='black',zorder=-1)\n",
    "    ax.text(getXTextLocation([row_data['Right_Sentiment'], row_data['Left_Sentiment']]), index+.3, row_data['Topic'],horizontalalignment='center',fontsize=12) #add labels\n",
    "    index += 1.5\n",
    "plt.xlim(-.25, .25) \n",
    "ax.get_yaxis().set_visible(False)\n",
    "\n",
    "# Incorporate the other graph properties\n",
    "plt.title(\"Sentiment and Frequency of Most Used Words\", fontsize=15, fontweight='bold')\n",
    "plt.xlabel(\"Compound VADER Score\", fontsize=12, fontweight='bold')\n",
    "plt.axvline(x=0,linewidth=100,color='gray',alpha=.1)\n",
    "plt.axvline(x=0,linewidth=.5,color='gray')\n",
    "plt.ylim(0,19)\n",
    "textstr = \"Size corresponds with the amount of times the topic was used in a tweet.\"\n",
    "plt.text(-.22,-3,textstr, fontsize=11, style='italic')\n",
    "\n",
    "# show plot and save as a png\n",
    "plt.show()\n",
    "plt.savefig(\"Images/sentiment_leftandright_2.png\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.6.5"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}