03_did_an.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 53,
   "id": "ba95554b",
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "import matplotlib.pyplot as plt\n",
    "from datetime import timedelta\n",
    "import matplotlib.dates as mdates\n",
    "from helpers.helpers import set_size\n",
    "import matplotlib.ticker as ticker\n",
    "import statsmodels\n",
    "import statsmodels.api as sm\n",
    "import numpy as np\n",
    "import seaborn as sns\n",
    "df = pd.read_csv(\"./data/banned_final_.csv\")\n",
    "df.banDate = pd.to_datetime(df.banDate)\n",
    "df_att = pd.read_csv(\"./data/joined_final.csv\")\n",
    "df_att.value_date = pd.to_datetime(df_att.value_date)\n",
    "df_att = df_att[df_att.source != \"mediacloud\"]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 54,
   "id": "9c531d16",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/var/folders/pn/m1444lgd3wj53304zyp1c63w0000gq/T/ipykernel_92990/588156721.py:16: FutureWarning: 'm' is deprecated and will be removed in a future version, please use 'ME' instead.\n",
      "  .groupby(pd.Grouper(key=\"value_date\", freq=\"m\")).value.sum()\n",
      "/var/folders/pn/m1444lgd3wj53304zyp1c63w0000gq/T/ipykernel_92990/588156721.py:36: FutureWarning: 'm' is deprecated and will be removed in a future version, please use 'ME' instead.\n",
      "  .groupby(pd.Grouper(key=\"value_date\", freq=\"m\")).value.sum()\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0.03\n",
      "0.021666666666666667\n",
      "2017-12-18T00:00:00.000000000\n",
      "2019-04-18T00:00:00.000000000\n",
      "0.061666666666666675\n",
      "0.07833333333333332\n",
      "2021-01-13T00:00:00.000000000\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 452.811x368.833 with 4 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "from datetime import timedelta\n",
    "\n",
    "fs = (5, 3.5)\n",
    "fig, axs = plt.subplots(2, 1, figsize=fs, gridspec_kw={\"hspace\": 0.6, \"wspace\": 0.5})\n",
    "axs = axs.flatten()\n",
    "\n",
    "bd = pd.to_datetime(\"2019-04-18\")\n",
    "for idx, entity, l, u in [\n",
    "    (0, \"Jayda Fransen\",  bd - timedelta(days=365), bd + timedelta(days=365)),\n",
    "    (1, \"Tom Fitton\", bd - timedelta(days=365), bd + timedelta(days=365))]:\n",
    "    ax1 = axs[idx]\n",
    "    ax2 = ax1.twinx()\n",
    "\n",
    "\n",
    "    x1 = df_att[(df_att.entity == entity) & (df_att.source == \"gtab\")]\\\n",
    "        .groupby(pd.Grouper(key=\"value_date\", freq=\"m\")).value.sum()\n",
    "    \n",
    "    x1 = x1[x1.index.values >= pd.to_datetime(l)]\n",
    "    x1 = x1[x1.index.values < pd.to_datetime(u)]\n",
    "    x_ban = df[(df.entity == entity)]\n",
    "    mean_pre = x1[x1.index.values < pd.to_datetime(bd)].mean()\n",
    "    print(mean_pre)\n",
    "    mean_pos = x1[x1.index.values >= pd.to_datetime(bd)].mean()\n",
    "    print(mean_pos)\n",
    "\n",
    "    ax1.plot(\n",
    "    [pd.to_datetime(l), pd.to_datetime(bd)],\n",
    "    [mean_pre, mean_pre], color=\"gray\")\n",
    "    ax1.plot(\n",
    "    [pd.to_datetime(bd),  pd.to_datetime(u)],\n",
    "    [mean_pos, mean_pos], color=\"gray\")\n",
    "    \n",
    "    ax1.scatter([v + timedelta(days=3) for v in x1.index] , x1.values, marker=\"v\", s=15, color=\"tab:blue\")\n",
    "\n",
    "    x2 = df_att[(df_att.entity == entity) & (df_att.source == \"wikimedia\")]\\\n",
    "        .groupby(pd.Grouper(key=\"value_date\", freq=\"m\")).value.sum()\n",
    "    x2 = x2[x2.index.values >= pd.to_datetime(l)]\n",
    "    x2 = x2[x2.index.values < pd.to_datetime(u)]\n",
    "    x_ban = df[(df.entity == entity)]\n",
    "    ax2.scatter([v - timedelta(days=3) for v in x2.index], x2.values, marker=\"^\", s=15, color=\"tab:red\")\n",
    "        \n",
    "    for ax, x in zip([ax1, ax2], [x1, x2]):\n",
    "        ax.set_ylim(0 - max(x)/10, max(x) + max(x)/10)\n",
    "        ylim_ax = ax.get_ylim()\n",
    "        ylim_ax_mid = (ylim_ax[1] + ylim_ax[0])/2\n",
    "        ylim_ax_d = (0 - ylim_ax[0])\n",
    "        ax.set_yticks([0, ylim_ax_mid, ylim_ax[1] - ylim_ax_d])\n",
    "        \n",
    "    def y_fmt(x, y):\n",
    "        if x == 0:\n",
    "            return \"0\"\n",
    "        if x < 0.5:\n",
    "            return '{:1.2f}'.format(round(x, 2))\n",
    "        if x < 1:\n",
    "            return '{:1.1f}'.format(round(x, 1))\n",
    "        if x < 10:\n",
    "            return '{:1.1f}'.format(round(x, 1))\n",
    "        if x < 1000:\n",
    "            return \"{:.0f}\".format(round(x))\n",
    "        else:\n",
    "            num = float('{:.3g}'.format(x))\n",
    "            magnitude = 0\n",
    "            while abs(num) >= 1000:\n",
    "                magnitude += 1\n",
    "                num /= 1000.0\n",
    "            return '{}{}'.format('{:f}'.format(num).rstrip('0').rstrip('.'), ['', 'K', 'M', 'B', 'T'][magnitude])\n",
    "\n",
    "    ax1.yaxis.set_major_formatter(ticker.FuncFormatter(y_fmt))\n",
    "    ax1.yaxis.set_major_formatter(ticker.FuncFormatter(y_fmt))\n",
    "    ax2.yaxis.set_major_formatter(ticker.FuncFormatter(y_fmt))\n",
    "    ax1.set_ylabel(\"G-Trends ▼\", color=\"tab:blue\")\n",
    "    ax2.set_ylabel(\"Wikipedia ▲\", color=\"tab:red\", rotation=270, labelpad=15)\n",
    "    \n",
    "    ax1.xaxis.set_major_locator(mdates.YearLocator())\n",
    "    ax1.xaxis.set_minor_locator(mdates.MonthLocator())\n",
    "    ax1.set_xlim([  pd.to_datetime(l),  pd.to_datetime(u)])\n",
    "    ax2.set_xlim([  pd.to_datetime(l),  pd.to_datetime(u)])\n",
    "    ax1.xaxis.set_minor_locator(mdates.MonthLocator())\n",
    "    #     ax.plot(x.index, x.values, marker=\".\")\n",
    "    #     ax.set_xlabel(\"Date\")\n",
    "    ax1.yaxis.grid(True, ls=\":\")\n",
    "    flag = 1\n",
    "    tmp__ = x_ban.drop_duplicates(\"bgid\")\n",
    "    for date, isTemp in zip(tmp__.banDate.values, tmp__.isTemporary.values):\n",
    "        print(date)\n",
    "        if date >  pd.to_datetime(u) or date < pd.to_datetime(l):\n",
    "            continue\n",
    "        ax1.axvspan(date, pd.to_datetime(u), color=\"tab:gray\", alpha=0.15)\n",
    "        ax1.axvline(date, zorder=-3, color=\"tab:grey\", lw=2, alpha=1, ls=\":\" if isTemp else \"-\",\n",
    "                    label=\"Deplatforming\" if flag == 1 else None)\n",
    "        flag = 0\n",
    "\n",
    "    #ax1.legend()\n",
    "    ax1.set_title(entity + \" (Treatment)\" if idx == 0 else entity + \" (Control)\")\n",
    "\n",
    "set_size(fig, fs)\n",
    "fig.savefig(\"./images/example_did.pdf\", bbox_inches=\"tight\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 55,
   "id": "08d694d3",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 510.059x262.733 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "fs = (4.75, 3.5)\n",
    "fig, axs = plt.subplots(1, 1, figsize=fs, gridspec_kw={\"hspace\": 0.6, \"wspace\": 0.5})\n",
    "\n",
    "plt.plot([1, 2], [0.0616, 0.078], marker=\"o\", color=\"tab:purple\", label=\"Control\")\n",
    "plt.plot([1, 2], [0.03, 0.0216], marker=\"o\", color=\"tab:green\", label=\"Treatment\")\n",
    "plt.plot([1, 2], [0.03, 0.03 + 0.016399999999999998], marker=\"x\", color=\"tab:green\", label=\"Counterfactual\", zorder=-1, ls=\":\")\n",
    "plt.plot\n",
    "plt.xlim([0.8, 2.3])\n",
    "plt.ylim([0, 0.1])\n",
    "plt.xticks([1, 2], [\"Pre\", \"Post\"])\n",
    "\n",
    "\n",
    "plt.legend(loc='upper center', bbox_to_anchor=(0.5, 1.45), fontsize=9, ncol=3)\n",
    "# Here is the label and arrow code of interest\n",
    "plt.annotate('$\\delta_{DD}$', xy=(2.05, 0.035), xytext=(2.15, .035), #xycoords='axes fraction', \n",
    "            fontsize=9, ha='center', va='center',\n",
    "            arrowprops=dict(arrowstyle='-[, widthB=2.1, lengthB=.6', lw=0.5, color='k'))\n",
    "plt.title(\"Difference-in-differences (intuition)\")\n",
    "plt.ylabel(\"Attention\")\n",
    "plt.xlabel(\"Period\")\n",
    "set_size(fig, fs)\n",
    "fig.savefig(\"./images/example_did2.pdf\", bbox_inches=\"tight\")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "cbfe2a72",
   "metadata": {},
   "source": [
    "# Gets dataframe\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 56,
   "id": "3881bddc",
   "metadata": {},
   "outputs": [],
   "source": [
    "df = pd.read_csv(\"./data/banned_final_.csv\")\n",
    "df.banDate = pd.to_datetime(df.banDate)\n",
    "df_att = pd.read_csv(\"./data/joined_final.csv\")\n",
    "df_att.value_date = pd.to_datetime(df_att.value_date)\n",
    "ban_dups = [1, 3, 39, 40, 42, 47, 56, 58, 66, 78, 94, 104, 106, 109, 127, 128, 134, 139]\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 62,
   "id": "920ee442",
   "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>entity</th>\n",
       "      <th>bgid</th>\n",
       "      <th>banDate</th>\n",
       "      <th>politician</th>\n",
       "      <th>media_personality</th>\n",
       "      <th>internet_personality</th>\n",
       "      <th>fringe_movements</th>\n",
       "      <th>platform_facebook</th>\n",
       "      <th>platform_instagram</th>\n",
       "      <th>platform_twitter</th>\n",
       "      <th>platform_youtube</th>\n",
       "      <th>reason2_HH</th>\n",
       "      <th>reason2_IV</th>\n",
       "      <th>reason2_MM</th>\n",
       "      <th>reason2_Other</th>\n",
       "      <th>isTemporary_False</th>\n",
       "      <th>isTemporary_True</th>\n",
       "      <th>banDate_months_since_2015</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>Alex Jones</td>\n",
       "      <td>2</td>\n",
       "      <td>2018-09-06</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>True</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>44</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>Ali Alexander</td>\n",
       "      <td>4</td>\n",
       "      <td>2021-01-10</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>72</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>Ali Khamenei</td>\n",
       "      <td>5</td>\n",
       "      <td>2019-04-16</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>51</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>Amanda Chase</td>\n",
       "      <td>6</td>\n",
       "      <td>2021-01-08</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>72</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>Andy Ngo</td>\n",
       "      <td>7</td>\n",
       "      <td>2019-11-25</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>58</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>131</th>\n",
       "      <td>Tommy Robinson</td>\n",
       "      <td>133</td>\n",
       "      <td>2018-03-28</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>38</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>133</th>\n",
       "      <td>Tommy Robinson</td>\n",
       "      <td>135</td>\n",
       "      <td>2019-06-14</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>53</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>134</th>\n",
       "      <td>Tommy Robinson</td>\n",
       "      <td>136</td>\n",
       "      <td>2020-08-07</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>67</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>135</th>\n",
       "      <td>Tyra Sanchez</td>\n",
       "      <td>137</td>\n",
       "      <td>2018-04-25</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>39</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>136</th>\n",
       "      <td>Win Gatchalian</td>\n",
       "      <td>138</td>\n",
       "      <td>2019-09-04</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>56</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>120 rows × 18 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "             entity  bgid    banDate  politician  media_personality  \\\n",
       "1        Alex Jones     2 2018-09-06       False               True   \n",
       "3     Ali Alexander     4 2021-01-10       False              False   \n",
       "4      Ali Khamenei     5 2019-04-16        True              False   \n",
       "5      Amanda Chase     6 2021-01-08        True              False   \n",
       "6          Andy Ngo     7 2019-11-25       False               True   \n",
       "..              ...   ...        ...         ...                ...   \n",
       "131  Tommy Robinson   133 2018-03-28       False              False   \n",
       "133  Tommy Robinson   135 2019-06-14       False              False   \n",
       "134  Tommy Robinson   136 2020-08-07       False              False   \n",
       "135    Tyra Sanchez   137 2018-04-25       False               True   \n",
       "136  Win Gatchalian   138 2019-09-04        True              False   \n",
       "\n",
       "     internet_personality  fringe_movements  platform_facebook  \\\n",
       "1                    True              True              False   \n",
       "3                    True              True              False   \n",
       "4                   False             False              False   \n",
       "5                   False             False               True   \n",
       "6                    True             False              False   \n",
       "..                    ...               ...                ...   \n",
       "131                 False              True              False   \n",
       "133                 False              True              False   \n",
       "134                 False              True              False   \n",
       "135                 False             False              False   \n",
       "136                 False             False              False   \n",
       "\n",
       "     platform_instagram  platform_twitter  platform_youtube  reason2_HH  \\\n",
       "1                 False              True             False        True   \n",
       "3                 False              True             False       False   \n",
       "4                  True             False             False       False   \n",
       "5                 False             False             False       False   \n",
       "6                 False              True             False        True   \n",
       "..                  ...               ...               ...         ...   \n",
       "131               False              True             False        True   \n",
       "133               False             False              True        True   \n",
       "134                True             False             False        True   \n",
       "135               False              True             False       False   \n",
       "136               False              True             False       False   \n",
       "\n",
       "     reason2_IV  reason2_MM  reason2_Other  isTemporary_False  \\\n",
       "1         False       False          False               True   \n",
       "3          True       False          False               True   \n",
       "4         False       False           True               True   \n",
       "5          True       False          False               True   \n",
       "6         False       False          False              False   \n",
       "..          ...         ...            ...                ...   \n",
       "131       False       False          False               True   \n",
       "133       False       False          False               True   \n",
       "134       False       False          False              False   \n",
       "135        True       False          False               True   \n",
       "136       False        True          False              False   \n",
       "\n",
       "     isTemporary_True  banDate_months_since_2015  \n",
       "1               False                         44  \n",
       "3               False                         72  \n",
       "4               False                         51  \n",
       "5               False                         72  \n",
       "6                True                         58  \n",
       "..                ...                        ...  \n",
       "131             False                         38  \n",
       "133             False                         53  \n",
       "134              True                         67  \n",
       "135             False                         39  \n",
       "136              True                         56  \n",
       "\n",
       "[120 rows x 18 columns]"
      ]
     },
     "execution_count": 62,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "def diff_month(d1, d2):\n",
    "    return (d1.year - d2.year) * 12 + d1.month - d2.month\n",
    "\n",
    "\n",
    "df_simplified = df.drop([\"isRelevant\", \"kind\", \"searchTerm\", \"reasonComment\", \"reason\", \"gid\", \"article\", \n",
    "         \"qid\", \"descriptions\", \"sources\"], axis=1)\n",
    "df_simplified = pd.get_dummies(df_simplified, columns=[\"platform\", \"reason2\", \"isTemporary\"])\n",
    "\n",
    "df_simplified.columns\n",
    "\n",
    "df_simplified = df_simplified.groupby([\"entity\", \"bgid\"]).max().reset_index()\n",
    "df_simplified[\"banDate_months_since_2015\"]  = df_simplified.banDate.apply(lambda x: \n",
    "                                                                          diff_month(x, pd.to_datetime(\"2015\")))\n",
    "df_simplified = df_simplified.loc[~df_simplified.bgid.isin(ban_dups)].copy()\n",
    "\n",
    "df_simplified"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 63,
   "id": "1e156b63",
   "metadata": {},
   "outputs": [],
   "source": [
    "\n",
    "df_att[\"months_since_2015\"]  = df_att.value_date.apply(lambda x: diff_month(x, pd.to_datetime(\"2015\")))\n",
    "df_att_simplified = df_att.groupby([\"entity\", \"google_id\", \"source\", \"months_since_2015\"]\n",
    "                                  ).value.mean().reset_index()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 64,
   "id": "1f607895",
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "16 o% Azealia Banks 50\n",
      "17 o% Mark Kern 49\n",
      "18 o% Milo Yiannopoulos 49\n",
      "18 o% Hamza Abbasi 49\n",
      "18 o% Hafiz Saeed 49\n",
      "21 o% Jonathan Daniel Brown 48\n",
      "24 o% Carl Benjamin 42\n",
      "24 o% Martin Shkreli 43\n",
      "26 o% David Duke 42\n",
      "29 o% Anthony Cumia 38\n",
      "33 o% Roger Stone 35\n",
      "33 o% Guo Wengui 35\n",
      "33 o% Austin Petersen 35\n",
      "33 o% Kamaal R. Khan 35\n",
      "33 o% Rose McGowan 35\n",
      "34 o% Baked Alaska 35\n",
      "35 o% James Allsup 32\n",
      "35 o% Jayda Fransen 32\n",
      "36 o% Beatrix von Storch 32\n",
      "36 o% David Clarke 32\n",
      "36 o% Paul Nehlen 32\n",
      "38 o% Tommy Robinson 32\n",
      "39 o% Tyra Sanchez 32\n",
      "39 o% Owen Benjamin 32\n",
      "39 o% Betzalel Smotrich 32\n",
      "39 o% Richard B. Spencer 32\n",
      "41 o% Azealia Banks 28\n",
      "41 o% Gavin McInnes 28\n",
      "43 o% TJ Kirk 26\n",
      "44 o% Kris Paronto 26\n",
      "44 o% Alex Jones 26\n",
      "44 o% James Woods 26\n",
      "46 o% Meghan Murphy 25\n",
      "46 o% Jesse Kelly 25\n",
      "46 o% Laura Loomer 25\n",
      "46 o% Khadim Hussain Rizvi 25\n",
      "47 o% Titania McGrath 25\n",
      "51 o% Jayda Fransen 24\n",
      "51 o% Ali Khamenei 24\n",
      "51 o% Sara Duterte 24\n",
      "51 o% Carl Benjamin 24\n",
      "51 o% Ben Garrison 24\n",
      "52 o% Louis Farrakhan 23\n",
      "52 o% Geert Wilders 23\n",
      "52 o% Candace Owens 23\n",
      "53 o% Tommy Robinson 22\n",
      "53 o% Gemma O'Doherty 22\n",
      "54 o% Lindsay Shepherd 18\n",
      "55 o% Pauline Hanson 18\n",
      "55 o% Mitch McConnell 18\n",
      "56 o% Saoud Al Qahtani 17\n",
      "56 o% Win Gatchalian 17\n",
      "56 o% Briscoe Cain 17\n",
      "58 o% Andy Ngo 14\n",
      "60 o% Katie Hopkins 12\n",
      "61 o% Nicholas J. Fuentes 2\n",
      "62 o% Cory Doctorow 1\n",
      "62 o% James Woods 1\n",
      "62 o% Jack Burkman 1\n",
      "63 o% Leland Sklar 1\n",
      "63 o% Diamond and Silk 1\n"
     ]
    }
   ],
   "source": [
    "\n",
    "def get_tuples(months, data, diff=12):\n",
    "    \n",
    "    acc = []\n",
    "    \n",
    "    data_permanent = data.loc[data.isTemporary_False].copy(deep=True)\n",
    "    for idx, rowx in data.sort_values(\"banDate_months_since_2015\").iterrows():\n",
    "        rowx = dict(rowx)\n",
    "        entity_x = rowx[\"entity\"]\n",
    "        ban_date_x = rowx[\"banDate_months_since_2015\"]\n",
    "        \n",
    "        previous_bans = data[(data[\"entity\"] == entity_x) & \n",
    "                  (data[\"banDate_months_since_2015\"] < ban_date_x)].banDate_months_since_2015\n",
    "        \n",
    "        # if previously banned very recently, remove\n",
    "        match_ban = True\n",
    "        for ban_date_y in previous_bans:\n",
    "            if ban_date_y < ban_date_x + diff and ban_date_y > ban_date_x - diff:\n",
    "                match_ban=False\n",
    "        \n",
    "        if not match_ban:\n",
    "            continue\n",
    "#         print(rowx)\n",
    "        \n",
    "        \n",
    "        pass_criteria_1 = []\n",
    "        for entity_y, df_y in data_permanent.groupby(\"entity\"):\n",
    "            if entity_y == entity_x:\n",
    "                continue\n",
    "                \n",
    "            min_date_ban_y = min(df_y.banDate_months_since_2015.values)\n",
    "            max_date_ban_y = max(df_y.banDate_months_since_2015.values)\n",
    "            if min_date_ban_y >= ban_date_x + diff:\n",
    "                pass_criteria_1.append(entity_y)\n",
    "        \n",
    "        pass_criteria_2 = []\n",
    "        for entity_y, df_y in data.groupby(\"entity\"):\n",
    "            \n",
    "            if entity_y == entity_x:\n",
    "                continue\n",
    "                \n",
    "            match_ban = True\n",
    "            \n",
    "            for ban_date_y in df_y.banDate_months_since_2015:\n",
    "                if ban_date_y < ban_date_x + diff and ban_date_y > ban_date_x - diff:\n",
    "                    match_ban=False\n",
    "            \n",
    "            if match_ban:\n",
    "                pass_criteria_2.append(entity_y)\n",
    "        \n",
    "        tmp = list(set(pass_criteria_1).intersection(set(pass_criteria_1)))\n",
    "        flag = \"o%\" if len(tmp) > 0 else \"x%\"\n",
    "        \n",
    "        if len(tmp) > 0:\n",
    "            for entity_y in tmp:\n",
    "                acc.append({\n",
    "                \"experiment\": ban_date_x,\n",
    "                \"entity\": entity_y,\n",
    "                \"condition\": \"control\",\n",
    "                })\n",
    "            acc.append({\n",
    "                \"experiment\": ban_date_x,\n",
    "                \"entity\": entity_x,\n",
    "                \"condition\": \"treatment\",\n",
    "                })\n",
    "            print(ban_date_x, flag, entity_x,  len(tmp))\n",
    "\n",
    "    return pd.DataFrame(acc)\n",
    "\n",
    "\n",
    "df_tuples = get_tuples(1, df_simplified).drop_duplicates([\"experiment\", \"entity\"])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 65,
   "id": "306eb8a0",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "37484\n",
      "[  1   3   4   5   6   7  10  12  13  17  18  19  21  22  23  24  26  28\n",
      "  29  30  31  32  35  36  39  41  42  43  46  47  49  50  51  52  53  55\n",
      "  57  58  60  62  63  65  66  67  68  69  71  72  73  74  75  77  78  79\n",
      "  80  81  84  85  86  88  89  90  91  94  95  96  97  99 100 101 103 105\n",
      " 108 109 111 112 114 115 117 118 119 120 121 122 123 124 127 128 129 130\n",
      " 131 135 136]\n",
      "37484\n"
     ]
    }
   ],
   "source": [
    "df_tuples_all = df_tuples.merge(df_att_simplified, on=\"entity\")\n",
    "df_tuples_all[\"diff\"] = (df_tuples_all[\"months_since_2015\"] - df_tuples_all[\"experiment\"]).astype(int)\n",
    "df_tuples_all = df_tuples_all.loc[df_tuples_all[\"diff\"].abs() <= 12].copy(deep=True)\n",
    "df_tuples_all[\"treat\"] = (df_tuples_all[\"condition\"] == \"treatment\").astype(int)\n",
    "df_tuples_all.loc[df_tuples_all.value == 0, \"value\"] =  0.00001\n",
    "df_tuples_all[\"logval\"] = np.log(df_tuples_all[\"value\"])\n",
    "print(len(df_tuples_all))\n",
    "\n",
    "df_simplified[\"isFirstBan\"] = 0\n",
    "first_ban = df_simplified.drop_duplicates(\"entity\").index.values #\n",
    "print(first_ban)\n",
    "df_simplified.loc[[x in first_ban for x in df_simplified.index.values],  \"isFirstBan\"] = 1\n",
    "\n",
    "\n",
    "df_tuples_all_treat = df_tuples_all[df_tuples_all.treat == 1].merge(df_simplified, \n",
    "    left_on=[\"entity\", \"experiment\"], \n",
    "    right_on=[\"entity\", \"banDate_months_since_2015\"])\n",
    "\n",
    "df_tuples_all_ctrl = df_tuples_all[df_tuples_all.treat != 1]\n",
    "\n",
    "print(len(df_tuples_all_ctrl) + len(df_tuples_all_treat))\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 66,
   "id": "eae64f1d",
   "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>experiment</th>\n",
       "      <th>entity</th>\n",
       "      <th>condition</th>\n",
       "      <th>google_id</th>\n",
       "      <th>source</th>\n",
       "      <th>months_since_2015</th>\n",
       "      <th>value</th>\n",
       "      <th>diff</th>\n",
       "      <th>treat</th>\n",
       "      <th>logval</th>\n",
       "      <th>...</th>\n",
       "      <th>platform_youtube</th>\n",
       "      <th>reason2_HH</th>\n",
       "      <th>reason2_IV</th>\n",
       "      <th>reason2_MM</th>\n",
       "      <th>reason2_Other</th>\n",
       "      <th>isTemporary_False</th>\n",
       "      <th>isTemporary_True</th>\n",
       "      <th>banDate_months_since_2015</th>\n",
       "      <th>isFirstBan</th>\n",
       "      <th>entity_exp</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>16</td>\n",
       "      <td>Azealia Banks</td>\n",
       "      <td>treatment</td>\n",
       "      <td>/m/0hnfbjv</td>\n",
       "      <td>gtab</td>\n",
       "      <td>4</td>\n",
       "      <td>1.307692</td>\n",
       "      <td>-12</td>\n",
       "      <td>1</td>\n",
       "      <td>0.268264</td>\n",
       "      <td>...</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>16.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>Azealia Banks_16</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>16</td>\n",
       "      <td>Azealia Banks</td>\n",
       "      <td>treatment</td>\n",
       "      <td>/m/0hnfbjv</td>\n",
       "      <td>gtab</td>\n",
       "      <td>5</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>-11</td>\n",
       "      <td>1</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>...</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>16.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>Azealia Banks_16</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>16</td>\n",
       "      <td>Azealia Banks</td>\n",
       "      <td>treatment</td>\n",
       "      <td>/m/0hnfbjv</td>\n",
       "      <td>gtab</td>\n",
       "      <td>6</td>\n",
       "      <td>1.076923</td>\n",
       "      <td>-10</td>\n",
       "      <td>1</td>\n",
       "      <td>0.074108</td>\n",
       "      <td>...</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>16.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>Azealia Banks_16</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>16</td>\n",
       "      <td>Azealia Banks</td>\n",
       "      <td>treatment</td>\n",
       "      <td>/m/0hnfbjv</td>\n",
       "      <td>gtab</td>\n",
       "      <td>7</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>-9</td>\n",
       "      <td>1</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>...</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>16.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>Azealia Banks_16</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>16</td>\n",
       "      <td>Azealia Banks</td>\n",
       "      <td>treatment</td>\n",
       "      <td>/m/0hnfbjv</td>\n",
       "      <td>gtab</td>\n",
       "      <td>8</td>\n",
       "      <td>1.769231</td>\n",
       "      <td>-8</td>\n",
       "      <td>1</td>\n",
       "      <td>0.570545</td>\n",
       "      <td>...</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>16.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>Azealia Banks_16</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>139021</th>\n",
       "      <td>63</td>\n",
       "      <td>James O'Keefe</td>\n",
       "      <td>control</td>\n",
       "      <td>/m/07s76_k</td>\n",
       "      <td>wikimedia</td>\n",
       "      <td>71</td>\n",
       "      <td>2279.322581</td>\n",
       "      <td>8</td>\n",
       "      <td>0</td>\n",
       "      <td>7.731634</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>James O'Keefe_63</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>139022</th>\n",
       "      <td>63</td>\n",
       "      <td>James O'Keefe</td>\n",
       "      <td>control</td>\n",
       "      <td>/m/07s76_k</td>\n",
       "      <td>wikimedia</td>\n",
       "      <td>72</td>\n",
       "      <td>1129.032258</td>\n",
       "      <td>9</td>\n",
       "      <td>0</td>\n",
       "      <td>7.029116</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>James O'Keefe_63</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>139023</th>\n",
       "      <td>63</td>\n",
       "      <td>James O'Keefe</td>\n",
       "      <td>control</td>\n",
       "      <td>/m/07s76_k</td>\n",
       "      <td>wikimedia</td>\n",
       "      <td>73</td>\n",
       "      <td>1019.892857</td>\n",
       "      <td>10</td>\n",
       "      <td>0</td>\n",
       "      <td>6.927453</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>James O'Keefe_63</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>139024</th>\n",
       "      <td>63</td>\n",
       "      <td>James O'Keefe</td>\n",
       "      <td>control</td>\n",
       "      <td>/m/07s76_k</td>\n",
       "      <td>wikimedia</td>\n",
       "      <td>74</td>\n",
       "      <td>520.096774</td>\n",
       "      <td>11</td>\n",
       "      <td>0</td>\n",
       "      <td>6.254015</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>James O'Keefe_63</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>139025</th>\n",
       "      <td>63</td>\n",
       "      <td>James O'Keefe</td>\n",
       "      <td>control</td>\n",
       "      <td>/m/07s76_k</td>\n",
       "      <td>wikimedia</td>\n",
       "      <td>75</td>\n",
       "      <td>2810.400000</td>\n",
       "      <td>12</td>\n",
       "      <td>0</td>\n",
       "      <td>7.941082</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>James O'Keefe_63</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>37484 rows × 29 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "        experiment         entity  condition   google_id     source  \\\n",
       "0               16  Azealia Banks  treatment  /m/0hnfbjv       gtab   \n",
       "1               16  Azealia Banks  treatment  /m/0hnfbjv       gtab   \n",
       "2               16  Azealia Banks  treatment  /m/0hnfbjv       gtab   \n",
       "3               16  Azealia Banks  treatment  /m/0hnfbjv       gtab   \n",
       "4               16  Azealia Banks  treatment  /m/0hnfbjv       gtab   \n",
       "...            ...            ...        ...         ...        ...   \n",
       "139021          63  James O'Keefe    control  /m/07s76_k  wikimedia   \n",
       "139022          63  James O'Keefe    control  /m/07s76_k  wikimedia   \n",
       "139023          63  James O'Keefe    control  /m/07s76_k  wikimedia   \n",
       "139024          63  James O'Keefe    control  /m/07s76_k  wikimedia   \n",
       "139025          63  James O'Keefe    control  /m/07s76_k  wikimedia   \n",
       "\n",
       "        months_since_2015        value  diff  treat    logval  ...  \\\n",
       "0                       4     1.307692   -12      1  0.268264  ...   \n",
       "1                       5     1.000000   -11      1  0.000000  ...   \n",
       "2                       6     1.076923   -10      1  0.074108  ...   \n",
       "3                       7     1.000000    -9      1  0.000000  ...   \n",
       "4                       8     1.769231    -8      1  0.570545  ...   \n",
       "...                   ...          ...   ...    ...       ...  ...   \n",
       "139021                 71  2279.322581     8      0  7.731634  ...   \n",
       "139022                 72  1129.032258     9      0  7.029116  ...   \n",
       "139023                 73  1019.892857    10      0  6.927453  ...   \n",
       "139024                 74   520.096774    11      0  6.254015  ...   \n",
       "139025                 75  2810.400000    12      0  7.941082  ...   \n",
       "\n",
       "        platform_youtube reason2_HH reason2_IV reason2_MM reason2_Other  \\\n",
       "0                  False       True      False      False         False   \n",
       "1                  False       True      False      False         False   \n",
       "2                  False       True      False      False         False   \n",
       "3                  False       True      False      False         False   \n",
       "4                  False       True      False      False         False   \n",
       "...                  ...        ...        ...        ...           ...   \n",
       "139021                 0          0          0          0             0   \n",
       "139022                 0          0          0          0             0   \n",
       "139023                 0          0          0          0             0   \n",
       "139024                 0          0          0          0             0   \n",
       "139025                 0          0          0          0             0   \n",
       "\n",
       "       isTemporary_False isTemporary_True banDate_months_since_2015  \\\n",
       "0                  False             True                      16.0   \n",
       "1                  False             True                      16.0   \n",
       "2                  False             True                      16.0   \n",
       "3                  False             True                      16.0   \n",
       "4                  False             True                      16.0   \n",
       "...                  ...              ...                       ...   \n",
       "139021                 0                0                       0.0   \n",
       "139022                 0                0                       0.0   \n",
       "139023                 0                0                       0.0   \n",
       "139024                 0                0                       0.0   \n",
       "139025                 0                0                       0.0   \n",
       "\n",
       "       isFirstBan        entity_exp  \n",
       "0             1.0  Azealia Banks_16  \n",
       "1             1.0  Azealia Banks_16  \n",
       "2             1.0  Azealia Banks_16  \n",
       "3             1.0  Azealia Banks_16  \n",
       "4             1.0  Azealia Banks_16  \n",
       "...           ...               ...  \n",
       "139021        0.0  James O'Keefe_63  \n",
       "139022        0.0  James O'Keefe_63  \n",
       "139023        0.0  James O'Keefe_63  \n",
       "139024        0.0  James O'Keefe_63  \n",
       "139025        0.0  James O'Keefe_63  \n",
       "\n",
       "[37484 rows x 29 columns]"
      ]
     },
     "execution_count": 66,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_tuples_all_ex = pd.concat([df_tuples_all_treat, df_tuples_all_ctrl]).fillna(0)\n",
    "df_tuples_all_ex[\"entity_exp\"] = df_tuples_all_ex[\"entity\"] + \"_\" + df_tuples_all_ex[\"experiment\"].astype(str)\n",
    "df_tuples_all_ex"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 67,
   "id": "498bdece",
   "metadata": {},
   "outputs": [],
   "source": [
    "df_tuples_all_ex[\"diff2\"] = df_tuples_all_ex[\"diff\"]\n",
    "df_tuples_all_ex = pd.get_dummies(df_tuples_all_ex, columns=[\"diff\"])\n",
    "for column in df_tuples_all_ex.columns:\n",
    "    if \"diff_\" in column:\n",
    "        df_tuples_all_ex[column] = df_tuples_all_ex[column].astype(int)\n",
    "df_tuples_all_ex.columns = [x.replace(\"diff_-\", \"diff_m\") for x in df_tuples_all_ex.columns]\n",
    "df_tuples_all_ex.columns = [x.replace(\" \", \"_\") for x in df_tuples_all_ex.columns]\n",
    "df_tuples_all_ex.columns = [x.replace(\"/\", \"_\") for x in df_tuples_all_ex.columns]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 68,
   "id": "7f7e4957",
   "metadata": {},
   "outputs": [],
   "source": [
    "to_int = ['politician', 'media_personality', 'internet_personality',\n",
    "       'fringe_movements', 'platform_facebook', 'platform_instagram',\n",
    "       'platform_twitter', 'platform_youtube',\n",
    "#        'reason_Hate_Harassment_Incitement_to_Violence',\n",
    "#        'reason_Manipulation_Misinformation', 'reason_Other',\n",
    "       'reason2_HH', \"reason2_IV\",\n",
    "       'reason2_MM', 'reason2_Other',\n",
    "       'isTemporary_False', 'isTemporary_True', 'banDate_months_since_2015',\n",
    "       'isFirstBan', 'diff_m12', 'diff_m11', 'diff_m10',\n",
    "       'diff_m9', 'diff_m8', 'diff_m7', 'diff_m6', 'diff_m5', 'diff_m4',\n",
    "       'diff_m3', 'diff_m2', 'diff_m1', 'diff_0', 'diff_1', 'diff_2', 'diff_3',\n",
    "       'diff_4', 'diff_5', 'diff_6', 'diff_7', 'diff_8', 'diff_9', 'diff_10',\n",
    "       'diff_11', 'diff_12']\n",
    "\n",
    "for col in to_int:\n",
    "    df_tuples_all_ex[col] = df_tuples_all_ex[col].astype(int)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 69,
   "id": "6d3f4527",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "'logval ~  + diff_m12:treat  + diff_m11:treat  + diff_m10:treat  + diff_m9:treat  + diff_m8:treat  + diff_m7:treat  + diff_m6:treat  + diff_m5:treat  + diff_m4:treat  + diff_m2:treat  + diff_m1:treat  + diff_0:treat  + diff_1:treat  + diff_2:treat  + diff_3:treat  + diff_4:treat  + diff_5:treat  + diff_6:treat  + diff_7:treat  + diff_8:treat  + diff_9:treat  + diff_10:treat  + diff_11:treat  + diff_12:treat '"
      ]
     },
     "execution_count": 69,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "ref = -3\n",
    "acc_form = \"logval ~ \"\n",
    "for i in range(-12, 13):\n",
    "    if i == ref:\n",
    "        continue\n",
    "    acc_form += \" + diff_{0}:treat \".format(str(i))\n",
    "acc_form = acc_form.replace(\"diff_-\", \"diff_m\")\n",
    "acc_form"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 70,
   "id": "2f7e820f",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.07320456728204033"
      ]
     },
     "execution_count": 70,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "(df_tuples_all_ex.isFirstBan == 1).sum() / len(df_tuples_all_ex)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 71,
   "id": "b0a678b0",
   "metadata": {
    "scrolled": false
   },
   "outputs": [],
   "source": [
    "from linearmodels import PanelOLS\n",
    "\n",
    "formula = acc_form + \"\"\"  + EntityEffects + TimeEffects \"\"\"\n",
    "mod = PanelOLS.from_formula(formula, \n",
    "                            df_tuples_all_ex[(df_tuples_all_ex.source == \"gtab\")].set_index(\n",
    "                                [\"entity_exp\", \"months_since_2015\"]), \n",
    "                            drop_absorbed=True)\n",
    "resgt = mod.fit(cov_type='clustered', cluster_entity=True)\n",
    "\n",
    "\n",
    "formula = acc_form + \"\"\"  + EntityEffects + TimeEffects \"\"\"\n",
    "mod = PanelOLS.from_formula(formula, \n",
    "                            df_tuples_all_ex[(df_tuples_all_ex.source == \"wikimedia\")].set_index(\n",
    "                                [\"entity_exp\", \"months_since_2015\"]), \n",
    "                            drop_absorbed=True)\n",
    "reswm = mod.fit(cov_type='clustered', cluster_entity=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 72,
   "id": "4b4d40fd",
   "metadata": {
    "scrolled": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "-36.49 % 95\\% CI [ -52.81 ,  -14.51 ]\n",
      "85.79 % 95\\% CI [ 38.92 ,  148.47 ]\n",
      "-55.62 % 95\\% CI [ -69.74 ,  -34.93 ]\n",
      "139.61 % 95\\% CI [ 52.99 ,  275.26 ]\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1160.41x396.032 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "import numpy as np\n",
    "fs = (10, 4)\n",
    "fig, axs = plt.subplots(1, 2, figsize=fs, gridspec_kw={\"hspace\": 0.5, \"wspace\": 0.1}, sharey=True)\n",
    "axs = axs.flatten()\n",
    "\n",
    "for res, ax, color, lss in zip([reswm, resgt], [axs[0], axs[1]], [\"tab:red\", \"tab:blue\"],\n",
    "                              [\"--\", \":\"]):\n",
    "\n",
    "    pes = (np.exp(res.params) - 1) * 100\n",
    "    low = (np.exp((res.params - res.std_errors  * 1.96)) - 1) * 100\n",
    "    hig = (np.exp((res.params + res.std_errors  * 1.96)) - 1) * 100\n",
    "    \n",
    "    print(round(pes.values[-1],2), \"% 95\\% CI [\",round(low.values[-1],2), \", \",\n",
    "          round(hig.values[-1], 2), \"]\")\n",
    "    \n",
    "    print(round(pes.values[-13],2), \"% 95\\% CI [\",round(low.values[-13],2), \", \",\n",
    "          round(hig.values[-13], 2), \"]\")\n",
    "\n",
    "    df_ef = pd.DataFrame(\n",
    "        {\n",
    "            \"pes\": pes,\n",
    "            \"low\": low,\n",
    "            \"hig\": hig,\n",
    "        }\n",
    "    ).reset_index()\n",
    "\n",
    "    df_ef[\"index\"] = df_ef[\"index\"].apply(lambda x: x.replace(\"diff_\", \"\").replace(\":treat\", \"\").replace(\"m\", \"-\"))\n",
    "\n",
    "    df_ef = pd.concat(\n",
    "    [\n",
    "        df_ef,\n",
    "         pd.DataFrame(\n",
    "        {\n",
    "            \"index\": [\"-3\"],\n",
    "            \"pes\": [0],\n",
    "            \"low\": [0],\n",
    "            \"hig\": [0]\n",
    "        } )\n",
    "    ])\n",
    "    df_ef[\"index\"] = df_ef[\"index\"].astype(int)\n",
    "    df_ef = df_ef.sort_values(\"index\").set_index(\"index\")\n",
    "    ax.plot(df_ef.pes, color=color, ls=lss)\n",
    "    ax.scatter(df_ef.index.values, df_ef.pes, color=color)\n",
    "    ax.fill_between(df_ef.index.values, df_ef[\"low\"], df_ef[\"hig\"], color=color, alpha=0.3, zorder=2)\n",
    "    ax.axhline(0, ls=\":\", color=\"black\", zorder=-5)\n",
    "    ax.axvline(0, ls=\":\", color=\"black\", zorder=-5)\n",
    "    \n",
    "axs[0].set_title(\"Wikipedia\", size=14)\n",
    "axs[1].set_title(\"Google Trends\", size=14)\n",
    "axs[0].set_ylabel(\"Effect (% change)\", size=14)\n",
    "axs[0].set_xticks([-12, -9, -6, -3, 0, 3, 6, 9, 12])\n",
    "axs[1].set_xticks([-12, -9, -6, -3, 0, 3, 6, 9, 12])\n",
    "axs[1].set_ylim([-100, +250])\n",
    "axs[0].set_xlabel(\"Month (Relative to Banning)\", size=14)\n",
    "axs[1].set_xlabel(\"Month (Relative to Banning)\", size=14)\n",
    "axs[0].xaxis.set_tick_params(labelsize=12)\n",
    "axs[0].yaxis.set_tick_params(labelsize=12)\n",
    "axs[1].xaxis.set_tick_params(labelsize=12)\n",
    "axs[1].yaxis.set_tick_params(labelsize=12)\n",
    "set_size(fig, fs)\n",
    "\n",
    "fig.savefig(\"./images/didres1.pdf\", bbox_inches=\"tight\")\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "id": "5fb20289",
   "metadata": {
    "scrolled": false
   },
   "outputs": [],
   "source": [
    "x = df_tuples_all_ex[\n",
    "    (df_tuples_all_ex.source == \"wikimedia\") & \n",
    "    (df_tuples_all_ex.treat==1) &\n",
    "    (df_tuples_all_ex.diff_m12 ==1)  \n",
    "                ][[\"experiment\", \"entity\", \"value\", \"isTemporary_False\"]].drop_duplicates().sort_values(by=\"value\").reset_index()\n",
    "\n",
    "\n",
    "lowv = {(y,z): \"low\" for y,z in x.iloc[0:2*(len(x)//2)][[\"experiment\", \"entity\"]].values}\n",
    "highv = {(y,z): \"high\" for y,z in  x.iloc[2*(len(x)//2):][[\"experiment\", \"entity\"]].values}\n",
    "df_tuples_all_ex[\"exp_entity\"] = [(y,z) for y, z in df_tuples_all_ex[[\"experiment\", \"entity\"]].values]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "id": "fdf4506d",
   "metadata": {},
   "outputs": [],
   "source": [
    "x = df_tuples_all_ex[\n",
    "    (df_tuples_all_ex.source == \"wikimedia\") & \n",
    "    (df_tuples_all_ex.treat==1) &\n",
    "    (df_tuples_all_ex.diff2 == -11)  \n",
    "                ][[\"experiment\", \"entity\", \"value\"]\n",
    "                 ].groupby([\"experiment\", \"entity\"]).value.sum().reset_index().sort_values(by=\"value\")\n",
    "    \n",
    "\n",
    "\n",
    "lowv = {(y,z): \"low\" for y,z in x.iloc[0:2*(len(x)//3)][[\"experiment\", \"entity\"]].values}\n",
    "highv = {(y,z): \"high\" for y,z in  x.iloc[2*(len(x)//3):][[\"experiment\", \"entity\"]].values}\n",
    "df_tuples_all_ex[\"exp_entity\"] = [(y,z) for y, z in df_tuples_all_ex[[\"experiment\", \"entity\"]].values]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "id": "e8bd7c8e",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "48089.96551724138"
      ]
     },
     "execution_count": 18,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "x.iloc[2*(len(x)//2):][\"value\"].mean()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "id": "fffbbd02",
   "metadata": {},
   "outputs": [],
   "source": [
    "df_tuples_all_ex[\"low_attention\"] = 0\n",
    "df_tuples_all_ex[\"low_attention\"] = df_tuples_all_ex[\"exp_entity\"].apply(lambda x: 1 if x in lowv else 0)\n",
    "df_tuples_all_ex[\"high_attention\"] = 0\n",
    "df_tuples_all_ex[\"high_attention\"] = df_tuples_all_ex[\"exp_entity\"].apply(lambda x: 1 if x in highv else 0)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "id": "21954fe2",
   "metadata": {},
   "outputs": [],
   "source": [
    "df_tuples_all_ex[\"mult_bans\"] = 0\n",
    "df_tuples_all_ex[\"mult_bans\"] = df_tuples_all_ex[\"bgid\"].apply(lambda x: 1 if x in ban_dups else 0)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "id": "08566209",
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/var/folders/pn/m1444lgd3wj53304zyp1c63w0000gq/T/ipykernel_92990/1360805091.py:33: AbsorbingEffectWarning: \n",
      "Variables have been fully absorbed and have removed from the regression:\n",
      "\n",
      "isTemporary_True\n",
      "\n",
      "  res = mod.fit(cov_type='clustered', cluster_entity=True)\n",
      "/var/folders/pn/m1444lgd3wj53304zyp1c63w0000gq/T/ipykernel_92990/1360805091.py:33: AbsorbingEffectWarning: \n",
      "Variables have been fully absorbed and have removed from the regression:\n",
      "\n",
      "isTemporary_True\n",
      "\n",
      "  res = mod.fit(cov_type='clustered', cluster_entity=True)\n",
      "/var/folders/pn/m1444lgd3wj53304zyp1c63w0000gq/T/ipykernel_92990/1360805091.py:33: AbsorbingEffectWarning: \n",
      "Variables have been fully absorbed and have removed from the regression:\n",
      "\n",
      "high_attention\n",
      "\n",
      "  res = mod.fit(cov_type='clustered', cluster_entity=True)\n",
      "/var/folders/pn/m1444lgd3wj53304zyp1c63w0000gq/T/ipykernel_92990/1360805091.py:33: AbsorbingEffectWarning: \n",
      "Variables have been fully absorbed and have removed from the regression:\n",
      "\n",
      "high_attention\n",
      "\n",
      "  res = mod.fit(cov_type='clustered', cluster_entity=True)\n",
      "/var/folders/pn/m1444lgd3wj53304zyp1c63w0000gq/T/ipykernel_92990/1360805091.py:33: AbsorbingEffectWarning: \n",
      "Variables have been fully absorbed and have removed from the regression:\n",
      "\n",
      "platform_facebook, platform_instagram, platform_twitter, platform_youtube\n",
      "\n",
      "  res = mod.fit(cov_type='clustered', cluster_entity=True)\n",
      "/var/folders/pn/m1444lgd3wj53304zyp1c63w0000gq/T/ipykernel_92990/1360805091.py:33: AbsorbingEffectWarning: \n",
      "Variables have been fully absorbed and have removed from the regression:\n",
      "\n",
      "platform_facebook, platform_instagram, platform_twitter, platform_youtube\n",
      "\n",
      "  res = mod.fit(cov_type='clustered', cluster_entity=True)\n",
      "/var/folders/pn/m1444lgd3wj53304zyp1c63w0000gq/T/ipykernel_92990/1360805091.py:33: AbsorbingEffectWarning: \n",
      "Variables have been fully absorbed and have removed from the regression:\n",
      "\n",
      "reason2_HH, reason2_MM, reason2_IV\n",
      "\n",
      "  res = mod.fit(cov_type='clustered', cluster_entity=True)\n",
      "/var/folders/pn/m1444lgd3wj53304zyp1c63w0000gq/T/ipykernel_92990/1360805091.py:33: AbsorbingEffectWarning: \n",
      "Variables have been fully absorbed and have removed from the regression:\n",
      "\n",
      "reason2_HH, reason2_MM, reason2_IV\n",
      "\n",
      "  res = mod.fit(cov_type='clustered', cluster_entity=True)\n",
      "/var/folders/pn/m1444lgd3wj53304zyp1c63w0000gq/T/ipykernel_92990/1360805091.py:33: AbsorbingEffectWarning: \n",
      "Variables have been fully absorbed and have removed from the regression:\n",
      "\n",
      "politician, media_personality, fringe_movements\n",
      "\n",
      "  res = mod.fit(cov_type='clustered', cluster_entity=True)\n",
      "/var/folders/pn/m1444lgd3wj53304zyp1c63w0000gq/T/ipykernel_92990/1360805091.py:33: AbsorbingEffectWarning: \n",
      "Variables have been fully absorbed and have removed from the regression:\n",
      "\n",
      "politician, media_personality, fringe_movements\n",
      "\n",
      "  res = mod.fit(cov_type='clustered', cluster_entity=True)\n",
      "/var/folders/pn/m1444lgd3wj53304zyp1c63w0000gq/T/ipykernel_92990/1360805091.py:33: AbsorbingEffectWarning: \n",
      "Variables have been fully absorbed and have removed from the regression:\n",
      "\n",
      "mult_bans\n",
      "\n",
      "  res = mod.fit(cov_type='clustered', cluster_entity=True)\n",
      "/var/folders/pn/m1444lgd3wj53304zyp1c63w0000gq/T/ipykernel_92990/1360805091.py:33: AbsorbingEffectWarning: \n",
      "Variables have been fully absorbed and have removed from the regression:\n",
      "\n",
      "mult_bans\n",
      "\n",
      "  res = mod.fit(cov_type='clustered', cluster_entity=True)\n"
     ]
    }
   ],
   "source": [
    "catsv = [\n",
    "    ['*isTemporary_True'],\n",
    "    [\"*high_attention\"],\n",
    "    ['*platform_facebook', '*platform_instagram', '*platform_twitter', '*platform_youtube'],\n",
    "    ['*reason2_HH', '*reason2_MM', \"*reason2_IV\"],\n",
    "    ['*politician', '*media_personality', ':internet_personality', '*fringe_movements'],\n",
    "    ['*mult_bans'],\n",
    "]\n",
    "\n",
    "sources = [\"gtab\", \"wikimedia\"]\n",
    "\n",
    "acc = []\n",
    "\n",
    "for cats in catsv:\n",
    "    for source in sources:\n",
    "        ref = -3\n",
    "        acc_form = \"logval ~ \"\n",
    "        for i in range(-12, 12):\n",
    "            if i == ref:\n",
    "                continue\n",
    "            acc_form += \" + diff_{0}:treat \".format(str(i))\n",
    "        acc_form = acc_form.replace(\"diff_-\", \"diff_m\")\n",
    "\n",
    "        for i in cats:\n",
    "            acc_form +=  \" + diff_after:treat\" + i\n",
    "        df_tuples_all_ex[\"diff_after\"] = df_tuples_all_ex[[ 'diff_'+ str(i) for i in range(12,13)]].max(axis=1)\n",
    "        \n",
    "        formula = acc_form + \"\"\" + EntityEffects + TimeEffects \"\"\"\n",
    "        mod = PanelOLS.from_formula(formula, \n",
    "                                    df_tuples_all_ex[(df_tuples_all_ex.source == source)].set_index(\n",
    "                                        [\"entity_exp\", \"months_since_2015\"]), \n",
    "                                    drop_absorbed=True)\n",
    "        res = mod.fit(cov_type='clustered', cluster_entity=True)\n",
    "        #print(cats)\n",
    "        #print(res)\n",
    "        \n",
    "        for cat in cats:\n",
    "            cat=cat.replace(\":\",\"\").replace(\"*\",\"\")\n",
    "            low = (res.params - res.std_errors  * 1.96)\n",
    "            hig = (res.params + res.std_errors  * 1.96)\n",
    "            acc.append({\n",
    "                \"cat\": cat,\n",
    "                \"source\": source,\n",
    "                \"pes\": res.params[\"diff_after:treat:\" + cat],\n",
    "                \"low\": low[\"diff_after:treat:\" + cat],\n",
    "                \"hig\": hig[\"diff_after:treat:\" + cat],\n",
    "                \"pva\": res.pvalues[\"diff_after:treat:\" + cat],\n",
    "            }\n",
    "            )\n",
    "\n",
    "acc_ = pd.DataFrame(acc)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 47,
   "id": "20252e83",
   "metadata": {
    "scrolled": false
   },
   "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>Variable</th>\n",
       "      <th>Source</th>\n",
       "      <th>Effect</th>\n",
       "      <th>$p$-value</th>\n",
       "      <th>Model</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>Temporary</td>\n",
       "      <td>G-Trends</td>\n",
       "      <td>13.2\\% 95\\% CI [-45.5\\%, 135.1\\%]</td>\n",
       "      <td>0.739</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>Temporary</td>\n",
       "      <td>Wikipedia</td>\n",
       "      <td>-5.3\\% 95\\% CI [-36.5\\%, 41.3\\%]</td>\n",
       "      <td>0.791</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>High attention</td>\n",
       "      <td>G-Trends</td>\n",
       "      <td>-7.8\\% 95\\% CI [-55.3\\%, 90.3\\%]</td>\n",
       "      <td>0.827</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>High attention</td>\n",
       "      <td>Wikipedia</td>\n",
       "      <td>-35.2\\% 95\\% CI [-53.8\\%, -9.1\\%]</td>\n",
       "      <td>0.012</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>Facebook</td>\n",
       "      <td>G-Trends</td>\n",
       "      <td>-51.9\\% 95\\% CI [-80.0\\%, 15.4\\%]</td>\n",
       "      <td>0.101</td>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>Facebook</td>\n",
       "      <td>Wikipedia</td>\n",
       "      <td>-45.3\\% 95\\% CI [-76.1\\%, 25.0\\%]</td>\n",
       "      <td>0.153</td>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>Instagram</td>\n",
       "      <td>G-Trends</td>\n",
       "      <td>-74.3\\% 95\\% CI [-91.7\\%, -20.7\\%]</td>\n",
       "      <td>0.018</td>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>Instagram</td>\n",
       "      <td>Wikipedia</td>\n",
       "      <td>-51.4\\% 95\\% CI [-72.8\\%, -13.2\\%]</td>\n",
       "      <td>0.015</td>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>Twitter</td>\n",
       "      <td>G-Trends</td>\n",
       "      <td>-71.8\\% 95\\% CI [-89.3\\%, -25.7\\%]</td>\n",
       "      <td>0.010</td>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10</th>\n",
       "      <td>Twitter</td>\n",
       "      <td>Wikipedia</td>\n",
       "      <td>-48.1\\% 95\\% CI [-79.0\\%, 28.2\\%]</td>\n",
       "      <td>0.155</td>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>YouTube</td>\n",
       "      <td>G-Trends</td>\n",
       "      <td>-48.3\\% 95\\% CI [-84.1\\%, 68.2\\%]</td>\n",
       "      <td>0.273</td>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11</th>\n",
       "      <td>YouTube</td>\n",
       "      <td>Wikipedia</td>\n",
       "      <td>13.6\\% 95\\% CI [-57.6\\%, 204.0\\%]</td>\n",
       "      <td>0.800</td>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12</th>\n",
       "      <td>Hate/Harassment</td>\n",
       "      <td>G-Trends</td>\n",
       "      <td>-13.2\\% 95\\% CI [-55.9\\%, 70.7\\%]</td>\n",
       "      <td>0.681</td>\n",
       "      <td>4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>15</th>\n",
       "      <td>Hate/Harassment</td>\n",
       "      <td>Wikipedia</td>\n",
       "      <td>-19.3\\% 95\\% CI [-46.3\\%, 21.3\\%]</td>\n",
       "      <td>0.303</td>\n",
       "      <td>4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>13</th>\n",
       "      <td>Misinformation</td>\n",
       "      <td>G-Trends</td>\n",
       "      <td>-59.8\\% 95\\% CI [-82.5\\%, -7.8\\%]</td>\n",
       "      <td>0.031</td>\n",
       "      <td>4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>16</th>\n",
       "      <td>Misinformation</td>\n",
       "      <td>Wikipedia</td>\n",
       "      <td>-48.7\\% 95\\% CI [-67.5\\%, -19.2\\%]</td>\n",
       "      <td>0.004</td>\n",
       "      <td>4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>14</th>\n",
       "      <td>Incitement to Violence</td>\n",
       "      <td>G-Trends</td>\n",
       "      <td>-39.2\\% 95\\% CI [-78.2\\%, 69.2\\%]</td>\n",
       "      <td>0.340</td>\n",
       "      <td>4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>17</th>\n",
       "      <td>Incitement to Violence</td>\n",
       "      <td>Wikipedia</td>\n",
       "      <td>-26.1\\% 95\\% CI [-62.9\\%, 47.5\\%]</td>\n",
       "      <td>0.391</td>\n",
       "      <td>4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>18</th>\n",
       "      <td>Politician</td>\n",
       "      <td>G-Trends</td>\n",
       "      <td>-5.5\\% 95\\% CI [-63.5\\%, 144.7\\%]</td>\n",
       "      <td>0.907</td>\n",
       "      <td>5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>22</th>\n",
       "      <td>Politician</td>\n",
       "      <td>Wikipedia</td>\n",
       "      <td>-0.9\\% 95\\% CI [-44.5\\%, 76.7\\%]</td>\n",
       "      <td>0.975</td>\n",
       "      <td>5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>19</th>\n",
       "      <td>Media pers.</td>\n",
       "      <td>G-Trends</td>\n",
       "      <td>32.2\\% 95\\% CI [-44.7\\%, 215.8\\%]</td>\n",
       "      <td>0.530</td>\n",
       "      <td>5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>23</th>\n",
       "      <td>Media pers.</td>\n",
       "      <td>Wikipedia</td>\n",
       "      <td>22.8\\% 95\\% CI [-24.1\\%, 98.5\\%]</td>\n",
       "      <td>0.403</td>\n",
       "      <td>5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>20</th>\n",
       "      <td>Internet pers.</td>\n",
       "      <td>G-Trends</td>\n",
       "      <td>25.1\\% 95\\% CI [-45.1\\%, 184.8\\%]</td>\n",
       "      <td>0.594</td>\n",
       "      <td>5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>24</th>\n",
       "      <td>Internet pers.</td>\n",
       "      <td>Wikipedia</td>\n",
       "      <td>35.6\\% 95\\% CI [-24.4\\%, 143.2\\%]</td>\n",
       "      <td>0.308</td>\n",
       "      <td>5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>21</th>\n",
       "      <td>Fringe mov.</td>\n",
       "      <td>G-Trends</td>\n",
       "      <td>-38.6\\% 95\\% CI [-71.8\\%, 33.8\\%]</td>\n",
       "      <td>0.220</td>\n",
       "      <td>5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25</th>\n",
       "      <td>Fringe mov.</td>\n",
       "      <td>Wikipedia</td>\n",
       "      <td>-1.4\\% 95\\% CI [-34.0\\%, 47.2\\%]</td>\n",
       "      <td>0.945</td>\n",
       "      <td>5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>26</th>\n",
       "      <td>Multiple Bans</td>\n",
       "      <td>G-Trends</td>\n",
       "      <td>-80.6\\% 95\\% CI [-94.1\\%, -36.1\\%]</td>\n",
       "      <td>0.007</td>\n",
       "      <td>6</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>27</th>\n",
       "      <td>Multiple Bans</td>\n",
       "      <td>Wikipedia</td>\n",
       "      <td>-55.5\\% 95\\% CI [-73.8\\%, -24.3\\%]</td>\n",
       "      <td>0.003</td>\n",
       "      <td>6</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                  Variable     Source                              Effect  \\\n",
       "0                Temporary   G-Trends   13.2\\% 95\\% CI [-45.5\\%, 135.1\\%]   \n",
       "1                Temporary  Wikipedia    -5.3\\% 95\\% CI [-36.5\\%, 41.3\\%]   \n",
       "2           High attention   G-Trends    -7.8\\% 95\\% CI [-55.3\\%, 90.3\\%]   \n",
       "3           High attention  Wikipedia   -35.2\\% 95\\% CI [-53.8\\%, -9.1\\%]   \n",
       "4                 Facebook   G-Trends   -51.9\\% 95\\% CI [-80.0\\%, 15.4\\%]   \n",
       "8                 Facebook  Wikipedia   -45.3\\% 95\\% CI [-76.1\\%, 25.0\\%]   \n",
       "5                Instagram   G-Trends  -74.3\\% 95\\% CI [-91.7\\%, -20.7\\%]   \n",
       "9                Instagram  Wikipedia  -51.4\\% 95\\% CI [-72.8\\%, -13.2\\%]   \n",
       "6                  Twitter   G-Trends  -71.8\\% 95\\% CI [-89.3\\%, -25.7\\%]   \n",
       "10                 Twitter  Wikipedia   -48.1\\% 95\\% CI [-79.0\\%, 28.2\\%]   \n",
       "7                  YouTube   G-Trends   -48.3\\% 95\\% CI [-84.1\\%, 68.2\\%]   \n",
       "11                 YouTube  Wikipedia   13.6\\% 95\\% CI [-57.6\\%, 204.0\\%]   \n",
       "12         Hate/Harassment   G-Trends   -13.2\\% 95\\% CI [-55.9\\%, 70.7\\%]   \n",
       "15         Hate/Harassment  Wikipedia   -19.3\\% 95\\% CI [-46.3\\%, 21.3\\%]   \n",
       "13          Misinformation   G-Trends   -59.8\\% 95\\% CI [-82.5\\%, -7.8\\%]   \n",
       "16          Misinformation  Wikipedia  -48.7\\% 95\\% CI [-67.5\\%, -19.2\\%]   \n",
       "14  Incitement to Violence   G-Trends   -39.2\\% 95\\% CI [-78.2\\%, 69.2\\%]   \n",
       "17  Incitement to Violence  Wikipedia   -26.1\\% 95\\% CI [-62.9\\%, 47.5\\%]   \n",
       "18              Politician   G-Trends   -5.5\\% 95\\% CI [-63.5\\%, 144.7\\%]   \n",
       "22              Politician  Wikipedia    -0.9\\% 95\\% CI [-44.5\\%, 76.7\\%]   \n",
       "19             Media pers.   G-Trends   32.2\\% 95\\% CI [-44.7\\%, 215.8\\%]   \n",
       "23             Media pers.  Wikipedia    22.8\\% 95\\% CI [-24.1\\%, 98.5\\%]   \n",
       "20          Internet pers.   G-Trends   25.1\\% 95\\% CI [-45.1\\%, 184.8\\%]   \n",
       "24          Internet pers.  Wikipedia   35.6\\% 95\\% CI [-24.4\\%, 143.2\\%]   \n",
       "21             Fringe mov.   G-Trends   -38.6\\% 95\\% CI [-71.8\\%, 33.8\\%]   \n",
       "25             Fringe mov.  Wikipedia    -1.4\\% 95\\% CI [-34.0\\%, 47.2\\%]   \n",
       "26           Multiple Bans   G-Trends  -80.6\\% 95\\% CI [-94.1\\%, -36.1\\%]   \n",
       "27           Multiple Bans  Wikipedia  -55.5\\% 95\\% CI [-73.8\\%, -24.3\\%]   \n",
       "\n",
       "   $p$-value Model  \n",
       "0      0.739     1  \n",
       "1      0.791     1  \n",
       "2      0.827     2  \n",
       "3      0.012     2  \n",
       "4      0.101     3  \n",
       "8      0.153     3  \n",
       "5      0.018     3  \n",
       "9      0.015     3  \n",
       "6      0.010     3  \n",
       "10     0.155     3  \n",
       "7      0.273     3  \n",
       "11     0.800     3  \n",
       "12     0.681     4  \n",
       "15     0.303     4  \n",
       "13     0.031     4  \n",
       "16     0.004     4  \n",
       "14     0.340     4  \n",
       "17     0.391     4  \n",
       "18     0.907     5  \n",
       "22     0.975     5  \n",
       "19     0.530     5  \n",
       "23     0.403     5  \n",
       "20     0.594     5  \n",
       "24     0.308     5  \n",
       "21     0.220     5  \n",
       "25     0.945     5  \n",
       "26     0.007     6  \n",
       "27     0.003     6  "
      ]
     },
     "execution_count": 47,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "mapt = {\n",
    "    'high_attention': 'High attention',\n",
    "    'low_attention': 'Low attention',\n",
    " 'isTemporary_False': 'Permanent',\n",
    " 'isTemporary_True': 'Temporary',\n",
    " 'platform_facebook': 'Facebook',\n",
    " 'platform_instagram': 'Instagram',\n",
    " 'platform_twitter': 'Twitter',\n",
    " 'platform_youtube': 'YouTube',\n",
    " 'reason2_HH': 'Hate/Harassment',\n",
    " 'reason2_MM': 'Misinformation',\n",
    " 'reason2_IV': 'Incitement to Violence',\n",
    " 'politician': 'Politician',\n",
    " 'media_personality': 'Media pers.',\n",
    " 'internet_personality': 'Internet pers.',\n",
    " 'fringe_movements': 'Fringe mov.',\n",
    " 'mult_bans': 'Multiple Bans',\n",
    "}\n",
    "\n",
    "acc = acc_.copy(deep=True)\n",
    "fround = lambda x: round(( np.exp(x)  -1 ) * 100, 1)\n",
    "acc[\"cat\"] = acc[\"cat\"].apply(lambda x: mapt[x.replace(\":\",\"\").replace(\"*\",\"\")])\n",
    "\n",
    "mapt = {'gtab': 'G-Trends',\n",
    " 'wikimedia': 'Wikipedia'}\n",
    "\n",
    "acc[\"source\"] = acc[\"source\"].apply(lambda x: mapt[x])\n",
    "\n",
    "acc[\"Effect\"] = list(zip(acc[\"pes\"], acc[\"low\"], acc[\"hig\"]))\n",
    "\n",
    "acc[\"Effect\"] = acc[\"Effect\"].apply(lambda x: \"{:.1f}\\% 95\\% CI [{:.1f}\\%, {:.1f}\\%]\".format(fround(x[0]), \n",
    "                                                              fround(x[1]),\n",
    "                                                              fround(x[2])))\n",
    "\n",
    "acc[\"$p$-value\"] = acc[\"pva\"].apply(lambda x: \"{:.3f}\".format(round(x, 3)))\n",
    "acc_f = acc[[\"cat\", \"source\", \"Effect\", \"$p$-value\"]]\n",
    "acc_f.columns = [\"Variable\", \"Source\", \"Effect\", \"$p$-value\"]\n",
    "acc_f = acc_f.iloc[[0,1,2,3,4,8,5,9,6,10,7,11,12,15,13,16,14,17,18,22,19,23,20,24,21,25,26,27]]\n",
    "acc_f[\"Model\"] = [\n",
    "    \"1\", \"1\",\n",
    "    \"2\", \"2\",\n",
    "    \"3\", \"3\", \"3\", \"3\", \"3\", \"3\", \"3\", \"3\", \n",
    "    \"4\", \"4\", \"4\", \"4\", \"4\", \"4\",\n",
    "    \"5\", \"5\", \"5\", \"5\", \"5\", \"5\", \"5\", \"5\", \n",
    "    \"6\", \"6\"\n",
    "]\n",
    "acc_f"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 48,
   "id": "471ecb5a",
   "metadata": {
    "scrolled": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\\begin{tabular}{lllll}\n",
      "\\toprule\n",
      " &  &  & Effect & $p$-value \\\\\n",
      "Model & Variable & Source &  &  \\\\\n",
      "\\midrule\n",
      "\\multirow[t]{2}{*}{1} & \\multirow[t]{2}{*}{Temporary} & G-Trends & 13.2\\% 95\\% CI [-45.5\\%, 135.1\\%] & 0.739 \\\\\n",
      " &  & Wikipedia & -5.3\\% 95\\% CI [-36.5\\%, 41.3\\%] & 0.791 \\\\\n",
      "\\midrule\n",
      "\\multirow[t]{2}{*}{2} & \\multirow[t]{2}{*}{High attention} & G-Trends & -7.8\\% 95\\% CI [-55.3\\%, 90.3\\%] & 0.827 \\\\\n",
      " &  & Wikipedia & -35.2\\% 95\\% CI [-53.8\\%, -9.1\\%] & 0.012 \\\\\n",
      "\\midrule\n",
      "\\multirow[t]{8}{*}{3} & \\multirow[t]{2}{*}{Facebook} & G-Trends & -51.9\\% 95\\% CI [-80.0\\%, 15.4\\%] & 0.101 \\\\\n",
      " &  & Wikipedia & -45.3\\% 95\\% CI [-76.1\\%, 25.0\\%] & 0.153 \\\\\n",
      "\\cline{2-5}\n",
      " & \\multirow[t]{2}{*}{Instagram} & G-Trends & -74.3\\% 95\\% CI [-91.7\\%, -20.7\\%] & 0.018 \\\\\n",
      " &  & Wikipedia & -51.4\\% 95\\% CI [-72.8\\%, -13.2\\%] & 0.015 \\\\\n",
      "\\cline{2-5}\n",
      " & \\multirow[t]{2}{*}{Twitter} & G-Trends & -71.8\\% 95\\% CI [-89.3\\%, -25.7\\%] & 0.010 \\\\\n",
      " &  & Wikipedia & -48.1\\% 95\\% CI [-79.0\\%, 28.2\\%] & 0.155 \\\\\n",
      "\\cline{2-5}\n",
      " & \\multirow[t]{2}{*}{YouTube} & G-Trends & -48.3\\% 95\\% CI [-84.1\\%, 68.2\\%] & 0.273 \\\\\n",
      " &  & Wikipedia & 13.6\\% 95\\% CI [-57.6\\%, 204.0\\%] & 0.800 \\\\\n",
      "\\midrule\n",
      "\\multirow[t]{6}{*}{4} & \\multirow[t]{2}{*}{Hate/Harassment} & G-Trends & -13.2\\% 95\\% CI [-55.9\\%, 70.7\\%] & 0.681 \\\\\n",
      " &  & Wikipedia & -19.3\\% 95\\% CI [-46.3\\%, 21.3\\%] & 0.303 \\\\\n",
      "\\cline{2-5}\n",
      " & \\multirow[t]{2}{*}{Misinformation} & G-Trends & -59.8\\% 95\\% CI [-82.5\\%, -7.8\\%] & 0.031 \\\\\n",
      " &  & Wikipedia & -48.7\\% 95\\% CI [-67.5\\%, -19.2\\%] & 0.004 \\\\\n",
      "\\cline{2-5}\n",
      " & \\multirow[t]{2}{*}{Incitement to Violence} & G-Trends & -39.2\\% 95\\% CI [-78.2\\%, 69.2\\%] & 0.340 \\\\\n",
      " &  & Wikipedia & -26.1\\% 95\\% CI [-62.9\\%, 47.5\\%] & 0.391 \\\\\n",
      "\\midrule\n",
      "\\multirow[t]{8}{*}{5} & \\multirow[t]{2}{*}{Politician} & G-Trends & -5.5\\% 95\\% CI [-63.5\\%, 144.7\\%] & 0.907 \\\\\n",
      " &  & Wikipedia & -0.9\\% 95\\% CI [-44.5\\%, 76.7\\%] & 0.975 \\\\\n",
      "\\cline{2-5}\n",
      " & \\multirow[t]{2}{*}{Media pers.} & G-Trends & 32.2\\% 95\\% CI [-44.7\\%, 215.8\\%] & 0.530 \\\\\n",
      " &  & Wikipedia & 22.8\\% 95\\% CI [-24.1\\%, 98.5\\%] & 0.403 \\\\\n",
      "\\cline{2-5}\n",
      " & \\multirow[t]{2}{*}{Internet pers.} & G-Trends & 25.1\\% 95\\% CI [-45.1\\%, 184.8\\%] & 0.594 \\\\\n",
      " &  & Wikipedia & 35.6\\% 95\\% CI [-24.4\\%, 143.2\\%] & 0.308 \\\\\n",
      "\\cline{2-5}\n",
      " & \\multirow[t]{2}{*}{Fringe mov.} & G-Trends & -38.6\\% 95\\% CI [-71.8\\%, 33.8\\%] & 0.220 \\\\\n",
      " &  & Wikipedia & -1.4\\% 95\\% CI [-34.0\\%, 47.2\\%] & 0.945 \\\\\n",
      "\\midrule\n",
      "\\multirow[t]{2}{*}{6} & \\multirow[t]{2}{*}{Multiple Bans} & G-Trends & -80.6\\% 95\\% CI [-94.1\\%, -36.1\\%] & 0.007 \\\\\n",
      " &  & Wikipedia & -55.5\\% 95\\% CI [-73.8\\%, -24.3\\%] & 0.003 \\\\\n",
      "\\bottomrule\n",
      "\\end{tabular}\n",
      "\n"
     ]
    }
   ],
   "source": [
    "print(acc_f.set_index([\"Model\", \"Variable\", \"Source\"]).to_latex(\n",
    "    ).replace(\"0.000\", \"$<$0.001\")\n",
    "     .replace(\"\\cline{1-5} \\cline{2-5}\", \"\\midrule\")\n",
    "     .replace(r'\\midrule' +'\\n' + r'\\bottomrule', r'\\bottomrule'))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 51,
   "id": "7b16b214",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th>Effect</th>\n",
       "      <th>$p$-value</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Model</th>\n",
       "      <th>Variable</th>\n",
       "      <th>Source</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th rowspan=\"2\" valign=\"top\">1</th>\n",
       "      <th rowspan=\"2\" valign=\"top\">Temporary</th>\n",
       "      <th>G-Trends</th>\n",
       "      <td>13.2\\% 95\\% CI [-45.5\\%, 135.1\\%]</td>\n",
       "      <td>0.739</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Wikipedia</th>\n",
       "      <td>-5.3\\% 95\\% CI [-36.5\\%, 41.3\\%]</td>\n",
       "      <td>0.791</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th rowspan=\"2\" valign=\"top\">2</th>\n",
       "      <th rowspan=\"2\" valign=\"top\">High attention</th>\n",
       "      <th>G-Trends</th>\n",
       "      <td>-7.8\\% 95\\% CI [-55.3\\%, 90.3\\%]</td>\n",
       "      <td>0.827</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Wikipedia</th>\n",
       "      <td>-35.2\\% 95\\% CI [-53.8\\%, -9.1\\%]</td>\n",
       "      <td>0.012</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th rowspan=\"8\" valign=\"top\">3</th>\n",
       "      <th rowspan=\"2\" valign=\"top\">Facebook</th>\n",
       "      <th>G-Trends</th>\n",
       "      <td>-51.9\\% 95\\% CI [-80.0\\%, 15.4\\%]</td>\n",
       "      <td>0.101</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Wikipedia</th>\n",
       "      <td>-45.3\\% 95\\% CI [-76.1\\%, 25.0\\%]</td>\n",
       "      <td>0.153</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th rowspan=\"2\" valign=\"top\">Instagram</th>\n",
       "      <th>G-Trends</th>\n",
       "      <td>-74.3\\% 95\\% CI [-91.7\\%, -20.7\\%]</td>\n",
       "      <td>0.018</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Wikipedia</th>\n",
       "      <td>-51.4\\% 95\\% CI [-72.8\\%, -13.2\\%]</td>\n",
       "      <td>0.015</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th rowspan=\"2\" valign=\"top\">Twitter</th>\n",
       "      <th>G-Trends</th>\n",
       "      <td>-71.8\\% 95\\% CI [-89.3\\%, -25.7\\%]</td>\n",
       "      <td>0.010</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Wikipedia</th>\n",
       "      <td>-48.1\\% 95\\% CI [-79.0\\%, 28.2\\%]</td>\n",
       "      <td>0.155</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th rowspan=\"2\" valign=\"top\">YouTube</th>\n",
       "      <th>G-Trends</th>\n",
       "      <td>-48.3\\% 95\\% CI [-84.1\\%, 68.2\\%]</td>\n",
       "      <td>0.273</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Wikipedia</th>\n",
       "      <td>13.6\\% 95\\% CI [-57.6\\%, 204.0\\%]</td>\n",
       "      <td>0.800</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th rowspan=\"6\" valign=\"top\">4</th>\n",
       "      <th rowspan=\"2\" valign=\"top\">Hate/Harassment</th>\n",
       "      <th>G-Trends</th>\n",
       "      <td>-13.2\\% 95\\% CI [-55.9\\%, 70.7\\%]</td>\n",
       "      <td>0.681</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Wikipedia</th>\n",
       "      <td>-19.3\\% 95\\% CI [-46.3\\%, 21.3\\%]</td>\n",
       "      <td>0.303</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th rowspan=\"2\" valign=\"top\">Misinformation</th>\n",
       "      <th>G-Trends</th>\n",
       "      <td>-59.8\\% 95\\% CI [-82.5\\%, -7.8\\%]</td>\n",
       "      <td>0.031</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Wikipedia</th>\n",
       "      <td>-48.7\\% 95\\% CI [-67.5\\%, -19.2\\%]</td>\n",
       "      <td>0.004</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th rowspan=\"2\" valign=\"top\">Incitement to Violence</th>\n",
       "      <th>G-Trends</th>\n",
       "      <td>-39.2\\% 95\\% CI [-78.2\\%, 69.2\\%]</td>\n",
       "      <td>0.340</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Wikipedia</th>\n",
       "      <td>-26.1\\% 95\\% CI [-62.9\\%, 47.5\\%]</td>\n",
       "      <td>0.391</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th rowspan=\"8\" valign=\"top\">5</th>\n",
       "      <th rowspan=\"2\" valign=\"top\">Politician</th>\n",
       "      <th>G-Trends</th>\n",
       "      <td>-5.5\\% 95\\% CI [-63.5\\%, 144.7\\%]</td>\n",
       "      <td>0.907</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Wikipedia</th>\n",
       "      <td>-0.9\\% 95\\% CI [-44.5\\%, 76.7\\%]</td>\n",
       "      <td>0.975</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th rowspan=\"2\" valign=\"top\">Media pers.</th>\n",
       "      <th>G-Trends</th>\n",
       "      <td>32.2\\% 95\\% CI [-44.7\\%, 215.8\\%]</td>\n",
       "      <td>0.530</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Wikipedia</th>\n",
       "      <td>22.8\\% 95\\% CI [-24.1\\%, 98.5\\%]</td>\n",
       "      <td>0.403</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th rowspan=\"2\" valign=\"top\">Internet pers.</th>\n",
       "      <th>G-Trends</th>\n",
       "      <td>25.1\\% 95\\% CI [-45.1\\%, 184.8\\%]</td>\n",
       "      <td>0.594</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Wikipedia</th>\n",
       "      <td>35.6\\% 95\\% CI [-24.4\\%, 143.2\\%]</td>\n",
       "      <td>0.308</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th rowspan=\"2\" valign=\"top\">Fringe mov.</th>\n",
       "      <th>G-Trends</th>\n",
       "      <td>-38.6\\% 95\\% CI [-71.8\\%, 33.8\\%]</td>\n",
       "      <td>0.220</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Wikipedia</th>\n",
       "      <td>-1.4\\% 95\\% CI [-34.0\\%, 47.2\\%]</td>\n",
       "      <td>0.945</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th rowspan=\"2\" valign=\"top\">6</th>\n",
       "      <th rowspan=\"2\" valign=\"top\">Multiple Bans</th>\n",
       "      <th>G-Trends</th>\n",
       "      <td>-80.6\\% 95\\% CI [-94.1\\%, -36.1\\%]</td>\n",
       "      <td>0.007</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Wikipedia</th>\n",
       "      <td>-55.5\\% 95\\% CI [-73.8\\%, -24.3\\%]</td>\n",
       "      <td>0.003</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                                                    Effect  \\\n",
       "Model Variable               Source                                          \n",
       "1     Temporary              G-Trends    13.2\\% 95\\% CI [-45.5\\%, 135.1\\%]   \n",
       "                             Wikipedia    -5.3\\% 95\\% CI [-36.5\\%, 41.3\\%]   \n",
       "2     High attention         G-Trends     -7.8\\% 95\\% CI [-55.3\\%, 90.3\\%]   \n",
       "                             Wikipedia   -35.2\\% 95\\% CI [-53.8\\%, -9.1\\%]   \n",
       "3     Facebook               G-Trends    -51.9\\% 95\\% CI [-80.0\\%, 15.4\\%]   \n",
       "                             Wikipedia   -45.3\\% 95\\% CI [-76.1\\%, 25.0\\%]   \n",
       "      Instagram              G-Trends   -74.3\\% 95\\% CI [-91.7\\%, -20.7\\%]   \n",
       "                             Wikipedia  -51.4\\% 95\\% CI [-72.8\\%, -13.2\\%]   \n",
       "      Twitter                G-Trends   -71.8\\% 95\\% CI [-89.3\\%, -25.7\\%]   \n",
       "                             Wikipedia   -48.1\\% 95\\% CI [-79.0\\%, 28.2\\%]   \n",
       "      YouTube                G-Trends    -48.3\\% 95\\% CI [-84.1\\%, 68.2\\%]   \n",
       "                             Wikipedia   13.6\\% 95\\% CI [-57.6\\%, 204.0\\%]   \n",
       "4     Hate/Harassment        G-Trends    -13.2\\% 95\\% CI [-55.9\\%, 70.7\\%]   \n",
       "                             Wikipedia   -19.3\\% 95\\% CI [-46.3\\%, 21.3\\%]   \n",
       "      Misinformation         G-Trends    -59.8\\% 95\\% CI [-82.5\\%, -7.8\\%]   \n",
       "                             Wikipedia  -48.7\\% 95\\% CI [-67.5\\%, -19.2\\%]   \n",
       "      Incitement to Violence G-Trends    -39.2\\% 95\\% CI [-78.2\\%, 69.2\\%]   \n",
       "                             Wikipedia   -26.1\\% 95\\% CI [-62.9\\%, 47.5\\%]   \n",
       "5     Politician             G-Trends    -5.5\\% 95\\% CI [-63.5\\%, 144.7\\%]   \n",
       "                             Wikipedia    -0.9\\% 95\\% CI [-44.5\\%, 76.7\\%]   \n",
       "      Media pers.            G-Trends    32.2\\% 95\\% CI [-44.7\\%, 215.8\\%]   \n",
       "                             Wikipedia    22.8\\% 95\\% CI [-24.1\\%, 98.5\\%]   \n",
       "      Internet pers.         G-Trends    25.1\\% 95\\% CI [-45.1\\%, 184.8\\%]   \n",
       "                             Wikipedia   35.6\\% 95\\% CI [-24.4\\%, 143.2\\%]   \n",
       "      Fringe mov.            G-Trends    -38.6\\% 95\\% CI [-71.8\\%, 33.8\\%]   \n",
       "                             Wikipedia    -1.4\\% 95\\% CI [-34.0\\%, 47.2\\%]   \n",
       "6     Multiple Bans          G-Trends   -80.6\\% 95\\% CI [-94.1\\%, -36.1\\%]   \n",
       "                             Wikipedia  -55.5\\% 95\\% CI [-73.8\\%, -24.3\\%]   \n",
       "\n",
       "                                       $p$-value  \n",
       "Model Variable               Source               \n",
       "1     Temporary              G-Trends      0.739  \n",
       "                             Wikipedia     0.791  \n",
       "2     High attention         G-Trends      0.827  \n",
       "                             Wikipedia     0.012  \n",
       "3     Facebook               G-Trends      0.101  \n",
       "                             Wikipedia     0.153  \n",
       "      Instagram              G-Trends      0.018  \n",
       "                             Wikipedia     0.015  \n",
       "      Twitter                G-Trends      0.010  \n",
       "                             Wikipedia     0.155  \n",
       "      YouTube                G-Trends      0.273  \n",
       "                             Wikipedia     0.800  \n",
       "4     Hate/Harassment        G-Trends      0.681  \n",
       "                             Wikipedia     0.303  \n",
       "      Misinformation         G-Trends      0.031  \n",
       "                             Wikipedia     0.004  \n",
       "      Incitement to Violence G-Trends      0.340  \n",
       "                             Wikipedia     0.391  \n",
       "5     Politician             G-Trends      0.907  \n",
       "                             Wikipedia     0.975  \n",
       "      Media pers.            G-Trends      0.530  \n",
       "                             Wikipedia     0.403  \n",
       "      Internet pers.         G-Trends      0.594  \n",
       "                             Wikipedia     0.308  \n",
       "      Fringe mov.            G-Trends      0.220  \n",
       "                             Wikipedia     0.945  \n",
       "6     Multiple Bans          G-Trends      0.007  \n",
       "                             Wikipedia     0.003  "
      ]
     },
     "execution_count": 51,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "acc_f.set_index([\"Model\", \"Variable\", \"Source\"])."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "c9e5e3ca",
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "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": "3.11.4"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}