02_pride_and_predjudice.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Pride & Prejudice analysis"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Real text analysis\n",
    "\n",
    "We got familiar with Spacy. In the next section we are going to analyse a real text (Pride & Prejudice). \n",
    "\n",
    "We would like to:\n",
    "* Extract the names of all the characters from the book (e.g. Elizabeth, Darcy, Bingley)\n",
    "* Visualize characters' occurences with regards to relative position in the book\n",
    "* Authomatically describe any character from the book\n",
    "* Find out which characters have been mentioned in a context of marriage\n",
    "* Build keywords extraction that could be used to display a word cloud ([example](http://www.cytora.com/data-samples.html))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Load text file"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "metadata": {},
   "outputs": [],
   "source": [
    "def read_file(file_name):\n",
    "    with open(file_name, 'r') as file:\n",
    "        return file.read()  # .decode('utf-8')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Process full text"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "metadata": {},
   "outputs": [],
   "source": [
    "import spacy\n",
    "\n",
    "nlp = spacy.load('en')\n",
    "\n",
    "# Process `text` with Spacy NLP Parser\n",
    "text = read_file('data/pride_and_prejudice.txt')\n",
    "processed_text = nlp(text)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### How many sentences are there in the book (Pride & Prejudice)?"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 44,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "6469\n"
     ]
    }
   ],
   "source": [
    "sentences = [s for s in processed_text.sents]\n",
    "print(len(sentences))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Print sentences from index 10 to index 15\n",
    "... to make sure that we have parsed the correct book"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 45,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[Chapter 1\n",
      "\n",
      "\n",
      "It is a truth universally acknowledged, that a single man in possession\n",
      "of a good fortune, must be in want of a wife.\n",
      "\n",
      ", However little known the feelings or views of such a man may be on his\n",
      "first entering a neighbourhood, this truth is so well fixed in the minds\n",
      "of the surrounding families, that he is considered the rightful property\n",
      "of some one or other of their daughters.\n",
      "\n",
      ", \"My dear Mr. Bennet,\" said his lady to him one day, \"have you heard that\n",
      "Netherfield Park is let at last?\"\n",
      "\n",
      ", Mr. Bennet replied that he had not.\n",
      "\n",
      ", \"But it is,\" returned she; \"for Mrs. Long has just been here, and she\n",
      "told me all about it.\"\n",
      "\n",
      "]\n"
     ]
    }
   ],
   "source": [
    "print(sentences[10:15])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Exercise: Find all the personal names\n",
    "\n",
    "Extract all the personal names from Pride & Prejudice and count their occurrences. \n",
    "\n",
    "Expected output is a list in the following form: [('elizabeth', 622), ('darcy', 312), ('jane', 286), ('bennet', 266) ...].\n",
    "\n",
    "*Hint: you can use a Counter: a container that keeps track of how many times equivalent values are added*\n",
    "\n",
    "*SuperHint: iterate over the entities checking if the label of each entity is PERSON*"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 46,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[]\n"
     ]
    }
   ],
   "source": [
    "from collections import Counter, defaultdict\n",
    "\n",
    "def find_character_occurences(doc):\n",
    "    \"\"\"\n",
    "    Return a list of actors from `doc` with corresponding occurences.\n",
    "    \n",
    "    :param doc: Spacy NLP parsed document\n",
    "    :return: list of tuples in form\n",
    "        [('elizabeth', 622), ('darcy', 312), ('jane', 286), ('bennet', 266)]\n",
    "    \"\"\"\n",
    "    \n",
    "    characters = Counter()\n",
    "    \n",
    "    # your code here\n",
    "            \n",
    "    return characters.most_common()\n",
    "\n",
    "print(find_character_occurences(processed_text)[:20])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Plot characters personal names as a time series "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 47,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Matplotlib Jupyter HACK\n",
    "%matplotlib inline\n",
    "\n",
    "import matplotlib\n",
    "import numpy as np\n",
    "import matplotlib.pyplot as plt"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 48,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Plot characters' mentions as a time series relative to the position of the actor's occurrence in a book.\n",
    "\n",
    "def get_character_offsets(doc):\n",
    "    \"\"\"\n",
    "    For every character in a `doc` collect all the occurences offsets and store them into a list. \n",
    "    The function returns a dictionary that has actor lemma as a key and list of occurences as a value for every character.\n",
    "    \n",
    "    :param doc: Spacy NLP parsed document\n",
    "    :return: dict object in form\n",
    "        {'elizabeth': [123, 543, 4534], 'darcy': [205, 2111]}\n",
    "    \"\"\"\n",
    "    \n",
    "    character_offsets = defaultdict(list)\n",
    "    for ent in doc.ents:\n",
    "        if ent.label_ == 'PERSON':\n",
    "            character_offsets[ent.lemma_].append(ent.start)\n",
    "            \n",
    "    return dict(character_offsets)\n",
    "\n",
    "character_occurences = get_character_offsets(processed_text)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 49,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "from matplotlib.pyplot import hist\n",
    "from cycler import cycler\n",
    "\n",
    "NUM_BINS = 10\n",
    "\n",
    "def normalize(occurencies, normalization_constant):\n",
    "    return [o / float(len(processed_text)) for o in occurencies]\n",
    "\n",
    "def plot_character_timeseries(character_offsets, character_labels, normalization_constant=None):\n",
    "    \"\"\"\n",
    "    Plot characters' personal names specified in `character_labels` list as time series.\n",
    "    \n",
    "    :param character_offsets: dict object in form {'elizabeth': [123, 543, 4534], 'darcy': [205, 2111]}\n",
    "    :param character_labels: list of strings that should match some of the keys in `character_offsets`\n",
    "    :param normalization_constant: int\n",
    "    \"\"\"\n",
    "    x = [character_offsets[character_label] for character_label in character_labels] \n",
    "        \n",
    "    with plt.style.context('fivethirtyeight'):\n",
    "        plt.figure()\n",
    "        n, bins, patches = plt.hist(x, NUM_BINS, label=character_labels)\n",
    "        plt.clf()\n",
    "        \n",
    "        ax = plt.subplot(111)\n",
    "        for i, a in enumerate(n):\n",
    "            ax.plot([float(x) / (NUM_BINS - 1) for x in range(len(a))], a, label=character_labels[i])\n",
    "            \n",
    "        matplotlib.rcParams['axes.prop_cycle'] = cycler(color=['r','k','c','b','y','m','g','#54a1FF'])\n",
    "        ax.legend(loc='center left', bbox_to_anchor=(1, 0.5))\n",
    "\n",
    "plot_character_timeseries(character_occurences, ['darcy', 'bingley'])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Spacy parse tree in action"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 50,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "['-PRON-', '-PRON-', 'late', 'late', 'late', 'late', 'intimate', 'confidential', 'much', '-PRON-', '-PRON-', 'late', 'superior', 'whose', '-PRON-', 'whose', 'evident', 'grave', 'late', 'abominable', '-PRON-', '-PRON-', 'late', 'few', 'least', '-PRON-', '-PRON-', 'poor', '-PRON-', '-PRON-', '-PRON-', 'clever', 'worth', 'engross', 'politeness', 'delight', 'bewitch', 'studious', 'sorry', 'unworthy', 'answerable', 'impatient', 'ashamed', 'ill', 'punctual', 'engaged', 'fond']\n"
     ]
    }
   ],
   "source": [
    "# Find words (adjectives) that describe Mr. Darcy.\n",
    "\n",
    "def get_character_adjectives(doc, character_lemma):\n",
    "    \"\"\"\n",
    "    Find all the adjectives related to `character_lemma` in `doc`\n",
    "    \n",
    "    :param doc: Spacy NLP parsed document\n",
    "    :param character_lemma: string object\n",
    "    :return: list of adjectives related to `character_lemma`\n",
    "    \"\"\"\n",
    "    \n",
    "    adjectives = []\n",
    "    for ent in processed_text.ents:\n",
    "        if ent.lemma_ == character_lemma:\n",
    "            for token in ent.subtree:\n",
    "                if token.pos_ == 'ADJ': # Replace with if token.dep_ == 'amod':\n",
    "                    adjectives.append(token.lemma_)\n",
    "    \n",
    "    for ent in processed_text.ents:\n",
    "        if ent.lemma_ == character_lemma:\n",
    "            if ent.root.dep_ == 'nsubj':\n",
    "                for child in ent.root.head.children:\n",
    "                    if child.dep_ == 'acomp':\n",
    "                        adjectives.append(child.lemma_)\n",
    "    \n",
    "    return adjectives\n",
    "\n",
    "print(get_character_adjectives(processed_text, 'darcy'))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 51,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[('Elizabeth', 41), ('Bennet', 28), ('Jane', 13), ('Miss Bingley', 9), ('Bingley', 8), ('Lady Catherine', 6), ('Gardiner', 6), ('Collins', 5), ('Wickham', 4), ('Lizzy', 4)]\n"
     ]
    }
   ],
   "source": [
    "# Find characters that are 'talking', 'saying', 'doing' the most. Find the relationship between \n",
    "# entities and corresponding root verbs.\n",
    "\n",
    "character_verb_counter = Counter()\n",
    "VERB_LEMMAS = ['say', 'talk']\n",
    "\n",
    "for ent in processed_text.ents:\n",
    "    if ent.label_ == 'PERSON' and ent.root.head.lemma_ in VERB_LEMMAS:\n",
    "        character_verb_counter[ent.text] += 1\n",
    "\n",
    "print(character_verb_counter.most_common(10))"
   ]
  }
 ],
 "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": "3.4.3"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 1
}