Machine_Learning_02.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {
    "_cell_guid": "dcc6c619-c02a-4bea-b322-4f32779ae672",
    "_uuid": "819c6b57207e720308ea29491d85a6e8af8c253e"
   },
   "source": [
    "# DATA SCIENTIST\n",
    "**In this tutorial, I only explain you what you need to be a data scientist neither more nor less.**\n",
    "\n",
    "**Content:**\n",
    "\n",
    "1. [Machine Learning](#1)\n",
    "    1. [Supervised Learning](#2)\n",
    "        1. [EDA(Exploratory Data Analysis)](#3)\n",
    "        1. [K-Nearest Neighbors (KNN)](#4)\n",
    "        1. [Regression](#5)\n",
    "        1. [Cross Validation (CV)](#6)\n",
    "        1. [ROC Curve](#7)\n",
    "        1. [Hyperparameter Tuning](#8)\n",
    "        1. [Pre-procesing Data](#9)\n",
    "    1. [Unsupervised Learning](#10)\n",
    "        1. [Kmeans Clustering](#11)\n",
    "        1. [Evaluation of Clustering](#12)\n",
    "        1. [Standardization](#13)\n",
    "        1. [Hierachy](#14)\n",
    "        1. [T - Distributed Stochastic Neighbor Embedding (T - SNE)](#15)\n",
    "        1. [Principle Component Analysis (PCA)](#16)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 196,
   "metadata": {
    "_cell_guid": "5ee3a7aa-eca4-411b-9f84-d14c09e13730",
    "_uuid": "2b90d6250c8f9c2c302c849bffa132bd3483e893"
   },
   "outputs": [],
   "source": [
    "# This Python 3 environment comes with many helpful analytics libraries installed\n",
    "\n",
    "import numpy as np # linear algebra\n",
    "import pandas as pd # data processing, CSV file I/O (e.g. pd.read_csv)\n",
    "import matplotlib.pyplot as plt\n",
    "import seaborn as sns"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 197,
   "metadata": {
    "_cell_guid": "32af03f6-41be-41ec-9023-8cd519040984",
    "_uuid": "a9c5426e9e5cef81c1e1639ebe57e9b45dfd2c43"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "['seaborn-dark', 'seaborn-darkgrid', 'seaborn-ticks', 'fivethirtyeight', 'seaborn-whitegrid', 'classic', '_classic_test', 'fast', 'seaborn-talk', 'seaborn-dark-palette', 'seaborn-bright', 'seaborn-pastel', 'grayscale', 'seaborn-notebook', 'ggplot', 'seaborn-colorblind', 'seaborn-muted', 'seaborn', 'Solarize_Light2', 'seaborn-paper', 'bmh', 'tableau-colorblind10', 'seaborn-white', 'dark_background', 'seaborn-poster', 'seaborn-deep']\n"
     ]
    }
   ],
   "source": [
    "# read csv (comma separated value) into data\n",
    "data = pd.read_csv('/Users/garvitkhurana/Desktop/Projects/Data_Science/column_2C_weka.csv')\n",
    "print(plt.style.available) # look at available plot styles\n",
    "plt.style.use('ggplot')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "_cell_guid": "89a724e2-426d-427b-9107-06835010cf59",
    "_uuid": "f4fa42e9a6cf069d54459be42a1726ab03c2f1e5"
   },
   "source": [
    "<a id=\"1\"></a> <br>\n",
    "# 6. MACHINE LEARNING (ML)\n",
    "In python there are some ML libraries like sklearn, keras or tensorflow. We will use sklearn.\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "_cell_guid": "10f3b719-c44b-451a-b464-0adf4e1e1522",
    "_uuid": "f6a2b205e1e3fc647bc2ee88d702b8572bf1cc75"
   },
   "source": [
    "<a id=\"2\"></a> <br>\n",
    "## A. SUPERVISED LEARNING\n",
    "* Supervised learning: It uses data that has labels. Example, there are orthopedic patients data that have labels *normal* and *abnormal*.\n",
    "    * There are features(predictor variable) and target variable. Features are like *pelvic radius* or *sacral slope*(If you have no idea what these are like me, you can look images in google like what I did :) )Target variables are labels *normal* and *abnormal*\n",
    "    * Aim is that as given features(input) predict whether target variable(output) is *normal* or *abnormal*\n",
    "    * Classification: target variable consists of categories like normal or abnormal\n",
    "    * Regression: target variable is continious like stock market\n",
    "    * If these explanations are not enough for you, just google them. However, be careful about terminology: features = predictor variable = independent variable = columns = inputs. target variable = responce variable = class = dependent variable = output = result"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "_cell_guid": "65e897a1-8259-44c5-9cb7-e5e653f9032d",
    "_uuid": "a0e671bf2ef8dbe81da2705ad70b69401bb7af16"
   },
   "source": [
    "<a id=\"3\"></a> <br>\n",
    "### EXPLORATORY DATA ANALYSIS (EDA)\n",
    "* In order to make something in data, as you know you need to explore data. Detailed exploratory data analysis is in my Data Science Tutorial for Beginners\n",
    "* I always start with *head()* to see features that are *pelvic_incidence,\tpelvic_tilt numeric,\tlumbar_lordosis_angle,\tsacral_slope,\tpelvic_radius* and \t*degree_spondylolisthesis* and target variable that is *class*\n",
    "* head(): default value of it shows first 5 rows(samples). If you want to see for example 100 rows just write head(100)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 198,
   "metadata": {
    "_cell_guid": "c1ecd622-67cc-485f-bfa7-8c682d30a5eb",
    "_uuid": "9a5993f4962882e1156f2062b7abf706a0739d51"
   },
   "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>pelvic_incidence</th>\n",
       "      <th>pelvic_tilt numeric</th>\n",
       "      <th>lumbar_lordosis_angle</th>\n",
       "      <th>sacral_slope</th>\n",
       "      <th>pelvic_radius</th>\n",
       "      <th>degree_spondylolisthesis</th>\n",
       "      <th>class</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>63.027818</td>\n",
       "      <td>22.552586</td>\n",
       "      <td>39.609117</td>\n",
       "      <td>40.475232</td>\n",
       "      <td>98.672917</td>\n",
       "      <td>-0.254400</td>\n",
       "      <td>Abnormal</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>39.056951</td>\n",
       "      <td>10.060991</td>\n",
       "      <td>25.015378</td>\n",
       "      <td>28.995960</td>\n",
       "      <td>114.405425</td>\n",
       "      <td>4.564259</td>\n",
       "      <td>Abnormal</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>68.832021</td>\n",
       "      <td>22.218482</td>\n",
       "      <td>50.092194</td>\n",
       "      <td>46.613539</td>\n",
       "      <td>105.985135</td>\n",
       "      <td>-3.530317</td>\n",
       "      <td>Abnormal</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>69.297008</td>\n",
       "      <td>24.652878</td>\n",
       "      <td>44.311238</td>\n",
       "      <td>44.644130</td>\n",
       "      <td>101.868495</td>\n",
       "      <td>11.211523</td>\n",
       "      <td>Abnormal</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>49.712859</td>\n",
       "      <td>9.652075</td>\n",
       "      <td>28.317406</td>\n",
       "      <td>40.060784</td>\n",
       "      <td>108.168725</td>\n",
       "      <td>7.918501</td>\n",
       "      <td>Abnormal</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   pelvic_incidence  pelvic_tilt numeric  lumbar_lordosis_angle  sacral_slope  \\\n",
       "0         63.027818            22.552586              39.609117     40.475232   \n",
       "1         39.056951            10.060991              25.015378     28.995960   \n",
       "2         68.832021            22.218482              50.092194     46.613539   \n",
       "3         69.297008            24.652878              44.311238     44.644130   \n",
       "4         49.712859             9.652075              28.317406     40.060784   \n",
       "\n",
       "   pelvic_radius  degree_spondylolisthesis     class  \n",
       "0      98.672917                 -0.254400  Abnormal  \n",
       "1     114.405425                  4.564259  Abnormal  \n",
       "2     105.985135                 -3.530317  Abnormal  \n",
       "3     101.868495                 11.211523  Abnormal  \n",
       "4     108.168725                  7.918501  Abnormal  "
      ]
     },
     "execution_count": 198,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# to see features and target variable\n",
    "data.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 199,
   "metadata": {
    "_cell_guid": "1631690c-bb9d-4460-a7d9-a335aa914b4f",
    "_uuid": "b7b9addc824de1a35b67d96d3092ffcb10869947"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'pandas.core.frame.DataFrame'>\n",
      "RangeIndex: 310 entries, 0 to 309\n",
      "Data columns (total 7 columns):\n",
      "pelvic_incidence            310 non-null float64\n",
      "pelvic_tilt numeric         310 non-null float64\n",
      "lumbar_lordosis_angle       310 non-null float64\n",
      "sacral_slope                310 non-null float64\n",
      "pelvic_radius               310 non-null float64\n",
      "degree_spondylolisthesis    310 non-null float64\n",
      "class                       310 non-null object\n",
      "dtypes: float64(6), object(1)\n",
      "memory usage: 17.0+ KB\n"
     ]
    }
   ],
   "source": [
    "# Well know question is is there any NaN value and length of this data so lets look at info\n",
    "data.info()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 200,
   "metadata": {
    "_cell_guid": "137897ca-b519-4ac3-afdd-6c4136447e39",
    "_uuid": "69d132068cce6a915aac7678b4e9fbcf3e365643"
   },
   "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>pelvic_incidence</th>\n",
       "      <th>pelvic_tilt numeric</th>\n",
       "      <th>lumbar_lordosis_angle</th>\n",
       "      <th>sacral_slope</th>\n",
       "      <th>pelvic_radius</th>\n",
       "      <th>degree_spondylolisthesis</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>count</th>\n",
       "      <td>310.000000</td>\n",
       "      <td>310.000000</td>\n",
       "      <td>310.000000</td>\n",
       "      <td>310.000000</td>\n",
       "      <td>310.000000</td>\n",
       "      <td>310.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>mean</th>\n",
       "      <td>60.496653</td>\n",
       "      <td>17.542822</td>\n",
       "      <td>51.930930</td>\n",
       "      <td>42.953831</td>\n",
       "      <td>117.920655</td>\n",
       "      <td>26.296694</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>std</th>\n",
       "      <td>17.236520</td>\n",
       "      <td>10.008330</td>\n",
       "      <td>18.554064</td>\n",
       "      <td>13.423102</td>\n",
       "      <td>13.317377</td>\n",
       "      <td>37.559027</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>min</th>\n",
       "      <td>26.147921</td>\n",
       "      <td>-6.554948</td>\n",
       "      <td>14.000000</td>\n",
       "      <td>13.366931</td>\n",
       "      <td>70.082575</td>\n",
       "      <td>-11.058179</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25%</th>\n",
       "      <td>46.430294</td>\n",
       "      <td>10.667069</td>\n",
       "      <td>37.000000</td>\n",
       "      <td>33.347122</td>\n",
       "      <td>110.709196</td>\n",
       "      <td>1.603727</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>50%</th>\n",
       "      <td>58.691038</td>\n",
       "      <td>16.357689</td>\n",
       "      <td>49.562398</td>\n",
       "      <td>42.404912</td>\n",
       "      <td>118.268178</td>\n",
       "      <td>11.767934</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>75%</th>\n",
       "      <td>72.877696</td>\n",
       "      <td>22.120395</td>\n",
       "      <td>63.000000</td>\n",
       "      <td>52.695888</td>\n",
       "      <td>125.467674</td>\n",
       "      <td>41.287352</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>max</th>\n",
       "      <td>129.834041</td>\n",
       "      <td>49.431864</td>\n",
       "      <td>125.742385</td>\n",
       "      <td>121.429566</td>\n",
       "      <td>163.071041</td>\n",
       "      <td>418.543082</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "       pelvic_incidence  pelvic_tilt numeric  lumbar_lordosis_angle  \\\n",
       "count        310.000000           310.000000             310.000000   \n",
       "mean          60.496653            17.542822              51.930930   \n",
       "std           17.236520            10.008330              18.554064   \n",
       "min           26.147921            -6.554948              14.000000   \n",
       "25%           46.430294            10.667069              37.000000   \n",
       "50%           58.691038            16.357689              49.562398   \n",
       "75%           72.877696            22.120395              63.000000   \n",
       "max          129.834041            49.431864             125.742385   \n",
       "\n",
       "       sacral_slope  pelvic_radius  degree_spondylolisthesis  \n",
       "count    310.000000     310.000000                310.000000  \n",
       "mean      42.953831     117.920655                 26.296694  \n",
       "std       13.423102      13.317377                 37.559027  \n",
       "min       13.366931      70.082575                -11.058179  \n",
       "25%       33.347122     110.709196                  1.603727  \n",
       "50%       42.404912     118.268178                 11.767934  \n",
       "75%       52.695888     125.467674                 41.287352  \n",
       "max      121.429566     163.071041                418.543082  "
      ]
     },
     "execution_count": 200,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data.describe()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "_cell_guid": "3776dd3d-d0aa-419e-b788-e75454e94b86",
    "_uuid": "c8961b1f3a3a73547a0b7d27955f9844f6ad43eb"
   },
   "source": [
    "pd.plotting.scatter_matrix:\n",
    "* green: *normal* and red: *abnormal*\n",
    "* c:  color\n",
    "* figsize: figure size\n",
    "* diagonal: histohram of each features\n",
    "* alpha: opacity\n",
    "* s: size of marker\n",
    "* marker: marker type "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 201,
   "metadata": {
    "_cell_guid": "fb106765-bb47-452b-8d6e-3578b479873c",
    "_uuid": "5dc0763dde8b2638a5289f0b4496f384f85aca85"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x1a1e222358>"
      ]
     },
     "execution_count": 201,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 576x648 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "d=data.corr()\n",
    "plt.figure(figsize=(8,9))\n",
    "sns.heatmap(data=d,xticklabels=d.columns,yticklabels=d.columns,fmt=\"0.2f\",annot=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 202,
   "metadata": {
    "_cell_guid": "36243fa5-1fa6-4f8b-bc16-43449b0dc898",
    "_uuid": "e1bb9fd338e6900888e2e4717f54be46cee848a2"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Abnormal    210\n",
       "Normal      100\n",
       "Name: class, dtype: int64"
      ]
     },
     "execution_count": 202,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "sns.countplot(x=\"class\", data=data)\n",
    "data['class'].value_counts()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "_cell_guid": "24a5d90f-3e7d-4733-a6f9-ff4f51145155",
    "_uuid": "9263c479815bb729dad40bf01b68aa18a3c946ac"
   },
   "source": [
    "<a id=\"4\"></a> <br>\n",
    "###  K-NEAREST NEIGHBORS (KNN)\n",
    "* KNN: Look at the K closest labeled data points\n",
    "* Classification method.\n",
    "* First we need to train our data. Train = fit\n",
    "* fit(): fits the data, train the data.\n",
    "* predict(): predicts the data\n",
    "<br> If you do not understand what is KNN, look at youtube there are videos like 4-5 minutes. You can understand better with it.\n",
    "<br> Lets learn how to implement it with sklearn\n",
    "* x: features\n",
    "* y: target variables(normal, abnormal)\n",
    "* n_neighbors: K. In this example it is 3. it means that Look at the 3 closest labeled data points\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 234,
   "metadata": {
    "_cell_guid": "c717491d-2bd5-4dc7-ac13-b9f581b1cddd",
    "_uuid": "854f0a3898a928640b9714fcd584e48c9b377f9f"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Prediction: ['Abnormal' 'Abnormal' 'Abnormal' 'Abnormal' 'Abnormal' 'Normal' 'Normal'\n",
      " 'Abnormal' 'Normal' 'Abnormal' 'Abnormal' 'Abnormal' 'Abnormal'\n",
      " 'Abnormal' 'Abnormal' 'Abnormal' 'Abnormal' 'Abnormal' 'Abnormal'\n",
      " 'Abnormal' 'Normal' 'Normal' 'Normal' 'Abnormal' 'Abnormal' 'Abnormal'\n",
      " 'Abnormal' 'Abnormal' 'Abnormal' 'Abnormal' 'Normal' 'Abnormal' 'Normal'\n",
      " 'Normal' 'Abnormal' 'Abnormal' 'Abnormal' 'Abnormal' 'Abnormal'\n",
      " 'Abnormal' 'Abnormal' 'Abnormal' 'Abnormal' 'Abnormal' 'Abnormal'\n",
      " 'Abnormal' 'Abnormal' 'Abnormal' 'Abnormal' 'Abnormal' 'Abnormal'\n",
      " 'Abnormal' 'Abnormal' 'Abnormal' 'Abnormal' 'Abnormal' 'Abnormal'\n",
      " 'Abnormal' 'Abnormal' 'Abnormal' 'Abnormal' 'Abnormal' 'Abnormal'\n",
      " 'Abnormal' 'Abnormal' 'Abnormal' 'Abnormal' 'Abnormal' 'Normal'\n",
      " 'Abnormal' 'Abnormal' 'Abnormal' 'Abnormal' 'Abnormal' 'Abnormal'\n",
      " 'Abnormal' 'Abnormal' 'Abnormal' 'Abnormal' 'Abnormal' 'Abnormal'\n",
      " 'Abnormal' 'Abnormal' 'Abnormal' 'Abnormal' 'Abnormal' 'Abnormal'\n",
      " 'Abnormal' 'Abnormal' 'Abnormal' 'Abnormal' 'Abnormal' 'Abnormal'\n",
      " 'Abnormal' 'Abnormal' 'Abnormal' 'Abnormal' 'Abnormal' 'Abnormal'\n",
      " 'Abnormal' 'Abnormal' 'Abnormal' 'Abnormal' 'Abnormal' 'Abnormal'\n",
      " 'Abnormal' 'Abnormal' 'Abnormal' 'Abnormal' 'Abnormal' 'Abnormal'\n",
      " 'Abnormal' 'Abnormal' 'Abnormal' 'Abnormal' 'Abnormal' 'Abnormal'\n",
      " 'Abnormal' 'Abnormal' 'Abnormal' 'Abnormal' 'Abnormal' 'Abnormal'\n",
      " 'Abnormal' 'Abnormal' 'Abnormal' 'Abnormal' 'Abnormal' 'Abnormal'\n",
      " 'Abnormal' 'Abnormal' 'Normal' 'Abnormal' 'Abnormal' 'Abnormal'\n",
      " 'Abnormal' 'Abnormal' 'Abnormal' 'Abnormal' 'Abnormal' 'Abnormal'\n",
      " 'Abnormal' 'Abnormal' 'Abnormal' 'Abnormal' 'Abnormal' 'Abnormal'\n",
      " 'Abnormal' 'Abnormal' 'Abnormal' 'Abnormal' 'Abnormal' 'Abnormal'\n",
      " 'Abnormal' 'Abnormal' 'Abnormal' 'Abnormal' 'Abnormal' 'Abnormal'\n",
      " 'Abnormal' 'Abnormal' 'Abnormal' 'Abnormal' 'Abnormal' 'Abnormal'\n",
      " 'Abnormal' 'Abnormal' 'Abnormal' 'Abnormal' 'Abnormal' 'Abnormal'\n",
      " 'Normal' 'Abnormal' 'Abnormal' 'Abnormal' 'Abnormal' 'Abnormal'\n",
      " 'Abnormal' 'Abnormal' 'Abnormal' 'Abnormal' 'Abnormal' 'Abnormal'\n",
      " 'Abnormal' 'Abnormal' 'Abnormal' 'Abnormal' 'Abnormal' 'Abnormal'\n",
      " 'Abnormal' 'Abnormal' 'Abnormal' 'Abnormal' 'Abnormal' 'Abnormal'\n",
      " 'Abnormal' 'Abnormal' 'Abnormal' 'Abnormal' 'Abnormal' 'Abnormal'\n",
      " 'Abnormal' 'Abnormal' 'Abnormal' 'Abnormal' 'Abnormal' 'Abnormal'\n",
      " 'Abnormal' 'Abnormal' 'Abnormal' 'Abnormal' 'Normal' 'Normal' 'Normal'\n",
      " 'Normal' 'Normal' 'Normal' 'Normal' 'Normal' 'Normal' 'Normal' 'Normal'\n",
      " 'Abnormal' 'Normal' 'Normal' 'Normal' 'Normal' 'Normal' 'Normal' 'Normal'\n",
      " 'Normal' 'Normal' 'Normal' 'Normal' 'Abnormal' 'Normal' 'Normal' 'Normal'\n",
      " 'Normal' 'Normal' 'Normal' 'Abnormal' 'Normal' 'Normal' 'Abnormal'\n",
      " 'Normal' 'Normal' 'Normal' 'Normal' 'Abnormal' 'Normal' 'Normal' 'Normal'\n",
      " 'Normal' 'Normal' 'Abnormal' 'Normal' 'Normal' 'Normal' 'Normal' 'Normal'\n",
      " 'Normal' 'Normal' 'Normal' 'Normal' 'Normal' 'Normal' 'Normal' 'Normal'\n",
      " 'Normal' 'Normal' 'Normal' 'Normal' 'Normal' 'Normal' 'Normal' 'Normal'\n",
      " 'Normal' 'Normal' 'Normal' 'Normal' 'Normal' 'Normal' 'Normal' 'Normal'\n",
      " 'Normal' 'Normal' 'Normal' 'Normal' 'Normal' 'Normal' 'Normal' 'Normal'\n",
      " 'Abnormal' 'Normal' 'Normal' 'Normal' 'Abnormal' 'Normal' 'Normal'\n",
      " 'Normal' 'Normal' 'Normal' 'Normal' 'Normal' 'Abnormal' 'Normal'\n",
      " 'Abnormal' 'Normal' 'Normal']\n"
     ]
    }
   ],
   "source": [
    "# KNN\n",
    "from sklearn.neighbors import KNeighborsClassifier\n",
    "knn = KNeighborsClassifier(n_neighbors = 3)\n",
    "x,y = data.loc[:,data.columns != 'class'], data.loc[:,'class']\n",
    "knn.fit(x,y)\n",
    "prediction = knn.predict(x)\n",
    "print('Prediction: {}'.format(prediction))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "_cell_guid": "b5d85f4e-ab30-4c49-b2bc-6265b6baea9d",
    "_uuid": "6d9c3eacd279ddf7c0ef33b9e1814cd549d0feaa"
   },
   "source": [
    "* Well, we fit the data and predict it with KNN. \n",
    "* So, do we predict correct or what is our accuracy or the accuracy is best metric to evaluate our result? Lets give answer of this questions\n",
    "<br> Measuring model performance:\n",
    "* Accuracy which is fraction of correct predictions is commonly used metric. We will use it know but there is another problem\n",
    "\n",
    "<br>As you see I train data with x (features) and again predict the x(features). Yes you are reading right but yes you are right again it is absurd :)\n",
    "\n",
    "<br>Therefore we need to split our data train and test sets.\n",
    "* train: use train set by fitting\n",
    "* test: make prediction on test set.\n",
    "* With train and test sets, fitted data and tested data are completely different\n",
    "* train_test_split(x,y,test_size = 0.3,random_state = 1)\n",
    "    * x: features\n",
    "    * y: target variables (normal,abnormal)\n",
    "    * test_size: percentage of test size. Example test_size = 0.3, test size = 30% and train size = 70%\n",
    "    * random_state: sets a seed. If this seed is same number, train_test_split() produce exact same split at each time\n",
    "* fit(x_train,y_train): fit on train sets\n",
    "* score(x_test,y_test)): predict and give accuracy on test sets"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 205,
   "metadata": {
    "_cell_guid": "79865658-c89f-43c8-a1a9-75acc4feab6a",
    "_uuid": "4702429fdfa62650937b09fde5a8fd3136da8c55"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "With KNN (K=3) accuracy is:  0.8709677419354839\n"
     ]
    }
   ],
   "source": [
    "# train test split\n",
    "from sklearn.model_selection import train_test_split\n",
    "x_train,x_test,y_train,y_test = train_test_split(x,y,test_size = 0.3,random_state = 54)\n",
    "knn = KNeighborsClassifier(n_neighbors = 3)\n",
    "x,y = data.loc[:,data.columns != 'class'], data.loc[:,'class']\n",
    "knn.fit(x_train,y_train)\n",
    "prediction = knn.predict(x_test)\n",
    "#print('Prediction: {}'.format(prediction))\n",
    "print('With KNN (K=3) accuracy is: ',knn.score(x_test,y_test)) # accuracy"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "_cell_guid": "a5665258-3f7f-435a-a634-49eb0c0d66e0",
    "_uuid": "544f51ef05efe0b3ae4b02da806778bcfa715f35"
   },
   "source": [
    "Accuracy is 87% so is it good ? I do not know actually, we will see at the end of tutorial.\n",
    "<br> Now the question is why we choose K = 3 or what value we need to choose K. The answer is in model complexity\n",
    "\n",
    "<br> Model complexity:\n",
    "* K has general name. It is called a hyperparameter. For now just know K is hyperparameter and we need to choose it that gives best performace. \n",
    "* Literature says if k is small, model is complex model can lead to overfit. It means that model memorizes the train sets and cannot predict test set with good accuracy.\n",
    "* If k is big, model that is less complex model can lead to underfit. \n",
    "* At below, I range K value from 1 to 25(exclude) and find accuracy for each K value. As you can see in plot, when K is 1 it memozize train sets and cannot give good accuracy on test set (overfit). Also if K is 18, model is lead to underfit. Again accuracy is not enough. However look at when K is 6(best performance), accuracy has highest value almost 88%. \n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 206,
   "metadata": {
    "_cell_guid": "db2c7062-ce1b-4e8b-9b2f-0ee0cd679a91",
    "_uuid": "18d8739373085a9964071f38b8f2adcb64f25491"
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 936x576 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Best accuracy is 0.8817204301075269 with K = 6\n"
     ]
    }
   ],
   "source": [
    "# Model complexity\n",
    "\n",
    "train_accuracy = []\n",
    "test_accuracy = []\n",
    "# Loop over different values of k\n",
    "for k in range (1,25):\n",
    "    # k from 1 to 25(exclude)\n",
    "    knn = KNeighborsClassifier(n_neighbors=k)\n",
    "    # Fit with knn\n",
    "    knn.fit(x_train,y_train)\n",
    "    #train accuracy\n",
    "    train_accuracy.append(knn.score(x_train, y_train))\n",
    "    # test accuracy\n",
    "    test_accuracy.append(knn.score(x_test, y_test))\n",
    "\n",
    "# Plot\n",
    "plt.figure(figsize=[13,8])\n",
    "plt.plot(test_accuracy, label = 'Testing Accuracy')\n",
    "plt.plot(train_accuracy, label = 'Training Accuracy')\n",
    "plt.legend()\n",
    "plt.title('-value VS Accuracy')\n",
    "plt.xlabel('Number of Neighbors')\n",
    "plt.ylabel('Accuracy')\n",
    "plt.savefig('graph.png')\n",
    "plt.show()\n",
    "print(\"Best accuracy is {} with K = {}\".format(np.max(test_accuracy),1+test_accuracy.index(np.max(test_accuracy))))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 231,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "With RF (max_depth=3) accuracy is:  0.8709677419354839\n"
     ]
    }
   ],
   "source": [
    "# RF\n",
    "from sklearn.model_selection import train_test_split\n",
    "x_train,x_test,y_train,y_test = train_test_split(x,y,test_size = 0.3,random_state = 93)\n",
    "\n",
    "from sklearn.ensemble import RandomForestClassifier\n",
    "rf = RandomForestClassifier(max_depth = 4)\n",
    "x,y = data.loc[:,data.columns != 'class'], data.loc[:,'class']\n",
    "rf.fit(x_train,y_train)\n",
    "prediction = rf.predict(x_test)\n",
    "print('With RF (max_depth=3) accuracy is: ',rf.score(x_test,y_test)) # accuracy"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 236,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "With LogisticRegressionaccuracy is:  0.8387096774193549\n"
     ]
    }
   ],
   "source": [
    "from sklearn.linear_model import LogisticRegression\n",
    "logreg = LogisticRegression()\n",
    "logreg.fit(x_train,y_train)\n",
    "prediction = logreg.predict(x_test)\n",
    "print('With LogisticRegressionaccuracy is: ',logreg.score(x_test,y_test)) # accuracy"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 237,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAxUAAAH0CAYAAABLmgAEAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADl0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uIDIuMi4yLCBodHRwOi8vbWF0cGxvdGxpYi5vcmcvhp/UCwAAIABJREFUeJzs3XlcVPX6B/DPLDALMywzw74qbqAiKi5p7qilmWaL1lXbVzPN6620brdfq5al93bbNLWsTLvlUrmVJblvuKHgDgiCbAMMOwzz/f2BjiCoKDADzOf9evWKmTlznucMCOc53+/zPRIhhAAREREREdEtkto7ASIiIiIiatlYVBARERERUYOwqCAiIiIiogZhUUFERERERA3CooKIiIiIiBqERQURERERETUIiwoiohYuKSkJEokEO3bssHcqRETkoFhUEBFRvSxYsAAqlQpGo7HO18eMGYPbb78dAGCxWDB//nx06dIFLi4ucHd3R7du3fDaa6/VK1ZsbCxkMhl69OjRaPkTEVHTYVFBRET18vDDDwMAvvnmm1qvXbhwARs3bsRTTz0FAHjzzTfxzjvv4JVXXsHRo0exc+dOzJkzB0VFRfWK9cUXX+DZZ59FUlISDhw40HgH0QDl5eX2ToGIqNliUUFEZCenT5+GRCLBrl27ajy/d+9eSCQSnDhxAgDw73//G5GRkdBoNPDx8cHEiRORnp5+zf1eazpUu3bt8MYbb1gfFxYWYvr06fD394darUb37t2xevXqa+5Xp9Phvvvuw+LFi2u9tnTpUmi1Wtx///0AgLVr1+Lxxx/HpEmTEBoais6dO2PChAlYsGDBDT+XgoICfP/993jqqacwceJELFq0qNY2hYWFmDFjBgIDA6FQKBASEoJ3333X+npmZiYeffRReHt7Q6lUomPHjli6dCkAICYmBhKJBKmpqTX2KZfL8dVXXwG48hl+9913GDVqFFxcXDBnzhwIIfDkk08iNDQUKpUKbdu2xZw5c1BWVlZjX1u2bMGAAQOgVqvh5uaGQYMG4ezZs9i6dStkMhlSUlJqbP/1119Dq9WioKDghp8PEVFzxKKCiMhO2rdvj759++Lrr7+u8fw333yD3r17o1OnTtbn5s+fj7i4OKxZswbnz5/HxIkTGxRbCIExY8bgyJEjWLVqFY4dO4Znn30WEydOxB9//HHN9z399NM4fvw4du/ebX3OYrFgyZIlmDx5MlQqFQDA19cXf/31Fy5cuHDTuX333Xdo3749IiIi8Mgjj+D7779HYWFhjdzvuusu/Pzzz/j444+RkJCA5cuXw9PTEwBQUlKCQYMG4ciRI/juu+8QHx+Pjz/+GGq1+qZzefnll/HQQw8hLi4OU6dOhRAC3t7eWLFiBRISErBw4UIsW7asRkGzZcsWjBw5Ej179sTu3buxd+9eTJkyBRUVFRgyZAjat29vLXAu+/LLLzFx4kRotdqbzpGIqFkQRERkN5999plwd3cXpaWlQgghysvLhcFgEP/973+v+Z6DBw8KACI1NVUIIURiYqIAILZv317n48tCQ0PFv/71LyGEEFu3bhUKhULk5eXV2ObRRx8VY8eOvW7O4eHh4tFHH7U+3rRpkwAg4uLirM8lJCSIzp07C4lEIjp06CCmTJkivv32W1FRUXGDT0SI7t27i4ULF9aI98UXX1gfb9myRQAQ+/fvr/P9X375pVAoFCIlJaXO17du3SoA1HpdJpOJZcuWCSGufIZvvvnmDfP96KOPRLt27ayPb7/9djF69Ohrbv/hhx+KoKAgUVlZKYQQ4sSJEwKA2Ldv3w1jERE1VxypICKyge3bt0Oj0Vj/u3xle8KECSgpKcHPP/8MANiwYQNMJlONkYiYmBiMHDkSgYGB0Gq11mbo5OTkW85n//79KC8vh7+/f428vv32W5w+ffq6733qqaewatUqmEwmAMDixYtx2223oUuXLtZtOnXqhLi4OMTGxuL5559HeXk5nnjiCfTt2xclJSXX3Pe+ffsQFxeHhx56yPrcww8/XGMKVGxsLDw8PBAVFVXnPmJjYxEeHo6AgIB6fRbX07t371rPLV68GH369IG3tzc0Gg1mz55d43sRGxuLESNGXHOfjzzyCDIzM7F582br/rp164ZevXo1OF8iInthUUFEZANRUVE4fPiw9b9nnnkGAODh4YExY8Zg+fLlAIDly5dj9OjR0Ov1AIDz589j1KhRCAkJwcqVK3HgwAFrAXKtxmGptOpXuxCixvMVFRXWry0WC9zc3GrkdPjwYcTHx2Pjxo3XPZYpU6ZACIHvvvsOGRkZ+Pnnn60N2tVJJBJ0794d06ZNw/fff4/ff/8dsbGx+OGHH66570WLFsFsNsPX1xdyuRxyuRyzZ89GbGwsDh48WGPf13O91+v6fCorK2GxWGpt6+LiUuPx//73P0ydOhUTJkzAhg0bcOjQIbz++us1Ptsbxa/em1JRUYHly5fX+fkREbUkcnsnQETkCFQqFdq1a1fna1OmTMH48eNx8uRJrF+/HqtWrbK+tn//fpSUlGDhwoXWfoXY2NjrxrrcW5CWlmZ9LjMzs0Z/Q1RUFPLy8lBaWlpjhKE+PDw8cP/992Px4sUwmUxQq9V44IEHbvi+sLAway51MZlMWLlyJT755BMMHDiwxmsvvPACFi1ahM8//xw9e/aE0WjEgQMH6hyt6NmzJ5YuXYrU1NQ6Ryu8vLwAVH0+gYGBAIDDhw/XKsLqsm3bNnTv3h0zZ860PpeUlFQr/ubNmzFt2rRr7ufpp5/GkCFD8Pnnn6OoqAh/+9vfbhibiKhZs+/sKyIiqqioEF5eXiIyMlIYDAZRXl5ufe3IkSNCIpGIt956S5w7d06sWbNGdOzYUQAQW7duFULU3UPRv39/0aNHD3H48GFx4MABMXLkSKFWq609FRaLRURHR4v27duL1atXi7Nnz4oDBw6I//znP2LRokU3zHnnzp0CgPDw8BBTp06t9fr48ePF/Pnzxa5du0RSUpLYuXOnuPPOO4WTk5M4fvx4nfv85JNPhEajEcXFxbVeW7JkidBqtaKwsFBYLBYxYMAA0bZtW7F27Vpx7tw5sWPHDrF48WIhhBBFRUWiQ4cOonv37uL3338X586dE1u2bBErV660ft7BwcHijjvuEAkJCWL79u1iwIABQiKR1OqpuLov5eOPPxYqlUqsXbtWnDlzRixcuFDo9XpR/c/p5s2bhVQqFdOnTxdHjhwRJ06cEMuWLRMnTpyosa/OnTsLZ2dn8cgjj9zw8yYiau5YVBARNQMzZswQAMTzzz9f67X//ve/IiAgQCiVStG/f3+xcePGGxYVJ0+eFAMHDhRqtVq0a9dO/PTTTzUatYUQori4WLz88ssiJCREODk5CW9vbzFy5Ejxxx9/1CvnLl26CADiyJEjtV5btGiRiI6OFj4+PsLZ2Vn4+fmJsWPHil27dl1zf926dRMTJ06s8zWj0SicnJyshYPJZBLPP/+88PHxEU5OTiIkJES899571u3T09PF5MmThV6vFwqFQnTs2NFaMAghxJ49e0SPHj2EUqkUERERYtu2bXU2al9dVJSXl4unnnpKeHh4CK1WKx588EHx8ccfi6uv0W3atEn07dtXKJVK4erqKgYPHizOnj1bY5uFCxcKANf9TIiIWgqJEPUY7yUiIqJG9dJLL2Hjxo2Ii4uzdypERA3GngoiIiIbys/PR1xcHBYvXlyvmwESEbUEHKkgIiKyocGDB2Pv3r2YMGECli5dal2NioioJWNRQUREREREDcLLI0RERERE1CAsKoiIiIiIqEFYVBARERERUYO02tWfqt9J1h4MBgOys7MZs4XHc5SYjnCM9ojpCMdoj5iOcIyOEtMRjtEeMR3hGO0R0xGOsa6Yfn5+9XofRyqIiIiIiKhBWFQQEREREVGDsKggIiIiIqIGabU9FVcTQqC0tBQWiwUSiaTJ42VkZKCsrKzJ4zhazJuJJ4SAVCqFUqm0yfeciIiIyFE5TFFRWloKJycnyOW2OWS5XA6ZTGaTWI4U82bjmc1mlJaWQqVSNWFWRERERI7NYaY/WSwWmxUU1HzI5XJYLBZ7p0FERETUqjlMUcHpL46L33siIiKipuUwRYW9GY1GDB8+HMOHD0dkZCR69uxpfVxeXl7v/axcuRKZmZnWxy+++CLOnDnTaHn+8ssv8Pf3R2JiYqPtk4iIiIhaN84HshGdTofff/8dAPDhhx/CxcUFzzzzzE3vZ+XKlejSpQu8vLwAAAsWLGjUPNetW4fevXtj3bp1mDFjRqPuuzqz2czpaEREREStBEcqmoEffvgBo0ePxvDhwzF79mxYLBaYzWZMmzYNw4YNw9ChQ7FkyRKsW7cOx48fx7PPPmsd4Rg3bhyOHTsGs9mM9u3b491330V0dDTGjBljvRviuXPncNddd2H06NF4//33ERYWVmceBQUFOHToED744AOsW7euxmv/+c9/MGzYMERHR2Pu3LkAgLNnz2L8+PGIjo7GyJEjkZKSgm3btuGxxx6zvu/ll1/GTz/9BADo2bMnFixYgLFjx2Ljxo1Yvnw5Ro0ahejoaDz11FMoKSkBAGRmZuLRRx9FdHQ0oqOjcfDgQbz33nv46quvrPt95513ajwmIiIiIvtxyEvFlpWLIVIad3qPJLANpBOfvOn3nThxAps2bcK6desgl8vx0ksvYd26dQgODkZubi7++OMPAEB+fj7c3NywbNkyvP322+jSpUutfZlMJvTt2xdz5szBG2+8gZUrV+L555/HP//5Tzz99NMYM2YMli1bds1cNmzYgGHDhqFdu3ZQq9WIj49HeHg4fvvtN2zduhW//vorVCoVcnNzAQBTp07FP/7xDwwbNgylpaUQQtxw2pRarbYWLEajEVOmTAEAvPvuu/jhhx/w8MMP49VXX8XAgQPx6KOPwmw2o6SkBDqdDs899xyeeOIJVFZW4tdff8XGjRtv+vMmIiIiosbHkQo72759O44cOYI777wTw4cPx+7du5GUlISQkBCcPXsWr7/+OmJiYuDq6nrDfalUKgwdOhQAEBERgZSUFADAoUOHMHr0aADAuHHjrvn+devWYezYsQCAu+++23ryv2PHDkycONG6LKuHhwfy8vJgNBoxcuRIAIBSqazXsq1333239euEhATcc889GDZsGNatW4dTp04BAHbt2oVJkyYBqFq9SavVIiQkBC4uLoiPj8eff/6JyMhIuLu73zAeERERETU9hxypuJURhaYihMCECRPw0ksv1Xpty5Yt+PPPP7FkyRJs2LAB77///nX35eTkZP1aJpOhsrKy3nnk5ORgz549OHPmDCQSCcxmM5ycnPDKK69ACFHne+paVenqJVyvvlGdWq22fj19+nR8++236NSpE1asWIGDBw9ed98PPvggVq1ahfPnz1uLDiIiIiKyP45U2NmAAQPwyy+/wGg0AqiaEnThwgXk5ORACIExY8Zg1qxZiIuLAwBoNBoUFRXdVIzIyEjrVKGreyUu++WXXzBx4kTs27cPe/fuRWxsLLy9vREbG4uBAwdi5cqV1p6H3NxcuLu7Q6fTYfPmzQCqbi5YUlKCgIAAnDp1CuXl5cjLy8POnTuvmVdJSQm8vLxQUVGBNWvWWJ/v168fvvnmGwBAZWUlCgoKAACjRo3Cb7/9huPHj2PAgAE39RkQERERUdOxyUjFp59+ioMHD8LNzQ0ffvhhrdeFEFi2bBkOHToEhUKB5557Dm3btgUAxMTEYPXq1QCA8ePHY/DgwbZI2WbCwsIwc+ZMTJgwAUIIyOVyzJ07FzKZDH//+98hhIBEIsGrr74KAHjggQcwa9YsKJVKrF+/vl4x3nzzTUyfPh2ffvophg4dCq1WW2ubtWvXYubMmTWeGzVqFNauXYu3334b8fHxGDVqFORyOYYPH46XXnoJH3/8MWbPno333nsPTk5OWLx4MYKCgjBy5EgMGzYMbdu2RdeuXa+Z16xZszB69Gj4+/ujY8eO1lGNd955B//4xz/w7bffQiaTYd68eejevTuUSiX69u0Lg8EAqZT1MBEREVFzIRHXmtvSiOLj46FUKvHJJ5/UWVQcPHgQmzZtwuzZs3H69Gl89dVXePfdd1FYWIhXXnnFutrQ5a81Gs0NY6alpdV4XFxcXGPqTVOTy+Uwm802i3e9mMXFxVCpVJBIJPjpp5+wadMmLF68uEljNgWLxYIRI0ZgyZIlCA4Orvf7Gvq9NxgM1pW0bMXWMR3hGO0R0xGO0R4xHeEYHSWmIxyjPWI6wjHaI6YjHGNdMf38/Or1Pptc7g0PD79uIXDgwAEMHDgQEokEHTp0QFFREXJzc3H48GFERERAo9FAo9EgIiIChw8ftkXKrcrhw4cxYsQIREdHY8WKFXjttdfsndJNS0hIQL9+/TB06NCbKiiIqGlVVApYmv7aFBERNXPNolHbaDTCYDBYH+v1ehiNRhiNRuj1euvzOp3O2ntA9devXz/rjfdaqrCwMOzZs8cuI0BEVDVNNbOoAkl5ZUjOLav6f14Z0grKoZCfQZCbM0LcFQh2VyDEo+r/GmeZvdMmIiIbaRZFRV0zsOpa/ed6z2/ZsgVbtmwBAMydO7dGkQIAGRkZNr+Dsz3uGO0IMW82nkKhqPXzcLPxGvL+lhDTEY7RHjFb6jEWlZlxNqcYZ7OLcCa7COeyi3E2pwhF5VdWlPNzU6KdpwbDOrmgpMKC05kF2JVSiM1n8qzbeGsVaGdQI9TgglC9C0INLgj0UEEurfv3eH211M+VMe0fz1FiOsIx2iOmIxxjQ2I2i6JCr9fXmLuVk5MDDw8P6HQ6xMfHW583Go0IDw+vcx+X77582dXzz8rKyiCT2e6qWXPqqWhNMW8lXllZWYPmIzaH+YytLZ6jxGzux1hpEUgvKEdSXhmScsuQnF/1/8yiCus2Lk5SBLsrMChEixB3JUI8FAh0c4ba6crv08sxhRDIKTFX7Svv0miGsRh7knJReenakZNUgkA3Z4R4KBDirqwa2XBXwF1V/z9Hzf1zZczmG89RYjrCMdojpiMcY10x69tT0SyKiqioKGzatAn9+/fH6dOnoVar4eHhgcjISHz//fcoLCwEABw5cgQPPfSQnbMlImp58kvNV4qHSyf8KfllKL90ti+VAP6uzuhoUGJkO3frFCaDWn7NEeKrSSQSGNROMKidEOV/pY+uotKCVFM5kqpNmzqUXow/z5ms27gpZQi5VGCEeFQVG4FuznCWcaU3IqKWwCZFxcKFCxEfH4+CggI888wzeOCBB6xXm0eMGIHu3bvj4MGDeOGFF+Ds7IznnnsOQNU9Ge69917Mnj0bAHDffffVa+UnIiJHVW624Jyx1HryXtUDUYrc0itTl9wvncDf2d4dIR5KhLgrENCEJ/BOMinaeCjRxkNZ4/n8UrM1x8sFx8bTeTUKHT/t5VGNKyMbej0bw4mImhubFBUzZsy47usSiQRPPPFEna8NHToUQ4cObYq0bMpoNGLChAkAgKysLMhkMuh0OgDA+vXr4ezsfMN9vPjii5g6dSratWt3zW2++uoruLq6Yvz48Y2Sd1ZWFnr27Il58+bhwQcfbJR9ElHjKiyvxO7zBYhJMiEh6yQqLVUn3ZenGnX3c7nlqUZNyU0pR4SPHBE+LtbnKi0C6YXl1mbwpLwynMouxY7kAus2WkUy+gS4YFCIK7p4qyGt50gK2Y8QAsZLU+OSrMVuGYylZyCExaa5eGrPI1AjR/DlYtVDAb2q/iNyZHvFFZU4n1d+6SJEKZJyy2AsMWNSrwrc7svvXXPRPP6yOACdTmddgenDDz+Ei4sLnnnmmRrbCCEghLjmjd0WLFhwwziPPPJIg3Ot7ueff0bPnj2xdu3aJi0qzGazXZrMiVqqikoLDqQV4a9EEw5cKESFRcBP64wHu/vDR2lBiIcCflpnyBrYFG1rMqkEAa4KBLgq0L/a6tHFFZVVJxS5ZUgqsCDmTA62nM2HXi3HwGBXDG7jipCrRkLIPkrNFpzPq148lCI5rwwF5VeKB0+1HMHuCvQM1qGsrNRmuQkhkF8hQUJGAbYlX5l+p3GWVq1edmnkLvjSf0o5p9/ZUqVFIKOwoqpwqDZd82LhlT4vlbyqz8tNKcf8rWexLUCDaX184KrkOYS98TtgZ4mJiXj88cfRq1cvHDp0CF9//TUWLFiAuLg4lJaW4u6778aLL74IABg3bhzefvttdOrUCV27dsXkyZPx559/QqVSYdmyZfDx8cG8efOg0+nw5JNPYty4cejduzd27twJk8mEjz76CL169UJxcTGmT5+OxMREdOjQAYmJifjggw/QpUuXWvmtXbsW77zzDp5++mlkZmbCy8sLQNVqWx988AEsFgsMBoO19+XVV1/FsWPHIJFIMGvWLERHR6Nr165ISEgAAKxbtw7bt2/H/PnzMW3aNBgMBsTFxSEyMhJ33nkn3njjDZSVlUGlUmHBggVo27YtzGYz3nrrLWzfvh1SqRSTJk1CSEgIVqxYgUWLFgEA/vzzT6xatQpffPGFjb5zRLZnEQIJmSX4K8mEHedNKCq3wE0pwx3t3TGojSva6ZTw9PS0eVOfLaidZAjzVCPMUw2DwYDHumViX2oh/krKx88njFiTYLzUUO6KgSGu8HRxsnfKrZ5FXD4BrL7McCnSCypweYKaUi5BsLsC/YJc61xu2J5NqIXllVeKn0v5/3E2H6XmquJHAsBH63Rl6t2lkQ1vjRNHxxqBqawSyZdGHS5P10zOq9nn5at1RqhOiWFt3ayfv5eLEyQSCSxCYGtqOT7dkYQXNiRhxm2+iPR1uUFUakoOWVR8eSADibmNe2WkjYcST0R539J7T506hY8++gjz5s0DAMyePRseHh4wm824//77MXr0aHTo0KHGe0wmE/r27Ys5c+bgjTfewMqVK+ucZiaEwPr16/Hbb79h4cKF+O6777B06VJ4enpi8eLFOH78OO64444680pJSUF+fj4iIiIwevRo/PLLL3j88ceRmZmJ2bNnY82aNQgJCUFWVhaAqhEYvV6PP/74o+pqUH7+DY89OTkZP/zwA6RSKUwmE9auXQuZTIatW7fi/fffx+eff47ly5cjIyMDv//+OxQKBbKysuDm5oZ//vOfMBqN0Ol0WLVqlXV6GVFrcz6vDH8lmfBXYj6yis1QyCS4LVCLQW1c0c3HpcWNRjQGhVyKASGuGBDiClOpGTvOFyAm0YTlh7PwzeEsdPZWY3CIK24L0vJ+GY2gsPzKSNHlKSjJeWUoNVedAEoA+GqdEOyuxKA2btame69mfAKucZYh3EuNcC+19TmLEMgsrLjS63PpmPekFNYolILcFNYVzC6PbGgU/DmrS0WlwAXTlcLh8s9QTsmVlRxdFVV9XiPbu1uLuEA3ZyiuM1IklUgwobs/2moE5u9Iw7/+TMG4MB0mdTPAiQs82IVDFhXNTXBwMCIjI62P161bh++//x6VlZW4ePEiTp06VauoUCqV1l6TiIgI7N27t85933nnnQCArl27IiUlBQCwb98+TJ06FQDQuXNndOzYsc73rl27FnfffTcAYOzYsZgzZw4ef/xxxMbGol+/fggICAAAeHh4AAC2b9+OpUuXAqjqk3F3d7/h8q933XWXdbqXyWTC9OnTkZycXGOb7du344knnrAuCXw53j333IO1a9finnvuQVxcHD799NPrxiJqSXKKK7A92YSYRBMSc8sglQCRPi6YFOmJPgFaqJz4R/MyV6Ucozp4YFQHD6QXlGNbUtXn9t+9F/HF/gxE+WswqI0rovxceLJxA2aLwPn8asVDbtU0lOziK7/Ltc5SBHsoER3qbu1JCHRrHVOFpBIJfLTO8NE6o0+g1vp8mdlS83PJK8Pu8wX47cyVi2eGS1O6Lq9gFuKugJ+rc4Pvy9JSCCGQU1xxVfFZhlRTGS4N/kAuBQLdFOjqo67xObkrZbfcF9HGQ4mP7gzBsoOZWJtgxNGLRfh7fz8EuCka8eioPhyyqLjVEYWmolZfuUpy7tw5fPnll1i/fj3c3Nwwbdo0lJWV1XpP9cZumUyGysrKWttU3676NnXdbLAu69atQ25uLv73v/8BqLqB4Pnz5yGEqPfNCaVSaY14paU1R4iqH/u8efMwaNAgPPLII0hMTMSkSZOs+dYVb8KECXjqqacAAGPGjLHpfUhak/SCcszfHY/RoRqEVbti15pkF1fgq4OZkMiz4KuG9UqYt8apWV3lL6640nAdd7EYAkB7vRJP9PTCgGDXZtNg3Zz5ap0xoasBD3TR44yxFH8lmrAt2YTdKQVwcZbi9iBXDApxRZiXqtleQbe1i5cKsT2phTiffxIVl6afyCRAgJsCnb3U1U6WFdA5YFOzQi5Fe70K7fUq63OXm8+Tq02fSsorw5GLRdVOoqsWS+jonQMfFRrlJLo5uFxkVT/28/lnkF96pfjUq+UIcVegp5+LdZlo/yYqshRyKZ7p7YPufi74eM9FvLgxCU/09MaIdm4t+nNuafgXqpkpLCyERqOBVqtFRkYGYmJiMHjw4EaN0bt3b/zyyy/o06cPEhIScOrUqVrbnDp1CpWVlYiNjbU+N3fuXKxbtw4TJ07Ev/71L6SmpiIkJAS5ubnw8PDAwIEDsWzZMrz++uvW6U/u7u5wc3PDuXPnEBISgk2bNkGv19eZl8lkgq+vLwDghx9+sD4/aNAgLF++HH369IFcLrfG8/f3h06nwyeffFJje7o5q+KysT3RhJ2JRjzQRY8Huhia1Yl2Q+06b8Iney/CbBHw0iqx41wJLi2OBIVMgqBLJ0tX5nsr4WrDaQwVlQKH0gsRk2jC/guFKK8U8NE44YGuegwKcYO/641XhqPaJBKJ9STw0R5eOHKxqqk9JjEfm8/kwVMtx6A2bhgU4oogd8e7omkqq8TOSyNhJ7JLAADhnircH+kHb4UFIe4K+Lsq4CRrPb8LGptEIoFe7QS92gk9/Krfl6XmdJ/E3DIcSMlHdlG5dRs3hQzBl/pL6jvdxx4uTwervkR1Um4Z0gvKrdPBLv8eHdRODx8lrE3uWjtMB+sToEW7UUr8e3c6Pt13EQfTCzG1j69Nf6c7MhYVzUzXrl1P6qNTAAAgAElEQVTRvn17DB06FEFBQejVq1ejx3jssccwffp0REdHo0uXLujYsSNcXV1rbLNmzZpavRajR4/GjBkzMG3aNLz33nt49NFHAQDe3t749ttvMXPmTMyePRtDhw6FVCrFSy+9hBEjRmDOnDmYNGkS/Pz80KFDB5SXl6MuU6dOxcyZM/Hpp5+iX79+1ucnTZqExMREREdHQy6XY/LkyZgyZQqAqub1goIChIaGNuZH5DCyiiqwLcmEuzp7o6i4BCvjcnA4vRgz+/vCW9OyT2ZLzRZ8eSADv5/NR3u9En/v74eubfxw4WImUvLLrauLJOeWYW9qIX4/e2Uag14ltzaUXv6j35gnWEIInMguwV+JJuw4X4CCskpoFTJEh7phcBs3dNAreXWtEcmkEvTw06CHnwbPVFiwN7UA25JMWB2fgx+P56CNhwKD27hiQLAr9OrW2+BdZrZg/4WqAvZgWiEqBRDk5ozJkZ4YdKm53R6N062Nk0xSNSJRbTUyg8GAcxcyajUmb67jviyXf+dc3Zjc1Iou981UG32o6pup3bg+KORK4/3lxvXm8rOjVzvhjaGBWJdgxLdHsjBjfSJm9POtsXQ1NQ2JqO9cmBYmLS2txuPi4uIaU22amlwuv2E/gb1ims1mmM1mKJVKnDt3Dg899BB27NhxS0u62vo4r4738ssvo2fPnnjggQeu+Z6Gfu/tuTpJU/vyQAY2nMrFD49EQV5eiG1JJny27yIA4Jle3hjUxq3JYjflMZ7JKcWHO9OQXlCOezvr8WCEAXLptf/oCSGQV1p56Y/plRvHpeTXnAsc4Frzj32w+/WnglwdLzW/quF6W5IJFwsr4CyToE+ABoPbuCHS16VRpgW05p/Xxo6XV2LG9mQT/koy4XROKSQAuvpcafBWO125utlSP9dKi8CxzGLEJJqw+3wBSswW6FRyDAy5tAyvu6LGz29L/V621JiVFoGLhRVVxcY1llBVO0mvFBrVCo7qP5/1jXc5ZlpB9TvcVxU6WdX6ZupaYjfITXHdXq7m9LledtZY9bcgzVSOe8J1eCjCs0EXh5rjMdoipp+fX73ex5EKB1RUVIQJEyZYT87nzZvXIu8RMXz4cLi5ueGtt96ydyotkqmsEr+dycOAEFf4uCqRnV2IgSGu6GhQYsGudHy0Kx0H04rwdG/v6/7xak4sQmBtvBHfHc2Cm1KOt6ID0dX7xlenJBIJPFRyeKjk6F5tSUKzRSDNVF6j2IjLLEZM0pX17bWXVi2pPoUqyE1hncaQe/nENdGEM8ZSSCVAhLcaE7oa0DdQ02I+29bIXSXHmE46jOmkwwVTOf5KysdfiSb8Z89FfL4/A738NRjcxhXdfTU33lkzIoRAYu6VAtZYYoZKLkW/IC0Gt3FFZy91q5ri2JLJpBL4uzrD39UZ/YKuPF/Xzd62JZlQVHHlXh9eLk7Wu81fLjZ8r7o3TV6pucaKXUm5ZUjJL0eFpVrfjKsCYV5q3FGtb6a13AwwVFfVxL0kNgOr4404crEYf+/vx2mlTaTlnUlSg7m5uWHTpk32TqPBLt9MkG7NhlO5KKsUGB9es8fFW+OMd6KD8L9jOVh1LBsJ2SWY2c8PnTxV19hT85BTXIGFu9Nx9GIxbgvUYmofnwbP6ZVLq+YKB7krMDDkyhTBwrJq0wQu/aH+7UweyiqrL6/pDIMmDccummARQKhOgcd6eGFAiCt0bLhudvxdnfFQhCce7GrAqZxSxCTmY0dyAXaeL4DWWYrbQ3Phr8alwtG2fTf1lVVUYV16+Hx+OWQSoKe/BoNDXBHlr2l28/Xp2tROMnTyVNX4vSuEQHZx9cbwqgsdBy4UWvvEnGUSBLop4K5Ox5nsQuSXXlnExUNV1Tgd4eNiLR4CXJ1b/YpoSrkUU/v4ooefBp/sScfMjYl4Msobw9o23yZuIQRO5ZTiryQTwv0rcLtvy5iWyb9sRA6ozGzB+pO5iPJzQXAdTaoyqQQTIwzo5qvGRzvTMfv3ZEzsasB9nfXN8grnnpQC/HdPOsorBZ7v44Po0Kb9Y6FRyNDZW43O3jXXt88orKhxVTCv3IJ7w/UY1MYVgVzesEWQSCToaFCho0GFx3t643B6EWIS87E7KRd5JVempOgunaBdHp2yV2NzYVkldqUUICYxH8czqxquwzxVeKaXN/oHuzbL4odujUQigaeLEzxdnBDlf2X0rLzSgtT88mr3gShFcUUlovw0NUYy3Bz8jtO3BWrRXq/Ewl3p+HjPRcSmFWFqb59mdX+R9ILyqgUlkvKRXlA1RdZD6wKwqGheWmnrCNUDv/e1bTmbD1NZJcZ3rnslrsvCPNVYOCoEX+zPwIqj2TicXoQX+/nBS9M8fsGVmS1YEpuJzWfyEKpTYGZ/PwS42ufkXSqRwFfrDF+tM24Lqlrfvrk0LtKtkUsliPLXIMpfA4PBgDMpF2uMTiXlleHoyWKYLTWXYL16OlxjL8FaUWnBgQtFiEnKx4ELRTBbBAJcnfG3bgYMCnFt8Yss0M1xlknRVqdEW13NxnD+7qnNoHbC/w0NxNoEI747koVT2SV4sZ8funjbbzn1/FIzdiRXXRg4Va23677OetwWqEWwn3eL+V46TFEhlUphNptbZO8A3Tqz2Wy9uR5VqbQIrE0woqNBhfB6TGlycZZhZn8/9PBzwef7MjBjQyKe7e2DASGuN3xvUzp3qQHvgqkc4xuhAY/oRtxVckSq5Ii8uu/mUtPr5avExzOL8Vf1vptLN4sLqTZnvXrfTX1YhEB8ZgliEvOxK6UAReUWeChlGNXBHYNC3BCqUzTbqRxEzYlMKsG9nfWI8FHjw51peG3L+RoLethCqdmCfamFiEnMx6H0IlgE0MZDgUe6e2JgSMtdhc5hzrCVSiVKS0tRVlZmk1+8CoWizpvWMabt4gkhIJVKoVQqb7yxA9mRbEJmUQWeiPK6qX8Lg9u4oZNBhY92pWH+zjQcTC/Ck1FeNm80tgiBX07kYvnhLLgqZPi/YYHoxqUCyU7kUgmC3KqKhOqq991cng635WweSs3V+26cEOyutE5RCXFXwOvS8pyXJeeVISYxH9uSTMguNkMpl+K2wKoVw7p6s+Ga6Fa116uw4M42+DI2Az8ez8GRS3fi9tU2zUhfpUXgaEYxYhLzsSelAKVmAYNajnvCdBjUxq3OqcgtjcMUFRKJBCqV7RpNm8MSYK0xJod0G0YIgdXxRgS6OaOX/82vaOOjdcZ7w4OxMi4bPx7PQXxm1UoaHQy2+bdlLDHj37vTcTi9CH0CNHi+L29qRM3TdftuLt0f5fJynntSCqw3ElPKpQh2d0agmwJJ+Sk4k10EqQTo4euCh7t7oU8AG66JGovKSYppfX3Rw9cFn+y7iBkbkvB0L28MaePaKBeghRA4Y6xquN6eZEJeaSVcnKVVSzqHuCHMS1XjIkJL5zBFBREBB9OKkJRXhhf6+tzyLzKZVIK/dfNEpK8LFuxMwyu/JeOhCE/cE65r0qum+1ML8Z896Sg1W/Bcbx+MaNd8V+4gqkuNvptArfX5UrMF562riZUhObcU+1ILEeChxlNR3rg9WOvwTbZETal/sCs6GFRYsCsN/96djoNphXimtw80zrd20epiQTm2JZkQk2TCBVM55FIJevm7YFAbN0T5ubTaFbf4W4rIgayOz4FeLcfAkIbf1K6zlxoLR7fBZ/su4psjWTiUXogZ/fzg6dK4c0HLzBZ8dSgTG07loY2HAn/v78eVlKhVUcql6GBQ1Rrx48gske14ujjhrWFBWB2fgxVHs3HyUhN3uFf9mrhNZZXYmWxCTKIJJ7KrVmLr4qXCuDAf9AvUNqtVppoKiwoiB3EyuwTHMkvwWA+vRmto1jjLMKu/H3r6mfDF/qom7uf6+KB/UOM0cSflVjVjn88vx9hOHpgc6dlqr/AQEZF9yaQS3N/FgAgfF3y0Mw2vbjmP+7voMaGLoc6R+DKzBfsvFCIm0YSDaYWoFECQmzMmR3piUIhro19ka+5YVBA5iNXxOXBxlmJ4u4aPUlQnkUgwtK0bwjxV+HBnGt7fnobo0CI80dMbKqdbKwCEEPj1ZC6+PpQFjbMUbwwNrHGnayIioqbS0aDCglEhWLQ/A6vicnA4vRh/7+8Lg6Gq4fpYZjFiEk3Yfb4AJWYLdCo5xnTSYXAbV4S4O+5KbCwqiBxAqqkMe1MKcV9nfZOt1uSrdcbcEcH4/mg2frI2cfujnf7mVt/KKzHjP3vSEZtWhF7+LpjW15fzyYmIyKbUTjLM6OeHHn4afLbvIqavT0J0xyLsPJcDY4kZKrkU/YK0GNzGFZ29uBIbwKKCyCGsiTfCSSbBXZ08mjSOXCrB5EhPRPqqsWBXOl7+LQl/i/DEuHBdvRrDD1yoasYuqbDg6V7euLO9u8Ne8SEiIvsbGOKKjoaqO3FvSMhETz8XDA5xRZQ/V2K7GosKolYup7gCMYn5GB7qDncbXfHv6u2Cf49qg0/3XcTXh7Nw6GIRZtzme80b+pRXWvD1oSz8ejIXwe4KvD3MD0GtYM1uIiJq+bw1znhvRDDcPHTIzzXaO51miyUWUSv384lcWAQwLkxn07hahQwv3e6H5/v44GRWCaavT8SelIJa2yXnlWHWpmT8ejIXYzp6YP4dwSwoiIio2eFCIdfHkQqiVqywvBKbT+ehf5AWPk10l9DrkUgkGN7OHeFeany4Mw3vbbuAke3c8VhPLwghsP5kLr46lAmVkxSvDw5Az1u4IR8RERHZH4sKolZs06k8lJgtGB+ut2se/q7OmDciGCuOZmFNvBHHM4sR4JGFPcm56Onnghf6+sJdxV9HRERELRX/ihO1UuWVFvxy0ohIXxe01d3cCkxNwUkmwcPdvdDd1wULdqXjYGoenozywugOHmzGJiIiauFYVBC1Un+ey0deaSXuDbdtL8WNRPi44JMxbaDQuENWVrvHgoiIiFoedpwQtUKVFoG1CUa01yvR1Vtt73RqUTvJ4K1lMzYREVFrwaKCqBXak1KA9IIKjA/XcWoRERERNTkWFUStjBACP8Ub4ad1Qp8Arb3TISIiIgfAooKolTlysRhnjaW4J1wPmZSjFERERNT0WFQQtTKr43PgoZRhcBtXe6dCREREDoJFBVErcianFEcuFmNMJx2ceedPIiIishGedRC1Iqvjc6B2kuKO9u72ToWIiIgcCIsKolYivaAcu1MKcEd7d7g4y+ydDhERETkQFhVErcTaBCOkEgnGdGpeN7sjIiKi1o9FBVErkFdixh9n8zG0rSt0Krm90yEiIiIHw6KCqBX45WQuzBaBcWF6e6dCREREDohFBVELV1xRiY2nctE3UAt/V2d7p0NEREQOiEUFUQu3+XQeiiosuLczeymIiIjIPlhUELVgFZUW/HwiFxHearTXq+ydDhERETkoFhVELdhfSSYYS8wY35m9FERERGQ/LCqIWiiLEFgdb0RbDwUifdT2ToeIiIgcGIsKohZqX2ohLpjKcU+4HhKJxN7pEBERkQNjUUHUAgkh8NPxHHhrnNA/SGvvdIiIiMjBsaggaoGOZ5bgVE4pxoXpIJNylIKIiIjsi0UFUQu0Oj4HbgoZhrV1s3cqRERERJDbKtDhw4exbNkyWCwWDBs2DOPGjavxelZWFj777DOYTCZoNBpMmzYNen3VijYTJkxAUFAQAMBgMODll1+2VdpEzU5Sbili04rwt24GKOS8LkBERET2Z5OiwmKxYMmSJXjttdeg1+sxe/ZsREVFISAgwLrNN998g4EDB2Lw4ME4duwYVqxYgWnTpgEAnJ2d8cEHH9giVaJmb3W8EUq5FKPae9g7FSIiIiIANpr+dObMGfj4+MDb2xtyuRz9+vXD/v37a2yTmpqKrl27AgA6d+6MAwcO2CI1ohYlo7Ac25NNGNnODRqFzN7pEBEREQGw0UiF0Wi0TmUCAL1ej9OnT9fYJjg4GHv37sWoUaOwb98+lJSUoKCgAFqtFhUVFXjllVcgk8kwduxY9O7du1aMLVu2YMuWLQCAuXPnwmAwNO1B3YBcLrd5Do4Q09bx9ibnYuuFdNwV7gWVk+1O4q91nN8cOwupRIJH+rWDQato8nhNyRFiOsIx2iOmIxyjo8R0hGO0R0xHOEZ7xHSEY2xITJsUFUKIWs9dva7+5MmTsXTpUsTExCAsLAw6nQ4yWdVJ3KeffgqdToeMjAy8+eabCAoKgo+PT433R0dHIzo62vo4Ozu7CY6k/gwGg81zcISYtoy38VQuFh3IgEUAX+9JwsQIA6JD3SG3wWpLdR2nqdSMn49dxMAQV0jLCpBdVtCk8ZqaI8R0hGO0R0xHOEZHiekIx2iPmI5wjPaI6QjHWFdMPz+/er3PJkWFXq9HTk6O9XFOTg48PGrOB9fpdJg1axYAoLS0FHv37oVarba+BgDe3t4IDw9HUlJSraKCqLEIIbDiaDZ+OJaDKD8XTO7TBl/sOIvP9mVgXYIRk7p5ol+Q1uY3nFt/KhfllQLjw3U2jUtERER0IzbpqQgNDUV6ejoyMzNhNpuxa9cuREVF1djGZDLBYrEAANasWYMhQ4YAAAoLC1FRUWHd5uTJkzUavIkaU6VF4L97L+KHYzmIDnXDnEEBiApyx7vDg/DqIH/IpRK8vyMN/9icjKMXi2yWV6nZgvUnc9EnQINAt8ab9kRERETUGGwyUiGTyfDYY4/hnXfegcViwZAhQxAYGIhVq1YhNDQUUVFRiI+Px4oVKyCRSBAWFobHH38cAHDhwgUsWrQIUqkUFosF48aNY1FBTaLUbMEH2y/gQFoRHuiix0MRButohEQiQe8ALXr6aRCTmI8VR7Pxzz9S0N3XBVMiPdFWp2zS3H4/k4eCcgvGh+tvvDERERGRjdnsPhU9evRAjx49ajw3YcIE69d9+/ZF3759a72vY8eO+PDDD5s8P3Js+aVmvBWTirPGUjzb2xt3XGO5VplUgmGh7hgQ4ooNp3Lxv2M5eHFjEgaGuOJvEQb4aJ0bPTezRWBdghHhnip08lQ1+v6JiIiIGspmRQVRc5VRWI43/kxBdrEZLw/wR99A7Q3f4yyTYlyYHtGh7lgTb8TPJ4zYdd6Eke098EAXPdyVjfdPa3uSCVnFZjzTm31ERERE1DyxqCCHds5Yiv/bmgKzReDNoYEI81Lf1Ps1zjJMjvTEqA7uWBWXg42ncvHH2XzcE6bD3WEeUDdwGVohBNbEGxHspkBPP5cG7YuIiIioqdikUZuoOTqcXoTZv5+Hk1SCuSOCb7qgqE6vdsJzfXzw8V1t0N3XBd/HZeOZn89h/clcVFTWXlK5vmLTipCcX4Z7wnU2X22KiIiIqL5YVJBDiknMx5tbU+CtccK8kcGNtqJSgKsCrwz0x/sjgxHo6oxFBzLw/K/nsC3JBEsd92u5kZ+O58BTLceAENdGyY+IiIioKbCoIIezNiEHC3alI8xThXeHB0Gvdmr0GB0NKrwdHYTXBwdAKZfiw51p+PvGJBxOr/8ytAlZxYjPKsHYMJ1NbrhHREREdKvYU0EOwyIElh3MxM8nctE/SIsZ/XzhLGu6uloikaCnvwbd/VywLcmE745k4V9/piDCR40pkZ5or7/+Sk6r443QKmQY3s69yXIkIiIiagwsKsghVFRa8O/d6dieXIC7Onrg8Z5ekNqoR0EqkWBwGzf0D9Ji0+k8rDqWg1mbktE/SItJ3Tzh51p7GdrEnGLsSy3ExK56KOUcUCQiIqLmjUUFtXpF5ZV4b9sFxGUU4+FIT7s1PTvJpBjTSYdhoW5YE2/EugQj9qQUYEQ7d0zoaoCH6so/xxWxqXCWSTC6Q933yyAiIiJqTlhUUKuWU1yBN7emIiW/DDNu88WQtm72TglqJxn+1s0Tozp4YFVcNn47k4c/z+VjbJgO94TrUFxhwW8nszCyvTtcG/F+F0RERERNhWcs1Gql5pfh/7amwFRWiX8OCUR33+Z1nwcPlRzP9PbB3Z10+PZIFn44loNNp/MQ6OYMIQTGduIoBREREbUMLCqoVTqRVYK3Y1IglUrwTnQw2umV9k7pmvxcnfHSAH+czinB8sNZOHqxGCM7ecJbU7vXgoiIiKg5YlFBrc7e1ALM35EGvVqOfw0JhK+2ZZyct9er8NawIJw1lqJLiA9KTHn2TomIiIioXlhUUKvy25k8fLbvIkJ1Srw2OADuLbAnIVSnhIuzHCX2ToSIiIionlreGRdRHYQQWBmXjZVxOejp54J/3O4PlROXYiUiIiKyBRYV1OJVWgQ+338Rv53Jx9C2bpjax4d3oCYiIiKyIRYV1KKVmS34YEca9l8oxH2d9ZjUzWCXe1AQEREROTIWFdRimcoq8XZMCk5ll+KpKG+M7sglWImIiIjsgUUFtUgZheX4v62pyCyswEsD/NAvyNXeKRERERE5LBYV1OIk5pbi/7amorzSgv8bFojOXmp7p0RERETk0FhUUIsSm5KHOb+fh8pJirnDgxHkrrB3SkREREQOj0UFtRiH04vwVkwq/LROeH1IIDxdnOydEhERERGBRQW1EBmF5Zi/4wKCPFR4a4g/NAqZvVMiIiIiokt4dzBq9srMFry37QIsAN67K4wFBREREVEzw6KCmjUhBP675yKScsswq78fAtxV9k6JiIiIiK7CooKatZ9P5GJbsgmTunmih5/G3ukQERERUR1YVFCzdfRiEb46lInbArW4t7PO3ukQERER0TWwqKBmKbOwAu/vSIO/qzNeuM0HEonE3ikRERER0TWwqKBmp6oxOxUWi8CcgQFQO7Exm4iIiKg5Y1FBzYoQAp/svYjE3DLM7O8HP1dne6dERERERDfAooKalV9P5uKvJBMeijAgyp+N2UREREQtAYsKajbiMoqw9GAm+gRocF8Xvb3TISIiIqJ6YlFBzUJWUQXe354GP60zZvTzhZSN2UREREQtBosKsrvLd8w2WwRmD/JnYzYRERFRC8OiguxKCIHP9l3EWWMpXuzniwBXhb1TIiIiIqKbxKKC7GrDqTxsTTThwa4G9A7Q2jsdIiIiIroFLCrIbo5nFGNJbAZ6+WvwQFc2ZhMRERG1VCwqyC6yiiowb/sFeGuc8SIbs4mIiIhaNBYVZHPllRbM234B5ZUCcwb5w8WZjdlERERELRmLCrIpIQQ+35eB0zmlmNHPF4FubMwmIiIiaulYVJBNbTydhz/O5eOBLnr0DWRjNhEREVFrwKKCbCY+sxhfHshATz8XPBhhsHc6RERERNRIWFSQTeQUX27MdsLM/n5szCYiIiJqRVhUUJOrqLRg7rYLKDULzB4UAA0bs4mIiIhaFRYV1KSEEPhifwZO5ZRixm2+CGJjNhEREVGrw6KCmtTmM3n4/Ww+7uusx21BbMwmIiIiao1YVFCTScgqxuJLjdkPsTGbiIiIqNWS2yrQ4cOHsWzZMlgsFgwbNgzjxo2r8XpWVhY+++wzmEwmaDQaTJs2DXq9HgAQExOD1atXAwDGjx+PwYMH2yptukU5xRWYt+0CDGonzOznB5mUjdlERERErZVNRiosFguWLFmCOXPmYMGCBdi5cydSU1NrbPPNN99g4MCBmD9/Pu677z6sWLECAFBYWIgff/wR7777Lt599138+OOPKCwstEXadIsqKi2Ytz0NJWYL5gwKgEbBxmwiIiKi1swmRcWZM2fg4+MDb29vyOVy9OvXD/v376+xTWpqKrp27QoA6Ny5Mw4cOACgaoQjIiICGo0GGo0GEREROHz4sC3Splu0+EAmTmaX4IW+vgh2Z2M2ERERUWtnk6LCaDRapzIBgF6vh9ForLFNcHAw9u7dCwDYt28fSkpKUFBQUOu9Op2u1nup+fjtTB42n8nD+HAd+ge72jsdIiIiIrIBm/RUCCFqPSe56uZnkydPxtKlSxETE4OwsDDodDrIZHVPm7n6vQCwZcsWbNmyBQAwd+5cGAz2bQyWy+U2z8HeMY+lm7DoQAZ6B7ljxrCwJumjsPcxttaYjnCM9ojpCMdoj5iOcIyOEtMRjtEeMR3hGO0R0xGOsSExbVJU6PV65OTkWB/n5OTAw8OjxjY6nQ6zZs0CAJSWlmLv3r1Qq9XQ6XSIj4+3bmc0GhEeHl4rRnR0NKKjo62Ps7OzG/swborBYLB5DvaMmVtixuyNSdCp5HihtydyjTk3fnMD4tmSI8R0hGO0R0xHOEZ7xHSEY3SUmI5wjPaI6QjHaI+YjnCMdcX08/Or1/tsMv0pNDQU6enpyMzMhNlsxq5duxAVFVVjG5PJBIvFAgBYs2YNhgwZAgCIjIzEkSNHUFhYiMLCQhw5cgSRkZG2SJvqqaJSYN72Cygqr8Scgf7QsjGbiIiIyKHYZKRCJpPhsccewzvvvAOLxYIhQ4YgMDAQq1atQmhoKKKiohAfH48VK1ZAIpEgLCwMjz/+OABAo9Hg3nvvxezZswEA9913HzQajS3SpnpaEpuBhKwS/L2/H0I8lPZOh4iIiIhszGb3qejRowd69OhR47kJEyZYv+7bty/69u1b53uHDh2KoUOHNml+dGt+PX4RG0/nYVyYDgND2JhNRERE5IhsVlRQ63MquwTzt55HNx81pkR62jsdIiIiIrITm/RUUOtjtgi8v/0CDC7OmHW7P++YTUREROTAWFTQLTl6sQhZxWZMG9AGrmzMJiKimyTMFag02nelRiJqPCwq6Jb8lWSCi7MUt4Xo7J0KERG1MEIIWD59DznPT4TIbZolyInItlhU0E0rM1uwJ6UQtwVq4SznjxAREd2kI3uBuAMQJcUQ/1tq72yIqBHwjJBu2oG0QpSaLVztiYiIbpooL4Nl5ZeAXxBc7p0CsX87xImj9k6LiBqIRQXdtG1JJngoZejipbZ3KkRE1MKIjT8BOZmQPvQMXO5/FNB7wbLiCwiz2d6pEVEDsKigm1JUXonYC0W4PdiVKz4REdFNEZnpEJt+gqT3IPXqsOoAACAASURBVEg6doFEoYB04pNAegrEn7/aOz0iagAWFXRT9qQUoMIiMIBTn4iI6CZZVn0JyOSQ3P/IlSe79Qa6RkH88j1EntFuuRFRw7CooJuyLckEH40TOuiV9k6FiIhaEHFkH3B0PyR3T4TEXW99XiKRQDrxCcBcAfHjMjtmSEQNwaKC6i2vxIyjGcUYEOwKiYRTn4iIqH6qmrMXA76BkAwdU+t1iZcfJCPHQ+z9C+LkMTtkSEQNxaKC6m3HeRMsAlz1iYiIborYtBrIzoD0oachkcvr3EZy5/1VTdvfs2mbqCViUUH1ti2pAMHuCgS5K+ydChERtRAi62JVc3avAZB0irjmdhKFAtIJTwAXkiFi1tswQyJqDCwqqF4yCstxMruEoxRERHRTLKu+BKQySO5/7MYbR/YBuvSA+Pl7iPzcpk+OiBoNiwqql+1JBQCAAcFaO2fSugkhIM6fg6WkyN6pEBE1mDi6HziyD5IxEyDx0N9w+6qm7aeAinKIH79q+gSJqNHUPbGR6Crbkk3oaFDBW+Ns71RaJVFWWtWguHUDkJoIo38QxLNzIPH2s3dqRES3RFSUVzVn+wRAMqx2c/a1SLz9IBlxD8SG/0EMGAFJh85NmCURNRaOVNANJeeVITmvDIM49anRiYw0WFZ9Ccs/HoX45hMAApJxk2ApMMHy7iyIhCP2TpGI6JaIzav/n737Doyyyv4//r5PQho9iSR0JIBlsSEqIChNZVERFQU7gm3tDQVWV11FUER/fq3oIoprwd7AhtJRF1dl7ZggIIEkJAHSSH3u748HopEWSGaemczn9Y/OZGbueUggc+becw5szNpWnN1or55rhp4FifvhvvgktqoqQBGKSH3SToXs0aLVBTgGju2go0/1wbpV8L8vcefPhR++hqgoTI8+mAEnQ5eDMMbQcshwcv95I+7Dd2LOuRzn+CF+hy0iUms2Nxs79zXMkcdiDjpsr59vYuNwRo7FfWIKdsHcvdrpEBF/KKmQ3bLWsmRNAYemNqZFvH5c6sIWbsEu+Ri78APIy4EWSZjTzsX0OwnTvGWNx0altMEZfz/u0w9g//047obfMGeNwURF+RS9iEjtubNngONgzh677y9yRG84+Ajs2y9gj+qLadZyz88REd/oXaLs1sq8UrKKKji7+54L7GRH1lr4dSV2/lzsl4uhshIOOATnrDFw2NG77NcOYOITcK7+O/a1Z7Efv43NWodz2S2YhMZBvAIRkb1jv/0vfPM55oyLMInJ+/w6xhiccy7DvfMa7GvPYcZcX49Rikh9U1Ihu7VodQGNHEOv9jr6tDdseRl2+WKv8HpNOsTFY/qdiOk/FNOmQ61fxzhRmLPH4rZuj33hCdwpt+BcfRumVesARi8ism9sRQXuy09BalvMCcPq/HomtS3mxNOw77+OPe4kTJeD6iFKEQkEJRWyS1WuZemaAnq2bUzjGB27qQ2bswG78H3sknlQUgSt22POvQLTuz8mLmGfX9fpdyK2VWvcJ6bg3nszzt8mYA7oXo+Ri4jUnf3oTcjZgHPDXXtdnL0r5uSR2C8W4r74JM7fH9QxUJEQpaRCdum7nBI2lVbRT12fdsu6Lnz3X6/w+vuvwBg4ohfOgJOhW3eMMfWyjjngEJyJD+A+cjfuQ7djzvsbTr8T6+W1RUTqyublYOe+Aj36YA4+ot5e18TG4Zw9FvfJ+7AL38cMPKXeXltE6o+SCtmlRasLiI926Nmmid+hhCRbVIBdOg+74H3IzYbmLTEnj8Qcd1KthjztC9OqNc6EqbhP3Y+d9ahXwD1iNMbRJ3ci4i/3lRmAwalLcfau9OgDBx2GfesFbM++mGYt6n8NEakTJRWyUxVVLp+tLaRX+ybERmucyR/Z1b94hdfLF0NFOXQ9GHPGhZgjetXbdv/umITGONf8A/vqM9sKuDNxLr0ZE7/vx6tEROrCfv81fPUZ5vQLMEn71fvre0Xbl+PedS32jecwo6+r9zVEpG6UVMhOfbW+mOIKl+N09AnwJsPa5UuwC+bCryshNg7TeyBmwFBMu05Bj8dERWFGXYqb2g770vTfC7j3Sw16LCIS2WxFBe5LT0GrNpgThgdsHdO6HWbwMOyHb2D7nYRJOzBga4nI3lNSITu1cHUBzWKjODQ1stuX2oLNFM59Bffjd6CowOtoMupSL6EIgdauTv+/YlPa4D55n1fAfeVETNeD/Q5rj2z+RiorSqFRnN+hNCi2qICyX3/CFhQEdd2qI44O6nqRwGZl4sbF+B1GrdiP34LsTJzr7sQ0CuxurTlle9H2dJy/P6CjnyEinH5eJXCUVMgOtla4LM8sYlDn5kQ79VNkHI6s6+I+cjclazPg0KNxBgyFgw6rt8Lr+mIOOsyrs3j0Htxpt2EuuArn2EF+h7UD67rw0wqvoH3FcvLj4zH/fHyHwX+yb+yadNxH72Hz5vygr50bFYU5ojdmwFDo+peQ+zsSbuyvv+DeP568Folw5URM+/39DmmXbN5G7JxX4IhemO49Ar6eiYvHnD0W+9T92EUfYvoPDfiasnvuog+xLz7JpnadsOMmY2L1YVGkUlIhO/hiXSHlVTbijz7ZJR/D6l9odt0/KO7e0+9wdsuktvUSi+n3YZ99GHfDWq/OIwQ+xbMlRdhln3oF7dmZ0LQ5ZtAp3u3Xn8WMucHvEMOe/e9S3GcegibNaXHHQxRU2eAtXllJ3I9fU/LxO9gvl0Dbjt48ll79MXHxwYujgbAFm3AfvxeatQDr4t53K84lN2IO7+V3aDvlvjoDsDgjLwnamqbnsdhFh2LffB575LGYps2Dtrb8zrpV2FdnYue9A/t3o3JNOubZ/4PLxumDhQilpEJ2sGh1AckJ0Ry4X+S+IbDFhdg3Z0G3vxB3/EkU5+X5HdIemcZNcK69Azv7X9gP3/QKuC+5sU7zMerCrvvVK2j/fAGUl0HagZhTbsQceSymUSNim7eg5PVZ2wZahf6RrVBkrcXOmY19+0VIOxDnyonEdu6Cyc0NahxNj+lL6QlnYP+zELtgLvaFJ7xi2t4DvQSjdbugxhOubGUl7vT7oaQQ59b7SOyURu7dN+E+Phlz+oWYIWeE1Js1+8M38N9lmNPOwyS1Ctq6xhicc7cXbc/CXHRN0NYWjy0pxn36Afjuv5jBwzAjLiZh6UcUPf8EdOqCOekMv0MUHyipkBoKSiv5ZkMxpx2UiBNCv7yCzb75PJQU45xzeUj9Et8TEx2NOe8K3DbtsS8/jTvlVpxrbg/aL3xbWYH96jNvknj6DxATgzn6eO+NZce0Go9tMuIiSubPxX1hOs5tGmi1t2x5GfbZ/8MuX4zpNQBz4VWYRv6daTaxsZh+J2L7ngCrfsbOn4Nd9AH20/fgoMNw+v8VDjtG3+fdsK/NhJXfYy65CdMhjajEZJxx93rf5zeegw2/wQVXBbxuoVaxVlbgvjQdWrXGnHR60Nc3rdtjBg3DfvQmtt+JmM4HBD2GSGU3ZuE+cjfkrMdccCXOcUMASDj9fIp/+B/29VnY9vvX66wSCQ/qFSo1LF1bSJWFfh0j9+iTXf2Ld1Z3wMm+dHaqD86Ak3GuuwPyc3En3YRN/zGg69n8XNy3X8C9dSz26QdgSz7mrDE498/EueiaHRIK8M5GO2ePhXW/ekehpNbs5nzcB/6O/XIJ5oyLMGOu9zWh+CNjDCbtQJxLbsK57xnM6RdA9npvGvyES3Hfm40t2OR3mCHHXfYp9pN3MYNPwznm+Or7TUws5tKbMcPOxX72Ke6Dt2ELNvsYqcd+/A5kZeKMusy3nz1z6khokYj74nSsW+VLDJHGrvwO996bYMsmnOvvqk4oYNvf/dHXQpv2uE89gN2Y5WOk4gclFVLD4jUFtGsWw/4tY/0OxRfWdXFfnO6d+x92rt/h1Ik5+AicCVMhPgF32t9xP5tfr69vrcX+uIKqJybjTrjEK9bs1BXnujtw7nkS58ThmMZNd/8iPfrAwYdj335BbzRrya7NwL33Zshcg/O3CTh/PTNkd9NMsxY4Q8/CufcpnKsmQut22LdfwL1lLO7T07DpP2BtEOs/QpRdk459/jE44BDMiNE7fN0Yg3PqKMxlt8Aa7/tv160Oepzb2fxc7JzZcPgxmEOO9C0OE5eAOWsMrEnHLv7YtzgihbvkY9wH/wFNmuFMfABz4KE7PMbExeNcOcGrB3p8MraszIdIxS9KKqTaxuIKfsjZynGdmoXsm5RAs0vnwa8rMWeODomWsXVlWrfDmfgApB2EfeYh3DdmeV2Y6sBuLcH99D3cO67GffB2WPkd5oThOJOmE3XN7ZjuR2Kc2v3T4g20ugzKy7Cvz6pTXJHAfvUZ7n3jwYBz632YI0KzePfPTFQU5vBeRN3wT5y7H8f0/yv22y9x7xuPe/f1XveYslK/w/SFLdyyrTC7Oc7lt+z2eJhzVF+ccZOhshJ3yq3YFcuDGOnv7KvPgOsGtTh7V8xR/eCAQ7yi7cLgtlKOFNatwn11Jva5R+CA7jgTpmJS2uzy8aZVG5xLbobM1dhZj+iDgwiipEKqLVlTgIWI7fpkiwuxb8yCLgdjeg/wO5x6Yxo3xbn+LsxxJ2Hffw33iSnY0q17/To2cy3uC0/gjrsY+9JT3gDAi6/Due8ZnBGj93nwnklthznhNOyyT7AZP+3TazR01lrcua/iPjEZ2nbEmTgN06Gz32HtE5PaDmfUpThTZ2IuuBKsxT7/GO64i3Fn/wubvd7vEIPGVlV5hdmFBd6MmVp0MTL7d8X5+zRIaYP72D24H70Z1Ddt9scV3rG7v47AJKcEbd1d2T5pm9ISr7mG1CtbWoL72L3Yj97EDDgZ59o7MAlN9vg8c8iRmNPOw/5nEfbjt4MQqYQCFWpLtcVrCuiaFEfrpqFxNjvY7FsvQHERzrnhVZxdGyY6Gs6/Etp0wM6egXv/eG8Cd+J+u32erayEbz73Zkus/A6iG2GO6ufVm+zftf7iO/lsb6DVC094Rdsh0Ao3VNiKcuysR7GfL/CK3kdfEzL1E3VhYuMwxw3B9jsJ0n/0Crvnz/HaUx58hDcX5tCeDfpnwb72LPz8LWbMDZiOXWr9PNMyCeeWKbgzH8K+OhPW/wbn/w0THdgCbq84+ynYLxUzJHS6+5i2HTADT8HOe8ebtF2P/zZFMpubjfvoPbDhN8y5V3h/J/eCGXoWdu0q7GvPeoXbBx0WoEglVGinQgBYV1BGRn5ZxBZo2zUZ2IUfYAYMDelBU3VhjMEZdCrOtbdDbrZXwL3q550+1m7Ox33nJdwJl3ifpOblYM68yCu8HnN9vf/SNnHx3tno337FLvygXl87nNmCTbjTbvMSiuHnYy65sUEkFH9kjMF0PRjnsnFeYfdp58L6tbiPTcKdeDnu+69hC7f4HWa9cz+fj533NmbQqTj7sDNqYmNxLrvFmzC9dB7uQ/8I+PEf+8m7sOE3nFGXhtzPoTn1HGjWAvfFJ+t8xFPApv/g1W5tysW57s69Tihg2++ci6+F1La4T92PzcsJQKQSSpRUCACLVxdggL4d91BY2wB5xdlPQpOm3huaBs50PxJn/P0QG4c7dSLuFwuBbYXXK7/DnX4/7vix2HdfgnadcK6+Hefe6ThDzsQ0DVzSaXoeCwcein3r3w3yTeTesut+xZ10M/y2CueK8Tgnn93gdtD+zDRviXPKKJwp/8K5Yjwkp2DfmIV7y8W4Mx7Crvq5QZzPtmszsLMeg25/wYy4eJ9fxzgOzmnnYS65CVatxJ18M3b92nqM9Hd2Ux723ZfhsKMxhx4VkDXqwsRvK9pe/Ys3uFT2mbvsU9xpt0F8Y69+4uDD9/m1TFwCzlV/hyoX9/F7VbjdwOn4k2CtZdHqQrqnJJCU4H//82Czn30Kq37GjL6uVmdFGwLTpgPOhAdwn5yM/dc0tqz6Cffn7yBzDSQ0xgw8BdP/r5hWuy7Gq/eY/jjQ6vXnvNaEEcp+8wXuv6Z5v9RvuW+nLXkbMhMVBUf2IerIPtj1a72Bep/Nx34+Hzp2Yeuwkdju4Xk0yhYW4D4+2eugc/kt3tHEOnKOOR67X6q3uzPlFpxLx9V7Vyb76jNQVRUSxdm7Yo4+DrvoQ+ybs7A9emOaRObO+76yrusVvH/wOhx4KM4Vt+65g18tmJQ2OJfciPvoPdjnH4WxNzb4D0gilXYqhFWbylhfWB6RBdq2uAj7+nPetOcGVJxdG6ZpM5wb/ok5djCln86BqCjMhVfj3P8sztljg5pQVMfUuj1m8DDs0nkRWbRtrcX94HWvG1Dr9jh/nxZxCcWfmTYdcM69wivsPvcKKC+j4JFJXrvK0hK/w9srtqoK9+mpXo//v03ANGtZb69tOh/gFXAnpeA+cjfuvHfqbVfH/vQ/b8jiX8/c54YMwbD9gwlKirFv/tvvcMKKLd2K+8Rk7AevY44bgnPdnfWSUGxnDj0KM+wc7BcLvWN00iApqRAWrS4g2oHe7SPw6NPbL0BRIc65V9S6DWpDYqIbYS66huSn38K57SGcfidiYv2dUWJOicyBVraiwpuc/PpzmJ59ccbdi2mR6HdYIcPEJeAMGIpz16M0vfRG+PZLr61qGJ3Ttm88Bz+uwJx/ZUCKiU3ifji3ToHDjsbO/hf23497zRbqwFZWerN7klMwQ86sp0gDx7Tt6BVtL/4Qu/oXv8MJCzZvo9eqesVyzKjLMOf/rV520P7MDD0bDu+FffUZ7M/f1vvri/8i712U1OBay+I1BRzRuglNY8PvKEFd2LWrsAvex/QfErbtOeuDMYao5FYhsx1dPdBqbQZ20Ud+hxMUtnAL7oO3Y5d9gjn1HG+CckxkDqDcE2MMCUNH4FwbvInx9cH9YiH2o7cwA4biHDsoYOuYuHicv433Ou8s+hD3/92BLdr3Am776R+Ks8PkZ9Kceg40bb7tgwkVbe+OzfjJm5Cdl41z7T9wBp0SsN8FxnFwxlwPrdrgTr8fm7cxIOuIf5RURLgfc7aSV1IZcUefahZnn+93OPInkTTQymauwZ10E6xJx1x2C86wc0ImwQtl5i+BnRhfn+xvv2JnPeLNwDl7bMDXM46Dc/oFmDE3QMaPuJPHYTes2+vXsZvzsO+8DIf0xBx2dAAiDQyT0Bhz1sXw60pvoKnslPv5AtwH/g6xcV5BdvceAV/TxCfgXDURKsq941blKtxuSJRURLiFqwuIjTIc3S4yCpS3s5/Ph4yfMGdciGkcWdceDqoHWpVtbdADrez/luNOvgUqK3HGTcY5qq/fIYWVQEyMr2+2qMCrkUlo4hW+BniWxB85vQfg3DQJtpZ4icX3X+/V8+2rz0JVJc6oSwMTYACZY/pD14Oxb8zCFhf6HU5Isa6L++a/sTMehM4H4Ex8ANO6fdDWN6ntcMbeCGvSvSN6DaCjm3iUVESwStey7LdCjmnXlLjoyPlRsCVF3tCp/bth+gTuGILUjWnbATPoVOySj7G/rvQ7nHplrcX96C1vsFRKG68gWwO79kn1xPh+J9ZpYnwgWLcK9+kHYHOeV5jdvP4Ks2vLdDnIK+BOTMb9v7twP32vVs+zP3+H/c9CzJAzMK1aBzjK+vd70XaRN9hUALBlpbjT78POfQXT70ScG+7ypUuWOfwYzKmjvK5u8+cEfX0JjMh5Jyk7+GZDMYVlVfTrFFkF2vadl6CoAOe8v0VkcXY4MaeOgmYtcV94ssEUbdvKCm9C9qvPQI/eOLdMwbRM8jussGaiozEXXIUZeQms+A/u/eOx+f6f17ZvPA8/fIM59wpM5wN8i8MktcIZfx8c0hP70lO4Lzyx2wJuW1mJ+9J0SGqFGTIiiJHWL9Nuf8yAk7EL38euyfA7HN/Z/Fzc+8fD119gzh7r/Z0J4s7Zn5lTRnlNBV6ZgV35nW9xSP3RO6oItmh1AU1iHI5oHTnHf+y6X7GfzsEcPyTiW3WGA69o+2Jvm7wBDLSyhQXe5OMlH2NOGelNRPa521ZDYYzBGTwM55o9T4wPBnf5EuyHb2COH4LT70Tf4tjOxCXgXDkBc9IZ2AXv4/7fXdjiop0+1i6YA5lrcEZeEvY/n2bYOdCkWcRP2ra/rvQKsnM24FxzG84Jp/leu+UVbt8Ayam4T96Hzc/1NR6pu6AlFd988w3XXXcd11xzDW+99dYOX8/NzeWuu+7illtu4eabb+arr74CICcnh/POO49x48Yxbtw4nnrqqWCF3KCVVbp8sa6QPh2a0igqMopCrbW4L0yHxo0xw1WcHS7M0cdBt+7YN56vUxcbv9n1a3En3wyrVmIuucmbhKydsnpnDtn5xPhgsutWY5992Jt/E0L1CMaJwhkxGjP6Olj5vVdnkZVZ4zFV+bnYt1+E7kfC4cf4FGn9MQlNMCNGw6qfvUGnEcj9zyLcqROhUQzO+KmYQ3r6HVI1k9DYK9wu31a4XVHud0hSB0H5jea6LjNmzGDixIk89NBDLF26lHXranaieP311+nduzf3338/119/PTNmzKj+WmpqKlOnTmXq1KlcdtllwQi5wfvPuiJKKy39OkZO1yf7+QJI/wFzxkX1OtRHAqv6bPTWYuybz/sdzj6x3/0Xd8otUFaKM+5enGOO9zukBm37xHj274r91zTct/4dtE+pbXGhV5gd3xjnivG+Hi/ZFefYQTg33QPFhbiTb8b+uKL6a0WzHoPKCpxzLvX9k+z6YnoNgC4HYV9/bpe7Mw2RdV3ct1/EPv0AdOziFWS37eB3WDswrdt7rWZX/4J94QkVboexoCQV6enppKamkpKSQnR0NH369GH58uU1HmOMoaTEm45aUlJCy5bBL2iLJIvXFNAyPpq/tErwO5SgsCXF2NdmesXZxw72OxzZS95Aq1Oxiz8Kq4FW1lrcT97F/b+7ITnFK8j28Wx9JDFNm+HceDfm2MHYOa94ffHLSgO6pnWrcP81DfJzvU5PITy80HQ92Ouc1SIJ9//dgbvgfezK7yld+CHmxDMwrdr4HWK9MY6Dc+4VUFToDTyNALasFPv0A9j3XsYcO8j7u9C0ud9h7ZLp0Rtz8tnYpZ9gF77vdziyj+p/ZOJO5Ofnk5T0eyFiUlISv/xS843BWWedxT333MMHH3xAWVkZt99+e/XXcnJyuOWWW4iPj2fUqFEcdNBBwQi7wSoqq+K/64sZ2q0FUU7D+CRqT+y7L0HhFpxrbteRkzBlhp2DXb4I98XpOOPvD/nvo62spPDJqdiP3oIjeuGMvRETG+d3WBHFRDeCi66BNh2wr83EvT8b5+rbAlYYb996Ab77ypuY3SX0f0+Z/VJxxt+P+/QD3ifE8Y1x9kuBoWf5HVq9M+33x/T/K3bB+5Qc8BdcN7ifhpc2bYpbGKTWttaSP/89bMbPmBEXY04cHha7TmbYOd5Q2pefxrbthOl6sN8h1WDzcij9eQVuQXCP4VakdYPElKCuua+MDcI+02effcaKFSu44oorAFi0aBHp6emMGTOm+jHvvfce1lpOPfVUVq5cyRNPPMG0adOoqqqitLSUpk2bsmrVKqZOncq0adNISKj5Cfu8efOYN88bcjNlyhTKy/09lxcdHU3lbrpr+Lnmu99lMeWTdP418jAOSq3bMaBgX+e+rFexJoP8G0cTP/hUmv3tlqCsWVfh8Ofqx5pbF3xAwcP/pOmV40k4YVhQ1twXbmEBm++fSMV3X5Fw5oU0OfeyoCVB4fK9DPZ6ZcuXsuXBOzDxCbSYMIVGe/mGZU9rli77lC1TbyP+hGE0u3L8Xr32vq5ZX2xVFUWzHqPk3dkkTriPRkGclxLMnx23uJC86y7AzcsJynp+MvEJNL/hTmLD7HvpFheSP24sdmsJiQ/MJCppv4CvuTvWdSlf8R9K3n+D8v8uAx+K/eMHDqXZNbcFdc0//7nGxMTU7nmBCuiPkpKSyMvLq76dl5e3w/GmTz/9lIkTJwLQrVs3KioqKCwspHnz5jRq5J1J7dy5MykpKWzYsIG0tJqdewYPHszgwb8fa8nN9beLQHJyctBjqO2ac79fT2qTRiRHlZKbW7dplsG+zr1dz1qL+/gUiE+g7K8j9inWUP5ehut6+7qm/cuR0PVgCmc9RnG3Q/a6NiYY12mz1uE+cjfkb6TZdbdT3P0oyvLzA7rmH4XL9zLo6+1/AObWKbiP3kP+36/EXHz9Xg0b3N2aNnMt7v/dA/t3o+z0i+rtzyKof66nnosz6DQadegY+t/Lurjz/0iyLps2bwremkDLFi2DumbS/mnkl5ZTGIbfS3v5eNzJN5N77y04N9+LabTruqRA/fzY4iLssk+wC96HnPXQtDlmyAhaDh7K5qLiel9vdxq3bef7v+lt2tTuOGRQkoq0tDQ2bNhATk4OiYmJLFu2jGuvvbbGY5KTk/nuu+/o378/69ato6KigmbNmlFQUECTJk1wHIfs7Gw2bNhASkp4bAOFovytlXyXXcKIvySFxXZoXdn/LIKV32MuuNKXAT9Sv7YXbbt334B983nM+Vf6HVIN9oevcZ+8H6KjcW6+l/hj+lLs8wcc8jvTrhPOxAe8LjNP3Y+74TfMqaPq9G+hLSnCfXwSxMZ5A+528wYo1JmExn6HEHAmLoHo5GRMfHBbqQd7TadJMygNz397TNsOOBdfj/vkFOxL0zEXXh20te3aVdgFc7FfLIDycq+D27BzMD36YBo1olFyMibI/6ZHJSZDmPweCUpSERUVxZgxY5g0aRKu6zJgwADat2/P7NmzSUtLo2fPnlx44YVMnz6dOXO8yYpXXnklxhh++OEHXnnlFaKionAch0svvZQmTSJnrkJ9W7qmANfCcZ0a/htsu7XEGzDWsQum7wl+hyP1pHqg1afvYfudiOnYxe+QAHDnz8G+/DS0bu/V7iS18jsk2QnTrAXOjfdgn3/Uq7XKWgejr8XE7P08Bq8w+0HIy8G5aZKGGIrUE3NkH8xfR2Dffw23UxeceoLC3wAAIABJREFU44YEbC1bWYH97zJvsnfGTxATgzmmP6b/UEyHzgFbtyEKSlIB0KNHD3r06FHjvpEjR1b/f7t27bj77rt3eF6vXr3o1atXwOOLFItWF7B/y1jaNw/vgUa1Yd99CQo241z1d4wT5Xc4Uo+8ou3FuC886XvRtq2sxM7+F3bBXDjsaJxLbsTERUZXtXBlGjWCi6/3CrjfmIXdmIVz1URMi71LCuw7L8G3X3oTs0OsqFQk3Jnh52F/W4V98SmvcDvtwHp9fZu/EbvwQ+ziD6FwC7Rq7U0a7zMI01gfXu+LWv0mnjt3LgVBrnaX+pdVWM7KvFKOi4DZFDZzLfaTdzF9T8Ds383vcKSemYQmmDNHw68rscs+8S0OW1zkTSZeMBdz0hnexGIlFGHBGIMz5EycKyfAht9wJ92MXZNR6+fbrz7DznkFc+xgTP+/BjBSkchknCicS26GxGTcJ6ZgN9e9Ns1ai/1xBVWP34s7/lLs+69C5wNwrrsT5+4nvEnjSij2Wa2Sim+//ZarrrqKKVOmsGzZMioqKgIdlwTAojVeYtivgR99stbivjQd4htjTr/Q73AkQExvfwda2ez1uJPHeTU7o6/zJhVrRyzsmMN74dx6HzgG9/5bsf9dtsfn2A2/4T7z/6BTV8x5V0REfZqIH0zjJjhXToStxV6NReW+vf+0JcW4n7yH+4+rcB+8HX75HnPS6Tj3PkXU1bdhuvcI+Tbl4aBWf4K33norjz/+OIcffjhz5szhsssu48knn+SHH34IdHxSjxavLuDg/eLZr3H4FhLWhl2+GH7+FjP8fEzThp1ARTKvaPsKKC7Cvv3voK5tf1yBe+/NUFyIc9M9OMcOCur6Ur9M+/1xJk6DdvvjPjkF973Zu5zqa0uKcR+7F2JicP42HtOodq0WRWTfmHadMKOvg4yfvLq1vWAz1+D++3HcWy7GvvwUxCd4nd/un4lz5kWYZDX+qU+1rqlo2rQpQ4YMYciQIaxZs4ZHH32U+fPnk5yczKBBgxg6dChxcRrsFKpWbypl7ZZyLj+qYf8FsqXbirM7pGGOO9HvcCTA/jjQyvY9AdMhbc9PqiN34QfYF5+E1HbeILX9UgO+pgSead4S5+ZJ2Oce8aYub1gHF11do4Dbui7ujAchN8ubUJy4+x76IlI/nKP64q5Jx374Bm7HLjj9dv373VZWYr/+zKtzW/k9NIrBHNUPM2AoplPX4AUdgfaqUPvbb79l8eLFLF++nLS0NK6++mqSk5OZO3cu9957L//85z8DFafU0aLVBTgGju1Qt2F3oc6+Oxs25+NcMV5HUSKEGX4e9ssl3qTtW6YEbAvbVlVhX30G+8m7cEhPnEtvxsSrfqIhMY1iYOyN0Lo99q1/Yzdu8Bo9NPfmKtn3Xob/LceMugzTrbvP0YpEFnPGBdsKt5/Etu2I6XxAja/bzXm/F15v2QTJKZgRo726J7WUD4paJRWzZs1i2bJlJCQkcNxxxzFt2jQSExOrv961a1cuvvjigAUpdWOtZfGaQg5PbUzzuKA1/Ao6u34t9pN3vOLseu4SIaFre9G2ffZh7GfzMQE4imRLinCfmgrff4054TTvF5WS1gbJGIM5+Wxs63a4Mx7CvfcmnKtuo3TVj9h3X8b0HogZeLLfYYpEHONE4Vw2DveeG3GfmIJz24PYpCTsz9/izp8DX38O1sJfeuBcOBS699C/00FWq3eYFRUV3HzzzXTpsvN+8NHR0UyZMqVeA5P683NuKTnFFZxzaLLfoQSMV5z9FMTGYc5QcXakMb0HYBd/iH39WewRx2AS6q97h81Zj/vIPbBxA+bCq3e77S4Nh+nRByc5BffRSbj33UpBVJQ38+b8v6kwW8QnpnFTnCsn4k4Zh/vwneQB7m+/QkITzODTMMcPwbRq7XeYEatW5wROP/10UlNrnhsuKioiP//39l5t27at38ik3ixavYWYKEOv9g23TZr9cin89D/M8AswTZv7HY4EmXEcnHMvh6JC7Nsv1tvr2p+/w713HBRuwbnhbiUUEcZ0SMOZ+AC07YjZPjF7H4bkiUj9Me33x1x0LaxbjWkUgxl9Lc7UmThnXayEwme1SiqmTp1aI4EAyM/P54EHHghIUFJ/qlzLkrWF9GzbhIRGDXMb0JZuxb4yAzp0xhx/kt/hiE9MhzTM8UOw8+dif/u1zq/nLv4I96HboVkLnIkPYA7QGfpIZFok4oy/n+QnXsUkqTBbJBQ4Rx+H89ALJD7wDM6xg5Xsh4haJRXr16+nQ4cONe7r0KEDmZmZAQlK6s+32SVsKa1q0APv7JxXYHMezrlX6PxkhDPDz4fGTXBffHKXLUH3xLpVuLNnYGc9Cgce6k3s1qdfEc04DiYu3u8wROQPTOMmOooYYmqVVDRr1oysrKwa92VlZdG0acPuJNQQLFxdQEIjhyPbNvY7lICwG9ZhP34Lc+wgFWeL90vmzIsg/UfsZ/P3+vl2awnuo5Ow897GDDoV55p/YBIa5t8dERGR+lSrQu0BAwYwbdo0Ro0aRUpKCllZWcyePZuBAwcGOj6pg/Iql89/K6RX+6bERDW8SZHVk7Nj4zBnXOR3OBIiTJ9B2EXbirYPP6bWSYHdmIX76D2QnYk5/0qc44cEOFIREZGGo1ZJxfDhw4mOjub5558nLy+PpKQkBg4cyCmnnBLo+KQO/ptZTEmFy3GdGujRp6+WwY8rMOdchmnWwu9oJEQYx8E57wrcSTdh33kRM+rSPT7Hrvwe94nJ4Lo4192JOeiwIEQqIiLScNQqqXAch2HDhjFs2LBAxyP1aNGaAprHRXFoSsMb0GXLSnFfmQHt9scc/1e/w5EQYzp2wRx3Enb+HG/SdrtOu3ysu/QT7POPQXIKzjW3Y1LaBC9QERGRBqLWk9AqKytZv349BQUFNe7v3l0dUUJRSUUVX2YWcUJac6KchlfIZOe8Avm53lTjKBVny47M6Rdg/7sU98UnccZN3qGgz7pV2DdmYT98Ew46DOfyWzGNG27bZRERkUCqVVLx008/8eCDD1JRUcHWrVuJj4+ntLSUpKQkHn300UDHKPvg89+KKK+y9GuAR59s1jrsR29heg/AdDnY73AkRJnGTTGnX4h9/jHsFwswvQZUf82WluD+60FY8R/MgKGYkZcqORUREamDWlXvPvfccwwbNoyZM2cSHx/PzJkzOfPMMznxRA2CClWLVxfQqnE0ByY3rDaIXnH20xATgxkx2u9wJMSZvifA/t2wrz2L3VoCgM3Nxp1yK3z7Jebcy71WxEooRERE6qTWcyqGDh1a477hw4czZ86cgAQldbOltJJvsorp17FZg+vhXPb5Qvjha8xp52GatfQ7HAlx1ZO2CzZj33mJ8h//h3vvzbApF+e6O3AGnOx3iCIiIg1CrY4/JSQksHXrVho3bkyLFi1Yt24dTZo0obS0NNDxyT5YurYQ19Lguj7Zgs0UznwY2nXC9B+65yeIAKZTV0y/E7GfvsumBXMhcT+ca27DpLbzOzQREZEGo1ZJxTHHHMPXX39N3759GThwIHfddRdRUVH07t070PHJPli8uoAOzWPo2CL8x9Zba2HVz9gFc7FfLgHX4tw8ScdVZK+Y0y/ArlhOTIf9qRx7I6axBneKiIjUp1olFaNHj67+/1NPPZWuXbuydetWDjtMvdxDzcbiCn7YuJXzDksO66NPtqwMu3wRdv5cWJsBcfGYfieRePp5bI5Xhx7ZO6ZJM5zJT9EitTV5eXl+hyMiItLg7DGpcF2X6667jgcffJBGjRoBcOCBBwY8MNk3i1d7LX/7dQzPo082ZwN24fvYJfOgpAjadMCcdwWmV39MXALRycmQm+t3mBKGTKOYsE60RUREQtkekwrHcXAch4qKiuqkQkLXojUFdEuKo3XTGL9DqTXrVsF3X+HOnwvffwWOgzm8F2bAydDtL3ojKCIiIhLianX8aejQoTz00EOcfvrpJCYm1niTl5KSErDgZO+szi/h101lXHJkK79DqRVbVIBdOg+74H3IzYbmiZhTRmKOOwnTIsnv8ERERESklmqVVDzzzDMA/O9//9vha7Nnz67fiGSfffzzRhwDx4b40Se7+hfs/LnY5Yuhohy6dceccRHmiF6Y6FoPeRcRERGREFGrd3BKHEKftZZ5P2+ke0oCifGh98bcVpRjly/2diV+XQmxcZg+AzEDTsa07eh3eCIiIiJSB6H37lP2SXp+Keu2lDL8wFS/Q6nB5mZjF36AXfIRFBVCajvMqMswvQdgEhr7HZ6IiIiI1INaJRX/+Mc/dlkse9ddd9VrQLJvlqwpJNox9G7vf/9967rww9e4C96H/y0HY+DwY3D6D4UDD1XhtYiIiEgDU6ukYuDAgTVub968mfnz59OvX7+ABCV77/ucEg5p04wmsf4NhbPFRdhln2AXzIWcDdC0OWboWZjjhmASk32LS0REREQCq1ZJRf/+/Xe4r1evXjz++OOMGDGivmOSvVTpWlZvKmNER386Jtm1GdgF72O/WADl5dDlIMywczFH9sFEqw2xiIiISEO3zzUViYmJrFmzpj5jkX20dnMZFa6lU9aPuBvzg7dweTn533yO+/N3EBOLOaY/pv9QTIfOwYtBRERERHxXq6Ti008/rXG7vLycL774gm7dugUkKNk7GfmlALR980ns1uBOm3Zbt8eMHIvpMwiT0CSoa4uIiIhIaKhVUrF48eIat2NjYznggAM4+eSTAxKU7J2M/FLiq8rpcHBX3FFTg7iyIalzF/Lyg7g7IiIiIiIhp1ZJxR133BHoOKQO0vO20rlwHTH7p1EW5EnUxnGCup6IiIiIhJ5avSNcuHDhDvUTq1evZtGiRQEJSmqv0rWs3lxGWuFvRLfp4Hc4IiIiIhKBapVUzJ49m6Skmp+AJycn8/LLLwckKKm937aUUeFCWmEmUW3a+x2OiIiIiESgWiUVW7duJSEhocZ9CQkJFBcXByQoqb3tRdppheuIbqudChEREREJvlolFe3atePzzz+vcd9//vMf2rVrF5CgpPbS80qJp5JUsxXTrIXf4YiIiIhIBKpVofZ5553H5MmTWbZsGampqWRlZfHtt98yYcKEQMcne5CRX0rn8jyclLYYY/wOR0REREQiUK12Kg488ECmTZtGly5dKC0tpUuXLkybNo0DDzww0PHJblRtK9LuvHktJqWt3+GIiIiISISq1U5FRUUFLVq0YPjw4dX3VVZWUlFRQaNGjQIWnOzeb1vKKK+ypOX+Al3/4nc4IiIiIhKharVTcc8997Bq1aoa961atYpJkyYFJCipnfTqIu1M0E6FiIiIiPikVknF2rVr6dq1a437unTpssPsCgmu9LxS4h2X1ltzMalKKkRERETEH7VKKhISEtiyZUuN+7Zs2UJsbGxAgpLaycgvZX9TgoOFVq39DkdEREREIlStkopjjjmGhx9+mLVr11JWVsbatWt59NFH6dWrV6Djk13YXqSdVrYRWiZjYuP8DklEREREIlStCrVHjRrFrFmzmDhxIhUVFcTExDBgwABGjRoV6PhkF7YXaXfevBpS2vgdjoiIiIhEsFolFTExMVxyySWMHTuWwsJCNm3axMKFC7nuuuuYPn16oGOUnagu0l7/PeZQdX4SEREREf/UKqkAKCgoYMmSJSxcuJDVq1dz0EEHMXr06Fov9M033zBz5kxc12XQoEE12tMC5Obm8thjj1FcXIzrupx77rn06NEDgDfffJNPP/0Ux3G4+OKLOfzww2u9bkOVkV9KXJShTf5vkHKC3+GIiIiISATbbVJRWVnJl19+yYIFC1ixYgWpqakce+yx5OTkcMMNN9C8efNaLeK6LjNmzOC2224jKSmJCRMm0LNnT9q1a1f9mNdff53evXtz4oknsm7dOiZPnkyPHj1Yt24dy5Yt48EHH2TTpk3cfffdPPzwwzhOrcpBGqyM/FI6J7g4WA2+ExERERFf7TapuPTSS3Ech+OPP56zzz6bzp07A/DRRx/t1SLp6emkpqaSkpICQJ8+fVi+fHmNpMIYQ0lJCQAlJSW0bNkSgOXLl9OnTx8aNWpEq1atSE1NJT09nW7duu1VDA1JlWv5dVMZJyUUeXcoqRARERERH+324/6OHTtSXFxMeno6GRkZFBUV7dMi+fn5JCUlVd9OSkoiPz+/xmPOOussFi9ezBVXXMHkyZMZM2bMTp+bmJi4w3MjTXWRdmk2REVDUiu/QxIRERGRCLbbnYo777yTjRs3snDhQt59911mzpzJoYceSllZGVVVVbVexFq7w33GmBq3ly5dSv/+/Tn11FNZuXIljzzyCNOmTdvpc3dm3rx5zJs3D4ApU6aQnJxc6/gCITo6OmAxfJGTDcCBhZlEpbYledsOUCDX3JVgrxkJ1+jHmpFwjX6sGQnX6MeakXCNkbJmJFyjH2tGwjX6sWYkXGNd1txjofZ+++3HiBEjGDFiBD/99BMLFy7EGMO4ceMYMGAA559//h4XSUpKIi8vr/p2Xl5e9fGm7T799FMmTpwIQLdu3aioqKCwsHCH5+bn55OYmLjDGoMHD2bw4MHVt3Nzc/cYVyAlJycHLIYVazYSF+2QvOZbqlqlVq8TyDV3JdhrRsI1+rFmJFyjH2tGwjX6sWYkXGOkrBkJ1+jHmpFwjX6sGQnXuLM127Sp3eiCvap2PvDAA7n88st56qmnuPjii1m7dm2tnpeWlsaGDRvIycmhsrKSZcuW0bNnzxqPSU5O5rvvvgNg3bp1VFRU0KxZM3r27MmyZcuoqKggJyeHDRs20KVLl70Ju8FJzy+jc8sYojaux6SqnkJERERE/FXrlrJ/FBMTQ9++fenbt2+tHh8VFcWYMWOYNGkSrusyYMAA2rdvz+zZs0lLS6Nnz55ceOGFTJ8+nTlz5gBw5ZVXYoyhffv29O7dmxtvvBHHcRg7dmxEd37yirRLOaltDFRWqkhbRERERHy3T0nFvujRo0f13IntRo4cWf3/7dq14+67797pc8844wzOOOOMgMYXLtYVlFNeZUljCwBG07RFRERExGeR+5F/mMrYNkm7c0mWd4eOP4mIiIiIz5RUhJn0/FLiog1tcldDfAI0beF3SCIiIiIS4ZRUhJmMvFI6t4zDyVkPrdrs0JpXRERERCTYlFSEkSrXsmpTKWmJcZCVqc5PIiIiIhISlFSEke1F2p2bR0H+RnV+EhEREZGQoKQijGwv0u7iFnh3qPOTiIiIiIQAJRVhpLpIu3ADAEY7FSIiIiISApRUhJGMvFL2bxmHk53p3aGdChEREREJAUoqwsT2SdppiXGQvR5aJGLi4v0OS0RERERESUW4yCwop6zKkpYYh83OVJG2iIiIiIQMJRVhIn17kfa2nQrVU4iIiIhIqFBSESYy8kuJjTK0iSqF4kLVU4iIiIhIyFBSESYy8r0i7agcdX4SERERkdCipCIMVLmWVfmldEmKw2ap85OIiIiIhBYlFWEgs/D3Im2yMyEqCpJT/A5LRERERARQUhEWMvJ+L9K22eshORUTHe1zVCIiIiIiHiUVYWB7kXbbZjHeToWOPomIiIhICFFSEQa2F2k7WK+dbKqKtEVEREQkdCipCHFVrmXVplLSkuJgUy5UVminQkRERERCipKKELe+sJzSSrtt6J3X+UntZEVEREQklCipCHHp24q00xLVTlZEREREQpOSihCXkV9KTJShXbMYyF4PsfHQPNHvsEREREREqimpCHHVk7Qdg93W+ckY43dYIiIiIiLVlFSEsO1F2l0SY707sjIxOvokIiIiIiFGSUUI216knZYYh60oh/yNoHayIiIiIhJilFSEsIz834u0yckCa0Gdn0REREQkxCipCGHp24q02zeP/UM7WR1/EhEREZHQoqQihGXklbJ/y9jfi7RBOxUiIiIiEnKUVIQo11pWbSrzjj6Bt1PRvCUmPsHfwERERERE/kRJRYhaX1BOaaVbnVTY7PUaeiciIiIiIUlJRYhK31ak3WX7TkVWJkZHn0REREQkBCmpCFEZfyjStsWFUFSgegoRERERCUlKKkKUN0nbK9Imez2gzk8iIiIiEpqUVIQg11pW5ZfVrKcA7VSIiIiISEhSUhGC1heWs/UPRdpkZYLjwH4p/gYmIiIiIrITSipCUEben4q0szMhOQUT3cjHqEREREREdk5JRQiqMUmb7e1kdfRJREREREKTkooQlJFfSqcW2yZpuy7kZKpIW0RERERClpKKEPPnIm0250F5uXYqRERERCRkKakIMduLtLskba+nUDtZEREREQltSipCzPYi7ep2slmZ3he0UyEiIiIiIUpJRYjJyC+lkfN7kTbZmRATCy2T/A1MRERERGQXlFSEmIz8Ujq1jCXaMcD2zk9tMMb4HJmIiIiIyM4pqQghrrVk5Jf9Pp8CIDsTo6NPIiIiIhLClFSEkA2FFTUmaduKCsjNgVQlFSIiIiISupRUhJCM/JpF2uRmgXVBnZ9EREREJIQpqQgh24u0O7T4Q5E26PiTiIiIiIS06GAt9M033zBz5kxc12XQoEEMHz68xtefffZZvv/+ewDKy8vZsmULzz77LAAjR46kQ4cOACQnJ3PrrbcGK+ygSv9zkXZ1O1ntVIiIiIhI6ApKUuG6LjNmzOC2224jKSmJCRMm0LNnT9q1a1f9mNGjR1f///vvv8+vv/5afTsmJoapU6cGI1TfeJO0Szm+U7Pf78xeD02bYxKa+BeYiIiIiMgeBOX4U3p6OqmpqaSkpBAdHU2fPn1Yvnz5Lh+/dOlS+vbtG4zQQkZWYQUlFb8XaQPY7EwNvRMRERGRkBeUnYr8/HySkn4f3paUlMQvv/yy08du3LiRnJwcunfvXn1fRUUF48ePJyoqitNOO42jjz56h+fNmzePefPmATBlyhSSk5Pr+Sr2TnR09F7F8E3eRgCOTEslOdnbmdi4MYuYI/vQvJavs7dr1odgrxkJ1+jHmpFwjX6sGQnX6MeakXCNkbJmJFyjH2tGwjX6sWYkXGNd1gxKUmGt3eG+XQ1zW7p0Kb169cJxft9Eefzxx0lMTCQ7O5t//vOfdOjQgdTU1BrPGzx4MIMHD66+nZubW0/R75vk5OS9iuHrNRtp5Bia2a3k5pZiS4pxN+dT1iKp1q+zt2vWh2CvGQnX6MeakXCNfqwZCdfox5qRcI2RsmYkXKMfa0bCNfqxZiRc487WbNOmdrW9QTn+lJSURF5eXvXtvLw8WrZsudPHLlu2jGOPPbbGfYmJiQCkpKRw8MEHs3r16oDF6pc/T9Imez2gzk8iIiIiEvqCklSkpaWxYcMGcnJyqKysZNmyZfTs2XOHx61fv57i4mK6detWfV9RUREVFRUAFBQU8PPPP9co8G4IvEnapTvWU4A6P4mIiIhIyAvK8aeoqCjGjBnDpEmTcF2XAQMG0L59e2bPnk1aWlp1grFkyRL69OlT42hUZmYmTz31FI7j4Louw4cPb3BJxc6KtMnOBOPAfq39C0xEREREpBaCNqeiR48e9OjRo8Z9I0eOrHH77LPP3uF5BxxwANOmTQtobH5L3zZJu0uNpGI9JLfCNGrkU1QiIiIiIrWjidohICO/lGjH0L55bPV9XjtZHX0SERERkdCnpCIEZOSX0qlFLI2itk3Sthay16tIW0RERETCgpIKn9ltk7Rr1FNszoeyUg2+ExEREZGwoKTCZ1lFFRRXuHRJ+lORNmB0/ElEREREwoCSCp+l53lF2jXbyXozKrRTISIiIiLhQEmFz7YXaXf4Q5E2WZkQEwMtk/wLTERERESklpRU+Cwjv5SOfyjShm2dn1q1wTj69oiIiIhI6NO7Vh9Za8nYVFpzPgV4MypUTyEiIiIiYUJJhY+yiiooLq85SdtWVkJultrJioiIiEjYUFLho4ztk7T/2PkpNwtcV0XaIiIiIhI2lFT4aKdF2ts6P6mdrIiIiIiECyUVPkrfVZE2QKp2KkREREQkPCip8Im1loz8nRRpZ2VCk2aYxk39CUxEREREZC8pqfDJzoq0YdvgOx19EhEREZEwoqTCJ9uLtP+cVJC9Xp2fRERERCSsKKnwiVekDR1bxFTfZ7eWwJZ87VSIiIiISFhRUuGT34u0//AtyNnW+UlF2iIiIiISRpRU+MBay6r80h3rKbK2dX7S8ScRERERCSNKKnyQXVRB0U6KtMleD8bAfqn+BCYiIiIisg+UVPhg10XamZC4HyYmdifPEhEREREJTUoqfJC+rUi7U4uayYPXTlZHn0REREQkvCip8EFGfikdmtcs0rbWQnYmRp2fRERERCTMKKkIsu2TtHc4+rRlE5Ru1U6FiIiIiIQdJRVBllO8myJt1E5WRERERMKPkoogS99WpN0l6U/tZLO3t5PV8ScRERERCS9KKoIsI6+UKAMd/1SkTfZ6iG4Eicn+BCYiIiIiso+UVARZRn4pHVrEEhNV84/eZmdCShuME+VTZCIiIiIi+0ZJRRBtL9Lu8ud6CvBmVOjok4iIiIiEISUVQZRTXEHhToq0bVUVbMxSO1kRERERCUtKKoIoYxdF2uRmQ1UVpLTzISoRERERkbpRUhFE6bss0vY6P2mnQkRERETCkZKKINp1kbY3o0KD70REREQkHCmpCJJdTtIGb6cioQk0aRr8wERERERE6khJRZBsL9LeWecnm5UJqW0xxvgQmYiIiIhI3SipCJLtRdo736lYr3oKEREREQlbSiqCJCO/jCgDnVrWLNK2pVthc57qKUREREQkbCmpCJL0XRRpk+MVaZtUJRUiIiIiEp6UVATB7oq0f+/8pONPIiIiIhKelFQEwcbiSgrLqnbd+QlgPyUVIiIiIhKelFQEwZ6KtElMxsTG7vg1EREREZEwoKQiCNLzS3EMdPrzJG22tZNVkbaIiIiIhDElFUGQkV9Kh+axxEb/aZK2tdvaySqpEBEREZFXmIEMAAAgAElEQVTwpaQiwHY7SbtwC2wtVpG2iIiIiIQ1JRUBlltSScGuirSzvCJttZMVERERkXCmpCLA0rcVaXdJ2lk72W2dn3T8SURERETCmJKKAMvI23WRNtnrIToakvYLfmAiIiIiIvUkOlgLffPNN8ycORPXdRk0aBDDhw+v8fVnn32W77//HoDy8nK2bNnCs88+C8CCBQt44403ADjjjDPo379/sMKus10VacO2nYr9WmOcKB8iExERERGpH0FJKlzXZcaMGdx2220kJSUxYcIEevbsSbt27aofM3r06Or/f//99/n1118BKCoq4rXXXmPKlCkAjB8/np49e9KkSZNghF4n1lrS80s5qu0uYs1er6NPIiIiIhL2gnL8KT09ndTUVFJSUoiOjqZPnz4sX758l49funQpffv2BbwdjkMPPZQmTZrQpEkTDj30UL755ptghF1nuyvStm4V5GzAqPOTiIiIiIS5oCQV+fn5JCUlVd9OSkoiPz9/p4/duHEjOTk5dO/efafPTUxM3OVzQ83uirTJ2whVlWonKyIiIiJhLyjHn6y1O9xnjNnpY5cuXUqvXr1wnF3nOzt77rx585g3bx4AU6ZMITk5eR+jrR/R0dFs2OoQZaBnlzbERtesmyhb8wubgRYH/IWYeoo1Ojo66Ncd7DUj4Rr9WDMSrtGPNSPhGv1YMxKuMVLWjIRr9GPNSLhGP9aMhGusy5pBSSqSkpLIy8urvp2Xl0fLli13+thly5YxduzY6tuJiYn88MMP1bfz8/M5+OCDd3je4MGDGTx4cPXt3Nzc+gh9nyUnJ/PtunzaNY+lcPMmCv/0dfeXHwHYEtcYU0+xJicnB/26g71mJFyjH2tGwjX6sWYkXKMfa0bCNUbKmpFwjX6sGQnX6MeakXCNO1uzTZvanaoJyvGntLQ0NmzYQE5ODpWVlSxbtoyePXvu8Lj169dTXFxMt27dqu87/PDDWbFiBUVFRRQVFbFixQoOP/zwYIRdJ7udpA1ekXZ8Y2jaPLiBiYiIiIjUs6DsVERFRTFmzBgmTZqE67oMGDCA9u3bM3v2bNLS0qoTjCVLltCnT58ax5uaNGnCmWeeyYQJEwAYMWJEWHR+yikqZ0tZFV12kVTY7ExIbbvLY2AiIiIiIuEiaHMqevToQY8ePWrcN3LkyBq3zz777J0+d+DAgQwcODBgsQXCzzlFALvZqcjEdOsexIhERERERAJDE7UD5KecIhwD+7fccZK2LSuD/Fx1fhIRERGRBkFJRYD8nF1E+2Y7n6TNxvXefzX4TkREREQaACUVAWCt5eecItKSdtylACArEwCjpEJEREREGgAlFQGQt7WSTVsrdllPYbO37VS0ah3EqEREREREAkNJRQBk5HmTtHdXpE2LJExcfBCjEhEREREJjKB1f4ok7ZvHckWfjnTeSZE2gM3y2smKiIiIiDQE2qkIgDbNYrjgqPY7LdK21nrtZNX5SUREREQaCCUVwVZUCCXF6vwkIiL/v707j4rqvN8A/swCLkxEZkAQFBHEGGu1PwV3TRWXc6KJthWMW7AeYwJujTZBU43WJcYYDmqKYkWsoSduPa51yxFjMDWJLNogBgWjRgHZFRBZZu77+8MwlQBxmbl3zMzz+WvmzvK87yjv3O+9932HiMhusKhQWkH9yk88U0FERERE9oFFhcLEj0UF51QQERERkb1gUaG0glxAowEMnrZuCRERERGRVbCoUJgoyAM8vKDSaGzdFCIiIiIiq2BRobTbuZykTURERER2hUWFgoRkAgrzoWJRQURERER2hEWFkkqLAWMdwJWfiIiIiMiOsKhQUkEeAEDFlZ+IiIiIyI6wqFCQuP3jcrK8/ImIiIiI7AiLCiUV5AItWwFt2tq6JUREREREVsOiQkGiIA/w9IFKpbJ1U4iIiIiIrIZFhZIKcrnyExERERHZHRYVChG1NUBpEVd+IiIiIiK7w6JCKUW3ASFYVBARERGR3WFRoZQfV35SeXWwcUOIiIiIiKyLRYVCREH9crLtbdsQIiIiIiIrY1GhlII8wFUPVcvWtm4JEREREZFVsahQiCjI5XwKIiIiIrJLLCqUUpALlReXkyUiIiIi+8OiQgGishyorOCZCiIiIiKySywqlFCQBwD84TsiIiIiskssKhTwv5WfWFQQERERkf1hUaGEgjxArQbcPW3dEiIiIiIiq2NRoQBRkAu4e0Gl1dq6KUREREREVseiQgkFeQBXfiIiIiIiO8WiQmZCkoDCPKi48hMRERER2SkWFXIrKwFqazlJm4iIiIjsFosKuf248hPPVBARERGRvWJRITPzcrKcU0FEREREdopFhdwK8oAWLQFXva1bQkREREQkCxYVMhMFuYCnN1Qqla2bQkREREQkCxYVcivIg4qTtImIiIjIjrGokJGoqwOKC7nyExERERHZNRYVcirKB4QEcOUnIiIiIrJjLCrkVJAHALz8iYiIiIjsGosKGYnbPy4nyzMVRERERGTHWFTIqSAXaNMWqtYutm4JEREREZFsWFTISBTk8SwFEREREdk9FhVyKsiFyquDrVtBRERERCQrrVJBFy5cwPbt2yFJEkJCQjB+/PhGzzl79iz27t0LlUqFTp06Yf78+QCAiRMnwtfXFwDg7u6OqKgopZr91KR7FUDFXZ6pICIiIiK7p0hRIUkStm3bhiVLlsBgMGDx4sUICgpChw7/O4qfn5+PAwcOYOXKldDpdLh79675MWdnZ6xbt06JplqNKfcmAEDFooKIiIiI7Jwilz/l5OTAy8sLnp6e0Gq1GDhwIFJSUho8JykpCaNHj4ZOpwMAuLq6KtE02Rjzfnhwg8vJEhEREZGdU+RMRWlpKQwGg/m+wWBAdnZ2g+fk5T34TYelS5dCkiSEhobiN7/5DQCgrq4OixYtgkajwbhx49C3b18lmm0RU94PgEoNeHjZuilERERERLJSpKgQQjTaplKpGtyXJAn5+flYtmwZSktL8d577yE6OhouLi7YtGkT9Ho9CgoKsGLFCvj6+sLLq+HO+smTJ3Hy5EkAwAcffAB3d3f5OvQYyvNvQePZHu5e7RXL1Gq1ivdb6UxH6KMtMh2hj7bIdIQ+2iLTEfroKJmO0EdbZDpCH22R6Qh9tCRTkaLCYDCgpKTEfL+kpARubm4NnqPX69G1a1dotVq0a9cO3t7eyM/PR5cuXaDX6wEAnp6e6N69O65fv96oqBgxYgRGjBhhvl9cXCxjjx5NlXsDJncvRdvh7u6ueL+VznSEPtoi0xH6aItMR+ijLTIdoY+OkukIfbRFpiP00RaZjtDHpjK9vR9vfrAicyoCAgKQn5+PwsJCGI1GnD17FkFBQQ2e07dvX1y8eBEAUF5ejvz8fHh6eqKyshJ1dXXm7ZcvX24wwftZJISAMe8mVF6cT0FERERE9k+RMxUajQYzZszA6tWrIUkShg0bho4dO2L37t0ICAhAUFAQevXqhf/+97946623oFarMXXqVDz33HO4fPky/v73v0OtVkOSJIwfP/6ZLypQVgLUVHM5WSIiIiJyCIr9TkXv3r3Ru3fvBtsmTpxovq1SqRAeHo7w8PAGz3n++ecRHR2tSButpiAXAKDiyk9ERERE5AD4i9pyEBK0/s8D/DVtIiIiInIAip2pcCSq7v8Hw9CRNp8sTkRERESkBJ6pICIiIiIii7CoICIiIiIii7CoICIiIiIii7CoICIiIiIii7CoICIiIiIii7CoICIiIiIii7CoICIiIiIii7CoICIiIiIii7CoICIiIiIii7CoICIiIiIii7CoICIiIiIii7CoICIiIiIii7CoICIiIiIii7CoICIiIiIii7CoICIiIiIii7CoICIiIiIii7CoICIiIiIii7CoICIiIiIii6iEEMLWjSAiIiIiol8unqmQyaJFi5hpB3mOkukIfbRFpiP00RaZjtBHR8l0hD7aItMR+miLTEfooyWZLCqIiIiIiMgiLCqIiIiIiMgimuXLly+3dSPslb+/PzPtIM9RMh2hj7bIdIQ+2iLTEfroKJmO0EdbZDpCH22R6Qh9fNpMTtQmIiIiIiKL8PInIiIiIiKyiNbWDbBHFy5cwPbt2yFJEkJCQjB+/HhZ8zZt2oT09HS4uroiOjpa1iwAKC4uRmxsLO7cuQOVSoURI0bgpZdekjWztrYWy5Ytg9FohMlkQv/+/REWFiZrJgBIkoRFixZBr9crsgLD7Nmz0bJlS6jVamg0GnzwwQeyZ967dw9xcXG4efMmVCoVIiIi0LVrV9ny8vLyEBMTY75fWFiIsLAwjBkzRrbMf//73zh16hRUKhU6duyIyMhIODs7y5YHAEePHkVSUhKEEAgJCZGlf0397VdWViImJgZFRUXw8PDAW2+9BZ1OJ2vmV199hb179yI3Nxfvv/8+AgICZM1LTExEWloatFotPD09ERkZCRcXF1kzd+3ahdTUVKhUKri6uiIyMhJ6vV7WzHqHDh3CP//5T8THx6NNmzay5e3ZswdJSUnmjEmTJqF3795WyWsuEwCOHTuG48ePQ6PRoHfv3pg6daqsmTExMcjLywMAVFVVoXXr1li3bp1sedevX8fWrVtRW1sLjUaDmTNnokuXLlbJe1RmdXU1PDw8MG/ePLRu3doqec3tA8g59jSXKefY01ymXONPc3lyjj2P2p974rFHkFWZTCYxZ84ccfv2bVFXVyf+/Oc/i5s3b8qamZmZKa5evSoWLFgga0690tJScfXqVSGEEFVVVWLevHmy91GSJHH//n0hhBB1dXVi8eLF4vLly7JmCiHE4cOHxfr168WaNWtkzxJCiMjISHH37l1Fsup9/PHH4uTJk0KIB59tZWWlYtkmk0nMnDlTFBYWypZRUlIiIiMjRU1NjRBCiOjoaPH555/LlieEEDdu3BALFiwQ1dXVwmg0ihUrVoi8vDyr5zT1t5+YmCj2798vhBBi//79IjExUfbMmzdvitzcXLFs2TKRk5Mje96FCxeE0WgUQjzorxJ9vHfvnvn2kSNHxJYtW2TPFEKIoqIisWrVKhEREWHVsaGpvN27d4uDBw9aLeNxMjMyMsSKFStEbW2tEEKIO3fuyJ75sB07doi9e/fKmrdy5UqRnp4uhBAiLS1NLFu2zGp5zWUuWrRIZGZmCiGESEpKEjt37rRaXnP7AHKOPc1lyjn2NJcp1/jTXJ6cY8/P7c89zdjDy5+sLCcnB15eXvD09IRWq8XAgQORkpIia2b37t2teiTyUdzc3MwTeFq1agUfHx+UlpbKmqlSqdCyZUsAgMlkgslkgkqlkjWzpKQE6enpCAkJkTXHlqqqqvDdd99h+PDhAACtVmvVI76PkpGRAS8vL3h4eMiaI0kSamtrYTKZUFtbCzc3N1nzcnNzERgYiBYtWkCj0eCFF17AuXPnrJ7T1N9+SkoKXnzxRQDAiy++aPXxp6nMDh06wNvb26o5P5fXq1cvaDQaAEDXrl2tPv40lfnwUd6amhqrjz/NjeM7duzAlClTFMuTU1OZn332GcaNGwcnJycAgKurq+yZ9YQQ+OqrrzBo0CBZ81QqFe7fvw/gwZhr7fGnqcy8vDy88MILAICePXvim2++sVpec/sAco49zWXKOfY0lynX+NNcnpxjz8/tzz3N2MPLn6ystLQUBoPBfN9gMCA7O9uGLZJXYWEhrl27ZtVTuc2RJAlRUVG4ffs2Ro8ejcDAQFnz/vGPf2Dq1KnmLwOlrF69GgAwcuRIjBgxQtaswsJCtGnTBps2bcKNGzfg7++P6dOnmws4uf3nP/+x6hd6U/R6PV5++WVERETA2dkZvXr1Qq9evWTN7NixI3bt2oWKigo4Ozvj/PnzVj0t/3Pu3r1r3mlxc3NDeXm5Irm2curUKQwcOFCRrJ07dyI5ORmtW7fGsmXLZM9LTU2FXq+Hn5+f7Fn1Tpw4geTkZPj7++O1116TvfDIz89HVlYWdu3aBScnJ0ybNk2R7xMA+O677+Dq6or27dvLmhMeHo7Vq1cjMTERkiRh1apVsuYBD8ag1NRUBAcH4+uvv0ZJSYksOQ/vAyg19ii53/GoTLnGn5/mKTH2PJz5tGMPz1RYmWhiMS25j6jbSnV1NaKjozF9+nSrXav5c9RqNdatW4e4uDhcvXoVP/zwg2xZaWlpcHV1VXwZt5UrV2Lt2rV49913ceLECVy6dEnWPJPJhGvXrmHUqFH48MMP0aJFCxw4cEDWzHpGoxFpaWno37+/rDmVlZVISUlBbGwstmzZgurqaiQnJ8ua2aFDB4wbNw6rVq3C+++/j06dOkGt5nBrbfv27YNGo8GQIUMUyZs0aRI2b96MwYMH4/jx47Jm1dTUYN++fZg4caKsOQ8bNWoUPv74Y3z44Ydwc3PDJ598InumJEmorKzE6tWrMW3aNMTExDT5PSoHJQ5qAA/OxoSHh2Pz5s0IDw9HXFyc7JkRERE4ceIEoqKicP/+fWi11j+GrPQ+wLOWKdf401Se3GPPw5kajeapxx5+y1mZwWBocESgpKRE9kstbMFoNCI6OhpDhgxBv379FM12cXFB9+7dceHCBdkyLl++jNTUVMyePRvr16/HxYsXsXHjRtny6tVPvnJ1dUVwcDBycnJkzTMYDDAYDOazPv3798e1a9dkzax3/vx5dO7cGW3btpU1JyMjA+3atUObNm2g1WrRr18/XLlyRdZMABg+fDjWrl2Lv/71r9DpdLIfDa3n6uqKsrIyAEBZWZnVJvY+a06fPo20tDTMmzdP8QM3gwcPturlJE0pKChAYWEh3n77bcyePRslJSWIiorCnTt3ZMts27Yt1Go11Go1QkJCcPXqVdmy6un1evTr1w8qlQpdunSBWq1GRUWF7Lkmkwnnzp1T5CzXF198Yf6eHDBggOzjOgD4+PhgyZIlWLt2LQYNGgRPT0+rvn9T+wByjz222O9oLlOu8edRfZRj7PlppiVjD4sKKwsICEB+fj4KCwthNBpx9uxZBAUF2bpZViWEQFxcHHx8fDB27FhFMsvLy3Hv3j0AD1aCysjIgI+Pj2x5kydPRlxcHGJjY/GnP/0JPXr0wLx582TLAx4cKai/1Kq6uhrffvstfH19Zc1s27YtDAaDeSWUjIwMdOjQQdbMekodJXR3d0d2djZqamoghJD9/069u3fvAniwusa5c+cU6SsABAUF4YsvvgDwYGcmODhYkVwlXbhwAQcPHkRUVBRatGihSGZ+fr75dmpqqmzXcdfz9fVFfHw8YmNjERsbC4PBgLVr18pahNfvEALAuXPn0LFjR9my6gUHB+PixYsAHswDMBqNeO6552TPzcjIgLe3d4PLleWi1+vNZ50vXrwILy8v2TPrxx9JkrBv3z6MHDnSau/d3D6AnGOPLfY7msuUa/xpLk/OsaepTEvGHv74nQzS09OxY8cOSJKEYcOG4fe//72seevXr8elS5dQUVEBV1dXhIWFmSfeyiErKwvvvfcefH19zRW6tZce/KkbN24gNjYWkiRBCIEBAwZgwoQJsuU9LDMzE4cPH5Z9SdmCggJ89NFHAB4cRRs8eLDs/3eAB0sPxsXFwWg0ol27doiMjJT9OuqamhpERETgb3/7myKnsPfs2YOzZ89Co9HAz88Pb775pnliqFzee+89VFRUQKvV4rXXXsOvf/1rq2c09bcfHByMmJgYFBcXw93dHQsWLLDqv2dTmTqdDgkJCSgvL4eLiwv8/Pzwl7/8Rba8/fv3w2g0mvsVGBiIWbNmWSWvucz09HTk5+dDpVLB3d0ds2bNsuqSso8ax2fPno01a9ZY7ehvU3mZmZm4fv06VCoVPDw8MGvWLKueaW8qc+jQoeY5XVqtFtOmTUOPHj1kzRw+fDhiY2MRGBiIUaNGWS2ruTxvb2/zMvNOTk6YOXOmVS+tbSqzuroaJ06cAAD07dsXkydPttoR9eb2AQIDA2Ube5rLNBqNso09zWVu375dlvGnubxTp07JNvY8zv7ck4w9LCqIiIiIiMgivPyJiIiIiIgswqKCiIiIiIgswqKCiIiIiIgswqKCiIiIiIgswqKCiIiIiIgswqKCiMiBxMbGYteuXTbJFkJg06ZN+OMf/4jFixdb/f2Li4sxbdo0SJL0yOcWFhYiLCwMJpOpycf37NmjyA9eEhHZCxYVREQ2NHv2bLz++uuorq42b0tKSsLy5ctt1yiZZGVl4dtvv8XmzZuxZs2aRo+fPn0aYWFhOHToUIPtb775JjIzMx/5/u7u7khMTIRaza82IiKlceQlIrIxk8mEo0eP2roZT+xxzgg8rKioCB4eHmjZsmWzz9HpdDh48CCqqqosbd4z40k/JyKiXyKtrRtAROToXnnlFRw8eBCjR4+Gi4tLg8cKCwsxZ84c7Ny5ExqNBgCwfPlyDBkyBCEhITh9+jSSkpIQEBCA06dPQ6fTYe7cucjPz8fu3btRV1eHqVOn4re//a35PcvLy7Fy5UpkZ2ejc+fOmDNnDjw8PAAAubm5SEhIwPfff482bdpg4sSJGDhwIIAHl045OzujuLgYly5dwttvv42ePXs2aG9paSm2bt2KrKws6HQ6jBs3DiNGjMCpU6ewbds2GI1GTJs2DS+//DLCwsIafRY+Pj5wcXHBkSNHEBoa2uhxSZJw6NAhJCUl4d69e+jRowdmzZoFnU7X6LMqLCxEbGwsrl27hsDAQLRv3x5VVVWYN2+e+f3OnDmD3bt3o7a2FmPGjGnwK/Z1dXWIiYnB+fPn0b59e0RERMDPzw8AcOvWLcTHx+P69evQ6/WYPHkygoKCmv2cjEYjEhMTUVJSglatWmHMmDF45ZVXHve/CBHRM49nKoiIbMzf3x+/+tWvcPjw4ad6fXZ2Njp16oSEhAQMHjwY69evR05ODjZu3Ii5c+ciISGhweVVX375Jf7whz9g27Zt8PPzM88dqK6uxqpVqzB48GDEx8dj/vz52LZtG27evNngtb/73e+wY8cOdOvWrVFbNmzYAIPBgC1btmDhwoXYuXMnMjIyMHz4cLz++uvo2rUrEhMTmywo6r366qs4cuQIKisrGz127NgxpKSkYPny5diyZQt0Oh3i4+ObfJ8NGzYgICAACQkJCA0NxZkzZxo9JysrCxs2bMDSpUvxr3/9C7du3TI/lpqaigEDBiAhIQGDBg3CunXrYDQaYTQasXbtWvTs2RPx8fGYMWMGNm7ciLy8vGY/p7i4OMyaNQuffPIJoqOj0aNHj2b7T0T0S8SigojoGRAWFoZjx46hvLz8iV/brl07DBs2DGq1GgMHDkRJSQkmTJgAJycn9OrVC1qtFrdv3zY/v3fv3ujevTucnJwwadIkXLlyBcXFxUhPT4eHhweGDRsGjUYDf39/9OvXD19//bX5tcHBwejWrRvUajWcnZ0btKO4uBhZWVmYMmUKnJ2d4efnh5CQECQnJz9Rf/z8/NCzZ08cOHCg0WMnT57Eq6++CoPBACcnJ4SGhuKbb75pNOG6uLgYV69excSJE6HVatGtWzf06dOn0fuFhoaa29qpUyfcuHHD/Ji/vz/69+8PrVaLsWPHoq6uDtnZ2cjOzkZ1dTXGjx8PrVaLHj16oHfv3vjyyy+b/Zw0Gg1u3bqFqqoq6HQ6+Pv7P9FnQkT0rOPlT0REzwBfX1/06dMHBw4cgI+PzxO91tXV1Xy7fke/bdu2DbY9fKbCYDCYb7ds2RI6nQ5lZWUoKipCdnY2pk+fbn7cZDJh6NChTb72p8rKyqDT6dCqVSvzNnd3d1y9evWJ+gMAEydOxLvvvouxY8c22F5UVISPPvoIKpXKvE2tVuPu3bsNnldaWgqdTocWLVo0aEtxcXGD5z38ObVo0aLZz0mtVsNgMKCsrMz8Xg9PCPfw8EBpaWmTrwWAhQsXYt++ffj000/h6+uLKVOmoGvXro/+IIiIfiFYVBARPSPCwsIQFRXVYEe6flJzTU0NWrduDQC4c+eORTklJSXm29XV1aisrISbmxsMBgO6d++OpUuXNvvah3fmf8rNzQ2VlZW4f/++ubAoLi6GXq9/4jb6+Pigb9++2LdvX4PtBoMBERERTV56VVhY2KgtNTU15sLipwXFozz8OUmShJKSEri5uZnfS5Ikc2FRXFyM9u3bm5//08+pS5cueOedd2A0GnH8+HHExMRg8+bNT9QeIqJnGS9/IiJ6Rnh5eWHAgAE4duyYeVubNm2g1+tx5swZSJKEU6dOoaCgwKKc8+fPIysrC0ajEbt27UJgYCDc3d3Rp08f5OfnIzk52Tx3ICcnp8E8g5/j7u6O559/Hp9++ilqa2tx48YNfP755xgyZMhTtTM0NBSnT5/GvXv3zNtGjhyJXbt2oaioCMCDSecpKSmNXuvh4YGAgADs3bsXRqMRV65cQVpa2hPlf//99+ZLq44ePQonJycEBgYiMDAQLVu2xKFDh2A0GpGZmYm0tDQMGjSoyfcxGo04c+YMqqqqoNVq0bp1ay57S0R2h2cqiIieIRMmTGg0ofiNN95AfHw8du7cieHDh1t82cygQYOwd+9eXLlyBf7+/ubVkFq1aoUlS5Zgx44d2LFjB4QQ6NSpE8LDwx/7vefPn4+tW7fijTfegE6nQ2hoaKMVoh5Xu3btMHToUHz22WfmbS+99BIAYNWqVSgrK4OrqysGDBiA4ODgRq+fO3cuNm3ahBkzZqBLly4YOHDgEy3vGhQUhLNnzyI2NhZeXl5YuHAhtNoHX5vvvPMO4uPjsX//fuj1esyZM+dnL1tLTk5GQkICJEmCt7c35s6d+9jtICL6JVAJIYStG0FERCS3mJgY+Pj4/OzKU0RE9HR4/pWIiOxSTk4Obt++DUmScOHCBaSmpjZ5RoOIiCzHy5+IiMgu3blzB9HR0aioqIDBYMDMmTPRuXNnWzeLiMgu8fInIiIiIiKyCC9/IiIiIiIii7CoICIiIiIii7CoINbeyGkAAAAnSURBVCIiIiIii7CoICIiIiIii7CoICIiIiIii7CoICIiIiIii/w/wHy6CIzNKm8AAAAASUVORK5CYII=\n",
      "text/plain": [
       "<Figure size 936x576 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Best accuracy is 0.8924731182795699 with max_depth = 17\n"
     ]
    }
   ],
   "source": [
    "# Model complexity\n",
    "\n",
    "train_accuracy = []\n",
    "test_accuracy = []\n",
    "# Loop over different values of k\n",
    "for d in range (1,25):\n",
    "    # d from 1 to 25(exclude)\n",
    "    rf = RandomForestClassifier(max_depth=d)\n",
    "    # Fit with rf\n",
    "    rf.fit(x_train,y_train)\n",
    "    #train accuracy\n",
    "    train_accuracy.append(rf.score(x_train, y_train))\n",
    "    # test accuracy\n",
    "    test_accuracy.append(rf.score(x_test, y_test))\n",
    "    \n",
    "\n",
    "# Plot\n",
    "plt.figure(figsize=[13,8])\n",
    "plt.plot(test_accuracy, label = 'Testing Accuracy')\n",
    "plt.plot(train_accuracy, label = 'Training Accuracy')\n",
    "plt.legend()\n",
    "plt.xticks(np.arange(0,25))\n",
    "plt.title('-value VS Accuracy')\n",
    "plt.xlabel('Number of Neighbors')\n",
    "plt.ylabel('Accuracy')\n",
    "plt.savefig('graph.png')\n",
    "plt.show()\n",
    "print(\"Best accuracy is {} with max_depth = {}\".format(np.max(test_accuracy),1+test_accuracy.index(np.max(test_accuracy))))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "_cell_guid": "b598ee81-e535-49c0-b53c-b13b0a5058db",
    "_uuid": "96423b4f710966c8071647874623c139c1c79bb7"
   },
   "source": [
    "### Up to this point what you learn:\n",
    "* Supervised learning\n",
    "* Exploratory data analysis\n",
    "* KNN and RF\n",
    "    * How to split data\n",
    "    * How to fit, predict data\n",
    "    * How to measure medel performance (accuracy)\n",
    "    * How to choose hyperparameter (K)\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "_cell_guid": "f9d427a9-faa5-46cf-9e3c-2c8cea2571ad",
    "_uuid": "d075fd2a7c05e5414e33b7b1314d81a6b945e7b3"
   },
   "source": [
    "<a id=\"5\"></a> <br>\n",
    "### REGRESSION\n",
    "* Supervised learning\n",
    "* We will learn linear and logistic regressions\n",
    "* This orthopedic patients data is not proper for regression so I only use two features that are *sacral_slope* and *pelvic_incidence* of abnormal \n",
    "    * I  consider feature is pelvic_incidence and target is sacral_slope \n",
    "    * Lets look at scatter plot so as to understand it better"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 238,
   "metadata": {
    "_cell_guid": "6b072c42-059f-4e45-9cfa-8ed39b274f72",
    "_uuid": "d2655c140423b1228c42d2e8dfe54344ba43dcb0"
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x720 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# create data1 that includes pelvic_incidence that is feature and sacral_slope that is target variable\n",
    "data1 = data[data['class'] =='Abnormal']\n",
    "\n",
    "x = data1[\"pelvic_incidence\"]\n",
    "y = data1[\"sacral_slope\"]\n",
    "x=x.values.reshape(-1,1)\n",
    "y=y.values.reshape(-1,1)\n",
    "# Scatter\n",
    "plt.figure(figsize=[10,10])\n",
    "plt.scatter(x=x,y=y)\n",
    "plt.xlabel('pelvic_incidence')\n",
    "plt.ylabel('sacral_slope')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "_cell_guid": "874ac5e0-5bc0-4429-b6c5-707690b5dd77",
    "_uuid": "c84719d363ff736e96a0a575dfd0381bcbc549fb"
   },
   "source": [
    "Now we have our data to make regression. In regression problems target value is continuously varying variable such as price of house or sacral_slope. Lets fit line into this points.\n",
    "\n",
    "<br> Linear regression\n",
    "* y = ax + b       where  y = target, x = feature and a = parameter of model\n",
    "* We choose parameter of model(a) according to minimum error function that is lost function\n",
    "* In linear regression we use Ordinary Least Square (OLS) as lost function.\n",
    "* OLS: sum all residuals but some positive and negative residuals can cancel each other so we sum of square of residuals. It is called OLS\n",
    "* Score: Score uses R^2 method that is ((y_pred - y_mean)^2 )/(y_actual - y_mean)^2"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 239,
   "metadata": {
    "_cell_guid": "fb7991f3-5869-4df0-bf6c-30f61e8215c6",
    "_uuid": "7cdc74efa8c46dab5f14f6cc2779928c11a4fa62"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "R^2 score:  0.6458410481075871\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# LinearRegression\n",
    "from sklearn.linear_model import LinearRegression\n",
    "reg = LinearRegression()\n",
    "# Predict space\n",
    "predict_space = np.linspace(min(x), max(x)).reshape(-1,1)\n",
    "# Fit\n",
    "reg.fit(x,y)\n",
    "# Predict\n",
    "predicted = reg.predict(predict_space)\n",
    "# R^2 \n",
    "print('R^2 score: ',reg.score(x, y))\n",
    "\n",
    "# Plot regression line and scatter\n",
    "plt.plot (predict_space,predicted, color='black', linewidth=3)\n",
    "plt.scatter(x,y)\n",
    "plt.xlabel('pelvic_incidence')\n",
    "plt.ylabel('sacral_slope')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "_cell_guid": "242dd057-bb8a-463a-b04c-1072d5de1e0a",
    "_uuid": "3cf254d2feef8f9f886b0e1de12ec2b155083d2b"
   },
   "source": [
    "<a id=\"6\"></a> <br>\n",
    "### CROSS VALIDATION\n",
    "As you know in KNN method we use train test split with random_state that split exactly same at each time. However, if we do not use random_state, data is split differently at each time and according to split accuracy will be different. Therefore, we can conclude that model performance is dependent on train_test_split. For example you split, fit and predict data 5 times and accuracies are 0.89, 0.9, 0.91, 0.92 and 0.93, respectively. Which accuracy do you use? Do you know what accuracy will be at 6th times split, train and predict. The answer is I do not know but if I use cross validation I can find acceptable accuracy.\n",
    "<br> Cross Validation (CV)\n",
    "* K folds = K fold CV.\n",
    "* Look at this image it defines better than me :)\n",
    "* When K is increase, computationally cost is increase\n",
    "* cross_val_score(reg,x,y,cv=5): use reg(linear regression) with x and y that we define at above and K is 5. It means 5 times(split, train,predict)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 240,
   "metadata": {
    "_cell_guid": "bf504792-66f4-411e-bc1d-5b049da959ac",
    "_uuid": "1f1a77c5d5e6ca52c0264875362665f228e66078"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "CV Scores:  [0.43000935 0.59581689 0.5819292  0.47603177 0.20981054 0.23291668]\n",
      "CV scores average:  0.4210857377256712\n"
     ]
    }
   ],
   "source": [
    "# CV\n",
    "from sklearn.model_selection import cross_val_score\n",
    "reg = LinearRegression()\n",
    "k = 6\n",
    "cv_result = cross_val_score(reg,x,y,cv=k) # uses R^2 as score \n",
    "print('CV Scores: ',cv_result)\n",
    "print('CV scores average: ',np.sum(cv_result)/k)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "_cell_guid": "27ffec30-4951-4479-9d86-32495a80c08d",
    "_uuid": "be425f142c0acc1ab3009fafafa3616948f5c8a5"
   },
   "source": [
    "### Regularized Regression\n",
    "As we learn linear regression choose parameters (coefficients) while minimizing lost function. If linear regression thinks that one of the feature is important, it gives high coefficient to this feature. However, this can cause overfitting that is like memorizing in KNN. In order to avoid overfitting, we use regularization that penalize large coefficients.\n",
    "\n",
    "* Ridge regression: First regularization technique. Also it is called L2 regularization. \n",
    "    * Ridge regression lost fuction = OLS + alpha * sum(parameter^2)\n",
    "    * alpha is parameter we need to choose to fit and predict. Picking alpha is similar to picking K in KNN. As you understand alpha is hyperparameter that we need to choose for best accuracy and model complexity. This process is called hyperparameter tuning.\n",
    "    * What if alpha is zero? lost function = OLS so that is linear rigression :)\n",
    "    * If alpha is small that can cause overfitting\n",
    "    * If alpha is big that can cause underfitting. But do not ask what is small and big. These can be change from problem to problem.\n",
    "* Lasso regression: Second regularization technique. Also it is called L1 regularization. \n",
    "    * Lasso regression lost fuction = OLS + alpha * sum(absolute_value(parameter))\n",
    "    * It can be used to select important features od the data. Because features whose values are not shrinked to zero, is chosen by lasso regression\n",
    "    * In order to choose feature, I add new features in our regression data\n",
    "    \n",
    "<br> Linear vs Ridge vs Lasso\n",
    "First impression: Linear\n",
    "Feature Selection: 1.Lasso 2.Ridge\n",
    "Regression model: 1.Ridge 2.Lasso 3.Linear \n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 241,
   "metadata": {
    "_cell_guid": "bdc1a379-07a4-4b61-8ac1-d7007ae33783",
    "_uuid": "85fa872e3988295a8fbe752bf96319518ca3595b"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Ridge score:  0.5608287918841997\n"
     ]
    }
   ],
   "source": [
    "# Ridge\n",
    "from sklearn.linear_model import Ridge\n",
    "x_train,x_test,y_train,y_test = train_test_split(x,y,random_state = 2, test_size = 0.3)\n",
    "ridge = Ridge(alpha = 0.1, normalize = True)\n",
    "ridge.fit(x_train,y_train)\n",
    "ridge_predict = ridge.predict(x_test)\n",
    "print('Ridge score: ',ridge.score(x_test,y_test))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 242,
   "metadata": {
    "_cell_guid": "de031d9f-b2f8-4fb1-a305-36cbc4dc970f",
    "_uuid": "57f91f4b4e267bd3eb22adfcdf719778c1901c92"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Lasso score:  0.9640334804327547\n",
      "Lasso coefficients:  [ 0.82498243 -0.7209057   0.         -0.        ]\n"
     ]
    }
   ],
   "source": [
    "# Lasso\n",
    "from sklearn.linear_model import Lasso\n",
    "x = np.array(data1.loc[:,['pelvic_incidence','pelvic_tilt numeric','lumbar_lordosis_angle','pelvic_radius']])\n",
    "x_train,x_test,y_train,y_test = train_test_split(x,y,random_state = 3, test_size = 0.3)\n",
    "lasso = Lasso(alpha = 0.1, normalize = True)\n",
    "lasso.fit(x_train,y_train)\n",
    "lasso_predict = lasso.predict(x_test)\n",
    "print('Lasso score: ',lasso.score(x_test,y_test))\n",
    "print('Lasso coefficients: ',lasso.coef_)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "_cell_guid": "84f9d88d-28e9-4b26-8824-ae32de7e143c",
    "_uuid": "d70e2366ccc8797c92d2edacb4ab3b59fad4506d"
   },
   "source": [
    "As you can see *pelvic_incidence* and *pelvic_tilt numeric* are important features but others are not important\n",
    "\n",
    "<br> Now lets discuss accuracy. Is it enough for measurement of model selection. For example, there is a data that includes 95% normal and 5% abnormal samples and our model uses accuracy for measurement metric. Then our model predict 100% normal for all samples and accuracy is 95% but it classify all abnormal samples wrong. Therefore we need to use confusion matrix as a model measurement matris in imbalance data.\n",
    "<br> While using confusion matrix lets use Random forest classifier to diversify classification methods.\n",
    "* tp = true positive(20), fp = false positive(7), fn = false negative(8), tn = true negative(58) \n",
    "* tp = Prediction is positive(normal) and actual is positive(normal). \n",
    "* fp = Prediction is positive(normal) and actual is negative(abnormal).\n",
    "* fn = Prediction is negative(abnormal) and actual is positive(normal).\n",
    "* tn = Prediction is negative(abnormal) and actual is negative(abnormal)\n",
    "* precision = tp / (tp+fp)\n",
    "* recall = tp / (tp+fn)\n",
    "* f1 = 2 * precision * recall / ( precision + recall)\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 227,
   "metadata": {
    "_cell_guid": "c71a33f5-5784-461b-949d-cb83f23dace6",
    "_uuid": "19fb6fb0f651e249835037a4f4b2f0b4a2619a27"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Confusion matrix: \n",
      " [[66  0]\n",
      " [ 0 27]]\n",
      "Classification report: \n",
      "              precision    recall  f1-score   support\n",
      "\n",
      "   Abnormal       1.00      1.00      1.00        66\n",
      "     Normal       1.00      1.00      1.00        27\n",
      "\n",
      "avg / total       1.00      1.00      1.00        93\n",
      "\n"
     ]
    }
   ],
   "source": [
    "# Confusion matrix with random forest\n",
    "from sklearn.metrics import classification_report, confusion_matrix\n",
    "from sklearn.ensemble import RandomForestClassifier\n",
    "x,y = data.loc[:,data.columns != 'class'], data.loc[:,'class']\n",
    "x_train,x_test,y_train,y_test = train_test_split(x,y,test_size = 0.3,random_state = 1)\n",
    "rf = RandomForestClassifier(random_state = 4)\n",
    "rf.fit(x_train,y_train)\n",
    "y_pred = rf.predict(x_test)\n",
    "cm = confusion_matrix(y_test,y_pred)\n",
    "print('Confusion matrix: \\n',cm)\n",
    "print('Classification report: \\n',classification_report(y_test,y_pred))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "_cell_guid": "4552b965-e15d-4e05-9e02-224159e8d508",
    "_uuid": "dbe175099c0c8151c16ad0c78f1414de8fd9ebdc"
   },
   "source": [
    "<a id=\"7\"></a> <br>\n",
    "### ROC Curve with Logistic Regression \n",
    "* logistic regression output is probabilities\n",
    "* If probability is higher than 0.5 data is labeled 1(abnormal) else 0(normal)\n",
    "* By default logistic regression threshold is 0.5\n",
    "* ROC is receiver operationg characteristic. In this curve x axis is false positive rate and y axis is true positive rate\n",
    "* If the curve in plot is closer to left-top corner, test is more accurate.\n",
    "* Roc curve score is auc that is computation area under the curve from prediction scores\n",
    "* We want auc to closer 1\n",
    "* fpr = False Positive Rate\n",
    "* tpr = True Positive Rate\n",
    "* If you want, I made ROC, Random forest and K fold CV in this tutorial. https://www.kaggle.com/kanncaa1/roc-curve-with-k-fold-cv/"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 250,
   "metadata": {
    "_cell_guid": "4f4a8f76-9792-485f-83b0-79e2524ab83c",
    "_uuid": "c7fce3a219764388088ec5f3d57ab913f5c05f35"
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# ROC Curve with logistic regression\n",
    "from sklearn.metrics import roc_curve\n",
    "from sklearn.linear_model import LogisticRegression\n",
    "from sklearn.metrics import confusion_matrix, classification_report\n",
    "# abnormal = 1 and normal = 0\n",
    "data['class_binary'] = [1 if i == 'Abnormal' else 0 for i in data.loc[:,'class']]\n",
    "x,y = data.loc[:,(data.columns != 'class') & (data.columns != 'class_binary')], data.loc[:,'class_binary']\n",
    "x_train, x_test, y_train, y_test = train_test_split(x, y, test_size = 0.3, random_state=42)\n",
    "logreg = LogisticRegression()\n",
    "logreg.fit(x_train,y_train)\n",
    "y_pred_prob = logreg.predict_proba(x_test)[:,1]\n",
    "fpr, tpr, thresholds = roc_curve(y_test, y_pred_prob)\n",
    "# Plot ROC curve\n",
    "plt.plot([0, 1], [0, 1], 'k--')\n",
    "plt.plot(fpr, tpr)\n",
    "plt.xlabel('False Positive Rate')\n",
    "plt.ylabel('True Positive Rate')\n",
    "plt.title('ROC')\n",
    "plt.show()\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "_cell_guid": "ceb21da5-4ced-400c-a017-60f735802b69",
    "_uuid": "db2a0e5e83aee71a9585ee113cb01170325c96f1"
   },
   "source": [
    "<a id=\"8\"></a> <br>\n",
    "### HYPERPARAMETER TUNING\n",
    "As I mention at KNN there are hyperparameters that are need to be tuned\n",
    "* For example: \n",
    "    * k at KNN\n",
    "    * alpha at Ridge and Lasso\n",
    "    * Random forest parameters like max_depth\n",
    "    * linear regression parameters(coefficients)\n",
    "* Hyperparameter tuning: \n",
    "    * try all of combinations of different parameters\n",
    "    * fit all of them\n",
    "    * measure prediction performance\n",
    "    * see how well each performs\n",
    "    * finally choose best hyperparameters\n",
    "* This process is most difficult part of this tutorial. Because we will write a lot of for loops to iterate all combinations. Just I am kidding sorry for this :) (We actually did it at KNN part)\n",
    "* We only need is one line code that is GridSearchCV\n",
    "    * grid: K is from 1 to 50(exclude)\n",
    "    * GridSearchCV takes knn and grid and makes grid search. It means combination of all hyperparameters. Here it is k.\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 264,
   "metadata": {
    "_cell_guid": "d1a56a4e-a307-4376-8d06-53fc30447ce6",
    "_uuid": "20fb485285b4d27da2e3eb89e8e43797db7c457a"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Tuned hyperparameter k: {'n_neighbors': 7}\n",
      "Best score: 0.8225806451612904\n"
     ]
    }
   ],
   "source": [
    "# grid search cross validation with 1 hyperparameter\n",
    "from sklearn.model_selection import GridSearchCV\n",
    "grid = {'n_neighbors': np.arange(1,50)}\n",
    "knn = KNeighborsClassifier()\n",
    "knn_cv = GridSearchCV(knn, grid, cv=5) # GridSearchCV\n",
    "knn_cv.fit(x,y)# Fit\n",
    "\n",
    "# Print hyperparameter\n",
    "print(\"Tuned hyperparameter k: {}\".format(knn_cv.best_params_)) \n",
    "print(\"Best score: {}\".format(knn_cv.best_score_))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "_cell_guid": "a4401913-229c-4bcc-8dd0-3563b824f6e9",
    "_uuid": "bdc9c267b635db5696e73cb6ba36d093c8f127c5"
   },
   "source": [
    "Other grid search example with 2 hyperparameter\n",
    "* First hyperparameter is C:logistic regression regularization parameter\n",
    "    * If C is high: overfit\n",
    "    * If C is low: underfit\n",
    "* Second hyperparameter is penalty(lost function): l1 (Lasso) or l2(Ridge) as we learnt at linear regression part."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 265,
   "metadata": {
    "_cell_guid": "8988b749-3bf5-448f-830e-edce366348d0",
    "_uuid": "eb5961ae83511410ae491366bd12f70bef9f440b"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Tuned hyperparameters : {'C': 100.0, 'penalty': 'l2'}\n",
      "Best Accuracy: 0.8525345622119815\n"
     ]
    }
   ],
   "source": [
    "# grid search cross validation with 2 hyperparameter\n",
    "# 1. hyperparameter is C:logistic regression regularization parameter\n",
    "# 2. penalty l1 or l2\n",
    "# Hyperparameter grid\n",
    "param_grid = {'C': np.logspace(-3, 3, 7), 'penalty': ['l1', 'l2']}\n",
    "x_train, x_test, y_train, y_test = train_test_split(x,y,test_size = 0.3,random_state = 12)\n",
    "logreg = LogisticRegression()\n",
    "logreg_cv = GridSearchCV(logreg,param_grid,cv=3)\n",
    "logreg_cv.fit(x_train,y_train)\n",
    "\n",
    "# Print the optimal parameters and best score\n",
    "print(\"Tuned hyperparameters : {}\".format(logreg_cv.best_params_))\n",
    "print(\"Best Accuracy: {}\".format(logreg_cv.best_score_))\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "_cell_guid": "37171c8f-21ad-4628-8279-5cfedb27aafb",
    "_uuid": "7cd0bf0ed62c1858f5de74ee4f85c84765f8bc06"
   },
   "source": [
    "<a id=\"9\"></a> <br>\n",
    "### PRE-PROCESSING DATA\n",
    "* In real life data can include objects or categorical data in order to use them in sklearn we need to encode them into numerical data\n",
    "* In data, class is *abnormal* and *normal*. Lets convert them into numeric value (actually I did it in logistic regression part with different method)\n",
    "* 2 different feature is created with the name *class_Abnormal* and *class_Normal*\n",
    "* However we need to drop one of the column because they are duplicated"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 266,
   "metadata": {
    "_cell_guid": "342e6f67-afeb-42b0-9e56-6521fd15e3ad",
    "_uuid": "8168cf770134142252e92022307815e31a931a34"
   },
   "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>pelvic_incidence</th>\n",
       "      <th>pelvic_tilt numeric</th>\n",
       "      <th>lumbar_lordosis_angle</th>\n",
       "      <th>sacral_slope</th>\n",
       "      <th>pelvic_radius</th>\n",
       "      <th>degree_spondylolisthesis</th>\n",
       "      <th>class_Abnormal</th>\n",
       "      <th>class_Normal</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>63.027818</td>\n",
       "      <td>22.552586</td>\n",
       "      <td>39.609117</td>\n",
       "      <td>40.475232</td>\n",
       "      <td>98.672917</td>\n",
       "      <td>-0.254400</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>39.056951</td>\n",
       "      <td>10.060991</td>\n",
       "      <td>25.015378</td>\n",
       "      <td>28.995960</td>\n",
       "      <td>114.405425</td>\n",
       "      <td>4.564259</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>68.832021</td>\n",
       "      <td>22.218482</td>\n",
       "      <td>50.092194</td>\n",
       "      <td>46.613539</td>\n",
       "      <td>105.985135</td>\n",
       "      <td>-3.530317</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>69.297008</td>\n",
       "      <td>24.652878</td>\n",
       "      <td>44.311238</td>\n",
       "      <td>44.644130</td>\n",
       "      <td>101.868495</td>\n",
       "      <td>11.211523</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>49.712859</td>\n",
       "      <td>9.652075</td>\n",
       "      <td>28.317406</td>\n",
       "      <td>40.060784</td>\n",
       "      <td>108.168725</td>\n",
       "      <td>7.918501</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>40.250200</td>\n",
       "      <td>13.921907</td>\n",
       "      <td>25.124950</td>\n",
       "      <td>26.328293</td>\n",
       "      <td>130.327871</td>\n",
       "      <td>2.230652</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>53.432928</td>\n",
       "      <td>15.864336</td>\n",
       "      <td>37.165934</td>\n",
       "      <td>37.568592</td>\n",
       "      <td>120.567523</td>\n",
       "      <td>5.988551</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>45.366754</td>\n",
       "      <td>10.755611</td>\n",
       "      <td>29.038349</td>\n",
       "      <td>34.611142</td>\n",
       "      <td>117.270068</td>\n",
       "      <td>-10.675871</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>43.790190</td>\n",
       "      <td>13.533753</td>\n",
       "      <td>42.690814</td>\n",
       "      <td>30.256437</td>\n",
       "      <td>125.002893</td>\n",
       "      <td>13.289018</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>36.686353</td>\n",
       "      <td>5.010884</td>\n",
       "      <td>41.948751</td>\n",
       "      <td>31.675469</td>\n",
       "      <td>84.241415</td>\n",
       "      <td>0.664437</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   pelvic_incidence  pelvic_tilt numeric  lumbar_lordosis_angle  sacral_slope  \\\n",
       "0         63.027818            22.552586              39.609117     40.475232   \n",
       "1         39.056951            10.060991              25.015378     28.995960   \n",
       "2         68.832021            22.218482              50.092194     46.613539   \n",
       "3         69.297008            24.652878              44.311238     44.644130   \n",
       "4         49.712859             9.652075              28.317406     40.060784   \n",
       "5         40.250200            13.921907              25.124950     26.328293   \n",
       "6         53.432928            15.864336              37.165934     37.568592   \n",
       "7         45.366754            10.755611              29.038349     34.611142   \n",
       "8         43.790190            13.533753              42.690814     30.256437   \n",
       "9         36.686353             5.010884              41.948751     31.675469   \n",
       "\n",
       "   pelvic_radius  degree_spondylolisthesis  class_Abnormal  class_Normal  \n",
       "0      98.672917                 -0.254400               1             0  \n",
       "1     114.405425                  4.564259               1             0  \n",
       "2     105.985135                 -3.530317               1             0  \n",
       "3     101.868495                 11.211523               1             0  \n",
       "4     108.168725                  7.918501               1             0  \n",
       "5     130.327871                  2.230652               1             0  \n",
       "6     120.567523                  5.988551               1             0  \n",
       "7     117.270068                -10.675871               1             0  \n",
       "8     125.002893                 13.289018               1             0  \n",
       "9      84.241415                  0.664437               1             0  "
      ]
     },
     "execution_count": 266,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Load data\n",
    "data = pd.read_csv('/Users/garvitkhurana/Desktop/Projects/Data_Science/column_2C_weka.csv')\n",
    "# get_dummies\n",
    "df = pd.get_dummies(data)\n",
    "df.head(10)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 267,
   "metadata": {
    "_cell_guid": "e2541cfc-7d99-45ce-8486-80f7fec79257",
    "_uuid": "032c522e3f423822402c7f75fcfe3818894f68c8"
   },
   "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>pelvic_incidence</th>\n",
       "      <th>pelvic_tilt numeric</th>\n",
       "      <th>lumbar_lordosis_angle</th>\n",
       "      <th>sacral_slope</th>\n",
       "      <th>pelvic_radius</th>\n",
       "      <th>degree_spondylolisthesis</th>\n",
       "      <th>class_Abnormal</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>63.027818</td>\n",
       "      <td>22.552586</td>\n",
       "      <td>39.609117</td>\n",
       "      <td>40.475232</td>\n",
       "      <td>98.672917</td>\n",
       "      <td>-0.254400</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>39.056951</td>\n",
       "      <td>10.060991</td>\n",
       "      <td>25.015378</td>\n",
       "      <td>28.995960</td>\n",
       "      <td>114.405425</td>\n",
       "      <td>4.564259</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>68.832021</td>\n",
       "      <td>22.218482</td>\n",
       "      <td>50.092194</td>\n",
       "      <td>46.613539</td>\n",
       "      <td>105.985135</td>\n",
       "      <td>-3.530317</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>69.297008</td>\n",
       "      <td>24.652878</td>\n",
       "      <td>44.311238</td>\n",
       "      <td>44.644130</td>\n",
       "      <td>101.868495</td>\n",
       "      <td>11.211523</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>49.712859</td>\n",
       "      <td>9.652075</td>\n",
       "      <td>28.317406</td>\n",
       "      <td>40.060784</td>\n",
       "      <td>108.168725</td>\n",
       "      <td>7.918501</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>40.250200</td>\n",
       "      <td>13.921907</td>\n",
       "      <td>25.124950</td>\n",
       "      <td>26.328293</td>\n",
       "      <td>130.327871</td>\n",
       "      <td>2.230652</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>53.432928</td>\n",
       "      <td>15.864336</td>\n",
       "      <td>37.165934</td>\n",
       "      <td>37.568592</td>\n",
       "      <td>120.567523</td>\n",
       "      <td>5.988551</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>45.366754</td>\n",
       "      <td>10.755611</td>\n",
       "      <td>29.038349</td>\n",
       "      <td>34.611142</td>\n",
       "      <td>117.270068</td>\n",
       "      <td>-10.675871</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>43.790190</td>\n",
       "      <td>13.533753</td>\n",
       "      <td>42.690814</td>\n",
       "      <td>30.256437</td>\n",
       "      <td>125.002893</td>\n",
       "      <td>13.289018</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>36.686353</td>\n",
       "      <td>5.010884</td>\n",
       "      <td>41.948751</td>\n",
       "      <td>31.675469</td>\n",
       "      <td>84.241415</td>\n",
       "      <td>0.664437</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   pelvic_incidence  pelvic_tilt numeric  lumbar_lordosis_angle  sacral_slope  \\\n",
       "0         63.027818            22.552586              39.609117     40.475232   \n",
       "1         39.056951            10.060991              25.015378     28.995960   \n",
       "2         68.832021            22.218482              50.092194     46.613539   \n",
       "3         69.297008            24.652878              44.311238     44.644130   \n",
       "4         49.712859             9.652075              28.317406     40.060784   \n",
       "5         40.250200            13.921907              25.124950     26.328293   \n",
       "6         53.432928            15.864336              37.165934     37.568592   \n",
       "7         45.366754            10.755611              29.038349     34.611142   \n",
       "8         43.790190            13.533753              42.690814     30.256437   \n",
       "9         36.686353             5.010884              41.948751     31.675469   \n",
       "\n",
       "   pelvic_radius  degree_spondylolisthesis  class_Abnormal  \n",
       "0      98.672917                 -0.254400               1  \n",
       "1     114.405425                  4.564259               1  \n",
       "2     105.985135                 -3.530317               1  \n",
       "3     101.868495                 11.211523               1  \n",
       "4     108.168725                  7.918501               1  \n",
       "5     130.327871                  2.230652               1  \n",
       "6     120.567523                  5.988551               1  \n",
       "7     117.270068                -10.675871               1  \n",
       "8     125.002893                 13.289018               1  \n",
       "9      84.241415                  0.664437               1  "
      ]
     },
     "execution_count": 267,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# drop one of the feature\n",
    "df.drop(\"class_Normal\",axis = 1, inplace = True) \n",
    "df.head(10)\n",
    "# instead of two steps we can make it with one step pd.get_dummies(data,drop_first = True)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "_cell_guid": "f55c171e-36d4-4873-a0f6-e8ee411d0b64",
    "_uuid": "5465db779fed984dd89ad591889d1dadb88fc056"
   },
   "source": [
    "Other preprocessing step is centering, scaling or normalizing \n",
    "* If you listen my advice and watch KNN in youtube, you have noticed that KNN uses form of distance for classificaiton like some oher methods. Therefore, we need to scale data. For this reason, we use \n",
    "    * standardization: ( x - x.mean) / x.variance   or   x - x.min / x.range\n",
    "* pipeline: The purpose of the pipeline is to assemble several steps like svm(classifier) and  standardization(pre-processing)\n",
    "* How we create parameters name: for example SVM_ _C :  stepName__parameterName\n",
    "* Then grid search to find best parameters"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 268,
   "metadata": {
    "_cell_guid": "5d4aef88-f75e-4485-b481-bdeabbc0e4c7",
    "_uuid": "4690eac9c778ba7230fcee5a782a46396aa8c4ef"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Accuracy: 0.8548387096774194\n",
      "Tuned Model Parameters: {'SVM__C': 100, 'SVM__gamma': 0.01}\n"
     ]
    }
   ],
   "source": [
    "# SVM, pre-process and pipeline\n",
    "from sklearn.svm import SVC\n",
    "from sklearn.preprocessing import StandardScaler\n",
    "from sklearn.pipeline import Pipeline\n",
    "steps = [('scalar', StandardScaler()),\n",
    "         ('SVM', SVC())]\n",
    "pipeline = Pipeline(steps)\n",
    "parameters = {'SVM__C':[1, 10, 100],\n",
    "              'SVM__gamma':[0.1, 0.01]}\n",
    "x_train, x_test, y_train, y_test = train_test_split(x,y,test_size=0.2,random_state = 1)\n",
    "cv = GridSearchCV(pipeline,param_grid=parameters,cv=3)\n",
    "cv.fit(x_train,y_train)\n",
    "\n",
    "y_pred = cv.predict(x_test)\n",
    "\n",
    "print(\"Accuracy: {}\".format(cv.score(x_test, y_test)))\n",
    "print(\"Tuned Model Parameters: {}\".format(cv.best_params_))\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "_cell_guid": "b343630f-75c8-4362-8cb0-858e71a2f4b0",
    "_uuid": "3a8fede9eadd9bd0438f7cb50c0155491d3f776b",
    "collapsed": true
   },
   "source": [
    "<a id=\"10\"></a> <br>\n",
    "## UNSUPERVISED LEARNING\n",
    "* Unsupervised learning: It uses data that has unlabeled and uncover hidden patterns from unlabeled data. Example, there are orthopedic patients data that do not have labels. You do not know which orthopedic patient is normal or abnormal.\n",
    "* As you know orthopedic patients data is labeled (supervised) data. It has target variables. In order to work on unsupervised learning, lets drop target variables and to visualize just consider *pelvic_radius* and *degree_spondylolisthesis*\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "_cell_guid": "e492f7d0-5d05-4717-a149-e046d9435607",
    "_uuid": "f4927a23d072a9416b57f0bd611d71411407a907"
   },
   "source": [
    "<a id=\"11\"></a> <br>\n",
    "### KMEANS\n",
    "* Lets try our first unsupervised method that is KMeans Cluster\n",
    "* KMeans Cluster: The algorithm works iteratively to assign each data point to one of K groups based on the features that are provided. Data points are clustered based on feature similarity\n",
    "* KMeans(n_clusters = 2): n_clusters = 2 means that create 2 cluster"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 272,
   "metadata": {
    "_cell_guid": "fa3dce86-b3cd-4a8f-b689-e5bc5f2fe551",
    "_uuid": "53175af6bce4acd6f743250c240033db04921ad4"
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# As you can see there is no labels in data\n",
    "data = pd.read_csv('/Users/garvitkhurana/Desktop/Projects/Data_Science/column_2C_weka.csv')\n",
    "plt.scatter(data['pelvic_radius'],data['degree_spondylolisthesis'])\n",
    "plt.xlabel('pelvic_radius')\n",
    "plt.ylabel('degree_spondylolisthesis')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 282,
   "metadata": {
    "_cell_guid": "63a3245d-0056-46a0-ba65-a1f61d84441f",
    "_uuid": "9550966d8637f2d5c61633d597cd3746fb294e7d"
   },
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAXoAAAENCAYAAAABh67pAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADl0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uIDIuMi4yLCBodHRwOi8vbWF0cGxvdGxpYi5vcmcvhp/UCwAAIABJREFUeJzs3WeAVOXd9/HvOWfazvYGy+7SO4iEIiCKIKwaxYImGktijOlEDJrHxDRNzG1uo/HWoCYmMdEkmqImosGOCNjQpSm9t+29zE6fcz0vhi3DzDaYBXb8f97onJ05c81h9zfXuaqmlFIIIYRIWPqpLoAQQoi+JUEvhBAJToJeCCESnAS9EEIkOAl6IYRIcBL0QgiR4CTohRAiwUnQCyFEgpOgF0KIBCdBL4QQCc5yqgvQqqys7FQX4bjl5ORQU1NzqotxSsk1CJPrINeg1cm4Dvn5+T16ntTohRAiwUnQCyFEgpOgF0KIBCdBL4QQCU6CXgghEtxpM+pGiM48u2cnrx06SLPfT7Yjia9MOIOZeYNOdbGE6DekRi9OaysO7ONP27awu6GecncLW+tquG/9hxxubjrVRROi35CgF6e1Vw8doCUYjDhW7fXwt53bT1GJhOh/JOjFac0XCsU83hzwn+SSCNF/SdCL01phSkrUMQONswbknYLSCNE/SdCL09p3J09jdHpG2y+qXdeZNnAgV4wYdUrLJUR/IqNuxGkt0+Hg9/Mv5OWD+9ldX8fs/ALOGVSArmmnumhC9BsS9OK0ZzMMrhw5+lQXQ4h+q8dBb5omd955J1lZWdx5551UVVXx8MMP43K5GD58OEuWLMFisRAIBHj00UfZv38/qampLF26lAEDBvTlZxBCCNGFHrfRv/LKKxQUFLQ9fvrpp1m4cCHLli0jOTmZVatWAbBq1SqSk5N55JFHWLhwIc8880z8Sy2EEKLHehT0tbW1bNy4kQULFgCglGLbtm3MmjULgHnz5lFcXAzA+vXrmTdvHgCzZs1i69atKKX6oOhCCCF6okdB/9RTT/HFL34R7WgHWHNzM06nE8MwAMjKyqKurg6Auro6srOzATAMA6fTSXNzc1+UXQghRA9020a/YcMG0tPTGTFiBNu2bev2hLFq71qMERIrV65k5cqVANx3333k5OT0pLynJYvF0q/LHw9yDcLkOsg1aHU6XYdug37Xrl2sX7+eTZs24ff78Xg8PPXUU7jdbkKhEIZhUFdXR1ZWFgDZ2dnU1taSnZ1NKBTC7XaTEmPSS1FREUVFRW2P+/PWY7J1mlyDVnId5Bq06ldbCV5//fU8/vjjPPbYYyxdupQzzjiDW2+9lYkTJ7Ju3ToAVq9ezfTp0wGYNm0aq1evBmDdunVMnDgxZo1eCCHEyXHcM2NvuOEGVqxYwZIlS3C5XMyfPx+A+fPn43K5WLJkCStWrOCGG26IW2GFEEL0nqZOkyExZWVlp7oIx01uVeUatJLrINegVb9quhFCCNG/SdALIUSCk6AXQogEJ0EvhBAJToJeCCESnAS9EEIkOAl6IYRIcBL0QgiR4CTohRAiwUnQCyFEgpOgF0KIBCdBL4QQCU6CXgghEpwEvRBCJDgJeiGESHAS9EIIkeAk6IUQIsFJ0AshRIKToBdCiAQnQS+EEAlOgl4IIRKcBL0QQiQ4CXohhEhwEvRCCJHgJOiFECLBSdALIUSCk6AXQogEJ0EvhBAJToJeCCESnAS9EEIkOAl6IYRIcBL0QgiR4CTohRAiwUnQCyFEgpOgF0KIBCdBL4QQCU6CXgghEpyluyf4/X7uvvtugsEgoVCIWbNmcc0111BVVcXDDz+My+Vi+PDhLFmyBIvFQiAQ4NFHH2X//v2kpqaydOlSBgwYcDI+ixBCiBi6rdFbrVbuvvtuHnjgAe6//342b97M7t27efrpp1m4cCHLli0jOTmZVatWAbBq1SqSk5N55JFHWLhwIc8880yffwghhBCd6zboNU3D4XAAEAqFCIVCaJrGtm3bmDVrFgDz5s2juLgYgPXr1zNv3jwAZs2axdatW1FK9VHxhRBCdKfbphsA0zT5wQ9+QEVFBRdddBEDBw7E6XRiGAYAWVlZ1NXVAVBXV0d2djYAhmHgdDppbm4mLS2tjz6CEEKIrvQo6HVd54EHHqClpYVf//rXlJaWdvrcWLV3TdOijq1cuZKVK1cCcN9995GTk9PTMp92LBZLvy5/PMg1CJPrINeg1el0HXoU9K2Sk5OZMGECe/bswe12EwqFMAyDuro6srKyAMjOzqa2tpbs7GxCoRBut5uUlJSocxUVFVFUVNT2uKam5gQ/yqmTk5PTr8sfD3INwuQ6yDVodTKuQ35+fo+e120bfVNTEy0tLUB4BM6WLVsoKChg4sSJrFu3DoDVq1czffp0AKZNm8bq1asBWLduHRMnToxZoxdCCHFydFujr6+v57HHHsM0TZRSnH322UybNo3CwkIefvhh/vnPfzJ8+HDmz58PwPz583n00UdZsmQJKSkpLF26tM8/hBBCiM5p6jQZElNWVnaqi3Dc5FZVrkEruQ5yDVr1q6YbIYQQ/ZsEvRBCJDgJeiGESHAS9EIIkeAk6IUQIsFJ0AshRIKToBdCiAQnQS+EEAlOgl4IIRKcBL0QQiQ4CXohhEhwEvRCCJHgJOiFECLBSdALIUSCk6AXQogEJ0EvhBAJToJeCCESnAS9EEIkOAl6IYRIcBL0QgiR4CTohRAiwUnQCyFEgpOgF0KIBCdBL4QQCU6CXgghEpwEvRBCJDgJeiGESHAS9EIIkeAk6IUQIsFJ0AshRIKToBdCiAQnQS+EEAlOgl4IIRKcBL0QQiQ4CXohhEhwEvRCCJHgJOiFECLBWbp7Qk1NDY899hgNDQ1omkZRURGXXHIJLpeLhx56iOrqanJzc7nttttISUlBKcWTTz7Jpk2bsNvtLF68mBEjRpyMzyKEECKGbmv0hmHwpS99iYceeoh7772X119/nZKSEpYvX86kSZNYtmwZkyZNYvny5QBs2rSJiooKli1bxje+8Q2eeOKJPv8QQgghOtdt0GdmZrbVyJOSkigoKKCuro7i4mLmzp0LwNy5cykuLgZg/fr1nHfeeWiaxpgxY2hpaaG+vr4PP4IQQoiu9KqNvqqqigMHDjBq1CgaGxvJzMwEwl8GTU1NANTV1ZGTk9P2muzsbOrq6uJYZCGEEL3RbRt9K6/Xy4MPPshNN92E0+ns9HlKqahjmqZFHVu5ciUrV64E4L777ov4cuhvLBZLvy5/PMg1CJPrINeg1el0HXoU9MFgkAcffJA5c+Ywc+ZMANLT06mvryczM5P6+nrS0tKAcA2+pqam7bW1tbVtNf+OioqKKCoqanvc8TX9TU5OTr8ufzzINQiT6yDXoNXJuA75+fk9el63TTdKKR5//HEKCgq49NJL245Pnz6dNWvWALBmzRrOOuustuNr165FKcXu3btxOp0xg14IIcTJ0W2NfteuXaxdu5YhQ4Zwxx13AHDdddexaNEiHnroIVatWkVOTg633347AFOmTGHjxo3ceuut2Gw2Fi9e3LefQAghRJc0FatR/RQoKys71UU4bnKrKteglVwHuQat+lXTjRBCiP5Ngl4IIRKcBL0QQiQ4CXohhEhwEvRCCJHgJOiFECLBSdALIUSCk6AXQogEJ0EvhBAJToJeCCESnAS9EEIkOAl6IYRIcBL0QgiR4CTohRAiwUnQCyFEguvxnrFCCNFT7zbs4dXaLXhCfnKsqXwl/xwG2TNOdbE+tSTohRBx9WbFFn5fuobmkBeA/d4aSg/Uc9+oz5NqccTtfYIqxCfNJZgoJqcMxqobcTt3opGgF0LE1QulH7WFfKtSfwPLqzfxpUFnx+U9drZU8FjJKsp8DSgUg+zpfCN/LpNTB8fl/IlG2uiFEHHlCvpiHi/3N8bl/KZSPF76Nkd8dYQwMVGU+hp4omwtQRWKy3skGgl6IURcZdtTo44ZaJyR3LP9TbtzyFtLha8p6niZr5FdLRVxeY9EI0EvhIirb48sYqCtPew1YKwzjwuyJsbl/BZNx9C0qOOGpmHVpTU6FrkqQoi4GpM6iF+O/BzPV66nLuhmvHMQC3POjFtnaaE9kwJ7Jrs9lRHHC+yZjEoaEJf3SDQS9EKIuMu2pvDNwnl9cm5N07h9yIU8dOTNo52xMMiWxq2Di9Bj1PSFBL0Qoh/Ks6fzq1Gfp9LfhKkUebY0NAn5TknQCyH6rYG2tFNdhH5BOmOFECLBSY1eCHHaerF6M+807MZrBsi1pvK1/DkUODJPdbH6HanRCyFOSy9UbeQflR+y11NFia+eTa7D/PLgy7SEYk/IEp2ToBdCnJbebdiD1wxEHCv1N7Ci5uNTVKL+S5puhPiU2+uu4pmKdTQE3SQbduZnjqMx6CHD4uTcjNGnbLEwzzEh36rcF5+lFD5NJOiF+BSr8DXyq0OvUh1obju2raUUBeho/Lt6A3cMvYihjpyTXrZsazJl/oaIY1bNYFb6iJNelv5Omm6E+BT7V+VHESEPoI7+10RR4qvn9yVrTn7BgJsGnUNeh+GTVs1gcspgZqRJ0PeW1OiF+BRrDHq7fU6lvxlPyE+SYYv5c58ZZE39TmoDLczNGEsO8an9j3QO4FejrmZ51UZqAi5mpY/g7PRRMvv1OEjQC/EpNjwpm42uQ10+x6LpWLTY7fRHvHX86tCrlPrqUcBrtVu51DuVqzOmxKV86ZYkvpx/TlzO9WkmTTdCfIp9fsB0RibldvmcMc6BnXbIPl66mpKjIQ/QGPKwonwjJd76OJdUnAip0QvRBXfIz3OVxRz21ZFpcXLtwJnk2FJOdbHiJsmw8cuRV/FS9cfs81Rh1yyU+BuoD7Rg0Q1GJw3glsELYr42qEJU+ZujjjcGPKys285NUhM/bUjQC9EJT8jPT/b9h/3emrZjW1yl/Hj4QoY4sk9hyeLLrlu5euB0lFIc8tYC4REvNt2KvYv13XV0rJ006WRZk/ukrOL4dBv0v/3tb9m4cSPp6ek8+OCDALhcLh566CGqq6vJzc3ltttuIyUlBaUUTz75JJs2bcJut7N48WJGjJAectE//btqQ0TIA1QGmvhb+Qf8ePilffKee91VvFzzCZoGl+ZMZkQ3zSo9UeFr5PHS1VT4G7FqBpOSC/lqwRwMrb3l9pC3hmVH3qLMGx7OmGdPZ0nhAkY4O39/XdOYkJxPub+hrekGYHBSFhdkTTjhcov46baNft68efzoRz+KOLZ8+XImTZrEsmXLmDRpEsuXLwdg06ZNVFRUsGzZMr7xjW/wxBNP9E2phTgJWmu3x6oPuvvk/Z6tLOZnB17k7YadrKrfyc/2v8jyqk0ndM6AGeKXB19ms+sIFf4mjvjqea1uK78vXd32HKUUvzn8Fvs81XhUAI8KcMBbwyMlb2Eq1fnJgW8VzmV+5ngG2dLJtqQwzpnHXRM/1+kIHXFqdBv0EyZMICUlsk2yuLiYuXPnAjB37lyKi4sBWL9+Peeddx6apjFmzBhaWlqor5dOGdE/5Vij9z4FSDbscX8vd8jPW/U7cHVYx6Up5OWNuq1RywD0xjsNuyn1Rf4Nmii2usraNtI+6K2lzNcQ9dpSXwP7PVVdnt+iGSwZvIBHx97AY+Nu4L5Rn2d8WsFxl1f0jeMaddPY2EhmZngFuczMTJqawhv11tXVkZPTPoY2Ozuburq6OBRTiJPvCwPPYpAtPeJYpsXJ5JRC/lC6hrfqtreF5Yna56mi0h+94XWVv5nDndxZ9ESVv4kQ0bVynxkgYIbLrlCgRT9HoTBjvDYWQ9Nx6NbjLqfoW3HtjFUxbvM62/Vl5cqVrFy5EoD77rsv4guiv7FYLP26/PGQiNcgB3go/UZ+t28l1b4mkg0HLSEvz1atx2sGMNB5vWEH//eZL5JhC3c+Hu91GO3USD+cRGPQE3E8zeZk1IDB5DjSO3ll2Ee1e3m2ZB3eUIBRKQP5xogFOC12LkuayWv122gIRDY35TkzGTwwH4AslUVheQ57XRURzxnizGbm4AkRbfk9kYi/C8fjdLoOxxX06enp1NfXk5mZSX19PWlp4WnK2dnZ1NS0d17V1ta21fyPVVRURFFRUdvjjq/rb3Jycvp1+eMhUa+BA7htUHh44Ws1W/hj2dq2GnIIk92ucu7f+iK3D7kQOP7r4ABGOHLZ5DoccXyUIxfdFaDG1fk5367bwZ/L36M5FJ7lurnhEJ/UHuKXo64iWTM4J20Uqxt2tS3vO9CaxvU5Z0WU81t55/FIyVvhPViVIt+RwTfzzqO+tvd35PH4XVBK4VNB7Jql324ReDL+JvLz83v0vOMK+unTp7NmzRoWLVrEmjVrOOuss9qOv/baa5xzzjns2bMHp9PZadAL0d9sdB2O2QxS5otPP9Sdwy7mD6Vr2eepAjTGOAfytfw53b7u5dotbSHfap+nmtX1uyjKmsDXC85jfuY43qzbTobFycKcM0m1OCKeP8o5gIdGX8teTyVKhR/3tiYfL8urNvF2/U5cIR/pFgeX5EyiKGviKSlLoug26B9++GG2b99Oc3Mz3/rWt7jmmmtYtGgRDz30EKtWrSInJ4fbb78dgClTprBx40ZuvfVWbDYbixcv7vMPIMTJ4tRjjySxafFpAbXrVpZ0MjmpK80x1qsJYbKjpZyio8McRzoHMNI5oMvz6JrGGGder98/lkPeGp4se4+agIsk3crczLFcmjO529e9U7+bZ6uKcZt+AGqDLv5S/gH5tgwmpEgn7/Hq9jd06dKlMY/fddddUcc0TeNrX/vaiZdKiOOiMDgMWAgR/1D43IBpfOwqoT7Y0nYsSbcyL3Ns3N+rN9ItSVQGIjtyLehMSik8JeVpDLi57+CrlPvb140v8dWjlOKy3M90+do367e3hXyr5pCXl2o+lqA/ATIzVpwyOlWkar/HoAqTFFzqSwQ5vok2FvaQpj2AhTJAJ0ghDeqnmAyKW3kHO7L4dsFcnq/eQH3ATbJh47yMMVyYfcYJndfgIE5ewiQFN1ehyOjV6y/P/Qx/LFsb0ZE7yjmQORmjT6hcx+sfh96PCHkIbyKytmF3t0HfOhLoWCcyxFRI0ItTRKOBLO3/YdHaOx+t7KZe3U2QM3t5thDp2i+xagfajtjYTga/oE79Nk4lDpuRPoIZ6SMIKTMubdgp/Jkk7UUMLRyMSeoNmtRt+JnZ43OcmzGabGsyL1RvwmcGGJU0kGsGTj9lbewV3ugx+dD5jlEdjUzKZYe7POKYBqfs7iRRSNCLKE6ew6G9iY4Lk2xc6iv4mRrX90jh6YiQBzC0WlL4Kw3q1706l5UtGJREHbdQgkEJIeIfEp2GqPKSzBPYtB0o7LSoawgQuxarU0uS9nJbyANYtApSeYJaNYNwxPXM+OR8xif3bARGPDh4A6f2EhouTLJoVl9ruxs7J2csa6q2E8CMeE2OtfvF4L446Gz2earZ66kioEI4dAvjnIO4PKfrOwHRNQn6OGlq9LDy1R24XD7mXziOwiH9c7RREitI1p7C0FrboctI5z7q1G8IxbEZxNDKYh8nejXE7mgE0Yh1y29CzOPxZ7CPFO0vGPWbSNGaaR0RaGUbzWoxXi6Kek0S/8bQoidDGRwkQ/s+XrUALxcSa16jThU6LQQZApzcPV3trCZNexRda+0XOIjBz6lTv8Ekj/kDJ/Lfw8VscZW0hf0gWzpf6cFqlg7dyr0jr6K46QA73RVMThnM5JTCfjvE8nQhQR8HGz7az7IHXqOm2gXAh+/t5/wLxvL566ef4pL1nkN7vUPIhxlaFU71DM38v7i9T1ANBe39qOOhXrZPA/iZTIhCLETeIQQZRIghx13GnrKymXTtXixadfhAh0wytEac/Aevig56h/ZOzPPpWgAHxdjZjIP3aFaL0aklwCg0AmRoP8fCPjT8hBhIs/oafmb3xUeLyakt7xDyYRatkhT1N5q4A0PT+fHwS3m/cS8fNR5ggC2NRblTooZ0dkbXNGamj2Cm7A0bNxL0J0gpxV/++E5byAO4mn288/Ye5l80nqzs/rNca021C8NaR25W9M8MrZEezobvkRZuwK4+xKrtbzsWVANxqa8ex9msNKlbSWPZ0SYcnSCDaVJLSNd+gZU9gEaA0TSp76FwxutjAJCi/bU95GPQaSB8d9FeM9doQMfV6WsANC2AXb2PTduEhocQeShsEX0ROvtJ4xFq1SQUsdfmiTcNT8zjutbQ9jtiaDpzMsYwJ2PMSSmT6JoE/QlqqPdQWxPd3NBQ72HDh4e44JL+sVzrf/61kdVv7uSWrxnkHlM5VMrAp86K6/spUqhTD5OinsKiHSakMmjhy4QYfFzn8zOdOnU/qTyCrrnwqItJ1Z7Crm1se46Fw+g0Ut+hD0CjhVSWYdX2ozDwqRm0cBO9WQYqHOSdM0mPcT4biu5XeNS0ENrRLwQLpSgVXS6LVk6SehU31/S0yCckxEBgV8QxpTT86sRGH4m+I0F/gpKSrNjt0ZfRYtXJye0ftfmaaher39xJY4OX3/3pLArzmxg6uAHDAFPZ8DMZD5fE/X0VaTRza1zuFCxsI0P7BRYtvF6LjW2dPG8PBmWEyAcUmdqd2LQtbT+3shddNdFM7PkjsYSDvJOfqTTc6vKo4wonAcZgobLH7wOgaWbs45y84YfN6hYsHMKqhfeaVcogwATcXHXSytBbOlUk8SomaXi5KO53dac7CfoT5EiyMv6MQiort0cEVn5hBpOnHl/t9GR7d/UeGhvCsyvr6pP57p2XcMmFuxk/tpnx065F2eZyum8vnKr9oS3kATQtSKyl1HVa0KknRD5WNmNhT8TPNS2InWKalQ+wA0FsbAbAz2eI9SfTom7AwiEMrX1dGKU0QgykSd3aaft5k/oR8L/Y2BDVL9IbQZWLh4uP+/W9ZTKAOvUYTvUcFu0wfjXl6PufnqtXOvk7yVq441spSOZ5GtSdBJl0qot20kjQx8HSH1yMUkH27KwkGDTJy0/nK9+cjW6c3uHYKiPTiabRFox+v4XlKybwfnEq/zv9PCynecgDGEQvHhVroEaIPAKMAsDCAXQtevkAjRZ0GtGpIF17EMvRoZtBBtOobsPGduzaB2iY+NUkXHyVenUPmfwQQ2s++t4KQ1Xg5K+kaP9ApwGTVAJqFLrWREjl4uZ6GtU9ZGh3YFDco88ZUqmYZGNQgYaXEPm0qOswidGx0ocUKbTwlbj22/QFnXqc2gtto5s0LdwElsaj1KnH6c0Q1v5Mgj4OrFaDm799LhDunO1vQ8FmzxnJa//dSnlp5GzGUWMHYLH2fOieq9nLm69ux9XkY96F4xh8EoeYmkQ3kykFiiR0Ldx5GFJZtKjPE66pg58ZhFQGhhbZxm6SisJGuvZQW/MEgJWDZPITdDxoWvDosW0YHMGtrozqpNQ0sLMz4gvHyvbwYw0c6j3q1T2YKjtm3igFIfKPfvE0ESIft7oMN18In4cGAkz51DVD9IadNTE7yg2q0KnDJHH2/u2KBH2c9XXIK6VY/uwmNq0/jM8bJC3dwVXXTWP8xOMf426zW1i8dB5//dMH1Fa7sFgNRo0ZwFe+dQ5KKTzuADab0WXob95wkId/9QrVVeGOw3Xv7WfugjFc88UT68Q1KMXOe4QoxMcsOmtCCqgRWNkdEaomadSph3CqFYCGm6s7zAUIYOMjFHaUMtC0o5twKB2DanK1q2O2e+s0R7yHpilsagtBhqMfDf+Ojv116PjYopWRxuM0qe9hU5uwaO3t9Upp+NRUGrgfgwp0aggyui3UA5yK1RxN7KzBob1LSA2ghS/0ermGSD6s7MEk/bg74btjkhnx79tKYUPRs+GeiUCCvp95/u8beP3lbQT84V/civImfvWz1xg5OofvfG/+cQ/nHDI8m5/8z6V4PQEsFh2L1WDbJ2U898x6Gurd2OwWxk/M48avz8Y4pklKKcWTf1jTFvIQHmL67uo9LLhoPNm53c+IPJZGA1ksxaIdQdNCmMogyCga1H2YhO8UDErQaSTASGza3hhNNTqKdJr57jHHg2Rqd2Djk7bOTVPZAIWuBdCIbs7piqE1ksSrKKWhxdipqSs6NYQYRJP6Fmk8ho6HEJm41E34CO/XEKKgTxZp6ykLW0nR/oWVLWg0oWvm0TuSNTSouwnS+0XdHLxCivYPDMqOdkyPpEH9Iu5DRH2cQ5AhWDkQcTzAWFSMu8BEdfo3vooIm9cfbgv5Vkop9u6u5vGHV5/w+R1JVixWg6ZGD39+/F0O7Kuhvs5NZXkTa9/ew1+f+ICK8kZCofbRH81NXmqro8eENzZ4+fCD8B+Y3x9kzcpdLH92E9WVXc9+1WgkW/sWVv1gW01M10LYtF2kav+HRjOZ2lKytcVkad8lW7sZg9Ko8xhaA3aiJ2Ul8Qo2tkSMYNE1P7rWk5Erse9qLFoNx9NgrUjGwcuka8uwaDXoWgsGtZ2OGjrZHLxBpvYTHNo7GFpDOOSPsmhlpGqP9/qcOjWkaH9u+xLXtWbs2mbStfviWfTWUtKg7sGnJhNUAwiqfDxqPo3qR33wXqcvqdH3I0opvN7o5oFW5WWN1NW29LhWHwqZ7NlRiVKKMRPyMAydspIGnv7zOg7sq8HdErlcrBlSrHlrF2tX7cZi0RkzbiC3fr8IhyP2EFPDopOTm0LJkXoee3AV5aWNKAVvvb6D8+aP4eobYs8cTuFvESNoOrJwmHTtPuza5rZjVspiji83lZ1g1LINJkna61G38j2lAKVs6Jo/6metdxRKEdG53fqz1uOtQioFvxpDmvY7dK39i1LXPDhYe7STtes15PuaU/t3VB9GRwadTxTrTBIvHv1ijGRhP8dOLIuHEIOpV78BvIS/qE/P0UF9SYK+H9E0jayc5IhZuB253QEC/s6/CDrau6uKPz/+LhXl4anseXlpXPeVmTz9p3VUlDV2+jplhjeN9odCbP24jHt/+jI/+9/LmDCpkMqKxohwKyhMZ9qMofzq569SVtJ+zqZGL2tW7ebc80cxKD+6jdfQohcoa2fBcsxtOITHlx8bpEGGEWBah2eZZGg/wcKOLs7fNV0LEVS5mErH0sl6PZoGprKgEYpoylHo4aUfCHcSgw9j6ErTAAAgAElEQVSn9nLMLw1Dq8WmivGy8LjLeuKC6HS9e5YZ147gvh7E8Olpkz+WNN30M1/44nRyOmnzDgZCvPV69yFmmoqn/vAepUcaCAVNQkGT0pIG/rBsTZchH8vhA7Ws//AQt97xWeYuGMuggnRyB6ZyxuQCvvuDIpSpqKmK/mJqbvTy7tt7Y54zpGLvcqQU+NQUOmsi6RjySml4VeT4fztrsVEc0fzQmVhj8NveBx9Naimm6rw9WSMY1V6vayYhBlCrnsStLsfKoZghD2CqJEIM7bacfcuCIq3Tn5rKgVfN6/VZPSwiqHKjjgcZgURS35AafT8zauxAfvjzi/npHctxt0S3KW/7JHYts6PDB2upLI9uJ29u6l0nJIQDcdeOCi653OAr3zrn6LH2IaZmyIw9WkeD9Iyktof1dW7MkEl2bgot3IhdbcCiHenwPjoedTYuvo0ZLMftbiInxYPF6GSmqKawsxG3ur7tmEN7p0ft8CGVSUANwc7HMcfiK5LxMxmvmoODt9A1X4yzGMReOTN8Qrv2ftsQzVgCjD5FI2siudUlpPIkutb++2IqG0GG41XzcfOFXp8zvPT1zaTw94jO2Eb1g3gWXXQgQd9HlFJ89P4Bij84SJLTysJFZ5KX3/lU+Z6or3NTXdnEP/9aHDPkAfz+7tueNU2DGKNDdEPH0HX8xzT/OJ02TGXi9UQHk2HRGTchsgbecYipboTb8isrmiIq4nl5aZw3fwz1dS08/ps1lJc2YpomuQNT+eq358CQX5PKHzCoxsSBS91IkDN4evUW3t1+Ni2ecWQ43Vz6mR18afaOmGEbCKqIvtOQGtjWOhA0NV7fOopPjuRx1rAy5k/Yh65BSGXgUl/Cw1UkqZdI4fcRs1ZNZcOrzgbsNHEHHnUB6fwSg+pjmo3ysajyiDA3lQOPuvDo/6fFbKkwlR2vmnt0pNCpn4/h4SpMlY2Tl9DwE1AjcfGNEx677+VivGo+VvZiktZnwytFmAR9H/ndQ6vZWHyIQCBc4/xkUwlf+urZTJ81rNfnCgZNfv+bNezaUUFzkxfT7LxdYcDAroenbfzoEC/+ezOhYHRN2Om0MXrsALZtKcN3tNPX4bDg8wUIhWK/58jRuUybOazL97zpG7NRCvbsrMQfCJKdk8L1X5nJxoMV/PFfH9FS0YKjMYimoLnJx+O/Wc09919Bo/HTiPO8uXk/y9ftxhsIAjbq3Tb+/M455GWfzbxRf8RuCS97EFIa/qDBX94dxFVz2+8uWvgCDrUGn7+KW/52KTsrcgiELLy0aQJ/ef88fnVtFdbkzxNiOAAeLserirAHl9HYtIUWH3x0YCQfl32GWy8LkGSzEmAKNepp0ngEKzuwGAae4EiaWEIay7Cz4eiSCzl41Rx8zGfDvnLe2jyVFEuIL539IaPzwksnmCqJZvV1PKfZmjE+5uJTc/vgzPbT4q7l00CCvg/s31vNls0lbSEP4dUsV7zwMdNmDu31pKrn/r6e4nUHumw3BsgvTOfGr3e+LnllRRN/feID6uvcMX/e3OSl9EgD1315Bnt3VZGS6uDD9/bjPeb5mgZ5+elMPWsIV1w9BV3v+vNYrAZfv2UOwaBJMBDCYjP4+T/XsvVgNQHDhAIbvmwLafu96EEoL21k5/ZKJkwaxN7yOtZsPcygrBTe2Xb4aMi3a/GZ/N8rSbw77Dyqm5xUu1Jw+61YjRBlDWlMGFnD+MHh9mBFOvXqHv64cgVbStvvQgIhnR1lDq57bCRzJtSweOEw9KP/Rgon3/3bWWw7MrzDu5bg9gX5+fWt4Wenif8HCnIycmiqCY8oaeJOdFWPQRlBhqBI5fFXN/DWJwfx+IPAMD7Ym883z9/KFdNqeW7DfF7amILLs4KMZAefmz2Os8f1bHesQDDEy+v3su1INelOO4auU17XTLLDxrVzJjB0wIlMbBL9nQR9H9hYfBi3O7pppaHeg88bxJHUu+Fdu7dXdBvy2TnJ/Pz+K7DZOv8nffmFTzoN+VaVFU2sX3eQO376Wfbvreat17ZHPUcpqKlqZs/OKoKBUNvQSqUUHn8Qh83SFpQdWSw6FovOf4v38PGBStpuTDSNUJJBS76d1MM+QiETny/AIys+4r0dJbi8ATTAZok9hr222cOrW8YSbhuKfN/3d5S0BT1AiJHsqjwDYqyN0+ILsPKTg2SlJnH93PCSuwcqGzhQFT28cF9FPdWNbnLTu27CMMlsm+D10e4y3t5y6GjIh9W7bfz9o3lojon84a31uLzhUS4VDS389pUN5KY7GTWo63VsAsEQP356NTuO1MTspt5+pIbvLZrJmcMGdnkekbgk6PvA4CGZGBY9qnnEkWTFZuv9tm/dhXxKqp0rrp7SZcgDeNyxR3gcq7Ym3CY9KD+djKzkmBOcAgGT3TsrefL373HL9+bz8oc7eOatjTS6faQ4bMw9YwjXnBt7Lf7N+zuEfAchezikB+aloWdaeWfVEdy+cCgqwBfsrv8h+stl04EK/MFQ25eEUgqrpfORHcGQyaotBynITqWhxUtGsr2tDB15fEEqG1y8unEvDS4fV509hsKc2H0wHn+Ae/71LrtKavDHaDKra/bwzNptuLyRlYP6Fi/PvbeDH36+6y34Xt24j50lsUMewl+E/1i7XYI+Tvas388LD75KU62LlIxkFn6niElzx53qYnVJgr4PTJ81jFde3MLB/e37gdpsBtNmDDmuFS1HjxvAgf01EZ2ZhkUnvzAdu81CyFS8/eZONnx4kM9dO5WhI3JinmfmOSPYtOFI1MzaYzXUezBNRZLTxszZw1j56o5OJ2qVHK5nZ2kNj7z0Pg2u8Kid2mYPz723g8wUBxd8Jno7uNSk2BtuaCbkDkzl6i+exdvbj8QMWEPXCHXRR3Gsg1VN3LzsvwxIT8ai69S3ePEFglh0jWAn56mob+GBF9a1vZ/N0PDH6KO4869vt/3/m5v3s2DyMP7n5uhx74+/upEtB6s6LWOyw0ZVQ+xliuuaY+/m1NG2w9Uxvzg7anT3fkSViFa6u4LfLv4LdWXtd3lleyu55fGbGDNj5CksWdeMn/3sZz871YUAaG7u/abQpwun04nb3d4kousa02cOo7bahWGEZ4eef+FYFl095bgWPZtwxiBKDjfQ4vLh8wfD57tgLJddNZnVb+6mvLSRhqPLFHyyqYQzpxSSmhY9OWRQQTplJQ3U1bYQCITQdGIOSTdDIQoKMykYnMnEMwsoKMxg357qqJmyEB4iuV/zsqesLuJ4MGTi8vqxWy1UNrgYkJ5MWZ2LQ1UNDB2QzscHKiOaMDAVBBRNOQaTJgwiFFJsOxw967IgK5UWX6Dbu5yOvIEQtc0eqpvcuLx+PP4gpgo3BXX3pRHu3I08ZjV0/KHomvnBqkYunDoa6zHj9P++dhsNLbGD1tA1zho1iMPVTTHLkpeRTFGML8uOdpfWsbMkepPxjgqzU2N+6UL4juMPr2/i2Xe28/aWg5imychumou6cuzfQyJ5+q5/s3d95IQ9X4uPpppmzr4ycqb3ybgOqak9WxtIavR9JDXNweLbz4/LuSxWg+9+fwE11S5qqpoZPDSL5BQ7v7l/JXW1kTXB2poWXnh2E4tvi35vTdP49tJ5HNhXw4fv72dT8WEqypqinmeasP7DQ8yYHe6AnDpjKEOGZ3PPj1ZQbvUTSLOgAGuLSZMlxIFOaqs7jtSw7XC4Ldyia5hKYarw5s+ZKXYKs1MpqT36Ba9rmCkGPtPk969t5Hffvpg1Ww9R0aGma9E1zhiSS3l913ut9lRvb65SHDamjszj3e1HYv7cVIpvP/IC/3PDXApz2icaGZ10VtusBrPHFvCdhdPYUVIb83OdP2lYt+X6/Dnj+XB3KWV1sa9LhtPOlWfHblowleLuv69l+5H2Pou95fXUubxcO0dGxByrpT72nZfnOOagnEwyDe04hGLU5k6GnNwUxk0cRHJKeD31FlesiTrdT3waPjIHM6Rihnyr7JzkqMf65HQ8eXaCyQahZANvroWDqSGaPZ3N7mz//6Cp2h6bSlHb7KXOFbtZwuMPUlLTxPcWzWR8YTa5aU4yUxykOGys3X64V003XfEFerfeTYvPz9ZDlZhd3E7UNLm597l38Xa4W5k0dACxst4fCPHO9iNc/+CLNMVoWhmam8b8ycMwleKDnSX87tUNrN5ykJAZ+fuX7rRz17VzmDEmnxEDMxhbkM0ZQ3KZMDiHaaPy+P5VZzNrbOzVL4v3lLG3PPJuzOMPsnbb4aj3EZA/Jvas7dyhp/e69lKj74XVK3ex6o2duJq8pKQ6mH/hOOZd0PslWuMlOycFYuw5OiCv82nrwZBJTZOb/Xs7X4xqwMBULr68faPnkGlyzz/fYV99U2R/5wmuvR+rDb5Vkt3K2IJsHvhKEe9uP8yjK9bj8sV3X1RD1zF78aWtFNR18uXaUUlNM69v2scVM8O/GzctOJOGFi/bDldT7/IS6PCeIVMRMkO0ntVuNchOTSI/K5XFF08DBT95ejU7j3bkvr5pPy99tIf/+eI8nPb20VuF2Wnc9YU5Pf4srfaU1sXsIG7xBvAFQjjtUhfs6MrbL2bXun0c2tq+HlPB2Dyu+VH0vsCnEwn6HtqyuYTnnlmPqzn8J1lb08Jzz6wnOzeZ84tid372tatvmM6BfTURO0MVDM7gc9dOjfn8Fz/cxWsb99PQ4sVeFfsWNCXVzne+dz5p6e3LE7y6YR+b9lf0aBHerBR7j8KwK0k2C2Py29uIX9u4Py4hr2vtdxm5aU5SkqwcqOzd2j49oSCiGcXQdb63aBYuj59fPPtOW3NW7DJqfG/RLMYWhGuI/3pnO1sOVrVd+2DIZHdZHU++9THfuST26p+9MWXkQF76aDfuY2ZDpzvtJB0dxWWGTD5YvoGPV21nwNBsLv7GfJIzPp27WjnTkvjxf27ltT+u5vC2UgaNGsgl35qPMy2p+xd30FjdxLO//C/Vh2tJSkvisiUXMGrqsL4pNBL0PfbmK9vbQr6Vy+XjjZe3c37RZ05JmbKyk/nxLxby0vObqapsJm9QOpd/fnJb005H2w5V8c93trc1s3jTNFIaQe/QepGSaueW781n2DGjdjbsK+92VAeAoWl8Zngeb2851OOV2c8cmsuOkhoCR3s8HVaDu66dE9Fp7e1kRU5D18hMSWJQZjJuX4B9FZ0vpwtw9ewJ+IJBbBaDy2aMZuO+Ch566aOYz9U4/u1Qk+1Wzp8UvSBZSpKN1KTof5uOPP4gVY0tbUG/7Uh1zHIcitMX1ITBuZwxdAAb9pYTOtokle60cdmM0WiaRigY4v9u/AM7PthD6GhTV/GKzXz3z18jf1TsZoxEZ3faueK7Fx3361sa3PzqC49Rurt9Ke6DnxzmW4/cyPjZo+NRxCgS9D107Pov3R0/WVLTHNxw86xun/dS8Z6ItvRAugV3QJHRrEi32UhNc1B08QTGnxG9JaHd2rNfkxGDMvn6hVM67ViMJTnJxh9vuZS3Pj5AZrKD8ycPw6WO8IH7UYLKS45lDAU5A9hZGj2q5LIZo7n23ImkJNk4UtPED/6yiiZ353cT+yrr+dl157U9njdpKP9+fyeHayL7KrJSHdyxaBb3Pv8+rk76Hzqj6xozxxYwrjD2Xd5lZ41m26Fqmr2xz5ublsTkDuPdHZ1ce1sv9vLtiqZp/Piac3h5/V427qvAbjW48uyxjCvI4cOXNvLcr1ZQfSjy2lfsr+Yf97zI9/76zbiU4dPm5d+tjAh5gIbKJl76zRsS9KfaiFG57NgavRnG8JGnptmmt4Ix2qJ9OVYyJmdx/5cXRG0P2NGVs8aw5WAVjR1CVNdg2shwjc4bCDF+aB6fnzUap93K4kum8cSbmymrbSYYMrusGac6bDhTPHzu3JFYNAelgfVs9P4FrwrXWCtCnzBu9kQOVn6GA5UNmCo8vHFsYTY3nn9m20SowTlpXDRlBG9u3k9DS+ywP7ZHwdB1Hv76hTy6opjiveWgFCMHZbHk0ukMzEjhgsnDWVG8J6JNPRarLYDFGsLTYmfaqEJuu3xGp8+dPHwg186ZwOub91Pd6MYXCLbdLdlMyK5qoeFgDWkTCqgrq0dbsYWsrSUELBreM/IJDMsi2WHl4mnxG7Nt6DqXzxjD5TPGALDt3V3c8+2/cfCTwzHXRAKoLe16nXrRufJ9sfvHXJ3MpYiHfhn0SilWLP+E4vcP4PEEyMpK5uovTmfUmL7bjeeKq6ewd3cV+/dUEwiYWK06w0flcuUXYreHn25mjytk476KqNAanZ/VZcgDjC3I4eaiybz40W4aW3wkO6zMmTCE685rH36Xk5NDzdE1XqaMyOOBr85mbdkr+KjDbBxBxZFs3vrkEN4OI10yUg3yprzKypYqDBzkWsbSFCxrC/lWbttuln5xHju2j2RXaS1TR+Rx7oTBGHpkub88/0w+O3Ukj728no8PVESMf7daQuSN3cT77o8Zb7uUTEt46KjNYnD7oth3RF+94DMMH5jBoy8Xx+ywNIwQ51y4mZyBDVgsJp6WJC4efEu3cyWumDWWS2eMpqy8npJ9Vfz3o90c2F6Gtv4Q1XUefvXfLcy7YTYbXv2E8r2V4eUfAFudh6TUJC69YSrnjO+b1R7ffvp9nv/VClydDCNsldShebCuvIH68gYKxg7Ckdx105SAwnGD2PjaJ1HH07J7v7dyT2lK9WbqSd8pK+t+HfVWr7y0heXPbmpbYREgd0AKP/3lZRFrnMebaSo2fnSIndsrGDchj6kzhqLrWkTIna6UUvzfix+yYV8FTW4fTruFMfnZ/PQL5/a4aaZ1LRu71YgK2Y7XoDFUyvvuh3Gp8IggDYMcYwwVGy/j3e0ltHgDZKRaGDV9HQOGdZx8omNgIxRjc+7h1rlMT/oqQeVDQ0NDZ79/NdWhXaQbhYy2XYhVC//bm0rx6Ipi1u8rocHtxZnsZejoCqafuxMAh5bOVMdNFFinRb1PLC99tJvXN+6nyeMj2W6l2eOn0e1jzkUbGT62PGLwkYGNLGMkw6xzGGY7N+pcPrePsr1VrPn7B2xZs4OGikZMU0WN/nEk2/HGuDMZMWUoP3r+Vqwxtm48UUop7vrsAxzeFr3/bkdJqQ6u+fHlnHfNTH73nb+wZ8NBWupbyCrIZO51s7jp7utO+7+Hk6GzXPC4vNx71W84sr098zIHZXDL725i1PThUc/vSn5+fo+e1y+D/ud3vsT+vdEX8KJLJ3L9TTPjWawe6Q9B36q0tomPD1YxYmAm4wp7N/bXZVZRF9xPUHlRmBRYp+HQw+u7dLwG77T8mopQZI1Fw2CK/UZG2s/HVIpPfM+wx/9G1HvoWDEJRL32DNvVVIa20GxWgIIgHgJ4aO0yTdcLOc/5faqDu9gXeBstpFPlK6euyU9KmgebPbIvJVsfxaykW/jY93dazGosmp3hlrkoTMpCm45+mRgElIckPYsUNYgWX5Bxqefw0ntlvPDRNq68+RUs1s6bdTL0IRQl34Omhb8UX/zN67z73EfUltR12iTSHU3XGDA0h/HnjOaGn12JzRF7OYmOlFJsfnMrH/53M2PPHsl518zE6LBAnKfZy7oXNxDwBXj5t6to6KKjN31AGld890IWfHkOf/vp86x86p2IXuuUDCe/WHEnWcOi1/3xuLxsXbMDZ7qT8bNHo+uJPXSzq1xw1bfwnwdfpWJfFckZThbd9lkKOhmj35WEDvofLf0PpSXRIyxmzh4et9movdGfgv54KGXyoedxKoJbCdDeyeogk1G2BYx3XB5xDV53/YgmM3rf10LLWZztXAJAseePHAy8E/UcO2mYBAnQPnU8Qx9G0PTiIvaG4a0sJBGk+7Vhwgw0FIrjCVydUFChG6rbqQQ2UnHq2bRsyOD1xZW4G+M3gzIpzUFOQRbO9CQcyXbm33gun1kQOZu1pdHNXRc9QE1J+6QoR6qdEWcOwbAapA9IY+e6vdQcDv9ct+iYsb6ENCgYlceAYTkEfAFGTR/Oxte3xKz9n3/9Odx0/zURx9b8cx3/XfYG1Ydrsdgt5I8ayC2Pf4WBw6O3FEwUJyMXehr0/bKNPmdASlTQW636cW3qIbq32/8aJcFi1DFb43mpZ0/gTfKtU9lR/SzlLTvRNIOgGTvMnHp7x/UI63xKAxsiAh3AoWViEsBUJjo6ucY4UvV8dpkrui1nz0MeIHTcwyfBxOjhX46fZvxmMxv+loy7sfvad294mrwcaWqvIO1ct5fCsYNwN3kYPX04n/v+Qh771lMRIQ/gbfax/b09Mc8ZK+RtDiuzrpzG5je3Ubon/GW7de0u7M7Yn0cdMxbXVd/Ciw+91taBG/QFObytlD/9v3/wo3/f2vMPLI5bnwT95s2befLJJzFNkwULFrBo0aK4nv+Gm2dRXfUmZSXhW0y73cKEM/P7RdD7VQvbvP+mySzDqjkZb7ucTMuwU12sLlUEt0SFfCufauQd94N4VPsQPB0DHQsm7c0lKdoAxtkWopRJo1mCVXMwzHoeh4Mf4OvQ+dqoDrX9fwhoMA9RHzoc/w91khUu8mEGFJVv2eirLQJ9LX72bQxfv/K9Vaz914fHPxmgg+zCLEp2lNFUE7nwoN8bPYnNkWxn4TcviDj2wQsbYo7SqTxUQ0uD+1M7+epkinvQm6bJn/70J37yk5+QnZ3ND3/4Q6ZPn05hYc92yumJgXlp3P2/l7PytR2UlzUyY9ZQzpw6+LhWhjyZgsrHmpb7aDDbw6w2tI+ZSd9kgCX22u2nA62bYPIdM0rGJISddJK1HIL4cOqZnOm4FpdZzQbv/bjMKnR0UvSBnO34Dpt9z0Rck47cqhaDntSET2SKU9/LmBjkk7ucnNR9YON0OdzNHjwxlktWporad0Gh8B0z98DmtMb859ENHb2LvQFE/MQ96Pfu3UteXh4DB4YnfcyePZvi4uK4Bj2EN/G49Moz43rOvrbH/0ZUoHlVPdt9L57WQT/EejbVod2YxJrko0XU3NsPK2Yn30qSHt5dyVQh3mz5CU1muE03BNSbB9jse4Zms7zL9w8Re/kDB5lYtSRsejIoRa25tzcf66SqKbbgq4rPJKeTraXeTWddecd2Kvta/Pzj3v/w/X8tbjs264ppPP+/K2iqjZxEl5RsJyklejltEX9xD/q6ujqys9tHc2RnZ7NnT+z2wE+bhtDBmMd96vRei3+o9VwOB9ZRGdoS9TMDGzpGVFu7XUvDrrUvrlYT3B0eMXOMZrOCUMwvkI6iQyZTH8785LvQtXB4BpSbFc239bKd/uQJ1J1Gd5vH1K6d6U5sSVY0TSPg9eOqj/y3DPZy9vfBbZHLOGuahhFj9zO/14/P7cPulLH3fS3uQR/rmz9Wk8rKlStZuXIlAPfddx85Of1jhmksFoulR+UfVD+WkvriqOMptszT/vOf6VrIm1VbOTZ07UYKA+xjOOwubqvZ2/VUJmdezoD09qn8XncqeKJ/N3QMdM1JQHW/QYOBjQxrIX6zBUO3UBx8jOmZX2SgIzyj8yLnD3m58u6oMp4OBn02yN4/hfCWn/pa/cTZY7HYLAR9QcbOHMWX7v48gaMrif73sdf52z3Pn9D5/W5/xO/zzo/20hxjrfyaknrqDjYx6bzxvTp/wBdg86qtGFYLk+dNiBgqejrpaS6cDHEP+uzsbGpr2zvmamtryczMjHpeUVERRUVFbY/78/DEng6jKlBz2K2vodFsr/E4tAxGap897T9/qhpFqp4X1cySQh7TjW8yNOcsdtevRdesjLZdSG5gbMRnsqk8UvQ8ms3IYbQp+kCceg6lwegvwGgaLYE6vDTQHKqEAFR793Nu0m1kWIbiZBj5xjTKQuvj8ZHjyp6lmHCHhx3/58BTYnBS2+qPUXWkhl+/f1fb4yZX+1o/jkwbutG75ZuPlZaTEvFvrywhklLsNB+zLLUz1QF2s1e/+1tW7+CZu/9D5YFqNEMnb3guX33wekZOiV5E7lQ7nYZXxr0nZOTIkZSXl1NVVUUwGOT9999n+vQTX041EVi1JOY672S4dR45xjgGGVM4O2kJedZJp7po3TI0G9McXyVDH4IVJ3YtjVxjHDOTvo2m6UxI+yznJt/ObOcSci3Ra/TrmoWpjhtJ0wvQMNCxkq4PZrrja8xKWsxQy7l0F34aGl4ih9V6VB3b/MvbHs92LmGoNXpG6smi0XntsuCzJje8PJ6RM09tKPk9nS/5PPPyaQwaecxSIhoxO03tybE7yW99PHKxs9wh2QydGN1HN+SMQgaN7PmG5UF/kGd+9gLl+6owTUUoEKJ0dwVP3fmvTvsQRFjca/SGYXDzzTdz7733Ypom559/PoMH9826HP2RXU9letLNp7oYxyXXMoai5F/gMiswNBtOvXczawdYJnBB8v9QG9yLphlkGyPaZo2elfR1fJ4mKoJbIcYkJitObFoKQRU9Rt9vtjcLaJrGVMeN1Ab34lJdT7A6+gpStDxGWuezJ/AabtX13qvdGeqcQbY5gS3eZ/HT3vdiJZnx9isYm/ZZpj/l5fZZP8Pd2HV/gqaB1WEjFAqhlMIwdAIxNmm3J9vRNPC5/W1j2J1pDrxuf8xx8UYXK19a7RZufeKrPHXns+HJTTYLY2YMJ+APsumNrXiP7jWQNyJck/7bT56nZFc5ZtDE6rCy4MvnMv3CM6Nqsrf84Wae/MG/OLSlBIViyIQCbr7/2i4//7F2F++n8mD0gmDVh2so3V1B4djolVdFWJ+Mo586dSpTp/aPxb5E72iaRqpx/H9QumaQa42u8WuaxjlJS9njf4Oq4PZwjd8YQmPoIJpmMMq6gCPBj9gfiN6fNkmP3MjaojmYkfQNNnufaQtur4q1Vr3OuY6lDLKF9xMYbJvB++7fUGfu7/ZzOMgghD+iE9qhZTA14xoMVzYjbHMpD3zCwcBaNAzG2C4kyxJecTIp1cGP/3MrD930BOL8VM8AABGxSURBVA2VDYSCZtQkI4DcoTnc++YP0C0GFqvBlrU7+csPn21bNtiWZGPh4gUsuu2zAGxdu5PVz3yApsH5N57Ls/e+yIGPo/e3nX1l13+beSMGcOeztxAKhtB0DV3XUUqxp3g/7z5fTO7gbBZ8+VycaUn84vXvU32kluZaF0MmFGCJ0ekKkJTiYPFjX+72unbFYrNgGNEzd3VD///t3XtUVFX7B/DvnDkMyGW4XyRvCbK8JeINsxRNWysj8hcp2g8pzQQFtATzQvn2rresfH8LcamgrIWWkEqZkJaJuuKmoYYopqCClChLbg7DrYEZ5szz+4PFSQRBDJxx2J//mHPOnuds9jxzZs+Z/YDvpWWbjdVTuQSCoTH2JRAexZPoA7WuHpmqL8RbNAHASuKC6ebrYSHt/NOFWteABl05slX/7XB3jwnM8ZLFJsilf9dTbRAqkaH6FGp6eD1dM9jiFasvoRBKcF19FGpdI0w5S4yU+WHswFk97of0pDNI2nS4QwLzmDIcH6W83+6xpsZm/Ho4Fy1NGrywwLvLFQ/TE3/FoS9/hKr+708OLm5O+PdPkRhg1Xe3NfbVWNDpdPjXK//XbjEwAHCbMAz/Orqm15/vnzKkOfqncgkEpn8y5eSYaR6Fa5ojaBAqYcE5YLTp/4gLq3V+jBVkEktYca6o1d1qt82Kc4EV1/6FYiV1xlCTF3FLkw0NOt4pwoHHMNmLMJEMgAs/Fi782A779NSMhVNx5rvfUHLp799Y2DjJ8Vr4nA77DrA0w5x3Hq027EtvvwDeVIrs5PNQ/6WGw2A7/O+/3+jTJN+XOI5DyPYg7Ik8iKrSe5BwErgMd8KKHUH6Ds3gsSv6XsCu6A2/DxqFSvzWHI9GXevUjyXnAm+zEFhIO19Uq05bhlvabBARlEIpmkgJXiKDK++FMaZvPvRX2I/bD81/qXF0+0ncunIH5lZmeC38ZQx77un8bquvxwIRoerWPXA8B8fBPfue6EkypCt6luh7gaEnuSfhaekDlU4BCSQd5vV7y9PSD32J9UErQ0r0bOqG6Vd6eqcQwxgDtqIQwzCMkWOJnmEYxsixRM8wDGPkWKJnGIYxcizRMwzDGDmDub2SYRiG6Rvsir4XbNiwQd8h6B3rg1asH1gftDGkfmCJnmEYxsixRM8wDGPkWKLvBfdXyuqvWB+0Yv3A+qCNIfUD+zKWYRjGyLEreoZhGCPHFjXroZ9++gnp6emQSCQYPHgwQkNDUVtbi23btqGxsRHPPvssVq1aBZ43rq6Ni4vDxYsXYW1tjejoaABAY2MjYmJiUF1dDUdHR6xZswaWlpYgInz11Ve4dOkSTE1NERoaiuHDh+v5DP65zvogKSkJeXl54Hkezs7OCA0NhYWFBQAgNTUV6enp4DgOS5cuxfjx4/UZfq/prB/aHD16FN988w0SEhIgl8v71VgAgOPHjyMtLQ1SqRQTJkzA4sWLARjAWCDmkSkUCgoNDSW1Wk1ERNHR0ZSRkUHR0dF05swZIiKKj4+nEydO6DPMPlFQUEAlJSUUEREhPpaUlESpqalERJSamkpJSUlERJSXl0ebN28mnU5HN27coI0bN+ol5t7WWR/k5+eTVqslotb+aOuDO3fu0Nq1a0mj0VBlZSWFh4eTIAh6ibu3ddYPRETV1dX02Wef0cqVK6muro6I+tdYuHLlCv3nP/8hjUZDRES1tbVEZBhjgU3d9JBOp4NGo4EgCNBoNLCxsUFBQQGmTp0KAJg5cyZyc3P1HGXvGz16NCwt25ety83NhY+PDwDAx8dHPO8LFy5gxowZkEgk8PDwwF9//QWlUvnEY+5tnfWBp6cnpNLWeqUeHh6oqakB0No306ZNg4mJCZycnODi4oKbN28+8Zj7Qmf9AAD79u1DYGBgu6Is/WksnDx5EvPmzYOJiQkAwNq6tfKZIYwF45pf6GN2dnbw8/PDypUrIZPJ4OnpieHDh8Pc3Fx8sdvZ2YkvdmNXV1cHW1tbAICtrS3q61vrrNbU1MDBwUHcz97eHjU1NeK+xio9PR3Tpk0D0NoHI0aMELcZ+7i4cOEC7OzsMGzYsHaP96exUF5ejuvXryM5ORkmJiYICgqCu7u7QYwFdkXfA42NjcjNzUVsbCzi4+PR3NyM/Px8fYdlcKiTG7keVnrPWKSkpEAqlWL69NZ6rp31gbFSq9VISUnBwoULO2zrT2NBp9OhsbERmzdvRlBQEGJiYkBEBjEWWKLvgStXrsDJyQlyuRw8z8Pb2xs3btyASqWCIAgAWq9g7Oz6pkydobG2thY/hiuVSsjlcgCtV233l1BTKBRGeQXXJjMzE3l5eVi9erWYxOzt7aFQKMR9jHlcVFZWoqqqCh9++CHCwsKgUCiwfv161NbW9quxYGdnB29vb0gkEri7u4PjODQ0NBjEWGCJvgccHBxQXFwMtVoNIsKVK1cwaNAgjBkzBufOnQPQ+qKfNGmSniN9MiZNmoSsrCwAQFZWFiZPniw+np3dWli7qKgI5ubmRvvizs/Px5EjR7B+/XqYmpqKj0+aNAk5OTloaWlBVVUVysvL4e7ursdI+86QIUOQkJCA2NhYxMbGwt7eHlu2bIGNjU2/GguTJ0/G1atXAbTWwNZqtbCysjKIscB+MNVD3333HXJyciCVSjFs2DCsWLECNTU1HW6vbPtCxlhs27YNhYWFaGhogLW1NQICAjB58mTExMTg3r17cHBwQEREhHh75Z49e3D58mXIZDKEhobCzc1N36fwj3XWB6mpqdBqteIXcyNGjEBwcDCA1umcjIwMcByHJUuWwMvLS5/h95rO+uGll14St4eFheGLL74Qb6/sL2NhxowZiIuLQ2lpKXieR1BQEMaOHQtA/2OBJXqGYRgjx6ZuGIZhjBxL9AzDMEaOJXqGYRgjxxI9wzCMkWOJnmEYxsixRM8AAGJjY5GcnKzvMJ4amZmZ2LRp0yPtGxYWht9//73b/QoKCrBixQrx74iICBQUFPRpbD0RFBSEysrKXm+X6Xss0TOMgdq6dSvGjBnT5T5VVVUICAgQf5ndl5KSkuDs7Nznz8P0PpbomT7zJJIPwzDdY6tX9lN//vkndu/ejfLycnh5ebVbaCovLw/Jycmorq7GoEGDsHz5cgwdOhQA8Mcff2D37t2oqKjA+PHjIZFIMHDgQCxatAgFBQXYsWMHXnnlFRw7dgzjxo3DqlWrumyvpqYGe/fuxbVr12BmZgZfX1+8+uqrXcZ+8+ZNJCQkoLy8HDKZDC+++CLeeecdVFVVITw8HMHBwTh06BCICH5+fvDz8wMAtLS0YP/+/Th79iwA4Pnnn0dgYCBMTEzE2H19fXHkyBFwHIe33noLs2bNAgA0NDQgLi4OhYWFcHV1haenpxhPQkICZDIZ3n77bfGxL7/8Es899xx8fX3bxd5VDA8KCwtDSEgIxo0b99Bz/uSTTwAAS5YsAYB2UzaJiYnIyMiAubk53nvvPfHXmCqVCvv27cOlS5cgkUgwa9YsBAQEgOM4VFRUYNeuXbh16xZ4nsfYsWOxZs0aAEBAQAC2b98OFxcXXLx4EUlJSVAoFBgwYAB8fX3x+uuvd/l/Y/Toia5+zxiElpYWWrlyJf3444/U0tJCZ8+epUWLFtHBgweppKSEli1bRkVFRSQIAmVkZFBoaChpNBrxuGPHjlFLSwudO3dOPI6I6OrVq7Rw4UJKSkoijUZDarW6y/YEQaB169bRoUOHqKWlhSoqKigsLIwuXbrUZfxRUVGUlZVFRERNTU1048YNIiKqrKykBQsWUExMDDU1NVFpaSm9++67dPnyZSIiSk5OpqioKKqtraW6ujr66KOPOsSenJxMLS0tlJeXR4GBgdTQ0EBERDExMRQdHS22GxwcTB9//DERERUXF1NwcLBYTKKuro4CAwNJqVQSEVFoaOgjxxASEiKe5/3HdXfObcVPiIgyMjJo0aJFdOrUKRIEgU6cOEHBwcGk0+mIiGjLli0UHx9PTU1NVFtbSxs2bKCTJ0+K53n48GESBIHUajVdu3ZNbHfBggVUXl5ORETLly+nwsJCIiJqaGigkpKSLv9njH6xqZt+qKioCIIgwNfXFzzPY+rUqeL6I7/88gvmzJmDESNGgOM4zJw5EzzPo7i4WDxu7ty54uqdDy7OJJFIEBAQABMTE8hksi7bKykpQX19PebPny+W4ps9ezZycnK6jJ/neVRUVKC+vh5mZmbw8PBot33BggUwMzPDkCFDMGvWLPz6668AgDNnzuDNN9+EtbU15HI55s+fj9OnT4vHSaVSMZYJEybAzMwMd+/ehU6nw/nz57Fw4UKx3baCKwDg7u4Oc3NzcUGrnJwcjBkzBjY2Nh1i7y6Gxz3nBzk4OGDOnDngOA4+Pj5QKpWoq6tDbW0t8vPzsWTJEpiZmcHa2hq+vr5in/M8j+rqaiiVSshkMowcObLT9qVSKcrKyqBSqWBpaWkU5QGNGZu66YeUSiXs7OzaTde0FYe4d+8esrKykJaWJm7TarWoqamBRCLpcJy9vX27tuVyOWQymfh3V+1xHAelUilOOwCta3qPGjWqy/hXrFiBb7/9FmvWrIGTkxPmz5+PiRMndhqTg4MDbt++DaB1msjR0VHc5ujo2K4AhJWVlVhABgBMTU3R3NyM+vp6CILQrl1HR0dcu3ZN/NvHxwfZ2dkYN24cTp8+jblz53Yae3cxPO45P+j+N5m2VTWbm5vR2NgIQRDEhdeA1jXj285t8eLFSE5ORlRUFCwsLPDaa6+1W7CsTWRkJFJSUnDgwAEMGTIEgYGB3b75MPrDEn0/ZGtri5qaGhCRmLQVCgVcXFxgb28Pf39/+Pv7dziusLDwoce1ebCoRFftFRUVwcnJCdu3b+9R/AMHDsQHH3wAnU6H3377DVu3bsWePXvE7QqFAs888wyA1jeatmVx7ezsUF1djcGDB4vbHmVdcLlcDqlU2qHd+02fPh2RkZG4desWysrKMGXKlE7betwYHnbOPS3iYW9vD57nsWfPnnZvam1sbGzEWzyvX7+OTz/9FKNHj273PwZaP8WsW7cOWq0WaWlpiImJwa5du3oUC/PksKmbfsjDwwMcx+H48eMQBAHnz58Xa1jOnj0bp06dQnFxMYgIzc3NuHjxIpqamsTj0tLSIAgCcnNzu6192VV77u7uGDBgAH744QdoNBrodDrcvn272zazs7NRX18PjuNgbm4OAOC4v4fy4cOHoVarcefOHWRmZorl/V544QWkpKSgvr4e9fX1+P7778WKUF3hOA5TpkzBoUOHoFarUVZWJq7D38be3h5ubm7YuXMnvL29232qud/jxvCwc5bL5ZBIJI98f7utrS08PT2RmJgIlUoFnU6HiooKFBYWAgDOnj0rFsmwsLAQn+d+Wq0Wp0+fhkqlAs/zMDc377APY1jYFX0/xPM81q5di/j4eCQnJ8PLy0u8AnVzc0NISAj27t0r3uExcuRIjBo1Sjxu9+7dOHDgALy8vDBx4kTw/MOHUVftcRyH9evXIzExEWFhYdBqtXB1de20JN398vPzkZiYCLVaDUdHR7z//vvtEuvo0aOxevVq6HQ6+Pn5iXfI+Pv7Q6VSYe3atQCAqVOndvpJozPLli1DXFwcgoOD4erqipkzZ3b4MZOPjw927tzZbirqQY8bQ1fn7O/vj02bNkEQBERFRXXbVnh4OPbv34+IiAg0NTXB2dkZ8+bNAwCUlJTg66+/hkqlgo2NDZYuXQonJ6cObWRnZ2Pv3r3Q6XRwdXXFqlWrun1eRn/YevTMPxIVFYWXX35ZvA1Rn9purzx48GCn0xJ9rbCwEDt27EBsbCy7wmUMChuNTI8UFhaitrYWgiAgMzMTpaWlGD9+vL7D0jutVouff/4Zs2fPZkmeMThs6obpkbt37yImJgbNzc1wdnZGZGRkn9QA/fzzz9vd1dLmjTfeeOTplielrKwMGzduxNChQ7v9sRfD6AObumEYhjFy7DMmwzCMkWOJnmEYxsixRM8wDGPkWKJnGIYxcizRMwzDGDmW6BmGYYzc/wNMv/7R9fbN4AAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# KMeans Clustering\n",
    "data2 = data.loc[:,['degree_spondylolisthesis','pelvic_radius']]\n",
    "from sklearn.cluster import KMeans\n",
    "kmeans = KMeans(n_clusters = 7)\n",
    "kmeans.fit(data2)\n",
    "labels = kmeans.predict(data2)\n",
    "plt.scatter(data['pelvic_radius'],data['degree_spondylolisthesis'],c = labels)\n",
    "plt.xlabel('pelvic_radius')\n",
    "plt.xlabel('degree_spondylolisthesis')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "_cell_guid": "fc405787-c073-4377-a4d8-0dcb83371888",
    "_uuid": "bb760dd8801ea5d0b7d98d1e22f784a9c999b900"
   },
   "source": [
    "<a id=\"12\"></a> <br>\n",
    "### EVALUATING OF CLUSTERING\n",
    "We cluster data in two groups. Okey well is that correct clustering? In order to evaluate clustering we will use cross tabulation table.\n",
    "* There are 7 clusters that are *0*, *1* , . . . *6*\n",
    "* First class *0* includes 138 abnormal and 100 normal patients\n",
    "* Second class *1* includes 72 abnormal and 0 normal patiens\n",
    "*The majority of two clusters are abnormal patients."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 283,
   "metadata": {
    "_cell_guid": "df624885-2e0e-474f-82cb-8a498049d8cc",
    "_uuid": "063968754033b4510dcb6f90245a1a9688d4c201"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "class   Abnormal  Normal\n",
      "labels                  \n",
      "0             19      54\n",
      "1             25       0\n",
      "2             59       2\n",
      "3              1       0\n",
      "4             14       0\n",
      "5             48      44\n",
      "6             44       0\n"
     ]
    }
   ],
   "source": [
    "# cross tabulation table\n",
    "df = pd.DataFrame({'labels':labels,\"class\":data['class']})\n",
    "ct = pd.crosstab(df['labels'],df['class'])\n",
    "print(ct)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "_cell_guid": "f5c33609-4573-443a-82c5-5dfe9852c582",
    "_uuid": "03df9051e892dc064c706619e796b5c9128c0a68"
   },
   "source": [
    "The new question is that we know how many class data includes, but what if number of class is unknow in data. This is kind of like hyperparameter in KNN or regressions. \n",
    "* inertia: how spread out the clusters are distance from each sample\n",
    "* lower inertia means more clusters\n",
    "* What is the best number of clusters ?\n",
    "    *There are low inertia and not too many cluster trade off so we can choose elbow"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 293,
   "metadata": {
    "_cell_guid": "710c918b-481b-4f32-94f6-adecd992a97e",
    "_uuid": "6c7e94ea35a4f535857aa2551ed646ff8c28c1ac"
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# inertia\n",
    "inertia_list = np.empty(8)\n",
    "for i in range(1,8):\n",
    "    kmeans = KMeans(n_clusters=i)\n",
    "    kmeans.fit(data2)\n",
    "    inertia_list[i] = kmeans.inertia_\n",
    "plt.plot(range(0,8),inertia_list,\"-o\")\n",
    "plt.xlabel('Number of cluster')\n",
    "plt.ylabel('Inertia')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "_cell_guid": "6449fdb8-7e0b-4432-9ea3-6d866d899704",
    "_uuid": "06031b6c13b5f700bb2e3d53081d990f66c5e11b"
   },
   "source": [
    "<a id=\"13\"></a> <br>\n",
    "### STANDARDIZATION\n",
    "* Standardizaton is important for both supervised and unsupervised learning\n",
    "* Do not forget standardization as pre-processing\n",
    "* As we already have visualized data so you got the idea. Now we can use all features for clustering.\n",
    "* We can use pipeline like supervised learning."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 295,
   "metadata": {
    "_cell_guid": "0ae27970-8319-45a7-ba7c-a8dd9d9b427e",
    "_uuid": "97a7e3a3fa1722970ff34935470be85f2ec7977b"
   },
   "outputs": [],
   "source": [
    "data = pd.read_csv('/Users/garvitkhurana/Desktop/Projects/Data_Science/column_2C_weka.csv')\n",
    "data3 = data.drop('class',axis = 1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 302,
   "metadata": {
    "_cell_guid": "65a46d3b-45b5-4bcd-8137-bdd5b6507ab0",
    "_uuid": "6883975e95325db10ebed6703c372cbe4da29a48"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "class   Abnormal  Normal\n",
      "labels                  \n",
      "0             97      90\n",
      "1            113      10\n"
     ]
    }
   ],
   "source": [
    "from sklearn.preprocessing import StandardScaler\n",
    "from sklearn.pipeline import make_pipeline\n",
    "scalar = StandardScaler()\n",
    "kmeans = KMeans(n_clusters = 2)\n",
    "\n",
    "pipe = make_pipeline(scalar,kmeans)\n",
    "pipe.fit(data3)\n",
    "labels = pipe.predict(data3)\n",
    "df = pd.DataFrame({'labels':labels,\"class\":data['class']})\n",
    "ct = pd.crosstab(df['labels'],df['class'])\n",
    "print(ct)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "_cell_guid": "15b4f98e-0c9e-49de-b69f-f583524c19ac",
    "_uuid": "4166039422a3ed14b4ac779da29401378dfb4dc2"
   },
   "source": [
    "<a id=\"15\"></a> <br>\n",
    "### T - Distributed Stochastic Neighbor Embedding (T - SNE)\n",
    " * learning rate: 50-200 in normal\n",
    " * fit_transform: it is both fit and transform. t-sne has only have fit_transform\n",
    " * Varieties have same position relative to one another"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 316,
   "metadata": {
    "_cell_guid": "b4ddf817-57ec-45eb-aa03-1ad385cd4f2e",
    "_uuid": "311a16a96b2ef3e9ca3f8e91db48f28837577291"
   },
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAY0AAAENCAYAAADzFzkJAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADl0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uIDIuMi4yLCBodHRwOi8vbWF0cGxvdGxpYi5vcmcvhp/UCwAAIABJREFUeJzs3Xd4U2X7wPHvGVlt05myEZkioCiCooIgP1CcoO/rfF04ETdOVEBUEEWmILw4UHlx4KCiogwRUHGgIiplCCJ7tWmbrowzfn+kDUmTlCAtFHg+1+V1kZMzntPW3DnPuG/JNE0TQRAEQUiAfLgbIAiCIBw5RNAQBEEQEiaChiAIgpAwETQEQRCEhImgIQiCICRMBA1BEAQhYSJoCIIgCAkTQUMQBEFImAgagiAIQsJE0BAEQRASph7uBtSGHTt2HO4mHBCXy0VeXt7hbsYhJe752CDu+cjRqFGjhPYTTxqCIAhCwkTQEARBEBImgoYgCIKQMBE0BEEQhISJoCEIgiAk7KicPSUIxzzTxPnii9gWLkTyetGbNKHo2WfRW7Q43C0TjnAiaAjCEUDNzSV11CiUPXswnE5KbrsNX9++SMXFSB4PRqNGIEmh/Z2jR5P8yivIPh8Alo0bUW6+mbzPP8d0OA7XbQhHARE0BKGOUDZuJOmddzCysym79lpMpxMAeccOMm+5BXXLltC+6vr1aNOno2zfjlRejt6gAcWPPIKvd28A7AsXhgJG6JgNG3DMmkXZrbceupsSjjoiaAhCHeB87jmS3n4bxe3GBJJnzsQ9aRJap044J06MCBgAituN/MMPVD5bKPn5pD3xBHmnnoqRmYlUVhZ1Dck0UTdtqv2bEY5qYiBcEA4zZcsWkt57D8XtBkAC1E2bSBs+HEwTec+emMdJVV6r27aR/PrrIEnoMVb3Gk4n5f361XDrhWONeNIQhEPM/sknJL/6KnJhIYbLhda0KcrevVH7WVetol63bhgOBybRQSIWqagIgKInnyTz7rtRN28GwEhKwtu7N4EuXWrwToRjkQgagnAIWZcsIe2JJ1Dy84MbNmxAzc3FlGUkw4jYV9J11L//BsC02TB9vlDgiBVE9MxMSm+8EQCtUyfyPv2U5P/+F2XnTsr+9S/855wTMVguCP+ECBqCcKjs2kX6o4/uCxgVFI8HIyUFqaQk/rGBQESQkADDYgFJQvb70Ro0oOyaa9Bbtw7tY2RmUjxkSM3eg3DME0FDEA6UaaJu2ACmida6dULf3i0rVmC5/36kbdvi7qNnZICmoRQXR79Z5SkEQDIMih5/HJKT8Z5/Pka9egd0G4LwT4iBcEE4AMqGDbguugjXpZcG/7vgAtQ1a/Z7XOpzzyFVdDXFIpeUoBQUgM2GlpUV48JK1CbTZsP2448YTidGrGMEoRaIoCEIiTJNMu6+G+uqVcgeD3JxMdbffyf9rrtiPgmEH6fs3Bn7rSqvlbw8TIsFMyxImLKM7nJhynLEcXJZGY7580l/4AEyb7gBdP0gbk4QEiOChiAkSF27NuY6B8u6dSTNmhX/QEnCSEmJ2hxvRpRSVIQUFgAkw0AKBCi98kr8nTqhp6ZGHCf7/diWL8eek5P4zQjCPySChiDEYprR39x1Pea3eQlIfvXV4DFxePv2xYwx9hHziBhPLYrbjeJ2kz9zJkbDhtFt8PuxL1oU9/qCUFNE0BCEcJpG6pAhZPfoQb2zziLrX/8KjVlo7dphpKbGPEwuKkKuWJwXS+Dkk0GO/N9NArBYIi9fvz5GZmbU8SbgWLCA7PPPjzvLSm/cOP59CUINEUFDEMKkDR1K8qxZWDZuRN22Ddv335MxcGAwLYcsUzpgQMynA9PhwExKinteZceOiC6nSkZKCqVXXYWvWzfKL7oI91tv4evRI+qppPKVum1bcFFgcnLE+4FmzSi9884DvV1BOGBiyq0gVDIMrN9+G/Xhrm7ciOPddym7+WZKBw7EkZODde3a0PumJOE/7bRqs8cGWrfGSEpCrpITSm/cmKKxYyOm7RY99xwYBrYff0TZujWqPXJpKd6zz0YuLUUqKUFv2BDP0KFiBpVwSNSJoOH3+xk+fDiapqHrOl27duXKK69kz549TJgwgZKSEpo3b84999yDqtaJJguHiLJ9O5gmepMmtX8xTUPyeqM2S6aJfcEC1I0bKb/4Ygpef520Rx9F3bQJ02rF37kzRaNHxz1tyvjxJM+YERUwtKZN8Tz2WPQ6D6uVovHjwecj+9xzsVSkAgnnP/10Sh566J/dpyAchDrxCWyxWBg+fDh2ux1N0xg2bBinnHIKn376KRdddBFnn30206dPZ/HixZx33nmHu7nCIaBs3kz6PfeE0mhozZpROHFi7RYRslrRmzRB3b49YrMpy9i++Qb711/j+OADfD174n7nneCAtSxHDVzLe/aQ8vLLyLt34zv7bJJmzYpaBa67XOR98glGdnb89thsBE49NSpoaI0bU3bTTQd1q4LwT9WJMQ1JkrDb7QDouo6u60iSxOrVq+natSsAPXv2ZMWKFYezmcKhYppkDBqE7eefUfLzUfLzsf3yCxmDBlU7Q6kmFD39NIEWLULjFqYkBae8VlxXKSnBvmgRtvnzkYqLybjlFup17069bt3IvPFGLEuW4Lr0UlJeeYWkuXNJe/xx1FhrNPx+7HPmYP3uu2rvqWjMGMrPOw+9fn2M1FQCbdtS9NRTGC5XLdy9IOxfnXjSADAMg0cffZRdu3Zx/vnnU79+fZKSklAqFjllZmbirmZ2inBkkvPzSXr9dWSPh9IbbkBv3Rp13TrUjRuj9lU3bkRdvRqtQ4daa4/WoQMld95J2ogRSCUloWAR0WavF8fcuSS//jr2b7/d174tW7D+8ANyWBoQWddjrseQS0tJHzECw+FAa9MG94wZGPXrR13LTEqiYMYM5Lw85MJCtObNY64OF4RDpc4EDVmWGTNmDKWlpbz44otsr9JFUJ1FixaxqGKO+ujRo3EdYd/CVFU94tp8sFRVJfunn1DvvRepovslee5cjDvuwLj4YiRNizpG0nUyHA7M2vxZ6TrqW28hV5c8ELClpCB/803U9ljTYasGDBNCg9tyeXkwBfqQIWhz58a/YKx71rRgAKnDmWuP1b/to/me60zQqJScnEy7du34888/KSsrQ9d1FEXB7XaTGWP+OkDv3r3pXVHmEiAvL+9QNbdGuFyuI67NB8uVlQVPPBEKGABSXh7y6NEEliyBRo2wVHna0Jo1Y2+zZlAbPyvTJPmVV3B89BFSbm71uwLGsmVIRUXRK7pjPJmEP2nEWwVu5OaSt2dP1FqOWKzff4/zuedQdu3CTErCe955FMcaUK8Djsm/7SP0nhvFKNwVS50Y0/B4PJSWlgLBmVS///47jRs3pn379nz//fcALFmyhM6dOx/OZgo1ads21L/+itos6Xqwy0fTCLRsiakomIpCoFUrip55Bmpp9lzKhAk4x4zB+vvvMddThIcCCbBs3Rqzm0jPyMCosl6jakrzWExVTehDX96zh/T778f200+o27ZhWb+e5FdeIWX8+P0eKwg1oU48aRQUFDBlyhQMw8A0Tc4880xOO+00mjRpwoQJE3j33Xdp3rw5vXr1OtxNFWpKfn7M6a2V1B07KHjhBcyMDDBNfD16gM1WO20xTRyffBI1JXa/rFa09HTU3bsB0Bo2pPjBB5E9Hhxz5yIXFATXWVSXzLBCoGPHhIJG8tSpqFu3RmyTfT7sCxZQMnjwgbVfEP6BOhE0mjVrxgsvvBC1vX79+jz33HOHoUVCbVPeeKPaD1MpEEDdupWSK6+s/cZoWvzUHBkZaM2bY/3116iptUZyMvnvvovjk0/ANCm77rrQYHbpHXcg79pF9gUXoFSp8W04HOgNG6Lk52MkJRE46SSKnnsO9Y8/MNPTq12TEqssLBAKwLYvvyTpvfcwVZXSW24hcNppCf8YBCERdSJoCEJVemYm5ZdeWjMnKy8Hq3Vfd5JpkjJ+PPaFC5F8PrTmzYP5nqpMvtBTUvCefz7WH36IeVqtdWv01q3jfsM3GjRAy85G3rMnolvKf+qpuN97D+XvvzHT07F++y2ufv1Qtm3DdDgItGtHwfTpmDHyXJVfdhn2L75ALi+PbMtxx5E6dChJ774bemKyLV1Kyb33UnrHHQn+oARh/+rEmIZw7NHvuQc9TqU5w+mk/OKLI0qX/hPqypVk9etHvXPOIbtnT9IefRQ0DeeoUaRMnoz1t9+wrFuH44svkNxutLApr4bFgta6NfZPP8WyaVPoqcgkmC/Ke/bZFE6ZUu31bZ9/jrp9e0TAMOx2PE8+CbIcXKhomqQ+8wyW9euRy8pQ8vOxf/016XECka9Xr2CVPqcz2B5FIdC2LSUPPIB93ryILjalsDCYsr2abkBBOFDiSUM4PFq2pPi++0h+7TXUrVsxkpMxUlLwn346ZddfT+D00w/q9JLHQ+bdd4dWlAOomzcHV3d/9x2yzxexv7p9O0ZmJv4TTkBv0YLyCy4gKScHpUq3lQT42rXDPXv2ftuQPGMGSmFhxDbZ6yXltdconDQJgJTnnotagQ5gyc0NpmGvOtguSRROmYJlxQocn3yC1rIlZVddhX3hQtRdu6LOo+zejbp5M9oJJ+y3vYKQCBE0hMOm7KabKLvqKix//IFRrx56s2b7PUbesQN161YCbdoEB8njSJo5MyJgQHBmlu3bb4MZa6uQCNaskEpL8V56KVrLllhWrox5bklRUH/7jdTnn0feuxczNZWSgQPxhU37BuIOrMsVgSR16FCSPvww9g2YZrUrxQNduhDo0iX0WmvVCiMtDbmoKGI/Iz0dPcaiQUH4p0TQEA4vhyPiwy8uTSP97rux/fADcl4eeqNGlF9yCcVPPhlzd6ViRlNVkt8fHKyOU35V9vmwf/ppMF9UQUHU+6YsE2jblszbbkPdti20Xd2wgYKJE/H36LGvyccfj7VK4DEBf+fOKJs24fj4Y2S/P/bttm59QNOLtRNPxH/KKdiWLt23JkRV8Z15JmZ6esLnEYT9EWMawhHB+eKLOObNQ9mzB8kwULdtI+mtt7AtXBhz/7Krr0ZPS4varh93HJ6HHkKrZiGTun076o4dUdtNVcXbqxdScXFEwIDgrKaUadMitnmGD8d/4omh2himxRLMTnv77dg//zwqiSEEc135Onak8B+su3C/9hql11+Pv2NH/KeeSsmgQcG065X8fpSNG+POFBOERIgnDaFuKS/H+vvv6FlZ6C1bhjZbly+PWnSnlJaS9O67+Pr0iTqNunEjqGow4aBpYhLswikcNQq9VSvycnJIe+QR7MuWRU/9rTIzqVLgxBMpmDaNzNtvj/l+1dQjRnY2+XPnkvTmm1hWr8Zy/vnk9+0LFgv68cdjWCzIgUDEMXrDhuR/+CFUU5sjLocDT5wU7cnTpwefnvbuxUhLw9ezJ0WjRtXJVeRC3SaChlBnOGbNImXaNNTNmzGcTrR27XC/+ipmjCeG6igbNpA2bFjkN3lVpaxfP/RWrQAwGjemYOZMMq++GuuKFch+fyjFR9UP8tB5t26lXvfuEOd9PcbTi5mUFKqo53K5QilQvOefj9a2Ldbff9+3r6ri69nznwWMalh+/JGUiRNDg/JyURHyu++iN2xIyb331ui1hKOf6J4S6gR5xw6c48Zh+esvJF1HKSzEtnw56RWFhvQGDaLKrJqyjP/EE3E+8wz2uXODs42AlClTohbUSZqGffHiiG1pDz+M9ddfQ+MK1X3nNmUZpbAQdedO1Ly8YHqTsPcDrVpRNHRo4jesKLjffJPyPn0ItGqF/8QTKb3ppmqLOf1TKa+9Fj2Ly+/H9uWXNX4t4egnnjSEOiF5xoyYU0YtublgmsGxjKpvGgbOKVOQ/X4MqxVt+nTyZ81CrshjVpVcWkrGjTcGp96aZrDe9n7WMARatEAqK4tqm6TrBFq2RGvTBu344ykZNAizMqFmeTnK7t3oDRtWm/rEqF+fgjfeqPb6NSLOk1GsHFuCsD8iaAh1w3761qtOJYXgk4FU8ZQg+/1YV64k9emnKe/bF9v8+chV0qvLO3fiWL8+4SbpKSn4TzmFpDgpy40GDSh49dV9G0wT56hROD7/HKmgAMPlovzyyym5776Er1kbfF27Yp8/PyLomkAgbMxIEBIluqeEOqF0wAC0Bg2itgfatQNJwsjKSug81p9/xjlxIpKmhbqPDFVFa9Ik7hNILKYsEzj5ZOxffx2ztocJ+Dt1itjmmDWL5DfeQN20CaWwEMuGDaRMnYpt/vyEr1sbbD/9FPWUJkF0wkjDQCosDHXzCUIsImgIdYJRkSE20KIFpqqiZ2QEU3VUTBktfuCBmEGlKmX7diwbNgQ/FCvPnZVFee/e+802awJ6ejreHj0ofvRRvOeeGzNBYOVai+L774/Y7pg7N2pBn1xcTNLbb++33bVJjlPxMnwdStLrr+M67zzq9epFdp8+JL/88qFqnnCEEd1TQp1Rfu21lF9+OZbff8dwudCbNw+95z/7bNwzZ+KcNAmpqAi9Xj3sX30VMUNKT0qKGRiUvXuDU1xTU5E9nrjXlwAzLY2CV17BTE7G/tlnmFZrqAuskpGVhfudd6Cirn3o+Djf0CVNC2bS/eIL7Lt24evVC7NKzY3aFGtWF4DWuDEAtiVLSH3xxVAXoLJ7NykvvYTetCneSy45ZO0UjgwiaAh1i90ed4W41q4dBWEL6Hw5OaT897/IeXmYqamU9+xJ8uzZUQn6TIuFwMkn4+3dO5ghtuJpIFYVPWXLFiwrV+Lv1i04LbZlSyxr1uw7lyTh69EDqayM9PvuQ924EdNqxdezJ/4uXbD++GNE4DItFgLt25N93nmof/1FhqahN2uG5+GH8fbvf3A/qwR5HnsM68qVEUWvAi1bBqv9AUkzZkSNGSkeD0nvvCOChhBFBA3hiOXt3x9vv37g8wVnKUkStl9+Qfnxx4j9tJYtCXTpQuHpp5P6+OMkvf02sqbFnGJrpqaGamKgqrhnzCDtkUdQN23CtFrxn346Rc88g+uyyyLWWFjWraP0mmvw9uqFbcUK5KIi9MxMfGedhX3RIizr1oX2Vf/+m9QXXsDXpw9mcnJt/GgiGI0akffBBzjHjkXZvh29SROKH3wQoyLLsFQleWOlqk9YggAiaAhHOkmK6CYqmDaNjLvuCq4INwy0446jcNy4UO1t68qVUbOqwvk7dEBr3Rpl82aSZs3CSE2lYOrUYG0LSQJJwj5vHuratZHN8Puxff01e7/8EvXPP7H89ht6djapo0ejhgWMSsrmzdgWLMB72WWJ36vfH6wBUlqK94ILMCvSoyfCqF+fohiFzgD8Xbti++YbpLAEiSbBuh+CUJUIGsJRxahfn/wPPkDeswc0DaNKf368MQ3D6cR77rkUPf88yRMnkjJjRmgQPHnWLArHjMHfrRsA6vr1MVeNy6WlSKWlaO3bo7Vrh+uCC7CGdW1FUNUDWvmtrlpFxn33oVYsftTGj6d48GDKr7gi4XPEUzJoENYVK7CsWIFSWoqRnEzglFMofvDBgz63cPQRs6eEo5JRr15UwADQs7Oj97XbKRw7lsKpU5F8PpJnzoyYNaVu2ULqs8+GUpV7L7wQPUZadr1+/VBGWXXtWtRNm+K2T2vRAm+iNe9Nk/QhQ7D8+WdosF3dsgXn+PE1k3zQasX9v//hfvNNPA8+iPvVV8l/772ogX5BABE0hGNM8eDBaA0bhl6bioK/Sxe8F1wAgP2LL1BjpE1XduxArtiutWmDt29fjLAZUFrDhhTfd9++RYqGEVVTHPYtqiuYMCFYghZQf/uNtMGDSR0yBKVKDZDKaytVsurCvi4uTBPbkiWk33UXqY8/jhKjqFM8cn4+9o8+wvLTTwS6dqVk8GD855wjEhkKcYnuKeGY4u/Rg/z33gtN3fV160bZjTeGxjz0Bg1iZp817XbMlJTQ66IxYyi/5BKSZs/GSEuj9K670CumsEKwvoV2/PFYc3MjzqO1aMHexYtDtTJSxowh+Y03QrmhHF98geexxyi/6qp917bZYtfWsFgwU1JIe/jhYG2Oillh9gULKBo1Ct9551X7s0gZO5akd95B3bkTIykJ7cQTyX/jjX3pUAQhBvGkIRxz9JYtKZw4kYI33qDs1lvBYgm95+vVC61Nm4j9TSBw6qnBwfBKkoS/Rw8Kp0zBM2pURMAAQJYpfPFFAm3bYlitGBYLRrt2wUH5igAg5+eTNHt2RDJBZc+eYF2OsKBluFwEqrQJguss9CZNIqYRA6g7d+IcP77ayn9qbm4w31fF05NcVob1559Jf/TR+D84QUAEDUGIpCgUvP463u7d0Zo0QWvenPLLLw92Jx0grWNH9i5YQP777+OePRvtp58i1qBYly+PWexJ2bkzajyk4OWX8XbrhmG1htKjyHl5pN91V8wKg8ru3Ugx8nVVSp4xI+ZxaryBe0GoILqnBKEKvUkT3O++C5oW7LaSD+K7laIQ6Nw59O+I6zRtipGSEl28yenEcLn2bfD7SXn5ZSzr10eUh1VKSpA3bsRUlKjV6HJ+Ps6xY/E8/XTM8Ql5y5aYzY2XckQQKtWJoJGXl8eUKVMoLCxEkiR69+7NhRdeSElJCePHj2fv3r1kZ2fzwAMPkBLWrywIteoAanT/E4GOHQm0a4ctbDGiKUkEOnXCCBtXyLj9duyLFkWso6gk6Tq604lSXBy5XdOC60yysiipkiMLQIo30F3L9ywc+epE95SiKFx//fWMHz+ekSNHMn/+fLZt20ZOTg4nnXQSkyZN4qSTTiInJ+dwN1UQao4k4X7jDcouuYRAmzYE2ral7D//oWDSpNAu6vr1WFesiBkwKvlPPz2URyqc7POR/MYbwSemKowYU4YBtLB8X4IQS50IGhkZGbRo0QIAh8NB48aNcbvdrFixgh49egDQo0cPVqxYcTibKQg1zkxLo3DaNPZ+9RV7v/ySouefjyjcpObmRlXdC6dnZFBy990ETjkl5vvy3r2kVVQ/DFdyxx0YYRMAIPiU4634/00Q4qlzz6J79uxh06ZNtGrViqKiIjIqvhFlZGTgibOad9GiRSxatAiA0aNHB2sxH0FUVT3i2nywxD0n6LzzMOrXR969O2KzKcuYJ5yAefvtpF14IZLPh7loUVQeKQlwrFiBxWqFsNlfUkFB1NOLZJo4163DUYO/F/F7PvrUqaDh9XoZO3YsN910E0kHkDq6d+/e9O7dO/Q6Ly+vNppXa1wu1xHX5oMl7jlBSUmknXsujjlzkCsCguF0UnrddcEstaoKeXnQrRtZJ52E7aefok5hFhZSsH49+vHHh7ZlvPcelhjdVvqGDTX6exG/5yNHozgp9KuqM0FD0zTGjh1L9+7dOeOMMwBIS0ujoKCAjIwMCgoKSA2fJy8Ix4iiF1/E17Urjk8/BUWh7Prr8Z17buROkkTBtGlk9+kTNZXWqF8fvUmTiG16nG/CZgKpQ6SCAqSysmCaFrFy/JhTJ8Y0TNNk2rRpNG7cmIsvvji0vXPnzixduhSApUuX0iVOnQVBOKpJEt4rrqDgzTcpeP316IBRwWjYkPJ//xsjLPutXq8eJXfcETUryt+5M2aVD3wT0KtZDS4VF5N5441k9+lD9gUX4LrwQixinPGYUyeeNNatW8eyZcs47rjjePjhhwG45ppr6N+/P+PHj2fx4sW4XC4GDx58mFsqCHWb56mnKD//fJJnzcK02ym58070li2j9nN88kn0mAYgl5Sg/vYbzqlTkcrK8PboQdkNN4Cqkn7ffdgrxg4BlPx80gcPZu+CBQeUsVc4skmmWc1cviPUjhirbOuyI7UP9GCIe64hpknKxInBOhs+H1qLFhSNGhW5ODBsX8t332FftAjbN99gXb06apdA8+bIxcUoFe00FQVvz54UTp1K9rnnolZJhmjKMoVjxlB+9dUxmyd+z0eOI25MQxCEA+ccNYrk114LDZJb1qxB2bKFvE8/jeyS0nUybr0V67ffopSWBmdfSVJU4aXwgAHBxYPW77/Hsnw5xKiBLhkGUpWFhcLRrU6MaQjCkUzOz0fNzY2qTV7rdB37ggWhgFHJsm4d9k8/jdjmmD0b+5dfopSWAhUf9qaJUbEmRE9Px3fuuZgxVoQrpaXYVqxAP+64qPe0Ro3wXn55Td2RcAQQQUMQYtH1arPEAhAIkH7nnbjOPx9Xv35kn3ceya+9dmjaB0heL1JFEIjY7vdjqVKONmXatKj8VADa8cfjnjKF/A8/xP3WW6EiUuEMq5VAx47B6oVt22JWLArUmjSh5K67MLKyauiOhCOB6J4SjinKxo1Yf/yRQMeOaO3aRb2v/vEHacOHo2zbhulw4OveHc9TT0UlGwRIffppHJ9+ilRRbEneuJGUiRPxnXlmzHPXNDMpCaNePahSNMpIS6O8oqgUgFRUhBJnnM90OvH27x96XdavHylbt4aeSAC09u3x9u0LikLe/PnYFi9Gdrvx9u0bM8gIRzcRNIRjg2GQfs892JYtQ3G7MdLS8HfpgvuVV0IV9KSSEjLuvBPLX3+FDlM3bQJZxjNiRNQprT/+GAoYlZT8fJKnT6dowgSk4mJsS5cGr3X22TV/T5JE8QMPkDZkSKguhmG34+3VC6NpU2xffIHerFmw6mBYvY1KJuCvWBNVqfTeezFcLpI++gjJ6yXQti2eYcP2BU1V3W9xJ+HoJoKGcExIeustHPPmIVWkFpeLirAtXoxz3LjgyuqKfdSwgAHBbLG2ZctinzRGOdfKYxxvv41z8mSUzZsxbTa0Vq3ggw8iUnnUBF+fPuS3bUvKlClIRUWU9+uHdeVKXOefj7pjB0ZaGoEWLTCcTuQqA9ZmcjKlt9wSdc7ya6+l/Npra7SdwtFDjGkIRzRl61ZSJk3C/sEHEFZroir7woWhgFFJMgzsc+eGPvyV7duJtb5ZKi+PGSACHTpEbTNSUynr3x/nhAmomzcH1z74fFhXr0a97bYDurdE6U2bUjR6NIVTp2JkZZE8c2aouJNcVIRt5UqMKiu9TcDXrRtG/fq10ibh6CWChnDEco4ejeuSS0h9/nkyBg8mu29flI0bY+4ba1YQgLp5M+kFliBYAAAgAElEQVSDBgFQdtVVGGlpUftIpaXYFi6M2u4ZORJv9+6hY7TGjSm94QbUv/+OWs8AwKZNSHGSbtaU5JkzkWNU7DMzMijr3z+Ygv2EEyi75hoKXn45+GZF8Ey/806cI0Yg791b7TWUrVvJvOEGsnv1wnXhhSRPmbL/SQPCUeMfd0/t3r0bWZbJzs6uyfYIQkKUDRtImjULpaLSnKTrWNatI33IEPJnz47av+y667B9/31UlTwJsH3zDeq6dWgnn0z5pZfimDsXuagIs+J9xe0m4557KLv66mAlvAqmw4GvRw9ktxspMxNf164U338/SXPmRK2BCDZaiTmgXmN0HSnOojLTYqFwypQYb5hk3HILtiVLQlUBHV98gfvll9FOPTVqd6m8nMwbbsCyfn1om7p+PVIgELPYk3D0SfhJY8KECaxbtw6Ar776isGDBzN48GAWL15ca40ThHiS3n47FDDCKZs3xyw65OvTh9L//Ccq3xKAUlCAdflyAIpGjyb/zTfRnc6Iriq5tBT7558jh30opw4fjnPMGKyrV2PZtInkd94hc8AAyvv3RwvLKBvSoQNmcvIB32silI0bcfXti335cqp+5zdlGf/pp8c8zrZoEbavv44oI6tu2ULayJEx93f873+of/4ZsU0uL8f+2WcH1X7hyJFw0Pjjjz9oWZHD5tNPP2Xo0KGMGjVKVNMTaoZh4HzxRVwXXkh2nz6kDxqEVCVba8Tu8bK0Wq1xv80XP/IIZowcSSaEVjU75swh7Yknop5IANQdO7CsWgWAVFaG7csvIxbWSYD1119R162j6IUX8Ldvj56ejl6/Pt4ePdBmzIh7Pwcr/aGHsObmIgUCoWBnAlr9+pRfdBGe4cNjHmefNw+5vDxqu1JlGi+A5PFg/fXXmFUE5eLimumiMk3knTuRqik8JRxeCXdPaZqGqqq43W5KSkpo27YtAEUx+k8F4UClDhtG8qxZocFqS24u6ubN5H3yCcjR322UHTtC3UeVTILZW+Om67bbg0GjyvRTCbCuXo3x4YekDhsWt1KenpmJVvHFSd67FznGfnJxMZbVqym77jry5s9H2bQJMyUFo149XOnpwdoXNUx2u1E3b47aLgGlt99O6cCBcY/VW7aM+jkCGCkp+16Ul5Nx//1YVq5E9nhidr3pNZAmXf3lF9KHDg2ukbFa0dq3p+CllzDDsvZKpaXIhYXoDRvG/LsQal/CP/Xjjz+eOXPm8MEHH9CpUycA3G43DpHdUjhYPh/2JUuiZjepa9Zgi9X9qevYvv46eqaTJOE788xqL6XFSIUBwQVxSTNnxg0Ypizj79IlVMhIb9QouLCuatOysvBVrsmQJPQWLWLuV5NMRcGM8QFqynLcuhmVSm+6KTgdOIyRnEz5pZeGXqc/9hj2Tz9F3b4dubgYyTQjusC0pk3xDBlyUPdQGZisv/6KkpeHumMH9oULSb/vvoqLaKTdfz/Z//d/uC64ANd55wVnvgmHXMJBY+DAgWzZsgW/38/VFRkt169fT7du3WqtccKxQS4pQYrRHST7fKgV42jhJJ8vdvoM08SyaVO11/JecAFGlS86WsOGlNx1V8wuKQjmZSq57TYKpk0LbjBNbEuXYlqtGBULAwEMmw1vr17ozZtX24aaZqaloZ1wQtR2rXlzvH37IrvdcdeUmCkpuN98k/L/+z8CLVqgp6djpKVh/+qr4IeyaWL55ZeoAC0B/rZtKbntNvI+/pjAQda6ccydG1xIWYVl9WqksjJSn36apI8+Qt26FSU/H+uaNaSOGIGybdtBXVc4cAl3TzVo0ID7KqN+ha5du9K1a9cab5RwbDEyM9Hr1UOpMtVTz8zEG2P1selwBNcXxEqfceGF1V6rdNAg5OJi7PPnI5WUYNSrR/H996M3a4aenY1lzZrIaykKxQ8/TNlNN4W2pd93X8RYgGG3E2jXjtIBA/BedtmB3Hq15D17SJkyBWXXLry9elH+r39FFVOqVPDSS2QMGoRl3Trw+9GbNMHfvj3ZF1+MXFSEnpVF2fXXU3bjjVHH6scfj+fpp3FdemnwSauwMDh+s2YNRSUlSIFAzGv6zj2X4iefjNgmeTykTJ2Kun49WuvWMHRoQvcqeTxRq+uB4KQGvx/rd99F5c5Sd+0i+eWX8YwaldA1hJqRcNCobpZUr169aqQxwjFKkii56y5SR4xA3b0bACMpCe9556G3bh1z/+LBg0l/9NHQgK3hcODt3RvtpJP2e63ixx6j+JFHgosBKxa9qWvWoPz1V0T/fuUYSVnY6mh11SrsixZFDB7LXi+SadZotld11SoyBw5E3bIFAPv8+Tg+/hj3zJkxB/rNzEzc776LvGsXlpUrSX7jDZI++gi54gNf2bMHZcyY4DqNKl14jg8+IPXJJ1GqrBiXCwtJeucdtBYtULdujXhPd7kou+66yP3dbjKvuAJrZbLEL77AWLwY+Z139pvU0HvZZWjTp4cWJYau07QpZno6UowZcUDMJ06hdiUcNL7++uuI14WFhezatYu2bduKoHEsKC8n6X//w5Kbi69Pn2ACuxociPT260fgpJNImToVqaSE8n//G181f1e+//s/8j7+mOQpU5ALC/F1745cWIj988+DTyf7Ww8hy6GAAZA2dCiWKl0dpt1O0fDhodxUAI5582IunpO3bUMqL485O+ufSBs5MhQwAKRAANuyZaQ++WTwm3WcQWfn+PE4PvkkZhuVggJSXnuNgvCg4fORMmFCVMCoJHs85E+bRubNN2PJzUUyDAxVRXe5MKqkRHE+//y+gFF5/OrVOEePpmjMmGrv13C5KBk0iJRp01C3bcO0WNBatKDwhRcA0Fq1ilgbAhWr76sELqH2JRw0hseYsrd48WK2x1r5KhxV5N27yfzPf7CsW4dkGDjmzsV/+um433oLKtJk1wS9RYv9frhE7N+4MZ5Ro0gdPpzU0aNR8vIwLBa0E07A/eabGA0axD1W2bYNyw8/oLVrh9amDUqVb9IQfIJIfuUViiZPDm0LtGoVc/aQXFSEGda9IpWV4Xj7bZRduyi/4oqYYw7VkSueuMJJpkny//4HkhSzS8a6aBGO994LPV3EInm9wbTvfj+SppF53XUxxxIqGdnZGI0b4+/SBXXNmmBaFE3DunYtWdddFyz2VPHloWrerkrVnT9c2YABlPfrh+PzzzEyM/H27h36+yp8/nmUnTtR165FLi9Hr1eP8v79D3osRThwB5WwsGfPntxyyy1cf/31NdUeoS7w+0mZPBnrDz9g2u3IHg/WsL5+2evF9u23JL39dsw+8tpkXb6clMmTkYuL0Rs0oOyyy0h6//3QN2s5EMD6xx+kPfooBW++GX0C0yRt8GDsX32FsndvMANt586YFcWIqnJ8/jnm44+HPqS9/fvDo49ClcJHkt9P0jvvUHbrrairV5Nx552of/2FZJokvfdecDwiLPjsjxk+5TX8OoaBfeFCSh58MKLLR121isy77qo2YJiKAmVlZJ97LlJZGVJ5edzZYgB6/fqhbjzbsmXIVccU1q7FtmhRKOutESdNeqzULHHbmJlJ2X/+E3N73iefYP3mG9S//sLbuzdG48YJn1eoOQkHDaPKIJXf72fZsmUk19IKV+EwMU0yb7gB2zffhL5NmzG6eiRdx7ZsWUTQkHfvxvHRRxhZWcEpm1WS5MWirlqF88UXUdxu9MxMih96CK1jx5j7Wr/5hox77kHZs2fftmXLYi/Ei/Pt1v7BBzg+/ji0KK8y223gxBMxVTWq71z2eknKyaHs2mvROnQAVUXPyorqe5cg1H2SNnQolrAcWIrbjeODD9DvugsSTLtT9u9/o65fHzOlubJzJ+r69fjDuplSR42KO/sLgl05ekYGtl9+iTuwHU7LziY/Jwf9uOOQ3O6Y55Z9Pixr14aCRnHFWg41bIKC0agRxQ88sN/rJUSS8Hfvjr9795o5n/CPJBw0rrnmmqhtmZmZ3HHHHTXaIOHQsSxfTsprryH5fHh79KDs5puxfvcd1l9+ieh+iVXxDcDIyAj9O3nKFJJffx111y5MRSFlyhQKpkwJftDGoaxbR+Ztt0Uk97OsXYt75ky0isWj4VImT44IGL80gKG9StibBK4yGLEEulR+lsfoNrN+/TVpI0dGlUeVTBPT4aD0mmtIfvvtqPuVi4pImj0bT4cOwbUXTZpEBQ0jKYnyiy+GQCDmNFCloABmzoTBg+P+PMKVDRiA5PWS+vzzUR/yRr16UWlKlBjdWbBvMN/z1FOkP/xwQgEj0KoVeTk5mBW/XzMjAz07O+oaekYG3vPPD73W2renYOpUnGPHIhcUYGRkoIwYccBdc0LdlnDQmFzl0dpms5Faw7UBhEMneepUnGPGhD5AbUuXYvvuO7QTTkCOMSOl6qphw+Gg5M47geATRmXAgIrkgRs2kPb44+RXswArddy4qGyw6o4dOMeNo2D69Kj95bAMsWuy4LKrYUtYj8gf9eGzWdAhX8F35pmoq1bhyMlBb9YMqagI53//G3OAOHhhFc9zz2H7/nssVXIrmVSseK5QPGQIyr33hmYUGTYbvh49gt+ATTNmV5cpSZhNmsT9WVSSCgpImT4d5e+/8XXvTlm/fiR9/HHow95UVXxnnYXRsGHEcUbYqulw/lNOIf+jj0BRguMZ+2FKEmVXXBEKGMFGSZTcfTepw4fvm91WOVutSkAIdOmC+913Q69dLletrIIXDp+Eg4bIZnv0kPfujQgYEOwrt33zTfADyeGIzkckSRG5hSSfj+SZM/E89RSOjz4KBYxwyvbtSIWFcUuCqlVmw1SyLVmCq08ftFat8IwciZGZCYAe9jc4/NzIgAGwNQ2GX5jETO+lSKWluK65JjhALcvBD80437INqxVvnz7BFeXdu6Nu2hTRTaW1aBGapSN5PCTNmIGpqujp6ZhOJ57HHsN76aXBn5Ek4T/rrGAtjbAnFq1FC8wbb4QYeZ5CP69Nm8i88cZQ15Zj3rzgU8IDD2BftgxME98551Byzz1Rx5Zdcw2WjRsjgmKgZUvcb78dmkmmHX98VLdd1S8DkmniyMmhdNCgiNlx3ksuIdC+/b7ZbZddhq9Pn7j3Ihy9qg0aI0eO5IknngBg2LBhSHGm+Y2IUQrzQL388sv88ssvpKWlMXbsWABKSkoYP348e/fuJTs7mwceeICUOAOEQuKcMbpoIJjJVS4tJXDqqViXLw99mBgWS9QAq2QYWL//Pvh+VhamokR3Y1ksEGeAGUDOz4+9vbQUa24u1txc1E2bgjN0VBXPkCFY1q9H3bKFvKTY59x1xkmU1buCrJtuClWqkwwj7opoPTOT8ksuCeVn8jz1VHDF99dfI3m96E2aUDRiRGhgOvOWW7BVZMQFMAsLcXzySUSd7aKhQ6G8HOvKlUg+X7BI0jPPkJ6cXG3QSB0xImIsRNI0rCtWUH7xxeR/+GHc4yBYbQ+LhaS330YqLQ3OLHvqKcywQeiiUaNQBgxA/fNPJF3HsNli/h0objdSUVHk0wYHPrtNODpVGzR69OgR+ndtr8Xo2bMnffv2ZUpYzv+cnBxOOukk+vfvT05ODjk5OVwn5mUftKr98ZVMWcZ/2mmU3H47znHjsK5cGexqKSvD/uOPUftXfhsvv/RSUqZMwbJhQ8T7gY4dq123YKSkoMQJHJUs69Zh/+wzvP36obdtS94HH+B84QVaeObwFdFjLU0djYKzqeKsO4i4fnIy/tNOQ926ldThwym+/37MzEw8zz4bfKoyjIj1HmpuLpY//oj8GQDWlSuR9+zByM4m9ZlnsC1ciFxUhOFyUTxgAOU337zftkDs34uk69iWL6dswID9Hl9+xRWUX3FF3Pf1445j77x5wSfDv/9Ga9qUtBEjogbbjbS0iCSBghCu2qARnleqZ8+etdqQdu3asSdskBNgxYoVPPXUU0AwgD311FMiaNSAeKtzjbQ0/GefHVw1XfGECWD57jusAwZEfRAH2rUL/sNup+Dll0kbMgRl+3awWAiccgqF48dX34769SFGdtZwkt+PZe1avP36BY9p3Jiy665j5PUf8V0jyA3LBXjiHhjW9iL0Bn/EPFf4+goTIBDAEVaRz/b11+R/+GGwO0ySohYIKtu2RYyrhNpYXIy8Zw+Ojz4i6c03kSvGDpT8fFLHj0c75RQCFUk+q2PEmYmo12RJVpuN8spJLaZJUk4Otu++23et5GTK+/ePm65EEKr9y0i0wFJtPYUUFRWRUfGInJGRgaeWS2UeKzwPPojl558jBqFNWUavV4/kadOCXTVhXZGBM8+k7NprsX/8MequXRgpKQQ6dKAorFCP1r49+XPnBusg2GwJrYwuuftu1IceipgRVZVhsewLTgCaRsq0adhLYdkMeLY7bMiClm4Y8ns6XN+e0vZnkDRnTkS6cNNiwXvGGciahuTzoebmRnXNWNavJ2XMGDzPPRezLf7TT0dr2jQ6pUajRmht2pD25JOhgFFJcbtJmT59X7LDapRVLKAMH5fQjjsu5hhGjZAk3DNn4nz+eSy//oppsVB+5ZXVPq0IQrVBIzx1iGmarFu3jvT0dLKyssjPz6ewsLBOpBFZtGgRixYtAmD06NHBGRtHEFVVD22bXS7MDz7AeOoppG+/RSopCY5RrFuH5cUXSfF40Kv2XU+ahDFkCIGvvoLWrZE6dyYr1hhXgvehqirOq67CaNsWXnghOGCemor03XfIYfP85UCA9GHD0A0D85ZbUB55BHnhQiTDIKscxi8I7mcC5gVnolV8ozdffRVj0CBwuyErC/OCC1BGjwZZRrnhBuSVK2O2K2nHDqzx7sHlgltvxZw0CamiW81o0ADuvJPsJUtiVhIEsJkmLpdr/7/ngQPR7XaYMQM8HmjYEHPUKDLbt6/+h3mwKmZGSkByxX815ZD/bdcBR/s9Vxs0wlOHvP7663Tp0oWLLrootG3evHnsijFrpqakpaVRUFBARkYGBQUFcaf49u7dm969e4de5x1hU/xcLtehb/Nxx2EZOJDMb78lvBNG8noxP/mEvHvvhapPCxYLVGad3c9YRCxbi7fyzA/PsLNkJ5kpmdx24m10a9wNJk7cd4kVK8i85hqU8ISAe/agT5hA3oUX4lq8GCXGuhGteXP2vvQS5OVhXb6ctMceQ9q0CUwTzenEc/LJ+Co+1LN//ZV4mam86ekUVve7uPVW1LPOIrlyBlWzZiRPmhQsMxsjiBpWK4W9e+PNy0vs93zxxcH/wh1hf8/hDsvf9mF2pN5zo7Bp5dVJOOPc119/zQUXXBCxrW/fvlGJDGtS586dWbp0KQBLly6li8gzU6Msf/yBEmOlr+TxBGsw1CC3181/Pv8Pn236jF/2/sKiTYu4+6u7WbJ1ScR+ypYtEQEjtH33bpQtW+LOggqcfHIwyBkGacOGYdm4EckwgjU2/v6b1FGjQqk/zDj99XpyMsUPPbTfe9HataNozBiKn3iC5LfeCk6vhX3jJRXBQ8/IwHvRRXj//e/9nlMQjhQJB4309HR++umniG0//fRTjS3wmzBhAk8++SQ7duxg4MCBLF68mP79+/Pbb79x77338ttvv9E/bFqjcPB855yDXrEGIpzhcgUHqQ+QnJ9P2iOPkHXVVWTceSdK2CK5SSsnsbFoY8T+e8v3Mu2rZyO26c2bR5YarWyT04nhchGIscLccDoprch/pq5dGwwuVah//411xQoA/GecEVXpzrBaKZg2Db1p0wTvFmwLFgSfMGK0p+iBB8j/6CMKJ08+6DKoglCXJDxFYsCAAYwdO5a5c+eSlZVFXl4e27ZtY3CCaRH25/7774+5fdiwYTVyfiGa3qoVvp49sX/6KXJFqVU9MzP4AXyAs2ekwkKyrrgiWASoguXnn3G/8gpax45sLY7OIgtQsu1P5B07MCoejQOnnkqgQwdsFWtAIJj7yt+1K2ZaGp5nn0XZsgXrTz8hBwKYFguBE08kUFEMzHQ4MK1WqLqqXZJCg/OeYcOQCwqw/vgjckkJev36lNxxB/4DHJszU1KCP6eqtR4cDspvugnjKO7XFo5dCX8ynHzyyUyePJmVK1fidrvp1KkTnTp1winmcx/RCidNwn7OOTjmzQObjZJbbyXQufMBnydl0qSIgAGgbt9O6osv4p45kw6uDnzx9xdUrRvaoEAjZepUPM88E9wgSbjfeIO0J57AkpuLqar4u3bFU1EhznQ4gum5KxYbSoEAllWrSB0+HM/TT6M3b47WsiVKladidB37ggUETjsNLBYKJ09GKixEzssL1v3+B1NMfeeei9ayZdR9B044QQQM4ah1QP+nOJ1OzjnnnNpqi3A4SBLeK67Ae5DTLC1xailUrvq+46Q7WLzgJX7J2jfN9fgCeO5LkLpF5oMynU4KJ02KfZ3vv8eyalXkNXw+7IsX4xkyBBwOyi69FOtPP0Wlx7AvWEDxo4+G0mOY6enocVKcJMRioeCll0h/5JFgPQ5VJdCmDQVhC1QF4WiTcNDQdZ358+eTm5tLcZVFXjWRRkQ4slXNulqpMm9UkiWJzzd0Y+LqL1lVH+qXwrCl0FRPIT9GBuV4LLm5MdOFSx4PytatOD7/HEdOTtUHmuA+xcXBwfAaqq4HwfUpeZ99hrx9O9jt+y1rKghHuoQHwt98800WLVpEu3bt+OuvvzjjjDMoKiqifW3PIReOCMX33kugSj1vrWFDPGFjXsZzYxm+9yTmfmDhtbnQRMmk7PLLo2pWV8fXo0fswfu0NDJvuw3nCy9gWb8eM8axRv36NRowIs7duLEIGMIxIeEnjR9++IGRI0ficrmYPXs2F154IR07dmR6jBTWwrHHzMwkf/ZsnKNGoW7dipGeTvHgwWhhXyqM7GzyPvsM+/z5pO7cSX7PnugtWx7QdfRWrfD26hWsg10xhVbPzMS02SKqC0pEZnDVGjakOM5kC0EQEpdw0PD7/WRVfJOyWq34fD4aN27M33//XVttE44wRr16FE2YUP1OioL3wgtJcbnQ/+ECqKIJE/D16oVj7lxMm42SO+4gI8YsPolgsPD16kXJXXehN2v2j64nCMI+CQeNxo0bs3HjRlq1akWLFi14//33cTgcZMboKhCEWiVJePv1CyUxhGDlvFi8F12ER4y5CUKNSXhM46abbkKpyPp54403smnTJn7++Wduv/32WmucICSq/PLL0assCtSOO46Su+46TC0ShKNTQk8ahmGwZcsWulcUdG/YsCFDhw6t1YYJwoEoGzAAKRDAMWdOcMFegwYUPf44Rr16+z9YEISEJRQ0ZFnmrbfeOuzZbAWhOqW3306pePIVhFqVcPfUaaedFpV7ShAEQTi2JDwQHggEGDduHG3atCErKyuiXvjdd99dK40TBEEQ6paEg0bTpk1pegAZQAVBEISjT8JB44oEchPl5OSI9OWCIAhHsYTHNBIxZ86cmjydIAiCUMfUaNAwzVgZfwRBEISjRY0GDUlUKBMEQTiq1WjQEARBEI5uontKEARBSFiNBo0TTzyxJk8nCIIg1DEJB42cnBw2bNgQsW3Dhg18/PHHoddDhgypuZYJgiAIdU7CQWPevHk0adIkYluTJk2YN29ejTdKEARBqJsSDhqapqGqkWsBVVXF7/fXeKMEQRCEuinhoNGiRQvmz58fsW3BggW0aNGixhtV1a+//sp9993HPffcQ05OTq1fTxAEQYgt4TQiN954I88++yzLli2jfv367N69m8LCwlqvq2EYBq+99hpPPvkkWVlZDBkyhM6dO0d1lQmCIAi174ASFk6cOJGff/6Z/Px8zjjjDE477TTsdnttto8NGzbQoEED6tevD8BZZ53FihUrRNAQBEE4DBIOGgB2u52zzz67ttoSk9vtJisrK/Q6KyuLP//885C2QRAEQQiqNmiMHDmSJ554AoBhw4bFTRMyYsSImm9ZhVgLBqu2Y9GiRSxatAiA0aNH43K5aq09tUFV1SOuzQdL3POxQdzz0afaoNGjR4/Qvw9XqdesrCzy8/NDr/Pz88nIyIjYp3fv3vTu3Tv0Oi8v75C1rya4XK4jrs0HS9zzsUHc85GjUaNGCe1XbdDo1q1b6N/nnHMOsnzoU1W1bNmSnTt3smfPHjIzM1m+fDn33nvvIW+HIAiCcABjGrfddhtnnnkm3bp1o23btrXZpgiKonDzzTczcuRIDMPg3HPPFRUEBUEQDpOEg8aTTz7Jt99+y8SJE5FlmbPPPptu3bpx3HHH1Wb7AOjUqROdOnWq9esIgiAI1Us4aDRv3pzmzZtz3XXXkZubyzfffMPTTz9Neno6L774Ym22URAEQagj/tEgRaNGjWjSpAlZWVns3bu3ptskCIIg1FEJP2mUlpbyww8/8M033/Dnn39y8skn069fPzp37lyb7RMEQRDqkISDxh133MEJJ5xAt27deOihh0hKSqrNdgmCIAh1UMJB46WXXopaHyEIgiAcW6oNGn/88UfE6+3bt8fcr0OHDjXXIkEQBKHOqjZoTJ06db8nkCSJyZMn11iDBEEQhLqr2qAxZcqUQ9UOQRAE4QhwQFNuNU1jzZo1LF++HACv14vX662VhgmCIAh1T8ID4Vu2bOH555/HYrGQn5/PWWedRW5uLkuXLuWBBx6ozTYKgiAIdUTCTxqvvPIKV111FRMmTAjVCm/Xrh1r166ttcYJgiAIdUvCQWPbtm107949Ypvdbsfv99d4owRBEIS6KeGgkZ2dzV9//RWxrbIUqyAIgnBsSHhM46qrrmL06NH06dMHTdOYM2cOCxYsYODAgbXZPkEQBKEOSfhJ47TTTuPxxx/H4/HQvn178vLyePjhh+nYsWNttk8QBEGoQxJ+0tA0jR9//JFVq1ZRUFBAZmYmTqeTJk2aYLVaa7ONgiAItUr9+WdSx45Fzs/HTEuj+N578YdVLhX2SThovPLKK+zYsYObb76Z7Oxs8vLymO58TegAACAASURBVDNnDq+++iqDBg2qzTYKgiDUGmXDBjIHDkTdsSO0Td2wAff06QREFu8oCQeNFStW8NJLL5GcnAxAkyZNaNWqFffcc0+tNU4QBKG2OceNiwgYAMru3aSMH4924olYcnMx0tMpvv9+9DZtDlMr646Eg0Z6ejo+ny8UNAD8fr/IfCsIwhFNLiiIud22YgWOJUtCr60//kjh5Mn4u3aN3O/zz0l55RXkoiIMlwvp2WehdevabPJhlXDQOOeccxg1ahR9+/YlKyuL/Px85s+fzznnnBORDVdkvBUE4UiitWkDy5ZFbZdKSyNeqzt34hw3jvzZs0PbrEuWkP7ooyj5+aFt5nXXobz9NnrLlrXW5sMp4aCxcOFCAObMmRO1vfI9kfFWEISDoaxfT+pzz6Hs3YuRlhbsEmrWjJQpU1B27sTbsyflV14JasIfXftVPHgw1u++w7J6NVLFNiMlBbmkJGpfOSw4AKRMnx4RMACkbdtwTphA4UsvVX9hTUMqL8dMSQHDQN24ESM9HaNevQNqv1RaSsrEiVj++OOQdKMl/JMXGW8FQahNyrZtZN14I+qWLaFtlt9/x1RV1N27AbDPn49j7lzcs2aBolR7PuvSpTjHjUPJy8NwOim7+mrKbropaj8zLY38OXNInjYNy5o1aC1bYpomzilTQkGkkpGWFvE6VmCB+F1ewQuapD79NLYvv0QqLcW02ZA0DdnjwUxKItChAwVTp2KGDQXE5fWSdeWVWH/9dd99x+lGqyk1F64FQRAOQsq4cREBA4j+Fq9pWH/4AXtODt5//SvuuZSNG0l/8EHUnTv3bdu8GTMlhfJ//xsA2e0m9amnUDduxHQ4KLv6akoefDB4neJiHF9+iSUst56enU1JlcXMWtOmWH/+Oer6gZNOin+fY8eS9MYbyLFSMBUXo+zeTdpDD1GYQD2jpJkzsfz2W8Q2dedOnGPHkv/++/s9/p847EHju+++4/3332f79u2MGjWKlmH9gHPmzGHx4sXIssyAAQM45ZRTDmNLBUGoTcqePQntJ/v92BcvjggaUnExsseD3rAhyDLOSZMiAgaA4vGQ9O67waDh85F5zTVYw8Zj1dWrkTweym6+GdPpJP/dd0l99tlQsCm58078Z58dcU7P0KFY1qzBsm4dAKYsY3buTEk1s0rtX34ZM2DkO2B7KrTOB+tvv4FhgFz9+mvbL78gGUb0z6hKsK1Jhz1oNG3alIceeojp06dHbN+2bRvLly9n3LhxFBQU8MwzzzBx4kTk/fwQBUE4MmmtWsFXXyW27/HHB//h95M+eDDWn35CKi3FaNAAz4MPIhUVxTxOKi8HIOn99yOeIqAiqMyeTdnNNwNgZGdTOHFite0wGjQg7+OPSf7vf7Fs2IDv1FP5f/bOO7ypsv3jn3Oy03SXttAWaIECZW8QZQj4shTBwXwVRARBQEFfFMQFIjhQcaG4UBBFGbKUvVfL3hRaWkoH3W2ajjTJ+f0RmjacBIogoL9+rovrImef0+S5z3OP762fPBnJjdsKQCgpcfpsFeCpfrC1NuTooIYRnriQw4ir9lMeOYJuzRoskZEUPfwwqNWURkaiheu60W4ld9xohIaGulweExPDPffcg0qlIjAwkODgYM6fP09kVZ50FVX8KzE+/zzqXbtQnz7tWGb194fiYhQVMplK69TB9MwzAHhPn45u5UoESQJAkZ2N92uvUThwINotW5BsVv6sA2cC4KGzUOOKsVEdPYpgsciuQczNLf9gNqNftAjNvn1YwsIwjRuHzc9Pto/k6UnBiy86Puu1WriG0bDUqeOYmQC83gV+agyWK6NxrAY+8DbRKP0AbYPbgiThM2EC2k2b7HEPhQLDggVk/fADplGj0K5Zg7qiGy0ggIIrz+fv4I4bDXdkZ2dTr0Kus5+fH9nZ2XfwiqqooorKYrFZmHd4HtEZ0VgtVnqH92Z41HAE4ep34nIkHx+yli3DMG8eqnPnsISEUPDCC2g2bkS/dCmiyYQlJIT8119HuvImrY6O5s86EvNbQ6kI/c/A04eSEQoKSO7chgF193M8QKJYDbM7K+gdpWYWUNytG7rlyxGv6jzqyFwqKcF/0CDUBw8iWK0A6DZsIGvhQqwREde8d5PZxOr41eiUOrqEdkEpOg+zee+8gyIpCdXZswhmM5vqCliUktM2OSoL3578lrbBbdFs2oT2jz8Qr8ySBKsV1alTeE+bRs6335Jd0Y3m4UHB6NGYO3W67t/or3JbjMaMGTPIrWjBrzBo0CDatGnjch9Jklwud8WmTZvYtGkTALNnzyYgIOCvXegdQqlU/uOu+Wapuud/NwOXDWT1udVYJfuAeyjjEGnmND7o8cG1dwwIgCsuIfXp0wROnoxw8SKSwYBt0CAUU6bgW8HwzGmYyfsNIE9n/7w5AnbUgu9tNoaPqkHM2fJxJENnZfmlPxhSMJIuQ4bATz8hbdvmMAq26tURXn2VgIAAxI8/RnHggFO8QBkfT7U5c7AsW+b28n87/Ruv//Y6cTlxKEUl9fzq8f1D39MsyC7saraaOS+lI21dS8D6nQinTmELWAF5p2THUigVBAQEoKhgMCqiSUqyf58CAmDxYsDupvK69hO+aW6L0Zg+ffoN71NWQFhGdnY2fi6mhgDdu3ene/fujs+ZmZk3fpF3kICAgH/cNd8sVff8zyMxP5EPD39IdlE2zas159lmz6JT6mTbxeXGsSNxh8NgABRbill1dhXjG43HoDZc91xCfj4BAwYgXunhIwCcOUNhQQEFzz/vOOYP9YsdBgPArIQNdQUO3ncfJy/Mkh3XWGpkfvR8GhsawzffoP/+ezS7dyN5eWEcPx5rvXqQmYnv1q0oXQSYLQkJbv+GhaWFTNsyjYT8BABKbaWcyjzFqFWjWPvwWn44/QPfnfyOVFMqXmov2ga35aMxH9FkXyoxVxkNnVJHYk4izb9sjiE8k6HtYOL+q65Fqbyl36caNWpUaru71j3VunVr5s2bR9++fcnJySE1NZW6deve6cuqoop/JcpDh/D87DNEo5HSqCiMkycjeXo61h9OP8zozaNJLkgGYHPSZrZe2sqvfX5Fq9Q6trPYLBzPPE52idyVnF2cTXJBMvX96svWRadF88mRT8g35xNqCOWdY0FUv6rpm1hUhHbdOofRSC5IJt1bAaXOx0r3kNjrmYNadK2+7TBaKhWFo0ZROGqUbBurm+ZyksG9wduStIXE/ETZ8iRjEjuSdzD30Fwyi+yDfEFpASvjVrI3dS86pQ4vtReltlKKLEUE64MpthZzOOOw/QBaON0NipTw8m77IptGQ3GPHm6v5e/kjhuN6Ohovv32W/Lz85k9eza1a9dm2rRphIWF0aFDByZNmoQoiowcObIqc6qKO0aJtYQtF7dQaiulR60eLt+w/6loNmywS2FcSXnV7N6NOiaGzOXLQaMBYE7MHIfBKONI+hF+PP0jo5qMIrUglck7JxOXG4eEhFpUY7Y5p5UG6AII8wyTnX/HpR1M3DaR9CL7+Q9cPsBphSd7dOB3lVdGrBAQD/YIxkfrR16p0Wkbb7U3jQMac1/IfZzOPo1FKg94B+uDGdv0+qrcBePHo92yBWViuRGw+vpievJJt/tolVqUopJSm7MVU4pKfj77s8NgVORy4WXH/z1Vnjzb5Fk81B58cNDZjVeghiVtdbx4wRfJw4PiHj0oeOGF697H38EdNxpt27albdu2LtcNGDCAAQMG3OYrqqIKZ2LSYnhx54tcyLuATbJR26s209pOo1d4r0ofI6c4h8sZl/Gx+aBRaP7Gq3VDURFes2ahPnIESamkuHt3TGPHgiBg+PxzWY2E6vhx9EuWOCqoM4oyZIe0YePA5QM83fhpntr4FMcynYvMBAQk7DEFD6UHD0U8hF6lL99AktBs3Mj8M1NJ1zmf/6zWyDudFbz3p9VpuVShCtxD5UGPWj1YdHoRxVZ7QFtEpHVQaxr6NeTlNi+Tb85nV/IuTBYTgfpAJjSfQE2vmuX3uXs3HosXI6lUmEaPxhIVZb+3oCCyv/4ar1mzUKSlYTMYMA0fTvFDD7l9xF1CuxDhFcHZ3LNOy+v51EMhXLt6Heyus2NZx6jlWcvx3CqSG1qN9J1bQa2+bv3G38kdNxpVVHE3Y5NsTN0zlfO55x3LLuRfYFbMLO6vef91DYDFZmHy9snsSd2DsdRIoC6Q/zb4L6Oayl0iQl4ekloNuls/i/EbORLN9u2OfH7V0aMoUlPJnzkT0UVWomC1ol+0CN2aNUiiiG8PeSAWoI53Hfan7edczjnZOk+1Jy2CW6CUlDwe+Th9I/qWrzSb8XviCdQxMeQ/UQwubjk2wgeJLKcaBEVaGrqVK+11CsAb7d+gpmdN/kj4A6vNSpvgNrzYyp7+qhAVzLlvDsWWYvLN+QToAhCF8sHWa8YM9IsXIxrtMxXt1q0YJ01yGEpLVBTZixa5eaJylKKSeV3nMXXfVC5kX7AHwn3q8Xm3z4nLjWNz0mbyzfnXPEaRpYhe4b1YEbeCIovzMw/WB4NW62bP20eV0aiiimtwPvc8SflJsuUJeQnsTd2L1WZl0ZlFWG1W+kb05bF6jzmllc6Omc3KuJUOF4nRbGTekXm0qd6G5tXsCgfKEyfwnjoVRXIyqFSYW7Qgd+7cW2Y8lAcPoo6JcRp8xZISNFu2IBQWYnPR3kACVGfOOOofXjCqONdPS6aiPEW1nk89RjcdzfZL2ymyyo2KgMCnPT/FT5InsHh8/TWa3bsRbDaCTLLVADSxBSLgXNksmkzoFi92GA1BEBjZeCQjG490e/9apdYp7gIgZmSg/f13h8EAu2SJx/ffUzhkiP1t/i/QOKAxu5/czZGEI6gVaoL0QYDdNTes4TBWnl9JiikFjaihxFYi2z/CK4KuoV3pXFyDLcRhvjJC18wTeCO7Orif6Nw2qoxGFVVcA61Ci0pUyZarFCpWx61m7YW1GK/41Hcl72Jf6j7mdp7r2G5Pyh4nnzpAdkk2Xx//mk/v/xSKi/EdNw7V+fKZjDLJbqQqoz10PZQHDuA3ahRiYaFsnZiXh5ieTsHo0SgTElBUzMRRKBypqACPHStF4+HD+0+2pMBaRJhnGNPbTcdb402n0E6EGcJIKnA2rjUMNajrV5fsLPlMRrNnjyOddeYWOFUNLvqUr2/i34SJewOA07J9y+4luzib32J/Q6VQ8Wi9R/FUe8q2dYd63z6ZzAiAIjUV5YULWOrLg/WVRRAEl7GbaW2nMbrJaPam7iXUI5SZ0TM5lH4Is82MiEhD/4a8XudZ/IcMYfWeCyxtACsaQlABvLxbItjzAJmjsl0WGN5OqoxGFXclqaZU8kvyqedbz+FSUFy4gPdrr6FITkbS6yl66CFHZfDfRU2vmtT1qUv05Win5eFe4exP2+8wGAAlthK2Jm0luSAZAYG3o9/mXK7cbQN2txWAbuVKlFdlCQF21dLi4ptzR9hs+Lz8Msqr4hVWAX5pBOtaWqmTs44nHhiBze9LDF98gVhQgE2jQbd9u+xwfY8X0T5ypmxA9dH4MLLxSL449oUjsBtmCGNK6ylO7iDBZMLju+9Qnj2LUEEFtlUqbF4Ib3aByyE+RHQdyPMtn0edsQRp43aZtpKldm2WxS5jzsHy4PyC4wt465636F6zO5XBEhGBzcsLMd/ZXWTz8sJ6pcAvrySPDw99SGxOLH5aPya1mkSE97UL+65HgC6AByMeBOCXPr+w4vwKdibvpElAE4Y1GEpYv0dRHz0KwJAT9n8O8pNRHT1KSdeuN3UNN0uV0ajiriLfnM+YTWM4lX2KYksxIYYQpradSne/dviNGIHqXPkgrIyNBYvFHtC9CkVcHF7vvouYlYU1OJj8V17BFhLyl67pi25fMG7rOOJz47FJNmp61WR8s/FM2DZBtm16UTrbk7Yz//h84vLiXB7PQ+XB4AaDARAzM10KzlFaimA2I92E0VCeOuWU/QNgVsB/hsGeMDArjQgHZvFr/HKW9FpCtYULAVAkJaHu29d55gHYAgKwhsnfoAFGNRlF7/De/HTmJ3RKHUMbDMVXW+72ElNS8B82zG4wAEmpRFKpEErtmUZ1c2DhZi9yZ8+iuH0/AEwjRqDdvBlVTAyi2Ywkiljq1yft1Sl8uGOIUzZXojGRWdGz6BraFYV4/aCzpVEjzE2bot21y7FMUigwt2+P5OtLgbmAx9c+zoms8lE7+nI0X3X/yuFWvFmUopLHIh/jscjHAFDv3Wv/TrvB6udXrrl1B6kyGlXcVTy/7Xm2J5e/5Z7JOcOre16lS+YjKM85v7WLJhO6VatkRkMRG4v/E0843Dxgf3PP/O03bG7y7wHSTGksPLUQCYknGj5BDYO92CnYI5hlfZeRXpiOxWahhqEG+eZ8fDQ+TjMNAF+NL/tS97k1GNX11Xkw4kE6h3YGoOiRR/D49ltHv4gyrCEhSF43V9trU4i80qmUjTWhRAn1s6BJmr1i2nZlAiAhcTr7NG/se4PP7rf3zLGGhVHSuTO61asRrqixWj08KHroISS93t3pCDGE8FLrl1yu8379dSe9JcFiQRIELDVqgChi8/amcNAgivv1K99JrSZryRK0a9ag3bIFc+PGFA4bxr6s/W7rIc7knKGRf6NKPZ+c777Da9o0VCdPgkJBSfv2GKdNA+CLY184GQyw14W8f+B9FvWqfHD8RhAvX3ZZ+Q32GFNpy5ZYw8P/lnPfCFVGo4q7hhJrCaey5HIKF40X6SZ9RYPB8Nx+eKCCN8dVExyv995zMhgAygsX8HzvPfI+cC1j8cvZX3jv4Hukmux+7l9jf2Vyq8kMaTDEsU2gvryjmpfai65hXfk59mfMVvvAKiLSJqgN+aWuM2QifCJY1meZ03Fs1atjqV0bRXo6gmRPtJTUaowT5LOYq8k357Ps3DJsko1H6j2Cj8bHaf0r2YtZ3KaU0isv3qcCYXPtcoNRkbIq5jJyP/oIc6tWaDdsAFGkcOBAivv2le/oArPVzHcnv2Nv6l6CvYMZ1WAU7ZLkyQSCJFHapAk5Cxa4baikPH0a7fr19voMQQClEh+1D1qllkKLc5ymrEjO6RzZ2SiTkrBERDgVKwJIej15H37o8rxnss+4XO6q1uJWUdK5M5bQUJSXLjktt2m1FA4cSP7rr/9t574RqoxGFXcNNsmGTXLhqgHOaU2cqw+7wuDNrTA+xr7cVeWumCGvKQBQpKSgWb8e/bJlSFotBc88g6VxY4otxXx29DOHwQBIK0zji6Nf0L9uf7eFfLM6zqKmZ002XdyEVbLSJqgN/2vzP+bEzGHjxY2y7ev713cyGADKM2dQxcY6spQEQDCbMXz9Ndndurk8L8D6xPW8sfcNLhrtTYsWnFjAy61f5uG69qwis9XMjuQdDoNRRoGbDGG98qoZhChS+OSTFF6jmM0VVpuVYX8OY2/KXmzY/5ab4jfxTag3vY7Lt7cFBLg1GNrVq/GePh3Flb+nZutWtNu20eyHH4j0jeRIxhGn7Rv4NigPQEsS3v/7H5rt2xGzsrAFBlLUt69jJnE9anvXdrk84PQFqt1/P9bQUPJmzMBaq1aljlcZJF9fTCNGYJg/33HPlivnKXnggVt2npulymhUcdegU+qI8I4g2ZTsdpscPSxoBWMPgBRWkzwXg4C7HsuK5GR8n3vOkX2j2b6d/Jde4mCPho7BtyKJxkSOZRyjXfV2Lo8nCALPNnuWZ5s967R8fPPxbLu0jdPZ5Zk/oYZQ3ur8FkiS/Y35CvpFi1C4aA2qTEiQbVuGxWZhdvRsp2tOMiYx99BceoX3QqPQYCo1YSqV57LaRDCoDBSUls/QvNXeDG041OU93iir4lcRkxbjMBhgT2qYcY+O/+zxRqzQ58ISEkLBNWZUhk8/dQyeYK8dUUdHo92xgwXdF/DC9heIz4tHRCTSL5KPu5T3vvCYPx/db785mh2JFy/i8f33mFu0oKR3b9m5Lhkv8dXxr7Bh4+lGT/Ncs+fYdHGTU31ODaPAtD8KUF04i+rsWRQXL5K5dm3l2rJWEtOYMRT36YP+hx+Q9HoKBw1CkZyM4vx5rHeJjFKV0ajirmJu57k8velpYnNiZcVNZWT4aUgaPQjtmBewVasmW5/zvxf5wrqbTX65qK3wXDR0sdZEzMlxSj1VZGZi+PZb/Hp/hUFlIKfEefA2qAz46/xv+B68Nd4s7bOUDw5+QHxePP5af14WutBy4ARsSUnYPD0p7tuXgokTsfm7Pr6kVrs0GACnsk65NHIX8i5w6PIhOtTogI/GhyB9kJNMBYCfxo9pbafxy7lfyCjMwFvjzeD6g3m4zsM3fJ+u2Jq0VSYfApDmIZE3dSr6n39GzM/HFhhI/pQpWENDUR4/jmb7dkqbNMHcqROKCxfwefFFVKfkrkrRZEKzbRs1unThlz6/UGAuQBRE50pzsPeeuKo7nlhYiP6XX2RGY9HpRXxw6APSC+1ZZmvi1zC++Xh+7vUzs2JmkWRMIuDwaaatK6BjBc+R8tw59AsXukzEuBmsYWEYp01Du2IF/oMGoUxMtPcOb9CAnK+/rkq5raKKitQw1GBtv7UcvHyQXcm7+OzoZxRanX3XXoG1EJ99C5vo+us7Mn4OG9sbKasy2F5Pxf+I5JWZm2Tbiunp1DVpiPKLYnfqbqd1UX5R1PX5a293flo/3u74tv0cKSkEPPwwYnIyZeEEZWIiklaLafhw9EuXOvXGlkQRc/v2bo/tqfZEp9Q5pDPK0Kl0eGnsPn1BEJjUchKv7H7F4XbzUHrQs3ZPBjUYxKAGgwDIKspySou9WRr5N2L5+eUyGQxvjTdFA4dRNGxY+UKbDZ/Ro9Hu3ImYl4dNr6e0cWPEvDynoHlFbGo15qZNHZ/dKua6aa0gSBI2ycbi04vZeHEjoiByPPO4w2CAXTLl25PfMrj+YD7p+gkAgXM6orzkHD8TQJaccasQs7LweucdlMn2WbeQl4dm/358XniB7CtZbneKKqNRxR0nMT+R6Xumc6ngEjqljj7hfXi26bO0CmrFnrQ97E4pH8wNKgMD6g6QNbYp41jGMXulNuWFabnKUhZqzjLJoEdT4GyAbF5e2Pz9WdBjAS/ueJGT2ScBaOjbkA86X6f3QyUxfPqp48dfhlhYiHbNGkxjxpA7d669iU5KCpJOR0m7duTNnOn2eOHe4UT6RrI/zVkrO9Inkii/KMfnHrV6EOUfxRdHvyCvJI/H6z/OvTXuBezy5ZN3TOai8SIiIvV86/HZ/Z/hp725t9gnop7g13O/OrnmPFWeDK4/WLatfskSdBs2ODK0xMJCNNHRSNfQVbJERV1T/6mMknvvtTdQqtCdz6bVUvjgg4zbMo4/E/50OSMqIzE/kcPph+kYYu8Jbq1e3e4yrIAEqE6dQigqQqpE9f7prNN8efxLSqwlDKw/kC6hXdxuq//pJ9l3BkB59iyYzX+5Yv1WUGU0sFt1w9y5KBMSsPr6YnzpJWy3MMB1NVlFWUSnRVPLqxZR/lHX3+FfTGFpIcPXDyc2tzw/PTYnFovNwoQWE1j4n4V8cPADDqcfRqVQMTByIP3r9nd7vL2pe8kzy/tDZwlFpLZqQO3thxzLJLUac6dOSHo93sCC+z5BvBCPNSgIwe/G3VLuULjpOFmWXmnu0IHMNWsQcnPtg08lajMWdF/AhG0TOJd7Dptko653XT7q8pGsM16IIYSZHZ0NkE2yMWazvRamjNTCVMZsGsPSvkuve26hsBDDhx+iOnbM3iluzBhKr4iO6pQ6wr3COZN9xjHbEBBcVmtrNm50GAyn47uoW5GUSop69yZv9mxQXn/YKpg4EWVsLJq9e1FkZmIJDqake3eOdmvCzjVvXtNggH02F+QR5PicP2YMfidOoKggOyIA6hMn8Hr1VbdZeWUsPrOYdw+868i+2nxxM49HPi7725TfcOWb0N1u/t8bDTEzE/9HH3UqGtOvXk3hQw+R9/HHt1xNcnb0bJadX0aKKQUvtRdNApqwavCqW3qOfxI/nP5BVjVdaClkTfwaJrSYgE6p49V2r1b6eM2rNZcFesHuHhHm/4DpzZmoTpwAlYrirl0d8tIeX32FftEixMuXkby9Md9zD7kffOA2s+dGKOrVC82GDYglzlpDToVagoDkQgPKHf46fxb3Woyp1IQkSZVqbFTG/tT9xOXK60jO5p4lpSDFUZ/iEosFv8GD0Rw44FikPniQvLfeorhfP05mnmRPyh4n91R+aT5fHP2ChyIecjZqGtepXDa1WhaPKG3UiNzPP3fEeaw2K3MOzGFX8i5sko2m1ZryZoc3yzPdFApy589HcekSynPnKI2KwhYUxO4T38piV65o4t/E4Zo0fPAB+iVLnAxGRcoquF3fjA0LNr4+/rVTuq7JYmLthbU82+xZQgzyotPCoUPRL1okm21YIiPv6CwDqowGnu++62QwwF54pF+xAmutWk4N42+WmLQYfjj9g+NNON+cz+6U3bzwUiPe7zmfUjetb//NnMs951IG+upBv7K0DW5Li2ot2Jmy07HMQ+nBgxEPovLydflGqDp8GMO8eeVZTAUFKFaswBoYiHHq1L90HRUpfvBBSlatQrtrF0JBAZIoUhoZSd6MGTd8LLPVzJwDczhw+QAKQUH3sO6y7K3rkW/Od/mmXWotldU+XI12xQrZIKnIzMTw9dcU9+vHxosbyTXLWzunF6WTW5LrVCVe8MwzqPfscZqJ2fR6Sjp1QnX4MMrLl5EAa3g4+W+8gQ2J3OIcvNReTN4+2UkI8njWcZKMSSzpvcTpvNbQUKyhoY7Pzao1w1PtidHsbAA8lB4E6gMRBZFG/o2Yc98cwJ4S7fH9925ni4DcnSZJGD78EO3atYhGI7ERPmT0SHH5THYk73DpurP5+2N8+WUMH31UHghv2JDcjz5yex23i//3RkNxUZ6FAvaAmXbrFFpyrQAAIABJREFU1ltqNCoajIrsV6TxyYJhBBU9R5+Oo1Er7uybxO2kb3hfVsWtkg1W1T2q/6XjCYLA9//5nncPvMuh9EOoRBX96/Z3KtK7Go+vv5alvQoWC5rdu3H9bnmDiCI5X39NtRMnKFm8GEuDBhQOGvSXdKWe3vg0W5K2OAzt4fTDJBUk8c6971T6GJ1CO1HLq5asoC/UM5Q63nWcN5YklGfOgChiiYxEs3evQ/rD6RavtGZu6NfQpYKrQWWQzYZKW7Ui/+WX7YNyZiY2Ly+Ke/XCOGUK4uXL6JcuxeblRdFjj/HjxRUsXD6NzKJMPFQeZBZlyoQgj2Yc5WTWSXtFeGkp+oUL0ezaheTtbW/lWrcuLQNb0qJaC3Yk73Dsp1PoeDLqSaa1k6dv6xcuvLbBEATMrVo5P4sPP8Tw2WcO92NY+iW8WyvIuarA36AyUM+nnttjFw0YQFGfPnY3oK8vlqqU27sDW0CA+5Uufhw3gyu1VIDzfjCjWgHK03P4LON3vn/ge0I9Q11u+2+jS2gXOoV0YmvSVsdAE+YZxittX/nLx9QqtbzW/rVKb18xWOqE1ep6+V9BEJC6diW/SZO/fIjY7Fj2pe5zmpmZbWa2JG3BaDZWWuVVp9QxueVk3jv4niMQHu4dzowOM5zcR8qTJ/GZNMkeABYELOHhFPXsiaRUyp5Zmbx6WfDd0aoU0Cg09KjVg9SCVC4XXqaRfyNHimzR0KEUDR6MmJODzcsLVPbfiC042FHDEXM5hjkH5pBdfGXwdp2JTZ45j/i8eBr5NsTviSfs0utX/obqXbvIe/ddSrp147sHvmN2zGwOZxxGJaroU7sPwxsNlx1Pu2oV2vXrXZ5Lwt6oydyyJflvvOG0Tly+3EkOxNMM3c5ZWdhSwCKU/+2aBDShVaCzwZGh0dx1Hoj/90bD+OKLaHbulImzAVgaNLil5xrdZDSbLm4iq7hCjwAJrFfc5hbRrgP08u6XWdTz79G3+Vux2dBs2IDq5EmKu3fH0qzZdXcRBIEFPRaw7sI61l1YR6A+kLHNxsoqp/9OCgcNQr1tG4oKkiQSUFrh+rdf2s68w/PIKs7CR+PDmKZj6Fm75227RoD9afsxWeQFe6mmVNJMadc0GpIkEZ8Xj06po4ahBgPqDaB7re6siV+DRqGhd3hv58p3mw2f559HXaFWQn3sGNhslDZpgvpwuVGw+vlheuopAERB5MeePzJ9z3TO5Z5Dr9FzX/B9HEo/RJ/f+5Bfkk+oZyjDGw4vb0Qlim7rVQA+O/JZucG4BsH6YNoGt0WzYQPq6GgnaXdlWhqeH39MSbduaJVa3ujwxjWPpdmyBe9XX0WRlSVbJwkCxT16kPfOO661zFzoR325Bnxr1GFjEw8skoUovyhm3jNTlrjwT+D/vdGw1q5N5tKl+D/xBIrUVASrFZtGQ2mTJtdMe/wr1Perz6RWk1i4eTaXBCOSAEUuPFEJeQm39Ly3AyEvz65ievIkYkkJhm++obhTJ3vw8jrJBKIg0jeir3Nnt9tISdeuFA0ahHbtWpSpqdi8vTE3b07elTfI45nHeWH7C06Fcgm7EvDWeNOheodbcg3ZxdlYbJZrGsuKGWYVERBcBlPLiEmL4dU9r5JkTEIlqoj0jWR+t/n46/zduu1UR4+ijJMHy5UJCWT99BO6339HdfIkkl5PwahRmDt1cmzjq/W19woBAgICGPbbMLZe2upYn5CfwCdHP6FzWGcifSPdXneaKY3xW8dz4PIBl+sVggKrZDcMBpWBvhF9CdIHod28GbG4WL59XJw9Q83HR7buajwWLHBpMGw6HSXdupEzb57bQD7h4XDCWexQUGt4pcWLvPDgg9c8b6mtlIzCDPx1/nemLXAl+H9vNACs9euTvn8/ivh4NFu2YGnQAHPHjm4rcm+GEI8QCAiEHCO4llm6pTENMSsLMTsbS3h4pVIVK0OJtQRTqQlfja/jTclr+nTUh8rTWcW8PLTr16NduZLif0Cf9/w336Rg3DjU0dFY6tZ1mmV+cvgTWWV1RlEGXxz9wsloiNnZeE2fjvL8eSSNhuLevTGNHn3N71F2cTbjtowjNicWq2SlpmdN3u/0vsvB1J17U6/SyzrTlVFsKWbyjslOqruZqZmM2zqOn3v/7Pa6sFpdS7ZLEpJWS/5bb7nf9yqu7h0OkFWcxdfHv+bdTu+63W/0ptEcSHdtMAC6hXXDoDZQai1lUINBjrqH0iZNkK5qIgWgyM2lWs+e5E+ZQnF/92nblwsvY7JlUxe4+i9nbtWKnC+/dFpmsVmwSlbHIG/56CM4dw7VuXMIVitWDw/MnTpR3KeP23MCfHnsS34++zNZxVn4anzpE9GH/7X+3zX3uRNUGY0KWCMiKIy4uSYr1yK3JJfX9r7GxYKL4Pr3j1JQcl/Ifdc9lmAyIalU7tPviovxHT8e1eHDiAUFZNQO4rWhIZwNgGraakxuNZmaXjVv6Pptko1pu6exM3knhZZCAnWBTGo1iQdqPSDLQAMQzWZ069bdlUajxFrCa3te40D6ASRJorF/Y2Z1nIXBhZKrO9VapwwviwW/oUPtLpwrqE6dQjAaKXjJtVw4wLObn2VXSnlPh4yiDMZuGcv6/usr1RcCwBst1R59DEVKCjYPD4p79aJg0iQQBP5M+NPlzPV8znmyi7NdF/OZzRg+/xxcGA1L7dpYom5NbZGrrDnH9eWed9vAqoz4/Hh+7vUz1Q3OSROFAweiX7QI9cmTsn2USUl4zp1Lca9eskSEzKJMnt38LOdzz2PtnEudBvDVamh2uex6wdy6tWP7IksRk3dM5mjGUSw2C7W8avHefe8RENGKzLVr0f/yC8ozZyjq25fSe+655svDtqRtfHz4Y0eiTJlRrWmo6ajev1uoMhq3kV/O/uJSM0gtqvHV+OKh9qBjjY681s59EFd59Khd+TMlBTQazG3akPvuuzLj4T11Ktp16xCAXC080NHIseLzcEU7Z3/afr77z3c09GtY6eufFT2Ln8785MhauVx4mam7p9LIvxH+KtdW8FaKud1Kxm0Zxx8Jfzg+n805y+XCy/zS5xcAVDEx6JcuxRoSQsMWEexkp+wYFTONtKtXozrjLKctFhWhW7eOgsmTQRSx2qx8eexLtl7aikJQ0LNWT2Jz5C6n87nn2ZmyU1YxXLHKuiLVL+Wg2bfP8VmZkICkVGKaOJFia7FTdXwZVsmK1eY60O81YwbaDRscyrtgHzCttWqR+/77NzwDbxLQhLM5zrIg/lp/nm78tNt9TKUmSqzyHtoVOZ97nhn7Z/B5t8+dV2g0ZC9ZgvfEiWi3bXO6D7A/H/XBg3ZvQgXGbB7D3tS99g8qyAqFYQPg8Jd2V1hp8+aYnnvOsf2ErRNYl7DO8flSwSWe3vg0B0YdAK1WrhBsseD1xhto9uyB0lIsERHkzZ6NrXp1vj/1vSyz0mQxsTxueZXR+P+Mux9BsCGY1Q+txlvj7dYFYbFZyMtJod74cWjiLjiWKxITkQTBqS+AJEnEnd+HEAiN0+Gde+HYVfG6pIIkZsfMZuF/Kq9jsyN5hyzNMdWUyvyj8/k4Kgr1kSNOLgFrtWoUPHtjNQS3g/TCdA6lH5ItP555nNicWNq8/RW6NWsQrxRzvVOvJtGj6nGsMA4bNgQEovyjnDK8VKdPO6qbC1WwOxQCiqCxMR+hpARJp+O/v/+X38/+7niG0WnR2Fy9zdssFJjldSqulgF4G52z/MSiInR//olp4kT6hPdh3uF5JBqdmxbV9KxJNb1c7BFAfeCAbKAVAHPz5lgq6D5Vlnc6vkNWURbHMo+RU5xDTa+aPNHwCer7ue/D3ci/EbW8asmMzdVc3Ze8DJu/P/kzZ9o7EF6VTi3pdNi8vZ2WpZpSXc5szvrDxnoi92sbk7V0qWN2km/O52imvKjvXO45/jj/B+395Nph3lOmoP/1V8dvRBUfj7pLF7IXLqTU5jpTs6wt8N1EldH4m4nNiSXJmESroFYMqj+IH07/4NS3AaCeXz0CdO5Tfz8+/DErz68kLzeFgAcKePoQPH9FdkiQJNQxMfb0UIWC+Nx4xm8bz4UHLmKToE4OGOQxQQAyC2+soUyp1fUXu3DHn+iW5yFYrfYmQgoFlrp1KRg//pZnoN0K0gvTZcVdYE/ZzJv4BLrdGU6BVL9zF9m2si0fv/Y6h9IP0TigMSMajXDKNip+4AE8fviBhRFG3rkP4vzAwwxRpnzm2/Ix5qawLXGbk9EttBSilkSZ47yWVy26hjn3gZYkiewSeQaRIMFjci8MwhU1X0+1Jy+1fokPDn5AQn4CalFNhHcE/er0Y9TGUeiVesY0HUND/8rPOG8UvUrPol6LSMxPJM2URpOAJjJV2qtRikpea/car+55lYT8BLeuLIPKfSW8tXZtLPXro6gwCwOw1KuHpZFzd78Sa4nLmZdFhCLBRkn8Gb5fNZXkWv4Mqj8IvUpPiUX+ElhqK7X/vq/2+hUX22tcro6zFBTgN2IE3Wc/zC5B6fT9EBCun5J7B7jjRuPHH3/k4MGDKJVKgoKCGDt2LB5XXBorVqxgy5YtiKLIiBEjaN781vTmvR2YSk08teEpjmceJ8+cR5ghjMcjH2di84l8cfwLEvMT0Sv1RPpGsqDPAooKi5h/bD5HM44SqA9kYouJhBhCWH5uOZ8f/dzhP78cBG91gYYZ8J8rHewEsxlKS5FEkfHbxtub01xJvDhSHfzlWZoA+Givn0VSkdpetWUZPJ6CllF/pqMw2b/sAva+B9Zata4ZbLydXC68zILjC+w6V42GU8+3HtU9qstastbKgU67kxBdGFmPiyk8HTUCmrh2qZS2bk1S17a8XmczSVdeYo1a2K8t5PkdL/BwxMNkFcmzcTQlNkJKFFz0tmFFoqZnTSa1nISHytmtl2JKcWnoJAEOhoiMPGxDrDCuWmuWx6v61+1Pz9o92X5pO55qT3479xuzomdRZLWnhm6/tJ2p7abyeOTjgD3Yqzp+3Gm2YfPywjR8uMt7ryy1vGpRy6vymm5dwrqwYcAG1l5YS4Ypg29Pf+v0wuWv9Wd0k9HXPEbOggX4TJhg771ts2GpW5fcDz+UudhqedYizDNMJjFSNxuCTNDmyRJic3+BPPg59mcGRQ6ihqEGmcXOL17VParTL7KfrJZENJkQ3LRyVeTn88JPcex5+gF2p+wmz5yHp8qTloEtebH1rSsuvlXccaPRtGlThgwZgkKhYNGiRaxYsYJhw4Zx6dIl9uzZw9y5c8nJyWHGjBl8/PHHiLdYC+rv4pVdrzgFOJMKkvjm5Dcs6bWEDf03sCd1D/5af1oGtsSgM3D/r/c7uUx2JNsbzSw9t1QmqZGjg8/alRsNa1gYaLXE5Z7nQt4FrsaohlCjwCXP8kEgWB/M8y2er9S9rDi/gq+Of0VGYQZahRaz1YwNG35aP/onedL1vLxfszJRvuxOsCZ+DW/ue5MUk13GYfWF1TwV9RRPNXqKDw996PjR+xXCU4fB182sTNLrr5s6/OUzbUg6sFm2PC43jij/KPx1/jLDEWSC6K+s/NExiLRXJtOn7kMu6y281F5uOwh+11zidDWRP3+wobUKWOrWlUmU6JQ6etbuSVxuHJsSNjgMBkBGcQYLDn7Go/UeRRRE8l97DUVqKup9+xBzc7EZDBQ+8ohDlPB2olfpeSzyMQDa1WjHewfeI6PI3gfkmSbP0CWsyzX3t/n5kb1oEYLJZM/8MriemQiCwOx7Z9sbO2WexSLYZ+mzNsOU7hBbwRGQXZzN0tilvN7+dT449IGjsj5QF8jwqOEEegTK2sLa/PywBgejSE/HFcrcPBb0WMrJzJPsTd1Li8AWtAq6+2YZcBcYjWYVCqgiIyPZd2UqGRMTwz333INKpSIwMJDg4GDOnz9PZKT7vO67iaub0oM9e+q7U9/xYecPeaBWefvGLw5+weH0w07bJhntMQd3LqEShV3zxlKnDrmzZgH27CbJhTqmpBKZes87rM/aQ4opBW+1NxNbTKRlUMvr3sfBywd5Y+8bTm9UHhaB3pZ6jH94Pi0++hGB7+Xn1Omw2qysT1zP0YyjdA3rSrvgdre1mMkm2fjo8EcOgwH2H/xPZ3/ij/5/0LFGRxZ+8zTEnWdcDLRKdX0cSaWipEuX6waAVddIla7nW497w+5lTexqrFdyrT1KYMAp8C2BQXvzybI2pdRNgZ6n2v7meangkmydWSGxo6bE9GejmO77KIX//a/dyLlgc9Jmsl1oQ2VkXiC9II1gzxqgVlPauDGaPXsQsLtQ9GvXYomKomiIezmWv5tWQa34uc810oSvQWUSMppVa8b6AevZ/+kL2P5czYOn7LPnF1zUcGYWZxKfH8+f/f9kyZkl5JnzGNJgiPt6GUEg/3//w2/sWMR8eTZeWXFjo4BGNApoJFt/N3HHjUZFtmzZwj333ANAdnY29eqV67L4+fmR7UYDZtOmTWzaZG+wM3v2bAKuJQ1ym1CrXA8gBp1Bdn0xO2Nc+myzzdn0CO/BvjRnn6wCBV0jumD5/DGkwYPxvRKc8/f3p45/HQ6nORuguv71efL+5xgpXn9mEZ0czc6kndwbei/tQtvx7fZvZVNwk1Ii++I5Ok6bi2XePKQtWxAqNhHS6Sh4pC+P/fkYR9OOUmwtZuHphXSq2YmljyxFKSpRKBSgBy+N19+mtRWXEyerrwC7q+dA7gGGNRnGPV0+QPnlYEcMwHEP3t7g74+k0yE98ADqWbMIuM5MY2yHsSw8s5DEPOdZVrPgZoQEhbD0saXM/vV5tm76EmWpjaHHYfiVNteCwYB3RARc47u76JFFjF8/nqWnlrrsaniwZSC6IdO4VmeH9h510Jfag/UV8Sm0Ui85G829TSErC+XSpU6DmyI9He9vv8Xj2WcdUh+VQalU3hW/xxvh4TeXoCieglj0B+aCPLSKHOCqLoCI1A+uT3iNcKbWcBa1dHvPjz2GpWVLlJ07O/Wxl8LCULz99j/mOd0WozFjxgxyc+VvN4MGDaLNFV2V5cuXo1AouO8+e42Cqzdmd3Tv3p3u3bs7Pme6kAS53TT3a86J9BNOxiBAF8AT9Z6QXV8tb9d+Xr2oZ2zUWKIvRXPo8iFMFhOeKk/aVW/HkB7fkCEqoaDA/u8KczrMYeL2icTnxWOTbIR7hTOj/Qxys+XPvyKltlJGbhhJdFo0xlIjnipP2gS1kWVLlVGsAGn3bnITEhDffx/Pd99FkZaG5OFBUe/ejK8bz/6z5U2CjGYj6+PW89HOj/DT+jH/5HxS81MxqA3cH3Y/b7R/4y/PQk5mnWTe4XnkmfNo6NuQSa0m4an2RCqR0CrkRW86pQ4vycv+d2jVCr/27dHs2OHQU7JUr07u++9j7tKlfKdriNYBHE0/ypwDc7BZbeiU9lmWp9qTBn4NmNNhDpmZmQQEBPB0p2lMmRvjJC0OUNKwIdkeHnCd7+6sdrNIykliS9IW2TqFTXHd734LSxgtLovsDi3P2tJYoFcsmJNSMWZmolm3Dn9XQp4XL5K7bx+WhpUPmgcEBNwVv8cb5n//g+efRywooNXxmcTH/urU97yOTx16BPVweW/XvGdvb4Tt2/H88EOUsbHYfH0xTpqENTz8un/7v5saNa4hiV+B22I0pk+ffs3127Zt4+DBg7z22muOgcPf35+sCmX82dnZ+N3h3rg3wpv3vElmcSaH0g+Rb86nhkcNnmr0FPV865FmSuPt6LdJMibhqfbkuXbPserMKqfArL/Wn5GNR6JValnSawn70/Zz8PJBOtToQMtA926lRgGN2DBgA4fSD2G1WWkV1Mptl7uKfHbkM7YmbXX8MIylRrZd2kb76u0REGQzoaaX7RW2Kcd2MMV7L5eeKESvDOLxyIEMbTiU0yvlRXIWycKaC2uIy4tztNdML0rnkvES3hpv+kX0Iy4vjhbVWrhNB72a3cm7mbBtAmmFaQDsTN7JvrR9LH9wOQaVgSCPICf3FEB93/rlldyCQPZ336FbuhTt5s3YfHwoGD8ea8VeF9chMT+RUZtGkWwq733go/Hh9Xav80jkI84bCwI5336Lzwsv2FuFCgKlUVHkzp1b6fONiBrBofRD5JaUvwj4aHx4qtFT193XWrcuK2Pq8WLGWY4HgdoKvc/ByxdrkXHPPWxI3MCyooV4Pa5k0k6Lk8vO5unpsif7vxa1GpufH7PvnY2AQHRaNGabmVBDKLM6znJbhX89JG9vmcjhPwlBupFX+r+BI0eOsHDhQt588028vMq1g5OSkpg3bx6zZs0iJyeHt956i3nz5lUqEJ6SIteuv1OkF6aTXphOXZ+6aJVa8kryeHjVw05ZSMEewbzc+mXWJ64npSAFg9rAGL8+9NuchDUoiKIhQ/72IrmBawc6Be7LaB/cHiUih5L2UqiSUFugZSqsX2QXqms/yYtYU4Jje0+1J1NaT2FV/Cqi06JlxwszhLnMrfdSeaEQFeSU5FBdX50+EX14s8Ob173uR9c8Wl6QdQUBgYnNJ7IpaRNnc85SaitFFEQ8lB60r96e9zu970hxLvv630ys5fltz/PruV9ly7uEdmFxr8WOz7I3UIvFHif5C42efjz1Iz+e+ZGc4hz8tH78t+F/GdZw2PV3BNS7d+M9ZQrKhAQEScISFobxxRd5Pvggy84tc4giBhTAm9thbMwVkb7evcn56qsbus5/7EzDBTbJhsVmua479Z96z3fVTONafPPNN1gsFmZcyfaoV68ezzzzDGFhYXTo0IFJkyYhiiIjR478x2ROlZFktA+OjQMaO5Z9fvRzWdpqmimNpbFL+bWvfeDxmjED3dL3UWRnIwEeP/xAzuefY7kJWe3r4e6HoFFqWNxzMbs+ncCu46vpGFfKY6dAUKl566FgYk2nnLY3mo38Gvury9mNgIC/1t+l0ago1ZFamMqSM0voFNKJbjW7AfYkguXnlqMUlfSv29+RYeRK/VRCYknsEqd4hk2yz6DGNB1DgC6AEmsJU3dNJeZyDBabhXDvcN69791rCv+54+qYj+OezK7lRxzchBbYf6PsRsJsM6MW1Tdk9MwdO5K5YQPa339HMJsp6tePS0oTf6yc6aSim2mATzuqGW6sjdSqzV9qGvVvQhTE/1e9btxxx43GJ5984nbdgAEDGHAX6hZdjwt5F5i4baK9KEmSqO1dm487f0yETwTxefEu9zmUfoiYtBjam/zQLV3qaPwiYK8c9Z4+nayVK/+2ax7aYCgxaTEYS8trATxVnjxc52EskoX7xn9C980Po//xR0oDLRT37MnxGvsh7pTsWMZSo8vKdgkJD5UHOqXOZSC3IiaLiZ/P/ky3mt1YcX4Fc2LmOIzNl8e+5LUOrxGTFkNyQbLL/V1V0hpLjSw5u4T21dszafskfo/73eF2SzQm8tSGp1j38LpKaz6VEeUXxdakrbLloYa/tyeKIAh/WQlV0uspGlzeMW7n2T/JKMqQbZcSqOPIiu+o7VX7r15mFf8y/lmv7v8AJEli3JZxHEw/SFZxFtkl2RxKP8TYrWORJIkof9dib8XWYl7Y/gK2335y2SlMkZwMRUVkFWXx1r63GL1pNItOLXIrP3Cj9Kzdk/EtxhPpE0k1XTVCPELwUHkwJ2YO9/92P1N3T6Xo/q7kfP892YsWUThsGN1rdUctyt+8gvRBbs9zNPMoNQ018dPZ41NahRa90nV6qEJUUGQpYu7BuU6zkwRjAhO2TmD+sfku28I29GvoNibiq/GlwFzAwcsHZXGa2JxYJwnvyjKh+QSaBjjLa9T1qcv0dteO5d1N1Pet77K62kvtRYD2n5HVU8Xtocpo3GLO5JyRVRqDvcDrZPZJnmnyjL0dpQsu5F9gcbUU4n1gRD/oOximdwWTCiStljOmC/Rb1Y8vj3/JmgtrmLpnKkPWDbllhmNcs3FsfGQji3stRkIirTCNtMI04vPiWXx6MW9Hv+20fd/wvrSv3h6lUD5hDTOE8UqbV2gW4LoBU0FpAQnGBIY2GsqU1lOY320+TzV6yukYAN5qb0ZEjeBA2gFZW9Ky41w96KsEFf0i+rG0z1I6h3aWHTPEI4QxTcdgLDW6nOmYbWaHS/FGMKgNLOu7jCmtp9C7dm/GNh3LygdXUsNQOR/x3UDzas1lhk8lqugU2knWprWK/9/ccffUv4ECcwEzo2dyKusUZquZYou8rNhis2C2mvFQeTCzw0z6r3EtsXGotoZPnlZy0WB3r6ytDxvqwNqi1syMmcWF/PKKb6tkJTotml/O/lLpIOj1UIpKfjv3myzjyCJZ2JnsrPSqEBX82PNHlsYuZWvSVgL1gTzX/Dmqe1QnxBDCvrR9JBmTZIN7ibWE6NRolvdeDkBOSQ6eak/ySvKQkBwKqO2qt+N45nG0Sq2sh7grJCQGNxiMVqFlWptpFJUWOWTcg/RBvNjqRYI9gpEkieoe1WWxiGB9ML1q9/orjw29Ss+EFhP+0r53A2W91d/Y+wbHs46jFJR0Du3M5FaT7/SlVXGXUWU0bhKbZGPoH0OdmsUIstYtEO4d7nj7vmSSV/WW7XcqN9ZhMMqICYGPWkeQGi9vZlM2mN8qowGug8vgWqVXKSoZ0mCIUwe4NFMag9YNcikDX4bZYi+W2p2ymxn7Zzhp/lglK82r2XXGGvs3JtI30q6ndR2UopJJ2ychCAJ1vOvwcZePebPDm5hKTfho7DpbOcU5eKg8eLnNy7y862WH28tP68dgz05EjXkJMTcXa7Vq5L/yCtYKBab/djxUHrzX6b07fRlV3OVUGY2bZNPFTTLJEAkJtajGYrMgCALhXuHMvGemI8DaoloLvNXeMv18H40PJTb5wCwJcDjrmFtFz1DPWxtwHVR/EH8m/CmLF9T2rF2p/d+Ofvu6DXSiqtljOwuOL5AZqZySHBacWMB9offZe4h3X2DXBMqLJ7MoE7PN7OqQFFuLHTPx0BOaAAAUvElEQVSk5IJknt74NL8/9Du+Wl82XdzE3INzSStMQ6/U0ym0E2seXsPCUwvJK8ljOK1oP/4tlGlpjuOpTp8m87ffsIU4Z1QlGZNIL0wnyj/KrR5UFVX8W6kyGjfJkfQjFFvl7ig/rR9vtH8DrVJL59DOTql64d7hdAntwtoLax0V13qFnv82/C8xl2NcnifcO5zuNbtzLveck7Gp7VWbcc3G3dJ76lC9A5E+kRzOOOxwLQkIeGm8rrnfwlML+S32N05lyzOqytAqtET5RzG3x1wsBRa3bqeK6ao1DDX4pc8vGM1GRqwfwd60vbLtRUSnil2wB7bP5pzFS+3FK7tecXK5XTpzCQ+lB9PaTQPA98knnQwGgPLiRUo/nEX05GHU9KyJr8aXZzY9w9HMoxjNRsIMYYxsMpLhUcOv+VyqqOLfRJXRuEk6h3bmm5PfyN7KA3QB9I3o6zZ//tP7P6XFiRZsSdqCVqPl4doP069OP9Ynrudczjknf3tdn7qMbTbW4WJZcnYJRrORGoYaTG833XXLzpugxFpCnjnPKRYhIbHz0k6SC5Jd1jJ8f+p7ZsfMdinfXUZDv4a83Ppl7q95Pz5aHzILMmka0JTdKbtl214yyl14nmpP+kb05VDGISdXmVpUu+yFUGQpIt+cz+Izi2UxmlJbKdsvbXcYDdGFzM2U7vBzzT9IWvM7AboA1KLaqeo7Pj+ejw59ROeQzoR7h7u97yqq+DdRlT11k7QNbkuH6h1QCOW5/WUSINcquBIFkVFNRrGk9xJWD1xNvzr9APhPrf/wUZeP6BTSiRbVWtA3vC8/9fzJYTAG1h/IyodWsvnRzfzY80cifW+96u/F/Iukm+QSzhnFGWy+KJf+BlgWu+yaBiPEI4Qvu31J91rdEYXyr92IqBEuazqyS7LZn7JftvzJqCcZGDmQEEMIalFNqCGUwQ0Guxy0a3nVolm1ZuSV5MnWAU5uLluwc2vDtfXgi9ZwUVeChERGUYaTwSgjoyiDBccXuL3vKqr4t1E107hJBEHgmx7fsPDUQrZd2oZOqWNM0zG0CGzxl4/ZNayrrHPb7STIIwilQgkutAo3JG7giagnZMtd1UuAPXW2ebXmvNT6Jer41JGtN1lMaBVaWdpwibWE0VtGM/OemfSNKNexEgSBd+59h+zibBLzEwn3DsdH48O6C+uYsX+GI/ge5hnGpJaT0Cg0PB75OH8m/omp1LkbVYR3hOP/+dOmoTpxAmVCAgDftLA3UaoM7rrKVVHFv5Eqo3ELUIgKnmr8FE81vr5g3D8BL7UXHioPWRczgMyiTI6kH2FWzCwumy7jqfZkYORAqntUl8mjqEU1b3V4i0cjH3V7rnDvcII9gjHmymcpGUUZzDkwh241u8kCzn5aPye3XO/w3nSs0ZHfzv2GTbLxWORjjtlZxxod6V+nP+surCO7JBu1qCbSN5LZ98527G+tWZPM5cvxnDsXRWoqlvA4IOG6zypAG8DTjV1386uiin8jVUajCpd0qN7BpQifQlAwZssYpyK4uLw4nmn8DInGREchnlpU0756e/rXvXbLV5WoYlSTUcyOnu2yB3ZCXgI7k3c6Na1yh7fGm5GNR8qWC4LAu81f5bnE6qzSnKB22170jHxYJhdiCwoib84cAAZe3Mq2Lc86yaqIiIQYQigoLSCvJI+aXjUZHjXc5Qyqiir+rVQZjSpcMr75eHYk73AS/fNQeaBRakjKdK6azjfnsz15O6v7rWb+sfkkGZPoFtaN/nX7V0rHaWiDoYR7hTPsz2GyWhClqHQrM1JZNGvW4D1rFsGJibQVBCy1z5A3qxrmTp3c7tO1ZleejHqSVfGrSClIwV/nT/vg9szrOo+LxoukF6bTNKApetXNXVsVVfzTuOPS6H8Hd5M0emW4W6WU1yeu55PDn5BemI6n2pOH6jzEicwTrEtYJ9s2yi+KjY9srPSxXd3zgNUD2J/mHPxu6NeQ9f3X37CIoIOSEqr16IEqzlnapbRBAzI2bPi/9u41qIlzjQP4P0vACJb7TWqxI+AotiqKl1o9SNG2ltoLIuLQWq0arWBbkbEW63hmbB2dUel4xRmlFUbFUlGno9U6imALVY5oHYmWS0VFuUkCweGWbJ7zweMeEQ1RiYv4/D6xSfbN/2UDT/bN5n07nJZc36pHka4IfXr1gbeDt9nHdqSrHmdr4j4/O56ZqdFZ1/VW37fwVt+3YDAZoFQooVAokKpJxdGrRyFS20tcO2NG1+SwZMSeiEVJXQlEEuH7gi/W/mvt4xcMAHb/+Q+UV660u93m2jUoL1+GcZD59Zgd7RwR7BX82M/PWHfDRYN16N5LYqcPmI7DVw7jdOVp6ZJVf2d//Pu1fz/x83jaeyLj3QxUN1bDYDI81toW9yMHB5BK1X4N8B49rL6wFWPdERcN9khsBVvsmrQLB0oP4MT1E/Bz8sOcV+fA0c78t8Ufhae9Z6e1ZRgyBEZ/f9hdaDtvlzEg4JGWdGWM3cFFgz0yG8EGUwKmYErAlI4fLDeFArpt2+D85ZdQlpWBFAoY/f1Rt2mT3MkYeyZx0WDdnujri9rMTAhaLUihALm4yB2JsWcWFw323DC5du4cXYw9j3juKcYYYxbjosEYY8xiXDQYY4xZjIsGY4wxi3HRYIwxZrFuOfcUY4wx6+AzjS5g6dKlckd46rjPzwfuc/fDRYMxxpjFuGgwxhizGBeNLmDChAlyR3jquM/PB+5z98MfhDPGGLMYn2kwxhizGE9YKKO8vDxkZGTgxo0bWLVqFfz8/KT79u/fjxMnTkAQBMyaNQtDhw6VMal1/PTTTzh+/DgcHe+sxTF9+nQMGzZM5lTWcf78efzwww8wmUwICwvDBx98IHckq4uNjYVKpYIgCLCxscHq1avljtTptmzZgoKCAjg5OWHdunUAgNu3byMpKQk1NTXw8PDAokWL0KtXL5mTdiJisrl+/TrduHGDVqxYQSUlJW1uT0hIoNbWVqqqqqK4uDgSRVHGpNaxd+9eOnjwoNwxrE4URYqLi6PKykoyGAyUkJBA169flzuW1S1YsIDq6+vljmFVhYWFVFpaSvHx8dJtaWlptH//fiIi2r9/P6WlpckVzyp4eEpGffr0eeBi7vn5+RgzZgxsbW3h6ekJb29vlJSUyJCQdYaSkhJ4e3vDy8sLSqUSY8aMQX5+vtyxWCcIDAxsdxaRn5+PkJAQAEBISEi3O9Y8PNUFabVaBAQESNuurq7QarUyJrKeo0ePIicnB/369cOMGTO612n8/2i1Wri5uUnbbm5uKC4uljHR0/Pdd98BACZOnNjtryq6q76+Hi7/W+jLxcUFer1e5kSdi4uGla1cuRJ1dXXtbo+OjsaIESMeuA91owvazPX/zTffRGRkJABg7969SE1NxYIFC552RKt70PFUKBQyJHm6Vq5cCVdXV9TX1+Pbb7+Fj48PAgMD5Y7FnhAXDStbvnz5I+/j5uaG2tpaaVur1cL1GV11ztL+h4WFYc2aNVZOI4/7j2dtba30TrQ7u/uadXJywogRI1BSUvJcFA0nJyfodDq4uLhAp9NJF3p0F/yZRhcUHByM3NxcGAwGVFdXo6KiAv7+/nLH6nQ6nU76+cyZM3jppZdkTGM9fn5+qKioQHV1NYxGI3JzcxEcHCx3LKtqbm5GU1OT9POFCxfg6+src6qnIzg4GNnZ2QCA7Ozsh44oPKv4y30yOnPmDFJSUqDX6+Hg4ICXX34Zy5YtAwBkZmYiKysLgiBg5syZCAoKkjlt59u4cSPKysqgUCjg4eEBtVrdbd+BFxQUYOfOnTCZTAgNDUVERITckayqqqoKa9euBQCIooixY8d2yz5///330Gg0aGhogJOTE6KiojBixAgkJSXh1q1bcHd3R3x8fLf6rI6LBmOMMYvx8BRjjDGLcdFgjDFmMS4ajDHGLMZFgzHGmMW4aDDGGLMYFw3W6TZv3oz09HS5YzwzTp48afGXIGNjY3HhwoUOH1dYWIj58+dL2/Hx8SgsLLRqtkfx8ccfo6qqqtPbZdbHRYOx58D69esxaNAgs4+prq5GVFQURFG0ep60tDR4eXlZ/XlY5+OiwZ4JT+MfGWOsYzz3FHtiV65cQXJyMioqKhAUFNRmMr6zZ88iPT0dNTU16NOnD+bOnYu+ffsCAP755x8kJyejsrISQ4cOhUKhQO/evREdHY3CwkJs3LgRb7/9Ng4dOoTBgwdj4cKFZtvTarVISUnBpUuXoFKpEB4ejnfeecds9pKSEmzfvh0VFRWws7PD2LFj8cknn6C6uhpxcXFQq9XIyMgAEWHy5MmYPHkyAMBgMGDXrl3Iy8sDALz22muIiYmBra2tlD08PBwHDx6EIAiYPn06QkNDAQANDQ3YsmULNBoNfHx8MGTIECnP9u3bYWdnhxkzZki3rV69Gq+++irCw8PbZDeX4X6xsbGYN28eBg8e/NA+r1ixAgAwc+ZMAG3nDUtNTUVWVhbs7e0xZ84caYaCxsZG7Ny5E+fOnYNCoUBoaCiioqIgCAIqKyuxdetWlJWVQalU4pVXXsGiRYsAAFFRUdiwYQO8vb1RUFCAtLQ01NbWomfPnggPD8d7771n9rgxGcm4lgfrBgwGA3322Wf0yy+/kMFgoLy8PIqOjqY9e/ZQaWkpzZ49m4qKikgURcrKyqIFCxZQa2urtN+hQ4fIYDDQn3/+Ke1HRHTx4kWaNm0apaWlUWtrK7W0tJhtTxRFWrJkCWVkZJDBYKDKykqKjY2lc+fOmc2fmJhI2dnZRETU1NREf//9NxERVVVV0dSpUykpKYmampro6tWr9Omnn9Jff/1FRETp6emUmJhIdXV1VF9fT8uWLWuXPT09nQwGA509e5ZiYmKooaGBiIiSkpJo3bp1UrtqtZq++eYbIiIqLi4mtVotLbpVX19PMTExpNPpiOjOwkaWZpg3b57Uz3v366jPRqNR2i8rK4uio6Pp2LFjJIoiHT16lNRqNZlMJiIiWrNmDW3bto2ampqorq6Oli5dSr/99pvUz3379pEoitTS0kKXLl2S2p06dSpVVFQQEdHcuXNJo9EQEVFDQwOVlpaaPWZMXjw8xZ5IUVERRFFEeHg4lEolRo8eLS1be/z4cUyYMAEBAQEQBAHjx4+HUqlEcXGxtN+kSZOgVCoxatSodpMyKhQKREVFwdbWFnZ2dmbbKy0thV6vR2RkJJRKJby8vBAWFobc3Fyz+ZVKJSorK6HX66FSqdC/f/8290+dOhUqlQq+vr4IDQ3FH3/8AQD4/fffMWXKFDg5OcHR0RGRkZE4deqUtJ+NjY2UZdiwYVCpVLh58yZMJhNOnz6NadOmSe3eXbAHAPz9/WFvb4+LFy8CAHJzczFo0CA4Ozu3y95Rhsft8/3c3d0xYcIECIKAkJAQ6HQ61NfXo66uDufPn8fMmTOhUqng5OSE8PBw6XeuVCpRU1MDnU4HOzs7DBgw4IHt29jYoLy8HI2NjejVqxf69evXYR+YfHh4ij0RnU4HV1fXNkNS7u7uAIBbt24hOzsbR44cke4zGo3QarVQKBTt9rt3oSIAcHR0hJ2dnbRtrj1BEKDT6aShFQAwmUwYOHCg2fzz58/H3r17sWjRInh6eiIyMhLDhw9/YCZ3d3dcu3YNwJ2hMA8PD+k+Dw+PNgtlvfDCC7CxsZG2e/TogebmZuj1eoii2KZdDw8PXLp0SdoOCQlBTk4OBg8ejFOnTmHSpEkPzN5Rhsft8/3uLVg9evQAcGfm2tu3b0MURajVaul+IpL69tFHHyE9PR2JiYlwcHDAu+++izfeeKNd+4sXL0ZmZiZ2794NX19fxMTEdFjImHy4aLAn4uLiAq1WCyKSCkBtbS28vb3h5uaGiIiIB85uqtFoHrrfXfcvVGSuvaKiInh6emLDhg2PlL9379748ssvYTKZcObMGaxfvx47duyQ7q+trcWLL74I4E7RujsLr6urK2pqaqTp3G/dumXRmieOjo6wsbFp1+69xo0bh8WLF6OsrAzl5eUYOXLkA9t63AwP6/OjLgzl5uYGpVKJHTt2tCmQdzk7O0uX/V6+fBkrV65EYGBgm2MM3Dm7WrJkCYxGI44cOYKkpCRs3br1kbKwp4eHp9gT6d+/PwRBwK+//gpRFHH69GlpPfOwsDAcO3YMxcXFICI0NzejoKAATU1N0n5HjhyBKIrIz8/vcB10c+35+/ujZ8+eOHDgAFpbW2EymXDt2rUO28zJyYFer4cgCLC3twcACML//yz27duHlpYWXL9+HSdPnsSYMWMAAK+//joyMzOh1+uh1+vx888/Y9y4cR3+vgRBwMiRI5GRkYGWlhaUl5dLay/c5ebmBj8/P2zatAmjRo1qc7Z1r8fN8LA+Ozo6QqFQWPz9CRcXFwwZMgSpqalobGyEyWRCZWUlNBoNACAvL09afMrBwUF6nnsZjUacOnUKjY2NUCqVsLe3b/cY1rXwmQZ7IkqlEgkJCdi2bRvS09MRFBQkvTP28/PDvHnzkJKSIl2pM2DAAAwcOFDaLzk5Gbt370ZQUBCGDx8OpfLhL0lz7QmCgK+++gqpqamIjY2F0WiEj48Ppk2bZjb/+fPnkZqaipaWFnh4eOCLL75o8086MDAQn3/+OUwmEyZPnixd6RQREYHGxkYkJCQAAEaPHm3xehGzZ8/Gli1boFar4ePjg/Hjx7f74l1ISAg2bdrUZrjtfo+bwVyfIyIisHz5coiiiMTExA7biouLw65duxAfH4+mpiZ4eXnh/fffBwCUlpbixx9/RGNjI5ydnTFr1ix4enq2ayMnJwcpKSkwmUzw8fHBwoULO3xeJh9eT4N1GYmJiZg4caJ0aaqc7l5yu2fPngcOvVibRqPBxo0bsXnzZn7nzboUfjUy2Wg0GtTV1UEURZw8eRJXr17F0KFD5Y4lO6PRiMOHDyMsLIwLButyeHiKyebmzZtISkpCc3MzvLy8sHjxYqss97pq1ao2Vyfd9eGHH3a5JUjLy8vx9ddfo2/fvh1+MZExOfDwFGOMMYvxuS9jjDGLcdFgjDFmMS4ajDHGLMZFgzHGmMW4aDDGGLMYFw3GGGMW+y9QIRoVqVXX0gAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "from sklearn.manifold import TSNE\n",
    "model = TSNE(learning_rate=100)\n",
    "transformed = model.fit_transform(data2)\n",
    "x = transformed[:,0]\n",
    "y = transformed[:,1]\n",
    "plt.scatter(x,y,c = color_list )\n",
    "plt.ylabel('pelvic_radius')\n",
    "plt.xlabel('degree_spondylolisthesis')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "_cell_guid": "00555f09-d381-494e-8bb9-7afb768d912a",
    "_uuid": "71d32fd64448612635a015c3a4e41871bddc337d"
   },
   "source": [
    "<a id=\"16\"></a> <br>\n",
    "### PRINCIPLE COMPONENT ANALYSIS (PCA)\n",
    "* Fundemental dimension reduction technique\n",
    "* first step is decorrelation:\n",
    "    * rotates data samples to be aligned with axes\n",
    "    * shifts data samples so they have mean zero\n",
    "    * no information lost\n",
    "    * fit() : learn how to shift samples\n",
    "    * transform(): apply the learned transformation. It can also be applies test data\n",
    "* Resulting PCA features are not linearly correlated\n",
    "* Principle components: directions of variance"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 317,
   "metadata": {
    "_cell_guid": "93b827bf-616e-436f-a514-86fafbf44512",
    "_uuid": "04e33157c4c218d873892090649b169babfd03a8"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Principle components:  [[ 3.23645647e-01  1.13192291e-01  3.03674740e-01  2.10453357e-01\n",
      "  -2.99598300e-02  8.63153779e-01]\n",
      " [-4.76634849e-01 -9.85632787e-02 -5.32783979e-01 -3.78071570e-01\n",
      "   3.21809199e-01  4.82438036e-01]\n",
      " [-1.54481282e-03 -2.64657410e-01 -4.96541893e-01  2.63112598e-01\n",
      "  -7.74612852e-01  1.18940778e-01]\n",
      " [ 3.73677251e-01  7.54113757e-01 -3.39411757e-01 -3.80436506e-01\n",
      "  -1.75106042e-01 -3.29143086e-02]\n",
      " [-4.41703869e-01  7.35414748e-02  5.12024113e-01 -5.15245344e-01\n",
      "  -5.14639730e-01  8.35992525e-02]\n",
      " [ 5.77350269e-01 -5.77350269e-01 -1.08930766e-11 -5.77350269e-01\n",
      "  -3.59066811e-12  3.06727015e-12]]\n"
     ]
    }
   ],
   "source": [
    "# PCA\n",
    "from sklearn.decomposition import PCA\n",
    "model = PCA()\n",
    "model.fit(data3)\n",
    "transformed = model.transform(data3)\n",
    "print('Principle components: ',model.components_)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 318,
   "metadata": {
    "_cell_guid": "fc2fdc3e-a5f6-4d67-979e-b9e9cd2e7198",
    "_uuid": "409100f13e8c3e7bea30dab77ad7ea56e2e338e8"
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# PCA variance\n",
    "scaler = StandardScaler()\n",
    "pca = PCA()\n",
    "pipeline = make_pipeline(scaler,pca)\n",
    "pipeline.fit(data3)\n",
    "\n",
    "plt.bar(range(pca.n_components_), pca.explained_variance_)\n",
    "plt.xlabel('PCA feature')\n",
    "plt.ylabel('variance')\n",
    "plt.show()\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "_cell_guid": "3b6f44f2-0add-4f82-a965-859215a9fea9",
    "_uuid": "ec0ad8f2341933207e26f9eaced8a45e16f2bacf"
   },
   "source": [
    "* Second step: intrinsic dimension: number of feature needed to approximate the data essential idea behind dimension reduction\n",
    "* PCA identifies intrinsic dimension when samples have any number of features\n",
    "* intrinsic dimension = number of PCA feature with significant variance\n",
    "* In order to choose intrinsic dimension try all of them and find best accuracy"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 319,
   "metadata": {
    "_cell_guid": "fa5ec66c-2338-4298-a6f8-364a0703ca1f",
    "_uuid": "0f1768fd650fe0b85cdf21140c877471c8baf2bc"
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# apply PCA\n",
    "pca = PCA(n_components = 2)\n",
    "pca.fit(data3)\n",
    "transformed = pca.transform(data3)\n",
    "x = transformed[:,0]\n",
    "y = transformed[:,1]\n",
    "plt.scatter(x,y,c = color_list)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.6.5"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 1
}