multiclass_classification_with_nn.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "provenance": [],
      "collapsed_sections": [],
      "toc_visible": true
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    },
    "language_info": {
      "name": "python"
    },
    "accelerator": "GPU",
    "gpuClass": "standard"
  },
  "cells": [
    {
      "cell_type": "markdown",
      "source": [
        "# Multiclass Classification With Neural Networks\n",
        "## Predicting NBA player position from advanced player stats"
      ],
      "metadata": {
        "id": "9AxRhRriRcg3"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "The following notebook trains a multiclass classification model using a neural network.\n",
        "\n",
        "Specifically, we will be training a model to classify whether we can determine an NBA player's listed position based on advanced statistics describing their play."
      ],
      "metadata": {
        "id": "535CYk6jSL92"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "## Background Context"
      ],
      "metadata": {
        "id": "lbNjKE_3yhLd"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "In basketball, there are five positions: `PG`, `SG`, `SF`, `PF`, `C`. Although the positions are usually mapped very closely to height/weight, each position has a traditional set of expected skills. For example `PGs` are expected to have high assist rates, `SGs` are expected to score, `SFs` are usually well-rounded or good defenders, while `PFs` and `Cs` tend to accumulate more rebounds and blocks and are usually offensive- or defensive-minded respectively.\n",
        "\n",
        "For each player, we will scrape advanced stats that describe how the player plays. We will then determine if how a player plays alone can be used to predict player position."
      ],
      "metadata": {
        "id": "fanL2StYymuj"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "## Global Configurations"
      ],
      "metadata": {
        "id": "PaD1MZI89weP"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "import matplotlib.pyplot as plt\n",
        "import pandas as pd\n",
        "import seaborn as sns\n",
        "import torch\n",
        "\n",
        "from torch import nn"
      ],
      "metadata": {
        "id": "ygfwLSSE9l5O"
      },
      "execution_count": 1,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "ORDERED_LABELS = ['PG', 'SG', 'SF', 'PF', 'C']\n",
        "POS_COLORS = {'PG': 'g', 'SG': 'c', 'SF': 'b', 'PF': 'm', 'C': 'r'}"
      ],
      "metadata": {
        "id": "OLOHHGUt3k4I"
      },
      "execution_count": 2,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "device = 'cuda' if torch.cuda.is_available() else 'cpu'"
      ],
      "metadata": {
        "id": "r-SRYSk5SIqb"
      },
      "execution_count": 3,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "def reorder_handles_labels(ax):\n",
        "  handles, labels = ax.get_legend_handles_labels()\n",
        "  order = list(map(lambda label: labels.index(label), ORDERED_LABELS))\n",
        "  return [handles[idx] for idx in order], [labels[idx] for idx in order]"
      ],
      "metadata": {
        "id": "zbR5_pqE7GGo"
      },
      "execution_count": 4,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "source": [
        "## Data Exploration"
      ],
      "metadata": {
        "id": "g5NgbAgnTQ8z"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "Set seed for reproducability"
      ],
      "metadata": {
        "id": "7ygAWROZe4bw"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "SEED = 0\n",
        "if SEED is not None:\n",
        "  torch.manual_seed(SEED)\n",
        "  torch.cuda.manual_seed(SEED)\n",
        "  print(f'Random seed: {SEED}')"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "3ypYZWxUNy5r",
        "outputId": "3b05ccbc-313b-451c-b876-84c6a975e5c3"
      },
      "execution_count": 5,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Random seed: 0\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "### Scraping the data"
      ],
      "metadata": {
        "id": "OlQEG_ieTUXA"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "All data are scraped from https://www.basketball-reference.com/leagues/NBA_2022_advanced.html.\n",
        "* There are **600+** raw data points (NBA players)\n",
        "* There are **11** input features (advanced stats: `TS%`, `3PAr`, `Ftr`, `ORB%`, `DRB%`, `TRB%`, `AST%`, `STL%`, `BLK%`, `TOV%`, `USG%`)\n",
        "  * See [glossary](https://www.basketball-reference.com/about/glossary.html) for advanced stat descriptions\n",
        "* There are **5** classes (NBA positions: `PG`, `SG`, `SF`, `PF`, `C`)"
      ],
      "metadata": {
        "id": "_2J2wrWeT8xO"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "import requests\n",
        "from bs4 import BeautifulSoup"
      ],
      "metadata": {
        "id": "J7pSB1ntWAzI"
      },
      "execution_count": 6,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "ADVANCED_STATS_URL = \\\n",
        "  'https://www.basketball-reference.com/leagues/NBA_2022_advanced.html'\n",
        "DATA_STATS = ['player', 'pos', 'team_id', 'ts_pct', 'fg3a_per_fga_pct', \n",
        "         'fta_per_fga_pct', 'orb_pct', 'drb_pct', 'trb_pct', 'ast_pct',\n",
        "         'stl_pct', 'blk_pct', 'tov_pct', 'usg_pct']"
      ],
      "metadata": {
        "id": "Z41-7lee_lwm"
      },
      "execution_count": 7,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "html = requests.get(ADVANCED_STATS_URL).content\n",
        "soup = BeautifulSoup(html, 'html.parser')\n",
        "data = []\n",
        "for row in soup.find_all('tr', {'class': 'full_table'}):\n",
        "  sample = map(lambda stat: row.find('td', {'data-stat': stat}).text, DATA_STATS)\n",
        "  # Convert numerical stats in sample to float or None if empty\n",
        "  sample = map(lambda x: x[1] if x[0] < 3 else float(x[1]) if x[1] else None,\n",
        "               enumerate(sample))\n",
        "  data.append(list(sample))"
      ],
      "metadata": {
        "id": "xYyUaA7VWB-o"
      },
      "execution_count": 8,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "source": [
        "### Cleaning the data"
      ],
      "metadata": {
        "id": "_wWi3LdEv43E"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "COLUMNS = ['Player', 'Pos', 'Tm', 'TS%', '3PAr', 'Ftr', 'ORB%', 'DRB%', 'TRB%',\n",
        "           'AST%', 'STL%', 'BLK%', 'TOV%', 'USG%']\n",
        "NUMERICAL_DATA = COLUMNS[3:]"
      ],
      "metadata": {
        "id": "PtZx0fLfWj5J"
      },
      "execution_count": 9,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "df = pd.DataFrame.from_records(data, columns=COLUMNS)\n",
        "df = df[df['Tm'] != 'TOT'] # Filter out aggregated TOT row for multiteam players\n",
        "df = df.dropna() # Drop rows with missing stat values\n",
        "print('First 10 rows:')\n",
        "display(df.head(10))"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 381
        },
        "id": "4YLmAbqByUOO",
        "outputId": "94bfbaa9-eb08-42e2-cd86-73c79956ba64"
      },
      "execution_count": 10,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "First 10 rows:\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "                   Player Pos   Tm    TS%   3PAr    Ftr  ORB%  DRB%  TRB%  \\\n",
              "0        Precious Achiuwa   C  TOR  0.503  0.259  0.217   8.7  21.7  14.9   \n",
              "1            Steven Adams   C  MEM  0.560  0.003  0.518  17.9  22.0  19.9   \n",
              "2             Bam Adebayo   C  MIA  0.608  0.008  0.466   8.7  26.1  17.5   \n",
              "3            Santi Aldama  PF  MEM  0.452  0.364  0.242   9.4  16.1  12.6   \n",
              "4       LaMarcus Aldridge   C  BRK  0.604  0.100  0.223   7.8  18.7  13.4   \n",
              "6           Grayson Allen  SG  MIL  0.609  0.684  0.130   1.9  10.9   6.5   \n",
              "7           Jarrett Allen   C  CLE  0.698  0.018  0.428  12.0  24.5  18.4   \n",
              "8           Jose Alvarado  PG  NOP  0.520  0.374  0.180   3.2  10.2   6.6   \n",
              "10          Kyle Anderson  PF  MEM  0.506  0.232  0.224   4.8  21.6  13.0   \n",
              "11  Giannis Antetokounmpo  PF  MIL  0.633  0.194  0.615   6.6  30.4  18.7   \n",
              "\n",
              "    AST%  STL%  BLK%  TOV%  USG%  \n",
              "0    6.9   1.1   2.3  11.3  18.5  \n",
              "1   16.1   1.6   2.7  19.6  12.0  \n",
              "2   17.5   2.2   2.6  14.4  25.0  \n",
              "3    7.7   0.8   2.5   9.9  18.4  \n",
              "4    6.3   0.6   4.0   8.0  22.4  \n",
              "6    7.6   1.2   1.0   6.7  15.2  \n",
              "7    8.2   1.2   3.7  12.7  18.1  \n",
              "8   26.9   4.2   0.8  11.2  18.4  \n",
              "10  16.2   2.5   2.7  12.1  16.2  \n",
              "11  31.7   1.6   4.0  12.2  34.9  "
            ],
            "text/html": [
              "\n",
              "  <div id=\"df-ce22938a-0eec-422c-bac8-c63ac2bd8532\">\n",
              "    <div class=\"colab-df-container\">\n",
              "      <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>Player</th>\n",
              "      <th>Pos</th>\n",
              "      <th>Tm</th>\n",
              "      <th>TS%</th>\n",
              "      <th>3PAr</th>\n",
              "      <th>Ftr</th>\n",
              "      <th>ORB%</th>\n",
              "      <th>DRB%</th>\n",
              "      <th>TRB%</th>\n",
              "      <th>AST%</th>\n",
              "      <th>STL%</th>\n",
              "      <th>BLK%</th>\n",
              "      <th>TOV%</th>\n",
              "      <th>USG%</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>Precious Achiuwa</td>\n",
              "      <td>C</td>\n",
              "      <td>TOR</td>\n",
              "      <td>0.503</td>\n",
              "      <td>0.259</td>\n",
              "      <td>0.217</td>\n",
              "      <td>8.7</td>\n",
              "      <td>21.7</td>\n",
              "      <td>14.9</td>\n",
              "      <td>6.9</td>\n",
              "      <td>1.1</td>\n",
              "      <td>2.3</td>\n",
              "      <td>11.3</td>\n",
              "      <td>18.5</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>Steven Adams</td>\n",
              "      <td>C</td>\n",
              "      <td>MEM</td>\n",
              "      <td>0.560</td>\n",
              "      <td>0.003</td>\n",
              "      <td>0.518</td>\n",
              "      <td>17.9</td>\n",
              "      <td>22.0</td>\n",
              "      <td>19.9</td>\n",
              "      <td>16.1</td>\n",
              "      <td>1.6</td>\n",
              "      <td>2.7</td>\n",
              "      <td>19.6</td>\n",
              "      <td>12.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>Bam Adebayo</td>\n",
              "      <td>C</td>\n",
              "      <td>MIA</td>\n",
              "      <td>0.608</td>\n",
              "      <td>0.008</td>\n",
              "      <td>0.466</td>\n",
              "      <td>8.7</td>\n",
              "      <td>26.1</td>\n",
              "      <td>17.5</td>\n",
              "      <td>17.5</td>\n",
              "      <td>2.2</td>\n",
              "      <td>2.6</td>\n",
              "      <td>14.4</td>\n",
              "      <td>25.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>Santi Aldama</td>\n",
              "      <td>PF</td>\n",
              "      <td>MEM</td>\n",
              "      <td>0.452</td>\n",
              "      <td>0.364</td>\n",
              "      <td>0.242</td>\n",
              "      <td>9.4</td>\n",
              "      <td>16.1</td>\n",
              "      <td>12.6</td>\n",
              "      <td>7.7</td>\n",
              "      <td>0.8</td>\n",
              "      <td>2.5</td>\n",
              "      <td>9.9</td>\n",
              "      <td>18.4</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>LaMarcus Aldridge</td>\n",
              "      <td>C</td>\n",
              "      <td>BRK</td>\n",
              "      <td>0.604</td>\n",
              "      <td>0.100</td>\n",
              "      <td>0.223</td>\n",
              "      <td>7.8</td>\n",
              "      <td>18.7</td>\n",
              "      <td>13.4</td>\n",
              "      <td>6.3</td>\n",
              "      <td>0.6</td>\n",
              "      <td>4.0</td>\n",
              "      <td>8.0</td>\n",
              "      <td>22.4</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>6</th>\n",
              "      <td>Grayson Allen</td>\n",
              "      <td>SG</td>\n",
              "      <td>MIL</td>\n",
              "      <td>0.609</td>\n",
              "      <td>0.684</td>\n",
              "      <td>0.130</td>\n",
              "      <td>1.9</td>\n",
              "      <td>10.9</td>\n",
              "      <td>6.5</td>\n",
              "      <td>7.6</td>\n",
              "      <td>1.2</td>\n",
              "      <td>1.0</td>\n",
              "      <td>6.7</td>\n",
              "      <td>15.2</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>7</th>\n",
              "      <td>Jarrett Allen</td>\n",
              "      <td>C</td>\n",
              "      <td>CLE</td>\n",
              "      <td>0.698</td>\n",
              "      <td>0.018</td>\n",
              "      <td>0.428</td>\n",
              "      <td>12.0</td>\n",
              "      <td>24.5</td>\n",
              "      <td>18.4</td>\n",
              "      <td>8.2</td>\n",
              "      <td>1.2</td>\n",
              "      <td>3.7</td>\n",
              "      <td>12.7</td>\n",
              "      <td>18.1</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>8</th>\n",
              "      <td>Jose Alvarado</td>\n",
              "      <td>PG</td>\n",
              "      <td>NOP</td>\n",
              "      <td>0.520</td>\n",
              "      <td>0.374</td>\n",
              "      <td>0.180</td>\n",
              "      <td>3.2</td>\n",
              "      <td>10.2</td>\n",
              "      <td>6.6</td>\n",
              "      <td>26.9</td>\n",
              "      <td>4.2</td>\n",
              "      <td>0.8</td>\n",
              "      <td>11.2</td>\n",
              "      <td>18.4</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>10</th>\n",
              "      <td>Kyle Anderson</td>\n",
              "      <td>PF</td>\n",
              "      <td>MEM</td>\n",
              "      <td>0.506</td>\n",
              "      <td>0.232</td>\n",
              "      <td>0.224</td>\n",
              "      <td>4.8</td>\n",
              "      <td>21.6</td>\n",
              "      <td>13.0</td>\n",
              "      <td>16.2</td>\n",
              "      <td>2.5</td>\n",
              "      <td>2.7</td>\n",
              "      <td>12.1</td>\n",
              "      <td>16.2</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>11</th>\n",
              "      <td>Giannis Antetokounmpo</td>\n",
              "      <td>PF</td>\n",
              "      <td>MIL</td>\n",
              "      <td>0.633</td>\n",
              "      <td>0.194</td>\n",
              "      <td>0.615</td>\n",
              "      <td>6.6</td>\n",
              "      <td>30.4</td>\n",
              "      <td>18.7</td>\n",
              "      <td>31.7</td>\n",
              "      <td>1.6</td>\n",
              "      <td>4.0</td>\n",
              "      <td>12.2</td>\n",
              "      <td>34.9</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>\n",
              "      <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-ce22938a-0eec-422c-bac8-c63ac2bd8532')\"\n",
              "              title=\"Convert this dataframe to an interactive table.\"\n",
              "              style=\"display:none;\">\n",
              "        \n",
              "  <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\"viewBox=\"0 0 24 24\"\n",
              "       width=\"24px\">\n",
              "    <path d=\"M0 0h24v24H0V0z\" fill=\"none\"/>\n",
              "    <path d=\"M18.56 5.44l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94zm-11 1L8.5 8.5l.94-2.06 2.06-.94-2.06-.94L8.5 2.5l-.94 2.06-2.06.94zm10 10l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94z\"/><path d=\"M17.41 7.96l-1.37-1.37c-.4-.4-.92-.59-1.43-.59-.52 0-1.04.2-1.43.59L10.3 9.45l-7.72 7.72c-.78.78-.78 2.05 0 2.83L4 21.41c.39.39.9.59 1.41.59.51 0 1.02-.2 1.41-.59l7.78-7.78 2.81-2.81c.8-.78.8-2.07 0-2.86zM5.41 20L4 18.59l7.72-7.72 1.47 1.35L5.41 20z\"/>\n",
              "  </svg>\n",
              "      </button>\n",
              "      \n",
              "  <style>\n",
              "    .colab-df-container {\n",
              "      display:flex;\n",
              "      flex-wrap:wrap;\n",
              "      gap: 12px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert {\n",
              "      background-color: #E8F0FE;\n",
              "      border: none;\n",
              "      border-radius: 50%;\n",
              "      cursor: pointer;\n",
              "      display: none;\n",
              "      fill: #1967D2;\n",
              "      height: 32px;\n",
              "      padding: 0 0 0 0;\n",
              "      width: 32px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert:hover {\n",
              "      background-color: #E2EBFA;\n",
              "      box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
              "      fill: #174EA6;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert {\n",
              "      background-color: #3B4455;\n",
              "      fill: #D2E3FC;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert:hover {\n",
              "      background-color: #434B5C;\n",
              "      box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n",
              "      filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n",
              "      fill: #FFFFFF;\n",
              "    }\n",
              "  </style>\n",
              "\n",
              "      <script>\n",
              "        const buttonEl =\n",
              "          document.querySelector('#df-ce22938a-0eec-422c-bac8-c63ac2bd8532 button.colab-df-convert');\n",
              "        buttonEl.style.display =\n",
              "          google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
              "\n",
              "        async function convertToInteractive(key) {\n",
              "          const element = document.querySelector('#df-ce22938a-0eec-422c-bac8-c63ac2bd8532');\n",
              "          const dataTable =\n",
              "            await google.colab.kernel.invokeFunction('convertToInteractive',\n",
              "                                                     [key], {});\n",
              "          if (!dataTable) return;\n",
              "\n",
              "          const docLinkHtml = 'Like what you see? Visit the ' +\n",
              "            '<a target=\"_blank\" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'\n",
              "            + ' to learn more about interactive tables.';\n",
              "          element.innerHTML = '';\n",
              "          dataTable['output_type'] = 'display_data';\n",
              "          await google.colab.output.renderOutput(dataTable, element);\n",
              "          const docLink = document.createElement('div');\n",
              "          docLink.innerHTML = docLinkHtml;\n",
              "          element.appendChild(docLink);\n",
              "        }\n",
              "      </script>\n",
              "    </div>\n",
              "  </div>\n",
              "  "
            ]
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "### Visualizing the clustered data"
      ],
      "metadata": {
        "id": "n78z8bO11Ft8"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "import warnings\n",
        "warnings.filterwarnings(action='ignore')\n",
        "\n",
        "from sklearn.manifold import TSNE"
      ],
      "metadata": {
        "id": "_DAVvXpV1ItN"
      },
      "execution_count": 11,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "N = df[COLUMNS[3:]].values # Get numerical data values\n",
        "T = TSNE(n_components=2, random_state=SEED).fit_transform(N)\n",
        "df['T1'] = list(map(lambda t: t[0], T))\n",
        "df['T2'] = list(map(lambda t: t[1], T))"
      ],
      "metadata": {
        "id": "9xsjNN2IvZjZ"
      },
      "execution_count": 12,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "sns.scatterplot(x='T1', y='T2', data=df, hue='Pos', palette=POS_COLORS)\n",
        "handles, labels = reorder_handles_labels(plt.gca())\n",
        "plt.legend(handles, labels, ncol=2)\n",
        "plt.title('All Observations (in 2-dimensions)')\n",
        "plt.show()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 295
        },
        "id": "dF3xo4vS2eVW",
        "outputId": "f0008c66-a9b4-43cf-c4a1-59ef57841530"
      },
      "execution_count": 13,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "Even when visualizing the data in lower-dimensional 2D space, we can see evidence of light clustering."
      ],
      "metadata": {
        "id": "0g9X-NWz7vMs"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "### Creating train/test data split"
      ],
      "metadata": {
        "id": "BdvK1nV-9gW1"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "from sklearn.model_selection import train_test_split"
      ],
      "metadata": {
        "id": "iY-SjMyM9kpi"
      },
      "execution_count": 14,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "train_df, test_df = train_test_split(df, random_state=SEED)\n",
        "x_train, x_test = train_df[NUMERICAL_DATA].values, test_df[NUMERICAL_DATA].values\n",
        "y_train, y_test = train_df['Pos'].values, test_df['Pos'].values"
      ],
      "metadata": {
        "id": "UceabRPL9taM"
      },
      "execution_count": 15,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "source": [
        "## Model Fitting"
      ],
      "metadata": {
        "id": "2wzcMp2z_V5z"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "### Define neural network model"
      ],
      "metadata": {
        "id": "q23i5-mGgDS4"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "INPUT_SIZE = 11\n",
        "HIDDEN_1_SIZE = 16\n",
        "HIDDEN_2_SIZE = 16\n",
        "HIDDEN_3_SIZE = 8\n",
        "OUTPUT_SIZE = 5"
      ],
      "metadata": {
        "id": "cDLLZh2u_Q3U"
      },
      "execution_count": 16,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "model = nn.Sequential(\n",
        "  nn.Linear(in_features=INPUT_SIZE, out_features=HIDDEN_1_SIZE),\n",
        "  nn.ReLU(),\n",
        "  nn.Linear(in_features=HIDDEN_1_SIZE, out_features=HIDDEN_2_SIZE),\n",
        "  nn.ReLU(),\n",
        "  nn.Linear(in_features=HIDDEN_2_SIZE, out_features=HIDDEN_3_SIZE),\n",
        "  nn.ReLU(),\n",
        "  nn.Linear(in_features=HIDDEN_3_SIZE, out_features=OUTPUT_SIZE)\n",
        ").to(device)"
      ],
      "metadata": {
        "id": "c0stK_FfHYbF"
      },
      "execution_count": 17,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "source": [
        "### Transforming the data for the model"
      ],
      "metadata": {
        "id": "g-Wz3Fqsd0BD"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "from sklearn import preprocessing"
      ],
      "metadata": {
        "id": "HtXql3RSf-ZJ"
      },
      "execution_count": 18,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "le = preprocessing.LabelEncoder()\n",
        "le.fit(ORDERED_LABELS)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "zpZe60GajGUO",
        "outputId": "ca3e43a2-06a6-4a00-d1e8-f618d9b85f88"
      },
      "execution_count": 19,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "LabelEncoder()"
            ]
          },
          "metadata": {},
          "execution_count": 19
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "X_train = torch.tensor(x_train).type(torch.float).to(device)\n",
        "Y_train = torch.tensor(le.transform(y_train)).type(torch.float).to(device)\n",
        "X_test = torch.tensor(x_test).type(torch.float).to(device)\n",
        "Y_test = torch.tensor(le.transform(y_test)).type(torch.float).to(device)"
      ],
      "metadata": {
        "id": "e46Ny6j1d2GK"
      },
      "execution_count": 20,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "source": [
        "### Defining the `fit()` function"
      ],
      "metadata": {
        "id": "3aicl6GHUJtu"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "from sklearn.metrics import accuracy_score\n",
        "from torch.optim import SGD"
      ],
      "metadata": {
        "id": "iyk9B82fG1wD"
      },
      "execution_count": 21,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "def fit(X_train, Y_train, X_test, Y_test,\n",
        "        epochs = 100000,\n",
        "        loss_fn = nn.CrossEntropyLoss(),\n",
        "        accuracy_fn = accuracy_score,\n",
        "        optimizer = SGD(params=model.parameters(), lr=0.001),\n",
        "        verbose = 1000,\n",
        "        early_stopping = 1000,\n",
        "        stop_after_k = 10):\n",
        "  best = {\n",
        "    'state_dict': None,\n",
        "    'epoch': None,\n",
        "    'loss_train': None,\n",
        "    'accuracy_train': None,\n",
        "    'loss_test': None,\n",
        "    'accuracy_test': None\n",
        "  }\n",
        "  early_stop_k = 0\n",
        "  losses_train = []\n",
        "  losses_test = []\n",
        "  accuracies_train = []\n",
        "  accuracies_test = []\n",
        "  for epoch in range(epochs):\n",
        "    # Training\n",
        "    model.train()\n",
        "    loss_train, accuracy_train = forward(X_train, Y_train, loss_fn, accuracy_fn)\n",
        "    losses_train.append(loss_train.item())\n",
        "    accuracies_train.append(accuracy_train)\n",
        "    optimizer.zero_grad()\n",
        "    loss_train.backward()\n",
        "    optimizer.step()\n",
        "    # Testing\n",
        "    model.eval()\n",
        "    with torch.inference_mode():\n",
        "      loss_test, accuracy_test = forward(X_test, Y_test, loss_fn, accuracy_fn)\n",
        "      losses_test.append(loss_test.item())\n",
        "      accuracies_test.append(accuracy_test)\n",
        "    # Verbose Print\n",
        "    if verbose and epoch % verbose == 0:\n",
        "      print_results(epoch, loss_train, accuracy_train, loss_test, accuracy_test)\n",
        "    # Early Stopping\n",
        "    if early_stopping and epoch % early_stopping == 0:\n",
        "      if not best['loss_test'] or loss_test < best['loss_test']:\n",
        "        early_stop_k = 0\n",
        "        best['state_dict'] = model.state_dict()\n",
        "        best['epoch'] = epoch\n",
        "        best['loss_train'] = loss_train\n",
        "        best['accuracy_train'] = accuracy_train\n",
        "        best['loss_test'] = loss_test\n",
        "        best['accuracy_test'] = accuracy_test\n",
        "      else:\n",
        "        early_stop_k += 1\n",
        "        if early_stop_k >= stop_after_k:\n",
        "          print(f'\\nStopping Early\\nBest', end=' ')\n",
        "          print_results(best['epoch'], best['loss_train'],\n",
        "                        best['accuracy_train'], best['loss_test'],\n",
        "                        best['accuracy_test'])\n",
        "          break\n",
        "  model.load_state_dict(best['state_dict'])\n",
        "  return losses_train, losses_test, accuracies_train, accuracies_test\n",
        "\n",
        "def forward(X, Y, loss_fn, accuracy_fn):\n",
        "  logits = model(X)\n",
        "  pred = torch.softmax(logits, dim=1).argmax(dim=1)\n",
        "  loss = loss_fn(logits.cpu(), Y.type(torch.LongTensor))\n",
        "  accuracy = accuracy_fn(Y.cpu(), pred.cpu())\n",
        "  return loss, accuracy\n",
        "\n",
        "def print_results(epoch, loss_train, accuracy_train, loss_test, accuracy_test):\n",
        "  print(f'Epoch: {epoch}', end=' | ')\n",
        "  print(f'Train Loss: {loss_train:.4f}', end=', ')\n",
        "  print(f'Train Acc: {accuracy_train:.1%}', end=' | ')\n",
        "  print(f'Test Loss: {loss_test:.4f}', end=', ')\n",
        "  print(f'Test Acc: {accuracy_test:.1%}')"
      ],
      "metadata": {
        "id": "UXcPBDaYUNM4"
      },
      "execution_count": 22,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "source": [
        "### Fitting the model"
      ],
      "metadata": {
        "id": "5jmwGgakcAtB"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "%%time\n",
        "losses_train, losses_test, accuracies_train, accuracies_test = fit(\n",
        "    X_train, Y_train, X_test, Y_test)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "5VQ67rY4cDHX",
        "outputId": "6074d96c-2dee-4075-c55a-80db5ca40a4c"
      },
      "execution_count": 23,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Epoch: 0 | Train Loss: 1.7928, Train Acc: 18.2% | Test Loss: 1.7471, Test Acc: 22.4%\n",
            "Epoch: 1000 | Train Loss: 1.4163, Train Acc: 37.7% | Test Loss: 1.4507, Test Acc: 40.8%\n",
            "Epoch: 2000 | Train Loss: 1.3188, Train Acc: 39.8% | Test Loss: 1.3797, Test Acc: 42.4%\n",
            "Epoch: 3000 | Train Loss: 1.2686, Train Acc: 41.4% | Test Loss: 1.3402, Test Acc: 40.8%\n",
            "Epoch: 4000 | Train Loss: 1.2248, Train Acc: 42.0% | Test Loss: 1.3022, Test Acc: 41.6%\n",
            "Epoch: 5000 | Train Loss: 1.1779, Train Acc: 46.8% | Test Loss: 1.2595, Test Acc: 43.2%\n",
            "Epoch: 6000 | Train Loss: 1.1374, Train Acc: 48.9% | Test Loss: 1.2371, Test Acc: 45.6%\n",
            "Epoch: 7000 | Train Loss: 1.1130, Train Acc: 50.8% | Test Loss: 1.2215, Test Acc: 48.0%\n",
            "Epoch: 8000 | Train Loss: 1.0888, Train Acc: 52.7% | Test Loss: 1.2056, Test Acc: 48.0%\n",
            "Epoch: 9000 | Train Loss: 1.0715, Train Acc: 51.6% | Test Loss: 1.1939, Test Acc: 48.0%\n",
            "Epoch: 10000 | Train Loss: 1.0591, Train Acc: 50.8% | Test Loss: 1.1855, Test Acc: 48.8%\n",
            "Epoch: 11000 | Train Loss: 1.0486, Train Acc: 51.6% | Test Loss: 1.1808, Test Acc: 47.2%\n",
            "Epoch: 12000 | Train Loss: 1.0392, Train Acc: 51.3% | Test Loss: 1.1795, Test Acc: 47.2%\n",
            "Epoch: 13000 | Train Loss: 1.0316, Train Acc: 51.6% | Test Loss: 1.1789, Test Acc: 48.8%\n",
            "Epoch: 14000 | Train Loss: 1.0246, Train Acc: 52.1% | Test Loss: 1.1775, Test Acc: 48.8%\n",
            "Epoch: 15000 | Train Loss: 1.0181, Train Acc: 52.7% | Test Loss: 1.1765, Test Acc: 48.8%\n",
            "Epoch: 16000 | Train Loss: 1.0120, Train Acc: 52.9% | Test Loss: 1.1748, Test Acc: 49.6%\n",
            "Epoch: 17000 | Train Loss: 1.0062, Train Acc: 53.2% | Test Loss: 1.1729, Test Acc: 50.4%\n",
            "Epoch: 18000 | Train Loss: 1.0005, Train Acc: 52.9% | Test Loss: 1.1716, Test Acc: 50.4%\n",
            "Epoch: 19000 | Train Loss: 0.9945, Train Acc: 53.5% | Test Loss: 1.1700, Test Acc: 49.6%\n",
            "Epoch: 20000 | Train Loss: 0.9882, Train Acc: 52.9% | Test Loss: 1.1689, Test Acc: 50.4%\n",
            "Epoch: 21000 | Train Loss: 0.9824, Train Acc: 53.2% | Test Loss: 1.1687, Test Acc: 51.2%\n",
            "Epoch: 22000 | Train Loss: 0.9767, Train Acc: 54.0% | Test Loss: 1.1684, Test Acc: 51.2%\n",
            "Epoch: 23000 | Train Loss: 0.9710, Train Acc: 54.0% | Test Loss: 1.1682, Test Acc: 51.2%\n",
            "Epoch: 24000 | Train Loss: 0.9653, Train Acc: 54.5% | Test Loss: 1.1676, Test Acc: 51.2%\n",
            "Epoch: 25000 | Train Loss: 0.9595, Train Acc: 55.9% | Test Loss: 1.1673, Test Acc: 51.2%\n",
            "Epoch: 26000 | Train Loss: 0.9535, Train Acc: 56.7% | Test Loss: 1.1662, Test Acc: 51.2%\n",
            "Epoch: 27000 | Train Loss: 0.9475, Train Acc: 56.7% | Test Loss: 1.1651, Test Acc: 50.4%\n",
            "Epoch: 28000 | Train Loss: 0.9413, Train Acc: 56.7% | Test Loss: 1.1638, Test Acc: 51.2%\n",
            "Epoch: 29000 | Train Loss: 0.9351, Train Acc: 57.2% | Test Loss: 1.1616, Test Acc: 51.2%\n",
            "Epoch: 30000 | Train Loss: 0.9290, Train Acc: 57.5% | Test Loss: 1.1590, Test Acc: 52.0%\n",
            "Epoch: 31000 | Train Loss: 0.9232, Train Acc: 57.5% | Test Loss: 1.1580, Test Acc: 52.0%\n",
            "Epoch: 32000 | Train Loss: 0.9177, Train Acc: 59.1% | Test Loss: 1.1576, Test Acc: 51.2%\n",
            "Epoch: 33000 | Train Loss: 0.9125, Train Acc: 59.1% | Test Loss: 1.1573, Test Acc: 50.4%\n",
            "Epoch: 34000 | Train Loss: 0.9075, Train Acc: 59.6% | Test Loss: 1.1575, Test Acc: 52.0%\n",
            "Epoch: 35000 | Train Loss: 0.9025, Train Acc: 60.4% | Test Loss: 1.1576, Test Acc: 52.0%\n",
            "Epoch: 36000 | Train Loss: 0.8978, Train Acc: 61.5% | Test Loss: 1.1570, Test Acc: 52.0%\n",
            "Epoch: 37000 | Train Loss: 0.8934, Train Acc: 61.8% | Test Loss: 1.1566, Test Acc: 52.0%\n",
            "Epoch: 38000 | Train Loss: 0.8892, Train Acc: 62.0% | Test Loss: 1.1576, Test Acc: 52.0%\n",
            "Epoch: 39000 | Train Loss: 0.8848, Train Acc: 62.3% | Test Loss: 1.1592, Test Acc: 52.0%\n",
            "Epoch: 40000 | Train Loss: 0.8808, Train Acc: 62.6% | Test Loss: 1.1600, Test Acc: 52.0%\n",
            "Epoch: 41000 | Train Loss: 0.8772, Train Acc: 62.6% | Test Loss: 1.1597, Test Acc: 51.2%\n",
            "Epoch: 42000 | Train Loss: 0.8737, Train Acc: 62.6% | Test Loss: 1.1592, Test Acc: 51.2%\n",
            "Epoch: 43000 | Train Loss: 0.8704, Train Acc: 62.8% | Test Loss: 1.1589, Test Acc: 51.2%\n",
            "Epoch: 44000 | Train Loss: 0.8673, Train Acc: 63.1% | Test Loss: 1.1587, Test Acc: 51.2%\n",
            "Epoch: 45000 | Train Loss: 0.8642, Train Acc: 63.4% | Test Loss: 1.1586, Test Acc: 50.4%\n",
            "Epoch: 46000 | Train Loss: 0.8612, Train Acc: 63.4% | Test Loss: 1.1578, Test Acc: 50.4%\n",
            "Epoch: 47000 | Train Loss: 0.8584, Train Acc: 64.4% | Test Loss: 1.1571, Test Acc: 50.4%\n",
            "\n",
            "Stopping Early\n",
            "Best Epoch: 37000 | Train Loss: 0.8934, Train Acc: 61.8% | Test Loss: 1.1566, Test Acc: 52.0%\n",
            "CPU times: user 1min 57s, sys: 3 s, total: 2min\n",
            "Wall time: 2min 6s\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "## Visualizing the Results"
      ],
      "metadata": {
        "id": "fh8e9GrMwGPh"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "### Plotting `loss` and `accuracy`"
      ],
      "metadata": {
        "id": "-cRPMKOPPaAV"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "x = list(range(len(losses_train)))\n",
        "fig, ax = plt.subplots(1, 2, figsize=(12.8, 4.8))\n",
        "# Subplot (1, 1)\n",
        "sns.lineplot(ax=ax[0], x=x, y=losses_train)\n",
        "sns.lineplot(ax=ax[0], x=x, y=losses_test)\n",
        "ax[0].legend(['training loss', 'test loss'])\n",
        "ax[0].set_xlabel('epoch')\n",
        "ax[0].set_ylabel('loss')\n",
        "ax[0].set_title('Learning Curves')\n",
        "# Subplot (1, 2)\n",
        "sns.lineplot(ax=ax[1], x=x, y=accuracies_train)\n",
        "sns.lineplot(ax=ax[1], x=x, y=accuracies_test)\n",
        "ax[1].legend(['training accuracy', 'test accuracy'])\n",
        "ax[1].set_xlabel('epoch')\n",
        "ax[1].set_ylabel('accuracy')\n",
        "ax[1].set_title('Classifier Accuracies')\n",
        "plt.show()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 339
        },
        "id": "PGA28ID4P-Go",
        "outputId": "d9d685f3-3788-4d4e-f18f-908425f1e52a"
      },
      "execution_count": 24,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 921.6x345.6 with 2 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "### Comparing `test` vs `pred` observations"
      ],
      "metadata": {
        "id": "dSdowfKbwiI8"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "Make predictions with the model"
      ],
      "metadata": {
        "id": "b2CQMjm6XOQc"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "model.eval()\n",
        "with torch.inference_mode():\n",
        "  logits_test = model(X_test)\n",
        "  y_pred = torch.softmax(logits_test, dim=1).argmax(dim=1).cpu().numpy()\n",
        "Y_pred = le.inverse_transform(y_pred)\n",
        "Y_test = test_df['Pos']"
      ],
      "metadata": {
        "id": "t6Q5-QjgwIpt"
      },
      "execution_count": 25,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "source": [
        "Plot `test` vs `pred`"
      ],
      "metadata": {
        "id": "pmJIfZAyXRyB"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "test_df['Pred'] = Y_pred"
      ],
      "metadata": {
        "id": "enrcFgUHgJSx"
      },
      "execution_count": 26,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "fig, ax = plt.subplots(1, 2, figsize=(12.8, 4.8))\n",
        "# Subplot (1, 1)\n",
        "sns.scatterplot(ax=ax[0], x='T1', y='T2', data=test_df, hue='Pos',\n",
        "                palette=POS_COLORS)\n",
        "handles, labels = reorder_handles_labels(ax[0])\n",
        "ax[0].legend(handles, labels, ncol=2)\n",
        "ax[0].set_title('Test Observations (in 2-dimensions)')\n",
        "# Subplot (1, 2)\n",
        "sns.scatterplot(ax=ax[1], x='T1', y='T2', data=test_df, hue='Pred',\n",
        "                palette=POS_COLORS)\n",
        "handles, labels = reorder_handles_labels(ax[1])\n",
        "ax[1].legend(handles, labels, ncol=2)\n",
        "ax[1].set_title('Predicted Observations (in 2-dimensions)')\n",
        "plt.show()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 339
        },
        "id": "B2OIhne3ws_o",
        "outputId": "609771eb-e06d-4112-8ba5-238b0c941b0a"
      },
      "execution_count": 27,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 921.6x345.6 with 2 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "Visually we can see that our predictions (right) captures the overall patterns in the test data (left)."
      ],
      "metadata": {
        "id": "ARcC_UbLCMOq"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "### Confusion matrix"
      ],
      "metadata": {
        "id": "7guTAwijyMWS"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "from sklearn.metrics import ConfusionMatrixDisplay, confusion_matrix"
      ],
      "metadata": {
        "id": "R40B6AhpyPQa"
      },
      "execution_count": 28,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "C = confusion_matrix(Y_test, Y_pred, normalize='true')\n",
        "ConfusionMatrixDisplay(C, display_labels=ORDERED_LABELS).plot()\n",
        "plt.title('Confusion Matrix - p(pred|true)')\n",
        "plt.show()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 295
        },
        "id": "6PpGi2mkIGMx",
        "outputId": "899cfb72-f0f0-451c-96c8-6f2453aa6565"
      },
      "execution_count": 29,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 2 Axes>"
            ],
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAT0AAAEWCAYAAAAU8EBoAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4yLjIsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy+WH4yJAAAgAElEQVR4nOzdeXhU1fnA8e87kz0QIAsBwr6LyCIogoq4gnXBVq2obdXW4gLWXWtp1aK1am2tVdyX+lMR644KghsCIggoi2GXJfu+EAhkmXl/f9ybZBKyTGAmM2HO53nuA3PvmXPfuZm8Oeeee88VVcUwDCNUOAIdgGEYRlsySc8wjJBikp5hGCHFJD3DMEKKSXqGYYQUk/QMwwgpJun5iIhEi8hHIlIqIm8fQT1XishiX8YWCCKyUESuCtC+h4nIGhGRNtrfbhE5y/5/XxHZ3Rb79dj/uyJyblvusz0LuaQnIlfYvxD7RCTb/uU8xQdVXwIkAwmqeunhVqKqb6jqOT6Ipx4RmSQiKiLvN1g/0l6/xMt67heR11sqp6rnquqrhxnukXoAeEyD4CJUb4/XEXoEeNDP+zhqhFTSE5HbgH8DD2ElqN7A08BUH1TfB9imqtU+qMtf8oHxIpLgse4qYJuvdiCWgH2vRKQ7cDrwwWG+P8y3EbW4vyM+Xqr6HRAnImN9FNbRTVVDYgE6AfuAS5spE4mVFLPs5d9ApL1tEpAB3A7kAdnANfa2vwKVQJW9j98B9wOve9TdF1AgzH59NbATKAN2AVd6rF/u8b4JwGqg1P53gse2JVitmm/sehYDiU18tpr4nwVm2OucQCZwL7DEo+wTQDqwF1gLnGqvn9Lgc673iONvdhwHgIH2umvt7c8A73rU/wjwBSCH8XOsOY7T7Z9RNnCHx/bfAJ83eM9u4B5gE1AMvAJENTgudwM5wGtYjYE/Aj8BhcD/gHiP+n4N7LG3zbLrP8sjvt2Hcbxq67DL3E/9789JwAqgBFgPTGrwGV8A7gv071l7WEKppTceiALeb6bMLKwv1yhgJHAi8GeP7d2wkmcKVmKbIyJdVPU+rNbjW6raQVVfai4QEYkF/gOcq6odsRLbukbKxQOf2GUTgH8BnzRoqV0BXAN0BSKAO5rbN/B/WIkBYDLwI1by8LQa6xjEA3OBt0UkSlU/bfA5R3q859dYiagjVkLwdDtwnIhcLSKnYh27q9T+bT1MpwODgHOAu2vOqQHHAVsbKX8l1ucdAAzm0J9rPFZrfTpwE3ARcBrQAytRzgHrfCFWEv+1vS0B6NlYgEdwvOoRkRSs78GDdpx3AO+KSJJHsc1Y31mjBaGU9BKAAm2++3klMFtV81Q1H6sF92uP7VX29ipVXYD113vIYcbjBoaLSLSqZqtqaiNlzgO2q+prqlqtqm8CW4ALPMq8oqrbVPUAVotkVHM7VdUVQLyIDMFKfv/XSJnXVbXQ3uc/sVrALX3O/6pqqv2eqgb1lWMdx38BrwM3qWpGC/W15K+qul9VN2K13C6313fGavU29JSqpqtqEVYr63KPbW6sVlKFfRyvB2apaoaqVmC1ui6xu76XAB+r6lJ721/s97dWk8erEb8CFqjqAlV1q+pnwBrgZx5lyrA+u9GCUEp6hUBiC+dselD/r+4ee11tHQ2SZjnQobWBqOp+4DKsX65sEflERIZ6EU9NTCker3MOI57XgJlYraVDWr4icoeIbLZHokuwWreJLdSZ3txGVV2F1Z0XrOTcKBFJtQeZ9tmtQm/25/lzKsZqPXlbHiBfVQ96vO4DvC8iJfbn3wy4sM4D9/Csy/5ZFjYTpzfxt6QPcGlNPHZMpwDdPcp0xOr6Gi0IpaT3LVCB1W1pShbWF6xGbw7t+nlrPxDj8bqb50ZVXaSqZ2N9cbdgnZNpKZ6amDIPM6YarwE3YrUeyj032InmLuCXQBdV7Yx1PrHm8o+muqTNdlVFZAZWizHLrr/xSlSPtbuCHVR1WTNV9vL4v+fPaQNW99Xb8o3Fno516qGzxxKlqplY5xBr6xKRGKxeRJMfycv1zX1f0oHXGsQTq6oPe5Q5Butcn9GCkEl6qlqKdcJ+johcJCIxIhIuIueKyKN2sTeBP4tIkogk2uUP93KDdcBEEektIp2wTqQDICLJIjLVPrdXgdVNbqyLtAAYbF9mEyYilwHDgI8PMyYAVHUX1vmqWY1s7ghUY430honIvUCcx/ZcoG9rRhxFZDDW+ahfYXVz7xKRZrvhXviL/TM8Fuuc5lv2+s+A40UkqkH5GSLS0z5POsujfGOeBf4mIn3s+JNEpGaE/x3gfBE5RUQigNk0/3vk7fFaB0yzv5NjsbrRNV4HLhCRySLiFJEo+xIkz3OJpwELW9iHQQglPQD7/NRtWCex87H+gs6k7vKGB7HOlWwANgLfc5jXP9nnXd6y61pL/UTlsOPIAoqwvrA3NFJHIXA+1kBAIVYL6XxVLTicmBrUvVxVG2vFLgI+xbqMZQ9wkPpdsZoLrwtF5PuW9mOfTngdeERV16vqduBPwGsiEnkEH+FrYAfWKPBjqroYQFVzgS859DKkuVij2zuxRmWb+7k+AcwHFotIGbASGGfXnwrMsOvLxupON3d+0tvj9ResQZZirHPJc2s2qGq6/Xn+RN339k7s318ROQHYp9alK0YL5MgG0AyjbYlIX6xLfMKbGpSyR1hfBU5UVbXvkLhWVT9vo/iWqGpff+/LY5/vAi/Zg2tGC9r0QkzDaAuqugk4IdBxtBVVvTjQMbQnIdW9NYw2UIJ1UbsRpEz31jCMkGJaeoZhhJR2c04vMd6pvXu1j3C370lquVAQkdLylgsFEQlzBjoEr2m1K9AhtEoZxQWqethf4Mmnx2phkXefee2GikWqOuVw93W42kcWAXr3CuObT7u3XDAITJ5+Y6BDaJXIT1YHOoRWcXZp7lrg4OIqOJybNQLnc32n2fuAW1JY5OK7Rb29Kuvsvr2lu3z8ot0kPcMwgp8C7sO6FbntmKRnGIbPKEqVBneX3iQ9wzB8yrT0DMMIGYriCvLL4EzSMwzDp9zNT7gTcCbpGYbhMwq4TNIzDCOUmJaeYRghQ4Eqc07PMIxQoajp3hqGEUIUXMGd80zSMwzDd6w7MoKbSXqGYfiQ4Kp9hlRwMknPMAyfsQYyTNIzDCNEWNfpmaRnGEYIcZuWnmEYocK09AJo7VdxvHhvb1xu4ZzL87lkZk697S/e14uNK6xnWFcccFBaGMabm39gwzcdeen+ukkQM36K4s6nf+KkKSV+jffEY9OZOW0lTofyybIhzP10ZL3tIwZlM/OylQzoWcTs58/g6+/71W677uJVnDQiHYcoazal8OS88RDAL97YSXu5/oEsnA5l4Zvx/O+p5DaPYczJhVx393YcDmXRe915++W+9baHhbu542+bGDisjLLScP5+57HkZUXXbk/qdpBnP1jFG8/0471Xre/DKwtXcKDcicsluF3CzZe3/QPXguHYNkcRXEH+FAq/JT0RcWE9MDsM2AxcparlIpIMPA6chPVg40rgUVV931f7drnguVl9mP3mNhK6V3L7z4Zx4jkl9B58sLbMtX+te371xy935acfYwAYcXIZT3yWCkBZsZPrThnB6NP2+iq0RjnEzc1XrOCOx88lvziWZ2d9yDfre7Mnu0ttmbyiDjz8ykQum7yx3nuPHZDL8IG5/O7+XwDw5N0fM2pwNuu29fBrzE1xOJQZD2Vyz7T+FGSH8+SC7axc1Im07VFtGsONf9rKrOmjKciN5N9vrmHlkiTSd8bWlpn8iyz27Q3j2vPHM3FKLr+95Scevmt47fbf37mdNcvjD6n7j78bzd6SiDb5HA0Fw7H1RrB3b/2Zkg+o6ihVHY6V2K4XEQE+AJaqan9VHQNMA3r6csfbf4ile98KuvWpIDxCOXVqEasWdWmy/NIPEph4UdEh67/5JJ4xp5cSGe3fK4+G9ssnMz+O7II4ql1Ovlzdn5NH1Z+1O6ewIzszE9AGXyhViAh3ERbmJjzcTZjTTdHeaAJlyOhysnZHkJMWSXWVgyUfdmb85NI2jWHw8L1kpcWQkxlNdbWDpZ92Zfzp+fXKnDSpgM/nW48fWP5ZEiPHFYN9J8H40/PJyYwm7afYhlUHVDAc25YoQqU6vVq8ISJTRGSriOwQkT82sv1xEVlnL9tEpMUuWVu1Q5cBA4EzgEpVfbZmg6ruUdUnfbmzwpwIEntU1r5O7F5JYU54o2XzMiLITY9gxMmHtuaWfRjPxKn+f8ZBUudy8ovqfsHyi2NJ6uzdw3o27Uxm3ZYevPfYXN79xxt8l5pCWk7TCd7fErpVkZ9V1xIqyA4nsXtV28aQXEFBbmRdDLmRJHStOKRMvl3G7XJQvs9JXOcqoqKrueS3e5j7TN9D6lXgwefW8cS81Uy5ONOfH6FRwXBsW2JdnOzwammJiDiBOcC5wDDgchEZVm9/qrfajatRwJPAey3V6/dzeiIShhX0p8CxwPeteO90YDpArxT/PAFr2YfxTDivGGeD6otyw9mzJZrRk/zbtT1SKUml9O5ewqV3XQ7AY7cu5LtBOWzc3i3AkbVPV964iw9e68XBA4f+atx51RgK8yLpFF/J355bR8buGH5cG7g/MMHKhwMZJwI7VHUngIjMA6YCm5oofzlwX0uV+rOlFy0i64A1QBrwUsMCIjJHRNaLSKOP41LV51V1rKqOTUzwPukldKukoN5fxAgSujX+F3Hph/FMnHpo13b5R/GcdG4xYeH+v5EwvySGpPj9ta+TuuwnvyTGq/eeMnoPm3Z25UBFOAcqwln1Y0+O7Z/rr1BbVJgTTlK9VnYVBdmNt7L9FkNuJInJdS27xOQKCvMiDymTZJdxON3EdHCxtyScIcft5be3/sQrC1cw9coMLrt2N+dPy7DeY9dRWhTBt18mMnh4WRt9IjvmIDi2LVEVXOrwagESRWSNxzK9QXUpQLrH6wx73SFEpA/QD/iypRjb4pzeKFW9SVUrgVTg+JoCqjoDOBPw6YNiB43aT9auSHLSIqiqFJZ9GM+4c4oPKZexI4r9pWEMHbvvkG1LP2g8GfrD1t1J9Oy6l26JZYQ5XZxxwk5WrO/j1XvzimIZNTgbp8ON0+lm5OAc9mR39nPETdu6LoaUfpUk96ogLNzNpKklrFzcqU1j2JbakR59yklOOUBYmJuJU/JYuaT+0wZXLUnkrAuzATjl7Hw2fNcFEO66egzXnDuBa86dwIdv9OStF/vy8byeREa7iI6pBiAy2sXo8UXs2dG25/yC4dh6w414tQAFNY0ae3n+CHY7DXhHteWnErX1JStfAg+JyA2q+oy9zrsmTSs4w+C6B9O4/4ohuN1w1mUF9B5ykDf+0YOBI8sZd451rnPph/GcOrUIadAaz02PoCA7guHj2+Yvucvt4Im5E/jHLQtxiLLwm8HszurCNReuZeueRFas78OQvvk8eONndIipZPyINK6eupZr7ruEr9f2Y/TQbF6+/11Uhe9Se/LtBu8Spj+4XcKcWSk8NHcnDicsnhfPnm1tO7rodjl45qHBPPjMOhxOZfEHPUj7qQO/unEn2zd1ZNWSJBa93507HtrEix9/S1lpGI94jNw2pkt8JX/+tzVy7nQqSxYms/abtn3+bjAc25ZYAxk+SyuZQC+P1z3tdY2ZBszwplJRP034JyL7VLVDI+u7Y12yMg7IB/YDz6rqW83Vd/zISDUP+/aPdvew70TzsG9/+VzfWauqYw/3/QOPi9F/fjjYq7IXDVjf7L7s8YBtWL3BTGA1cIWqpjYoNxRrzKCfepHQ/NbSayzh2euzsbKyYRhHIZePrtNT1WoRmQksApzAy6qaKiKzgTWqOt8uOg2Y503Cg6P4jgzDMNqer+/IUNUFwIIG6+5t8Pr+1tRpkp5hGD7l1hC9Dc0wjNBjTThgkp5hGCFCEaq8vMUsUEzSMwzDZ1SpufA4aJmkZxiGD9VeeBy0TNIzDMNnFNPSMwwjxJiBDMMwQoYiQT+JqEl6hmH4jPUIyOBOK8EdnWEY7Yx52LdhGCFEMXdkGIYRYkxLzzCMkKEqpqVnGEbosAYyzG1ohmGEDDEXJ/vKjoOdOX/LLwIdhlciiitbLmQcturBPn1Msl9Vjusf6BBa5+N3jujt1kCGOadnGEYIMXdkGIYRMswdGYZhhBx3kLf0gjs6wzDaFVWocju8WrwhIlNEZKuI7BCRPzZR5pcisklEUkVkbkt1mpaeYRg+Y3VvfdOWEhEnMAc4G8gAVovIfFXd5FFmEHAPcLKqFotI15bqNS09wzB8ymXff9vS4oUTgR2qulNVK4F5wNQGZX4PzFHVYgBVzWupUpP0DMPwmZpLVrxZgEQRWeOxTG9QXQqQ7vE6w17naTAwWES+EZGVIjKlpRhN99YwDB9qVfe2QFXHHuEOw4BBwCSgJ7BURI5T1ZKm3mBaeoZh+JTbfk5GS4sXMoFeHq972us8ZQDzVbVKVXcB27CSYJNM0jMMw2es0VunV4sXVgODRKSfiEQA04D5Dcp8gNXKQ0QSsbq7O5ur1HRvDcPwGV9enKyq1SIyE1gEOIGXVTVVRGYDa1R1vr3tHBHZBLiAO1W1sLl6TdIzDMOnfPkISFVdACxosO5ej/8rcJu9eMUkPcMwfMZMOGAYRsgxk4gahhEyVIVqk/QMwwglpnsbKN8dQOaUgBv0Z7FweVz97Z/uR54vgURr6FyndoDzOgAgZ6dDv3CrXFcn+mCS38MdOyqTG65ZjcOhfPrFQN764Lh62487Jpfrr1lN/z7FPPT4RJat7FNve0x0JS/8ez4rvuvFnJfG+T3e5oydtJfrH8jC6VAWvhnP/55KDmw87ejYnnhsOjOnrcTpUD5ZNoS5n46st33EoGxmXraSAT2LmP38GXz9fb/abdddvIqTRqTjEGXNphSenDce2vghPSF/Tk9EZgFXYA0lu4HrgLXAbOBSYL9d9G1V/ZvPduxS5D/F6KNdIcmJ3JiLjo+GvuH1y02KQf/Q5dD3Rwj6fDefhdMSh8PNzGtX8cfZZ1NQFMOTDy/g2zW9SMvoXFsmryCWx+aczCUXpjZax1XT1rFxU4v3Wvudw6HMeCiTe6b1pyA7nCcXbGflok6kbY8KUDzt59g6xM3NV6zgjsfPJb84lmdnfcg363uzJ7vuO5pX1IGHX5nIZZM31nvvsQNyGT4wl9/db80u/uTdHzNqcDbrtvXwe9wNBXvS81vnW0TGA+cDx6vqCOAsrPvoHgR6AMep6ijgVCC8yYoOx5ZKSAmHHmEQLujpMbDigE934UtDBhaSldORnLyOVFc7+fqbvkw4Ib1emdz8Duza0wV1H/qFGtS/kC6dD7J2fdt/wRsaMrqcrN0R5KRFUl3lYMmHnRk/uTRw8bSjYzu0Xz6Z+XFkF8RR7XLy5er+nDxqT70yOYUd2ZmZgDZILKoQEe4iLMxNeLibMKebor3Rfo+5oZrr9Ly89zYg/HnGsTvWvXUVAKpaAJRgzYpwk6oetNeXqer9Pt1zgQuSPK74TnIiBa5Dyy0rR67NQe4vgLzquvWVityQg8zMheXlPg2tMYnx5eQXxNa+zi+MISHeu/2KKNOvWsPzr47xV3itktCtivysiNrXBdnhJHavClg87enYJnUuJ7/II9biWJI6exfrpp3JrNvSg/cem8u7/3iD71JTSMtppBfTBnx4G5pf+LN7uxi4V0S2AZ8DbwHFQJqqlvlxv94ZH4We0QMiBD7ahzxShP7T6sLo3O6QFAZZ1cgdeWj/CKvVGIQumLyV775PocDjl8XwjfZ0bFOSSundvYRL77ocgMduXch3g3LYuL3tTtOA1eKs9nKC0EDx22+yqu4TkTFY3dfTsZLeQ55lROQa4GYgAZigqukNtk8HpgNEJnf0fueJTsj3aNnlu9DEBvf6dfJ4/bNYeMFjUoYk+7D0CIORkbC90q9Jr6AohqTE/bWvkxLKKSyK8eq9w4bkM3xoLhdM3kp0VDVhYW4OHAzj5TcC0/IrzAknqUfd0+ASu1dRkO3bsxet0Z6ObX5JDEnxHrF22U9+iXexnjJ6D5t2duVAhXWsV/3Yk2P757Z50oPgP6fn1+aLqrqAJcASEdmINZDRW0Q62t3aV4BXRORHrHvrGr7/eeB5gLghyer1jodGQGYVZFdDohP5qhydlVC/TKELEuxdfnsAetu/mGVuiBSrBVjqgtRKuKzByK+Pbd2RQEr3Mrp1LaOgKIbTTt7Nw/8+1av3PvxEXbmzJ+1g8IDCgCU8gK3rYkjpV0lyrwoKc8KZNLWEh2f0afmN/oqnHR3brbuT6Nl1L90SyygojuGME3by4Iune/XevKJYzj91K3MXjgSBkYNzeOfzY/0Wa1NC+sFAIjIEcKvqdnvVKGAr8APwlIhcp6oH7SmhI5qq57A4Bb2pC3J3PrgVPbcD9A1HXilFh0TAhGjk/TJrcMMp0NGB3hVvvTetCnm82BrpV9BpHQ8d9fUxt9vBUy+eyEN//hyHQ1n05UD2ZHTmN5etY9tPCaxc04vBAwq4764ldIyt5KSx6fz6snVMv7XhJLKB53YJc2al8NDcnTicsHhePHu2BWbkFtrXsXW5HTwxdwL/uGUhDlEWfjOY3VlduObCtWzdk8iK9X0Y0jefB2/8jA4xlYwfkcbVU9dyzX2X8PXafowems3L97+LqvBdak++3RCYPzYNB1mCjVj36/qhYqtr+yTQGagGdmB1VUuBB4BLgDLgAPAJ8A97SuhGxQ1J1hOevdIvsfqa/CU+0CG0iqxYH+gQWkUnjGy5UJCo7OLbv+f+tuzju9ceycSeHYd009FP/9q7fZ312BHt63D585zeWmBCE5v/aC+GYRxFVEP8nJ5hGKFGcIXq6K1hGKEp2M/pmaRnGIbPhPy9t4ZhhBi1zusFM5P0DMPwqUDeYuaN4D7jaBhGu6L2QIY3izdEZIqIbBWRHSJyyBUfInK1iOSLyDp7ubalOk1LzzAMn/JV99a+cWEOcDbW821Xi8h8Vd3UoOhbqjrT23pNS88wDJ9SFa8WL5wI7FDVnfaNC/OAI75VxiQ9wzB8RrVVSS9RRNZ4LNMbVJeCNQdnjQx7XUMXi8gGEXlHRHq1FKPp3hqG4VOtuGSlwAe3oX0EvKmqFSJyHfAqcEZzbzAtPcMwfErVu8ULmYBny62nvc5jX1pYM1Ex8CLQ4jQ4JukZhuEziuB2O7xavLAaGCQi/UQkApgGzPcsICLdPV5eCGxuqVLTvTUMw6d8dW2yqlaLyExgEdZ8my+raqqIzAbWqOp84A8iciHWTE5FwNUt1WuSnmEYvqO+vfdWVRcACxqsu9fj//cA97SmTpP0DMPwLXMbmmEYoaTdzrIiIk/STM5W1T/4JaKmbKvCcWZ6y+WCQnuJ07Ioa12gQ2iVyYF/vK/XKqedFOgQ2pQC7kaeHxxMmmvprWmzKAzDODoo0F5beqr6qudrEYlRVf8/+dowjHYt2KeWavFiGREZLyKbgC3265Ei8rTfIzMMo31SL5cA8eYKwX8Dk4FCAFVdD0z0Z1CGYbRX3t13G8jBDq9Gb1U1XaRekC7/hGMYRrsX5N1bb5JeuohMAFREwoGb8eJWD8MwQpCCBvnorTfd2+uBGVhTumQBo+zXhmEYjRAvl8BosaWnqgXAlW0Qi2EYR4Mg7956M3rbX0Q+suehzxORD0Wkf1sEZxhGO3QUjN7OBf4HdAd6AG8Db/ozKMMw2qmai5O9WQLEm6QXo6qvqWq1vbwORPk7MMMw2icfTiLqF83dextv/3eh/ei1eVh5/DIaTPViGIZRK8hHb5sbyFiLleRqPsF1HtuUVs5hZRhGaJAgH8ho7t7bfm0ZiGEYR4EAD1J4w6s7MkRkODAMj3N5qvp//grKMIz2KrCDFN5oMemJyH3AJKyktwA4F1gOmKRnGMahgryl583o7SXAmUCOql4DjAQ6+TUqwzDaL7eXixdEZIqIbBWRHfaAalPlLhYRFZEWn6PrTff2gKq6RaRaROKAPOo/i7JdGjtpL9c/kIXToSx8M57/PZUc6JCaFUzxrv6qI8/+JQWXWzj38kIuuymv3vZn7+vB+m86AlBxUCgpCOe9LRsBePGB7qz6Ig51C8dPLOOGBzKRAPeGgunYjhuaxi0XrcDpUD5aOZTXvhxdb/u00zZwwbjNuNwOSvZF8dBbk8gp7sigHgXceckyYqKqcLuFVz8fzRfrBrb9B/DhJKIi4gTmAGcDGcBqEZmvqpsalOuINSfAKm/q9SbprRGRzsALWCO6+4BvWxE7IjILuAJrdhY31kjwI1gXPB+wiz2oqu+0pt7D5XAoMx7K5J5p/SnIDufJBdtZuagTaduD8/LDYIrX5YI5f+rJ3+f9RGL3Km762WBOmlxKn8EVtWWu/2tW7f8/fCmRHT9GA5C6OobU1bE8+8VWAG6/aBAbvu3AyAn72vZDeAimY+sQN3f84htufvY88kpjeenW91iW2pfduV1qy2zLTOC3j/+Ciqpwfj4hlRvPX8m9r53NwaowZs89g4yCTiTG7efl295j1ZZe7DsY2eafw4ejtycCO1R1J4CIzAOmApsalHsAK5/c6U2lLXZvVfVGVS1R1WexMu5VdjfXKyIyHjgfOF5VRwBnUfcQiStVdZS9tEnCAxgyupys3RHkpEVSXeVgyYedGT+5tK1232rBFO/WH2Lo0beC7n0qCY9QJk0t5ttFTZ/t+OqDLky6qBgAEaiscFBdKVRVCNVVQpekqrYKvVHBdGyH9c4joyCOrKI4ql1OPv9hIKcO312vzPc7UqioCgcgdU8yXTvvByA9vzMZBdbPoWBvLMX7oujc4WCbxl/L+9vQEkVkjccyvUFNKdR/4EyGva6WiBwP9FLVT7wNr7mLk49vbpuqfu/lProDBapaAbUTGCAB7NMkdKsiPyui9nVBdjhDjw/emfCDKd7CnHCSetQlqsTuVWz5PqbRsrkZ4eSmRzDqFKslN2xsOSMn7OPy0cNRhQuvyaf3oIpG39tWgunYJnUqJ7ekQ+3r/JJYhvXJa7L8+eO2sHJz70PWH9M7j3Cnm8zCOL/E6UMFqtriObimiIgD+BdePODbU3Pd2382s02BMzr93p8AACAASURBVLzcx2LgXhHZBnwOvKWqX9vb3hCRmu7tmapa6PlGO/NPB4ii8V8sI3gt+aALp5xXgtNpvc7cFUH6jkjeWJsKwD3TBrBxVRnHjdsfwCjbp8ljtjG0Vz4znrqw3vqEjvu594ovefDN0wM2O7EPu7eZ1B8/6Gmvq9ERGA4ssRtR3YD5InKhqjb5YLPmLk4+/YjCratnn4iMAU4FTgfe8hiFubK54FT1eeB5gDiJ99mhtForlbWvE7tXUZAd7qvqfS6Y4rVaRnX7LsgOJ7F7413Urz/szIyHMmpfr1jYiaHHlxMdaw3djT19L5vXxAY06QXTsc0vjSG5c935zaTO+8kvjT2k3NhBGVx11g/MmHMhVS5n7fqYyEoe+/2nPL/gRFL3BGgwRvHlbWirgUEi0g8r2U3DGhuwdqVaCiTWvBaRJcAdzeUU8O6SlSOmqi5VXaKq9wEzgYvbYr9N2bouhpR+lST3qiAs3M2kqSWsXBy8V+EEU7xDRpWTuSuSnLQIqiqFJR924aRz9h5SLm17JPtKwxg2tq6rmJRSxYZvO+Cqhuoq2LiyA70HBei8ky2Yju3m9K70TCqle/xewpwuzhq9g+U/9qlXZnBKAXdfuoy7XppC8b7o2vVhThcPX7OIhWsG8dWGAM/85qOppVS1GitfLMKarf1/qpoqIrNF5MLm3900r+7IOBIiMgRwq+p2e9UoYA9WszQg3C5hzqwUHpq7E4cTFs+LZ8+24By5heCK1xkGM/6WwZ+u6I/bJZwzrYi+Qw7y6qPdGDyynPGTrQT49YddOG1qcb3LUU49v4T133TgujOGImK19BpLmG0pmI6ty+3gX++dwuPTF+B0KB9/N4RdufFcO2U1W9KTWJ7alxkXrCQ6sooHr/oMgNziDtz98hTOHPUTowbkEBdbwc9O2AbA396cxPasxOZ26Re+vPdWVRfQYIITVb23ibKTvKlT1M9zvNhd2yeBzkA1sAPrPN07eNEUrREn8TpOzvRbnKFsUda6QIfQKpN7jAp0CF4rm3ZSoENolVVv3rH2SAYXInv10p633OpV2Z133H5E+zpc3tyGJljTxfdX1dki0hvopqrfebMDVV0LTGhk06TWBGoYRjtxFNyG9jQwHrjcfl2GdZW0YRhGPaLeL4HizTm9cap6vIj8AKCqxSIS0dKbDMMIUe14EtEaVfY9cAogIkl4fbuwYRihJtgnEfWme/sf4H2gq4j8DWtaqYf8GpVhGO1XkD8NzZvn3r4hImuxppcS4CJV3ez3yAzDaH8CfL7OG96M3vYGyoGPPNepapo/AzMMo51q70kP+IS6BwRFAf2ArcCxfozLMIx2SoL8jL833dvjPF/bs6/c6LeIDMMw/KjVt6Gp6vciMs4fwRiGcRRo791bEbnN46UDOB7IaqK4YRih7GgYyMCas6pGNdY5vnf9E45hGO1ee0569kXJHVX1jjaKxzCM9q69Jj0RCVPVahE5uS0DMgyj/RLa9+jtd1jn79aJyHzgbaB2iltVfc/PsRmG0d4cJef0ooBCrGdi1Fyvp4BJeoZhHKodJ72u9sjtj9QluxpB/rEMwwiYIM8OzSU9J9CB+smuRpt/LImIICzl0MfdBaPicT0CHUKrTG5f4VL02/GBDsFrxWcG9hkgrfbmkVfhy+6tiEwBnsDKRy+q6sMNtl8PzABcwD5guqo2fBh4Pc0lvWxVnX1kIRuGEXJ8lPTsq0fmAGdjPeh7tYjMb5DU5qrqs3b5C7GegzuluXqbm1oquGcCNAwj+Kg1euvN4oUTgR2qulNVK4F5wNR6u1P1fLJULF6k3OZaeuYpPIZhtJ73Lb1EEfF8MNjz9rOua6QA6R6vM4BDboEVkRnAbUAE1oBrs5p72HdRS282DMNoqBXn9Ap88TQ0VZ0DzBGRK4A/A1c1V75NHvZtGEYI8d3MyZlAL4/XPe11TZkHXNRSpSbpGYbhO94mPO+S3mpgkIj0sx9GNg2Y71lARAZ5vDwP2N5Spa2eWsowDKMpgu8uWbFvg50JLMK6ZOVlVU0VkdnAGlWdD8wUkbOAKqCYFrq2YJKeYRg+5svr9FR1AbCgwbp7Pf5/c2vrNEnPMAzfasd3ZBiGYbSeSXqGYYSMo2SWFcMwDO+ZpGcYRihpz5OIGoZhtJrp3hqGETq8v/A4YEzSMwzDt0zSaztjxuUx/ZYfcTiVxR/15u3XBtXbHhbu4va/rGPg0BLKSiN4+C9jyMuJYdI5GVx8xU+15foO3MvN10xk5/ZO/Oa6zZwxJYMOHau45Kyf+S32cUPTuOXnK3CK8tGqobz2xeh626edtoELTtqMy+2gZF8UD82bRE6x9XTOf03/hGP75rFhZzfufPFcv8XorbGT9nL9A1k4HcrCN+P531PJAY1n/KA0bv/ZNzgcyodrj+HVpfWP7RUT1jN17BZcbqFkfzSz359ETol1bFfOfo6fcuMByCnpwO1v+Pf4Rq/fS+JrGYhb2TspgZILuzVaLva7Ero9sYuMB4ZQ0T8GgIi0AyS9lIbjgBsVyHxgCBrRtnea+vKODH/xe9ITERew0d7XZuAqVS33WF/jIlXdfbj7cTiUG+7YyJ9vPomCvGgef2kZK5d1I3133WN7J1+Qzr6ycH7/yzOZeFYm19y4mUfuHcOSxT1ZsrgnAH367+Uvj6xm5/ZOAKxa3o2P3unHC299ebihtRy7uLnj4m+4+dnzyCuJ5aVb32PZj33Zndultsy2zAR++69fUFEVzs8npHLjBSu59//OBuCNr0YSFVHNReM3+y1GbzkcyoyHMrlnWn8KssN5csF2Vi7qRNr2qMDEI27uumA5M185n9y9sbx6/Xss3dyHXfnxtWW2Zifym2esY3vxian8YfJK/vSWdWwrqpxcOefStgnWrST9N52sewZSHR9Oz79sZf/xnajqGV2vmBxw0enTPA4OiKlb6VK6Pr2bvBv6UNknBkdZNRoWmCkxxR3cWa8t/gwcUNVRqjocqASub7C+Ztl9JDsZPKyYrIxYcrJiqa52sPTzHpx0ak69MuNOzeGLhVZyW/5Vd0aOzadhW/y0szNZ+nnd/OlbU7tQXOjfX9hhvfPIKIgjqzCOapeTz38YyKnDd9cr8/2OFCqqwgFI3ZNM1861D6Zj7faelB+M8GuM3hoyupys3RHkpEVSXeVgyYedGT+5NGDxHNszj/TCODKLrWP72cYBnHbM7npl1u6qO7Yb05PpGrcvAJFC5E/lVCVHUt01EsIc7DupC7FrDz128e9kU3JBcr1WXMzGvVT2jqayj5UI3R3DwBGApOfbCQf8oq1nWVkGDPRHxQlJBynIrfuLWJAfRULSwUPK5Ntl3C4H5fvDietUWa/MxLOy+PqzFH+E2KSkzuXklnSofZ1fGktSp/1Nlj9/3BZWbg7O54UkdKsiP6suARdkh5PYvSpg8STF7Se3tO7Y5u7tQFJc08d26pjNrNhed2wjwly8esO7vHzde5x2zC6/xhpWVEl1Qt2xq46PIKy4/rGL2FVOWGEl5aM71Vsfnl0BQPeHd9Bz1hY6f5Tr11ibI+rdEihtdk5PRMKAc4FP7VXRIrLO/v8uVf15W8XSlCHDiqk46GTPzrhAh9KkyWO2MbRXPjOeujDQoRx1zh25jWNS8rnuxboZyS987EryyzqQ0mUvT/92Pjty48ks6tRMLX7kVhLfyCTvukb+4LmVqG37ybDP4/V4aDsV/WI4MLzjoWX9Lbh7t22S9DyT2zLgJfv/B1R1VHNvFJHpwHSAKGfzP7zC/CgSkw/Uvk5MOkhhftQhZZKSD1CYH43D6SYmtoq9pXV/WSeeldnmrTyA/JIYkjvXdamSOu0nvzT2kHJjB2dw1dk/MOOpC6lyOdsyRK8V5oST1KOu9ZzYvYqC7PCAxZO/N5bkTnXHNjluH/l7Dz22Jw7I4JrTvue6l6bWO7b5ZVYrMbM4ju939WBI9wK/Jb3q+AjCCuuOXVhRJdVd6o6d46CbiPQD9HhwBwDO0iq6/fMncm4fgCs+goNDO1jdWqB8VCcid5cHJOkF+0BGW57TG6WqN9kP+PCKqj6vqmNVdWyEM6bZsts2dyal536Su5cTFuZm4llZrFpef+Rr1bJkzjw3A4BTTs9mw9pEap5/JKKccmZ2vfN5bWVzeld6JpXSPX4vYU4XZ43ewfLUPvXKDE4p4O5Ll3HXi1Mo3hfdRE2Bt3VdDCn9KknuVUFYuJtJU0tYuThALSNgU2ZXeieU0qOLdWzPPu4nlm7pW6/M4O4F3DN1Kbe/MYXi/XXHtmNUBeFOFwCdYg4woncOu/K64C8V/WMIz6kgLK8Cqt10WFnM/jF1x84d42T3cyNIe+JY0p44loqBseTcPoCK/jGUj+hIRPoBpMINLiVqcxmVKYEZPAr2c3pHzSUrbpeDZ/41nAceX4nDqXz2cS/SdnXkV9duYfuWzqxa3o3FH/fmjnt/4IX/fUHZ3ggevff42vcPH1VIQW4UOVn1WwHX3LiJSedkEhnl4tUPPmPRR72Z+9IQn8bucjv417un8Ph1C3A6lI9XDWFXTjzXTlnNlvQklqf2ZcaFK4mOrOLBqz8DILe4A3e/ZD3p7umbPqRP1xJiIqr44L7X+fu801i1tVdzu/Qbt0uYMyuFh+buxOGExfPi2bMtQL98WMf20Y9P4T9XfYLTocxfO4SdefFcd+ZqNmcmsXRLX26e8i3REVU8PM06tjWXpvRLKuaeqUtxq+AQ5dVlo+uN+vqcUyi4uifdH/nJumTltASqekbT5Z1sKvrFUD6m6T8e7tgwSs7tSs+/bAWB/SPjDjnv1yY0+G9DE1X/plwR2aeqHbxd35ROkd10QsqVvg3OT9rbw747vrUy0CG0innYt//sumLW2iN5WE+HhF46/NxbvSq76o3bj2hfh8vvLb2mEltrEp5hGO2InxtSR+qo6d4ahhEczECGYRihw8cXJ4vIFBHZKiI7ROSPjWy/TUQ2icgGEflCRPo0Vo8nk/QMw/ApcXu3tFiPiBOYg3V97zDgchEZ1qDYD8BYVR0BvAM82lK9JukZhuFTvkp6wInADlXdaV/qNg+Y6llAVb9S1XL75UqsB4I3yyQ9wzB8R7EGMrxZIFFE1ngs0xvUlgKke7zOsNc15XfAwpZCNAMZhmH4VCsGMgp8dcmKiPwKGAuc1lJZk/QMw/At343eZgKeV9n3tNfVIyJnAbOA01S1oqVKTffWMAyfqZlE1EezrKwGBolIPxGJAKYB8+vtT2Q08BxwoarmeVOpaekZhuE7qj6bRFRVq0VkJrAIcAIvq2qqiMwG1qjqfOAfQAfgbREBSFPVZqcgMknPMAzf8uHFyaq6AFjQYN29Hv8/q7V1mqRnGIZPBfsdGSbpGYbhOwoE+TMyTNIzDMO3gjvnmaRnGIZvme6tYRghJdgfAWmSnmEYvhPgqeC90X6SnqsaLSoJdBRe6bIiyOfLbkCP9e309/4WVdJ+ju85gwP/APbWeO4I329dnBzcWa/9JD3DMNqHIP+bZJKeYRg+ZVp6hmGEDnNOzzCM0OK7e2/9xSQ9wzB8y3RvDcMIGe3gYd8m6RmG4VumpWcYRkgJ7pxnkp5hGL4l7uDu35qkZxiG7yjm4mTDMEKHoObiZMMwQkyQJz3zNDTDMHzL+4d9t0hEpojIVhHZISJ/bGT7RBH5XkSqReQSb+o0Sc8wDN+pOafnzdICEXECc4BzgWHA5SIyrEGxNOBqYK63IZrurWEYPuXD0dsTgR2quhNAROYBU4FNNQVUdbe9zeudmpaeYRg+5GXX1ureJorIGo9leoPKUoB0j9cZ9rojYlp6hmH4jtKagYwCVR3rx2gadVQlvTGnFHHdrJ04HMqid7rx9gu96m0PC3dzxyNbGXjsPspKwvn7bUPJy4xi8HFl3DR7OwAi8MZTvfn280TCI9w8+vp6wiMUp1NZvjiRN57s47t4T8pj+m2bcDiUxfN78fb/DWwQr4vb71vPwKGllJVG8PCfR5OXHQNA34F7mfnHjcTEVqNu4ZZrTiYsTHn0uW9r35/Q9QBffZrCC48f67OYG/0cY7O57sZ11nFf2I+33zqm3vbhx+Uz/YYf6Ne/lIf/dhLfLOvVRE3+d+Ix6dx8yQocDuXjFUN547NR9bZfdsYGzh+/BZfbQcm+KP7++mnkFncMULRQvsJFwT+rUTfETXXS5epDf2X3feai6IVqACIHC8kPRrR1mPX57jq9TMDzy9LTXndEApb0RKQb8G/gBKAEyAVuUdVth1Ofw6HceO9PzPrtcApyI/n32+tY+WU86T/F1paZfEkO+/aGce3kE5j4szx+e/suHr7tGPZsj+HmS0bjdgldkiqZ88H3rPoqgapK4Z6rR3Cw3IkzzM1jb2xgzdIubF0fd8Sf3+FQbrgzlT/fNI6CvCge/+9yVi5LJn1X3S/Y5AvT2VcWzu8vOZ2JZ2dxzYwtPPLn43E43dxx/zr++ddR7NoeR8e4SlzVDqoqhZt+fWrt+594dRkrvup2xLE2/znc3HjT98y6+zQKCqL591Ofs/LbHqSndaotk5cXw7/+cSIXX7rVr7G0xCFubvvlcm596jzyS2J54c73+WZjH3bndKktsy09kWuX/YKKqjAuOmUTN1y0ivtfOSsg8apLyX+0mh5PhROWLGRcVUnsRAcR/evOSlWmuSn+bzUpL0bgjBOqiwJ/uYgPr9NbDQwSkX5YyW4acMWRVhqQc3oiIsD7wBJVHaCqY4B7gOTDrXPwiDKy0qLIyYimusrB0gVJjD+zqF6Zk84s5PMPrF0sX5TEyPElgFJx0InbJQBERLg9WufCwXInAGFhijPM7bP7CgcPKyErI4acrBiqqx0s/awHJ03MrVdm3MRcvvikpxXvl90YeUIBoBw/roDdOzqya7uVfMv2RuB2S7339ui1j05dKkldF++bgJv6HEOKyMrqQE5OB6qrnSxd0pvxE7LqlcnLjWX3rs64VZqopW0c0zefzIJOZBfGUe1y8sX3AzhlxO56ZX7Y3oOKKqstkLq7K1077w9ApJaKVCW8lxDe04GECx3OdrL/6/rNqLIPXHS61Ikzzjq2YfGBPcaAzy5ZUdVqYCawCNgM/E9VU0VktohcCCAiJ4hIBnAp8JyIpLZUb6BaeqcDVar6bM0KVV1/JBUmJFdQkB1Z+7ogJ4IhI8vql+laSb5dxu0SysvCiOtczd6ScIaM2Mstf9tO1x4HeezuIbVJ0OFQnnj3B3r0PsDHc3uwdcORt/KsWA5SkBtdF29eFEOOrf/go4Skg+TnRdnxOijfF05cpypSeu9HEWY/sYpOnStZ+lkP3n19QL33nnZONss+7471qBb/SUg8QEF+TN3nKIhmyNCiZt4ROEmd9pNXXNfyzy+O5Zi+eU2WP2/8FlZuClxXvDpfCUuu+/mFJQsHf6yf9CrTrOSR+bsK1A3xvw8jZoKzTeOsRxVcvuvfquoCYEGDdfd6/H81VrfXa4EavR0OrG2pkIhMrxnZqXQf9GtAWzfEccMFY7jl0tH8cno64RHWD87tFm76+fH8ZtI4Bo8oo8+gwP3lr+F0uhk2sojH7h3NXdMnMH5SDiPHFtQrM/HsLL5efMQDXSHrnBO2M7R3AW9+MTLQoTTPBVXpSo/nIkh+MJy8v1XhKgtwF9eHFyf7Q1BfsqKqz6vqWFUdG+GIarZsYW4kid0ral8ndqukMDeyfpm8CJLsMg6nEtOxmr0l9Ru76TtjOFjupO/g+sltf1kYG1Z1YsypxUfykTxiiSIx+UBdvF0PUphf/zMW5keR1PWgHa+bmA5V7C0NpyAvmh9/iGdvaQQVFU7WrOjKgKGlte/rN2gvTqeyY0sn/K2wIJrEpPK6z5F4gMKC6GbeETj5pbF07VL3c03qsp+C0thDyo0ZksGvJ//AH5+bTFV14FpNYUlCdW5dcqjOVcKS6rfcw7oKMac6kDAhPMVBRG8HVWkm6TUnUEkvFRjjywq3bexIjz4HSU45SFi4m4k/y2fll/XPZ636MoGzLrLOm50yOZ8NKzsDQnLKQRxO64fQtcdBevY/QG5GFHFdKontaI2KRUS6GD2hhIydvvmF3ra5Eym99pPcvZywMDcTz85i1dL6pzRXLUvmzPMyrHjPyGHDmkRA+H5lEn0HlBEZ6cLhdHPc6ELSd3Wofd9pZ2fx9eIePomzxc+xNZ4eKftI7raPsDAXEyelsfLbttl3a23Zk0TPpFK6J+wlzOnizON/YvmG+qPxg3oWcOe0Zdzz3GRK9gU2eUcOE6rSlKpMN1ql7PvMRezE+r+ysac5OPC91StxlSiVaW7CUwJ4Xk8Bt3q3BEigzul9CTwkItNV9XkAERkBdFLVZYdTodslPPPAAB586UfrEpB3k0nbEcuvbtrN9h87suqrBBa90407Ht3Ki4tWU1YaxiO3DQXg2DGlXPr7DKqrBXXD038dwN6ScPoO3s/tD2/F4VREYNmniXy3JMEnB8DtcvDMY8N54D/f4XAon33Uk7RdHfnV9K1s39yZVcuSWTy/F3fcv44X3vmKsr3hPPrn4wHYVxbOB2/24/H/LkcV1qzoyupv6hLmqWdlcd+tJ/okzhY/h9vBM08dz4N/X2od90X9SNvTiV9d9SPbt3Vh1bcpDBpcxF/u/4YOHSoZd1IWv/pNKjf8fkqbxOfJ5Xbw+P9O5p8zFuIQN5+sHMLunHh+d94atqQl8s3Gvtx40SqiI6uZ/bvPAcgtjuWe59o+VgAJExLvCiP7D1WoC+IudBIxwEHRs1VEHuMg9jQn0eMdlK9yk/bLCsQBCTeH4+wcyMEMBQ3uuaVEA9TMFJEeWJesjAEOAruxLlnZ3lj5TmGJOr7D1LYL8AhIJ98MdrQVjTu0ixfMyoZ0DnQIXjvuniMan2tzz419fe2RXDDcKSJZJ3S73Kuyn6Y/cUT7OlwBu05PVbOAXwZq/4Zh+EmQTy11VN2RYRhGEDBJzzCM0BHYkVlvmKRnGIbvKGAeDGQYRkgxLT3DMEKHb29D8weT9AzD8B0FDfLr9EzSMwzDtwJ4t4U3TNIzDMO3zDk9wzBChqoZvTUMI8SYlp5hGKFDUZcr0EE0yyQ9wzB8p2ZqqSBmkp5hGL4V5JesBPXMyYZhtC8KqFu9WrwhIlNEZKuI7BCRPzayPVJE3rK3rxKRvi3VaZKeYRi+o/Ykot4sLRARJzAHOBcYBlwuIsMaFPsdUKyqA4HHgUdaqtckPcMwfEpdLq8WL5wI7FDVnapaCcwDGs4kPBV41f7/O8CZ9iNmmxSwmZNbS0TygT1+qDoRKGixVHBoT7FC+4q3PcUK/ou3j6omHe6bReRTrNi8EYU1a3qN52seH2HXdQkwRVWvtV//GhinqjM9yvxol8mwX/9kl2ny2LSbgYwj+UE0R0TWBGLK6sPRnmKF9hVve4oVgjdeVQ3MA0VawXRvDcMIVpmA59PWe9rrGi0jImFAJ6CwuUpN0jMMI1itBgaJSD8RiQCmAfMblJkPXGX//xLgS23hnF276d760fMtFwka7SlWaF/xtqdYof3F22qqWi0iM4FFgBN4WVVTRWQ2sEZV5wMvAa+JyA6gCCsxNqvdDGQYhmH4guneGoYRUkzSMwwjpIRM0hMRl4isE5EfReRtEYmx1yeLyFwR2Skia0XkWxH5eRDEO0tEUkVkgx33OBEJE5GHRGS7vW6diMwKdKw1moh5iX0bUU28lwQ6Tmj2++DyiHWdN7c1tSUR6SYi80TkJ/v7ukBEBgc6rvYklAYyDqjqKAAReQO4XkQeBz4AXlXVK+xtfYALAxcmiMh44HzgeFWtEJFEIAJ4EOgGHKeqB0WkI3B7AEOt1UzMAFeq6prARdeoQ74PwL881wcb+06D97G+r9PsdSOBZGBbIGNrT0Ip6XlaBowAzgAqVfXZmg2qugd4MlCB2boDBapaAaCqBXZL5PdAX1U9aK8vA+4PWJT1HRIzQAt3BAWLmu9DsDsdqGrwfV0fwHjapZDp3tawL2A8F9gIHAt8H9iIGrUY6CUi20TkaRE5DRgIpNmJLhg1FnONNzy6iwmBCrAxDb4PANEesb4fwNAaMxxYG+gg2rtQSnrRIrIOWAOkYV3fU4+IzBGR9SKyus2j86Cq+4AxwHQgH3gLmORZRkSusX8x00Wk16G1tK3GYhaRq+3NV6rqKHtp9mr5NtTU9+GAR6wBP7dr+F4odW8POVcjIqnAxTWvVXWGfS4q4OefVNUFLAGWiMhG4Dqgt4h0VNUyVX0FeMW+4doZwFBrNRLzVc2/I6CC9txdM1Kx7jowjkAotfQa8yUQJSI3eKyLCVQwNURkiIgM8lg1CtiK1Rp5SkSi7HJO6gYLAqqJmP0xK04o+xKIFJHpNStEZISInBrAmNqdUGrpHUJVVUQuAh4XkbuwumX7gbsDGxkdgCdFpDNQDezA6jaWAg8AP4pIGXAAay6xrEAF6qGpmN8JaFRHEfv7+nPg3yJyN9a0TLuBWwIaWDtjbkMzDCOkhHr31jCMEGOSnmEYIcUkPcMwQopJeoZhhBST9AzDCCkm6R1Fmpo55DDr+m/NjCgi8qIc+rxRz7KTRGTCYexjt30xuFfrG5TZ18p93S8id7Q2RuPoY5Le0aXmFqrhQCXWzCG17PtMW01Vr1XVTc0UmQS0OukZRiCYpHf0WgYMtFthy0RkPrBJRJwi8g8RWW3Pe3cdWNMWichT9tx3nwNdayqy58Qba/9/ioh8b9+j/IU939z1wK12K/NUEUkSkXftfawWkZPt9yaIyGKx5tx7EWhxChYR+cCeNy7V804Ee9vj9vovRCTJXjdARD6137NMRIb64mAaR4+QviPjaOUxc8in9qrjgeGqustOHKWqeoKIRALfiMhiYDQwBBiGNT/bJuDlBvUmAS8AE+264lW1SESeBfap6mN2ubnA46q6XER6Yz3Y5RjgPmC5qs4WkfOAT3p8pQAAAf9JREFU33nxcX5r7yMaWC0i79qTFsRiPRzmVhG51657JtYDc65X1e0iMg54GmsKMcMATNI72tTMHAJWS+8lrG7nd6q6y15/DjBC6mYw7gQMAiYCb9qTBmSJyJeN1H8SsLSmLlUtaiKOs4BhUjeXXpyIdLD38Qv7vZ+ISLEXn+kPUjeTdS871kLAjTX7DMDrwHv2PiYAb3vsO9KLfRghxCS9o0tjM8nA/7dzvyoRBXEUx79nURBERJvRIj6AiBYRmxYxiKDRZFjfQZOvIJh8AA0W1yRbNAmG1WAQrAZF0TqGmdkrF2Q3iCBzPu3OzuXO3nCYP9xf/J642wQ0QwitWr+VXxxHA5jLxU5rY+mbpEVigM6HED4lXQJDP3QP6bmv/7B6iv0h7+mVpwXsSBoEkDQlaRhoAxtpz2+CWKW37hpYkDSZ7h1P7e/AyLd+F0AzX0jKIdQGcln+ZWCsx1hHgZcUeNPEmWbWoCqztElcNr8Bj5LW0zOkWE7drMuhV54j4n7dTarFd0ic8Z8CD+m3Y+CqfmMI4ZlYOeVE0i3V8vIMWMsHGcAuMJMOSu6oTpH3iKHZIS5zn3qM9RwYkHQPHBBDN/sAZtN/WAL2U/sWsJ3G1wFW+3gnVhBXWTGzonimZ2ZFceiZWVEcemZWFIeemRXFoWdmRXHomVlRHHpmVpQvB1av9Kk5VO0AAAAASUVORK5CYII=\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "## Conclusion"
      ],
      "metadata": {
        "id": "5HSqAkilyb2S"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "The model performed well at predicting the `PG` and `SF` labels with accuracies over `0.75`. This may be helped by the fact that `PGs` typically have unique skillsets (specializing in assists) making them easier to distinguish. However, I'm surprised at both the high precision and recall for the `SF` class, especially since `SFs` tend to not have any distinguishing characteristics.\n",
        "\n",
        "We see a lot of confusion predicting true `SG` and `PF` classes where `SGs` were most incorrectly classified as `PGs` and `PFs` were most incorrectly classified as `Cs`. A lot of times a tall `PG` may get classified as a `SG` or a smaller `C` as a `PF`, so this confusion is expected especially absent of any height/weight data.\n",
        "\n",
        "In general, adding height/weight data as input features will almost certainly boost the classification accuracy, as those two features are by far the most important in determining NBA position. However, this exercise still shows that using advanced stats alone, NBA player position can be predicted with above `50%` accuracy (well above `20%` chance)."
      ],
      "metadata": {
        "id": "I-MBa3gbNr3n"
      }
    }
  ]
}