notebooks/03-final-dataset-analysis.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "# enable automatic reloading of the notebook\n",
    "%load_ext autoreload\n",
    "%autoreload 2"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "import ast\n",
    "import numpy as np\n",
    "import pandas as pd\n",
    "\n",
    "from src.utils.NewsArticle import NewsArticle\n",
    "from src.utils.NewsEvent import NewsEventBase"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "##### Constants"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "DATASET_FILE_PATH = \"../data/final/og2021.csv\""
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "##### Helper functions"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "def literal_converter(val):\n",
    "    try:\n",
    "        return ast.literal_eval(val)\n",
    "    except Exception:\n",
    "        return val"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "def load_dataset(input_file):\n",
    "    df = pd.read_csv(\n",
    "        input_file,\n",
    "        names=[\n",
    "            \"id\",\n",
    "            \"title\",\n",
    "            \"body\",\n",
    "            \"lang\",\n",
    "            \"source\",\n",
    "            \"dateTime\",\n",
    "            \"url\",\n",
    "            \"uri\",\n",
    "            \"eventUri\",\n",
    "            \"concepts\",\n",
    "            \"clusterId\",\n",
    "        ],\n",
    "        dtype={\n",
    "            \"title\": \"str\",\n",
    "            \"body\": \"str\",\n",
    "            \"lang\": \"str\",\n",
    "            \"source\": \"str\",\n",
    "            \"dateTime\": \"str\",\n",
    "            \"url\": \"str\",\n",
    "            \"uri\": \"str\",\n",
    "            \"eventUri\": \"str\",\n",
    "            \"concepts\": \"str\",\n",
    "            \"clusterId\": \"str\",\n",
    "            \"id\": \"Int64\",\n",
    "        },\n",
    "        parse_dates=[\"dateTime\"],\n",
    "        on_bad_lines=\"warn\",\n",
    "        engine=\"python\",\n",
    "        skiprows=1,\n",
    "    )\n",
    "\n",
    "    return df"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Load OG2021 dataset"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "label_df = load_dataset(DATASET_FILE_PATH)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>id</th>\n",
       "      <th>title</th>\n",
       "      <th>body</th>\n",
       "      <th>lang</th>\n",
       "      <th>source</th>\n",
       "      <th>dateTime</th>\n",
       "      <th>url</th>\n",
       "      <th>uri</th>\n",
       "      <th>eventUri</th>\n",
       "      <th>concepts</th>\n",
       "      <th>clusterId</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0</td>\n",
       "      <td>Iran's Basketball Team Held At Airport Over 'U...</td>\n",
       "      <td>Incident comes weeks ahead of Tokyo Olympics, ...</td>\n",
       "      <td>eng</td>\n",
       "      <td>Haberler.com</td>\n",
       "      <td>2021-07-01 06:01:00</td>\n",
       "      <td>https://en.haberler.com/iran-s-basketball-team...</td>\n",
       "      <td>6628239427</td>\n",
       "      <td>NaN</td>\n",
       "      <td>['olympic_games', 'japan', 'basketball']</td>\n",
       "      <td>wn-1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1</td>\n",
       "      <td>Olympics latest: Samoa participation in doubt ...</td>\n",
       "      <td>TOKYO -- The July 23 opening ceremony of the d...</td>\n",
       "      <td>eng</td>\n",
       "      <td>Nikkei Asia</td>\n",
       "      <td>2021-07-01 08:51:00</td>\n",
       "      <td>https://asia.nikkei.com/Spotlight/Tokyo-2020-O...</td>\n",
       "      <td>6628362266</td>\n",
       "      <td>eng-6899867</td>\n",
       "      <td>['olympic_games', 'japan', 'basketball']</td>\n",
       "      <td>wn-2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>2</td>\n",
       "      <td>What Do the Olympics Have Against Women?</td>\n",
       "      <td>Yes, I am talking about how Laurel Hubbard, a ...</td>\n",
       "      <td>eng</td>\n",
       "      <td>Townhall</td>\n",
       "      <td>2021-07-01 08:54:00</td>\n",
       "      <td>https://townhall.com/tipsheet/rebeccadowns/202...</td>\n",
       "      <td>6628364007</td>\n",
       "      <td>eng-6906275</td>\n",
       "      <td>['olympic_games', 'japan', 'basketball']</td>\n",
       "      <td>wn-3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>3</td>\n",
       "      <td>Breastfeeding Mother Granted Permission To Bri...</td>\n",
       "      <td>Breastfeeding mothers competing at the Olympic...</td>\n",
       "      <td>eng</td>\n",
       "      <td>www.sportbible.com</td>\n",
       "      <td>2021-07-01 10:10:00</td>\n",
       "      <td>https://www.sportbible.com/australia/news-brea...</td>\n",
       "      <td>6628423216</td>\n",
       "      <td>eng-6906275</td>\n",
       "      <td>['olympic_games', 'japan', 'basketball']</td>\n",
       "      <td>wn-4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>4</td>\n",
       "      <td>Sports News | Breastfeeding Olympians Allowed ...</td>\n",
       "      <td>Canadian basketball player Kim Gaucher, who is...</td>\n",
       "      <td>eng</td>\n",
       "      <td>LatestLY</td>\n",
       "      <td>2021-07-01 10:20:00</td>\n",
       "      <td>https://www.latestly.com/agency-news/sports-ne...</td>\n",
       "      <td>6628431262</td>\n",
       "      <td>eng-6906275</td>\n",
       "      <td>['olympic_games', 'japan', 'basketball']</td>\n",
       "      <td>wn-4</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",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10935</th>\n",
       "      <td>10935</td>\n",
       "      <td>Juegos Olímpicos ganaron aceptación entre much...</td>\n",
       "      <td>Antes, muchos japoneses mostraron su reticenci...</td>\n",
       "      <td>spa</td>\n",
       "      <td>www.diariolibre.com</td>\n",
       "      <td>2021-08-13 18:36:00</td>\n",
       "      <td>https://www.diariolibre.com/deportes/olimpismo...</td>\n",
       "      <td>6686899243</td>\n",
       "      <td>spa-2454564</td>\n",
       "      <td>['olympic_games', 'japan', 'judo']</td>\n",
       "      <td>wn-1347</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10936</th>\n",
       "      <td>10936</td>\n",
       "      <td>Japón, al final, se dejó 'seducir' por los Jue...</td>\n",
       "      <td>Por ahora, hay en muchos japoneses un sentimie...</td>\n",
       "      <td>spa</td>\n",
       "      <td>El Financiero</td>\n",
       "      <td>2021-08-13 21:09:00</td>\n",
       "      <td>https://www.elfinanciero.com.mx/tokio-2020/202...</td>\n",
       "      <td>6687117130</td>\n",
       "      <td>spa-2454564</td>\n",
       "      <td>['olympic_games', 'japan', 'judo']</td>\n",
       "      <td>wn-1347</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10937</th>\n",
       "      <td>10937</td>\n",
       "      <td>Películas esenciales por si te clavaste con el...</td>\n",
       "      <td>La brasileña Rayssa Leal se robó el corazón de...</td>\n",
       "      <td>spa</td>\n",
       "      <td>Chilango</td>\n",
       "      <td>2021-08-13 23:08:00</td>\n",
       "      <td>https://www.chilango.com/cine-y-tv/peliculas-e...</td>\n",
       "      <td>6687247369</td>\n",
       "      <td>NaN</td>\n",
       "      <td>['olympic_games', 'japan', 'skateboarding']</td>\n",
       "      <td>wn-1350</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10938</th>\n",
       "      <td>10938</td>\n",
       "      <td>El alcalde japonés que le tocó ofrecer disculp...</td>\n",
       "      <td>Por una escena transmitida por televisión, que...</td>\n",
       "      <td>spa</td>\n",
       "      <td>Récord</td>\n",
       "      <td>2021-08-14 02:47:00</td>\n",
       "      <td>https://record.acento.com.do/polideportivo/el-...</td>\n",
       "      <td>6687444123</td>\n",
       "      <td>spa-2455119</td>\n",
       "      <td>['olympic_games', 'japan', 'judo']</td>\n",
       "      <td>wn-1349</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10939</th>\n",
       "      <td>10939</td>\n",
       "      <td>Juegos Olímpicos fueron aceptados por muchos j...</td>\n",
       "      <td>En los momentos en que comenzaban los Juegos O...</td>\n",
       "      <td>spa</td>\n",
       "      <td>El Informador Venezuela</td>\n",
       "      <td>2021-08-14 04:25:00</td>\n",
       "      <td>https://www.elinformadorvenezuela.com/deportes...</td>\n",
       "      <td>6687517349</td>\n",
       "      <td>spa-2454564</td>\n",
       "      <td>['olympic_games', 'japan', 'judo']</td>\n",
       "      <td>wn-1347</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>10940 rows × 11 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "          id                                              title  \\\n",
       "0          0  Iran's Basketball Team Held At Airport Over 'U...   \n",
       "1          1  Olympics latest: Samoa participation in doubt ...   \n",
       "2          2           What Do the Olympics Have Against Women?   \n",
       "3          3  Breastfeeding Mother Granted Permission To Bri...   \n",
       "4          4  Sports News | Breastfeeding Olympians Allowed ...   \n",
       "...      ...                                                ...   \n",
       "10935  10935  Juegos Olímpicos ganaron aceptación entre much...   \n",
       "10936  10936  Japón, al final, se dejó 'seducir' por los Jue...   \n",
       "10937  10937  Películas esenciales por si te clavaste con el...   \n",
       "10938  10938  El alcalde japonés que le tocó ofrecer disculp...   \n",
       "10939  10939  Juegos Olímpicos fueron aceptados por muchos j...   \n",
       "\n",
       "                                                    body lang  \\\n",
       "0      Incident comes weeks ahead of Tokyo Olympics, ...  eng   \n",
       "1      TOKYO -- The July 23 opening ceremony of the d...  eng   \n",
       "2      Yes, I am talking about how Laurel Hubbard, a ...  eng   \n",
       "3      Breastfeeding mothers competing at the Olympic...  eng   \n",
       "4      Canadian basketball player Kim Gaucher, who is...  eng   \n",
       "...                                                  ...  ...   \n",
       "10935  Antes, muchos japoneses mostraron su reticenci...  spa   \n",
       "10936  Por ahora, hay en muchos japoneses un sentimie...  spa   \n",
       "10937  La brasileña Rayssa Leal se robó el corazón de...  spa   \n",
       "10938  Por una escena transmitida por televisión, que...  spa   \n",
       "10939  En los momentos en que comenzaban los Juegos O...  spa   \n",
       "\n",
       "                        source            dateTime  \\\n",
       "0                 Haberler.com 2021-07-01 06:01:00   \n",
       "1                  Nikkei Asia 2021-07-01 08:51:00   \n",
       "2                     Townhall 2021-07-01 08:54:00   \n",
       "3           www.sportbible.com 2021-07-01 10:10:00   \n",
       "4                     LatestLY 2021-07-01 10:20:00   \n",
       "...                        ...                 ...   \n",
       "10935      www.diariolibre.com 2021-08-13 18:36:00   \n",
       "10936            El Financiero 2021-08-13 21:09:00   \n",
       "10937                 Chilango 2021-08-13 23:08:00   \n",
       "10938                   Récord 2021-08-14 02:47:00   \n",
       "10939  El Informador Venezuela 2021-08-14 04:25:00   \n",
       "\n",
       "                                                     url         uri  \\\n",
       "0      https://en.haberler.com/iran-s-basketball-team...  6628239427   \n",
       "1      https://asia.nikkei.com/Spotlight/Tokyo-2020-O...  6628362266   \n",
       "2      https://townhall.com/tipsheet/rebeccadowns/202...  6628364007   \n",
       "3      https://www.sportbible.com/australia/news-brea...  6628423216   \n",
       "4      https://www.latestly.com/agency-news/sports-ne...  6628431262   \n",
       "...                                                  ...         ...   \n",
       "10935  https://www.diariolibre.com/deportes/olimpismo...  6686899243   \n",
       "10936  https://www.elfinanciero.com.mx/tokio-2020/202...  6687117130   \n",
       "10937  https://www.chilango.com/cine-y-tv/peliculas-e...  6687247369   \n",
       "10938  https://record.acento.com.do/polideportivo/el-...  6687444123   \n",
       "10939  https://www.elinformadorvenezuela.com/deportes...  6687517349   \n",
       "\n",
       "          eventUri                                     concepts clusterId  \n",
       "0              NaN     ['olympic_games', 'japan', 'basketball']      wn-1  \n",
       "1      eng-6899867     ['olympic_games', 'japan', 'basketball']      wn-2  \n",
       "2      eng-6906275     ['olympic_games', 'japan', 'basketball']      wn-3  \n",
       "3      eng-6906275     ['olympic_games', 'japan', 'basketball']      wn-4  \n",
       "4      eng-6906275     ['olympic_games', 'japan', 'basketball']      wn-4  \n",
       "...            ...                                          ...       ...  \n",
       "10935  spa-2454564           ['olympic_games', 'japan', 'judo']   wn-1347  \n",
       "10936  spa-2454564           ['olympic_games', 'japan', 'judo']   wn-1347  \n",
       "10937          NaN  ['olympic_games', 'japan', 'skateboarding']   wn-1350  \n",
       "10938  spa-2455119           ['olympic_games', 'japan', 'judo']   wn-1349  \n",
       "10939  spa-2454564           ['olympic_games', 'japan', 'judo']   wn-1347  \n",
       "\n",
       "[10940 rows x 11 columns]"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "label_df"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Dataset analysis"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [],
   "source": [
    "from datachart import bar_chart, heatmap\n",
    "from datachart.constants import FIG_SIZE, FIG_FORMAT, FONT_WEIGHT, FONT_STYLE, LINE_STYLE, LINE_DRAW_STYLE, NORMALIZE, COLORS\n",
    "from datachart.config import config\n",
    "from datachart.utils import save_figure"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [],
   "source": [
    "THEME_CONFIG = {\n",
    "    # general color style\n",
    "    \"color.general.singular\": COLORS.Blue,\n",
    "    \"color.general.multiple\": COLORS.Spectral,\n",
    "    # general font style\n",
    "    \"font.general.family\": \"sans-serif\",\n",
    "    \"font.general.sans-serif\": [\"Inter\"],\n",
    "    \"font.general.color\": \"#06102B\",\n",
    "    \"font.general.size\": 11,\n",
    "    \"font.general.style\": FONT_STYLE.NORMAL,\n",
    "    \"font.general.weight\": FONT_WEIGHT.NORMAL,\n",
    "    # title size style\n",
    "    \"font.title.size\": 12,\n",
    "    \"font.title.color\": \"#06102B\",\n",
    "    \"font.title.style\": FONT_STYLE.NORMAL,\n",
    "    \"font.title.weight\": FONT_WEIGHT.NORMAL,\n",
    "    # subtitle size style\n",
    "    \"font.subtitle.size\": 11,\n",
    "    \"font.subtitle.color\": \"#06102B\",\n",
    "    \"font.subtitle.style\": FONT_STYLE.NORMAL,\n",
    "    \"font.subtitle.weight\": FONT_WEIGHT.NORMAL,\n",
    "    # xlabel size style\n",
    "    \"font.xlabel.size\": 10,\n",
    "    \"font.xlabel.color\": \"#06102B\",\n",
    "    \"font.xlabel.style\": FONT_STYLE.NORMAL,\n",
    "    \"font.xlabel.weight\": FONT_WEIGHT.NORMAL,\n",
    "    # ylabel size style\n",
    "    \"font.ylabel.size\": 10,\n",
    "    \"font.ylabel.color\": \"#06102B\",\n",
    "    \"font.ylabel.style\": FONT_STYLE.NORMAL,\n",
    "    \"font.ylabel.weight\": FONT_WEIGHT.NORMAL,\n",
    "    # plot axes style\n",
    "    \"axes.spines.top.visible\": True,\n",
    "    \"axes.spines.right.visible\": True,\n",
    "    \"axes.spines.bottom.visible\": True,\n",
    "    \"axes.spines.left.visible\": True,\n",
    "    \"axes.spines.width\": 0.5,\n",
    "    \"axes.spines.zorder\": 100,\n",
    "    \"axes.ticks.length\": 2,\n",
    "    \"axes.ticks.label.size\": 9,\n",
    "    # plot legend style\n",
    "    \"plot.legend.shadow\": False,\n",
    "    \"plot.legend.frameon\": True,\n",
    "    \"plot.legend.alignment\": \"left\",\n",
    "    \"plot.legend.font.size\": 9,\n",
    "    \"plot.legend.title.size\": 10,\n",
    "    \"plot.legend.label.color\": \"#06102B\",\n",
    "    # plot area style\n",
    "    \"plot.area.alpha\": 0.3,\n",
    "    \"plot.area.color\": \"#4D69B2\",\n",
    "    \"plot.area.line.width\": 0,\n",
    "    \"plot.area.hatch\": None,\n",
    "    \"plot.area.zorder\": 3,\n",
    "    # plot grid style\n",
    "    \"plot.grid.alpha\": 1,\n",
    "    \"plot.grid.color\": \"#E6E6E6\",\n",
    "    \"plot.grid.line.width\": 0.5,\n",
    "    \"plot.grid.line.style\": LINE_STYLE.SOLID,\n",
    "    \"plot.grid.zorder\": 0,\n",
    "    # plot line style\n",
    "    \"plot.line.color\": \"#4D69B2\",\n",
    "    \"plot.line.style\": LINE_STYLE.SOLID,\n",
    "    \"plot.line.marker\": None,\n",
    "    \"plot.line.width\": 1,\n",
    "    \"plot.line.alpha\": 1.0,\n",
    "    \"plot.line.drawstyle\": LINE_DRAW_STYLE.DEFAULT,\n",
    "    \"plot.line.zorder\": 3,\n",
    "    # plot bar style\n",
    "    \"plot.bar.color\": \"#4D69B2\",\n",
    "    \"plot.bar.alpha\": 1.0,\n",
    "    \"plot.bar.width\": 0.8,\n",
    "    \"plot.bar.zorder\": 3,\n",
    "    \"plot.bar.hatch\": None,\n",
    "    \"plot.bar.edge.width\": 0.5,\n",
    "    \"plot.bar.edge.color\": \"#06102B\",\n",
    "    \"plot.bar.error.color\": \"#06102B\",\n",
    "    # plot hist style\n",
    "    \"plot.hist.color\": \"#4D69B2\",\n",
    "    \"plot.hist.alpha\": 1.0,\n",
    "    \"plot.hist.zorder\": 3,\n",
    "    \"plot.hist.fill\": None,\n",
    "    \"plot.hist.hatch\": None,\n",
    "    \"plot.hist.type\": \"bar\",\n",
    "    \"plot.hist.align\": \"mid\",\n",
    "    \"plot.hist.edge.width\": 0.5,\n",
    "    \"plot.hist.edge.color\": \"#06102B\",\n",
    "    # plot vline style\n",
    "    \"plot.vline.color\": \"#2FDD92\",\n",
    "    \"plot.vline.style\": LINE_STYLE.SOLID,\n",
    "    \"plot.vline.width\": 1,\n",
    "    \"plot.vline.alpha\": 1.0,\n",
    "    # plot hline style\n",
    "    \"plot.hline.color\": \"#2FDD92\",\n",
    "    \"plot.hline.style\": LINE_STYLE.SOLID,\n",
    "    \"plot.hline.width\": 1,\n",
    "    \"plot.hline.alpha\": 1.0,\n",
    "    # plot heatmap style\n",
    "    \"plot.heatmap.cmap\": COLORS.GnBu,\n",
    "    \"plot.heatmap.alpha\": 1.0,\n",
    "    \"plot.heatmap.font.size\": 10,\n",
    "    \"plot.heatmap.font.color\": \"#06102B\",\n",
    "    \"plot.heatmap.font.style\": FONT_STYLE.NORMAL,\n",
    "    \"plot.heatmap.font.weight\": FONT_WEIGHT.NORMAL,\n",
    "}"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [],
   "source": [
    "config.update_config(config=THEME_CONFIG)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [],
   "source": [
    "def get_article_group_count(groups, label_func=lambda x: x):\n",
    "    return [{\"label\": label_func(group), \"y\": len(values)} for group, values in groups.items()]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Article Analysis"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [],
   "source": [
    "iso_to_lang_mapping = {\n",
    "    \"eng\": \"English\",\n",
    "    \"por\": \"Portuguese\",\n",
    "    \"spa\": \"Spanish\",\n",
    "    \"fra\": \"French\",\n",
    "    \"rus\": \"Russian\",\n",
    "    \"deu\": \"German\",\n",
    "    \"slv\": \"Slovenian\",\n",
    "    \"ara\": \"Arabic\",\n",
    "    \"zho\": \"Chinese\",\n",
    "    #\"hrv\": \"Croatian\",\n",
    "    #\"ita\": \"Italian\",\n",
    "}\n",
    "def get_lang(x):\n",
    "    return iso_to_lang_mapping.get(x, \"Undefined\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Temporal Analysis"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [],
   "source": [
    "def format_date(x):\n",
    "    def beautify(x):\n",
    "        return f\"0{x}\"[-2:]\n",
    "    return \"{}-{}-{}\".format(x.year, beautify(x.month), beautify(x.day))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [],
   "source": [
    "label_df[\"date\"] = label_df[\"dateTime\"].apply(lambda x: format_date(x))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [],
   "source": [
    "count_by_date = get_article_group_count(label_df.groupby([\"date\"]).groups)\n",
    "xlabels = [x[\"label\"] for x in count_by_date]\n",
    "xticks = [0, 21, 38, len(xlabels)-1]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 820x240 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "figure = bar_chart(\n",
    "    {\n",
    "        \"charts\": {\n",
    "            \"data\": count_by_date,\n",
    "            \"xticks\": xticks,\n",
    "            \"xticklabels\": [xlabels[idx] for idx in xticks],\n",
    "            \"xtickrotate\": 0,\n",
    "            \"vlines\": [\n",
    "                {\n",
    "                    \"x\": 21,\n",
    "                    \"ymax\": 1000,\n",
    "                    \"label\": \"start of the Olympics\",\n",
    "                },\n",
    "                {\n",
    "                    \"x\": 38,\n",
    "                    \"ymax\": 1000,\n",
    "                    \"label\": \"end of the Olympics\",\n",
    "                },\n",
    "            ],\n",
    "        },\n",
    "        \"ymax\": 1000,\n",
    "        \"title\": \"article distribution by date\",\n",
    "        \"xlabel\": \"date\",\n",
    "        \"ylabel\": \"number of articles\",\n",
    "        \"show_legend\": True,\n",
    "        \"show_grid\": \"y\",\n",
    "        \"figsize\": FIG_SIZE.A4_NARROW,\n",
    "    }\n",
    ")\n",
    "save_figure(figure, \"../results/og2021-article-distribution-by-date.png\", dpi=300, format=FIG_FORMAT.PNG, transparent=True)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Language Analysis "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [],
   "source": [
    "count_by_lang = get_article_group_count(label_df.groupby([\"lang\"]).groups, label_func=get_lang)\n",
    "count_by_lang = sorted(count_by_lang, key=lambda x: x[\"y\"], reverse=True)\n",
    "\n",
    "avg_count = sum([v[\"y\"] for v in count_by_lang]) / len(count_by_lang)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 820x240 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "figure = bar_chart({\n",
    "    \"charts\": {\n",
    "        \"data\": count_by_lang,\n",
    "        \"hlines\": [\n",
    "                {\n",
    "                    \"y\": avg_count,\n",
    "                    \"xmin\": -0.5,\n",
    "                    \"xmax\": len(iso_to_lang_mapping.keys()) - 0.5,\n",
    "                    \"label\": \"avg number of articles\",\n",
    "                },\n",
    "            ],\n",
    "    },\n",
    "    \"title\": \"article distribution by language\",\n",
    "    \"xlabel\": \"language\",\n",
    "    \"ylabel\": \"number of articles\",\n",
    "    \"xmin\": -0.5,\n",
    "    \"xmax\": len(iso_to_lang_mapping.keys()) - 0.5,\n",
    "    \"show_grid\": \"y\",\n",
    "    \"show_legend\": True,\n",
    "    \"figsize\": FIG_SIZE.A4_NARROW # (4.1, 2.4) # half the width of the A4\n",
    "})\n",
    "save_figure(figure, \"../results/og2021-article-distribution-by-lang.png\", dpi=300, format=FIG_FORMAT.PNG, transparent=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [],
   "source": [
    "def get_length_stats(df, column):\n",
    "    vals = [len(x.split()) for x in df[column]]\n",
    "    return {\n",
    "        \"min\": min(vals) if len(vals) > 0 else 0,\n",
    "        \"max\": max(vals) if len(vals) > 0 else 0,\n",
    "        \"avg\": sum(vals) / len(vals) if len(vals) > 0 else 0,\n",
    "        \"std\": np.std(vals) if len(vals) > 0 else 0,\n",
    "    }\n",
    "\n",
    "def get_cluster_stats(df, column):\n",
    "    vals = [len(v) for v in df.groupby([column]).groups.values()]\n",
    "    return {\n",
    "        \"count\": len(vals) if len(vals) > 0 else 0,\n",
    "        \"min\": min(vals) if len(vals) > 0 else 0,\n",
    "        \"max\": max(vals) if len(vals) > 0 else 0,\n",
    "        \"avg\": sum(vals) / len(vals) if len(vals) > 0 else 0,\n",
    "        \"std\": np.std(vals) if len(vals) > 0 else 0,\n",
    "    }"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Min, mean, and max article title and body lengths (in characters) by language:\n",
      "Language   & Arts   & tavg (tstd) & bavg  (bstd ) & cls   & cavg (cstd) \\\\\n",
      "Overall    & 10,940 & 12   (4   ) & 792   (862  ) & 1,350 & 8    (20  ) \\\\\n",
      "English    & 4,009  & 11   (3   ) & 1,231 (1,147) & 729   & 5    (11  ) \\\\\n",
      "Portuguese & 2,410  & 13   (3   ) & 527   (374  ) & 368   & 7    (12  ) \\\\\n",
      "Spanish    & 2,049  & 13   (4   ) & 562   (427  ) & 381   & 5    (9   ) \\\\\n",
      "French     & 845    & 13   (4   ) & 565   (464  ) & 170   & 5    (8   ) \\\\\n",
      "Russian    & 553    & 10   (3   ) & 301   (358  ) & 152   & 4    (6   ) \\\\\n",
      "German     & 516    & 9    (3   ) & 833   (1,011) & 100   & 5    (6   ) \\\\\n",
      "Slovenian  & 331    & 9    (3   ) & 450   (370  ) & 102   & 3    (3   ) \\\\\n",
      "Arabic     & 218    & 10   (3   ) & 405   (269  ) & 71    & 3    (3   ) \\\\\n",
      "Chinese    & 9      & 2    (1   ) & 56    (36   ) & 5     & 2    (1   ) \\\\\n"
     ]
    }
   ],
   "source": [
    "langs = [None] + list(iso_to_lang_mapping.keys())\n",
    "\n",
    "print(\"Min, mean, and max article title and body lengths (in characters) by language:\")\n",
    "print(f\"Language   & Arts   & tavg (tstd) & bavg  (bstd ) & cls   & cavg (cstd) \\\\\\\\\")\n",
    "for lang in langs:\n",
    "    df = label_df[label_df[\"lang\"] == lang] if lang else label_df\n",
    "    language = iso_to_lang_mapping.get(lang, \"Overall\")\n",
    "\n",
    "    title_stats = get_length_stats(df, \"title\")\n",
    "    body_stats = get_length_stats(df, \"body\")\n",
    "\n",
    "    cluster_stats = get_cluster_stats(df, \"clusterId\")\n",
    "    print(f\"{language:<10} & {len(df):<6,.0f} & {title_stats['avg']:<4.0f} ({title_stats['std']:<4,.0f}) & {body_stats['avg']:<5,.0f} ({body_stats['std']:<5,.0f}) & {cluster_stats['count']:<5,} & {cluster_stats['avg']:<4.0f} ({cluster_stats['std']:<4.0f}) \\\\\\\\\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "None                                         Overall    & 10,940 & 1,350 \\\\\n",
      "None                                         English    & 4,009  & 729   \\\\\n",
      "None                                         Portuguese & 2,410  & 368   \\\\\n",
      "None                                         Spanish    & 2,049  & 381   \\\\\n",
      "None                                         French     & 845    & 170   \\\\\n",
      "None                                         Russian    & 553    & 152   \\\\\n",
      "None                                         German     & 516    & 100   \\\\\n",
      "None                                         Slovenian  & 331    & 102   \\\\\n",
      "None                                         Arabic     & 218    & 71    \\\\\n",
      "None                                         Chinese    & 9      & 5     \\\\\n",
      "['olympic_games', 'japan', 'basketball']     Overall    & 2,898  & 368   \\\\\n",
      "['olympic_games', 'japan', 'basketball']     English    & 963    & 189   \\\\\n",
      "['olympic_games', 'japan', 'basketball']     Portuguese & 293    & 83    \\\\\n",
      "['olympic_games', 'japan', 'basketball']     Spanish    & 774    & 148   \\\\\n",
      "['olympic_games', 'japan', 'basketball']     French     & 279    & 67    \\\\\n",
      "['olympic_games', 'japan', 'basketball']     Russian    & 168    & 60    \\\\\n",
      "['olympic_games', 'japan', 'basketball']     German     & 164    & 30    \\\\\n",
      "['olympic_games', 'japan', 'basketball']     Slovenian  & 216    & 73    \\\\\n",
      "['olympic_games', 'japan', 'basketball']     Arabic     & 39     & 18    \\\\\n",
      "['olympic_games', 'japan', 'basketball']     Chinese    & 2      & 2     \\\\\n",
      "['olympic_games', 'japan', 'judo']           Overall    & 3,868  & 487   \\\\\n",
      "['olympic_games', 'japan', 'judo']           English    & 1,096  & 234   \\\\\n",
      "['olympic_games', 'japan', 'judo']           Portuguese & 981    & 150   \\\\\n",
      "['olympic_games', 'japan', 'judo']           Spanish    & 634    & 138   \\\\\n",
      "['olympic_games', 'japan', 'judo']           French     & 498    & 81    \\\\\n",
      "['olympic_games', 'japan', 'judo']           Russian    & 247    & 60    \\\\\n",
      "['olympic_games', 'japan', 'judo']           German     & 202    & 40    \\\\\n",
      "['olympic_games', 'japan', 'judo']           Slovenian  & 56     & 13    \\\\\n",
      "['olympic_games', 'japan', 'judo']           Arabic     & 150    & 45    \\\\\n",
      "['olympic_games', 'japan', 'judo']           Chinese    & 4      & 2     \\\\\n",
      "['olympic_games', 'japan', 'skateboarding']  Overall    & 2,203  & 219   \\\\\n",
      "['olympic_games', 'japan', 'skateboarding']  English    & 1,048  & 159   \\\\\n",
      "['olympic_games', 'japan', 'skateboarding']  Portuguese & 822    & 88    \\\\\n",
      "['olympic_games', 'japan', 'skateboarding']  Spanish    & 263    & 40    \\\\\n",
      "['olympic_games', 'japan', 'skateboarding']  French     & 0      & 0     \\\\\n",
      "['olympic_games', 'japan', 'skateboarding']  Russian    & 65     & 15    \\\\\n",
      "['olympic_games', 'japan', 'skateboarding']  German     & 0      & 0     \\\\\n",
      "['olympic_games', 'japan', 'skateboarding']  Slovenian  & 4      & 3     \\\\\n",
      "['olympic_games', 'japan', 'skateboarding']  Arabic     & 1      & 1     \\\\\n",
      "['olympic_games', 'japan', 'skateboarding']  Chinese    & 0      & 0     \\\\\n",
      "['olympic_games', 'japan', 'rowing']         Overall    & 1,217  & 197   \\\\\n",
      "['olympic_games', 'japan', 'rowing']         English    & 670    & 138   \\\\\n",
      "['olympic_games', 'japan', 'rowing']         Portuguese & 108    & 21    \\\\\n",
      "['olympic_games', 'japan', 'rowing']         Spanish    & 249    & 40    \\\\\n",
      "['olympic_games', 'japan', 'rowing']         French     & 36     & 14    \\\\\n",
      "['olympic_games', 'japan', 'rowing']         Russian    & 54     & 16    \\\\\n",
      "['olympic_games', 'japan', 'rowing']         German     & 66     & 16    \\\\\n",
      "['olympic_games', 'japan', 'rowing']         Slovenian  & 3      & 1     \\\\\n",
      "['olympic_games', 'japan', 'rowing']         Arabic     & 28     & 10    \\\\\n",
      "['olympic_games', 'japan', 'rowing']         Chinese    & 3      & 2     \\\\\n",
      "['olympic_games', 'japan', 'table_tennis']   Overall    & 378    & 50    \\\\\n",
      "['olympic_games', 'japan', 'table_tennis']   English    & 138    & 15    \\\\\n",
      "['olympic_games', 'japan', 'table_tennis']   Portuguese & 100    & 16    \\\\\n",
      "['olympic_games', 'japan', 'table_tennis']   Spanish    & 0      & 0     \\\\\n",
      "['olympic_games', 'japan', 'table_tennis']   French     & 23     & 9     \\\\\n",
      "['olympic_games', 'japan', 'table_tennis']   Russian    & 19     & 8     \\\\\n",
      "['olympic_games', 'japan', 'table_tennis']   German     & 73     & 12    \\\\\n",
      "['olympic_games', 'japan', 'table_tennis']   Slovenian  & 25     & 9     \\\\\n",
      "['olympic_games', 'japan', 'table_tennis']   Arabic     & 0      & 0     \\\\\n",
      "['olympic_games', 'japan', 'table_tennis']   Chinese    & 0      & 0     \\\\\n",
      "['olympic_games', 'japan', 'swimming']       Overall    & 275    & 52    \\\\\n",
      "['olympic_games', 'japan', 'swimming']       English    & 31     & 7     \\\\\n",
      "['olympic_games', 'japan', 'swimming']       Portuguese & 106    & 25    \\\\\n",
      "['olympic_games', 'japan', 'swimming']       Spanish    & 129    & 27    \\\\\n",
      "['olympic_games', 'japan', 'swimming']       French     & 9      & 5     \\\\\n",
      "['olympic_games', 'japan', 'swimming']       Russian    & 0      & 0     \\\\\n",
      "['olympic_games', 'japan', 'swimming']       German     & 0      & 0     \\\\\n",
      "['olympic_games', 'japan', 'swimming']       Slovenian  & 0      & 0     \\\\\n",
      "['olympic_games', 'japan', 'swimming']       Arabic     & 0      & 0     \\\\\n",
      "['olympic_games', 'japan', 'swimming']       Chinese    & 0      & 0     \\\\\n",
      "['olympic_games', 'japan', 'sport_climbing'] Overall    & 101    & 21    \\\\\n",
      "['olympic_games', 'japan', 'sport_climbing'] English    & 63     & 15    \\\\\n",
      "['olympic_games', 'japan', 'sport_climbing'] Portuguese & 0      & 0     \\\\\n",
      "['olympic_games', 'japan', 'sport_climbing'] Spanish    & 0      & 0     \\\\\n",
      "['olympic_games', 'japan', 'sport_climbing'] French     & 0      & 0     \\\\\n",
      "['olympic_games', 'japan', 'sport_climbing'] Russian    & 0      & 0     \\\\\n",
      "['olympic_games', 'japan', 'sport_climbing'] German     & 11     & 5     \\\\\n",
      "['olympic_games', 'japan', 'sport_climbing'] Slovenian  & 27     & 5     \\\\\n",
      "['olympic_games', 'japan', 'sport_climbing'] Arabic     & 0      & 0     \\\\\n",
      "['olympic_games', 'japan', 'sport_climbing'] Chinese    & 0      & 0     \\\\\n"
     ]
    }
   ],
   "source": [
    "concepts = [None] + list(label_df[\"concepts\"].unique())\n",
    "\n",
    "for concept in concepts:\n",
    "    for lang in langs:\n",
    "        df = label_df[label_df[\"concepts\"] == concept] if concept else label_df\n",
    "        df = df[df[\"lang\"] == lang] if lang else df\n",
    "        language = iso_to_lang_mapping.get(lang, \"Overall\")\n",
    "\n",
    "        title_stats = get_length_stats(df, \"title\")\n",
    "        body_stats = get_length_stats(df, \"body\")\n",
    "\n",
    "        cluster_stats = get_cluster_stats(df, \"clusterId\")\n",
    "        print(f\"{str(concept):<44} {language:<10} & {len(df):<6,.0f} & {cluster_stats['count']:<5,} \\\\\\\\\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Cluster Analysis"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Min, max, mean, and std of cluster duration by language:\n",
      "Overall    & 0.00  & 19.41 & 0.78  & 1.21  \\\\\n",
      "English    & 0.00  & 19.41 & 0.56  & 1.32  \\\\\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Portuguese & 0.00  & 11.49 & 0.67  & 1.10  \\\\\n",
      "Spanish    & 0.00  & 12.11 & 0.68  & 1.06  \\\\\n",
      "French     & 0.00  & 3.77  & 0.54  & 0.78  \\\\\n",
      "Russian    & 0.00  & 3.21  & 0.34  & 0.60  \\\\\n",
      "German     & 0.00  & 7.07  & 0.60  & 0.93  \\\\\n",
      "Slovenian  & 0.00  & 7.80  & 0.51  & 0.99  \\\\\n",
      "Arabic     & 0.00  & 3.35  & 0.48  & 0.65  \\\\\n",
      "Chinese    & 0.00  & 2.59  & 0.56  & 1.14  \\\\\n"
     ]
    }
   ],
   "source": [
    "def get_time_stats(df, column):\n",
    "    vals = df.groupby(column)['dateTime'].agg(['min', 'max'])\n",
    "    return {\n",
    "        \"mean\": (vals[\"max\"] - vals[\"min\"]).mean() / np.timedelta64(1, 'D'),\n",
    "        \"std\":  (vals[\"max\"] - vals[\"min\"]).std()  / np.timedelta64(1, 'D'),\n",
    "        \"min\":  (vals[\"max\"] - vals[\"min\"]).min()  / np.timedelta64(1, 'D'),\n",
    "        \"max\":  (vals[\"max\"] - vals[\"min\"]).max()  / np.timedelta64(1, 'D'),\n",
    "    }\n",
    "\n",
    "langs = [None] + list(iso_to_lang_mapping.keys())\n",
    "print(\"Min, max, mean, and std of cluster duration by language:\")\n",
    "for lang in langs:\n",
    "    df = label_df[label_df[\"lang\"] == lang] if lang else label_df\n",
    "    language = iso_to_lang_mapping.get(lang, \"Overall\")\n",
    "    title_stats = get_time_stats(df, \"clusterId\")\n",
    "    print(f\"{language:<10} & {title_stats['min']:<5.2f} & {title_stats['max']:<5.2f} & {title_stats['mean']:<5.2f} & {title_stats['std']:<5.2f} \\\\\\\\\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [],
   "source": [
    "count_by_cluster_id = get_article_group_count(label_df.groupby([\"clusterId\"]).groups)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [],
   "source": [
    "cls_sizes = [x[\"y\"] for x in count_by_cluster_id]\n",
    "min_cls_size = min(cls_sizes)\n",
    "max_cls_size = max(cls_sizes)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {},
   "outputs": [],
   "source": [
    "number_of_cls_size = []\n",
    "for cls_size in range(min_cls_size, max_cls_size + 1):\n",
    "    count = len([x for x in count_by_cluster_id if x[\"y\"] == cls_size])\n",
    "    number_of_cls_size.append({\"label\": cls_size, \"y\": count})"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {},
   "outputs": [],
   "source": [
    "PERCENTILE = 0.95\n",
    "\n",
    "cumm_dist = [\n",
    "    sum([v[\"y\"] for v in number_of_cls_size[: idx + 1]])\n",
    "    for idx in range(len(number_of_cls_size))\n",
    "]\n",
    "quart_th = len(list(filter(lambda x: x <= cumm_dist[-1] * PERCENTILE, cumm_dist)))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {},
   "outputs": [],
   "source": [
    "number_of_cls_size = number_of_cls_size[:40] + [{\"label\": \"\", \"y\": 0 }]*100 + [number_of_cls_size[-1]]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {},
   "outputs": [],
   "source": [
    "xlabels = [x[\"label\"] for x in number_of_cls_size]\n",
    "xticks = [0, quart_th, len(xlabels) - 1]\n",
    "\n",
    "ymax = 300"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 820x240 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "figure = bar_chart({\n",
    "    \"charts\": {\n",
    "        \"data\": number_of_cls_size,\n",
    "        \"xticks\": xticks,\n",
    "        \"xticklabels\": [xlabels[idx] for idx in xticks],\n",
    "        \"vlines\": {\n",
    "            \"x\": quart_th,\n",
    "            \"ymax\": ymax,\n",
    "            \"label\": f\"{PERCENTILE*100:.0f}th percentile\",\n",
    "        }\n",
    "    },\n",
    "    \"ymax\": ymax,\n",
    "    \"title\": \"cluster distribution by size\",\n",
    "    \"xlabel\": \"cluster size (number of articles in cluster)\",\n",
    "    \"ylabel\": \"number of clusters\",\n",
    "    \"figsize\": FIG_SIZE.A4_NARROW,\n",
    "    \"show_legend\": True,\n",
    "    \"show_grid\": \"y\",\n",
    "})\n",
    "save_figure(figure, \"../results/og2021-cluster-distribution-by-size.png\", dpi=300, format=FIG_FORMAT.PNG, transparent=True)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Mono- vs multi-lingual cluster analysis"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {},
   "outputs": [],
   "source": [
    "distinct_langs = label_df.groupby([\"clusterId\"])[\"lang\"].nunique().to_list()\n",
    "min_dist_langs = min(distinct_langs)\n",
    "max_dist_langs = max(distinct_langs)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {},
   "outputs": [],
   "source": [
    "number_of_dist_langs = []\n",
    "for dist_lang_count in range(min_dist_langs, max_dist_langs + 1):\n",
    "    count = len([x for x in distinct_langs if x == dist_lang_count])\n",
    "    number_of_dist_langs.append(\n",
    "        {\n",
    "            \"label\": dist_lang_count,\n",
    "            \"y\": count,\n",
    "        }\n",
    "    )"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.2777777777777778"
      ]
     },
     "execution_count": 32,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "sum([x[\"y\"] for x in filter(lambda x: x[\"label\"] != 1, number_of_dist_langs)]) / sum([x[\"y\"] for x in number_of_dist_langs])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 820x240 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "figure = bar_chart({\n",
    "    \"charts\": { \"data\": number_of_dist_langs },\n",
    "    \"title\": \"cluster distribution per languages\",\n",
    "    \"xlabel\": \"number of languages in cluster\",\n",
    "    \"ylabel\": \"number of clusters\",\n",
    "    \"show_grid\": \"y\",\n",
    "    \"figsize\": FIG_SIZE.A4_NARROW,\n",
    "    \"ymax\": 1200,\n",
    "})\n",
    "save_figure(figure, \"../results/og2021-cluster-distribution-by-lang.png\", dpi=300, format=FIG_FORMAT.PNG, transparent=True)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Language coocurrence in clusters analysis"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {},
   "outputs": [],
   "source": [
    "iso_to_idx_mapping = { key: idx for idx, key in enumerate(iso_to_lang_mapping.keys()) }\n",
    "n_langs = len(iso_to_lang_mapping.keys())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "metadata": {},
   "outputs": [],
   "source": [
    "coocurrence = np.array([[0] * n_langs for _ in range(n_langs)])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "metadata": {},
   "outputs": [],
   "source": [
    "examples = label_df.groupby([\"clusterId\"])[\"lang\"].unique().to_list()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "metadata": {},
   "outputs": [],
   "source": [
    "for langs in examples:\n",
    "    for idx in range(len(langs)):\n",
    "        lang_idx = iso_to_idx_mapping[langs[idx]]\n",
    "        for idy in range(len(langs)):\n",
    "            lang_idy = iso_to_idx_mapping[langs[idy]]\n",
    "            coocurrence[lang_idx][lang_idy] += 1"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "metadata": {},
   "outputs": [],
   "source": [
    "coocurrence = coocurrence / len(examples)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 88,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x640 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "figure = heatmap(\n",
    "    {\n",
    "        \"charts\": {\n",
    "            \"data\": coocurrence,\n",
    "            \"norm\": NORMALIZE.LOG,\n",
    "            \"valfmt\": \"{x:.1%}\",\n",
    "            \"xticks\": [idx for idx in range(n_langs)],\n",
    "            \"xticklabels\": [iso_to_lang_mapping[key] for key in iso_to_idx_mapping.keys()],\n",
    "            \"xtickrotate\": 45,\n",
    "            \"yticks\": [idx for idx in range(n_langs)],\n",
    "            \"yticklabels\": [iso_to_lang_mapping[key] for key in iso_to_idx_mapping.keys()],\n",
    "        },\n",
    "        \"title\": \"the language co-occurence matrix\",\n",
    "        \"figsize\": FIG_SIZE.SQUARE,\n",
    "        \"show_heatmap_vals\": True,\n",
    "    }\n",
    ")\n",
    "save_figure(figure, \"../results/og2021-cluster-language-cooccurence.png\", dpi=300, format=FIG_FORMAT.PNG, transparent=True)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Final save of the dataset"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 90,
   "metadata": {},
   "outputs": [],
   "source": [
    "import uuid\n",
    "import math"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Prepare the final dataset"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 91,
   "metadata": {},
   "outputs": [],
   "source": [
    "final_df = pd.DataFrame(label_df[[\"id\", \"title\", \"body\", \"lang\", \"source\", \"dateTime\", \"url\", \"clusterId\"]])\n",
    "final_df.rename(columns={ \"dateTime\": \"published_at\", \"clusterId\": \"cluster_id\" }, inplace=True)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Create the mapping between the cluster ID and the unique identifier"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 92,
   "metadata": {},
   "outputs": [],
   "source": [
    "id_mapping = { cluster_id: str(uuid.uuid4()).split(\"-\")[0] for cluster_id in final_df[\"cluster_id\"].unique()}"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 93,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "True"
      ]
     },
     "execution_count": 93,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(set(id_mapping.keys())) == len(set(id_mapping.values()))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 94,
   "metadata": {},
   "outputs": [],
   "source": [
    "final_df[\"cluster_id\"] = final_df[\"cluster_id\"].map(lambda x: f\"cls-{id_mapping[x]}\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 95,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>id</th>\n",
       "      <th>title</th>\n",
       "      <th>body</th>\n",
       "      <th>lang</th>\n",
       "      <th>source</th>\n",
       "      <th>published_at</th>\n",
       "      <th>url</th>\n",
       "      <th>cluster_id</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0</td>\n",
       "      <td>Iran's Basketball Team Held At Airport Over 'U...</td>\n",
       "      <td>Incident comes weeks ahead of Tokyo Olympics, ...</td>\n",
       "      <td>eng</td>\n",
       "      <td>Haberler.com</td>\n",
       "      <td>2021-07-01 06:01:00</td>\n",
       "      <td>https://en.haberler.com/iran-s-basketball-team...</td>\n",
       "      <td>cls-28ced65f</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1</td>\n",
       "      <td>Olympics latest: Samoa participation in doubt ...</td>\n",
       "      <td>TOKYO -- The July 23 opening ceremony of the d...</td>\n",
       "      <td>eng</td>\n",
       "      <td>Nikkei Asia</td>\n",
       "      <td>2021-07-01 08:51:00</td>\n",
       "      <td>https://asia.nikkei.com/Spotlight/Tokyo-2020-O...</td>\n",
       "      <td>cls-8eb8df54</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>2</td>\n",
       "      <td>What Do the Olympics Have Against Women?</td>\n",
       "      <td>Yes, I am talking about how Laurel Hubbard, a ...</td>\n",
       "      <td>eng</td>\n",
       "      <td>Townhall</td>\n",
       "      <td>2021-07-01 08:54:00</td>\n",
       "      <td>https://townhall.com/tipsheet/rebeccadowns/202...</td>\n",
       "      <td>cls-3ee1cc3c</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>3</td>\n",
       "      <td>Breastfeeding Mother Granted Permission To Bri...</td>\n",
       "      <td>Breastfeeding mothers competing at the Olympic...</td>\n",
       "      <td>eng</td>\n",
       "      <td>www.sportbible.com</td>\n",
       "      <td>2021-07-01 10:10:00</td>\n",
       "      <td>https://www.sportbible.com/australia/news-brea...</td>\n",
       "      <td>cls-49c24cb8</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>4</td>\n",
       "      <td>Sports News | Breastfeeding Olympians Allowed ...</td>\n",
       "      <td>Canadian basketball player Kim Gaucher, who is...</td>\n",
       "      <td>eng</td>\n",
       "      <td>LatestLY</td>\n",
       "      <td>2021-07-01 10:20:00</td>\n",
       "      <td>https://www.latestly.com/agency-news/sports-ne...</td>\n",
       "      <td>cls-49c24cb8</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10935</th>\n",
       "      <td>10935</td>\n",
       "      <td>Juegos Olímpicos ganaron aceptación entre much...</td>\n",
       "      <td>Antes, muchos japoneses mostraron su reticenci...</td>\n",
       "      <td>spa</td>\n",
       "      <td>www.diariolibre.com</td>\n",
       "      <td>2021-08-13 18:36:00</td>\n",
       "      <td>https://www.diariolibre.com/deportes/olimpismo...</td>\n",
       "      <td>cls-4e17beb8</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10936</th>\n",
       "      <td>10936</td>\n",
       "      <td>Japón, al final, se dejó 'seducir' por los Jue...</td>\n",
       "      <td>Por ahora, hay en muchos japoneses un sentimie...</td>\n",
       "      <td>spa</td>\n",
       "      <td>El Financiero</td>\n",
       "      <td>2021-08-13 21:09:00</td>\n",
       "      <td>https://www.elfinanciero.com.mx/tokio-2020/202...</td>\n",
       "      <td>cls-4e17beb8</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10937</th>\n",
       "      <td>10937</td>\n",
       "      <td>Películas esenciales por si te clavaste con el...</td>\n",
       "      <td>La brasileña Rayssa Leal se robó el corazón de...</td>\n",
       "      <td>spa</td>\n",
       "      <td>Chilango</td>\n",
       "      <td>2021-08-13 23:08:00</td>\n",
       "      <td>https://www.chilango.com/cine-y-tv/peliculas-e...</td>\n",
       "      <td>cls-a2208557</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10938</th>\n",
       "      <td>10938</td>\n",
       "      <td>El alcalde japonés que le tocó ofrecer disculp...</td>\n",
       "      <td>Por una escena transmitida por televisión, que...</td>\n",
       "      <td>spa</td>\n",
       "      <td>Récord</td>\n",
       "      <td>2021-08-14 02:47:00</td>\n",
       "      <td>https://record.acento.com.do/polideportivo/el-...</td>\n",
       "      <td>cls-83f2a7a0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10939</th>\n",
       "      <td>10939</td>\n",
       "      <td>Juegos Olímpicos fueron aceptados por muchos j...</td>\n",
       "      <td>En los momentos en que comenzaban los Juegos O...</td>\n",
       "      <td>spa</td>\n",
       "      <td>El Informador Venezuela</td>\n",
       "      <td>2021-08-14 04:25:00</td>\n",
       "      <td>https://www.elinformadorvenezuela.com/deportes...</td>\n",
       "      <td>cls-4e17beb8</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>10940 rows × 8 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "          id                                              title  \\\n",
       "0          0  Iran's Basketball Team Held At Airport Over 'U...   \n",
       "1          1  Olympics latest: Samoa participation in doubt ...   \n",
       "2          2           What Do the Olympics Have Against Women?   \n",
       "3          3  Breastfeeding Mother Granted Permission To Bri...   \n",
       "4          4  Sports News | Breastfeeding Olympians Allowed ...   \n",
       "...      ...                                                ...   \n",
       "10935  10935  Juegos Olímpicos ganaron aceptación entre much...   \n",
       "10936  10936  Japón, al final, se dejó 'seducir' por los Jue...   \n",
       "10937  10937  Películas esenciales por si te clavaste con el...   \n",
       "10938  10938  El alcalde japonés que le tocó ofrecer disculp...   \n",
       "10939  10939  Juegos Olímpicos fueron aceptados por muchos j...   \n",
       "\n",
       "                                                    body lang  \\\n",
       "0      Incident comes weeks ahead of Tokyo Olympics, ...  eng   \n",
       "1      TOKYO -- The July 23 opening ceremony of the d...  eng   \n",
       "2      Yes, I am talking about how Laurel Hubbard, a ...  eng   \n",
       "3      Breastfeeding mothers competing at the Olympic...  eng   \n",
       "4      Canadian basketball player Kim Gaucher, who is...  eng   \n",
       "...                                                  ...  ...   \n",
       "10935  Antes, muchos japoneses mostraron su reticenci...  spa   \n",
       "10936  Por ahora, hay en muchos japoneses un sentimie...  spa   \n",
       "10937  La brasileña Rayssa Leal se robó el corazón de...  spa   \n",
       "10938  Por una escena transmitida por televisión, que...  spa   \n",
       "10939  En los momentos en que comenzaban los Juegos O...  spa   \n",
       "\n",
       "                        source        published_at  \\\n",
       "0                 Haberler.com 2021-07-01 06:01:00   \n",
       "1                  Nikkei Asia 2021-07-01 08:51:00   \n",
       "2                     Townhall 2021-07-01 08:54:00   \n",
       "3           www.sportbible.com 2021-07-01 10:10:00   \n",
       "4                     LatestLY 2021-07-01 10:20:00   \n",
       "...                        ...                 ...   \n",
       "10935      www.diariolibre.com 2021-08-13 18:36:00   \n",
       "10936            El Financiero 2021-08-13 21:09:00   \n",
       "10937                 Chilango 2021-08-13 23:08:00   \n",
       "10938                   Récord 2021-08-14 02:47:00   \n",
       "10939  El Informador Venezuela 2021-08-14 04:25:00   \n",
       "\n",
       "                                                     url    cluster_id  \n",
       "0      https://en.haberler.com/iran-s-basketball-team...  cls-28ced65f  \n",
       "1      https://asia.nikkei.com/Spotlight/Tokyo-2020-O...  cls-8eb8df54  \n",
       "2      https://townhall.com/tipsheet/rebeccadowns/202...  cls-3ee1cc3c  \n",
       "3      https://www.sportbible.com/australia/news-brea...  cls-49c24cb8  \n",
       "4      https://www.latestly.com/agency-news/sports-ne...  cls-49c24cb8  \n",
       "...                                                  ...           ...  \n",
       "10935  https://www.diariolibre.com/deportes/olimpismo...  cls-4e17beb8  \n",
       "10936  https://www.elfinanciero.com.mx/tokio-2020/202...  cls-4e17beb8  \n",
       "10937  https://www.chilango.com/cine-y-tv/peliculas-e...  cls-a2208557  \n",
       "10938  https://record.acento.com.do/polideportivo/el-...  cls-83f2a7a0  \n",
       "10939  https://www.elinformadorvenezuela.com/deportes...  cls-4e17beb8  \n",
       "\n",
       "[10940 rows x 8 columns]"
      ]
     },
     "execution_count": 95,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "final_df"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 96,
   "metadata": {},
   "outputs": [],
   "source": [
    "final_df.to_csv(\"../data/final/og2021.full.csv\", encoding=\"utf-8\", index=False)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Trim the body to contain at most 15% of the original text"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 97,
   "metadata": {},
   "outputs": [],
   "source": [
    "def trim_body(x):\n",
    "    max_id = math.floor(len(x) * 0.15)\n",
    "    return x[:max_id]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 98,
   "metadata": {},
   "outputs": [],
   "source": [
    "final_df[\"body\"] = final_df[\"body\"].map(lambda x: trim_body(x))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 99,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>id</th>\n",
       "      <th>title</th>\n",
       "      <th>body</th>\n",
       "      <th>lang</th>\n",
       "      <th>source</th>\n",
       "      <th>published_at</th>\n",
       "      <th>url</th>\n",
       "      <th>cluster_id</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0</td>\n",
       "      <td>Iran's Basketball Team Held At Airport Over 'U...</td>\n",
       "      <td>Incident comes weeks ahead of Tokyo Olympics, ...</td>\n",
       "      <td>eng</td>\n",
       "      <td>Haberler.com</td>\n",
       "      <td>2021-07-01 06:01:00</td>\n",
       "      <td>https://en.haberler.com/iran-s-basketball-team...</td>\n",
       "      <td>cls-28ced65f</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1</td>\n",
       "      <td>Olympics latest: Samoa participation in doubt ...</td>\n",
       "      <td>TOKYO -- The July 23 opening ceremony of the d...</td>\n",
       "      <td>eng</td>\n",
       "      <td>Nikkei Asia</td>\n",
       "      <td>2021-07-01 08:51:00</td>\n",
       "      <td>https://asia.nikkei.com/Spotlight/Tokyo-2020-O...</td>\n",
       "      <td>cls-8eb8df54</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>2</td>\n",
       "      <td>What Do the Olympics Have Against Women?</td>\n",
       "      <td>Yes, I am talking about how Laurel Hubbard, a ...</td>\n",
       "      <td>eng</td>\n",
       "      <td>Townhall</td>\n",
       "      <td>2021-07-01 08:54:00</td>\n",
       "      <td>https://townhall.com/tipsheet/rebeccadowns/202...</td>\n",
       "      <td>cls-3ee1cc3c</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>3</td>\n",
       "      <td>Breastfeeding Mother Granted Permission To Bri...</td>\n",
       "      <td>Breastfeeding mothers competing at the Olympic...</td>\n",
       "      <td>eng</td>\n",
       "      <td>www.sportbible.com</td>\n",
       "      <td>2021-07-01 10:10:00</td>\n",
       "      <td>https://www.sportbible.com/australia/news-brea...</td>\n",
       "      <td>cls-49c24cb8</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>4</td>\n",
       "      <td>Sports News | Breastfeeding Olympians Allowed ...</td>\n",
       "      <td>Canadian basketball player Kim Gaucher, who is...</td>\n",
       "      <td>eng</td>\n",
       "      <td>LatestLY</td>\n",
       "      <td>2021-07-01 10:20:00</td>\n",
       "      <td>https://www.latestly.com/agency-news/sports-ne...</td>\n",
       "      <td>cls-49c24cb8</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10935</th>\n",
       "      <td>10935</td>\n",
       "      <td>Juegos Olímpicos ganaron aceptación entre much...</td>\n",
       "      <td>Antes, muchos japoneses mostraron su reticenci...</td>\n",
       "      <td>spa</td>\n",
       "      <td>www.diariolibre.com</td>\n",
       "      <td>2021-08-13 18:36:00</td>\n",
       "      <td>https://www.diariolibre.com/deportes/olimpismo...</td>\n",
       "      <td>cls-4e17beb8</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10936</th>\n",
       "      <td>10936</td>\n",
       "      <td>Japón, al final, se dejó 'seducir' por los Jue...</td>\n",
       "      <td>Por ahora, hay en muchos japoneses un sentimie...</td>\n",
       "      <td>spa</td>\n",
       "      <td>El Financiero</td>\n",
       "      <td>2021-08-13 21:09:00</td>\n",
       "      <td>https://www.elfinanciero.com.mx/tokio-2020/202...</td>\n",
       "      <td>cls-4e17beb8</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10937</th>\n",
       "      <td>10937</td>\n",
       "      <td>Películas esenciales por si te clavaste con el...</td>\n",
       "      <td>La brasileña Rayssa Leal se robó el corazón de...</td>\n",
       "      <td>spa</td>\n",
       "      <td>Chilango</td>\n",
       "      <td>2021-08-13 23:08:00</td>\n",
       "      <td>https://www.chilango.com/cine-y-tv/peliculas-e...</td>\n",
       "      <td>cls-a2208557</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10938</th>\n",
       "      <td>10938</td>\n",
       "      <td>El alcalde japonés que le tocó ofrecer disculp...</td>\n",
       "      <td>Por una escena transmitida por televisión, que...</td>\n",
       "      <td>spa</td>\n",
       "      <td>Récord</td>\n",
       "      <td>2021-08-14 02:47:00</td>\n",
       "      <td>https://record.acento.com.do/polideportivo/el-...</td>\n",
       "      <td>cls-83f2a7a0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10939</th>\n",
       "      <td>10939</td>\n",
       "      <td>Juegos Olímpicos fueron aceptados por muchos j...</td>\n",
       "      <td>En los momentos en que comenzaban los Juegos O...</td>\n",
       "      <td>spa</td>\n",
       "      <td>El Informador Venezuela</td>\n",
       "      <td>2021-08-14 04:25:00</td>\n",
       "      <td>https://www.elinformadorvenezuela.com/deportes...</td>\n",
       "      <td>cls-4e17beb8</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>10940 rows × 8 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "          id                                              title  \\\n",
       "0          0  Iran's Basketball Team Held At Airport Over 'U...   \n",
       "1          1  Olympics latest: Samoa participation in doubt ...   \n",
       "2          2           What Do the Olympics Have Against Women?   \n",
       "3          3  Breastfeeding Mother Granted Permission To Bri...   \n",
       "4          4  Sports News | Breastfeeding Olympians Allowed ...   \n",
       "...      ...                                                ...   \n",
       "10935  10935  Juegos Olímpicos ganaron aceptación entre much...   \n",
       "10936  10936  Japón, al final, se dejó 'seducir' por los Jue...   \n",
       "10937  10937  Películas esenciales por si te clavaste con el...   \n",
       "10938  10938  El alcalde japonés que le tocó ofrecer disculp...   \n",
       "10939  10939  Juegos Olímpicos fueron aceptados por muchos j...   \n",
       "\n",
       "                                                    body lang  \\\n",
       "0      Incident comes weeks ahead of Tokyo Olympics, ...  eng   \n",
       "1      TOKYO -- The July 23 opening ceremony of the d...  eng   \n",
       "2      Yes, I am talking about how Laurel Hubbard, a ...  eng   \n",
       "3      Breastfeeding mothers competing at the Olympic...  eng   \n",
       "4      Canadian basketball player Kim Gaucher, who is...  eng   \n",
       "...                                                  ...  ...   \n",
       "10935  Antes, muchos japoneses mostraron su reticenci...  spa   \n",
       "10936  Por ahora, hay en muchos japoneses un sentimie...  spa   \n",
       "10937  La brasileña Rayssa Leal se robó el corazón de...  spa   \n",
       "10938  Por una escena transmitida por televisión, que...  spa   \n",
       "10939  En los momentos en que comenzaban los Juegos O...  spa   \n",
       "\n",
       "                        source        published_at  \\\n",
       "0                 Haberler.com 2021-07-01 06:01:00   \n",
       "1                  Nikkei Asia 2021-07-01 08:51:00   \n",
       "2                     Townhall 2021-07-01 08:54:00   \n",
       "3           www.sportbible.com 2021-07-01 10:10:00   \n",
       "4                     LatestLY 2021-07-01 10:20:00   \n",
       "...                        ...                 ...   \n",
       "10935      www.diariolibre.com 2021-08-13 18:36:00   \n",
       "10936            El Financiero 2021-08-13 21:09:00   \n",
       "10937                 Chilango 2021-08-13 23:08:00   \n",
       "10938                   Récord 2021-08-14 02:47:00   \n",
       "10939  El Informador Venezuela 2021-08-14 04:25:00   \n",
       "\n",
       "                                                     url    cluster_id  \n",
       "0      https://en.haberler.com/iran-s-basketball-team...  cls-28ced65f  \n",
       "1      https://asia.nikkei.com/Spotlight/Tokyo-2020-O...  cls-8eb8df54  \n",
       "2      https://townhall.com/tipsheet/rebeccadowns/202...  cls-3ee1cc3c  \n",
       "3      https://www.sportbible.com/australia/news-brea...  cls-49c24cb8  \n",
       "4      https://www.latestly.com/agency-news/sports-ne...  cls-49c24cb8  \n",
       "...                                                  ...           ...  \n",
       "10935  https://www.diariolibre.com/deportes/olimpismo...  cls-4e17beb8  \n",
       "10936  https://www.elfinanciero.com.mx/tokio-2020/202...  cls-4e17beb8  \n",
       "10937  https://www.chilango.com/cine-y-tv/peliculas-e...  cls-a2208557  \n",
       "10938  https://record.acento.com.do/polideportivo/el-...  cls-83f2a7a0  \n",
       "10939  https://www.elinformadorvenezuela.com/deportes...  cls-4e17beb8  \n",
       "\n",
       "[10940 rows x 8 columns]"
      ]
     },
     "execution_count": 99,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "final_df"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 100,
   "metadata": {},
   "outputs": [],
   "source": [
    "final_df.to_csv(\"../data/final/og2021.research.csv\", encoding=\"utf-8\", index=False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 80,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Min, mean, and max article title and body lengths (in characters) by language:\n",
      "Language   & Arts   & tavg (tstd) & bavg  (bstd ) & cls   & cavg (cstd) \\\\\n",
      "Overall    & 10,940 & 12   (4   ) & 118   (128  ) & 1,350 & 8    (20  ) \\\\\n",
      "English    & 4,009  & 11   (3   ) & 183   (171  ) & 729   & 5    (11  ) \\\\\n",
      "Portuguese & 2,410  & 13   (3   ) & 79    (56   ) & 368   & 7    (12  ) \\\\\n",
      "Spanish    & 2,049  & 13   (4   ) & 84    (64   ) & 381   & 5    (9   ) \\\\\n",
      "French     & 845    & 13   (4   ) & 85    (70   ) & 170   & 5    (8   ) \\\\\n",
      "Russian    & 553    & 10   (3   ) & 45    (54   ) & 152   & 4    (6   ) \\\\\n",
      "German     & 516    & 9    (3   ) & 125   (153  ) & 100   & 5    (6   ) \\\\\n",
      "Slovenian  & 331    & 9    (3   ) & 66    (54   ) & 102   & 3    (3   ) \\\\\n",
      "Arabic     & 218    & 10   (3   ) & 60    (40   ) & 71    & 3    (3   ) \\\\\n",
      "Chinese    & 9      & 2    (1   ) & 9     (5    ) & 5     & 2    (1   ) \\\\\n"
     ]
    }
   ],
   "source": [
    "langs = [None] + list(iso_to_lang_mapping.keys())\n",
    "\n",
    "print(\"Min, mean, and max article title and body lengths (in characters) by language:\")\n",
    "print(f\"Language   & Arts   & tavg (tstd) & bavg  (bstd ) & cls   & cavg (cstd) \\\\\\\\\")\n",
    "for lang in langs:\n",
    "    df = final_df[label_df[\"lang\"] == lang] if lang else final_df\n",
    "    language = iso_to_lang_mapping.get(lang, \"Overall\")\n",
    "\n",
    "    title_stats = get_length_stats(df, \"title\")\n",
    "    body_stats = get_length_stats(df, \"body\")\n",
    "\n",
    "    cluster_stats = get_cluster_stats(df, \"cluster_id\")\n",
    "    print(f\"{language:<10} & {len(df):<6,.0f} & {title_stats['avg']:<4.0f} ({title_stats['std']:<4,.0f}) & {body_stats['avg']:<5,.0f} ({body_stats['std']:<5,.0f}) & {cluster_stats['count']:<5,} & {cluster_stats['avg']:<4.0f} ({cluster_stats['std']:<4.0f}) \\\\\\\\\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Prepare the data set for the public"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 101,
   "metadata": {},
   "outputs": [],
   "source": [
    "final_df = final_df[[\"id\", \"title\", \"lang\", \"source\", \"published_at\", \"url\", \"cluster_id\"]]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 102,
   "metadata": {},
   "outputs": [],
   "source": [
    "final_df.to_csv(\"../data/final/og2021.public.csv\", encoding=\"utf-8\", index=False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "worldnews",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.8.12"
  },
  "orig_nbformat": 4
 },
 "nbformat": 4,
 "nbformat_minor": 2
}