{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "78729"
      ]
     },
     "execution_count": 1,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "\"\"\"https://www.science.org/doi/10.1126/science.1199295\"\"\"\n",
    "\n",
    "import http.cookiejar\n",
    "import io\n",
    "import pathlib\n",
    "import re\n",
    "import shutil\n",
    "import subprocess\n",
    "import urllib.parse\n",
    "import urllib.request\n",
    "import zipfile\n",
    "\n",
    "DOI = '10.1126/science.1199295'\n",
    "\n",
    "PDF = pathlib.Path('1199295-atkinson-som.pdf')\n",
    "\n",
    "TXT = PDF.with_suffix('.txt')\n",
    "\n",
    "URL = 'https://www.science.org/action/downloadSupplement'\n",
    "\n",
    "URL = (urllib.parse.urlparse(URL)\n",
    "       ._replace(query=urllib.parse.urlencode({'doi': DOI,\n",
    "                                               'file': PDF})))\n",
    "\n",
    "USER_AGENT = ('Mozilla/5.0 (X11; U; Linux i686)'\n",
    "              ' Gecko/20071127 Firefox/2.0.0.11')\n",
    "\n",
    "ENCODING = 'utf-8'\n",
    "\n",
    "\n",
    "if not PDF.exists():\n",
    "    cookies = http.cookiejar.CookieJar()\n",
    "    processor = urllib.request.HTTPCookieProcessor(cookies)\n",
    "    opener = urllib.request.build_opener(processor)\n",
    "\n",
    "    request = urllib.request.Request(URL.geturl(),\n",
    "                                     headers={'User-Agent': USER_AGENT})\n",
    "    with opener.open(request) as u, PDF.open('wb') as f:\n",
    "        shutil.copyfileobj(u, f)\n",
    "\n",
    "if not TXT.exists():  # requires one of popper-utils, miktex-poppler-bin, xpdf\n",
    "    cmd = ['pdftotext', '-f', '21', '-l', '33', '-layout', '-nopgbrk', PDF, TXT]\n",
    "    subprocess.run(cmd, check=True)\n",
    "\n",
    "_, _, S1 = TXT.read_text(encoding=ENCODING).strip().partition('\\n\\n\\n')\n",
    "\n",
    "assert ',' not in S1\n",
    "\n",
    "len(S1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(12, 504)"
      ]
     },
     "execution_count": 2,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "S1_HEAD = ['Language Name',\n",
    "           'WALS code',\n",
    "           'Family',\n",
    "           'Latitude',\n",
    "           'Longitude',\n",
    "           'Normalized Vowel Diversity',\n",
    "           'Normalized Consonant Diversity',\n",
    "           'Normalized Tone Diversity',\n",
    "           'Total Normalized Phoneme Diversity',\n",
    "           'ISO codes',\n",
    "           'Estimated Speaker Pop. Size',\n",
    "           'Distance from best fit origin']\n",
    "\n",
    "S1_FEAT = [h for h in S1_HEAD if h.startswith('Normalized ')]\n",
    "S1_OUTC = S1_HEAD[-2:]\n",
    "\n",
    "S1_DATA = S1[S1.index('\\nAbkhaz '):].strip()\n",
    "\n",
    "S1_DATA = re.sub(r'(\\))([a-z]{3} )', r'\\1 \\2', S1_DATA)  # fix missing space\n",
    "S1_DATA = re.sub(r'\\n +([a-z]{3})\\n(.+?)  ', r'\\n\\2 \\1 ', S1_DATA)  # fix splitted lines\n",
    "S1_DATA = S1_DATA.replace('(San Mateo del Mar)', '(San Mateo Del Mar)')  # simplify iso detection\n",
    "\n",
    "S1_DATA = [l.strip() for l in S1_DATA.splitlines()]\n",
    "\n",
    "\n",
    "len(S1_HEAD), len(S1_DATA)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "'Language Name,WALS code,Family,Latitude,Longitude,Normalized Vowel Diversity,Normalized Consonant Diversity,Normalized Tone Diversity,Total Normalized Phoneme Diversity,ISO codes,Estimated Speaker Pop. Size,Distance from best fit origin\\nAbkhaz,abk,Northwest Caucasian,43.08,41,-1.2345266,-1.5544112,-0.7687792,-1.185905651,abk,105952,5856.362\\nAcoma,aco,Keresan,34.92,-107.58,-0.4846364,-0.7169629,1.86204304,0.220147906,kjq,3391,18601.19\\nAndoke,adk,Andoke,-0.67,-72,1.3900889,-1.5544112,0.54663194,0.'"
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "ROW = re.compile(r'''^\n",
    "(.+?)\\ +\n",
    "([a-z]{2,3})\\ +\n",
    "(.+?)\\ +\n",
    "(-?\\d+(?:\\.\\d+)?)\\ +\n",
    "(-?\\d+(?:\\.\\d+)?)\\ +\n",
    "(-?\\d+\\.\\d+)\\ +\n",
    "(-?\\d+\\.\\d+)\\ +\n",
    "(-?\\d+\\.\\d+)\\ +\n",
    "(-?\\d+\\.\\d+)\\ +\n",
    "(.+?)\\ +\n",
    "(\\d+)\\ +\n",
    "(\\d+(?:\\.\\d+)?)\n",
    "$''', flags=re.VERBOSE)\n",
    "\n",
    "\n",
    "csv = '\\n'.join(','.join(r) for r in [S1_HEAD] + [ROW.match(l).groups() for l in S1_DATA])\n",
    "\n",
    "csv[:500]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'pandas.core.frame.DataFrame'>\n",
      "Index: 504 entries, abk to zun\n",
      "Data columns (total 11 columns):\n",
      " #   Column                              Non-Null Count  Dtype  \n",
      "---  ------                              --------------  -----  \n",
      " 0   Language Name                       504 non-null    object \n",
      " 1   Family                              504 non-null    object \n",
      " 2   Latitude                            504 non-null    float64\n",
      " 3   Longitude                           504 non-null    float64\n",
      " 4   Normalized Vowel Diversity          504 non-null    float64\n",
      " 5   Normalized Consonant Diversity      504 non-null    float64\n",
      " 6   Normalized Tone Diversity           504 non-null    float64\n",
      " 7   Total Normalized Phoneme Diversity  504 non-null    float64\n",
      " 8   ISO codes                           504 non-null    object \n",
      " 9   Estimated Speaker Pop. Size         504 non-null    int64  \n",
      " 10  Distance from best fit origin       504 non-null    float64\n",
      "dtypes: float64(7), int64(1), object(3)\n",
      "memory usage: 47.2+ KB\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Language Name</th>\n",
       "      <th>Family</th>\n",
       "      <th>Latitude</th>\n",
       "      <th>Longitude</th>\n",
       "      <th>Normalized Vowel Diversity</th>\n",
       "      <th>Normalized Consonant Diversity</th>\n",
       "      <th>Normalized Tone Diversity</th>\n",
       "      <th>Total Normalized Phoneme Diversity</th>\n",
       "      <th>ISO codes</th>\n",
       "      <th>Estimated Speaker Pop. Size</th>\n",
       "      <th>Distance from best fit origin</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>WALS code</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>abk</th>\n",
       "      <td>Abkhaz</td>\n",
       "      <td>Northwest Caucasian</td>\n",
       "      <td>43.08</td>\n",
       "      <td>41.00</td>\n",
       "      <td>-1.234527</td>\n",
       "      <td>-1.554411</td>\n",
       "      <td>-0.768779</td>\n",
       "      <td>-1.185906</td>\n",
       "      <td>abk</td>\n",
       "      <td>105952</td>\n",
       "      <td>5856.362</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>aco</th>\n",
       "      <td>Acoma</td>\n",
       "      <td>Keresan</td>\n",
       "      <td>34.92</td>\n",
       "      <td>-107.58</td>\n",
       "      <td>-0.484636</td>\n",
       "      <td>-0.716963</td>\n",
       "      <td>1.862043</td>\n",
       "      <td>0.220148</td>\n",
       "      <td>kjq</td>\n",
       "      <td>3391</td>\n",
       "      <td>18601.190</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>adk</th>\n",
       "      <td>Andoke</td>\n",
       "      <td>Andoke</td>\n",
       "      <td>-0.67</td>\n",
       "      <td>-72.00</td>\n",
       "      <td>1.390089</td>\n",
       "      <td>-1.554411</td>\n",
       "      <td>0.546632</td>\n",
       "      <td>0.127437</td>\n",
       "      <td>ano</td>\n",
       "      <td>619</td>\n",
       "      <td>23780.160</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>aea</th>\n",
       "      <td>Aleut (Eastern)</td>\n",
       "      <td>Eskimo-Aleut</td>\n",
       "      <td>54.75</td>\n",
       "      <td>-164.00</td>\n",
       "      <td>-1.234527</td>\n",
       "      <td>0.957934</td>\n",
       "      <td>-0.768779</td>\n",
       "      <td>-0.348457</td>\n",
       "      <td>ale</td>\n",
       "      <td>490</td>\n",
       "      <td>14629.420</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>aeg</th>\n",
       "      <td>Arabic (Egyptian)</td>\n",
       "      <td>Afro-Asiatic</td>\n",
       "      <td>30.00</td>\n",
       "      <td>31.00</td>\n",
       "      <td>-0.484636</td>\n",
       "      <td>0.120485</td>\n",
       "      <td>-0.768779</td>\n",
       "      <td>-0.377643</td>\n",
       "      <td>arz</td>\n",
       "      <td>46321000</td>\n",
       "      <td>4153.443</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "               Language Name               Family  Latitude  Longitude  \\\n",
       "WALS code                                                                \n",
       "abk                   Abkhaz  Northwest Caucasian     43.08      41.00   \n",
       "aco                    Acoma              Keresan     34.92    -107.58   \n",
       "adk                   Andoke               Andoke     -0.67     -72.00   \n",
       "aea          Aleut (Eastern)         Eskimo-Aleut     54.75    -164.00   \n",
       "aeg        Arabic (Egyptian)         Afro-Asiatic     30.00      31.00   \n",
       "\n",
       "           Normalized Vowel Diversity  Normalized Consonant Diversity  \\\n",
       "WALS code                                                               \n",
       "abk                         -1.234527                       -1.554411   \n",
       "aco                         -0.484636                       -0.716963   \n",
       "adk                          1.390089                       -1.554411   \n",
       "aea                         -1.234527                        0.957934   \n",
       "aeg                         -0.484636                        0.120485   \n",
       "\n",
       "           Normalized Tone Diversity  Total Normalized Phoneme Diversity  \\\n",
       "WALS code                                                                  \n",
       "abk                        -0.768779                           -1.185906   \n",
       "aco                         1.862043                            0.220148   \n",
       "adk                         0.546632                            0.127437   \n",
       "aea                        -0.768779                           -0.348457   \n",
       "aeg                        -0.768779                           -0.377643   \n",
       "\n",
       "          ISO codes  Estimated Speaker Pop. Size  \\\n",
       "WALS code                                          \n",
       "abk             abk                       105952   \n",
       "aco             kjq                         3391   \n",
       "adk             ano                          619   \n",
       "aea             ale                          490   \n",
       "aeg             arz                     46321000   \n",
       "\n",
       "           Distance from best fit origin  \n",
       "WALS code                                 \n",
       "abk                             5856.362  \n",
       "aco                            18601.190  \n",
       "adk                            23780.160  \n",
       "aea                            14629.420  \n",
       "aeg                             4153.443  "
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "%matplotlib inline\n",
    "\n",
    "import itertools\n",
    "\n",
    "import numpy as np\n",
    "import pandas as pd\n",
    "import scipy.stats\n",
    "\n",
    "\n",
    "def pearsonr(df, left, right, func=scipy.stats.pearsonr):\n",
    "    df = df[[left, right]].dropna()\n",
    "    name = '%s & %s' % (left, right)\n",
    "    return pd.Series(func(df[left], df[right]), index=('r', 'p'), name=name)\n",
    "\n",
    "\n",
    "with io.StringIO(csv) as f:\n",
    "    s1 = pd.read_csv(f, na_values='', keep_default_na=False, index_col='WALS code')\n",
    "\n",
    "s1.info()\n",
    "assert s1.index.is_unique\n",
    "assert s1.index.is_monotonic_increasing\n",
    "s1.head()"
   ]
  },
  {
   "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>r</th>\n",
       "      <th>p</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>Estimated Speaker Pop. Size &amp; Normalized Vowel Diversity</th>\n",
       "      <td>0.03153</td>\n",
       "      <td>0.48006</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Estimated Speaker Pop. Size &amp; Normalized Consonant Diversity</th>\n",
       "      <td>0.08969</td>\n",
       "      <td>0.04415</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Estimated Speaker Pop. Size &amp; Normalized Tone Diversity</th>\n",
       "      <td>0.03651</td>\n",
       "      <td>0.41345</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Estimated Speaker Pop. Size &amp; Total Normalized Phoneme Diversity</th>\n",
       "      <td>0.08219</td>\n",
       "      <td>0.06523</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Distance from best fit origin &amp; Normalized Vowel Diversity</th>\n",
       "      <td>-0.39421</td>\n",
       "      <td>0.00000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Distance from best fit origin &amp; Normalized Consonant Diversity</th>\n",
       "      <td>-0.26014</td>\n",
       "      <td>0.00000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Distance from best fit origin &amp; Normalized Tone Diversity</th>\n",
       "      <td>-0.39063</td>\n",
       "      <td>0.00000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Distance from best fit origin &amp; Total Normalized Phoneme Diversity</th>\n",
       "      <td>-0.54466</td>\n",
       "      <td>0.00000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                                          r        p\n",
       "Estimated Speaker Pop. Size & Normalized Vowel ...  0.03153  0.48006\n",
       "Estimated Speaker Pop. Size & Normalized Conson...  0.08969  0.04415\n",
       "Estimated Speaker Pop. Size & Normalized Tone D...  0.03651  0.41345\n",
       "Estimated Speaker Pop. Size & Total Normalized ...  0.08219  0.06523\n",
       "Distance from best fit origin & Normalized Vowe... -0.39421  0.00000\n",
       "Distance from best fit origin & Normalized Cons... -0.26014  0.00000\n",
       "Distance from best fit origin & Normalized Tone... -0.39063  0.00000\n",
       "Distance from best fit origin & Total Normalize... -0.54466  0.00000"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "PAIRS = list(itertools.product(S1_OUTC, S1_FEAT + ['Total Normalized Phoneme Diversity']))\n",
    "\n",
    "\n",
    "pd.DataFrame([pearsonr(s1, x, y) for x, y in PAIRS]).round(5)"
   ]
  },
  {
   "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>r</th>\n",
       "      <th>p</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>Population &amp; Total Normalized Phoneme Diversity</th>\n",
       "      <td>0.38454</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                                       r    p\n",
       "Population & Total Normalized Phoneme Diversity  0.38454  0.0"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "s1.plot.scatter(x='Estimated Speaker Pop. Size', y='Total Normalized Phoneme Diversity', logx=True);\n",
    "\n",
    "(pearsonr(s1.assign(Population=s1['Estimated Speaker Pop. Size'].apply(np.log10)),\n",
    "         'Population', 'Total Normalized Phoneme Diversity')\n",
    " .to_frame().T.round(5))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>r</th>\n",
       "      <th>p</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>Distance &amp; Total Normalized Phoneme Diversity</th>\n",
       "      <td>-0.54466</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                                     r    p\n",
       "Distance & Total Normalized Phoneme Diversity -0.54466  0.0"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "s1.plot.scatter(x='Distance from best fit origin', y='Total Normalized Phoneme Diversity');\n",
    "\n",
    "(pearsonr(s1.assign(Distance=s1['Distance from best fit origin']),\n",
    "         'Distance', 'Total Normalized Phoneme Diversity')\n",
    " .to_frame().T.round(5))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'pandas.core.frame.DataFrame'>\n",
      "Index: 2679 entries, aab to zzo\n",
      "Data columns (total 7 columns):\n",
      " #   Column                        Non-Null Count  Dtype  \n",
      "---  ------                        --------------  -----  \n",
      " 0   Name                          2679 non-null   object \n",
      " 1   latitude                      2679 non-null   float64\n",
      " 2   longitude                     2679 non-null   float64\n",
      " 3   family                        2679 non-null   object \n",
      " 4   1A Consonant Inventories      563 non-null    object \n",
      " 5   2A Vowel Quality Inventories  564 non-null    object \n",
      " 6   13A Tone                      527 non-null    object \n",
      "dtypes: float64(2), object(5)\n",
      "memory usage: 167.4+ KB\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Name</th>\n",
       "      <th>latitude</th>\n",
       "      <th>longitude</th>\n",
       "      <th>family</th>\n",
       "      <th>1A Consonant Inventories</th>\n",
       "      <th>2A Vowel Quality Inventories</th>\n",
       "      <th>13A Tone</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>wals_code</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>aab</th>\n",
       "      <td>Arapesh (Abu)</td>\n",
       "      <td>-3.450000</td>\n",
       "      <td>142.950000</td>\n",
       "      <td>Torricelli</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>aar</th>\n",
       "      <td>Aari</td>\n",
       "      <td>6.000000</td>\n",
       "      <td>36.583333</td>\n",
       "      <td>Afro-Asiatic</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>aba</th>\n",
       "      <td>Abau</td>\n",
       "      <td>-4.000000</td>\n",
       "      <td>141.250000</td>\n",
       "      <td>Sepik</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>abb</th>\n",
       "      <td>Arabic (Chadian)</td>\n",
       "      <td>13.833333</td>\n",
       "      <td>20.833333</td>\n",
       "      <td>Afro-Asiatic</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>abd</th>\n",
       "      <td>Abidji</td>\n",
       "      <td>5.666667</td>\n",
       "      <td>-4.583333</td>\n",
       "      <td>Niger-Congo</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                       Name   latitude   longitude        family  \\\n",
       "wals_code                                                          \n",
       "aab           Arapesh (Abu)  -3.450000  142.950000    Torricelli   \n",
       "aar                    Aari   6.000000   36.583333  Afro-Asiatic   \n",
       "aba                    Abau  -4.000000  141.250000         Sepik   \n",
       "abb        Arabic (Chadian)  13.833333   20.833333  Afro-Asiatic   \n",
       "abd                  Abidji   5.666667   -4.583333   Niger-Congo   \n",
       "\n",
       "          1A Consonant Inventories 2A Vowel Quality Inventories 13A Tone  \n",
       "wals_code                                                                 \n",
       "aab                            NaN                          NaN      NaN  \n",
       "aar                            NaN                          NaN      NaN  \n",
       "aba                            NaN                          NaN      NaN  \n",
       "abb                            NaN                          NaN      NaN  \n",
       "abd                            NaN                          NaN      NaN  "
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "URL = 'http://cdstar.shh.mpg.de/bitstreams/EAEA0-7269-77E5-3E10-0/wals_language.csv.zip'\n",
    "\n",
    "ARCHIVE = pathlib.Path(URL.rpartition('/')[2])\n",
    "\n",
    "EXTRACT = 'language.csv'\n",
    "\n",
    "INFO = ['wals_code', 'Name', 'family', 'latitude', 'longitude']\n",
    "\n",
    "FEAT = ['2A Vowel Quality Inventories', '1A Consonant Inventories', '13A Tone']\n",
    "\n",
    "\n",
    "if not ARCHIVE.exists():\n",
    "    urllib.request.urlretrieve(URL, ARCHIVE)\n",
    "\n",
    "with zipfile.ZipFile(ARCHIVE) as archive:\n",
    "    wf = pd.read_csv(archive.open(EXTRACT), encoding=ENCODING,\n",
    "                     na_values='', keep_default_na=False,\n",
    "                     index_col=INFO[0], usecols=INFO + FEAT)\n",
    "\n",
    "wf.info()\n",
    "assert wf.index.is_unique\n",
    "wf.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'pandas.core.frame.DataFrame'>\n",
      "Index: 567 entries, abi to zun\n",
      "Data columns (total 7 columns):\n",
      " #   Column                        Non-Null Count  Dtype  \n",
      "---  ------                        --------------  -----  \n",
      " 0   Name                          567 non-null    object \n",
      " 1   latitude                      567 non-null    float64\n",
      " 2   longitude                     567 non-null    float64\n",
      " 3   family                        567 non-null    object \n",
      " 4   1A Consonant Inventories      563 non-null    object \n",
      " 5   2A Vowel Quality Inventories  564 non-null    object \n",
      " 6   13A Tone                      527 non-null    object \n",
      "dtypes: float64(2), object(5)\n",
      "memory usage: 35.4+ KB\n"
     ]
    }
   ],
   "source": [
    "wf = wf.dropna(how='all', subset=FEAT)\n",
    "\n",
    "wf.info()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'pandas.core.frame.DataFrame'>\n",
      "Index: 504 entries, abk to zun\n",
      "Data columns (total 18 columns):\n",
      " #   Column                              Non-Null Count  Dtype  \n",
      "---  ------                              --------------  -----  \n",
      " 0   Language Name                       504 non-null    object \n",
      " 1   Family                              504 non-null    object \n",
      " 2   Latitude                            504 non-null    float64\n",
      " 3   Longitude                           504 non-null    float64\n",
      " 4   Normalized Vowel Diversity          504 non-null    float64\n",
      " 5   Normalized Consonant Diversity      504 non-null    float64\n",
      " 6   Normalized Tone Diversity           504 non-null    float64\n",
      " 7   Total Normalized Phoneme Diversity  504 non-null    float64\n",
      " 8   ISO codes                           504 non-null    object \n",
      " 9   Estimated Speaker Pop. Size         504 non-null    int64  \n",
      " 10  Distance from best fit origin       504 non-null    float64\n",
      " 11  Name                                504 non-null    object \n",
      " 12  latitude                            504 non-null    float64\n",
      " 13  longitude                           504 non-null    float64\n",
      " 14  family                              504 non-null    object \n",
      " 15  1A Consonant Inventories            504 non-null    object \n",
      " 16  2A Vowel Quality Inventories        504 non-null    object \n",
      " 17  13A Tone                            504 non-null    object \n",
      "dtypes: float64(9), int64(1), object(8)\n",
      "memory usage: 91.0+ KB\n"
     ]
    }
   ],
   "source": [
    "df = s1.join(wf)\n",
    "\n",
    "df.info()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "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>Name</th>\n",
       "      <th>latitude</th>\n",
       "      <th>longitude</th>\n",
       "      <th>family</th>\n",
       "      <th>1A Consonant Inventories</th>\n",
       "      <th>2A Vowel Quality Inventories</th>\n",
       "      <th>13A Tone</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>abi</th>\n",
       "      <td>Abipón</td>\n",
       "      <td>-29.000000</td>\n",
       "      <td>-61.000000</td>\n",
       "      <td>Guaicuruan</td>\n",
       "      <td>2 Moderately small</td>\n",
       "      <td>2 Average (5-6)</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>abm</th>\n",
       "      <td>Alabama</td>\n",
       "      <td>32.333333</td>\n",
       "      <td>-87.416667</td>\n",
       "      <td>Muskogean</td>\n",
       "      <td>1 Small</td>\n",
       "      <td>1 Small (2-4)</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>ach</th>\n",
       "      <td>Aché</td>\n",
       "      <td>-25.250000</td>\n",
       "      <td>-55.166667</td>\n",
       "      <td>Tupian</td>\n",
       "      <td>1 Small</td>\n",
       "      <td>2 Average (5-6)</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>acm</th>\n",
       "      <td>Achumawi</td>\n",
       "      <td>41.500000</td>\n",
       "      <td>-121.000000</td>\n",
       "      <td>Hokan</td>\n",
       "      <td>2 Moderately small</td>\n",
       "      <td>2 Average (5-6)</td>\n",
       "      <td>2 Simple tone system</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>adz</th>\n",
       "      <td>Adzera</td>\n",
       "      <td>-6.250000</td>\n",
       "      <td>146.250000</td>\n",
       "      <td>Austronesian</td>\n",
       "      <td>2 Moderately small</td>\n",
       "      <td>1 Small (2-4)</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "         Name   latitude   longitude        family 1A Consonant Inventories  \\\n",
       "abi    Abipón -29.000000  -61.000000    Guaicuruan       2 Moderately small   \n",
       "abm   Alabama  32.333333  -87.416667     Muskogean                  1 Small   \n",
       "ach      Aché -25.250000  -55.166667        Tupian                  1 Small   \n",
       "acm  Achumawi  41.500000 -121.000000         Hokan       2 Moderately small   \n",
       "adz    Adzera  -6.250000  146.250000  Austronesian       2 Moderately small   \n",
       "\n",
       "    2A Vowel Quality Inventories              13A Tone  \n",
       "abi              2 Average (5-6)                   NaN  \n",
       "abm                1 Small (2-4)                   NaN  \n",
       "ach              2 Average (5-6)                   NaN  \n",
       "acm              2 Average (5-6)  2 Simple tone system  \n",
       "adz                1 Small (2-4)                   NaN  "
      ]
     },
     "execution_count": 11,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "wf.loc[wf.index.difference(s1.index)].head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "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>mismatches</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>2A Vowel Quality Inventories</th>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1A Consonant Inventories</th>\n",
       "      <td>86</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>13A Tone</th>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                              mismatches\n",
       "2A Vowel Quality Inventories           1\n",
       "1A Consonant Inventories              86\n",
       "13A Tone                               0"
      ]
     },
     "execution_count": 12,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "COLS = list(itertools.chain.from_iterable(zip(S1_FEAT, FEAT)))\n",
    "\n",
    "MAP = dict(zip(S1_FEAT, FEAT))\n",
    "\n",
    "\n",
    "_category = pd.api.types.CategoricalDtype(ordered=True)\n",
    "\n",
    "df[COLS] = df[COLS].apply(lambda x: x.astype(_category).cat.codes)\n",
    "\n",
    "mism = df[S1_FEAT].rename(columns=MAP) != df[FEAT]\n",
    "\n",
    "mism.sum().to_frame('mismatches')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "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>Normalized Vowel Diversity</th>\n",
       "      <th>2A Vowel Quality Inventories</th>\n",
       "      <th>Normalized Consonant Diversity</th>\n",
       "      <th>1A Consonant Inventories</th>\n",
       "      <th>Normalized Tone Diversity</th>\n",
       "      <th>13A Tone</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>WALS code</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>abk</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>4</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>aco</th>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>4</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>aea</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>3</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>aeg</th>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>3</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>agh</th>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>3</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "           Normalized Vowel Diversity  2A Vowel Quality Inventories  \\\n",
       "WALS code                                                             \n",
       "abk                                 0                             0   \n",
       "aco                                 1                             1   \n",
       "aea                                 0                             0   \n",
       "aeg                                 1                             1   \n",
       "agh                                 2                             2   \n",
       "\n",
       "           Normalized Consonant Diversity  1A Consonant Inventories  \\\n",
       "WALS code                                                             \n",
       "abk                                     0                         4   \n",
       "aco                                     1                         4   \n",
       "aea                                     3                         2   \n",
       "aeg                                     2                         3   \n",
       "agh                                     3                         2   \n",
       "\n",
       "           Normalized Tone Diversity  13A Tone  \n",
       "WALS code                                       \n",
       "abk                                0         0  \n",
       "aco                                2         2  \n",
       "aea                                0         0  \n",
       "aeg                                0         0  \n",
       "agh                                1         1  "
      ]
     },
     "execution_count": 13,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.loc[mism.any(axis=1), COLS].head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "-4      1\n",
       "-3      7\n",
       "-2     10\n",
       "-1     21\n",
       " 0    418\n",
       " 1     33\n",
       " 2      9\n",
       " 3      4\n",
       " 4      1\n",
       "Name: count, dtype: int64"
      ]
     },
     "execution_count": 14,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "diff = df['Normalized Consonant Diversity'] - df['1A Consonant Inventories']\n",
    "\n",
    "diff.value_counts().sort_index()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1500x300 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "nf = df[diff != 0].assign(diff=diff)\n",
    "\n",
    "nf['diff'].sort_values().plot.bar(figsize=(15, 3));"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "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>r</th>\n",
       "      <th>p</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>diff &amp; Estimated Speaker Pop. Size</th>\n",
       "      <td>0.076365</td>\n",
       "      <td>0.484652</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>diff &amp; Distance from best fit origin</th>\n",
       "      <td>0.417043</td>\n",
       "      <td>0.000065</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                             r         p\n",
       "diff & Estimated Speaker Pop. Size    0.076365  0.484652\n",
       "diff & Distance from best fit origin  0.417043  0.000065"
      ]
     },
     "execution_count": 16,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "pd.DataFrame([pearsonr(nf, 'diff', x) for x in\n",
    "              ['Estimated Speaker Pop. Size', 'Distance from best fit origin']])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "nf.plot.scatter(x='Distance from best fit origin', y='diff');"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "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>r</th>\n",
       "      <th>p</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>Diversity &amp; Estimated Speaker Pop. Size</th>\n",
       "      <td>0.06618</td>\n",
       "      <td>0.1379</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Diversity &amp; Distance from best fit origin</th>\n",
       "      <td>-0.58884</td>\n",
       "      <td>0.0000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                                 r       p\n",
       "Diversity & Estimated Speaker Pop. Size    0.06618  0.1379\n",
       "Diversity & Distance from best fit origin -0.58884  0.0000"
      ]
     },
     "execution_count": 18,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "nf = df[FEAT + S1_OUTC + ['Total Normalized Phoneme Diversity']].copy()\n",
    "\n",
    "nf[FEAT] = nf[FEAT].apply(scipy.stats.zscore)\n",
    "\n",
    "nf['Diversity'] = nf[FEAT].mean(axis=1)\n",
    "\n",
    "pd.DataFrame([pearsonr(nf, 'Diversity', x) for x in\n",
    "              ['Estimated Speaker Pop. Size', 'Distance from best fit origin']]).round(5)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "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>r</th>\n",
       "      <th>p</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>Population &amp; Diversity</th>\n",
       "      <td>0.41379</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                              r    p\n",
       "Population & Diversity  0.41379  0.0"
      ]
     },
     "execution_count": 19,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "(pearsonr(nf.assign(Population=s1['Estimated Speaker Pop. Size'].apply(np.log10)),\n",
    "          'Population', 'Diversity')\n",
    " .to_frame().T.round(5))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "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>r</th>\n",
       "      <th>p</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>Distance &amp; Diversity</th>\n",
       "      <td>-0.58884</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                            r    p\n",
       "Distance & Diversity -0.58884  0.0"
      ]
     },
     "execution_count": 20,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "(pearsonr(nf.assign(Distance=s1['Distance from best fit origin']),\n",
    "          'Distance', 'Diversity')\n",
    " .to_frame().T.round(5))"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3 (ipykernel)",
   "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": "3.12.3"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}