unsupervised_learning.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Unsupervised Learning\n",
    "This notebook demonstrates unsupervised learning techniques. \n",
    "\n",
    "If you are interested in identifying hidden patterns in data, specifically unlabelled (or unstructured) data, then this notebook will help guide you through unsupervised learning from data with Python. We'll use some key Python statistics and machine learning libaries:\n",
    "\n",
    "* `sci-kit learn`\n",
    "* `statsmodels`\n",
    "\n",
    "For data, we've already collected over 200 speeches from the U.S. Secretary of Defense from the last 4 years. If you're curious to learn about webscraping or how this data was acquired, check out the notebook `data_collection.ipynb`. This is _text_, so it is inherently unstructured data. We'll use some basic natural language processing (NLP) techniques to convert this text into structured data.\n",
    "\n",
    "Let's get started by importing some packages we'll need:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "import pandas as pd\n",
    "import re\n",
    "\n",
    "import nltk\n",
    "from nltk.corpus import stopwords\n",
    "from nltk.stem.snowball import SnowballStemmer\n",
    "\n",
    "from sklearn import feature_extraction\n",
    "from sklearn.preprocessing import MinMaxScaler\n",
    "from sklearn.decomposition import PCA\n",
    "from sklearn.manifold import TSNE\n",
    "from sklearn.cluster import DBSCAN\n",
    "from sklearn.metrics.pairwise import cosine_similarity\n",
    "\n",
    "import hdbscan\n",
    "\n",
    "import seaborn as sns\n",
    "import matplotlib.pyplot as plt\n",
    "%matplotlib inline"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Dealing with stopwords"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Now, let's make sure that we have loaded NLTK's (a key Python NLP library) stopwords:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[nltk_data] Downloading package stopwords to\n",
      "[nltk_data]     /Users/brandon/nltk_data...\n",
      "[nltk_data]   Package stopwords is already up-to-date!\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "True"
      ]
     },
     "execution_count": 2,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "nltk.download('stopwords')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Stopwords are words that are common to language and are important in context, but don't carry significant semantic meaning with respect to an overall text. For example, if you were looking at a news article about nuclear weapons in North Korea you could determine what the article was about based on the frequency of words like \"nuclear\", \"weapons\", and \"North Korea\". You wouldn't be able to learn much from words like \"who\", \"when\", \"our\", \"me\", etc.\n",
    "\n",
    "Let's load nltk's English stopwords as variable called `stopwords`:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "['i', 'me', 'my', 'myself', 'we', 'our', 'ours', 'ourselves', 'you', \"you're\"]\n"
     ]
    }
   ],
   "source": [
    "stopwords = stopwords.words('english')\n",
    "print(stopwords[:10])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Tokenization and stemming"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Next, we are going to define some functions to help us resolve words like \"ran\", \"running\", \"runs\" into a canonical (or _stemmed_) form (e.g. \"run\"). We'll use the NLTK [Snowball Stemmer](https://www.nltk.org/_modules/nltk/stem/snowball.html)."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "stemmer = SnowballStemmer(\"english\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "def tokenize_and_stem(text):\n",
    "    # first tokenize by sentence, then by word to ensure that punctuation is caught as it's own token\n",
    "    tokens = [word for sent in nltk.sent_tokenize(text) for word in nltk.word_tokenize(sent)]\n",
    "    filtered_tokens = []\n",
    "    # filter out any tokens not containing letters (e.g., numeric tokens, raw punctuation)\n",
    "    for token in tokens:\n",
    "        if re.search('[a-zA-Z]', token):\n",
    "            if len(token) >= 4:\n",
    "                filtered_tokens.append(token)\n",
    "    stems = [stemmer.stem(t) for t in filtered_tokens]\n",
    "    return stems\n",
    "\n",
    "\n",
    "def tokenize_only(text):\n",
    "    # first tokenize by sentence, then by word to ensure that punctuation is caught as it's own token\n",
    "    tokens = [word.lower() for sent in nltk.sent_tokenize(text) for word in nltk.word_tokenize(sent)]\n",
    "    filtered_tokens = []\n",
    "    # filter out any tokens not containing letters (e.g., numeric tokens, raw punctuation)\n",
    "    for token in tokens:\n",
    "        if re.search('[a-zA-Z]', token):\n",
    "            if len(token) >= 4:\n",
    "                filtered_tokens.append(token)\n",
    "    return filtered_tokens"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Note that we'll also need to **_tokenize_** the speeches. This is the process of converting text into words.\n",
    "\n",
    "For example, we would take \n",
    "```\n",
    "\"The quick brown fox jumped over the lazy dog.\"\n",
    "```\n",
    "\n",
    "and tokenize it into:\n",
    "```\n",
    "[\"The\", \"quick\", \"brown\", \"fox\", \"jumped\", \"over\", \"the\", \"lazy\", \"dog\", \".\"]\n",
    "```\n",
    "\n",
    "This lets us generate frequency counts of words within speeches."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Parsing our data\n",
    "Let's read in the speeches as a dataframe and take a look at what we're dealing with!"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "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>date</th>\n",
       "      <th>speech</th>\n",
       "      <th>title</th>\n",
       "      <th>url</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>March 30, 2015</td>\n",
       "      <td>Good morning everyone! What a great looking cr...</td>\n",
       "      <td>Remarks by Secretary Carter on the Force of th...</td>\n",
       "      <td>https://www.defense.gov/News/Speeches/Speech-V...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>Jan. 11, 2017</td>\n",
       "      <td>Good evening everyone. Thanks  for that kind i...</td>\n",
       "      <td>Remarks on \"Defense at A Time Of Strategic Tra...</td>\n",
       "      <td>https://www.defense.gov/News/Speeches/Speech-V...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>Jan. 14, 2016</td>\n",
       "      <td>Ladies and gentlemen, excellencies, Congressma...</td>\n",
       "      <td>Remarks at SOUTHCOM Change of Command</td>\n",
       "      <td>https://www.defense.gov/News/Speeches/Speech-V...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>Nov. 14, 2014</td>\n",
       "      <td>Good morning. Earlier this year, following rev...</td>\n",
       "      <td>Statement on the Nuclear Enterprise Review &amp; R...</td>\n",
       "      <td>https://www.defense.gov/News/Speeches/Speech-V...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>April 27, 2016</td>\n",
       "      <td>I. PURPOSE OF THIS TESTIMONY Chairman Cochran,...</td>\n",
       "      <td>Submitted Statement -- Senate Appropriations C...</td>\n",
       "      <td>https://www.defense.gov/News/Speeches/Speech-V...</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "             date                                             speech  \\\n",
       "0  March 30, 2015  Good morning everyone! What a great looking cr...   \n",
       "1   Jan. 11, 2017  Good evening everyone. Thanks  for that kind i...   \n",
       "2   Jan. 14, 2016  Ladies and gentlemen, excellencies, Congressma...   \n",
       "3   Nov. 14, 2014  Good morning. Earlier this year, following rev...   \n",
       "4  April 27, 2016  I. PURPOSE OF THIS TESTIMONY Chairman Cochran,...   \n",
       "\n",
       "                                               title  \\\n",
       "0  Remarks by Secretary Carter on the Force of th...   \n",
       "1  Remarks on \"Defense at A Time Of Strategic Tra...   \n",
       "2              Remarks at SOUTHCOM Change of Command   \n",
       "3  Statement on the Nuclear Enterprise Review & R...   \n",
       "4  Submitted Statement -- Senate Appropriations C...   \n",
       "\n",
       "                                                 url  \n",
       "0  https://www.defense.gov/News/Speeches/Speech-V...  \n",
       "1  https://www.defense.gov/News/Speeches/Speech-V...  \n",
       "2  https://www.defense.gov/News/Speeches/Speech-V...  \n",
       "3  https://www.defense.gov/News/Speeches/Speech-V...  \n",
       "4  https://www.defense.gov/News/Speeches/Speech-V...  "
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df = pd.read_csv('data/SecDef_Speeches.csv')\n",
    "df.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Now we can tokenize and stem the speeches:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [],
   "source": [
    "totalvocab_stemmed = []\n",
    "totalvocab_tokenized = []\n",
    "\n",
    "for i in df.speech:\n",
    "    allwords_stemmed = tokenize_and_stem(i)\n",
    "    totalvocab_stemmed.extend(allwords_stemmed)\n",
    "\n",
    "    allwords_tokenized = tokenize_only(i)\n",
    "    totalvocab_tokenized.extend(allwords_tokenized)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We need to generate a vocabulary: this will be a mapping from stemmed version of words back to their original form. That way we can lookup a stemmed word and back into what it actually was (in human readable form)."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "There are 302045 items in our vocabulary.\n"
     ]
    }
   ],
   "source": [
    "vocab = pd.DataFrame({'words': totalvocab_tokenized}, index = totalvocab_stemmed)\n",
    "print('There are ' + str(vocab.shape[0]) + ' items in our vocabulary.')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### tf-idf and calculating document similarity\n",
    "To get a ff-idf matrix, first count word occurrences by document. This is transformed into a document-term matrix (dtm). This is also just called a term frequency matrix.\n",
    "\n",
    "Then apply the term frequency-inverse document frequency weighting: words that occur frequently within a document but not frequently within the corpus receive a higher weighting as these words are assumed to contain more meaning in relation to the document. \n",
    "\n",
    "We will use this tf-idf matrix as a structured representation of the relationship between documents. We can use this to perform unsupervised learning: "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "CPU times: user 11.9 s, sys: 102 ms, total: 12 s\n",
      "Wall time: 12 s\n",
      "(204, 1593)\n"
     ]
    }
   ],
   "source": [
    "from sklearn.feature_extraction.text import TfidfVectorizer\n",
    "\n",
    "#define vectorizer parameters\n",
    "tfidf_vectorizer = TfidfVectorizer(max_df=0.8, max_features=200000,\n",
    "                                 min_df=0.1, stop_words='english',\n",
    "                                 use_idf=True, tokenizer=tokenize_and_stem, ngram_range=(1,3))\n",
    "\n",
    "%time tfidf_matrix = tfidf_vectorizer.fit_transform(df.speech) #fit the vectorizer to speeches\n",
    "\n",
    "print(tfidf_matrix.shape)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Note that the tf-idf matrix has 204 rows (one for each speech) and 1,593 columns (one for each word which met the thresholds we specified for our `TfidfVectorizer`.\n",
    "\n",
    "We can get all the terms from our tf-idf matrix:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [],
   "source": [
    "terms = tfidf_vectorizer.get_feature_names()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### K-means clustering\n",
    "Now we're ready to perform clustering on our speeches. We'll pick an arbitrary, but human understandable, number of clusters. Let's try for 12 clusters."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "CPU times: user 8.9 s, sys: 1.9 s, total: 10.8 s\n",
      "Wall time: 2.99 s\n"
     ]
    }
   ],
   "source": [
    "from sklearn.cluster import KMeans\n",
    "\n",
    "num_clusters = 12\n",
    "\n",
    "km = KMeans(n_clusters=num_clusters)\n",
    "\n",
    "%time km.fit(tfidf_matrix)\n",
    "\n",
    "clusters = km.labels_.tolist()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We need to add the cluster labels to our dataset. This let's us see how many speeches ended up in each cluster:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "cluster\n",
       "0     14\n",
       "1     18\n",
       "2     13\n",
       "3     16\n",
       "4     23\n",
       "5     11\n",
       "6     41\n",
       "7      4\n",
       "8     25\n",
       "9     21\n",
       "10    13\n",
       "11     5\n",
       "Name: cluster, dtype: int64"
      ]
     },
     "execution_count": 12,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df['cluster'] = clusters\n",
    "df.groupby('cluster')['cluster'].count()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Now we can inspect the top terms in each of these clusters and compare them with the speech titles. This should give us a good sense of whether or not our clustering was successful."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {
    "scrolled": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\u001b[1mTop terms per cluster:\u001b[0m\n",
      "\n",
      "\u001b[94m\u001b[1mCluster 0 Words:\u001b[0m\n",
      "\t-president\n",
      "\t-budget\n",
      "\t-health\n",
      "\t-sequestration\n",
      "\t-family\n",
      "\t-care\n",
      "\t-reforming\n",
      "\t-secretary\n",
      "\t-secretary\n",
      "\t-want\n",
      "\n",
      "\u001b[92m\u001b[1mCluster 0 Speech Titles:\n",
      "\u001b[0m\n",
      "\t- Remarks at Armed Forces Farewell Parade for President Obama\n",
      "\t- Farewell Message to the Department of Defense From Secretary of Defense Chuck Hagel\n",
      "\t- Independence Day Message\n",
      "\t- Pentagon All-Hands Meeting\n",
      "\t- Armed Forces Day Message\n",
      "\t- Ceremonial Swearing-In\n",
      "\t- Remarks at the Global Health Security Agenda\n",
      "\t- Opening Summary -- House Armed Services Committee (Budget Request)\n",
      "\t- Remarks by Secretary Hagel at the White House\n",
      "\t- Opening Summary -- Senate Armed Services Committee (Budget Request)\n",
      "\n",
      "\n",
      "\n",
      "\u001b[94m\u001b[1mCluster 1 Words:\u001b[0m\n",
      "\t-innovative\n",
      "\t-technologies\n",
      "\t-companies\n",
      "\t-future\n",
      "\t-things\n",
      "\t-board\n",
      "\t-investing\n",
      "\t-business\n",
      "\t-america\n",
      "\t-better\n",
      "\n",
      "\u001b[92m\u001b[1mCluster 1 Speech Titles:\n",
      "\u001b[0m\n",
      "\t- \"Defense Innovation Days\" Opening Keynote (Southeastern New England Defense Industry Alliance)\n",
      "\t- Remarks on \"Building the First Link to the Force of the Future\" (George Washington University)\n",
      "\t- Remarks Announcing a New Manufacturing Innovation Institute in Cambridge, Massachusetts\n",
      "\t- Remarks on \"The Path to an Innovative Future for Defense\" (CSIS Third Offset Strategy Conference)\n",
      "\t- Remarks on “Securing the Oceans, the Internet, and Space: Protecting the Domains that Drive Prosperi\n",
      "\t- Remarks before Fireside Chat at the University of Texas\n",
      "\t- Remarks on Receiving the Distinguished Service Award from the Center for the National Interest\n",
      "\t- Remarks at DARPA's \"Wait, What?\" Future Technology Forum\n",
      "\t- Remarks at the World Economic Forum dinner presenting the Secretary of Defense IDEAS Award to Dr. Wi\n",
      "\t- Remarks Announcing a New Manufacturing Innovation Institute in Silicon Valley\n",
      "\n",
      "\n",
      "\n",
      "\u001b[94m\u001b[1mCluster 2 Words:\u001b[0m\n",
      "\t-nato\n",
      "\t-ministers\n",
      "\t-welcomed\n",
      "\t-russia\n",
      "\t-ukraine\n",
      "\t-relationship\n",
      "\t-member\n",
      "\t-united\n",
      "\t-ambassador\n",
      "\t-alliances\n",
      "\n",
      "\u001b[92m\u001b[1mCluster 2 Speech Titles:\n",
      "\u001b[0m\n",
      "\t- Remarks at Bilateral Meeting with Portuguese Defense Minister Azeredo Lopes\n",
      "\t- Remarks at Bilateral Meeting with Ukrainian Minister of Defense Poltorak\n",
      "\t- Remarks at a Bilateral Meeting with Montenegrin Minister of Defense Boskovic\n",
      "\t- Bilateral Meeting with Australian Defence Minister Payne\n",
      "\t- Bilateral Meeting with Kuwaiti Deputy Prime Minister and Defense Minister al-Sabah\n",
      "\t- Remarks at Spanish Center for Advanced Studies of National Defense: \"Strong and Principled Security \n",
      "\t- Remarks at Atlantik Brücke: \"U.S., Germany, & NATO Are Moving Forward Together\"\n",
      "\t- Secretary of Defense James Mattis Remarks with President Petro Poroshenko \n",
      "\t- Meeting with Polish Minister of National Defence Macierewicz\n",
      "\t- Remarks at Bilateral Meeting with Secretary of State for Defence Williamson\n",
      "\n",
      "\n",
      "\n",
      "\u001b[94m\u001b[1mCluster 3 Words:\u001b[0m\n",
      "\t-budget\n",
      "\t-investing\n",
      "\t-billions\n",
      "\t-capable\n",
      "\t-future\n",
      "\t-sequestration\n",
      "\t-technologies\n",
      "\t-fiscal\n",
      "\t-need\n",
      "\t-fund\n",
      "\n",
      "\u001b[92m\u001b[1mCluster 3 Speech Titles:\n",
      "\u001b[0m\n",
      "\t- Submitted Statement -- Senate Appropriations Committee-Defense (FY 2017 Budget Request)\n",
      "\t- Opening Statement -- Senate Armed Services Committee (FY 2017 Budget Request)\n",
      "\t- Opening Statement -- House Appropriations Committee-Defense (FY 2017 Budget Request)\n",
      "\t- Submitted Statement -- Senate Appropriations Committee-Defense (Budget Request)\n",
      "\t- Submitted Statement -- Senate Armed Services Committee (Budget Request)\n",
      "\t- Remarks Previewing the FY 2017 Defense Budget\n",
      "\t- Reagan National Defense Forum Keynote\n",
      "\t- Submitted Statement -- House Armed Services Committee (FY 2017 Budget Request)\n",
      "\t- Opening Statement -- House Armed Services Committee (FY 2017 Budget Request)\n",
      "\t- Opening Statement -- Senate Appropriations Committee-Defense (FY 2017 Budget Request)\n",
      "\n",
      "\n",
      "\n",
      "\u001b[94m\u001b[1mCluster 4 Words:\u001b[0m\n",
      "\t-isil\n",
      "\t-campaigns\n",
      "\t-iraqi\n",
      "\t-coalition\n",
      "\t-syria\n",
      "\t-iraq\n",
      "\t-defeat\n",
      "\t-local\n",
      "\t-partners\n",
      "\t-operating\n",
      "\n",
      "\u001b[92m\u001b[1mCluster 4 Speech Titles:\n",
      "\u001b[0m\n",
      "\t- Remarks on \"Defense at A Time Of Strategic Transition\" (CSIS Sam Nunn National Security Leadership P\n",
      "\t- Counter ISIL Campaign Remarks at the Ecole Militaire, Paris\n",
      "\t- Remarks to the XVIII Airborne Corps On the Counter-ISIL Campaign\n",
      "\t- Opening Remarks at Counter-ISIL Defense Minister Meeting\n",
      "\t- Remarks at the NORAD/NORTHCOM Change of Command\n",
      "\t- Remarks on the \"U.S.-UK Alliance and the Future of the Principled International Order\"\n",
      "\t- Secretary of Defense James Mattis Remarks with Brett McGurk and LTG Townsend\n",
      "\t- Opening Remarks at Counter-ISIL Foreign and Defense Minister Meeting\n",
      "\t- Statement on Counter-ISIL Operations and U.S. Military Strategy in the Middle East before the Senate\n",
      "\t- Statement on the Administration's Strategy and Military Campaign Against ISIL Before the House Armed\n",
      "\n",
      "\n",
      "\n",
      "\u001b[94m\u001b[1mCluster 5 Words:\u001b[0m\n",
      "\t-army\n",
      "\t-strategic\n",
      "\t-secretary\n",
      "\t-challenges\n",
      "\t-leaders\n",
      "\t-soldiers\n",
      "\t-ready\n",
      "\t-family\n",
      "\t-john\n",
      "\t-department\n",
      "\n",
      "\u001b[92m\u001b[1mCluster 5 Speech Titles:\n",
      "\u001b[0m\n",
      "\t- Welcoming Ceremony for Senior DoD Leaders\n",
      "\t- Remarks before Administering the Oath of Office to Secretary of the Army Eric Fanning\n",
      "\t- Remarks at the Arrival Ceremony for Secretary of the Army Eric Fanning\n",
      "\t- Remarks on the Women-in-Service Review\n",
      "\t- Army Chief of Staff Change of Responsibility\n",
      "\t- Remarks Honoring Zbigniew Brzezinski with the DOD Distinguished Public Service Award\n",
      "\t- Remarks Honoring Secretary Madeleine Albright with the DOD Distinguished Public Service Award\n",
      "\t- Remarks Honoring Senator Carl Levin and Senator John Warner\n",
      "\t- Association of the United States Army (AUSA)\n",
      "\t- Remarks at the National Guard Bureau Change of Responsibility \n",
      "\n",
      "\n",
      "\n",
      "\u001b[94m\u001b[1mCluster 6 Words:\u001b[0m\n",
      "\t-family\n",
      "\t-serve\n",
      "\t-every\n",
      "\t-veterans\n",
      "\t-women\n",
      "\t-service\n",
      "\t-want\n",
      "\t-future\n",
      "\t-american\n",
      "\t-honor\n",
      "\n",
      "\u001b[92m\u001b[1mCluster 6 Speech Titles:\n",
      "\u001b[0m\n",
      "\t- Remarks by Secretary Carter on the Force of the Future\n",
      "\t- Remarks to USAFA Upper Class Cadets\n",
      "\t- Remarks at Pentagon Vietnam War Exhibit Opening Ceremony\n",
      "\t- Remarks on \"Forging Two New Links to the Force of the Future\"\n",
      "\t- Veterans Day Message\n",
      "\t- Remarks on \"The Next Two Links to the Force of the Future\"\n",
      "\t- USO Gala\n",
      "\t- Remarks at the Fisher House Foundation's 25th Anniversary Gala\n",
      "\t- Remarks to the U.S. Military Academy Corps of Cadets\n",
      "\t- Retirement Ceremony of General John \"Jay\" Paxton\n",
      "\n",
      "\n",
      "\n",
      "\u001b[94m\u001b[1mCluster 7 Words:\u001b[0m\n",
      "\t-sexual\n",
      "\t-assault\n",
      "\t-sexual\n",
      "\t-report\n",
      "\t-prevent\n",
      "\t-responsibilities\n",
      "\t-courage\n",
      "\t-every\n",
      "\t-trusted\n",
      "\t-issue\n",
      "\n",
      "\u001b[92m\u001b[1mCluster 7 Speech Titles:\n",
      "\u001b[0m\n",
      "\t- Statement on Sexual Assault in the Military\n",
      "\t- Remarks by Secretary Carter on Sexual Assault Prevention and Response (Georgetown University)\n",
      "\t- Remarks at Sexual Assault Response Coordinators of the Year Awards\n",
      "\t- Message on Sexual Assault Awareness and Prevention Month\n",
      "\n",
      "\n",
      "\n",
      "\u001b[94m\u001b[1mCluster 8 Words:\u001b[0m\n",
      "\t-honor\n",
      "\t-family\n",
      "\t-secretary\n",
      "\t-senator\n",
      "\t-serve\n",
      "\t-service\n",
      "\t-remembering\n",
      "\t-soldiers\n",
      "\t-president\n",
      "\t-american\n",
      "\n",
      "\u001b[92m\u001b[1mCluster 8 Speech Titles:\n",
      "\u001b[0m\n",
      "\t- Remarks at the Opening Ceremony of the 2015 Warrior Games\n",
      "\t- Medal of Honor Ceremony for Lieutenant Colonel Charles Kettles\n",
      "\t- September 11th Memorial Observance\n",
      "\t- John Glenn College of Public Affairs - The Ohio State University (2017 Excellence in Public Service \n",
      "\t- U.S. Military Academy Graduation and Commissioning\n",
      "\t- Portrait Unveiling for Secretary Leon Panetta\n",
      "\t- Remarks by Secretary Mattis at Portrait Unveiling in Honor of Former Secretary of Defense Ashton B. \n",
      "\t- Remarks at the 25th Anniversary of the Cooperative Threat Reduction\n",
      "\t- Medal of Honor Hall of Heroes Ceremony For Specialist Five James C. McCloughan\n",
      "\t- Remarks at Medal Ceremony for Train Heroes\n",
      "\n",
      "\n",
      "\n",
      "\u001b[94m\u001b[1mCluster 9 Words:\u001b[0m\n",
      "\t-region\n",
      "\t-states\n",
      "\t-asia-pacific\n",
      "\t-united\n",
      "\t-united\n",
      "\t-china\n",
      "\t-peace\n",
      "\t-principle\n",
      "\t-relationship\n",
      "\t-philippines\n",
      "\n",
      "\u001b[92m\u001b[1mCluster 9 Speech Titles:\n",
      "\u001b[0m\n",
      "\t- \"Networking Defense in the 21st Century\" (Remarks at CNAS)\n",
      "\t- Secretary of Defense Joint Remarks with South Korean Defense Minister Han Min Koo, Seoul, Republic o\n",
      "\t- Remarks on the Next Phase of the U.S. Rebalance to the Asia-Pacific (McCain Institute, Arizona State\n",
      "\t- Remarks at XII Conference of Defense Ministers of the Americas\n",
      "\t- Statement on the Impacts of the Joint Comprehensive Plan of Action before the Senate Armed Services \n",
      "\t- Secretary of Defense Ash Carter Dinner Remarks UN Peacekeeping Defense Ministerial\n",
      "\t- Remarks on \"Asia-Pacific's Principled Security Network\" at 2016 IISS Shangri-La Dialogue\n",
      "\t- Shangri-La Dialogue\n",
      "\t- IISS Shangri-La Dialogue: “A Regional Security Architecture Where Everyone Rises”\n",
      "\t- Remarks at F-35I Arrival Ceremony\n",
      "\n",
      "\n",
      "\n",
      "\u001b[94m\u001b[1mCluster 10 Words:\u001b[0m\n",
      "\t-command\n",
      "\t-general\n",
      "\t-marines\n",
      "\t-family\n",
      "\t-operating\n",
      "\t-admiring\n",
      "\t-marked\n",
      "\t-leadership\n",
      "\t-john\n",
      "\t-joint\n",
      "\n",
      "\u001b[92m\u001b[1mCluster 10 Speech Titles:\n",
      "\u001b[0m\n",
      "\t- Remarks at SOUTHCOM Change of Command\n",
      "\t- U.S. Special Operations Command Change of Command\n",
      "\t- Remarks at Chief of Naval Operations Change of Responsibility\n",
      "\t- Remarks at the Chief of Staff of the Air Force Retirement Ceremony\n",
      "\t- Army & Navy Personnel Announcement\n",
      "\t- U.S. Transportation Command Assumption of Command\n",
      "\t- Remarks at Chairman of the Joint Chiefs of Staff Change of Responsibility\n",
      "\t- Marine Corps Passage of Command\n",
      "\t- Remarks at Commandant of the Marine Corps Passage of Command\n",
      "\t- Remarks at CENTCOM Change of Command\n",
      "\n",
      "\n",
      "\n",
      "\u001b[94m\u001b[1mCluster 11 Words:\u001b[0m\n",
      "\t-nuclear\n",
      "\t-deterrence\n",
      "\t-enterprise\n",
      "\t-space\n",
      "\t-nuclear\n",
      "\t-nuclear\n",
      "\t-weapons\n",
      "\t-review\n",
      "\t-missions\n",
      "\t-missile\n",
      "\n",
      "\u001b[92m\u001b[1mCluster 11 Speech Titles:\n",
      "\u001b[0m\n",
      "\t- Statement on the Nuclear Enterprise Review & Reforms\n",
      "\t- Remarks at STRATCOM Change of Command\n",
      "\t- Remarks on \"Sustaining Nuclear Deterrence\"\n",
      "\t- A Message to the Force on Our Nuclear Enterprise\n",
      "\t- Remarks at Secretary of the Air Force Farewell Ceremony\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n"
     ]
    }
   ],
   "source": [
    "class color:\n",
    "   PURPLE = '\\033[95m'\n",
    "   CYAN = '\\033[96m'\n",
    "   DARKCYAN = '\\033[36m'\n",
    "   BLUE = '\\033[94m'\n",
    "   GREEN = '\\033[92m'\n",
    "   YELLOW = '\\033[93m'\n",
    "   RED = '\\033[91m'\n",
    "   BOLD = '\\033[1m'\n",
    "   UNDERLINE = '\\033[4m'\n",
    "   END = '\\033[0m'\n",
    "\n",
    "print(\"{0}Top terms per cluster:{1}\".format(color.BOLD,color.END))\n",
    "print()\n",
    "#sort cluster centers by proximity to centroid\n",
    "order_centroids = km.cluster_centers_.argsort()[:, ::-1] \n",
    "\n",
    "for i in range(num_clusters):\n",
    "    print(\"{0}Cluster {1} Words:{2}\".format(color.BLUE + color.BOLD, i, color.END))\n",
    "    \n",
    "    for ind in order_centroids[i, :10]:\n",
    "        word = str(vocab.loc[terms[ind].split(' ')].values.tolist()[0][0])\n",
    "        print('\\t-{0}'.format(word))\n",
    "    print('')\n",
    "    \n",
    "    print(\"{0}Cluster {1} Speech Titles:\\n{2}\".format(color.GREEN + color.BOLD, i, color.END))\n",
    "    titles = df[df['cluster'] == i].title\n",
    "    for title in titles[:10]:\n",
    "        print('\\t- {0}'.format(title[:100]))\n",
    "    print('\\n\\n')\n",
    "    \n",
    "print('\\n')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Visualizing our clusters"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Let's use t-Distributed Stochastic Neighbor Embedding, or [t-SNE ](https://lvdmaaten.github.io/tsne/), for dimensionality reduction. This let's us project highly dimensional data, such as our tfidf matrix, into 2 dimensional space. This will let us visualize the clustering in 2-dimensions. Remember that much is lost projecting data with over 1500 features into 2d space!"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [],
   "source": [
    "data = tfidf_matrix.toarray()\n",
    "projection = TSNE().fit_transform(data)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAsQAAAHZCAYAAACB9S1bAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADl0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uIDIuMi4yLCBodHRwOi8vbWF0cGxvdGxpYi5vcmcvhp/UCwAAIABJREFUeJzs3edzXFmaJvbnuvQWSCDhDWEJgABIgqZsl7cz1VVd1TM9babHaHZ2QhEbsSH9A5IiFCFpJa1CH6TQamY1ZmfaVVdV93SxyC7Hcix6C+99ZsKld9fpAwgQINLcBGEygfcX0RFdmSfvPSBB4MmT73kPo6oqCCGEEEIIOazY/Z4AIYQQQggh+4kCMSGEEEIIOdQoEBNCCCGEkEONAjEhhBBCCDnUKBATQgghhJBDjQIxIYQQQgg51CgQE0IIIYSQQ40CMSGEEEIIOdQoEBNCCCGEkEONAjEhhBBCCDnUKBATQgghhJBDjQIxIYQQQgg51CgQE0IIIYSQQ40CMSGEEEIIOdQoEBNCCCGEkEONAjEhhBBCCDnUKBATQgghhJBDjd+pC1VWVt4EUA8gDGBkp65LCCGEEELIQxoBWACMz87OHn/Ui+1YIMZqGLbf/1/lDl6XEEIIIYSQVOp34iI7GYjDAOw2mw3t7e07eFlCCCGEEEIe6O3tRTAYBFbz5yPbyUA8AqCyvb0dv/rVr3bwsoQQQgghhDzwzjvv4NKlS8AOlenSpjpCCCGEEHKoUSAmhBBCCCGHGgViQgghhBByqFEgJoQQQgghhxoFYkIIIYQQcqhRICaEEEIIIYcaBWJCCCGEEHKoUSAmhBBCCCGHGgViQgghhBByqFEgJoQQQgghh9pOHt1MCCF5SZEVzE4HEIsmwXEsyiptMFv0+z0tQggheYICMSHkwFIUFXevz2Cw14tYVFx/nGGA6roinDhbA7vTuI8zJIQQkg8oEBOS55YWwhgdWEA0mgTPcyivsqOusRgcRxVPmSiKis/ODWBm0r/lOVUFpsaX4Z0L4qXvtqHIZd6HGRJCCMkXFIgJyVORUAJf/H4IPk940+Ojgwu49vUETj1ZhyPNJfs0u/x39/pMyjC8USIh4dMPB/C9H58AyzJ7NDNCCCH5hpaYCMlD0UgS5967tyUMr4nHJXz58QhGBnx7PLPCoMgKBnu9msZGwklMjS3t8owIIYTkMwrEhOShG99OIRJOZh13+YtxJBPSHsyosKxuoBOzD7xvdGhxF2dDCCEk31EgJiTPxGMiJka0BTRJUmiVOIVoJPubiY1iOY4nhBBysFAgJiTPzE37Icuq5vHTEyu7OJvCxPO5/WjjOKofJoSQw4wCMSF5RkzKuzr+MCirtIHJIeOWVzl2bzKEEELyHgViQvKM3pBb85dcxx8GZose1XVFmsayLIPmdvcuz4gQQkg+o9+khOSZylonBB2neeW3rtG1yzPaKhGXMDrow+ToEpIJGTo9h9qGYjS0lO5oQFdVFXPTfnjngpBlFWaLDkeaS2AwCllfe+JsDTxzASQTmf8cO3uqYDLrdmrKhBBCChAFYkLyjCBwaGgpwcBdT9axej2P+qbiPZjVA2NDi/jm0xGEQwlEo0lIogIA6LvjgdU6gmdeaUZze9kj32d6fBlXv55AKJjY9Pj1S1M40uzCmafqwQtc2tfbnUa89EYbPjs3mLJjB8sy6DxZia6eqkeeKyGEkMJGgZiQPHTiTA0WPCEsLUTSjmFZBk+/1ASeTx8Kd9rk2BIunh/E8mIUkrR55VUSZawsR/GbX9zBq28paO+u2PZ9xocX8eXHw1BT7C1UFBUjAwsI+mN48Y22jF9/cYkF3/vRcUyNL2N0aBGxSBIcx6C8yoHmdjetDBNCCAFAgZiQvCToOLz0RhuufDWB8eFFKMrmZFhUbMKpp+pRVmHbszmpiorLX4xjeTECSVLSjpMlBRd+04cjTS4YtxE4E3ER33w2mjIMb+TzhHH3+iyOn6nJOI7lWNQ1uvaltIQQQkhhoEBMSJ7S6Xk8+XwjTj5Wg4mRJUQjSXAci4oaB0rLrHs+n+nJFSz5whnD8JpkQsI3n4/i+deP5nyf4X6fpnusje3qqQLL0f5gQggh20eBmJA8ZzTpcLSzfL+nAc9MANEcTn8bGVjAs6+2gmVz6/E7MaL9GOVYVIRnLoiKamqbRgghZPtoWYUQokkyKUEStfc8TiYkRMOJ7ANTvC4XCTq6mhBCyCOiQEwI0cRo1uc0nuOYrHXAqQgZOkekotPt3aZCQgghBxMFYkKIJs1tpeA01uoyDGCxGWA0Z+8X/LDqem0HagCrbefce7ixkBBCyMFEgZgQoonVZkBlrbZaXb1BQFObe1st4Zrb3ZrrjhtaS/a07RwhhJCDiQIxIUSzV97sgF6feS+uIHBwFptw9Nj2DucwmXU49URt1nGOIiMdqkEIIWRHUCAmhGjmLDbh7T89AZvdsGUVl2UZmCw6lJRZ8OwrLXAUmbZ9n9Zj5XjsmSNpj4GurHHglTfbocsSzgkhhBAt6LcJISQn1XVF+NG/OYO712cxcM8DUZTBsgyMJgF1jS60d1egyGV+5Ps0t7nR0FyC8ZFFeOeCkGUVZosOja2lsDuNWV8fiyYhJmXo9DwMxtxrmQkhhBweFIgJyTOqqmJmYgUjAwuIhBNgWQbuChua292w2gz7PT0AgM1hxBPPN+L0U/UIBmJQ1dUa43QrutvF8SwaW0vR2FqqabyqqBgbXsTAXQ8WfeH1x8sqbWg9VobaI8U7Oj9CCCEHAwViQvJI0B/DJx8OIOiPb3p8wRtG7605HD1Wjp4nasEwuR12sVsEHYfiEst+TwMAoMgKPj8/hOmJlS3PeWaD8MwG0XS0FI89cyRv/vwIIYTkBwrEhOSJSCiB8x/0IhpJfRqcqgJ9d+ahqipOP1W/x7PLP7KkwOcJQUzKMBh5jAwspAzDGw33+2C26mkzHiGEkE0oEBOSJ25fm0kbhjfqv+tBc7sbjiITEnERY0OLCAbiYBjAVWpFXUMRWI39ggtRIi7h7o0ZjPQvrJ9SJ8sKVhYjMJp1MFt0ANKvAPffnkd7dzm1ayOEELKOAjEheSCZkDA+vKh5/MDdeTAMg+H+BciysuEZD659I6D7dDWa29w7P9F9Fo+JOP9+L/wrsU2Px6IiJElBKBBHMiHBWWRaPR0khURCwsTIkua6ZEIIIQcfBWJC8oDPE4IkKdkHAgBU3Lw8DZMl9VHKsaiIS5+PIRmX0HGicucmmQcuXhjaEoYBQJLk9f+fiEsIBROw2tNvQAykuAYhhJDDiwIxIXlA1hyGgWhERCScTBuI19y4PIWqOucj9QPORJEVTI0vY7DXi5WlKADAWWRCU5t7V8o2lhcj8MwGUz738FpwLJqExaanzXOEEEI0oUBMSB4wmXWax0YjSbBc9qCnqsBgrxdndmEDXjSSxMf/2r8ehNd45oLwzAVx97oRL/zBUZitmUN7Lkb6fWmf4wUOwIP6a0VREY+JMJpS/7nuRJ9kQgghB8fB3XlDSAFxuS2wO7L3GJYkGZIopw16D5scXXrUqW0hijJ+/9u+LWF4I/9KDBd+24fk/U1vOyEcSqR9zmgStqwGb66t3jy29kjRjs2LEEJI4aNATEgeYBgGrcfKs45TFRUsy2g+eS0Rzx5Io5EkZiZXMD2+jKA/e23t6IAP/uXs44L+OIb70q/q5irTqjjLsltW2Zk0nSaOnag80F04CCGE5I5KJgjJEy0dbix4QxgbSt9tghc4OIpMYFlttbE6ffrWYksLYdy5PouZiRUoirr+eFmlDR3HK1FZ40j5usF7Xk33BoDBXg/aust3pJbXXW7D5Ohy2uctNj0URUUsmgQACCm+9mMnKnG0M/sbD0IIIYcLBWJC8gTDMHjy+UYUuczovzOPSDi54TmgotqB7tPV+PLj4S0n2aVTk+ao4tmpFXx2bihlWYFnNgjvXBCnn6xH67GyTc8lE1LKLg9r1gJpPCpCVhQseBhcvDCEjuOVcJU+2ol2DS0luHF5CpKYuhSCYRjYnUYYjAIURYFez0NVAZZlUNtQjNYON0rLbY80B0IIIQcTBWJC8gjDMGjvrkBbZznmZgKIhBNgWQbuChusttUa45Z2N65+Panpeq0dW3sRR8IJfH4+dRheo6rAla/G4XSZ4N4QIjeuJD8smZDgX44+NEbF+NAiJkeX0dhagseeadC8uv0wnZ5Hd081rl3K/LUbjDyee60VFdUOSKIMQeDAbPOehBBCDgcKxITkIYZl0pYstHaUYXpiJW0LsjVdPVVwFm/tpjDU6027yrqRqq6e6rYxEOv1PHR6DsmEvGmsmJSxshSFqm4OzCzLrAfgkYEFsCyDx55pyHrvdNqPV0CSFdy+Og01RTbnOBZPPt+AqlongNUQTQghhGRDvy0IKTAsx+L511tx+YtxjA0tblm11ek5dPVUoa2rIuXrRwa0b3SbnlhBIi5Cb1jdxMewDBpaStB/x7NpXDgU3xKGAaxu/ttQPzzU50NbVwXsTqPmOTysq6cKR5pcGOz1YnpiGZIoQ28QUN/kQtPRUs0bDgkhhJA1FIgJKUA8z+GJ5xpx/EwNRgd8CAbiYFkGLrcF9U0u8HzqzXSKoiIaEVM+l258JJxcD8QA0HqsDMN9vvWT9WRJSdnNgmEYmCxb28MN3vPg9CP2RrbaDeh5vBY9j9c+0nUIIYQQgAIxIQXNZNbh2MkqzeMZZvV/qcoN0uEealFmsxvx9ItNuHhhCLKsIpmUt7xmbYNbqmDu84S035wQQgjZA9SMk5A8IIky5qb9mBxbgnc+mLL8YCcwDIPScqvm8SazAKt964Eh1fVFePV7HahtKNpYEQGGWS2TKHKZ0pYuZNqYRwghhOwHWiEmZB/FYyLuXJvB6ODCppVWi1WPlg432roqNHVlkCQZU2MriEYS4DgWZZV2OItNKce2tJfBO6dtlbbpqDvt/YtLLHjm5RbUN7pw4Td9AACeZ7MeemHZweOcCSGEkJ1AgZiQfRINJ/HR+/cQCm49kjgcSuD6pSl450N49pWWtKFUkRXcvDKN4T4fEg8dk+yusOLkY7UocW9eEa49UgR3uRXe+cyh2GrTb+lDnErNkSK4Si0IBrT1Rm5sLdU0jhBCCNkrVDJByD65eGEwZRjeaGZiBbeuTKV8TpEVfHpuEPduzm0JwwDgnQvh/Pt9mJ8JbHqc5Vg891oryirTH1Jhdxjw4httmjo2MAyDti5tp7/ZHAZU1zk1jSWEEEL2Cq0QE7LH5mcC6L8zj+H+1b68BqMAQZf+iOXBXi86e6q2bFC7e3MOs1P+jPeSZQUXzw/hnT89AV548HqdnsdLb7RhfiaAwV4vlnxhAKvdG1ra3aipL0pZ+pCIS5gaX0IsKoLnWVTWOGF3GtHSUYaVpSgGe9Mf62wy6/D8a610SAYhhJC8Q4GYkD3i84TwzWejCKzEEPTHEIuuHs0cCScg6Li0XRmSCRlTY8s40lyy/piiqBjq9WwZm0oiIWFseBHNbZtPrWMYBhXVDlRUpz4AZCMxKePaNxMYG1pcb7cGAFe/nkRZpQ2nn6zD2e8cgcttQf/teSwvRdfHCDoOR5pd6DxRlbINGyGEELLfKBATsgd880H8/rf9D3r3PnRsspiUsbwYQZHLnDIUR0KbSys8s4Gc+gmPDS1sCcRaiUkZ5z/oxdJCJOXzntkgzr3Xi5e/24bG1lI0tpZieTGCaCQJjmPhKrVkXAEnhBBC9hvVEBOyy1RVxVefjGxaWWWwtWxAkVWE/Kk3pj1cvhCLag/DABDPcfxG1y9Npg3Da8SkjM8/GoJ6v6VakcuMqlonyqvsFIYJIYTkPVohJmSXzU75t2yeE/Qc4vGtITWRkCBJ8pZVYnf55g1wvJDbe9mN9cPprCxFMXjPg9kpPyRJhtGkQ80RJ4b709cFbxQOJTA9sYKaI0U5zY2Q/aaqKsbngvAuR5AUZdRV2FHjtoJhqN6dkMOCAjEhu2xydGnLYyaTgHAwkfIAjnhMgsX6IMC6Ss1wuS2bxpRV2MHz7KZV50wqa9LXCauqimtfT6LvzvyWecxN+xEKxGF3Gjcd35zO+PAiBWJSMFRVxdd35nD+0iRmfGHE7ndrYRjA5TDi1cfr8MKpGug0vKEkhBQ2KpkgZJelPNqYZdMeUKFuOMmNZRmcOFu7ZYzewKOusVjT/RkGGeuHb3w7tSUMr1FkFYqiwr8cg5jc2trtYbFYUtOcCNlvqqri5x8P4Z/ODWB42r8ehlefAxZWYvincwP4jz+7iWiKT3MIIQcLBeJDIi4lMLgwiivTt3B97i6mA3P7PaVDQ69P/UGM2aqHxabHw5/Krh3CwQssnnm5GeVV9pSv7z5dDbOGrg3dp6phThO+o+Ek+m6nDsMA1lukqaqKUCBzz2QAKTcEEpKPvrw1i69uzcEfSv99rSgq7ows4L98NLCHMyOE7AcqmTjAZEXGPd8gLgx/geHlcSRlEQwYmHVGWPUWVNsr8HTtGXS4W/Z7qgdazZEiDPf7Uj5nsRpgNAqIRkQk4iJUFSirtKGlowyNraXQGzb/E11ZitzvGxwBoMJZbIIsK4jHtq7echyL7lNV6DhRmXZuQ31eKMrWso01ug1hPpmUIIlyxnpkOnSDFAJFWS2V8Iezv8mTJBX3Rhcx7Q2h+qFTHwkhBwcF4gMqKYv42Z0PcNvbj+XoCtYijwoV4WQU4WQUCSmBhcgSgokwHq85ua/zPcgqaxyw2Q1pjzbmeA5WOwer3YCKajte/MO2LWMkScbXn4xiIkU9MgBYrHoUlZiRTEjgOBZllTY0HS3NWvfrnQ9mfJ7nWegNPBLx1cAtJtMH4tV+wyUpnyMkn4zO+uFdimquwQ9FRVzt81AgJuQAo0B8QL3f/xEGFkc3heGHLccC4Fgevx/9AtX2clTbK/Z0jocFwzB48oVGXPhNHyQx/S9go0nAY985suVxVVVx8fwQZibTn0oXDiWg03F45XsdEHLYAJRqU9/DrHYDxGQEiqJmHH/6ybqsLdZW65GjkCUFJrMubSkHIbvJH0pA1BiGAUCSFSysxHZxRoSQ/UaB+ADyRZbQvzCCYCKcNgyv8ccDsOhMuDxzkwLxLipxW/HyG+345vNRrGw4xW2Nu8KKJ55rhMVm2PLc9MRKxjC8Zvl+27SO4+lLJB5mtugBhDKO4XkOTpcZgeUoOH7rtgNBx+H0k3VobC1Ne41EXLp/XLV304EiZZU2HO0sR009daYge4fn2C21+5kwDENHjhNywFEgPoCuz96BqqqIiluD18NEWUJMjKN/YQRJWYSOy95ai2yPy23BG3/cBd98ENMTK0gmZRgMPOoaXXAWm9K+bvCetiOaAWCo14v27grN/VMbW0sxNrSYdZwgcKiqc+KJZxswPrKEeEwEz3OorHXgSJMrY11xNJLEhQ96EUhx6IhnNgjPbBAdxytw8rGt3TQI2Q31FXboczgwxqjnUOky7+KMCCH7jQLxAeQNL0BRFSgaPg4HAFEWoagKYmKMAvEeKC23ofShgzbSUVUV8zMBzdcOBRMIBeKwOYyaxpdX2VFcYs56Eh0AHO0sR1VdEarqtK/mqqqKz84NpAzDG927OQebw4imo+lXmQnZKQ6rHl1NJVhYiSGZoYwJAMAANrMeZzrK92ZyhJB9QW3XDqjtnLDEs/T+KN+s1u3m9hpZ1l4bCQDPvtICqy1zLW99UzGOZehWkc78TACLvuxhGwB6b87mfH1Ctuvls3VwF6X/ZGaN3aLDqTY3XBrfZBJCChMF4gOo1OICy7DQ89l71AKAwAlwW0pg1mX/5UD2FsexafsYp8IwgNGk7e99jdmqx2tvH8PRzjLo9Js/RrY7jTj7dD2eeqFpW2+y0rWbSyXgj8M7l7nrBSE7xV1kwr/74+OoLrMAqb61GcBu1eGprkq89Uzjns+PELK3aEnwAOqp6MTV2duw6SxYkJYzjhU4HkbBgFOVXXs0O5KrhpaStCfJPayi2gGDMfeyF4NRwOkn63HibA0WPGFIogyjWQdXqSX7izMIpWk1l04wEIO7Qls5CSGPqtptxf/4N0/iq1uz+P2VKcwthqEoKkwGAV1NLjzbU42WGue23gwSQgoLBeIDqNTiQqurEf0LwwgnI4hJ6ZvPO/Q2VNnK0VW2tfctyQ8tHW4M3PNkPEBjTVvXo9U58jyX9mS87cg1R1DwIHtN4Fk821ONZ3uqAayWHLEsQ9+LhBwyVDJxQL3V9grqnTUotbhgElLXvjmNdnSVt+FHXW+BZ+nI3Xxlcxjx+LMNWcNl96kqVFQ79mZSGhXnuML8qCvShDwqjmMpDBNyCNEK8QGl4wT8uOst3PUO4ursLYyvTCOUDEOUJTiNdhwtacTzR55Epa1sv6dKNGhoKYHRJODWlWkseMObnnM4jeg4UYmGlvw7Ja6lvQyD97yaxrorrHBo2ORECCGE7DQKxAcYx3LoLm9Dd3kbYmIcoizCKBggUGu1glRR7UBFtQPLixEsLUSgqirsDmNe19w6i02ob3JhfDhzr2OWZdB9umaPZkUIIYRsRoH4kDAKBhiFraegkcJT5DKjqIAOCXjiuQbIsoKpsdQbPDmOxVMvNqIsj4M9IYSQg40CMSFkV3Eci2debsb8TAAD9zyYnwlAlhSYzDo0tJSguc0NszVzH2RCCCFkN1EgJoTsOoZh1ks+CCGEkHxDXSYIIYQQQsihRoGYEEIIIYQcalQyQcghsOgLY/CeB1NjyxBFGTodj9rGYrR2uOEsLpwNeoQQQshuoEBMyAF3/dIk7t2c2/RYIiFhqNeL4T4vjp+uxrGTVfs0O0IIIWT/UckEIQfY3RuzW8LwRqoK3Lg8jcFebYdnEEIIIQcRBWJCDihRlHH3xqymsbevTkORlV2eESGEEJKfKBATckCNDS5ATMqaxsaiIibHUx+cQQghhBx0FIgJOaAWfeHcxntzG08IIYQcFBSICTmgVDXX8Tm+gBBCCDkgKBATckDZHIacxtsdxl2aCSGEEJLfqO1aCovRZVyduY3+xREkpCSMvB5tpU3oqexCkZGOniWFobG1FLevzkBRsq/88jyL+ibXHsyKEEIIyT8HIhBHxRgGFkYRFaPQcTo0F9fDYbRv61pfTlzBp+Nfb3osKSdxafoGLk3fwCtNz+BM1fGdmDYhu8pk1uFIswsjAwtZxza3u6HT5/bjQFFVREUZLAMYeQ4Mw2x3qoQQQsi+KuhAHBfjOD/yBe75BiApD3bTnxv+DE3F9Xil8RkUmbSv6F6eubklDD/so+HPoed06C5v3/a8ye6Y8s+ib2EYMTEOo2DA0ZJG1DoO94ETZ56uRyScxPxMIO2YmvoinHysVvM1/XERN71+9C4EEZdW/93Z9AK6Su3octth4LlHnjchhBCylwo2EMfEOP7zjV9gIbqU8vnhpXHMBD348+PfR4m5OOv1RFnE5+OXNN37k7Gv0ek+CpbVXoIdE+MYW5lCUk7CJBjR4KwFzxXsH39e8YQX8H7fR/BGFjc9fnnmJkrNLrx59GWUW0sf+T6SJGNqbAWRcAIcx8BdYUNxieWRr7ubeJ7DC6+3YqjPi8F7XvhXYuvPFRWb0HKsDE1HSzWv7k74I/hgeB7iQz2LgwkRX04v4rYvgHdaK1Fk1O3o10EIIYTspoJNZOeGP0sbhtfExBh+1fsh/ub0T7Je755vEHEpoene4WQEA4sjaCttzjo2GA/ho5HP0b8wAlVV14OHkTfgeHkHnqk/C4ETNN2XbOUJL+A/3/gFknIy5fO+yCL+v5u/xJ8d//62Q7EiK7h1dfU0t2Ric1/fErcFJx+rhbvCtq1r7wWWY9F6rBytx8oRWIkhmZSg1/Ow5biJbimWSBmGNwomRLw7MIufdtZCx9GeXUIIIYWhIANxOBlBr29I01hfZBETKzOoc2b+6Hw6MJ/THKYD8xkDcVyM48upK/h130eIiXEAAMswMAtm2Ayrq4rfTF/DVGAWP+l+GzoKxdvyQf+FtGF4TVJO4oP+8/i3Gt4YPUxRVHx6bhCzU/6Uzy94w7jwmz48+2oLqmqdOV9/r9md2+8kcW3enzEMrwkkRPQtBtHtPtgbUOVEAv4btxC4ew9yLApWp4O1tQVFp3rAW/L7kwNCCCGbFeQSTq9vCIqq/ZjZu96BrGNyuV628f54EP/p+r/gV70frofh1deoCCXDmAt6EUlGAQAzwXmcH76Y073JqpnAPDxhn6ax3sgipvzajjHeqO/WXNowvEZRVHzx+2EkE1LO1y8UCVlB/2JI8/hb3vQ1ywdBcGAQg//Tf8Dcv/4OkclJxH0LiM7Mwvvxpxj4n/9XLH17eb+nSAghJAcFGYgjyVj2QRuEk5GsY5yG3LpSOIypPyKXFRn/dPvXmA95kZBSr1yqULEQWV4v0bjj7d8UnHeDoijwhBcwHZiDP1aYYSWUCGMmMI+5oAeiLKJ/YSSn1+c6XlVUDNzzaBorJmWMDmbv5lCoAvEkJEX7m8alWPLAHvQRHhnF1D//DHIidYmVKsuY++3vsHz12h7PjBBCyHYVZMmELsfNaFrKEbrL23Bx4luoyP5LnGM4dLnbUj7X6xvCUnQF4fsrwOmoUBGMh2Cw6CEpEnp9Q+ip7Mx671zFpQS+nb6BG3P3EEo+OJq3xl6JM1Xdmuqg99vEyjS+mb6O4aXx9cf0vB4cw0JSJPCstu8HrTXia7yeECLhzOUYG40NLeJoZ3lO9yCFZ/7Dj6DKctZxnvO/h6O7C6xA5VCEEJLvCjIQNxbX45OxzO3RNmoqrs86xm6wob20Gfd8g1nHdpe3w6RLXYt5fe4uAGxqA5dOVIxBUmTwLIdgIpx1fK7CyQj+4ea7KTcfTgVmMRWYxZnAHF5pembH771Trszcwrnhz7Y8npASWI75EU5GUWYp0fSmxyjkdnJbPKo9DANAPCbmNL6Q2A068CyreZW42KgryL7EqqIgNDiEyNgYFFGCYLfBcfw4dI7VT5AiExOIe72ariXHYgjcuQvnyRO7OWVCCCE7oCADcZmlBNW2CkwH57KONQlGTaugS9EVGAUDoskYgskwdKwAq94C4aHV6CPOmowB0ne/9ZeWMKBitd37wPqiAAAgAElEQVQbz3Lg2Z3v3frzu7/N2onj8sxNFJucOFXZteP3f1Rjy5Mpw/Aas2BCIB6CN7yASlsZWCZzBVBbaVNO9+eF3P5OeKEgK5A00XMsjrqsuOvTVm7T7d7ewTj7KdDbh/nfnYMY2Pw1+j75DLa2o6h487uIjI2neXVq4dExCsSEEFIACjIQA8Crzc/i7278DCuxICJiFIqqgAULk2CARW9eD0evNT+XMWyKsogPBi6sd60wCAYkFRGhRASBRAgWnRkukxNFJidOVXbidGU3OA3h1cjr1zfOaXGkqEbzWC0m/TOYCWrrnPHN1DX0VHTm3Yre11OZazD1vA56ToeEnEQkGYVVn35nf7mlFFW23MoZ3OU2CAILUdS2KloIXSYeRU+5AwNLoaydJux6AW2u/G1Dl4r/9h3M/PLdlHXPqqoi0NuHxNIyzEfqcrquIh7cTw0IIeQgKdhAHBVjkBUF/nhgU3lCTIpjJR5AmdWNH3W+ifYMq8OyIuNnd3+DsZWp9cdYhoHDYINdb0VSFqGoCipsbvzliR9oCsJlllJM+Kdh1pmwHAtk7EbBgIGOE7YV1rK5MXdP81h/PIixlSk0FGk/rWy3+WOBTX8v6bjMRZgP+RBKRtIGYj2vx5ttr+Q8B0HH4UhzCQZ7s39EzjCrxx/vFlVVMTftx+jgImLRJHiBQ0WVHY2tpRB0e3MyXLFRj+82lWfsRWzTC3i7tbKgehBL0Rhm3/sg6ybAuMcDJofDeABAsNsRHhtHYmF1w6WxvAymmp1980sIIeTRFWQgnliZwT/feR8qVFTZyhEV44gkI5BVBSzDwiQYoWP5rL/g7noH0oYuhmGg51dP25oP+TCwOJoxXK85WXEME/5psAyLIqMdi9GVtGNNggECJ+Dlpu9kvW6ulmOZW4WlGt+A/AnEyxo7Yeg4AeXWUvhjwZTPV1rL8MbRl1Cq4bTCVLpOVWF2yo9wKPOGvM6TVbDacqtR1sq/HMXn54cQWNncXWVmYgU3L0/h5ON1aNnFML5RncOMnx6rxS2vH/c2HN1s1wvodNvRVVp4RzevXL+heSU3sbgAhuM0baqTwmH4b9/B0qVvNz1uKHOj9LlnYW9PvTGXEELI3ivIQHxu+LP1lVeGYWDWGWFOscnto+HPcbSkMe3K7tXZ25rveXX2tqZAfLSkEaVmF3yRRVj1FqgAlqP+Ld0rGIZBqcWFPzn2XdQ6Mh8ash25lj9kq7/da7nMR8cJqHNW4cdd30OfbwhxKQEDr0dbafMjH9lsNOnwypvt+Pz8IBZ9W9v3cRyLrp5KHDu583+HABAMxHD+/V7E46s9jiVJRiwqQpFVgAF0Oh6XPh8FVBUtHWW7MoeHOQwCnqktwdM1LsREGQwDGHku70putAr29mkeqyRFmGqqEZ2azjhO9Pshx2IpD+iIe7yY/pefQ/qD11B89kzO8yWEELLzCi4QT/ln1zeuZRMRo+hbGMYxd+uW56LJGOZC2naLA6s1uUlZzNrNgGM5/LjrLfzj7V9jIbIEm94Cs2BEKBlBVIxBVVXwLI9Xmr6DV5qe3bUT6iqtZZgOZN90uKbCujcrjFq5LS7wLKepWwcAVNrKUG4tfeQAnIrZqsfr73TCOx/ESL8PkXASLMugrNKGpqOl0Bt2r63W9W+mEI9LUGQFAX8Mifjmwz9ikSRCQQZf/H4YdY0u6A1790+aZRiYdQX3I2QLOaatr7kqSZCjUbA6PTijEXI0ulor8/D14nGIoTAM5enfoKiqivl//RCm2hoYy6lVHyGE7LeC+202ujyZ8/hUgTip5L7ZRdQQiAHAqrfgvzr5J7g134urs7exGF2Gw2BDmaUE3WXtOFXVhSLj7h5r21PZiW9nbmgauxYm84lRMKC9tAW3PdlX7xQZMMfq8cEXo1AUFSVOI060lMK0w0HVXW6Du3zvNotFwglMTyxDUVQsL0UhianfHCiyiuXFCL78ZBgvvH50z+Z3UHDGzKUuiigiubR8Pzir8N+5A85ggBQOgwED3mpZD8acXgeG48AKAhguc+mIqqpY/vYKKt/67k59KYQQQrap4AKxmGOQlZTUx+kaeQMYMJoO4gBWD+Mw8HrN99VxAk5XdeN0VTcSUhKyKsPA6/esNKHY5ER3WTtueXozjmPA4Ln6x/dkTrl6svYU+hdGkJTT9wNenjUj6nNBNIQAPOjlfOHbSfS0ufH64/XgCmiD10Zz036oKhAOxdOG4Y16b8zhyecaYTDSQRC5sLW1ITo9k/I5JZlEfH4e6v3+ywzLgjMYwLAsBJtt9RMfqwXu558DZzbDWF2Fof/lf8sahtf4b99BxXf/MOfNeoQQQnZWwf0UztRaKxWLzpzycT2vQ7PriObrtJU2aeoyke5eJsG453W6f9DyPNpLW9I+zzIsvtf2Co7kUXeJjVymIvyo8820b0QWJq0IzpWgWO8CsPmja0lW8e1dD/75wuC+HyGsqiqiooyoKOc0F0lUoKoq4lFtbwIVRcFwn/YyILLK2XMCLJ96bSDh9a2HYQDgLZZN4ZVhGEjhCOJeL+ztbVBFadP4bBRRhBzP7QTF/SYrKu6NLuIXHw/hH8/1491PhzE0tbLv/84IIeRRFNwKcUdpCz4e/RKKxh++XWXpd3KfqTqOwcVRTdc5XdmtaVw+4VgO77S/hpMVx3B19hZGlychyhIsejO63EdxsrITDkN+94utcVTi3539C9ycv4cbc/ewEg+AZVjYUYHlgA1lFiMYpN/M1T++jOsDPvQc3fsa6UhSwi1vAHcWAogkVz+psOh4dJba0eW2wyxk/udnMAlIJiQoirbvdZZjMTW+vGsb/A4q3mRCxZtvYPbd9zaFOjkahSI9eDPCCjoIztS9pv03b8H90ovbOqaZzfJ98DA5kYD/1h1EJyehyjJ0TiecJ49DX1KS871zNTy9gnc/HUEwsvlTm+sDPhQ7DPjBiy2oLMlt0YIQQvJBwQViq96CthJtRyxX2yoy1sbWO6vxVO1pfDl5JeN1nqt/AlX2wt34Uu+sRr2zer+nsW1GwYDHa3rweE3P+mO//GQIc8KCptd/e29+zwOxJxzHu4OziD1U6hBOSvhmZgm3vH683VoJtzl9/Wp1rTOncg+DUUAyoW0TItnMebwbLM9j/sOPIAZXW/hJ4bWuIgw4kxF6lyttaYOcSCLY1w/n8W4YytyIe7St1Jvr63IK0YtffwPfJ59BTmxeVV748itYm5tQ9f13wJtSHyv/qEam/fiH3/VDTvMGbckfx//7wT381ZsdqHBRKCaEFJaCK5kAgNebn0Op2ZVxjE1vxdvtr2a91nNHnsDrzc/Dprduec5hsOG7rS/hqbrT254r2R33RjMfSb3R3EIEK8H4Ls5ms3BSShmGN4qKMt4dmEVETF3jDqweHV1TX6TpnhzHwmDkodMX3HvcvGE/1oGW//bfo+aHP0Dx2TPQuYogOJwwVlXC4HZnrQteC9BFp09pvmfxGe0/W3yfX8T8hx9tCcNrQkPDGP/bv0v7/KNQFBXvfjacNgyvSSRlvH9R26duhBCSTwryt6dBMODPT/wRPhn9Cne8/UjKDz7W5BgObaVNeLHhKc31xj2VnThZcQzDS+NYiK4GLbe5BA1FtQXbW/Ugk2UFoqS9ThMAYkkJe3Ww8k2vP2MYXhMVZdz2BvB4VfpDQ55+qQn9d+cRy1BHzLIMHEVGAAzqGrZ3AAlZxXAc7O1tsLe3Ibm8jNDQsObXsrrVlV7niePw37qdtVextakRNo2HcySWluD7+NOs4+IeLxYufomyl17QdF2tBiaXEQin39y60Yw3jLmFMCqodIIQUkAKMhADgIHX4/WW5/FCw1MYWhpDTIxDz+vQWFQHs86U8/UYhkGz6wiaoX2jHdkfHMdCEFiIovZQbNyjfrmKquKuT9spewBwxxfAY5VFad946Q0Czn7nCC5/OY54VNxST6zX87DaDeAFDjzPovFofrXPK2TW5ibNgZhhGFibmgAArCCg7qc/wfTPf5n29faOdlS98z3N3SWWL1/VvGlt5dp1uJ9/VnOnCy0GJpZzGt8/sUyBmBBSUAo2EK/R87qUfYbJwdZxpBg3B7XVEFeUmOHcpWOVH7bWTUKrcFJCTFJgEtKHlxNnarDgCWPBG0I8thqKGWY1DHP3j0lmGODx5xr29GCOg85x4jg8Fz6Gksy+MmppaoSu6MFnEJzBgLqf/gSxuTksX7mGhM8HgIGhvAxFp0/B4M7tjUtoIPueiTVSJILo1DTM9XU53SOTRDK32vR4juMJIWS/0W/Ph0wH5uANr56E57a4UG2v2OcZkVTOdpRrDsRnO/J9Q2TmlT9e4PDSG2349osxTIwsbVklttr0OPVEHao11hsTbTi9HuWvv4rZ9z7IPM5oQNmrL6d8zlhRgco333jkucjx3Grgcx2fjSnH3tZm6oVNCCkwFIjv6/MN4+LEt1uOhS41u/CdujNoK23ep5mRVKrdVnznRCUu3pjNOK6tvggnW/eujMAkcDAKnKYaYgAw63gY+ewfbQs6Dk+90ISTj9VifGgRkXACHM/CXWFDZY2Dat13SVHPSUBV4Tn3EeTE1pViXVERan74AxhKd/d7jDebIEUi2QduGL+TOhtduHzPo2ksw6yOJ4SQQkKBGMC30zdwfuRiyud8kUX8svd3eDkRxtnqE3s8M5LJy2frYNIL+OzGNOIPtRvjOQan2srw2uN1exoWWYZBZ6kdl2e11Vx2lthzmp/JrEP7cfrUYi8VneqBvfMY/DdvITQ0DCUpgreY4ejugrWleU++v+zHjiH+SfZNdQCgczpgrN7ZNov1FXaUFZvgWYpmHdtc40TRHpUoEULITjn0gXg26Ekbhjc6P3IRVfZyVNny/eP3w+Wp45U4e6wMt4cXMbsQhqoCLocBJ1pKYTLsz8e23W4HbnsDiEuZV4mNAocut32PZkUeBafXo/jsGRSfPbMv93f2nITv84tQ5eyfPBSdPr0rIf0HL7bg/3n/LqLx9K0CHVY93nqmccfvTQghu60g+xDvpMszNzWPvTJzaxdnQrZL4Dn0HHXju0834M3vNODJrsp9C8MAYNXxeLu1AoYMpRBGgcP3Wiph2aPuF6SwCTYrKt/6btaga2lsgOuJx3ZlDqVFJvz1W51oqNr6Jo5hgKP1Rfi33+uEzazblfsTQshuOtS/jSVZQp9Pe5/RPt8w3mh5ETx3qP/YiAblFiN+2lmLW14/7vgC6zXFJoHDsVI7jrsdBROGw9EkgpEkeI6Fy2EEy1K98n5wHu8Gp9fDc/4CEoubD6ZhBQHOnpMoe+WlHW239rASpxF/+UYHfCtR9I8vI56UYTby6DjigsOq37X7EkLIbiuM38i7JCbFIava2wPJqoyYFIeVo/6aJDurjsdT1S48UVW83orNJHBgC2QD3PD0Cr6+PYfhaT/WWuDaLDqcbnPjsWMVMNKpeHvO1nYU1qOtCI+MIjo5BVWRIdgdcHR2gDOmP7I57vFg5fpNJP1+MBwHc30dHN1d4PTbC7GlThNKnTu7cY8QQvbTof6NJrC5f/n8Nl5DDjeWYQpmNXjN59enceHy1JbHg+EkPr4yjVtDC/jLNzpgt9Cq4F5RVRVKPA5GEGBtaoS1KXutrhSOYPqXv0J4ZPNxyoG79+A9/3uUPv/crpVYEEJIISms39I7zCAYUG51Yz7k1TS+3OqGUaDd0+RguzuymDIMb7Toj+Mfz/Xjv36ni1q+7bK4z4flb6/Af+s25EQCAGCurUHRmdOwd7SnLZGQYzGM/+3fIe5L3a9bTiQw/+E5KKKI0mee3rX5k50hBnyIDl1DwjMOKBI4kw3GhuMw1naAoTI+Qh7Zof9XdKqyE78Z+L3msYTshKgoYXApjKgog2cZ1DvMKDXnx2rrxZszmsbNLUQwPO1Hc40z+2CyLSs3bmL2vQ+gKpuPKY9MTiEyOYXlq9dR+5Mfpix98H3+RdowvGncx5/A0Xls00l7JH+oqoLg1Q8RGbqy6XEptISEdxyhmx+j6NkfQiiidoyEPIpD32Wi030UNfbKrONq7JXodB/dgxmRgywuyfhwxIP/+8Y4Ppnw4dLsEr6cXsQ/3J3EP/dOwxPe2RPGcjXtDWFuQfsBEFf6tB3WQHIXGh7B7K/f3xKGN4qMj2P6Z7/Y8rgiili5fkPTfVRVxfKVq9ueJ9ldqcLwRnIsiKWP/x5ScCntGEJIdoc+EHMshx92vomm4vq0Y5qK6/HDzjfBsTu7e3s6MIcvJ67g8/FLuDF3F3EpsaPXJ/klIcn4Wd8M+haDUNStxzXPhWL4ef8M5kKxfZjdqkV/bvdeWNm/uR50vk8/g5ri++RhoaFhRGc2r+pHJ6cgx7T/3QQHBnOeH9l9YsCXMQyvUZIxhO58tgczIuTgOvQlEwCg53X4YeebmA/5cG32DnyRRahQ4TaXoKeyE+XWnT2WdXxlGhdGvoAn7Nv0+EfDn+N4eQdebHwa/A6Hb7L/Lk4tYjGa+U2PKCv4zfA8/s3x+n3pRpHrPTlqwbYr4h4PolPTWx5XVRVyNAopFIKSSAAqwAgCZn71azT8zV+vl07IicR6mNZS463E9/eTCZJadEj7yn18qhdy/FVwBvMuzoiQg4sC8Qbl1lL8YesLu3qPocUx/Pzeb6GoWz8GFRUJV2ZvwRdZwo+73trxFWmyf+KSjL7FkKax4aSEkZUwmousuzyrrSpLLWAYQMPC5Or4EmpBuBti81tLUVRZRtzrXQ3CGx9PJuC/dQfD//H/RNUffR+xmRksfH4R0YlJAACr14G3WsGbzWDY1B8Kcua9aaGmqOr9ExwZGHmWNmRmkfBMaB6rKjKSvkkYa9p2b0KEHGAUiPdQXErg3b5zKcPwRhP+aXw5eQXP1FM7pINiZCUMKUMt6MP6F0P7EohdDiOOVNoxOhPQNP50e9kuz+hgSC6vrJY1KAp0RU6Yampyer2qqinD8IYRSCwsou+/+x+gd5eC5XkwHAdVlqAkEkgmEpACQejL3GD5rT/27ceObeOr0i6YEHHLG8DdhQeH1Nj0AjpL7egqtcMo0Jv/lJT0x2SnJOc4PuUtk4iN3UZ8ZgBqMg5Wb4Shph3Gug4w3P6dAErIbqNAvIdue/qQlJOaxl6fu4una8+ATbOiQwrL2sEcWsWk3MbvpOdP1WBi7h5kJfMycWudE9XuvQ/thSQ6MwPfJ58hPDyyqR7YUFoK19NPwnm8O+Xr9CWuTf8tR6IZwjDAcDziHg9URYYUCEJXXATBZkNyZXl9jCImkfB4Yago37RSzPI8inpObvdLzGomGMV7Q3NISJvfEAYTIr6aXsQtrx/fP1qJYmN+dFnJJ5zJBimkfbMcZ3Y80v1iU70IXPoAiri5hCY+O4TgjfNwPvEO9BXZ+18TUogobe2hu94BzWPDyQgm/FtrCElh0nG5/VPT7eMbobpyG37wUgt4Lv3H2U3VDvzgxZY9nFXhCQ2PYPw//R1CQ8NbNsfFfT7M/OrX8FxI3fLRVFUFY0X5+n9LoWDmmzGrH5kDgBQOQ1UUCA77ltPrFDEJKfKgiwjDsqh8+y3wlt2pO/XHRfx6cGsY3iiclPCrgdn7pRRkI2PDcc1jeWsxhJLqbd8rPj2AlS9+sSUMr1ESUSx//l+Q8Ixt+x6E5DMKxHsomsxtR35EpB38B8URhzmneskG5/5ujGk/Uox//8MT+M6JSlhMqx+TsizQWO3Aj19txU9fb4OOPuZOSwpHMP0vP4MiZf4Ie+Hilwj09qV8rmTDYRlyPP3qMKc3bNoUpyoyVFEEABjcbgg2OxjmwY96Kbhay24oLUHtj38IR+fulUtc96wgKWcvFQolJNxb2Br6FUlBeHQJvo9HMf/bAXg+HIL/xhzEwOHYBGis7QBn1PYpjLn1sW3XZKuKjMCVfwWQ+VMhVZERuPq7bd2DkHxHJRN7SJdj/VWu4wudqqqYCsawEE1ABeAy6lBnNx2IjTc2vYAjDhNGV7L3+NVxLNpctj2YVWZOqwEvn63Dy2frICsqdZTIwfK165AT2sqjlr65BHv71o1Q9vZ2lL/6MjwfXUj7Wlanh1DkRHx+ftPj67GGYaArLoLO6YAUiUCRJDAMg7q/+CmsDQ1av5xtEWUFvSlCbjq3vQH0lD84HCSxEMHS15NQEptXjsOhBMLDS7A0FcN+vPxA/HxIh+F4FD37Iyx9/PdQHl5QURTI8TBUMQmhqByKGIccC2kO0BvFZwYgx7T9XUmBBSQ849CXpW9VSkghokC8h5pdDfBGFjWN1XE61Dm2//FXoelbDOLSzDJW4ptDhF0v4GxlEY6V2vdpZjvn2dpSzIWn1jcVpcQAL9aXQsixxGK3URjOjf/mTc1jIxOTSC6vpDwpzvXkEzBWVWHof/8/kFhYxFrUZQUBvM0GwWqFcn81+AFm68Y5lgVvXQ1KDMPAUr/7YSaUlDStDq9ZiSfX33gl/TEsfjEBNVOpxfASwDBwHC9PO+YgEIoq4HrlrxC6/Sni0/1QZRlyxA85EgA4DpzJBlUWEbr1McJ3PoPxyHHYT72W03HOibnRnOaUmB+hQEwOnPz6rXvA9VQcAwNtwaLT3Qo9r8s6LpgQMROMYT4cz6mLQT65MreMD0c8W8IwAAQSIs6PefH1TOGfwuQwCPiTtmq4TKk3Dxl4Dq83lOFoHqwOk0cj+rV16VgfH0w/3lxXi6q334K5rhamqiqYqmtgrKqCYLMBzGr43bhKyplMYLj05SyC05m2/dp+W1vZDt7xZgzDa8JDi5DC2lbiCxlvc8H51B+h9K3/BrrSGoDjwTvdEIorwZke/LxQFRnRkWtYvvgvULN0M9pIlXL7M1Slh9+EEVL4aIV4D9kMVjzf8AQ+Hv0q4ziHwZa15droShjX5v2YDkbXHzPwHDpKbDhV7oRZVxh/tfPhOL6Yzr5qfmlmCdVWI2rse9MvdbcUGXX4s85aTAej6FsMISpKEFgW9Q4zWoot4PM0qJDcMIIAZKkf3jQ+RSu0jZwnT2Dxy68hRaNbn2RZcGYzpHAYAAPBnvkNVdGpHs3zehRWHQ8dx2peJXYadOBZBlIkifi8tp7dABAZXYK962CvEq+RgotILkyBM2bu/52YG0Zs5CZMTdq6h2Qts1ABJR6BIq7Wsot+L5RkHKzOoOn6hBSCwkhNB8gTNafAszw+HfsaSXnru+xqWwXe6XgdZl364PfNzBK+SbFiGpdkXJtfweByCH90tApOQ/YV5v12w+PPto9j3U2vv+AD8ZpqmwnVtoPxteSjqCjBHxfBMAyKjDro97gExVxfh2Bfv6axvMkEQ1nmfs682Yzan/wQE3//T5BTnCon2O2QIhHoiovBGdKHFN5shrPnhKZ5PSqBY9FeYsNNj1/T+C73allUYiF7nf1GCV9u4/eLFFpBdPgqYhN3oMQjYHgd9BVNMLechq5EW1/q6GD2Y5zXRIauaA7ExiNdCPd/nfI5ORqCHF4BlLVSLwaJuWF43/0PMDf3wHr8RTB0iBQ5ACgQ74MzVcfRXdaO254+TPinISky7HorusvbUWnL/ItxcCmUMgxvFEpI+PXAHP68q3Zfjv/VSlZUDC1rXwkaWYkgISt7Hm5I4ZgPx3F1bhkjKxEo91udCRyLo8VW9JQ7UWTcmzeJxWfPaA7EzpMnUh6W8TBTTQ0a/uavsfDFlwjcubteO8wwDBydx+B+8QUsfvU1VDl1jTpvMqH2T38E3rR3b8ROljnQuxDMukps0fHoKLm/si1rfId8n5qlX3Y+iI7eRODbDzaVMajJGGITdxCbuANTYw/sZ/4w6wbB+Iz21p3iyjyksB+8JXtvYsFZBl1pLZK+yU2Py2H/ahjegDGYwHA8VDmJcP83kMJ+OJ/+4wO9uZEcDhSI94me1+F0VTdOV6VuzJ/O5bnl7IOwukFleDmMluL8PTghIctZD3/YSL1/7CsFYpJK/2IQ50a960F4jSgruOMLYGAphLdbK1FpNaa5ws6xNByBre1o1lCsczrgevIJzdfVu4pR9b03Uf7qy4j7FgBVga5o9RAOALB3tGHx60sI3utdb/nGGQ1wHu9G8ROPQ+d4tIMbcuUw6PBWSwXeG5xLG4otOh7vtFbCwK+uMnLm3LrrcKb87sYTnxmA/9L7yPRRWHTkGlhBB9vJV9KOUWURao4n16lpegqn4nziHSxe+FvIEf/91ya3hGHwAnhb8aaH4tN9iI3ehKlxbz55IGS3UCAuIN5IHL5I+n6kD7u7EMjrQCxso15WoG4HJAVvJJ4yDG+UlBW8NziHv+iqhUnY/R991X/8fcy++x78d+6mfN5QWoLan/x4W4dicEYjzLVbP2Y3VlSg+vtvQ/6D15D0+8EwDHTFxWCF/QuN1TYT/qyzFje9ftz1BdcP4LDoeHSW2tHttm/6+9C7LeBMAuSoto1b5vqt3Tm2S5UlKPEIwPHgDDvTCzx482NoqQuLDF6G+egT4Eypf2YznACGE6CmKLVLh9Vpf/PHme1wvfxXCN44j/hUL6TohjZsDAvWYAZndaYsj4gMXqZATAoeBeIC4o/ntrN3Jcfxe03gWFTZjJgJajuAxG02pA0yqqrSR3aH2LX5lYxheE1cknHHF8DZyuKsYx8Vy/Oo/uPvw/X0k1i+fBXR6WlAWV3RdfachLW1Zde+ZzmjEUbj7q+Ea2XTC/hOTQmeqnYhKspgGMDEcym/foZhYGkqRuC2J+t1eYsOhspH78oirngQGbyM2Pid9cDJ20pgau6BqeEEWGF7x0onvBOQAj5NY1c7RFyHtfOZtGMMNe2Ijd/SdD2dqxqcObd2lZzJCueT70COvwrPP//3YHgBYDmwemPGOmFxZR5SaAW8defenBCy1ygQF5Bcf3nmc/3wmm63Q3Mg7nY/+OG+Wn8cxm2fH3OhOBSosOsEHCu1o7PUticrgCQ/JCQZQ8thzePv+IJ7EvVWmNgAACAASURBVIjXGMvLUfnmG3t2v3zGMgwsGjrgWFpcEP1xRCfTb8hjDTyKn6p95DcV0bHbCHz7/vrR12uk4AKC184hOnwDxc//adqV20zExZncxi9lHm9uOa05EJtaTud0741YvQmMTg9Op/2NwOrBIRSISeGiYswCUmbW5/TDv9yc/y1xmossOOLI/tFktc20fnpbTJTxs75p/G5kHjPB2OrKoLras/ir6UX83e1JzARTtKciB1IwKeVUix5MiJpWk8n+YRgGzjNVcJwoB2/dvBGS4RiY6p0ofaEBgu3RfsYlvBPwX3pvSxjeSAp4sfzZP0HdzvdMjq9Rs/SS17mqYD32TNbrGOs6YazrzOneGzEMA0ZDH/yNtruKTki+OHDLaKqqYswfxXw4BkUFnAYBLcVW6A7ARqzV43/NGF3Rthp2vCz/T3djGQZvNJfj/JgX/UuhlKV2TUUWvNZQBo5loKgq3huaw3w4/WaRuCTj14Nz+FFHDYr3qKsA2T85l5Uz0Hg8DtlPq6UTLpgbi5FcikKOimB4FnqXCewO9VkP37sIaDjAQlyZR2J2EIaq1pyuz+VYQsBbi7KOsXY9B1ZvQujuRSiJBy3nVFUFZBmcyQrR78Xi7/4vCMUVMDWfgq64Mqd5AIChuk3zarTgcG/ZbEdIoTlQgbh/MYgvp5cQTGyunf1scgFdbjueqnYVRBlBJo9XFWEyEM16Kl1TkQXlluw1hLKiYngljJGVMERZgZHn0FJsRZ3dtGc1uTzL4vXGcjxWWYw7vgAWoqsbB4uNOnS57Sg2Plh5GFuJYC6UvcQiKSu4MreMVxsyt7Ejhc+u18EocJmPxN6gzGygevMCwjAM9K6d2eC2kRRaQWJ+TPP4yNDVnAOxofooWL15U3DNxNSk7dAUc+tZmJpPIT7ZC3F5HnIihvjUPahqAkoydr98ARD9HkRHb8BQ2QLHU98Hm8Oqb07lGU2nNF+XkHx1YALxLa8fH4+n3ryQlBVcnVuBPy7ijabygv5l6DYb8FZLBT4YSt/G6IjTjNc0BMExfwTnR72IiJtb+dxbCMJp0OEPmsrg3sOyiyKjDs/UlmQcc8unrck/AAwshf5/9u77ubHrzBv899yEi4sMEARzDp1zkFottZJlSZY1lu13xnHy66nZXdduze4v+ze8VfvT1u5szbwznuwwDvOObUmWLMlqSS21Ws1udSbZbOaMHC9u2h/QZBMECFwwgzifKlWJwOG9h00QeO65z3kePNPuXy7nRO1NHENwyO/Ep9Ph8oMBHK3f/XdOqK2X2+xmPqVBjZjbHLcSYVjY9j+O+LW3y461NHaD95i/gCcMC2vnEVha+rH4xt/k2i+vcbskM3UP4d/9G7zPfheEmLtbKtS1wL7/HBJ3Pioz7x7TDUAoajer/jwC5Gru/nZ0oey4oVACn89Ht2FGW6vdJeHPj3XgydY6eEQBLEMgsAy6PXZ8fV8zXutrAl8mRWQkksQv7k0XBMNLwpksfnR7cnm1dreYTZifj6YbWNxl86e2xskGD2wmNlL6JQv21+3eUoTU3mM/+CSkruMlx/CeRnjO//66jp+894mpShbyzH1kJsw39gAA58kX4Tj2PAhXmB9MGBZS9wl4n/4m7VRH7Ql7YoX42lzU9IaHgbkojga2tzj9VpB4DmebvTjbXD7nbDXdMPDmSOmarUBuZf3tB/P45sHW9U5z01W6GaoKmlhRm8AucPj6/mb89O4UEtniF3n1Ngu+2t8Mbh31r7eLEosjNTYGXVXBu1ywdXZU9R2t3Yxz1yOXTW7uTYL3BNZ1HkII3Odeg9DQieTdT6CEppafYyUXpN5TsO17bF2b0gzDQGroiunxqcHLsLYdqOgcjkNPwdZ/FumR61BC04Chg3XWQeo6vq7KGxS1W1V1QKzpBsZjKVyeDkFWdVg4BmyZD4/FlIxgWs7LS601Q6EEkmsEDatNxdNYSMnwS7vj38trFTBbYkNdHgK4xd3dxYraPH7Jgj852oE7izFcn4silMmCgKDBbsHRehf6vA6wu7Sxi7ywiLnfvoPY7Tt5rZcFrwe+xx+D7/HHdkVgrKXTiA8OQctkwIoiHP19YMXdX82mGM7ugaWpB/L0kKnxG82TlbqOQeo6BjUWhJ5JgHACOE/AdApDMVo8tNxZzgx59sG6arYzvAW2DZRxo6hqUJUBsaYb+GQ6hOvzUSSzKqbjGRjAcrF3l4UHV+KDL61owO6pWb/t7ofNbfB4ND6xawLiw36X6YC402WD00ID4lpiYRkcC7hxrIruAqWnp/Hg7/4BWrpws2g2FMbMr15HZnoGzV97bTmQMXQdsTt3kRx5AF1RILhdcB87BsG7NXVg1VQKc2++hcj1z6ErjzYtMzwP97GjCLzwBXBS9b2pOg49hezMfRhlKk3w3mZYmvs25Zyc0wdsUkWGSrrW5RbCDUDXALYqP/opaktV3V+FpufKbo1GHgV1DAE0I1fyMaloyGg66iXLmm1+90IJto1QylSoWC2r7Z68gwN1Dnw8HURcLrPCTYDTTbRIPLW76aqKsX/616LB8ErhgWuwNjfB9/hjiN68hZlfvw4lGssbM//Oe3Ds60fzV1/b1OBUTSQx8jd/C3kxWDh/RUHo0ytIPhhF13/9s3W1od5JQn073E98DZGPfrZmLWLe3QDvM9/ethV6LRmFGgsCBODcgZItpFmrA6XSPgxdh56OQ0vFAE0FGBaLv/nvsPWehrXzMAhLFwwoaknVBcQfTgbzgmEAsHIsEitKLi1tpmqwiwX1Rl0Wftesdu4Ua4VVF6z8+jdMLKRkDIcTyD4s6dbndWwojYFnGXytvxk/uTu1dtoHAb7QEUCbU1r3eShqO8Ru3oISi5UfCGDxo4/BWK2Y+vefFd0zYRgGYnfuIvu3f4eu7/3ZpqUyTP70Z0WD4ZXkxUVM/fwXaP/utzflnNvJ2nEYnCeA1L3LSI1ch6HmNuJyrgBsfachdR/PtTDeYtn5MSRufYDM1CCWAlzCsBBb98N+6KmiFSgY0QaxuQ+ZqXsFzxlZGUpkLrcivDTeaocSnEIkOIXk3Y/hffa7NA+Yoh6qqoBY0fSiVSLsApcXEAOAohuQVa2g5NaRgGtX5OLtpH0+h+lqG4QQ9HvtFZ9jISXj7QfzmFpVM/j9iUV0uW14oTNgqoVrMXWSBd891IYrM2HcXIgho+Z+9wwh6PbYcKrRg2ZH9d2+pWpP+LOreV8bmgZD00AYBoTL//uQFxYw8W8/Knh8tczcHObefgdNr7y84fnJi0EkhoZNjY3fG0Q2FN6ytI2txLvq4TrzCpynXoIup0EYFoxl+95DUiPXEL30i4LUDUPXkB67iczkPXie+gbE5t6C77Xte7wgIDZUBUpkFlh5N5CQhyvKOUpkFqF3/gl1L/8FrRJBUaiygHgkklwOflYSWAZOgUNs1YphUskPiBvtIk5UUW7hVmlzSfBZBQTT2bJjezyV5+HOJ2X88PZE8TrJRq65xr+mJvCtg63rDortAoen2/043+pDOKNA0w04LRwkE6W3KGq3WFod1pJJKLEYtMyj/HhGEMA7nOCcuSBGTSRAWBacvfwFamTgGhpeeB6MsLFOjZFr101X8DEMA5Fr11D/7DMbOudOIgwL1lr5AsBGKKHposHwSoamIPz+j1D/6vfB2vLraFsau+A49nxerWMtGSsMhl3+gpVuJTKLzNgtWDuPQEvFkRr+DPLsfRiqAtbqgLX7GMSWfTRgpmpCVUUPqwPeldwiD0JyY5bev9Wl/yFAr8eOF7sbytbnrRWv9DbiR7cni15gLHGJPJ7vrK/42L8cnlmzaciSmKzgrQdzeK2/8paiK3EMU/MpMFT1IiyH7GIQSrwwbULPZiEHF6GmkhADAajJFASX09RxtUwGifsjcO6vrLPaamo8XtF4JVbZ+LXoigw9HQdAwNrdezogS9z5uOymPgAwtCySg5fhPP6Fgucch54CZ/cgcesDKMFp6JnE8nNEEMHa3WCE4iveycFPoSZCSNz4XV4etYJcQw/W5ob36W9V1DSEoqpRVQXEa22SW+Ky8HAIHJKKhqymw2nhcKbJi8P1TnjEja2U7DV+yYJvHmzB2w8WMBFL5T1HCEGPx4bnO+tNNTtYaSyaQsjEyjMA3I8kEZUVuGglCKpGGbpeNBheSUunIS8uAoYB5mFesKFpUBMJGKoKEAJGFMFarXnpYOU26plB+Mr+NpkN3qFRwrNI3rmE9NjN5QoKjGiH1HMCtv7Htn31dqvpiozM+E3T41PDV4sGxEAuF9racRjJoc8Qeu9fAAAML4CUadecHruJ7MLYms9ryQiCb/096l78Xq5CBkXtUVUVELea2CTFEALHw9vwT7bV4WxT5Y0raoXPasEfHGhBMC3jfjgJWdMhcSz6vHY41hmk3gtWsEJk5Mafob8jqgpo6TSykSgIQyB4vWAqDBZX0xUFSjgMM80h1EQSYqAeIATy4iLUeCL/e6JREI6D4PWCs+WqErDWjefA2ru7ELz0sfnxPT3rPld67CYiH/60oNqDnkkgcfN9pO9fg/f5PwTvqvyu1W6lpxMwNHM14QFAl5PQFblkEw/eEzB94WBoGrR4qOzGOj2bRmzgLXgvfMP0XCmq2lRVQOyzCmh1SgUrmsWwDMFhv6vsOCoXGG9Wo5J0iRSMYjJqZSXgKGq7pWdmsPjBR4jdvAVdzQUvrFWE5/hx+J44B8G9vveZ2O07uY50TkfZShOEYSA2NyN++3ZenvFKhqpCnl8A6gxY/HWw93Sva14rOfb1Q3C7kY2Ub/4geD2w9xVu+jIjuzhZNBheSUvHEPrtP8H/6vfBlFn1rBrraMpRLn2EtXtACGMqDUNPx0FM1iSWJ+9CS8XASubSdiiq2lRVQAwAF9rq8MPbk1DL1NI91+KDtIFyYbXMMAyMRVMYDicha7mNift8jqKVGzKqhuFwbrOjZR352bVeE5ra3aI3b2Hix/+e1z0OALR0BosfXULk+nV0/NEfwtrcVPGx5YVFAADvzd0hyeXfFq4UE5aDJVAPLZWELpdLRzIgLy6i7sKTG17BBnLpU41f/hLG/+XfYJR4zyUMg6ZXvrTuCj7J2x+VDIaXaKko0iPXYevbWNe4rWToGuSZ+9ASERCWgxDoBOcoXnmDtbvASs5cnWATeE8jsgsT0LNpMLwIob6tIKBlRRssrfuRGb9V9ni6nAZjsuyaYeiQZx9A6jpqanzBuRQZ6ZHrUGOLAGHA+xphbTtoOiCnqK1Wda/EBruIr+9rwn8MzeQ6zq1CCMETLV6aKrFOU/E03hiZQ3hVHvDAbAT1Ngte6m6AX7Igo2p4b2wBd4OJvIuTjKohoWhwizw4Ex+O3Z7qKuRP1Y709HTRYHglNZnC6D/+M3r/1+9X3gzj4Z8HIQSCzwfO6YQai0PLpAHDAGE5cA472IcpEPL8AoS6OsgLCyiVYsEIPHjn5tWWde7rR+s3fh9TP/05NFkueJ61WND8tdfg6F9fJzctk0Rm8o7p8amhK7syIDYMHYmbF5EavAwtvTJ1jMDS1APn8S8UbEwjhIHUcwrxz98pd3BoySgMTUXw7b9ffpix2CD1nIT90JN5aRT2A09Anrxb9iKDsBwY0XxedkWd8ZanriM+8DaS9y7D0PI/V2JX3oDj8AXY9j1W8XEparNVXUAMAC1OCd871om7wThuL8aRVFTwDIMutw1H6p3rzn+tdVPxNH5yZ2rN1felcmpf7WvGmw/mim6es7AMQhkFc0kZAclSsoV2i9NKK0RUMUNTkB67BTU8CwDgXPWwdhzeliYG22Hxg49KBsNL1EQCkatXUXf+iYqOb23KX1VmeB6Cr/iFvJbOAAbA2W0ghEAOBovknhJwkgTBX4f43Xto+MLzFc2nFNfBA7D3dCNy7Tpit3JpG6wownnwINzHjoC1rP/vWIuHTK0OL1Fji+s+11YxDAPh93+MzMTtYs9Cnh7C4vwYfM/9IQR/W96ztv4zSI0MQEuE1zo4lPBsblOlLT9dQZeTSNx6H/L0EHxf+OPlShJCXQvcj7+GyKWfr/lvy4h2WLuaoISmTf+cK+sYm2EYBiIX/x3pNTYO6nIS0Su/hi6n4Dj6bEXHpqjNVpUBMZDrWHa43oXD9TRPeDMYhoE37s+VTUWRVR1/f2MM4hqpDoQQeEUeC6ksguksArbiH5QWjsHzHXtnc0ytSdz6AIlbF6Fn8ysZxD57A7YD52A/dKGqG+CoqTRit4oFN8WFPr1ScUDs6O8D73JBiZprksM5cit5rE2CZJOgpVIPA2UDhMvVJ15q2qEly++zqBRrscB39gx8Z89s7oErfp3svtdV8vaHawTDjxhqFqH3/g2B1/4q76KRsUjwPffHCL3zj1DjhR0B1VgQMAxw7gDIGjnHSngGkUu/gPfCN5cfs3YeAeusQ/LOR7mqHaoCQhgwoi1XtWPfY8jOjyH8wU9M/YyMaIelqbJNk+kH19cMhleK33gPlpZ+CL6NleGkqI2gCZwUgFy5tHCmfLk0VTewkJRL1hkWORZ+SYBmGEXH+awCvnGgFXV0dbgqRa+8jtjAbwqCYQDQlQzi199B9PIvd2Bmm0eJhJc30JmRDYZMN7BYQhgGgefMNbFwHzsCwubviWAlCYLPC6HOB97tzutgx1o3p23zduCcPhDW/F0F3ru76uEahoHkvU9MjdXlJNKjNwoe5xwe+L/8v8D9xNcg+NvBCBIYiw1CXRsYqx2ct7Hg979aZuIu1Hj+KrORTcNQszB07eF/Kjh3Pfi6FjBWB8S2A6Y3ydn6TldcD9rsvwsApO5drujYFLXZqnaFmNpcQ+FE+UEAUqoGA7lqEqU2xIkci0Ybg3a3DU6Bg6zpsHIs9tU50FamfJ5uGBiJJDGfzOUr+qwCejx2sGXqUFNbT54ZQfLupbLjUkOfQmzug9jSvw2z2nyEqXCtYJ2r4Z6TJ6Amkph76+01A2rPieNofPUVDP63/wtqMmnquI79+9c1n53ACFZYOw4jdf9q+cEApF2WP5ydewAtZW6VHwBSD65D6jlR8DhhWEidRyF1Ptq0lrj9IbKL4yaPbCA9MrCcehC98gaSdz/KHZsQED5XmSM7P4rs/Cis7YfhfuKr8Fz4FkJv/wC6Urx6CQBYmnphP/SUyXnkaMkolOCU6fHpsZtwn3utonNQ1GaiATEFAJDLdJZboum5D23dxGoYIQQSx+LFbvMrOjfmo/hoKoi4nL86J/EszjR5caqx+G5tanskB82v+CQHL1dtQCz4fGBFcc0SZ6tZmxrXnSLiv/Ak7L3dCH58GdHPb0BXFBCGgWNfP7xnTsPRm7tN7Tl1Agu/u1j2eIRh4D2zu4LGcuwHnkB67BYMtXDT3kq8OwBr+6FtmpU5+RvoytNNVpQAckFlJdRErjxe4vaHy8HwWtJjN8CINrhOvwzfi3+O+NW3kJkaxMoNm4zFBlvfadgPX6h4dbjYHaRSDE2BoWt7uishtbvRgJgCAIhlbsctWfrMZ0x++FfSKvvjqRA+mCi+YSal5KpaRGUFz9Hc4x1haAoyk/dMj5enh6FnM2CE6rl9v4ThebiPHzPdlMK7wbxaa1MTWr76FbR89Su5gJjjCgJs/4WnkBgcRnpmpuSxGl58Yd21kXcK5/LD+8y3EX7vX9dcqeRcAXif/e6uC5gqSfeodHylPythWRiaisSt8hdOQO5Ojv3wBfCuenif+Ta0ZBTy7AgMVQFrc8LS1Lvuf29SonlI0fEst+t+t1RtoTnENULRdKQUFapefGW332eu9I6Vy71hSZy5N64OV/nuggAwk8isGQyvNDAbwX2T6R3U5tKVLGCi2P8jRsWrRLtJ3flz4KTyr18xEIDr8OatWjI8X3S1mbVY0PGnfwzn/n1Fn+ckCc2/92XUPXFu0+aynSyBDvi//H04Dj+dV82A9zTCdfZV+F/63q5sCiHUt4Mw5teWLI3mG6YIgfaK55IZvw1dNrep0tA1pIcfpaqwNhek7uOw9Z+B2LJvQwEqZ/eA9zSaHi+2Hlj3uShqM9AV4j3ufjiBgbkoRqNJwMit7HZ7bDgecKNtRbDa6pRQJ1mwmCp9y9LCMvCIfNGVX8MwIGs64lkVGU0HQwjeGV3AXFLG0YALrhLl8AZmy3fCWh47F0G3x3ztTGr9DMNAdu4B1OgCDE2DocgVrfwwfPWtDi8R3G60/9F3MfaP/7xm7q4YCKDjj7+7KU0wzOAkK9q/8y3IwSAiVweQDUdAOA62jna4Dh/atnlsFVZywHH0WTiOPgtDVQCG2fWrhqxog9h2AOnRz02MJpD6zN9NsDT3g7W5oSXLvz8yFgnW9kOIXy9T03gVJTJX0fhK2PrOIPLJf5geS+0+iqJA0zQIggCm0r0VVYYGxHuU/rCM2u3FWMHjQ6EEhkIJnGr04Ol2//JzL3c34Ie3J0pWkJB4Fl/qacAbI3PL+cRALnAKprNIrWjF7LHySCoqLk+HcGUmjJe6G7C/rrCOpWEYuBcyn4c3Gk0hrWiw0k6EWyp1fwCJm+/nlYJSE2FA08DaPWDE0qunQn0HGEuFzSp2GamlGb3/2/cRvnIV4SufIRsKAQCszU3wnjkN15HDOxKEWnw+BDaxzvBuVE31rB1Hn4M8cx+6XHrTo/3AE2t2rSuGEALniRcQvvgTlGrGAgDOY1/YdV3frD3HkZ64DXl6qOQ4W/9jEOrbSo6hto+u65iYmMDw8DCCwdz7P8MwaG1tRU9PD3w+H8LhMIaGhjA7OwtN0yCKItrb29HV1QVRrM6FkN3110NtmvfHFwuC4dWuzIRhEzicfrhRrd5mwTcPtuI3I3OYSRTm8bU6JXyxKwC3yEPiOfxy+FG3wFBGWQ6GGQJ4RSEvrUI3DLx+fxYSz6J9VRpFVtPzguuyjFyVCxoQb5345+8i/vm7BY+zkhNqeA5qZB6s0we2RNtXW//eWPHhJAn+p87D/9T5XPtiQqq6xjK1+TiHB74X/gTh9/6taC1hEAb2A+fhOPZcxce2th+CoSqIfvI/ijfZIAxcJ1+E1HsyNxd3oLK5u7duTwYhDLwXvonop79GemSgYP6EE3L/Lkee3rI5UJVRVRUffvgh5uby7xzouo6xsTGMjY3B4XAgHs9fxFIUBTdv3sSdO3dw9uxZtLS0bOe0NwUNiKuEbhhYTGWh6DocAgdnifSDtKJhYM5cCsLl6RCOB9zLHeX8kgXfPtSG2UQGQ+EEspoOkWOxz2eHz/roVnm7S8JfHO/CYDCOa3MRzCVlWDkGVo6FxHMoViFNNwx8OBksCIg5hsnV2q8gJi5V8o3aGHlmpGgwDORuyzJWO/R0Alo8CIa3LJdzWsnadghi28Gtnuq2q7gcG1UzeFc9/K9+H/LkIFL3B6ClYiAsCyHQCVvvKbC29W90lLqPw9LUg9TQZ7lqHNk0iGCB2Loftt7Tece2th9A7Mrr0LPl84gJw0LqLiwBt5kIy8H92KtwHH0W6fsDUGMLAGHAe5tg7Tqa13Ka2nmXL18uCIZXisfjmJ+fh9PpLLoSrGkaLl26hKeeegqBQGUXZzuNBsS7XEbV8NlMBJ8vRJHMPipF1uK04kSDB33ewlzamwtR0yuuaUXDYCiOA3X5m1Ua7CIa7KVve3AMwQG/E7PJDAJFVpSLmY6nMZ/MoN726NgsQ9DpsuFBxFyN1YBNhF2gL92tUq7OMOusAxgWeioGLR0Hx/uWnyMMC6n3NJwnv7gnV1GTaQVpWQHHMnA7qvO2ILV1CGEgtu6D2Lpv04/NWh1wHHm67GoqYXnYDz6B2MBbZY8p9ZwEa92e/Ris1Q77oSe35VzU+sRiMUxOTq75vKZpSKdzG6WTyeSaqRGGYeDGjRs0IKY2TyKr4sd3JhFKF3aQm4ylMRlLF+QBA8B8qnzHuZUWimykG4+lcHM+hqisgCEELQ4rjtQ74SiyMj2bLL0Rr9j4lQExABwLuEwHxMcC1VVSqprochqZMvl+hBBwDi8Mmwt6NgNr51EABLy7Htbu42BF2/ZMdhvdGF7Arz8axd2xMNKyChiAzcrh9IEAXrvQg4Bv7/3MVPWyH3wSWiqO5L21ywaKbQfhPPXSNs6K2u3u3r0LWZZhGAZYlgW/an/EUjAM5ILjbDYLQSi8QwgAoVAI4XAYHk/19A6gAfEu9ovB6aLB8EpXZsLwWgUcqV9/kLiyx0ZcVvCLwRnMJfNXfCdiKXw8HcLJBjcutNXlrf6ZadKxUrHxXW4b9tU5cHex9Oa6dpdUsJq9lxmGAXnqHpL3LiM7PwZD18DaXLB2HQNnc0PPJAAQ8J4GCI3dG16V1eWU6dJqhGHBija4zryyZ297GoaBH789iDc/GUMqnd8sJplW8d5nU7hyZx7f+8ohnDlYWGIqkVZw5fYsxmbj0HQDXqcFJ/cF0BpYO/eaojaD6/TLsDT3IXXvk7yGG0KgE7a+0xDbDu7JuzhU5ZLJJG7duoXr169Dlh8tcLEsC0mSIIoiCCFQFCXv+1RVXTMgBoDFxUUaEFMbNx5NYdZkGsKn02Ec9juX39zqrGu/QIupk3LBTErR8MM7k4hmlKLjDMPAlZkwVN3A852PNmJ4RMH0XIHchrvVCCF4ubsBIsvi8/loYdBMgH0+B77YGaiZFs66IiP8ux9Cnr2f93h2fhzpB58Dhg7OVb9cyYFz+GA//DSkrqPFDmfKenap77ad7ZvpN5+M463L4wXB8EqJlIK/+89bcEgW7O/0Asj9rbx9eRzvD0wVpC9dvjWHjiYnvvVCP+xSZX+rFFUJsakHYlMPdEWGkc2A8JaqbJRDbZ1YLIZ3330XsixD1/MXQzRNQzweh6IocDgqv4hffbzdju4Q2aU+nzffsjOcyWIi/uhWxqF6p+lOciLHov9hHvIn06E1g+GVrs1FML8iTeJovfkVW7fIo9VZvBQXQwie76zH94534lyLa3vcEwAAIABJREFUDz1eO7o9dpxp8uLPj3bilZ7GijrfVbvwxR8XBMNaPAwttghoKqDrUCPz0JXc70KNBxH56KdI3P5w3edkbS5wLn/5gQ9ZAp27vk7semWyKt67OoFEqvzfRDSRxZufjC5//fqlUbz72eSaufyj0zH8zX/cRMrE3xtFbRTDW8DaXDQYpvLouo6LFy8urwpzXPHFjUwmg3Q6DXZVR1tCCGRZLhpMA4DdXl39Avbu0k6VC1f4QRlOK2h7GJfaeA5H6l24ZqLSxKlGN3iWgaLpuLVQukzbStfmInihK5cw3+KU0OSwYjpevivZ6UZv2dt0doHDuRZfyTF7nTw3WlC7U5fThQX6DR1aMgpmRemk2NXfQAh0QPA1PxqmKUiP3kRq6AqUyDwAgPc2wNZ7GmL7wbyg1tZ7BtErvzI1T2mPlFYr5trgAkJRc3c+DAMYGo9gYi4OnmPwwbXpst+zEE7j/YEpvPh4xwZnSlEUVbnp6WkkVzQdslqteXnCK6XTadjtdmQyGei6DsMwEIvF8j7PLRYLbDYbOI6DKIpobDTfqXA3qJ3ltipTaVbA6vHPtPvRW6QCxUpHAy6cbcrd4p1NZpBRi9S4XMODaP4GuN/ra4SvTKrGqUYPjtINcaakBi8XPKanil+wGHIKhrbylr6B1L1H36/Gw1j45f+DyKWfI7s4AUOVYagysvNjCH/471j41f8LLfnojoTUcwJCXWvZOVqaeiG27jf/Q1WZuVAKsmL+byKraphaSODjm7Omv+ezu3NQSzTCoSiK2ioPHjzI+5rjuDVzgjVNWw5+VVUFKVKPXZZlhMNhZLNZ9Pb2Vl1nu+qabQ0pV/Ks3HiWIXi1txFf6mlEy8oUBQJ0uCV8pb8JX+gMLL+glUoaYwBQtPzxNp7Dtw624lyLr6AkWqtTwu/1NRVUw6DWlg1O5T+g69DlNVbgDQOGmr/5Mj12A4aR+57gb39QvFnAQ2p0HsF3/nE59YJwPLzPfgeWxu41v8fafgjep74BQvbuWwgBKqqNDTzsujgWMj0+mVYxOZ+o7CQURVGbIFmkJb3T6SyoLrEkk8nAMAzwPL9msGsYBhRFQVdX16bOdTvQlIld6ljAhYFZc801WpxW+KXCXf6EEOyvc2B/nQMZVYOs6RBZBhauMOfTXmHXt2J1gC0ci3MtPjzW7EUwnYWqG7DxbMkmIqstpGTcmI8illXBPiz3dsDvhKWGcocBAKvysXIdnkpEZ6ueMjQVhiIjNXQFWiJc9nRqdAHp+9dg23cWAMAIVvie+yNkg1NIDV2BGl0ACAHvaYTUdwq8a+u6W+0WjXU28DyDTNbcKrGFZ+H3SMgqla34ZitYhaYoitosq3OCgVyLZrfbvZw3rKpq3niPxwOO45BKpZBOp2Gs2ADPcRysVitEUcTo6Cj6+/u35efYLDQg3qV8VgsO1DnLtl9mCDGVbytyLMQigfCSepuIOsmCxSI1iYvZX7f2jlOGkKIBeikpRcUvh2cxHs3vrnQvGMf7E4s43+LDycbqKd+yUZzDCy21YmNlmZVYUrAZggAsj+TQp6bPmRy6vBwQLxF8zXm5yLXkaK8fdS4r4sny+fyEAdoaHOhudsEu8blaxSbZreYvGCmKojaL3+9HOFy4YEIIgdVqhdVqXc4XXrIURNvtdthsNqiqCsMwwDBM3qa8Bw8eVF1AXGPLbtXli12Bop3olrAMwSu9DWhzSmuOqcTxgNvUOJ5lcNi/ebnAGVXDD29PFgTDSxRNx7tjC/h4yvyt6Gon9Z7K+5qwLMgatX6JIIJw+XlflsYuGMU24ZWgRheW0yYoQOBZvPBYO6yW8ndPXDYLnjnZCkIIjvaaTw3ye6xo8lfXTmyKovaG7u610+KWMAwDlmXh9XqLVpngeR6CIBRUqEilyrcO321oQLyLsQzBl3sb8V/2t6DXawfPMiCEwGHhcLbZiz872oE+7+YV+D9S7yzb9IIhBK/0NECqMMWilEtTobINSADgg8lFRGqkTJXYdgCcI3/ln7EW/12zUuHFia3vjOkGG3l0evt+pWdOtuKVJ7tgEdZ+vTttAr5yoRunDzQAAM4cCIBjze2KfexQde3Cpihq73A4HOjp6Sk7juM49PX1VXTsamz6QlMmdjlCCNpdEtpdm7MKXO5cL3UH4JMEXJ2NIJnNv+3b5LDiqVYfWjZpRRpAZeXeDOD6fAQX2vb+5jzCsPA+8x0E3/7BcuoEa3VAzyRhZB9trmMdXjBi/u9DbN0PS8s+QFdBOKFgw91aGMEKQuuUFvjaM7041OXDz94bxu0HISiKDhDAJnI41O3Dy+c60d/uXR5vlwT8l+f68MO37qFUE8fD3T48dqhhG34CiqKo4o4fPw7DMHD//v2izwuCgPPnz8PtdmNgYCAvp7gUv7/6PqdpQEzlIYTgbJMXpxs9GIkkEZMVMISg2VF8495GTcXTFZV7ux9O1kRADACc04e6l/4CyTsfIXX/KnQ5Bd4TgBpdyPWal5zLXepyCKTu43Cd/XLu6pzlYe08gtTQFVPnk7qP7+mqERvR3+7F//lHZ6CqGubCaRACOG2WNfN/D/fUQbSwePPjMUwv5O/klkQOjx1qxLOnWqtyFYWiqL2DEIKTJ0+ip6cHw8PDmJubg6ZpkCQJHR0daG9vX06H6OjowPDwsKnjmll53m1oQEwVxRCCHs/W5zZmKyz3lq2xmq2s1Q7niRfgOPos1Og8DE0Da3PDUDJIDn0GNTr/sPpDA6TeU+Ds+RsPbfseQ/r+NRh66at6wgqQ+s6WHEMBHMei2WTOb2+rB72tHozNxjA+G4emG/A6Ldjf4QPP0QsPiqJ2D5fLhZMnT5Ycs2/fPkxOTiKTKd2wqKGhAYFAYDOnty1oQEztKKnCwGAzc5erCWE58N6mFY844Dr1Ytnv4131cJ//OiIf/ORh6bZix+bheer3wTlqp4rHdmpvcKK9wXx7c4qiqN1IkiRcuHAB77///pod7RoaGnDu3LmqvPtFA2JqRzU5rHBYOMRNlqnq923eJsJaYW07APaLf47krQ+RmbyzHBgThoXYdhD2g+fBe2guK1UblLgMLZEFCMC7rWBF+jG4mpZSkLgfQmYqBl3RwPAsrK1O2Lq8YGmZwJrmcrnw8ssvY2JiAiMjI0gkEiCEwOfzoaenB/X11Vujnr4T1LBwJovrc1FMxNLQDANOgcOheid6PHYw23R1xxCCo/VufDCxWHYsy5BNLfdWSwRfM4Snfh9aOgE1tghCCDhnHRjRttNTo6htkZ6OIXF3AfLCinJQDIHU4oTjQD14F91QCgCJ4SAiV6fzmv1oUKBEM4jdXoDnZBNsXd61D0DteSzLoqOjAx0dHTs9lU1FA+IaZBgG3hlbwMBcJO9NbzElYySShEvk8dX+Jvism7+JrpjTjR5MxFIYW6MOMZBL/H+5e3PLvdUi1moHa6V1b6m9I52MIzQ3BV3XITlc8AUKG8nE7y0iem2m8Jt1A6nxKNLTcdQ91QGLv7YvEJOjYUQ+m157gG4g/OkUCMdAajNXt56iqgUNiGvQb0cXcG1u7YYN0YyCH92exHcOtVXUdnm9WIbgtf5m/G58ATcXYlBWbZzzWgU83e5Hl7u2P6w2k2EYSEbHkIpOQtcVcLwEh68PFivNI6aqQyy0gDuffYCZsWEYK2pu211e9Bw+jc79xwAAmdl48WB4BUPVEfxgDIGX+8BaavNj0dANRK/PmhobvT4La6urKvNEKWottfmXX8MWU3LJYHhJStHw0WQQL3ZvT24pxxA811GP8y0+3A0mEMsqYAlBi9O6aZ34qJxkZBxz4xehyPG8x4MzA5CczWjsfBacQC8+qJ2lqzr0jArCkoK81eDcFD56/cdQlcIa24loCNc+eBOx0DyOPvECEvfKp2MBgJ7VkHoQhmNfbZR1XC09EYWeMbeXQ0spyEzFYG2hKWzU3kED4hpzbS5qeuzdYAJPt2sQue1LU7BwLI4G6JvsVklERjE9/GZeb/qVUrEpjN/9Bdr2vwaOpxci1PbLBlNIDC4iNRkDHpZl5BwW2Hq8sHV5oRsqPv7NT4sGwyuN3B6AyxkAN2u+kk1ypHYDYnlVvWwz42lATO0ltBhmjRmLme8vruo6puLFS6tQ1UfXVcw+eHfNYHiJIsexMHFpm2ZFUY8khoOYf/s+UuPR5WAYANS4jOjADBZ+ex9jt28gmzH3vjRx80ZF59dS5ro6bgXDMKDLaeiKvDPnr7AmfKXjKWq3oyvENUar8E1MpW96e0Y8OAxNNfdhGw/dh7/1HDjeWn4wRW2C9HSs9IYuAEokg9i4uTxXAIhFQ1DtWXC8YO4bdiAnVktGkRy8jNTwVehybpWWc9ZB6j0NqecEGH57NjdzUmX7RTjJ5L8pRVUJukJcY+xCZddAlY6ndq9EZMT0WMPQkYyMbeFsKCpf/PaCqXFMmoWomctxV5gMdGK+NbylbnvThOTZEcz/5/+NxK2Ly8EwAKixRcQ+ex2Lv/5rqInwtsxF6qxgQy0BpA5aZYLaW2hAXGMO1plvbOGxCmiy09qce4XZ1eElurYzt26p2qNEMsgGzaVzEQAOxVwdXIMY4BvN3+Ww9fhMj90oNbqA0Hv/CqPE36UaDyL0zj/BUJUtnw9nE2BtMddRUWpz0wYd1J5DA+Iac6DOCavJWr4nG9y0rM4ewnKVXdww7PbcqqUoJZoxPZa3iBB0c69NluXgP9kJxkQ3Oku9DWLT9nXCTNz+EIZaPmdZjS0iPVpZLvR6eU63gPeUfp8QvFa4TzaVHENR1YgGxDWGZxl8pa8JAlv6V3+gzomj9XQH8V5i93SaHksYBnZ3+xbOhqJWqOC622ozH7S29ByA6LLD/3Qn2BI5spaAHb7z7du2AKArckVBbnLw0y2czSOMwML/TBcc+/1gLPkLJ4zIwXHAj7pnusDQBknUHkQTRGtQs8OKbx5sxQcTi7gfSeZ1q3NaeJxscOMEXR3ecxzeHixMXDKVOuHw9IClG+qobVJJ22SOF2D1O4EyVcIsVgn7jp9bPn7Dy31IjUeRHAlBTWRBCCB4Jdh6fLAEbNv6fqfGgjA082kQasT8RsKNYngWriMNcB6sh7yQgq5oYAQWFr8NhKGfCdTeRQPiGuWXLHitvxkxWcFkPA1NN+C08GhzWmkgvEcxDIfGzucwNfxGXmev1QTRCX/r49s4M6rW8S4RFr8EecFcHnHPs2eh3FMwNlh8lVWyu/D4i1+H5Hh0l4uwDGydHtgq2Ty2ZXZ/9R7CMhAbaJt3qnbQgLjGOS08DmxDe2Zqd7C529Dc9zLmxy4im8lv0kIASK42NHQ+Tcut1YBsMIXEcBCZ6XhuFdDCQWpzwdbjA+/Y/vxxx8EA5N89KBsrWuptsDY4caLhZfQcPo2R2wMIzk5A1zRIDhfa+4+gqaMPDLt7b+tzdi8Iy8HQzHWG45y12SyEorYTDYgpqsbYnC3oOPQNpGKTSEYnYOgKON4Gh68Xgkjzxvc6wzAQHZhBYiiY97ieUZEYDCIxGIT7ZBPs21hxAQDEgB2eMy0IfzqV15RjJcEnwffEo9x2p9ePY+df2K4pbhrGYoXYdgjpB9dMjZf6Tm/xjCiKogExRdUgQghsrlbYXK07PRVqm8VuzhUEw6tFPpsGI7CQ2ra31qytwwPBY0ViOIjUaASGmkvtEbxW2Hp8kNpcIGU2BFcL+4FzyIzfLLtKzNrcsHYeWf5aV2TIk/egyUkwnABLUy9YyVy5tGI0WYWh6GAEBgytO0/VMPrqpyiKqhGarCJxd9HU2Ojnc7C2urZ9TwHvEuE52Qz3iSboWQ2EZcBweyMIXon3NMDz5B8gfPHHa26wYyUXfM/9IRjeAl2RER94C6mRa3nl2ghhYGnph/PEi+Ac5vKjDcNAeiKKxFAQ2cVHeduWgB32Xh+szesPsCmqWtGAmKIoqkakHoRhmGzHriWzkGcTEBu3rzbvSoQQsJa9/REltvTD/6W/RPLux0iNXF9u0sFKTkg9p2DrOw1GtEFXZATf+gGU0FTBMQxDR2biDrILE6h74c/AOUunuhi6gdClcaQnYwXPyXMJyHMJ2Lo8cJ9qphusqZqyt99tKIqiqGVmu8Etjw+ldiwgrhWcsw6uM6/AefKL0NJJEELASA4Q8mhVPHbl9aLB8Ep6JoHw+z+E/5X/ueS4yMB00WB4peRIGKzEw3kwYP4Hoagqt/fuQ1EURVHFVVjtq0R1PmqTEZYHZ3eDtbnygmFdTiE9+rmpYyiROcizD9Z8XsuoSI6ETR0rcW8RukpfAFTtoAExRVFUjeAcwpaOpzZfeuyW6fJsAJB+cH3N51IPwmtW8FhNV3SkxyOmz0tR1Y4GxBRFUTVC6vKaHsvwDKwttAzfTtPT8YrGayXGK9FMRcdSYuW7WlLUXkEDYoqiqBrBOyymKwjYeuv2ZHWHakO4ylbpS46ne+Qoak303Y6iKKqGeM62QPBJJcdIbS44D9Vv04yoUixNPeUH6YCWVqCE0khNiph9fRCRazNQE9m8YbxLrOjcvLuy8RRVzWhATFEUVUYsmcX4bAwTc3FkFW2np7MhDM+i7ulOuI42gLXlrybyHhGe083wPNZKS27tErynAULd2g109KyGbDAJNZ6FrnEwmDaoMRmJe4uY/dU9RG/MLo+VOj0grLnfKyOwkFppygxVO2jZNYqiqDU8mI7i4rUp3BsLw3i4F8kisDje78dTx1rgdlh2doLrxHAMHPv8sPfXQY3K0FUNjIUDX6U/z16kJrPIzMRhqDr4xvPIBn8CGPmb63RFhxJ5lBfMus6AkPyP9fjtBRBC4DwUAGvhYO/1IW6iOYtjv3/PdAWkKDNoQExRFFXEJzdn8D8ujiwHwkvkrIaPb8zixvAi/uTLB9FUZ9+ZCW4CQgi9Lb7LqHEZkWszyEznb44zcB5IfwDWoi7f29USDze9EQ6s6ywYW/H0itidBdh6fGBFDs4jDdAyKlKja1eQsPfXwbHPvyk/D0VVCxoQUxRFrTIyFS0aDK+UTKv4h1/dxl996yQsPLt9k6P2LCWawcK7I9DlwrQcwvlhOF4F2BlYvEFoiTiy0RRYVwuI1APClNhMpxtIjoTgPFAPQgi8Z1shtbmRGAoiM/Mw8CaAtdkJe68PlvrqvcijqPWiATFFUdQq7w9MlgyGl8STCgbuzeOxQ41bPylqzwtdGi8aDC8hhIWut4DYD0Fqk5DNTJs+tjyfBA48+lpsdEBsdMDQdOiqDoZnQRiaN07VLhoQUxS1rRYjaVwfWkAyrUDgWezv8KK9MVcKLJVRMDmfgKYZ8DgtaPDZNnw+wzAwPBnBJzdnMTEfh64b8LmsOH0ggCM9fvCrSotFEzKGJsw3JPj09hwNiKkNy8wloETN1f1Nj0cLNkSWpRfvOkdYBuwezxXWNQWx4CAS4QfQNBksJ8Lh6YLD1wuGoWEQlUNfCRRFbYtoQsZP3x3G/clI3urr+wNTcDsE+FxWjM3EoGqPnmyut+P8kSYc7VtfPmMmq+Jf3riL+5PRvMeT6TjGZ+N4+/I4/uSVg6j3PipDFoplTK0OLwlG0+uaG0WtlBozfxFm6AbURGVNM1ipNrsOxkMjmB19F7qm5D2ejE5gYfJjNHQ+C7u7fYdmR+0mNCCmKGrLRRMy/vpnnyO6qi4qkNukdm1wEYZhoLHOBlF49LY0NZ/Aj94exHQwiZce76jonIZh4J9fv4uRqeiaY6KJLP77f97E//S1o3DZcxUWmArLjTH0NjO1CfSM+fbMAMBJAhiBhZ41VwZQ6vSsZ1pFaRkVyQchpMci0DIqCMtAbHTA3uPbVZs0E+EHmLn/G6x1faupMqaH30Bz78uwudYubUfVhr19n4SiqF3hPy+OFA2GdcPAbCgJXTdgGMB8OAUU+fi6ODCF64MLFZ3z7mi4ZDC8JJ5U8MG1qeWvA16pII2ilJZ6R0Xz2isycwmEP5tC6NI4wp9NITOX2OkpVTVSYVdARmBh6/GZGsu7RIiBzdkol56KYfaXdxH7fA5KVIYua9BSCpL3Q5h7cwiRgWkYldxi2SKGoWN+/IM1g+FH44zcuF0wZ2pn0YCYoqgtFYnLuDMaKvpcMqVAW5EioaoGkmuslF28PlX08bV8cmvG9NjP7s1DUXMrbaKFw9Fe8ykajx1qqGhe1S4bSmP29UEsvvcAyeEQUuNRJIdDWHzvAWZ/PYhsMLXTU6xKYkNlAavY4IDzYD3EMq24WSsP3/nNSQnIzCUQ/GgchrZ28JgYDCJ2Y25TzrcRicgolGzS1NhsJopUvLL3F2rvoQExRVFb6taD4Jo5ubFU4apxMq0UGQlMLyQxs2juAw4ARmdipsdmZA1zoUeB3IUTzbBaymeUtTc4sK/da/o81S4bTmPh3RGoseL5q2pcxsK7D2hQvA7WNjcYwVz5PotfAu8WQRgC3xNtua6DEp83hrAMbF0e1H+hG5x9c/KHY5/PAnr5ldT4vUVoa/wdb5dUbLKy8dGJLZoJVS1oDjFFUVsqnVn7g1HVCne+6yU+cCNxGY115ipPaCVWsYqOX3Fen8uKP37lAP7hV7eRWmPFur3Bge++vL+mcojDn07CUItXK1hiaDrCn04i8GLfNs1qb2A4Bu4TTQh9XDowIxwD1/GmR18Tstx1MLuQXM7ptfhtYAQWajKL2M05qHEZYAh4jxW2Do/p4HtJNpxGNmRyA+lS3eODgYrOsZkMvbKcbL3C8dTeQwNiiqK2lFhipbXY/rVSASZTwT0tj9OCxRVtbUshBHDb89sWtwYc+KtvncDVe/P49PYcQtEMGJagLeDAY4casa/DC7aGgmF5MQklbO7fU4nKkBeSsPg3XjavlkjtbhiGgchn00UvPFiJh+9cGwSPteA5QkheQw1d1RH6ZCJXvWLlteFoBLHPZ+HY54fjYK5RhxnZBfN3ZwBAXtjZuwQsX9lrj+Ol8oOoPY0GxBRFbakDnT68/tFo0bQJUeCQUPNXkG1i8bclliUVbWA7tT+ANy6NmRrb2+perjKxkiTyOH+0GeePNps+716VmTKfggIA6ckoDYjXwdbhgbXZidRYBJnpOAxVB2NhIbW7ITY5TTXPMHQDwYujuWYcxZ7XDMRuzUPPanCfaCo6ptgxK7LDm9Scvj6EZgZMjSUPx1O1jQbEFEVtKa9TRG+rG4PjhXVWnTYBidSjgJhlCSQrXzAOAI701MG2xnPFnNofwO+uTiEtl74VSghowGuCrpROlTAzXs9qSI2GkRhaRGY+CUMzwEk8pA433CeawIrmf797GcOzsPf4YDdZRWK15EhozWB4pcRQENY2Fywm0pC4ChuBVNw4ZJNZrB7YnM1IxspvlrN5OsBbarNaDPUI3VRHUdSWe/WpbtilwmBHFLjlIJcQoN5jBUHhCpjVwuGZk5XVCZVEHt95aR8sJXIlCQFeOteBnlZ3RceuRZXmnK4er0QymP31PSz8bhTJkQi0RBZ6WkE2mELks2mM/2AAkWvmK4NQa0sMBU2PTQ4XrwCzmtjsBGMx/xqwbWLd4/Vq6Hy2bKAriG40tF/YphlRuxkNiCmK2nJep4i/eO0wWovUQq33WFHnFtHgk2C1FAbNdonHn756EHXuwrzJcjqbXPjLrx7BsT4/WDY/0O5sduIPX95PV4dNsra6KhovrRivpRXMvzcCeS4JPaOgWK1pXdEe1jSe3uhUa5oSl9esAlJMetJcKgxhCOy95lasBa91V6TLcIINbfu/ClddHxgmP5hnGBYu/z607f8KWL7y9xZq76EpExRFbQufy4q//NpRTC0kcH1oAcm0AoFnsb/Di95WN8bn4vjk5ixGZ2LQdB1uuwUn9wdwvM8PnqtsdXKleq+E33++Dy8/0YnZxSR0w4DXKa4rwK5lgscKwSeZKqnGe3NjlyQGF6FGMtCzpdNXDFVH7MYspHZzt/E3mxKdR2rwCrIL44Cug3V4IPWchKWp1/Tms51mVJjaYmg6DMMw9fM5DtRDicpIT6zd8IazC/A9sXtaIXO8FQ2dz8Lfeg7JyBg0LQuWs8DmagfLFe4boGoXDYgpitpWzX47mv2FK8XtDU60N5RuMrARditfs6kRmppFLHgPscV7UJUECOFgdTTBXX8QVrv50liesy1Y+O196PLa7YIZgYX3TMvy14amI/kgbLourZZWkRwObWtAbOgaop/8J1L3r+Y9rkRmkZm4A84VgPeZb4Gz73waQDmVpDUAAOEZ08E+IQTex1uRrLchMRTMW4lmBBZSpweO/X6wJmp4bzeWE+Gs69/paVC72O571VIURVGbJhWfxvTwm9DU/NvoSnAQseAgnL5eNHQ8DcKUD6R4hwX+57oR+XSyaFktoU6C53QzeKe4/JiayEJLKWXrFy8xVA3pySiAynLGNyLy0c+RHv18zefV6ByCb/096l76C7DizqcClMLZBAheq+mawVKFqTCEkOUNf9lQGrqsgrAEvFcCU2H7aYraTWhATFEUtUfJqUVMDb0OXVt7dTYWHAIhDBo6nzF1TN5hgf/ZbiiRDNKTUehZDYzAwtriAu8WC7/BqLxkl6HqMHTDVImxjZLnRksGw0u0ZATJWxfhPPkiACCraLjzIIRYKgueY9Db6obPtTvScGy9PmQ/MdepzWYyL7gYwbs7fl6K2gw0IKYoitqjFqc+LRkML4ku3oMncBQWyXwbat4tFg+AV2FtPBje/MohYRgQnt2WYBgAUoOfmh97fwDioWfw1qdTuHJ3HnL2UeoIIUBPixsvP9GJgHdnmzxI7W5kpuMlc30BwHk4AIHm0lMUABoQU9SmiyzOIhENgxACt78RNkdltyQpajMocgLJ6Ljp8ZGFWwi0P7np82B4FlKHB9lQBoa6du7x8ngrV3FFi42QZ0dMj9XkNH7y8w9wK1R4IWAYwNBEBH/9s8/x5793qGie/HZZyvWN2QUkhoIF6SrZ6sW3AAAgAElEQVSsyMFxKAB7t/kLoGqjxGUkh4O51B7dAGvjYevyQmxyVM0GSWp70YCYojbJ5MhdDF3/GJHFubzHA61d2H/iPDz1jTs0M6oWZZJzMCroFpZObF0NYHtfHeL3FqFGywTEhIC18rD3bF+gZmilK1+sFIplMKvFAG7tlXE5q+GfX7+D/+M7p3a0tTchBK4jDXDs9yM1HoUal0EIAe+1wtpsruNdNTJ0A5Gr00jez6+vrEQzyEzHwTkE+J7sAO+gFSaofDQDnqI2wd2rH+HT3/5HQTAMAHMTI3j/l/+C2fH7OzAzqlYZRvnV2LzxemXluioheKyou9ABZo223AAAQsC7RTgPBSBsY8oBK5mrbGIYQDyZRYYpn2IQTWRxd9Rcw4utxvAs7N1euI81wnW0AVKra88GwwAQuTJVEAyvpMazWHxnBFrKXNUTqnbQgJiiNmhuYgR3PrtY9DlVUSCnU0gn4rj05r8jnUxs8+yoWsULlZWw4y1bV/IOAGztHjS9dgBiszOXcLvk4aqw2GCH90wLXIfNl4HbDNauo6bGJTMKQowXScZci9+BwfmNTItah2wojeSDcNlxWkZF7A79/VD5aMoERW3Q8I3CTTnZTBqpRAxKdkWpqzDwmx/9fzj34tfhb9o9heupvcnqaIAgupHNREyNd21DjVZLnQ3NXz0IJSEjORiEmlLAihyEOim3cslu/xqN1HMSiZvvw1CzJcdpmoFRvtf0cePJ0sejNl9i2HzL6tRoBK4jDWD49Tf9ofYWukJMURuQTsYxPzW66rEEoqGF/GD4ocXpMXz46x9hYvjWNs2QqmWehiOmxgkWJ+yezi2ezSO83QL3iSbUnW+H51QzbB2eHQmGAYAVbfA8+QcgbOn1IbX5GKb5NtPHFWigte3kOfN34AxVRzZorlYzVRtoQExRG5BOxvO+VrMyktG189d0TYNhGLj6u18jFl7c6ulRNc7tPwBP4FDJMRwvobn3JRBSux8HYnMvfM//CSyN3QDy82s5hw+us6+i9enXUElxgr623d/Vbq8xtArbVm9h3jxVfWjKBEVtALOqu1c6mcCa+/oNAwYBdD232Wnk1lUcO//C1k6Qqnn1bechSn6E524gk3p0EcawPJy+Pngbj4MXdq5E2G4h+Fvhe+6PoMaCyC5OALoO1uGFUN++XKarv92Du6Plc1R5jsHJffVbPWVqFVbiS7YVLxhv5bdwNlS1oQExRW2A01MHiyhBzqRgGDrkTHLVCAO6rkPXVBiGAUIYBGenIFhEDN+4jKPnngdhandljtoezrp+OOv6IaeCULNJEIaFaPODYYWdntquwzl94JzFu7e99HgnxmbiSMulS7V98bF2SCINtrab1OFBNGyufCDvEiF4aFMS6hH6SUxRG8CwLNr35fI0dU1HftlXA5qiQFOV5XqwDJtbUc7KGYTmZ3D14pvbPGOqllkkH2zuNkjOZhoMr4PfY8V//coheJ3F6xDzHIMvP9mJc0eatnlmFADYOjymuyLa+9bfspram+gKMUVtUM/h05i8fweJSH7usKaq0I1HOWqEMAV5mmN3r8Hp8aH3yJltmStFURvT4LPhf//2CdwbC2Pg3jziaQU8y6C3zY0T/fV0ZXgHMQIL3/l2LL4/CkNbuymNrcsDW9fe7dJHrQ8NiClqgyyihPNf+iYuvfETRIJz0FQVMIzlXGEgFwyzHJ/XMpTjBRCGwdDnn6D74Mnl1WOKonY3Qgj2dXixr6PyoGp8NobFaAYMAVrqHahz09v2m8lSb4f/uW7EbswhM5O/6ZmzC7D31cHeS1eHqUI0IKaoTWBzuPDc1/4UV977FW5dfg9yJgUCAsIwYBi2aJ6w1ZbbyCSnU5geG0JL177tnjZFUdvk2uA83h+YwmwwtfwYIUBXswvPn25De+PWNkbZiNz+h+rpbid4rKh7qgNqIotsMAVD08HaBFjqbVX1c1DbiwbEFLVJCMPg+JMvIhZewMTwbWTJ2jUueUGAxWpb/joRMV9QnlofTdNxdyyMcDwDliFob3SiqY5WV6C23tuXx/DOlcmCxw0DuD8Zxej0TXzjhX4c7NodK5eGYSAzE0dyOITMXALQDTAiB6ndDXuPD5y9OvLPObtQNXOldh4NiClqE3E8jyde/gP8x9/+N2QzxQNiXrDA6a1btVJBVy22imEYeO/qJC7dmEEipeQ91xqw48XHO9DZ5Nqh2VF73d2xUNFgeCVNN/DjtwfxV986AZfdsk0zK87QdAQvTSAzFct7XM+oSNxbRGIoCO+ZFkjt7h2aIUVtDVplgqI2mWi14fDjz8Hl9cMiWsFyHFiOg8UqweWrh7suUFC/2OWrzZqlhrH2xpfNOv6P3v7/2buz57juLE/s399dct9XILESAEFwp0RS1F5SSbWX1DXVU9XL1Hgc3e3w2OHwiyP8X/jNEZ6JGNvjmd7C057uUpWqVatKJZa2ksSdBEAQIPY1E7nnzbv6IQkSIHK5CSA34HyepMyDvD+QSeDcX57fOZP45adzu5JhAJhfzeL/eucO7s1UHqZCyH78/saSqThF1fGHuysNXk1tm58t7EqGd9ANJD6dh7SSqRxDSAeiHWJCGuDYyQuYvP4JLLbaB2YcLg+6+oebsKr2sJbI45M7y7h5fwN5SYUocBgb8OPKmW4M9RzsTu2nd1Zw8371iYCabuDvfzWB//VHl+DswEb9SjGLTOI+VCUHxglwenrh8PS2elkEQCpbxPRiynT8F+NrePO5gQauqDolJSE/Z2K9BpC+vQZbl7vxiyKkSSghJqQB7E43BsfOY+betZqxY8++dGQOenx8axk/vTq9o1+zouq49SCOWw/iuHQygn/x2siB/Xl8fMtck35F0fHF+CpefaZzEklNlbD68HfIJmd27LQnlq/DYvMhMvAynJQYt1Q6J6OeD0HSOblxizEhO2X+LIMcz0NOFmChLhnkkKCEmJAGOffim1CKEham71WMOXXpVQycONfEVbXOzal1/OTD6aoxn99bg9Ui4DsvHasat5EsIJOXYRF5dAWd4LndCfTiehbrm5UPNj7t+uR6xyTEmlrE/Pg7KBbKl3rIUhKLkz9DbOQbcHr7oRRTMHQNvOiAIB7+BEZJSsje30BhMQ1d1sBZeNh7vXCNBCH6yg/VaARRqK8qUeBbW8WoJMz/e9mKb7eEOJ4qoFBUYbcKCHrba22kvVFCTEiDcByHS199C33HT2P67pdYnS8lgxzPo3foJIZOPQt/pLvFq2wOwzDwy8/mTMV+cmsZrz7TA7dj9+nw65Nr+OjmMhbWso8fcztFXD4Zxcvne2CzPvmRlsnXt9vW6t25emwsflYxGd5i6Crm7v0jrI4gVLk0UpwBcHj74Y+ehdPb14SVNl/69irSd9Z2PKYXNeQeJJB7kIDnbBSeU82p2Q/77HA7RWRyu+vXyxnube3hznor+ht8BACbGQkTs5soyhqcdhGnjgXKDj7RdQNfjK/i0zsrWFrPPX68K+jA82e7cWksCq7MTTMh21FCTEgDMcbQ1T+Mrv5h6LoOTVVKAzmOSInElqmFJOJJyVSsphv4w91VfPXSzoTtnz6Ywmd3VnfFZ3IKfvP5Am5Px/FXf3QWrkd1wGKdu20WsTMGo+iajHR8smqMYWgo5uMwdAWGoUEQSy3+DAC51BxyqTmEYpcQ7LnUhBU3T2ZiY1cy/LT0rVVwFh6ukca3OON5DpdPRmt2mdhy5XRXg1dUnei21rVLLLob09IskZbw06vTmJjd3JF0v/M7hnPHw/j2i4OPE2NN0/G3v5goezB2JZ7HP/32Ae7NJPCjb46B5zkk0hI+vb2MG/c3kCsosIg8xgb9uHK6C/1d7dsLmjQedZkgpEk4joNosR65ZBgAFrft6JqxsLbzBPvVG4tlk+Ht1hIF/M17T8pT+rvcsFvN3/OP9vvrWmOrZJNz0LXqO45yIQFDL8VoSvkEZ2Ppc6Q3Jkxd09A1FLIryKcXau5Mt4qu6sjcrZ4Mb0nfXoWhN3h785EXz8UQNFGmMTbob/l70Dlk/vqC2wJL2Fk7sE7rmwX8H//fDYw/3Ny1A61qBr4cX8O//8dbyEul9/fPPnpYs0vMxOwmfnJ1Bn+4u4L/7W++wIfXl5DOydB0A4WiimsT6/h3//UWfvy7Bw3vfEPaF+0QE0IaTq/zl8z2cE03cPW6udZVs8sZzK6kMdDlgSjweHYsYrrt1fNnWrs7Z5auVd9p17UidO1J+Ue1X/CJlWvwhE5UfF5TJSRWbiC9cQ+q8uS6NkcIvugZeEPtM12xMJeELmu1A1EqoSjMp5rSS9dhE/FXb5/Bf/rZPSxv5MrGnBkO4gdvHG/5zbI14oIl6IC8bZpeJe6xcEPW+3e/GEeuoFaNWd8s4CcfTuOtV4ZMt6r74Mt5WEUeXJmpoVs+vb0Cm4XHN54frGfJ5JCghJgQ0nBhn6Ou+NC2gzr35zbrqu/9/O4qBh599Pnas724N5NAIl09iXzxXDe6gge/29UIHF/9Y2pV2ZnMVEtaioUkCpll2N27a9lVOYf5iXcgS7vbcEn5DazM/BaFzDKig6+1PJEDALnOA2FyPN+04RJelxX/0w/O4/58Ep/dXUE8KYHjGHojLlw53YVYuH0mJgZfHsDGb2egpCr/m3GfDMM5FDjwa88spXaMtq7m9oM4Qj47VM3MzbaBjaQEt8MCn7v64JOr15fw0vmex6VX5OighJgQ0nCnjgXgtAs1d362XD4Vffzf8Sq/mMuJb0t+nXYR/933zuBv3hvfcRBvC8cBL5/vwTeeb13v13o5vf3gOB66Xn43dKtUYgsvVP+4XpaSZRPipamfl02Gt0ttTMBiDyDQdb7GqpugzhKIZn80zhjDaH/ryyJq4W0Cwm8MIXs/jtyDBLRtA22sURdco0HYY42ptb0+uW46VtMN3HpQvcf4lkJRhaLqULTanyBouoHP767gtYuH89ApqYwSYkJIw/E8h5fOxfCLT2t3mjh5LICI/8mOcpVPOMvintqt9Lqs+B//5XnMLKXw+b1VbGaK4DmGgS4PnjvdBY+zMQeDGoUXbHAHRpCqWP+7/fvnwIu1dud37+7mM0so5MzV426u3oQ/ehaMtfZIilBj529X/FMjkpVMEbmpOAoLaRiKBmYR4OjzwjkSgNBh75H94kQenlMRuE+GoaaLMHQDvE0A3+Bd02zBXDeOLbKim4orPi6lMfdJRrmbZ3L4UUJMCGmKrzzbi0Rawuf3Kida/V1u/OCN4zse643UNw2rL1o+/ljMi2Ox1ra1Oiih3ueRzyxBKe4en8tx4uMaYtHmrZmo2pyhXY+l1sdNr0WVc8il5uDyDZr+mkZwDPqQur1qaqeY8QyOwSflEum7a0jfeurQpiIjM76OzMQ6fM/GmtKVot0wxiB6m9e32VJn3+aI345kplgzbusdYTXZSabeMw/kcKCEmBDSFIwxfP/14xju9eGTW8uYXXmSzAV9Nlw53YUrp7sgCjt/afVF3eiJuEx1qmAMeO50tGZcpxNEO/rHvoflmV8jn955aFCwOKGqEixWL/gaQzjsriisjt0JsSKn61pPucS82Xi7COeAD7mZzZqxjmN+8I86kGQnN3Ynw9sZQPKLJTCBg3OwOeUOhmFg/OEmJuYSKCoaXHYLLoyG0dNGtcaNcGLAjxs1Rq1v99XLfVhYyyIvVS/FEgUOHAe4HOZ2uIOe5t0EkPZBCTEhpKnOHw/j/PEwkpkisgUZVlFAyGerejDr61f68R9/erfmIIArp7vgdx+NX2aCxYm+E2+jWEggvTEBVcmD4wQ4PL1IJ6aQ3Zyp+vWMMQRj5fsQszo7crbDoToA8F2MQc0rKK5Wvnmydbvhu1CqmdZVHenb1dv5bUnfXIVjwNfw73VybhP/9MGDXTufv7+xhP4uN3745igChzRhOzMcwru/nzF11qAv6kJ/1IPXL/Xh3avV3+tOmwg+wHaVU1Vy+VRndJwhB4v6EBNCWsLntqI34kbYb9+VZCTSEm5MruPLiTXMLqdxvM+PH745WnZE85ZnxyL47stDjV5227HaAwj3vYDuoTcQHfwK3IFhdA+9AZevv+LXMMah69jrFafV2V317bLbnO2xK894DqFXB+F7pntXTbHgscL3bAzBlwfAHg1tKcwloZusQ9UKCqTF+nbO6zUxm8B/+tndimUAcysZ/Pt/vInNTH0HTTuFwHN4+9Vh1MpbRbEUBwAvnYvhq5eqj1x/7WIv3nrF3M+GEwN+RAL1dcUhhwPtEBNC2sb8aga//sMc7s8nd+wGh/12vHKhB//Lv7qIz+6u4PrkOjI5GRaRx2i/D1fOdGOwm6ZMbeE4AbGRbyGXmkVy7Tby6UUYhgFesMETHIUvchoWW+V6am/4NBLL10yN8rW7IrA5wwe3+H1iHINrNATXaAhKSoIua+AsfNla2OKGuRZf2+PtDRqvrGo6/uE396HXyM8zOQU/vTqDf/2tkw1ZR6udHQ4BXwP+8YMpSMXdXSG8Lgv+7OsndpSPvPncAE4dC+KT28u49SCOoqzBauFxeiiIK6e70Bd1wzBKrdeqdbLoibjwwzdHG/J9kfZHCTEhpC3cn9/Ef/7ZvbJ9Rdc3C/iv70/h5QsxfPvFY/j6lc5pk9YqjDG4fINw+QZLLcYMHYwzd6hItLrgi57B5urtGtfgEOq5chDLbYiaB8LqPTzVwOl2t6Y2TLclHH+YQDJTrNlTt1OdHQlhbNCP65PruPcwAVnR4bQJOHc8jLHBQNlPimJhF77/+nF8//XjMAxj16dOjDH84I3jGOrx4pPby1hafzIkxee24rlTUbx4LtYxI9zJwaOEmBDSctmCgr/9+UTNJvtXry+hL+LG2ZHdB8FIZYwxgNX3iz7c9yIMXUNy/V7Z5zmOR9fQG3B4eg5iiU1naDrUvAIlKZXGODOAtwrgbAJYhdIcwdW49mt3a4wf3s4wgHsP43jhbKxh62k1UeBx+VTXnup5K9V5M8Zw6WQUl05GsZEsIFdQYBF5dAUdbVMHT1qHEmJCSMt9fm91W6/Q6n5/c4kS4iZgjEN08Cvwhk8iuXYHudQcdF2FIDrgCY7CGz4JoWaP4/ZUXMsi/tEctLwCXX6yK6sqGpCTIbgsu3ruMp7B3sDJdpJsbnf4Sby5fy+kvJDPvmMiJiGUEBNCWu7ahLkhEEDpYNFGskC/zJrE5oyg61ik1cs4MMWNPDZ+9xCGZoDxHDirAL24LRk1DKiPDrVtT4odg09atTWCo87XtjdwLYQcRdRlghDScsls7eb626XqjCdkS+raEoxtpTmCxwpWpm5UzcqPxztboy74ntk93vognanjUw+eYzg9dPQGhRDSSJQQE0JaTuTr+1Ek1DnRihAAkDcLkBOFHY8xxiD6bOCdFrDtc8IfJcPe810IvTr4uFVbo5w6FjQ9RvzUUBBux9EaJ01Io9FnLoSQlhvu8eLWg7ipWLtVQCx0uCd2HXWGYaC4koWSLoIxQAzYYQ059/26xZXs1gWgFbVSqYRuAIyBs/IQg3YYql5KhhmDo88L91hzWsrxHMOffG0U//Gnd6GolXuv+T1WvPXyMTxcTuOzOytYiZe6JXQFnXjudBe1HyRkjyghJoS03PNnu00nxBdPRiDSDvGhlZ2KI3NvHVpe2fG46LPBcyYKe8/eEz5d1aErGtTUo84S25+TVSArQ3BbwdtKvxqfjmm0YzEv/uLt0/jxBw+wEt/ZI5kxYLTfj2+9MIC//+UkphdTO55fiedxfXIdw71e/Pk3xqjGmJA60b8YQkjLHYt5cXY4WDMp9rmtePVC57X50nUVjHFgjBL5apLXl5Gd2Cj7nJKUEL86C//lHjiHAnt6fUPVoCSlyv2HDQNqWgKYDbxVeJwYN9NAlwf/8588g9nlNMZnNyErGlwOEeePh+F2WPAffnwL81VGUz9YSOH/efcu/vLtM3TjSEgdKCEmhLSFH745Cp6fqjhJKuy349985xRcHVI7qRTTSK7dRmpjAppaOgRod3fBFz4Nt3/I9JCMo6KwmK6YDG+3+fkiLGEnxD0MpZCqJJLbqZkieAsPx6C/7msclIFuDwaeKn/47O5K1WR4y9xKBtcn1/bUw5eQo4oSYkJIW+B5Dj98cxSvPtODT++sYG4lA03XEfDYcPlkF8YG/R3TPD+TeIDlmV/DeGoObyGzgkJmBSn3XcRGvgVe6Izkvhmyk7WTYQCAAeTux+F7tr6hFHI8DzVVLLVZk5TqwboBJvKwhvdft3yQPruzYjr20zsrlBATUgdKiAkhbaUr6MQfvTrc6mXsWSGzjOXpX8MwKh+MymeWsfzgF+g98d0mrqx9aQUFxbVc7cBH8rPJuhNiaSUDoDRtTlF1GGqVwRYcgzV8MENH5M0Ccg8SpVIMAKLPDudIAKKnxljpp+QlZce44VqW1nPISwocNrF2MCGEEmJCCDlI8aUvqibDW3LpBeQzS3C4D+/4XbM0qb4pbbqswdCNiiOWyzHUUt0w4xhEvw1aVi5dd0c9canbhOB6qgXbHuiyhsQn85CWMzseL67nkb0fh6PfC/9zvabbuVXrPFFJrVHohJAnKCEmhJADIksp5NMLpuNT63dbmhCvL81hefY+VLkIi82OnqGT8Ieb/zE7q/fwF8fqSoYBgNt2QI4xVuom4bJAL2qP2qwBnEV4/Lqcfe+/Hg1Nx8bvHkJ+qlPEdvm5FHRFR/CVAVOlQA6bCIFnppNcUeDgaPKhQDWvwJA1cBYevIN2pklnoYSYEHLgFLmIlbkpFAt58IKIrr4h2F2Hvz9qMb+BevbkpJzJutkDllhbwpe/+2dkNnde//7Nz+APd+PS69+Fy7u3Tg57Ibgs4J0WaDnZVLytq/4+1I5+L1I3lrH9L4gxVrGThGPAV/c1tuRmNqsmw1uk5QykhTTsfd6asaLA4exICNcmyh86fdrZkRCEBg8TAUo9o/MPk8hOxaFsG3piCTrgHAnAMeDrmNp/crRRQkwIOTCKXMSdzz7A/P3bUNUnB5cYY+jqH8HZ578Kp2fvicbh0/yPtBOri7j67t9D08qXKWyuL+ODd/4aX3n7R01LihljcI0EkLph7tCYa6T+scW8XYS9z4vCXKpmrOi37WsQSG7KXE9toNR32UxCDAAvnovh+uR6xa5xWxgDXjzb2FHTQKlPc/yjOUiL6V3PyfE85Hge0lIGgRf6KCkmbY8SYkLIY5pSgGHo4AVb3W3BFLmIq+/+HZIbq7ueMwwDy7P3kVhdxCtv/TncvvoTmk5gsdWX7Ncbv1+GYeCL375bMRneIksFXPvw53jlu3/WpJUBzpEg8nMpKJuFqnH2Pi+se9ghBgD/xRiUpAQ1XawYw9kEBF/o39PrA6WyASVV+fWfVlzLma6H7gm78NYrQ/jJh9MVk2LGgLdfGUIs3PhpjslrS2WT4e0K8ymknSK85xufoJPailIes+M3sTgzAblYgGixITZ4HINj52FzHO0JoJQQE3LE6bqK9MY4kmt3UCxsAgA4joc7MAJ/9CysjpCp17n1yW/KJsPbFaU8PvvVP+GNf/mX+153O7I6grC7Iihk10zF+8KnG7yindYWZpBNb5qK3VieQ3pzAx6/ub///eIEDuHXBhH/aB7FCr12Hcf88F/q2fNuI2cREHljCMnrKyjMJWFsr8dlgD3mgfeZbgjOvbfDM7T6D78Zqg5mMXcD+vyZbnidVrz/xTwW1nb+OfVGXfjqxT6MDTZ+Z1+TVOSmzb2XsvcTcJ+MgDP5PZLGmJ+6iy9/9zPo2vYOKymk4quYuPYRzr34NRw7eaFl62s1SogJOcJUpYDFyXch5XfWkuq6htTGBNLxSUQHvwJvaKzq68jFAuan7pi6ZnpzA+tLswjHBva87nYW6H4WS/ffq1kMYXOE4PD2NWVNW5YeTtYXPzPZtIQYKCWs4deOQU7kS63KMkWAMYh+O1wjAQiu+odxlLtG4Lle6Be6UFhIQ5c1MJGDrdsD4QAOgvFWAWAwXQ3DeA5MrK/W9+SxAE4eC2BpPYuVRKlWuSvgaMqu8Jb8w03A5GhrQ9ORn0vuqdSFHIylmUl8/v5PKj6v6zquX/05eF5A/+iZJq6sfVBCTMgRZRgGlqbe25UMPx2zOvNbCBYXnJ7einGL0xNP7TpUN3//zqFNiF2+QYT7X8L63O8r5kRWuw89x7+177pKXVOQzyxCU4vgBSscnl5wXOUf60pRquv1Fbm++INiCThgCRxMH+BKOIuw5xHQ1V+Xh63bDWkpUzsYgGPAu+f3QSzsamoSvF21spODiCcHxzAM3Pr0N6Zib3/6PnqHT4Ljj95uPiXEhBxR+fQCCtnqJQ5AaaMrsfRl1YRYypsbift0vKFrkPLr0DUFguiA1XE4dpD80bOwOULYXL2FbHIGxqOCT9Hqhi98Ct7w6X1NqdPUIuJLf0BqYwK69uTwIi9Y4QmeQKjnEjh+9+uLVntd1xEt9Q2PICWu0ZDphNh5vEPf83RGrmOszk8jn6l9mBQolbUtzkygb+RUg1fVfighJuSISq3fNR2bzyxBLmzCYveXfZ4X6vuomXEM6wufIL0xDlV5sgtpdQTgj5yFJzTW8afS7e5u2N3d0FQJqlIAx/EQLO59f1+aKmF+/MeP6713PlfE5upNFDKL6D3xNnhhZ4lB7NgoHo5fN32tnqET+1rrUWWLuuA5HUH6TvVact+zMVh89d2ktAvRbwdgroYYAMRAZ36fh8HG8lzd8UcxIW58k0JCSFsqFhJ1xlf+5RftO2b6dQxdAzOWkVi+viMZBoBiPoGVhx9gZeb9x7uqnY4XbLDa/RCtngNJ8ldm3q/6dwEAUj6O1dkPdj0e6Rk03UotHBs4tN1AmsFzJgr/5R4IrjI79V4rgi/2w9Wpu8Mo9Wk2O1CFs/BwmGwtRw6eVkc5Wym+vsmRhwXtEBNC9s0biCAY7QughGMAACAASURBVEV8tfaUNk1Nw+ur3m4sHZ+Exe5DsPvZg1rioSBLKeSSs6Zis5szUIpZiNYnNaaMMVx67bv48N2/g7atT/TTrDYHLrzyjX2v96hzDgXgOOZHcTUHNS2VDgj6bLCG997juF1wIg/3WBjp27XLrtynIqZHVJO9U7Mycg8SkB+1LhQ9VjiHA7A73XW9jsN5+IcolUPvUEKOKKu9vgNFVkf1+HMvvlGzdELXZMQGfOBM/HJMrt6Code3s3HYpeOTpkd5GIaBdGJ3Vwl/pBuvfPfP4A2Ey35dINqDV9/+EVye8uUxpD6MMdi6XHCNhuA6HjwUyfAWz+lIzV1u98kw3Cea16nkKDJ0A4nPFrDy7gQy4+sormZRXM0iez+O1ffuwxn3gNVR9N0/eraBq21ftENMyBHlDZ9CZnPGVKzDE6s5RMIX6sJL3/4TfParfyp7yE4QRPSO9sDhMHcAT1UKyKZm4fYPmYo/ClS5vsOLqpwr+7g/3I2v/vFfIL6ygOXZ+1DkIiw2O3qHTsIbjBzEUskR4Xs2BnufF9mpOAoLaUA3wHgGe68XzpEgrKHGdgs56gzDQOKjORSqDEgx1lX4uSgS2nJpcksVXf0jcHmP5s0wJcSEHFEOTy/srmjNThOMMdOlC8FoD77xp/8WS7P3sTB1F8VCDoLFiq7+YfQfP4OV6Z8jnzGf1KlFcyf1jwpWpaVa2XhWvXVSsKsXwa7K3UOOCkM3oKQkGJoO3i7uazDHUWQNO2ENO2EYBgzNAGeytpjsn7SUqZoMbxlwnIUsF5GVK58/8PhDuPjatw9yeR2FEmJCjijGGHpGvomF++9CypXvRcwYQ3TwNTiqtFx7Gsfz6B0aQ+/Q7mEedR8qY/SLdUsyU8S9BQ4sU9r1FQUOHqcFolA56XV4Ys1aXkfSJBXZyQ3kphPQi0/Kc6wRJ1yjIdh7jmYt5V4xxsCEzu4O0wl0VUdhNonczCbys0noRRWcyINzCOAq/DwQOBGnw69gM7SG2YmbKEr5x8+JVhsGjp/F2MWXIFr2P/ymU1FCTMgRxot29I19D+mNCSTX76CYL3We4DgBnuBx+CJnDrQ3sM3VhVx60XS83RU9sGt3Kk038NOr0/jszgoMw8CVLg4WTgYApHMynDYRYb99182GaHXD6T2cw08OgpqTsf7+DLScvOu54loOxbUc3Kci8J6l9yBpH3I8j42rs9AlFYZhQMuX3r+apkOTFHA2EaLbUrY0Qs+oOPHaixi7+BI2luYgFyWIFivCsf66W2ceRpQQE3LEcZwAX+Q0fJHT0FQJhq6BF2xg3MFPKvKFTyGx/GXVlmqGrkFV8gB0LEy+C44X4fQOwBc5XfdBwE5nGAb+y68ncfP+1g4+w3TqGMb8E4+eB7IFBZquozvofPxLkAGI9L3Y8b2cG8UwDMQ/fFg2Gd4uc3cNgtsC5+DRrKkk7UVJS9j4YAa6opceKPNjVJcUKCh1mCjHUHUIDiuifXQ242n0eSQh5DFesEGwOBuSDAOAYHHCH618gllVCpByq1DlDHjBBk0tQilmkVy7g4e3/1+sz3/ckHW1q8m5zW3JcMl6IYyJzVEY2358F4oa0o92ijiOR9fQG3D5zfeGPmqkpQyUlLlRwtnxyqPNO4GhG1BzMtScDEM/HL29j6r07bUnyTBQuvMttxMsKTBUfdfjAMDbjt5IZrNoh5gQ0lSh3heg6xqSa3d2PK6pEhSpdODDYveD43fvcCRWboBxAkI9l5uy1lb75PZK2cfXChEkJD+6nKsI2eMQmIrVtA0jY1fgCZ0AL9DI5Wpy0+aH0igpCXI8D0uws7olqDkZuak4cjObj+ujOQsPxzE/XMeDdHCww2gFZdfhOcYYOKsAXdrdU1wrKBDcO3+G2rrd4CyU9lVCfzKEkKZijCE68Aq8oTEk124jl5qHoatQpCREqxu86KjaHSGxcg3+6NlDn/QZhoHJuconwlVDxEK2FwvZJwceX359bNe4ZrKbmq1eKlEuvpMS4uJGHvEPH0KXd/bx1mUN2YkN5Gc2EXxlANbQ4emJfNjJ8TxQZoeft5dPiJ/+uwfQ0ZMRm4FKJgghLWFzhtF17HUMX/hv0DP6bYg2HwSLu2arMEPXkVofb9IqW0dRddQ7vVpWaJCJGfV3O2nMOhpBKyhlk+HtdFlD/MNZqDVqqEn7qFTuwok8BFftm2D3WAi27vom1h01lBATQlqukClfGlAxPrvcoJW0D4vIQxTN/4hmDHDY6aS4GZaAvaHxrZSdildNhrfosobcA/OlI6S1eEflEhfeIULw2HaMx2Yce/yc72IM3vPdDV9jp6OSCUJIyxlGfTubhlH+wMhhM9rnx4fXF5GXFOi6AY5jcNlFuB0W8E+Nvx7u9cFFCbEpzuNB5GYql6NsZ+tymdqBaweGYSA3be77Akq11J4z0cfJE2lf1pADgscKNV3+MChvE8DbBOiyBl3V4BoJwjkcgK3bTd1mTKIdYkJIy4nW+j7KE62Hf2DC1RuLuDa5hmS2CFnRoWoGZEVHIl3E3GoG2fzOj7tfOEM7QGZZ/HbYe2u/hxjH4DnTOX2I9aIGXVLriy+ajyet5R4N1YzhLDysISeCL/bDHvNQMlwHSogJIS3n8g/VdRjMG9o9Be8w+eT2Mn72+4ewCDyCnt2HBw0DWNssIPfoMM2L57px8tjR6tG8X/4rfbDFKt+IMZ5D8OWBjjpMVy/DMMq1siVtyjkcgHOk+r9zzsoj+MrAjvIJYg6VTBBCWo7jBPjCpxFf/rJmrMPdDZsz3IRVtYaiavjFp7OP/9/rsoLnGTbTRShP9RbN5GT84KvH8fKFngO7/sxSCp/dWcHcaga6biDotePyySjODAd3lWl0Mk4oJbzF1SyyUwkUV7MwdAO8XYTzmA/OoQD4fZagKGkJqZur0PIKOJGDaywMewMPNnFWHrxdhFbY3XVgi65q0PMqtKIKMGD5nXuwxzxwjQRh7XK1xY6iruoormSgFTVwAgdrlwu8ldIVAPBf7IHFb0d2Mg4lJT1+nPEM9j4vPKejEFzUUm8v6B1GCGkLwZ5LkKVNZDZnKsZY7T50D3+9iatqvuuT65CKO2uqXXYLXHYR+aJa6iRhAKLIw2ETED2gHUxZ0fB3v5jAxOzOGtRUVsb0Ygq//MyG//a7pxDydc4Bs1oYY7B1uWHrOtgkVc3KWH1vEtJKFttbhaRvr0L02RF+/Rjsvd4DvSZQ+n6cw36kb6+VfV7LK1CzT2pQeYcFDAzSUgbSUgb2fi8CV/paVlOsqzrSt1aRn0nsGECxlex5z3Xt+yblMHAOBeAcCkBO5KHlFTCegxiw003DPh2e231CSEdjjEP38NcR6X8JFtvOZIEXbAh0X0Df2PcgiIcnISvn4XK6wjMMDqsIn8sGn9sGp00EA8PMUqV48wzDwN/+fHxXMrxdIi3hP7xzG2lq1VWVmpWx8Pc3IS1nUK5vnpIsYPmdceTmkg25vnM4CN62OzHSJHVHMgyO7UouC3MpJL9casi6atEVDRvvTyM7ubFzGhsAQzOQf5jE2q8eUKu4bSwBB+y9Xti63ZQMHwD6EySEtA3GGPzRs/BHz0LKrUNTCmC8ALsz2rBx0u1GrTBytRJN23/Hjcm5TUyaSNDSWRkfXl/Ed16isdCVrL43WbVkAQAMTcfaz+9j4C8vguMOdl+KtwkIfeUYNj6Ygfb4gJ0BbXsiyTGIPlvZneDcdALuk+GmT7JLXluGnChUjdHyChIfzyPy5nCTVkWOEtohJoS0JZszDKevHw537MgkwwDgK3OIrmq8e/8twSqNiC7ny/E1KCoNAClHSUulMgkTdElF5s56Q9Yh+myIfPM4PGej4B0idFmDoelgHAfeaYEl4AAnVPg3ZdQ32vogaEUVhVlzO+ZyPF+a2kbIAaOEmBBC2sjFExHTsQLPcP74/g8YVi7T2K1QVLFCCUlZqZurZcskKsmMNyYhBgDeKsBzKoLut8bgeyYGS8gJMWiH4LTUrBFWNqWqzx+0wlyq4iS2cnIPzfdaJsQsSogJIaSNRAIOjPb7TMU+OxaBw7b/Q0ZqnWUXWh3Jy1Gi5auXSjytnp7B+8IxMI61RQeJcmqVmDxNL1DvZHLwKCEmhJA284M3RhGpMS54MOY5sFreesouGAN8HTK5rdm4OkZtAwATmvMrWPTU9/clHEAZTj3q7ZnL+PZM7Elno4SYEELajNMu4t/+i3N4+UIM9qdOj7udIt58rg9/8dZpiJXqQOt0ccz8NLahHu+B1C3XQ9dVqEoBht7etcuusfrKVxwDB996rRxbzFO280QlzmF/A1ezmzXqamg8IWZQlwlCCGlDNquAb794DF97bgAPl1OQZA0Oq4DBmBf8AfeJvXwyig++XEBRrp1wvnw+dqDXrsQwDGSTM0iu3UE+vQig1JrP5R+EL3IGDndz1lEPe7cbos8GJVm7BpfxHHzPHtxAlarX4hhcY2Gkri/XjLX3eCDWebBzv6whB0S/zVTtMmfhYTdZUkRIPWiHmBBC2pgocDje58fZ4RCGe30HngwDpR3pH31zDGKNj/y/fqUfJwYaPyLa0DUsP/gllqZ+8TgZBgDD0JFJTGN+/B1sLP6h4evYi/DrQ6ZKAPyXe+ratd0v94kQXMeDVWOsYQf8V3qbtKKdfM/ETA0E8V7oBtekUhNytNC7ihBCCIZ7ffgfvn8OZ4aDeLo1bl/UhX/1zTG8drGvKWtZm7uKzOZ01Zj40hdIrt1uynrqYe/1IvqdE+AqDEpgPIfA833wX25+4ul7NobgywO7Sg5ErxW+izGEvnIMnNiaFofWsBPBVwbAWcpfn/EMvks9cB5rbjkHOTqoZIIQQggAoCvoxJ9/YwzpnIyljSx03UDAY0NX0Nm0NahyDqmNcVOx8eVr8IZPgbH22ttxDvgw8FcXkbmzjuz4OjRJBRM4OAa88D3b3J3hp9l7PLD3eKBJKvRiaV3NHsJRia3Lja63xlCYSyI/l4JeVMGJHGwxD5xDfnAWSllI49C7ixBCyA4epwUeZ+NLI8pJrt+FYbKXryrnkE0+hNs/1OBV1Y/jOHjPRuE9a/7AYjPxNqGliXklnMDBORSAc6g17z9ydLXXbTUhhJAjTS7UNyVNzjd3qhoh5HCihJgQQgghhBxplBATQghpGxZ7fYem6o0nhJByKCEmhBDSNryhUzDbWE4Q7XD5Bhu5HELIEUEJMSGEkLYhWl3whE6Yig10PQPGtaZNGCHkcKGEmBBCSFuJDrwKl6+/akyg6xz8XeeatCJCyGHXfj1XCCGEHGmM4xEb+SYyiSkk1+6gkF0tPQ7A6RuAL3IGTm9zhoQQQo4GSogJIeSAGLoGKbcOXVfACzbYnOFWL6ljMcbBExyFJzgKTZWgazJ4wQaOb48hEoSQw4USYkII2SdNLSKxch2p9buQpRQ0VQJgQLS6EYpdRrDnOTBm9qgYeRov2MALtlYvgxByiFFCTAgh+6AqBSxM/Bj5zAoUKQXDUB8/pyl5LEz+FBtLn2P4wr+BaHG1cKWEEEIqoUN1hBCyD8sPfol8ehlyIbEjGd5Oyq7gwbX/G6pSaPLqCCGEmEEJMSGE7JGUW0MuvQBZSgIwqsfmN7A+/1FzFkYIIaQulBATQsgepTbGoSkFAHrtYENHcu32o/piQvZGzckobuShJCUYRvWbMEKIeVRDTAghe6RIaWiq+TIIXZORS83BExxt4KrIYVRYSCE7uYHiev7xY7xdhHPYD9fxEDhLaUCJrmiQ43kYmgHeLsISsLdqyYR0FEqICSFkjxjH1blLx6BrcsPWQw6n5LUlZCfjux7XCgrSt9eQn0sh8Hwfcg8SyM8mYahPPrEQPFa4RkNwDQeauWRCOg4lxIQQskc2VxcYYzWqh5/geMuh7aOrazIyiQdQ5CwYx8PhjsHu6mr1shpOyq1BllJgjIPNGYFodR/o62en4mWT4e2URAGL/3AHose66zk1XUTy80Uo8Tz8z/Ue6NoIOUwoISaEkD3yhk5idea3kLVizVjGiRBEO5zegSasrHkMXcP6wsdIrY9D13d22bA6goj0vQCHpz0TMU2VUMiuwNA1CBYnbM6o6X7R6fgkEivXUcwnHj9WmqTXj2Ds8oEMZTEMA5nxjZoxSkqCoenQ7QI4kS8bl5vZhOizwTUa2ve6CDmMKCEmhOxZIZfFytwUFLkIi9WG7sHjsNocrV5W0wiiHeG+F7A49R5gVDtYx2CxeuAOHAcv7N7F61SGrmHh/s+QTy+Wfb6Yj2Nh8l3ERr4Bl2+wuYurQimmEV/6HJnEA+i69vhxi80Lf/QsfJEzVb9+Y+FTxJev7XrcAJBNziGfXkRs5Jv7Hi9dXM1By1UvsdFlDYZWeu9pBaViQgwAmck4nMeDNCSGkDIoISaE1K2QTePmJ7/B8sP7MLYlgjd+/0v0DI/h7PNfPTKJcaj3eUi5DcSX/oDyrdcYLHY/bK4own3PN3t5DZVYuVYxGd5iGAaWp3+NoXM/aoubgWI+gYXJd6Aqu7t9yFIKq7NXIeXWER18rWzimN2cKZsMb6frGpYe/ALHzv45BHHvh9rUbO1PHnTpya68oVUv3tFyMoprOdiiNCCGkKdRQkwIqUsuk8Lv3vlr5NJJSPks5KIEGAY4nofV7sTc5G1sri7h1bd/BKv98CfFjDH0jb0Fl38Qy9O/Kk2rgwHGOAiiA4LogMs/hOjga4dq/LBh6Eiu3TUVq2sK0vEJ+KPnGryq6gxdw+LUP5dNhrdLbUzA6gjBHz2767nN1ZumrqVrCtIb4wh0P7OntQIo1WDUsLU7bFatHWdCjipKiAkhdfn8/Z8gvrKAfDa98wlVgVyUkMskoSkyrn34Hp7/+vdbs8gW2PqoPZ+eRy61AMNQIYgOuAPHYbF5W728A1fILEFV8rUDH8nEp1qeEGeTD6EUM6ZiN1dvwhc5s2OXWCmmkc8sm75eap8JsSVg5obyyfo4wcRoAY7KJQgphxJiQohpyY0VzE/d2Z0Mb6NrGpLxNTDuLvKZN+BwH75ksBLGGJzefji9/a1eSsPV2mXdFV9Hv+ZGSW2Mm45VihkUMktweHqePCZn67qeWmf80yx+OywBO+RE5T87zsJBU0t10LxDrPma1pBzX2si5LCiSXWEENMe3P4ChUzlZHiLYRjIppOYnzL3kTrpPBxfO/naT3wjqLK53eEtTyfAjFU+sFYO4/a/5+Q521W1dIK3lf5cOZsIxlf/lW7rdkNwHc62f4TsFyXEhBDTVuanTffclaUCMsnq/VNJ53K4Y3UluS7vYOMWYxJj9SWoTyfAVkewroOBDnesruuVY+tyIfB8X8VSByZwsHW7IbqrJ7qMZ/Ccje57PYQcVpQQE0JMK1QplSgnn6svnnQOjhdNj6BmjMEbPtXgFdVmd3ebjmWMwfFUPMcJ8IZOmH4NX+S06dhqHP0+dH17FO6TYfB2AWClRNje40HotWOI/fFpOEeCFb+eEzmEXh2ExU9jnAmphGqICSGmWR311R96A5EGrYS0g1DsEvLpechS9RufUM9zEK2tb/Xli5xGcvWWqU85XL5BCJbd73d/1wVkEtM164ld/sEDHUgiOC3wnuuC91z56X/+Sz1wjgSQux+HtJoDNB2cQ4Rz0AfHoL9qf2JCCCXEhJA6BKO9WJ2fhq5pNWMFUUS091gTVkVahRft6DvxR1ia/iUKmZVdz3O8iFDP5ZZ3l9hisfngjZxGcu1O1TiOFxHsuVz2OUF0oPfEW1i8/8+QpSSAUs28rhWhKnkYugqLzQNesKNY2ITV7j/w76MSi88Oy+X2nArYaIZhQFrKQM2UejeLfjv1WyZ1oYSYEGLawImzmLt/C5nERtVdNsaAQLQHkZ7BZi2NtIhgcaJ/7HuQcmtIbUxAlbNgjIfDE4MnOAqOb69DXJH+lwCgYlIsiDbERr4Fqz1Q8TUsNi8Gz/wQ2eRDbK7cRGpjHLomg+NFiDYfON6C1Po9pNbvwRs6gejAq2Ac7dA2SvZ+HJnxdWh5ZcfjgtsCz5koHP2+Fq2MdBJKiAkhpnUPHIc/XKqrzCYT0PXdQwE4joPbH8LYMy+B4ykJOCpszghszvYvkWGMQ3TglVL5xNod5NOLMAwNgsUJb/AE3MHj4Ex0h2CMg93VjXXlY1hslROu1MYEDEND99CbB/ltwDB0SNk1aJoEjrfA7oweyaQ7eX0Z2YmNss+pGRmJj+ehSypco6Emr4x0GkqICSGmcRyH57/+fVx99+9hsdpQLORRlAowdB0cx8Fqd8BicyA2eBwnLhyuMcXkcLHaA4gOvLKv19hcuW5q0Ec6PgVv+NSBdJ0wdA2J1RtIrd3dUccsiA54wycR6LrQFi3umqGwlK6YDG+XvLYMa8QF0Xd4JkWSg0cJMSGkLt5ABF95+1/j7ue/w/LDSdgcT+r0bA4Xhk49i9HzV8A4amJDDi9dV+sa9JFau7PvhFjXVSxOvlt2Wp6q5BFf+gK51Bz6TrzVdqUqjZCd3NnW0dB0GLoBxrFdPZmzU3H4L/WAkEooISaE1M3l9eO5N/4IUiGHjaU5aKoKm9OFSGyAEmFyJBTzcWhq0XR8Lr2472uuzV2tOTpayq1jZea3iI18fd/Xa2eapKK4mgVgQCuo0AoKDPVJCRcTOPB2sdSmDgz52SQlxKQqSogJIXtmszvRO3yy1csgpOkMXW1o/NNUpYD0xqSp2OzmNJRiGqLVs69rtjO9qAKGASVdLP33UwxVh5opQi9qEL1WGKoOXdXBCXTDTsqjdwYhpGMYug7DMDsrj5DGKdej+CDjn5aJT8Iwdh9iLcdA6TDfYcYEDmpWLpsMb6fLKtSsXPoavsoMbHLk0Q4xIaStycUCHo7fwPgXv0cmlQAAeINRnLr0CgZGz4AXjsYBItJeON0Jq9GNYiEO2BSAq36j5qljwl05ilz78N6OeBOH/ToZJ3LQVXM3CFpBhWMoAMYoISaVUUJMCGlbyY0V/Oq//J9IxdegaU92gnLpJFbnphCI9uIbf/rfw+H2tnCV5CiRVjLITmxAWsmCKd1gBSsg6DD8WSCSAsTdQ2tKI5/3V1rEWH0f6B72Wv7cTBK8VYCq1B4SBBiPaokJqexw/4shhHSsQjaNn//dv0NibWlHMrxF0zSsL83iZ3/9v0NVlDKvQMjByoyvY+ODh5BWSu3OeNEOweICVA5s3QM2GQOknZ9YMI5DbPjrEET7vq5td3U3NL7TqJkiOLsA9mgktc6KULkMVC4FlUtDYwUYKO0gc1aByiVITZQQE0La0o2PfoVUYr1mXGJtCXc++23jF0SOtMJSGqkbu8dTW2xeWOy+0jAPhQebjgI6wAA4PT3oO/E2nL7+fV/f6RuAaDE3ipgXrHAHhvd9zbbGUCqBCEiQXQvQuBwMpsBgGgymQucKUPkkNFcSgsd66HfMyf7RZwiEkLajaSqmbv3BdPy9L6/i/Etfa+CKDidVKSCfnoeuyeAFG5zegSMz1KFe2fHKN2eC6IQgOqBpMgxdg9M2AO+JvqoT7OrFGIdQ7xUsT/+6Zmyo54qpaXudzBJwIMU9RM7xBQynBqZYwGX8YIoVYIAhFqC5k4CggMkSAoHeVi+ZtLnD/S+GENKRMptxFHLpOuI3oMhFiBZrw9ak6yoKmSVoqgxesMLhjnXsqFxVzmF94RNkNh/A2DZ+m+NFeIMnEOp97kgMdjBLSRdRXM/XiGLgeSvAA4g7DjQZ3uIJHoeuK1ib/bBstxUGINz3AnyRUwd+7XZj7/NCunsfBko1xIYoQwuslo2VrDMQuw92dDY5fCghJoS0HVmWUE93NcMAilK+IQmxrsmIL32B1Mb4jkEMgmiDN3QSgdjFjtqNU4oZzI//eMfY3y26pmBz7TYK2RX0jb1NSfEjWk6uK17N1BdfD1/4FFzeASTX7yKTeABdk8DxVrj8x+ALnzrUvYe3U9U0dFcG2P023oV3CshsTiBov9T4hZGO1Tk/xQkhR4bbGwRjzHTPYZ7nYbHu79BSOZoqYX7iJyjm47ueUxUJ8eVryKUXOmpU7vL0L8smw9tJ+Q2szl5F99BXm7SqQ4Zr7AEuweJEqOcyQj2XG3qddpbPLIF3iIBhQK1ywyI4LeAdFuTTSwjub3I2OeSoypwQ0nbsThf8YfOn5EPdfbBYbQe+jpWHvy2bDG8n5daxOvvhgV+7EaTcGgrZNVOxmcQUVKVWmcDRIPrtdSW5lsDB35yRnQy9VCrBOy2wBOzg7eLjPsOMMfA2ARa/HbyzdKNqGGbas5GjjBJiQkhbOvv86zDTR5/jOJx9/o0Dv74spZDbfGgqNpOYgirnDnwNB83s6F8AMAwdmcRUA1fTOXibAEev+VIE50iwgashAHZ03GACD8FthSXshDXihCXshOCxPW7JBux/UiA5/CghJoS0pdHzz2PgxPmqSTHHcTh28hkMnrxw4NdPxydhtozZMAyk4+aTzVZRlepJu2EY0HUV+qNuCXvZIdY1GbKU6ogbhHq4T0XAhNq/Mq1RF2xRc+3RyN45vf0QxHKfCpX/geENjjV2QaTjUQ0xIaQtMY7DG3/8F/j45/+A6btfoljIP64p5jgGq92F4dMXceVr3wPXgB6jar2jcmvU5bYDVuHwn2HoUJUcNDkPw3gyBCW5egt2VxdcvsGar13IrmBz9RaymzMwjFLnCqvdB2/4NLzhkx118LAc0WtD6NVBxK/OQpfLf/xujTgRfGn/PYdJbYzj4QufwcbS5zVjrXYfHN6+JqyKdLLO/glFCDnUeEHAy9/5U1x4+Rt4cPtzJNYWAQDBaC+GTl+Ey+tv2LUZq6+lWr3xreD09CIdv7/jMcPQUcxvwNB3T/tTlQIW77+HQNc5hPterPi6mys3sD7/8a4d9WIhibW53yMdn0Dv6HfBCwdf591M1rAT0W+PIj+zidx0AlpOATgGa9gJ50gAEhB7hQAADLZJREFUtm734zpW0niB2LOQChvIViltEkQHYiPfpL8XUhMlxISQtufy+ps+eMPujiG5fs90vMPT/kfY3YERrM1/tKN9nFxIlE2GOU583DkjsXITotVXtr9tZnMaa/MfV72ulNvA0tTP0Tf2R/v8DlqPtwpwj4XhHgu3eilHHmOlsdibKzeRXLu941MajuPhDowgGLsM0UolLKQ2SogJIaQMt38I6+JHUJVCzVjR6oLTO9CEVVVnGDo0pQAwBl6wgbGdpSSM4xHpfwnL078BUKr31bVimVdiEG3eHY8kVq7BGz65a6ctsfSlqbXlM8vIZ5bgcLf/jQPpHIxxCHRfgL/rHAqZZahKHhwnwO7u7vhPJEhzUUJMCCFlMI5HuPd5LM+8Xz0OQLjvxZZ+JKsUs0iu3UJqYwKaKgEonar3hU7CGzkNQXzSBswTHIVhGFib/R1kqczBN8bBag+A43cOOVGKGeTTC3Buq8WUcmuQ8hum15lav0cJMWkIxjg4PD2tXgbpYJQQE0JIBZ7QCeiGhrXZq48Pim3HOA5dA1+B2z/UgtWV5DNLWLr/HjRt53ACVc5hY+lzJNfvonf0O7A6nrQC84ZOwOUbxPTN/4xCZgWADjAOguAAL9pR6aS+LCV3JMTFwmZda5ULibriCSGkWSghJoSQKnzhU3D5BpFav4fs5gw0rQhesMLtH4Y3NPYogWwNpZgumwxvpyp5LEy+i8EzP9zxETIvWGFzhKFrasWvfdrTu+D174nTwSZCSHuihJgQQmoQRAeCsYsIxi62eik7bK7eqpoMb1GVPFLr4wh07+zXbHWEkc8sm76ezRl56uvrG0BRbzwhhDQLDeYghJAOZOga0hsTpuOT63d2PeaLnDK9Z2tzhMokxCHYXZEKX7GbL3LadCw5etScDCVdrNjnmZBGoh1iQgjpQIqcNbU7/Di+mIGuqzsGZFhsPnhCJ5CqkVgzAMGey2WfC8YuYXHyZzWn+jm9/bsSakJ0VUduOoHcVAJq5lHHEwbYYx64RoOwRqhlGmkO2iEmhJAjLDrwatVDgYwxRI+9BpevfFs5p7cf0cHXqnbZcLi7ERt+c99rJYeLVlSx/ptppK4tP0mGAcAACotprL8/g/S9tdYtkBwptENMCCEdSLS4wAvWHUM2qsZb3WXHJzOOR/fw1+BNzyO5dge51DwMQwcv2OAJHocvchoWm6/qa3vDY7A5I0iu3UI6MQVdKw36sLui8EVOw+0fBuPaf5Ifaa7ER3NQNqv3+U7fXIXgsMAxUP09SMh+UUJMCCEdiHE8PMET2Fy9aSreF65cv8sYg9PbD6e3H0BpwMfTQz1qsToCiA5+BZGBV6BrMhgnlE3ACQGA4noOxbUyfbDLyNxbo4SYNByVTBBCSIfyR8+CfzReuRrB4oQ3PGb6detNhp/+Wl6wUTJMqspNm+9JraSKKK6bS54J2StKiAkhpEOJVjd6jn8bvGCtGCNYnOgd/Q6NsSVtRUmbK/XZotYZT0i96BaeEEI6mN3dhcHTP0Ry7TZSG/egKqXRzaLFBW/kFHyhky0dHkJIOXWPaKGZLqTBKCEmhJAOJ1icCPVeQbDnMjS1CAaAE2xVOz8Q0kqizwY5Uf1A3Y54P93UkcaihJgQQg4JxjgItBtMDpChGygspFCYS0GXVTCRhy3mgWPAB07Ye9WlcySI3PSmqVhLwA4LJcSkwSghJoQQQsguxfUcEh/NQZPUHY9LSxmkbyzDd6kHjv69dX+w+O2w93hQWEzXjPWcie7pGoTUgxJiQgghpMNpkgpD1cBZBHCW+ns+a5KK/Mwm5EQeMABwDIW5JFCh7EZXdCQ+ngfjGOy93j2t2f98H/SrsyiuZssHMMB/qQe2bveeXp+QelBCTAghhHQgQzeQn00iNxXfUY9r63bDdTxoKpE0DAPpmyvITsZh6E8GcMuJPAxVB2cVIHisFevRN79Ygi3mAePqr1fnBA6hVwdRWEghN5V43FqNCRwcgz64RoIQvdQdhTQHJcSEEEJIhzE0HfGrs5BWdu+uSssZSMsZuEaD8D0Tq/o6yc8Xd9Xy6ooGQ9VL/11UoSQNiL7yhzR1SUVhIbXn0gnGMTj6fXD0+2DoBgxNBxM4OhBKmo76EBNCCCEdZvPzxbLJ8HbZyTgyExsVny+uZcsebNNlbcf/G4oGvaDuitsiLWdqrNYcxjFwIk/JMGkJSogJIYSQDqLmZOQfJk3FZsbXd5RCbJedqjAtztgdrxWUitfY2k0mpJNRQkwIIYR0kNwD82OPdUmt2MlBWir/OON2pwaGpldMfDkbVV+SzkcJMSGEENJB6h1jrKakXY+V6nXL7xxz1vJdKowyO8cA4Bjw17UeQtoRJcSEEELIYVamJJdxDEwsnwIwngNn3b3rW66ThBiwwxpy7HuJhLQaJcSEEEJIBxF99bUiE33lp7xV6wwhuK1g/JMUgYn8jv8HSqUSwRf66loLIe2KEmJCCCGkgziHAmV3fcvhHSJssfL9iF0jwYpfxzgG0W8HZxMBxsDbxW1PAvYeDyJvDkNwWetZOiFtiyrhCSGEkA7CO0Q4hwPIVeoSsY37VKRiGzPRZ4P3fBdSN1bKPs84BtFjhb0vDFuXG7qsgYkc7N0e8A6x7NcQ0qkoISaEEEI6jO+ZGPSihsJ8qmKM50wEruFA1ddxj4XBWQWkb69Cy+9srcZZeLhGg1WTakIOC0qICSGEkA7DOIbAC32QBnzI3o+juPpoSAfH4Oj1wHk8CGvIaeq1nMf8cAz6IC1noCQKMAxAcFvg6PPuqhsm5LCihJgQQgjpQIwx2Hs8sPd4Sn2CNR1M4Mt2gzD1WjEP7DFPA1ZKSPujhJgQQgjpcIznaDeXkH2gfz2EEEIIIeRIo4SYEEIIIYQcaZQQE0IIIYSQI40SYkIIIYQQcqRRQkwIIYQQQo40SogJIYQQQsiRRm3XCCGEEBOUtIT8zCbUrAzGMVjCTjgGfOBEvtVLI4TsEyXEhBBCSBW6rCLxyQKk5cyOx/NzKaRurMBzOgL3WLhFqyOEHARKiAkhhJAKdEXD+vszUJJS2ecNVUfqxgoMVYfnTLTJqyOEHBSqISaEEEIqyIxvVEyGt0vfWYOSLjZhRYSQRqAdYkIIIQTA7EoayXQRHAN6g0547CKyD+Kmvz43FYfv2VgDV0gIaRRKiAkhhBxpf7i7gqs3lpBK5NGnAX0asGkAbgsPlwGITgt4uwjGsaqvU1hKU0JMSIeihJgQQsiR9ZMPp/HxrWV4dAMvK4DFePKcVNQgGgD0IrSCAtFrq9pRwlD0xi+YENIQVENMCCHkSPpyYg0f31qGxTBw6alkGACMRxvCsqrD0AwoKQmGVjnp5az/fzv30xpXFcdx+Hvnz83M5MaZtvlTSSNqBzeaKupafAdusnLtK/BFunOhOK5EQZoqlEJLW03TzFwXaW3QNhY6CTbneVYzcDmc5Ycf5xzPr8HrShADUKRvvr+VJHlr/u8YTpKjJIskbZvMF22yaDP/8+iF6w13JmezUeDMCWIAivP7nYe5dfth0rbZmb/4u4MnU+L5k8nw/OBx2vY59dyp0ly/fAY7Bc6DIAagOHcfHD+RVidZeU7fPnVQJY+r5O9PFu2JP89c+nQ73VF/2dsEzolLdQAUp9d9+XnQ/SppqiqDqjo+P9G2SY5Hx721OuMbVzO8Nj6jnQLnQRADUJxrm03qfieHh/M8qk6fEidJhr3Uk2HaeZu19zdTdTpZ2RhlZatJVZ3+HBvw/+fIBADFGdS9fPTeRlJVufkSj0OMV+tUVZXxja1MPnwz492tDK6uiWG4IAQxAEX6/OOdNKN+fu0mh6d0bTPqZ6XupVN301y/cn4bBM6NIAagSJO1lXz1xQcZjQf5tn98ee6fmlE/G5NhOnU365+97eIcXFDOEANQrM1Lo3z95Sf58ec7+e6H3zK4/Ue2DhdZ63XzxmqdQVNn9d3LaaZXxDBcYIIYgKJ1O1V2p+vZna4nSdpFm8Wjo6RNOoNeqo5zwnDRCWIAOKHqVOkOTYOhJM4QAwBQNEEMAEDRBDEAAEUTxAAAFE0QAwBQNEEMAEDRBDEAAEUTxAAAFE0QAwBQNEEMAEDRBDEAAEXrLXGtaZLMZrPs7e0tcVkAAHhmNps9/TldxnpV27bLWCfb29t3k4yXshgAAPy3e/v7+5NXXWSZE+JfkryT5EGSn5a4LgAAnDRN0uS4P1/Z0ibEAADwOnKpDgCAogliAACKJogBACiaIAYAoGiCGACAogliAACKJogBACiaIAYAoGiCGACAogliAACKJogBACiaIAYAoGiCGACAogliAACKJogBACiaIAYAoGh/ATCROM7FrW4GAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<Figure size 864x576 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "sns.set_context('poster')\n",
    "sns.set_color_codes()\n",
    "plot_kwds = {'marker': 'o', 'alpha' : 0.75, 's' : 150, 'linewidths':0}\n",
    "\n",
    "labels = clusters\n",
    "palette = sns.color_palette(palette='deep', n_colors=np.unique(labels).max() + 1)\n",
    "colors = [palette[x] if x >= 0 else (0.75, 0.75, 0.75) for x in labels]\n",
    "\n",
    "fig, ax = plt.subplots(figsize=(12, 8))\n",
    "ax.margins(0.05)\n",
    "\n",
    "plt.scatter(*projection.T, c=colors, **plot_kwds)\n",
    "frame = plt.gca()\n",
    "frame.axes.get_xaxis().set_visible(False)\n",
    "frame.axes.get_yaxis().set_visible(False)\n",
    "\n",
    "\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### DBSCAN\n",
    "We can also try using Density-based spatial clustering of applications with noise (DBSCAN) to cluster the speeches. Note that DBSCAN doesn't require globular clusters, so we should expect a slightly different clustering."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "First, let's calculate the cosine distance between each document. Let's use cosine distance instead of Euclidean distance as the measure of \"closeness\" between speeches. This is a common practice in NLP."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [],
   "source": [
    "dist = 1 - cosine_similarity(tfidf_matrix)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "The average cosine distance between speeches is 0.8026433604271683\n"
     ]
    }
   ],
   "source": [
    "print(\"The average cosine distance between speeches is {0}\".format(np.average(dist)))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Now we can actually run the model, using our precomputed distance metrics. We can set epsilon (`eps`) to be 0.5 which is less than the average distance we calculated above. We'll also require that at least 3 speeches be in a cluster."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [],
   "source": [
    "model = DBSCAN(eps=0.5, min_samples=3, metric='precomputed')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [],
   "source": [
    "labels = model.fit_predict(dist)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Let's see how many clusters were found and how many speeches were left unclustered:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "DBSCAN found 7 clusters\n",
      "96 speeches were considered 'noise' and were not clustered.\n"
     ]
    }
   ],
   "source": [
    "if -1 in labels:\n",
    "    num_clusters = len(set(labels)) - 1\n",
    "else:\n",
    "    num_clusters = len(set(labels))\n",
    "    \n",
    "num_noise = len([i for i in labels if i == -1])\n",
    "    \n",
    "print(\"DBSCAN found {0} clusters\".format(num_clusters))\n",
    "print(\"{0} speeches were considered 'noise' and were not clustered.\".format(num_noise))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Visualizing DBSCAN results:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {
    "scrolled": false
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 864x576 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "palette = sns.color_palette(palette='deep', n_colors=np.unique(labels).max() + 1)\n",
    "colors = [palette[x] if x >= 0 else (0.75, 0.75, 0.75) for x in labels]\n",
    "\n",
    "fig, ax = plt.subplots(figsize=(12, 8))\n",
    "ax.margins(0.05)\n",
    "\n",
    "plt.scatter(*projection.T, c=colors, **plot_kwds)\n",
    "frame = plt.gca()\n",
    "frame.axes.get_xaxis().set_visible(False)\n",
    "frame.axes.get_yaxis().set_visible(False)\n",
    "\n",
    "\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### HDBSCAN\n",
    "We can use the hierarchical form of DBSCAN, which doesn't require we specify a threshold for distance (`epsilon`) and see how it performs as well."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Here we can set the minimum number of speeches in a cluster. Let's select 4, which would dictate that we have at least 4 speeches in a cluster. `min_samples` is a measure of how conservative we would like the clustering to be--setting this value to `1` means we will not look at just high density clusters."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [],
   "source": [
    "model = hdbscan.HDBSCAN(min_cluster_size=4, min_samples=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [],
   "source": [
    "labels = model.fit_predict(dist)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "HDBSCAN found 18 clusters\n",
      "54 speeches were considered 'noise' and were not clustered.\n"
     ]
    }
   ],
   "source": [
    "if -1 in labels:\n",
    "    num_clusters = len(set(labels)) - 1\n",
    "else:\n",
    "    num_clusters = len(set(labels))\n",
    "    \n",
    "num_noise = len([i for i in labels if i == -1])\n",
    "    \n",
    "print(\"HDBSCAN found {0} clusters\".format(num_clusters))\n",
    "print(\"{0} speeches were considered 'noise' and were not clustered.\".format(num_noise))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Visualizing HDBSCAN results"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {
    "scrolled": false
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 864x576 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "palette = sns.color_palette(palette='deep', n_colors=np.unique(labels).max() + 1)\n",
    "colors = [palette[x] if x >= 0 else (0.75, 0.75, 0.75) for x in labels]\n",
    "\n",
    "fig, ax = plt.subplots(figsize=(12, 8))\n",
    "ax.margins(0.05)\n",
    "\n",
    "plt.scatter(*projection.T, c=colors, **plot_kwds)\n",
    "frame = plt.gca()\n",
    "frame.axes.get_xaxis().set_visible(False)\n",
    "frame.axes.get_yaxis().set_visible(False)\n",
    "\n",
    "\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": "3.7.0"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}