Align Experiments.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Aligning Parameters\n",
    "\n",
    "In this notebook we are taking one language (for example Spanish), reading the output from ExtractNamedTempates and:\n",
    "\n",
    "* Counting templates frequency\n",
    "* Counting parameters frequency\n",
    "* Getting the pages where templates occurs\n",
    "* Getting a second language (Ex: Catalan), trying to find the equivalent template (using Wikidata)\n",
    "* Comparing the paramers and values in those pairs of language"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Define basic functions"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 50,
   "metadata": {},
   "outputs": [],
   "source": [
    "%matplotlib inline\n",
    "from matplotlib import pyplot\n",
    "import requests\n",
    "import json\n",
    "import mwparserfromhell\n",
    "from scipy.spatial import distance\n",
    "from collections import Counter\n",
    "\n",
    "def getWikidataPair(titles,lang,target):\n",
    "    \"\"\"\n",
    "    titles: list of pages titles\n",
    "    lang:source lang (same of the titles)\n",
    "    target: target lang\n",
    "    returns a list: page_title_source,page_title_target\n",
    "    \"\"\"\n",
    "    response= requests.get(\"https://www.wikidata.org/w/api.php?action=wbgetentities&sites=%swiki&titles=%s&props=sitelinks&format=json\" % (lang,'|'.join(titles)))\n",
    "    output = []\n",
    "    for entity,data in  response.json()['entities'].items():\n",
    "        if data['sitelinks'].get(target+'wiki'):\n",
    "            s=data['sitelinks'][lang+'wiki']['title']\n",
    "            t=data['sitelinks'][target+'wiki']['title']\n",
    "            output.append([s,t])\n",
    "    return output\n",
    "\n",
    "def getWikidataID(titles,lang):\n",
    "    \"\"\"\n",
    "    titles: list of pages titles\n",
    "    lang:source lang (same of the titles)\n",
    "    target: target lang\n",
    "    returns a list: page_title_source,page_title_target\n",
    "    \"\"\"\n",
    "    response= requests.get(\"https://www.wikidata.org/w/api.php?action=wbgetentities&sites=%swiki&titles=%s&props=sitelinks&format=json\" % (lang,'|'.join(titles)))\n",
    "    output = []\n",
    "    for entity,data in  response.json()['entities'].items():\n",
    "        if data['sitelinks'].get(target+'wiki'):\n",
    "            s=data['sitelinks'][lang+'wiki']['title']\n",
    "            t=data['sitelinks'][target+'wiki']['title']\n",
    "            output.append([s,t])\n",
    "    return output\n",
    "\n",
    "\n",
    "#Get articles\n",
    "def getContent(title,lang):\n",
    "    \"\"\"\n",
    "    title: page title\n",
    "    lang: lang\n",
    "    returns wikitext\n",
    "    \"\"\"\n",
    "    url = \"https://%s.wikipedia.org/w/api.php?action=query&prop=revisions&rvprop=content&format=json&formatversion=2&titles=%s\" % (lang,title)\n",
    "    response = requests.get(url)\n",
    "    content = response.json()['query']['pages'][0]['revisions'][0]['content']\n",
    "    return content\n",
    "\n",
    "#get all the templates with named parameters\n",
    "def extract_templates(text):\n",
    "    wikicode = mwparserfromhell.parse(text)\n",
    "    tmpdict = {}\n",
    "    for template in wikicode.filter_templates():\n",
    "        if template.params:\n",
    "            values = dict([[t.name.strip(),t.value.strip()] for t in template.params if t.showkey])\n",
    "            if values:\n",
    "                tmpdict[template.name.strip()] = values\n",
    "    return tmpdict\n",
    "\n",
    "def getTemplateData(template,lang):\n",
    "    url =  \"https://%s.wikipedia.org/w/api.php?action=templatedata&titles=%s&formatversion=2&redirects=1\" % (lang,template)\n",
    "    #print(url)\n",
    "    resp = requests.get(url)\n",
    "    data = list(resp.json()['pages'].values())[0]\n",
    "    return data\n",
    "\n",
    "\n",
    "def apply_transform(vec, transform):\n",
    "        \"\"\"\n",
    "        Apply the given transformation to the vector space\n",
    "\n",
    "        Right-multiplies given transform with embeddings E:\n",
    "            E = E * transform\n",
    "\n",
    "        Transform can either be a string with a filename to a\n",
    "        text file containing a ndarray (compat. with np.loadtxt)\n",
    "        or a numpy ndarray.\n",
    "        \"\"\"\n",
    "        transmat = np.loadtxt(transform) if isinstance(transform, str) else transform\n",
    "        return np.matmul(vec, transmat)\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Aggregating templates coming from ExtractNamedTempates.ipynb\n",
    "\n",
    "Note that all templantes name are capitalized and parameters are in low case"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "lang = 'es'\n",
    "f = open('templates-articles_%s.json' % lang)\n",
    "templates = {}\n",
    "params = {}\n",
    "for l in f:\n",
    "    tmp = json.loads(l)\n",
    "    page = tmp[1]\n",
    "    for template,param in tmp[2].items():\n",
    "        template = template.capitalize()\n",
    "        templates[template] = templates.get(template,{'Tcount':0,'Tpages':[],'Params':{}})\n",
    "        templates[template]['Tcount'] += 1\n",
    "        templates[template]['Tpages'].append(page)\n",
    "        for name,val in param.items():\n",
    "            name = name.lower()\n",
    "            templates[template]['Params'][name] =templates[template]['Params'].get(name,0)\n",
    "            templates[template]['Params'][name] +=1\n",
    "            params[name] = params.get(name,0)\n",
    "            params[name] += 1"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "targetLang = 'en'\n",
    "f = open('templates-articles_%s.json' % targetLang)\n",
    "templatesTarget = {}\n",
    "paramsTarget = {}\n",
    "for l in f:\n",
    "    tmp = json.loads(l)\n",
    "    page = tmp[1]\n",
    "    for template,param in tmp[2].items():\n",
    "        template = template.capitalize()\n",
    "        templatesTarget[template] = templatesTarget.get(template,{'Tcount':0,'Tpages':[],'Params':{}})\n",
    "        templatesTarget[template]['Tcount'] += 1\n",
    "        templatesTarget[template]['Tpages'].append(page)\n",
    "        for name,val in param.items():\n",
    "            name = name.lower()\n",
    "            templatesTarget[template]['Params'][name] =templatesTarget[template]['Params'].get(name,0)\n",
    "            templatesTarget[template]['Params'][name] +=1\n",
    "            paramsTarget[name] = paramsTarget.get(name,0)\n",
    "            paramsTarget[name] += 1"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Getting most frequent templates and parameters"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [],
   "source": [
    "templateFreq = [(a[0],a[1]['Tcount'],len(a[1]['Params'])) for a in sorted(templates.items(), key =  lambda x:x[1]['Tcount'],reverse=True)]\n",
    "templateFreq = [(a[0],a[1]['Tcount'],len(a[1]['Params'])) for a in sorted(templates.items(), key =  lambda x:x[1]['Tcount'],reverse=True)]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[('Listaref', 618206, 17467),\n",
       " ('Cita web', 474422, 770),\n",
       " ('Ficha de entidad subnacional', 256644, 2003),\n",
       " ('Ficha de taxón', 166540, 464),\n",
       " ('Cita libro', 159579, 687),\n",
       " ('Ficha de deportista', 120360, 1181),\n",
       " ('Cita publicación', 89720, 925),\n",
       " ('Cita noticia', 81953, 260),\n",
       " ('Ficha de persona', 68685, 2285),\n",
       " ('Coord', 67812, 81),\n",
       " ('Referencias', 66815, 484),\n",
       " ('Commonscat', 66446, 4925),\n",
       " ('Enlace roto', 64063, 212),\n",
       " ('Sucesión', 47983, 144),\n",
       " ('Altura', 46017, 14),\n",
       " ('Wayback', 38548, 15),\n",
       " ('Cite web', 36285, 255),\n",
       " ('Evolución demográfica', 36199, 214),\n",
       " ('Peso', 33365, 16),\n",
       " ('Ficha de artista musical', 28863, 1256),\n",
       " ('Control de autoridades', 27633, 3083),\n",
       " ('Ficha de álbum', 26938, 495),\n",
       " ('Wikispecies', 26675, 678),\n",
       " ('Gráfica de evolución', 25393, 141),\n",
       " ('Ficha de película', 25351, 1334),\n",
       " ('Ficha de actor', 23706, 1134),\n",
       " ('Ficha de autoridad', 22795, 1458),\n",
       " ('Esp', 22403, 5693),\n",
       " ('!', 21445, 1614),\n",
       " ('Traducido ref', 20612, 34),\n",
       " ('Iucn', 19424, 137),\n",
       " ('Biology catalog', 17671, 3),\n",
       " ('Usa', 17412, 5967),\n",
       " ('Ficha de localidad de españa', 17070, 652),\n",
       " ('En', 14898, 3645),\n",
       " ('Ficha de cuerpo celeste\\n<!-- encabezado -->', 13853, 133),\n",
       " ('Cite news', 13230, 157),\n",
       " ('Bandera', 13164, 125),\n",
       " ('Ficha de sencillo', 13107, 357),\n",
       " ('Lista de canciones', 13085, 1399),\n",
       " ('Sfn', 12901, 53),\n",
       " ('Arg', 12232, 4482),\n",
       " ('Ficha de organización', 12132, 1512),\n",
       " ('Harvnp', 11939, 23),\n",
       " ('Cite book', 11718, 315),\n",
       " ('Cita requerida', 10659, 4607),\n",
       " ('Ficha de serie de televisión', 10595, 783),\n",
       " ('Ficha de estación', 10385, 293),\n",
       " ('Mex', 9583, 4126),\n",
       " ('Ficha de cuerpo de agua', 9545, 431)]"
      ]
     },
     "execution_count": 34,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "templateFreq[0:50]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 121,
   "metadata": {
    "scrolled": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[7028454, \"Oswaldo de Oliveira\", {\"JPN\": {\"\u001b[01;31m\u001b[K05.22.2012 acessodata\u001b[m\u001b[K\": \"\", \"url\": \"http://esporte.uol.com.br/futebol/ultimas-noticias/2012/05/22/campeao-mundial-em-2000-oswaldo-de-oliveira-revela-torcida-pelo-corinthians-na-libertadores.htm\", \"t\\u00edtulo\": \"Campe\\u00e3o mundo en 2000, Oswaldo de Oliveira revela animando para el Corinthians en la Libertadores\", \"fecha\": \"22 de mayo de 2012\", \"UOL autor\": \"\"}, \"cita web\": {\"\u001b[01;31m\u001b[K05.22.2012 acessodata\u001b[m\u001b[K\": \"\", \"url\": \"http://esporte.uol.com.br/futebol/ultimas-noticias/2012/05/22/campeao-mundial-em-2000-oswaldo-de-oliveira-revela-torcida-pelo-corinthians-na-libertadores.htm\", \"t\\u00edtulo\": \"Campe\\u00e3o mundo en 2000, Oswaldo de Oliveira revela animando para el Corinthians en la Libertadores\", \"fecha\": \"22 de mayo de 2012\", \"UOL autor\": \"\"}, \"BRA\": {\"\u001b[01;31m\u001b[K05.22.2012 acessodata\u001b[m\u001b[K\": \"\", \"url\": \"http://esporte.uol.com.br/futebol/ultimas-noticias/2012/05/22/campeao-mundial-em-2000-oswaldo-de-oliveira-revela-torcida-pelo-corinthians-na-libertadores.htm\", \"t\\u00edtulo\": \"Campe\\u00e3o mundo en 2000, Oswaldo de Oliveira revela animando para el Corinthians en la Libertadores\", \"fecha\": \"22 de mayo de 2012\", \"UOL autor\": \"\"}, \"Ficha de deportista\": {\"fecha nacimiento\": \"{{fecha|5|12|1950|edad}}\", \"debut internacional\": \"\", \"selecci\\u00f3n\": \"\", \"posici\\u00f3n\": \"[[Entrenador]]\", \"veces internacional\": \"\", \"imagen\": \"\", \"equipos\": \"\", \"lugar nacimiento\": \"[[R\\u00edo de Janeiro]], [[Brasil]]\", \"equipo_retiro\": \"\", \"tama\\u00f1o de imagen\": \"\", \"pie de imagen\": \"<!-- Secci\\u00f3n de datos personales -->\", \"retiro\": \"\", \"nac_deportiva\": \"\", \"altura\": \"1,85 [[metro]]s\\n<!-- Secci\\u00f3n de carrera -->\", \"n\\u00famero_selecci\\u00f3n\": \"\", \"nacionalidad\": \"[[Archivo:Flag of Brazil.svg|20px|brasile\\u00f1o]]\", \"nombrecompleto\": \"Oswaldo de Oliveira Filho\", \"liga\": \"[[J1 League]]\\n<!-- Secci\\u00f3n de trayectoria -->\", \"goles_clubes\": \"\", \"nombre\": \"Oswaldo de Oliveira\", \"deporte\": \"F\\u00fatbol\", \"inicio\": \"[[1999]]\", \"equipo_debut\": \"[[Sport Club Corinthians Paulista|Corinthians]]\", \"goles internacional\": \"\", \"torneos\": \"\", \"t\\u00edtulos\": \"\", \"club\": \"{{bandera|JPN}} [[Urawa Red Diamonds]]\"}, \"JAP\": {\"\u001b[01;31m\u001b[K05.22.2012 acessodata\u001b[m\u001b[K\": \"\", \"url\": \"http://esporte.uol.com.br/futebol/ultimas-noticias/2012/05/22/campeao-mundial-em-2000-oswaldo-de-oliveira-revela-torcida-pelo-corinthians-na-libertadores.htm\", \"t\\u00edtulo\": \"Campe\\u00e3o mundo en 2000, Oswaldo de Oliveira revela animando para el Corinthians en la Libertadores\", \"fecha\": \"22 de mayo de 2012\", \"UOL autor\": \"\"}, \"QAT\": {\"\u001b[01;31m\u001b[K05.22.2012 acessodata\u001b[m\u001b[K\": \"\", \"url\": \"http://esporte.uol.com.br/futebol/ultimas-noticias/2012/05/22/campeao-mundial-em-2000-oswaldo-de-oliveira-revela-torcida-pelo-corinthians-na-libertadores.htm\", \"t\\u00edtulo\": \"Campe\\u00e3o mundo en 2000, Oswaldo de Oliveira revela animando para el Corinthians en la Libertadores\", \"fecha\": \"22 de mayo de 2012\", \"UOL autor\": \"\"}, \"listaref\": {\"\u001b[01;31m\u001b[K05.22.2012 acessodata\u001b[m\u001b[K\": \"\", \"url\": \"http://esporte.uol.com.br/futebol/ultimas-noticias/2012/05/22/campeao-mundial-em-2000-oswaldo-de-oliveira-revela-torcida-pelo-corinthians-na-libertadores.htm\", \"t\\u00edtulo\": \"Campe\\u00e3o mundo en 2000, Oswaldo de Oliveira revela animando para el Corinthians en la Libertadores\", \"fecha\": \"22 de mayo de 2012\", \"UOL autor\": \"\"}}]\r\n"
     ]
    }
   ],
   "source": [
    "### looking for errors or unsual parameters\n",
    "!grep --color=always \"05.22.2012 acessodata\" templates-articles_es.json"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "6"
      ]
     },
     "execution_count": 22,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "params['05.22.2012 acessodata']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[('título', 1564249),\n",
       " ('url', 1380220),\n",
       " ('nombre', 1259267),\n",
       " ('fechaacceso', 989166),\n",
       " ('imagen', 800260),\n",
       " ('fecha', 562732),\n",
       " ('idioma', 559555),\n",
       " ('editorial', 525474),\n",
       " ('año', 338468),\n",
       " ('país', 328587),\n",
       " ('tipo', 301166),\n",
       " ('mapa_loc', 297770),\n",
       " ('autor', 287528),\n",
       " ('obra', 275163),\n",
       " ('tipo_superior_1', 274437),\n",
       " ('superior_1', 274310),\n",
       " ('familia', 250417),\n",
       " ('ubicación', 249046),\n",
       " ('tipo_superior_2', 244119),\n",
       " ('superior_2', 244032)]"
      ]
     },
     "execution_count": 35,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "sorted(params.items(), key =  lambda x:x[1],reverse=True)[0:20]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[('title', 12129321),\n",
       " ('date', 10136985),\n",
       " ('url', 9735142),\n",
       " ('accessdate', 6061055),\n",
       " ('publisher', 5732969),\n",
       " ('name', 3450763),\n",
       " ('website', 2726258),\n",
       " ('image', 2659807),\n",
       " ('last', 2644079),\n",
       " ('first', 2563686),\n",
       " ('year', 2097265),\n",
       " ('work', 1952569),\n",
       " ('caption', 1818576),\n",
       " ('display', 1703860),\n",
       " ('df', 1637443),\n",
       " ('language', 1625473),\n",
       " ('location', 1563582),\n",
       " ('isbn', 1290803),\n",
       " ('access-date', 1187202),\n",
       " ('pages', 1169205)]"
      ]
     },
     "execution_count": 24,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "sorted(paramsTarget.items(), key =  lambda x:x[1],reverse=True)[0:20]\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Experiments with some (arbitrary) selected templates"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "metadata": {},
   "outputs": [],
   "source": [
    "exampleTemplate = 'Ficha de persona'\n",
    "sourceLang ='es'\n",
    "targetLang = 'en'\n",
    "templatesTwoLang = getWikidataPair(['Plantilla:'+exampleTemplate],sourceLang,targetLang)[0]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "metadata": {},
   "outputs": [],
   "source": [
    "import random\n",
    "random.seed(1)\n",
    "pagesSource = random.sample(templates[exampleTemplate][\"Tpages\"],50)\n",
    "pagesPairs = getWikidataPair(pagesSource,sourceLang,targetLang)\n",
    "templatesPairsPerPage = {}\n",
    "for pages in pagesPairs:\n",
    "    templateSource = [(name,data) for name,data in  extract_templates(getContent(pages[0],sourceLang)).items() if name.lower() in templatesTwoLang[0].lower()] \n",
    "    templateTarget = [(name,data) for name,data in  extract_templates(getContent(pages[1],targetLang)).items() if name.lower() in templatesTwoLang[1].lower()] \n",
    "    if templateSource:\n",
    "        if templateTarget:\n",
    "            templatesPairsPerPage[pages[0]] = templatesPairsPerPage.get(pages[0],[])\n",
    "\n",
    "            templatesPairsPerPage[pages[0]].append((templateSource,templateTarget))\n",
    "    \n",
    "    "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## TemplateData (metadata) based aligments\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Using Josifoski et al alignments [https://arxiv.org/abs/1904.03922]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 192,
   "metadata": {},
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "from cr5 import Cr5_Model\n",
    "from scipy.spatial import distance\n",
    "from collections import Counter\n",
    "\n",
    "model = Cr5_Model('./','joint_28') # path_to_pretrained_model, model_prefix\n",
    "model.load_langs(['en', 'ca','es']) # list_of_languages\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 57,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "['Plantilla:Ficha de persona', 'Template:Infobox person']"
      ]
     },
     "execution_count": 57,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "templatesTwoLang"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 58,
   "metadata": {},
   "outputs": [],
   "source": [
    "\n",
    "templateDataLang1 = getTemplateData(templatesTwoLang[0],sourceLang)\n",
    "templateDataLang2 = getTemplateData(templatesTwoLang[1],targetLang)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 59,
   "metadata": {},
   "outputs": [],
   "source": [
    "distances = {}\n",
    "for param1,data1 in templateDataLang1['params'].items():\n",
    "    try:\n",
    "        vec1 = model.get_document_embedding(param1.split(),sourceLang)\n",
    "        distances[param1] = []\n",
    "        for param2,data2 in templateDataLang2['params'].items():\n",
    "            if data1['type'] == data2['type']:\n",
    "\n",
    "                try:\n",
    "                    vec2 = model.get_document_embedding(param2.split(),targetLang)\n",
    "                    dist = distance.cosine(vec1,vec2)\n",
    "                    distances[param1].append([dist,param2])\n",
    "                except:\n",
    "                    pass\n",
    "    except:\n",
    "        pass\n",
    "for param,words in distances.items():\n",
    "    try:\n",
    "        print(param,sorted(words)[0:3])\n",
    "    except:\n",
    "        pass"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 60,
   "metadata": {},
   "outputs": [],
   "source": [
    "distances2= {}\n",
    "for param1,data1 in templateDataLang1['params'].items():\n",
    "    try:\n",
    "        try:\n",
    "            vec1 = model.get_document_embedding(data1['label'][sourceLang].split(),sourceLang)\n",
    "        except:\n",
    "            vec1 = model.get_document_embedding(param1.split(),sourceLang)\n",
    "\n",
    "        distances2[param1] = []\n",
    "        for param2,data2 in templateDataLang2['params'].items():\n",
    "            if data1['type'] == data2['type']:\n",
    "                try:\n",
    "                    vec2 = model.get_document_embedding(data2['label'][targetLang].split(),targetLang)\n",
    "                except:\n",
    "                    vec2 = model.get_document_embedding(param2.split(),targetLang)\n",
    "                try:\n",
    "                    dist = distance.cosine(vec1,vec2)\n",
    "                    distances2[param1].append([dist,param2])\n",
    "                except:\n",
    "                    pass\n",
    "    except:\n",
    "        pass\n",
    "for param,words in distances2.items():\n",
    "    try:\n",
    "        print(param,sorted(words)[0:2])\n",
    "    except:\n",
    "        pass"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Using Wikidata based alignemnts https://github.com/digitalTranshumant/wmf-interlanguage"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {},
   "outputs": [],
   "source": [
    "## Trying my alignments\n",
    "\n",
    "import fastText\n",
    "myModel = fastText.load_model('fastText/wiki.es.bin')\n",
    "myModel2 = fastText.load_model('fastText/wiki.en.bin')\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 61,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "educación [[0.21443717063835965, 'education'], [0.6178690504391597, 'citizenship'], [0.6280876465303372, 'children']]\n",
      "ocupación [[0.2631526201422385, 'occupation'], [0.6787827437526144, 'nationality'], [0.67897376154512, 'years_active']]\n",
      "cónyuge [[0.24983593109192836, 'spouse'], [0.40392868186760367, 'birth_name'], [0.5124027167893581, 'other_names']]\n",
      "influenciado por [[0.33104337127144423, 'influenced'], [0.4706785758481643, 'influences'], [0.6523160108643545, 'landscape']]\n",
      "movimiento [[0.250243910083511, 'movement'], [0.5778743841110583, 'party'], [0.6147539468328521, 'organization']]\n",
      "nombre [[0.26619029141547823, 'name'], [0.6145292268664284, 'native_name'], [0.6246814849355296, 'callsign']]\n",
      "influyó [[0.36215121638317693, 'influenced'], [0.45778910238385484, 'influences'], [0.6678311460816266, 'landscape']]\n",
      "imagen [[0.35426880401306926, 'image'], [0.4961995987604414, 'caption'], [0.5666827976810916, 'image_upright']]\n",
      "ciudadanía [[0.3015432342644394, 'citizenship'], [0.653580951322187, 'nationality'], [0.6600960568858325, 'education']]\n",
      "premios [[0.1996928100588924, 'awards'], [0.6319804378746378, 'honours'], [0.746109625629121, 'honorific_prefix']]\n",
      "peso [[0.36017355935521966, 'weight'], [0.49174267895300294, 'weight_kg'], [0.549960720560015, 'weight_lb']]\n",
      "residencia [[0.22379335319260452, 'residence'], [0.5994219355354917, 'resting_place_coordinates'], [0.6178789690657341, 'resting_place']]\n",
      "sitio web [[0.3162818377780624, 'website'], [0.7412872971725771, 'organization'], [0.7445792200812413, 'residence']]\n",
      "nombre completo [[0.3834951969338314, 'name'], [0.6662905921800288, 'native_name'], [0.6776359587662848, 'pronunciation']]\n",
      "partido político [[0.2710686742438828, 'party'], [0.6210887194532322, 'movement'], [0.628641284713549, 'native_name']]\n",
      "salario [[0.34686267984318064, 'salary'], [0.5329042061310687, 'employer'], [0.5965979002841617, 'net_worth']]\n",
      "altura [[0.38311822276064533, 'height'], [0.5832588073048197, 'height_m'], [0.6220852773203933, 'height_ft']]\n",
      "empleador [[0.3348148659869248, 'employer'], [0.5446088862241445, 'partner'], [0.5862241792475072, 'salary']]\n",
      "padres [[0.2876338169892533, 'parents'], [0.4058660453093854, 'mother'], [0.4181208590909826, 'father']]\n",
      "sucesor [[0.3365330269688752, 'successor'], [0.456283892848407, 'predecessor'], [0.618620802171814, 'father']]\n"
     ]
    }
   ],
   "source": [
    "distances3 = {}\n",
    "transmat = np.loadtxt('fastText_multilingual/my_alingments/apply_in_en_to_es.txt')\n",
    "for param1,data1 in templateDataLang1['params'].items():\n",
    "        vec1 = myModel.get_sentence_vector(param1)\n",
    "        distances3[param1] = []\n",
    "        for param2,data2 in templateDataLang2['params'].items():\n",
    "          #  if data1['type'] == data2['type']:\n",
    "                vec2= myModel2.get_sentence_vector(param2)\n",
    "                vec2T = apply_transform(vec2,transmat)\n",
    "                dist = distance.cosine(vec1,vec2T)\n",
    "                distances3[param1].append([dist,param2])\n",
    "\n",
    "for param,words in distances3.items():\n",
    "    try:\n",
    "        if sorted(words)[0][0] < .4:\n",
    "            print(param,sorted(words)[0:3])\n",
    "    except:\n",
    "        pass"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Template data limitations\n",
    "* Big amount of 'es' templates, have no templatedata info"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Listaref ok\n",
      "Cita web ok\n",
      "Ficha de entidad subnacional ok\n",
      "Ficha de taxón no\n",
      "Cita libro ok\n",
      "Ficha de deportista no\n",
      "Cita publicación ok\n",
      "Cita noticia ok\n",
      "Ficha de persona ok\n",
      "Coord no\n",
      "Referencias ok\n",
      "Commonscat ok\n",
      "Enlace roto ok\n",
      "Sucesión ok\n",
      "Altura no\n",
      "Wayback no\n",
      "Cite web ok\n",
      "Evolución demográfica no\n",
      "Peso no\n",
      "Ficha de artista musical ok\n",
      "Control de autoridades no\n",
      "Ficha de álbum ok\n",
      "Wikispecies ok\n",
      "Gráfica de evolución no\n",
      "Ficha de película ok\n",
      "Ficha de actor ok\n",
      "Ficha de autoridad ok\n",
      "Esp no\n",
      "! no\n",
      "Traducido ref no\n",
      "Iucn no\n",
      "Biology catalog no\n",
      "Usa no\n",
      "Ficha de localidad de españa no\n",
      "En no\n",
      "Ficha de cuerpo celeste\n",
      "<!-- encabezado --> no\n",
      "Cite news ok\n",
      "Bandera ok\n",
      "Ficha de sencillo ok\n",
      "Lista de canciones no\n",
      "Sfn ok\n",
      "Arg no\n",
      "Ficha de organización no\n",
      "Harvnp ok\n",
      "Cite book ok\n",
      "Cita requerida ok\n",
      "Ficha de serie de televisión ok\n",
      "Ficha de estación no\n",
      "Mex no\n",
      "Ficha de cuerpo de agua ok\n",
      "27 ok 23 No\n"
     ]
    }
   ],
   "source": [
    "withTemplateData = 0\n",
    "withoutTemplateData = 0\n",
    "for t in templateFreq[0:50]:\n",
    "    try:\n",
    "        tmp = getTemplateData('Plantilla:'+t[0],lang='es')\n",
    "        print(t[0],'ok')\n",
    "        withTemplateData +=1\n",
    "    except:\n",
    "        print(t[0],'no')\n",
    "        withoutTemplateData +=1\n",
    "print(withTemplateData,'ok',withoutTemplateData,'No')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Without Template Data\n",
    "This is working pretty good. \n",
    "We need to:\n",
    "* Play with the similiarty threshold\n",
    "* Map 1 to 1 (keep just the most similar)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 62,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "['Plantilla:Ficha de persona', 'Template:Infobox person']\n",
      "almamáter [[0.20313075649282697, 'alma_mater'], [0.40838973184531924, 'university'], [0.4787241033567037, 'image_name']]\n",
      "ocupación [[0.2631526201422385, 'occupation'], [0.6600671853620486, 'period'], [0.6787827437526144, 'nationality']]\n",
      "hijos [[0.32078906070348034, 'siblings'], [0.4237291785050009, 'parents'], [0.44422598496494703, 'mother']]\n",
      "salario [[0.34686267984318064, 'salary'], [0.5329042061310687, 'employer'], [0.5914404969325433, 'networth']]\n",
      "religión [[0.1918389519777819, 'religion'], [0.5063039178227149, 'ethnicity'], [0.5418350265150662, 'ethnic']]\n",
      "página web [[0.28040716269534793, 'website'], [0.39915367366575616, 'homepage'], [0.5409951249382179, 'url']]\n",
      "ciudadanía [[0.3015432342644394, 'citizenship'], [0.5893396413425851, 'ethnicity'], [0.6132553383436963, 'ethnic']]\n",
      "obras [[0.3723628392673217, 'works'], [0.5072776723566816, 'notable works'], [0.6917973714188009, 'influences']]\n",
      "partido político [[0.2710686742438828, 'party'], [0.5555918771322552, 'elected'], [0.6210887194532322, 'movement']]\n",
      "altura [[0.38311822276064533, 'height'], [0.6268790213404272, 'size'], [0.6356246068183318, 'height_cm']]\n",
      "profesión [[0.2907516493035255, 'profession'], [0.6062218171219675, 'occupation'], [0.6481471582822792, 'father']]\n",
      "tamaño [[0.37431834555687193, 'size'], [0.41338675119865964, 'image size'], [0.6053846158681382, 'height']]\n",
      "movimiento [[0.250243910083511, 'movement'], [0.5778743841110583, 'party'], [0.6147539468328521, 'organization']]\n",
      "empleador [[0.3348148659869248, 'employer'], [0.5446088862241445, 'partner'], [0.5862241792475072, 'salary']]\n",
      "padres [[0.2876338169892533, 'parents'], [0.3318735405300991, 'siblings'], [0.4058660453093854, 'mother']]\n",
      "web [[0.28798770556143627, 'website'], [0.4095679844204547, 'homepage'], [0.5692907818495073, 'url']]\n",
      "partido [[0.2765007207344614, 'party'], [0.6017298525914572, 'elected'], [0.6286008653139599, 'opponents']]\n",
      "sucesor [[0.3365330269688752, 'successor'], [0.456283892848407, 'predecessor'], [0.596444413828311, 'birthname']]\n",
      "educación [[0.21443717063835965, 'education'], [0.5491864642800925, 'university'], [0.5631942038119764, 'training']]\n",
      "sucesor2 [[0.38024225945413925, 'successor'], [0.5209480783535644, 'predecessor'], [0.6205342973430593, 'spouse']]\n",
      "cónyuge [[0.24983593109192836, 'spouse'], [0.40392868186760367, 'birth_name'], [0.4092985624124069, 'domesticpartner']]\n",
      "conocido [[0.3359444275967298, 'known'], [0.4397517801871975, 'known for'], [0.5681550678209043, 'nickname']]\n",
      "tamaño de imagen [[0.3487460935340897, 'image size'], [0.44899283918994626, 'image caption'], [0.46018400256375347, 'image']]\n",
      "alma máter [[0.3910620676540274, 'alma mater'], [0.47251254267794773, 'university'], [0.4729891746963797, 'alma_mater']]\n",
      "sucesor3 [[0.3923570462103766, 'successor'], [0.5411234273905198, 'predecessor'], [0.6156393623717037, 'spouse']]\n",
      "nombre de nacimiento [[0.36015477905009385, 'birth name'], [0.428332985713629, 'other name'], [0.44078376047482226, 'name']]\n",
      "otros premios [[0.3566963188347183, 'awards'], [0.4269460563112747, 'prizes'], [0.49343071972289976, 'baftaawards']]\n",
      "nombre [[0.26619029141547823, 'name'], [0.38294882716225065, 'birth name'], [0.3887340448708989, 'other name']]\n",
      "imagen [[0.35426880401306926, 'image'], [0.3695230158584034, 'image caption'], [0.44648244574729823, 'image size']]\n",
      "premios [[0.1996928100588924, 'awards'], [0.36031485835432775, 'prizes'], [0.36977659925063355, 'baftaawards']]\n",
      "etnia [[0.34014606600360775, 'ethnicity'], [0.34867307048601637, 'ethnic'], [0.5098585953152114, 'language']]\n",
      "peso [[0.36017355935521966, 'weight'], [0.6609215027394688, 'height_cm'], [0.6805149870931443, 'size']]\n",
      "sitio web [[0.3162818377780624, 'website'], [0.4391092135402731, 'homepage'], [0.6131980278300186, 'url']]\n",
      "nombre completo [[0.3834951969338314, 'name'], [0.4488191480151389, 'birth name'], [0.4540852631947522, 'other name']]\n",
      "causa muerte [[0.3906710959670078, 'death'], [0.5230927214892079, 'death_cause'], [0.5804853362146993, 'criminal_penalty']]\n",
      "apodo [[0.2508204685816622, 'nickname'], [0.5267847117341005, 'name'], [0.573542126756905, 'birth name']]\n",
      "años activo [[0.2560523897523098, 'years active'], [0.5423190894009016, 'period'], [0.5719775227860848, 'age']]\n",
      "posgrado [[0.3396015193960188, 'university'], [0.4851736171837069, 'education'], [0.5389821894688279, 'alma_mater']]\n",
      "residencia [[0.22379335319260452, 'residence'], [0.5764792058191653, 'deathplace'], [0.5838530613268462, 'home town']]\n",
      "otros nombres [[0.32951607687855056, 'other names'], [0.3858069244413459, 'other name'], [0.581711504984381, 'name']]\n",
      "60 Without Mapping\n"
     ]
    }
   ],
   "source": [
    "distancesNoTemplateData = {}\n",
    "print(templatesTwoLang)\n",
    "template1 = templates[templatesTwoLang[0].split(':')[1]]\n",
    "template1['params'] = sorted(template1['Params'].items(), key= lambda x: x[1],reverse=True)[:100]\n",
    "template2 = templatesTarget[templatesTwoLang[1].split(':')[1]]\n",
    "template2['params'] = sorted(template2['Params'].items(), key= lambda x: x[1],reverse=True)[:200]\n",
    "\n",
    "\n",
    "transmat = np.loadtxt('fastText_multilingual/my_alingments/apply_in_en_to_es.txt')\n",
    "for param1,data1 in template1['params']:\n",
    "    try:\n",
    "        vec1 = myModel.get_sentence_vector(param1.strip())\n",
    "        distancesNoTemplateData[param1] = []\n",
    "    except: pass\n",
    "    for param2,data2 in template2['params']:\n",
    "            try:\n",
    "                vec2= myModel2.get_sentence_vector(param2.strip())\n",
    "                vec2T = apply_transform(vec2,transmat)\n",
    "                dist = distance.cosine(vec1,vec2T)\n",
    "                distancesNoTemplateData[param1].append([dist,param2])\n",
    "            except:\n",
    "                pass\n",
    "\n",
    "cNotFound = 0\n",
    "for param,words in distancesNoTemplateData.items():\n",
    "    try:\n",
    "        if sorted(words)[0][0] < .4:\n",
    "            print(param,sorted(words)[0:3])\n",
    "        else:\n",
    "            cNotFound +=1\n",
    "    except:\n",
    "        pass\n",
    "print(cNotFound,'Without Mapping')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[('fecha de nacimiento', 48008),\n",
       " ('lugar de nacimiento', 46031),\n",
       " ('nombre', 42222),\n",
       " ('imagen', 39398),\n",
       " ('nacionalidad', 38977),\n",
       " ('fecha de fallecimiento', 36024),\n",
       " ('lugar de fallecimiento', 33473),\n",
       " ('ocupación', 33152),\n",
       " ('nombre de nacimiento', 23697),\n",
       " ('cónyuge', 21299),\n",
       " ('hijos', 17867),\n",
       " ('pie de imagen', 15154),\n",
       " ('padres', 15017),\n",
       " ('tamaño', 14278),\n",
       " ('alma máter', 14107),\n",
       " ('firma', 13678),\n",
       " ('descripción', 13666),\n",
       " ('residencia', 13649),\n",
       " ('área', 12665),\n",
       " ('premios', 12549),\n",
       " ('tamaño de imagen', 11100),\n",
       " ('años activo', 9415),\n",
       " ('otros nombres', 9261),\n",
       " ('educación', 9154),\n",
       " ('predecesor', 9151),\n",
       " ('empleador', 9050),\n",
       " ('sucesor', 8669),\n",
       " ('religión', 7618),\n",
       " ('causa muerte', 7280),\n",
       " ('abreviatura en botánica', 6965),\n",
       " ('familiares', 6751),\n",
       " ('partido político', 6671),\n",
       " ('conocido por', 6604),\n",
       " ('cargo', 6534),\n",
       " ('inicio', 6454),\n",
       " ('twitter', 6400),\n",
       " ('notas', 6377),\n",
       " ('conocido', 5981),\n",
       " ('pareja', 5938),\n",
       " ('patrimonio', 5718),\n",
       " ('final', 5656),\n",
       " ('posgrado', 5649),\n",
       " ('facebook', 5539),\n",
       " ('nombre nativo', 5367),\n",
       " ('etnia', 5345),\n",
       " ('escudo', 5184),\n",
       " ('altura', 5168),\n",
       " ('apodo', 4864),\n",
       " ('web', 4856),\n",
       " ('peso', 4721),\n",
       " ('obras', 4670),\n",
       " ('nombre completo', 4550),\n",
       " ('salario', 4239),\n",
       " ('título', 4095),\n",
       " ('página web', 3982),\n",
       " ('obras destacadas', 3939),\n",
       " ('ciudadanía', 3809),\n",
       " ('id', 3506),\n",
       " ('inicio2', 3488),\n",
       " ('término', 3448),\n",
       " ('final2', 3315),\n",
       " ('empresa', 3228),\n",
       " ('t', 3225),\n",
       " ('cargo2', 3208),\n",
       " ('programas', 3053),\n",
       " ('medio de comunicación', 3036),\n",
       " ('wikidata', 3033),\n",
       " ('escudo2', 2416),\n",
       " ('sitio web', 2396),\n",
       " ('cabecera', 2335),\n",
       " ('partido', 2239),\n",
       " ('predecesor2', 2130),\n",
       " ('movimiento', 2123),\n",
       " ('sucesor2', 2097),\n",
       " ('fuentebiográfica', 2018),\n",
       " ('tiempo', 2016),\n",
       " ('inicio3', 1859),\n",
       " ('final3', 1818),\n",
       " ('cargo3', 1691),\n",
       " ('ciudadania', 1606),\n",
       " ('presidente', 1605),\n",
       " ('imdb', 1553),\n",
       " ('profesión', 1515),\n",
       " ('año debut', 1458),\n",
       " ('estatura', 1453),\n",
       " ('titulo', 1433),\n",
       " ('pieimagen', 1365),\n",
       " ('género', 1360),\n",
       " ('termino', 1358),\n",
       " ('distrito', 1350),\n",
       " ('almamáter', 1273),\n",
       " ('escudo3', 1262),\n",
       " ('año retiro', 1217),\n",
       " ('sucesor3', 1111),\n",
       " ('predecesor3', 1100),\n",
       " ('otros premios', 1083),\n",
       " ('inicio4', 1007),\n",
       " ('final4', 971),\n",
       " ('pie de foto', 951),\n",
       " ('cargo4', 910)]"
      ]
     },
     "execution_count": 41,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "template1['params']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "IOPub data rate exceeded.\n",
      "The notebook server will temporarily stop sending output\n",
      "to the client in order to avoid crashing it.\n",
      "To change this limit, set the config variable\n",
      "`--NotebookApp.iopub_data_rate_limit`.\n",
      "\n",
      "Current values:\n",
      "NotebookApp.iopub_data_rate_limit=1000000.0 (bytes/sec)\n",
      "NotebookApp.rate_limit_window=3.0 (secs)\n",
      "\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0 Without Mapping\n"
     ]
    }
   ],
   "source": [
    "#Same before, just trying with other Template\n",
    "exampleTemplate2 = 'Ficha de deportista'\n",
    "sourceLang2 ='es'\n",
    "targetLang2 = 'en'\n",
    "templatesTwoLang2 = getWikidataPair(['Plantilla:'+exampleTemplate2],sourceLang2,targetLang2)[0]\n",
    "\n",
    "\n",
    "distancesNoTemplateData = {}\n",
    "print(templatesTwoLang2)\n",
    "template1 = templates[templatesTwoLang2[0].split(':')[1]]\n",
    "template1['params'] = sorted(template1['Params'].items(), key= lambda x: x[1],reverse=True)[:100]\n",
    "template2 = templatesTarget[templatesTwoLang2[1].split(':')[1]]\n",
    "template2['params'] = sorted(template2['Params'].items(), key= lambda x: x[1],reverse=True)[:200]\n",
    "\n",
    "transmat = np.loadtxt('fastText_multilingual/my_alingments/apply_in_en_to_es.txt')\n",
    "for param1,data1 in template1['params']:\n",
    "    try:\n",
    "        vec1 = myModel.get_sentence_vector(param1.strip())\n",
    "        distancesNoTemplateData[param1] = []\n",
    "    except: pass\n",
    "    for param2,data2 in template2['params']:\n",
    "            try:\n",
    "                vec2= myModel2.get_sentence_vector(param2.strip())\n",
    "                vec2T = apply_transform(vec2,transmat)\n",
    "                dist = distance.cosine(vec1,vec2T)\n",
    "                distancesNoTemplateData[param1].append([dist,param2])\n",
    "            except:\n",
    "                pass\n",
    "\n",
    "cNotFound = 0\n",
    "for param,words in distancesNoTemplateData.items():\n",
    "    try:\n",
    "        if sorted(words)[0][0] < .4:\n",
    "            print(param,sorted(words)[0:3])\n",
    "        else:\n",
    "            cNotFound +=1\n",
    "    except:\n",
    "        pass\n",
    "print(cNotFound,'Without Mapping')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## With Graph approach\n",
    "Notes:\n",
    "* Play with amount of candidates per language. By now, what works the best is to have few elements in the source lang and a lot in the target"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 51,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "['Plantilla:Ficha de película', 'Template:Infobox film']\n"
     ]
    }
   ],
   "source": [
    "import networkx as nx\n",
    "G= nx.Graph()\n",
    "exampleTemplate2 = 'Ficha de película'\n",
    "sourceLang2 ='es'\n",
    "targetLang2 = 'en'\n",
    "templatesTwoLang2 = getWikidataPair(['Plantilla:'+exampleTemplate2],sourceLang2,targetLang2)[0]\n",
    "\n",
    "\n",
    "distancesNoTemplateData = {}\n",
    "print(templatesTwoLang2)\n",
    "template1 = templates[templatesTwoLang2[0].split(':')[1]]\n",
    "#template1['params'] = sorted(template1['Params'].items(), key= lambda x: x[1],reverse=True)[:400]\n",
    "template1['params'] = [x for x in template1['Params'].items() if x[1]/len(template1['Params']) > .05] #appears at least X% of times\n",
    "template2 = templatesTarget[templatesTwoLang2[1].split(':')[1]]\n",
    "#template2['params'] = sorted(template2['Params'].items(), key= lambda x: x[1],reverse=True)[:400]\n",
    "template2['params'] = [x for x in template2['Params'].items() if x[1]/len(template2['Params']) > .05] #appears at least 1X of times\n",
    "\n",
    "\n",
    "transmat = np.loadtxt('fastText_multilingual/my_alingments/apply_in_en_to_es.txt')\n",
    "for param1,data1 in template1['params']:\n",
    "    try:\n",
    "        vec1 = myModel.get_sentence_vector(param1.strip().replace('_',' '))\n",
    "        distancesNoTemplateData[param1] = []\n",
    "    except: pass\n",
    "    for param2,data2 in template2['params']:\n",
    "            try:\n",
    "                vec2= myModel2.get_sentence_vector(param2.strip().replace('_',' '))\n",
    "                vec2T = apply_transform(vec2,transmat)\n",
    "                dist = distance.cosine(vec1,vec2T)\n",
    "                distancesNoTemplateData[param1].append([dist,param2])\n",
    "                if dist < .45:\n",
    "                    node1= '%s_%s' % (sourceLang2,param1)\n",
    "                    node2= '%s_%s' % (targetLang2,param2)\n",
    "\n",
    "                    G.add_edge(node1,node2)\n",
    "                    G[node1][node2]['w'] = dist\n",
    "            except:\n",
    "                pass\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 52,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/srv/home/dsaez/3.6/lib/python3.5/site-packages/networkx/drawing/nx_pylab.py:126: MatplotlibDeprecationWarning: pyplot.hold is deprecated.\n",
      "    Future behavior will be consistent with the long-time default:\n",
      "    plot commands add elements without first clearing the\n",
      "    Axes and/or Figure.\n",
      "  b = plt.ishold()\n",
      "/srv/home/dsaez/3.6/lib/python3.5/site-packages/networkx/drawing/nx_pylab.py:522: MatplotlibDeprecationWarning: The is_string_like function was deprecated in version 2.1.\n",
      "  if not cb.is_string_like(edge_color) \\\n",
      "/srv/home/dsaez/3.6/lib/python3.5/site-packages/networkx/drawing/nx_pylab.py:543: MatplotlibDeprecationWarning: The is_string_like function was deprecated in version 2.1.\n",
      "  if cb.is_string_like(edge_color) or len(edge_color) == 1:\n",
      "/srv/home/dsaez/3.6/lib/python3.5/site-packages/networkx/drawing/nx_pylab.py:724: MatplotlibDeprecationWarning: The is_string_like function was deprecated in version 2.1.\n",
      "  if not cb.is_string_like(label):\n",
      "/srv/home/dsaez/3.6/lib/python3.5/site-packages/networkx/drawing/nx_pylab.py:138: MatplotlibDeprecationWarning: pyplot.hold is deprecated.\n",
      "    Future behavior will be consistent with the long-time default:\n",
      "    plot commands add elements without first clearing the\n",
      "    Axes and/or Figure.\n",
      "  plt.hold(b)\n",
      "/srv/home/dsaez/3.6/lib/python3.5/site-packages/matplotlib/__init__.py:805: MatplotlibDeprecationWarning: axes.hold is deprecated. Please remove it from your matplotlibrc and/or style files.\n",
      "  mplDeprecation)\n",
      "/srv/home/dsaez/3.6/lib/python3.5/site-packages/matplotlib/rcsetup.py:155: MatplotlibDeprecationWarning: axes.hold is deprecated, will be removed in 3.0\n",
      "  mplDeprecation)\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAdgAAAFCCAYAAABSJMy8AAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nSUVORK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f4f59effeb8>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "%matplotlib inline\n",
    "\n",
    "nx.draw(G,with_labels=True)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 53,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "producer -> productor\n",
      "music -> música\n",
      "director -> director\n",
      "production -> producción\n",
      "awards -> premios\n",
      "screenplay -> guion\n",
      "language -> idioma\n",
      "country -> país\n",
      "preceded by -> precedida_por\n",
      "preceded_by -> sucedida_por\n",
      "narrator -> narrador\n",
      "name -> nombre\n",
      "website -> página web\n",
      "distributor -> distribuidora\n",
      "budget -> presupuesto\n",
      "sound -> sonido\n",
      "image size -> tamaño imagen\n",
      "executive producer -> producción ejecutiva\n",
      "production company -> diseño producción\n",
      "production companies -> diseño de producción\n",
      "image -> imagen\n",
      "photographer -> fotografía\n",
      "imdb_rating -> imdb\n",
      "producers -> productora\n",
      "animator -> animador\n",
      "distributors -> distribuidor\n",
      "cinematography -> cinematografía\n",
      "based on -> basado en\n",
      "amg_id -> amg_id\n",
      "associate director -> ayudante dirección\n",
      "image_size -> tamaño\n",
      "based_on -> basada en\n",
      "script -> guión\n",
      "image  caption -> nombre imagen\n",
      "0.3695652173913043\n"
     ]
    }
   ],
   "source": [
    "c = 0\n",
    "while G.edges():\n",
    "    p = sorted(G.edges(data=True), key=lambda x: x[2]['w'])[0]\n",
    "    psorted = sorted(list(p[:2]))\n",
    "    print(psorted[0][3:],'->',psorted[1][3:])\n",
    "    G.remove_node(p[0])\n",
    "    G.remove_node(p[1])\n",
    "    c+=1\n",
    "print(c/len(template1['params']))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Download top 100 templates in 'es' to 'en"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 63,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "['Plantilla:Listaref', 'Template:Reflist']\n",
      "Listaref ======================\n",
      "['Plantilla:Cita web', 'Template:Cite web']\n",
      "Cita web ======================\n",
      "['Plantilla:Ficha de taxón', 'Template:Taxobox']\n",
      "Ficha de taxón ======================\n",
      "['Plantilla:Cita libro', 'Template:Cite book']\n",
      "Cita libro ======================\n",
      "['Plantilla:Ficha de deportista', 'Template:Infobox sportsperson']\n",
      "Ficha de deportista ======================\n",
      "['Plantilla:Cita publicación', 'Template:Cite journal']\n",
      "Cita publicación ======================\n",
      "['Plantilla:Cita noticia', 'Template:Cite news']\n",
      "Cita noticia ======================\n",
      "['Plantilla:Ficha de persona', 'Template:Infobox person']\n",
      "Ficha de persona ======================\n",
      "['Plantilla:Coord', 'Template:Coord']\n",
      "Coord ======================\n",
      "['Plantilla:Referencias', 'Template:Unreferenced']\n",
      "Referencias ======================\n",
      "['Plantilla:Commonscat', 'Template:Commons category']\n",
      "Commonscat ======================\n",
      "['Plantilla:Enlace roto', 'Template:Dead link']\n",
      "Enlace roto ======================\n",
      "['Plantilla:Sucesión', 'Template:State detail page browse']\n",
      "Sucesión ======================\n",
      "['Plantilla:Altura', 'Template:Height']\n",
      "Altura ======================\n",
      "['Plantilla:Evolución demográfica', 'Template:Demography']\n",
      "Evolución demográfica ======================\n",
      "['Plantilla:Ficha de artista musical', 'Template:Infobox musical artist']\n",
      "Ficha de artista musical ======================\n",
      "['Plantilla:Control de autoridades', 'Template:Authority control']\n",
      "Control de autoridades ======================\n",
      "['Plantilla:Ficha de álbum', 'Template:Infobox album']\n",
      "Ficha de álbum ======================\n",
      "['Plantilla:Wikispecies', 'Template:Wikispecies']\n",
      "Wikispecies ======================\n",
      "['Plantilla:Gráfica de evolución', 'Template:Vertical bar chart']\n",
      "Gráfica de evolución ======================\n",
      "['Plantilla:Ficha de película', 'Template:Infobox film']\n",
      "Ficha de película ======================\n",
      "['Plantilla:Ficha de autoridad', 'Template:Infobox officeholder']\n",
      "Ficha de autoridad ======================\n",
      "['Plantilla:!', 'Template:!']\n",
      "! ======================\n",
      "['Plantilla:En', 'Template:En icon']\n",
      "En ======================\n",
      "['Plantilla:Bandera', 'Template:Flagdeco']\n",
      "Bandera ======================\n",
      "['Plantilla:Lista de canciones', 'Template:Track listing']\n",
      "Lista de canciones ======================\n",
      "['Plantilla:Ficha de organización', 'Template:Infobox organization']\n",
      "Ficha de organización ======================\n",
      "['Plantilla:Harvnp', 'Template:Sfn']\n",
      "Harvnp ======================\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/srv/home/dsaez/3.6/lib/python3.5/site-packages/scipy/spatial/distance.py:505: RuntimeWarning: invalid value encountered in double_scalars\n",
      "  dist = 1.0 - np.dot(u, v) / (norm(u) * norm(v))\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "['Plantilla:Cita requerida', 'Template:Citation needed']\n",
      "Cita requerida ======================\n",
      "['Plantilla:Ficha de estación', 'Template:Infobox station']\n",
      "Ficha de estación ======================\n",
      "['Plantilla:Ficha de cuerpo de agua', 'Template:Infobox body of water']\n",
      "Ficha de cuerpo de agua ======================\n"
     ]
    }
   ],
   "source": [
    "import networkx as nx\n",
    "G= nx.Graph()\n",
    "output = {}\n",
    "transmat = np.loadtxt('fastText_multilingual/my_alingments/apply_in_en_to_es.txt')\n",
    "\n",
    "for exampleTemplate2 in  templateFreq[0:50]:\n",
    "    try:\n",
    "        exampleTemplate2 = exampleTemplate2[0]\n",
    "        sourceLang2 ='es'\n",
    "        targetLang2 = 'en'\n",
    "        templatesTwoLang2 = getWikidataPair(['Plantilla:'+exampleTemplate2],sourceLang2,targetLang2)[0]\n",
    "\n",
    "\n",
    "        distancesNoTemplateData = {}\n",
    "        print(templatesTwoLang2)\n",
    "        template1 = templates[templatesTwoLang2[0].split(':')[1]]\n",
    "        #template1['params'] = sorted(template1['Params'].items(), key= lambda x: x[1],reverse=True)[:400]\n",
    "        template1['params'] = [x for x in template1['Params'].items() if x[1]/len(template1['Params']) > .3] #appears at least X% of times\n",
    "        template2 = templatesTarget[templatesTwoLang2[1].split(':')[1]]\n",
    "        #template2['params'] = sorted(template2['Params'].items(), key= lambda x: x[1],reverse=True)[:400]\n",
    "        template2['params'] = [x for x in template2['Params'].items() if x[1]/len(template2['Params']) > .05] #appears at least 1X of times\n",
    "\n",
    "\n",
    "        for param1,data1 in template1['params']:\n",
    "            try:\n",
    "                vec1 = myModel.get_sentence_vector(param1.strip().replace('_',' '))\n",
    "                distancesNoTemplateData[param1] = []\n",
    "            except: pass\n",
    "            for param2,data2 in template2['params']:\n",
    "                    try:\n",
    "                        vec2= myModel2.get_sentence_vector(param2.strip().replace('_',' '))\n",
    "                        vec2T = apply_transform(vec2,transmat)\n",
    "                        dist = distance.cosine(vec1,vec2T)\n",
    "                        distancesNoTemplateData[param1].append([dist,param2])\n",
    "                        if dist < .45:\n",
    "                            node1= '%s_%s' % (sourceLang2,param1)\n",
    "                            node2= '%s_%s' % (targetLang2,param2)\n",
    "                            G.add_edge(node1,node2)\n",
    "                            G[node1][node2]['w'] = dist\n",
    "                    except:\n",
    "                        pass    \n",
    "        output[exampleTemplate2] = []\n",
    "        print(exampleTemplate2,'======================')\n",
    "        while G.edges():\n",
    "            p = sorted(G.edges(data=True), key=lambda x: x[2]['w'])[0]\n",
    "            psorted = sorted(list(p[:2]))       \n",
    "            \n",
    "            output[exampleTemplate2].append({psorted[0][:2]:psorted[0][3:],\n",
    "                                            psorted[1][:2]:psorted[1][3:],\n",
    "                                            'd':p[2]['w']})\n",
    "            G.remove_node(p[0])\n",
    "            G.remove_node(p[1])\n",
    "\n",
    "    except:\n",
    "        pass"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 64,
   "metadata": {},
   "outputs": [],
   "source": [
    "with open('templatesEsEN-top100.json','w') as o:\n",
    "    json.dump(output,o)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Value based - In progress"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 551,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "['Plantilla:Ficha de persona', 'Template:Infobox person']\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/srv/home/dsaez/3.6/lib/python3.5/site-packages/scipy/spatial/distance.py:505: RuntimeWarning: invalid value encountered in double_scalars\n",
      "  dist = 1.0 - np.dot(u, v) / (norm(u) * norm(v))\n"
     ]
    }
   ],
   "source": [
    "import networkx as nx\n",
    "G= nx.Graph()\n",
    "exampleTemplate2 = 'Ficha de persona'\n",
    "sourceLang2 ='es'\n",
    "targetLang2 = 'en'\n",
    "templatesTwoLang2 = getWikidataPair(['Plantilla:'+exampleTemplate2],sourceLang2,targetLang2)[0]\n",
    "\n",
    "import random\n",
    "random.seed(1)\n",
    "pagesSource = random.sample(templates[exampleTemplate2][\"Tpages\"],50)\n",
    "pagesPairs = getWikidataPair(pagesSource,sourceLang,targetLang)\n",
    "templatesPairsPerPage = {}\n",
    "for pages in pagesPairs:\n",
    "    templateSource = [(name,data) for name,data in  extract_templates(getContent(pages[0],sourceLang)).items() if name.lower() in templatesTwoLang[0].lower()] \n",
    "    templateTarget = [(name,data) for name,data in  extract_templates(getContent(pages[1],targetLang)).items() if name.lower() in templatesTwoLang[1].lower()] \n",
    "    if templateSource:\n",
    "        if templateTarget:\n",
    "            templatesPairsPerPage[pages[0]] = templatesPairsPerPage.get(pages[0],[])\n",
    "\n",
    "            templatesPairsPerPage[pages[0]].append((templateSource,templateTarget))\n",
    "            \n",
    "            \n",
    "\n",
    "\n",
    "transmat = np.loadtxt('fastText_multilingual/my_alingments/apply_in_en_to_es.txt')\n",
    "\n",
    "for pages, data in templatesPairsPerPage.items():\n",
    "    pageSourceLang = data[0][0][0][1]\n",
    "    pageTargetLang = data[0][1][0][1]\n",
    "    for param1,value in pageSourceLang.items():\n",
    "        try:\n",
    "            vec1 = myModel.get_sentence_vector(value.strip())\n",
    "            distancesNoTemplateData[param1] = []\n",
    "            for param2,value2 in pageTargetLang.items():\n",
    "                    vec2= myModel2.get_sentence_vector(value2.strip())\n",
    "                    vec2T = apply_transform(vec2,transmat)\n",
    "                    dist = distance.cosine(vec1,vec2T)\n",
    "                    distancesNoTemplateData[param1].append([dist,param2])\n",
    "        except:\n",
    "            pass\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 556,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "tiempo [[0.8237192646115215, 'years_active'], [0.8297616419425919, 'name'], [0.8372941016593429, 'children']]\n",
      "pie de imagen [[0.595959560523716, 'image'], [0.6550599889689077, 'other_names'], [0.7264075447081769, 'movement']]\n",
      "nombre [[0.4041053071307944, 'image'], [0.518457326659189, 'other_names'], [0.7447650042056859, 'citizenship']]\n",
      "nacionalidad [[0.6498525538332381, 'death_date'], [0.6839225568797711, 'death_place'], [0.6960114454594034, 'birth_place']]\n",
      "alma máter [[0.6260514271654607, 'image'], [0.6876603840579052, 'other_names'], [0.6956217498859136, 'movement']]\n",
      "tamaño [[0.6252674964760204, 'death_date'], [0.6543020957305372, 'birth_date'], [0.6881550150885685, 'alma_mater']]\n",
      "lugar de nacimiento [[0.6006641143067445, 'movement'], [0.651721071660169, 'citizenship'], [0.749902622171667, 'image']]\n",
      "website [[0.14517355546728405, 'website'], [0.33303012814333954, 'residence'], [0.4358272801220364, 'caption']]\n",
      "nombre de nacimiento [[0.5268444887815582, 'image'], [0.561234729287486, 'other_names'], [0.7458845651968874, 'citizenship']]\n",
      "fecha de fallecimiento [[0.405637195020605, 'death_date'], [0.5586009123247648, 'birth_date'], [0.5769374843869859, 'birth_place']]\n",
      "seudónimo [[0.3431979248660232, 'other_names'], [0.5191122790661059, 'image'], [0.7383798853791346, 'movement']]\n",
      "imagen [[0.4066104016008, 'image'], [0.5735632346552102, 'other_names'], [0.7130810176711444, 'citizenship']]\n",
      "lugar de fallecimiento [[0.4353487507509327, 'death_date'], [0.4372675858983963, 'birth_date'], [0.4903421781752545, 'birth_place']]\n",
      "ocupación [[0.6054499455608767, 'image'], [0.6647989597977009, 'movement'], [0.6799747139150318, 'other_names']]\n",
      "residencia [[0.6862214922582682, 'citizenship'], [0.7706700544198648, 'image'], [0.78425004764957, 'movement']]\n",
      "relacionados [[0.5925765268114909, 'image'], [0.5988486690224618, 'name'], [0.6005501826684927, 'caption']]\n",
      "medio de comunicación [[0.6175320186849692, 'image'], [0.6475602640619512, 'other_names'], [0.7843196163492381, 'movement']]\n",
      "conocido por [[0.37402284024828025, 'known_for'], [0.5769164000381919, 'caption'], [0.6158629193219431, 'residence']]\n",
      "abreviatura en zoología [[0.5300942275088152, 'parents'], [nan, 'style'], [0.7068505382194628, 'death_place']]\n",
      "fecha de nacimiento [[0.7068943512676559, 'movement'], [0.7435227743074442, 'image'], [0.7604446107799168, 'citizenship']]\n"
     ]
    }
   ],
   "source": [
    "cNotFound = 0\n",
    "for param,words in distancesNoTemplateData.items():\n",
    "    try:\n",
    "        if sorted(words)[0][0] < .9:\n",
    "            print(param,sorted(words)[0:3])\n",
    "        else:\n",
    "            cNotFound +=1\n",
    "    except:\n",
    "        pass"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "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": "3.5.3"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}