Earthquake_Challenge.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "PwAdMj-gZ_Gu"
   },
   "source": [
    "## Lecture Content\n",
    "- Earthquake prediction background & helpful resources\n",
    "- Step 1 - Installing dependencies\n",
    "- Step 2 - Importing dataset\n",
    "- Step 3 - Exploratory data analysis\n",
    "- Step 4 - Feature engineering (statistical features added)\n",
    "- Step 5 - Implement \"Catboost\" model\n",
    "- Step 6 - Implement Support Vector Machine + Radial Basis Function model\n",
    "- Step 7 - Future Directions (Genetic Programming, Recurrent Networks, etc.)\n",
    "- Step 8 - Freestyle Rap\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "yutaK7z6exYj"
   },
   "source": [
    "## EarthQuake Prediction Background\n",
    "\n",
    "- Predicting earthquakes has long been thought to be near-impossible. \n",
    "- Being able to predict earthquakes could allow us to better protect human life and property.\n",
    "\n",
    "![alt text](https://i.imgur.com/rTadNbm.png)\n",
    "\n",
    "- There are many ways to compute the magnitude of an earthquake!\n",
    "- A0 is the seismometer reading produced by an Earthquake of standard size (i.e., a calibration earthquake). Generally A0 is 0.001 mm. \n",
    "- This equation assumes that a distance of 100 km separates the seismometer and the epicentre\n",
    "![alt text](http://1.bp.blogspot.com/-wX1aS8tcJ6U/Tby9LlQkLiI/AAAAAAAAAZI/zA16efEl_qc/s200/Picture1.png)\n",
    "![alt text](http://exponential-logarithm-project.weebly.com/uploads/6/5/6/0/65601499/8320544.png?1447690096)\n",
    "![alt text](https://geology.com/records/largest-earthquake/global-seismic-moment-release.gif)\n",
    "![alt text](https://www.sott.net/image/s15/315072/full/ryzproductionsfuture_blogspot_.jpg)\n",
    "\n",
    "\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "x-mw6pCMXirQ"
   },
   "source": [
    "- Given seismic signals we are asked to predict the time until the onset of laboratory earthquakes \n",
    "- In the lab, 2 plates are put under pressure, resulting in shear stress\n",
    "- The training data is a single sequence of signal and seems to come from one experiment alone.\n",
    "- The test data consists of several different sequences, called segments, that may correspond to different experiments. \n",
    "- For each test data segment with its corresponding seg_id we are asked to predict it's single time until the lab earthquake takes place."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 321
    },
    "colab_type": "code",
    "id": "-2jkuVtBfyRK",
    "outputId": "e0b0f0f7-c227-4d63-c5d3-1c8d26a45b29"
   },
   "outputs": [
    {
     "data": {
      "image/jpeg": "/9j/4AAQSkZJRgABAQAAAQABAAD/2wCEABALDA4MChAODQ4SERATGCgaGBYWGDEjJR0oOjM9PDkz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\nBCEIBCEIBCEIBCEIBCEIBCEIBCEIP//Z\n",
      "text/html": [
       "\n",
       "        <iframe\n",
       "            width=\"400\"\n",
       "            height=\"300\"\n",
       "            src=\"https://www.youtube.com/embed/m_dBwwDJ4uo\"\n",
       "            frameborder=\"0\"\n",
       "            allowfullscreen\n",
       "        ></iframe>\n",
       "        "
      ],
      "text/plain": [
       "<IPython.lib.display.YouTubeVideo at 0x7f79860e2d68>"
      ]
     },
     "execution_count": 2,
     "metadata": {
      "tags": []
     },
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from IPython.display import YouTubeVideo\n",
    "YouTubeVideo('m_dBwwDJ4uo')\n",
    "\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "he7SDXkgeqj0"
   },
   "source": [
    "## Before we begin, check out these helpful resources\n",
    "\n",
    "- https://www.kaggle.com/kernels  (For Ideas)\n",
    "- https://paperswithcode.com (For the latest ML models)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "uT-B0AyTbFVY"
   },
   "source": [
    "# Step 1 - Install & Import Dependencies"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 530
    },
    "colab_type": "code",
    "id": "QBzPAcuLoaew",
    "outputId": "56b85956-c85c-4820-c960-a6b93214b6bc"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Requirement already satisfied: kaggle in /usr/local/lib/python3.6/dist-packages (1.5.2)\n",
      "Requirement already satisfied: urllib3<1.23.0,>=1.15 in /usr/local/lib/python3.6/dist-packages (from kaggle) (1.22)\n",
      "Requirement already satisfied: six>=1.10 in /usr/local/lib/python3.6/dist-packages (from kaggle) (1.11.0)\n",
      "Requirement already satisfied: certifi in /usr/local/lib/python3.6/dist-packages (from kaggle) (2018.11.29)\n",
      "Requirement already satisfied: python-dateutil in /usr/local/lib/python3.6/dist-packages (from kaggle) (2.5.3)\n",
      "Requirement already satisfied: requests in /usr/local/lib/python3.6/dist-packages (from kaggle) (2.18.4)\n",
      "Requirement already satisfied: tqdm in /usr/local/lib/python3.6/dist-packages (from kaggle) (4.28.1)\n",
      "Requirement already satisfied: python-slugify in /usr/local/lib/python3.6/dist-packages (from kaggle) (2.0.1)\n",
      "Requirement already satisfied: chardet<3.1.0,>=3.0.2 in /usr/local/lib/python3.6/dist-packages (from requests->kaggle) (3.0.4)\n",
      "Requirement already satisfied: idna<2.7,>=2.5 in /usr/local/lib/python3.6/dist-packages (from requests->kaggle) (2.6)\n",
      "Requirement already satisfied: Unidecode>=0.04.16 in /usr/local/lib/python3.6/dist-packages (from python-slugify->kaggle) (1.0.23)\n",
      "Requirement already satisfied: numpy==1.15.0 in /usr/local/lib/python3.6/dist-packages (1.15.0)\n",
      "Collecting catboost\n",
      "\u001b[?25l  Downloading https://files.pythonhosted.org/packages/98/03/777a0e1c12571a7f3320a4fa6d5f123dba2dd7c0bca34f4f698a6396eb48/catboost-0.12.2-cp36-none-manylinux1_x86_64.whl (55.5MB)\n",
      "\u001b[K    100% |████████████████████████████████| 55.5MB 657kB/s \n",
      "\u001b[?25hRequirement already satisfied: numpy>=1.11.1 in /usr/local/lib/python3.6/dist-packages (from catboost) (1.15.0)\n",
      "Collecting enum34 (from catboost)\n",
      "  Downloading https://files.pythonhosted.org/packages/af/42/cb9355df32c69b553e72a2e28daee25d1611d2c0d9c272aa1d34204205b2/enum34-1.1.6-py3-none-any.whl\n",
      "Requirement already satisfied: pandas>=0.19.1 in /usr/local/lib/python3.6/dist-packages (from catboost) (0.22.0)\n",
      "Requirement already satisfied: six in /usr/local/lib/python3.6/dist-packages (from catboost) (1.11.0)\n",
      "Requirement already satisfied: pytz>=2011k in /usr/local/lib/python3.6/dist-packages (from pandas>=0.19.1->catboost) (2018.9)\n",
      "Requirement already satisfied: python-dateutil>=2 in /usr/local/lib/python3.6/dist-packages (from pandas>=0.19.1->catboost) (2.5.3)\n",
      "Installing collected packages: enum34, catboost\n",
      "Successfully installed catboost-0.12.2 enum34-1.1.6\n"
     ]
    },
    {
     "data": {
      "application/vnd.colab-display-data+json": {
       "pip_warning": {
        "packages": [
         "enum"
        ]
       }
      }
     },
     "metadata": {
      "tags": []
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "#to access kaggle datasets\n",
    "!pip install kaggle\n",
    "#Math operations\n",
    "!pip install numpy==1.15.0\n",
    "#Machine learning\n",
    "!pip install catboost"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 0,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "reHWaDX-pGlx"
   },
   "outputs": [],
   "source": [
    "#data preprocessing\n",
    "import pandas as pd\n",
    "#math operations\n",
    "import numpy as np\n",
    "#machine learning\n",
    "from catboost import CatBoostRegressor, Pool\n",
    "#data scaling\n",
    "from sklearn.preprocessing import StandardScaler\n",
    "#hyperparameter optimization\n",
    "from sklearn.model_selection import GridSearchCV\n",
    "#support vector machine model\n",
    "from sklearn.svm import NuSVR, SVR\n",
    "#kernel ridge model\n",
    "from sklearn.kernel_ridge import KernelRidge\n",
    "#data visualization\n",
    "import matplotlib.pyplot as plt"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "QMtu1W7ebhRD"
   },
   "source": [
    "# Step 2 - Import Dataset from Kaggle"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 89,
     "resources": {
      "http://localhost:8080/nbextensions/google.colab/files.js": {
       "data": "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",
       "headers": [
        [
         "content-type",
         "application/javascript"
        ]
       ],
       "ok": true,
       "status": 200,
       "status_text": ""
      }
     }
    },
    "colab_type": "code",
    "id": "KvuWAvb9obkg",
    "outputId": "6447ade6-f289-418b-becf-e8c61f0e9d57"
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "\n",
       "     <input type=\"file\" id=\"files-0f578877-5618-42f1-8d52-bafe9a1e4663\" name=\"files[]\" multiple disabled />\n",
       "     <output id=\"result-0f578877-5618-42f1-8d52-bafe9a1e4663\">\n",
       "      Upload widget is only available when the cell has been executed in the\n",
       "      current browser session. Please rerun this cell to enable.\n",
       "      </output>\n",
       "      <script src=\"/nbextensions/google.colab/files.js\"></script> "
      ],
      "text/plain": [
       "<IPython.core.display.HTML object>"
      ]
     },
     "metadata": {
      "tags": []
     },
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Saving kaggle.json to kaggle.json\n",
      "User uploaded file \"kaggle.json\" with length 62 bytes\n"
     ]
    }
   ],
   "source": [
    "# Colab's file access feature\n",
    "from google.colab import files\n",
    "\n",
    "#retrieve uploaded file\n",
    "uploaded = files.upload()\n",
    "\n",
    "#print results\n",
    "for fn in uploaded.keys():\n",
    "  print('User uploaded file \"{name}\" with length {length} bytes'.format(\n",
    "      name=fn, length=len(uploaded[fn])))\n",
    "  \n",
    "# Then move kaggle.json into the folder where the API expects to find it.\n",
    "!mkdir -p ~/.kaggle/ && mv kaggle.json ~/.kaggle/ && chmod 600 ~/.kaggle/kaggle.json"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 391
    },
    "colab_type": "code",
    "id": "sGcveM2up-04",
    "outputId": "b0688bfa-3225-4d18-e24d-2aafb8af766a"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "ref                                            deadline             category            reward  teamCount  userHasEntered  \n",
      "---------------------------------------------  -------------------  ---------------  ---------  ---------  --------------  \n",
      "digit-recognizer                               2030-01-01 00:00:00  Getting Started  Knowledge       2625           False  \n",
      "titanic                                        2030-01-01 00:00:00  Getting Started  Knowledge       9896           False  \n",
      "house-prices-advanced-regression-techniques    2030-01-01 00:00:00  Getting Started  Knowledge       4145           False  \n",
      "imagenet-object-localization-challenge         2029-12-31 07:00:00  Research         Knowledge         33           False  \n",
      "competitive-data-science-predict-future-sales  2019-12-31 23:59:00  Playground           Kudos       2287           False  \n",
      "two-sigma-financial-news                       2019-07-15 23:59:00  Featured          $100,000       2897           False  \n",
      "LANL-Earthquake-Prediction                     2019-06-03 23:59:00  Research           $50,000        991            True  \n",
      "tmdb-box-office-prediction                     2019-05-30 23:59:00  Playground       Knowledge         31           False  \n",
      "dont-overfit-ii                                2019-05-07 23:59:00  Playground            Swag         51           False  \n",
      "gendered-pronoun-resolution                    2019-04-22 23:59:00  Research           $25,000        101           False  \n",
      "histopathologic-cancer-detection               2019-03-30 23:59:00  Playground       Knowledge        568           False  \n",
      "petfinder-adoption-prediction                  2019-03-28 23:59:00  Featured           $25,000       1011           False  \n",
      "vsb-power-line-fault-detection                 2019-03-21 23:59:00  Featured           $25,000        817           False  \n",
      "microsoft-malware-prediction                   2019-03-13 23:59:00  Research           $25,000       1540           False  \n",
      "humpback-whale-identification                  2019-02-28 23:59:00  Featured           $25,000       1769           False  \n",
      "elo-merchant-category-recommendation           2019-02-26 23:59:00  Featured           $50,000       3583           False  \n",
      "quora-insincere-questions-classification       2019-02-26 23:59:00  Featured           $25,000       4037           False  \n",
      "ga-customer-revenue-prediction                 2019-02-15 23:59:00  Featured           $45,000       1104           False  \n",
      "reducing-commercial-aviation-fatalities        2019-02-12 23:59:00  Playground            Swag        157           False  \n",
      "pubg-finish-placement-prediction               2019-01-30 23:59:00  Playground            Swag       1534           False  \n"
     ]
    }
   ],
   "source": [
    "#list competitions\n",
    "!kaggle competitions list"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 170
    },
    "colab_type": "code",
    "id": "zHB6DdwrqAav",
    "outputId": "07c84fd9-e61c-4cf2-cf4b-0a0cf09e65ee"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Downloading sample_submission.csv to /content\n",
      "\r",
      "  0% 0.00/33.3k [00:00<?, ?B/s]\n",
      "100% 33.3k/33.3k [00:00<00:00, 13.0MB/s]\n",
      "Downloading test.zip to /content\n",
      " 95% 231M/242M [00:01<00:00, 148MB/s]\n",
      "100% 242M/242M [00:01<00:00, 159MB/s]\n",
      "Downloading train.csv.zip to /content\n",
      "100% 2.02G/2.03G [00:36<00:00, 24.4MB/s]\n",
      "100% 2.03G/2.03G [00:36<00:00, 59.9MB/s]\n"
     ]
    }
   ],
   "source": [
    "#download earthquake data, will take 30-60 seconds\n",
    "!kaggle competitions download -c LANL-Earthquake-Prediction"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 85
    },
    "colab_type": "code",
    "id": "oxOrB-axqCOw",
    "outputId": "7d6b03f8-7e36-4131-bc7d-08668e08acfa"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "sample_data  sample_submission.csv  test.zip  train.csv.zip\n",
      "Archive:  train.csv.zip\n",
      "  inflating: train.csv               \n",
      "sample_data  sample_submission.csv  test.zip  train.csv  train.csv.zip\n"
     ]
    }
   ],
   "source": [
    "#unzip training data for usage, will take about 5 minutes (its big)\n",
    "!ls\n",
    "!unzip train.csv.zip\n",
    "!ls"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "afvCmHVDcPe-"
   },
   "source": [
    "# Step 3 - Exploratory Data Analysis"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 0,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "IPTChUFWpJKY"
   },
   "outputs": [],
   "source": [
    "#Extract training data into a dataframe for further manipulation\n",
    "train = pd.read_csv('train.csv', nrows=6000000, dtype={'acoustic_data': np.int16, 'time_to_failure': np.float64})"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 359
    },
    "colab_type": "code",
    "id": "oUXVLqoSxt01",
    "outputId": "56525b15-0519-420f-ff3a-4db8ff958df3"
   },
   "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>acoustic_data</th>\n",
       "      <th>time_to_failure</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>12</td>\n",
       "      <td>1.4691</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>6</td>\n",
       "      <td>1.4691</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>8</td>\n",
       "      <td>1.4691</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>5</td>\n",
       "      <td>1.4691</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>8</td>\n",
       "      <td>1.4691</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>8</td>\n",
       "      <td>1.4691</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>9</td>\n",
       "      <td>1.4691</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>7</td>\n",
       "      <td>1.4691</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>-5</td>\n",
       "      <td>1.4691</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>3</td>\n",
       "      <td>1.4691</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   acoustic_data  time_to_failure\n",
       "0             12           1.4691\n",
       "1              6           1.4691\n",
       "2              8           1.4691\n",
       "3              5           1.4691\n",
       "4              8           1.4691\n",
       "5              8           1.4691\n",
       "6              9           1.4691\n",
       "7              7           1.4691\n",
       "8             -5           1.4691\n",
       "9              3           1.4691"
      ]
     },
     "execution_count": 24,
     "metadata": {
      "tags": []
     },
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#print first 10 entries\n",
    "train_df.head(10)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 498
    },
    "colab_type": "code",
    "id": "YBkgDQ0by_Q9",
    "outputId": "0ec28b3c-dcb2-4df5-f11b-115201755c0f"
   },
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAv0AAAHhCAYAAAD9OgFpAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz\nAAALEgAACxIB0t1+/AAAADl0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uIDMuMC4yLCBo\ndHRwOi8vbWF0cGxvdGxpYi5vcmcvOIA7rQAAIABJREFUeJzs3Xd4FEUfB/Dv1fSElAsI0ixIValS\npYYkFA28SO9FkC4dBClBEaQJ0kREFJASAQERkCIC0kMXBOkEyF16L3c37x9HNjlyKWCSO47v53l4\nuNud3Znd2ez9dnZ2ViaEECAiIiIiIrslt3YBiIiIiIiocDHoJyIiIiKycwz6iYiIiIjsHIN+IiIi\nIiI7x6CfiIiIiMjOMegnIiIiIrJzDPqJCADQuXNnvPfee1bJ+/Dhw3jw4AEAYN68efjpp5+eeV2f\nfPIJFi9enGe6TZs2PXMe/8Xp06fRrFkzi/OylqlXr164fPlyoZThWbZ93759aNiwIaZOnZprunHj\nxuHAgQO4f/8+Kleu/KxFfGqJiYkYPXp0tjwPHDiAli1bol27drhz5440/d69e+jcuTMMBkORlTEv\nJ06cgJ+f31MvV7lyZdy/f79A0kVERGD//v1PXQYisn0M+okI165dg5ubG0qWLImzZ88Wef7ff/+9\nFPSPHj0aXbp0KdT8DAYD5syZU6h5PK0ny7RmzRpUqVKl0PPJrwMHDqBDhw6YPn16runmzJmT40VN\nYercuTNKlSqVbfqCBQvw448/on///vj++++l6Z999hnGjx8PhUJRhKW0fSdOnMCBAwesXQwiKgQM\n+okIW7duRUBAANq0aYNt27aZzdu2bRv8/f3h7++PsWPHIi0tDQDw22+/oU2bNggICEDPnj1x9+5d\nAMCECROwdOlSafms39euXYvAwEAEBASgQ4cOuH79OhYuXIjjx49j7Nix2LVrl1n6S5cuoX379vD3\n90f37t1x7969bGWPjo5G37590axZM3z44YeIj4+X5p09exbt27dHQEAAWrVqhb/++gsA0KdPH8TH\nxyMgIAD37t3DzZs30aVLFwQGBsLPzw87d+60uJ8iIiLQr18/BAQEoFmzZli9erU0r1mzZtiwYQM6\ndOiAhg0b4osvvpDmLV26FI0bN0ZQUJBUhic9WaZmzZrh9OnTuH//Pho2bIiVK1dK9XDu3Dl8+OGH\naNSoESZOnCitY9++fWjbti2aN2+Ovn37IioqKs98Hjx4gH79+sHf399i/QOmC5A9e/Zgw4YNmDx5\nMoxGI6ZPnw5/f380a9YMY8eORXp6OgCgR48e+OWXX8yWX7x4MT755BOL33v06IEFCxYgMDAQoaGh\niIuLw9ixY+Hv74/mzZvj559/lpbLuItgyYwZM9CxY8ds0xMSElC8eHFUqlRJaunft28fvLy8UL16\ndYvrAkwXwp06dULr1q3RsmVLrF27FgCQnJyMkSNHSts+e/ZsaZkePXrgm2++QadOnVC3bl2sW7cO\nS5culY6/jOO3WbNmWLlyJdq3b4+6deti4cKF2fJPS0vDzJkzpXyWL18uzTt06BD8/PwQGBiIb7/9\nNsdtyC3dkiVL4O/vjxYtWmDgwIGIi4vD5cuXMWPGDOzZswcff/wxAGDz5s0IDAxEy5Yt0a1bN4SF\nheWYHxHZOEFELzS9Xi+aN28u4uPjRVJSkmjSpIlITU0VQghx7949UbduXfHo0SNhNBrFkCFDxMqV\nK0VYWJioWbOmuH37thBCiFWrVolevXoJIYQYP368WLJkibT+jO/x8fGiVq1aIj4+XgghxK5du8Q3\n33wjhBCiadOm4tSpU9mW9/PzE3/88YcQQojVq1eLAQMGZCv/7NmzxahRo6TyVq9eXSxatEgIIUSb\nNm3Ezp07hRBCbN26VbRo0UJKV6lSJWkdAwcOFCtWrBBCCHHy5Enx5ptvirS0tGx5zZgxQ3z66adC\nCCHu3r0rqlSpIh48eCBtw6hRo4RerxePHj0SVapUEQ8fPhTXr18XtWvXFjqdTuj1ejF48GDRtGnT\nbOt+skwZ++TevXuicuXKYuvWrUIIIYYNGyaaNGkiIiMjRVRUlKhataq4c+eOuHv3rqhevbr4559/\nhBBCLF++XAwbNizPfPr27SuWL18uhBDi/v37ombNmuLevXvZlstaL7t37xZt2rQRaWlpIiUlRQQG\nBopt27YJIYTo3r272LZtm1k+ixYtEpMmTZLWlfV79+7dRd++fYXBYBBCCDFx4kQxbtw4YTAYRGRk\npGjcuLG0TXl5ctsy9qMQQly/fl3069dPJCcni3bt2omzZ8+KIUOGiCFDhoi7d+9mW9ewYcPEli1b\nhBBCREZGio8++kikpqaKVatWif79+wuj0ShiYmJEnTp1pGO3e/fuon///iI9PV0cOHBAvPXWW+Ln\nn3+W1rdgwQKpTIMHDxZ6vV5ERESI2rVriytXrojjx49Lx+jXX38tevXqJVJTU0ViYqIICgoSBw4c\nEHq9XjRo0EAcPnxYCGH626tQoUK2Osst3cWLF0W9evVEfHy8MBgMonfv3lLdZq2biIgIUbVqVfHw\n4UMhhBATJkwwq0cier6wpZ/oBXfkyBFUq1YNrq6ucHJyQp06dXDw4EEAwNGjR1G9enUUL14cMpkM\n8+bNQ+/evXH06FG88847KFu2LADggw8+wIkTJ6DX63PMx8HBATKZDCEhIYiIiEBgYCAGDBiQY/pb\nt24hOjoajRs3BgB0797dYl/906dPIzAwEADw8ssvo06dOtK8bdu2SfNq1qxp8U4BYGqJ79evn5Qu\nNTUVOp0uW7rJkydjypQpAIDSpUtDo9GY9ZFu27YtFAoFihcvDm9vbzx8+BCnTp1C7dq14ePjA4VC\n8UzPTej1egQEBAAAKlSogGrVqsHLywuenp7QaDTQarX4888/UadOHVSoUAGAqbvLgQMHcu2znp6e\njr/++gtdu3YFAJQqVQrvvPMOjh8/nmt5/P398fPPP0OlUsHBwQHVqlXLcd/mR+PGjSGXm36ODh48\niJ49e0Iul8PLywt+fn7Yu3fvM6/b19cXt27dwsmTJ1GlShUsW7YMHTt2xI8//oi+ffuif//+Fo8r\nb29v7NmzB5cvX4anpyeWLl0KtVqNvn37YunSpZDJZPDw8MDrr79udgw0bdoUSqUSFSpUQHJyMvz9\n/QGY6k2r1UrpgoKCoFAo4O3tjZo1ayI0NNQs/4MHD6Jr165Qq9VwdnbG+++/j7179+L27dtIS0tD\nw4YNAQDt2rWzuN25patatSr++OMPuLq6Qi6Xo3r16hbrz9vbG2fOnEGJEiUAALVq1fpP9UxE1qW0\ndgGIyLq2bNmCP//8E7Vq1QJg6vMdGxsLf39/REdHw93dXUrr4OAAANmmu7m5QQiB6OjoHPNRqVT4\n/vvvsXz5cixevBhvvPEGpk6dijfeeMNi+ujoaLi5uUnflUollMrsp6zY2FizdFnLtWPHDvzwww9I\nTEyE0WiEEMJiXocPH8ayZcsQHR0NmUwGIQSMRmO2dBcvXsS8efPw8OFDyOVy6HQ6s3Surq7SZ4VC\nIe3LnMqXXwqFAo6OjgAAuVwOZ2fnbPnEx8fj9OnT0sVBRnliYmLg7e1tcb0xMTEQQmQrn6VuQVlF\nRUUhODgYf//9N2QyGSIiItCrV6+n3q4MHh4e0uf4+HiMHDlS6mufmppqtk1Pa/z48Rg5ciQ8PT0x\nbNgwzJ07F+vWrcP333+PqlWrQgiBS5cuZVtuzJgxWLFiBUaOHInU1FQMHDgQ3bp1w+3bt/HFF1/g\n5s2bkMvlePToEdq3by8t5+LiAgBS+TO+y+Vys2Ml6zZ7eHggLi7OLP/4+HjMmjUL8+fPB2Dq7vPm\nm28iNjbW7DjLup6sckuXnJyMWbNm4cSJE1LaJk2aZFuHwWDAokWLpIvHxMRElC9f3mJ+RGT7GPQT\nvcBiY2Nx8uRJnDhxAmq1GoCpVblx48aIioqCp6en2YO9CQkJSElJgbe3t9n02NhYyOVyeHp6Zgtu\nYmNjpc+VK1fGokWLkJaWhm+//RZTp07Fhg0bLJbN09MTMTExMBqNkMvlSE9PR3h4OF5++WWzdO7u\n7mb9+KOiolC6dGmEh4dj8uTJ2Lx5MypVqoTbt29Lra5ZpaenY+TIkVi4cCEaN24sBVeWjB07Fr16\n9UKXLl0gk8nQqFGj3HavxfLldmH0X/j6+qJ+/fpYtGhRvpfJqK/Y2FgpKMztIiHDggULoFQqsWPH\nDqjVaowePTrX9LkdE5a2Y8mSJdIdi/+qevXq0jMGAwYMwMSJEyGXy6ULQCGExbshLi4uGDVqFEaN\nGoULFy5gwIABqF+/PoKDg1GlShUsWbIECoUCnTt3fqZyZT0OYmJisgXvvr6+6Nu3L5o2bWo2/caN\nG0hISJC+53SB5uHhkWO6NWvW4Pbt29iyZQtcXFywYMEChIeHZ1vHrl27cODAAaxduxZeXl7YtGkT\nduzY8XQbSkQ2g917iF5gv/76K+rWrSsF/ICpRb1hw4bYuXMnGjdujNDQUNy/fx9CCEydOhUhISFo\n0KABTp8+Ld3q37BhAxo0aAClUgmNRoOrV68CMA2LmNFt4Z9//sHw4cORlpYGtVqNqlWrQiaTSXlm\nDYwBoFy5cihRooTUtSMkJASffvpptm14++23sW/fPgDA3bt3cebMGQCmIMfZ2RmvvPIK9Ho9Nm7c\nCMA0tKNKpYLRaERCQgKSk5ORlJSEqlWrAjAFRCqVCklJSdnyioyMlMq9detWadncVK9eHWfOnEFU\nVBQMBgO2b99uMV3WMj2Lhg0bmtXJhQsXMHPmzFzzyajrjH1z9+5dnD59GvXr1881r8jISFSoUAFq\ntRpXr17F2bNnc90Pvr6+uHbtGoxGI6KiovDnn3/mmDbjgWjAdAH6+eefF8jQpbt370bJkiWlC7pX\nX30VFy5cwPnz5y3ebRo0aBCuX78OwNQ1x9XVFTKZDJGRkahUqRIUCgWOHj2KO3fu5HkMWLJr1y4Y\njUZEREQgNDRUutOWoXnz5ti8eTMMBgOEEFi6dCn+/PNPlClTBgqFQmql37Jli/R3lFVu6SIjI/HK\nK6/AxcUFYWFhOHTokLQNWf8WIyMjUapUKXh5eSE6Ohq//fYbEhMTn3pbicg2MOgneoFt27YNLVq0\nyDbdz88P27ZtQ4kSJTBjxgz06tVLaiXv06cPSpQogZkzZ2Lw4MEICAjAqVOnMGPGDABAx44dERYW\nhpYtW2LevHlmfZpffvlltGnTBq1bt8bXX38tjeDi7++PUaNGmY2GI5PJ8NVXX2H58uVo2bIldu7c\niWnTpmUr68CBAxEWFoZmzZohODgYLVu2BABUrFgR7777Lvz9/dGpUyc0a9YMb7/9Nnr06AGNRoOa\nNWuiadOm+Pfff9G/f38EBQUhKCgIZcqUQYsWLTBo0KBswdyIESMwZMgQtG3bFklJSejUqROmTJki\njVxkSaVKldC5c2e0a9cO7du3R40aNSymy1qmJ/t354evry+Cg4MxZMgQBAYGYsaMGWjVqlWe+Uyf\nPh0nTpxAQEAAhgwZgpkzZ+Kll17KNa++fftiw4YNCAwMxLp16zB+/Hhs3rwZv/32m8X0AQEBcHZ2\nRosWLTBu3Lhcu+uMHDkS8fHx8Pf3R+vWrWE0GqWgPKfRey5fvoyAgAD07t0bBoMBAQEBZnkkJSVh\nxYoVGDVqlDRt8ODBmDJlCmbMmIGPPvoo2zq7d++O0aNHIzAwEO3atUPXrl1Rrlw5fPTRR5g9ezba\ntGmDkydPYujQoVi8eLF0sZlfr7/+Ojp06IDWrVujR48eeP31183md+3aFSVLlkTr1q0REBCAGzdu\noGbNmlCpVAgODsakSZMQGBgImUxm1t0rQ27pOnfujFOnTsHf3x+zZ8/GhAkTcOzYMXz//fdo0KAB\njh8/jv/9739o06YNYmJi4Ofnh9GjR2PkyJF49OiR2chURPT8kImcOrkSERFRgWvWrBnmzJmTrXWf\niKgwsaWfiIiIiMjOMegnIiIiIrJz7N5DRERERGTn2NJPRERERGTnGPQTEREREdm5F+LlXDpdfN6J\nCoGnpzOio59+/GYqPKwT28R6sT2sE9vEerE9rBPbZK160Wjc8kxz7do1DB48GL1790b37t3x8OFD\nTJw4EXq9HkqlEl9++SU0Gk2Bl40t/YVIqVRYuwj0BNaJbWK92B7WiW1ivdge1oltstV6SUpKQnBw\nMOrVqydNW7hwITp27Ii1a9fCz8/P7J01BYlBPxERERFREVCr1Vi5ciV8fX2laVOnTpVeZOnp6YmY\nmJhCyZtBPxERERFREVAqlXB0dDSb5uzsDIVCAYPBgPXr16Nt27aFk3ehrNXGeHo6W+02T376dlHR\nYp3YJtaL7WGd2CbWi+1hndim56leDAYDxo0bh7p165p1/SlIL0TQb60HbDQaN6s9REyWsU5sE+vF\n9rBObBPrxfawTmyTterlWS80Jk6ciLJly2Lo0KEFXKJM7N5DRERERGQl27dvh0qlwvDhwws1nxei\npZ+IiIiIyNouXbqE2bNnIywsDEqlEnv27EFkZCQcHBzQo0cPAMCrr76KadOmFXjeDPpfEOfOhaJs\n2XLw9PTChAmj8MUX859q+cmTx6F9+46oUaOWxflHjhzCO+/Uh0qlKojiEhEREdmdqlWr4scff7RK\n3uze84L49dftiI6OAoCnDvjzY8OGdUhPTy/w9RIRERHRf8eWfitKTEzA9OmTkZycjJSUFHz88VhU\nrlwVp04dx4oVSyGXy9GiRUt07NgVoaGn8c03S6FUKqHR+GLixE+xb98e3Lx5A0OHjkRSUhJ69uyE\nkJAdWLv2exw6dBByuRwNGjRCpUqVcfjwH7h16yZmzpyDfv264ddf9+PatauYN2825HIZqlZ9C0OG\njDAr37p1a7Bv3x6UKPESEhMTAQBabTiCgz8FAOj1ekyePB0XL57H339fwpgxw/HVV8uwfPli/P33\nZaSlpSEo6H9o2zaoyPctEREREWVi0P+Yy7TJcNixrWBX2qkjMO7THGdHRkaiTZsgvPtuE5w5cwrr\n1q3BzJlzMG/ebCxb9h3c3d0xceJovP9+e8ydOwsLFixB8eIlMH/+bPz++27IZDKL692wYS22bdsN\nhUKBbdt+Ru3adfHaaxUwatQ4lChRQkq3cOFcjB07Ca+99jqCgz/Fo0cPUaLESwCA+Ph4bN0agnXr\nQmAw6NGxY9DjMkegT58BqFGjFnbu/AVbtmzGsGEf49tvl2Pu3EUwGo0oUaIkhg0bhdTUFHTsGMSg\nn4iIiMjKGPRbkZeXN9as+RY//fQj0tPT4ejoiJiYaKjVanh6egIA5sxZiLi4WMhkMhQvbgrYa9So\nhXPnQlGhQkWL623SpDlGjhwMP78AtGwZkGP+d+/ewWuvvQ4AmDJlhtm8sLB7KF/+FTg4OABwwBtv\nVJLKvHDhXKxatQLx8XHS9AwODg6Ii4vFoEF9oVQqERMT/Uz7hoiIiIgKDoP+xxKnzUTitJkFuk6N\nxg3IZYzYTZvWw8fHF1OmBOPq1b/x9dcLIZfLYTSKJ1LKIETmtPT0dMhkcrOWfr1eL30eM2Yi7ty5\njQMHfsewYQPxzTdrLOYvl+f8SIcQAjKZPMt3IwBg1aoVeOeduggK6oCDB/fhr7+OmC139uwZhIae\nxtdffwOlUgk/v0Y55kFERERERYMP8lpRbGwMSpV6GQBw6NBB6PV6eHgUg9FogE6nhRAC48aNhEwm\ng0wmw6NHjwCYRuKpWLESnJ1dEBkZAQC4cOEcACAhIQGrV69E2bLl0KfPALi5eSApKRFyuRwGg8Es\n/3LlyuPy5UsAgFmzZuD27VvSvFKlXsadO7eQnp6OxMQE/PPPFQBATIypzEIIHDlySHp4VyYzrT82\nNga+vsWhVCpx5MghGAxGPuBLREREZGUM+q0oIKA1Nm5ch48/HoIqVaoiMjISv/66HaNHT8DkyeMx\naFBf1KxZG25ubhg3bjKmT/8EQ4d+CL1ej+bNW6JWrdq4e/cOhg79EHfv3oZMJoerqytiYqIxYEBP\nDB8+CFWqVIW7uwfefrsGJk8ej5s3b0j5jxgxBl9/vQAffdQPbm7uKFeuvDTP3d0DgYFtMHBgH8ya\nFYyKFasAAN5/vz0WLPgSo0cPR/Pm/jh3LhQnTx5H9eo1MHhwP1SsWBn379/F0KEfIizsPurXb4i5\nc2cV+b4lIiIiokwykbXfiJ2y1uux+Wpu28M6sU2sF9vDOrFNrBfbwzqxTdaqF43GrcjzzC+29BMR\nERER2TkG/URERFTwhIAsIsLapaAiEhsLXL8uR2SkDPbfh+T5xNF7iIiIqMC5jhoGp3U/IGr/ERiq\nvWnt4lAha97cBXfvmtqSFQoBb2/Tv+LFTf/7+gr4+Aj4+BhRvLiAl5eARmOa5uBg5cK/IBj0ExER\nUYFzWvcDAEB18jiD/hfAw4emYcQDAtIRGSmHVivD/ftyXLli+UWiWbm5CRQvboS3d+aFQNYLBo1G\nwNfXCC8vAQ8PIId3k1IeGPQTERER0X8ilwM1axrwww8pZtOTkwGdTobISBnCw2WIipJBq5UjIkKG\niAgZtFrTPJ1Ohps35TAac4/oHR0z7xJoNBl3EYzS3YSM6e++W5hb+3xi0E9EREREhcLJCShTRqBM\nmbw7+uv1QEyM6eIg40Ig44JBq5WbTbt+XY7z53O+QAgOBgYOLMgtef4x6CciIiIiq1Mq8bjff/6e\nBI6Lg3THQKcz3T0ID5chPl6GoCB1IZf2+cOgn4iIiIieO+7ugLu7wCuvCABGs3kajRo6nXXKZas4\nZCcRERERkZ1j0E9EREREZOcY9BMRERER2TkG/UREREREdo5BPxERERGRnWPQb2V//LEfALBr1w4c\nOnSwQNb56NEj/P33pXyljYmJQffuHbF8+dcW50dGRmDOnM8AAB06tEVSUlKBlJGIiIiIig6Dfit6\n+PAB9u3bAwBo1aotGjduWiDrDQ09hStXLucr7e3bN1G6dGkMGjTU4nxvbx+MG/dJgZSLiIiIiKyD\n4/Rb0fz5s3HlymWsXr0SRqMRxYoVQ/nyr2Lz5g1QKBS4du0qevbsixMnjuH69X8wePAIvPtuExw6\ndAAbNqyFQqHEG29UwrBhH0vrjI6OxnfffQOlUonixUvgpZdKYf782ZDJZHB2dsHkydPg7u4hpV+0\naD602kdYvvxrNG/eEvPnz4ZSqYRcLkdw8BdITEzE5MnjsWrVj9Iyn302DU2aNEeDBo1w9Ohh/PHH\nfvTt+yFmzJgCJydn/O9/HeHq6ooVK5ZAqVTC17c4xo+fDJVKVaT7l4iIrE/95x9I6fehtYtB9MKz\nStB/7do1DB48GL1790b37t3x8OFDjBs3DgaDARqNBl9++SXUajW2b9+ONWvWQC6Xo2PHjvjggw+Q\nnp6OCRMm4MGDB1AoFJg1axZKly79n8s0bZoDduwo2N3RqRMwblzO87t06YEtWzahT58BWLVqhTT9\n33+vYd26EJw/H4rp06dg8+btuHz5In7+eSNq1aqDNWtWYfny1VCr1ZgyZQIuXDiHN998GwDg6emJ\nwMA2KFasGBo2bIzhwwdh8OARqFKlKtav/xGbN29Av36Z76UeOnQktmzZhEGDhuLUqeP4+OOxqFCh\nIr79djn27v0NDRq8m+/tvX79H/z88054eBRDnz5d8dVXy+Du7oGlS7/CwYP70LJl4NPvRCIieq45\n/LbT2kUgIlgh6E9KSkJwcDDq1asnTVu0aBG6du2KwMBAzJ8/HyEhIQgKCsKSJUsQEhIClUqFDh06\nwM/PDwcPHoS7uzvmzZuHI0eOYN68eVi4cGFRb0aheu2116FWq+Ht7YPSpcvAyckJXl5eSEhIwK1b\nNxEe/gijRpm64yQmJuDRo0d4803L67p9+xaqVKkKAKhRoxZWr/4mx3w9Pb2xbNlipKamICJCBz+/\ngKcqd6lSL8PDoxiioiJx//49TJo0FgCQkpICD49iT7UuIiIiIio4RR70q9VqrFy5EitXrpSmnThx\nAtOnTwcANG3aFN999x3Kly+PatWqwc3NDQBQo0YNhIaG4tixYwgKCgIA1K9fH5MmTSqQck2blopp\n01ILZF0ZNBq3Z3oFtEKhsPhZCAGVytSlZ/58yw/e5kavT4dcnvNjHF99NRfduvVC3br1sX79j0hO\ntvzQrkwmy7JOvfRZqVRJ//v4aPD11zlfYBARERFR0SnyB3mVSiUcHR3NpiUnJ0OtVgMAvL29odPp\nEBERAS8vLymNl5dXtulyuRwymQxpaWlFtwEFSC6Xw2AwPNUyZcqUw+3btxAdHQUAWLVqBXQ6bY7r\nLV/+VVy6dAEAcPZsKN54o1KO646NjUGpUi8jLS0Nx48fNQvos3J2dkFkZAQA4MKFc9nmu7u7AwBu\n3boJAAgJ2YB//73+NJtJRERERAXI5h7kFUIUyPSsPD2doVQq8kxXGDQatxzn1axZDZ99dg0rVy6G\nm5sbXF0dUayYMxwcVNBo3BAd7QK1Wmn2uXRpDaZMmYwJEz6GWq1G5cqVUanSK2at7w0avIPx48ej\ndOmXMGPGVEyfPh0ymQweHh6YNWsWXF1dpbRZ8+vduxemTBmH0qVLo2/f3pgxYwY6dAiCUimHRuMG\nhUIOHx9XdO7cAWPGjMFffx1CpUqVkJaWBC8vFykdAHzxxSzMnj0TKpUKvr6+6Nevl3RhZ2251QlZ\nD+vF9rBObNPzWC/PY5mfhr1vX34plQqb2he2VBZbIBP5iZoLweLFi+Hp6Ynu3bujefPm+PXXX+Ho\n6IiTJ09i7dq16NatGzZu3Ij58+cDACZOnIiWLVtiz549aN26NRo1aoT09HQ0a9YMhw8fzjUvnS6+\nKDYpG1P3HuvkTZaxTmwT68X2sE5s0/NULxpfd+mzThtnxZIUruepTgpT6dKuqFrViN9+s433+Vir\nXmz5QsMmxumvX78+9uwxjVe/d+9eNGrUCG+99RYuXryIuLg4JCYmIjQ0FLVq1UKDBg2we/duAMDB\ngwfxzjvvWLPoREREREQ2r8i791y6dAmzZ89GWFgYlEol9uzZg7lz52LChAnYuHEjSpYsiaCgIKhU\nKowePRr9+vWDTCbDkCFD4OYw2BK/AAAgAElEQVTmhlatWuGvv/5Cly5doFar8cUXXxT1JhARERER\nPVes1r2nKLF7D2Vgndgm1ovtYZ3YpuepXti958XC7j2Z+doqm+jeQ0REREREhYdBPxERERGRnWPQ\nT0RERERk5xj0ExERERHZOQb9RERERER2jkE/EREREZGdY9BPRERERGTnGPQTEREREdk5Bv1ERERE\nRHaOQT8RERERkZ1j0E9EREREZOcY9BMRERER2TkG/UREREREdo5BPxERERGRnWPQT0RERERk5xj0\nExERERHZOQb9RERERER2jkE/EREREZGdY9BPRERERGTnGPQTEREREdk5Bv1ERERERHaOQT8RERER\nkZ1j0E9EREREZOcY9BMRERER2TkG/UREREREdo5BPxERERGRnWPQT0RERERk5xj0ExEREREVkWvX\nrqFFixZYu3YtAODhw4fo0aMHunbtihEjRiAtLa1Q8mXQT0RERERUBJKSkhAcHIx69epJ0xYtWoSu\nXbti/fr1KFu2LEJCQgolbwb9RERERERFQK1WY+XKlfD19ZWmnThxAs2bNwcANG3aFMeOHSuUvJWF\nslYb4+npDKVSYZW8NRo3q+RLOWOd2CbWi+1hndim56ZehJA+aqxYjKLw3NRJIVMqFTa1L2ypLBmU\nSiWUSvPwOzk5GWq1GgDg7e0NnU5XOHkXylptTHR0klXy1WjcoNPFWyVvsox1YptYL7aHdWKbnqd6\n0fi6S5912jgrlqRwPU91UrhcodcbodNZJ+Z6krXq5b9eaIgsF8sFjd17iIiIiIisxNnZGSkpKQCA\n8PBws64/BYlBPxERERGRldSvXx979uwBAOzduxeNGjUqlHxeiO49RERERETWdunSJcyePRthYWFQ\nKpXYs2cP5s6diwkTJmDjxo0oWbIkgoKCCiVvBv1EREREREWgatWq+PHHH7NNX716daHnze49RERE\nRER2jkE/EREREZGdY9BPRERERGTnGPQTEREREdk5Bv1ERERERHaOQT8RERERkZ1j0E9EREREZOcY\n9BMRERER2TkG/UREREREdo5BPxERERGRnWPQT0RERERk5xj0ExERERHZOQb9RERERER2jkE/ERER\nEZGdY9BPRERERGTnGPQTEREREdk5Bv1ERERERHaOQT8RERERkZ1TWrsAAHDixAmMGDECr7/+OgCg\nQoUK6N+/P8aNGweDwQCNRoMvv/wSarUa27dvx5o1ayCXy9GxY0d88MEHVi49EREREZFts4mgHwDq\n1KmDRYsWSd8nTpyIrl27IjAwEPPnz0dISAiCgoKwZMkShISEQKVSoUOHDvDz80OxYsWsWHIiIiIi\nIttms917Tpw4gebNmwMAmjZtimPHjuH8+fOoVq0a3Nzc4OjoiBo1aiA0NNTKJSUiIiIism0209L/\n77//YtCgQYiNjcXQoUORnJwMtVoNAPD29oZOp0NERAS8vLykZby8vKDT6axVZCIiIiKi54JNBP3l\nypXD0KFDERgYiHv37qFnz54wGAzSfCGExeVymv4kT09nKJWKAinr09Jo3KySL+WMdWKbWC+2h3Vi\nm57Henkey/w07H378kupVNjUvrClstgCmwj6ixcvjlatWgEAypQpAx8fH1y8eBEpKSlwdHREeHg4\nfH194evri4iICGk5rVaLt99+O8/1R0cnFVrZc6PRuEGni7dK3mQZ68Q2sV5sD+vENj1P9aLJ8vl5\nKfOzeJ7qpHC5Qq83QqezTsz1JGvViy1faNhEn/7t27dj1apVAACdTofIyEi0b98ee/bsAQDs3bsX\njRo1wltvvYWLFy8iLi4OiYmJCA0NRa1ataxZdCIiIiIim2cTLf3NmjXDmDFjsH//fqSnp2PatGmo\nVKkSxo8fj40bN6JkyZIICgqCSqXC6NGj0a9fP8hkMgwZMgRubrZ7RUVEREREZAtsIuh3dXXF8uXL\ns01fvXp1tmkBAQEICAgoimIREREREdkFm+jeQ0REREREhYdBPxERERGRnWPQT0RERERk5xj0ExER\nERHZOQb9RERERER2jkE/EREREZGdY9BPRERERGTnGPQTEREREdk5Bv1ERERERHaOQT8RERERkZ1j\n0E9EREREZOcY9BMRERER2TkG/UREREREdo5BPxERERGRnWPQT0RERERk5xj0ExERERHZOQb9RERE\nRER2jkE/EREREZGdY9BPRERERGTnGPQTEREREdk5Bv1ERERERHaOQT8RERERkZ1j0E9EREREZOcY\n9BMRERER2TkG/UREREREdo5BPxERERGRnWPQT0RERERk5xj0ExERERHZOQb9RERERER2jkE/ERER\nEZGdY9BPRERERGTnGPQTEREREdk5pbULQERERET0IkhMTMT48eMRGxuL9PR0DBkyBI0aNSqSvBn0\nExEREREVga1bt6J8+fIYPXo0wsPD0atXL+zevbtI8mb3HiIiIiKiIuDp6YmYmBgAQFxcHDw9PYss\nb7b0ExEREREVgdatW2PLli3w8/NDXFwcVqxYUWR5y4QQoshysxK93gClUmHtYhARERHZJUdH4O23\ngePHrV0S2/bLL7/g9OnTCA4OxtWrVzFp0iRs2bKlSPJ+IVr6o6OTrJKvRuMGnS7eKnmTZawT28R6\nsT2sE9v0PNWLxtdd+qzTxlmxJIXreaqTwuUKvd4Inc46MdeTrFUvGo1brvNDQ0PRsGFDAEDFihWh\n1WphMBigUBR+4zT79BO9yNLSIH8QZu1SkJ1xWrIIrh8PtXYxiIhsTtmyZXH+/HkAQFhYGFxcXIok\n4AcY9D9X5OGPgLQ0axeD7ITi6hV41XkL3m9Xgiw83NrFITviOn0ynNb9YO1iEBHZnE6dOiEsLAzd\nu3fH6NGjMW3atCLL+4Xo3mMPZNFR8K5WAenV3kLM/sPWLg7ZAa9335E+K+7fhb54cSuWhihvrmNG\nwvDGG0ge8JG1i0LPi4QEyAx6CI9i1i4JEQDAxcUFX331lVXyZtD/nJA/fAgAUF08b+WSkF2y/+f5\nyQ44/fAdADDop3zTvFISgH0/U0CUX+zeQ0XGadF8OH8+w9rFIEsY9BMREdk1Bv1UZFxnToPLwrnW\nLgYRERHRC4dB//OCLbH0lJwWzUexVi0Ao9HaRSGiF5zr6OHZpqn37YF6/17+vhEVEQb9BACQRUZC\ndWCftYvxwpPFxRbYD6DrzGlQnT4JWUx03on5m0s2THVwP5yWLLJ2MZ6Zwy9boDwXmn3GCxTsOv34\nPZCaKn2X37wBj64fwKNLBzhsXG+9ghG9QBj0EwDAM6ApinVuD+X5sxbny2/dhHvfHpA/fFDEJXtx\nKK78DZ/XSsN10tiCXfGLE1cUCOWZU3Ab0BtITLR2UeixYp3awXX6ZOm7PPyRFUuTSfXXEagO7s89\nkcEA9wG94dmyidlk5fFj0BT3gPq3XwuvgHlJT4f6991AcnKRZCeLy3yYVh4ZIX1WHz5UJPlT4fov\n17DqnduhvHCu4ApDFjHof14UcouQ4s5tAJmjBD3JfdggOOz8BS5TJxVqOV5kqr+OAACcVn1TsCt+\ngVoTC4JnYHM4/rIFjjbc+igLD4csKtK6hbDiceX4XQH/jTyjYkGtUKxTu9wT5bCfnFcsAQC4fBFc\n0MXKN6cVS+HRrSNcPy2a87pni0aZXwzsdkiPCQGPvt3h2eJda5fE7jHoJ6iOZhn3Xy6znCjJ9Fpt\nWUpKwWSaZBuv6X5qKSlmfeRdx4yE8xczM+cbDJDfvPFs65blsO//I+WlC3knKowALiUFjj+sNnVZ\neh4ZDXmnSU01tfSmpxd8/kLk+HfiU+11+FQsX/B55pP85g1oinvAYcO63NPdvVM4BSikv5XCIEuI\nz1c65YnjcBs6sEhfwKi8YLqzqzr6Z5Hkp8hyp9htTJY+/hnnn6QkNlK8iFjnRYZBP8Fhx7bML0Xw\nKmiHjeuhKVcC6l93FHpeBU1TxhfFsrRGOP3wHVzmz5G+u3w6Ed51q8Nx1YrMadMm529b5YXz56jI\nR+Dl2bYl5PfvZU5ITwf0+mfP1GiEpowv3MaMgOvoEflaRHniOGTRUc+eZy6cVi6D8vixbNMdNq6H\n85ezLC8ky7s+XD6fgWKd2sF5ScG/aMV1zAhoypWAPOx+ga/7v3IM2QgAcB+e+3j5LlkviAtSPuqm\nsChPn4TG1z3vbj2P+VQom3uCxwGPZ9uWcNz0E9S7i6a7j8OWzVDcf3xsWeFhf+W1fzK/CAH5gzBo\nypWA65j8nS/IjjDoLzIM+p8XT9GyJX/44Klai0SWQF8UwY+p0zfLAACO6394uuUWzYfruI8Lo0hP\nRZVLy7nj+rUAALeJY6G4dBGymGg4L10Ejz7d8l6xhaBfceki5I8sd7nKN0M+WqxhCmAz+LxSEt5v\nvpFreoetITkGpIrr16TPyssX8878yhV4tm2JYgHN8lVWidGYZyu7TKuF6yfj4fmef7Z57sMGwSWn\noD8fF2GqY6YuWcqTx/Mu61Ny+vF707pDT5smCFGkrcC4dg0e7wdKdSkPfwTFxcfH/uMgUeR1XsrP\n3ZJnYcWWfudF8wEALsFT/9uKHm+D7Im7ObInL7aftc71ejhsWGfxQlpx5W+4D+oH1emTpu93bsN5\n9mdwKYpuPjkEeMpzprsOGcc9PedSUuBdoYz5nfCcMOgvMgz6C4n80UNgw4ZnX4EQ5gFVPv8o5I8e\nwvutiijWvk3+85Jnad2Xy035PsMfoeOqb+C4+ts800k/ak+Rh0yrhevMaXD6fpXZdNdxH8Ozfs2n\nKaZJerp5/gXUPUOWZZ0OO7ZCce+u9N1p2de5LpvtAV4h4NWsQZ7Bd9b0FoPw/AZeWVr7ZKmpkEfo\npHUqjx8ze9hPeS4U7gP7wrt6ZfiUewkync40I2M/Zr3QyE+Adsd0N0J562b+yvqY57vvQFPKO/c0\nbfzMvjts+gk+5Urk/VB6fu56ZQRuGftOCChPHM+xW45L8FTTg5NPwWmd6eLYvWdnaF72MXUxK0ip\nqXDYuD57N5SPPoL62FG4jh0JAPCuVgFezRua6lY83t68LowKq992IdwVU9y4Dp8yvlDv3G42XRYe\nDue5XwAJCebTky3XsfL82fzdsXp8rsh2Jy7L34t69y5oXvaBw5bN+dgCc45rvoP78I/gPrBvtnlZ\nH6gFAJnBAJd5s+G8PPdzVFaejepYHIYzL6pDB7NNU/59CYr7dy2kpueV4t/rkMfEmN0JBwDHH1bD\n8dvl5okZ9BcZBv2FpFhgc6BLF6iOHbWcwGiEe7+ecPh5k8XZTksXw7t65Tz7zEIIQK+H4uIFyKKj\noHjcn1yVW8vj42WyTJA+OezZBe/qlaEp7gHl41YgwDyYfZLzrBlQ/74bbhPHwG38qOwJ9HqzP2rF\nzX/NZsuiIk0/ahmBk14PmVYLCAG3QX3hsGEdfKq+ZjFvp+9XQfnvdcji4+DxwfvS/vbo3B7ufXtI\n6VR/HIDD1hBpuzWlvFHsvQAAgMvk8dCU8obixnWzbX42mdvpsmAuPJtnPrjmOnUSHJ+4aMlK9mSL\n3tOcCPV6OM+eCe/qlbONBqI+ekT6XCygKVxHDcuh6NkDtIzj0PM9f7gPHShNz/rAtywpEe7DBkL9\n+25oSnnD4Zct5t0FZDIgLc0UDGeZrrh+DR7/e880YkNgYP63NQuzLgI5UNy+JX12XP0t3IcOhCwp\nCQ7btpgnTEuDLDLLw7H5Cfozgk8hAIMBXrXfhGfblnAfZCHQioiA8+IF8OjWEU5LFkF+7655XaWm\nmuf/mPrAPsBohMOe30xZ6rR5l8sC5ckT8H6lFJRnTgEwBV8endrBed5suA8bBNcxI80XyLiAe/Ki\nzWCAzPj42LQQfKuzdBeUZVz8paTAecGXkD98YGoQ+a8/8v8h6Hf6ah40vu6mc3QWjmtWQ5aSYnac\nA4D7sIFwmfM5XBZ8aZrwuOzKG+bnMcB0N8TTrzG8GtQ2fX88QMJTkcmA5GQorvwNxzWm84XT44d+\nn0bGca88ZeGc5qB++nI9QfnP1WdqlZdrw7M1siiv/A3XyRPyvQ7V0cNQ/H0523TF5UuZf0NCwGHb\nz09dPso/1bGjUFzK+06u01fzpLtIbmNGwG3SOPMEeZ0PMmICALKYaCgvnn+m8hKD/kKjeNziKg+7\nDxiNcFq0APJbN4H0dDhsDYHy4nk47NgG94/6Sz/CWWW07Dj8uj3bvKzce3SCpqQXvJo3hM8b5bL3\nw7bQtaNYi3fh82opAKY/IOcVS6V5WUeOKfa/tpCFh+e8jf9chXrXTrgsmAuPbh0zp9+4bhpSLzER\nDltDoCnphWKtMn9gZVnGalbv3A6fiuXhPqgfHLZvBQB4fPA+fKq+BsXfl+G4JSTPfsMA4Lj2B6gP\nHUSx9wPh+e47UB/YB4edv2RuS8cgU4uXSiU9WKo6cQye9WrA+XF3I696NeHZqkW++uqq/vwD6sfl\nzerJW/VPchv3sXTyAkw/5j5li8N59md55pkTx9XfQlPSCy7zTUHJky3JqiOm4fBk8XFQhZ6B09o1\nFtejOnMmWwu146bMEWwcdmyDe5/upqAvNsYsnfrAPqnblttH/VGsXevMmXI5NC/7wLNtS7Ox1t1G\nDoH68B/ZR2x44piV37xhWu7xBYP80UN4dAyC4vKl7Ntw4Heod+20uH0AzC9Kszy0rrxwDp6N68Kn\nUubDsVLrqsGQ7UdJ/duvkD96CNUZU9cb9cH9UB06KLXaOuzelT3zLNvlOn0yvGtWhUevLnBauhgA\n4FW/pil/C89SuA3un1nsvIarNBqlOy+qI3/CZcanUJ49A882fpAnxMNtzEhAr0exD96H+uB+6S3Z\nqsfdiOQ3b8CjYxBwzdStR330sHnwKoRUF7IswZv8QRicFs2HR7+e2bbZadU3cJkVDO+3KsL7zTfg\n3quLqYyPL5ayplXc/BfOcz6H8uwZOK7+Fq5jLXTpk8uh3vsb5GH3IYuMhMbXHW6PzxOy8HAoblyH\nZ9MG8K6U/WFn18+mm7b3zCnI79w2vRwKyLzIy3JnTHniONR/HAAAOK5bA5eZ03LY6Y+LpTWdL+UR\nOkCvh3ftN7OnuXUTTosX5nhuV4aegaZscXg1rguH/b+byno2FI6rvjG9SyUyEh6d2kG94xezelFv\n3wqNrzvUu3fBackiaXQg2RN3+mQJ8fnr8mfhuM+JetdOqRuiLDbGFGznkIf70IFP3Y1PcfmSqWvn\n49+OYu1aw6tJPSAxEV5vVzI9Q5WQAK+m9eFdozI8OreHd+VX4P5hn6fKR3niOBw2/fRUyxQU1fG/\n4LA5j94BaWnwqv0mnBYtKJpC5aHY+4HwatbA4jx5ltHFXD+bnvtdpDyOM49O7eBT9TXIHz2EZ9MG\n8GzeyOKdWsW1f+BdsRxUj/9myQLxHPrss89Ex44dRadOncT58+fzTK/VxhX5P2E6jIW+ZCnpswBE\n/IzPzb5n/NP9fVNow2Ol5dPeqi4EINKq18iWNn7G56Z04bHZ5sUuX5VtmlEmEyktA0TMmp9EUt8B\nmXleuSVS/Pwtlifrv5hVP4j0ylWFAERKQKts2/is/wzFipl9T323qYhd+X3mfB8fy/vq+l2REtBK\nJIybJE1LadXWYlqtNk5Eb9hivj3rNuVarvjJ00Ts8lUiYdwkEbtomUh/9TWhL1NW6K7eynGZiNDL\nInrDz/na7qjf9ouYNT/lug+TO3YxK6f2UYzQauNEwqhxpvzOXRGxi5aJqF37LC/fqavZ94QxE0TM\nDxsy98GPG5+53qJ+/V3Efxqcbbq+xEv5Wj5y/xER88MGi8d2xr+kvgOE9kGU0P1zWxjcPYQAROKI\n0SLyyKnM48PbO3P/nzhndiwnv99exE+ZIZJ698s5jz798y7rgaNCACKt2ltCq40TEeevirgv5llM\nG7PmJ7Pv0dt3i6gde0VS3wEiatc+EfnniZzzOfiX9Dlx+CihDYvM377864zZeSe5/QfSvLgvFz51\n3SYOHi5SGzXJNj3t7eq5LpfarIXF6Xrf4kL37z2ROHBwrsvHrvhOCEDET52Zr3ImdeuZeRz4aKTP\nls6J2tuPTOeBrb/meDxE/nFMpJd/JbM8Xy0VSb0sp02vWCnbNN2N+yIlsI35PqtZO9d6Nlv+2p2n\nritp+7RxZus1OjiYzTeq1dL5I/6z2flan1YbJ4wKhUh/7XUReTxUOraTuvYQWm2cVF8CEPHTTb9p\nRplMJA4cknksfTTsmbfJ4OkpUlq/JwRM566s86I3/yJ9jvkpJHOZJ35PLG7Xoxihu/SviPlurYje\nsEWk1msg4idPE7ob983Ox9pHMSLi1AURceEfaX/ErN8sIs5dyfx7C4sU8Z8Gi4hzV549zngUI/3u\nS8fIX2dE4kfDREqb94X2QZRZ+ozzkQBEUr8PhfZ+ROb88FgRtfcP82lP/Is4+7eIOH3xv8U3j4+l\nrGXW3bgvtLceCrXaKGrW1JuOH5XKYh08WR9abZzQ3tVmO/4sxVPR23ZJn6N2HzArU3K7/wmDq5sQ\ngNCXLmNazlrxnw2TCSGEtS88nsbJkyexatUqrFixAjdu3MCkSZOwcePGXJfR6fI3ZFpB0vi6P9Ny\nBt/iUGhzbl2np2MoXcasXz0RvVgSZnxeZOPQk31Ia9oc6sd3fNOaNjc915SlK2Hsuk2mOyoZ3Zs6\ndABCQpBepy4SPpsN9e5dUJ4/i9R2HaC8eAGyuFgo7t2F+vAhJEz/HE7LFkORZYCGxJFjpDtuT0qc\nOAWKmzfguHE90mvWhuqJngEp77eH4y9bkPK/jnD8eRPSq72F9CbNIJydYfT2geLObejffAuKa//A\nZd5sAIBwckLsmp8g12mhr1QFolgxOK5dA0OlyjCUKw/h6gqnxQvhtP5H6N+oiKQxE+A+oLdZvvpK\nVaC8Yt7FykGehrdeCseBQWvhOmVitm1Jbf1etjtchuIlYKhYCeosz3qkv1MPiZ9MheKfq1BeOCft\n51Q/fzj8vkfaBllyMuJWfAfXUcMhTzR/5kZfpRqURw9Dl1b0HVo0GrcizzO/nrug/6uvvkLJkiXx\nwQcfAAACAgIQEhICV1fXHJd5noJ+IiIioueNA1JQA6E4hvrWLopEp43LO1EBs+Wg/7nr0x8REQFP\nT0/pu5eXF3QZI4cQEREREVE2SmsX4L/Kz40KT09nKJWF/9IpIiIiIrINttzqbg3PXdDv6+uLiIgI\n6btWq4VGo8l1mejo3EdUKQy5l8j69K+8ithN2+Bdq9ozLR+7el3+XjhVCFID28Dht5xHaCkM+oqV\noLx6pUjzFDKZNFSqoUw56N962/ztyXYuaeBgOPy2C4q7ty3Oj131g/lIMYVIyGRI6dkXTmtyHnLV\nWhJHjYPi9k3oa9aG6yfj/9O60t98G6oL5wqoZC+OlPYfwDEfY+nHrvkJyiuXC+xNxYmjxkJx5w4c\ncxj6OavkHr2R5h8Ij+6dCiRvW5Beq470grG8GDW+EE7OOZ5PhKMjZBbegWHwLY7YkO1wnfYJUtp/\ngNT32kGWngbF7VvQl38V8ugoeE+dgOjhY6B/3fROFUXYfahOn0Sqfyuof9+NtMDWkMXEwOijgePW\nEKTXqgPDK69CFh8P9eE/zEYZip8932yUsdi1G2Eo/yoMpV6GplyJzPLKZIjefwSu0yZD/edBRJ66\nAO/ab8Lo7Y2YnXuhuHIFzovnQ7h7IPaHDfBs0QjKxy/ai12/GfJ795D6QSfTaF+xsRCurpAlJkL+\n8AFU50JheLkMVH8dgbFkKQhHR6TXqQtDlapS1+WET6bC6dsV0Nd+B4aSJU2j4CkUMBYvBd1xLRy3\nbIbbyCFm+1J36yE05V8CAETcDIPi8mUoL19EWkAreL9dySxt0uDhSO7TH5DJLMYp0dv3QKZPh+MP\n3yG1zftIa94Syqt/m4a8NRrhtPpbOM7/0jrdu235QsOqjxE/gzNnzojevXsLIYS4dOmS6Ny5c57L\nWOXp7cdPmKe0aitiv16ROaLBiNFCACLui3kidulKETdvkdDeCc9c9mG0iN6yU8TNmmt6Qv33Q8Kg\n8TV7Aj7Fz98sjyf/6Uu9nO8RGrT3dCL91dfyTJ/6eEQPo7OLiDhzSWi1cSL9jYpmaSL/OpM5YoRM\nJuKnfy7iJ0/PfNp+x16R2qSZNBqLAIT2QVTmvhn2sYhZu1HErN0odFduicRBQ03b6x8oYpd9a7Z/\nLY2MYXE7H8WIlIDWmdNGjBCpDd/NHHli5hfZlombPV9EnLogIg+flKZFnDiXfYQJF9ds+/PJNFlH\nm5FGIPjlN6HVxonYRctyLHfihx9lH+kgy3eDp6dIaRuUuR0zPhfau1oRHbJd6G7cF9FbfzXPc9M2\nEfXbful7XqOxPPkvxT8wc//MWSBil32bLU16lWo510OW0Wi0d7W5Hr8CEHGfz5G2Of2114UAROw3\nq00jU9zVmka8yDJSi/ZBlEgYO1H6rrtyS+iu3hJR2/eIpG49Rdzs+SJ+ygzT+h6PvhL79Qqz0SC0\n2jizEZMSRo6RRn7RauNE1L4/pXmxXy0VEacumNf1/iMirVYd87+beg2kz0ndelrc7qhd+4RWG5dt\n9Jd8Hd9ZRvx68nixlF7/UkmhL1NWJEz6VBjlcpEw6VOz+bp/7+V7FCazuq/whnSei/l+vdlxE3H+\nqsXRniyNWGNwcRX6MuXMpiV8PEZEHj5pNvKM9mG00N4JFyl+/iJ607bc99E9nenfo5jcj7nZ8031\nmGWUJaOzs/Q5/rPZQvfvPZH8QWfTvvQtLrRhkSLi1AWhDY81G03syX8RF68J3fW7OZ4ntPd0uZYt\nNeuISsOHm9WzVhsnlcmsTqpUEwmTPhXamw9E9LZdZsdKzPrN5vvd21saRSXi5HkRcfGalFZ35ZbZ\nqG9xn8+xWMaECZNFxOmL0vk1ud3/nvo4MttnZy6Z/X5YzHPSpyKpT38R9+XCXNNG/nnC4u90xIV/\npPORpfPt04zW8l/iBbsARIQAACAASURBVN3F6yJ20bLMOrr9SEScuyIiQi+bpYve+qvp3PjE331+\n/8Ws3yxil3xTILFN7FdLzefd05mN3qN9FCNS/AOl0Z2ePC/ltF7tw2ihDYu0OC8/dVqQ9fLM+8iG\nPXct/TVq1ECVKlXQuXNnyGQyTJ061dpFylVKl+5I8w+Erm0Q4OQEwNQyl/E5G4UC6Q3fRXrDd5HS\n70MAQOTlfyELD4dPtdcBAHHrzFuU0uo1gPrxS6nSa7+D+PmL4dWojjQ/fs4CuI2zMNY1ADg4IL1e\nA+lFM4ljJ8Lly1kAgKhjZ+DerSMMVapJb5jUV3sTxtJlTGmnTDdrNTK89rr0OeHLhUjpaWq9cJ1p\nqiP9O3URu2kbXKZ+Audli5HcpTugzDwEDWXLIa1l5kuaEqdMR1pzP6TXawCozV8mE/Pr70BaGoq9\n5w/VubNIa9YCScM+hv7NtyB/+BBeDWsjrUEjQC5H3A+Z4y5rNG6I1cVDef4s5GFhSGvVxuJLYYxl\nywEAYn/cCNWhAzCWK4/4WXOhOnUc6r17IE+IhyhWDHhixACjj8Y0RjeAhE+DYXi5tDTvyQeKUjt3\nQ5xCAceQjYhdtxmakl6ZM1WWX56TMPMLuE6egKijZyCcnKDq3guG8q/AWM40Hnn6u01M/zfIfClY\n0kfDkN7ENC52RitSapsgqB6/9j43hjLloLh7G8LNHVH7j8Bh3x6k9OpreudEFkKlQvRB0zHoXflV\naR8AQKp/IKBSIeLaHdMbWh0dAQCRJ89DFXoaspQUqUUoZstO02gS3plv2Y3ZugsO+/ci9b122V7K\npH/lVShv3gCUSiSNnQjVkT8BIaTl9XXrIaFuPbNlYkJ2wOH33Ujt0AmKxy1fGdL8AqTPxuIlAGfn\nzH3xeB+nBrRCapfupmnFS0AR/ghGHx8Yqr0JfeWqZi2Pces2wecV0zsxEqdMNy/Hpm1w2PEL9DVq\nAQBSOnTMdgcrfuESs9ay5G49YXitApKHPP2bUAEgud9AJA83nQuSRo4BALh8PiMzgV5v8cVXsd+u\ngUf/XqblBgyCcHOT3guR9bjWdexitpxwcoLxpZJI7j8QrjOmSNNjtuyEKJb5bFbE5RtwnTIBieMm\nwaObaZCGlKD2iP9qmXSuTH6jYuadDIUCcHIynQufeK+BcHaBLCkRUX8cg+LObcDBQZqX2jZIulsW\n/fshuI4aDtXF80j1D0RK735I6WN6H0LEP7cBowBkgOLaNRh9fWEs/wogkyHh02DAYEDi+E8AlUo6\nVySN+X97dx4eVXn///81k0kgK8mEmYABlE20VnBBZFVwQUWqdQkiAtpiES1oq6KIVFDriloruLIV\nrRYKWvXzlVLcoNZGLNWfC2pZZBMkmZAEErJnzu+Pw2zZZhKSmcnwfFxXrsycOWfOfebO8r7Pvbxn\nqfrMs5R+zRXez8VzV9Sd5btD2yC/MvqrvPAilU+9RdVDh/syT/vNafM4fM/vFLdzhywHCmT7frtq\nevX2/j5KUvXQ4YGfUZ06NpKP3Jns2NH7t8T7Wmamyn47U4lLF6nkkfmqmHKTKn45VZbDpTKSU+To\nkm7uaLXK3eN4lfxhoZIfeVCH739IHf/W8uRY7u49VH7rb73/PySpdO7vVT14iDKOJFYz4hNU+thT\n5un37A7Y11/tSSc3uN3dpWuLy9eajKwsVY736zlPSpLb72+PR/WwEQF/25ur6oKLWnxsPXWT9tX9\nGbZadegVc2XFlLm+1bMObPz/AnL1+Kv5yU/N3+1GEiO6MzNlbSCBIULX7oJ+SbrzzjsjXYTQeX4v\n/IP8xgL+pt6mgWypHgdfW63OfbqpbMZvVTb7PklS0Zr3lDHmAklSxQ1TAoL+mr4n1nlz3y9vbbfu\nKnrnXbl7HC93VhcVfWIGhp0uN4Nxw++XserCi3XomecDkmdVXnypOqx9RzX9B3i3Fa/8mwy/1ZUO\n3zVbtT17qfJqX0Iv882NwOfx8ao+d1TDF221msGjX0ZUzx/D2tQ0uXbleYPLhtQMOF0acHrg6Y90\n79YMOM13jRddoqqLzGuvmDJVFVOmKvXG69Xx7b/JSE5W0dv/kLtbN/938Su/rV5inLoqc8arMmd8\n/ReO1EnZrberpt9J3s3lU29R+dRbvM+rR51f79C6jETf51BxwxRVn3Ouanv1qfcPsvKiS+TO7qaK\nayYo4yLzc6+8/AolLfiDqs67QLWn9lfZqWayIXev3ip+6+9KeOdtJb30vIzk5Hplr/vcP8iTJPcJ\nPVV5Qk91fO0V3/UMr5OoS+Y/xIoJk+ptl6SijzcFZPc8+NbfG/sYfO/ndKriuskNl7WBcnuPS+sk\n1879Ab+/tX1PVFzeftX0MX+njIT4wGNSUnXgs81y2zMDGhCSVD3yPG9j7MgJ6xWhYsIkb9BfefEY\nlT48P+S/H4dnzWn+MJKa2sC6lFQ18jxVXXaFJDPoP/zQ4+qw2vxn7k5ufNU0STp8/8PmA7/fRde+\nwoDGviQZDodKXjCHTtWc2l+27dvkPr5naNdap54OfL1FlpISubsep9qfnBLw2qEXlighZ7xUVama\nAaer+P2PGnxLI8PXAK+p02g0srK8Za2r9kgD4GiUPPmMEhe/INu330i2Bv4Gjhkj3X+/DL/rdnfr\nruJ33lXCu2vV6bpxAX8jQmM0+ao7q0vgTQurVUaq2aA59NIypdx+qyquNm8AuY8/odHPR2r+Msql\nDz3mbezVbexWXO8bFmMkBf7com14lsCtGnVBi4539+zV4HbX7vx6N/fqKn7nXdkHn9Gi88LULoP+\nY1ID2Q2L1rwna1GhlJysgn2FAf/8agYOUm12N9Wc/BNJkmvvATOLpttd7xer7LczfesNWyyqOevs\neueyeM4f5/cjY7HU+4d0aPFyxe3Zpdrevrv+9QLT5GRV3DCl/jW6G2/YNMbd5ThJ/5W7c51ZFC1o\nWB166U+q7dM3oMeiQZ7P2e2uFxD4X0PFNRNk2AKDwKaU/u4BpTx4X8C2w3PmhXx8Y2pP8Psja7EE\n1I0/IyVVpY8+aZ737nvlTs8w735eM0G1dRuKkqqHDFPNT0+Vdf9+ld/mG4Nq1Aukmwisj1YTd4VC\n0sQEfyOtgWV36wTu3sasp2fGr4fmwGfmGtZuv96eJvndfS34dke9O9iHnlvcrJ/rstvv8gb9Naec\nKtvmr4L/bHfsICUE3rEru9Ws29rsbt5M45U/v0pl332rijp39r1lfeZ52bZukbvrcd5th+/5ndyp\nqfUC/rpK5z+t6sHDVDE+xDlDdRtnKakyUhoZUxsfr6qLx4T2vi3g7tVbhxa+qJozzd6b0ocek+Xw\n4Xr7Ff2/d5Xw8T9l+/wzVQ8cFPCakZSkw/fOVaeJ16j85une7QeX/0Xurl2VMWiQXN/vkxpYprrq\nwosbfS1Qnd/JujdcmqHy51ep8udXhb7/z36upOeeCdhW+sDDSnj/XR1c9qocvY4L3P+SsUq5926V\nPvRYvffyr2cjM1OHXlyqmpNPkf0c83+Ytycw1LJdPEa2z/4b8v7HovJp01V+06+bvmFSh2dN/SY1\ncZPOq9YvRmigRxLBEfS3Fw0ExDX+/ywa+AUs/Gyzb3t848FnQFDS2B//2iMBSFyQX7SEhEaDyqBa\nEPSXPP4H1fY4XmW33dGyc/qLtwUPiiTJeuQzbeizOnINZTfe5L2zXfR/6+TOzg76tu7uIQaHISqd\nc79Sfj+38d6Suvyup+wO34TQ2hP7NX5IappKFi8P2HZo8ctKves3MpKTFf/fTapt5M5ONKjt2Vvl\n101usNs7lECm9JEnJGucSueZwXVtj+MlSdVnnBl6sH+Ep4ep4sqcgOFNHhYZQe7HNu7gq3+V7Zuv\nVXX+6Eb3KX7r7zLSOvk2XH65XC+97P0bUvifL6WqKvM1m63JBmnl+OtUtwO/7LczQyqr0SldFb/8\nVUj7SmpW8BEOlX4NofJf3dzgPjWDzlbNoPo3VyRJVquqRl8iV97BgGuruuRS3z5NBfVBA37JqDtE\n6CiC/marE6xVjTzPDCSnTW9wd3e37vU+i8ZUXnF1wPOif/1HamQoSUMOvbwivJ9Fe9XM37mCrXta\n9P+9npoaHXx5hRL+/v8avBGF4Aj624uW/MK04j/Dst/OVKeJ16jslpaNJQ5N8//YGk6nDj/wcBuU\npQkW85+WpYHeF7mPXIN/r8vZg0N628qxl3sfG0Huhoai/NbfqnzGb0L+OWhukNqYmsFDVPTPjbIU\nF6njytdUPvGGpg+I5D9Zi0Wlf1jY8Gsh1IE7u5sO/elV7/OKyb+QpbZGlZdf2eyiuLt1V8HW3YGB\ndzOV3Xq7arO71dvuTs8IOp63esgw84F/ffj/7NhsIX0mYedXxqK1H0SwIK3D2zvYho2ZmtPPVMkT\nf1THVSsUvzE3vL+DR3rmjPh4Hfj2+8Z7Zfy19LNoyc9slDUiY0KQYTshq6lR1cVj2rS3LtZF4V9w\nNKihALMVue12WQsLZdjr32GUZN552l9c7y6N0cBko5aytMadgHDwDCdxN/CP0qgf9IfMZlPRe/9U\n0vxHArr1j0qI5SifMlWHQ7wTGyojPcPsBj6WxMe3YDy1j9Ep/ahO3+jd9yZ6+hrVjoKfquHnqHrg\nIO+k6PaoeNVbSvzTEu8corZWMfkXsn22yQz6w/i31/D8LMYnNNjAPfTSMlkKC8NWHrQfwebIITiC\n/rbWSv84m5rI2xqK1n6oDmvfUdWFTdwNbGAMndEpXcWr3vKubNIShxb9Scnz5qji51cH37kNuVND\nvMPqnTzcQJ14g/6WjTes6X+ad8WDcCqbNl1KZiKcZI7rdTfS+I2EQ888r/hPPwntjmgdRX9/X3Hf\nb28y6C/88N+yVPjG2x6+faa5Us+v277BVn7dZFkPFATfMYiDb4Q3b0dbqD53VOhD8VpLBBp25VNv\nlu2LzwOGEPprzvwAHBuqhg5Xwr//pdpuPSJdlHaPoL+dqO1xgmr69A0YL9qa3Cf0bHRMZTBH+4+q\n8vIrWzQcorUUfrxJCR+8q5qzBgXfWVLFdZPVceVrOnxn/aU+vb0V7eguaaS5j6yU4j6KYS2t6dDL\nKyJdhACV468LXM6vGWrOPEs1Z57V5D61p/w04HnVZVfI9eNl5lKMbZzYptGhVQiL8l/fpg7/WKOS\np58N2zmN9Ix6y04DTTn4+v/JUlzc4FwnNA9BfxtxZ2TIWlQkt8PZOm8YH6+if7OqQFuo7Xuiypsx\nKah68NAGlx00HbnTz8oCIau6eIz02GMqOrfxSaYIs6NZEQntRm2fvjrwzffBdwQiKS6OgL+VEJm0\nkaIP/y399a+qOY01ZWNSI5PDKo+Mx/Vf679diGTPhNUq3XWX3L16R64MYVTbrbuMUJanA2JM5cWX\nBt8JQJvhTn8bcR+XLQ04qc27xxFdSp5+ThXXT/GthALUUfifL1kWEMek8qkNL2EKIDwI+oHWlJhY\nL+U9EIChMzhGtdZqb0Vr3pOloqLR14tX/o3hIEADCPoBAECbKrvlVtWcfmarvFfNwKYXXaiXBR6A\nJIJ+ABKrDQFoU4ePZK0GEDlM5AUAAABiHEE/cIwqeeKPkS4CAAAIE4J+4BhVMfkXkS4CAAAIE8b0\nAwDaBXd6umqPkXwOANDaCPoBAO3Cge92MukcAFqIoB8A0D5YGZEKAC0V2l/Q0lJp927za+tWaVDT\na+QCAAAAaD1ffCENHCiddJL5/MEHpY0bQz8+eND/+ONSt25Sv37SmWdKp59ufgGIHQyZAAAgqk2f\nLi1dKnXtaj6/5hrp9ttDPz540L96tZSfLw0eLLlc0muvST/9aQuLCwAAAKC54uOl/v19z088UbI1\nY6B+8KA/NVVKSJCqqsznl10mvfVWM4sJAAAAoKVsNmnHDl/n/N//LhlGM44PukdGhvTqq+bd/V/8\nQvrJT6R9+1pYXAAAAADN9eST0uWXS//7n9Spk3TCCdLLL4d+fPCg/+WXzeE9V1whPf209MMP0ooV\nLS8xgOjDmH4AAKJa587Sl1+ao+07dJDS0pp3fPDhPQ88YDYlkpKk2bOlP/5ReuaZlpUWAAAAQLNd\nd5353eFofsAvNXWn/29/k954Q3rvvcDhPNXV0j//2fwzAQAAAGiRE0+UJk+Whg41p9t6/PKXoR3f\neNB/8cWS0ylt2iSdf75vu9UqzZvXstICiCqHFv1J8R++L/dx2ZEuCgAAaEJlpRQXV39t/qMP+hMT\npWHDpM8/lzp2DHxt5kxp/vxmFhVAtKm8/EpVXn5lpIsBAACCWLbs6I4PPpH3o4/MsfwHDpjPKysl\nu52gHwAAAAiT7t0bXndj9+7Qjg8e9M+ZIy1YIP3mN9KSJdLKldKIEc0sJgAAAICW+te/fI+rqqT3\n35fKykI/PnjQn5ZmZuNNSJBOOcVczeeSS6QLL2xBcQEAAAA01/HHBz7v21e66CLp9ttDOz540F9d\nbTYtMjKk5cvN5Fw7drSgqAAAAABa4oMPAp/v2SNt3x768cGD/hdflPbvN8fwT58u5eWZY/wBAAAA\nhMWDD/oeWyzmYJwXXgj9+OBBf79+5pckrVvXzOIBAAAAOFoffnh0xzce9Pfs2fAUYY/vvz+6MwMA\nAABo0ogRTYfkoebMbTzof+898/tLL0ldukjnnSfV1krvviuVljajqAAAAABa4ve/b533aTzo793b\n/P7ZZ2ag73HGGdLYsa1zdgAAAACNSkuTTj+9/kTe5go+pj8/3xzLP2yYZLVKubnSrl1Hd1YAAAAA\nQb3yihn0+0/k9bBYzME4oQge9L/wgnTnndJXX5nPf/ITaeHCZhQVAAAAgCS9/fbbWrx4sWw2m269\n9VaNHDmyyf2fesr83tBE3tdfD/28wYP+IUOkjz8O/R0BAAAA1FNUVKRnn31Wr7/+usrKyrRgwYKg\nQb/H7t3mffeCAvN5ZaU55Oeqq0I7t7VlRQYAAADQHLm5uRoyZIhSUlLkdDr1YENjdhoxaZJkt5sj\n7c88U3K5zKE/oSLoBwAAAMLghx9+UEVFhaZNm6YJEyYoNzc35GNtNmnWLCkrS/r1r6W335aefTb0\ncwcf3iNJhw6ZU4clMyNvVlboZ4gCGRlJstniInJuhyM1IudF46iT6ES9RB/qJDq1m3oxDO9DRwSL\nEQ7tpk7aWHx8XFR9FtFUFn/FxcVauHCh9u3bp8mTJ+vDDz+UpamF+I8oL5d++MFcV+f776Xjj5d2\n7gz9vMGD/mefNZfsfPNN8/m110pXXilNnx76WSKsqKgsIud1OFLlcpVE5NxoGHUSnaiX6EOdRKf2\nVC8OZ5r3sSv/UARL0rbaU520rRRVV7vlckUm5qorUvUSrKGRmZmp008/XTabTT169FBycrIKCwuV\nmZkZ9L3vukt6/31p5kzptNOkuDhpwoTQyxZ8eM+f/yytXu17vm6d9NproZ8BAAAAgIYPH65PPvlE\nbrdbRUVFKisrU0ZGRpPHrFljfrfZpOuvly65RCosNO/2t+7wntpa8yweFktAlx0AAACA4LKysnTR\nRRdp3LhxkqQ5c+bIam36Hvztt5t39X/3Oyk5uX4Y3nrr9F92mTR0qDRihOR2m/0Koa4NBAAAAMBr\n/PjxGj9+fMj733yzNH++OX7/gQcCX2vd5Fxz5kgjR0obN5rv/Nxz0uDBIRcUAAAAQMvcdpv59eyz\n5qo9LdV4f8Lnn5vfP/hAqqoy8/+edppUVmZuAwAAABAWRxPwS03d6X/5ZTPQbyhpQHP6EgAAAABE\nVONB/x/+YH6/7z5p1KjA1zzLdwIAAACIeo0H/Tt3Stu3S3feKT31lG+qcHW19JvfSD//eXhKCAAA\nABzjSkvNe/L/+Y856GbwYDMkT0wM7fjGg/4ff5RWrqw/VdhqlaZNO7pSAwAAAAjZr34ldesm3XST\neS/+vffMbX/+c2jHNx70Dxlifo0Zw119AAAAIILy8qS//MX3fOxYc4HNUAXPyNuhg68Jcd11Ut++\n0htvNK+UTXjjjTd07rnnatKkSZo0aZKef/55SdJ3333nXcd07ty53v0XL16sq6++Wjk5OdqwYUOr\nlQMAAACIVocPm4to+j+vqAj9+ODr9D/wgPR//yf9/e9mdt7PPzebFlde2YLiNmzMmDG6++67A7Y9\n9NBDmj17tvr376877rhDGzZsUK9evbRmzRqtWLFCpaWlmjBhgoYPH664uLhWKwsAAAAQbW66STrp\nJGngQPP5f//b8CKbjQke9CclSZ07S++8I02aJKWkmLmA21BVVZX27t2r/v37S5JGjRql3NxcuVwu\njRgxQgkJCbLb7crOzta2bdvUr1+/Ni0PAAAAEEm//KV0wQXm/XeLRVqwQLLbQz8++PCeigoz9+/a\ntdL550tbt0oHDx5Fkev79NNPNWXKFF1//fX65ptvVFRUpLS0NO/rmZmZcrlcKigokN3v6ux2u1wu\nV6uWBQAAAIg2F18s9eghXX65dNllUna2dM45oR8f/E7/Sy9JixZJy5ZJHTtK//iH9OijLSrsqlWr\ntGrVqoBtl156qWbMmKGRI0fq888/1913363FixcH7GN4lguto7HtdWVkJMlmi8wQIIcjNSLnReOo\nk+hEvUQf6iQ6tcd6aY9lbo5Yv75QxcfHRdVnEU1lORqvvmqOtt+1ywz6Paqrpays0N8neNCfmCjd\neqv5+PvvzdV8WignJ0c5OTmNvn766aersLBQGRkZKi4u9m7Py8uT0+mU0+nUjh076m0PpqioLOg+\nbcHhSJXLVRKRc6Nh1El0ol6iD3USndpTvTj8HreXMrdEe6qTtpWi6mq3XK7IxFx1Rape2qKhcd11\n0vjx0pQp0v33+7ZbrdJxx4X+PsGD/vPPNwcOGYZUVSW5XNIpp5gDilrBokWL1LVrV40dO1ZbtmyR\n3W5XQkKCevXqpU2bNmngwIFat26dJk2apBNOOEHLli3TjBkzVFRUpPz8fPXp06dVygEAAABEo7g4\n6U9/Orr3CB70+91ZlyRt3iwtWXJ0Z/Xzs5/9TDNnztSKFStUU1Ojhx56SJI0e/Zs3XfffXK73Row\nYICGDh0qSRo3bpwmTpwoi8WiefPmyWoNPi0BAAAAOJYFD/rrOuUUc42gVtKlSxe98sor9bb36dNH\nr732Wr3tnvX8AQAAAIQmeNB/332Bz/fskfzG2wMAAABoe++8Yw7CmT5d2r5d6tXLHIUfiuBjY+Li\nfF82mzRggLRmzVEWGQAAAECo7r7bHGG/bJn5/LXXfGvthCL4nf65c808v//7n9mU6NfPTNgFAAAA\nICw2bJA++UQaNcp8/rvfScOGhX588KD/zTelm2+WuneX3G5p/35z3f5LLmlhkQEAAAA0R2Ki+d0z\nnKe2VqqpCf344EH//PnSl19KjiMr7u7bJ119NUE/AAAAECZDh0q/+IUZij/1lPTGG9LIkaEfHzzo\nT0jwBfySmQWgQ4fmlxQAAABAizz0kLR6tTnK/ocfpNtvl668MvTjgwf9KSnSk09KF15oPl+7VkqN\njbTGAAAAQHsxerR09tlmzlxJ2r1b6tEjtGODB/1LlpjLdv75z+YgoiFDWjU5FwAAAICm3XKLmZW3\nc2fzuWGYofnu3aEdHzzodzqlO+6Q+vY1n3/+eeBwHwAAAABt6l//kgoLpY4dW3Z88HX6771XeuQR\n3/NHH5VmzWrZ2QAAAAA0W//+UnV1y48Pfqd//Xrp4499z1eulIYPb/kZAQAAADTLz35mZuA9+WQz\nX67HBx+EdnzwoL+qyvxKSDCfl5YeXTMDAAAAQLPcc4/0xBNSt24tOz540D9tmtmkGDjQzALwn/9I\n8+a17GwAAAAAmu0nP5Guv77lxwcP+qdMMZfr/M9/zCnCf/iDlJbW8jMCAAAAaJaTTzaD/mHDAof3\n/PKXoR0fPOiXzCE9nhV7vvtOuvVW6dtvm1lUAAAAAC1RUCBZrVJubuD21gv6b7tNWrdO2r9f6tNH\n2r5duvPOFhQVAAAAQEssW3Z0xwdfsvPTT827+qedZg7xefddqazs6M4KAAAAIKhrrjG/d+9uZt+t\n+xWq4Hf6O3Qwv1dWmqm/zjyTO/0AAABAGDzzjPn9X/+q/9rhw6G/T/Cgv18/6bnnpHPOMSf09usn\nFReHfgYAAAAALZKVZX6/6SZp7drA1846yxyIE4rgQf8LL0hFRVJ6urRihZSXZy4UCgAAAKBNvfqq\n9MAD0q5dgcN5qqt9DYJQBA/6LRbJbjcfT5jQzGICAAAAaKnrrpPGjzdX0b//ft92q1U67rjQ3ye0\nJTsBAAAARERcnPSnPx3dewRfvQcAAABAu0bQDwAAAMQ4gn4AAAAgxhH0AwAAADGOoB8AAACIcQT9\nAAAAQIwj6AcAAABiHEE/AAAAEOMI+gEAAIAYR9APAAAAxDiCfgAAACDGEfQDAAAAMY6gHwAAAIhx\nBP0AAABAjCPoBwAAAGIcQT8AAAAQ4wj6AQAAgBhH0A8AAADEOIJ+AAAAIMYR9AMAAAAxjqAfAAAA\niHEE/QAAAECYVFRU6IILLtAbb7wR1vMS9AMAAABh8vzzz6tTp05hPy9BPwAAABAG27dv17Zt2zRy\n5Miwn5ugHwAAAAiDxx57TLNmzYrIuW0ROWuYZWQkyWaLi8i5HY7UiJwXjaNOohP1En2ok+jUburF\nMLwPHREsRji0mzppY/HxcVH1WURTWSTpzTff1Gmnnabu3btH5PzHRNBfVFQWkfM6HKlyuUoicm40\njDqJTtRL9KFOolN7qheHM8372JV/KIIlaVvtqU7aVoqqq91yuSITc9UVqXppqqGxfv167dmzR+vX\nr9f+/fuVkJCgLl26aOjQoWEp2zER9AMAAACR9PTTT3sfL1iwQNnZ2WEL+CXG9AMAAAAxjzv9AAAA\nQBjNmDEj7OfkTj8AAACOit+8bUQpgn4AAAAcNYsl0iVAUwj6AQAAgBgX9qD/008/1ZAhQ/Thhx96\nt3333XcaP3682vUNhAAAH3lJREFUxo8fr7lz53q3L168WFdffbVycnK0YcMGSVJJSYmmTp2qa6+9\nVlOmTFFxcXG4LwEAAABoV8Ia9O/evVvLli3TGWecEbD9oYce0uzZs7VixQqVlpZqw4YN2rNnj9as\nWaPXXntNL774oh555BHV1tZq+fLlGjRokP7yl79o9OjRWrRoUTgvAQAAAGh3whr0OxwOLVy4UKmp\nvsQFVVVV2rt3r/r37y9JGjVqlHJzc7Vx40aNGDFCCQkJstvtys7O1rZt25Sbm6sLL7wwYF8AAAAA\njQvrkp2JiYn1thUVFSktzZe1LzMzUy6XS+np6bLb7d7tdrtdLpdLBQUF3u2ZmZnKz88Pet6MjCTZ\nbHGtcAXNF20poEGdRCvqJfpQJ9GpPdZLeyxzc8T69YUqPj4uqj6LaCpLNGizoH/VqlVatWpVwLYZ\nM2ZoxIgRTR5nNLLmU0PbG9u3rqKiyKSEJjV39KFOohP1En2ok+jUnurF4fe4vZS5JdpTnbStFFVX\nu+VyRSbmqitS9RLNDY02C/pzcnKUk5MTdD+73R4wGTcvL09Op1NOp1M7duxocLvL5VJqaqp3GwAA\nAIDGRXzJzvj4ePXq1UubNm2SJK1bt04jRozQ4MGDtX79elVVVSkvL0/5+fnq06ePhg0bprVr1wbs\nCwAAAKBxYR3Tv379ei1ZskTff/+9Nm/erFdeeUVLly7V7Nmzdd9998ntdmvAgAEaOnSoJGncuHGa\nOHGiLBaL5s2bJ6vVqkmTJmnmzJmaMGGC0tLSNH/+/HBeAgAAANDuWIxQB8a3Y5Eaa8c4v+hDnUQn\n6iX6UCfRqT3Vi8PpW6TDlX8ogiVpW+2pTtpSdnaKBgxwa80axvRHq4gP7wEAAADQtgj6AQAAgBhH\n0A8AAADEOIJ+AAAAIMYR9AMAAAAxjqAfAAAAiHEE/QAAAECMI+gHAAAAYhxBPwAAABDjCPoBAACA\nGEfQDwAAAMQ4gn4AAAAgxhH0AwAAADGOoB8AAACIcQT9AAAAQIwj6AcAAABiHEE/AAAAEOMI+gEA\nAIAYR9APAAAAxDiCfgAAACDGEfQDAAAAMY6gHwAAAIhxBP0AAABAjCPoBwAAAGIcQT8AAAAQ4wj6\nAQAAgBhH0A8AAADEOIJ+AAAAIMYR9AMAAAAxjqAfAAAAiHEE/QAAAECMI+gHAAAAYhxBPwAAABDj\nCPoBAACAGEfQDwAAAMQ4gn4AAAAgxhH0AwAAADGOoB8AAACIcQT9AAAAQIwj6AcAAABiHEE/AAAA\nEOMI+gEAAIAYR9APAAAAxDiCfgAAACDGEfQDAAAAMY6gHwAAAIhxBP0AAABAjCPoBwAAbcbtcEa6\nCABE0A8AANpATa/ekqSqIcMiXBIgujz++OO65pprdNVVV2ndunVhO68tbGcCAADHDhshBlDXJ598\noq1bt2rlypUqKirSFVdcodGjR4fl3PxGAgAAAGFw1llnqX///pKktLQ0lZeXq7a2VnFxcW1+bob3\nAAAAAGEQFxenpKQkSdLq1at1zjnnhCXglySLYRhGWM4UQTU1tbLZwvOBAgAAHGsSEqSBA6V//zvS\nJWkf3nvvPb344otaunSpUlNTw3LOY2J4T1FRWUTO63CkyuUqici50TDqJDpRL9GHOolO7aleMoaf\nJduW/6nisitUsnh5pIvTZtpTnbStFFVXu+VyRSbmqitS9eJwBA/gP/roI73wwgtavHhx2AJ+KQLD\nez799FMNGTJEH374oXfbpEmTdNVVV2nSpEmaNGmSvv76a0nS4sWLdfXVVysnJ0cbNmyQJJWUlGjq\n1Km69tprNWXKFBUXF4f7EgAAAIBmKykp0eOPP64XX3xR6enpYT13WO/07969W8uWLdMZZ5xR77VH\nHnlEJ554ovf5nj17tGbNGq1YsUKlpaWaMGGChg8fruXLl2vQoEG68cYbtXLlSi1atEgzZ84M52UA\nAAAAzbZmzRoVFRXpN7/5jXfbY489puOOO67Nzx3WoN/hcGjhwoW69957g+67ceNGjRgxQgkJCbLb\n7crOzta2bduUm5urhx9+WJI0atQoTZs2ra2LDQAAWsgS+1MHgZBdc801uuaaayJy7rAG/YmJiY2+\n9swzz6ioqEi9e/fW7NmzVVBQILvd7n3dbrfL5XIFbM/MzFR+fn7Q82ZkJEVsIm8oY7sQXtRJdKJe\nog91Ep3aTb0c+b/boYOt/ZS5hWL9+kIVHx8XVZ9FNJUlGrRZ0L9q1SqtWrUqYNuMGTM0YsSIevtO\nnjxZ/fr1U48ePTR37ly9+uqr9fZpaJGhUBceYiIvPKiT6ES9RB/qJDq1p3rJqHXLJqmyskaH2kmZ\nW6I91UnbYiKv57zRqs2C/pycHOXk5IS074UXXuh9fN5552nNmjU6++yztWPHDu/2vLw8OZ1OOZ1O\nuVwupaamercBAAAAaFzEk3MZhqEbbrhBhw4dkmSO5e/bt68GDx6s9evXq6qqSnl5ecrPz1efPn00\nbNgwrV27VpK0bt26BnsOAAAAAPiEdUz/+vXrtWTJEn3//ffavHmzXnnlFS1dulTjxo3TDTfcoMTE\nRGVlZWnGjBlKTEzUuHHjNHHiRFksFs2bN09Wq1WTJk3SzJkzNWHCBKWlpWn+/PnhvAQAAACg3Qlr\n0D9y5EiNHDmy3vYxY8ZozJgx9bZ71u33l5ycrOeee66tiggAAADEnIgP7wEAAADQtgj6AQAAgBhH\n0A8AAADEOIJ+AAAAIMYR9AMAAAAxjqC/jVRVSXv2SKWlkS4JAAAAjnVhXbLzWDJhQqL++U9JSlVS\nkqHOnc0vp9OtzExDWVmGMjPNL6fT9z093VBcXKRLDwBAKzGMSJcAgAj628wNN1Sre3eb9u2r0YED\nFuXnW/T111ZVVTUd0dtshjIyfA0Bh8NsLDgcZoPBbjdfczjM1zt2DNMFAQDQHBZLpEsAwA9BfxsZ\nO7ZGv/iF5HKVe7cZhlRcLBUUWFVQYJHLZVFBgdkgKCjwPLaqsNCinTut2rw5+B/MTp2MIw0Bt19v\nghHQOOjc2S2n01BqKn+DAQAAjkUE/WFksUgZGVJGhlt9+wbfv7xcKiy0KC/PcqS3wKoDB8zGgqfB\n4Hm8a1ec3O6mI/r4eF/Pgf8QI0+DwX9b586GbPx0AAAAxATCuiiWmChlZxvKzvaMh6xtdF+3Wyoo\nsHgbBZ7HngaDy2X19iZs2WJVRUXwW/52u9s7tMgz58DhMHsRsrLc3uFHDoeh5ORWumgAAAC0OoL+\nGGG1Sk6nGZiffHLw/Q8flvLzzQbCgQNW5edbvHMP/BsMLpdVW7YEbyAkJgbOP3A43AGTlD2NA08v\nAsOMAAAAwoeg/xiVnCz17GmoZ09DkrvJfaurFTCsyNM4cLms3see7199ZVV1ddMRfVyc2Vvg3xDw\nDCtyONwBPQt2u6HExFa8cAAAgGMQQT+Cio+XunQx1KVL8GXX3G4zN8H+/eaEZJfL13uQn29RYaE5\nN8HlsmjPHqu++Sb4Lf/UVF8DoO6wIk/joHNncyhSWprZ6wEAAAAfgn60KqtVSkuT0tKa7j3wKC+v\n34uQl2dtsGdh506ramubXvK0QwffykWexoH/Uqd9+kg2m1VOp7k0akJCa1w1AABAdCPoR0QlJkrd\nuhnq1i14L0JtrVRUFDicKLBx4JubsGWLVeXljfUi+GYdZ2a6vfMM/JOkNdRgSEpiyVMAANA+EfSj\n3YiLkzcXQShKS80VjTw9BwUFFpWVddTu3VXeVY08uRFCmazsyazsm4fgrpdV2TM3wW43GGYEAACi\nBkE/YlZKipSSYuiEE3xLnTocHeVyVdbbt6ZG3qVOPfMP/Ica+c9J+OILq2pqmm4kWK2+fAjmnANf\ng6FzZ7e3ceB5vUOHVr98AAAAL4J+QJLNJnXtaqhr1+C9CIYhHTyoI8nRrA1kVfYlU9uzx6pvvw3e\ni5CS4uklcNdZzci3FKrTab7WqRPDjAC0I0ZovbMA2hZBP9BMFouUni6lpxvq06fxhGkeFRUKmJBs\nNg6s3p4F/wbDjh3BMysnJARmVfZf1Sgw07L5Wnx8a105ADQDdyeAqELQD7Sxjh2l7t0Nde8e2mTl\n4mJfJmX/rMpmD4L1yDwFi7Zvt+qrr5r+p2qxGEpPl7eXwNNzYDYUPI0Dt3f4UXIy/6cBAIhFBP1A\nFImLk/eOfTCGYWZW9vUYWAN6EwoL/RsMVm3ZIhlG0xF9UlLdeQjuBrMqO52G0tMNxTW9gioAAIgS\nBP1AO2Wx+CYrh5JZuarKXPLU0xComxvBTJxmNhi++sqqqqrQMyvXn4fgDhh+5HAY6tixFS8eAAA0\nC0E/cIxISJCyssxAPBjDkIqLPYnTAucf+BoMZs/Crl2hZVZOS/MNJ6rbc3DiiZLNFieHwy2Hw8ys\nzDAjAABaD0E/gHosFikjQ8rIaN5kZc+wov37A7Mq+zcYdu6Ma2SYUZL3UXy8L/eBJ3Ga/4TlukON\nGGYEAEDTCPoBHLX6k5Ubbyi43ao3SbmiIlE7dlTWy42wdatVX34Z/JZ/Rkbg/APPUCPPUqeerMqd\nOxtKSWmliwYAoB0h6AcQVlarvHfqTzrJ3OZwSC5XVb19DUMqK5NfBmWrXx6E+nMTtm4Nfss/MbF+\nVmVPb0LdBkNGBr0IAIDYQNAPIGpZLFJystSrl6FevUKbrFxYWHeSsm+pU/8GwzffWPX5501H9Dab\nGfh7EqT5sirXbzBkZBhKSmry7QAAiBiCfgAxIyFB6tLFUJcuwScru91SSYkCljr1n3/gmZtQWGjR\nnj2hT1b2JE7zTFj2DCvyn4eQleVWaiqTlQEA4UPQD+CYZLVKnTpJnTq51bdv8P3LyuRd1tQz56Bu\nbgRPg2HnTqvc7qZ7ERISAjMn+4YVNZwbgczKAICjQdAPACFISjKTl3XrFlovgqdh4N9z0NDchP/9\nz6qKiuC3/DMzAyck+89D8PQqeOYkJCXRiwAACETQDwCtzH+ycihKS+U358DaYFblvDyzsbB1a/Bo\n3jNZ2dc4qL/UqadnwW43ZLUe7RUDTTBC+z0A0LYI+gEgwjyZlU84Ifhk5ZoaX04Ez/eGhhoVFlr0\n5ZdW1dSEllnZv+fA4TDUs6fUsaMtYDWjzp3JrIzmoLsJiCYE/QDQjthsUteuhrp2DS2z8qFD5mRl\n/6FGnse+FY6s2rvXqm+/rRukJdZ7z5QUs5fAbje8GZQ9PQe+Cctmz0KnTgwzAoBoQdAPADHKYmne\nZOXycnl7CcrLk7VjR3m9BoMnN8LOnVbV1jY9WblDB0/jwD+DcsNDjTIyDCUktNKFAwDqIegHAEiS\nEhN9mZXNhGk1je5bUyMdPOg/Obn+UqeefAnbtgXPrGyxGEpPl7eXIDCrsmeFI1+DITmZXgQAaA6C\nfgBAs9ls8q4YFIqSEvllULb6JU6ru8KRVVu3SobRdESflORb6tScb+CuN0nZ02BITyezMgAQ9AMA\n2lxqqpSaGtpk5epqXwPhwAGL9u+3HOk5sHq3eYYaffmlVdXVTTcQrFZf0rTApU59jQXf8CNDHTq0\n4oUDQJQg6AcARJX4+NAzKxuGVFwsby+Bf3Zlc6iR+b2gwKo9exqarFxfamr9pU79cyN4Xuvc2VBa\nGsOMALQPBP0AgHbLYpEyMqSMDEO9e9cG3b+yUgETks1lT325ETzLoB44YNGnn8YFHWYUHx/YS+Bp\nHHjmJni2eRoONv7rAogQ/vwAAI4ZHTr4JisHU1srFRXVnaTs6TkIzI2wfbtVX30VfLJyp07yTkiu\nP9TIUGam2/s4JaW1rhoACPoBAGhQXJy8d+5POqnpfQ1DKiuT8vLMRkHdpU79syu7XBb973/BZxYn\nJvp6DjwNheOPl5KT470NBc93JisDCIagHwCAo2SxSMnJUq9ehnr1Cj5ZuarK7EXwT5bmGXbkaTB4\nGgrffGPV55/7R/T10yLHxZm5DgKHGAU2GDzPMzMNJdbPuwYgxhH0AwAQZgkJUlaWubxoMJ7JygcO\nWFRTk6ItW8q98w98qxmZQ41277bqm2+CzyxOSwtMlua/1KnT6Z84za3UVCYrA7GAoB8AgCjmP1nZ\n4ZBOOqnxpGmSmVnZ00tgDjUyJyt7ehP8Gwy7dsXJ7Q4+WdnTc+BpHJiZlt3e3gP/HoZ6k5WN0HI5\nAGhbBP0AAMSQxEQpO9tQdnbwYNvtNnsQ/Fcz8iROqzs3YcsWqyoqgt/yt9vdstsNZe9bKqd2qvM3\nier0VILfPASzseB0mpmVAYQHQT8AAMcoq1Xe5GTBJitL0uHDqtM4sDbaYNhWNkjSIGmXpEcbfr+O\nHQOXNXU43N5My3VXOMrMNGS1tubVA8cWgn4AABCS5GQpOdmTWVmSGs+NkHLuOSr6Nk97h12l72c8\n5tc4sHobBwcOmL0IX30VWmZlc1hR4FAjT14Eh8PwNhjsdiYrA3UR9AMAgFYXb6nRcfpRmWk7dcJ5\nTSdOc7ulkhIpL8+qwkKzceCfG8HTOHC5rNq7N7TMyikp/sua+nIj+E9Y7tzZzKzcqROTlRH7CPoB\nAEBEWa1Sp05Sp05NL3XqUVYmvyRpgUudeoYaeSYz79plVW1t00kMEhKMBoYVuQPyIXjmIWRkGEpI\naI2rBsKLoB8AALQrSUlSUpKhbt1Cy6xcXNxwz4EvcZo55GjLFqvKy0ObrOxZrahbNyklpUOdHAlu\n7zyF5GR6ERAdCPoBAEDMiouTdyJwKEpLzRWN9u+3elc2qttg8PQsbNniieYbv/Xvyazs6zFwe+ck\n+GdV7tzZ7EUgszLaCkE/AADAESkp5nyA449veh6CJNXUSIaRqu++O1yv56CgILDB8MUXVtXUNH3L\n32LxTVL2TFj2NRjcAZOYnU5DHTq01lXjWEDQDwAA0AI2m+RwSPHxweciGIZ08KACljr1JEvzPPb0\nIPzwQ/MmK3fu7K7Xc+CfVTkzk8nK0eThhx/WF198IYvFotmzZ6t///5hOW9Yg/6amhrde++92r17\nt2pra3XXXXdp4MCB+u677zRv3jxJUr9+/XT//fdLkhYvXqy1a9fKYrFo+vTpOvfcc1VSUqI77rhD\nJSUlSkpK0pNPPqn09PRwXgYAAECzWCxSerqUnm6od+/gvQgVFfLmPvA0CFwuq1yuwKzKBw5YtHNn\n6JmV/YcaOZ3m3AP/IUZOpzmpOT6+ta4c/j799FPt2rVLK1eu1Pbt2zV79mytXLkyLOcOa9D/1ltv\nKTExUX/5y1+0detW3XPPPVq9erUeeughb0vnjjvu0IYNG9SrVy+tWbNGK1asUGlpqSZMmKDhw4dr\n+fLlGjRokG688UatXLlSixYt0syZM8N5GQAAAG2qY0epe3dD3bs3b7Kyp+fAl1XZooICq3f40fbt\nVn31VfBhRp06ydtL4JmU7MmNYLebDQbPtpSU1rrq2Jebm6sLLrhAktS7d28dPHhQpaWlSgnDhxjW\noP+yyy7T2LFjJUl2u13FxcWqqqrS3r17vV0bo0aNUm5urlwul0aMGKGEhATZ7XZlZ2dr27Ztys3N\n1cMPP+zdd9q0aeG8BAAAgKjSnMnKhuHLrOwZauTJjVA3q7I5WVkyjKYbCUlJRtDkajAVFBTolFNO\n8T632+1yuVyxF/TH+/UVLV++XGPHjlVRUZHS0tK82zMzM+VyuZSeni673e7d7vlQCgoKvNszMzOV\nn58f9LwZGUmy2SIzHd7hSI3IedE46iQ6US/RhzqJTu2mXjZ/JUnqIMkR2ZK0uXZTJ3569gxtv6oq\nqbBQ+vFHqaBAysuT9u83H+/bJ7lcUn6+RS6X9POfx0XVZxFNZWmMYYS2qlRraLOgf9WqVVq1alXA\nthkzZmjEiBF69dVXtXnzZr3wwgsqLCwM2Kexi29oezg/KAAAgGNNQoLUpYv5haPndDpVUFDgfZ6f\nny+HIzzN4jYL+nNycpSTk1Nv+6pVq/TBBx/oueeeU3x8vHeYj0deXp6cTqecTqd27NjR4HaXy6XU\n1FTvtmCKispa56KayeFIlctVEpFzo2HUSXSiXqIPdRKdqJfoQ51Ep0jVS7DehWHDhmnBggUaP368\nNm/eLKfTGZahPZJkDctZjtizZ49WrFihhQsXqsORxWXj4+PVq1cvbdq0SZK0bt06jRgxQoMHD9b6\n9etVVVWlvLw85efnq0+fPho2bJjWrl0bsC8AAAAQ7c444wydcsopGj9+vH7/+99r7ty5YTt3WMf0\nr1q1SsXFxZo6dap325IlSzR79mzdd999crvdGjBggIYOHSpJGjdunCZOnCiLxaJ58+bJarVq0qRJ\nmjlzpiZMmKC0tDTNnz8/nJcAAAAAtNidd94ZkfNajGNgYHykut3o8os+1El0ol6iD3USnaiX6EOd\nRKdoHd4TSWEd3gMAAAAg/Aj6AQAAgBhH0A8AAADEOIJ+AAAAIMYR9AMAAAAxjqAfAAAAiHEE/QAA\nAECMI+gHAAAAYhxBPwAAABDjCPoBAACAGEfQDwAAAMQ4gn4AAAAgxhH0AwAAADGOoB8AAACIcRbD\nMIxIFwIAAABA2+FOPwAAABDjCPoBAACAGEfQDwAAAMQ4gn4AAAAgxhH0AwAAADGOoB8AAACIcbZI\nFyBWPfzww/riiy9ksVg0e/Zs9e/fP9JFiklbtmzRLbfcohtuuEETJ07Ujz/+qLvuuku1tbVyOBya\nP3++EhIS9Pbbb2v58uWyWq0aN26ccnJyVF1drVmzZmnfvn2Ki4vTI488ou7du+u7777TvHnzJEn9\n+vXT/fffH9mLbGcef/xx/fe//1VNTY1uuukmnXrqqdRJhJWXl2vWrFk6cOCAKisrdcstt+ikk06i\nXqJARUWFxo4dq1tuuUVDhgyhTiJs48aNuu2229S3b19J0oknnqgbb7yReomwt99+W4sXL5bNZtOt\nt96qfv36USctYaDVbdy40Zg6daphGIaxbds2Y9y4cREuUWw6fPiwMXHiRGPOnDnGK6+8YhiGYcya\nNctYs2aNYRiG8eSTTxqvvvqqcfjwYWP06NHGoUOHjPLycuPSSy81ioqKjDfeeMOYN2+eYRiG8dFH\nHxm33XabYRiGMXHiROOLL74wDMMwbr/9dmP9+vURuLr2KTc317jxxhsNwzCMwsJC49xzz6VOosA7\n77xjvPTSS4ZhGMYPP/xgjB49mnqJEk899ZRx5ZVXGq+//jp1EgU++eQTY8aMGQHbqJfIKiwsNEaP\nHm2UlJQYeXl5xpw5c6iTFmJ4TxvIzc3VBRdcIEnq3bu3Dh48qNLS0giXKvYkJCRo0aJFcjqd3m0b\nN27U+eefL0kaNWqUcnNz9cUXX+jUU09VamqqOnbsqDPOOEOfffaZcnNzdeGFF0qShg4dqs8++0xV\nVVXau3evt2fG8x4IzVlnnaU//vGPkqS0tDSVl5dTJ1FgzJgx+tWvfiVJ+vHHH5WVlUW9RIHt27dr\n27ZtGjlypCT+fkUr6iWycnNzNWTIEKWkpMjpdOrBBx+kTlqIoL8NFBQUKCMjw/vcbrfL5XJFsESx\nyWazqWPHjgHbysvLlZCQIEnKzMyUy+VSQUGB7Ha7dx9Pffhvt1qtslgsKigoUFpamndfz3sgNHFx\ncUpKSpIkrV69Wueccw51EkXGjx+vO++8U7Nnz6ZeosBjjz2mWbNmeZ9TJ9Fh27ZtmjZtmq699lp9\n/PHH1EuE/fDDD6qoqNC0adM0YcIE5ebmUictxJj+MDAMI9JFOCY19rk3Zzt11zLvvfeeVq9eraVL\nl2r06NHe7dRJZK1YsULffvutZs6cGfA5Ui/h9+abb+q0005T9+7dG3ydOomME044QdOnT9cll1yi\nPXv2aPLkyaqtrfW+Tr1ERnFxsRYuXKh9+/Zp8uTJ/P1qIe70twGn06mCggLv8/z8fDkcjgiW6NiR\nlJSkiooKSVJeXp6cTmeD9eHZ7mnZV1dXyzAMORwOFRcXe/f1vAdC99FHH+mFF17QokWLlJqaSp1E\nga+//lo//vijJOnkk09WbW2tkpOTqZcIWr9+vd5//32NGzdOq1at0nPPPcfvShTIysrSmDFjZLFY\n1KNHD3Xu3FkHDx6kXiIoMzNTp59+umw2m3r06KHk5GT+frUQQX8bGDZsmP7xj39IkjZv3iyn06mU\nlJQIl+rYMHToUO9nv27dOo0YMUIDBgzQV199pUOHDunw4cP67LPPNHDgQA0bNkxr166VJH344Yc6\n++yzFR8fr169emnTpk0B74HQlJSU6PHHH9eLL76o9PR0SdRJNNi0aZOWLl0qyRx+WFZWRr1E2NNP\nP63XX39df/3rX5WTk6NbbrmFOokCb7/9tpYsWSJJcrlcOnDggK688krqJYKGDx+uTz75RG63W0VF\nRfz9OgoW41jp0wizJ554Qps2bZLFYtHcuXN10kknRbpIMefrr7/WY489pr1798pmsykrK0tPPPGE\nZs2apcrKSh133HF65JFHFB8fr7Vr12rJkiWyWCyaOHGiLrvsMtXW1mrOnDnauXOnEhIS9Oijj6pr\n167atm2b7rvvPrndbg0YMED33HNPpC+13Vi5cqUWLFignj17erc9+uijmjNnDnUSQRUVFbr33nv1\n448/qqKiQtOnT9dPf/pT3X333dRLFFiwYIGys7M1fPhw6iTCSktLdeedd+rQoUOqrq7W9OnTdfLJ\nJ1MvEbZixQqtXr1aknTzzTfr1FNPpU5agKAfAAAAiHEM7wEAAABiHEE/AAAAEOMI+gEAAIAYR9AP\nAAAAxDiCfgAAACDGEfQDAAAAMY6gHwAAAIhxBP0AAABAjPv/Afsq6Jcm3ZIyAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<Figure size 864x576 with 2 Axes>"
      ]
     },
     "metadata": {
      "tags": []
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "#visualize 1% of samples data, first 100 datapoints\n",
    "train_ad_sample_df = train['acoustic_data'].values[::100]\n",
    "train_ttf_sample_df = train['time_to_failure'].values[::100]\n",
    "\n",
    "#function for plotting based on both features\n",
    "def plot_acc_ttf_data(train_ad_sample_df, train_ttf_sample_df, title=\"Acoustic data and time to failure: 1% sampled data\"):\n",
    "    fig, ax1 = plt.subplots(figsize=(12, 8))\n",
    "    plt.title(title)\n",
    "    plt.plot(train_ad_sample_df, color='r')\n",
    "    ax1.set_ylabel('acoustic data', color='r')\n",
    "    plt.legend(['acoustic data'], loc=(0.01, 0.95))\n",
    "    ax2 = ax1.twinx()\n",
    "    plt.plot(train_ttf_sample_df, color='b')\n",
    "    ax2.set_ylabel('time to failure', color='b')\n",
    "    plt.legend(['time to failure'], loc=(0.01, 0.9))\n",
    "    plt.grid(True)\n",
    "\n",
    "plot_acc_ttf_data(train_ad_sample_df, train_ttf_sample_df)\n",
    "del train_ad_sample_df\n",
    "del train_ttf_sample_df"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "nosoupBSzMZV"
   },
   "source": [
    "![alt text](https://www.kaggleusercontent.com/kf/10364757/eyJhbGciOiJkaXIiLCJlbmMiOiJBMTI4Q0JDLUhTMjU2In0..onsybg1U7xvMy-GjJdYtTA.RUVJoAZ7sxCnRNmM1WTjXMjmsl7raHI_FzIYc0vh0poV3hcKVMrr2GRPCwlr_urWqByjmahePNQlEFEsBBUO2meVZXzX-HjRQuU40ySqElezoc9HLF7nGBMCKPSYt-HPZAsGHnXBtkXV8qVOZlu5BWddQp7wthNJ5JkSftT41GM.vS5mBigypyc7eNri4zC2rQ/__results___files/__results___12_0.png)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "ylsjYactdeUN"
   },
   "source": [
    "#### Also, shoutout to Anton Loss. He turned this data into sound, then compared it with dubstep.\n",
    "\n",
    "[Anton's work here](https://www.kaggle.com/avloss/audio-analysis-with-animation)\n",
    "\n",
    "#####  Anton's hypothesis: perhaps just like a dubstep song, we can hear the build-up of an earthquake before the drop\n",
    "1. The data looks like sound waves, it oscillates! Thus, he converted it into an audio file.\n",
    "2. He then denoised the audio file\n",
    "3. Next, he animated the wave\n",
    "4. Finally, he compared it to a dubstep song\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 63,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 261
    },
    "colab_type": "code",
    "id": "vvfw6YRZ0MoE",
    "outputId": "ac5dc80c-2a92-4553-a9a5-558e5bcaa59c"
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<video width=\"320\" height=\"240\" controls>\n",
       "  <source src=\"https://www.kaggleusercontent.com/kf/10317300/eyJhbGciOiJkaXIiLCJlbmMiOiJBMTI4Q0JDLUhTMjU2In0..i3D--jDnUYfbn8vMyX3RVg.wV-Ek57ve9b_f8UthEd5YAE5rqBUgeMkXYmWJqcoaJgldO5d7afEsXTYoyC395ZJcMiXqrqeuxoaStyStAQLJFGH0KtzE-4sGt41imaqW5Y9tcu7XH8UqCgDlrUjWlK_PqM1akZHI9nohY3UNahoMN_2Z4u1ORWZ3u1iLtQv9jk.Cyv4iN3TfPCuUP5G7v47Bw/animation_with_sound.mp4\" type=\"video/mp4\">\n",
       "</video>\n",
       "\n",
       "<video width=\"320\" height=\"240\" controls>\n",
       "  <source src=\"https://www.kaggleusercontent.com/kf/10317300/eyJhbGciOiJkaXIiLCJlbmMiOiJBMTI4Q0JDLUhTMjU2In0..i3D--jDnUYfbn8vMyX3RVg.wV-Ek57ve9b_f8UthEd5YAE5rqBUgeMkXYmWJqcoaJgldO5d7afEsXTYoyC395ZJcMiXqrqeuxoaStyStAQLJFGH0KtzE-4sGt41imaqW5Y9tcu7XH8UqCgDlrUjWlK_PqM1akZHI9nohY3UNahoMN_2Z4u1ORWZ3u1iLtQv9jk.Cyv4iN3TfPCuUP5G7v47Bw/animation_with_dubstep.mp4\" type=\"video/mp4\">\n",
       "</video>"
      ],
      "text/plain": [
       "<IPython.core.display.HTML object>"
      ]
     },
     "metadata": {
      "tags": []
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "%%HTML\n",
    "<video width=\"320\" height=\"240\" controls>\n",
    "  <source src=\"https://www.kaggleusercontent.com/kf/10317300/eyJhbGciOiJkaXIiLCJlbmMiOiJBMTI4Q0JDLUhTMjU2In0..i3D--jDnUYfbn8vMyX3RVg.wV-Ek57ve9b_f8UthEd5YAE5rqBUgeMkXYmWJqcoaJgldO5d7afEsXTYoyC395ZJcMiXqrqeuxoaStyStAQLJFGH0KtzE-4sGt41imaqW5Y9tcu7XH8UqCgDlrUjWlK_PqM1akZHI9nohY3UNahoMN_2Z4u1ORWZ3u1iLtQv9jk.Cyv4iN3TfPCuUP5G7v47Bw/animation_with_sound.mp4\" type=\"video/mp4\">\n",
    "</video>\n",
    "\n",
    "<video width=\"320\" height=\"240\" controls>\n",
    "  <source src=\"https://www.kaggleusercontent.com/kf/10317300/eyJhbGciOiJkaXIiLCJlbmMiOiJBMTI4Q0JDLUhTMjU2In0..i3D--jDnUYfbn8vMyX3RVg.wV-Ek57ve9b_f8UthEd5YAE5rqBUgeMkXYmWJqcoaJgldO5d7afEsXTYoyC395ZJcMiXqrqeuxoaStyStAQLJFGH0KtzE-4sGt41imaqW5Y9tcu7XH8UqCgDlrUjWlK_PqM1akZHI9nohY3UNahoMN_2Z4u1ORWZ3u1iLtQv9jk.Cyv4iN3TfPCuUP5G7v47Bw/animation_with_dubstep.mp4\" type=\"video/mp4\">\n",
    "</video>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "pgPdreNUefZl"
   },
   "source": [
    "# Step 4 - Feature Engineering"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 0,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "7QfB_uNr01V8"
   },
   "outputs": [],
   "source": [
    "# Step 4 - Feature Engineering and signifiance of these statistical features\n",
    "\n",
    "#lets create a function to generate some statistical features based on the training data\n",
    "def gen_features(X):\n",
    "    strain = []\n",
    "    strain.append(X.mean())\n",
    "    strain.append(X.std())\n",
    "    strain.append(X.min())\n",
    "    strain.append(X.max())\n",
    "    strain.append(X.kurtosis())\n",
    "    strain.append(X.skew())\n",
    "    strain.append(np.quantile(X,0.01))\n",
    "    strain.append(np.quantile(X,0.05))\n",
    "    strain.append(np.quantile(X,0.95))\n",
    "    strain.append(np.quantile(X,0.99))\n",
    "    strain.append(np.abs(X).max())\n",
    "    strain.append(np.abs(X).mean())\n",
    "    strain.append(np.abs(X).std())\n",
    "    return pd.Series(strain)\n",
    "\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "D8Kc1RZg4-pk"
   },
   "source": [
    "![alt text](https://qph.fs.quoracdn.net/main-qimg-29a4925034e075f16e1c743a4b3dda8b)\n",
    "![alt text](https://qph.fs.quoracdn.net/main-qimg-cb5a6703bdfd5e2a15cf8865607a1590)\n",
    "![alt text](https://encrypted-tbn0.gstatic.com/images?q=tbn:ANd9GcTkRGxxt76XIth4pjG5Itp-k6MNt5qPaejI_gyENu17KkGC2oFq)\n",
    "![alt text](https://cramster-image.s3.amazonaws.com/definitions/DC-2195V1.png)\n",
    "![alt text](https://i.stack.imgur.com/KBQLN.jpg)\n",
    "- Kurtosis is a measure of whether the data are heavy-tailed or light-tailed relative to a normal distribution. \n",
    "- Data sets with high kurtosis tend to have heavy tails, or outliers. Data sets with low kurtosis tend to have light tails, or lack of outliers\n",
    "![alt text](https://study.com/cimages/multimages/16/skewness_posnegpict.jpg)\n",
    "- skewness is a measure of the asymmetry of the probability distribution of a real-valued random variable about its mean. The skewness value can be positive or negative, or undefined.\n",
    "![alt text](https://upload.wikimedia.org/wikipedia/commons/5/5e/Iqr_with_quantile.png)\n",
    "- Quantiles are cut points dividing the range of a probability distribution into continuous intervals with equal probabilities, or dividing the observations in a sample in the same way. \n",
    "\n",
    "\n",
    "![alt text](https://www.researchgate.net/publication/304403489/figure/fig4/AS:427142036103170@1478850005427/Statistical-features.png\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 0,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "OtK1gGWE2sTF"
   },
   "outputs": [],
   "source": [
    "train = pd.read_csv('train.csv', iterator=True, chunksize=150_000, dtype={'acoustic_data': np.int16, 'time_to_failure': np.float64})\n",
    "\n",
    "X_train = pd.DataFrame()\n",
    "y_train = pd.Series()\n",
    "for df in train:\n",
    "    ch = gen_features(df['acoustic_data'])\n",
    "    X_train = X_train.append(ch, ignore_index=True)\n",
    "    y_train = y_train.append(pd.Series(df['time_to_failure'].values[-1]))\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 64,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 317
    },
    "colab_type": "code",
    "id": "_01NoTlx4cWL",
    "outputId": "5f8f6b4c-2940-4802-87e7-3510266de60b"
   },
   "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>0</th>\n",
       "      <th>1</th>\n",
       "      <th>2</th>\n",
       "      <th>3</th>\n",
       "      <th>4</th>\n",
       "      <th>5</th>\n",
       "      <th>6</th>\n",
       "      <th>7</th>\n",
       "      <th>8</th>\n",
       "      <th>9</th>\n",
       "      <th>10</th>\n",
       "      <th>11</th>\n",
       "      <th>12</th>\n",
       "      <th>13</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>count</th>\n",
       "      <td>4195.000000</td>\n",
       "      <td>4195.000000</td>\n",
       "      <td>4195.000000</td>\n",
       "      <td>4195.000000</td>\n",
       "      <td>4195.000000</td>\n",
       "      <td>4195.000000</td>\n",
       "      <td>4195.000000</td>\n",
       "      <td>4195.000000</td>\n",
       "      <td>4195.000000</td>\n",
       "      <td>4195.000000</td>\n",
       "      <td>4195.000000</td>\n",
       "      <td>4195.000000</td>\n",
       "      <td>4195.000000</td>\n",
       "      <td>4195.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>mean</th>\n",
       "      <td>4.519475</td>\n",
       "      <td>6.547788</td>\n",
       "      <td>-149.190942</td>\n",
       "      <td>163.522288</td>\n",
       "      <td>3.482899</td>\n",
       "      <td>68.297997</td>\n",
       "      <td>0.125830</td>\n",
       "      <td>-11.224603</td>\n",
       "      <td>-2.184779</td>\n",
       "      <td>11.231716</td>\n",
       "      <td>20.321890</td>\n",
       "      <td>170.046246</td>\n",
       "      <td>5.547367</td>\n",
       "      <td>5.750165</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>std</th>\n",
       "      <td>0.256049</td>\n",
       "      <td>8.503939</td>\n",
       "      <td>265.087984</td>\n",
       "      <td>272.930331</td>\n",
       "      <td>1.621138</td>\n",
       "      <td>70.532565</td>\n",
       "      <td>0.477901</td>\n",
       "      <td>14.106852</td>\n",
       "      <td>2.346558</td>\n",
       "      <td>2.358067</td>\n",
       "      <td>14.225526</td>\n",
       "      <td>296.887015</td>\n",
       "      <td>1.517038</td>\n",
       "      <td>8.339211</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>min</th>\n",
       "      <td>3.596313</td>\n",
       "      <td>2.802720</td>\n",
       "      <td>-5515.000000</td>\n",
       "      <td>23.000000</td>\n",
       "      <td>2.199321</td>\n",
       "      <td>0.648602</td>\n",
       "      <td>-4.091826</td>\n",
       "      <td>-336.000000</td>\n",
       "      <td>-39.000000</td>\n",
       "      <td>9.000000</td>\n",
       "      <td>11.000000</td>\n",
       "      <td>23.000000</td>\n",
       "      <td>4.147707</td>\n",
       "      <td>2.589085</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25%</th>\n",
       "      <td>4.349497</td>\n",
       "      <td>4.478637</td>\n",
       "      <td>-154.000000</td>\n",
       "      <td>92.000000</td>\n",
       "      <td>2.820130</td>\n",
       "      <td>28.090227</td>\n",
       "      <td>-0.040779</td>\n",
       "      <td>-14.000000</td>\n",
       "      <td>-3.000000</td>\n",
       "      <td>10.000000</td>\n",
       "      <td>15.000000</td>\n",
       "      <td>94.000000</td>\n",
       "      <td>5.061843</td>\n",
       "      <td>3.862810</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>50%</th>\n",
       "      <td>4.522147</td>\n",
       "      <td>5.618798</td>\n",
       "      <td>-111.000000</td>\n",
       "      <td>123.000000</td>\n",
       "      <td>3.324506</td>\n",
       "      <td>45.816625</td>\n",
       "      <td>0.085620</td>\n",
       "      <td>-10.000000</td>\n",
       "      <td>-2.000000</td>\n",
       "      <td>11.000000</td>\n",
       "      <td>19.000000</td>\n",
       "      <td>127.000000</td>\n",
       "      <td>5.380853</td>\n",
       "      <td>4.781513</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>75%</th>\n",
       "      <td>4.693350</td>\n",
       "      <td>6.880904</td>\n",
       "      <td>-79.000000</td>\n",
       "      <td>170.000000</td>\n",
       "      <td>3.805475</td>\n",
       "      <td>78.664202</td>\n",
       "      <td>0.253930</td>\n",
       "      <td>-6.000000</td>\n",
       "      <td>-1.000000</td>\n",
       "      <td>12.000000</td>\n",
       "      <td>23.000000</td>\n",
       "      <td>175.000000</td>\n",
       "      <td>5.748553</td>\n",
       "      <td>5.887947</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>max</th>\n",
       "      <td>5.391993</td>\n",
       "      <td>153.703569</td>\n",
       "      <td>-15.000000</td>\n",
       "      <td>5444.000000</td>\n",
       "      <td>31.724966</td>\n",
       "      <td>631.158927</td>\n",
       "      <td>4.219429</td>\n",
       "      <td>-2.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>50.000000</td>\n",
       "      <td>337.000000</td>\n",
       "      <td>5515.000000</td>\n",
       "      <td>32.762073</td>\n",
       "      <td>150.432368</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                0            1            2            3            4   \\\n",
       "count  4195.000000  4195.000000  4195.000000  4195.000000  4195.000000   \n",
       "mean      4.519475     6.547788  -149.190942   163.522288     3.482899   \n",
       "std       0.256049     8.503939   265.087984   272.930331     1.621138   \n",
       "min       3.596313     2.802720 -5515.000000    23.000000     2.199321   \n",
       "25%       4.349497     4.478637  -154.000000    92.000000     2.820130   \n",
       "50%       4.522147     5.618798  -111.000000   123.000000     3.324506   \n",
       "75%       4.693350     6.880904   -79.000000   170.000000     3.805475   \n",
       "max       5.391993   153.703569   -15.000000  5444.000000    31.724966   \n",
       "\n",
       "                5            6            7            8            9   \\\n",
       "count  4195.000000  4195.000000  4195.000000  4195.000000  4195.000000   \n",
       "mean     68.297997     0.125830   -11.224603    -2.184779    11.231716   \n",
       "std      70.532565     0.477901    14.106852     2.346558     2.358067   \n",
       "min       0.648602    -4.091826  -336.000000   -39.000000     9.000000   \n",
       "25%      28.090227    -0.040779   -14.000000    -3.000000    10.000000   \n",
       "50%      45.816625     0.085620   -10.000000    -2.000000    11.000000   \n",
       "75%      78.664202     0.253930    -6.000000    -1.000000    12.000000   \n",
       "max     631.158927     4.219429    -2.000000     0.000000    50.000000   \n",
       "\n",
       "                10           11           12           13  \n",
       "count  4195.000000  4195.000000  4195.000000  4195.000000  \n",
       "mean     20.321890   170.046246     5.547367     5.750165  \n",
       "std      14.225526   296.887015     1.517038     8.339211  \n",
       "min      11.000000    23.000000     4.147707     2.589085  \n",
       "25%      15.000000    94.000000     5.061843     3.862810  \n",
       "50%      19.000000   127.000000     5.380853     4.781513  \n",
       "75%      23.000000   175.000000     5.748553     5.887947  \n",
       "max     337.000000  5515.000000    32.762073   150.432368  "
      ]
     },
     "execution_count": 64,
     "metadata": {
      "tags": []
     },
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X_train.describe()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "GqkMwNkaGLVm"
   },
   "source": [
    "# Step 5 - Implement Catboost Model\n",
    "\n",
    "- Yandex is Russian Google\n",
    "- Yandex uses \"gradient boosting\" a lot to power their services (music streaming, search, everything really)\n",
    "- Gradient boosting on decision trees is a form of machine learning that works by progressively training more complex models to maximize the accuracy of predictions. \n",
    "- It's particularly useful for predictive models that analyze ordered (continuous) data and categorical data. \n",
    "- It's  one of the most efficient ways to build ensemble models. The combination of gradient boosting with decision trees provides state-of-the-art results in many applications with structured data.\n",
    "- On the first iteration, the algorithm learns the first tree to reduce the training error, shown on left-hand image in figure 1. \n",
    "- This model usually has a significant error; it’s not a good idea to build very big trees in boosting since they overfit the data.\n",
    "- The right-hand image in figure 1 shows the second iteration, in which the algorithm learns one more tree to reduce the error made by the first tree. \n",
    "- The algorithm repeats this procedure until it builds a decent quality model\n",
    "\n",
    "![alt text](https://devblogs.nvidia.com/wp-content/uploads/2018/12/first-second-trees-625x357.png)\n",
    "\n",
    "\n",
    "- Gradient Boosting is a way to implement this idea for any continuous objective function. \n",
    "\n",
    "### Each step of Gradient Boosting combines two steps:\n",
    "\n",
    "- Step 1 - Computing gradients of the loss function we want to optimize for each input object\n",
    "- Step 2 - Learning the decision tree which predicts gradients of the loss function\n",
    "\n",
    "### ELI5 Time\n",
    "\n",
    "- Step 1 - We first model data with simple models and analyze data for errors. \n",
    "- Step 2 - These errors signify data points that are difficult to fit by a simple model. \n",
    "- Step 3 - Then for later models, we particularly focus on those hard to fit data to get them right. \n",
    "- Step 4 -  In the end, we combine all the predictors by giving some weights to each predictor.\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 54,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 34
    },
    "colab_type": "code",
    "id": "zUvIKHcA1JZ8",
    "outputId": "ac6ca531-334d-4226-dafe-5af61b9e2a84"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{'learn': {'MAE': 1.74357544138688}}"
      ]
     },
     "execution_count": 54,
     "metadata": {
      "tags": []
     },
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#Model #1 - Catboost\n",
    "\n",
    "train_pool = Pool(X_train, y_train)\n",
    "m = CatBoostRegressor(iterations=10000, loss_function='MAE', boosting_type='Ordered')\n",
    "m.fit(X_train, y_train, silent=True)\n",
    "m.best_score_"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "m-xmqVpvM2lM"
   },
   "source": [
    "# Step 6 - Implement Support Vector Machine + Radial Basis Function Kernel \n",
    "\n",
    "you might be thinking WTF YOOOOoooo. thats ok, breathe. here we go. \n",
    "\n",
    "### Need to learn a nonlinear decision boundary? Grab a Kernel\n",
    "\n",
    "- A very simple and intuitive way of thinking about kernels (at least for SVMs) is a similarity function. \n",
    "- Given two objects, the kernel outputs some similarity score. The objects can be anything starting from two integers, two real valued vectors, trees whatever provided that the kernel function knows how to compare them.\n",
    "\n",
    "- The arguably simplest example is the linear kernel, also called dot-product. Given two vectors, the similarity is the length of the projection of one vector on another.\n",
    "\n",
    "- Another interesting kernel examples is Gaussian kernel. Given two vectors, the similarity will diminish with the radius of σ. The distance between two objects is \"reweighted\" by this radius parameter.\n",
    "\n",
    "- The success of learning with kernels (again, at least for SVMs), very strongly depends on the choice of kernel. You can see a kernel as a compact representation of the knowledge about your classification problem. It is very often problem specific.\n",
    "\n",
    "![alt text](https://i.stack.imgur.com/FsYi6.png)\n",
    "![alt text](http://www.statsoft.com/textbook/graphics/SVMIntro3.gif)\n",
    "![alt text](https://slideplayer.com/slide/3397099/12/images/26/Support+vector+machines.jpg)\n",
    "![alt text](https://i.stack.imgur.com/rZyn7.png)\n",
    "![alt text](https://slideplayer.com/slide/5745650/19/images/95/Support+Vector+Machines.jpg)\n",
    "![alt text](https://image.slidesharecdn.com/svm-140807035301-phpapp01/95/support-vector-machine-without-tears-5-638.jpg?cb=1407384107)\n",
    "\n",
    "## grid search : pick a bunch of values of α -- (α1,α2,…), pick a bunch of values of β -- (β1,β2,…) and for each pair of values, evaluate the validation error function. Then pick the pair that gives the minimum value of the validation error function.\n",
    "\n",
    "The pairs (α1,β1),(α1,β2),…,(α2,β1),(α2,β2),… when plotted in space look like a grid, hence the name."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 56,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 51
    },
    "colab_type": "code",
    "id": "4g4VQh8vFtjH",
    "outputId": "2d20fa28-cf7a-4701-dde1-5952e42bd7c2"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Best CV score: -2.1651\n",
      "{'C': 2, 'gamma': 0.02}\n"
     ]
    }
   ],
   "source": [
    "#Model #2 - Support Vector Machine w/ RBF + Grid Search\n",
    "\n",
    "from sklearn.preprocessing import StandardScaler\n",
    "from sklearn.model_selection import GridSearchCV\n",
    "from sklearn.svm import NuSVR, SVR\n",
    "\n",
    "\n",
    "scaler = StandardScaler()\n",
    "scaler.fit(X_train)\n",
    "X_train_scaled = scaler.transform(X_train)\n",
    "\n",
    "parameters = [{'gamma': [0.001, 0.005, 0.01, 0.02, 0.05, 0.1],\n",
    "               'C': [0.1, 0.2, 0.25, 0.5, 1, 1.5, 2]}]\n",
    "               #'nu': [0.75, 0.8, 0.85, 0.9, 0.95, 0.97]}]\n",
    "\n",
    "reg1 = GridSearchCV(SVR(kernel='rbf', tol=0.01), parameters, cv=5, scoring='neg_mean_absolute_error')\n",
    "reg1.fit(X_train_scaled, y_train.values.flatten())\n",
    "y_pred1 = reg1.predict(X_train_scaled)\n",
    "\n",
    "print(\"Best CV score: {:.4f}\".format(reg1.best_score_))\n",
    "print(reg1.best_params_)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "tInD81CRMeM8"
   },
   "source": [
    "# Future Ideas \n",
    "- Recurrent Networks\n",
    "- Genetic Algorithms\n",
    "- Ordinary Differential Equation Networks (YOOOOOO)\n",
    "\n",
    "## Time to Rap"
   ]
  }
 ],
 "metadata": {
  "colab": {
   "name": "Earthquake_Challenge.ipynb",
   "provenance": [],
   "version": "0.3.2"
  },
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.7.1"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 1
}