1-3-2_scott_byron.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<div style='color:white;font-size:30px;background-color:#196363;text-align:center;padding-top:20px; padding-bottom:20px;margin-bottom:40px'>VERSIFICATION & AUTHORSHIP ATTRIBUTION</div>\n",
    "<h1>2.3.1 Burrows' Delta<br>\n",
    "<br>model situation: recognizing G. G. Byron and W. Scott</h1>\n",
    "<h2 style='margin-bottom: 50px'>Replication code</h2>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [],
   "source": [
    "import json\n",
    "import matplotlib.pyplot as plt\n",
    "import numpy as np\n",
    "import pandas as pd\n",
    "from collections import Counter\n",
    "from collections import defaultdict\n",
    "from scipy.stats import zscore"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<h3>Read data from JSON</h3>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [],
   "source": [
    "#----------------------------------------------------------------------------------------------------------#\n",
    "# File contains a dict with following structure:                                                           #   \n",
    "# {                                                                                                        #\n",
    "#     'marmion': list of words (tokens) in Scott's Marmion,                                                #\n",
    "#     'lady': list of words (tokens) in Scott's Lady of the Lake                                           #\n",
    "#     'childe': list of words (tokens) in Byron's Childe Harold,                                           #\n",
    "# }                                                                                                        #\n",
    "# Dataset comes from Gutenberg project (https://www.gutenberg.org/).                                       #\n",
    "# Tokenization provided by TreeTagger (https://www.cis.uni-muenchen.de/~schmid/tools/TreeTagger/).         #\n",
    "#----------------------------------------------------------------------------------------------------------#\n",
    "\n",
    "with open('data/scott_byron.json', 'r') as f:\n",
    "    data = json.load(f)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<h3>Relative frequencies of the 20 most common words</h3>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1440x576 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Concat the three lists and get 20 most frequent words in entire corpus\n",
    "corpus = data['marmion'] + data['lady'] + data['childe']\n",
    "c = Counter(corpus)\n",
    "most_frequent = [ x[0] for x in c.most_common(20) ]    \n",
    "\n",
    "# Count their relative frequencies in each particular poem \n",
    "frequencies = defaultdict(lambda: defaultdict(int))\n",
    "for poem in data:\n",
    "    for word in data[poem]:\n",
    "        if word in most_frequent:      \n",
    "            frequencies[poem][word] += 1/len(data[poem])\n",
    "\n",
    "# Print the frequencies chart\n",
    "df = pd.DataFrame.from_dict(frequencies).reindex(most_frequent).reindex(\n",
    "    sorted(frequencies.keys(), reverse=True), axis=1\n",
    ")\n",
    "ax = df.plot.bar(figsize=(20,8), color=['#1f77b4', '#ff7f0e', '#2ca02c'], xticks=range(20));\n",
    "ax.set_xlabel(\"20 most frequent words\")\n",
    "ax.set_ylabel(\"frequency\");"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<h3>Relative frequencies to z-scores</h3>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1440x576 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "zscores = defaultdict(dict)\n",
    "\n",
    "# Iterate over most frequent words\n",
    "for word in most_frequent:\n",
    "\n",
    "    # Calculate mean and standard deviation of its frequencies across poems\n",
    "    distribution = [frequencies['childe'][word], frequencies['marmion'][word], frequencies['lady'][word]]\n",
    "    mean = sum(distribution) / len(distribution)\n",
    "    stddev = (sum([(x-mean)**2 for x in distribution]) / len(distribution))**0.5\n",
    "\n",
    "    # Store its z-scores\n",
    "    for poem in frequencies:\n",
    "        zscores[poem][word] = (frequencies[poem][word] - mean) / stddev\n",
    "        \n",
    "# Print z-scores chart\n",
    "df = pd.DataFrame.from_dict(zscores).reindex(most_frequent).reindex(sorted(\n",
    "    frequencies.keys(), reverse=True), axis=1\n",
    ")\n",
    "ax = df.plot.bar(figsize=(20,8), color=['#1f77b4', '#ff7f0e', '#2ca02c'], xticks=range(20));\n",
    "ax.set_xlabel(\"20 most frequent words\")\n",
    "ax.set_ylabel(\"frequency transformed to z-score\");"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<h3>Delta measure</h3>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "DELTA(childe) = 1.6702187476717127)\n",
      "DELTA(marmion) = 1.3752470162010222)\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1440x576 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "diffs = defaultdict(lambda: defaultdict(int))\n",
    "sums = defaultdict(int)\n",
    "\n",
    "# Iterate over most frequent words\n",
    "for w in most_frequent:\n",
    "    # Calculate and sum up the differences between Childe Harold / Marmion and Lady of the Lake\n",
    "    for poem in ('childe', 'marmion'):\n",
    "        diffs[poem][w] = abs(zscores[poem][w] - zscores['lady'][w])\n",
    "        sums[poem] += diffs[poem][w] \n",
    "\n",
    "# Print delta measures\n",
    "print('DELTA({0}) = {1})'.format('childe', sums['childe'] / len(most_frequent)))\n",
    "print('DELTA({0}) = {1})'.format('marmion', sums['marmion'] / len(most_frequent)))\n",
    "        \n",
    "# Print z-scores chart\n",
    "df = pd.DataFrame.from_dict(diffs).reindex(most_frequent).reindex(\n",
    "    sorted(sums.keys(), reverse=True), axis=1\n",
    ")\n",
    "ax = df.plot.bar(figsize=(20,8), color=['#1f77b4', '#2ca02c'], xticks=range(20));\n",
    "ax.set_xlabel(\"20 most frequent words\")\n",
    "ax.set_ylabel(\"distance from target\");"
   ]
  }
 ],
 "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": "3.6.8"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}