Tweets - Hashtag Analysis.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "d25e18a8",
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "import numpy as np\n",
    "\n",
    "from collections import Counter\n",
    "\n",
    "import statsmodels.api as sm\n",
    "\n",
    "from tqdm.auto import tqdm\n",
    "tqdm.pandas()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "52c9499f",
   "metadata": {},
   "source": [
    "## Reading the data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "9140491a",
   "metadata": {},
   "outputs": [],
   "source": [
    "df = pd.read_csv('Tweets - Sentiment Analysis (RoBERTa) Raw Values.csv', lineterminator='\\n')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "9bb3e0bc",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'pandas.core.frame.DataFrame'>\n",
      "RangeIndex: 3735462 entries, 0 to 3735461\n",
      "Data columns (total 12 columns):\n",
      " #   Column    Dtype  \n",
      "---  ------    -----  \n",
      " 0   tweetid   int64  \n",
      " 1   text      object \n",
      " 2   hashtags  object \n",
      " 3   language  object \n",
      " 4   date      object \n",
      " 5   negative  float64\n",
      " 6   neutral   float64\n",
      " 7   positive  float64\n",
      " 8   anger     float64\n",
      " 9   joy       float64\n",
      " 10  optimism  float64\n",
      " 11  sadness   float64\n",
      "dtypes: float64(7), int64(1), object(4)\n",
      "memory usage: 342.0+ MB\n"
     ]
    }
   ],
   "source": [
    "df.info()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "d26e52d6",
   "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>tweetid</th>\n",
       "      <th>text</th>\n",
       "      <th>hashtags</th>\n",
       "      <th>language</th>\n",
       "      <th>date</th>\n",
       "      <th>negative</th>\n",
       "      <th>neutral</th>\n",
       "      <th>positive</th>\n",
       "      <th>anger</th>\n",
       "      <th>joy</th>\n",
       "      <th>optimism</th>\n",
       "      <th>sadness</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1499174584720969730</td>\n",
       "      <td>Map situation in #Ukraine after the seventh da...</td>\n",
       "      <td>['Ukraine', 'RussiaUkraineConflict']</td>\n",
       "      <td>en</td>\n",
       "      <td>MAR03</td>\n",
       "      <td>0.237624</td>\n",
       "      <td>1.357651</td>\n",
       "      <td>-1.783151</td>\n",
       "      <td>1.320624</td>\n",
       "      <td>-1.717873</td>\n",
       "      <td>-0.575291</td>\n",
       "      <td>1.707901</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1499174584976826368</td>\n",
       "      <td>#Ukraine: Let's just say it's not just the TB-...</td>\n",
       "      <td>['Ukraine']</td>\n",
       "      <td>en</td>\n",
       "      <td>MAR03</td>\n",
       "      <td>0.947360</td>\n",
       "      <td>0.889939</td>\n",
       "      <td>-2.091296</td>\n",
       "      <td>1.598669</td>\n",
       "      <td>-0.535707</td>\n",
       "      <td>0.341185</td>\n",
       "      <td>-0.105763</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>1499174585073242116</td>\n",
       "      <td>⚡️The SWIFT company confirmed that it will dis...</td>\n",
       "      <td>['EU', 'Russian']</td>\n",
       "      <td>en</td>\n",
       "      <td>MAR03</td>\n",
       "      <td>-0.295086</td>\n",
       "      <td>1.447142</td>\n",
       "      <td>-1.220792</td>\n",
       "      <td>2.702571</td>\n",
       "      <td>-1.684210</td>\n",
       "      <td>-0.772874</td>\n",
       "      <td>0.606500</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>1499174585987600384</td>\n",
       "      <td>#Ukraine: Ukrainian forces recovered a Eniks E...</td>\n",
       "      <td>['Ukraine']</td>\n",
       "      <td>en</td>\n",
       "      <td>MAR03</td>\n",
       "      <td>-1.926291</td>\n",
       "      <td>1.267447</td>\n",
       "      <td>0.810379</td>\n",
       "      <td>1.197331</td>\n",
       "      <td>0.001189</td>\n",
       "      <td>0.045890</td>\n",
       "      <td>0.067511</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>1499174586159665155</td>\n",
       "      <td>Volunteers needed for a rapid-response #DH #Di...</td>\n",
       "      <td>['DH', 'DigitalHumanities', 'CulturalHeritage']</td>\n",
       "      <td>en</td>\n",
       "      <td>MAR03</td>\n",
       "      <td>-1.135949</td>\n",
       "      <td>1.682476</td>\n",
       "      <td>-0.762521</td>\n",
       "      <td>0.115792</td>\n",
       "      <td>-0.286950</td>\n",
       "      <td>0.583140</td>\n",
       "      <td>0.516610</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "               tweetid                                               text  \\\n",
       "0  1499174584720969730  Map situation in #Ukraine after the seventh da...   \n",
       "1  1499174584976826368  #Ukraine: Let's just say it's not just the TB-...   \n",
       "2  1499174585073242116  ⚡️The SWIFT company confirmed that it will dis...   \n",
       "3  1499174585987600384  #Ukraine: Ukrainian forces recovered a Eniks E...   \n",
       "4  1499174586159665155  Volunteers needed for a rapid-response #DH #Di...   \n",
       "\n",
       "                                          hashtags language   date  negative  \\\n",
       "0             ['Ukraine', 'RussiaUkraineConflict']       en  MAR03  0.237624   \n",
       "1                                      ['Ukraine']       en  MAR03  0.947360   \n",
       "2                                ['EU', 'Russian']       en  MAR03 -0.295086   \n",
       "3                                      ['Ukraine']       en  MAR03 -1.926291   \n",
       "4  ['DH', 'DigitalHumanities', 'CulturalHeritage']       en  MAR03 -1.135949   \n",
       "\n",
       "    neutral  positive     anger       joy  optimism   sadness  \n",
       "0  1.357651 -1.783151  1.320624 -1.717873 -0.575291  1.707901  \n",
       "1  0.889939 -2.091296  1.598669 -0.535707  0.341185 -0.105763  \n",
       "2  1.447142 -1.220792  2.702571 -1.684210 -0.772874  0.606500  \n",
       "3  1.267447  0.810379  1.197331  0.001189  0.045890  0.067511  \n",
       "4  1.682476 -0.762521  0.115792 -0.286950  0.583140  0.516610  "
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "9ec51d7f",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "9e549038ebd6483eaf5d4280c507f2c1",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/3735462 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "df['hashtags'] = df.hashtags.progress_apply(eval)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "ec0967a3",
   "metadata": {},
   "source": [
    "## Getting popular hashtags"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "ef1fa25f",
   "metadata": {},
   "outputs": [],
   "source": [
    "hashtags = df.hashtags.to_list()\n",
    "hashtags = [j for i in hashtags for j in i]  # Flattening"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "id": "0665b4e5",
   "metadata": {},
   "outputs": [],
   "source": [
    "hashtag_occurances = Counter(hashtags)\n",
    "hashtag_occurances = Counter({k: c for k, c in hashtag_occurances.items() if c >= 500})  # Only keeping popular hashtags"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "id": "47f015ef",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Top 10 hashtags\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "[('Ukraine', 1361248),\n",
       " ('Russia', 652633),\n",
       " ('Putin', 414755),\n",
       " ('StopPutin', 201149),\n",
       " ('SafeAirliftUkraine', 159124),\n",
       " ('Russian', 152752),\n",
       " ('UkraineRussianWar', 147350),\n",
       " ('UkraineRussiaWar', 127101),\n",
       " ('StandWithUkraine', 126355),\n",
       " ('Kyiv', 109992)]"
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "print(\"Top 10 hashtags\")\n",
    "hashtag_occurances.most_common(10)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "id": "79fddc43",
   "metadata": {},
   "outputs": [],
   "source": [
    "popular_hashtags = list(dict(hashtag_occurances.most_common(40)).keys())"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "4807deec",
   "metadata": {},
   "source": [
    "## Transforming the dataframe"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "id": "89ded1f7",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Index(['tweetid', 'text', 'hashtags', 'language', 'date', 'negative',\n",
       "       'neutral', 'positive', 'anger', 'joy', 'optimism', 'sadness'],\n",
       "      dtype='object')"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.columns"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "id": "d19468a3",
   "metadata": {},
   "outputs": [],
   "source": [
    "df = df.explode('hashtags')\n",
    "df = df[['date', 'hashtags', 'negative', 'neutral', 'positive', 'anger', 'joy', 'optimism', 'sadness']]\n",
    "df.rename(columns={\"hashtags\": \"hashtag\"}, inplace=True)\n",
    "df = df[df.hashtag.isin(popular_hashtags)]\n",
    "df.reset_index(drop=True, inplace=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "id": "429548eb",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>date</th>\n",
       "      <th>hashtag</th>\n",
       "      <th>negative</th>\n",
       "      <th>neutral</th>\n",
       "      <th>positive</th>\n",
       "      <th>anger</th>\n",
       "      <th>joy</th>\n",
       "      <th>optimism</th>\n",
       "      <th>sadness</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>MAR03</td>\n",
       "      <td>Ukraine</td>\n",
       "      <td>0.237624</td>\n",
       "      <td>1.357651</td>\n",
       "      <td>-1.783151</td>\n",
       "      <td>1.320624</td>\n",
       "      <td>-1.717873</td>\n",
       "      <td>-0.575291</td>\n",
       "      <td>1.707901</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>MAR03</td>\n",
       "      <td>Ukraine</td>\n",
       "      <td>0.947360</td>\n",
       "      <td>0.889939</td>\n",
       "      <td>-2.091296</td>\n",
       "      <td>1.598669</td>\n",
       "      <td>-0.535707</td>\n",
       "      <td>0.341185</td>\n",
       "      <td>-0.105763</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>MAR03</td>\n",
       "      <td>EU</td>\n",
       "      <td>-0.295086</td>\n",
       "      <td>1.447142</td>\n",
       "      <td>-1.220792</td>\n",
       "      <td>2.702571</td>\n",
       "      <td>-1.684210</td>\n",
       "      <td>-0.772874</td>\n",
       "      <td>0.606500</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>MAR03</td>\n",
       "      <td>Russian</td>\n",
       "      <td>-0.295086</td>\n",
       "      <td>1.447142</td>\n",
       "      <td>-1.220792</td>\n",
       "      <td>2.702571</td>\n",
       "      <td>-1.684210</td>\n",
       "      <td>-0.772874</td>\n",
       "      <td>0.606500</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>MAR03</td>\n",
       "      <td>Ukraine</td>\n",
       "      <td>-1.926291</td>\n",
       "      <td>1.267447</td>\n",
       "      <td>0.810379</td>\n",
       "      <td>1.197331</td>\n",
       "      <td>0.001189</td>\n",
       "      <td>0.045890</td>\n",
       "      <td>0.067511</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4765915</th>\n",
       "      <td>MAR05</td>\n",
       "      <td>NATO</td>\n",
       "      <td>0.844373</td>\n",
       "      <td>0.767706</td>\n",
       "      <td>-1.969419</td>\n",
       "      <td>2.737618</td>\n",
       "      <td>-1.661515</td>\n",
       "      <td>0.054990</td>\n",
       "      <td>-0.094986</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4765916</th>\n",
       "      <td>MAR05</td>\n",
       "      <td>Ukraine</td>\n",
       "      <td>1.888199</td>\n",
       "      <td>0.310074</td>\n",
       "      <td>-2.671582</td>\n",
       "      <td>2.163848</td>\n",
       "      <td>-2.166087</td>\n",
       "      <td>-0.164755</td>\n",
       "      <td>0.844673</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4765917</th>\n",
       "      <td>MAR05</td>\n",
       "      <td>Putin</td>\n",
       "      <td>1.888199</td>\n",
       "      <td>0.310074</td>\n",
       "      <td>-2.671582</td>\n",
       "      <td>2.163848</td>\n",
       "      <td>-2.166087</td>\n",
       "      <td>-0.164755</td>\n",
       "      <td>0.844673</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4765918</th>\n",
       "      <td>MAR05</td>\n",
       "      <td>SafeAirliftUkraine</td>\n",
       "      <td>1.888199</td>\n",
       "      <td>0.310074</td>\n",
       "      <td>-2.671582</td>\n",
       "      <td>2.163848</td>\n",
       "      <td>-2.166087</td>\n",
       "      <td>-0.164755</td>\n",
       "      <td>0.844673</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4765919</th>\n",
       "      <td>MAR05</td>\n",
       "      <td>Ukraine</td>\n",
       "      <td>-0.767866</td>\n",
       "      <td>1.070801</td>\n",
       "      <td>-0.351321</td>\n",
       "      <td>0.298714</td>\n",
       "      <td>-0.168255</td>\n",
       "      <td>-0.014217</td>\n",
       "      <td>0.867072</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>4765920 rows × 9 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "          date             hashtag  negative   neutral  positive     anger  \\\n",
       "0        MAR03             Ukraine  0.237624  1.357651 -1.783151  1.320624   \n",
       "1        MAR03             Ukraine  0.947360  0.889939 -2.091296  1.598669   \n",
       "2        MAR03                  EU -0.295086  1.447142 -1.220792  2.702571   \n",
       "3        MAR03             Russian -0.295086  1.447142 -1.220792  2.702571   \n",
       "4        MAR03             Ukraine -1.926291  1.267447  0.810379  1.197331   \n",
       "...        ...                 ...       ...       ...       ...       ...   \n",
       "4765915  MAR05                NATO  0.844373  0.767706 -1.969419  2.737618   \n",
       "4765916  MAR05             Ukraine  1.888199  0.310074 -2.671582  2.163848   \n",
       "4765917  MAR05               Putin  1.888199  0.310074 -2.671582  2.163848   \n",
       "4765918  MAR05  SafeAirliftUkraine  1.888199  0.310074 -2.671582  2.163848   \n",
       "4765919  MAR05             Ukraine -0.767866  1.070801 -0.351321  0.298714   \n",
       "\n",
       "              joy  optimism   sadness  \n",
       "0       -1.717873 -0.575291  1.707901  \n",
       "1       -0.535707  0.341185 -0.105763  \n",
       "2       -1.684210 -0.772874  0.606500  \n",
       "3       -1.684210 -0.772874  0.606500  \n",
       "4        0.001189  0.045890  0.067511  \n",
       "...           ...       ...       ...  \n",
       "4765915 -1.661515  0.054990 -0.094986  \n",
       "4765916 -2.166087 -0.164755  0.844673  \n",
       "4765917 -2.166087 -0.164755  0.844673  \n",
       "4765918 -2.166087 -0.164755  0.844673  \n",
       "4765919 -0.168255 -0.014217  0.867072  \n",
       "\n",
       "[4765920 rows x 9 columns]"
      ]
     },
     "execution_count": 12,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "id": "4448324d",
   "metadata": {},
   "outputs": [],
   "source": [
    "df = pd.get_dummies(df,columns=['hashtag'],drop_first=True)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "b8d86473",
   "metadata": {},
   "source": [
    "## Getting the top hashtags for each sentiment"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "id": "ffea28f5",
   "metadata": {},
   "outputs": [],
   "source": [
    "sentiment_cols = ['negative', 'neutral', 'positive', 'anger', 'joy', 'optimism', 'sadness']\n",
    "hashtag_cols = [col for col in df if col.startswith('hashtag')]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "id": "d01789ae",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "bd20a15b1c77464ea6b0332ade814e7e",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/7 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "**********************\n",
      "Analyzing negative\n",
      "                                 OLS Regression Results                                \n",
      "=======================================================================================\n",
      "Dep. Variable:               negative   R-squared (uncentered):                   0.273\n",
      "Model:                            OLS   Adj. R-squared (uncentered):              0.273\n",
      "Method:                 Least Squares   F-statistic:                          4.599e+04\n",
      "Date:                Thu, 17 Mar 2022   Prob (F-statistic):                        0.00\n",
      "Time:                        17:59:45   Log-Likelihood:                     -8.0505e+06\n",
      "No. Observations:             4765920   AIC:                                  1.610e+07\n",
      "Df Residuals:                 4765881   BIC:                                  1.610e+07\n",
      "Df Model:                          39                                                  \n",
      "Covariance Type:            nonrobust                                                  \n",
      "===============================================================================================\n",
      "                                  coef    std err          t      P>|t|      [0.025      0.975]\n",
      "-----------------------------------------------------------------------------------------------\n",
      "hashtag_BREAKING                0.7484      0.008     88.370      0.000       0.732       0.765\n",
      "hashtag_China                   0.3194      0.008     37.807      0.000       0.303       0.336\n",
      "hashtag_EU                     -0.0326      0.008     -4.204      0.000      -0.048      -0.017\n",
      "hashtag_Europe                  0.6621      0.008     79.815      0.000       0.646       0.678\n",
      "hashtag_Kharkiv                 1.0138      0.005    192.564      0.000       1.004       1.024\n",
      "hashtag_Kyiv                    0.4079      0.004    103.241      0.000       0.400       0.416\n",
      "hashtag_Mariupol                1.8757      0.006    329.385      0.000       1.865       1.887\n",
      "hashtag_NATO                    0.5315      0.004    120.871      0.000       0.523       0.540\n",
      "hashtag_Putin                   1.2325      0.002    605.767      0.000       1.228       1.236\n",
      "hashtag_Russia                  0.7925      0.002    488.591      0.000       0.789       0.796\n",
      "hashtag_RussiaUkraine           0.6399      0.005    119.181      0.000       0.629       0.650\n",
      "hashtag_RussiaUkraineWar        0.8111      0.006    147.225      0.000       0.800       0.822\n",
      "hashtag_Russian                 1.0711      0.003    319.485      0.000       1.065       1.078\n",
      "hashtag_RussianUkrainianWar     0.8406      0.005    181.614      0.000       0.832       0.850\n",
      "hashtag_SafeAirliftUkraine      1.4387      0.003    437.985      0.000       1.432       1.445\n",
      "hashtag_SlavaUkraini           -0.4402      0.008    -53.028      0.000      -0.456      -0.424\n",
      "hashtag_StandWithUkraine       -0.1311      0.004    -35.570      0.000      -0.138      -0.124\n",
      "hashtag_StandWithUkraine️       0.0111      0.008      1.404      0.160      -0.004       0.027\n",
      "hashtag_StopPutin               1.2841      0.003    439.519      0.000       1.278       1.290\n",
      "hashtag_StopPutinNOW            0.8832      0.006    144.843      0.000       0.871       0.895\n",
      "hashtag_StopRussia              0.8979      0.005    175.606      0.000       0.888       0.908\n",
      "hashtag_UKRAINE                 0.8833      0.007    131.262      0.000       0.870       0.896\n",
      "hashtag_US                      0.7462      0.008     90.050      0.000       0.730       0.762\n",
      "hashtag_Ukraine                 0.3542      0.001    315.408      0.000       0.352       0.356\n",
      "hashtag_UkraineRussia           0.7094      0.008     87.048      0.000       0.693       0.725\n",
      "hashtag_UkraineRussiaWar        0.7322      0.004    199.219      0.000       0.725       0.739\n",
      "hashtag_UkraineRussianWar       0.9197      0.003    269.428      0.000       0.913       0.926\n",
      "hashtag_UkraineUnderAttack      1.2719      0.005    255.278      0.000       1.262       1.282\n",
      "hashtag_UkraineUnderAttaсk      0.7062      0.008     91.001      0.000       0.691       0.721\n",
      "hashtag_UkraineWar              0.7553      0.005    137.999      0.000       0.745       0.766\n",
      "hashtag_Ukraine️                0.2255      0.006     39.793      0.000       0.214       0.237\n",
      "hashtag_Ukrainian               0.5795      0.006    104.209      0.000       0.569       0.590\n",
      "hashtag_Ukriane                 0.4857      0.008     60.025      0.000       0.470       0.502\n",
      "hashtag_Zelensky                0.2215      0.007     30.521      0.000       0.207       0.236\n",
      "hashtag_Zelenskyy              -0.1206      0.009    -13.959      0.000      -0.137      -0.104\n",
      "hashtag_putin                   0.9803      0.008    120.548      0.000       0.964       0.996\n",
      "hashtag_russia                  0.6200      0.007     90.531      0.000       0.607       0.633\n",
      "hashtag_ukraine                 0.1634      0.005     32.385      0.000       0.153       0.173\n",
      "hashtag_war                     0.7532      0.007    106.978      0.000       0.739       0.767\n",
      "==============================================================================\n",
      "Omnibus:                   300755.637   Durbin-Watson:                   1.401\n",
      "Prob(Omnibus):                  0.000   Jarque-Bera (JB):           360999.594\n",
      "Skew:                          -0.669   Prob(JB):                         0.00\n",
      "Kurtosis:                       3.168   Cond. No.                         7.69\n",
      "==============================================================================\n",
      "\n",
      "Notes:\n",
      "[1] R² is computed without centering (uncentered) since the model does not contain a constant.\n",
      "[2] Standard Errors assume that the covariance matrix of the errors is correctly specified.\n",
      "**********************\n",
      "Analyzing neutral\n",
      "                                 OLS Regression Results                                \n",
      "=======================================================================================\n",
      "Dep. Variable:                neutral   R-squared (uncentered):                   0.686\n",
      "Model:                            OLS   Adj. R-squared (uncentered):              0.686\n",
      "Method:                 Least Squares   F-statistic:                          2.676e+05\n",
      "Date:                Thu, 17 Mar 2022   Prob (F-statistic):                        0.00\n",
      "Time:                        17:59:56   Log-Likelihood:                     -2.9141e+06\n",
      "No. Observations:             4765920   AIC:                                  5.828e+06\n",
      "Df Residuals:                 4765881   BIC:                                  5.829e+06\n",
      "Df Model:                          39                                                  \n",
      "Covariance Type:            nonrobust                                                  \n",
      "===============================================================================================\n",
      "                                  coef    std err          t      P>|t|      [0.025      0.975]\n",
      "-----------------------------------------------------------------------------------------------\n",
      "hashtag_BREAKING                0.8850      0.003    306.996      0.000       0.879       0.891\n",
      "hashtag_China                   0.8135      0.003    282.861      0.000       0.808       0.819\n",
      "hashtag_EU                      0.8149      0.003    308.569      0.000       0.810       0.820\n",
      "hashtag_Europe                  0.5888      0.003    208.561      0.000       0.583       0.594\n",
      "hashtag_Kharkiv                 0.7144      0.002    398.671      0.000       0.711       0.718\n",
      "hashtag_Kyiv                    0.7681      0.001    571.230      0.000       0.766       0.771\n",
      "hashtag_Mariupol                0.3269      0.002    168.668      0.000       0.323       0.331\n",
      "hashtag_NATO                    0.7813      0.001    521.948      0.000       0.778       0.784\n",
      "hashtag_Putin                   0.4627      0.001    668.126      0.000       0.461       0.464\n",
      "hashtag_Russia                  0.7449      0.001   1349.359      0.000       0.744       0.746\n",
      "hashtag_RussiaUkraine           0.7014      0.002    383.836      0.000       0.698       0.705\n",
      "hashtag_RussiaUkraineWar        0.7074      0.002    377.247      0.000       0.704       0.711\n",
      "hashtag_Russian                 0.6977      0.001    611.426      0.000       0.695       0.700\n",
      "hashtag_RussianUkrainianWar     0.6911      0.002    438.657      0.000       0.688       0.694\n",
      "hashtag_SafeAirliftUkraine      0.3627      0.001    324.426      0.000       0.361       0.365\n",
      "hashtag_SlavaUkraini            0.6567      0.003    232.432      0.000       0.651       0.662\n",
      "hashtag_StandWithUkraine        0.6446      0.001    513.758      0.000       0.642       0.647\n",
      "hashtag_StandWithUkraine️       0.6324      0.003    235.128      0.000       0.627       0.638\n",
      "hashtag_StopPutin               0.4163      0.001    418.626      0.000       0.414       0.418\n",
      "hashtag_StopPutinNOW            0.5005      0.002    241.157      0.000       0.496       0.505\n",
      "hashtag_StopRussia              0.5312      0.002    305.202      0.000       0.528       0.535\n",
      "hashtag_UKRAINE                 0.5924      0.002    258.658      0.000       0.588       0.597\n",
      "hashtag_US                      0.8032      0.003    284.782      0.000       0.798       0.809\n",
      "hashtag_Ukraine                 0.7063      0.000   1847.751      0.000       0.706       0.707\n",
      "hashtag_UkraineRussia           0.7623      0.003    274.783      0.000       0.757       0.768\n",
      "hashtag_UkraineRussiaWar        0.6717      0.001    536.989      0.000       0.669       0.674\n",
      "hashtag_UkraineRussianWar       0.6583      0.001    566.609      0.000       0.656       0.661\n",
      "hashtag_UkraineUnderAttack      0.5988      0.002    353.110      0.000       0.595       0.602\n",
      "hashtag_UkraineUnderAttaсk      0.6567      0.003    248.625      0.000       0.652       0.662\n",
      "hashtag_UkraineWar              0.6503      0.002    349.087      0.000       0.647       0.654\n",
      "hashtag_Ukraine️                0.6697      0.002    347.143      0.000       0.666       0.673\n",
      "hashtag_Ukrainian               0.7333      0.002    387.387      0.000       0.730       0.737\n",
      "hashtag_Ukriane                 0.6795      0.003    246.732      0.000       0.674       0.685\n",
      "hashtag_Zelensky                0.7594      0.002    307.438      0.000       0.755       0.764\n",
      "hashtag_Zelenskyy               0.6614      0.003    224.994      0.000       0.656       0.667\n",
      "hashtag_putin                   0.5218      0.003    188.537      0.000       0.516       0.527\n",
      "hashtag_russia                  0.7025      0.002    301.402      0.000       0.698       0.707\n",
      "hashtag_ukraine                 0.7314      0.002    425.936      0.000       0.728       0.735\n",
      "hashtag_war                     0.6275      0.002    261.827      0.000       0.623       0.632\n",
      "==============================================================================\n",
      "Omnibus:                    82172.490   Durbin-Watson:                   1.377\n",
      "Prob(Omnibus):                  0.000   Jarque-Bera (JB):            97791.345\n",
      "Skew:                           0.266   Prob(JB):                         0.00\n",
      "Kurtosis:                       3.458   Cond. No.                         7.69\n",
      "==============================================================================\n",
      "\n",
      "Notes:\n",
      "[1] R² is computed without centering (uncentered) since the model does not contain a constant.\n",
      "[2] Standard Errors assume that the covariance matrix of the errors is correctly specified.\n",
      "**********************\n",
      "Analyzing positive\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "                                 OLS Regression Results                                \n",
      "=======================================================================================\n",
      "Dep. Variable:               positive   R-squared (uncentered):                   0.613\n",
      "Model:                            OLS   Adj. R-squared (uncentered):              0.613\n",
      "Method:                 Least Squares   F-statistic:                          1.938e+05\n",
      "Date:                Thu, 17 Mar 2022   Prob (F-statistic):                        0.00\n",
      "Time:                        18:00:08   Log-Likelihood:                     -7.9437e+06\n",
      "No. Observations:             4765920   AIC:                                  1.589e+07\n",
      "Df Residuals:                 4765881   BIC:                                  1.589e+07\n",
      "Df Model:                          39                                                  \n",
      "Covariance Type:            nonrobust                                                  \n",
      "===============================================================================================\n",
      "                                  coef    std err          t      P>|t|      [0.025      0.975]\n",
      "-----------------------------------------------------------------------------------------------\n",
      "hashtag_BREAKING               -1.8653      0.008   -225.235      0.000      -1.882      -1.849\n",
      "hashtag_China                  -1.3335      0.008   -161.407      0.000      -1.350      -1.317\n",
      "hashtag_EU                     -0.9627      0.008   -126.887      0.000      -0.978      -0.948\n",
      "hashtag_Europe                 -1.5114      0.008   -186.331      0.000      -1.527      -1.495\n",
      "hashtag_Kharkiv                -1.9305      0.005   -374.988      0.000      -1.941      -1.920\n",
      "hashtag_Kyiv                   -1.3466      0.004   -348.565      0.000      -1.354      -1.339\n",
      "hashtag_Mariupol               -2.4674      0.006   -443.100      0.000      -2.478      -2.456\n",
      "hashtag_NATO                   -1.5655      0.004   -364.056      0.000      -1.574      -1.557\n",
      "hashtag_Putin                  -2.0177      0.002  -1014.184      0.000      -2.022      -2.014\n",
      "hashtag_Russia                 -1.7562      0.002  -1107.337      0.000      -1.759      -1.753\n",
      "hashtag_RussiaUkraine          -1.5293      0.005   -291.297      0.000      -1.540      -1.519\n",
      "hashtag_RussiaUkraineWar       -1.7274      0.005   -320.653      0.000      -1.738      -1.717\n",
      "hashtag_Russian                -2.0152      0.003   -614.735      0.000      -2.022      -2.009\n",
      "hashtag_RussianUkrainianWar    -1.7610      0.005   -389.088      0.000      -1.770      -1.752\n",
      "hashtag_SafeAirliftUkraine     -2.2332      0.003   -695.277      0.000      -2.239      -2.227\n",
      "hashtag_SlavaUkraini           -0.3249      0.008    -40.033      0.000      -0.341      -0.309\n",
      "hashtag_StandWithUkraine       -0.7176      0.004   -199.092      0.000      -0.725      -0.711\n",
      "hashtag_StandWithUkraine️      -0.8575      0.008   -110.972      0.000      -0.873      -0.842\n",
      "hashtag_StopPutin              -2.1046      0.003   -736.695      0.000      -2.110      -2.099\n",
      "hashtag_StopPutinNOW           -1.7062      0.006   -286.148      0.000      -1.718      -1.695\n",
      "hashtag_StopRussia             -1.7618      0.005   -352.382      0.000      -1.772      -1.752\n",
      "hashtag_UKRAINE                -1.7860      0.007   -271.428      0.000      -1.799      -1.773\n",
      "hashtag_US                     -1.7972      0.008   -221.798      0.000      -1.813      -1.781\n",
      "hashtag_Ukraine                -1.2263      0.001  -1116.673      0.000      -1.228      -1.224\n",
      "hashtag_UkraineRussia          -1.6602      0.008   -208.326      0.000      -1.676      -1.645\n",
      "hashtag_UkraineRussiaWar       -1.5977      0.004   -444.566      0.000      -1.605      -1.591\n",
      "hashtag_UkraineRussianWar      -1.7795      0.003   -533.155      0.000      -1.786      -1.773\n",
      "hashtag_UkraineUnderAttack     -2.0880      0.005   -428.582      0.000      -2.098      -2.078\n",
      "hashtag_UkraineUnderAttaсk     -1.5604      0.008   -205.636      0.000      -1.575      -1.546\n",
      "hashtag_UkraineWar             -1.6058      0.005   -300.027      0.000      -1.616      -1.595\n",
      "hashtag_Ukraine️               -1.0500      0.006   -189.452      0.000      -1.061      -1.039\n",
      "hashtag_Ukrainian              -1.5074      0.005   -277.185      0.000      -1.518      -1.497\n",
      "hashtag_Ukriane                -1.3575      0.008   -171.572      0.000      -1.373      -1.342\n",
      "hashtag_Zelensky               -1.1637      0.007   -163.991      0.000      -1.178      -1.150\n",
      "hashtag_Zelenskyy              -0.6825      0.008    -80.813      0.000      -0.699      -0.666\n",
      "hashtag_putin                  -1.7193      0.008   -216.219      0.000      -1.735      -1.704\n",
      "hashtag_russia                 -1.5060      0.007   -224.894      0.000      -1.519      -1.493\n",
      "hashtag_ukraine                -1.0341      0.005   -209.630      0.000      -1.044      -1.024\n",
      "hashtag_war                    -1.5950      0.007   -231.669      0.000      -1.608      -1.582\n",
      "==============================================================================\n",
      "Omnibus:                  1181757.281   Durbin-Watson:                   1.466\n",
      "Prob(Omnibus):                  0.000   Jarque-Bera (JB):          2675748.633\n",
      "Skew:                           1.418   Prob(JB):                         0.00\n",
      "Kurtosis:                       5.330   Cond. No.                         7.69\n",
      "==============================================================================\n",
      "\n",
      "Notes:\n",
      "[1] R² is computed without centering (uncentered) since the model does not contain a constant.\n",
      "[2] Standard Errors assume that the covariance matrix of the errors is correctly specified.\n",
      "**********************\n",
      "Analyzing anger\n",
      "                                 OLS Regression Results                                \n",
      "=======================================================================================\n",
      "Dep. Variable:                  anger   R-squared (uncentered):                   0.650\n",
      "Model:                            OLS   Adj. R-squared (uncentered):              0.650\n",
      "Method:                 Least Squares   F-statistic:                          2.273e+05\n",
      "Date:                Thu, 17 Mar 2022   Prob (F-statistic):                        0.00\n",
      "Time:                        18:00:19   Log-Likelihood:                     -7.6418e+06\n",
      "No. Observations:             4765920   AIC:                                  1.528e+07\n",
      "Df Residuals:                 4765881   BIC:                                  1.528e+07\n",
      "Df Model:                          39                                                  \n",
      "Covariance Type:            nonrobust                                                  \n",
      "===============================================================================================\n",
      "                                  coef    std err          t      P>|t|      [0.025      0.975]\n",
      "-----------------------------------------------------------------------------------------------\n",
      "hashtag_BREAKING                1.5721      0.008    202.248      0.000       1.557       1.587\n",
      "hashtag_China                   1.5436      0.008    199.052      0.000       1.528       1.559\n",
      "hashtag_EU                      1.4060      0.007    197.423      0.000       1.392       1.420\n",
      "hashtag_Europe                  1.9806      0.008    260.145      0.000       1.966       1.996\n",
      "hashtag_Kharkiv                 1.4726      0.005    304.750      0.000       1.463       1.482\n",
      "hashtag_Kyiv                    1.1998      0.004    330.883      0.000       1.193       1.207\n",
      "hashtag_Mariupol                1.7101      0.005    327.182      0.000       1.700       1.720\n",
      "hashtag_NATO                    1.7553      0.004    434.878      0.000       1.747       1.763\n",
      "hashtag_Putin                   1.9560      0.002   1047.454      0.000       1.952       1.960\n",
      "hashtag_Russia                  1.7047      0.001   1145.165      0.000       1.702       1.708\n",
      "hashtag_RussiaUkraine           1.7495      0.005    355.047      0.000       1.740       1.759\n",
      "hashtag_RussiaUkraineWar        1.9069      0.005    377.121      0.000       1.897       1.917\n",
      "hashtag_Russian                 1.8572      0.003    603.562      0.000       1.851       1.863\n",
      "hashtag_RussianUkrainianWar     2.0743      0.004    488.274      0.000       2.066       2.083\n",
      "hashtag_SafeAirliftUkraine      1.8346      0.003    608.536      0.000       1.829       1.841\n",
      "hashtag_SlavaUkraini            0.9097      0.008    119.407      0.000       0.895       0.925\n",
      "hashtag_StandWithUkraine        1.1984      0.003    354.221      0.000       1.192       1.205\n",
      "hashtag_StandWithUkraine️       1.4167      0.007    195.326      0.000       1.402       1.431\n",
      "hashtag_StopPutin               2.1430      0.003    799.220      0.000       2.138       2.148\n",
      "hashtag_StopPutinNOW            2.0351      0.006    363.620      0.000       2.024       2.046\n",
      "hashtag_StopRussia              2.7688      0.005    589.996      0.000       2.760       2.778\n",
      "hashtag_UKRAINE                 2.9734      0.006    481.426      0.000       2.961       2.986\n",
      "hashtag_US                      1.8512      0.008    243.394      0.000       1.836       1.866\n",
      "hashtag_Ukraine                 1.1913      0.001   1155.800      0.000       1.189       1.193\n",
      "hashtag_UkraineRussia           1.7516      0.007    234.166      0.000       1.737       1.766\n",
      "hashtag_UkraineRussiaWar        1.8181      0.003    538.984      0.000       1.812       1.825\n",
      "hashtag_UkraineRussianWar       1.7497      0.003    558.489      0.000       1.744       1.756\n",
      "hashtag_UkraineUnderAttack      1.8491      0.005    404.357      0.000       1.840       1.858\n",
      "hashtag_UkraineUnderAttaсk      1.9660      0.007    276.022      0.000       1.952       1.980\n",
      "hashtag_UkraineWar              1.9243      0.005    383.047      0.000       1.914       1.934\n",
      "hashtag_Ukraine️                1.3054      0.005    250.937      0.000       1.295       1.316\n",
      "hashtag_Ukrainian               1.4313      0.005    280.407      0.000       1.421       1.441\n",
      "hashtag_Ukriane                 1.6190      0.007    218.012      0.000       1.604       1.634\n",
      "hashtag_Zelensky                1.3530      0.007    203.137      0.000       1.340       1.366\n",
      "hashtag_Zelenskyy               1.0523      0.008    132.746      0.000       1.037       1.068\n",
      "hashtag_putin                   2.0805      0.007    278.760      0.000       2.066       2.095\n",
      "hashtag_russia                  1.7907      0.006    284.905      0.000       1.778       1.803\n",
      "hashtag_ukraine                 1.2454      0.005    268.962      0.000       1.236       1.254\n",
      "hashtag_war                     1.8735      0.006    289.908      0.000       1.861       1.886\n",
      "==============================================================================\n",
      "Omnibus:                   131707.005   Durbin-Watson:                   1.378\n",
      "Prob(Omnibus):                  0.000   Jarque-Bera (JB):           116216.400\n",
      "Skew:                          -0.325   Prob(JB):                         0.00\n",
      "Kurtosis:                       2.596   Cond. No.                         7.69\n",
      "==============================================================================\n",
      "\n",
      "Notes:\n",
      "[1] R² is computed without centering (uncentered) since the model does not contain a constant.\n",
      "[2] Standard Errors assume that the covariance matrix of the errors is correctly specified.\n",
      "**********************\n",
      "Analyzing joy\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "                                 OLS Regression Results                                \n",
      "=======================================================================================\n",
      "Dep. Variable:                    joy   R-squared (uncentered):                   0.459\n",
      "Model:                            OLS   Adj. R-squared (uncentered):              0.459\n",
      "Method:                 Least Squares   F-statistic:                          1.038e+05\n",
      "Date:                Thu, 17 Mar 2022   Prob (F-statistic):                        0.00\n",
      "Time:                        18:00:30   Log-Likelihood:                     -7.3337e+06\n",
      "No. Observations:             4765920   AIC:                                  1.467e+07\n",
      "Df Residuals:                 4765881   BIC:                                  1.467e+07\n",
      "Df Model:                          39                                                  \n",
      "Covariance Type:            nonrobust                                                  \n",
      "===============================================================================================\n",
      "                                  coef    std err          t      P>|t|      [0.025      0.975]\n",
      "-----------------------------------------------------------------------------------------------\n",
      "hashtag_BREAKING               -0.9189      0.007   -126.112      0.000      -0.933      -0.905\n",
      "hashtag_China                  -1.0549      0.007   -145.108      0.000      -1.069      -1.041\n",
      "hashtag_EU                     -0.7825      0.007   -117.220      0.000      -0.796      -0.769\n",
      "hashtag_Europe                 -1.3419      0.007   -188.018      0.000      -1.356      -1.328\n",
      "hashtag_Kharkiv                -0.8119      0.005   -179.240      0.000      -0.821      -0.803\n",
      "hashtag_Kyiv                   -0.4968      0.003   -146.164      0.000      -0.503      -0.490\n",
      "hashtag_Mariupol               -1.3539      0.005   -276.336      0.000      -1.364      -1.344\n",
      "hashtag_NATO                   -0.9569      0.004   -252.916      0.000      -0.964      -0.950\n",
      "hashtag_Putin                  -1.4397      0.002   -822.459      0.000      -1.443      -1.436\n",
      "hashtag_Russia                 -0.9626      0.001   -689.818      0.000      -0.965      -0.960\n",
      "hashtag_RussiaUkraine          -0.7177      0.005   -155.383      0.000      -0.727      -0.709\n",
      "hashtag_RussiaUkraineWar       -0.9064      0.005   -191.237      0.000      -0.916      -0.897\n",
      "hashtag_Russian                -1.1126      0.003   -385.735      0.000      -1.118      -1.107\n",
      "hashtag_RussianUkrainianWar    -0.9390      0.004   -235.805      0.000      -0.947      -0.931\n",
      "hashtag_SafeAirliftUkraine     -2.1444      0.003   -758.796      0.000      -2.150      -2.139\n",
      "hashtag_SlavaUkraini            0.0562      0.007      7.864      0.000       0.042       0.070\n",
      "hashtag_StandWithUkraine       -0.7165      0.003   -225.910      0.000      -0.723      -0.710\n",
      "hashtag_StandWithUkraine️      -0.7680      0.007   -112.962      0.000      -0.781      -0.755\n",
      "hashtag_StopPutin              -1.9977      0.003   -794.757      0.000      -2.003      -1.993\n",
      "hashtag_StopPutinNOW           -1.3835      0.005   -263.711      0.000      -1.394      -1.373\n",
      "hashtag_StopRussia             -1.5875      0.004   -360.868      0.000      -1.596      -1.579\n",
      "hashtag_UKRAINE                -1.4554      0.006   -251.377      0.000      -1.467      -1.444\n",
      "hashtag_US                     -1.2322      0.007   -172.834      0.000      -1.246      -1.218\n",
      "hashtag_Ukraine                -0.6169      0.001   -638.468      0.000      -0.619      -0.615\n",
      "hashtag_UkraineRussia          -0.7026      0.007   -100.200      0.000      -0.716      -0.689\n",
      "hashtag_UkraineRussiaWar       -0.7535      0.003   -238.285      0.000      -0.760      -0.747\n",
      "hashtag_UkraineRussianWar      -0.8643      0.003   -294.294      0.000      -0.870      -0.859\n",
      "hashtag_UkraineUnderAttack     -1.0344      0.004   -241.315      0.000      -1.043      -1.026\n",
      "hashtag_UkraineUnderAttaсk     -0.8188      0.007   -122.632      0.000      -0.832      -0.806\n",
      "hashtag_UkraineWar             -0.8551      0.005   -181.586      0.000      -0.864      -0.846\n",
      "hashtag_Ukraine️               -0.5408      0.005   -110.908      0.000      -0.550      -0.531\n",
      "hashtag_Ukrainian              -0.6728      0.005   -140.618      0.000      -0.682      -0.663\n",
      "hashtag_Ukriane                -0.6847      0.007    -98.355      0.000      -0.698      -0.671\n",
      "hashtag_Zelensky               -0.2766      0.006    -44.304      0.000      -0.289      -0.264\n",
      "hashtag_Zelenskyy               0.0171      0.007      2.304      0.021       0.003       0.032\n",
      "hashtag_putin                  -0.8904      0.007   -127.269      0.000      -0.904      -0.877\n",
      "hashtag_russia                 -0.6801      0.006   -115.430      0.000      -0.692      -0.669\n",
      "hashtag_ukraine                -0.5613      0.004   -129.324      0.000      -0.570      -0.553\n",
      "hashtag_war                    -1.0455      0.006   -172.597      0.000      -1.057      -1.034\n",
      "==============================================================================\n",
      "Omnibus:                   977010.901   Durbin-Watson:                   1.418\n",
      "Prob(Omnibus):                  0.000   Jarque-Bera (JB):          1807691.298\n",
      "Skew:                           1.290   Prob(JB):                         0.00\n",
      "Kurtosis:                       4.565   Cond. No.                         7.69\n",
      "==============================================================================\n",
      "\n",
      "Notes:\n",
      "[1] R² is computed without centering (uncentered) since the model does not contain a constant.\n",
      "[2] Standard Errors assume that the covariance matrix of the errors is correctly specified.\n",
      "**********************\n",
      "Analyzing optimism\n",
      "                                 OLS Regression Results                                \n",
      "=======================================================================================\n",
      "Dep. Variable:               optimism   R-squared (uncentered):                   0.090\n",
      "Model:                            OLS   Adj. R-squared (uncentered):              0.090\n",
      "Method:                 Least Squares   F-statistic:                          1.209e+04\n",
      "Date:                Thu, 17 Mar 2022   Prob (F-statistic):                        0.00\n",
      "Time:                        18:00:41   Log-Likelihood:                     -5.6887e+06\n",
      "No. Observations:             4765920   AIC:                                  1.138e+07\n",
      "Df Residuals:                 4765881   BIC:                                  1.138e+07\n",
      "Df Model:                          39                                                  \n",
      "Covariance Type:            nonrobust                                                  \n",
      "===============================================================================================\n",
      "                                  coef    std err          t      P>|t|      [0.025      0.975]\n",
      "-----------------------------------------------------------------------------------------------\n",
      "hashtag_BREAKING               -0.5610      0.005   -108.723      0.000      -0.571      -0.551\n",
      "hashtag_China                   0.4182      0.005     81.241      0.000       0.408       0.428\n",
      "hashtag_EU                      0.2994      0.005     63.345      0.000       0.290       0.309\n",
      "hashtag_Europe                  0.0032      0.005      0.643      0.520      -0.007       0.013\n",
      "hashtag_Kharkiv                -0.4168      0.003   -129.955      0.000      -0.423      -0.411\n",
      "hashtag_Kyiv                   -0.1311      0.002    -54.475      0.000      -0.136      -0.126\n",
      "hashtag_Mariupol               -0.7178      0.003   -206.904      0.000      -0.725      -0.711\n",
      "hashtag_NATO                    0.1687      0.003     62.958      0.000       0.163       0.174\n",
      "hashtag_Putin                   0.0183      0.001     14.751      0.000       0.016       0.021\n",
      "hashtag_Russia                 -0.1988      0.001   -201.140      0.000      -0.201      -0.197\n",
      "hashtag_RussiaUkraine          -0.2209      0.003    -67.537      0.000      -0.227      -0.214\n",
      "hashtag_RussiaUkraineWar       -0.3078      0.003    -91.700      0.000      -0.314      -0.301\n",
      "hashtag_Russian                -0.3712      0.002   -181.751      0.000      -0.375      -0.367\n",
      "hashtag_RussianUkrainianWar    -0.2685      0.003    -95.226      0.000      -0.274      -0.263\n",
      "hashtag_SafeAirliftUkraine      0.2433      0.002    121.576      0.000       0.239       0.247\n",
      "hashtag_SlavaUkraini            0.2150      0.005     42.510      0.000       0.205       0.225\n",
      "hashtag_StandWithUkraine        0.6170      0.002    274.756      0.000       0.613       0.621\n",
      "hashtag_StandWithUkraine️       0.5015      0.005    104.169      0.000       0.492       0.511\n",
      "hashtag_StopPutin               0.1298      0.002     72.911      0.000       0.126       0.133\n",
      "hashtag_StopPutinNOW            0.1482      0.004     39.899      0.000       0.141       0.156\n",
      "hashtag_StopRussia             -0.1408      0.003    -45.205      0.000      -0.147      -0.135\n",
      "hashtag_UKRAINE                -0.3727      0.004    -90.910      0.000      -0.381      -0.365\n",
      "hashtag_US                      0.0106      0.005      2.097      0.036       0.001       0.020\n",
      "hashtag_Ukraine                 0.0257      0.001     37.520      0.000       0.024       0.027\n",
      "hashtag_UkraineRussia          -0.4220      0.005    -84.999      0.000      -0.432      -0.412\n",
      "hashtag_UkraineRussiaWar       -0.2971      0.002   -132.688      0.000      -0.301      -0.293\n",
      "hashtag_UkraineRussianWar      -0.4330      0.002   -208.237      0.000      -0.437      -0.429\n",
      "hashtag_UkraineUnderAttack     -0.7778      0.003   -256.235      0.000      -0.784      -0.772\n",
      "hashtag_UkraineUnderAttaсk     -0.2782      0.005    -58.848      0.000      -0.287      -0.269\n",
      "hashtag_UkraineWar             -0.3272      0.003    -98.135      0.000      -0.334      -0.321\n",
      "hashtag_Ukraine️                0.0183      0.003      5.309      0.000       0.012       0.025\n",
      "hashtag_Ukrainian              -0.1743      0.003    -51.454      0.000      -0.181      -0.168\n",
      "hashtag_Ukriane                -0.0463      0.005     -9.391      0.000      -0.056      -0.037\n",
      "hashtag_Zelensky                0.1022      0.004     23.106      0.000       0.093       0.111\n",
      "hashtag_Zelenskyy               0.1406      0.005     26.716      0.000       0.130       0.151\n",
      "hashtag_putin                  -0.2933      0.005    -59.208      0.000      -0.303      -0.284\n",
      "hashtag_russia                 -0.3203      0.004    -76.764      0.000      -0.328      -0.312\n",
      "hashtag_ukraine                -0.1898      0.003    -61.740      0.000      -0.196      -0.184\n",
      "hashtag_war                    -0.1357      0.004    -31.635      0.000      -0.144      -0.127\n",
      "==============================================================================\n",
      "Omnibus:                   385112.926   Durbin-Watson:                   1.322\n",
      "Prob(Omnibus):                  0.000   Jarque-Bera (JB):           486568.541\n",
      "Skew:                           0.749   Prob(JB):                         0.00\n",
      "Kurtosis:                       3.452   Cond. No.                         7.69\n",
      "==============================================================================\n",
      "\n",
      "Notes:\n",
      "[1] R² is computed without centering (uncentered) since the model does not contain a constant.\n",
      "[2] Standard Errors assume that the covariance matrix of the errors is correctly specified.\n",
      "**********************\n",
      "Analyzing sadness\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "                                 OLS Regression Results                                \n",
      "=======================================================================================\n",
      "Dep. Variable:                sadness   R-squared (uncentered):                   0.122\n",
      "Model:                            OLS   Adj. R-squared (uncentered):              0.122\n",
      "Method:                 Least Squares   F-statistic:                          1.691e+04\n",
      "Date:                Thu, 17 Mar 2022   Prob (F-statistic):                        0.00\n",
      "Time:                        18:00:52   Log-Likelihood:                     -6.1267e+06\n",
      "No. Observations:             4765920   AIC:                                  1.225e+07\n",
      "Df Residuals:                 4765881   BIC:                                  1.225e+07\n",
      "Df Model:                          39                                                  \n",
      "Covariance Type:            nonrobust                                                  \n",
      "===============================================================================================\n",
      "                                  coef    std err          t      P>|t|      [0.025      0.975]\n",
      "-----------------------------------------------------------------------------------------------\n",
      "hashtag_BREAKING                0.5920      0.006    104.664      0.000       0.581       0.603\n",
      "hashtag_China                   0.0208      0.006      3.686      0.000       0.010       0.032\n",
      "hashtag_EU                     -0.0717      0.005    -13.835      0.000      -0.082      -0.062\n",
      "hashtag_Europe                  0.1861      0.006     33.583      0.000       0.175       0.197\n",
      "hashtag_Kharkiv                 0.5038      0.004    143.265      0.000       0.497       0.511\n",
      "hashtag_Kyiv                    0.2368      0.003     89.734      0.000       0.232       0.242\n",
      "hashtag_Mariupol                1.0005      0.004    263.054      0.000       0.993       1.008\n",
      "hashtag_NATO                   -0.0334      0.003    -11.376      0.000      -0.039      -0.028\n",
      "hashtag_Putin                   0.3058      0.001    225.062      0.000       0.303       0.308\n",
      "hashtag_Russia                  0.2775      0.001    256.156      0.000       0.275       0.280\n",
      "hashtag_RussiaUkraine           0.0781      0.004     21.778      0.000       0.071       0.085\n",
      "hashtag_RussiaUkraineWar        0.1813      0.004     49.265      0.000       0.174       0.188\n",
      "hashtag_Russian                 0.4397      0.002    196.362      0.000       0.435       0.444\n",
      "hashtag_RussianUkrainianWar     0.0609      0.003     19.708      0.000       0.055       0.067\n",
      "hashtag_SafeAirliftUkraine      0.7927      0.002    361.342      0.000       0.788       0.797\n",
      "hashtag_SlavaUkraini           -0.3523      0.006    -63.555      0.000      -0.363      -0.341\n",
      "hashtag_StandWithUkraine       -0.2331      0.002    -94.695      0.000      -0.238      -0.228\n",
      "hashtag_StandWithUkraine️      -0.2323      0.005    -44.007      0.000      -0.243      -0.222\n",
      "hashtag_StopPutin               0.4679      0.002    239.786      0.000       0.464       0.472\n",
      "hashtag_StopPutinNOW            0.0589      0.004     14.463      0.000       0.051       0.067\n",
      "hashtag_StopRussia             -0.2506      0.003    -73.394      0.000      -0.257      -0.244\n",
      "hashtag_UKRAINE                -0.4153      0.004    -92.397      0.000      -0.424      -0.406\n",
      "hashtag_US                      0.2258      0.006     40.803      0.000       0.215       0.237\n",
      "hashtag_Ukraine                 0.2013      0.001    268.442      0.000       0.200       0.203\n",
      "hashtag_UkraineRussia           0.2205      0.005     40.510      0.000       0.210       0.231\n",
      "hashtag_UkraineRussiaWar        0.1143      0.002     46.551      0.000       0.109       0.119\n",
      "hashtag_UkraineRussianWar       0.3987      0.002    174.910      0.000       0.394       0.403\n",
      "hashtag_UkraineUnderAttack      0.7452      0.003    223.938      0.000       0.739       0.752\n",
      "hashtag_UkraineUnderAttaсk      0.0363      0.005      7.010      0.000       0.026       0.046\n",
      "hashtag_UkraineWar              0.1127      0.004     30.838      0.000       0.106       0.120\n",
      "hashtag_Ukraine️                0.0558      0.004     14.737      0.000       0.048       0.063\n",
      "hashtag_Ukrainian               0.2217      0.004     59.696      0.000       0.214       0.229\n",
      "hashtag_Ukriane                 0.0026      0.005      0.478      0.633      -0.008       0.013\n",
      "hashtag_Zelensky               -0.2072      0.005    -42.745      0.000      -0.217      -0.198\n",
      "hashtag_Zelenskyy              -0.3274      0.006    -56.758      0.000      -0.339      -0.316\n",
      "hashtag_putin                   0.0025      0.005      0.457      0.648      -0.008       0.013\n",
      "hashtag_russia                  0.0401      0.005      8.766      0.000       0.031       0.049\n",
      "hashtag_ukraine                 0.2274      0.003     67.506      0.000       0.221       0.234\n",
      "hashtag_war                     0.1619      0.005     34.432      0.000       0.153       0.171\n",
      "==============================================================================\n",
      "Omnibus:                   348918.267   Durbin-Watson:                   1.357\n",
      "Prob(Omnibus):                  0.000   Jarque-Bera (JB):           443515.600\n",
      "Skew:                           0.677   Prob(JB):                         0.00\n",
      "Kurtosis:                       3.632   Cond. No.                         7.69\n",
      "==============================================================================\n",
      "\n",
      "Notes:\n",
      "[1] R² is computed without centering (uncentered) since the model does not contain a constant.\n",
      "[2] Standard Errors assume that the covariance matrix of the errors is correctly specified.\n"
     ]
    }
   ],
   "source": [
    "top_hashtags = {}\n",
    "for sentiment in tqdm(sentiment_cols):\n",
    "    print(\"**********************\")\n",
    "    print(f\"Analyzing {sentiment}\")\n",
    "    result = sm.OLS(df[sentiment], df[hashtag_cols]).fit()\n",
    "    print(result.summary())\n",
    "    top_hashtags[sentiment] = result.params.sort_values(ascending=False).head(5).to_dict()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "id": "165afe9d",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "negative: ['Mariupol', 'SafeAirliftUkraine', 'StopPutin', 'UkraineUnderAttack', 'Putin']\n",
      "neutral: ['BREAKING', 'EU', 'China', 'US', 'NATO']\n",
      "positive: []\n",
      "anger: ['UKRAINE', 'StopRussia', 'StopPutin', 'putin', 'RussianUkrainianWar']\n",
      "joy: ['SlavaUkraini', 'Zelenskyy']\n",
      "optimism: ['StandWithUkraine', 'StandWithUkraine️', 'China', 'EU', 'SafeAirliftUkraine']\n",
      "sadness: ['Mariupol', 'SafeAirliftUkraine', 'UkraineUnderAttack', 'BREAKING', 'Kharkiv']\n"
     ]
    }
   ],
   "source": [
    "for sentiment, hashtags in top_hashtags.items():\n",
    "    print(f\"{sentiment}: {[k.split('_')[1] for k,v in list(hashtags.items()) if v > 0]}\")"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3 (ipykernel)",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.8.8"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}