topic_modeling.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Topic modeling of Svevo Letter Corpus"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Imports"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "# General\n",
    "import string\n",
    "import numpy as np\n",
    "import pandas as pd\n",
    "import warnings\n",
    "\n",
    "# Gensim\n",
    "import gensim\n",
    "from gensim import corpora, models\n",
    "from gensim.utils import simple_preprocess\n",
    "from gensim.models import CoherenceModel, KeyedVectors\n",
    "from gensim.models.wrappers import LdaMallet\n",
    "from gensim.test.utils import datapath\n",
    "\n",
    "# Lemmatization\n",
    "import spacy\n",
    "\n",
    "# Silhouette index and shuffle method\n",
    "from sklearn.metrics import silhouette_samples, silhouette_score\n",
    "from sklearn.utils import shuffle\n",
    "\n",
    "# Plotting tools\n",
    "import pyLDAvis\n",
    "import pyLDAvis.gensim\n",
    "import matplotlib.pyplot as plt\n",
    "import matplotlib.cm as cm\n",
    "%matplotlib inline"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Global variables and settings"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Ignore warnings in libraries\n",
    "warnings.filterwarnings(\"ignore\")\n",
    "\n",
    "# Import Italo Svevo corpus dataset\n",
    "data = pd.read_csv('datasets/carteggio.svevo3.csv', sep=';', parse_dates=['date'])\n",
    "\n",
    "# Spacy languages\n",
    "# Load modules using sudo python -m spacy download <lang>\n",
    "nlp_en = spacy.load(\"en\", disable=['parser', 'ner'])\n",
    "nlp_it = spacy.load(\"it\", disable=['parser', 'ner'])\n",
    "nlp_de = spacy.load(\"de\", disable=['parser', 'ner'])\n",
    "nlp_fr = spacy.load(\"fr\", disable=['parser', 'ner'])\n",
    "\n",
    "# Language dictionary for scipy language format\n",
    "dict_spacy = { \"ENG\" : nlp_en, \"ITA\" : nlp_it, \"GER\" : nlp_de, \"FRE\" : nlp_fr}\n",
    "\n",
    "# Custom stopwords for our purposes\n",
    "stop_words = set(['schmitz', 'signore', 'signora', 'mano', 'ettore', 'lettera', 'parola', 'fare', 'cosa'])\n",
    "\n",
    "# Enables pyLDAvis visualization for the notebook\n",
    "pyLDAvis.enable_notebook()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Custom NLP toolbox"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "def tokenize(texts):\n",
    "    \"\"\"\n",
    "    Tokenizes a list of texts, returning a list of tokens.\n",
    "    \"\"\"\n",
    "    token_texts = []\n",
    "    for text in texts:\n",
    "        token_texts.append(simple_preprocess(text, deacc=True))\n",
    "    return token_texts\n",
    "     \n",
    "def preprocess_text(text, nlp, stop_words = [], pos = ['PROPN', 'NOUN', 'VERB', 'ADJ']):\n",
    "    \"\"\"\n",
    "    Performs preprocessing on a text through spaCy, langauge based on nlp parameter.\n",
    "    Filter default stopword plus those in the list stop_words,\n",
    "    accepts only word with POS tag in the pos list,\n",
    "    filters out non-alpha words, performs lemmatization.\n",
    "    Returns a list of lemmatized, filtered tokens for the text.\n",
    "    \"\"\"\n",
    "    doc = nlp(text.lower())\n",
    "    tokens = []\n",
    "    for token in doc:\n",
    "        if token.lemma_ not in stop_words and token.pos_ in pos and not token.is_stop and token.is_alpha:\n",
    "            tokens.append(token.lemma_)\n",
    "    return tokens\n",
    "\n",
    "def create_lda_model(dictionary, corpus, num_topics, num_words = 20, passes = 20):\n",
    "    \"\"\"\n",
    "    Creates an LDA model using gensim.\n",
    "    Dictionary and corpus should be obtained with corpora.\n",
    "    Number of topics and number of passes should be tweaked depending on context.\n",
    "    \"\"\"\n",
    "    lda_model = models.LdaModel(corpus, num_topics=num_topics, id2word=dictionary, passes=passes, random_state=42)\n",
    "    for topic,words in lda_model.show_topics(formatted=True, num_topics=num_topics, num_words=num_words):\n",
    "        print(str(topic)+ \": \" + words + \"\\n\")\n",
    "    return lda_model\n",
    "\n",
    "def get_lda_keywords(model, num_words = 20):\n",
    "    \"\"\"\n",
    "    Extract the firt num_words keywords from all topics of a gensim LDA model.\n",
    "    \"\"\"\n",
    "    num_topics = len(model.print_topics())\n",
    "    return [[tup2[0] for tup2 in tup[1]]for tup in model.show_topics(formatted=False, num_topics=num_topics, num_words=num_words)]\n",
    "    \n",
    "def create_paper_points_data(lda_model, corpus, num_topics):\n",
    "    \"\"\"\n",
    "    Creates a dataframe with topic scores for each text in corpus using a gensim LDA model.\n",
    "    \"\"\"\n",
    "    x = [[] for i in range(num_topics)]\n",
    "    for text in corpus:\n",
    "        l_tup = lda_model[text]\n",
    "        for val in range(num_topics):\n",
    "            added = False\n",
    "            for tup in l_tup:\n",
    "                if tup[0] == val:\n",
    "                    x[val].append(tup[1])\n",
    "                    added = True\n",
    "            if added == False:\n",
    "                x[val].append(0)\n",
    "    df = pd.DataFrame(x)\n",
    "    df = df.transpose()\n",
    "    df.columns = ['Topic' + str(n) for n in range(num_topics)]\n",
    "    return df\n",
    "\n",
    "def compute_silhouette(lda_model, corpus, num_topics):\n",
    "    \"\"\"\n",
    "    Computes silhouette index for an LDA model based on topics it classified.\n",
    "    \"\"\"\n",
    "    points = create_paper_points_data(lda_model, corpus, num_topics).values\n",
    "    lda_labels = points.argmax(axis=1) \n",
    "    return silhouette_score(points, lda_labels)\n",
    "    \n",
    "def compute_perplexity(lda_model, corpus):\n",
    "    \"\"\"\n",
    "    Computes perplexity score for an LDA model.\n",
    "    \"\"\"\n",
    "    perplexity = lda_model.log_perplexity(corpus)\n",
    "    return perplexity\n",
    "\n",
    "def compute_coherence(lda_model, texts, corpus, dictionary, coherence = \"c_v\"):\n",
    "    \"\"\"\n",
    "    Computes coherence score for an LDA model.\n",
    "    \"\"\"\n",
    "    score = CoherenceModel(model=lda_model, texts=texts, corpus=corpus, dictionary=dictionary, coherence=coherence).get_coherence()\n",
    "    return score\n",
    "\n",
    "def assess_model_goodness(lda_model, texts, corpus, dictionary, num_topics):\n",
    "    \"\"\"\n",
    "    Assess the goodness of an LDA model using coherence and silhouette scores.\n",
    "    \"\"\"\n",
    "    points = create_paper_points_data(lda_model, corpus, num_topics).values\n",
    "    lda_labels = points.argmax(axis=1)\n",
    "    fig, ax1 = plt.subplots(1, 1)\n",
    "    fig.set_size_inches(18, 7)\n",
    "    ax1.set_xlim([-0.1, 1])\n",
    "    ax1.set_ylim([0, len(points) + (n + 1) * 10])\n",
    "    silhouette_avg = silhouette_score(points, lda_labels)\n",
    "    print(\"For {} topic the average silhouette score is : {}\\n\".format(num_topics, silhouette_avg))\n",
    "    #print(\"Perplexity: {}\\n\".format(compute_perplexity(lda_model, corpus)))\n",
    "    print(\"Coherence: {}\\n\".format(compute_coherence(lda_model, texts, corpus, dictionary, \"c_v\")))\n",
    "    sample_silhouette_values = silhouette_samples(points, lda_labels)\n",
    "    y_lower = 10\n",
    "    for i in range(num_topics):\n",
    "        ith_topic_silhouette_values = sample_silhouette_values[lda_labels == i]\n",
    "        ith_topic_silhouette_values.sort()\n",
    "        size_topic_i = ith_topic_silhouette_values.shape[0]\n",
    "        y_upper = y_lower + size_topic_i\n",
    "        color = cm.nipy_spectral(float(i) / num_topics)\n",
    "        ax1.fill_betweenx(np.arange(y_lower, y_upper),\n",
    "                          0, ith_topic_silhouette_values,\n",
    "                          facecolor=color, edgecolor=color, alpha=0.7)\n",
    "        ax1.text(-0.05, y_lower + 0.5 * size_topic_i, str(i))\n",
    "        y_lower = y_upper + 10\n",
    "    ax1.set_title(\"Silhouette plot for \" + str(num_topics) + \" topics.\")\n",
    "    ax1.set_xlabel(\"Silhouette coefficient\")\n",
    "    ax1.set_ylabel(\"Topic number\")\n",
    "    ax1.axvline(x=silhouette_avg, color=\"red\", linestyle=\"--\")\n",
    "    ax1.set_yticks([])\n",
    "    ax1.set_xticks([-0.1, 0, 0.2, 0.4, 0.6, 0.8, 1])\n",
    "    plt.show()\n",
    "\n",
    "def assess_model_certainity(lda_model, corpus, num_topics, low_bound = 0.20, up_bound = 0.80):\n",
    "    \"\"\"\n",
    "    Assess the certainity of assignation of textes in a corpus to a certain topic based on pre-defined bounds.\n",
    "    \"\"\"\n",
    "    n = 0\n",
    "    c_topics = []\n",
    "    f_topics = []\n",
    "    f_topics_bad = []\n",
    "    f_topics_good = []\n",
    "    sum_bad = 0\n",
    "    for text in corpus:\n",
    "        c_topics.append((lda_model[text], n))\n",
    "        n += 1 \n",
    "    print(\"Total texts:\", len(c_topics))\n",
    "    for n in range(num_topics):\n",
    "        f_topics.append([[(tup[1], t[0], t[1]) for t in tup[0] if t[0] == n] for tup in c_topics])\n",
    "        f_topics[n] = [x for x in f_topics[n] if x != []]\n",
    "        print(\"Texts with topic\", n, \"assigned:\", len(f_topics[n]))\n",
    "        f_topics_bad.append([x for x in f_topics[n] if x[0][2] > low_bound and x[0][2] < up_bound])\n",
    "        print(\"Texts with topic\", n, \" uncertainly assigned:\", len(f_topics_bad[n]))\n",
    "        f_topics_good.append([x for x in f_topics[n] if x[0][2] > up_bound])\n",
    "        print(\"Texts with topic\", n, \" prevalent:\", len(f_topics_good[n]))        \n",
    "        sum_bad += len(f_topics_bad[n])\n",
    "    print(\"Total uncertain texts:\", sum_bad)\n",
    "    print(\"Uncertainty percentage:\", float(sum_bad) / float(len(c_topics)))\n",
    "    good_text_ids = [[x[0][0] for x in topic] for topic in f_topics_good]\n",
    "    bad_text_ids = [[x[0][0] for x in topic] for topic in f_topics_bad]\n",
    "    return good_text_ids, bad_text_ids"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# N-grams custom toolbox\n",
    "\n",
    "### Not in use, not working nor implemented \n",
    "\n",
    "This was disregarded because of the difficult integration with spaCy, which doesn't supports POS tagging for ngrams and thus can include lots of stopwords and non-alpha words in our dictionary."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "def remove_stopwords(text, language, stop_words = []):\n",
    "    nlp = spacy.load(language, disable=['parser', 'ner'])\n",
    "    doc = nlp(text.lower())\n",
    "    filtered = []\n",
    "    for token in doc:\n",
    "        if token.text not in stop_words and token.is_alpha and not token.is_stop:\n",
    "            filtered.append(token)\n",
    "    return \" \".join([token.text for token in filtered])\n",
    "\n",
    "def lemmatize_pos_filter(text, language, pos):\n",
    "    nlp = spacy.load(language, disable=['parser', 'ner'])\n",
    "    doc = nlp(text.lower())\n",
    "    filtered = []\n",
    "    for token in doc:\n",
    "        if token.pos_ in pos:\n",
    "            filtered.append(token)\n",
    "    return [token.lemma_ for token in filtered]\n",
    "    \n",
    "def generate_ngrams(text, bigram, trigram, is_token = False):\n",
    "    if is_token:\n",
    "        token_texts = text\n",
    "    else:\n",
    "        token_texts = tokenize(text)\n",
    "    return trigram[bigram[token_texts]]\n",
    "    \n",
    "def preprocess_text_ngrams(text, language, stop_words = [], pos = ['PROPN', 'NOUN', 'VERB', 'ADJ'], ngrams = False, bigram = None, trigram = None):\n",
    "    text = remove_stopwords(text, language, stop_words)\n",
    "    if ngrams and bigram and trigram:\n",
    "        tokens = generate_ngrams(text, bigram, trigram)\n",
    "    print(tokens.corpus)\n",
    "    return lemmatize_pos_filter(\" \".join(tokens), language, pos)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Data exploration"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>n</th>\n",
       "      <th>corpus</th>\n",
       "      <th>index</th>\n",
       "      <th>date</th>\n",
       "      <th>year</th>\n",
       "      <th>sender</th>\n",
       "      <th>senderLocation</th>\n",
       "      <th>recipient</th>\n",
       "      <th>recipientLocation</th>\n",
       "      <th>languages</th>\n",
       "      <th>mainLanguage</th>\n",
       "      <th>text</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0</td>\n",
       "      <td>Schmitz Joyce</td>\n",
       "      <td>1</td>\n",
       "      <td>08/02/1909</td>\n",
       "      <td>1909</td>\n",
       "      <td>Ettore Schmitz</td>\n",
       "      <td>Trieste</td>\n",
       "      <td>James Joyce</td>\n",
       "      <td>Trieste</td>\n",
       "      <td>ENG</td>\n",
       "      <td>ENG</td>\n",
       "      <td>Trieste, 8th. of February 1909 Dear Mr. Joyce,...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1</td>\n",
       "      <td>Schmitz Joyce</td>\n",
       "      <td>2</td>\n",
       "      <td>15/06/1910</td>\n",
       "      <td>1910</td>\n",
       "      <td>Ettore Schmitz</td>\n",
       "      <td>Charlton</td>\n",
       "      <td>James Joyce</td>\n",
       "      <td>Trieste</td>\n",
       "      <td>ENG</td>\n",
       "      <td>ENG</td>\n",
       "      <td>Charlton, 15 June  1910 My dear Mr.  Joyce, I ...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>2</td>\n",
       "      <td>Schmitz Joyce</td>\n",
       "      <td>3</td>\n",
       "      <td>26/07/1912</td>\n",
       "      <td>1912</td>\n",
       "      <td>James Joyce</td>\n",
       "      <td>Galway</td>\n",
       "      <td>Ettore Schmitz</td>\n",
       "      <td>Trieste</td>\n",
       "      <td>ENG,ITA</td>\n",
       "      <td>ENG</td>\n",
       "      <td>Hector   Schmitz   Esq[uire] presso la Fabbric...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>3</td>\n",
       "      <td>Schmitz Joyce</td>\n",
       "      <td>4</td>\n",
       "      <td>26/06/1914</td>\n",
       "      <td>1914</td>\n",
       "      <td>Ettore Schmitz</td>\n",
       "      <td>Trieste</td>\n",
       "      <td>James Joyce</td>\n",
       "      <td>Parigi</td>\n",
       "      <td>ENG</td>\n",
       "      <td>ENG</td>\n",
       "      <td>Trieste, 26 June 1914 Dear Mr. Joyce, Many tha...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>4</td>\n",
       "      <td>Schmitz Joyce</td>\n",
       "      <td>5</td>\n",
       "      <td>05/01/1921</td>\n",
       "      <td>1921</td>\n",
       "      <td>James Joyce</td>\n",
       "      <td>Parigi</td>\n",
       "      <td>Ettore Schmitz</td>\n",
       "      <td>Trieste</td>\n",
       "      <td>ITA,UND</td>\n",
       "      <td>ITA</td>\n",
       "      <td>Boulevard Raspail 5 Parigi VII Caro signor Sch...</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   n         corpus  index        date  year          sender senderLocation  \\\n",
       "0  0  Schmitz Joyce      1  08/02/1909  1909  Ettore Schmitz        Trieste   \n",
       "1  1  Schmitz Joyce      2  15/06/1910  1910  Ettore Schmitz       Charlton   \n",
       "2  2  Schmitz Joyce      3  26/07/1912  1912     James Joyce         Galway   \n",
       "3  3  Schmitz Joyce      4  26/06/1914  1914  Ettore Schmitz        Trieste   \n",
       "4  4  Schmitz Joyce      5  05/01/1921  1921     James Joyce         Parigi   \n",
       "\n",
       "        recipient recipientLocation languages mainLanguage  \\\n",
       "0     James Joyce           Trieste       ENG          ENG   \n",
       "1     James Joyce           Trieste       ENG          ENG   \n",
       "2  Ettore Schmitz           Trieste   ENG,ITA          ENG   \n",
       "3     James Joyce            Parigi       ENG          ENG   \n",
       "4  Ettore Schmitz           Trieste   ITA,UND          ITA   \n",
       "\n",
       "                                                text  \n",
       "0  Trieste, 8th. of February 1909 Dear Mr. Joyce,...  \n",
       "1  Charlton, 15 June  1910 My dear Mr.  Joyce, I ...  \n",
       "2  Hector   Schmitz   Esq[uire] presso la Fabbric...  \n",
       "3  Trieste, 26 June 1914 Dear Mr. Joyce, Many tha...  \n",
       "4  Boulevard Raspail 5 Parigi VII Caro signor Sch...  "
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'pandas.core.frame.DataFrame'>\n",
      "RangeIndex: 894 entries, 0 to 893\n",
      "Data columns (total 12 columns):\n",
      "n                    894 non-null int64\n",
      "corpus               894 non-null object\n",
      "index                894 non-null int64\n",
      "date                 894 non-null object\n",
      "year                 894 non-null int64\n",
      "sender               894 non-null object\n",
      "senderLocation       894 non-null object\n",
      "recipient            894 non-null object\n",
      "recipientLocation    894 non-null object\n",
      "languages            894 non-null object\n",
      "mainLanguage         894 non-null object\n",
      "text                 894 non-null object\n",
      "dtypes: int64(3), object(9)\n",
      "memory usage: 83.9+ KB\n"
     ]
    }
   ],
   "source": [
    "data.info()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>n</th>\n",
       "      <th>corpus</th>\n",
       "      <th>index</th>\n",
       "      <th>date</th>\n",
       "      <th>year</th>\n",
       "      <th>sender</th>\n",
       "      <th>senderLocation</th>\n",
       "      <th>recipient</th>\n",
       "      <th>recipientLocation</th>\n",
       "      <th>languages</th>\n",
       "      <th>mainLanguage</th>\n",
       "      <th>text</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>count</th>\n",
       "      <td>894.000000</td>\n",
       "      <td>894</td>\n",
       "      <td>894.000000</td>\n",
       "      <td>894</td>\n",
       "      <td>894.000000</td>\n",
       "      <td>894</td>\n",
       "      <td>894</td>\n",
       "      <td>894</td>\n",
       "      <td>894</td>\n",
       "      <td>894</td>\n",
       "      <td>894</td>\n",
       "      <td>894</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>unique</th>\n",
       "      <td>NaN</td>\n",
       "      <td>45</td>\n",
       "      <td>NaN</td>\n",
       "      <td>827</td>\n",
       "      <td>NaN</td>\n",
       "      <td>8</td>\n",
       "      <td>38</td>\n",
       "      <td>52</td>\n",
       "      <td>34</td>\n",
       "      <td>30</td>\n",
       "      <td>4</td>\n",
       "      <td>894</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>top</th>\n",
       "      <td>NaN</td>\n",
       "      <td>Schmitz Livia</td>\n",
       "      <td>NaN</td>\n",
       "      <td>21/11/1925</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Ettore Schmitz</td>\n",
       "      <td>Trieste</td>\n",
       "      <td>Livia Veneziani</td>\n",
       "      <td>Trieste</td>\n",
       "      <td>ITA</td>\n",
       "      <td>ITA</td>\n",
       "      <td>Alla moglie [Murano,]  1.3.901 Carissima Livia...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>freq</th>\n",
       "      <td>NaN</td>\n",
       "      <td>639</td>\n",
       "      <td>NaN</td>\n",
       "      <td>3</td>\n",
       "      <td>NaN</td>\n",
       "      <td>816</td>\n",
       "      <td>310</td>\n",
       "      <td>634</td>\n",
       "      <td>517</td>\n",
       "      <td>515</td>\n",
       "      <td>826</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>mean</th>\n",
       "      <td>446.500000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>22.399329</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1910.778523</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>std</th>\n",
       "      <td>258.219868</td>\n",
       "      <td>NaN</td>\n",
       "      <td>18.241949</td>\n",
       "      <td>NaN</td>\n",
       "      <td>10.506871</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>min</th>\n",
       "      <td>0.000000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1885.000000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25%</th>\n",
       "      <td>223.250000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>8.000000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1901.000000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>50%</th>\n",
       "      <td>446.500000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>17.500000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1908.000000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>75%</th>\n",
       "      <td>669.750000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>32.000000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1922.000000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>max</th>\n",
       "      <td>893.000000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>83.000000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1928.000000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                 n         corpus       index        date         year  \\\n",
       "count   894.000000            894  894.000000         894   894.000000   \n",
       "unique         NaN             45         NaN         827          NaN   \n",
       "top            NaN  Schmitz Livia         NaN  21/11/1925          NaN   \n",
       "freq           NaN            639         NaN           3          NaN   \n",
       "mean    446.500000            NaN   22.399329         NaN  1910.778523   \n",
       "std     258.219868            NaN   18.241949         NaN    10.506871   \n",
       "min       0.000000            NaN    1.000000         NaN  1885.000000   \n",
       "25%     223.250000            NaN    8.000000         NaN  1901.000000   \n",
       "50%     446.500000            NaN   17.500000         NaN  1908.000000   \n",
       "75%     669.750000            NaN   32.000000         NaN  1922.000000   \n",
       "max     893.000000            NaN   83.000000         NaN  1928.000000   \n",
       "\n",
       "                sender senderLocation        recipient recipientLocation  \\\n",
       "count              894            894              894               894   \n",
       "unique               8             38               52                34   \n",
       "top     Ettore Schmitz        Trieste  Livia Veneziani           Trieste   \n",
       "freq               816            310              634               517   \n",
       "mean               NaN            NaN              NaN               NaN   \n",
       "std                NaN            NaN              NaN               NaN   \n",
       "min                NaN            NaN              NaN               NaN   \n",
       "25%                NaN            NaN              NaN               NaN   \n",
       "50%                NaN            NaN              NaN               NaN   \n",
       "75%                NaN            NaN              NaN               NaN   \n",
       "max                NaN            NaN              NaN               NaN   \n",
       "\n",
       "       languages mainLanguage  \\\n",
       "count        894          894   \n",
       "unique        30            4   \n",
       "top          ITA          ITA   \n",
       "freq         515          826   \n",
       "mean         NaN          NaN   \n",
       "std          NaN          NaN   \n",
       "min          NaN          NaN   \n",
       "25%          NaN          NaN   \n",
       "50%          NaN          NaN   \n",
       "75%          NaN          NaN   \n",
       "max          NaN          NaN   \n",
       "\n",
       "                                                     text  \n",
       "count                                                 894  \n",
       "unique                                                894  \n",
       "top     Alla moglie [Murano,]  1.3.901 Carissima Livia...  \n",
       "freq                                                    1  \n",
       "mean                                                  NaN  \n",
       "std                                                   NaN  \n",
       "min                                                   NaN  \n",
       "25%                                                   NaN  \n",
       "50%                                                   NaN  \n",
       "75%                                                   NaN  \n",
       "max                                                   NaN  "
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data.describe(include = 'all')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "mainLanguage\n",
       "ENG     10\n",
       "FRE     30\n",
       "GER     28\n",
       "ITA    826\n",
       "dtype: int64"
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# We see that most of the texts are in italian\n",
    "data.groupby(\"mainLanguage\").size()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "corpus\n",
       "Schmitz Livia           639\n",
       "Svevo Montale            62\n",
       "Svevo Comnène            30\n",
       "Svevo Crémieux           19\n",
       "Schmitz Joyce            19\n",
       "Schmitz Henri Michel     17\n",
       "Svevo Jahier             15\n",
       "Svevo Larbaud            15\n",
       "Schmitz Prezzolini       12\n",
       "Schmitz Pasini           11\n",
       "Schmitz Letizia           5\n",
       "Schmitz Ottavio           5\n",
       "Schmitz Benco             3\n",
       "Schmitz Tribel            3\n",
       "Schmitz Treves            3\n",
       "dtype: int64"
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# The main corpus is Schmitz Livia, between Svevo and his wife Livia Veneziani.\n",
    "data.groupby(\"corpus\").size().sort_values(ascending=False)[:15]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Tokenized texts, corpus and dictionaries"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Despite creating dictionaries and corpus for all the languages, we finished by focusing on the italian letters only since the others were not enough to perform a decent analysis with our methods."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Schmitz Livia letter corpus dataframe, corpus and dictionary\n",
    "data_livia = data[data.corpus == \"Schmitz Livia\"]\n",
    "texts_livia = [preprocess_text(text, nlp_it, stop_words=stop_words, pos = ['PROPN', 'NOUN']) for text in data_livia[\"text\"]]\n",
    "d_livia = corpora.Dictionary(texts_livia)\n",
    "c_livia = [d_livia.doc2bow(text) for text in texts_livia]\n",
    "\n",
    "# Italian letters dataframe, corpus and dictionary\n",
    "data_it = data[data.mainLanguage == \"ITA\"]\n",
    "texts_it = [preprocess_text(text, nlp_it, stop_words=stop_words, pos = ['PROPN', 'NOUN']) for text in data_it[\"text\"]]\n",
    "d_it = corpora.Dictionary(texts_it)\n",
    "d_it.filter_extremes(no_below=5, no_above=0.5)\n",
    "c_it = [d_it.doc2bow(text) for text in texts_it]\n",
    "\n",
    "# English letters dataframe, corpus and dictionary\n",
    "data_en = data[data.mainLanguage == \"ENG\"]\n",
    "texts_en = [preprocess_text(text, nlp_en, stop_words=stop_words, pos = ['PROPN', 'NOUN']) for text in data_en[\"text\"]]\n",
    "d_en = corpora.Dictionary(texts_en)\n",
    "c_en = [d_en.doc2bow(text) for text in texts_en]\n",
    "\n",
    "# French letters dataframe, corpus and dictionary\n",
    "data_fr = data[data.mainLanguage == \"FRE\"]\n",
    "texts_fr = [preprocess_text(text, nlp_fr, stop_words=stop_words, pos = ['PROPN', 'NOUN']) for text in data_fr[\"text\"]]\n",
    "d_fr = corpora.Dictionary(texts_fr)\n",
    "c_fr = [d_fr.doc2bow(text) for text in texts_fr]\n",
    "\n",
    "# German letters dataframe, corpus and dictionary\n",
    "data_de = data[data.mainLanguage == \"GER\"]\n",
    "texts_de = [preprocess_text(text, nlp_de, stop_words=stop_words, pos = ['PROPN', 'NOUN']) for text in data_de[\"text\"]]\n",
    "d_de = corpora.Dictionary(texts_de)\n",
    "c_de = [d_de.doc2bow(text) for text in texts_de]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Topic Modeling with LDA\n",
    "\n",
    "We then created a range of LDA models with 1-10 topics and assessed their goodness using our custom evaluation functions."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0: 0.018*\"cuore\" + 0.013*\"sera\" + 0.010*\"trieste\" + 0.010*\"venezia\" + 0.010*\"livia\" + 0.010*\"marcare\" + 0.010*\"baciare\" + 0.009*\"notte\" + 0.008*\"mattina\" + 0.008*\"fabbricare\" + 0.008*\"domenica\" + 0.008*\"letizia\" + 0.007*\"olga\" + 0.007*\"abbracciare\" + 0.007*\"viaggiare\" + 0.006*\"leggere\" + 0.006*\"notizia\" + 0.006*\"settimana\" + 0.006*\"sabato\" + 0.006*\"bisognare\"\n",
      "\n",
      "1: 0.017*\"amicare\" + 0.017*\"senilità\" + 0.015*\"librare\" + 0.015*\"parigi\" + 0.012*\"trieste\" + 0.011*\"romanzare\" + 0.009*\"articolare\" + 0.009*\"joyce\" + 0.008*\"villa\" + 0.008*\"scrivere\" + 0.008*\"zeno\" + 0.007*\"traduzione\" + 0.007*\"maestro\" + 0.007*\"coscienza\" + 0.007*\"larbaud\" + 0.006*\"saluto\" + 0.006*\"libro\" + 0.006*\"piacere\" + 0.006*\"italia\" + 0.006*\"copiare\"\n",
      "\n",
      "For 2 topic the average silhouette score is : 0.9053335659430272\n",
      "\n",
      "Coherence: 0.5687491659414248\n",
      "\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1296x504 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0: 0.012*\"cuore\" + 0.012*\"venezia\" + 0.011*\"baciare\" + 0.010*\"sera\" + 0.009*\"livia\" + 0.008*\"bisognare\" + 0.008*\"olga\" + 0.008*\"marcare\" + 0.007*\"abbracciare\" + 0.007*\"addio\" + 0.007*\"notte\" + 0.006*\"viaggiare\" + 0.006*\"letizia\" + 0.006*\"mattina\" + 0.006*\"trieste\" + 0.006*\"curare\" + 0.006*\"ritornare\" + 0.006*\"salso\" + 0.005*\"piacere\" + 0.005*\"notizia\"\n",
      "\n",
      "1: 0.018*\"amicare\" + 0.018*\"senilità\" + 0.016*\"parigi\" + 0.015*\"librare\" + 0.012*\"romanzare\" + 0.012*\"trieste\" + 0.010*\"articolare\" + 0.009*\"joyce\" + 0.008*\"zeno\" + 0.008*\"villa\" + 0.008*\"scrivere\" + 0.008*\"traduzione\" + 0.007*\"larbaud\" + 0.007*\"coscienza\" + 0.007*\"libro\" + 0.007*\"saluto\" + 0.006*\"piacere\" + 0.006*\"maestro\" + 0.006*\"copiare\" + 0.006*\"editore\"\n",
      "\n",
      "2: 0.026*\"cuore\" + 0.018*\"sera\" + 0.017*\"fabbricare\" + 0.015*\"trieste\" + 0.014*\"domenica\" + 0.013*\"notte\" + 0.013*\"sabato\" + 0.013*\"londra\" + 0.013*\"marcare\" + 0.011*\"mattina\" + 0.010*\"livia\" + 0.009*\"letizia\" + 0.009*\"settimana\" + 0.008*\"leggere\" + 0.007*\"baciare\" + 0.007*\"notizia\" + 0.007*\"signorina\" + 0.007*\"titina\" + 0.007*\"bepi\" + 0.007*\"abbracciare\"\n",
      "\n",
      "For 3 topic the average silhouette score is : 0.6840145176590988\n",
      "\n",
      "Coherence: 0.5212737581648504\n",
      "\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1296x504 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0: 0.019*\"cuore\" + 0.016*\"venezia\" + 0.015*\"sera\" + 0.011*\"letizia\" + 0.011*\"marcare\" + 0.010*\"olga\" + 0.010*\"mattina\" + 0.009*\"notte\" + 0.009*\"livia\" + 0.008*\"bisognare\" + 0.007*\"abbracciare\" + 0.007*\"notizia\" + 0.007*\"trieste\" + 0.007*\"curare\" + 0.007*\"domenica\" + 0.006*\"baciare\" + 0.006*\"settimana\" + 0.006*\"leggere\" + 0.006*\"piacere\" + 0.006*\"ritornare\"\n",
      "\n",
      "1: 0.019*\"amicare\" + 0.019*\"senilità\" + 0.015*\"parigi\" + 0.015*\"librare\" + 0.013*\"romanzare\" + 0.012*\"trieste\" + 0.010*\"joyce\" + 0.010*\"articolare\" + 0.009*\"zeno\" + 0.009*\"villa\" + 0.009*\"scrivere\" + 0.008*\"traduzione\" + 0.008*\"larbaud\" + 0.007*\"coscienza\" + 0.007*\"saluto\" + 0.007*\"piacere\" + 0.007*\"copiare\" + 0.007*\"editore\" + 0.007*\"maestro\" + 0.006*\"crémieux\"\n",
      "\n",
      "2: 0.026*\"cuore\" + 0.023*\"fabbricare\" + 0.016*\"trieste\" + 0.016*\"sera\" + 0.015*\"marcare\" + 0.014*\"notte\" + 0.014*\"domenica\" + 0.012*\"sabato\" + 0.011*\"londra\" + 0.010*\"signorina\" + 0.010*\"mattina\" + 0.009*\"livia\" + 0.009*\"settimana\" + 0.009*\"leggere\" + 0.009*\"operare\" + 0.008*\"notizia\" + 0.008*\"letizia\" + 0.007*\"testare\" + 0.007*\"olga\" + 0.007*\"scrivere\"\n",
      "\n",
      "3: 0.020*\"baciare\" + 0.013*\"viaggiare\" + 0.011*\"livia\" + 0.010*\"londra\" + 0.010*\"scriverti\" + 0.010*\"addio\" + 0.009*\"trieste\" + 0.008*\"affare\" + 0.007*\"donna\" + 0.007*\"abbracciare\" + 0.007*\"maritare\" + 0.007*\"titina\" + 0.006*\"ritornare\" + 0.006*\"cuore\" + 0.006*\"compagnia\" + 0.006*\"occhio\" + 0.006*\"madre\" + 0.006*\"salsomaggiore\" + 0.006*\"desiderio\" + 0.006*\"fondere\"\n",
      "\n",
      "For 4 topic the average silhouette score is : 0.627183071382828\n",
      "\n",
      "Coherence: 0.4760018846018731\n",
      "\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1296x504 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0: 0.024*\"cuore\" + 0.018*\"sera\" + 0.015*\"venezia\" + 0.012*\"letizia\" + 0.011*\"trieste\" + 0.011*\"marcare\" + 0.010*\"mattina\" + 0.010*\"livia\" + 0.010*\"domenica\" + 0.010*\"notte\" + 0.010*\"olga\" + 0.009*\"sabato\" + 0.009*\"settimana\" + 0.008*\"notizia\" + 0.007*\"abbracciare\" + 0.007*\"bisognare\" + 0.007*\"piacere\" + 0.007*\"baciare\" + 0.007*\"ritornare\" + 0.007*\"leggere\"\n",
      "\n",
      "1: 0.021*\"amicare\" + 0.020*\"senilità\" + 0.016*\"librare\" + 0.016*\"parigi\" + 0.013*\"romanzare\" + 0.012*\"trieste\" + 0.010*\"joyce\" + 0.010*\"articolare\" + 0.009*\"zeno\" + 0.009*\"villa\" + 0.009*\"scrivere\" + 0.009*\"traduzione\" + 0.008*\"larbaud\" + 0.007*\"coscienza\" + 0.007*\"libro\" + 0.007*\"maestro\" + 0.007*\"editore\" + 0.007*\"piacere\" + 0.007*\"copiare\" + 0.007*\"saluto\"\n",
      "\n",
      "2: 0.024*\"fabbricare\" + 0.019*\"cuore\" + 0.016*\"marcare\" + 0.015*\"notte\" + 0.014*\"sera\" + 0.013*\"londra\" + 0.013*\"trieste\" + 0.012*\"domenica\" + 0.012*\"titina\" + 0.010*\"tchap\" + 0.010*\"operare\" + 0.010*\"macchina\" + 0.008*\"mattina\" + 0.008*\"leggere\" + 0.008*\"scrivere\" + 0.008*\"testare\" + 0.008*\"sabato\" + 0.008*\"abbracciare\" + 0.008*\"livia\" + 0.007*\"olga\"\n",
      "\n",
      "3: 0.024*\"baciare\" + 0.013*\"viaggiare\" + 0.012*\"londra\" + 0.010*\"scriverti\" + 0.010*\"trieste\" + 0.009*\"addio\" + 0.009*\"abbracciare\" + 0.009*\"affare\" + 0.009*\"livia\" + 0.008*\"salsomaggiore\" + 0.008*\"cuore\" + 0.008*\"donna\" + 0.008*\"ritornare\" + 0.007*\"maritare\" + 0.007*\"compagnia\" + 0.007*\"tolone\" + 0.007*\"titina\" + 0.007*\"marsiglia\" + 0.007*\"salso\" + 0.007*\"sera\"\n",
      "\n",
      "4: 0.015*\"dottore\" + 0.013*\"dolore\" + 0.012*\"curare\" + 0.010*\"fondere\" + 0.010*\"mamma\" + 0.009*\"febbre\" + 0.009*\"madre\" + 0.009*\"livia\" + 0.008*\"malattia\" + 0.008*\"riccione\" + 0.008*\"restare\" + 0.007*\"bisognare\" + 0.007*\"leggere\" + 0.007*\"montecatini\" + 0.007*\"dubbio\" + 0.006*\"colorire\" + 0.006*\"frase\" + 0.006*\"desiderio\" + 0.006*\"pensiero\" + 0.006*\"notte\"\n",
      "\n",
      "For 5 topic the average silhouette score is : 0.6028707959166416\n",
      "\n",
      "Coherence: 0.4501067668609\n",
      "\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1296x504 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0: 0.018*\"cuore\" + 0.016*\"venezia\" + 0.015*\"sera\" + 0.011*\"marcare\" + 0.011*\"letizia\" + 0.009*\"olga\" + 0.008*\"livia\" + 0.008*\"bisognare\" + 0.008*\"abbracciare\" + 0.008*\"notte\" + 0.007*\"mattina\" + 0.007*\"baciare\" + 0.007*\"ritornare\" + 0.007*\"salso\" + 0.007*\"viaggiare\" + 0.007*\"restare\" + 0.006*\"notizia\" + 0.006*\"pranzare\" + 0.006*\"domenica\" + 0.006*\"addio\"\n",
      "\n",
      "1: 0.021*\"senilità\" + 0.020*\"amicare\" + 0.017*\"parigi\" + 0.017*\"librare\" + 0.013*\"romanzare\" + 0.011*\"trieste\" + 0.011*\"joyce\" + 0.010*\"articolare\" + 0.010*\"zeno\" + 0.010*\"villa\" + 0.009*\"scrivere\" + 0.009*\"traduzione\" + 0.008*\"larbaud\" + 0.008*\"coscienza\" + 0.007*\"maestro\" + 0.007*\"editore\" + 0.007*\"piacere\" + 0.007*\"saluto\" + 0.007*\"crémieux\" + 0.007*\"libro\"\n",
      "\n",
      "2: 0.025*\"fabbricare\" + 0.018*\"cuore\" + 0.014*\"marcare\" + 0.014*\"notte\" + 0.012*\"sera\" + 0.011*\"operare\" + 0.011*\"tchap\" + 0.010*\"trieste\" + 0.009*\"abbracciare\" + 0.009*\"macchina\" + 0.008*\"leggere\" + 0.008*\"venezia\" + 0.008*\"londra\" + 0.008*\"titina\" + 0.007*\"signorina\" + 0.007*\"tchaperoff\" + 0.007*\"mattina\" + 0.007*\"lavoro\" + 0.006*\"caffè\" + 0.006*\"scrivere\"\n",
      "\n",
      "3: 0.026*\"baciare\" + 0.025*\"trieste\" + 0.021*\"londra\" + 0.017*\"viaggiare\" + 0.015*\"affare\" + 0.012*\"livia\" + 0.011*\"tolone\" + 0.010*\"marsiglia\" + 0.010*\"chatham\" + 0.009*\"scriverti\" + 0.009*\"titina\" + 0.009*\"ritornare\" + 0.009*\"compagnia\" + 0.009*\"inghilterra\" + 0.008*\"addio\" + 0.007*\"abbracciare\" + 0.007*\"donna\" + 0.007*\"plymouth\" + 0.007*\"francia\" + 0.006*\"potere\"\n",
      "\n",
      "4: 0.017*\"dottore\" + 0.013*\"fondere\" + 0.012*\"dolore\" + 0.010*\"sigaretta\" + 0.009*\"pensiero\" + 0.009*\"mamma\" + 0.009*\"bisognare\" + 0.009*\"malattia\" + 0.009*\"dubbio\" + 0.009*\"madre\" + 0.009*\"occhio\" + 0.009*\"livia\" + 0.008*\"curare\" + 0.008*\"maritare\" + 0.008*\"riccione\" + 0.008*\"febbre\" + 0.007*\"desiderio\" + 0.007*\"restare\" + 0.007*\"ideare\" + 0.007*\"speranza\"\n",
      "\n",
      "5: 0.032*\"cuore\" + 0.020*\"sera\" + 0.019*\"domenica\" + 0.017*\"trieste\" + 0.017*\"sabato\" + 0.015*\"mattina\" + 0.014*\"notte\" + 0.014*\"fabbricare\" + 0.013*\"marcare\" + 0.013*\"livia\" + 0.012*\"olga\" + 0.011*\"settimana\" + 0.010*\"notizia\" + 0.010*\"giornata\" + 0.009*\"londra\" + 0.009*\"baciare\" + 0.009*\"leggere\" + 0.009*\"scrivere\" + 0.009*\"letizia\" + 0.008*\"milano\"\n",
      "\n",
      "For 6 topic the average silhouette score is : 0.5648470918382676\n",
      "\n",
      "Coherence: 0.45165299442257223\n",
      "\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAABBAAAAG5CAYAAAA3TsdxAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADl0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uIDMuMC4wLCBodHRwOi8vbWF0cGxvdGxpYi5vcmcvqOYd8AAAIABJREFUeJzs3XmgZGdZ5/Hfc5Za79639+6ks2+dJktDQqYTEiAskSABjMguKOA6yCIIgkTZdJBJVNQRZUBQxJEZh4HIwIA9IzAGg4CAggbECWQn6fR6l6rzzh/vOVWnqusu3bl1z61b34+WZz/13JtO7Od5N3POCQAAAAAAYDFB0QEAAAAAAIC1jwICAAAAAABYEgUEAAAAAACwJAoIAAAAAABgSRQQAAAAAADAkiggAAAAAACAJVFAAAAMNTN7npl9KnfszOzMdP/9ZvbW4qJbHWb2YjP73Aq+7wYzu9PMDpvZxSv13iKY2e+b2ZuKjgMAgLWAAgIAYN0zs31m9gUze9jMHjSzz5vZoyXJOfcnzrknFR1jnpntN7Of6DrXKmwUqVdsPbxL0s8650acc19eoe99opn9vZkdMbPvmdmNC9x3tZl9byW+U5Kcc69wzv3aSr0PAIBBFhUdAAAA/WRmY5I+LumnJP25pJKkKyXNFhnXOneqpG+czINmFjrnml3nzpf0p5JeJOnTksYlTTzSIAEAwImhBwIAYL07W5Kccx92zjWdc8ecc59yzv2DtKzu+5Nm9gkzO2Rmt5nZGdkFM7vCzP4u7dnwd2Z2Re7ad83sibnjt5jZh3LHl6e9Ig6Y2VfN7Or0/NvkCxy/kw4B+B0z+z/pY19Nz/1oeu/TzOwr6Tu+YGZ7Fvoh0h4MP29m3zGzB8zsP5hZz78HLPRz9Yqt67mymR2WFKaxfjs9f17ac+GAmX3DzJ6ee+b9ZvZ7ZnarmR2RdE2PkH5Z0n9yzv2Vc67hnPuBc+7bPeKuS/orSdvS+A6b2bY0rpvN7K70c7OZldNnrk57NLwh/b1818ye1xXfW3PHP5z+zg+a2bfN7Cnp+Renv9tDZvav+XcAALBeUEAAAKx3/yypaWYfMLOnmtnkCT7/HEk3SZqUdIekt0mSmU1J+oSk35K0QdK7JX3CzDYs9UIz254++1ZJU5JeI+mjZrbROfdGSX+j9hCAn3XOXZU++qj03EfSuQXeJ+nl6ff/J0kfyxLjBdwgaa+kSyT9sKSX9IhtwZ+rV2z5Z51zs865kVysZ5hZLOl/SPqUpE2Sfk7Sn5jZOblHnyv/ex2V1KuYc3ka29fM7G4z+1AaZwfn3BFJT5V0VxrfiHPuLklvTN9xkaRHSXqMfFEis0XStKTt8r0c/qArvux38xhJfyzptfI9IK6S9N20cPFbkp7qnBuVdIWkr/T4OQAAGGgUEAAA65pz7qCkfZKcpPdKut/MPmZmm5f5iv/mnPuic64h6U/kk1BJ+iFJ/+Kc+2DaKv5hSd+UdP0y3vl8Sbc65251ziXOuU9Lul3SdSfwo71MvlX+trRnxQfkh2Vcvsgzv+6ce9A59/8k3Szpx3rc80h+rl4ulzQi6Z3OuTnn3Gflh5Tkv/u/O+c+n/4uZnq8Y4ekF0h6lqSzJFUl/fYJxPA8Sb/qnLvPOXe/fEHoBV33vCktgPxv+QJKrzkWXirpfc65T6exft859830WiJpt5lVnXN3O+dOaggHAABrGQUEAMC655z7J+fci51zOyTtlrRNPoFejnty+0flk2Gl7/i3rnv/Tb4VeymnSvqRtEv/ATM7IF/k2LrMmLJ3vLrrHTvTuBZyZ1esve59JD9XL9sk3emcSxZ5351a3DFJ/9k598/OucOS3q4TK7Z0/0zdP/tDae+Fha5ndko6buhE+uyPSnqFpLvTIS/nnkB8AAAMBAoIAIChkrYYv1++kPBI3CWfxOedIun76f4RSbXctS25/TslfdA5N5H71J1z78zCXMb33ynpbV3vqKU9BhaysyvWu3rcs9TPtZzYut+3s2u+hfz7lvPOf+i6Z7H7e13r/pm6f/bJdBjCQtczd0o6o8d5Oef+p3PuWvki0Dfle7sAALCuUEAAAKxrZnaumb3azHakxzvlu8//7SN89a2Szjaz55pZlE5seL5893zJj4F/jpnFZrZX0rNzz35I0vVm9mQzC82skk7mtyO9fq+k07u+r/vceyW9wswuM69uZj9kZqOLxPxaM5tMfwf/XtJHTuLn6hXbYm6T77nxi+nv4mr54RB/dgLv+M+SftzMTjezmqTX5+Lpdq+kDWY2njv3YUm/bGYbzWxa0pvl/xnk3WRmJTO7UtLTJP2XHu/+ozSOJ5hZYGbb0z9fm9PJFevyw0gOyw9pAABgXaGAAABY7w5JukzSbeks/38r6euSXv1IXuqc+4F8ovlqST+Q9IuSnuaceyC95U3yrdUPyY+5/9Pcs3fKT2L4Bkn3y7dsv1bt/798i6Rnm9lDZvZb6bm3SPpAOlzhRufc7ZJ+UtLvpN9xh6QXLxH2f5f0JfnixifkE+IT/bl6xbYg59ycfMHgqZIekPS7kl6YmztgSc6598lPXnib/PCCWUk/v8C935QvGHwn/V1tk5+s8nb5ngxfk/T36bnMPfK/w7vk57l4Ra/4nHNflPTjkv6jpIcl/W/5ng2BpFelzz8o6XHyy4bKzK40vzIFAAADz5w70Z6IAABg0JiZk3SWc+6OomNZS9IeER9K58cAAACLoAcCAAAAAABYEgUEAAAAAACwJIYwAAAAAACAJdEDAQAAAAAALCnqx0unp6fdrl27+vFqAAAAAADwCHzpS196wDm38USf60sBYdeuXbr99tv78WoAAADkfetbfnvOOcXGAQAYGGb2byfzXF8KCAAAAFglL3+53+7fX2gYAID1jzkQAAAAAADAkuiBMOick+Zmi44CAAAUJUn8dnam2DiAxUSxFIZFRwHgEaKAMOj2f0x6z69IEf8oAQAYSt9M50B40b5i4wAW0mhIlz9Bes1vFh0JgEeIrHPQzRyTgkCa3lp0JAAAoAjxv/otfxfAWnXogDQ3V3QUAFYABQQAAIBBtvesoiMAFuecFMdFRwFgBVBAAAAAGGQ7T3gZb2B1zc1KW3YWHQWAFcAqDAAAAIPs/of9B1irXCJV60VHAWAFUEAAAAAYZJ/7hv8Aa1mpUnQEAFYABQQAAAAA/RPF0vbTio4CwAqggAAAAACgf5yTJjYUHQWAFUABAQAAAEB/JInUmJemNhUdCYAVQAEBAAAAQH88dL908b+jBwKwTrCMIwAAwCC7/NyiIwAW5hLp0ddIZkVHAmAFUEAAAAAYZFunio4A6M0lfnvKmcXGAWDFMIQBAABgkN39oP8Aa83sjFQflc7eU3QkAFYIBQQAAIBB9rff9B9grUia0gP3SIcOSNc8o+hoAKwghjAAAAAAWDk/uFc68wLpx1/ntwDWDQoIAAAAAFaGS/yEiT//Dmnz9qKjAbDCGMIAAAAAYGUc+IF04WXSpm1FRwKgDyggAAAAAFgZjXnpkitZthFYpxjCAAAAMMj2McYca0SzKYWhtO+pRUcCoE8oIAAAAAyyjeNFRwB4B+73qy6M8mcSWK8YwgAAADDI7rzff4CiJYl0/qVFRwGgj+iBAAAAMMhu/xe/3bmx2Dgw3Brzft6DPZcXHQmAPqIHAgAAAIBH5uAB6YonS2MTRUcCoI8oIAAAAAB4ZFwi7X1c0VEA6DMKCAAAAABOXpL4zylnFR0JgD6jgAAAAADg5B14QDp7j7R9V9GRAOgzJlEEAAAYZFfvKToCDLO5WWl+TnrFm/0kigDWNQoIAAAAg2xypOgIMIxcIj14n1StSy/4BXofAEOCAgIAAMAg+9d7/fa0zcXGgeFx+KA0c1R67BOln75JiktFRwRglVBAAAAAGGRf+bbfUkDAanj4B5IF0utuli66gmELwJChgAAAAABgaccOS1EsvfNPpc3bi44GQAEoIAAAAABYnEukQw9Lv/gfKR4AQ4wCQg+7du3S6OiowjBUFEW6/fbbiw4JAAAAWH3ZZInOSY+9Vtr7uKIjAlAgCggL+Ou//mtNT08XHQYAAACwupxrfw48IG05RXrlO6VTzyo6MgAFo4AAAAAwyJ54cdERYC1JmtLcrDQ7I83NSGHkJzrsNdmhc5JcZ8Eg+0SRf3b7Lunn3kbxAIAkCgg9mZme9KQnycz08pe/XC972cuKDgkAAKC30WrREaCfnPPDCJLcx3XvN6VmQ5JJQSBt3Cqdc5F05gXS1lOlStVPfhhGfhvFUhwff671iVhdAUBPFBB6+NznPqft27frvvvu07XXXqtzzz1XV111VdFhAQAAHO9f7vLbs7YVGwc69WrdP3xQasz5ZRADk5T1DLD0XuWeSQsEklQqp5+KVK1J5YpUrvpPpeY/510i7b1KGpsk+QfQNxQQeti+3c8su2nTJt1www364he/SAEBAACsTV//rt9SQHjkmg2f5Le6/gc+t+/mWv+nq1Agn/i79FoYSkGYtvSHfvWCJ/+oLwLE5VxhoOyP41LuON2PYgoCANYMCghdjhw5oiRJNDo6qiNHjuhTn/qU3vzmNxcdFgAAwHDLWvGTtMt+kkiu2dm1v3XetRP5bPy/BelW8v/Hji8OJIl00RXS2Xukbbt8y34U+SJAGPkiQJjODZA/3/OekMQfwLpDAaHLvffeqxtuuEGS1Gg09NznPldPecpTCo4KAABgHXLOt/ofOyLNHO3d6p+N8282faJeSrvv1+p+bH+lJlXq/riafuKy7+afteTHpR7j/OPj5wMYnZAmNhT26wCAtY4CQpfTTz9dX/3qV4sOAwAAYG3LegTISUm+G3+inrP6OyfNHvMt8/meBLVRadc50uVPkMY3+OJANR3X3xrjn473j/irKwAUif8KAwAArDcdXf2b7f1ms7Orf9L03ezDUK0Z/FvN/7nEX+oc258k7efCtPt+FElhflb/9DjKzfJfqUnXPlvastO39tdH0+8GAAwCCggAAABrmXN+Ur+ke7x/2rX/MTt9D4D770qT+khqzPvu++VaOmt/VaqP+O79tVGfuFdHpNEx3/2/Uk2HBqTDA7LjUnnhsf9BwBh/ABgyFBAAAACKNnNUmjkmJQ3fS0DyCXoQ+sLByLg0vaWrW39aHKhkY/9rPuGf3iqdudsfAwCwgiggAAAArBaXSAcPtJcJDAK/9F+1Ll32eN+tf2xSGhnz5yo1vz3jgoULAu9/vzQr6Rk3ruIPAgAYRhQQAAAAliM/aaBbYH4Al+435qX5Ob/CQBimPQnSlQTOv9QXC8an/GdsUtq4zQ8fOBnvf7/fvvjFj/xnBABgERQQAADA+tRsSEcO5RL5qPe4/awIoLQAoHxhIFcoMPOFgDBsFwWC0E8OmE0cGIZ+4sCpjX6iwK2nSJMbpalN/jOxwS8pCADAAKKAAAAA1r5secCOFQS6j3PbKPbJ/QWXSttP88n7+JSfS6A+6pP4IFcIyIoL2YSBrUJB7rwZkwYCAIYaBQQAAFC8Y0ekY4d94p7pKBokPunPVgmopxMJZqsM1Eb8cX1Eqo1JZ18onXUhrf0AAKwgCggAAGBlNJt+uMD8rNRoSC7XU8AsHT4QdLXkm+8xUKlKz3qZT/orVV8kKFfbBYNyxfcIAAAAhaGAAAAAlifrEdBo+CLB3JzUmPPDBZpNKYr8UIFNZ0kbNkvVkfYqApWqX2IwLvtt9onLfsWBLTsZHnCybr216AgAAEOCAgIAAMPs2BG/pGBj3vcUyMb/Z/LDCFwiRSU/h8C2Xf6z43Rp1znSrrP9agIUAVZfbYHlHQEAWGEUEAAAWOuypQMT114uMEvoXdL7fOLaQwicaw8fCAJJJpnaQwsefbU0vVWa3uInGqyNtocR5IcTVKrp81hTfvd3/fanf7rYOAAA6x4FBAAA+sE5P7a/2ezs7t+aC8Dks/g0mc8/J6UFAbVb/8N0ucC45HsBlMt+/7hhAZX0U/YTC9ZG2kMI4nJ7ToFSuj+9VRodX/3fD1bOn/+531JAAAD0GQUEAABORrMhHXzI72fd9jt6BjiftFfrfhnBraf67fhkWggo+22p3C4MLPSJYiYQBAAAhaOAAABAN+f8agKzx6S5WV8sCKN2zwHn/PbCy6SnPsd3+a/WOrv6xyXmAwAAAOsKBQQAwGDJT+qXb+1vfRY77rqWJFIYSAqkwPzkgUniCwbjU9JZu6VTzmp386+P+RUD6mPSpu1SHBf92wAAAFg1FBAAAMVpNnxLf/ZpzPvu+tlEfa35ALqS/yj24/zLFSmu+EQ+iqVwoSEBcTpvQMVvSyU/hCAbHhBFfqjBxLQ0uVGanPb3AgAAoIUCAgBgZWQt+kkzt222ewpkEwqa0pb+pk/eN2yWNu2QtuyQtuz0CfzYZLtAkJ8ksFxhaADQbf/+oiMAAAwJCggAMOxaQwKyVv7cfmtJwAX2sxUFgtAXB+JYKtekSkWq1P28ANURqT7ityOj0rbTpJ1n+GJBbYRiAAAAwICggAAAg6J77H++td+l4/abTb8NQj9rf2upQNd7OECSFgGyrvz5bv/ldDnAbFLASjpJYLXm96v19nKBY5PSmRf4LYDV9a53+e1rXlNsHACAdY8CAgCsVc5JD92fS/iT9jj+clWq1dPkvuaT+vpYe/z+SDrhX300HfcftwsE+W22PCC9AIDB9fGP+y0FBABAn1FAAICi9OpR4JrSXDqh4MwR6fTzpV+82RcFSmUSfQAAABSGAgIArISZo9LRw34IgRIpiPyygJatJiBfLJDz+0m6H6erAZTLfs6ASk2a3iJtP03avEO6+N/5SQYBAACAglFAAIA85/wcAg/en64WELSLANnSgq17lRYF0mtXX++HEIxO+KED1bqfJLBS61xJIPtEMT0KAAAAMDAoIABYP1w6UWC2XGC2bcxLjTl/TxhJMl8cyD+Tn1jQTLp0n3TJVb6HQJbsl8q5SQZz+6WyLxZUqkX95ACGWZX/9gAAVgcFBAArzzm1Zv133ce5jxbYb51T53GSrjIQRmlvgNzqAi7x16I4XTEgXUJwZNwPCdi03W9HJ9o9AsrVdJutNlBhngEAg+ev/qroCAAAQ4ICAoBOWRf+ZjOd4K95/JKB2b4FUpT+Z6S1nGCa+IdheynB1jby2yjy56LIn4uy83F6b9w+F5d8wSCKpNFJaWqTLwKMjvviQG2kvcxgudqOBwAAAMCK4m/awHqVdeXvSPzzXfrn24m9Wef9tZF2l/xyzbfY1+q+Vb9Wl2qj/p76qJ/gb3qLT+xLZf/OIKAVHwBWy6/9mt++6U3FxgEAWPcoIADrwdFD0tEjPnHPlgWUfJf8ctoyX6n5gsD4lE/2N2yRJjb4VvyxCb8dGfdFgTAs9ucBACzfZz7jtxQQAAB9RgEBWE3ZWP1sGEA2aV/3JH49j7uuJYkUBn6OgYlp6effIJ1zUVo0qDDDPwAAAIAVRQEBeKSck44ckmaO+oQ933rfK/nPZu2Py354QBT5Mf9xSYpjn/i3ZvmPu2b7z86X/X5U8kMJHnONfycAAAAA9AkFBKy+bJI9pTP0d2xd5z2t2fkXuK/ns4/wXS6RkvTeIPATBZq1W/PzrfpOfmWAU86SHnW5dMrZ0tRGP6N/qZwrFFTaSwDSKwAAAADAAKKAgJXVmsE/ncW/2Wwvvdds+HuyCfvyyXkQtCffy863ttlEf0H7epAeh/nnwtz59Fz+uTBqrwjQ8Wy6GkD2XKnSntU/a/nv/kS5/Wpd2rCp2N87AGB4bdhQdAQAgCFBAQFtzYZ07IjfJk5y2cz9aWt8K6nvao3PWvazJfyqdT8R38i4VB/zE/SNTkjjG6SRMWnTdumiK2iJBwBgJXz0o0VHAAAYEhQQBsHcrDRzTJqbkWZnpNljfjs3I33xs9KD9/nEPknH2ge5Vn2ZZEq3PRL2fPf+MJTO3O1n5s9m7q/W27P358fut8bkl32xoDYiVUekao3CAAAAAACsQxQQijY/L93z/3zLf/fn4AHpO/8offVv0y765nsB5N33fems3dLVP+wT/mxivTj2XfajdFK+bL/7fHYtjn0PgYg/EgAADJRf+iW/fcc7io0DALDuDUa2+J1/kj77l8fPZp91me+11F2ywDl1PZs0c/vpsnpJM+22n7S78CfNziX48vd377cm5VPnBH2tTW4Cv2NHpKOH23MG5O8LQ18MGJs4vnCQiWLpuudJ179gxX/tAABgAPzf/1t0BACAITEYBYRv3C595QvqSLxbm65zrWP1nqk//1y/ZEMHgqA9fKDjnLXvO/NCqT5y8t/VbEpnXfjIYwYAAAAAYBGDUUC4/gW0sAMAAAAAUKAF+sUDAAAAAAC0DUYPBAAAAPS2Y0fREQAAhgQFhAF3NPkHPZj8iYzOJAAADKf37/Tbxi8VGwdWRVNH1XQHdGr0+wqsWnQ4AIYMBYQBN+P+UQ8lf6FYm4sOBQAAAH3S1CE1dVChxjUaXClTpeiQAAwhCgjrQKhxRbah6DAAAEABpl/9DUnSA795QcGR4JFI3IwaekhmoUwmOb9qlzO/tHiourYEb9BYcLVi21porACGFwUEAACAAVb+6sNFh4AT4FwiqSmnhpyacpqTs1k5SzRuT9NosE+BagpUV2C1dD87ptcBgGJRQOih2Wxq79692r59uz7+8Y8XHQ4AAAAGjHMNNfSQZIlMgZyc5JqSTIGNKNSoAo0qti2q22M1Elymku0sOmwAWBQFhB5uueUWnXfeeTp48GDRoQAAAGBAOJco0UEldkyySGP2BI0EV6psZyrSlALVZSrJzIoOFQBOCgWELt/73vf0iU98Qm984xv17ne/u+hwAAAAsIY559TUQ0o0I7NAVV2o0eAJGg+eqsimig4PAFYUBYQur3zlK/Ubv/EbOnToUNGhAAAALGnurJGiQ+gb55yk/EeSnB8O0OPjcvcs/3x2LUmPE0mBzLIlsgNJaY8BZ2r1HTD/BllTFTtP43adqrZb1WD3iv4OAGAtoYCQ8/GPf1ybNm3SpZdeqv379xcdDgAAwJLu/709RYdwQnxRIFFnEp/IaU6JHZVcmObrTlJDslCmUFK2DWSK0k/+vD/OrkmhzHL7inPPZZ/YP2exAsUyVWSqKLA4vb+Ubrs+1nk+1haZhUX8OgFgVVFAyPn85z+vj33sY7r11ls1MzOjgwcP6vnPf74+9KEPFR0aAADACXHOyWlGiWbUTti1ROu9U77F3Y/Vzz4unQwwO1K63GDXF1v7Wzq+w/kWfpNJFrcS+CxRDzWlqeA5Ktl2hTapUBMKNa7ASv36FQEATpD5KvDK2rt3r7v99ttX/L2raf/+/XrXu9615ldheLD5Z7qn+Rsq2faiQwEAAF38kn0+eW8n6EsdZy3yCx0HaWLvE/0tP/UdmTPd83un++80lyb4TUWaVsXOUaCq1ErYcy3waUu6ci3y/nwgtVrzs1b/MNea376ev2a5wkDn9+XPB0wiCAAFM7MvOef2nuhz9EAAAABrQvd4d9dqNe9OsJW7Tx1bt8i19h29ri30DpPJ0nMmn/e2W+Tb+yZzaVJsaewuSZ8pdyXRldxx2hXe8t3lS639QGWZygosO1dTYGWZSum1kmp3/IIkUzl6b1oQ8NcC1RTa+Mn8owAAoCcKCAu4+uqrdfXVVxcdBgAAhXAukdNs2v29mZtcrrO7u0+uraOre0fCLUnO1ErMLd/zMd+93SfdpkCyKB2PniXTlVby3TkOPmsFD3Lj5IPcuPj8/vHb1n5ujH37Wrvl3RcBct/V+m5Tu/W9rEB1hVZXoPZndZbsG5MkJu8DAPQdBQQAAArkE/WGpKZ8N/Uk3eaPm63W+NYY9I6x6fkEtZ3AmyTnWqPZ/fcp7d5u6Vj0jhb3VlSSnGLboopdoEiTuZbycqvlu7P1PJvALkv0OyeqM4WS5ce8d4+BzyanCwQAANYmCggAAJykzi73+THq+S737QJAVihwStIZ2wPJEj9RnCoyVRWopsBqvvt51pJto2mLdll+/HivWem7W9Z7tb5nLfWBjmuFz52TwjQOJq8DAABtFBAAAGtWe7m37pb4dgv98ZPNdY6VP76LveS7pWfXcivGpWcl19mo3z3BXau7vaWt6vmW9Eimase50EYVaEyhxhSan1k+tBEFqqlsZ6lsu/rx68OwuOiioiMAAAwJCggAsEa1V8np7lre+5xPcJtympdTo7VdeAK47Nkgt9/dFT67ktvv6BLvWvdmEWRf4XJPtbrNd8TceSY/aV1Hkm4lBaooUDUdU15rtdKHqslUk59srpSOnc8mrEu31rk+fOfM8mHP82YLrRufn60+nX2e2eRRtJtvLjoCAMCQoIAAYN3xiXcjTaLbH6kh37IcpuPI20mxf647Mc4n2blk13pddQuk0gsl/9k117WGusu9KUvhA8myydqy5D9oX1M2cV2oQCOKNKXIphVrsyLbpMAq6T1h1/PBSe9n39f+7kCyoL2vE93v9X1+YjoSdAAAgLWBAgKwznW2Ynd3817ofO9nFr+/97Jox0/OliacHS3h3QlidxpvueQ+97Nl91g+gfdd3EMbUaisq/ikIk0rtmmFmlJo42lLdtiRBPdOytN960zeOxPfhZPgbEK7Xs8s+n1ZF3uSZwBLef7z/fZDHyo2DgDAukcBAVhl+RnXs7Hcvcd3pzOuW7ultt2lPG0Cz1Js607U/frj2Tk/TrtzWTK/zbp15yda69HV27Lj9rVsbHe7C3h+lvXouHe1u3t3L4XWYzm2Ja51TgTXvtcUK9SYAo0wkzuA4fG97xUdAQBgSFBAAE6Ac9ka6ElXot+e1C3rKu8nbUu7jDtLk/ymJFNoY+mM6n7W9dD8OO7WrOu5dcTN/Nhuv1xalriX1F7yLD7+XG4yN7/GOa3YAAAAAB4ZCgjrQFMPat7xj/KRSdLJ2tKWc+fXSG+PUU+vK5BZSaZsHfRqaz9QVYGNKdIGRbZJkU0q1KhvEbcx35Ve4wpUI6EHAAAAMHDIOgdcLbhEm9wrdfzocJwIs1iRNii0cb/MmsbSdderMpXaPQCMf2UAAAAADCeyoQFXsbNVic4uOgwAAFCUxz626AgAAEOCAgLBiqFxAAAgAElEQVQAAMAge8c7io4AADAkmKYcAAAAAAAsiR4IA+6LD39Rv33nb6sUlIoOBQAAFOCn//1nJEm/e8sTCo4Eq8nJ6f65+3WweVCRRdo7tle/fuavFx0WgHWOAsKA+86x7+i2g7dpa3lr0aEAAIACBA8ekCTdceyOgiPBanFyunv2bk3H09pZ2ak9I3v0jI3PKDosAEOAAsI6MBKOaCwaKzoMAABQgMhCSeLvAuvcseYxHUuOaTaZVdM1deXElXrf+e8rOiwAQ4YCQpeZmRldddVVmp2dVaPR0LOf/WzddNNNRYcFAACAIfDQ/EOaTWaVuESRRQoskJPTaDiqK8av0OnV07WzslP7xvcVHSqAIUQBoUu5XNZnP/tZjYyMaH5+Xvv27dNTn/pUXX755UWHBgAAgHXEOac5N6e5ZE5zbk6zyawCBXrh1hdqY7xR49G4xqIxjUVjOrt2tspBueiQAQw5CghdzEwjIyOSpPn5ec3Pz8vMCo4KAACgt69eNl10CDgBzjkdTY7qSPOIjjaPajwa1xnVM7S1vFU7yjt05cSVetToo4oOEwB6ooDQQ7PZ1KWXXqo77rhDP/MzP6PLLrus6JAAAAB6+sjLzi46BCzTw42HdbhxWKdWTtXjJx+v3fXdumbqGo1H40WHBgDLQgGhhzAM9ZWvfEUHDhzQDTfcoK9//evavXt30WEBAACgT5xzfqseW+e32blEiZquqcQlairduqY/p0SSFCpUYIFM1nrmWPOY3nTam/Qjm39ktX88AFgRFBAWMTExoWuuuUaf/OQnKSAAAIA16S0/c5vfvmdwe0w65zTv5jXn5nS4cViRLfxXVJO1EvnW87njbN9kvihgvYsDTq71ruy6mSlQoMACBfLJf2jtQkBggUKFioNYE9GE6kFd9bCu0WjUr4oV+vkK6mFdI+GI/0Qj7f30AwCDigJCl/vvv19xHGtiYkLHjh3Tpz/9ab3uda8rOiwAAICeSjPNRa8vlDwvdm3BZ1zvFvperfS9viP7n2bSVGSRzEyJS5Qo0Vg0po3xRj1l6im6evLqVhLfSt7T/dDCVjKfT/a772ttl3mfyZj3CgCWQAGhy913360XvehFajabSpJEN954o572tKcVHRYAABgQWWt60/nEPlFyXIKdT9Rb3eKd7/re63r+uJXspknwvJuXJP1g7gcd92bd6qXjW9a7t6GFnQl3mqhnLe7d94UWdny6z7WOFR53b2ihNsebtbOyU9vK27Sx5FcbCC0s4J8WAOBEUEDosmfPHn35y18uOgwAANaEXq3HvZLajnu6r3e1XGf3du9n7+k417Vd6D3HnUvvyxJt/78mOSkdkt4amy75BFtOrXtzL1uwVdp0/Hknp4ZraCwa00g4otBCRRb5jyKF5ru/Rxa1usJn12PL7afnS1ZqnSsHZUUWKbDAP2+hykFZp1dfr0CBfv/c31c1rKoSVFQNqqqGVZWt7FvsaVkHAKwACggAABQscYnmkjnNulnNJrNqumZnYpwmseZ8F+tWkptLbluJcpfsXPeY8excr27p3YWBrDU635qcJbCRRYqC9n6oXMKc3pNdy8aTS2p1Q89aq7P9rKXczNot37nW8lCd3dezZ/Kt5Nk78l3dWy32ua7qWQt+6zt7nF/snvz7AgskSZPRpLaVt6kW1lb2D8liondKkvaM7lm97wQADCUKCAAALEO+q/mSY77T5Hvezeto82grac4n81kXcyenQIE2ljbq9PLpOqVyiqbiqY7W6CwR707ks4R6sa7kWWKdP87fm2/N7tXlPEuesYYx1BIAsEooIAAA1ows6T6WHGstkZYokXOuNY48cYlvhU5bhCW1W+UXeu8Cre/5782f6y4EZN+x4NjxHuPEwyDUaDiqp298ujbGGzu6lleCSuu4ElQUW0ySjpP3mtcUHQEAYEhQQACANSKfJHe3bEsrMzv6cde6xq63ri/Qsp64pKNF2nL/kx9X3p3Md3ejzy/Dlv8Zm66peljXubVztSHeoHJQbiXZWeJdDsoqBSXFFisOYsUWq2Sl1n6rtX6BGdjz1zru69p2XyPBBwAAw44CAgD0gXNOc25O827et6KnLemJS9RUu2W9Nb46TaizCdJWYnb0jtbw3LXsOzq6wfeaLV1hR9f2LDmPg7jnLOy9ur7nn1/OM+WgrLFwjGQdOBFXX+23+/cXGQUAYAhQQACAJWRj37MCwELb2WRWkfn/rDZcQ1PRlLaUtqge1FUNq6oGVY2EI6qFNY2Go6qFNVWDqspBWfWwrt0ju7WptKngnxYAAADojQICgGVbzhJr+WS71SW/a/x69wRyrW16r6TWjOZZ1/fjusVb57X8dtk/T9e4+FbXffNbSa3iQGSRSkFJlaDit2Glo2t9NazqlPIp2ju2V7uqu7Qx3qg4iE8oHgAAAGAto4CAobPQZGmt/QXGnmfbhZLo/Brli7VUO+faY6rT5cCk9pjw/FjxjjXKF7jea3K4ha4fd+8S66l3rM2eWzJusSXOQgtVspIfox7EKgWl1nE5KKtkuW1Y9mPcrdIa656fbT4/Qd1yx6kvdf9yxr7n78liz2bRBwAAAIYVBYQCZK2xCyWY+ZnHj9vvuv8bR76hQ81DOtg4eNxkaQtNlNY6l0t4F1qGLIs3v+61nFqtv8dNlGadCW/HtQXWJ88nsAslxPlkuCMxTmPJ4szf3zMxzs2mniW++TW888cd65PnZl83s1aS2msN80BBq3W6GlRVC2uqhTXVg7rqoe/Knk+ks7XRe61Tni80LHQ9f657rfLFnmlte6xvnn8+f737nzEAAACA4UEBoc/unb1XP/FPP7HshH6hGdEXmyl9W3mbzqmdc9wkZa3JyxaYKO24464JzronY8sn3Mcdp0lmfmm1LKHOJ6sne9y9353YLpRwL5hQkwQDANaLG28sOgIAwJCwfKvtStm7d6+7/fbbV/y9g+rhxsN+6bOuFu6VUg2qJMQAAAAAgGUxsy855/ae6HP0QFgF49F40SEAAID16uhRv63Vio0DALDuUUAAAAAYZNdd57f79xcaBgBg/aOAMOAOzTd0x6GjRYcBAAAKcmajKUm648GDBUeCYTQWRzpjlN4vwLCggDDgbv3+A3rjV7+tjeVS0aEAAIACvOegb0j4mS9+q+BIMGzmkkTVMNDnnvzookMBsEooIAw4J2ksDjVdiYsOBQAAFCAO/ETK/F0Aq+2B2TntmRwpOgwAq4gCAgAAAIBlmWkmemhuXqFM8y7Rk7ZuKDokAKuIAgIAAMAA+8xTnl50CBgiD83N6+pNk/qh7dO6aHJUk2V6vgDDhAICAADAAKOAgNVytNFUaKY37zldYzFpBDCMgqIDAAAAwMkbe/ghjT38UNFhYJ2bbSY6NN/Ur198JsUDYIjxbz8AAMAAe/2vvFaS9Iab/7DgSLBezTYTPTA7r9edf6qu3jxVdDgACkQBAQAAAEBPRxtNHZhr6LXnn6IbT91cdDgACkYBAQAAAECHxDndNzOnWhjqXZecpWu20PMAAHMgHOfOO+/UNddco/PPP18XXHCBbrnllqJDAgAAAFZFM3G659is7puZ0xO3TOkvH/coigcAWuiB0CWKIv3mb/6mLrnkEh06dEiXXnqprr32Wp1//vlFhwYAAAD0zXyS6N5jc3rGzo164enbtKtekZkVHRaANYQCQpetW7dq69atkqTR0VGdd955+v73v08BAQAArEm3Pv1Hig4BA6yRON0/O6dAUiUM9YStU3rVeadqlJUWAPTAfxkW8d3vfldf/vKXddlllxUdCgAAQE+fe/yTiw4Ba0jinJrOKXHpfnouO06U23fSwfmGrtk8qbdedKY2lmN6HABYFAWEBRw+fFjPetazdPPNN2tsbKzocAAAAHqavu8eSdIDm7YUHAlOlHNOTpJzUiKXbn1y79RO8pvOqZF+JCkyk0kKZHJq39t0UmCmehSoGoaqR6EqYaBqGKgahaqFgWpRqGro9+vpuUdPj2tTpVTcLwLAwKCA0MP8/Lye9axn6XnPe56e+cxnFh0OAADAgl719l+WJL3h5j8sOJL1x6Wt+Q0nNZxTM/FJvG/h98m9k1Mgk5lP3rP2++52fJdus2dcuh+aKQ78pxwEKgWmUhD6/dBUCQJNlGJNliJNl2NNlmONRKFGokj1yBcEfCHAb0uB0YsAQN9QQOjinNNLX/pSnXfeeXrVq15VdDgAAABYIa0Wf6Wt/GmL/0yzqWPNRKGZAvM3NNOW/3oUajQONR3HmihFmoh9Ml+PQlWjQNUgUDn0iXspzAoAQfqx1jbusY0DU0CyD2CAUEDo8vnPf14f/OAHdeGFF+qiiy6SJL397W/XddddV3BkAAAAgyGfqLtci7vfT4/d4tfz4/Szcftmfg1yk0/0W6395p/1Z9rfLeVa/NNtZIGioN3qH1ugCydG9OipMe2sVzSZtvZPlWONxREJPgDkUEDosm/fPjnnlr4RAABgncgS/mbaPT9rfW99En8uCnziniX6/tnjCwCSH6cfmikM0m766XGUJu1B2nU/Ss9F1t6vhIFqYTpWP+umH4aqREHazT9QOfSt++W0JT97b1YciFpDA4LWfmR07weAR4ICAgAAwBrlnNN84jSbJJpPnOaTRM103HyWB88nPmV/YGau/WDaDT/btpJ9tVv9JaektS/Fgflx9HGgkTDSaOy77o/FkcbiSFsqJW2ulrWhFGtDOVY5DBSaFFmgwNoFgyjw8wCQqAPA+rNoAcHMQklvc869fpXiAQAAGHjO+eTct9y3J+LrtYxefhI+M/Nd9NNW/qZzGo0j7apXtalS0pZqSVsrJY2X4larfem1r1Eg6ea953Qk8aF1tu6HPfbDXOJPwg8AWMqiBQTnXNPMrlmtYAAAAB6pLHnPWtlb4+FdrgU+Pegep6+O49zzaSt9fvx++7g9BGA+cYrM5CSVAtNIOkP+aJy26EeRaukSe9XcMnqVKGxNtlcOApXTyfi21yraWI4XT+5f8GMr+NsDAGBhyxnC8CUz+6+S/oukI9lJ59zH+hYVAAAYSEna5b7h3AlOpJcuayf/QCInU9YV3vfEN/mDQJ1L5LVn1W+/M05b4Nsfde37MfjZcWSmIGudlylstc6n3fQD69jP3tU9fr8Whbpu+7SmSpGiIFidX/q3vuW355yzOt8HABhayykgjMoXDvLLEDhJFBAAAFjHXNrFvtnR5d5PppfIqZH4YkEc+HS+6SST02Qp1nQ5Vmy5CexCy01kF3TMgJ/tl9LJ8NpL3HWNrU+T9XIQqBL6TzkMVGkdh6qkLfdD1R3/5S/32/37Cw0DALD+LVlAcM69YDUCAQAAvSXOt+g3s4Q+n8jnZs2XnAJLW+3TNvogl0dbfqdrgr38+kPZ2HxJPkEPA1XDQNUwaq17X49CTZZjnTNa0yn1irZUy5oqxRqLw+FK3gEAGCJLFhDM7ExJ75G0xTn3KDPbI+mHnHPv6Ht0AACsUdmSv/nu+e3u913d89VuzU/S89lEes45NdN35LvrZ4WA7P1jcahqOl6+HoUaiaJ0bL2fJX80ClUO0yXs8svlpT0AOrra57axmaLAd83PL3eXLY1HMQAAAGSWM4ThDyW9Qb6IIElfk/RhSRQQAABrnktb7+fT7vbzLlEz8a34khRaO1mXOlvis+P8snf5goGpc7b70PLrzwcds95nE+NVQr9ufS3KWvbDtHW/3f0+Tu8di0Ntr1a0uVJSGJDIAwCAYi2ngFB3zn0ha4Fwzjkzm+9vWACAXvKt3tk2n9i2rnWt+966ftz57H533Hm13t05O33vd7hF3t31fO5862eQa3W7b0+Y5///Tnac7fvn0jPWDnKh34lzUj0OtaEUa1OlpI3lkibLkSZLkUbjSPUwVC0KVYt8y34pbYXvXgqvNdle7hyt8wAAYJgsp4DwAzM7Tdlf2cyeIemevkYFYChlybHUuxU428nf5Ra4p/sdHYnuEve09nNZbj45fXiuoWoYtMeRd+s+nzvOv6f7tl7JenfCnr0uNFNgPskO0hb0wNRaSz5MZ6oPzB8H6czylt4TtJ7P3pU+p/Z7ApPCwFrd6bN3Wjozffb9Ye6+bGb67B1Rr/Np8p3dE5p1fE/clbS316oPjlu3Pk673i+8xr2pGgarNxs+UIRf/uWiIwAADInlFBB+VtIfSTrXzP5N0t2SntPXqAA8Yolzmkuc5pKktaRaNoO6pUmbTyblM8IeyXIvi15zPi3O7rHuVnHlEufjEvjsSnsceNA1CVxg7ePO1mrLPSNZaxI5ta/nftaw9d7292Vd2PPflT+fxeMkjU9Ees6uzf53qM7kvVcyH6RJdtB9fy7Jbz/fec1y7w2zeGn1BpD3xCcWHQEAYEgsZxWGOyQ93szGJZlz7kD/wwLgnJ+QrZkWAmabiWaTxCegarcet+7PPulEbYGZNpRjbR+palu1rOlyrHoUqh6GqkZ+ubNyYH4btpdFi4PguMT5+K7lCyTp6kzYu+9tJec9CgPtLuokxwBwQr7yFb+96KJi4wAArHvLWYVhUtKbJO2T5Mzsc5Le6px7qN/BAaup19hyf94fLTx2O9t3C5zXguPLEyfNNJutVufsXLbeehyYalGoLZWSTq1XdMZITZsqcWut82wytt7roQck4wAwDF75Sr/dv7/QMAAA699yhjD8maS/lfS89Pi5kj4i6Un9Cgroll/nvNlaC90fzzaT1jjr3uPKs7nSXWdi39GF3j+ZjdXOL6PWMc78uPHl+W7oQce9+e7t2Tt7jS8/Z7Sm3RMj2lgpqRaGvpdAFKoeMW4bAAAAwNqxnALCdufcr+SObzKzr/crIAyW+XR8vZNP8v3Wt8Zn65z7bbtLvkvH4FsuIc+6umf92LO105O0O76ZVI9C1cJQE3GkkTjUaBRpLA41WYr1lG0bNFWO20m/2l3pjysIqHNceZbw01oPAAAAAAtbTgHhM2b2bOfcX0iSmT1T0qf7GxZOxMH5pupz8z1b1rNEvj2be7sVPlverLV8Wm5MeyY/lr19ru1oM9HpI1WVAlMpDFS2QKWwvd55OdfFPr8thYHiwFQKenTDz+1nx3HAcmkAAAAAUKQFCwhm9pDaPcJ/zszm0/1I0gFJv7AqEWJR54zV9OStGxSlA+jDtOt8ZEFrvfI4t/RZnK5vni111rFcWtBepi1b1i1b97y9BFu7e35opo3lWKeOVIv9JQAAAAAA+m6xHgjTqxYFTtrFU2N6z2PGig4DAAAU5e1vLzoCAMCQWLCA4JxrZvtmdr6kXV33f6x/YQEAAGBZrrii6AgAAENiOcs4vlfSXkn/KD8XnuSHNlBAAAAAKNoXvuC3FBIAAH22nEkU90k63znnlrwTq+4b3zisj370HiYYBABgSL3gfX5aqg++5E8LjgRA0ebmEj3zmZt18cUMcUZ/LKeAcJuksyV9q8+x4CR87WuH9P73363Nm+OiQwEAAAX4oR/MS5JuvfX+giMBULS77prThReOUkBA3yyngPBHkm4zs+9LmpVficE55y7pa2RYtomJSNPTpaLDAAAABYjjQJL4uwAAzc87lUpB0WFgHVtOAeF9kl4i6Wtqz4EAAAAAAFhDgsAURQxtRv8sp4DwA+fcf+17JAAAAACAk2Ymlcv0QED/LKeAcLuZ/bGk/yE/hEGS5JxjFQYAAICCvfeCVxcdAoA1wswPbwb6ZTl/usbT7dNz51jGEQAAYA341/Fzig4BwBphZopjhjCgf5YsIDjnXrAagQAAAODEPer+2yRJX914WcGRAFgLKpWw6BCwji1ZQDCzP+h13jn3spUPBwAAACfiR//ljyRRQAAgNRpOk5MMYUD/LOdP12dy+xVJN0i6sz/hAAAAAABOlHNOzaZjSVf01XKGMHwkf2xmH5T0ub5FBAAAAAA4ITMzibZvLysMmQMB/XMya3ycJmnzSgcCAAAAADg5jYbTxERcdBhY55YzB8JD8qsuSL7g8KCk1/czKAAAAADA8jWbTuPjzH+A/lrOn7Dp3H7inHML3gkAAIBV9Z49byg6BABrwMyM0/bt5aLDwDq3nDkQmma2RdIpkiIzy85/oc+xAQAAYAnfH9lVdAgA1oAwlC67bLzoMLDOLWcIw9slPV/SNyU109NO0nV9jAsAAADL8Oh7/48k6e82X1VwJACKlCTSmWfWig4D69xyhjA8S9LZzrmZfgcDAACAE3PDtz8kiQICMOyShCUc0X/LWYXhXyWF/Q4EAAAAAHDiksQpCEyVyskssgcs33J6IByS9Pdm9r8kzWYnnXOv6ltUAAAAAIBlOXy4qTPOqCkIrOhQsM4tp4DwyfQDAAAAAFhjDh9u6olPnCo6DAyB5azC8EerEQgAAAAA4MTFsWn37pGiw8AQWE4PBAAAAKxR7774V4sOAUCBZmYShaHpoovGig4FQ4ACAgAAwAB7oLql6BAAFOjhh+f17GdvUbnMBIroP/6UAQAADLB9d31K++76VNFhAChIsylt3szyjVgdSxYQzOyTZjaRO540s0/0NywAAAAsx3Xf/Qtd992/KDoMAAXIlm98/OM3FB0KhsRyeiBsds4dyA6ccw9J2ta/kAAAAAAAS3n44YbOO6+unTsrRYeCIbGcAkJiZjuyAzM7pY/xAAAAAACW4dixRI9/PMs3YvUsZxLFN0v6vJl9VpJJulrST/UzKAAAAADA4oLAtGVLuegwMESWLCA45z5hZo+R9Nj01C865+7rb1gAAAAAgIXMzSWKY9PevSzfiNWzYAHBzM5yzv2Lme1JT30n3W4xsy3OuX/of3gAAABYzDv3/nrRIQAowOHDTV1++bimp1mBAatnsR4Ir5f0Uknv6XHNSbqqLxEBAABg2Q6WJosOAUABZmac9u3j33+srgULCM65l6bbK1cvHAAAAJyIJ9z5MUnSZ3Y+veBIAKwW55wkp0c/erzoUDBklpwDwczKkl4uaZ98z4O/kfRe59xsn2MDAADAEp5w58clUUAAhsncnNPUVMzyjVh1y1nG8QOSLpX0Xkl/mO5/oJ9BFeklL3mJNm3apN27dxcdCgAAAAAc58EHG7r++o1Fh4EhtJwCwh7n3Iucc59OPz8uac+STw2oF7/4xfrkJz9ZdBgAAAAA0ME5pwcfnFeSOP3kT+4oOhwMoeUUEL5qZo/ODszsUklf7l9Ixbrqqqs0NTVVdBgAAAAA0DI7m+juu+d01lk13XLLOapWw6JDwhBacg4ESRdKus3MsmUcT5P0T2b2ZUnOOXdJ36IDAAAAgCF36FBDDz/c1CtesUMve9kOmVnRIWFILaeA8MN9jwIAAAAn5abLbik6BAAroNFIdPRoovl5p0bDf8LQFARSsynddNMZuv76jRQPUKglCwjOuW+b2QWSsuUc/8Y5943+hgUAAIDlmA2rRYcA4AQlidPcnNPcXKLZ2USNhlQuBzr77Jq2bClr48ZY09MlTUxEGh2NdO65dW3dWi46bGBZyzj+rKSflvSX6ak/N7P3OOd+t6+RAQAAYEnXfffPJUm37rqx4EgALMQ5p0OHmjp2LJGZ5Jy0YUOsnTtr2r69rOuv36i9e8cVBPQuwNq2nCEML5P0GOfcYUkys7dL+oKkdVlA+LEf+zHt379fDzzwgHbs2KGbbrpJL33pS4sOCwAAoKd9d/0vSRQQgLVqZibRgw/Oa/v2il772u268MJR7dxZoViAgbScAoJJmssdz6fn1qUPf/jDRYcAAAAAYIAdODCvY8ec4tiUJE6veMVO/cRPbGf+Agy8BQsIZhY55xqSPii/CsNH00s3SPrAagQHAAAAAEVxzunIkaYOHWoqCPyEhpKpuw7gnL83Sfz8BrVaqNe//lSddlpNp51W1dRUXET4wIpbrAfCFyVd4pz7DTPbL2lfev4Vzrm/63tkAAAAAFCAw4cbOnw4kSTt2lXRs589pTPOqKlcDlQuByqVLN0GrXPZ+VIpUBwbvQ2wLi1WQGj9iXfOfVG+oAAAAAAAA885v1Ritmzi/LxfFSEMTbVaqLe97UxdccWERkaWM+obGA6L/duw0cxetdBF59y7+xAPAAAATsAbrviDokMABk6z6XTvvXOq10Nt2OCXTNy0qaStW0vas2dUF188ptFRCgdAt8X+rQgljWgdT5gIAAAAYH1qNp0OHmy0ehiYmcJQMjM1m07PfOYmveENpzPUADgBixUQ7nbO/eqqRQIAAIATdsO3/1iS9N/OeGHBkQBrx8GDDR0+3NRll41r166qNmyINTUVa3Q00uhoqE2bSjrttFrRYQIDZ1lzIAAAAGBtevS9n5NEAQHDI0mcmk3fq2B2NtHRo0lr0sLsWhiabrnlXO3bN1l0uMC6slgB4QmrFgUAAACAoZckTgcONDQ7m8g5KYoCBUG2TKIflmAmjYyEGh+PdNZZZV122bjOPLOmiYlIk5OxJiYi1WohQxOAPliwgOCce3A1AwEAAACw/s3MJHrwwXmFoSkI1Er0/aoI0mMeM6bdu0e0cWOpVRAYH480NhZpZCSkOAAUiKlFAQAAAPSVc04HDzZ15EhTc3OJXv7ynTrjjKpGRnxRoF4PNTISpnMUkKIAaxX/dgIAAAywubBcdAhAT82m0wMPzEvyww/OPbeuxz52XLt3j+jKKyfpRQAMIAoIAAAAA+wtl/120SEAPd1335we97hJXXvttM45p8aqB8A6QAEBAAAAwElzzml21unYsabm5vwqCEFgmpqK9da3nqVqNSw6RAArhAICAADAAPvRf36vJOkjZ/9kwZFgWBw71tSBAw0FgSkMTY2G04YNsS65ZEw7d1a0Y0dF09MlnX9+neIBsM5QQAAAABhgj3rg7yRRQEB/NJtOx44lOny4qTA0JYnT5GSs5z1vq849t67t2ys69dSKxsfjokMFsAooIAAAAADrzNGjTc3MJHLODzFIkvzWf8ykIDCZKf34SadO60UAACAASURBVA2zuQ2bTSczaceOim68cYsuvnhUp51W1aZNJSZABIYUBQQAAABgjcmS/WbTKUmcmk11bLP9ZtNJksLQFATtIkCj4XTVVROqVEKVy4HK5UCVSvtTLgeK40BxbOkn6NiOjITavLmsycmIYgGAFgoIAAAAQB80Gk5zc4kaDT+xYPZpNKQg8K3/+aTfuXZxwDmpXA5Uq4WqVELVaqFqNX88MhKpXg81OhpqbCzSxESkkZFIIyNh+om0cWPMsAIAK44CAgAAwAA7VBovOoShMDfn5wHwPQKy3gAubfn3H8m1igHNplMUmbZuLWt8PGol+uPjsSYn/XGtFqpebxcH2vuhqtWAln8Aaw4FBAAAgAH2jr3/oegQBlKjkWhuzvcQmJtzajTyQwHa8wJkvQLK5UCPfeyExsd963+97lv7q9WwNTwgGypQq4XaupXu/wDWHwoIAAAAGBrOOd1zz7wk6dRTK9qypazt28vatq2s0dGoNUwg6wWQ7U9NxSqVgoKjB4BiUUAAAAAYYC/8p9+WJP3xeT9XcCRrT7PpND/f7mkwP++HGDzqUSN6y1vO1M6dlaJDBICBQgEBAABggJ370NeKDqEvnHPp5IPt+Qby+9kqBFJ+BQJJMjnnrwWBND1d0s6dJe3YUdGF/7+9u4+SrK7vPP751lNX11NXPw7dPT0z3UPDAPPQAzPADIyIDAiomAAuEEQ0gGYFj1EwosSNGOQkMVFPjjEHjKLJOUYS92Q1WRajRjbJEhGyKEazPgV3BRWHcZ56prur6t7f/nHrVlfX9Ez3DN19u6rfr3PuuQ91H75V3DP073t/9/vblNPwcLu2bMkrmaQ3AQCcKBIIAAAAiIRzwWgD9aMTjI97isdNlYpTLhcWGYxV6w5MjzTQ0REsFwozRyAIahNMr1ODAAAWDgkEAAAAnLSwR0Cl4mrDFdYvx+Ox2nCF08dMD1eYSFhdnYG4Xve6Tl16abcGB9PKZOLRfTEAwFFIIAAAAOC4fN+pXHaamvI1MeFXG/4xeV4wckHYE6CjIxiqsLMzqe7upDo7k9UeAtO9AxqHKkwkeJUAAJoFCQQAAIAm9mJ730s+h3NOpZLT+LinctlXIhGTc5pRS6BYTGp0NKPR0axGRzNavz6j4eF25fO8JgAAKwUJBAAAgCb24a33zVivVJyOHPE0OelXiw0GxQVjMZOZ6iabcUyxmNDOnUVdeGFRa9YEiYGOjoTy+YTS6RhJAgAACQQAAICFFjy5D+oABE/yg/f+w6KB4dP96c+CVwGCfY7eT7JaEiAcbaC+QR/u5/tSMmkaHc1ow4ZstdBgQu3tMbW1TU/p9Mz1zs6kurqSEf1aAIBmQQIBAAAsmKCxGzRop7fVz13D+tH7uLoPGved3zlcw/lmxhVO4Xmm16f3NZt+Wi/pqPVQsN3JOat95pxUqfhqbw/qAiSTpmTSlEjElEiYUqlgOZWy6mcxJZOmVCrYlkrNth5TPB4c294eVzodq81X/+G7FY+bSn/wkep2egsAABYHCQQAAFYY54In4+Vy8N57ueyrUpESCVPQqA7m0/tPz81mPhWvb4SH6rvLB0/Lg6fm013ow/VgOR6ffqo+vTxzHo/bjPOEx08vB+eRgu8RrEuJRNDwTiaDfYKG+PT2eFy1xnkyKcXjMSUSqo4cEMYT7Bdeo/56jfvE46bu7qT6+lLKZpfoz6xnvxPM6UEAAFhkJBAAAFhGwqJ15bJ/VLf3mcszu7/XL0833q22LIXD5gXzfD6h7u6kVq1Kqb+/TWeckdXYWEHZbHxGY78+EVD/Dv3Ma0w39CXx9BsAgBZFAgEAgEUSvpMeNvDrG/DBPOjqXi47JZOxas8Aqb09pq6u5Czd2ae7tKdSwbvrqZTNeJc9/DyRCI7LZuPK56eH0QuH0mPoPAAAcKJIIAAAml5jwbmwEF39k/m5t8/8vL67fX3lemn2J+wzY5i+RtjoDxr48VoBu3Dq6kppbCynkZGMentT6u5Oqr09vsS/IAAAwNxIIAAAFsSxq8vP7ILvea42VSpBoz14v11HdYEP37dvvIZ0dLX64B12qxapC4rVtbXFZi1QFzTo47X14Ml9TG1tVmvgh+cIjw/PHa4f/dnshfDozo9Fd9ppUUcAAFghSCAAwCzqG7z13c/ru6E3PskOj5s+x+zz2SrIH11JPixQN7Pqe9AYdTMa2OE8LHBX3SIzV1sOC+IdrzHbeI7ZYpqtsn24HBbASySCKWzQB43ruBIJUzodUz6fUEdHMHV2JlQsJpTNBsPMNTbOg7nNuq1xnYY6VqwHH4w6AgDACkECAcBJaRyqLZx7ntP+/RWVy642FJrUOBTasRrHjcv1leBtlkby9L7h9RufWE/HG37m6panz1//PcKkQPg0OZkMnkq3t8drT67b2oLP0unpRm/YgJ6uJF+/PF1FPqjmPl3FffZt0/s3VrKfbVtjwbvZquCf2LbpWOqTAfXr9YmCsOo9AAAAWhcJhBZw5Iin8XEv6jBwgnw/6MLd2KW7vuE5/UTV6VhPV+dqPNcPuzb9VDuYu4adZ38a7maco3GotrCxGzZkk0nT5Zf3aHQ0c1RDurFae+Px08szh2abeezMyvKz79O477HPX3++MP6w6zpPtAE0hTe/OZjTEwEAsMhIIDS5tWvbtXFjTqkU1bSbTXt7TLlcUB09n0+oUEioUIgrk4mrvT0otDZXgzl8ct341Hi2xvSxG+JHj7c+v8Y5jWsAWBa+//2oIwAArBAkEJrcjh1Fff7zY1GHAQAAAABocTy2BgAAAAAAcyKBAAAAAAAA5sQrDE3uRz/6kb7yla/wPjoAACvUjnhckvQvFFHEAjl48KDe/OY3q1AoRB0KgGWGBEKTe/LJJ3X//fert7c36lAAAEAEamkDEghYAPv379cpp5yit7zlLVGHAmAZIoHQArq7u9Xf3x91GAAAAGhynufpt37rt5TP56MOBcAyRAIBAACgib3z6aclSR/eujXiSNDMKpWKfv7znyufz2sr9xKAYyCBAAAA0MR6JiaiDgFNrlKp6Kc//al27dql22+/XUNDQ1GHBGCZIoEAAAAArFDOOT3//PO66qqrdO+99ypeLcoJALMhgQAAAACsQOVyWXv27NHpp5+ue+65h+QBgDmRQAAAAABWkD179qhcLiuTyeiaa67R2972NrW3t0cdFoAmQAIBAACgif2fzs6oQ0CT8X1fH/nIR7Rr1y6ZWdThAGgiJBAAAACa2J+fcUbUIaCJlEoltbW1adu2bSQPAJywWNQBAAAAAFga+/bt06ZNm5TJZKIOBUATIoEAAADQxN7z1FN6z1NPRR0GmsCBAwfU2dmpD3zgA1GHAqBJkUAAAABoYvlSSflSKeow0ATGx8d16623qqenJ+pQADQpEggAAABAi6tUKkqlUrrggguiDgVAEyOBAAAAALS4ffv26cILL9TAwEDUoQBoYiQQAAAAgBbneZ62bdsWdRgAmhzDOAIAADSxb/E+O+ZQKpWUSCR08cUXRx0KgCZHAgEAAKCJPXzaaVGHgGVuz549uvrqq7Vq1aqoQwHQ5HiFAQAAAGhR5XJZyWRSt9xyS9ShAGgBJBAAAACa2PufeELvf+KJqMPAMnXw4EFt2bJF/f39UYcCoAXwCgMAAEATS3le1CFgGSuXy3rTm94UdRgAWgQ9EAAAAIAWNDExoXw+r61bt0YdCoAWQQIBAAAAaDG+72vPnj3auXOn0ul01OEAaBEkEAAAAIAW4pzTz372M5199tl6/etfH3U4AFoINRAAAACa2JMMzYcGe/bs0Y4dO/Sxj31MZhZ1OABaCAkEAACAJvY369dHHQKWkVKpJEm66667SB4AWHC8wgAAAAC0iH379umKK67Q8PBw1KEAaEEkEAAAAJrY/Y8/rvsffzzqMLAMOOdUKpW0ffv2qEMB0KJIIAAAAAAtYO/evTr77LO1e/fuqEMB0KJIIAAAAABNzjmnqakpvfWtb1VbW1vU4QBoUSQQAAAAgCZ3+PBhDQ8Pa9u2bVGHAqCFkUAAAAAAmtiBAwd08OBBah8AWHQM4wgAANDE/nlgIOoQEBHP8zQ+Pq5Dhw7p3nvv1ate9aqoQwLQ4kggAAAANLFH1q2LOgQssX379qlUKsnMdMYZZ+jaa6/Vq171KplZ1KEBaHEkEAAAAJpYm+dJkqbi8YgjwVL45S9/Kd/39eEPf1jnnnuuUqlU1CEBWEFIIAAAADSx33niCUnSe3fujDgSLCbnnPbu3av29nZ96lOf0jp6ngCIAEUUAQAAgGVs7969euGFFzQyMqKHHnqI5AGAyNADAQAAAFhmyuWyDh8+rCNHjiiVSumhhx7Sxo0bow4LwApHAgEAAABYYr7v1ybP8+R5nnzf1/j4uGKxmNLptDZt2qRt27bpFa94Bb0OACwLJBAAAACARVKpVGo9CeLxuJxz8jyvliRob29XNptVJpNRNptVe3u7rrnmGl100UWMqgBg2SGBAAAA0MS+OjQUdQgryuTkpMrlcq33gHNuRm+CcIrH44rFYhofH9f555+vc889Vy972cs0NDSkdDqtRCJBggBA0yGBAAAA0MRIICyN8fFxHTp0SIVCQWvWrKn1Hqifwp4E6XRa6XRaqVRKXV1d2rFjR9ThA8CCIIEAAADQxAqlkiTpYCoVcSStw/d9TU1N1SbnnHK5nN73vvfp0ksvVSaTiTpEAIgECQQAAIAmdvdTT0mS3rtzZ8SRNL9SqaS9e/cqFotp1apVGh0d1cjIiM466yzt2rVL+Xw+6hABIFIkEAAAALDieZ6nF154QVdffbVuu+029ff3Rx0SACw7JBAAAACwYjjntHfvXnmep3g8XtvmeZ6uv/563XXXXUok+BMZAGbDv44AAABoec457d+/X5OTkzr99NN14403qqOjQx0dHSoWiyoWi8rlclGHCQDLGgkEAAAAtCTnXG30hHg8rqGhId1222265JJLlEwmow4PAJoOCQQAAIAm9si6dVGHsCzt27dPU1NTWr9+vW666SZt3LhRmzdvVnt7e9ShAUDTIoEAAADQxP55YCDqEJYN3/dVKpW0f/9+ZTIZfeITn9DY2FjUYQFAyyCBAAAA0MR6JiYkSS+u4CfrYW8DSerr69MFF1yge+65R729vRFHBgCthQRCg0cffVRvf/vb5Xmebr31Vt19991RhwQAAHBM73z6aUnSe3fujDiSpeWc06FDhzQ5OalYLKYHH3xQmzdvro2sAABYeCQQ6niep9tvv11f/vKXtXr1am3fvl1XXXWVzjzzzKhDAwAAgKSDBw9qampKlUpFXV1dete73qWLLrpI+Xw+6tAAoOWRQKjzjW98Q6eeeqpGRkYkSddff72+8IUvkEAAAACIkHNOk5OTOnDggBKJhO69915deOGFymazUYcGACsKCYQ6zz//vIaGhmrrq1ev1hNPPBFhRAAAAM3F932Vy2VVKhV5niff9+X7fm05nMfjccViMcViMZnZjHM452pTeHyxWNQdd9yhSy+9VAMUjgSASJBAAAAAwEmbnJzU/v37FY/H5XmezEydnZ3q7e1VNptVJpNRLpdTNptVLpdToVBQNptVe3u72tralE6n1dbWdtwplUopFotF/VUBYMUjgVBncHBQP/nJT2rrzz33nAYHByOMCAAA4Pj+Zv36Jb1epVLRxMSEJicn5Xme2tradOedd+r8889XV1eX8vn8UT0KAACtgQRCne3bt+sHP/iBnn32WQ0ODupzn/ucPvvZz0YdFgAAwDE9uWrVol/DOafDhw/r0KFDamtr0/DwsEZHR7VlyxZdfPHF6ujoWPQYAADRI4FQJ5FI6GMf+5he+cpXyvM8/fqv/7rOOuusqMMCAAA4psHxcUnS87ncgp7XOae9e/fWahAMDg7q9ttv16tf/Wql0+kFvRYAoDmQQGhw5ZVX6sorr4w6DAAAgHm5/ZlnJEnv3blzQc7neZ7Gx8d16NAhbdu2TTfccIO2bNmi7u7uBTk/AKB5kUAAAABYgcLREsKpVCrVRkjYvHmzXvGKV+iGG25QKpWKOlQAwDJBAgEAAKDF+b6vAwcOaGpqasZoCR0dHerr69PAwIDWrFmjoaEhrVmzRtu3b6cQIgDgKCQQAAAAWozv+5qYmNDExITK5bJisZi2bt2q17zmNdq0aZO6urqUy+VIEgAATggJBAAAgCYVFjj0fV8vvPCCPM9TIpGQ7/saGhrShRdeqLGxMb3sZS9TT09P1OECAJocCQQAAIBlzDmnqakpTU5OampqSs65WpLAOaf/esYZ6uvr03W7dmloaEh9fX0688wz1dfXF3XoAIAWQwIBAABgGXLOaf/+/ZqYmFBfX582btyo0dFRrV+/XoODgxoYGFBfX58SCf6cAwAsDf6PAwAAEAHnnDzPU6VSqU1TU1Pyfb/Ww6C3t1ef/OQndfrppx/7RN/8ZjAfG1uawAEAKxYJBAAAgCUwOTmpffv21ZIDnucpnU6rUCiot7dX3d3dGh4e1sjIiAYGBjQyMqLu7u65Cx3+5m8G88ceW/TvAABY2UggAAAALJKJiQkdOHBAsVhM6XRat912m84991ytXbtW+XxeyWQy6hABAJg3EggAAAAnwfO8WmHD8FUE3/cVj8cVj8clSel0Wtdff7127dqlsbExpVKpiKMGAODkkUAAAAA4jkqlosnJSZVKJU1OTiqRSMg5p1gsptWrV2vt2rXq7u5WZ2enisWi8vm8stmscrmc1q9fr66urqi/AgAAC4IEAgAAgIJEwcTEhMrlskqlUi1JEI/HNTQ0pOHhYZ166qnasGGDTj31VK1atUqxWCzqsAEAWDIkEAAAQEtzzsn3fZXLZZXLZVUqFZVKJVUqFSUSCcXj8drIB8PDwxocHNTq1as1MDCgtWvXamxsbHkPlXj//VFHAABYIZbx/w0BAAACzjlVKhV5nifP82qjGDSuJxIJxWIxmZmcc7WhElOplIrFok455RT19PSov79fg4OD6u3tVVdXl/r7+zUwMDD3iAfL0c6dUUcAAFghSCAAAIBl7eDBg5qYmFA6nVYul1OxWFQ2m1U+n1cul1M+n6/VHigWiyoUCioUCsrn87XldDod9ddYPI8/HsxJJAAAFhkJBAAAsCw557Rv3z6Vy2V98IMf1O7du5uzh8Bie+97g/ljj0UaBgCg9ZFAAAAASyasQRC+cuD7/oxlz/NkZkomk6pUKtq8ebPe8Y53aOPGjVGHDgDAikcCAQAAnLSwxkBYlLCxLoHv+7WRDJxzamtrU2dnp9rb25XNZmvDHYavIuTzeXV0dGjt2rUaGRlRZ2dn1F8RAABUkUAAAADz5pzT5OSkDhw4oFgsJs/zlE6n1dvbq56enhl1CcL6A5lMRul0WtlsVtu2bWvtegQAALQwEggAAGBOnudpz549cs7plFNO0a5du3TjjTdq9erVJAQAAFghSCAAALCChEMbOufk+35tXr/snFO5XNbU1JTi8bhisZhKpZIuu+wy3Xnnnert7Y36a6DeRz8adQQAgBWCBAIAAMtYWGMgLD4YTmGdATNTLBZTLBarjVAQzsNEQf15nHOKx+NKJpNKJpNqa2tTKpVSKpVSW1tbbb2vr08jIyMaHBxUf3+/Tj31VGWz2ch+BxzH2FjUEQAAVggSCAAALIH6QoP1U9igDxMAZnZUz4BsNqtisaiOjg51dXWpp6dHnZ2dyufztfoC6XRabW1tRy2HCYEwYZBMJhWPx6P+ObCQvvKVYL57d7RxAABaHgkEAAAWmHNOpVJJ+/fvVywWq41E0Nvbq66uLnV1dam7u7u2HiYCMpmMstlsbYSCTCaj9vZ2xWKxqL8SlrP77gvmJBAAAIuMBAIAACepVCrp4MGDKpVKSiQStWSB53nq7u7W2WefrZtvvlkbNmxQsVisvVoAAADQjEggAAAwB9/3deTIEZXLZZVKJXmep3g8rnQ6rYsuukhjY2O1YQx7enrU29urVCoVddgAAAALigQCAACz8DxPBw4cUKlUkiStW7dOq1ev1sDAgFavXq3Vq1fr/PPPVzKZjDhSAACApUECAQCABi+++KJ839e5556ryy67TDt27FBPT0/UYQEAAESKBAIAYNkol8uampo6avjBxuWTWQ9HOKgf6rB+XZJisZg8z1Mmk9FDDz2k4eHhJfvuwEl74IGoIwAArBAkEAAAkXDOaWpqShMTE5qYmKgVIRwZGVE8Hlc8Hq9tSyQStfXZ5vF4vDY8YSKRmPF5/Xo4xWKxWee9vb0aGBhQd3c3BQ/RPE4/PeoIAAArBAkEAMCS8n1fv/jFL+ScU09Pj7Zu3aotW7ZodHRUZ511Fq8KACfqb/82mL/mNdHGAQBoeSQQAAALrlKpaHJyUocPH5ZzTolEQs45+b6vSqWiLVu26KMf/ag6OjqiDhVofn/0R8GcBAIAYJGRQAAAnDDf9+V53lHToUOHlEwmlUqltG7dOm3atEmbNm1Sf3+/Ojs71dnZqUKhoHg8HvVXAAAAwAkigQAAqPUOCOee56lcLqtUKqlUKtXqEDjnVKlUZGbKZDLKZDLq7OxULpdToVBQZ2enrr76am3cuJEaAgAAAC2GBAIANJHDhw+rXC7L8zz5vl+bPM+rFQJsHGmgXv0IBWHCIFxPJpNKJpNKJBIqFApatWqV+vv7NTg4qFNOOaU29fb2Kp1OkyAAAABYYUggAMAy43me9u7dq3K5XBshIBaLyTmnYrGojRs3KpvNKpPJKJvNKpvNKpfLKZ1OK51Oq62tTW1tbUqlUrXXCcLkQJggaNwWJh4AAACAYyGBAACLLCwmODU1NaOHQNhgD3sAhK8P+L6vSy65RJs2bVJHR4cKhYLy+bwKhYLWrFmjtra2KL8OgOXmL/4i6ggAACsECQQAWADOOR06dEilUkmVSkWe580YeWD37t3asmWLisXiUT0F6tfDeSqVivorAWgWQ0NRRwAAWCFIIADAS1Qul7Vnzx4NDAzonHPOUXd3t3p6emq9BtavX6/BwcGowwTQqh5+OJhfd120cQAAWh4JBADQzNcHnHNHDU8Y9irwfV/xeHxG74JEIqFrr71Wd9xxh/L5fNRfBcBK86d/GsxJIAAAFhkJBAAtzTmnqakpjY+Py/f9WjHC2UYiSCaTtdcIwgKFHR0dKhaLKhaL6uzsVLFYVD6fV1dXl7q7u9Xd3a1isahYLBbxNwUAAAAWFwkEAE0tHMIw7CFQP4VFC3t6erRjxw5dffXV6unpUSqVqtUfCOsNpFIpRiEAAAAAjoMEAoCm4XmeSqVSbSqXy6pUKurq6lI+n1cul1OhUFChUFCxWFR3d7d+9Vd/VX19fVGHDgAAADQ9EggAlp3Dhw/r0KFDcs4pkQj+mfI8r9abYGRkRGvXrtW6deu0YcMGnXPOORFHDAAAALQ+EgjACuL7fq2r/2xd/sN6AGZWmyTNWA41bjtW9/9we1hzIFS/3rhcLBZ14403anh4WP39/err61N3d7dyuRyvGQBAo89/PuoIAAArBAkEoImFxf/qCwHWr/u+r8OHD9cKB0pSoVA4qrt/R0eHOjs7lc1mlUgkFIvFZkzxePyEtx1reT7H5nI5JZPJiH9dAGgSPT1RRwAAWCFIIADLgO/72rdvX63oXywWk5nVGv318/okgZkpmUwqkUgokUgomUweNZ1xxhm6/PLLNTQ0pFWrVjFaAAC0mk9/Opi/8Y1RRgEAWAFIIAB1wm79cz3Zn229fllS7Wl64+sAofqkgOd52r59u8477zytWrVKPT09KhaLRyUHGpdJBgAASCAAAJYKCQSsKL7va2Jiola9P6wFEHah931fXV1dSqVSRz3JD7eF83D4v8ahANva2mr71W+vn8J96j/n/X4AAAAAyxkJBLQs3/d15MgRHT58WJJqrwQMDQ1pYGBAnZ2d6urqUldXlzo6OpTL5bRhwwYNDg5GHDkAAAAALD8kELAslUqlGSMDNM5935ckxePxo14RCD83Mw0PD2tsbExbt27V6Oio1qxZQ3E+AAAAADgJJBCwLPi+r3K5rImJCU1OTiqbzaqzs1OZTKY25fN5ZTIZFQoF5XK52vZsNjtjv/ptiQS3OAAAAAAsBFpXeMkmJiZ08ODB2pN/59xRPQPC3gGzFRIMj+nq6tLIyIje8IY36OKLL6ZAIAAA8/HII1FHAABYIUggNLljjQ7QOFLAbK8BhHMzq40YIGnWEQPCa812zUKhoOuuu065XE7pdFrpdHrW4QQbCxKGowmEvQ3oLQAAwEnIZKKOAACwQtBiaxA2rMMprNQfLvu+P2Nb4z6e56lcLqtUKtWmcrmsqakpTU5OamJiYsY0OTmpqamp2n7hucrlssrlcu169SMGNF4vFotpz549SiQSisfjisfjteH+4vG40un0jK79uVxOuVxO+Xy+9ipA/cgBYSM/bODPNiJB/ZRIJBg9AACAqHz848H8rW+NNg4AQMtbcQmEBx54QA899NCsT+ql6S714fLJTuFT//ApvaTak/76J/6N28KGeGNvgGNNL774ou677z7ddNNNEfyaAAAgcn/1V8GcBAIAYJGtuATCpk2bdMUVV8x4Wl//1L4Zn6Sfd955UYcAAAAAAGhxKy6BsHPnTu3cuTPqMAAAAAAAaCqUuQcAAAAAAHMigQAAAAAAAOZkYYG/BT2p2SFJ31vwE+NYeiS9GHUQwALjvkar4t5GK+K+Rqvi3karOt05lz/RgxarBsL3nHPbFuncaGBmT/F7o9VwX6NVcW+jFXFfo1Vxb6NVmdlTJ3McrzAAAAAAAIA5kUAAAAAAAABzWqwEwoOLdF7Mjt8brYj7Gq2KexutiPsarYp7G63qpO7tRSmiCAAAAAAAWguvMAAAAAAAgDmRQAAAAAAAAHNakASCmXWZ2ZfN7AfVeecx9nvUzPab2d8txHVXEjO7rZ5XYwAACTdJREFU3My+Z2Y/NLO7Z/m8zcwern7+hJmtW/oogRM3j3v7nWb2XTN7xsy+amZro4gTOBFz3dd1+11jZs7MGCIMTWE+97aZ/afqv9vfMbPPLnWMwMmYx98ja8zsa2b2dPVvkiujiBM4EWb2KTP7hZn92zE+NzP74+p9/4yZnT3XOReqB8Ldkr7qnBuV9NXq+mw+JOmmBbrmimFmcUl/IukKSWdKusHMzmzY7RZJ+5xzp0r6iKTfX9oogRM3z3v7aUnbnHObJX1e0h8sbZTAiZnnfS0zy0t6u6QnljZC4OTM5942s1FJ75F0gXPuLEm/ueSBAidonv9u/7akv3LObZV0vaSPL22UwEn5tKTLj/P5FZJGq9ObJf3pXCdcqATCayV9prr8GUm/MttOzrmvSjq0QNdcSc6V9EPn3H8450qSPqfgN69X/9/g85IuMTNbwhiBkzHnve2c+5pz7kh19euSVi9xjMCJms+/2ZL0uwqSvZNLGRzwEszn3r5N0p845/ZJknPuF0scI3Ay5nNvO0mF6nKHpJ8uYXzASXHO/aOkXx5nl9dK+nMX+Lqkopn1H++cC5VAWOWc+1l1+eeSVi3QeREYlPSTuvXnqttm3cc5V5F0QFL3kkQHnLz53Nv1bpH0PxY1IuClm/O+rnYRHHLO/felDAx4iebzb/Zpkk4zs/9lZl83s+M9+QKWi/nc2++X9Hoze07SI5LetjShAYvqRP8WV2K+Zzazr0g6ZZaP7qlfcc45M2NsSAALysxeL2mbpIuijgV4KcwsJunDkt4YcSjAYkgo6Ar7cgU9xv7RzDY55/ZHGhXw0t0g6dPOuT8ysx2S/sLMNjrn/KgDA5bSvBMIzrndx/rMzF4ws37n3M+qXR7orrawnpc0VLe+urpttn2eM7OEgq5Ve5cmPOCkzefelpntVpCsvMg5N7VEsQEna677Oi9po6THqm+anSLpi2Z2lXPuqSWLEjhx8/k3+zlJTzjnypKeNbPvK0goPLk0IQInZT739i2qvkvunPsXM0tL6hHtHjS3ef0tXm+hXmH4oqSbq8s3S/rCAp0XgScljZrZsJmlFBRu+WLDPvX/Da6V9A/OOXqCYLmb8942s62SHpB0Fe/Sokkc9752zh1wzvU459Y559YpqO1B8gDNYD5/j/w3Bb0PZGY9Cl5p+I+lDBI4CfO5t/+fpEskyczOkJSWtGdJowQW3hclvaE6GsP5kg7UlSaY1UIlEH5P0qVm9gNJu6vrMrNtZvZn4U5m9k+S/lpBgb/nzOyVC3T9llataXCHpC9J+ncFFWC/Y2YfMLOrqrt9UlK3mf1Q0jt17JEwgGVjnvf2hyTlJP21mX3TzBr/hw4sK/O8r4GmM897+0uS9prZdyV9TdK7nHP0iMSyNs97+05Jt5nZtyT9paQ38rAOy52Z/aWkf5F0erX9fYuZ/YaZ/UZ1l0cUJHl/KOkTkt465zm57wEAAAAAwFwWqgcCAAAAAABoYSQQAAAAAADAnEggAAAAAACAOZFAAAAAAAAAcyKBAAAAAAAA5kQCAQCABWZm95jZd8zsmerwo+dVt/+ZmZ1ZXf6xmfWY2Toz+7dFjmedmf1a3fqYmV25mNc8Tiy9ZvaEmT1tZrvM7HVm9u9m9rXq8M9/PMfxj5hZ8SSv/Svh7w8AAE5cIuoAAABoJWa2Q9KrJZ3tnJsysx5JKUlyzt0aUVjrJP2apM9W18ckbVMw/vNSu0TSt8PfwswelXSbc+6fq58/dbyDnXMvJfHxK5L+TtJ3X8I5AABYseiBAADAwuqX9KJzbkqSnHMvOud+Kklm9piZbZvlmLiZfaLaa+Hvzay9uv+YmX292pPhb8yss/E81V4MP64ux83sQ2b2ZPWYt1TP/3uSdlV7Q7xb0gckXVddv87Msmb2KTP7RrVnwGtn+2Jm9m4z+7aZfcvMfm+OGNeb2aNm9q9m9k9mtsHMxiT9gaTXVq/9O5IulPTJatwvN7O/qx6fM7OHqtd7xsyuqW7/cTUpIzN7fTXmb5rZA2YWr24fN7MPVuP8upmtMrOdkq6S9KHq/utP8r8vAAArFgkEAAAW1t9LGjKz75vZx83sonkcMyrpT5xzZ0naL+ma6vY/l/Ru59xmSd+W9DtznOcWSQecc9slbZd0m5kNS7pb0j8558acc78v6b9Ieri6/rCkeyT9g3PuXEkXK2hkZ+tPbGZXSHqtpPOcc1sUJAKOF+ODkt7mnDtH0l2SPu6c+2bDte9V0OPgRufcuxq+y/uq32VT9dz/0BDPGZKuk3SBc25MkifpxurHWUlfr8b5jwp6ODwu6YuS3lW99o/m+C0BAEADXmEAAGABOefGzewcSbsUNMYfNrO7nXOfPs5hz1Yb15L0r5LWmVmHpKJz7n9Wt39G0l/PcfnLJG02s2ur6x0KkhOleRx3lZndVV1PS1oj6d/r9tkt6SHn3JHq9/zlsWI0s5ykndXl8Pi2OWJotFvS9eGKc25fw+eXSDpH0pPVa7RL+kX1s5KCVxWk4Pe89ASvDQAAZkECAQCABeac8yQ9JukxM/u2pJslffo4h0zVLXsKGsPHU9F0L8J03XZT8NT/S/U7m9nL5zifSbrGOfe9Ofabr5ik/dWeAYvFJH3GOfeeWT4rO+dcddkTf+8AALAgeIUBAIAFZGanm9lo3aYxSf/3RM/jnDsgaZ+Z7apuuklS+KT/xwqevkvStXWHfUnSfzazZDWW06qvIhySlK/br3H9S5LeZtVH+Wa2dZaQvizpTWaWqe7TdawYnXMHJT1rZq+r7mtmtuWEfoDgereHK2FthTpflXStmfWF8ZjZ2jnO2fi9AQDACSCBAADAwspJ+oyZfdfMnpF0pqT3n+S5blZQj+AZBYmID1S3/6GCRMHTknrq9v8zBSMM/G8LhoZ8QMHT92ckedWigu+Q9DVJZ4ZFFCX9rqSkpGfM7DvV9Rmcc48qqCHwlJl9U0Fdg+PFeKOkW8zsW5K+o6B+wom4T1Knmf1b9RwXN8TzXUm/Lenvq9f+soIClsfzOUnvqhaKpIgiAAAnyKZ7+AEAAAAAAMyOHggAAAAAAGBOJBAAAAAAAMCcSCAAAAAAAIA5kUAAAAAAAABzIoEAAAAAAADmRAIBAAAAAADMiQQCAAAAAACY0/8HDVYlAouiqBsAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<Figure size 1296x504 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0: 0.020*\"venezia\" + 0.015*\"sera\" + 0.015*\"cuore\" + 0.010*\"marcare\" + 0.010*\"olga\" + 0.009*\"letizia\" + 0.008*\"salso\" + 0.007*\"livia\" + 0.007*\"bisognare\" + 0.007*\"notizia\" + 0.007*\"ragione\" + 0.007*\"abbracciare\" + 0.007*\"pranzare\" + 0.006*\"baciare\" + 0.006*\"restare\" + 0.006*\"martedì\" + 0.006*\"ottavio\" + 0.006*\"frizzo\" + 0.006*\"salsomaggiore\" + 0.006*\"trieste\"\n",
      "\n",
      "1: 0.024*\"amicare\" + 0.024*\"senilità\" + 0.019*\"librare\" + 0.016*\"parigi\" + 0.016*\"romanzare\" + 0.012*\"joyce\" + 0.011*\"zeno\" + 0.011*\"villa\" + 0.011*\"articolare\" + 0.010*\"traduzione\" + 0.010*\"trieste\" + 0.009*\"larbaud\" + 0.009*\"scrivere\" + 0.008*\"maestro\" + 0.008*\"coscienza\" + 0.008*\"piacere\" + 0.008*\"editore\" + 0.008*\"copiare\" + 0.008*\"italia\" + 0.007*\"saluto\"\n",
      "\n",
      "2: 0.022*\"cuore\" + 0.018*\"notte\" + 0.016*\"fabbricare\" + 0.014*\"tchap\" + 0.013*\"marcare\" + 0.011*\"sera\" + 0.010*\"fumare\" + 0.010*\"testare\" + 0.009*\"operare\" + 0.009*\"domenica\" + 0.009*\"macchina\" + 0.009*\"trieste\" + 0.008*\"signorina\" + 0.008*\"leggere\" + 0.008*\"bepi\" + 0.008*\"titina\" + 0.008*\"abbracciare\" + 0.008*\"pomeriggio\" + 0.007*\"venezia\" + 0.007*\"charlton\"\n",
      "\n",
      "3: 0.053*\"trieste\" + 0.018*\"prezzolini\" + 0.016*\"parigi\" + 0.012*\"scheifling\" + 0.010*\"crémieux\" + 0.010*\"londra\" + 0.010*\"scrivere\" + 0.010*\"fotografia\" + 0.009*\"ufficiare\" + 0.008*\"roma\" + 0.008*\"violare\" + 0.008*\"riccione\" + 0.007*\"giornale\" + 0.007*\"bambino\" + 0.007*\"carta\" + 0.007*\"pensiero\" + 0.007*\"affare\" + 0.007*\"lingua\" + 0.007*\"scuola\" + 0.007*\"rispondere\"\n",
      "\n",
      "4: 0.015*\"dottore\" + 0.014*\"malattia\" + 0.013*\"dolore\" + 0.013*\"mamma\" + 0.011*\"febbre\" + 0.011*\"speranza\" + 0.009*\"bisognare\" + 0.009*\"livia\" + 0.009*\"maritare\" + 0.009*\"dubbio\" + 0.009*\"occhio\" + 0.008*\"fondere\" + 0.008*\"coiazzi\" + 0.008*\"baciare\" + 0.007*\"desiderio\" + 0.007*\"sigaretta\" + 0.007*\"convinzione\" + 0.007*\"pensiero\" + 0.007*\"ideare\" + 0.007*\"senso\"\n",
      "\n",
      "5: 0.031*\"cuore\" + 0.026*\"marcare\" + 0.022*\"sera\" + 0.020*\"sabato\" + 0.017*\"domenica\" + 0.016*\"mattina\" + 0.016*\"settimana\" + 0.016*\"olga\" + 0.014*\"fabbricare\" + 0.014*\"milano\" + 0.014*\"notte\" + 0.012*\"trieste\" + 0.012*\"livia\" + 0.011*\"abbracciare\" + 0.010*\"leggere\" + 0.010*\"venezia\" + 0.009*\"piacere\" + 0.009*\"notizia\" + 0.009*\"giornata\" + 0.009*\"baciare\"\n",
      "\n",
      "6: 0.016*\"cuore\" + 0.015*\"baciare\" + 0.013*\"livia\" + 0.012*\"viaggiare\" + 0.011*\"fabbricare\" + 0.011*\"londra\" + 0.010*\"sera\" + 0.010*\"addio\" + 0.009*\"ritornare\" + 0.009*\"trieste\" + 0.009*\"mattina\" + 0.008*\"letizia\" + 0.008*\"scriverti\" + 0.008*\"notte\" + 0.008*\"affare\" + 0.007*\"titina\" + 0.007*\"bisognare\" + 0.007*\"curare\" + 0.007*\"giornata\" + 0.007*\"domenica\"\n",
      "\n",
      "For 7 topic the average silhouette score is : 0.5198586888822634\n",
      "\n",
      "Coherence: 0.4213633920958399\n",
      "\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1296x504 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0: 0.035*\"sera\" + 0.025*\"venezia\" + 0.024*\"cuore\" + 0.020*\"sabato\" + 0.015*\"marcare\" + 0.014*\"letizia\" + 0.014*\"domenica\" + 0.012*\"olga\" + 0.012*\"trieste\" + 0.010*\"giovedì\" + 0.010*\"mattina\" + 0.009*\"bepi\" + 0.009*\"settimana\" + 0.008*\"notizia\" + 0.008*\"martedì\" + 0.008*\"salutami\" + 0.008*\"frizzo\" + 0.007*\"arrivare\" + 0.007*\"pranzare\" + 0.007*\"treno\"\n",
      "\n",
      "1: 0.025*\"amicare\" + 0.025*\"senilità\" + 0.021*\"librare\" + 0.016*\"romanzare\" + 0.013*\"parigi\" + 0.012*\"joyce\" + 0.012*\"zeno\" + 0.011*\"articolare\" + 0.011*\"villa\" + 0.011*\"traduzione\" + 0.010*\"larbaud\" + 0.009*\"scrivere\" + 0.009*\"coscienza\" + 0.008*\"maestro\" + 0.008*\"editore\" + 0.008*\"piacere\" + 0.008*\"copiare\" + 0.007*\"italia\" + 0.007*\"copia\" + 0.007*\"libro\"\n",
      "\n",
      "2: 0.020*\"cuore\" + 0.017*\"fabbricare\" + 0.016*\"marcare\" + 0.016*\"notte\" + 0.013*\"signorina\" + 0.012*\"titina\" + 0.009*\"fumare\" + 0.009*\"domenica\" + 0.009*\"operare\" + 0.009*\"scrivere\" + 0.008*\"coiazzi\" + 0.008*\"caldaia\" + 0.008*\"tchap\" + 0.008*\"restare\" + 0.008*\"abbracciare\" + 0.007*\"trieste\" + 0.007*\"febbre\" + 0.007*\"grigio\" + 0.007*\"dottore\" + 0.007*\"lavoro\"\n",
      "\n",
      "3: 0.049*\"trieste\" + 0.029*\"parigi\" + 0.020*\"londra\" + 0.018*\"prezzolini\" + 0.012*\"crémieux\" + 0.012*\"bambino\" + 0.010*\"violare\" + 0.010*\"venezia\" + 0.009*\"riccione\" + 0.008*\"cuore\" + 0.008*\"correre\" + 0.008*\"signorina\" + 0.007*\"rendere\" + 0.007*\"scrivere\" + 0.007*\"ufficiare\" + 0.007*\"viaggiare\" + 0.006*\"baciare\" + 0.006*\"settimana\" + 0.006*\"direttore\" + 0.006*\"roma\"\n",
      "\n",
      "4: 0.018*\"trieste\" + 0.016*\"macchina\" + 0.015*\"fotografia\" + 0.009*\"operare\" + 0.009*\"pitturare\" + 0.009*\"giornale\" + 0.008*\"maritare\" + 0.008*\"sigaretta\" + 0.008*\"speranza\" + 0.008*\"pensiero\" + 0.008*\"salute\" + 0.007*\"nome\" + 0.007*\"malattia\" + 0.007*\"saluto\" + 0.007*\"baciare\" + 0.007*\"donna\" + 0.007*\"famiglia\" + 0.006*\"leggere\" + 0.006*\"storia\" + 0.006*\"simpatia\"\n",
      "\n",
      "5: 0.028*\"cuore\" + 0.023*\"notte\" + 0.022*\"marcare\" + 0.016*\"milano\" + 0.013*\"abbracciare\" + 0.012*\"mattina\" + 0.012*\"olga\" + 0.012*\"settimana\" + 0.011*\"bagnare\" + 0.011*\"trovare\" + 0.010*\"livia\" + 0.010*\"leggere\" + 0.009*\"baciare\" + 0.009*\"giornata\" + 0.009*\"trieste\" + 0.009*\"notizia\" + 0.009*\"testare\" + 0.009*\"porro\" + 0.008*\"festa\" + 0.008*\"domenica\"\n",
      "\n",
      "6: 0.018*\"cuore\" + 0.014*\"fabbricare\" + 0.012*\"livia\" + 0.011*\"sera\" + 0.011*\"viaggiare\" + 0.011*\"letizia\" + 0.011*\"mattina\" + 0.010*\"londra\" + 0.010*\"baciare\" + 0.009*\"bisognare\" + 0.009*\"giornata\" + 0.008*\"scriverti\" + 0.008*\"trieste\" + 0.008*\"curare\" + 0.007*\"leggere\" + 0.007*\"ritornare\" + 0.007*\"piacere\" + 0.007*\"notte\" + 0.007*\"addio\" + 0.006*\"domenica\"\n",
      "\n",
      "7: 0.014*\"baciare\" + 0.013*\"livia\" + 0.009*\"venezia\" + 0.009*\"addio\" + 0.009*\"restare\" + 0.008*\"dolore\" + 0.008*\"bisognare\" + 0.008*\"notte\" + 0.008*\"occhio\" + 0.007*\"madre\" + 0.007*\"cuore\" + 0.007*\"ritornare\" + 0.007*\"abbracciare\" + 0.007*\"affare\" + 0.007*\"padre\" + 0.006*\"notizia\" + 0.006*\"donna\" + 0.006*\"viaggiare\" + 0.006*\"olga\" + 0.006*\"maritare\"\n",
      "\n",
      "For 8 topic the average silhouette score is : 0.5040126353351535\n",
      "\n",
      "Coherence: 0.3768908217802961\n",
      "\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1296x504 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0: 0.030*\"sera\" + 0.025*\"cuore\" + 0.021*\"venezia\" + 0.014*\"olga\" + 0.013*\"sabato\" + 0.012*\"domenica\" + 0.012*\"letizia\" + 0.011*\"marcare\" + 0.011*\"giovedì\" + 0.010*\"frizzo\" + 0.009*\"trieste\" + 0.009*\"bepi\" + 0.008*\"operare\" + 0.008*\"mattina\" + 0.008*\"abbracciare\" + 0.008*\"settimana\" + 0.007*\"martedì\" + 0.007*\"fumare\" + 0.007*\"pranzare\" + 0.007*\"testare\"\n",
      "\n",
      "1: 0.023*\"senilità\" + 0.022*\"amicare\" + 0.019*\"librare\" + 0.016*\"parigi\" + 0.015*\"romanzare\" + 0.012*\"joyce\" + 0.012*\"articolare\" + 0.011*\"zeno\" + 0.011*\"villa\" + 0.010*\"trieste\" + 0.010*\"traduzione\" + 0.009*\"larbaud\" + 0.009*\"scrivere\" + 0.008*\"coscienza\" + 0.008*\"maestro\" + 0.008*\"piacere\" + 0.008*\"editore\" + 0.008*\"copiare\" + 0.007*\"italia\" + 0.007*\"saluto\"\n",
      "\n",
      "2: 0.031*\"fabbricare\" + 0.024*\"notte\" + 0.018*\"cuore\" + 0.015*\"marcare\" + 0.014*\"macchina\" + 0.014*\"tchap\" + 0.013*\"domenica\" + 0.012*\"caldaia\" + 0.010*\"scrivere\" + 0.008*\"titina\" + 0.008*\"olga\" + 0.008*\"dirti\" + 0.008*\"pom\" + 0.007*\"signorina\" + 0.007*\"abbracciare\" + 0.007*\"salutami\" + 0.007*\"sera\" + 0.007*\"peschiera\" + 0.007*\"pitturare\" + 0.007*\"bottone\"\n",
      "\n",
      "3: 0.032*\"prezzolini\" + 0.031*\"trieste\" + 0.022*\"parigi\" + 0.013*\"bambino\" + 0.013*\"ufficiare\" + 0.013*\"fotografia\" + 0.012*\"consenso\" + 0.012*\"crémieux\" + 0.012*\"londra\" + 0.011*\"rendere\" + 0.011*\"signorina\" + 0.010*\"roma\" + 0.010*\"scrivere\" + 0.009*\"trattativa\" + 0.009*\"church\" + 0.008*\"riccione\" + 0.007*\"pratico\" + 0.007*\"stimare\" + 0.007*\"germania\" + 0.007*\"direttore\"\n",
      "\n",
      "4: 0.017*\"malattia\" + 0.011*\"fotografia\" + 0.010*\"dottore\" + 0.010*\"acqua\" + 0.010*\"fondere\" + 0.010*\"tono\" + 0.010*\"coiazzi\" + 0.010*\"pensiero\" + 0.009*\"uomini\" + 0.008*\"bravin\" + 0.008*\"sigaretta\" + 0.008*\"dr\" + 0.008*\"trieste\" + 0.008*\"vecchio\" + 0.008*\"leggere\" + 0.008*\"fuoco\" + 0.007*\"macchina\" + 0.007*\"ant\" + 0.007*\"vaso\" + 0.007*\"bambino\"\n",
      "\n",
      "5: 0.037*\"cuore\" + 0.020*\"notte\" + 0.017*\"settimana\" + 0.017*\"sera\" + 0.011*\"mattina\" + 0.011*\"trieste\" + 0.011*\"livia\" + 0.011*\"leggere\" + 0.010*\"giornata\" + 0.010*\"marcare\" + 0.009*\"letizia\" + 0.009*\"violare\" + 0.009*\"domenica\" + 0.009*\"gilda\" + 0.008*\"olga\" + 0.008*\"piacere\" + 0.008*\"sabato\" + 0.008*\"trovare\" + 0.007*\"notizia\" + 0.007*\"sonno\"\n",
      "\n",
      "6: 0.015*\"cuore\" + 0.011*\"fabbricare\" + 0.010*\"baciare\" + 0.010*\"sera\" + 0.010*\"livia\" + 0.010*\"letizia\" + 0.009*\"curare\" + 0.009*\"viaggiare\" + 0.009*\"londra\" + 0.009*\"giornata\" + 0.009*\"scriverti\" + 0.008*\"mattina\" + 0.007*\"piacere\" + 0.007*\"bisognare\" + 0.007*\"ritornare\" + 0.007*\"addio\" + 0.006*\"posto\" + 0.006*\"notte\" + 0.006*\"bagno\" + 0.006*\"leggere\"\n",
      "\n",
      "7: 0.015*\"baciare\" + 0.012*\"cuore\" + 0.012*\"livia\" + 0.011*\"marcare\" + 0.011*\"notizia\" + 0.009*\"olga\" + 0.009*\"dolore\" + 0.009*\"settimana\" + 0.009*\"restare\" + 0.008*\"affare\" + 0.008*\"milano\" + 0.007*\"occhio\" + 0.007*\"sera\" + 0.007*\"donna\" + 0.007*\"londra\" + 0.007*\"mamma\" + 0.007*\"notte\" + 0.006*\"voce\" + 0.006*\"padre\" + 0.006*\"dubbio\"\n",
      "\n",
      "8: 0.022*\"trieste\" + 0.019*\"venezia\" + 0.015*\"cuore\" + 0.014*\"livia\" + 0.013*\"marcare\" + 0.012*\"titina\" + 0.012*\"baciare\" + 0.011*\"notte\" + 0.011*\"mattina\" + 0.011*\"sera\" + 0.011*\"abbracciare\" + 0.010*\"bisognare\" + 0.010*\"viaggiare\" + 0.009*\"domenica\" + 0.008*\"ritornare\" + 0.007*\"leggere\" + 0.007*\"addio\" + 0.007*\"bagnare\" + 0.007*\"fabbricare\" + 0.007*\"compagnia\"\n",
      "\n",
      "For 9 topic the average silhouette score is : 0.5430431459845674\n",
      "\n",
      "Coherence: 0.3954755574909938\n",
      "\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1296x504 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAZIAAAEKCAYAAAA4t9PUAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADl0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uIDMuMC4wLCBodHRwOi8vbWF0cGxvdGxpYi5vcmcvqOYd8AAAIABJREFUeJzt3Xl4VOX5//H3Jys7CAREdgVUBGUJiwLaqrRYLShaBRFFQbQVl6pt7U/bWu232sW91o3d4oJrcUHcqVhZAiI7iKACKgSQRbaQ5P79MSd0moZkIJmcmeR+XddcmfOcc565h0tz5zyrzAznnHPucKWEHYBzzrnk5onEOedcuXgicc45Vy6eSJxzzpWLJxLnnHPl4onEOedcuXgicc45Vy6eSJxzzpWLJxLnnHPlkhbPyiUNAB4AUoGxZnZ3sfMjgL8AG4Kiv5nZWEnfB+6LuvQ4YIiZvSxpInAasD04N8LMFpYWR+PGja1Nmzbl/DbOOVe9zJ8/f7OZZZV1XdwSiaRU4GGgP7AemCdpmpktK3bps2Y2JrrAzN4DugT1NARWA29GXfILM3s+1ljatGlDTk7OYXwL55yrviR9Ect18Wza6gmsNrM1ZpYHPAMMOox6LgCmm9nuCo3OOedchYhnImkOrIs6Xh+UFXe+pEWSnpfUsoTzQ4Cni5X9X3DPfZIyKyhe55xzhyHszvZXgDZmdiLwFjAp+qSkZkBnYEZU8a+J9Jn0ABoCvyqpYkmjJeVIysnNzY1H7M4554hvItkARD9htOA/neoAmNkWM9sXHI4Fuher40LgJTPbH3XP1xaxD5hApAntf5jZ42aWbWbZWVll9hU555w7TPFMJPOA9pLaSsog0kQ1LfqC4ImjyEBgebE6hlKsWavoHkkCzgWWVHDczjnnDkHcRm2ZWb6kMUSapVKB8Wa2VNIdQI6ZTQOukzQQyAe2AiOK7pfUhsgTzcxiVU+RlAUIWAhcHa/v4JxzrmyqDjskZmdnmw//dc65QyNpvplll3Vd2J3tCe2d5RuZmrOu7Audc64ai+vM9mRmZjw150veX5VL4zoZnH5c07BDcs65hORPJAchiQeHdqVjs3pcM+VjPlm3LeyQnHMuIXkiKUXtzDTGj+hB47oZXDFxHp9v3hV2SM45l3A8kZQhq24mky7vSaEZl02Yy+bv9pV9k3POVSOeSGJwdFYdxo3owcYdexk5cR678/LDDsk55xKGJ5IYdWt1BA8N7cbiDdu5ZsoC8gsKww7JOecSgieSQ9C/Y1PuPLcT763M5daXllAd5uA451xZfPjvIRrWqzXfbN/LQ++uplmDGtxwZoewQ3LOuVB5IjkMN/bvwNfb93L/259yZL0aDOnZKuyQnHMuNJ5IDoMk7hrcmdyd+7j15SU0qZfpExadc9WW95EcpvTUFP4+rJtPWHTOVXueSMrBJyw655wnknLzCYvOuerOE0kF8AmLzrnqzBNJBfEJi8656soTSQXyCYvOueoorolE0gBJKyWtlnRLCedHSMqVtDB4jYo6VxBVPi2qvK2kOUGdzwb7wSeMYb1ac+3p7Xg2Zx0PvPNp2OE451zcxW0eiaRU4GGgP7AemCdpmpktK3bps2Y2poQq9phZlxLK/wTcZ2bPSHoUGAk8UpGxl5dPWHTOVSfxfCLpCaw2szVmlgc8AwwqT4WSBJwOPB8UTQLOLVeUcVA0YfHUDlnc+vIS3l2xMeyQnHMubuKZSJoD0Ruerw/Kijtf0iJJz0tqGVVeQ1KOpNmSipJFI2CbmRUNizpYnaFLT03hEZ+w6JyrBsLubH8FaGNmJwJvEXnCKNLazLKBi4H7JR1zKBVLGh0kopzc3NyKi/gQ+IRF51x1EM9EsgGIfsJoEZQdYGZbzKxoBt9YoHvUuQ3BzzXA+0BXYAvQQFJR387/1Bl1/+Nmlm1m2VlZWeX/NofJJyw656q6eCaSeUD7YJRVBjAEmBZ9gaRmUYcDgeVB+RGSMoP3jYE+wDKLjKd9D7gguOcy4J9x/A4VwicsOueqsrglkqAfYwwwg0iCmGpmSyXdIWlgcNl1kpZK+gS4DhgRlB8P5ATl7wF3R432+hVwo6TVRPpMxsXrO1Qkn7DonKuqVB0mzWVnZ1tOTk7YYQAwZc4X3PrSEi7Kbsnd53cmMhDNOecSj6T5QV91qXw/kkrmOyw656oaTyQh8AmLzrmqxBNJCIomLG7yHRadc1VA2PNIqq2iCYvHN6vrExadc0nNE0mIfMKic64q8EQSsiZ1a/iERedcUvNEkgB8wqJzLpl5IkkQPmHROZesPJEkEN9h0TmXjHz4b4LxCYvOuWTjiSQB+YRF51wy8USSgHzConMumXgfSYLyCYvOuWThiSSB+YRF51wy8ESS4HzConMu0XkiSQI+YdE5l8g8kSQJn7DonEtUcU0kkgZIWilptaRbSjg/QlKupIXBa1RQ3kXSR8E2vIskXRR1z0RJa6Pu6RLP75BIfMKicy4RxW34r6RU4GGgP7AemCdpWtTe60WeNbMxxcp2A5ea2aeSjgLmS5phZkVDl35hZs/HK/ZE5hMWnXOJJp7zSHoCq81sDYCkZ4BBQPFE8j/MbFXU+68kbQKyAB8Di09YdM4llng2bTUH1kUdrw/Kijs/aL56XlLL4icl9QQygM+iiv8vuOc+SZklfbik0ZJyJOXk5uaW42sknqIJi6d2yOLWl5fw7oqNYYfknKvGwu5sfwVoY2YnAm8Bk6JPSmoGPAlcbmZFvcu/Bo4DegANgV+VVLGZPW5m2WaWnZWVFa/4Q+MTFp1ziSKeiWQDEP2E0SIoO8DMtphZ0cSIsUD3onOS6gGvAbea2eyoe762iH3ABCJNaNWST1h0ziWCeCaSeUB7SW0lZQBDgGnRFwRPHEUGAsuD8gzgJWBy8U71onskCTgXWBK3b5AEik9YzN3pExadc5UrbonEzPKBMcAMIgliqpktlXSHpIHBZdcFQ3w/Aa4DRgTlFwKnAiNKGOY7RdJiYDHQGPhDvL5DsoiesHjZ+Lls37M/7JCcc9WIqsNchOzsbMvJyQk7jLibuSqXUZPm0aVlAyZf0YuaGalhh+ScS2KS5ptZdlnXhd3Z7irQaR2yuO+iLuR88S0/nTKfvHyf/e6ciz9PJFXMOScexR/P68z7K3O5cepCCgqr/hOncy5cvrFVFTS0Zyt27NnPXdNXULdGOn88rxORsQnOOVfxPJFUUVeddgzb9uznkfc/o0GtdH414LiwQ3LOVVGeSKqwX/7wWHYEyaR+zXSuPu2YsENyzlVBnkiqMEncMagTO/bmc/f0FdSrkc7FvXxdLudcxfJEUsWlpoh7fnISO/fu59aXF1O3Rho/PumosMNyzlUhMY3aklRT0rHxDsbFR0ZaCo8M606P1g25cepC3l+5KeyQnHNVSJmJRNKPgYXAG8FxF0nTSr/LJZqaGamMHZFN+yZ1ufof85n3+dawQ3LOVRGxPJHcTmRhxG0AZrYQaBvHmFyc1KuRzuSRPTmqfk2umDiPpV9tDzsk51wVEEsi2W9mxX/j+Cy3JNW4TiZPjupF3cw0Lhs/lzW534UdknMuycWSSJZKuhhIldRe0kPAv+Mcl4uj5g1q8uSoXhQaDB83l6+27Qk7JOdcEoslkVwLnADsA54CtgM3xDMoF3/HZNVh8hU92bFnP8PHzWHLd778vHPu8JSaSCSlAneY2a1m1iN43WZmeyspPhdHnZrXZ+xl2az/dg8jJsxj515fft45d+hKTSRmVgD0raRYXAh6Hd2IRy7pxvKvdzBqUg579xeEHZJzLsnE0rT1saRpkoZLGlz0intkrtKcflxT7rnwJOZ+vpVrpixgf4EvP++ci10siaQGsAU4Hfhx8DonlsolDZC0UtJqSbeUcH6EpNyoXRBHRZ27TNKnweuyqPLukhYHdT4oX9a2Qgzq0pw7B3XinRWbuPm5Tyj05eedczEqc4kUM7v8cCoO+lceBvoD64F5kqaZ2bJilz5rZmOK3dsQ+B2QTWSo8fzg3m+BR4ArgTnA68AAYPrhxOj+2yW9W7N9z37+MmMl9Wum8/uBJ/jy8865MsUys72FpJckbQpeL0hqEUPdPYHVZrbGzPKAZ4BBMcb1Q+AtM9saJI+3gAGSmgH1zGy2RfYIngycG2OdLgY/+94xjD71aCZ/9AX3vrUq7HCcc0kglqatCcA04Kjg9UpQVpbmwLqo4/VBWXHnS1ok6XlJLcu4t3nwvqw63WGSxK/POo6Lslvy0LurGfvBmrBDcs4luFgSSZaZTTCz/OA1EciqoM9/BWhjZicSeeqYVEH1Imm0pBxJObm5uRVVbbUgiT8O7syPOh/JH15bztR568q+yTlXbcWSSLZIukRSavC6hEjne1k2AC2jjlsEZQeY2RYzK5oJNxboXsa9G4L3B60zqu7HzSzbzLKzsioq71UfqSnivou60K99Y255cRHTF38ddkjOuQQVSyK5ArgQ+Ab4GrgAiKUDfh7QXlJbSRnAECJNZAcEfR5FBgLLg/czgB9IOkLSEcAPgBlm9jWwQ1LvYLTWpcA/Y4jFHYbMtFQeG96dLi0bcP0zC/ngU3+yc879rzITiZl9YWYDzSzLzJqY2blm9mUM9+UDY4gkheXAVDNbKukOSQODy66TtFTSJ8B1wIjg3q3AnUSS0Twis+uL1j3/GZGnl9XAZ/iIrbiqlZHGhBE9OTqrNqMnz2f+F9+GHZJzLsEoMviplAukScD1ZrYtOD4CuMfMrqiE+CpEdna25eTkhB1GUtu0cy8XPvoRW3flMfXqkznuyHphh+ScizNJ880su6zrYmnaOrEoiQAEw3G7lic4l3ya1K3BkyN7UTMjleHj5vL55l1hh+ScSxCxJJKU4CkEODBZ0Pd6r4ZaNqzFP0b2Ir+gkEvGzeGb7b52p3MutkRyD/CRpDsl/YHIXiR/jm9YLlG1b1qXiZf35NtdeQwfN4dvd+WFHZJzLmSxdLZPBgYDG4mM3BpsZk/GOzCXuE5q2YAnLsvmi627GTFhLt/tyw87JOdciGJZIuUY4DMz+xuwBDhTUoO4R+YS2inHNObhi7ux5KsdjJ7sy887V53F0rT1AlAgqR3wGJGJgk/FNSqXFPp3bMpff3Ii//5sC9c9/TH5vvy8c9VSLImkMJgTMhj4m5n9AmhWxj2umjivawtu/3FH3ly2kV+9sNiXn3euGopl9NV+SUOJzCL/cVCWHr+QXLIZ0act2/fkc9/bq6hXM43fntPRl593rhqJJZFcDlwN/J+ZrZXUFvDOdvdfrjujHdv37Gf8h2upXzOdG87sEHZIzrlKEsvGVsuILF9SdLwW+FM8g3LJRxK3nX08O/bu5/63P6V+zXQu79M27LCcc5XAJxa6CpOSIu4e3Jmde/fz+1eWUa9GOud3j2UPNOdcMouls925mKWlpvDAkK70adeIX76wiDeXfhN2SM65OIs5kUiqFc9AXNVRIz2Vx4dn07l5fcY89TH/Xr057JCcc3EUy4TEUyQtA1YExydJ+nvcI3NJrXZmGhMv70GbxrW4cnIOC9dtK/sm51xSiuWJ5D7ghwS7IprZJ8Cp8QzKVQ0NamXw5MheNKqTyYgJc1m1cWfYITnn4iCmpi0zK75pt6+H4WLStF4N/jGyFxmpKQwfN4d1W3eHHZJzroLFkkjWSToFMEnpkm7mP1viOlemVo1q8eTIXuzdH1l+ftMOX37euaoklkRyNXAN0BzYAHQJjsskaYCklZJWS7qllOvOl2SSsoPjYZIWRr0KJXUJzr0f1Fl0rkkssbhwHXtkXSZe3oPcnfu4dPxctu325eedqypiWUZ+s5kNM7OmwZ7tl5jZlrLuk5QKPAycBXQEhkrqWMJ1dYHrgTlRnznFzLqYWRdgOLDWzBZG3Tas6LyZbSrzW7qE0LXVETw+PJs1ubu4fOI8duf58vPOVQWxjNqaFL1svKQjJI2Poe6ewGozW2NmecAzwKASrruTyEz5g7V3DA3udVVA3/aNeXBoFz5Zt42hj89m+dc7wg7JOVdO8dyzvTkQ3Um/Pig7QFI3oKWZvVZKPRcBTxcrmxA0a/1Gvjpg0hnQqRkPX9yNdd/u4ZyHZvGHV5f55ljOJbHQ9myXlALcC9xUyjW9gN1mtiSqeJiZdQb6Ba/hB7l3tKQcSTm5ubnlDddVsLM6N+Pdm07jwuyWjJ21ljPvmcn0xV9j5svQO5ds4rln+wYim2AVaRGUFakLdALel/Q50BuYVtThHhhCsacRM9sQ/NxJZIOtniV9uJk9bmbZZpadlZUVQ7iusjWolcFdgzvzwk9P4YjaGfx0ygKumDiPL7f4EGHnkkmse7afz6Hv2T4PaC+praQMIklhWlS9282ssZm1MbM2wGxgoJnlwIEnlguJ6h+RlCapcfA+HTiHyPa/Lol1b30Er4zpw21nH8/ctVvpf99MHnrnU/bl+3Ql55JBrGttrQBeJJIIvpPUqqwbgl0VxwAziMw7mWpmSyXdIWlgDJ95KrDOzNZElWUCMyQtAhYSecJ5Isbv4BJYWmoKo/odzds3ncYZxzfhnrdWcdYDH/g6Xc4lAZXVJi3pWuB3RJ5ICgABZmYnxj+8ipGdnW05OTlhh+EOwXsrN/G7fy7ly627ObfLUdx6dkey6maGHZZz1Yqk+WaWXeZ1MSSS1UCvWOaOJCpPJMlp7/4C/v7eah6duYbM9BR++cNjubhXa1JTfKCec5Uh1kQS0xIpwPbyh+TcoamRnsqNPziW6Tf048QW9fnNP5cy+O8fsni9/+foXCKJ5YlkHHAs8Bqwr6jczO6Nb2gVx59Ikp+ZMe2Tr7jz1eVs3bWP4b1bc9MPj6VejfSwQ3Ouyor1iSSW+SBfBq+M4OVcpZPEoC7N+d6xTbjnzZVMnv0Fry/5htvOPp6BJx2Fz0t1LjxlPpEcuFCqZWZJOcDfn0iqnkXrt3HrS0tYvGE7fds15o5BJ3B0Vp2ww3KuSqmwPhJJJ/sOiS7RnNiiAS9f04c7Bp3AJ+u2MeD+D7j3rVXs3e9zT5yrbLF0tt+P75DoElBqirj05Da8c/NpnNX5SB5851N+eP+/mLnKl8RxrjL5Doku6TWpW4MHhnRlyqhepEpcNn4u10xZwDfbfQMt5yqD75Doqow+7Roz/YZ+3Ni/A28t38gZ97zPuFlryS8oDDs056q0uO6Q6Fxly0xL5boz2vPWz08lu01D7nx1GQP/9iELvvw27NCcq7JKTSTBLofDD2eHROfC1LpRbSZe3oNHhnVj6648zn/k3/z6xcW+xa9zcVBqIjGzAuDiSorFuQolibM6N+Ptm07jij5tmZqzjjPumcnz89f7vifOVaBYmrZmSfqbpH6SuhW94h6ZcxWkTmYavzmnI6+M6UurRrW4+blPuOjx2Xy6cWfYoTlXJcSyRMp7JRSbmZ0en5Aqnk9IdEUKC41nc9Zx9/QV7NqXz5WnHs11p7enZkZq2KE5l3AqbPXfqsATiStuy3f7uGv6Cp6fv57mDWry+4EncGbHpmGH5VxCqciZ7U0ljZM0PTjuKGlkRQTpXFga1cnkrz85ialXnUztzFRGTc7hysk5bNi2J+zQnEs6sfSRTCSyy+FRwfEq4IZ4BeRcZerZtiGvXdePW846jlmfbubMe2by6MzP2O9zT5yLWSyJpLGZTQUK4cAWujHNbJc0QNJKSasl3VLKdedLMknZwXEbSXskLQxej0Zd213S4qDOB+XLvrpySk9N4erTjuGtG0+lT7vG3D19BWc/+AFz124NOzTnkkIsy8jvktQIMABJvYlho6tgDsrDQH9gPTBP0jQzW1bsurrA9cCcYlV8ZmZdSqj6EeDK4PrXgQHA9Bi+h3OlanFELcZels1byzZy+7SlXPjYR1zQvQW/Pus4GtWp2G1+CwuNvfkF7N1fyN79BezZX8DeA6/ossKo8shx0bV79hew73/ujxx3a30EvzmnI/Vr+n4tLv5iSSQ3AtOAYyR9CGQBF8RwX09gtZmtAZD0DDAIWFbsujuBPwG/KKtCSc2AemY2OzieDJyLJxJXgfp3bEqfdo148J3VjP1gDW8v38gNZ7TnqAY1D/zy3lPCL/Z9+QXsyQt+mRe9zy9kb15BkDT+U5aXf3hNZymCmump1DjwSqFGeio101OplZFGw9qppAhe+ngDH322hfuHdKFHm4YV/C/k3H8rM5GY2QJJpxHZJVHASjPbH0PdzYls01tkPdAr+oJgPkpLM3tNUvFE0lbSx8AO4DYz+yCoc32xOpvHEItzh6RWRhq3nHUcg7s157aXl3D7K8X//vmPjNQUMqN+oddIT6FmeiqZ6anUr5nOkfUyI7/001KpmZEauTZ4XyMtJfIzPZXMqLIa6UXvU6mRkXLg/vRUxbSJ18dffsv1zyzkosc+4trT23Pt6e1IS41pjVbnDlksTyQQebpoE1zfTRJmNrk8HywpBbgXGFHC6a+BVma2RVJ34GVJJxxi/aOB0QCtWrUqT6iuGuvQtC7Pju7N0q92ABx4Coh+KkhNSbxuuq6tjuC16/ryu2lLeeCdT5m1ejP3X9SFlg1rhR2aq4LKTCSSngSOARbyn052A8pKJBuAllHHLYKyInWBTsD7wV9YRwLTJA00sxyC/eHNbL6kz4AOwf0tSqnzADN7HHgcIvNIyojVuYOSRKfm9cMO45DVrZHOvRd24bQOWdz20hJ+9MAH/OG8Tgzq4g/xrmLF8kSSDXS0Q5+5OA9oL6ktkV/2Q4hat8vMtgONi44lvQ/cbGY5krKArWZWIOlooD2wxsy2StoRdPjPAS4FHjrEuJyrVgZ1aU63Vkdww7MLuf6ZhcxclcsdgzpRJzPWBgnnShdLo+kSIk8LhyQYJjyGyByU5cBUM1sq6Q5JA8u4/VRgkaSFwPPA1WZWNBbzZ8BYYDXwGd7R7lyZWjasxbOje3P9Ge15+eMNnP3gByxcty3ssFwVcdAlUiS9QqQJqy6RPUjmEjQ3AZhZWckgYfgSKc79x7zPt3LDMwvZuGMvP+/fgatPOyYh+3lc+GJdIqW0Z9u/VmA8zrkE0aNNQ16/vh+3vrSYv8xYyb9W5XLfRV04qkHNsENzSeqgTVtmNrPoBawg8mRSF1gelDnnklT9muk8NLQrf/3JSSzesJ2zHviA6Yu/Djssl6RiWbTxQiLNWj8BLgTmSIplQqJzLoFJ4oLuLXjtun60blSLn05ZwC0vLGJ3Xn7YobkkE8t+JJ8A/c1sU3CcBbxtZidVQnwVwvtInCtdXn4h9729ikdnfkbbRrV5cGjXpBzy7CpWhS0jD6QUJZHAlhjvc84liYy0FH414DimjOrF7rwCzvv7hzzxrzUUFvoULFe2WBLCG5JmSBohaQTwGj7k1rkq6ZRjGjP9+n6cflwT/u/15Vw2YS6bduwNOyyX4GLaIVHSYKBvcPiBmb0U16gqmDdtOXdozIyn567jjleXUisjjT+ff6LvIFkNlbtpS1I7SX0AzOxFM7vRzG4EciUdU4GxOucSjCQu7tWKV6/ty5H1ajBqcg6//ecS9u6PaSsiV82U1rR1P5GVd4vbHpxzzlVx7ZrU5aVrTmFU37ZM/ugLBv5tFiu+KenXgqvOSkskTc1scfHCoKxN3CJyziWUzLRUbjunI5Ou6MnWXfsZ+LcPmfjhWg59+T1XVZWWSBqUcs6nwDpXzZzWIYs3buhH33aNuf2VZVwxcR6bv9tX9o2uyistkeRIurJ4oaRRwPz4heScS1SN62Qy7rJsfj/wBD78bAsD7v+Amatyww7Lhay0RRubAi8BefwncWQDGcB5ZvZNpURYAXzUlnMVb8U3O7ju6Y9ZtfE7RvZtyy8HHEtmWmrYYbkKFOuorVhmtn+fyAZUAEvN7N0KiK9SeSJxLj727i/grteXM+mjLzi+WT0eGtqFdk3qhh2WqyAVlkiqAk8kzsXXO8s38ovnI+t0/eacjlzcs1VMe8u7xFaRS6Q451ypzji+KW9c348ebRpy60tLuOrJ+Xy7Ky/ssFwl8UTinKsQTerVYNLlPbnt7ON5b+UmBjzwL/69enPYYblKENdEImmApJWSVku6pZTrzpdkkrKD4/6S5ktaHPw8Pera94M6FwavJvH8Ds652KWkiFH9jualn/WhdmYaw8bN4e7pK8jLLww7NBdHcUskklKBh4GzgI7AUEkdS7iuLnA9MCeqeDPwYzPrDFwGPFnstmFm1iV4bcI5l1A6Na/Pq9f2ZUiPVjw68zMuePTfrN28K+ywXJzE84mkJ7DazNaYWR7wDDCohOvuBP4EHFhi1Mw+NrOvgsOlQE1JmXGM1TlXwWplpHHX4M48ekk3vtiym7Mf/IDnctb5jPgqKJ6JpDmwLup4fVB2gKRuQEsze62Ues4HFphZ9BTaCUGz1m90kKEhkkZLypGUk5vrE6acC8uATs1444Z+nNiiPr94fhFjnv6Y7Xv2hx2Wq0ChdbZLSgHuBW4q5ZoTiDytXBVVPCxo8uoXvIaXdK+ZPW5m2WaWnZWVVXGBO+cOWbP6NZkyqje/HHAsM5Z8w48e+IC5a7eGHZarIPFMJBuAllHHLYKyInWJTHR8X9LnQG9gWlSHewsiM+svNbPPim4ysw3Bz53AU0Sa0JxzCS41Rfzse+14/qenkJYqhjz+Efe+tYr8Au+IT3bxTCTzgPaS2krKAIYA04pOmtl2M2tsZm3MrA0wGxhoZjmSGhDZifEWM/uw6B5JaZIaB+/TgXOAJXH8Ds65CtalZQNeu64f53VtwYPvfMqFj33EFl/8ManFLZGYWT4wBpgBLAemmtlSSXdIGljG7WOAdsBviw3zzQRmSFoELCTyhPNEvL6Dcy4+6mSmcc+FJ/Hg0K4s/WoHl46f6/0mScyXSHHOheq9lZsYPTmHE1s04MmRPamVkRZ2SC7gS6Q455LC949twoNDuvLxl99y5eQc3843CXkicc6F7qzOzfjzBSfx4eotjHlqAfu9Az6peCJxziWEC7q34M5BJ/D28k3cOPUTCgqrfrN7VeGNkc65hDH85DZ8t6+AP72xgtoZqdw1uLMvR58EPJE45xLKT793DLv25fO391ZTKyON35xzvCeTBOeJxDmXcG76QQe0vuwZAAAP70lEQVS+25fP+A/XUqdGGjf27xB2SK4UnkiccwlHEr89pyO78/J58J1PqZOZyuhTjwk7LHcQnkiccwkpJUXcNfhEducV8MfXV1ArI41LercOOyxXAk8kzrmElZoi7ruoC3vyCvjNP5dQOzOV87q2CDssV4wP/3XOJbT01BQeHtaN3m0bcfNzi3hjyTdhh+SK8UTinEt4NdJTGXtZNie2qM+1Ty9g5irfYyiReCJxziWF2plpTBzRk/ZN6nLVkznMWbMl7JBcwBOJcy5p1K+VzuSRPWneoCYjJ+XwybptYYfk8ETinEsyjetk8o9RvWhQK53LJsxl5Tc7ww6p2vNE4pxLOs3q1+SpUb3JTEth2Ng5rN28K+yQqjVPJM65pNSqUS2mjOpFoRnDnpjNhm17wg6p2oprIpE0QNJKSasl3VLKdedLsqL92oOyXwf3rZT0w0Ot0zlX9bVrUpfJV/Rk5758hj0xm00794YdUrUUt0QiKRV4GDgL6AgMldSxhOvqAtcDc6LKOhLZ4/0EYADwd0mpsdbpnKs+OjWvz8TLe7Bp5z6Gj53Lt7vywg6p2onnE0lPYLWZrTGzPOAZYFAJ190J/AmI/lNiEPCMme0zs7XA6qC+WOt0zlUj3Vs35IlLs1m7ZRcjJsxl517f/70yxTORNAfWRR2vD8oOkNQNaGlmr8V4b5l1Oueqpz7tGvP3i7ux9KsdjJyUw54837K3soTW2S4pBbgXuClO9Y+WlCMpJzfXZ8E6Vx2c2bEp917UhXmfb+Xqf8xnX74nk8oQz0SyAWgZddwiKCtSF+gEvC/pc6A3MC3ocD/YvWXVeYCZPW5m2WaWnZWVVc6v4pxLFgNPOoq7B3dm5qpcrn96Ifm+/3vcxTORzAPaS2orKYNI5/m0opNmtt3MGptZGzNrA8wGBppZTnDdEEmZktoC7YG5ZdXpnHMAF/VoxW/P6cgbS7/hl88votD3f4+ruC0jb2b5ksYAM4BUYLyZLZV0B5BjZgdNAMF1U4FlQD5wjZkVAJRUZ7y+g3MueV3Rty279uVzz1urqJWZyp2DOlW7LXu3795P/Vrpcf8cmVX9TJ2dnW05OTlhh+Gcq2Rmxt3TV/DYv9Zw1WlHc8uA46pFMvn4y28ZN2stby7byLs3nUaLI2odVj2S5ptZdlnX+cZWzrkqSxK3nHUcu/LyeWzmGupkpHHtGe3DDisu8gsKmbF0I+NmrWHBl9uom5nGZSe3JjMtNe6f7YnEOVelSeKOgZ3Yva8gaOZKY2TftmGHVWG279nPs/O+ZNK/v2DDtj20blSL23/ckQuyW1Ins3J+xXsicc5VeSkp4s8XRPZ/v/PVZdTJTOWiHq3CDqtcPt+8iwkfruW5+evZnVdA76MbcvvAEzj9uCakplRu850nEudctZCWmsIDQ7swevJ8bnlxMTUz0hh40lFhh3VIzIyP1mxh/KzPeWfFRtJSxI9POoor+rSlU/P6ocXlicQ5V21kpqXy6CXduWzCXG58diG10lM5s2PTsMMq0778Al755GvGz1rLsq930LB2Btd+vx2X9G5Nk3o1wg7PR20556qfnXv3c8nYOSz/ZicTRvSgT7vGYYdUos3f7WPK7C95cvYXbP5uHx2a1uGKPm05t2tzaqTHvxM91lFbnkicc9XSt7vyGPL4bL7cupt/jOpJ99YNww7pgJXf7GT8rLW8tHADefmFfO/YLEb2bUvfdo0rdfiyJ5IonkiccyXZtHMvFz02m83f7ePpK3uH2s9QWGjMXJXLuFlrmbV6MzXSUzi/Wwsu79OGdk3qhhKTJ5IonkiccwezYdseLnz0I/bsL2DqVb0r/Zf27rx8XliwgQkfrmVN7i6a1svk0pPbcHHPVhxRO6NSYynOE0kUTyTOudKs3byLnzz6Eakp8NxVp9Cq0eHNBD8U32zfy6SPPuepOV+yfc9+TmxRn5F92/Kjzs1IT02MXdA9kUTxROKcK8vKb3Zy0eMfUbdGGs9ddQpH1o/PaKhP1m1j3Ky1vL74awrN+EHHIxnZry3ZrY9IuOVbPJFE8UTinIvFovXbuPiJOTStl8mzV51M4zqZFVJvfkEhby7byLhZa5n/xbfUyUzjoh4tGXFKG1o2jP/Tz+HyRBLFE4lzLlZz127l0vFzOLpxHZ4e3Zv6NQ9/9dwde/czdd46Jnz4ORu27aFlw5pcfkpbfpLdgro14r8qb3l5IoniicQ5dyhmrspl1KR5dGpen3+M7EXtQ1yz6ostu5jw4ec8l7OOXXkF9GzTkCv6tqV/x6aVvnxJefjqv845d5hO65DFQ0O7cs1TH3Pl5BzGj+hR5gRAM2PO2q2Mn7WWt5ZvJFX/Wb6kc4vwhhVXBk8kzjlXggGdmvGXCwq4ceonXDNlAY8O717iaKq8/EJeXfQV42atZelXO2hQK52ffe8YLj25DU0TYPmSyuCJxDnnDmJwtxbsyivgNy8v4efPLuSBIV0PNE1t3ZXHlNlfMHn2F+Tu3Ee7JnX443mdOa9rc2pmxH/5kkQS10QiaQDwAJFtccea2d3Fzl8NXAMUAN8Bo81smaRhwC+iLj0R6GZmCyW9DzQD9gTnfmBmm+L5PZxz1dfw3q3ZvS+fu6avoFZGKqP6Hc2ED9fy4oIN7MsvpF/7xvzlghM5tX0WKUnU/1GR4tbZLikVWAX0B9YD84ChZrYs6pp6ZrYjeD8Q+JmZDShWT2fgZTM7Jjh+H7jZzGLuPffOdudced375koefHc1AJlpKQzu1pzL+7SlQ9Nwli+pDInQ2d4TWG1ma4KAngEGAQcSSVESCdQGSspqQ4Fn4hinc86V6ef9O1C3Rjp5BYUM7dmKhiEvX5JI4plImgProo7XA72KXyTpGuBGIAM4vYR6LiKSgKJNkFQAvAD8warDGGbnXKgkceWpR4cdRkIKfUEXM3s4aLb6FXBb9DlJvYDdZrYkqniYmXUG+gWv4SXVK2m0pBxJObm5uXGK3jnnXDwTyQagZdRxi6DsYJ4Bzi1WNgR4OrrAzDYEP3cCTxFpQvsfZva4mWWbWXZWVtYhhu6ccy5W8Uwk84D2ktpKyiCSFKZFXyCpfdTh2cCnUedSgAuJ6h+RlCapcfA+HTgHiH5acc45V8ni1kdiZvmSxgAziAz/HW9mSyXdAeSY2TRgjKQzgf3At8BlUVWcCqwr6qwPZAIzgiSSCrwNPBGv7+Ccc65svtaWc865EsU6/Df0znbnnHPJzROJc865cvFE4pxzrlyqRR+JpFzgi8O8vTGwuQLDibdkitdjjZ9kijeZYoXkire8sbY2szLnT1SLRFIeknJi6WxKFMkUr8caP8kUbzLFCskVb2XF6k1bzjnnysUTiXPOuXLxRFK2x8MO4BAlU7wea/wkU7zJFCskV7yVEqv3kTjnnCsXfyJxzjlXLp5IDkJSS0nvSVomaamk68OO6WAk1ZA0V9InQay/DzumskhKlfSxpFfDjqUskj6XtFjSQkkJv9aOpAaSnpe0QtJySSeHHVNJJB0b/JsWvXZIuiHsuA5G0s+D/7+WSHpaUo2wYyqNpOuDWJfG+9/Vm7YOQlIzoJmZLZBUF5gPnBu9VXCikCSgtpl9FyxoOQu43sxmhxzaQUm6EcgG6pnZOWHHUxpJnwPZZpYUcwckTQI+MLOxwcrbtcxsW9hxlSbYmnsD0MvMDnfOV9xIak7k/6uOZrZH0lTgdTObGG5kJZPUicjK6T2BPOAN4GozWx2Pz/MnkoMws6/NbEHwfiewnMiujwnHIr4LDtODV8L+hSCpBZFtA8aGHUtVI6k+kZWzxwGYWV6iJ5HAGcBniZhEoqQBNSWlAbWAr0KOpzTHA3PMbLeZ5QMzgcHx+jBPJDGQ1AboCswJN5KDC5qKFgKbgLfMLGFjBe4HfgkUhh1IjAx4U9J8SaPDDqYMbYFcIttRfyxprKTaYQcVg//ZxC6RBBvq/RX4Evga2G5mb4YbVamWAP0kNZJUC/gR/73RYIXyRFIGSXWI7A1/g5ntCDuegzGzAjPrQmQnyp7Bo23CkXQOsMnM5ocdyyHoa2bdgLOAaySdGnZApUgDugGPmFlXYBdwS7ghlS5ofhsIPBd2LAcj6QhgEJFEfRRQW9Il4UZ1cGa2HPgT8CaRZq2FQEG8Ps8TSSmC/oYXgClm9mLY8cQiaMZ4DxgQdiwH0QcYGPQ7PAOcLukf4YZUuqjtnTcBL3GQ7Z0TxHpgfdQT6fNEEksiOwtYYGYbww6kFGcCa80s18z2Ay8Cp4QcU6nMbJyZdTezU4lsHLgqXp/lieQggg7sccByM7s37HhKIylLUoPgfU2gP7Ai3KhKZma/NrMWZtaGSHPGu2aWsH/ZSaodDLYgaCL6AQm8vbOZfQOsk3RsUHQGkHADRIoZSgI3awW+BHpLqhX8bjiDSL9pwpLUJPjZikj/yFPx+qy4bbVbBfQBhgOLg74HgP9nZq+HGNPBNAMmBSNfUoCpZpbww2qTRFPgpcjvDtKAp8zsjXBDKtO1wJSgyWgNcHnI8RxUkJz7A1eFHUtpzGyOpOeBBUA+8DGJP8P9BUmNiGxlfk08B1348F/nnHPl4k1bzjnnysUTiXPOuXLxROKcc65cPJE455wrF08kzjnnysUTiXNRJJmke6KOb5Z0ewV/xuVRK97mRa0sfPdh1NVS0rMVGZ9zh8qH/zoXRdJeImsp9TCzzZJuBuqY2e1x+rzPSaKVhZ0riT+ROPff8olMNPt58ROSJkq6IOr4u+Dn9yTNlPRPSWsk3S1pWLBHzGJJx8T64ZIaS5omaZGkfxetmSbpD5ImSZot6VNJVwTl7YomzEpKk3RfsAfFIkk/C8r/osi+Oosk/ak8/zjOlcRntjv3vx4GFkn68yHccxKRpbu3EplNPtbMeiqyIdq1QKwbC91JZPnvgZJ+AEwksm8LQGci6zvVAxZIeq3YvT8lsqDgSWZWIKmhpKZEVn49wcysaCkd5yqSP5E4V0ywyvNk4LpDuG1esIfNPuAzIquuAiwG2hxCPX2BJ4M43gSOiloG/mUz2xssHvkvoEexe88EHjWzguD+rUQSWyHwhKTziKwG7FyF8kTiXMnuB0YC0Xt55BP8PyMpBciIOrcv6n1h1HEhFffkX7xDs8wOzmCl2mzgZeBcoPhTjHPl5onEuRIEf81PJZJMinwOdA/eDySyE2VF+wAYBiDpTGCDmRU9RZwrKVNSFtAPKL5//FvA1cHinQRNW3WJbGf8KpF+n65xiNlVc95H4tzB3QOMiTp+AvinpE+IbBYUj2ai3wLjJS0CvuO/V+5dQmTL1EbA78xsY9ES94HHgPZE+nfygUeAV4EXJWUS+cPxxjjE7Ko5H/7rXBKQ9Adgs5ndH3YszhXnTVvOOefKxZ9InHPOlYs/kTjnnCsXTyTOOefKxROJc865cvFE4pxzrlw8kTjnnCsXTyTOOefK5f8DNxDpIwc5A2QAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "n_topics = 10\n",
    "models_it = []\n",
    "for n in range(2, n_topics):\n",
    "    model_it = create_lda_model(d_it, c_it, n, passes = 200)\n",
    "    assess_model_goodness(model_it, texts_it, c_it, d_it, n)\n",
    "    models_it.append(model_it)\n",
    "\n",
    "silhouette_list = [compute_silhouette(model_it, c_it, len(model_it.print_topics())) for model_it in models_it]\n",
    "perplexity_list = [compute_perplexity(model_it, c_it) for model_it in models_it]\n",
    "coh_v = [compute_coherence(model_it, texts_it, c_it, d_it, coherence = \"c_v\") for model_it in models_it]\n",
    "\n",
    "x = range(2,n_topics)\n",
    "plt.plot(x, coh_v)\n",
    "plt.xlabel(\"Num Topics\")\n",
    "plt.ylabel(\"Coherence score\")\n",
    "plt.show()\n",
    "plt.plot(x, silhouette_list)\n",
    "plt.xlabel(\"Num Topics\")\n",
    "plt.ylabel(\"Silhouette index\")\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We chose the model with five topics for various reasons:\n",
    "\n",
    "- Even though it is evident from the graphs that there are two well-divided macro-topics in the corpus, we are looking for more niche, unnoticed topics. For this reason, we decide to use a model which includes at least 5 topics for our analysis.\n",
    "\n",
    "\n",
    "- The model with 5 is the one with the most meaningful division between topical words in our opinion, and since the silhouette and the coherence basically degrade by adding more topic (see also extra at the end with 40 topics), it is the safest pick given the previous constraint.\n",
    "\n",
    "\n",
    "- The coherence score and silhouette index are acceptable, provided the corpus and the constraints on which we are working (heavily unbalanced set, use of dialects and multiple languages, etc.)\n",
    "\n",
    "We set the threshold to an arbitrary high value (in our case, it was to obtain n = 500) to filter out uncertain classifications."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "\n",
       "<link rel=\"stylesheet\" type=\"text/css\" href=\"https://cdn.rawgit.com/bmabey/pyLDAvis/files/ldavis.v1.0.0.css\">\n",
       "\n",
       "\n",
       "<div id=\"ldavis_el47461402366019459285528783847\"></div>\n",
       "<script type=\"text/javascript\">\n",
       "\n",
       "var ldavis_el47461402366019459285528783847_data = {\"mdsDat\": {\"x\": [-0.09600545925739586, 0.25486776473098727, -0.053765986608430794, -0.12210908463074602, 0.017012765765585403], \"y\": [-0.01655012186040343, -0.07370013344284218, -0.015328026118768323, -0.10476126618972796, 0.21033954761174178], \"topics\": [1, 2, 3, 4, 5], \"cluster\": [1, 1, 1, 1, 1], \"Freq\": [38.735355377197266, 20.007619857788086, 16.238815307617188, 14.04901123046875, 10.969197273254395]}, \"tinfo\": {\"Category\": [\"Default\", \"Default\", \"Default\", \"Default\", \"Default\", \"Default\", \"Default\", \"Default\", \"Default\", \"Default\", \"Default\", \"Default\", \"Default\", \"Default\", \"Default\", \"Default\", \"Default\", \"Default\", \"Default\", \"Default\", \"Default\", \"Default\", \"Default\", \"Default\", \"Default\", \"Default\", \"Default\", \"Default\", \"Default\", \"Default\", \"Topic1\", \"Topic1\", \"Topic1\", \"Topic1\", \"Topic1\", \"Topic1\", \"Topic1\", \"Topic1\", \"Topic1\", \"Topic1\", \"Topic1\", \"Topic1\", \"Topic1\", \"Topic1\", \"Topic1\", \"Topic1\", \"Topic1\", \"Topic1\", \"Topic1\", \"Topic1\", \"Topic1\", \"Topic1\", \"Topic1\", \"Topic1\", \"Topic1\", \"Topic1\", \"Topic1\", \"Topic1\", \"Topic1\", \"Topic1\", \"Topic1\", \"Topic1\", \"Topic1\", \"Topic1\", \"Topic1\", \"Topic1\", \"Topic1\", \"Topic1\", \"Topic1\", \"Topic1\", \"Topic1\", \"Topic1\", \"Topic1\", \"Topic1\", \"Topic1\", \"Topic1\", \"Topic1\", \"Topic1\", \"Topic1\", \"Topic1\", \"Topic1\", \"Topic1\", \"Topic1\", \"Topic1\", \"Topic1\", \"Topic1\", \"Topic1\", \"Topic1\", \"Topic1\", \"Topic1\", \"Topic1\", \"Topic1\", \"Topic1\", \"Topic1\", \"Topic1\", \"Topic1\", \"Topic1\", \"Topic1\", \"Topic1\", \"Topic1\", \"Topic1\", \"Topic1\", \"Topic1\", \"Topic1\", \"Topic2\", \"Topic2\", \"Topic2\", \"Topic2\", \"Topic2\", \"Topic2\", \"Topic2\", \"Topic2\", \"Topic2\", \"Topic2\", \"Topic2\", \"Topic2\", \"Topic2\", \"Topic2\", \"Topic2\", \"Topic2\", \"Topic2\", \"Topic2\", \"Topic2\", \"Topic2\", \"Topic2\", \"Topic2\", \"Topic2\", \"Topic2\", \"Topic2\", \"Topic2\", \"Topic2\", \"Topic2\", \"Topic2\", \"Topic2\", \"Topic2\", \"Topic2\", \"Topic2\", \"Topic2\", \"Topic2\", \"Topic2\", \"Topic2\", \"Topic2\", \"Topic2\", \"Topic2\", \"Topic2\", \"Topic2\", \"Topic2\", \"Topic2\", \"Topic2\", \"Topic2\", \"Topic2\", \"Topic2\", \"Topic2\", \"Topic2\", \"Topic3\", \"Topic3\", \"Topic3\", \"Topic3\", \"Topic3\", \"Topic3\", \"Topic3\", \"Topic3\", \"Topic3\", \"Topic3\", \"Topic3\", \"Topic3\", \"Topic3\", \"Topic3\", \"Topic3\", \"Topic3\", \"Topic3\", \"Topic3\", \"Topic3\", \"Topic3\", \"Topic3\", \"Topic3\", \"Topic3\", \"Topic3\", \"Topic3\", \"Topic3\", \"Topic3\", \"Topic3\", \"Topic3\", \"Topic3\", \"Topic3\", \"Topic3\", \"Topic3\", \"Topic3\", \"Topic3\", \"Topic3\", \"Topic3\", \"Topic3\", \"Topic3\", \"Topic3\", \"Topic3\", \"Topic3\", \"Topic3\", \"Topic3\", \"Topic3\", \"Topic3\", \"Topic3\", \"Topic3\", \"Topic3\", \"Topic3\", \"Topic3\", \"Topic3\", \"Topic3\", \"Topic3\", \"Topic3\", \"Topic3\", \"Topic3\", \"Topic3\", \"Topic3\", \"Topic3\", \"Topic3\", \"Topic3\", \"Topic3\", \"Topic4\", \"Topic4\", \"Topic4\", \"Topic4\", \"Topic4\", \"Topic4\", \"Topic4\", \"Topic4\", \"Topic4\", \"Topic4\", \"Topic4\", \"Topic4\", \"Topic4\", \"Topic4\", \"Topic4\", \"Topic4\", \"Topic4\", \"Topic4\", \"Topic4\", \"Topic4\", \"Topic4\", \"Topic4\", \"Topic4\", \"Topic4\", \"Topic4\", \"Topic4\", \"Topic4\", \"Topic4\", \"Topic4\", \"Topic4\", \"Topic4\", \"Topic4\", \"Topic4\", \"Topic4\", \"Topic4\", \"Topic4\", \"Topic4\", \"Topic4\", \"Topic4\", \"Topic4\", \"Topic4\", \"Topic4\", \"Topic4\", \"Topic4\", \"Topic4\", \"Topic4\", \"Topic4\", \"Topic4\", \"Topic4\", \"Topic4\", \"Topic4\", \"Topic4\", \"Topic4\", \"Topic4\", \"Topic4\", \"Topic4\", \"Topic4\", \"Topic4\", \"Topic4\", \"Topic4\", \"Topic4\", \"Topic4\", \"Topic4\", \"Topic5\", \"Topic5\", \"Topic5\", \"Topic5\", \"Topic5\", \"Topic5\", \"Topic5\", \"Topic5\", \"Topic5\", \"Topic5\", \"Topic5\", \"Topic5\", \"Topic5\", \"Topic5\", \"Topic5\", \"Topic5\", \"Topic5\", \"Topic5\", \"Topic5\", \"Topic5\", \"Topic5\", \"Topic5\", \"Topic5\", \"Topic5\", \"Topic5\", \"Topic5\", \"Topic5\", \"Topic5\", \"Topic5\", \"Topic5\", \"Topic5\", \"Topic5\", \"Topic5\", \"Topic5\", \"Topic5\", \"Topic5\", \"Topic5\", \"Topic5\", \"Topic5\", \"Topic5\", \"Topic5\", \"Topic5\", \"Topic5\", \"Topic5\", \"Topic5\", \"Topic5\", \"Topic5\", \"Topic5\", \"Topic5\", \"Topic5\", \"Topic5\", \"Topic5\", \"Topic5\", \"Topic5\", \"Topic5\", \"Topic5\", \"Topic5\", \"Topic5\", \"Topic5\", \"Topic5\", \"Topic5\", \"Topic5\"], \"Freq\": [123.0, 122.0, 184.0, 219.0, 222.0, 108.0, 109.0, 82.0, 177.0, 61.0, 169.0, 215.0, 418.0, 63.0, 137.0, 130.0, 57.0, 65.0, 307.0, 70.0, 171.0, 110.0, 52.0, 41.0, 40.0, 60.0, 40.0, 163.0, 48.0, 223.0, 30.019609451293945, 24.24529266357422, 17.507164001464844, 40.4366569519043, 13.655035018920898, 12.69591999053955, 12.692255973815918, 12.689340591430664, 12.67697525024414, 11.734540939331055, 11.7239408493042, 10.768753051757812, 10.765618324279785, 9.805695533752441, 58.79637908935547, 9.787900924682617, 24.882917404174805, 21.336158752441406, 8.851112365722656, 8.842717170715332, 8.829765319824219, 32.42710876464844, 7.887973308563232, 7.8873066902160645, 7.884014129638672, 7.879846572875977, 7.879405975341797, 7.878042697906494, 7.870179176330566, 6.926471710205078, 55.14948272705078, 30.937376022338867, 26.57402229309082, 18.113018035888672, 38.636470794677734, 15.482878684997559, 14.620536804199219, 140.55018615722656, 60.47926712036133, 23.442607879638672, 177.3338623046875, 60.41656494140625, 20.09077262878418, 103.26405334472656, 274.1497497558594, 207.4686737060547, 64.24187469482422, 35.45583724975586, 29.189128875732422, 63.08535385131836, 110.9646224975586, 61.845550537109375, 114.80152893066406, 118.90117645263672, 60.735042572021484, 99.95188903808594, 43.62605667114258, 126.0286865234375, 53.57359313964844, 40.11359786987305, 117.44741821289062, 113.13805389404297, 91.95977783203125, 43.826114654541016, 78.53825378417969, 131.69168090820312, 82.61909484863281, 84.06489562988281, 76.62247467041016, 81.8118667602539, 75.730712890625, 61.31924819946289, 78.68315124511719, 58.52904510498047, 121.51205444335938, 122.4537582397461, 62.28261184692383, 56.5517463684082, 51.77528762817383, 47.95518493652344, 42.2236328125, 39.35699462890625, 34.58123016357422, 32.66759490966797, 30.759864807128906, 30.756254196166992, 29.795549392700195, 26.93865966796875, 25.980850219726562, 24.0731201171875, 18.34149742126465, 18.341402053833008, 18.34127426147461, 17.38619041442871, 17.3858642578125, 17.385025024414062, 16.430442810058594, 16.430240631103516, 15.475341796875, 15.473953247070312, 14.520259857177734, 32.56822204589844, 13.56456184387207, 13.563309669494629, 43.140235900878906, 75.8342514038086, 61.144325256347656, 55.36983108520508, 27.828413009643555, 97.19634246826172, 27.656055450439453, 96.0017318725586, 24.074594497680664, 39.60361862182617, 34.86812973022461, 39.496517181396484, 31.57719612121582, 42.37511444091797, 44.2059211730957, 37.18443298339844, 74.11583709716797, 53.20668411254883, 41.9796028137207, 33.27004623413086, 33.7990837097168, 32.839054107666016, 30.916837692260742, 28.997541427612305, 22.27802848815918, 12.625555038452148, 10.754761695861816, 8.834168434143066, 8.830584526062012, 8.822840690612793, 7.872188568115234, 6.908631801605225, 6.908587455749512, 5.947662830352783, 5.9459919929504395, 5.942955017089844, 4.9893083572387695, 4.988136291503906, 4.985495090484619, 4.977391242980957, 4.969594955444336, 11.390928268432617, 15.405536651611328, 7.868127822875977, 8.527175903320312, 9.664539337158203, 5.951415538787842, 14.55448055267334, 26.151870727539062, 10.805977821350098, 39.49724578857422, 29.947654724121094, 25.248628616333008, 117.47382354736328, 11.682226181030273, 37.11618423461914, 48.7348518371582, 19.67962074279785, 63.831634521484375, 17.158658981323242, 44.227577209472656, 45.82711410522461, 20.468067169189453, 57.447723388671875, 32.17673110961914, 27.322750091552734, 34.10888671875, 35.98002243041992, 23.185087203979492, 18.00398826599121, 45.481109619140625, 29.885425567626953, 36.53191375732422, 44.21030807495117, 33.40420913696289, 46.57469177246094, 24.646486282348633, 27.859529495239258, 38.59800720214844, 31.958080291748047, 27.842437744140625, 27.971698760986328, 26.13673210144043, 40.914859771728516, 39.96575927734375, 21.970134735107422, 15.320272445678711, 13.444607734680176, 14.32852554321289, 12.50040340423584, 12.499052047729492, 10.600210189819336, 8.707023620605469, 6.816422939300537, 5.865329265594482, 5.861260890960693, 19.028343200683594, 4.923140048980713, 4.922359943389893, 4.92099142074585, 4.919461250305176, 4.917466163635254, 4.902091979980469, 9.475106239318848, 24.030527114868164, 8.426950454711914, 6.693550109863281, 12.536238670349121, 5.857862949371338, 6.541055679321289, 6.475930213928223, 15.777544021606445, 5.729186534881592, 20.57532501220703, 19.11482048034668, 7.861207962036133, 28.3383846282959, 99.0191879272461, 11.98633098602295, 22.935354232788086, 9.006131172180176, 40.454933166503906, 13.10092544555664, 17.69972038269043, 49.180843353271484, 67.71942138671875, 55.59592056274414, 64.95433807373047, 51.078121185302734, 80.37635803222656, 25.734786987304688, 33.49188232421875, 58.37299728393555, 27.24085807800293, 54.59775161743164, 33.227569580078125, 34.01509094238281, 33.50593948364258, 34.574710845947266, 33.095176696777344, 26.534765243530273, 33.058441162109375, 28.687196731567383, 25.474241256713867, 26.056917190551758, 21.95442771911621, 25.725616455078125, 15.323196411132812, 22.661771774291992, 8.69607162475586, 8.688126564025879, 7.749632358551025, 5.859468936920166, 5.854790687561035, 4.916005611419678, 4.913034439086914, 4.910374641418457, 21.216089248657227, 18.624576568603516, 5.657010078430176, 32.97574996948242, 28.742765426635742, 8.289700508117676, 19.52939796447754, 48.44328689575195, 8.157058715820312, 13.19438648223877, 5.840169429779053, 11.329926490783691, 4.870851516723633, 4.8165106773376465, 4.800195217132568, 8.667474746704102, 5.259029865264893, 3.912428855895996, 7.175586700439453, 42.833892822265625, 15.345311164855957, 20.612117767333984, 14.405050277709961, 16.13498306274414, 16.655717849731445, 11.766271591186523, 39.70797348022461, 34.13462829589844, 26.703144073486328, 19.607675552368164, 28.649333953857422, 19.231538772583008, 15.447959899902344, 18.587244033813477, 16.353124618530273, 21.61905860900879, 18.46029281616211, 17.853059768676758, 24.57787322998047, 18.725786209106445, 28.19987678527832, 24.005043029785156, 23.099485397338867, 20.431516647338867, 20.123287200927734, 19.552759170532227, 20.084684371948242, 18.018545150756836, 18.73141098022461, 18.5347900390625, 18.470687866210938], \"Term\": [\"amicare\", \"senilit\\u00e0\", \"fabbricare\", \"baciare\", \"marcare\", \"librare\", \"parigi\", \"romanzare\", \"domenica\", \"dottore\", \"londra\", \"notte\", \"cuore\", \"joyce\", \"sabato\", \"titina\", \"zeno\", \"articolare\", \"sera\", \"dolore\", \"letizia\", \"curare\", \"traduzione\", \"tchap\", \"macchina\", \"villa\", \"mamma\", \"viaggiare\", \"larbaud\", \"livia\", \"colera\", \"lloyd\", \"luzzatto\", \"frizzo\", \"zeline\", \"temperatura\", \"gorizia\", \"bortolo\", \"divertimento\", \"tosse\", \"sior\", \"pittura\", \"chiedere\", \"nube\", \"marted\\u00ec\", \"paulina\", \"velare\", \"dente\", \"padova\", \"baule\", \"timore\", \"natalia\", \"andrea\", \"corona\", \"graziadio\", \"dille\", \"toilette\", \"nonna\", \"schedare\", \"neve\", \"ottavio\", \"pepe\", \"aurelio\", \"fausto\", \"banca\", \"gatto\", \"barba\", \"letizia\", \"gioved\\u00ec\", \"vederti\", \"venezia\", \"salutami\", \"novit\\u00e0\", \"sabato\", \"cuore\", \"sera\", \"pranzare\", \"paola\", \"mezzod\\u00ec\", \"bepi\", \"olga\", \"bagnare\", \"domenica\", \"mattina\", \"arrivare\", \"settimana\", \"treno\", \"marcare\", \"trovare\", \"murare\", \"livia\", \"notte\", \"notizia\", \"mercoled\\u00ec\", \"ritornare\", \"trieste\", \"bisognare\", \"abbracciare\", \"leggere\", \"piacere\", \"viaggiare\", \"giornata\", \"baciare\", \"fabbricare\", \"senilit\\u00e0\", \"amicare\", \"joyce\", \"zeno\", \"traduzione\", \"larbaud\", \"editore\", \"cr\\u00e9mieux\", \"copia\", \"benco\", \"prezzolini\", \"pubblicazione\", \"critico\", \"convegno\", \"criticare\", \"piaggiare\", \"edizione\", \"traduttore\", \"fiero\", \"commerce\", \"gallimard\", \"rue\", \"henri\", \"prefazione\", \"preg\", \"jahier\", \"saba\", \"novello\", \"taglio\", \"burlare\", \"libro\", \"romanzare\", \"articolare\", \"villa\", \"aff\", \"librare\", \"svevo\", \"parigi\", \"bobi\", \"copiare\", \"succedere\", \"saluto\", \"marzo\", \"maestro\", \"coscienza\", \"italia\", \"trieste\", \"scrivere\", \"piacere\", \"londra\", \"tolone\", \"marsiglia\", \"chatham\", \"porro\", \"plymouth\", \"scena\", \"devonport\", \"room\", \"lambert\", \"sandrini\", \"rosely\", \"insolenza\", \"riso\", \"piccolezza\", \"dopopranzo\", \"austria\", \"occhiato\", \"ufficiale\", \"fermo\", \"documento\", \"pasto\", \"tavolo\", \"frantzen\", \"miss\", \"giovanotto\", \"chiesa\", \"merito\", \"descrizione\", \"potere\", \"rimprovero\", \"salsomaggiore\", \"capra\", \"inghilterra\", \"baciare\", \"via\", \"donna\", \"scriverti\", \"revoltella\", \"viaggiare\", \"hotel\", \"affare\", \"addio\", \"francia\", \"londra\", \"salso\", \"puntare\", \"compagnia\", \"maritare\", \"bocca\", \"levico\", \"abbracciare\", \"occhio\", \"ritornare\", \"livia\", \"titina\", \"trieste\", \"proposito\", \"curare\", \"cuore\", \"sera\", \"olga\", \"marcare\", \"fabbricare\", \"tchap\", \"macchina\", \"nicoletto\", \"dormire\", \"male\", \"bordare\", \"meggie\", \"grigio\", \"bottone\", \"avventizii\", \"tonn\", \"magazzino\", \"salariare\", \"merce\", \"operaio\", \"pasqua\", \"mastelloni\", \"domattina\", \"fatturare\", \"caro\", \"toccare\", \"caldaia\", \"maglia\", \"barile\", \"bottiglia\", \"disperazione\", \"motore\", \"polire\", \"colazione\", \"trovarla\", \"tchaperoff\", \"corpo\", \"programmare\", \"pitturare\", \"fabbricare\", \"gondola\", \"lavoro\", \"pelle\", \"operare\", \"latta\", \"charlton\", \"titina\", \"marcare\", \"londra\", \"notte\", \"domenica\", \"cuore\", \"fumare\", \"testare\", \"sera\", \"pomeriggio\", \"trieste\", \"sabato\", \"leggere\", \"scrivere\", \"mattina\", \"abbracciare\", \"affare\", \"livia\", \"olga\", \"piacere\", \"venezia\", \"restare\", \"riccione\", \"coiazzi\", \"montecatini\", \"spaziare\", \"fisonomia\", \"sostanza\", \"cervello\", \"confusione\", \"cugina\", \"inquietudine\", \"metodo\", \"colorire\", \"dr\", \"fanciullo\", \"mamma\", \"febbre\", \"dilazione\", \"convinzione\", \"dottore\", \"giuoco\", \"scommettere\", \"bellezza\", \"commedia\", \"animale\", \"carezza\", \"carissimo\", \"vestito\", \"bicicletta\", \"fantasia\", \"andare\", \"dolore\", \"penna\", \"frase\", \"sorella\", \"verit\\u00e0\", \"istante\", \"sogno\", \"curare\", \"fondere\", \"malattia\", \"questione\", \"madre\", \"acqua\", \"medicare\", \"speranza\", \"sentimento\", \"dubbio\", \"voce\", \"bagno\", \"restare\", \"padre\", \"livia\", \"bisognare\", \"leggere\", \"desiderio\", \"pensiero\", \"giornata\", \"notte\", \"occhio\", \"letizia\", \"mattina\", \"cuore\"], \"Total\": [123.0, 122.0, 184.0, 219.0, 222.0, 108.0, 109.0, 82.0, 177.0, 61.0, 169.0, 215.0, 418.0, 63.0, 137.0, 130.0, 57.0, 65.0, 307.0, 70.0, 171.0, 110.0, 52.0, 41.0, 40.0, 60.0, 40.0, 163.0, 48.0, 223.0, 30.784400939941406, 25.010805130004883, 18.274934768676758, 42.32929992675781, 14.425861358642578, 13.46360969543457, 13.463592529296875, 13.463479995727539, 13.46336555480957, 12.50132942199707, 12.501303672790527, 11.539013862609863, 11.53902816772461, 10.576764106750488, 63.43671798706055, 10.576679229736328, 26.91514015197754, 23.079010009765625, 9.614587783813477, 9.614541053771973, 9.614511489868164, 35.55802917480469, 8.652316093444824, 8.652300834655762, 8.652314186096191, 8.652207374572754, 8.652244567871094, 8.652251243591309, 8.652013778686523, 7.690054893493652, 61.54965591430664, 34.60542297363281, 29.817033767700195, 20.17735481262207, 44.23603439331055, 17.294841766357422, 16.343425750732422, 171.73171997070312, 72.93309783935547, 26.92973518371582, 232.36192321777344, 75.77820587158203, 23.081645965576172, 137.06919860839844, 418.7816162109375, 307.770751953125, 84.48711395263672, 44.18157196044922, 35.50113296508789, 86.23603057861328, 167.87721252441406, 85.29093933105469, 177.17770385742188, 185.78268432617188, 86.32128143310547, 160.12632751464844, 58.51007843017578, 222.1038055419922, 75.72216033935547, 52.74663162231445, 223.10733032226562, 215.095703125, 168.5932159423828, 59.41147232055664, 139.1373748779297, 311.1578369140625, 157.99002075195312, 165.810302734375, 150.07748413085938, 168.37892150878906, 163.1538543701172, 111.99482727050781, 219.7298583984375, 184.08010864257812, 122.27642822265625, 123.23172760009766, 63.04826354980469, 57.31650161743164, 52.5400505065918, 48.7188720703125, 42.98711395263672, 40.12123107910156, 35.344764709472656, 33.43415451049805, 31.523595809936523, 31.523569107055664, 30.56831932067871, 27.702421188354492, 26.747133255004883, 24.836549758911133, 19.10479164123535, 19.104793548583984, 19.104793548583984, 18.149497985839844, 18.14950180053711, 18.149492263793945, 17.194204330444336, 17.194196701049805, 16.238914489746094, 16.238901138305664, 15.28361701965332, 34.37941360473633, 14.328322410583496, 14.328336715698242, 45.834537506103516, 82.10079956054688, 65.93505859375, 60.14799880981445, 29.618349075317383, 108.87622833251953, 29.601581573486328, 109.85266876220703, 25.782745361328125, 44.895633697509766, 39.13205337524414, 46.76876449584961, 36.26640701293945, 53.50103759765625, 61.965694427490234, 61.199256896972656, 311.1578369140625, 157.77163696289062, 168.37892150878906, 169.07052612304688, 34.56441116333008, 33.6041374206543, 31.683549880981445, 29.762969970703125, 23.040929794311523, 13.437371253967285, 11.517434120178223, 9.596853256225586, 9.596811294555664, 9.596717834472656, 8.636542320251465, 7.676241397857666, 7.676240921020508, 6.715938091278076, 6.7159624099731445, 6.715916156768799, 5.755640506744385, 5.7556657791137695, 5.755603790283203, 5.755523681640625, 5.755422592163086, 13.420490264892578, 18.21121597290039, 9.598688125610352, 10.5398588180542, 12.448665618896484, 7.6712327003479, 19.142518997192383, 34.46653747558594, 14.356656074523926, 52.818389892578125, 40.271541595458984, 34.47545623779297, 219.7298583984375, 16.298681259155273, 69.79710388183594, 98.71407318115234, 31.673534393310547, 163.1538543701172, 27.812339782714844, 108.92566680908203, 115.96138763427734, 36.40560531616211, 169.07052612304688, 75.94510650634766, 59.37250900268555, 90.23319244384766, 101.42853546142578, 49.87835693359375, 32.53238296508789, 165.810302734375, 85.23665618896484, 139.1373748779297, 223.10733032226562, 130.97300720214844, 311.1578369140625, 69.95793914794922, 110.74502563476562, 418.7816162109375, 307.770751953125, 167.87721252441406, 222.1038055419922, 184.08010864257812, 41.682403564453125, 40.7352409362793, 22.73908233642578, 16.109228134155273, 14.214613914489746, 15.162797927856445, 13.267410278320312, 13.267436981201172, 11.373117446899414, 9.478780746459961, 7.584424018859863, 6.637253284454346, 6.637299060821533, 21.819534301757812, 5.690074443817139, 5.690087795257568, 5.690103530883789, 5.690101146697998, 5.690176010131836, 5.690270900726318, 11.38277530670166, 29.442106246948242, 10.445518493652344, 8.546390533447266, 16.153636932373047, 7.5926432609558105, 8.551090240478516, 8.552056312561035, 20.913238525390625, 7.599072456359863, 27.556995391845703, 25.66071128845215, 10.450804710388184, 43.76877975463867, 184.08010864257812, 17.124624252319336, 38.969139099121094, 12.353767395019531, 93.5257568359375, 20.9414119720459, 31.406883239746094, 130.97300720214844, 222.1038055419922, 169.07052612304688, 215.095703125, 177.17770385742188, 418.7816162109375, 65.7608413696289, 108.87567901611328, 307.770751953125, 76.38790130615234, 311.1578369140625, 137.06919860839844, 150.07748413085938, 157.77163696289062, 185.78268432617188, 165.810302734375, 108.92566680908203, 223.10733032226562, 167.87721252441406, 168.37892150878906, 232.36192321777344, 113.6677474975586, 26.49957275390625, 16.09217071533203, 24.60858154296875, 9.46939468383789, 9.469539642333984, 8.523260116577148, 6.631042957305908, 6.6310882568359375, 5.684871673583984, 5.684878826141357, 5.6849751472473145, 24.64657974243164, 21.791912078857422, 6.634468078613281, 40.80982971191406, 36.07073211669922, 10.438630104064941, 24.672191619873047, 61.616912841796875, 10.420225143432617, 17.041759490966797, 7.593602180480957, 15.17584228515625, 6.64749813079834, 6.646686553955078, 6.632147312164307, 12.356593132019043, 7.603433609008789, 5.694620609283447, 10.455156326293945, 70.3690414428711, 24.64948272705078, 34.25965881347656, 24.76823616027832, 29.511127471923828, 31.453584671020508, 20.001760482788086, 110.74502563476562, 89.5545425415039, 63.925865173339844, 41.904781341552734, 80.27513122558594, 43.866092681884766, 31.417726516723633, 42.86776351928711, 36.210777282714844, 63.94917297363281, 54.41411209106445, 52.51628112792969, 113.6677474975586, 61.21001434326172, 223.10733032226562, 157.99002075195312, 150.07748413085938, 93.74950408935547, 100.41091918945312, 111.99482727050781, 215.095703125, 85.23665618896484, 171.73171997070312, 185.78268432617188, 418.7816162109375], \"loglift\": [30.0, 29.0, 28.0, 27.0, 26.0, 25.0, 24.0, 23.0, 22.0, 21.0, 20.0, 19.0, 18.0, 17.0, 16.0, 15.0, 14.0, 13.0, 12.0, 11.0, 10.0, 9.0, 8.0, 7.0, 6.0, 5.0, 4.0, 3.0, 2.0, 1.0, 0.92330002784729, 0.9172999858856201, 0.9054999947547913, 0.9027000069618225, 0.8934999704360962, 0.8896999955177307, 0.8894000053405762, 0.88919997215271, 0.8881999850273132, 0.8851000070571899, 0.8841999769210815, 0.8792999982833862, 0.8790000081062317, 0.8726999759674072, 0.8725000023841858, 0.8708999752998352, 0.8698999881744385, 0.8698999881744385, 0.8657000064849854, 0.8647000193595886, 0.8633000254631042, 0.8561999797821045, 0.85589998960495, 0.8557999730110168, 0.855400025844574, 0.8549000024795532, 0.8549000024795532, 0.8547000288963318, 0.8536999821662903, 0.8438000082969666, 0.8385999798774719, 0.8363999724388123, 0.833299994468689, 0.840499997138977, 0.8130999803543091, 0.8377000093460083, 0.8370000123977661, 0.7480000257492065, 0.7612000107765198, 0.8097000122070312, 0.6782000064849854, 0.7218999862670898, 0.8095999956130981, 0.6651999950408936, 0.5246999859809875, 0.5540000200271606, 0.6744999885559082, 0.7283999919891357, 0.7526000142097473, 0.6358000040054321, 0.5343999862670898, 0.6269999742507935, 0.5145000219345093, 0.5020999908447266, 0.5968999862670898, 0.4771000146865845, 0.6549000144004822, 0.38179999589920044, 0.602400004863739, 0.6746000051498413, 0.3068000078201294, 0.3059000074863434, 0.3422999978065491, 0.6442000269889832, 0.3765000104904175, 0.08860000222921371, 0.3000999987125397, 0.26919999718666077, 0.27619999647140503, 0.22660000622272491, 0.1808999925851822, 0.34610000252723694, -0.07859999686479568, -0.19740000367164612, 1.6028000116348267, 1.6026999950408936, 1.5967999696731567, 1.5956000089645386, 1.5944000482559204, 1.5932999849319458, 1.5910999774932861, 1.5898000001907349, 1.5872000455856323, 1.5858999490737915, 1.5844999551773071, 1.5844000577926636, 1.5835000276565552, 1.5810999870300293, 1.5800000429153442, 1.5778000354766846, 1.5683000087738037, 1.5683000087738037, 1.5683000087738037, 1.566100001335144, 1.566100001335144, 1.565999984741211, 1.563599944114685, 1.563599944114685, 1.5608999729156494, 1.5607999563217163, 1.557800054550171, 1.554900050163269, 1.5542999505996704, 1.5542000532150269, 1.5484999418258667, 1.5297000408172607, 1.5335999727249146, 1.5262999534606934, 1.5467000007629395, 1.4955999851226807, 1.541100025177002, 1.4743000268936157, 1.5405000448226929, 1.4836000204086304, 1.4937000274658203, 1.4400999546051025, 1.4706000089645386, 1.3759000301361084, 1.271299958229065, 1.11080002784729, 0.17440000176429749, 0.5220999717712402, 0.2199999988079071, -0.016599999740719795, 1.7954000234603882, 1.794700026512146, 1.7933000326156616, 1.791700005531311, 1.78410005569458, 1.7553999423980713, 1.7493000030517578, 1.7350000143051147, 1.7345999479293823, 1.7337000370025635, 1.725100040435791, 1.712399959564209, 1.712399959564209, 1.6963000297546387, 1.6959999799728394, 1.6955000162124634, 1.6749000549316406, 1.6746000051498413, 1.6741000413894653, 1.6725000143051147, 1.6710000038146973, 1.6538000106811523, 1.6505000591278076, 1.61899995803833, 1.6059000492095947, 1.5645999908447266, 1.5638999938964844, 1.5437999963760376, 1.541700005531311, 1.5336999893188477, 1.5270999670028687, 1.5216000080108643, 1.5062999725341797, 1.19159996509552, 1.4847999811172485, 1.1862000226974487, 1.111899971961975, 1.3418999910354614, 0.8792999982833862, 1.3348000049591064, 0.9164000153541565, 0.8894000053405762, 1.2418999671936035, 0.7383000254631042, 0.9589999914169312, 1.041700005531311, 0.8449000120162964, 0.7814000248908997, 1.0516999959945679, 1.226099967956543, 0.5242000222206116, 0.7696999907493591, 0.4805000126361847, 0.19910000264644623, 0.4514999985694885, -0.08150000125169754, 0.7745000123977661, 0.4377000033855438, -0.5663999915122986, -0.4472000002861023, 0.0210999995470047, -0.2542000114917755, -0.13429999351501465, 1.944000005722046, 1.94350004196167, 1.9282000064849854, 1.9124000072479248, 1.9069000482559204, 1.906000018119812, 1.9031000137329102, 1.902999997138977, 1.892199993133545, 1.8776999711990356, 1.8559000492095947, 1.8389999866485596, 1.8382999897003174, 1.825700044631958, 1.8178000450134277, 1.8177000284194946, 1.8173999786376953, 1.8171000480651855, 1.8166999816894531, 1.8135000467300415, 1.7791999578475952, 1.7595000267028809, 1.7479000091552734, 1.7182999849319458, 1.7091000080108643, 1.7031999826431274, 1.694700002670288, 1.684499979019165, 1.680799961090088, 1.6801999807357788, 1.6705000400543213, 1.6680999994277954, 1.677899956703186, 1.527899980545044, 1.3425999879837036, 1.6059000492095947, 1.4325000047683716, 1.6466000080108643, 1.1246000528335571, 1.4936000108718872, 1.3890999555587769, 0.9830999970436096, 0.7748000025749207, 0.8503999710083008, 0.7652000188827515, 0.7188000082969666, 0.31200000643730164, 1.024399995803833, 0.7836999893188477, 0.3000999987125397, 0.9315000176429749, 0.2222999930381775, 0.5454999804496765, 0.478300005197525, 0.4131999909877777, 0.28119999170303345, 0.35120001435279846, 0.5504000186920166, 0.053199999034404755, 0.19580000638961792, 0.07410000264644623, -0.22540000081062317, 0.3183000087738037, 2.1803998947143555, 2.161099910736084, 2.127700090408325, 2.1249001026153564, 2.124000072479248, 2.1149001121520996, 2.086400032043457, 2.085599899291992, 2.0648000240325928, 2.064199924468994, 2.0636000633239746, 2.0601999759674072, 2.052999973297119, 2.0506999492645264, 1.996899962425232, 1.9830000400543213, 1.979599952697754, 1.9763000011444092, 1.969499945640564, 1.9651999473571777, 1.954200029373169, 1.9474999904632568, 1.917799949645996, 1.8990999460220337, 1.8880000114440918, 1.8868000507354736, 1.8554999828338623, 1.8414000272750854, 1.8346999883651733, 1.8336999416351318, 1.7137000560760498, 1.7360999584197998, 1.7020000219345093, 1.6680999994277954, 1.6062999963760376, 1.5743000507354736, 1.6794999837875366, 1.184399962425232, 1.2454999685287476, 1.3371000289916992, 1.450600028038025, 1.1797000169754028, 1.3854999542236328, 1.5002000331878662, 1.374400019645691, 1.4150999784469604, 1.125599980354309, 1.12909996509552, 1.131100058555603, 0.678600013256073, 1.0256999731063843, 0.14169999957084656, 0.32580000162124634, 0.33869999647140503, 0.6865000128746033, 0.6026999950408936, 0.46470001339912415, -0.16099999845027924, 0.656000018119812, -0.00570000009611249, -0.09480000287294388, -0.9110999703407288], \"logprob\": [30.0, 29.0, 28.0, 27.0, 26.0, 25.0, 24.0, 23.0, 22.0, 21.0, 20.0, 19.0, 18.0, 17.0, 16.0, 15.0, 14.0, 13.0, 12.0, 11.0, 10.0, 9.0, 8.0, 7.0, 6.0, 5.0, 4.0, 3.0, 2.0, 1.0, -5.953100204467773, -6.166800022125244, -6.492400169372559, -5.655200004577637, -6.740900039672852, -6.813700199127197, -6.814000129699707, -6.814199924468994, -6.815199851989746, -6.892399787902832, -6.893400192260742, -6.978300094604492, -6.978600025177002, -7.072000026702881, -5.280900001525879, -7.073800086975098, -6.1407999992370605, -6.294600009918213, -7.1743998527526855, -7.1753997802734375, -7.1768999099731445, -5.875999927520752, -7.289599895477295, -7.289700031280518, -7.29010009765625, -7.2906999588012695, -7.2906999588012695, -7.290900230407715, -7.291900157928467, -7.419600009918213, -5.344900131225586, -5.922999858856201, -6.075099945068359, -6.458399772644043, -5.700799942016602, -6.615300178527832, -6.672599792480469, -4.40939998626709, -5.252699851989746, -6.200399875640869, -4.177000045776367, -5.253699779510498, -6.354700088500977, -4.717700004577637, -3.741300106048584, -4.019999980926514, -5.192299842834473, -5.7866997718811035, -5.981200218200684, -5.2104997634887695, -4.6458001136779785, -5.230299949645996, -4.611800193786621, -4.576700210571289, -5.248499870300293, -4.75029993057251, -5.5792999267578125, -4.518499851226807, -5.373899936676025, -5.663300037384033, -4.589000225067139, -4.626399993896484, -4.833600044250488, -5.57480001449585, -4.991399765014648, -4.4745001792907715, -4.940700054168701, -4.923399925231934, -5.01609992980957, -4.9506001472473145, -5.0278000831604, -5.238900184631348, -4.98960018157959, -5.2855000495910645, -3.8942999839782715, -3.8866000175476074, -4.562699794769287, -4.659200191497803, -4.747399806976318, -4.824100017547607, -4.951399803161621, -5.021699905395508, -5.151000022888184, -5.208000183105469, -5.268099784851074, -5.2683000564575195, -5.300000190734863, -5.4008002281188965, -5.436999797821045, -5.513299942016602, -5.785200119018555, -5.785200119018555, -5.785200119018555, -5.838699817657471, -5.838699817657471, -5.838699817657471, -5.895199775695801, -5.895199775695801, -5.955100059509277, -5.9552001953125, -6.018799781799316, -5.210999965667725, -6.086900234222412, -6.086999893188477, -4.929900169372559, -4.365799903869629, -4.581099987030029, -4.680300235748291, -5.368299961090088, -4.117599964141846, -5.374499797821045, -4.130000114440918, -5.513199806213379, -5.015399932861328, -5.1427998542785645, -5.018099784851074, -5.2418999671936035, -4.947800159454346, -4.9054999351501465, -5.078499794006348, -4.388700008392334, -4.720200061798096, -4.957200050354004, -5.189700126647949, -4.965199947357178, -4.99399995803833, -5.054299831390381, -5.1184000968933105, -5.381999969482422, -5.949900150299072, -6.110300064086914, -6.307000160217285, -6.307400226593018, -6.308300018310547, -6.422299861907959, -6.5528998374938965, -6.5528998374938965, -6.702600002288818, -6.702899932861328, -6.703400135040283, -6.878300189971924, -6.878600120544434, -6.8790998458862305, -6.88070011138916, -6.882299900054932, -6.052800178527832, -5.750899791717529, -6.422800064086914, -6.342400074005127, -6.217199802398682, -6.702000141143799, -5.807700157165527, -5.221700191497803, -6.105500221252441, -4.8094000816345215, -5.08620023727417, -5.256899833679199, -3.719399929046631, -6.027599811553955, -4.871600151062012, -4.5991997718811035, -5.506100177764893, -4.329400062561035, -5.643099784851074, -4.696300029754639, -4.660799980163574, -5.466800212860107, -4.434800148010254, -5.014400005340576, -5.1778998374938965, -4.956099987030029, -4.902699947357178, -5.342100143432617, -5.59499979019165, -4.668300151824951, -5.0883002281188965, -4.887499809265137, -4.696700096130371, -4.9770002365112305, -4.644599914550781, -5.281000137329102, -5.1585001945495605, -4.832399845123291, -5.021200180053711, -5.15910005569458, -5.154399871826172, -5.222300052642822, -4.629300117492676, -4.6528000831604, -5.251100063323975, -5.611599922180176, -5.742199897766113, -5.678500175476074, -5.815000057220459, -5.815100193023682, -5.979899883270264, -6.176700115203857, -6.42140007019043, -6.571700096130371, -6.572400093078613, -5.394899845123291, -6.746799945831299, -6.747000217437744, -6.747300148010254, -6.747600078582764, -6.748000144958496, -6.751100063323975, -6.092100143432617, -5.161499977111816, -6.209400177001953, -6.439599990844727, -5.81220006942749, -6.572999954223633, -6.462699890136719, -6.472700119018555, -5.582200050354004, -6.595200061798096, -5.316699981689453, -5.3902997970581055, -6.278800010681152, -4.996600151062012, -3.745500087738037, -5.85699987411499, -5.208099842071533, -6.142899990081787, -4.640600204467773, -5.768099784851074, -5.467199802398682, -4.445300102233887, -4.125400066375732, -4.322700023651123, -4.167099952697754, -4.407400131225586, -3.9540998935699463, -5.092899799346924, -4.829500198364258, -4.273900032043457, -5.036099910736084, -4.340799808502197, -4.837399959564209, -4.814000129699707, -4.829100131988525, -4.797699928283691, -4.841400146484375, -5.062300205230713, -4.84250020980835, -4.984300136566162, -5.103099822998047, -5.08050012588501, -5.251800060272217, -4.845799922943115, -5.363999843597412, -4.972599983215332, -5.930500030517578, -5.931399822235107, -6.0457000732421875, -6.325300216674805, -6.326099872589111, -6.500800132751465, -6.501399993896484, -6.501999855041504, -5.038599967956543, -5.168799877166748, -6.360400199890137, -4.597599983215332, -4.734899997711182, -5.978300094604492, -5.121399879455566, -4.212900161743164, -5.9944000244140625, -5.513500213623047, -6.32859992980957, -5.665900230407715, -6.5100998878479, -6.521299839019775, -6.524700164794922, -5.933700084686279, -6.4334001541137695, -6.7291998863220215, -6.122600078582764, -4.335999965667725, -5.362500190734863, -5.067399978637695, -5.4257001876831055, -5.312300205230713, -5.280600070953369, -5.6280999183654785, -4.411799907684326, -4.563000202178955, -4.808499813079834, -5.117400169372559, -4.7382001876831055, -5.1367998123168945, -5.355800151824951, -5.17080020904541, -5.298900127410889, -5.019700050354004, -5.177700042724609, -5.211100101470947, -4.891499996185303, -5.163400173187256, -4.754000186920166, -4.91510009765625, -4.953499794006348, -5.076200008392334, -5.091400146484375, -5.120200157165527, -5.093400001525879, -5.201900005340576, -5.163099765777588, -5.173699855804443, -5.17710018157959]}, \"token.table\": {\"Topic\": [1, 2, 3, 4, 5, 1, 2, 3, 4, 5, 1, 3, 4, 5, 2, 3, 1, 2, 3, 4, 5, 2, 1, 3, 5, 1, 1, 5, 1, 3, 4, 2, 3, 1, 3, 3, 4, 1, 3, 4, 5, 1, 3, 4, 5, 1, 2, 3, 4, 5, 1, 2, 3, 1, 2, 3, 4, 1, 1, 5, 2, 1, 4, 5, 1, 5, 1, 2, 3, 4, 5, 2, 5, 1, 2, 3, 5, 4, 1, 1, 4, 4, 2, 1, 4, 1, 3, 4, 5, 3, 5, 4, 5, 4, 5, 1, 2, 4, 3, 1, 3, 4, 5, 5, 1, 4, 1, 3, 5, 2, 5, 2, 1, 2, 3, 4, 5, 2, 3, 5, 2, 1, 2, 4, 1, 1, 3, 4, 1, 2, 3, 5, 2, 2, 2, 5, 1, 2, 3, 4, 5, 1, 2, 3, 5, 1, 2, 3, 5, 1, 2, 3, 4, 5, 3, 1, 5, 1, 2, 4, 1, 3, 1, 2, 3, 4, 5, 4, 1, 3, 4, 1, 2, 3, 4, 5, 3, 4, 1, 4, 5, 2, 5, 1, 2, 3, 4, 5, 2, 2, 1, 3, 4, 5, 2, 5, 4, 1, 5, 1, 5, 3, 2, 5, 1, 2, 3, 4, 5, 2, 3, 3, 4, 1, 2, 5, 1, 3, 1, 3, 4, 5, 2, 1, 5, 1, 2, 3, 4, 5, 3, 4, 1, 4, 5, 4, 5, 1, 4, 1, 1, 4, 2, 1, 2, 3, 2, 3, 4, 5, 3, 1, 3, 5, 1, 2, 4, 2, 2, 3, 2, 3, 4, 2, 3, 4, 5, 1, 2, 3, 4, 5, 1, 4, 5, 1, 3, 4, 2, 3, 4, 2, 5, 1, 3, 4, 5, 1, 1, 2, 3, 4, 1, 4, 1, 2, 3, 5, 1, 2, 3, 4, 5, 4, 1, 4, 1, 2, 3, 4, 5, 4, 1, 5, 1, 3, 4, 1, 3, 4, 5, 3, 1, 4, 5, 2, 4, 4, 1, 3, 4, 5, 1, 2, 3, 5, 4, 3, 4, 1, 4, 5, 2, 3, 5, 1, 5, 1, 3, 4, 5, 1, 4, 1, 3, 4, 1, 4, 1, 4, 1, 1, 2, 3, 4, 5, 1, 3, 4, 5, 2, 5, 1, 3, 1, 3, 1, 2, 3, 5, 1, 3, 4, 4, 1, 2, 3, 4, 1, 3, 4, 1, 1, 2, 3, 5, 1, 3, 4, 5, 2, 3, 4, 5, 4, 3, 1, 1, 2, 4, 2, 4, 5, 1, 2, 3, 4, 5, 1, 3, 4, 1, 2, 3, 4, 5, 2, 3, 1, 1, 2, 3, 4, 3, 1, 4, 1, 3, 4, 5, 3, 1, 3, 4, 1, 2, 3, 4, 5, 2, 2, 2, 3, 4, 1, 2, 3, 4, 5, 2, 1, 2, 3, 4, 5, 1, 2, 4, 5, 1, 2, 3, 4, 5, 1, 3, 5, 5, 3, 5, 3, 1, 2, 3, 4, 5, 2, 5, 3, 3, 2, 2, 1, 4, 4, 1, 3, 1, 3, 4, 1, 4, 2, 3, 4, 3, 3, 1, 4, 5, 1, 2, 3, 4, 5, 1, 3, 4, 5, 2, 1, 2, 4, 5, 1, 2, 3, 4, 5, 1, 2, 3, 4, 5, 1, 1, 5, 1, 5, 5, 5, 1, 2, 3, 4, 5, 2, 5, 2, 5, 2, 3, 4, 4, 1, 2, 3, 4, 1, 1, 2, 4, 5, 1, 1, 3, 4, 2, 4, 1, 3, 4, 1, 2, 2, 1, 3, 4, 1, 2, 3, 4, 5, 1, 2, 3, 4, 5, 3, 4, 3, 1, 3, 1, 4, 1, 2, 3, 4, 1, 2, 3, 5, 1, 5, 2, 3, 1, 2, 3, 4, 2, 4, 1, 2, 3, 4, 5, 1, 2], \"Freq\": [0.5066030025482178, 0.006030988413840532, 0.2713944613933563, 0.19902262091636658, 0.012061976827681065, 0.36474642157554626, 0.06838995218276978, 0.06838995218276978, 0.06838995218276978, 0.43313637375831604, 0.4053073227405548, 0.3966837525367737, 0.12072984129190445, 0.07761204242706299, 0.9453599452972412, 0.0337628535926342, 0.17443087697029114, 0.05508343502879143, 0.4039452075958252, 0.24787546694278717, 0.11016687005758286, 0.9900047779083252, 0.1912931650876999, 0.09564658254384995, 0.6695261001586914, 0.9246079325675964, 0.1504325419664383, 0.7521626949310303, 0.7066623568534851, 0.10426165908575058, 0.1853540539741516, 0.9251527190208435, 0.06066575273871422, 0.9055226445198059, 0.10061363130807877, 0.8934000730514526, 0.949489176273346, 0.3595323860645294, 0.5324720144271851, 0.05461251363158226, 0.05461251363158226, 0.7269236445426941, 0.02344914898276329, 0.19931776821613312, 0.058622874319553375, 0.285625696182251, 0.057125139981508255, 0.26658400893211365, 0.057125139981508255, 0.3427508473396301, 0.8816341757774353, 0.04521200805902481, 0.06781801581382751, 0.91780024766922, 0.06118668243288994, 0.11700846254825592, 0.8190592527389526, 0.936082124710083, 0.13168980181217194, 0.790138840675354, 0.987014651298523, 0.7305530905723572, 0.23192162811756134, 0.03478824347257614, 0.2630390524864197, 0.6575976610183716, 0.5253496170043945, 0.12659026682376862, 0.12659026682376862, 0.07595416158437729, 0.15190832316875458, 0.9308550953865051, 0.038785628974437714, 0.36087796092033386, 0.0801951065659523, 0.4611218571662903, 0.10024388134479523, 0.9233124256134033, 0.9655750393867493, 0.1857166886329651, 0.8047723174095154, 0.9671930074691772, 0.9770848155021667, 0.16982480883598328, 0.8151590824127197, 0.12415715306997299, 0.7449429035186768, 0.024831430986523628, 0.07449429482221603, 0.1504509001970291, 0.7522545456886292, 0.15078072249889374, 0.7539036273956299, 0.8786928057670593, 0.904835045337677, 0.1592007726430893, 0.28656139969825745, 0.5731227993965149, 0.9784257411956787, 0.9532865285873413, 0.8032989501953125, 0.08032989501953125, 0.08032989501953125, 0.9321302771568298, 0.23908300697803497, 0.765065610408783, 0.9745195508003235, 0.12172074615955353, 0.8520451784133911, 0.19768260419368744, 0.7248362302780151, 0.9366649985313416, 0.47654303908348083, 0.07757677137851715, 0.376801460981369, 0.06649437546730042, 0.9048288464546204, 0.9746440649032593, 0.1621258556842804, 0.8106292486190796, 0.9902456402778625, 0.0668216422200203, 0.8909552097320557, 0.04454776272177696, 0.9246095418930054, 0.03897007927298546, 0.1948504000902176, 0.740431547164917, 0.01613796129822731, 0.7100703120231628, 0.01613796129822731, 0.24206942319869995, 0.9720668196678162, 0.9814082384109497, 0.9720539450645447, 0.8795273303985596, 0.6542789340019226, 0.01671515591442585, 0.0931272953748703, 0.191030353307724, 0.0429818294942379, 0.2167140245437622, 0.17156526446342468, 0.2528330385684967, 0.3611900508403778, 0.9099177122116089, 0.04332941398024559, 0.7835959196090698, 0.20895890891551971, 0.3946687579154968, 0.17066757380962372, 0.18133428692817688, 0.03200016915798187, 0.21333445608615875, 0.9550738334655762, 0.19159601628780365, 0.7663840651512146, 0.9246195554733276, 0.13170643150806427, 0.790238618850708, 0.9655832052230835, 0.8687306642532349, 0.11368636041879654, 0.07105397433042526, 0.15631873905658722, 0.04263238236308098, 0.6110641956329346, 0.8787189722061157, 0.6490658521652222, 0.062084559351205826, 0.2878465950489044, 0.11461793631315231, 0.05730896815657616, 0.5301079750061035, 0.12894517183303833, 0.17192690074443817, 0.8933939337730408, 0.9311432838439941, 0.09737586230039597, 0.09737586230039597, 0.7790068984031677, 0.1376657485961914, 0.8718830943107605, 0.26583611965179443, 0.14073677361011505, 0.20328643918037415, 0.06254967302083969, 0.3440232276916504, 0.9770369529724121, 0.9421719908714294, 0.3205126225948334, 0.1412428617477417, 0.5378093123435974, 0.9043679237365723, 0.17560432851314545, 0.7024173140525818, 0.8787074685096741, 0.8920891880989075, 0.049560509622097015, 0.19406315684318542, 0.8039759397506714, 0.8687185645103455, 0.9421719312667847, 0.9504157900810242, 0.13399654626846313, 0.2009948343038559, 0.17866207659244537, 0.11166379600763321, 0.37965691089630127, 0.41202446818351746, 0.5493659377098083, 0.8236682415008545, 0.10982242971658707, 0.2918884754180908, 0.1167553961277008, 0.6129658222198486, 0.9449719190597534, 0.023624297231435776, 0.3801654577255249, 0.03041323646903038, 0.39537206292152405, 0.19768603146076202, 0.9366648197174072, 0.867310643196106, 0.05782070755958557, 0.5446680188179016, 0.03571593388915062, 0.13393475115299225, 0.10714780539274216, 0.1785796731710434, 0.8539013862609863, 0.0948779284954071, 0.8226717710494995, 0.10968957096338272, 0.05484478548169136, 0.09596721827983856, 0.7677377462387085, 0.291977196931839, 0.7007452845573425, 0.9655669331550598, 0.9246081113815308, 0.9044700860977173, 0.9305461049079895, 0.10786578804254532, 0.25168684124946594, 0.611239492893219, 0.08701843023300171, 0.7251535654067993, 0.14503070712089539, 0.8795261979103088, 0.9119046330451965, 0.25434303283691406, 0.19075727462768555, 0.5404789447784424, 0.2777811586856842, 0.6045824885368347, 0.11438047140836716, 0.9237078428268433, 0.9833736419677734, 0.9378114938735962, 0.9852445125579834, 0.33426591753959656, 0.6207795143127441, 0.1539679914712906, 0.1539679914712906, 0.5902106165885925, 0.10264532268047333, 0.5130683183670044, 0.033316124230623245, 0.0732954740524292, 0.2265496402978897, 0.1532541662454605, 0.8210480809211731, 0.06987643241882324, 0.11063768714666367, 0.15369302034378052, 0.553294837474823, 0.2766474187374115, 0.8909199237823486, 0.04592370614409447, 0.0551084466278553, 0.9381571412086487, 0.04363521561026573, 0.5244112610816956, 0.1972144991159439, 0.14791087806224823, 0.1255001276731491, 0.9595852494239807, 0.1360379010438919, 0.19518481194972992, 0.3371374011039734, 0.3312227129936218, 0.9849556088447571, 0.9819507598876953, 0.41108623147010803, 0.06228579208254814, 0.1744002103805542, 0.3612575829029083, 0.09345613420009613, 0.7850314974784851, 0.018691226840019226, 0.05607368052005768, 0.03738245368003845, 0.9039884209632874, 0.09573483467102051, 0.7658786773681641, 0.35979175567626953, 0.06257247924804688, 0.10950183868408203, 0.046929359436035156, 0.4223642349243164, 0.9145517349243164, 0.17152729630470276, 0.8086287379264832, 0.5673022866249084, 0.1260671764612198, 0.30616313219070435, 0.3450705409049988, 0.35492971539497375, 0.14788737893104553, 0.14788737893104553, 0.9820219278335571, 0.9300607442855835, 0.015763740986585617, 0.04729122295975685, 0.8823592662811279, 0.11029490828514099, 0.8787186145782471, 0.6405333280563354, 0.0753568634390831, 0.18839214742183685, 0.10227002948522568, 0.1273166537284851, 0.2864624857902527, 0.09548749774694443, 0.47743746638298035, 0.9798445701599121, 0.0916609838604927, 0.8707793354988098, 0.7405977249145508, 0.1851494312286377, 0.06732706725597382, 0.13035714626312256, 0.7821428775787354, 0.879511296749115, 0.8168753385543823, 0.16900868713855743, 0.10418090224266052, 0.8334472179412842, 0.04063623026013374, 0.9346333146095276, 0.11694414913654327, 0.8186090588569641, 0.7583422660827637, 0.0379171147942543, 0.20854412019252777, 0.8999373912811279, 0.08436913043260574, 0.9102665781974792, 0.9674972891807556, 0.9246148467063904, 0.5456922054290771, 0.15421736240386963, 0.10676585882902145, 0.100834421813488, 0.08897155523300171, 0.5253475308418274, 0.07903458923101425, 0.3021910786628723, 0.09298186749219894, 0.9598767757415771, 0.02908717468380928, 0.8664893507957458, 0.08664893358945847, 0.9454687833786011, 0.8687130212783813, 0.3871573805809021, 0.05866020917892456, 0.35196125507354736, 0.21117675304412842, 0.6611975431442261, 0.16678856313228607, 0.17274530231952667, 0.87872314453125, 0.3528439700603485, 0.18176810443401337, 0.032076723873615265, 0.42768967151641846, 0.8935874700546265, 0.08123522251844406, 0.016247045248746872, 0.9360775947570801, 0.4574414789676666, 0.065348781645298, 0.1633719652891159, 0.31040671467781067, 0.7921854853630066, 0.0905354842543602, 0.02263387106359005, 0.067901611328125, 0.8738977313041687, 0.07282481342554092, 0.03641240671277046, 0.009103101678192616, 0.8787210583686829, 0.8687459230422974, 0.9454763531684875, 0.16189393401145935, 0.08094696700572968, 0.7285227179527283, 0.16227521002292633, 0.20284400880336761, 0.6085320115089417, 0.46807658672332764, 0.07967261224985123, 0.12946799397468567, 0.11950891464948654, 0.19918152689933777, 0.8958133459091187, 0.05779441073536873, 0.05779441073536873, 0.4869968295097351, 0.2494373917579651, 0.023755943402647972, 0.14847464859485626, 0.08908478170633316, 0.9663177728652954, 0.8933971524238586, 0.9532877206802368, 0.18277868628501892, 0.11423667520284653, 0.0685420036315918, 0.639725387096405, 0.9548225998878479, 0.11693093925714493, 0.7015856504440308, 0.4320055842399597, 0.10472862422466278, 0.35345911979675293, 0.10472862422466278, 0.9743651151657104, 0.029013648629188538, 0.7543548345565796, 0.20309554040431976, 0.7575119733810425, 0.011836124584078789, 0.10652512311935425, 0.047344498336315155, 0.07101674377918243, 0.9305465221405029, 0.9237070679664612, 0.9833903312683105, 0.1913728266954422, 0.7654913067817688, 0.34306326508522034, 0.18582594394683838, 0.35735759139060974, 0.04288290813565254, 0.07147151976823807, 0.9833911657333374, 0.21895655989646912, 0.15158531069755554, 0.4547559320926666, 0.11789968609809875, 0.05052843689918518, 0.28636351227760315, 0.16704536974430084, 0.07159087806940079, 0.47727251052856445, 0.45747366547584534, 0.0703805610537529, 0.06158299371600151, 0.19354654848575592, 0.21993926167488098, 0.2841489017009735, 0.6314420104026794, 0.06314419955015182, 0.9811478853225708, 0.7661951184272766, 0.2089623063802719, 0.9119046926498413, 0.5677841901779175, 0.05749713256955147, 0.26592424511909485, 0.07905855774879456, 0.03593570739030838, 0.9256913661956787, 0.07308089733123779, 0.9378073811531067, 0.9262966513633728, 0.9366652965545654, 0.9814430475234985, 0.7514452338218689, 0.24075430631637573, 0.9039821624755859, 0.5661984086036682, 0.4213569760322571, 0.22719359397888184, 0.738379180431366, 0.018932798877358437, 0.791784405708313, 0.19794610142707825, 0.833889901638031, 0.021381791681051254, 0.12829075753688812, 0.9378206133842468, 0.9674511551856995, 0.9246402382850647, 0.17603816092014313, 0.7628320455551147, 0.34226682782173157, 0.3359285593032837, 0.08873584866523743, 0.2155013382434845, 0.01901482418179512, 0.31403830647468567, 0.4963831305503845, 0.1722145527601242, 0.020260535180568695, 0.997739315032959, 0.3313930630683899, 0.138080433011055, 0.055232174694538116, 0.44185739755630493, 0.6725785136222839, 0.016245858743786812, 0.10397349298000336, 0.18845196068286896, 0.016245858743786812, 0.624506950378418, 0.1124112457036972, 0.10616617649793625, 0.10616617649793625, 0.049960553646087646, 0.95989990234375, 0.3999647796154022, 0.5999472141265869, 0.4037429094314575, 0.5652400851249695, 0.9386079907417297, 0.9504303336143494, 0.13996531069278717, 0.18662041425704956, 0.18662041425704956, 0.06998265534639359, 0.4432235062122345, 0.8944074511528015, 0.10221799463033676, 0.9458954334259033, 0.033781979233026505, 0.9770857691764832, 0.819642186164856, 0.07451292872428894, 0.983628511428833, 0.10886527597904205, 0.0725768506526947, 0.03628842532634735, 0.7620569467544556, 0.9655657410621643, 0.4684241712093353, 0.11940223723649979, 0.30309799313545227, 0.10103266686201096, 0.9360849857330322, 0.3664877414703369, 0.25196030735969543, 0.37412288784980774, 0.08785203844308853, 0.7906683087348938, 0.9246155619621277, 0.9836707711219788, 0.922944188117981, 0.9598979353904724, 0.9421719312667847, 0.9897211790084839, 0.7520071864128113, 0.05127321928739548, 0.18800179660320282, 0.4242219924926758, 0.23782142996788025, 0.15104874968528748, 0.17675916850566864, 0.012855212204158306, 0.7131333947181702, 0.03961852192878723, 0.03961852192878723, 0.07923704385757446, 0.13206173479557037, 0.13159500062465668, 0.7895700335502625, 0.8687092065811157, 0.8540744781494141, 0.1114010214805603, 0.928845226764679, 0.037153810262680054, 0.7617427110671997, 0.021518155932426453, 0.0989835187792778, 0.1118944063782692, 0.16942761838436127, 0.135542094707489, 0.135542094707489, 0.542168378829956, 0.24278536438941956, 0.728356122970581, 0.24541862308979034, 0.7362558841705322, 0.46581798791885376, 0.05516265705227852, 0.39226776361465454, 0.09193775802850723, 0.9144111275672913, 0.0665026307106018, 0.31241896748542786, 0.091887928545475, 0.18377585709095, 0.07351034134626389, 0.3307965397834778, 0.9704793095588684, 0.9944779872894287], \"Term\": [\"abbracciare\", \"abbracciare\", \"abbracciare\", \"abbracciare\", \"abbracciare\", \"acqua\", \"acqua\", \"acqua\", \"acqua\", \"acqua\", \"addio\", \"addio\", \"addio\", \"addio\", \"aff\", \"aff\", \"affare\", \"affare\", \"affare\", \"affare\", \"affare\", \"amicare\", \"andare\", \"andare\", \"andare\", \"andrea\", \"animale\", \"animale\", \"arrivare\", \"arrivare\", \"arrivare\", \"articolare\", \"articolare\", \"aurelio\", \"aurelio\", \"austria\", \"avventizii\", \"baciare\", \"baciare\", \"baciare\", \"baciare\", \"bagnare\", \"bagnare\", \"bagnare\", \"bagnare\", \"bagno\", \"bagno\", \"bagno\", \"bagno\", \"bagno\", \"banca\", \"banca\", \"banca\", \"barba\", \"barba\", \"barile\", \"barile\", \"baule\", \"bellezza\", \"bellezza\", \"benco\", \"bepi\", \"bepi\", \"bepi\", \"bicicletta\", \"bicicletta\", \"bisognare\", \"bisognare\", \"bisognare\", \"bisognare\", \"bisognare\", \"bobi\", \"bobi\", \"bocca\", \"bocca\", \"bocca\", \"bocca\", \"bordare\", \"bortolo\", \"bottiglia\", \"bottiglia\", \"bottone\", \"burlare\", \"caldaia\", \"caldaia\", \"capra\", \"capra\", \"capra\", \"capra\", \"carezza\", \"carezza\", \"carissimo\", \"carissimo\", \"caro\", \"cervello\", \"charlton\", \"charlton\", \"charlton\", \"chatham\", \"chiedere\", \"chiesa\", \"chiesa\", \"chiesa\", \"coiazzi\", \"colazione\", \"colazione\", \"colera\", \"colorire\", \"colorire\", \"commedia\", \"commedia\", \"commerce\", \"compagnia\", \"compagnia\", \"compagnia\", \"compagnia\", \"confusione\", \"convegno\", \"convinzione\", \"convinzione\", \"copia\", \"copiare\", \"copiare\", \"copiare\", \"corona\", \"corpo\", \"corpo\", \"corpo\", \"coscienza\", \"coscienza\", \"coscienza\", \"coscienza\", \"criticare\", \"critico\", \"cr\\u00e9mieux\", \"cugina\", \"cuore\", \"cuore\", \"cuore\", \"cuore\", \"cuore\", \"curare\", \"curare\", \"curare\", \"curare\", \"dente\", \"dente\", \"descrizione\", \"descrizione\", \"desiderio\", \"desiderio\", \"desiderio\", \"desiderio\", \"desiderio\", \"devonport\", \"dilazione\", \"dilazione\", \"dille\", \"disperazione\", \"disperazione\", \"divertimento\", \"documento\", \"dolore\", \"dolore\", \"dolore\", \"dolore\", \"dolore\", \"domattina\", \"domenica\", \"domenica\", \"domenica\", \"donna\", \"donna\", \"donna\", \"donna\", \"donna\", \"dopopranzo\", \"dormire\", \"dottore\", \"dottore\", \"dottore\", \"dr\", \"dr\", \"dubbio\", \"dubbio\", \"dubbio\", \"dubbio\", \"dubbio\", \"editore\", \"edizione\", \"fabbricare\", \"fabbricare\", \"fabbricare\", \"fanciullo\", \"fantasia\", \"fantasia\", \"fatturare\", \"fausto\", \"fausto\", \"febbre\", \"febbre\", \"fermo\", \"fiero\", \"fisonomia\", \"fondere\", \"fondere\", \"fondere\", \"fondere\", \"fondere\", \"francia\", \"francia\", \"frantzen\", \"frantzen\", \"frase\", \"frase\", \"frase\", \"frizzo\", \"frizzo\", \"fumare\", \"fumare\", \"fumare\", \"fumare\", \"gallimard\", \"gatto\", \"gatto\", \"giornata\", \"giornata\", \"giornata\", \"giornata\", \"giornata\", \"giovanotto\", \"giovanotto\", \"gioved\\u00ec\", \"gioved\\u00ec\", \"gioved\\u00ec\", \"giuoco\", \"giuoco\", \"gondola\", \"gondola\", \"gorizia\", \"graziadio\", \"grigio\", \"henri\", \"hotel\", \"hotel\", \"hotel\", \"inghilterra\", \"inghilterra\", \"inghilterra\", \"inquietudine\", \"insolenza\", \"istante\", \"istante\", \"istante\", \"italia\", \"italia\", \"italia\", \"jahier\", \"joyce\", \"lambert\", \"larbaud\", \"latta\", \"latta\", \"lavoro\", \"lavoro\", \"lavoro\", \"lavoro\", \"leggere\", \"leggere\", \"leggere\", \"leggere\", \"leggere\", \"letizia\", \"letizia\", \"letizia\", \"levico\", \"levico\", \"levico\", \"librare\", \"librare\", \"librare\", \"libro\", \"libro\", \"livia\", \"livia\", \"livia\", \"livia\", \"lloyd\", \"londra\", \"londra\", \"londra\", \"londra\", \"luzzatto\", \"macchina\", \"madre\", \"madre\", \"madre\", \"madre\", \"maestro\", \"maestro\", \"maestro\", \"maestro\", \"maestro\", \"magazzino\", \"maglia\", \"maglia\", \"malattia\", \"malattia\", \"malattia\", \"malattia\", \"malattia\", \"male\", \"mamma\", \"mamma\", \"marcare\", \"marcare\", \"marcare\", \"maritare\", \"maritare\", \"maritare\", \"maritare\", \"marsiglia\", \"marted\\u00ec\", \"marted\\u00ec\", \"marted\\u00ec\", \"marzo\", \"marzo\", \"mastelloni\", \"mattina\", \"mattina\", \"mattina\", \"mattina\", \"medicare\", \"medicare\", \"medicare\", \"medicare\", \"meggie\", \"merce\", \"merce\", \"mercoled\\u00ec\", \"mercoled\\u00ec\", \"mercoled\\u00ec\", \"merito\", \"merito\", \"metodo\", \"mezzod\\u00ec\", \"mezzod\\u00ec\", \"miss\", \"miss\", \"montecatini\", \"montecatini\", \"motore\", \"motore\", \"murare\", \"murare\", \"murare\", \"natalia\", \"natalia\", \"neve\", \"nicoletto\", \"nonna\", \"notizia\", \"notizia\", \"notizia\", \"notizia\", \"notizia\", \"notte\", \"notte\", \"notte\", \"notte\", \"novello\", \"novello\", \"novit\\u00e0\", \"novit\\u00e0\", \"nube\", \"occhiato\", \"occhio\", \"occhio\", \"occhio\", \"occhio\", \"olga\", \"olga\", \"olga\", \"operaio\", \"operare\", \"operare\", \"operare\", \"operare\", \"ottavio\", \"ottavio\", \"ottavio\", \"padova\", \"padre\", \"padre\", \"padre\", \"padre\", \"paola\", \"paola\", \"paola\", \"paola\", \"parigi\", \"parigi\", \"parigi\", \"parigi\", \"pasqua\", \"pasto\", \"paulina\", \"pelle\", \"pelle\", \"pelle\", \"penna\", \"penna\", \"penna\", \"pensiero\", \"pensiero\", \"pensiero\", \"pensiero\", \"pensiero\", \"pepe\", \"pepe\", \"pepe\", \"piacere\", \"piacere\", \"piacere\", \"piacere\", \"piacere\", \"piaggiare\", \"piccolezza\", \"pittura\", \"pitturare\", \"pitturare\", \"pitturare\", \"pitturare\", \"plymouth\", \"polire\", \"polire\", \"pomeriggio\", \"pomeriggio\", \"pomeriggio\", \"pomeriggio\", \"porro\", \"potere\", \"potere\", \"potere\", \"pranzare\", \"pranzare\", \"pranzare\", \"pranzare\", \"pranzare\", \"prefazione\", \"preg\", \"prezzolini\", \"programmare\", \"programmare\", \"proposito\", \"proposito\", \"proposito\", \"proposito\", \"proposito\", \"pubblicazione\", \"puntare\", \"puntare\", \"puntare\", \"puntare\", \"puntare\", \"questione\", \"questione\", \"questione\", \"questione\", \"restare\", \"restare\", \"restare\", \"restare\", \"restare\", \"revoltella\", \"revoltella\", \"revoltella\", \"riccione\", \"rimprovero\", \"rimprovero\", \"riso\", \"ritornare\", \"ritornare\", \"ritornare\", \"ritornare\", \"ritornare\", \"romanzare\", \"romanzare\", \"room\", \"rosely\", \"rue\", \"saba\", \"sabato\", \"sabato\", \"salariare\", \"salso\", \"salso\", \"salsomaggiore\", \"salsomaggiore\", \"salsomaggiore\", \"salutami\", \"salutami\", \"saluto\", \"saluto\", \"saluto\", \"sandrini\", \"scena\", \"schedare\", \"scommettere\", \"scommettere\", \"scrivere\", \"scrivere\", \"scrivere\", \"scrivere\", \"scrivere\", \"scriverti\", \"scriverti\", \"scriverti\", \"scriverti\", \"senilit\\u00e0\", \"sentimento\", \"sentimento\", \"sentimento\", \"sentimento\", \"sera\", \"sera\", \"sera\", \"sera\", \"sera\", \"settimana\", \"settimana\", \"settimana\", \"settimana\", \"settimana\", \"sior\", \"sogno\", \"sogno\", \"sorella\", \"sorella\", \"sostanza\", \"spaziare\", \"speranza\", \"speranza\", \"speranza\", \"speranza\", \"speranza\", \"succedere\", \"succedere\", \"svevo\", \"svevo\", \"taglio\", \"tavolo\", \"tavolo\", \"tchap\", \"tchaperoff\", \"tchaperoff\", \"tchaperoff\", \"tchaperoff\", \"temperatura\", \"testare\", \"testare\", \"testare\", \"testare\", \"timore\", \"titina\", \"titina\", \"titina\", \"toccare\", \"toccare\", \"toilette\", \"tolone\", \"tonn\", \"tosse\", \"traduttore\", \"traduzione\", \"treno\", \"treno\", \"treno\", \"trieste\", \"trieste\", \"trieste\", \"trieste\", \"trieste\", \"trovare\", \"trovare\", \"trovare\", \"trovare\", \"trovare\", \"trovarla\", \"trovarla\", \"ufficiale\", \"vederti\", \"vederti\", \"velare\", \"velare\", \"venezia\", \"venezia\", \"venezia\", \"venezia\", \"verit\\u00e0\", \"verit\\u00e0\", \"verit\\u00e0\", \"verit\\u00e0\", \"vestito\", \"vestito\", \"via\", \"via\", \"viaggiare\", \"viaggiare\", \"viaggiare\", \"viaggiare\", \"villa\", \"villa\", \"voce\", \"voce\", \"voce\", \"voce\", \"voce\", \"zeline\", \"zeno\"]}, \"R\": 30, \"lambda.step\": 0.01, \"plot.opts\": {\"xlab\": \"PC1\", \"ylab\": \"PC2\"}, \"topic.order\": [1, 2, 4, 3, 5]};\n",
       "\n",
       "function LDAvis_load_lib(url, callback){\n",
       "  var s = document.createElement('script');\n",
       "  s.src = url;\n",
       "  s.async = true;\n",
       "  s.onreadystatechange = s.onload = callback;\n",
       "  s.onerror = function(){console.warn(\"failed to load library \" + url);};\n",
       "  document.getElementsByTagName(\"head\")[0].appendChild(s);\n",
       "}\n",
       "\n",
       "if(typeof(LDAvis) !== \"undefined\"){\n",
       "   // already loaded: just create the visualization\n",
       "   !function(LDAvis){\n",
       "       new LDAvis(\"#\" + \"ldavis_el47461402366019459285528783847\", ldavis_el47461402366019459285528783847_data);\n",
       "   }(LDAvis);\n",
       "}else if(typeof define === \"function\" && define.amd){\n",
       "   // require.js is available: use it to load d3/LDAvis\n",
       "   require.config({paths: {d3: \"https://cdnjs.cloudflare.com/ajax/libs/d3/3.5.5/d3.min\"}});\n",
       "   require([\"d3\"], function(d3){\n",
       "      window.d3 = d3;\n",
       "      LDAvis_load_lib(\"https://cdn.rawgit.com/bmabey/pyLDAvis/files/ldavis.v1.0.0.js\", function(){\n",
       "        new LDAvis(\"#\" + \"ldavis_el47461402366019459285528783847\", ldavis_el47461402366019459285528783847_data);\n",
       "      });\n",
       "    });\n",
       "}else{\n",
       "    // require.js not available: dynamically load d3 & LDAvis\n",
       "    LDAvis_load_lib(\"https://cdnjs.cloudflare.com/ajax/libs/d3/3.5.5/d3.min.js\", function(){\n",
       "         LDAvis_load_lib(\"https://cdn.rawgit.com/bmabey/pyLDAvis/files/ldavis.v1.0.0.js\", function(){\n",
       "                 new LDAvis(\"#\" + \"ldavis_el47461402366019459285528783847\", ldavis_el47461402366019459285528783847_data);\n",
       "            })\n",
       "         });\n",
       "}\n",
       "</script>"
      ],
      "text/plain": [
       "<IPython.core.display.HTML object>"
      ]
     },
     "execution_count": 33,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "num_topics = 5\n",
    "threshold = 0.733\n",
    "\n",
    "vis = pyLDAvis.gensim.prepare(topic_model=models_it[num_topics - 2], corpus=c_it, dictionary=d_it)\n",
    "pyLDAvis.display(vis)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We then created a new dataframe by adding to the original one the topic to which each letter was classified under the column `lbl` and the classification score for that topic under the column `max`."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {},
   "outputs": [],
   "source": [
    "X = create_paper_points_data(models_it[num_topics - 2], c_it, num_topics)\n",
    "df = data_it\n",
    "df[\"lbl\"] = pd.Categorical(X.values.argmax(axis=1))\n",
    "df[\"max\"] = X.values.max(axis = 1)\n",
    "df = df[df[\"max\"] > threshold]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We can visualize that by splitting letters by topics to which they were assigned inside different dataframes and analyze those."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "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>n</th>\n",
       "      <th>corpus</th>\n",
       "      <th>index</th>\n",
       "      <th>date</th>\n",
       "      <th>year</th>\n",
       "      <th>sender</th>\n",
       "      <th>senderLocation</th>\n",
       "      <th>recipient</th>\n",
       "      <th>recipientLocation</th>\n",
       "      <th>languages</th>\n",
       "      <th>mainLanguage</th>\n",
       "      <th>text</th>\n",
       "      <th>lbl</th>\n",
       "      <th>max</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>count</th>\n",
       "      <td>225.000000</td>\n",
       "      <td>225</td>\n",
       "      <td>225.000000</td>\n",
       "      <td>225</td>\n",
       "      <td>225.000000</td>\n",
       "      <td>225</td>\n",
       "      <td>225</td>\n",
       "      <td>225</td>\n",
       "      <td>225</td>\n",
       "      <td>225</td>\n",
       "      <td>225</td>\n",
       "      <td>225</td>\n",
       "      <td>225.0</td>\n",
       "      <td>225.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>unique</th>\n",
       "      <td>NaN</td>\n",
       "      <td>4</td>\n",
       "      <td>NaN</td>\n",
       "      <td>222</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1</td>\n",
       "      <td>9</td>\n",
       "      <td>6</td>\n",
       "      <td>14</td>\n",
       "      <td>14</td>\n",
       "      <td>1</td>\n",
       "      <td>225</td>\n",
       "      <td>1.0</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>top</th>\n",
       "      <td>NaN</td>\n",
       "      <td>Schmitz Livia</td>\n",
       "      <td>NaN</td>\n",
       "      <td>05/06/1913</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Ettore Schmitz</td>\n",
       "      <td>Murano</td>\n",
       "      <td>Livia Veneziani</td>\n",
       "      <td>Trieste</td>\n",
       "      <td>ITA</td>\n",
       "      <td>ITA</td>\n",
       "      <td>Alla moglie Murano,  18.3.1913 Carissima Livia...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>freq</th>\n",
       "      <td>NaN</td>\n",
       "      <td>221</td>\n",
       "      <td>NaN</td>\n",
       "      <td>2</td>\n",
       "      <td>NaN</td>\n",
       "      <td>225</td>\n",
       "      <td>124</td>\n",
       "      <td>219</td>\n",
       "      <td>143</td>\n",
       "      <td>144</td>\n",
       "      <td>225</td>\n",
       "      <td>1</td>\n",
       "      <td>225.0</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>mean</th>\n",
       "      <td>535.702222</td>\n",
       "      <td>NaN</td>\n",
       "      <td>21.275556</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1906.911111</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.907587</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>std</th>\n",
       "      <td>177.685480</td>\n",
       "      <td>NaN</td>\n",
       "      <td>17.570637</td>\n",
       "      <td>NaN</td>\n",
       "      <td>5.669555</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.078950</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>min</th>\n",
       "      <td>164.000000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1895.000000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.734079</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25%</th>\n",
       "      <td>413.000000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>8.000000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1903.000000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.850349</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>50%</th>\n",
       "      <td>564.000000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>16.000000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1908.000000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.941927</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>75%</th>\n",
       "      <td>676.000000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>29.000000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1911.000000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.969514</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>max</th>\n",
       "      <td>823.000000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>69.000000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1922.000000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.994531</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                 n         corpus       index        date         year  \\\n",
       "count   225.000000            225  225.000000         225   225.000000   \n",
       "unique         NaN              4         NaN         222          NaN   \n",
       "top            NaN  Schmitz Livia         NaN  05/06/1913          NaN   \n",
       "freq           NaN            221         NaN           2          NaN   \n",
       "mean    535.702222            NaN   21.275556         NaN  1906.911111   \n",
       "std     177.685480            NaN   17.570637         NaN     5.669555   \n",
       "min     164.000000            NaN    1.000000         NaN  1895.000000   \n",
       "25%     413.000000            NaN    8.000000         NaN  1903.000000   \n",
       "50%     564.000000            NaN   16.000000         NaN  1908.000000   \n",
       "75%     676.000000            NaN   29.000000         NaN  1911.000000   \n",
       "max     823.000000            NaN   69.000000         NaN  1922.000000   \n",
       "\n",
       "                sender senderLocation        recipient recipientLocation  \\\n",
       "count              225            225              225               225   \n",
       "unique               1              9                6                14   \n",
       "top     Ettore Schmitz         Murano  Livia Veneziani           Trieste   \n",
       "freq               225            124              219               143   \n",
       "mean               NaN            NaN              NaN               NaN   \n",
       "std                NaN            NaN              NaN               NaN   \n",
       "min                NaN            NaN              NaN               NaN   \n",
       "25%                NaN            NaN              NaN               NaN   \n",
       "50%                NaN            NaN              NaN               NaN   \n",
       "75%                NaN            NaN              NaN               NaN   \n",
       "max                NaN            NaN              NaN               NaN   \n",
       "\n",
       "       languages mainLanguage  \\\n",
       "count        225          225   \n",
       "unique        14            1   \n",
       "top          ITA          ITA   \n",
       "freq         144          225   \n",
       "mean         NaN          NaN   \n",
       "std          NaN          NaN   \n",
       "min          NaN          NaN   \n",
       "25%          NaN          NaN   \n",
       "50%          NaN          NaN   \n",
       "75%          NaN          NaN   \n",
       "max          NaN          NaN   \n",
       "\n",
       "                                                     text    lbl         max  \n",
       "count                                                 225  225.0  225.000000  \n",
       "unique                                                225    1.0         NaN  \n",
       "top     Alla moglie Murano,  18.3.1913 Carissima Livia...    0.0         NaN  \n",
       "freq                                                    1  225.0         NaN  \n",
       "mean                                                  NaN    NaN    0.907587  \n",
       "std                                                   NaN    NaN    0.078950  \n",
       "min                                                   NaN    NaN    0.734079  \n",
       "25%                                                   NaN    NaN    0.850349  \n",
       "50%                                                   NaN    NaN    0.941927  \n",
       "75%                                                   NaN    NaN    0.969514  \n",
       "max                                                   NaN    NaN    0.994531  "
      ]
     },
     "execution_count": 35,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_n = [df[df[\"lbl\"] == n] for n in range(num_topics)]\n",
    "df_n[0].describe(include=\"all\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "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>n</th>\n",
       "      <th>corpus</th>\n",
       "      <th>index</th>\n",
       "      <th>date</th>\n",
       "      <th>year</th>\n",
       "      <th>sender</th>\n",
       "      <th>senderLocation</th>\n",
       "      <th>recipient</th>\n",
       "      <th>recipientLocation</th>\n",
       "      <th>languages</th>\n",
       "      <th>mainLanguage</th>\n",
       "      <th>text</th>\n",
       "      <th>lbl</th>\n",
       "      <th>max</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>count</th>\n",
       "      <td>145.000000</td>\n",
       "      <td>145</td>\n",
       "      <td>145.000000</td>\n",
       "      <td>145</td>\n",
       "      <td>145.000000</td>\n",
       "      <td>145</td>\n",
       "      <td>145</td>\n",
       "      <td>145</td>\n",
       "      <td>145</td>\n",
       "      <td>145</td>\n",
       "      <td>145</td>\n",
       "      <td>145</td>\n",
       "      <td>145.0</td>\n",
       "      <td>145.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>unique</th>\n",
       "      <td>NaN</td>\n",
       "      <td>26</td>\n",
       "      <td>NaN</td>\n",
       "      <td>139</td>\n",
       "      <td>NaN</td>\n",
       "      <td>5</td>\n",
       "      <td>11</td>\n",
       "      <td>28</td>\n",
       "      <td>16</td>\n",
       "      <td>9</td>\n",
       "      <td>1</td>\n",
       "      <td>145</td>\n",
       "      <td>1.0</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>top</th>\n",
       "      <td>NaN</td>\n",
       "      <td>Svevo Montale</td>\n",
       "      <td>NaN</td>\n",
       "      <td>30/06/1926</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Ettore Schmitz</td>\n",
       "      <td>Trieste</td>\n",
       "      <td>Ettore Schmitz</td>\n",
       "      <td>Trieste</td>\n",
       "      <td>ITA</td>\n",
       "      <td>ITA</td>\n",
       "      <td>71 Rue du Cardinal Lemoine. V 20 febbraio 25 E...</td>\n",
       "      <td>1.0</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>freq</th>\n",
       "      <td>NaN</td>\n",
       "      <td>54</td>\n",
       "      <td>NaN</td>\n",
       "      <td>2</td>\n",
       "      <td>NaN</td>\n",
       "      <td>105</td>\n",
       "      <td>84</td>\n",
       "      <td>40</td>\n",
       "      <td>48</td>\n",
       "      <td>92</td>\n",
       "      <td>145</td>\n",
       "      <td>1</td>\n",
       "      <td>145.0</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>mean</th>\n",
       "      <td>361.779310</td>\n",
       "      <td>NaN</td>\n",
       "      <td>21.096552</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1926.068966</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.930819</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>std</th>\n",
       "      <td>367.341452</td>\n",
       "      <td>NaN</td>\n",
       "      <td>15.497009</td>\n",
       "      <td>NaN</td>\n",
       "      <td>2.985277</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.070038</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>min</th>\n",
       "      <td>5.000000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1897.000000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.733215</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25%</th>\n",
       "      <td>91.000000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>9.000000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1926.000000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.899014</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>50%</th>\n",
       "      <td>133.000000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>17.000000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1926.000000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.966178</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>75%</th>\n",
       "      <td>844.000000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>30.000000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1927.000000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.980556</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>max</th>\n",
       "      <td>891.000000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>62.000000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1928.000000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.991852</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                 n         corpus       index        date         year  \\\n",
       "count   145.000000            145  145.000000         145   145.000000   \n",
       "unique         NaN             26         NaN         139          NaN   \n",
       "top            NaN  Svevo Montale         NaN  30/06/1926          NaN   \n",
       "freq           NaN             54         NaN           2          NaN   \n",
       "mean    361.779310            NaN   21.096552         NaN  1926.068966   \n",
       "std     367.341452            NaN   15.497009         NaN     2.985277   \n",
       "min       5.000000            NaN    1.000000         NaN  1897.000000   \n",
       "25%      91.000000            NaN    9.000000         NaN  1926.000000   \n",
       "50%     133.000000            NaN   17.000000         NaN  1926.000000   \n",
       "75%     844.000000            NaN   30.000000         NaN  1927.000000   \n",
       "max     891.000000            NaN   62.000000         NaN  1928.000000   \n",
       "\n",
       "                sender senderLocation       recipient recipientLocation  \\\n",
       "count              145            145             145               145   \n",
       "unique               5             11              28                16   \n",
       "top     Ettore Schmitz        Trieste  Ettore Schmitz           Trieste   \n",
       "freq               105             84              40                48   \n",
       "mean               NaN            NaN             NaN               NaN   \n",
       "std                NaN            NaN             NaN               NaN   \n",
       "min                NaN            NaN             NaN               NaN   \n",
       "25%                NaN            NaN             NaN               NaN   \n",
       "50%                NaN            NaN             NaN               NaN   \n",
       "75%                NaN            NaN             NaN               NaN   \n",
       "max                NaN            NaN             NaN               NaN   \n",
       "\n",
       "       languages mainLanguage  \\\n",
       "count        145          145   \n",
       "unique         9            1   \n",
       "top          ITA          ITA   \n",
       "freq          92          145   \n",
       "mean         NaN          NaN   \n",
       "std          NaN          NaN   \n",
       "min          NaN          NaN   \n",
       "25%          NaN          NaN   \n",
       "50%          NaN          NaN   \n",
       "75%          NaN          NaN   \n",
       "max          NaN          NaN   \n",
       "\n",
       "                                                     text    lbl         max  \n",
       "count                                                 145  145.0  145.000000  \n",
       "unique                                                145    1.0         NaN  \n",
       "top     71 Rue du Cardinal Lemoine. V 20 febbraio 25 E...    1.0         NaN  \n",
       "freq                                                    1  145.0         NaN  \n",
       "mean                                                  NaN    NaN    0.930819  \n",
       "std                                                   NaN    NaN    0.070038  \n",
       "min                                                   NaN    NaN    0.733215  \n",
       "25%                                                   NaN    NaN    0.899014  \n",
       "50%                                                   NaN    NaN    0.966178  \n",
       "75%                                                   NaN    NaN    0.980556  \n",
       "max                                                   NaN    NaN    0.991852  "
      ]
     },
     "execution_count": 36,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_n[1].describe(include=\"all\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "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>n</th>\n",
       "      <th>corpus</th>\n",
       "      <th>index</th>\n",
       "      <th>date</th>\n",
       "      <th>year</th>\n",
       "      <th>sender</th>\n",
       "      <th>senderLocation</th>\n",
       "      <th>recipient</th>\n",
       "      <th>recipientLocation</th>\n",
       "      <th>languages</th>\n",
       "      <th>mainLanguage</th>\n",
       "      <th>text</th>\n",
       "      <th>lbl</th>\n",
       "      <th>max</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>count</th>\n",
       "      <td>65.000000</td>\n",
       "      <td>65</td>\n",
       "      <td>65.000000</td>\n",
       "      <td>65</td>\n",
       "      <td>65.000000</td>\n",
       "      <td>65</td>\n",
       "      <td>65</td>\n",
       "      <td>65</td>\n",
       "      <td>65</td>\n",
       "      <td>65</td>\n",
       "      <td>65</td>\n",
       "      <td>65</td>\n",
       "      <td>65.0</td>\n",
       "      <td>65.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>unique</th>\n",
       "      <td>NaN</td>\n",
       "      <td>2</td>\n",
       "      <td>NaN</td>\n",
       "      <td>65</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1</td>\n",
       "      <td>8</td>\n",
       "      <td>2</td>\n",
       "      <td>6</td>\n",
       "      <td>9</td>\n",
       "      <td>1</td>\n",
       "      <td>65</td>\n",
       "      <td>1.0</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>top</th>\n",
       "      <td>NaN</td>\n",
       "      <td>Schmitz Livia</td>\n",
       "      <td>NaN</td>\n",
       "      <td>11/08/1904</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Ettore Schmitz</td>\n",
       "      <td>Murano</td>\n",
       "      <td>Livia Veneziani</td>\n",
       "      <td>Trieste</td>\n",
       "      <td>ITA</td>\n",
       "      <td>ITA</td>\n",
       "      <td>Alla moglie [Pola,  1.10.901] Non ho ricevuta ...</td>\n",
       "      <td>2.0</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>freq</th>\n",
       "      <td>NaN</td>\n",
       "      <td>64</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1</td>\n",
       "      <td>NaN</td>\n",
       "      <td>65</td>\n",
       "      <td>36</td>\n",
       "      <td>64</td>\n",
       "      <td>58</td>\n",
       "      <td>34</td>\n",
       "      <td>65</td>\n",
       "      <td>1</td>\n",
       "      <td>65.0</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>mean</th>\n",
       "      <td>503.446154</td>\n",
       "      <td>NaN</td>\n",
       "      <td>18.707692</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1905.846154</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.918932</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>std</th>\n",
       "      <td>138.926918</td>\n",
       "      <td>NaN</td>\n",
       "      <td>18.622670</td>\n",
       "      <td>NaN</td>\n",
       "      <td>5.181791</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.072044</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>min</th>\n",
       "      <td>183.000000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1896.000000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.733602</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25%</th>\n",
       "      <td>438.000000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>4.000000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1903.000000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.864287</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>50%</th>\n",
       "      <td>500.000000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>12.000000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1905.000000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.948873</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>75%</th>\n",
       "      <td>559.000000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>28.000000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1908.000000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.970759</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>max</th>\n",
       "      <td>829.000000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>83.000000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1924.000000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.991863</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                 n         corpus      index        date         year  \\\n",
       "count    65.000000             65  65.000000          65    65.000000   \n",
       "unique         NaN              2        NaN          65          NaN   \n",
       "top            NaN  Schmitz Livia        NaN  11/08/1904          NaN   \n",
       "freq           NaN             64        NaN           1          NaN   \n",
       "mean    503.446154            NaN  18.707692         NaN  1905.846154   \n",
       "std     138.926918            NaN  18.622670         NaN     5.181791   \n",
       "min     183.000000            NaN   1.000000         NaN  1896.000000   \n",
       "25%     438.000000            NaN   4.000000         NaN  1903.000000   \n",
       "50%     500.000000            NaN  12.000000         NaN  1905.000000   \n",
       "75%     559.000000            NaN  28.000000         NaN  1908.000000   \n",
       "max     829.000000            NaN  83.000000         NaN  1924.000000   \n",
       "\n",
       "                sender senderLocation        recipient recipientLocation  \\\n",
       "count               65             65               65                65   \n",
       "unique               1              8                2                 6   \n",
       "top     Ettore Schmitz         Murano  Livia Veneziani           Trieste   \n",
       "freq                65             36               64                58   \n",
       "mean               NaN            NaN              NaN               NaN   \n",
       "std                NaN            NaN              NaN               NaN   \n",
       "min                NaN            NaN              NaN               NaN   \n",
       "25%                NaN            NaN              NaN               NaN   \n",
       "50%                NaN            NaN              NaN               NaN   \n",
       "75%                NaN            NaN              NaN               NaN   \n",
       "max                NaN            NaN              NaN               NaN   \n",
       "\n",
       "       languages mainLanguage  \\\n",
       "count         65           65   \n",
       "unique         9            1   \n",
       "top          ITA          ITA   \n",
       "freq          34           65   \n",
       "mean         NaN          NaN   \n",
       "std          NaN          NaN   \n",
       "min          NaN          NaN   \n",
       "25%          NaN          NaN   \n",
       "50%          NaN          NaN   \n",
       "75%          NaN          NaN   \n",
       "max          NaN          NaN   \n",
       "\n",
       "                                                     text   lbl        max  \n",
       "count                                                  65  65.0  65.000000  \n",
       "unique                                                 65   1.0        NaN  \n",
       "top     Alla moglie [Pola,  1.10.901] Non ho ricevuta ...   2.0        NaN  \n",
       "freq                                                    1  65.0        NaN  \n",
       "mean                                                  NaN   NaN   0.918932  \n",
       "std                                                   NaN   NaN   0.072044  \n",
       "min                                                   NaN   NaN   0.733602  \n",
       "25%                                                   NaN   NaN   0.864287  \n",
       "50%                                                   NaN   NaN   0.948873  \n",
       "75%                                                   NaN   NaN   0.970759  \n",
       "max                                                   NaN   NaN   0.991863  "
      ]
     },
     "execution_count": 37,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_n[2].describe(include=\"all\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "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>n</th>\n",
       "      <th>corpus</th>\n",
       "      <th>index</th>\n",
       "      <th>date</th>\n",
       "      <th>year</th>\n",
       "      <th>sender</th>\n",
       "      <th>senderLocation</th>\n",
       "      <th>recipient</th>\n",
       "      <th>recipientLocation</th>\n",
       "      <th>languages</th>\n",
       "      <th>mainLanguage</th>\n",
       "      <th>text</th>\n",
       "      <th>lbl</th>\n",
       "      <th>max</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>count</th>\n",
       "      <td>43.000000</td>\n",
       "      <td>43</td>\n",
       "      <td>43.000000</td>\n",
       "      <td>43</td>\n",
       "      <td>43.000000</td>\n",
       "      <td>43</td>\n",
       "      <td>43</td>\n",
       "      <td>43</td>\n",
       "      <td>43</td>\n",
       "      <td>43</td>\n",
       "      <td>43</td>\n",
       "      <td>43</td>\n",
       "      <td>43.0</td>\n",
       "      <td>43.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>unique</th>\n",
       "      <td>NaN</td>\n",
       "      <td>2</td>\n",
       "      <td>NaN</td>\n",
       "      <td>40</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1</td>\n",
       "      <td>9</td>\n",
       "      <td>3</td>\n",
       "      <td>7</td>\n",
       "      <td>10</td>\n",
       "      <td>1</td>\n",
       "      <td>43</td>\n",
       "      <td>1.0</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>top</th>\n",
       "      <td>NaN</td>\n",
       "      <td>Schmitz Livia</td>\n",
       "      <td>NaN</td>\n",
       "      <td>12/07/1901</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Ettore Schmitz</td>\n",
       "      <td>Trieste</td>\n",
       "      <td>Livia Veneziani</td>\n",
       "      <td>Salsomaggiore</td>\n",
       "      <td>ITA</td>\n",
       "      <td>ITA</td>\n",
       "      <td>Alla moglie Trieste, 17.5.99 Mia cara Livia, R...</td>\n",
       "      <td>3.0</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>freq</th>\n",
       "      <td>NaN</td>\n",
       "      <td>41</td>\n",
       "      <td>NaN</td>\n",
       "      <td>2</td>\n",
       "      <td>NaN</td>\n",
       "      <td>43</td>\n",
       "      <td>12</td>\n",
       "      <td>41</td>\n",
       "      <td>16</td>\n",
       "      <td>24</td>\n",
       "      <td>43</td>\n",
       "      <td>1</td>\n",
       "      <td>43.0</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>mean</th>\n",
       "      <td>358.116279</td>\n",
       "      <td>NaN</td>\n",
       "      <td>40.651163</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1901.162791</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.914950</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>std</th>\n",
       "      <td>108.170106</td>\n",
       "      <td>NaN</td>\n",
       "      <td>28.078024</td>\n",
       "      <td>NaN</td>\n",
       "      <td>3.054507</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.087611</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>min</th>\n",
       "      <td>170.000000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1895.000000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.745749</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25%</th>\n",
       "      <td>301.500000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>11.000000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1900.500000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.802893</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>50%</th>\n",
       "      <td>363.000000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>47.000000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1901.000000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.963052</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>75%</th>\n",
       "      <td>380.500000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>66.500000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1901.000000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.986270</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>max</th>\n",
       "      <td>807.000000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>82.000000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1915.000000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.993800</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                 n         corpus      index        date         year  \\\n",
       "count    43.000000             43  43.000000          43    43.000000   \n",
       "unique         NaN              2        NaN          40          NaN   \n",
       "top            NaN  Schmitz Livia        NaN  12/07/1901          NaN   \n",
       "freq           NaN             41        NaN           2          NaN   \n",
       "mean    358.116279            NaN  40.651163         NaN  1901.162791   \n",
       "std     108.170106            NaN  28.078024         NaN     3.054507   \n",
       "min     170.000000            NaN   1.000000         NaN  1895.000000   \n",
       "25%     301.500000            NaN  11.000000         NaN  1900.500000   \n",
       "50%     363.000000            NaN  47.000000         NaN  1901.000000   \n",
       "75%     380.500000            NaN  66.500000         NaN  1901.000000   \n",
       "max     807.000000            NaN  82.000000         NaN  1915.000000   \n",
       "\n",
       "                sender senderLocation        recipient recipientLocation  \\\n",
       "count               43             43               43                43   \n",
       "unique               1              9                3                 7   \n",
       "top     Ettore Schmitz        Trieste  Livia Veneziani     Salsomaggiore   \n",
       "freq                43             12               41                16   \n",
       "mean               NaN            NaN              NaN               NaN   \n",
       "std                NaN            NaN              NaN               NaN   \n",
       "min                NaN            NaN              NaN               NaN   \n",
       "25%                NaN            NaN              NaN               NaN   \n",
       "50%                NaN            NaN              NaN               NaN   \n",
       "75%                NaN            NaN              NaN               NaN   \n",
       "max                NaN            NaN              NaN               NaN   \n",
       "\n",
       "       languages mainLanguage  \\\n",
       "count         43           43   \n",
       "unique        10            1   \n",
       "top          ITA          ITA   \n",
       "freq          24           43   \n",
       "mean         NaN          NaN   \n",
       "std          NaN          NaN   \n",
       "min          NaN          NaN   \n",
       "25%          NaN          NaN   \n",
       "50%          NaN          NaN   \n",
       "75%          NaN          NaN   \n",
       "max          NaN          NaN   \n",
       "\n",
       "                                                     text   lbl        max  \n",
       "count                                                  43  43.0  43.000000  \n",
       "unique                                                 43   1.0        NaN  \n",
       "top     Alla moglie Trieste, 17.5.99 Mia cara Livia, R...   3.0        NaN  \n",
       "freq                                                    1  43.0        NaN  \n",
       "mean                                                  NaN   NaN   0.914950  \n",
       "std                                                   NaN   NaN   0.087611  \n",
       "min                                                   NaN   NaN   0.745749  \n",
       "25%                                                   NaN   NaN   0.802893  \n",
       "50%                                                   NaN   NaN   0.963052  \n",
       "75%                                                   NaN   NaN   0.986270  \n",
       "max                                                   NaN   NaN   0.993800  "
      ]
     },
     "execution_count": 38,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_n[3].describe(include=\"all\") "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "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>n</th>\n",
       "      <th>corpus</th>\n",
       "      <th>index</th>\n",
       "      <th>date</th>\n",
       "      <th>year</th>\n",
       "      <th>sender</th>\n",
       "      <th>senderLocation</th>\n",
       "      <th>recipient</th>\n",
       "      <th>recipientLocation</th>\n",
       "      <th>languages</th>\n",
       "      <th>mainLanguage</th>\n",
       "      <th>text</th>\n",
       "      <th>lbl</th>\n",
       "      <th>max</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>count</th>\n",
       "      <td>22.000000</td>\n",
       "      <td>22</td>\n",
       "      <td>22.000000</td>\n",
       "      <td>22</td>\n",
       "      <td>22.000000</td>\n",
       "      <td>22</td>\n",
       "      <td>22</td>\n",
       "      <td>22</td>\n",
       "      <td>22</td>\n",
       "      <td>22</td>\n",
       "      <td>22</td>\n",
       "      <td>22</td>\n",
       "      <td>22.0</td>\n",
       "      <td>22.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>unique</th>\n",
       "      <td>NaN</td>\n",
       "      <td>7</td>\n",
       "      <td>NaN</td>\n",
       "      <td>22</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1</td>\n",
       "      <td>3</td>\n",
       "      <td>7</td>\n",
       "      <td>4</td>\n",
       "      <td>5</td>\n",
       "      <td>1</td>\n",
       "      <td>22</td>\n",
       "      <td>1.0</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>top</th>\n",
       "      <td>NaN</td>\n",
       "      <td>Schmitz Livia</td>\n",
       "      <td>NaN</td>\n",
       "      <td>14/01/1901</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Ettore Schmitz</td>\n",
       "      <td>Trieste</td>\n",
       "      <td>Livia Veneziani</td>\n",
       "      <td>Trieste</td>\n",
       "      <td>ITA</td>\n",
       "      <td>ITA</td>\n",
       "      <td>Al fratello Elio Carissimo Elio, Non a pena ri...</td>\n",
       "      <td>4.0</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>freq</th>\n",
       "      <td>NaN</td>\n",
       "      <td>16</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1</td>\n",
       "      <td>NaN</td>\n",
       "      <td>22</td>\n",
       "      <td>15</td>\n",
       "      <td>16</td>\n",
       "      <td>19</td>\n",
       "      <td>16</td>\n",
       "      <td>22</td>\n",
       "      <td>1</td>\n",
       "      <td>22.0</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>mean</th>\n",
       "      <td>316.590909</td>\n",
       "      <td>NaN</td>\n",
       "      <td>9.363636</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1899.636364</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.914722</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>std</th>\n",
       "      <td>203.982880</td>\n",
       "      <td>NaN</td>\n",
       "      <td>9.742130</td>\n",
       "      <td>NaN</td>\n",
       "      <td>6.918199</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.071808</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>min</th>\n",
       "      <td>160.000000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1885.000000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.744916</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25%</th>\n",
       "      <td>174.250000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>2.250000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1896.000000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.876179</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>50%</th>\n",
       "      <td>191.500000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>8.000000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1896.500000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.918428</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>75%</th>\n",
       "      <td>476.750000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>9.000000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1904.000000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.979197</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>max</th>\n",
       "      <td>742.000000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>41.000000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1913.000000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.995723</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                 n         corpus      index        date         year  \\\n",
       "count    22.000000             22  22.000000          22    22.000000   \n",
       "unique         NaN              7        NaN          22          NaN   \n",
       "top            NaN  Schmitz Livia        NaN  14/01/1901          NaN   \n",
       "freq           NaN             16        NaN           1          NaN   \n",
       "mean    316.590909            NaN   9.363636         NaN  1899.636364   \n",
       "std     203.982880            NaN   9.742130         NaN     6.918199   \n",
       "min     160.000000            NaN   1.000000         NaN  1885.000000   \n",
       "25%     174.250000            NaN   2.250000         NaN  1896.000000   \n",
       "50%     191.500000            NaN   8.000000         NaN  1896.500000   \n",
       "75%     476.750000            NaN   9.000000         NaN  1904.000000   \n",
       "max     742.000000            NaN  41.000000         NaN  1913.000000   \n",
       "\n",
       "                sender senderLocation        recipient recipientLocation  \\\n",
       "count               22             22               22                22   \n",
       "unique               1              3                7                 4   \n",
       "top     Ettore Schmitz        Trieste  Livia Veneziani           Trieste   \n",
       "freq                22             15               16                19   \n",
       "mean               NaN            NaN              NaN               NaN   \n",
       "std                NaN            NaN              NaN               NaN   \n",
       "min                NaN            NaN              NaN               NaN   \n",
       "25%                NaN            NaN              NaN               NaN   \n",
       "50%                NaN            NaN              NaN               NaN   \n",
       "75%                NaN            NaN              NaN               NaN   \n",
       "max                NaN            NaN              NaN               NaN   \n",
       "\n",
       "       languages mainLanguage  \\\n",
       "count         22           22   \n",
       "unique         5            1   \n",
       "top          ITA          ITA   \n",
       "freq          16           22   \n",
       "mean         NaN          NaN   \n",
       "std          NaN          NaN   \n",
       "min          NaN          NaN   \n",
       "25%          NaN          NaN   \n",
       "50%          NaN          NaN   \n",
       "75%          NaN          NaN   \n",
       "max          NaN          NaN   \n",
       "\n",
       "                                                     text   lbl        max  \n",
       "count                                                  22  22.0  22.000000  \n",
       "unique                                                 22   1.0        NaN  \n",
       "top     Al fratello Elio Carissimo Elio, Non a pena ri...   4.0        NaN  \n",
       "freq                                                    1  22.0        NaN  \n",
       "mean                                                  NaN   NaN   0.914722  \n",
       "std                                                   NaN   NaN   0.071808  \n",
       "min                                                   NaN   NaN   0.744916  \n",
       "25%                                                   NaN   NaN   0.876179  \n",
       "50%                                                   NaN   NaN   0.918428  \n",
       "75%                                                   NaN   NaN   0.979197  \n",
       "max                                                   NaN   NaN   0.995723  "
      ]
     },
     "execution_count": 39,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_n[4].describe(include=\"all\") "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "By having a look at the pyLDAvis graph and the topical dataframe descriptions, we chose the following names for the topics:\n",
    "\n",
    "- The topic `family` contains matters related with the social life of Svevo, mostly with members of his family (Livia, Letizia, Olga) concerning day-to-day topics.\n",
    "\n",
    "\n",
    "- The topic `literature` includes comments or discussion on the literary work of other authors or Svevo himself.\n",
    "\n",
    "\n",
    "- The topic `travel` is related with the travelling of Svevo at the beginning of the 20th century, mostly for work matters. It includes dates, destinations and impressions (adjectives).\n",
    "\n",
    "\n",
    "- The topic `work` concerns the period in which Svevo started working in his father-in-law industry after quitting his job at a bank. It is tightly related with travel, since he travelled mostly for work.\n",
    "\n",
    "\n",
    "- The topic `health` contains a typical unformal medical lexicon (doctor, health, illness, fever, etc.). It matches in time with the period in which Svevo lost his brother and his father to illnesses, and in which Letizia was born."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "metadata": {},
   "outputs": [],
   "source": [
    "df['lbl'] = df['lbl'].replace(0.0, 'family')\n",
    "df['lbl'] = df['lbl'].replace(1.0, 'literature')\n",
    "df['lbl'] = df['lbl'].replace(2.0, 'travel')\n",
    "df['lbl'] = df['lbl'].replace(3.0, 'work')\n",
    "df['lbl'] = df['lbl'].replace(4.0, 'health')\n",
    "\n",
    "df.describe(include='all')\n",
    "df.to_csv(\"datasets/svevo_with_topics.csv\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "n_topics = 40\n",
    "models_test = []\n",
    "d_test = corpora.Dictionary(texts_it)\n",
    "d_test.filter_extremes(no_below=5, no_above=0.5)\n",
    "c_test = [d_test.doc2bow(text) for text in texts_it]\n",
    "\n",
    "for n in range(2, n_topics):\n",
    "    model_test = create_lda_model(d_test, c_test, n, passes = 200)\n",
    "    assess_model_goodness(model_test, texts_it, c_test, d_test, n)\n",
    "    models_test.append(model_test)\n",
    "\n",
    "silhouette_list = [compute_silhouette(model_test, c_test, len(model_test.print_topics())) for model_test in models_test]\n",
    "perplexity_list = [compute_perplexity(model_test, c_test) for model_test in models_test]\n",
    "coh_v = [compute_coherence(model_test, texts_it, c_test, d_test, coherence = \"c_v\") for model_test in models_test]\n",
    "\n",
    "x = range(2,n_topics)\n",
    "plt.plot(x, coh_v)\n",
    "plt.xlabel(\"Num Topics\")\n",
    "plt.ylabel(\"Coherence score\")\n",
    "plt.show()\n",
    "plt.plot(x, silhouette_list)\n",
    "plt.xlabel(\"Num Topics\")\n",
    "plt.ylabel(\"Silhouette index\")\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Extra: Large LDA Models\n",
    "\n",
    "We tried to evaluate models up to 40 topics with our previous approach, to prove that a low number of topic is best in this context.\n",
    "\n",
    "### DO NOT EXECUTE IF NOT STRICTLY NECESSARY, LONG RUNTIME!"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "n_topics = 40\n",
    "models_test = []\n",
    "d_test = corpora.Dictionary(texts_it)\n",
    "d_test.filter_extremes(no_below=5, no_above=0.5)\n",
    "c_test = [d_test.doc2bow(text) for text in texts_it]\n",
    "\n",
    "for n in range(2, n_topics):\n",
    "    model_test = create_lda_model(d_test, c_test, n, passes = 200)\n",
    "    assess_model_goodness(model_test, texts_it, c_test, d_test, n)\n",
    "    models_test.append(model_test)\n",
    "\n",
    "silhouette_list = [compute_silhouette(model_test, c_test, len(model_test.print_topics())) for model_test in models_test]\n",
    "perplexity_list = [compute_perplexity(model_test, c_test) for model_test in models_test]\n",
    "coh_v = [compute_coherence(model_test, texts_it, c_test, d_test, coherence = \"c_v\") for model_test in models_test]\n",
    "\n",
    "x = range(2,n_topics)\n",
    "plt.plot(x, coh_v)\n",
    "plt.xlabel(\"Num Topics\")\n",
    "plt.ylabel(\"Coherence score\")\n",
    "plt.show()\n",
    "plt.plot(x, silhouette_list)\n",
    "plt.xlabel(\"Num Topics\")\n",
    "plt.ylabel(\"Silhouette index\")\n",
    "plt.show()"
   ]
  }
 ],
 "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": "3.7.3"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}