notebooks/Analysis-Synsets_Cross-linguality.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import json\n",
    "import glob\n",
    "import time\n",
    "import pickle\n",
    "import jsonlines\n",
    "import numpy as np\n",
    "import pandas as pd\n",
    "from tqdm import tqdm\n",
    "from collections import Counter, defaultdict\n",
    "\n",
    "import seaborn as sns\n",
    "import matplotlib.pyplot as plt\n",
    "from matplotlib import rcParams\n",
    "rcParams['font.family'] = 'Times New Roman'\n",
    "\n",
    "colors = ['#0072B2', '#56B4E9', '#E69F00', '#CC79A7', '#009E73', '#F0E442']"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Load Synsets"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(1000, 2)\n"
     ]
    },
    {
     "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>concept</th>\n",
       "      <th>en wiki</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>English setter</td>\n",
       "      <td>https://en.wikipedia.org/wiki/English_Setter</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>grey whale, gray whale, devilfish, Eschrichtiu...</td>\n",
       "      <td>https://en.wikipedia.org/wiki/Gray_whale</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>Egyptian cat</td>\n",
       "      <td>https://en.wikipedia.org/wiki/Egyptian_Mau</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>Staffordshire bullterrier, Staffordshire bull ...</td>\n",
       "      <td>https://en.wikipedia.org/wiki/Staffordshire_Bu...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>miniature poodle</td>\n",
       "      <td>https://en.wikipedia.org/wiki/Poodle</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                             concept  \\\n",
       "0                                     English setter   \n",
       "1  grey whale, gray whale, devilfish, Eschrichtiu...   \n",
       "2                                       Egyptian cat   \n",
       "3  Staffordshire bullterrier, Staffordshire bull ...   \n",
       "4                                   miniature poodle   \n",
       "\n",
       "                                             en wiki  \n",
       "0       https://en.wikipedia.org/wiki/English_Setter  \n",
       "1           https://en.wikipedia.org/wiki/Gray_whale  \n",
       "2         https://en.wikipedia.org/wiki/Egyptian_Mau  \n",
       "3  https://en.wikipedia.org/wiki/Staffordshire_Bu...  \n",
       "4               https://en.wikipedia.org/wiki/Poodle  "
      ]
     },
     "execution_count": 2,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# ImageNet 1k concepts manually mapped onto Wikipedia pages\n",
    "synsets_df = pd.read_csv(\"synset_to_en-wiki.csv\", \"\\t\")\n",
    "print(synsets_df.shape)\n",
    "synsets_df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "name2wiki = {e['concept']: e['en wiki'] for e in synsets_df.to_dict('records')}"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "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>concept</th>\n",
       "      <th>en wiki</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>cheetah</td>\n",
       "      <td>https://en.wikipedia.org/wiki/Cheetah</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>flute</td>\n",
       "      <td>https://en.wikipedia.org/wiki/Flute</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>goose</td>\n",
       "      <td>https://en.wikipedia.org/wiki/Goose</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>miniature schnauzer</td>\n",
       "      <td>https://en.wikipedia.org/wiki/Miniature_Schnauzer</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>knee pad</td>\n",
       "      <td>https://en.wikipedia.org/wiki/Knee_pad</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "               concept                                            en wiki\n",
       "0              cheetah              https://en.wikipedia.org/wiki/Cheetah\n",
       "1                flute                https://en.wikipedia.org/wiki/Flute\n",
       "2                goose                https://en.wikipedia.org/wiki/Goose\n",
       "3  miniature schnauzer  https://en.wikipedia.org/wiki/Miniature_Schnauzer\n",
       "4             knee pad             https://en.wikipedia.org/wiki/Knee_pad"
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# NLVR2 synsets\n",
    "nlvr_synsets = set()\n",
    "with jsonlines.open('../data/en/annotations/train.jsonl') as reader:\n",
    "    for obj in reader:\n",
    "        nlvr_synsets.add(obj['synset'])\n",
    "len(nlvr_synsets)\n",
    "\n",
    "nlvr_wikis = defaultdict(list)\n",
    "s_synsets = set(synsets_df['concept'])\n",
    "for syn in nlvr_synsets:\n",
    "    for s in s_synsets:\n",
    "        if syn in {e.strip() for e in s.split(',')}:\n",
    "            nlvr_wikis[syn].append(name2wiki[s])\n",
    "            \n",
    "nlvr_synsets_df = pd.DataFrame.from_dict(nlvr_wikis, orient='index')\n",
    "nlvr_synsets_df = nlvr_synsets_df.reset_index().rename(columns={'index': 'concept', 0: 'en wiki'})\n",
    "nlvr_synsets_df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "id 95\n",
      "sw 78\n",
      "ta 83\n",
      "tr 79\n",
      "zh 94\n"
     ]
    },
    {
     "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>concept</th>\n",
       "      <th>lang</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>Rambutan</td>\n",
       "      <td>id</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>Renang_(olahraga)</td>\n",
       "      <td>id</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>Agama_Buddha</td>\n",
       "      <td>id</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>Patung</td>\n",
       "      <td>id</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>Bajak</td>\n",
       "      <td>id</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "             concept lang\n",
       "0           Rambutan   id\n",
       "1  Renang_(olahraga)   id\n",
       "2       Agama_Buddha   id\n",
       "3             Patung   id\n",
       "4              Bajak   id"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# MaRVL synsets\n",
    "\n",
    "# id\n",
    "with jsonlines.open(\"../data/id/annotations/marvl-id.jsonl\") as reader:\n",
    "    id_concepts = set([item['concept'].split(\"-\")[-1].strip() for item in reader])\n",
    "print('id', len(id_concepts))\n",
    "    \n",
    "# sw\n",
    "with jsonlines.open(\"../data/sw/annotations/marvl-sw.jsonl\") as reader:\n",
    "    sw_concepts = set([item['concept'].split(\"-\")[-1].strip() for item in reader])\n",
    "print('sw', len(sw_concepts))\n",
    "\n",
    "# ta\n",
    "with jsonlines.open(\"../data/ta/annotations/marvl-ta.jsonl\") as reader:\n",
    "    ta_concepts = set([item['concept'].split(\"-\")[-1].strip() for item in reader])\n",
    "print('ta', len(ta_concepts))\n",
    "    \n",
    "# tr\n",
    "with jsonlines.open(\"../data/tr/annotations/marvl-tr.jsonl\") as reader:\n",
    "    tr_concepts = set([item['concept'].split(\"-\")[-1].strip() for item in reader])\n",
    "print('tr', len(tr_concepts))\n",
    "\n",
    "# zh\n",
    "with jsonlines.open(\"../data/zh/annotations/marvl-zh.jsonl\") as reader:\n",
    "    zh_concepts = set([item['concept'].split(\"-\")[-1].strip() for item in reader])\n",
    "print('zh', len(zh_concepts))\n",
    "\n",
    "# Synsets DataFrame\n",
    "marvl_synsets_df = pd.DataFrame()\n",
    "marvl_synsets_df['concept'] = list(id_concepts) + list(sw_concepts) + list(ta_concepts) + list(tr_concepts) + list(zh_concepts)\n",
    "marvl_synsets_df['lang'] = ['id']*len(id_concepts) + ['sw']*len(sw_concepts) + ['ta']*len(ta_concepts) + ['tr']*len(tr_concepts) + ['zh']*len(zh_concepts)\n",
    "marvl_synsets_df.head()\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Get cross-lingual stats"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "import requests\n",
    "\n",
    "def get_json(url):\n",
    "    u = requests.get(url=url)\n",
    "    return u.json()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# ImageNet\n",
    "synset2langs = defaultdict(list)\n",
    "synset2nlangs = dict()\n",
    "for ix, row in tqdm(synsets_df.iterrows(), total=len(synsets_df)):\n",
    "    s, en_url = row\n",
    "    api_url = f\"https://en.wikipedia.org/w/api.php?action=query&titles={en_url.split('/')[-1]}&prop=langlinks&lllimit=500&format=json\"\n",
    "    data = get_json(api_url)\n",
    "    try:\n",
    "        result = list(data[\"query\"][\"pages\"].values())[0][\"langlinks\"]\n",
    "\n",
    "        count = len(result)\n",
    "        synset2nlangs[s] = count\n",
    "\n",
    "        for res in result:\n",
    "            synset2langs[s].append(res['lang'])\n",
    "    except:\n",
    "        print(s)\n",
    "        synset2nlangs[s] = 0\n",
    "        synset2langs[s] = []\n",
    "\n",
    "with open(\"synset2langs.pkl\", \"wb\") as f:\n",
    "    pickle.dump(synset2langs, f)\n",
    "    \n",
    "with open(\"synset2nlangs.pkl\", \"wb\") as f:\n",
    "    pickle.dump(synset2nlangs, f)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# NLVR2\n",
    "nlvr_synset2langs = defaultdict(list)\n",
    "nlvr_synset2nlangs = dict()\n",
    "for ix, row in tqdm(nlvr_synsets_df.iterrows(), total=len(nlvr_synsets_df)):\n",
    "    s, en_url = row\n",
    "    api_url = f\"https://en.wikipedia.org/w/api.php?action=query&titles={en_url.split('/')[-1]}&prop=langlinks&lllimit=500&format=json\"\n",
    "    data = get_json(api_url)\n",
    "    try:\n",
    "        result = list(data[\"query\"][\"pages\"].values())[0][\"langlinks\"]\n",
    "\n",
    "        count = len(result)\n",
    "        nlvr_synset2nlangs[s] = count\n",
    "\n",
    "        for res in result:\n",
    "            nlvr_synset2langs[s].append(res['lang'])\n",
    "    except:\n",
    "        print(s)\n",
    "        nlvr_synset2nlangs[s] = 0\n",
    "        nlvr_synset2langs[s] = []\n",
    "\n",
    "with open(\"nlvr_synset2langs.pkl\", \"wb\") as f:\n",
    "    pickle.dump(nlvr_synset2langs, f)\n",
    "    \n",
    "with open(\"nlvr_synset2nlangs.pkl\", \"wb\") as f:\n",
    "    pickle.dump(nlvr_synset2nlangs, f)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# MaRVL\n",
    "marvl_synset2langs = defaultdict(list)\n",
    "marvl_synset2nlangs = dict()\n",
    "for ix, row in tqdm(marvl_synsets_df.iterrows(), total=len(marvl_synsets_df)):\n",
    "    s, l = row\n",
    "    api_url = f\"https://{l}.wikipedia.org/w/api.php?action=query&titles={s}&prop=langlinks&lllimit=500&format=json\"\n",
    "    data = get_json(api_url)\n",
    "    try:\n",
    "        result = list(data[\"query\"][\"pages\"].values())[0][\"langlinks\"]\n",
    "\n",
    "        count = len(result)\n",
    "        marvl_synset2nlangs[s] = count + 1\n",
    "\n",
    "        marvl_synset2langs[s].append(l)\n",
    "        for res in result:\n",
    "            marvl_synset2langs[s].append(res['lang'])\n",
    "    except:\n",
    "        time.sleep(1)\n",
    "        try:\n",
    "            result = list(data[\"query\"][\"pages\"].values())[0][\"langlinks\"]\n",
    "\n",
    "            count = len(result)\n",
    "            marvl_synset2nlangs[s] = count + 1\n",
    "\n",
    "            marvl_synset2langs[s].append(l)\n",
    "            for res in result:\n",
    "                marvl_synset2langs[s].append(res['lang'])\n",
    "        except:\n",
    "            print(s)\n",
    "            marvl_synset2nlangs[s] = 0\n",
    "            marvl_synset2langs[s] = []\n",
    "\n",
    "with open(\"marvl_synset2langs.pkl\", \"wb\") as f:\n",
    "    pickle.dump(marvl_synset2langs, f)\n",
    "    \n",
    "with open(\"marvl_synset2nlangs.pkl\", \"wb\") as f:\n",
    "    pickle.dump(marvl_synset2nlangs, f)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Load cross-lingual stats"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "ImageNet: 999\n",
      "NLVR2: 124\n",
      "MaRVL: 445\n"
     ]
    }
   ],
   "source": [
    "# Load ImageNet cross-lingual stats\n",
    "inet_synset2langs = pickle.load(open(\"synset2langs.pkl\", \"rb\"))\n",
    "inet_synset2nlangs = pickle.load(open(\"synset2nlangs.pkl\", \"rb\"))\n",
    "print(\"ImageNet:\", len(inet_synset2nlangs))\n",
    "\n",
    "# Load NLVR2 cross-lingual stats\n",
    "nlvr_synset2langs = pickle.load(open(\"nlvr_synset2langs.pkl\", \"rb\"))\n",
    "nlvr_synset2nlangs = pickle.load(open(\"nlvr_synset2nlangs.pkl\", \"rb\"))\n",
    "print(\"NLVR2:\", len(nlvr_synset2nlangs))\n",
    "\n",
    "# Load MaRVL cross-lingual stats\n",
    "marvl_synset2langs = pickle.load(open(\"marvl_synset2langs.pkl\", \"rb\"))\n",
    "marvl_synset2nlangs = pickle.load(open(\"marvl_synset2nlangs.pkl\", \"rb\"))\n",
    "print(\"MaRVL:\", len(marvl_synset2nlangs))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Add English language\n",
    "for s in inet_synset2langs:\n",
    "    inet_synset2langs[s] += ['en']\n",
    "    inet_synset2nlangs[s] += 1\n",
    "    \n",
    "for s in nlvr_synset2langs:\n",
    "    nlvr_synset2langs[s] += ['en']\n",
    "    nlvr_synset2nlangs[s] += 1\n",
    "    \n",
    "for s in marvl_synset2langs:\n",
    "    marvl_synset2langs[s] += ['en']\n",
    "    marvl_synset2nlangs[s] += 1"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "ImageNet: 305\n",
      "NLVR2: 273\n",
      "MaRVL: 311\n"
     ]
    }
   ],
   "source": [
    "# Get list of languages spanned by each dataset\n",
    "inet_langs = set()\n",
    "for s, lst in inet_synset2langs.items():\n",
    "    for l in lst:\n",
    "        inet_langs.add(l)\n",
    "print(\"ImageNet:\", len(inet_langs))\n",
    "\n",
    "nlvr_langs = set()\n",
    "for s, lst in nlvr_synset2langs.items():\n",
    "    for l in lst:\n",
    "        nlvr_langs.add(l)\n",
    "print(\"NLVR2:\", len(nlvr_langs))\n",
    "\n",
    "marvl_langs = set()\n",
    "for s, lst in marvl_synset2langs.items():\n",
    "    for l in lst:\n",
    "        marvl_langs.add(l)\n",
    "print(\"MaRVL:\", len(marvl_langs))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "ImageNet: 292\n",
      "NLVR2: 262\n",
      "MaRVL: 298\n"
     ]
    }
   ],
   "source": [
    "# Map each language to its ISO codes\n",
    "from iso639 import Lang  # pip install iso639-lang\n",
    "\n",
    "inet_lang2codes = dict()\n",
    "for l in sorted(inet_langs):\n",
    "    try:\n",
    "        inet_lang2codes[l] = Lang(l)\n",
    "    except:\n",
    "        pass\n",
    "print(\"ImageNet:\", len(inet_lang2codes))\n",
    "\n",
    "nlvr_lang2codes = dict()\n",
    "for l in sorted(nlvr_langs):\n",
    "    try:\n",
    "        nlvr_lang2codes[l] = Lang(l)\n",
    "    except:\n",
    "        pass\n",
    "print(\"NLVR2:\", len(nlvr_lang2codes))\n",
    "\n",
    "marvl_lang2codes = dict()\n",
    "for l in sorted(marvl_langs):\n",
    "    try:\n",
    "        marvl_lang2codes[l] = Lang(l)\n",
    "    except:\n",
    "        pass\n",
    "print(\"MaRVL:\", len(marvl_lang2codes))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "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>Name</th>\n",
       "      <th>Macroarea</th>\n",
       "      <th>Latitude</th>\n",
       "      <th>Longitude</th>\n",
       "      <th>Glottocode</th>\n",
       "      <th>ISO639P3code</th>\n",
       "      <th>Family</th>\n",
       "      <th>Subfamily</th>\n",
       "      <th>Genus</th>\n",
       "      <th>GenusIcon</th>\n",
       "      <th>ISO_codes</th>\n",
       "      <th>Samples_100</th>\n",
       "      <th>Samples_200</th>\n",
       "      <th>Country_ID</th>\n",
       "      <th>Source</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>aab</td>\n",
       "      <td>Arapesh (Abu)</td>\n",
       "      <td>NaN</td>\n",
       "      <td>-3.450000</td>\n",
       "      <td>142.950000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Torricelli</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Kombio-Arapesh</td>\n",
       "      <td>c0000dd</td>\n",
       "      <td>NaN</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>PG</td>\n",
       "      <td>Nekitel-1985</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>aar</td>\n",
       "      <td>Aari</td>\n",
       "      <td>Africa</td>\n",
       "      <td>6.000000</td>\n",
       "      <td>36.583333</td>\n",
       "      <td>aari1239</td>\n",
       "      <td>aiw</td>\n",
       "      <td>Afro-Asiatic</td>\n",
       "      <td>Omotic</td>\n",
       "      <td>South Omotic</td>\n",
       "      <td>ccccccc</td>\n",
       "      <td>aiw</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>ET</td>\n",
       "      <td>Hayward-1990a</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>aba</td>\n",
       "      <td>Abau</td>\n",
       "      <td>Papunesia</td>\n",
       "      <td>-4.000000</td>\n",
       "      <td>141.250000</td>\n",
       "      <td>abau1245</td>\n",
       "      <td>aau</td>\n",
       "      <td>Sepik</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Upper Sepik</td>\n",
       "      <td>ddd0000</td>\n",
       "      <td>aau</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>PG</td>\n",
       "      <td>Bailey-1975</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>abb</td>\n",
       "      <td>Arabic (Chadian)</td>\n",
       "      <td>Africa</td>\n",
       "      <td>13.833333</td>\n",
       "      <td>20.833333</td>\n",
       "      <td>chad1249</td>\n",
       "      <td>shu</td>\n",
       "      <td>Afro-Asiatic</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Semitic</td>\n",
       "      <td>cdd0000</td>\n",
       "      <td>shu</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>TD</td>\n",
       "      <td>Abu-Absi-1995</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>abd</td>\n",
       "      <td>Abidji</td>\n",
       "      <td>Africa</td>\n",
       "      <td>5.666667</td>\n",
       "      <td>-4.583333</td>\n",
       "      <td>abid1235</td>\n",
       "      <td>abi</td>\n",
       "      <td>Niger-Congo</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Kwa</td>\n",
       "      <td>d009900</td>\n",
       "      <td>abi</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>CI</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "    ID              Name  Macroarea   Latitude   Longitude Glottocode  \\\n",
       "0  aab     Arapesh (Abu)        NaN  -3.450000  142.950000        NaN   \n",
       "1  aar              Aari     Africa   6.000000   36.583333   aari1239   \n",
       "2  aba              Abau  Papunesia  -4.000000  141.250000   abau1245   \n",
       "3  abb  Arabic (Chadian)     Africa  13.833333   20.833333   chad1249   \n",
       "4  abd            Abidji     Africa   5.666667   -4.583333   abid1235   \n",
       "\n",
       "  ISO639P3code        Family Subfamily           Genus GenusIcon ISO_codes  \\\n",
       "0          NaN    Torricelli       NaN  Kombio-Arapesh   c0000dd       NaN   \n",
       "1          aiw  Afro-Asiatic    Omotic    South Omotic   ccccccc       aiw   \n",
       "2          aau         Sepik       NaN     Upper Sepik   ddd0000       aau   \n",
       "3          shu  Afro-Asiatic       NaN         Semitic   cdd0000       shu   \n",
       "4          abi   Niger-Congo       NaN             Kwa   d009900       abi   \n",
       "\n",
       "   Samples_100  Samples_200 Country_ID         Source  \n",
       "0        False        False         PG   Nekitel-1985  \n",
       "1        False        False         ET  Hayward-1990a  \n",
       "2        False        False         PG    Bailey-1975  \n",
       "3        False        False         TD  Abu-Absi-1995  \n",
       "4        False        False         CI            NaN  "
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Load WALS features\n",
    "langs_df = pd.read_csv(\"languages.csv\")\n",
    "langs_df.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Analysis per Language"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 576x432 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "f, ax = plt.subplots(figsize=(8,6))\n",
    "\n",
    "x, y = np.unique(sorted([v for v in inet_synset2nlangs.values() if v > 0]), return_counts=True)\n",
    "sns.kdeplot(x=[v for v in inet_synset2nlangs.values() if v > 0], color=colors[0], lw=4, ls='-', label='ImageNet', ax=ax)\n",
    "\n",
    "x, y = np.unique(sorted([v for v in nlvr_synset2nlangs.values() if v > 0]), return_counts=True)\n",
    "sns.kdeplot(x=[v for v in nlvr_synset2nlangs.values() if v > 0], color=colors[1], lw=4, ls='--', label='NLVR2', ax=ax)\n",
    "\n",
    "x, y = np.unique(sorted([v for v in marvl_synset2nlangs.values() if v > 0]), return_counts=True)\n",
    "sns.kdeplot(x=[v for v in marvl_synset2nlangs.values() if v > 0],  color=colors[2], lw=4, ls='-.', label='MaRVL', ax=ax)\n",
    "\n",
    "ax.set_xlabel(\"# Languages per synset\", fontsize=32)\n",
    "ax.set_ylabel(\"Density\", fontsize=32)\n",
    "ax.set_xlim(0)\n",
    "ax.tick_params(axis='both', which='major', labelsize=24)\n",
    "ax.grid(alpha=0.2, axis='y')\n",
    "ax.legend(fontsize=24)\n",
    "\n",
    "f.savefig(\"synsets_languages.pdf\", bbox_inches=\"tight\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Analisys per Macroarea"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "'Eurasia'"
      ]
     },
     "execution_count": 12,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Map language names onto macroarea\n",
    "lang2area = {r['ISO639P3code']: r['Macroarea'] for _, r in langs_df.iterrows()}\n",
    "lang2area['ita']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "'Eurasia'"
      ]
     },
     "execution_count": 13,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Map language ids onto macroarea\n",
    "id2area = {r['ID']: r['Macroarea'] for _, r in langs_df.iterrows()}\n",
    "id2area['ita']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "ImageNet: 59\n",
      "NLVR2: 51\n",
      "MaRVL: 67\n"
     ]
    }
   ],
   "source": [
    "# number of ImageNet synsets per macroarea\n",
    "inet_area2synsets = defaultdict(list)\n",
    "missed_langs = set()\n",
    "for s, lst in inet_synset2langs.items():\n",
    "    synset_areas = set()  # counts 1 even if in many languages\n",
    "    for l in lst:\n",
    "        try:\n",
    "            a = lang2area[inet_lang2codes[l].pt3]\n",
    "            synset_areas.add(a)\n",
    "        except:\n",
    "            try:\n",
    "                a = id2area[inet_lang2codes[l].pt3]\n",
    "                synset_areas.add(a)\n",
    "            except:\n",
    "                missed_langs.add(l)\n",
    "    for a in synset_areas:\n",
    "        inet_area2synsets[a].append(s)\n",
    "print(\"ImageNet:\", len(missed_langs))\n",
    "\n",
    "# number of NLVR2 synsets per macroarea\n",
    "nlvr_area2synsets = defaultdict(list)\n",
    "missed_langs = set()\n",
    "for s, lst in nlvr_synset2langs.items():\n",
    "    synset_areas = set()  # counts 1 even if in many languages\n",
    "    for l in lst:\n",
    "        try:\n",
    "            a = lang2area[nlvr_lang2codes[l].pt3]\n",
    "            synset_areas.add(a)\n",
    "        except:\n",
    "            try:\n",
    "                a = id2area[nlvr_lang2codes[l].pt3]\n",
    "                synset_areas.add(a)\n",
    "            except:\n",
    "                missed_langs.add(l)\n",
    "    for a in synset_areas:\n",
    "        nlvr_area2synsets[a].append(s)\n",
    "print(\"NLVR2:\", len(missed_langs))\n",
    "\n",
    "# number of MaRVL synsets per macroarea\n",
    "marvl_area2synsets = defaultdict(list)\n",
    "missed_langs = set()\n",
    "for s, lst in marvl_synset2langs.items():\n",
    "    synset_areas = set()  # counts 1 even if in many languages\n",
    "    for l in lst:\n",
    "        try:\n",
    "            a = lang2area[marvl_lang2codes[l].pt3]\n",
    "            synset_areas.add(a)\n",
    "        except:\n",
    "            try:\n",
    "                a = id2area[nlvr_lang2codes[l].pt3]\n",
    "                synset_areas.add(a)\n",
    "            except:\n",
    "                missed_langs.add(l)\n",
    "    for a in synset_areas:\n",
    "        marvl_area2synsets[a].append(s)\n",
    "print(\"MaRVL:\", len(missed_langs))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "<ipython-input-15-c946bed78da9>:12: UserWarning: FixedFormatter should only be used together with FixedLocator\n",
      "  ax.set_yticklabels([''] + names, fontsize=28, rotation=0);\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 576x432 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "f, ax = plt.subplots(figsize=(8,6))\n",
    "w = 0.3\n",
    "\n",
    "names = ['Eurasia', 'Papunesia', 'South America', 'Africa', 'North America', 'Australia']\n",
    "xs = np.array(range(len(names)))\n",
    "ax.barh(xs - w, [100*len(inet_area2synsets[k])/len(inet_synset2langs) for k in names], height=w, color=colors[0], ec='k', label='ImageNet')\n",
    "ax.barh(xs, [100*len(nlvr_area2synsets[k])/len(nlvr_synset2langs) for k in names], height=w, color=colors[1], ec='k', hatch='.', label='NLVR2')\n",
    "ax.barh(xs + w, [100*len(marvl_area2synsets[k])/len(marvl_synset2langs) for k in names], height=w, color=colors[2], ec='k', hatch='/', label='MaRVL')\n",
    "\n",
    "ax.set_xlabel(\"Synsets per macro-area [%]\", fontsize=32)\n",
    "ax.tick_params(axis='x', which='major', labelsize=24)\n",
    "ax.set_yticklabels([''] + names, fontsize=28, rotation=0);\n",
    "ax.grid(alpha=0.2, axis='x')\n",
    "ax.legend(fontsize=18)\n",
    "\n",
    "f.savefig(\"synsets_macroareas.pdf\", bbox_inches=\"tight\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Analysis per Family"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "'Indo-European'"
      ]
     },
     "execution_count": 16,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Map language names onto families\n",
    "lang2fam = {r['ISO639P3code']: r['Family'] for _, r in langs_df.iterrows()}\n",
    "lang2fam['ita']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "'Indo-European'"
      ]
     },
     "execution_count": 17,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Map language ids onto families\n",
    "id2fam = {r['ID']: r['Family'] for _, r in langs_df.iterrows()}\n",
    "id2fam['ita']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "ImageNet: 35\n",
      "NLVR2: 32\n",
      "MaRVL: 35\n"
     ]
    }
   ],
   "source": [
    "# number of ImageNet synsets per family\n",
    "inet_fam2synsets = defaultdict(list)\n",
    "for s, lst in inet_synset2langs.items():\n",
    "    synset_fams = set()  # counts 1 even if in many languages\n",
    "    for l in lst:\n",
    "        try:\n",
    "            a = lang2fam[inet_lang2codes[l].pt3]\n",
    "            if type(a) == str:\n",
    "                synset_fams.add(a)\n",
    "        except:\n",
    "            try:\n",
    "                a = id2fam[inet_lang2codes[l].pt3]\n",
    "                if type(a) == str:\n",
    "                    synset_fams.add(a)\n",
    "            except:\n",
    "                pass\n",
    "    for a in synset_fams:\n",
    "        inet_fam2synsets[a].append(s)\n",
    "print(\"ImageNet:\", len(inet_fam2synsets))\n",
    "        \n",
    "# number of NLVR2 synsets per family\n",
    "nlvr_fam2synsets = defaultdict(list)\n",
    "for s, lst in nlvr_synset2langs.items():\n",
    "    synset_fams = set()  # counts 1 even if in many languages\n",
    "    for l in lst:\n",
    "        try:\n",
    "            a = lang2fam[nlvr_lang2codes[l].pt3]\n",
    "            if type(a) == str:\n",
    "                synset_fams.add(a)\n",
    "        except:\n",
    "            try:\n",
    "                a = id2fam[nlvr_lang2codes[l].pt3]\n",
    "                if type(a) == str:\n",
    "                    synset_fams.add(a)\n",
    "            except:\n",
    "                pass\n",
    "    for a in synset_fams:\n",
    "        nlvr_fam2synsets[a].append(s)\n",
    "print(\"NLVR2:\", len(nlvr_fam2synsets))\n",
    "        \n",
    "# number of MaRVL synsets per family\n",
    "marvl_fam2synsets = defaultdict(list)\n",
    "for s, lst in marvl_synset2langs.items():\n",
    "    synset_fams = set()  # counts 1 even if in many languages\n",
    "    for l in lst:\n",
    "        try:\n",
    "            a = lang2fam[marvl_lang2codes[l].pt3]\n",
    "            if type(a) == str:\n",
    "                synset_fams.add(a)\n",
    "        except:\n",
    "            try:\n",
    "                a = id2fam[marvl_lang2codes[l].pt3]\n",
    "                if type(a) == str:\n",
    "                    synset_fams.add(a)\n",
    "            except:\n",
    "                pass\n",
    "    for a in synset_fams:\n",
    "        marvl_fam2synsets[a].append(s)\n",
    "print(\"MaRVL:\", len(marvl_fam2synsets))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "<ipython-input-19-2496d7eebf66>:13: UserWarning: FixedFormatter should only be used together with FixedLocator\n",
      "  ax.set_xticklabels([0, 20, 40, 60, 80, 100])\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 576x432 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "f, ax = plt.subplots(figsize=(8,6))\n",
    "\n",
    "names = list(inet_fam2synsets.keys())\n",
    "\n",
    "sns.kdeplot(x=[len(inet_fam2synsets[k])/len(inet_synset2langs) for k in names], color=colors[0], lw=4, ls='-', label='ImageNet', ax=ax)\n",
    "sns.kdeplot(x=[len(nlvr_fam2synsets[k])/len(nlvr_synset2langs) for k in names], color=colors[1], lw=4, ls='--', label='NLVR2', ax=ax)\n",
    "sns.kdeplot(x=[len(marvl_fam2synsets[k])/len(marvl_synset2langs) for k in names], color=colors[2], lw=4, ls='-.', label='MaRVL', ax=ax)\n",
    "\n",
    "ax.set_xlabel(\"Synsets per language family [%]\", fontsize=32)\n",
    "ax.set_ylabel(\"Density\", fontsize=32)\n",
    "ax.set_xlim(0, 1)\n",
    "ax.tick_params(axis='both', which='major', labelsize=24)\n",
    "ax.set_xticklabels([0, 20, 40, 60, 80, 100])\n",
    "ax.grid(alpha=0.2, axis='y')\n",
    "ax.legend(fontsize=24)\n",
    "\n",
    "f.savefig(\"synsets_families.pdf\", bbox_inches=\"tight\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "x"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.8.8"
  },
  "latex_envs": {
   "LaTeX_envs_menu_present": true,
   "autoclose": false,
   "autocomplete": true,
   "bibliofile": "biblio.bib",
   "cite_by": "apalike",
   "current_citInitial": 1,
   "eqLabelWithNumbers": true,
   "eqNumInitial": 1,
   "hotkeys": {
    "equation": "Ctrl-E",
    "itemize": "Ctrl-I"
   },
   "labels_anchors": false,
   "latex_user_defs": false,
   "report_style_numbering": false,
   "user_envs_cfg": false
  },
  "toc": {
   "base_numbering": 1,
   "nav_menu": {},
   "number_sections": true,
   "sideBar": true,
   "skip_h1_title": false,
   "title_cell": "Table of Contents",
   "title_sidebar": "Contents",
   "toc_cell": false,
   "toc_position": {
    "height": "calc(100% - 180px)",
    "left": "10px",
    "top": "150px",
    "width": "241.4705810546875px"
   },
   "toc_section_display": true,
   "toc_window_display": true
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}