Delta-Intro.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "## pyΔ.next"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "['/home/tv/git/pydelta', '/usr/lib/python39.zip', '/usr/lib/python3.9', '/usr/lib/python3.9/lib-dynload', '', '/home/tv/.virtualenvs/pydelta/lib/python3.9/site-packages', '/home/tv/git/pydelta', '/home/tv/.virtualenvs/pydelta/lib/python3.9/site-packages/IPython/extensions', '/home/tv/.ipython']\n"
     ]
    }
   ],
   "source": [
    "import sys; print(sys.path)\n",
    "%matplotlib inline\n",
    "import pandas as pd\n",
    "import numpy as np\n",
    "try:\n",
    "    import seaborn     # sieht halt besser aus\n",
    "except ImportError:\n",
    "    pass               # dann halt nicht\n",
    "import delta"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "## 1. Corpus vorbereiten"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "slideshow": {
     "slide_type": "-"
    }
   },
   "outputs": [],
   "source": [
    "whole_corpus = delta.Corpus(\"test/corpus3/\")\n",
    "corpus = whole_corpus.top_n(2500)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "### 2. Corpus → Distanzmatrix"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {
    "slideshow": {
     "slide_type": "-"
    }
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Raabe,-Wilhelm_Im-alten-Eisen</th>\n",
       "      <th>Fontane,-Theodor_Der-Stechlin</th>\n",
       "      <th>Raabe,-Wilhelm_Stopfkuchen</th>\n",
       "      <th>Marlitt,-Eugenie_Das-Geheimnis-der-alten-Mamsell</th>\n",
       "      <th>Raabe,-Wilhelm_Die-Chronik-der-Sperlingsgasse</th>\n",
       "      <th>Marlitt,-Eugenie_Die-Frau-mit-den-Karfunkelsteinen</th>\n",
       "      <th>Fontane,-Theodor_Frau-Jenny-Treibel</th>\n",
       "      <th>Fontane,-Theodor_Effi-Briest</th>\n",
       "      <th>Marlitt,-Eugenie_Das-Heideprinzesschen</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>Raabe,-Wilhelm_Im-alten-Eisen</th>\n",
       "      <td>0.000000</td>\n",
       "      <td>1.343237</td>\n",
       "      <td>0.624532</td>\n",
       "      <td>1.200938</td>\n",
       "      <td>0.732207</td>\n",
       "      <td>1.215720</td>\n",
       "      <td>0.886441</td>\n",
       "      <td>1.258676</td>\n",
       "      <td>1.395609</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Fontane,-Theodor_Der-Stechlin</th>\n",
       "      <td>1.343237</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>1.246155</td>\n",
       "      <td>1.331243</td>\n",
       "      <td>1.402441</td>\n",
       "      <td>1.257329</td>\n",
       "      <td>1.017926</td>\n",
       "      <td>0.640157</td>\n",
       "      <td>1.119296</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Raabe,-Wilhelm_Stopfkuchen</th>\n",
       "      <td>0.624532</td>\n",
       "      <td>1.246155</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>1.268060</td>\n",
       "      <td>0.793546</td>\n",
       "      <td>1.247171</td>\n",
       "      <td>0.909698</td>\n",
       "      <td>1.209208</td>\n",
       "      <td>1.387786</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Marlitt,-Eugenie_Das-Geheimnis-der-alten-Mamsell</th>\n",
       "      <td>1.200938</td>\n",
       "      <td>1.331243</td>\n",
       "      <td>1.268060</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>1.133478</td>\n",
       "      <td>0.801244</td>\n",
       "      <td>1.267795</td>\n",
       "      <td>1.304392</td>\n",
       "      <td>0.725335</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Raabe,-Wilhelm_Die-Chronik-der-Sperlingsgasse</th>\n",
       "      <td>0.732207</td>\n",
       "      <td>1.402441</td>\n",
       "      <td>0.793546</td>\n",
       "      <td>1.133478</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>1.203421</td>\n",
       "      <td>0.837908</td>\n",
       "      <td>1.280661</td>\n",
       "      <td>1.329303</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Marlitt,-Eugenie_Die-Frau-mit-den-Karfunkelsteinen</th>\n",
       "      <td>1.215720</td>\n",
       "      <td>1.257329</td>\n",
       "      <td>1.247171</td>\n",
       "      <td>0.801244</td>\n",
       "      <td>1.203421</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>1.354223</td>\n",
       "      <td>1.260774</td>\n",
       "      <td>0.722380</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Fontane,-Theodor_Frau-Jenny-Treibel</th>\n",
       "      <td>0.886441</td>\n",
       "      <td>1.017926</td>\n",
       "      <td>0.909698</td>\n",
       "      <td>1.267795</td>\n",
       "      <td>0.837908</td>\n",
       "      <td>1.354223</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.897811</td>\n",
       "      <td>1.506644</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Fontane,-Theodor_Effi-Briest</th>\n",
       "      <td>1.258676</td>\n",
       "      <td>0.640157</td>\n",
       "      <td>1.209208</td>\n",
       "      <td>1.304392</td>\n",
       "      <td>1.280661</td>\n",
       "      <td>1.260774</td>\n",
       "      <td>0.897811</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>1.215294</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Marlitt,-Eugenie_Das-Heideprinzesschen</th>\n",
       "      <td>1.395609</td>\n",
       "      <td>1.119296</td>\n",
       "      <td>1.387786</td>\n",
       "      <td>0.725335</td>\n",
       "      <td>1.329303</td>\n",
       "      <td>0.722380</td>\n",
       "      <td>1.506644</td>\n",
       "      <td>1.215294</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                                    Raabe,-Wilhelm_Im-alten-Eisen  \\\n",
       "Raabe,-Wilhelm_Im-alten-Eisen                                            0.000000   \n",
       "Fontane,-Theodor_Der-Stechlin                                            1.343237   \n",
       "Raabe,-Wilhelm_Stopfkuchen                                               0.624532   \n",
       "Marlitt,-Eugenie_Das-Geheimnis-der-alten-Mamsell                         1.200938   \n",
       "Raabe,-Wilhelm_Die-Chronik-der-Sperlingsgasse                            0.732207   \n",
       "Marlitt,-Eugenie_Die-Frau-mit-den-Karfunkelsteinen                       1.215720   \n",
       "Fontane,-Theodor_Frau-Jenny-Treibel                                      0.886441   \n",
       "Fontane,-Theodor_Effi-Briest                                             1.258676   \n",
       "Marlitt,-Eugenie_Das-Heideprinzesschen                                   1.395609   \n",
       "\n",
       "                                                    Fontane,-Theodor_Der-Stechlin  \\\n",
       "Raabe,-Wilhelm_Im-alten-Eisen                                            1.343237   \n",
       "Fontane,-Theodor_Der-Stechlin                                            0.000000   \n",
       "Raabe,-Wilhelm_Stopfkuchen                                               1.246155   \n",
       "Marlitt,-Eugenie_Das-Geheimnis-der-alten-Mamsell                         1.331243   \n",
       "Raabe,-Wilhelm_Die-Chronik-der-Sperlingsgasse                            1.402441   \n",
       "Marlitt,-Eugenie_Die-Frau-mit-den-Karfunkelsteinen                       1.257329   \n",
       "Fontane,-Theodor_Frau-Jenny-Treibel                                      1.017926   \n",
       "Fontane,-Theodor_Effi-Briest                                             0.640157   \n",
       "Marlitt,-Eugenie_Das-Heideprinzesschen                                   1.119296   \n",
       "\n",
       "                                                    Raabe,-Wilhelm_Stopfkuchen  \\\n",
       "Raabe,-Wilhelm_Im-alten-Eisen                                         0.624532   \n",
       "Fontane,-Theodor_Der-Stechlin                                         1.246155   \n",
       "Raabe,-Wilhelm_Stopfkuchen                                            0.000000   \n",
       "Marlitt,-Eugenie_Das-Geheimnis-der-alten-Mamsell                      1.268060   \n",
       "Raabe,-Wilhelm_Die-Chronik-der-Sperlingsgasse                         0.793546   \n",
       "Marlitt,-Eugenie_Die-Frau-mit-den-Karfunkelsteinen                    1.247171   \n",
       "Fontane,-Theodor_Frau-Jenny-Treibel                                   0.909698   \n",
       "Fontane,-Theodor_Effi-Briest                                          1.209208   \n",
       "Marlitt,-Eugenie_Das-Heideprinzesschen                                1.387786   \n",
       "\n",
       "                                                    Marlitt,-Eugenie_Das-Geheimnis-der-alten-Mamsell  \\\n",
       "Raabe,-Wilhelm_Im-alten-Eisen                                                               1.200938   \n",
       "Fontane,-Theodor_Der-Stechlin                                                               1.331243   \n",
       "Raabe,-Wilhelm_Stopfkuchen                                                                  1.268060   \n",
       "Marlitt,-Eugenie_Das-Geheimnis-der-alten-Mamsell                                            0.000000   \n",
       "Raabe,-Wilhelm_Die-Chronik-der-Sperlingsgasse                                               1.133478   \n",
       "Marlitt,-Eugenie_Die-Frau-mit-den-Karfunkelsteinen                                          0.801244   \n",
       "Fontane,-Theodor_Frau-Jenny-Treibel                                                         1.267795   \n",
       "Fontane,-Theodor_Effi-Briest                                                                1.304392   \n",
       "Marlitt,-Eugenie_Das-Heideprinzesschen                                                      0.725335   \n",
       "\n",
       "                                                    Raabe,-Wilhelm_Die-Chronik-der-Sperlingsgasse  \\\n",
       "Raabe,-Wilhelm_Im-alten-Eisen                                                            0.732207   \n",
       "Fontane,-Theodor_Der-Stechlin                                                            1.402441   \n",
       "Raabe,-Wilhelm_Stopfkuchen                                                               0.793546   \n",
       "Marlitt,-Eugenie_Das-Geheimnis-der-alten-Mamsell                                         1.133478   \n",
       "Raabe,-Wilhelm_Die-Chronik-der-Sperlingsgasse                                            0.000000   \n",
       "Marlitt,-Eugenie_Die-Frau-mit-den-Karfunkelsteinen                                       1.203421   \n",
       "Fontane,-Theodor_Frau-Jenny-Treibel                                                      0.837908   \n",
       "Fontane,-Theodor_Effi-Briest                                                             1.280661   \n",
       "Marlitt,-Eugenie_Das-Heideprinzesschen                                                   1.329303   \n",
       "\n",
       "                                                    Marlitt,-Eugenie_Die-Frau-mit-den-Karfunkelsteinen  \\\n",
       "Raabe,-Wilhelm_Im-alten-Eisen                                                                1.215720    \n",
       "Fontane,-Theodor_Der-Stechlin                                                                1.257329    \n",
       "Raabe,-Wilhelm_Stopfkuchen                                                                   1.247171    \n",
       "Marlitt,-Eugenie_Das-Geheimnis-der-alten-Mamsell                                             0.801244    \n",
       "Raabe,-Wilhelm_Die-Chronik-der-Sperlingsgasse                                                1.203421    \n",
       "Marlitt,-Eugenie_Die-Frau-mit-den-Karfunkelsteinen                                           0.000000    \n",
       "Fontane,-Theodor_Frau-Jenny-Treibel                                                          1.354223    \n",
       "Fontane,-Theodor_Effi-Briest                                                                 1.260774    \n",
       "Marlitt,-Eugenie_Das-Heideprinzesschen                                                       0.722380    \n",
       "\n",
       "                                                    Fontane,-Theodor_Frau-Jenny-Treibel  \\\n",
       "Raabe,-Wilhelm_Im-alten-Eisen                                                  0.886441   \n",
       "Fontane,-Theodor_Der-Stechlin                                                  1.017926   \n",
       "Raabe,-Wilhelm_Stopfkuchen                                                     0.909698   \n",
       "Marlitt,-Eugenie_Das-Geheimnis-der-alten-Mamsell                               1.267795   \n",
       "Raabe,-Wilhelm_Die-Chronik-der-Sperlingsgasse                                  0.837908   \n",
       "Marlitt,-Eugenie_Die-Frau-mit-den-Karfunkelsteinen                             1.354223   \n",
       "Fontane,-Theodor_Frau-Jenny-Treibel                                            0.000000   \n",
       "Fontane,-Theodor_Effi-Briest                                                   0.897811   \n",
       "Marlitt,-Eugenie_Das-Heideprinzesschen                                         1.506644   \n",
       "\n",
       "                                                    Fontane,-Theodor_Effi-Briest  \\\n",
       "Raabe,-Wilhelm_Im-alten-Eisen                                           1.258676   \n",
       "Fontane,-Theodor_Der-Stechlin                                           0.640157   \n",
       "Raabe,-Wilhelm_Stopfkuchen                                              1.209208   \n",
       "Marlitt,-Eugenie_Das-Geheimnis-der-alten-Mamsell                        1.304392   \n",
       "Raabe,-Wilhelm_Die-Chronik-der-Sperlingsgasse                           1.280661   \n",
       "Marlitt,-Eugenie_Die-Frau-mit-den-Karfunkelsteinen                      1.260774   \n",
       "Fontane,-Theodor_Frau-Jenny-Treibel                                     0.897811   \n",
       "Fontane,-Theodor_Effi-Briest                                            0.000000   \n",
       "Marlitt,-Eugenie_Das-Heideprinzesschen                                  1.215294   \n",
       "\n",
       "                                                    Marlitt,-Eugenie_Das-Heideprinzesschen  \n",
       "Raabe,-Wilhelm_Im-alten-Eisen                                                     1.395609  \n",
       "Fontane,-Theodor_Der-Stechlin                                                     1.119296  \n",
       "Raabe,-Wilhelm_Stopfkuchen                                                        1.387786  \n",
       "Marlitt,-Eugenie_Das-Geheimnis-der-alten-Mamsell                                  0.725335  \n",
       "Raabe,-Wilhelm_Die-Chronik-der-Sperlingsgasse                                     1.329303  \n",
       "Marlitt,-Eugenie_Die-Frau-mit-den-Karfunkelsteinen                                0.722380  \n",
       "Fontane,-Theodor_Frau-Jenny-Treibel                                               1.506644  \n",
       "Fontane,-Theodor_Effi-Briest                                                      1.215294  \n",
       "Marlitt,-Eugenie_Das-Heideprinzesschen                                            0.000000  "
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "distances = delta.functions.cosine_delta(corpus)\n",
    "distances"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "1.851976045085546"
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "distances.simple_score()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {
    "slideshow": {
     "slide_type": "-"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Metadata(complete=False, corpus='test/corpus3/', delta='cosine_delta', delta_descriptor='cosine-z_score', delta_title='Cosine Delta', features='words', frequencies=False, lower_case=False, normalization=('z_score',), ordered=True, words=2500)"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "distances.metadata"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "### Nun ein hierarchisches Clustering"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/tv/git/pydelta/delta/graphics.py:88: MatplotlibDeprecationWarning: Passing the pad parameter of tight_layout() positionally is deprecated since Matplotlib 3.3; the parameter will become keyword-only two minor releases later.\n",
      "  plt.tight_layout(2)\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAbkAAAEGCAYAAAD4yOuIAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjQuMiwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy8rg+JYAAAACXBIWXMAAAsTAAALEwEAmpwYAABOaUlEQVR4nO2debgUxdW431IRF6AxYjS4MCrutpKEqDEaSYxJdCTyS+KG6GBEPvcYjZ+jorZGdGLiGjQKLrQirgm4jHsUt88N9Eq7Ji6DIKCo0AqiYanfH6cG+g5z7527zr3DeZ/nPnTXeqpmqNPnVE0fY61FURRFUWqRNaotgKIoiqK0F6rkFEVRlJpFlZyiKIpSs6iSUxRFUWoWVXKKoihKzaJKTlEURalZVMkpiqIoNctqq+SMMUONMVONMQuNMXOMMQ8ZY/aqtlzNwRgzxRgzog3aGWSMmdVA3iPGmJ+3to/2xhjTxxjznDHmM2PMAmPM88aYH5Up1yXGoyhK27BaKjljzGnAlcDFwMbAFsC1wEEtaGutNhWuE2GMWR8YCDzVRu0ZY0x7fecWAr8DNgI2AP4M3J/8fNp6PIqidH5WOyVnjPGAC4ETrbX/tNYustYusdbeb609w5Xpboy50hgz2/1daYzp7vIGGWNmGWPONMbMBW5OpJ1tjPnUGFMwxhyR6LOexWWMGW6MedZdG2PMFcaYT4wxXxhjImPMzhWMYzSwNzDGWaNjXPr2xpjHjDGfG2PeMcYckqhzgDHmTWPMl8aYj4wxf3QL/0NAX9fOQmNMX1dlX+A5a+03xpg13fjec/WnGWM2d+3uaYx52RgTu3/3LBn7aGPMc8BXwFYu7RJjzEtuzPcaY76VnN+SsRaMMT9z17s5C/wLY8zHxpjLAay1X1tr37HWLgcMsAxRdt9KNLViPE3Nr6IotcFqp+SAHwLrAJMaKXMOsAcwANgV2A0YlcjfBFk8+wEjE2l9gE2BDDDWGLNdBfL8HPgxsC3gAYcAn8EKl+r0cpWstecAzwAnWWt7WGtPcgrrMWAi8G3gMOBaY8yOrtqNwP9Ya3sCOwNPWGsXAfsDs107Pay1s135A4C8uz4NONyl9UKspq+ccsoDVwMbApcDeWPMhglxj3Tz1BOY4dKOcm18B1jq6lfCVcBV1tpewNbAXclMN19fA/cBN1hrP0lkJ8ejKMpqwOqo5DYEPrXWLm2kzBHAhdbaT6y184ALkIW6yHLgfGvtN9baxYn0c13aU8hieghNswRZ/LcHjLX2LWvtHABr7URr7S6VD40DgYK19mZr7VJr7avAP4CDE33taIzpZa2db619pYn2DgAedNcjgFHOWrLW2testZ8BaeA/1tpbXZ+3A28DgxPtjLfWvuHyl7i0W621rzsley5wiDFmzQrGuATob4zpY61daK19IZnp5qsXMBR4tpHxKIqyGrA6KrnPgD5N7KX1ZaXFgbvum7ifZ639uqTOfLdgN1SnLNbaJ4AxwDXAJ8aYscaYXk3Va4B+wO7u4MUCY8wCRGFv4vJ/gyz0M4wxTxljfthQQ8YYH4ittTNd0ubAe2WKls4V7n7TxP1MViWZNgPohljCTXEMYvW+7VyjB5YWcK7L24GsMWbXBsajKMpqwOqo5J4HvgGGNFJmNqIwimzh0oqUC92wgXMXlquzCFgvkbdJ4hpr7dXW2u8DOyIL+BmNyFavasn9TOApa23vxF8Pa+3xrp+XrbUHIa7Myax09ZUbT6nVMxNxD5ZSOlcgY/+oETlBlGay/BLgU0rmyll3G61oyNr/WGsPd2P4M3BPybwn6QZs1cB4FEVZDVjtlJy1NgbOA64xxgwxxqxnjOlmjNnfGHOpK3Y7MMoYs5Expo8rP6GC5i8wxqxtjNkbcR3e7dLrgF+7vvoj1ggAxpgfGGN2N8Z0Qxb4rxF3aCV8zMpFHOABYFtjzJFuTN1c+zs4uY4wxnjOZfhFop+PgQ2NHMopUrp/dQPwJ2PMNu6wzC5u3+1B1+dQY8xaxphDEWX9QBOyDzPG7GiMWQ85CHSPtXYZ8G9gHWNM2s3JKKB7Yr6GGWM2cgdMFrjk5caYPYwxe7lxrmuMORM5OftiA+NRFGV1wFq7Wv4hbrypiGKZiyyAe7q8dZCDEHPc39XAOi5vEDCrpK1BwCzkwMqnwIfAkYn8PsCjwJfAc0AAPOvy9gWmI0fgPwVuA3okZHyjkTH8EFEK84GrXdp2bizzENfsE8gBmrWBh13ZL4CXgb0Sbd3kyi9AXJDzgLUS+WsiCucDN46Xgc1c3l7ANCB2/ybbnQKMKJF7CnAJ8JKT5X6gTyJ/uJv3T4A/AgXgZy5vgktfCLwBDHHp+wCvOdk+R34m8GOX17t0PPqnf/q3evwZazVoamsxxgwCJlhrN6uyKG2C+9nBb621lRycaUn7U5D5uqE92i/TX7uOR1GUzstq565UKmIBcEW1hWhDFlBb41EUpUJq9m0dSsux1j5abRnaklobj6IolaPuSkVRFKVmUXeloiiKUrOoklMURVFqli61J9enTx+bSqWqLYaiMG3atE+ttRs1XVJRlGrSpZRcKpVi6tSp1RZDUTDGlL7KTFGUToi6KxVFUZSaRZWcoiiKUrOoklMURVFqFlVyiqIoSs3SpQ6eKIoCfugXgBFRJnq82rK0Bj/0twPuREI4nQN8D5gVZaJRbdhHgRqYK6XlqJJTlHbALa4bA8uQiAkPAydFmWhhNeVqCj/0DwIuQEI4/ReJkHFMlIk+8EM/APpHmWhYG3X3v8CTUSYa4Poe30btKsoK1F2pKO3H4CgT9UBCHX0XOKu64jSOH/r9gVuA0wEP2BKJWL+snbrsh4RLUpR2Qy05RWlnokw01w/9RxBlB4Af+lngWCTC+UzgnCgTTXJ5WwPjgF2RqOqPACdGmWhBotkf+KF/NfAdJMr78VEm+trVPxC4CEgBbwLHRZloegWiDgA+iDLRv9z9l8A/XJu/BM4GjB/6Q4D3oky0qx/6fYHrkJiCnwN/jjLROFcnAHZGlOQBwH+Ao6NM9Jof+k8gMQD38kP/SsRVuQI/9HsC9wEREkHifaBblImWuvwpwIQoE93g7o8FTgM2c/M5LMpErxTH5Yf+5YhSfRjIVDJXzhofAxxVrq7SNVAlp3Q4Y8eOZeLEidUWo8PwQ38zYH8kgG2R94C9kYC9BwMT/NDvH2WiOYBBgso+DfRCFE0AnJqofwTwCyTo7/1IQNtRfuh/FwmAOxgJCjwMuM8P/e2iTPSNH/rXAkSZ6IQyor4CbO+H/hWIgnm56F6NMtHDfuhfzKruyjuA15FAu9sDj/mh/16UiYpjPQg43Mnxe2CyH/rbRpnop2UUVXG+NgQeAh6NMtEoP/RTTczvwW5+hrgxbw0sSRQ5BPgl8DUStHg4cF1Tc9VY3cbkUToXquSUDmfixInU1dUxYMCAaovS3kz2Q98CPRAFd34xI8pEdyfK3emH/lnAbsC9USZ6F3jX5c1zVsj51GdMlIlmAvihPxr4G6LoRgLXR5noRVcu9EP/bGAP4KkGlFtRpvf90B+EWER3AT390L+DBvYS/dDfHPgRkHbWTZ0f+jcglk9RyU2LMtE9rvzliCt0D+CZBsToi0R1D6NM9JeGZC1hBHBplIledvfvluRfHWWi2U6G+1lpUTc6V03UVboIquSUqjBgwACmTJlSbTFajDGmkmJDokz0uB/6+wATgT5IAFf80D8KUSYpV7aHy8cP/Y2BqxBLryeydz6/pO2ZiesZiHIAcatl/NA/OZG/diK/UaJM9AJiveCH/g+Q04/nUH4/sS/weZSJviyRZWA5OaNMtNwP/VlNyJJGDuo0x1raHLGMG2Ju4vormjdXDdVVugiq5BSlnYky0VPu5OBfgSF+6PdD9tz2BZ6PMtEyP/TrEDclwMXIXpwfZaLP3R7YmJJmN09cbwHMdtczgdFRJhrdBnK/7If+P5F9NZxMSWYD3/JDv2dC0W0BfFROTj/010D2zGbTMOOADYAH/dD/ZZSJFiEuWYD1gC/c9SaJOjMRF2VzabO5UjovquQUpWO4Eij4ob8rsl9kgXkAfugfzUpFAmK9xUDsh/6mwBll2jvRD/0HEOviHMTiAlESk/zQfxx4CVEMg4CnSyyuVfBDfy9gB8Rl+okf+tsDvwJCV+RjYD8/9NeIMtHyKBPN9EP//4BL/ND/I7AtcAyyX1jk+37o/xrZ4zsF+AZ4oTE5gJPcOO73Qz8dZaJ5fuh/BAzzQ/96IEN9pXYDcLkf+s8i+4pbA0uiTNTUS7RbPFdK10F/QqAoHUCUieYhx/PPizLRm8BlwPOI4vCRQw1FLkBOG8ZAHvhnmSYnAo8ipw7fQ04IEmWiqcipzTGIi/Nd5LAEAH7oX+eHfkOuwAWIUov80C/+tm8ScKnLL+4jfuaHfvHk4uGIy3W2K3t+yQ+v7wUOdbIcCfw6ykTJQyGrEGUii+yXzQLu9UN/HTemM4DPgJ2A/0uUvxsY7ebkS+S06bca68PVa3SulNrAWFvqgei8DBw40Gqona7PoEGDALr6ntw0a+3ApkuuvrTDj8cVpdmoJacoiqLULKrkFEVRlJpFD54oitIuRJkoqLYMiqKWnKIoilKzqCWntJiWvp5rNXnbiaIonQC15JQWU3w9l6IoSmdFLTmlVbTk9VzFnxAoiqK0N2rJKYqiKDWLKjlFURSlZlElpyiKotQsquQURVGUmkWVnKIoilKzqJJTlC6GH/oFP/R/Vm05WoMf+sNdaJyq44f+Gy4iulKD6E8IFKUd8EO/AGwMLEMiXT8MnBRlooXVlKsp/NC3wDZRJnq3Vvr1Q384cCOwuCRr2ygTzY4y0U5t3afSeVAlpyjtx+AoEz3uh/4mwCPAWUiAU6XjeT7KRHtVWwil41Elp1SFurq61eZH4VEmmuuH/iPAgGKaH/pZJGDnt4GZwDlRJprk8rZGolbvikQQfwQ4McpECxLN/sAP/auB7yBBQo+PMtHXrv6BSBDVFPAmcFyUiaY3V24XD24nJJr3QUAB+I37+4NLPybKRI82UL/BMZaUe9pdvuYsumOiTHRnY+NwlvIY4CigH2IpZ4pz0MxxFoAR7oFkN+BaJMr5YuC2KBOd5srtAVwO7AjMAH4fZaIpLm8K8AzwU2AXJCDu0CgTfdpceZS2RffklA5n6NChq9W7K/3Q3wzYH4k8XeQ9YG/AQyKBT/BD/zsuzwCXAH2BHYDNgaCk2SOAXwBbIwvyKNfXd4GbgP8BNgSuB+7zQ7+7y7/WD/1rmyH+YOBWYAPgVUThrgFsClzo2m+Ixsa4gigT/dhd7hploh5OwTU6DschwC+BLRHFMrwZ42qIq4CrokzUC5nbuwD80N8UidJ+ERJ1/I/AP/zQ3yhRdyhwNKLU13ZllCqjlpzS4YwcOZKRI0dWW4xWYYyppNhkZ5n0AJ4Azi9mRJno7kS5O/3QPwvYDbjX7UsVFeI8P/QvT9Z1jIky0UwAP/RHA39DFN1I4PooE73oyoV+6J8N7AE8FWWiE5oxTIBnokz0iOvnbuDXQC7KRMv80L8DGOuHfu8SK7PJMVbQb6PjcGlXR5lotpPtfhKWchn28EM/KeNnUSbauky5JUB/P/T7OCvsBZc+DHgwykQPuvvH/NCfChwAhC7t5igT/dvJcxfwqwrGqbQzquQUpf0Y4lxg+wATgT7AAgA/9I8CTkNccSCKsI/L2xixKPYGeiKW0/yStmcmrmcgVh+I6y7jh/7Jify1E/nN5ePE9WLg0ygTLUvcF2VfUFqxsTFWQCXjmJu4/orGx/hChXtyxyAW6tt+6H8AXBBlogecPAf7oT84UbYb8GQj8vSooD+lnVElpyjtTJSJnvJDfzzwV2CIH/r9kD23fZEDEcv80K9D3JQAFyN7cX6UiT73Q38Isv+UZPPE9RbAbHc9ExgdZaLR7TGWSqlgjE1RlXFEmeg/wOF+6K+BWK33+KG/oZPn1igTHduR8iitR5WconQMVwIFP/R3RVxiFpgH4If+0cDOibI9gRiI3V7QGWXaO9EP/QcQi+Ec4E6XPg6Y5If+48BLwHrAIODpKBN92cZjaoz1aXyMpXwMbMVKN21VxuGH/jDgkSgTzUu4N5cDE4CX/dD/BfA4YsXtAbwbZaJZ7SWP0npUySn1aE4gVA1+Wjlu0bwFOC/KRL/xQ/8y5ATecuAW4LlE8QtcWows+rcipxmTTAQeRVx09yIHIogy0VQ/9I9FLL9tEJfis8DTAH7oX+fKHdcOw1xBlInebGKMpQTIvtu6wMgoE93V2DhawA/90C/9jeJPokz0cknaL4HL/dBfD3EDHxZlosXATD/0DwIuBW5Hfv/4EnB8C+VROghjra22DBUzcOBAO3Xq1GqLUdMMGjSoYuVVLNfceHK1gDFmmrV2YLXlUBSlcdSSU1ahUsW1uvzOTVGUrov+Tk5RFEWpWVTJKYqiKDWLKjlFURSlZlElpyiKotQsquQURVGUmkWVnKIoilKzqJJTFEVRahZVcoqiKErNokpOURRFqVlUySmKoig1iyo5Reli+KFf8EP/Z1Xqe7wf+hd1UF9H+KH/aOLe+qHfv4VtVW3OWosf+lP80B/hruvNidI0+u5KRWkH/NAvABsjb6tfCDwMnBRlotI34Xcq/NA3wMlIZO4tkWCtzwMXRpko6khZokx0G3BbR/ZZKS5K+bHARkjA2OeiTHRoe/fbmeeks6KWnKK0H4OjTNQDGAB8FziruuJUxFXA74FTgG8B2wKTgXRzG/JDv8s/RJcbgx/6GeBI4Gfu8x0I/Ksd+jYueKvSCrr8l1BROjtRJprrh/4jiLIDwA/9LGIJfBuJOn1OlIkmubytkaChuyKBRx8BTowy0YJEsz/wQ/9q4DuIEjo+ykRfu/oHIvHlUsCbwHFRJprelJx+6G8DnAj8MMpELyWySi2HDfzQzwM/du0PjTLRe64NC5wEnIqsL1u6uHBnIkrzWSfP7ET544HTEavoNsTitX7oDwdGRJlorzKy7oXEdTsyykRTyuQf6eagB3B5Sd4awP8i898bUVDHuSjsKeADYARwPlBw40zyAySw6nsgny8wNtH+FMT63RfYHngSODrKRJ+7/D2cTDsiMet+XxyDq/scEiD2e4BfInu9OWli/tZE4t9lgC+By4C/Ad2iTLTUtXWeq/cpMCrKRLdVUO9oN3+bIUFx/xxlouudPH2A8cBeSBzBN4B9oky0nCqhSk5pFXV1dRpypwn80N8M2B94IpH8HrA3MBc4GJjgh37/KBPNAQxwCRIgtBfwDySo6KmJ+kcAvwAWAfcDo4BRfuh/F7gJGAxMBYYB9/mhv12Uib7xQ/9agCgTnVBG1H2BWSUKrhyHufG8AoTAaJdWZAiwO7DYD/2furH8HFnw/grcQX3FcSCiOHoB09x4Hm6ocz/0f4k8BPymnKx+6O8I/B04AHjR9b9ZosjJTsZ9kEX6auAa4PBEmX2AHZCFupQXgKv90P8IUWCvRploWUmZo5DP5wMkYOzVwDAX6T2PWIIPI3P+Dz/0t48y0TxX90hkft9BvgtN0dD8HevaGYB8T+4uVvBDf30n0w+iTPSOH/rfQR5CaKye4xPX5/vI5/iQH/ovR5noFUTZzkIUJ0j09KoGLVUlp7SYoUOHVluEzs5k96TdA1Fw5xczokyUXDju9EP/LGA34N4oE72LRAQHmOeH/uXJuo4xUSaaCeCH/mjkSXsUspd2fZSJXnTlQrd/tAfwVAPKrciGwJwKxjWpqFz80L+NEksJuCRhtRwB3OQWQNw45/uhn4oyUcGVzzkrdYEf+k8ii2tDSu5g4Dhg/ygTvd5Amd8CD0SZqBgN/VzEuixyHGLtzHL5AfChs/6KBFEmWlSu8SgTTXCf69HIw8fXfuhfGmWiPyeK3VqUz/Vf59ycw4AHo0z0oCv3mB/6UxGFHLq08VEmeqPYkB/WM+bK0dD8HQJclRhnDlGqRZYDO/uh/6F7uCp+9o3WizJRPtHGU+4gzN7IQ88SxLvQz32Pn2lK+PZGlVwDjB07lokTJ1ZbjA6n0qjgACNHjmTkyJHtK1AnxZhKHrAZEmWix/3Q3weYCPRBDingh/5RwGmISxFEEfZxeRsje2N7Az2RvfP5JW3PTFzPAPq6635Axg/9kxP5ayfyG+MzZIFqirmJ66+c7A3J1hdZ/ACIMtFCP/Q/AzZFXIGVtJfkVOCWRhRcsc8VMkSZaJHrs0g/YJIf+kkrbRlyUKjcGFaheADED/1uiFV4mx/6dVEmeqRM/RlAN+Tz7Qcc7If+4ER+N8QirKjvMjQ0f/XmgVXn5FDgj8CNfug/B5weZaK3G6sH4If+/shD17bId3M9oHgo6S+I4n/UKeexUSbKNXM8bYoquQaYOHFisxZ8RWmIKBM95Yf+eMRVN8QP/X6Iu21f4PkoEy3zQ7+Ola6pixEXj+/2iYYAY0qa3TxxvQUw213PBEZHmWh0C0T9F3CNH/oDo0w0tQX1iyTdU7ORhR1Y4SbbEPiohW0fjCzKs6JMdFUDZeYgrsZin+u5PovMBH4XZaLnSiu6PTmo0MUWZaIlwN1+6J8J7Izsn8Kqn88SZN9rJmLlHdtIs23l3ptDfTdtUiacQn7ED/11kf3LcciDVYP1/NDvjrjPj0K8Dkv80J+M++5GmehLxGV5uh/6OwNPOFdmmx/MqRRVco0wYMAApkyZUm0xOhTdX2s3rgQKfujviix4FtkPwm3k75wo2xOIgdjt4ZxRpr0T/dB/AHlyPwe406WPQ6yUx4GXkKfsQcDTbgFqkCgT/cft2d3uDov8H/KkPgRItfCJ/HbX3kTgLUSBv5hwVTaX2cjDwRQ/9P8bZaK/lylzD/CiO5zyEnAh9U+SXweM9kM/E2WiGX7obwTsGWWieysRwB3YmIfsmS5C9t52Qvb/igzzQ/8WxFq9ELjHPcxMAF72Q/8XwOOIFbcH8G7RPdiG3AX83h0SWoQc/imOYWPX7+PAYuRnLsubqod4Bboj41/qrLqfA0XX7IHA28iec4xYyFU7dAL6EwJF6RDcoYJbgPOiTPQmcmLteeBj5ARd0qq4ADlZFyOHFP5ZpsmJwKPI5v97yJM4zgI7FrH85iN7e8OLlfzQv84P/esaEfUUV/caxLX6HvD/kMMMzSbKRI8D5yJP/3OAral/SKUlbX6IKLps8UfSJflvIKdEJ7o+5yOHIYpcBdyHuNS+RA6S7N4MEb4AzgY+ROboUuR067OJMrcipwznAusg84rbRz3I1Z+HWHZn0D5r8TjkOzIdeBV4EFiKKJ41EHf5bOBz5KDN8U3Vcw9KpyCKcD4wFJnLItsginMh8v2+NspET8Kq372m7tsKY21VD740i4EDB9qpU1vjRamcokWzulpyq9u4m4sxZpq1dmC15VA6H+5nABOiTHRDtWVJ4qyu66JM1K/Jwm1Qr7Og7kpFUZQaxO21/QSxyjZGDotMaq96nRV1VyqKotQmBnF9z0fcjm8hP/5ur3qdErXkFEVR2pAoEw2qtgwAUSb6CvmReIfU66yoJacoiqLULKrkFEVRlJpFlZyiKIpSs6iSUxRFUWoWVXKKoihKzaJKTlEURalZVMkpiqIoNYsqOUVRFKVmUSWnKIrSCUhl89elsvlz3fWgVDbfoqgEralbi+gbTxSlHUhl8wXkvX/LEsnbFnLp2eVrVNTmeGBWIZce1TrpWk8qmw+QED/fJJIvLOTSl3ZAv/0LufSw9uynUlLZ/NlIRAGQ9bQbEroGYEYhl96p0rYKufRxbSyegio5RWlPBhdy6cerLUQ7cmdTyiaVza9ZyKWXNVamK1PIpS9GYuSRyuaHAyMKufRepeVS2bwBTCGXrmpstdURVXKK0oGksvnuwJ+BQ1zSXcCZhVz6m1Q2PwiYAFyBBKpcBpxdyKVvTmXzI4EjAJvK5k8Fnizk0oNT2XwWiR/3bSQ22TmFXHqS62s4MAKJl3YMEvvshEIu/ZDL94DLgQOQwJY3A+e3Rik5a3MxEg18H+AgN+aLkFhyMXBjIZcOXPlBwIRCLr1Zoo0CoiyafEBIZfPbA38Dvo/EZzu3kEvflZBlEZACfgy8CQwt5NLvuXyLxFA7HdgIuA04CbHG5gL7FHLpyJX9NhIAtV8hl55X4VxMQeIEDkLiA/qpbH6tJuStZ6k7S/E0JD7bOYVc+jaX3h0YjXyPuiNRAv5QyKWLVqTiUCWnrEJdXZ1GCG8/zkEiMg9AooPfC4xCAosCbAJ4wKbAfsA9qWx+ciGXHpvK5vdkVXfle8DeyKJ8MDAhlc33L+TSc1z+7kAI9AFGAjemsvlNC7m0RYJ6fgL0B9YHHkAU5fWpbH4LJGjmLoVc+sNmjnEoojgPRCJJ7wEcBbyBREB/LJXN1xVy6cnNbLceqWx+feAx5A35+yPBZx9LZfOvF3LpN12xw1zeK8g8jKZ+0NYDkZcR9wKmAfcXcumHU9n8HcAwVkbFPhz4V6UKLsGRrv93kDl+vQl5k2yCfG6bInP4YCqbn1rIpd8BcshDwwAk0vxE1+5ZzZSv5lElp9Rj6NCh1RahlpicyuaXuusphVx6CGKNnVzIpT8BSGXzFwDXs1LJLUH2tpYii9pCYDvEGluFQi59d+L2zlQ2fxawG6I8QfaFxrm+QuBaYGNnxRwA9HZP/4tS2fwViCK83im23k2M75BUNn9g4n5H9++9hVy6GOn8a2BKosz0VDZ/O2LlTW6i/aY4ECgUcumb3f2rqWz+H4iyv8ClTSrk0i8BpLL52xDLNUmukEsvABaksvknEaXxMKIQ705l81n3QHAkEgG8uYwv5NJvuP5/WYG8pZxbyKW/AZ5KZfN5ZM4vQj6nXQq59Oeu7YsRRadKrgRVcko9Ro4cyciRI6stRqfHGFNJsSFlXG59gRmJ+xkurchnTsEV+Qro0VAHqWz+KMSdlXJJPZCn/yJzixeFXPqrVDZfLPMtxC03x6WBnLae2eiI6nNX6Z6ca2tmSdruiOWxM2LZdQeSyrml9AN2T2XzCxJpawG3Ju7nJq7LzWXZ/EIu/WIqm/8KGJTK5ucg1u59LZAxOReVyJtkfiGXXpS4L35XNgLWA6YlPjsDrNkC+WoeVXKK0rHMRha7N9z9Fi6tEmzyJpXN9wPGAfsCzxdy6WWpbL4OWfCaYiZyMrJPiVJtC2zJ/URgDLB/IZf+OpXNX8lKRbwIWbABOaiCLOKVMBN4qpBL79c6cRskRFyWc4F7Crn01y1oIzkXzZV3g1Q2v35C0W2BuDs/RfY9dyrk0h+1QKbVClVyitKx3A6MSmXzLyML4HnIYZNK+BjYKnG/vmtjHkAqmz8asZaapJBLz0ll848Cl7nfZi0EtgQ2K+TST1UoT6X0BD53Cm43ZM/uUZf3b2CdVDafdmlnI5ZeJTwA5FLZ/JHAHS5tALCwkEu/1QZyTwBeA75E3JWtpSXyXuAOn+yOuGfPL+TSy1PZ/DjgilQ2f1Ihl/4klc1vCuxcyKUfaQM5awr9MbiidCwXAVORQx0RciDiogrr3gjsmMrmF7jDKG8ClwHPIwrQR07zVcpRiPvwTWA+cA/wHYBUNr9FKptf6A6gtJYTgAtT2fyXiFK/q5hRyKVjl38D8BFi2TX1Q2br6n4J/Bw5SDIbsbj+TOVKslEKufRM5POxwDNt0F5z5Z2LfC6zkZOfxxVy6bdd3pnAu8ALqWz+C+BxZO9WKcFYW+pZ6LwMHDjQTp06tUP6Kp4unDJlSof0p3QtjDHTrLUDqy3H6kYqm78cWKOQS5/aQf3dBMzuDD/AV1pGTbgrx44dy8SJE9u0zbq6OgYMGNCmbSqK0nJS2Xxv4BfAJR3UXwr4NfDdjuhPaR9qwl05ceJE6urqqi2GoijthPupwnvAiyTcne3Y35+QQx5/KeTSH7R3f0r7URPuyvZwLaq7UmkMdVcqStegJiw5RVEURSmHKjlFURSlZlElpyiKotQsquQURVGUmkWVnKIoilKzqJJTFEVRahZVcoqiKErNokpOURRFqVlUySmKoig1S028u1JRlM5FKpv/ETAeiWowDImUcDfyHsixwGfAVoVcekQb9rkQiZb9flu1qXR9VMkpSjuQyuYLwMbAskTytoVcutIAqeXaHA/M6gxvxE9l8wFwDhJ4tcjSQi7d211fCIwp5NJXufLnIsE+exVy6UbfJZjK5gcBTyCRugFi4IZCLn1+Y/UKuXSDEdQrxX1uI8pEdFe6KKrkFKX9GFzji+WdhVx6WAN5yejnxfs3m1JwCWYXcunNAFLZ/JbAM6ls/tVCLj25tGAqm1+rHaKbKzWCKjmlQ2mPsEhdiVQ23x0JlHmIS7oLOLOQS3/jLJgJwBVIUMxlwNmFXPrmVDY/EjgCsKls/lTgyUIuPTiVzWeBY4FvAzOBcwq59CTX13BgBPACcAywADihkEs/5PI94HLgAGA5cDMSeTppfbZkjO8hUcbvT2Xzy4D7gd8mZB8C7AX0b0RJrqCQS3+Qyub/D9gRmOz6sMBJwKnIOralS9umkEu/6+Z5NDLP3YFJwB8KufTiVDbfB3Gl7uXG/QawDxACWyTkvrCQS1/amrlQqo8ePFE6FA2LxDnAHsAAYFdgNyDpftwE8IBNEcV0TSqb36CQS49FokNfWsilexRy6cGu/HvA3q7OBcCEVDb/nUR7uwPvAH2AS4EbU9m8cXnjgaVAf2Sv7OeIUixGBl/QksjghVx6a+BDxJLtUcilDy+RvVnWbSqb3wb4EaKskwxx49uxTLUcsC0yz/2R+TzP5Z2ORB/fCHEpnw3YQi59ZIncquBqALXklA5nwIABXT6EkTGm6UIwOZXNF91oUwq59BDEGju5kEt/ApDK5i8ArgfOdeWWIBbEUuBBd5hiO1Zd4AEo5NJ3J27vTGXzZyGK816XNqOQS49zfYXAtcDGzuo5AOhdyKUXA4tS2fwVwEjg+kIu/SHQu4nxHeLivBV5tZBL/6SJOpXSN5XNL0AexHsiltizJWUuKeTSn5dWdEp8JHII5XOXdjEwETgLmePvAP0KufS7wDNtJLPSCVElpyjtx5AyVktfYEbifoZLK/JZyf7SV0CDBypS2fxRwGlAyiX1QKy2InOLF4Vc+qtUNl8s8y2gGzDHpYEolJmNjqg+d1XibmwhyT05D1HOIXB4okxDsm4ErAdMS4zNAGu6678AAfCoyx9byKVzbSm80nlQJacoHcts6h/K2MKlVUK9QxupbL4fMA7YF3i+kEsvS2XzdciC3hQzkZORfTr7oY1CLh2nsvmJwJ0lWQ0dYvkUWAzsVMilPyrT3peIy/L0VDa/M/BEKpt/uZBL/6uRNpUuiio5RelYbgdGpbL5l5EF9TzksEklfAxslbhf37UxDyCVzR8N7FxJQ4Vcek4qm38UuMwd71+IHBbZrJBLP1WhPB1CKpvvARxG/dOaDVLIpZensvlxwBWpbP6kQi79SSqb3xTYuZBLP+JcrG8j+5kxcsBnuateOsdKF0eVnKJ0LBcBvYDp7v5ul1YJNwJ3u72qKYVcekgqm78M+aH1cuAW4LlmyHIUckDjTWTf633k5CfuwMmbwI5uf64ch6ay+SElaVsV9xtbSV+3Hwlicb6A7GdWypnIA8QL7jTlR8DfgUeAbYAxiFtzPnBtIZd+0tW7BPhbKpu/FLiokEv/tdUjUaqKsbbrWOcDBw60U6dOXSV90KBBAG16mKE92lRqZ16NMdOstQOrLYeiKI2jPyFQFEVRahZVcoqiKErNokpOURRFqVn04Ekj1NXVrdhDUtqGuro6BgwYUG0xFEVZTVAl1wBDhw6ttgiKoihKK1El1wAjR45k5MiR1Raj5lDLWFGUjkT35BRFUZSaRZWcoiiKUrOoklMURVFqFlVyiqIoSs2iSk5RFEWpWVTJKYrSaUll88NT2XxpsNRk/pRUNl+MZn6Ei6ygKCvQnxAoSjuQyuYLwMZIGJci2xZy6Upjx5Vrczwwq5BLj2qddK0nlc33Bi5HoouvD8wBbioGH3WRx7dxkbc7hEIufRtwW0f1p3QNVMkpSvsxuExk8FrhCkS57YDEZNuWCmPZKUpHokpOUTqQVDbfHYnZdohLugs4s5BLf5PK5gchAVSvQOKhLQPOLuTSN6ey+ZFIPDWbyuZPBZ4s5NKDU9l8FjgW+DYS7fucQi49yfU1HBiBxGI7BlgAnFDIpR9y+R4rrbHlwM3A+YVcOml9NsQPgFGFXHq+u3/b/ZHK5p92aa85i+6YQi59pwtWehGQQmLVHVfIpae7OpsDVwF7I9sotxdy6ZMS8/bXcmMomdvhwIhCLr2Xu7fA8UgU8I0QK++kQi7ddeKLKa1GlZzS4azm7wQ9B9gDGIBE9b4XGAWc6/I3ATxgU2A/4J5UNj+5kEuPTWXze7Kqu/I9RDHMBQ4GJqSy+f6FXHqOy98dCIE+wEjgxlQ2v6lb6McDnwD9EavsAURRXu+Cpk4HdmkgaOoLwOhUNr8B8Gwhl/5PMaOQS//YKZhdi+7KVDb/XeAmYDAwFRgG3JfK5rcDlrq+nwCORJR7MlZfY2NoigMRhdwLmAbcDzxcQT2lRlAlp3Qoq9k7QSensvml7npKIZceglhjJxejZ6ey+QuA61mp5JYAFxZy6aXAgy469naIUlmFQi59d+L2zlQ2fxawG6I8AWYUculxrq8QuBbY2CmhA4DehVx6MbAolc1fgSiR651i693I2E4G/gCcBIxNZfMz3LhWsbAcxXZfdPdhKps/G1H4/wX6Ame4cQMkD5uUHQOi2JsiV8ilFwALUtn8k8jDhSq51QhVckqHUivvBDXGVFJsSJk9ub7AjMT9DJdW5LPEQg/wFdCjoQ5S2fxRwGmICxBXtk+iyApFUMilv0pl88Uy3wK6AXNcGoibcGajI1rZ1mLgYuDiVDbfC8gCd6ey+S0KufTnZar0AzKpbP7kRNrayNiXIYpsaZl6jY2hEpKKsNG5VGoTVXKK0rHMRhb8N9z9Fi6tEuq551LZfD9gHLAv8Hwhl16WyubrgEo08EzgG6BPI8qlIgq59BepbP5i4CxgS6CckpsJjC7k0qNLM1LZ/A+BLVLZ/FqtlUVRSlElpygdy+3AqFQ2/zKitM5DDptUwsfAVon79V0b8wBS2fzRVHjCsZBLz3G/Kbsslc2fCyxEFNRmhVz6qabquzoPA68hFuDvkUMh75TIWvwJwThgUiqbfxx4CVgPGAQ87e7nALlUNn8+Ytl9v5BLP1fJWBSlMfTH4IrSsVyEHLyYDkTAKy6tEm4Edkxl8wvcYZQ3gcuA5xGl4gPNUQxHIS7DN4H5wD3AdwBS2fwWqWx+oTuAUg6LnMb8FLFE9wPShVx6ocsPkH23Bals/pBCLj0VOQU6xvX1LjAcwJ3mHIwcgPkQmAUc2oxxKEqDGGu7zmnagQMH2qlTp66SXjypN2XKlI4VSFltMcZMs9YObLqkoijVRC05RVEUpWZRJacoiqLULKrkFEVRlJpFlZyiKIpSs6iSUxRFUWoWVXKKoihKzaJKTlEURalZVMkpiqIoNYsqOUVRFKVmUSWnKIqi1Cyq5BRFUZSaRZWconQlAi9F4FkCby13/xCBl6myVPUJvOsIvHMbybcEXv+q9N3ZCLwpBN6IaotRy2ioHUVpDwKvgAQE7UsQf5pIfxWJTr0lQVxofT/x/om2hwMjCOK9EmnjgVkE8ajK2/QssA1B/G4iLQD6E8TDKpDpuIr7amuq2bfSKVFLTlHajw+Aw1fcBZ6PxFFrGUXrTSlP4K1ZbRGUzof+p1GU9uNWJGbb39x9BriFZPy4wEu7+62BGLiRIA5cXgpRlCOA84GCa49E/SlI0NXngOuAbgTeQmAp8L/AEYAl8E4FniSIB7fJyAJvezeu7yNBW88liO9yeeNJWo+BdwZwGhKDblRJO92B0cAhQHdgEvAHgngxgTfIje1aV38hcA5BfFuin8VIpPV9gIMIvGEr+l5Z/wrgTCQY69kE8c0EXl/g3wlJ1gDWJYiNa/t3wBnAJkhQ15EE8QwCzwCXI/O6DjADOJwgfp3AOwD4K7A58AVwBUH8V9feQcAFSCDZecCJBPHDru9+BN5zwC5IbMChK6z/wNvD9bej6+v3BPEUlzcFeAb4adm6CqBKTmkHxo4dy8SJE6stRmfgBeBIAm8HZEE9DPgR9YOkLkIU1xtIVO/HCLw6gnhyosw+wA7AcmDjsj0F8VsE3nGs6q7ck1J3ZeBd6+qc0KJRBd76wGNIVPP9kWCtjxF4rxPEb5aU/SXwR2BfRGGPK2kthyj4AcASYKJr9yyXvwnQB9gU2AN4kMCbShAXI5APBQ4ADkQCwJa6UzcBPFd/P+AeAm8yQTwb6JGQ8zaKni1RSGcjgVz/A2SRiO57Aj8HfgxsizyUbI9ERAcJansIQfwMgbcBEmkdAm835OHmt8C/kMC0PRMyDnXzOBN4yM1XlsDbFMgDRyJR2PcF/kHgbU8Qz2u0rrICVXJKmzNx4kTq6uoYMGBAtUXpDBStuaeAt4CP6uUWn8qF6QTe7YhSm5wsRRAvkiuv9RJVptxeIfCWJ+7XQSKHgyiUAkF8s7t/lcD7B3AwYq0kOQS4mSB+Xfr2AoouXLGKRgK7EMSfu7SLEUV3VqKNcwnib4CnCLy8a/NPLu9egrgYDf3rMvOzBLiQIF6KKMiFwHbIAwiuzzMRZVV8ODgOuIQgfish09kEXj/XXk9X/qUVZVb2tSOB9xpBPB+JgA5wDHATQfyYu6//HZD5+bfr6y7gVy59mMgcP+juHyPwpiJKPWyiruJQJae0CwMGDKjpSO3GmEqL3go8jTzV37JKbuDtjlgzOyOWSHfg7pJSM1sqZyv4XtmDJ0I/YHcCb0Gi/FrIWEvpC0xL3M9IXG+E7FFOSygnAyT31uavUPAr6/dN3Dc1N585BVfkK+pbcPsDvwd2J4gXu9R+wFUE3mWJegbYlCB+gsAbA1yDuBn/CfyRIP4C+A3ijs0ReNOBLEH8POK+fJCGmduAfP2Agwm8pIu5G/BkBXUVhyo5RWlPZB/nA+Tp+5gyJSYCY4D9CeKvCbwrEfdcElthb+XKVVq3OcwEniKI96ug7BxkkS+yReL6U2RPbSeCuNS6KbIBgbd+QtFtAbyeyG/5+AJvO8Qi+jVBnFSWM4HRK/b+VqkXXw1cTeB9G7gL2bs7lyB+GdkX7Aac5PI2d+1t3QIJZwK3EsTHtqCu4lAlpyjtzzHABgTxojInJHsCnzsFtxuyx/JoC/v5GNiMwFubIP5vIm2rFrbXEA8g1sqRwB0ubQCwsMR9B7LQ30zg3YIcnDl/RU4QLyfwxgFXEHgnEcSfuH2onQniRxJtXEDgnQ3sjrhKz6e1BF4v4F7kIMuzJbnXAX9ye6NvEHge8HOC+G4C7wfI3t0ryH7q18ByAm9txF37AEEcE3hfIHuoIHt1jxJ4DyBWmOzJBfHbTUg5AXiZwPsF8Dhixe0BvEsQz2rdBKw+6E8IFKW9CeL3COKpDeSeAFxI4H2JHLi4qxU9PYEcYJlL4BVP2N2I7BMtIPAmizzedQTedS3uJYi/RA5gHAbMRlxmf0ZcraVlHwKudLK96/5NcqZLf8EphseRPbMic5G9rdnAbcBxFSiHSvie6+cKAm/hij+ReZIbzx1OpteRwx0AvZDDM/MR1+lnwF9c3pFAwdU5DjmBCUH8EnA0csozRvZn+zUpoViXxUMw8xDL7gx03W4Wxtr28Ga0DwMHDrRTp666VgwaNAigpveAuhKrw+dhjJlmrR1YbTlqmuJPAIJ4s2qLonRd9IlAURRFqVlUySmKoig1ix48URSlcyK/IVRXpdIq1JJTFEVRaha15JRWU/oaL33biaIonQW15JRWU3yNl6IoSmdDLTmlTUi+xqv4EwJFUZRqo5acoiiKUrOoklMURVFqFlVyiqIoSs2iSk5RFEWpWVTJKYqiKDWLKjlF6UoEXorAsytC9gTeQwRepspSdTyBdwSB19KQRKvOY2cm8LaTsD/elwTeKa1saziBVxpaqP3qdQI6/wesKF2RwCsgEaz7EsSfJtJfRWKvbUkQF1rfT7z/ymtvODCCIN4rkTYemEUQj6q8Tc8iUaaLIUqWEsS9Wy1rWyIBTVcGNRWZt6kXzbxayGc/giB+3N0fBvwdGEIQP9WCFv8XeJIgHtBWIrYbLfm+tTNqySlK+/EBcPiKu8DzgfVa3FrHWh27EsQ93F/vKsvSdREr+xog3WwFt3KO+yFxApUWoF9UpV2oq6vTH4XDrcBRwN/cfQa4BbhoRYnAS7v7rZGAmjcSxIHLSyGKcgQSDbvg2iNRfwoSQfo5JKJ1Nxf8cyliARwBWALvVMQaGNyikZSX5ccE3t3A3sC6wGvA8QTxG/VkC+Ib3P1wSi3NVdv/HXAh0AM4C5iGBH7dwrV10iptBd7TrpXXnEV3DEF8Z0n7ayKBUIcDXwCXleR7wOXAAUhE75uB8wniZSv6gheQKO8LgBNcQNjG5ux/gNHAL1YEzQ28rZGgq7silvIjwIkE8QKXX0CsviOA7Qi854B9gL0IvCuRYK9jaWxeZQ6OB04HNkIs3pMI4lWDhwbeX4AfAmmXUn4O6tcxrtwRwDpI8NjDgT1pq+9bG6KWnNLmDB06VN9dKbwA9CLwdnCL7GGIQkqyCFFcvZGF5ngCb0hJmX2AHYBfNNhTEL+FRKN+foX1FcRjkQXuUpcmC07gXUvgXdvCMZXK8hCwDfBt4BWSLsSWsbtr71Akovg5wM+AnYBDCLx9VqkRxD92V0Xr885VysCxwIHAd4GBwG9L8scjDwb9XZmfI4otKdc7QB/gUuBGt9g3xPGIst63JCq8AS5BXNk7AJsDQUndw5HvQm+C+KfAM4iS6kEQ/7uRPpMcCPwA2AU4hNLvTuCtQeCNc/k/J4hjmp6DIj8HfgxsC3iu/c8a/L5VGbXklDZn5MiRjBw5stpitCvGNLa+1aNozT0FvAV8VC9XwskUmU7g3Y4oksnJUgTxIrnyWiJufYL4hApKvULgLXfXtyBP7vVlkbZuWnntBcB8As9zi2ZL+BNB/DXwKIG3CLidIP7Etf8Msvi2ZF/rEOBKgnima+sSYJC73hixXnoTxIuBRQTeFcBI4HpXfwZBPM6VD4FrgY2BuQ30tx/wJBDVS5U9w+K+4TwC73LEMk5y9Qo5W07OWYcLCLwnkX3gh11eN+B2ZP0fTBD/t8I5KLIE6AlsD7zkHrA6LarkFKV9uRV4GtgSURb1CbzdgRywM7A20B24u6RUaxe8lvC9eoc4xJ1YXxaxTkcDByNusaJS7IO4XlvCx4nrxWXue7Sw3b7Un8cZiet+yMI/J/EQsUZJ+ZXKLIi/cuUak+V4YBRwA4F3zApXoSiTqxAXb0/Xz/ySum3xeSeV71clsvZH3KW7EcT/dWmVzIEQxE8QeGOQvcZ+BN4/gT8SxF+0gdxtjrorFaU9CeIZyF7TAcA/y5SYCNwHbE4Qe8i+WqmZuOpeSnnKlau0bqUk2xsKHIS4Ez0g5dKL8i+i/kGbTdpYluYwB3ENFtkicT0T+Abo49y8vQniXgTxTq3o72NgX0SZJV3DFyNz6BPEvYBhNP/zbu28vgUcDTxE4G3n0po3B0F8NUH8fWBHxG15RoWydziq5BSl/TkG+Gk9N99KegKfE8RfE3i7IYqjpXwMbEbgrV2StlUr2myMnsjC+Bmy6F5ckl8H/JrAW4/A64/MQ3vR1DjvAk4h8DYj8DYAsitygngO8ChwGYHXy+1XbV12/685BPFsRNH90rn+QOZsIRATeJuyUjk0hzpaO69BfDtwNvC4jLUZcxB4PyDwdifwuiEK92tWWvHt+X1rETXjrtTTfNVDg6Q2QRC/10juCcjCMgbZa7oLOYTSEp5AjprPJfCWE8R9kJOJdxN4C4ApBPEQAu86J9dxLeynyC3IgYaPgM+BcxE3XZErkMMPHwPTkUMJP2tlnw0RACGBty4wkiC+qyR/HGJxvIacrvwr8NNE/lGI2/hNRBG9j5zGbKVU8YcE3k+Bpwm8r4ELkHmLkb25W4E/NLPVtpnXIA7dA9ETTplVOge9nAxbIQruEeAvLq/p71tT922MsbbTWZcNMnDgQDt16tRV0ksjUysdS1HJFePJrQ4YY6ZZawdWWw5FURqnJiy51eE0X2dGLWhFUToruienKIqi1Cyq5BRFUZSaRZWcoiiKUrOoklMURVFqFlVyiqIoSs2iSk5RFEWpWVTJKYqiKDWLKjlFURSlZlElpyiKotQsquQURVGUmkWVnKJ0JQIvReBZAm8td/8QgZepslT1CbzhBN6zLax7BIH3aBtLVGz7OgLv3DZoZwqBVy5idtsTeOMJvIs6pK8apSbeXakonY7AKyCBOvsSxJ8m0l9FojRvSRAXWt9PvH+i7eHACIJ4r0TaeGAWQTyq8jY9A5yIRIXuj7y1/23gOoL4jlbL3Gjf8W3IW/Xbo+12ect9hxF4g4AJBPFm7dT+eCADDCGI702kXwGcChxNEI9vl77bEbXkFKX9+AA4fMVd4PnUD3bZPIrWW/tzNbKonQ5sCGyKRLn+ZQf1X5t03OfXGv6NhNwRROZDgMbCRXVqusKkK0pX5VZkwfibu88gscRWup8CL+3ut0ZijN1IEAcuL4UoyhHA+UCB5AIkZaYAE4DnkKji3Qi8hcBS4H+BIwBL4J0KPEkQD25U4sDbFolxtztBnIxr9az7K5bzgMuRiOfLgZuB8wniZYkyf0UCei4ATiCIH2qybqk1GngWsSr/gETAvhIYj8ztzsDDwDCC+L8rLB2JdXYmsAw4myC+2bU1nqJVG3h9XDt7ORneAPYhiIvBP5Nzsh/yGX7H9WtK8n+HBD/dBHgJiWc3IyH/SchDw1rAlmXavxuJIL4uEu/ueIL4jZIy6wMPAd3d5wsSH28u8jkfi8Qh/BdwHEH8eeL7Mxz4E/KAdQVBPHoVGVZyP3AkgbcBQTwfebCZjsSXK8qyNRKfb1ckEvgjwIkE8QKXXwCuAY5Evtd3IAFaxyPz/SJwMEE8n8BbB7gB2B9YE/gPcCBB/HEjMjYLVXJKm6BBa8vyArJg7IA8IR8G/IikkpPIykchi+zOwGMEXh1BPDlRZh9gB2Qx3rhsT0H8FoF3HKu6K/ek1F0ZeNe6OieUaemnwMwSBVeO8cAniDtzfeABYCZwvcvfHQiBPojb80YCb1OC2FZQt5RfAN8HNgdeAfYEhiERyZ9HrOXQld0E8BDrcz/gHgJvsluwk5wOzAI2cvd7IAt2fUQZ/hM4GrgXUVjHIcoOAu8gZAEfjCzQWeB2J2ORIW4+FjcwvoeA3wH/RYKU3oa4tBNyxIsIvP0pdVcG3u9d+/sA8xAr/BqSHgRRLNshSvElAu+fBPFbDcjytRvnYcDfke/mLciDRhEDXAI8jQRQ/QcStPbURJnfIPO/FvAq8F3kgect4EHgFCSAbAb5vDZHoswPoOF5ahGq5JRWM3To0GqL0JkpWnNPIf/BP6qXG8RTEnfTCbzbkQVrcrIUQbxIrrzWS1ReuRXpg1gHifLeLKAHsA6yWH6NWGG9CeLFwCK3bzOSlYpqBkE8ztUPgWuBjZ1l01TdUi4liL8A3iDwXgceJYjfd20/hCygRSW3BLiQIF4KPOisnu2QB44kSxDLrB9B/C7wTAN9HyD9xve4/q5EFGSR44BLViiNwLsYOJvA67fCmpP8zxtoH4L4ppXXXgDMJ/A8gjhusE79/k8iiGcl6n9I4B2ZKHOBm+vXCLzXEAusISUHotT+kvguZkgqOZmvd93dPALvcsTTkORvK6yxwHsG+IQgftXdTwL2deWWIC7x/gTxdGBaBWNuFqrklFazOgatNcY0XUi4FXni3RJZPOoTeLsDOcSKWxvoDtxdUmpmS+VsAZ8hi/9KgngztzezBHmK7wd0A+YklO4a1JdzbqL+V65cD+BbFdQtJem6WlzmfpN68ouCK/KV67eUvyDWx6NOjrEEca5Mub71ZAtiS+AlZe0HXEXgXZZIM4glWVRyDY8t8NYERgMHI1Zl0V3aB3FfN0U/YBKBl3SzLqO+xZ98aGloPhIyxc8SeBsB5wAPEMSL6z1cBd7GwFWIi7Un8vmVWspNfWZFGW5FrLg7CLzeiLv5HIJ4SaMyNgNVcorSngTxDALvA8QiOKZMiYnAGGB/gvhrZyn0KSmzqhutPOXKVVq3yBPAGAJvYCMuy5mIa6lPiUKphNbUbTuC+EvEIjudwNsZeILAe5kg/ldJyTnIIuzqeabevYxntDsV2hCNfQZDgYOAnyF7rh6iMMo9RZVrZybwO4L4uVVyZE+upUwAzgN+UibvYieL7/b+hiDf4eYjyuwC4AIn74PAO8CNLWqvDHq6UlHan2OAn65wOdanJ/C5U3C7IYteS/kY2IzAW7skbauKWwjidxC34R0E3n4E3rrO2tgzUWYO8ChwGYHXi8Bbg8DbmsDbp4L2W163LQm8Awm8/k5pxYj1s+qhE8gDOxF4v3bW7CnUtxyvA84i8HZy7XoE3sHNkKQnovQ/Qw6GXNxI2Y+BDd3BnWT/owm8fq7/jdw+YWu5GtlTe7oBmRcCMYG3KXLopmUE3k8IPN99x75AvAXlPocWo0pOUdqbIH6vEavoBOBCAu9L5Mn5rlb09ARygGUugVf8bd6NwI4E3gICb7LI411H4F3XSDsnIovc5cDnyAGNPwGHAh+6Mkch7tU3EcvjHkrdnA3TmrptxTbA48hi/TxwLUH85Cql5DeOByMu5c9cvecS+ZOQwyJ3EHhfAK8jJwUr5RbErfkRMh+le4dJWd5GDrW87z7Pvojb8D7E7fqlq797M/pvqK/PCeJ/uYNCpVwAfA95OMgjB3NayibI5/8Fsk/4FCsP9Zzt9lyp6L4BjLXN9WZUj4EDB9qpU5s69KUo7Y8xZpq1dmC15VAUpXHUklMURVFqFlVyiqIoSs2iSk5RFEWpWVTJKYqiKDWLKjlFURSlZulSpyuNMfNY+RaBcvQBPm0kv5qobC2js8q2nbW2Z9PFFEWpJl3qjSfW2o0ayzfGTO2sx7pVtpbRWWUzxuhvWRSlC6DuSkVRFKVmUSWnKIqi1Cy1puTGVluARlDZWkZnla2zyqUoSoIudfBEURRFUZpDrVlyiqIoirKCLqfkjDGbG2OeNMa8aYx5wxjz+zJljDHmamPMu8aY6caY73WQbDcZYz4xxrzeQP4gY0xsjKlzf+d1hFwVylaVOXN9/9IY847rO1smf7gxZl5i3kZ0Itm6G2PudPkvGmNSHSWboihN0+WUHLAUON1auyOwB3CiMWbHkjL7IyExtgFGAn/vINnGA79soswz1toB7u/CDpCpyHgal60qc2aMWRO4xvW/I3B4mc8T4M7EvN3QiWQ7Bphvre0PXIGEXVEUpZPQ5ZSctXaOtfYVd/0lEoNo05JiBwG3WOEFoLcxpt3jVVlrn0bib3U6KpCtKnMG7Aa8a61931r7X+AOJ0tnoBLZDgJCd30PsK8xplxUZ0VRqkCXU3JJnGvou8CLJVmbImHhi8xiVUVYLX5ojHnNGPOQMWanaguToFpzVmm/v3Fu1HuMMZt3gFxQmWwrylhrlyKBJDfsEOkURWmSLqvkjDE9gH8Ap1prv6i2PBXyCtDPWrsr8DdgcnXF6TLcD6SstbsAj7HSclIURWmULqnkjDHdEAV3m7W2XOj1j4Dk0/5mLq2qWGu/sNYudNcPAt2MMX2qLFaRas1Zk/1aaz+z1n7jbm8Avt8BclUkW7KMMWYtwAM+6xDpFEVpki6n5Nx+x43AW9bayxsodh9wlDsxuAcQW2vndJiQDWCM2aS4X2OM2Q2Z/86yIFZrzl4GtjHGbGmMWRs4zMmygpK9wV8h+7AdQZOyufuMu/4t8ITVH58qSqehS72g2fEj4EggMsbUubSzgS0ArLXXAQ8CBwDvAl8BR3eEYMaY24FBQB9jzCzgfKBbQq7fAscbY5YCi4HDOmpBrEC2qsyZtXapMeYk4BFgTeAma+0bxpgLganW2vuAU4wxv0JO1n4ODO9Est0I3GqMedfJdlhHyKYoSmXoG08URVGUmqXLuSsVRVEUpVJUySmKoig1iyo5RVEUpWZRJacoiqLULKrkFEVRlJqlSSVnjFnm3vz+hnsd1enGmEbrGWNSxbfdG2MGGGMOaI2QxpgexpjrjTHvGWOmGWOmGGN2b0E7NzTw8t/mtlOMJvCqe0P908aYAyuoN9wYM8ZdD2mNLMaY/dxcRO7fnybypji5im/t/7ZLb/CN+caYs1z6O8aYX7RUrmaOobcx5oRG8k8xxrxljLmtI+SpFPf9HtrBfQbGmD92ZJ+KUgtU8ju5xdbaAQBusZwI9EJ+Z1UJA4CByO+wWsoNwAfANtba5caYLZG3wjcLa21bhmh5xlp7IIgiByYbYxZba/9VYf0hwAPAmy3s/1NgsLV2tjFmZ+S3XMn3Kh5hrZ1aUmfFG/ONMYchb8w/1Cnbw4CdgL7A48aYba21y1ooW6X0Bk4Arm0g/wTgZ9baWclEY8xa7j2R1SIFDEX+L7Q57oUBxlq7vD3aV5TVCmtto3/AwpL7rZC3dBjkB7J/Qd4MMR34H1cmBbwOrA18CMwD6oBDkTe7Pw+8CvwfsF0T/W+NKLg1G8g/zfX1OvIeS4D1gTzwmks/1KVPAQYWxwWMdmVeADZ26Rshrwx72f39qEyfg4AHStJ+B0xqrA3kR8xjgD2RHw5/4OZla+BYV/Y1V3e9pj6bRN/Gtde9dJwl5R4Bfuiu10IUpQHOAs4qV670u+A+7zeAx91nOQV4H/iVK7MOcDMQuc/4Jy59J+AlN97pSEifO5AfxdcBfynp6zrgv66dPwABcCvwHHB7I3O8IfCok/EGYAbQB/edTLT/RyBIfMceBqYBzwDbu/TxwNXI9/R94Lcu/QXkRcx1wB9K5L4mMReTkB+QF78foxv5zqaAd4BbnOz9gHOAfwPPujH/0ZU9BXk4mg7cUen3RP/0b3X8a7pAiZJzaQuAjZG4Y6NcWndgKrBlckHBLeyJur2Atdz1z4B/uOu+wINl+voVTnmUyfu+WwTXB3q4xeG7wG+AcYlynvt3CiuVnEUsIYBLE+OYCOzlrrdAXh9W2u8gVlVyA4plG2ojORduAf1tov6GieuLgJMr/hDlTSqPJ+6nuHmpA85l5Y/+Xwc2S5R7D1EAY4BhifQbk7Il0i2wv7uehCiTbsCuQJ1LP52VC/v2yEPOOsgLqY9w6WsD61KieMr0VwD6uOsAUULrNjHHVwPnueu0k7kpJfcvxEsAsDvyaq7iZ3Q34tbfEQm7U/bzT7R7GE5hI0r9BXd9M/ALGv7OpoDlwB4l3+31kP8z77JSyc1m5QNN745YKPRP/7rqX2tf6/VzYBdjzG/dvYc8of+7kToeEBpjtkEWoOKrpWYjr5VqDnshCnARgDHmn8DeyFP5ZcaYPyOL0TNl6v4XcReCLJ77ueufATsmQoL1Msb0sO7Fyo2QjCFWto0m6u9sjLkIceH1QKypJnHhev6MfBZFjrDWfmSM6YlYO0ciFkJr+S8ytyAL8DfW2iXGmAhZpEE+k78BWGvfNsbMALZFrPdzjDGbAf+01v7HND/s2n3W2sXuuqE5/jHwa9d/3hgzv7EGXZ09gbsTbXVPFJlsxW34pjFm4wpkfAY41bmA3wQ2cO/e/CFigRUt/tLv7H3ADCux/HBpk6y1X7lyyXdmTgduM8ZMRiNZKEqjNFvJGWO2ApYBnyAL+8nW2kdKyqQaaeJPwJPW2v/nyk1poss3gF2NMWvaCveIrLX/NsZ8D1GaFxlj/mVXjcK9xFpbfKfZMlbOxRrI0/TXlfSV4LusfHFw2TaaWNTHA0Osta8ZY4Yj1kKjOIUxCTjKWvteMd1a+5H790tjzETErXgLK9+YP8vUf2N+pREIknO2HPjG9bPctdcg1tqJxpgXEevqQWPM/yAuwOawKHHd3DleSv2DVusk2llg3b5zGb5JXDepld3DRW8kCvvTwLeAQxCPyJdNfAcWNZaZII0o88HIg4Nvq7tHqSidlmb9hMAYsxGyVzLGLXaPIC8c7ubytzXGrF9S7UugZ+LeY+UCOrypPt3iPRW4wG3IF0+3pZGn5iHGmPVcv/8PeMYY0xf4ylo7AdlD+l4zhvkocHJizAOaqmCM2QVxC17TjDZK56UnMMfN5RGJuv/PGHNJmT57I/uOWWvtc4n0tYwL3+PaOhBxU0LDb8y/DzjMnb7cErHGX2pq3A3wTFF+Y8y2iCvxHfdw9L619mrgXmCXMnPQHBqa46eRQyEYY/YHNnDpHwPfNsZsaIzpjswLVmIRfmCMOdjVMcaYXZvouym5XwBOdbI8g7hGi96Est/ZMm087cqt6yzywU6+NYDNrbVPAmci/5+a8hIoympLJUpuXXcMvXjY4FHgApd3A+KSecXITwauZ1Xr8EnErVRnjDkU2f+6xBjzarKsMaavMaahE5gjkD3Ad10/44FPrLWvuOuXkOjgN1hrXwV84CUjUQrOR/a4KuUUYKCRKNRvAsc1UG5v435CgCi3U+zKk5WVtHEHcIZrY2tESb6IHKx4O1Fua6BcUNiTgP7Aeab+TwW6A48YY6Yje3IfAeNcnRuBDY28Mf80IAtgrX0DuAv5LB8GTqzUai7DtcAazoV5JzDcSiy4Q4DX3WeyM3CLtfYz4DljzOvGmL80s5+G5vgC4Mfu+/prZE8Qa+0S4ELku/IY9ef4COAYY8xriOfgoCb6ng4sM/KTmj+UyX8G2Xd+FwmU+y2XRiPf2Xq4cnciB5EeQg7XgBz2muDm91XgamvtgibkVZTVFo1C0MkxxkxATvDNq7YsXRFjTAE5bPRptWVRFKXj6Yrx5FYrrLXDqi2DoihKV0UtOUVRFKVm0XdXKoqiKDWLKjlFURSlZlElpyiKotQsquQURVGUmkWVnKIoilKzqJJTFEVRapb/D5aEb/r7aT2bAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "hclust = delta.Clustering(distances)\n",
    "delta.Dendrogram(hclust).show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "source": [
    "### Flachklopfen"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {
    "slideshow": {
     "slide_type": "-"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "3 clusters of 9 documents (ground truth: 3 groups):\n",
      "{1: ['Marlitt: Geheimnis der alten Mams…', 'Marlitt: Frau mit den Karfunkelst…',\n",
      "     'Marlitt: Heideprinzesschen'],\n",
      " 2: ['Fontane: Stechlin', 'Fontane: Effi Briest'],\n",
      " 3: ['Raabe: Im alten Eisen', 'Raabe: Stopfkuchen',\n",
      "     'Raabe: Chronik der Sperlingsgas…', 'Fontane: Frau Jenny Treibel']}\n",
      "\n",
      "Cluster Errors         1.000000\n",
      "Adjusted Rand Index    0.642857\n",
      "Homogeneity            0.772507\n",
      "Completeness           0.800000\n",
      "V Measure              0.786013\n",
      "Purity                 0.888889\n",
      "Entropy                0.227493\n",
      "dtype: float64\n"
     ]
    }
   ],
   "source": [
    "fclust = hclust.fclustering()\n",
    "print(fclust.describe())\n",
    "print(fclust.evaluate())"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "### Deltafunktionen\n",
    "\n",
    "Cosine Delta kann jeder. Was gibts denn sonst so?"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {
    "slideshow": {
     "slide_type": "-"
    }
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<h4>Delta Functions</h4><ol><li><code>Linear Delta</code> = linear</li><li><code>distance</code> = linear2</li><li><code>_classic_delta</code> = burrows2</li><li><code>manhattan</code> <em>Manhattan Distance</em></li><li><code>euclidean</code> <em>Euclidean Distance</em></li><li><code>sqeuclidean</code> <em>Squared Euclidean Distance</em></li><li><code>cosine</code> <em>Cosine Distance</em></li><li><code>canberra</code> <em>Canberra Distance</em></li><li><code>braycurtis</code> <em>Bray-Curtis Distance</em></li><li><code>correlation</code> <em>Correlation Distance</em></li><li><code>chebyshev</code> <em>Chebyshev Distance</em></li><li><code>burrows</code> <em>Burrows' Delta</em> = manhattan-z_score</li><li><code>quadratic</code> <em>Quadratic Delta</em> = sqeuclidean-z_score</li><li><code>eder</code> <em>Eder's Delta</em> = manhattan-z_score-eder_std</li><li><code>eder_simple</code> <em>Eder's Simple</em> = manhattan-sqrt</li><li><code>cosine_delta</code> <em>Cosine Delta</em> = cosine-z_score</ol><h4>Normalizations</h4><ol><li>z_score (Z-Score)</li><li>eder_std</li><li>binarize</li><li>length_normalized</li><li>diversity_scaled</li><li>sqrt</li><li>clamp</li><li>ternarize"
      ],
      "text/plain": [
       "<delta.deltas._FunctionRegistry at 0x7f12ebed1eb0>"
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "delta.functions"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "source": [
    "Da fehlt Everts Ternarisierung! Erstmal eine Normalisierungsfunktion schreiben:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {
    "slideshow": {
     "slide_type": "fragment"
    }
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Registering ternarize as ternarize, replacing existing function with this name\n"
     ]
    }
   ],
   "source": [
    "@delta.normalization\n",
    "def ternarize(z):    \n",
    "    result = pd.DataFrame(np.zeros_like(z), index=z.index, columns=z.columns)\n",
    "    result[z < -2/3] = -1\n",
    "    result[z > +2/3] = +1\n",
    "    return result"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "fragment"
    }
   },
   "source": [
    "Nun in eine Deltafunktion einbauen."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {
    "slideshow": {
     "slide_type": "fragment"
    }
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<code>ternarized_delta</code> <em>Ternarized Delta</em> = euclidean-z_score-ternarize"
      ],
      "text/plain": [
       "<delta.deltas.CompositeDeltaFunction at 0x7f12de64bdc0>"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "delta.deltas.CompositeDeltaFunction(descriptor=\"euclidean-z_score-ternarize\", name=\"ternarized_delta\", title=\"Ternarized Delta\")\n",
    "#  heute schrübe ich das nicht mehr so -----------´"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/tv/git/pydelta/delta/graphics.py:88: MatplotlibDeprecationWarning: Passing the pad parameter of tight_layout() positionally is deprecated since Matplotlib 3.3; the parameter will become keyword-only two minor releases later.\n",
      "  plt.tight_layout(2)\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "distances_t = delta.functions.ternarized_delta(corpus)\n",
    "hclust_t = delta.Clustering(distances_t)\n",
    "delta.Dendrogram(hclust_t).show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "F-Ratio         0.703738\n",
      "Fisher's LD     7.042422\n",
      "Simple Score    1.537760\n",
      "dtype: float64\n"
     ]
    }
   ],
   "source": [
    "print(distances_t.evaluate())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {
    "slideshow": {
     "slide_type": "-"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Metadata(complete=False, corpus='test/corpus3/', delta='ternarized_delta', delta_descriptor='euclidean-z_score-ternarize', delta_title='Ternarized Delta', features='words', frequencies=False, lower_case=False, normalization=('z_score', 'ternarize'), ordered=True, words=2500)"
      ]
     },
     "execution_count": 13,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "distances_t.metadata"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<AxesSubplot:>"
      ]
     },
     "execution_count": 14,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 648x144 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "delta.graphics.spikeplot(delta.functions.ternarize(delta.functions.z_score(corpus.top_n(1000))), docs=[0,1,3])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "celltoolbar": "Slideshow",
  "kernelspec": {
   "display_name": "pydelta",
   "language": "python",
   "name": "pydelta"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.9.5"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}