notebook.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Group 15 Notebook"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Tanzania_waterFall_Predictions"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Objective of the study"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Tanzania, as a developing country, struggles with providing clean water to its population of over 57,000,000. There are many water points already established in the country, but some are in need of repair while others have failed altogether.\n",
    "\n",
    "Through this analysis we are going to answer the following questions:\n",
    "\n",
    "1. To identify which factors contribute to functionality of the well:\n",
    "\n",
    "-  Does source of water influence the functionality of the wells?\n",
    "\n",
    "-  Does Age effect the condition of the wells?\n",
    "\n",
    "-  Does having Public Meetings help to keep the wells functional?\n",
    "\n",
    "2. What is the best mode of payment for maintainence of wells?\n",
    "\n",
    "3. To analyse the dataset using various classification methods and identify the best among them"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**End User:** Government Ministry of water & NGO's"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Importing libraries"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "import pandas as pd\n",
    "import matplotlib.pyplot as plt\n",
    "import seaborn as sns\n",
    "from sklearn.metrics import accuracy_score, classification_report, confusion_matrix\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Read and understanding  the Dataset "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "df1 = pd.read_csv('d.1.csv') # the target class dataset\n",
    "df3 = pd.read_csv('d.3.csv') # all our features dataset"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>id</th>\n",
       "      <th>status_group</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>69572</td>\n",
       "      <td>functional</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>8776</td>\n",
       "      <td>functional</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>34310</td>\n",
       "      <td>functional</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>67743</td>\n",
       "      <td>non functional</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>19728</td>\n",
       "      <td>functional</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "      id    status_group\n",
       "0  69572      functional\n",
       "1   8776      functional\n",
       "2  34310      functional\n",
       "3  67743  non functional\n",
       "4  19728      functional"
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df1.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>id</th>\n",
       "      <th>amount_tsh</th>\n",
       "      <th>date_recorded</th>\n",
       "      <th>funder</th>\n",
       "      <th>gps_height</th>\n",
       "      <th>installer</th>\n",
       "      <th>longitude</th>\n",
       "      <th>latitude</th>\n",
       "      <th>wpt_name</th>\n",
       "      <th>num_private</th>\n",
       "      <th>...</th>\n",
       "      <th>payment_type</th>\n",
       "      <th>water_quality</th>\n",
       "      <th>quality_group</th>\n",
       "      <th>quantity</th>\n",
       "      <th>quantity_group</th>\n",
       "      <th>source</th>\n",
       "      <th>source_type</th>\n",
       "      <th>source_class</th>\n",
       "      <th>waterpoint_type</th>\n",
       "      <th>waterpoint_type_group</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>69572</td>\n",
       "      <td>6000.0</td>\n",
       "      <td>2011-03-14</td>\n",
       "      <td>Roman</td>\n",
       "      <td>1390</td>\n",
       "      <td>Roman</td>\n",
       "      <td>34.938093</td>\n",
       "      <td>-9.856322</td>\n",
       "      <td>none</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>annually</td>\n",
       "      <td>soft</td>\n",
       "      <td>good</td>\n",
       "      <td>enough</td>\n",
       "      <td>enough</td>\n",
       "      <td>spring</td>\n",
       "      <td>spring</td>\n",
       "      <td>groundwater</td>\n",
       "      <td>communal standpipe</td>\n",
       "      <td>communal standpipe</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>8776</td>\n",
       "      <td>0.0</td>\n",
       "      <td>2013-03-06</td>\n",
       "      <td>Grumeti</td>\n",
       "      <td>1399</td>\n",
       "      <td>GRUMETI</td>\n",
       "      <td>34.698766</td>\n",
       "      <td>-2.147466</td>\n",
       "      <td>Zahanati</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>never pay</td>\n",
       "      <td>soft</td>\n",
       "      <td>good</td>\n",
       "      <td>insufficient</td>\n",
       "      <td>insufficient</td>\n",
       "      <td>rainwater harvesting</td>\n",
       "      <td>rainwater harvesting</td>\n",
       "      <td>surface</td>\n",
       "      <td>communal standpipe</td>\n",
       "      <td>communal standpipe</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>34310</td>\n",
       "      <td>25.0</td>\n",
       "      <td>2013-02-25</td>\n",
       "      <td>Lottery Club</td>\n",
       "      <td>686</td>\n",
       "      <td>World vision</td>\n",
       "      <td>37.460664</td>\n",
       "      <td>-3.821329</td>\n",
       "      <td>Kwa Mahundi</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>per bucket</td>\n",
       "      <td>soft</td>\n",
       "      <td>good</td>\n",
       "      <td>enough</td>\n",
       "      <td>enough</td>\n",
       "      <td>dam</td>\n",
       "      <td>dam</td>\n",
       "      <td>surface</td>\n",
       "      <td>communal standpipe multiple</td>\n",
       "      <td>communal standpipe</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>67743</td>\n",
       "      <td>0.0</td>\n",
       "      <td>2013-01-28</td>\n",
       "      <td>Unicef</td>\n",
       "      <td>263</td>\n",
       "      <td>UNICEF</td>\n",
       "      <td>38.486161</td>\n",
       "      <td>-11.155298</td>\n",
       "      <td>Zahanati Ya Nanyumbu</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>never pay</td>\n",
       "      <td>soft</td>\n",
       "      <td>good</td>\n",
       "      <td>dry</td>\n",
       "      <td>dry</td>\n",
       "      <td>machine dbh</td>\n",
       "      <td>borehole</td>\n",
       "      <td>groundwater</td>\n",
       "      <td>communal standpipe multiple</td>\n",
       "      <td>communal standpipe</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>19728</td>\n",
       "      <td>0.0</td>\n",
       "      <td>2011-07-13</td>\n",
       "      <td>Action In A</td>\n",
       "      <td>0</td>\n",
       "      <td>Artisan</td>\n",
       "      <td>31.130847</td>\n",
       "      <td>-1.825359</td>\n",
       "      <td>Shuleni</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>never pay</td>\n",
       "      <td>soft</td>\n",
       "      <td>good</td>\n",
       "      <td>seasonal</td>\n",
       "      <td>seasonal</td>\n",
       "      <td>rainwater harvesting</td>\n",
       "      <td>rainwater harvesting</td>\n",
       "      <td>surface</td>\n",
       "      <td>communal standpipe</td>\n",
       "      <td>communal standpipe</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 40 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "      id  amount_tsh date_recorded        funder  gps_height     installer  \\\n",
       "0  69572      6000.0    2011-03-14         Roman        1390         Roman   \n",
       "1   8776         0.0    2013-03-06       Grumeti        1399       GRUMETI   \n",
       "2  34310        25.0    2013-02-25  Lottery Club         686  World vision   \n",
       "3  67743         0.0    2013-01-28        Unicef         263        UNICEF   \n",
       "4  19728         0.0    2011-07-13   Action In A           0       Artisan   \n",
       "\n",
       "   longitude   latitude              wpt_name  num_private  ... payment_type  \\\n",
       "0  34.938093  -9.856322                  none            0  ...     annually   \n",
       "1  34.698766  -2.147466              Zahanati            0  ...    never pay   \n",
       "2  37.460664  -3.821329           Kwa Mahundi            0  ...   per bucket   \n",
       "3  38.486161 -11.155298  Zahanati Ya Nanyumbu            0  ...    never pay   \n",
       "4  31.130847  -1.825359               Shuleni            0  ...    never pay   \n",
       "\n",
       "  water_quality quality_group      quantity  quantity_group  \\\n",
       "0          soft          good        enough          enough   \n",
       "1          soft          good  insufficient    insufficient   \n",
       "2          soft          good        enough          enough   \n",
       "3          soft          good           dry             dry   \n",
       "4          soft          good      seasonal        seasonal   \n",
       "\n",
       "                 source           source_type  source_class  \\\n",
       "0                spring                spring   groundwater   \n",
       "1  rainwater harvesting  rainwater harvesting       surface   \n",
       "2                   dam                   dam       surface   \n",
       "3           machine dbh              borehole   groundwater   \n",
       "4  rainwater harvesting  rainwater harvesting       surface   \n",
       "\n",
       "               waterpoint_type waterpoint_type_group  \n",
       "0           communal standpipe    communal standpipe  \n",
       "1           communal standpipe    communal standpipe  \n",
       "2  communal standpipe multiple    communal standpipe  \n",
       "3  communal standpipe multiple    communal standpipe  \n",
       "4           communal standpipe    communal standpipe  \n",
       "\n",
       "[5 rows x 40 columns]"
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df3.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'pandas.core.frame.DataFrame'>\n",
      "Int64Index: 59400 entries, 0 to 59399\n",
      "Data columns (total 41 columns):\n",
      " #   Column                 Non-Null Count  Dtype  \n",
      "---  ------                 --------------  -----  \n",
      " 0   id                     59400 non-null  int64  \n",
      " 1   status_group           59400 non-null  object \n",
      " 2   amount_tsh             59400 non-null  float64\n",
      " 3   date_recorded          59400 non-null  object \n",
      " 4   funder                 55765 non-null  object \n",
      " 5   gps_height             59400 non-null  int64  \n",
      " 6   installer              55745 non-null  object \n",
      " 7   longitude              59400 non-null  float64\n",
      " 8   latitude               59400 non-null  float64\n",
      " 9   wpt_name               59400 non-null  object \n",
      " 10  num_private            59400 non-null  int64  \n",
      " 11  basin                  59400 non-null  object \n",
      " 12  subvillage             59029 non-null  object \n",
      " 13  region                 59400 non-null  object \n",
      " 14  region_code            59400 non-null  int64  \n",
      " 15  district_code          59400 non-null  int64  \n",
      " 16  lga                    59400 non-null  object \n",
      " 17  ward                   59400 non-null  object \n",
      " 18  population             59400 non-null  int64  \n",
      " 19  public_meeting         56066 non-null  object \n",
      " 20  recorded_by            59400 non-null  object \n",
      " 21  scheme_management      55523 non-null  object \n",
      " 22  scheme_name            31234 non-null  object \n",
      " 23  permit                 56344 non-null  object \n",
      " 24  construction_year      59400 non-null  int64  \n",
      " 25  extraction_type        59400 non-null  object \n",
      " 26  extraction_type_group  59400 non-null  object \n",
      " 27  extraction_type_class  59400 non-null  object \n",
      " 28  management             59400 non-null  object \n",
      " 29  management_group       59400 non-null  object \n",
      " 30  payment                59400 non-null  object \n",
      " 31  payment_type           59400 non-null  object \n",
      " 32  water_quality          59400 non-null  object \n",
      " 33  quality_group          59400 non-null  object \n",
      " 34  quantity               59400 non-null  object \n",
      " 35  quantity_group         59400 non-null  object \n",
      " 36  source                 59400 non-null  object \n",
      " 37  source_type            59400 non-null  object \n",
      " 38  source_class           59400 non-null  object \n",
      " 39  waterpoint_type        59400 non-null  object \n",
      " 40  waterpoint_type_group  59400 non-null  object \n",
      "dtypes: float64(3), int64(7), object(31)\n",
      "memory usage: 19.0+ MB\n"
     ]
    }
   ],
   "source": [
    "# merging target data frame with our main dataframe\n",
    "data = pd.merge(df1, df3, on='id', how='inner')\n",
    "data.info()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The target feature in this dataset is the 'status_group' feature. This feature is an indicator of whether or not the water well is 'functional', 'functional needs repairs', or 'non functional'.\n",
    "\n",
    "Also this are some of the features we are going to use:\n",
    "\n",
    ".amount_tsh: Amount water available to waterpoint\n",
    "\n",
    ".date_recorded: The date the row was entered\n",
    "\n",
    ".funder: Who funded the well\n",
    "\n",
    ".gps_height: Altitude of the well\n",
    "\n",
    ".installer: Organization that installed the well\n",
    "\n",
    ".longitude: GPS coordinate\n",
    "\n",
    ".latitude: GPS coordinate\n",
    "\n",
    ".wpt_name: Name of the waterpoint if there is one\n",
    "\n",
    "Among others."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "id                           0\n",
       "status_group                 0\n",
       "amount_tsh                   0\n",
       "date_recorded                0\n",
       "funder                    3635\n",
       "gps_height                   0\n",
       "installer                 3655\n",
       "longitude                    0\n",
       "latitude                     0\n",
       "wpt_name                     0\n",
       "num_private                  0\n",
       "basin                        0\n",
       "subvillage                 371\n",
       "region                       0\n",
       "region_code                  0\n",
       "district_code                0\n",
       "lga                          0\n",
       "ward                         0\n",
       "population                   0\n",
       "public_meeting            3334\n",
       "recorded_by                  0\n",
       "scheme_management         3877\n",
       "scheme_name              28166\n",
       "permit                    3056\n",
       "construction_year            0\n",
       "extraction_type              0\n",
       "extraction_type_group        0\n",
       "extraction_type_class        0\n",
       "management                   0\n",
       "management_group             0\n",
       "payment                      0\n",
       "payment_type                 0\n",
       "water_quality                0\n",
       "quality_group                0\n",
       "quantity                     0\n",
       "quantity_group               0\n",
       "source                       0\n",
       "source_type                  0\n",
       "source_class                 0\n",
       "waterpoint_type              0\n",
       "waterpoint_type_group        0\n",
       "dtype: int64"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# checking missing values\n",
    "data.isna().sum()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Visualizing our Target compared to other features"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0    32259\n",
       "2    22824\n",
       "1     4317\n",
       "Name: target, dtype: int64"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Convert the 'status_group' column to a categorical data type\n",
    "data['status_group'] = data['status_group'].astype('category')\n",
    "\n",
    "# Convert the categorical values into numerical codes\n",
    "data['target'] = data['status_group'].cat.codes\n",
    "\n",
    "# Display the counts of each numerical code in the 'target' column\n",
    "data['target'].value_counts()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 576x360 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Plotting the value counts of the 'status_group' column in descending order using a bar plot\n",
    "data['status_group'].value_counts(ascending=False).plot(kind='bar', figsize=(8,5), color=['darkgreen','midnightblue','purple'])\n",
    "\n",
    "# Adding a title to the plot with a larger font size and bold font weight\n",
    "plt.title(\"Water Pump Status\", fontweight='bold', fontsize=18)\n",
    "\n",
    "# Adding a label to the x-axis with a larger font size, bold font weight, and descriptive text\n",
    "plt.xlabel('Status of the Water Pump', fontweight='bold', fontsize=10)\n",
    "\n",
    "# Adding a label to the y-axis with a larger font size, bold font weight, and descriptive text\n",
    "plt.ylabel('Value Counts', fontweight='bold', fontsize=10)\n",
    "\n",
    "# Adding gridlines to the plot along the y-axis with a dashed linestyle and reduced opacity\n",
    "plt.grid(axis='y', linestyle='--', alpha=0.6)\n",
    "\n",
    "# Displaying the plot\n",
    "plt.show()\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We see that there is a lot of imbalance among our classes, hence we will combine \"functional needs repair\" and \"non functional\" into one for modelling purpose. Then We will convert our target variable to 0 and 1, so now we can drop status group while modelling.\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "functional      32259\n",
       "needs repair    27141\n",
       "Name: status_group, dtype: int64"
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Replace \"functional needs repair\" and \"non functional\" with \"needs repair\" in the 'status_group' column\n",
    "data['status_group'] = data['status_group'].replace(to_replace=[\"functional needs repair\", \"non functional\"], value=\"needs repair\")\n",
    "\n",
    "# Display the value counts of each category in the 'status_group' column after replacement\n",
    "data['status_group'].value_counts()\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 576x360 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Plotting the value counts of the 'status_group' column in descending order using a bar plot\n",
    "data['status_group'].value_counts(ascending=False).plot(kind='bar', figsize=(8,5), color=['darkgreen','midnightblue','purple'])\n",
    "\n",
    "# Adding a title to the plot with a larger font size and bold font weight\n",
    "plt.title(\"Water Pump Status\", fontweight='bold', fontsize=18)\n",
    "\n",
    "# Adding a label to the x-axis with a larger font size, bold font weight, and descriptive text\n",
    "plt.xlabel('Status of the Water Pump', fontweight='bold', fontsize=10)\n",
    "\n",
    "# Adding a label to the y-axis with a larger font size, bold font weight, and descriptive text\n",
    "plt.ylabel('Value Counts', fontweight='bold', fontsize=10)\n",
    "\n",
    "# Adding gridlines to the plot along the y-axis with a dashed linestyle and reduced opacity\n",
    "plt.grid(axis='y', linestyle='--', alpha=0.6)\n",
    "\n",
    "# Displaying the plot\n",
    "plt.show()\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0    32259\n",
       "1    27141\n",
       "Name: target, dtype: int64"
      ]
     },
     "execution_count": 11,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Convert the 'status_group' column to a categorical data type\n",
    "data['status_group'] = data['status_group'].astype('category')\n",
    "\n",
    "# Convert the categorical values into numerical codes\n",
    "data['target'] = data['status_group'].cat.codes\n",
    "\n",
    "# Display the counts of each numerical code in the 'target' column\n",
    "data['target'].value_counts()\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>id</th>\n",
       "      <th>status_group</th>\n",
       "      <th>amount_tsh</th>\n",
       "      <th>date_recorded</th>\n",
       "      <th>funder</th>\n",
       "      <th>gps_height</th>\n",
       "      <th>installer</th>\n",
       "      <th>longitude</th>\n",
       "      <th>latitude</th>\n",
       "      <th>wpt_name</th>\n",
       "      <th>...</th>\n",
       "      <th>water_quality</th>\n",
       "      <th>quality_group</th>\n",
       "      <th>quantity</th>\n",
       "      <th>quantity_group</th>\n",
       "      <th>source</th>\n",
       "      <th>source_type</th>\n",
       "      <th>source_class</th>\n",
       "      <th>waterpoint_type</th>\n",
       "      <th>waterpoint_type_group</th>\n",
       "      <th>target</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>69572</td>\n",
       "      <td>functional</td>\n",
       "      <td>6000.0</td>\n",
       "      <td>2011-03-14</td>\n",
       "      <td>Roman</td>\n",
       "      <td>1390</td>\n",
       "      <td>Roman</td>\n",
       "      <td>34.938093</td>\n",
       "      <td>-9.856322</td>\n",
       "      <td>none</td>\n",
       "      <td>...</td>\n",
       "      <td>soft</td>\n",
       "      <td>good</td>\n",
       "      <td>enough</td>\n",
       "      <td>enough</td>\n",
       "      <td>spring</td>\n",
       "      <td>spring</td>\n",
       "      <td>groundwater</td>\n",
       "      <td>communal standpipe</td>\n",
       "      <td>communal standpipe</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>8776</td>\n",
       "      <td>functional</td>\n",
       "      <td>0.0</td>\n",
       "      <td>2013-03-06</td>\n",
       "      <td>Grumeti</td>\n",
       "      <td>1399</td>\n",
       "      <td>GRUMETI</td>\n",
       "      <td>34.698766</td>\n",
       "      <td>-2.147466</td>\n",
       "      <td>Zahanati</td>\n",
       "      <td>...</td>\n",
       "      <td>soft</td>\n",
       "      <td>good</td>\n",
       "      <td>insufficient</td>\n",
       "      <td>insufficient</td>\n",
       "      <td>rainwater harvesting</td>\n",
       "      <td>rainwater harvesting</td>\n",
       "      <td>surface</td>\n",
       "      <td>communal standpipe</td>\n",
       "      <td>communal standpipe</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>34310</td>\n",
       "      <td>functional</td>\n",
       "      <td>25.0</td>\n",
       "      <td>2013-02-25</td>\n",
       "      <td>Lottery Club</td>\n",
       "      <td>686</td>\n",
       "      <td>World vision</td>\n",
       "      <td>37.460664</td>\n",
       "      <td>-3.821329</td>\n",
       "      <td>Kwa Mahundi</td>\n",
       "      <td>...</td>\n",
       "      <td>soft</td>\n",
       "      <td>good</td>\n",
       "      <td>enough</td>\n",
       "      <td>enough</td>\n",
       "      <td>dam</td>\n",
       "      <td>dam</td>\n",
       "      <td>surface</td>\n",
       "      <td>communal standpipe multiple</td>\n",
       "      <td>communal standpipe</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>67743</td>\n",
       "      <td>needs repair</td>\n",
       "      <td>0.0</td>\n",
       "      <td>2013-01-28</td>\n",
       "      <td>Unicef</td>\n",
       "      <td>263</td>\n",
       "      <td>UNICEF</td>\n",
       "      <td>38.486161</td>\n",
       "      <td>-11.155298</td>\n",
       "      <td>Zahanati Ya Nanyumbu</td>\n",
       "      <td>...</td>\n",
       "      <td>soft</td>\n",
       "      <td>good</td>\n",
       "      <td>dry</td>\n",
       "      <td>dry</td>\n",
       "      <td>machine dbh</td>\n",
       "      <td>borehole</td>\n",
       "      <td>groundwater</td>\n",
       "      <td>communal standpipe multiple</td>\n",
       "      <td>communal standpipe</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>19728</td>\n",
       "      <td>functional</td>\n",
       "      <td>0.0</td>\n",
       "      <td>2011-07-13</td>\n",
       "      <td>Action In A</td>\n",
       "      <td>0</td>\n",
       "      <td>Artisan</td>\n",
       "      <td>31.130847</td>\n",
       "      <td>-1.825359</td>\n",
       "      <td>Shuleni</td>\n",
       "      <td>...</td>\n",
       "      <td>soft</td>\n",
       "      <td>good</td>\n",
       "      <td>seasonal</td>\n",
       "      <td>seasonal</td>\n",
       "      <td>rainwater harvesting</td>\n",
       "      <td>rainwater harvesting</td>\n",
       "      <td>surface</td>\n",
       "      <td>communal standpipe</td>\n",
       "      <td>communal standpipe</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 42 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "      id  status_group  amount_tsh date_recorded        funder  gps_height  \\\n",
       "0  69572    functional      6000.0    2011-03-14         Roman        1390   \n",
       "1   8776    functional         0.0    2013-03-06       Grumeti        1399   \n",
       "2  34310    functional        25.0    2013-02-25  Lottery Club         686   \n",
       "3  67743  needs repair         0.0    2013-01-28        Unicef         263   \n",
       "4  19728    functional         0.0    2011-07-13   Action In A           0   \n",
       "\n",
       "      installer  longitude   latitude              wpt_name  ...  \\\n",
       "0         Roman  34.938093  -9.856322                  none  ...   \n",
       "1       GRUMETI  34.698766  -2.147466              Zahanati  ...   \n",
       "2  World vision  37.460664  -3.821329           Kwa Mahundi  ...   \n",
       "3        UNICEF  38.486161 -11.155298  Zahanati Ya Nanyumbu  ...   \n",
       "4       Artisan  31.130847  -1.825359               Shuleni  ...   \n",
       "\n",
       "   water_quality quality_group      quantity quantity_group  \\\n",
       "0           soft          good        enough         enough   \n",
       "1           soft          good  insufficient   insufficient   \n",
       "2           soft          good        enough         enough   \n",
       "3           soft          good           dry            dry   \n",
       "4           soft          good      seasonal       seasonal   \n",
       "\n",
       "                 source           source_type source_class  \\\n",
       "0                spring                spring  groundwater   \n",
       "1  rainwater harvesting  rainwater harvesting      surface   \n",
       "2                   dam                   dam      surface   \n",
       "3           machine dbh              borehole  groundwater   \n",
       "4  rainwater harvesting  rainwater harvesting      surface   \n",
       "\n",
       "               waterpoint_type  waterpoint_type_group target  \n",
       "0           communal standpipe     communal standpipe      0  \n",
       "1           communal standpipe     communal standpipe      0  \n",
       "2  communal standpipe multiple     communal standpipe      0  \n",
       "3  communal standpipe multiple     communal standpipe      1  \n",
       "4           communal standpipe     communal standpipe      0  \n",
       "\n",
       "[5 rows x 42 columns]"
      ]
     },
     "execution_count": 12,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "functional      0.543081\n",
       "needs repair    0.456919\n",
       "Name: status_group, dtype: float64"
      ]
     },
     "execution_count": 13,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Calculate the relative frequency of each unique value in the 'status_group' column\n",
    "data['status_group'].value_counts(normalize=True)\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We are going to use these numbers as a baseline when comparing subgroups. For example,if a region has less than 54% functionality, we know they are below average and some features within that region are effecting the functionality of the wells.\n",
    "\n",
    "This will help us identify important features more easily, and give us references for further data exploration.\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'pandas.core.frame.DataFrame'>\n",
      "Int64Index: 59400 entries, 0 to 59399\n",
      "Data columns (total 44 columns):\n",
      " #   Column                 Non-Null Count  Dtype   \n",
      "---  ------                 --------------  -----   \n",
      " 0   id                     59400 non-null  int64   \n",
      " 1   status_group           59400 non-null  category\n",
      " 2   amount_tsh             59400 non-null  float64 \n",
      " 3   date_recorded          59400 non-null  object  \n",
      " 4   funder                 55765 non-null  object  \n",
      " 5   gps_height             59400 non-null  int64   \n",
      " 6   installer              55745 non-null  object  \n",
      " 7   longitude              59400 non-null  float64 \n",
      " 8   latitude               59400 non-null  float64 \n",
      " 9   wpt_name               59400 non-null  object  \n",
      " 10  num_private            59400 non-null  int64   \n",
      " 11  basin                  59400 non-null  object  \n",
      " 12  subvillage             59029 non-null  object  \n",
      " 13  region                 59400 non-null  object  \n",
      " 14  region_code            59400 non-null  int64   \n",
      " 15  district_code          59400 non-null  int64   \n",
      " 16  lga                    59400 non-null  object  \n",
      " 17  ward                   59400 non-null  object  \n",
      " 18  population             59400 non-null  int64   \n",
      " 19  public_meeting         56066 non-null  object  \n",
      " 20  recorded_by            59400 non-null  object  \n",
      " 21  scheme_management      55523 non-null  object  \n",
      " 22  scheme_name            31234 non-null  object  \n",
      " 23  permit                 56344 non-null  object  \n",
      " 24  construction_year      59400 non-null  int64   \n",
      " 25  extraction_type        59400 non-null  object  \n",
      " 26  extraction_type_group  59400 non-null  object  \n",
      " 27  extraction_type_class  59400 non-null  object  \n",
      " 28  management             59400 non-null  object  \n",
      " 29  management_group       59400 non-null  object  \n",
      " 30  payment                59400 non-null  object  \n",
      " 31  payment_type           59400 non-null  object  \n",
      " 32  water_quality          59400 non-null  object  \n",
      " 33  quality_group          59400 non-null  object  \n",
      " 34  quantity               59400 non-null  object  \n",
      " 35  quantity_group         59400 non-null  object  \n",
      " 36  source                 59400 non-null  object  \n",
      " 37  source_type            59400 non-null  object  \n",
      " 38  source_class           59400 non-null  object  \n",
      " 39  waterpoint_type        59400 non-null  object  \n",
      " 40  waterpoint_type_group  59400 non-null  object  \n",
      " 41  target                 59400 non-null  int8    \n",
      " 42  functional             59400 non-null  uint8   \n",
      " 43  needs repair           59400 non-null  uint8   \n",
      "dtypes: category(1), float64(3), int64(7), int8(1), object(30), uint8(2)\n",
      "memory usage: 21.3+ MB\n"
     ]
    }
   ],
   "source": [
    "# create dummies for status group to make plotting easier\n",
    "dummies_status = pd.get_dummies(data['status_group'])\n",
    "data = data.join(dummies_status)\n",
    "\n",
    "data.info()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Data  Exploration \n",
    "#### Plotting functions for categorical variables\n",
    " \n",
    " we run a couple of categorical features which recommend what values/columns to maintain or drop"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Define a function named plot_percent with one parameter: col (column name)\n",
    "def plot_percent(col):\n",
    "    \"\"\"Plot the mean functionalities of pumps compared to 'col'\n",
    "    \n",
    "    Args:\n",
    "        col (str): column name\n",
    "    \"\"\"\n",
    "    # Determine the layout of subplots based on the number of unique values in the specified column\n",
    "    if data[col].nunique() > 4:\n",
    "        rows = 2\n",
    "        cols = 1\n",
    "        width = 12 if data[col].nunique() > 12 else data[col].nunique()\n",
    "        height = 14\n",
    "    else:\n",
    "        rows = 1\n",
    "        cols = 2\n",
    "        width = 12\n",
    "        height = 3\n",
    "    \n",
    "    # Create a new figure with subplots based on the determined layout\n",
    "    fig, ax = plt.subplots(rows, cols, figsize=(width, height))\n",
    "        \n",
    "    # Iterate over each unique status in the 'status_group' column\n",
    "    for idx, status in enumerate(data['status_group'].unique().tolist()):\n",
    "        # Plot the mean value of the specified status for each category in the specified column\n",
    "        data.groupby(col).mean()[status].sort_values().plot.bar(ax=ax[idx])\n",
    "        # Add a horizontal line representing the overall mean of the specified status\n",
    "        ax[idx].axhline(y=data[status].mean(), color='r', linestyle='-')\n",
    "        # Set title, y-axis label, and adjust the y-axis limits for each subplot\n",
    "        ax[idx].set_title(f\"{status.title()} pumps by {col.title()}\", fontsize=14)\n",
    "        ax[idx].set_ylabel('Percent')\n",
    "        ax[idx].set_xlabel('')\n",
    "        ax[idx].set_ylim(0,data.groupby(col).mean()[status].max()*1.2)\n",
    "    \n",
    "    # Adjust the layout of subplots to prevent overlap\n",
    "    fig.tight_layout()\n",
    "    "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Unique Values: 1897\n",
      "Missing Values: 3635/59400\n",
      "Possible Missing Values: 777 (wells with '0' as the value)\n",
      "Unique Values: 2145\n",
      "Missing Values: 3655/59400\n",
      "Possible Missing Values: 777 (wells with '0' as the value)\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "DWE                   17402\n",
       "Government             1825\n",
       "RWE                    1206\n",
       "Commu                  1060\n",
       "DANIDA                 1050\n",
       "KKKT                    898\n",
       "Hesawa                  840\n",
       "0                       777\n",
       "TCRS                    707\n",
       "Central government      622\n",
       "CES                     610\n",
       "Community               553\n",
       "DANID                   552\n",
       "District Council        551\n",
       "HESAWA                  539\n",
       "World vision            408\n",
       "LGA                     408\n",
       "WEDECO                  397\n",
       "TASAF                   396\n",
       "District council        392\n",
       "Name: installer, dtype: int64"
      ]
     },
     "execution_count": 16,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Analyzing 'funder' variable\n",
    "print(f\"Unique Values: {data['funder'].nunique()}\")  # Counting the number of unique values in the 'funder' column\n",
    "print(f\"Missing Values: {data['funder'].isna().sum()}/{data.shape[0]}\")  # Counting missing values in the 'funder' column\n",
    "print(f\"Possible Missing Values: {(data['funder'] == '0').sum()} (wells with '0' as the value)\")  # Counting occurrences of '0' as a possible placeholder for missing values\n",
    "data['funder'].value_counts().head(20)  # Displaying the top 20 most frequent values in the 'funder' column\n",
    "\n",
    "# Analyzing 'installer' variable\n",
    "print(f\"Unique Values: {data['installer'].nunique()}\")  # Counting the number of unique values in the 'installer' column\n",
    "print(f\"Missing Values: {data['installer'].isna().sum()}/{data.shape[0]}\")  # Counting missing values in the 'installer' column\n",
    "print(f\"Possible Missing Values: {(data['installer'] == '0').sum()} (wells with '0' as the value)\")  # Counting occurrences of '0' as a possible placeholder for missing values\n",
    "data['installer'].value_counts().head(20)  # Displaying the top 20 most frequent values in the 'installer' column\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We see that DWE (District Water Engineering) and Govt are the major installers which is similar to our funders.\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Basins: 9\n",
      "Region: 21\n",
      "Districts: 20\n",
      "LGAs: 125\n",
      "Wards: 2092\n",
      "Subvillages: 19287\n"
     ]
    }
   ],
   "source": [
    "# Analyzing the 'basin' category\n",
    "print(f\"Basins: {data['basin'].nunique()}\")  # Counting the number of unique values in the 'basin' column\n",
    "print(f\"Region: {data['region'].nunique()}\")  # Counting the number of unique values in the 'region' column\n",
    "print(f\"Districts: {data['district_code'].nunique()}\")  # Counting the number of unique values in the 'district_code' column\n",
    "print(f\"LGAs: {data['lga'].nunique()}\")  # Counting the number of unique values in the 'lga' (Local Government Authority) column\n",
    "print(f\"Wards: {data['ward'].nunique()}\")  # Counting the number of unique values in the 'ward' column\n",
    "print(f\"Subvillages: {data['subvillage'].nunique()}\")  # Counting the number of unique values in the 'subvillage' column\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead tr th {\n",
       "        text-align: left;\n",
       "    }\n",
       "\n",
       "    .dataframe thead tr:last-of-type th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th colspan=\"2\" halign=\"left\">functional</th>\n",
       "      <th colspan=\"2\" halign=\"left\">needs repair</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th>mean</th>\n",
       "      <th>sum</th>\n",
       "      <th>mean</th>\n",
       "      <th>sum</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>basin</th>\n",
       "      <th>region</th>\n",
       "      <th>region_code</th>\n",
       "      <th>district_code</th>\n",
       "      <th>lga</th>\n",
       "      <th>ward</th>\n",
       "      <th>subvillage</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th rowspan=\"5\" valign=\"top\">Internal</th>\n",
       "      <th rowspan=\"5\" valign=\"top\">Arusha</th>\n",
       "      <th rowspan=\"5\" valign=\"top\">2</th>\n",
       "      <th rowspan=\"5\" valign=\"top\">1</th>\n",
       "      <th rowspan=\"5\" valign=\"top\">Monduli</th>\n",
       "      <th rowspan=\"5\" valign=\"top\">Engaruka</th>\n",
       "      <th>Hyhh</th>\n",
       "      <td>0.0</td>\n",
       "      <td>0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Madukani</th>\n",
       "      <td>0.0</td>\n",
       "      <td>0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Mkaoo</th>\n",
       "      <td>0.0</td>\n",
       "      <td>0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Mula</th>\n",
       "      <td>0.0</td>\n",
       "      <td>0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Mwembeni</th>\n",
       "      <td>0.0</td>\n",
       "      <td>0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                                                      functional  \\\n",
       "                                                                            mean   \n",
       "basin    region region_code district_code lga     ward     subvillage              \n",
       "Internal Arusha 2           1             Monduli Engaruka Hyhh              0.0   \n",
       "                                                           Madukani          0.0   \n",
       "                                                           Mkaoo             0.0   \n",
       "                                                           Mula              0.0   \n",
       "                                                           Mwembeni          0.0   \n",
       "\n",
       "                                                                           \\\n",
       "                                                                      sum   \n",
       "basin    region region_code district_code lga     ward     subvillage       \n",
       "Internal Arusha 2           1             Monduli Engaruka Hyhh         0   \n",
       "                                                           Madukani     0   \n",
       "                                                           Mkaoo        0   \n",
       "                                                           Mula         0   \n",
       "                                                           Mwembeni     0   \n",
       "\n",
       "                                                                      needs repair  \\\n",
       "                                                                              mean   \n",
       "basin    region region_code district_code lga     ward     subvillage                \n",
       "Internal Arusha 2           1             Monduli Engaruka Hyhh                1.0   \n",
       "                                                           Madukani            1.0   \n",
       "                                                           Mkaoo               1.0   \n",
       "                                                           Mula                1.0   \n",
       "                                                           Mwembeni            1.0   \n",
       "\n",
       "                                                                           \n",
       "                                                                      sum  \n",
       "basin    region region_code district_code lga     ward     subvillage      \n",
       "Internal Arusha 2           1             Monduli Engaruka Hyhh         1  \n",
       "                                                           Madukani     1  \n",
       "                                                           Mkaoo        3  \n",
       "                                                           Mula         1  \n",
       "                                                           Mwembeni     1  "
      ]
     },
     "execution_count": 18,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Grouping the data by multiple categorical variables and performing aggregation on target variables\n",
    "data.groupby(['basin', 'region', 'region_code', 'district_code', 'lga', 'ward', 'subvillage'])[\n",
    "    ['functional', 'needs repair']].agg(['mean','sum']).head()\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "For this group, the mean and sum of the 'functional' variable are 0.0 and 0, respectively, indicating that none of the water pumps in this group are functional. Similarly, the mean and sum of the 'needs repair' variable are 1.0 and 1, respectively, indicating that all water pumps in this group need repair."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 576x360 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Grouping the data by 'basin' and 'status_group', and then unstacking the results to make it suitable for plotting\n",
    "sub_df = data.groupby('basin')['status_group'].value_counts(ascending=True).unstack()\n",
    "\n",
    "# Plotting the DataFrame using a bar plot\n",
    "sub_df.plot(kind='bar', figsize=(8,5))\n",
    "\n",
    "# Adding labels to the x-axis, y-axis, and title\n",
    "plt.xlabel(\"Water Source (Basin)\", fontweight='bold')\n",
    "plt.xticks(rotation=45, fontweight='bold')\n",
    "plt.ylabel(\"Pump Functionality\")\n",
    "plt.title('Pump Functionality based on Basin', fontsize=14, fontweight='bold')\n",
    "\n",
    "# Adding a legend with a custom location and title\n",
    "plt.legend(bbox_to_anchor=(1.0, 0.9), title='Water Quantity');\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "we noted that reliable basins are Lake nyasa, pangani and Rufiji "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x360 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Grouping the data by 'basin' and calculating the mean of the 'target' variable (encoded status_group)\n",
    "source_df = data.groupby('basin')['target'].mean()\n",
    "\n",
    "# Sorting the values and plotting the DataFrame using a bar plot with a specified color and figsize\n",
    "source_df.sort_values().plot(kind='bar', color=['brown'], figsize=(10,5))\n",
    "\n",
    "# Adding labels to the x-axis, y-axis, and a title\n",
    "plt.xlabel(\"Quality of Water\")\n",
    "plt.xticks(rotation=90)  # Rotating x-axis labels for better readability\n",
    "plt.axhline(y=data['target'].mean(), color='gray')  # Adding a horizontal line representing the overall mean likelihood of needing repairs\n",
    "plt.ylabel(\"Likelihood of Needing Repairs\")\n",
    "plt.title('Likelihood of Needing Repairs vs. Basin (Water Source)', fontsize=14, fontweight='bold');\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Lake Nyasa, Pangani, Rufiji seem to have more than 60% of their pumps functioning with a less likelihood of needing repaires."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAm0AAAGECAYAAABkqBdpAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjMuMSwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy/d3fzzAAAACXBIWXMAAAsTAAALEwEAmpwYAABwC0lEQVR4nO2dd7gVxfnHP6+gYAMLahBUSGKhSBE0CBZsgIqKGmPvigVr1IjGRNRoTGJFo8bYjcafXWMsUSOxYAkgFkQUIyqKCgiIgAjc+f3xznL2Hs6999x7T+Fwv5/n2efsztndeXd2duadd96ZsRACQgghhBBi+WalcgsghBBCCCHqRkqbEEIIIUQFIKVNCCGEEKICkNImhBBCCFEBSGkTQgghhKgApLQJIYQQQlQAUtqEEEXHzO4ws2BmU8osx6gox6hUWIjbiCLGu1w8fz6UIj2WZyrpXYmmh5Q20eRIVdzJtsTMPjezf5hZ33LLlw9m1iHrGaptZZSrpgr/I+B14M3SS1Unr8dtahLQ1BWX5YEc3+liM/vSzB4ws45FjHp5zquiidO83AIIUUZ+wAvmFkBXYDAwyMz6hRDeKKtk9eNzUgrH8kgI4RLgknLLkYsQQp9yyyBqJflO1wE2BX4OdMK/2YKzPOdVIWRpE02ZaSGEPiGEnsCQGNYcOARyd5OY2VGpln+H7PPM7LD4O9fMbjCzlmZ2mZl9Y2ZfmNlvU/dKW8vOMrO/m9l3Zva1mV1oZpbnc9wSn2PpFu8/ItvyZmb9U3H2zz7PzHYys3FmtiD+VlNozKyXmT1qZjPMbKGZfRJl7Z9l4bswnXY1pGWz+NwT4r2+NbN/m9kuNch7lJk9YWbzzexjMzs2dd7qUa6PzWxevN+HZnaxma1SW+KlrWq1PYeZnRD3F5jZ2qnrL47hn5tZnWWqmQ0xs/fN7Hsze9nMuqT+G2RmL8U88ENMkxfNbFDWPc40s/fis34b0/D2rHMOMrPX4jnzYtr2yzpnezMbH2V5M/v/Op6jq5k9HPPCDzHtrzCzNVLnLO2ONrNhqW/jCTP7UZ5RJd/pZsDdMayLma2bimdNM7sqyvCDmU0zs5vMbK3UOWZmvzWzr6IMd5vZ6VbL95y6tmB5VYhGEULQpq1JbcAoIABTUmGDY1gArolhd+Q476jUeR2yzlsIfAdMTp0zAZgDfJoK2y1e1yEV9j1uLfsiFXZyLc+QvnZEDeeMSM5JhfVPXdc/+7wox/vAouTZgebxvL7xGQNu/XgX+Dqm51bAa6n7TI3Hj9SSlrekzp8MzIj7S4Ddc8j7A/BxTM/kvC3ieW1i2Je4Veaz1HV/yvHuR6XClqZjbc8BrA7MjuHDUte/F8Mur+V9Jc//PTA/5oskjT8FWsbzzo7P+REwDpgbz1kEdI/n7JWS7714r3nA4lR8Z2WlbZL/fgC2jedskLr/gnivOanrcuareG2n1LXfxWuXxONXgJWy0vuHGMcHqfvf04Dv9O4YNhtYOYatAowl8w2+FWUKMTw57+RU3NPiu/0uFZb9PRclr2rT1pit7AJo01bqLVUZLMQr5DdTFegi4GfxvFyF91G1FPIB6IdbsNOVZAdgjVhpLa3cqa54vYBb+VbB/WkC8Ektz5C+Nnt7NJ4zIglLXZeuWPpnnwecGsNOS4UlitG/yVSYnWKYAT1T989Z4WenJfBjoCqGXR/D1iRTqY/NIe+DMb5uqbAT43mrAJ2z4kwq+M9yvPtRtclcy3NcG8PHxeNOqXM71fK+0nlk1xg2JBV2dOq9rpW6bm3g23jOJTEsUcieS53XHNgh7q9GRhm5LIatBDwTw56NYRfF4ypgqxh2XE3PnvU8d8Zz5gGbxLATU9fulZXeS4AeMezhGPZlPb/TSfF4JrBv6rwjyHy7XWLYJsDiGH5oDEu+yf/iLhErAy+mZM7+nouSV7Vpa8ym7lHRlFkF+BlesE4H/gnsGEJ4vYH3mxVCeCWEUIVXEADvhhCmhBC+w61S4BaObB4OISwOIfwAPBrDNjazVnnE+zkZZ/rX8cqtoSTdT++lwhJ5fxZ/HwkhTATXBkMIbzYgnt54pQZwb7zXXOCJGNbDzJplXfO3EEKoQbYlwGFm9kHsvgrAYfG/DRsgX03cgFfAPc2sB+5fBfBGkiZ18E0I4bm4/xiukEDGP2sV4I7YPboE+AZXECDzHM/gjYFdYtfkaGAkrrQAdMGtggDnxbRYAgyIYUmX95bx98MQwri4/395PAPA1vH3lRDCJ3H/3tT/vbPOfyeEMD7uJ+9v/TzjSr7TzVLXv5L6P8mXzYF34/NOAZL80yd+RxvF40dCCAtDCItw5aouCp1XhWgwGoggmjKfhBA61PJ/iL/pArl1Led/m9pfnCMsuV9dvmrp/0ONZ2W4JYQwIkf40mvNrFkIYQm1y08IYXbcXZwKzte3rqHk84zgFj5CCIst4+6X7AwHzov7n+DdpO2BdhTQdzeEMMnM/g3sAhwLbB//uqOBt8xO2ydwZ/vFwDt4d2pPXHFpFmV4N/rBHRL/6w6cBAw190FM3/N9vIuu2mM0UNZc1OvdRZK8lW+++gS3du2GK7rbAbfi3cTp+yzCu5Sz+SpLzsY8fyHyqhANRpY2IWomsYytHx2dVwL2KVJc+5pZczNbGdg7hn0aW/QN5evU/k+SeBpxv8QCOcTMEqsHZtY9dc6C+Ls6tZP4IAEcGu+zJu5bCDA+Kpn5kliPPoiKeF/ct6mh1PYcf46/R+MK00Lgvjzvu46Z7Rz398KVMXAL0bq4wgbw2xBCD+AgshQFM9sUN3JeHELYF+iM+8k1A3bEfQ3nx9P/DfQNmQEqRwEXJnHG359GqyHAAXk+x3/jbz8z2yTuH5L6f0ye98mLEEJVCOEZMmk/2My2jfvJSO/mwBmpZ90O7wL+W/yOEuv33ma2cvzWEktpbRQ6rwrRYKS0CVEzz8ffVfAW/Hi8UiwG2+COy1PIdPf8sZH3fAH3xQH4t5m9DhzeiPtdgHfLrYUrGe+Y2Ze4n1fC+/H3NDP7r5ldlutGIYSPgNvi4TAzm4w//6ZR5gvqKdvb8XczM/sYr6AbM5VHbc/xOD7QIVHoHg8hzMrzvguBJ8zsXTJdc1OBv+NdocnULReZ2Tt4vstWCHYEJpuPRh6HD1pYLf73dghhPq6sgDvffx5Hhn4NTCSjXN2A+6StBIw2swnAdXk+x+W439zqwIR4baJQjcZdDYrBlXgeBDg//v4d/zaN+BxmNhG3MD6J+wkC/CH+9sG/s4+BXnVFWIS8KkSDkdImRA2EEP6FF8hfAG3xirxYBfSvccfrVvjItEvwSrXBhBDeB4biFdS6wCy8Em/o/UbjFqzH8G7fzXElZFTqtNPwbj1wX6DNqJkTgHNwv5+NcOfwF4ABIYSn6ineZbhz/Gw8De+jcelX43NEq8pfUufeWY/7fokrTc1x681oYI8QwvfRB2p/3Iq1BLecHYr7W6Z5E3fmX4gPhFgzhh0bQng2yvjHeO1reHpshqfNnfhISEIIXwJ74gpvM1wBycuSHP33tsVH1S6M9/8MV6oGRr/OghNC+IKM3+VgM+seQliIDwK4Cs/rmwLr4aNqf0fGongjbmWcjrsJvAT8PnX7BdRMIfOqEA3GvJwQQpSaOC/Ux/Hw6BDCHeWTRtQHM9sPeAifOmIjdY8t/5hZa3xqla/icTPgKdxXbhrQLqhCFMs5GogghBB5YmbbA8OAXWPQVVLYKoaOwOtm9l/cmt2dTNfp+VLYRCWg7lEhhMifnwAH4g3eG4CryyuOqAfT8a78zfBu4bWA54A9ZeUWlYK6R4UQQgghKgBZ2oQQQgghKgApbUIIIYQQFcAKOxChTZs2oUOHDuUWQwghhBCiTsaOHTsjhLBebeessEpbhw4dGDOmoJNyCyGEEEIUBTP7pK5z1D0qhBBCCFEBSGkTQgghhKgApLQJIYQQQlQAK6xPmxBCCCEyLFq0iKlTp/L999+XW5QmTcuWLWnfvj0rr7xyva+V0iaEEEI0AaZOncqaa65Jhw4dMLNyi9MkCSEwc+ZMpk6dSseOHet9vbpHhRBCiCbA999/z7rrriuFrYyYGeuuu26DrZ1S2oQQQogmghS28tOYdyClTQghhBDVOPPMM7nmmmuWHg8cOJDjjjtu6fFZZ53FVVddVeP1d9xxB1988UWj5Xj00Ufp1q0bW2yxBV27duXBBx9s9D1zcc011zB//vylx3vssQezZ89m9uzZ3HDDDUWJsyFIaRNCCCFENfr27cvo0aMBqKqqYsaMGUyYMGHp/6NHj6Zfv341Xt8QpW3x4sXVjt966y3OPvtsHnvsMd5//33+8Y9/cO655zJ27Nh63TcfspW2J598krXWWktKmxBCCCGWb/r167dUaZswYQJdu3ZlzTXXZNasWSxcuJCJEyfSs2dPLr74Yrbeemu6du3K0KFDCSHw4IMPMmbMGA499FB69OjBggULGDt2LDvuuCO9evVi4MCBTJs2DYD+/ftz/vnns+OOO3LttddWk+GKK67g/PPPX+qw37FjR84//3yuvPLKpdcmKx/NmDGDZOnKKVOmsP3227PVVlux1VZbLX2OUaNG0b9/f37+85+zxRZbcOihhxJCYOTIkXzxxRfstNNO7LTTToCvqjRjxgyGDx/ORx99RI8ePTjnnHM4/PDDeeyxx5bKeOihh/L4448X6S3kIISwQm69evUKQgghhHDee++9ep2/ySabhE8++STcdNNN4cYbbwwXXHBB+Oc//xlefvnlsP3224cQQpg5c+bS8w877LDw+OOPhxBC2HHHHcN///vfEEIIP/zwQ9h2223D119/HUII4b777gtHH3300vNOOumknPH37NkzjB8/vlrY+PHjQ/fu3ZeJY/r06WGTTTYJIYQwb968sGDBghBCCB988EFI9IEXXnghtGrVKnz22WdhyZIloU+fPuGll15a+qzTp0+v9uzTp08PH3/8cejSpcvS8FGjRoV99tknhBDC7NmzQ4cOHcKiRYvyTdKl5HoXwJhQh26jKT9Evegw/J81/jfl8j1LKIkQQohikljbRo8ezS9/+Us+//xzRo8eTevWrenbty8AL7zwAn/84x+ZP38+33zzDV26dGGvvfaqdp9Jkybx7rvvsttuuwGwZMkS2rZtu/T/Aw88MGf8IYRlnPZdt6mdRYsWccoppzB+/HiaNWvGBx98sPS/bbbZhvbt2wPQo0cPpkyZwnbbbZdHajg77rgjw4YN4+uvv+bhhx9m//33p3nz0qlSUtqEEEIIsQyJX9s777xD165d2Wijjbjyyitp1aoVxxxzDN9//z0nn3wyY8aMYaONNmLEiBE5p7IIIdClSxdeffXVnPGsvvrqOcO7dOnCmDFj6Nat29KwcePG0bt3bwCaN29OVVUVQLV4r776ajbYYAPeeustqqqqaNmy5dL/WrRosXS/WbNmy/jR5cPhhx/OPffcw3333cdtt91W7+sbg3zahBBCCLEM/fr144knnmCdddahWbNmrLPOOsyePZtXX32Vbbfddqmi1KZNG7777rtqIzvXXHNN5s6dC8Dmm2/O9OnTlyptixYtqjaooSbOPvtsfv/73zNlyhTAfdWuueYazjnnHMD9zpJBCem458yZQ9u2bVlppZW4++67WbJkSZ1xpeWtK/yoo45aOrK2S5cudd67kEhpE0IIIcQybLnllsyYMYM+ffpUC2vdujVt2rRhrbXW4vjjj2fLLbdkyJAhbL311kvPO+qoozjxxBPp0aMHS5Ys4cEHH+Tcc8+le/fu9OjRY+nggNro0aMHf/jDH9hrr73YbLPN2GyzzbjxxhvZfPPNAVfqbrzxRvr27cuMGTOWXnfyySdz55130qdPHz744IMaLXlphg4dyu677750IELCuuuuS79+/ejatetSZXGDDTagU6dOHH300XXet9BYPv3DlUjv3r1DMqpEFA75tAkhRGUyceJEOnXqVG4xGszw4cN5/fXXeeaZZ1hllVXKJsf8+fPZcsstGTduHK1bt27QPXK9CzMbG0LoXdt1Rbe0mVkzM3vTzJ6Ix+uY2bNm9mH8XTt17nlmNtnMJpnZwFR4LzN7J/430jSlsxBCCNGkuPzyy3nhhRfKqrA999xzbLHFFpx66qkNVtgaQym6R08HJqaOhwPPhxA2BZ6Px5hZZ+AgoAswCLjBzJrFa24EhgKbxm1QCeQWQgghhFjKrrvuyqeffsoZZ5xRlviLqrSZWXtgT+CWVPA+wJ1x/05gSCr8vhDCwhDCx8BkYBszawu0CiG8GucxuSt1jRBCCCFEk6DYlrZrgF8BVamwDUII0wDi7/oxvB3wWeq8qTGsXdzPDl8GMxtqZmPMbMz06dML8gBCCCGEEMsDRVPazGww8HUIId9FwnL5qYVawpcNDOHmEELvEELv9dZbL89ohRBCCCGWf4o5uW4/YG8z2wNoCbQys78BX5lZ2xDCtNj1+XU8fyqwUer69sAXMbx9jnAhhBBCiCZD0SxtIYTzQgjtQwgd8AEG/w4hHAY8DhwZTzsSSFZefRw4yMxamFlHfMDBG7ELda6Z9YmjRo9IXSOEEEKICmHkyJF06tSJQw89tCD3mzJlCvfee+/S4zFjxnDaaacV5N5pkgXky005lrG6HLjfzI4FPgUOAAghTDCz+4H3gMXAsBBCMo3xScAdwKrAU3ETQgghRAOpbd7NhpDPXJ033HADTz31FB07dixMnFFpO+SQQwDo3bv30mWuVkRKsiJCCGFUCGFw3J8ZQtglhLBp/P0mdd6lIYSfhBA2DyE8lQofE0LoGv87JayoMwILIYQQKygnnngi//vf/9h7771p3bo1V1xxxdL/unbtypQpU5gyZQqdOnXi+OOPp0uXLgwYMIAFCxYAMHnyZHbddVe6d+/OVlttxUcffcTw4cN56aWX6NGjB1dffTWjRo1i8ODBAHzzzTcMGTKEbt260adPH95++20ARowYwTHHHEP//v358Y9/zMiRI5fKMWTIEHr16kWXLl24+eabS5g6+aFlrIQQQghRdG666SY23HBDXnjhBc4888waz/vwww8ZNmwYEyZMYK211uKhhx4C4NBDD2XYsGG89dZbjB49mrZt23L55Zez/fbbM378+GXueeGFF9KzZ0/efvttLrvsMo444oil/73//vs888wzvPHGG1x00UUsWrQIgNtuu42xY8cyZswYRo4cycyZM4uQEg2nHN2jQgghhBA56dixIz169ACgV69eTJkyhblz5/L555+z7777AtCyZcs67/Pyyy8vVfh23nlnZs6cyZw5cwDYc889adGiBS1atGD99dfnq6++on379owcOZJHHnkEgM8++4wPP/yQddddtwhP2TCktAkhhBCipDRv3pyqqswUrt9///3S/RYtWizdb9asGQsWLKAhXlG5rklWwcyOY/HixYwaNYrnnnuOV199ldVWW43+/ftXk2t5QN2jQgghhCgpHTp0YNy4cQCMGzeOjz/+uNbzW7VqRfv27Xn00UcBWLhwIfPnz2fNNddk7ty5Oa/ZYYcduOeeewAYNWoUbdq0oVWrVjXGMWfOHNZee21WW2013n//fV577bUGPFlxkdImhBBCiJKy//77880339CjRw9uvPFGNttsszqvufvuuxk5ciTdunWjb9++fPnll3Tr1o3mzZvTvXt3rr766mrnjxgxgjFjxtCtWzeGDx/OnXfeWcOdnUGDBrF48WK6devGb37zG/r06dOoZywGtqIOxOzdu3cYM2ZMucVY4ahtiHg+w72FEEKUh4kTJ9KpU6dyiyHI/S7MbGwIodb5SmRpE0IIIYSoAKS0CSGEEEJUAFLahBBCCCEqACltQgghhBAVgJQ2IYQQQogKQEqbEEIIIUQFIKVNCCGEEBVNhw4dmDFjRlllGDNmDKeddlpR49AyVkIIIURTZETrAt9vTmHvVwIWL15M8+aFUYV69+5N797LTrNWyDhkaRNCCCFE0ZkyZQqdOnXi+OOPp0uXLgwYMIAFCxYA8NFHHzFo0CB69erF9ttvz/vvvw/A9OnT2X///dl6663ZeuuteeWVVwCYOXMmAwYMoGfPnpxwwglL1xmdN28ee+65J927d6dr16783//93zJy9O/fn/PPP58dd9yRa6+9lrFjx7LjjjvSq1cvBg4cyLRp05aed8YZZ9C3b1+6du3KG2+8AcAbb7xB37596dmzJ3379mXSpEmAL5U1ePBgwFdjGDp0KAMGDOCII44oWBpKaRNCCCFESfjwww8ZNmwYEyZMYK211uKhhx4CYOjQoVx33XWMHTuWK664gpNPPhmA008/nTPPPJP//ve/PPTQQxx33HEAXHTRRWy33Xa8+eab7L333nz66acAPP3002y44Ya89dZbvPvuuwwaNCinHLNnz+Y///kPp512GqeeeioPPvggY8eO5ZhjjuHXv/710vPmzZvH6NGjueGGGzjmmGMA2GKLLXjxxRd58803ufjiizn//PNzxjF27Fgee+wx7r333sIkHuoeFUIIIUSJ6NixIz169ACgV69eTJkyhe+++47Ro0dzwAEHLD1v4cKFADz33HO89957S8O//fZb5s6dy4svvsjDDz8MwJ577snaa68NwJZbbsnZZ5/Nueeey+DBg9l+++1zynHggQcCMGnSJN5991122203AJYsWULbtm2XnnfwwQcDvvj8t99+y+zZs5k7dy5HHnkkH374IWbGokWLcsax9957s+qqq9Y7jWpDSpsQQgghSkKLFi2W7jdr1owFCxZQVVXFWmutxfjx45c5v6qqildffTWn8mNmy4RtttlmjB07lieffJLzzjuPAQMG8Nvf/naZ81ZffXUAQgh06dKFV199Nae82XGYGb/5zW/YaaedeOSRR5gyZQr9+/fPeW0SRyFR96gQQgghykarVq3o2LEjDzzwAOCK1FtvvQXAgAEDuP7665eemyh2O+ywA/fccw8ATz31FLNmzQLgiy++YLXVVuOwww7j7LPPZty4cbXGvfnmmzN9+vSlStuiRYuYMGHC0v8Tn7iXX36Z1q1b07p1a+bMmUO7du0AuOOOOxr59PVDSpsQQgghyso999zDrbfeSvfu3enSpQuPPfYYACNHjmTMmDF069aNzp07c9NNNwFw4YUX8uKLL7LVVlvxr3/9i4033hiAd955h2222YYePXpw6aWXcsEFF9Qa7yqrrMKDDz7IueeeS/fu3enRowejR49e+v/aa69N3759OfHEE7n11lsB+NWvfsV5551Hv379WLJkSTGSo0YsGXGxotG7d+8wZsyYcouxwtFh+D9r/G/K5XuWUBIhhBD1YeLEiXTq1KncYlQM/fv354orrsg5jUdjyfUuzGxsCKHWyGRpE0IIIYSoAIqmtJlZSzN7w8zeMrMJZnZRDB9hZp+b2fi47ZG65jwzm2xmk8xsYCq8l5m9E/8babm8D4UQQgghCsSoUaOKYmVrDMUcPboQ2DmE8J2ZrQy8bGZPxf+uDiFckT7ZzDoDBwFdgA2B58xssxDCEuBGYCjwGvAkMAh4CiGEEEKIJkLRLG3B+S4erhy32hzo9gHuCyEsDCF8DEwGtjGztkCrEMKrwR3w7gKGFEtuIYQQYkVlRfVjryQa8w6K6tNmZs3MbDzwNfBsCOH1+NcpZva2md1mZmvHsHbAZ6nLp8awdnE/O1wIIYQQedKyZUtmzpwpxa2MhBCYOXMmLVu2bND1RZ1cN3Zt9jCztYBHzKwr3tV5CW51uwS4EjgGyOWnFmoJXwYzG4p3oy4d/iuEEEIIaN++PVOnTmX69OnlFqVJ07JlS9q3b9+ga0uyIkIIYbaZjQIGpX3ZzOyvwBPxcCqwUeqy9sAXMbx9jvBc8dwM3Aw+5Ueh5BdCCCEqnZVXXpmOHTuWWwzRCIo5enS9aGHDzFYFdgXejz5qCfsC78b9x4GDzKyFmXUENgXeCCFMA+aaWZ84avQI4LFiyS2EEEIIsTxSTEtbW+BOM2uGK4f3hxCeMLO7zawH3sU5BTgBIIQwwczuB94DFgPDYvcqwEnAHcCq+KhRjRwVQgghRJOiaEpbCOFtoGeO8MNrueZS4NIc4WOArgUVUAghhBCigtCKCEIIIYQQFYCUNiGEEEKICkBKmxBCCCFEBSClTQghhBCiApDSJoQQQghRAUhpE0IIIYSoAKS0CSGEEEJUAFLahBBCCCEqACltQgghhBAVgJQ2IYQQQogKQEqbEEIIIUQFIKVNCCGEEKICkNImhBBCCFEB1Km0mdnpZtbKnFvNbJyZDSiFcEIIIYQQwsnH0nZMCOFbYACwHnA0cHlRpRJCCCGEENXIR2mz+LsHcHsI4a1UmBBCCCGEKAH5KG1jzexfuNL2jJmtCVQVVywhhBBCCJGmeR7nHAv0AP4XQphvZuviXaRCCCGEEKJE5GNpezaEMC6EMBsghDATuLqoUgkhhBBCiGrUaGkzs5bAakAbM1ubjB9bK2DDEsgmhBBCCCEitXWPngCcgStoY8kobd8Cfy6uWEIIIYQQIk2NSlsI4VrgWjM7NYRwXQllEkIIIYQQWdTp0xZCuM7M+prZIWZ2RLLVdZ2ZtTSzN8zsLTObYGYXxfB1zOxZM/sw/q6duuY8M5tsZpPMbGAqvJeZvRP/G2lmmnJECCGEEE2KfFZEuBu4AtgO2DpuvfO490Jg5xBCd3z06SAz6wMMB54PIWwKPB+PMbPOwEFAF2AQcIOZNYv3uhEYCmwat0F5Pp8QQgghxApBPlN+9AY6hxBCfW4cz/8uHq4ctwDsA/SP4XcCo4BzY/h9IYSFwMdmNhnYxsymAK1CCK8CmNldwBDgqfrII4QQQghRyeQz5ce7wI8acnMza2Zm44Gv8alDXgc2CCFMA4i/68fT2wGfpS6fGsPaxf3scCGEEEKIJkM+lrY2wHtm9gbe5QlACGHvui4MISwBepjZWsAjZta1ltNz+amFWsKXvYHZULwblY033rgu8YQQQgghKoZ8lLYRjY0khDDbzEbhvmhfmVnbEMI0M2uLW+HALWgbpS5rD3wRw9vnCM8Vz83AzQC9e/euV3euEEIIIcTyTD6jR/+Ta6vrOjNbL1rYMLNVgV2B94HHgSPjaUcCj8X9x4GDzKyFmXXEBxy8EbtQ55pZnzhq9IjUNUIIIYQQTYI6LW1mNpdMd+Qq+ICCeSGEVnVc2ha4M44AXQm4P4TwhJm9CtxvZscCnwIHAIQQJpjZ/cB7wGJgWOxeBTgJuANYFR+AoEEIQgghhGhS1Km0hRDWTB+b2RBgmzyuexvomSN8JrBLDddcClyaI3wMUJs/nBBCCCHECk0+o0erEUJ4FNi58KIIIYQQQoiayKd7dL/U4Ur4vG1y8hdCCCGEKCH5jB7dK7W/GJiCT4QrhBBCCCFKRD4+bUeXQhAhhBBCCFEz+aw92t7MHjGzr83sKzN7yMza13WdEEIIIYQoHPkMRLgdn0NtQ3z5qH/EMCGEEEIIUSLyUdrWCyHcHkJYHLc7gPWKLJcQQgghhEiRj9I2w8wOi4u/NzOzw4CZxRZMCCGEEEJkyGf06DHA9cDV+FQfo2OYEEIIISqUDsP/WeN/Uy7fs4SSiHzJZ/Top8DeJZBFCCGEEELUQI3do2b2RzM7MUf4mWb2h+KKJYQQQggh0tTm0zYYuDlH+LWA7KZCCCGEECWkNqUthBCqcgRWAVY8kYQQQgghRDa1KW3zzWzT7MAYtqB4IgkhhBBCiGxqG4jwW+ApM/sdMDaG9QbOA84oslxCCCGEECJFjUpbCOEpMxsCnAOcGoPfBfYPIbxTAtmEEEIIIUSk1ik/QgjvAkeWSBYhhBBCCFED+UyuK4QoEDVNZqmJLIUQQtRFPstYCSGEEEKIMlOn0mZm65RCECGEEEIIUTP5WNpeN7MHzGwPM9P8bEIIIYQQZSAfpW0zfGWEw4HJZnaZmW1WXLGEEEIIIUSaOpW24DwbQjgYOA4fTfqGmf3HzLYtuoRCCCGEECIvn7Z1zex0MxsDnI3P2dYGOAu4t5brNjKzF8xsoplNMLPTY/gIM/vczMbHbY/UNeeZ2WQzm2RmA1PhvczsnfjfSHXTCiGEEKKpkc+UH68CdwNDQghTU+FjzOymWq5bDJwVQhhnZmsCY83s2fjf1SGEK9Inm1ln4CCgC7Ah8JyZbRZCWALcCAwFXgOeBAYBT+UhuxBCCCHECkE+Pm0XhBAuSStsZnYAQAjhDzVdFEKYFkIYF/fnAhOBdrXEsw9wXwhhYQjhY2AysI2ZtQVahRBeDSEE4C5gSB5yCyGEEEKsMOSjtA3PEXZefSIxsw5AT+D1GHSKmb1tZreZ2doxrB3wWeqyqTGsXdzPDs8Vz1AzG2NmY6ZPn14fEYUQQgghlmtq7B41s92BPYB2ZjYy9VcrvOszL8xsDeAh4IwQwrdmdiNwCRDi75XAMUAuP7VQS/iygSHcjI90pXfv3jnPEUIIIYSoRGrzafsCGAPsDYxNhc8Fzszn5ma2Mq6w3RNCeBgghPBV6v+/Ak/Ew6nARqnL20cZpsb97HAhhBBCiCZDjUpbCOEt4C0zuyeEkLdlLSGO8LwVmBhCuCoV3jaEMC0e7gu8G/cfB+41s6vwgQibAm+EEJaY2Vwz64N3rx4BXFdfeYQQQgghKpnaukfvDyH8AnjTzJbpagwhdKvj3v3wCXnfMbPxMex84GAz64F3cU4BToj3m2Bm9wPv4d2vw+LIUYCTgDuAVfFRoxo5KoQQQogmRW3do6fH38ENuXEI4WVy+6M9Wcs1lwKX5ggfA3RtiBxCCCGEECsCtXWPTou/n5ROHCGEEEIIkYvaukfnknuUpuGrW7UqmlRCCCGEEKIatVna1iylIEIIIYQQombyWcYKADNbH2iZHIcQPi2KREIIIYQQYhnyWTB+bzP7EPgY+A8+4lOjN4UQQgghSkg+lrZLgD7AcyGEnma2E3BwccUSQojy0WH4P2v8b0rLQ3L/MWJOkaQRQggnn7VHF4UQZgIrmdlKIYQXgB7FFUsIIYQQQqTJx9I2O64f+iJwj5l9TT3WHl3eqLUFffmeJZRECCGEECJ/8rG07QMswNcbfRr4CNirmEIJIYQQQojq1GlpCyHMSx3eWURZhBBCCCFEDeQzenQ/M/vQzOaY2bdx8fZvSyGcEEIIIYRw8vFp+yOwVwhhYrGFEUIIIYQQucnHp+0rKWxCCCGEEOUlH0vbGDP7P+BRYGESGEJ4uFhCCSGEEKKMjGhdQ7jmIywn+ShtrYD5wIBUWACktAkhhBBClIh8Ro8eXQpBhBBCCCFEzeQzerS9mT1iZl+b2Vdm9pCZtS+FcEIIIYQQwslnIMLtwOPAhkA74B8xTAghhBBClIh8fNrWCyGklbQ7zOyMIskjhBBNipqW1tOyekKIbPKxtM0ws8PMrFncDgNmFlswIYQQQgiRIR+l7RjgF8CXwDTg5zFMCCGEEEKUiHxGj34K7F0CWYQQQgghRA3UaGkzs1/F3+vMbGT2VteNzWwjM3vBzCaa2QQzOz2Gr2Nmz8b1TJ81s7VT15xnZpPNbJKZDUyF9zKzd+J/I83MGvfYQgghhBCVRW3do8nSVWOAsTm2ulgMnBVC6AT0AYaZWWdgOPB8CGFT4Pl4TPzvIKALMAi4wcyaxXvdCAwFNo3boHwfUAghhBBiRaDG7tEQwj/i7vwQwgPp/8zsgLpuHEKYhvvAEUKYa2YT8SlD9gH6x9PuBEYB58bw+0IIC4GPzWwysI2ZTQFahRBejXHfBQwBnsrrCYUQQgghVgDyGYhwXp5hNWJmHYCewOvABlGhSxS79eNp7YDPUpdNjWHt4n52eK54hprZGDMbM3369PqIKIQQQgixXFOjpc3Mdgf2ANpl+bC1wrs+88LM1gAeAs4IIXxbiztarj9CLeHLBoZwM3AzQO/evXOeI4QQQghRidQ2evQL3J9tb6r7sM0Fzszn5ma2Mq6w3RNCSBaY/8rM2oYQpplZW+DrGD4V2Ch1efsow9S4nx0uhBBCCNFkqM2n7S3gLTN7BJgXQlgCEAcHtKjrxnGE563AxBDCVam/HgeOBC6Pv4+lwu81s6vwJbM2Bd4IISwxs7lm1gfvXj0CuK5+jymEEEIIUdnk49P2L2DV1PGqwHN5XNcPOBzY2czGx20PXFnbzcw+BHaLx4QQJgD3A+8BTwPDEkUROAm4BZgMfIQGIQghhBCiiZHP2qMtQwjfJQchhO/MbLW6LgohvExufzSAXWq45lLg0hzhY4CuecgqhBBCCLFCko+lbZ6ZbZUcmFkvYEHxRBJCCCGEENnkY2k7A3jAzBLn/7bAgUWTSAghBIxoXct/c0onhxBiuSGftUf/a2ZbAJvj3Z3vhxAWFV0yIYQQQgixlHwsbQBbAx3i+T3NjBDCXUWTSgghhBBCVKNOpc3M7gZ+AowHktGcAZDSJoQQQghRIvKxtPUGOocQtMKAEEIIIUSZyGf06LvAj4otiBBCCCGEqJl8LG1tgPfM7A1gYRIYQti7aFIJIYQQQohq5KO0jSi2EEIIIYQQonbymfLjP6UQRAghhBBC1Ew+o0fn4qNFAVYBVsYXkG9VTMGEEEIIIUSGfCxta6aPzWwIsE2xBBIiocPwf+YMn3L5niWWpARo9nshhBB1kM/o0WqEEB4Fdi68KEIIIYQQoiby6R7dL3W4Ej5vm+ZsE0IIIYQoIfmMHt0rtb8YmALsUxRphBBCCCFETmpU2sxsvxDCwyGEo81s7RDCrFIKJoQQQgghMtTm03ZBav/5YgsihBBCCCFqpjalzWrYF0IIIYQQJaY2n7ZVzawnrti1jPtLlbcQwrhiCyeEEEIIIZzalLZpwFVx/8vUPvjoUU37IYQQIn80H6EQjaJGpS2EsFMpBREib1TwCyGEaILUe3JdIYQQQghRevKZp61BmNltwGDg6xBC1xg2AjgemB5POz+E8GT87zzgWGAJcFoI4ZkY3gu4A1gVeBI4PYSgyX2FEKKJU9NSd7CCLncnmjzFtLTdAQzKEX51CKFH3BKFrTNwENAlXnODmTWL598IDAU2jVuuewohhBBCrNDkZWmLS1lthw9AeDmE8Ehd14QQXjSzDnnKsQ9wXwhhIfCxmU0GtjGzKUCrEMKrUY67gCHAU3neVwghhBBihSCftUdvAH4K/D0GnWBmu4YQhjUwzlPM7AhgDHBWXGmhHfBa6pypMWxR3M8OF0IIsZxSU7fllJYlFkSIFYx8ukd3BAaGEG4PIdwO7AH0b2B8NwI/AXrgU4pcGcNzTd4bagnPiZkNNbMxZjZm+vTpNZ0mhBBCCFFx5NM9OgnYGPgkHm8EvN2QyEIIXyX7ZvZX4Il4ODXeN6E98EUMb58jvKb73wzcDNC7d28NVhBCCFFRaHCFqI18LG3rAhPNbJSZjQLeA9Yzs8fN7PH6RGZmbVOH+wLvxv3HgYPMrIWZdcQHHLwRQpgGzDWzPmZmwBHAY/WJUwghhBBiRSAfS9tvG3JjM/s73o3axsymAhcC/c2sB97FOQU4ASCEMMHM7scVwsXAsBDCknirk8hM+fEUGoQgSklNE/lqEl8hhBAlpk6lLYTwHwAza5U+P4TwTR3XHZwj+NZazr8UuDRH+Biga11yCiGEEEKUihoH3BSxGzuf0aNDgUuABUAVPjggAD8umlRCCCGEEKIa+XSPngN0CSHMKLYwQgghhBAiN/kMRPgImF9sQYQQQgghRM3kY2k7DxhtZq8DC5PAEMJpRZNKiBJS6xB7TQYqhBBiOSEfpe0vwL+Bd3CfNiGEEEIIUWLyUdoWhxB+WXRJhBBCCCFEjeTj0/ZCXB6qrZmtk2xFl0wIIYQQQiwlH0vbIfH3vFSYpvwQQgghhCgh+Uyu27EUggghhBBCiJrJZ3LdI3KFhxDuKrw4QgghhBAiF/l0j26d2m8J7AKMA6S0CSGEEEKUiHy6R09NH5tZa+DuokkkhBBCCCGWIZ/Ro9nMBzYttCBCCCGEEKJm8vFp+wc+WhRcyesM3F9MoYQQQpSQEa1rCJ9TWjmEELWSj0/bFan9xcAnIYSpRZJHFJqaCmNQgSyEEEJUEDUqbWbWEjgR+Cm+hNWtIYTFpRJMCCGEEEJkqM3SdiewCHgJ2B3vFj29FEKJCkVdLEIIUTzUc9LkqU1p6xxC2BLAzG4F3iiNSEIIIYQQIpvaRo8uSnbULSqEEEIIUV5qs7R1N7Nv474Bq8ZjA0IIoVXRpRNCCCGEKASlcuEpYjd2jUpbCKFZo+4shFg+kV+MEIVF/ryiROQz5YcQQizfSBEV2ShPVAZ6T/WiaEqbmd0GDAa+DiF0jWHrAP8HdACmAL8IIcyK/50HHAssAU4LITwTw3sBdwCrAk8Cp4cQAsVAmUcIIYQQyykNWcYqX+4ABmWFDQeeDyFsCjwfjzGzzsBBQJd4zQ1mlnTP3ggMxZfO2jTHPYUQQgghVniKprSFEF4EvskK3gef/434OyQVfl8IYWEI4WNgMrCNmbUFWoUQXo3WtbtS1wghhBBCNBlK7dO2QQhhGkAIYZqZrR/D2wGvpc6bGsMWxf3scJGiw/B/1vjflJYlFEQIIYTIQU31lOqo+lHM7tH6YDnCQi3huW9iNtTMxpjZmOnTpxdMOCGEEEKIclNqS9tXZtY2WtnaAl/H8KnARqnz2gNfxPD2OcJzEkK4GbgZoHfv3sUZrCCEqEatlt7L9yyhJEIIsWJTakvb48CRcf9I4LFU+EFm1sLMOuIDDt6IXalzzayPmRlwROoaIYQQQogmQzGn/Pg70B9oY2ZTgQuBy4H7zexY4FPgAIAQwgQzux94D1gMDAshLIm3OonMlB9PxU0IIYQQoklRNKUthHBwDX/tUsP5lwKX5ggfA3QtoGhCCCGEEBXH8jIQQQghhBBC1IKUNiGEEEKICkBrjwohKgLNRyiEaOrI0iaEEEIIUQHI0iaEEEKIFYIV3SIvS5sQQgghRAUgpU0IIYQQogKQ0iaEEEIIUQFIaRNCCCGEqACktAkhhBBCVABS2oQQQgghKgApbUIIIYQQFYCUNiGEEEKICkCT6wohhBB1sKJP2ioqA1nahBBCCCEqACltQgghhBAVgJQ2IYQQQogKQEqbEEIIIUQFIKVNCCGEEKICkNImhBBCCFEBaMoPIUTxGNG6hvA5pZVDCCFWAGRpE0IIIYSoAKS0CSGEEEJUAGVR2sxsipm9Y2bjzWxMDFvHzJ41sw/j79qp888zs8lmNsnMBpZDZiGEEEKIclJOS9tOIYQeIYTe8Xg48HwIYVPg+XiMmXUGDgK6AIOAG8ysWTkEFkIIIYQoF8vTQIR9gP5x/05gFHBuDL8vhLAQ+NjMJgPbAK+WQUYhhKhItHamEJVPuSxtAfiXmY01s6ExbIMQwjSA+Lt+DG8HfJa6dmoMWwYzG2pmY8xszPTp04skuhBCCCFE6SmXpa1fCOELM1sfeNbM3q/lXMsRFnKdGEK4GbgZoHfv3jnPEUIIIYSoRMpiaQshfBF/vwYewbs7vzKztgDx9+t4+lRgo9Tl7YEvSietEEIIIUT5KbmlzcxWB1YKIcyN+wOAi4HHgSOBy+PvY/GSx4F7zewqYENgU+CNUsvdEGr1Ibl8zxJKIpoa8l8SQogVj3J0j24APGJmSfz3hhCeNrP/Aveb2bHAp8ABACGECWZ2P/AesBgYFkJYUga5hRBCCCHKRsmVthDC/4DuOcJnArvUcM2lwKVFFk0IIYQQYrlFKyIIIYQQQlQAUtqEEEIIISoAKW1CCCGEEBWAlDYhhBBCiApgeVrGqmkxonUN4XNKK4cQQgghKgJZ2oQQQgghKgApbUIIIYQQFYCUNiGEEEKICkBKmxBCCCFEBSClTQghhBCiApDSJoQQQghRAUhpE0IIIYSoAKS0CSGEEEJUAFLahBBCCCEqACltQgghhBAVgJQ2IYQQQogKQEqbEEIIIUQFIKVNCCGEEKICkNImhBBCCFEBSGkTQgghhKgApLQJIYQQQlQAUtqEEEIIISqAilHazGyQmU0ys8lmNrzc8gghhBBClJKKUNrMrBnwZ2B3oDNwsJl1Lq9UQgghhBCloyKUNmAbYHII4X8hhB+A+4B9yiyTEEIIIUTJsBBCuWWoEzP7OTAohHBcPD4c+FkI4ZSs84YCQ+Ph5sCkekbVBpjRSHGXt7j0TJUR14r4TKWMS89UGXGtiM9Uyrj0TJURV0Pj2SSEsF5tJzRvmDwlx3KELaNthhBuBm5ucCRmY0IIvRt6/fIYl56pMuJaEZ+plHHpmSojrhXxmUoZl56pMuIqZjyV0j06Fdgoddwe+KJMsgghhBBClJxKUdr+C2xqZh3NbBXgIODxMsskhBBCCFEyKqJ7NISw2MxOAZ4BmgG3hRAmFCGqBnetLsdx6ZkqI64V8ZlKGZeeqTLiWhGfqZRx6ZkqI66ixVMRAxGEEEIIIZo6ldI9KoQQQgjRpJHSJoQQQghRAUhpE0IAYGa5ptYRQgixnCClbQVDFa+oL2a2v5kNDHJwFU0EM6uIQXhCZCOlbQXCzCypeM2sbT2vXWHyQjkV10pTms3sp8ADQC8z2zSGVdQz5CJ5hnI/SzniL/czF4JiPoOZHQI8bWYrlyCusryL5SkPLE+yLG80JG1WmIq6mNSk0CxvmTGlsJ0IvGBm95tZr7palWa2UgihKu73N7MjzGwLM2vZWJlSlWezxt4r3/hCCMHMepvZMDPby8xaFDO++LuRmW0ArF6suOohyyrpPFtTPjWzdUIIk4HewDrAJDPbPabfcqFsNKQxEfNzYjVcpdGC5R9vkv5dzeznZtap1NbLVP7vaWanxm+5ewHuW7K6Iv3+zKyNmRXkmzKnJbA/8DpwhJm1KlZ+z3qOrcysk5ltVNd1BY63YOlXj/iT72ATM2sHrFmg+20Ry/Uu2f81lqzysmXWf4WKI3mONZM6Oea9+n1bIQRttWzASqn9w4E7gT8B3ZPvYjmTcXPgM2AyUAWMAfoDzfK4z6+AmcD3wPvAscCajZArmVJmO+BK4JfApkVMhyS+fYG5+KoZVcDvgPZFjG8w8C6+cseNwDZlyAOJLLsAjwKjgN9l/5863gWYCOwRj8+OaTUfGJjrmhLJvz3wW2A4sE4MW6ke90l/C0fEtLgC2LFE8g8BvgFmxTx4DNC6xHlhzxj/nPhO3wR2bcT90ml6EHALcAbQtQiyp+M6GRgPvAKcV8A41gAOjWkzKinjCpnf0/eK5d7n8X38o5h5MSveoqRfPvHHMvE9YBpwK9CvkffbJ5bn38b7HprrmQuQ5w4H/g78EehRwDiS5xgEvIjPOTs8lwx13qsUL3JF2HCFZlbMMFXAWKB3IV5oAWUcHCuqB4GuwHl4JfwasBu1KG7A8bGS+Q3wcXzGqcBxjal0gP2iDEm6PQBsV8Q06BcLx98BL8Q4q4DL8cV4Cx3f3riSexswG1gAPEZ5FLf+wDxgNK58VwG35DhvlZifkwp9lxh+SgxbCAwoodxphWcu3uioAl4F2sT/6mx0ZN1zeCzgX4v3+jewV5GfY5eY/n/CV3GpinliKLB2idJyA7zBdg+wE3BBlON+oFVjyirg3PhtTYr3/L9i5fNUefv3GNdcUo2QBtxvpazj9sAj8d7PAa3SebGAz3FWzIeXxN8fYp7cocj5oKDpV8+494pl4i2xHFoIPEnDFbc943d1acx7S+J9Dy6w3OfFeJJv98mkbCxE3ojl23d4Q7Iq5oc/1JRHa7xPKV5ipW8xsb8FhsUPLqkQxwF9yyhXuoXwCzIKSlVSKAC/jscvAKulzk+3yFrjSs25wBPAp/G6hbHSObIhGRbvdvs6FlhPxAJrcSwsdypgOiSV/obA6fHj/nOU/2wySsBIoEUB410vPssfgTtiIfnPWGA9DfQpYV5ohivr7wAb40r7f+Jz756dZ/DVUH6DK9SvEy0xwGmpPLRzCeXvhyu8v45p+F2UYTSwXr55IO7vCHwFnIkrftNjvvsP8PMCy53kvWbAhbhV70a88P9d/P0BV4jrpXg2QIa1gZ/gitVtqf+fBz4A1mrIfeP+HvG+v8QbrIkVryCNsKy4toxl0DlR9inx2/oBuLaR8eyBWzvWw8u9++NzPE0BLKJUL5N/BNyLW46fAT4E/hbz4ns0UInJQ4aewCd42VfQ9Msj7rWBh6leJj6Kl4nPA9vncY8uVO8luA64AfgrXg9fG9/ZAuDIAuW5XXFr6Ol4HT89ptdTxN6IRqZLR+B/eD17Z3yOmbgCOrJe9yrmC6zUjWVbZVvFD+DRmNDb4ybOKuBloCVltLbhpv57YsZOWtYz8Za2xcJvi1zPBzSPv+vjpuE5uHIzEDerjwU2q4cs6Q9hO2BEvN9i3Jr3CpnWbaOsD/HZ0t1qo4G+eMH8dnw3O+DKzL8a+/Gl44vHK+PWlbPix3cucFT84KuIltgivvfk2XsDbYG74vtLunuOiwXPMsoX3vX9TCyoqoAJZBS3X+HKXKcSyf+TWKhdhbfOf8AtxomV7O2Y1jm/sax3sgtwCHA0cB/e6NidjAL7PLBGgZ9jMK4ArwIciFuoH4p5cRSuSA8qRvql0nAXXNndEfgyvtej8Qrwufjd1an81pCm/WKanohXyNPx7qp3YpreC7RsxHOky6MuwCa4Bft2vLLfBbdeVuHKeH2eI33vA+I9fojvpTtufUysUY/UlMcamCe2wBWoc/Hyb2j8Jj/GGyUdChRXdn21anzWRqdfffIK0C7+bo93ny8GzgcOi3myijoss3hdmnz3V8awDfBydS7eGN8NeCPev0Hd/ln5u2/8Vk6OeXkO7mLzBhmL22qNiSMeH4KXcVXAwTFtkgbyFnnft5Avb0XYiEpM3N8C+Enc3xh4idifHj+I+yhyxZaHvN3xFkcVcHUMO4Usi1sN1w7DC/pLgA54l0Fy3Ti8m2+dBsi0E3BK3N8MVw4+ws3ct+AtjYJVYnhFdX2Ud23c6pEo1R/ivltd4rmNLpSBbXGlYrt4vFeM69lYqFwBbFnk954UlFvHtD0LuIyMknNsLIC+JMv3COiBK2U3AX1w60nSrZr4uBW1Oy8l/7Yx/w7CK+vRuDKwf8wrL9SWV6he+O4X3/X+QOeY7z7BrdC3A38o5PeKD+RaE28czIvyr0/Gsjsat2JuXqi8V4McfXBrxIN4pTcEV1arcIvVd9SjwUJ1RecQ3C92CNAr7o+J7+se3C2gc2PSMLV/GpnG5kq4teZ73HJ+B3ARDfSJjXngJtzSdUdMm9fx8rN1zB8N/maznqNHvP918XhvMt39j+FKXNsi5IN+VLeq/wnvbWh0+uUR98CY506Kx4PiM7+IW56vpg4fSDJlQg+8DKsC/hTDOgIzcKvhY7iBokFlOtXLjH3wOmI3XMF+AS9PD8QbwZdTD4UqR1x74MrfyvE46cFKeoOupp69TgV/eZW84T5Bx8f942JhtwA4FWiB+3BU4QX0m0R/mxLLmN2qao1X2J/i5tfdY/g5UdbNc10bP4zpuFWgClfefoS3aD4B3qrrI8uSI/ngfhzTbSbecjHgL2T8e6YB3dLXNDI9OpNRNK9LhW+DK1HjKIB5O3XfrVLxzYv5pGX80Ktwi1u64CyaBRZvQT+MtwTXxq1Rd8ZCYW58t/tky4ArSlXAPak8NJZMxVLvVmUD5d8OeBzvouqMd9neRsaf8u0kD9aVjvE9PBO/0TVi2L2pd/UpBaooc6Tnz2Mcp8XjI3G/sinA4CKnYc9UvhuRCv8ZrvReSnR8r29exC0kj+GLX6+KWxITK8hcvAFbEKsNrsgk7+pXMexkMt3kM2jgYCJc4azCB4jsHcP+GMMm4d2xhbKwbYgrnI/E8mEffGT2GzHNplOghgPVlY9hqfR7Fe/+PQJXeqtivAUdjEWmzO8c81oVPljgqJhXJsSwxaTK4JrSmozrxmZ4ObYoXv/HGH4L3gidD+xZ1/3ykP8EvPvzITI+jXenvqePidbDRqTL62R8alfGrf7Jcy0i5Ref73MU7AVW+oYrLJ/hfc2nx0R+DO+Tr8JN293w1srtRC2/xDKmla4d8NZN53h8WpT1fWBIDFs7dX76A98GN18/gVfgd8Vn/AduLViVBvh3xAz5CG7tSBS34+L9ro73L2g3Ubz3LbiiMhn4cTq9yDiyN8YBO/kIf4S32P8OnESmkNwvfpA7U0CFtA6ZkoqgCrdMbRTDW+KWx11joZHuQmsbf5vh3cWz8K7QPXFl5xrq0RXeSPnbptLv/lT4QLwF+hKpgrmOe/2UzOCZp4i+W/HZR+AWhwa3lmuIc3BMwx1ivrgD75JMuojaECvJYuYF3EL+LJmu31VrysP1kQMv6xKF6U4yloKN8dGAd9EIC1tWXMm3dCHeNXVZDF8D7+Y7GPhpPe6XrVRvgFtDkzRKXEL+jJf3HQv0HIfFOI7ELd2z8TJvjfi9dgE2KFBc6fJ8Ndwt5soYXxXe/btuzJ/H1if96inHjvGd3YmXi0twZXU/vBG2A/UoE3Gr8dz4PZ1Pxn1jeOpd/iTf+9USTye8q7gKt1CvnPpvBO7S0xgL2254Q/h+3J80cQky3JJ8Fg0cRVzwl1ipW8z4h+MWkxm4ybQT3j//YUz0c+K5eQ/PLZKs50QZE0Vl/xietLbGAaunzk8re0nBsghX8DaIz560MB4g1UVcD5nWwy10T8YPdb8o4+fACfGcpOAvhAK1Ka58dojHic/GBIqgeJCZIuNz4JlUWKJ4nJRLzgLLkC6oV8IVsxfw1vQFZLUKs84fik9L8iDe7b8v3iJP5P+YIoyureN5jiLjqH9c1n8ta0rH7O8vpsUv8G67ObjvVZuazm/sO8Aro/uj7AtxBfPZmP+PL8a7z36neHdRZ1z5XQO3uCYNr3r77OVI01b4QI6PcQv+XmR8JZuTquQaE1dMz+PxhnKH+P3ejE9d9BfgiEbcewtixQushSvZiY9SUhat39h3kXqOxOL6GT56+SNcKTyqwHkg/Yyn4L0ic/GGu5GxVr9DkXuDcF/ORcQuPrw7fQk+AO3grHNzfcs9SE1HgvuWLQYOj8c/I+P+M7Ku++Wbv2PYoXgdOB+3SrZI/VfvOjCVD1rgA1s+wxX11fHGQtJoWDnr/PpZwIv5Qithy/oAVsL7sifGBD4uhu+CKyDzgHXLLO/peAvuLFyBWEJ0nIz/DyXVJZp17d64de2mVCZ6DC/4V4+FZb3mXiJTiWyIW9bGEbtMcOvNIrzQ3y59fiPTYB+86+mr+IGfjFu6rojP9BE1WBwa+Gw9cP/F93Dz/KxU3tg9xpmXVagAsvSKeXRQPO4X0/wb3DfxR9npjCt3E8l0hb+Mt2g3xa0cF1KklngO+TfD3RASOQ+KeeQz4OQ87pP+XnfH50HbLR7vhbdqp+MKR+siyN8Gt+62xBsqv8F91xLl9/kSpOEeeENydsyLh0d5kpGQ/6IeimpWmu4Y0zXplh4Wv+v3cKWkwQMOcuTLrcn4JbWIv6NxC8UD8Vny7krMuveZMY3m4P5+7XH/wyfjfR8p4HOcgvtYrYJbab7FRw4ncX2EV+SFnk6kF979Np6Mj95WeCP8wRhWECtirueOx2fFeG7Cy+D++JQZS3BrX41KY/yOvsGVzJ/FsP7xfuPjc2we0/ETYJ8CvKvBuNHjl/Gb2Q8vF2fjdWeDlLUccd4en2PHeHw4GavhzY3JCwV7mZW4sawFar+4fyCuFHxJbCXFwuwnZZZ3NXy486UxI/8v7icVxpBaru2fOu8X8V73kDGl18tsT6oSS4XdmfqAB+EOqNeTslIWIA264BX8H3Cfpy/i8/TEle7rKOB8XLiZO2m17oArjB/hLclksMW66TQpVj7FfXNmkemy/10M3xb3zfke7ypcJXXtMLw1+QBumbkYV3RfIU5JUiy5c+SVfXBFew6uWCWjVQ8k43NUq99N6l6/iucviL/nx/A9472nUM8pLvKIcy+8Mvk0pt82MfxHeNf4P+Nz7F/EtOwZn+9h3LqcTL0xCB8E8QQNr9zSFvz/Ab+I4Un35VgaMfKW6uXtcTEfVgG/J9PQe5TM6PceDYznLLxr9894g2AR3oBsj/tvPkTssmtMfoj7rch0vU7CFbincWW+K14m9ixCPjiFzNRCO8fvugq3+v4MV0gKPtghxt0H77beFFe8Er+t0TFf3ha/hQ+ow+cRtxgnA+d+E8P+Go+nxe/tWeIgERqj7Lhv+lxcqa7C/TNb4eXSPFypatXQ/ICPgt8CV16HxDi+j+/mXlxZmwY83aj0L8ZLrbQNrwDm4F1N68eMeGDMdItIzb5cYrlymXQ3wLsTFuCtuf1x36YvqD7oILtFZHg/fRXejbQ5bjX4R8ywGzVAvgF4a/ZR3Iq3OZlCdwludj4Nr+TOKFCa7IErTK/Fd3YEbgF4Cu9eST6gQjkW9yOj7P46hu0bZfgunW6FijN1v7S5fkCM77excPwhynRxSs7tcP+fV2MebkbGX/Gr+H5a4ha5RFkv2Lx1dTzLoJjPrsa7aatiGiYWw0NJDeDIcf1AMkreKTEtTo35byHV/V4GUuAGFt4VPwevUGbghf9OpFYMwZXn2dSzS6+echwT40isi3vGZ78lHq/akLyIW/CTedimxu93LnFeO1zJymnBb8Az9Mf9e+4j46x+HV62/Qq36uXtM0x1ZXATfEDKL3GL4/9wRfZ7vKzbsDHfKdXdTn6D+4E2xxtHyWCeBfEb3LiA7z27SzlRbD7FFZ/meH2QWFoLOR9l2i92KzLK9kS83G+DK49f4+XMMbjV91lqsHZnPc/WZMrYpIv3YlwBnUBh5knrSWbFmm1xw0IV8Lf4/yHUs7eB6sr7gJjHZsVnb477EiZlxXxcp3gGH8G8SoOfpVAvtlI3XMv+HveDORRXiC7CWxFH4haWkjhnZ8mVztT74T4SyXDqxEn4E7zF/0dSlrKsa3viFeZmuF9HMi3Ef3Cr1TrUY4QMcYLQeG1itv4e9607Am+5HI2PBjsG7y6YRgPN9KnC4qfxdwcyIzdPxBWRW+J7anRFTfVusLXjfqcYXxVwVuqdFGyC4BxydIvvKOlGPAF3AP9LLADOIWNxuyqes3IsEH4X02Z93LJwTTzvadzathpe4ZRqHrZkBPH1eKGZWEGq4v5e2ddk3SetOO+BW1Z/n8p/J+AFY8EsujlkGIorE0/jys3B+AzqN5JxbD8kylCw8iKVhsmghsQn9a54vDFuSb0ppnN+s6pXLyPWjOl5GW4h+R8Zxb6K2ANRoOc5iMxo0F54OfssriRej7siNMjPLOaNZrhSeHG85zF4V+nMuDV4BCXeGLiLaAHELTM3pf7vhJd3VXhDYsMi5MN9cSWnOT7woCq+/x/H7/83FGiASI64++M9PQ/hrigLcKVxcPy/Nd6IuDp+j3VN8bErcGzc70XGd+201DkN6sXIPh+3Ds4DLojHa+FK5xuNSI/0KNGbcFejcWTcjlbBrbv98TLqipgHGzUoquAvdnnfsgs13B/qB7wwvp/M1A1Jq73Ba28WSN5z8YotWUfzejL+Zx/GjybnR4qPEE1Wb/g6FmTrk3Haf5o8HYpjnInCNhhvtfwL7zY+HO+y+RBX3JrHzHoB3pW5VSPTYBfc4pl0Bybdul/iVsZ5NLBLKCue5CMcgCtMSbfi5nFLuqLSa8bV25E0DzmaxQ/8L3gl1zmG7RLT/Z9x/xFckUgPf1+VzOoYb+DK55qxUFmCWwAKYjHJ81n2xCvOVeO39g2ufO6AWyY+o4Yls2LaNsO7wqvIWBcPxefXmoN3xewd8/L7FG46hWZZx6fGuOfjBXBHfPDOa2QU61UpoHUlFfeg+H0dhVtKk4phXPwG5wH9G3jvw+L33A63DMzHlbf94rN93pj8kv1t4Nbg8VH+y3Bfry1xP6hvafgEwNtT3WXguHj8LF6uX568pwY+x+pkJuG9HW9UvUO0dGeduysF8hHNesZkqp5X8cZ4c1xBSizpHQqd91JxdyGjxCd147Hxnc0iWpfxBtafiWtz1/Q8uIL7CV6uJ4MOepMpYy+KYfUeRJSVZivjZciP8TpwAa58bx3T8R80YnJ8vCs6qV8Pxsusf5PpyUga/nvEfN+oujCEJqi0pRL7UNyRuEtMzPF4i+wk3DfhZoq07Ewdcq1LZmTTCXgldxLuv5X0xf8O1+JbkZr8luqt513xwvyP8bmSeYL2JKMU5DXoAPfNGE20IJAxwy8h40B6BG5x+4zM1BI9aMA8N/Ha9CCApGD6jtjNivvyPY/7UOyavqaR6b9dTKf/kJlF/y3cIrlNLGT2LkE+WBNXiJP4O8WwV3DlZQyuqHTBrS3pgqoN3vVUFc9vg48wvCfmoYIrFjnkb4bPHZfMH9gSLzi/wq0sz+JdZJ1qendkLFibxGu+ImNdPDk+2/e44norBZgzjJSjPV4xHogrzT3xrv+FUfZn4ve1R03yN1KOdOV2R3zmz/FBAq1wR/N38S6k3fOVgeplRHu8Evt3PO5CZu3X0XgZ0eApKrLiakPGarINXp4txAfBtIxp3KB8iXerro4rUZ/iFek6ZKaB+ZICKPN4Q+FveLl3Q7z/B3ij4jy8TNqlsfFk54G43yJuV8U8/yLeVdkcr6t+oMg+12RmGPiCzKj9k/Hu0p3TstZxn91wV6S34v3mAEfH/36Gl7EN8kvGfXpXi/vH4/NAJiPmz43vLvGXm0Y9rZKp7zL53R3XHRaSmQh4NTLTMSV+r0YD/OVyylDMl7w8bVkfwE4xQT/EK8KVY6GyN94q+4YSWiNScu1CynKCd7fcgFcW0/Fux8Ry9uesa3fEFadkYtHjYwZNTNdJK+2aesrUDO+avQ5vlSejQJM1TT8h49B+NNHXpkDp0RevVG7DW2/Js5+TOmf15P1SGKXtMrzgT0bP/TbGmRQqa2Xnp0Lm0ax8uj7e/ZZYyLbEfT7+jVesg3Hl/DvihKRZ1yaF7IvxeA0aYW3I9xmyjhMLVTK6eQRe6M+kdh+2vnghm0zZcCluYUwGtiTPdVf8Zuo16rmGOLfAHda7k/FhSwr4P5BZGutrXKkZnOuZC5iW/XEl9S4yowHnkJmHsRmp7qP6yIFbipL1Un/AreIr4RbZifhAjgbPRcmy01KMw3sxTo9hPyNjMTyvEfGcRJw0N76zH4C/xP9ax++jwd2UOfJz29S7SPzXkl6QJRRoPsCscmB4zOvr4orb72NcT+LlejMKNP9bdvy4n/BWqfDEL3oScXUFMnNE1mkVw30KP4jfTy/cCv85Xh4cGc9ZJ1fa53HvPmSMGrvgjbxPYtjH8f+BeBl0Eak5PRuQLm1S+9vijegq4JIY1pIirT1d8BsujxvLtlhWJuOgPZHM4sG34y3A7mWQsTmZlsBbZKxH28aM9y/c/+7BuHVOXdsOr8Tn4IpeCzJdZNfjfiNb4y3Oi/P5uLIyaDO8QJyJW036xvARZPxTtk1fV98Prob4T6B6hb893n3zLVlKSmPzBpkZsZPVGxIlbUfc8fv0JC2y81Sh8yneTbAHXmC2IGNpfB6vbFcm0yWXOKLPSGRM3W8d3P8k6SYqyfyCeOvzNnw01Y9xi9p/cT+SZriFZ9P0M2ddn57/bha+7u/NeEWxLV7QJz6Ge9DIaShinGviFunEH+XvuLP3cLxAnk2cgy1+b2sVMq/XINPFeAt++3h8TJRvKllzYOWbt7LyzLW4peRN3OG7B14OrUOBljGL7242melIqsiM9O2Hj7xssBUM76JL7ntTzGdVwIEFfhfn4o2GxE8pGSl/T8yD61KEFTeo3t32D7yeakVmWqrHKfBgIjLl0N5kppZ5FNg6hifl0We4S0Deg79iOk2OeW6VGJa4u8wBeuVKhzzlHhDz8WK8F+IO3PI2NJYXE2jgVFB4b9OBcX933JL2Bq78rYrXj8lqIZfX9D4L8n4KebPlccv6AE7Gu9Pa4JXhrTGR34wfw3oUuMVST1nXwYerz4of5W4x7BkyZv4XyGo5xkxzFF6pfYb7mLUnY2mZiiuCX5OnrwXVRwz1wyuvS/BW7dtAv/jf5TGOgjvl45bDqlhg/AhXPP+Dd+G8Q4GWZcFbX4kJvXt8xs9xRSFZhL1oi79npXXSWqzCfbR2wiuKa8gMq0+sK8lUILvjiuwc4MzUfQfEdzWKIjkn53iWFmTmLZuNdye9HPNmXlZYMusWLiGznFWyluzz8Zxfx/s3+rnwrrkZuB/U8FT6n5X6/1vgiVKkYUquRJaRMY/sSsY69Qx5KlZULwNb4l2us3AL0R/xhtj3eNdbo+aporqFrVtM19PxRsPHZLosE9+zvBVuvBHzY7z8PgBXCDfHy8xn8IZtMmH0f0kpFI18jlXIzGp/L95o2oiMj9tfGhpPHTKcBdwW95PpZJ7BrdD340pJUQYT4Yr9D3gZmMwv9jKZ7r4/ExvT+eQ9XFlbO3VtYlDYDy/bkiXnrmiArNmjOEfjZcezyf9443U+0bpWn/eFuw3Mj9/hUXH/XrwbdxbeG7Iq3qAcT7Hn7Czmzcu9Zb3MbVMf2W14a6UZmWHab1AiS0SWjNkDI9aKhdEc4nBn3Bn9r/GjqWnQwSFk/Jg+whW3nnj33st45VnvDzwWipPwIfRd8G6UuTHddozn1Gg1qW86xDh2JbMcUDLJ5od4ofwQrox+RiMmjkwVJluQUYrvwU32e+IV2vcxLYvuwxZl2Q7v+niQjJI2FVfc0k7HyxRsMZ8kittp8fidWMAUdcm1VFpuRMaatnn8ziaR8SN5oB73HBTz2WS8kN8zvqd7yXSfrFMA2dcm07XxA14oJxOGvoD7DXXGFZuXKdDcb3Xkx+1ivKuSKZ9eiN/CX+N3MJF6WsPwaSnG4BbQofG9DCNjXZ5BgSYkJlO57xe/q4X4yMdkPrH51GO2frxhMhlvuI4nM0JzVMwf1+C9CwfgXdwFyfNEyw/eeE4UwvvIKG63U7jlvNI9G6vG+OaTKQsTi1vS3dcgf+E84l8bb6xehhs2ZuNl7iK8nvxZ9jV13Hcg3vB6EzcAdCOjhC7Ge3BOwrvkz6yPvOSwdOPlZWL1uhIfSf8kXmfUa3J8fNaFmTHtd8TnpLuCTGP+fbyeuAP3ZWudb7o0+D0V68bl3qjeUjoG71o6howp9u74kd8dX2i51xLdHp8kMXH2Pxv3VXqdzCLHNa0reDBu/r0kFiJT4wdwKJnuvLxa0Lg5OVl2aju8oJxEZsDBeviacFW4L9FaqWvrnVFjAZWefPWz+NzjcEV0NbxSmYwrUMfGD+RlGllp45bMj/Buh2TU8P14xbkm3rJfun5kUT9EL8iSQiwZln4GmW6DXciMYL2RHKN+cUVtbswLM+K5W5YoLw+Kcc4A/pAK3xpvdLyN+//0q8c9B8e88A2ZqQHWpYADhHAL+0sxnZOBPkfhjuWJtfAx3FJd0HVza3h/n5JZ8/P3uFXp3zFdp+Nl2IN490+tjs3xW9047nfFLdZz43YuXgHfGtPg/MbkFao3kA+I8g+Jx5fhFf7NeHn7Z+oxLySZ7txRZBpYY+O9PibTKHgqfqcNthbiPmuJq8Q+8b7JpK9rkJlj8BniyMQi5INkubDt8IZE+ns6F/evLOhauqn774Q3XLvj/qgL4vvbN+bNxfl8w2Tqnc1iPku6c+cSG+X4HKO/wX0ef4P7juY9mILUsntRvgti3lsF764cgyv2n8dnqO+ggy7xu0sMPT1xS95FMexovJ5ORrU32q82L7lKEUk5N7wgfAF3uF0Db8EmFqnFuBm9oC2WPOVKF3Jnx8zxPa4gJQ7OiY/Na8QRMdn3wC0w18WCa4cYfjKZCudA8lQ2cAVpSrz2HFwJnIW39gaSUa5+FNNzYCPTYMtUgTgwfmBXk5lw8xPgkPj/yvFjvwZXSuo9oznuBJs4zbbAK+Mvohzt8NbT4pg/ijqHWQ7Z1sOtiEtiYZMsdJ90ke2CK3af1SYb1bsWS2Vh+yk+/cjTZBZh/kvWubvEfDS4nnHsiRf0syiQH2O875pkupl3JDO4IPGVOzz1/b1EPRa9bqA8XfFK60HcUppUcsPid94WV2KvJr85sFrirgsP4Fb2SbjFcCCZHofv8XJn5wI+R3PcAvU57su4Er6SRNLF9hn1cAAn49/4MhmL09N42dYOr0QT5/gqGjfadSBuDZqIl31n40raD8C58ZwzYhmxiMK5Z6TrgrNw680+eFfwXTFf9EudU5QeoZjHZgH/SX0XVXh5fCeuJG9axz3Skw/vFvPdWLzsOhpvGM0ms152ssrPl9RjOgxc4ZsdZT6bzDx8s/BGwar4YJeJ8bx6DcDCFbbv8Lr32fjO/xSfI/EtvT7KfgFxgF4ptpJEUq4Nt1xVc36N4Z1xp+LbKXHlnEPGM2KmOjt+nFUx8yUjSE+h+koHuVZJOJHMqJk2ZEahvk09J/rEl9/4BldmrsK7V2fipvptyUzDkDiRNsS6ZrgC/RXwaAz7Pd6CvAW3EP0Br5ymkbH89cC7V+ptEcAVi3m4dTUZcfpoTLfEknlafNbFFFBBqCkN4m9HMt3La5BZs240mcEGHVPX1dmlhCu3JRn9jHd3LMAr5V1xa+A78RluT73voTGs3tMSkFnf9RMK0yXaJebnJ3HLQreYFy4iY2H7Afdxu4Ai+aik8sBKMR3nkZmzagu80vhHKg23x61VPeq470Bc2d0v9Sx/S/2frPgyK35rHQr0PCfhqw/shSubC8hYjbfCy9x6vf/Uu59DXJsWVwLeJ+NS0RJXEBozoCGx5o3Drfrf4z61g8g0Im/DG3q/pwjzsMXjP5Jp+DyOd4V/DpxazDwY91fBe6LmkulZeTbKMo9Ugytb7hjWE28EbB6Pz8WV28VkpsZJBo4sIjM11AakrGb1eFdX4I3a7/AG1gQyE/Qm85n2pgFuNPG6h/F642d4b8ziGOdJZBp3VaTW1c6VLgV/Z8WOoJQb3jq5I37oq+LWiz+TWaNtm6zzy+HDthnRqT3KeD2ZxYU/IjPx7RKypkSgendqX3wKjh3i/gPxmgl4a/EpGjhvVfwovsMVpj/hitLXeGWcd/dWHvF8BdwX97vgCtU3eHfozrjFaS6puY/IczLgrHi64j4Uv8Sdoq/Hu9l2ioXMBLywvAH3GZqKWwkavNRIHfKk17KchHd73YcrpauSWaJmHF6QGmWYMzDPZzgoPsMi4JcxbEsyFtsuUf7NaMAQ+1R8u1GALiG8Zf8/Mpb2hfiow2lR5rXItKSryFh3imVhG4BXBEmcL+H+U21j/n+AzGLqRh2O+1RfPSKZUyyx1idT8yRLXXWhkdYiqlf6ydJAr8fvZxSuADVqkBIZ/8Zv8a62GWSUikaX4VS35m2Mj7afSmaVicG4MroQb9Q12Je2FhnOxS3tyfyGv8K7fhML5RgaUPbVIw8ehzca28Q4b0v9vwMZRazG7wBXZu7Ge3qSGQZ+idcl44nlOK68D4j79SrXyCjxH+CWwANiWj0Z39kvcGtuFXBlI9Nl1dT+tjEPLMIbdsfjZV9RpvWoVa5SR1i0B/GWaVJYfRYLja3xvu0/xMLrIQqodDRAxsExsz1AZlLRjXFL2ff4FBo/x1uSX1KDlQxvVSTL9vyAV/KH4IX/R7HQbJRzLBnF7Qu8myWxfPVt5H03BU6M+5OIloR4vGV8T3PITIrYMf7XoJmxY0H0RbzXmrgyW4VbEbeMH18y/9t8vCX4Km7RKEohGWXbK77zv+DK6yK8kuuJKxZ3UYBVHoogd6KsrZEKS5TPOVRfam2HcstbwzPsjVf8s2Oa305mos9E8TyH4lvYNsNHQ/8O7+p7JMowDVcg5pPnxNEsu3rEt/Fb+jsZH9QX8cr5z8l9G/EM6bVgj8XL2n640vlXvBJNyuMraWQDCHdzSebNSy8d1ejGDBlF4FtgaAwbhzdU38e7wC7HlYSCTExNzaNT/05UzHG/siOTcqFIeXEDXLFOFKGRuF/vV+ToNs+VD3Fr1BW4stk59d63jf//CrfqfkRqBDn1n1tw33jfr3CL2p9wi9i2eIPr0ZgvH8LL8IKOmMctbo9HGa6h+kTcRbewLY2rVBEV/UFcOfswfmT/wZWeebiCdG7MOFWxQGn0vE4NkG/XmNHuIc53k/pvK7yA/igWejUuuRIzzjx8stGDcGVkYaqw+RGFmnnZFbdZuLJ2JXX4M+Rxv7TP3HC8ov8brsglFdlt8d3NIKW0NOajwC1Zf4/72+FdHFXxo28X06w/3h2S+PEVrdsct6Tcg1v3biezyPDCWNhsRabrp2SFQR0yL7X2xYLxBVwBHhbDBpPpRjo167rl4hmynmcPvFHyMe4C8CPcRaF/jucuxnQO/cgsMH1EKnwkPmjqCTLWiHxG6GWvHjGTTDfbvriFuSq1NaYrMW3NOyeWW4vjN30/bpFvhSuLEynchLPJ1DYzKbD7Ahlr3ly80TsXH2yQ5OnF1GO0az3izTU69WGqW3kKamUnU9ZugHfNt8PL+pdSeWYJ8Ot4Xo0NZtxa+0m87swYlky8Po/M1FC/ITXnYANkXjte/+dY1iTdlVfjlsBJMc4p8dtZq9Dplcr7T1GARewbLE+5Ii7Kw3i3VzLBbGcyEy0uwv27PqZE81WlXzhu7XkqFgYbxuPL8dbo72OGvD3KN5XqE+duQmbR9x/FAnghGWtVx1iQ5T2dQj3lH4JXLukuysYoUPviXaBfkfEL+DR+bK/iLdwHyCyT0tB5lrItek+k/tsW9xX5Hlfi28XCayiu8Be0VcuyKx0Y3qV9Wsybv8V9jKaSaqEuLxvVK5D9cGXnCTKjwZLCfV+86zHnWqLL2xYL/7nxeY7OfmdFijPdrfw+0X+S6svRGZkuzDqVRmpePSKZD6sKr5QH4o2FBvk7sqw1bz6Zxk8yoi6ZT+/2eE1B126mwP6NWfdOW/OuJ2PxGkzhfNjyHZ36PAWeNDdLjmTk/Hhi4xh3Dega881kvEFR44TB+CorH+PK7cb4IJR2eB2XLGY/i8zk0PWeIy0rvo7A+nF/m1gGLcYHxf0Gt7T9myKMmKd6+d26Mc/RaFnKEWmxEhT3YZuCK0iH4RXi/bg1agolGpJbg4wPxQLhAjKtgkXx99cp+ddNXZNMjrqQjF/KI3H//3Ctf6v44VxH8Xwe1kuncwHutyeZ6RVm411Br8RCZDpxZFEj7l+TRW+z1Dl9yawPl3RVt6DAFUGWXNvjA0uStSL7x/jfwLtjr6PEjYo8ZE4mnh2Ed+NMxSvtO8nMC7eEzPItSaG63FnXani+Pch0jyeWgmKudtGKjI/aPrhVajbeYFgn+/w87lvX6hFHpf7v2ZhnY1lr3vS4VeH+obuRGYTyWbG+JQrk31jDvRNr3jcU3ppXn9GpCynQ6NRU/In1vjNuVXsvfruzYz5pkTr3NLyb/me13K9fzHMX464nt8Z0m4ErUcmo3oU0cmqoGuL/GW74mI93V3aiuIruclGmlV2AAidqC9z/IPH1+gtxSRYKNGlkQ1827qSZrFH3HT5gYkAs9N4k5SMUr9kD7059Oj7TvfG6H/CKc0nc/wj3vSj6aMFCZlpcCZgT0+SsVPjahYiLmi16/8N9Bt+KhWMy9Lyog1Jw/7l099SvYr54nYwCv2fq/LIXEPG7SazVP+BOvtvhymUVbh28MPVMBbFGlOE5947fWlEshGQqy0F4w2sq3ujaHPdhfTd+zydTf8fsfFaPOA+3KDa4QUB+1rxusby9oBTlURHzQ8GteZRpdGqMO+1/uhveoH0Tb/Dvjzf6kxVBkpH1yYTLNa7bivtkLo7lbDLP5Se4b9x3+OouZ1Ng31yqW722wRv9MyjjakYlzZ/lFqDgD+ROkbPih1HyCXNrkatF/ECPiwWgxQLuXby1kHZqTFrPo8lMjrke7vA7B7ce7oNb766p1AIyFmTfUN0BvJATp9Zk0ZuMtyKTqT6K5beUWFfa4ZXzffiEx0m3UuKHuDWxQi2GHI2QvzWZiWeTgS9DcOfoRbi1IJmrqGBzfZXpWQtqTY73Sncr94sV2T/IKOrPxTTeP1Z0DVIaKfLqEdRtzTsy9X/JR9MVKT8UzJpHGUen4o3FmWR6E5KVMKrIzIH5i3hOevWF3uSxBjeueM7FFc+/4PXUMbgyd1zqvIKWsVRX3HpTj8maK30ruwBFeSh3jq0Cjiy3LFGeZRQRXAEbhVvaumT9l7T0ZpNaBBx3HE2c6LuQWk2gUjeKb+WoyaKXVGYFV9iyCpSdcevel8DLMawHGd+ZX9d0bRnfSU0TzybK5r1kWtaLyDhTL5eDDhr67hp5n6RbOZko+yq8odAFt2BeHdMvWYS+bWPip4irR5CfNW843kBarrr3l4eNMoxOTcV9Oj6V0TAyU24ciSta/wP2i2EHEadmqW+ewX3YEv+/1ng36RyKuF5zjKuiy5oGP3e5BSjSy/wpbgIu99JUzbP+SywvB+KtrtdqkjF+6HNj5v9lKvxOvLVb8LmCyvi+Cm7lyLp/US16tcTbE3f8fjPmx3nA2fG/raPCU5J1Teshcz4Tzy6M+fAXVLiFrUhpmN2tnPgyVqUqxt1ifji2gPEWZfWIeO+SrAW7om6UaXRqjLtXvH86/x0X88l04IDUuQ1ueOHd5+/F++5X7jRfUbeyC1C0ByvPtB5phe0X+GCIZabuwLtKt6rrI6X6IuBnxgr1DdxCV6/FoithK5bSFu9dVItejGMDMo74O8VC8iPc2rYj7kc3h8xosWSNweWixUh+E88eT6YrbKn/4fLyDMvDRvVu5WQd0ZtjhfYNbll5MOaFrQscdzFHVxZ9LdgVeaMEo1Pj/ZaZjBtfHmsh3lDYOYadjPvWNWry43ivlWKd93fitDkqF4qUj8otwIq44d2z82KFtyHVu8vqa3pOFgGfg486WuqfoK3e76VoFj3cuvstPtHsSvjw9KTCviaes3NU4haRWh1geSrYyG/i2bOp5xqiTWGj5m7l2TH9nonbt7g/WFGsrBR3dGXRrHlNYaOIo1Pj/dN+lDvg3aPbxOOTyIwWTeYATNZiLpRbwGrJ/Zancm1F2pLuOlEgzGwPfKDAZbh1Yl3cv+VvIYRXGnjPQXh3FfjC1e8WQNQmi5lZKHDGN7Ne+Gjgebh/0nlmtjHeBf4j4PIQwvlmtgvwQwjhpULGX0hSeXg6rrTdjFuNx4QQRqXOM4BCp2UlYmZd8HVzZ+Hzls3Ep0J4C+8ivQivMM/Au8zXCSFMqcQ0NLPd8ZGjn+FzGn5TZpEqimKln5l1whsFh+Ld2B/h67J+BVwUQrjJzIbixoRF+MjQmfHagpeJojhIaWskSWY3s5VCCFVmdjI+P80ruMWiCz40+uoQwlmNiGdX4LMQwqSCCC4KQqysj8On8FgVLzR/BlwWQrjAzDrgTtvtgatCCGfH65brytrMBuNdHYavcHB76j8V8CnMbDV8FHgHMg77t+MjvBfiA0/2xydyBp9+aG7JBS0gZrYbXh69X25ZKpFipJ+ZHYc3sObgvoZtcH/a0/HF038VFbdhwMchhCdrvJlYblmp3AJUMlFRSyqv1ePvU3gl3QZ36D4P7wpZx8yaNTSuEMJzUtiWL6LidTOZEVrzcEvKK8C5ZnZZCGEK3lU2M4YDrqwtz4pPCOEJfMDMQuBWMzsTpLDlIoQwH3/v3+Bd4q/ifqtf41M8HBNCuBV3m9i70hU2gBDCs1LYGk4h08/MfmpmV+KK2hm4T+UJuFJ4OXA0rsjdYGanhBD+HEJ40iKFkEGUDlnaGki68jKz4/H5q8AHClyBt7h/jjuZ7oavwTaxDKKKImBm7fBKejV8Hc6u+FD3E/FRo9fG35tCCL80s5YhhO8rTekxs73xiWD3CSH8q9zyLM+oW1mUmmjpfxIvhy4OIVxnZufiE/RW4YMOXozf8W3AoSGEZ8onsWgsUtoaSfQR+BPuo3AQvnblMDKTUO4F/DyE8HbZhBQFxcx+ivsqnRBC+JuZrY1b0bYgo7j1wAvJLfGBIxVrJTWz9UII0ytN4SwH6lYWpcLM1senD/kAH9X9Be4/WwX8Mm5z8PrneX3HKwZS2hqBma2Kz2E1C/dZaosPo+6OO5n+C58GZHq5ZBSFxcy64mtvPo/7Lm2BF44tyShuN+OK+1b4XH2vlkfawqLCPj+ixe1ufL62s0IIVyvtRKExs374SgrX4YaDa/Dl+6pwS/96+PyAi3BlbrbyYOXTvNwCrACsg3d/fgIcgq9ScCDwMP6RLCmjbKJAxC6t1XFFfDTuVH4/vgj0EnwepH74NBlDgVVCCMekr6/0ArPS5S8V0V/oaLxbeUIMU9qJQjMdn2bmFHzEaEfcWLAQn+h2V3y95Q9DCLPKJaQoLBqIkCdph82UT8oCfLQguBP6IPxjaQvcL4VtxSGOG/gOWBlYFJ3JL8XXkjwZX5ZoPj6p7mS8BVzt+tJKLMpJCOFxfEmif8nZWxSDEMIH+DqfK+OT5P4VX4fzctzy3zWEcEUI4TENOlhxkKUtD7KtJHGKj+YhhMVxWoe18YkL2+MtnYEhhA/LJa8oLGa2Kb5u30344IM1AUIIL5vZEryAPCb+noKvv7h4RbCuiYYjtwhRbEIId5nZc8A3caBTa9ziPw8YnzpP5dAKgnza6iCZfy3un4aPxhkSj9MjSH+GfyhfhxC+Lpe8orDEObjew6duOB8fPv9ffKTWB/Gcvrg/SW98LVmNEhZClJQ4avRIvKfn2BDCw2UWSRQBKW15YmYD8TnX3sYr7BkxfKlSJ1ZMzGxffFToImANfBLdqfj6nLNwi3UX4MAQwkPlklMI0TQxs5XwKab2Bf4SQhilqWVWTKS05YGZHYqPBpuFD59+Qcpa08LM9sSnclgDXzvwXXxahw3wAQonhBAeV0EphCgXZrZaCGG+yqEVFw1EyEEOh823cefytYGDzWz14EtWKf2aCCGEfwK/wBfLng88EkLoF0L4Kd4lKoVNCFFW4uocy/2KK6LhSOnIIr00lZmtb2ZtQwjvAHvjvk3HASPMbA1Z2poWIYSn8WldWgJ/MrNfxr/mxP9VUAohhCgaUtpSxIEFyaCDU/F1RN+M67W9DxyAd4udBZynIdRNj2hxOwqfC+ndGKapXYQQQhQd+bTlwMzOBi7E59o6MAaPAC7BHc5vwReBfq8sAoqyoyVhhBBClBopbVmYWU98JvPrgH2An+ATF24O/D6E8GszWyWE8EMZxRTLCVLahBBClIomr7SZ2Sb4UlQbAONCCF+b2f7AnsDhuA9TD3y6jznAZpo0UwghhBClpkn7tJnZAHwZoteAJ4FnzezkONfWR/g6onvgE6teD2wphU0IIYQQ5aDJWtrMbA/gIeA/wCSgDbAXrsieA3wI3A60Az4HdgohTC6PtEIIIYRo6jRJpc3MdgGexS1sB4UQPjWz9fBpPa7B12w7EJ+XbVvg+RDCx+WRVgghhBCi6XaPrhJ/O+PLfiSLO/8DeB5fcLdjCGFCCOEWKWxCCCGEKDdNUmkLITyFDzRoBlycTJIaF3qfA8zGu0SFEEIIIZYLmqTSBksVtwPx9SMvNLMzzawL0AlftmpOOeUTQgghhEjTJH3a0sQBCf8HVAFfAOsD24UQJpZVMCGEEEKIFE1eaQMws0H4lB8A3UII75ZTHiGEEEKIbKS0RcxsV+CzEMKkcssihBBCCJGNlDYhhBBCiAqgyQ5EEEIIIYSoJKS0CSGEEEJUAFLahBBCCCEqACltQgghhBAVgJQ2IcQKj5l1MLMQtyoz+8rMRppZg8tAMxsR7/fzQsoqhBA1IaVNCNGUeBM4AvgIOBXYvxH3ehA4GHitAHIJIUSdSGkTQjQlvggh/A34czz+sZm1NrPbzOxrM5thZjeb2WoAZtbNzN4xs2lmdkm0rI2K1/4c+DvQJ567Tzx3npm9a2b7xPD+8brHzOxFM5tjZn8q7WMLIVYEpLQJIZoSK5vZBkD/ePxf4BrgcOAO4BbgWODi+P8d+HrE1wDda7qpmW0OPACsDJwJNAceiOEJuwIPATOBs81s48Y/jhCiKdG83AIIIUQJGQB8GfevCSH828z+Dy8Lz0mfZ2atgJ7AKyGEP5jZpsBeNdx3N1xhuzKE8FczC8DNuKI2IZ7zWAjhWjPrBJwAbAJ8WsiHE0Ks2EhpE0I0JV7HrWa/B4aZ2R0x/Evc2pawELC4X59lY2o795v4uzj+NqvHfYUQQkqbEKJJMSOEcJ+ZzQcew7tBnwCOAvYG3gJ6AVUhhJfM7E1gWzM7B9i+lvs+CywCzjIzw7tIFwHPAW2L9TBCiKaFfNqEEE2OEMLjwFi8u/Ma3JftF8C1+MCCV+KpRwETgbOBd2PY7Bz3mwQcgFvRrgWqgF/EcCGEKAhaMF4IIWrAzPoDG+CDB44CDgVODSFcXz6phBBNFXWPCiFEzawLXB1/PwcuAm4oq0RCiCaLLG1CCCGEEBWAfNqEEEIIISoAKW1CCCGEEBWAlDYhhBBCiApASpsQQgghRAUgpU0IIYQQogKQ0iaEEEIIUQH8P8VpvB7MNQ5HAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<Figure size 720x360 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Grouping the data by 'region' and 'status_group', and then unstacking the results to make it suitable for plotting\n",
    "sub_df = data.groupby('region')['status_group'].value_counts(ascending=True).unstack()\n",
    "\n",
    "# Plotting the DataFrame using a bar plot\n",
    "sub_df.plot(kind='bar', figsize=(10,5))\n",
    "\n",
    "# Adding labels to the x-axis, y-axis, and a title\n",
    "plt.xlabel(\"Region\", fontweight='bold')\n",
    "plt.xticks(rotation=45, fontweight='bold')  # Rotating x-axis labels for better readability\n",
    "plt.ylabel(\"Pump Functionality Counts\")\n",
    "plt.title('Pump Functionality based on Region', fontsize=14, fontweight='bold')\n",
    "\n",
    "# Adding a legend with a custom location and title\n",
    "plt.legend(bbox_to_anchor=(1.0, 1.0), title='Water Quantity');\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "we noted a good number of functioning pumps in Iringa region "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Unique Values: 2\n",
      "Missing Values: 3334/59400\n",
      "Possible Missing Values: 0 (wells with '0' as the value)\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 360x360 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Analyzing the 'public_meeting' variable\n",
    "print(f\"Unique Values: {data['public_meeting'].nunique()}\")  # Counting the number of unique values in the 'public_meeting' column\n",
    "print(f\"Missing Values: {data['public_meeting'].isna().sum()}/{data.shape[0]}\")  # Counting missing values in the 'public_meeting' column\n",
    "print(f\"Possible Missing Values: {(data['public_meeting'] == '0').sum()} (wells with '0' as the value)\")  # Counting occurrences of '0' as a possible placeholder for missing values\n",
    "data['public_meeting'].value_counts().head(20)  # Displaying the top 20 most frequent values in the 'public_meeting' column\n",
    "\n",
    "# Grouping the data by 'public_meeting' and 'status_group', and then unstacking the results to make it suitable for plotting\n",
    "sub_df = data.groupby('public_meeting')['status_group'].value_counts(ascending=False).unstack()\n",
    "\n",
    "# Sorting the DataFrame by the 'public_meeting' column in descending order and plotting the result using a bar plot\n",
    "sub_df.sort_values(by='public_meeting', ascending=False).plot(kind='bar', figsize=(5,5))\n",
    "\n",
    "# Adding labels to the x-axis, y-axis, and a title\n",
    "plt.xlabel(\"Public Meeting\", fontweight='bold')\n",
    "plt.xticks(rotation=0, fontweight='bold')  # Rotating x-axis labels for better readability\n",
    "plt.ylabel(\"Public Meetings\")\n",
    "plt.title('Pump Functionality based on Public Meetings', fontsize=14, fontweight='bold')\n",
    "\n",
    "# Adding a legend with a custom location and title\n",
    "plt.legend(bbox_to_anchor=(1.0, 1.0), title='Status Group');\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We noted that meeting have an effect on functionality of the pumps "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Unique Values: 12\n",
      "Missing Values: 0/59400\n",
      "Possible Missing Values: 0 (wells with '0' as the value)\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "vwc                 40507\n",
       "wug                  6515\n",
       "water board          2933\n",
       "wua                  2535\n",
       "private operator     1971\n",
       "parastatal           1768\n",
       "water authority       904\n",
       "other                 844\n",
       "company               685\n",
       "unknown               561\n",
       "other - school         99\n",
       "trust                  78\n",
       "Name: management, dtype: int64"
      ]
     },
     "execution_count": 23,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Analyzing the 'management' variable\n",
    "print(f\"Unique Values: {data['management'].nunique()}\")  # Counting the number of unique values in the 'management' column\n",
    "print(f\"Missing Values: {data['management'].isna().sum()}/{data.shape[0]}\")  # Counting missing values in the 'management' column\n",
    "print(f\"Possible Missing Values: {(data['management'] == '0').sum()} (wells with '0' as the value)\")  # Counting occurrences of '0' as a possible placeholder for missing values\n",
    "data['management'].value_counts().head(20)  # Displaying the top 20 most frequent values in the 'management' column\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th>functional</th>\n",
       "      <th>needs repair</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>management_group</th>\n",
       "      <th>management</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th rowspan=\"4\" valign=\"top\">commercial</th>\n",
       "      <th>company</th>\n",
       "      <td>0.389781</td>\n",
       "      <td>0.610219</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>private operator</th>\n",
       "      <td>0.748858</td>\n",
       "      <td>0.251142</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>trust</th>\n",
       "      <td>0.589744</td>\n",
       "      <td>0.410256</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>water authority</th>\n",
       "      <td>0.493363</td>\n",
       "      <td>0.506637</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th rowspan=\"2\" valign=\"top\">other</th>\n",
       "      <th>other</th>\n",
       "      <td>0.598341</td>\n",
       "      <td>0.401659</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>other - school</th>\n",
       "      <td>0.232323</td>\n",
       "      <td>0.767677</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>parastatal</th>\n",
       "      <th>parastatal</th>\n",
       "      <td>0.576923</td>\n",
       "      <td>0.423077</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>unknown</th>\n",
       "      <th>unknown</th>\n",
       "      <td>0.399287</td>\n",
       "      <td>0.600713</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th rowspan=\"4\" valign=\"top\">user-group</th>\n",
       "      <th>vwc</th>\n",
       "      <td>0.504234</td>\n",
       "      <td>0.495766</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>water board</th>\n",
       "      <td>0.739857</td>\n",
       "      <td>0.260143</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>wua</th>\n",
       "      <td>0.690730</td>\n",
       "      <td>0.309270</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>wug</th>\n",
       "      <td>0.599540</td>\n",
       "      <td>0.400460</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                   functional  needs repair\n",
       "management_group management                                \n",
       "commercial       company             0.389781      0.610219\n",
       "                 private operator    0.748858      0.251142\n",
       "                 trust               0.589744      0.410256\n",
       "                 water authority     0.493363      0.506637\n",
       "other            other               0.598341      0.401659\n",
       "                 other - school      0.232323      0.767677\n",
       "parastatal       parastatal          0.576923      0.423077\n",
       "unknown          unknown             0.399287      0.600713\n",
       "user-group       vwc                 0.504234      0.495766\n",
       "                 water board         0.739857      0.260143\n",
       "                 wua                 0.690730      0.309270\n",
       "                 wug                 0.599540      0.400460"
      ]
     },
     "execution_count": 24,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Grouping the data by 'management_group' and 'management', and then calculating the mean of 'functional' and 'needs repair' variables\n",
    "data.groupby(['management_group', 'management'])[['functional', 'needs repair']].mean()\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "\"Pumps managed by private operators and the water board have a good percentage of functioning, which is approximately 74.88% and 73.98%, respectively, whereas other pumps need repair.\""
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Unique Values: 7\n",
      "Missing Values: 0/59400\n",
      "Possible Missing Values: 0 (wells with '0' as the value)\n",
      "Unique Values: 7\n",
      "Missing Values: 0/59400\n",
      "Possible Missing Values: 0 (wells with '0' as the value)\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "never pay     25348\n",
       "per bucket     8985\n",
       "monthly        8300\n",
       "unknown        8157\n",
       "on failure     3914\n",
       "annually       3642\n",
       "other          1054\n",
       "Name: payment_type, dtype: int64"
      ]
     },
     "execution_count": 25,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Analysis of the 'payment' variable\n",
    "print(f\"Unique Values: {data['payment'].nunique()}\")  # Counting the number of unique values in the 'payment' column\n",
    "print(f\"Missing Values: {data['payment'].isna().sum()}/{data.shape[0]}\")  # Counting missing values in the 'payment' column\n",
    "print(f\"Possible Missing Values: {(data['payment'] == '0').sum()} (wells with '0' as the value)\")  # Counting occurrences of '0' as a possible placeholder for missing values\n",
    "data['payment'].value_counts().head(20)  # Displaying the top 20 most frequent values in the 'payment' column\n",
    "\n",
    "# Analysis of the 'payment_type' variable\n",
    "print(f\"Unique Values: {data['payment_type'].nunique()}\")  # Counting the number of unique values in the 'payment_type' column\n",
    "print(f\"Missing Values: {data['payment_type'].isna().sum()}/{data.shape[0]}\")  # Counting missing values in the 'payment_type' column\n",
    "print(f\"Possible Missing Values: {(data['payment_type'] == '0').sum()} (wells with '0' as the value)\")  # Counting occurrences of '0' as a possible placeholder for missing values\n",
    "data['payment_type'].value_counts().head(20)  # Displaying the top 20 most frequent values in the 'payment_type' column\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The analysis reveals that both the 'payment' and 'payment_type' variables have seven unique values, with no missing values or occurrences of '0' as a placeholder. The most common payment methods include 'never pay', 'per bucket', 'monthly', and 'unknown', indicating varied payment structures within the dataset, with 'never pay' being the most prevalent method across both variables."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 3. Exploration On  Numeric variables \n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Unique Values: 2428\n",
      "Missing Values: 0/59400\n",
      "Possible Missing Values: 0 (wells with '0' as the value)\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       " 0       20438\n",
       "-15         60\n",
       "-16         55\n",
       "-13         55\n",
       "-20         52\n",
       " 1290       52\n",
       "-14         51\n",
       " 303        51\n",
       "-18         49\n",
       "-19         47\n",
       " 1269       46\n",
       " 1295       46\n",
       " 1304       45\n",
       "-23         45\n",
       " 280        44\n",
       " 1538       44\n",
       " 1286       44\n",
       "-8          44\n",
       "-17         44\n",
       " 1332       43\n",
       "Name: gps_height, dtype: int64"
      ]
     },
     "execution_count": 26,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# GPS-HEIGHT\n",
    "# Counting the number of unique altitude values in the 'gps_height' column\n",
    "print(f\"Unique Values: {data['gps_height'].nunique()}\")\n",
    "\n",
    "# Counting missing values in the 'gps_height' column and displaying the count out of the total number of records\n",
    "print(f\"Missing Values: {data['gps_height'].isna().sum()}/{data.shape[0]}\")\n",
    "\n",
    "# Counting occurrences where the altitude is recorded as '0', possibly indicating missing or unknown values\n",
    "print(f\"Possible Missing Values: {(data['gps_height'] == '0').sum()} (wells with '0' as the value)\")\n",
    "\n",
    "# Displaying the top 20 most frequent altitude values\n",
    "data['gps_height'].value_counts().head(20)\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The analysis of the 'gps_height' variable indicates a diverse range of altitude values with no missing or potentially missing data, with the most frequent altitude recorded at 0 meters and a variety of other altitudes present in the dataset."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 576x432 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig, ax = plt.subplots(figsize=(8, 6))  # Create a figure and axes object\n",
    "fig.suptitle('Logistic Regression', fontsize=16)  # Add a title to the figure\n",
    "sns.scatterplot(x='gps_height', y='target', data=data[(data['gps_height'] != 0)], ax=ax)  # Scatter plot of GPS height vs. target\n",
    "sns.regplot(x='gps_height', y='target', data=data[(data['gps_height'] != 0)], logistic=True, scatter=False, color='black', ax=ax)  # Logistic regression line\n",
    "ax.set_xlabel('GPS Height', fontsize=14)  # Set x-axis label\n",
    "ax.set_ylabel('Well Status', fontsize=14)  # Set y-axis label\n",
    "plt.show()  # Show the plot\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Unique Values: 98\n",
      "Missing Values: 0/59400\n",
      "Possible Missing Values: 0 (wells with '0' as the value)\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "0.0       41639\n",
       "500.0      3102\n",
       "50.0       2472\n",
       "1000.0     1488\n",
       "20.0       1463\n",
       "200.0      1220\n",
       "100.0       816\n",
       "10.0        806\n",
       "30.0        743\n",
       "2000.0      704\n",
       "250.0       569\n",
       "300.0       557\n",
       "5000.0      450\n",
       "5.0         376\n",
       "25.0        356\n",
       "3000.0      334\n",
       "1200.0      267\n",
       "1500.0      197\n",
       "6.0         190\n",
       "600.0       176\n",
       "Name: amount_tsh, dtype: int64"
      ]
     },
     "execution_count": 28,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#AMOUNT-TSH\n",
    "print(f\"Unique Values: {data['amount_tsh'].nunique()}\")\n",
    "print(f\"Missing Values: {data['amount_tsh'].isna().sum()}/{data.shape[0]}\")\n",
    "print(f\"Possible Missing Values: {(data['amount_tsh'] == '0').sum()} (wells with '0' as the value)\")\n",
    "data['amount_tsh'].value_counts().head(20)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Since 90% of the data is 0's, hence we will be dropping this when modelling."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Unique Values: 55\n",
      "Missing Values: 0/59400\n",
      "Possible Missing Values: 0 (wells with '0' as the value)\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "0       20709\n",
       "2010     2645\n",
       "2008     2613\n",
       "2009     2533\n",
       "2000     2091\n",
       "2007     1587\n",
       "2006     1471\n",
       "2003     1286\n",
       "2011     1256\n",
       "2004     1123\n",
       "2012     1084\n",
       "2002     1075\n",
       "1978     1037\n",
       "1995     1014\n",
       "2005     1011\n",
       "1999      979\n",
       "1998      966\n",
       "1990      954\n",
       "1985      945\n",
       "1980      811\n",
       "Name: construction_year, dtype: int64"
      ]
     },
     "execution_count": 29,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# CONSTRUCTION YEAR\n",
    "# Counting the number of unique construction years in the dataset\n",
    "print(f\"Unique Values: {data['construction_year'].nunique()}\")\n",
    "\n",
    "# Counting missing values in the 'construction_year' column and displaying the count out of the total number of records\n",
    "print(f\"Missing Values: {data['construction_year'].isna().sum()}/{data.shape[0]}\")\n",
    "\n",
    "# Counting occurrences where the construction year is recorded as '0', possibly indicating missing or unknown values\n",
    "print(f\"Possible Missing Values: {(data['construction_year'] == '0').sum()} (wells with '0' as the value)\")\n",
    "\n",
    "# Displaying the top 20 most frequent construction years\n",
    "data['construction_year'].value_counts().head(20)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "2024    20709\n",
       "14       2645\n",
       "16       2613\n",
       "15       2533\n",
       "24       2091\n",
       "17       1587\n",
       "18       1471\n",
       "21       1286\n",
       "13       1256\n",
       "20       1123\n",
       "12       1084\n",
       "22       1075\n",
       "46       1037\n",
       "29       1014\n",
       "19       1011\n",
       "25        979\n",
       "26        966\n",
       "34        954\n",
       "39        945\n",
       "44        811\n",
       "28        811\n",
       "40        779\n",
       "42        744\n",
       "30        738\n",
       "52        708\n",
       "50        676\n",
       "27        644\n",
       "32        640\n",
       "31        608\n",
       "23        540\n",
       "36        521\n",
       "41        488\n",
       "49        437\n",
       "38        434\n",
       "48        414\n",
       "54        411\n",
       "33        324\n",
       "35        316\n",
       "37        302\n",
       "43        238\n",
       "47        202\n",
       "45        192\n",
       "51        184\n",
       "11        176\n",
       "53        145\n",
       "64        102\n",
       "57         88\n",
       "61         85\n",
       "56         77\n",
       "55         59\n",
       "60         40\n",
       "62         30\n",
       "63         21\n",
       "59         19\n",
       "58         17\n",
       "Name: age_of_well, dtype: int64"
      ]
     },
     "execution_count": 30,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from datetime import datetime\n",
    "data['age_of_well']=datetime.now().year - data['construction_year']\n",
    "\n",
    "data['age_of_well'].value_counts()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>id</th>\n",
       "      <th>status_group</th>\n",
       "      <th>amount_tsh</th>\n",
       "      <th>date_recorded</th>\n",
       "      <th>funder</th>\n",
       "      <th>gps_height</th>\n",
       "      <th>installer</th>\n",
       "      <th>longitude</th>\n",
       "      <th>latitude</th>\n",
       "      <th>wpt_name</th>\n",
       "      <th>...</th>\n",
       "      <th>quantity_group</th>\n",
       "      <th>source</th>\n",
       "      <th>source_type</th>\n",
       "      <th>source_class</th>\n",
       "      <th>waterpoint_type</th>\n",
       "      <th>waterpoint_type_group</th>\n",
       "      <th>target</th>\n",
       "      <th>functional</th>\n",
       "      <th>needs repair</th>\n",
       "      <th>age_of_well</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>69572</td>\n",
       "      <td>functional</td>\n",
       "      <td>6000.0</td>\n",
       "      <td>2011-03-14</td>\n",
       "      <td>Roman</td>\n",
       "      <td>1390</td>\n",
       "      <td>Roman</td>\n",
       "      <td>34.938093</td>\n",
       "      <td>-9.856322</td>\n",
       "      <td>none</td>\n",
       "      <td>...</td>\n",
       "      <td>enough</td>\n",
       "      <td>spring</td>\n",
       "      <td>spring</td>\n",
       "      <td>groundwater</td>\n",
       "      <td>communal standpipe</td>\n",
       "      <td>communal standpipe</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>25</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>8776</td>\n",
       "      <td>functional</td>\n",
       "      <td>0.0</td>\n",
       "      <td>2013-03-06</td>\n",
       "      <td>Grumeti</td>\n",
       "      <td>1399</td>\n",
       "      <td>GRUMETI</td>\n",
       "      <td>34.698766</td>\n",
       "      <td>-2.147466</td>\n",
       "      <td>Zahanati</td>\n",
       "      <td>...</td>\n",
       "      <td>insufficient</td>\n",
       "      <td>rainwater harvesting</td>\n",
       "      <td>rainwater harvesting</td>\n",
       "      <td>surface</td>\n",
       "      <td>communal standpipe</td>\n",
       "      <td>communal standpipe</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>14</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>34310</td>\n",
       "      <td>functional</td>\n",
       "      <td>25.0</td>\n",
       "      <td>2013-02-25</td>\n",
       "      <td>Lottery Club</td>\n",
       "      <td>686</td>\n",
       "      <td>World vision</td>\n",
       "      <td>37.460664</td>\n",
       "      <td>-3.821329</td>\n",
       "      <td>Kwa Mahundi</td>\n",
       "      <td>...</td>\n",
       "      <td>enough</td>\n",
       "      <td>dam</td>\n",
       "      <td>dam</td>\n",
       "      <td>surface</td>\n",
       "      <td>communal standpipe multiple</td>\n",
       "      <td>communal standpipe</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>15</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>67743</td>\n",
       "      <td>needs repair</td>\n",
       "      <td>0.0</td>\n",
       "      <td>2013-01-28</td>\n",
       "      <td>Unicef</td>\n",
       "      <td>263</td>\n",
       "      <td>UNICEF</td>\n",
       "      <td>38.486161</td>\n",
       "      <td>-11.155298</td>\n",
       "      <td>Zahanati Ya Nanyumbu</td>\n",
       "      <td>...</td>\n",
       "      <td>dry</td>\n",
       "      <td>machine dbh</td>\n",
       "      <td>borehole</td>\n",
       "      <td>groundwater</td>\n",
       "      <td>communal standpipe multiple</td>\n",
       "      <td>communal standpipe</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>38</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10</th>\n",
       "      <td>49056</td>\n",
       "      <td>functional</td>\n",
       "      <td>0.0</td>\n",
       "      <td>2011-02-20</td>\n",
       "      <td>Private</td>\n",
       "      <td>62</td>\n",
       "      <td>Private</td>\n",
       "      <td>39.209518</td>\n",
       "      <td>-7.034139</td>\n",
       "      <td>Mzee Hokororo</td>\n",
       "      <td>...</td>\n",
       "      <td>enough</td>\n",
       "      <td>machine dbh</td>\n",
       "      <td>borehole</td>\n",
       "      <td>groundwater</td>\n",
       "      <td>other</td>\n",
       "      <td>other</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>13</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 45 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "       id  status_group  amount_tsh date_recorded        funder  gps_height  \\\n",
       "0   69572    functional      6000.0    2011-03-14         Roman        1390   \n",
       "1    8776    functional         0.0    2013-03-06       Grumeti        1399   \n",
       "2   34310    functional        25.0    2013-02-25  Lottery Club         686   \n",
       "3   67743  needs repair         0.0    2013-01-28        Unicef         263   \n",
       "10  49056    functional         0.0    2011-02-20       Private          62   \n",
       "\n",
       "       installer  longitude   latitude              wpt_name  ...  \\\n",
       "0          Roman  34.938093  -9.856322                  none  ...   \n",
       "1        GRUMETI  34.698766  -2.147466              Zahanati  ...   \n",
       "2   World vision  37.460664  -3.821329           Kwa Mahundi  ...   \n",
       "3         UNICEF  38.486161 -11.155298  Zahanati Ya Nanyumbu  ...   \n",
       "10       Private  39.209518  -7.034139         Mzee Hokororo  ...   \n",
       "\n",
       "    quantity_group                source           source_type source_class  \\\n",
       "0           enough                spring                spring  groundwater   \n",
       "1     insufficient  rainwater harvesting  rainwater harvesting      surface   \n",
       "2           enough                   dam                   dam      surface   \n",
       "3              dry           machine dbh              borehole  groundwater   \n",
       "10          enough           machine dbh              borehole  groundwater   \n",
       "\n",
       "                waterpoint_type  waterpoint_type_group target functional  \\\n",
       "0            communal standpipe     communal standpipe      0          1   \n",
       "1            communal standpipe     communal standpipe      0          1   \n",
       "2   communal standpipe multiple     communal standpipe      0          1   \n",
       "3   communal standpipe multiple     communal standpipe      1          0   \n",
       "10                        other                  other      0          1   \n",
       "\n",
       "    needs repair age_of_well  \n",
       "0              0          25  \n",
       "1              0          14  \n",
       "2              0          15  \n",
       "3              1          38  \n",
       "10             0          13  \n",
       "\n",
       "[5 rows x 45 columns]"
      ]
     },
     "execution_count": 31,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Filtering the DataFrame 'data' to include only rows where 'construction_year', 'population', and 'gps_height' columns have values greater than 0\n",
    "data = data[(data['construction_year'] > 0) & (data['population'] > 0) & (data['gps_height'] > 0)]\n",
    "data.head()\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "sns.boxplot(x='status_group',y='age_of_well', data=data, showfliers=False)\n",
    "plt.xlabel('Functionality of Well', fontsize=12, fontweight='bold');\n",
    "plt.ylabel('Age of Well', fontsize=12, fontweight='bold');\n",
    "plt.title('Age of well vs Functionality', fontsize=16, fontweight='bold');\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "0 is Funtional 1 is Non Funtional or Needs repair\n",
    "\n",
    "So we see, that newer the well is, more chances of it being functional"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {},
   "outputs": [],
   "source": [
    "data = data.drop_duplicates()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {},
   "outputs": [],
   "source": [
    "to_drop = ['wpt_name','id','date_recorded','recorded_by','region_code',\n",
    "           'lga','scheme_name','funder','subvillage','latitude',\n",
    "           'longitude','extraction_type_class','extraction_type',\n",
    "           'payment_type','quantity_group','source_type','waterpoint_type_group',\n",
    "           'district_code','amount_tsh','num_private','construction_year',\n",
    "           'status_group','population','functional','needs repair','management'\n",
    "          ]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'pandas.core.frame.DataFrame'>\n",
      "Int64Index: 29361 entries, 0 to 59399\n",
      "Data columns (total 19 columns):\n",
      " #   Column                 Non-Null Count  Dtype \n",
      "---  ------                 --------------  ----- \n",
      " 0   gps_height             29361 non-null  int64 \n",
      " 1   installer              29361 non-null  object\n",
      " 2   basin                  29361 non-null  object\n",
      " 3   region                 29361 non-null  object\n",
      " 4   ward                   29361 non-null  object\n",
      " 5   public_meeting         29361 non-null  object\n",
      " 6   scheme_management      29361 non-null  object\n",
      " 7   permit                 29361 non-null  object\n",
      " 8   extraction_type_group  29361 non-null  object\n",
      " 9   management_group       29361 non-null  object\n",
      " 10  payment                29361 non-null  object\n",
      " 11  water_quality          29361 non-null  object\n",
      " 12  quality_group          29361 non-null  object\n",
      " 13  quantity               29361 non-null  object\n",
      " 14  source                 29361 non-null  object\n",
      " 15  source_class           29361 non-null  object\n",
      " 16  waterpoint_type        29361 non-null  object\n",
      " 17  target                 29361 non-null  int8  \n",
      " 18  age_of_well            29361 non-null  int64 \n",
      "dtypes: int64(2), int8(1), object(16)\n",
      "memory usage: 4.3+ MB\n"
     ]
    }
   ],
   "source": [
    "data = data.drop(to_drop, axis=1)\n",
    "data = data.dropna()\n",
    "data.info()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### MODELING\n",
    "#### Train Test Split\n",
    ". The train-test split is a technique for evaluating the performance of a machine Learning algoritham.\n",
    "train Dataset Used to fit the machine learning model\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "metadata": {},
   "outputs": [],
   "source": [
    "X = data.drop(['target'], axis=1)  # Assigning features to X by dropping the 'target' column\n",
    "y = data['target']  # Assigning the 'target' variable to y\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "['gps_height', 'age_of_well'] ['installer', 'basin', 'region', 'ward', 'public_meeting', 'scheme_management', 'permit', 'extraction_type_group', 'management_group', 'payment', 'water_quality', 'quality_group', 'quantity', 'source', 'source_class', 'waterpoint_type']\n"
     ]
    }
   ],
   "source": [
    "# Extracting numerical columns from the feature dataset 'X' and converting them to a list\n",
    "numerical_cols = list(X.select_dtypes(include=np.number).columns)\n",
    "\n",
    "# Extracting categorical columns from the feature dataset 'X' and converting them to a list\n",
    "categorical_cols = list(X.select_dtypes(exclude=np.number).columns)\n",
    "print(numerical_cols, categorical_cols)\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Preprocessing  Data \n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.compose import ColumnTransformer\n",
    "from sklearn.pipeline import Pipeline\n",
    "from sklearn.impute import SimpleImputer\n",
    "from sklearn.preprocessing import OneHotEncoder, StandardScaler\n",
    "\n",
    "# Preprocessing for numerical data\n",
    "numerical_transformer = Pipeline(steps=[\n",
    "    ('scale', StandardScaler())\n",
    "])\n",
    "\n",
    "# Preprocessing for categorical data\n",
    "categorical_transformer = Pipeline(steps=[\n",
    "    ('imputer', SimpleImputer(strategy='most_frequent')),\n",
    "    ('onehot', OneHotEncoder(handle_unknown='ignore'))\n",
    "])\n",
    "\n",
    "# Bundle preprocessing for numerical and categorical data\n",
    "preprocessor = ColumnTransformer(\n",
    "    transformers=[\n",
    "        ('num', numerical_transformer, numerical_cols),\n",
    "        ('cat', categorical_transformer, categorical_cols)\n",
    "    ])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.metrics import mean_absolute_error\n",
    "from sklearn.model_selection import GridSearchCV\n",
    "from imblearn.over_sampling import SMOTE\n",
    "from imblearn.under_sampling import RandomUnderSampler\n",
    "from imblearn.pipeline import Pipeline\n",
    "from sklearn.decomposition import TruncatedSVD\n",
    "import xgboost as xgboost\n",
    "\n",
    "def fit_predict(model, X_train, X_test, y_train, y_test):\n",
    "    '''fit pipeline using given model, and return predictions'''\n",
    "    \n",
    "    param_grid = model['params']\n",
    "    model = model['model']\n",
    "        \n",
    "    \n",
    "    my_pipeline = Pipeline(steps=[('preprocessor', preprocessor),\n",
    "                                  ('model', model)\n",
    "                                 ])\n",
    "\n",
    "    search = GridSearchCV(estimator=my_pipeline,\n",
    "             param_grid=param_grid, n_jobs=-1, verbose=2, cv=10)\n",
    "    \n",
    "    search.fit(X_train, y_train)\n",
    "    \n",
    "    best_estimator = search.best_estimator_._final_estimator\n",
    "    \n",
    "    print(\"Best parameter (CV score=%0.3f):\" % search.best_score_)\n",
    "    print(search.best_params_)\n",
    "\n",
    "    # Preprocessing of validation data, get predictions\n",
    "    test_preds = search.predict(X_test)\n",
    "    train_preds = search.predict(X_train)\n",
    "    \n",
    "    return test_preds, train_preds, search"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.metrics import accuracy_score\n",
    "from sklearn.metrics import classification_report\n",
    "from sklearn.metrics import confusion_matrix\n",
    "import itertools\n",
    "import matplotlib.pyplot as plt\n",
    "%matplotlib inline\n",
    "\n",
    "def plot_confusion_matrix(y_true, y_preds):\n",
    "    # Print confusion matrix\n",
    "    cnf_matrix = confusion_matrix(y_true, y_preds)\n",
    "    # Create the basic matrix\n",
    "    plt.imshow(cnf_matrix,  cmap=plt.cm.Blues)\n",
    "    # Add title and axis labels\n",
    "    plt.title('Confusion Matrix')\n",
    "    plt.ylabel('True label')\n",
    "    plt.xlabel('Predicted label')\n",
    "    # Add appropriate axis scales\n",
    "    class_names = set(y) # Get class labels to add to matrix\n",
    "    tick_marks = np.arange(len(class_names))\n",
    "    plt.xticks(tick_marks, class_names, rotation=0)\n",
    "    plt.yticks(tick_marks, class_names)\n",
    "    # Add labels to each cell\n",
    "    thresh = cnf_matrix.max() / 2. # Used for text coloring below\n",
    "    # Here we iterate through the confusion matrix and append labels to our visualization\n",
    "    for i, j in itertools.product(range(cnf_matrix.shape[0]), range(cnf_matrix.shape[1])):\n",
    "            plt.text(j, i, cnf_matrix[i, j],\n",
    "                     horizontalalignment='center',\n",
    "                     color='white' if cnf_matrix[i, j] > thresh else 'black')\n",
    "    # Add a legend\n",
    "    plt.colorbar();\n",
    "    plt.show();\n",
    "\n",
    "def metrics(model_name, y_train, y_test, y_train_pred, y_test_pred):\n",
    "    '''Print out the evaluation metrics for a given models predictions'''\n",
    "    print(f'Model: {model_name}', )\n",
    "    print('-'*60)\n",
    "    plot_confusion_matrix(y_test,y_test_pred)\n",
    "    print(f'test accuracy: {accuracy_score(y_test, y_test_pred)}')\n",
    "    print(f'train accuracy: {accuracy_score(y_train, y_train_pred)}')\n",
    "    print('-'*60)\n",
    "    print('\\ntest report:\\n' + classification_report(y_test, y_test_pred))\n",
    "    print('~'*60)\n",
    "    print('\\ntrain report:\\n' + classification_report(y_train, y_train_pred))    \n",
    "    print('-'*60)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "metadata": {},
   "outputs": [],
   "source": [
    "def roc(X_test, y_test, pred_y, model):\n",
    "    # Extracting the name of the model from the pipeline\n",
    "    name = str(model.best_estimator_.named_steps[\"model\"])[:str(model.best_estimator_.named_steps[\"model\"]).index('(')]\n",
    "    \n",
    "    # Predicting probabilities for positive class\n",
    "    y_pred_proba = model.predict_proba(X_test)[:,1]\n",
    "    \n",
    "    # Calculating true positive rate (TPR) and false positive rate (FPR)\n",
    "    fpr, tpr, threshold = roc_curve(y_test, y_pred_proba)\n",
    "    \n",
    "    # Plotting ROC curve\n",
    "    plt.plot(fpr, tpr, label=model)\n",
    "    plt.xlabel('False Positive Rate')\n",
    "    plt.ylabel('True Positive Rate')\n",
    "    plt.title(f'ROC curve for {name}')\n",
    "    plt.show()\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Calculate the smallest number of samples among the two classes\n",
    "smallest_num = data['target'].value_counts().sort_values().values[0]\n",
    "\n",
    "# Randomly sample the subset of data where the target label is 0 to match the smallest number of samples\n",
    "target_0 = data[data['target'] == 0].sample(smallest_num)\n",
    "\n",
    "# Randomly sample the subset of data where the target label is 1 to match the smallest number of samples\n",
    "target_1 = data[data['target'] == 1].sample(smallest_num)\n",
    "\n",
    "# Concatenate the sampled subsets for both target labels to create a balanced dataset\n",
    "sampled_df = pd.concat([target_0, target_1])\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "1    12664\n",
       "0    12664\n",
       "Name: target, dtype: int64"
      ]
     },
     "execution_count": 43,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Display the count of each class (0 and 1) in the 'target' column of the 'sampled_df' DataFrame\n",
    "sampled_df['target'].value_counts()\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 44,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Separate the features and the target variable from the 'sampled_df' DataFrame\n",
    "X_sampled_df = sampled_df.drop('target', axis=1)\n",
    "y_sampled_df = sampled_df['target']\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 45,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>gps_height</th>\n",
       "      <th>installer</th>\n",
       "      <th>basin</th>\n",
       "      <th>region</th>\n",
       "      <th>ward</th>\n",
       "      <th>public_meeting</th>\n",
       "      <th>scheme_management</th>\n",
       "      <th>permit</th>\n",
       "      <th>extraction_type_group</th>\n",
       "      <th>management_group</th>\n",
       "      <th>payment</th>\n",
       "      <th>water_quality</th>\n",
       "      <th>quality_group</th>\n",
       "      <th>quantity</th>\n",
       "      <th>source</th>\n",
       "      <th>source_class</th>\n",
       "      <th>waterpoint_type</th>\n",
       "      <th>age_of_well</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>33529</th>\n",
       "      <td>1518</td>\n",
       "      <td>DWE</td>\n",
       "      <td>Internal</td>\n",
       "      <td>Manyara</td>\n",
       "      <td>Gidas</td>\n",
       "      <td>True</td>\n",
       "      <td>VWC</td>\n",
       "      <td>True</td>\n",
       "      <td>nira/tanira</td>\n",
       "      <td>user-group</td>\n",
       "      <td>never pay</td>\n",
       "      <td>soft</td>\n",
       "      <td>good</td>\n",
       "      <td>enough</td>\n",
       "      <td>shallow well</td>\n",
       "      <td>groundwater</td>\n",
       "      <td>hand pump</td>\n",
       "      <td>20</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12218</th>\n",
       "      <td>1527</td>\n",
       "      <td>Oikos E .Africa</td>\n",
       "      <td>Internal</td>\n",
       "      <td>Arusha</td>\n",
       "      <td>Ngarenanyuki</td>\n",
       "      <td>True</td>\n",
       "      <td>VWC</td>\n",
       "      <td>False</td>\n",
       "      <td>gravity</td>\n",
       "      <td>user-group</td>\n",
       "      <td>unknown</td>\n",
       "      <td>soft</td>\n",
       "      <td>good</td>\n",
       "      <td>enough</td>\n",
       "      <td>machine dbh</td>\n",
       "      <td>groundwater</td>\n",
       "      <td>communal standpipe</td>\n",
       "      <td>14</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>26614</th>\n",
       "      <td>1844</td>\n",
       "      <td>Chamavita</td>\n",
       "      <td>Pangani</td>\n",
       "      <td>Tanga</td>\n",
       "      <td>Kwai</td>\n",
       "      <td>True</td>\n",
       "      <td>VWC</td>\n",
       "      <td>True</td>\n",
       "      <td>gravity</td>\n",
       "      <td>user-group</td>\n",
       "      <td>pay monthly</td>\n",
       "      <td>soft</td>\n",
       "      <td>good</td>\n",
       "      <td>enough</td>\n",
       "      <td>spring</td>\n",
       "      <td>groundwater</td>\n",
       "      <td>communal standpipe</td>\n",
       "      <td>36</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>49474</th>\n",
       "      <td>1327</td>\n",
       "      <td>TWESA</td>\n",
       "      <td>Lake Tanganyika</td>\n",
       "      <td>Kigoma</td>\n",
       "      <td>Busagara</td>\n",
       "      <td>True</td>\n",
       "      <td>VWC</td>\n",
       "      <td>False</td>\n",
       "      <td>gravity</td>\n",
       "      <td>user-group</td>\n",
       "      <td>unknown</td>\n",
       "      <td>soft</td>\n",
       "      <td>good</td>\n",
       "      <td>enough</td>\n",
       "      <td>river</td>\n",
       "      <td>surface</td>\n",
       "      <td>communal standpipe</td>\n",
       "      <td>21</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>46475</th>\n",
       "      <td>2014</td>\n",
       "      <td>Active MKM</td>\n",
       "      <td>Lake Victoria</td>\n",
       "      <td>Arusha</td>\n",
       "      <td>Arash</td>\n",
       "      <td>True</td>\n",
       "      <td>Parastatal</td>\n",
       "      <td>False</td>\n",
       "      <td>gravity</td>\n",
       "      <td>parastatal</td>\n",
       "      <td>never pay</td>\n",
       "      <td>soft</td>\n",
       "      <td>good</td>\n",
       "      <td>seasonal</td>\n",
       "      <td>rainwater harvesting</td>\n",
       "      <td>surface</td>\n",
       "      <td>communal standpipe</td>\n",
       "      <td>13</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "       gps_height        installer            basin   region          ward  \\\n",
       "33529        1518              DWE         Internal  Manyara         Gidas   \n",
       "12218        1527  Oikos E .Africa         Internal   Arusha  Ngarenanyuki   \n",
       "26614        1844        Chamavita          Pangani    Tanga          Kwai   \n",
       "49474        1327            TWESA  Lake Tanganyika   Kigoma      Busagara   \n",
       "46475        2014       Active MKM    Lake Victoria   Arusha         Arash   \n",
       "\n",
       "      public_meeting scheme_management permit extraction_type_group  \\\n",
       "33529           True               VWC   True           nira/tanira   \n",
       "12218           True               VWC  False               gravity   \n",
       "26614           True               VWC   True               gravity   \n",
       "49474           True               VWC  False               gravity   \n",
       "46475           True        Parastatal  False               gravity   \n",
       "\n",
       "      management_group      payment water_quality quality_group  quantity  \\\n",
       "33529       user-group    never pay          soft          good    enough   \n",
       "12218       user-group      unknown          soft          good    enough   \n",
       "26614       user-group  pay monthly          soft          good    enough   \n",
       "49474       user-group      unknown          soft          good    enough   \n",
       "46475       parastatal    never pay          soft          good  seasonal   \n",
       "\n",
       "                     source source_class     waterpoint_type  age_of_well  \n",
       "33529          shallow well  groundwater           hand pump           20  \n",
       "12218           machine dbh  groundwater  communal standpipe           14  \n",
       "26614                spring  groundwater  communal standpipe           36  \n",
       "49474                 river      surface  communal standpipe           21  \n",
       "46475  rainwater harvesting      surface  communal standpipe           13  "
      ]
     },
     "execution_count": 45,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from sklearn.model_selection import train_test_split\n",
    "\n",
    "X_train, X_test, y_train, y_test = train_test_split(X_sampled_df, y_sampled_df, test_size=.3, random_state=42)\n",
    "\n",
    "X_sampled_df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 46,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "25328"
      ]
     },
     "execution_count": 46,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# no of raws \n",
    "X_sampled_df.shape[0]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Classification Algorithms\n",
    "### 1) LogisticRegression\n",
    "Logistic regression is one of the most popular machine learning algorithms,and its used for predict\n",
    "the categorical dependent variable using a given set of independent variable"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 47,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Accuracy: 0.7849717068035268\n",
      "ROC AUC: 0.7847149742189303\n",
      "Best hyperparameters: {'model__C': 1}\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Users\\johnn\\anaconda3\\envs\\learn-env\\lib\\site-packages\\sklearn\\linear_model\\_logistic.py:762: ConvergenceWarning: lbfgs failed to converge (status=1):\n",
      "STOP: TOTAL NO. of ITERATIONS REACHED LIMIT.\n",
      "\n",
      "Increase the number of iterations (max_iter) or scale the data as shown in:\n",
      "    https://scikit-learn.org/stable/modules/preprocessing.html\n",
      "Please also refer to the documentation for alternative solver options:\n",
      "    https://scikit-learn.org/stable/modules/linear_model.html#logistic-regression\n",
      "  n_iter_i = _check_optimize_result(\n"
     ]
    }
   ],
   "source": [
    "from sklearn.linear_model import LogisticRegression\n",
    "from sklearn.pipeline import Pipeline\n",
    "from sklearn.model_selection import GridSearchCV\n",
    "from sklearn.metrics import accuracy_score, roc_auc_score\n",
    "\n",
    "# Define logistic regression model\n",
    "logistic_regression = LogisticRegression(random_state=42)\n",
    "\n",
    "# Define hyperparameters to tune\n",
    "param_grid = {\n",
    "    'model__C': [0.001, 0.01, 0.1, 1, 10, 100]  # regularization parameter\n",
    "}\n",
    "\n",
    "# Create a pipeline\n",
    "pipeline = Pipeline(steps=[\n",
    "    ('preprocessor', preprocessor),  # assuming you have defined preprocessor\n",
    "    ('model', logistic_regression)\n",
    "])\n",
    "\n",
    "# Grid search for hyperparameter tuning\n",
    "grid_search = GridSearchCV(estimator=pipeline,\n",
    "                           param_grid=param_grid,\n",
    "                           cv=5,  # cross-validation folds\n",
    "                           scoring='accuracy',  # evaluation metric\n",
    "                           n_jobs=-1  # use all available CPU cores\n",
    "                          )\n",
    "\n",
    "# Fit the grid search to the training data\n",
    "grid_search.fit(X_train, y_train)\n",
    "\n",
    "# Make predictions on the test data\n",
    "y_pred = grid_search.predict(X_test)\n",
    "\n",
    "# Evaluate the model\n",
    "accuracy = accuracy_score(y_test, y_pred)\n",
    "roc_auc = roc_auc_score(y_test, y_pred)\n",
    "\n",
    "print(f\"Accuracy: {accuracy}\")\n",
    "print(f\"ROC AUC: {roc_auc}\")\n",
    "\n",
    "# Best hyperparameters\n",
    "print(\"Best hyperparameters:\", grid_search.best_params_)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 48,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Confusion Matrix:\n",
      "[[3117  717]\n",
      " [ 917 2848]]\n"
     ]
    }
   ],
   "source": [
    "from sklearn.metrics import confusion_matrix\n",
    "\n",
    "# Calculate confusion matrix\n",
    "conf_matrix = confusion_matrix(y_test, y_pred)\n",
    "\n",
    "# Print confusion matrix\n",
    "print(\"Confusion Matrix:\")\n",
    "print(conf_matrix)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 49,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 576x432 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "import matplotlib.pyplot as plt\n",
    "import seaborn as sns\n",
    "\n",
    "# Plot confusion matrix\n",
    "plt.figure(figsize=(8, 6))\n",
    "sns.heatmap(conf_matrix, annot=True, cmap='Blues', fmt='d', cbar=False)\n",
    "plt.xlabel('Predicted labels')\n",
    "plt.ylabel('True labels')\n",
    "plt.title('Confusion Matrix')\n",
    "plt.show();"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 2) Support Vector Machine\n",
    "Here the Machine Learning models learn from the past input data and predict the output.support\n",
    "vector machines are basically supervised learning models used for classification and regression\n",
    "analysis.\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 50,
   "metadata": {},
   "outputs": [],
   "source": [
    "lsvc_param = {\n",
    "    'model__C': [1, 10], #the higher the C, the more penalty SVM was given when it misclassified\n",
    "    'model__max_iter': [10000], # maximum number of iterations to be run.\n",
    "    'model__dual':[False],  # dual=False when n_samples > n_features.\n",
    "    'model__penalty': ['l1','l2'],# Specifies the norm used in the penalization. The ‘l2’ penalty is the standard used in \n",
    "                                  # SVC. The ‘l1’ leads to coef_ vectors that are sparse.\n",
    "}"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 51,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fitting 10 folds for each of 4 candidates, totalling 40 fits\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "[Parallel(n_jobs=-1)]: Using backend LokyBackend with 4 concurrent workers.\n",
      "[Parallel(n_jobs=-1)]: Done  33 tasks      | elapsed:  3.7min\n",
      "[Parallel(n_jobs=-1)]: Done  40 out of  40 | elapsed:  4.2min finished\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Best parameter (CV score=0.788):\n",
      "{'model__C': 1, 'model__dual': False, 'model__max_iter': 10000, 'model__penalty': 'l1'}\n"
     ]
    }
   ],
   "source": [
    "from sklearn.svm import LinearSVC\n",
    "\n",
    "lsvc = { 'model': LinearSVC(random_state=42), 'params': lsvc_param }\n",
    "\n",
    "lsvc_test_preds, lsvc_train_preds,lsvc_pipeline = fit_predict(lsvc, X_train, X_test, y_train, y_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 52,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Model: LinearSVC\n",
      "------------------------------------------------------------\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "test accuracy: 0.7857612843795236\n",
      "train accuracy: 0.8343392182300187\n",
      "------------------------------------------------------------\n",
      "\n",
      "test report:\n",
      "              precision    recall  f1-score   support\n",
      "\n",
      "           0       0.77      0.82      0.79      3834\n",
      "           1       0.80      0.75      0.78      3765\n",
      "\n",
      "    accuracy                           0.79      7599\n",
      "   macro avg       0.79      0.79      0.79      7599\n",
      "weighted avg       0.79      0.79      0.79      7599\n",
      "\n",
      "~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~\n",
      "\n",
      "train report:\n",
      "              precision    recall  f1-score   support\n",
      "\n",
      "           0       0.81      0.86      0.84      8830\n",
      "           1       0.86      0.81      0.83      8899\n",
      "\n",
      "    accuracy                           0.83     17729\n",
      "   macro avg       0.84      0.83      0.83     17729\n",
      "weighted avg       0.84      0.83      0.83     17729\n",
      "\n",
      "------------------------------------------------------------\n"
     ]
    }
   ],
   "source": [
    "# Call the metrics function to evaluate the LinearSVC model\n",
    "metrics('LinearSVC', y_train, y_test, lsvc_train_preds, lsvc_test_preds);\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 3) XgBoost\n",
    "XgBoost is a gradient boosting algorithm for supervised learning. It’s a highly efficient and scalable\n",
    "implementation of the boosting algorithm, with performance comparable to that of other state-ofthe-art machine learning algorithms in most cases."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 53,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Importing the libraries\n",
    "import pandas as pd\n",
    "import numpy as np\n",
    "\n",
    "from scipy import stats\n",
    "import math\n",
    "\n",
    "#Visual/Graphs\n",
    "import matplotlib.pyplot as plt\n",
    "%matplotlib inline\n",
    "import seaborn as sns\n",
    "plt.style.use('seaborn')\n",
    "\n",
    "#Sklearn\n",
    "from sklearn.ensemble import RandomForestClassifier\n",
    "from sklearn.linear_model import LogisticRegression\n",
    "from sklearn.neighbors import KNeighborsClassifier as KNN\n",
    "from sklearn.preprocessing import LabelEncoder\n",
    "from sklearn.preprocessing import StandardScaler\n",
    "from sklearn.model_selection import train_test_split\n",
    "from sklearn.model_selection import cross_val_score\n",
    "from sklearn.model_selection import StratifiedShuffleSplit\n",
    "from sklearn.decomposition import PCA\n",
    "from sklearn.linear_model import LogisticRegression\n",
    "from sklearn.metrics import accuracy_score\n",
    "from sklearn import metrics\n",
    "from sklearn.metrics import roc_curve, auc\n",
    "from sklearn.model_selection import GridSearchCV\n",
    "from sklearn.model_selection import RandomizedSearchCV\n",
    "from sklearn.svm import SVC\n",
    "from sklearn.tree import DecisionTreeClassifier\n",
    "from sklearn.metrics import classification_report, accuracy_score, roc_auc_score\n",
    "\n",
    "from IPython.core.interactiveshell import InteractiveShell\n",
    "InteractiveShell.ast_node_interactivity = \"all\"\n",
    "\n",
    "#Import warnings\n",
    "import warnings\n",
    "warnings.filterwarnings(\"ignore\")\n",
    "\n",
    "import pandas.util.testing as tm"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 54,
   "metadata": {},
   "outputs": [],
   "source": [
    "# This function finds the top features of a model using eli5 library\n",
    "def top_feat(model_pipe):\n",
    "    '''This function is used to find the best features of our models\n",
    "    \n",
    "    Args:\n",
    "        model_pipe (GridSearchCV): model_pipe is a pipeline\n",
    "        \n",
    "    Returns:\n",
    "        the top features of the model\n",
    "    '''\n",
    "    \n",
    "    # Extracting the one-hot encoded column names\n",
    "    onehot_columns = list(model_pipe.best_estimator_.named_steps['preprocessor']\n",
    "                          .named_transformers_['cat']\n",
    "                          .named_steps['onehot']\n",
    "                          .get_feature_names(input_features=categorical_cols))\n",
    "    \n",
    "    # Combining numerical and one-hot encoded column names\n",
    "    numeric_features_list = list(numerical_cols)\n",
    "    numeric_features_list.extend(onehot_columns)\n",
    "    \n",
    "    # Returning the top features using eli5 library\n",
    "    return eli5.explain_weights(model_pipe.best_estimator_.named_steps['model'], top=20, feature_names=numeric_features_list)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 55,
   "metadata": {},
   "outputs": [],
   "source": [
    "xgb_param = {\n",
    "    'model__eta': [.3, .2, .1, .05, .01, .005], #Learning rate\n",
    "    'model__max_depth': [10], #The maximum depth of a tree.Used to control over-fitting\n",
    "    'model__min_child_weight': [6], # minimum sum of weights of all observations required in a child\n",
    "    'model__subsample': [0.8]      # Subsample ratio of the training instances.\n",
    "}"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 56,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fitting 10 folds for each of 6 candidates, totalling 60 fits\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "[Parallel(n_jobs=-1)]: Using backend LokyBackend with 4 concurrent workers.\n",
      "[Parallel(n_jobs=-1)]: Done  33 tasks      | elapsed:  2.3min\n",
      "[Parallel(n_jobs=-1)]: Done  60 out of  60 | elapsed:  4.1min finished\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Best parameter (CV score=0.800):\n",
      "{'model__eta': 0.2, 'model__max_depth': 10, 'model__min_child_weight': 6, 'model__subsample': 0.8}\n"
     ]
    }
   ],
   "source": [
    "# Importing XGBClassifier from xgboost library\n",
    "from xgboost import XGBClassifier\n",
    "\n",
    "# Defining the XGBClassifier model and its parameters\n",
    "xgb = { 'model': XGBClassifier(random_state=42), 'params': xgb_param }\n",
    "\n",
    "# Fitting and predicting using the XGBClassifier model\n",
    "xgb_test_preds, xgb_train_preds, xgb_pipeline = fit_predict(xgb, X_train, X_test, y_train, y_test)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 57,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "XGBoost Accuracy: 0.795499407816818\n",
      "XGBoost Precision: 0.8118476727785614\n",
      "XGBoost Recall: 0.7644090305444887\n",
      "XGBoost F1-score: 0.7874145006839945\n"
     ]
    }
   ],
   "source": [
    "from sklearn.metrics import accuracy_score, precision_score, recall_score, f1_score\n",
    "\n",
    "# Calculate and print various metrics using the same data provided to the `metrics` function\n",
    "XGBoost_accuracy = accuracy_score(y_test, xgb_test_preds)\n",
    "XGBoost_precision = precision_score(y_test, xgb_test_preds)\n",
    "XGBoost_recall = recall_score(y_test, xgb_test_preds)\n",
    "XGBoost_f1 = f1_score(y_test, xgb_test_preds)\n",
    "\n",
    "# Printing the calculated metrics for XGBoost\n",
    "print(f\"XGBoost Accuracy: {XGBoost_accuracy}\")\n",
    "print(f\"XGBoost Precision: {XGBoost_precision}\")\n",
    "print(f\"XGBoost Recall: {XGBoost_recall}\")\n",
    "print(f\"XGBoost F1-score: {XGBoost_f1}\")\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 68,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 576x396 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Plot ROC curve\n",
    "roc(X_test, y_test, xgb_test_preds, xgb_pipeline)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 60,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 576x396 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plot_confusion_matrix(y_test, xgb_test_preds);\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "import eli5\n",
    "\n",
    "# Define a function to extract and return the top features of an XGBoost model\n",
    "def top_feat(model_pipe):\n",
    "    '''This function extracts the top features of a model trained using XGBoost.\n",
    "    \n",
    "    Args:\n",
    "        model_pipe (GridSearchCV): The pipeline resulting from the GridSearchCV process.\n",
    "    \n",
    "    Returns:\n",
    "        The top features of the XGBoost model.\n",
    "    '''\n",
    "    \n",
    "    # Extract the feature names from the one-hot encoded categorical columns\n",
    "    onehot_columns = list(model_pipe.best_estimator_.named_steps['preprocessor'].named_transformers_['cat'].named_steps['onehot'].get_feature_names(input_features=categorical_cols))\n",
    "    \n",
    "    # Combine the feature names from categorical and numerical columns\n",
    "    numeric_features_list = list(numerical_cols)\n",
    "    numeric_features_list.extend(onehot_columns)\n",
    "    \n",
    "    # Return the weights of the features using eli5.explain_weights()\n",
    "    return eli5.explain_weights(model_pipe.best_estimator_.named_steps['model'], top=20, feature_names=numeric_features_list)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "top_feat(xgb_pipeline)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 4) RandomForest Classifier\n",
    "A Random forest is a machine learning technique that is used to solve regression and classification\n",
    "problems.it utilizes ensemble learning,which is a technique that combines many classifiers to provide\n",
    "solution to complex problems.easy to use machine learning algorithm that produce,even without\n",
    "hyper-parameter tuning,a great result most of the time."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 65,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fitting 10 folds for each of 4 candidates, totalling 40 fits\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "[Parallel(n_jobs=-1)]: Using backend LokyBackend with 4 concurrent workers.\n",
      "[Parallel(n_jobs=-1)]: Done  33 tasks      | elapsed:  3.3min\n",
      "[Parallel(n_jobs=-1)]: Done  40 out of  40 | elapsed:  4.0min finished\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Best parameter (CV score=0.751):\n",
      "{'model__criterion': 'gini', 'model__max_depth': 10, 'model__min_samples_leaf': 5, 'model__min_samples_split': 10, 'model__n_estimators': 1000}\n"
     ]
    }
   ],
   "source": [
    "from sklearn.tree import DecisionTreeClassifier\n",
    "from sklearn.ensemble import BaggingClassifier, RandomForestClassifier\n",
    "\n",
    "forest_params = {'model__n_estimators' : [500,1000],  # number of decision trees to be created\n",
    "                 'model__criterion' : ['gini','entropy'], # function that is to be used for splitting\n",
    "                 'model__max_depth': [10],          # maximum depth of the trees\n",
    "                 'model__min_samples_split' : [10], # minimum number of samples required in a leaf node before a split is attempted\n",
    "                 'model__min_samples_leaf' : [5]} # minimum number of samples required to be at a leaf node. \n",
    "                                            # Smaller leaf size makes the model more prone to capturing noise in train data.\n",
    "rfc = { 'model': RandomForestClassifier(random_state=42), 'params': forest_params }\n",
    "\n",
    "rfc_test_preds, rfc_train_preds,rfc_pipeline = fit_predict(rfc, X_train, X_test, y_train, y_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 66,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fitting 10 folds for each of 4 candidates, totalling 40 fits\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "[Parallel(n_jobs=-1)]: Using backend LokyBackend with 4 concurrent workers.\n",
      "[Parallel(n_jobs=-1)]: Done  33 tasks      | elapsed:  3.6min\n",
      "[Parallel(n_jobs=-1)]: Done  40 out of  40 | elapsed:  4.4min finished\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Best parameter (CV score=0.751):\n",
      "{'model__criterion': 'gini', 'model__max_depth': 10, 'model__min_samples_leaf': 5, 'model__min_samples_split': 10, 'model__n_estimators': 1000}\n",
      "Random Forest Accuracy: 0.73917620739571\n",
      "Random Forest Precision: 0.7513391598533973\n"
     ]
    }
   ],
   "source": [
    "from sklearn.metrics import accuracy_score, precision_score, recall_score, f1_score\n",
    "\n",
    "rfc_test_preds, rfc_train_preds, rfc_pipeline = fit_predict(rfc, X_train, X_test, y_train, y_test)\n",
    "\n",
    "random_forest_accuracy = accuracy_score(y_test, rfc_test_preds)\n",
    "random_forest_precision = precision_score(y_test, rfc_test_preds)\n",
    "# ... calculate other metrics\n",
    "\n",
    "print(f\"Random Forest Accuracy: {random_forest_accuracy}\")\n",
    "print(f\"Random Forest Precision: {random_forest_precision}\")\n",
    "# ... print other metric values\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 67,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Accuracy: 0.73917620739571\n",
      "Classification Report:\n",
      "              precision    recall  f1-score   support\n",
      "\n",
      "           0       0.73      0.77      0.75      3834\n",
      "           1       0.75      0.71      0.73      3765\n",
      "\n",
      "    accuracy                           0.74      7599\n",
      "   macro avg       0.74      0.74      0.74      7599\n",
      "weighted avg       0.74      0.74      0.74      7599\n",
      "\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 576x432 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "\n",
    "# Calculate accuracy\n",
    "accuracy = accuracy_score(y_test, rfc_test_preds)\n",
    "print(f'Accuracy: {accuracy}')\n",
    "\n",
    "# Generate and print classification report\n",
    "print('Classification Report:')\n",
    "print(classification_report(y_test, rfc_test_preds))\n",
    "\n",
    "# Generate confusion matrix\n",
    "cm = confusion_matrix(y_test, rfc_test_preds)\n",
    "\n",
    "# Plot confusion matrix\n",
    "plt.figure(figsize=(8, 6))\n",
    "sns.heatmap(cm, annot=True, fmt='d', cmap='Blues', cbar=False)\n",
    "plt.xlabel('Predicted')\n",
    "plt.ylabel('Actual')\n",
    "plt.title('Confusion Matrix')\n",
    "plt.show();\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 69,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAe8AAAGACAYAAACEK+pCAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjMuMSwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy/d3fzzAAAACXBIWXMAAAsTAAALEwEAmpwYAACDg0lEQVR4nO3dd3Rc1bX48e9U9Wp12bK65C7LDVdkCdOMaQYM/EJJIIQXAoQQ2gMcqiEEkgB5EMhLCJCXQEIxxTR3G+NuybKNm3q3eh1Nvff3h6SxhC1Ltsqo7M9arCVpZu7dc5C155y77z4aVVVVhBBCCDFsaF0dgBBCCCHOjiRvIYQQYpiR5C2EEEIMM5K8hRBCiGFGkrcQQggxzEjyFkIIIYYZvasDEGIoSkpKIjExEa1Wi0ajobW1FW9vb5544gmmTJkCgMlk4tVXX2XDhg0YjUYA0tPT+a//+i/c3d2dx/r444957733MJvN2Gw2ZsyYwQMPPICvr+9pz322z3eFEydO8Ic//IFDhw6h0Whwc3PjZz/7GRdccIGrQ3M6ceIE9957L++9956rQxGi/6lCiFMkJiaqNTU1XX72v//7v+p1112nqqqq2mw29brrrlNXrVqlmkwmVVVV1WQyqU8//bR64403qjabTVVVVX399dfVG264Qa2qqlJVVVWtVqv6xBNPqDfccMNpz3u2z3eFmpoaNS0tTf34449VRVFUVVXVw4cPq+edd5767bffujg6IUYHmXkL0Qt2u53y8nL8/PwA+Oqrr1AUhUceecT5HA8PDx599FGuvPJK1q5dy/nnn88bb7zBxx9/TFBQEAAGg4EHH3yQtWvXYrVanTN2aJvJ9/T8N954g7q6OlauXAnAq6++6vz+pptuws/Pj7y8PFasWMFrr73G1q1bMRqNOBwO0tLS+Pvf/05ISAjPPvssx44dw2azMXfuXB588EH0ej2vvPIKa9euxWAwEBAQwHPPPUdISEiXsfjnP/9JamoqV155pfNnycnJvPLKK87VgT179vDCCy/Q2tqKwWDgl7/8JYsWLeKjjz7im2++QVEUysrKCA0N5brrruMf//gHBQUF/PjHP+YnP/kJH330kXOMO573/PPPExoaSlZWFr/73e+wWq1UVVUxb948Vq1aRUlJCf/v//0/4uLiKC0t5fnnn+cnP/kJmZmZ5Obm8uijj2K1WlFVlWuuuYb/9//+Hzabjeeff57t27ej0+mYOnUqjzzyCN7e3qSnp3PVVVexfft2ysvLueKKK/jlL3/Z779bQpwLueYtRDduueUWli1bxoIFC7jooosAeO655wDIzMxk5syZp7xGo9Ewd+5c9u7dS15eHu7u7kRHR3d5joeHB5dffnmXxA2c9fNPx9fXly+++IJbbrmFhIQENmzYAMC3337L2LFjiYuLY9WqVUyaNImPPvqI1atXU1dXx1tvvUV5eTlvv/02H374IR999BHz588nOzv7lHMcPHiQ1NTUU34+a9YskpKSqKur45577uHRRx/ls88+47e//S0PPPAAxcXFQFtif/LJJ/n000+pqKhgzZo1vP322/zlL3/hj3/8I4qiALB7924effRRvvjiCyZNmsSzzz4LwDvvvMM999zDf/7zH9asWcOGDRs4ePAgABUVFfz85z/n66+/Jjg42BnbX//6V9LT0/noo49488032bNnD4qi8Prrr1NZWcknn3zCJ598gqIovPDCC87XmUwm/vnPf/Lee+/xt7/9zfkehHA1mXkL0Y23336bwMBADh06xB133MGcOXMYM2aM83G73X7a11mtVnQ6HVqt1pmIeuNsn386nT9QXHPNNXz88cdcfPHFfPTRR1x33XUAbNq0iQMHDvDBBx8AYDabAQgNDSU5OZmrrrqKRYsWsWjRIubOnXvKOTQaDeoZuipnZ2cTFRXFtGnTAEhISCA1NZVdu3ah0WiYMmUK4eHhAIwdO5YFCxag1WoZN24cFouF1tZWAObPn09MTAwA1113HVdccQUAzz//PFu2bOHPf/4zeXl5WCwWTCYT/v7+6PV6UlJSTolpyZIlPPTQQ2RnZzN37lwee+wxtFotW7Zs4b777sNgMABw0003cddddzlfl5GR4RybMWPG0NDQwLhx43r63yDEgJOZtxA9mDRpEo888ggPP/wwJSUlAKSmpjpnb50pisLu3buZPn068fHx2O12CgoKujzHYrHw05/+lBMnTnT5eW+e/8PEabPZujzX09PT+fUll1zC/v37yc3NZffu3Vx88cXOGF9++WXnbPM///kPK1euRKvV8o9//IPnnnsOf39/Vq1a1WUW2iElJYWsrKxTfv7ee+/x1ltv4XA40Gg0XR5TVdX5YeeHKwh6/ennEDqdzvm1oijO73/0ox+xefNmYmNjueuuuwgJCXGOidFoPO3xFi9ezNdff80ll1zC4cOHWbZsGRUVFSiK0iVWRVG6jKmbm5vz654+tAgxmCR5C9ELl112GVOnTnUum1900UV4eHiwatUq58zVbDbz9NNP4+XlxZIlSzAajfz0pz/l0Ucfpbq6Gmibla9atYrW1lZCQ0O7nKM3zw8ICODQoUOoqkpzczMbN27sNmY3NzeWLl3Kww8/zIUXXoiHhwcACxYs4O9//zuqqmK1Wvmv//ov/vGPf3DkyBEuu+wy4uLi+NnPfsatt97KgQMHTjnuihUr2LVrF59++qkzmR08eJBXXnmFxMREUlJSyMvLcy65Hz9+nN27dzN79uyzGvMdO3Y4P+C89957LF68mMbGRg4cOMCvf/1rLrzwQioqKigqKupxxeL+++/niy++YOnSpfzmN7/B29uboqIiFi5cyL/+9S9sNhuKovB///d/zJ8//6ziFMIVZNlciF56/PHHufzyy9m6dSsLFy7kb3/7G6+99hpXX301Wq0Wh8NBeno6f/vb35zLsHfeeSceHh7cdtttQNssevbs2bz22munPUdPz+84/4UXXkhoaCizZ88+42zw2muv5R//+AdPPPGE82ePPvoozz77LMuWLcNmszFv3jxuv/12DAYDl1xyCcuXL8fT0xN3d3cee+yxU47p7+/Pu+++y+9+9zveeOMNtFotHh4ePPvss87E9/LLL/P0009jNpvRaDQ899xzxMTEkJmZ2evxDg0N5YEHHqCqqor4+HieeuopfH19ueOOO7jqqqvw9PQkNDSU1NRUCgsLz7ic/fOf/5xHH32U999/H51OxwUXXMCsWbOYOnUqv/3tb7nyyiux2+1MnTqVxx9/vNcxCuEqGlXWgYQQQ8xHH33E119/zRtvvOHqUIQYkmTZXAghhBhmZOYthBBCDDMy8xZCCCGGGUneQgghxDAjyVsIIYQYZobNrWLl5eX9erygoCDnvbTi3Mk49p2MYd/JGPadjGHfDcQYdnQj/KFRO/PuuA9X9I2MY9/JGPadjGHfyRj23WCO4ahN3kIIIcRwJclbCCGEGGYkeQshhBDDjCRvIYQQYpiR5C2EEEIMM5K8hRBCiGFGkrcQQggxzAybJi1D0Z49e3j99dd57bXXcHNzo7q6mgceeIAXXniB7OxsVq9eDYBWqyU+Pp4777wTg8HAihUrCA0NRaPR0NrayiWXXMJVV10FQH5+Pm+88QZms5nW1lbOO+88br31VrKysvj000/5zW9+06eYa2pqeOedd7jvvvvYunUrf/3rX1m2bBlZWVk8/fTTfR0SIYQQg0CSdx/MnDmTWbNm8dprr3H33Xfz5JNPctddd5Gbm8uaNWtYtWoVPj4+qKrK//zP//D1119z2WWXAfC73/0ONzc3bDYbt9xyC2lpaej1ep566imefvppxo4di8Ph4IknnuDTTz8lKiqqX2IeM2YM9913HwDbt2/njjvuYN68eSxfvrxfji+EEGLgjZjk/cGmIvYeq+3183W6AzgcjjM+Z0ZiINeknTlp3n777dx99908+uijzJgxg5kzZ/Lggw9y55134uPjA4BGo+Guu+5Co9Gc8nqz2YzRaMTb25v169eTmprK2LFj22PU8cgjj2AwGDh48KDzNR999BFbt27Fbrfj5eXF008/TUVFBc8//zx6vb7L65588kkURcHhcPCrX/0KDw8PnnrqKX70ox+xfft2jhw5gp+fH4899hgff/wxeXl5vPLKK6iqip+fHw8++CDHjx/nzTffRK/Xs2zZMi688MJej7MQQoj+N6DJe//+/bz44ou8++67XX6+YcMG/ud//ge9Xs/y5cu57rrrBjKMAaXX61m6dCl/+MMf+NWvfgW09WGPjIwE4NChQ7z55ps4HA6Cg4Ody94PPPAAGo2GoqIiFixYgF6vp6am5pQ+tp6enl2+VxSFxsZGXnrpJbRaLQ888ABHjhwhJyeHxMRE7rrrLrKzs2lqauLEiRN4eXnx+OOPU1BQQEtLCx4eHgDMnz+fLVu2kJ6ezqRJk5zH/93vfsdDDz1EdHQ0a9as4b333mPGjBlYrVZef/31ARtHIYQQvTdgyfsvf/kLn376qTNZdLDZbDz33HN88MEHeHh4cMMNN7B48WKCg4P7dL5r0qJ6nCV3Fh4e3i+bnVRUVPD+++9z55138uyzz/KHP/yBkJAQysvLiY+PZ9KkSbz88ssUFhby+9//3vm6zsvmDz30EGvXriU0NJRjx451OX55eTmVlZXO77VaLQaDgaeffhoPDw+qqqqw2+1ceuml/Otf/+LBBx/Ey8uL22+/nTlz5lBSUsKjjz6KXq/npptu6vH9FBUV8Yc//AEAh8PhXAUYN25cn8dKCCFGosaGJrJ3fIdRr2V2xpJBOeeAJe+oqCheffVVHnzwwS4/z83NJSoqCj8/PwBmzJjBnj17uOSSS854vKCgoH5v+t7dbi29ZbVaueeee3j88cc5//zzKSoq4sMPP+S2227jrbfe4uWXX3YunW/atAk3NzfCw8PR6/WEh4fj5uYGQEREBF5eXlx88cVcf/312Gw2oqKisNlsPPPMM8ybN4/4+Hg8PDxoaGhg586d/Oc//6G1tZWrr76awMBAvv/+e9LS0njkkUf4/PPP+eSTT7j88suJj4/nnnvuITMzk9///vc899xzGI1GwsPD8fT0JDAwkPDwcHQ6HeHh4cTGxvLHP/6RiIgI9u7dS1VVFQEBAXh6enY7Xn0dRyFj2B9kDPtOxrBnZqud3Kxj5O7ZTVPuAbwrK0DVcjxmPMlFpZjmziMuLn7A4xiw5H3RRRdRUlJyys+bm5udCQ3Ay8uL5ubmHo83ENus9XXm/fLLL5OcnExiYiLl5eXceeed3HHHHSQkJHDhhRdy++23A2AymYiPj+f++++nvLwcu93Oj370I7RaLYqiEBwczKxZs2hqauLXv/41Dz74IIqi0Nrayty5c1m8eDFZWVm0trbi5uaGTqdj2bJlGI1G/Pz8OHbsGBMnTuTZZ59Fp9Oh1Wq56667CAgI4JVXXuEvf/kLWq2WW265hcrKSqxWK+Xl5ZhMJmpraykvL8fhcFBeXs4vfvELfvnLX6IoCgAPPvgg1dXVtLa2nna8+msFYzSTMew7GcO+kzE8lUNRKSurp/zAcZqOH0d/Io+gxho8bHbGAGMAuw6Kg71x6LQ0LpiDp6dXv45jdx+oNKqqqv12lh8oKSnhV7/6Ff/+97+dPzty5AgvvfQSf/nLXwBYtWoVqampXHzxxWc8Vn//Uskvav+Qcew7GcO+kzHsu9E+hqqqUllvpvhwEQ1HclCKC/GtK2WMuaFLQ5RGLy0VY9xoDA3BKyqG1NkLCPMNxW63ExUVNSC56nQGvdo8Li6OwsJC6uvr8fT0ZM+ePdx2222DHYYQQohRrL7ZSkFRLdWHcrDmF+BeWUKoqZJAxUpg+3PsOigPNlAeZMAUPoYxSRNJGDeRqEoLO77bgVrrIMQ7GI1GM+j7oQ9a8v7ss88wmUysWLGChx9+mNtuuw1VVVm+fDmhoaGDFYYQQohRxmSxU1jeTOmxYlqO56ErKyao6QRB1np8Obn43OClpSjIjfIgA/WhPgTFJpEclMBFAbH4uflgtVrZunUrubm5GI1GZs2ahU6nc8l7GtDkPXbsWOeS+bJly5w/T09PJz09fSBPLYQQYhSy2RWKK00UFNVQfzQPpagQ37pyIs1VxCpW5/PsWg0VwQbKgvRUBBk4EWwkNGQ8yYFxXBAQx1jvcLSdenNUVlayfv16mpqaCAkJISMjo0v91mAbMU1ahBBCjC4ORaW8ppWC8mYqcoqx5OXjUVVGRGsl4631RHd6brOXO7nBPhSPUSkPMlAdoCfAM5DkgDgWB8SR4B+Nu97ttOdRFIWNGzfS1NRESkoKM2fORKt17dYgkryFEEIMeaqqUtVgoaCihaKiGppz8tGVtV2njjRXEdFpVu3Q6WgIG0NliIHj/hbKxmgxeegwag0k+sewKDCO5IA4gj0Cz3DGtqSt1WrRarWkpaVhs9mcvS9cTZK3EEKIIaehxUpBRQsFZc1U55XgKCpkTNMJIs1VTLfW07nZtM3Hj8bIsVSF6Tnk3UKhtxVFqwEUxnqPY15AW7KO8R2HXtu7a9QlJSVs27aNSy+9FB8fnyFXmyXJuw8yMzN58sknGT9+PBqNBpPJRHh4OI899lifKg+ffPJJLr/8cqZPn97nGL/88kveeuutLrcbXHfddcyfP7/Px+5s//79eHt7ExcX16/HFUKMfCaLncKKFgoqWigursWcf/I6dby5iimdZtWKTo8aFY1jfDDlQTqyvRrJVapRaesX4m3wIjUgmeSAOJICYvE1ep9VLIqisHv3bvbv349Wq6WystKl17a7I8m7j6ZPn95lm86nn36abdu2kZaW5rqgfiAjI4Of/exnA3qOL774gvT0dEneQogzchaUVTSTX95MXUEZbidK2pa+zVUk/mBWrfgH4BY3Bc34MIoDNRw01nG0qRCLoxAAnaolzm88yQFxTAiMI8IrrEuh2dlobGxk/fr1VFVV4evrS0ZGRp9bdw+UEZO8P8lbS1bV971+vk6n63FXsZTgiVwR2/s+tTabjZqaGnx8fHA4HLz00ktUVVXR0NDAnDlzuO2225ztSSsqKqipqeHhhx8mMTGRjz/+mDVr1jBmzBjq6uoAsNvt/Pa3v6WsrAxFUbj22mtJT0/n3nvvJT4+nvz8fDw8PJgyZQq7d++mubmZF198sVefEpuamnj22WcxmUw4HA5uu+02UlNTufXWWxk3bhwGg4Ff/epXvPDCCzQ2NgJwzz33EBsby3PPPUdZWRlWq5U777wTT09Pdu3axfHjx4mOjh5yy0tCCNdQFJWymta25e+KZkpK6lBLiolorSTCXMV55mo8Os2qVb0BQ2wsXvGxaKPHUTpGyxFbBYdrc6k27wYTYIIg90Bmh8aRHBBLvF/3hWZno6SkhLVr12Kz2YiPj2fBggUYjcY+H3egjJjk7SqZmZnce++91NfXo9FoWLZsGTNmzKC8vJyJEydy2WWXYbFYuPbaa53NaEJDQ7n//vv5/PPP+eyzz/jZz37Ghx9+yN/+9je0Wi133HEHAJ9++il+fn48+uijmEwmfvrTn5KamgpAcnIyd999Nw888ADu7u689NJLPPfcc2RlZbFw4cIuMa5fv57vv2/7YOPv78+TTz7Ju+++y8yZM7nmmmuoqqri7rvv5p///Cetra3cfPPNJCQk8MYbbzBjxgyuuOIKSkpKeP7553nhhRfIysrijTfeACAnJ4ekpCRmz55Nenq6JG4hRilVValusJBf0UJheVuybioqI6TlBBHmauLNVcz9waxaExiIZ3ws7rExGGNiqPLTcrSxgCN1ueQ3HsTR3Nam2U1nZMqYJJLbr10HeQT0e/z+/v64ubkxf/58EhMT+/34/W3EJO8rYpec1Sy5v1oBdiybNzQ08Otf/9p5bdnX15ejR4+SlZWFp6cnNpvN+ZqEhAQAgoODOXDgAEVFRURHRzs/5SUnJwNtO3zNmDEDaNsaNDo6mrKyMgDnL5e3tzfjx493fm21nvwU2+F0y+ZFRUUsWbLEGYeXlxf19fXAyR3E8vLy2LdvHxs2bADa+tJ7enpy77338uKLL2Iymbjmmmv6MnxCiGHKWVBW3natuqysDr/6Cufy91JzNe6dZtUYDLglxOMRG4NbbAxuMdGY3LUcrcvjcF0uR8v20FzY4nz6OO9wZ7KO8R2LrpeFZmejpqYGu91OaGgo3t7erFixwmVNV87WiEnertYxQ77vvvv43//9XzZt2oS3tzf3338/JSUlfP7553TXRj4iIoLCwkIsFgt6vZ6cnByWLFlCVFQU2dnZLFy4EJPJRF5eXr/t+tNx7ISEBKqqqmhqasLX1xcATfv1oqioKJYsWcIFF1xAXV0da9asoaamhqNHj/LMM89gsVi4/vrr+c9//oNWq+32/QkhhjeTxU7RCRP55c3tCbsZbW2VM1GfZ64m6Aezal1QEB5xMbjFxLTNrCMjcGggv7GIw3W5HDm+g9KWCufzfY3ezAqZSnJgHMn+sXgbvQbs/aiqyvfff8+OHTtwd3dnxYoV6PX6YZO4QZJ3v4qOjubqq6/mlVde4dZbb+Wpp54iOzsbd3d3IiMju90Zzd/fnx//+Mfcdddd+Pv74+7uDrR1pXvxxRf5xS9+gdVq5ZZbbiEgoH+Wi370ox/x29/+ls2bN2OxWLj//vvR6/WnPOeFF15wtra99dZbCQwMpLa2lttvvx0PDw9+8pOfoNfrmTBhAm+++Sbh4eHOlQAhxPDTuaCsoKKF/IoW6qsaCbdUE2muIsFcxfmWatwdnWbVRiPuCfG4t8+q3WNi0Pn6tN2bba5lf20uRw5/y/H6AqxK2yqkTqMjwT+aCe2z6wivUOfEYSCZzWY2b95MYWEhbm5uLFiw4JS/fcPBgO4q1p9kV7GhScax72QM+07G8Nx0LiirbFT5Pq+SkkoTAZZ656x6rKWaMZaus2p9cFBbou40q9a0z1rNdgvH6vM5UpfLkbpcasz1zteFeIxpXwqPJd4/Gjfd4BaElZeXs2HDBlpaWggPDyc9PR0vr/6b4Q/E7+GQ2VVMCCHE4OsoKHNepz7RQtGJFjCbiTBXEWGpZq65ikhrNUb7yVm1xmjELTGh06w6Gl2nO1oUVaW4ubwtWdfmkt9UgqK2FZq569yYGpTMhIA4kgLiGOPuP9hv+2ScisK2bdswmUzMnDmTlJQUl7c47QtJ3kIIMQL9sKCs4EQLLSYbY2wNzuYnGbYa/FrrfjCrDsY9NqZ9Zh3dZVbtPLal6WShWX0eLTYTABogyifCWWg23ncsOo1rE6TD4UCn06HVaklPT8dqtRIWFubSmPqDJG8hhBjmOgrKCsqbyW/vVFbXZMXNYSXC3FZYNt1RQ6ipCkPnWbWbEbekRNxjogmfkYrJ36/LrLqDXbGT21DUvhSeR1nLCedjfkYf5oSmtHc0i8HL4Dko77k3CgoKnC1OAwICCAw8cy/z4USStxBCDCM2u0JxVVuiLmhP1BW1ZlBVgqwNRJirWGSvYay1Gp+W2i6vNYQEO69Tu8XGYIwId86qA8PDsbRfr1VVlcrWGud165z6QmehmV6jI8k/tq0qPCCOcM/gQSk0Oxt2u52dO3dy6NAhdDoddXV1/VbsO1RI8hZCiCGqc0FZYUUL+RXNlFa14lBU56w62VbDJUoNQU2V6GwW52s7z6rd2pfBdd7d9/k2WVvZX32YI7VtCbvW0uB8LNQzqK39aEAccX7jMerOfe+GgVZfX8/69eupqakhICCAjIyMETXj7iDJWwghhoAuBWXt/xWdaMFiU5yz6nHWKs6jjnBTFR5NNV1ebwgJdiZpt9gYjOHhp1yr7kxRFYqb2grNDtflUthU6iw089C7kxI0oX0pPI5Ad78Bfe/9paSkhG+++Qa73c6ECROYO3fusLwNrDdG5rsaJJmZmXz66addNiaBtl3B/vu//7tPO4v1pL6+nt///ve0traiqiqhoaHcc889/N///R8AP/nJT5zP3bp1K5s3b+axxx4jOzubt99+G7vdjtls5pJLLuHKK68csDiFEKfX2GJr35yjU0FZqx0AN4eFSEs16Zo6oqw1+DVUoLV2nlW7tc2q24vKeppVd6i3NDqXwo/W5WOyt7YdDw3xY6KJ8xpHcmAcUT6RLi80OxdjxozBx8eHGTNmEBsb6+pwBpQk7wHww2Q+EN577z1n33GAV199lU8//ZSlS5dy33338eMf/9h5HeqLL77g+uuvp6ysjFdeeYUXXniBwMBALBYLv/zlLwkPD2fOnDkDHrMQo1WrxUHhiRbnder89oIyoH1WXU+Kpp44pYagphMYG34wqw4Nab9WHd1+rToCTS9uc7IpdnLrC9s6mtXlUmGqcj7mb/Rlaljb1pmJ/jHER8UOy3vlKysrsVqtjB07Fg8PD5YvXz6sbwHrrRGTvGs+/JiWvZm9fn6pXofDfuZdxbxmTGfM8qvOOpYVK1bwzjvv8Pvf//60O4ht2rSJf//732i1WqZMmcLPfvYzKisr+cMf/oDVaqWxsZGbb76ZhQsXdtnla+XKlc5zhIaGsnnzZiIjI5kyZQr/9V//hUajQafTMXbsWLKzs5k2bRo1NTVUVFQwbdo0/v73v3PRRRc5r/+4ubnxu9/9Dg8Pj7N+j0KI0ztZUNbi7FJ2otZMRzcsN4eFBLWOdG094a2VeNWUo+lmVt0xs+7NrBralt5PmKqdyTq3oRCb0jabN2j1zlu4JgTEEeoZNOQKzc6Gqqrs37+f3bt34+7uzvXXX4/BYBgViRtGUPIeqn64g9hPf/pT3nrrLd544w3c3d159tln2bNnDwDXXXcd06dP5+DBg7z11lssXLiwyy5fnV1xxRW4ubnx/vvv88QTTzBlyhTuu+8+QkJCuOyyy/jmm2+YNm0a33zzDZdeeinQ1oQ/Pj6+y3G8e/lHQQhxKkVRKXduedm+7WV7QRmARlWIUJu40NBIlK0a//pydLVd2ySfnFV3qgA/iwRksrVytFNHs3pLo/OxMM9gZ7KO9Ysa0oVmZ8NkMrFp0yZKSkrw9PRk8eLFA3qZcigaMcl7zPKrzmqWPFjtFH+4g1hpaSn19fU89NBDALS2tlJWVsaUKVN49913+eKLL9BoNNjtducxOnb56iwzM5MLL7yQSy+9FKvVynvvvcef/vQnnnrqKebNm8df/vIXLBYL69ev56WXXgLaPkhUVlZ2OU5OTg6qqp7y4UAI0ZWqqtQ0WMmvaD61oKydN1ZmuzUSp9QS3FyJW2UZWMzOxzVubrgnJzn7f7vFjO/1rLqDQ1Uoaip1VoUXNpWhts/rPfXuTA+e6Jxh+7v59s+bH0JKSkrYuHEjra2tjBs3jrS0tFG5ejhikvdwER4eTkhICC+99BJ6vZ4vv/yS+Ph4/va3v3HZZZcxZ84cvvzyS7766ivna063tPXhhx9SXl7OZZddhtFoJDo6msLCQgD0ej0LFizg3XffJTo6Gj+/tkrRjIwMHnvsMdLT0/H398dkMvHSSy9xyy23SPIW4gc6CsoKKlraiso6FZQBaFFI9jST7FZPZGsVXtVlUFPV5RiG0FDcYqc5+4Cf7ay6Q5254WShWX0+rXZzewwaon0jnck6yicC7TAsNOstVVXZvXs3FouF8847jylTpgzrpf++kOTdR3v27OGOO+5wfv/444+f8fn+/v5ce+213HvvvSiKQlhYGIsXLyYtLY1XXnmFwMBAQkJCaGhoOONxfvWrX/GHP/yB1atX4+bmhr+/P/fdd5/z8aVLl3Lrrbfy4osvOn8WHh7OnXfeyeOPP45Wq6W1tZWlS5dy3nnnneO7F2Jk+GFBWUFFC7VN1i7PifRSWRDQyHhbNf4NJ9CVFaNaOl2rdnfv86y6g9VhI7fhZKHZCdPJpfZANz+mB00kKSCOxIAYPPXu5/amhxGbzYbBYECj0ZCRkYHFYiE4ONjVYbmU7Com+kTGse9kDPvubMawc0FZYUVbO9HOBWUAfh46JntbiFNrCWk+gVtlKY4fXHJqm1VHOwvLDOHnNquGthlluanKublHbkMhdrWtoNaoNRDvP945uw7xGDMgs82h+nuYk5PDtm3buOSSSwgJCXF1OGcku4oJIUQ/6KmgDMDdqGVymIFkXVtrUe+aMpTjxajmk9eqFXd3PDpm1R0V4H3cSrLFZuJo3clCswZrk/OxCK9QkgNiSQ6II84vCr129P2pttlsfPfddxw9ehS9Xk9LS4urQxpSRt9vhBBiROooKCuoOLk5xw8LyvQ6DeND3En2MBNtryGgvhxNaTH270/Oqh2AISwUt9gU3Nvvre7LrNp5XFWhsLHEmayLmsqcs30vgyepwZPbt86Mxc/t1M1BRpPq6mrWr19PQ0MDY8aMISMjA39/f1eHNaRI8hZCDEudC8rK6wo4WlhLc6eCMo0Gwsd4EB+oJUGtI6S5EuOJEqy7C53Xqh20Xav2mJDs7FTWH7PqDrXmemf70eN1+bQ62s6r1WiJ9Y1ybu4x1jsc7SgtvPqhkpISvvrqKxRFYfLkycyZMwfdGdq8jlaSvIUQQ56zoKziZFHZDwvKgvzcSB7nTaLRRKSlCu+acuz5Bdh2tM2qVcACGMLC2q5Vx3Rcqw7r86y6g8VhJaeh0HkbV2XryU5pY9z9SQ2Z7Oxo5q5365dzjjQhISEEBQWRmppKVFSUq8MZsiR5CyGGFJtdoaTKRH57QVnHlpedC8p8PPVMjfUnzl/DeEct4+xNNB35HsvBk7PqVjrNqjuuVUdHo/Pqv/2mVVWlrKXSuRSe21CEo1Oh2aTAhLZCs8A4gt0DR+1tTT0pLy/HbDYTExOD0WjkiiuukLHqgSRvIYTL9LagLHGcD9EhnsQZWwhpPoG+rBhLVj62yrb7qjvurjaEhTk7lbnHxmAIC+23WXWHZmsLR+rzOFKby9H6PBqtzc7HIr3CSA6MZUJAHDG+40ZlodnZUBSFffv2kZmZidFoJDIyEqPRKIm7F+Q3qw8yMzN58sknGT9+PBqNBpPJRHh4OI899lifWvU9+eSTXH755UyfPr1f4tyzZw///Oc/sVqt6HQ6wsLCuPvuu6U1qhhUnQvKOjbnOF1B2bgQT2LCvYjx0zDWUo1HVSmW/AIs3xWgWqx01IBr3N3xmJiMW0wMETNn0OLr26+z6g4OxUF+p0KzkuZy5yqAt8GTGSFTnIVmvkb5N9Vbzc3NbNiwgYqKCry9vUlPT8doNLo6rGFDkncfTZ8+vcsuYk8//TTbtm0jLS3NdUF1kpOTwxtvvMGqVaucTQ3+85//8N5773H77be7ODoxkrUVlJ3cnKOgoqVrQRkQHuRBdJgX0SEeRGua8asvw1Z4GMvmtlm1Fei4sm0ID2trfnKaWXVAeDjmfry/trq1zpmsj9XnY3G0RaHVaInzO3nPdaR3mBSanYOCggI2b96MxWIhJiaGRYsW4eYmNQBnY0Ql73/+85+n/fm0adOYNGkSgPOTnk6nw+E4uatYSEgIF1xwAQCHDx8mMzOTG2+88azOb7PZqKmpwcfHB4fDwUsvvURVVRUNDQ3MmTOH2267jeeee+60O419/PHHrFmzhjFjxlBXVweA3W7nt7/9LWVlZSiKwrXXXkt6ejr33nsv8fHx5Ofn4+HhwZQpU9i9ezfNzc28+OKL+PicvM3k008/5aabburSjejaa691ft2xA5qbmxtvvPEGUVFRXHLJJbz55ptkZ2ejKArXXXcdaWlprF69mq+//hqNRuPcyeybb77htddec87o//u//3vU7OojTupSUFbRTEF5NwVlUb5tydoXQloqUYryMR8uwPJlAYrFSl37c7UeHs5ZtbMC3LP/Z9UdLA4rx+vzOVKXx+HaXKrNtSfjdg9gVshUkgPjSPCLlkKzPlJVlYMHD2K321mwYAETJkyQZfJzMKKStytkZmZy7733Ul9fj0ajYdmyZcyYMYPy8nImTpzIZZddhsVi4dprr+W2224DTt1p7Gc/+xkffvghf/vb39Bqtc52q59++il+fn48+uijmEwmfvrTn5KamgpAcnIyd999Nw888ADu7u689NJLPPfcc2RlZbFw4UJnfOXl5URGRjq/fv755wFwOBz86U9/Ou172rlzJ+Xl5fzpT3/CYrHw85//nBkzZvDll19yzz33MGnSJD755BPsdjuff/4511xzDRkZGXz99de0tLR0+fAgRp6OgrLOld+nKyibEutHTJg340PcGas2oSsrxpy3H/OufOxVVdR2er4hvNO16piBuVbdmaKqlLVUcLi9Kjy/sRiH2rZ876YzMmVMEkkBbdeugzwCByyO0cRsNuPu7o5Go2Hx4sVYLBbn9sTi7I2o5N2bmXJ6ejpw5jZ2EyZMYMKECb06Z8eyeUNDA7/+9a+drex8fX05evQoWVlZeHp6YrPZnK/54U5jRUVFREdHO6/3JCcnA1BUVMSMGTMA8PT0JDo6mrKyMgASExOBti09x48f7/zaau062wkJCaG8vJy4uDjCw8N5+eWXsVgs3Hzzzd2+p7y8PI4dO8a9994LtCX6EydO8NBDD/H+++/z5ptvMnHiRAAeeeQR/vCHP/Dpp58SFRXFggULejVuYnhQFJWK2taTm3NUtFBSZTp9QVmYN9FhXkT5qHhWlWLJP4p5Rz6WwkLqLCd/L9tm1ROcydotevyAzqo7NFlbnEvhR+ryaLad7Ng1zjvcuRQe4zsWnVbuK+4vqqpy9OhRtm/fzoUXXkhkZCReXl549dO99KPViErertQxQ77vvvv43//9XzZt2oS3tzf3338/JSUlfP7553TXRj4iIoLCwkIsFgt6vZ6cnByWLFlCVFQU2dnZLFy4EJPJRF5eXrd9brtz+eWX8/vf/54JEyYwZswYALKyspzLVEajkdraWsLCwsjJySEqKoqoqCimT5/Or3/9axRF4Z133iEiIoK//vWv/OpXv8LNzY0HHniAgwcPcvToUW699VYCAgJ46aWX2Lp1KxdffHHfBlO4hKqq1DRau2zOUdhNQVl0mBcx4d6MD3YnwFyLNa8AS94uzOsKaKqqwtnoU6PpdK06elBm1R3sioP8xuJOhWYVzsd8DF7OpfBk/1i8jZJIBoLVamXLli3k5eVhNBq7TGJE30jy7kfR0dFcffXVvPLKK9x666089dRTZGdn4+7uTmRkJNXV1ad9nb+/Pz/+8Y+566678Pf3x929bZegZcuW8eKLL/KLX/wCq9XKLbfcQkBAwFnFlJSUxJ133slzzz2H3W7HbDYTGhrqXD6/4YYbeOihhwgLC3NWn8+bN4+srCzuvvtuWltbWbhwIZ6ensTExHDnnXfi7+9PUFAQEyZMwM3NjV//+tf4+vri6enJ3Llz+zCCYjD1qqBsTHtBWbgX0WFehLs7sBcWYsnPxvx126y6vPOs2rPrrNo9JhrtIO61XNVa69zc43hDgbPQTKfRkeAfzYT22XWEV6hcZx1glZWVrF+/nqamJkJDQ0lPT5dLav1IdhUTfSLj2HeDMYZmq4PC9tuzCiqaKaxooaax6yWWMb5GYsLblr7Hh3kRFeSOtuoE5rx8LPn5mPPysVd1+gDaZVbdXgEeGjIos2rn+7JbOF5fQJGlnL2lB6gx1zkfC/EY074UHku8fzRuOrkN6Uz68/ewpKSEL7/8ElVVmT59OjNmzBgVhayyq5gQ4px1KShrLyrrrqCs4zp1dJgXng5zW6LOO4R5Sz4VhUWo1h/MqidNdFZ/D/asGtoKzUqby537XOc3lqC0F5q569yYOibZ2dFsjLv/oMYmTgoLC2Ps2LFMnTrVWTAr+pckbyGGsc4FZR1FZacrKEsY50NMp0Qd4KXDVlaOOS8Hy6YC6vLzqTrdrNq5Bebgz6o7NFqbnUvhR+vzaLaZ2kIExvlEkBwQx4L42Xhb3aXQzIVKSkowmUwkJiai1+u55JJLXB3SiCbJW4hhwllQ1ukWrTMVlHUUlYUGuqM2NbfNqr/fh/nzfApPmVV7unxW3cGu2MltKOJIXS5H6/IobTnhfMzX6M3s0GlMCIgjMSAWb0NblXp4sFy+cRWHw8Hu3bvJzs7GYDAwfvx4abgyCCR5CzFEnVVBWXtRWWSQJ3qNirW0DHPe91h2FVCSl4+9uuus2hgRfnILzNgYDCGumVVD24eSytYaZ1V4Tn0hVqWtKlmv0ZHoH9NWaBYYR7hniBSaDSGNjY2sX7+eqqoq/Pz8yMjIkMQ9SCR5CzEEdBSUtfX87r6gLGlcIDHhbQVl40O9cDfqcDQ2Yc7Lx7w9n6q8fCxnmlXHxuAePd5ls+oOrXYzx+rznVtn1loanI+FegS173MdS7xfNEbdue8TIAZOTk4OW7duxWazkZCQwPz586U3+SCS5C3EILPZFUqrTO2V3y2UVB+m5ERT14Iyj1MLynw8DagOB9aSUsw5mTR+k09lfsHpZ9XtncraZtXBLptVd1BUheKmco7U5XK4LpfCxhKU9nfsoXNjWtAEZ5OUQHc/l8YqeqaqKnl5eQAsXrzY2XhKDB5J3kIMoLaCMnOXpe+SKhN2x8lU7eGmI2GcT9s16vZkHejbti2ivbERS14+5v0FlOXlYykoRO3U6ELr5YnH5Eknr1UPgVl1hwZLk7Mq/GhdHiZ7KwAaNIz3iSApII4JAXFE+Uai04z824hGgpaWFry8vNBoNCxatAiLxYKfn3zYcgVJ3kL0ky4FZe2bc5yuoGxs8MmCsugwb1ImxVB5oqJ9Vl2Ced8BqvLyMefnY6+uOXkCjQZjRISzU5lbx33VQ+QasE2xk9tQ2F4Znke5qdL5mL/Rl6lhbbdxJfrH4GUYGh8wRO+oqsqhQ4fYuXMn6enpxMTE4O7u7mwoJQafJG8hzlGTydZ+e1Z75XdFC00/KCgLG+NBjDNRexEZ7IlB3zbLbJtVH6N43zZqsg+0XavuZlbt3t4DXDuE/liqqsqJ1mrndeuchkJsStv7N2j1zmXwCQFxhHoGDZkPGeLsmM1mNm/eTGFhIe7u7uj1kjaGAvm/IEQvdC4o65hZn66gbMa4QGfld0dBGeCcVZu27mlvhFKAvaabWXXnCvAhlvBMtta2QrP2a9f1lkbnY2Gewc5kHesXJYVmI0BZWRkbN26kpaWFiIgIFi9eLBuKDBGSvIX4gY6CMmfjk4oWKmpaTykomxzj52wn2lFQ1sHe0Ijl0EFq8vOxdFSAd5lVe+E5ZRJuMTFEzJpJi4/3kJpVd1BUhaKmMue168LGUtT2kfDUuzM9eKJzhu3v5uviaEV/Ki0t5YsvvgBg1qxZTJs2bVS0OB0uJHmLUa03BWVuBq2zoCw6zJuYTgVl0DarthSX0NDe//u0s+rICOdGHT+cVfuHh9M6hBqM1FsanftcH6vPw2Q3A22FZtG+kc5kHeUTgVYKzUas8PBwYmNjmTRpEmFhYa4OR/zAgCVvRVF44oknOHr0KEajkWeeeca57zTAp59+yltvvYVWq2X58uW92otbiL5QVZXaRqtzc46O69Q9FZSFBbqj1Z5cvrY3NGLKOuLcsONMs+qheK36h6wO28lCs7o8KkxVzscC3PyYFtQ2u04MiMFTP3Tfh+i7goICmpqamDJlClqtloyMDFeHJLoxYMl73bp1WK1W3n//fbKysnj++ed5/fXXnY+/8MILfP7553h6erJ06VKWLl0qtxyIftVRUNZRVNZdQVnbLVqnFpQBqHY7lqKittu18tqWwO21tSdP0nlW3X5vtT4keMhdq+5MVVXKTVXtt3DlkttQ1KXQbEJAPMmBbdeuQzzGDOn3IvqH3W5nzZo17NmzB4PBQEJCglSSD3EDlrz37t3LwoULAUhJSeHgwYNdHk9KSqKpqQm9Xo+qqvIHQvSJ2eqg8EQLBeW9LChr3/ayo6Csg72hgZaORJ1f0M2serIzWbuNjxrSs+oOLTYTR+vynS1IG6xNzsfCvUKc+1zH+kVh0MrVtNGkrq6O9evXU1tbS0BAABkZGZK4h4EB+1fa3NyMt7e383udTofdbnfeZpCQkMDy5cvx8PBgyZIl+PqeudglKCgIg6F/q1e72ydVnJ3BHkebXaGgvIFjRfUcL67jeHE9xSea6LwzvZ+3kZkTQkkY509iVADxY/3x9+nac1mx2WjJL6Dp6DGajh6l6egxLJUnl4zRavEaPx6f5ER8ktr+cw8PH5APmv09hg7FwfGaAvZXfM/+iu/JrS10Fpr5GL2YFzWTlLCJTA2bQKCHf7+e21Xk3/PZ27dvH19++SV2u50ZM2Zw0UUX9fvf2dFmsH4PByx5e3t709LS4vxeURRn4j5y5AibNm1i/fr1eHp68sADD/Dll1+ecQu56s4tIPvBQGyaPhoN9DgqqkpFTS8KyiJ9iA73crYTHdOpoAwUWptraSqpx5Jf0NYHPC8fa1Fx11m1tzeeUyefvFY9fjxa97aE7wDqASoq+v099tcY1prrnbdwHa/Lp9VhAUCLhhjfsc59rsd5hzsLzSz1rZTXt/b53K4m/57PzdGjR9FqtVxwwQXMnz9fxrCPBuL3sLsPAwOWvFNTU9m4cSOXXnopWVlZJCYmOh/z8fHB3d0dNzc3dDodgYGBNDY2nuFoYjToKCjr2JyjoKKFohMtmK0nC8p02q5bXp6uoAzarlWbi0varlXnd1yrrjv5BK0WY2SEs1OZe2wM+uDh1UjE4rCS01DobJJS2Xqywj3QzZ/pIZPbO5pF4yGFZqJdQ0MDvr6+aDQa5s+fz4wZM/Dx8XF1WOIsDVjyXrJkCdu2beP6669HVVVWrVrFZ599hslkYsWKFaxYsYIbb7wRg8FAVFQUV1111UCFIoaozgVlHe1Ez7agrIO9vt5ZUGbOK8BaVIRqP3ks56y6Y8OOTrPq4UJVVcpaKp3XrXMbinCoDgCMWgOTAhOct3EFewQOqw8iYuCpqsr+/fvZvXs3ixYtIikpCYPBIMvkw9SAJW+tVstTTz3V5WdxcXHOr2+44QZuuOGGgTq9GGI6CsoKK1rIL+++oCx1XIBzc47TFZRBewV4b2bVsZ1m1UHDa1bdodlm4mhdnjNhN1qbnY9FeoWRHBjLhIA4YnzHoZdCM9ENk8nExo0bKS0txdPTU2baI4D8axf9zu5QKOnoUNZe/V3+gw5l3u0dyjqWvqPDvPD1Ov0MoMdZtY83nlOnOHfWcosej9ZteM2qOzgUBwVNJRypzeVwXR4lzWXOcfM2eDIjZAoTAuJICojF1+h9xmMJAVBcXMymTZtobW0lKiqK888/H48hsvOcOHeSvEWfKIpKeU2rc3OObgvKxrZ3KAs/XUHZSV1m1e3/Oep+MKseG+nsVDacZ9UdKpur+bZsD0fqcjleX4C5o9BMoyXOb7xzKTzSOwztMH6fYvCVlZXx5ZdfotVqmTt3LpMnTx7W/1bESZK8Ra91LijrKCorrtxHq+XkLFinbetQFhN+cietsECPUwrKOtjr6p2dypwV4N3Nqjvuqx6ms+oOFoeV4/UFzqXwqtaTTV+C3AOYGTKF5MA4EvyicdcP7/cqXCs8PJwJEyYwYcIEgoKCXB2O6EeSvEW3Tikoq2ihydS1oGxsqA9jg9ycRWXdFZQBqDYbluISZwOU7mfVna9VD/8OX4qqUtZyon2f61zyGotwqG0V9G46IzMjphLtGcmEgDiCPAJdHK0Y7nJycmhoaGDGjBloNBpnsywxskjyFl0cLmxga3YVBRUtVDdYujw2xtdIamKAc3OOqFAvYqPHdntfo72uDnNeQbezap2PD57Tpp68Vj0CZtUdmqwt7e1H24rNmmwnex6M9Q4nOaCt0CzadxzjIrsfQyF6y2azsW3bNo4dO4bBYGDixIlybXsEk+QtnFRV5e9f5VPXZD2rgjL4way6vQrcUVd/8gkds+rYGGcTlJEwq+5gVxzkNxY7l8JLmk82c/ExeDErZCrJ7YVmPkbZD1n0r+rqatavX09DQwNBQUFkZGRI4h7hJHkLp6ITJuqarMxODuS2pXFnTKxts+p88k9UUnPgIJbiEuhuVt1xrdpoHIy3MWiqWmudS+HHGwqwONpufdNptCT4RZMc2FZoFuEVKoVmYsAcPHiQHTt2oCgKU6dOZdasWeh0p95iKUYWSd7CKSun7frz9ISuDT7aZtXFzr2qzzSr7kjW+jEjZ1bdwWy3cLy+gMPts+sa88nr9cEegc6q8AT/aNx0I+uDihi6GhsbMRqNpKWlERUV5epwxCCR5C2csnLq0Os0TIrxQ2ltpe6LrzHn5Jw6q/Y9OauOmDWTZm+vETerhrZCs9Lmcmeyzm8sQelUaDZlTBLJAW1bZ47xCHBxtGI0qa6uZkz7B+Q5c+aQkpKCp6enq8MSg0iStwCgst5MaXUrU2L9cDfqqF/7HQ1r14FWi9u4sSe3wIyJ7jKr9gsPxzSCiq0arc3OpfCj9Xk020xAW2X9OJ8Ikv1jSQ6MI9pnLDqtLE2KwaUoCvv27SMzM5M5c+YwdepUdDqdJO5RSJK3AGB/+5J5SnzbDLJlXyZotUQ99wx6vzNv1zqc2RU7eQ0nC81KW044H/M1ejM7dBoTAuJIDIjF2yB/IIXrNDc3s2HDBioqKvD29iY0NNTVIQkXkuQtAMjKqUcDTIsLwF5bhyW/AI/kpBGXuFVVdRaaHa7LJae+AKvSti2oTqMj0T+GCe1bZ4Z7hoy46/ZieMrPz2fLli1YLBZiY2NZuHAhbiPktkpxbiR5C5pMNnJKm4iN8MbXy0DDjiwAvFJTXBpXf2m1mznW0dGsNpdaS73zsVCPoPaq8Fji/aIx6mSHJTG0VFRUsHbtWnQ6HQsXLiQ5OVk+VApJ3gKyc+tR1U5L5plZoNHgmTLNtYGdI0VVKG4ud+5zXdBYgtK+vYeHzo1pQRPaK8NjCXT3d22wQvQgNDSUlJQUEhISCAiQwkjRRpK3ICu3/Xp3QkDbDl65ebjHx6H3HT5L5g2WJud166N1ebTYW4G2QrMon0hnVXiUbyQ6zenbtwoxFKiqytGjR6mtrWXevHloNBpmz57t6rDEECPJe5Sz2Bx8X9BIeKA7oQHuNGzaCao65JfMbYqd3IZCjtblcbgul/KWSudjfkYf5oSmMCEwjkT/WLwM0mlKDA8Wi4WtW7eSl5eH0WiUW8BEtyR5j3LfFzRisyukJHRUmWcB4JWS4rqgTkNVVU60VjuXwnMaCrEpbfeeG7R6kgNinU1SwjyD5ZqgGHZOnDjBhg0baGpqIjQ0lIyMDEncoluSvEe5rE63iNkbGzEfz8EtNhZ9gL9rAwNMtlaO1ee3L4fnUWdpcD4W5hnsTNhxfuOl0EwMa/v372fXrl2oqkpqaiqpqalotXJ5R3RPkvco5lBUsvPq8fMyMD7Mi+at21y6ZK6oCkVNZc6OZoWNpajthWaeendSgiY6C80C3P1cEqMQA8Fut+Ph4UF6ejoRERGuDkcMA5K8R7Gc0iZaWu2cPy0ErUbT1pgF8B7E5F1vaeRw+1L4sfo8THYzABo0jPftVGjmE4FWCs3ECFJRUUFISAharZbp06czadIk3N3dXR2WGCYkeY9iJ5fM/XE0N9N67Dhu0ePRBwYO2DmtDhu5DYXOpfAKU5XzMX83X+dtXIn+MXhKoZkYgRwOB7t37yY7O5sZM2YwY8YMtFqtJG5xViR5j1KqqpJ1vA53o46kKF9aduwARcErdfqAnE9RVd4//jl7Kw90KTSb0F5klhwYR6hHkBSaiRGtoaGBDRs2UFVVhZ+fH+PHj3d1SGKYkuQ9SpVWt1LTaGVmUiB6nfZklfn0lAE53+Ha4+yoyGSMuz9Tg5KdhWYGrfwKitHh+PHjfPvtt9hsNhITE5k/fz4GgxRainMjfzlHqazjJxuzOFpMtB4+gjFqHIbgoAE538bSHQDcNvE6Ir3DBuQcQgxVlZWVbNy4EYPBwOLFi0lISHB1SGKYk+Q9SmXm1KHTapgc44dp3572JfOUATlXSXMFx+sLSPSPkcQtRhVVVdFoNISEhDBnzhyio6Px85M7JUTfSfnuKFTTYKG40kRylC+ebvoBXzLfVNI2604be96AHF+IoUZVVQ4ePMjGjRtR1bbbHadNmyaJW/QbmXmPQvvbe5lPi/dHaW3FdPgIxshIjAOwP3CDpYl9VQcJ9QhiQkB8vx9fiKHGbDazadMmioqKcHd3p6WlBW9vb1eHJUYYSd6jUFZOPdC2d3dL9n6w2wdsyXxr2W4cqsL5kXPQSiW5GOHKysrYsGEDJpOJyMhIFi9eLC1OxYCQ5D3KtLTaOVbcSHSYFwE+RiraG7MMxC1iVoeNbeV78dJ7MCt0ar8fX4ihZN++fezZs8e5C9i0adPk1kcxYCR5jzLZefUoKkxPCEAxm2k99D2G8DCM4f1fSLbrxH5M9lYujFoovcfFiGc0GvHx8SE9PZ3QAbgEJURnkrxHmf3tXdWmxQVgOngI1W4fkFm3oqpsLt2JTqNjYcSsfj++EENBcXExERER6HQ6Jk2aRGJiIkaj0dVhiVFAqs1HEatN4VBBAyEBboSPcXdWmQ9EL/Pva49T2VrDjJDJ+BqlWEeMLHa7nW+//ZYvv/ySXbt2AaDRaCRxi0EjM+9R5EhRAxabQkp8AKrViungIQyhIRgGYBcj5+1hkXJ7mBhZ6urqWL9+PbW1tQQGBpKcnOzqkMQoJMl7FOmoMp8eH4Dp0PeoVite01P6vaimpLmc4w0dTVnk2p8YGVRV5ciRI3z33Xc4HA4mTpzIeeedh14vf0bF4JPfulFCUVT259bh46knJtybqi+zgIGpMt9UshOAxdKURYwgNTU1bN26FTc3N9LT04mJiXF1SGIUk+Q9SuSVN9NksrNgSjDYbZgOHEQfFIRx3Nh+PU+DpYm97U1ZkqUpixgBOlqcBgUFsXDhQsaNGydNV4TLScHaKOHciCQ+gNbvj6BaLHil9v+S+day3SiqQtpYacoihjdFUcjMzOSrr75ytjidMGGCJG4xJMjMexRQVZWsnDrcDFomjPel7t2BacxicVidTVlmhkhTFjF8mUwmNm7cSGlpKZ6enjQ3N+Pj4+PqsIRwkuQ9CpTXmKmst5CaEIBeddCy/wD6wADcxkf163l2n8jGZG/lImnKIoaxoqIiNm3ahNlsJioqirS0NNzd3V0dlhBdSPIeBbJyTu7d3XrkKKrZjNeCef26ZK6oKptKd6DT6FggTVnEMLVnzx727duHVqtl3rx5TJo0SVqciiGpx2veDQ0NPPbYY9x8883U19fzyCOP0NDQMBixiX6SlVOHVgNTYv1pdvYyT+nXc3xfe4yq1lpmhkyRpixi2PL19cXPz48rr7ySyZMnS+IWQ1aPyfvxxx9nypQp1NfX4+npSUhICA888MBgxCb6QV2TlYKKFhLH+eKpB9P+bHT+/rhFR/freTY69+ye06/HFWKg5eXlYbPZAEhISGD58uUEBQW5OCohzqzH5F1SUsKKFSvQarUYjUbuu+8+KioqBiM20Q869u5Oifen9egxFFNrW2MWbf/daFDcVE5OQyFJ/rFEeElTFjE82Gw2Nm7cyLp169ixo+3Dp0ajkaYrYljo8bdUp9PR1NTkXD4qKChA249/+MXA6rjePS0+gJZPNgD9v2S+qVRm3WJ4qa6uZv369TQ0NBAcHMy0adNcHZIQZ6XH5H333Xdz0003UV5ezs9//nOysrJYtWrVYMQm+shksXO0qImoUE8CvfQU7s9G5+uLe1xsv52j1lTPvqpDhHpKUxYx9KmqysGDB9m5cyeKojB16lRmzZqFTqdzdWhCnJUek/eiRYuYPHky2dnZOBwOnnrqKXx9fQcjNtFHB/MbcCgqKXEBmI/noDQ343v+wn5dMv8qZ1NbU5bI86Qpixjy6urq2LFjB25ubixevJhx48a5OiQhzkmPyXvFihW8//77pKWlAW1dh6644go+++yzgY5N9JGzq1pCAC1rvwX6tzGLxWFlbe5WvAyezAyZ0m/HFaK/KYqCVqslMDCQ9PR0wsPD8fT0dHVYQpyzbpP3zTff7NynNjk52XnNW6fTkZ6ePjjRiXNmsysczK8nyM+NiEA3irOy0Pp44x4f12/n2HViPy1WExdFLZKmLGJIUhSFvXv3UlZWxrJly9BqtcTF9d+/ASFcpdvk/c477wDwzDPP8Nhjjw1aQKJ/HC1uxGxVWDDFH0tePo7GJnwWzkfTT9f2FFVlU8lO9Fo9CyJm9ssxhehPzc3NrF+/nhMnTuDj40Nzc7Nc8hMjRo/L5g888ABr166lpaUFAIfDQUlJCffee++AByfO3f72vbunxQfQ8t3XAHhNT+m34x+qPUa1uZbFMfOkKYsYcvLz89m8eTNWq5XY2FgWLlyIm5ubq8MSot/0mLzvv/9+GhoaKCoqYubMmezcuZPU1NTBiE2cI6V9IxIvDz1x4V6U7stC6+WJR1Jiv51jU3tTlqWJ6dDab4cVos927dpFVlYWOp2ORYsWkZSUJJ3SxIjTY9nx0aNHeeedd1iyZAm33347//rXvygtLe3xwIqisHLlSlasWMFNN91EYWFhl8ezs7O58cYbueGGG7jnnnuwWCzn/i5EFwUVLTS02JgW64+9sBBHQwNe06b125J556YsUf6R/XJMIfpLSEgIgYGBXH311V3qdYQYSXpM3mPGjEGj0RATE8PRo0cZN26cs5Xgmaxbtw6r1cr777/P/fffz/PPP+98TFVVHn/8cZ577jn+9a9/sXDhwl59IBC9s7/TRiQtA9DLvKMpy+Kx5/XbMYU4V6qqkpWVhdlsBiA6Opqrr76agIAAF0cmxMDpcdk8ISGBp59+mhtuuIFf//rXVFZWOjemP5O9e/eycOFCAFJSUjh48KDzsfz8fPz9/Xn77bc5duwY559/PrGx/dc4ZLTLOl6HQa8lOcqHE3/NQuvhgUdyUr8cu97SyL6qQ4R5BpMcIFW7wrUsFgtbt24lLy+PhIQEFi9eDCBdIMWI12PyfuKJJ8jMzCQ+Pp67776b7du38/vf/77HAzc3N+PtfbKQSafTYbfb0ev11NXVkZmZyeOPP8748eO58847mTx5MnPnzu32eEFBQRgM/Xs7Unh4eL8ebygoqWyivNbMnElhBFnMlNXVEbw4jYh+akaxMXsniqpwxcQLiYiIAEbmOA42GcOzV1xczOrVq2loaCAqKoqlS5fi5+fn6rCGNfk97LvBGsMzJu+8vDy8vLyYObPtVqCMjAymTZvGK6+8wlNPPXXGA3t7ezsr1KHtGnhHw39/f3/Gjx9PfHxbO82FCxdy8ODBMybv6urq3r2jXgoPD6e8vLxfjzkUrN3V9p4mjPOg8Jt1AGgnJPXLe7U4rHx9fDPeBk/i3aIoLy8fseM4mGQMz46iKOzfv589e/YAkJqaytKlSzlx4gQmk8nF0Q1f8nvYdwMxht19GOh2benVV19l+fLlXHzxxXz33XcA/PWvf+Wiiy7q1fXp1NRUtmzZAkBWVhaJiScrnceNG0dLS4uziG3Pnj0kJCT0/t2Ibu3PqUOjgckxfrRkZqJxd8djQnK/HHvXif202s0sCJ8pTVmEyzQ1NbFv3z48PT1ZunQpM2fOlGVyMep0O/NevXo1X3/9NZWVlbzyyiv87W9/48SJE/zxj390Xss+kyVLlrBt2zauv/56VFVl1apVfPbZZ5hMJlasWMGzzz7L/fffj6qqTJ8+3dl+VZy7hhYreWXNxI/1wa3mBPbqGrxnzUTbD5cbFFVpa8qi0TFfmrIIF+i47Obn58eSJUsICQnB3d3d1WEJ4RLdJm8vLy9CQkIICQkhOzubK6+8kjfeeKPXu+9otdpTltY7tyWcO3cuH3zwwTmGLU4nO7ceFUiJD6BlX1tr2/7qZX6o5jjV5lrOC5suTVnEoHI4HOzatYvS0lKuvPJK9Ho9UVFRrg5LCJfqNnl3XoYKCAjg4YcfHpSAxLnLbN+IZFqcH82fZKJxM+IxaUK/HHtj6XYAzo+UPbvF4GloaGD9+vVUV1fj5+eHyWSSFqdCcIbk3bmxgSxNDX1mq4MjRY2MDfbAr6WW5qoqvGakojUa+3zs4qYychuKSA6IJcIrpB+iFaJnx44dY9u2bdhsNhITE5k/f36/33EixHDVbfI+fvw4GRkZAJw4ccL5taqqaDQa1q9fPzgRil45lN+A3aEyLa5TY5Z+6mW+sXQnAGmR0pRFDI4dO3aQnZ2NwWAgPT3deWeKEKJNt8n766+/Hsw4RB9lde6q9noWGoMBz8mT+nzceksjmdKURQyyqKgoKioqSE9Pl2VyIU6j2+QdGSk9q4cLu0PhQF49AT5GwhwNlFZU4JkyDa1733dR2lK2C0VVSIs8T3pEiwGjqiqHDh0iJiYGLy8vIiIiuOKKK+R3TohuyM2RI8DxkiZMFgcp8f6YMvcD4N0PvcwtDivfle/D2+DJzNApfT6eEKfT2trKV199xXfffcf27dudP5fELUT3emyPKoY+55J5fAAt72SCXo/nlMl9Pu7Oiixa7WYuHn8+Bq38qoj+V1paysaNGzGZTERGRjJv3jxXhyTEsNCrmffevXv517/+hdVqZffu3QMdkzgLqqqSlVOPp5uOaIMJa2kZnhOT0Xp49Om4iqqwubStKcuCcGnKIvqXoijs2rWLNWvW0NrayuzZs7n00kvx9PR0dWhCDAs9Tqfefvtt1q1bR2VlJRdffDErV67kmmuu4bbbbhuM+EQPik6YqGuyMmfiGMz7s4H+aczS1pSljvPCpuNj9Orz8YTorKWlhUOHDuHj40NGRgYhIXILohBno8eZ98cff8xf//pXPDw8CAgI4IMPPuDDDz8cjNhELziXzDtuEdPp8JzS9+vTHU1Z0qQpi+hHVqsVAB8fHy655BKWL18uiVuIc9Bj8tZqtRg7Nfpwc3PrdYtUMfCycurQ6zQkeduwFpfgkZyEzqtvS49FzqYscYRLUxbRD+x2O1u3buXDDz90JvCwsLAuf1uEEL3X47L57Nmz+e1vf0trayvr1q3j/fff57zzpFnHUFBZb6a0upUpsX7YD/bfkvmmkh0ALJamLKIf1NbWsn79eurq6ggMDMRsNkvSFqKPekzeDz74IP/+979JSkpi9erVnH/++Vx//fWDEZvowf7OVeZrPgWtFq9pfVsyrzM3kFn9PWGewSQFxPZHmGKUUlWVw4cPs337dhwOBxMnTuS8885Dr5c7F4Toqx7/FT3//PNcfvnlkrCHoKycejTA5ACVusJCPCYko/Pu245fW8t2o6gKi8dKUxbRNzt37iQ7Oxs3NzcyMjKIjo52dUhCjBg9Ju+oqCieffZZGhoaWLZsGcuWLWPs2LGDEZs4gyaTjZzSJmIjvNEcPQiAVx8bs7Q1ZdmLt8GLGSHSlEX0TVxcHDU1NZx//vl49/FDpRCiqx4L1n70ox/xr3/9i7/+9a8YjUbuuusubrzxxsGITZxBdm49qtqxd3cWaDR4TZvWp2PurMii1WFhYcRMacoizpqiKGRlZdHQ0ABAcHAwS5culcQtxADoVZOWpqYmtm3bxrZt23A4HMyfP3+g4xI9yMptu949NUSDJS8f98QEdL4+53y8zk1Z5ktTFnGWWlpa+OKLL9i1axc7duxwdThCjHg9Tq/uvPNODh06xIUXXsi9997LtD7O7kTfWWwOvi9oJDzQHc+8I7TS9+0/D9Yco9pcx1xpyiLOUlFREZs2bcJsNhMVFcX555/v6pCEGPF6TN7XXXcdixYtkgrRIeT7gkZsdqVt+8+d69uWzKf37UNVx+1h50tTFtFLDoeDXbt2ceDAAbRaLfPmzWPSpElS6CjEIOg2I7/66qvcfffdrF27lrVr157y+HPPPTeggYnudXRVmxaqw5ybh3tcLHo/v3M+XmFTKbmNRUyQpiziLFgsFo4fP46fnx8ZGRkEBQW5OiQhRo1uk/ekSZOAtiYtPySfrF3Hoahk59Xj720gsOw4tara58Ysm0p2ApA2VpqyiJ6ZTCY8PT3x9PTk0ksvxc/PD4PB4OqwhBhVuk3e6enpAFRWVvKzn/2sy2O///3vBzYq0a2c0iZaWu2cPy0EU+ZWgD4tmdeZG8iqOkS4ZwhJ/tKURXTParWybds2SkpKuOaaa/Dw8JDZthAu0m3yfvHFF6mpqWHDhg0UFBQ4f+5wONi/fz+/+tWvBiM+8QPOjUjC9Zg/Oo5bbAz6gIBzPt6Wsl0oqKRJUxZxBlVVVWzYsIGGhgaCg4NxOByuDkmIUa3b5H3hhReSm5vLjh07uiyd63Q6fv7znw9KcKIrVVXJOl6Hu1FHZE1++5J5yjkfz+Kwsr18X3tTlsn9F6gYMVRV5cCBA+zatQtFUZg2bRozZ86UzYmEcLFuk/fUqVOZOnUqS5YskSYLQ0RpdSs1jVZmJQfSmtW2ZafX9HO/3t3RlOWSsedJUxZxWjt27ODAgQN4eHiwePFi6a4oxBDR7V/sq666io8//piZM2d2WU5VVRWNRsPhw4cHJUBxUtbx9iXzSDdavziG2/jxGMYEntOxFFVhU3tTlgXSlEV0Izk5maamJhYsWICnZ9+2mhVC9J9uk/fHH38MwJEjRwYtGHFmmTl16LQaYhoLaVSUPi2ZH6g5Ro25jrlhqXhLUxbRTlEU9u7dS2xsLGPGjCEgIIALL7zQ1WEJIX6gx/aoRUVFfPrpp6iqysqVK1m+fDkHDx4cjNhEJzUNFoorTSRH+WLL7ti7O+Wcj7eppG3ZPU2asoh2TU1NfPbZZ2RmZrJr1y5XhyOEOIMek/cjjzyCoiisX7+e/Px8HnnkEZ555pnBiE10sr+9l3nKODdajxzFOG4shuDgczpWYWMpeY3FTAiIJ8zr3I4hRpa8vDw+/PBDTpw4QVxcHBkZGa4OSQhxBj1WKVksFq688koeffRRli1bxsyZM7FarYMRm+gkK6cegCRTKSaHo0+NWTaVtrVCXSxNWUY9u93Od999x5EjR9Dr9SxatIikpCS5bVCIIa7HmbdOp+Prr79m06ZNpKWlsW7dOrTaXm1GJvpJS6udY8WNRId5weEDwLkvmdeaG8iq+p5wrxAS/WP6MUoxHNntdoqLiwkMDOSqq64iOTlZErcQw0CPM++nnnqKv//97/zmN78hJCSENWvWyLL5IMvOq0dRYcZ4T0zvH8EYGYExNPScjrW1oylLpDRlGa1UVaWpqQlfX1/c3d2de27L5kNCDB89TqGTkpK49dZbqays5O9//zt33HEHycnJgxGbaLe/vavaRGs52O3nvGRutlvYXr4PH4MXM6Upy6hksVhYt24dH330Ec3NzQD4+/tL4hZimOkxea9evZq77rqLkpISysrK+MUvfsEHH3wwGLEJwGpTOFTQQEiAG4bjh4Bz37t754m2piwLI2ahl6Yso05FRQUffvgh+fn5jBkzRlZehBjGevwL/tZbb/Gf//yHgPb+2XfeeSc333wz11xzzYAHJ+BIUQMWm0LqeC9aP/keQ1gYxojwsz6OoipsLt2JQatnfviMAYhUDFWKopCVlcXevXsBmDFjBtOnT5faFSGGsR6Tt6IozsQNEBgYKJ/YB1FHlfk0tRLVZjvnQrUDNUepMdczT5qyjDo7d+7kwIEDeHl5kZ6eTnj42X/4E0IMLT0m76SkJJ599lnnTPuDDz6Qa96DRFFU9ufW4eOpxyv/ACY45+vdG0vabg87f6w0ZRltJk+ejNlsZu7cubi7u7s6HCFEP+hx3eyZZ57BaDTy3//93zzyyCMYDAZ+85vfDEZso15eeTNNJjvTx3vTeugQhpBgjJERZ32cgsYS8huLmRgYT5inNGUZ6RwOB9u3b6eiogIAHx8fFi9eLIlbiBHkjDPv5uZmioqK+MUvfsEDDzwwWDGJdh0bkUzXVqNarHilTj+nSxabSncCkBYpTVlGuvr6etavX09NTQ11dXVceumlrg5JCDEAuk3eX375JQ899BCenp5oNBpefvnlLvt6i4GlqipZOXW4GbSMKT16zkvmteYG9ld9T4RXqDRlGeGOHTvGt99+i91uJykpiXnz5rk6JCHEAOk2eb/++ut88MEHJCYmsnXrVl599VXefffdwYxtVCuvMVNZb2FmrA+tWw6hDxqDcdzZ76V8sinLHCk0HKFsNhtbt24lJycHg8FAeno68fHxrg5LCDGAur3mrdFoSExMBGDhwoXU19cPVkwCyGpvzDLTWItqNuM1/eyXzM12C9+1N2WZIU1ZRrSqqiqCg4NZvny5JG4hRoFuZ94/vAdUOjANrqycOrQaCDuRg5lz62W+40QWZoeF9LFzpSnLCKOqKnV1dQQGBmIwGFi6dCmenp5y77YQo0S3f9FbWlrYs2cPqqoCYDKZunw/a9aswYlwFKprslJQ0cKEsV5Ydx5EFxCAW/T4szqGoipsLmlvyhIxc4AiFa7Q2trKpk2bKC8v5+qrr8bf3x9vb29XhyWEGETdJu/Q0FBefvll5/chISHO7zUaDe+8887ARzdKdezdfZ57HUprK37zzn4Tkezqo9Ra6pkXPgNvg+dAhClcoLS0lA0bNtDa2srYsWMxGo2uDkkI4QLdJm8pTnOdjuvdY2vysHFuVeabSrcDkBYpTVlGAkVR2LNnD1lZWWg0GubMmcPUqVOlCFGIUUouhA4xJoudo0VNjA92w5F9EJ2fH24x0Wd1jLamLCVMDEwg1DNoYAIVg2rXrl1kZ2fj4+NDRkYGISEhrg5JCOFCkryHmIP5DTgUlXleTSgtJnwXn4/mLIuQTjZlkVn3SDFt2jTsdjuzZ8+WpXIhRM/tUcXg6uiqFlOfD5z9knmtuV6asowAdrudrVu3UlRUBICHhwcLFiyQxC2EAHqRvBsaGnjssce4+eabqa+v55FHHqGhoWEwYht1bHaFg/n1BPsa4MghdD4+uMfFntUxtrQ3ZVkcefZFbmJoqK2t5eOPP+bw4cMcOHDA1eEIIYagHpP3448/zpQpU6ivr8fT05OQkBDpcz5AjhY3YrYqLPBtRmluxnP6tLNaMjfbLWwvz8TX6E1qyKQBjFQMBFVV+f777/n444+pq6tj0qRJXHTRRa4OSwgxBPWYGUpKSlixYgVarRaj0ch9993n3K3oTBRFYeXKlaxYsYKbbrqJwsLC0z7v8ccf58UXXzz7yEeg/e17d8c3FQDgfZZL5jsqMjE7LCyMmCVNWYYZq9XK2rVr+fbbb9Hr9Vx44YXMnz9fmiMJIU6rx+St0+loampyLsEWFBT0qovTunXrsFqtvP/++9x///08//zzpzznvffe49ixY+cQ9sijtG9E4u2uxXj8MFovL9wTet/m0qEqbC7dhUGrZ174jAGMVAwEnU5Hc3Mz4eHhLF++nOjoaFeHJIQYwnr8WH/33Xdz0003UV5ezs9//nOysrJYtWpVjwfeu3cvCxcuBCAlJYWDBw92eTwzM5P9+/ezYsUK8vLyzjH8kaOgooWGFhsXh5lxHGzEZ/48NDpdr19/oPoItZZ65ktTlmFDURSqq6sJDw9Hp9NxySWX4ObmJi1OhRA96jF5L1q0iMmTJ5OdnY3D4eCpp54iKKjne4ebm5u7tGzU6XTY7Xb0ej2VlZX86U9/4k9/+hNffvllrwINCgrCYDD06rm9FR4e3q/H64u1md8DMN1RiQqMuyCdgLOI738O/QOAa1MuI9w3bCBC7NZQGsfhorGxkY8//piioiKCg4OJiIhwdUjDnvwe9p2MYd8N1hj2mLz/9Kc/dfn+8OHDAPziF7844+u8vb1paWlxfq8oivP63VdffUVdXR133HEHVVVVmM1mYmNjufrqq7s9XnV1dU+hnpXw8HDKy8v79Zh98W1mMQadBs3hTDSenrSGBGPuZXwFjSUcq8ljYmACmhaV8pbBe19DbRyHg8LCQjZt2oTFYiE6OpqAgAAZwz6S38O+kzHsu4EYw+4+DJxVNUzHvsHTpk3r8bmpqals3LiRSy+9lKysLOf2ogA333wzN998MwAfffQReXl5Z0zcI11FbSvltWbOH9OKcrQB77lzzmrJfFPpDgAWR543UCGKfuBwONi5cycHDx5Ep9Mxf/58Jk6ciIeHh2y5K4Q4Kz0m7x/OsO+66y5+8pOf9HjgJUuWsG3bNq6//npUVWXVqlV89tlnmEwmVqxYce4Rj0BZ7VXmUy3FwNlVmdeY68mqOkykVygJ/tEDEJ3oL3v37uXgwYP4+/uTkZHBmDFjXB2SEGKYOuv7UFpaWigrK+vxeVqtlqeeeqrLz+Li4k553miecXfYn1OHBhWfwqOo7u54JCf1+rVbSnehopI2VpqyDHUpKSmoqkpqamq/128IIUaXHpN3enq6MymoqkpDQwO33377gAc2WjS0WMkra2aWbytKTh3ec2ah6eUfdrPdwvaKfW1NWYInD3Ck4mxZrVa+/fZboqKiiI+Px2g0MmeO9JsXQvRdj8n7j3/8o3N5T6PR4Ovr26WKXPRNdm49KjDdXgqA1/TeL5lvr8jE4rBywbj56LW9v0YuBl5VVRXr16+nsbERs9lMXFycrIwIIfpNj8n7oYce6vXtXOLsZR6vA1UlsOQYuLnhMTG5V69zqApbSndi0OqZL01ZhgxVVTlw4AC7du1CURSmTZvGrFmzJHELIfpVj8k7OTmZ1atXM3XqVNzd3Z0/l/tS+85sdXCkqJEpXibU3Bq8Zs5A28tdo9qasjQwP3wmXtKUZUiwWq2sX7+e4uJiPDw8WLx4MWPHjnV1WEKIEajH5L1//37279/f5WcajYb169cPWFCjxaH8BuwOlVlKWwGgV2pKr1+7sf32sPMjZw9EaOIc6PV67HY7Y8eOJS0tDU9P+VAlhBgY3Sbvjz/+mKuuuooNGzYMZjyjSlZO25J5SPlxNEYjnpN7txNYfmMxBY0lTApMINSz5253YuAoikJZWRljx45Fq9Vy0UUXYTAYZJlcCDGgum2i/M477wxmHKOO3aFwIK+eOEML1FTjOXlSr5fMN5XsBGDxWGnK4kpNTU18+umnfPnll86uSkajURK3EGLAyX6DLnK8pAmTxcE8Y9sf/d4umdeY69lffZhIrzDi/aIHLkBxRnl5eWzZsgWr1UpcXJw0XBFCDKpuk/fx48fJyMg45eeqqso1736QlVMHQHhlLhqDoddL5ltKd6KisliasriE3W7nu+++48iRI+j1es4//3wSExPl/4UQYlB1m7zHjx/Pm2++OZixjBqqqpKVU89YtQltdSUe06ai7VTJ351Wu5ntFZn4Gr2ZHty7ZC/6V1ZWFkeOHGHMmDFkZGTg7+/v6pCEEKNQt8nbYDAQGRk5mLGMGkUnTNQ1WVmirwDAq5e9zHe0N2VZMm6BNGUZRKqqAm13WaSkpKDVapk6dapzlzwhhBhs3RaspaamDmYco0rHknlUTR7o9XhN7bm1aVtTll0YtQbmSVOWQWM2m1m7dq1zK1y9Xk9qaqokbiGES3X7F2jlypWDGceokpVTR7CjEX31CTynTEbr4dHja7K7NGXp+fmi7yoqKli/fj0tLS04HA4mTJgg17aFEEOCTB8GWVW9mdLqVq7UnQB6X2W+qWQ7AGmRsrHFQFMUhaysLPbu3QvAzJkzSUlJkcQthBgyJHkPso69u2MbCkCrxXPq1B5fk99YTEFTKZMDEwnxlFuSBpLVauXrr7+mvLwcLy8vMjIyCAsLc3VYQgjRhSTvQZaVU4e/rQljVTkekyai8+q5heamkrZWqGnSlGXAGQwGDAYD0dHRLFq0qEs/fyGEGCokeQ+iJpONnNImLtZ2VJmn9PiamtY69lcfYax3GPF+4wc4wtHJ4XBQVFRETEwMGo2GCy64AJ1OJ8vkQoghq9tqc9H/snPrUVVIaioErRavaT0vmW8u24WKSlqkNGUZCPX19axevZq1a9dSXFwMtFWUy1gLIYYymXkPoqzcOnxtzbhXl+GRnITO2/uMz2+1m9lRkYmf0UeasvQzVVU5duwY27Ztw263k5ycTHh4uKvDEkKIXpHkPUgsNgffFzSyUO3oZd5zYxZpyjIwrFYr3377LTk5ORgMBjIyMoiLi3N1WEII0WuSvAfJ9wWN2OwKk1qLQKPBM+XMS+YOVWGzNGUZEN9//z05OTmEhISQnp6Or6+vq0MSQoizIsl7kGTl1OFtN+FVVYp7YgL6HhJGdvVh6iwNLJCmLP2ic4vTjtamEydORKuVsg8hxPAjf7kGgUNRyc6rZ5qtFOi5ylxVVTaWbEcDnC9NWfrMZDLx1VdfsX//fgC0Wi2TJ0+WxC2EGLZk5j0IckqbaGm1M9VSDBoNXinTzvj8gsYSCpvKmDxGmrL0VUlJCRs3bqS1tRWNRuPc0lYIIYYzSd6DICunDk97K77VpbjHx6LvYRvJjaVtTVkWR0pTlnOlKAq7d+9m//79aLVazjvvPKZMmSKJWwgxIkjyHmCqqpJ1vI5J5hI0qHhNTznj82ta68iuPsJY73DipCnLObHZbKxZs4bKykp8fX3JyMggODjY1WEJIUS/keQ9wEqrW6lptLLcXgLQY/LuaMqyWJqynDODwYCfnx++vr4sWLAAo9Ho6pCEEKJfSfIeYFnH6/BwmAmoKcUtJhp9YEC3z+3clCUleOIgRjn82e12cnNzSUpKAuD8889Ho9HIByAhxIgkyXuAZebUkWQqRqMqPTZm2S5NWc5JbW0t69ato76+HqPRSExMjFSSCyFGNEneA6imwUJxpYkl9vZbxM6wZN7WlGWnNGU5C6qq8v3337Njxw4cDgeTJ08mKirK1WEJIcSAk+Q9gPbn1uHusBBcV4Lb+CgMQd3f9rW/6jD1lkYWRsySpiy9YDab2bJlCwUFBbi5uXHBBRcwfrwU+AkhRgdJ3gMoK6eehJZiNMqZl8xVVWVjaVtTlkWRswcvwGEsJyeHgoICwsPDSU9Px8vLy9UhCSHEoJHkPUBaWu0cK27k/9l6XjLPbyymqKmMKWOSCPGQpizdURQFaOuQNmnSJIxGI/Hx8XJ9Wwgx6shfvQGSnVePwW4loqEY47ixGEK6v8+4oylLmjRl6VZzczNr1qxhz549QFuP8sTEREncQohRSWbeA2R/Th3xppK2JfMzzLqrW+s4UH2Ucd7hxPlJsdXpFBQUsHnzZiwWC+7u7tLiVAgx6knyHgBWm8KhggauthQDZ967e0t7U5a0sdKU5Yfsdjs7d+7k0KFD6HQ6FixYwIQJE2SchBCjniTvAXCkqAHVYmFcYwmGiHCMYaGnfZ6pU1OW6UHSlKUzu93OJ598Qk1NDQEBAWRkZBAYGOjqsIQQYkiQ5D0AsnLqiWspQas4zrhkvqN8HxaHlQujFqKTpixd6PV6wsLCCA4OZt68eej18qsqhBAd5C9iP1MUlf25dVxkblsy9+5mydyhONhctqutKUtY6mCGOGRZrVaOHj3K5MmT0Wg0zJs3T5bIhRDiNCR597O88mbMza3ENJdgCA3FEBF+2uftrz7ZlMVTmrJQWVnJ+vXraWpqwsPDg/j4eEncQgjRDUne/SzreB2xpjJ0DjteqSmnTUCqqrKhZAca4PzIOYMf5BCiqirZ2dns2rULVVVJSUkhNjbW1WEJIcSQJsm7H6mqSlZOHfNNhUD3VeZ5jcUUN7c1ZQn2GL1FWCaTiU2bNlFSUoKHhwfp6elERka6OiwhhBjyJHn3o/IaM7W1LcS1lKIPDsI49vSJaFN7U5bFY0d3U5aioiJKSkoYN24caWlpeHjI5QMhhOgNSd79KCunjhhTGXqHDa/U6addMq9ureVA9RHGeUcQ6zv6mrIoioKiKOj1epKSknB3d2f8+PFyfVsIIc6C9JbsR1k5dSS3tC2Ze6emnPY5m0t3oQJpY+eMuoTV2NjIJ598wvbt24G2FqfR0dGjbhyEEKKvZObdT+qarBSXN7LcVII+MBDjafaVHs1NWXJzc9myZQs2mw0/Pz8URZG+5EIIcY4kefeT/bl1RJvKMThs3VaZby/fh1WxcXHkolHTlMVms/Hdd99x9OhR9Ho9aWlpJCYmujosIYQY1iR595OsnDqSm7uvMncoDraUtjVlmTtKmrI4HA5Wr15NXV0dY8aMISMjA39/f1eHJYQQw54k735gstg5XljPxaZidAH+uEWPP+U5WdWHqbc2sihi9qhpyqLT6YiJiSEyMpI5c+ag042O1QYhhBhokrz7wcH8BsY2l2N0WPGaPg/ND67lqqrKRmdTltmuCXKQmM1mDh48SGpqKlqtlhkzZkhBmhBC9DNJ3v0g63jnKvNTl8w7mrJMHZNM0AhuylJeXs6GDRtoaWnBx8eHpKQkSdxCCDEAJHn3kc2u8H1eLXe0FKPz88UtNuaU52wqaWvKkjZCm7IoikJmZib79u0DYObMmSQkJLg4KiGEGLkGLHkrisITTzzB0aNHMRqNPPPMM4wff/Ja8Oeff87bb7+NTqcjMTGRJ554YljeOnS0uJHghjLcHRa8UuacsmRe3VrLgZqOpizjXBTlwGlsbGTNmjWUl5fj7e1Neno6YWFhrg5LCCFGtAHLluvWrcNqtfL+++9z//338/zzzzsfM5vN/PGPf+Sdd97hvffeo7m5mY0bNw5UKANqf049yc1FAHidpjFLR1OWxWPPG5FLyMXFxZSXlxMdHc3y5cslcQshxCAYsJn33r17WbhwIQApKSkcPHjQ+ZjRaOS9995z9rK22+24ubkNVCgDRlFV9h+v4eaWIrTe3rgnxHd53GRrZUdFJv5GX1KCJrgoyv5nt9tRFAWj0cikSZMwm82Eh4ePyA8nQggxFA1Y8m5ubsbb29v5vU6nw263o9fr0Wq1BAUFAfDuu+9iMpmYP3/+GY8XFBSEwWDo1xjDw0+/13ZvHS2sxbe6FE+HmeD5FxLxgx2xPjn8DVbFxnUTLmNs5Ng+nWuoqK6u5pNPPiEoKIjly5cDMGPGDBdHNfz19XdRyBj2BxnDvhusMRyw5O3t7U1LS4vz+47NKDp//7vf/Y78/HxeffXVHmdt1dXV/RpfeHg45eXlfTrG+p3FJLVXmWuSE7scz6E4+PzIOoxaA5O8Evp8LldTVZVjx46xbds27HY7AQEBlJWVERkZOezfm6v1x+/iaCdj2Hcyhn03EGPY3YeBAbvmnZqaypYtWwDIyso6pSXmypUrsVgsvPbaa8N2K8isY7UkNReh9fLCI7FrdXVW9WEarE2cFzYdT727iyLsH1arlQ0bNrB582a0Wi0XXHABixYtGpYFhkIIMRIM2Mx7yZIlbNu2jeuvvx5VVVm1ahWfffYZJpOJyZMn88EHHzBz5kxuueUWAG6++WaWLFkyUOH0u4raVnRlRXg7WvFKmYumU/ewrk1Z5rguyH6gKAqrV6+mvr6ekJAQMjIy8PHxcXVYQggxqg1Y8tZqtTz11FNdfhYXF+f8+siRIwN16kGRlVNPkrPKvGtjlrzGoramLEHJBHkEuCK8fqPVapkwYQKtra3MnDlTZttCCDEESJOWc7T/eC0XtRSi8fDAI6nrJYGNHU1ZIodnUxaTyURWVpazH/mUKVNcHZIQQohOJHmfg4YWK+b8AnztJrxmzkHTqRCvqrWWgzVHifIZnk1ZSkpK2LhxI62trQQEBDBhwsi5xU0IIUYKSd7nIDu3nkTn9p8pXR7bXLqzrSlL5PBqyqIoCrt372b//v1otVrOO+88kpOTXR2WEEKI05DkfQ4yj9WysLkQ3NzwnHAywZlsreysyMLfzZdpw6gpS2NjI+vXr6eqqgpfX18yMjIIDg52dVhCCCG6Icn7LJmtDuqO5+Nnb8E7dRaaTo1jvqvYh1WxcUnE+ei0w2fv6vr6eqqqqkhISGD+/PkYjUZXhySEEOIMJHmfpUP5DcQ3nLpk7lAcbCndhZvOyHnhqS6KrvdsNhsOhwN3d3eioqK46qqrZLYthBDDhNz3c5ayjte2dVUzGvGYeHJpPLP6+2HTlKW6upqPPvqIjRs3oqoqgCRuIYQYRmTmfRbsDoXyw/kstjXhNSMVbfvycuemLIsiZrs2yDNQVZVDhw6xY8cOFEVh3LhxqKo6rArrhBBCSPI+K8dLmhhfmwd0bcyS21BESXP5kG7KYjab2bx5M4WFhbi7u5OWlkZUVJSrwxJCCHEOJHmfhazjtSQ3F6LqDXhOnuj8+abStqYsi4doUxZFUfjkk09oaGggIiKCxYsX4+Xl5eqwhBBCnCNJ3r2kqiqFBwuYY2vEI2Ua2vb9xytbazhYc5TxPhHEDNGmLFqtlpSUFFpaWkhJSZEWp0IIMcxJ8u6lohMmwitzAPCZcXLJfEvpLlQgLXLukLp23NzczJ49e5g/fz4Gg4GkpCRXhySEEKKfSPLupaycOpKaC1F1ejynTAagpb0pS4CbH9OCh05TloKCAjZv3ozFYiE4OJhJkya5OiQhhBD9SJJ3L+UdyGOKtR73KVPQurfdCra9oylL5Gx0GtcvRdvtdnbu3MmhQ4fQ6XQsWLBAepMLIcQIJMm7F6rqzfgXHwPAZ2bbkrm9U1OWuWHTz/TyQVFfX8+6deuora0lICCAjIwMAgMDXR2WEEKIASDJuxeycurbqsy1Wrymtm2PmVXV1pTl/Mg5eAyBpixms5m6ujomTJjA3Llz0evlf60QQoxU8he+F45n53OxtQ7DhIloPTzamrKU7kCDxqVNWaxWKzabDS8vL8LCwrj22mvx9/d3WTxCCCEGh+sv1A5xTSYb7rmHAfCf1dazPLeh0OVNWSorK/nwww9Zt24diqK0xSeJWwghRgWZefcgO7e+rcpco8VzWtuS+UYXNmVRVZX9+/eze/duVFUlLi5u0GMQQgjhWpK8e3D0QCFplhp0CUnovLyobK3hUM0xxvtEEu07dlBjMZlMbNq0iZKSEjw9PVm8eDGRkZGDGoMQQgjXk+R9BhabAw4fBCBgzgzgZFOWxWPPG9SmLIqi8Pnnn1NfX8+4ceNIS0vDw8Nj0M4vhBBi6JDkfQbfFzSS0FiAqtHglTK1S1OWqUGDe/+0Vqtl1qxZNDU1MWXKlCHVzU0IIcTgkoK1Mzh8oIhISzWa6Dh03t58V74Xq2Lj/EFqytLY2Mg333yDxWIBICYmhqlTp0riFkKIUU5m3t1wKCrWA9kABM5Jxa442Fq2GzedkfMGoSlLTk4OW7duxWazERkZKS1OhRBCOEny7kZOaRMxdfmogPf0FDKrDg1KUxabzcZ3333H0aNH0ev1pKWlkZiYOGDnE0IIMfxI8u7GoQPFTDNXooyLRufrw6bctqYs5w9gU5ba2lrWrVtHfX09QUFBZGRk4OfnN2DnE0IIMTxJ8j4NVVVpydyPBgg6bwY5DYWUNFeQEjSBMQPYlEVRFGdB2uzZs9HpdAN2LiGEEMOXJO/TKK1uZWx1LgA+qdP5oOQrANLG9n9TFrPZjNVqxdfXl6CgIFasWIG3t3e/n0cIIcTIIdXmp3Egu4RxrZXYwsdR6+bgUO0xon0iifEd16/nKS8v58MPP+Sbb77BbrcDSOIWQgjRI5l5n0b93iy0qPjNmcGm0p2oQNrYuf12fEVR2LdvH5mZmQBMnDgRrVY+RwkhhOgdSd4/UNNgIaT8OABu0yey8/jfCXTzY2pQcr8cv7m5mQ0bNlBRUYG3tzfp6emEhYX1y7GFEEKMDpK8f+DAoVLGt1ZgCYlgpyUPm2JnUT81ZVFVlS+//JK6ujpiYmJYtGgRbm5u/RC1EEKI0USS9w/U7MoiChXfmdPZUrarX5uyaDQa5s2bR2NjI8nJydIpTQghxDmRC62dtLTa8Ss6AkBFgg+N1mbmhk3vU1OWuro61qxZg8lkAiAyMpIJEyZI4hZCCHHOZObdycHD5Yw3lWMeE8Z60/do0LAocs45HUtVVY4ePcq2bdtwOBzk5eUxefLkfo5YCCHEaCTJu5MT2/cRjErrlBhKW3JJCZrIGHf/sz6O1Wply5Yt5OXlYTQaWbx4MbGxsf0fsBBCiFFJknc7q03BM/8wANmRFlDb9uw+W9XV1axdu5ampiZCQ0NJT0/Hx8env8MVQggxiknybnfk+Amimsto8Q9kt1pCtO9Yon3HnvVxDAYDFouF6dOnM2PGDLl/WwghRL+T5N2u9LtMklAoSvAHmlkc2ftZt8lkwmw2ExgYiJ+fH9dffz3u7gO385gQQojRTZI3oCgqhmMHAdgZ1kSgmz9TetmUpbi4mE2bNmEwGFi+fDkGg0EStxBCiAElyRvIK6ghqqmEBl8vKn00XBk5p8emLA6Hg927d5OdnY1WqyUlJQW9XoZTCCHEwJNsAxR+u5cYVSFnvB43nZG5PTRlaWxsZP369VRVVeHn50dGRgZBQUGDFK0QQojRbtQnb1VV4XDbkvnhcTrmhqXiru++Zamqqqxdu5aamhoSEhKYP38+RqNxsMIVQgghJHmXlTcSWV9EvY+BWj8950fOPu3zVFVFo9Gg0WhYuHAh9fX1JCYmDnK0QgghhLRHJW/LXoyqg2NRBlJCJhJ4mqYs1dXVrF69msbGRgBCQkIkcQshhHCZUZ+87QezAcgZ50baD24PU1WVgwcPsnr1aqqqqiguLnZFiEIIIUQXo3rZvLa2mbDaAuq9tXiPj+7SlMVsNrN582YKCwtxd3cnLS2NqKgoF0YrhBBCtBnVyfv45n0EK3YOjPMkbexc58+rqqr45ptvaGlpISIigsWLF+Pl5eXCSIUQQoiTRnXybs3eB0BF3BiWd2rK4unpiaIozJo1i2nTpkmLUyGEEEPKqE3eTY0mQivzafTUMnXaIlpbTJhMJkJCQvDy8uL666/HYDC4OkwhhBDiFKN2Srl3zSbcHA7yojyJsAbx4Ycf8s0332C1WgEkcQshhBiyBix5K4rCypUrWbFiBTfddBOFhYVdHt+wYQPLly9nxYoV/Pvf/x6oMLpVtmM9Do2GltBENq3fiN1uZ8aMGZK0hRBCDHkDtmy+bt06rFYr77//PllZWTz//PO8/vrrANhsNp577jk++OADPDw8uOGGG1i8eDHBwcEDFU4XFrMVn+oT7JwQT2uTnYCAAC644AICAgIG5fxCCCFEXwzYzHvv3r0sXLgQgJSUFA4ePOh8LDc3l6ioKPz8/DAajcyYMYM9e/YMVCin2LNlEzmREZjcPZgwYQJXXXWVJG4hhBDDxoDNvJubm/H29nZ+r9PpsNvt6PV6mpub8fHxcT7m5eVFc3PzGY8XFBTUb0va0clJmLdvIW7ePC6/8sp+OeZoFh4e7uoQhj0Zw76TMew7GcO+G6wxHLDk7e3tTUtLi/N7RVGcW2b+8LGWlpYuyfx0qqur+y22yKjxzPzzXykvL6e8vLzfjjsahYeHyxj2kYxh38kY9p2MYd8NxBh292FgwJbNU1NT2bJlCwBZWVldeoHHxcVRWFhIfX09VquVPXv2MH36mbfhFEIIIUSbAZt5L1myhG3btnH99dejqiqrVq3is88+w2QysWLFCh5++GFuu+02VFVl+fLlhIaGDlQoQgghxIgyYMlbq9Xy1FNPdflZXFyc8+v09HTS09MH6vRCCCHEiDVqm7QIIYQQw5UkbyGEEGKYkeQthBBCDDOSvIUQQohhRpK3EEIIMcxI8hZCCCGGGUneQgghxDAjyVsIIYQYZiR5CyGEEMOMJG8hhBBimNGoqqq6OgghhBBC9J7MvIUQQohhRpK3EEIIMcxI8hZCCCGGGUneQgghxDAjyVsIIYQYZiR5CyGEEMPMiE/eiqKwcuVKVqxYwU033URhYWGXxzds2MDy5ctZsWIF//73v10U5dDW0xh+/vnnXHvttVx//fWsXLkSRVFcFOnQ1dMYdnj88cd58cUXBzm64aGnMczOzubGG2/khhtu4J577sFisbgo0qGtp3H89NNPueqqq1i+fDn//Oc/XRTl0Ld//35uuummU34+aDlFHeG+/vpr9aGHHlJVVVUzMzPVO++80/mY1WpVL7jgArW+vl61WCzq1VdfrVZWVroq1CHrTGPY2tqqZmRkqCaTSVVVVb3vvvvUdevWuSTOoexMY9jhX//6l3rdddepv/vd7wY7vGHhTGOoKIp6+eWXqwUFBaqqquq///1vNTc31yVxDnU9/S7Onz9fraurUy0Wi/Pvo+jqzTffVC+77DL12muv7fLzwcwpI37mvXfvXhYuXAhASkoKBw8edD6Wm5tLVFQUfn5+GI1GZsyYwZ49e1wV6pB1pjE0Go289957eHh4AGC323Fzc3NJnEPZmcYQIDMzk/3797NixQpXhDcsnGkM8/Pz8ff35+233+ZHP/oR9fX1xMbGuirUIa2n38WkpCSampqwWq2oqopGo3FFmENaVFQUr7766ik/H8ycMuKTd3NzM97e3s7vdToddrvd+ZiPj4/zMS8vL5qbmwc9xqHuTGOo1WoJCgoC4N1338VkMjF//nyXxDmUnWkMKysr+dOf/sTKlStdFd6wcKYxrKurIzMzkxtvvJG33nqLHTt2sH37dleFOqSdaRwBEhISWL58OUuXLiUtLQ1fX19XhDmkXXTRRej1+lN+Ppg5ZcQnb29vb1paWpzfK4riHPQfPtbS0tJl4EWbM41hx/e//e1v2bZtG6+++qp8Uj+NM43hV199RV1dHXfccQdvvvkmn3/+OR999JGrQh2yzjSG/v7+jB8/nvj4eAwGAwsXLjxlRinanGkcjxw5wqZNm1i/fj0bNmygtraWL7/80lWhDjuDmVNGfPJOTU1ly5YtAGRlZZGYmOh8LC4ujsLCQurr67FarezZs4fp06e7KtQh60xjCLBy5UosFguvvfaac/lcdHWmMbz55pv56KOPePfdd7njjju47LLLuPrqq10V6pB1pjEcN24cLS0tzuKrPXv2kJCQ4JI4h7ozjaOPjw/u7u64ubmh0+kIDAyksbHRVaEOO4OZU06d948wS5YsYdu2bVx//fWoqsqqVav47LPPMJlMrFixgocffpjbbrsNVVVZvnw5oaGhrg55yDnTGE6ePJkPPviAmTNncssttwBtyWjJkiUujnpo6en3UPSspzF89tlnuf/++1FVlenTp5OWlubqkIeknsZxxYoV3HjjjRgMBqKiorjqqqtcHfKQ54qcIruKCSGEEMPMiF82F0IIIUYaSd5CCCHEMCPJWwghhBhmJHkLIYQQw4wkbyGEEGKYGfG3igkxFJSUlHDxxRcTFxfX5ed//vOfCQ8PP+1rOtov3n333ed83o8++ojnn3/eeQ6z2czs2bP5zW9+c9oOUWfy8ssvM3nyZDIyMrjpppt49913Abjiiiv45JNPzjlGgJtuuomKigo8PT2Btk5V48aN48UXX3R28Dudf//733h6enLZZZf16fxCDDeSvIUYJCEhIX1OcuciPT2d559/HgCHw8H111/PBx98wPXXX39Wx7n33nudX+/atcv5dX+9p2eeeYY5c+YAbV2/7rnnHt566y0eeOCBbl+zb98+Zs+e3S/nF2I4keQthIsdO3aMp59+GpPJRG1tLXfccQc33HCD83GbzcZ///d/c/z4cQBuvPFGrrvuOqqrq1m5ciUVFRVoNBruv/9+5s2bd8Zz6XQ6Zs6c6TzWhx9+yFtvvYVGo2HSpEk8/vjjGI3G057v4YcfZvbs2Xz//fcAXHvttfznP/8hKSmJQ4cOkZaWxurVqwkKCqK+vp7LLruMjRs3sn37dl555RXsdjtjx47l6aefJiAg4Ixxmkwm6urqmDp1KgBffvklb731FmazGavVyqpVqzCbzWzYsIEdO3YQHBzMhAkTzno8hBiu5Jq3EIOksrKSK664wvnf//7v/wLwn//8h5///Od8+OGHvPPOO7zwwgtdXpeZmUlDQwOrV6/mjTfecO5S9Oyzz7J8+XI++ugjXn/9dVauXNnjJgh1dXV8++23pKSkcPToUf785z/z7rvv8tlnn+Hh4cGf/vSnbs/X4bHHHnPG3UGv13PxxRfz1VdfAfDNN9+wZMkSmpqaeOmll/jrX//K6tWrWbBgQbf7lT/22GNcfvnlLFiwgBUrVjBv3jxuvfVWFEXhvffe489//jOffvopt99+O2+++Sbz5s0jPT2de+65h4ULF57TeAgxXMnMW4hB0t2y+cMPP8zWrVt54403OHbsGCaTqcvjCQkJ5Ofnc9ttt7Fo0SIefPBBAL777jvy8vJ45ZVXgLbtWIuLi5kwYUKX12/YsIErrrgCVVVRVZUlS5Zw2WWX8X//938sXrzYOQtesWIFjzzyCHfcccdpz9eTyy+/nOeee44f/ehHfP7559x3333s37+f8vJybr75ZqBtOdzPz++0r+9YNt+3bx/33HMPS5YswWg0AvA///M/bNiwgfz8fHbt2oVWe+q8o7fjIcRIIMlbCBf75S9/ia+vL4sXL+bSSy/l888/7/J4QEAAa9asYdu2bWzevJmrrrqKNWvWoCgKb7/9Nv7+/kDbzH7MmDGnHL/zNe/OFEXp8r2qqtjt9m7P15OpU6fS0NBAdnY2J06cYPr06axbt47U1FT+/Oc/A2CxWLrsunQ6qamp3HTTTdx///18/PHHWCwWrrnmGi6//HJmzZpFUlIS//d//3fa99Ob8RBiJJBlcyFcbNu2bdxzzz1ccMEFzt2eHA6H8/H169fzwAMPkJaWxmOPPYanpyfl5eWcd955/POf/wQgJyeHZcuW0dra2uvzzp49mw0bNlBfXw+0VW7PmTOn2/N19sM9oDssW7aM3/zmNyxduhSAadOmkZWVRX5+PgCvvfbaKZcFTufHP/4xLS0tvP/++xQUFKDRaLjzzjuZM2cOa9eudY6PTqdzft3X8RBiOJGZtxAudvfdd3PjjTfi5uZGcnIykZGRlJSUOB9ftGgR33zzDUuXLsXNzY3LL7+cpKQkHnvsMVauXMmyZcsAeOGFF/D29u71eZOTk/nZz37GTTfdhM1mY9KkSTz55JO4ubmd9nydZWRkcMUVV5yy7/jll1/Oyy+/zB/+8AcAgoODWbVqFb/85S9RFIXQ0FB+97vf9Rib0Wjkl7/8JatWrWLt2rVMmDCBSy65BI1Gw4IFC9i7dy8A8+bN4/e//z0+Pj59Hg8hhhPZVUwIIYQYZmTZXAghhBhmJHkLIYQQw4wkbyGEEGKYkeQthBBCDDOSvIUQQohhRpK3EEIIMcxI8hZCCCGGGUneQgghxDDz/wHUX6mCrVPX5wAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<Figure size 576x432 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "import matplotlib.pyplot as plt\n",
    "from sklearn.metrics import roc_curve\n",
    "\n",
    "def plot_roc_curves(models, X_test, y_test):\n",
    "    plt.figure(figsize=(8, 6))\n",
    "    for model_name, y_pred_proba in models.items():\n",
    "        # Calculating true positive rate (TPR) and false positive rate (FPR)\n",
    "        fpr, tpr, _ = roc_curve(y_test, y_pred_proba)\n",
    "        \n",
    "        # Plotting ROC curve for each model\n",
    "        plt.plot(fpr, tpr, label=model_name)\n",
    "\n",
    "    # Plotting the random guess line\n",
    "    plt.plot([0, 1], [0, 1], linestyle='--', color='grey', label='Random Guess')\n",
    "\n",
    "    plt.xlabel('False Positive Rate')\n",
    "    plt.ylabel('True Positive Rate')\n",
    "    plt.title('ROC Curves Comparison')\n",
    "    plt.legend()\n",
    "    plt.grid(True)\n",
    "    plt.show()\n",
    "\n",
    "# Define the models and their corresponding predicted probabilities\n",
    "models = {'XGBClassifier': xgb_test_preds,\n",
    "          'Random Forest': rfc_test_preds,\n",
    "          'Linear SVC': lsvc_test_preds}\n",
    "\n",
    "# Plot ROC curves for the models\n",
    "plot_roc_curves(models, X_test, y_test)\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Conclusion and Recommendation\n",
    "Groundwater is very important to maintain the functionality of the wells.Almost, the entire water supply to the wells is dependent on grounwater. Hence we would look into different methods such as rainwater harvesting and soil conservation which would also help sustain more water in the lakes.\n",
    "\n",
    "We see that when we make payment based on bucket or monthly payment, the wells are maintained better. So payment is an important factor.\n",
    "\n",
    "Definitely, age of the wells is an important factor to predict the functionality of the wells. The newer the wells are, more water they have or better is their condition.\n",
    "\n",
    "We can see that having a public meeting helps in functioning of the wells. More than 50% wells are functional when there is a public meeting held for the same. Thus, Public meeting is an important factor for the functioning of wells.\n",
    "\n",
    "Further send out designated people to inspect the pumps detected by the model and assess what needs to be done.\n",
    "\n",
    "Future Work\n",
    "Need to examine the missing values or Zero values of Population and gps. Also, try binning the highly cardinal columns. Most static head values were 0. Need to examine those too."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.8.5"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}