Vaida_CW3.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Assignment 3"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "import numpy as np\n",
    "import re\n",
    "import seaborn as sns\n",
    "import matplotlib.pyplot as plt\n",
    "from wordcloud import WordCloud, STOPWORDS\n",
    "import csv\n",
    "%matplotlib inline"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "Number of text files: 195\n"
     ]
    }
   ],
   "source": [
    "# reading in index file\n",
    "\n",
    "inventory = pd.read_csv(\"encyclopaediaBritannica-inventory.csv\", header=None)\n",
    "inventory.columns = ['file','volume']\n",
    "print(\"\\nNumber of text files: \" + str(len(inventory)))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Selecting volumes for each edition"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "# ignore supplements & index volumes\n",
    "\n",
    "first_ed = inventory[:3]\n",
    "first_ed_reprint = inventory[3:7]\n",
    "second_ed = inventory[6:16]\n",
    "third_ed = inventory[16:34]\n",
    "third_ed_reprint = inventory[34:52]\n",
    "fourth_ed = inventory[56:96]\n",
    "fifth_ed = inventory[96:116]\n",
    "sixth_ed = inventory[116:136]\n",
    "seventh_ed = inventory[152:172]\n",
    "eighth_ed = inventory[173:194]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "['Encyclopaedia Britannica; or, A dictionary of arts and sciences, compiled upon a new plan … - First edition, 1771, Volume 1, A-B - EB.1',\n",
       " 'Encyclopaedia Britannica; or, A dictionary of arts and sciences, compiled upon a new plan … - First edition, 1771, Volume 2, C-L - EB.1',\n",
       " 'Encyclopaedia Britannica; or, A dictionary of arts and sciences, compiled upon a new plan … - First edition, 1771, Volume 3, M-Z - EB.1']"
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "list(first_ed['volume']) # list() to print whole text"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "# doesn't include reprints \n",
    "all_editions = [first_ed, second_ed, third_ed, fourth_ed, fifth_ed, sixth_ed, seventh_ed, eighth_ed]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "# for testing\n",
    "two_editions = [first_ed, second_ed]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Basic statistics"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Reading in text files and generate volume, length (in words and characters) counts\n",
    "\n",
    "def basic_data(editions):\n",
    "    \n",
    "    data = []\n",
    "    \n",
    "    for ed in editions: \n",
    "        length = 0\n",
    "        words = 0\n",
    "        volumes = 0\n",
    "\n",
    "        for index, row in ed.iterrows():\n",
    "            print(\"Reading: \" + row['file'])\n",
    "\n",
    "            f = open('text/' + row['file'], 'r', encoding=\"utf8\")\n",
    "            content = f.read()\n",
    "            content = clean_up(content)\n",
    "            length += len(content)\n",
    "            volumes += 1\n",
    "\n",
    "            words += len(content.split())\n",
    "\n",
    "            f.close()\n",
    "            \n",
    "        data.append([volumes, length, words])\n",
    "        data_df=pd.DataFrame(data,columns=['volumes', 'length_chars','length_words'])\n",
    "    return data_df"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 231,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Reading: 144133901.txt\n",
      "Reading: 144133902.txt\n",
      "Reading: 144133903.txt\n",
      "Reading: 144850370.txt\n",
      "Reading: 144850373.txt\n",
      "Reading: 144850374.txt\n",
      "Reading: 144850375.txt\n",
      "Reading: 144850376.txt\n",
      "Reading: 144850377.txt\n",
      "Reading: 144850378.txt\n",
      "Reading: 144850379.txt\n",
      "Reading: 190273289.txt\n",
      "Reading: 190273290.txt\n",
      "Reading: 190273291.txt\n",
      "Reading: 149977338.txt\n",
      "Reading: 149977873.txt\n",
      "Reading: 149978642.txt\n",
      "Reading: 149979156.txt\n",
      "Reading: 149979622.txt\n",
      "Reading: 149981189.txt\n",
      "Reading: 149981670.txt\n",
      "Reading: 149982181.txt\n",
      "Reading: 149982692.txt\n",
      "Reading: 149983206.txt\n",
      "Reading: 190273372.txt\n",
      "Reading: 191253798.txt\n",
      "Reading: 192200061.txt\n",
      "Reading: 191253799.txt\n",
      "Reading: 191319917.txt\n",
      "Reading: 191253800.txt\n",
      "Reading: 191253817.txt\n",
      "Reading: 191253818.txt\n",
      "Reading: 191253819.txt\n",
      "Reading: 192545783.txt\n",
      "Reading: 192545784.txt\n",
      "Reading: 191809865.txt\n",
      "Reading: 191809866.txt\n",
      "Reading: 191253825.txt\n",
      "Reading: 191253826.txt\n",
      "Reading: 191320556.txt\n",
      "Reading: 191320557.txt\n",
      "Reading: 192547776.txt\n",
      "Reading: 192547777.txt\n",
      "Reading: 192692191.txt\n",
      "Reading: 192692190.txt\n",
      "Reading: 191320560.txt\n",
      "Reading: 191320561.txt\n",
      "Reading: 193109110.txt\n",
      "Reading: 193109111.txt\n",
      "Reading: 192692192.txt\n",
      "Reading: 192692193.txt\n",
      "Reading: 191679019.txt\n",
      "Reading: 191679020.txt\n",
      "Reading: 192545785.txt\n",
      "Reading: 192545786.txt\n",
      "Reading: 191678902.txt\n",
      "Reading: 191678903.txt\n",
      "Reading: 191253838.txt\n",
      "Reading: 191253839.txt\n",
      "Reading: 191253836.txt\n",
      "Reading: 191253837.txt\n",
      "Reading: 191689060.txt\n",
      "Reading: 191689061.txt\n",
      "Reading: 191320559.txt\n",
      "Reading: 191320558.txt\n",
      "Reading: 192692194.txt\n",
      "Reading: 192692195.txt\n",
      "Reading: 193108316.txt\n",
      "Reading: 193108317.txt\n",
      "Reading: 191678900.txt\n",
      "Reading: 191678901.txt\n",
      "Reading: 192016229.txt\n",
      "Reading: 191689063.txt\n",
      "Reading: 193108318.txt\n",
      "Reading: 191679033.txt\n",
      "Reading: 192984254.txt\n",
      "Reading: 192984255.txt\n",
      "Reading: 191678897.txt\n",
      "Reading: 191678898.txt\n",
      "Reading: 193109108.txt\n",
      "Reading: 193109109.txt\n",
      "Reading: 192547785.txt\n",
      "Reading: 192200901.txt\n",
      "Reading: 192200898.txt\n",
      "Reading: 192200900.txt\n",
      "Reading: 192200897.txt\n",
      "Reading: 192692638.txt\n",
      "Reading: 191689062.txt\n",
      "Reading: 192984256.txt\n",
      "Reading: 192200899.txt\n",
      "Reading: 191678899.txt\n",
      "Reading: 192547781.txt\n",
      "Reading: 192015835.txt\n",
      "Reading: 193696080.txt\n",
      "Reading: 192200662.txt\n",
      "Reading: 192692756.txt\n",
      "Reading: 192200322.txt\n",
      "Reading: 192015838.txt\n",
      "Reading: 192015836.txt\n",
      "Reading: 192547782.txt\n",
      "Reading: 191679036.txt\n",
      "Reading: 192984257.txt\n",
      "Reading: 191689064.txt\n",
      "Reading: 192200323.txt\n",
      "Reading: 191689065.txt\n",
      "Reading: 192015837.txt\n",
      "Reading: 192547783.txt\n",
      "Reading: 192547784.txt\n",
      "Reading: 192015834.txt\n",
      "Reading: 193057497.txt\n",
      "Reading: 193696081.txt\n",
      "Reading: 192984259.txt\n",
      "Reading: 193057500.txt\n",
      "Reading: 193108322.txt\n",
      "Reading: 193696083.txt\n",
      "Reading: 193322690.txt\n",
      "Reading: 193819043.txt\n",
      "Reading: 193322688.txt\n",
      "Reading: 193696084.txt\n",
      "Reading: 193469090.txt\n",
      "Reading: 193638940.txt\n",
      "Reading: 192693199.txt\n",
      "Reading: 193108323.txt\n",
      "Reading: 193322689.txt\n",
      "Reading: 193819044.txt\n",
      "Reading: 194474782.txt\n",
      "Reading: 193469091.txt\n",
      "Reading: 193469092.txt\n",
      "Reading: 193057501.txt\n",
      "Reading: 193913444.txt\n",
      "Reading: 193819045.txt\n",
      "Reading: 192984260.txt\n",
      "Reading: 193322698.txt\n",
      "Reading: 193696085.txt\n",
      "Reading: 193696086.txt\n",
      "Reading: 193108324.txt\n",
      "Reading: 193109113.txt\n",
      "Reading: 193109114.txt\n",
      "Reading: 193108325.txt\n",
      "Reading: 193322700.txt\n",
      "Reading: 193109115.txt\n",
      "Reading: 193469392.txt\n",
      "Reading: 193696087.txt\n",
      "Reading: 193916150.txt\n",
      "Reading: 193696088.txt\n",
      "Reading: 193592632.txt\n",
      "Reading: 193322699.txt\n",
      "Reading: 193819046.txt\n",
      "Reading: 193108326.txt\n",
      "Reading: 193322701.txt\n",
      "Reading: 193469393.txt\n",
      "Reading: 193819047.txt\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>volumes</th>\n",
       "      <th>length_chars</th>\n",
       "      <th>length_words</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>3</td>\n",
       "      <td>14191040</td>\n",
       "      <td>2567260</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>10</td>\n",
       "      <td>56733498</td>\n",
       "      <td>10172846</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>18</td>\n",
       "      <td>96884019</td>\n",
       "      <td>17414509</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>40</td>\n",
       "      <td>103931985</td>\n",
       "      <td>18672098</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>20</td>\n",
       "      <td>103686840</td>\n",
       "      <td>18617826</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>20</td>\n",
       "      <td>103513075</td>\n",
       "      <td>18597305</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>20</td>\n",
       "      <td>119701529</td>\n",
       "      <td>21056156</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>21</td>\n",
       "      <td>131624089</td>\n",
       "      <td>22878318</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   volumes  length_chars  length_words\n",
       "0        3      14191040       2567260\n",
       "1       10      56733498      10172846\n",
       "2       18      96884019      17414509\n",
       "3       40     103931985      18672098\n",
       "4       20     103686840      18617826\n",
       "5       20     103513075      18597305\n",
       "6       20     119701529      21056156\n",
       "7       21     131624089      22878318"
      ]
     },
     "execution_count": 231,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "basic_df = basic_data(all_editions)\n",
    "basic_df"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 234,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "   Unnamed: 0  volumes  length_chars  length_words\n",
      "0           0        3      14191040       2567260\n",
      "1           1       10      56733498      10172846\n",
      "2           2       18      96884019      17414509\n",
      "3           3       40     103931985      18672098\n",
      "4           4       20     103686840      18617826\n",
      "5           5       20     103513075      18597305\n",
      "6           6       20     119701529      21056156\n",
      "7           7       21     131624089      22878318\n"
     ]
    }
   ],
   "source": [
    "# testing saving & loading from CSV\n",
    "\n",
    "basic_df.to_csv('csv/basic_data.csv')\n",
    "basic_reimport = pd.read_csv('csv/basic_data.csv')\n",
    "print(basic_reimport) # edition 4 is actually 20 volumes because text files separate its volumes into part 1 & 2"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Most referenced topics"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**Note:** the terms entries/headers/words are used interchangeably, they all refer to the `ENTRY` part of an `ENTRY, definition`. References are links to other parts of the Encyclopaedia in a definition, of the form `See TOPIC`."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Helper functions"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### For cleanup"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 57,
   "metadata": {},
   "outputs": [],
   "source": [
    "def replace_by(s, a, b):\n",
    "    \"\"\" perform a regex replacement, prints number of occurrences found and returns a string.\n",
    "    s: string to make changes in\n",
    "    a: string to remove \n",
    "    b: string to add\n",
    "    returns: a string\"\"\"\n",
    "    \n",
    "    # print(\"Replacing \\\"\" + a + \"\\\" by \\\"\" + b + \"\\\", found \" + str(len(re.findall(a, s))) + \"...\")\n",
    "    return re.sub(a, b, s)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 58,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Using regex cleanup ideas from: https://sites.temple.edu/tudsc/2014/08/12/text-scrubbing-hacks-cleaning-your-ocred-text/\n",
    "\n",
    "def clean_up(s):\n",
    "    \"\"\" Does a minimal cleanup of a string of text\n",
    "    returns: a string\n",
    "    \"\"\"\n",
    "    # print(\"Initial length: \" + str(len(s)))\n",
    "    s2 = replace_by(s, 'tbe', 'the')\n",
    "    s3 = replace_by(s2, 'tiie', 'the')\n",
    "    s4 = replace_by(s3, 'liis', 'his')\n",
    "    s5 = replace_by(s4, 'bis', 'his')\n",
    "    s6 = replace_by(s5, '■', '')\n",
    "    s7 = replace_by(s6, 'ib','in') # was noticed by Ava in .txt files\n",
    "    s8 = replace_by(s7, '¬','')\n",
    "    s9 = replace_by(s8, '.(\\.\\.+)', '') # multiple periods; what about ellipsis though?\n",
    "    # print(\"Clean up done!\")\n",
    "    return s9 # CAREFUL to always pass the right one to next, and return"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 59,
   "metadata": {},
   "outputs": [],
   "source": [
    "# This uses the list of header/definition/refs dataframes for each edition, and performs a basic cleanup\n",
    "\n",
    "def clean_up_definitions(editions):\n",
    "    editions_cleaned = []\n",
    "    for ed in editions:\n",
    "        # realistically, a definition can't be shorter than 5 letters (\"See x\" is itself 5 characters long)\n",
    "        ed = ed.loc[ed['def_length'] > 5]\n",
    "        # remove headers with two letters as they are likely noise\n",
    "        ed = ed.loc[ed['header_length'] > 2]\n",
    "        # remove headers longer than 40 characters, as these are likely noise\n",
    "        ed = ed.loc[ed['header_length'] < 40]\n",
    "        \n",
    "        # We might also want to drop header duplicates, as each entry should only be defined once\n",
    "        # A good portion of these duplicates are probably noise (such as annotations in roman numerals)\n",
    "        # We're not selecting the correct definition out of the duplicates here, just dropping them for the sake of simplicity\n",
    "        ed.drop_duplicates(subset =\"headers\", keep = False, inplace = True)\n",
    "        \n",
    "        editions_cleaned.append(ed)\n",
    "        \n",
    "    return editions_cleaned"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### For extracting words and their definitions, and references (\"See x\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 60,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Uses RegEx to identify the definition of a word as being the content between the uppercase word and the next uppercase word\n",
    "\n",
    "def find_definition(word, volume):\n",
    "    start = re.search(word, volume).start()\n",
    "    start_next = re.search(word, volume).end()\n",
    "    \n",
    "    second_word = re.search(\"[A-Z][A-Z]+\", volume[start_next:])\n",
    "    \n",
    "    if second_word: # checking that second capitalised word exists (prevents NoneType exception)\n",
    "        end = re.search(\"[A-Z][A-Z]+\", volume[start_next:]).start()\n",
    "        return volume[start_next+2:start_next+end] # +2 to ignore comma and space before a definition\n",
    "    else:\n",
    "        return \"\""
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 61,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Extracts entries (headers), their definitions, their respective lengths, and all references (\"See x\") from an edition\n",
    "\n",
    "def extract_info(edition):\n",
    "    \n",
    "    headers = []\n",
    "    definitions = []\n",
    "    refs = [] # for counting up all instances of \"See x\" in volumes  \n",
    "    \n",
    "    for index, row in edition.iterrows():\n",
    "        current_headers = []\n",
    "        \n",
    "        print(\"Reading: \" + row['file'])\n",
    "        f = open('text/' + row['file'], 'r', encoding=\"utf8\")\n",
    "        content = f.read()\n",
    "        content = clean_up(content)\n",
    "        \n",
    "        # using https://stackoverflow.com/questions/9525993/get-consecutive-capitalized-words-using-regex\n",
    "        # note: this ReGex is not entirely correct, as the first ReGex will select \"MARTIAL\" and \"LAW\" as separate entries\n",
    "        # (while the second one will add a correct entry for \"MARTIAL LAW\")\n",
    "        # \"MARTIAL\" will be cleaned out later due to having an empty definition, but \"LAW\" will not, leading to duplicates\n",
    "        \n",
    "        current_headers = re.findall('[A-Z][A-Z]+', content) # at least two uppercase letters following each other\n",
    "        current_headers += re.findall('([A-Z][A-Z]+(?=\\s[A-Z])(?:\\s[A-Z][A-Z]+)+)', content) # two or more uppercase words\n",
    "        refs += re.findall('See [^,\\.]*', content) # matches \"See x\" until a comma or a period\n",
    "        \n",
    "        for word in current_headers:\n",
    "            definitions.append(find_definition(word, content))\n",
    "            \n",
    "        headers = headers + current_headers\n",
    "        \n",
    "        f.close()\n",
    "\n",
    "    data = pd.DataFrame(headers, columns =['headers'])\n",
    "    data['definition'] = definitions\n",
    "    data['header_length']  = data['headers'].str.len()\n",
    "    data['def_length']  = data['definition'].str.len()\n",
    "    return data, refs"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 62,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Generates a list of dataframes with words and their definitions for each edition\n",
    "\n",
    "def words_and_definitions(editions):\n",
    "    \n",
    "    all_edition_data = []\n",
    "    all_references = []\n",
    "    \n",
    "    for ed in editions:\n",
    "        edition_data, references = extract_info(ed)\n",
    "        all_edition_data.append(edition_data)\n",
    "        all_references.append(references)\n",
    "    return all_edition_data, all_references"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Generating headers, definitions and references for all editions and saving to CSV"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Reading: 192984259.txt\n",
      "Reading: 193057500.txt\n",
      "Reading: 193108322.txt\n",
      "Reading: 193696083.txt\n",
      "Reading: 193322690.txt\n",
      "Reading: 193819043.txt\n",
      "Reading: 193322688.txt\n",
      "Reading: 193696084.txt\n",
      "Reading: 193469090.txt\n",
      "Reading: 193638940.txt\n",
      "Reading: 192693199.txt\n",
      "Reading: 193108323.txt\n",
      "Reading: 193322689.txt\n",
      "Reading: 193819044.txt\n",
      "Reading: 194474782.txt\n",
      "Reading: 193469091.txt\n",
      "Reading: 193469092.txt\n",
      "Reading: 193057501.txt\n",
      "Reading: 193913444.txt\n",
      "Reading: 193819045.txt\n",
      "Reading: 192984260.txt\n",
      "Reading: 193322698.txt\n",
      "Reading: 193696085.txt\n",
      "Reading: 193696086.txt\n",
      "Reading: 193108324.txt\n",
      "Reading: 193109113.txt\n",
      "Reading: 193109114.txt\n",
      "Reading: 193108325.txt\n",
      "Reading: 193322700.txt\n",
      "Reading: 193109115.txt\n",
      "Reading: 193469392.txt\n",
      "Reading: 193696087.txt\n",
      "Reading: 193916150.txt\n",
      "Reading: 193696088.txt\n",
      "Reading: 193592632.txt\n",
      "Reading: 193322699.txt\n",
      "Reading: 193819046.txt\n",
      "Reading: 193108326.txt\n",
      "Reading: 193322701.txt\n",
      "Reading: 193469393.txt\n",
      "Reading: 193819047.txt\n"
     ]
    }
   ],
   "source": [
    "# function takes in a list of editions and extracts the relevant data\n",
    "# here, we processed two editions at a time to perform the processing in chunks (it takes around a minute for each txt file)\n",
    "seventh_eighth_data, seventh_eighth_refs = words_and_definitions([seventh_ed, eighth_ed])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "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>headers</th>\n",
       "      <th>definition</th>\n",
       "      <th>header_length</th>\n",
       "      <th>def_length</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>55</th>\n",
       "      <td>AAHUS</td>\n",
       "      <td>a little town of Germany, in the circle of\\nWe...</td>\n",
       "      <td>5</td>\n",
       "      <td>210</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>56</th>\n",
       "      <td>AALBORG</td>\n",
       "      <td>one of the four sees (stiffs) into which\\nthe ...</td>\n",
       "      <td>7</td>\n",
       "      <td>2275</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>58</th>\n",
       "      <td>AALEN</td>\n",
       "      <td>a bailiwick in the circle of Jaxt, in the king...</td>\n",
       "      <td>5</td>\n",
       "      <td>773</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>59</th>\n",
       "      <td>AALSMEER</td>\n",
       "      <td>a town in the arrondissement of Am\\nsterdam, i...</td>\n",
       "      <td>8</td>\n",
       "      <td>205</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>60</th>\n",
       "      <td>AALTEN</td>\n",
       "      <td>a town in the arrondissement of Zutphen,\\nand ...</td>\n",
       "      <td>6</td>\n",
       "      <td>119</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>123231</th>\n",
       "      <td>PLATE CCCCXCV</td>\n",
       "      <td>Fin. 17.\\n</td>\n",
       "      <td>13</td>\n",
       "      <td>9</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>123232</th>\n",
       "      <td>PLATE CCCCXCVH</td>\n",
       "      <td>4\\nn „ - .\\nI .\\nh\\n———\\n©</td>\n",
       "      <td>14</td>\n",
       "      <td>21</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>123233</th>\n",
       "      <td>OTMISIEM IPAIM\\nJPJLATE</td>\n",
       "      <td>cccxcix.\\nSf Bloomsbrn^\\n</td>\n",
       "      <td>22</td>\n",
       "      <td>23</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>123234</th>\n",
       "      <td>ZOOPHYTES\\nPLATE</td>\n",
       "      <td>)11.\\nAsatrLs hunbricoicUis.\\nTetjurhynckus Un...</td>\n",
       "      <td>15</td>\n",
       "      <td>216</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>123237</th>\n",
       "      <td>PLATE DIV</td>\n",
       "      <td>Holothuna papiUosa.\\nPhpscdus pelcujica.\\nHolo...</td>\n",
       "      <td>9</td>\n",
       "      <td>82</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>24072 rows × 4 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "                        headers  \\\n",
       "55                        AAHUS   \n",
       "56                      AALBORG   \n",
       "58                        AALEN   \n",
       "59                     AALSMEER   \n",
       "60                       AALTEN   \n",
       "...                         ...   \n",
       "123231            PLATE CCCCXCV   \n",
       "123232           PLATE CCCCXCVH   \n",
       "123233  OTMISIEM IPAIM\\nJPJLATE   \n",
       "123234         ZOOPHYTES\\nPLATE   \n",
       "123237                PLATE DIV   \n",
       "\n",
       "                                               definition  header_length  \\\n",
       "55      a little town of Germany, in the circle of\\nWe...              5   \n",
       "56      one of the four sees (stiffs) into which\\nthe ...              7   \n",
       "58      a bailiwick in the circle of Jaxt, in the king...              5   \n",
       "59      a town in the arrondissement of Am\\nsterdam, i...              8   \n",
       "60      a town in the arrondissement of Zutphen,\\nand ...              6   \n",
       "...                                                   ...            ...   \n",
       "123231                                         Fin. 17.\\n             13   \n",
       "123232                         4\\nn „ - .\\nI .\\nh\\n———\\n©             14   \n",
       "123233                          cccxcix.\\nSf Bloomsbrn^\\n             22   \n",
       "123234  )11.\\nAsatrLs hunbricoicUis.\\nTetjurhynckus Un...             15   \n",
       "123237  Holothuna papiUosa.\\nPhpscdus pelcujica.\\nHolo...              9   \n",
       "\n",
       "        def_length  \n",
       "55             210  \n",
       "56            2275  \n",
       "58             773  \n",
       "59             205  \n",
       "60             119  \n",
       "...            ...  \n",
       "123231           9  \n",
       "123232          21  \n",
       "123233          23  \n",
       "123234         216  \n",
       "123237          82  \n",
       "\n",
       "[24072 rows x 4 columns]"
      ]
     },
     "execution_count": 15,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "seventh_eighth_cleaned = clean_up_definitions(seventh_eighth_data)\n",
    "seventh_eighth_cleaned[0]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [],
   "source": [
    "# save the two dataframes to CSV\n",
    "\n",
    "seventh_eighth_cleaned[0].to_csv('csv/seventh_ed_data.csv')\n",
    "seventh_eighth_cleaned[1].to_csv('csv/eighth_ed_data.csv')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 63,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Generates a cleaned dataframe with counts of references, from the list of all references picked up in each edition\n",
    "\n",
    "def generate_ref_counts(refs):\n",
    "    \n",
    "    all_ref_counts = []\n",
    "    \n",
    "    for ref in refs:\n",
    "        # Converting unique counts to a DataFrame for easy visualisation\n",
    "        ref_df = pd.DataFrame(ref)\n",
    "        ref_counts = ref_df[0].value_counts()\n",
    "        ref_counts = pd.DataFrame(ref_counts)\n",
    "        ref_counts['references'] = ref_counts.index\n",
    "        ref_counts.columns = ['count', 'references']\n",
    "#         ref_counts = ref_counts.reset_index(True)\n",
    "\n",
    "        # cleaning ref counts (removing non-topics such as \"See the next article\")\n",
    "        ref_counts = ref_counts.loc[ref_counts['references'] != \"See the next article\"]\n",
    "        ref_counts = ref_counts.loc[ref_counts['references'] != \"See there\"]\n",
    "        ref_counts = ref_counts.loc[ref_counts['references'] != \"See thefe articles\"]\n",
    "        ref_counts = ref_counts.loc[ref_counts['references'] != \"See these articles\"]\n",
    "        ref_counts = ref_counts.loc[ref_counts['references'] != \"See Vol\"]\n",
    "        ref_counts = ref_counts.loc[ref_counts['references'] != \"See p\"]\n",
    "        ref_counts = ref_counts.loc[ref_counts['references'] != \"See fig\"]\n",
    "        ref_counts = ref_counts.loc[ref_counts['references'] != \"See that article\"]\n",
    "        ref_counts = ref_counts.loc[ref_counts['references'] != \"See the preceding article\"]\n",
    "        ref_counts = ref_counts.loc[ref_counts['references'] != \"See Supplement\"]\n",
    "        ref_counts = ref_counts.loc[ref_counts['references'] != \"See SUPPLEMENT\"]\n",
    "        ref_counts = ref_counts.loc[~ref_counts['references'].str.contains(\"Plate\")]\n",
    "        ref_counts = ref_counts.loc[~ref_counts['references'].str.contains(\"plate\")]\n",
    "        ref_counts = ref_counts.loc[~ref_counts['references'].str.contains(\"Index\")]\n",
    "        ref_counts = ref_counts.loc[~ref_counts['references'].str.contains(\"In\\ndex\")]\n",
    "        ref_counts = ref_counts.loc[~ref_counts['references'].str.contains(\"index\")]\n",
    "        \n",
    "        # Change formatting to not have \"See\" in topic name\n",
    "        ref_counts['references'] = ref_counts['references'].str.replace(r'See ', '')\n",
    "        \n",
    "        all_ref_counts.append(ref_counts)\n",
    "    \n",
    "        \n",
    "    \n",
    "    return all_ref_counts"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {
    "scrolled": true
   },
   "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>count</th>\n",
       "      <th>references</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>See Ornithology</th>\n",
       "      <td>43</td>\n",
       "      <td>Ornithology</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>See Mammalia</th>\n",
       "      <td>35</td>\n",
       "      <td>Mammalia</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>See Music</th>\n",
       "      <td>31</td>\n",
       "      <td>Music</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>See Metaphysics</th>\n",
       "      <td>27</td>\n",
       "      <td>Metaphysics</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>See Anatomy</th>\n",
       "      <td>26</td>\n",
       "      <td>Anatomy</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>See Chemist^\\nZINZENDORF</th>\n",
       "      <td>1</td>\n",
       "      <td>Chemist^\\nZINZENDORF</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>See the article Arabia</th>\n",
       "      <td>1</td>\n",
       "      <td>the article Arabia</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>See Hales’\\nChronology</th>\n",
       "      <td>1</td>\n",
       "      <td>Hales’\\nChronology</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>See the I ransia\\ntion of Bayle’s Dictionary</th>\n",
       "      <td>1</td>\n",
       "      <td>the I ransia\\ntion of Bayle’s Dictionary</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>See Saxii Onomasticon ; Biographic Universelle;\\nGesner’s Isagoge</th>\n",
       "      <td>1</td>\n",
       "      <td>Saxii Onomasticon ; Biographic Universelle;\\nG...</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>2657 rows × 2 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "                                                    count  \\\n",
       "See Ornithology                                        43   \n",
       "See Mammalia                                           35   \n",
       "See Music                                              31   \n",
       "See Metaphysics                                        27   \n",
       "See Anatomy                                            26   \n",
       "...                                                   ...   \n",
       "See Chemist^\\nZINZENDORF                                1   \n",
       "See the article Arabia                                  1   \n",
       "See Hales’\\nChronology                                  1   \n",
       "See the I ransia\\ntion of Bayle’s Dictionary            1   \n",
       "See Saxii Onomasticon ; Biographic Universelle;...      1   \n",
       "\n",
       "                                                                                           references  \n",
       "See Ornithology                                                                           Ornithology  \n",
       "See Mammalia                                                                                 Mammalia  \n",
       "See Music                                                                                       Music  \n",
       "See Metaphysics                                                                           Metaphysics  \n",
       "See Anatomy                                                                                   Anatomy  \n",
       "...                                                                                               ...  \n",
       "See Chemist^\\nZINZENDORF                                                         Chemist^\\nZINZENDORF  \n",
       "See the article Arabia                                                             the article Arabia  \n",
       "See Hales’\\nChronology                                                             Hales’\\nChronology  \n",
       "See the I ransia\\ntion of Bayle’s Dictionary                 the I ransia\\ntion of Bayle’s Dictionary  \n",
       "See Saxii Onomasticon ; Biographic Universelle;...  Saxii Onomasticon ; Biographic Universelle;\\nG...  \n",
       "\n",
       "[2657 rows x 2 columns]"
      ]
     },
     "execution_count": 24,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "seventh_eighth_refs_counts = generate_ref_counts(seventh_eighth_refs)\n",
    "seventh_eighth_refs_counts[0]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {},
   "outputs": [],
   "source": [
    "seventh_eighth_refs_counts[0].to_csv('csv/seventh_ed_refs.csv')\n",
    "seventh_eighth_refs_counts[1].to_csv('csv/eighth_ed_refs.csv')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Example of manual cleaning"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# seventh edition had \"the article Optics\" references in addition to Optics\n",
    "# this code checks the counts, manually updates the ones for Optics, and removes \"the article Optics\" entry"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 55,
   "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>Unnamed: 0</th>\n",
       "      <th>count</th>\n",
       "      <th>references</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>17</th>\n",
       "      <td>See Optics</td>\n",
       "      <td>23</td>\n",
       "      <td>Optics</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "    Unnamed: 0  count references\n",
       "17  See Optics     23     Optics"
      ]
     },
     "execution_count": 55,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# check counts\n",
    "# seventh_eighth_refs_counts[0].loc[seventh_eighth_refs_counts[0].references == \"the article Optics\"]\n",
    "seventh_eighth_refs_counts[0].loc[seventh_eighth_refs_counts[0].references == \"Optics\"]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 54,
   "metadata": {},
   "outputs": [],
   "source": [
    "# add counts\n",
    "seventh_eighth_refs_counts[0].loc[seventh_eighth_refs_counts[0].references == \"Optics\", \"count\"] = 11 + 12"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 56,
   "metadata": {},
   "outputs": [],
   "source": [
    "# remove \"the article Optics\" entry\n",
    "seventh_eighth_refs_counts[0] = seventh_eighth_refs_counts[0].loc[seventh_eighth_refs_counts[0].references != \"the article Optics\"]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 62,
   "metadata": {},
   "outputs": [],
   "source": [
    "# re-sort the dataframe, as updated entry might be at a different position in popularity\n",
    "seventh_eighth_refs_counts[0] = seventh_eighth_refs_counts[0].sort_values(by=['count'], ascending=False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 58,
   "metadata": {},
   "outputs": [],
   "source": [
    "# save to CSV\n",
    "seventh_eighth_refs_counts[0].to_csv('csv/seventh_ed_refs_fixed.csv')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<seaborn.axisgrid.FacetGrid at 0x1a65f490>"
      ]
     },
     "execution_count": 32,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1440x720 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# visualise top 20\n",
    "\n",
    "sns.set(font_scale=1.5)\n",
    "sns.set_style(\"whitegrid\")\n",
    "ax = sns.catplot(x='count', y='references', data=seventh_eighth_refs_counts[0].head(20), kind='bar', height=10, aspect=2, color=\"#cfbd93\")\n",
    "ax.fig.suptitle(\"Top 20 most referenced topics in seventh edition\", y=1)\n",
    "ax.set_axis_labels('counts of references', 'topic')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Overall number of references in each edition"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "8"
      ]
     },
     "execution_count": 28,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# load in references from CSV\n",
    "\n",
    "references = []\n",
    "references.append(pd.read_csv('csv/first_ed_refs.csv'))\n",
    "references.append(pd.read_csv('csv/second_ed_refs.csv'))\n",
    "references.append(pd.read_csv('csv/third_ed_refs.csv'))\n",
    "references.append(pd.read_csv('csv/fourth_ed_refs.csv'))\n",
    "references.append(pd.read_csv('csv/fifth_ed_refs.csv'))\n",
    "references.append(pd.read_csv('csv/sixth_ed_refs.csv'))\n",
    "references.append(pd.read_csv('csv/seventh_ed_refs_fixed_2.csv'))\n",
    "references.append(pd.read_csv('csv/eighth_ed_refs.csv'))\n",
    "len(references)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[4909, 7020, 9363, 8811, 8687, 8895, 3489, 6032]"
      ]
     },
     "execution_count": 29,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# calculate sum of references for each edition\n",
    "\n",
    "sum_refs = []\n",
    "for refs in references:\n",
    "    sum_refs.append(refs['count'].sum())\n",
    "sum_refs"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Text(0, 0.5, 'counts of references')"
      ]
     },
     "execution_count": 30,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# visualise sums\n",
    "\n",
    "# dip might be due to a change in format?\n",
    "# more longform articles, with content within, rather than entries with references to main topic?\n",
    "\n",
    "ed_count = list(range(1, 9))\n",
    "sns.set_style(\"whitegrid\")\n",
    "ax = sns.barplot(x= ed_count, y=sum_refs, color=\"#cfbd93\")\n",
    "ax.set_title('Total number of references in each edition', fontsize=16)\n",
    "ax.set_xlabel('editions', fontsize=14)\n",
    "ax.set_ylabel('counts of references', fontsize=14)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### One of the longest words in each edition"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "8"
      ]
     },
     "execution_count": 33,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# load in headers for each edition\n",
    "\n",
    "headers = []\n",
    "headers.append(pd.read_csv('csv/first_ed_data.csv'))\n",
    "headers.append(pd.read_csv('csv/second_ed_data.csv'))\n",
    "headers.append(pd.read_csv('csv/third_ed_data.csv'))\n",
    "headers.append(pd.read_csv('csv/fourth_ed_data.csv'))\n",
    "headers.append(pd.read_csv('csv/fifth_ed_data.csv'))\n",
    "headers.append(pd.read_csv('csv/sixth_ed_data.csv'))\n",
    "headers.append(pd.read_csv('csv/seventh_ed_data.csv'))\n",
    "headers.append(pd.read_csv('csv/eighth_ed_data.csv'))\n",
    "len(headers)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "metadata": {},
   "outputs": [],
   "source": [
    "# this is mostly manual inspection, as the longest headers are often errors\n",
    "# selected not necessarily the longest valid word, but one of the longest ones that has a more readable/identifiable description"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 84,
   "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>Unnamed: 0</th>\n",
       "      <th>headers</th>\n",
       "      <th>definition</th>\n",
       "      <th>header_length</th>\n",
       "      <th>def_length</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>4419</th>\n",
       "      <td>7779</td>\n",
       "      <td>EXTENSOR MINIMI DIGITI PROPRIUS</td>\n",
       "      <td>This is a kind of collateral or auxiliary mufc...</td>\n",
       "      <td>31</td>\n",
       "      <td>383</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4483</th>\n",
       "      <td>7861</td>\n",
       "      <td>THE PARTS OF GENERATION IN\\nMALES</td>\n",
       "      <td>The fpermatic arteries go out moft commonly fr...</td>\n",
       "      <td>32</td>\n",
       "      <td>4066</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>15803</th>\n",
       "      <td>29601</td>\n",
       "      <td>ENCYCLOPAEDIA BRITANNICA\\nVOLUME</td>\n",
       "      <td>he T H I R D,\\nEncyclopaedia Britannica;\\nO R,...</td>\n",
       "      <td>31</td>\n",
       "      <td>47</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "       Unnamed: 0                            headers  \\\n",
       "4419         7779    EXTENSOR MINIMI DIGITI PROPRIUS   \n",
       "4483         7861  THE PARTS OF GENERATION IN\\nMALES   \n",
       "15803       29601   ENCYCLOPAEDIA BRITANNICA\\nVOLUME   \n",
       "\n",
       "                                              definition  header_length  \\\n",
       "4419   This is a kind of collateral or auxiliary mufc...             31   \n",
       "4483   The fpermatic arteries go out moft commonly fr...             32   \n",
       "15803  he T H I R D,\\nEncyclopaedia Britannica;\\nO R,...             31   \n",
       "\n",
       "       def_length  \n",
       "4419          383  \n",
       "4483         4066  \n",
       "15803          47  "
      ]
     },
     "execution_count": 84,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "headers[0].loc[headers[0]['header_length'] > 30]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 85,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "'This is a kind of collateral or auxiliary mufcle of the extenfor communis, of which it appears almoit al ways to be more or lefs a portion. It is fixed along the fuperior external half of the ul na, from whence its long fmall tendon runs down in company with the fourth tendon of the extenfor commu nis, all the way to the little finger, where it joins it, and is infer ted with it. '"
      ]
     },
     "execution_count": 85,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "re.sub(\"\\n\", \" \", headers[0].iloc[4419].definition) # remove \\n for simpler formatting"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Words that reference a specific topic"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [],
   "source": [
    "# again, requires manual inspection of what definitions are not just references to the main article\n",
    "# as well as having concise and fairly clean, interesting defs"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "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>Unnamed: 0</th>\n",
       "      <th>headers</th>\n",
       "      <th>definition</th>\n",
       "      <th>header_length</th>\n",
       "      <th>def_length</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>10870</th>\n",
       "      <td>20411</td>\n",
       "      <td>TALS</td>\n",
       "      <td>in natural biftory, are defined to be foffile\\...</td>\n",
       "      <td>4</td>\n",
       "      <td>180</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10967</th>\n",
       "      <td>20831</td>\n",
       "      <td>MINIUM</td>\n",
       "      <td>or red-lead. See Chemistry, p. 84.\\nand 136.\\n</td>\n",
       "      <td>6</td>\n",
       "      <td>44</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10998</th>\n",
       "      <td>20871</td>\n",
       "      <td>MIXTURE</td>\n",
       "      <td>a compound, or affemblage of feveral dif\\nfere...</td>\n",
       "      <td>7</td>\n",
       "      <td>88</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11149</th>\n",
       "      <td>21172</td>\n",
       "      <td>MUFFLE</td>\n",
       "      <td>in Chemittry. See Chemistry, p. 109.\\n</td>\n",
       "      <td>6</td>\n",
       "      <td>37</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11462</th>\n",
       "      <td>22446</td>\n",
       "      <td>NOCTILUCA</td>\n",
       "      <td>a fpecies of phofphorus, fo called be-\\ncaufe ...</td>\n",
       "      <td>9</td>\n",
       "      <td>170</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12501</th>\n",
       "      <td>24221</td>\n",
       "      <td>PLUMBUM</td>\n",
       "      <td>lead See Chemistry, p. 84.\\nPlume, in botany S...</td>\n",
       "      <td>7</td>\n",
       "      <td>68</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12657</th>\n",
       "      <td>24520</td>\n",
       "      <td>PRECIPITANT</td>\n",
       "      <td>in chemiftry, is applied to any liquor,\\nwhich...</td>\n",
       "      <td>11</td>\n",
       "      <td>364</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12895</th>\n",
       "      <td>24816</td>\n",
       "      <td>PYROTECHNY</td>\n",
       "      <td>the art of fire, or afcience which teaches\\nth...</td>\n",
       "      <td>10</td>\n",
       "      <td>121</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12928</th>\n",
       "      <td>24862</td>\n",
       "      <td>QUARTATION</td>\n",
       "      <td>a method of purifying gold by melting\\nthree p...</td>\n",
       "      <td>10</td>\n",
       "      <td>149</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12944</th>\n",
       "      <td>24881</td>\n",
       "      <td>QUINQUATRIA</td>\n",
       "      <td>in Roman antiquity, feftivals celebra-\\nyoung ...</td>\n",
       "      <td>11</td>\n",
       "      <td>249</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>13103</th>\n",
       "      <td>25096</td>\n",
       "      <td>REFINING</td>\n",
       "      <td>in general, is the art of purifying a thing;\\n...</td>\n",
       "      <td>8</td>\n",
       "      <td>207</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>13201</th>\n",
       "      <td>25336</td>\n",
       "      <td>RESOLUTION</td>\n",
       "      <td>in chemiHry, &amp;c. the reduftion of a\\nmixed bod...</td>\n",
       "      <td>10</td>\n",
       "      <td>448</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>13236</th>\n",
       "      <td>25381</td>\n",
       "      <td>RETORT</td>\n",
       "      <td>in cbemiftry. See Chemistry, p. 110.\\n</td>\n",
       "      <td>6</td>\n",
       "      <td>37</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>13583</th>\n",
       "      <td>25840</td>\n",
       "      <td>SATURN</td>\n",
       "      <td>in aftronomy. See Astronomy, p. 442.\\nSaturn,\\...</td>\n",
       "      <td>6</td>\n",
       "      <td>247</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>14040</th>\n",
       "      <td>26672</td>\n",
       "      <td>SMALT</td>\n",
       "      <td>a preparation of arfenic. See Chemistry, p;\\n1...</td>\n",
       "      <td>5</td>\n",
       "      <td>49</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>14064</th>\n",
       "      <td>26726</td>\n",
       "      <td>SOAP</td>\n",
       "      <td>a kind of pafte, fometimes hard and dry, and\\n...</td>\n",
       "      <td>4</td>\n",
       "      <td>1139</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>14107</th>\n",
       "      <td>26780</td>\n",
       "      <td>SOLUTION</td>\n",
       "      <td>in chemiHry, &amp;c. the reduftion of a\\nmixed bod...</td>\n",
       "      <td>8</td>\n",
       "      <td>448</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>14354</th>\n",
       "      <td>27089</td>\n",
       "      <td>STIBIUM</td>\n",
       "      <td>or Antimony. See Chemistry, p. 87, 139.\\n</td>\n",
       "      <td>7</td>\n",
       "      <td>40</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>15361</th>\n",
       "      <td>29001</td>\n",
       "      <td>VITRIOL</td>\n",
       "      <td>a compound body formed of the particles of\\nme...</td>\n",
       "      <td>7</td>\n",
       "      <td>313</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>15426</th>\n",
       "      <td>29080</td>\n",
       "      <td>VOLATILE</td>\n",
       "      <td>in phyfics, is commonly ufed to denotea\\nmixed...</td>\n",
       "      <td>8</td>\n",
       "      <td>265</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "       Unnamed: 0      headers  \\\n",
       "10870       20411         TALS   \n",
       "10967       20831       MINIUM   \n",
       "10998       20871      MIXTURE   \n",
       "11149       21172       MUFFLE   \n",
       "11462       22446    NOCTILUCA   \n",
       "12501       24221      PLUMBUM   \n",
       "12657       24520  PRECIPITANT   \n",
       "12895       24816   PYROTECHNY   \n",
       "12928       24862   QUARTATION   \n",
       "12944       24881  QUINQUATRIA   \n",
       "13103       25096     REFINING   \n",
       "13201       25336   RESOLUTION   \n",
       "13236       25381       RETORT   \n",
       "13583       25840       SATURN   \n",
       "14040       26672        SMALT   \n",
       "14064       26726         SOAP   \n",
       "14107       26780     SOLUTION   \n",
       "14354       27089      STIBIUM   \n",
       "15361       29001      VITRIOL   \n",
       "15426       29080     VOLATILE   \n",
       "\n",
       "                                              definition  header_length  \\\n",
       "10870  in natural biftory, are defined to be foffile\\...              4   \n",
       "10967     or red-lead. See Chemistry, p. 84.\\nand 136.\\n              6   \n",
       "10998  a compound, or affemblage of feveral dif\\nfere...              7   \n",
       "11149             in Chemittry. See Chemistry, p. 109.\\n              6   \n",
       "11462  a fpecies of phofphorus, fo called be-\\ncaufe ...              9   \n",
       "12501  lead See Chemistry, p. 84.\\nPlume, in botany S...              7   \n",
       "12657  in chemiftry, is applied to any liquor,\\nwhich...             11   \n",
       "12895  the art of fire, or afcience which teaches\\nth...             10   \n",
       "12928  a method of purifying gold by melting\\nthree p...             10   \n",
       "12944  in Roman antiquity, feftivals celebra-\\nyoung ...             11   \n",
       "13103  in general, is the art of purifying a thing;\\n...              8   \n",
       "13201  in chemiHry, &c. the reduftion of a\\nmixed bod...             10   \n",
       "13236             in cbemiftry. See Chemistry, p. 110.\\n              6   \n",
       "13583  in aftronomy. See Astronomy, p. 442.\\nSaturn,\\...              6   \n",
       "14040  a preparation of arfenic. See Chemistry, p;\\n1...              5   \n",
       "14064  a kind of pafte, fometimes hard and dry, and\\n...              4   \n",
       "14107  in chemiHry, &c. the reduftion of a\\nmixed bod...              8   \n",
       "14354          or Antimony. See Chemistry, p. 87, 139.\\n              7   \n",
       "15361  a compound body formed of the particles of\\nme...              7   \n",
       "15426  in phyfics, is commonly ufed to denotea\\nmixed...              8   \n",
       "\n",
       "       def_length  \n",
       "10870         180  \n",
       "10967          44  \n",
       "10998          88  \n",
       "11149          37  \n",
       "11462         170  \n",
       "12501          68  \n",
       "12657         364  \n",
       "12895         121  \n",
       "12928         149  \n",
       "12944         249  \n",
       "13103         207  \n",
       "13201         448  \n",
       "13236          37  \n",
       "13583         247  \n",
       "14040          49  \n",
       "14064        1139  \n",
       "14107         448  \n",
       "14354          40  \n",
       "15361         313  \n",
       "15426         265  "
      ]
     },
     "execution_count": 21,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "headers[0].loc[(headers[0]['definition'].str.contains(\"See Chemistry\")) & (headers[0]['def_length'] > 35)].iloc[50:100]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "'a compound, or affemblage of feveral dif ferent bodies in the fame mafs. See Chemistry. '"
      ]
     },
     "execution_count": 27,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "re.sub(\"\\n\", \" \", headers[0].iloc[10998].definition)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Words with most references to other topics"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {},
   "outputs": [],
   "source": [
    "headers_with_refs = headers.copy()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {},
   "outputs": [],
   "source": [
    "# find all references of the form of \"See x\" in each definition\n",
    "\n",
    "for ed in range(0, len(headers_with_refs)):\n",
    "    refs_in_desc = []\n",
    "    for index, row in headers_with_refs[ed].iterrows():   \n",
    "        refs_in_desc.append(len(re.findall('See [^,\\.]*', row['definition']))) # matches \"See x\" until a comma or a period\n",
    "    headers_with_refs[ed] = headers_with_refs[ed].assign(refs_in_desc=refs_in_desc)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "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>Unnamed: 0</th>\n",
       "      <th>header_length</th>\n",
       "      <th>def_length</th>\n",
       "      <th>refs_in_desc</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>count</th>\n",
       "      <td>26070.000000</td>\n",
       "      <td>26070.000000</td>\n",
       "      <td>26070.000000</td>\n",
       "      <td>26070.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>mean</th>\n",
       "      <td>55320.488953</td>\n",
       "      <td>7.520483</td>\n",
       "      <td>725.914423</td>\n",
       "      <td>0.242079</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>std</th>\n",
       "      <td>36662.148468</td>\n",
       "      <td>3.147801</td>\n",
       "      <td>1557.828398</td>\n",
       "      <td>0.486152</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>min</th>\n",
       "      <td>5.000000</td>\n",
       "      <td>3.000000</td>\n",
       "      <td>6.000000</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25%</th>\n",
       "      <td>21171.500000</td>\n",
       "      <td>5.000000</td>\n",
       "      <td>96.250000</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>50%</th>\n",
       "      <td>54119.500000</td>\n",
       "      <td>7.000000</td>\n",
       "      <td>246.000000</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>75%</th>\n",
       "      <td>87723.750000</td>\n",
       "      <td>9.000000</td>\n",
       "      <td>698.000000</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>max</th>\n",
       "      <td>122922.000000</td>\n",
       "      <td>39.000000</td>\n",
       "      <td>33766.000000</td>\n",
       "      <td>7.000000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "          Unnamed: 0  header_length    def_length  refs_in_desc\n",
       "count   26070.000000   26070.000000  26070.000000  26070.000000\n",
       "mean    55320.488953       7.520483    725.914423      0.242079\n",
       "std     36662.148468       3.147801   1557.828398      0.486152\n",
       "min         5.000000       3.000000      6.000000      0.000000\n",
       "25%     21171.500000       5.000000     96.250000      0.000000\n",
       "50%     54119.500000       7.000000    246.000000      0.000000\n",
       "75%     87723.750000       9.000000    698.000000      0.000000\n",
       "max    122922.000000      39.000000  33766.000000      7.000000"
      ]
     },
     "execution_count": 40,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "headers_with_refs[3].describe()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 238,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Again, the checking is manual, as some of the headers/definitions are noise that is picked up \n",
    "# (eg. a \"CCCCLXXXVIL\"-type header)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 44,
   "metadata": {},
   "outputs": [],
   "source": [
    "# prints out the words with most references in a specific edition, with a count of those words and the number of refs \n",
    "# if there are several, word parameter can be used to print specific ones\n",
    "# adjuster is for easily looking at words with less references, if most referenced words are noise \n",
    "\n",
    "def most_refs_word(ed, word, adjuster):\n",
    "    most_refs = np.max(headers_with_refs[ed]['refs_in_desc']) - adjuster \n",
    "    longest = headers_with_refs[ed].loc[headers_with_refs[ed]['refs_in_desc'] == (most_refs)]\n",
    "    print(\"Number of words with most refs: \" + str(len(longest)) + \", with number of refs: \"\\\n",
    "          + str(most_refs) + \"\\n\")\n",
    "    print(list(longest.headers)[word])\n",
    "    print(re.sub(\"\\n\", \" \", list(longest.definition)[word]))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 45,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Number of words with most refs: 2, with number of refs: 7\n",
      "\n",
      "MOULD\n",
      "or Mold, in the mechanic arts, &c. a cavity artificially cut, with a defign to give its form or impreflion to fome fofter matter applied therein. Moulds are implements of great ufe in fculpture, foundery, &c. The workmen employed in melting the mineral or metallic ore dug out of mines, have their feveral moulds to receive the melted metal as it comes out of the fur nace ; but thefe are different according to the diverfity of metals and works. In gold mines, they have moulds for ingots ; in filver mines, for bars; in copper and lead mines, for pigs or falmons; in tin mines, for pigs and ingots; and in iron mines, for fows, chimney backs, anvils, caldrons, pots, and other large utenfils and mer- chandifes of iron ; which are here caft, as it were, at firft hand. Moulds of founders of large works, as ftatues, bells, guns, and other brazen works, are of wax, fupported within-fide by what we call a core, and covered with- out-fide with a cape or cafe. It is in the fpace which the wax took up, which is afterwards melted away to leave it free, that the liquid metal runs, and the work is formed ; being carried thither through a great num- Moulds, ber of little canals, which cover the whole mould. See Foundery. Moulds of moneyers are frames full of fand, where in the plates of metal are caft that are to ferve for the ftriking of fpecies of gold and filver. See Coining. A fort of concave moulds made of clay, having within them the figures and inferiptions of ancient Roman coins, are found in many parts of England, and fuppofed to have been ufed for the calling of money. Mr Baker having been favoured wdth a fight of fome of thefe moulds found in Shropftiire, bearing the fame types and inferiptions with fome of the Roman coins, gave an account of them to the Royal Society. They w’ere found in digging of fand, at a place called Ryton in Shroplhire, about a mile from the great Watling- ftreet'road. They are all of the fize of the Roman de narius, and of little more than the thicknefs of oor halfpenny. They are made of a fmooth pot or brick clay, which feems to have been firft well cleanfed from dirt and fand, and well beaten or kneaded, to render it fit for taking a fair impreffion. There were a great many of them found together, and there are of them npt unfrequently found in Yorklhire; but they do not feem to have been met wdth in any other kingdom, ex cept that fome have been faid to be once found at Lyons. They have been fometimes found in great numbers joined together fide by fide, on one flat piece of clay, as if intended for the calling of a great num ber of coins at once ; and both thefe, and all the others that have been found, feem to have been of the empe ror Severus. They are fometimes found impreffed on both fides, and fome have the head of Severus on one fide and fome well known reverfe of his on the other. They feem plainly to have been intended for the coin age of money, though it is not eafy to fay in what manner they can have been employed for that purpofe, efpecially thofe which have impreflions on both fides, unlefs it may be fuppofed that they coined two pieces at the fame time by the help of three moulds, of which this was to be the middle one. If by difpofing thefe into fome fort of iron frame or cafe, as our letter- founders do the brafs moulds for calling their types, the melted metal could be eafily poured into them, it would certainly be a very eafy method of coining, as fuch moulds require little time or expence to make, arid therefore might be fupplied with new ones as often as they happen to break. Thefe moulds feem to have been burnt or baked fufficiently to make them hard ; but not fo as to ren der them porous like our bricks, whereby they would have loft their fmooth and even furface, which in thefe is plainly fo clofe, that whatever metal Ihould be form ed in them would have no appearance like the fand- holes by which counterfeit coins and metals are ufually detefted. Moulds of founders of fmall works are like the frames of coiners : it is in the'e frames, which are likewife filled* with fand, that their feveral works are falhioned ; into w'hich, when the two frames of which the mould is compofed, are rejoined, the melted brafs is run. Moulds of letter founders are partly of fteel and partly of wood. The wood, properly fpeaking, ferves only to covet the real mould which is within, and to 3 N 2 prevent M O U t46B] M O U Moulds. prevent the workman, who holds it in his hand, from being incommoded by the heat of the melted metal. Only one letter or type can be formed at once in each mould. See Letter Foundery. Moulds, in the manufatlure of paper, are little frames compofed of feveral brafs or iron wires, faflen- ed together by another wire flill finer. Each mould is of the bignefs of the fheet of paper to be made, and has a rim or ledge of wood to which the wires are faftened. Thefe moulds are more ufually called frames or forms. See Pafkr-Making. Moulds, with furnace and crucinle makers, are made of wood, of the fame form with the crucinles j that is, in form of a truncated cone : they have handles of wood to hold and turn them with, when, being covered with the earth, the workman has a mind to round or flatten his veflfel. Moulds for leaden bullets are little iron pincers, each of whofe branches terminates in a hemifpherical concave, which when {hut form an entire fphere. In the lips or fides where the branches meet, is a little jet or hole, through which the melted lead is conveyed. Laboratory Moulds -are made of w'ood, for filling and driving all forts of rockets and cartridges, &c. Glaciers Moulds. The glaziers have two kinds of moulds, both ferving to caff their lead : in the one they cafl: the lead into long rods or canes fit to be drawn through the vice, and the grooves formed therein 5 this they fometimes call ingot-mould. In the other, they mould thofe little pieces of* lead a line thick and tw7o lines broad, faftened to the iron bars. Thefe may be alfo caft in the vice. Goldfmiths Moulds. The goldfmiths ufe the bones of the cuttle fifti to make moulds for their fmall works; which they do by prefling the pattern between two bones, and leaving a jet or hole to convey the metal through, after the pattern has been taken out. Mould, among mafons, is a piece of hard wood or iron, hollowed within fide, anfwerable to the contours of the mouldings or cornices, &c. to be formed. This is othervvife called caliner. Moulds, among plumbers, are the tables on which they caft ftieets of lead. Thefe they fometimes call {imply tables. Befides which they have other real moulds, wherewith they caft pipes without foldering. See each defcrined under Plumbery. Moulds, among the glafs grinders, are wooden frames, whereon they make the tubes wherewith they fit their perfpedlives, telefcopes, and other optic ma chines. Thefe moulds are cylinders, of a length and diameter according to the ufe they are to be applied to, but always thicker at one end than the other, to facilitate the Aiding. The tubes made on thefe moulds are of two kinds j the one Amply of pafteboard and paper ; the other of thin leaves of wood joined to the pafteboard. To make thefe tubes to draw out, only the laft or innermoft is formed on the mould j each tube made afterwards ferving as a mould to that which is to go over it, but without taking out the mould from the firft. See Grinding. Moulds ufed in bafket-making are very Ample, con fiding ordinarily of a willow or ofier turned or bent in to an oval, circle, fquare, or other figure, according to the balkets, panniers, hampers, and other utenfils in tended. Ou khefe moulds they make, or more properly meafure, all their wrork j and accordingly they have Mould them of all fizes, fhapes, &c. li Mould, in Ihip-building, a thin flexinle piece of ^0Ulinet,l timber, uftd by ihipwrights as a pattern whereby to ”~V form the different curves of the timbers, and other com- pafting pieces in a {flip’s frame. There are two forts of thefe, viz, the bend mould and hollow mould ; the form er of thefe determines the convexity ot the timbers, and the latter their concavity on the outfide, where they ap proach the heel, particularly towards the extremities of the veffel. The figure given to the timbers by this pattern is called their bevelling. Moulds, among tallow chandlers, are of twTo kinds: the firft for the common dipped candles, being the vef fel wherein the melted tallow is difpofed, and the wick dipped. This is of wood, of a triangular form, and fupported on one of its angles; fo that it has an open ing of near a foot a-top : the other, ufed in the fabric of mould candles, is of brafs, pewter, or tin.—Here each candle has its feveral mould. See Candle. Mould, among gol^-beatersT1 a certain number of leaves of vellum or pieces of gut, cut fquare, of a cer tain fize, and laid over one another, between which they put the leaves of gold and filver which they beat on the marble with the hammer, See Gold leaf. They have four kinds of moulds j two whereof are of vellum and two of gut : the fmallelt of thofe of vellum confifts of 40 or 50 leaves j the largelt contains 100 : for the others, each contains 500 leaves. The moulds have all their feveral cafes, confiding of two pieces of parchment, ferving to keep the leaves of the mould in their place, and prevent their being diforder- ed in beating. Mould, in Agriculture, a general name for the foft earthy fubftance with which the dry land is generally covered, and in which all kinds of vegetables take root and grown It is far from being an homogeneous fub ftance 5 being compofed of decayed animal and vege table matters, along with calcareous, argillaceous, and filiceous earths, mixed together in various proportions, and with the different degrees of moiftme, conftituting every variety of soil. \n"
     ]
    }
   ],
   "source": [
    "most_refs_word(3, 1, 0)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 131,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "['AXIS', 'CALCULUS']"
      ]
     },
     "execution_count": 131,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# print out words with specific number of references\n",
    "\n",
    "list(headers_with_refs[0].loc[headers_with_refs[0]['refs_in_desc'] == 6].headers)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Proportional topic popularity changes"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [],
   "source": [
    "# calculate proportional popularity of topics (count of references to topic, adjusted to total number of references in that ed)\n",
    "# and visualise the popularity across editions\n",
    "\n",
    "def change_in_topic_vis(topic, editions_ref_counts):\n",
    "    \n",
    "    counts = []\n",
    "    sum_refs = []\n",
    "    \n",
    "    for i in range(0, len(editions_ref_counts)):\n",
    "        try: \n",
    "            counts.append(int(editions_ref_counts[i].loc[editions_ref_counts[i].references == topic]['count']) / editions_ref_counts[i]['count'].sum())\n",
    "        except: # handle case where a topic was never referenced in early editions\n",
    "            counts.append(0) \n",
    "    # proportional popularity, by dividing count by total number of references in an edition\n",
    "    ed_count = list(range(1, 9))\n",
    "    sns.set_style(\"whitegrid\")\n",
    "    ax = sns.barplot(x= ed_count, y=counts, color=\"#cfbd93\")\n",
    "    ax.set_title('Popularity of ' + topic + ' across editions', fontsize=16)\n",
    "    ax.set_xlabel('editions', fontsize=14)\n",
    "    ax.set_ylabel('proportion of references', fontsize=14)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "change_in_topic_vis(\"Chemistry\", references)\n",
    "# a decrease in popularity could mean entries were added to the long-form article,\n",
    "# so doesn't necessarily mean topic itself became less popular;\n",
    "# all that we can state from this is that proportion of entries referencing that topic decreased"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Total number of references for top 20 topics for all editions"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "                                    ref count\n",
      "0                               Anatomy  1235\n",
      "1                                   Law   497\n",
      "2                              Medicine   429\n",
      "3                             Chemistry   608\n",
      "4                             Astronomy   397\n",
      "...                                 ...   ...\n",
      "22761       Hamilton’s edition\\nof Reid     1\n",
      "22762                            Quilon     1\n",
      "22763  the\\n(Euvres Completes de Volney     1\n",
      "22764       the historical\\npart of art     1\n",
      "22765                            Police     1\n",
      "\n",
      "[22766 rows x 2 columns]\n"
     ]
    }
   ],
   "source": [
    "# goes through each dataframe and checks whether that reference exists in total_refs already\n",
    "# if it does, skips it (it was already calculated)\n",
    "# if does not yet, goes through all further editions and sums the references to that topic\n",
    "\n",
    "total_refs = pd.DataFrame(columns=['ref', 'count'])\n",
    "\n",
    "for i in range(0, len(references)):\n",
    "    for index, row in references[i].iterrows():\n",
    "        #only process a word in an edition if it was not processed by a previous edition\n",
    "        if (row.references not in total_refs.values):          \n",
    "            count = 0\n",
    "            for ed in references[i:]:\n",
    "                try: \n",
    "                    count += int(ed.loc[ed.references == row.references]['count'])\n",
    "                except: # handle case where there's not a reference to a topic in an edition\n",
    "                    pass\n",
    "            total_refs = total_refs.append({'ref': row.references, 'count': count}, ignore_index=True)\n",
    "            \n",
    "print(total_refs)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [],
   "source": [
    "# sort by most referenced\n",
    "\n",
    "total_refs = total_refs.sort_values(by=['count'], ascending=False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<seaborn.axisgrid.FacetGrid at 0x152cce50>"
      ]
     },
     "execution_count": 18,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1440x720 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# visualise top 20 most referenced topics across all editions\n",
    "\n",
    "sns.set(font_scale=1.5)\n",
    "sns.set_style(\"whitegrid\")\n",
    "ax = sns.catplot(x='count', y='ref', data=total_refs.head(20), kind='bar', height=10, aspect=2, color=\"#cfbd93\")\n",
    "ax.fig.suptitle(\"Top 20 most referenced topics across all editions\", y=1)\n",
    "ax.set_axis_labels('counts of references', 'topic')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Unused code"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "['or Entablement, in Archi\\ntecture, is that part of an order of a column which is\\nover the capital, and comprehends the architrave, frize,\\nand corniche. See Architecture, chap. i.\\n']"
      ]
     },
     "execution_count": 37,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# find definition of a specific word in a specific edition\n",
    "\n",
    "list(headers[5].loc[headers[5]['headers'].str.contains(\"ENTABLATURE\")].definition)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "['a compound, or affemblage of feveral dif\\nferent bodies in the fame mafs. See Chemistry.\\n']\n",
      "['a compound, or affemblage of feveral\\ndifferent bodies in the fame mafs. Simple mixture,\\nconfifts only in the Ample appofition of parts of diffe\\nrent bodies to each other. Thus, when powders of\\ndifferent kinds are rubbed together, the mixture is on\\nly Ample, and each of the powders retains its particu\\nlar charafters. In like manner, when oil and water\\nare mixed together, though the parts of both are con\\nfounded, fo that the liquor may appear to be homoge\\nneous, we cannot fay that there is any more than a\\nAmple appofition of the parts, as the oil and water may\\nvery eafily be again feparated from each other. But\\nthe cafe is very different when bodies chemically\\nmixed ; for then one or both bodies affume new pro\\nperties, and can by no means be difeovered in their\\nproper form without a particular chemical procefs ad\\napted to this purpofe. Hence chemical mixture is\\nattended with many phenomena which are never obfer-\\nved in Ample mixtures ; fuch as heat, effervefcence,\\n&c. To chemical mixture belongs the union of acids\\nand alkalies, the amalgamation of metals, folution of\\ngums, &c. and upon it depend many of the principal\\noperations of chemiftry. See that article pajfim.\\n']\n",
      "['a compound or affemblage of feveral\\ndifferent bodies in the fame mafs. Simple mixture,\\nconfifts only in the Ample appofition of parts of diffe\\nrent bodies to each other. Thus, when powders of\\ndifferent kinds are rubbed together, the mixture is on-\\nly Ample, and each of the powders retains its particu\\nlar chara&ers. In like manner, when oil and water\\nare mixed together, though the parts of both are con\\nfounded, fo that the liquor may appear to be homoge\\nneous, we cannot fay that there is any more than a\\nAmple\\nM N I [ i £\\nAmple appofitlon of the parts, as the oil and water may\\nvery eafily be again feparated from each, other. But\\nthe cafe is very different when bodies are chemically\\nmixed ; for then one or both bodies affume new pro\\nperties, and can by no means be difcovered in their\\nproper form without a particular chemical procefs ad\\napted* to this purpofe. Hence chemical mixture is\\nattended with many phenomena which are never obfer-\\nved in Ample mixtures; fuch as heat, effervefcence,\\n&c. To chemical mixture belongs the union of acids\\nand alkalies, the amalgamation of metals, folution of\\ngums, &c. and upon it depend many of the principal\\noperations of Chemistry. See that article pajftm.\\nMixture, in pharmacy, a medicine which differs\\nfrom a julep in this i dpe<ft, that it receives into its\\ncompofition not only falls, extracts, and other fub-\\nItances diffoluble in water; but alfo earths, powders,\\nand fuch fubftances as cannot be diffolved.\\n']\n",
      "[]\n",
      "[]\n",
      "['a compound or assemblage ot several\\ndifferent bodies in the same mass. Chemical mixture is\\nattended with many phenomena which are never observ\\ned in simple mixtures ; such as heat, effervescence, &c.\\nTo chemical mixture belong the union of acids amt\\nalkalies, the amalgamation of metals, solution of gums,\\n&e. and upon it depend many of the principal opera\\ntions of Chemistry. See that article, passim.\\nMixture, in Pharmacy, a medicine which diiiers\\nfrom a julep in this respect, that it receives into its\\ncomposition not only salts, extracts, and othei su -\\nstances dissoluble in water; but also earths, powders,\\nand such substances as cannot be dissolved.\\n']\n",
      "['a compound or assemblage of several dif\\nferent bodies in the same mass. Chemical mixture is at\\ntended with many phenomena which are never observed\\nin simple mixtures; such as heat, effervescence, &c. To\\nchemical mixture belong the union of acids and alkalies,\\nthe amalgamation of metals, the solution of gums, &c.; and\\nupon it depend many of the principal operations of che\\nmistry.\\nM1']\n",
      "[]\n"
     ]
    }
   ],
   "source": [
    "# show definition of a specific word across all editions\n",
    "\n",
    "for i in range(0, len(headers)):\n",
    "    print(list(headers[i].loc[headers[i]['headers'].str.match(\"MIXTURE\")].definition))"
   ]
  }
 ],
 "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": "3.6.1"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}