{
 "cells": [
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Eksperimentelle tjenester"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "from dhlab.future.corpus_conc_coll import Corpus, Concordance, Collocations, Frequencies"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "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>dhlabid</th>\n",
       "      <th>urn</th>\n",
       "      <th>title</th>\n",
       "      <th>authors</th>\n",
       "      <th>oaiid</th>\n",
       "      <th>sesamid</th>\n",
       "      <th>isbn10</th>\n",
       "      <th>city</th>\n",
       "      <th>timestamp</th>\n",
       "      <th>year</th>\n",
       "      <th>publisher</th>\n",
       "      <th>langs</th>\n",
       "      <th>subjects</th>\n",
       "      <th>ddc</th>\n",
       "      <th>genres</th>\n",
       "      <th>literaryform</th>\n",
       "      <th>doctype</th>\n",
       "      <th>ocr_creator</th>\n",
       "      <th>ocr_timestamp</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>100495998</td>\n",
       "      <td>URN:NBN:no-nb_digibok_2014061628001</td>\n",
       "      <td>Ulysses &amp; Outinus Unus &amp; Idem , Sive Disqvisit...</td>\n",
       "      <td>Ramus , Jonas / Lynov , Ove / Bircherod , Jens...</td>\n",
       "      <td>oai:nb.bibsys.no:990402670084702202</td>\n",
       "      <td>ae76cdc850ddfd619290b3e6dd860513</td>\n",
       "      <td></td>\n",
       "      <td></td>\n",
       "      <td>17130101</td>\n",
       "      <td>1713</td>\n",
       "      <td>Typis Ovidii Lynnovii</td>\n",
       "      <td>lat</td>\n",
       "      <td>Geografi / Historie / Oddysevs / Odin</td>\n",
       "      <td></td>\n",
       "      <td></td>\n",
       "      <td>Faglitteratur</td>\n",
       "      <td>digibok</td>\n",
       "      <td>nb</td>\n",
       "      <td>20060101</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>100327307</td>\n",
       "      <td>URN:NBN:no-nb_digibok_2017030348025</td>\n",
       "      <td>Regjeringsadvokat Johannes Bergh : 18. februar...</td>\n",
       "      <td>Bergh , Johannes</td>\n",
       "      <td>oai:nb.bibsys.no:999510918244702202</td>\n",
       "      <td>41d4e1afe9f378fbc01a15582cff6c56</td>\n",
       "      <td></td>\n",
       "      <td>&lt;S.l.&gt;</td>\n",
       "      <td>19370101</td>\n",
       "      <td>1937</td>\n",
       "      <td></td>\n",
       "      <td>nob</td>\n",
       "      <td></td>\n",
       "      <td></td>\n",
       "      <td></td>\n",
       "      <td>Uklassifisert</td>\n",
       "      <td>digibok</td>\n",
       "      <td>nb</td>\n",
       "      <td>20060101</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "     dhlabid                                  urn  \\\n",
       "0  100495998  URN:NBN:no-nb_digibok_2014061628001   \n",
       "1  100327307  URN:NBN:no-nb_digibok_2017030348025   \n",
       "\n",
       "                                               title  \\\n",
       "0  Ulysses & Outinus Unus & Idem , Sive Disqvisit...   \n",
       "1  Regjeringsadvokat Johannes Bergh : 18. februar...   \n",
       "\n",
       "                                             authors  \\\n",
       "0  Ramus , Jonas / Lynov , Ove / Bircherod , Jens...   \n",
       "1                                   Bergh , Johannes   \n",
       "\n",
       "                                 oaiid                           sesamid  \\\n",
       "0  oai:nb.bibsys.no:990402670084702202  ae76cdc850ddfd619290b3e6dd860513   \n",
       "1  oai:nb.bibsys.no:999510918244702202  41d4e1afe9f378fbc01a15582cff6c56   \n",
       "\n",
       "  isbn10    city  timestamp  year              publisher langs  \\\n",
       "0                  17130101  1713  Typis Ovidii Lynnovii   lat   \n",
       "1         <S.l.>   19370101  1937                          nob   \n",
       "\n",
       "                                subjects ddc genres   literaryform  doctype  \\\n",
       "0  Geografi / Historie / Oddysevs / Odin             Faglitteratur  digibok   \n",
       "1                                                    Uklassifisert  digibok   \n",
       "\n",
       "  ocr_creator ocr_timestamp  \n",
       "0          nb      20060101  \n",
       "1          nb      20060101  "
      ]
     },
     "execution_count": 2,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "c = Corpus.build(doctype=\"digibok\")\n",
    "c.head(2)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "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>Ut och leta växter ! : om bär och blommor i vår natur</th>\n",
       "      <th>Nye utfordringer i klagesaksbehandlingen : fra et nordisk seminar på Lillehammer den 5. - 7 . Marts 1997</th>\n",
       "      <th>Erasmus Montanus , eller Rasmus Berg : Komedie i fem Akter</th>\n",
       "      <th>Ulysses &amp; Outinus Unus &amp; Idem , Sive Disqvisitio Historica &amp; Geographica , Qva Ex collatis inter se Odyssea Homeri &amp; Edda Island . Homerizante , Outini fraudes deteguntur , ac , detracta larva in lucem protrahitur Ulysses</th>\n",
       "      <th>De norske Livsforsikringsselskapers Korrespondansekursus for yrkesakkvisitører</th>\n",
       "      <th>Det spirituelle og okkulte</th>\n",
       "      <th>Regjeringsadvokat Johannes Bergh : 18. februar 1837 - 18. februar 1937 : utg. på 100-årsdagen 18. februar 1937</th>\n",
       "      <th>Drama i Russland</th>\n",
       "      <th>Eksamensoppgaver i mekanikk ved universitetet i Oslo</th>\n",
       "      <th>Ord om fred</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>.</th>\n",
       "      <td>662.0</td>\n",
       "      <td>3318.0</td>\n",
       "      <td>1387.0</td>\n",
       "      <td>1673.0</td>\n",
       "      <td>3459.0</td>\n",
       "      <td>1780.0</td>\n",
       "      <td>307.0</td>\n",
       "      <td>2569.0</td>\n",
       "      <td>440.0</td>\n",
       "      <td>65.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>,</th>\n",
       "      <td>347.0</td>\n",
       "      <td>2059.0</td>\n",
       "      <td>1222.0</td>\n",
       "      <td>3120.0</td>\n",
       "      <td>2705.0</td>\n",
       "      <td>1493.0</td>\n",
       "      <td>312.0</td>\n",
       "      <td>1304.0</td>\n",
       "      <td>222.0</td>\n",
       "      <td>96.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>och</th>\n",
       "      <td>293.0</td>\n",
       "      <td>395.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>år</th>\n",
       "      <td>229.0</td>\n",
       "      <td>341.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>289.0</td>\n",
       "      <td>27.0</td>\n",
       "      <td>21.0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>du</th>\n",
       "      <td>227.0</td>\n",
       "      <td>9.0</td>\n",
       "      <td>85.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>179.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>4.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>NYE</th>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>3.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>UKJENT</th>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>3.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>WILLIAM</th>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>3.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>GAMLE</th>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>6.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>TESTAMENTET</th>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>9.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>49061 rows × 10 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "             Ut och leta växter ! : om bär och blommor i vår natur  \\\n",
       ".                                                        662.0       \n",
       ",                                                        347.0       \n",
       "och                                                      293.0       \n",
       "år                                                       229.0       \n",
       "du                                                       227.0       \n",
       "...                                                        ...       \n",
       "NYE                                                        0.0       \n",
       "UKJENT                                                     0.0       \n",
       "WILLIAM                                                    0.0       \n",
       "GAMLE                                                      0.0       \n",
       "TESTAMENTET                                                0.0       \n",
       "\n",
       "             Nye utfordringer i klagesaksbehandlingen : fra et nordisk seminar på Lillehammer den 5. - 7 . Marts 1997  \\\n",
       ".                                                       3318.0                                                          \n",
       ",                                                       2059.0                                                          \n",
       "och                                                      395.0                                                          \n",
       "år                                                       341.0                                                          \n",
       "du                                                         9.0                                                          \n",
       "...                                                        ...                                                          \n",
       "NYE                                                        0.0                                                          \n",
       "UKJENT                                                     0.0                                                          \n",
       "WILLIAM                                                    0.0                                                          \n",
       "GAMLE                                                      0.0                                                          \n",
       "TESTAMENTET                                                0.0                                                          \n",
       "\n",
       "             Erasmus Montanus , eller Rasmus Berg : Komedie i fem Akter  \\\n",
       ".                                                       1387.0            \n",
       ",                                                       1222.0            \n",
       "och                                                        0.0            \n",
       "år                                                         0.0            \n",
       "du                                                        85.0            \n",
       "...                                                        ...            \n",
       "NYE                                                        0.0            \n",
       "UKJENT                                                     0.0            \n",
       "WILLIAM                                                    0.0            \n",
       "GAMLE                                                      0.0            \n",
       "TESTAMENTET                                                0.0            \n",
       "\n",
       "             Ulysses & Outinus Unus & Idem , Sive Disqvisitio Historica & Geographica , Qva Ex collatis inter se Odyssea Homeri & Edda Island . Homerizante , Outini fraudes deteguntur , ac , detracta larva in lucem protrahitur Ulysses  \\\n",
       ".                                                       1673.0                                                                                                                                                                               \n",
       ",                                                       3120.0                                                                                                                                                                               \n",
       "och                                                        0.0                                                                                                                                                                               \n",
       "år                                                         0.0                                                                                                                                                                               \n",
       "du                                                         3.0                                                                                                                                                                               \n",
       "...                                                        ...                                                                                                                                                                               \n",
       "NYE                                                        0.0                                                                                                                                                                               \n",
       "UKJENT                                                     0.0                                                                                                                                                                               \n",
       "WILLIAM                                                    0.0                                                                                                                                                                               \n",
       "GAMLE                                                      0.0                                                                                                                                                                               \n",
       "TESTAMENTET                                                0.0                                                                                                                                                                               \n",
       "\n",
       "             De norske Livsforsikringsselskapers Korrespondansekursus for yrkesakkvisitører  \\\n",
       ".                                                       3459.0                                \n",
       ",                                                       2705.0                                \n",
       "och                                                        0.0                                \n",
       "år                                                       289.0                                \n",
       "du                                                         0.0                                \n",
       "...                                                        ...                                \n",
       "NYE                                                        0.0                                \n",
       "UKJENT                                                     0.0                                \n",
       "WILLIAM                                                    0.0                                \n",
       "GAMLE                                                      0.0                                \n",
       "TESTAMENTET                                                0.0                                \n",
       "\n",
       "             Det spirituelle og okkulte  \\\n",
       ".                                1780.0   \n",
       ",                                1493.0   \n",
       "och                                 0.0   \n",
       "år                                 27.0   \n",
       "du                                  2.0   \n",
       "...                                 ...   \n",
       "NYE                                 0.0   \n",
       "UKJENT                              0.0   \n",
       "WILLIAM                             0.0   \n",
       "GAMLE                               0.0   \n",
       "TESTAMENTET                         0.0   \n",
       "\n",
       "             Regjeringsadvokat Johannes Bergh : 18. februar 1837 - 18. februar 1937 : utg. på 100-årsdagen 18. februar 1937  \\\n",
       ".                                                        307.0                                                                \n",
       ",                                                        312.0                                                                \n",
       "och                                                        0.0                                                                \n",
       "år                                                        21.0                                                                \n",
       "du                                                         3.0                                                                \n",
       "...                                                        ...                                                                \n",
       "NYE                                                        0.0                                                                \n",
       "UKJENT                                                     0.0                                                                \n",
       "WILLIAM                                                    0.0                                                                \n",
       "GAMLE                                                      0.0                                                                \n",
       "TESTAMENTET                                                0.0                                                                \n",
       "\n",
       "             Drama i Russland  \\\n",
       ".                      2569.0   \n",
       ",                      1304.0   \n",
       "och                       0.0   \n",
       "år                        5.0   \n",
       "du                      179.0   \n",
       "...                       ...   \n",
       "NYE                       0.0   \n",
       "UKJENT                    0.0   \n",
       "WILLIAM                   0.0   \n",
       "GAMLE                     0.0   \n",
       "TESTAMENTET               0.0   \n",
       "\n",
       "             Eksamensoppgaver i mekanikk ved universitetet i Oslo  Ord om fred  \n",
       ".                                                        440.0            65.0  \n",
       ",                                                        222.0            96.0  \n",
       "och                                                        0.0             0.0  \n",
       "år                                                         0.0             1.0  \n",
       "du                                                         0.0             4.0  \n",
       "...                                                        ...             ...  \n",
       "NYE                                                        0.0             3.0  \n",
       "UKJENT                                                     0.0             3.0  \n",
       "WILLIAM                                                    0.0             3.0  \n",
       "GAMLE                                                      0.0             6.0  \n",
       "TESTAMENTET                                                0.0             9.0  \n",
       "\n",
       "[49061 rows x 10 columns]"
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "c.get_freqs().display_names()"
   ]
  },
  {
   "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>frequencies</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>.</th>\n",
       "      <td>15660.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>,</th>\n",
       "      <td>12880.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>och</th>\n",
       "      <td>688.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>år</th>\n",
       "      <td>913.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>du</th>\n",
       "      <td>512.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>NYE</th>\n",
       "      <td>3.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>UKJENT</th>\n",
       "      <td>3.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>WILLIAM</th>\n",
       "      <td>3.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>GAMLE</th>\n",
       "      <td>6.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>TESTAMENTET</th>\n",
       "      <td>9.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>49061 rows × 1 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "             frequencies\n",
       ".                15660.0\n",
       ",                12880.0\n",
       "och                688.0\n",
       "år                 913.0\n",
       "du                 512.0\n",
       "...                  ...\n",
       "NYE                  3.0\n",
       "UKJENT               3.0\n",
       "WILLIAM              3.0\n",
       "GAMLE                6.0\n",
       "TESTAMENTET          9.0\n",
       "\n",
       "[49061 rows x 1 columns]"
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "c.get_freqs().sum_freqs()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "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>index</th>\n",
       "      <th>docid</th>\n",
       "      <th>urn</th>\n",
       "      <th>concordance</th>\n",
       "      <th>link</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0</td>\n",
       "      <td>100279411</td>\n",
       "      <td>URN:NBN:no-nb_digibok_2016011848084</td>\n",
       "      <td>... Mwaash eller Woot &lt; Mashamboy &gt; er Kubas f...</td>\n",
       "      <td>&lt;a target='_blank' href = 'https://urn.nb.no/U...</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   index      docid                                  urn  \\\n",
       "0      0  100279411  URN:NBN:no-nb_digibok_2016011848084   \n",
       "\n",
       "                                         concordance  \\\n",
       "0  ... Mwaash eller Woot < Mashamboy > er Kubas f...   \n",
       "\n",
       "                                                link  \n",
       "0  <a target='_blank' href = 'https://urn.nb.no/U...  "
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "Concordance.get_concordances(corpus=c, words=\"konge\").head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "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>counts</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>,</th>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>første</th>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>den</th>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Kubas</th>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>er</th>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>alle</th>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>&gt;</th>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>i</th>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Mashamboy</th>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Kuba-kulturen</th>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>&lt;</th>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>kan</th>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Woot</th>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>føre</th>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>sin</th>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>eller</th>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Mwaash</th>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>avstamning</th>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>.</th>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>tilbake</th>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>blir</th>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Kubas</th>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>der</th>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>seremonier</th>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>mytiske</th>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>under</th>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>opprinnelse</th>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>båret</th>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>frem</th>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>De</th>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>sti</th>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>lt</th>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>.</th>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>,</th>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>)</th>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>men</th>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Zaire</th>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>tidligere</th>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>også</th>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "               counts\n",
       ",                   1\n",
       "første              1\n",
       "den                 1\n",
       "Kubas               1\n",
       "er                  1\n",
       "alle                1\n",
       ">                   1\n",
       "i                   1\n",
       "Mashamboy           1\n",
       "Kuba-kulturen       1\n",
       "<                   1\n",
       "kan                 1\n",
       "Woot                1\n",
       "føre                1\n",
       "sin                 1\n",
       "eller               1\n",
       "Mwaash              1\n",
       "avstamning          1\n",
       ".                   1\n",
       "tilbake             1\n",
       "blir                2\n",
       "Kubas               1\n",
       "der                 1\n",
       "seremonier          1\n",
       "mytiske             1\n",
       "under               1\n",
       "opprinnelse         1\n",
       "båret               1\n",
       "frem                1\n",
       "De                  1\n",
       "sti                 1\n",
       "lt                  1\n",
       ".                   1\n",
       ",                   1\n",
       ")                   1\n",
       "men                 1\n",
       "Zaire               1\n",
       "tidligere           1\n",
       "også                1"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "Collocations.get_collocations(corpus=c, words=[\"konge\", \"kongeriket\"])"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "dhlab",
   "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": "3.10.12"
  },
  "orig_nbformat": 4,
  "vscode": {
   "interpreter": {
    "hash": "1d1df0d064732dbd8ae09ceab87be2790d7eed279040ea9405054873f855fb6c"
   }
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}