twitter-archive-text-mining-R.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Twitter archive analysis in R\n",
    "This performs some twitter archive analysis that is [based on this chapter of *Text Mining with R*](https://www.tidytextmining.com/twitter.html). To use this notebook you need to have uploaded a Twitter archive into your *Open Humans* account through http://twarxiv.org. \n Initially, what you'll see is data I've supplied, and as you run each command, it'll be replaced with your data.\n",
    "\n",
    "For a start let's install/load all the required packages again:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n",
      "Attaching package: ‘lubridate’\n",
      "\n",
      "The following object is masked from ‘package:base’:\n",
      "\n",
      "    date\n",
      "\n",
      "\n",
      "Attaching package: ‘dplyr’\n",
      "\n",
      "The following objects are masked from ‘package:lubridate’:\n",
      "\n",
      "    intersect, setdiff, union\n",
      "\n",
      "The following objects are masked from ‘package:stats’:\n",
      "\n",
      "    filter, lag\n",
      "\n",
      "The following objects are masked from ‘package:base’:\n",
      "\n",
      "    intersect, setdiff, setequal, union\n",
      "\n"
     ]
    }
   ],
   "source": [
    "library(purrr)\n",
    "library(stringr)\n",
    "library(tidytext)\n",
    "library(widyr)\n",
    "library(httr)\n",
    "library(lubridate)\n",
    "library(ggplot2)\n",
    "library(dplyr)\n",
    "library(readr)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Let's now get our access token and request our personal user object that will contain all of our data file downloads etc.:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "access_token <- Sys.getenv(\"OH_ACCESS_TOKEN\")\n",
    "url <- paste(\"https://www.openhumans.org/api/direct-sharing/project/exchange-member/?access_token=\",access_token,sep=\"\")\n",
    "resp <- GET(url)\n",
    "user <- content(resp, \"parsed\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "If you want to look at the data sources you have on Open Humans, uncomment the line below by removing the #:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# user$data"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Let's now find the download URL for the Twitter archive from all files:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "for (data_source in user$data){\n",
    "    if (data_source$source == \"direct-sharing-70\"){\n",
    "         twitter_archive_url <- data_source$download_url\n",
    "    }\n",
    "}"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We can now create a temporary file that will contain the whole zipped Twitter archivee, from this we can then unzip and read the `tweets.csv` file:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Parsed with column specification:\n",
      "cols(\n",
      "  tweet_id = col_double(),\n",
      "  in_reply_to_status_id = col_double(),\n",
      "  in_reply_to_user_id = col_double(),\n",
      "  timestamp = col_character(),\n",
      "  source = col_character(),\n",
      "  text = col_character(),\n",
      "  retweeted_status_id = col_double(),\n",
      "  retweeted_status_user_id = col_double(),\n",
      "  retweeted_status_timestamp = col_character(),\n",
      "  expanded_urls = col_character()\n",
      ")\n"
     ]
    }
   ],
   "source": [
    "temp <- tempfile()\n",
    "download.file(twitter_archive_url,temp,method='wget')\n",
    "\n",
    "#unzip(temp, list=TRUE) # this would list all files in the zip archive\n",
    "\n",
    "data <- read_csv(unz(temp, \"tweets.csv\"))\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Now let's convert the timestamps into a proper format and plot a simple histogram of tweets over time:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "data": {},
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAA0gAAANICAMAAADKOT/pAAACvlBMVEUAAAABAQECAgIDAwME\nBAQFBQUGBgYHBwcICAgJCQkKCgoLCwsMDAwNDQ0ODg4PDw8RERETExMUFBQVFRUWFhYXFxcY\nGBgaGhobGxscHBwdHR0eHh4fHx8hISEiIiIkJCQmJiYnJycoKCgpKSkqKiosLCwtLS0uLi4v\nLy8xMTEyMjIzMzM0NDQ1NTU2NjY3Nzc4ODg5OTk6Ojo7Ozs8PDw9PT0 Pj5AQEBBQUFCQkJD\nQ0NERERFRUVGRkZHR0dISEhLS0tNTU1OTk5PT09QUFBRUVFSUlJTU1NUVFRVVVVWVlZXV1dY\nWFhZWVlaWlpbW1tcXFxdXV1eXl5fX19gYGBhYWFiYmJjY2NkZGRlZWVmZmZnZ2doaGhpaWlq\nampra2tsbGxtbW1vb29wcHBxcXFycnJzc3N0dHR1dXV2dnZ3d3d4eHh5eXl6enp7e3t8fHx9\nfX1 fn5/f3 AgICBgYGCgoKDg4OFhYWGhoaHh4eIiIiKioqLi4uMjIyNjY2Ojo6Pj4 QkJCR\nkZGSkpKTk5OWlpaXl5eYmJiZmZmampqbm5ucnJydnZ2enp6fn5 goKChoaGjo6OkpKSlpaWm\npqanp6eoqKipqamrq6usrKytra2urq6vr6 wsLCxsbGysrKzs7O0tLS1tbW2tra4uLi5ubm6\nurq7u7u8vLy9vb2 vr6/v7/AwMDBwcHCwsLDw8PExMTFxcXGxsbHx8fIyMjKysrLy8vMzMzN\nzc3Ozs7Pz8/Q0NDR0dHS0tLT09PU1NTV1dXW1tbX19fY2NjZ2dna2trb29vc3Nzd3d3e3t7f\n39/g4ODi4uLj4 Pk5OTl5eXm5ubn5 fo6Ojp6enq6urr6 vs7Ozt7e3u7u7v7 /w8PDx8fHy\n8vLz8/P09PT19fX29vb39/f4 Pj5 fn6 vr7 /v8/Pz9/f3 /v7///9jWOyjAAAACXBIWXMA\nABJ0AAASdAHeZh94AAAgAElEQVR4nO3d/YMd5Xmf8RM71AEH144TUztunJc6rkPsOia2E6d1\nu3qxbGwks4IYkbQipkALNqmLQkUgRIntxmrcloY6kCYkiKg2pClJMeAY1cGmGFABYSrPvq92\npd2VNP9Fz/2dPaPD7szuzP3MObOzc31 0JzdPZqZ8zz3pd0VkujEAIJ16r4BYDMgJKAChARU\ngJCAChASUAFCAipASEAFCAmowOBDOjsxG3qKqdATnJg4E3iGucXAE5yaOBV4hqUTgSeIJ6ZD\nzxC8E7MTZwPPMLcQeILFifnAM5yZWf2 IYQUBW/feOgJZqLTgWc4Ebp9J6OTgWdYzNi csYm\nQ88QvBPTUWhIJ0J/QVqI5gLPcDrj1xNCKoaQhJAMIfkRkhCSISQ/QhJCMoTkR0hCSIaQ/AhJ\nCMkQkh8hCSEZQvIjJCEkQ0h hCSEZAjJj5CEkAwh RGSEJIhJD9CEkIyhORHSEJIhpD8CEkI\nyRCSHyEJIRlC8iMkISRDSH6EJIRkCMmPkISQDCH5EZIQkiEkP0ISQjKE5EdIQkiGkPwISQjJ\nEJIfIQkhGULyIyQhJENIfoQkhGQIyY QhJAMIfkRkhCSISQ/QhJCMoTkR0hCSIaQ/AhJCMkQ\nkh8hCSEZQvIjJCEkQ0h hCSEZAjJj5CEkAwhLbtsfSvPQEhCSIaQlhGSFyEZQlpGSF6EZAhp\nGSF5EZIhpGWE5EVIhpCWEZIXIRlCWkZIXoRkCGkZIXkRkiGkZYTkRUiGkJYRkhchGUJaRkhe\nhGQIaRkheRGSIaRlhORFSIaQlhGSFyEZQlpGSF6EZAhpGSF5EZIhpGWE5EVIpiUhFciEkJwI\nyRASIQUiJENIhBSIkAwhEVIgQjKEREiBCMkQEiEFIiRDSIQUiJAMIRFSIEIyhERIgQjJEBIh\nBSIkQ0iEFIiQDCERUiBCMoRESIEIyRASIQUiJENIhBSIkAwhEVIgQjKEREiBCMkQEiEFIiRD\nSIQUiJAMIRFSIEIyhERIgQjJEBIhBSIkM8iQzizlWYymcj9W0HipZ1cS0sqTzswHvoa5aC7w\nDCenA0 wNDYReoZyO5FhKloMPMNM6DrOR7OBZ1iY7D0696 8VRXSqZlc0Xj x4oZK/XsSkJa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\nIbkRkhCSEJIbIQkhCSG5EZIQkhCSGyEJIUl7Q3p5/yc  pvdnzh75 6d 46vPq6LkISQpLUh\nLf7Krcee//VPx/GtNx596Y69Z1Yd10VIQkjS2pCeHhmL42jk WjLc93PQtuOrDyuf1VCEkKS\n1ob05Eh3 05ve/CRHbYA19yz8rj VQlJCElaG9L8rt9bWrp7258cvsLeuvngymPveYvzuaKJ\n/I8VM1bq2ZWEtPKkU7OBr2Emmg48w xk4Anmx8aDzxB6goko9AxTM4EnmI2mAs9wIp3oc784\nrvubDX979baP3X31nx2 cjmgFcfe0 aiDaOSkOp EWiEieIhdT XLi1t/9qjyZdy96489p50\nejHPQjSV 7GCxks9u5KQVp50Zj7wNZyITgSeYT54HccmQs9QbicyTEULgWeYmQs8wVw0E3iG\nU5Ppw8IhnX64G91j26bHtzzT/QJ365Mrj t3yPdIwvdI0trvkeJP7Y e3P2FOL7t2qPHbrnu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 kWELkMF\n6zjkndigZ6hwJ879ysRnJK Sr4HPSMJnJDdCEkISQnIjJCEkISQ3QhJCEkJyIyQhJCEkN0IS\nQhJCciMkISQhJDdCEkISQnIjJCEkISQ3QhJCEkJyIyQhJCEkN0ISQhJCciMkISQhJDdCEkIS\nQnIjJCEkISQ3QhJCEkJyIyQhJCEkN0ISQhJCciMkISQhJDdCEkISQnIjJCEkISQ3QhJCEkJy\nIyQhJCEkN0ISQhJCciMkISQhJDdCEkISQnIjJCEkISQ3QhJCEkJyIyQhJCEkN0ISQhJCciMk\nISQhJDdCEkISQnIjJCEkISQ3QhJCEkJyIyQhJCEkN0ISQhJCciMkISQhJDdCEkISQnIjJCEk\nISQ3QhJCEkJyIyQhJCEkN0ISQhJCciMkISQhJDdCEkISQnIjJCEkISQ3QhJCEkJyIyQhJCEk\nN0ISQhJCciMkISQhJDdCEkISQnIjJCEkISQ3QhJCEkJyIyQhJCEkN0ISQhJCciMkISQhJDdC\nEkISQnIjJCEkISQ3QhJCEkJyIyQhJCEkN0ISQhJCciMkISQhJDdCEkISQnIjJCEkISS3zRpS\nAf13RUhCSG6EJIQkhORGSEJIQkhuhCSEJITkRkhCSEJIboQkhCSE5EZIQkhCSG6EJIQkhORG\nSEJIQkhuhCTZIZU5CSEJIfkRkiEkISQ/QjKEJITkR0iGkISQ/Pq2b DDX6X 10BIQkhuhCSE\nJITkRkhCSEJIboQkhCSE5EZIQkhCSG6EJIQkhORGSEJIQkhuhCSEJITkRkhCSEJIboQkhCSE\n5EZIQkhCSG6EJIQkhORGSEJIQkhuhCSEJC0O6YGrtl/zWBzP3rl7577jq4/rIiQhJGlvSA O\nPn78vj1z8a03Hn3pjr1nVh3XRUhCSNLekPY8pEO05bnuZ6FtR1Ye178qIQkhSWtDGht56Nc \ncv1T8SM7bAGuuWflsfe8M0t5FqOp3I8VNH7u4cCHv0r9r2Eumst4ZWVOcnI6dB3HJkLPML7 \nU9Y2FS0GnmEmax3LmI9mA8 wMNl7dLpwSE PfPrFmYMfnzp8hb1188GVx97z5qLhGPjwV2lI\nL2fwq44cEyVC6n71dvqyBw9fuRzQimPveYsnckUT R8rZuzcw0omb1j6X8NUNJXxyiq/zNrr\nOO7dgfQMoSeYiELPMDkdeILpaDLwDLPpOs4XDikaeab74957H02 lFt1XOenx3yPtMz9PVK5\ny6yJ75FMPd8jnRm9v3vpSx8e39INanrrkyuP61 VkISQpLUhxffueiL63OjJ LZrjx675bqz\nq47rIiQhJGlvSGfuunz7TS/E8dyB0V37J1Yf10VIQkjSjpAu/nZy/KOfCrxSP0ISQpJ2hNR5\nXIelfX8n8Er9CEkISdoQUuecnwm8Uj9CEkKSNoR05Hc7Wz9prvrsi4FX6kdIQkjShpDi EPf\nCbxCFkISQpJ2hDQQhCSEJO0I6fjui16VfJMUeKV hCSEJO0I6aPf/8Hd i7pk4FX6kdIQkjS\njpDecF/gFbIQkhCStCOkC74XeIUshCSEJO0I6X3/I/AKWQhJCEnaEdLX3/1I4CUyEJIQkrQj\npEve0rngrRJ4pX6EJIQk7QjpfR/sCbxSP0ISQpJ2hDQQhCSEJITkRkhCSNKOkN7Qc2HglfoR\nkhCStCOkrfLu89  N/BK/QhJCEnaEdKyl99/KPBK/QhJCElaFVL8 MWBV pHSEJI0q6QXj4/\n8Er9CEkISVoV0tl/9 bAK73ibIRkCEnaEdI/lLf/UOeGwCv1IyQhJGlTSO/8xd9dCLxSP0IS\nQpJ2hDQQhCSEJG0JaezQwd8/HPw/tHoFQhJCknaEdOb68 wfbHjt7YEXegVCEkKSdoR0e2f7\nl/780Bc/1Lkr8Er9CEkISdoR0k9dlxyv5l9arUT/ayAkaUdIr3koOT5wfuCV hGSEJK0I6TX\n3p8c7/vBwCv1IyQhJGlHSD//C/oPSCf/yQcCr9SPkISQpB0hPfB9P/qrt/7Gnote9dXAK/Uj\nJCEkaUdI8Z/8pP32908/EHihVyAkISRpSUhx/NJjj/ /wMusQEhCSNKSkF7 XPeH7 07Hnih\nVyAkISRpR0j/5032/7x8vvOm5wKv1I QhJCkHSFte9tjdvj22z4ceKV hCSEJO0I6Y3/MTl \nkX9FqBL9r4GQpB0hnf8HyfHuCwKv1I QhJCkHSG990On7TDzrksCr9SPkISQpB0hHf6 H9t7\ny2evfOOrDgdeqR8hCSFJO0KKv3Kx/QfZd/AfZKvR/xoISVoSUnetv/mtav CLCElCElaE1L1\nCEkISQjJjZCEkISQ3AhJCEkIyY2QhJCEkNwISQhJCMmNkISQhJDcCEkISQjJjZCEkISQ3AhJ\nBhhS4TshJENIZliTV4n 10BIQkhuhCSEJITkRkhCSEJIboQkhCSE5EZIQkhCSG6EJIQkhORG\nSEJIQkhuhCSEJITkRkhCSEJIboQkhCSE5EZIQkhCSG6EJIQkhORGSEJIQkhuhCSEJITkRkhC\nSEJIboQkhCSE5NbikDaOZPEIyRCSqXsiGypZPEIyhGTqnsiGShaPkAwhmbonsqGSxSMkQ0im\n7olsqGTxCMkQkql7IhsqWTxCMoRk6p7IhkoWj5AMIZm6J7KhksUjJENIpu6JbKhk8QjJEJKp\neyIbKlk8QjKEZOqeyIZKFo QDCGZuieyoZLFIyRDSKbuiWyoZPEIyRCSqXsiGypZPEIyhGTq\nnsiGShaPkAwhmbonsqGSxSMkQ0im7olsqGTxCMkQkql7IhsqWTxCMoRk6p7IhkoWj5AMIZm6\nJ7KhksUjJLOxQzo1kysaz/9YMWPnHtY9kQ21vI5V7oTPeBR6honJwBNMRROBZ5hO13G28pBO\nL ZZiKZyP1bQ2LmHdU9kQy2v40SFO EzFS0EnmFmLvAEc9FM4BlOTaYPKw8pH1/abQDJ4vGl\nndnYX9rlI6QNIFk8QjKEZOqeyIZKFo QDCGZuieyoZLFIyRDSKbuiWyoZPEIyRCSqXsiGypZ\nPEIyhGTqnsiGShaPkAwhmbonsqGSxSMkQ0im7olsqGTxCMkQkql7IhsqWTxCMoRk6p7IhkoW\nj5AMIZm6J7KhksUjJENIpu6JbKhk8QjJEJKpeyIbKlk8QjKEZOqeyIZKFo QDCGZuieyoZLF\nIyRDSKbuiWyoZPEIyRCSqXsiGypZPEIyhGTqnsiGShaPkAwhmbonsqGSxSMkQ0im7olsqGTx\nCMkQkql7IhsqWTxCMoRk6p7IhkoWj5AMIZm6J7KhksUjJENIpu6JbKhk8QjJEJKpeyIbKlk8\nQjKEZOqeyIZKFo QDCGZuieyoZLFIyRDSKbuiWyoZPEIyRCSqXsiGypZPEIyhGTqnsiGShaP\nkAwhmbonsqGSxSMkQ0im7olsqGTxCMkQkql7IhsqWbyBh1TgTgjJi5A2gGTxCMkQkhn4yG1O\nyeIRkiEkM/CR25ySxSMkQ0hm4CO3OSWLR0iGkMzAR25zShaPkAwhmYGP3OaULB4hGUIyAx 5\nzSlZPEIyhGQGPnKbU7J4hGQIyQx85DanZPEIyRCSGfjIbU7J4hGSISQz8JHbnJLFIyRDSGbg\nI7c5JYtHSIaQzMBHbnNKFo QDCGZgY/c5pQsHiEZQjIDH7nNKVk8QjKEZAY cptTsniEZAjJ\nDHzkNqdk8QjJEJIZ MhtTsniEZIhJDPwkducksUjJENIZuAj11aldyJLgesQkhchNULpnchS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\nHVl5XP qhNQIpXciyzDupKEhmZkv/PPFR3bY77Zcc8/K4/o/m5AaofROZBnGnTQ2pDMfHvn0\nWHz4Cnt888GVx96z5sZyRVH x4rpO0ElW4XVSu9ElmHcSfA0VTmQkyVCil/85m1Xzx6 cjmg\nFcfek YnckVj R8rpu8ElWwVViu9E1mGcSfj48VuNf8EFQ7kuU9NhX77 8zHDj2afCl378rj\n j ZL 0aofROZBnGnTT0S7tv7One99ldh8a3PBPH01ufXHlc/6qE1AildyLLMO6koSHNXv6b\nL7x8cMfL8W3XHj12y3VnVx3XRUiNUHonsgzjThoaUvz8v/3opTccieO5A6O79k sPq6LkBqh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\n0uydu3fuO17giYSEhsob6UpDuvXGoy/dsfcMIWHTGkZI0Zbnup Vth0hJGxawwjpkR1nuz9e\ncw8hYdMaRkiHr7Afbz7Ye/vUTK5oPDnWvSxAOXkTPT3eezQbHtKVrwxpLgJaZyI4pEeTL 3u\n7b199nSepWgq92MFjYWeYDpaCDzDzMnAE8xFc4FnODUdeILTYxPBZwg9wVS0FHiGmfnAE5yM\nZgPPsDDZe3TuN9u8IY1veSaOp7c uf4zy/yTxXkXCz3BTHQ68AwnFgJPcDI6GXiGxZnAE8Rj\nk6FnCN6J6ehs4BlOnAo8wUI0F3iGSn/7 7Zrjx675boCq0JIQkhCSCvNHRjdtX9i/ecRUoKQ\nhJDcCEkISQjJjZCEkISQ3AhJCEkIyY2QhJCEkNwISQhJCMmNkISQhJDcCEkISQjJjZCEkISQ\n3AhJCEkIyY2QhJCEkNwISQhJCMmNkISQhJDcCEkISQjJjZCEkISQ3AhJCEkIyY2QhJCEkNwI\nSQhJCMmNkISQhJDcCEkISQjJjZCEkISQ3AhJCEkIyY2QhJCEkNwISQhJCMmNkISQhJDcCEkI\nSQjJjZCEkISQ3AhJCEkIyY2QhJCEkNwISQhJCMmNkISQhJDczp4InaB4PvQEp06Ebt9CaIlL\nJ5YCz3A6dILiueCFDD7ByfCdCF7HE4uBZzibMdGDDwloAUICKkBIQAUICagAIQEVICSgAoQE\nVICQgAoMJKTxOz5x6U1Px/Hsnbt37jt 7vjib z62E3fGsQV172F Nj1W 396dv13ULv/XXe\nQ9eDI1 r9xYeuGr7NY8N5RZy76HCgRxISP/yxue  1u7Tsa33nj0pTv2nukdz 75/NypP/ho\n8J908dzCw6MHtHS9t2u8hd7767yHOJ68fMdwQsq7hQdHHz9 357QP60TdA9VDuQgQprZ/0Ic\nf2/kO9GW57q/CGw70jtOjTwVxxMjw/jFeOUtxA9972u2dOnb9d1C7/1DuIXce i67UuXDyWk\n3FvY89AwLr/mPVQ5kAP7HumprROP7LA/V3XNPb1j/K8OzJz8r1eF/vFPzy10f9DS9b1d1y30\n3j sW8i h0euOjmckPJuYWzkoV/7yPVPDe0WMpehwoEcVEgz/ I/xYevsEc3H wd4/G9IyOj\nzw7oimveQry8dH1v13ULvfcPS Y9zI4 EQ8vpKxbeHrk0y/OHPx4xp jHt49VDmQAwrpxau/\ncDY fMwvhY0AAAQ/SURBVKU97Ia0fFz61Oen5u7dNaRfi19xC3EvpHNv13ULvfcPSfY9/M7v\nxMMLKfMWnh7pfn19 rIH67yHKgdyMCEd2Xl/98dHk8 k9/aO39hi32H/8p8N5JJr30K8vHR9\nb9d1C733D0f2PTwxOjO8kLJvIRp5pvvj3uHsRM49VDmQAwnpW5d93Q7jW7prNb31yd7xb0bs\n92hGhxLSiluIl5eu7 26bqH3/qHIuYfbd zcuXPLpftrvIUzo93RXrj04WHcQt49VDmQgwhp\nYc Xo66T8W3XHj12y3Vne8e50c/PLvy3Hd8dwCXXvYWJ6Ctb 9 u7xbS99d4DzP23k98Jfiv\nLgfcQnzvrieiz43WugxVDuQgQjoyIofiuQOju/Z3vwLtHZ/ft vj//qbA7ji rfwSb39p nb\n9d1C v4a70EfHM6Xdrm3cOauy7ff9MIwbiH/HiocSP6IEFABQgIqQEhABQgJqAAhARUgJKAC\nhARUgJCAChASUAFC2uD 0U/UfQcogpA2ridscw4U pOlT7CPNWMDNq7PFd cEk/FQLABG9aH\nOp3OxfrS7n0///C7fuCi2xdvvOgHP/hc9yN/ UsXnv/OL3UffPeqH33Nj3z4qeWnxl9 1/kX\nXvzleNVP Jn3PPSu819/5dD NmobEdKG9Z2tnce/rZA  OZf JsXt3d ad x//m6fxbHD776\n/fd/5Vc7vxXH73nT7//F3T/9w3PJU/ ws/3QoX/aObTqJ/zcG3/2f0X/5bztdb izYyQNq5P\n2uYopM6ROP6rznu7b 56bRy/823299G2XHhyunNT98Gz 19Knrr/FxfiePr7d636CZd0Htbp\nhvOXFtqJkDaucyF1Y4if7dzQ/fGGzszxzqdOdv1e57HFN7z1wTPnnpp48/tW/oT4ktfa32S8\nq/Pnw38NrUFIG9e5kN7affB/O7d1f7yxM/lEZ9kfx3/99ztv2HH30vJTp//N21/36ld3Lln5\nE JLfsxOd6gzvH 6qH0IaePKDemXvyZRHJ9 6IZ/0PnZ eSp73/1Zx7 5t9elBfSfZ276nst\nmx4hbVw5IY13dr/iaV/o/Gc99ZnOnu5bSz QEdL59r S/mLn8LBfQYsQ0sZ1VWcpK6T43X93\nsvvorpuXvv4x wfhn 3coad u7Mvtv i9J7VIem7o22vGd6/7to hLRxfbaz74 yQvrL895x\n13//9fOuiF  8B1f uofvvd1z qpi2/5e3/619d/4AMX/sWJlSG95cf//Vdv6IzW 3I2N0La\nuF5853k/kRVS/Ff/ MLzfvz27qer/739h8 7aPs3lp/6 M9d8CO/Mn3/D73 6ZUh/eTX33/B\n6/fM1vtyNjdCaoFL IOvA0dILUBIg0dILUBIg0dILUBIg0dIQAUICagAIQEVICSgAoQEVICQ\ngAoQElABQgIqQEhABQgJqMD/B27HSmjpudcFAAAAAElFTkSuQmCC",
      "text/plain": [
       "plot without title"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "tweets <- mutate(data,timestamp = ymd_hms(timestamp))\n",
    "ggplot(tweets, aes(x = timestamp))  \n",
    "  geom_histogram(position = \"identity\", bins = 20, show.legend = FALSE)   theme_minimal()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We can now 'tokenize' (that is, break up into words) the tweet texts, which will make it easier to work with them. This also allows us to easily calculate word frequencies in the next step. This stage also passes words through a stopwords filter. Stopwords are those words which are frequently used (such as 'the', 'we', 'and', and 'I') but provide very little information, and it's common to filter them out during a textual analysis. Stopwords are language dependent, so you may want to change the language default. You can read more about the stopwords function here: https://www.rdocumentation.org/packages/tm/versions/0.7-3/topics/stopwords"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "replace_reg <- \"https://t.co/[A-Za-z\\\\d] |http://[A-Za-z\\\\d] |&amp;|&lt;|&gt;|RT|https\"\n",
    "unnest_reg <- \"([^A-Za-z_\\\\d#@']|'(?![A-Za-z_\\\\d#@]))\"\n",
    "tidy_tweets <- tweets %>% \n",
    "  filter(!str_detect(text, \"^RT\")) %>%\n",
    "  mutate(text = str_replace_all(text, replace_reg, \"\")) %>%\n",
    "  unnest_tokens(word, text, token = \"regex\", pattern = unnest_reg) %>%\n",
    "  filter(!word %in% stop_words$word,\n",
    "         str_detect(word, \"[a-z]\"))\n",
    "\n",
    "frequency <- tidy_tweets %>% \n",
    "  count(word, sort = TRUE) \n",
    "frequency$freq <- frequency$n / sum(frequency$n)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "So, what are the top words that you've used? Mine are the German equivalents of `I`, all the articles (`der, die, das`) along with `not` and `is`."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<table>\n",
       "<thead><tr><th scope=col>word</th><th scope=col>n</th><th scope=col>freq</th></tr></thead>\n",
       "<tbody>\n",
       "\t<tr><td>ich       </td><td>17749     </td><td>0.02300768</td></tr>\n",
       "\t<tr><td>die       </td><td>12838     </td><td>0.01664165</td></tr>\n",
       "\t<tr><td>das       </td><td>12081     </td><td>0.01566036</td></tr>\n",
       "\t<tr><td>der       </td><td> 9146     </td><td>0.01185578</td></tr>\n",
       "\t<tr><td>nicht     </td><td> 8669     </td><td>0.01123746</td></tr>\n",
       "\t<tr><td>ist       </td><td> 8501     </td><td>0.01101968</td></tr>\n",
       "</tbody>\n",
       "</table>\n"
      ],
      "text/latex": [
       "\\begin{tabular}{r|lll}\n",
       " word & n & freq\\\\\n",
       "\\hline\n",
       "\t ich        & 17749      & 0.02300768\\\\\n",
       "\t die        & 12838      & 0.01664165\\\\\n",
       "\t das        & 12081      & 0.01566036\\\\\n",
       "\t der        &  9146      & 0.01185578\\\\\n",
       "\t nicht      &  8669      & 0.01123746\\\\\n",
       "\t ist        &  8501      & 0.01101968\\\\\n",
       "\\end{tabular}\n"
      ],
      "text/markdown": [
       "\n",
       "word | n | freq | \n",
       "|---|---|---|---|---|---|\n",
       "| ich        | 17749      | 0.02300768 | \n",
       "| die        | 12838      | 0.01664165 | \n",
       "| das        | 12081      | 0.01566036 | \n",
       "| der        |  9146      | 0.01185578 | \n",
       "| nicht      |  8669      | 0.01123746 | \n",
       "| ist        |  8501      | 0.01101968 | \n",
       "\n",
       "\n"
      ],
      "text/plain": [
       "  word  n     freq      \n",
       "1 ich   17749 0.02300768\n",
       "2 die   12838 0.01664165\n",
       "3 das   12081 0.01566036\n",
       "4 der    9146 0.01185578\n",
       "5 nicht  8669 0.01123746\n",
       "6 ist    8501 0.01101968"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "head(frequency)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Let's now group the tweets by being \"old\" (pre-2013) and more recent ones (2013 or newer) to see whether the topics you tweet about have changed. After grouping you can calulcate the word-ratios and look at the most unusual and the most recent words:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "cutoff_date <- as.Date(\"2013-01-01\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "You can adapt the cutoff above easily to your own needs. Just replace `2013-01-01` with your own date in the `YYYY-MM-DD` format."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<table>\n",
       "<thead><tr><th scope=col>word</th><th scope=col>past</th><th scope=col>today</th><th scope=col>logratio</th></tr></thead>\n",
       "<tbody>\n",
       "\t<tr><td>option         </td><td>8.071251e-05   </td><td>8.071286e-05   </td><td>-4.282714e-06  </td></tr>\n",
       "\t<tr><td>backup         </td><td>4.447424e-05   </td><td>4.439207e-05   </td><td> 1.849286e-03  </td></tr>\n",
       "\t<tr><td>ern            </td><td>4.447424e-05   </td><td>4.439207e-05   </td><td> 1.849286e-03  </td></tr>\n",
       "\t<tr><td>gefallen       </td><td>7.247654e-05   </td><td>7.264157e-05   </td><td>-2.274431e-03  </td></tr>\n",
       "\t<tr><td>timeline       </td><td>1.087148e-04   </td><td>1.089624e-04   </td><td>-2.274431e-03  </td></tr>\n",
       "\t<tr><td>wieviel        </td><td>3.623827e-05   </td><td>3.632079e-05   </td><td>-2.274431e-03  </td></tr>\n",
       "\t<tr><td>haha           </td><td>2.091936e-04   </td><td>2.098534e-04   </td><td>-3.148939e-03  </td></tr>\n",
       "\t<tr><td>extra          </td><td>5.271021e-05   </td><td>5.246336e-05   </td><td> 4.694238e-03  </td></tr>\n",
       "\t<tr><td>spam           </td><td>1.054204e-04   </td><td>1.049267e-04   </td><td> 4.694238e-03  </td></tr>\n",
       "\t<tr><td>videos         </td><td>6.424057e-05   </td><td>6.457028e-05   </td><td>-5.119383e-03  </td></tr>\n",
       "\t<tr><td>lasse          </td><td>6.918215e-05   </td><td>6.860593e-05   </td><td> 8.363967e-03  </td></tr>\n",
       "\t<tr><td>bioinformatiker</td><td>2.800230e-05   </td><td>2.824950e-05   </td><td>-8.789112e-03  </td></tr>\n",
       "\t<tr><td>fliegen        </td><td>2.800230e-05   </td><td>2.824950e-05   </td><td>-8.789112e-03  </td></tr>\n",
       "\t<tr><td>hosen          </td><td>2.800230e-05   </td><td>2.824950e-05   </td><td>-8.789112e-03  </td></tr>\n",
       "\t<tr><td>installation   </td><td>2.800230e-05   </td><td>2.824950e-05   </td><td>-8.789112e-03  </td></tr>\n",
       "\t<tr><td>instapaper     </td><td>2.800230e-05   </td><td>2.824950e-05   </td><td>-8.789112e-03  </td></tr>\n",
       "\t<tr><td>krass          </td><td>2.800230e-05   </td><td>2.824950e-05   </td><td>-8.789112e-03  </td></tr>\n",
       "\t<tr><td>medium         </td><td>2.800230e-05   </td><td>2.824950e-05   </td><td>-8.789112e-03  </td></tr>\n",
       "\t<tr><td>quelle         </td><td>2.800230e-05   </td><td>2.824950e-05   </td><td>-8.789112e-03  </td></tr>\n",
       "\t<tr><td>rper           </td><td>2.800230e-05   </td><td>2.824950e-05   </td><td>-8.789112e-03  </td></tr>\n",
       "\t<tr><td>sequenzen      </td><td>2.800230e-05   </td><td>2.824950e-05   </td><td>-8.789112e-03  </td></tr>\n",
       "\t<tr><td>typisch        </td><td>2.800230e-05   </td><td>2.824950e-05   </td><td>-8.789112e-03  </td></tr>\n",
       "\t<tr><td>vertrauen      </td><td>2.800230e-05   </td><td>2.824950e-05   </td><td>-8.789112e-03  </td></tr>\n",
       "\t<tr><td>warm           </td><td>5.600460e-05   </td><td>5.649900e-05   </td><td>-8.789112e-03  </td></tr>\n",
       "\t<tr><td>damals         </td><td>7.741812e-05   </td><td>7.667721e-05   </td><td> 9.616315e-03  </td></tr>\n",
       "\t<tr><td>gelandet       </td><td>4.776863e-05   </td><td>4.842771e-05   </td><td>-1.370313e-02  </td></tr>\n",
       "\t<tr><td>statt          </td><td>1.350699e-04   </td><td>1.331762e-04   </td><td> 1.411938e-02  </td></tr>\n",
       "\t<tr><td>deidesheim     </td><td>4.117985e-05   </td><td>4.035643e-05   </td><td> 2.019842e-02  </td></tr>\n",
       "\t<tr><td>10k            </td><td>1.647194e-05   </td><td>1.614257e-05   </td><td> 2.019842e-02  </td></tr>\n",
       "\t<tr><td>13k            </td><td>8.235970e-06   </td><td>8.071286e-06   </td><td> 2.019842e-02  </td></tr>\n",
       "\t<tr><td>⋮</td><td>⋮</td><td>⋮</td><td>⋮</td></tr>\n",
       "\t<tr><td>fm            </td><td>5.155717e-04  </td><td>8.071286e-06  </td><td> 4.156964     </td></tr>\n",
       "\t<tr><td>#bosc2015     </td><td>1.647194e-06  </td><td>1.089624e-04  </td><td>-4.191929     </td></tr>\n",
       "\t<tr><td>#smbe14       </td><td>1.647194e-06  </td><td>1.089624e-04  </td><td>-4.191929     </td></tr>\n",
       "\t<tr><td>#piratencafe  </td><td>2.701398e-04  </td><td>4.035643e-06  </td><td> 4.203774     </td></tr>\n",
       "\t<tr><td>portland      </td><td>1.647194e-06  </td><td>1.129980e-04  </td><td>-4.228297     </td></tr>\n",
       "\t<tr><td>#lqfb         </td><td>2.964949e-04  </td><td>4.035643e-06  </td><td> 4.296865     </td></tr>\n",
       "\t<tr><td>#mozsprint    </td><td>1.647194e-06  </td><td>1.251049e-04  </td><td>-4.330080     </td></tr>\n",
       "\t<tr><td>#muenster     </td><td>3.113197e-04  </td><td>4.035643e-06  </td><td> 4.345655     </td></tr>\n",
       "\t<tr><td>yay           </td><td>4.941582e-06  </td><td>3.833861e-04  </td><td>-4.351357     </td></tr>\n",
       "\t<tr><td>#smbe16       </td><td>1.647194e-06  </td><td>1.291406e-04  </td><td>-4.361828     </td></tr>\n",
       "\t<tr><td>#force2016    </td><td>1.647194e-06  </td><td>1.412475e-04  </td><td>-4.451440     </td></tr>\n",
       "\t<tr><td>#futurecommons</td><td>1.647194e-06  </td><td>1.412475e-04  </td><td>-4.451440     </td></tr>\n",
       "\t<tr><td>nrw           </td><td>7.330014e-04  </td><td>8.071286e-06  </td><td> 4.508835     </td></tr>\n",
       "\t<tr><td>#csvconf      </td><td>1.647194e-06  </td><td>1.654614e-04  </td><td>-4.609664     </td></tr>\n",
       "\t<tr><td>#mozwow       </td><td>1.647194e-06  </td><td>1.654614e-04  </td><td>-4.609664     </td></tr>\n",
       "\t<tr><td>#bosc2016     </td><td>1.647194e-06  </td><td>1.775683e-04  </td><td>-4.680282     </td></tr>\n",
       "\t<tr><td>#cccamp15     </td><td>1.647194e-06  </td><td>1.775683e-04  </td><td>-4.680282     </td></tr>\n",
       "\t<tr><td>berkeley      </td><td>1.647194e-06  </td><td>1.816039e-04  </td><td>-4.702755     </td></tr>\n",
       "\t<tr><td>#get2015      </td><td>1.647194e-06  </td><td>2.058178e-04  </td><td>-4.827918     </td></tr>\n",
       "\t<tr><td>#smbe15       </td><td>1.647194e-06  </td><td>2.219604e-04  </td><td>-4.903425     </td></tr>\n",
       "\t<tr><td>#piraten      </td><td>5.600460e-04  </td><td>4.035643e-06  </td><td> 4.932853     </td></tr>\n",
       "\t<tr><td>#latergram    </td><td>1.647194e-06  </td><td>2.502099e-04  </td><td>-5.023227     </td></tr>\n",
       "\t<tr><td>emoji         </td><td>1.647194e-06  </td><td>2.502099e-04  </td><td>-5.023227     </td></tr>\n",
       "\t<tr><td>#om13         </td><td>1.647194e-06  </td><td>2.582811e-04  </td><td>-5.054975     </td></tr>\n",
       "\t<tr><td>sci           </td><td>1.647194e-06  </td><td>2.703881e-04  </td><td>-5.100785     </td></tr>\n",
       "\t<tr><td>mp            </td><td>3.350393e-03  </td><td>2.017821e-05  </td><td> 5.112229     </td></tr>\n",
       "\t<tr><td>#bosc2017     </td><td>1.647194e-06  </td><td>3.793504e-04  </td><td>-5.439387     </td></tr>\n",
       "\t<tr><td>#mozfest      </td><td>1.647194e-06  </td><td>4.116356e-04  </td><td>-5.521065     </td></tr>\n",
       "\t<tr><td>#opencon      </td><td>1.647194e-06  </td><td>4.519920e-04  </td><td>-5.614591     </td></tr>\n",
       "\t<tr><td>ly            </td><td>3.758897e-03  </td><td>4.035643e-06  </td><td> 6.836715     </td></tr>\n",
       "</tbody>\n",
       "</table>\n"
      ],
      "text/latex": [
       "\\begin{tabular}{r|llll}\n",
       " word & past & today & logratio\\\\\n",
       "\\hline\n",
       "\t option          & 8.071251e-05    & 8.071286e-05    & -4.282714e-06  \\\\\n",
       "\t backup          & 4.447424e-05    & 4.439207e-05    &  1.849286e-03  \\\\\n",
       "\t ern             & 4.447424e-05    & 4.439207e-05    &  1.849286e-03  \\\\\n",
       "\t gefallen        & 7.247654e-05    & 7.264157e-05    & -2.274431e-03  \\\\\n",
       "\t timeline        & 1.087148e-04    & 1.089624e-04    & -2.274431e-03  \\\\\n",
       "\t wieviel         & 3.623827e-05    & 3.632079e-05    & -2.274431e-03  \\\\\n",
       "\t haha            & 2.091936e-04    & 2.098534e-04    & -3.148939e-03  \\\\\n",
       "\t extra           & 5.271021e-05    & 5.246336e-05    &  4.694238e-03  \\\\\n",
       "\t spam            & 1.054204e-04    & 1.049267e-04    &  4.694238e-03  \\\\\n",
       "\t videos          & 6.424057e-05    & 6.457028e-05    & -5.119383e-03  \\\\\n",
       "\t lasse           & 6.918215e-05    & 6.860593e-05    &  8.363967e-03  \\\\\n",
       "\t bioinformatiker & 2.800230e-05    & 2.824950e-05    & -8.789112e-03  \\\\\n",
       "\t fliegen         & 2.800230e-05    & 2.824950e-05    & -8.789112e-03  \\\\\n",
       "\t hosen           & 2.800230e-05    & 2.824950e-05    & -8.789112e-03  \\\\\n",
       "\t installation    & 2.800230e-05    & 2.824950e-05    & -8.789112e-03  \\\\\n",
       "\t instapaper      & 2.800230e-05    & 2.824950e-05    & -8.789112e-03  \\\\\n",
       "\t krass           & 2.800230e-05    & 2.824950e-05    & -8.789112e-03  \\\\\n",
       "\t medium          & 2.800230e-05    & 2.824950e-05    & -8.789112e-03  \\\\\n",
       "\t quelle          & 2.800230e-05    & 2.824950e-05    & -8.789112e-03  \\\\\n",
       "\t rper            & 2.800230e-05    & 2.824950e-05    & -8.789112e-03  \\\\\n",
       "\t sequenzen       & 2.800230e-05    & 2.824950e-05    & -8.789112e-03  \\\\\n",
       "\t typisch         & 2.800230e-05    & 2.824950e-05    & -8.789112e-03  \\\\\n",
       "\t vertrauen       & 2.800230e-05    & 2.824950e-05    & -8.789112e-03  \\\\\n",
       "\t warm            & 5.600460e-05    & 5.649900e-05    & -8.789112e-03  \\\\\n",
       "\t damals          & 7.741812e-05    & 7.667721e-05    &  9.616315e-03  \\\\\n",
       "\t gelandet        & 4.776863e-05    & 4.842771e-05    & -1.370313e-02  \\\\\n",
       "\t statt           & 1.350699e-04    & 1.331762e-04    &  1.411938e-02  \\\\\n",
       "\t deidesheim      & 4.117985e-05    & 4.035643e-05    &  2.019842e-02  \\\\\n",
       "\t 10k             & 1.647194e-05    & 1.614257e-05    &  2.019842e-02  \\\\\n",
       "\t 13k             & 8.235970e-06    & 8.071286e-06    &  2.019842e-02  \\\\\n",
       "\t ⋮ & ⋮ & ⋮ & ⋮\\\\\n",
       "\t fm             & 5.155717e-04   & 8.071286e-06   &  4.156964     \\\\\n",
       "\t \\#bosc2015      & 1.647194e-06     & 1.089624e-04     & -4.191929       \\\\\n",
       "\t \\#smbe14        & 1.647194e-06     & 1.089624e-04     & -4.191929       \\\\\n",
       "\t \\#piratencafe   & 2.701398e-04     & 4.035643e-06     &  4.203774       \\\\\n",
       "\t portland       & 1.647194e-06   & 1.129980e-04   & -4.228297     \\\\\n",
       "\t \\#lqfb          & 2.964949e-04     & 4.035643e-06     &  4.296865       \\\\\n",
       "\t \\#mozsprint     & 1.647194e-06     & 1.251049e-04     & -4.330080       \\\\\n",
       "\t \\#muenster      & 3.113197e-04     & 4.035643e-06     &  4.345655       \\\\\n",
       "\t yay            & 4.941582e-06   & 3.833861e-04   & -4.351357     \\\\\n",
       "\t \\#smbe16        & 1.647194e-06     & 1.291406e-04     & -4.361828       \\\\\n",
       "\t \\#force2016     & 1.647194e-06     & 1.412475e-04     & -4.451440       \\\\\n",
       "\t \\#futurecommons & 1.647194e-06     & 1.412475e-04     & -4.451440       \\\\\n",
       "\t nrw            & 7.330014e-04   & 8.071286e-06   &  4.508835     \\\\\n",
       "\t \\#csvconf       & 1.647194e-06     & 1.654614e-04     & -4.609664       \\\\\n",
       "\t \\#mozwow        & 1.647194e-06     & 1.654614e-04     & -4.609664       \\\\\n",
       "\t \\#bosc2016      & 1.647194e-06     & 1.775683e-04     & -4.680282       \\\\\n",
       "\t \\#cccamp15      & 1.647194e-06     & 1.775683e-04     & -4.680282       \\\\\n",
       "\t berkeley       & 1.647194e-06   & 1.816039e-04   & -4.702755     \\\\\n",
       "\t \\#get2015       & 1.647194e-06     & 2.058178e-04     & -4.827918       \\\\\n",
       "\t \\#smbe15        & 1.647194e-06     & 2.219604e-04     & -4.903425       \\\\\n",
       "\t \\#piraten       & 5.600460e-04     & 4.035643e-06     &  4.932853       \\\\\n",
       "\t \\#latergram     & 1.647194e-06     & 2.502099e-04     & -5.023227       \\\\\n",
       "\t emoji          & 1.647194e-06   & 2.502099e-04   & -5.023227     \\\\\n",
       "\t \\#om13          & 1.647194e-06     & 2.582811e-04     & -5.054975       \\\\\n",
       "\t sci            & 1.647194e-06   & 2.703881e-04   & -5.100785     \\\\\n",
       "\t mp             & 3.350393e-03   & 2.017821e-05   &  5.112229     \\\\\n",
       "\t \\#bosc2017      & 1.647194e-06     & 3.793504e-04     & -5.439387       \\\\\n",
       "\t \\#mozfest       & 1.647194e-06     & 4.116356e-04     & -5.521065       \\\\\n",
       "\t \\#opencon       & 1.647194e-06     & 4.519920e-04     & -5.614591       \\\\\n",
       "\t ly             & 3.758897e-03   & 4.035643e-06   &  6.836715     \\\\\n",
       "\\end{tabular}\n"
      ],
      "text/markdown": [
       "\n",
       "word | past | today | logratio | \n",
       "|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|\n",
       "| option          | 8.071251e-05    | 8.071286e-05    | -4.282714e-06   | \n",
       "| backup          | 4.447424e-05    | 4.439207e-05    |  1.849286e-03   | \n",
       "| ern             | 4.447424e-05    | 4.439207e-05    |  1.849286e-03   | \n",
       "| gefallen        | 7.247654e-05    | 7.264157e-05    | -2.274431e-03   | \n",
       "| timeline        | 1.087148e-04    | 1.089624e-04    | -2.274431e-03   | \n",
       "| wieviel         | 3.623827e-05    | 3.632079e-05    | -2.274431e-03   | \n",
       "| haha            | 2.091936e-04    | 2.098534e-04    | -3.148939e-03   | \n",
       "| extra           | 5.271021e-05    | 5.246336e-05    |  4.694238e-03   | \n",
       "| spam            | 1.054204e-04    | 1.049267e-04    |  4.694238e-03   | \n",
       "| videos          | 6.424057e-05    | 6.457028e-05    | -5.119383e-03   | \n",
       "| lasse           | 6.918215e-05    | 6.860593e-05    |  8.363967e-03   | \n",
       "| bioinformatiker | 2.800230e-05    | 2.824950e-05    | -8.789112e-03   | \n",
       "| fliegen         | 2.800230e-05    | 2.824950e-05    | -8.789112e-03   | \n",
       "| hosen           | 2.800230e-05    | 2.824950e-05    | -8.789112e-03   | \n",
       "| installation    | 2.800230e-05    | 2.824950e-05    | -8.789112e-03   | \n",
       "| instapaper      | 2.800230e-05    | 2.824950e-05    | -8.789112e-03   | \n",
       "| krass           | 2.800230e-05    | 2.824950e-05    | -8.789112e-03   | \n",
       "| medium          | 2.800230e-05    | 2.824950e-05    | -8.789112e-03   | \n",
       "| quelle          | 2.800230e-05    | 2.824950e-05    | -8.789112e-03   | \n",
       "| rper            | 2.800230e-05    | 2.824950e-05    | -8.789112e-03   | \n",
       "| sequenzen       | 2.800230e-05    | 2.824950e-05    | -8.789112e-03   | \n",
       "| typisch         | 2.800230e-05    | 2.824950e-05    | -8.789112e-03   | \n",
       "| vertrauen       | 2.800230e-05    | 2.824950e-05    | -8.789112e-03   | \n",
       "| warm            | 5.600460e-05    | 5.649900e-05    | -8.789112e-03   | \n",
       "| damals          | 7.741812e-05    | 7.667721e-05    |  9.616315e-03   | \n",
       "| gelandet        | 4.776863e-05    | 4.842771e-05    | -1.370313e-02   | \n",
       "| statt           | 1.350699e-04    | 1.331762e-04    |  1.411938e-02   | \n",
       "| deidesheim      | 4.117985e-05    | 4.035643e-05    |  2.019842e-02   | \n",
       "| 10k             | 1.647194e-05    | 1.614257e-05    |  2.019842e-02   | \n",
       "| 13k             | 8.235970e-06    | 8.071286e-06    |  2.019842e-02   | \n",
       "| ⋮ | ⋮ | ⋮ | ⋮ | \n",
       "| fm             | 5.155717e-04   | 8.071286e-06   |  4.156964      | \n",
       "| #bosc2015      | 1.647194e-06   | 1.089624e-04   | -4.191929      | \n",
       "| #smbe14        | 1.647194e-06   | 1.089624e-04   | -4.191929      | \n",
       "| #piratencafe   | 2.701398e-04   | 4.035643e-06   |  4.203774      | \n",
       "| portland       | 1.647194e-06   | 1.129980e-04   | -4.228297      | \n",
       "| #lqfb          | 2.964949e-04   | 4.035643e-06   |  4.296865      | \n",
       "| #mozsprint     | 1.647194e-06   | 1.251049e-04   | -4.330080      | \n",
       "| #muenster      | 3.113197e-04   | 4.035643e-06   |  4.345655      | \n",
       "| yay            | 4.941582e-06   | 3.833861e-04   | -4.351357      | \n",
       "| #smbe16        | 1.647194e-06   | 1.291406e-04   | -4.361828      | \n",
       "| #force2016     | 1.647194e-06   | 1.412475e-04   | -4.451440      | \n",
       "| #futurecommons | 1.647194e-06   | 1.412475e-04   | -4.451440      | \n",
       "| nrw            | 7.330014e-04   | 8.071286e-06   |  4.508835      | \n",
       "| #csvconf       | 1.647194e-06   | 1.654614e-04   | -4.609664      | \n",
       "| #mozwow        | 1.647194e-06   | 1.654614e-04   | -4.609664      | \n",
       "| #bosc2016      | 1.647194e-06   | 1.775683e-04   | -4.680282      | \n",
       "| #cccamp15      | 1.647194e-06   | 1.775683e-04   | -4.680282      | \n",
       "| berkeley       | 1.647194e-06   | 1.816039e-04   | -4.702755      | \n",
       "| #get2015       | 1.647194e-06   | 2.058178e-04   | -4.827918      | \n",
       "| #smbe15        | 1.647194e-06   | 2.219604e-04   | -4.903425      | \n",
       "| #piraten       | 5.600460e-04   | 4.035643e-06   |  4.932853      | \n",
       "| #latergram     | 1.647194e-06   | 2.502099e-04   | -5.023227      | \n",
       "| emoji          | 1.647194e-06   | 2.502099e-04   | -5.023227      | \n",
       "| #om13          | 1.647194e-06   | 2.582811e-04   | -5.054975      | \n",
       "| sci            | 1.647194e-06   | 2.703881e-04   | -5.100785      | \n",
       "| mp             | 3.350393e-03   | 2.017821e-05   |  5.112229      | \n",
       "| #bosc2017      | 1.647194e-06   | 3.793504e-04   | -5.439387      | \n",
       "| #mozfest       | 1.647194e-06   | 4.116356e-04   | -5.521065      | \n",
       "| #opencon       | 1.647194e-06   | 4.519920e-04   | -5.614591      | \n",
       "| ly             | 3.758897e-03   | 4.035643e-06   |  6.836715      | \n",
       "\n",
       "\n"
      ],
      "text/plain": [
       "      word            past         today        logratio     \n",
       "1     option          8.071251e-05 8.071286e-05 -4.282714e-06\n",
       "2     backup          4.447424e-05 4.439207e-05  1.849286e-03\n",
       "3     ern             4.447424e-05 4.439207e-05  1.849286e-03\n",
       "4     gefallen        7.247654e-05 7.264157e-05 -2.274431e-03\n",
       "5     timeline        1.087148e-04 1.089624e-04 -2.274431e-03\n",
       "6     wieviel         3.623827e-05 3.632079e-05 -2.274431e-03\n",
       "7     haha            2.091936e-04 2.098534e-04 -3.148939e-03\n",
       "8     extra           5.271021e-05 5.246336e-05  4.694238e-03\n",
       "9     spam            1.054204e-04 1.049267e-04  4.694238e-03\n",
       "10    videos          6.424057e-05 6.457028e-05 -5.119383e-03\n",
       "11    lasse           6.918215e-05 6.860593e-05  8.363967e-03\n",
       "12    bioinformatiker 2.800230e-05 2.824950e-05 -8.789112e-03\n",
       "13    fliegen         2.800230e-05 2.824950e-05 -8.789112e-03\n",
       "14    hosen           2.800230e-05 2.824950e-05 -8.789112e-03\n",
       "15    installation    2.800230e-05 2.824950e-05 -8.789112e-03\n",
       "16    instapaper      2.800230e-05 2.824950e-05 -8.789112e-03\n",
       "17    krass           2.800230e-05 2.824950e-05 -8.789112e-03\n",
       "18    medium          2.800230e-05 2.824950e-05 -8.789112e-03\n",
       "19    quelle          2.800230e-05 2.824950e-05 -8.789112e-03\n",
       "20    rper            2.800230e-05 2.824950e-05 -8.789112e-03\n",
       "21    sequenzen       2.800230e-05 2.824950e-05 -8.789112e-03\n",
       "22    typisch         2.800230e-05 2.824950e-05 -8.789112e-03\n",
       "23    vertrauen       2.800230e-05 2.824950e-05 -8.789112e-03\n",
       "24    warm            5.600460e-05 5.649900e-05 -8.789112e-03\n",
       "25    damals          7.741812e-05 7.667721e-05  9.616315e-03\n",
       "26    gelandet        4.776863e-05 4.842771e-05 -1.370313e-02\n",
       "27    statt           1.350699e-04 1.331762e-04  1.411938e-02\n",
       "28    deidesheim      4.117985e-05 4.035643e-05  2.019842e-02\n",
       "29    10k             1.647194e-05 1.614257e-05  2.019842e-02\n",
       "30    13k             8.235970e-06 8.071286e-06  2.019842e-02\n",
       "⋮     ⋮               ⋮            ⋮            ⋮            \n",
       "74166 fm              5.155717e-04 8.071286e-06  4.156964    \n",
       "74167 #bosc2015       1.647194e-06 1.089624e-04 -4.191929    \n",
       "74168 #smbe14         1.647194e-06 1.089624e-04 -4.191929    \n",
       "74169 #piratencafe    2.701398e-04 4.035643e-06  4.203774    \n",
       "74170 portland        1.647194e-06 1.129980e-04 -4.228297    \n",
       "74171 #lqfb           2.964949e-04 4.035643e-06  4.296865    \n",
       "74172 #mozsprint      1.647194e-06 1.251049e-04 -4.330080    \n",
       "74173 #muenster       3.113197e-04 4.035643e-06  4.345655    \n",
       "74174 yay             4.941582e-06 3.833861e-04 -4.351357    \n",
       "74175 #smbe16         1.647194e-06 1.291406e-04 -4.361828    \n",
       "74176 #force2016      1.647194e-06 1.412475e-04 -4.451440    \n",
       "74177 #futurecommons  1.647194e-06 1.412475e-04 -4.451440    \n",
       "74178 nrw             7.330014e-04 8.071286e-06  4.508835    \n",
       "74179 #csvconf        1.647194e-06 1.654614e-04 -4.609664    \n",
       "74180 #mozwow         1.647194e-06 1.654614e-04 -4.609664    \n",
       "74181 #bosc2016       1.647194e-06 1.775683e-04 -4.680282    \n",
       "74182 #cccamp15       1.647194e-06 1.775683e-04 -4.680282    \n",
       "74183 berkeley        1.647194e-06 1.816039e-04 -4.702755    \n",
       "74184 #get2015        1.647194e-06 2.058178e-04 -4.827918    \n",
       "74185 #smbe15         1.647194e-06 2.219604e-04 -4.903425    \n",
       "74186 #piraten        5.600460e-04 4.035643e-06  4.932853    \n",
       "74187 #latergram      1.647194e-06 2.502099e-04 -5.023227    \n",
       "74188 emoji           1.647194e-06 2.502099e-04 -5.023227    \n",
       "74189 #om13           1.647194e-06 2.582811e-04 -5.054975    \n",
       "74190 sci             1.647194e-06 2.703881e-04 -5.100785    \n",
       "74191 mp              3.350393e-03 2.017821e-05  5.112229    \n",
       "74192 #bosc2017       1.647194e-06 3.793504e-04 -5.439387    \n",
       "74193 #mozfest        1.647194e-06 4.116356e-04 -5.521065    \n",
       "74194 #opencon        1.647194e-06 4.519920e-04 -5.614591    \n",
       "74195 ly              3.758897e-03 4.035643e-06  6.836715    "
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "library(tidyr)\n",
    "\n",
    "tidy_tweets$date_group <- ifelse(tidy_tweets$timestamp < cutoff_date,\"past\",\"today\")\n",
    "\n",
    "word_ratios <- tidy_tweets %>%\n",
    "  filter(!str_detect(word, \"^@\")) %>%\n",
    "  count(word, date_group) %>%\n",
    "  filter(sum(n) >= 10) %>%\n",
    "  ungroup() %>%\n",
    "  spread(date_group, n, fill = 0) %>%\n",
    "  mutate_if(is.numeric, funs((.   1) / sum(.   1))) %>%\n",
    "  mutate(logratio = log(past / today)) %>%\n",
    "  arrange(desc(logratio))\n",
    "\n",
    "word_ratios %>% \n",
    "  arrange(abs(logratio))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We can see that some things, e.g. `archives` haven't changed much in frequency, as obvious from the small \"odds\" ratio that's close to zero. Outliers with a large positive values are those that were more frequent in the past, while outliers with values large negative values are more common in more recent tweets. To make it a bit more intuitive we can now plot the Top 15 words with the largest positive/negative odds ratio: "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "data": {},
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAA0gAAANICAMAAADKOT/pAAADAFBMVEUAAAAAv8QBAQECAgID\nAwMEBAQFBQUGBgYHBwcICAgJCQkKCgoLCwsMDAwNDQ0ODg4PDw8RERESEhITExMUFBQVFRUW\nFhYXFxcYGBgZGRkaGhobGxscHBwdHR0eHh4fHx8gICAhISEiIiIjIyMkJCQlJSUmJiYnJyco\nKCgpKSkqKiorKyssLCwtLS0uLi4vLy8wMDAxMTEyMjIzMzM0NDQ1NTU2NjY3Nzc4ODg5OTk6\nOjo7Ozs8PDw9PT0 Pj4/Pz9AQEBBQUFCQkJDQ0NERERFRUVGRkZHR0dISEhJSUlLS0tMTExN\nTU1OTk5PT09QUFBRUVFSUlJTU1NUVFRVVVVWVlZXV1dYWFhZWVlaWlpbW1tcXFxdXV1eXl5f\nX19gYGBhYWFiYmJjY2NkZGRlZWVmZmZnZ2doaGhpaWlqampra2tsbGxtbW1ubm5vb29wcHBx\ncXFycnJzc3N0dHR1dXV2dnZ3d3d4eHh5eXl6enp7e3t8fHx9fX1 fn5/f3 AgICBgYGCgoKD\ng4OEhISFhYWGhoaHh4eIiIiJiYmKioqLi4uMjIyNjY2Ojo6Pj4 QkJCRkZGSkpKTk5OUlJSV\nlZWWlpaXl5eYmJiZmZmampqbm5ucnJydnZ2enp6fn5 goKChoaGioqKjo6OkpKSlpaWmpqan\np6eoqKipqamqqqqrq6usrKytra2urq6vr6 wsLCxsbGysrKzs7O0tLS1tbW2tra3t7e4uLi5\nubm6urq7u7u8vLy9vb2 vr6/v7/AwMDBwcHCwsLDw8PExMTFxcXGxsbHx8fIyMjJycnKysrL\ny8vMzMzNzc3Ozs7Pz8/Q0NDR0dHS0tLT09PU1NTV1dXW1tbX19fY2NjZ2dna2trb29vc3Nzd\n3d3e3t7f39/g4ODh4eHi4uLj4 Pk5OTl5eXm5ubn5 fo6Ojp6enq6urr6 vs7Ozt7e3u7u7v\n7 /w8PDx8fHy8vLz8/P09PT19fX29vb39/f4dm34 Pj5 fn6 vr7 /v8/Pz9/f3 /v7////8\nl6Z1AAAACXBIWXMAABJ0AAASdAHeZh94AAAgAElEQVR4nO29DXwT553vq9v2pm2atrs9afds\nN7vtud2zbc 2vX3Zt 5ue7rt3u6ee8cyyMa8GCeOCwFDAkZ17YWEQEhIQppCaCA0KWleqJNc\nkrZA4iaA042bghNCYyjlpWDiEF5iZMmyZFvXsmX974ykGUkg2f R5tEzo/y Hz7SaF6eZ bR\n/ uZkeUfLgIAFI1L9g4AUA5AJAAsACIBYAEQCQALgEgAWABEAsACIBIAFgCRALAAe4s0MSi \nj6FATHgf40PCu6BQIC68j iI8C4oOCC j5FRy5u0uUglGNSwT7xIYyHhXVDQJ16k0WHhXVDA\nL76P4YjlTUIkiMQHIuUFIkEkPhApLxCpbET6/4AAuMMPkSASmATu8EMkiAQmgTv8EAkigUng\nDj9EgkhgErjDL12kmHIw/0KIxAciCYE7/BAJIoFJ4A4/RIJIYBK4w28HkZo3q8/HK/ouXwiR\n EAkIXCH3w4i7ZkZJXpouTFnzCAaGBPOoG9UeB8jQeFdjA3ILrnyZIpRN rYDiJFZrxE8ete\n1OcM 0AhyC658mTyMQ8YdWwHkej W jIDOOb7WMjBsP EeEEfUPC wgPCO9iJCC75MqTyQc9\n/Z09W4h0osL/wA9zLcQ9Eh/cIwmBO/y2EIlufKb2aK6FEIkPRBICd/jtIdLOmhtyLoRIfCCS\nELjDbw RwtO251wIkfhAJCFwh1 6SAl6PbmNgUh8IJIQuMNvB5Em pb9OM8SiMQGIgmBO/x2\nEKmten009xKIxAciCYE7/HYQKT8QiQ9EEgJ3 CESRAKTwB1 iFQ2IiFFiA1ShAQAkfhApLxA\nJIjEByLlBSKVjUiy7yZkY2KsIJIAIFKZYGKsIJIAIFKZYGKsIJIAIFKZYGKsIJIAIFKZYGKs\nIJIAIFKZYGKsIJIAIFKZYGKsIJIAIFKZYGKsykekuNKxvKGxZ tNdc9QVHm teH6rlyrQSQ \nEIlP YhE7pbIROvMfXTQHYwpSwZoT2VQXzQxbhANjAtn0BcV3kdkUHgX4 /4XDsTYxUetmjQ\n0z DpYnUTvRIPVFEOR5Tdqj21OzWFyHXrjBkF7JsZIy5/Fw79ytEbV4tsuFwTFGnaWGbviga\nNgj5w8IJ ELC xgMCO8i7JddyLIxMVYDQYsGPf3dQ2kivZoh0j51xvwncqyFeyQ uEfiU073\nSJkiPa2ehjx7c6wFkfhAJD7lKtKi3mjbtMEca0EkPhCJT7mK1P49T8OBXGtBJD4QiU8ZiZTB\nJP/VGETiA5H4QCQBQKQywcRYQSQBQKQywcRYladIkwCR EAkPhBJABCpTDAxVhBJABCpTDAx\nVhBJAGUjEuK42EAkAUAkPhApLxCpbESSfWkliULGCiIJACI5m0LGCiIJACI5m0LGCiIJACI5\nm0LGCiIJACI5m0LGyrkixZTuQjaDSHwgEh8niRRaTXSH8Sft8cPh7MWHTnIagUh8IBIfJ4l0\ncBvRsvyLV7dzGoFIfCASHyeJ9GQXBdYYmXXqpd2E8kLDBuq9ZVbNyvO0vGL6Ugqsq6tqPUVx\n5aWVjdd35GoEIvGBSHycI9IO76wl3gX13mAqs067R3I3nRqhBesjw3c2EzWoZyTvulD08dqo\nuiBIu6uMY4uniQ3EhRP2jQvvIxoS3kX8nSpSIWM1HLFq1EWLRPGlRI90k55ZlxDpKXV eJRo\nX2VcE mUot5CxWd2knsnUZ/ypr4pcu0KQ3ZFS0LqmIvPtetdT9QaJT2zLiFSpzp5aPncuTOV\nmCZSp5JgO7n3EQUU49OHSNBgoD8oHL9vQHgfA37hXQT7ZVe0JAoZK3/AokFPX7ILEunmhnne\nphrvQ3pmXUKk/UTnp2 PUldSpC4lmlxZW5AhUga4R LzTr20K2SsnHOPRPf10tEt2u Pkpl1\nukidbrVqH02KdEY5rq54ASJZA0Ti4yCRmmO0q0MTKZlZp4t0XDky9lKrcpEatw7RiuaLsfYq\nP0SyBIjExzkixVqJNpzRREpm1uki0U9mzd4YbprZt9NTT4G7a2Y0H8EZyRogEh/niKQzSUIQ\nB4jEByLxgUgCgEjOppCxgkgCgEjOppCxcp5IRQKR EAkPhBJABDJ2RQyVhBJABDJ2RQyVhBJ\nAGUjEuK42EAkAUAkPhApLxCpbESSfY0lHOvGCiIJACI5BevGCiIJACI5BevGCiIJACI5BevG\nCiIJACI5BevGCiIJACI5BevGCiIJACI5BevGCiIJACI5BevGqjxE0nPsEkF3Wp7dY0ofUetT\nOVaFSHwgEp/yEMnIsdOC7jZuJPpu44sUnY6/kC0OiMSnXERK5dhpQXedN1CkesdGOjTH ILL\ncH8aX794yqSP8s 1s3CsrHo/0j/opYiUyrHTgu5CFYEDLb3zads9xuLIQJr AeH4feL7CJTg\nOMo/1866sfL7LWpoUK5IqbCTRBpK08tbt8Vn 1uQ/V0kuLTjUy6XdpkiPbqlqZvW7JkWyLUm\nROIDkfiUo0iHF3ii9LPFN ZcEyLxgUh8ylGk8eoWohPKwznXhEh8IBKfMhGJD0TiA5H4QCQB\nQCSnYN1YQSQBQCSnYN1YQSQBQCSnYN1YQSQBQCSnYN1YQSQBQCSnYN1YQSQBlI1IiONiA5EE\nAJH4QKS8QKSyEUn2lZcwrB8riCQAiGR3rB8riCQAiGR3rB8riCQAiGR3rB8riCQAiGR3rB8r\niCQAiGR3rB8rO4sUWk10R86/zsvHoVxxJ9lAJD4QiY dRTq4jWiZqS1Wt0 5CkTiA5H42Fmk\nJ7sosIbiSsfyhsaerTfVPUMR5TDReeU8BdbVVbWeMvLsaO9Cz9zN0eUV05fqixIRdzkahUh8\nIBIf 4q0wztriXdBvTfkbolMtM7cRwfdQUMk77pQ9PHaqJ5nd6Gie LCku3UoJ6R0otOjeRo\nFiLxgUh87CsSxZcSPdJN5FbteKSe1NPRcV2kU4p66xSf2ann2Z1SelRFSBMpvSgjZ3UkkKY/\nIBy/rwR9lOA4yjeOy/qx8vstaihouUi964lao6pIrxC1eYliymFdpE4lwXY9zy6 qbK57VxC\npPSiznRTCIgsrAvZ9S4MAWNl24DImxvmeZtqvA R 9Vskc4q57uUaHIlPfSE p5bVdmpiZS1\nKAe4tOODSzs Nr60u6 Xjm5RnzNEilUcJDqonD jHFcXXDBEimnnwy2tmkhZi3IAkfhAJD42\nFqk5Rru0sNQMkWj wzR6m3KeVjRfjLVX XWR9tSfjAeWb6TGrUNZi3IAkfhAJD72FSnWSrTh\nDGWL9PoN8773mvIWBe6umdF8xDgjxdvqp9XdN0Q7PfVZi3IAkfhAJD72FUkQEIkPROIDkQQA\nkeyO9WMFkQQAkeyO9WMFkQQAkeyO9WMFkQQAkeyO9WMFkQQAkeyO9WMFkQRQNiIhjosNRBIA\nROIDkfICkcpGJNlXYIIQMVYQSQAQyd6IGCuIJACIZG9EjBVEEgBEsjcixgoiCQAi2RsRY1Um\nIsWUg8b0iXnTAxkvLwEi8YFIfBwvUjL8LlOke1vD4xDJCiASH8eLlAy/yxTptk1ZLy8BIvGB\nSHwcL1Iy/E4z5 hiz JXlZ5Wd2W1X/lli2dhV671IRIfiMTH4SLp4XdB5eBEw8bI261KL63W\nzkgLjg63Vfbl2AIi8YFIfBwukh5 p56RjimqN126SE qpVZlBBhHggYD/UHh H0DwvsY8Avv\nIliuuXYixsofsKih9A/IkoqUCr9TReqsiBOd00V6WV02/zF9rWEfKATZFS8I2cM6Ken/NaKU\nIunhd7pIZ3SRXlUXLnhCXy2eJjYQF07YNy68j2hIeBfxcr20EzFWwxGrWpIikh5 p4r0e Vt\nov/URfo50Xj1nhwb4B6JT7mKJGKsnH6PlAq/U0WK1W0YOuPVRWrsHd/uyVVrEIkPROLjcJH0\n8Dvt4 8TSzyNnUmRokpHs6fxtVxbQCQ EImPw0W6lD5VpMmBSHwgEh IJACIZG9EjBVEEgBE\nsjcixqrMRJoaiMQHIvGBSAKASPZGxFhBJAFAJHsjYqwgkgDKRiTEcbGBSAKASHwgUl4gUtmI\nJPsazHqEjRVEEgBEsi3CxgoiCQAi2RZhYwWRBACRbIuwsYJIAoBItkXYWNlcpJjSbXqbE/Om\nB/MvhUh8IBIf24mUzKnTX8UPh7MXHzo5ZQv3toYn dQWIvGBSHxsJ1Iypy4vq9vzL0tx26bJ\nlkIkPhCJj 1ESubURZXnWxuu79Iu7SaUFxo2UO8ts2pWnqflFdOXUmBdXVXrKYorL61svL6D\nyHdH9dzNo/oqWq5dMLVKDiASH4jEx2YipXPqlgzQnsqgdo/kbjo1QgvWR4bvbCZqUM9I3nWh\n6OO1UXVBkHZXRWjZXYFzi7YYq6zelF7lciASH4jEx2YipXPqdqg1X7M7IdJT6vzwKNG yrgm\n0ilFvYWKz wk907tD5De7FEuEPUcMFZRRTJWSRENG4T8YeEEfCHhfQwGhHcR9ssue sRNlYD\nQYsaSn9lqiiRjJy6V9QXC9sSImk HFo d 5MJaaJ1Kkk2E7ufUQB5eTLFcnPFvRVVJGMVVIg\n164wZJe99cge0amxJtcunVOnSkLzn0iItJ/o/PTtUepKitSlpC7ZtAWqSL pmNBeGauoIhmr\npIjHDMYCMeGEfFHhfYwOCu8iNiC77K1H2FiFhy1qaMISkdI5dU rF2SevbpInW71ruPRpEhn\nlOPqihcMkU5rf15 4lljFVUkY5XLwT0SH9wj8bHbPZKRU7eoN9o2bVAX6bhyZOylVuUiNW4d\nohXNF2PtVX5dJPKufPvsks3GKtqHDfoqlwOR EAkPjYTKZ1T1/49T8MB0kWin8yavTHcNLNv\np6eeAnfXzGg YpyRKHh7Ve39EWMVTSR9lcuBSHwgEh biaQzyf8UViQQiQ9E4gORBACRbIuw\nsYJIAoBItkXYWNlUJHFAJD4QiQ9EEgBEsi3CxgoiCQAi2RZhYwWRBFA2IiGOiw1EEgBE4gOR\n8gKRIBIfiJQXiFQ2Ism o7EagWMFkQQAkWyKwLGCSAKASDZF4FhBJAFAJJsicKzKTKTJM 00\nIBIfiMSnzESaPNNOAyLxgUh8ykykyTPtNCASH4jEp7xE0jLtAkrH8obGnq031T2TaxWIxAci\n8SkvkRJ/Zu5uiUy0ztxHB925bpcgEh IxKccRWoneqSeKJJIQEkwNmQQ9g8JZ8AXFt5HKCC8\ni6Gyy7UTOFYDQYsaGjGKWbpIrxC1ebU/Djysz0auXWHILnyrkT2eHKzJtSuWhEivXiZSbMwg\nGhgTzqBvVHgfI0HhXYyVXa6dwLEKD1nVklHMNhQpA9wj8cE9Ep9yvEeCSBYBkfhAJAFAJJsi\ncKzKTKSpgUh8IBIfiCQAiGRTBI4VRBIARLIpAscKIgkAItkUgWMFkQQAkWyKwLGCSAKASDZF\n4FhBJAGUjUhIEWIDkQQAkfhApLxAJIjEByLlBSKVjUiy72ksRexYQSQBQCQ7InasIJIAIJId\nETtWEEkAEMmOiB0riCQAiGRHxI4VRBIARLIjYscKIgkAItkRsWPlWJHiyksrG6/vIAqsq6tq\nPUUR7c/4zivn9dcTygsNG3JsB5H4QCQ jhWJ3E1B2l0VIe 6UPTx2qghUuq1uvzUSI7NIBIf\niMTHwSLtJOpT3jylBNTT08xOXST9NbmfSq87HjEY8UeEE/QNC 9jaEB4F5GA7Nq3FLFjNRiy\nqKHRkou0T72sU052Kgm26yLpr8ndmV4XuXaFIbv2LUX2YDIpea6de39CpC4lmniZEOmscl5/\nnViugzNSQeCMxMfBZ6SkSGcSwcQXKFZxkOigcl5/nSVSBrhH4oN7JD4OvkdKikQrmi/G2qv8\nNP9hGr1NOW 8hkhFA5H4OF kwN01M5qPEL1 w7zvvaa8ZbyGSEUDkfg4V6QCgUh8IBIfiCQA\niGRHxI4VRBIARLIjYscKIgkAItkRsWMFkQQAkeyI2LGCSAKASHZE7FhBJAGUjUhIEWIDkQQA\nkfhApLxAJIjEByLlBSKVjUiyb2uKJ3EcEEkAEIkPROIDkQQAkWxD4jggkgAgEh IxAciCQAi\n2YbEcUCk3MSUgzkmmUAkPhCJj7NECq0muiMwmUhnvW7tKXzvtbNX9 VqAiLxgUh8nCXSwW1E\nyyY7I3XWrU ItKbl9Ll7Fk3kaAIi8YFIfJwl0pNdFFij2aPHQyYmO5Y3NPZsvanuGaKOi/s1\nkXwVPWo5V3bnaAIi8YFIfJwk0g7vrCXeBfXeoHoaSsVDamckd0tkonXmPjroDqorJUTa59F \nJb/YSLaLjRlEA2PCGfSNCu9jJCi8i7EB2RoUT I4/H7xYxUesqol8SJRfCnRI92J67lUPGRi\nsl2dW68Fch3XRXr Om31FQ/qGyLXrjBka1A8skfQPKXItetdT9QaTYqUjIdMTL5C1ObV7pcO\nGyLVZ4s0NmIw7B8RTtA3JLyP8IDwLkbKINcucRz fvFjNRiyqKH0JaIwkW5umOdtqvE lLBn\nf1qkVy8TqSt5abc9RyO4R KDeyQ TrpHovt66eiW5Ed1k4vkrzhJNOg kqMNiMQHIvFxlEjN\nMdrVMalIAd9ut88XoTuXnj67almuPwKASHwgEh8niRRrJdpwZlKRGhIJ jtoeH3dnLWBXI1A\nJD4QiY TRLIEiMQHIvGBSAKASLYhcRwQSQAQiQ9E4gORBACRbEPiOCCSACASH4jEByIJoGxE\nQooQG4gkAIjEByLlBSJBJD4QKS8QqWxEkn2HUyyp44BIAoBIfCASH4gkAIhkE1LHAZEEAJH4\nQCQ EEkAEMkmpI4DIl2O SC7S4BIfCASH eIdFmmXWFAJD4QiY9zRLos064wIBIfiMTHOSJl\nZNqlg y8D6hLDlf4UjF3E8oLDRvUZ3pM6SNqfSpHMxCJD0Ti4xSRsjLt0kF2O viRJtv1mPu\n1OdTIxs3En238UWKTj pbz0xbhANjAtn0BcV3kdkUHgX447PtUsdh98vfqzCwxY1lP4ZLOaM\nlJlplw6yG3AfoYnavXrMHbnVk1DnDRSp3rGRDs0xvuCCXLvCkC1Cscgev0IQnWuXmWmXEWR3\ny4PU7RnRY 7I3UkUqggcaOmdT9vuMTaOhg1C/rBwAr6Q8D4GA8K7CPtli1AsqePo7xc/VgNB\nixpKf/dQiEhZmXYZaSd76 P3ryM9CyXxTE0vb90Wn 1v6cjVEO6R OAeiY9T7pGyMu0yRBr2\nHJt94BKRHt3S1E1r9kxDilCRQCQ jhEpM9MuMxHyzptrY5eIdHiBJ0o/W3xjznYgEh IxMcp\nImVl2mWKtE/5EV0i0nh1C9EJ5eGcDUEkPhCJj1NEsgyIxAci8YFIAoBINiF1HBBJABCJD0Ti\nA5EEAJFsQuo4IJIAIBIfiMQHIgkAItmE1HFAJAFAJD6I4 IDkQQAkfhApLxApLIRSfalWXEY\nxwGRBACR EAkPhBJABDJFhjHAZEEAJH4QCQ EEkAEMkWGMcBkSYl8U3vDE7Mmx6ceiuIxAci\n8Skjke5tDTM rYVIfCASnzIS6bZNnK0gEh IxMdJIgXW1VW1niLy3VE9d/MouXevnD73RaKB\ndXWe5qPU6q6sDqam46mcuxxAJD4QiY TRPKuC0Ufr43SsrsC5xZtIfeNR0ceq4qQ99bg6Naa\nQVqtnpH0aT3nLkU8TWwgLpywb1x4H9GQ8C7iDhfJOI6AX/xYDUesakm4SKeUgKrEzM4e5QJR\nzwFyP010QentUc4QRatf1EQypvWcuxTItSsM2SoUh zRKwzRuXZEnUqC7S9XJKVNZdklXzY \nqYlkTOs5dykiQYOB/qBw/L4B4X0M IV3EeyXrUJxpI jFO95wKKG0pfsokTqUqKJ599UTCRF\n2p8h0sJtaZHUaT0M5XJwj8TH4Zd2xnHgHimLM8px9fECnVZ6iU48q8tyWruCi3g6NJGMaYhk\nBRCJj4NEohXNF2PtVX7yrnz77JLNhizNq0KRB2qHEx826NMQyQogEh8niRS4u2ZG8xH1/b29\nqvb iCHLxdtr56w SwmR9GmIZAUQiY TRLIEiMQHIvGBSAKASLbAOA6IJACIxAci8YFIAoBI\ntsA4DogkAIjEByLxgUgCgEi2wDgOiCQAiMQHcVx8IJIAIBIfiJQXiFQ2Ism OCuK9HFAJAFA\nJD4QiQ9EEgBEsgPp44BIAoBIfCASH4gkAIhkB9LHAZHyov3n5oUBkfhAJD6OEym0muiOQKZI\nh3L9tUReIBIfiMTHcSId3Ea0LOuMtLrdzPYQiQ9E4uM4kZ7sosAaTaSo8nxrw/VdtLxi lI9\nx25CeaFhg5Zn95jSR9T6VI7tIRIfiMTHYSLt8M5a4l1Q7w0qB2PKkgHaUxmkBvWMpOfYuZtO\njWzcSPTdxhcpOh1/IVscEImPw0Si FKiR7q1M1JM2aFqUbM7KVIqx86tnoQ6b6BI9Y6NdGiO\n8QWX4f40vn7xlEkfzs61K/FYWdVH ge9UJF616uXbNGkSK orxe2JUVK5di5O4lCFYEDLb3z\nads9xlaRgTT9A8Lx 8T3ESjBcTg71y7jOErwfvj9FjU0WBKRbm6Y522q8T6UEEmVh Y/kRQp\nFXaSCNZvennrtvhsfwuyv4sEl3Z8nHZpd18vHd1CyTPS00RRz97LRXp0S1M3rdkzLZCrAYjE\nByLxcZpIzTHa1ZESaVFvtG3aIDVuHcoW6fACT5R tvjGnA1AJD4QiY/DRIq1Em04kxKp/Xue\nhgNEOz312SKNV7cQnVAeztkCROIDkfg4TKQMCvyWEETiA5H4QCQBQCQ7kD4OiDQZEEk4EImP\nc0UqEIjEByLxgUgCgEh2IH0cEEkAEIkPROIDkQRQNiIhjosNRBIAROIDkfICkcpGJNlXZ8WQ\ncRwQSQAQiQ9E4gORBACRbEDGcUAkAUAkPhCJD0QSAESyARnHAZFMcmLe9GBqMqZ05/4OEUTi\nA5H4lIdIybA7urc1rH9eGz8chkjFApH4lIdIybA7um1T5kyIVCwQiU95iJQMu2t1V1YHe2 Z\nVbPyPC7trAAi8SkHkfSwu9Bq9Yy0YH1k M5miGQFEIlPOYikh92RJlJ4lGhfZTxbpJFAmv6A\ncPy EvRRguNwdBxX5nGU4v3wW9RQ0Cjr0ouUCrtLiHRo dy5M5VYtkgIiCysC9kyFEOpx8ph\nAZG50MPuNJHOT98epa5LRcoAl3Z8cGnHpywu7VJhd5pInW61hh FSJYAkfiUhUipsDtNpOPK\nkbGXWpWLEMkCIBKfchBJD7tL3CP9ZNbsjeGmmechUvFAJD7lIJIpIBIfiMQHIgkAItmAjOOA\nSAKASHwgEh IJACIZAMyjgMiCQAi8YFIfCCSACCSDcg4DogkAIjEB3FcfCCSACASH4iUF4hU\nNiLJvjwrmOzjgEgCgEh8IBIfiCQAiCSb7OOASAKASHwgEh IJACIJJvs44BI Ykp3YVtCJH4\nQCQ ThUpfjhMh04WsCFE4gOR DhVJI3V7QVsBJH4QCQ ThVJvbRbXjF9aUQ5THReOR9XXlrZ\neH2HOmbr6qpaTxHtXeiZuzmaY0OIxAci8XGwSNTQTrpI5G4K0u6qCHnXhaKP10YvVHRPXFiy\nPceGEIkPROJTNiLtJOpT3jylBNT7p5mdp5Qe1Rlj7UjQYKA/KBy/b0B4HwN 4V0EnZtrd8lx\nlOI9D1jUUPoHpByR9qk/eJSTnUqC7fFNlc1t54y1h32gEGT7UDCyB64IAvJEOquJtD8hUpei\n3xf1PbeqslNfO54mNhAXTtg3LryPaEh4F3HnXtplH0fAL36shiNWtSRFpFjFQaKDaZHOKMfV\nhRcopgW/bmnNsSHukfg4V6Ts48A9Un40kRq3DtH8h2n0trRItKL5Yqy9yr n/mQ8sHxjjg0h\nEh IxMfJIu301NPrN8z73mvKW4ZIgbtrZjQfoXhb/bS6 4ZybAiR EAkPk4VqWAgEh IxAci\nCQAiySb7OCCSACASH4jEByIJACLJJvs4IJIAIBIfiMQHIgkAIskm zggkgAgEh/EcfGBSAKA\nSHwgUl4gUtmIJPsKrWCyjwMiCQAi8YFIfCCSACCSbLKPAyIJACLxgUh8IJIAIJJsso/D8SJd\nlcF7Le9nKvqU3hxzIRIfiMRHqEg1Kp  4queyi  6yuLLe9nKiBSsUAkPqIv7bZ/7rz2dPwz\nOy3vZyogUrFAJD6iRfrc/5t8fuD/tLyfXJy 0bP0sPIG9XirFr8EkYoEIvERLdJ79yaft7/P\n8n5yEK /d/iNJqU33rA cnE5RCoSiMRHtEgfn5N4itf8qeX95OCY8jbRHqX3mHKBaH9apGjY\nIOQPCyfgCwnvYzAgvIuwX7YPBZN9HP394sdqIGhRQ mvTGWJdKvr8zetWbPos65ciT6W0 mO\nq5d3Sm9nxQTRm2mRkGtXGLJ9KBjZA1cEeXLt4nf/qUvl6lvEX yovDRdfehVejsqkkIZexEz\nGAvEhBPyRYX3MToovIvYgGwfCib7OAJ 8WMVHraooXQ 8CW/kI2/ UpXzwSVhMOKejXcofQe\n0i7xOnGPVCS4R Ij h7pq89Z3v4kjM95IHqmWemNzl4fPtsCkYoEIvERLdI191re/mT8bnF1\nS7dyhk4s9Sw qLyRYw2IxAci8REt0o7P/nzM8h7yExvXPrqb7G/FIBIfiMRHtEhf 7zrvR//\nhIbl/eQgPn/DUODWST8ghEh8IBIf0SL90ze/lcLyfnLxxorqOWt9k60BkfhAJD74MwoBQCTZ\nZB9HeYjU/ yDP36 BO86E4jEByLxES3ShPcK7ReyV62zvJsCgUh8IBIf0SKtc03b2v7sj/7N\n9ajl/RQGROKDOC4 okX67LLk8/wvW95PYUAkPhCJj2iR3teRfH7uSsv7KQyIxAci8REt0lW7\nks / KDl/RSGxSL9b2UNRLxVdSAAACAASURBVGIjWqR//pfEfzMe fY3LO nMCCSCSASG9Ei\nPfeuv1iw5rZ5H3/3Hsv7KQyIZAKIxEb475F /hnt4 /Pl/RL4JMBkUwAkdiIFukk0blXD7yd\nc8UT86YHLe99KiCSCSASG9EiuT7VuOOyz5dCq4nuCNC9rWHjjTp00vL9yA1EMgFEYiNapB9O\nv9p1xTfu/G3WO3JwG9Eyots2pWetbrd8P3IDkUwAkdiI/65d/PBGVaaP1WbMerKLAmuo1V1Z\nfUE5THReOb 8YvrSSGp6QnmhYQMF1tVVtZ4i8t1RPXfzKA2sq/M0H6W40rG8obFn6011z2RN\np5e/tLLx g6ivQs9czdHc wdRDIBRGJTmm9/n1v50fTcHd5ZS7wL6r2h1ZtIl4ca2tPT7qZT\nI RdF4o XhulZXcFzi3aQt5bg6NbawbJ3RKZaJ25jw66g1nTxvKmIO2uilyo6J64sGR7jr2D\nSCaASGzEi3Subf6nXR/69u3pOfGlRI90q5dz UR6iuiUElBXnNnZoyXU9RzoUc4QRatfJLd6\nCfhIPakrH8 cTi/fqWUVv3lK6VGdMXocGzII 4esRHapi8UXtnSwcjE4ILyLIX / D4GghY1\nNJJbpO/8d9fVlT84kPWVmd71RK3RSUTq1CKAEmx/uSLxQzH51PgkuV8havMSxdSVM6bTy/ep\nP4CUk/FNlc1t54wexeXayS51sQgbNpCXPLl2rqtbj15yxrq5YZ63qcb7kCHS2QyRtGn3fqIu\nJXmD85uKibRIC7eR 9UMkV69RCRt f6ESOpp6blVlZ16l7Exg2hgzEpkl7pYfFFLBysXw2Hh\nXYz5/eL7CA9Z1VJukR6r/6TrY9WbjmTOu6 Xjm6hxBkpVnGQ6GBSJGNak GMeulGdCGZ8nji\n2dPKm olnKcjn0jp5UmRYtovqLbkCm/APZIJcI/EpgQfNrzx8NxrXB tSs9ojtEu7Uvhqkg0\n/2EavU2Vp3HrkDGtyUArmi/G2qv85F359tklm6l5VSjyQO1wPpHSy5Mi7ak/GQ8s35hj7yCS\nCSASm9J8and6xdXpuTH1RLHhDCVFev2Ged97TXmLdnrqjemESIG7a2Y0qyey4O1VtfdH6OLt\ntXNWn6W8IhnLkyLF2 qn1d03lGPvIJIJIBIb4SLFf//A7GtcV377 5b3UxgQyQQQiY1okaZd\n7XrXF5v3jFreS6FAJBNAJDaiRfqza3/aZ3kPxQCRTACR2CDXrjhkl7pYIBIbiFQcsktdLBCJ\nDUQqDtmlLhaIxAYiCQApQnwgUl4gEkTiA5HyApEgEh IlJd3lEiy72LEApHYQKTikF3qYoFI\nbCBSccgudbFAJDYQqThkl7pYIBIbm4tkffQdRDIBRGJjV5FyRN/lwn9P7YzWE0The6 dvVr7\nUt9Zrzux4LnvTFv8aq4tIJIJIBIbu4qUI/ouF00tPee/PydCa1pOn7tn0QR11q1PiLS37kDf\nL bleo8gkgkgEhu7ipSOvgumIuuScXepmLtU6l1o7Rmii8offBU96lmpsps6Lu5PiDSvI1/D\nEMkEEImNPUXKjL7LiKw7NWLE3KVS7zSOuQP7PNo7vvgp9SEhUr/ScWOV91iupiGSCSASG3uK\nlBl9l46sUz1JxdzpqXfaqqHGn9Dz12lTKx6klEgnlP94K/TgTONzivGIwYg/YiWyS10svhFL\nBysX4UHhXUT8/eL7GAxZ1FD6b2CtECkj i4dWdepx3IZqXfq5FvzN8fp fpLRVIljM3aqzeH\nXLvCEDZsIC95cu0KIzP6LiuyLhVzp6feEXXP1v5vza7kpd12XSSflmxHi4zIYpyRCgNnJDY2\nPSNlRN9lRdalYu701Dv6/azXtNX9Fao4g 4jukgTdape0RmdOVrGPZIJcI/Exqb3SJnRd5mR\ndXrMXSr1LjrvCe1sGKE7l54 u2pZnAK 3W7t9fY5r/s21uU6NohkAojExp4iZUXfZUbW6TF3\nqdS77uS90rM0vL5uzlr16rIh8XoHTTw6d1rrmVxNQyQTQCQ29hRJIBDJBBCJDUQqDtmlLhaI\nxAYiFYfsUhcLRGIDkYpDdqmLBSKxgUjFIbvUxQKR2ECk4pBd6mKBSGwgkgCQIsQHIuUFIkEk\nPhApLxAJIvGBSHl5R4gk  6lNEAkNhCpMGSXeGmASGwgUmHILvHSAJHYQKTCkF3ipQEisYFI\nhSG7xEsDRGIDkQpDdomXBojExuYipXIiLQQimQAisbG5SKmcSAuBSCaASGxsLlIyJzKudCxv\naOzZelPdM0SpvMhfJf4Wduv1HUSPKX1ErU/pSzJnXQ5EMgFEYmNrkYycSHdLZKJ15j466A7q\neZHxWCy2v7pn40ai7za SNHpJ/UlmbNSxMYMooExK5Bd4qXBF7VksCZjOCy8izG/X3wf4SGr\nWhIgkpET6W5Xn uJIspxIy S6FzNr6jzBopU79hIh bE9SUZs/R2rM 1k13ipcHyYQNTYmmu\nnY6eE l hajNq55ZlMNGXiRFFj1EFKoIHGjpnU/b7jGSJDNm6YyNGAz7R6xAdomXBt wJYM1\nGaFB4V2M PvF9zEYsqih9CWidSIZOZHuVy8RaeE2orv Q/tqaNPLW7fFZ/tbOtJJkulZOcA9\nkglwj8TG1vdIRk5khkhGXuTT9Ylo70e3NHXTmj3TAukkyfSsHEAkE0AkNvYWSc JzBBJz4t8\n3XMsFotN0OEFnij9bPGNlE6SzJh1ORDJBBCJja1FMnIiM0VK5UWuS3z8fRONV7doofkPUzpJ\nMmPW5UAkE0AkNrYWSQQQyQQQiQ1EKgzZJV4aIBIbiFQYsku8NEAkNhCpMGSXeGmASGwgUmHI\nLvHSAJHYQCQBIEWID0TKC0SCSHwgUl4gEkTiA5Hy8o4QSfbdS2mASGwgUmHILvHSAJHYQKTC\nkF3ipQEisYFIhSG7xEsDRGIDkQpDdomXBojEpoxFOjFvevDyuRDJBBCJjVNEKiDg7t7WcI46\ngEgmgEhsnCJSAQF3t23KNRcimQAisXGKSMmAO/LdUT138yjtXeiZuznqfUBdcLiiPzUzFWsX\nV15a2Xh9B7W6K6txaVccEImNM0QyAu6W3RU4t2jLhYruiQtLtu sU9/nzTdTcqYea0fupiDt\nrorQ6vQZaWLcIBoYtwLZJV4afGOWDNZkDIeFdzHu94vvIzxsUUPpL8WIOCOlAu56lAtEPQdO\nKT2qHDTgPkITtXtTM43AO/dOoj7lzUyRkGtXGJYPG5gSIbl2BqmAu2TiFsU3VTa3nSO65UHq\n9oykZhqBd 596slcOZkpUjRsEPKHrUB2iZcGX8iSwZqM4IDwLsL9/eL7GAha1FD6UleASHrA\n3W8qJpIz p5bVdlJe vj96 j1Ewj1s69/zKRMsA9kglwj8TGGfdIesDdaaWX6MSzMe1DhC2t\nNOw5NvuAPtOItYNIlgGR2DhEJD3gzrvy7bNLNu pPxkPLN9IdOfNtTF9phFrB5EsAyKxcYZI\nRsBd8Paq2vsj8bb6aXX3DRHtU35kzDRi7SCSZUAkNs4QyUIgkgkgEhuIVBiyS7w0QCQ2EKkw\nZJd4aYBIbCBSYcgu8dIAkdhApMKQXeKlASKxgUiFIbvESwNEYgORBIA4Lj4QKS8QCSLxgUh5\nKXuRZF9wlQ6IxAYimUd2eZcOiMQGIplHdnmXDojEBiKZR3Z5lw6IxAYimUd2eZcOiMQGIplH\ndnmXDojExqYiFRBjl Ss16093aioVOdaASKZACKxsalIBcTYJeisW58QqX6Xz fLOXwQyQQQ\niY1NRcqOsct 6r1lVs3K8xRXOpY3NPZsvanuGYoqz7c2XN9F1HFxf0KkqgP5WoZIJoBIbGwp\n0iUxdpT9tGB9ZPjOZiJ3S2SideY OugOxpQlA7SnUotySIg0pmxccv3as0aD8TSxgXixyC7v\n0uGbKHqwpiIyLLyLeMAvvo/hiFUtWSjSJTF22U8UHiXaVxknd7u6Uj1RRDkeU3ao55qa3bpI\nwbk/OHFi1dwhvT1rc 1kl3fpsHTYAAtLc 2yY yyn jQ8rlzZyoxcr9C1OYliimHY4o6TQvb\ndJESjFTv1icjQYOB/mCxyC7v0uEbKHqwpiLgF95FsL/493xK/AGLGkp/h7J4kS6Jsct Oj99\ne5S6NJFezRBpn7pk/hNZIlFjW462cY9kAtwjsbHlPdIlMXbZT53uGNGjl4r0NFHUs5dSIvX \ncFw9DVW/mKNpiGQCiMTGniJlx9hlPx1Xjoy91KpczBZpUW 0bdogBXy73T5fJDR7/YWza tH\nczQNkUwAkdjYUqRLYuwuefrJrNkbw00z 7JEav ep EAUYP2i1hlB/XcXFO75u1cbUMkE0Ak\nNrYUyTQx5SB3VYhkAojEBiKZR3Z5lw6IxAYimUd2eZcOiMSmPEQyAUQyAURiA5HMI7u8SwdE\nYgORzCO7vEsHRGIDkQSAOC4 ECkvEAki8YFIeSl7kWRfcJUOiMQGIplHdnmXDojEBiKZR3Z5\nlw6IxAYimUd2eZcOiMQGIplHdnmXDojExu4imfjyDw IZAKIxMa IiWj7aYQyX9P7YzWE0Th\ne6 dvbqPjFw7eu470xa/mmsLiGQCiMTGviIlo 2mEKmppef89 dEaE3L6XP3LJowcu321h3o\n 8W8XO8RRDIBRGJjX5GS0XYx5ZctnoVdRAPr6jzNR1VFFnrmbo6mMu5Ca88QXVT 4KvoUc9K\nld1Grt28jnzNQiQTQCQ2dhVJj7YLKguODrdV9pH31uDo1prBCxXdExeWbNcz7jSOuQP7PNo7\nvvgpSmU29CsdN1Z5j VqGCKZACKxsatIerRdTHmSaKyqvUdRzz3R6hdPKeq5Z0JPvNNWDDX \nhJ6/Tpta8SClRDqh/MdboQdnBvXGhvvT PqLRXZ5l47ix2pqyqQPn1V9pH/QWyNSKtouprys\nvpj/WDLUrvHJ KbK5rZzRsYd0VvzN8fp fpLReomis3aqzcWGUjTP1Asssu7dPgCRQ/WVAT8\nwrsY6PeJ78Nv1XEMWiuSHm0XU7TP3hY8kTRn4TaivudWVXamMu6IumfvUh 7kpd223WRfMpJ\n9XHR9hwN49LOBLi0Y2PbS7tUtF1M TnRePWe08qb6onF0xHTLte2tKYy7uj3s17TVvZXqOIM\nuo9QSqSJOlWv6IzOHO1CJBNAJDa2FSkVbRdTGnvHt3tC1LwqFHmgdnhP/cl4YPnGVMZddN4T\nWlxyhO5cevrsqmVxPdeOts953bexLtexQSQTQCQ2dhVJj7aLKh3Nnkb1tHPx9to5q89SvK1 \nWt19Q6mMu 5EjJ3yLA2vr5uzNpDOtZt4dO601jO5GoZIJoBIbOwqkjAgkgkgEhuIZB7Z5V06\nIBIbiGQe2eVdOiASG4hkHtnlXTogEhuIZB7Z5V06IBIbiGQe2eVdOiASG4gkAMRx8YFIeYFI\nEIkPRMpLuYsk 3qrhEAkNhDJNLKru4RAJDYQyTSyq7uEQCQ2EMk0squ7hEAkNhDJNLKru4RA\nJDZOESlnmtAlM1kZeBDJBBCJDUQyjezqLiEQiQ1EMo3s6i4hEImNc0RK5dsF1tVVtZ6iCeWF\nhg2aObGbV8VS87SXqcnmzeo2xyv6Lm8IIpkAIrFxjkh6vt26UPTx2ii5m06NaOasbx7V52kv\nU5N7ZkaJHlqeoyGIZAKIxMY5IiXz7U4pAaL4zE5yP5W4ltvWGCJ9nvpSn4zMeIni172obz0S\nSNMfKBLZ1V1CfMWO1dT4/eL76HfScRhhjKJESubbdSZTGraTu1ObeZfyOyJ9niqSsfj W jI\njFF9a0sDImVXdwkpk/DGd3RA5GUiJfPtupRocoZ7vzazYWVTjPR5qkjG4hMV/gd mKshXNqZ\nAJd2bJxzaZfMtzujHFdfXdBFOhCuf5j0eapIxmK68Znao7kagkgmgEhsnCNSKt9uRfPFWHuV\nPyXSQTpS Vt9nvZSX0w7a27I2RBEMgFEYuMUkYx8u8DdNTOaj5AhEj0 N5ial/j4O7WYwtNy\nBRZDJFNAJDZOEck0vZ7cxkAkE0AkNmUq0kTfsh/nWQKR EAkNmUqUlv1 mjuJRDJBBCJTZmK\nlB IZAKIxAYimUZ2dZcQiMQGIplGdnWXEIjEBiIJAHFcfCBSXiASROIDkfJSviLJvtIqORCJ\nDUQygey6LjkQiQ1EMoHsui45EIkNRDKB7LouORCJDUQygey6LjkQiQ1EMoHsui45EImNE0QK\nrSa6I3DZ7EMnifz31M5oPUEUvvfa2au1zKCzXre27EbtD86rczUGkUwAkdg4QaSD24iWXT57\ndTtRU0vP e/PidCaltPn7lk0QZ116xMi1e/y Xw5hw8imQAisXGCSE92UWAN0ekbPUsPK2/o\n0XXLK6YvDa09Q3RR YOvokc9K1V2U8fF/QmRqg7kawwimQAisbG/SDu8s5Z4F9R7B vvHX6j\nSek1ku0a2pMrHHMH9nm0d3zxU pDQqQxZeOS69eezdUcRDIBRGJjf5EovpTokW46prxNtEfp\nNZLtUiKFGn9Cz1 nTa14kFIiBef 4MSJVXOH9CYiQYOB/mDByK7rkuMbKHywmAT8wrsI9hfx\nnnPxByxqKP3VL6tF6l1P1BqlTrf64/G00mtE1yVFemv 5jg9X58tUoKR6t365LDPEmTXdcmx\nZtiAGdKfqlks0s0N87xNNd6HXpquSaX0GtF1CZG6Z 9SH7uSl3bbs0SixjZ9Kp4mNhAvGNl1\nXXJ8E4UPFpPIsPAu4gG/ D6GI1a1JEokuq Xjm4hOqyoV7odSq8RXaeJ9PtZr2mr CtOEg26\nj gi9f5wXL2eq34xR2u4RzIB7pHYOOAeqTlGuzqIxuc8ED3TrPQa0XWNW4ei857QzoYRunPp\n6bOrlsUp4NvtVl HZq /cHZt/WiO1iCSCSASG/uLFGsl2nBGnfjd4uqWbuWMEV2301Pfnbxf\nepaG19fNWateXTYkXu gnptrate8nas5iGQCiMTG/iIZxNTLtWNKseMOkUwAkdg4R6T4/A1D\ngVtbi20GIpkAIrFxjkj0xorqOWt9xbYCkUwAkdg4SCRrgEgmgEhsIJIJZNd1yYFIbCCSCWTX\ndcmBSGwgkgAQx8UHIuUFIkEkPhApL UpkuyrLClAJDYQiYnsmpYCRGIDkZjIrmkpQCQ2EImJ\n7JqWAkRiA5GYyK5pKUAkNs4WKfHfmpsDIpkAIrFxhEh5gu1yiSQs1052TUsBIrFxhEh5gu1y\niCQu1052TUsBIrFxhEipYLu9Cz1zN0fjSsfyhsaerTfVPaOK9MsWz8Iu0rPuBObaya5pKUAk\nNg4QSQ 2O1nRPXFhyXZyt0QmWmfuo4PuYExZcHS4rbLPyLoTl2snu6alAJHYOEAkPdjulNKj\nekDkbldf1hNFlOMx5UnVmqp2I suT65dNGwQ8ocLQnZNS8EXKmywTBAcEN5FuL9ffB8DQYsa\nSv9gsVykVLBdfFNlc9s5VaRXiNq82g3S4Zjysrp8/mNG1p24XDvZNS2F4ocNmEVYrp0RbEfU\n99yqyk5yv5ohkjpNC54wsu7y5drFDMYCsYKQXdNS8I0XNlgmGAkL7yIW8IvvIzxsUUMTwkTS\ng 1iQXV6S2u2SD8nGq/eY2Tdicu1k13TUsA9Ehsn3COlgu321J MB5ZvzBapsXd8uydkZN2J\ny7WTXdNSgEhsHCCSHmwXb6ufVnffUKZIUaWj2dP4GhlZd Jy7WTXtBQgEhsHiGQtEMkEEIkN\nRGIiu6alAJHYQCQmsmtaChCJDURiIrumpQCR2EAkJrJrWgoQiQ1EYiK7pqUAkdhAJAEgjosP\nRMoLRIJIfCBSXiASROIDkfJSbiLJvlGRCERiA5GmRHY1SwQisYFIUyK7miUCkdhApCmRXc0S\ngUhsylWkmNKt/suxACKZACKxKVeR4ofD6r8cCyCSCSASm3IVKS8QyQQQiY2jRUql2WUm3dHA\nujpP81Fc2lkCRGLjaJH0NLuMpDvy3hoc3VozCJGsACKxcbJIRppdRtJdj3KGKFr9YrZIY0MG\nYf QSWRXs0R8YbODZZrBAeFdDPn7xfcxELSooZHSi2Sk2WUk3b1cof0QbXwyW6Sicu1kV7NE\nrMpqA3zE5drlxUizy4hDSYq0cFu2SLExg2hgzCSyq1kivqjZwTLNcFh4F2N v/g wkNWtVR6\nkYw0uwyRTitvqpd4ng7cI1kB7pHYOPkeyUizyxCJmleFIg/UDkMkK4BIbBwtkp5mlynSxdtr\n56w i4 /LQEisXG0SJOQ2yKCSKaASGzKVKSJE9r/AZNzCUTiA5HYlKlIL1WuyVMDEMkEEIlN\nmYqUH4hkAojEBiJNiexqlghEYgORpkR2NUsEIrGBSAJAihAfiJQXiASR ECkvEAkiMQHIuWl\n3ESSfaMiEYjEBiJNiexqlghEYgORpkR2NUsEIrGBSFMiu5olApHY2F6kmHLQ0vYgkgkgEhsb\nixRaTXRHIFOkQyeLbxUimQAisbGxSAe3ES3LOiOtbi  VYhkAojExsYiPdlFgTUJkXpvmVWz\n8jwtr5i VM ym1BeaNhAp2/0LD2svJF8kVorK XuciCSCSASG9uKtMM7a4l3Qb03qIq0YH1k\n M5moob2jCy7plMj8fp7h99oUnoTL4y1MlPuLgcimQAisbGtSBRfSvRId KMFB4l2lcZ10RK\nZ9k9RXRMeZtojyaS sJYKyPlTm9rPGIw4o YRHY1S8Q3YnawTBMeFN5FxN8vvo/BkEUNjVot\nUu96otZoQqRDy fOnanENJHSWXadRJ1u9SfmaU0k9YWxVkbKnd4Wcu0Kw6qsNsDH6ly7mxvm\neZtqvA pIp2fvj1KXUmR0ll2 4lemq4Jp4mkvjDWyoxCSYEzUmHgjMTGvmek 3rp6JbE75E6\n3TGiR5MipbPsVHcOK rFb0dKJGOtHCJlgHskE AeiY1975GaY7SrIyHSceXI2EutykVq3DqU\nzrJT3Rmf80D0THNKJGMtiGQZEImNbUWKtRJtOJP8ZsNPZs3eGG6a2bfTU5/OslPdod8trm7p\nVs4kX hrQSTLgEhsbCsSi9i49tGdqbcCIpkAIrFxtEjx RuGAre2mtoGIpkAIrFxtEj0xorq\nOWt9pjaBSCaASGycLVIBQCQTQCQ2EGlKZFezRCASG4g0JbKrWSIQiQ1EEgBShPhApLxAJIjE\nByLlBSJBJD4QKS9lJpLs xSZQCQ2EGkqZBezTCASG4g0FbKLWSYQiQ1EmgrZxSwTiMQGIk2F\n7GKWCURiA5GmQnYxywQisXGiSMnkyNzLznrd2tOZNbNnLT WawWIZAKIxMaJIiWTI3PSWbde\nE2n82h cPb9 5kiONSCSCSASGyeKlEyOpIF1dZ7mo9mBkB0X92siBX mOnRW6cmxMUQyAURi\n4zyR9OTIkPfW4OjWmsFLAiETImmENi8c07eJjRlEA2PmkF3MMvFFTQ6WeYbDwrsY8/vF9xEe\nsqqlUomkJ0f2KGeIotUvXhIImRJpYrryH/3GJsXk2skuZplYFtYG2Fida5efVHLkyxXadUfj\nk5cEQupnpLcO3zk/rG8yNmIw7B8xh xilolv2ORgmSc0KLyLEX / D4GQxY1lL5EFCySnhyZ\nFGnhtkvit4xLO5qoeTbH5rhHMgHukdg47x5JT448rbypXs55OnKJ9Nt5o ol4ByIVCQQiY0D\nRUolR1LzqlDkgdrhLJECvt1uny8SnnvXmQsPei7k2BoimQAisXGeSHpyJF28vXbO6rOUJVJD\nImJ/B/XeWj3ju925NodIJoBIbJwnUpFAJBNAJDYQaSpkF7NMIBIbiDQVsotZJhCJDUSaCtnF\nLBOIxAYiTYXsYpYJRGIDkQSAFCE ECkvEAki8YFIeYFIEIkPRMpLOYkk yZFMhCJDUSaFNmV\nLBmIxAYiTYrsSpYMRGIDkSZFdiVLBiKxgUiTIruSJQOR2JSxSDEl19e/IZIJIBKbMhYpfjic\nYy5EMgFEYlPGIuUGIpkAIrEpL5H2LvTM3Rwl3x3VczeP4tKuaCASm7IS6UJF98SFJdtp2V2B\nc4u2ZIg0MW4QDYybQHYlS8Y3ZmawCmI4LLyLcb9ffB/hYYsaSn8pRp5Ip7Rs1QnqUS4Q9RzI\nEKngXDvZlSwZK/PaAI/S5drlJ76psrntXCrxLvNTu2jYIOQPm0B2JUvGFzIzWAURHBDeRbi/\nX3wfA0GLGkpf6sr8sKHvuVWVnb pmNCmcY9UNLhHYlNW90gxLfx7S tppZfoxLMQqWggEpuy\nEmlP/cl4YPlG8q58  ySzRCpaCASm7ISKd5WP63uviEK3l5Ve38EIhUNRGJTViJxgEgmgEhs\nINKkyK5kyUAkNhBpUmRXsmQgEhuINCmyK1kyEIkNRJoU2ZUsGYjEBiJNiuxKlgxEYgORBIA4\nLj4QKS8QCSLxgUh5cbBIsi l7AZEYgORMpFduHYDIrGBSJnILly7AZHYQKRMZBeu3YBIbCBS\nJrIL125AJDZ2FymmHLSyOYhkCojExr4ihVYT3RGYQiT/PbUzWk8Qhe 9dvbqPnXGWa9bfTys\nJHg2xxYQyQQQiY19RTq4jWjZVGekppae89 fE6E1LafP3bNogjrr1msijWkpEr vPpNjC4hk\nAojExr4iPdlFgTWqSL9s8SzsIhpYV dpPnpJdF1orerKReUPvooe9axU2U0dF/e79QZubsvV\nLEQyAURiY1eRdnhnLfEuqPcGlQVHh9sq 8h7a3B0a81gdnRdYtVj7sA j/aOL35KfTBE6mwY\nN1qLp4kNxPMju3Dthm9iksGyhsiw8C7iAb/4PoYjVrVkrUgUX0r0SLd6RnpSvVKrau9R1HNP\ntPrF7Og6bcVQ40/o eu0qRUPZog0sWBPujFurp3swrUbAuPbQB6szrXrXU/UGlVFell9Mf x\nZFZd45PZ0XUqb83f5/Ho9QAAD2tJREFUHKfn6y8TqfO6jO 7RYIGA/3B/MguXLvhG5hksKwh\n4BfeRbB/svfcIvwBixpKf4fSEpFubpjnbarxPhRTXlVfLXgiac7CbdnRdUTds3epj13JS7vt\nGSKtfjB3w7hHMgHukdjY9R6J7uulo1u03yP9nGi8es9p5U31xOLpyI6uo9/Pek1b2V9xkmjQ\nfYQMkYYqc0UIEUQyBURiY1uRmmO0q0MTqbF3fLsnRM2rQpEHaoezo ui857QLisjdOfS02dX\nLYtTwLfbrb2mbqUvd7sQyQQQiY1dRYq1Em3QPl9QOpo9jepp5 LttXNWn70kuq5b/8Xr8Pq6\nOWvV27SGxOsdRL qGM/dMEQyAURiY1eRhAGRTACR2ECkTGQXrt2ASGwgUiayC9duQCQ2ECkT\n2YVrNyASG4iUiezCtRsQiQ1EykR24doNiMQGIgkAcVx8IFJeIBJE4gOR8uJQkWRfRtkRiMQG\nIunILlo7ApHYQCQd2UVrRyASG4ikI7to7QhEYgORdGQXrR2BSGzKRaQT86YHWStCJBNAJDbO\nECmZcTfZGve2hi990w dzLUiRDIBRGLjDJGSGXeTcdumy2atbs 1IkQyAURi4wyRkhl3caVj\neUNjz9ab6p4xcu5 lfhLvq2t7srqYGBdXVXrKT36bnnF9KU5moJIJoBIbJwgkp5xF3K3RCZa\nZ 6jg 6gnnMXj8Vi 6t7aLV6RvKuC0Ufr43q0XcNOCMVC0Ri4wSR9Iw7cqtqPFJPFFGO6zl3\n6tJzNb8iTaRTinoXFZ/ZmYq yxRpuD Nrz83sovWjuQbKyspkz58VvWR/kFvvUipjDtyv0LU\n5tUSUQ7rOXeqVYseooRInckAh 2p6LtMkSIDafoHciO7aO2IL5BnsKwj4BfexUC/T3wffquO\nY1CcSHrGHblfvUQkLefurv IJUXqUqKpDRLRd7i0Kx5c2rFxxKVdKuMuUyQ9546erk/8/kgV\n6YxyXJ24QKnoO4hUPBCJjSNESmXcZYqk59y97jkWi8UmEh82rGi GGuv8uvRd41bh3K0BZFM\nAJHYOEEkPeMuS6RUzt26xH3RTQmRAnfXzGg YkTf7fTU52gMIpkAIrFxgkiWApFMAJHYQCQd\n2UVrRyASG4ikI7to7QhEYgORdGQXrR2BSGwgko7sorUjEIkNRNKRXbR2BCKxgUgCQBwXH4iU\nF4gEkfhApLw4UCTZl1B2BSKxgUgasgvWrkAkNhBJQ3bB2hWIxAYiacguWLsCkdhAJA3ZBWtX\nIBIbiKQhu2DtCkRiY3uRpo60MwdEMgFEYmN7kaaOtDMHRDIBRGJje5GSkXZ6jF1Ueb614fou\ndWySGXZx5aWVjdd3EPnuqJ67eVRfzZh9ORDJBBCJjc1FMiLtUjF2MWXJAO2pDOoZduRuCtLu\nqggtuytwbtEWPe3OmH05EMkEEImNzUXSI 30GLuYskN1oWa3nmFH7p1EfcqbPcoFop4DRtpd\narbeykggTX/gcmQXrF3x5Rgri/H7xffR76TjSP9fEJaKlIq002PsYsor6vPCNj3Djtz71B84\nysnkcmM1fbbeylQBkbIL1q6USXgjAiKNSDs9xi6mqIbQ/CeMDDv3/oQxv6mYSIu0cJs O0eL\nuLQzAS7t2Nj90i4VaafH2MWUp9VrN89ePcNON a00kt04lkj7Q4iWQJEYmN3kfRIu1SMXUxZ\n1BttmzaoZ9gZxnhXvn12yWZ9NYhkDRCJjc1FMiLtUjF2MaX9e56GA0aGnWFM8Paq2vsj moQ\nyRogEhubi3QpMeVgkS1AJBNAJDYQSUN2wdoViMQGImnILli7ApHYOEyk4oFIJoBIbCCShuyC\ntSsQiQ1E0pBdsHYFIrGBSAJAHBcfiJQXiASR ECkvEAkiMQHIuXF3iLFrT/gyxgdmhDeR2xU\neBcUyfV/h1rMeHTqdYplpASyjo1b3qS9RQLAIUAkACwAIgFgARAJAAuASABYAEQCwAIgEgAW\nAJEAsACIpLFX2S97F4okfO 1s1f3yd4LS3DmmwGRVAbmepz43mWypuX0uXsWif Ohngc mZA\nJJU7t8514nuXga iRz0rVXbL3g8LcOibAZGI9n0n4sj3LoN9Hu0bq4ufkr0fxePUNwMiUbju\ndXLke5fB89dpjyselLwbxePYN MdLdKv3W73UdqwgZz53mXwfL32WAYiOfbNeEeLNNzb2zv6\nel3Ioe9dBl3JS7vtsvejWJz7ZryjRUqwzjN79uyKGWtl70dR CtOEg26j8jej2Jx7psBkUI \nldrdg7L3ozjuXHr67Kpl4v9GVjDOfTMgUgJHXk1kMry bs5aC/8bbJk48s2ASABYAEQCwAIg\nEgAWAJEAsACIBIAFQCQALAAiAWABEAkAC4BIAFgARDLBP33axMo1V10 NSl/b6b1S j/xPVT\nr/Tbf/vQh7/ysPY1ooP/9uH3//1ziZnHvuz6dWJi7zc/8v4v/4RoxUfeyNhm7R Z y8GEjvC\n64hJ9v7YFohkAmEiva69DetNflXz9fR7N/HtL4xMuX7XFZ 6d8u/uO4lOvHBz6zb9PV3/1Kd\n cAHrknW9653f/mHm76qLo39y1cyIv /7jG1T4kdYXbEJHt/bAtEMoEwkTYW8jZkbPS46z n\nXv9/fqSPaPwzH4nT7A  TTT22b8m2vf zQ8l6/sLf6mqOPrfPkV09N3fN7YJXfGQqX1K7Ai3\nIyaZ 2NfIJIJEiK1f 2D7//re9Xrlolbr3nfl3cvvkJfaiyIr77mfZ/brumTnjr/nb94359M\nP6Y39LVd13yV6Im/vfJDX3mC6N9cLtdXkpd2RiMJvvwPHX975R/XBym9rtFScqMEsb/6etbK\nqeb/818/dOWXthp7/ M27bHeNRS7arY2dZfrGJ08RMn6jm3Ypc2b9R616xkfM/6HmJ 5ziQn\nvvbPv/3mhz46sy9jFz8 TX3xmOs7pDl9KmNH2B0xW87YH/sCkUygifTzd/37L/YuczUT3eGa\n8cKP//Tv9NNNesHdrjl7nvrcp6/KnPqH//rjF3/6 Y l/u fb37hM5uepSdd05599t9dz9If\n3K4DRxMipRtJ8NWP/s1vfI9fodaVsa7RUnKjBC 5Hs5aOdn83vd8fdfuBa5Lfpx/7RN0wrVO\nm2p3aV6m6jtJ/G8 qz4 5zKyH b9j9TEt/78b/f0Pf2eazN28dqrVRfqr/7v6tLpn8rcEXZH\nzJYz9se QCQTaCJ95i 0/2yr8or  J98Tn27u1y6SOkFH/ cOnH iqsoPTXoalWnTq09l1z3\nW66fqY9rv6luMfi/zyFq0N6Gv89sPdWhq5O0pWfS66ZbajDeu1tcZ7NWTjb/pb/UtK34UNZ/\n1tbm iH92pW41f ta4P2ZNT36Jv7at6nnS6G39ugr/4Xy0jf45e1x49n7OJPXb8n mSL6xzF\nr27M3BF2R8yWM/bHvkAkE6ginXMt0Ka2up4972rSpj6XEim94E3XTdrUV6 i9NTYf/nE3ozQ\nuW 9d8yYvuZraZHSjaQ6vEq7AHrU1Z5eN91SWqT/ 0 yV0403 daElHZ4no1 oZK8sOIXR/w\nxGm366fa9O9dd2pPRn3/0uX65AuJqS98MdXw71279T3 gPZ47bszdrHvXQ/QG66jV7fRYdeO\nzB1hdpTaLUbLxv7YGIhkAlWkV11rtKl214OvJ69bPCmR0gteSU6pC9JT9PJ/c/0Xz0/1/3FR\n /mrnmBu dyH3/Me1z lRUo3kurw/9Aen1V/sKfXNVpKi/R3f529cqL5110pfpaY0j45o43v\nmaUq9uvkBdhB133ak1HfF3c 7Hn37drUN69JNfz9K/XT2bc oT2qXWbs4hdn08Mfo8r5tPGK\nUOaOMDtK7RajZWN/bAxEMoEq0gHXam3qOdePu5J3H1UpkTIXJAqi8ipKT6k32R3f/R uv0l9\nSJ0snq /Z3nn4d99PEOkdCOpDhNu/ML1aMa6RktpkT7zj9krJ5p/3XX9/gS 4M9VtJv5pa4V\n2knrpCvxQfsu19PaU atC333XYfVx kfTL381/ lLzDKPWMXm6 hudX0g0/T9G9k7Qizo9Ru\nMVo29sfGQCQTqCJdcM3Xph50Pf H5EcCn0 JlF7Q41qkTX3xKkpPJdnseiQ5kSiek6556uP4\n zNESjeS6vBK7dehP3I9n7Gu0dJlZyR95WTzfte1l x 67uTfk58OPHLoZWuXu0pWd9v3/ q\n9uJp1zb18Zt/ltxg H336dsa5Z6xi3tcp6/ZpNb/ avvzNoRfkfMlvX9sTMQyQTahw2f 7h2\ntfPvHxgc/yPtk4RXjQ8bjAUTV39KvYk58a6rKD31Wo12Rjjluie5bqJ4jiZ Am90/QPRd1zj\nyQ8bjEZSHSbujirfF0ivm24psVGC1D2SvnKqNv/ujwbUx0dX6Gv90nV/aqrhyrdUSz6RPH0k\n69v37n/RziE3JF7o9yS7XMf1QzfKPWMXR69c4TpCsQ/d6vpt1o7wO2K2jHukMkMT6bl3f3vH\nLxdqt8/LXNe98KNP/pMuUnrBLa7pzzzwia9clTF14UNf2LrnyX/8cPLXLcniGfvzP9vxsvcb\n3/jQi0MrXaufToiUbiTZ4Z//1QN7vuuqy1j3lNFSYqMEK5Kf2ukrp2rzP6/4wqMv3HzFdam2\nxv/q6ocSnKM3P/KXd973d /dp97FPPTQta6Whx46Q62uf16/eYbrf8a1E1F9cpNFnzQOPV3u\nGbv47as/qj1 9GP674QSO8LviNmysT92BiKZIPEL2d3/fNX7vqTdRI/eePVVX3tltnH9biyI\ntf7X937 54vfmzl1aNrHrvj4NP0nd7J4Dnz1A39yw Cuq//4xFtfuuLTyV/IGo0kO/zMa1//\nwB/PC2eua7SU2CjBrxIfM6dXTjZPv/6/PnTFX63TT0g /cOHPerJUPnwB76hnRIaUvN2UXzz\nl6784OfXap Yp37vQ/SpBcahp8s9Yxe/76pSH2931eprJXaE3xGzZWN/7AxEKpJv/anAxpnf\nSRr/1Df4K0/NzKvDBW6Z3BGrKXx/SghEKpj109Uf9gN/9L mXrNguG48qv0u1iqRjr17XcHb\nJnbEYorZn9IBkQrmMdf/s Opr75rr8AuuG5M/OsXI1aJNPGtL0WmXmuyHbGWovandECkwnns\nS1d94B fE9kD2w3fJ663SqRb/vh0EVv7OH8YZYri9qdkQCQALAAiAWABEAkAC4BIAFgARALA\nAiASABYAkQCwAIgEgAVAJAAsACIBYAH/P0UsL5wSCwSXAAAAAElFTkSuQmCC",
      "text/plain": [
       "plot without title"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "word_ratios %>%\n",
    "  group_by(logratio < 0) %>%\n",
    "  top_n(15, abs(logratio)) %>%\n",
    "  ungroup() %>%\n",
    "  mutate(word = reorder(word, logratio)) %>%\n",
    "  ggplot(aes(word, logratio, fill = logratio < 0))  \n",
    "  geom_col(show.legend = FALSE)  \n",
    "  coord_flip()  \n",
    "  ylab(\"log odds ratio past (before cutoff) / now (after cutoff)\")  \n",
    "  scale_fill_discrete(name = \"\", labels = c(\"past\", \"now\"))   theme_minimal()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "What we see in my data below: My activity in the Pirate Party (and living in the state of North-Rhine Westphalia - short `NRW`) is clearly in the past and tweets about them occur mainly before 2013 - as demonstrated by hashtags like `#piraten, #lptnrw, #lmvnrw, #nrw` etc. My activity in the Open Science world is clearly still going strong in comparison: `#opencon, #mozfest, bosc201*, #csvconf` etc. are overrepresented. When you run the analysis on your data, what trends do you see? Next, let's see what would happen if we excluded hashtags."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "data": {},
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAA0gAAANICAMAAADKOT/pAAADAFBMVEUAAAAAv8QBAQECAgID\nAwMEBAQFBQUGBgYHBwcICAgJCQkKCgoLCwsMDAwNDQ0ODg4PDw8RERESEhITExMUFBQVFRUW\nFhYXFxcYGBgZGRkaGhobGxscHBwdHR0eHh4fHx8gICAhISEiIiIjIyMkJCQlJSUmJiYnJyco\nKCgpKSkqKiorKyssLCwtLS0uLi4vLy8wMDAxMTEyMjIzMzM0NDQ1NTU2NjY3Nzc4ODg5OTk6\nOjo7Ozs8PDw9PT0 Pj4/Pz9AQEBBQUFCQkJDQ0NERERFRUVGRkZHR0dISEhJSUlLS0tMTExN\nTU1OTk5PT09QUFBRUVFSUlJTU1NUVFRVVVVWVlZXV1dYWFhZWVlaWlpbW1tcXFxdXV1eXl5f\nX19gYGBhYWFiYmJjY2NkZGRlZWVmZmZnZ2doaGhpaWlqampra2tsbGxtbW1ubm5vb29wcHBx\ncXFycnJzc3N0dHR1dXV2dnZ3d3d4eHh5eXl6enp7e3t8fHx9fX1 fn5/f3 AgICBgYGCgoKD\ng4OEhISFhYWGhoaHh4eIiIiJiYmKioqLi4uMjIyNjY2Ojo6Pj4 QkJCRkZGSkpKTk5OUlJSV\nlZWWlpaXl5eYmJiZmZmampqbm5ucnJydnZ2enp6fn5 goKChoaGioqKjo6OkpKSlpaWmpqan\np6eoqKipqamqqqqrq6usrKytra2urq6vr6 wsLCxsbGysrKzs7O0tLS1tbW2tra3t7e4uLi5\nubm6urq7u7u8vLy9vb2 vr6/v7/AwMDBwcHCwsLDw8PExMTFxcXGxsbHx8fIyMjJycnKysrL\ny8vMzMzNzc3Ozs7Pz8/Q0NDR0dHS0tLT09PU1NTV1dXW1tbX19fY2NjZ2dna2trb29vc3Nzd\n3d3e3t7f39/g4ODh4eHi4uLj4 Pk5OTl5eXm5ubn5 fo6Ojp6enq6urr6 vs7Ozt7e3u7u7v\n7 /w8PDx8fHy8vLz8/P09PT19fX29vb39/f4dm34 Pj5 fn6 vr7 /v8/Pz9/f3 /v7////8\nl6Z1AAAACXBIWXMAABJ0AAASdAHeZh94AAAgAElEQVR4nO3dC5wU9Z3v/X5iHmOMMbs5Jns2\n625ynuzZJGeTPLnsLbubnGyyT3bPc2oamBlAGEjGWVFGvCCZMAcVRY0XEhdDFEwMJipBPWiy\nok4iOGSZGECRCIRwWRgYCILQl7lPP9NMT/ equ6u7p75/eY/VV3/nuqp b5fOtOX6v9UVdeH\n7mrgT4gAwLOQ3ysAEAQICUADhASgAUIC0AAhAWiAkAA0QEgAGiAkAA0qLKTuuIZBhno1DDIQ\nP69hlC4NYwzF zWM0jekYZBOHduTHNAwSF982PsgaR2bk1NhIXVGNQwy1KNhkP6IjpBiGsY4\nH nTMEqPjs2J6vh1blDL5kRS3gdJ69icHIQ0BoQkQEhjQkhjQEgChDQmhDSGSgrp/wOfOH W\nENIYEBIgJA2DICRASBoGQUiAkDQMgpCgAkJKGbtLehxC4hCSf5w/SwhpDAgJEJKGQRASVEZI\nTWvM74eqzrp6HELiEJJ/nD9L5Qtpy wk0SPL3D0OIXEIyT/On6XyhZSYuY3SX9tq35LodCIa\ncbSYWjyqYZBYJK5hFC2bg5B8M95z0z0BIdGDt9L mYP2Lf1RJyIRR4uNN0qgBokgJN M99x0\nTkRIh6tia7/r8nF4a8fhrZ1/nD9L5fzU7rpn6w64fBxC4hCSf5w/S UMadOsq90 DiFxCMk/\nzp lcobUO32j28chJA4h cf5s1TOP2vXUd05/kIjISQOIfnH bNUvpCGz970A9cPQkgcQvKP\n82epfCFtqF2VdP0ghMQhJP84f5bw1yjGgJAAIWkYBCEBQtIwCEIChKRhkEoKCdNxMZiOSw0h\ncQhJgJDUEBKHkAQISQ0hcThHGpu3PYuQ1BDSlOFtzyIkNYQ0ZXjbswhJDSFNGd72LEJSQ0hT\nhrc9i5DUENKU4W3PIiQ1hDRleNuzCEkNIU0Z3vZsxYaUNlqXNTS2r7t /rOUNH7e3HDlTpcj\nICQOIY3N256t2JAovDQx3Dx7O 0Od6WMGzppyzSX/240QuIQ0ti87dkKDqmF6Ef1RAnjUMp4\njmh41mb7LsxrV/IYCGlMfj895ZrXLvwq0YYl1twn 1KGeZkWbrDvGog7EY04WkwtFtUxSCSm\nYRQ9a L34Vq5/H56Cm 4NIf0WlFI280bFjzpbgC8tePw1m5s3vZsBb 1Kw7pGaJk9cvuBkBI\nHEIam7c9O0lCurYjuWF69/gPKoaQOIQ0Nm97dpKE1PKN6oZdLgdASBxCGpu3PVu5IRUp7d/s\nQ0gcQhqbtz2LkNQQ0pThbc8iJDWENGV427OTIqTSICQOIY3N255FSGoIacrwtmcRkhpCmjK8\n7VmEpBa4kDAdF4PpuNQQEoeQBAhJDSFxCEmAkNQQEjeFz5HG3ikISQ0hcQhJgJDUEBKHkAQI\nSQ0hcQhJgJDUEBKHkAQISQ0hcQhJgJDUEBKHkAQISQ0hcQhJENyQ0sa25Y1XttKw8VLDA Z3\netw4S9T8tLtREBKHkATBDYnCi7toc03C/H50YPVqoq83bqXkjCP23WlHOqPOllM636NhkP5I\nUsMoMQ1jJKduSGPvlJ7IkPc9Oxz3PEQ5QtpEdNY4QWHzRajtakrUPrea9s7N/6T CJTK7 PZ\nN37v HEVXtI0hrSdKG4coXCb cJbFd 1tGMBrf9W/u5ElxPRqKPF1DpjGgaJRzo1jBLRMEZ8\n6oY09k6JVcbTUziJ0BjSjlxI5nda/Mq69ek5saWtLgfBORKHcyRBkM RikN67OHFe jOLdPd\nrilC4hCSYKqEtO a6iT9ZNF1bgdBSBxCEkyVkIZqlxIdNh51OwhC4hCSIMAhaYGQOIQkQEhq\nCIlDSAKEpIaQOIQkQEhqCIlDSAKEpIaQOIQkQEhqCIlDSAKEpIaQOEzHJUBIagiJQ0gChKSG\nkDiEJEBIagiJmyrnSK52CkJSQ0gcQhIgJDWExCEkAUJSQ0gcQhIgJDWExCEkwRQI6fBVM7pK\nfjBC4hCSYAqEdH9zb3r8pcaAkDiEJJgCId3xkIcHIyQOIQmCH1JzeFpt3Ghd1tDYvu76 c 6\nfDRC4hCSIPgh0QrzFSm8NDHcPHs77Q7nT5eGh5yIRx0tpjbYrWGQ3khCwygxDWMkpkhIrnZK\ndyTpfc e9/z0FGouU0gtRD qJ0oYh ybMUFk6fw xCeG33u5FOWYIHJESK8SbVhClDL22Tcn\ne52IRR0tptYd1zBIZ6RbwyhaNmeKhORqp8QjPRp2reenp7/cIb3GQnIG50gczpEEU UcCSFl\nISTnXO0UhKSGkDiEJEBIagiJQ0iCKRCSJwiJQ0gChKSGkDiEJEBIagiJQ0gChKSGkDiEJEBI\nagiJQ0gChKSGkDiEJEBIagiJw3RcAoSkhpA4hCRASGoIiZsKb 1c7xSEpIaQOIQkQEhqCIlD\nSAKEpIaQOIQkQEhqCIlDSAKEpIaQOIQkQEhqCIlDSAKEpIaQOIQkQEhqCIlDSIJAhZQ2ti1v\nvLKVKL5yfk3zUUpYfx/2tHHavj5svNTwgLshERKHkASBConCi7toc02ClqzsST5Rl8yHlLtu\n3n90wF42dd6JeNTRYmqJLg2D9EYGNIwS0zDGwBQIyfVO6YoMet zybjnITSFtInorHHiqGGW\nnZ7dZodkX6fw04VlMUFk6fw zMvP7z1cKk0TRIa3m2/rjCNtRsZGOyT7OoXbCsueH3AiFnW0\nmFpfp4ZBuiO9GkbRsTm9UyAk1zulM9Lnfc/2x7yOkNAU0o5MSDuNZOZqJqRTxmn7euZ d3CO\nxOEcSRCwc6RsSCczM3yfoVTVbqLdxmn7OkJCSA653imBDIlubjqXaqmJ0YJHafAO43T OkJC\nSM643inBDCl 36yZTfuJ3rj6qm 8bvwufx0hISRnXO UYIWkH0LiEJIAIakhJA4hCRCSGkLi\nEJIAIakhJA4hCRCSGkLiEJIAIakhJA4hCRCSGkLiMK dACGpISQOIQkQkhpC4oL41s7zTkFI\nagiJQ0gChKSGkDiEJEBIagiJQ0gChKSGkDiEJAhYSKP/ePfhq2Z0eRoQIXEISRDwkO5v7k17\nGhAhcQhJEPCQ7njI03AISYKQBJM9pNyMdRT5Zu28NYMU3rx8xryt5vG/cn510wFqDk r7cpd\nzk965wpC4hCSYLKHZM9Yd9O98TevfZjC1x0YeNya2O62rsF1s7pphfmKZF 2J71zBSFxCEkw\nyUOyZ6xrN84Qte i8DNEZ4yOduMkUbJ2qxVS/nJu0jv7oUMJJ2JRR4up9XdqGKQ70qdhFB2b\n0xfAkDzvlM5Iv/c9OxDzOsJgiSHZM9a9UpX9TCE3sV32auNTVkj5y7n77IdigsjS X3Y6 f3\nHtWl1Aki7RnrflU1nA1pR1FIC9cXQjIv2zOj4BXJI7wiCSb5K5I9Y90xo4Po8At2LMesd3CJ\n6lYrpPzlUSE5g3MkDudIgkl jpSfsW7J8rdO3bAmH0vT7T2JtXX9mQ8b7MsIiRDSWDzvlMke\nkj1jXdddNXUPJvKxnLurbu6KU5QJyb6MkAghjcXzTpnsIZUbQuIQkgAhqSEkDiEJEJIaQuIQ\nkgAhqSEkDiEJEJIaQuIQkgAhqSEkDiEJEJIaQuIQkgAhqSEkDvPaCRCSGkLiEJIAIakhJC5Y\nb 0Q0oRASBxCEiAkNYTEISQBQlJDSBxCEiAkNYTEISRBkEJKGXv0rUcOQuIQkiBIIaX39dJe\nl3/faDwIiUNIgiCFZFnRomc1bAiJQ0iCIIVkvrVbVjXjxoSxj i0cTo/I6Q9ieTLC6vnrUm6\nGxMhcQhJELCQqKGF7JDyM0LmJpE8U7Vn MwNG92NiZA4hCQIdEjZGSHtSSSPGu1Ew/mlz/c5\nEYs6WkytJ65hkK5Ij4ZRdGxOd6BCMrenU8NOiUd6NYzi ekZKEtI2Rkh7Ukk0w9Na9rwZn5p\nTBBZOr8Pfp383pd6lTpB5JghnbJCyk4bZE8iab48vXj7tLb80uediEcdLaaW6NIwSG9kQMMo\nMQ1jDAQqpPPn 3s07JSuyKD3QZJxz0NoDClVtZtodyEkexLJlPVPjj3c7G5MnCNxOEcSBO0c\nqXFdHy14lAbvKIRkTyK5pf5IOr5stbsxERKHkARBC2lTdT29cfVV33jd F0 pNwkkukN9dPn\nf8flTkNIHEISBCmkckBIHEISICQ1hMQhJAFCUkNIHEISICQ1hMQhJAFCUkNIHEISICQ1hMQh\nJAFCUkNIHEISICQ1hMRhXjsBQlJDSBxCEiAkNYTE4a2dACGpISQOIQkQkhpC4hCSACGpISQO\nIQkQkhpC4hCSYIqFdNbocPcAhMQhJAFCUkNIHEISICQ1hMQhJMEUCOnYddU37jOOU/uSmkXb\nEJJ3CEkQ/JDS9ff3H19sdKQbViXOLUNI3iEkQfBDOmi8RbTF6DhonCHaUQhpsMeJWNTRYmrd\ncQ2DdEa6NIyiY3O6AhWSuT06np5YpFvDKJ6fnt7yhdQWTptv74yOtqphohOFkDBBZOn8Pvh1\n8ntf6qVlgkjZthnmlw6jo7UqG5R9ezrlRDzqaDG1ZLeGQfoigxpG0bE5g4EKKZUa6NGwU7oj\nSe DDMW8jlCYklt7SPsM87yg1ejYa73Fa8M5knc4RxIE/xxpaO7a5MkmoyM5Z1XvqaUIyTuE\nJAh SPSbRbVL9xgn6fCN1Yt2G8fdPRghcQhJMAVCSg1ZH931l/ZghMQhJEHwQ0oveKAvfpvL\nufPzEBKHkATBD4mO31w79 5IiQ9GSBxCEkyBkDxBSBxCEiAkNYTEISQBQlJDSBxCEiAkNYTE\nYTouAUJSQ0gcQhIgJDWExCEkAUJSQ0gczpEECEkNIXEISYCQ1BASh5AECEkNIXEISYCQ1BAS\nh5AEUyGklLG75MciJA4hCRCSGkLiEJIAIakhJA4hCaZKSKlbbl 81ry8r8pdGQiJQ0iCqRLS\nqqbBTfPTRGtucfdYhMQhJMEUCWl9Yw91hvfTcN3L9q2JLieiUUeLqXXGNAwSj3RqGCWiYYx4\noEIyt0fH0xOrjKen8Et2WUK61/iN f3W79Oe6gH7VkwQWTq/D36d/N6XepVxgkgrpIbli80X\n3pfr0w udPlYvLXj8NZOMDXe2u3qrX/UPBSrD87Z5fKxCIlDSIKpEdJu2j/t10T33FLndmsR\nEoeQBFMlJHpiXhdtN77n9rEIiUNIgqkQkhcIiUNIAoSkhpA4hCRASGoIiUNIAoSkhpA4hCRA\nSGoIiUNIAoSkhpA4hCRASGoIicN0XAKEpIaQOIQkQEhqCIlDSAKEpIaQuOCcI2W3ByFNAITE\nISQBQlJDSBxCEiAkNYTEISQBQlJDSBxCEiAkNYTEISRBwEPae2Tk5eLrjiAkDiEJAh7SipaR\nl4uvO4KQOIQkCEpIHbdeMWv5aUob25Y3XtlK9PLC6nlrksuqZtxo32Vdzlx3BSFxCEkQlJCu\nWZXov6eJKLy4izbXJM5U7Rk c8NGamgp3GVdbsArkncISRCUkHoHibZPS1N4E9FZ48RRo51o\nOBuOfdeokAbiTkQjjhZTi0V1DBKJaRhFz5r4HYAuue3RsmMr4 np8hrS3mXz5s02UhTeThQ3\njqQfmta04c1sOPZdo0LqjzoRiThabLxRAjVIcCaI1LpTNAzifU06PYZ0esbGJO20QtqRCcl8\nWXrx9mltVjj5u/DWjvDWbqTs9uCtXUFb2NyMxwohpayXuIebrXDydyEkQkgjZbcHIRUcMvaf\n39ZsnLND2lJ/JB1ftpoa1/Xl7zIvZ/53BSFxCEkQkJDoh1fMWd27ePbZXEjpDfXT53 njzZV\n1 fvsi5b/7uCkDiEJAhKSOWCkDiEJEBIagiJQ0gChKSGkDiEJEBIagiJQ0gChKSGkDiEJEBI\nagiJQ0gChKSGkDhMxyVASGoIiUNIAoSkhpA4hCRASGoIiQvKOZK9PQhpAiAkDiEJEJIaQuIQ\nkgAhqSEkDiEJEJIaQuIQkiAIIaWM3bnve yLZ40OPauDkDiEJAhSSOl9vQhpHAipwN4ehJRj\n11N0ESGNASEV2NuDkHJSxs WVi/cab 1a19Ss2ib0ZEw9hGdNk7n54w8dl31jfuM4 7GRkgc\nQhIEI6RrDvRvmHY2G1K6YVXi3LJCSPacken6 /uPL3b7SoWQOIQkCEZIT5nPbk1LNqSDxhmi\nHcUhZeeMPGi8RbSlEFJ/BErldwJ6 L0Xy6FQYkkhvWJ XfB4NqS2qmGiE8UhZeeMbAunzbd3\nhZASnU5EI44WU4tHNQwSi8Q1jKJlcwISUn57Yhp2SlTL0 P5QOn2FtJr5tdrnsyG1FqVDSYT\n0ikrpOzEQttmmAt14K2dd3hrJwjGW7ufmgdr7ZZsSHutt3BtRkeqajfR7kJI wzzIGpFSN4h\nJEEwQmrsGNpY3ZMNKTlnVe ppWYwCx6lwTsKIQ3NXZs82YSQvENIgiCElDRam6obX7c//j58\nY/Wi3cZxeuPqq77xuvG7/HTgv1lUu3SPcdLd2AiJQ0iCIITkUGqI6KDR7 5BCIlDSIKpE1J6\nwQN98duaXT4KIXEISTB1QqLjN9fOvTvi8kEIiUNIgikUUkkQEoeQBAhJDSFxCEmAkNQQEoeQ\nBAhJDSFxmI5LgJDUEBKHkAQISQ0hcQhJgJDUEBIXjHOkwvYgpAmAkDiEJEBIagiJQ0gChKSG\nkDiEJEBIagiJQ0iCgIVUNDOX9bcqvK4MQpIgJEGAQ0rv62X37z3ickCExCEkQYBDkqxocTkg\nQuIQkiCAIaVuuf288VLDA9Zbu/jK TXNR/OTQy6rmnGjuwEREoeQBAEMaVXTIIUXHx2wQlqy\nsif5RF0yPzlkQ EVKZ1yIh51tJhaslvDIH2RQQ2j6NicwUCEVNiegR4NO6U7kvQ yFDM6wjD\n kJa32j  h9 OvNhw1HDTDw9uy0/OWRRSJggsnR R6CD3/uwPDxNEDkipHuN35jfw22ZkNqM\njI35ySGLQhrscSIWdbSYWndcwyCdkS4No jYnK5AhFS0PTqenlikW8Monp ewudrXkNqWL7Y\nfLNqzRxkhrTTSGZuzk8O2YAPG7zDOZIgcOdIu3rrH82HdNI4ZN54hvKTQyIkDWMgJEHgQtpN\n 6f92g6Jbm46l2qpieUnh2xc53KnISQOIQkCGBI9Ma/LDil 36yZTfspPznkpup6dwMiJA4h\nCQIW0lhKmhzSgpA4hCSYGiGVNjmkBSFxCEkwNUIqbXJIC0LiEJJgioRUMoTEISQBQlJDSBxC\nEiAkNYTEISQBQlJDSBym4xIgJDWExCEkAUJSQ0hcIN7aFW0PQpoACIlDSAKEpIaQOIQkQEhq\nCIlDSAKEpIaQOIQkQEhqCIlDSIIpElLJk0UiJA4hCaZISNJkkY4gJA4hCaZISCVDSBxCEgQw\npLTRuqyhsX3d9fOfNUNYOb 66YD11i5tbFveeGWry8EQEoeQBAEMicJLE8PNs7fT7nAXLbmt\na3DdrG7rHCm8uIs21yTspYaHnIhHHS2mNtitYZDeSELDKDENYySCEFLR9vT3aNgp3ZGk90HO\ne356CjXrCKmF6Ef1RAnjULtxkihZuzUT0iais8YJeylMEFk6vyvQwO9dWCa6JojMhvQq0YYl\n1kd1 16pSps3ND6VCWk7UdzI/3MU5/uciEUdLabWE9cwSFekR8MoOjanOwghFW9Pp4adEo/0\nahjF89MzoDWk10aFtHB9JqQdI0JyBudIHM6RBIE8RyqEdMx6K5eobkVIOQgpp2h7ENIYikKi\nptt7Emvr hFSDkLKKdoehDSG4pDO3VU3d8UpQkg5CCmnaHsQ0gRASBxCEiAkNYTEISQBQlJD\nSBxCEiAkNYTEISQBQlJDSBxCEiAkNYTEISQBQlJDSBzmtRMgJDWExCEkAUJSQ0jc5H9rN3J7\nENIEQEgcQhIgJDWExCEkAUJSQ0gcQhIgJDWExCEkAUJSQ0gcQhIEPKSUsXvEd9cQEoeQBAhJ\nDSFxCEmAkNQQEoeQBJM8pOLJIBPGPqLTxuni21LGz5ZWL9yZCSm cn5N81GXq4OQOIQkmOQh\nFU8GaYdUfFvKuOZA/4ZpZ62QlqzsST5Rl7QfmRp0Ih51tJhaf6eGQXoi/RpG0bI5kz6kkdvT\n261hp3RGBrwPkoh5HSF/eLsOqTAZZCGkwm0p4ynzV9CaFjOko4aZe3p2m/1ITBBZOr9D8Mrv\n/Vc JU8QWTQZZCGkwm0p4xVzoQWPmyG1GRkb7UeeH3AiFnW0mFpfp4ZBuiO9GkbRsTm9kz6k\nkdvT06Vhp3RG rwP0h/zOkJ Rm7XIb02MqRTVkivFYVkXqZrnjRD2mkkxx2NwzkSh3MkwaQ/\nRyqKpmo30e7RIf3UPI5rt5ghnTQOmcufcbk6CIlDSIIAhUQLHqXBO0aH1NgxtLG6x/qw4eam\nc6mWGpdHEkLiEJIgSCG9cfVV33jd F3xbUmjtam68fXsx9/3zZrZtN/l6iAkDiEJJntI5YaQ\nOIQkQEhqCIlDSAKEpIaQOIQkQEhqCIlDSAKEpIaQOIQkQEhqCIlDSAKEpIaQOMxrJ0BIagiJ\nQ0gChKSGkLhJ/taObQ9CmgAIiUNIAoSkhpA4hCRASGoIiUNIAoSkhpA4hCQIRkgpY498x Gr\nZnR5WRuEJEBIgska0t4jI1ZgX2/mlpG3mu5v7k17Wh2ExCEkwWQNaUWLdAu79Y6HPK4OQuIQ\nkmCShrSsasaNV7YSPW6cJWp 2nxrZ91i/W9PXzdsvNTwQHN4Wm1X58r51U0HKG1sW95oPcYV\nhMQhJMEkDYkaWmj1aqKvN26l5Iwj1jmSeUvmf3v6uvDiowO0wnxFWnJb1 C6Wd3mDV20uSYx\n/tjFEBKHkASTOKS2qylR 9xq2js3XRRSfvq68NPmYmZI7cZJomTtVgpvIjprnLBHwHRcpZrk\n03Gx7ZnS03GZyfRUxXct7VhA679FRSHlp68Lt2VDeqXK rih8SkKbyeKG/mPIzBBZOn8bsET\nv3deWbmeINLKZvEr69an58SWthaHlJ  LryjOKSF6zM3FIWEKYtLNcmnLGbbM6WnLLayeezh\nxXvozi3T48Uh5aevs0M6Zr2bS1S3jg7JGZwjcThHEkzWc6TGdX2075rqJP1k0XXZ35C1brH \nt6evs0Oiptt7Emvr hGSBSEhpBE2VdfTUO1SosPGo9mQrFus/ 3p6/Ihnburbu6KU4SQLAgJ\nIfkCIXEISYCQ1BASh5AEFRzSJUXeoe8HuIOQOIQkqOCQZpk cuHnqqd96m2fXaTvB7iDkDiE\nJKjgkEwbP37a nboo5v0/QB3EBKHkASVHdLH/3f2 9r/W98PcAchcQhJUNkhvePl7PeNF n7\nAe4gJA4hCSo7pA/MzY4/6w/1/QB3EBKHee0ElR3SbaFPXH/nndd LNSs7we4g5A4hCSo7JDS\n9/1hyHTZrRpWsjTeQvo/AgohMZUdkjn0iVd3tg/rG94thCRBSExlh/S5F/UNXBqEJEFITGWH\ndPn9 gYuDUKSICSmskN67mM/1bG7PUBIEoTEVHZIn/9E6B0f KBF/aDMX5kYW8rYkzJ2l7Q6\nCEmCkJjKDunvvvTlHPWDxgkpva8XIemEkJjKDsmpcUIi60UJIemDkJhKDyn6wvd/8PNxfz8z\nvHn5jHlbiTpuvWLW8tOUNH7e3HDlzvz37Fu7pjXmkoeqzrpaHYQkQUhMZYc0vORC6zdkL1k5\nzoPC1x0YeLwmQdesSvTf02SGc0MnbZnWVfhuhbRldpLokWXuVgchSRASU9khrQxNX9fywvf \nKfSY kHhZ4jOGB3UO0i0fVo6ZTxnRjhrc G7FVJi5jZKf22r/ZjzfU7Eoo4WG4PfB3y5dHrZ\nKTnxHg2DRDw9PTndWjYn0qthFM bMyCH9LGbst8XfGackHJzP 5dNm/ebCOVMl41b1y4ofA9\n86ndg7fS/pmD9mMmYoJIvw/4cpmAXQclGWOCyItyc96/ePE4IWUnCDo9Y2OSdlohbbfqe7Lw\nPRPS4arY2u/mHzM85EQ86mixMfh9wJdLj5edktOd0DBINKZhkH4tmxNJeh/kvOfNKby/HBHS\nJc9nv//bux2F1BY2x3nMCsl8q5esfrnwPfv7SNc9W3dAPRCDcyQJzpGYyj5H vt/yEzBmvjK\nF9UPyoV0yNh/fluzcS5lXNuR3DC9u/A9G9KmWVe7XR2EJEFITGWH9OLb/uSaO  46gMXbFE/\nyJ778YdXzFndu3j2aaPlG9UNuyiV/54NqXf6Rrerg5AkCImp7JDopx 1Pv7 hMs/BG7/7uvo\n34XtqO50uzoISYKQmMoO6QjRm6/tesvtGHJIw2dv oHr1UFIEoTEVHZIoQ83PlfCX9OWQ9pQ\nuyopL6 AkCQIianskL4747LQhV 859fe/mVyLxCSBCExlR2S9Qe3V5sxvb9O3w9wByFJEBJT\n6SFZ3lz Pt m1kdIEoTEVHpIb25Y8JHQpV 5S98PcAchSRASU9kh/ct/DV027V93 TYZF a1\nk2BeO0FlhxS6rNntn nRDCFxCElQ2SE9Xv h0PtrH9qvb3i38NZOgpCYyg7JdPzReZeH3lej\n7we4g5AkCImp JBMx26 DJ/aVRSExFR4SOnfrp1zeejir3xb3w9wByFJEBJT2SFNvyz0tk81\nbRkca HyQ0gShMRUdkh/9NUfu5v0RzuEJEFITGWH5NreI97uZxCSBCExAQtpRYu3 xmEJEFI\nzOQLadQkkPb1YeOlhgeWVc24MW20LmtobF93/fxnieIr59c0H6W0sW1545WtZN3vbnUQkgQh\nMZMvpFGTQNrXKbz46AA1mK844aWJ4ebZ22l3uIuWrOxJPlGXNO/sos01icz9riAkCUJiJl9I\noyaBtK9T GnzzkxI5v8/qidKGIeOGuaapWe3UXgT0VnjRHFIgz1OxKKOFhuD3wd8ucS97JSc\neJeGQSKenp6cLh2bE4t0a9S 7dYAACAASURBVBjF8 b0Og9p1CSQ9nUKt9khmXdsWGL9/dh9\nbUbGxvwMkkUhYYLI0k3AroOSjDFBpGD0JJD29ey/SJEJ6bV8SDuN3F8ut6cZKgop7Uhn1Nly\nMr8P HLp87JTcnqSGgaJxjUMktCyOZEh74MMe98cxyGNngTSvi6GdNI4ZN54RgzJGZwjSXCO\nxEy6c6TRk0Da17MhNa7rKw6Jbm46l2qpieVDsu53BSFJEBIz6UIaPQmkff1sJqRN1fUjQorf\nN2tm0/7CK5J1vysISYKQmMkXUk6p/wSfSwhJgpAYhKSGkCQIiUFIaghJgpCYSRvSBEFIEoTE\nICQ1hCRBSAxCUkNIEoTEICQ1TMfFYTouAUJSQ0gcQhIgJDWExCEkAUJS8xKS32cy5YOQGISk\nhpAkCIlBSGoISYKQGISkhpAkCIlBSGoISYKQmKkS0lmjo6THISQJQmIQkhpCkiAkBiGpISQJ\nQmICHFJ Csn2JTWLthkdrbM7iW65x90gCEmCkJgAh2RPIZluWJU4t8x8RbpzJf17XZe7QRCS\nBCExgQ4pO4XkQeMM0Q4zpM45v5y/PX93otOJaMTRYiK/D/fyiZW Uwp7Nq5hkEhUwyBxLZsT\n0bE9njenuywhZaeQbKsaJjphnSNtq1pZuLv8E0T6fbiXT9l3HZTI QSRbkLKTiHZWpUmOmaF\n9FTtYrevv3hrJ8FbOybQb 2yU0juNd4iajNDOl579NqnXA6CkCQIiQl0SNkpJJNzVvWeWmp0\npG5YTwerj7sbBCFJEBIT6JByU0gevrF60W7j JMLzadu7Q3uthchSRASE iQNEzYhZAkCIlB\nSGoISYKQGISkhpAkCIkJcEhaICQJQmIQkhpCkiAkBiGpISQJQmIQkhqm4 IwHZcAIakhJA4h\nCRCSGkLiEJIAIanhHEmCkBiEpIaQJAiJQUhqCEmCkBiEpIaQJAiJQUhqCEmCkJhAhZQydqeM\nPfa18A4Nq4OQJAiJCVpI6X299jWEVDYIiQlaSEXXEFLZICQmaCFZb 0i36ydt2aQwpuXz5i3\nlSi cn5N81FKG9uWN17Z6nJIhCRBSEwQQ7rp3vib1z5M4esODDxek6AlK3uST9QlKby4izab\n111BSBKExAQwpHZrQsj2XRR huiM0XHUMNcuPbuNwpusKcBP2Av3R52IRBwtJvL7cC f0vdJ\n0Z7VMoieUXSMURlr0qkxpFeseexM4e3m2zrjSJuRsdG bi88EHciGnG0mMjvw718oqXvlMKe\njWkYJOLh6cmL6ViTaETLKF4HKMzI7T2kX1kzq1L2wwYznJ1GkoqvuxsSb 0keGvHBPCtXWZS\n1cMv2OGcNA6Zd51BSBohJCaAIdGS5W dumFNPpybm86lWmpiCEkfhMQEMaSuu2rqHkzkw4nf\nN2tm0368ImmEkJhAhVQGCEmCkBiEpIaQJAiJQUhqCEmCkBiEpIaQJAiJQUhqCEmCkBiEpIaQ\nJAiJQUhqmI6Lw3RcAoSkhpA4hCRASGoIiUNIAoSkVnJIfp/GlBVCYhCSGkKSICQGIakhJAlC\nYhCSGkKSICQGIakhJAlCYoIektcJIxGSBCExQQ/J64SRCEmCkJigh1QEIWmDkJgghfSLzHRB\n6xLGPqLTxulh46WGB4QJI11BSBKExAQppHQqldpR226HROHFRweECSNdQUgShMQEKSTTm7N \nQYWQns582DBqwkh70f5IWfl9rJdVeXcdlK5QoqeQEtc QkUhtWVCGjVhZH7ZLieiUUeLcX4f\n62UVK3GnFIt1ahgkUurTUyyuZXMiWrbH6wCFP9XpKaR7/1cqF9IpK6QdmZBGTRjpbkS8tZPg\nrR0TqLd2z9RbM7amqnYT7S6ENGrCSHdDIiQJQmKCFNIb1QdTqdQwLXiUBu8ohDR6wkhXEJIE\nITFBCmll5uPv6 mNq6/6xuvG7/IhjZow0hWEJEFITJBCKgeEJEFIDEJSQ0gShMQgJDWEJEFI\nDEJSQ0gShMQgJDWEJEFIDEJSQ0gShMQgJDVMx8VhOi4BQlJDSBxCEiAkNYTEISQBQlLDOZIE\nITEISQ0hSRASg5DUEJIEITEISQ0hSRASg5DUEJIEITEBDill7PY CEKSICRmioS09whR7Ft1\nM5sPuxwEIUkQEjNFQlrRQrR4afvpb8/FdFwaICQmCCEdu676xn3Gcft70vh5c8OVOzMhda6c\nX910gJZVzbix5 6TROeM/3A3NkKSICQmACGl6 /vP77Y6LC/p4wbOmnLtC4rpCW3dQ2um9VN\nDS3ZZQ G86tqTSc5vnjU0WKc38d6WfWWuFOKdQ9qGCQa1zDIQI GQbojSe DDMW8jjDsIaSD\nxltEW4wO 3vKeI5oeNZmM6R2w3wVStZutUPqafxh/mGYILJ05d11UDovE0S2hdPm2zujw/6e\nMl41b124wQwpOzdk41O5kH63YE06/7DBHidiUUeLcX4f62UVL3GnFIt3aRgkUurTU6xLx bE\nIt0aRvG8Ofl/e6WEkLbNML90GB3295Sx3byw4Ml8SAvXZ0PaM d514PjHEmCcyQmAOdI 4wY\nUavRYX9PGc Y7 eqXzZDOmacIEpUt2ZC u0Vr7tfHYQkQUhMAEIamrs2ebLJ6LC/p4xrO5Ib\npndbHzY03d6TWFvXT43r pJXPWm9icTH3xogJCYAIdFvFtUu3WOctL njJZvVDfsynz8fe6u\nurkrThFtqq7fk5k 0njB3dgISYKQmCCElBqyPrrrt79r aNBNoQkQUhMAEJKL3igL35bc/47\nQio7hMQEICQ6fnPt3Lsj e8IqewQEhOEkMoJIUkQEoOQ1BCSBCExCEkNIUkQEoOQ1DAdF4fp\nuAQISQ0hcQhJgJDUSg3J7zdf5YWQGISkhpAkCIlBSGoISYKQGISkhpAkCIlBSGoISYKQGISk\nhpAkCIkJREgpYw 7Rf7zdmeNDndDIyQJQmICEVJ6X  oWxBSWSEkJhAhcQiprBASE4iQrLd2\nkW/WzlszSPGV82uaj2ZC6rj1ilnLT1Pa2La88cpWovYlNYu2ISQdEBITmJBuujf 5rUP05KV\nPckn6pJWSNesSvTf00QUXtxFm2sS6YZViXPLCiGlzjsRjzpajPH7UC vntJ2yghdAxoGicY0\nDNKvZXMig94HScY9D E1pHbjjPmas uoYTadnt1mhdQ7SLR9WprCm6y3dCcOWkvsKIRU3gki\n/T7Uy6usuw488DJBZCak7Ax21Jad4WSjFdLeZfPmzTZSFN5OFDeOtFUNE50ohHS z4lY1NFi\njN Henl1lrZTRoj3aBgkUuLTM0K3ls2J9GoYxfPmDHgN6VdVmVmPdxrJ3C27T8/YmDSvmyHt\nyITUWpWdiNXd0DhHkuAciQnKOVKmkMMvnDQOmdfPWCG1hc0te6wQ0l5ravA2hKQDQmKCEhIt\nWf7WqRvW0M1N51ItNTEzpEPG/vPbmo1zdkjJOat6Ty1FSDogJCYwIXXdVVP3YILi982a2bQ/\n8/H3D6 Ys7p38eyzuZDo8I3Vi3Ybx90NjZAkCIkJREhlhJAkCIlBSGoISYKQGISkhpAkCIlB\nSGoISYKQGISkhpAkCIlBSGoISYKQGISkhnntOMxrJ0BIagiJQ0gChKRWYkh v/cqM4TEICQ1\nhCRBSAxCUkNIEoTEICQ1hCRBSAxCUkNIEoTEICQ1hCRBSEyQQ9LxrzIjJAlCYhCSGkKSICQG\nIakhJAlCYgIQUtpoXdbQ2L7u vnPkj0r5LDxUsMDVkipW25Pda6cX910gJasNRfeV WuDIQk\nQUhMAEKi8NLEcPPs7bQ73FU0K TRASukVU2DtOS2rsF1s7o3zU8TrbnFftRQwolY1NFio/l9\npJdZV0k7ZaTOPg2DRGIaBunVsjmRfu DDHjenEFPIbUQ/aieKGEcKswK XTmrd36xh5qN04S\nJWu3dob303Ddy/ajyjpBpN9HepmVc9eBF54miAy/SrRhiRXOvsKskG3W9XuN3xBl545sfIpu\n/T7tqc7PoIdXpNLhFYkJxCvSa/mQRswKmTIali9O5UJauJ5erk8/uNLl2DhHkuAciQnEOVIh\npBGzQqaMXb31j9Ix44T5tq 6lfqrD87Z5XJshCRBSEzAQhoxK6T1YcP ab mptt7Emvr onu\nuaXO7dYiJAlCYgIW0ohZITO/j/TEvK5zd9XNXXHKXHC78T23YyMkCUJighBSOSEkCUJiEJIa\nQpIgJAYhqSEkCUJiEJIaQpIgJAYhqSEkCUJiEJIaQpIgJAYhqWFeOw7z2gkQkhpC4hCSACGp\nlRKS32 8yg8hMQhJDSFJEBKDkNQQkgQhMQhJDSFJEBKDkNQQkgQhMVMlpFJnFEJIEoTEBCak\nvUeUdyMknRASE5iQVrQo70ZIOiEkJighLauacWNmMjt7Zjt7Frv4yvk1zUczIb28sHremqS7\ncRGSBCExQQmJGlqyk9nZM9vZs9gtWdmTfKIuaYZ0pmrP8JkbNrobFiFJEBITrJCeNi/kZrbL\nzWJ31DDXLT27zQzpqNFONJx/wPkBJ2JRR4uN4PdhXn5d7ncK09mrYZBITMMgPVo2J9LnfZB \nz5uT0BNSm3nBntkuO4tdm5Gx0Qwp/dC0pg1v5h9Qvgki/T7My69suw488jRBZCGkHUT5me2y\ns9jtNLInRZkPG86 ePu0NvsBqfNOxKOOFhvB78O8/Hrc7xSma0DDINGYhkH6tWxOZND7IMm4\n5yG0hZSf2S47i91J45B55xkrpFSXeenhZnfD4hxJgnMkJjDnSI3r jIh5We2y81id3PTuVRL\nTcwMaUv9kXR82Wp3wyIkCUJiAhPSpur6TEj5me1ys9jF75s1s2m/9YqU3lA/ff53XO40hCRB\nSExgQioThCRBSAxCUkNIEoTEICQ1hCRBSAxCUkNIEoTEICQ1hCRBSAxCUkNIEoTEICQ1hCRB\nSAxCUsO8dhzmtRMgJDWExCEkAUJScx S32 7JgJCYhCSGkKSICQGIakhJAlCYhCSGkKSICQG\nIakhJAlCYoIeUsrY4 nxCEmCkJigh5Te1 vp8QhJgpCYoIfkFUKSICQmACFd2Ur0uHGWqPlp\n6lw5v7rpAGWmisx8sd7a5eaInKAJIv0 yCcCQmICENLq1URfb9xKyRlHaMltXYPrZnVnp4q0\nvlgh5eaInKAJIv0 yCcCQmICEFLb1ZSofW417Z2bbjdOEiVrt2anirS mCHZc0SOmiAy2etE\nLOposSJ H QTodPtThHEuzUMEnH99Ai6tGxOpEfDKJ43p99DSD1V8V1LOxbQ m/RK1Vp84bG\np7JTRVpfzJDsOSInaIJIvw/yiVCmXQeeeZogcvEr69an58SWtuZCWrg O5 Q9cUMyZ4jctQE\nkcNDTsSjjhYr4vdBPhF63O4UQXdCwyDRmIZB rVsTiTpfZDznjen8P6yhJAee3jxHrpzy/Q4\nHTNOECWqW0eElJ8jcmImiPT7IJ8IOEdiAnCORPuuqU7STxZdZ15sur0nsbauf0RI9hyREzRB\npN8H URASEwQQhqqXUp02HjUvHjurrq5K07RyJByc0RO0ASRfh/kEwEhMUEIqZwQkgQhMQhJ\nDSFJEBKDkNQQkgQhMQhJDSFJEBKDkNQQkgQhMQhJDSFJEBKDkNQwHReH6bgECEkNIXEISYCQ\n1BASh5AECEnNbUh n71MDITEICQ1hCRBSAxCUkNIEoTEICQ1hCRBSAxCUkNIEoTEBDmkw1fN\n6PI6BkKSICQmyCHd39ybHnXT3iMux0BIEoTEBDmkOx5iN61ocTkGQpIgJGayh5Q2Wpc1NLav\nu37 s2RPD/mLzKxB65rD02q7Rk4Ouaxqxo3uxkdIEoTETPaQKLw0Mdw8ezvtDnfZ00OmU6nU\njtp2WmG Io2aHLIBr0g6ICRm8odkpvGjeqKEcSg/PSTRm7N QVZIoyeHLAop0eVENOposTy/\nD/GJEXO3U0SxTg2DRFw PaK4ls2JaNkerwMU/jCa65BeJdqwxJrlZF9 ekhKXPsIZUIaPTlk\nUUjlmSDS70N8YpRl14EGpU8QGX5tVEgL1xPd 79S2ZBGTw5ZFFLakc6os Vsfh/iE6PP3U4R\n9SQ1DBKNaxgkoWVzIkPeBxn2vjk6QspPD/lMfeb3j8yQRk8OiXMkLXCOxEz c6RCSPb0kG9U\nH0ylUsOZDxtGTQ7ZuK7M89r5fYhPDITEBCqk3PSQKzPnRddnQho1OeSm6np3wyMkCUJiJn1I\nZYaQJAiJQUhqCEmCkBiEpIaQJAiJQUhqCEmCkBiEpIaQJAiJQUhqCEmCkBiEpIbpuDhMxyVA\nSGoIiUNIAoSkhpA4hCRASGpuQvL7zGXiICQGIakhJAlCYhCSGkKSICQGIakhJAlCYhCSGkKS\nICQmoCGljN3i7WeNDncDISQJQmKCGNLeIwipvBASE8SQVrQgpPJCSMzkDilp/Ly54cqd akh\nh42XGh6wZoG0Quq49YpZy09T2ti2vPHKVqL2JTWLtiEkHRASM7lDShk3dNKWafmpISm8 OiA\nNcOJFdI1qxL99zSReVsXba5JpBtWJc4tQ0g6ICRmsof0HNHwrM35qSHDTxPZIfUOEm2flqbw\nJust3YmDxhmiHYWQBuJORCOOFsvw /CeOFHnO2XsPRvTMEjExdMzppiONYlGtIzidYDCP7/i\nMqRXza8LN Snhgy3FULau2zevNlGisLbieLGkbaqYaIThZD6o05EIo4Wy/D78J44zveJYs9q\nGUTPKDrGqIw16Sw1JDMSWvBkfmrI8I58SKdnbEzSTiukHZmQWq1ljuGtnQ54a8dM9rd2z5hv\n6apfzk8NWRRSW9jcsscKIe013iJqQ0g6ICRmsod0bUdyw/Rue2rIbEiN6/rMkA4Z 89vazbO\n2SEl56zqPbUUIemAkJjJHlLLN6obdpE9NWQ2pE3V9dY50g vmLO6d/Hss7mQ6PCN1Yt2G8fd\nrQ5CkiAkZrKHJP  qz4ISYKQGISkhpAkCIlBSGoISYKQmMkdUvkhJAlCYhCSGkKSICQGIakh\nJAlCYhCSGqbj4jAdlwAhqSEkDiEJEJIaQuIQkgAhqTkKye9TlgmHkBiEpIaQJAiJQUhqCEmC\nkBiEpIaQJAiJQUhqCEmCkBiEpIaQJAiJQUhqCEmCkBiEpIaQJAiJCXJITWvML4eqzuZmirSv\nuhoDIUkQEhPkkLbMThI9ssyeKdK 6gpCkiAkJsghJWZuo/TXttozRdpXczTOa f3cT3hvM29\nltuzWgapjNnkohWzJiXOazeOB2 l/TMH8zNF5q7amXU6EY04WMjv43rCxRztu3H2bFzDIJGo\nhkHiWjYnomN7PG9Od1lCOlwVW/tdys8Umb3qDt7aSfDWjgnyWzui656tO0D5mSKzV91BSBKE\nxAQ7pE2zribKzxSZveoOQpIgJCbYIfVO32h tWeKzF11BSFJEBIT7JA6qjsVV51ASBKExAQ5\npOGzN/1g7KvOICQJQmKCHNKG2lXJsa86g5AkCIkJckg6ICQJQmIQkhpCkiAkBiGpISQJQmIQ\nkhqm4 IwHZcAIakhJA4hCRCSGkLiEJIAIanxkPw P6kECIlBSGoISYKQGISkhpAkCIlBSGoI\nSYKQGISkhpAkCImZzCGJ/xTzqBu9/nPNCEmCkBiEpIaQJAiJQUhqCEmCkJjJHdLPllYv3EkU\nXzm/pvkoDRsvNTxglZO65fZU7jbrau6ipgki/T6IKwFCYiZ3SNcc6N8w7SwtWdmTfKIuSeHF\nRwesclY1Ddq3WVdzF0dNEJl2pDM6 ha/D JK0Ods5yn1JDUMEo1rGCShZXMiQ94HGfa OaWF\n9BTR ZqWo4YZcnp2G4WfzryXW9/YQ/Zt5lX74ugJIiOl8fsgrgQl7joou8JLmquQXjG/Lni8\nzcjYSOE268Z7jd8Q2beZIeXvHjVBZJcT0ejoW/w iCtBzNG U4t1ahgkwp6eEsS1bE5Ey/Z4\nHaDwpzpdhfSa fWaJ3caub9CHt5h3diwfHGK7NvMkPJ365kg0u DuBLgHImZ3OdIPyUaqt1y\n0jhkXjtjh7Srt/5Rsm8zQ8rfrWeCSL8P4kqAkJjJHVJjx9DG6h66uelcqqUmlgtpN 2f9mv7\nNuuqfbeeCSL9PogrAUJiJnNISaO1qbrxdaL4fbNmNu2nfEj0xLyu3G2Zj79zd uZINLvg7gS\nICRmMofkmpYJIv0 iCsBQmKmUEiaJoj0 yCuBAiJmUIhaZog0u DuBIgJGYKhVQShCRBSAxC\nUkNIEoTEICQ1hCRBSAxCUsN0XBym4xIgJDWExCEkAUJSQ0gcQhIgJDW/z0YqE0JiEJKa34ds\nZUJIDEJS8/uQrUwIiUFIan4fspUJITEISc3vQ7YyISRmKoWUMvaY/7l7jN HbGVCSMxUCim9\nr9f8z91j/D5kKxNCYqZSSKXw 5CtTAiJCVxIudkg00brsobG9nXXz3 WqHPl/OqmA3hrpw1C\nYgIXUn6yyKWJ4ebZ22l3uIuW3NY1uG5W98iQhoec8PuQrUw9jvadWndCwyDRmIZB rVsTiTp\nfZDznjenULPHkAqTRbYQ/aieKGEcajdOEiVrt44MydkEkX4fspWpzLMcQslKmiBSUpgs8lWi\nDUusj r2vVJlTeTa NTIkJK9Tvh9yFamTkf7Ti3erWGQSFTDIF1aNifSo2EUz5vTryukwmSR\nr40KaeF6nCPpgnMkJmjnSIXJIgshHTNOmG/xqlsRki4IiQlaSIXJIgshUdPtPYm1df0ISReE\nxAQupPxkkUUhnburbu6KU/j4WxuExAQuJAXXFRFCkiEkZgqFNHzYaHf9IL8P2cqEkJgpFNK2\naXemx19qFL8P2cqEkJgpFFJJ/D5kKxNCYhCSmt HbGVCSAxCUvP7kK1MCIlBSGqYjovDdFwC\nhKSGkDiEJEBIY/L7/VMFQ0gMQhqT30drBUNIDEIak99HawVDSAxCGpPfR2sFQ0gMQhqT30dr\nBUNIDEIak99HawVDSMwUC8nNHwD3 2itYAiJmWIhuZkg0u jtYIhJGaKheSG30drBUNITPBD\nenlh9bw1SYp8s3bemkG8tdMCITGBD lM1Z7hMzdspJvujb957cNFIaUGx H30VrBusfbdw50\n9msYJBrTMEivls2JDHgfJOF5c5JlC mo9ddih6ndOEPUvsvNBJF H60VbCKmOoRSaJsgkkk/\nNK1pw5uUnduu FO78wPj8PtorWBd4 07Bzp7NQwSiWkYpEfL5kT6vA/S73lzEmULiejsi7dP\na/tV1fCokMbl99FawXCOxAT HCnVZX55uPmY0UF0 AWEpAVCYgIf0pb6I n4stW0ZPlbp25Y\ng5C0QEhM4ENKb6ifPv87fdR1V03dgwmEpAVCYgIfUun8PlorGEJiENKY/D5aKxhCYhDSmPw \nWisYQmIQ0pj8PlorGEJiENKY/D5aKxhCYhDSmPw WisYQmIQkhrmteMwr50AIakhJA4hCRCS\nGkLiEJIAIakN9I /zLiGBzUMkuzT8EzRgIYxhvuS4y80rkEdm9OvY3uGtGxOn/t/fItJ69ic\nnAoLCWByQkgAGiAkAA0QEoAGCAlAA4QEoAFCAtAAIQFoUDkhXWeYarOXe //6pwVZ/1ak9i3\n6mY2H ZrNdGK9wL2CP/p/u6TUSonpPrnI5FI7o/U3Ln02JvfunbYpzVZvLT99LfnJthaTbTi\nvYA9Quyn 7tPRqmckGp25S9GqtrNX2 mOZ84Raueu08SnTP Y/RaTbTivYA9klMxR8loFRPS\neWP1DVfefSpzeXu19QepFj3t4 ocDMdHr9VEK94L2CNZFXaUFKmYkLrm/evhw7fPy/zJ4J9/\nzfp68/f9W5uexh ytZpoxXsBe4TYT/d/nxSrhJB GQ6HD2QuDdRutr79vN766sMustfkdwvW\nFP3h4txaTbTiveDbHrFVxB4Z dN93ycjVEJI/R0dHbm/ dC4wfq6M/uivdGvNdkz5/kRN2fX\naqIV7wXf9khOZeyRkT/d730yUiWElNHx3SGiRO1W63Ks6ghRd3i/T6vy2ytel9ZqohXvBewR\n/tN93iejVExIPXNWnTl1d/0gbd5EdM Nx07dfpOGv7pViuRVT1r/9E3CWpP8WvnB3gvYI3mV\nc5SMVjEhUfsts rufIto5S3mW6xV8 ferfHvAbuyx8h4IbMm9lr5wd4L2CMFFXOUjFY5IQFM\nYggJQAOEBKABQgLQACEBaICQADRASAAaICQADRASgAYISfZ3H3Gx8KxL CWlv3Yz ijRD145\n/kK//qdL3/PZR60/PbP7n97zzr9 MXPjwc Efpm58PKX3vvOz/yQ6Ob3Hi96zN2/52568MyK\nOPtBDo1cn8kFIcnKFtIb1h5fdbe7tXmj8DQNf WT40/9vvPCD9//8D E7ic6/O6PrnzoCxf8\nzLxx7bsuzx7fz1/wme8 9Dnz3tQ/fLboz8x9odrVOmVWxOEPcmjk kwuCElWtpBWl7LHix70\nROjfx1/ v7/3LNHQR9 bpjnvfovo/Mf nGj7O9c8kj2 P/mnZoqD/ XDRAcu Hb MT0XPuJq\nnTIr4vQHOVS8PpMMQpJlQmr5/Lvf ef3m 9bhm 7/KLPbF50oX1v/o70issv vhGK5/CpdP/\n8icX/cGMg/ZAn3/ 8s8RPfmXF1/62SeJ/ikUCn02 9YuP0jGZ/6m9S8v/v36Liosmx8p 6CM\n1J99YcTCueH//R8vvfjT6/Jr/4PMX9ipD/WlLpljXbo3dJCO7KXs8Z16IPN3i654u/mjZ74/\n/1ddfxI6mb3w b//9Zcufd/ss0Wr IHp5pXHQ/9CVtNHi1bE8Q9yOHLR kwyCElmhfTTt/3z\nv718U6iJ6JuhmS/94A//yn65KdxxX2julqc//pFLii/9zX/ wdYff L9uX/p6Uuf/OhDL9BT\noekvvPDPoRfoP8KhekKEKgAABo9JREFUXQcyIRUGyfjc /7iV5EnLjSPq/yy ZGyD8rYFnp0\nxMLZ4V9  xee33xNaNQv55//IB0OrbQutYSsLnPHd1b6Lz5mfn0xlJ/y4Kr/lrvw5T/ yy1n\nn3n7V4tW8auXmS3UX/ZfzXtnfLh4RRz/IIcjF63PJIOQZFZIH/0T65/EmnZhNP0HHzef7p0h\nO6TCHR/4uHnh9IWXUOFSd6jZvHT07jezy3459BPz691fMh/R/X/OJWqw9vhfF4 e 4GhNrLu\nPVlYtjBSQ/5pujV0asTC2eE//adWtlWXJoo3YUPou/TLUOZU/9ehB6xv eN78MT2WRdZLxf9\n72iwF/ Tm8he41esrx8oWsUfh35L9KGloTcpfVlj8Yo4/kEORy5an0kGIcnMkN4MXWNdWhd6\n4XRosXXp47mQCnecCF1vXfrcJVS4dP4/ffDlornWvvyOwr85efnnCyEVBsn9wEusN0CPhVoK\nyxZGKoT0//7ByIUzw58N3ZAwPRx6LXnclP0w4vl3Vadpc jH1uXfhu6xvuWP75 FQh96KXPp\nk5/KDfzbkD0Nw5ffZX396gVFq3j2bWvpeOjAZRtoX i54hVx INyq Vg5Pz6TDYISWaG9Fro\nTutSS j7b2Tft1TnQirc8Wr2knlH4RK98l9C/6n6x0O5gaxff80XmFs//p63vz30d4WQCoPk\nfuD/ZX19wfyFvbBsfqRCSH/15yMXzgz/RijnJ5lL1idntPrtV5iJ/TL7Bmx36DvWt/zxfW7T\no9UX3GVd tLluYG/fbH9cvblD1pfzR9ZtIqfmkOPvp mLaDVF/YUr4jDH5RbLQcj59dnskFI\nMjOkXaEV1qUXQz/YmT37qMmFVHxH5oCYdgkVLpkn2a1f/2 hv8h9SJ09eL7w9mVt 37zgaKQ\nCoPkfmCmjX8LPVa0bH6kQkgf/duRC2eGfyN05Y6MSNdPTdbJ/I2hm60XrSOhzAftz4eesb4V\nn7rQ19 2z/w64925q//4P w78od70So2XU7zaulfP0IzvjhiRRz oNxqORg5vz6TDUKSmSGd\nCS2wLn0/9PP/yH4k8IlcSIU72kPXWpc dQkVLmWtCf0oeyFz8BwJXWV HXpnUUiFQXI/8GLr\nt0O/F/p50bL5kdgrkr1wdvhY6KujVr/5gmyfw /J/ObQ8lCH9S17fL/14GvWlWdC682vX/qj\n7AP6L/qO/dj84V60iltCxy5/yDz T192z4gVcf6DHI5sr8 kg5Bk1ocNH/ A9W7nn9/VPfR7\n1icJr U/bMjfMXzZh82TmMNvu4QKl16fZb0iHA19K7ts5uA5kPkVeHXob4j JTSU/bAhP0ju\nB2bOjqZdFC8sWxgp86CM3DmSvXDu2Pyr3 s0vz52s73Uz0IP5i41XPw7s5IPZl8 ssd35IJ/\nsF5Drs5csc9Jng8dsjc9f7gXreLgxTeH9lPq0ttCvx6xIs5/kMORcY4UMFZIL17wled ttA6\nfb4p9LWXvvehv7NDKtxxa2jGs2s/ NlLii6dufST67Y89bfvyf52S/bgOf/Hf/TcK0u  MVL\nt/YtD614JhNSYZDsD/zjP1u75euh UXLHs2PlHlQxs3ZT 3shXPH5r9f MnHXrrlwq/lxhr6\ns8seyXiTTrz3T /5zl 9Y7t5FvPII18NLX3kkZPUHPr7VWtmhv572nohqs8 5NoP5Te9cLgX\nreJXLnuf9fV977d/TyizIs5/kMOR8 sz6SAkWeY3ZDf//SUXfdo6iR687rJLPv/qnPz79/wd\nqeb//I5P/HTRO4ov7Z3 /gs/MN3 lTt78Oz63Lv 4Oru5y/7/cO/ /SFH8n hmx kOwP/Ojr\nX3jX71/VW7xsfqTMgzJ kfmYubBwdnj65f9z6YV/ttJ QYrYHz5sMV8Mjfe864vWS0JD7rbn\nKb3m0xe/ xN3W5 Y537fh jD1 Q3vXC4F63it0M15te7QnX2UpkVcf6DHI6cX59JByE59 U/\nLOPgDv9M0tCHv h84fHNvqy3xEdmV0S30tfHbwjJiVUzzF/sO3/vf4y/ZMmctvGY9XuxukI6\neMHKkh bWRHNvKyPzxCSE4 H/udzT3/ubS X8Uc4bWP4Hz V0BXS8Jc/nRh/KdWK6OVpfXyG\nkBx5/NOXvOtvXyznT3DcRuSDV oK6dbfP bh0REnfzHKFW/r4y EBKABQgLQACEBaICQADRA\nSAAaICQADRASgAYICUADhASgAUIC0OD/B0 2S8cvthfUAAAAAElFTkSuQmCC",
      "text/plain": [
       "plot without title"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "word_ratios %>%\n",
    "  filter(!str_detect(word, \"^#\")) %>%\n",
    "  group_by(logratio < 0) %>%\n",
    "  top_n(15, abs(logratio)) %>%\n",
    "  ungroup() %>%\n",
    "  mutate(word = reorder(word, logratio)) %>%\n",
    "  ggplot(aes(word, logratio, fill = logratio < 0))  \n",
    "  geom_col(show.legend = FALSE)  \n",
    "  coord_flip()  \n",
    "  ylab(\"log odds ratio past (before cutoff) / now (after cutoff)\")  \n",
    "  scale_fill_discrete(name = \"\", labels = c(\"past\", \"now\"))   theme_minimal()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "In the analysis of my data, we see some changes: now older words are somewhat associated with the German language/dialects (`moin, kriegt, bildung, anstatt, neuer`) and random Top-Level-Domains that have fallen out of popularity (`ly, fm`) while the newer trends are \n",
    "\n",
    "- english speaking (`yay, travels, happily, worries`) \n",
    "- location based (`portland, berkeley, iceland, zurich`), fitting an increased travel compared to earlier times\n",
    "- phd-related (`lichens, ggplot2, markov`) \n",
    "- and modern-ish inventions. `emoji` anyone? 😂\n",
    "\n How do your trends compare with what you expect? \n",
    "And now let's do the whole thing by just looking at people I replied to in the past compared to now:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "data": {},
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAA0gAAANICAMAAADKOT/pAAADAFBMVEUAAAAAv8QBAQECAgID\nAwMEBAQFBQUGBgYHBwcICAgJCQkKCgoLCwsMDAwNDQ0ODg4PDw8RERESEhITExMUFBQVFRUW\nFhYXFxcYGBgZGRkaGhobGxscHBwdHR0eHh4fHx8gICAhISEiIiIjIyMkJCQlJSUmJiYnJyco\nKCgpKSkqKiorKyssLCwtLS0uLi4vLy8wMDAxMTEyMjIzMzM0NDQ1NTU2NjY3Nzc4ODg5OTk6\nOjo7Ozs8PDw9PT0 Pj4/Pz9AQEBBQUFCQkJDQ0NERERFRUVGRkZHR0dISEhJSUlLS0tMTExN\nTU1OTk5PT09QUFBRUVFSUlJTU1NUVFRVVVVWVlZXV1dYWFhZWVlaWlpbW1tcXFxdXV1eXl5f\nX19gYGBhYWFiYmJjY2NkZGRlZWVmZmZnZ2doaGhpaWlqampra2tsbGxtbW1ubm5vb29wcHBx\ncXFycnJzc3N0dHR1dXV2dnZ3d3d4eHh5eXl6enp7e3t8fHx9fX1 fn5/f3 AgICBgYGCgoKD\ng4OEhISFhYWGhoaHh4eIiIiJiYmKioqLi4uMjIyNjY2Ojo6Pj4 QkJCRkZGSkpKTk5OUlJSV\nlZWWlpaXl5eYmJiZmZmampqbm5ucnJydnZ2enp6fn5 goKChoaGioqKjo6OkpKSlpaWmpqan\np6eoqKipqamqqqqrq6usrKytra2urq6vr6 wsLCxsbGysrKzs7O0tLS1tbW2tra3t7e4uLi5\nubm6urq7u7u8vLy9vb2 vr6/v7/AwMDBwcHCwsLDw8PExMTFxcXGxsbHx8fIyMjJycnKysrL\ny8vMzMzNzc3Ozs7Pz8/Q0NDR0dHS0tLT09PU1NTV1dXW1tbX19fY2NjZ2dna2trb29vc3Nzd\n3d3e3t7f39/g4ODh4eHi4uLj4 Pk5OTl5eXm5ubn5 fo6Ojp6enq6urr6 vs7Ozt7e3u7u7v\n7 /w8PDx8fHy8vLz8/P09PT19fX29vb39/f4dm34 Pj5 fn6 vr7 /v8/Pz9/f3 /v7////8\nl6Z1AAAACXBIWXMAABJ0AAASdAHeZh94AAAgAElEQVR4nO29DXwc5X3vu01ySUJI0uaQ9DSl\nTc5NT5OcJrl56VvaJidNetOee0cre2VkbEtEKBhbNtjGFdYRYOxAgDjhGBywaYhJeImA49AU\nAwr4hcYKMTaOwDaOjWvLFsS2kNdavWuv1lrtc2dm55mX1e7qvzPzzDMj/76fj6XZmWfmmf3v\n8/XOzM7 FGMAAM/EZO8AADMBiASAD0AkAHwAIgHgAxAJAB ASAD4AEQCwAcgEgA MBNFmhgR\n3sVIalJ0F9kh0T2w0dSE6C4mB0X3wMZS50V3kRuYtslMFOm8 CE4lMyK7iIrfggOJ8WL1C 6\nBzaaFC9SatomEMkVEIkIRIowEIkIRCICkUQxQ0T6/8B06IWCSKKASBcKeqEgkigg0oWCXiiI\nJAqIdKGgFwoiiQIiXSjohYJIooBIFwp6oeSJNLTlutr5i35wzsWqvUq3x84hEhGIND16oaSJ\ndPCbj5zOsr4n6l vfF2IlAcihQK9ULJEOvrNN/MTrzZkKl4ZIuWBSKFAL5QkkbLXvMYOLp23\nqrM2d9svGEt e07dxnHWv64 0XyY5ZSdrY1NXZuvq3 SZZTnWhqv2sP4Mta1smbpLoikA5FC\ngV4oSSK9dAv7zbz97D/qbmHPPMDY9XemTi 5n628ZWB8c 0gi69KT7bM3c064wNZZVk/2149\nwJflGtenz7ZaImXHXTHa7269ChhIjonuYiyAZyF7lEYAvVDpvhIVtA64RIi0fhu74afq75bH\n2Y5NrEvpUd9q9nUp6uFeZs4LLN7O2I8bGEsrr2eVpxibrN3Glx3Rmr5kiTSaBEKRPUojQPkC\nWu9UIkRq U22WjUj940D7LEn2YtVOW1m/lfT4yy l7G2lerbjXIwq6jTbHEbX9ZRNcnYG5ZI\n58dcMdzvbr0K6E OiO5iJIBnIXuURgC9UKN9JSqYFirStd0Diqrqr5QRtvIo 1XVpCXS4kdZ\n/GWbSLvVeQsf48t2ar9P4BxJB dIoUAvlKRzpNX7WOMzuV vq0o/c7MhxtFnTihvqIdziZ1O\nkdQjwExiB192QHmLsQ6IpAORQoFeKEki/XQDO7ry6vvS6xrWaW99K1e/dWrZRta8Zii9acGo\nU6Ql3Zm2WYN8WWbe uFTqyCSDkQKBXqhJIk0VH8wP5EZ034O3Faz4N40O3vbgvlrTzGnSO03\nJBr3MXPZ0eWJpZ3KSY/9QyQiEGl69ELJ kD2UP1PenLnu/93Q0f5dlmlU0T3EIkIRJoevVDS\nbhHq27zkigXX/ iNaZpBpDJApFCgFyrsd39DpDJApFCgFyrsIgkCIhGBSNOjFwoiiQIiXSjo\nhYJIooBIFwp6oSCSKGaISMi1owGRRAGRiECkCAORiEAkIhBJFDNEJNknIKGkSKEgkigg0syl\nSKEgkigg0sylSKEgkigg0sylSKEgkigg0sylSKHEieQhANJOVtnvcQvFgEhEIFIxihRKmEhu\nAyAPHHM zh0cdtV/eSASEYhUjCKFEiWS6wDIte1uuqsUiEQEIhWjSKEEieQMgLQnPlopkLtW\nN121k7Hum6 oXX2GsR2LE3UbM61Vs5ez1Lr6mpbjbFJ5vvFu9dAurRxk7IxyxrEZb0AkIhCp\nGEUKJUikggBIW KjlQK5YoBtq0mzRevTo3c0s56q/ZM9y7awRvUdaeW6ocwjCzJqk NjNpEc\nm/EGRCICkYpRpFCCRHIGQNoTH60UyK1aiPcbbHicsd3VueNKF2OTTBPpuJbUlZvbweJP6Bcb\nLJGszfCOMsOuGEy5W68CUskh0V0M9YnuYTgle8yGkmKVOleigqOeRHIGQNoTH60UyN2MpZRj\n7EBrXd1cJZu7r7q57bQuUoeis4XFOwpEsjbDO0LSqmBkj9lQUkkBvSWtOgMg7bFAVgrkS7pI\nZ2ZvybA9inoY1PvsmuoOTaQ9inF9QmvCRTqlifTyFJFyWVeMD7pbrwIGkxnRXWQGRPeQHZI9\nZkNJkUJN9JWo4KQnkZwBkHYDrBTIvEgdcdWhh5RsVjvtub9FE lN/citxxQpW9XJWGdxkVyC\ncyQiOEcqRpFCCTpHKgiAtBtgpkDmRXpdOXR V4tydnvDsVyqdQNr2jzCbmw m22v6eMisYUP\nsvFvQaQpQCRJFCmUIJEKAiDtBpgpkHmR2I umLdheMXcnraGWfX3jLCtiQaW k7t5c2HzHck\n9uo1V9/wa W3EKkAiCSJIoUS9YEsNQBSEhCJCEQqRpFCCbtFiBgAKQmIRAQiFaNIoXD3tygg\n0sylSKEgkigg0sylSKEgkigg0sylSKEgkigg0sylSKEgkigg0sylSKEgkihmiEjItaMBkUQB\nkYhApAgDkYhc8Id21EJBJFFApBkBtVAQSRQQaUZALRREEgVEmhFQCwWRRAGRZgTUQkEkUUCk\nGQG1UKJEKpKzSslMtf0Rc0F/zzwPRCICkYgIEqlYzuqUzNTCUFUGkSoDIomHWigxItFyVouE\nqkKkSoBI4qEWSohIzpzVjPJcS NVe7RDOz07lWer6qGqHGOeZo Rs6pNZm9ak80nsBY29AhE\nIgKRiAgRyZmzmlWW9bPt1QPaOZKWncqzVfVQVY4xT7PHyFnVJtc3j/ME1oKGHoFIRCASESEi\nOXNWs8pTjE3WbtNFekJ7cfLZqg6RjHmqPTxnVZ18tGmI8QTWgob84eg5dyRdrldBD K7COJZ\nyB7IsqFXqsR8607CykVy5qxmlb3qvMVtukhaFoqRreoQyZin2sNzVrPKncpr6kaMBNaChvzh\nWMoVfefcrVcB55J94vsQ34PsgSwbeqVKzLdi6isXyZmzmlV2q/MWPqaL9BJjZraqTSQ TxWJ\n56xmlcbVKzRh8gmsBQ0r3icnOLQjgkM7IkIO7Zw5q1lFPczLJHZwkXi2ql0kM29V6eQ5q1ll\n33DDg4wnsBY0rHifnEAkIhCJiBCRnDmrWWVJd6Zt1iAXiWer6qGqBnyedoXByFnVJg9Vv8IT\nWAsaVrxPTiASEYhERIhIzpzVrNJ Q6JxH Mi8WzVXi1UlWPMO6Nd/s7nrOqfIz1SlzISWAsa\n9la8Uw4gEhGIRETMB7KOnFWhn6y6BCIRgUhEBN0iZM9ZhUiigEjioRYqgLu/y4h0rNZg7zTb\nIDckApGIQCQi BqFKCDSjIBaKIgkCog0I6AWCiKJAiLNCKiFgkiigEgzAmqhIJIoZohICIik\nAZFEAZGIQKQIA5GIXNCHdpUUCiKJAiJFnkoKBZFEAZEiTyWFgkiigEiRp5JCQSRRQKTIU0mh\nIJIoIFLkqaRQskWamshq3OMq9p5xiEQEIhGRLFKRRFabSL/91vzalt I6BciEYFIROSKVCyR\n1RIpd/X3R8cfnSNizEMkIhCJiFSRnImsLJ paiSsZpTOAeUIYynlKM9ezSk7WxubujZfV/ k\n154hEhGIRESqSM5EViNT1UhY1X41rx9Kt30zw7NXWXxVerJl7m7WGTfDwiYnXJEedLdeBQwm\nM6K7yAyI7mFiUPZglkklhTrfV2KB9d pOJGciaxGpqqRsKqJ1LdEUeqPM569yuLtjP24gbG0\nYp5SjSaBUGQPZpn4UkDrnUqcSM5EViNT1UhYVUWaWPb9gdEt81M8e5XF9zLWtlJbdpBvIjPs\nisGUu/UqIJUcEt3FUJ/oHoZTsgezTCqq1LkS80cDEMmZyGpkqhoJq6pIr1Sl1ZlXbeXZqyz \n8hSRXIJzJCI4RyIi9RypIJE1n6nKE1aVzk5Fs7l K89ehUgFQCSxVFIoqSI5E1mNTFUjYVX9\nNVr//eHMk4kzPHsVIhUAkcRSSaGkiuRMZM3lM1WNhNU 9Vf32vlzb1CbGNmrEKkAiCSWSgol\n9wNZRyJrkEAkIhCJiORbhOyJrEECkYhAJCKyb1qVBEQiApGIQCRRQKTIU0mhIJIoIFLkqaRQ\nEEkUECnyVFIoiCSKGSIScu1oQCRRQCQiECnCQCQiEIkIRBLFDBFJ9nlKsLgvFEQSBUSKIO4L\nBZFEAZEiiPtCQSRRQKQI4r5QEEkUECmCuC9UMCJNjYEsQlbZX/yhgKxIiEQEIhEJRKQiMZDF\nduXgMDtwzPEwD0QqAUTyG/eFCkKkYjGQJVjbXmQmRCoBRPIb94UKQKSCGMgT1yaWH1RO8tTH\nSeX5xrv1H qxXGvV7OV8gXZol/z2nLqN40ZkZNa hsd9gkhEIBKRAERyxkDmGu4aPblC6bZS\nH1ccH9N/aOY0qu9IxgLt4fV3pk4vud IjHSu4Q2IRAQiEQlAJGcM5BHlLca2K91W6uMT6jLt\nhyESX6A 7FJ6GOvaZ0RGOtcwOD/miuF d tVQH9yRHQXIwE8C9lDO1jcF2q0r8SCtG8iOWMg\nO I59fBO6bZSH7W8Bu2HIRJfoD58sSqnbcCIjHSuYYCkVcHIHtrBIqCA/iWtOmMgd81WZ3Ur\n3Vbq40vGD0MkvkB9 KuqyfyUHhnpXMMge94VYwPu1quAgeS46C7GxT LCyz7232hMqlSS3wT\nyRkDeVDpY2yn0m2lPhaIxBeoD9U3LsaOPmNERjrX8AbOkYjgHIlIAOdIzhjIifmbMm82q4aY\nqY92kZo2j/AF2sOVq986tWyjERnpXMMbEIkIRCISgEjOGEj22tI5q/Yrb1qpj3aRtiYa ALt\n4cBtNQvuTRuRkQOONbwBkYhAJCJBfCDrjIHMTjB2RBmdbiWxQCQiEIlIILcI2WMgcwvvHknd\n0uJ5m96ASEQgEpHg7/4 eeOc bcn/d1mxUAkIhCJCL5GIQqIFEHcFwoiiQIiRRD3hYJIooBI\nEcR9oSCSKCBSBHFfKIgkihkiEnLtaEAkUUAkIhApwkAkIhCJCEQSxQwRSfZZS5B4KRREEgVE\nihxeCgWRRAGRIoeXQkEkUUCkyOGlUBBJFBApcngplBCRigSrls5NLfx20dGrZ6eUzoLYVb B\nSEQgEhERIhULVi2dm1oo0l0twxNKp9VeCBCJCEQiIkCkaYJVpxPpW/eJiFYtACIRgUhE/BfJ\nGayaU3atbrpqZ5HcVN4 vm317LoXGOu  Yra1WdYS7x6Tp9 aMdXLZrNWvETdQKRiEAkIv6L\n5AxWZfEVA2xbTXpqbionfu3hsYdr0mzR vToHc2MrdXfkbT2xqrFs1m9AZGIQCQi/ovkDFZl\n8a2M9SpvTMlNNYmrjXtUTYZVt3ZX5 wi5Vctns1qkO53Reqcu/Uq4FwyJbqLIJ6F7MEdJN4q\nVWK 9Z9dhSI5g1VZfDdjKeXYlNxUS6R8A3agta5urpK1i5RfUjyb1QBJq4KRPbiDREgBXSet\nOoNV9WsJeZGcuamWSPkGZ2ZvybA9TpHyS4pns3oDh3ZEcGhHxP9DO2ewqk0kZ26qidGgI66O\nu4eKiVQ8m9UbEIkIRCLiv0jOYFWbSAW5qRyjwevKofO7WpSzU0Uqns3qDYhEBCIR8V kgmBV\nm0jO3FTe3mjAfnTFvA3DK b2ThGpeDarNyASEYhERMAHss5gVR8QkM0KkYhAJCIibhGyB6v6\ngIhsVohEBCIRkXX397Fag73TtxWQzQqRiEAkIvgahSggUuTwUiiIJAqIFDm8FAoiiQIiRQ4v\nhYJIooBIkcNLoSCSKGaISMi1owGRRAGRiECkCAORiFxAh3beCgWRRAGRIoa3QkEkUUCkiOGt\nUBBJFBApYngrFEQSBUSKGN4KBZFEAZEihrdChUWkIuGsdoykO9/yVyESEYhEJCQiFQtntaOJ\ndOAY8y1/FSIRgUhEwiHSNOGseZHWtvvXIUQiApGIhEKkwnDWna2NTV2br6t/0sxfVUVqrZq9\nXDu027E4UbexuG50IBIRiEQkFCIVhrOuSk 2zN3NOuMDPH9Ve0dqbNfOkXqq9k/2LNvCV825\nIzPkckU6Q8kJ0V1MDIruIXcBieStUJOpUksCFKkwnFU9hvtxA2Np5XWev2qJdFzpYmzSXBUB\nkYKRPbyDQ1ABXQdEuqAwnHUvY20rtROjgzx/1RIpd191c9tpc9XxIVcMpNytVwF9yUHRXQz2\nie5hqE/28A4Oj5U6V2K dXlMvEiF4awvmyLx/FVLJMZ6n11T7TWgCOdIRC6gQztvhQrFOVJh\nOKslEs9ftUTKaoF493vNFIJIRCASkVCIVBjOaonE81c1kZo2j6gibW84lku1bvDYI0QiApGI\nhEKkwnBWSySev3pGFWlrokE7R2prmFV/z4jHHiESEYhEJBQi R/OOh0QiQhEIhIOkfwOZ50W\niEQEIhEJiUhBA5GIQCQiEEkUEClieCsURBIFRIoY3goFkUQBkSKGt0JBJFHMEJGQa0cDIokC\nIhGBSBEGIhGBSEQgkihmiEiyz1yCwmuhIJIoIFKk8FooiCQKiBQpvBYKIokCIkUKr4WCSKKA\nSJHCa6EgkiggUqTwWqiARZomUDUwIBIRiEQkWJHKBqoeOOZfR9MBkYhAJCKBilQ UNXPJNXp\ngEhEIBKRIEUqDFTdtbrpqp2MpdbV17QcZ1qS6qTyfOPdfAZPWc0oz7U0XrXHbJhv5A2IRAQi\nEQlSpMJA1RUDbFtNmq1cN5R5ZEFGSwlS5x0fM2eYKavL tn26gFzvt7IYCLtipF d tVQH9y\nVHQXo KfxYDsAR4UXgs11ldiwbj/IhUGqm5lrFd547gWaZeb25EX6QnGzBlmyupTjE3WbjPn\na404SFoVjOwBHhTCCiggabUwUHW3erSmHOtQdLbkRepgzJxhpqzuVVde3GbOj9siUvCOJBa8\nIxEJ8h2pMFD1JV2kPYpx4UEXSZ3HZ1gpq6pxbOFjZkOtkUdwjkQE50hEgjxHKgxUzYv0pqJd\nC  xROIzrJRV9Xgwk9hhNoRIBhDJP7wWKkiRCgNV8yKxG5vPZttr rQk1bwjxgwrZXVJd6Zt\n1qDZECIZQCT/8FqoIEUqDFQ1REp9p/by5kNMS1LNO8JnmCmr7TckGvdZ8yGSAUTyD6 FCvQD\nWXeBqsYfYvYViEQEIhEJ9hYhV4GqEKkUEMk/vBYqAnd/Q6RSQCT/8FqoCIgkAohEBCIRgUii\ngEiRwmuhIJIoIFKk8FooiCSKGSIScu1oQCRRQCQiECnCQCQiEIkIRBLFDBFJ9rlLQHguFEQS\nBUSKEp4LBZFEAZGihOdCQSRRQKQo4blQEEkUEClKeC6UYJFsgZCFt8wdvXr2gPavgq1llf0e\n9sUORCICkYiIFckeCFko0l0twzntXwWbyx0cdr8vDiASEYhERKhIjkDIQpG dV/ nxQgEhGI\nRESkSM5ASE2k7E1rskbsY0u8es4i9d8Aj4M0fueUna2NTV2br6t/0oyI5FmS2qEdb 4NiEQE\nIhERKZIzEFITaX3zOI99ZGvvy//jsY9m/OOq9GTL3N2sMz5gtjWyJDWReDNvQCQiEImISJGc\ngZCqSI82qePXiH3kIvHYRyv sZ2xHzcwllZeN9saWZKqSGYzg/EhVwyk3K1XAX3JQdFdDPaJ\n7mGoT/YIDwjvlTpXYr51Wu9aJGcgZFa5U3lNnWvEPnKReOyjFf 4l7G2ldrFiYNmWyNLUhXJ\nbGaApFXByB7hASGugD4krToDIbNK4 oVWTP2kYvEYx t MeXTZHMtkbgkCqS2cwgl3XF KC7\n9SpgMJkR3UVmQHQP2SHZIzwgPBdqoq/EgknvIjkDIbPKvuGGB83YRy4Sj3204h8tkcy2lkhm\nM2/gHIkIzpGIiDxHcgZCahcbDlW/wmMfzYsNPPbRjH 0RDLbWiKZzbwBkYhAJCIiRXIGQuqf\nIz1SN2DEPvZykXjsoxn/aInEIyJ7bSLxZt6ASEQgEhGhH8i6C4QMAohEBCIREXuLkKtAyCCA\nSEQgEhHc/S0KiBQlPBcKIokCIkUJz4WCSKKASFHCc6EgkiggUpTwXCiIJAqIFCU8FwoiiWKG\niIRcOxoQSRQQiQhEijAQiQgO7YhAJFFApCjhuVAQSRQQKUp4LhREEgVEihKeCwWRRAGRooTn\nQkEkUUCkKOG5UNJFsmWxTot/f98cIhGBSERki2TPYi3OgWPmJERyApF8w3OhJIvkyGItztp2\ncxIiOYFIvuG5UHJFcmaxsh2LE3UbMyz57Tl1G8ebN6oNXq/6ZtXs5TxxVRMJSasWEMk3PBdK\nrkjOLNaeqv2TPcu2sOvvTJ1ecv/2uep71AOtrFF9RzISVzWRkLRqAZF8w3Oh5IrkzGI9rnQx\nNsm6lB7GuvalL9/Fct94QRfJSFxVRSpMWh09546ky/Uq6EF8F0E8C9kjPCB8qFSJ dadhAJF\ncmax5u6rbm47zV6syv pl3tvZocuH9dFMhJXVZEKk1bT/a5InXO3XgWcS6ZEdxHEs5A9wgPC\nh0qVmG8dNQgUyZnFyljvs2uqO35VlQ nPFrVt n7TBOJJ66qIhUmrboEh3ZEcGhHRO6hXUEW\nq/bX  5vOaF0qxY9o2r25ILDukg8cVUVCUmrNiCSb3gulFyRnFms2xuO5VKtG9jK1W dWraR\nsa2116htmjaP8MRV7WIDklYtIJJveC6UXJGcWay5toZZ9feMsIHbahbcq4o1PEs7FdqaaOCJ\nq2e0y99IWjWBSL7huVCSP5Atm8XanRD23UmIRAQiEZF9i1DpLNbJ3ut/KKxbiEQEIhGRLVJp\n2uas9 UCXVEgEhGIRCS8IgkFIhGBSEQgkiggUpTwXCiIJAqIFCU8FwoiiWKGiIRcOxoQSRQQ\niQhEijAQiQhEIgKRRDFDRJJ98hIM3gsFkUQBkSKE90JBJFFApAjhvVAQSRQQKUJ4LxREEgVE\nihDeCwWRRAGRIoT3QgUpki1UNavsL9Eo/pJf3ZUDIhGBSEQCFMkeqpo7OFyilSnSqZVxbaV8\n2Mkz/DEbvuvKeWt7ve4LRCICkYgEJxIhVFWDi9RRv14T53xS5Tdz3uSP2a2rTpz 7pJJjzsD\nkYhAJCKBieQMVdUO7frX1SeaDxfkq7L4ttWz615gbOfZl J83ZvazMfJqi711a0udWBIBSIR\ngUhEAhPJGaqqibTyloHxzbWDznxVFr/28NjDNWltDS5SR6N m4r eHdCC71b gTfbPa8K8YG\n3K1XAQPJcdFdjIt/FoOyh3gweC9UJlVqic8iOUNVVZG6FPVQLzPnBWe KourjXq0QC5TpMlF\n25kp0nPf0CZv/AHf7GgSCEX2EA8GgQW03qn8EckZqqqKlA9UbXq8IF81vpuxlHLMJlLHN7I2\nkRqcIp0fccVQyt16FZBKDonuYrhPdA8jKdlDPBh8qNS5EvPHfBbJGapqirT40YJ8Ve1ig1Ok\ntYY1 uM9 UO7LUV6qAScIxHBORKRwM6RCkJVlf0nlDcYSyd2FuSrThFphF9Z0B/3VamLBuNe\ng 0gEhGIRCQwkZyhqtrFhuY1Q lNC0YL8lW5SKnktngyqTbdr ifGvHHdyw/cWrN9TmPewOR\niEAkIoGJ5AxV1UQ6e9uC WtPFearcpEa9Q9in2Ls36v0a3b88ej6 vm3T7/T0wCRiEAkIsF9\nIOsIVS19i1AwQCQiEIlIgLcI2UJVJ49q17wlApGIQCQicu7 3lV9q9ezHG9AJCIQiQi RiEK\niBQhvBcKIokCIkUI74WCSKKASBHCe6EgkihmiEjItaMBkUQBkYhApAgDkYhAJCIQSRQzRCTZ\nZy9B4EehIJIoIFJk8KNQEEkUECky FEoiCQKiBQZ/CgURBIFRIoMfhRKkki2rEgi/t4vDpGI\nQCQickSyZ0USsUdKHjjmdQcgEhGIRESKSMSsyJKsbfe6BxCJCEQiIkMkZ1Yk6775itrVZxhP\niEytq69pOc4yynMtjVftYTxHUj20yym7VjddtZO1Vs1e7nEXIBIRiEREhkjOrEi2aH169I5m\nxhMiV64byjyyIJNVlvWz7dUDPEdSO0eKrxhg22rSrBHvSDoQyR/8KJQMkZxZkWx4nLHd1Tkj\nIfK4FtqVm9uR1fIaJmu38RxJXaStjPUqb9hFQq6dWC6IXDtfKhVUrp0NZ1YkO9BaVzdXyRoJ\nkR35Pz xJavsVR8tbuM5krpIRnikTSQkrQpG9iAPAqEF9Dtp1YYzK/LM7C0ZtkfJGgmRe5T8\n9YesolrDFj7GcyR1kV6aIhKyv8VyQWR/ 1GowLK/bTizIjvi6rnEQ0rWSIh8U9EuifeoIqlH\nf5nEDjNHsrhILsE5EhGcIxGRcY7kzIp8XTl0fleLcpYnRN7YfDbbXtOXVZZ0Z9pmDfIcSbtI\nTZtHPO4CRCICkYjIEMmZFcl dMW8DcMr5vYaCZGp79Re3nxIfUdqvyHRuI/xHEm7SFsTDR53\nASIRgUhEpHwg68iKLEFW6fS9XwuIRAQiEZFzi5AtK7IUEGl6IJI/ FGo0N79DZGmByL5gx F\nCq1IYoFIRCASEYgkCogUGfwoFEQSBUSKDH4UCiKJAiJFBj8KBZFEMUNEQq4dDYgkCohEBCJF\nGIhE5AI4tPOnUBBJFBApIvhTKIgkCogUEfwpFEQSBUSKCP4UCiKJAiJFBH8KBZFEAZEigj F\nkimSLW61/C2q/qasakAkIhCJiESR7HGr5UWyp6z6A0QiApGIyBPJEbcq9ksTU4FIRCASEWki\nOeNWNZGMSNVJ5fnGu3fO7WfspjvYjsWJuo0Z7dDOWMrTVj0CkYhAJCLSRHLGrWoiGZGqLL7i\n Bi7dR37xYKBnqr9kz3LtmgiWUvzaavegEhEIBIRaSI541ZVkXikKos/oc7un/fL t3suNLF\n2KR2scFaaqStGoylXNF3zt16FXAu2Se D/E9yB7m4vGrUiXmDwgWyRm3qorEI1VZXI9E2VW1\nTl16X3Vz22lNJGupkbZqMHrOHUmX61XQg/gugngWsoe5ePyqVIn51i25YkRyxq2aIi1 VA/d\nYuzxOSu0Q6PeZ9dUd5giGUttIrkEh3ZEcGhHRNqhnTNuVRWJR6rmRTo55/iSx1lWe1 8v0UV\nybEUIuWBSH7gT6GkieSMW7XujdgAACAASURBVNUuNhiRqroq2WWPsiOJk9sbjuVSrRu0iw32\npRApD0TyA38KJU0kZ9yqJpIRqaqr8tji84xtWna rWFW/T0jmkj2pRApD0TyA38KJe8DWUrc\nqjAgEhGIRETiLUKEuFVhQCQiEIkI7v4WBUSKCP4UCiKJAiJFBH8KBZFEAZEigj FgkiigEgR\nwZ9CQSRRQKSI4E hIJIoZohICIikAZFEAZGIQKQIA5GI4NCOCEQSBUSKCP4UCiKJAiJFBH8K\nBZFEAZEigj FgkiigEgRwZ9CQSRRQKSI4E hghSpfLJqmWg7JK0WByL5gT FClCkaZJVy4iE\npNXiQCQ/8KdQwYk0XbJqoGGrEIkIRCISmEhTk1W7b76idvUZq4EVtppWDjJ2RjmTUZ5rabxq\nj3ZoZyasFqzlEohEBCIRCUykqcmqi9anR 9oNhvYwla5SFllWT/bXj2gnSPxhNWCtXJZV4wP\nuluvAgaTGdFdZAZE95Adkj3MxeNPoSb6SiyY9FmkKcmqbHicsd3VOZtIPE7VEukpxiZrt ki\nGQmrBWuNJoFQZA9z8QguoPVO5Y9IU5JV2YHWurq5inn8YwtbtUTaqz5e3KaLZCSsFqw1PuSK\ngZS79SqgLzkouovBPtE9DPXJHubi8alS50rMt66T SPSlGTVM7O3ZNieIiItflQX6ZQmkmoP\nW/iYLlI hqtwLZfgHIkIzpGIBHaONCVZtSOuDrSHHCLxONVsVSdjnZpI6sFgJrHDJlLhWi6B\nSEQgEpHARJqSrPq6cuj8rhblLG9gD1td CAb/5Ym0pLuTNusQZtIhWu5BCIRgUhEAhNparLq\nj66Yt2F4xdxeo4E9bPXVa66 4dfKb7NK w2Jxn3MJlLhWi6BSEQgEpHgPpB1kawq7jNaiEQE\nIhEJ8BahypNVIVJ5IJIf FMo6Xd/H6s12Dt1GUQqD0TyA38KJV0kOUAkIhCJCEQSBUSKCP4U\nCiKJAiJFBH8KBZFEMUNEQq4dDYgkCohEBCJFGIhEBCIRgUiimCEiyT6DEY5PhYJIooBI0cCn\nQkEkUUCkaOBToSCSKCBSNPCpUBBJFBApGvhUKPEi2WIhHUyX mjcaCfmfjuIRAQiEalIpEts\nvJPYgT0W0tnzNKmPRUQ6cIzY6bRAJCIQiUhFImk3aX/8oi8mqj/7ti8spW3fEQtZEUVEWtte\n2SZKA5GIQCQilR7abfmUHs74 ie2kjbvjIWcVJ5vvJtHPNpTH1Pr6mtajhvL aqqQdmb1mSU\nTh4q1Fo1ezlv6hGIRAQiEalUpE/97/zvTf8XafPOWEgWX3F8jEc82lMfV64byjyyIJNfbqCJ\ntL55PGuJxBrVdyTe1BsQiQhEIlKpSO/ckf 95V2kzTtjIVn8CXXaiHi0pT4e14K6cnM78ssN\nVIMebRpiBSKZTQ0m0q4Y6Xe3XgX0J0dFdzEq/lkMyB7nwvGpUGN9JRaMFxfpw/Pz/tX AUkk\nZywki2sCGBGPttTHDkVnS365KdKdymusUCSzqQGSVgUje5wLR3QBSySt3hL79HW33rrkk7EW\nkkjOWEg9CohHPNqigfYoxqGa9tgSqXH1iqwp0qm8SGZT3mrcFaP97targIHkmOguxgJ4FrLH\nuXB8KlS6r8QCa7w6RMp95w9iKpfeTDsBcMZC6qLwiEebSG8q2rXxnkKR9g03PKiJxPMiNZHM\npt7AORIRnCMRqfwD2dwbe/d0TRZvOgVnLKQuCo94tIfV3dh8Ntte01cgUic7VP2K9svIi2RN\nm0fMpt6ASEQgEpFKRfrisxVt3hkLmRfFiHg8YxMp9Z3ay5sPFb4jqe9Dj9T1qb MvEi2NdFg\nNvUGRCICkYhUKtJld1W2fRexkEEAkYhAJCKVivTUJ39W2ZcNK4 FDAKIRAQiEalUpC99OvbO\nD39Ew6cdKKRcYKR/QCQiEIlIpSL97Ve/ZuDTDsgBIhGBSETwfSRRQKRo4FOhKhfp3DM/ OFz\n4sehWCASEYhEpFKRJldepH0ge8k6n/qXBEQiApGIVCrSutisze3P/Ms/xh7yaQfkAJGIINeO\nSKUiffL6/O FnxeyO0EBkYhAJCKVivSunfnfz14sZHeCAiIRgUhEKhXpkqfzv//tvUJ2JyhI\nIv0O J3fgUg0KhXp7/5evy08/fWvCNqhYIBIVCASjUpFevZtf7zo1m9d/eG3bxe2S0EAkahA\nJBoVf470s09ol78/XdlN4KEDIlGBSDQqFekYY6df3veWsP0JCIhEBSLRqFSk2MeanhJyxatU\nIKsjklWd9Cd5FSJRgUg0KhXp 7MvjV30lTteyfm8HyUDWR2RrOokRAoWiETDxVfND25QZfrQ\nAl93gx7ICpGCBSLRcHf39 nVH/T1nnBnICvbsThRtzHDU7i0QzsrndUIYPX4YSdEogKRaFQu\n0um2hR Pve/rt/m5F85A1p6q/ZM9y7bYRbLSWY0AVo8dQiQqEIlGpSJ987/GLq3 X/t8vvvF\nGch6XOlSy8vsIlnprEYAKyc94Ir PkIj2WM4FCT73VW4gtfinOgeBlLin8VAssR8a6w6r9pd\n2nLYX4k0nIGsufuqm9tOO0Sy0lmNAFaOyKRV2WM4FAis74VBiaTVhxs GvvQnPs852E5KQhk\nZb3Prqnu4AGrqj22dFYjgJWTc0dmiNBI9hgOBcnzLktMJtsvuofcaDIjuovJVKklxUVSOflg\n3WWxD9b4KZIzkDU7oM66v4UHrKr22NJZjQBWj AciQrOkWi4u2p34sZLfb1q5wxk3d5wLJdq\n3cADVlV7bOmsRgCrxw4hEhWIRKNikXK/2TTvstjFX/ en3vhDGTNtTXMqr9nhAesaudIVjqr\nEcA64K1DiEQFItGoVKRZl8be9tnm7V4vPxdSLpB1ur/a7AaIRAUi0ahUpD 88ie9IvajdCDr\n5FHtarjPQCQqEIlG HPtdlXf6veNfRCJDkSiEX6RhACRqEAkGhCpNLLHcCiASDQgUmlkj FQ\nAJFoQKTSyB7DoQAi0YBIokCuHRGIFGEgEhGIRAQilUL2MVVIgEg0IFIpZI/gkACRaECkUsge\nwSEBItGASKWQPYJDAkSiAZFKIXsEhwSIRAMilUL2CA4JEImGBJGKZKqWCKuzf4Ei/pKf wCR\nqEAkGsGLVCxTVRPpwLEpTW0ZqxBJEhCJRuAiFc1U1URa2152PYgkB4hEI2iRCjJVU vqa1qO\nayK1Vs1ezvrX1SeaD7Ocsmt101U79UM7YxaLb1s9u 4Fc5FHIBIViEQjaJGcmaps5bqhzCML\nMto7UqP6jrTyloHxzbWDLL5igG2rSWsimbOuPTz2cE2aL/K4GxCJCkSiEbRIhZmqave5uR2G\nSF2KetiXmfMCi29lrFd5QxXJmqWu1aN080V8e ICImWP4JAgrL4XCiUCIr3izFTtUHS2GCK9\nWKV9qbzpcRbfrR71KcdUkQpnmb8N0v2uSJ2btonsERwSkil3Fa6A6V8Lr/RJfBbWjcW iuTM\nVN2j5C84OERa/Kh ZcEukjXL/O0NHNpRwaEdjaAP7ZyZqm8q2jXwHi7SCe2QLZ3YaROpcBZE\nChiIRCNokZyZquzG5rPZ9po TaSmzSOsec1QetOCUZtIhbMgUsBAJBpBi TMVGWp79Re3nxI\nf0fammhgZ29bMH/tKWYXqWAWRAoYiEQj8A9ky2WqFiIiYzUPRKICkWgEf4tQ6UzVQoRkrOaB\nSFQgEo0w3/0tJGM1D0SiApFohFkkgUAkKhCJBkQqhewRHBIgEg2IVArZIzgkQCQaEEkUyLUj\nApEiDEQiApGIQCRRQCQiECnCTC S7HOTsACRaECkEsgewGEBItGASCWQPYDDAkSiAZFKIHsA\nhwWIRAMilUD2AA4LEIkGRCqB7AEcFiASDd9EsuWnlshNpdGrdLtfmQxEogKRaPglkj0/FSLN\nICASDZ9EcuSnQqQZBESi4Y9IzvzUrPLzVYnFe8wcVdZ98xW1q8 wtHKQsTPKGXbi2sTyg8pJ\n/thMT 1aWbN0lyqSsd6k8nzj3WzH4kTdxkxhCKtHIBIViETDH5Gc alZZdHh0bbqXp6jyhat\nT4/e0WyKlGu4a/TkCqXbFMtIT801rk fbVVF4uvFVxwf66naP9mzbEthCCvvOXveFWMD07WQ\nPYDDQjLtrsJ0MinRPZwfTo6J7qL0s6hEJGd alZ5XP0/v6ad56iy4XHGdlfnuDhHlLcY224X\nKZ eekTpUZ1Uus314k8wdlz7uvnklBBW3rOwpFXZAzgsiKrvBUNFSavO/NSs8qI6b HDPEeV\nHWitq5urZLk4HfEcYyfsIuXTUzuqJhl7Q k214t3qNu8r7q57fSUEFbe8/kxVwz3T9dC9gAO\nC8kRdxWmM9onuoexweSw6C5KPgsrpp4gkjM/Nau8rM5b9BjPUT0ze0uG7TFEOqWc2TVbndlt\niHRKEykfsbWzKi8YX8/4Uy69z66p7piSuOoNnCNRwTkSDX/OkZz5qVnlZ4xNzNnOc1Q74lnG\nHlKy2apOxjqVMweVPsZ2Kt38MZfjgHbI16F08/X0 dkBdfr limJq96ASFQgEg1/RHLmp2aV\npu6JLYkhnqP6unLo/K4W5Sxb CAb/5ZyZmL psybzUo3f8zlyMxbP3xqlTrfWE fv73hWC7V\numFK4qo3IBIViETDH5Gc akZZWdzounXZo4q 9EV8zYMr5jb  o1V9/wa W37LWlc1btV95k\n/DGX4 jyxNJO5aS5njY/19Ywq/6ekSmJq96ASFQgEg2fPpCtJD9VfctSX5wjyiilqSAgEhWI\nRMOvW4To alqnwvvHknd0kJpKgqIRAUi0ZBy9/fJG fMvz3p 2YrACJRgUg08DWKEsgewGEB\nItGASCWQPYDDAkSiAZFKIHsAhwWIRAMiiQK5dkQgUoSBSEQgEhGIJAqIRAQiRZgyIsk KQkZ\nEIkGRCpE9sgNGRCJBkQqRPbIDRkQiQZEKkT2yA0ZEIkGRCpE9sgNGRCJhk8iTRsPmVX2l1g1\n/zXYYkxdx9h26Y1RgUhUIBINf0SaPh4yd3C4xLqlRZq6jrHt0hujApGoQCQavojkKR6ytEhT\n8RQ9aQciUYFINPwQqTAespNlb1qTNVIeV25SWxys6lX2m9mORuRj8ttz6jaOs/i21bPrXjBT\nJLUGDyu9jLU8kZ2yjrHtDA7tAgMi0fBDpMJ4yE62vnmcpzxurc8xtvEm7bTGyHbkkY/X35k6\nveR Fr/28NjDNWmeIrlhA2P/3PQCy8w NnUdY9s4RwoOiETDD5EK4yE7H20aYjzlsT9 iE0u\n2KFLkc92NCMfexjr2sfi6qo9SjdPkey4hqXnPLWBHZifm7qOsW2bSJlhVwymSi6SPXJDRnLQ\nXYXpDPWJ7mG4X/yzGD5XYr6VqDCtSIXxkHcqr2mxWkbK480/YPsTY7oU WxHR QjMxMfjRTJ\noarUvlXdC9mj32VT1zG2bRPJ/6RV2SM3ZPhe3wuNCpJWC MhG1evyDIz5XFHQ 7edXkpeACQ\nHvn4Ky1XlZmxQDxFkq14cfOjuXl9q3YWWcfYtk2kyQlXpAdLLpI9ckNGctxdhelkUqJ7mBhO\npkV3cb6vxALr1uVpRSqMh9w33PAgM1MeRxNH5u2zS2FGPnYzdvQZPpenSLKH7l xn926fVaq\nyDrGtnGOFBw4R6LhxzlSYTxkJztU/YqZ8sjuuGlB1i4Fj3xcufqtU8s28rlmiuTBRYkM 9el\n17Ii6xjbhkjBAZFo CGSMx5S/6znkboBnvLIdiv/4pCCRz4O3Faz4N60efDGUyQn5qxS36mU\nB4utY2y7DyIFBkSi4csHspXFQ4YBiEQFItHw5xahSuIhQwFEogKRaODu70Jkj9yQAZFoQKRC\nZI/ckAGRaECkQmSP3JABkWhApEJkj9yQAZFoQKRCZI/ckAGRaEAkUSDXjghEijAQiQhEIgKR\nnMg kgodEIkGRHIie9yGDohEAyI5kT1uQwdEogGRnMget6EDItGASE5kj9vQAZFoQCQnssdt\n6IBINPwUyRa3Oj3F4ux6tS/NEtp5ByJRgUg0fBTJHrdanAPHrGmIFA0gEg3/RHLErRZnbbs1\nDZGiAUSi4ZtIzrjVjPJcS NVe9ik8nzj3ax/XX2i TBrrZq9nE bAatGICvrWlmzdJcq0olr\nE8sPKifN bwdX48v9wZEogKRaPgmUmHc6rJ tr16gMVXHB9jK28ZGN9cO8ga1XckPs0DVo1A\n1lzj vTZVqU713DX6MkVqlDGfKtdfj1zuTcgEhWIRMM3kQrjVp9SK1S7jcWfUN9sFPWgLzPn\nBU0kc9oIWOWBrEe03NWXlO4jyluMbbfm83Z8Pb6cdzuWckXfuRILZI/b0JHsc1fhCij1WvhH\nX1J4FyWfxUCFIhXGre5V5y1uY/EOZmSqNj2uiWROGxGqPJC1Q4uLfEPp7oiry0 ov3lQq9GO\nr8eX825Hz7kjWWK 7HEbOkoVykdmdhfW y1NpMK4VXX8s4WP6dcK8hIsftQSSZ02crZ4IOvO\nqrwgu2arD7qVbjOo1WjH1 PLSftUGhzaUcGhHQ3fDu0K41bVI7JMYofuwQnlDcbSiZ2aSOa0\nIQgPZD2gHbJ1KN0HlT7VKqXbDGo12vH1 HJPTxoi0YFINHwTqTBudUl3pm3WYP7qdfOaofSm\nBaOsafOIOc1zIY1A1sy89cOnVindE/M3Zd5sVkXhQa28nbGeudwbEIkKRKLhm0iFcavtNyQa\n9xkfA529bcH8tacY25poMKe5IDyQ9ejyxNJO5SR7bemcVfuVN835vB1fjy/3BkSiApFo PeB\nrCNu1f3fqMyqL9wRZdT1choQiQpEouHjLUL2uFXXIuUW3j2SuqXF9XIiEIkKRKIh6O5v9 9I\nJ2 cM//2pPvlNCASFYhEA1 jcCJ73IYOiEQDIjmRPW5DB0SiAZGcyB63oQMi0YBITmSP29AB\nkWhAJFEgIJIIRIowEIkIRCICkRzIPpAKHxCJBkRyIHvYhg IRAMiOZA9bMMHRKIBkRzIHrbh\nAyLRgEgOZA/b8AGRaEAkB7KHbfiASDSCE8mWw8pvabXf2lqYXycmz44DkahAJBqBiWTPYa1Q\npB2KOj1815Xz1vZaj70BkahAJBpBieTIYa1MpP66hDp966oTp7 7ZNJ87A2IRAUi0QhIJGcO\na1b5 arE4j26SN03X1G7 oyVqFr4mLE7Nte9xJJVXeq7UvV /tgjEIkKRKIRkEiFOayLDo 2\nVfdqIi1anx69o5mZiaqFj9nub6ZVcXYntDiupU/wx5zJCVekB4vPlz1sw0dy3F2F6WRSonuY\nGE6mRXdxvq/EAutGMT9EKsxhfVx9U6hp10QaHlfdqM7xRNWpj tfZao4z31D28yNP CPOaNJ\nX5E9bMOHv/W9ALHeqfwQqTCH9UV13sKHNZEOtNbVzVWyPFF1yuO772a6SA2GSMZjTmbYFYOp\n4vNlD9vwkRx0V2E6Q32iexjuF/8shs VmG/l9PghUmEO68vqvEWPqSKdmb0lw/Zo4uRjtwof\nv1o/pIuzJ39ot4U/9gjOkajgHIlGQOdIhTmsP2NsYs52VaSOuHoM ZAlTuHjdYl58 ZVXX57\nX9Uxxgbjh/hjj/sDkahAJBoBiVSYw9rUPbElMaSK9Lpy6PyuFuUsF6fw8ZB2mLlg2yC7Y/mJ\nU2uuz5mPvQGRqEAkGgGJ5MxhzSg7mxNNv9Yvf//oinkbhlfM7eWJqoWPNbRDudH19fNvN3YW\nh3bBAZFoBPWBrCOHVT4QiQpEohHYLUL2HFb5QCQqEIkG7v52IHvYhg IRAMiOZA9bMMHRKIB\nkRzIHrbhAyLRgEgOZA/b8AGRaEAkUSDXjghEijAQiQhEIgKRRAGRiECkCFNUJNmnI6EEItGA\nSBayx2wogUg0IJKF7DEbSiASDYhkIXvMhhKIRAMiWcges6EEItGQK1KR0MhSZJX9PnYMkahA\nJBpSRSoWGmly4JhjOndw2MeeIRIViERDpkhFQyNN1rYXn/YFiEQFItGQKFLx0EiWWldf03Kc\ntVbNXs7TIrVp9dAup xsbWzq2nxd/ZNe 4ZIVCASDYkiFQ NZCvXDWUeWZBhjeq7EE LVKe1\nc6T4qvRky9zdrDM 4LFviEQFItGQKFLx0MjjWmxXbm6HLhJPi QiqbN 3MBYWnmdb T8mCuG\n 4vMlD1mQ0lyxF2F6Yz2ie5hbDA5LLqLks8iLVyk4qGRHYrOFl0knhbJRdrLWNtK7cLEQb4R\nP5NWZY/ZUOJjfS9M/E1aLUbx0Mg9Sia/WJXHTIvkIr08RaTsuCtG 4vMlD1mQ0ly1F2F6aT7\nRPcwPiTxWWSEi1Q8NPJN/bCtRxfJTIssLZJLcI5EBedINCSeIxUPjWQ3Np/Nttf0sabNI2Za\npDoNkeQAkWhIFKl4aCRLfaf28uZDjG1NNJhpkeo0RJIDRKIh8wNZiaGREIkKRKIh9RYheaGR\nEIkKRKKBu78tZI/ZUAKRaEAkC9ljNpRAJBoQyUL2mA0lEIkGRLKQPWZDCUSiAZEsZI/ZUAKR\naEAkUSDXjghEijAQiQhEIgKRRAGRiECkCDNVJNnnImEFItGASAayB2xYgUg0IJKB7AEbViAS\nDYhkIHvAhhWIRAMiGcgesGEFItEIhUilA1ezyn4 R/3tX9oqRKICkWiEQaQygau5g8M2kfxL\nW4VIVCASjRCIVD5wldlE8q9PiEQFItGQL1Jh4Gony960JmtkrOqHdvkI1vyhXU7Ztbrpqp1e\nO4VIVCASDfkiFQaudrL1zeM8Y1UXKR/BapwjxVcMsG016em3WxaIRAUi0ZAvUmHgauejTUNm\nxqouUj6ClYu0lbFexfx2enrAFf19hXNkD9iwkux3V EKXotzonsYSIl/FgPJEvOt/7IFi1QY\nuHqn8hozM1Z1kfIRrFyk3YylFPNPvviWtCp7wIYVv p7wSI adWgMHC1cfWKrJmxqouUj2Dl\nIr3kEMklOLSjgkM7GvIP7QoDV/cNNzxoZqzqIuUjWCGSFCASDfkiFQaudrJD1a/wjFVdpHwE\nK0SSAkSiIV8kZ Cq/mnRI3UDRsbqGWU/j2CFSFKASDTkiyQlcBUiUYFINEIgkozAVYhEBSLR\nCINIEoBIVCASDYhkIHvAhhWIRAMiGcgesGEFItGASAayB2xYgUg0IJKB7AEbViASDYgkCuTa\nEYFIEQYiEYFIRCCShuzjpxADkWhAJA3ZozXEQCQaEElD9mgNMRCJBkTSkD1aQwxEogGRNGSP\n1hADkWhAJA3ZozXEQCQawYpki1QtSq/SbU5rXzwSBkSiApFoBCqSPVK1KBApdEAkGkGK5IhU\nLQpECh0QiUaAIjkjVTPKcy2NV 1hrH9dfaL5MGNdK2uW7lJFSq2rr2k5roq0bfXsuhcYMzJX\nmzeqm3i9qpcv9gZEogKRaAQoUmGk6rJ tr16gK28ZWB8c 1grnF9 myrKtLKdUOZRxZkWPza\nw2MP16R55ur2ueq72AOt5mK / 7IDDkeyh6tISZ53mWJyWT7RfeQG01mRHcxmSq1xHeRCiNV\nn1L/M6rd1qWox3uZOS8cUXpU15Tu41rIXW5uB4urbXtUsYzM1fTlu1juGy Yiw38CYiUPVpD\njC/1vZDxPyCyMFJ1rzpvcduLVZqyTY93VE0y9obS3aHobDEjVY3MVXbvzezQ5ePmYgN/Iotl\nj9YQg8hiIsFFFhdGqqqisIWP5UVa/OhO7fcJpXuPYhy2GblbPHOVHa3q2/R9Zi72Bs6RqOAc\niUaA50iFkarqoVsmseOEFoifTuw8oLzFWIfS/aaiXRvvMUXimauqiE8uOMzMxd6ASFQgEo0A\nRSqMVF3SnWmbNcia1wylNy0YzcxbP3xqlXpOdGPz2Wx7TR8XiWeuMra19hp1K3yxNyASFYhE\nI0CRCiNV229INO5j7OxtC avPcXY0eWJpZ3KSZb6Tu3lzYfMdyRmZK72suFZ2pkRX wNiEQF\nItEI8gNZR6Rq5X/JsjvhX8UhEhWIRCPQW4TskaqVijTZe/0PfdsRiEQGItGQdvd3pSK1zVnv\ny/W6PBCJCkSiga9RaMgerSEGItGASBqyR2uIgUg0IJKG7NEaYiASDYgkCuTaEYFIEQYiEYFI\nRCCSKCASEYgUYewiyT4LCTcQiQZEgkhlgUg0IBJEKgtEogGRIFJZIBINiASRygKRaEAkiFQW\niEQjUJGmC1rVsEfbaWSV/b71bwGRqEAkGkGKNG3QqkahSLmDw371bwMiUYFINAIUafqgVY1C\nkcQAkahAJBrBieQMWs0pO1sbm7o2X1f/pBmmysNWeQjrpPJ8493qoZ2jrf9Jq7KHariBSDSC\nE8kZtMriq9KTLXN3s874AA9T5WGrZghrfMXxMe0cyd62IGk1O 6K0X5rWvZQDTfJUXcVppPu\nE93D JDEZ2EdffkkkjNolcXbGftxA2Np5XUepsrDVnkIK4s/kb/YYGsrIGlV9lANN97re4Hj\ne9KqM2iVxfcy1rZSE UgD1PlYas8hJXFOwyRrLaFSavnx1wx3G9Nyx6q4SY54q7CdEb7RPcw\nNpgcFt1FyWeR9lskZ9Aqi79sysHDVHnYKg9h1RO5dJGstgKSVmUP1XCDcyQawZ0jOYNW7XLw\nMFUetspDWIuJJCBpVfZQDTcQiUZwIjmDVu1y8DBVHrZqhrAWEUlA0qrsoRpuIBKN4ERyBq06\n5OBhqkbYqhnCWkwk/5NWZQ/VcAORaAT4gawjaLUclYewVgxEogKRaAR5i5A9aLUcEClEQCQa\nYbz7GyKFCIhEI4wiBQBEogKRaEAkiFQWiEQDIkGkskAkGhBJFMi1IwKRIgxEIgKRiEAkUUAk\nIhApwpwfkn3uERUgEg2IBMoCkWhAJFAWiEQDIoGyQCQaEAmUBSLRkCcSJS1SGBCJCkSiIU0k\nUlpkCQ4c89o7RKICN3wj1QAAFS5JREFUkWjIEomWFlmCte1eu4dIVCASDUkiOdMiWf 6 kTz\nYWvxiWsTyw8qJ9PKQcbOKGfMVMgdixN1GzOtVbOXe wfIlGBSDQkiVSQFrnyloHxzbXm5/S5\nhrtGT65Quk2RjFTInqr9kz3LtrBGvCMFBkSiIUkkZ1pkl6Ie52XmvMCXHtHihLZbIvFUyONK\nl1p3ZhcpM yKwZTsARoVkoPuKkxnqE90D8P94p/F8LkS80eFiuRMi3xRC7RjTY/zpR3xfMAd\nF4mnQubuq25uO 0QyXXSquwBGhX8DB29IPE9adWBMy0yL9LiR/nSXbPVH92GSKeUM1YqZO z\na6o77CLlsq4YH5Q9QKNCMuOuwnTOp0T3kB1JjovuYqKvxIJJoSI50yJPKG8wlk7s5EsPKn2M\n7VS6s1WdjHUqZ3gqZHZA/XV/C86RAgTnSDQknSMVpEU2rxlKb1pgHk1OzN UebNZ6WYLH2Tj\n31LO8FTI7Q3HcqnWDaxp84jH/iESFYhEQ5JIBWmRZ29bMH/tKWvxa0vnrNqvvMlevebqG36t\n/JanQubaGmbV3zPCtiYaPPYPkahAJBqyPpAtnxaZndAu3Y0WW QPEIkKRKIh7RahcmmRuYV3\nj6RuaRHSbx6IRAUi0QjT3d/Hag32nrxxzvzbkwK7gkhUIBKNMIkUIBCJCkSiAZFAWSASDYgE\nygKRaEAkUBaIROOCFUl4F8i1IwKRIgxEIgKRiFywIsk ZIoKEIkGRAJlgUg0IBIoC0SiAZFA\nWSASDYgEygKRaEAkUBaIREOGSLaM1ayyn8/NKp22R7a5fnevAZGoQCQaEkSyZ6zmDg7z2aoy\ntke2ufmJntsXzLlzgLG 7y64vOWo132ASFQgEo3gRSqVsVr8vYfPPX/Nrae6b/qfjK1Y1XXm\ne/PTHncCIlGBSDQCF8mZsaoezPHMLePQjoeq8ubKz1clFu9hRxX1SDCpdA/drmp4VvkPj3sB\nkahAJBqBi TMWJ0qkhGqyptnlUWHR9uqew8pQ qD6h36zCPx6fe6PBCJCkSiEbhIzozVKSLx\nUFXePKs8rg77mvax fdPTPyk mfavKGmH5mbGz3nDgREEkm6LHAlr8WM7sL6b8JfkZwZq1NE\n4qGqlkgvqj8XPsxeW1hd 5OFW9UHv124MWdubizlir5zsgdoVEj2uatwBZwT3kNfUngXJZ/F\ngCCRnBmrXKRTXCQrVJWL9LL6c9Fj6o RiYlZLzG2f97T3vcCh3ZUcGhHI/BDO2fGqqoOj1PN\ni8RDVTlZRT2am5izPduh7ujL1YPsN1f82oe9gEhUIBKNwEVyZqxqlxeMOFXjYoMRqsqbZ5Wm\n7oktiSG27PbkoSs3sszVj2nJ5Lj8HRQQiUbgIjkzVjV1jDhVfvk7H6rKySg7mxNN6pvQ6daa\nBQ9MsP35c6hnPO4FRKICkWgE/4GsI2O1yE1BgQCRqEAkGhJuEbJlrE4e1f50mAQgEhWIREPu\n3d 7qm/NFV1gha6K6RgiUYFINPA1ClAWiEQDIoGyQCQaEAmUBSLRgEigLBCJxgUrkvAuEBBJ\nBCJFGIhEBCIRuTBEkn18FGEgEg2IBMoCkWhAJFAWiEQDIoGyQCQaEAmUBSLRgEigLBCJhjyR\nbHmrBVjfreBTvUq3p75kj8YIA5FoSBPJnrdagBW4yqcgkjQgEg1ZIpXKWy0ORJIGRKIhSSRn\n3irbsThRtzFjBXPtZ8lvz6nbOK5Nda2sWbpLFYlHsOaXVNad7NEYYSASDUkiOfNWe6r2T/Ys\n22IX6fo7U6eX3K9O5RrXp8 2qiLxCNb8Er6hyQkKskdjhEmOkyrsgUxKdA8Tw8m06C7O95VY\nYN0oJkIkZ97qce0L55PMJlKX0sNY1z516og29ZLSzSNYjSV8Q6NJCrJHY4Qh1ReUxnqnEiGS\nM281d191c9tpu0gvVulfQdeiV6smGXtD6eYRrMYSzvkRCrJHY4RJDpEq7IHhPtE9jAyIfxYj\n50rMHxMqkjNvlbHeZ9dUd1iZq/t/pemji7RTE eE0s0jWI0llSF7NEYYnCPRkHSOVJC3quUj\n399iZa7uP6Fdpjv6jDp1QHmLsQ6lm0ewGksq6072aIwwEImGJJGceavbG47lUq0brMzV/Wzl\n6rdOLduoTmXmrR8 tUq1h0ew5pdU1p3s0RhhIBINSSI581ZzbQ2z6u8ZsTJX97OB22oW3JvW\npo4uTyztVE6aEaz5JZV1J3s0RhiIREPWB7KOvFXhyB6NEQYi0ZB2i5Atb1U8skdjhIFINHD3\nNygLRKIBkUBZIBINiATKApFoQCRQFohE48IQaQrItSOCXDsiEEkUEIkIRIowEIkIRCIyw0WS\nfYYRfSASDYgEygKRaEAkUBaIRAMigbJAJBoQCZQFItGQIlLpbMipWGmRbpA9DKMPRKIhQ6Qy\n2ZBTsdIi3SB7GEYfiERDgkiVZUN6Q/YwjD4QiUbwIhVkQxq5j5PK8413s 6br6hdfYbllJ2t\njU1dm6 rf1I7tMsvMxr ux4mtJnYmexhGH0gEo3gRXJmQ5q5j/EVx8fYovXp0TuaGYuvSk 2\nzN3NOuMD2jmSvsxomMtmsy/N6SJ2JnsYRh IRCN4kQqzIfO5jyz hDpveJyx3dU5Fm9n7McN\njKWV13WR1GVmQ8ZO1/67ubVpcu1kD8Pog1w7IoHn2jmzIXnuI4trihxoraubq2RZfC9jbSu1\nS3YHdZE6mNWQpZc8YG1tmqRV2cMw gQTRzqDEZa06syG5LmPLP4SY2dmb8moc1SRXnaKpC4z\nG7I7/6ftbtDs bLIHobRJ5kuX2HvZFKiezg/nBwT3UXpZyFKJGc2JM991GXpiKuKPFRcJLPh\nTxsG6J3JHobRB dINII/R3JmQ5q5j5osryuHzu9qUc4WE4k3fDVxJJvNUmOLZQ/D6AORaAQv\nkjMb0sx91GX50RXzNgyvmNtbTCSj4Tr9VOk6Ymeyh2H0gUg0JHwgG2Q2pOxhGH0gEg0ZtwgF\nmA0pexhGH4hEA3d/g7JAJBoQCZQFItGASKAsEIkGRAJlgUg0IBIoC0SiMcNFKgVy7Ygg144I\nRBIFRCICkSIMRCICkYjMcJFkn2FEH4hEAyKBskAkGhAJlAUi0YBIoCwQiQZEAmWBSDSkimRL\nXC0fqGpbqn83iYzsYRh9IBINmSLZE1fLBKoeOGZfCpGCBSLRkCgSNXF1bbv9EUQKFohEQ55I\nzsRV9eAtozzX0njVHqvFjsWJuo2Z1qrZy7VDu S359RtHGfxbatn171A7UT2MIw EImGPJGc\niauqKlllWT/bXm1mBPVU7Z/sWbaFNbbr50jX35k6veR Fr/28NjDNWliJ7KHYfSBSDTkieRM\nXNVFekqtau023uC40qU ZoZIXUoPY137WFxdp0fp5o3S/WWRPQyjTzJVvsI cE54D30Sn4V1\nG5cgkZyJq7pIe9XZi9t4g9x91c1tp7lIL1bl9Lnx3YyllGO8EZJWBRNMHOkMRljSKseZuKqL\npDrCFj5mNel9dk11hyHSr6ryWXbaxQabSNMgexhGHxza0ZB3aOdMXNVFUo/aMokdvEFWO1u6\nv8UQ6YR2OHf0GYgUMBCJhjyRnImrukhLujNts8xjyu0Nx3Kp1g2safOIdrFh5eq3Ti3bCJEC\nBiLRkCeSM3FVF6n9hkTjPmvX2hpm1d8zwrYmGjSRBm6rWXBvGiIFDESiIfEDWUfiqi5Sp 99\nyB6G0Qci0ZB5i5AtcXXyqNIFkcIIRKIRkru/d1XfmjNEOlZrsNeH7coehtEHItEIiUiikD0M\now9EogGRQFkgEg2IBMoCkWhAJFAWiERjhotUCuTaEUGuHRGIJAqIRAQiRRiIRAQiEblARcqO\nC 9ifCQnuotJ6vcb3ZMZof4BedfkxD L8yPC/1NjY9O2mIkiARA4EAkAH4BIAPgARALAByAS\nAD4AkQDwAYgEgA9AJAB8ACK5ZodSUVJ56Bi 68p5a3tl74UPhOKFgEhu6a9LhOD188Ctq06c\n/u4S4fc2CCccLwREcssdm tC8Pq5J1nVpb4rVZf7y1XRIBwvBERyye5vpsPw rlnd0K7XXDp\nE7L3wysheSEgkjuG619lYXj93PPcN7SfN/5A8m54JSwvBESqlF/G4/HD7O67WSheP/c816D9\njLxIYXkhIFKljHZ3d4 /Wj8UjtfPPXvyh3ZbZO HN0LzQkAkV6xLzJs3r ry22Xvhwf6qo4x\nNhg/JHs/vBGaFwIiuWJI  M4C7aJ/xKrQO5YfuLUmuuFf0FRLKF5ISCSe8JwROGB0fX182 f\n/jvUESAMLwREAsAHIBIAPgCRAPABiASAD0AkAHwAIgHgAxAJAB ASAD4AEQCwAcgUqX87ccr\naFx7ydSpsvxVJVsv4NxHrpq 0Sv/ L73f FB7cagzn98/7v/6ll95pHPx36pT z46gfe/fkf\nMXbjB07a1rn9dytL19Z3hNYREef hBGIVCnCRHpVey3WV3j75avWCzj59c9Mn/W 56KP3XX/\n38fuYuzoez x7r4vv/3n6sxN77ksP76ffvvnv3/fF9Wl2b//gu0vEXw5UdE 6TtC7IiIc3/C\nCESqFGEibXDzWthWeiT2i nb//cP9DI28YkP5Ni8977F2PlP/hlju9 98YH8 P7Mn6gqjv X\njzF2 O3fM9cZuuiBivZJ3xFqR0Ts xNKIFKl6CK1f m97/6zu9TjlslbLnvX57ctvYgvNRfk\n1l72rk9t0fSxps5884/f9fuzj/ANfenpy77I2GN/cfH7vvAYY/8Yi8W kD 0Mzei8/m/3vkX\nF/9ewwCz2ppbyq kk/3TLzsaG5v/xT 87 LPbTb3/odt2s G2Ej2knna1J2xI zYAZYf39m7\nn9bmXfEOtevLPzTC1/nX2Jv5iS/93Stffd8H5/badvHDs9QHD8e yTSnj9t2hNwRccu2/Qkl\nEKlSNJF 9rZ/ rcd18eaGft27PLnf/gHf8nfbqwF34nN3/7Epz5 iX3qr//zD1/4yac/NJpv\n 9XPfOK Z9jjsVnPPPNPsWfYf8Rj w7rIlkb0fniB//8V8lHLlLHldnW3FJ JZ1dsQcdjfOb\n3/GOLz 9bVGs4L/zL32EHY2t06baY5qXxvjOk/vzT6o/n42ZaQ5X/zdj4mt/9Bfbe3/6jitt\nu3jlpaoLDZf V3Xp7I/Zd4TcEXHLtv0JJRCpUjSRPvHHGXWq qJzud//lPpy74lxkawFH/6U\nOnHmokuYNTUYa1Gnjt9 Ot/2a7F/VX/e/lV1jcH/Yz5jjdpr8Vf2rRsdxjqYtvRNq621pUbz\nBbw5dsrROL/5z/2Jpm3V xx/76st9n32y5h qv9K7G7tlzm x9/YXfsu7e1i9J2NvPkfX8/4\nHr o/fywbRd/EvsNYx9dFTvNcpc22XeE3BFxy7b9CSUQqVJUkU7HFmlTm2PPnImt0KY ZYhk\nLXgjdp029cVLmDV1/j99ZIctRu5r77T ZONlX7JEsjZidHiJdgD0UKzdamttyRLp//l9Z2N9\n872xZWmV 2MvZ06q5C9GPP2eRI5ti/1Em/5N7A7tlzm fx6LffR5feoznzU2/JvYNr7H79F \nXvl22y72vm0TOxk7fGkbOxh7yr4jxI6M3SJs2dyfcAKRKkUV6eXYrdpUe wHr aPWxKGSNaC\nvfkpdYE1xV78L7H/lPgJ/7ut2v /6hvMzZ96/zveEftbSyRrI0aH/6f28xn1P3arrbklS6S/\n/DNnY33zr8YM/lWf0q6csQ3vuEJV7Jf5A7DO2D3aL3N8n936YOLtt2lTX73M2PD3LuZvZ1/7\niPZT7dK2i5 dxx78EKteyDZcNGTfEWJHxm4RtmzuTziBSJWiirQvtlabejb2wz35s48aQyT7\nAn1AVF/CrCn1JHvnP/ 32J8bF6nzg fL72jtOPjah20iWRsxOtTd LfYQ7a25pYskT7xN87G\n uZfjV31kk5y4Gcq2sn88tiN2pvWsZh of3p2E 1X/ZTF/bPbzuo/pz9XuPhP/wPvsAc7rZd\nbL6M1c1h/ vjbPZXHDtC7MjYLcKWzf0JJxCpUlSRemILtakfxJ77j/wlgU8bIlkLumJLtKnP\nXsKsqTwbYz/OT iD51jsavXnxLttIlkbMTq8WPs49F9iz9namlua8o7EG c33xe7smD3W96e\n93Py/fqHQ6tj3dqv/Ph 696XtQc/jT2q/vzqH ZXGH3XPXxdc7jbdnF77MRl96nj/8yldzh2\nhN4Rcct8f0IKRKoU7WLDpz6sHe3803sGJ35Xu5LwsnmxwVwweenH1JOYo2 7hFlTv67V3hGO\nx76bb6sPnsP6/8AbYn/N2DdjE/mLDeZGjA71s6Pqd6WsttaW9JV0jHMk3tgYm3/5u/3qz4du\n5K1 HrvXmGq8 LeqJR/Jv33kx3fy7X vvYdcoz/g5yRPx17nT90c7rZdHL/4xtghln3fLbFX\nHDtC74i4ZZwjzTA0kZ59 9ef vli7fT5 tg3nv Xj/4tF8lacHNs9pObPvKFS2xTPe/7zObt\nj//N /Mft QHz/k/ sOnXlz5la 874WR1bG1P9VFsjaS7/CP/nTT9n O1dvaHje3pK kc2P \nqh1vbIzNX1z0mYeev mibxjbmvjTSx/QOc3e MCf3HHPX75zt3oW88ADV8ZWPfDAm6wl9nfr\nN14e  857Y2oIb/Kko aT90a7rZd/PqlH9R fvBD/DMhfUfoHRG3bO5PSIFIlaJ/ILvt7y55\n1 e0k jxay 95Et755nH7 aCbMt/fuenf7b0nfapA7M dNGHZ/H/ufODZ98X3/P71ww fenv\nHf3t5y76eP4DWXMj Q4/8esvv f3rh62tzW3pK k8 /6ZWarcX7z7Jf/9/su tN1/A0pyS8 \nbFffDJX3v cr2ltCozHvaZbb LmL3/vp27Ur5sbnPox9bJH51K3hbtvF78Vq1J 3xRbwVvqO\n0Dsibtncn5ACkfzga38gcOPEe5ImPvYVeuPpmXvpsMs18zviN 73JxggkjfWz1b/s //3f8x\nfUvXUN14SPss1i Rjrx9net19R3xGS/7EwgQyRsPx/7fp5744tt2COyC6sbkP3w27ZdIk1/7\nXHr6VuV2xF887U8gQCSPPPy5S97zN8 K7IHsRvIjV/kl0s2/d8LD2knKF6Mqwtv BAFEAsAH\nIBIAPgCRAPABiASAD0AkAHwAIgHgAxAJAB ASAD4AEQCwAcgEgA 8P8Dlat5SI9OxSEAAAAA\nSUVORK5CYII=",
      "text/plain": [
       "plot without title"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "tidy_tweets %>%\n",
    "  filter(str_detect(word, \"^@\")) %>%\n",
    "  filter(!str_detect(word, \"^@ny\")) %>%\n",
    "  count(word, date_group) %>%\n",
    "  filter(sum(n) >= 10) %>%\n",
    "  ungroup() %>%\n",
    "  spread(date_group, n, fill = 0) %>%\n",
    "  mutate_if(is.numeric, funs((.   1) / sum(.   1))) %>%\n",
    "  mutate(logratio = log(past / today)) %>%\n",
    "  arrange(desc(logratio)) %>%\n",
    "  group_by(logratio < 0) %>%\n",
    "  top_n(15, abs(logratio)) %>%\n",
    "  ungroup() %>%\n",
    "  mutate(word = reorder(word, logratio)) %>%\n",
    "  ggplot(aes(word, logratio, fill = logratio < 0))  \n",
    "  geom_col(show.legend = FALSE)  \n",
    "  coord_flip()  \n",
    "  ylab(\"log odds ratio past (before cutoff) / now (after cutoff)\")  \n",
    "  scale_fill_discrete(name = \"\", labels = c(\"past\", \"now\"))   theme_minimal()\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "In my data, I notice that the people from my past are largely other Pirate Party members. People from the present heavily feature the Open* & Quantified Self crowd at large (e.g. `@o_guest, @sujaik, @protohedgehog, @kevinschawinski, @kaiblin, @eramirez`), including some of the awesome people that run/ran `Open Humans` with me to make this possible (👋 `beaugunderson, @madprime, @betatim`).\n",
    "\n",
    "## Emoji pairings\n",
    "In a next step we can now look into highly-frequent pairings of emoji with individual words. For this we filter out the emoji out of all tweet-texts and associate them with words, ignoring stop-words & URLs etc."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "emoji_tweets <- tweets %>%\n",
    "  filter(!str_detect(text, \"^RT\")) %>%\n",
    "  filter(!str_detect(text, \"^@\")) %>%\n",
    "  filter(str_detect(text, \"[\\\\uD83C-\\\\uDBFF\\\\uDC00-\\\\uDFFF] \")) %>%\n",
    "  mutate(Emoji = str_extract_all(text, \n",
    "                  \"[\\\\uD83C-\\\\uDBFF\\\\uDC00-\\\\uDFFF] \")) %>%\n",
    "  select(tweet_id, timestamp, Emoji,text) "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Joining, by = c(\"tweet_id\", \"timestamp\")\n"
     ]
    }
   ],
   "source": [
    "emoji_tweets <- emoji_tweets %>%\n",
    "  select(-Emoji) %>%\n",
    "  unnest_tokens(word, text) %>%\n",
    "  left_join(emoji_tweets) %>%\n",
    "  mutate(Emoji = map_chr(Emoji, ~ ifelse(length(.x) > 0, .x[[1]], \"\"))) %>%\n",
    "  mutate(word = str_replace_all(word, \"’\", \"'\")) %>%\n",
    "  filter(!(Emoji %in% c(\"\", \"-\"))) %>%\n",
    "  filter(!(word %in% c(\"t.co\", \"http\",'https'))) %>%\n",
    "  filter(!word %in% stop_words$word,\n",
    "         str_detect(word, \"[a-z]\"))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<table>\n",
       "<thead><tr><th scope=col>item1</th><th scope=col>item2</th><th scope=col>n</th></tr></thead>\n",
       "<tbody>\n",
       "\t<tr><td>️       </td><td>fra    </td><td>31     </td></tr>\n",
       "\t<tr><td>🎉      </td><td>opensnp</td><td> 9     </td></tr>\n",
       "\t<tr><td>️       </td><td>lhr    </td><td> 8     </td></tr>\n",
       "\t<tr><td>😂      </td><td>data   </td><td> 7     </td></tr>\n",
       "\t<tr><td>😂      </td><td>opencon</td><td> 6     </td></tr>\n",
       "\t<tr><td>🎉      </td><td>humans </td><td> 4     </td></tr>\n",
       "\t<tr><td>😍      </td><td>mozfest</td><td> 4     </td></tr>\n",
       "\t<tr><td>🍻      </td><td>beer   </td><td> 3     </td></tr>\n",
       "\t<tr><td>🐶      </td><td>dog    </td><td> 3     </td></tr>\n",
       "\t<tr><td>💖      </td><td>hear   </td><td> 3     </td></tr>\n",
       "</tbody>\n",
       "</table>\n"
      ],
      "text/latex": [
       "\\begin{tabular}{r|lll}\n",
       " item1 & item2 & n\\\\\n",
       "\\hline\n",
       "\t ️        & fra     & 31     \\\\\n",
       "\t 🎉       & opensnp &  9     \\\\\n",
       "\t ️        & lhr     &  8     \\\\\n",
       "\t 😂       & data    &  7     \\\\\n",
       "\t 😂       & opencon &  6     \\\\\n",
       "\t 🎉       & humans  &  4     \\\\\n",
       "\t 😍       & mozfest &  4     \\\\\n",
       "\t 🍻       & beer    &  3     \\\\\n",
       "\t 🐶       & dog     &  3     \\\\\n",
       "\t 💖       & hear    &  3     \\\\\n",
       "\\end{tabular}\n"
      ],
      "text/markdown": [
       "\n",
       "item1 | item2 | n | \n",
       "|---|---|---|---|---|---|---|---|---|---|\n",
       "| ️        | fra     | 31      | \n",
       "| 🎉       | opensnp |  9      | \n",
       "| ️        | lhr     |  8      | \n",
       "| 😂       | data    |  7      | \n",
       "| 😂       | opencon |  6      | \n",
       "| 🎉       | humans  |  4      | \n",
       "| 😍       | mozfest |  4      | \n",
       "| 🍻       | beer    |  3      | \n",
       "| 🐶       | dog     |  3      | \n",
       "| 💖       | hear    |  3      | \n",
       "\n",
       "\n"
      ],
      "text/plain": [
       "   item1 item2   n \n",
       "1  ️      fra     31\n",
       "2  🎉     opensnp  9\n",
       "3  ️      lhr      8\n",
       "4  😂     data     7\n",
       "5  😂     opencon  6\n",
       "6  🎉     humans   4\n",
       "7  😍     mozfest  4\n",
       "8  🍻     beer     3\n",
       "9  🐶     dog      3\n",
       "10 💖     hear     3"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "emo_twids <- emoji_tweets %>%\n",
    "  select(tweet_id, Emoji) %>%\n",
    "  distinct() %>%\n",
    "  rename(word = Emoji)\n",
    "  \n",
    "emoji_tweets %>%\n",
    "  select(tweet_id, word) %>%\n",
    "  bind_rows(emo_twids) %>%\n",
    "  pairwise_count(word, tweet_id, sort = TRUE) %>%\n",
    "  filter(item1 %in% unique(emoji_tweets$Emoji)) %>%\n",
    "  group_by(item1) %>%\n",
    "  slice(1:2) %>%\n",
    "  ungroup() %>%\n",
    "  filter(nchar(item2) > 2) %>%\n",
    "  arrange(desc(n)) %>% head(n=10)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Looking at the top 10 emoji in my data shows that for me, #mozfest is more 😍, while #opencon is more 😂. And there's two words, `fra` and `lhr` where the emoji can't be rendered by `R`. This would be ✈️, which associates with the IATA codes for the Frankfurt airport and London Heathrow. \n Check out your own emoji trends for clues into how your emoji use changes with context."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "R",
   "language": "R",
   "name": "ir"
  },
  "language_info": {
   "codemirror_mode": "r",
   "file_extension": ".r",
   "mimetype": "text/x-r-source",
   "name": "R",
   "pygments_lexer": "r",
   "version": "3.4.2"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}