Assignment_6_ML_FeatureSelectionAndEngineering.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "name": "Assignment_6_ML_FeatureSelectionAndEngineering.ipynb",
      "provenance": [],
      "collapsed_sections": [],
      "include_colab_link": true
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    }
  },
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "view-in-github",
        "colab_type": "text"
      },
      "source": [
        "<a href=\"https://colab.research.google.com/github/sp1999/SHALA2020/blob/master/Assignment_6_ML_FeatureSelectionAndEngineering.ipynb\" target=\"_parent\"><img src=\"https://colab.research.google.com/assets/colab-badge.svg\" alt=\"Open In Colab\"/></a>"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "blXAIWH0h5QX",
        "colab_type": "text"
      },
      "source": [
        "# Assignment on Feature Engineering (L6)\n",
        "\n",
        "This assignment has been adapted from the course **Feature Engineering for Machine Learning in Python** On DataCamp. \n",
        "\n",
        "We will explore what feature engineering is and how to get started with applying it to real-world data. We will be working with a modified subset of the [Stackoverflow survey response data](https://insights.stackoverflow.com/survey/2018/#overview). This data set records the details, and preferences of thousands of users of the StackOverflow website. "
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "V8OJHHSFhkVo",
        "colab_type": "text"
      },
      "source": [
        "## Creating Features \n",
        "---"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "xxrLsTxvi4Ef",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 71
        },
        "outputId": "ba2f56a1-6048-4aa2-a116-432911fde7aa"
      },
      "source": [
        "import pandas as pd \n",
        "import numpy as np \n",
        "import matplotlib.pyplot as plt\n",
        "import seaborn as sns "
      ],
      "execution_count": 1,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.6/dist-packages/statsmodels/tools/_testing.py:19: FutureWarning: pandas.util.testing is deprecated. Use the functions in the public API at pandas.testing instead.\n",
            "  import pandas.util.testing as tm\n"
          ],
          "name": "stderr"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "lw55w5YsB6jE",
        "colab_type": "text"
      },
      "source": [
        "---\n",
        "### Inspect your data\n",
        "---\n"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "fCpuZS41h35f",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "# Load the data which has been stored as a CSV on the URL given below:\n",
        "so_survey_csv = 'https://assets.datacamp.com/production/repositories/3752/datasets/19699a2441073ad6459bf5e3e17690e2cae86cf1/Combined_DS_v10.csv'\n",
        "\n",
        "# Import so_survey_csv into so_survey_df\n",
        "so_survey_df = pd.read_csv(so_survey_csv)"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "yO7TR7EGBgDP",
        "colab_type": "text"
      },
      "source": [
        "Instructions: For the `so_survey_df` DataFrame, \n",
        "* Print its shape and its first five rows. \n",
        "* Print the data type of each column.  "
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "-Kfzf3yVOuGk",
        "colab_type": "code",
        "outputId": "1c4fbf77-352f-4e9e-b69f-cb94b3b89e02",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 649
        }
      },
      "source": [
        "print(so_survey_df.shape)   # prints the shape of so_survey_df\n",
        "print(so_survey_df.dtypes)   # prints the data type of each column\n",
        "so_survey_df.head()   # prints first 5 rows"
      ],
      "execution_count": 3,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "(999, 11)\n",
            "SurveyDate                     object\n",
            "FormalEducation                object\n",
            "ConvertedSalary               float64\n",
            "Hobby                          object\n",
            "Country                        object\n",
            "StackOverflowJobsRecommend    float64\n",
            "VersionControl                 object\n",
            "Age                             int64\n",
            "Years Experience                int64\n",
            "Gender                         object\n",
            "RawSalary                      object\n",
            "dtype: object\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "execute_result",
          "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>SurveyDate</th>\n",
              "      <th>FormalEducation</th>\n",
              "      <th>ConvertedSalary</th>\n",
              "      <th>Hobby</th>\n",
              "      <th>Country</th>\n",
              "      <th>StackOverflowJobsRecommend</th>\n",
              "      <th>VersionControl</th>\n",
              "      <th>Age</th>\n",
              "      <th>Years Experience</th>\n",
              "      <th>Gender</th>\n",
              "      <th>RawSalary</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>2/28/18 20:20</td>\n",
              "      <td>Bachelor's degree (BA. BS. B.Eng.. etc.)</td>\n",
              "      <td>NaN</td>\n",
              "      <td>Yes</td>\n",
              "      <td>South Africa</td>\n",
              "      <td>NaN</td>\n",
              "      <td>Git</td>\n",
              "      <td>21</td>\n",
              "      <td>13</td>\n",
              "      <td>Male</td>\n",
              "      <td>NaN</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>6/28/18 13:26</td>\n",
              "      <td>Bachelor's degree (BA. BS. B.Eng.. etc.)</td>\n",
              "      <td>70841.0</td>\n",
              "      <td>Yes</td>\n",
              "      <td>Sweeden</td>\n",
              "      <td>7.0</td>\n",
              "      <td>Git;Subversion</td>\n",
              "      <td>38</td>\n",
              "      <td>9</td>\n",
              "      <td>Male</td>\n",
              "      <td>70,841.00</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>6/6/18 3:37</td>\n",
              "      <td>Bachelor's degree (BA. BS. B.Eng.. etc.)</td>\n",
              "      <td>NaN</td>\n",
              "      <td>No</td>\n",
              "      <td>Sweeden</td>\n",
              "      <td>8.0</td>\n",
              "      <td>Git</td>\n",
              "      <td>45</td>\n",
              "      <td>11</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>5/9/18 1:06</td>\n",
              "      <td>Some college/university study without earning ...</td>\n",
              "      <td>21426.0</td>\n",
              "      <td>Yes</td>\n",
              "      <td>Sweeden</td>\n",
              "      <td>NaN</td>\n",
              "      <td>Zip file back-ups</td>\n",
              "      <td>46</td>\n",
              "      <td>12</td>\n",
              "      <td>Male</td>\n",
              "      <td>21,426.00</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>4/12/18 22:41</td>\n",
              "      <td>Bachelor's degree (BA. BS. B.Eng.. etc.)</td>\n",
              "      <td>41671.0</td>\n",
              "      <td>Yes</td>\n",
              "      <td>UK</td>\n",
              "      <td>8.0</td>\n",
              "      <td>Git</td>\n",
              "      <td>39</td>\n",
              "      <td>7</td>\n",
              "      <td>Male</td>\n",
              "      <td>£41,671.00</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "      SurveyDate  ...   RawSalary\n",
              "0  2/28/18 20:20  ...         NaN\n",
              "1  6/28/18 13:26  ...   70,841.00\n",
              "2    6/6/18 3:37  ...         NaN\n",
              "3    5/9/18 1:06  ...   21,426.00\n",
              "4  4/12/18 22:41  ...  £41,671.00\n",
              "\n",
              "[5 rows x 11 columns]"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 3
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "1MzXp3T8Cl8W",
        "colab_type": "text"
      },
      "source": [
        "---\n",
        "### Selecting specific data types\n",
        "---\n",
        "Often a data set will contain columns with several different data types (like the one we are working with). The majority of machine learning models require us to have a consistent data type across features. Similarly, most feature engineering techniques are applicable to only one type of data at a time. "
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "glnEQlrVCobL",
        "colab_type": "text"
      },
      "source": [
        "Instructions:\n",
        "\n",
        "* Create a subset of `so_survey_df` consisting of only the numeric (int and float) columns and save it as `so_numeric_df`. \n",
        "* Print the column names contained in `so_numeric_df`. "
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "x3fl7fdsPUQ7",
        "colab_type": "code",
        "outputId": "33bacfe3-d748-4e1a-8dbc-edb9c583bc99",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 68
        }
      },
      "source": [
        "# Creates a subset of so_survey_df consisting of only the numeric (int and float) columns \n",
        "so_numeric_df = so_survey_df.select_dtypes(include=np.number)   \n",
        "so_numeric_df.columns    # prints the column names in so_numeric_df"
      ],
      "execution_count": 4,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "Index(['ConvertedSalary', 'StackOverflowJobsRecommend', 'Age',\n",
              "       'Years Experience'],\n",
              "      dtype='object')"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 4
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "O_4kpncQDXuj",
        "colab_type": "text"
      },
      "source": [
        "---\n",
        "### One-hot encoding and dummy variables\n",
        "---\n",
        "To use categorical variables in a machine learning model, we first need to represent them in a quantitative way. The two most common approaches are to one-hot encode the variables using or to use dummy variables. "
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "oRQM6bVaDhnT",
        "colab_type": "text"
      },
      "source": [
        "Instructions:\n",
        "\n",
        "* One-hot encode the `Country` column of `so_survey_df` DataFrame, adding \"OH\" as a prefix for each column.\n",
        "* Create dummy variables for the `Country` column, adding \"DM\" as a prefix for each column."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "bKu3ggZbQRFz",
        "colab_type": "code",
        "outputId": "2d226841-7b42-48c8-f606-c49680d5728a",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 119
        }
      },
      "source": [
        "# One-hot encoding the Country column of so_survey_df DataFrame, adding \"OH\" as a prefix for each column\n",
        "OHE = pd.get_dummies(so_survey_df[\"Country\"], prefix=\"OH\")   \n",
        "print(OHE.columns)\n",
        "\n",
        "# One-hot encoding the Country column of so_survey_df DataFrame, adding \"DM\" as a prefix for each column\n",
        "dummy_variable = pd.get_dummies(so_survey_df[\"Country\"], prefix=\"DM\")\n",
        "print(dummy_variable.columns)"
      ],
      "execution_count": 5,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Index(['OH_France', 'OH_India', 'OH_Ireland', 'OH_Russia', 'OH_South Africa',\n",
            "       'OH_Spain', 'OH_Sweeden', 'OH_UK', 'OH_USA', 'OH_Ukraine'],\n",
            "      dtype='object')\n",
            "Index(['DM_France', 'DM_India', 'DM_Ireland', 'DM_Russia', 'DM_South Africa',\n",
            "       'DM_Spain', 'DM_Sweeden', 'DM_UK', 'DM_USA', 'DM_Ukraine'],\n",
            "      dtype='object')\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "UZ56-6j9oz4H",
        "colab_type": "text"
      },
      "source": [
        "---\n",
        "### Dealing with uncommon categories\n",
        "---\n",
        "Some features can have many different categories but a very uneven distribution of their occurrences. Take for example Data Science's favorite languages to code in, some common choices are Python, R, and Julia, but there can be individuals with bespoke choices, like FORTRAN, C etc. In these cases, we may not want to create a feature for each value, but only the more common occurrences."
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "1z1TCKQ7EHcb",
        "colab_type": "text"
      },
      "source": [
        "Instructions:\n",
        "\n",
        "* Extract the `Country` column of `so_survey_df` as a series and assign it to `countries`.\n",
        "* Find the counts of each category in the newly created `countries` series."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "-5yjbaZzV3RH",
        "colab_type": "code",
        "outputId": "c366dc26-9b48-44fa-b115-01a1710de451",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 204
        }
      },
      "source": [
        "countries = pd.Series(so_survey_df[\"Country\"])\n",
        "country_counts = countries.value_counts()\n",
        "print(country_counts)   # prints the counts of each category in countries"
      ],
      "execution_count": 6,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "South Africa    166\n",
            "USA             164\n",
            "Spain           134\n",
            "Sweeden         119\n",
            "France          115\n",
            "Russia           97\n",
            "India            95\n",
            "UK               95\n",
            "Ukraine           9\n",
            "Ireland           5\n",
            "Name: Country, dtype: int64\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "lKxW1Z70EW5W",
        "colab_type": "text"
      },
      "source": [
        "Instructions:\n",
        "\n",
        "* Create a mask for values occurring less than 10 times in `country_counts`.\n",
        "* Print the first 5 rows of the mask."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "zmpeDMxnWabu",
        "colab_type": "code",
        "outputId": "c6c288a4-99fc-435a-87df-5c6d70219024",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 119
        }
      },
      "source": [
        "mask = country_counts[country_counts>=10]\n",
        "mask.head()"
      ],
      "execution_count": 7,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "South Africa    166\n",
              "USA             164\n",
              "Spain           134\n",
              "Sweeden         119\n",
              "France          115\n",
              "Name: Country, dtype: int64"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 7
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "6Zro-imOEiwt",
        "colab_type": "text"
      },
      "source": [
        "Instructions:\n",
        "\n",
        "* Label values occurring less than the `mask` cutoff as 'Other'.\n",
        "* Print the new category counts in `countries`."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "LxWk4nLAYfGE",
        "colab_type": "code",
        "outputId": "052a8483-cf63-4266-f428-d6509f7fa9e6",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 221
        }
      },
      "source": [
        "countries[\"Other\"] = np.sum(country_counts[country_counts<10])  # Labels all other categories as \"Other\"\n",
        "countries    # prints the new category counts in countries"
      ],
      "execution_count": 8,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "0        South Africa\n",
              "1             Sweeden\n",
              "2             Sweeden\n",
              "3             Sweeden\n",
              "4                  UK\n",
              "             ...     \n",
              "995             Spain\n",
              "996               USA\n",
              "997           Sweeden\n",
              "998            Russia\n",
              "Other              14\n",
              "Name: Country, Length: 1000, dtype: object"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 8
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "w0j7IlzLpke5",
        "colab_type": "text"
      },
      "source": [
        "---\n",
        "### Binarizing columns\n",
        "---\n",
        "While numeric values can often be used without any feature engineering, there will be cases when some form of manipulation can be useful. For example on some occasions, we might not care about the magnitude of a value but only care about its direction, or if it exists at all. In these situations, we will want to binarize a column. In the `so_survey_df` data, we have a large number of survey respondents that are working voluntarily (without pay). We will create a new column titled `Paid_Job` indicating whether each person is paid (their salary is greater than zero)."
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "HdCPaTzFExzS",
        "colab_type": "text"
      },
      "source": [
        "Instructions:\n",
        "\n",
        "* Create a new column called `Paid_Job` filled with zeros.\n",
        "* Replace all the `Paid_Job` values with a 1 where the corresponding `ConvertedSalary` is greater than 0."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "mp_SR3VnmLHw",
        "colab_type": "code",
        "outputId": "2ab55cdf-80cf-4d4a-ca5e-623c84ef7df3",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 428
        }
      },
      "source": [
        "so_survey_df[\"Paid_Job\"] = 0  # creating a new column called Paid_Job filled with zeros\n",
        "# replacing all the Paid_Job values with a 1 where the corresponding ConvertedSalary is greater than 0\n",
        "so_survey_df.loc[so_survey_df[\"ConvertedSalary\"]>0,[\"Paid_Job\"]] = 1 \n",
        "so_survey_df.head()"
      ],
      "execution_count": 9,
      "outputs": [
        {
          "output_type": "execute_result",
          "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>SurveyDate</th>\n",
              "      <th>FormalEducation</th>\n",
              "      <th>ConvertedSalary</th>\n",
              "      <th>Hobby</th>\n",
              "      <th>Country</th>\n",
              "      <th>StackOverflowJobsRecommend</th>\n",
              "      <th>VersionControl</th>\n",
              "      <th>Age</th>\n",
              "      <th>Years Experience</th>\n",
              "      <th>Gender</th>\n",
              "      <th>RawSalary</th>\n",
              "      <th>Paid_Job</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>2/28/18 20:20</td>\n",
              "      <td>Bachelor's degree (BA. BS. B.Eng.. etc.)</td>\n",
              "      <td>NaN</td>\n",
              "      <td>Yes</td>\n",
              "      <td>South Africa</td>\n",
              "      <td>NaN</td>\n",
              "      <td>Git</td>\n",
              "      <td>21</td>\n",
              "      <td>13</td>\n",
              "      <td>Male</td>\n",
              "      <td>NaN</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>6/28/18 13:26</td>\n",
              "      <td>Bachelor's degree (BA. BS. B.Eng.. etc.)</td>\n",
              "      <td>70841.0</td>\n",
              "      <td>Yes</td>\n",
              "      <td>Sweeden</td>\n",
              "      <td>7.0</td>\n",
              "      <td>Git;Subversion</td>\n",
              "      <td>38</td>\n",
              "      <td>9</td>\n",
              "      <td>Male</td>\n",
              "      <td>70,841.00</td>\n",
              "      <td>1</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>6/6/18 3:37</td>\n",
              "      <td>Bachelor's degree (BA. BS. B.Eng.. etc.)</td>\n",
              "      <td>NaN</td>\n",
              "      <td>No</td>\n",
              "      <td>Sweeden</td>\n",
              "      <td>8.0</td>\n",
              "      <td>Git</td>\n",
              "      <td>45</td>\n",
              "      <td>11</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>5/9/18 1:06</td>\n",
              "      <td>Some college/university study without earning ...</td>\n",
              "      <td>21426.0</td>\n",
              "      <td>Yes</td>\n",
              "      <td>Sweeden</td>\n",
              "      <td>NaN</td>\n",
              "      <td>Zip file back-ups</td>\n",
              "      <td>46</td>\n",
              "      <td>12</td>\n",
              "      <td>Male</td>\n",
              "      <td>21,426.00</td>\n",
              "      <td>1</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>4/12/18 22:41</td>\n",
              "      <td>Bachelor's degree (BA. BS. B.Eng.. etc.)</td>\n",
              "      <td>41671.0</td>\n",
              "      <td>Yes</td>\n",
              "      <td>UK</td>\n",
              "      <td>8.0</td>\n",
              "      <td>Git</td>\n",
              "      <td>39</td>\n",
              "      <td>7</td>\n",
              "      <td>Male</td>\n",
              "      <td>£41,671.00</td>\n",
              "      <td>1</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "      SurveyDate  ... Paid_Job\n",
              "0  2/28/18 20:20  ...        0\n",
              "1  6/28/18 13:26  ...        1\n",
              "2    6/6/18 3:37  ...        0\n",
              "3    5/9/18 1:06  ...        1\n",
              "4  4/12/18 22:41  ...        1\n",
              "\n",
              "[5 rows x 12 columns]"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 9
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "7lJTmpvzq_NJ",
        "colab_type": "text"
      },
      "source": [
        "---\n",
        "### Binning values\n",
        "---\n",
        "For many continuous values we will care less about the exact value of a numeric column, but instead care about the bucket it falls into. This can be useful when plotting values, or simplifying your machine learning models. It is mostly used on continuous variables where accuracy is not the biggest concern e.g. age, height, wages.\n",
        "\n",
        "Bins are created using `pd.cut(df['column_name'], bins)` where bins can be an integer specifying the number of evenly spaced bins, or a list of bin boundaries."
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "GglGS3NkFtU3",
        "colab_type": "text"
      },
      "source": [
        "Instructions:\n",
        "\n",
        "* Bin the value of the `ConvertedSalary` column in `so_survey_df` into 5 equal bins, in a new column called `equal_binned`.\n",
        "* Print the first five rows of both columns: `ConvertedSalary` and `equal_binned`. "
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "uUqb-xn3oH10",
        "colab_type": "code",
        "outputId": "fd03322f-d4cd-442f-a5ae-9df2197ad226",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 204
        }
      },
      "source": [
        "# binning the values of the ConvertedSalary column in so_survey_df into 5 equal bins\n",
        "so_survey_df[\"equal_binned\"] = pd.cut(so_survey_df[\"ConvertedSalary\"],bins = 5)\n",
        "so_survey_df[[\"ConvertedSalary\",\"equal_binned\"]].head()  # prints the first five rows of both columns"
      ],
      "execution_count": 10,
      "outputs": [
        {
          "output_type": "execute_result",
          "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>ConvertedSalary</th>\n",
              "      <th>equal_binned</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>70841.0</td>\n",
              "      <td>(-2000.0, 400000.0]</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>21426.0</td>\n",
              "      <td>(-2000.0, 400000.0]</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>41671.0</td>\n",
              "      <td>(-2000.0, 400000.0]</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "   ConvertedSalary         equal_binned\n",
              "0              NaN                  NaN\n",
              "1          70841.0  (-2000.0, 400000.0]\n",
              "2              NaN                  NaN\n",
              "3          21426.0  (-2000.0, 400000.0]\n",
              "4          41671.0  (-2000.0, 400000.0]"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 10
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "-o_1_UiiF1dc",
        "colab_type": "text"
      },
      "source": [
        "Instructions:\n",
        "\n",
        "* Bin the `ConvertedSalary` column using the boundaries in the list bins and label the bins using `labels` in a new column called `boundary_binned`. \n",
        "* Print the first 5 rows of the `boundary_binned` column. "
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "UZyNvV6krVLJ",
        "colab_type": "code",
        "outputId": "47d84d34-2a54-4add-9b9d-a7556ee9a48c",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 136
        }
      },
      "source": [
        "# Specify the boundaries of the bins\n",
        "bins = [-np.inf, 10000, 50000, 100000, 150000, np.inf]\n",
        "\n",
        "# Bin labels\n",
        "labels = ['Very low', 'Low', 'Medium', 'High', 'Very high']\n",
        "\n",
        "# binning the ConvertedSalary column using the boundaries in the bins and labelling them using labels\n",
        "so_survey_df[\"boundary_binned\"] = pd.cut(so_survey_df[\"ConvertedSalary\"], bins = bins, labels = labels)\n",
        "so_survey_df[\"boundary_binned\"].head() # prints the first 5 rows of the boundary_binned column"
      ],
      "execution_count": 11,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "0       NaN\n",
              "1    Medium\n",
              "2       NaN\n",
              "3       Low\n",
              "4       Low\n",
              "Name: boundary_binned, dtype: category\n",
              "Categories (5, object): [Very low < Low < Medium < High < Very high]"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 11
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "cIkFDAVUGhcQ",
        "colab_type": "text"
      },
      "source": [
        "## Dealing with Messy Data\n",
        "---"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "m8o-Z4YEsYF9",
        "colab_type": "text"
      },
      "source": [
        "---\n",
        "### How sparse is my data?\n",
        "---\n",
        "Most data sets contain missing values, often represented as NaN (Not a Number). If we are working with Pandas, we can easily check how many missing values exist in each column.\n",
        "\n",
        "Let's find out how many of the developers taking the survey chose to enter their age (found in the `Age` column of `so_survey_df`) and their gender (`Gender` column of `so_survey_df`)."
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "1n0JDxzMHMwD",
        "colab_type": "text"
      },
      "source": [
        "Instructions:\n",
        "\n",
        "* Subset the `so_survey_df` DataFrame to only include the `Age` and `Gender` columns.\n",
        "* Print the number of non-missing values in both columns."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "HMfXb8gSqVRu",
        "colab_type": "code",
        "outputId": "a5de99c4-52c5-427e-c04a-1d8b05596dd6",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 68
        }
      },
      "source": [
        "sub_df = so_survey_df[[\"Age\",\"Gender\"]]  # including only the Age and Gender columns in so_survey_df dataframe\n",
        "sub_df.shape[0]-sub_df.isna().sum()  # prints the number of non-missing values in both columns"
      ],
      "execution_count": 12,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "Age       999\n",
              "Gender    693\n",
              "dtype: int64"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 12
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "STDb03R3swA4",
        "colab_type": "text"
      },
      "source": [
        "---\n",
        "### Finding the missing values\n",
        "---\n",
        "While having a summary of how much of your data is missing can be useful, often we will need to find the exact locations of these missing values. Using the same subset of the StackOverflow data from the last exercise (`sub_df`), we will show how a value can be flagged as missing."
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "E5XbMPSrHhY4",
        "colab_type": "text"
      },
      "source": [
        "Instructions:\n",
        "\n",
        "* Print the first 10 entries of the `sub_df` DataFrame.\n",
        "* Print the locations of the missing values in the first 10 rows of this DataFrame.\n",
        "* Print the locations of the non-missing values in the first 10 rows."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "-d90iN3MrCxd",
        "colab_type": "code",
        "outputId": "56a1dc16-c91b-4901-e956-85eabb66a5fd",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 476
        }
      },
      "source": [
        "print(sub_df.head(10))  # prints the first 10 rows of the sub_df dataFrame\n",
        "print(\"\\n\")\n",
        "# prints the locations of the missing values in the first 10 rows \n",
        "print(sub_df.head(10).loc[sub_df.head(10)[\"Gender\"].isna() == 1])  \n",
        "print(\"\\n\")\n",
        "# prints the locations of the non-missing values in the first 10 rows\n",
        "print(sub_df.head(10).loc[sub_df.head(10)[\"Gender\"].isna() == 0])"
      ],
      "execution_count": 13,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "   Age  Gender\n",
            "0   21    Male\n",
            "1   38    Male\n",
            "2   45     NaN\n",
            "3   46    Male\n",
            "4   39    Male\n",
            "5   39    Male\n",
            "6   34    Male\n",
            "7   24  Female\n",
            "8   23    Male\n",
            "9   36     NaN\n",
            "\n",
            "\n",
            "   Age Gender\n",
            "2   45    NaN\n",
            "9   36    NaN\n",
            "\n",
            "\n",
            "   Age  Gender\n",
            "0   21    Male\n",
            "1   38    Male\n",
            "3   46    Male\n",
            "4   39    Male\n",
            "5   39    Male\n",
            "6   34    Male\n",
            "7   24  Female\n",
            "8   23    Male\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "CgR_uodEtS2o",
        "colab_type": "text"
      },
      "source": [
        "---\n",
        "### Listwise deletion\n",
        "---\n",
        "The simplest way to deal with missing values in our dataset when they are occurring entirely at random is to remove those rows, also called 'listwise deletion'.\n",
        "\n",
        "Depending on the use case, we will sometimes want to remove all missing values in our data while other times we may want to only remove a particular column if too many values are missing in that column."
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "kIioVg70H8fY",
        "colab_type": "text"
      },
      "source": [
        "Instructions:\n",
        "\n",
        "* Print the number of rows and columns in `so_survey_df`.\n",
        "* Drop all rows with missing values in `so_survey_df`.\n",
        "* Drop all columns with missing values in `so_survey_df`.\n",
        "* Drop all rows in `so_survey_df` where `Gender` is missing."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "-XAJomoFvD7A",
        "colab_type": "code",
        "outputId": "253f59b1-65fa-4c69-ef02-af23680b6a1f",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 85
        }
      },
      "source": [
        "print(so_survey_df.shape)  # prints the number of rows and columns in so_survey_df\n",
        "no_NA_row = so_survey_df.dropna()  # drops all rows with missing values in so_survey_df\n",
        "print(no_NA_row.shape)  # prints the number of rows and columns in no_NA_row\n",
        "no_NA_column = so_survey_df.dropna(axis=1)  # drops all columns with missing values in so_survey_df\n",
        "print(no_NA_column.shape)  # prints the number of rows and columns in no_NA_column\n",
        "no_NA_gender = so_survey_df.dropna(subset=[\"Gender\"]) # drops all rows where Gender is missing\n",
        "print(no_NA_gender.shape) # prints the number of rows and columns in no_NA_gender"
      ],
      "execution_count": 14,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "(999, 14)\n",
            "(264, 14)\n",
            "(999, 8)\n",
            "(693, 14)\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "9C1H8JwhtxHU",
        "colab_type": "text"
      },
      "source": [
        "---\n",
        "### Replacing missing values with constants\n",
        "---\n",
        "While removing missing data entirely maybe a correct approach in many situations, this may result in a lot of information being omitted from your models.\n",
        "\n",
        "We may find categorical columns where the missing value is a valid piece of information in itself, such as someone refusing to answer a question in a survey. In these cases, we can fill all missing values with a new category entirely, for example 'No response given'."
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "cPq_Ae40IUnY",
        "colab_type": "text"
      },
      "source": [
        "Instructions:\n",
        "\n",
        "* Print the count of occurrences of each category in `so_survey_df`'s `Gender` column.\n",
        "* Replace all missing values in the `Gender` column with the string 'Not Given'. Make changes to the original DataFrame.\n",
        "* Print the count of occurrences of updated category in `so_survey_df`'s `Gender` column."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "w2WSEXWn4pKF",
        "colab_type": "code",
        "outputId": "3c39f74c-ae77-4ef8-b157-c0c64aa18129",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 374
        }
      },
      "source": [
        "# prints the count of occurrences of each category in so_survey_df's Gender column\n",
        "print(so_survey_df[\"Gender\"].value_counts())\n",
        "print(\"\\n\")\n",
        "# Replace all missing values in the Gender column with the string 'Not Given'\n",
        "so_survey_df[\"Gender\"] = so_survey_df[\"Gender\"].fillna(\"Not Given\")\n",
        "# prints the count of occurrences of updated category in so_survey_df's Gender column\n",
        "print(so_survey_df[\"Gender\"].value_counts())"
      ],
      "execution_count": 15,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Male                                                                         632\n",
            "Female                                                                        53\n",
            "Female;Male                                                                    2\n",
            "Transgender                                                                    2\n",
            "Female;Transgender                                                             1\n",
            "Non-binary. genderqueer. or gender non-conforming                              1\n",
            "Female;Male;Transgender;Non-binary. genderqueer. or gender non-conforming      1\n",
            "Male;Non-binary. genderqueer. or gender non-conforming                         1\n",
            "Name: Gender, dtype: int64\n",
            "\n",
            "\n",
            "Male                                                                         632\n",
            "Not Given                                                                    306\n",
            "Female                                                                        53\n",
            "Female;Male                                                                    2\n",
            "Transgender                                                                    2\n",
            "Female;Transgender                                                             1\n",
            "Non-binary. genderqueer. or gender non-conforming                              1\n",
            "Female;Male;Transgender;Non-binary. genderqueer. or gender non-conforming      1\n",
            "Male;Non-binary. genderqueer. or gender non-conforming                         1\n",
            "Name: Gender, dtype: int64\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "Rn2h-F8IuI0C",
        "colab_type": "text"
      },
      "source": [
        "---\n",
        "### Filling continuous missing values\n",
        "---\n",
        "Earlier, we dealt with different methods of removing data missing values and filling in missing values with a fixed string. These approaches are valid in many cases, particularly when dealing with categorical columns but have limited use when working with continuous values. In these cases, it may be most valid to fill the missing values in the column with a value calculated from the entries present in the column."
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "_S9YZ2GsIyLx",
        "colab_type": "text"
      },
      "source": [
        "Instructions:\n",
        "\n",
        "* Print the first five rows of the `StackOverflowJobsRecommend` column of `so_survey_df`.\n",
        "* Replace the missing values in the `StackOverflowJobsRecommend` column with its mean. Make changes directly to the original DataFrame.\n",
        "* Round the decimal values that we introduced in the `StackOverflowJobsRecommend` column."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "NuQTxChw8GkM",
        "colab_type": "code",
        "outputId": "361a7332-d554-47ef-aa4a-c1976ccccdce",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 391
        }
      },
      "source": [
        "# prints the first five rows of the StackOverflowJobsRecommend column of so_survey_df\n",
        "print(so_survey_df[\"StackOverflowJobsRecommend\"].head())\n",
        "print(\"\\n\")\n",
        "# replaces the missing values in the StackOverflowJobsRecommend column with its mean\n",
        "so_survey_df[\"StackOverflowJobsRecommend\"] = so_survey_df[\"StackOverflowJobsRecommend\"].fillna(so_survey_df[\"StackOverflowJobsRecommend\"].mean())\n",
        "\n",
        "# prints the first five rows of the StackOverflowJobsRecommend column of so_survey_df\n",
        "print(so_survey_df[\"StackOverflowJobsRecommend\"].head())\n",
        "print(\"\\n\")\n",
        "# rounds the decimal values in the StackOverflowJobsRecommend column\n",
        "so_survey_df[\"StackOverflowJobsRecommend\"] = round(so_survey_df[\"StackOverflowJobsRecommend\"])\n",
        "# prints the first five rows of the StackOverflowJobsRecommend column of so_survey_df\n",
        "print(so_survey_df[\"StackOverflowJobsRecommend\"].head())"
      ],
      "execution_count": 16,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "0    NaN\n",
            "1    7.0\n",
            "2    8.0\n",
            "3    NaN\n",
            "4    8.0\n",
            "Name: StackOverflowJobsRecommend, dtype: float64\n",
            "\n",
            "\n",
            "0    7.061602\n",
            "1    7.000000\n",
            "2    8.000000\n",
            "3    7.061602\n",
            "4    8.000000\n",
            "Name: StackOverflowJobsRecommend, dtype: float64\n",
            "\n",
            "\n",
            "0    7.0\n",
            "1    7.0\n",
            "2    8.0\n",
            "3    7.0\n",
            "4    8.0\n",
            "Name: StackOverflowJobsRecommend, dtype: float64\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "SKcC12f0vCuh",
        "colab_type": "text"
      },
      "source": [
        "---\n",
        "### Dealing with stray characters (I)\n",
        "---\n",
        "In this exercise, we will work with the `RawSalary` column of so_survey_df which contains the wages of the respondents along with the currency symbols and commas, such as $42,000. When importing data from Microsoft Excel, more often that not we will come across data in this form."
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "tz0CGSQkJI03",
        "colab_type": "text"
      },
      "source": [
        "Instructions:\n",
        "\n",
        "* Remove the commas (,) from the `RawSalary` column.\n",
        "* Remove the dollar ($) signs from the `RawSalary` column.\n",
        "* Print the first five rows of updated `RawSalary` column. "
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "U_FmbC-P9LrF",
        "colab_type": "code",
        "outputId": "843b776e-b2e9-482a-baf5-9cb57122fc88",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 119
        }
      },
      "source": [
        "# removes the commas in the RawSalary column\n",
        "so_survey_df[\"RawSalary\"] = so_survey_df[\"RawSalary\"].str.replace(',', '')\n",
        "# removes the dollar signs in the RawSalary column\n",
        "so_survey_df[\"RawSalary\"] = so_survey_df[\"RawSalary\"].str.replace('$','')\n",
        "# prints the first five rows of updated RawSalary column\n",
        "so_survey_df[\"RawSalary\"].head()"
      ],
      "execution_count": 17,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "0          NaN\n",
              "1     70841.00\n",
              "2          NaN\n",
              "3     21426.00\n",
              "4    £41671.00\n",
              "Name: RawSalary, dtype: object"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 17
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "UpayqY5IwMBl",
        "colab_type": "text"
      },
      "source": [
        "---\n",
        "### Dealing with stray characters (II)\n",
        "---\n",
        "In the last exercise, we could tell quickly based off of the `df.head()` call which characters were causing an issue. In many cases this will not be so apparent. There will often be values deep within a column that are preventing us from casting a column as a numeric type so that it can be used in a model or further feature engineering.\n",
        "\n",
        "One approach to finding these values is to force the column to the data type desired using `pd.to_numeric()`, coercing any values causing issues to `NaN`, Then filtering the DataFrame by just the rows containing the `NaN` values.\n",
        "\n",
        "Try to cast the `RawSalary` column as a float and it will fail as an additional character can now be found in it. Find the character and remove it so the column can be cast as a float."
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "Z7Duf4v7JpPv",
        "colab_type": "text"
      },
      "source": [
        "Instructions:\n",
        "\n",
        "* Attempt to convert the `RawSalary` column of `so_survey_df` to numeric values coercing all failures into null values.\n",
        "* Find the indexes of the rows containing `NaN`s.\n",
        "* Print the rows in `RawSalary` based on these indexes."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "AYuh53vbv5_d",
        "colab_type": "code",
        "outputId": "c90447cc-1be9-47d5-a5b0-4a1540211783",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 221
        }
      },
      "source": [
        "# Attempt to convert the column to numeric values\n",
        "numeric_vals = pd.to_numeric(so_survey_df[\"RawSalary\"], errors='coerce')\n",
        "indices = numeric_vals.isna()  # finds the indices of the rows containing NaNs\n",
        "print(so_survey_df[\"RawSalary\"][indices]) # prints the rows based on these indices"
      ],
      "execution_count": 18,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "0            NaN\n",
            "2            NaN\n",
            "4      £41671.00\n",
            "6            NaN\n",
            "8            NaN\n",
            "         ...    \n",
            "989          NaN\n",
            "990          NaN\n",
            "992          NaN\n",
            "994          NaN\n",
            "997          NaN\n",
            "Name: RawSalary, Length: 401, dtype: object\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "vyllnJ5NKEwy",
        "colab_type": "text"
      },
      "source": [
        "Instructions:\n",
        "\n",
        "* Did you notice the pound (£) signs in the `RawSalary` column? Remove these signs like we did in the previous exercise."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "nhLE6z---zey",
        "colab_type": "code",
        "outputId": "28bc029f-4a47-42e6-cd18-65582e0d369a",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 221
        }
      },
      "source": [
        "so_survey_df[\"RawSalary\"] = so_survey_df[\"RawSalary\"].str.replace('£','')  # replaces the pound (£) signs in the RawSalary column\n",
        "print(so_survey_df[\"RawSalary\"]) # prints the RawSalary column"
      ],
      "execution_count": 19,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "0             NaN\n",
            "1        70841.00\n",
            "2             NaN\n",
            "3        21426.00\n",
            "4        41671.00\n",
            "          ...    \n",
            "994           NaN\n",
            "995      58746.00\n",
            "996      55000.00\n",
            "997           NaN\n",
            "998    1000000.00\n",
            "Name: RawSalary, Length: 999, dtype: object\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "feXhsa94wtBz",
        "colab_type": "text"
      },
      "source": [
        "---\n",
        "### Method chaining\n",
        "---\n",
        "When applying multiple operations on the same column (like in the previous exercises), you made the changes in several steps, assigning the results back in each step. However, when applying multiple successive operations on the same column, you can \"chain\" these operations together for clarity and ease of management. This can be achieved by calling multiple methods sequentially:\n",
        "```\n",
        "# Method chaining\n",
        "df['column'] = df['column'].method1().method2().method3()\n",
        "\n",
        "# Same as \n",
        "df['column'] = df['column'].method1()\n",
        "df['column'] = df['column'].method2()\n",
        "df['column'] = df['column'].method3()\n",
        "```"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "yo-JUG77KTRi",
        "colab_type": "text"
      },
      "source": [
        "Instructions:\n",
        "\n",
        "* Remove the commas (`,`) from the `RawSalary` column of `so_survey_df`.\n",
        "* Remove the dollar (`$`) signs from the `RawSalary` column.\n",
        "* Remove the pound (`£`) signs from the `RawSalary` column.\n",
        "* Convert the `RawSalary` column to float."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "7jlu4P3i_TaN",
        "colab_type": "code",
        "outputId": "0473a6ec-6fc9-49d2-86e6-df3310dcd65d",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 221
        }
      },
      "source": [
        "# using method chaining\n",
        "so_survey_df[\"RawSalary\"] = so_survey_df[\"RawSalary\"].str.replace(',','').str.replace('$','').str.replace('£','')\n",
        "print(so_survey_df[\"RawSalary\"]) # prints the RawSalary column         "
      ],
      "execution_count": 20,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "0             NaN\n",
            "1        70841.00\n",
            "2             NaN\n",
            "3        21426.00\n",
            "4        41671.00\n",
            "          ...    \n",
            "994           NaN\n",
            "995      58746.00\n",
            "996      55000.00\n",
            "997           NaN\n",
            "998    1000000.00\n",
            "Name: RawSalary, Length: 999, dtype: object\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "CqpCNv72Kl_1",
        "colab_type": "text"
      },
      "source": [
        "## Conforming to Statistical Assumptions\n",
        "---"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "hYwlN2FyyBml",
        "colab_type": "text"
      },
      "source": [
        "---\n",
        "### What does your data look like?\n",
        "---\n",
        "Up until now we have focused on creating new features and dealing with issues in our data. Feature engineering can also be used to make the most out of the data that we already have and use it more effectively when creating machine learning models.\n",
        "Many algorithms may assume that our data is normally distributed, or at least that all our columns are on the same scale. This will often not be the case, e.g. one feature may be measured in thousands of dollars while another would be number of years. In this exercise, we will create plots to examine the distributions of some numeric columns in the `so_survey_df` DataFrame. "
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "LHMvrvwxyTQu",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "so_numeric_df = so_survey_df[[\"ConvertedSalary\", \"Age\", \"Years Experience\"]]"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "i8Ch5slQy6B-",
        "colab_type": "text"
      },
      "source": [
        "Instructions:\n",
        "\n",
        "* Generate a histogram of all columns in the `so_numeric_df` DataFrame.\n",
        "* Generate box plots of the `Age` and `Years Experience` columns in the `so_numeric_df` DataFrame.\n",
        "* Generate a box plot of the `ConvertedSalary` column in the `so_numeric_df`.\n",
        "* Plot pairwise relationships (using `sns.pairplot`) in the `so_numeric_df`."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "2c37ANBR_y2L",
        "colab_type": "code",
        "outputId": "5c2de998-8232-496c-832d-69085d7277d8",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 499
        }
      },
      "source": [
        "so_numeric_df.hist(figsize=(15,8)) # Creating a histogram of all columns\n",
        "plt.show()"
      ],
      "execution_count": 22,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 1080x576 with 4 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "xSiPTqyKCoyb",
        "colab_type": "code",
        "outputId": "fb13ce92-27e8-4127-9d13-b3688ba52414",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 483
        }
      },
      "source": [
        "plt.figure(figsize=(15,8))\n",
        "so_numeric_df[[\"Age\", \"Years Experience\"]].boxplot() # creating a boxplot of Age and Years Experience columns\n",
        "plt.show()"
      ],
      "execution_count": 23,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 1080x576 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "3S_W1U7eCzyk",
        "colab_type": "code",
        "outputId": "2c544232-7513-4e54-de57-45c79cd2e456",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 494
        }
      },
      "source": [
        "plt.figure(figsize=(15,8))\n",
        "so_numeric_df[[\"ConvertedSalary\"]].boxplot() # creating a boxplot of ConvertedSalary\n",
        "plt.show()"
      ],
      "execution_count": 24,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 1080x576 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "umfmYK4uDERq",
        "colab_type": "code",
        "outputId": "efceb20f-11b2-49ae-c5ac-9dc5983cd79d",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 578
        }
      },
      "source": [
        "sns.pairplot(so_numeric_df) # plotting pairwise relationships\n",
        "plt.show()"
      ],
      "execution_count": 25,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAh8AAAIxCAYAAADt+9qXAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4yLjEsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy+j8jraAAAgAElEQVR4nOzde3xU1b0//M93zy2TCySEgJegICI09YCQAAF+rRdOLVaO1ANqlYCgcpHTetp67TmH1nNsn0dFH39Vy7WVuwqi1lZbL7Vae/AG8UIVRatCCSKEkEAuk7nt9fwxF2bP7CGTZO75vF+vvCA7e89eM7Nm7e+s9V1ri1IKREREROmiZboARERE1Lcw+CAiIqK0YvBBREREacXgg4iIiNKKwQcRERGlFYMPIiIiSqucDD5E5BEROSwiHyS4/5UisltEPhSRR1NdPiIiIopPcnGdDxH5JoA2ABuUUud2se8IAFsBXKSUahaRQUqpw+koJxEREcXKyZ4PpdRrAI5GbhOR4SLyvIjUi8hfRWRU8E8LAPxKKdUcPJaBBxERUQblZPARx2oAP1BKVQO4BcDy4PZzAJwjIttF5E0RmZaxEhIRERGsmS5AMohIMYDJAJ4QkdBmR/BfK4ARAC4AUAngNRH5J6VUS7rLSURERHkSfCDQg9OilDrP5G8NAN5SSnkBfCEinyAQjOxIZwGJiIgoIC+GXZRSxxEILK4AAAkYE/zzbxHo9YCIDERgGObzTJSTiIiIcjT4EJHHALwBYKSINIjI9QBmA7heRN4H8CGAGcHdXwDQJCK7AbwC4FalVFMmyk1EREQ5OtWWiIiIcldO9nwQERFR7mLwQURERGmVc8HHtGnTFAD+8CfZPynB+sqfFPykBOsqf1LwE1fOBR9HjhzJdBGIEsb6SrmCdZXSKeeCDyIiIsptDD6IiIgorVK2wqmIDAGwAcBgBMZ+Viulfhm1jwD4JYDvAOgAME8p9U6qykTUF+i6QlO7Bx6fHzarBqsmcHn8sFstKC+yQ9MkoWMT2T9ZxyZLNpSB4jN7fwBk5D1Ldl3J9sfLNqlcXt0H4Gal1DsiUgKgXkReUkrtjtjnEgSWOh8BYCKAFcF/iagHdF1hz6FWLNiwEw3NLlSWObFs1mjc+/weNLa5sWZuDUYOLjFtxMyOPdn+yTo2WbKhDBSf2fuz4boJcPv0tL9nya4r2f542Shlwy5KqYOhXgylVCuAjwCcHrXbDAAbVMCbAEpF5NRUlYko3zW1e8INFgA0NLtw67ZdWHzBcDQ0u7Bgw040tXsSPvZk+yfr2GTJhjJQfGbvz76mjoy8Z8muK9n+eNkoLTeWE5GhAMYCeCvqT6cD2B/xe0Nw28Go4xcCWAgAZ5xxRtzzDL3juR6Vb+/dl/boOCIzidbXVPD4/OEGK6Sh2YVSpy38f4/P361j4+2frGOTJRvKkGvSWVfN3p9CuyUj71my60q2P142SnnCafB2908C+GHwBnDdppRarZSqUUrVVFRUJLeAREmWyfpqt1pQWeY0bKssc6LF5Q3/3261dOvYePsn69hkyYYy5Jp01lWz96fD48/Ie5bsupLtj5eNUhp8iIgNgcBjs1LqKZNdDgAYEvF7ZXAbEfVAeZEda+bWhBuuUM7Hylc/C48bh5L8Ejn2ZPsn69hkyYYyUHxm78+Z5YUZec+SXVey/fGyUcpuLBecybIewFGl1A/j7HMpgO8jMNtlIoAHlVITTva4NTU1aufOnaZ/47AL9UJKsrhOVl9ThbNd8neGQFDO1lXOdsnc42VI3AKnMudjCoA5AP4mIu8Ft/0HgDMAQCm1EsAfEAg8/o7AVNv5KSwPUZ+gaYKKEodxY1Evju3NedMsG8pA8cV7fzLxniW7rmT742WblAUfSqn/RRcRugp0u/xbqspARERE2YcrnBIREVFaMfggIiKitGLwQURERGnF4IOIiIjSisEHERERpRWDDyIiIkorBh9ERESUVgw+iIiIKK0YfBAREVFaMfggIiKitGLwQURERGnF4IOIiIjSisEHERERpRWDDyIiIkorBh9ERESUVgw+iIiIKK0YfBAREVFaMfggIiKitGLwQURERGnF4IOIiIjSisEHERERpRWDDyIiIkorBh9ERESUVgw+iIiIKK0YfBAREVFaMfggIiKitGLwQURERGmVsuBDRB4RkcMi8kGcv18gIsdE5L3gz09TVRYiIiLKHtYUPvY6AA8D2HCSff6qlJqewjIQERFRlkmo50NEnhKRS0Uk4Z4SpdRrAI72uGRERESUlxINJpYDuAbApyJyt4iMTNL5J4nI+yLyRxH5epIek4iIiLJYQsGHUupPSqnZAMYB2AvgTyLyuojMFxFbD8/9DoAzlVJjADwE4LfxdhSRhSKyU0R2NjY29vB0ROnB+kq5gnWVMiXhYRQRKQcwD8ANAN4F8EsEgpGXenJipdRxpVRb8P9/AGATkYFx9l2tlKpRStVUVFT05HREacP6SrmCdZUyJaGEUxF5GsBIABsB/ItS6mDwT1tEZGdPTiwipwA4pJRSIjIBgUCoqSePRURERLmjy+AjmGRar5S63OzvSqmaOMc9BuACAANFpAHAzwDYgsesBDALwI0i4gPgAvA9pZTqyZMgIiKi3NFl8KGU0kVkJoCfd+eBlVJXd/H3hxGYiktERER9SKI5Hy+LyEwRkZSWhoiIiPJeosHHIgBPAHCLyHERaRWR4yksFxEREeWphBJOlVIlqS4IERER9Q0JL68uImUARgAoCG0LrmJKRERElLBEp9reAODfAVQCeA9ALYA3AFyUuqIRERFRPko05+PfAYwHsE8pdSGAsQBaUlYqIiIiyluJBh+dSqlOABARh1LqYwQWHSMiIiLqlkRzPhpEpBSB+6+8JCLNAPalrlhERESUrxKd7RJa3fROEXkFQH8Az6esVERERJS3Thp8iMgAk81/C/5bDOBo0ktEREREea2rno96AAqA2cqmCsBZSS8RERER5bWTBh9KqWHpKggRERH1DVxkjIiIiNKKi4wRERFRWnGRMSIiIkorLjJGREREacVFxoiIiCituMgYERERpVVXi4wVAvAqpbzB30cCGIdA7ocnDeUjogzQdYWmdg88Pj/sVgvKi+zQNLPlfojyGz8LqdFVz8fzAK4H8KmInI3ADJfNAKaLyASl1B2pLiARpZeuK+w51IoFG3aiodmFyjIn1sytwcjBJWx0qU/hZyF1uko4LVNKfRr8/7UAHlNK/QDAJQAuTWnJiCgjmto94cYWABqaXViwYSea2tnZSX0LPwup01XwoSL+fxGAlwAgOOSip6pQRJQ5Hp8/3NiGNDS74PH5M1QioszgZyF1ugo+donIfSLyIwBnA3gRAIIzX4goD9mtFlSWOQ3bKsucsFstGSoRUWbws5A6XQUfCwAcATAUwMVKqY7g9ioA96WwXEQUpOsKja1uHGjuQGOrG7quuj6oF8qL7Fgztybc6IbGucuL7Ck9L1GmRX/Wypw2fhZSpKsby7kA3G2y/XUAr6eqUEQUkImEN00TjBxcgqeXTGGGP/UZ8T5rIyqK+VlIga6m2v4NxrwPA6XU6KSXiIjC4iW8Pb1kCipKHCk7r6ZJSh+fKNtk6rPWV3U11XZ68N9/C/67MfhvHU4SlBBRcjDhjSg9+FlLr5PmfCil9iml9gH4llLqNqXU34I/twO4OD1FJOq7mPBGlB78rKVXojeWExGZEvHL5K6OFZFHROSwiHwQ7wFF5EER+buI7BKRcYkXm6hvYPInUXrws5Zeid5Y7noAj4hI/+DvLQCu6+KYdQAeBrAhzt8vATAi+DMRwIrgv0R5q7tLNWc6+bM75c33Zajz/fklKlOvQ6rPG++zBgCNre6UPt++WLcSvbFcPYAxoeBDKXUsgWNeE5GhJ9llBoANSikF4E0RKRWRU5VSBxMpE1Gu6enMlUwlf3anvPm+DHW+P79EZep1SNd5oz9r6ThvX61bCQ27iMhgEfkNgMeVUsdEpEpEru/luU8HsD/i94bgNqK8lGtLNXenvLn23Lor359fojL1OuTzeftq3Uo052MdgBcAnBb8/RMAP0xFgcyIyEIR2SkiOxsbG9N1WqIeiVdfsymbPpGFy7pT3mx6bqmQr8+vu21rpl6H7pw3mYvypeP55mvd6kqiwcdApdRWBO/nopTyAejtK3MAwJCI3yuD22IopVYrpWqUUjUVFRW9PC1RasWrr9mSTR/q5r18+XZMuecVXL58O/Ycao1ppLtT3mx5bqmSr8+vu21rpl6HRM+baN1O9nl7I1/rVlcSDT7aRaQcwbU9RKQWQJd5H134HYC5wVkvtQCOMd+D8lm2ZNMn2s3bnfJmy3NLlXx/fonK1OuQ6HmTPYSRjufbV+tWorNdfoxAsDBcRLYDqABwxckOEJHHAFwAYKCINAD4GQAbACilVgL4A4DvAPg7gA4A83tQfqKckemZKyGJdvN2p7y9eW65kOmfLe9dpmXqdUh0JkqyhzDS8Xz7at1KNPj4EMD5AEYCEAB70PUCZVd38XeFEyunEvUJ2bBseaibN7KRTkY3b0+eWy5l+mfDe5cNMvU6JDIT5dEbJprWbZtV6/F02XQ8375YtxIddnlDKeVTSn2olPpAKeUF8EYqC0ZEqZFoN2+yx8/N9NVMf+o9s7rz8+d2Y9Wc6pi63dbpS2k9pu7r6sZypyAw/dUpImMR6PUAgH4AClNcNiJKgUS7edNxo62+mulPvWdWd17cfRh3zTjXULctGnDZw9t5w7gs09Wwy7cBzENgJsr9OBF8HAfwH6krFhGlUiLdvOkIDFI1BET5L17d0TTNULcPNHcwwM1CXeVtrAcwFcBipdRFSqkLgz8zlFJPpaeIRJQJ6ZgC2Fcz/an3Eq07fXUqa7brMuFUKaWLyI8AbE5DeYgoS4Qa9+hk0GQGBn010596L9G6k456TN2X6GyXP4nILQC2AGgPbVRKHU1JqYgo49IVGPTFTH9KjkTqDgPc7JRo8HFV8N/IqbEKwFnJLQ4RZRMGBpQPWI+zT6J3tR2W6oIQERFR35DoXW0LReS/RGR18PcRIjI9tUUjIiKifJToImNrAXgATA7+fgDAz1NSIiIiIspriQYfw5VS9wLwAoBSqgMn1vwgIiIiSliiwYdHRJw4cVfb4QDcKSsVERER5a1EZ7vcCeB5AENEZDOAKQisfEpElNdy4a67lDp8/1Mj0dkuL4pIPYBaBIZb/l0pdSSlJSMiyrBcuusuJR/f/9RJdLbL7wFcDOBVpdSzDDyIqC/gXXf7Nr7/qZNozsd9AL4BYLeIbBORWSJSkMJyERFlHO+627fx/U+dhIIPpdRflFJLEFjRdBWAKwEcTmXBiIgyjTcl69v4/qdOoj0fCM52mQlgMYDxANanqlBERNmAd93t2/j+p05CCacishXABARmvDwM4C9KKT2VBSMiyjTelKxv4/ufOolOtf0NgKuVUhzoIqI+hTcl69v4/qdGolNtXxCRySIyNPIYpdSGFJWLqM/gOgLUV7CuU0iiwy4bAQwH8B6AUO+HAsDgg6gXuI4A9RWs6xQp0WGXGgBVSimVysIQ9TXx1hF4esmUpHb18hsnZVq66roZ1v/sk2jw8QGAUwAcTGFZiPqcdKwjwG+clA0ytWYG6392SnSq7UAEFhh7QUR+F/pJZcGI+oJ0rCPAVRopG2RqzQzW/+zUnRvLEVGShdYRiP5Wlsx1BLhKI2WDdNR1M6z/2SnR2S5/EZHBCCwuBgBvK6W4wilRL6VjHYHQN87IBpirNFK6ZWrNDNb/7JTobJcrASwD8CoCd7V9SERuVUptS2HZiPqEVK8jUF5kx4brJmBfUwcK7RZ0ePw4s7yQqzRS2vWmrvc0aTRTPS50cokOu/wngPGh3g4RqQDwJwAnDT5EZBqAXwKwAPi1UuruqL/PQyCoORDc9LBS6tcJl56IEuL26Vj6zAeGxjfbcEZC35TI+96bpFGuUpqdEk041aKGWZq6OlZELAB+BeASAFUArhaRKpNdtyilzgv+MPAgSrJcSLgLXVwuX74dU+55BZcv3449h1qh65zdn88Sfd97W4dDPS6nlxWiosTBwCMLJBp8PB+c6TIv2FvxHIA/dHHMBAB/V0p9rpTyAHgcwIyeF5WIeiIXEu5yIUCi5Ev0fc+FOkzd01XvxdkiMkUpdSuAVQBGB3/eALC6i8c+HcD+iN8bgtuizRSRXSKyTUSGxCnHQhHZKSI7GxsbuzgtUWZlW33NhduC8+KSGZmuq4m+77lQh6l7uur5+L8AjgOAUuoppdSPlVI/BvB08G+99XsAQ5VSowG8BGC92U5KqdVKqRqlVE1FRUUSTkuUOtlWX3PhtuC8uGRGputqou97LtRh6p6uEk4HK6X+Fr1RKfW34E3mTuYAgMiejEqcSCwNPU5TxK+/BnBvF49J1Od1NzEzFxLuOCOhb0r0fc+FOsyE6e7pKvgoPcnfnCf5GwDsADBCRIYhEHR8D8A1kTuIyKlKqdCS7ZcB+KiLxyTq03qa9Z/ttwXPhYsLJV933vdsrsNcwr37uhp22SkiC6I3isgNAOpPdqBSygfg+wBeQCCo2KqU+lBE/kdELgvudpOIfCgi7wO4CcC87j4Bor4knxMzOSOhb8qH9z2fP5ep0lXPxw8BPC0is3Ei2KgBYAdweVcPrpT6A6JmxSilfhrx/58A+El3CkzUlzExkyj78HPZfScNPpRShwBMFpELAZwb3PycUurPKS8ZEcXgUtFE2Yefy+5LaJ0PpdQrSqmHgj8MPIgyhFn/RNmHn8vuS3R5dSLKAkzMJMo+/Fx2H4MPohyTzVn/RH0VP5fdk+jy6kRERERJweCDiIiI0orDLkSUVbhSJCWKdSV3MfggoqRIxoWAK0VSonpbVxi4ZBaHXYhylK4rNLa6caC5A42tbui6ymhZ9hxqxeXLt2PKPa/g8uXbsedQa7fLxJUis0s21bFovakryaqv1HMMPohyULY1nskKGrhSZPbItjoWrTd1hUFu5jH4IMpyZt8+M9V4xvsmnKygIdFbrFPqdaeOZaKHpDd1hUFu5jHngyiLxRvX7ldg7VHj2d1x7sj9nXYLDh13m46xJ2t56URvsU6pl+gFOhV5Oj6fjsNtbnj9OmwWDYOKHbBajd+Ve1NXbFbNtL7arPw+ni58pYmyWLxvnyLS7W993e1Gj97//f3H4n4TTtby0pErRW6//UI8vWQKk00zJNGehWT3wvl8Oj4+1IorV72B85e9iitXvYGPD7XC59MN+/Wmrlg1wbJZow31ddms0bCynqUNez6I0sis5wFA3N6IeN8+LYJuf+szu0g88NIe3HnZuVBKxZw7ev9CuyXuN+FkLi/NlSKzQ6I9C8kewjjc5sbiTfWGerp4Uz2eunEyRCSmfiVSV6I/dx6fH/c+vwdLp1eh1GlDi8uLe5/fg4evGQsU9ajYaZFPM3QYfBCliVn39Ko51XBYNcxbu8O0yzrecIamad2+2EdfJMYOKcW1k4fhylVvmJ47ev8Wl/ekQyvdDRryqSHNR4kGlMm+o6vXr8cEMxXFDhxuPRGUdGdox+xz9+gNE9HY5saijfVdlrm7XxhSJd+moXPYhShNzHoeFm2sx/6jrrhd1icbzghd7E8vK0RFiaPL3A0RwbbFk7BqTjWurK7EvbNGw2HVsHR6FcYOKY05d3S3+8pXP4vpqu5pPka2z6SggETqWLLv6GqzaDHDPTdNHRHTG5Lo0E5TuwcPvBTo5diysBZLp1dh85t7sWpOdZdlNqune5vaM1J3822GDns+iNLE4/Ojothh6Opd+epnKC20YdWcasO2UJd1MoYzor8xXVw1CN+/aATmrzvR23LPzNG474U9eHd/S/jc0d3ujW1uVJY5sXXRJHj9OqyawGHVwjkf8cpk9s0xXkP69JIpHHLJMcm+o+ugYgdW1lUbejmGDixMeGgnur4JFJZceDaa270AALtFwyWjT0NFsR3r5k+AJoCuAIc1trxm9XRfUweWPvNBTN19aslkCGKHhXrKbKgon2boMPggShOn3YI7L6vC0YhG8M7LqjCgyI4fPPZuuKFdNms0nPYT3b+9zYGIbkBnVg/Bks3vGBrP25/chbXzxuOYywsRga6rmIvKidkubxmClvWvf4EffWtkuPs3stG0WTW0dfow95G3kzJbJ5k47NMz8V63ntZRs5ktowaXYOuiSfD5dVgtGqyamA7tOO0WNLa6w2Upc9rwaWObYWhi3fzx8Pr0cMAQ+oz5FTBv7duGbUUOKwYUnXgeZhf8eLlPHW4/6n7zlqGe93RIJN5QUTKHtzKNwQdRmigodHj8MY2giNcQCNy6bReeWjI55vhQoy9QcPt0+HRlmIYYb2za4/Pj/ivGoMXlxcu7D2F4RVFM4zn5rHIU2CywaIIOjx8HWjpwemlh+O9WTeDy+OHx6Vg6vQorX/0M7+5vwe1P7sLS6VXhXovyIntMo/mra8bi7n/9J/R32lBgs8DrD8xayGRD2p3xcwYpJyT7dQvNbIns5VhZV42Rg4phs2hQKlDHy5w2bLhuAvY1daDQbkGHx4+zBxWhpd2LfUdPbBt1SnFMT8X+o66Ynopbt+3CowsmGnoh127/AnfNOBdftrjCgVCBXcOibwzFrJozYNEEfl2h3e0zrbtfHGlPqCcvkRwSBRXzPH7+3G6smlONRRuNeS9lTpshAMuV+snggyhNOj06bt22K6YR3HT9RMN+Dc0udHp1+Hw6ml1eeHx+iAiUUmjp8EJXCjcGey4iG+u/H2k3JrPWVWNAsR1ftnQCCPS03PDNYTjS5sHFVYMws3oIBpU4UFpog9Omoc2twyIKDqsFulJobO3E0XYPHvjTJ7h28jDc/uSu8GM/fM1YtHX6AsFPiQMVxQ64PD585dfD4+uDShzo57TBbhG4vDr8SoWHei6uGoTls8eFe2BC5VVK4dAxFzRNMzSikQ22iMAigKYFLkqh18is4Y1u6EP7u7w+fHWsExXFDjQ0u+JeLPItya+3mto9+O07+7F23vjwxXjbzn9g4QXD4faq8EW7osgeUx/XzK3B2QOL0NjuCe8nAtOZLVsX1WLPV22GQMNmEQwZUBgeJim0W/BJc5shmF8xe1z4PQ2J11OhFDC8ohiaAOXFDvzkO6NwpM2DRRGB0Ko51bhi/BmGhPB188fHBAEr66qh635sv/1C+HQFqybY/eUx6LreZc/Mmrk1cFg0zI3ohdl0/UTTIdqBRXbD8Fa8x8uF+snggyhN/LoybQQlqo2oLHPC59exv6UDe4904MGXP0Vjmxv3zByNU/sXhIcwQscv3lSPLQtrY5NZN9XjrhnnosCm4d7n96CxzY1ls0ZjQJEN379oBJZsfgcVxQ7cNHUEhg4shK50uDw6btwcaGjXzhuPpc98gKXTq8KBB4BgoOHHHU/9zdDoF9g06Ar4wUUjDMFRaGhm/pRh4QvDi7sPY1h5IR5bUAtdKXj9Cn/c9SXOrSzFGeWFaGztQGunA2eUFeKoy4MOtx9fHGk3vBav7TmEfzmvMmYGxOB+gfKZLYq2sq4aD778CV7cfTgm18Vs2Kev56ZEB28WTeHSMacb8oVWzB6Hdrcf16w5MeSw+YaJpq/b4wtr0enVoQng9ftht4j58JvfmLypK4XWTn9MD8na7V8YznHj5ndw14xzMX/djvCxHR6/aU8FYBx22XzDRCzatCMmIfyuGecats1buwNPLK7FXTPODQdHxQUWtHUCV61+88TrUlcNq1Uw4+HthqETs9cl+hyNrW7cNm1k+MtKqJfUZtUMw0KNre6E62e29eAx+CBKk3jj1jZNcHHVoPAFcfnscbj7jx+h1GnHwvOH44HvnQeLCBQUlIJpY+2LE9gU2i24+Yn38diCWnR6/di28x+omzQM169/ExXFDtzy7ZHhwOJPPz4fN26OXdej1GkzPPbiC4bH9ODcuPkdPLagFm6fHg48xg4pxeILhsNh1XDrt0dh2QsfY/EFw7FoYz3GDinFN0cOxtVr3jQkwUb2hDxw5RjoShm+dYaChVCOSugiGCpHqCGfv25HOHiKDtSWTq/Ci7sPh3Ndlk6vwqKN9abDPvmW5NcdZr0+jy+sjckXunHzO1g3f0LMxdPsdfP69ZgLvtlnQteVoUdj+exxeO79A4aegAdf/gQzq4fgxd2HDecYVlEUfszKMieGDHDGJLCurKvGo2/uTajMhXZLzLYOt24IcMzq2o2b6vHk4kmGMre6faY9GtHncPtOBPehx7t12y48deNkQ0+KrsdOSzarn9nYg8fggyhNRID7rxiDm594P9wA3H/FGECAm6aegzsv+zp8OvD+P5rwH9/5GhRg6PlYO3887BYN2xZPQlO7J5x3UVnmhF9Xpo14iyuQT/Jliws3P/E+HplXAy1YjgFFdmzb+Y/wEEmBTcOv51bDomno9PpRYAtMtfX69fAwTanTFh5miTxX6BwWTcKBR2RgEwochgxwYsvCWgwoshsCB7Mk2B9tfT/mG2Eo6Lht267wuSI1NLvC5ygvtmPyWeVY8M2zwkMEa177HKVOm2H/Uqct3BhbNOBAc0f4m2Gy17DIJZFTVEMXSp/fPMgtsIphxpZFM8/p2Xukw/B+/uK53THDbyvrqnH3Hz8y7Ldk8zvYcN0EQ+LyPTNHY3A/h+G8T9bvh00Tw7BQf6cNbm+noadCoPC9iWdg9qRh0JWCJgIo889Q9ATayjInoq/XZkM7FcUONLZ5cNezuw3P7c7LqvBvjxoTzLWo7s8BRXbT17nDY0xqXTWnOvzFJbJ8NqtmCFLMckgy3YPH4IMoTZQC/vzRVzHj5XMnD0NjqxtOuwWFdg1DK/qh7jfGRvaZdw/gSKvb0A0bGs5Y8I2z0NrpxbJZo3Hrtl3hoZQzygtxsCXQq9Di8qKi2IHWTh+uW3diyu2t00ah4WigQer06ihyWHGgpROD+9kh0LD5homwWzXcNPUcwzfHZbNG497nA8MVQKDBO73MCQHCCXrRvRK3P7kLjy+shd2qwWoRQwAzqMRh2tiafes85vLitmkjw+eN7GUpL7KjwGbBgCIb+jltqJt0pmGIYPnscRhYbPx/la0AACAASURBVMeWhbXhi1VlmRNPLZmMtk4fLovoIl8ztwYjKorjrvKZbd3YyabrOq7/P2cZguXNN0w0BKItLi/e2duEYy5fzEX2N9dW48sWd/iCP2SAE6v/8nnMtPIBRXZDYDCgyIbGVk/Mfi6v3xAIvbbnEC4bW2k474rZ42CzCgqUBf5gsqpSCr9774AhafRgSzt0OHHjph2GC/nG68fDr0s4r8Sv+1HssBl6UpbNGo0Oj99QPoXYYMtsbZLFm+rxyLU1MW3A7EnDDOfo77SZBkLRSa2LNtbj0RsmYvfBVkP9jJ5htun6iVnXgydK5daiPjU1NWrnzp2mfxt6x3M9esy9d1/amyJRfkjJVSOyvrZ0dKKxzYuGo65wQ1s5wImKYhtu2/Y3vLj7MP704/PD3dIhlWVOrJ03Hste+Dim0Z8zeRj8uoJFE7g8PhTYLDjm8hkChRV11dj4+l5cOX4IfrjlvfDF+oGrxsCvgH80GfNKPv3qGKqHDcSNwceI7lIOlSk0vBEZjFSU2PGDqedA1xVm/Gp7zOvxyi3n48db3g/nn4QCmD/ffL4hlyV0juWzx+HgsU7Dc55VcwaOubw4tX8BOjw+LHthT0xC7IrZ41BaaMPVwTyEyMfccN0EXHT/X8IXyVGDS9Ds8uLy5dtjApnTSp0YVOyISWoFkMlu7JTXVQA42OLCFcHVb0MWfWMoZo0/w1CHR55SjCtXvWnY7+KqQTG5P4HVfAX7j3aGjx1UYsfgfgVocfnCF/x+BVZ81thmCLR/dc1YaCKGx9t43QTMMakzWxdNQofHH348u1Vg0QCXR4W3lTqt+O7y12OO3XzDRMz+9VuGejS4nwMfftkaLvNZg4rQGvUZ27xgIo67fOHPTGWZExuvn4AVr3wW0/N244VnY05E78Xy2eNQUmDFJ4fawvX866eVYF+TK6ZO//SZD8MBf8hfb7sQFk3CSbyFdg3TH9pueG5r543HY2/vM7QfT9bvxy8uH53qno+4dZU9H0Rp4vEBR1rdMdn5bq8VSy48G42tHmhintNhtYjhAhvKkfheRILbPTNHw27R8KOt78WMPT96w0So4O+hIZHo3pVQLsWjC2pxzZoTF5N4swWGDSzCn378Tew/6goHEavmVOPGTfVYO2+86Tc3vw7c8u2RuO+FPVi7/Qssu2IMmtoCvT6hnpvIC45FxPjNtq4ay174OJwfs7KuGkunVxmCjFAewmMLak3LHVJR7EBjqxv9nFZYJNATE50HE5nEag028AePuSAieOClPVnVjZ1sHpNlzi8YNdi0DkcPw82sHhIOFIDA6/PLP32C7180wnDsA1cGpoBH5vVsXVQbk1N0tN0bk1PR1O4xfX/dPmNeyaYbJuBYu98QGKyoq8bks8qxtb7BcGxk3keoHm1ZWAuPX0chLPD4dRxs6cQtwd6g0H6fH27Hm581Gno0oJRpz1uH2xczpPTYglrDUu+v3XYhXttzyPB4Hp+Oxja34fmGEmcjb5Gwqq465v34498OxvRerqyrRlnEEGS6cXl1ojTx+GOn2t64+R3sPtgKl8eP26aNDOduRKosc8KqaYYZJ/EWCjulfwEqio0Xv4ZmFw4e68Tnje1Y9I2h+OXVYw2PFTp28QXD0dDsgq6M4/qhe7pEl6mhuQNNbR7MX7cj/G0slJza6fXjnpnGpdjvmTkaLo8Ptz+5CzdffA6unTwM89a+jatWv4kDza7wjb5CS2C3dvrC0x7Dr9emesysHhL+ffGmeoiY537oyvy19OsqHIAtfeYDfPPeV3HV6jdx27SRuG3aSKx//QssnV6Fp5dMxtp546EHpzgfPN6JK1e9ge8/+i72fNWK2y/5GjZdPwFjh5SGz5lPiagWLfbOyaf0LzCtwzdNHWHYr9wkZyFeXk9zu9fwvof+FsksAG5q95i+v3ujhiZ8foQDj3CZN9Vj4fnDY47t9Brfv4bmQDL3oo31uGr1m1i0sR5iUr6BxXZ8c+RgzF+3Axfd/xfMX7cDNosl5vku2fwOCkyGEnWlsGpONbYsrMWqOdUosAmmn1dpeDwFYN388YbP1LJZo/HVsU7jUMym+pj341+rK02HgI66er40u64rNLa6caC5A42t7m4vL8/ggyhN4k21LbRbcOu2XRgyoBBFDgtWzB5naGBWzB4HwHhs9AyU0GMdOt6J26aNDF8QQ4/R4vLig4YWTD+vEs1xvjGGEi81MV50zO7p8sCVY3D/i5/EBCah37881hm+iIcuKutf/wJfBhvKU/oXGAKgpnZP+EZfoUa+wGbe4xKdMOqLE7A1t3uwos54/47ls8dhzWufY/EFw2MCsNB7cO3kYbjr2d24fPnrmL9uB1o6vLj7jx/B5fFj8lnl4aBl6v1/wR1P/Q23fDvweudbIqrdosW87/ECvaEDiwz7VZQ4Yt4Ts4CkojiQ6HzXs7tx1eo3cdezu9Hc7sXFVYMM+4Wmy0Z6sn4/lpt8Vh58+VPDfvF6E20WibmQO6Lev8oyJyyaGAKDUH5HJKfdGlOfzHqOGppdMeMQgS8XguEVxagocWB4RTF8fhUTMC3eFPhMRH6m7n1+DwpsmqF8FcUOnFVhfD9O7V9gWpZOr46eSMa9mVI67CIi0wD8EoAFwK+VUndH/d0BYAOAagBNAK5SSu1NZZmIMsUWZ6ptaEbKgebAjJRHF0zEloW14cWK/vv3H2Jm9RDDsfHuMNvUHsiuj8zHCA2p3DtrNOav24H1100wPba82IH1101Aq8uLtfPHY36wK7wxOCxy3xVjMKjEAYdVQ6vbi8Y2dzgwCX0bfrJ+P1bUVeOhl2MXJguVI9SgR55/5auf4Z6Zow37Dyiyx329In8XIGbI5p6Zo/Hfv9+NCUNL8fjCWvj1wIwGmwVYeP5wWOOsL+HTlWmv0NLpVcFVMWtx+Hgnlk6vwsu7D2Fq1WA4rBqWXTEmsFhVD2+mlq0K7RZDMqhF4qxMazHOMHH7/DHvSXlx7Pt509QRMcMzizbVY8N1EwxJlAOKbDELe/37P5+DIoexfOXFdlSU2A3JoFaL+efOpytDAuu9z+/B/3fVeYbEzxV11dCVMgz9rZs/PmbqrtVk5lW8GWh2q2Y4x8q6avh041DRxusnmNdPvzIMz1xcNQgKMJRv2azRcNoshsXIPD7ztU4sPcweSsb6NykLPkTEAuBXAL4FoAHADhH5nVJqd8Ru1wNoVkqdLSLfA3APgKtSVSaiTOrn1LCirtow9rx89jhsemOfIQi5Zs1b2Hj9BBxpdePMAYWYP2UY1m7/wnBxDn3rWxK1mNd9LwTyEM6qKMJfbr0Anze2hxfRCl3wbRaJudCvqKvGvc9/FM6lePB7Y/Hk4knYHxxb/+/f7Q5P6318YS1aXT48sXgSOr063F4fHl9Yi6+OdaKsyI6twcS2fgVWrL8ucOOuvUc6cN8LgYXOVtZVh6fxhhqvd/e3YP3rX2DtvPE42u5Bi8uLLW/vi2nkl88eh4f/HPhmG2pov2wJDNk8tqAWX7a4cFqpEzc99i7e3d+Cd/e34FtfPxWzVr6BRd8YiunnVeLG4FofZo2xWZd6qLelodmFw8c7MWvlG4ayhF6zNXNr0laX0sHl9ePO3+3G4guGh/MdOjw+rJg9zpD4uWL2ONgsYshtePiasagocRgCgyKTvJ4zys1vGCcCw7EKwIAim2FbWaENd/4uEJifKJ8/Zr2YlXXVhmA6VN+3vr0Pq/66N3zeUBARnWdxTVQ+0by1O/DAlecZApeWDk9Mfdq28x8xn/cVdYEhlcjn4bAKfvGccWrx3iMdpvXTFhFIVZY58Z+XVoUTZEPH3rptF7YurMVpZSduj3C03R3z2kffQ6o7krH+TSp7PiYA+LtS6nMAEJHHAcwAEBl8zABwZ/D/2wA8LCKicm0KDlECjrl0PPTyJ4ZG6+E/f4qrJ5yJGWNPx30v7AEQ+BBrIvjR1vex4boJuPf5PVh8wXD0K7Di0QW1aOv0oqTAhk6vL/xNvKndEw4yAt9oBBaLoMCmhZPUdBXMefCr8JBIqdOGAUX2cBJn6Pw3Pf4uHl9Yi1kr3zA8h4ZmF/y6wo+2vo9N109Eh8eHSx/8Xzy2YCIAoLndE2jQIxr1sUNK8dA1Y3H/lWPCGf8AYoKnH0w9x5BMuqKuGq98dChcTq9fh90quHrCmbj+/5yFjuAqpv/9u91obHPD59dx8xPvY+N1EwxTgEO3HI+c/mvW07Js1ujwMWa9LZGPFRq/j1ywLN8STi2ahIfCQl655Xzc/+JHhjr80J8/xU++8zXDBfD7j76LRxdMNCRqdnr1cF5P6NiDLS7T1/vQcbfh2OWv/B1XTzjTsLDXtsWT8OLuw4Y1Ll65+fyYPIvQCsCRd7Dt57TgglGD8dwHhwzvf0NzB65e81b48bYsNE9a1pWxB+KxBRNjLu4XjBqMkgILHl1QC6UURAR/3n0QZw4sMTyPyjJnuB6FPPjypzFB3j0zR0PTYLjhnj/OImO+qCtoqTMwqygy6BncrwClzp711CVj/ZtUBh+nA9gf8XsDgInx9lFK+UTkGIByAEcidxKRhQAWAsAZZ5yRqvISJUW8+urTVUxjCQD/eWkVfvj4e4YLphYcW7dbNcMFYMvCWly1+k2smlONu57dHZ6dEd3tur+5A/c+vwc/u6wq3OC4vX6sqKvGkTYP5k8ZFm4oQ414pFCQYdo7ECzboeOd4YuyJoK7//gx7g3mCEQe09jmhsvjN6z7AQBlhdbw8JImgjf+3ohbvz0Kd1zyNfh1hQ6PD1vqGwzHXFw1CHf+y9fR6dOxr6kjHHgsmzUaR9o84X9DZQ31BgEwDPW8u78F970QuBCOOqUEH3/Vinuf34P/mfH1mKAktJ5K5GOFXqPo/JNcSzg9Wdtq0yTmgqqJmNbh26aNMvze0OxCU5vHcIH+620XxgQzF1cNwqq66pj7qdgtEp5REuq9WPrbDwznCCWcRtYP/0lWAN4fcQO6syoKUeywRq0vYsf9L+4xHBtvafYOj/F9tlk0/OK5j2KGce64ZBSuWv2mYd8tC2tjyhc9XNfY5kanVzc83vrXv8BP/+XrhnyTw62dpuUrsBnTOTVNMLS8CCUFtqSsS1NeZI+7/k2icmKqrVJqNYDVQGAueoaLQ3RS8eprvOXVrcFvmKHfAxfQzvD4cOTQQ6gxjPzmft8Le3DXjHOD90Rxo7zYjmXPf4x397dg+St/x39eWgVNABEBoGAptqPd7Qs3vKWF5rkVIjDtNn7mnYZwb0Bo+KepLZAwetu2XaYXb7Mu6OnnVaKfU8NXxzwQ0fDgK5+hofnEBebiqkExwy4/mHoOCh0a2jw+AMAdl4xCh8ePihIHOjx+FIqG452+8Cqqy174OBzURQdT7+5vwV3P7sbaeePDy6sfafPgwZc/xX1XjMEp/Qvg8yu4PD7cNu1ruPf5jwxrLJjln+RawunJ2la7TVBebFwAzBYnfyIU8EVua+nwGrY1d3higpn5U4ah0K4Zhjq2f3oYF1WdYuypKLDGTDN9sn5/TOBit8a5hUFUcoPNouGhP39qGLJ5qn4/vn/RiJhck+h6uyqYoxE5/DGgyB4TWJkFKfG2hRJ0DcMzVglPmz8xTGIMKgYWOUyDgIFFsb1vmiZJ65XTNMHIwSWGvJLuBjMpW2RMRCYBuFMp9e3g7z8BAKXU/xuxzwvBfd4QESuArwBUnGzYhYuMUYqkfOGmzk4fPm1qj7mYn1bqwK79x8MN/MBiOx7686e4aeo5GDW4BABwODisUGCzoKnNgwUbdxpuCmfVNFi0wCqqogEdbt2wSuNv/roXV9QMgdNuweJN9eFjh1UUocCqBb6lRpXroZc/wT+d1h/fHVcZXoL6t+80YEt9Q7g3YOE3h2NgiR2Nxz0osGm4MeJmdcMGFsFuEYgG6DpQ7AjcOTeUSFtg19DfYcfB4y4oAMejFm5aVVeNin52uL0qfH6rBagoKsDB4y64fQoWAawWLXAeERzvPLFmxMVVg/CDqeeEX+/InI/onJvXP28y3JQu8g64Hl9iN6nLt0XGdF3hQEsH3L4Ti3MVOTQcbfcaEj8Di4dphrU6VtVVw2HTYu4GCwD7IxYoO2OAEzoQk48xuJ8dHx44sbDX6CH9cLDFHbNORXmxDR8fPHH321GnFqOpzRuz36mlDnS49XA9ctgEja2emOdRUmDFZ4fbDeWzWjV8HrFt1KnFaHP7DQutnTWoCG2dvi5fl0C+U+y28mIb2iM+sw5r4C3+e8R5zywvxNDyopj6leUr7cYtSCqDDyuATwBMBXAAwA4A1yilPozY598A/JNSanEw4fRflVJXnuxxGXxQiqSlQe/s9KHJ5QlfgIsdGnx6YGnz0EqlmgYoJRhU7IDVGjsbPrKxARBusJx2DS5PRNARXF5aA9Dp02HVBBVFdhxz+8O3pi+wCVweHQU2DZ3eE4FBiVNDq+vE7+VOO5pcnnAZLZpAV4BFAIdNQ4dHh0AhNNMukdvbh/4e2q7rgfP5dQWrRcOgYgc0TU56TPR2n08PB2rW4K3dWzp94f1KC6xobPeE/15o19DuTqzRjj5nZHCS5kY/LXUVMH/PdF0ZXuNBwXVlerpN1/XAexKsaxVFdmiaFrOfz6cbPjvlTjssFgnsF9xmOEfENqVUzDmUAo50nNg2sDDw/iVyrK6jx2Ux22axaDGvM4BsDioSlf7gAwBE5DsA/i8CU20fUUr9QkT+B8BOpdTvRKQAwEYAYwEcBfC9UIJqPAw+KEXS1qAT9RLrKuWKzCyvrpT6A4A/RG37acT/OwFckcoyJKInQQsDFiIiop7JiYTTbJQLvSz5GlTlwmtPRETx5dxdbUWkEcC+OH8eiKhpun0cXw+jk70eR5RS05J9wi7qa7RceL9yoYxAfpczE3U1V17PrvB5pFfcuppzwcfJiMhOpVR+LTPYC3w9jLL99cj28gG5UUaA5Uy2XClnV/g8sgdvLEdERERpxeCDiIiI0irfgo/VmS5AluHrYZTtr0e2lw/IjTICLGey5Uo5u8LnkSXyKueDiIiIsl++9XwQERFRlmPwQURERGnF4IOIiIjSisEHERERpVXOBR/Tpk1TAPjDn2T/pATrK39S8JMSrKv8ScFPXDkXfBw5kgsryhIFsL5SrmBdpXTKueCDiIiIchuDDyIiIkorBh9ERESUVtZMF4CIiCiX6LpCU7sHHp8fdqsF5UV2aJpkulg5hcEHERFRgnRdYc+hVizYsBMNzS5UljmxZm4NRg4uYQDSDRx2ISIiSlBTuycceABAQ7MLCzbsRFO7J8Mlyy3s+SDKgKF3PNej4/befWmSS0JE3eHx+cOBR0hDswsenz9DJcpN7PkgIiJKkN1qQWWZ07CtsswJu9WSoRLlJgYfRERECSovsmPN3JpwABLK+Sgvsme4ZLklbcMuIvIjADcgsOTq3wDMB3AqgMcBlAOoBzBHKcWBMyIiykqaJhg5uARPL5nC2S69kJaeDxE5HcBNAGqUUucCsAD4HoB7ADyglDobQDOA69NRHiIiop7SNEFFiQOnlxWiosTBwKMH0jnsYgXgFBErgEIABwFcBGBb8O/rAXw3jeUhIiKiDEhL8KGUOgDgPgD/QCDoOIbAMEuLUsoX3K0BwOnpKA8RERFlTrqGXcoAzAAwDMBpAIoATOvG8QtFZKeI7GxsbExRKYmSg/WVcgXrKmVKuoZd/hnAF0qpRqWUF8BTAKYAKA0OwwBAJYADZgcrpVYrpWqUUjUVFRXpKTFRD7G+Uq5gXaVMSVfw8Q8AtSJSKCICYCqA3QBeATAruM+1AJ5JU3mIiIgoQ9KV8/EWAoml7yAwzVYDsBrA7QB+LCJ/R2C67W/SUR4iIiLKnLSt86GU+hmAn0Vt/hzAhHSVgYiIiDKPK5wSERFRWjH4ICIiorRi8EFERERplbacDyIiomTSdYWmdg/vsZKDGHwQEVHO0XWFPYdasWDDTjQ0u8J3lx05uIQBSA7gsAsREeWcpnZPOPAAgIZmFxZs2Immdt4YPRcw+CAiopzj8fnDgUdIQ7MLHp8/QyWi7siLYRezcT8AHAskIsoT0e28zaqhssxpCEAqy5ywWy0pPS+vJcmR88FHvHE/h1XD3Efe5lggEVGOM2vnN1w3AWvm1sS0/aEvn6k6L68lyZHzwy7xxv32NXVwLJCIKA+YtfNzH3kbg/s58PSSKdh++4V4esmUpAcFzCtJnZzv+Yg37ldaaIvZxrFAIqLcE6+dd3n8OL2sMO3n1XUdja1uDsX0Qs73fNitFlSWOQ3bKsuc6O+0YeyQUsO2ZI8FEhFR6sVr51Pdppud9+KqQTjS7sHly7djyj2v4PLl27HnUCt0XaW0LPkm53s+yovsWDd/PPYfdaHQbkGHx4+yIhvu/uNHWHbFGNz6xPtobHMnfSwwmzFBiohyWXQbVua0meZ3lDltKe2BKC+yx5z3vy6twjW/fitmKObpJVNQUeLo0Xl8Ph2H29zw+nXYLBoGFTtgtcb2DeRT257zwQcAuH06lj7zQbhy3H/FGDS2etDS4cHPv3suTi0tQKkzd9+k7mCCFBHlsnht2IiKYjy9ZIohIPm0sS2lbZ2mCUYOLjGcN9lTfH0+HR8fasXiTfXh57GyrhqjBpcYApB8a9vTMuwiIiNF5L2In+Mi8kMRGSAiL4nIp8F/y7r72E3tHizaWG+IQm9+4n3cNHVE4G+b6uHXA/s2trpxoLkDja3urOsi03WVlPIxQYqIclm8NqzZ5UVFiQOnlxWiosSBZpe3V21dT9tcpz25Q0CH29zhwAMIPI/Fm+pxuM1t2C/f2va09HwopfYAOA8ARMQC4ACApwHcAeBlpdTdInJH8Pfbu/PY8aLQM8oLccvW98MRaTZHjMmMaLnwDhHlskTbsN60dYm2uab7zanBuvnjMW/tDkNPRZnTdpIzxuf166bPwxf61pyE55uNMpFwOhXAZ0qpfQBmAFgf3L4ewHe7+2CRCUFjh5Ri1ZxqbFs8Kfz30N+yOWJMZkSbqcQsIqJkSLQNExFcXDUIq+ZUY8vCWqyaU42LqwZBpOsvbIm2uab7bdyJtk4f1s4bjz/ffD7WzhuP37/XgGaXt0fP12bRTJ+v1WK8POdb256JnI/vAXgs+P/BSqmDwf9/BWCw2QEishDAQgA444wzDH8LJQQ98NIeXDt5GG5/clc4Gl02azScdgtEkNURYzIjWrMEqb6UbJsNTlZfibJJNtbVRNuwApvgB1PPwY0RuRIr6qpRYOs6+Oht70qJ04ZrIxaxvGfmaOi6saciUYOKHVhZVx2T8zGo2Ji8mm9te1qDDxGxA7gMwE+i/6aUUiJiOuimlFoNYDUA1NTUGPbRNMHgfg7cNu1rmLf2bUOEeuu2XbjvijEAJC3L8PZUKKJNRvnMEqRyOSM6F52svhJlk2ysq4m2YS6PHg48gECbf+OmemxZWAsUnfwciba58fb7R9Qilrc/uQtbF01CT1itGkYNLsHWRZPg8+uwxpntkm9te7qHXS4B8I5S6lDw90MicioABP893JMHdXn8cHl8WDq9Ktz9NnZIKRqaXTi1fwEGFTuwZm5NuMsq2yLGUESbrPJpmhgSs3K1chIRxePTlWmvhD+BxNFE21yz/VbVVePBlz+NOa9SicVuZomuVquG00qdOKO8CKeVOk2n2QL51bane9jlapwYcgGA3wG4FsDdwX+f6cmDOu0WqHbgrmd3G7rB1r/+BQTZHzFme/mIiNIl0WTQUK5EdK9EdK6EmUTbXLP9LBrQGDUTJdGe6nybLtsbaev5EJEiAN8C8FTE5rsBfEtEPgXwz8Hfu82nKyzZ/E5MN9gdl3wNP39uN7463gkAWR0x5lNES0TUU4kmg4ZyJSJ7JcxyJeIxa3PNeiWi9yt19rynOt+my/ZG2no+lFLtAMqjtjUhMPulV7w+86lKx1xevLj7MBZ+M/D/vhhdEhHlkkSTQRPNlUhUor0Svempzrfpsr2RFyucipgnlB5udaOyzImmdg9+uOW9Xi1/S5TLht7xXLeP2Xv3pSkoCdHJxWvPzabQhnIlkiFer4TZdSPUG9JdyZxckOty/sZyAGAR4J6Zow3dYPfMHI0n6/fjnpmjsfLVz/psdElElEviteeWFHdap6NXItmTC3JZXvR8aJqG9a9/gaXTqzCoxIFihxUev45bvz0Kt23bhXf3t/TZ6JKIKJdEtuelThtaXF6sf/0L/OLy0Sk9bzp6JTi54IS8CD7KnDb816VVONzqRqfXD7dPx+B+Dhw6HshIjowu8+mugERE6ZbsNtTsDrY/+tbItC+m1dtFvBJ9XXo6ZJNvcj740HUVvrNhRbEDt00biVueeD9ceVbVVYfvaguA05yIiHoo2VNFE72DbTq+JPamV4JTaLsv53M+mto9+O07+7F23ng8ePVY3LptlyFhaNGmenR6dei6wlfHO9HuDixGFlqErK9OcyIi6q5kTxWN93jH3bH3STGbBpusu4GH9HTJg3jP40hboGz7mtpxoLkDXi/zDkNyvudDoHDpmNMxf90O3H/FGNOEoS9bXDja7sGDL3+CF3cfDicw3ffCHry7v4WJqERECUh2UqbZ41UUO3CwpROLIu51suG6CXD79JieBYdVw9yIe6xkqrch3uvS7vFjzm/eMtx7ZtSgYthszD/M+Z4Pt08PLzDW4vKa3vWvqd2DxZvqMbN6CIATi5AtvmA4E1GJiBKU7Durmj3eTVNHhAMPINBe72vqMO1Z2Bd1j5UFG3biSLs7qb0hPX0elWVO7D3SHnPvmcNRq6P2VTkffESu77/y1c9Mp2iFptqWOm3h4xqaXeEEo744zYmIqLuSPVXU7PGGDSyK6UUotFtMexYK7ZaYbR1uPy5fvh1T7nkFly/fjj2HWlMegJg9j5Vx7gHjS0MwlAtyftglcn3/d/e34L4X9uCuGedimFj6GwAAIABJREFUWEURjrS6UVHiwP1XjgEAFNg0bFlYixaXF0/W78dppU6c0q+ACUFERAlI9lRRTROMqCjG1kWT4PXrsFk0WLXYRcY6PH7TabAdHuNwT2WZE42tbsM03Qde2oO7Z46GX4ehzABiZqeYbUvkuZm9Lrqum94DxqoJGlvdSTlvLsv54KOiyI4VddXhWys3trkxsMSBAqvAbtUw+9cnxtuWzx6HJ+sb8PrnTeF7AOT7G0xElEzJnCoaOVsxMr9jZV01FkfkfAwZ4MT9V4zBzREzGe+/YgzKi+3hoKSyzIl188ejqc1juMno/VeMwdF2D+at3dFlvkhvckiiXxePx2e4NoVyPnSlcPny7V3ms+T7TJmcDz5aOn149r0GrJ03HhZN4NcVtu38B+omDYu52dySze9g7bzx2FrfgMWb6rncOhFRBpnNEpn7yNt44MrzDL0Xze1e/OZ/Pzds+83/fo67Zpxr6G3w+XXMe2KH4fFufuJ93DXj3JjckES39fQ6ccztR/0XR/DogloopSAi6PT6cPWat2LyWZY+80HSzpsrcj748Pj8WPXXvVj1172G7dfUDjUdI7QEI0kut05E1H3JXGQs3iwRXSks2lgf3jZ2SCnu+u65ht6QFXXVGFBoN8wcOdDcYfp4ZrkhpYU2rJpTHQ5mVr76mel+ZteJRF4Dj8+PO5/9GHc++3F425aFtQnns+T79SltwYeIlAL4NYBzASgA1wHYA2ALgKEA9gK4UinV3J3Hjbckrtm4YWWZE/5gsg9nuRARdU+yF9OK135H53I0trlxan8HtiyshU9XsGqCQcWOmCmrNquWcG5If6cNP3js3fDzWDZrNLSom9eZXScSfQ3MnptZ7kq8fJZ8vz6lc7bLLwE8r5QaBWAMgI8A3AHgZaXUCAAvB3/vlvIiO9bMCWQZjx1SirXzxmPj9RPQ7vFhxexxhuzj5bPHYc1rn/fpm/n0RcleiIior+rtImPRn8Uyp8109syZ5YUx28oKA4t/nVlehNPLCk3XyrBqgmWzjDMeH7hyDIYMcBq2rZpTjbv/+JHhedy6bRdO6V9gPO+c2OtEoq+B2QyYM8sLE9q2Zm4Nypy2vG630tLzISL9AXwTwDwAUEp5AHhEZAaAC4K7rQfwKoDbu/PYuq5gswoeuPI8FNg03BjM86gsc+LX11Zj/XUTIAh0tZQUWDG79kx8/6KzcVp/Z14n81AAlz0mSp7eLDLWnaXUAfRoRo3L48e9z+8x5Ib8P3/4GA9fMxZ3zTgXhXYLOjx+lBXa8OLuwzHPQyll2M9mFei6Mpw70dcg3swgs+cWva3MaYtJxM23ditdwy7DADQCWCsiYwDUA/h3AIOVUgeD+3wFYHB3H/hwmxvz1u7A0ulV4QznsUNKsfiC4Wjr9KO/U8Nt23ahsc2Nu2acC49fx789+k7eJ/NQQLxvKXz/iboWndsQb1jDbIgg+liLhm59FhP5fMaUz6Khsc1tyBepLHPik0NtmL9uR3jbX2+70PR5iIhhv8oyJ7YumoTTSk8sICZiPqQvkvhN5Lra1tjqzvt2K13DLlYA4wCsUEqNBdCOqCEWpZRCoIMihogsFJGdIrKzsbHR8DevXw8vIBYKPG759kjc9exuzFr5Buav24Hbpo3Euvnj8eDLn4b3y/dkHgpI9nLQiThZfSXKJierq6GeisgFu9o6fQktMmZ27MGWTlQUGy+cvfksmp2jsc2NR+Z1vdgXoEwXpIy+BDU0u+Dz64ZtFoHpsZYkdkhkot1Kt3T1fDQAaFBKvRX8fRsCwcchETlVKXVQRE4FcNjsYKXUagCrAaCmpsZQO0KLjOlK4YlFk3BK/wJcvebNmLG8rYsm4fZLRqG82I6188bDac/vZB4KiJfQlspkrpPVV6JscrK6Gm8a7O++P6XLIRGzYxdtqsequmrYrZphWQSzz6LPp+Nwmzu88NigYgesVq3rc2ysx31XjDEMu7g8/pjFvtw+hfWvf2HYb/3rX+C2aV8z7FdZ5oTVYjyvpml4bc+hmOUdvnbq2TGLh/V0iCQT7Va6pSX4UEp9JSL7RWSkUmoPgKkAdgd/rgVwd/DfZ7r72IOKHVg3fzzcXj+Od/pw6HinacTY6fXjlogFatbMrUGpM/9XkevrQklf0WOnTDYmOrl4375dHj9OLyvs9rGTzyqHDmD+uhOLfa2oq0ZpgfEy5PPp+PhQq2Fa7cq6aowaXGIIQOKVT4CYabqr6qoNN6pzWAU3TT0n5hx264khldC2QVG9NWVOG/7lvErD81g3fzwOHXNjwcbk5Gj0hXYrnet8/ADAZhGxA/gcwHwEhn22isj1APYBuLK7D2q1auhXYMOHR11Y+swHWDq9yjRiNLsBUT6Nn5G5ZC8HTdRX9Obbt9mxC88fjnlr3za0wzduqo/JqTjc5g4HBaH9Fpvs161puqUFMW3Aqf0Uti6aBJ9fhzXYuwIgZlt0j0uzyxtTvv3B60+yrjF9od1KW/ChlHoPQI3Jn6b29rG9fh2n9ndg2azRGBKMVg2L0cweh58+86HhmHwbP6P4krkcNFFf0Ztv3+VFdmy4bgL2NXWcmDliEdOeiuicilAeX/R+Vg34ssUVHoqpiFM+R0RS7Ml6uTVNDMEMEJw9adGgVOBfTZOYISCLgAuFJUHOr3AaqCwSHpf73pq3UFHswF0zzsWZ5YGuwUPHO01v8JNP42dERMnU22/fbp8e7g2oLHPi8YW15gtCRuVURN4sNGTRN4biUKvHcJ+UlXXVGDkoedN0zaYCP7ZgIo65fDHDMxdXDTJM1U32QmF9YYmAdC4ylhJN7R60uf3YfzSQWFpR7MBt00ZiaHkhdBVYdGbnF0exPGrBsXwbPyMiSrZQr+HpZYWoKEn8RpxN7R488FJgvY0tC2uxdHoV/vThQayoq46ZiRKdUzGo2IGVUfvNmTwsHHgAJ4Zijru98Pp1+HQFr18Pr8kRXWazhQZ9Ph1ftriwr6kdX7a40OyKnd7q9inTIaD/vLQqdqGwOcYymy1Qlqh4SwQcaXfnzcJjOd/zEerWKrRbUFHswM8uq4LL48eciDsTrqyrxu/fa8DS6VUoL7JjUImDi4wREaWIruu4dvIw3P7krnA7fM/M0Ti91NFlToXVqmHU4BLDfmZDMZPPKseXx9wxvSHRiammvQhzamCziuFOtyvrqlFR7DCcRzMZYgkltUb2rvR3WLCvxWVYoMxqFfj9OjSt+70f8ZJpO9x+1P3mrbzoDcn54MNutcDl1dHh8eOmqSPQ3O6NSfxZvKkeG66bgMZWN5raPVj92mf4xeWjmQdAlAJD73iu28fsvfvSFJSEUiGRm6r5FcKBBxBoh29/cldM0mg8miaG3AsBEkpgDSWm2ixauHwKKrYXYePOmCm5D778CW6aOsKwyJiuYs8bGiqKvH4caO7A/LU7YvbbsrC2y5lBZuIl035xpD1vJk7kfPBRXmRHp9eH08sKYNE0HGl1m0aMAHD3Hz9GY5sb98wcDV3XzR6OiIjiSDQXQSll2g7//+ydeXgUVbr/v6eql3S6A2lCgkjYRRQxQDqEAM64zXh13IZhUwhIUBYd9I7juMy9OuNzce5PRK6KXkhglFWQzWUuXh1nUBwVGSHioDeyDBAgCCSEhKSTTi9V5/dHd1f6VJ3ORtLpDufzPDyak1N1TlWqT739nvf9vkEtydaPsXZ2riGRIFoAq19RMaXoS63f+vvHcPulp1gZ+YVFE7O0mjLhNrtVwvJ8F+NdWZ7vQs9kM3O+gMq/XqWN2yK8YN+ifBeefvc7wxiJGtSa8MaHJBH07maD2+dDvTfoAbllWAYmuvpqFu224pNQKbB4chaOVNRhza5jePau4Z09dYEgrmmLB0PQtWlpuYJo39xtFrlZIa5o4mbv/XIcU9U2fE79GBIhjEfjTE0Dt98JnfzCk9v2Y/PcPEY8rKLWh/f/ccogKNZz/CCYvIp2HdGqqFtNEpOhw9tm4sEL9pUldKnEiYQ3PlSVoqKuAefcfry64xB+fcuVWHDTEDwUUWBueb4LW/YcR9FnpR0ihSsQCASXAi2V/eZ9c187Oxdna7zNek2ijVHbwMY7bJmfx/VKyBLV6nyFpRYM8gv5LqzbVWoYw6tQVgRteja+Kq1G0Wds33vGDMCMCO/K5nl5WDQxi4lxeW3aKJxz+xhxM15MSjT0EgGqSruU8FjCGx+VdT4EFGgP4MyxA/DU298ahGyeuWMY8Fkps/coECQawhsh6ExaKjzG++ZOQTFz2a42e0308Q4HTrux8avjjJfj1R2HcG9uf3b9f/NrvDRlJLff5uIyZoxS3RgPvvk1Ft493FBsTt/v4BnjXBr8KhZs2desWFpL6WrCYwlvfPgCCpSI/UWzLHGt5lSbmfm5qb3HlgRUCQQCwaUGTzysf1oynDYzdztFH5SZ7rAyL+jCnUda5DXhxTskW2R8VFLO6G0AwP3XDWJ+LqvyQKWUkVwHgN/+rFENO+yVeIYTUzGgp73Zfkt3HMZzPx/OeDnW3Z/bIlG11tCVBBNbbXwQQnoB+E8Al1NKbyOEDAMwllL6ervPrgVYTDJUf6PAS7XHz7Waqz1+5udo+2SXgriLQCAQtBW9eNja2bk4XOFuds20WWQ8cetQPL61cWti8aQsQ5HPlsY7RBP20surR2szSYRJje1hN3NjKuwWmUn7tZqJoR9Pwj2gqC0SVbtUaYvnYzWAVQD+PfTzIQCbAHSK8ZFmt6DK49X29A6drjHs7xXlu2CSCYpmuLCt+CQe/enQqPtkLQ2o4iE8JgKBoCvDWx+PV9Zz65q8/dA4EJDGbRdKNcMj3O/xrfvx9oPjDF4TNSQaFlApiKIiNcnoDenbw4YVM104e8GrGRC9ulthkVl59ZemjIDTbmHaivJd2FFyGuOHZGiBpB99dxqrC0bj5HmP0atT50PYV97NYuZ6f7pZzaiK+JLb024xvIt4omodQSK8i9pifPSklG4mhPwWACilAUJIp+X6qCrFmWovlu44hP+eNgqSJOF/vinD2tm5qG0IICXJhOc/+B4flZQj02nDsunZcFijRwe3NKCKNw/hMREIBF0Z3voYra6Jx6dg+h8bA0SjpbzW+9hA0rWzc1HvUwwv7e42E+OpcCab8UN1A+OFWT49G90cJmZr5z//9wAK87MZr4REKLIHpDHBpW/MyjF4dVbOyMHxqnpGjGx1wWj4A9TQ70RVPWZGiFuunJljmHOyRe7w90GivIva4v+pI4SkAUFDkBCSB+BCu86qFZS7vZi3vhgflZQj2WLCg+uLkT0gDTPf+Apnahow842vtD3BsioPHnrza3gDFNUeH/d84WCnSFqSzhTNY1JZxx+nPeBJBgsEAkF74fcrOFVVj+OVdThVVY8ks2RYH8PbH5FkOm3wK5RZD4+dq+P20weSHq+s50qan6pugC8UL+FTVHj8Kh4MZTWG+z345tfwBYLxHVNX7Ma8dcWocHshSRIjud7gV7WMyPCxp6oaMG8dO+6cdXtx8ryHaTt5Ptiu78ernP7P8joUrN6DqSt2o2D1Hsx84yvuO6E91/LOeBe1hbYYH78G8CcAgwkhXwBYC+Dhdp1VKwhEyO7KEtGCSyP/G0lZlQfV9T6crm7g/oHDwU6trQPTVo9JWwlbtxOWfYHxiz7BhGVf4ODZWmGACASCdsHvV3Cg3I2pK3bj+sU7MXXFbvxwwYtVs9j1MdOZhCWTRzBtiyZmweMLMOf74Ft+bZelOw4z/aJ5UjJSrFi4vQRTV+zGwu0lUJoQ9mpu/eYdG23cZF1MSkv7RWvTvxPaey2P9buorbR624VS+jUh5HoAQwEQAAcppf5mDgMhpBRALQAFQIBSmkMI6YFgvMgAAKUAplBKq1ozHzlC3EVRKW4ZloEeof29aMGnlXU+bCs+iWfvGg5KKbMn1tZ0ppamoLUXFxObIhAIBM1R7vYairk9uL4Yb+mEuDx+Ba9/fpTZ6liz6xjuze3PnO+2a3vj1R2HmH4en8INJOUJRZbXepljw0aGIaBTItxKt5FxJVaTsXJutABWCqBohksblwLc+bU00NVskpi5yBLadS2P9buorbQl2+UXuqYrCSEXAHxLKS3nHRPBjZTScxE/PwVgB6X0eULIU6Gfn2zNfEwS0YJ69h0/j4dvvhKL/3wAiyZmYc2uYwbhl+XTs7HzQDnuGzeQkeCN3BNrSzoTLz2sIwVgEsW6FQgEiUlTkuGRsRKrC0bj0Z8M1bYiwmtf5As+02nDwJ52Q2rsFFemQShsUIYdD980RNtSCYuCmWUwcuh/vM9lOHbZ9GxYTRJ66sS59DEQRTNceG3aKCzYsE9rc9rN3GQFs4kw426YMwYP33ylUXLdYTak5CZbZKZt5cwcuBsCTGxIEaeg3cWs5bF+F7UV0hKtfeYAQt4HMBbAJ6GmGwAUAxgI4D8opeuiHFcKICfS+CCEHARwA6X0NCGkN4CdlNKhTY2fk5ND9+7dq/1cUeuFxx+AL0BhNUm4d+VulFV5MKpvKubfMBiZThu628zwhconr/2yFI//y1XahydMptN20V6DWEYYV9R6MWHZF+1+DZcwHfKH0j+vYYRYGIsoLNcqYvKsnqqqx9QVuw1rzMY5efjRC58wbX9aMB6KCoO3IXI9pKD4RYTIGACsmjUaG786zngRutvM2ss+cgye2NeW+XlwNyiM9PkDP76CWQOjrZUvTh6BCx4/472Y86PBOF/v09pGZHbHpMIvDXOOzO6JvC8lp2uY8z0/MYu5L7IE3PWacS68a7uYtTyOsl2iDtqWbBcTgKsppWcBTfdjLYAxAP4GgGt8IBig+hEhhAIoopSuANCLUno69PszAHpxZ0/IXABzAaBfv37M79LsFpyqDkClKhTaWGho38lqTVRm09w8TF2xWzvmqduu7hCvQSwFYBLFur0Uaep5FQjiiaae1QyHlStf/u7XZUy/cGYLr3prc/LgPG/Iprl5LY6p8AWo9rt6n4Kpuf0NgmfRvMSXdUtiPBorZ+TAapbw6OZvtLY3HzBm6ESL+QgoqkHI7Pd3svflVFU999iBOiGz1qzl0QyNeP8S2hbjo2/Y8AhRHmo7TwhpKvbjOkrpKUJIBoC/EEIORP6SUkpDhomBkKGyAgha55G/kySCPqnJqPX64G7g79npBcZMsnG/Lx73xJqiq0ntdiWael4FgniiqWfVbJZxVYaDKebmSJKxqZg1Plq6dkaTXG9p7AUvfkIipNmU1w0PjOGeT9aJjFnNEvo5k5n5+QLGuUSbn6L7pPPuS7R4jGSr3Ka1PFHSanm0xfjYSQjZDmBL6OeJoTY7gOpoB1FKT4X+W04IeQdALoCzhJDeEdsuzcWMRIcGt130lnphvgsen4JRfVNR4fZi+fRsVNV5sWTyCDwWafUmoNcgEaxbgUCQuJjNMvPNXVUpV2CLt3b6/QrK3V7NcMlwWEEIYcTD0jly7Vdk2FGU72KkyotmuBjF0LA66pkLDYaU14V3D2fannu/BMunZ7MxJNOzsf7LY8gekIZkyPApKp7/4Hs8N+FakIidAkeSZIgD6dvDxi1U180m46+/vh4SAVQKWE3E4IVx2vgCZT3t1jYZC9ESD3jbYPFmjLTF+PglgF8AuC70814Et0/qANzIOyBkmEiU0trQ/98C4D8QTNm9D8Dzof++19rJqCpFlceLH6qDkdnjBqVh3f25AAhKz9XhmXe/04yO1GQzXt3xT+w6WonXpo3C87+4Fv16JMNmMcXlH0cgEAjiDYMQ18wcQ59wmq7+i2CSWWIEuwrzXUg2ywbvRTedOJfVJGHNF6VMtssLHx7EU7ddxYzL256pqPUhySwx5+tmM+Gmqy9jvoAumpgFf0DV4lzCMSVW3bEAkOYwY3VBrmZoWEwENZ6AQYzscLmbG4jb3P1rKbwtpXSHFaerGxjjLR69IW1JtaWEkKMA8gBMBnAMwLZmDusF4B1CSHjMDZTSDwkhewBsJoTcD+A4gCmtnU9wr0vVHvLNxWW4b/wAlNcEJXfn3zAYhTuPaNUJbx7WC5uLy7Bgwz4svHs4TLIkDA+BQCDgoI8noKDcb9p6KXVfQDGk6c5fX2zwSsxfX4wXJ49gjIqX/noQ9+b2NwRg8oIyJUKYNNhtxSfhsJrwl0d/rAWh+gIqClbvZV7S+qDRsqpgtfPVBblMW0ABFn94ABNdfTUPyQsfHog6v8hjT573cGXn9f0uNq1Wn/bbI9miGR7tMUZH0WLjgxByJYB7Q//OIajPQSilXG9HJJTSowBGcNorAdzc4tly8AUUJiXssZ8MAaVgLMtFE7Pw4p8PItkiIxlBq7isyoMBPZPx7J++w6M/HRp3VqFAIBB0Jrx4gqgS6V5WIp0XqBktaDQ9xcoEfi6amIVuSSZDv35pycy2y/Lp2SCEYOHmksZx54xBjSfApAIvz3dh3KA0bI6IVYkWNFrbwIYtyhJw37iBjFxDtPldjBhZW5MdnDYzHrn5SoMUfXum7nYUrfF8HADwGYA7KKX/BABCyKMdMqtWYDHJQERQ0N3ZmZi2crdWuvny7kmwWUx4+Z6RkAiBSqkW/6FS4KOScpScro07q1AgEAg6E148QVgiXR8wqZdI9yt8ATBe0OgJnSx52AOh71ehExmr8ylMSm5ZlQf+AOUKo60uyGWMj2hBo96AygqKUWDNrmMGAbUnbr3aML+LESPTC4+11Btf5fFzpeh5XqJ4S6hojfHxCwD3APiEEPIhgLfQQfnmrSHNbkFNg1cLNFUpRbrDit/8y1Cs2XUM940byAQaLZ6UhWfvGgYKoN4blP+NR6tQIBAIOhNePMHSHYeNwaD5Ljz97ndMv69LK7lpumYZrPci34Xf6Y4tq/LALBNDv242E+Mh4XlXJAKut0F/vswewSDWcC2XsMfAapaYMbbMz+N6PmwWVkCtMN8Fq06MbNO8louRef0K7l3Z6DlqaYxGtDTii0ndjRUtNj4ope8CeDcUMHo3gF8ByCCELAfwDqX0ow6aY5NIEkGdV9Vke00SwSM3D8GT2/bjmTuGaQ8N0FjCeeHdw2GzyLgQSsGNR6tQIBAIOhNCiME7UOH2Ij3FyqSFyhIMEunZA9LwwoffMx6DV3ccwuzrBjFt3W0mw7GZzqBEenPHRpbWCKNScD0aEmHPt/jDA3jmjmEGmfiCVXuY90VAgeEd8uS2/dg0N49JQTbLEn6xfBfTz+Pje2HWzs5l5rJ0xyHcm9u/TTEa7Z26G0vaEnBaB2ADgA2EECeCQadPAugU4wMIFgmqqA1W7PP4FQzsaUdZVfTCcn172JBklvB2cVncWoUCgUDQmcgEBkmCJZNHQCKhkuYhulnNBvEws0wM4mEA8NufDWO8A2/NHWMogbFoYhYAGAS7Hr5piPaFEQDMMsHSe0bhkbcaJdItJmJIq102PRsqVZlzhd8XR8/VaUbA5d2TtO36cFtDFM8CBWANbflbTDIUVTX0i+aF4V3b/dcNMvRriTc+mtgkL3U3jlRPAbQt1VYjVAROE6npLJIsMp64dSge37of6Q4r/mvqCE1cjGcVnjzvQZrDgrnXD8JTP7sKqbb4swoFAoGgMzGbJEOKamqyCWdrvcx2xcqZORiS7mC+aUdqcoTJdAa/9G2eNxYBRYVJlgBKWxRTccuwDFAAC7c3BpcWzXAhLcXCzM8XUFFcWsl4NL44XI7sAWnMsS9NGQFvQGXa3piVg3/72VV4dPM/GMPllmEZjBGV6bRBVakm2R6ei75fNC+MnnC8iL6trcJtPKMiHsXIWl3bpbPh1co4Xe3B5FCRuKIZLmwrPon7xg3UYj70VvWLfz6ICrcXC+8ejuF9uhskgOPJOmwLXeEa2pMW3g9R26UTEbVdWkVMntXy2ga8/rcjmJTTT3uR13kDWLBxn+GFqt8iOHvBg6Pn6vD41sa1d/GkLFx1mQP1PlXbrkizWXC4so4bF3HgtDtCeMyh1e2KHJdXn+XZu67BwTONxw5Ot2PaH/9uOFYflBmtZsuGB8Zox4cNjVf+eshgkOj7bZgzBhfq/QZxs4wUK0pO12rz62E3I6NbEg5EtPVPS0Zm9yRU1PkYkTazuW3hAR1RC+xi19WL8nzEA6pKGddYqs2Mj0rKUVHrw/wbBqNbkgmrZo2GSSY4dNaNF/98EPtOBoVYky0y49qKR+uwtXSFa2hPxP0QCNqGBOD2EX2YtNV19+dytxL0WwQNARUvfHiQ8WjsPXYejiSzwdDo18OKt+bmQVEpZInAZpFQfsHHyCWsj5K626sbm6b72rRRqKrzM8dGm3NLU2NVgPGudLeZDNtJZVUeyBJhPBAEFPWmACNGFlAV+BRWpG3J5BFQVcq0bZwzBgcrjEbZVRmONhkg7V0FvT3WValNI8cRlXU+lJ6r19xZ4a2WcGG5O1/7AgWr96D0XD3mrSvWDI9w2lekayuaVG1lnS/2F9ZGusI1tCfifggEbcMbUPFQ6Fs7EPzsRK61YSJTRU9V1aOi1oskk4T0FDaO7u7sPtwAzFqPCo9PCQZ9+hScrPTg3X0nsWrWaHz82PVYNWs0CIxbFplOGxQVzPmq6vwGga1oc9anvIbTb/X9qut8yHTakJ5iRabThnO1Pm4/SWqUjvcrKvwqxQNrivGT//oUNy35FD/5r0/xwJpieAOUmd9jW/4Bj19l2rxRUobLdcG5LSUcmKqfc1sTLdpjXU1448MbULB0x2EsmpiFTKcNO0rOYtn0bO1Gh11dl6daDW36mgTtbR12Bl3hGtoTcT8EgrahRIg3hlm64zAK813MWrpyZg7cDQFMWPYFxi/6BBOWfYE6r4JHbr4SC7eXYOqK3Vi4vQQ1ngDSHayLv6zKA4VSFKzeg5uWfIqC1XtwWXer5nEJt/kUFS9PHcmMu2hiFjy+AHM+nveCN+eifBecdjPT5rSbUaTrV5jvgt0qM3MZDPLtAAAgAElEQVRJtsoomsH2e3nqSLi9AUwp+hLXL96JKUVfosHPX3v0Qma8tmjBqgG1bWES4cBU/d+trYkW7bGuJvy2C0EwzevFPx/Utlf+3//qUrw+Pozf/uxqbJ6bh3q/gpPnPchIsSI9JYlxEUVLW0qkNNyucA3tibgfAkHb4KWyVri9SHdYDKm2d732BfMt+Pj5eoO0eDTxK4kQpp9fgcHjUrBqDzbMGWMITL03tz8zZ554WIXbC49PYY6t9ylY+dkRpm3ZJ//E03cMY9q62UyYtvLvzFxmr96LrfPHMv0yulkN/cIeF/3aU13PGhq8tmjBqqY2bhW3dxX09lhXE9/4INBSq57Yuh8vThlhSPEa1TcVBAT+0J5i3x42KDQYLBP5B4iWtpRIabhd4RraE3E/BIK2QQi4abAqpegVEaR4qqre8C04WvzEAJ341fJ8F979uozpF1CMaatlVR5Qyma7FOa7EFDZSrdOu5lb2fyZd7/TttyB4Dth4c+HGyrTvrbjn4wS6sePXW9Ivy3cecSQKfPmA2MwblAa5vx4kBac++G3pw3Vb5dMHoE0h8UgMpZskZk2q4lwRdoyHPGhwt0e62rCGx8AQYNf1QKCwnuD4Yd3VN9UPHHrUKbuQGG+Cys+PYpdRyuZIJn2tg47g65wDe2JuB8CQduIJi3++zuvYfrxvgVHky+3yAQb5+RBpRQSITDJwLc/XGDOp0T51i8TwgR+WkwE/7P3B4P3Ys6PBhu8F3ohswq3Fz3sZiYl1xdQsetoJdPPLBNNxiEya8dmZkW8JEKRP7Y/E5y7bHo2MrpZmLm8/vlR/L+J1zLpxhmOoCaHfo26LEVlhMwuJtulvQPv22NdTfhU2+p6L6o9ftR4Anjoza+R7rDiP+6+RktvipY+tXFOHkpO12Bb8Uk8PzELigrxcrq0Eam2nYhItW0VMXlWL3i8OHG+gZOdkoTuNlaeQP9iWzs7F16/ypSTL5rhwp6j5zB+SIb2wt+69wTuGz+QSY29qrcDlW6/wSux/ZsyFH1Wqo0b9jhMj0hvXTnDBXuSCUfK65jznav1GYTH+jiTsP9kjdavb48kAAQnz3saj70sRZNxiBx3y7yx6J3aGMD5Q7UHUzj9Ns3Nw9QVuzs9066i1ot/f2e/oc7MHyZkdXRNs/hItSWEyAD2AjhFKb2DEDIQwRoxaQCKAcyglLYqDcHrV3Gu1ocVfwvu32U6behuM2nWZprDwnXh/VDtwcLtJVgyeQTO1/kwa9WeTn9ABAKBIF5wNyjY+f1ZbJiTB0opCCF47+sy/MKVie4RiRO8b8FOmxknquoZT0UPuxmuAWlsxdnp2aj3KUya6coZLqQ5zEyKqsMqMYYHEFzHCdg02FS7GZVuNtW2KN+FnQfKGS/Hh9+exo1X92L6vTx1JFKTzS1K0/UprGKqP8pWkaLSuPC6qqrK1bxSVbX5gzuIWG+7/CuA7wF0C/28CMBLlNK3CCGFAO4HsLw1J1QosOJvRzSLzm4xQZYkLNxegnSHFS9OGYGt88eiss6Hwp1HsO9kNTKdNu1heWzLP7Dw7uGGlCFR5VYgEFzKWEwyvv3hAq7s3U37tvztDxcwfWx/QwVWPTVeP2a+8RXzQv7rr6/XvA9AKH30za+N6++6Yrw0ZSTO1/u0cXskWzDvRwMYwbOte09AlggGZzi0bRx/gBqqvM4LBbr+9KW/aXNZNWu0od+vNn2Dl6aMZLZJwmm1eo+GWZZwqqqeqe3CDRCVW5ZQGgioKHd74VdUmCO2YtpLLFKh/Bo1m+eNbdP52oOYGR+EkEwAtwP4A4BfE0IIgJsATAt1WQPgWbTS+JAIDBbdS1NG4I/3uVDrCeC+0AcgbOmt2XUMBeMHwmaRMapvKvadrDaIzYhUTIFAcKnjtJnxyM1XMtsfqwtG4+wFL7OdsnZ2LrwBldl2Kcp3Id1hZV7G0dJH9etvusMalHWPCOhcNSsHd4zMZL0m+S4kWSTc/dquZj0V+kDXAT2TDf144748dSRWFYzWCs6Fxw2oqpbdkum04c05Y7gBooqqatsx0bzqgYCKA2drDffZH6DMfb4YjzylxrTpYBBv54VdxNLz8TKAJwCkhH5OA1BNKQ0napcB6NPakyoqxZPb9mOqKxN3Z2dCpUFLlAB4YE2xwdJbNWs0nti6HxVuL565YxgWbi8xiM0kQipmU9K2Ql5dIBBcLFUev8E7cPK8x5BCe7zSmFY7j5NWq9JgjRZ93IF+/X3k5iF49ePDjAfCapZRsJpNZX1wfTE2zc3DhogA1nqvnzuG1x8w9NN7Kh65eYjBM/OrTd9g2fRsZstm694TyBuczvSbvvLvePuhcYYA1oXbS5jreOkvBw1xFuVuL5aGqrJrHhe3T1NuDY9xMR75eJQciInxQQi5A0A5pbSYEHJDG46fC2AuAPTr14/5XUClmOrKxA1X98K0lY2BPWtm8y3g83U+LeXqsm5JWDkjB2YTYaziwnwXnDZz2y42BjQVuQxAyIl3Mk09rwJBPNHUs8oTkkpNNlYKj5ZWO1DnbbBbJTx885UG70CqzcT0u7KX3eDNLuR4UsqqgqJbkQGnqwtGG7w1hfku+BXKvB+WTB6BVbNyULC6cZ2M5g0hAONxWTQxC92STIbr9fgUZmvnfxaMb2GcBTX0i/b+aqtHPh4lB2KlcDoewF2EkFIEA0xvAvAKgFRCSPivmAngFO9gSukKSmkOpTQnPT2d+Z1JIvh5dqZBivZEJV9StzpUkjnTaUNPhwW9ulvxwocH8Mwdw7Bpbh6euWMYlu44hCoPK/oSTzQlbSvkxDufpp5XgSCeaOpZ5UlypySZDW3RZMmtJgnvPDQeXzx5I955aDxkSeJKhltDaavhfgAxxCfMX1+MR24eYhjDr7BS5SfPewzemvnri3HO7WPaHtvyD5hkiVn3w3EbkfC8IU9u2w+bhTU+wmJpkdgsJm6chaLb6aCceIxo76+2eioig4LD97mzv5DGxPNBKf0tgN8CQMjz8RtK6XRCyBYAkxA0SO4D8F5rz53hsOKHmgaDlfjBt6dRlO/SdP7D+dkvfHhQs0BNEoHHpxhEyQDg93fGb8xHc9K2LbGYxdaMQCBoCt63ZZtZMgiP8YS9Fk3MgkTAbBHwxMjKqjzwB1QQQhBQKYiighAYBLtW/u0oV6BsxadHmPNF88Lw4vpkQjA43QGJAGkOK5JMxmvrn2b0hpRVeWCzSFg1a7SWZZPZwwZZBjM/k0S4x8oETMCuzOm3dMdhw/ursz0V7U1ni4w9CeAtQshzAPYBeL21J6ChGA+9sNiE7D54JbSPlma3ID3FCptZwu/uHIbyWi/W7DqGP0zIgkXiC9rEc8xHc/t3zV2PqPQqEAiag5dCSyk1CI8t++SfWHDTEIMY2R8mZDHnM5v4GSHBrZPGLZHVBaMxgyPY1d0mM9Vvk0ySQRSMgr/+6cMqw22zVn3FGDOHz1xgrqO81ms43y3DMlCtq5y7YoYLxT9UM2nJDb4A99hzdT7MW9doVBTNcOGWYRnMF+AKtxe9U5PaLU03Htf8mBeWo5TupJTeEfr/o5TSXErpFZTSyZTSVpfsq6jz4S//dxrLIwoCPXLzEDy+dT8+KinHvHXFmFT4Jab/8e+oaVAwYdkuLNxegkd/MhRpdku7F9yJBU3NuSXXc6ltzagqZSpuqm0sziQQXGpIEkF6ihV9nMlIT7HCJBP88sYrYAmlkFpkCb+88QqkOyxMEblHfzrUsIaaJIKXpoxg1qZl07Pxh/dLDFsn+q2Oh978Gh6fisNn3ThzoQGHz7ohy2DW/UynDf16BD3ckW2LJ2WhT2oS01aY78KG3aXMtsurOw7hJ9f0Zq6jp8OCV+8dxRz7259dbaicO3ddMUb2S8OR8uD8jpS7kZpsNszlqduu1gyP8LHz1hXj6duHGdbsbtb2izuMxzW/sz0fF01ApXh2+wF8+vj1mtUJ8LcezDLBe78cj5QkE8xy0NpLRPnt5ubc3PVcSpVe49HiFwgSFX9ARYNfZb71L5k8AhIxyoPrP1/+gAqzSWJEwZzJZsOWd7Stk4BKdWJkOejVzcLItQcUihc+PMh4L1748CB+dydbMC7NYcaPh/YyBINKhBUts1uD3p7INgL+dopKqUHc7J2vTzHjev18MTK9vLrTZsbhCne7rVvxuOYnvPFhkghuGZYBX4DigqcBlXU+XHVZCtf15lcoztQ04JcbSrDw7uGwmmVkhCrbJpqgWFNzbu564jHtqqOIZvELETmBoHn0sWGEELz++VFDvZJn7xoORVG1uI2wdzHyWJUCCzbsY9adVbNGG9aiaHVhImMjyqo8mLNuryGdd9Ws0ahwezFvXTFzrEWWcHXvblApRe9UG1SVL7q1Zf5Y+BQVyZDhU1T4FaplxDQ153DAaeS20L7jlbhv3ABYTBJkiaCH3cIUkIs8Vm9OnPe0fN1qSfxePK75Md92aW9SbMH0rYLVezCp8Ess3F6CqnofVhWMZtxYy/Nd+Lq0Eqk2sxaA5PEpl6QLPhG3mtpKPFr8AkEiEPYaTlj2BcYv+gQTln2BCrcXD914BbM1cd+4gaCgmFL0Ja5fvBNTir5E6fk6HDzDHusNGL/1L91xGMunZzNrUd8ewW0R/fqt6FJUy6o8SE02N3u+16aNgsUsYdrK3bhh8U5MW7kbPo4cerrDinNuH3NtNZ4A0nWVZJfuOGyY38tTR6IhoOCeFbtx/eKduGfFbgxM7waLWULB6j24acmnKFi9Bx6/giWT2a2nRROzQAHmXtV7W7Zu8f5GB8/WGt5r8bjmJ7zno9ajGtK3FmzYh6J8FxP883HJaVx3ZS8cqXAj02lDvU+BRAjO1DSAUpoQ2y3tRSJuNbWVeLT4BYJEgOc1nLeu2CCH/uS2/dgwZwzjDamq82PlZ0eYtgpO8GaF2xvcxonoZ7ea8PmhHwzrd/+eKcz8Mp02pDksKJrhYgTFenW3MoJidmujCmp4zqXn6vkiY5w0Xb13pcLtRa9uVsbLIUvA5MLdhmNfnDxCtwV0wFB1d82uY5h93SCm7UxNA99DokvnbalnNx7X/IQ3PgIqXzY2ySIzojKLJmYh2SJhW/FJLJ6UhZ4OCwCKH6obMKnwy0suFiARt5raQjyK6wgEiUA0ryFPDv2CJ8DIki/Pd+GhG6/QtlnC3gG9VHlhvisYUxFx7Ftzx2DIZd2Z9bswJGkemcr62rRRqNWNu/6BXJyt8RmEzMYNSsPm4jJtzmHvRaQYGU9krKzKKM2+qmA0ymu9TMbK+vvHcI/t1c2qKZWG30O9U5Pw6OZvmNgQv6riN1si7t/0bMP8lk3PhkVm302t8ezG25qf8MaHPs0WCFqIJyrrGWvwyW37sWluHn77s6shh4wLb4Bq0b4iFqBrEo8Wv8DIgKfeb9Nxpc/f3s4zEYSJ5jXkyaHzxMP0HpJfbfrG4AlYuuOQwRNQVtVgSOftZjPhOZ1UeYNfxYIt+5gxAgq4c1ldkMsYHxVuL9IcFuZ84evTX++Feh8jm17XEMCCjbpxVco9VlHBfQ9tnjcWAUWFSZYgE2Bi4ZfsnN/82lDk7rWPD+PZu4a36G+UCJ7dhDc+ksySwUJcXTAaZy40YNPcPFR7/Fo1W79KceOLn+KT31wPsyzBbiWwRxSYS+RYACEaFp14s/gFgkSA5zUsmuGCRSaMwNYVGfYWeUjKqjywmtgww4paH9JTrIwnYHXBaDx8U6OyaNgb0hIJ8mjF68wyW0Jj0cQsQKf+sf2bHwzvkqIZLjz9zndaSQ4A2DQ3zzCGxxcwCJQtmpgFjy/A9AtnxSSFNFPMsgSPPxA1eyYycBYAfn8nZQTKnDZzwnp2E9748AYoutlMWH//GKiUotLtQ4NfxVNvf8s8BGt2HYM/oGoBN+4GP2595XPt9y/++SAq3N6EsBj1iHRSgUDQEVh1qbE97Gacc7MCW4X5RpGsaMJe3W1mPLyxcStm8aQsWM3sGMkWGVv2nGC8DTKRDN6QWo+xOJxK+d4LRaXMsX87eBbdbZnMls0r94xEis3EzMVikpA7IJUxPnjZOOfcPmz86rghluPe3P6Ge6DSYHBpeNyNc/K4c+bdP0WlzLErZ+ZgSLojIT27pDNL6raFnJwcunfvXu3n6nov6n0KDp5x45n3vtMq1er/kG8+MAaffH8GQy7rjjW7jmHm2AHIf/0r7fcL7x6OjBQreqcmIdWWGH+8MBW1Xu2BDJPptIktpNbRIX9w/fMapq3bDAKWS3TbJSbPKm9d+euvr9dUQcNkOm1YOzsXM9/4ijEqBqQlY3LRbsaL8MpfDxmMlGXTs3H6QoP20v66tBKTR/fDyfMezQgY1jsFZ2oacL7Or7Vd1duB0nP1eHxro7dh/QO5qPUEGK/J8unZSEux4OBpt3bs4Aw7pq38u+E69MGlmU4bts4fi+9P12rHDsqwwx9QmfkNzrCjxhMwFLRLMkuYFRHjUpTvwis72HuwZd5YBFSVuY7Fk7LQPy0ZU1pw/+J8nY/6rCa+58OvwhtQNWGacCptJGVVHtR5A3ANSMO6L4/jo5Jy3H/dIOb3/Xok4zdb/oEKtzfhvAYinVQgELQ3vHUl2rbGBY/fIOz12rRRzDdyVVUNgmLhqrGRHoglk0cAAONd2TwvzyButnx6tkHEq9YTwLovjzNekw+/PY0br+7FHLs2StVYXjDtObePOXZ1wWjDXArzXeim85pQSuGwsm1Ws4SKWlZVVKV8YbRX7hnZ7P1L5HU+4Y0PhQKl5+oBNFat5bmwTlZ5sDAUsLTraKUWYBT+/ekLHs21djGBp7zYCwAdGo+RyEFHAoEgPuGtK9G2Narr/UZhL5PMrKG8VFte1djHtvzDEKyqUhjEzV79+DDuze1vEBnbdbSSCS5dNWu0odLt8Upjqm20YFr9sSfPezTDI9w2f30x1s7OZQTKXv34MJ649WqDJ0XvXan3KVxhNJMsNXv/EnmdT3iRMUoplu44jL4hTf9txSexaCKrp79oYhYKdx5BWZUHaXYLlk/Pxrbik9rvF0/KgiPJhFF9UwG03ZqMJvhSWlnXrAjMxRCPAjICgSCx4a0rVhMxCGwV5bvQt4eNaSvMd8FpMzd7vmjprXoPhEop7hs30CBuNqSXgzlf3x42Q70X3hg8MbIlk0cYroN3bDT5d1kihvkFFMXQb2AodTc8Rv+0ZMM9Lcx3IUMnbtbV1vmE93xYTDIq3F7YrSZU1fnx29uuhiwTrJmdC3dDAE67Bf+6cR/2naxGptOGy7snQaEUC24agiduvQpHKurwwofBYNNn7hiGeeuK22xNRhN80Vvx7Z3SK9JJm0ZkAgkErSfautK7G2VSRa1mgt9u+9aQQvvchGtBQJhj9efzB/hS6noPhEQIVw5909w8RuzLIhNs2XOU2XaRiFGOgSdu9vrnR/H7O69hxM14Qa3R5N/9CjXMb3VBLnMdmU4bkq2y4Z6qKntPMxxWSBJhMlt49y+R17KENz66W2UUhqR3kywyZrzBlkh+88tjmuHxyj0j4fYFsG3vSdwxMhMf7D+NJX89rJ0r1Wa+KGuypaI8HbFPJ9JJ+YhMIIGg7fDWFUkiuDzVpv18qqoeH5WUG+IR/u1nCvJf/7vhcxd5vqq6BizPdxlEwXp1szCpsSrli0kqKsW0PzaOsbpgNG66+jIUrG4M8lw1KwdLJo/AYxFiX2G59shYk0UTs6BSqgWiZjqDQmZF+S6tim3Yu6JPyV2e78KKT48Y5mfSpfiunJmDnnarYe3R39Om1q2uss7HxPgghCQB+BsAa2jMrZTS3xNCBgJ4C0AagGIAMyilrarxW1Hng8enIDXZDAnA6oJcSARaoFHe4HRMyxuAilovntv+PSrcXqyaNRoFq/dg1azRmvGR6bRpkcNttSZbKsqTyPt0iYYoLCcQdCyE41nIdNpw7Fxds5+7ep+KV3ccYmM5dhzC7+68hmmTo4xBCFtsjhePUbB6r0Gwq8GnGOTfw6mxkccu2LAPW+aPZTwpKVaTwbviC6jYdbSSuS+ZThssUvPVfnlcCutWrDwfXgA3UUrdhBAzgM8JIR8A+DWAlyilbxFCCgHcD2B5a04cUCm27D2JBTdfAbc3gNtf/RwAMKpvKubfMFjzOmzecxLzbxiMVJsZSWYZ6Q6rpnQatip7d7dd1LfhaFLeVpNksH4TdZ8u0fAFFKQ7rMwiU7jzSMJGiAsEsaQlW5YygdGzMD0bv3vv/5h+ZVUeqKrKbCUEVMr1mvz77cMwqKddqwYLGMdYMnkECKFMbZeeDgvXQ9I71Qq3Nyj4ZZElZHSz4F9/ciUjkV6Y78Iz735nOJaGvC7hjJWhlzlQVc+Kh338/VmsLhjNpN9mOpNgko3vk5bc00shgzEmxgcNiom4Qz+aQ/8ogJsATAu1rwHwLFppfCSZJNz/owGgtHEvLt1hxW/+ZSijNrd8ejZe/fgwPiop14JM7RYZXzx5Y7vtnUXbIwXQZfbpEg2bRcYTtw415NDbLMLzJBA0RUu3LM0mCUk6obDUZDMq3F7mfLcMy8C5Oh/zwt80ly+wJUuE2TrZOn+sYYwkswSJELamzPRsg+DZLcMyUF0fYFJjl03PRvdkM3O+NIeFO+eqer8h1XbG2P7M/JbnuwCw6cHL813oDlZQbO3sXHgDarP39FLIYIxZzAchREZwa+UKAP8N4AiAakpp2IQsA9CntedVAciSjD+8X4LHbhmKxZOy0OBXDcFJD775NVbNGo37rxuEao8fq744hud+fi36OG1ND9BKosVedBVXWaIRUKlmeADBZ+Hxrfvx9kPjOnlmAkF8U1nnw0t/YfUnXvrLQfxhQhazngVUil9u2Me8KG8ZloGiGS7G0Hj69mFafAYQ/Cz+5f9Oc2M+1u06xvQ7c6GBqacCBF/GL04eYVjn187ORcnpWu18T912NZ7/4HtDnZTZ1w1iUmPPu31YPCmL+aLCO/ac26cVi9PG5dSyeTCUfqtP8dVvC81ZuxdvPzSOCc6NJpvutJkNQaiJ+kU2ZsYHpVQBMJIQkgrgHQBXtfRYQshcAHMBoF+/fszv/IoKiQAflZTj4ZuGoK/TBoVGF8KZumK3FlyUaOqugtbjD6jcZ8EfUDtszKaeV4EgnmjqWVVV1VBPZdHELKgq+9nhfcY+KinHwruHMx5f3lbCs9sP4NPHr2cyTBRFQdFnpUw/b5TPsf61G17nI40FX0DhXkemM4mpOMsTLeMdy6spEy2xQE9qMl8Es95rDM7Vy6Y7bWYcrnB3meD5mGe7UEqrCSGfABgLIJUQYgp5PzIBnIpyzAoAK4CgBHDk72RC4KMUtwzLQDebGRQEAMWqWaOxdMdhTTgs02ljKtg+uW0/Ns8b21GXKYgTOsN92dTzKhDEE009qwoFN71Vv25G+4yZTRKUCDvFZuH3czcosJhoKHhThS9Ug0uf3jrvRwMwKaefFuS5de8JSIQwMR/bik8aBM/++uvr8eS2YsN1bJyT16xoGe/YEy0UKAtm6bD3OyXJzD22otZr8DC9MCkLfkVFQKUgiorz9V0rCDVW2S7pAPwhw8MG4KcAFgH4BMAkBDNe7gPwXuvPDSRbJTx+61X4odpj2NsPa3iEi8eFCQcSCbo20YKARcCvQNA0NEp6q37d5H3G1s7Oxdkar+Fzp68Bs2HOGFzwBAzxExvmjGFSXof2dqBnipXpF66dElkRd3m+C2a5UYU1GD/C94QrKjWk2g7omcwca5aJ4dilOw4bUm0L812wmtnEguXTs2Exsam2SWZiqH67alYOztf5mbm8PHUkKtw+FETUhVl3P9/jkqhBqLHyfPQGsCYU9yEB2Ewp3U4IKQHwFiHkOQD7ALze2hNTCoAClW6fZniEM13MsoSl946CWSb43XtsWeRo336FIFXXQgiwCQRtoymPRnPiVxQUM5ft4sY26LdiHtTJl4djJdjUWNXQb36UOIvX7h3FpMFeqOeX3NCnAj+5bT+2zR+LDXPyoNKgOJlEjHLyFW4vnMlmZqvIF1Dw+aFypu3jktP4eXYmNs8bC7+iwixLMEnEUJ3XJEtaFk94Lr/a9I3h2krP8T0uNouckHEgscp22Q9gFKf9KIBc4xEtx2aR0OBXkZ5i1QwPfaZLUb4Lv7zxCiYIifftVwhSdU2EAJtA0HqcNjP3G767IcB4L3jiV6eq6qPGWvVxJmttxyvruP0AMFsnnzx2fYvjLLonWzAjIn5iVcFoQ/ArL6023WFFhdvHXO9bc8cYPBWLJmYBAKatbKw4u3HOGOQM7Mm0rS4YjdPVXsxZF/E+mZGDx2+9ivForL9/TIuujedxieZhSoR3VsIrnHp8KlQKbR9u/g2DDfuU80KW9IuTR6B39ySYJMLV9LgUhF0EAoGgJVSFZNIjv6V7fIqh0BpvjWxprJVZlrj99LESpij9eHEWpTqPRsGqPdgwZ4zhOvRptbwicofP1mHjV8cNYmTPTbjW4OmZv5LN5OEJns1ZtxdvPzgOm+bmIaBSmCQS9R7or63C7UW6w9IiD1MivLMS3vhQVIo6r4KlOw5j0cQsWE0S14qsbQjAbpFhkggUSnGuzmuQuW2tsIvYohEIOpcBT73f6mNKn7+9A2bS9fAFFIMA2Ka5eS1aI6PFWulTRdPtFq5UeUBtrJ+S6bTBJAPLpmfjoVAF3HA/my7Ooijfhac5QmEBhTKelFF9Uw3jRitAt/Dnw5l+K2fmINVqQkWdDwGVAgEFVpNkEDOMltniV9SgEdZEWm3RDBfMOmn2ZdOzQQha5GHivbPi7X2V8MZHkjmoklfh9uK9fafw8M1DuFak0x586Keu2M08RJHuqdZkRogtGoFA0JUxm4zfyKMVVeOtkVYTKwpmNUk4UVVv3LLJcDBF1ZItEp7Yup95kZ+94MUH3/7AxHJ8cbgc112ZwY5hlpCewm6n8zwpFcp6T7MAACAASURBVG4vethZkTFZMl5vhduLng6LTkDNhIMVdYw2SWG+C8/9/BrMW99oHK2/f0xUr06k8BgvrZZSitc/O8Jc79a9J3Df+EHMdbT0nRWP7yuSaBkfOTk5dO/evdrP5TUNePrdb3HfuIHwBVRs/Oo47r9uECPDW5jvAgG04kBhMp02xj3Vmj9QRa1Xe4CinU+QUHTIJ1D/vIZpyzd2QfvQBTwfMXlWz9d5cfJ8Pc7X+bUX7+WpVgRUMPETvDUy2vr44uQRuODxM6mxz0/MgqKC0bM4XeOBN0AhEUClgDPZhENn3Uw2Y2TmTOQYm+fl4eAZd6PMeQ8bTBJQeq5R+rxvDxu27DmB7AFp2ly+Lq3EXSMzmSJyhfku7D12Dr2ddq3flb0cmPG6cdyFdw9n0nS3zBuLgKoaMjBNkoTJRV8yx7794DgEVKoFpvZMNuNQRZ0h3uaqXikwmSTt2Ja+szrxfRX1WU14z4dfUfFRSTkqan1YPDkLFbU+JFtkxlJVVBXJlmA9l8ibr3dPtSYz4lLQ3hcIBJcu/oAaLDsfIRm+ZPIIDE63N7tGRlsf01OsjLDXkskjcL7Oh1kRAZgb54xBtSdgUD3de+w84w3x+o3CY+kOK87XGeXQq+pYefWiGS7clnU5FoSUWcOBpD1TWC9HN5sJV1+eynyZjZbyqg8QVSnFCx+yCrEvfHgQT912leHYOp/CBMkW5rtwZbqd8QhlOKyM4RGG52HSE4/vq4Q3PsJup30nq3Gkog6P3DwED4b2BcOErdJHbh7CWKaZzmBVRMC4H9ZckblE096Pt/0+gUAQ3ygUhhTQx7b8A28/OA5Et3To1xfelk2m04YTlfWG8+lTSr0Byk2/3TBnDL4/XaudL8VmMozBCxrlBX7OW2dM0w0Lj0W+I1bNGm04NlrKKwUYwTOK4LZNZKxJS4Nk568vxuZ5Y3F5atPlPyrrfFzvT1sDgGOJ0URKMNLsFqyckYNMpw07Ss5yg4bCVumAnnZkOoN/zLDVLZNG19WEZV9g/KJPMGHZFzh4thaqfqNQP+7MHOZ88Spe1ZbrEwgElzYqR2Qs3WHVXPiRa0lpZR3T5m4IGNbHonwXlu44zJyP5zGQiFEULN1hRY0ngIXbSzB1xW4s3F6CBr+CJZNHMGPw1v9ki9wiT0VZVVB4rLljwymvkeMW5ruQ5rAw80tzNL6bwv1WzsxB/7Rkw7G8+xJQmi8B0VKPRjy+rxLe8yFJJPhHv3s4BqQl42yNN2raklUmWHj3cAxKt4NS4D//twR/mJDVphTbRBKvEinEAoGgtUiEcD0L8ziptnovwsw3vsL7j4xnUkodSbIhvTXTaTNIpIfbm/NozFq1By9NGclsawDBonYTXX21Nt51RPNAmGR2LrxjK9xeqJQywaAev6Jpd4TnV7BqD95bMM6wdQKAaTNJBOkpFoNMvElu3jfQUo9GPL6vEt7zASD4h1+9B+W1Xiz64IDBGl48KQs97GbU+xUMSk/GH94vQW2DH0/ddjXS7JY274eFxav6OJORnmKNS8MDiM/9PoFAEN/IBKECbI1raf80vmc5NdnMtI0blIYT5xswdcVuXL94J6au2I2T5xuwYc4Y5nyrC0YHY/QiPAayTLB8enazHo2yKg96pliZY61mCQ/ffCXTZrPIKJph9FQ47WaDJ7zBrxiOXVUwmum3PN8Fi4mgYPUe3LTkUxSs3gO71cSdX61HwZSiL/HjxTsxpehLHK5w40RVPdOmUNUw54dvvhLpLfBKtMajEW/vq4T3fACNQjXVHj8q3F48/8EBrJ2dC0qDLrwzNQ149k8lqHB7sWluHj4qKUfJ6Vq8/dA4SBKJy/2w9qSrX59AIGh/JEkySIETjtx4ptOGlCTW+Jh7/WDMWvUV4wmYz5FN55Wnn77y7yjKdzGeBUsUIa6zNQ3M+U5Uegznm7++GC9OHsH062Yz4bntJUybRIgW+NrUsa/uOITZ1w1i2ipq+R53vYQ7z0t04LTbEFfy4PrioGfa3PQaHY8ejZbSJYyPDIc1tG92SJPCrfMG8MOFBqTazKhpCAAI7aOF9vTKqhrLqnf14mNd/foEAkH7k2a34NGfDmXWjU1z87hy48kWCatmjdYyLngF2cqqPDDL7EvRHiUeIzXZjENnG9Nlh12eYhh3+fRs7DxQjluv7Q1ZIuhht6AhYMyA4WXZbJwzBveNG8icb81sfhYLASv1DgD/9rNhzPlWF4zGkskjmKyY5dOz8bv3/o97bZFbLNHEyFrqmU7U8hFdwvgwmSRc1SsFv7/zGlR7/Hh7/lhUuH2GioVrdh2DTAhG9U1FhdurffNPZOuxJXT16xMIBO0Pb92goAZvyJpdx/D07cOYVNZNc/MMnoBbhmXgQihoNNxv2fRs3DIsg1FRvWVYBqo9fkNq7N8OnmXGtVlkXH9VBlPp9s0H+MJe+pL1p6obDNcRzXvBiw2RJcKkt9otMl74/ChbDM+vcmNcutvMeHhjY4rvuvtzueOaOSmzXYmEFxkDgEAg+Ef2K8E6LyaZ4J6QkmmY8IO5YXcp8ganI6ObFVdf1k28gAVhhMjYJYIQGeMT7VmNhCdqVTTDhVf+eogxIP60YDzqvIEWiYJtnJOHf5Y3ejmuyHDg3pX89ftoRZ3W78peDk2xOswtwzLwrzdfyQiF/fe0bCRbZaaY26qC0fAFVEYsbXXBaDT4VYOwV5JZYnRIeNd7y7AMPHHrVTh53hNxHXa4vYrhXv1pX5lB3Oz2EX0Y6fjFk7Iw7PIU1HlVTXgswxGM00gwyYSuKzIWCKg4cLaWeWDefIBfJZBSoOizUkzLGwC7RY73P5pAIBDEFTxviKqqzIsYADw+xSCwdcHj567LikoZL0e0Kq8qZfvx1vmPSsrx7F3XMF4JCop0h5nJvEm3W3DW7cXqglxNRbVbkgyPz8seSym6JZkMIl4VtT5m3IpaH7wBVpCNJ5tOQPHjob0M21Y2s8Tcq73HzsORZDZIuCdbZG5F4UR8l8XE+CCE9AWwFkAvABTACkrpK4SQHgA2ARgAoBTAFEppVWvOXe72YumOQ3j13pHo6UiCSoMPF8+N5VdUZDptKK/x4vLUJJyqqgchBDIJBld1tBUphL4EAkFXI1oNGL3A1qpZo7nr8hld0OiZmgZuv9JzrECZX6FRtklUg5jkprl56ONM1toqar3YUXIGNw3rHXxnEIKAQvHqx4cx0dUXyZDhU1S8+vFhLLx7OAZnOKBSCokQfFxyGosnZwUNr1BbvdePB9YaU5DfWzAOfoUioFIQRYVMYKi6/uS2/VhdkMvcq7/++npuwK4+WDWRJRNitakUAPAYpXQYgDwAvySEDAPwFIAdlNIhAHaEfm4lFA/fNAQNfhX3rgymda3ddQzLdSIwy/Nd+OJwOV6bNgoqpZi6YjfGL/oEU4q+xD8r6vDv7+zvUOEtIfQlEAgSHd46drbGi7Wzcw0pufoU0MEZdu66nJpsYtJMk8wSijgiXnohrhWfHuGe74P9PzD9yqo8UHThBbJE4RrYE9NW7sYNi3di2srdUEFx37iBzFweuvEKnHP7mH7jr8xAg19l2gJqUAgtknSHFaervZhS9CWuD6XVVrh9hn5lVR6YQhVsw9cRLWBXn9KcyJIJMfF8UEpPAzgd+v9aQsj3APoAuBvADaFuawDsBPBk684NnHP7mFSlos9KAQBvzc2DLxCMA+lmM2H8kAycrWnAU29/a7A8n7ljWIdakULoSyAQJDrR1rG3HxpnCGgHwLT5FRWv7jhkSFu9N7c/c74FG/Zhw5wxhtRYffDmrqOVuG/cAEPl17zB6Uy/TKcNsk4P3uNTDRLuDX5q8EpURdSJCbeVceTaw16JSI8LTxiN1y/TaYNVlph75QvwqwfrU5oTWTIh5jEfhJABAEYB+DuAXiHDBADOILgtwztmLoC5ANCvXz/mdxSsBO6ovqmYf8NgpNqCf6SzNQ144cODePmekbBZZAxKt/MtSpu5Q61IIfR16dDU8yoQxBOtfVZ9AQXpDitjGBTuPAJ/QGW2NQAYvLp+ReXGSvBkzivdPmYb4n8WjOem2j797nfYd7KaOX5Kbn/txR2OqTDptrcDKsW4QWmY8+NBmuFCYWwDjBLz0eTaw+U7wuNGE0bT9yvMdyFdVzTu7AUPN6XZZpaYYxNZMiGmxgchxAFgG4BfUUprSIQ1SimlhBDuHgSldAWAFUAwIjvydxZZCpZNdtqQ7rDiN/8ylPmDLZ6UhWfvGoYfqj14fOt+FOa7DKldYYGyjrQihdDXpUNTz6tAEE+09lm1WWQ8cetQQ5l4m76iKycrZuv8sdxjJZ1XItNpgzfA1jU55/Zh41fHGaMnXLhNf2ytx29IBX7u59cy/RxWGflj+zNpukUzXHjgxwOZtuX5Lsz70QDNmw5Ae9/o1/Jw+Y5wYKpJ4gujmXRpusmc5AeewNuaXcfw3IRru4xkQsxSbQkhZgDbAfyZUvpfobaDAG6glJ4mhPQGsJNSOrSp8+jTwcprGkAIhS8QDOopr/HCG1BgliVNI//3d14DsywhoKoABfxqMEiopsGPeq+CHg4LKmq96OmwYFBPR4f8MXkfxo6KVBaBrW1CpNoKmiSOUnRj9qz6fAFU1Pka67NYJbzz9SktUDMcgHnbtX0QUFWtX4pNxn/v+Ccm5fRr9CxQihc/OsjUXdlWfBLPTbgWvoAKRaWQJQIgeN4pRbuZNNg0hxn1XjWiVoyEU9VenL3g1V7kvbpb0SfVCndDYz9ZAuxWCTWeyDaCSYVfGgyDFyePwAWPn5nf7+68RpNuyHTa8MasHAQUirkRabpFM1ywmQCr2ayNcfZCPbolW5n02749bKip96JX92St35HyGozsn4raiPml2Sz4obYBxyvrtWP7pyWjnzMZVR6/YXtLv97z2jrpHdC5qbYk6OJ4HcD3YcMjxJ8A3Afg+dB/32vtua0m4ESVDw+uL0a6w4onbh2qxXSEXVXVHj8Cigq/QvGrTd8wVne3JBN+s/kfqHB7UTTD1S7XyyNWQl+xNHIEAkHXxecL4GBFHZPu+eacMVqgZqR3QJKAqYW7mX53jOjDeBE2zc0zqIq+PHUkLnj8jAZHMIjUyngH0lPMOHHey8xleb4LPR1mzF3b2PbegnGGfoX5Lri9rFbHuvv5aqYZOiXURROzIBEwcSWl59wYkO5g5peSZEKNJ4CZqxrvwaqC0Yb02xUzXEhJtmr6JOF+J897Gc2RsL4II7SW78KJqnpDqq3VJLWoLd7eAbHadhkPYAaAbwkh34Ta/g1Bo2MzIeR+AMcBTGntid3exsChZ+4Yprn0gMZg0lWzRqO8xmsIEnp8634svHs45t8wGPPWFWPeuuIODQCNhQyuCGwVCDqGtnqr4shj0ioq6nyGoEx/gBraHlxfjLfm5hn7hUSzwm0qNaaZ/mrTN4b00QfXF2PT3DwmKPOLJ2/kjrtxDjtuAyeQlJeiWnqunrslcryy3vD+2DQ3Dz996W9av8+euNEggrZq1ugWBabOXWecS1MBrJFt86Kk2ra0Ld7eAbHKdvkc0d0vN1/MuQNqY0BQOGg0krIqD9zeQNQgoWSLjGTI2s+JHgAqAlsFAkF7ELm2hpEIuOuLogsu5fULqPy6K7yA04DufLy5BA2alvXTj7F0x2Esn56tGUhN1WJRaaOeSKbTBpW2LAi1qXdOW/pdbFu8vQMSXuE0UlAsHDSqt2bLa71RqyLW+xT4FFX7OdEDQEVgq0AgaA94Yo0q5Ve11QeN8vopKv9YXu0UfXZKNOFI/bjR+unHqHB70eBXW1SLhYAw2y68MXhBqNECU/VzaWm/i22Lt3dAwleuSbNZNKGZwp1HsHhSFiPWsmhiFrYVn0QPuxkvTx3J/G7xpCxkpFhQuPNIwqcthQlXsI28zq5wXQKBILak2y0GES+ziXCFvWQZhn6Fun4WzrGF+S5k9rAZzudIkpi2ZKvEHdekGzfJYuxXmO9CX90YhfkuOO1mRlAs1R6sNsu7toLVe3DTkk9RsHoP9x707WEzXG9mD5vhfEUzjNfL6xdtzv3Tkg1re0vb4u0d0CUKyzU0BFDpCUZkJ5kkUAC+gApZIgga0CRktaqgNOiakyQCm1kCIQQeX6dHBLcrItulTYhsF0GH0AExH52W7ZJut4AQgnK3V2vLcFhBKTX0kyQp2E9RYQoVRgsEVG2tDmd1ADC0ybJxDEWhhn6EAOfqG9t6JltAqfF8vDF4bbzzmc2yYT31+ZQ2n483P14/WTbeP15hOUBku3QaSUkm9Em6iEuxt99c4oFYBLYKBIKuj8ViQh+LcW3VC4oB4Pa7PNXG/GwySdy1mtumG8NsjtLP2sLztbSNcz79ehrtnXNR8+P0098/3lxa0xZPJPy2i0AgEAgEgsSiS3g+BAKBIF5pyxZboqbnCgQtJeFiPgghFQhqgvDoCeBcDKcT74j7wdLU/ThHKb21vQds5nnVkwh/r0SYI9C159kZz2qi3M/mENcRW6I+qwlnfDQFIWQvpTSns+cRL4j7wRLv9yPe5wckxhwBMc/2JlHm2RziOuIHEfMhEAgEAoEgpgjjQyAQCAQCQUzpasbHis6eQJwh7gdLvN+PeJ8fkBhzBMQ825tEmWdziOuIE7pUzIdAIBAIBIL4p6t5PgQCgUAgEMQ5wvgQCAQCgUAQU4TxIRAIBAKBIKYI40MgEAgEAkFMSTjj49Zbb6UAxD/xr73/dQjieRX/OuBfhyCeVfGvA/5FJeGMj3PnEkFRViAIIp5XQaIgnlVBLEk440MgEAgEAkFiI4wPgUAgEAgEMcXU2RNoT1SVorLOB19AgcUkI81ugSSRzp6WQHBJID5/AoGgpXQZ40NVKQ6ercWctXtRVuVBptOGlTNzMLRXilgABYIORnz+BAJBa+gy2y6VdT5t4QOAsioP5qzdi8o6XyfPTCDo+ojPn0AgaA1dxvjwBRRt4QtTVuWBL6B00owEgksH8fkTCAStoctsu1hMMjKdNmYBzHTaYDHJnTgrgeDSQHz+BInIgKfeb9Nxpc/f3s4zufToMp6PNLsFK2fmINNpAwBtzznNbunkmQkEXR/x+RMIBK2hy3g+JIlgaK8UvPPQeBFtLxDEGPH5EwgEraHLGB9AcAFMT7F29jQEgksS8fkTtAdt2QoR2yCJR5fZdhEIBAKBQJAYCONDIBAIBAJBTBHGh0AgEAgEgpgSE+ODENKXEPIJIaSEEPJ/hJB/DbU/Swg5RQj5JvTvZ7GYj0AgEAgEgs4jVgGnAQCPUUq/JoSkACgmhPwl9LuXKKUvxmgeAoFAIBAIOpmYGB+U0tMATof+v5YQ8j2APrEYWyAQCAQCQXwR85gPQsgAAKMA/D3UtIAQsp8Q8gYhxBnlmLmEkL2EkL0VFRUxmqlA0DbE8ypIFMSzKugsYmp8EEIcALYB+BWltAbAcgCDAYxE0DOyhHccpXQFpTSHUpqTnp4es/kKBG1BPK+CREE8q4LOImbGByHEjKDh8Sal9G0AoJSepZQqlFIVwEoAubGaj0AgEAgEgs4hVtkuBMDrAL6nlP5XRHvviG4TAHwXi/kIBAKBQCDoPGKV7TIewAwA3xJCvgm1/RuAewkhIwFQAKUA5sVoPgKBQCAQCDqJWGW7fA6AV2Hqf2MxvkAgEAgEgvhBKJwKBAKBQCCIKcL4EAgEAoFAEFNiFfMRM1SVotrjg8enQKEUSSYZJpnA41NgMclIs1sgSbwdIIEgMVBViso6H3yBjnumo40R2U4IgUwASZKYOcRifgKBILHpUsaHqlKUVtbhbE0DHt+6H2VVHmQ6bXhpygj85/8eQIXbi5UzczC0V4pYDAUJiapSHDxbizlr92rPd3s/09HGGJLuwOEKN9O+aGIW1uw6hkd/OhRDe6UAQIfPTyAQJD5datulss6H45X1muEBAGVVHjy6+R+Yf8NglFV5MGftXlTW+Tp5pgJB26is82kvdgAd8kxHG6Pc7TW0P7ltPya6+mpziMX8BAJB4tOlPB++gIJki6wtfGHKqjxItZm1//cFlM6YnkBw0fgCCvf5bs9nOtoYfkWN+tmKnENHz08gECQ+XcrzYTHJqPcpyHTamPZMpw3VHr/2/xaT3BnTEwguGotJ5j7f7flMRxvDLEtRP1vhOcRifgKBIPHpUsZHmt2C/mnJWDwpS1sAwzEfhTuPaPvPaXZLJ89UIGgbaXYLVs7MYZ7v9n6mo42R4bAa2hdNzMK24pPaHGIxP4FAkPgQSmlnz6FV5OTk0L1790b9vT7bxSxLkAB4FYoks4SedqsIfBPw6JCHQv+8tkcmSKJmuzQ3b5El02Ji8qx2FgOeer/Vx5Q+f3vMxrqY8S5Boj6rXSrmAwAkiaCH3QrVFo7Y/zsTdd/Tbu3sKQouUdorU0WSCNJTOvY5jjZGS8bm9Wnu2mORxSMQCOKHLrXtEomIuhfEG5fyM9nctV/K90YguBTpssZHLLICBILWcCk/k81d+6V8bwSCS5Eua3yIqHtBvHEpP5PNXfulfG8EgkuRLhPzoaoU5+q8aPArMBECkyxh/QNjcKyiDkt3HEZ6igVP3z4MvoCCilqvCGYTxJw0uwVrZ+fieGU9ki3BtPD+acktzgT5/+y9eXgc1Znv/32relEvsrUb4x1iTDzEJm6ZOIEwDNuPTEgy+dlAgmViJYO3IUy4eUgyc4eZPJeZuSyXMIGMLduADdjMBezJMiQwSSBAEgLYMkuIwRjwIhtj7ZZavdR27h/dp1RdXdWLpJa65fN5Hj1Sbee8VXVO9VHX9/0euyCzNuBFX1wdE4FmKcWehsHAwLDjG5/Coe5Uf+yKJrG5JQJZSm3nWTJ2zYfIkhEIJieTYvDhJFa7e8Ui3PXMgdRDblUEfo+E6x94RYjZBBNKUjNw20/fymiHhWBv41cubMLNl52DdTvaR92mSyn2dCq7rSWCuKLjH37yVsaUBwumVePHGy4U2S4CwWnApHjt4iRWu3XXm6al+tpH29HRGxdiNsGEMhpRpf3Y5ZFZ5sCj2LLGMq6RlL1uRzt6Ywpe6+jPqItnycyoDaKxWqTECwSTmaIHH0R0ERG1pv9uJKJ5Yx9WcbiJ1ayW6kGfnLVdiNkE48loRJX2Y7ml+UjKGsu4Rlo275tjWZdAIKgcihp8ENE/AfgugL9Lr/IC2DHWQRWLm1jNaqkeU/Ss7ULMJhhPRiOqtB/LLc1HUtZYxjXSsnnfHMu6BAJB5VDsNx9fBvBFAEMAwBj7EEB1voOIaBYR/YaI9hPRn4job9Pr64joV0R0MP27ttgTAJztoO9esSjDUn1OfVBYPgsmlNFYj9uP3d3egbaWyJi06VJaojuV3dYSwe72jjGvSyAQVA5F2asT0auMsQuIaB9jbAkRhQD8gTG2KM9x0wFMZ4ztI6JqAO0A/grAagC9jLE7iOh7AGoZY9/NVZabBTDPdkmqBiQCiACJUs6JGmPwShJ8XkJCMUwxGwBh5yzgjItltaYZ6IwmoeoGvLKEprAfHk/qf4Bi7cft2S7FZL+UMnOmlHVNpAV7Gdm/C3t1G8JevWwZM3v1J4hoM4AaIroRwNcBbM13EGPsBIAT6b8HiehtADMAfAnAJendHgbwPFKvdYpGkggNIb+prG8M+/Gdqxbg1l1vZqjsz51WDY9HEnbOgnHHMBgOdkUd2xyAvO3RybacLxfTnse77eeKuxgmss+K54VAMLYU9dqFMfZ/AOwCsBvAAgD/yBi7v5gyiGgugE8CeAXAtPTABAA+AjCtmLLsWJX16y452xx4AMMq+85oMmtfvl1kwAhKSa42N9r2WMzxldr2JzLuSr1mAkG5UtQ3H+nMlt8yxn6VXg4Q0VzG2OECjw8jNXD5FmNsgGj4PwbGGCMix3dARLQGwBoAmD17tmv5VmW9WzaAphtZ+1q3C9W9YLS4tdd8bW407bGY9lypbX8i467Ua5aPQp+tAsFYU6zg9EkAhmVZT6/LCxF5kRp47GSM/Wd69cm0HoTrQjqdjmWMbWGMNTPGmhsbG13rsCrr3bIBPLKUta91u1DdC0aLW3vN1eZG2x6LOb5S2/5Exl2p1ywfhT5bBYKxpljNh4cxZn7PyBhTiCivTJ1SX3E8COBtxtgPLJt+BuBrAO5I//5pkfEASL2P7YsnoWgGHl+zDAYDDMaw868/hX/5+X50DSq4+bL5mNsQhEciaJoBWQI2t0Sw1uYQKVT3k49yEQq62avXBrzojSt4fM0yxFUDgwkV/TE1r/U6F696JIAxoK0lkuV4WhvwomswmXHuVivzxrAfN182H/MaQ1B1A92DCSg6A2Ms41oVcg1VVUdnNAnNYPBIhKawH17v2H04O8bdEAIDg2GwguMcbd3ieSEQjJ5iBx9dRPRFxtjPAICIvgSgu4DjLgSwCsAfiej19Lq/R2rQ8QQRfQPAEQDXFhkPDIPhcM8QTg4ksO33h/C1z8zDd3cPi0w3rVyCKq+E1u2Z9s73PfsuugYV3P6l8zCvIYSgX0ZDSLgqTjbKSShoGAwxRc+wV9/cEsHRvhjuePrtrLbb1hIxP1TtaJqBd04O4r9eP4bPL56BDTv3oTHsx+1fOg9zG0II+WTUBX2uAtcF06rxs5suxIn+hDkAv3JhE266dD427NyXsf/8xrBrOTw2VdXxTmcU6y2Dn00tEZzbFB6zAYgkkWPcxcQ52rqF/btAMDYU+9plHYC/J6KjRNSBVGbK2nwHMcZ+xxgjxtgixtj56Z9fMMZ6GGOXMcbmM8YuZ4z1FnsCPUMKjvTEcOuuN7E8Mst8eAOpd7Lrd+7Dsb5ElvB0eWQWXuvoR+v2PWh58BUQSDxIJiHlJBTsjCazLNHX7mjHkZ6YY9u1CqTdylrRPNscLPD2vOrBV6AZDH1x1fXcJYmgGzA/wIGUZTsvy7p/ZzSZ9xp2RpPmwIPvsz5H/CPFO7BzCAAAIABJREFUKe5i4hxt3cL+XSAYG4r65oMx9j6AZWnhKBhj0ZJEVQSKpiPok03LZidRmJO1urB3Pj0oJ6Ggqhuu7TMIOadA2q0sWSLX4xhjOc+9UMt2t7it11AznOvSjMJ9hArF7Z5qBcQpEAjKg2Lt1f1EdD2AmwH8DyL6RyL6x9KEVhg+T+rdObdsdhKFOVmrC3vn04NyEgp6Zcm1feYTSLuVpRvM9bh8516oZbtb3NZr6JHIOY4SfDvgdl6eAuIUCATlQbEOp88AOIWUQ6n5ic4Yu2fsQ3PG7sJnGAzH+2NIqDr6YxqqvFL6VUvqne+91y6G1yPhpsdey3gXff+z7+KX+zsnpVlQuQgsy4EiNB8ld43UNAOHe4fQ0Rs3Baez6gKQiBw1H9talyLs98AwUuLPmioPuoYUqLqBgFdG52AS//X6MSxvnoXjfQmzzJl1ATSGfQj7vK4aCMNg6IomoTMDCZVhMKFC0xkCPjlLtDp9qh/H+2wai1XNWHBGfs3H7LoqRBO522Gx7dXtnjppPh75+gUIV3mgasao+0Iud9pxRjic2hAOp2WLa1stdvDxFmPsvDEJaYS4DT764xrW72g3VfBzG0Ko8kiQCdAYA2PAqbiKjr449h3uwfXL5sIj0aT7cC4ngWW5UOCHW8kf6MmkhoPdQxkf7m0tEXysPoRBVYdhGNAMBt1gAAGdA0l86/HXXQfN21uXYkqVF13RJNY+Olzmv113Pra8+D5uuWIB5jeGs6zMDYPhnZODuO/Zdx0F2kOKjjOmVCHokxFNarjhoVfNfjW7PoiuwSQawj7MrQtlWMMf748hqTFIBBgM8HkI//zU/pyD/JG2V7d7al0f8Mk4OZAck77ABb72e8cdk8cZMfiwIQYfZYtrWy2217xERJ8YZTBjSs+QgqTGzP+4rKK7IUXHa8dO4Y/HB3Ddlpfxl/f9Dmsfbcfm3x7GygdeARFNOuFYOQksy4VyEQp2x5Qswem6He3oiStorPZj2tTUK4LrH3gFb58YNAcefN/1aaE0X169bQ9U3TAHHnz9tx5/Hcsjs3DjI3vRF1ezzp2LVd0E2qfiKlrSotUbHno1o1997aFXcSquYvW2PRli0p4hBV/d+gou/8ELuPSeF3D5D17A9VtfyYjXqR2OtL263VPret3AmPUFJ7FwLkGwQCDITbGpthcBWE1EhwAkkRrVsHwTy5USRdMhkbM7pEQwhaXFiPkqmXISWAoyKUSUye+fm/jTLpR2K5Mf73TfuYA0Vx25hKZ8u7X/uLW7fMLuUrbXsSzb7VpMxmeIQDAeFPvNx+cAzAdwJYAvALg6/XvC8HlkGAyOQjODpUR0xYr5KplyElgKMilElMnvn1ubtQul3crkxzvddy4gzVVHLqEp327tP27tLp+wu5TtdSzLdrsWk/EZIhCMB8VOLHcEwCwAl6b/jhVbxlhTH/LB7yFsWrnEfDjMrA1g48ol8MrA7vYOtD3/Pu5esShje1tLBE3h4mfWLHe4E6P1XIUTY3nQEPRhU0sk495saomgITh8b/j9293egTuXL8rad3d7h7nM27D9ft+5fBF2t3e43vemsB9t6bLsdViPzVW2vf84tbs2W7xO8ZSyvY5l2fyanQ7PEIFgPChWcPpPAJoBLGCMnUNEZwJ4kjF2YakCtOMkilIUDT0xFYpuQJYIXolQ5SNEE4Zp9VyVFoXFNQO6wRDwymgITy69B0dku4yIcRHxJRIaeuKK2S5rAhJiKplC0JQ9uQGZCF6ZoOoMOmPwSBIagl50x1QQGAwGMAB+j4zagBd9cRVJTQcB5nGMAX5vbit0opQ1u8EYJCL4PRKSmmGKRhkAiSgl3DZSYlIGcsz04GXqBoMsEXweCUq6v3lkCY0hH/oTmtkuedyKpoPSdUiSNKbtlWeoaLoBT4EZKm79ZyRllQghOLUx3oLTkXCailRd22qxmo8vA/gkgH0AwBj7kIiqRxHYqFFVHe92ZWYQbG5ZAgbKWLdx5RJMCXjwrz9/e9Km2HK46E5QXiQSGg72DGWlo/ZF45hZF0ZM0TPa7J3LF+Hhl1JTBjz80iEze8Upffbs+iAODCRwv0MGi72dGwbDe91DWWV8rCGEA51RxyyYfH1F0wwc6Ixm9bkfPXcQv9zfiSsXNuHmy87Jyha5r4Qp74bBirZbz5V94/FIOLMm4HicQCAojmKH7QpLfVXCAICIQmMfUnE4qdA7B7OzCjbs3AdNR171vUBQKnriiqMF+dlNU3CkJ5bVZr+7e3jKAJ694mYh3jWkmNkw9gwWezt3yzDJlQWTr6849cMNO/eZ/W15ZJZjtkgp++NIMmlEtphAMD4UO/h4gog2A6ghohsB/BrA1rEPq3Cc1P7cbt2KPfuFrxNZIILxIle2i1ub5Zkl+TJQeNluGSzWdu5qT15EGXZyZcYA7tbtpeyPI8l2EdliAsH4UKzg9P8A2AVgN4AFAP6RMXZ/KQIrFCe1P7dbt2LNfrGuE1kggvEiV7aLW5vlmSX5MlB42W4ZLNZ27mpPXkQZdnJlxgDu1u2l7I8jyXYR2WICwfhQlOC0HLCLolRVx5G+mGlZLRHhrKYg4ooBRTPAACQUHUG/DK8s4dGXDmHzbw9Pas2HYESUXMSXSGjoGIjjmMVefWZdAIOxJBqnBKEZBg53x3DfswfRFU3inmsWQyJCXdiHrsEkmqr9mF0bxNG+GI70xNAQ9iHg88AjEwJeGQMJFXc/805OvYZhMHQPJRFL6jjUPYSn/3gCn/vEdJzVGIJHIjAwnOhPwmAM337yDYuOKoLGaj8YY1miUMNg6Isn0TekoqM3jpqgF9VVXkwNyEioLDXJHYCAT8I7J6IZ1vJ3PfOOqflwskLnItyRWJo76Tf+48ZPQSIyBb9NYT+8XjnnMWX4nBCCUxtCcFq2jE5wSkS/Y4xdRESDSOs9LAUzxtiUUQY4KhKqgdt++hYaw37csfw8fNiXyJjfZePKJdj0m/fx0gc92LhyCb74yZmo9ntw5tRAOT1QBJMcIkDRUm2Vt80tqyKYEvTjq1tfzhBiNlX7MZBIOYlaPwQBIKkZ+I9Xj+Brn5mX0c63rmrGP//VJ6DoBra3XoDBhIr+mAq/xQLd+sF65cImfPOyczIEsFzk+p2rzsUTa5YhoRk40hPDP/zkLXRFk+b2W65YgAXTUlrzAycHce+vDuAbF51lntvaz87F1efPzBLX/serR4YFpquaccfyRfinLwxbod+w8SVz/0e+fkGWCLdYS3O/R8LtXzoPQZ+M2pDXnIbBGtO5TWFzACJJhAXTqvHjDReKbDGBoIQU1IMZYxelf1czxqZYfqoneuBhFbqtu+RsyJJsPpCBYeHbjRefZf59rC+O6x94BX2Wr3wFglLTHVOyrNDXPNqOjt54lhBT1Q1z4MHXWwWnjqLQR/dCMxi+suVlXP6DF/DljS+hdfse3PDQq+gZUrLElMsjs7IEsFzcunrbHmgMuOGhV9G6fQ9e6+jP2M5FmLzM5ZFZ5jclALCiebajuDZDYProXugGXK3QnUS4xVia9wwpZvzXbXkZVV6PY0z28srFjl8gmMwUrPkgIpmI3illMCPBKuKrCXhdrdbl9APEKqgTIjLBeOImOA365Kx1bvvms0bXXISfiqZniSnz2asbLL91u5sdvCxRUQJTJ6Gnmwi3UEtze5lu52O1txcIBONDwT4fjDGdiA4Q0WzG2NFSBlUMXCR3rC+O/riK+rDfXObMrA2kZgpFfutpgaBUWNsqZ2ZtADElcxBsFX/a97Vbo9u3e9Lb7et5W7ducyuDr5fIOQZ7/3GKRzeY67FOcXGhp3V/LsJ1OsdCsJfpdj4e8c1GxTOe2g3B2FBsqm0tgD8R0bNE9DP+U8iBRPQQEXUS0VuWdd8nouNE9Hr65y+LjCfD9rjt+fehG7qj1frWFz8oyHpaICgVTvbqbS0RzKoLONp2O1mD8/VO1uhului8rdvtxne3d2TFw/vHppYIntt/Iqf9urVMezy79h51tJJ3s1t3skKfUx8claW5vczn9p9wjElYpAsE40+x9up/7rSeMfZCAcdeDCAK4BHG2Hnpdd8HEE2n8BaEkyLbalkd8EiQZUJCzbRWH1J0SBLBI1HKgl0mDCVHJygTNuaVQYH3aULs1acGJCR1QFEZ1PT0AB6JIEkSaqo86BpSsuzUp/rl1Pq0NTq3X+f25bLEEFcs7d8rQTeYadUuE5n26UGfhJhiDJeVtkavDkiIJhh8MiGpGab9Ot+vPuBDb0I1s1B8MkHRDTCLJXvQRxhKDscR9kuIWpYbQz6cSurmfeHna81sAWBamlst293KsN9b+73n104zGKo8EogIqm5UWv8V2S4ViMh2yaQoe3XG2AtENAfAfMbYr4koCKCgdxeMsReJaG4x9RWCqup4rydlr94Y9uM7Vy3ArbuG0wzvuWYxqrwS/uax18x1d69YhNqgF3/3nykF/0hS6SokJe+0p5zuUz579aRmZMTZ1hLBf71+DBcvmJaROrupJYL7LbbkPAPl5svOwd5D3Vgyt94UVvP9qzyE1u3DZd+5fBFePHASV58/09GSfVNLBIe7BjCvcQoOdQ1gbuMUM26nTJaNK5dgxx+O4KUPetDWEsGCpnCWhbvVbp3X8dTrx7D5t4cd7df5fTpjSpWZUWON0ylbx35vnaYamOHzlFW7EAhOR4p67ZJ2Nd0FYHN61QwAPxllDDcR0Zvp1zK1xR5sz3bhAw8gJSb79pNvoHdIzVh36643ceJUEusuOdvMIijWPlnYMFcG5XSf8tmr2+Nct6MdK5pnZ2W12LNGeAbKuh3tuHThdHPgYd3/WF8iK6uFZ6Q4Zc6s39GOT86px7r0b2vcTpks1owynpFiPx+r3TqvY0XzbADO9utOGTXWOJ2ydQq9t+XULgSC05FiNR9/A+BCAAMAwBg7CKBpFPVvAnA2gPMBnABwj9NORLSGiPYS0d6urq6MbfZsl0KzCYI+2VTejyTzRdgwVwYTcZ/c2utI7NULzRrJl6Hi1Ad42W79Rk/Hq9vidovJmlHmdq72uPkxuSzd3TJqRmIDzxH9N0WuZ6tAUEqKHXwkGWPmvwZE5EGm6VhRMMZOMsZ0xpiB1BwxF7jst4Ux1swYa25sbMzYZrWsdrNwdsomiCm6qbwfSeaLsGGuDCbiPrm115HYq/OsEft6e9aIPUPFvr9TH+Blu/UbOR2vbIvbLSZrRpnbudrj5sfksnTn99C+z0hs4Dmi/6bI9WwVCEpJsYLTuwD0A7gBwDcBbACwnzH2Pws8fi6ApyyC0+mMsRPpv28B8CnG2FdyleFkr/5uVxRrHy1c87Fx5RLUhbw43B1DlVdGU7UfZ04NFOyaCJSXlkDgThH3aVzs1bk+yarrCHqBKq8XST3TXp1rIi79+BkZVufbWpeiJ6qAkEpHbQj7cP9zB03Nx4XnNGVYuM+qS33Irt62B585qx5r/vxseGQCITUh3F0OluzbWpeiyiNDNxgUTUdNyIukmhKsBtNupG6aj22tS1Ht9yChGjjUPWSez7bVzegZUs24z50eRlJlYIzBI0vwyMiwX59TH8Ts2iAGkipO9CfwQ5s2xVHzsaoZ06b6oWoGdAYwxlyFqPnahVWsSkSQCZAkCbUBL/riakmE5uUkjp4ohOB0UuHaVosdfEgAvgHgynSh/w3gAVZAIUT0HwAuAdAA4CSAf0ovn4/UtyeHAazlgxE37B0kmdTw0VASus7Qm1bKB7wyaoI+aAZDXNGQUA1Mm1qFviEFHw0ksO9wT5ZgbqSiU5HtUv6UywM9kdBwbCBuzkMUU3Sc3RTCYEIznU/5gMQjAQ/97jCuaZ6F2pAX3VEFjdV+DCU1SERZA5gZNX5EkzqSmoGEamRs39QSwfNvn8SlH58GA8ho920tEcQVHVVeyZwnxu8hHO9L4FuPv+4qct1546fglSRougGPLMHvIUSTOoiAzoFkxrGbWyI4Y6o/pbPKIVq117F1VTP8Xgk3PPQqGsN+3HzZfCw4I2xm5XhkCbph4L3OoYyB1pN7jmaJdJ36d6524TQ44SLdL5w/01EYO9q+X04D5YlEDD4mFWMz+AAAIvIBOBepAcMB62uY8cDeQY73xfDuyag5pwQAbF4Vwe1P7c8yE7rt6oVY+2i76/Yfb7gwSxkvOG0o+QP9eF8M1215OaPdbVu9NKPtAqm2+NiNy3DxXb8x2+ptVy/E7U/tx7bVS9G6fU/W/o+vWYbrtryMx25chuu3vpy1fdvqpQDgeCzvF3z5/65Zhq9syS7Dvt/ja5ZhRm0QANA1mMSXN/7ejNMtPr7+V7dcXFAst3/pPLRu35OxD++nvE6nc3Uqu5j+Xcqyi63ToWwx+KhAxOAjk6JSbYno8wDaALyfLnQeEa1ljD09uvhGjpNYL5dtdK7tp5vYTDC+OIkw3YSm/J8C3lb5bzexJy/bTXBqFYPat9lFoHaBqdt+VltyN1GoPT5OoUJaJ6FsLkv2XNeomP5dyrKLrVM8lwSTkaIGH0hlo/wFY+w9ACCiswH8HMCEDT6sYj3ecXPZRufafrqJzQTji5NlupuFOFFqsMDbKv/tZlvOy3azELeKQd36BV+WHeJ02s9qS24XhbrFx9cXar/uJJTNZcme6xoV079LWXaxdVbqc2myfoMhGBuKzXYZ5AOPNB8AGBzDeIqmIejDzLqUcRhXr+9u78iyWN+0colp7exkKy3s1gWlpj6Qba8+sy6AzauyLb+f25+SPu1u70Bb2pb8zuWLsGvvUWy0t+2WCBrTVuJOFuJ8eoFCLM83tUTw2pGeLFt1p/2stuRuNuu8bzWGMs+9UPv1OfVB137qZMne1hLBrr1HHWMopn87lc2vv93yfayeHU51iueSYLJSrOB0E4A5AJ5ASvNxDYCjAH4NAIyx/yxBjBm42atHVQ1Ji+1yyC+Z1s5emeD3SIirBnSDwSsR6tL20Fww1xT2F5XtMtaMpXi1kLKEWDaLCbFXr/JKkAiIWezQq6skDCYM+D0SFJ2Z1uZVXkJcMUxrdG6F3hD0YUDRoRup9i1JgGEMbw/4JAwmdHgkQk1AQn/cyLB3P2VZrg/40BMftm7ndfMy+fopAQkD6eN8sgRZIsRV3TynhJqyQzfSZRCQVYY9Frv9elPYD1mW0DOkwDBS67nQtCHoRXdMNW3XuSU9z0QxjNzZLoXA+4i9bm5jb8+CGYs+VC7i6LFAfPORidB8ZFLsa5cqpDJV+BwvXQACAL6A1GCk5IMPO4qioWMgjv4hBbc8MZyOyK2cuwYVfP+LCxFT9IwUXG5dvfm3hyc8TXYs03YLTSEUacLjj2EwHOqL4Sf7OvD5xTOwYec+fOaserR8ek6WHfphm6W5k/24NfviPgeLdG67fssVC+D3SHjs5cNmvU4ZJvYMFKdUVqstu3X93SsW4a5nDqArmsTGlUuw73APIvMaso59+KVD+Npn5jmWYbdfb2uJ4Nxp1agNePHOycGcmTJ8X49HGjPRuCQR6kM+x74yvzGMg13JMe9DTnbwAsFkpKh/9RljrfYfpNJjWxljXy9RjDnpGlJwrDduDjyATCvndZecjd4hNct2fZ3F2vlY38RaK4+l1XMhZQlr6YmBX/cVzbPNAcCNF5/laIdutzR3sh/nFunrXCzSue36jY/sxZGeWEa91rq45bndNt3Jvtxqy25df+uuN83pCjbs3IdLF053PJbH6WbRbrVf5zbt1ikUnOK07luqe2bvK0728aIPCQSFU+zcLs9bJ4cjoqUA9rgeMA7ksqauCXhRE/DmtK62Lk+UqnwsVe6FlCVU9RMDv+7WjAm37Al7xolbFkk+i3S+PuiT82aY2Lfnq9OtnGN97lk3+bJ2sjJqdAOqbhSUKaPpBsYat76i2WLi60UfEggKo1iRw/8G8AwRbSCifwGwBUDr2IdVOLmsqfvjKvrjak7rauvyRKnKx9LquZCyhLX0xMCvu9We3M2q3G5p7mYlns8ina+PKXpeq3b79nx1upVjzbpxi6dQ23iPLMErSxn7uh3rkcdes+XWVzy2mPh60YcEgsIYicnYJQB+BaAbwCcZYx+VIC5X7KIoRdFwIprEQFzLeAf82I2fgkSUEolJhCqfhI6eOJKakeGEeLpoPuY3hk1L6EDaHltoPjIouYjPMBhODsah64ABBpkIlBaPPvrSIbMtbm6JoLHaj4Sm4/20eycDUB/2oXXbHvOebV4VwdSAB/0xLafm428vPwe1QS+ICL1RBWttWokpAQ9+uu84PntOI6YGvabFe2O1L0vzsWnlEjAAAZ+cEYtV87FtdTMUjSHolzPs4vNpPh5a3Yxei/36rLoA5tQG0RdXMaRoZlkXzK1x1XwAqZmuVd2A1yYOzSc8dRJ7Asih+YhOVB8SgtMKRAhObRuKzHa5DcC1ANYAWATgFgDfZoyNWytzmtvlaF8MSd3AyVNJBH0ywn5Plo00nyeDP+A3rlyCxrAPBkNZZHuUMtulNuDNelA+8vULEK7yQNWMsjj/MqDkD/RkUsO73UOOIsxvXnYOqjyEY30J1Ia8ePrNDx0/YOOKbmagBHwy1u1oN63H508LAUgNuFWdQdNT9/+Op982RZw/uv6T8EoSgn4PDqfnXXEaZLS1pAY2skw42hNHU7UfR3qGBxL/dt35YAxoqPajyiNBMwyoOoNHJgzEtSx794awDwAzs110A/B7CZrGUhkzHgk9toGR1V7dPvhprPZhKDmc+eP3EM6orsKBzqhZt+PcLy4DhFz/AABw7JsTmDEmBh8ViBh82DYUOfj4NwB/xxiLp5fnIDW3yxWjDrFACrFXd7Nu3rZ6Ka6490Vz+fYvnYfzZkyd9OryImybT2cmxF59Zm0gwzr9intfNNtqLvtxtykCrPbqhewDuE9HwGNws0zn660W6LmmNuA28Vb79CfWfhrXbv6Dax1O9ur5LObznZdTu6+wPiIGHxWIGHxkUlSqLWPsW7blI2nL9QnDSXDqJkizC0yDPvm0EIgJgWl54GSvbhdh8nX5BJluYlBmEXq67WMXg46FmDVfWVabeOt6Ltx0O87JXj2fxXy+WJzavegjAsH4UpBCi4h+Z/n7Udvml8c0oiJxEpy6CdLsAtOYop8WAjEhMC0PPFJ+ESZfl0+Q6SYGJYvQ020fuxh0LMSs+cqy2sRb13PhpttxTvbqroLTAkW6Tu1e9BGBYHwpVB4esvz9Z7ZtEyoUaAr7MasugB9+5Xzz4eFm3bxr71Fz+d5rF2NOffC0sC4Wts3lQUMw2179zuWLsLu9w7RA5+vy2Y87TRGwceUSPLf/hGn/3fb8+xnTDvCy7RbsTmVZY3CyTOdxb14Vwcy6QN6y+Hq7fXpT2J/Tlt1ur57L4rwp7M9YX8w0CqKPCATjS0GaDyLaxxhbYv/babnUuNmrDyRVGEiNhFSDQU4L0ax26wmVQdENyJSynNb0lP2y1yNBMxhUzUDAJ5t/TyYh5ulsp15OltXcXp1bjHML9GDaAl2Whq3Mgz4JMcUw9/F6CKrGhpdlgmqxX+fiS1lKiU6JAJkIIAJjzFzHGODzSFAs0xGEqyREEwZ0lsrCsVuhV/kkJBTDrMtajk8GoknDjIf/5mXzZW7dzuNP/UioqfKga0iBR0LGcVVeCZrOQOn4NYMNx5c+V26rzu+pomjoGhq+voylziE1Tx/lnEbBrZ2UYd8Rmo8KRGg+MilU81FDRF9G6puSGiL6/y0FTx1lcKMimdTwXs8Q7nv2XXzz0vmIJjVs+/2hrLTDe69djJqgF1tfPISXPujBxpVL4PcQGIDBuIZbnngDjWE/vnPVggwb9smSgnq62jaXk5V8IqHhYM8Q7ndIi924cgl2/OEIXvqgB5taIuiLxlEbDpj78hRVeyqtU8qqNaX14ZcO4darzoWiGvhhAWXZ7det2Tjth7ox/4ypGcfdc81iVHkl/OLND/H5xTPwo+cOOpZ98KNTjnbruezhN65cgp+/cRzXLJ2NhGpkWcs//NIh3HzZOTg3fS81zcC7XUN5rebd7r1THymn9iMQTCYKfe3yAoAvArg6/fcX0j9XA3ixNKEVRndMMe2lu6MKbt31pqPVdMp+PYEbLz4Lx/pSVs6yJMMjyaY1+7pLzs6yYReWyZVNOVnJ98QV087c3j437Nxnts31O9pxdtOUjH3d7NOdrMat+y+PzMKx3jjWFliW3X6d779+RzsuXTg967hvP/kGeodU81i3st3s1nPZw2/YuQ8rmmejozfuaC3PLee5rTq3Yc9nNV/MvS+n9iMQTCYK+uYjPYdLWcIV7lZr51zKeWtGAf/HJV92gFC8Vy7llMVgbau5srGsmRvWLJFcGSn29fbskrEoy80y3WrdXkgWjlOduWLKNX0Cz5gBYNqw58u4Kebel1P7EQgmE+M2hzwRPUREnUT0lmVdHRH9iogOpn/XFlsuV7hbbdRzKeetGQWp6b6RNztAKN4rl3LKYrC21VzZWNbMDWuWSK6MFPt663H2fjHSstws063W7YVk4TjVmSumXNMn8IwZAKYNe76Mm2LufTm1H4FgMlG0vfqIKyK6GEAUwCOMsfPS6+4C0MsYu4OIvgegljH23Vzl2EVRXPPxw1+/i7+9/BwMxFUcOHEKly2cDlVPifV27T2KyxeegZqQDx4JODmgYNoUP7yyhIBPwon+pOkUWQrNRykEa2UogitLinhnX3IRXyKh4dhAHH1DKqq8EtZbtBVOmo/66gB++OthncaGv/gY+oZUBH0yYoqOhrAPEqV0S1an03kNIcgSgYjhWF8C02uqEE1o+OGv38WGv/gYVI1l1b+5JYLGKX50DiQztBX3XLMYEhEaqn3wSBISqobW7cPXsq0lguk1fiRVA71DKu579t2sOGfVBRDyy+gdUrH20Uyrdt1gmBr04uRAElVeCTc99lrGNdl3uAeXLZxUrEXGAAAgAElEQVQOzTBwuDuGp/94Ap/7xHTMrg+iazCJhrAPc+tC8HgkaJqBd04OZmg++DXh+zdV+zG3PlRwXylTzYcQnFYgQnBq2zBegw8ASM+I+5Rl8HEAwCWMsRNENB3A84yxBbnKcJrbpT+pQtMZVN1A0CfjowEly1p9Ro0fbx3rR22oKuOhe/eKRagNedEf03DGlCpUB2ToBsYs26UUD68yfSCWLeWS7ZJIpAbK1oHC3IYQAl4JXpkQTerwSKlMLFkidA4q6IkqmFFTBQZk2Za3tURw37PvomtQwd//5cdNu3W7yPKmS+djasCDzgEFVT4Z6231n4opAAE3PfZaxnq/TIipOlZb5nC555rFmBrwwO9N2bO/dawff35uEzbs3IfGsB9//5cfR9AnZ9ik33PNYjz39kdYsXQ2jvWmXtNY7eHNAdCqCBpCPiQ1A3I62+WjU0mzLEe79FXNWHBGtZmVcrw/hqTG4PcQPHLKsn1dAfbquSjDgb4YfFQgYvCRSVGvXYjoGiKqTv/9D0T0n0Q0mjTbaYyxE+m/PwIwrdgCuoYUJFWGAx9F8UFXDKfiepawbf2OdvQOaZjbOMUcePBtt+56Eyf6kzgVV9Hy4CswDEJTdRVm1AbRWO0f9UOmFII1IYIrDp7FMFb3dKT0xIc/CF/r6Efr9j1Y9eAr0A2G+nAV5tSHMKM2iLpQFaIJHa3b9uBUXMVXt76Ct08MZokuubjytY5+9MYUV1Hmhp378EFXDL2x4UG5tf6Q32N+42BdbwDmwIOX+e0n30DQ78WqB19B6/Y9uOoT002RKo9jrS2Obz/5BlY0z0brtj1o3b4H12152THetY+2Q5IkzE5fB4NRRllc+JrR7h8dbvc9Qwq+uvUVXP6DF/DZu57Hm8dOZdUxkn5SLu1HIJhMFKv5uI0xNkhEFwG4HMCDADaNRSAs9RWM49cwRLSGiPYS0d6urq6MbTz3P+iT0/9RDQtIOVxcmkswNxIxWiGUQrAmRHDljVt7dbNX14zsZl+o4DSf3brVAr1YoanuEq9VPGo/ttA6ChF329t5vmOK3V+Q+9kqEJSSYgcfvNd+HsCW9Gy2o7EAPJl+3YL0706nnRhjWxhjzYyx5sbGxoxtHokgU8piPaboGQJSDheX5hLMjUSMVgilEKwJEVx549Ze3ezVPQ7/SRcqOM1nt261QC9WaCq7xGsVj9qPLbSOQsTd9nae75hi9xfkfrYKBKWk2FltnwJwHMAVAJYAiAN4lTG2uMDj5yJT83E3gB6L4LSOMfadXGU4aT4GVQ1JNT2lt5R6T91qeU+9qSWC6ioZz799Msvo6N5rF8PrkbDxN+/lNCAaKeOl+Xjk6xcgXOWZdM6s48i4CU470roHLsacNsWLuIIMF0+vJEExDJzoT6Cx2o+uwSSqqzwZgs3NLRFUBzzQ0lPZDya0jO1WI66aoAcn+pMwGMO3n3zD3Gd761J4ZQkMgKYzbHnhfbz0QQ+2tS4FMxgCPhlJjWEwoaI/pqIh7INXJvi9Mg53x/DWsX5ccm4T1ls0H3YtR1tLBPVhHxQt1Ue3vPA++uOK63T3QOoVimEYGExq5vVy0olYj+mPKzjRn8DatKbl7//y46gNebOud13Ih5pARfcPofmoQITmw7ahyMFHEMBVAP7IGDuY/rbiE4yxXxZw7H8AuARAA4CTAP4JwE8APAFgNoAjAK5ljPXmKsfeQVRVx4HOaNbDbkr6oWwwwOsh1AZlxJMs42HGH0Y1AS8YqGQf2KXOdgn4ZJwcSAoB6ugYV8Gpta02hL34x5/+KcOV88qFTfjmpfMzxNFtLRHEFR0GS6ef1gVw9zPv4Jf7OzGzNoBtrUsxlNAwNeiFbgCanmpvdzz99vA+q5vh88gwGDCYUEFEWeLsM2v86OiNY+Nv3styCuUiV17e5pYI6sJeqDozBbFW0WosqcHrIXzdkiGzqSUCCcDDLx3G5z4xHfMaQgj6ZTSEUu6ifGCdGkCca5oA8gyZIUXHGVOqchyTGgDd9+y7+MZFZ2UMtu65ZjEe/N0HJflHYxwRg48KRAw+bBsKHXwQkQzgT4yxc8cqqpFg7yDH+2K4bsvLGe92Z9YGcNvVC7H20XZz+fE1y8AAfMVh3x9vuLCirce7BpP48sbfT7rzGmdK/kB3a6uPr1mGtz4cwO1P7Te3bV4VyVjm+9rbtX152+qlaN2e+tbPrQy+j3Vf6/bHblyG67e+jNuuXlhQDLd/6TwouuG4Ly/LKYYr7n3RXOZt1dqWc12D25/a73iM9drlit96fAUiBh8ViBh8ZFKw5oMxpgM4QESzxySkMcJNxMeFeHxZM5irgK7SBWhCgFoZ5BKc2sWR+QSmbstWYWc+8aeb0JQLSguNIZeY1U3kLVu+cXATjRbqVOomNB1Lp1OBQDC2FCs4rQXwJyJ6loh+xn9KEVihuIn4uBCPL3vSs2hORgGaEKBWBrkEp3ZxZD6BqduyVdiZT/zpJjTlgtJCY8glZnUTeeuWDB830WihTqVuQtOxdDoVCARjS9GptkhNJve/ANxj+ZkwmsJ+tLVEzIcMd0bc3d5hLm9qiSDol/DakZR7pHXfrTc0oz40moSdiac+5MPWG5on3XlNNuoDvqz2t6klAlkCdrd34M7li8xtu9s7sHHlkqx97e3avrxr71GznLbn38fdKxZllHHn8kXYtfcoNq5cYv621/Hc/hO4c/mirJic6mxriWBWXcBx340rl+C5/Scc69i196i5bG2r1rbsFv/u9g7XY6zXzikmp+MFAsH4M64Op2OB03vJREJDb0KFphuQJELQJ0HVGNT0qxa/R4JXIsS1lN26VyKQBMiShIZQtmlQGToaOmIXnWoGE9kuI2dc3qMnEhp64go0I5WZNTUgIZZk0FlqYjRZIlT5JCQUA0QAY4DOGGQi1AQk9MdTbViWCJIEGAbglQmqzmAwBolSzqAJ1YCWbuseWUJcTbmn+jwSFM1A2C8hmjQy9vWk+05MMdK+ODDr4HWGqyREE8MxNIZ8ICJ0RpNmHLwsvhz0S4glh+vgZWgGg1eW0BjyoT+hgcCQ1IyM4yWiVJaMLT5+bFPYb1qrd0aTUHXDdIlNqAwyATpLZRIREXwyQdEZGGN5+4lbmW7HjuNzQ2g+KhCh+cikoFltzVKIlgG4H8DHkfL3kAEMMcamjCq8UaCqenq+DAW3PPEGGsN+fP+LCxFT9Iw5WjatXIL7nztoqvTvvXYx6kN+Uy3PqRTr8kqJUzBMMqnhYM9QRnZJW0sEkgSseSS1bu1n5+Lq82dm7MPnOInMa8D9lnlLzKwYW8rqppYI7rdkpNy9YhHueuYAuqJJbGqJ4HDXAOY2Tskqi9f1o3Q/+f7V52alpm9qiaD9UDci8xrQfqgbF53TBFVjuPfXB7LKunP5Irx44GTW+dgzZtpaIth7qBtL5tabbqm8z0oErN2xz/Vc21oiWNAUxnvdQ3n7QjF9hs8Tsy7HdbUeK/qjQFAcxb52+RGArwI4CCAA4K8B/PtYB1UMndEkOnrjZjreukvORu+Qag48gLTF+s59WB6ZZS7f8sQbONIby7JarhTr8kqJUzBMt8XeHBi2SD95KmmuW9E8O2ufDTv34dKF07E+bafOP+ABZ8txvh9fvnXXm1h3ydnmtk/OqXcsi9fFj+V12svm6y9dOB0dvSmLc6eyvrv7TcfzWWeLb126rA22qQ/W79yHzkEl57mu29GOzmiyoL5QTJ/pjCazrNnt19V6rOiPAkFxFDv4AGPsPQAyY0xnjG1DyvdjwtCMlLW6VSFvXea4qfTtivdKyRyplDgFw7hluwR9w8LHYjNQis2KOdY3bJue71i3TBUeC2PDfa9Y63Z7fLmmPsh3rm7XdTR9W9WNguJ2y7jJVbZAICh+8BEjIh+A14noLiK6ZQRljCkeKWWtblXIW5c5bip9u+K9UjJHKiVOwTBu2S4xZfgDqtgMlGKzYmbWDtum5zvWLVOFx0I03PeKtW63x5dr6oN85+p2XUfTt72yVFDcbhk3ucoWCATFO5zOQcqd1AfgFgBTAWxMfxsyLjg5nHbHkuiNqqat8h3Lz0NSYxnvjze3RDA16EFHbxzbfn8IN372LNSHfKgJ+RBXhgViAEb17na8RGe53jEDqAjBbJlRchFfMqmhN6FA04dFpB4Z6IupWPNIqu3+81+dBwPI1FmsXAK/V4JXltA9qGRYpF+5sAnfuercLNfeviEVBmNgAGbXBVJ/MwJjKaEoA8OxvgSqvJI5oy3vJ9UBD3qiCqqrPIirBtbvaMdnzqrHmj8/G16Z4LGIYvmkeI+9fBifW3Qm+oZSFuwBnweyBJyKqQhVeTKmO+BOrVVeCQGfB16Z4PMQuqNq1nnLMpl6GDd9y8waPz46peDGR937rKrq6I0pWQ7Hc+qDmFsfGrHmY9oUP+LKuLsMC8FpBSIEp7YNxWa7EFEAwGzG2IHRRjUSnLIH3usZwn3PvovlkVmp/z4Y8MP0cn3Ih/qwD5t+k5qz4u4VizJcDe+y2FOP9sN7vEVnTgMdYHSDp9OYcbFXtwtON7VEMLvOjyM9CRCA9Tv3ZXzQyxJBNwysfOBV85jtrUsR8MpQdAOxpJ41WGlrieC/Xj+Gzb89jJm1Afz79Z9EQjUybMb5vC9/e9k5qakIDIaTA8mswciT65bBK0k4MZDMivspSx0PrW6GojHc5yBi3bwqgvpQam6XD08lEPLJuP+5g1n7PfC1CHyyDDntyRPwSuiLqRmDhXOnh2EYqVdYqmUumu2tSzGlygtVz872UlUd73RGcb+D3XquvsGzXbR0FpI128XrkRBNaLjhoeH7Mo7zK4374EMMJEaPGHxkUtQrEyL6AoDXATyTXj5/ok3GeuIK1u1oxy/3d5qTaq21LK9o+wNWPfgqLls4zRTfdfTGwRiwetseRwGZJBEaq/2YURtEY3V2Kq5rLOMsOnOKUwjfypeeeLbgdP2OdkQTBk6cSpjzuDzRfgyX/+AFrHzgFcQV3Rx48GNWb9uDmKJj1YOvwueRHEWYK5pnm8u9Q6r5gcvXfXf3m1gemYW1O9qRUA2sevBVnIqr5sCD73dN28tIaoZj3NY6jvclTCGpXXi69tF2DCY0XP/AKzgVV03xt32/v364HQc7o1j5wCv404cDePPYAFZv24PW7Xtw3ZaX0bp9D65pexkAsPKBV3D5D17AE+3HzGuiGcyxz3ZGk6ZY1H4dcvUNj0fCmTUBzK4PYUZtEHWhKpxZE8CM2iAIZA48eFk3PPQqCFT0c0MgOB0pVq/xfQAXAOgHAMbY6wDmjXFMRWEXm+UT0XERGxfCuQnIRkI5iM7KIQaBM8XYq/NtboLNfBbpVvvyXAJsaxnFCjqd6sh3HoVan+cSjrteR92AE1oege1I+oboZwLB6Ch28KEyxk7Z1k2oS5ldbJZPRMdFbFwI5yYgGwnlIDorhxgEzhRjr863uQk281mkW+3LcwmwrWUUK+h0qiPfeRRqfZ5LOO56HWXnxxnf363OkfQN0c8EgtFR0OCDiH5BRPOQmtflegAyEc0novsBvFTSCPNQH/Bl2Ks72VLfvWIR2p5/3/x7Zm0Vdu09ijabVfRoLZfLwea8HGIQOONmrx70S65W5la7dKf1bhbpVvvyupAX91yzOGMfbjPObdbd7NTbWiJ4v3PAMW6nOtwszXftPWpas7vVZV2/u70Dc+qDjm3ZaUqFtpYImsLOM9Q2hf0ZdY9F3xD9TCAYHQUJTonoGgD/AuBRpMzFrkhv+m8AtzPGkiWL0IabvfqAokHRU7bPYb+MpMZMu2qvRNAZA2OAzyPBMBg8HsJUvw99cdX8qpQIAMi0bB4J5WDNXg4xVCATYq9eE5Cg6EDMYqfOLc25XTq3ROfLfL+AT0JcMeD1EFRt2NZ8SkDCQDyVieKTJUgEs/3zMuzW7PZlu+06r4PXzS3TdYMh5Jcz9uXHMpbqU/wYuzU7j99+HN/eFPZDliX0DCkwDMO0Sfd5ZNRUedA1pJhiUK9M0Axk9V3eF3QjPbWCxQLeas9u3ddeF+8/9n5VG/Caz49x7mdCcFqBCMFpJgXZqzPGniSip5GaWO4qpAYhfNTyNwB+MNoIR4qmGeiMJXEqrmF9OtXWyV6dq/u/9pl5ePilQ7j5snMwdZoPtQEv3jmZyEipa2uJ4Nxp1SMagHAR6ERSDjEIslFVPSvbZdvqZiQ1ZrY/NwtxR/vxtNX5/DOm5rRb37hyCXb84Qj644pjqmp7uuwfpTNQDn50KstWfePKJfj5G8dx8YJpGf3oH65eiBOnkhn9h+/7hfNnmjbqVy5sws2XnZPzPK3W7tZ+WB/yOWZwfawhhAOdUde+a88+s8dgz3A7cHIQ9/4q2yZ+6w3NmN8YxsGuqMgiEwjGiGI+XRUAQwD8AMKWn+oSxFUwndEkFI2ZDzE3e3Wu7ue/uS2zk40y3yYQjCU868La1o71JTLan5uFuKP9eHp9Prv1DTv34caLz3K1Yudl8/7hZKu+Yec+rGiendWPFMvAyb6v1Uad97l8cdpt1zujSdcMrnx9136cPQZrtgvf1ykLh9clssgEgrGjoG8+iOgqpL7d+BmAJYyxWEmjKgJVT83AyR8K1qwWK05Ke003wFz2dVPOCwQjxSlLw57R4ZaRwfJYnec7XpbIdZths253szp3ylix9j23fZ3iKtQWXtMN13PPl/Viz0jJl+2SM+PHxW5dZLcIBCOj0Flt/yeAaxhjfyplMCPBK0tQ9ZQq/lhfHP1xFb60NbL1YeGktOfqeKd93ZTzAsFI4VkX1rbGMzr4Ot427e2RW5rnW+92vG4w122SzbpdcqnLKWOlPuzPu69TXG6x2LPPPLJkWp1n9VGH62ntuzwjJV+dPEPFem5OZeY6ViDIx0h1M5NVK1K0w2lJgiA6DGAQgA5AY4w1u+1rF0VpmoHeeBKdAwrue/Zd3PDpuZhRGwBjwB1Pv22+P+bTe193wRwMJjTUBL2YXu2HJElZNsoj1XwIoWd5UuB9KbmIT1V1xDQV0cSwyDLoT4kuj/bG4fdIqA/7MBjXsNbWHqu8EuKKbhqRWfUayz7WAI8kg5+SRyZcv/UVHOuLY+1n52LVZ+aZIk6fh/Cn44OmY+jMugB+/24nInPrTdfRgx+dQvO8hkxr8ZVL8JRF8/HNS+cjoRqYWVeFnqjqqPm4ZunsVJ3pwQB3JY0rGhKqgYBPzqrjfgfNhyTRiDUfh3uGcKQnhqBPhkSUVefmVRE0hHwgiRBNaLjj6bcrQfMhBKenERU++Bg7e/VSkB58NDPGuvPt6zS3yzudURzuGsBZjVMyHtqbWyKoDXnBGOD1EDoHFMcHFQDTRtljU8AXynhbqwsKo4j7MmH26rVBD76y5RVz3Y/SdujTplTh5EACdz79DrqiSey88VPQ9VR/NRgQ8kuo8ko41pdtfS4BCPokRJVhd1Inkee/XXc+Al4Zfq+E7qiCmbVVkCUJQ7Y5UGbVBVDtT9mwv3syivuePYiuaBJbV0VQE/JiKGlAJqS/qQBUneFUXHO0W+fi7+9cdS66owoIKfX69JoqfNA55DjvitsA0mqBbu+7TvfeaoGuGwz//PP9WXO1qFph2S4T+M+FGHycRojBRwkZzeDjeF8M1215GdtWL0Xr9j1ZX4vedvVC3P7UfmxvvQCrt72atf2JtZ/GmTWZZkEjoWswiS9v/H1W+T/ecKHIPJlAirgvJX+g87Zqj+X/rlmGi+78TcY63m5vu3oh1j7anrHeuvz4mmWOZW5bvRR+r4zrtw5v27wqgtuf2u+4b+v2PRl1Ou3nVtftXzoPrdv3OO7rVpb9/NxiG03/yXXvAVRyfxWDj9OIyTr4KBdhAwPwSyJqJ6I19o1EtIaI9hLR3q6uroxtXHTmZjPNBWRuwrixEpYKu+XyZCLui1t7dRNIWp1C+TqrqNO+3rqcy/rcLhwdrfW5W11Bn+y6byE26rliK9V0B6K/psj1bBUISkm5DD4uYowtAfA5AH9DRBdbNzLGtjDGmhljzY2NjRkHctGZm800F5AZDI7bx0pYKuyWy5OJuC9u7dXNFly2fXVvF3Xa11uXc1mfc+EoZ7TW5251xRTddd9CbNRzxVaq6Q5Ef02R69kqEJSSshh8MMaOp393AvgxUpPXFQS3Tt619yg22WymuU3zD79yPjwy0tbqhVkyF4uwWy5Pyum+uNmrA8yx3drt/52mA3CyGuf94af7jmXUt7u9I6t+bn3elsf6nPeVrGu5qhmz6gIZ6zauXIKTp2KuluZO58eXx/I+5br35dQuBILTkQnXfBBRCIDEGBtM//0rAP+LMfaM0/5O7yWTSQ2nEioMALrBUv/1SQRP+j9KKe3vrKUtpnXG4JUkNI7CRt2JfIK0MhKsnVbkEiVamBB79SqfBEVlkAhIaIa5LqEM25NLBDAQGoJedMdUqHpqP79HAgNlWY1zG3PNYKiukjMszIM+CTFl2KpdopR4NfWaBqjyEuIOVu8MgN8jY6pfTtWVLi/gk8DfVMgSMmzW6wM+9MQVswz+m2feNIZ8OJVMvQIhIvhkQlIzzJgYUtMheCRCXHHuM/zeqroBryyhMeRDf0IryAK9gvuj0HycRkxWzUehPh+lZBqAH1NqYhUPgMfcBh5OaJqBjlNxdA0mse33h1yV9Td+9iz86y9SWQObWiI4pyE0pgMPILetuciGmRgMg5VNiqRTtgtvn60XzsNdzxxAY7UvKyOFp4l/4fyZGdla/Nhbrlhgnk8h7WxqIHsfpzquXNiEmy6dn2Xpfn/aMt2aVrv6onk4eUrNyrqZXx/Cez1DjlkvbS0RLGgK473uZIYFur3Ou1cswl3PHEhl11jORdOMrDT5TS0RPPX6MWz+7eG891pMQyAQTBwT/tqFMfYBY2xx+ufPGGP/UszxndEkOnrjuHXXm47WyNwG+pYn3sC6S87Gsb6UpXR3bHxtkd0sooU9c2kpp+veE1eyLMV5+7x115tYd8nZjrbj3939pmlX7nSs9XwKOV+nfZzqWB6Z5WjpbrVA51bqmg5H6/aeuGLarNv7JrdCt1ug2+vk18Z+Lk726ut3tGNF82zXcxcIBOVBOXzzMSpU3TAtqgtV1nM1/ngi1PUTQzldd7dsEaf2ad8nXzYXP59CztdtH3sdhVqgy+lZo3NlyBSaQZOvTuu5qC6W51YBr+hjAkF5MuHffIwWryyZFtWFKuu5Gn88Eer6iaGcrrtbtoi1febLSHE7lp9PIefrto+9jlz9yX6cTM7nli/rxX5N8tVpPRduu+50Hm7nLhAIyoMJF5wWi5O9+uHeIVfNx6aVS5BQDQR9Eh5+6Qg+94npmNsQRNDrQWO1f9ze+wvNx8RQ7g6nhWg+2loimBrwgAG4fusraAz7cfNl8zG7PoiuwSRm1FTB75Whaga8HgnRhIYbHnrVPJ5biEuSZGZz8GtiLWswriJU5UHrtj0Faz42rVyC9sM9uOLPpkPRDRzujpnup9talyLs8yCpGfhoIIGaoAcn+pMZrqlzaoN4r3soQ/NhP3+75mPaFD/iig6vLKFzMJml+ZAA/MNP3srSiEwihOD0NGKyCk4nzeAjpuioSz9YVT2VKXCkZ/hBuHlVBH6PhNXpB+tEfPhXsLq+oimXuV00zcCHA3EoGoMsAR5JgiQBQwkNUwJeKFoqG0czDHzYn0BjtR9HLW24rSWCWXVVONabMKcRuHJhE26+7JyMD2DrB7SThfiC9JQC/XEFJ/oTGVMS/Nt154MxoLHaD79HglcmKDoDYwxejwTdMDCUHM5ICfkl9A6pWPto5mCpPuzF8b4EvvX4666DCi44lWUJ3UNJxJI6DnUP4ek/nsDnPjEd8xpCCPpkeORUtovTwGrTyiUIV3kgEUE3GNqefx8vfdCDzS0RTK+pQk1gUvYxMfg4jRCDjzLB3kE+7I/j2s1/wLG+uGnR7Gbp7GQDXSF2yoLSU/IHurWtcmbWDtubW22/C7U4z2VLzsty2tZY7Xe1H+d12/uG0/7bVi/FbT99y/Wc8sX5+JplmFEbLMgGP1+8duv5Sdy3xeDjNGKyDj4mheCUP4zy2UM72UALMZpgvMglkLS2xbEQaFrLctvmJjy1i1g5TvtzsbfbOeWLkwu/RyOUdbOiF31bIChfJoXglIvO8tlDO9lACzGaYLzIJZC0236PVqBZiIW423a7iJXjtD8Xe7udU744ufB7NEJZNyt60bcFgvKl4gcfVnvptuffx90rnO2h7712MWbUVgk7ZcGE4WSFvnHlEuzaezTL9rtQi/NctuT5LMSdtnPrc6e+4bT/nPpg1jpu726N38nafZNleoNC7M5zxetkPS/6tmAyMPd7Py/6pxKoeM0HACiKhq4hBbrB4PdI0AxmWlJrugFJInglgtdDprV0lVdGQ8g/GayWBWPDhNirVwckJFXKaG/cMpzAYLBhO3Jup263DM+1zO3bkS6L26TzfQwjZYfO6/DKBFVnaAr74fXKZr8wDAM6g2nLLhPM7BnDYBn29fUBL7pjimnRrrNUKm7IL2EoOWy/zuvg8PMGUtMg8FjdLNHJEkcuG/VJiNB8CHJSRjqRyav50DQD73YNZdlOv3jgJD6/eEZGmuC91y42Lda33tCMhlDqvy6RBisYDxQlO9V2U0sECxpDGR+udjv4tpYI7rOktzq1TS6sdGrL/379J5FQDXz7yTcyMmScLM956u/Nl52Ttj4fwr2/OpC1nzVrJpd9faF9i593rrokiXJaok9ScalAMCmp+NcuThbL3CrabtNstVgv1pJaIBgtXUPZ9urrd7SjK4/1+TqbpXmutul0fO+Qag48gJSFuZvlObdst1qfO+3HY8jXdwrtW3y/XHUJBILJQ8V/85Evg8C+3smmuZwsuAWTFzd7davVf66MDuuyW9ssJCMlX1YYX5/PGn00GTVumSMVNVcAABVOSURBVDT56hIIBJODiv/mI18GgX29k01zOVlwCyYvbvbqVqv/XBkd1mW3tllIRkq+rDC+Pp81+mgyatwyaXLVJRAIJg8VLzjlDqfdUQUNYR9kiRDyyVB1lmX3vLklgikBD473J9AQ9qEu5ENNINNuejJoPoR4dkSUXMSnKBq64wp0PSXcTIlICYaREkPbXTy59fn8aSEoGsOpuIr+mIo59UHMrQ8BQNZ9BrLbspPm4ztXnYvuqJLloropbeUuE5mCzqRuYCCuZbmocs2Hvb5Hvn4BwlWeLLt3fj7zGkII+mXUBXymSDTgk3HyVBL3/voAvnHRWWasvK75jeGiBKWF9IEK7idCcCrISSUITifF4ONA56Bp73zlwiZ889L5WG+dj2LlEtSHfXj494ew+beHTRvpLS++j1uuWGA+RCv0QZSBEM+OmJI/0FVVxzud0SzBafuhbiyYPjVj/pLpU/043pfADx1EofzD2E3oCQy3Za9HgkciqFoqW4UxhiqfhI/6kxm26m0tEcQVHf/6i7fRFU2awtObLp2Pn79xHJ9bdCb6hlRzXhY+AOKiUl5fwCfj5EAyKy5+PvY6rULa7a1LEVN0BLwSjvUlMuoCkGGrnqtNF9IHKryfiMGHICeVMPio+NcundGkOfAAUmK69Tah6fqd+3DgoyiWzK03133r8dexPDLLFLNxFf2M2uC4Tjg31gjxbPnSGU06Ck4vXTgdt+56M0MMHVMMrHURhd74yF5TDOp0n61tuam6CnUhP6ZNDeDMmgBm1AaRVJk5CODHrtvRjt6Ygtc6+jOEpxt27sOK5tm46bHX0Lp9D67b8jJat+/BDQ+9arYpa326Ace4+PnkEtKu3rYHAa+M1u17s+o60hMruE0X0gdEPxEIJpZJJzjNZa0ehJyxzs1GupIR4tnyxU1wytiwsJOv09Lt2tWW3EVoXch9dhNp20WtvG438bZTXW7tzy1ee51udRUzNcJorNpFPxEIxoeK/+bDLjjNZa1uF+252UhXMkI8W764CU6JhoWdfJ0n3a5dbcldhNaF3Gc3kbZb/3ATbzvV5db+3OK11+lWVzFTI4zGql30E4FgfCiLwQcRXUVEB4joPSL6XjHHNoX92Lxq2LZ5d3sHNq1ckmHBfPeKRZhZF8iwX77nmsWONtKGwdA1mMTxvhi6BpMwjMrSxBRiUy2YGJrCfkeL8ef2n8DdKxah7fn3zfvFbdSdbNY3t0Tg91BGu+dCTwaG430xfNgfx8lTccc27GTz3tYSQV3Qh8fXLMO21Uvxo+s/id3tHab9uz2GrauaIUvA8b4YTsWT6TpjMAwjq+zNq9LxOtS5u70D3758Pn77nb/Azr/+FAI+GdtWZ7ffOfXBnG3a2m9lCSOyahf9RCAYPyZccEpEMoB3AVwB4BiAPQC+yhjb77S/XRRlGAwnB+NgDFB1BoMxKJqOoM8DI20H/dGpBJ5/5ySuXzYXBMDrkSBh2B56kojQTCpYxT+RjEu2y4loEqqWsv83GOD1EGqqZOgsle1ivV+GwdA9lISqGVB0A4wBRyyZKQ+tbkbvkAoCIBEh4JOznH4ffumQKaq2tvPj/TEk03EQpSzcV2/bYx67eVUEjWE/JBruV3HVwGAilXEzqy6Au555B584cyou+fg03G8RxvKslrkNQcSSOrwewte373XMdhlSVRztTWSJcM+c6kdCNTKyeNzatFO/tWbciGyXwhCC08lDJQhOy2Hw8WkA32eM/X/p5b8DAMbY/3ba395BugaTOBVX0dEbw20/fQvH+uLYvCoCnyyZy5yZtQHc/qXzcN6MqY5WzF2DSXx54++zjvnxhguFdfPkp+QP9ON9MVy35eWs9vX4mmWYURt0PJ63yduuXojbn9qfdextVy/E2kfbsXlVxHX77U/tz2jD9nbudiw/xq1f3Hb1Qnx8+hRcv/Vl1/geu3EZrt+afc687JFcE7drdBr1WzH4EOSkEgYf5fDaZQaADsvysfQ6EyJaQ0R7iWhvV1dXxsGKpkOiTBfHmoA3y9URGBaujUaoJhDkw629FuJwaief8ycXbOZzK80ltsznKprLddVguV1QuZjWreyRXBM7ot+OnFzPVoGglJTD4CMvjLEtjLFmxlhzY2NjxjafR4bBMl0c++NqlqsjMCxcG41QTSDIh1t7LcTh1E4+508u2MznVppLbJnPVTSX66pEuV1QuZjWreyRXBM7ot+OnFzPVoGglJTD4OM4gFmW5ZnpdQVRH/LB7yHMqK1KCUtrA2h7/n3UhbzmMpB6GN177WLMqQ+6isqECE1QShpDPkfBaWOO9sXbpJPwdOPKJaaIend7R5bQ887lixxF1fZ27nSs9RinfsHFoj/ddwyb0n87xffc/hPYaBOAW8t2E+E2hQt/XSL6rUCQydzv/XxEP+NJOWg+PEgJTi9DatCxB8D1jLE/Oe3v9F7SMBhOJZJQVAbVYDAMhipv6r8eRTdgGAweiVDlk1ETGL0ts2BSMi7v0RVFQ9eQAi3dJhtDPvh8ue12eJs0jGGXUiKCTyYoOgNjDD6PjNqA17Qg59bodlG1vUzezq3HOrX7XPuHq2REE7opotUNbh0vQdFZVpz2slVVR2c0aV6TprAfXm9x31qcZv1WaD4EJaEEWhHXtjrhJmOMMY2IbgLw3wBkAA+5DTzckCRCbbBqTOLhbo0CQSnw+TyYkWewYaeYNlnofk5l5jo23/5TA/YjCsfrlQsWl7oh+q1AUFlM+OADABhjvwDwi4mOQyAQCAQCQekpB82HQCAQCASC04iy+OZDIBAIBALBxDISbc9IdSITLjgtFiLqAnDEZXMDgO5xDKfcEdcjk1zXo5sxdtVYV5invdqphPtVCTECkzvOiWirlXI98yHOY3xxbasVN/jIBRHtZYw1T3Qc5YK4HpmU+/Uo9/iAyogREHGONZUSZz7EeZQPQvMhEAgEAoFgXBGDD4FAIBAIBOPKZBt8bJnoAMoMcT0yKffrUe7xAZURIyDiHGsqJc58iPMoEyaV5kMgEAgEAkH5M9m++RAIBAKBQFDmTIrBBxFdRUQHiOg9IvreRMcz0RDRQ0TUSURvTXQsEw0RzSKi3xDRfiL6ExH9bbnGRER1RPQrIjqY/l1bBrHKRPQaET2VXp5HRK+k+9rjRFQWs7cRUQ0R7SKid4jobSL6dLldTyK6JX2/36L/196ZB9tRVHH4+wHZg2EVMSCBkIDIEjYhhsKoiAIqUGymQEVTIJayKJuUFCVLaaVwAYtVtkdBDJKwiMEKIMoStkBCWEIMCkEIJEgBAQKEJRz/OOeSycu9L+89kjtzH+ermro9Pd3TZ/r2nDnT3dNHmiCpb1Xrs0Yr69Z6erBqbaIztJK+6Aotb3xIWh04H9gL2AoYI2mrcqUqnTZgpa8D0KK8DxxvZlsBuwI/rkD7aCTTz4HbzWwYcHvsl82xwOzC/jjg92a2OfAqMLYUqZbnXGCKmW0JbIfLXJn6lDQYOAbYycy2xv1YfZvq1mdP0K1tLK8HK9MmukAr6YtO0/LGB/B54D9m9rSZvQtcA+xbskylYmZ3Aa+ULUcVMLP5ZjYjwm/gD6XBFZVpX+DKSHYlsF85EjqSNgL2AS6NfQFfBiZFktJlBJA0CNgduAzAzN41s4VUrD7xFaX7hSfv/sB8KlifBVpatzbQg1VrEyukVfRFV+kJxsdg4LnC/jxKfrgk1UTSEGB74IFyJVlKO5k2MLP5cWgBsEFJYtU4BzgJ+CD21wUWmtn7sV+Ve21T4CXgihgiulTSACpUn2b2PPAb4Fnc6HgNmE4167NGT9StlWkT3aHi+qJL9ATjI0lWiKSBwHXAcWb2etnyQMcymX+GVtqnaJK+AfzPzKaXJUMXWAPYAbjQzLYH3qRdF3QF6nNt/E11U+DTwAByaLRUym4TXaXK+qI79ATj43lg48L+RhGXJABI6oXftOPN7Pqy5YGGMr0oacM4viHwv7LkA0YB35L0DN7d/mV8XsVaMWwA1bnX5gHzzKzWozUJN0aqVJ97AHPN7CUzew+4Hq/jKtZnjZ6oW6vUJjpNC+iLLtMTjI8HgWExa7w3PonrppJlSipCzFO4DJhtZr8rWx7oUKabgO9F+HvAX5otWw0zO8XMNjKzIfg99Q8zOxT4J3BgJCtVxhpmtgB4TtIWEfUV4AkqVJ/4cMuukvrH/1+TsXL1WaAn6tYqtYlO0Qr6oluYWctvwN7Ak8BTwC/KlqfsDZiAjyu/h78Vji1bphLrYje8O/JRYGZse1dRJnxOxe3Av4G/A+uUXX8h72hgcoQ3A6YB/wEmAn3Kli/kGgE8FHV6I7B21eoTOB34F/A4cBXQp6r1WZC5ZXVrPT1YtTbRyetoKX3R2S1XOE2SJEmSpKn0hGGXJEmSJElaiDQ+kiRJkiRpKml8JEmSJEnSVNL4SJIkSZKkqaTxkbQ0XXWiJ+nggoOmP61q+ZIkSZLlSePjIyDpU5KukfSUpOmS/iZpeAlyHC7p013MM6T2wI61B8ZLeiw8bk6N1fQ6yv+MpPU+itwriTY6uVKkpGHAKcAoM/sccNwqlCtph6T9JJmkLcuWJakecqZK2qsQd5CkKau43DZJcyXNjO3eVVlelLnKy6g6aXx0k1j45QbgDjMbamY74g+2pq6vH54nD8eXbO4uxwIvmtk25h43x+Lfxq80Qs6VjtVxHiVpqKQpYRDeXXjYHQGcb2avRt6WWhGwBzAGmBq/SbIM5us+HAX8TlLfeAH6FfDj7pyvsHJsZzjRzEbE9oXulNcVmVZlGa1CGh/d50vAe2Z2US3CzB4Bpko6O3oQHpN0CICk0ZLukDRJ0r+ip0GSvi5pYu0ckW5yhPeUdJ+kGZIm1nojotdhnKQZuCLfCRgfVns/STtKujMevrcUluDdUdIjkh5h2Rt6QwrLJpvZHDN7J/LcGOeZJenIehXRKI2kRZJ+G+X9QtKNhWNflXRDN+t+RfwRODoMwhOACyJ+ODBc0j2S7peUvjWaRLTd3XDD9tsRt5qkC+J+uC16Dg+MY3XbcNKzMbPHgb8CJwOnAVfjumOa3GngvvBhz+3doRtnSPpCxI+O+JuAJyQNkHRz6L3Ha/q4M0g6V9JpEf6apLuizbZJukjSQ5KelPtBQtLqofsflPSopB/WkyniFhXKObGQ5/TC9c2WdEno1Vsl9Ytjm0v6e1zTDElDG52n0pS9ylmrbsAxwO/rxB8A3AasjveCPIs/3Efjniw3wo2++3BlvEakGRD5LwQOA9YD7irEnwycFuFngJMKZd4B7BThXsC9wPqxfwhweYQfBXaP8NnA4xEegfsFuA84CxhWOPc68dsPX5lx3YIM660gjQEHR1j46o41uf4EfHMl/RdDCtcyEHibpSsBzsSXJQaYjPdW9cIdfD0HrFV2W/o4bMChwGURvhfYEV9W/G9xP3wKeDXiGrbh3Hr+hjvdmwM8BvwaOCzi18JXWx0A9Af6Rvww4KEIj8YdC24a+wcAlxTOPahOeW3A3IK+GB/x/YFZ+IvmHGBoIf2UaLfD8NVT+wJHAqdGmj74irubtpcpji+K3z3xlyXF+SYDu4dOex8YEemuLdTDA8D+Ee4bctY9T9n/ZUdbV7qlks6xGzDBzJbgjn/uBHYGXgemmdk8AEkzgSFmNlU+pvlNSZOAfXA35l8EtgLukQTQGzcOavy5QflbAFsDt0W+1YH5ktbCH7R3RbqrgL0AzGympM3wBrwH8KCkkWY2GzhG0v6RZ2P8Znu5XZmN0izBnSFhZibpKuAwSVcAI4HvrqAuu8NquJvyEXWOzQMeMHfsNVfSkyHrg6tAjmRZxuCO6cAd1Y3BDe+JZvYBsEDSP+N43TbcXHGTsjCzNyX9GVgEHIzrxhPicF/gM8ALwHmSRuB6pjjXbpqZzY3wY8BvJY3DXQTc3aDYE81sUjs53pJ0BP4S+FMze6pw+Npot/+W9DSwJa4/t6313gGDcP3ybjuZiuwZ28OxPzDyPIs7IpwZ8dOBIZLWBAab2Q0h42LwXvIG56np+8qRxkf3mcVSh1Cd5Z1CeAlL6/8a4Cf43IWHzOwNuda9zcwajY+/2SBewCwzG7lMpBsfDTGzRbinzeslfQDsLWkD3BgZGTfiHfjNXzzv6A7SLA4jrMYVeJfqYvyh835HMnUHM3tdPnnsIDObGPW4rfmQ2I34Q+8K+WTZ4cDTK1uGZFkkrYN7xd1GkuHGhOG9UHWzUKcNJx8rPohNwAFmNqd4UNIvgReB7fAXjsWFwx/qRjN7UtIOuC+UsyTdbmZndEGObfAXqfZz6tr7JbGQ9Wgzu6WdrKPpWF//2swubpdnCMs/L/p1IGfd81SZnPPRff4B9Gk3x2FbYCFwSIz/rY93oU1bwbnuxF2AH4EbIgD3A6MkbR7nHqDGX9K8AawZ4TnA+pJGRr5ekj5nZguBhZJ2i3SHFuQeJWntCPfGe1z+i1vur4ZRsSWwa52yO5MGADN7AX9jORU3RD4ykibgPUJbSJonaWxc21j5XJNZwL6R/BbgZUk1b6Inmln7Xpxk5XMgcJWZbWJmQ8xsY7yb+xXggBhH3wDvnoYGbbgMwZPSuQU4Ol4ikLR9xA8C5kfvw3dwg3Y55F8BvmVmV+NDzTt0tmBJmwDHA9sDe0napXD4oGi3Q3HngHNC1h9J6hX5h0sa0Inr+4GWzucbLOmTjRKb2RvAPEn7Rfo+kvp39TxVIHs+ukkMI+wPnCPpZNzyfgb/fHMg8AhuDZ9kZgvUweeFZrZEPsn0cMJFspm9JOlwYIKkPpH0VHzMsz1twEWS3saHMw4E/iBpEP4fn4M/hL8PXB5vn7cW8g8FLowbfDXgZny4pDdwlKTZ+M11f52yp3QiTZHx+Fj+7BWk6xQd9AwtN5nUfJD0Z7ElzWMMMK5d3HXAZ/GhsCfw+TczgNfM7N3ouq7XhpOPF2fi//2jklbDjdZv4JPIr5P0XVwHNepZ2AY4O3pz3wN+1CDd2ZJOLezvgruxP8HMXoiXmjZJO8fxZ/GXyk8AR5nZYkmX4nM1ZoQufQnYr6OLM7NbJX0WuC/sq0X4nL8lHWT7DnCxpDPimg7q4DyV/aIvvdomTUXSecDDZnZZ2bIk5SNpoJktkrQursxHmdmCsuVKkkZIasPnj0xaUdqkMdnzkTQNSdPxN5Tjy5YlqQyTYz5Sb+DMNDyS5ONB9nwkSZIkSdJUcsJpkiRJkiRNJY2PJEmSJEmaShofSZIkSZI0lTQ+kiRJkiRpKml8JEmSJEnSVNL4SJIkSZKkqfwfhxY9DIgrF1IAAAAASUVORK5CYII=\n",
            "text/plain": [
              "<Figure size 540x540 with 12 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "op7uFM-70CW5",
        "colab_type": "text"
      },
      "source": [
        "---\n",
        "### Normalization\n",
        "---\n",
        "In normalization we linearly scale the entire column between 0 and 1, with 0 corresponding with the lowest value in the column, and 1 with the largest. When using scikit-learn (the most commonly used machine learning library in Python) we can use a `MinMaxScaler` to apply normalization. (It is called this as it scales our values between a minimum and maximum value.)"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "ME9djjbu0K83",
        "colab_type": "text"
      },
      "source": [
        "Instructions:\n",
        "\n",
        "* Fit the `MinMaxScaler` on the `Age` column of `so_numeric_df`.\n",
        "* Transform the same column with the scaler you just fit."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "jNgwBJRp0Zgl",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "# Import MinMaxScaler\n",
        "from sklearn.preprocessing import MinMaxScaler"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "iYeEJLVQ0HCU",
        "colab_type": "code",
        "outputId": "4ef14532-7c1a-4259-a12c-38c5c74e2e48",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 221
        }
      },
      "source": [
        "MM_scaler = MinMaxScaler() # instantiating MinMaxScaler\n",
        "MM_scaler.fit(so_numeric_df[[\"Age\"]]) # fitting MM_scaler to the Age column\n",
        "# transforming the Age column using the fitted scaler\n",
        "so_numeric_df[\"scaled_Age\"] = MM_scaler.transform(so_numeric_df[[\"Age\"]])\n",
        "# comparing the original and transformed Age columns\n",
        "print(so_numeric_df[[\"scaled_Age\", \"Age\"]].head())"
      ],
      "execution_count": 27,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "   scaled_Age  Age\n",
            "0    0.046154   21\n",
            "1    0.307692   38\n",
            "2    0.415385   45\n",
            "3    0.430769   46\n",
            "4    0.323077   39\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.6/dist-packages/ipykernel_launcher.py:4: SettingWithCopyWarning: \n",
            "A value is trying to be set on a copy of a slice from a DataFrame.\n",
            "Try using .loc[row_indexer,col_indexer] = value instead\n",
            "\n",
            "See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
            "  after removing the cwd from sys.path.\n"
          ],
          "name": "stderr"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "lBKJrNv60m2p",
        "colab_type": "text"
      },
      "source": [
        "---\n",
        "### Standardization\n",
        "---\n",
        "While normalization can be useful for scaling a column between two data points, it is hard to compare two scaled columns if even one of them is overly affected by outliers. One commonly used solution to this is called standardization, where instead of having a strict upper and lower bound, you center the data around its mean, and calculate the number of standard deviations away from mean each data point is."
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "gFyXLW3v0rU9",
        "colab_type": "text"
      },
      "source": [
        "Instructions:\n",
        "\n",
        "* Fit the `StandardScaler` on the `Age` column of `so_numeric_df`.\n",
        "* Transform the same column with the scaler we just fit."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "TL8N_V9n03jy",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "# Import StandardScaler\n",
        "from sklearn.preprocessing import StandardScaler"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "lOoRs_vR0qmJ",
        "colab_type": "code",
        "outputId": "816dd36b-0ffc-47e6-98af-f9f15bc02947",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 221
        }
      },
      "source": [
        "S_scaler = MinMaxScaler() # instantiating StandardScaler\n",
        "S_scaler.fit(so_numeric_df[[\"Age\"]]) # fitting S_scaler to the Age column\n",
        "# transforming the Age column using the fitted scaler\n",
        "so_numeric_df[\"S_scaled_Age\"] = S_scaler.transform(so_numeric_df[[\"Age\"]])\n",
        "# comparing the original and transformed Age columns\n",
        "print(so_numeric_df[[\"S_scaled_Age\", \"Age\"]].head())"
      ],
      "execution_count": 29,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "   S_scaled_Age  Age\n",
            "0      0.046154   21\n",
            "1      0.307692   38\n",
            "2      0.415385   45\n",
            "3      0.430769   46\n",
            "4      0.323077   39\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.6/dist-packages/ipykernel_launcher.py:4: SettingWithCopyWarning: \n",
            "A value is trying to be set on a copy of a slice from a DataFrame.\n",
            "Try using .loc[row_indexer,col_indexer] = value instead\n",
            "\n",
            "See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
            "  after removing the cwd from sys.path.\n"
          ],
          "name": "stderr"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "QLntVWcp1aZ1",
        "colab_type": "text"
      },
      "source": [
        "---\n",
        "### Log transformation\n",
        "---\n",
        "In the previous exercises we scaled the data linearly, which will not affect the data's shape. This works great if our data is normally distributed (or closely normally distributed), an assumption that a lot of machine learning models make. Sometimes we will work with data that closely conforms to normality, e.g the height or weight of a population. On the other hand, many variables in the real world do not follow this pattern e.g, wages or age of a population. \n",
        "\n",
        "Now, we will use a log transform on the `ConvertedSalary` column in the `so_numeric_df` DataFrame as it has a large amount of its data centered around the lower values, but contains very high values also. These distributions are said to have a long right tail."
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "NMQB-F4e1lMK",
        "colab_type": "text"
      },
      "source": [
        "Instructions:\n",
        "\n",
        "* Fit the `PowerTransformer` on the `ConvertedSalary` column of `so_numeric_df`.\n",
        "* Transform the same column with the scaler we just fit."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "N95T8tjM1h5A",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "# Import PowerTransformer\n",
        "from sklearn.preprocessing import PowerTransformer"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "9CNzVVOg1ueY",
        "colab_type": "code",
        "outputId": "c97d2d81-4799-427c-8565-78b81da8241e",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 396
        }
      },
      "source": [
        "pow_trans = PowerTransformer() # instantiating PowerTransformer\n",
        "pow_trans.fit(so_numeric_df[[\"ConvertedSalary\"]]) # fitting pow_trans to the ConvertedSalary column\n",
        "# applying the log transformation to the ConvertedSalary column \n",
        "so_numeric_df[\"log_ConvertedSalary\"] = pow_trans.transform(so_numeric_df[[\"ConvertedSalary\"]])\n",
        "# plotting the columns before and after the log transformation\n",
        "so_numeric_df[[\"ConvertedSalary\", \"log_ConvertedSalary\"]].hist()\n",
        "plt.show()"
      ],
      "execution_count": 31,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.6/dist-packages/ipykernel_launcher.py:4: SettingWithCopyWarning: \n",
            "A value is trying to be set on a copy of a slice from a DataFrame.\n",
            "Try using .loc[row_indexer,col_indexer] = value instead\n",
            "\n",
            "See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
            "  after removing the cwd from sys.path.\n"
          ],
          "name": "stderr"
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 2 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "VhlDOrtb18uG",
        "colab_type": "text"
      },
      "source": [
        "---\n",
        "### Percentage based outlier removal\n",
        "---\n",
        "One way to ensure a small portion of data is not having an overly adverse effect is by removing a certain percentage of the largest and/or smallest values in the column. This can be achieved by finding the relevant quantile and trimming the data using it with a mask. This approach is particularly useful if we are concerned that the highest values in our dataset should be avoided. When using this approach, we must remember that even if there are no outliers, this will still remove the same top N percentage from the dataset."
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "AhiOqx-v2CFB",
        "colab_type": "text"
      },
      "source": [
        "Instructions:\n",
        "\n",
        "* Find the 95th quantile of the `ConvertedSalary` column in `so_numeric_df` DataFrame.\n",
        "* Trim the `so_numeric_df` DataFrame to retain all rows where `ConvertedSalary` is less than it's 95th quantile and store this as `trimmed_df`.\n",
        "* Plot the histogram of `so_numeric_df[['ConvertedSalary']]`.\n",
        "* Plot the histogram of `trimmed_df[['ConvertedSalary']]`"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "G1_-eGVu1_JA",
        "colab_type": "code",
        "outputId": "59eac64c-f591-446f-ddac-5b7932c505ad",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 558
        }
      },
      "source": [
        "# finding the 95th quantile of the ConvertedSalary column in so_numeric_df DataFrame\n",
        "quantile = so_numeric_df[\"ConvertedSalary\"].quantile(0.95)\n",
        "# trimming the so_numeric_df DataFrame to retain all rows where ConvertedSalary is less than it's 95th quantile\n",
        "trimmed_df = so_numeric_df[so_numeric_df[\"ConvertedSalary\"] < quantile]\n",
        "so_numeric_df[[\"ConvertedSalary\"]].hist()  # plotting the histogram of so_numeric_df[['ConvertedSalary']]\n",
        "plt.show()\n",
        "trimmed_df[[\"ConvertedSalary\"]].hist()  # plotting the histogram of trimmed_df[['ConvertedSalary']]\n",
        "plt.show()"
      ],
      "execution_count": 32,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "Lgva617FNqZy",
        "colab_type": "text"
      },
      "source": [
        "## Dealing with Text Data\n",
        "---"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "IuHUqVR03PJV",
        "colab_type": "text"
      },
      "source": [
        "---\n",
        "### Cleaning up your text\n",
        "---\n",
        "Unstructured text data cannot be directly used in most analyses. Multiple steps need to be taken to go from a long free form string to a set of numeric columns in the right format that can be ingested by a machine learning model. The first step of this process is to standardize the data and eliminate any characters that could cause problems later on in your analytic pipeline.\n",
        "\n",
        "Here, we will be working with a new dataset containing the inaugural speeches of the presidents of the United States loaded as `speech_df`, with the speeches stored in the `text` column."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "dFmYNs7W3hYj",
        "colab_type": "code",
        "outputId": "5856a45c-54fd-41a1-b653-cb5eadec0896",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 204
        }
      },
      "source": [
        "speech_df = pd.read_csv('https://raw.githubusercontent.com/shala2020/shala2020.github.io/master/Lecture_Materials/Assignments/MachineLearning/L6/inaugural_speeches.csv')\n",
        "speech_df.head()"
      ],
      "execution_count": 33,
      "outputs": [
        {
          "output_type": "execute_result",
          "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>Name</th>\n",
              "      <th>Inaugural Address</th>\n",
              "      <th>Date</th>\n",
              "      <th>text</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>George Washington</td>\n",
              "      <td>First Inaugural Address</td>\n",
              "      <td>Thursday, April 30, 1789</td>\n",
              "      <td>Fellow-Citizens of the Senate and of the House...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>George Washington</td>\n",
              "      <td>Second Inaugural Address</td>\n",
              "      <td>Monday, March 4, 1793</td>\n",
              "      <td>Fellow Citizens:  I AM again called upon by th...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>John Adams</td>\n",
              "      <td>Inaugural Address</td>\n",
              "      <td>Saturday, March 4, 1797</td>\n",
              "      <td>WHEN it was first perceived, in early times, t...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>Thomas Jefferson</td>\n",
              "      <td>First Inaugural Address</td>\n",
              "      <td>Wednesday, March 4, 1801</td>\n",
              "      <td>Friends and Fellow-Citizens:  CALLED upon to u...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>Thomas Jefferson</td>\n",
              "      <td>Second Inaugural Address</td>\n",
              "      <td>Monday, March 4, 1805</td>\n",
              "      <td>PROCEEDING, fellow-citizens, to that qualifica...</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "                Name  ...                                               text\n",
              "0  George Washington  ...  Fellow-Citizens of the Senate and of the House...\n",
              "1  George Washington  ...  Fellow Citizens:  I AM again called upon by th...\n",
              "2         John Adams  ...  WHEN it was first perceived, in early times, t...\n",
              "3   Thomas Jefferson  ...  Friends and Fellow-Citizens:  CALLED upon to u...\n",
              "4   Thomas Jefferson  ...  PROCEEDING, fellow-citizens, to that qualifica...\n",
              "\n",
              "[5 rows x 4 columns]"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 33
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "qvbCog623SDx",
        "colab_type": "text"
      },
      "source": [
        "Instructions:\n",
        "\n",
        "* Print the first 5 rows of the `text` column in `speech_df` DataFrame to see the free text fields.\n",
        "* Replace all non letter characters in the `text` column with a whitespace and add it as a new column `text_clean` in the `speech_df` DataFrame. \n",
        "* Make all characters in the newly created `text_clean` column lower case.\n",
        "* Print the first 5 rows of the `text_clean` column. "
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "7kuZqMcEOvGN",
        "colab_type": "code",
        "outputId": "f7e7403d-24e1-4c53-9d08-7193a2c3a744",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 255
        }
      },
      "source": [
        "print(speech_df[\"text\"].head())  # prints the first 5 rows of the text column in speech_df DataFrame\n",
        "print(\"\\n\")\n",
        "# replaces all non letter characters in the text column with a whitespace \n",
        "speech_df[\"text_clean\"] = speech_df[\"text\"].str.replace('[^a-zA-Z]', ' ')\n",
        "speech_df[\"text_clean\"] = speech_df[\"text_clean\"].str.lower() # makes all characters lower case\n",
        "print(speech_df[\"text_clean\"].head())  # prints the first 5 rows of the text_clean column"
      ],
      "execution_count": 34,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "0    Fellow-Citizens of the Senate and of the House...\n",
            "1    Fellow Citizens:  I AM again called upon by th...\n",
            "2    WHEN it was first perceived, in early times, t...\n",
            "3    Friends and Fellow-Citizens:  CALLED upon to u...\n",
            "4    PROCEEDING, fellow-citizens, to that qualifica...\n",
            "Name: text, dtype: object\n",
            "\n",
            "\n",
            "0    fellow citizens of the senate and of the house...\n",
            "1    fellow citizens   i am again called upon by th...\n",
            "2    when it was first perceived  in early times  t...\n",
            "3    friends and fellow citizens   called upon to u...\n",
            "4    proceeding  fellow citizens  to that qualifica...\n",
            "Name: text_clean, dtype: object\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "Ag6i4ijT4tcM",
        "colab_type": "text"
      },
      "source": [
        "---\n",
        "### High level text features\n",
        "---\n",
        "Once the text has been cleaned and standardized we can begin creating features from the data. The most fundamental information we can calculate about free form text is its size, such as its length and number of words. "
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "q4jPqRmA4zRA",
        "colab_type": "text"
      },
      "source": [
        "Instructions:\n",
        "\n",
        "* Record the character length of each speech (`speech_df['text_clean']`) and store it in a new `char_count` column.\n",
        "* Record the word count of each speech in the `word_count` column.\n",
        "* Record the average word length of each speech in the `avg_word_length` column.\n",
        "* Print the first 5 rows of the columns: `text_clean`, `char_cnt`, `word_cnt`, `avg_word_length`\n"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "XhUyQjRdP3et",
        "colab_type": "code",
        "outputId": "feeb828b-4e68-4beb-c820-ec65fb134555",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 153
        }
      },
      "source": [
        "speech_df[\"char_count\"] = speech_df[\"text_clean\"].str.len() # records the character length of each speech\n",
        "speech_df[\"word_count\"] = speech_df[\"text_clean\"].str.split().str.len() # records the word count of each speech \n",
        "speech_df[\"avg_word_length\"] = speech_df[\"char_count\"] / speech_df[\"word_count\"] # records the average word length of each speech\n",
        "print(speech_df[[\"text_clean\", \"char_count\", \"word_count\", \"avg_word_length\"]].head()) # prints the first 5 rows"
      ],
      "execution_count": 35,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "                                          text_clean  ...  avg_word_length\n",
            "0  fellow citizens of the senate and of the house...  ...         6.016760\n",
            "1  fellow citizens   i am again called upon by th...  ...         5.829630\n",
            "2  when it was first perceived  in early times  t...  ...         5.971158\n",
            "3  friends and fellow citizens   called upon to u...  ...         5.843318\n",
            "4  proceeding  fellow citizens  to that qualifica...  ...         5.948363\n",
            "\n",
            "[5 rows x 4 columns]\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "Wi2hOh045bsm",
        "colab_type": "text"
      },
      "source": [
        "---\n",
        "### Counting words (I)\n",
        "---\n",
        "Once high level information has been recorded we can begin creating features based on the actual content of each text, as given below: \n",
        "\n",
        "* For each unique word in the dataset a column is created.\n",
        "* For each entry, the number of times this word occurs is counted and the count value is entered into the respective column.\n",
        "\n",
        "These \"count\" columns can then be used to train machine learning models."
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "b6sMPuHi5nB7",
        "colab_type": "text"
      },
      "source": [
        "Instructions:\n",
        "\n",
        "* Import `CountVectorizer` from `sklearn.feature_extraction.text`.\n",
        "* Instantiate `CountVectorizer` and assign it to 'cv'.\n",
        "* Fit the vectorizer to the `text_clean` column.\n",
        "* Print the feature names generated by the vectorizer and find the number of features.\n"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "nomeBUIc5lA2",
        "colab_type": "code",
        "outputId": "5ba8b5fc-f261-467f-c150-b4d9041250c2",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 51
        }
      },
      "source": [
        "from sklearn.feature_extraction.text import CountVectorizer # importing CountVectorizer\n",
        "cv = CountVectorizer() # instantiating CountVectorizer\n",
        "cv.fit(speech_df['text_clean']) # fitting the vectorizer to the text_clean column\n",
        "print(cv.get_feature_names()[:10]) # prints the feature names generated by the vectorizer\n",
        "print(len(cv.get_feature_names())) # prints the number of feature names generated by the vectorizer"
      ],
      "execution_count": 36,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "['abandon', 'abandoned', 'abandonment', 'abate', 'abdicated', 'abeyance', 'abhorring', 'abide', 'abiding', 'abilities']\n",
            "9043\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "bAm6_JHr6Cto",
        "colab_type": "text"
      },
      "source": [
        "---\n",
        "### Counting words (II)\n",
        "---\n",
        "Once the vectorizer has been fit to the data, it can be used to transform the text to an array representing the word counts. "
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "JcFMh9k_6Hu4",
        "colab_type": "text"
      },
      "source": [
        "Instructions:\n",
        "\n",
        "* Apply the vectorizer ('cv' in the previous exercise) to the `text_clean` column.\n",
        "* Convert this transformed (sparse) array into a `numpy` array with counts and print it. "
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "IQChK09eS3Ph",
        "colab_type": "code",
        "outputId": "ebf88925-c8f3-45de-e638-091193f7d2d7",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 136
        }
      },
      "source": [
        "cv1 = cv.transform(speech_df['text_clean']) # applying the vectorizer to the text_clean column\n",
        "cv2 = cv1.toarray() # Convert the array into a numpy array with counts\n",
        "print(cv2) # printing the array"
      ],
      "execution_count": 37,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "[[0 0 0 ... 0 0 0]\n",
            " [0 0 0 ... 0 0 0]\n",
            " [0 1 0 ... 0 0 0]\n",
            " ...\n",
            " [0 1 0 ... 0 0 0]\n",
            " [0 0 0 ... 0 0 0]\n",
            " [0 0 0 ... 0 0 0]]\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "vUQAwgOV73BB",
        "colab_type": "text"
      },
      "source": [
        "---\n",
        "### Limiting your features\n",
        "---\n",
        "As we have seen, using the `CountVectorizer` with its default settings creates a feature for every single word in our corpus. This can create far too many features, often including ones that will provide very little analytical value.\n",
        "\n",
        "For this purpose `CountVectorizer` has parameters that you can set to reduce the number of features:\n",
        "\n",
        "* `min_df` : Use only words that occur in more than this percentage of documents. This can be used to remove outlier words that will not generalize across texts.\n",
        "* `max_df` : Use only words that occur in less than this percentage of documents. This is useful to eliminate very common words that occur in every corpus without adding value such as \"and\" or \"the\"."
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "I9X2x50g8Dfd",
        "colab_type": "text"
      },
      "source": [
        "Instructions:\n",
        "\n",
        "* Limit the number of features in the `CountVectorizer` by setting the minimum number of documents a word can appear to 20% and the maximum to 80%.\n",
        "* Fit and apply the vectorizer on `text_clean` column in one step.\n",
        "* Convert this transformed (sparse) array into a `numpy` array with counts and print the dimensions of the new reduced array.\n",
        "* Did you notice that the number of features (unique words) greatly reduced from 9043 to 818?\n"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "RqNCtZu-8NbI",
        "colab_type": "code",
        "outputId": "b66b7a6c-c37b-435a-c449-0b1a7d8c52b3",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        }
      },
      "source": [
        "from sklearn.feature_extraction.text import CountVectorizer # importing CountVectorizer\n",
        "cv = CountVectorizer(min_df=0.2, max_df=0.8) # specifying arguments to limit the number of features generated\n",
        "cv1 = cv.fit_transform(speech_df[\"text_clean\"]) # fitting and applying the vectorizer on text_clean column in one step\n",
        "cv2 = cv1.toarray() # converting the array into a numpy array with counts\n",
        "print(cv2.shape) # prints the shape of cv2 array"
      ],
      "execution_count": 38,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "(58, 818)\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "KKZ4xX3L9IxL",
        "colab_type": "text"
      },
      "source": [
        "---\n",
        "### Text to DataFrame\n",
        "---\n",
        "Now that we have generated these count based features in an array we will need to reformat them so that they can be combined with the rest of the dataset. This can be achieved by converting the array into a pandas DataFrame, with the feature names you found earlier as the column names, and then concatenate it with the original DataFrame.\n"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "rvioGe8x9Olj",
        "colab_type": "text"
      },
      "source": [
        "Instructions:\n",
        "\n",
        "* Create a DataFrame `cv_df` containing the `cv_array` as the values and the feature names as the column names.\n",
        "* Add the prefix `Counts_` to the column names for ease of identification.\n",
        "* Concatenate this DataFrame (`cv_df`) to the original DataFrame (`speech_df`) column wise.\n"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "oyEJPkOpU1Jy",
        "colab_type": "code",
        "outputId": "95e7ed66-b8f2-424d-dc2f-b4766952d50c",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 153
        }
      },
      "source": [
        "# creating a DataFrame with the features and adding the prefix \"Counts_\" to column names\n",
        "cv_df = pd.DataFrame(cv2, columns=cv.get_feature_names()).add_prefix('Counts_') \n",
        "# concatenating cv_df to the speech_df DataFrame column wise\n",
        "speech_df = pd.concat([speech_df, cv_df], axis=1, sort=False)\n",
        "print(speech_df.head())"
      ],
      "execution_count": 39,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "                Name         Inaugural Address  ... Counts_young Counts_your\n",
            "0  George Washington   First Inaugural Address  ...            0           9\n",
            "1  George Washington  Second Inaugural Address  ...            0           1\n",
            "2         John Adams         Inaugural Address  ...            0           1\n",
            "3   Thomas Jefferson   First Inaugural Address  ...            0           7\n",
            "4   Thomas Jefferson  Second Inaugural Address  ...            0           4\n",
            "\n",
            "[5 rows x 826 columns]\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "Rpo0ek00-dsr",
        "colab_type": "text"
      },
      "source": [
        "---\n",
        "### Tf-idf\n",
        "---\n",
        "While counts of occurrences of words can be useful to build models, words that occur many times may skew the results undesirably. To limit these common words from overpowering your model a form of normalization can be used. In this lesson we will be using **Term frequency-inverse document frequency** (**Tf-idf**). Tf-idf has the effect of reducing the value of common words, while increasing the weight of words that do not occur in many documents."
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "c29xSq9Q-mcH",
        "colab_type": "text"
      },
      "source": [
        "Instructions:\n",
        "\n",
        "* Import `TfidfVectorizer` from `sklearn.feature_extraction.text`.\n",
        "* Instantiate `TfidfVectorizer` while limiting the number of features to 100 and removing English stop words.\n",
        "* Fit and apply the vectorizer on `text_clean` column in one step.\n",
        "* Create a DataFrame `tv_df` containing the weights of the words and the feature names as the column names.\n",
        "* Add the prefix `TFIDF_` to the column names for ease of identification."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "-F0vdfx9-yIq",
        "colab_type": "code",
        "outputId": "bec53163-5737-462d-e1d4-22157ed5f75a",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 153
        }
      },
      "source": [
        "from sklearn.feature_extraction.text import TfidfVectorizer # importing TfidfVectorizer\n",
        "tv = TfidfVectorizer(max_features=100, stop_words=\"english\") # instantiating TfidfVectorizer\n",
        "tv1 = tv.fit_transform(speech_df['text_clean']) # fitting and applying the vectorizer on text_clean column in one step\n",
        "tv_df = pd.DataFrame(tv1.toarray(), columns=tv.get_feature_names()).add_prefix('TFIDF_')\n",
        "print(tv_df.head()) # prints the first 5 rows of tv_df"
      ],
      "execution_count": 40,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "   TFIDF_action  TFIDF_administration  ...  TFIDF_world  TFIDF_years\n",
            "0      0.000000              0.133415  ...     0.045929     0.052694\n",
            "1      0.000000              0.261016  ...     0.000000     0.000000\n",
            "2      0.000000              0.092436  ...     0.063643     0.073018\n",
            "3      0.000000              0.092693  ...     0.095729     0.000000\n",
            "4      0.041334              0.039761  ...     0.054752     0.062817\n",
            "\n",
            "[5 rows x 100 columns]\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "qzcGgNmS_Yc5",
        "colab_type": "text"
      },
      "source": [
        "---\n",
        "### Inspecting Tf-idf values\n",
        "---\n",
        "After creating Tf-idf features we will often want to understand what are the most highest scored words for each corpus. This can be achieved by isolating the row we want to examine and then sorting the the scores from high to low."
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "MP0QjLZJ_cvX",
        "colab_type": "text"
      },
      "source": [
        "* Assign the first row of `tv_df` to `sample_row`.\n",
        "* `sample_row` is now a series of weights assigned to words. Sort these values to print the top 5 highest-rated words."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "g800sJg2_j2x",
        "colab_type": "code",
        "outputId": "cef5ba7c-c53a-4159-b936-61e9ed2ae1a6",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 119
        }
      },
      "source": [
        "sample_row = tv_df.iloc[0] # isolating the row to be examined\n",
        "print(sample_row.sort_values(ascending=False).head()) # prints the top 5 highest-rated words of sample_row"
      ],
      "execution_count": 41,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "TFIDF_government    0.367430\n",
            "TFIDF_public        0.333237\n",
            "TFIDF_present       0.315182\n",
            "TFIDF_duty          0.238637\n",
            "TFIDF_citizens      0.229644\n",
            "Name: 0, dtype: float64\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "A7WK0CHgH_3J",
        "colab_type": "text"
      },
      "source": [
        "---\n",
        "### Sentiment analysis\n",
        "---\n",
        "You have been given the tweets about US airlines. Making use of this data, your task is to predict whether a tweet contains \n",
        "\n",
        "* positive, \n",
        "* negative, or \n",
        "* neutral sentiment \n",
        "\n",
        "about the airline."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "mupsP8kXIFld",
        "colab_type": "code",
        "outputId": "84af8672-9e97-4fa1-b825-f83c9c8e05cd",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 445
        }
      },
      "source": [
        "data_source_url = \"https://raw.githubusercontent.com/shala2020/shala2020.github.io/master/Lecture_Materials/Assignments/MachineLearning/L6/Tweets.csv\"\n",
        "airline_tweets = pd.read_csv(data_source_url)\n",
        "airline_tweets.head()"
      ],
      "execution_count": 42,
      "outputs": [
        {
          "output_type": "execute_result",
          "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>tweet_id</th>\n",
              "      <th>airline_sentiment</th>\n",
              "      <th>airline_sentiment_confidence</th>\n",
              "      <th>negativereason</th>\n",
              "      <th>negativereason_confidence</th>\n",
              "      <th>airline</th>\n",
              "      <th>airline_sentiment_gold</th>\n",
              "      <th>name</th>\n",
              "      <th>negativereason_gold</th>\n",
              "      <th>retweet_count</th>\n",
              "      <th>text</th>\n",
              "      <th>tweet_coord</th>\n",
              "      <th>tweet_created</th>\n",
              "      <th>tweet_location</th>\n",
              "      <th>user_timezone</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>570306133677760513</td>\n",
              "      <td>neutral</td>\n",
              "      <td>1.0000</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>Virgin America</td>\n",
              "      <td>NaN</td>\n",
              "      <td>cairdin</td>\n",
              "      <td>NaN</td>\n",
              "      <td>0</td>\n",
              "      <td>@VirginAmerica What @dhepburn said.</td>\n",
              "      <td>NaN</td>\n",
              "      <td>2015-02-24 11:35:52 -0800</td>\n",
              "      <td>NaN</td>\n",
              "      <td>Eastern Time (US &amp; Canada)</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>570301130888122368</td>\n",
              "      <td>positive</td>\n",
              "      <td>0.3486</td>\n",
              "      <td>NaN</td>\n",
              "      <td>0.0000</td>\n",
              "      <td>Virgin America</td>\n",
              "      <td>NaN</td>\n",
              "      <td>jnardino</td>\n",
              "      <td>NaN</td>\n",
              "      <td>0</td>\n",
              "      <td>@VirginAmerica plus you've added commercials t...</td>\n",
              "      <td>NaN</td>\n",
              "      <td>2015-02-24 11:15:59 -0800</td>\n",
              "      <td>NaN</td>\n",
              "      <td>Pacific Time (US &amp; Canada)</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>570301083672813571</td>\n",
              "      <td>neutral</td>\n",
              "      <td>0.6837</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>Virgin America</td>\n",
              "      <td>NaN</td>\n",
              "      <td>yvonnalynn</td>\n",
              "      <td>NaN</td>\n",
              "      <td>0</td>\n",
              "      <td>@VirginAmerica I didn't today... Must mean I n...</td>\n",
              "      <td>NaN</td>\n",
              "      <td>2015-02-24 11:15:48 -0800</td>\n",
              "      <td>Lets Play</td>\n",
              "      <td>Central Time (US &amp; Canada)</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>570301031407624196</td>\n",
              "      <td>negative</td>\n",
              "      <td>1.0000</td>\n",
              "      <td>Bad Flight</td>\n",
              "      <td>0.7033</td>\n",
              "      <td>Virgin America</td>\n",
              "      <td>NaN</td>\n",
              "      <td>jnardino</td>\n",
              "      <td>NaN</td>\n",
              "      <td>0</td>\n",
              "      <td>@VirginAmerica it's really aggressive to blast...</td>\n",
              "      <td>NaN</td>\n",
              "      <td>2015-02-24 11:15:36 -0800</td>\n",
              "      <td>NaN</td>\n",
              "      <td>Pacific Time (US &amp; Canada)</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>570300817074462722</td>\n",
              "      <td>negative</td>\n",
              "      <td>1.0000</td>\n",
              "      <td>Can't Tell</td>\n",
              "      <td>1.0000</td>\n",
              "      <td>Virgin America</td>\n",
              "      <td>NaN</td>\n",
              "      <td>jnardino</td>\n",
              "      <td>NaN</td>\n",
              "      <td>0</td>\n",
              "      <td>@VirginAmerica and it's a really big bad thing...</td>\n",
              "      <td>NaN</td>\n",
              "      <td>2015-02-24 11:14:45 -0800</td>\n",
              "      <td>NaN</td>\n",
              "      <td>Pacific Time (US &amp; Canada)</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "             tweet_id  ...               user_timezone\n",
              "0  570306133677760513  ...  Eastern Time (US & Canada)\n",
              "1  570301130888122368  ...  Pacific Time (US & Canada)\n",
              "2  570301083672813571  ...  Central Time (US & Canada)\n",
              "3  570301031407624196  ...  Pacific Time (US & Canada)\n",
              "4  570300817074462722  ...  Pacific Time (US & Canada)\n",
              "\n",
              "[5 rows x 15 columns]"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 42
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "AYlKJmRzIHlu",
        "colab_type": "text"
      },
      "source": [
        "Instructions:\n",
        "\n",
        "* Draw a pie chart to see the number of tweets for each airline. \n",
        "* Draw a pie chart to see the distribution of sentiments across all the tweets. \n",
        "* Apply suitable data pre-processing steps to get rid of undesired symbols. This part has been done upto some extent.  \n",
        "* Using `TfidfVectorizer` class, convert text features into TF-IDF feature vectors.\n",
        "* `airline_sentiment` is the label and `text` is the feature. Apply suitable `train_test_split`, implement suitable machine learning classifier, and show the accuracy. \n",
        "\n",
        "Note: \n",
        "\n",
        "* For your convenience, the pre-processing of features have been done. \n",
        "* You are supposed to apply the vectorization on `processed_features` and implement ML algorithm. "
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "vVfLGp1fX1ks",
        "colab_type": "code",
        "outputId": "a30d78f8-790d-42e8-8f35-22169962f7e3",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 265
        }
      },
      "source": [
        "airline_tweets.airline.value_counts().plot(kind='pie', autopct='%1.0f%%')"
      ],
      "execution_count": 43,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<matplotlib.axes._subplots.AxesSubplot at 0x7f236a4c12b0>"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 43
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "Cfyjq4F6YA40",
        "colab_type": "code",
        "outputId": "45f165c0-6fc1-4215-d83e-10cc427c2014",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 265
        }
      },
      "source": [
        "airline_tweets.airline_sentiment.value_counts().plot(kind='pie', autopct='%1.0f%%', colors=[\"yellow\", \"blue\", \"red\"])"
      ],
      "execution_count": 44,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<matplotlib.axes._subplots.AxesSubplot at 0x7f236ab10e80>"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 44
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "g5W-u3BCBfRN",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "# Extracting features and label\n",
        "features = airline_tweets.iloc[:, 10].values\n",
        "labels = airline_tweets.iloc[:, 1].values"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "oeDEmJ8yBiQj",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "import re \n",
        "processed_features = []\n",
        "\n",
        "for sentence in range(0, len(features)):\n",
        "    # Remove all the special characters\n",
        "    processed_feature = re.sub(r'\\W', ' ', str(features[sentence]))\n",
        "    # remove all single characters\n",
        "    processed_feature= re.sub(r'\\s+[a-zA-Z]\\s+', ' ', processed_feature)\n",
        "    # Remove single characters from the start\n",
        "    processed_feature = re.sub(r'\\^[a-zA-Z]\\s+', ' ', processed_feature) \n",
        "    # Substituting multiple spaces with single space\n",
        "    processed_feature = re.sub(r'\\s+', ' ', processed_feature, flags=re.I)\n",
        "    # Removing prefixed 'b'\n",
        "    processed_feature = re.sub(r'^b\\s+', '', processed_feature)\n",
        "    # Converting to Lowercase\n",
        "    processed_feature = processed_feature.lower()\n",
        "    processed_features.append(processed_feature)"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "RHT6Vt8CZm2M",
        "colab_type": "code",
        "outputId": "72973472-5b42-4ccc-f0d2-fe5f5e2fe8d9",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 51
        }
      },
      "source": [
        "import nltk \n",
        "nltk.download('stopwords')\n",
        "from nltk.corpus import stopwords\n",
        "from sklearn.feature_extraction.text import TfidfVectorizer\n",
        "tf_vectorizer = TfidfVectorizer (max_features=2500, min_df=6, max_df=0.9, stop_words=stopwords.words('english'))\n",
        "processed_features = tf_vectorizer.fit_transform(processed_features).toarray()"
      ],
      "execution_count": 47,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "[nltk_data] Downloading package stopwords to /root/nltk_data...\n",
            "[nltk_data]   Unzipping corpora/stopwords.zip.\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "NSqWryj9Y7xt",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "from sklearn.model_selection import train_test_split\n",
        "X_train, X_test, y_train, y_test = train_test_split(processed_features,labels,test_size=0.2,random_state=1)"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "V_HxdhNCZHGD",
        "colab_type": "code",
        "outputId": "9fb4c94d-820f-4c77-9236-bd12058f56c0",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 187
        }
      },
      "source": [
        "from sklearn.ensemble import BaggingClassifier\n",
        "from sklearn.neighbors import KNeighborsClassifier\n",
        "\n",
        "text_classifier = BaggingClassifier(KNeighborsClassifier(), max_samples=0.5, max_features=0.5, random_state=1)\n",
        "text_classifier.fit(X_train, y_train)"
      ],
      "execution_count": 60,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "BaggingClassifier(base_estimator=KNeighborsClassifier(algorithm='auto',\n",
              "                                                      leaf_size=30,\n",
              "                                                      metric='minkowski',\n",
              "                                                      metric_params=None,\n",
              "                                                      n_jobs=None,\n",
              "                                                      n_neighbors=5, p=2,\n",
              "                                                      weights='uniform'),\n",
              "                  bootstrap=True, bootstrap_features=False, max_features=0.5,\n",
              "                  max_samples=0.5, n_estimators=10, n_jobs=None,\n",
              "                  oob_score=False, random_state=1, verbose=0, warm_start=False)"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 60
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "5JeI2d4HZZpN",
        "colab_type": "code",
        "outputId": "86d52019-23e1-401d-b497-7bbeef21e5ef",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 374
        }
      },
      "source": [
        "y_pred = text_classifier.predict(X_test)\n",
        "from sklearn.metrics import classification_report, confusion_matrix, accuracy_score\n",
        "print(confusion_matrix(y_test,y_pred))\n",
        "print(\"\\n\")\n",
        "print(classification_report(y_test,y_pred))\n",
        "print(\"\\n\")\n",
        "print(accuracy_score(y_test, y_pred))\n",
        "print(\"\\n\")\n",
        "print(len(y_pred[y_pred=='negative']))"
      ],
      "execution_count": 61,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "[[1708  103   15]\n",
            " [ 357  224   30]\n",
            " [ 242   71  178]]\n",
            "\n",
            "\n",
            "              precision    recall  f1-score   support\n",
            "\n",
            "    negative       0.74      0.94      0.83      1826\n",
            "     neutral       0.56      0.37      0.44       611\n",
            "    positive       0.80      0.36      0.50       491\n",
            "\n",
            "    accuracy                           0.72      2928\n",
            "   macro avg       0.70      0.55      0.59      2928\n",
            "weighted avg       0.71      0.72      0.69      2928\n",
            "\n",
            "\n",
            "\n",
            "0.7206284153005464\n",
            "\n",
            "\n",
            "2307\n"
          ],
          "name": "stdout"
        }
      ]
    }
  ]
}