nlp_shakespeare.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Intro"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "This is a project I worked on a few months ago as I was learning about NLP and found myself with a digital copy of all the Shakespeare's plays.\n",
    "\n",
    "There is some controversy around Shakespeare and whether he actually was the author of all the work it's been attributed to him. Despite this the experts do agree on a classification of the plays into comedies and tragedies. Beside the fact that one has a happy ending and one doesn't the two genres do differ for language, themes and writing style.\n",
    "\n",
    "So the objective of this analysis is to identify, if possible, a split of the plays using natural language processing and compare it to the traditional classification to see it's captured by it."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Required libraries"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "import nltk\n",
    "import math\n",
    "import re\n",
    "import lda\n",
    "import sklearn\n",
    "import my_kmeans\n",
    "import numpy as np\n",
    "import matplotlib.pyplot as plt\n",
    "import seaborn as sns\n",
    "import pandas as pd\n",
    "\n",
    "from nltk.corpus import stopwords\n",
    "from nltk.tokenize import RegexpTokenizer\n",
    "from sklearn.feature_extraction.text import CountVectorizer, TfidfTransformer\n",
    "from sklearn.cluster import KMeans\n",
    "%matplotlib inline\n",
    "from ggplot import *"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "stopwords = nltk.corpus.stopwords.words('english')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Pre-processing"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Before doing any analysis we need to perform a few operations to make the text machine-usable, so we first tokenize it (which simply means create a list where every element is a unigram = single word) and then use a process called stemming to remove suffixes so that words with the same root/meaning have the same representation. After this we remove all the stop-words (words like 'and', 'you', 'auxiliary verbs'.. that are very common but don't give any real insight on the text). At the end we obtain a matrix representation of the collection where each document is represented by a vector or word frequencies."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Input Data"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "shakespeare.txt includes all Shakespeare's plays and poems. Each line of the file starts with the expression title@row_number:"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    ">hamlet@56053\tHAMLET\tWords, words, words.\n",
    ">\n",
    ">hamlet@56081\n",
    ">\n",
    ">hamlet@56082\tLORD POLONIUS\tWhat is the matter, my lord?\n",
    ">\n",
    ">hamlet@56125\n",
    ">\n",
    ">hamlet@56126\tHAMLET\tBetween who?\n",
    ">\n",
    ">hamlet@56146\n",
    ">\n",
    ">hamlet@56147\tLORD POLONIUS\tI mean, the matter that you read, my lord."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Text normalization"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "- find all the titles as a unique list of title@ expressions and delete the histories (for now) and the poems\n",
    "- split the text into a list of lines\n",
    "- for each play create an element in _collection_, delete title@row_number and store as plain text"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "collection = {} # dictionary of raw texts\n",
    "\n",
    "with open('../../Datastore/shakespeare.txt', 'r') as f:\n",
    "    raw = f.read()\n",
    "    titles = [t for t in set(re.findall(r'[a-z]+@{1}', raw))]\n",
    "\n",
    "    histories = ['glossary@', 'various@', 'sonnets@', 'kinghenryvi@', 'kinghenryiv@', \n",
    "                 'venusandadonis@', 'rapeoflucrece@', 'loverscomplaint@', 'kingrichardiii@', \n",
    "                 'kingrichardii@', 'kingjohn@', 'kinghenryv@','kinghenryviii@']\n",
    "    titles = [title for title in titles if title not in histories]\n",
    "\n",
    "    lines = raw.splitlines()\n",
    "    for t in titles:\n",
    "        key = t[:-1]\n",
    "        text = [re.sub(r'[a-z]+@{1}[0-9]+[ \\t\\n]+','', l) for l in lines if re.match(t+'[0-9]+', l)]\n",
    "        collection[key] = ' '.join(text)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "- Initialize the tokenizer (disregards punctuation signs) and a stemmer (Porter algorithm)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "tokenizer = RegexpTokenizer(r'\\w+')\n",
    "st = nltk.stem.PorterStemmer()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Add glossary of victorian english words"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "with open('glossary.txt', 'r') as f:\n",
    "    glossary = f.read()\n",
    "\n",
    "glossary = tokenizer.tokenize(glossary)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "tokens = {}\n",
    "stemmed = {}\n",
    "documents = {} # dictionary of normalized texts\n",
    "\n",
    "for key in collection.keys():\n",
    "    collection[key] = re.sub(r'\\n?\\n[A-Z]*\\t|\\n|\\t', ' ', collection[key])\n",
    "    tokens[key] = tokenizer.tokenize(collection[key]) # delete punctuation\n",
    "    names = [w.lower() for w in set(tokens[key]) if w.isupper()]\n",
    "    tokens[key] = [w for w in tokens[key] if w.lower() not in names]\n",
    "    tokens[key] = [w.lower() for w in tokens[key] if w.lower() not in stopwords + glossary]\n",
    "    stemmed[key] = [st.stem(w.lower()) for w in tokens[key]]\n",
    "    documents[key] = ' '.join(stemmed[key])    "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "from gensim import corpora, models\n",
    "\n",
    "texts = []\n",
    "for doc in stemmed:\n",
    "    texts.append(stemmed[doc])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "words = corpora.Dictionary(texts)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "corpus = [words.doc2bow(text) for text in texts]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "hdpmodel = models.hdpmodel.HdpModel(corpus, id2word=words)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "150"
      ]
     },
     "execution_count": 11,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(hdpmodel.show_topics(topics=-1))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**Feature Extraction/Vectorization**: given a Vocabulary (list of unique words in the corpus) each play is represented as an array of frequencies"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "vectorizer = CountVectorizer()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "tfidfTransf = TfidfTransformer()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "CountVectorizer() requires as input a list of documents (raw text)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(27, 11959)"
      ]
     },
     "execution_count": 14,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "freq_matrix = vectorizer.fit_transform(list(documents.values())).toarray()\n",
    "freq_matrix.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(27, 11959)"
      ]
     },
     "execution_count": 15,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "tfidf_matrix = tfidfTransf.fit_transform(freq_matrix)\n",
    "tfidf_matrix.shape"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Latent Dirichlet Allocation"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Latent Dirichlet Allocation is an unsupervised technique for topic modelling.\n",
    "Given a collection of documents, and k probability distributions over a collection of words (which we're going to call topic) LDA describe the statistical relationship between random variables that specifies a probabilistic procedure to generate topics, and documents. In this prospective LDA is a generative model, that is it describes how document obtains its words.\n",
    "\n",
    "LDA describes documents generation as a 3 step process:\n",
    "1. for each document draw a topic distribution\n",
    "2. for each word in a document\n",
    "1. draw a specific topic\n",
    "2. draw a word from the multinomial probability conditioned on the topic\n",
    "\n",
    "We can graphically display these model:\n",
    "alpha and beta are corpus level parameters, where alpha is the parameter for the dirichlet distribution describing the topics within documents and beta is a metrics of word probabilities in a topic. This parameters are only selected once in the process of generating the corpus.\n",
    "then the outher box represent documents, with theta being a document-level parameter sampled once per each document.\n",
    "And the inner plate represents the repeated choice of topics and words within a document, where topics and words (z and w) are selected once per each word in each document.\n",
    "\n",
    "\n",
    "what we are most interested in for this analysis is the dirichlet distribution. This because it can be used to model the randomness of probability mass functions.\n",
    "A probability mass function gives the probability of a discrete variable to take a specific value x. A simple example of pmf is a 6-sided dice. If we have a bag full of dices and we randomly extract one we are effectively drawing from a random pmf. This can be modelled with a Dir. Of parameter alpha.\n",
    "In the text example, a document is a collection of words and therefore can be regarded as a multinomial, so if x is a multinomial of parameters n and q, the dirichlet serves as a prior for the parameter q, where Qi positive and the sum of them equals 1.\n",
    "In LDA topics are latent multinomials so we can use the dirichlet distribution to describe the probability of each topic in each document."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "titles = np.array(documents.keys())\n",
    "labels = np.array(['tragedy','comedy','comedy', 'tragedy','tragedy','tragedy','tragedy', \n",
    "                           'tragedy','comedy','comedy','comedy', 'comedy','tragedy','comedy', \n",
    "                           'tragedy', 'comedy', 'comedy','comedy', 'comedy', 'comedy','comedy', \n",
    "                           'comedy', 'tragedy', 'tragedy', 'comedy','comedy','comedy'])\n",
    "year = np.array([1591, 1604, 1601, 1606, 1599, 1606, 1605, 1599, 1610, 1596, 1594, 1609, \n",
    "                 1605, 1598, 1607, 1600, 1603, 1610, 1590, 1599, 1600, 1595, 1604, 1592, \n",
    "                 1594, 1589, 1607])\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "lookup = []\n",
    "for index in xrange(len(titles)):\n",
    "    lookup.append( (titles[index], labels[index], year[index]))\n",
    "\n",
    "lookup = np.array(lookup)\n",
    "unique_words = vectorizer.get_feature_names()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 90,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<lda.lda.LDA instance at 0x16a5b1248>"
      ]
     },
     "execution_count": 90,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "num_of_topics = 12\n",
    "topic_model = lda.LDA(n_topics= num_of_topics, n_iter=500, random_state=123)\n",
    "topic_model.fit(freq_matrix)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 91,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "-2235165.7166455593"
      ]
     },
     "execution_count": 91,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "topic_model.loglikelihood()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 92,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[<matplotlib.lines.Line2D at 0x169b8dd10>]"
      ]
     },
     "execution_count": 92,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAgsAAAFbCAYAAABbOc2kAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x1699c7990>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.plot(topic_model.loglikelihoods_)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "As a result, we obtain *num_of_topic* word distributions, which represent the topics (top 10 here) and, more interesting  *num_of_topics-dimentional* distribution of topics within documents."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Model with 9 topics"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 93,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Topic 0: well daughter come god father signior tell night old\n",
      "Topic 1: night fear death day done us dead live hand\n",
      "Topic 2: shall good enter let would make know give like\n",
      "Topic 3: son father eye queen hand tear sweet king see\n",
      "Topic 4: love one man sweet ladi fair wit faith therefor\n",
      "Topic 5: sir lord first one ti honour great gentleman whose\n",
      "Topic 6: lord men servant god fool man brother old first\n",
      "Topic 7: rome nobl citizen senat well roman first second honour\n",
      "Topic 8: come go see look must man us yet made\n",
      "Topic 9: master come well provost host good wife go woman\n",
      "Topic 10: sir master father well hous mistress one home pray\n",
      "Topic 11: well eye word three letter grace true still light\n"
     ]
    }
   ],
   "source": [
    "topic_word = topic_model.topic_word_\n",
    "\n",
    "n_top_words = 10\n",
    "\n",
    "for i, topic_dist in enumerate(topic_word):\n",
    "    topic_words = np.array(unique_words)[np.argsort(topic_dist)][:-n_top_words:-1]\n",
    "    print('Topic {}: {}'.format(i, ' '.join(topic_words)))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 94,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAjkAAAGpCAYAAABvZSezAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nMcghIiIiWWKQQ0RERLL0fyydrZl/fdI4AAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x16a749bd0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "f, ax= plt.subplots(5, 1, figsize=(8, 6), sharex=True)\n",
    "for i, k in enumerate([0, 1, 4, 6, 8]):\n",
    "    ax[i].stem(topic_word[k,:], linefmt='b-',\n",
    "               markerfmt='bo', basefmt='w-')\n",
    "    ax[i].set_xlim(-50,4350)\n",
    "    ax[i].set_ylim(0, 0.04)\n",
    "    ax[i].set_ylabel(\"Prob\")\n",
    "    ax[i].set_title(\"topic {}\".format(k))\n",
    "\n",
    "ax[4].set_xlabel(\"word\")\n",
    "\n",
    "plt.tight_layout()\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 95,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "pd.DataFrame(topic_word, columns=unique_words).to_csv('outputs/topic_word.csv')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 96,
   "metadata": {
    "collapsed": false,
    "scrolled": false
   },
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAjkAAAR4CAYAAADkP8eeAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nEZEs/R8d2bTKL5ANjgAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x1591600d0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "doc_topic = topic_model.doc_topic_\n",
    "\n",
    "f, ax= plt.subplots(13, 1, figsize=(8, 16), sharex=True)\n",
    "for i, k in enumerate([1, 3, 5, 7, 9, 11, 13, 15, 17, 19, 21, 23, 25]):\n",
    "    ax[i].stem(doc_topic[i,:], linefmt='b-',\n",
    "               markerfmt='bo', basefmt='w-')\n",
    "    ax[i].set_xlim(-1, 9)\n",
    "    ax[i].set_ylim(0, 0.4)\n",
    "    ax[i].set_ylabel(\"Prob\")\n",
    "    ax[i].set_title(titles[i])\n",
    "\n",
    "ax[4].set_xlabel(\"Topic\")\n",
    "\n",
    "plt.tight_layout()\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 97,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "np.savetxt(\"outputs/topics.csv\", topic_model.doc_topic_, delimiter=\",\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The distribution of topics over each document is used as input of the k-means. "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### K-Means"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "K means is an unsupervised clustering technique that creates a partition of the data such that every observation is assigned to the cluster with nearest mean. This guarantees minimal variability within observations in the same cluster and maximum variability between groups.\n",
    "\n",
    "The results of this last step is a group almost entirely made of comedies, except for 'Romeo and Juliet' and a group that is almost evanly split between comedies and tragedies. So it looks like nlp doesn't capture the differences between comedies and tragedies.\n",
    "However, experts do agree that Shakespeare's style has changed over the years."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### KM on 9 topics"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 98,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "KMeans(copy_x=True, init='random', max_iter=300, n_clusters=2, n_init=10,\n",
       "    n_jobs=1, precompute_distances='auto', random_state=15, tol=0.0001,\n",
       "    verbose=0)"
      ]
     },
     "execution_count": 98,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "km_topics = KMeans(n_clusters = 2, init = 'random', random_state = 15)\n",
    "km_topics.fit(doc_topic)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 99,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([1, 1, 0, 1, 1, 1, 1, 1, 1, 0, 0, 1, 1, 0, 1, 1, 1, 1, 0, 0, 0, 0, 1,\n",
       "       0, 0, 0, 1], dtype=int32)"
      ]
     },
     "execution_count": 99,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "km_topics.labels_"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 100,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([['twelfthnight', 'comedy', '1601'],\n",
       "       ['merchantofvenice', 'comedy', '1596'],\n",
       "       ['loveslabourslost', 'comedy', '1594'],\n",
       "       ['muchadoaboutnothing', 'comedy', '1598'],\n",
       "       ['tamingoftheshrew', 'comedy', '1590'],\n",
       "       ['asyoulikeit', 'comedy', '1599'],\n",
       "       ['merrywivesofwindsor', 'comedy', '1600'],\n",
       "       ['midsummersnightsdream', 'comedy', '1595'],\n",
       "       ['romeoandjuliet', 'tragedy', '1592'],\n",
       "       ['comedyoferrors', 'comedy', '1594'],\n",
       "       ['twogentlemenofverona', 'comedy', '1589']], \n",
       "      dtype='|S21')"
      ]
     },
     "execution_count": 100,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "lookup[ km_topics.labels_ == 0]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 101,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([['titusandronicus', 'tragedy', '1591'],\n",
       "       ['allswellthatendswell', 'comedy', '1604'],\n",
       "       ['macbeth', 'tragedy', '1606'],\n",
       "       ['hamlet', 'tragedy', '1599'],\n",
       "       ['antonyandcleopatra', 'tragedy', '1606'],\n",
       "       ['kinglear', 'tragedy', '1605'],\n",
       "       ['juliuscaesar', 'tragedy', '1599'],\n",
       "       ['tempest', 'comedy', '1610'],\n",
       "       ['winterstale', 'comedy', '1609'],\n",
       "       ['timonofathens', 'tragedy', '1605'],\n",
       "       ['coriolanus', 'tragedy', '1607'],\n",
       "       ['troilusandcressida', 'comedy', '1600'],\n",
       "       ['measureforemeasure', 'comedy', '1603'],\n",
       "       ['cymbeline', 'comedy', '1610'],\n",
       "       ['othello', 'tragedy', '1604'],\n",
       "       ['periclesprinceoftyre', 'comedy', '1607']], \n",
       "      dtype='|S21')"
      ]
     },
     "execution_count": 101,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "lookup[km_topics.labels_ == 1]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 102,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "lookup_df = pd.DataFrame({'title':titles,\n",
    "                          'label':labels,\n",
    "                          'year':year,\n",
    "                          'cluster':km_topics.labels_\n",
    "                         })\n",
    "groups = set(km_topics.labels_)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 103,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "lookup_df.to_csv('outputs/lookup_titles.csv')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "However, experts do agree that Shakespeare's style has changed over the years.\n",
    "\n",
    "More specifically, they talk about an elizabethan Shakespeare, more influenced by the classics, and a more mature Shakespeare, writing under the reign of King James. So looking at the groups and the year the plays were written, group 1 includes only plays that were written before the death of elizabeth the first in 1603 and group 2 includes all the plays written for James the first (except for one)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 104,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<seaborn.axisgrid.FacetGrid at 0x15865cc10>"
      ]
     },
     "execution_count": 104,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAARkAAAEZCAYAAACjEFEXAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nETFKkRERoxQZETFKkRERoxQZETHq/wHfJRxFLKKXRgAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x169b521d0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "sns.set(style=\"ticks\")\n",
    "\n",
    "g = sns.factorplot(x = 'cluster', y = 'year', data = lookup_df, kind='box')\n",
    "g.despine()\n",
    "g.set_axis_labels('Cluster', 'Year')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Conclusion"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "It was possible to create a two way split of shakespeare's works using nlp, the groups obtained this way don't reflect so much a difference in genre but rather a change in the style of Shakespeare's writing."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## LDA with gensim"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 105,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "ldamodel = models.ldamodel.LdaModel(corpus, num_topics=12, id2word=words, passes=10)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 106,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "titusandronicus [(0, 0.68614029418648959), (1, 0.3019738160848664), (8, 0.011811388172448825)]\n",
      "allswellthatendswell [(1, 0.9999146003555931)]\n",
      "twelfthnight [(1, 0.99872629638407606)]\n",
      "macbeth [(1, 0.99935582758058461)]\n",
      "hamlet [(1, 0.99763621610733388)]\n",
      "antonyandcleopatra [(1, 0.9980869854649228)]\n",
      "kinglear [(1, 0.9998170638663898)]\n",
      "juliuscaesar [(1, 0.024338987214502183), (8, 0.97556715724650556)]\n",
      "tempest [(0, 0.98247235042340564), (1, 0.017422747269538946)]\n",
      "merchantofvenice [(1, 0.99753825701373688)]\n",
      "loveslabourslost [(1, 0.99770636799763701)]\n",
      "winterstale [(1, 0.99974402556734432)]\n",
      "timonofathens [(1, 0.92669277939287897), (11, 0.073219002861269394)]\n",
      "muchadoaboutnothing [(1, 0.29346597180710948), (8, 0.70624831716751002)]\n",
      "coriolanus [(1, 0.98809163687122326), (8, 0.011715538083812364)]\n",
      "troilusandcressida [(1, 0.99197982734007883)]\n",
      "measureforemeasure [(1, 0.99987459366162001)]\n",
      "cymbeline [(1, 0.99920666825092608)]\n",
      "tamingoftheshrew [(1, 0.99581954398361483)]\n",
      "asyoulikeit [(11, 0.99550928887022672)]\n",
      "merrywivesofwindsor [(1, 0.99990556315542778)]\n",
      "midsummersnightsdream [(1, 0.10477753033432659), (6, 0.74286530649253668), (8, 0.1492277168711045)]\n",
      "othello [(1, 0.99849111751089026)]\n",
      "romeoandjuliet [(1, 0.011499837766398546), (8, 0.98842937190770463)]\n",
      "comedyoferrors [(1, 0.97457444349931355), (8, 0.025301731348655137)]\n",
      "twogentlemenofverona [(1, 0.98577625404886615), (8, 0.014049960118439345)]\n",
      "periclesprinceoftyre [(1, 0.99872278988586238)]\n"
     ]
    }
   ],
   "source": [
    "for i in range(len(titles)):\n",
    "    print titles[i], ldamodel.get_document_topics(corpus[i])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 107,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "document_topics = []\n",
    "for i in range(len(corpus)):\n",
    "    document_topics.append(\n",
    "        [x[1] for x in ldamodel.get_document_topics(corpus[i], minimum_probability=0)]\n",
    "        )"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 108,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "KMeans(copy_x=True, init='random', max_iter=300, n_clusters=2, n_init=10,\n",
       "    n_jobs=1, precompute_distances='auto', random_state=10, tol=0.0001,\n",
       "    verbose=0)"
      ]
     },
     "execution_count": 108,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "km_topics_gensim_input = KMeans(n_clusters = 2, init = 'random', random_state=10)\n",
    "km_topics_gensim_input.fit(np.array(document_topics))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 109,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([0, 1, 1, 1, 1, 1, 1, 0, 0, 1, 1, 1, 1, 0, 1, 1, 1, 1, 1, 0, 1, 0, 1,\n",
       "       0, 1, 1, 1], dtype=int32)"
      ]
     },
     "execution_count": 109,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "km_topics_gensim_input.labels_"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 110,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([['allswellthatendswell', 'comedy', '1604'],\n",
       "       ['twelfthnight', 'comedy', '1601'],\n",
       "       ['macbeth', 'tragedy', '1606'],\n",
       "       ['hamlet', 'tragedy', '1599'],\n",
       "       ['antonyandcleopatra', 'tragedy', '1606'],\n",
       "       ['kinglear', 'tragedy', '1605'],\n",
       "       ['merchantofvenice', 'comedy', '1596'],\n",
       "       ['loveslabourslost', 'comedy', '1594'],\n",
       "       ['winterstale', 'comedy', '1609'],\n",
       "       ['timonofathens', 'tragedy', '1605'],\n",
       "       ['coriolanus', 'tragedy', '1607'],\n",
       "       ['troilusandcressida', 'comedy', '1600'],\n",
       "       ['measureforemeasure', 'comedy', '1603'],\n",
       "       ['cymbeline', 'comedy', '1610'],\n",
       "       ['tamingoftheshrew', 'comedy', '1590'],\n",
       "       ['merrywivesofwindsor', 'comedy', '1600'],\n",
       "       ['othello', 'tragedy', '1604'],\n",
       "       ['comedyoferrors', 'comedy', '1594'],\n",
       "       ['twogentlemenofverona', 'comedy', '1589'],\n",
       "       ['periclesprinceoftyre', 'comedy', '1607']], \n",
       "      dtype='|S21')"
      ]
     },
     "execution_count": 110,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "lookup[km_topics_gensim_input.labels_ == 1]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 111,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([['titusandronicus', 'tragedy', '1591'],\n",
       "       ['juliuscaesar', 'tragedy', '1599'],\n",
       "       ['tempest', 'comedy', '1610'],\n",
       "       ['muchadoaboutnothing', 'comedy', '1598'],\n",
       "       ['asyoulikeit', 'comedy', '1599'],\n",
       "       ['midsummersnightsdream', 'comedy', '1595'],\n",
       "       ['romeoandjuliet', 'tragedy', '1592']], \n",
       "      dtype='|S21')"
      ]
     },
     "execution_count": 111,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "lookup[km_topics_gensim_input.labels_ == 0]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 112,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "lookup_df = pd.DataFrame({'title':titles,\n",
    "                          'label':labels,\n",
    "                          'year':year,\n",
    "                          'cluster':km_topics_gensim_input.labels_\n",
    "                         })\n",
    "groups = set(km_topics_gensim_input.labels_)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 113,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<seaborn.axisgrid.FacetGrid at 0x158a03150>"
      ]
     },
     "execution_count": 113,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAARkAAAEZCAYAAACjEFEXAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\njIgYpciIiFH/D7/SEhWgcuQNAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x159530310>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "sns.set(style=\"ticks\")\n",
    "\n",
    "g = sns.factorplot(x = 'cluster', y = 'year', data = lookup_df, kind='box')\n",
    "g.despine()\n",
    "g.set_axis_labels('Cluster', 'Year')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 2",
   "language": "python",
   "name": "python2"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 2
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython2",
   "version": "2.7.10"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 0
}