data_exploration_task1.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import json\n",
    "from pprint import pprint\n",
    "\n",
    "import jsonlines\n",
    "import matplotlib.pyplot as plt\n",
    "import numpy as np\n",
    "import pandas as pd\n",
    "import seaborn as sns\n",
    "from tqdm import tqdm, trange\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "example_1 = {\n",
    "    \"annotations\": [\n",
    "        {\n",
    "            \"annotation_id\": 9824473826482099305,\n",
    "            \"minimal_answer\": {\"plaintext_end_byte\": -1, \"plaintext_start_byte\": -1},\n",
    "            \"passage_answer\": {\"candidate_index\": 2},\n",
    "            \"yes_no_answer\": \"YES\",\n",
    "        },\n",
    "        {\n",
    "            \"annotation_id\": 9020802343333398197,\n",
    "            \"minimal_answer\": {\"plaintext_end_byte\": -1, \"plaintext_start_byte\": -1},\n",
    "            \"passage_answer\": {\"candidate_index\": 0},\n",
    "            \"yes_no_answer\": \"NO\",\n",
    "        },\n",
    "        {\n",
    "            \"annotation_id\": 17573111662051460994,\n",
    "            \"minimal_answer\": {\"plaintext_end_byte\": -1, \"plaintext_start_byte\": -1},\n",
    "            \"passage_answer\": {\"candidate_index\": 0},\n",
    "            \"yes_no_answer\": \"NO\",\n",
    "        },\n",
    "    ],\n",
    "    \"document_plaintext\": '\\nIn mainstream Nicene Christianity, there is no restriction on kinds of animals that can be eaten.[1][2] This practice stems from Peter\\'s vision of a sheet with animals, described in the Book of Acts, Chapter 10, in which Saint Peter \"sees a sheet containing animals of every description lowered from the sky.\"[3]  Nonetheless, the New Testament does give a few guidelines about the consumption of meat, practiced by the Christian Church today; one of these is not consuming food knowingly offered to pagan idols,[4] a conviction that the early Church Fathers, such as Clement of Alexandria and Origen preached.[5] In addition, Christians traditionally bless any food before eating it with a mealtime prayer (grace), as a sign of thanking God for the meal they have.\\nSlaughtering animals for food is  often done without the trinitarian formula,[6][7] although the Armenian Apostolic Church, among other Orthodox Christians, have rituals that \"display obvious links with shechitah, Jewish kosher slaughter.\"[8] The Bible, states Norman Geisler, stipulates one to \"abstain from food sacrificed to idols, from blood, from meat of strangled animals\".[9]\\nIn the New Testament, Paul of Tarsus notes that some devout Christians may wish to abstain from consuming meat if it causes \"my brother to stumble\" in his faith with God (cf. 1 Corinthians 8:13).[10] As such, some Christian monks, such as the Trappists, have adopted a policy of Christian vegetarianism.[11] In addition, Christians of the Seventh-day Adventist tradition generally \"avoid eating meat and highly spiced food\".[12] Christians in the Anglican, Catholic, Lutheran, Methodist, and Orthodox denominations traditionally observe a meat-free day, especially during the liturgical season of Lent.[13][14][15][16]\\nWith reference to medieval times, Jillian Williams states that \"unlike the Jewish and Muslims methods of animal slaughter, which requires the draining of the animal\\'s blood, Christian slaughter practices did not usually specify the method of slaughter\". In actual practice, states Williams, European Christians have flexibly practiced both the method of draining the blood, and wringing the animal\\'s neck to retain its blood as valuable food.[17] According to Basheer Ahmad Masri, the \"Jewish and the Christian methods of slaughter fulfill the Islamic condition of bleeding the animal\".[18] In contrast, David Grumett and Rachel Muers state that the Orthodox Christian Shechitah and Jewish Kosher methods of slaughter differ from the Muslim Halal (Dabh) method in that they require the cut to \"sever trachea, oesophagus and the jugular veins\" as this method is believed to produce meat with minimal suffering to the animal.\\nSome Christian denominations condone the moderate drinking of alcohol (moderationism), such as Anglicans, Catholics, Lutherans, and the Orthodox,[19] although others, such as  Seventh-day Adventists, Baptists, Methodists, and Pentecostals either abstain from or prohibit the consumption of alcohol (abstentionism and prohibitionism).[20] However, all Christian Churches, in view of the Biblical position on the issue, universally condemn drunkenness as sinful.[21][22]\\nSee also\\n\\nBible Diet\\nChristian vegetarianism\\nChristian views on alcohol\\nFasting and abstinence in the Catholic Church\\nKashrut (Jewish dietary laws)\\nKutha meat\\nIslamic dietary laws\\nPescetarianism\\nSeventh-Day Adventist Church, Health and Diet\\nWord of Wisdom\\n\\n\\n\\n\\n\\n\\n\\n\\n\\n',\n",
    "    \"document_title\": \"Christian dietary laws\",\n",
    "    \"document_url\": \"https://en.wikipedia.org/wiki/Christian%20dietary%20laws\",\n",
    "    \"example_id\": 6177847336319493473,\n",
    "    \"language\": \"english\",\n",
    "    \"passage_answer_candidates\": [\n",
    "        {\"plaintext_end_byte\": 766, \"plaintext_start_byte\": 1},\n",
    "        {\"plaintext_end_byte\": 1149, \"plaintext_start_byte\": 767},\n",
    "        {\"plaintext_end_byte\": 1768, \"plaintext_start_byte\": 1150},\n",
    "        {\"plaintext_end_byte\": 2692, \"plaintext_start_byte\": 1769},\n",
    "        {\"plaintext_end_byte\": 3161, \"plaintext_start_byte\": 2693},\n",
    "        {\"plaintext_end_byte\": 3417, \"plaintext_start_byte\": 3172},\n",
    "    ],\n",
    "    \"question_text\": \"Does Catholicism have any dietary restrictions?\",\n",
    "}\n",
    "example_2 = {\n",
    "    \"annotations\": [\n",
    "        {\n",
    "            \"annotation_id\": 896469894141137617,\n",
    "            \"minimal_answer\": {\"plaintext_end_byte\": -1, \"plaintext_start_byte\": -1},\n",
    "            \"passage_answer\": {\"candidate_index\": -1},\n",
    "            \"yes_no_answer\": \"NONE\",\n",
    "        },\n",
    "        {\n",
    "            \"annotation_id\": 16078811170855026478,\n",
    "            \"minimal_answer\": {\"plaintext_end_byte\": -1, \"plaintext_start_byte\": -1},\n",
    "            \"passage_answer\": {\"candidate_index\": 52},\n",
    "            \"yes_no_answer\": \"YES\",\n",
    "        },\n",
    "        {\n",
    "            \"annotation_id\": 16091152801123294719,\n",
    "            \"minimal_answer\": {\"plaintext_end_byte\": -1, \"plaintext_start_byte\": -1},\n",
    "            \"passage_answer\": {\"candidate_index\": 52},\n",
    "            \"yes_no_answer\": \"YES\",\n",
    "        },\n",
    "    ],\n",
    "    \"document_plaintext\": '\\n日本の貿易史（にほんのぼうえきし）では、gnored (help)\\n Text \" 和書\" ignored (help)\\n Text \" 和書\" ignored (help)\\n Text \" 和書\" ignored (help)\\n Text \" 和書\" ignored (help)\\n Text \" 和書\" ignored (help)\\n Text \" 和書\" ignored (help)\\n Text \" 和書\" ignored (help)\\n Text \" 和書\" ignored (help)\\n Text \" 和書\" ignored (help)（原書  Text \" 洋書\\n\" ignored (help)）\\n Text \" 和書\" ignored (help)\\n Text \" 和書\" ignored (help)\\n Text \" 和書\" ignored (help)\\n Text \" 和書\" ignored (help)\\n Text \" 和書\" ignored (help)\\n Text \" 和書\" ignored (help)\\n Text \" 和書\" ignored (help)\\n論文、記事\\n\\n関連項目\\n\\n経済史\\n山丹交易\\n朱印船貿易\\n商業史\\n長崎貿易\\n南蛮貿易\\n日系人\\n日本の貨幣史\\n日本の経済 - 日本の経済史\\n日元貿易\\n日明貿易\\n日宋貿易\\n日唐貿易\\n貿易理論\\n貿易 - 貿易史\\n北方貿易\\n琉球貿易\\n\\n外部リンク\\n\\n\\n\\n\\n\\n\\n\\n',\n",
    "    \"document_title\": \"日本の貿易史\",\n",
    "    \"document_url\": \"https://ja.wikipedia.org/wiki/%E6%97%A5%E6%9C%AC%E3%81%AE%E8%B2%BF%E6%98%93%E5%8F%B2\",\n",
    "    \"example_id\": -6001319775852403354,\n",
    "    \"language\": \"japanese\",\n",
    "    \"passage_answer_candidates\": [\n",
    "        {\"plaintext_end_byte\": 286, \"plaintext_start_byte\": 1},\n",
    "        {\"plaintext_end_byte\": 1058, \"plaintext_start_byte\": 301},\n",
    "        {\"plaintext_end_byte\": 1701, \"plaintext_start_byte\": 1059},\n",
    "        {\"plaintext_end_byte\": 2261, \"plaintext_start_byte\": 1709},\n",
    "        {\"plaintext_end_byte\": 3144, \"plaintext_start_byte\": 2262},\n",
    "    ],\n",
    "    \"question_text\": \"安土桃山文化は民間貿易が盛んでしたか？\",\n",
    "}\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "def utf8len(s):\n",
    "    return len(s.encode(\"utf-8\"))\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "def extract_metrics(\n",
    "    dataset_path, language=\"english\", desired_language=None, desired_amount=None\n",
    "):\n",
    "    length_shortest_question = np.inf\n",
    "    samples_with_shortest_question = []\n",
    "\n",
    "    length_longest_question = 0\n",
    "    samples_with_longest_question = []\n",
    "\n",
    "    metrics = []\n",
    "    amount = 0\n",
    "    with jsonlines.open(dataset_path) as dataset:\n",
    "        for sample in tqdm(dataset.iter(skip_invalid=True)):\n",
    "            length_annotation_answer_candidates = [\n",
    "                annotation[\"minimal_answer\"][\"plaintext_end_byte\"]\n",
    "                - annotation[\"minimal_answer\"][\"plaintext_start_byte\"]\n",
    "                for annotation in sample[\"annotations\"]\n",
    "            ]\n",
    "\n",
    "            yes_no_question = int(\n",
    "                np.mean(\n",
    "                    [\n",
    "                        annotation[\"yes_no_answer\"] != \"NONE\"\n",
    "                        for annotation in sample[\"annotations\"]\n",
    "                    ]\n",
    "                )\n",
    "            )\n",
    "            modified_language = sample[\"language\"][0].upper() + sample[\"language\"][1:]\n",
    "            if desired_language is not None and sample[\"language\"] != desired_language:\n",
    "                continue\n",
    "\n",
    "            if (\n",
    "                len(sample[\"question_text\"]) == length_shortest_question\n",
    "                and sample[\"language\"] == language\n",
    "            ):\n",
    "                samples_with_shortest_question += [sample]\n",
    "            if (\n",
    "                len(sample[\"question_text\"]) < length_shortest_question\n",
    "                and sample[\"language\"] == language\n",
    "            ):\n",
    "                length_shortest_question = len(sample[\"question_text\"])\n",
    "                samples_with_shortest_question = [sample]\n",
    "\n",
    "            if (\n",
    "                len(sample[\"question_text\"]) == length_longest_question\n",
    "                and sample[\"language\"] == language\n",
    "            ):\n",
    "                samples_with_longest_question += [sample]\n",
    "            if (\n",
    "                len(sample[\"question_text\"]) > length_longest_question\n",
    "                and sample[\"language\"] == language\n",
    "            ):\n",
    "                length_longest_question = len(sample[\"question_text\"])\n",
    "                samples_with_longest_question = [sample]\n",
    "\n",
    "            metric = {\n",
    "                \"document_length\": len(sample[\"document_plaintext\"]),\n",
    "                \"document_length_byte\": utf8len(sample[\"document_plaintext\"]),\n",
    "                \"unique_caracters_document\": len(\n",
    "                    np.unique(list(sample[\"document_plaintext\"]))\n",
    "                ),\n",
    "                \"question_length\": len(sample[\"question_text\"]),\n",
    "                \"question_length_byte\": utf8len(sample[\"question_text\"]),\n",
    "                \"unique_caracters_question\": len(\n",
    "                    np.unique(list(sample[\"question_text\"]))\n",
    "                ),\n",
    "                \"language\": modified_language,\n",
    "                \"nb_annotation_answer_candidates\": len(\n",
    "                    length_annotation_answer_candidates\n",
    "                ),\n",
    "                \"mean_annotation_answer_candidates\": np.mean(\n",
    "                    length_annotation_answer_candidates\n",
    "                ),\n",
    "                \"min_annotation_answer_candidates\": np.min(\n",
    "                    length_annotation_answer_candidates\n",
    "                ),\n",
    "                \"max_annotation_answer_candidates\": np.max(\n",
    "                    length_annotation_answer_candidates\n",
    "                ),\n",
    "                \"yes_no_question\": yes_no_question,\n",
    "            }\n",
    "            metrics.append(metric)\n",
    "\n",
    "            amount += 1\n",
    "            if desired_amount is not None and amount >= desired_amount:\n",
    "                break\n",
    "    metrics = pd.DataFrame.from_records(metrics).sort_values(\n",
    "        \"language\", ascending=False\n",
    "    )\n",
    "    return metrics, samples_with_shortest_question, samples_with_longest_question\n"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Smallest and longest question samples in the dataset for English and Arabic\n",
    "\n",
    "### English samples:\n",
    "\n",
    "- *Question*: Where is Belize?\n",
    "    \n",
    "    **Title**: Belize\n",
    "    \n",
    "    **Url**: [https://en.wikipedia.org/wiki/Belize](https://en.wikipedia.org/wiki/Belize)\n",
    "    \n",
    "- **Question**: 'What\\'s the real name of the guy called \"La Bamba\" in the Max Weinberg band on Late Night with Conan O\\'Brien?’\n",
    "    \n",
    "    **Title**: Richie Rosenberg\n",
    "    \n",
    "    **Url**: [https://en.wikipedia.org/wiki/Richie%20Rosenberg](https://en.wikipedia.org/wiki/Richie%20Rosenberg)\n",
    "    \n",
    "\n",
    "### Arabic samples:\n",
    "\n",
    "- **Question**: ما هو العثث؟ (Qu'est-ce que les acariens ?)\n",
    "    \n",
    "    **Title**: عثة (Les papillons)\n",
    "    \n",
    "    **Url**: [https://ar.wikipedia.org/wiki/العهد_الدولي_الخاص_بالحقوق_الاقتصادية_والاجتماعية_والثقافية](https://ar.wikipedia.org/wiki/%D8%A7%D9%84%D8%B9%D9%87%D8%AF_%D8%A7%D9%84%D8%AF%D9%88%D9%84%D9%8A_%D8%A7%D9%84%D8%AE%D8%A7%D8%B5_%D8%A8%D8%A7%D9%84%D8%AD%D9%82%D9%88%D9%82_%D8%A7%D9%84%D8%A7%D9%82%D8%AA%D8%B5%D8%A7%D8%AF%D9%8A%D8%A9_%D9%88%D8%A7%D9%84%D8%A7%D8%AC%D8%AA%D9%85%D8%A7%D8%B9%D9%8A%D8%A9_%D9%88%D8%A7%D9%84%D8%AB%D9%82%D8%A7%D9%81%D9%8A%D8%A9)\n",
    "    \n",
    "- **Question**: كم عدد الدول الموقعة على اتفاقية العهد الدولي الخاص بالحقوق الاقتصادية والاجتماعية والثقافية؟ (Combien de pays sont signataires du Pacte international relatif aux droits économiques, sociaux et culturels ?)\n",
    "    \n",
    "    **Title**: العهد الدولي الخاص بالحقوق الاقتصادية والاجتماعية والثقافية (Pacte international relatif aux droits économiques, sociaux et culturels)\n",
    "    \n",
    "    **Url**: [https://ar.wikipedia.org/wiki/العهد_الدولي_الخاص_بالحقوق_الاقتصادية_والاجتماعية_والثقافية](https://ar.wikipedia.org/wiki/%D8%A7%D9%84%D8%B9%D9%87%D8%AF_%D8%A7%D9%84%D8%AF%D9%88%D9%84%D9%8A_%D8%A7%D9%84%D8%AE%D8%A7%D8%B5_%D8%A8%D8%A7%D9%84%D8%AD%D9%82%D9%88%D9%82_%D8%A7%D9%84%D8%A7%D9%82%D8%AA%D8%B5%D8%A7%D8%AF%D9%8A%D8%A9_%D9%88%D8%A7%D9%84%D8%A7%D8%AC%D8%AA%D9%85%D8%A7%D8%B9%D9%8A%D8%A9_%D9%88%D8%A7%D9%84%D8%AB%D9%82%D8%A7%D9%81%D9%8A%D8%A9)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "166916it [02:09, 1293.05it/s]\n",
      "29336it [01:51, 262.53it/s]\n"
     ]
    }
   ],
   "source": [
    "# (\n",
    "#     train_metrics,\n",
    "#     train_samples_with_shortest_question,\n",
    "#     train_samples_with_longest_question,\n",
    "# ) = extract_metrics(\"train_samples/tydiqa-v1.0-train.jsonl\")\n",
    "(\n",
    "    arabic_metrics,\n",
    "    arabic_samples_with_shortest_question,\n",
    "    arabic_samples_with_longest_question,\n",
    ") = extract_metrics(\n",
    "    \"train_samples/tydiqa-v1.0-train.jsonl\",\n",
    "    desired_language=\"arabic\",\n",
    "    desired_amount=None,\n",
    ")\n",
    "(\n",
    "    multi_metrics,\n",
    "    multi_samples_with_shortest_question,\n",
    "    multi_samples_with_longest_question,\n",
    ") = extract_metrics(\n",
    "    \"train_samples/tydiqa-v1.0-train.jsonl\", desired_language=None, desired_amount=29337\n",
    ")\n"
   ]
  },
  {
   "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>document_length</th>\n",
       "      <th>document_length_byte</th>\n",
       "      <th>unique_caracters_document</th>\n",
       "      <th>question_length</th>\n",
       "      <th>question_length_byte</th>\n",
       "      <th>unique_caracters_question</th>\n",
       "      <th>language</th>\n",
       "      <th>nb_annotation_answer_candidates</th>\n",
       "      <th>mean_annotation_answer_candidates</th>\n",
       "      <th>min_annotation_answer_candidates</th>\n",
       "      <th>max_annotation_answer_candidates</th>\n",
       "      <th>yes_no_question</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>14891</th>\n",
       "      <td>1167</td>\n",
       "      <td>3211</td>\n",
       "      <td>90</td>\n",
       "      <td>45</td>\n",
       "      <td>129</td>\n",
       "      <td>25</td>\n",
       "      <td>Thai</td>\n",
       "      <td>1</td>\n",
       "      <td>49.0</td>\n",
       "      <td>49</td>\n",
       "      <td>49</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>22871</th>\n",
       "      <td>10295</td>\n",
       "      <td>27355</td>\n",
       "      <td>135</td>\n",
       "      <td>50</td>\n",
       "      <td>146</td>\n",
       "      <td>29</td>\n",
       "      <td>Thai</td>\n",
       "      <td>1</td>\n",
       "      <td>18.0</td>\n",
       "      <td>18</td>\n",
       "      <td>18</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1512</th>\n",
       "      <td>30854</td>\n",
       "      <td>83940</td>\n",
       "      <td>135</td>\n",
       "      <td>63</td>\n",
       "      <td>185</td>\n",
       "      <td>33</td>\n",
       "      <td>Thai</td>\n",
       "      <td>1</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>22803</th>\n",
       "      <td>10542</td>\n",
       "      <td>29970</td>\n",
       "      <td>114</td>\n",
       "      <td>42</td>\n",
       "      <td>122</td>\n",
       "      <td>28</td>\n",
       "      <td>Thai</td>\n",
       "      <td>1</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>13102</th>\n",
       "      <td>38099</td>\n",
       "      <td>104997</td>\n",
       "      <td>157</td>\n",
       "      <td>31</td>\n",
       "      <td>89</td>\n",
       "      <td>22</td>\n",
       "      <td>Thai</td>\n",
       "      <td>1</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</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",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>19501</th>\n",
       "      <td>1072</td>\n",
       "      <td>1781</td>\n",
       "      <td>75</td>\n",
       "      <td>30</td>\n",
       "      <td>55</td>\n",
       "      <td>16</td>\n",
       "      <td>Arabic</td>\n",
       "      <td>1</td>\n",
       "      <td>4.0</td>\n",
       "      <td>4</td>\n",
       "      <td>4</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10280</th>\n",
       "      <td>7227</td>\n",
       "      <td>12884</td>\n",
       "      <td>88</td>\n",
       "      <td>20</td>\n",
       "      <td>37</td>\n",
       "      <td>13</td>\n",
       "      <td>Arabic</td>\n",
       "      <td>1</td>\n",
       "      <td>8.0</td>\n",
       "      <td>8</td>\n",
       "      <td>8</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3137</th>\n",
       "      <td>19327</td>\n",
       "      <td>34736</td>\n",
       "      <td>107</td>\n",
       "      <td>42</td>\n",
       "      <td>78</td>\n",
       "      <td>19</td>\n",
       "      <td>Arabic</td>\n",
       "      <td>1</td>\n",
       "      <td>96.0</td>\n",
       "      <td>96</td>\n",
       "      <td>96</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1043</th>\n",
       "      <td>24669</td>\n",
       "      <td>44439</td>\n",
       "      <td>82</td>\n",
       "      <td>58</td>\n",
       "      <td>106</td>\n",
       "      <td>19</td>\n",
       "      <td>Arabic</td>\n",
       "      <td>1</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>17710</th>\n",
       "      <td>7229</td>\n",
       "      <td>12980</td>\n",
       "      <td>92</td>\n",
       "      <td>19</td>\n",
       "      <td>34</td>\n",
       "      <td>12</td>\n",
       "      <td>Arabic</td>\n",
       "      <td>1</td>\n",
       "      <td>36.0</td>\n",
       "      <td>36</td>\n",
       "      <td>36</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>29337 rows × 12 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "       document_length  document_length_byte  unique_caracters_document  \\\n",
       "14891             1167                  3211                         90   \n",
       "22871            10295                 27355                        135   \n",
       "1512             30854                 83940                        135   \n",
       "22803            10542                 29970                        114   \n",
       "13102            38099                104997                        157   \n",
       "...                ...                   ...                        ...   \n",
       "19501             1072                  1781                         75   \n",
       "10280             7227                 12884                         88   \n",
       "3137             19327                 34736                        107   \n",
       "1043             24669                 44439                         82   \n",
       "17710             7229                 12980                         92   \n",
       "\n",
       "       question_length  question_length_byte  unique_caracters_question  \\\n",
       "14891               45                   129                         25   \n",
       "22871               50                   146                         29   \n",
       "1512                63                   185                         33   \n",
       "22803               42                   122                         28   \n",
       "13102               31                    89                         22   \n",
       "...                ...                   ...                        ...   \n",
       "19501               30                    55                         16   \n",
       "10280               20                    37                         13   \n",
       "3137                42                    78                         19   \n",
       "1043                58                   106                         19   \n",
       "17710               19                    34                         12   \n",
       "\n",
       "      language  nb_annotation_answer_candidates  \\\n",
       "14891     Thai                                1   \n",
       "22871     Thai                                1   \n",
       "1512      Thai                                1   \n",
       "22803     Thai                                1   \n",
       "13102     Thai                                1   \n",
       "...        ...                              ...   \n",
       "19501   Arabic                                1   \n",
       "10280   Arabic                                1   \n",
       "3137    Arabic                                1   \n",
       "1043    Arabic                                1   \n",
       "17710   Arabic                                1   \n",
       "\n",
       "       mean_annotation_answer_candidates  min_annotation_answer_candidates  \\\n",
       "14891                               49.0                                49   \n",
       "22871                               18.0                                18   \n",
       "1512                                 0.0                                 0   \n",
       "22803                                0.0                                 0   \n",
       "13102                                0.0                                 0   \n",
       "...                                  ...                               ...   \n",
       "19501                                4.0                                 4   \n",
       "10280                                8.0                                 8   \n",
       "3137                                96.0                                96   \n",
       "1043                                 0.0                                 0   \n",
       "17710                               36.0                                36   \n",
       "\n",
       "       max_annotation_answer_candidates  yes_no_question  \n",
       "14891                                49                0  \n",
       "22871                                18                0  \n",
       "1512                                  0                0  \n",
       "22803                                 0                0  \n",
       "13102                                 0                0  \n",
       "...                                 ...              ...  \n",
       "19501                                 4                0  \n",
       "10280                                 8                0  \n",
       "3137                                 96                0  \n",
       "1043                                  0                0  \n",
       "17710                                36                0  \n",
       "\n",
       "[29337 rows x 12 columns]"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "multi_metrics"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 66,
   "metadata": {},
   "outputs": [],
   "source": [
    "def group_by_language(metrics):\n",
    "    metrics_per_language = metrics.groupby(\"language\").agg(\n",
    "        {\n",
    "            \"document_length\": \"mean\",\n",
    "            \"document_length_byte\": \"mean\",\n",
    "            \"unique_caracters_document\": \"mean\",\n",
    "            \"question_length\": \"mean\",\n",
    "            \"question_length_byte\": \"mean\",\n",
    "            \"unique_caracters_question\": \"mean\",\n",
    "            \"nb_annotation_answer_candidates\": \"mean\",\n",
    "            \"mean_annotation_answer_candidates\": \"mean\",\n",
    "            \"min_annotation_answer_candidates\": \"min\",\n",
    "            \"max_annotation_answer_candidates\": \"max\",\n",
    "            \"yes_no_question\": \"mean\",\n",
    "        }\n",
    "    )\n",
    "    metrics_per_language[\"count\"] = metrics.groupby(\"language\").count()[\n",
    "        \"document_length\"\n",
    "    ]\n",
    "    return metrics_per_language\n",
    "\n",
    "\n",
    "train_metrics_per_language = group_by_language(train_metrics)\n",
    "arabic_metrics_per_language = group_by_language(arabic_metrics)\n",
    "multi_metrics_per_language = group_by_language(multi_metrics)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "language\n",
       "Arabic    22592\n",
       "Name: count, dtype: int64"
      ]
     },
     "execution_count": 28,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "a[\"Arabic\"] = a[\"Arabic\"] - arabic_metrics_per_language[\"count\"]\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 71,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/tmp/ipykernel_1181/1722202348.py:3: SettingWithCopyWarning: \n",
      "A value is trying to be set on a copy of a slice from a DataFrame\n",
      "\n",
      "See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
      "  a[\"Arabic\"] = a[\"Arabic\"] - arabic_metrics_per_language[\"count\"]\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1600x800 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig, axs = plt.subplots(1, 2, figsize=(16, 8))\n",
    "a = train_metrics_per_language[\"count\"]\n",
    "a[\"Arabic\"] = a[\"Arabic\"] - arabic_metrics_per_language[\"count\"]\n",
    "document_number_metric = pd.DataFrame(\n",
    "    data=[\n",
    "        multi_metrics_per_language[\"count\"],\n",
    "        arabic_metrics_per_language[\"count\"]\n",
    "        - multi_metrics_per_language[\"count\"][\"Arabic\"],\n",
    "        a - multi_metrics_per_language[\"count\"],\n",
    "    ],\n",
    "    index=[\n",
    "        \"Reduced Multi-Language set\",\n",
    "        \"Reduced Arabic set\",\n",
    "        \"Training set\",\n",
    "    ],\n",
    ").T.sort_values(\"language\", ascending=False)\n",
    "\n",
    "document_number_metric.plot(\n",
    "    kind=\"barh\",\n",
    "    stacked=True,\n",
    "    color=[\n",
    "        \"indianred\",\n",
    "        \"coral\",\n",
    "        \"cornflowerblue\",\n",
    "    ],\n",
    "    ax=axs[0],\n",
    ")\n",
    "axs[0].set_xscale(\"log\")\n",
    "axs[0].set_ylabel(\"Language\")\n",
    "axs[0].set_xlabel(\"# of documents\")\n",
    "axs[0].legend()\n",
    "handles, labels = axs[0].get_legend_handles_labels()\n",
    "order = [2, 1, 0]\n",
    "axs[0].legend([handles[idx] for idx in order],[labels[idx] for idx in order], loc=\"lower right\") \n",
    "\n",
    "# sns.histplot(\n",
    "#     train_metrics[train_metrics[\"language\"] == \"Arabic\"],\n",
    "#     x=\"document_length\",\n",
    "#     bins=50,\n",
    "#     ax=axs[1],\n",
    "#     color=\"cornflowerblue\",\n",
    "#     log_scale=True,\n",
    "#     label=\"Training set\",\n",
    "# )\n",
    "sns.histplot(\n",
    "    arabic_metrics,\n",
    "    x=\"document_length\",\n",
    "    bins=50,\n",
    "    ax=axs[1],\n",
    "    color=\"coral\",\n",
    "    log_scale=True,\n",
    "    label=\"Reduced Arabic set\",\n",
    ")\n",
    "sns.histplot(\n",
    "    multi_metrics[multi_metrics[\"language\"] == \"Arabic\"],\n",
    "    x=\"document_length\",\n",
    "    bins=50,\n",
    "    ax=axs[1],\n",
    "    color=\"indianred\",\n",
    "    log_scale=True,\n",
    "    label=\"Reduced Multi-Language set\",\n",
    ")\n",
    "handles, labels = plt.gca().get_legend_handles_labels()\n",
    "order = [0, 1]\n",
    "axs[1].legend(loc=\"lower right\") \n",
    "axs[1].set_xlabel(\"Arabic document length (# of caracters)\")\n",
    "axs[1].set_yscale(\"log\")\n",
    "plt.tight_layout()\n",
    "plt.savefig(\"figures/number_documents_length_doc.jpg\", dpi=300)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1600x800 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig, axs = plt.subplots(1, 2, figsize=(16, 8))\n",
    "\n",
    "ordered_languages = train_metrics_per_language.index.sort_values(ascending=True)\n",
    "\n",
    "sns.barplot(\n",
    "    train_metrics,\n",
    "    x=\"document_length\",\n",
    "    y=\"language\",\n",
    "    ax=axs[0],\n",
    "    color=\"cornflowerblue\",\n",
    "    order=ordered_languages,\n",
    ")\n",
    "axs[0].set_xlabel(\"Document length (# of caracters)\")\n",
    "axs[0].set_ylabel(\"Language\")\n",
    "sns.barplot(\n",
    "    train_metrics,\n",
    "    x=\"mean_annotation_answer_candidates\",\n",
    "    y=\"language\",\n",
    "    ax=axs[1],\n",
    "    color=\"cornflowerblue\",\n",
    "    order=ordered_languages,\n",
    ")\n",
    "axs[1].set_xlabel(\"Average length of annotations\")\n",
    "axs[1].set_ylabel(\"Language\")\n",
    "plt.tight_layout()\n",
    "plt.savefig(\"figures/doc_annot_length.jpg\", dpi=300)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 226,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1600x800 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig, axs = plt.subplots(1, 2, figsize=(16, 8))\n",
    "\n",
    "sns.barplot(\n",
    "    train_metrics,\n",
    "    x=\"mean_annotation_answer_candidates\",\n",
    "    y=\"language\",\n",
    "    ax=axs[0],\n",
    "    color=\"cornflowerblue\",\n",
    "    order=ordered_languages,\n",
    ")\n",
    "axs[0].set_xlabel(\"Average length of annotations\")\n",
    "axs[0].set_ylabel(\"Language\")\n",
    "sns.barplot(\n",
    "    train_metrics,\n",
    "    x=\"yes_no_question\",\n",
    "    y=\"language\",\n",
    "    ax=axs[1],\n",
    "    color=\"cornflowerblue\",\n",
    "    order=ordered_languages,\n",
    ")\n",
    "axs[1].set_xlabel(\"Ratio of Yes/No questions\")\n",
    "axs[1].set_ylabel(\"Language\")\n",
    "plt.tight_layout()\n",
    "plt.savefig(\"figures/train_length_ratio_annotations.jpg\", dpi=300)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 242,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 800x800 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig, axs = plt.subplots(1, 1, figsize=(8, 8))\n",
    "\n",
    "sns.histplot(\n",
    "    train_metrics[train_metrics[\"language\"] == \"Arabic\"],\n",
    "    x=\"document_length\",\n",
    "    bins=50,\n",
    "    ax=axs,\n",
    "    color=\"cornflowerblue\",\n",
    "    log_scale=True,\n",
    ")\n",
    "axs.set_xlabel(\"Arabic document length (# of caracters)\")\n",
    "# axs.set_ylabel(\"PDF\")\n",
    "plt.tight_layout()\n",
    "plt.savefig(\"figures/hist_arabic_doc_length.jpg\", dpi=300)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "166916it [01:02, 2668.14it/s]\n"
     ]
    }
   ],
   "source": [
    "dataset_path = \"train_samples/tydiqa-v1.0-train.jsonl\"\n",
    "\n",
    "with jsonlines.open(dataset_path) as dataset:\n",
    "    for sample in tqdm(dataset.iter(skip_invalid=True)):\n",
    "        if sample[\"language\"] == \"arabic\":\n",
    "            with jsonlines.open(\n",
    "                \"train_samples/tydiqa-v1.0-train_arabic.jsonl\", mode=\"a\"\n",
    "            ) as writer:\n",
    "                writer.write(sample)\n"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "MVA",
   "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": "3.9.8"
  },
  "orig_nbformat": 4
 },
 "nbformat": 4,
 "nbformat_minor": 2
}