RegressionPredict.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "id": "c01e9bc0",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-06-11T09:24:53.643384Z",
     "start_time": "2021-06-11T09:24:53.622385Z"
    }
   },
   "source": [
    "# Regression Predict Student Solution\n",
    "\n",
    "© Explore Data Science Academy\n",
    "\n",
    "---\n",
    "### Honour Code\n",
    "\n",
    "I, **TEAM JM3**, confirm - by submitting this document - that the solutions in this notebook are a result of my own work and that I abide by the [EDSA honour code](https://drive.google.com/file/d/1QDCjGZJ8-FmJE3bZdIQNwnJyQKPhHZBn/view?usp=sharing).\n",
    "\n",
    "Non-compliance with the honour code constitutes a material breach of contract.\n",
    "\n",
    "### Predict Overview: Spain Electricity Shortfall Challenge\n",
    "\n",
    "The government of Spain is considering an expansion of it's renewable energy resource infrastructure investments. As such, they require information on the trends and patterns of the countries renewable sources and fossil fuel energy generation. Your company has been awarded the contract to:\n",
    "\n",
    "- 1. analyse the supplied data;\n",
    "- 2. identify potential errors in the data and clean the existing data set;\n",
    "- 3. determine if additional features can be added to enrich the data set;\n",
    "- 4. build a model that is capable of forecasting the three hourly demand shortfalls;\n",
    "- 5. evaluate the accuracy of the best machine learning model;\n",
    "- 6. determine what features were most important in the model’s prediction decision, and\n",
    "- 7. explain the inner working of the model to a non-technical audience.\n",
    "\n",
    "Formally the problem statement was given to you, the senior data scientist, by your manager via email reads as follow:\n",
    "\n",
    "> In this project you are tasked to model the shortfall between the energy generated by means of fossil fuels and various renewable sources - for the country of Spain. The daily shortfall, which will be referred to as the target variable, will be modelled as a function of various city-specific weather features such as `pressure`, `wind speed`, `humidity`, etc. As with all data science projects, the provided features are rarely adequate predictors of the target variable. As such, you are required to perform feature engineering to ensure that you will be able to accurately model Spain's three hourly shortfalls.\n",
    " \n",
    "On top of this, she has provided you with a starter notebook containing vague explanations of what the main outcomes are. "
   ]
  },
  {
   "cell_type": "markdown",
   "id": "dad0cdf8",
   "metadata": {},
   "source": [
    "<a id=\"cont\"></a>\n",
    "\n",
    "## Table of Contents\n",
    "\n",
    "<a href=#intro>Introduction</a>\n",
    "\n",
    "<a href=#one>1. Importing Packages</a>\n",
    "\n",
    "<a href=#two>2. Exploratory Data Analysis (EDA)</a>\n",
    "    \n",
    "&nbsp;&nbsp;&nbsp;&nbsp;<a href=#twoone>2.1 Data Collection</a>\n",
    "\n",
    "&nbsp;&nbsp;&nbsp;&nbsp;<a href=#twotwo>2.2. Data Cleaning</a>\n",
    "\n",
    "&nbsp;&nbsp;&nbsp;&nbsp;<a href=#twothree>2.3. Univariate Analysis</a>\n",
    "\n",
    "&nbsp;&nbsp;&nbsp;&nbsp;<a href=#twofour>2.4. Multivariate Analysis</a>\n",
    "\n",
    "\n",
    "<a href=#five>3. Modeling</a>\n",
    "\n",
    "<a href=#six>4. Model Performance</a>\n",
    "\n",
    "<a href=#seven>Conclusion</a>\n",
    "\n",
    "<a href=#app>Appendix</a>"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "d2350a7d",
   "metadata": {},
   "source": [
    "<a id=\"intro\"></a>\n",
    "## Introduction \n",
    "<a href=#cont>Back to Table of Contents</a>\n",
    "\n",
    "Continuous energy supply is fundamental to fulfilling basic social needs, driving economic growth and fueling human development. This is because energy services have an effect on productivity, health, education, safe water and communication services. \n",
    "\n",
    "In recent years, there has been a global concern and sufficient evidence that the use of purely non-renewable sources of energy is not sustainable, thus the urgent need to consider other options; renewable energy sources such as Solar, Wind, Geothermal and Hydro among others.\n",
    "\n",
    "This project by Team JM3 Digital Solutions Limited, therefore provides the Spanish government with the requisite information on the trends and patterns of the country's renewable sources and fossil fuel energy generation by modeling the shortfall between the two energy sources. \n",
    "\n",
    "This will guide the government of Spain in the expansion of its renewable energy resource infrastructure investments."
   ]
  },
  {
   "cell_type": "markdown",
   "id": "f4b30a53",
   "metadata": {},
   "source": [
    " <a id=\"one\"></a>\n",
    "## 1. Importing Packages\n",
    "<a href=#cont>Back to Table of Contents</a>\n",
    "\n",
    "The following packages will be improted for use within out implementation:\n",
    "1. [Pandas](https://pandas.pydata.org/) - an open source data analysis and manipulation tool, built on top of the Python programming language.\n",
    "2. [NumPy](https://numpy.org/) - a tool that offers arrays comprehensive mathematical functions, random number generators, linear algebra routines, Fourier transforms, and more.\n",
    "3. [Seaborn](https://seaborn.pydata.org/) -a Python data visualization library based on matplotlib.\n",
    "4. [sklearn](https://scikit-learn.org/) - set of tools for predictive data analysis built on NumPy, SciPy, and matplotlib.\n",
    "5. [statsmodels](https://www.statsmodels.org/stable/index.html) - provides classes and functions for the estimation of many different statistical models, as well as for conducting statistical tests, and statistical data exploration."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "831f4727",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-06-23T10:30:53.800892Z",
     "start_time": "2021-06-23T10:30:50.215449Z"
    }
   },
   "outputs": [],
   "source": [
    "# Libraries for data loading, data manipulation and data visulisation\n",
    "import pandas as pd\n",
    "import numpy as np\n",
    "import matplotlib.pyplot as plt\n",
    "import seaborn as sns\n",
    "from statsmodels.graphics.correlation import plot_corr\n",
    "\n",
    "# Libraries for data preparation and model building\n",
    "from sklearn.impute import KNNImputer\n",
    "from sklearn.preprocessing import StandardScaler\n",
    "# Import train/test splitting function from sklearn to split the data into training and testing data\n",
    "from sklearn.model_selection import train_test_split\n",
    "# Import the ridge regression module from sklearn\n",
    "from sklearn.linear_model import Ridge\n",
    "#Importing the linear model from Sklearn\n",
    "from sklearn.linear_model import LinearRegression\n",
    "# Import metrics module\n",
    "from sklearn import metrics\n",
    "# Setting global constants to ensure notebook results are reproducible\n",
    "#PARAMETER_CONSTANT = ###"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "f2dc5aab",
   "metadata": {},
   "source": [
    "<a id=\"two\"></a>\n",
    "## 2. Exploratory Data Analysis (EDA)\n",
    "<a class=\"anchor\" id=\"2\"></a>\n",
    "<a href=#cont>Back to Table of Contents</a>\n",
    "\n",
    "\"Exploratory Data Analysis refers to the critical process of performing initial investigations on data ... with the help of summary statistics and graphical representations.\"<sup>2\n",
    "    \n",
    "EDA unearths a lot of important information which can easily be missed, especially with large sets of data. This information is known to help the analysis process in the long run, from revealing insights on how to frame questions to displaying the results. \n",
    "\n",
    "Carrying out a thorough EDA will often lead to:\n",
    "\n",
    "<ul>\n",
    "<li>Identifying erroneous and missing data</li>\n",
    "<li>Understanding the basic structure of the data as well as mapping it out</li>\n",
    "<li>Spotting the most important variables in the dataset</li>\n",
    "<li>Assessing a hypothesis related to a specific model</li>\n",
    "<li>Establishing a model that can explain the data using minimum number of variables)</li>\n",
    "<li>Approximating parameters and computing the margins of error.</li>\n",
    "</ul>\n",
    "    \n",
    "The following are the steps involved in EDA:\n",
    "<ol type=\"1\">\n",
    "    <li> <b>Data Collection</b> - The data used here has been obtained from the government of Spain, a reliable source.</li> \n",
    "<li> <b>Data Cleaning</b> - this is the removal of anomalies, e.g. unwanted values or variables, from the dataset which can skew the data and thus negatively affect the results.</li> \n",
    "<li><b>Univariate Analysis</b> - this is the process of analyzing data of just one variable (a single feature/column). We will implement both graphical or non-graphical techniques.</li> \n",
    "<li><b>Multivariate Analysis</b> - This is the process of comparing two or more variables. We will implement both graphical or non-graphical techniques here as well.</li> \n",
    "</ol>\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "4cea7ef4",
   "metadata": {},
   "source": [
    "<a id=\"twoone\"></a>\n",
    "### 2.1. Data Collection\n",
    "<a class=\"anchor\" id=\"2.1\"></a>\n",
    "<a href=#cont>Back to Table of Contents</a>\n",
    "\n",
    "\n",
    "The government of Spain has been tracking the different energy sources available within the country out of which, they have provided two data sets in the form of comma-separated values (CSV) files. \n",
    "\n",
    "The first, the train dataset, will eventually be used to train the model(s) while the second, the test dataset, will be used to test the performance of the model(s).\n",
    "\n",
    "The two datasets will be loaded into Python DataFrames using the pandas package."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "4f51a4be",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-06-28T08:49:35.311495Z",
     "start_time": "2021-06-28T08:49:35.295494Z"
    }
   },
   "outputs": [],
   "source": [
    "# loading the data\n",
    "df_train = pd.read_csv('df_train.csv')\n",
    "df_test = pd.read_csv('df_test.csv', index_col=0)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "1534b746",
   "metadata": {},
   "source": [
    "<a id=\"twotwo\"></a>\n",
    "### 2.2. Data Cleaning\n",
    "<a class=\"anchor\" id=\"2.2\"></a>\n",
    "<a href=#cont>Back to Table of Contents</a>\n",
    "\n",
    "Data cleaning will start by checking for, and deleting, any columns that are purely made up of unique values. This is because they are not strongly predictive when it comes to supervised machine learning.\n",
    "\n",
    "**NOTE:** We will work on cleaning the train dataset first, then replicate all the steps followed to clean the test dataset."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "1a81bb36",
   "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>Unnamed: 0</th>\n",
       "      <th>time</th>\n",
       "      <th>Madrid_wind_speed</th>\n",
       "      <th>Valencia_wind_deg</th>\n",
       "      <th>Bilbao_rain_1h</th>\n",
       "      <th>Valencia_wind_speed</th>\n",
       "      <th>Seville_humidity</th>\n",
       "      <th>Madrid_humidity</th>\n",
       "      <th>Bilbao_clouds_all</th>\n",
       "      <th>Bilbao_wind_speed</th>\n",
       "      <th>...</th>\n",
       "      <th>Madrid_temp_max</th>\n",
       "      <th>Barcelona_temp</th>\n",
       "      <th>Bilbao_temp_min</th>\n",
       "      <th>Bilbao_temp</th>\n",
       "      <th>Barcelona_temp_min</th>\n",
       "      <th>Bilbao_temp_max</th>\n",
       "      <th>Seville_temp_min</th>\n",
       "      <th>Madrid_temp</th>\n",
       "      <th>Madrid_temp_min</th>\n",
       "      <th>load_shortfall_3h</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0</td>\n",
       "      <td>2015-01-01 03:00:00</td>\n",
       "      <td>0.666667</td>\n",
       "      <td>level_5</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.666667</td>\n",
       "      <td>74.333333</td>\n",
       "      <td>64.000000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>...</td>\n",
       "      <td>265.938000</td>\n",
       "      <td>281.013000</td>\n",
       "      <td>269.338615</td>\n",
       "      <td>269.338615</td>\n",
       "      <td>281.013000</td>\n",
       "      <td>269.338615</td>\n",
       "      <td>274.254667</td>\n",
       "      <td>265.938000</td>\n",
       "      <td>265.938000</td>\n",
       "      <td>6715.666667</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1</td>\n",
       "      <td>2015-01-01 06:00:00</td>\n",
       "      <td>0.333333</td>\n",
       "      <td>level_10</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.666667</td>\n",
       "      <td>78.333333</td>\n",
       "      <td>64.666667</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>...</td>\n",
       "      <td>266.386667</td>\n",
       "      <td>280.561667</td>\n",
       "      <td>270.376000</td>\n",
       "      <td>270.376000</td>\n",
       "      <td>280.561667</td>\n",
       "      <td>270.376000</td>\n",
       "      <td>274.945000</td>\n",
       "      <td>266.386667</td>\n",
       "      <td>266.386667</td>\n",
       "      <td>4171.666667</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>2</td>\n",
       "      <td>2015-01-01 09:00:00</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>level_9</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>71.333333</td>\n",
       "      <td>64.333333</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>...</td>\n",
       "      <td>272.708667</td>\n",
       "      <td>281.583667</td>\n",
       "      <td>275.027229</td>\n",
       "      <td>275.027229</td>\n",
       "      <td>281.583667</td>\n",
       "      <td>275.027229</td>\n",
       "      <td>278.792000</td>\n",
       "      <td>272.708667</td>\n",
       "      <td>272.708667</td>\n",
       "      <td>4274.666667</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>3</td>\n",
       "      <td>2015-01-01 12:00:00</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>level_8</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>65.333333</td>\n",
       "      <td>56.333333</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>...</td>\n",
       "      <td>281.895219</td>\n",
       "      <td>283.434104</td>\n",
       "      <td>281.135063</td>\n",
       "      <td>281.135063</td>\n",
       "      <td>283.434104</td>\n",
       "      <td>281.135063</td>\n",
       "      <td>285.394000</td>\n",
       "      <td>281.895219</td>\n",
       "      <td>281.895219</td>\n",
       "      <td>5075.666667</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>4</td>\n",
       "      <td>2015-01-01 15:00:00</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>level_7</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>59.000000</td>\n",
       "      <td>57.000000</td>\n",
       "      <td>2.0</td>\n",
       "      <td>0.333333</td>\n",
       "      <td>...</td>\n",
       "      <td>280.678437</td>\n",
       "      <td>284.213167</td>\n",
       "      <td>282.252063</td>\n",
       "      <td>282.252063</td>\n",
       "      <td>284.213167</td>\n",
       "      <td>282.252063</td>\n",
       "      <td>285.513719</td>\n",
       "      <td>280.678437</td>\n",
       "      <td>280.678437</td>\n",
       "      <td>6620.666667</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 49 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "   Unnamed: 0                 time  Madrid_wind_speed Valencia_wind_deg  \\\n",
       "0           0  2015-01-01 03:00:00           0.666667           level_5   \n",
       "1           1  2015-01-01 06:00:00           0.333333          level_10   \n",
       "2           2  2015-01-01 09:00:00           1.000000           level_9   \n",
       "3           3  2015-01-01 12:00:00           1.000000           level_8   \n",
       "4           4  2015-01-01 15:00:00           1.000000           level_7   \n",
       "\n",
       "   Bilbao_rain_1h  Valencia_wind_speed  Seville_humidity  Madrid_humidity  \\\n",
       "0             0.0             0.666667         74.333333        64.000000   \n",
       "1             0.0             1.666667         78.333333        64.666667   \n",
       "2             0.0             1.000000         71.333333        64.333333   \n",
       "3             0.0             1.000000         65.333333        56.333333   \n",
       "4             0.0             1.000000         59.000000        57.000000   \n",
       "\n",
       "   Bilbao_clouds_all  Bilbao_wind_speed  ...  Madrid_temp_max  Barcelona_temp  \\\n",
       "0                0.0           1.000000  ...       265.938000      281.013000   \n",
       "1                0.0           1.000000  ...       266.386667      280.561667   \n",
       "2                0.0           1.000000  ...       272.708667      281.583667   \n",
       "3                0.0           1.000000  ...       281.895219      283.434104   \n",
       "4                2.0           0.333333  ...       280.678437      284.213167   \n",
       "\n",
       "   Bilbao_temp_min  Bilbao_temp  Barcelona_temp_min  Bilbao_temp_max  \\\n",
       "0       269.338615   269.338615          281.013000       269.338615   \n",
       "1       270.376000   270.376000          280.561667       270.376000   \n",
       "2       275.027229   275.027229          281.583667       275.027229   \n",
       "3       281.135063   281.135063          283.434104       281.135063   \n",
       "4       282.252063   282.252063          284.213167       282.252063   \n",
       "\n",
       "   Seville_temp_min Madrid_temp  Madrid_temp_min  load_shortfall_3h  \n",
       "0        274.254667  265.938000       265.938000        6715.666667  \n",
       "1        274.945000  266.386667       266.386667        4171.666667  \n",
       "2        278.792000  272.708667       272.708667        4274.666667  \n",
       "3        285.394000  281.895219       281.895219        5075.666667  \n",
       "4        285.513719  280.678437       280.678437        6620.666667  \n",
       "\n",
       "[5 rows x 49 columns]"
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_train.head() # Checking the content of the uncleaned train data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "080d448c",
   "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>time</th>\n",
       "      <th>Madrid_wind_speed</th>\n",
       "      <th>Valencia_wind_deg</th>\n",
       "      <th>Bilbao_rain_1h</th>\n",
       "      <th>Valencia_wind_speed</th>\n",
       "      <th>Seville_humidity</th>\n",
       "      <th>Madrid_humidity</th>\n",
       "      <th>Bilbao_clouds_all</th>\n",
       "      <th>Bilbao_wind_speed</th>\n",
       "      <th>Seville_clouds_all</th>\n",
       "      <th>...</th>\n",
       "      <th>Barcelona_temp_max</th>\n",
       "      <th>Madrid_temp_max</th>\n",
       "      <th>Barcelona_temp</th>\n",
       "      <th>Bilbao_temp_min</th>\n",
       "      <th>Bilbao_temp</th>\n",
       "      <th>Barcelona_temp_min</th>\n",
       "      <th>Bilbao_temp_max</th>\n",
       "      <th>Seville_temp_min</th>\n",
       "      <th>Madrid_temp</th>\n",
       "      <th>Madrid_temp_min</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>8763</th>\n",
       "      <td>2018-01-01 00:00:00</td>\n",
       "      <td>5.000000</td>\n",
       "      <td>level_8</td>\n",
       "      <td>0.0</td>\n",
       "      <td>5.000000</td>\n",
       "      <td>87.000000</td>\n",
       "      <td>71.333333</td>\n",
       "      <td>20.000000</td>\n",
       "      <td>3.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>...</td>\n",
       "      <td>287.816667</td>\n",
       "      <td>280.816667</td>\n",
       "      <td>287.356667</td>\n",
       "      <td>276.150000</td>\n",
       "      <td>280.380000</td>\n",
       "      <td>286.816667</td>\n",
       "      <td>285.150000</td>\n",
       "      <td>283.150000</td>\n",
       "      <td>279.866667</td>\n",
       "      <td>279.150000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8764</th>\n",
       "      <td>2018-01-01 03:00:00</td>\n",
       "      <td>4.666667</td>\n",
       "      <td>level_8</td>\n",
       "      <td>0.0</td>\n",
       "      <td>5.333333</td>\n",
       "      <td>89.000000</td>\n",
       "      <td>78.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>3.666667</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>...</td>\n",
       "      <td>284.816667</td>\n",
       "      <td>280.483333</td>\n",
       "      <td>284.190000</td>\n",
       "      <td>277.816667</td>\n",
       "      <td>281.010000</td>\n",
       "      <td>283.483333</td>\n",
       "      <td>284.150000</td>\n",
       "      <td>281.150000</td>\n",
       "      <td>279.193333</td>\n",
       "      <td>278.150000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8765</th>\n",
       "      <td>2018-01-01 06:00:00</td>\n",
       "      <td>2.333333</td>\n",
       "      <td>level_7</td>\n",
       "      <td>0.0</td>\n",
       "      <td>5.000000</td>\n",
       "      <td>89.000000</td>\n",
       "      <td>89.666667</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>2.333333</td>\n",
       "      <td>6.666667</td>\n",
       "      <td>...</td>\n",
       "      <td>284.483333</td>\n",
       "      <td>276.483333</td>\n",
       "      <td>283.150000</td>\n",
       "      <td>276.816667</td>\n",
       "      <td>279.196667</td>\n",
       "      <td>281.816667</td>\n",
       "      <td>282.150000</td>\n",
       "      <td>280.483333</td>\n",
       "      <td>276.340000</td>\n",
       "      <td>276.150000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8766</th>\n",
       "      <td>2018-01-01 09:00:00</td>\n",
       "      <td>2.666667</td>\n",
       "      <td>level_7</td>\n",
       "      <td>0.0</td>\n",
       "      <td>5.333333</td>\n",
       "      <td>93.333333</td>\n",
       "      <td>82.666667</td>\n",
       "      <td>26.666667</td>\n",
       "      <td>5.666667</td>\n",
       "      <td>6.666667</td>\n",
       "      <td>...</td>\n",
       "      <td>284.150000</td>\n",
       "      <td>277.150000</td>\n",
       "      <td>283.190000</td>\n",
       "      <td>279.150000</td>\n",
       "      <td>281.740000</td>\n",
       "      <td>282.150000</td>\n",
       "      <td>284.483333</td>\n",
       "      <td>279.150000</td>\n",
       "      <td>275.953333</td>\n",
       "      <td>274.483333</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8767</th>\n",
       "      <td>2018-01-01 12:00:00</td>\n",
       "      <td>4.000000</td>\n",
       "      <td>level_7</td>\n",
       "      <td>0.0</td>\n",
       "      <td>8.666667</td>\n",
       "      <td>65.333333</td>\n",
       "      <td>64.000000</td>\n",
       "      <td>26.666667</td>\n",
       "      <td>10.666667</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>...</td>\n",
       "      <td>287.483333</td>\n",
       "      <td>281.150000</td>\n",
       "      <td>286.816667</td>\n",
       "      <td>281.816667</td>\n",
       "      <td>284.116667</td>\n",
       "      <td>286.150000</td>\n",
       "      <td>286.816667</td>\n",
       "      <td>284.483333</td>\n",
       "      <td>280.686667</td>\n",
       "      <td>280.150000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 47 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "                     time  Madrid_wind_speed Valencia_wind_deg  \\\n",
       "8763  2018-01-01 00:00:00           5.000000           level_8   \n",
       "8764  2018-01-01 03:00:00           4.666667           level_8   \n",
       "8765  2018-01-01 06:00:00           2.333333           level_7   \n",
       "8766  2018-01-01 09:00:00           2.666667           level_7   \n",
       "8767  2018-01-01 12:00:00           4.000000           level_7   \n",
       "\n",
       "      Bilbao_rain_1h  Valencia_wind_speed  Seville_humidity  Madrid_humidity  \\\n",
       "8763             0.0             5.000000         87.000000        71.333333   \n",
       "8764             0.0             5.333333         89.000000        78.000000   \n",
       "8765             0.0             5.000000         89.000000        89.666667   \n",
       "8766             0.0             5.333333         93.333333        82.666667   \n",
       "8767             0.0             8.666667         65.333333        64.000000   \n",
       "\n",
       "      Bilbao_clouds_all  Bilbao_wind_speed  Seville_clouds_all  ...  \\\n",
       "8763          20.000000           3.000000            0.000000  ...   \n",
       "8764           0.000000           3.666667            0.000000  ...   \n",
       "8765           0.000000           2.333333            6.666667  ...   \n",
       "8766          26.666667           5.666667            6.666667  ...   \n",
       "8767          26.666667          10.666667            0.000000  ...   \n",
       "\n",
       "      Barcelona_temp_max  Madrid_temp_max  Barcelona_temp  Bilbao_temp_min  \\\n",
       "8763          287.816667       280.816667      287.356667       276.150000   \n",
       "8764          284.816667       280.483333      284.190000       277.816667   \n",
       "8765          284.483333       276.483333      283.150000       276.816667   \n",
       "8766          284.150000       277.150000      283.190000       279.150000   \n",
       "8767          287.483333       281.150000      286.816667       281.816667   \n",
       "\n",
       "      Bilbao_temp  Barcelona_temp_min Bilbao_temp_max  Seville_temp_min  \\\n",
       "8763   280.380000          286.816667      285.150000        283.150000   \n",
       "8764   281.010000          283.483333      284.150000        281.150000   \n",
       "8765   279.196667          281.816667      282.150000        280.483333   \n",
       "8766   281.740000          282.150000      284.483333        279.150000   \n",
       "8767   284.116667          286.150000      286.816667        284.483333   \n",
       "\n",
       "      Madrid_temp  Madrid_temp_min  \n",
       "8763   279.866667       279.150000  \n",
       "8764   279.193333       278.150000  \n",
       "8765   276.340000       276.150000  \n",
       "8766   275.953333       274.483333  \n",
       "8767   280.686667       280.150000  \n",
       "\n",
       "[5 rows x 47 columns]"
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_test.head() # checking test data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "dc7630b3",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Unnamed: 0\n",
      "time\n"
     ]
    }
   ],
   "source": [
    "# interate through all train dataset columns \n",
    "# to see if the number of rows of unique values\n",
    "# is equal to the number of rows of the dataset\n",
    "# if true, the column is purely made of unqiue values\n",
    "for (columnName, columnData) in df_train.iteritems():\n",
    "    if len(df_train[columnName].value_counts()) == len(df_train):\n",
    "        # print name of column with uniue values\n",
    "        print(columnName)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "d315ac85",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "time\n"
     ]
    }
   ],
   "source": [
    "# interate through all test dataset columns \n",
    "# to see if the number of rows of unique values\n",
    "# is equal to the number of rows of the dataset\n",
    "# if true, the column is purely made of unqiue values\n",
    "for (columnName, columnData) in df_test.iteritems():\n",
    "    if len(df_test[columnName].value_counts()) == len(df_test):\n",
    "        # print name of column with uniue values\n",
    "        print(columnName)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "0d2fe4b4",
   "metadata": {},
   "source": [
    "On the train dataset:\n",
    "\n",
    "1. The **Unnamed: 0** column looks like the index column and will therefore be deleted (dropped).\n",
    "\n",
    "2. The **time** column, on the other hand has relevant data. A quick glance will reveal that the way it is captured, year-week-day-hour-minute-second, inevitably makes the column unique so it will not be deleted.\n",
    "\n",
    "On the test dataset, there is an indexing column that will be reset. The other unique column is time which, as explained above, is of value and will not be deleted."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "id": "786197a7",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Drop the Unnamed Column\n",
    "df_train = df_train.drop('Unnamed: 0', axis=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "id": "6d8e8e2c",
   "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>time</th>\n",
       "      <th>Madrid_wind_speed</th>\n",
       "      <th>Valencia_wind_deg</th>\n",
       "      <th>Bilbao_rain_1h</th>\n",
       "      <th>Valencia_wind_speed</th>\n",
       "      <th>Seville_humidity</th>\n",
       "      <th>Madrid_humidity</th>\n",
       "      <th>Bilbao_clouds_all</th>\n",
       "      <th>Bilbao_wind_speed</th>\n",
       "      <th>Seville_clouds_all</th>\n",
       "      <th>...</th>\n",
       "      <th>Madrid_temp_max</th>\n",
       "      <th>Barcelona_temp</th>\n",
       "      <th>Bilbao_temp_min</th>\n",
       "      <th>Bilbao_temp</th>\n",
       "      <th>Barcelona_temp_min</th>\n",
       "      <th>Bilbao_temp_max</th>\n",
       "      <th>Seville_temp_min</th>\n",
       "      <th>Madrid_temp</th>\n",
       "      <th>Madrid_temp_min</th>\n",
       "      <th>load_shortfall_3h</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>2015-01-01 03:00:00</td>\n",
       "      <td>0.666667</td>\n",
       "      <td>level_5</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.666667</td>\n",
       "      <td>74.333333</td>\n",
       "      <td>64.000000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>265.938000</td>\n",
       "      <td>281.013000</td>\n",
       "      <td>269.338615</td>\n",
       "      <td>269.338615</td>\n",
       "      <td>281.013000</td>\n",
       "      <td>269.338615</td>\n",
       "      <td>274.254667</td>\n",
       "      <td>265.938000</td>\n",
       "      <td>265.938000</td>\n",
       "      <td>6715.666667</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>2015-01-01 06:00:00</td>\n",
       "      <td>0.333333</td>\n",
       "      <td>level_10</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.666667</td>\n",
       "      <td>78.333333</td>\n",
       "      <td>64.666667</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>266.386667</td>\n",
       "      <td>280.561667</td>\n",
       "      <td>270.376000</td>\n",
       "      <td>270.376000</td>\n",
       "      <td>280.561667</td>\n",
       "      <td>270.376000</td>\n",
       "      <td>274.945000</td>\n",
       "      <td>266.386667</td>\n",
       "      <td>266.386667</td>\n",
       "      <td>4171.666667</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>2 rows × 48 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "                  time  Madrid_wind_speed Valencia_wind_deg  Bilbao_rain_1h  \\\n",
       "0  2015-01-01 03:00:00           0.666667           level_5             0.0   \n",
       "1  2015-01-01 06:00:00           0.333333          level_10             0.0   \n",
       "\n",
       "   Valencia_wind_speed  Seville_humidity  Madrid_humidity  Bilbao_clouds_all  \\\n",
       "0             0.666667         74.333333        64.000000                0.0   \n",
       "1             1.666667         78.333333        64.666667                0.0   \n",
       "\n",
       "   Bilbao_wind_speed  Seville_clouds_all  ...  Madrid_temp_max  \\\n",
       "0                1.0                 0.0  ...       265.938000   \n",
       "1                1.0                 0.0  ...       266.386667   \n",
       "\n",
       "   Barcelona_temp  Bilbao_temp_min  Bilbao_temp  Barcelona_temp_min  \\\n",
       "0      281.013000       269.338615   269.338615          281.013000   \n",
       "1      280.561667       270.376000   270.376000          280.561667   \n",
       "\n",
       "   Bilbao_temp_max Seville_temp_min  Madrid_temp  Madrid_temp_min  \\\n",
       "0       269.338615       274.254667   265.938000       265.938000   \n",
       "1       270.376000       274.945000   266.386667       266.386667   \n",
       "\n",
       "   load_shortfall_3h  \n",
       "0        6715.666667  \n",
       "1        4171.666667  \n",
       "\n",
       "[2 rows x 48 columns]"
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Let's ensure that we have truely dropped the unnamed column in our train dataset as observed before. View first 2 rows\n",
    "df_train.head(2)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "147183cd",
   "metadata": {},
   "source": [
    "To avoid using the column names as variable names later on, we will check if some of the columns have white space which will be replaced with an underscore."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "id": "5c0b4951",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Index(['time', 'Madrid_wind_speed', 'Valencia_wind_deg', 'Bilbao_rain_1h',\n",
       "       'Valencia_wind_speed', 'Seville_humidity', 'Madrid_humidity',\n",
       "       'Bilbao_clouds_all', 'Bilbao_wind_speed', 'Seville_clouds_all',\n",
       "       'Bilbao_wind_deg', 'Barcelona_wind_speed', 'Barcelona_wind_deg',\n",
       "       'Madrid_clouds_all', 'Seville_wind_speed', 'Barcelona_rain_1h',\n",
       "       'Seville_pressure', 'Seville_rain_1h', 'Bilbao_snow_3h',\n",
       "       'Barcelona_pressure', 'Seville_rain_3h', 'Madrid_rain_1h',\n",
       "       'Barcelona_rain_3h', 'Valencia_snow_3h', 'Madrid_weather_id',\n",
       "       'Barcelona_weather_id', 'Bilbao_pressure', 'Seville_weather_id',\n",
       "       'Valencia_pressure', 'Seville_temp_max', 'Madrid_pressure',\n",
       "       'Valencia_temp_max', 'Valencia_temp', 'Bilbao_weather_id',\n",
       "       'Seville_temp', 'Valencia_humidity', 'Valencia_temp_min',\n",
       "       'Barcelona_temp_max', 'Madrid_temp_max', 'Barcelona_temp',\n",
       "       'Bilbao_temp_min', 'Bilbao_temp', 'Barcelona_temp_min',\n",
       "       'Bilbao_temp_max', 'Seville_temp_min', 'Madrid_temp', 'Madrid_temp_min',\n",
       "       'load_shortfall_3h'],\n",
       "      dtype='object')"
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# View Column Names for possible White Spaces\n",
    "df_train.columns"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "id": "c899ebc0",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(8763, 48)"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_train.shape"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "6a4fc0bd",
   "metadata": {},
   "source": [
    "From the above observations of the train and test datasets, no column name has white space."
   ]
  },
  {
   "cell_type": "markdown",
   "id": "aff10288",
   "metadata": {},
   "source": [
    "<a id=\"twothree\"></a>\n",
    "### 2.3. Univariate Analysis\n",
    "<a class=\"anchor\" id=\"2.3\"></a>\n",
    "<a href=#cont>Back to Table of Contents</a>\n",
    "\n",
    "Univariate analysis is arguably the simplest form of statistical analysis<sup>3</sup>. It is the process of analysing one (\"uni\") variable/ column (\"variate\"). \n",
    "\n",
    "We will start by checking if there are any missing values per column using both graphical and non-graphical techniques."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "id": "ac021279",
   "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>time</th>\n",
       "      <th>Madrid_wind_speed</th>\n",
       "      <th>Valencia_wind_deg</th>\n",
       "      <th>Bilbao_rain_1h</th>\n",
       "      <th>Valencia_wind_speed</th>\n",
       "      <th>Seville_humidity</th>\n",
       "      <th>Madrid_humidity</th>\n",
       "      <th>Bilbao_clouds_all</th>\n",
       "      <th>Bilbao_wind_speed</th>\n",
       "      <th>Seville_clouds_all</th>\n",
       "      <th>...</th>\n",
       "      <th>Madrid_temp_max</th>\n",
       "      <th>Barcelona_temp</th>\n",
       "      <th>Bilbao_temp_min</th>\n",
       "      <th>Bilbao_temp</th>\n",
       "      <th>Barcelona_temp_min</th>\n",
       "      <th>Bilbao_temp_max</th>\n",
       "      <th>Seville_temp_min</th>\n",
       "      <th>Madrid_temp</th>\n",
       "      <th>Madrid_temp_min</th>\n",
       "      <th>load_shortfall_3h</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>2015-01-01 03:00:00</td>\n",
       "      <td>0.666667</td>\n",
       "      <td>level_5</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.666667</td>\n",
       "      <td>74.333333</td>\n",
       "      <td>64.000000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>265.938000</td>\n",
       "      <td>281.013000</td>\n",
       "      <td>269.338615</td>\n",
       "      <td>269.338615</td>\n",
       "      <td>281.013000</td>\n",
       "      <td>269.338615</td>\n",
       "      <td>274.254667</td>\n",
       "      <td>265.938000</td>\n",
       "      <td>265.938000</td>\n",
       "      <td>6715.666667</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>2015-01-01 06:00:00</td>\n",
       "      <td>0.333333</td>\n",
       "      <td>level_10</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.666667</td>\n",
       "      <td>78.333333</td>\n",
       "      <td>64.666667</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>266.386667</td>\n",
       "      <td>280.561667</td>\n",
       "      <td>270.376000</td>\n",
       "      <td>270.376000</td>\n",
       "      <td>280.561667</td>\n",
       "      <td>270.376000</td>\n",
       "      <td>274.945000</td>\n",
       "      <td>266.386667</td>\n",
       "      <td>266.386667</td>\n",
       "      <td>4171.666667</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>2015-01-01 09:00:00</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>level_9</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>71.333333</td>\n",
       "      <td>64.333333</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>272.708667</td>\n",
       "      <td>281.583667</td>\n",
       "      <td>275.027229</td>\n",
       "      <td>275.027229</td>\n",
       "      <td>281.583667</td>\n",
       "      <td>275.027229</td>\n",
       "      <td>278.792000</td>\n",
       "      <td>272.708667</td>\n",
       "      <td>272.708667</td>\n",
       "      <td>4274.666667</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>2015-01-01 12:00:00</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>level_8</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>65.333333</td>\n",
       "      <td>56.333333</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>281.895219</td>\n",
       "      <td>283.434104</td>\n",
       "      <td>281.135063</td>\n",
       "      <td>281.135063</td>\n",
       "      <td>283.434104</td>\n",
       "      <td>281.135063</td>\n",
       "      <td>285.394000</td>\n",
       "      <td>281.895219</td>\n",
       "      <td>281.895219</td>\n",
       "      <td>5075.666667</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>2015-01-01 15:00:00</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>level_7</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>59.000000</td>\n",
       "      <td>57.000000</td>\n",
       "      <td>2.0</td>\n",
       "      <td>0.333333</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>280.678437</td>\n",
       "      <td>284.213167</td>\n",
       "      <td>282.252063</td>\n",
       "      <td>282.252063</td>\n",
       "      <td>284.213167</td>\n",
       "      <td>282.252063</td>\n",
       "      <td>285.513719</td>\n",
       "      <td>280.678437</td>\n",
       "      <td>280.678437</td>\n",
       "      <td>6620.666667</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 48 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "                  time  Madrid_wind_speed Valencia_wind_deg  Bilbao_rain_1h  \\\n",
       "0  2015-01-01 03:00:00           0.666667           level_5             0.0   \n",
       "1  2015-01-01 06:00:00           0.333333          level_10             0.0   \n",
       "2  2015-01-01 09:00:00           1.000000           level_9             0.0   \n",
       "3  2015-01-01 12:00:00           1.000000           level_8             0.0   \n",
       "4  2015-01-01 15:00:00           1.000000           level_7             0.0   \n",
       "\n",
       "   Valencia_wind_speed  Seville_humidity  Madrid_humidity  Bilbao_clouds_all  \\\n",
       "0             0.666667         74.333333        64.000000                0.0   \n",
       "1             1.666667         78.333333        64.666667                0.0   \n",
       "2             1.000000         71.333333        64.333333                0.0   \n",
       "3             1.000000         65.333333        56.333333                0.0   \n",
       "4             1.000000         59.000000        57.000000                2.0   \n",
       "\n",
       "   Bilbao_wind_speed  Seville_clouds_all  ...  Madrid_temp_max  \\\n",
       "0           1.000000                 0.0  ...       265.938000   \n",
       "1           1.000000                 0.0  ...       266.386667   \n",
       "2           1.000000                 0.0  ...       272.708667   \n",
       "3           1.000000                 0.0  ...       281.895219   \n",
       "4           0.333333                 0.0  ...       280.678437   \n",
       "\n",
       "   Barcelona_temp  Bilbao_temp_min  Bilbao_temp  Barcelona_temp_min  \\\n",
       "0      281.013000       269.338615   269.338615          281.013000   \n",
       "1      280.561667       270.376000   270.376000          280.561667   \n",
       "2      281.583667       275.027229   275.027229          281.583667   \n",
       "3      283.434104       281.135063   281.135063          283.434104   \n",
       "4      284.213167       282.252063   282.252063          284.213167   \n",
       "\n",
       "   Bilbao_temp_max Seville_temp_min  Madrid_temp  Madrid_temp_min  \\\n",
       "0       269.338615       274.254667   265.938000       265.938000   \n",
       "1       270.376000       274.945000   266.386667       266.386667   \n",
       "2       275.027229       278.792000   272.708667       272.708667   \n",
       "3       281.135063       285.394000   281.895219       281.895219   \n",
       "4       282.252063       285.513719   280.678437       280.678437   \n",
       "\n",
       "   load_shortfall_3h  \n",
       "0        6715.666667  \n",
       "1        4171.666667  \n",
       "2        4274.666667  \n",
       "3        5075.666667  \n",
       "4        6620.666667  \n",
       "\n",
       "[5 rows x 48 columns]"
      ]
     },
     "execution_count": 11,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_train.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "id": "707711f9",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(8763, 48)"
      ]
     },
     "execution_count": 12,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#check the shape of the dateset\n",
    "df_train.shape"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "73519286",
   "metadata": {},
   "source": [
    "Our Shape looks good;\n",
    "\n",
    "- For the TRAIN dataset; we have 8763 rows and 48 columns\n",
    "- And for our TEST dataset; we have 2920 rows and 47 containing all other columns except our dependent variable \"load_shortfall_3h\"\n",
    "\n",
    "NEXT we move further to verify the DataTypes and Check for possible missing values PER Column"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "id": "a27a630b",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'pandas.core.frame.DataFrame'>\n",
      "RangeIndex: 8763 entries, 0 to 8762\n",
      "Data columns (total 48 columns):\n",
      " #   Column                Non-Null Count  Dtype  \n",
      "---  ------                --------------  -----  \n",
      " 0   time                  8763 non-null   object \n",
      " 1   Madrid_wind_speed     8763 non-null   float64\n",
      " 2   Valencia_wind_deg     8763 non-null   object \n",
      " 3   Bilbao_rain_1h        8763 non-null   float64\n",
      " 4   Valencia_wind_speed   8763 non-null   float64\n",
      " 5   Seville_humidity      8763 non-null   float64\n",
      " 6   Madrid_humidity       8763 non-null   float64\n",
      " 7   Bilbao_clouds_all     8763 non-null   float64\n",
      " 8   Bilbao_wind_speed     8763 non-null   float64\n",
      " 9   Seville_clouds_all    8763 non-null   float64\n",
      " 10  Bilbao_wind_deg       8763 non-null   float64\n",
      " 11  Barcelona_wind_speed  8763 non-null   float64\n",
      " 12  Barcelona_wind_deg    8763 non-null   float64\n",
      " 13  Madrid_clouds_all     8763 non-null   float64\n",
      " 14  Seville_wind_speed    8763 non-null   float64\n",
      " 15  Barcelona_rain_1h     8763 non-null   float64\n",
      " 16  Seville_pressure      8763 non-null   object \n",
      " 17  Seville_rain_1h       8763 non-null   float64\n",
      " 18  Bilbao_snow_3h        8763 non-null   float64\n",
      " 19  Barcelona_pressure    8763 non-null   float64\n",
      " 20  Seville_rain_3h       8763 non-null   float64\n",
      " 21  Madrid_rain_1h        8763 non-null   float64\n",
      " 22  Barcelona_rain_3h     8763 non-null   float64\n",
      " 23  Valencia_snow_3h      8763 non-null   float64\n",
      " 24  Madrid_weather_id     8763 non-null   float64\n",
      " 25  Barcelona_weather_id  8763 non-null   float64\n",
      " 26  Bilbao_pressure       8763 non-null   float64\n",
      " 27  Seville_weather_id    8763 non-null   float64\n",
      " 28  Valencia_pressure     6695 non-null   float64\n",
      " 29  Seville_temp_max      8763 non-null   float64\n",
      " 30  Madrid_pressure       8763 non-null   float64\n",
      " 31  Valencia_temp_max     8763 non-null   float64\n",
      " 32  Valencia_temp         8763 non-null   float64\n",
      " 33  Bilbao_weather_id     8763 non-null   float64\n",
      " 34  Seville_temp          8763 non-null   float64\n",
      " 35  Valencia_humidity     8763 non-null   float64\n",
      " 36  Valencia_temp_min     8763 non-null   float64\n",
      " 37  Barcelona_temp_max    8763 non-null   float64\n",
      " 38  Madrid_temp_max       8763 non-null   float64\n",
      " 39  Barcelona_temp        8763 non-null   float64\n",
      " 40  Bilbao_temp_min       8763 non-null   float64\n",
      " 41  Bilbao_temp           8763 non-null   float64\n",
      " 42  Barcelona_temp_min    8763 non-null   float64\n",
      " 43  Bilbao_temp_max       8763 non-null   float64\n",
      " 44  Seville_temp_min      8763 non-null   float64\n",
      " 45  Madrid_temp           8763 non-null   float64\n",
      " 46  Madrid_temp_min       8763 non-null   float64\n",
      " 47  load_shortfall_3h     8763 non-null   float64\n",
      "dtypes: float64(45), object(3)\n",
      "memory usage: 3.2+ MB\n"
     ]
    }
   ],
   "source": [
    "# summary of dataset\n",
    "df_train.info()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "7476a7f2",
   "metadata": {},
   "source": [
    "Our Outcome Includes; \n",
    "\n",
    "1. Identified 3(three) of our Potential features having datatype OBJECT (Categorical)\n",
    "\n",
    "       a. Valencia_wind_deg --> Covert to Numerical \n",
    "       b. Seville_pressure  --> Covert to Numerical \n",
    "       c. time              --> Covert to DateTime\n",
    "       \n",
    "2. Valencia_pressure  has   6695 non-null   Instead of 8763 (TOTAL MISSING VALUE = 2068)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "id": "49b30cfe",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-06-28T08:52:37.824204Z",
     "start_time": "2021-06-28T08:52:37.811206Z"
    }
   },
   "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>Madrid_wind_speed</th>\n",
       "      <th>Bilbao_rain_1h</th>\n",
       "      <th>Valencia_wind_speed</th>\n",
       "      <th>Seville_humidity</th>\n",
       "      <th>Madrid_humidity</th>\n",
       "      <th>Bilbao_clouds_all</th>\n",
       "      <th>Bilbao_wind_speed</th>\n",
       "      <th>Seville_clouds_all</th>\n",
       "      <th>Bilbao_wind_deg</th>\n",
       "      <th>Barcelona_wind_speed</th>\n",
       "      <th>...</th>\n",
       "      <th>Madrid_temp_max</th>\n",
       "      <th>Barcelona_temp</th>\n",
       "      <th>Bilbao_temp_min</th>\n",
       "      <th>Bilbao_temp</th>\n",
       "      <th>Barcelona_temp_min</th>\n",
       "      <th>Bilbao_temp_max</th>\n",
       "      <th>Seville_temp_min</th>\n",
       "      <th>Madrid_temp</th>\n",
       "      <th>Madrid_temp_min</th>\n",
       "      <th>load_shortfall_3h</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>count</th>\n",
       "      <td>8763.000000</td>\n",
       "      <td>8763.000000</td>\n",
       "      <td>8763.000000</td>\n",
       "      <td>8763.000000</td>\n",
       "      <td>8763.000000</td>\n",
       "      <td>8763.000000</td>\n",
       "      <td>8763.000000</td>\n",
       "      <td>8763.000000</td>\n",
       "      <td>8763.000000</td>\n",
       "      <td>8763.000000</td>\n",
       "      <td>...</td>\n",
       "      <td>8763.000000</td>\n",
       "      <td>8763.000000</td>\n",
       "      <td>8763.000000</td>\n",
       "      <td>8763.000000</td>\n",
       "      <td>8763.000000</td>\n",
       "      <td>8763.000000</td>\n",
       "      <td>8763.000000</td>\n",
       "      <td>8763.000000</td>\n",
       "      <td>8763.000000</td>\n",
       "      <td>8763.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>mean</th>\n",
       "      <td>2.425729</td>\n",
       "      <td>0.135753</td>\n",
       "      <td>2.586272</td>\n",
       "      <td>62.658793</td>\n",
       "      <td>57.414717</td>\n",
       "      <td>43.469132</td>\n",
       "      <td>1.850356</td>\n",
       "      <td>13.714748</td>\n",
       "      <td>158.957511</td>\n",
       "      <td>2.870497</td>\n",
       "      <td>...</td>\n",
       "      <td>289.540309</td>\n",
       "      <td>289.855459</td>\n",
       "      <td>285.017973</td>\n",
       "      <td>286.422929</td>\n",
       "      <td>288.447422</td>\n",
       "      <td>287.966027</td>\n",
       "      <td>291.633356</td>\n",
       "      <td>288.419439</td>\n",
       "      <td>287.202203</td>\n",
       "      <td>10673.857612</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>std</th>\n",
       "      <td>1.850371</td>\n",
       "      <td>0.374901</td>\n",
       "      <td>2.411190</td>\n",
       "      <td>22.621226</td>\n",
       "      <td>24.335396</td>\n",
       "      <td>32.551044</td>\n",
       "      <td>1.695888</td>\n",
       "      <td>24.272482</td>\n",
       "      <td>102.056299</td>\n",
       "      <td>1.792197</td>\n",
       "      <td>...</td>\n",
       "      <td>9.752047</td>\n",
       "      <td>6.528111</td>\n",
       "      <td>6.705672</td>\n",
       "      <td>6.818682</td>\n",
       "      <td>6.102593</td>\n",
       "      <td>7.105590</td>\n",
       "      <td>8.178220</td>\n",
       "      <td>9.346796</td>\n",
       "      <td>9.206237</td>\n",
       "      <td>5218.046404</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>min</th>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>8.333333</td>\n",
       "      <td>6.333333</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>...</td>\n",
       "      <td>264.983333</td>\n",
       "      <td>270.816667</td>\n",
       "      <td>264.483333</td>\n",
       "      <td>267.483333</td>\n",
       "      <td>269.483333</td>\n",
       "      <td>269.063000</td>\n",
       "      <td>270.150000</td>\n",
       "      <td>264.983333</td>\n",
       "      <td>264.983333</td>\n",
       "      <td>-6618.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25%</th>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>44.333333</td>\n",
       "      <td>36.333333</td>\n",
       "      <td>10.000000</td>\n",
       "      <td>0.666667</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>73.333333</td>\n",
       "      <td>1.666667</td>\n",
       "      <td>...</td>\n",
       "      <td>282.150000</td>\n",
       "      <td>284.973443</td>\n",
       "      <td>280.085167</td>\n",
       "      <td>281.374167</td>\n",
       "      <td>284.150000</td>\n",
       "      <td>282.836776</td>\n",
       "      <td>285.816667</td>\n",
       "      <td>281.404281</td>\n",
       "      <td>280.299167</td>\n",
       "      <td>7390.333333</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>50%</th>\n",
       "      <td>2.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>1.666667</td>\n",
       "      <td>65.666667</td>\n",
       "      <td>58.000000</td>\n",
       "      <td>45.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>147.000000</td>\n",
       "      <td>2.666667</td>\n",
       "      <td>...</td>\n",
       "      <td>288.116177</td>\n",
       "      <td>289.416667</td>\n",
       "      <td>284.816667</td>\n",
       "      <td>286.158333</td>\n",
       "      <td>288.150000</td>\n",
       "      <td>287.630000</td>\n",
       "      <td>290.816667</td>\n",
       "      <td>287.053333</td>\n",
       "      <td>286.083333</td>\n",
       "      <td>11114.666667</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>75%</th>\n",
       "      <td>3.333333</td>\n",
       "      <td>0.100000</td>\n",
       "      <td>3.666667</td>\n",
       "      <td>82.000000</td>\n",
       "      <td>78.666667</td>\n",
       "      <td>75.000000</td>\n",
       "      <td>2.666667</td>\n",
       "      <td>20.000000</td>\n",
       "      <td>234.000000</td>\n",
       "      <td>4.000000</td>\n",
       "      <td>...</td>\n",
       "      <td>296.816667</td>\n",
       "      <td>294.909000</td>\n",
       "      <td>289.816667</td>\n",
       "      <td>291.034167</td>\n",
       "      <td>292.966667</td>\n",
       "      <td>292.483333</td>\n",
       "      <td>297.150000</td>\n",
       "      <td>295.154667</td>\n",
       "      <td>293.884500</td>\n",
       "      <td>14498.166667</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>max</th>\n",
       "      <td>13.000000</td>\n",
       "      <td>3.000000</td>\n",
       "      <td>52.000000</td>\n",
       "      <td>100.000000</td>\n",
       "      <td>100.000000</td>\n",
       "      <td>100.000000</td>\n",
       "      <td>12.666667</td>\n",
       "      <td>97.333333</td>\n",
       "      <td>359.333333</td>\n",
       "      <td>12.666667</td>\n",
       "      <td>...</td>\n",
       "      <td>314.483333</td>\n",
       "      <td>307.316667</td>\n",
       "      <td>309.816667</td>\n",
       "      <td>310.710000</td>\n",
       "      <td>304.816667</td>\n",
       "      <td>317.966667</td>\n",
       "      <td>314.816667</td>\n",
       "      <td>313.133333</td>\n",
       "      <td>310.383333</td>\n",
       "      <td>31904.000000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>8 rows × 45 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "       Madrid_wind_speed  Bilbao_rain_1h  Valencia_wind_speed  \\\n",
       "count        8763.000000     8763.000000          8763.000000   \n",
       "mean            2.425729        0.135753             2.586272   \n",
       "std             1.850371        0.374901             2.411190   \n",
       "min             0.000000        0.000000             0.000000   \n",
       "25%             1.000000        0.000000             1.000000   \n",
       "50%             2.000000        0.000000             1.666667   \n",
       "75%             3.333333        0.100000             3.666667   \n",
       "max            13.000000        3.000000            52.000000   \n",
       "\n",
       "       Seville_humidity  Madrid_humidity  Bilbao_clouds_all  \\\n",
       "count       8763.000000      8763.000000        8763.000000   \n",
       "mean          62.658793        57.414717          43.469132   \n",
       "std           22.621226        24.335396          32.551044   \n",
       "min            8.333333         6.333333           0.000000   \n",
       "25%           44.333333        36.333333          10.000000   \n",
       "50%           65.666667        58.000000          45.000000   \n",
       "75%           82.000000        78.666667          75.000000   \n",
       "max          100.000000       100.000000         100.000000   \n",
       "\n",
       "       Bilbao_wind_speed  Seville_clouds_all  Bilbao_wind_deg  \\\n",
       "count        8763.000000         8763.000000      8763.000000   \n",
       "mean            1.850356           13.714748       158.957511   \n",
       "std             1.695888           24.272482       102.056299   \n",
       "min             0.000000            0.000000         0.000000   \n",
       "25%             0.666667            0.000000        73.333333   \n",
       "50%             1.000000            0.000000       147.000000   \n",
       "75%             2.666667           20.000000       234.000000   \n",
       "max            12.666667           97.333333       359.333333   \n",
       "\n",
       "       Barcelona_wind_speed  ...  Madrid_temp_max  Barcelona_temp  \\\n",
       "count           8763.000000  ...      8763.000000     8763.000000   \n",
       "mean               2.870497  ...       289.540309      289.855459   \n",
       "std                1.792197  ...         9.752047        6.528111   \n",
       "min                0.000000  ...       264.983333      270.816667   \n",
       "25%                1.666667  ...       282.150000      284.973443   \n",
       "50%                2.666667  ...       288.116177      289.416667   \n",
       "75%                4.000000  ...       296.816667      294.909000   \n",
       "max               12.666667  ...       314.483333      307.316667   \n",
       "\n",
       "       Bilbao_temp_min  Bilbao_temp  Barcelona_temp_min  Bilbao_temp_max  \\\n",
       "count      8763.000000  8763.000000         8763.000000      8763.000000   \n",
       "mean        285.017973   286.422929          288.447422       287.966027   \n",
       "std           6.705672     6.818682            6.102593         7.105590   \n",
       "min         264.483333   267.483333          269.483333       269.063000   \n",
       "25%         280.085167   281.374167          284.150000       282.836776   \n",
       "50%         284.816667   286.158333          288.150000       287.630000   \n",
       "75%         289.816667   291.034167          292.966667       292.483333   \n",
       "max         309.816667   310.710000          304.816667       317.966667   \n",
       "\n",
       "       Seville_temp_min  Madrid_temp  Madrid_temp_min  load_shortfall_3h  \n",
       "count       8763.000000  8763.000000      8763.000000        8763.000000  \n",
       "mean         291.633356   288.419439       287.202203       10673.857612  \n",
       "std            8.178220     9.346796         9.206237        5218.046404  \n",
       "min          270.150000   264.983333       264.983333       -6618.000000  \n",
       "25%          285.816667   281.404281       280.299167        7390.333333  \n",
       "50%          290.816667   287.053333       286.083333       11114.666667  \n",
       "75%          297.150000   295.154667       293.884500       14498.166667  \n",
       "max          314.816667   313.133333       310.383333       31904.000000  \n",
       "\n",
       "[8 rows x 45 columns]"
      ]
     },
     "execution_count": 14,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# look at data statistics\n",
    "df_train.describe()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "84d1acd1",
   "metadata": {},
   "source": [
    "Here we take a look at the summary statistics of the caterogorical variables"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "id": "2d6abe06",
   "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>time</th>\n",
       "      <th>Valencia_wind_deg</th>\n",
       "      <th>Seville_pressure</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>count</th>\n",
       "      <td>8763</td>\n",
       "      <td>8763</td>\n",
       "      <td>8763</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>unique</th>\n",
       "      <td>8763</td>\n",
       "      <td>10</td>\n",
       "      <td>25</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>top</th>\n",
       "      <td>2015-01-01 03:00:00</td>\n",
       "      <td>level_1</td>\n",
       "      <td>sp13</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>freq</th>\n",
       "      <td>1</td>\n",
       "      <td>1072</td>\n",
       "      <td>500</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                       time Valencia_wind_deg Seville_pressure\n",
       "count                  8763              8763             8763\n",
       "unique                 8763                10               25\n",
       "top     2015-01-01 03:00:00           level_1             sp13\n",
       "freq                      1              1072              500"
      ]
     },
     "execution_count": 15,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# look at data statistics of categorical data\n",
    "df_train.describe(include=['object'])"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "a1480223",
   "metadata": {},
   "source": [
    "Valencia_wind_deg are Categorized between \n",
    "   * Level_1 to Level_10   \n",
    "   \n",
    "Seville_pressure are Categorized between\n",
    "   * sp1  to  sp25"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "id": "82902a9e",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "time                       0\n",
       "Madrid_wind_speed          0\n",
       "Valencia_wind_deg          0\n",
       "Bilbao_rain_1h             0\n",
       "Valencia_wind_speed        0\n",
       "Seville_humidity           0\n",
       "Madrid_humidity            0\n",
       "Bilbao_clouds_all          0\n",
       "Bilbao_wind_speed          0\n",
       "Seville_clouds_all         0\n",
       "Bilbao_wind_deg            0\n",
       "Barcelona_wind_speed       0\n",
       "Barcelona_wind_deg         0\n",
       "Madrid_clouds_all          0\n",
       "Seville_wind_speed         0\n",
       "Barcelona_rain_1h          0\n",
       "Seville_pressure           0\n",
       "Seville_rain_1h            0\n",
       "Bilbao_snow_3h             0\n",
       "Barcelona_pressure         0\n",
       "Seville_rain_3h            0\n",
       "Madrid_rain_1h             0\n",
       "Barcelona_rain_3h          0\n",
       "Valencia_snow_3h           0\n",
       "Madrid_weather_id          0\n",
       "Barcelona_weather_id       0\n",
       "Bilbao_pressure            0\n",
       "Seville_weather_id         0\n",
       "Valencia_pressure       2068\n",
       "Seville_temp_max           0\n",
       "Madrid_pressure            0\n",
       "Valencia_temp_max          0\n",
       "Valencia_temp              0\n",
       "Bilbao_weather_id          0\n",
       "Seville_temp               0\n",
       "Valencia_humidity          0\n",
       "Valencia_temp_min          0\n",
       "Barcelona_temp_max         0\n",
       "Madrid_temp_max            0\n",
       "Barcelona_temp             0\n",
       "Bilbao_temp_min            0\n",
       "Bilbao_temp                0\n",
       "Barcelona_temp_min         0\n",
       "Bilbao_temp_max            0\n",
       "Seville_temp_min           0\n",
       "Madrid_temp                0\n",
       "Madrid_temp_min            0\n",
       "load_shortfall_3h          0\n",
       "dtype: int64"
      ]
     },
     "execution_count": 16,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#check the dataset for null values\n",
    "df_train.isnull().sum()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "id": "7ce52a15",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<AxesSubplot:>"
      ]
     },
     "execution_count": 17,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "#Plot heatmap of missing values\n",
    "sns.heatmap(df_train.isnull())"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "330c9811",
   "metadata": {},
   "source": [
    "The above heatmap shows how the missing value is spread with in the column (Valencia_pressure)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "ef67a5d7",
   "metadata": {},
   "source": [
    "<a id=\"twofour\"></a>\n",
    "### 2.4. Multivariate Analysis\n",
    "<a class=\"anchor\" id=\"2.4\"></a>\n",
    "<a href=#cont>Back to Table of Contents</a>\n",
    "\n",
    "Multivariate analysis is the process of analysing more than one type of variable/ column where by multiple measurements are made on each variable/ column and the relationships among the columns, as well as their structures, are established<sup>4</sup>.\n",
    "\n",
    "The first step will be to check on correlation, that is, find out if there is a mutual relationship or connection between the features and the target variable \"load_shortfall_3h\" by interpreting the **correlation coefficients**. \n",
    "\n",
    "**Correlation coefficients** have values of between *-1* and *1*. A value of *0* means there is no relationship between the variables at all, *-1* means that there is a perfect negative and *1* means that there is a perfect positive correlation. Negative or positive correlation here refers to the type of graph the relationship will produce."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "id": "5727e150",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "load_shortfall_3h       1.000000\n",
       "Valencia_temp_min       0.201863\n",
       "Barcelona_temp_min      0.195984\n",
       "Madrid_temp             0.186036\n",
       "Barcelona_temp          0.182673\n",
       "Madrid_temp_max         0.182548\n",
       "Bilbao_temp_max         0.180934\n",
       "Bilbao_temp             0.176411\n",
       "Madrid_temp_min         0.174744\n",
       "Seville_temp_min        0.174718\n",
       "Valencia_temp           0.173225\n",
       "Bilbao_temp_min         0.172920\n",
       "Barcelona_temp_max      0.164734\n",
       "Seville_temp            0.159343\n",
       "Bilbao_weather_id       0.145875\n",
       "Valencia_temp_max       0.144270\n",
       "Seville_temp_max        0.109834\n",
       "Madrid_pressure         0.092638\n",
       "Valencia_pressure       0.086022\n",
       "Barcelona_weather_id    0.063868\n",
       "Seville_weather_id      0.061104\n",
       "Valencia_humidity       0.043140\n",
       "Bilbao_pressure         0.039052\n",
       "Madrid_weather_id       0.030868\n",
       "Valencia_snow_3h       -0.021633\n",
       "Barcelona_rain_3h      -0.024720\n",
       "Madrid_rain_1h         -0.027159\n",
       "Seville_rain_1h        -0.031572\n",
       "Bilbao_snow_3h         -0.031721\n",
       "Seville_rain_3h        -0.032945\n",
       "Barcelona_pressure     -0.034385\n",
       "Seville_wind_speed     -0.048433\n",
       "Barcelona_rain_1h      -0.056633\n",
       "Bilbao_wind_speed      -0.081602\n",
       "Madrid_clouds_all      -0.081623\n",
       "Seville_clouds_all     -0.091804\n",
       "Barcelona_wind_speed   -0.103633\n",
       "Barcelona_wind_deg     -0.116133\n",
       "Bilbao_wind_deg        -0.120800\n",
       "Bilbao_clouds_all      -0.127293\n",
       "Valencia_wind_speed    -0.142791\n",
       "Madrid_humidity        -0.150536\n",
       "Madrid_wind_speed      -0.150981\n",
       "Bilbao_rain_1h         -0.152510\n",
       "Seville_humidity       -0.167290\n",
       "Name: load_shortfall_3h, dtype: float64"
      ]
     },
     "execution_count": 18,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# evaluate correlation - this shows how a given feature value correlates to other features in the dataframe\n",
    "correlation = df_train.corr()\n",
    "# evaluating the correlation btw the features and the target variable\n",
    "correlation['load_shortfall_3h'].sort_values(ascending=False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "id": "85ecb685",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAABDYAAAQwCAYAAAAeit5RAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjQuMywgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy/MnkTPAAAACXBIWXMAAAsTAAALEwEAmpwYAAEAAElEQVR4nOzdeZxU5Z3v8c+36Y2mZW8RRURccI1IWg0hCG5XjcQYx5tREw1ZhizjaHKTScx6jZO4zE0m0ZuV3Iy4xERHzSRqxCgKImKwBdxFDVHQILbIYrM1Tf/uH3U6VtqqpsGnaAq+79fLV516znO+5zlV1VXFz+ecUkRgZmZmZmZmZlaOKnp6AGZmZmZmZmZm28qFDTMzMzMzMzMrWy5smJmZmZmZmVnZcmHDzMzMzMzMzMqWCxtmZmZmZmZmVrZc2DAzMzMzMzOzsuXChpmZme2UJE2W9OA72P4uSR9LOabtTdJwSS2SevX0WMzMzErFhQ0zMzMrGUnnSmrK/nG9LCsWvK+nx9WZpEsk3ZDfFhGnRsS1JdjXNEkh6fRO7T/M2id3M+dFSSd21ScilkREfURsfgdDNjMz26G5sGFmZmYlIel/AT8ELgOGAMOBnwAf3Iasyu60lZHngL/NBsmO5X8Cf061gzJ/fMzMzLrNhQ0zMzNLTlI/4FLgnyPitohYGxGbIuL2iPjXrE9NNkvhr9l/P5RUk62bKOllSV+R9CpwTTar4hZJN0haA0yW1E/SL7PZIK9I+k6x0y4kXSVpqaQ1kh6VND5rPwX4GvCP2cySx7L2mZI+lS1XSPqGpJckvSbpuuwYkTQim2nxMUlLJL0u6etbeIhuB8ZJGpDdPwV4HHg1b7z7SbpP0oos81eS+mfrridXKLo9G/OX88bxSUlLgPvy2iolDcwe0w9kGfWSXpB0/lY8tWZmZjscFzbMzMysFMYCtcBvu+jzdeA9wGjgCOBo4Bt56/cABgL7AFOytg8CtwD9gV8B1wJtwP7AkcD/AD5VZH+PZPsaCNwI/Jek2oiYTm5WyU3ZaRtHFNh2cvbfccBIoB74Uac+7wNGAScA35J0cBfHvgH4PXB2dv984LpOfQRcDuwJHAzsDVwCEBHnAUuAD2Rj/ve87SZk/U/OD4uIN4BPAL+QtDvwA2BhRHTer5mZWVlxYcPMzMxKYRDwekS0ddHnI8ClEfFaRDQD3wbOy1vfDvzviNgYEeuztrkR8d8R0Q70BU4FPp/NCHmN3D/Wz6aAiLghIlZERFtEfB+oIVeI6I6PAP8REYsjogX4KnB2p9M9vh0R6yPiMeAxcsWarlwHnJ/N/JgA/Hen8b4QEfdkx98M/EfWb0suyR6P9Z1XRMQfgf8CZgCnAZ/uRp6ZmdkOzedempmZWSmsAAZLquyiuLEn8FLe/Zeytg7NEbGh0zZL85b3AaqAZZI62io69fkbSV8kN5tjTyDIFUYGb/lQio61kty1Qzq8mre8jtysjqIi4kFJDeRmqdwREevzjoNsVsXVwHhgN3LHtrIbYy14/HmmAhcAl0XEim7kmZmZ7dA8Y8PMzMxKYS650y3O6KLPX8kVJzoMz9o6RIFt8tuWAhuBwRHRP/uvb0Qc2nmj7HoaXwE+DAyIiP7AanKnexTb15bG2gYs38J2W3ID8EXefhoK5E5DCeBdEdEX+ChvjReKj7nosWTXH/l5tr/PStp/WwZtZma2I3Fhw8zMzJKLiNXAt4AfSzpDUp2kKkmnSuq4HsSvgW9IapA0OOt/Q7HMAvtYBvwR+L6kvtkFPveTVOh0jd3IFSKagUpJ3yI3Y6PDcmCEpGLfjX4NfEHSvpLqeeuaHF2datMdVwMnAQ8UGXMLsErSXsC/dlq/nNz1PrbG17LbTwDfA64rdrFVMzOzcuHChpmZmZVERPwH8L/InWrRTG6GxQW8dS2J7wBN5H4N5Algfta2Nc4HqoGnyZ2mcQswtEC/u4G7yP3M6kvkZpPkn7LxX9ntCknzC2z/n8D15AoQf8m2/5etHOvbRMQbETEjIgrNsvg2MIbczJI7gds6rb+cXGFolaQvbWlfkt5N7vk4PyI2A1eSm91x8Ts5BjMzs56mwp+jZmZmZmZmZmY7Ps/YMDMzMzMzM7Oy5cKGmZmZmZmZmZUtFzbMzMzMzMzMrGy5sGFmZmZmZmZmZauypwdg22Y39YoGqpJmDnz3YdDanDQTgOoG4vXFSSM1eCQb2tqTZtZWVrAucSZAXWUFbEr8uFY1QMvLaTPrh8HmN9JmAvQaCK2vpc2s3h02Lk+bCVAzhI2b074GanpVwKbXk2ZSNRhWv5Q2E6DfPgTrkkaKOuDNpJmwG8HaxJkg+rA5NibN7KWakmTSvippJgAV/WmL1qSRlaqmPTYkzQSoUC0tmzYnzayv6sWGxH//tb0qWL3xnf4a7Nv1q6lkfeLPq96VFaxN/Jj2qepFRNr3FACpLvnndV1lBcTqpJmoX/rMLDf166pfTSW5XxZOrZ431m9KmjiwdxXEmqSZqG/yzz8o5Wdg2udK1CfPfCu3FN8rUo+1nvYSvFdVqDTPf/rMUuXuxrpNad+r//ryS6xasUKF1rmwUaYaqOI77JM089ymJuLlnyTNBNCwz7HpF+ckzaz6p1/z9Btp34AOGVjHwub0b+qjG+qJZVOTZmroFNrnfCVpZsW4K4mVv06aCaAB5xBLrk6bOfxC2v/8/aSZABX7fZEXVq9Pmrl/v97E8l8mzdSQT9J+52eSZgJUnPYzNm5uSppZ06uRaJ+RNFMVJ7CpfW7STICqirGsan0+aWb/6gNYuTFtYXdAzUhi3e+TZgKo7nSaN6QtmDbUDuPNTc8kzQTYrepg5vw17T8Yx+3Zj+dWpf37P7B/b+5cnLiwCZw2cjBPrkhb3DtsUB/mvZr2H4tH79GXDW2Ffjn3namtHJP883p0Qz3RelfSTFWfmjyzI/f2P6d9XX1gv8G0x+ykmQAVGs8NT/w1aeZHD9+T2HR30kxVnZz88w+yz8CYlTRTmkBb+5ykmZUV45JnduSW4ntF6tdqhcazrm1h0kyAusrRJfkOlDrzb7kleK02LU9bLDn/f0wous6nopiZmZmZmZlZ2XJhw8zMzMzMzMzKlgsbZmZmZmZmZla2XNgwMzMzMzMzs7LlwoaZmZmZmZmZlS0XNszMzMzMzMysbJWksCFphKSQdFx2v1rSSkkXdGPbh4u0/7xA2xWSJm7l2E6R9KFu9t3q/BSKPQZmZmZmZmZm9vcqS5jdBJwJ3A+cCDy/rUGSKiLi0ykGFRHTU+SYmZmZmZmZWc8rZWHjJWC4JAEfAn4LIOkGYFi273MjYomkjwH/AjwN9Mn6XQLsC+wOfFnSNRHRKGk08AvgVaAdeFuhQtLngNeBO4GV2f6OzP57FaiPiB9JegaYDxwC/EdEXN/N/P2B64GNwHMRMUXSzCzrGGBWRHxN0mDg/wF9gb8CH8syrwYOA9qAj0fEy5K+BnwAeBTotVWPtJmZmZmZmdkuqtTX2JgLHAs0AMuytikRMRH4d+DTknoBXwDGAZ8H9snbfklEnBoRT+S1fQf4KHA6MLDIfh8AxgNjgRnZ8nhgdqd+ewCfzdZ9bivyJwC/yo7jM3nt0yNiHDBG0r7AxcDVEXE8sIBcgec0YGVEHJetv1jSHsDJwHvJFT0GFdqppCmSmiQ1vcnmIkMzMzMzMzMz23WUcsYGwK3ATcB12f1ewJXZrIga4ClyRY+lEbER2Cjpz3nbP1Igc0hELAKQ1FRkv08Bh5KbrfFd4CzgYOBS4IC8fosjYk2Wpa3Ivxn4uqTrgHvzju/R7HYBMJLcTJBjJH0L6E1ulkcd8CFJxwIClgIjgMcjIoDnJK0utNOImApMBRip2igyNjMzMzMzM7NdRklnbETE88CDwC1ZU39yhYPx5GZGCGgGhmUXGB0A7JcX0V4gdrmkA7JCxJgi+w1gFXBURDwIDAfaI6K1c9dtyQfaIuLiiDif3IyLjsfxyOx2NLAYeBb4WkRMjIhjgJ9nbTdnbROAjwMvAocpZ3+gX5H9mpmZmZmZmVmeUs/YICIuBMgmRKwChkq6h9z1NIiIzZJ+CDxE7h/9L20h8pvAjcBrWV4xc8hd7wJgBW+dCrMl3ck/PfuFlxpyp5+0Z8c3SdK3yV1j4y+Svgv8ImsD+DJwO3C8pPuzthsi4pfZYzKX3HU6VnRzrGZmZmZmZma7tJIUNiLiRXKnf+S3TcsWryrQ/1rg2k5tl3S635jdLgCO6sYYvp+3/E8FxvG3zGz5Pd3Nj4ibyJ1i09k3IqIlr98Kcr8M09nnC2ReBlzW1X7NzMzMzMzM7O+VfMZGqUnqB/yuU/PnI2JhOeSbmZmZmZmZ2bYr+8JGRKwGJvZ0fvYLKWZmZmZmZma2HZX6517NzMzMzMzMzErGhQ0zMzMzMzMzK1subJiZmZmZmZlZ2Sr7a2zsqgYePoRz/nBR8txf7/22H615x86Nz1F5yrjkufv3ezJx4tEcMehPiTMBTiBe/EvSRA0F3jspaSbA5n7Dk2dWAkQkz20bMT55ZjWwW/WaxKm9of8eiTNBJ34geSZARAnq3Ur/UVOp1uSZALUVbybP7KVNyTNpXZc+sw56qS157OYozVeN3pXpX6utm9uTZ44eUprjry3B8a/emP75n/3KiOSZJ+0DC19ZlTRzdEM9j61O+7kyugGeWJP+s+pdg+Gn//lI0swPfPdU5i57V9JMgHF7wtF7D0ie+8ybaR/XQwbCY80HJs0EOHoPePS1MUkzG4fAkyuOSJo5ugEefz398z9md5j36v5JM8fvBc+tenfSzIMGwKyleyfNBDh1X3hq5XuSZh42CB5bcUzSTMi9Bp5ZucUfHt0qhwxMGrdFihL8g8NKr7GxMZqamnp6GGZmZmZmZmYld8gRY3j6sfkqtM4zNspVazPx8k+SRmrY57hRo5JmApwbi4ilP0qaqb0voHXzvKSZ1b2OJtpnJM0EUMUJtM/9atLMirGX0x6z02ZqPG3tc5JmAlRWjCNeSjsTSPtclPz5h9xrYPn65Ukzh/QeQmy8M2mmak5LntmRu6FtftLM2soxRMxKmilNSJ7ZkVuK41/TuihpZt/qUcSq3yTNBFD/s3lj44tJMwfWjGBV6/NJMwH6Vx/A/NfSzq4Zs/tuPLlibdLMwwb14ZW1q5JmAuzVpz8vrF6fNHP/fr2556U3kmaetM/A5JkdudMWvpw0c/LoYSxsbkmaObqhnsdfT5sJ8K7B9bz/63clzfzDd09lzl9XJ80EGLdnP55blfa1emD/3jz9RtpZa4cMrGPeq6lnbMLRe/SlaXna96rGIbuV5LWa+j0Vcu+rsxPPrhq/V3+eXZn2+T9oQB13/WVF0kyAU/cdVJLPldTPP+ReA6X4u0r9+u+Kr7FhZmZmZmZmZmXLhQ0zMzMzMzMzK1subJiZmZmZmZlZ2XJhw8zMzMzMzMzKlgsbZmZmZmZmZla2XNgwMzMzMzMzs7LlwoaZmZmZmZmZla2yKmxIGiGpWdJMSU2SzpY0WdLYbN0tWb+Zkuq347gmSxq7Ff3PlPSspKZOGReUZoRmZmZmZmZmO6fKnh7ANpgVEWdJqgMejIgxkCt6lHrHkioior1ze0RM28qoWcC7gIe2cv9TgCkAw/cauJW7NDMzMzMzM9v5lNWMjU7qgHWSLpE0qcD6SyXNkXQZgKSTJN0naZ6ki7O2fpJulzRL0s2SqjuHZDNBZkv6L+BLks6TNEPSfEnnZX0ukTQp6/uQpFskPS7pxEIDj4gVEdFaYNXxku6U9IikPQtsNzUiGiOisWHgdpuQYmZmZmZmZrbDKsfCxgRJM4HHgWu66Dc9IsYBYyTtC8yJiOOBY4AzJPUmN/vhzoiYADwFnFMka0/gIxHx78CtEXECMBa4sEDfQcDZwD8An9vKY1sVEacBU4GztnJbMzMzMzMzs11OORY2ZkXERGAEMBnoXaTfo9ntAmAkcKSke4GZ2f3dgf2AR7J+fwL2L5L1WN4Mi5MkzQL+CBxYoO+TEdEGLAUGdOuI3rIgu92Wbc3MzMzMzMx2OeVY2AAgKzQIeNvpI5kjs9vRwGLgYnIzLI4DlmTbvgAclfU7Bni+SFb+dTX+N3A6cDLQUmhoecvq6hgSb2tmZmZmZma2yynHwsaE7FdP5gJ3AauL9JskaQ6wICL+AtwK3ATcAKzN+vwi6/cAcCjwm27s/1bgfuDHwMptOQBJE7PZIwdKulfSkG3JMTMzMzMzM9vVldWvokTEi0BDF13OyvpNLLDtNGBagW0KXXi08z7Pyrv/XeC7nfpcUmAMG4C3jSNbN5PcKTH5puWtnw5M72pcZmZmZmZmZlZmhY1Sk9QP+F2n5s9HxMJ3kHkR8KG8poUR8fltzTMzMzMzMzOzt7iwkSciVlNklsU7yLwKuCplppmZmZmZmZnllOM1NszMzMzMzMzMABc2zMzMzMzMzKyMKSK23Mt2OI2NjdHU1NTTwzAzMzMzMzMruUNGj+HphfNVaJ2vsVGm4vXFbPrFOUkzq/7p18TSHyXNBNDeF3CjRiXNPDcWsaRlXdLM4fV1NG94OWkmQEPtMGLF9UkzNeg82mf/a9LMivH/h2i5LWkmgOrPJF75WdrMvT5D+6PfSpoJUPHuS1na0pI0c+/6euK1a5JmaveP0/7M5UkzASoO/iotm55KmllfdSjr2hYmzayrHE203ZM0E0CVJ7GhbX7SzNrKMby56ZmkmbtVHUzErKSZANIEXl3XnDRzj7oGNm5OX4Sv6dXIc6vWJ808sH/vkhz/C6vTjhNg/369S/JeVYrjf239sqSZALv3Hsq8V9ckzTx6j760bp6XNLO619Fsap+bNBOgqmIsM5asTJp5wvABrNiwJGkmwKDa4XzzjrSfK/826dDkj2tVxdjkr38ozd/A7r2HliSzVM9/Kca6cuPipJkDakayfP3ypJkAQ3oPKclYS/Vcpf53UEPtMOa/9mbSzK74VBQzMzMzMzMzK1subJiZmZmZmZlZ2XJhw8zMzMzMzMzKlgsbZmZmZmZmZla2XNgwMzMzMzMzs7LlwoaZmZmZmZmZla1tLmxImi1pSN79T0p62+9PSnrHv/Mm6RRJH3qnOVnWz7vZ7yBJ01Lsc2tIukLSxO29XzMzMzMzM7Ny9E5mbNwKnJl3/x+AW97ZcAqLiOkR8dtEWZ9OkWNmZmZmZmZmPe+dFDZuIStsSOoP9Ad+LmmmpHsk9c3vLGmkpLuz9T/I2iZLuk3SnZIekbRn1v4JSQ9LekDS8Vm/C7J1N2QZD0oa3nlQknaT9Pts+UZJX86W789um7LbaZJ+IeleSb9TTqWkWyTdC/xzsQPPtp2djW9ENr7/kjQ9G1tD1u9rkmZl/Q7P2k7Jtn1I0jlZ2+js+G8HDt7G58PMzMzMzMxsl7PNhY2IeBmokTQY+CDwO+CDETERuB34x06bXAl8LltfKakxa18VEacBU4GzsqLAJ4HxEXEsMLNTzpQs49+Bt82+iIg3gT6SKoBK4F2S9gdeKHAYsyPiRKAFOBw4A3gua3u80HFLqiJXfDg2G9+SbNWmiDglO45/zgoZoyJiAvBh4NJsTN8CTgDeB3xGUi/gO8BHgdOBgYX2m+17iqQmSU2vv7mxWDczMzMzMzOzXUblO9z+t+SKGmcAFwO/yGZR9Cd3qkq+UcAvJQHsBszI2hdkt0uBY4CRwIKI2AQQEe3ZNmRFgCsljQZqgKeKjOsx4EPA88AIYCLwQIF++fseAOwPPJq1zQPGdd4gIjZJugr4T0mrgW9kqx7Nuz2VXPHjvZJmZu2bgcHAAcAfs7bBQAMwJCIWZcdY9JokETGVXOGEd48YGMX6mZmZmZmZme0q3umvotwCTCb3D/SDgL9msxj+H6BOfRcBH8tmWzQCd2Tt+f9AF7AYGC2pEiCb5dBhNLkiwHhysxw676PDbOBr2e2TwGez5c467/sF4MjsfuPbu/+tuPJfEfFx4DXeus7IkXm3fwaeBWZFxMTsmE8BXgeeAU7K2kZHxKvAckkHKFfBGVPkmMzMzMzMzMysk3c0YyMilmYFiNuBh4GvS7oTWAa83Kn7V4CfSaoB2oFPFMlslnQNMEfSWnIFjA7PAkMl3QM83cXQZgPvAuYAbwKfjYgXu3FI/w2cLWlGtq9CdgN+nxVc2oGzgf8B1Eq6m9xMkv+ZHcfzkmZl/e6JiMskfRe4V1I70EzuNJVvAjeSK5Ss6sY4zczMzMzMzIx3fioKETE27+7bZhtERGN2u5jcKRr5puX1mw5Mz5Z/CfyyyC7Hd2NMrwNV2d05wPC8dR3jmZzXdnHe5mdtIXsVcGx+W3aqzMyI+FGnvleSu7ZIftvdwN2d2hYAR3W1XzMzMzMzMzN7u3dc2Ohpki4H8osr0yPiinLJNzMzMzMzM7NtV/aFjYj4ak/nR8S0Uo7BzMzMzMzMzAp7pxcPNTMzMzMzMzPrMS5smJmZmZmZmVnZcmHDzMzMzMzMzMqWIqKnx2DboLGxMZqamnp6GGZmZmZmZmYld/ARR/LMYwtUaF3ZXzx0V7WhrZ2n31iXNPOQgXW0bp6XNBOgutfRLGlJO9bh9XXcqFFJM8+NRTx39CFJMwEOnPc0GzenLULV9Gpk8ZoNSTNH9q2l/f4vJM0EqDjuByUZ68LmlqSZAKMb6omW25Jmqv5M2uemvcZxxdjLmbFkZdJMgBOGD+CVtauSZu7Vpz+xvNivd28bDfkkb256JmkmwG5VByfP3a3qYFo2PZU0s77q0OSvU8i9Vte0Lkqa2bd6FKtan0+aCdC/+gDuXPx60szTRg4uyev/uVXrk2YCHNi/Nz/900tJMz97zD4lea/+h8tmJM0EuPVrJyR/XA/s35vjP/fbpJn3/eRDTPynW5NmAsz8xT+U5Pjf8/5rkmYCPPyHj7P51o8nzez1D9dw3GfSvgfe/7MzOeWrf0iaCTD98vfzvn/8VdLMB2/6CP/jK3cmzfzjlafx/q/flTQT4A/fPTV57h++eyqTvjk9aeYd/3YKZ1x6T9JMgP/+1km8e9zPkmY+OuczjDvrhqSZAHNu+SjHnn9T0swHrvtHHl62OmlmV3wqipmZmZmZmZmVLRc2zMzMzMzMzKxsubBhZmZmZmZmZmXLhQ0zMzMzMzMzK1subJiZmZmZmZlZ2XJhw8zMzMzMzMzK1g5T2JDUV9LtkmZKmifpA1ux7c+z20skTZI0QtItW7H9ZEkXbMu4i2SN7dRWK2lmtvxDSb0ljZZ0dIp9mpmZmZmZme2qKnt6AHnOA6ZHxI8lCejX3Q0j4tOlG9bWiYhpW1j/eQBJo4F6YF7JB2VmZmZmZma2k9phZmwA64CjJQ2JnFXZ7IfZkh6SdLykIZJu79hA0n2S6iU1FQuV1Cjp/iznS13s/3hJd0p6RNKe2bZ/y5X0cHY7TdIvspklP5X0LUmzJF2drb9E0qRs+ceSZgHfzsuZKake+CxwkaS7JH1B0jnZ+oMlTdv6h8/MzMzMzMxs17MjFTauBxYBd2eFjKOAc4BjgROBr0fEcqBK0kBJ+wDNEdGyhdwrgTMjYjwwTtKQIv1WRcRpwFTgrC1kzoyIicChwGMRMQF4r6Tajg6SGoEB2bp7C2T8FLgqIk4FbgT+MWv/KHBDoZ1KmiKpSVLTGyte38IQzczMzMzMzHZ+O8ypKBHRBlwGXCbpOOBS4BDg/qxLQ3Z7G3AmMAC4uRvRhwO/zZ3dwgBgb2B5gX4LstulwDEF1itv+fHs9q95y6/y96fP7A88mi13ebpJRCyXhKTdgYnAN4v0m0qu8MJho8dEV5lmZmZmZmZmu4IdprCRzcBYFhGtwGtANbmiwaSICElVWddbgeuA3sD7uxH9GHBWRKyW1AtoL9Ivv1DQUcSozbYZBgwu0rfQdgAvAJOy5cYC+9sE9Mq7/2vgKuChiCg2RjMzMzMzMzPLs8MUNsjNrLhJ0obs/gXAkcAsSZuBJ4ALI2KFpABejogNRbLyXQzcJqkCaAXOANZ3c0y/AuYC84E3un0kQEQ0SVoj6QHgTwW6zAWuk9QYEecBvwN+DkzYmv2YmZmZmZmZ7cp2mMJGRNwB3NGp+Uly197o3HdSp/uN2e0lec1nZW2PAidsYd/T8panA9Oz5cuByzv1nZy3fHaBMV2S1/a5AvuamC22AO/rtHphRDzW1VjNzMzMzMzM7C07TGFje5E0gbxfKYG/Kzb0CEkHAP8J/KAnx2FmZmZmZmZWbna5wkZEzCJ3gc4dRkQ8D4zv6XGYmZmZmZmZlZsd6edezczMzMzMzMy2igsbZmZmZmZmZla2XNgwMzMzMzMzs7K1y11jY2fRDrRubk+eW6U3k2cC9K7cql/L7YY6Go/qlTgTmh7ZnDzzQKAtapJm1gC7Vb+eNBOGQXV14sycqoq25JmD69K//gFoa00e2f6Xl5PmVYyFIX1K81xFKerd1XXJI0WJnv8SqKro7i+Mb4XVr6XPrE8fWUqLX038eTVycNq8TJ+qjSVI7U2f3um/wr22Nu1YR/atpb09kmZ2KMV3oNYV65Jntq3ZkDwTYN2m9N9X1EvJMwG0x+7JMzev25Q8c+3SNckzAda9uDp55oZlLckzW1vSf/8pVe6GNenfVzesT/+aAqioSv+9an2JXquleF1VVqQ9flH8fUoRpfnAsdJqbGyMpqamnh6GmZmZmZmZWckdcsQYnn5sfsHqhmdslKl1be0sbE5bVRvdUE+0z0iaCaCKE2jekPb/WDfUDuO5ow9JmnngvKe5UaOSZgKcG4tY2/ZE0sw+lYeX5DFtn/OVpJkAFeOuZGlL2tfq3vX1vLw2fbV6WJ++xKrfJM1U/7Npu/G8pJmV517PkyvWJs0EOGxQn+SP67A+fYmVv06aqQHn0LLpqaSZAPVVh/LmpmeSZu5WdTAbN6ctQtf0aiRe+VnSTADt9RnWtC5Kmtm3ehSrWp9PmgnQv/oA/u9Df0ma+S/v3ZdX1q5KmrlXn/7JMztyr3v8laSZ579rLx5elvb/LL9naD8+9J17k2YC/PYbJyZ/DzxsUB/e94+/Spr54E0fSZ7ZkVuK74BjPzAtaSbA3NsnJ/9uUTHuSo49/6akmQ9c94+M/2jaz3+A2TeczZhjfpI0c/6fPleS4z/xot8lzQS496oPJs8tVeYpX/1D0kyA6Ze/n6MmTk2a+cjMKclfU5B7XR190i+TZs6755M0LS/N2QCF+BobZmZmZmZmZla2XNgwMzMzMzMzs7LlwoaZmZmZmZmZlS0XNszMzMzMzMysbLmwYWZmZmZmZmZly4UNMzMzMzMzMytbO0xhQ9IISSHpuOx+taSVki7oxrYPF2n/eYG2KyRNLLL/W7Z+5AX3O1rSZ4uNU9JkSWMl9Zf04RT7NDMzMzMzM9sV7TCFjUwTcGa2fCLw/LYGSaqIiE8nGdVWioiFEfHTLtZPi4i5QH/AhQ0zMzMzMzOzbbSjFTZeAoZLEvAh4LcAkm6QNFPSg5KGZ20fk9Qk6TqgT9Z2iaRrJd0FHCqpKWsfLekRSbcDB3ex/z0l3SLpcUknZtvOlFSfLf8mm9kxWdJ/S7pT0hxJ52XLsyTVSJoo6XtbGOck4LPAhGwfx3TMMJHUS9Lc7HEwMzMzMzMzsyJ2tMIGwFzgWKABWJa1TYmIicC/A5+W1Av4AjAO+DywT972SyLi1Ih4Iq/tO8BHgdOBgV3sexBwNvAPwOe2MM7miDgNuA8Yky3PByZ0dNjCOAF+CsyKiIkR8SfgIEnVwPHAjIiI/M6SpmRFkqZVK17fwvDMzMzMzMzMdn6VPT2AAm4FbgKuy+73Aq6UNBqoAZ4iV/RYGhEbgY2S/py3/SMFModExCKAjlkcRTwZEW2SlgIDsrb84kL+DIrHs9tXgI15ywOA5dn9rsZZyB3A+4EzgCs6r4yIqcBUgENGj4nO683MzMzMzMx2NTvcjI2IeB54EOi4kGd/coWJ8eRmXghoBoZlFxgdAOyXF9FeIHa5pAOyUzvGdLX7vOWOIsZKYG9JlcC7ivQtVvzoapwAm8gVbjrcAHwcGB4Rz3YxTjMzMzMzMzNjx5yxQURcCJBdYmIVMFTSPcDT2frNkn4IPAQ8S+7aHF35JnAj8FqWtzV+AtwMPAds1fkf3RjnMqB39mssX46IxZJqgd9t5RjNzMzMzMzMdkk7TGEjIl4EzurUNi1bvKpA/2uBazu1XdLpfmN2uwA4qrv7j4gNwMRs+V7g8E7dp+Vt97O85e/l9ZnZzXGe0il7I/DrrsZqZmZmZmZmZjk7TGFje5LUj7fPivh8RCzsgeH8jaT/BhZFxGs9OQ4zMzMzMzOzcrFLFjYiYjXZjIwdSUSc0dNjMDMzMzMzMysnO9zFQ83MzMzMzMzMusuFDTMzMzMzMzMrWy5smJmZmZmZmVnZUkT09BhsGzQ2NkZTU1NPD8PMzMzMzMys5A4dPYanFs5XoXW75MVDdwqbmollU5NGaugU2ud+NWkmQMXYy4kV1yfN1KDz2Lg5bWGnplcja9ueSJoJ0KfycG7UqKSZ58YiWjY9lTSzvupQ1rUtTJoJUFc5mnjpbb/Y/I5on4uImJU0E0CawNKWlqSZe9fXE+0zkmaq4gQ2tc9NmglQVTGWDW3zk2bWVo5J/nfVp/Jw2trnJM0EqKwYV5Kxrmp9Pmlm/+oDaI/ZSTMBKjSeV9c1J83co66BNzc9kzQTYLeqg3mpZUPSzH3qa3lyxdqkmYcN6sPKjYuTZgIMqBnJ8vXLk2YO6T2kJI/pc6vWJ80EOLB/bxY2p32vHt1Qn/y5GlAzkjc2vpg0E2BgzQg+8oMHkmb+6gvH8sLq9M/V/v168/jraZ+rdw2uT/64DqwZkfxvCnJ/V8vWrUiaObRuEK+tX5Y0c/feQ5N//4Hcd6Cn31iXNPOQgXWs2LAkaeag2uHMe3VN0kyAo/foy5KWtMc/vL4u+WsKcq+rUrxWU//9d8WnopiZmZmZmZlZ2XJhw8zMzMzMzMzKlgsbZmZmZmZmZla2XNgwMzMzMzMzs7LlwoaZmZmZmZmZlS0XNszMzMzMzMysbLmwYWZmZmZmZmZlq8cLG5JGSGqWNFNSk6SzJU2WNDZbd0vWb6ak+hKO4xJJk97B9ldImvgOtp8m6TBJEyV9b1tzzMzMzMzMzHYllT09gMysiDhLUh3wYESMgVzRo2eHZWZmZmZmZmY7sh6fsdFJHbCui9kTl0qaI+kyAEknSbpP0jxJF2dt/STdLmmWpJslVRfakaRvSHoo63dIp3U/kPRgNktk36ytKW/9w9ntaEmPSLodODhrG5uNZ5akS4vs+1+zcT8q6aTuPjiSpmSzWpqaV7R0dzMzMzMzMzOzndaOUtiYIGkm8DhwTRf9pkfEOGBMVnCYExHHA8cAZ0jqDUwB7oyICcBTwDmdQyQdAbwbGJf1ezZv3VHA0Ih4H/C/gW91MZ7vAB8FTgcGZm3vBy7Nci8pst2Ps3GfDHyti/y/ExFTI6IxIhobBpXsrBwzMzMzMzOzsrGjFDZmRcREYAQwGehdpN+j2e0CYCRwpKR7gZnZ/d2B/YBHsn5/AvYvkHMQuaJIAEREe9667myv7HZIRCzKcjpmdPwYOEnSdcApRY7jI5JmA7cCexbpY2ZmZmZmZmZbsKMUNgCIiFZyRYOCp48AR2a3o4HFwMXAhcBxwJJs2xeAo7J+xwDPF8h5Fnhvxx1J+Y9Dse1rJfWStA8wOGtbLukASQLGZG2rI+Ii4FPAlUWO40vZmM8qst7MzMzMzMzMumFHuXhox6koNcAdQFWRfpMkfZvcDI+/SLoVuAl4Alib9fkF8CtJ5wKvApd3DomIxyQtkDQXWA9ckLeuSdIySQ8CbcDHs1W/AuYC84E3srZvAjcCrwGrsrZPSzoT6ANMK3Ic9wOzyc0IWVOkj5mZmZmZmZltQY8XNiLiRaChiy5nZf0mFth2GoWLB1v82daI+Dfg3/KaLslb94UC/S+nU5EkIhbw1uyODjOBH25h358p0Da5U4aZmZmZmZmZbUGPFzZKTVI/4Hedmj8fEQu30/4vB8bmNU2PiCu2x77NzMzMzMzMdnY7fWEjIlYDE3tw/1/tqX2bmZmZmZmZ7ex2qIuHmpmZmZmZmZltDRc2zMzMzMzMzKxs7fSnouy0Nm4kFv85aaSGAu95f9LMDvH040nzNB5eWXtY0syRfWFd24CkmQB9KuH01tuS5/6++sykeefGInq3L0uamTOaeLbQry5vO+0DLFmYNBOAfSZQW7kqcWg9LE/7t8rQE6hc8VzaTICGsazZNDRpZG0lVFe0JM0EWLNpr+SZA2tgU3tt8tzWzfXJM8tJ6+Y+6UOrYGNbJI/ds745cWIfXmkZkjgTBtTA2k390ob2hiG1T6fNZAwH9Hs4cSbAcRw66InEmWPpX700ceZIBlS/lDgTYAT/8ZlR6VN3m588E8bx5PI3kya+a3B9CR7XEdT0KsUPFQ6hvuq1xJmDqE0+1qEMq/tT4kyAEzi47r7EmZOo7rV2y9220lF105NnwofZvfbZxJlj2K3q1cSZAINK8lrtV5u23NCrQkXXKSL9FwMrvcbGxmhqaurpYZiZmZmZmZmV3LvGvJvH5z9asLrhGRvlquVl2ud8JWlkxbgraY/ZSTMBKjSe9tn/mjZz/P9h8ZoNSTNH9q2lecPLSTMBGmqH0bLpqaSZ9VWHcqPS/t+ac2MR0XpX0kwAVZ9K+90XJM2sOPlHxEtXJc0E0D4XJX8NNNQOI5ZNTZqpoVOI5muTZgKo4WO8tj7trJ3dew9lU/vcpJlVFWN5Y+OLSTMBBtaMYFVr2tlF/asPKMljWqr36lfXpZ2xsEddAys2LEmaCTCodjjPrVqfNPPA/r2Tv64G1ozgyRXp/8/iYYP6lOQzcENb2v9jX1s5hoj7k2YCSMeV5H0lYlbSTGlC8syO3OXrlyfNHNJ7CG3tc5JmAlRWjOPGp9K+B5576NCSPFep3/8h9xnw5qZnkmbuVnUwa1oXJc3sWz2KaJ+RNBNAFScQG+5Im1k7qSSPaay5OWkmgPp+uCTvq6n/XQG5f1uU4nF9qSXtZ1VXfI0NMzMzMzMzMytbLmyYmZmZmZmZWdlyYcPMzMzMzMzMypYLG2ZmZmZmZmZWtlzYMDMzMzMzM7Oy5cKGmZmZmZmZmZWtkhc2JI2Q1CxppqQmSWdLmixpbLbulqzfTEn122E8p0j6UDf7XiFpYmlHVHC/D2/vfZqZmZmZmZmVo8rttJ9ZEXGWpDrgwYgYA7mix3ba/99ExPTtvU8zMzMzMzMzK43tfSpKHbBO0iWSJhVYf6mkOZIuA5B0kqT7JM2TdHHW1k/S7ZJmSbpZUnXnEEmfk/RhSX0ktUraXdLJki7OZotckPV7RtKvJC2QdF7WNlrSI5JuBw4udBCS9pc0N5tlMjVrmynpPzqNf7Ck/86O4QZJvZTzfyXdL+keScOyvl/LMn8E9Cqy3ynZrJem5lXrt+6RNzMzMzMzM9sJba/CxgRJM4HHgWu66Dc9IsYBYyTtC8yJiOOBY4AzJPUGpgB3RsQE4CngnAI5DwDjgbHAjGx5PDC7U789gM9m6z6XtX0H+ChwOjCw2PEAv4qIicBnuhj/xcDV2TEsAD4EnAasjIjjsvUXS9oDOBl4L3A1MKjQTiNiakQ0RkRjQ//eRYZmZmZmZmZmtuvYXoWNWVkRYAQwGSj2r/JHs9sFwEjgSEn3AjOz+7sD+wGPZP3+BOxfIOcp4FDgWOC75AoXR+Vt12FxRKyJiBZAWduQiFgUEQE0FRnnzcAwSdeRK4IUG/8hwLezos6HyRVSDgE+lLX9B9Cf3OPyeOQ8B6wusl8zMzMzMzMzy7NdT0WJiFZyBYS3nT6SOTK7HQ0sJjej4ULgOGBJtu0L5IoUkJvJ8XyB/QSwCjgqIh4EhgPt2f7/rmuBMSyXdIAkAWOKjLMtIi6OiPPJzbjoeBw7j/9Z4GsRMTEijgF+nrXdnLVNAD4OvAgclp2msj/Qr8h+zczMzMzMzCzPdj0VRdJc4C6Kz0iYJGkOsCAi/gLcCtwE3ACszfr8Iuv3ALlZGb8pkjUHeDNbXsHbZ2sU803gRuAOcsWRQk6XNFvSPHKnn7QXGf93gS9k19i4DzgCuB0YlF1j437g/Ih4FbgHmAv8r2y8ZmZmZmZmZrYFJf9VlIh4EWjoostZWb+JBbadBkwrsE2hC4923vb7ecv/1CmzY7kxb/k92e0C3poRUiz7JnIFl86+kZ3W0tFvBXBmgX6fL5B5GXBZV/s1MzMzMzMzs7+3vX7utWQk9QN+16n58xGxsBzyzczMzMzMzGzblX1hIyJWAxN7Or/QjBMzMzMzMzMzK63tevFQMzMzMzMzM7OUXNgwMzMzMzMzs7LlwoaZmZmZmZmZlS1FRE+PwbZBY2NjNDU19fQwzMzMzMzMzEru8CPfzRMLHlWhdWV/8dBd1uY3iJW/ThqpAefQ1j4naSZAZcU4ouW2pJmqP5P2+7+QNLPiuB/QPucrSTMBKsZdybq2hUkz6ypHE613Jc1U9ancqFFJMwHOjUVsaJufNLO2cgyx8c6kmQCqOY3l65cnzRzSewgbv1foV5+3Xc2XbiM23JE0E0C1k1i5cXHSzAE1I2mfcVHSzIoTrko+TsiNdU3roqSZfatH0R6zk2ZWaDztf/7+ljtube5+X6Rl01NJM+urDmXFhiVJMwEG1Q7nihnPJc28+IQDeW7V+qSZB/bvzU//9FLSTIDPHrMPH/nBA0kzf/WFY7n7xRVJM08eMYjZr6xKmgkwfq/+3PWXtGM9dd9BzHp5VdLMCcP6J8/syC3Fa/Wel95Imglw0j4DiZiVNFOakPx1NX6v/iU7/tv//HrSzA/sN7gkr9WFzS1JMwFGN9TTtPzNpJmNQ3YrSebDy1YnzQR4z9B+zFiyMmnmCcMHcOfitK8pgNNGDi7JZ8DiNRuSZnbFp6KYmZmZmZmZWdlyYcPMzMzMzMzMypYLG2ZmZmZmZmZWtlzYMDMzMzMzM7Oy5cKGmZmZmZmZmZUtFzbMzMzMzMzMrGy5sGFmZmZmZmZmZavHChuS+kq6XdJMSfMkfWArtv15dnuJpEmSRki65R2OZ6ak+new/cPvcP9N2e0lkia9kywzMzMzMzOzXUVlD+77PGB6RPxYkoB+3d0wIj5dumGZmZmZmZmZWbnoyVNR1gFHSxoSOaskTZY0W9JDko6XNETS7R0bSLpPUn3H7IZCJDVKuj/L+VKRPr0l/VrSLEn3dlrXL5tJMkvSzZKqJU2U9L1s/UGSpmXLH5PUJOk6oE/W9llJj2Tbf6jI/m/IZog8KGl4dx8wSVOy/TU1v/5mdzczMzMzMzMz22n1ZGHjemARcHdWyDgKOAc4FjgR+HpELAeqJA2UtA/QHBEtW8i9EjgzIsYD4yQNKdBnCjAvIiYA/6PAujuzdU9lY3obSb2ALwDjgM8D+2Sr/hE4Mdv+d0XGOCUiJgL/DnR79klETI2IxohobBi8W3c3MzMzMzMzM9tp9dipKBHRBlwGXCbpOOBS4BDg/qxLQ3Z7G3AmMAC4uRvRhwO/zZ3dwgBgb2B5pz4HAb/MxtHead1+wC+y5T+RK1y8mLdeeeNbGhEbgY2S/py1fxn4nqRK4ApyxZu3Ns4VRK6UNBqoIVc8MTMzMzMzM7Nt0GOFjWwGxrKIaAVeA6qBx4FJERGSqrKutwLXAb2B93cj+jHgrIhYnRUROhcuAJ4hV7BoklTRqbjxAnAU8ChwDPA8sJJcgQTg3dltMzBMUjW501D2y9qfiIh/kvRe4CvAJzrtezQwJCLGSzqdXNHGzMzMzMzMzLZBT1489HDgJkkbsvsXAEcCsyRtBp4ALoyIFZICeDkiNhTJyncxcJukCqAVOANY36nPL4Bpkv4hW3dyp3W/knQu8CpwObAJqM2ux/ECQERslvRD4CHgWeClbPufSRoB1AP/WmB8zwJDJd0DPN2N4zEzMzMzMzOzInryVJQ7gDs6NT9J7tobnftO6nS/Mbu9JK/5rKztUeCELex7PblrYeS3Tcy7W+jnVj9YIOda4NpObR/bwr7XAuMLtBc6JjMzMzMzMzPrQk/O2NguJE0Avp3f1qmIUer930XuNJoOV0TE9O21fzMzMzMzM7Od2U5f2IiIWcDEHtz/qT21bzMzMzMzM7OdXU/+3KuZmZmZmZmZ2TviwoaZmZmZmZmZlS1FRE+PwbZBY2NjNDU19fQwzMzMzMzMzEru0NFjeGrhfBVat9NfY2On1foaseTqpJEafiHx0lVJMwG0z0XEKz9Lm7nXZ1i8pju//tt9I/vWsrSlJWkmwN719ckfV+1zEe13X5A0s+LkH7GhbX7STIDayjHcqFFJM8+NRcSam5NmAqjvh2ne8HLSzIbaYbTH7KSZFRpPbLo7aSaAqk7mzU3PJM3crepgWjY9lTSzvupQXl3XnDQTYI+6hpKMdVXr80kz+1cfwMbN6QvbNb0aSzLW19YvS5oJsHvvoclfA3vUNfDK2lVJM/fq059nV65Lmglw0IA6VmxYkjRzUO1wnlyxNmnmYYP6sKQl/fEPr6/j4WWrk2a+Z2g/WjfPS5pZ3evo5JkduTcvWp4088OjhrBy4+KkmQADakZy5+LXk2aeNnJwSZ6rUn0HKsXnSimOP3dZwrSkCSUZaym+V6V+niD3XKX+vK7p1ci6toVJMwHqKkcnz62rHJ38c6UrPhXFzMzMzMzMzMqWCxtmZmZmZmZmVrZc2DAzMzMzMzOzsuXChpmZmZmZmZmVLRc2zMzMzMzMzKxsubBhZmZmZmZmZmVrhylsSBohqVnSTElNks6WNFnS2GzdLVm/mZLqt8N4TpH0oW72vULSxC30qZU0M8HQzMzMzMzMzCxT2dMD6GRWRJwlqQ54MCLGQK7osb0HEhHTt/c+zczMzMzMzGzr7DAzNjqpA9ZJukTSpALrL5U0R9JlAJJOknSfpHmSLs7a+km6XdIsSTdLqu4cIulzkj4sqY+kVkm7SzpZ0sXZbJELsn7PSPqVpAWSzsvaRkt6RNLtwMHFDkTSjyXNAr6d19Yo6X5JsyV9KWsbnh3THyTdKGnyNj96ZmZmZmZmZruIHa2wMSE7XeNx4Jou+k2PiHHAGEn7AnMi4njgGOAMSb2BKcCdETEBeAo4p0DOA8B4YCwwI1seD8zu1G8P4LPZus9lbd8BPgqcDgwsNEhJjcCAbAz35q26EjgzIsYD4yQNAb4CfDsi3g+0F8mbkp2m09T8RkuRh8bMzMzMzMxs17GjFTZmRcREYAQwGehdpN+j2e0CYCRwpKR7gZnZ/d2B/YBHsn5/AvYvkPMUcChwLPBdcoWLo/K267A4ItZERAugrG1IRCyKiACaioxz/7yxzstrPxz4bVbEGQnsnY23o2/n/QMQEVMjojEiGhsGlvwyI2ZmZmZmZmY7vB2tsAFARLSSKyC87fSRzJHZ7WhgMXAxcCFwHLAk2/YFckUKyM3keL7AfgJYBRwVEQ8Cw4H2bP9/17XAGJZLOkCSgDFFxvlC3lgb89ofAz6YFXHGkCto/Dmv77uL5JmZmZmZmZlZnh3t4qEdp6LUAHcAVUX6TZL0bXIzPP4i6VbgJuAJYG3W5xfArySdC7wKXF4kaw65wgfACmBZN8f6TeBG4DVyxZG3iYgmSWskPUBu1kiHi4HbJFUArcAZwP8BbpT0xewYNnVzHGZmZmZmZma7rB2msBERLwINXXQ5K+s3scC204BpBbYpdOHRztt+P2/5nzpldiw35i2/J7tdwFszQrrK/1yBtkeBE/LbJL0cEe/Nlm8gN9vDzMzMzMzMzLqwwxQ2Sk1SP+B3nZo/HxELd5D8fSRNI/ecPBYRf9pCfzMzMzMzM7Nd3i5T2IiI1cDEHTU/Iv5M7uKlZmZmZmZmZtZNO+TFQ83MzMzMzMzMusOFDTMzMzMzMzMrWy5smJmZmZmZmVnZ2mWusbHTiSA2pf1FWAGbho1NmglQDcSrf02aqb1gzca2pJkAUnvyTACGj04eqYMOSJ5Zs7m7v3a8FSrhnNX/ljz21/2+mTzz3Pgw/auXJk4dhjauThtZC6x6NW0mQAP00sbksX16pR7rodRVvpE4E6CBTe29k6f27pX4+QeqNzcnz6QXtEexX1nfdqV5roayamOfpIl71EGFNifNBOhdVZr/h7S6dfekeYNqoW+NkmYC7F1xb/JMOJ39BryZOLMfVW2J/656QVX7yrSZWe7YvdK+/gH6x9PJM2EkrZvTf7eqqlifPLOGFckzAWp7rUqeWRXpP1do25A+s6oEfwO9QNGaNlOl+K4CcGj6z+teUNsr9ftfTilyq3ul/QxUFx9TioikO7Pto7GxMZqamnp6GGZmZmZmZmYld9iRY3hywfyC5Q3P2ChXG5fT/ufvJ42s2O+LtG6elzQToLrX0bQ/+q2kmRXvvpSFzS1JM0c31PPy2jVJMwGG9elLxKykmdIE4qWr0mbucxGx8c6kmQCqOY1Yc3PazL4f5kaNSpoJcG4sYlP73KSZVRVjiQ13JM1U7SSi+dqkmQBq+Bjr2hYmzayrHE20z0iaqYoTWNO6KGkmQN/qUazcuDhp5oCakWzcnLYIXdOrkWi9K2kmgKpP5Y2NLybNHFgzgpZNTyXNBKivOpRnV65LmnnQgDqWrUv7f2yH1g3ipZb0/xd0n/paFq9Jmzuyby1LWtI+psPr64h1v0+aCaC602ne8HLSzIbaYck/A1VzGrHp7qSZAKo6maUtab8D7V1fn/yzCnKfV799/rWkmR86YPfSfK9quydpJoAqT6KtfU7SzMqKccnHqsqTSvZaTZ2rqpNL8r0idebfchN/Xqv6VNpjdtJMgAqNT55bofE8tyr97Kqi+9tuezIzMzMzMzMzS8yFDTMzMzMzMzMrWy5smJmZmZmZmVnZcmHDzMzMzMzMzMqWCxtmZmZmZmZmVrZc2DAzMzMzMzOzsrXFwoakEZKaJc2U1CTp7FQ7z7JvSZXXzX2OlvTZbvb9jKTJJR5Sof0+vL33aWZmZmZmZlaOKrvZb1ZEnCWpDngQ+E1XnSVVRET7Ox5dCUTEQmBhDw/DzMzMzMzMzBLY2lNR6oB1kk6SdJ+keZIuBpA0WdJNku4ETpT0CUkPS3pA0vGSaiXdkG33e0l984MlnZD1/1PHLAlJ0yT9QtK9kn6nnCHZ/Qck3SKpV+dBSnq/pC9ny3+VdLSkUZJ+JmmipO9l6+ZL+mm2z69mbcMlPSjpD8CxhR4ESQOzGSwzJf0+b6w/y47v2qyt4DFL+pqkWdkxHJ61fSybEXMd0KfIfqdkfZqa31i7lU+dmZmZmZmZ2c6nu4WNCZJmAo8D1wBzIuJ44BjgDEm9s36tEXEasAD4JDA+Io4FZgKfAu7LtrsWmNJpH5cBk4DxwIV5mbMj4kSgBTgcWAmckuUuAY4vMN4HgfdK2h94LMscD8zu1K8/cAUwFug4xebLwKUR8X4gijweRwJNETEROCOvfX52fBslTSx0zFkhY1RETAA+DFyaFWe+AIwDPg/sU2inETE1IhojorFhYMHah5mZmZmZmdkuZWtPRakGZgDPSvo2UAWMBHbP+j2S3Y4EFkTEJoCIaJd0CHCUpPOz7ToXGSoi4nUASc8Be2btC7LbpcAAYCDwM0kDgKHkChd/JyLWSNoNmAhcDZwLtAPfAPbL67oyIl7K9rk+a9sfeDRbnlfs8SBXOLkWeAL4Xtb+aN7tfkChYz4423Zm1ncz0AAsjYiN5Ioify6yXzMzMzMzMzPL093CBgAR0SpJwHeAfwaeJfePf2VdOq6rsRgYLakyItokVWR950bE9QCSqoC98uLbJQ0G1gAHAn/t2G1eHwEfAf4YET+R9B95++7sSeCfgBOA84C9I2KppPzCRqEZGS+Qm5FxL9BIrpDTWVVE/Ft2HH+UdHPWfiS5osaR5K5D0qfAMR9MrlD0qby2dmBYVjjqw98XX8zMzMzMzMysiO4WNjpORakB7gBeAW4iN1vhbRd7iIhmSdcAcyStJVcImQpMlfTxrNv3gafyNvsacGe2fHVErM/VUN5mBnC9pJOzfT9eZMwPAGMjokXSU0BbN4/134EbJX0JWFWkz1GSLgN6AX8BXs7aj5b0EWBJRMzMft3k7445Iu6U9LykWeQKGvdExGWSfgg8RK4A9FI3x2pmZmZmZma2S9tiYSMiXiR3qkRn07q6HxG/BH7Zqc/5BXLOyvrPIHfNjvyMyXnLF+etOrz4iP/W/1bg1mz5u3ntM8ld84OIaMxrf092uwR43xayH+jcJyvCXB0RT+b120CBY46IK4ErO7VdS+46HGZmZmZmZmbWTVt1KsqOStJdQO+8pisiYnq55JuZmZmZmZnZttkpChsRcWpP5+fPLjEzMzMzMzOz7aO7P/dqZmZmZmZmZrbDcWHDzMzMzMzMzMqWCxtmZmZmZmZmVrYUET09BtsGjY2N0dTU1NPDMDMzMzMzMyu5w0aP4cmF81Vo3U5x8dBd0cbN7bywen3SzP379Wb5+uVJMwGG9B7C0paWpJl719cTLbclzVT9mcSq3yTNBFD/s0ty/M0bXk6a2VA7rGTPfynGuql9btJMgKqKsdyoUUkzz41FvLquOWnmHnUNyf/+IfceUIqxPrcq7VgP7N+b19YvS5oJsHvvobyx8cWkmQNrRrBs3YqkmUPrBvHsynVJMwEOGlDHig1LkmYOqh2e/O8fcu8BEbOSZkoTiHW/T5tZdzqx6e6kmQCqOpnWzfOSZlb3OrokmS+1bEiaCbBPfW1Jnv85f12dNHPcnv2YsWRl0kyAE4YPYEPb/KSZtZVjmPfqmqSZAEfv0Zd1bQuTZtZVjubOxa8nzTxt5GCalr+ZNBOgcchu3L807WvguL0HMOvlVUkzJwzrz+xX0mYCjN+rPw8vS/t39Z6h/Ury/M9/Lf3zP2b30jz/97z0RtJMgJP2GZg896R9BvL0G+m/rxTjU1HMzMzMzMzMrGy5sGFmZmZmZmZmZcuFDTMzMzMzMzMrWy5smJmZmZmZmVnZcmHDzMzMzMzMzMqWCxtmZmZmZmZmVraSFTYkjZDULGmmpCZJZyfOviVVXjf3OVrSZ7vZ9zOSJnej38PveGBmZmZmZmZm9jeVifNmRcRZkuqAB4HfdNVZUkVEtCceQxIRsRBY2MPDMDMzMzMzM7MulOpUlDpgnaSTJN0naZ6kiwEkTZZ0k6Q7gRMlfULSw5IekHS8pFpJN2Tb/V5S3/xgSSdk/f/UMUtC0jRJv5B0r6TfKWdIdv8BSbdI6tV5kJLeL+nL2fJfJR0taZSkn0maKOl72br5kn6a7fOrWdtwSQ9K+gNwbLEHQtLXJM2V9COgV9Y2UtLd2eyWH2Rt/SX9UdJ0Sb+UdMk7fRLMzMzMzMzMdnapCxsTJM0EHgeuAeZExPHAMcAZknpn/Voj4jRgAfBJYHxEHAvMBD4F3Jdtdy0wpdM+LgMmAeOBC/MyZ0fEiUALcDiwEjgly10CHF9gvA8C75W0P/BYljkemN2pX3/gCmAs0HGKzZeBSyPi/UAUejAk7QGcDLwXuBoYlK26EvhcREwEKiU1Av8E3BIRpwDLiuRNyU7zaXpjxeuFupiZmZmZmZntUlIXNmZl/1gfAUwGjpR0L7mCxUhg96zfI9ntSGBBRGwCyE5LOQT4bFYg+V/A4M5jjojXI6IVeA7YM2tfkN0uBQYAA4FbJM0iVwjZs1MOEbEG2A2YSK7wMJpcYeOBTl1XRsRL2fjWZ237A49my/OKPB4jgMcj5zlgddY+CvhldozvBYYB++XlPUIBETE1IhojonHgoM4Pi5mZmZmZmdmupySnomRFBwHfAS4EjiM3a0JZl47raiwGRkuqhNw1N4BngasjYmJEjAO+2Sm+XdJgSdXAgcBfO3ab10fAR4A/RsQE4I68fXf2JLnZErPJnSqyT0Qs7XxIBbZ7ATgyW24skv0icFh2asz+QL+sfRHwsawI1JiN7895ee8ukmdmZmZmZmZmeVJfPLTjVJQacv9YfwW4CXgCWNu5c0Q0S7oGmCNpLblCyFRgqqSPZ92+DzyVt9nXgDuz5asjYr1UsGYxA7he0snZvh8vMuYHgLER0SLpKaCtm8f678CNkr4ErCrUISJelXQPMBeYD6zIVn0F+JmkGnJFnk8A/w/4L0n/E3iNXIHHzMzMzMzMzLqQrLARES8CDQVWTevqfkT8Evhlpz7nF8g5K+s/g9w1O/IzJuctX5y36vDiI/5b/1uBW7Pl7+a1zyR3Cg0R0ZjX/p7sdgnwvm7kX0buuiD5bYuBU/PbstkqJ0fEZknfITcjxMzMzMzMzMy6kHrGxg5N0l1A77ymKyJi+g6S3xuYrtz0k+XApanGZWZmZmZmZraz2qUKGxFx6pZ79Ux+RKwld+FSMzMzMzMzM+umklw81MzMzMzMzMxse3Bhw8zMzMzMzMzKlgsbZmZmZmZmZla2FBE9PQbbBo2NjdHU1NTTwzAzMzMzMzMruXeNeTePz39UhdbtUhcP3alsep1Y3vlXct8ZDfkksfHOpJkAqjmNeO2atJm7f5z2uV9Nmlkx9nLabjwvaSZA5bnXE+0zkmaq4gRi2dS0mUOnsPF7ZybNBKj50m20x+ykmRUaT2y4I2kmgGon8eq65qSZe9Q1cKNGJc08NxYRLbclzQRQ/Zmsan0+aWb/6gNoX3Rl0syKUV9hbdsTSTMB+lQeXpLjb2ufkzSzsmIc7TMuSpoJUHHCVbyx8cWkmQNrRvDmpmeSZgLsVnUw512V9n3l+ovG8/Qb65JmHjKwjp8/siRpJsCnjxrOZfcsSpr5tZNGcddfViTNPHXfQcx6eVXSTIAJw/onz50wrD8zlqxMmnnC8AHJMzty5/x1ddLMcXv243cvpP38A/jg/g20bp6XNLO619HcvzTt43rc3gNKdvw3PrUsaea5hw4tyd/qw8vSvqYA3jO0X0leq6X4+089TijN39UH92/g5kXLk2YCfHjUkJKM9aWWDUkzu+JTUczMzMzMzMysbLmwYWZmZmZmZmZly4UNMzMzMzMzMytbLmyYmZmZmZmZWdlyYcPMzMzMzMzMypYLG2ZmZmZmZmZWtlzYMDMzMzMzM7Oy1SOFDUkjJIWk47L71ZJWSrqgG9s+XKT95wXarpA0sZtjukTSpO70LbJ9t/dVZPtpkg6TNFHS97Y1x8zMzMzMzGxX0pMzNpqAM7PlE4HntzVIUkVEfDrJqMzMzMzMzMysbPRkYeMlYLgkAR8Cfgsg6QZJMyU9KGl41vYxSU2SrgP6ZG2XSLpW0l3AoZKasvbRkh6RdDtwcLGdS/qGpIckzZJ0SKd1P8j2P1PSvllbU976h4vtS9JYSfOy3EuL7PtfJd0n6VFJJ3X3AZM0JXscmprfeLO7m5mZmZmZmZnttCp7eP9zgWOBBmAOUA9MiYh1kk4HPi3pW8AXgGPIFTVezNt+SUR8DCBXHwHgO8BHgeeABwrtVNIRwLuBcRERkiry1h0FDI2I90maAHwL+HiR8Rfa1/uBSyPijvzcTn4cEf9H0mDgv4B7ivT7OxExFZgK0HjEiOjONmZmZmZmZmY7s56+eOitwA+Amdn9XsCVkmYD3wD2JFf0WBoRGyPiDeDPeds/UiBzSEQsioggd7pLIQcBc7I+RER73rr98nL/BOxfYPuOKkqhff0YOCmbXXJKkf1/JDvGW7NjNDMzMzMzM7Nt0KOFjYh4HngQuCVr6k+uWDCe3GwIAc3AsOwCowPIFR465BckOiyXdEB2isuYIrt+Fnhvx51OMyteAI7Klo/hrWt/1ErqJWkfYHAX+1odERcBnwKuLLL/LwHHAWcVWW9mZmZmZmZm3dDTp6IQERfC304lWQUMlXQP8HS2frOkHwIPkStIvLSFyG8CNwKvZXmF9vmYpAWS5gLrgQvy1jVJWibpQaCNt05D+RW5U2fmA290sa9PSzqT3Gkz04qM8X5gNrkZIWu2cDxmZmZmZmZmVkSPFDYi4kU6zVaIiGnZ4lUF+l8LXNup7ZJO9xuz2wW8NeOiqzH8G/BveU2X5K37QoH+lwOXd2ortK+ZwA+3sO/PFGib3CnDzMzMzMzMzLagx2dslJqkfsDvOjV/PiIWbqf9Xw6MzWuaHhFXbI99m5mZmZmZme3sdvrCRkSsBib24P6/2lP7NjMzMzMzM9vZ9fSvopiZmZmZmZmZbTMXNszMzMzMzMysbO30p6LstNatJZoeSRqp0z6ZNC9frHg1aZ52h/v3+nLSzBOAZ0/+WdJMgMOANuqSZlYBVNYkzQSovuATyTMB1LYubWAV8OaKtJkAtdCyqT557DlvXr7lTlvp17ulP8vt3DiTPpWvJ049AI0cnTgTxObkmQCb29P/Xb25aWjSvAE1oKPGJc3ssDnSfy1QwV9mf+c+ceZhyTP71bQmTqxj7L4DE2fmnHLEnskzq3ul//9dx+71WPJMmEC/mvSv1eP3Xpg48bgSZOZyB/WuSp56+t7zkmfCady/dL+kiSePgInDFibNhOM4dd+/JM4EaOCcQ55NnDmUk0c8nThzPMfssTBxJsAEjtnjycSZ40rwvjKBo/dI/ZgCjOX0Pecmzjyd/3lA6scUYAgf2C/1a7WB6oq1SRO7+v6niEi6M9s+Ghsbo6mpqaeHYWZmZmZmZlZyR4wZzWPzF6rQOs/YKFerX6L9zrf9auw7UnHaz4iNdybNBFDNabQ/k/b/WFcc/FVmLFmZNPOE4QN4ckXaqiLAYYP6sKk9bbW2qmIs0XztljtuBTV8jNhwR9JMANVOIjbdnTaz6uTkxw+5x+CF1euTZu7frzfRclvSTNWfyY0alTQT4NxYVJrXagme/3VtC5NmAtRVjmbFhiVJMwfVDmflxsVJMwfUjCTW3Jw0E0B9P0zzhpeTZjbUDqNl01NJMwHqqw7l/qVpPwOO23sAr6xdlTRzrz79efz1lqSZAO8aXM/8195Mmjlm991K8rkaMStpJoA0gYXNaR/X0Q31RNyfNFM6LnlmR+6zK9POhDxoQF3JvgPe/WLaGZYnjxhUkueqdXP6GSvVvY4uyVjbY3bSzAqNL9nfalv7nKSZlRXjko9VmpD8+w9k34HW/T5ppupOJ9pnJM0EUMUJJXldLVtXghnWxfa33fZkZmZmZmZmZpaYCxtmZmZmZmZmVrZc2DAzMzMzMzOzsuXChpmZmZmZmZmVLRc2zMzMzMzMzKxsubBhZmZmZmZmZmWrZIUNSX0l3S5ppqR5kj6wFdv+PLu9RNIkSSMk3ZJoXBdL2rebfR9Osc+tIekgSdO2937NzMzMzMzMylFlCbPPA6ZHxI8lCejX3Q0j4tOlGlREXFGqbDMzMzMzMzPbvkp5Kso64GhJQyJnlaTJkmZLekjS8ZKGSLq9YwNJ90mql9RULFRSo6T7s5wvFelzk6TdJZ0s6ams7XJJ75E0TdJhkiZKmi7pd5Iek3R41u9jkpokXQf0KZJ/hqRHstkon83anpZ0vaQFks4oNlZJgyX9d3asN0jqJalS0i2S7gX+uYtjn5KNral59YauH30zMzMzMzOzXUApCxvXA4uAu7NCxlHAOcCxwInA1yNiOVAlaaCkfYDmiGjZQu6VwJkRMR4YJ2lIgT6zgfHZfy9JGgy8G3i0U7+qiPgg8CXg45J6AV8AxgGfB/YpMoZ/AD4eEROBn2dtw4GLsuP7ZhdjvRi4OiKOBxYAHwLOAJ6LiBOBx4sdeERMjYjGiGhs6FdbrJuZmZmZmZnZLqNkp6JERBtwGXCZpOOAS4FDgPuzLg3Z7W3AmcAA4OZuRB8O/DZ3dgsDgL2B5Z36PAB8EtgdmAacBLRFxKZsuw4Ls9ulWVYDsDQiNgIbJf25yBj+DfiCpD7AT4CHgb9ExBsAkjZKqiwy1kOAYyR9C+hNrgBUz1tFl3nkCitmZmZmZmZmtgUlK2xkMzCWRUQr8BpQTW42wqSICElVWddbgevI/SP//d2Ifgw4KyJWZzMs2gv0eZLcDI3nyc3euAO4vUC/yB8y0AwMk1RN7jSU/YqMYWlEfFbSXuQKE8cDIyT1BzYBNRHRJqnQWJ8FfhsRswGyx+GDwJHZY9HYjcfAzMzMzMzMzCjtxUMPB26S1HExiAvI/eN9lqTNwBPAhRGxQlIAL0dEdy4ccTFwm6QKoJXcaRzr8ztERLuklcAjEbFMUl9yszi6FBGbJf0QeIhcAeKlIl0vkTSW3EyL72VtS8nN3jiI3OyUYmP9LvALSd/O+nwZ+G/gbEkzsv2amZmZmZmZWTeU8lSUO8jNlMj3JLkZDp37Tup0vzG7vSSv+ays7VHghG7s/wN5ywfkLU/O6zYza3sWmJwtXwtcu4XsrxRoXhcR53bqV2is68mdetPZWV3t08zMzMzMzMzerpQzNrYLSROAb+e3ZRf1LIt8MzMzMzMzM9t2ZV/YiIhZwMSezu+YZWJmZmZmZmZm208pf+7VzMzMzMzMzKykXNgwMzMzMzMzs7LlwoaZmZmZmZmZlS1FRE+PwbZBY2NjNDU19fQwzMzMzMzMzEruXWOO5PH5C1RoXdlfPHRXFaxj4+a0hY2aXo1saJufNBOgtnIMLZueSppZX3Uor6xdlTRzrz79eXntmqSZAMP69E3+uNZWjuG19cuSZu7eeygrNy5OmgkwoGYkb256JmnmblUHs65tYdJMgLrK0by6rjlp5h51DaxqfT5pZv/qA9jUPjdpJkBVxVhu1KikmefGopK8/ls3z0uaCVDd6+iSjHXFhiVJMwfVDi/Z8TdveDlpZkPtsLL6Wz3//z6YNPO6f3lf8s8qyH1eleJ19cbGF5NmDqwZUVav1dRjre51dMle/6X4DlSq56oUn4FrWhclzexbPYq29jlJMwEqK8Ylz62sGFeSfwOk/q4Oue/rpfg3QCm+V5bq+EvxvlKq74ClGGsp/m1VjE9FMTMzMzMzM7Oy5cKGmZmZmZmZmZUtFzbMzMzMzMzMrGy5sGFmZmZmZmZmZcuFDTMzMzMzMzMrWy5smJmZmZmZmVnZcmHDzMzMzMzMzMrWditsSBohqVnSTElNks5OnH1Lqryt3Pcpkj60Ff3fK+lJSa/mtU2U9L3SjNDMzMzMzMxs51W5nfc3KyLOklQHPAj8pqvOkioion37DK1rxcYSEdO3Muop4GjggSQDMzMzMzMzM9uF9dSpKHXAOkknSbpP0jxJFwNImizpJkl3AidK+oSkhyU9IOl4SbWSbsi2+72kvvnBkk7I+v9J0uSsbZqkX0i6V9LvlDMku/+ApFsk9So0UElPS7oO+D9djPeCbPkZSb+StEDSeYXyImJ1RKwrsOqwbGyPSTq8yFimZLNdmpqbV3brgTYzMzMzMzPbmW3vwsYESTOBx4FrgDkRcTxwDHCGpN5Zv9aIOA1YAHwSGB8RxwIzgU8B92XbXQtM6bSPy4BJwHjgwrzM2RFxItACHA6sBE7JcpcAxxcZ8zDgooj4Yhfj7bAH8Nls35/r/sMCQFVEfBD4EvDxQh0iYmpENEZEY0PDgK2MNzMzMzMzM9v59NSpKNXADOBZSd8GqoCRwO5Zv0ey25HAgojYBBAR7ZIOAY6SdH623exO+6iIiNcBJD0H7Jm1L8hulwIDgIHAzyQNAIYCjxUZ8wsR0TE94sgi4+2wOCLWZPvWlh+Ov7Ow0/jMzMzMzMzMbAu2d2EDgIhozf7h/x3gn4FngXlARzGg41oWi4HRkiojok1SRdZ3bkRcDyCpCtgrL75d0mBgDXAg8NeO3eb1EfAR4I8R8RNJ/5G3787yr6txMXBhgfH+7dC2ePDFdR6fmZmZmZmZmW3B9i5sdJyKUgPcAbwC3AQ8Aazt3DkimiVdA8yRtJZcIWQqMFVSx+ka3yd3Qc4OXwPuzJavjoj1RSZPzACul3Rytu/HuzH+W7sab3dIOhj4v8CBku4FvrgtOWZmZmZmZma2HQsbEfEi0FBg1bSu7kfEL4FfdupzfoGcs7L+M8hdAyM/Y3Le8sV5qwpepLPTto15y9O6Gm+nvu8pkvcMcGKBVTOz9c8CkwusNzMzMzMzM7NOeuRUlB2VpLuA/AuCXrENP+ean/eP5C4m2uHViDh7W/PMzMzMzMzM7O+5sJEnIk5NnHcTuVNXzMzMzMzMzKwEtvfPvZqZmZmZmZmZJePChpmZmZmZmZmVLUW8k18otZ7S2NgYTU1NPT0MMzMzMzMzs5I7YswRPDb/sYI/eeprbJStN4n2GUkTVXECEbOSZgJIE1jXtjBpZl3laGJ55x/LeWc05JPEyl8nzQTQgHNY2/ZE0sw+lYezqX1u0syqirG0z7goaSZAxQlX0bLpqS133Ar1VYcmf/1D7m/guVXrk2Ye2L837YuuTJpZMeorxKa7k2YCqOpkNrTNT5pZWzmGGzUqaea5sYg3Nr6YNBNgYM0IVm5cnDRzQM1IVmxYkjRzUO1w2hdckjQToOLIS1jV+nzSzP7VB5Tsufq/D/0laea/vHdf5r26Jmnm0Xv05eFlq5NmArxnaD/Gf/Q3STNn33A2i9dsSJo5sm8tL6xO+54KsH+/3ixtaUmauXd9PU+/sS5p5iED63h2ZdpMgIMG1LGkJW3u8Pq65McPuccg9XdLaULyx/WgAaV7rp5csTZp5mGD+pTku0rq1xTkXlcvtaR9X9mnvrYkmanf/yD3HliK95VS/a2W4nW1fP3ypJld8akoZmZmZmZmZla2XNgwMzMzMzMzs7LlwoaZmZmZmZmZlS0XNszMzMzMzMysbLmwYWZmZmZmZmZly4UNMzMzMzMzMytbLmyYmZmZmZmZWdna4QsbkvpKul3STEnzJH1gK7b9eXZ7iaRJkkZIuiXx+H4oqfdW9P+OpL9K+l5e2zRJh6Ucl5mZmZmZmdmuoLKnB9AN5wHTI+LHkgT06+6GEfHpVIOQVBER7QX28fmtjPoRMAM4LcW4zMzMzMzMzHZlO/yMDWAdcLSkIZGzStJkSbMlPSTpeElDJN3esYGk+yTVS2oqFiqpUdL9Wc6XivSZLOkmSXcCJ0r6Xt7MkdFZn5nZviZLuk3SnZIekbRnocyIeBWIAqsuknSvpN9lBZxC45kiqUlSU3PzqmKHZmZmZmZmZrbLKIfCxvXAIuDurJBxFHAOcCxwIvD1iFgOVEkaKGkfoDkiWraQeyVwZkSMB8ZJGlKkX2tEnBYRfwS+FRETgU8B/1qg76qIOA2YCpy1lcc5OyJOBFqAwwt1iIipEdEYEY0NDf23Mt7MzMzMzMxs57PDn4oSEW3AZcBlko4DLgUOAe7PujRkt7cBZwIDgJu7EX048NtscsQAYG9geYF+j+Qtf1HSyUA7sLlA3wXZ7VLgmG6Modi2A7ZyWzMzMzMzM7Nd0g5f2MhmYCyLiFbgNaAaeByYFBEhqSrreitwHdAbeH83oh8DzoqI1ZJ6kStWFNKejWMQMAl4D7miyNUF+uafYlLwdJIuvJNtzczMzMzMzHZJO3xhg1wR4SZJG7L7FwBHArMkbQaeAC6MiBWSAng5IjYUycp3MXCbpAqgFTgDWN9F/5XkZnTcDzy0TUcCSLoIOB8YLGmPiPjotmaZmZmZmZmZ7ep2+MJGRNwB3NGp+Uly197o3HdSp/uN2e0lec1nZW2PAidsYd/T8pbbgdML9JmYLeb3nQ5ML5J5FXBVp7bJecsXdzUmMzMzMzMzM3vLDl/Y2F4kTQC+nd+WV7TY1sxrgH3zmqblF0vMzMzMzMzM7J1xYSMTEbOAiYkzP54yz8zMzMzMzMz+Xjn83KuZmZmZmZmZWUEubJiZmZmZmZlZ2XJhw8zMzMzMzMzKliKip8dg26CxsTGampp6ehhmZmZmZmZmJXfEmCN4bP5jKrTOFw8tU8FaNrXPTZpZVTGW3DVU05ImEG33pM2sPIk3Nz2TNHO3qoNp2fRU0kyA+qpDaWufkzSzsmIcb2x8MWnmwJoRrNy4OGkmwICakby6rjlp5h51DaxpXZQ0E6Bv9SheW78saebuvYeytu2JpJl9Kg9nXdvCpJkAdZWjad08L2lmda+jS/JavVGjkmYCnBuLSjLWUmSW6r0q9XvAgJqRtM/6YtJMgIoJ3y/JWJevX540c0jvIcnfUyD3vrKq9fmkmf2rDyjJe/WzK9clzQQ4aEAdC5tbkmaObqjnhdXrk2bu36938syO3NSP60ED6ko21mXrViTNHFo3iMVrNiTNHNm3NnlmR+5LLWlz96lPP9ZSHn8p/q5K8Zg+tyr96//A/r1Z0pL2b3V4fV3y44fcY1CKsab+XO2KT0UxMzMzMzMzs7LlwoaZmZmZmZmZlS0XNszMzMzMzMysbLmwYWZmZmZmZmZly4UNMzMzMzMzMytbLmyYmZmZmZmZWdlyYcPMzMzMzMzMytYOWdiQNEJSs6SZkpoknS1psqSx2bpbsn4zJdX39Hi3hqSTJD2Y/Xe9pF75x2RmZmZmZmZm3VfZ0wPowqyIOEtSHfBgRIyBXNGjZ4f1js2KiPcBSLoGeC+wtGeHZGZmZmZmZlaedsgZG53UAeskXSJpUoH1l0qaI+ky+NuMiPskzZN0cdbWT9LtkmZJullSdecQSftLmpvNApmatc2U9ENJD0j6SbEsSf8u6WhJoyStyWZhfFrS2Z33ExGtWY4AAX/JVu0p6RZJj0s6sdADIWlKNoOl6fXmlVv9QJqZmZmZmZntbHbkwsYESTOBx4Fruug3PSLGAWMk7QvMiYjjgWOAMyT1BqYAd0bEBOAp4JxC+wN+FRETgc/ktd8WEccC75LUr0jWA8D47L85wOhseXahAUs6L9u2AWjOmgcBZwP/AHyu0HYRMTUiGiOicXDDgC4eEjMzMzMzM7Ndw45c2JiVFRlGAJOB3kX6PZrdLgBGAkdKuheYmd3fHdgPeCTr9ydg/wI5NwPDJF0HfDSvfUF2+zLQv0jWg8A44D3AvwPHAntFxCuFBhwR10fEIcCLwIey5icjoo3caSmuWpiZmZmZmZl1w45c2AD+duqGgLedPpI5MrsdDSwGLgYuBI4DlmTbvgAclfU7Bni+QE5bRFwcEecDF0vqeGwir0/BrIhYRa7oUU9ulsYZwEuFBiupJu/uGmBtkf2YmZmZmZmZ2RbsyIWNCdk1LuYCdwGri/SbJGkOsCAi/gLcCtwE3MBbRYNfZP0eAA4FflMg53RJsyXNI3d6S3uR/RXLegpYks26aCV3ekoh52fHNQsYCNxZpJ+ZmZmZmZmZbcEO+asoEfEiuetPFHNW1m9igW2nAdMKbFPowqP5291EriCS3zYxbzn/QqBvy4qIf8lbPqmL/fyCXHEk34u8dUwbgImYmZmZmZmZ2RbtkIWNUssuAvq7Ts2fj4iFifczAfh2fluhYoyZmZmZmZmZbZtdsrAREavZDrMiImLW9tiPmZmZmZmZ2a5qR77GhpmZmZmZmZlZl1zYMDMzMzMzM7OypYjYci/b4TQ2NkZTU1NPD8PMzMzMzMys5I4YcwSPzX9MhdbtktfY2Blsjo2san0+aWb/6gPY0DY/aSZAbeWY5Lm1lWN4c9MzSTN3qzo4eWZH7tq2J5Jm9qk8vCTP/5rWRUkzAfpWj6Jl01NJM+urDmXlxsVJMwEG1IzkjY0vJs0cWDOiJM/Vig1LkmYCDKodXpK/1dTPVSmeJ8g9VzdqVNLMc2NRSV5TpTr+chprKV5Xqf+uBtUOL9nfavOGl5NmNtQOK8nxL1u3ImkmwNC6QSxesyFp5si+tSxfvzxp5pDeQ3ht/bKkmQC79x6a/HEdWjeIV9c1J80E2KOuoSTvK6kf1917D03+/EPuNVCKv9Vyeq2Ww3M1pPeQkj3/pRhr6tcU5F5X5fBa7YpPRTEzMzMzMzOzsuXChpmZmZmZmZmVLRc2zMzMzMzMzKxsubBhZmZmZmZmZmXLhQ0zMzMzMzMzK1subJiZmZmZmZlZ2SpJYUPSCEnNkmZKapJ0duLsW1LlmZmZmZmZmVn5KuWMjVkRMRE4FvjyljpL2qFnj5RyfDv6sZuZmZmZmZntqCq3wz7qgHWSTgK+CtQDt0XEFZImA6dmbVdJGgZMAVqBS4CHgP8H7Am0AB/ND5Z0AvBdQMBPI2KapGnAJmBfYC1wBrA78CugGngN+MeI2Nx5oJKeBh4FDgO+HRH/LWlmR5ukDxYYz+7A9cBG4LmImJKNYT8ggPOBiUB9RPxI0inAeyLikmx/TUCzpB8DPwVqgAUR8YWtfaDNzMzMzMzMdjWlnCkwISsKPA5cA8yJiOOBY4AzJPXO+rVGxGnAAuCTwPiIOBaYCXwKuC/b7lpyRY98lwGTgPHAhXmZsyPiRHLFh8OBlcApWe4S4PgiYx4OXERulsk389r/EBEnFxnPBOBX2eyUz0iqAg4Gjs3bXzHDgIsi4ovAlcDnspxKSY2dO0uakp3a0/T66290EWtmZmZmZma2ayjljI1ZEXGWpGpgBvCspG8DVcBIcjMdAB7JbkeSm6mwCSAi2iUdAhwl6fxsu9md9lEREa8DSHqO3EwKyBVJAJYCA4CBwM8kDQCGAo8VGfNfIuKNLG+jpI7Hp2OMhcbzc+Drkq4D7o2I6yRdBfynpNXAN8jN3OigvOUXImJltjwK+KUkgN2yx6wpf3ARMRWYCnDkuw/PzzQzMzMzMzPbJZX8VJSIaFXuX+vfAf4ZeBaYx1v/wG/PbhcDoyVVRkRbdt2JZ4G5EXE9QDYbYq+8+HZJg4E1wIHAXzt2m9dHwEeAP0bETyT9B39fXMg3QlJ/cqey1GTjyB9jofFURsTF2f2nJf0K+K+IuFHS14Azyc0YOSTLeHf++POWFwFfioiXsserV5ExmpmZmZmZmVmmlIWNjlNRaoA7gFeAm4AnyF374u9ERLOka4A5ktaSK4RMBaZK+njW7fvAU3mbfQ24M1u+OiLWZ4WIzmYA10s6Odv340XGvBT4CXAQcGmB9YXGUy/pguw4p5ObbfH7rDDTDpwNrAK+JGk60Az8uUD2V8jNKqnJtvsEXZ/GYmZmZmZmZrbLK0lhIyJeBBoKrJrW1f2I+CXwy059zi+Qc1bWfwa5a3bkZ0zOW744b9XhxUf8N+si4txOeRPzljcUGc9Nne4fW6DP29oiojFveTG5C6mamZmZmZmZWTdtj19F2SFJugvondd0RU+NxczMzMzMzMy2zS5b2IiIQrMjpm/3gZiZmZmZmZnZNivlz72amZmZmZmZmZWUCxtmZmZmZmZmVrZc2DAzMzMzMzOzsrXLXmNjZxDRK3lma3uf5Jm1wKbovcV+W5sp2pNmAlRXvO2XiJPY1F6bPLN1c33yzPqq15Jnwijaojp5al3lG8kzYSQbN++WPLW+MvXjegAVakucmbO2bXDSvNpKaI/y+ag5ZcPdyTOn156cNO/cWMRuVcuSZuaMYEA8mTyzVJa+uXvSvAE1UN0r/WdAr4qNyTMBqirWl0VmW3tN8kyAwbUvJU4cRevm9N+BSvG9CqC9BN8BN0dV8kwozXerUjyupfqsamtP+x0YICL9/5uWNifPBIgy+f/o5fRdpRSvKYDNJXq/3l4UET09BtsGjY2N0dTU1NPDMDMzMzMzMyu5I8YcwWPzH1OhdeVTmrK/szk2snLj4qSZA2pGsqZ1UdJMgL7Vo3hz0zNJM3erOpiWTU8lzayvOpSNm9MXi2p6NbKq9fmkmf2rD+C19Wn/j+3uvYfSHrOTZgJUaHxJjr9Uz9WydSuSZg6tG0Rb+5ykmZUV45L//UPuPWDFhiVJMwfVDi9J5hsbX0yaCTCwZkTy3IE1I7hRo5JmnhuL2NQ+N2kmQFXFWGLDHUkzVTupZM/V46+3JM181+D6knxWpX7/g9x7YCneV1N/B+hbPYqlLWmfJ4C96+vLYqx719fz8to1STMBhvXpyytrVyXN3KtP/+SZHbml+L6W+nEd1qdv8s9/yH0HKMX3ilfXNSfN3KOugeYNLyfNBGioHcby9cuTZg7pPaQkmaV6/stprKV4XaU+/q6Ux9wgMzMzMzMzM7MCXNgwMzMzMzMzs7LlwoaZmZmZmZmZlS0XNszMzMzMzMysbLmwYWZmZmZmZmZly4UNMzMzMzMzMytbO3xhQ1JfSbdLmilpnqQPbMW2P89uL5E0SdIISbckHt8PJfXeiv43SZqVHctxWds0SYelHJeZmZmZmZnZrqCypwfQDecB0yPix5IE9OvuhhHx6VSDkFQREe0F9vH5rYz6aERskrQP8J/A/SnGZ2ZmZmZmZrYr2uFnbADrgKMlDYmcVZImS5ot6SFJx0saIun2jg0k3SepXlJTsVBJjZLuz3K+VKTP5GyGxZ3AiZK+lzdzZHTWZ2a2r8mSbpN0p6RHJO1ZKDMiNmWLfYEn8lZdJOleSb/LCjiFxjNFUpOkpteb3+jqMTMzMzMzMzPbJZRDYeN6YBFwd1bIOAo4BzgWOBH4ekQsB6okDcxmQjRHRMsWcq8EzoyI8cA4SUOK9GuNiNMi4o/AtyJiIvAp4F8L9F0VEacBU4Gziu1Y0v3AvcD0vObZEXEi0AIcXmi7iJgaEY0R0Ti4YeAWDs/MzMzMzMxs57fDn4oSEW3AZcBl2TUpLgUO4a1TOBqy29uAM4EBwM3diD4c+G02OWIAsDewvEC/R/KWvyjpZKAd2Fyg74LsdilwTLEdR8RxkoYDv+et4kb+tgO6MX4zMzMzMzOzXd4OX9jIZmAsi4hW4DWgGngcmBQRIakq63orcB3QG3h/N6IfA86KiNWSepErVhTSno1jEDAJeA+5osjVBfpG/tALHIuAyux0lBZgTXe3NTMzMzMzM7O32+ELG+SKCDdJ2pDdvwA4EpglaTO561RcGBErJAXwckRsKJKV72LgNkkVQCtwBrC+i/4ryc3ouB94aJuOJPd435ONsxfw1W3MMTMzMzMzMzPKoLAREXcAd3RqfpLctTc6953U6X5jdntJXvNZWdujwAlb2Pe0vOV24PQCfSZmi/l9p/P318/oaN8ETCzQPjlv+eKuxmRmZmZmZmZmb9nhCxvbi6QJwLfz2/KKFtuaeQ2wb17TtPxiiZmZmZmZmZm9My5sZCJiFgVmU7zDzI+nzDMzMzMzMzOzv1cOP/dqZmZmZmZmZlaQCxtmZmZmZmZmVrZc2DAzMzMzMzOzsqWI6Okx2DZobGyMpqamnh6GmZmZmZmZWcmNHnMEC+c/pkLrfPHQctW+ilj3+6SRqjudWPWbpJkA6n82uWuzJszUBKLltrSZ9WcSr/wsaSaA9voM7TE7aWaFxpcm88/fT5oJULHfF9m4OW0RrqZXI9F6V9JMAFWfyrMr1yXNPGhAHe0zLkqaWXHCVcSam5NmAqjvh2ndPC9pZnWvo2lfcEnSzIojL6Fl01NJMwHqqw7ljY0vJs0cWDOCTe1zk2ZWVYzlRo1KmglwbixiTeuipJl9q0clf0wh97i+98y3/er7O/LQbecx/7U3k2aO2X03ZixZmTQT4IThA/jezBeSZn5p4v4sbG5Jmjm6oZ7FazYkzQQY2beWl1rS5u5TX5rMJS1pP1MAhtfXMe/VNUkzj96jb8meq1K8By5tSfta3bu+vmTP1dNvpM09ZGAdr6xdlTRzrz79eXVdc9JMgD3qGkoy1uXrlyfNHNJ7SPLXFOReV6V4X3lu1fqkmQAH9u/NC6vT5u7frzevrV+WNLMrPhXFzMzMzMzMzMqWCxtmZmZmZmZmVrZc2DAzMzMzMzOzsuXChpmZmZmZmZmVLRc2zMzMzMzMzKxsubBhZmZmZmZmZmXLhQ0zMzMzMzMzK1s7fGFD0ghJIem47H61pJWSLujGtg8Xaf95gbYrJE3cxjFOljR2K/qfKelZSU2dMrZ4TGZmZmZmZmb2lh2+sJFpAs7Mlk8Ent/WIEkVEfHpbd22UHtETIuIuVsRNQt417aMwczMzMzMzMzeUi6FjZeA4ZIEfAj4LYCkGyTNlPSgpOFZ28ckNUm6DuiTtV0i6VpJdwGHdsyUkDRa0iOSbgcOLrTjbMbIbEn/BXxJ0nmSZkiaL+m8vPxJWd+HJN0i6XFJJxbKjIgVEdFaYNXxku7MxrRngbFMyY6tqfn1NVv3CJqZmZmZmZnthCp7egBbYS5wLNAAzAHqgSkRsU7S6cCnJX0L+AJwDLmixot52y+JiI8B5OojAHwH+CjwHPBAF/veEzghIlol1UXE9ZJqgAeB6zv1HZSNc1/gSuDerTjGVRHxCUn/BJwFXJ2/MiKmAlMBGsfsH1uRa2ZmZmZmZrZTKqfCxq3ATcB12f1ewJWSRgM1wFPkih5LI2IjsFHSn/O2f6RA5pCIWASQf72LAh7Lm2FxkqT/lS0fWKDvkxHRJmkpMKAbx5VvQXa7lFxxxszMzMzMzMy6UC6nohARz5ObIXFL1tSfXGFiPLmZFwKagWHZBUYHAPvlRbQXiF0u6YDsFJcxXew+f9v/DZwOnAy0FBpq3rIKrO/KO9nWzMzMzMzMbJdTTjM2iIgL4W+nkqwChkq6B3g6W79Z0g+Bh4BnyV2boyvfBG4EXsvyuuNW4H5ysytWbs34O2S/vvIN4EBJ9wIf2ZYcMzOz/8/enYfbVdZ3/39/MicESIAQ5iEiQRANeAQRQwJIgUIdKHXWUm2x+lhtf4+1tHZAq1Y7WbV1wPKIExVkKAIFRCAhIgoHAgjIEYyMMgRIwBBCEvL9/bFX6vZ4ThJgnYSTvF/XxbXXXuten/tea++zc86Xe60tSZK0qXveFzaq6k4695voXndas/iZAdp/Ffhqv3Un93ve0zzOB17+TPqvqo8DH19D/vHNumXA7EEy5wBz+q0+rWv7xcDFaxqXJEmSJEkaBoWN9SnJlsB5/Vb/aVXd8BwyP0Dnm1xWu6Gq/vTZ5kmSJEmSpF+xsNGlqh5jkFkWzyHzMwwws0SSJEmSJD13w+bmoZIkSZIkSf1Z2JAkSZIkScNWqmrtrfS809PTU729vRt6GJIkSZIkDbkZ+7+UG66/MQNt8x4bw9TKWs7CZfe2mjll3E48+tSdrWYCbDV2Nx5YurDVzO0mTOHx5X2tZm4xZnrrmatzh+L4hyJzyYpbWs0EmDh6HxYvv73VzEljXjhk79VHlt3daubW43Zpfaxbjd2t9Z9/6HwGDMXnylC8/oueWtBqJsDksdOG5LWqZRe0mplxxw7ZZ9Xpmd5q5luqb8h+Vofitfrlip+0mrn56Be1nrk6dyh+Vtv+N2Di6H24Z8mSVjMBdp44kSdW/rjVzM1G7cu9TzzeauZOm23Reubq3KF4/7f9ewV0frdo+/N68thp3PfE4lYzd9xsUuuZq3PvX/pIq5nbT9h6SDIfevL+VjMBth2//ZCM9cEnH2w1c+r4qUP2szoUv1cO1c/qUJzXoXhfDcZLUSRJkiRJ0rBlYUOSJEmSJA1bFjYkSZIkSdKwZWFDkiRJkiQNWxY2JEmSJEnSsGVhQ5IkSZIkDVsWNiRJkiRJ0rC13gobSXZLsjDJnCS9Sd7UcvZZbeU9w76PSvL6Z9D+95PMS/LDJP/UrJud5J+HbpSSJEmSJG2cRq3n/uZW1fFJJgDfB761psZJRlTVqvUztDUbbCxVdfEzjDq9qr7aZF6eZOdWBihJkiRJ0iZoQ12KMgFYmuSI5o/7a5KcBJDkhCRnJLkQeHWSdzazG65McliScUm+0ez3nSRbdAcnObxp/6MkJzTrTkvy5STfS3JeOqY2z69MclaSkQMNNMmtSb4G/NMaxvu+ZvknSb6ZZH6Stw+UV1UrmrajgMeAR5pNL27GdmOSfQcZy4nNbJfeRxY++oxOuCRJkiRJG6P1XdiYlWQOcBPwFeCqqjoMOBB4XZLxTbvlVXUMMB94FzCzqg4B5gB/CFze7PdV4MR+fXwCOBaYCby/K3NeVb0aWALsCywCjmpy7wYOG2TMOwEfqKr/u4bxrrYd8J6m7/cOdhKS/DnwU+CRqlrarB5dVa8FPgj8wUD7VdUpVdVTVT1bT9lqsHhJkiRJkjYZG+pSlDHAZcBtST4CjAamAds27a5tHqcB81fPcqiqVUn2Bl6e5B3NfvP69TGiqh4GSPJTYIdm/fzm8R5gMrAV8MUkk4HtgRsHGfMdVbWoWd5vkPGutqCqHm/6zmAnoar+Kcm/AucmeUWz+oZ+45MkSZIkSWuxvgsbAFTV8uYP/48B/we4DbgGWF0MWH0viwXAjCSjqmplkhFN26ur6usASUYDO3bFr0qyDfA4sCfwi9XddrUJ8Fbgu1X1+abIMFghovu+GicB7x9gvP97aGs79iRjq+qpqno6yRPAUmDcAOOTJEmSJElrsb4LG6svRRkLXADcB5wB/Bh4on/jqlqY5CvAVU0R4GPAKcApSVZfrvEvwC1du/0VcGGz/NmqenKQyROXAV9PcmTT903rMP6z1zTedfShJIfTOfffq6qbksx+llmSJEmSJG3S1ltho6ruBKYMsOm0NT2vqlOBU/u1eccAOcc37S+jcw+M7owTupZP6to04E06++3b07V82prG26/tKxhAVf098Pf91s2hc/8Qquo24IT++0mSJEmSpN+0QS5Feb5KchHQfUPQTz6Lr3PtznsjnZuJrvZAVb3p2eZJkiRJkqRfZ2GjS1Ud3XLeGXQuXZEkSZIkSUNgfX/dqyRJkiRJUmssbEiSJEmSpGHLwoYkSZIkSRq2UlUbegx6Fnp6eqq3t3dDD0OSJEmSpCE3Y/+XcsP1N2agbd48dJhaVcv45YqftJq5+egXsXj57a1mAkwa80KeerrdIszYkT2tj3XSmBcO2fEPxWv1yLK7W83cetwurWeuzn3oyftbzdx2/PYsWXFLq5kAE0fvw8Jl97aaOWXcTkPy+g/V8S9deUOrmRNGzeDRp+5sNXOrsbuxau7/bTUTYMSsfxmSsQ6HzKHK3Wrsbpye6a1mAryl+lj01IJWMyePnTYkx9/2Zwp0PleG4viH4rP6wScfbDUTYOr4qdz3xOJWM3fcbNKQHH/bmatz2z6vU8dPHbKxDsW/q8PpvTocxjp1/NQhO/7h8F4dytd/OLz/YWg+r4bqM3AwXooiSZIkSZKGLQsbkiRJkiRp2LKwIUmSJEmShi0LG5IkSZIkadiysCFJkiRJkoYtCxuSJEmSJGnYsrAhSZIkSZKGrfVa2EgyL8nUrufvSvLnA7TrbaGvo5K8/rnmtC3JlUnmJPlBkhc36+YkmbihxyZJkiRJ0nCzvmdsnA0c1/X8d4GzhqKjqrq4qs4diuzn6PCqmg38JfBnz2THJCcm6U3S+/DDjw7J4CRJkiRJGk7Wd2HjLJrCRpJJwCTgS82MhUuTbNHdOMm0JJc02z/drDshyTlJLkxybZIdmvXvTPLDZkbEYU279zXbvtFkfD/JLgMNLMlpzYySK5PstoZ+Ppjk6mbGxcuS7J3k8822HyR5Q5LNkpw/UD9VtaJZ3AL4cdemjzV9f36wk1dVp1RVT1X1bLPNVms+05IkSZIkbQLWa2Gjqu4FxibZBngtcB7w2mYGw/nAG/vt8ingvc32UUl6mvWLq+oY4BTg+CRTgHcBM6vqEGBOv5wTm4x/BN7df1xJRgMvAg5p9r97kH62A14DHAy8DfhUVd0K7JVkPPAw8CrgIOCHA52DJFOSXAV8Hriya9M5Td8vSbLlQPtKkiRJkqRftyFuHnounaLG8cAFwJeTXAn8IbBDv7bTgVOTzAFeCezUrJ/fPN4DTAamAfNXz4aoqlWrA5KMBD6VZB7w1wP0sXoWxWeA/5fk34AJg/SzG3BjVa2qqgXA6gLEo8AxdIozU4BD+PWiRXdfC6vqYDqX4Xyia9Pqvu6lM5NFkiRJkiStxYYobJwFnABsA+wF/KKZqfCfQPq17QN+v5lt0UOnEAJQXW0CLABmJBkFkKT7uGYAU6tqJvCxAfpYXfz4dlX9AfAQv7oPSP9+7mz6GZFkGrC42XYVnXtmzAMWAb8FXDNAP6O6xvYY8ETX5v59SZIkSZKktRi1vjusqnuaAsT5dC7X+HCSC4H76cxW6PYXwBeTjAVWAe8cJHNhkq8AVyV5gk4BY7XbgO2TXArcOsiwNge+0xQdVgFvolOc6N/PA0nOo1PIKOBPmk1XAn9ZVbc1l5m8uKqeGqCfqcDpSVY1/fyfQcYjSZIkSZLWwXovbABU1UFdT/cfYHtP87gAOLrf5tO62l0MXNwsnwqcOkiXM9cynsV0Lh9Zl37+GfjnfvtfB2zbLH8T+OYg/dwHzBpg/eyu5TetaaySJEmSJOlXNkhhY0NL8g90bvC52sVV9ckh6OciYHzXqk82RRJJkiRJktSCTbKwUVV/uZ766T/bRJIkSZIktWhD3DxUkiRJkiSpFRY2JEmSJEnSsJWqWnsrPe/09PRUb2/vhh6GJEmSJElDbv+X7cP1192SgbZtkvfY2BgsWfE0V/3isVYzD95hS65/6JetZgLsv+3m/HTxk61m7jlpPBcueLjVzGOmbcPnfvDzVjMB/uSVu3PXkmWtZu46cdyQnNNPXvbTVjMBTjp8Tx5YurDVzO0mTOG2RUtbzQTYa/IEqua2mpnM4u2fmddq5tc/MJMr7lnUaibAoTtPHpLXqu2fqz955e4sempBq5kAk8dOaz138thp3PTwklYzX7LNRF553NdbzQT4wTlv59Gn7mw1c6uxuw3Za3V6prea+Zbqo566sNXMjD2Guv+UVjMBsv2JrLriz1rNHHHop6kn/rvVzGz2OqquaDUTIDl0SD6r2x7rUIyzkztraF6rlZe2mgmQUUdwx2Pt/r6yx5bjWbnqqlYzR404mFp1WauZABlxOLX0O+1mTngNq6rd3ytGZCZPPd3+/zAdO7KH5U9f02rmmJEHtD7WsSN7WLHq6lYzAUaPOKj191VGHN76v1XQ/Hu14pJ2M0cfyZIVt7SauSZeiiJJkiRJkoYtCxuSJEmSJGnYsrAhSZIkSZKGLQsbkiRJkiRp2LKwIUmSJEmShi0LG5IkSZIkadiysCFJkiRJkoatIS9sJNktSSU5tHk+JsmiJO9bh31/OMj6Lw2w7pNJZj/X8a5hLCd2Lc9JMrGFzBOSHNRv3bgkc55rtiRJkiRJm4L1NWOjFziuWX41cPuzDUoyoqre3cqonpkT195kcEl+41xX1WlVdfVzyZUkSZIkaVO2vgobdwG7JAnweuBcgCTfaGY/fD/JLs2630/Sm+RrwGbNupOTfDXJRcA+SXqb9TOSXJvkfOBFA3Wc5L1J3pBksyTLk2yb5MgkJzWzI76R5PIk30myRZIRSb7bjOvSZt17gOnNullN9MeSXJnk800/A2XtlmRekm8DHxxgbCcnObZZ/o8kc4GPDHYSk5zYnJvexY888sxfBUmSJEmSNjLr8x4bVwOHAFOA+5t1J1bVbOAfgXcnGQn8GXAw8KfArl37311VR1fVj7vWfQx4G/AaYKtB+r0SmAkcBFzWLM8E5gF/CFxeVYcBX23Gswp4bTOu84E3VtUXgL6qml1Vc5vcc6rqEOAlSbYcKKtptwPw1qr6x8FOTJIeYHJVzQK+N1i7qjqlqnqqqmfS1lsP1kySJEmSpE3GqPXY19nAGcDXmucjgU8lmQGMBW6hU/S4p6qeAp5K8rOu/a8dIHNqVfUBrJ7FMYBbgH2ARcDHgePpzO74KPBW4OVJ3gGMBuYl2Qz4UjODZFIz7oHMbx7vbdrt3T+r2X5jVS0fJGO1PYDrmuVr1tJWkiRJkiQ11tuMjaq6Hfg+cFazahKdwsRMOjMvAiwEdmpuMDoZeEFXxKoBYh9M8sLmEpf9B+m3gMXAy6vq+8AuwKqm2HAb8NlmJsbBwN8ARwG/aGZj/GczLoDqH921nEGyBht3f3cA+zXLPevQXpIkSZIksZ6/7rWq3l9V9zZPFwPbJ7kUOLzZ/jTwb8APgM/RuTfHmvwNcDpwQZM3mKuAXzbLj/Cr2R+nAEc098W4HPgt4IfAq5NcCLykK6MvydlJXjFIHwNlrZOq6gUeT3IlncKKJEmSJElaB0N+KUpV3Unn8o/udac1i58ZoP1X6dyjonvdyf2e9zSP84GXr8MY/qVr+Y+6lpcB7xhgl9+Y/VFVb+t6Ortr/Zu61g+UdfwA61bve3LX8nsHaydJkiRJkga2Pu+xMeSam3ie12/1n1bVDRtgOL8myRuB93SteqBfUUSSJEmSJD1DG1Vho6oeo2s2xfNJVZ1B5+apkiRJkiSpJev1HhuSJEmSJEltsrAhSZIkSZKGrXS+DVXDTU9PT/X29m7oYUiSJEmSNORe1vMiruv9SQbatlHdY2NTsuzpVfx08ZOtZu45aTw3P/JEq5kAL956Mx5YurDVzO0mTOG+Jxa3mrnjZpNaz1yd2/Z5ffHWm/HoU3e2mrnV2N1af09B5301FK/V/UsfaTUTYPsJW1NLv9NqZia8hlsfXdpq5t5bTRiy9+o7Pvf9VjO/9iev4poHHm8184DttuDBJx9sNRNg6vipPLLs7lYztx63C79c8ZNWMzcf/SKuf+iXa2/4DO2/7eZDMta2P6ug83lVT13YambGHsPpmd5q5luqj1p+UauZABlzNLXiknYzRx/J8qevaTVzzMgDhuz1X7LillYzJ47eh8eX97WaucWY6a2PEzpjHYrXv+3jh845uOOxdn+32GPL8SxdeUOrmRNGzRiy4x+Kz9WheP8vW3l9q5kA40bt33ruuFH7D8nrP1THPxSfK22/p6DzvhqK92rb/66siZeiSJIkSZKkYcvChiRJkiRJGrYsbEiSJEmSpGHLwoYkSZIkSRq2LGxIkiRJkqRhy8KGJEmSJEkatixsSJIkSZKkYesZFTaS7JZkYZI5SXqTvKmtgTTZZ7WV17YkJ3Ytn5bkxS1kHpXk9QOs/+FzzZYkSZIkaVPwbGZszK2q2cAhwIfW1jjJxjIr5MS1NxncQOehqi6uqnOfS64kSZIkSZuy51J0mAAsTXJEksuTXJPkJIAkJyQ5I8mFwKuTvDPJD5NcmeSwJOOSfKPZ7ztJtugOTnJ40/5HSU5o1p2W5MtJvpfkvHRMbZ5fmeSsJCP7DzLJbyf5ULP8iyQHJJme5ItNxueSXJHk0iQ7Ne2+0cxK+X6SXZpZFdObdW9soj/QbyyDZd2a5GvAPw0wthOSvK9Z/qskVyf5d+A3jqNpc2IzU6Z30cMPP4uXTJIkSZKkjcuzKWzMSjIHuAn4CnBVVR0GHAi8Lsn4pt3yqjoGmA+8C5hZVYcAc4A/BC5v9vsqvzkb4hPAscBM4P1dmfOq6tXAEmBfYBFwVJN7N3DYAOP9PvDKJHsANzaZM4F5wDHAoqo6FDip+Q/gxGZWyj8C725mVfRV1eyqOmOQsQyWtRPwgar6v4Od0CTbAUcCrwQ+C2w9ULuqOqWqeqqqZ/I22wwWJ0mSJEnSJmPUs9hnblUdn2QMcBlwW5KPAKOBacC2Tbtrm8dpwPyqWgFQVauS7A28PMk7mv3m9etjRFU9DJDkp8AOzfr5zeM9wGRgK+CLSSYD29MpXPyaqno8yebAbDpFg7cAq4C/Bt4MvD7JIUCAe5pZH59KMgMYC9wyyHnoP5a9+2c12++oqkWDZKy2G3BTVRXw0ySPraW9JEmSJEni2RU2AKiq5UkCfAz4P8BtwDV0/qiHTvEAYAEwI8moqlrZ3GviNuDqqvo6QJLRwI5d8auSbAM8DuwJ/GJ1t11tArwV+G5VfT7Jv3b13d/NwB8BhwNvB3auqnuS3AacWVV/3zWOGcDUqpqZ5DXAcQP0PdBYBsrqPg9rcifw4uZ8vgDYch32kSRJkiRpk/dsChurL0UZC1wA3AecAfwYeKJ/46pamOQrwFVJnqBTCDkFOCXJHzTN/oVfnxnxV8CFzfJnq+rJzt/8v+Ey4OtJjmz6vmmQMV8JHFRVS5LcAqxs1p8PHJbkiub5N4BvAdsnuRS4tSvjiiTfAb48SB8DZZ06SNtfU1UPNP1dDVwPPLIu+0mSJEmStKl7RoWNqroTmDLAptPW9LyqTuU3/8h/xwA5xzftL6Nzz47ujBO6lk/q2rTv4CP+3/ZnA2c3yx/vWl/Anw6wy8wBMrq/Aeb8QcbyG1lV1bOGcZ3WtfwJOvcWkSRJkiRJ6+hZX4ryfJXkImB816pPVtXFG2o8qyWZBXyke11zg1JJkiRJkvQsbXSFjao6ekOPYSBVNZfODUwlSZIkSVJLns3XvUqSJEmSJD0vWNiQJEmSJEnDloUNSZIkSZI0bKXzxSAabnp6eqq3t3dDD0OSJEmSpCG3/8v24frrbslA2za6m4duKh57aiUXLni41cxjpm3DfU8sbjUTYMfNJnHHY0+2mrnHluP56eJ2M/ecNH7Ijn/RUwtazZw8dho3P/JEq5kv3nozvvCju1rNBHjPgbty26KlrWbuNXkCdy1Z1momwK4Tx1ErLmk1M6OP5EvX3t1q5rtfvgs3Pbyk1UyAl2wzsfWfgR03m8QP73+s1cxXbL8lDz15f6uZANuO355HlrX7Wm09bhcWL7+91cxJY17IZXcvajUT4PBdJvPLFT9pNXPz0S9i4bJ7W80EmDJuJ+r+U1rNzPYnUssvajdzzNGcnumtZgK8pfqoRf/VamYmv5nl//57rWaOed+3h+zf1VUX/nGrmSOO+SJz713cauasnSZx1S/a/fwDOHiHLXnq6Xb/59bYkT3c+mi7/1YD7L3VhNY/r7cdvz3XPPB4q5kHbLcFl9z5SKuZAEfutjU3LGz33+sZUyYOyd8AZ/Y92GomwBumT+Xc2x9qNfP1L9x2SI7/op+3//ofvfvW9D74y1Yze6Zu3vr7Hzo/A0Pxc/XEyh+3mrkmXooiSZIkSZKGLQsbkiRJkiRp2LKwIUmSJEmShi0LG5IkSZIkadiysCFJkiRJkoYtCxuSJEmSJGnYel4WNpLslmRhkjlJepO8KckJSQ5qtp3VtJuTZOKGHq8kSZIkSdowRm3oAazB3Ko6PskE4PtVtT90ih4bYjBJRlTVquGWLUmSJEnSxux5OWOjnwnA0iQnJzl2gO0fTXJVkk8AJDkiyeVJrklyUrNuyyTnJ5mb5MwkY/qHNDNB5iU5O8n8JAc1669P8u/AV5Nsk+S/m/xvJBnZzCK5psn+aJIxSS5onl+ZZFz32JP8cTP7ZHV/3wY+mKQnyRXNug8OdCKSnNjMYOl97NFH2jm7kiRJkiQNY8/nwsasJHOAm4CvrKHdxVV1MLB/kt2Bq6rqMOBA4HVJxgMnAhdW1SzgFuDNg2Tt1Gx7LfD3zbrJwL9V1duBk4DPNvnzgdcDvw18tMk+GdgZWNY8n1VVy9Yw9h2At1bVPwKfAo6rqpnAwUmm9m9cVadUVU9V9Wy51dZriJUkSZIkadPwfC5szK2q2cBuwAnA+EHaXdc8zgemAfsl+R4wp3m+LfAC4Nqm3Y+APQbJurmqllfV3cCWzbpFVXVHs7w38JGm4PIGYDvgP4AjknwNOKqqfgbMTXIa8LEkI4Hq6iNdyzdW1fJmeV/g3CZ7Gp0CiSRJkiRJWoPn8z02AKiq5UkC/MblI439gO8BM4BTgM8C7wduA66hU0i4A3g5nSLIgcDtg2Ttk2Q0MBV4rFnXfe+L24Bzq2oeQNN2VFV9oLm85boklwH/UVWrkpwCHAws4leFipcB3x8g+0bg+Kp6rCmGeM8NSZIkSZLW4vlc2Fh9KcpY4AJg9CDtjk3yETozPH6e5GzgDODHwBNNmy8D30zyFuAB4B8GyboP+DawK/DeAbZ/HPhy0x/Ah4BXJTkO2Aw4rdn31CQjgMeB6+kUVs5r7rPx5CB9nwSc0+y3HHjdGtpKkiRJkiSep4WNqroTmLKGJsc37WYPsO9pdAoM/Q1049H+Hqyq4/vl9XQtPwIc12+fXuDf+q2b2e/5EjozRvr7376q6jrg8HUYoyRJkiRJajwvCxtDLcmWwHn9Vv/pBhiKJEmSJEl6DjbJwkZVPQbMHmDT8QOskyRJkiRJz1PP529FkSRJkiRJWiMLG5IkSZIkadiysCFJkiRJkoatVNWGHoOehZ6enurt7d3Qw5AkSZIkaci9rGdvruu9NQNt2yRvHroxeHLlKm5+5IlWM1+89Wbc8diTrWYC7LHleO5ZsqTVzJ0nTuQLP7qr1cz3HLgrX7vpvlYzAd7xkh158MkHW82cOn4qCx5f1mrmtC3G8dZPX9lqJsA3/+wQHll2d6uZW4/bpfXjh845WP70Na1mjhl5AJ+4tK/VzL86YjrXP/TLVjMB9t928yF5rWa+7VutZs77xptYvPz2VjMBJo15IQuX3dtq5pRxO7U+1kljXsg/z7mj1UyAD87eY0iOf9FTC1rNBJg8dhqrrvizVjNHHPppasUlrWZm9JHUov9qNRMgk9/M6ZneauZbqo9VNa/VzBGZSa28tNVMgIw6Ykg+q+uXZ7Wamc2Pbz1zde6KVVe3mjl6xEHUknNazQTIxOO46eF2fwd8yTYTWx9rJh5HPX5mq5kA2eINrX8GZPKbh+b4l1/UaiZAxhxNPXVhu5ljj6GWfqfdzAmvGbrjb/l9lS3eQC1u9/cqgEx605B8Bj719Pr7H/FeiiJJkiRJkoYtCxuSJEmSJGnYsrAhSZIkSZKGLQsbkiRJkiRp2LKwIUmSJEmShi0LG5IkSZIkadiysCFJkiRJkoat9VLYSLJFkvOTzElyTZLfeQb7fql5PDnJsUl2S9L+l4KveQy7JfmtruetfCFvkn9LMr7fuqOSnNxGviRJkiRJG7v1NWPj7cDFVTUbOBCYt647VtW7h2pQz8BuwG+trdGaJPmNc11Vf1pVTz6XXEmSJEmSNmXrq7CxFDggydTqWJzkhCTzkvwgyWFJpiY5f/UOSS5PMnFNsyOS9CS5osn54CBtzkiybZIjk9zSrPuHJK9IMi3JJc1Mkk832/Zt+v5Bkn9vYt4DvLFptyWwWZJvJpmf5O3NfgNlndD0fyHw6gHGNqc5xklJvpvkYuD31nC8JybpTdK76JGH13zGJUmSJEnaBKyvwsbXgT7gkqZg8HLgzcAhdP7g/3BVPQiMTrJVkl2BhVW1ZC25nwKOq6qZwMFJpg7QZh4ws/nvriTbAC8Drmv2f28zk2RUkh7gDuDwqnolsEOSFwJfAM6oqtlV9RiwHZ1ix0zgvV1j6Z8FsLyqjqmq767hOP4QOKuqjgIeGKxRVZ1SVT1V1TN5623WcmokSZIkSdr4jVofnVTVSuATwCeSHAp8FNgbuKJpMqV5PAc4DpgMnLkO0fsC5yah2Wdn4MF+ba4E3gVsC5wGHAGsrKoVSaYDpzb7bw5cBjwB/GuSCcDuwA4D9Lugqh4HSLMzMFAWwLXrcBx7AF9ulq8B9luHfSRJkiRJ2uStl8JGMwPj/qpaDjwEjAFuAo6tqkoyuml6NvA1YDzw2+sQfSNwfFU9lmQksGqANjfTmaFxO53ZGxcAqy956QM+WFV3NQWKkcCngc9V1f8kOQcIsKLZtloN0M9AWW8bZEz93UGnmHEd0AM8vQ77SJIkSZK0yVsvhQ06MyvOSLKsef4+On/Iz03yNPBj4P1V9UiSAu6tqmWDZHU7CTinuTHncuB1wK/djLOqViVZBFxbVfcn2YLOLA6AvwC+mGQsnQLEO+kUPT6d5F38qpjxY+Afmm9j+YNBxjJQ1rr6T+DMJL8H/AK46xnsK0mSJEnSJmt9XYpyAZ2ZEt1upnPvjf5tj+33vKd5PLlr9fHNuuuAw9eh/9/pWn5h1/IC4Oh+ze8G9hkg5pCu5dX3z6CqXrGGrNPWMq7ZXU+f07euSJIkSZK0KVpfMzbWiySzgI90r+tXPNhgkvwDcFDXqour6pMbajySJEmSJG0MNqrCRlXNBWZv6HEMpKr+ckOPQZIkSZKkjc36+rpXSZIkSZKk1lnYkCRJkiRJw1aqBvrmUj3f9fT0VG9v74YehiRJkiRJQ66nZy96e2/LQNs2qntsbEqeWPE01zzweKuZB2y3BZfe9WirmQBH7LoVDyxd2GrmdhOmsODxdflG4HU3bYtx/PD+x1rNBHjF9lty15J2x7rrxHEsW3l9q5njRu3PJXc+0momwJG7bc3NjzzRauaLt96Mu5csbTUTYJeJE1j+9DWtZo4ZeQAX/bzd83r07ltz2d2LWs0EOHyXyTz61J2tZm41drch+Vlt+zMFOp8rjyy7u9XMrcftwuPL+1rN3GLMdG5YuKTVTIAZUyayZMUtrWZOHL0PDz15f6uZANuO35564r9bzcxmrxuSn//l//57rWYCjHnft1lV81rNHJGZnJ7prWa+pfqoh77SaiZAtv0DfnrA3q1m7nnNrdTKS1vNzKgjWn+doPNa1cKvtpqZKb9Prbik1UyAjD6ShcvubTVzyridhuR3oKorWs0ESA5t/bxm9JH8csVPWs3cfPSLWPTUglYzASaPncaKVVe3mjl6xEHcs6TdfwN3njiRWtb/Czyfu4w7dkj+XV256qpWMwFGjTiYJ1b+uNXMzUbtOySfgYPxUhRJkiRJkjRsWdiQJEmSJEnDloUNSZIkSZI0bFnYkCRJkiRJw5aFDUmSJEmSNGxZ2JAkSZIkScPWei1sJJmXZGrX83cl+fMB2vW20NdRSV7/XHMkSZIkSdLz1/qesXE2cFzX898FzhqKjqrq4qo6t628JENyroYqV5IkSZKkTcH6/qP6LJrCRpJJwCTgS0nmJLk0yRbdjZNMS3JJs/3TzboTkpyT5MIk1ybZoVn/ziQ/THJlksOadu9rtn2jyfh+kl0GGliz/V+TXJXkE826k5N8NclFwD5J/irJ3KaPfZOMSXJB17pxSd7TjGtuktcn2S3JWU3euCRzuvr7F+CiZv03klye5Dv9z4MkSZIkSRrYei1sVNW9wNgk2wCvBc4DXltVs4HzgTf22+VTwHub7aOS9DTrF1fVMcApwPFJpgDvAmZW1SHAnH45JzYZ/wi8ew1DvLiqDgb2T7J7s+7uqjq6WZ5eVbOANwAfBXYGljXrZlXVsuYYXt2sO28tp+R/qupI4A+By6vqMOCrwIkDNU5yYpLeJL2LH3lkLdGSJEmSJG38Rm2APs+lU9R4HXAS8OVmFsUkOpeqdJsOnJoEYHPgsmb9/ObxHuBAYBowv6pWAFTVqmYfkowEPpVkBjAWuGUNY7uuK39as3xt8/gi4JWrZ1wAT1fVz5qZGacB9yX5W+BDwD8nGQV8EljWlZ9+/a3O3ht4eZJ3AKOBeQMNrqpOoVPM4UUv3a/WcBySJEmSJG0SNkRh4yzg9KbvvYBfVNXbkrwf2Kpf2z7gg1V1VzqVipHA24DuP+oDLABmJBlVVSv73bdiBjC1qmYmeQ2/fo+P/vYDvtfscwowE1jVbLsNmFtVfwiQZHSSscB/NIWUU4CDgWur6o+SvBL4C+DPgJ2ajJf16687++qq+vrq7DWMUZIkSZIkNdZ7YaOq7mlmM5wP/BD4cJILgfuBe/s1/wvgi00BYRXwzkEyFyb5CnBVkieAj3Vtvg3YPsmlwK1rGd6xST5Cp4Dx89WzPpo+bkpye5K5zVgupVOkObUppDwOXN+MdzdgIvDnVfVYkuuTzONXMzT6OwU4JckfNM//BbhwLWOVJEmSJGmTtyFmbFBVB3U93X+A7T3N4wLg6H6bT+tqdzFwcbN8KnDqIF3OXMeh/XVVLenKP7nfuD5F574fa8r+/f6hVfXeAdbN7lpeBrxjHccoSZIkSZIaG6SwsaEl+Qegu7hy8YYaiyRJkiRJevY2ycJGVf3lAKs/ud4HIkmSJEmSnpP1+nWvkiRJkiRJbbKwIUmSJEmShi0LG5IkSZIkadiysCFJkiRJkoatVNWGHoOehZ6enurt7d3Qw5AkSZIkacj19Eynt7cvA23bJL8VZWNQtZRlK69vNXPcqP259K5HW80EOGLXrXjoyftbzdx2/Pb87icuazXz7L86nNd/7HutZgKc+9ev5qeLn2w1c89J46m6otXM5FDm3be41UyAmTtO4u4lS1vN3GXiBGrpd1rNBMiE13DXkmWtZu46cRxz713cauasnSZRNbfVTIBkFsufvqbVzDEjD+COx9p9/++x5XhuW9Tuewpgr8kTuH/pI61mbj9ha+5ZsqTVzJ0nTmTB4+2+TwGmbTFuSMb64JMPtpoJMHX81CH5DHz0qTtbzdxq7G7c98TiVjMBdtxsErXy0lYzM+oI6qGvtJu57R9weqa3mgnwluqj7vpMq5nZ9QP8fPZLW83cfc6N3PPbL2s1E2Dn/7luSF6rK7Zv/7U69P4+Lrmz3c/VI3fbmoVvObjVzCmnX8V9x/a0mgmw4wW9Q/K+Gorjf/w9h7aaCbDFF67goTe9stXMbb/1A578u2NbzRz/kQv4xWte3momwA7fubb1z4Cd/+e6IftcWfTOQ1rNnPz/rqRWtfv32pp4KYokSZIkSRq2LGxIkiRJkqRhy8KGJEmSJEkatixsSJIkSZKkYcvChiRJkiRJGrYsbEiSJEmSpGFrvRY2kmyR5Pwkc5Jck+R3nsG+X2oeT05ybJLdkpz1DPafkeSAZzNuSZIkSZL0/DRqPff3duDiqvqPJAG2XNcdq+rdz7HvGcBE4JrnmCNJkiRJkp4n1velKEuBA5JMrY7FSU5IMi/JD5IclmRqkvNX75Dk8iQTk/QOFpqkJ8kVTc4HB2n2HuADSS5q9vmrJHOTXJlk32bd9Um+mOSGJCcmOS3J/CRvbrbPSfKvSa5K8olBxnJCkv9OcmHT7u3N8twkY5vj+17T71lJRiZ5cdMmSf4hydue5fmVJEmSJGmTsr4LG18H+oBLmkLGy4E3A4cArwY+XFUPAqOTbJVkV2BhVS1ZS+6ngOOqaiZwcJKpA7T5AvCZqjq6KWRMr6pZwBuAjzZtJgEfB14J/BPwIeBQ4H1dORdX1cHA/kl2H2Q8C6vqGOByYP9m+XpgFrAIOKqqDgHuBg6rqpuBK4EvAntU1TcGCm2KLb1Jehc+vGgtp0SSJEmSpI3fer0UpapWAp8APpHkUDoFhb2BK5omU5rHc4DjgMnAmesQvS9wbufqFiYDOwMPrqH9i4BXJpnTPH+6eVxUVfcAJLmtqh5qlsd27Xtd8zgfmAb8fID8m5rH+4CnupYnA1sBX0wyGdgeuLHZfgpwP53ix4Cq6pSmHS972d61huOTJEmSJGmTsF4LG80MjPurajnwEDCGThHg2KqqJKObpmcDXwPGA7+9DtE3AsdX1WNJRgKrBmizAhjZLN8GzK2qP2zGtbrf7mLBYIWD/YDv0blnxymDtBksJ8Bbge9W1eeT/GuzDuCfgT8DTk7y21Vl4UKSJEmSpLVY3zcP3Rc4I8my5vn76BQK5iZ5Gvgx8P6qeiRJAfdW1bJBsrqdBJyTZASwHHgd8GS/NlcDX0vSU1VvT3J7krl0iiCX0plJsi6OTfIROoWRgWZrrM1lwNeTHAk8AdyU5FhgWVV9oSmy/Cnw6WeRLUmSJEnSJmV9X4pyAXBBv9U307n3Rv+2x/Z73tM8nty1+vhm3XXA4WvpewHwqq7nn6Jzb47f6KNZfsVA64G/XtM9P6rqtK7lL3Yt/3NXs30H2PWCpt1n13QckiRJkiTpV9b3jI31Isks4CPd66pq9nDtR5IkSZIkDWyjLGxU1Vxg9hDkzu73fEj6kSRJkiRJ62Z9f92rJEmSJElSayxsSJIkSZKkYcvChiRJkiRJGrZSVRt6DHoWenp6qre3d0MPQ5IkSZKkIdfTM53e3r4MtG2jvHnopmDpylXcsHDQb519VmZMmchpN9zbaibACTN24poHHm8184DttuCni59sNXPPSeO5+ZEnWs0EePHWmw3Ja7Vi1dWtZo4ecRAX/fyRVjMBjt59a354/2OtZr5i+y1ZuKz99+qUcTvRuSdwe5JZzL13cauZs3aa1Pp7Cjrvq7bP65RxO3HPknbHuvPEiUN2/AseX9Zq5rQtxvH48r5WM7cYM527lrQ7ToBdJ47jiZU/bjVzs1H7ct8Ti1vNBNhxs0lD8rO6ZMUtrWZOHL0Pqy7841YzAUYc80WWP31Nq5ljRh7ATw/Yu9XMPa+5lbrrM61mAmTXD3B6prea+Zbqo1Ze2mpmRh3BqprXaibAiMyk7v18q5nZ6b3U0u+0mgmQCa8Zks/Ap55u93/ujR3Z0/rvVdD53WooPqsWL7+91cxJY17Yeubq3KE4/qH4vaIeP7PVTIBs8YYh+Xd12crrW80EGDdq/yH5WW379V8TL0WRJEmSJEnDloUNSZIkSZI0bFnYkCRJkiRJw5aFDUmSJEmSNGxZ2JAkSZIkScOWhQ1JkiRJkjRsPW8KG0l2S1JJDm2ej0myKMn71mHfHw6y/ksDrPtkktnPdbySJEmSJGnDe94UNhq9wHHN8quBZ/2FyklGVNW7WxlVk9dW1vrMliRJkiRpY/Z8+4P6LmCXJAFeD5wLkOQbSeYk+X6SXZp1v5+kN8nXgM2adScn+WqSi4B9kvQ262ckuTbJ+cCLBuq4mTEyL8nZSeYnOahZf32Sfwe+mmSbJP+d5PJmTCOTHJTkmiRzk3y0mWlyQfP8yiTjmnEd2+T9cZITuvr7NvDBJD1JrmjWfXAoT7IkSZIkSRuLURt6AAO4GjgEmAJcBUwETqyqpUleA7w7yd8CfwYcSKeocWfX/ndX1e8DdOojAHwMeBvwU+DKNfS9E3A4sB3w/+jMGpkM/FtV3ZHkn4HPVtXlSf4vneLLS4GPVtUFzcyL3YFlVXVsklRVdY2jvx2Aw6tqeZLLgOOqalGSc5N8vaoe7G6c5ETgRIDtd9p5DYchSZIkSdKm4fk2YwPgbODTwJzm+UjgU0nmAX9NpxgwBbinqp6qqkeBn3Xtf+0AmVOrqq+qis7lLoO5uaqWV9XdwJbNukVVdUezvDfwkSRzgDfQKYD8B3BEM3PkqKr6GTA3yWnAx5KMBKqrj+4qx41VtbxZ3hc4t8meBvxG5aKqTqmqnqrqmbT1Nms4DEmSJEmSNg3PuxkbVXV7ku8DZ9GZMTGJTmFiZjNj4zhgIbBTkjF0Zmy8oCti1QCxDyZ5IXAHsD9w3iDd75NkNDAVeGyAvNuAc6tqHkDTdlRVfaAZy3XNzIv/qKpVSU4BDgYW8atCxcuA7w+QfSNwfFU91hRDBjoOSZIkSZLU5XlX2ACoqvfD/15KshjYPsmlwK3N9qeT/BvwAzrFhrvWEvk3wOnAQ03eYO4Dvg3sCrx3gO0fB76c5CPN8w8Br0pyHJ0Cy2nNvqc2l6U8DlxPp6ByXnOfjScH6fsk4Jxmv+XA69bQVpIkSZIk8TwqbFTVncDx/dad1ix+ZoD2XwW+2m/dyf2e9zSP84GXr8MwHqyq/mPo6Vp+hF99a8tqvcC/9Vs3s9/zJYP0/799VdV1dO7vIUmSJEmS1tHzprCxPiXZkt+8HOVPN8BQJEmSJEnSc7BJFjaq6jFg9gCbjh9gnSRJkiRJep56Pn4riiRJkiRJ0jqxsCFJkiRJkoYtCxuSJEmSJGnYsrAhSZIkSZKGrVTVhh6DnoWenp7q7e3d0MOQJEmSJGnI9bzshfRed3sG2rZJfivKRqEeo5Zf1GpkxhzNDQuXtJoJMGPKRJY/fU2rmWNGHsBh7z231czLP/96XvXGb7aaCfD9M97KoqcWtJo5eew0qua2mpnMYu69i1vNBJi106Qhef3rqQtbzQTI2GO46hePtZp58A5bctndi1rNPHyXyVRd0WomQHLokLxWtz66tNXMvbeawB2PPdlqJsAeW47nwScfbDVz6vip3LOk3c/VnSdO5K4ly1rNBNh14jjufeLxVjN32mwLHnry/lYzAbYdv33rPwPJoTy+vK/VzC3GTB+yz9X65VmtZmbz46mVl7abOeoIfj77pa1mAuw+58YhGevpmd5q5luqj/8a0W4mwJtX9Q3J8c/dsf2xzrqvj49c/JNWM//uqBdx/hbtjvV3Hu/j0intH/8RC9vPPWJhH5dNbTfz8Af7uHHv9o//pbf2cd2e7ea+7Kd9zN+r3cz9buvjxy9u//j3vbmPq3ZtN/fgu9p//aHzHrh2j3ZzX35HH7XsglYz18RLUSRJkiRJ0rBlYUOSJEmSJA1bFjYkSZIkSdKwZWFDkiRJkiQNWxY2JEmSJEnSsGVhQ5IkSZIkDVutFjaSzEsytev5u5L8+QDtelvo66gkr38G7U9IMua59itJkiRJkp4/2p6xcTZwXNfz3wXa/aL1RlVdXFXnPoNdTgAsbEiSJEmStBFpu7BxFk1hI8kkYBLwpSRzklyaZIvuxkmmJbmk2f7pZt0JSc5JcmGSa5Ps0Kx/Z5IfJrkyyWFNu/c1277RZHw/yS79B5XkIGAGcFGSDyTZJsl/J7m82XdkktlJvpvkvCQ3Jjk+yXeaMWybZLdmRsrZSeY3mb+hGcdnkvwoyYeTfLYZ958329+e5LIk1yd5e7Pu4836Cc0xbDNI9olJepP0Lnz48Wf+6kiSJEmStJFptbBRVfcCY5s/zF8LnAe8tqpmA+cDb+y3y6eA9zbbRyXpadYvrqpjgFOA45NMAd4FzKyqQ4A5/XJObDL+EXj3AOO6GrgBOLqqPgOcBHy2qg4D5gOrL2lJVb0W+Czw1qp6DfAN4HXN9p2ANzfH9vdrOBVnAq8A/gj4f8DBwNuabWdX1eHAQcD7m3UfaY7vP4F/qqqHBwqtqlOqqqeqeqZss8VATSRJkiRJ2qSMGoLMc+n84f86OgWELzezKCbRuVSl23Tg1CQAmwOXNevnN4/3AAcC04D5VbUCoKpWNfuQZCTwqSQzgLHALeswxr2BA5P8LTAe+DrwMHBTs/2+fssvaJZvrqrlwN1JtlxD/k1VVUkeAG5sllc0245I8v81y3s2x7M8yRnA+6vqLeswfkmSJEmSxNAUNs4CTm+y9wJ+UVVvS/J+YKt+bfuAD1bVXelUKkbSmdlQXW0CLABmJBlVVSuTdM80mQFMraqZSV7Dr9/jo9uKJh/gNuDcqpoHkGQ0nVkV3f32HwPAPk3bqcBjazgH/7tvVVW/bX8HHAo8Bfys6X8bOjNBzk7yrqo6dQ3ZkiRJkiSp0Xpho6ruSTKKzqUnPwQ+nORC4H7g3n7N/wL4YpKxwCrgnYNkLkzyFeCqJE8AH+vafBuwfZJLgVvXMLTvAGcmORP4OJ2ZJB9ptn1oHQ/vPuDbwK7Ae9dxn/7OBq6gMytlUbPuP+ici2uAS5J8r6ruepb5kiRJkiRtMoZixgZV1X1jzf0H2N7TPC4Aju63+bSudhcDFzfLpwKDzWSYuQ5j+hzwua5VA83smDNAv2cBJNkNeLCqjl9LP7O7ll/Rtbz6mD9Op7DSvU/3vUdevbZjkSRJkiRJHUNS2NjQkvwDnZtzrnZxVX1yuPYjSZIkSZIGtlEWNqrqL4cg807g+H7rWu9HkiRJkiStu1a/7lWSJEmSJGl9srAhSZIkSZKGLQsbkiRJkiRp2EpVbegx6Fno6emp3t7eDT0MSZIkSZKGXM/+L6D3+p9loG0b5c1DNwn1GLX8olYjM+Zobnp4SauZAC/ZZiIrVl3dauboEQcx+4/ObjVzzpd/l1e98ZutZgJ8/4y38uhTd7aaudXY3aia22pmMou59y5uNRNg1k6TWP70Na1mjhl5ALXiklYzATL6SC67e1GrmYfvMnlIMquuaDUTIDmUpStvaDVzwqgZ3LZoaauZe02ewB2PPdlqJsAeW47noSfvbzVz2/Hbc+8Tj7eaudNmW3D3knbPKcAuEycMyVjbPqfQOa9D8Rm4ZMUtrWZOHL0PV/3isVYzAQ7eYUvql2e1mpnNj2dVzWs1c0Rmcs9vv6zVTICd/+e6IRnrf42Y3mrmm1f1cXrazQR4S/VRqy5rNTMjDueqXdsf68F39fE3F7T7c/X3x+7DGaPaHesbV/ZxxfbtH/+h9/dx6ZR2c49Y2MdlU9vNPPzBPubv1f7x73dbH9ft2W7uy37ax9W7t5t50M/7+PGL2z/+fW/uY97O7ebOvKf99xR03lc/mtZu7oEL+lr/t2pNvBRFkiRJkiQNWxY2JEmSJEnSsGVhQ5IkSZIkDVsWNiRJkiRJ0rBlYUOSJEmSJA1bFjYkSZIkSdKwNWwKG0nmJZna9fxdSf58gHa9LfR1VJLXP4P2JyQZ81z7lSRJkiRJz8ywKWwAZwPHdT3/XWBIvhi3qi6uqnOfwS4nABY2JEmSJElaz4ZTYeMsmsJGkknAJOBLSeYkuTTJFt2Nk0xLckmz/dPNuhOSnJPkwiTXJtmhWf/OJD9McmWSw5p272u2faPJ+H6SXfoPKslBwAzgoiQfSLJNkv9Ocnmz78gks5N8N8l5SW5McnyS7zRj2DbJbs2MlLOTzG8yJUmSJEnSWgybwkZV3QuMTbIN8FrgPOC1VTUbOB94Y79dPgW8t9k+KklPs35xVR0DnAIcn2QK8C5gZlUdAszpl3Nik/GPwLsHGNfVwA3A0VX1GeAk4LNVdRgwH1h9SUuq6rXAZ4G3VtVrgG8Ar2u27wS8uTm2vx/oHCQ5MUlvkt6FDz8+yJmSJEmSJGnTMWpDD+AZOpfOH/6vo1NA+HIzi2ISnUtVuk0HTk0CsDlwWbN+fvN4D3AgMA2YX1UrAKpqVbMPSUYCn0oyAxgL3LIOY9wbODDJ3wLjga8DDwM3Ndvv67f8gmb55qpaDtydZMuBgqvqFDoFGXpe9sJah7FIkiRJkrRRG26FjbOA0+mMey/gF1X1tiTvB7bq17YP+GBV3ZVOpWIk8DaguyAQYAEwI8moqlqZpHsWywxgalXNTPIafv0eH91WNPkAtwHnVtU8gCSjgYP79dt/DAD7NG2nAo+t6SRIkiRJkqSOYVXYqKp7koyic+nJD4EPJ7kQuB+4t1/zvwC+mGQssAp45yCZC5N8BbgqyRPAx7o23wZsn+RS4NY1DO07wJlJzgQ+TmcmyUeabR9ax8O7D/g2sCvw3nXcR5IkSZKkTdqwKmwAVFX3jTX3H2B7T/O4ADi63+bTutpdDFzcLJ8KnDpIlzPXYUyfAz7XtWqgmR1zBuj3LIAkuwEPVtXxa+tLkiRJkiT9yrArbGxoSf4B6C6uXFxVn9xQ45EkSZIkaVNmYeMZqqq/HILMOwFna0iSJEmS9AwNm697lSRJkiRJ6s/ChiRJkiRJGrYsbEiSJEmSpGErVbWhx6Bnoaenp3p7ezf0MCRJkiRJGnI9L3shvdfdnoG2efPQYeqxp1Zy/s8ebjXzd16wDb/94YtazQT4n48fzWV3L2o18/BdJvPTxU+2mrnnpPHcsHBJq5kAM6ZM5K2fvrLVzG/+2SE8+OSDrWZOHT+19XMKnfN6Zl+7Y33D9Kncs6T912rniRNZtvL6VjPHjdqfq37xWKuZB++wJbctWtpqJsBekydw3xOLW83ccbNJ3L2k3bHuMnHCkB3//UsfaTVz+wlbD8k5veaBx1vNBDhguy149Kk7W83cauxurX9WQefzqp7471Yzs9nrqBWXtJs5+kieerr9/wkxdmQPK1Zd3Wrm6BEHUQu/2mpmpvw+9dBXWs0EyLZ/QN37+XYzd3ovtfLSdjNHHUGtuqzVTICMOJzTM73VzLdUH7Xov1rNBMjkNw/Nea257WZmFrXsglYzATLuWGrpd9rNnPAaanm7v69nzNHUUxe2mgmQsce0/r7K5DdTdUW7mTmUeuA/W80EyHZ/2Pr7KuOOHbr36hD8GzgUYx2Ml6JIkiRJkqRhy8KGJEmSJEkatixsSJIkSZKkYcvChiRJkiRJGrYsbEiSJEmSpGHLwoYkSZIkSRq2LGxIkiRJkqRha70WNpLslmRhkjlJepO8KckJSQ5qtp3VtJuTZOL6HNvaJDmxa7mV8a0+9n7rxiWZ81yzJUmSJEnaFIzaAH3Orarjk0wAvl9V+0On6LEBxvJMnAic8mx3TjKiqlZ1r6uq057roCRJkiRJ2pRtyEtRJgBLk5yc5NgBtn80yVVJPgGQ5Igklye5JslJzbotk5yfZG6SM5OM6R+S5L1J3pBksyTLk2yb5MgkJzWzI77R5H4nyRZJRiT5bjMr49Jm3XuA6c26WU30x5JcmeTzTT8DZe2WZF6SbwMfHGBs/3vsSf4jyVzgI4OdsCQnNjNdeh979JFndrYlSZIkSdoIbYjCxqzmUoubgK+sod3FVXUwsH+S3YGrquow4EDgdUnG05lFcWFVzQJuAd48QM6VwEzgIOCyZnkmMA/4Q+DyJverwInNrIrXVtVs4HzgjVX1BaCvqmZX1dwm95yqOgR4SZItB8pq2u0AvLWq/nGwA03SA0xujuN7g7WrqlOqqqeqerbcaus1nDpJkiRJkjYNG/JSlDF0Cg0/GKTddc3jfGAasEOSjwCjm+fbAi8Avty0+xFw8AA5twD7AIuAjwPHAy8CPgq8FXh5knc0ufOSbAZ8KckuwCTg7EHGN795vLdpt3f/rGb7jVW1fJCM1fboOt5r1tJWkiRJkiQ1NtilKM0f+wF+4/KRxn7N4wxgAXAS8H7gUODuZt87gJc37Q4Ebh+gnwIWAy+vqu8DuwCrmv5vAz7bzMQ4GPgb4CjgF81sjP9s+gGo/tFdyxkkC2AVa3dH1/H2rEN7SZIkSZLEBrwUJcnVwEXAY4O0OzbJVcD8qvo5nZkTZwDfAJ5o2ny5aXclnVkZ3xok6yrgl83yI8C1zfIpwOp7d1wO/BbwQ+DVSS4EXtKV0Zfk7CSvGKSPgbLWSVX1Ao83x3HUuu4nSZIkSdKmbr1eilJVdwJT1tDk+Kbd7AH2PQ04bYB9BrrxaP99/6Vr+Y+6lpcB7xhgl/0HyHhb19PZXevf1LV+oKzj1zCuk7uW3ztYO0mSJEmSNLANcY+NIdPcxPO8fqv/tKpu2ADD+TVJ3gi8p2vVA/2KIpIkSZIk6RnaqAobVfUYXbMpnk+q6gw6l9JIkiRJkqSWbLCbh0qSJEmSJD1XFjYkSZIkSdKwZWFDkiRJkiQNW6mqDT0GPQs9PT3V29u7oYchSZIkSdKQ69l/D3qvvyMDbduobh66aVnCqprXauKIzOSqXzzWaibAwTtsySPL7m41c+txu/CK3/5Kq5k//J8/4KDfOa3VTICrzz+BOx57stXMPbYcz8pVV7WaOWrEwVx616OtZgIcsetWLHpqQauZk8dOo5Zd0GomQMYdyzUPPN5q5gHbbcF5dyxsNfO1e0yhnrqw1UyAjD2G5U9f02rmmJEHcOujS1vN3HurCa3/TEHn5+qBpe2+VttNmMJ9TyxuNXPHzSax4PFlrWYCTNti3JAc/0NP3t9qJsC247enVl7aamZGHcHjy/tazdxizPTW3//Q+RmoJee0mpmJx1ErLmk3c/SRXLH99FYzAQ69v49a+p1WMzPhNczdsd2xzrqvj6t2bf/4D76rj1r0X61mZvKbOT3tj/Ut1cfpt7T7GfCWfbbnuj3bHevLftrHj6a1f/wHLujj0int5h6xsG9Ijv8nM9o//hfd0MeNe7eb+9Jb+1hwyEtazZx25U3c8pL2j3+fm/o4b2K7ua9d0v57Cjrvq7Z/Bg5c0Ec98d+tZq6Jl6JIkiRJkqRhy8KGJEmSJEkatixsSJIkSZKkYcvChiRJkiRJGrYsbEiSJEmSpGHLwoYkSZIkSRq2NprCRpItkpyfZE6Sa5L8zjPY90vN48lJjk2yW5KznsH+M5Ic8GzGLUmSJEmSnr1RG3oALXo7cHFV/UeSAFuu645V9e7n2PcMYCJwzXPMkSRJkiRJz8BGM2MDWAockGRqdSxOckKSeUl+kOSwJFOTnL96hySXJ5mYpHew0CQ9Sa5ocj44SLP3AB9IclGzz18lmZvkyiT7NuuuT/LFJDckOTHJaUnmJ3lzs31Okn9NclWSTwwylhOT9CbpXbhw8bM7S5IkSZIkbUQ2phkbXwe2By5JshT4APBm4BBgPHB+VR2eZHSSrYDNgYVVtaQzwWNQnwKOq6pFSc5N8vWqerBfmy8AE6vq35tCxvSqmpVku2bb64FJwMeBR4D7gRcCy4ELgf9qci6uqv8vycVJdq+qn3d3UlWnAKcA9PTsVc/8FEmSJEmStHHZaAobVbUS+ATwiSSHAh8F9gauaJpMaR7PAY4DJgNnrkP0vsC5TfFjMrAz0L+w0e1FwCuTzGmeP908LqqqewCS3FZVDzXLY7v2va55nA9MA36tsCFJkiRJkn7dRlPYSLIrcH9VLQceAsYANwHHVlUlGd00PRv4Gp1ZHL+9DtE3AsdX1WNJRgKrBmizAhjZLN8GzK2qP2zGtbrf7hkWg8222A/4Hp17dpyyDmOTJEmSJGmTttEUNujMrDgjybLm+fvoFArmJnka+DHw/qp6JEkB91bVskGyup0EnJNkBJ1LR14HPNmvzdXA15L0VNXbk9yeZC6dIsildGaSrItjk3yETmHE2RqSJEmSJK3FRlPYqKoLgAv6rb6Zzr03+rc9tt/znubx5K7VxzfrrgMOX0vfC4BXdT3/FJ17c/xGH83yKwZaD/x1VS1ZU1+SJEmSJOlXNprCxvqSZBbwke51VTV7w4xGkiRJkqRNm4WNZ6iq5gKzhyC39UxJkiRJkjZ2Izb0ACRJkiRJkp4tCxuSJEmSJGnYsrAhSZIkSZKGrVTVhh6DnoWenp7q7e3d0MOQJEmSJGnI7f+yvbn+ulsz0DZvHjpMPfrkCr7x41+0mvm2fXfgp4ufbDUTYM9J4/mbC25pNfPvj92Hp8/+g1YzR/7uV1h11V+0mgkw4uBPcdPD7X6L70u2mcjpt9zfauZb9tmezr1x25XM4sIFD7eaecy0bTj39odazQR4/Qu3ZenKG1rNnDBqBsufvqbVzDEjD+CSOx9pNRPgyN22ZvHy21vNnDTmha2/r5JZ3L+0/ePffsLWPPrUna1mbjV2N5asaPfzb+LofVix6upWMwFGjziIRU8taDVz8thpLFx2b6uZAFPG7cQdj7X779UeW44fksyHnmz3sxpg2/HbD8m/K22/VlPG7TRkn1WPL+9rNXOLMdP5yMU/aTXz7456Ueu//0Dnd6BaeWmrmRl1ROu/V0Dnd4vTM73dzOobkt8rL/p5++/Vo3ffmjP7Hmw18w3Tp7b+O9DrX7gt3/rJA61mArzpRdsNyfEPReZ5dyxsNRPgtXtM4bQb2v1cPWHGTq0fPwzd+6rt36vXxEtRJEmSJEnSsGVhQ5IkSZIkDVsWNiRJkiRJ0rBlYUOSJEmSJA1bFjYkSZIkSdKwZWFDkiRJkiQNW+utsJFkXpKpXc/fleTPB2jX20JfRyV5/TNoPyfJxOfab5P1pQHW/XGSE5Jsl+QjzboTkoxpo09JkiRJkjZV63PGxtnAcV3Pfxc4ayg6qqqLq+rcocheh77fvYZtD1TV3zVPTwAsbEiSJEmS9Bysz8LGWTSFjSSTgEnAl5rZEpcm2aK7cZJpSS5ptn+6WXdCknOSXJjk2iQ7NOvfmeSHSa5McljT7n3Ntm80Gd9PsssaxvexZv/Pd/W1OuOoJCc3yz9pMm9O8uYk30pyU5JZzfbe5nGXps//AQ5p1u2W5KwkBwEzgIuSfCDJ3CTjmjb/mOTQ53iuJUmSJEnaJKy3wkZV3QuMTbIN8FrgPOC1VTUbOB94Y79dPgW8t9k+KklPs35xVR0DnAIcn2QK8C5gZlUdAszpl3Nik/GPwKCzKYBzmv1fkmTLNbTbDvjj5hj+hc7Mi7cPkP0h4KNV9dtAdW+oqquBG4Cjq+ozwHeAY5MEeBUwd6COk5yYpDdJ7+OLHlnDECVJkiRJ2jSMWs/9nUunIPA64CTgy80sikl0LlXpNh04tfO3PpsDlzXr5zeP9wAHAtOA+VW1AqCqVjX7kGQk8KkkM4CxwC1rGNvq3Hub8XQXI9K1vKCqliRZCdxeVcuS3AdM7pe3B3Bds3zNGvoF+Cbw78AjwJVVtWqgRlV1Cp2CDtP2eWkN1EaSJEmSpE3J+v5WlLPozHDYBtgL+EUzS+I/+fXiAUAf8PvNbIse4IJmff+CwwJgRpJRAEm6j2kGMLWqZgIfG6CPbv1zFwE7N89fNki7wYofAHcA+zXLPfymFcBI6Nx7g06R6U+Ar69hjJIkSZIkqct6LWxU1T10/oA/H/gh8OokFwIvGaD5XwBfTHI5cCmwwyCZC4GvAFc1bWd3bb4N2D7JpcDhz3C43wMOSnIxndkjz9Q/Aic3+48eYPt3gDOTvKt5fgawe1WtaVaJJEmSJEnqsr4vRaGqDup6uv8A23uaxwXA0f02n9bV7mLg4mb5VODUQbqcuQ5jmt21/KauTYesYXzLaIooVfUwcFS/7XfTuV9Gf8c32z8HfK5r/Srga2sbqyRJkiRJ+pX1XtjY0JL8A9BdXLm4qj65ocYDkOTddG5AesyGHIckSZIkScPNJlfYqKq/3NBj6K+qvgR8aUOPQ5IkSZKk4WZ93zxUkiRJkiSpNRY2JEmSJEnSsGVhQ5IkSZIkDVupqg09Bj0LPT091dvbu6GHIUmSJEnSkOuZMY3eGxZkoG2b3M1DNxr1OLXiklYjM/pIbn10aauZAHtvNYEVq65uNXP0iIM49I/PaTXzii8exyHvOKPVTIArv/ZGHn3qzlYztxq7G1VzW81MZjHvvsWtZgLM3HESy5++ptXMMSMPaP34oXMOLlzwcKuZx0zbhivuWdRq5qE7T6bqilYzAZJDeXx5X6uZW4yZzm2L2v1c2WvyBBY8vqzVTIBpW4zjoSfvbzVz2/Hbc+8Tj7eaudNmW3DPkiWtZgLsPHEi9z2xuNXMHTeb1Po5hc55XbnqqlYzR404mKUrb2g1c8KoGVzzQLuvP8AB221BLWn338BMPI5lK69vNXPcqP1Z+JaDW80EmHL6VTz1dLv/c2fsyB7O32J6q5m/83gfZ4xqNxPgjSv7huR3gOv2bH+sL/tpH39zwS2tZv79sftwetod61uqj2v3aP/4X35HH/N2bjd35j1Dk/njF7d//Pve3H7uvjf3cfXu7WYe9PM++npe1GomwPTen/Cjae2O9cAFfczdsf3XatZ97f8MvPyOPurRb7aauSZeiiJJkiRJkoYtCxuSJEmSJGnYsrAhSZIkSZKGLQsbkiRJkiRp2LKwIUmSJEmShi0LG5IkSZIkadgassJGknlJpnY9f1eSPx+g3XP+vq4kRyV5/TNof0KSMc+13+cqyUlJdt/Q45AkSZIkabgayhkbZwPHdT3/XeCsoeioqi6uqnOfwS4nABu8sFFVn6yqn2/ocUiSJEmSNFwNZWHjLJrCRpJJwCTgS0nmJLk0yRbdjZNMS3JJs/3TzboTkpyT5MIk1ybZoVn/ziQ/THJlksOadu9rtn2jyfh+kl36DyrJQcAM4KIkH0iyTZL/TnJ5s+/IJLOTfDfJeUluTHJ8ku80Y9g2yW7NjJSzk8xvMn9DM47PJPlRkg8n+Wwz7j9vtp+W5MVNfxd39bdvOy+BJEmSJEkbtyErbFTVvcDYJNsArwXOA15bVbOB84E39tvlU8B7m+2jkvQ06xdX1THAKcDxSaYA7wJmVtUhwJx+OSc2Gf8IvHuAcV0N3AAcXVWfAU4CPltVhwHzgdWXtKSqXgt8FnhrVb0G+Abwumb7TsCbm2P7+zWcijOBVwB/BPw/4GDgbQO0G93090HgDwYKSnJikt4kvQsffmwNXUqSJEmStGkYNcT559L5w/91dAoIX25mUUyic6lKt+nAqUkANgcua9bPbx7vAQ4EpgHzq2oFQFWtavYhyUjgU0lmAGOBW9ZhjHsDByb5W2A88HXgYeCmZvt9/ZZf0CzfXFXLgbuTbLmG/JuqqpI8ANzYLK8YoN0NXcc5eaCgqjqFToGHnpftWetwbJIkSZIkbdSGurBxFnB6089ewC+q6m1J3g9s1a9tH/DBqrornUrFSDozG7r/gA+wAJiRZFRVrUzSPetkBjC1qmYmeQ2/fo+PbiuafIDbgHOrah5AktF0ZlV099t/DAD7NG2nAmuaPvG/+1bVmooRA/UhSZIkSZLWYEgLG1V1T5JRdC49+SHw4SQXAvcD9/Zr/hfAF5OMBVYB7xwkc2GSrwBXJXkC+FjX5tuA7ZNcCty6hqF9BzgzyZnAx+nMJPlIs+1D63h49wHfBnYF3ruO+0iSJEmSpBYN9YwNqqr7xpr7D7C9p3lcABzdb/NpXe0uBi5ulk8FTh2ky5nrMKbPAZ/rWjXQzI45A/R7FkCS3YAHq+r4tfQzu2v5FV3Lq4/5hAH6u43Ot7ZIkiRJkqS1GPLCxoaW5B+A7uLKxVX1yeHajyRJkiRJ+pWNvrBRVX85BJl3Asf3W9d6P5IkSZIkac2G7OteJUmSJEmShpqFDUmSJEmSNGxZ2JAkSZIkScOWhQ1JkiRJkjRspao29Bj0LPT09FRvb++GHoYkSZIkSUOuZ/896L3+jgy0baP/VpSNVbGUp55ut7AxdmQP1zzweKuZAAdstwUPLF3YauZ2E6Zw1F/+T6uZF//DbzPzbd9qNRNg3jfexINPPthq5tTxU1m8/PZWMyeNeSGX3vVoq5kAR+y6FctWXt9q5rhR+1MrL201EyCjjqD3wV+2mtkzdXPOu6Pd9/9r95jC8qevaTUTYMzIA1i56qpWM0eNOJjbFi1tNXOvyRNY8PiyVjMBpm0xbkh+Vu9f+kirmdtP2Jq7l7R7TgF2mTiB+55Y3GrmjptNav2cQue81qrLWs3MiMN5fHlfq5lbjJnOJXe2+/oDHLnb1tTjZ7aamS3eQNUV7WbmUO47tqfVTIAdL+hlxaqrW80cPeIgLp0yvdXMIxb2ccX27WYCHHp/H7XsglYzM+5YfjSt/bEeuKCPi37e7s/A0btvzbV7tDvWl9/Rx+lp//jfUu2/Bw69v49vj2k38/eW93Hdnu0f/8t+2tf6++rABX1ctWu7mQff1ceNe7d//C+9tW9I3qtzd2x/rLPu62P+Xu3m7ndbH7XknFYz18RLUSRJkiRJ0rBlYUOSJEmSJA1bFjYkSZIkSdKwZWFDkiRJkiQNWxY2JEmSJEnSsGVhQ5IkSZIkDVvPqrCRZLckC5PMSdKb5E1tDajJPqutvH7ZJw5FriRJkiRJ2jCey4yNuVU1GzgE+NDaGid5PswOsbAhSZIkSdJGpI1iwwRgaZIjklye5JokJwEkOSHJGUkuBF6d5J1JfpjkyiSHJRmX5BvNft9JskV3cJLDm/Y/SnJCs+60JF9O8r0k56VjavP8yiRnJRnZf5BJ3gNMb2aZzErSk+SKJPOSfLBpc3KS05Nc0ozn/zTL53Qdz7eTXNzkTBnohCT5SXNcNyd5c5JvJbkpyaxm+z83+1+TZEaSEU0/05O8OMmFSTJA7onNDJnehQsXPZfXTJIkSZKkjcJzKWzMSjIHuAn4CnBVVR0GHAi8Lsn4pt3yqjoGmA+8C5hZVYcAc4A/BC5v9vsqvzmj4hPAscBM4P1dmfOq6tXAEmBfYBFwVJN7N3BY/8FW1ReAvqqaXVVzgU8Bx1XVTODgJFObprdU1ZHAY8CoZjlJ9my2r6iqo4BTgP8zyLnZDvhj4LXAvwAnAG8H3t1s/9tmtssfAn9eVauA9wD/AXwOeE9V1QDHcEpV9VRVz5QpkwfpWpIkSZKkTceo57Dv3Ko6PskY4DLgtiQfAUYD04Btm3bXNo/TgPlVtQKgqlYl2Rt4eZJ3NPvN69fHiKp6GCDJT4EdmvXzm8d7gMnAVsAXk0wGtgduXIfx7wuc20yMmAzs3Ky/qXm8r9/y6krCdV2PRw+SvaCqliRZCdxeVcuSdGf83yRHAquApwGqakGSR4BHq+rudRi/JEmSJEmbvOdS2ACgqpY3l018jM4MhtuAa4DVl1Ksah4XADOSjKqqlc09N24Drq6qrwMkGQ3s2BW/Ksk2wOPAnsAvVnfb1SbAW4HvVtXnk/xrV9+/Mdyu5RuB46vqsebSlVV0ZofUIO1XZ+7X9fizdejn1zKSbN308wo6xZXPNhtmAcuAHZPsU1W3DJItSZIkSZIaz6WwsfpSlLHABXRmNZwB/Bh4on/jqlqY5CvAVUmeoFMIOQU4JckfNM3+Bej+g/6vgAub5c9W1ZMD3HoCOjNGvt7MgniCX8206K8vydnAPwEnAec0BZblwOvW5aCBcUkuoXPcv7eO+3RbBDwIXAH8ACDJ5sDH6RQ8NgdOS3JkVa18FvmSJEmSJG0ynlVho6ruBAa6ceZpa3peVacCp/Zr844Bco5v2l9G554d3RkndC2f1LVp38FH/L/t39Zv1eH9np88UHZVvQ8gyV7AnKr697X009M8LgNmN8sPA0c1TV4zwG6vah4XDzAuSZIkSZI0gOd8KcrzVZKLgPFdqz5ZVRcP134kSZIkSdJv2mgLG1U12I09n0vmaeujH0mSJEmStG6ey9e9SpIkSZIkbVAWNiRJkiRJ0rBlYUOSJEmSJA1bqaoNPQY9Cz09PdXb27uhhyFJkiRJ0pDr2e8F9M7/WQbattHePHTj90uq5raamMyi98FftpoJ0DN1cx568v5WM7cdvz2veuM3W838/hlvZf8DP99qJsD1P3ov9y99pNXM7SdszS9X/KTVzM1Hv4jzf/Zwq5kAv/OCbViy4pZWMyeO3oeVq65qNRNg1IiDueKeRa1mHrrzZE6/pd33/1v22Z6qK1rNBEgObf28jhpxMDc/8kSrmS/eejPuWrKs1UyAXSeOY+Gye1vNnDJupyH5+b/10aWtZgLsvdWEIRlr25//0Pk3oJZ+p9XMTHjNkHyu3rBwSauZADOmTKQW/VermZn8ZmrFJe1mjj6Sn89+aauZALvPuXFIfge6dMr0VjOPWNjXeubq3KF4/w/VWM/se7DVzDdMn8q8ndsd68x7+rhi+/aP/9D7+zg97ea+pfqYu2O7mbPu6+PaPdo//pff0cePprWbe+CCPn784nYz9725j/l7tX/8+93WxyVbt5t75CN9/NeI9sf65lV9XDip3dxjFvdRj5/ZauaaeCmKJEmSJEkatixsSJIkSZKkYcvChiRJkiRJGrYsbEiSJEmSpGHLwoYkSZIkSRq2LGxIkiRJkqRha70WNpLslqSSHNo8H5NkUZL3rcO+Pxxk/ZcGWPfJJLMHWD8pyRue+cglSZIkSdLz0YaYsdELHNcsvxq4/dkGJRlRVe9+BrtMAixsSJIkSZK0kdgQhY27gF2SBHg9cC5Akm8kmZPk+0l2adb9fpLeJF8DNmvWnZzkq0kuAvZJ0tusn5Hk2iTnAy8apO/3ALOafqYnOSrJvCQ/SPLmJue0JF9u2nwhyd8mmZvks139fy3Jd5NckGRc/06amSlXJzkrya1JjktyTpKbkrxooONNMqFZ3ibJkUm+3N4plyRJkiRp47Sh7rFxNXAIMAW4v1l3YlXNBv4ReHeSkcCfAQcDfwrs2rX/3VV1dFX9uGvdx4C3Aa8Bthqk3y8Ac5t+bgf+FjgceBXwx02fAHOaNvsAN1bVLOCVXUWMe6vqt4AfAG8apK/JwBuBPwE+DBwP/A3w9oGOt6qWAn8OfLlp93/7ByY5sSn09C5c+Ngg3UqSJEmStOkYtYH6PRs4A/ha83wk8KkkM4CxwC10ih73VNVTwFNJfta1/7UDZE6tqj6A1bM41mIb4IXAd7ueT2mWb2oef9G1/ACwZbN8XdfjqwbJv6Wqnk5yH3BzVa1qlic3BZT+x0tVXZ1kB+Ccqnq8f2BVnQKcAtDTM73W4RglSZIkSdqobZAZG1V1O/B94Kxm1SQ6hYmZdGZeBFgI7NTcYHQy8IKuiFUDxD6Y5IXNJS77D9L1CjpFFICHgZ8ARzQzJ2ZU1QOrh9g93K7lNI/7dT12F1x+7TDXkDGD3zxekrwDmAMcnWTqILmSJEmSJKmxoWZsUFXvB+jUIVgMbJ/kUuDWZvvTSf6NzuUet9G5N8ea/A1wOvBQkzeQ+4HxSc4CPgR8HPheklV0CinremPR3ZqxPkXnEpNn6jb6HW+SnYB3AkcALwH+Hfi9Z5EtSZIkSdImY70WNqrqTvoVAqrqtGbxMwO0/yrw1X7rTu73vKd5nA+8fC39rwKO6lq1ALikX5sTupbf1LV8LPxvIeZbVXXBGvq5k+Y4q+o24IRmuZfOt8IAzBxg19nN43VY1JAkSZIkaa022IyNoZZkS+C8fqv/tKpuGI79SJIkSZKk37TRFjaq6jF+NQOizdyT10c/kiRJkiRp7TbU171KkiRJkiQ9ZxY2JEmSJEnSsGVhQ5IkSZIkDVupqg09Bj0LPT091dvbu/aGkiRJkiQNcz37vYDe+T/LQNs22puHbuyKJaxcdVWrmaNGHMwNC5e0mgkwY8pEHnry/lYztx2/Pb/1Fxe2mvndTx3DIe84o9VMgCu/9sYhOf7Hl/e1mrnFmOnMvXdxq5kAs3aaxPKnr2k1c8zIA6iVl7aaCZBRR7R+DmbtNImLfv5Iq5lH7741q2peq5kAIzKTp55ut2A6dmQPP138ZKuZe04az4LHl7WaCTBti3E8+OSDrWZOHT+VB5YubDVzuwlTuO+Jxa1mAuy42STuX9rue3X7CVu3fk6hc17b/hkYkZksWXFLq5kTR+/DhQsebjUT4Jhp21BLzmk1MxOP45crftJq5uajX8TCtxzcaibAlNOvYvHy21vNnDTmhVw2dXqrmYc/2Nd65urcWn5Rq5kZczTX7dn+WF/20z7Ovf2hVjNf/8Jtmbdzu2OdeU8f3x7T/vH/3vI+5u7Ybu6s+/o4Pe1mvqX6+NG09o//wAV9XLtHu7kvv6OPcye0m/n6pX38+MXtH/++N7d/Xg9c0McV27c/1kPvH5qx1uNntpq5Jl6KIkmSJEmShi0LG5IkSZIkadiysCFJkiRJkoYtCxuSJEmSJGnYsrAhSZIkSZKGLQsbkiRJkiRp2HreFTaS7JZkYZI5SXqTvKnl7LPayuuXfeJQ5EqSJEmSpME97wobjblVNRs4BPjQ2honeT4ch4UNSZIkSZLWs1EbegBrMQFYmuQI4C+BicA5VfXJJCcARzfrPpNkJzrFheXAycAPgP8EdgCWAG/rDk5yOPBxIMAXquq0JKcBK4DdgSeA1wHbAt8ExgAPAW+sqqf7Zb0HmJ5kDvB3zb7/ROf8nldV/5zkZGBPYGvgKeAS4DXAE1V1XHM8xwCbA+OA36uqhc/l5EmSJEmStLF7Psx0GMispkhwE/AV4KqqOgw4EHhdkvFNu+VVdQwwH3gXMLOqDgHmAH8IXN7s91V+c0bFJ4BjgZnA+7sy51XVq+kUQ/YFFgFHNbl3A4f1H2xVfQHoq6rZVTUX+BRwXFXNBA5OMrVpektVHQk8BoxqlpNkz2b7iqo6CjgF+D/9+0lyYnN5Tu/DCxevw2mUJEmSJGnj9nydsTG3qo5PMga4DLgtyUeA0cA0OrMoAK5tHqcB86tqBUBVrUqyN/DyJO9o9pvXr48RVfUwQJKf0pnZAZ0iCcA9wGRgK+CLSSYD2wM3rsP49wXOTUKTsXOz/qbm8b5+y5Ob5eu6Ho/uH1pVp9ApevCynr1qHcYhSZIkSdJG7fla2ACgqpanUx34GJ0ZDLcB19C5fARgVfO4AJiRZFRVrWzuuXEbcHVVfR0gyWhgx674VUm2AR6nc4nIL1Z329UmwFuB71bV55P8a1ffvzHcruUbgeOr6rEkI5txHtuvTf9+APbrevzZIP1IkiRJkqTG87WwsfpSlLHABXRmNZwB/JjO/St+TVUtTPIV4KokT9AphJwCnJLkD5pm/wLc0rXbXwEXNsufraonmxkW/V0GfD3JkU3fNw3UCOhLcjade2ucBJzTFFiW07lXx7oYl+QSOsf9e+u4jyRJkiRJm6znXWGjqu4Epgyw6bQ1Pa+qU4FT+7V5xwA5xzftL6Nzz47ujBO6lk/q2rTv4CP+3/Zv67fq8H7PTx4ou6reB5BkL2BOVf372vqSJEmSJEkdz7vCxnCQ5CJgfNeqT1bVxRtqPJIkSZIkbaosbDwLVfUbN/ZsIfO0tjMlSZIkSdrYPV+/7lWSJEmSJGmtLGxIkiRJkqRhy8KGJEmSJEkatixsSJIkSZKkYStVtaHHoGehp6enent7N/QwJEmSJEkacj37v4De63+Wgbb5rSjDVLGElauuajVz1IiDuf6hX7aaCbD/tpvzyLK7W83cetwu/PaHL2o1838+fjSv/sB5rWYCfO8zr+WeJUtazdx54kRq1WWtZmbE4dywsN1xAsyYMpGqua1mJrOoFZe0mgmQ0Ucy777FrWbO3HESP7z/sVYzX7H9lq2fU+ic1yUrbmk1c+Lofbh7ydJWM3eZOIEFjy9rNRNg2hbjeOjJ+1vN3Hb89ixcdm+rmVPG7cQDSxe2mgmw3YQpQ3L8Dz75YKuZAFPHT+Wpp9st7o8d2cOylde3mjlu1P6c2df+8b9h+lRqebv/BmbM0Sx6akGrmZPHTuPx9xzaaibAFl+4gsXLb281c9KYF3Lj3tNbzXzprX3M36vdTID9buujnrqw1cyMPYafzGh/rC+6oY9v/eSBVjPf9KLt+PGL2x3rvjf3cd2e7R//y37ax7V7tJv78jv6+NG0djMPXNDH6Wn/+N9SfZwxqt3cN67sG5LX/6LJ7R//0Yva/wzY77Y+rt69/bEe9POhGWv98qxWM9fES1EkSZIkSdKwZWFDkiRJkiQNWxY2JEmSJEnSsGVhQ5IkSZIkDVsWNiRJkiRJ0rBlYUOSJEmSJA1bG7ywkWS3JAuTzEnSm+RNSU5IclCz7aym3ZwkE1vqc1KSN7SR9Vwl+dKGHoMkSZIkScPVqA09gMbcqjo+yQTg+1W1P3SKHkPU3yTgDcCZQ5S/zqrq3Rt6DJIkSZIkDVfPl8LGahOApUlOBnqBm/tt/2iSA+kUQv4qyRHAXwITgXOq6pNJtgS+AWwBPAi8raqW98t5DzAryRzg3cDuwIeBkcDnquq/kpwGrABeCPwEuB84HLixqt7fjHEasB2wHDi+qpZ1d9IUZv4LuA/YG/hr4G3AHsAbq+onSXqrqqerv92BJ4DXVVU981MoSZIkSdKmY4NfitJYXWS4CfjKGtpdXFUHA/sn2R24qqoOAw4EXpdkPHAicGFVzQJuAd48QM4X6BRHZgO3A39Lp2jxKuCPk4xs2s1p2uxDp6AxC3hlknHN9nur6reAHwBvGmTMk4E3An9Cp3hyPPA3wNsHaDuvql4NLAH27b8xyYnN5Tq9Dy9cPEh3kiRJkiRtOp4vhY3VRYbdgBOA8YO0u655nE9ntsR+Sb4HzGmebwu8ALi2afcjOrMj1mQbOrMyvgtc3jyf0my7qXn8RdfyA8CW/cZzXdPvQG6pqqfpzNq4uapWNcuTB2g7v3m8Z6DtVXVKVfVUVc82Uyat5bAkSZIkSdr4PV8KGwA0l4wEGDNIk/2axxnAAuAk4P3AocDdzb53AC9v2h1IZ0ZGfyvoXHYC8DCdS02OaIorM6rqgdVD6h5e13L6jWc/4GeDHdZaMgZrO9B2SZIkSZLU5flS2JjVfOvJ1cBFwGODtDs2yVXA/Kr6OXA2cAade2o80bT5ctPuSjqXkHxrgJz7gfHNN67sBnwc+F6SK4BvPoNx75bkUjqXsAzUjyRJkiRJGkIb/OahVXUnv7r0YyDHN+1mD7DvacBpA+xz7Fr6XAUc1bVqAXBJvzYndC2/qWv5WIAkAN+qqgvW0M+dXeO/jc5lNlRVL52bo1JVPQP0d9Kaxi9JkiRJkjo2eGFjqDXfknJev9V/WlU3DMd+JEmSJEnSr2z0hY2qegyYPQS5J6+PfiRJkiRJ0uCeL/fYkCRJkiRJesYsbEiSJEmSpGHLwoYkSZIkSRq2UlUbegx6Fnp6eqq3t3dDD0OSJEmSpCHXs98L6J3/swy0baO/eejGqljKU0+3W9gYO7KHefctbjUTYOaOk3joyftbzdx2/Pb89ocvajXzfz5+NK/+QP8vtnnuvveZ13Lro0tbzdx7qwnUskG/afhZybhj6X3wl61mAvRM3ZzlT1/TauaYkQdQKy5Ze8NnKKOP5If3P9Zq5iu235KrftFu5sE7bMnKVVe1mgkwasTBLFlxS6uZE0fvw11LlrWauevEcdzx2JOtZgLsseX4IfmsevDJB1vNnDp+Kvc9sbjVTIAdN5vE/UsfaTVz+wlbt3780DkHQ/G5smzl9a1mjhu1P+fe/lCrmQCvf+G21FMXtpqZscewYtXVrWaOHnEQD73pla1mAmz7rR9QNbfVzGQW1+05vdXMl/20r/XM1bm16L9azczkN3Pj3u2P9aW39nFmX7ufAW+YPpUfv7jdse57cx8/mtb+8R+4oP3cAxf0ce0e7Wa+/I4+zhjV/vG/cWUfp6fd3LdUH7ftv1ermXtdf9uQHX/bP1cvvbWPq3dvf6wH/XxoxlqPn9lq5pp4KYokSZIkSRq2LGxIkiRJkqRhy8KGJEmSJEkatixsSJIkSZKkYcvChiRJkiRJGrYsbEiSJEmSpGFroytsJNktycIkc5L0JnlTkhOSHNRsO6tpNyfJxJb6nJTkDW1kSZIkSZKkdTdqQw9giMytquOTTAC+X1X7Q6foMUT9TQLeAKy/L+qVJEmSJEkbbWFjtQnA0iQnA73Azf22fzTJgXQKIX+V5AjgL4GJwDlV9ckkWwLfALYAHgTeVlXL++W8B5iVZA7wbmB34MPASOBzVfVfSU4DVgAvBH4C3A8cDtxYVe9vxjgN2A5YDhxfVctaOxOSJEmSJG2ENrpLURqriww3AV9ZQ7uLq+pgYP8kuwNXVdVhwIHA65KMB04ELqyqWcAtwJsHyPkCneLIbOB24G/pFC1eBfxxkpFNuzlNm33oFDRmAa9MMq7Zfm9V/RbwA+BN/TtJcmJzeU3vwoWL1vVcSJIkSZK00dpYCxuriwy7AScA4wdpd13zOJ/ObIn9knwPmNM83xZ4AXBt0+5HwB5r6XsbOrMyvgtc3jyf0my7qXn8RdfyA8CW/cZzXdPvr6mqU6qqp6p6pkyZvJZhSJIkSZK08dtYCxsANJeMBBgzSJP9mscZwALgJOD9wKHA3c2+dwAvb9odSGdGRn8r6Fx2AvAwnUtNjmiKKzOq6oHVQ+oeXtdy+o1nP+Bngx+ZJEmSJEmCjbewMav51pOrgYuAxwZpd2ySq4D5VfVz4GzgDDr31HiiafPlpt2VdC4h+dYAOfcD45tvXNkN+DjwvSRXAN98BuPeLcmldC5hGagfSZIkSZLUZaO7eWhV3cmvLv0YyPFNu9kD7HsacNoA+xy7lj5XAUd1rVoAXNKvzQldy2/qWj4WIAnAt6rqgjX1JUmSJEmSfmWjK2wMteZbUs7rt/pPq+qGDTAcSZIkSZI2aRY2nqGqegyYPQS5J7edKUmSJEnSxm5jvceGJEmSJEnaBFjYkCRJkiRJw5aFDUmSJEmSNGxZ2JAkSZIkScNWqmpDj0HPQk9PT/X29m7oYUiSJEmSNOR69ptG7/wFGWib34oybC1hVc1rNXFEZnLboqWtZgLsNXkCi55a0Grm5LHTOPZvLm4184K/P4pXf6D/N/k+d9/7zGt5ZNndrWZuPW4XfrniJ61mbj76RfQ++MtWMwF6pm4+JO/VWnVZq5kAGXE4Fy54uNXMY6Ztw9x7F7eaOWunSVTNbTUTIJk1JO+ru5YsazVz14njWs9cnfvgkw+2mjl1/NRhkTlUuVPHT+WhJ+9vNRNg2/Hb89TT7Rb3x47sYenKG1rNnDBqRuufKdD5XKml32k1MxNewz1LlrSaufPEiTz5d8e2mgkw/iMXDMlY5+81vdXM/W7r4+rd280EOOjnfVRd0WpmcigLDnlJq5kA0668iTP72v1cecP0qa2f14N+3sdVu7b/Wh18Vx8/fnG7ufve3Me5E9rNfP3S9scJnbHetv9erWbudf1tnJ52x/qW6uPOw2a0mgmw2+U3DMnrf92e7b9WL/tpHzfu3W7uS2/toxb9V6uZa+KlKJIkSZIkadiysCFJkiRJkoYtCxuSJEmSJGnYsrAhSZIkSZKGLQsbkiRJkiRp2LKwIUmSJEmShq3nXNhIsluShUnmJOlN8qY2BtaVfVZbef2yTxyK3GcqyZc29BgkSZIkSRqu2pqxMbeqZgOHAB9aW+Mkz4eZIs+LwkZVvXtDj0GSJEmSpOFqVMt5E4ClSY4A/hKYCJxTVZ9McgJwdLPuM0l2olNcWA6cDPwA+E9gB2AJ8Lbu4CSHAx8HAnyhqk5LchqwAtgdeAJ4HbAt8E1gDPAQ8Maqerpf1nuA6UnmAH/X7PtPdM7HeVX1z0lOBvYEtgaeAi4BXgM8UVXHNcdzDLA5MA74vapa2P+EJPkJcB0woxn/a4G94f9n76zD7aiuNv5bEZIAMSwEd3eCS4K7u7sWd4pT3IprKa4FihaKU6yFFGtLCS3UP6C0tKWUQoHs7493DWfucO69M+fOSbhkvc9znjNnzsyaPVuXb/ZPKT1lZmNTSqP8eXMDUwFTAGunlD4u0NrT64xZZhnRtAECgUAgEAgEAoFAIBCYlFCX58RoVxK8BlwDPJtSWhVYBtjYzAb5df9LKa0HvAzsBqyUUloZeBLYHXjc77uOr3pUnAasD6wEHJCj+XRKaXWkDFkY+AdSCqwM/BFYtVjYlNJlwLiU0piU0lPAmcCmKaWVgBXMLNMa/CqltBbwL6CfH5uZzeP/f5ZSWhu4EvhWJ3UzPbA3UmicC+wM7AA089QYl1JaB3gaWL1Jua9MKY1KKY2adtphnTwuEAgEAoFAIBAIBAKBSQd1eWw8lVLa3MwmAx4D3jCzk4D+wBzIiwLgRf+eA3g5pfQZQEppvJktACxlZjv6fU8XntEnpfQ3ADN7E3l2gJQkAH8ChiOPh8vNbDgwEni1RPkXBn5oZjiNmf38a/79l8LxcD/+ee57nU5ov51S+sjMPgd+k1L6xMzyNPIovksgEAgEAoFAIBAIBAKBLlBrKEpK6X8m7cApyIPhDeAFFD4CMN6/3wYWM7N+KaXPPefGG8DzKaUbAMysPzBjjvx4M5sG+BCFiPxf9tjcNQZsBzycUrrUzM7LPfsrxc0dvwpsnlL6l5n19XKuX7im+ByAxXPfb5V4TjManV3bWbkDgUAgEAgEAoFAIBAIOOpSbGShKAOA+5FXw23AL1D+ig5IKb1vZtcAz5rZf5Ai5ErgSjPbxS87F/hV7rZvAw/48YUppf+6h0URjwE3mNla/uzXml0EjDOzO1FujaOAu1zB8j+Uq6MMBprZj9F7b1HynkAgEAgEAoFAIBAIBAI1oceKjZTS74Fpm/x1bVe/U0pXA1cXrtmxCZ3N/frHUM6OPI2dc8dH5f5auPMSf3n99oVTqxV+n9iMdkppPwAzmw94MqV0cTfPGeXfnwBj/PhvwNqF//PPu7y78gcCgUAgEAgEAoFAIBCof1eUryXM7EFgUO7UGSmlh3rrcwKBQCAQCAQCgUAgEAgIk4Riw3caqZvmtRPiOYFAIBAIBAKBQCAQCAQ6R13bvQYCgUAgEAgEAoFAIBAITHCEYiMQCAQCgUAgEAgEAoFAr0UoNgKBQCAQCAQCgUAgEAj0WlhKaWKXIdACRo0alcaOHTuxixEIBAKBQCAQCAQCgUDbMWrxORn78lvW7L9JInnoNxHj08d8/PkrtdKcvN9iPPi7v9dKE2Cd2afmvf++VyvNEYNGsPHJj9RK8+7j12Dto39UK02Ah05flxfe/bBWmktPP4T04e210rQhW/LTd/5VK02AZUcO5aPPflUrzSn7L0ga/1itNAGsz2q89Nd/10pziekG8+z/1VuvK8wwlM/GP18rTYD+fZZrS1u9/eEntdKcY8hA3vznf2ulCTDPsEFtmave+bjeeXXk5FPzp48+qpUmwMxTTsmf/1PvXDXTFENqr1NQvdY9Bvr3WY5PPn+pVpoD+y3RtnU1/e/BWmnaZOuQPrm/XpoD1+f/NlyqVpoAM9z7YlvWwF8sNG+tNBf+5bjaaWZ007vfq5WmTb87v1qk/rIu+No47vnt+7XS3GiuaRk3av5aac479te8ukD977/o6+N4eb566S7+Rv39auFfjuPB4fW//zr/GMdt/eqlu9Xn4/j9qovVSnO2x1/hZqv//bdN43pVX31z6QVqpTnPC6/XPld3hQhFCQQCgUAgEAgEAoFAINBrEYqNQCAQCAQCgUAgEAgEAr0WodgIBAKBQCAQCAQCgUAg0GsRio1AIBAIBAKBQCAQCAQCvRah2AgEAoFAIBAIBAKBQCDQaxGKjUAgEAgEAoFAIBAIBAK9FhNNsWFms5nZ+2b2pJmNNbOtzWxnM1vO/7vDr3vSzKas6ZnDzGzLOmgFAoFAIBAIBAKBQCAQmPjoN5Gf/1RKaXMzmxx4JqW0BEjp0abnDQO2BCbchrqBQCAQCAQCgUAgEAgE2oavSyjK5MDHZnaima3f5P+TzexZMzsNwMzWMLPHzewFMzvKzw01s/vM7Ckzu93MJmtCZx9gtHuBzGtma5vZ02b2nJlt43SuNbOr/JrLzOx4p3mh/3+imV1vZg+b2f1mNrD4EPc4ed7M7jCz181sUzO7y8xeM7P5/Zob/RnPmNksZja5H09jZmuZ2VX1VG0gEAgEAoFAIBAIBALfXExsxcZoM3sSeA24povrHkoprQAsYWazA8+mlFYFlgE2NrNBwJ7AAyml0cCvgG2a0LkMeYmMAX4DHA+sBqwI7G1mff26J/2aBYFXnebyOSXGn1NKawLPAVt3UubhwFbA/sAxwObAccAO/v+e/oyzgL1SSh8DhwNX+XWHFgma2Z4etjP2b3/7ZxfVFQgEAoFAIBAIBAKBwKSBia3YyJQMswE7A4M6ue7n/v0yMAewuJk9Cjzpv6cD5gRe9Ot+BszVzbOnAeYGHgYe99/T+n+v+ff/5Y7fBYYWyvNzf24z/Cql9AXwF+CXKaXxfjzcFShnmtnTwLHADAAppef9+L6U0odFgimlK1NKo1JKo6aZZlg3rxcIBAKBQCAQCAQCgcA3HxNbsQFASul/gAHNwkcAFvfvxYC3gaOAA4BVgD/6vb8FlvLrlkEeGUV8BmReGX8Dfg2s4cqVxVJK72ZFyhcvd2yF8iwOvNXZa3VBYzFgREppJeCUjK6Z7YiUNeuY2YhO6AYCgUAgEAgEAoFAIBBwTOzkoVkoygDgfqB/J9etb2YnIQ+P35nZncBtwC+A//g1VwE3mdm2yLvi9CZ03gEG+Y4rRwCnAo+a2XjgfZRYtAxmM7NHgE9RiElVvAGMdBqvA5jZTMCuwBrAIsDFwBYt0A4EAoFAIBAIBAKBQGCSwURTbKSUfk8j9KMZNvfrxjS591rg2ib3NEs8mr9vPLB27tTbwI8L1+ycO946d7w+gJkB3JpSur+L5/w+V/43UJgNKaWxwFi/bKUmt47x758TSo1AIBAIBAKBQCAQCAS6xcT22GgbzGwocE/h9EEppVd643MCgUAgEAgEAoFAIBAIfBXfWMVGSulfNDwg6qR74oR4TiAQCAQCgUAgEAgEAoHu8bVIHhoIBAKBQCAQCAQCgUAg0ApCsREIBAKBQCAQCAQCgUCg1yIUG4FAIBAIBAKBQCAQCAR6LSylNLHLEGgBo0aNSmPHju3+wkAgEAgEAoFAIBAIBHo5Ri0xJ2Nfesua/feNTR76zce/SeMfq5Wi9VmNX/79P7XSBFho6in4x6dv10pz+IA5WHKFy2ul+fNn92apMVfWShPgxSf35I8ffVwrzVmmnJxPPn+pVpoD+y3BY3/8R600AVabZTifflGvEm5A31Gk/z1YK00Am2wdnvhTvXWwyszDuee379dKc6O5piV9fG+tNAFs8g353xcv1Epzsr5L8/oH9fb/BaaavPYxBRpX7/33vVppjhg0oi00//DRJ7XSBJh1yoH8/ZM/1kpz6oGz8P4nf66VJsC0A2dqyxr44f/G1UpzyGTzMva9f9dKE2DUiMGkD2+vlaYN2ZKPPvtVrTSn7L8gf1p3yVppAsz8o5/zn89/USvNKfotzLOzzlsrzRX+MI6nZ66XJsBKfxpH+uT+WmnawPW5Z8r6y7rRR+O49pV654CdF5uJn81Rb1mXeXscL85V//sv9dtx/Hjqeumu9fdxbXn/l+er//0Xf2Mcry5QL91FXx/HLxaql+bCvxzHuFHz10oTYN6xv+Zmq7es26ZxPDCs/rZa75/jeGJkvXRXeWcc6d931EqzK0QoSiAQCAQCgUAgEAgEAoFei1BsBAKBQCAQCAQCgUAgEOi1CMVGIBAIBAKBQCAQCAQCgV6LUGwEAoFAIBAIBAKBQCAQ6LUIxUYgEAgEAoFAIBAIBAKBXotQbAQCgUAgEAgEAoFAIBDotZhoig0zG2Jm95nZk2b2gpltUOHeK/z7RDNb38xmM7PSe8mY2WJmtnQr5a4TXo59JnY5AoFAIBAIBAKBQCAQ6K3oNxGfvQPwUErpEjMzYGjZG1NKe/Xw2YsBUwIv9JBOj5BSegV4ZWKWIRAIBAKBQCAQCAQCgd6MiRmK8jGwtJmNSMI/zWxnM3vazJ4zs1XNbISZ3ZfdYGaPm9mUZja2M6JmNsrMnnA6h3Vy2T7AgWb2oN/zbTN7ysx+YmYL+7mXzOxyM3vFzPY0s2vN7GUz28b/f9LMzjOzZ83stE7KsrOZ3W1mD/h1O/jxU2Y2wMzGmNk5ueddZmY/M7OjW6rRQCAQCAQCgUAgEAgEJjFMTMXGDcA44MeuyFgK2AZYGVgdOCal9B7Q38ymMrNZgfdTSh91Q/dMYNOU0krACmY2osk1lwEXpJTWcUXGvCml0cCWwMl+zTDgVGB54GzgCGAVYL8cnYdSSisAS5jZ7J2U5/2U0nrA48ASfvwSMLpw3TDgDGA5YOtmhFzBMtbMxr7//j87e/9AIBAIBAKBQCAQCAQmGUy0UJSU0ufAacBpZrYKUigsADzhl0zr33cBmwLDgdtLkF4Y+KGiWxgOzAy818X18wPLm9mT/vsL//5HSulPAGb2Rkrpr348IHfvz/37ZWAO4HdN6L/m338BPs0dDy+U6x8ppT/4M/7brKAppSuBKwFGjZo3dfFOgUAgEAgEAoFAIBAITBKYaIoN98B4J6X0P+CvwGRICbB+SimZWX+/9E7gemAQsG4J0q8Cm6eU/mVmfYHxTa75DOjrx28AT6WUdvdyZc/NKw46UyIsDjyKcnZc2ck1ndGxLq4LBAKBQCAQCAQCgUAgUAITM3nowsBtZvaJ/94PKQqeMrMvgF8AB6SU/m5mCfhzSumTTmjlcRRwl5n1Af4HbAwUPSCeB643s1EppR3M7Ddm9hRSgjyCPEnKYH0zOwkpRpp5awQCgUAgEAgEAoFAIBBoIyZmKMr9wP2F079EuTeK165f+D3Kv0/Mnd7cz/0cWK2bZ78NrJj7fSbKzfGVZ/jxss3OA8d2lfMjpXRt7vjy3PE5ucue7Op5gUAgEAgEAoFAIBAIBDrHxPTYmCAws9HASflzKaUxvfU5gUAgEAgEAoFAIBAIBBr4xis2UkpPAWPaQHdM4XdbnhMIBAKBQCAQCAQCgUCgc0zM7V4DgUAgEAgEAoFAIBAIBHqEUGwEAoFAIBAIBAKBQCAQ6LUIxUYgEAgEAoFAIBAIBAKBXotQbAQCgUAgEAgEAoFAIBDotbCU0sQuQ6AFjBo1Ko0dO3ZiFyMQCAQCgUAgEAgEAoG2Y9TiczD25bet2X/f+F1Rvrn4N2n8Y7VStD6r8cr7H9VKE2Cxaafk75/8sVaaUw+chRU2v7FWms/esT1LLHNprTQBXvrZvrzz8d9rpTly8qn56LNf1Upzyv4L8sDbf6uVJsB6c0zDx5+/UivNyfstxvj0dK00AfrYSjzyhw9qpbnGrFNx+7j3aqW55bwjah//oDngs/HP10qzf5/leP2Dj2ulucBUk/OHjz6plSbArFMO5P1P/lwrzWkHztSW8f/mP/9bK02AeYYN4t2P36+V5vSTT8tf//tOrTQBphs0kvTpA7XStAHr8e/Pfl0rzcH95+eFdz+slSbA0tMPIf3z1lpp2rCt+Xz8s7XS7NdnBf607pK10gSY+Uc/55PPX6qV5sB+S/DYiHlrpbnae+N4ZNp6aQKs8f440if310rTBq7ftrK2Yw18asZ6yzr6L+Nqp5nRvaVPvXS3GT+OJ0bWS3OVd8bx/Oz1v/9yv6uf7nK/G8fP56mX5pJvjuPVBep//0VfH8cDw+qlu94/x3Gz1V/WbdM4fjBZvXS3+N+42teqrhChKIFAIBAIBAKBQCAQCAR6LUKxEQgEAoFAIBAIBAKBQKDXIhQbgUAgEAgEAoFAIBAIBHotQrERCAQCgUAgEAgEAoFAoNciFBuBQCAQCAQCgUAgEAgEei1CsREIBAKBQCAQCAQCgUCg1+Ibodgws9nMLJnZKv57MjP7h5ntV+Len3Zy/oom584wszFNzg8zsy2rlzwQCAQCgUAgEAgEAoFAT/CNUGw4xgKb+vHqwG9aJWRmfVJKe1W4ZRgQio1AIBAIBAKBQCAQCAQmMPpN7ALUiD8As5iZAZsAPwQwsxuBmdC7bptS+qOZ7QTsD7wOTOHXnQjMDkwHHGFm16SURpnZYsBVwLvAeOChJs/eBxhtZk8CezmdY4C+wEUppVvM7FrgM2Bu4NfAO8BqwKsppQP8+XMA0wP/AzZPKX2Sf4iZ7QnsCTDLLNP1oKoCgUAgEAgEAoFAIBD4ZuCb5LEB8DywMjAtUhwA7JlSGgOcBexlZn2Bg4EVgIOAWXP3/zGltE5K6Re5c6cA2wMbAlN18tzLgKf8Ob8BjkdKixWBvf2ZAE/6NQsihcZoYHkzG+j//zmltCbwHLB18SEppStTSqNSSqOmnXZY97URCAQCgUAgEAgEAoHANxzfJI8NgDuB24Dr/Xdf4Ez3uhgA/AopPf6UUvoU+NTM3srd/2ITmiNSSuMAzGxsiTJMg7wyHs79ntaPX/Pv/8sdvwsM9eOf575XLPGsQCAQCAQCgUAgEAgEJml8ozw2Ukq/AZ4B7vBTw5BiYiXkeWHA+8BMnmB0ODBnjsT4JmTfM7O5PcRliU4e/RlSogD8DYWarOHeGYullN7Nipgvbu7Y/Hvx3Hde4RIIBAKBQCAQCAQCgUCgCb5pHhuklA4AkB6CfwIjzewRlE+DlNIXZnY+Cvd4A+Xm6ArHATcDf3V6zfAOMMjM7gCOAE4FHjWz8UiRUjax6Gxe1k+BzUveEwgEAoFAIBAIBAKBwCSLb4RiI6X0ewqKgJTStX54QZPrrwOuK5w7sfB7lH+/DCzVzfPHA2vnTr0N/Lhwzc65461zx+vDl4qYW1NK93f1rEAgEAgEAoFAIBAIBAINfCMUGxMSZjYUuKdw+qCU0isToTiBQCAQCAQCgUAgEAhM0gjFRkWklP4FjGkD3RPrphkIBAKBQCAQCAQCgcA3Hd+o5KGBQCAQCAQCgUAgEAgEJi2EYiMQCAQCgUAgEAgEAoFAr0UoNgKBQCAQCAQCgUAgEAj0WlhKaWKXIdACRo0alcaOHTuxixEIBAKBQCAQCAQCgUDbMWqJORn70lvW7L9IHtpr8W9SeqpWimajef2Dj2ulCbDAVJPz/id/rpXmtANnYuUdb6uV5k+u34ql17i6VpoALzyyG+98/PdaaY6cfGr+/dmva6U5uP/8/Pj39ZYTYK3Zpubjz1+plebk/RZjfHq6VpoAfWwlHvnDB7XSXGPWqbjnt+/XSnOjuaZt2/v/74sXaqU5Wd+lefOf/62V5jzDBvHHj+qfq2aZsj1z1bsf19v+008+Lb/9V711CjDX0EG899/3aqU5YtAI/vrfd2qlCTDdoJGkz37c/YUVYP3Xasu8+sK7H9ZKE2Dp6YeQ/n1HrTRt8Ob85/Nf1Epzin4L849dV66VJsDw7/+ED/83rlaaQyablxfnmrdWmkv9dhw/m6NemgDLvD2uLf2/XWX94W/+WivNTeaeri1t9fJ89b//4m+M44Fh9dJd75/196tl3m7f+7+6QL10F329PTTfXHqBWmkCzPPC6zwxst6yrvLOOH4wWf1ttcX/xnGz1Ut32zSu9rWqK0QoSiAQCAQCgUAgEAgEAoFei1BsBAKBQCAQCAQCgUAgEOi1CMVGIBAIBAKBQCAQCAQCgV6LUGwEAoFAIBAIBAKBQCAQ6LUIxUYgEAgEAoFAIBAIBAKBXotQbAQCgUAgEAgEAoFAIBDotZgoig0zm83Mkpmt4r8nM7N/mNl+Je79aSfnr2hy7gwzG9Pk/DAz27J6yetHs3IHAoFAIBAIBAKBQCAQKIeJ6bExFtjUj1cHftMqITPrk1Laq8Itw4CvhWKjYrkDgUAgEAgEAoFAIBAI5NBvIj77D8AsZmbAJsAPAczsRmAmL9u2KaU/mtlOwP7A68AUft2JwOzAdMARZnZNSmmUmS0GXAW8C4wHHmry7H2A0Wb2JLCX0zkG6AtclFK6xcyuBT4D5gZ+DbwDrAa8mlI6wJ8/BzA98D9g85TSJ/mHmNlswC3AX4AFgGOB7YG5gK1SSr82s7Fe7ux5swP/ATZOKaXKtRoIBAKBQCAQCAQCgcAkhImdY+N5YGVgWqQ4ANgzpTQGOAvYy8z6AgcDKwAHAbPm7v9jSmmdlNIvcudOQcqDDYGpOnnuZcBT/pzfAMcjpcWKwN7+TIAn/ZoFkUJjNLC8mQ30//+cUloTeA7YupNnDQe2QoqZY4DNgeOAHZpc+3RKaXXgI2Dh4p9mtqeZjTWzse+//69OHhcIBAKBQCAQCAQCgcCkg4npsQFwJ3AbcL3/7guc6V4XA4BfIaXHn1JKnwKfmtlbuftfbEJzREppHICZjS1RhmmQV8bDud/T+vFr/v1/ueN3gaF+/PPc94qd0P9VSukLM/sL8MuU0ng/Ht7k2pf9+0/N/k8pXQlcCTBq1LzhzREIBAKBQCAQCAQCgUkeE9VjI6X0G+AZ4A4/NQwpJlZCnhcGvA/M5AlGhwNz5kiMb0L2PTOb20Nclujk0Z8hJQrA31CoyRrunbFYSundrIj54uaOzb8Xz33nFS4dXrMbGp1d2+z/QCAQCAQCgUAgEAgEAjlMbI8NUkoHAEgPwT+BkWb2CMqngXs7nI/CPd5AuTm6wnHAzcBfnV4zvAMMMrM7gCOAU4FHzWw8UqSUTSw6m5f1UxRiEggEAoFAIBAIBAKBQGACYqIoNlJKv6egCEgpXeuHFzS5/jrgusK5Ewu/R/n3y8BS3Tx/PLB27tTbwI8L1+ycO946d7w+fKmIuTWldH8Xz/k9/p4ppTeAnf14LNoVJl/u/POO6qr8gUAgEAgEAoFAIBAIBISJ7rHRbpjZUOCewumDUkqv9MbnBAKBQCAQCAQCgUAgEGjgG6/YSCn9CxjTBronTojnBAKBQCAQCAQCgUAgEOgcE3u710AgEAgEAoFAIBAIBAKBlhGKjUAgEAgEAoFAIBAIBAK9FqHYCAQCgUAgEAgEAoFAINBrYSmliV2GQAsYNWpUGjt27MQuRiAQCAQCgUAgEAgEAm3HqCXmZOxLb1mz/77xyUO/qfj4s/GMfe/ftdIcNWJw7TQzui/9tV66S0w3mJ++869aaS47cmjb3v+1v31UK81FppmSP3z0Sa00Z51yIG9/WC9NgDmGDOSXf/9PrTQXmnoK3vznf2ulCTDPsEG8/sHHtdJcYKrJ29JW73z891ppAoycfGr+/J8Pa6U50xRDeO+/79VKc8SgEbXTbBfddtH863/fqZUmwHSDRtZOtx00M7offfarWmlO2X9B/vfFC7XSnKzv0vzn81/UShNgin4L8+kX9Ro3BvQdxfj0dK00+9hKpPGP1UoTwPqsRkpP1UvTRpM+ub9emgPXJ/37jlppAtjgzdtT1v/cXStNAJtiYz7+/JVaaU7ebzHSBzfVStOm2o700V210gSwKTclfXh7vTSHbNkemu3qq+0o6z9uqZfm8G1qLye0p15t8Oakf95aK00AG7Z1W8p6s81bK80P+Eun/0UoSiAQCAQCgUAgEAgEAoFei1BsBAKBQCAQCAQCgUAgEOi1CMVGIBAIBAKBQCAQCAQCgV6LUGwEAoFAIBAIBAKBQCAQ6LUIxUYgEAgEAoFAIBAIBAKBXotQbAQCgUAgEAgEAoFAIBDotQjFRiAQCAQCgUAgEAgEAoFei4mu2DCzMWZ2Tov3nmFmY0pe+6SZTdnCM3Y2s8n8eEYzG2tm13dx7X5+3HSDeTPbycyeNrOfmtnZfq7lOggEAoFAIBAIBAKBQGBSxkRXbHydYWZ9gJ2ByfzUysBNKaUde0D25pTSSimlZYElzWzmHhYzEAgEAoFAIBAIBAKBSRZfG8WGmW1rZj/zz9p+7nAze9zMfm5ma/i5xczsRTO7D5i/E1pzmdnz7qVxZe6vU8zsJ2Z2qV831MzuM7OnzOx2M5vMvSfuN7N7geOAxYAHzewE4ARgXzM7xszW8LK9YGZHlX3PlNJn/ux+wL+Av/tfC5nZPWb2qpkt3Ml77ekeI2P/8cHfyj4yEAgEAoFAIBAIBAKBbyz6TewCOPoCRwFLAYOAx4GHgEtSSmeb2TTAD4BHgFOA7YE3gZ90Qm808qy42L0uMtyVUjrIzJ4xs6HAnsADKaXLXXGxDfAHYAgwOqWUzGwVYP2U0kdm9gdgSqc7eUppVTMz4Hkzu6Dsy5rZ4cA+wOMppY9Fgv4ppbVdgbMLcEjxvpTSlcCVAAssukQq+7xAIBAIBAKBQCAQCAS+qfi6eGwMBv6QUvo0pfRP4FP3aNjOzJ4G7gRm8GtHpJTGpZQS0DSPBXA7MJPnwtg+d/5l//4zMAyYE3jRz/0MmMuPxzr9rrC4mT0KPAnMAUzX7Vs6UkpnA3MD05nZsn76Ff/+EzC8LK1AIBAIBAKBQCAQCAQmZXxdFBv/BmY1swFmNgyYLKX0OXAYsAqwee7a98xsbveUWKITep+nlI7yXBhH5bw28soKA36LvEQAlgF+48fjc9d9hjxKijgKOMDL90en1y3MbABASukL4D/Ax52ULRAIBAKBQCAQCAQCgUA3+LqEonwBnEEjtOQY/34CeBp5U3zo544Dbgb+CvyzE3ob+u4kA4CHUkrjPdyjiKuAm8xsW+Bd4HRg+cI19wK3m9ntXs4MdwK3Ab9ACoqyOMLMVkN1/2hK6bWyO7sEAoFAIBAIBAKBQCAQ6IiJrthIKT2JwjlACov8f3s3uf5lGl4WndG8DSkd8ufG5I63zv21fuH2fHlIKV0EXNTkGdcC1xZOX5v7f1QnZfsO8J3CuS+fmVJ6A+3EEggEAoFAIBAIBAKBQKAbTHTFRk/hSUDvKZw+KKX0ykQoTgeY2VYoSWiGdwtKlUAgEAgEAoFAIBAIBAI9QK9XbKSU/gWMmdjlaIZmniOBQCAQCAQCgUAgEAgE6sPXJXloIBAIBAKBQCAQCAQCgUBlhGIjEAgEAoFAIBAIBAKBQK9FKDYCgUAgEAgEAoFAIBAI9FpYSmlilyHQAszsfeAPJS+fBvhbzUXoLTTbRbe30GwX3Um9rPH+8f6T8vu3i25vodkuur2FZrvo9haa7aI7qZc13j/ef1J+/3bR7S00q9CdNaU0bbM/QrExCcDMxna2/ew3nWa76PYWmu2iO6mXNd4/3n9Sfv920e0tNNtFt7fQbBfd3kKzXXQn9bLG+8f7T8rv3y66vYVmXXQjFCUQCAQCgUAgEAgEAoFAr0UoNgKBQCAQCAQCgUAgEAj0WoRiY9LAlZMwzXbR7S0020V3Ui9rvH970FvKOqm/f7vo9haa7aLbW2i2i25vodkuupN6WeP924PedWhY9gAAjKRJREFUUtZJ/f3bRbe30KyFbuTYCAQCgUAgEAgEAoFAINBrER4bgUAgEAgEAoFAIBAIBHotQrHxDYGZ2cQuQ2DiIdo/EH0gEAgEAoFAIDCpIhQb3wCY2bQppWRm0Z6TKKL9J13B3sxWgy/7wCRZB4FAuxBjKhAIBCY95Of+WAd6DyZpQeibADNbEXjKzOZMKY3/ugq3ZraemS1Rd/na9b5m1jd3XNuEZmbzmNl6NdLbysweBqi7/c2sX120nF6f3PGgGumuAhLsa6TZrn41l5ntXSO9kcB2ZrYfhHKjLnxd59FAc5jZQm2guYyZrVTnmGrH2PQ5ZZm66bYDbXr/2sdqb6GZoz1HzfRqF+gKNGupizyfVjfayAPUWrcFvmqSW/vb+f55nrIO/rI3tU9vVuoE89aLYWZ9U0rPAGcAl5jZ7F9H5YaZTQYsDKwJLFDnIEkpjfdnbFcXTaf7hQmz1sXYej3MCaxjZuv0vJSQUroNeMXMrvLftbS/mc0NbGxm0/SUltPr62UzM/sWsLmZTV4D3amAc/PKgp62lZn1yfWrlXtYxDzdAcDUwBxmtkcN9PqklN4BzgSWMrNd4Ou7ALeLZjuUpbm+upOZzWtmQ2ug26uYg3aV18wGtoHsjmb2g5ppTonW1WXqWAN8Dkx+PLKOAvqcMgrYzMyWrIlm24TF3PvPUwe9wroyfx004ct11OqYp+Erc8r8db2/0+4HnGlm36mJXp9cO01Wl8EgR3N7YIQftzymvO2/MLM+ZraamS1YRzkz5NprjZrpJqc7pKd1W+hXA4Ba5tZ2zQF10y28/8xAbUq4rG+a2RF18OuFcVULX+20am+rfFmhXqPhhMDXSgAOVEM2qQOzA/8B7jOzuXoi3JrZTC4sYmYbmbwsZmu1jGbWL6X0P+B8pNzYA5i3BiYxr02cGljav3uEQrlWB64zs6lqWIAWAjYFXgPuA1Yzs3V7QjOHD4D5zexB6LlywxfIAcAKwEp11Guur14FjAZWAdYzsyl6UM7+KaUPgPWBPczsQH9Wy0JIYaE8GtjXzDZptYw5uvMC+wPvAHcAM5vZXj2hmSlfgP1QH1jNpDTKntnyGMstwHUxCnMA0/nxdmZ2upnNaVL29QSDc0qoWcxsRE/LmrU/cC8wD3AUsJOZTdkqzXa9v5nNlglJTncTq0nArasPmNniZja9H+8IXGZmK9VQxC+tdSmlI4B/mNnFNdA0p/kY8H3gYjMb3VPlRk4Iuw44zswONbPpelLOlNKnwN1oXtnKzBZtlZ7T7JMr5znOA6zdE5oZ3dzxbkAtCuPcunI7sJ+ZXWJSRLY6/+fv64/m1ME9XE8tN6fcBhwBfM9qMMZ4e30ObAksbGYH9JBeXlF0BXClma3eEx7AOnq/9gNWRHxgjwSmzPgEXA/si3iA/XLPapkHyP1cCFivyfme0t0V2NjPt9xXc/3qB8ClwMlmtnQPy9kvNwdcZmYHmtlMPaHpdItzyxZmtkIPy5m9/y3A1cDeZrZwT8uaHP5zKPDfwrMrtVm+rGZ2A/WO/6xOrzKzQ8xsY+uBsqNQ1mvN7EIzW6gGXmVec89KM9vRzLY297auG6HY6P04DBiYUtoMKQ9uNbN5ckJPaZjZcGA7YBMz2xU4BtgE2NXMFm+lcCmlz30xuwX4NTAY2AqYrxV6OboZ0z1dSunvwHhg2Z7QLNDtk1J6BAk3mUDSk/GSgI2A5YFfAI8Cq1oPNcFmdibQN6W0IvAbM7sHOgi9VWiZyZL4BPAhUkIsDYzpCWOTw85Av5TSlogZXx5Y21r03EgpfWZm/YFTgR8Dh1lOudEizWyhvBUxtj8FljCzzVuhl8NfkLfOZsD/AQ8AM5rZnq0Qy/qimR0OTJlSOhg4C1jEzPbxd6lcBya39r38eA8k1B3ic0Mr5TQv637Anma2JaqD/yJmdClr0YLvffX7JgXB1miOOcfnrlbo5cf3nMBPU0rHAAsArwOftUjXgL2BfWp+/ymBDYH1zWxfYDdgfjR/txzuZmZz19kHkBLzepOCcHXgEeAoM1uz1TJmyCm1tgX+gcbU91ul50xdtgbsjtr+h8hzI1NuVFoHCkzwMcBY4DzgYNxy3UNsCSwJDEPKjVGtEsrNfzcCfwWGA7uZ2Qw9KWCOUZ4TKWEX6wm9QhvsCDyH6nMDYGgP5v/Mmr4t8C3EI0/dynqap+mHFwLvppR2AfYE9jezMa3SddpZuTYEfou8LI/qAb1MWXCg07sBWBsZOVoSwnOC14quhDkIGGRmLfGAZrZwrv1PAD53/vciYE4zO8Sf2yMewMyWAJYCpjGzAT3pA3m6/nNa3Lsiz3NWpJe93ynAr9Aa+zxS8LTsEeY8uwHfQXPqdCjctUfe1rn3Pw+NqxHAuq3wwGY2tfclgOOAN4Ctne7oVuUVp515alxqZqciw9naJgX9HP4upfuWmU2fq9OTgWeBQ5EitkdGs1ydno6U239EyrgdekAzK+txwG+QMXZbYEmT0bMyvG+vAOzuPPpOSGG0rfNEtSIUG70M9lVN3AfANAAppe+hTnilmQ2oOgmllP4B/NzpbQzsllLKJo1NrEIcb+HZUwJ/TymdgjT1gxBjW1mzavIoyYS6rYBDTJ4PVyLBqaUJ3cxmzR1vDlxrZoOBKZAipifKAksp/QpNamsjj4XXaCg3Nq1Ar9j+nwOfevkOAIaZ2bVVy+n3p6TQhquQFeTfiMEdhRaLSsqNJmV9A02UpJTuBf6HhJ7F/fpWFs1lAEspfduPdzWzE6oSKTAVQ5AQ+13gYqSQW9PMNmiBrpmUZB8hJeSMaJH4C3A/MFPGiJWk1xc69MWfAH/wxfM14A/IyljJwpKVE439Oc3sbCSE3uaXHN+iYDvQy3osYuS2AK5PKZ0MvAisAyxvUlBVgvfVO4FzEVOzuj9nM/OwnLKwjp46J6E+v5iZ3Y/G7W+Bm6yiC6nJPXYw8vr41MtZ1/t/BDyN+upqwJkppdP83LLWujJuGDCXmZ1DD/qAM8KzoHH0fWSp/nFK6WakhNvZzDZupYBmNrO5J6HJG2pvFJK5OfAvk7W5Ks1ZgbVcCBsErARc4nV6GLKyrVxlHTAPP3GaqwD/At5F/XR/4GMz26xiORc1s+FOd3qk0DoB+DbwFrClmS1VkWa2phriC/+IvCC2QG3336pzSp6uY3UkfK6H5tNDTLm3lq9IM/Mq6GNmW6B5ZTCy2O4PfGBmR1Slmfs5EldooPX/LDM728yOtwrCZ5P17zfAKmY2f0rpDeA6oCWlvpmNNLNp/Xgp1D9PRX1hlJkdVpFe/r22QoqXn7jX0g+QcmtYRYFuiayMaHweb2ZHAkcj3qJyOIqP0QW8/acG3gZGmNkKKaW3gCvQ+jVrl4Sa087XwUrALsi7ZHbgbDPb3XnOqnTz73ecmV2O5v2VzWxTr6epys4r1tEDZn5gJjQGUkrpLuDvVcvotPbPjcXDkFL3HDRXfYoU8gu0QDf//vsCs6H5/xLgZf9dhd7SaB3N+sN8iJf8F1JCT4aMcYMr0s14q6yPH43G1PeA5dAafpSZLVKB5lKIlwBYBCmf3vS+ui8aE5UF+0Jf3QlYELgvpXQH8CTqE1Vp7pCTRbZHY+Belyt/idp/VFU+3cxmRP3zHiRfrgecl1K6Ao3XNarya90hFBu9CNbR7Wg7M1sWWb/eMLk2Lwd8ARyZUvq07CJkZquY4h5JKT2KGOM3US6IoUgA+zMlB4t1jCUeihaxmcxs85TSF8gLYiTwcYXXx6QtXAhZPn+ABsu1aBFeEzEiM/m1VRiQQTQYrfOAF9CCeQBibDa1FlzmzMwylzafgN9Gk+VqNJQbz6KJuAy9fPuvbGZzIYXONGa2mzP4v/RzlcuaHaeUrkHKjVtpKDeWQv2hlDu+dYx/PcCFrP9D3g8Hmyxiizj9dfy53fbXJszin5AAumxK6f+Q5nonM5uh7AScMcp+PAJZ038HZBamPyNl3DJWwXU81/7jzWwxJDAegWuqUX08Avy+JL18+x9nyv8xIxpHW5jZTshye1JK6c9ly+nIFBCvoVCZwcD7KaWfIBfXPwHnWwXvGjNbCzGES6aUPkYWoDeB5cxsaErpViSITZVSKu0NYWYr+Xy3EppLbkGeCnOklP6ArI37WoX8KAVL9Z+9bL9EbfZ34Brg+ymlv1Uo5xqI0boYOAkJ3r8EVuzh+69lHk+fUnoZKUhfRMrH2VCfegl5iZWGmS3oa8gvkTJjCC32ATNbFbgJMW8X+LveAGxvZiNSSk+hOq3sCWBSAq0E/M3k7fI7JDQu4evLFcCiprCXKhiMFIWzAIa8tfYwWWsfAl5F83Zp5CzgFzr9X6K6/DOa+6+i4jqIwqP6mtnAlNK7yMNuuPfN55G30exlifk8lfX/zEI/3o/vQn3pFrQWloZ1DGsYkuQBuSEybjyPrMELAP+sQjdXp5d7Oe9ECq2+aBxcjizYZcuZn1fXB0amlG5MKR2LPAHfQ4LYg1WUWjmaJ/lcdDdSkl5lZjsg4aFq22eeWmsD//F6+BOy1k6dUvodWv93sZIeUdZRqTu7j9ULkQfkrMh7czAyUJUt40zAJ8BtZvY06u+bIGWOoTX/HH9eWT61T0rpDyml28zsIGAN4EF/3/XMbIwrjA73daA0CnUwG/BySml/1Fd/hgTnt1E9V6HbN/9+KaWTU0p7o/VgOBoPmwBzl6RnqZH/bVOkgLwajaEDvJ+tQWueYNellJ5zGlcgj63NvPyXozr4b1cEmpQ3n1diAPA4qse1UX/6EHkcTl6WX0spvZBSusTkrZwQXzEU9am/Ib71rpTSvyuWMxuvN5rZ/sAGzrd8D3gkpbRVSmlPNyCVxbiU0v2m8Lt3kRLiTDNb0NfuXdD4LY2sr/rx9IjH+BniK+ZBxs61zGy6snUKkFK6IaV0l5ntnFK6Aa2Fm5vZYDdGvIr4gSrKzbXROn8TWldvB36EPAsXSimNRTJcrfm8rEIZA18DeEfNwjoMCfKnoAEyG3BzSunHFegthjRp/wauTil918+viCzg/0SCzv9SSqUnNZNi4VrEvLyEBsVliFHYCDg4pfRkBXp7AjejPB3XIAvq1imlj0xWxJFIozprSqk082myor2GFpm7gX+nlJbz/2ZGC8RJwD0ppcoKA6dzELAWWhRuQIqjMxBzd2dK6Z8VaBli5H4FTAW8j5jPQ/0dLqnS/hnN3OKzLppofwusiup0O3/WPCml2yvQ7YeUYg8hrfU7yPq3MjArasfpkRZ7tyQLdLfl9L51npfxNTSRn4PGxQbACSml50uWMWNosvwfn6IJfQhiur9AwsxByOX5ipTSz8vWgT9jP2T5/C3wl5TS8dZI9PbdpDwhHdqhC1pZnPZrSDE9EDELg4F1gdtdgKhSvk1Qf9wkpfS6C4sLozjgF1NK3/dzY1zAK0NzDi/no17O21NKPzcpEU9GDM1lVRQFTncNv/8B1IdOSyn9zqSYXRtZAl4yhaj9tSJtQ2PyNymlHfzcdsgS9J+U0hMVaC0GXAB8O6X0rMkL5ByUC+nbKMzpohbefx40H38A/DaltLufXxAYg5Sk9wFvVWRA1kFhEv8EfpZS+o7JFXsPYGxK6eqyfcCUePgOYF9/98OAW1NKfzYpNLcF9kkpVWLoCs/og4SBzdAcMydSev8BKfdeTild2wLdqZHV/100T6/qn0+AXyR5b5Shk59T10dzy9y+Xq2P3JsnQ/P/Ay2UM1Nqb4U8IZZGQv2awBsppTNK0skzyecjxvVU0xbSW6J5ZmM0T/2ohXL2QWve58AzwOMppbfM7C6kgH21Aq18nY5CSs0tvI/Niuboj4Hnqtapj/1H0PhfD7gypXS5mY0GFk8pnV+BVl9XsGFmJwPzIsu0oXGxEjLGfNeF9C+vr/CM/qj/jwYedpozorVgQ+DXKaXrS9DJ1lRD4+hDpMzdFNgdrVsvofn7uZJlmxtYNqV0g8nt/DBg9ZTSuNw1c6Ix8HKS4rQ7mvl+2h9Z6Q9GQt1LKKx1BdSn/lGmnM2egRQlv0DzyUUppcfN7FDgpSrzv9PL8yt3oj7wKXBhSum/ZnY6Wh+qKHayOrgM9atxSPCeBtXnLMDlKaUH89d3QzffX6dBCtcTEK9+Bmqjy8vwJ528vyFZYGrgeP97a/89FaqPbuu2OE5cWbAjsA0K7dkFzdl3VSlnvrxIxngNeRfviQTxW5Cy5EBk9Oh2rJpCGj/Pyg3sg9rmTGQkPAfYtqKSpIMCDhlhZkFryxPo/edGPOGlKaUnSvKUxXp9EHg1pXSU8y39kTfovyqWdSEk8+2H+ud2yNAxGK1ZSwA3VlkDSiOlFJ9e8KGhhFod5dIYgBaidYAB/t+gFuguDMzrx08ghUP232ikNBndAt3TvRPPhQSDYcg6twCyrFWhNQSY048nR4vt/mjiGQRMkbv24ux9StJeETEcyyN3yYuAg7Jn+fdMSHkwSwv1sCKyUvRBbm1/RIz9dEiYr9r+2wKnOb2HgLVy1wzrYR/bHcUqH4AUEFN5/3oTWYPK0NgfONGPB+eO70WWiQGFvvcqci2tUs4rvYxrIAXfIOQquDuwcgU6s/l3H+Quvz8SEF5HQvNMyPI5u/ePJ5DirEpZF0fWj37AIUjwOtf73BZV2t6PZ/YxOaWXZ01aGPf5PoPCji5GLozz+/mBSLF5Crk5oVieLugORB45g1C857nA0v7fIB9nm1Ys6+TI42V5/32A01nJ22tDpEiZtUpZc31/E2+nu5HCpCdjaTCwiB/PiUKPMq8PvL43boHuzDTm62uB7+X+WwR5hK1dkeYsSAm1hP++wet0uPeBk/E5sWy90pivZ0DKksu8fw1BytKDq5Sxk2eMRIzyHkipsxxixL9Ttf0LdFdGQtMeiGEejZTolWki77wZ0Jx9I/LOycbAZD18/6OQJWyYl3FnJNi19O6I2b6dxlozL1q/F+9BGa9FjO0qSAA9GpgDWS4rzfs5mhs4je2QUWaUn+9bkU5+Xt0M2MOPn0BeGvshxfvyFWj29e8+yNhwPJpTpgH2QkaSBZHS6F5grh7U7SjEZ23l9buRj92zy/QBoE/ueFXgCD8+3Ot1CqTc/kHZPuXlyNaQoxD/sAJSGs3t/Wl1//9I4Jyy7eR1ej3ixdZG89bVSAkzIzCih+PpOyivynCkIHrcx++qSKHbKt0bkdJtDPAYUhpMgfj4RfLt0F1bIb5hZxo86s7ep5ZFxrMjkSfQlFXGgNftSmiNHe7l3AopH64BZqr4zvm+tZO//8rIIDHK3/tcH2P9q9D1ejvQ239dtK7MiPL2lOb9nV7f3PEQpCQZiJSFpyCFxq5l67NJWU9GivbFkTLnTLSmrEeFeaVJnW6D+NVhSLm3pdfB8YhHKjUWCu2/HT4fIWXJ6X58Gi3MU94eC/nxMKQwvBzxWbMg+XCbVsdVl89uB9H41NhAhcUaLZCnolilDbwzX1114inQHOjfI9Civrf/ngpnUEvQyA+6vj7AtvZJcVO0oK3TQtmWRQzRAkhJ8AxiNuZFk/jZSEs7p5f/fuRK2h3dQ2gI3ZsghmteJDhcgeL/rkAKiFmQ9nraFso/g7fZLj6ol0EhCAu12P5zIQb0OSTUzoIWiFKKhy6eMztikgf6ZPwYYkpHAotWoDMYMR77+O8nkevhakj4/CkNxmcqYIaKfWs65Ca7GNKub+L9YdaK77sYUthkTPE+XsabnOZswHr+3xR+fuEW6nUQYg42Q4vFMGRlWSd3TVfMZ98CraFO5xG0mM2ErEyzt9ju/fC5AzFEz2fv6X1hJeSOWoWmFX6PQJaVU4FFUYhPq/10ylw/+7HXxaW44Iksha2UcQ3E2K6D5q8HgLNaLGOfwu9taQgNNyDPgn49qIO8cvA65KUFmrsqM/fep2bJ6hcp9u5A88wsSDDZssV3X47GevItYJVW37vZc7xvHeW05+uqjUvQzI+1FZGAdyA5xrs7mnQUlhfzPr8lmqP2Rta/oT1893w5D0NKqRk6K0eJcu6DPGpAiqGL6iib/54fzVs/RGvu5V6n01WgmS/rbMCJSOk0m4/X+6luLMmvKdmcMgDxUisiIfR3KMSnah0YmpNuRjzEMUhomhEJcpnSb2tgxh72hcW9PrZHAlOpvkpHQfketGY/n/v/eOR+D1IeX1Ec201ozuxjcR/Ep2zgfXMOtLbeg9aqFfz6behGYCq0054+JudCCXjHIL7lhir9qYu+Oh1a6+9Cgmdm6BncQ7prIIXBj3y8Xol4rIFV+r+31fe9Dm/CFYNIYXYW4i3W8fE1tEo5vW3u9bZex2n9DK1dldaqXN/qg3jT+3AjBjLMPorWhWWR5+2mXT2jUM7jkSfkVl6fsyOF3k/JrYsV6zVT6szgv3dDY2oqpEQpPbcUynocUoycjRRQyyI+6LtUUOYU6jQLwfsxCrkC8dUPIgPSbEh5sgvdKM4LZb0BGXTOBA7wczdSQvFYouz90dy/NxqzN6L1sFJ7VXpmuwjHp4bG6TjwLvLBsTmyVF7tE9pjwLo1PCsbOJP7xPMWsF+xLN3ca2jRmQ5Z499D2r5hTnONimXKrB9bIuZlqB/fgxjakT7Bjc7dM0VJ2n2c5u5IkDkIWVgW8PKfCayWu76yUiN3bz80wa+OFv0DKrZJH6Q9XhtN4sf45LAZUhxUstB28bwpkDvj9V7XVwB7VukDfnwIYgi2QovDy4gRfZach0nFPmDIdbs/WhzeRBrfqdACV2Xx6eff6wNPIeZ7NZTr4lAkzD0FrFm8p2LZ8wz50YipOQL4VkU6WejNGYiR3c3HwBZIEdmj8U/HBW4HFIozj/e1Ec3ep4VnDEaM/O9QzoUe0UTKwmX9eBXg3M7qvhs629LwVFjB63lNnxMqCUsln3cDOeG+B++fH29Xo5C/Hnkq+H2r4UoM5FWybuH/nvSBW4DtW2ij/DgaWjyPmKTT8mXtjnaB5oJd1MUGLb5rJryORsLYFl7Og6igIC2Uc9pOzp+Ie8BQwvrbyXPORO7LIMXxFT1o5z5ISbgMWqPnQcz9QkjQXaxFutk4nR8Jb3uhuXpDKniV0JGvuh6FCWzn525GPME9uADeQjn3Q4k3QcL+AWjNnpIWmPlCW39FIEJrwgV05IPKeNUZ4iv39d/XovDA7P+5m/W9ztrcvxdCwuBBSOBaw8f9Aoi/mKV4Txc0894v53mbbObnlkThuKtTUkHQRV893/vTUki59QDy0riwSr9q0re2o+G1tzqaT6fztqrsBeV96CI/PgDNK5mRaObcdZOXee9cH1gKOMx/j/I+sQHusddivRoS7I9FPMXNKJQZ7xNL+vGyXfWtQh84HHn+TevnNkR82hzd9c/O6PrxPf75LpKvFkBC/q+o4KlDR379aH//fsjoejhSzC1PCcNrF3V6tPeDxZDxIfN+miVXp7MB01Qo65ooJAqkfP1Orj+0tFY1+W9Y7vg2YKVW6qDsJ3Js9AKY2a3IkvoZ0nSei5LmzA/8IZXMJ1DhebcAH6eUdqtwjyEt4tNIK3eOl3cnNCBvSynd10JZMpfOJZCCZCenvxewS0rpj1Vp5mifjQSba1JKx3o86GCUp+Rtv+bLWLluaHWIZWvyez3kYfJySunA7N1SNzGQXq+3Iw+dPmhhvA0J+UsBL6QSMaply+75FtZFsaC/SymdW5HW9UiR8Qvkzn09WjRmQQxIpZhCp9kHab7HoX7/C7SYZUzzpSml+yvS7IsEgtnRYr4u8thZDXlBXJdSerACve7af2bEQP0mpXRU9l7dtb9fdx2ySPwbKUZGo8UyIUb3p2XLWab8pszvNyPhZt+e0s494zzUB2qj6XSvxnOXlLg2H088A1K+9UO7dbxmZmchpcnhSdtI11nOW9F2jwfVRC+LYT4fWZ13r4Nujv51KCfCdSXL0eG7cM0twP+llA7tQXk2RUzbRcmTrTZ7ZpVYcNOuLKsCx6aUPuzu3bqgk+9XGyOh/uaU0i987t8JWYLvSin9r/xbf0l/B6TMvzil9ImfKzV/NHsfPz4U9ZuT/PcliOHdx8zmS0rEWJZu/v2PQ8LB/6FQwT+jOXZz4KhUPk9PnubyaKzejYSZ+VAYxsMoN9gnJWnm8xTsiRT4d6H15XCUu2Yr4MlUMldRkxj1pZDy+XUkzM6BrNI/SEru2RK8Xw1Da1NxfE2TPF9PV321UKdToLp8JpuTzOxGYHxKaUdfdy2VzP/hvMoNKNny1Ihn/RHiU3cHdm42xkrQPBQpmn+L1r8dU0rjTDmABqWUnq1IM18Hx3tZn0YhB0ejkJkDgVPL9tUmdG9AvOSvaOSpOxLNNfsmJemvQm9yJNAOQnzEe17Ggciq/nHZea+Q++IHSJkzC3BgSulJH2s7oTnx/Qrvnx9bmyMj3okppadNic3XQ6FypRP7ZuVFHil9/LMwUpC/6/zqn1NKL1Ytp9MdhRSC55jy9myLElHfgbypXmmhrJehtv8Y8W03ozlgE+DaVCGxe6FOV0BKh++llG425UE8HLgq43+rrAle1mvRHDgdmleuMCXP3Rfx1b+vUNZpu+ovznNfgfI11soDfgVFTUd8Jv6Hr7rzHoMmnvvR5DCM1lyOy2jxZwa26qwsndFEWtlj/PhJ3I3ff5fyoshdPycNq/oZeLw7cq+6Hy1E2wJLVaQ7T+74YjQpTIOs3lnZD6dk+E2zOqJJjF+ujqZodk9X7eTlO9nb/BHktTEFLcZnd9cH0AR3MEoWVLrf+HVD0IKQWREWRAqZvXtSVmStOQgxs0/QsKzMXLWtcrT3x638iOF+mYZrZ6WcFYX2GlVsG5rEfnfT/sXxfxyyfl3vZZ2JnDWtrk+uvvdBi2XZPvOV/4vn8C3Yyrx/hfL2RV4w3y1T1tz79UHKwgW9D+3pdbwvcpmv6tZe5v2nBo4uW6cV6E5DR8+iuur1QtyCX+G+pmMRWarzOSoqlxFZlV7ELda5tswsekNRUtbSeRZ8rP4ND2HJ0c7TfIBuwvzoaFXcEc3PByAhaQFkUb+XTjxDSpRzGSTAZyFixXcfgq8RJcuZeVbOjdziM2+PNVFegfkqlq8zS+UcaD3dFVnyu7QidlGnJyJPva39PVdDlt57cIt4CzQ78wBYryy9Qlv0QeEnu6Jkq8t7XRzo/7ecB8nv3wit/8Xxn/FJQ8l5Q5Vop1W9/TMPhSNy11Wx0u6bvRsyut3ix4sgL52jvB+0FCqLPHR/ScNbZ1u0VpcK5e2mDxQ9ADIPzgWp6AVC1xbw45HH7RRUDO12ensgb8KZ0Ry3D+IH+uF5wlqsi8to8EBr+e81sv7USr368fT+vbePs5H+e9cq47XQx97J/f4WmqdaloHgy0S+96E8b2NyffhyWghtypXtPj9eHCnldkUKqKpyUH6+XhDNgasiD83F/b8VaTEHko/PLARxPm+rLIS8av+fHK3PneaN8z58aLEt2vHpQ+BrhYKGbnbXqs2AQg9+gDL2PoDcWqvQzVtqhpp2J8i0dl8ipfSnlNJtxbI0odfXr09+6g1gITN7CjEir5vZaaZtDf9ToZxLo3wBmZfE22ibKZK2IX0PxRfenqppaRdEFvkMCe0n/TekLDjczPZKKZ2dtMd0WbqWa689gAMttyWq11Pf3OVr5O9pQi+/PdZwtIvCSOQuezVa5O9FC1slFGgvbdqSrYi/pZS+m1I6Mvd+qcl1RdqWZI25F9jQ5KEwHjEKlbIeN+lbP0eWtLOR5v5dlKPiT1XaqoB/ANOb2WRJe38/BVxj2vv80wplzdfpPjijV3wXPx5s2qa5Xzft/+W2c6Yt0v6JxtfDKMfMbYhhqoTiWM+elx2nlJKX956U0h7F9+uMZu79lzSzYabtMTOLUOb19D/g2VzblrY0m9li1nzv+PHA3Smlg8uUNfffFchK+Su0W8OvkXVxReSK/VKFspV5//5J3h9nNylLGbozmtm02X05un18/no0d08ZD7DiuSIfMDXaWWLfzu4pnvdxc6Fb+/LX9E3a7ehOk4W4VNs3eeaHSEGyhdNIXq9fmNkQFHv+ZOrCutxkrRuLrPUXed195vPBF6Ztyq9ESvUuPXdSY/vRm1Cun/8gxfl/UO6Dm5DyrZSlssm7D0DW6tVN26YmH1PZu98A/Ch1s7uWX98HrSULp5R+gwwHo8zs+4gJPzhV8NRwupn181KkwJgFWXzfQbHfMwL/ShV2AMrV6Y3AJymlz5K2IR2HEjxfgbxXSnsA5mgegviI64EjzGzepJ2udqDC1rOFtTHbqa4PavMP0Lic2sxmSRV2lMtoF05NgfiXBbL/fWx97n3gGtRHOkWune5HiqEzkFJ3G2Ajk6cNKaVfVCjq60k7fAxGHiofmbaHfA3xbgsBX3Q3hroo8+2oDxxtjW0nz0UW8VboZX3gTMRDfwLcYmbTJ1m9z0ehHKU8gHJ0s7q9BvEA65v4yb+g8dkPKVBK8Ss5eg+jkJAjUCj63Ygf3BwptX5fpZwZfB74Db5FeNJOevcDO5rZVKniDhipsVXqj1Bb3YKMBO8D3zazmVJK368yXnO0LwV+ZmY3+e9LkAw0cwu0svF6KfJ23wDV7VZmtkaS9/uRqeKOavDlmP0D8Dsz2xp5Fj+L2muqKnKQlzWbrx9DCvOn0Lz1OHComS2TUnomadvYVvA4MJ+Zrelz/jXA4t5Wpfu/mc2OxuPOSP7ZuJP3eTa5B3h3vFqPkdqkMYlP9Q8dNXQ3o8G7N3Jhuh1pbp+gYp6CwjP2QZaK7+Oxrv68TPM/BLk7dRqnR0dt8gko2dCayHr6KMp+/BwVc2oUnrEDiktcDGnWv4WsIBfRQuw7DavCHihR0EJIW7sGmnhupuLuL8izZEmvh2WQJWEYslCsV7h2iNf7oiXb/1qkyFoThd3cjawVT9BCEtbCcw5GzO2Qwvn+ubLuShdWJjrRaCPN755O/1kqxtLl+pYhIWaEl+VOFEc/PxLux/SwDiZD1o8LkRb8WnzXjgo05vH+M4Ufv+B9aXHkXTRVof3v7arv0tGacBdi6H7i9x7mZX2olfano1fRSnTMH9O38N3K7krbI6XbKchFdopCnxqKhL3pW6B9CI3kY1kd5S2Vo7q5P29NMhrz6iXInf3QYh19Xd4fuXA/jZjj7XLn82P1onxf62EfyOq1tMUGzdHT+ljYGgnifWiM5aHel0slSsza2I839bIujqzA15CzCiFlx4OUTBrr9+yNYr+/5f3hBDru/DDQ+0iXWesL9bkGYjwn8z6wn9dLf1r0rvJ33w3Nqet5v9qBRrLLKdE8sVyF+jwBeYBmVvsNkLVtf1rMKeF0a7FUFup0PpTAcHakFDwFTzxcdRzlaNbiAYBbtL0eM2XJ7GiOXw8Jo3NTWGNb6P+rIz5jNpRT4CZyPAQSKH7UVf8nN597Pz3Dj++j4bEymAoesORyTnk/fdXb+kC0lh6A1q5KOz900ReORYL3sFbp5WjV5gFQoNtjCziNRP59ES+ReX48DGzox1PRoqdqk+ft6u2V5b8o7VVV7LNOK0s8+S4N+eIw3Iu3Bbr5tfsi4I4a3rmPj9exNHbGWw+tsd2uoyVor41klm29Xob2gNa6NPiTcaD8XD4XVPZ+afKM0f7eG/jvql4lA308nYC8vedH4Tcb567J80A9ruNS5Wr3A+JTsUE0EI5BrvcDaGxnNwItllXdRIfnOtaWSICfFrlLPkzHRXII0rJ2mzDIy3mZf3b3yXE75CZ6XDYAK5Sz6H6/C1rM1kA5NY5CC/GaFekWM1RvhfJ/rIzydjxEbttUyoddmA/oUxEjt4yX90ikGLoZT77q9Xo33TCgObqH+GcRZFVZDy2U89GNENcJzVE0FpwZEYPQD7le7khuyyUv6wN0zShthNxs58qXO1/ndNyWslKSQK+D7yOm/RzvC0ciBc8dVFCYQWP7uSb/TYMW3ctpQVno9XCe9/mFEeN9hLf/sfiuGjSUWmV37LiJRsbr/RCDOLl/Wk5i6/S2RAqnq5HXU3Z+QK6s99KNMEbHLVVHI+Z6JHJxPxTNX0NzNH9IRcEJKcgORC6MD9Jwv8wrYR+iC+aZjkrYq5CwMav3+9mQ0Pws8lIordho1/ujuXmwH2/ibTHIy/0gsG3u2oxuaaG+Qh+4r7s+4NdOjyzVt3j/P5mc0tNp/aBq2/u9u3i9boMU+0ugMLSH0HxraG6skjhyEzQXz+Z0dvDzl9GYI7sNcaOjEnorf89TkMV3ZZQg8LCq75yjv5e3/d7exisiJfepNIT7MXSzY1WhnObtcwya83ZClsuWktnlnmFoLbkQKbb6IY+AYyix61WTsvZB434YmgNOQ2vsETR2AOqJEvIo75PZONueEmtz7v7pEI+zBJrrs0STz3udzuzHXbZNiefsisJP9kQKvTmQpf5RNIf18ed1qpD360Yjr991EB95G3IbXwV511xJhVAROhofZsyV9QHEsy6B5phVq9Js0q/yyo2T8ASJPf2gOeCm3O9DaYG3KtBcB3gF51ERT3AlJXcs9H5zAlpHdkB828+c5qpIcfYsuUShNdRDFupyO+IvSocHIOPg3Lmy747mk1to5Cw7pJU+QIGfzB1/jxaTDzd5XstKnUL5rPDdFykejqO6IWN+GsqLhZHi+cdoPV4NzylHi7uqNWtf71t3It6jcngI4k32QjJBptx4jo7KjaGIdy89z/aobSfEQ+LTbcfID9xZUKxnfkeS+6mYT8LvG4kY+AH+ORF4Mff/ZkjTPh3S2N/bVccrTDy7ogU20zBvTCO7dtWtjPJxqhciRmYQUpQcgS+QVI/7y9O9BDFfSyKN6pk0tvoc1gpdPz7LPysg60qWmfhylCwLxEx2qiwqtP9KSDO7rv+eCzH3PdnOd1okHO/qvx9FC/kFPqmdlNUvEsC7VGx52xyPrHyzNqtv/65i9c3X6ck0Yj+XQwtEth/28Ao0Z6Kx/dyczZ7lvyvlKymU9Xi0OK7mfSDbNvNSfFFH1qtOxy9fVb7dgkIBst9nU9GbKHfvbLk+uQYKacnG7EUoYVR27VC0wHUpKKOF+2QkzBmyTNyOmBlD7tIn09jm7u7uaHbynM2QVWUDJIhcSyN+dyiar0plbfcyHOjH0/j3/MhaW2lXoXa9P7LE7erja4DTfYvGDhsLoLCsUchL6IeUU5bW3gf82hWQNXkRpIC82Mv3uL/3IKQ8r6p46YM8H65A89/haDz19zqaNXdtlwweX52T9kAus/s5zUG0sI1zrpz3AHsVzs+P5tsxFWgV56TD0Ly9F3BDrn8sUJWml/NhFEc/EAmeUyPB41VqsP5Sk6WShkfVvoXziyBmefWelDF33CMPAKTM+lc2dmhsl74n8tzpiVdtH6R0v82/j/dx2h/xaYvnru1ylxUv1xFoHs2UQvshYWlb5P1Yuqx0VGr8GHli3ez9aWenVymfRoHm/HTMhfalN7H/rsRbNnlW7R4ABfqj6ZkF/GDgIxRmBeLP3kJeMc/RwlpVos770IJyE83NdyDFy8ZozXoKKeBHoHWglHdpoTwDKHiN9rTdO3smLSh1CvNIMZ/alx6l2dxCNYPJUKTEfIuGPHUlSkEwD1pXSnvsdtf+ud+VPMua3D8t8nw6ytt+fn+HRRCvci898N6q3LYT6kHx6bqD+CD7rneClX1S2A4xIc9SgaEp0J/CO9kayAJyMmJAJ/f/l/LvNSkfJjGTT+Dnemfug4SPG8m5IJcsX36h2Q+FB5yCmNoRyGr3HaorH/J0v4UUDaciL4uVvbzn0WJSK6e7u0+KNyPhc1HEOB6LMrVn13W6DzodrVSZ2+GeSJkxCgkzP0F5R3rSv3ZAW23u48/KXOWvpKFE2JCSllVvnyeRgmiO3PlsYt8aKU+61SzzVcH+VGRdn9l/n4TiHr9ybYn3zrafO5iOAlGfXFkXpIVkUWhRvNv76/d8jE3p731J7rpOFSd0FD7OwhMMev0+hiyfP6U1xWZfpHAZjBbMASju8/TcNTf4e0yGLPhdtn+uvIaUTkd62TdHi9pq/n+2J/xgSsxddFQWrYzmvs297RfyMr5Fw1J/QFdlpeP4H4AUphsjZjbbW34GYOWKddqW98/RH+j9cRvkbbe9t9FCBbqLlazX2vpAoY0yb4lH0NyyFWK8VieXLJgW5i0aSpdvoTUlS0h4Bh6SlI2Zbuj0K5YD5We4BXjCf5+Lz7vFdyxRB5sgJnQaL+ulSNFzLO423sK7Z/Py0Wjev8J/X0JO2O+unAWaV9DwAFsRKboWQQruSltFF/tA4bslSyUdBYVlvE5n9Xq8GVnCt6dCYs9mfYMaPQB8PB2OhLhsS88F0JidoxWaTZ5xKFpXMuXJ8eS8IMr2AaQkexx5d2ReKosiJWppT6om7ZT1qZ2RgDs14teqhIXl59Pbve1/QmP73WJY2qmUDJUslLctHgDN2oAWLOC5elgazcGP0FDAL4c81Sp5lNBRtjgHhfaukPs/Xw8L0IKSD/HV+eSbQ/xZ3yk7Xgt94A5//7NxT6RCO26APGW75QM7u4YeKHXoKAddgYwtq+NyhJ/P3mcjWlAa+zh/jpzCHPE7+1HBY73Q/lchL/CNc++Qb/9lKBkqWnjG9sgQvQyS1/ZGPNFIGrzKvDTZVKGdnwn2oPh02TkMWX72z51bCGk/r6OxcFaaJHO/d0ZM4cr4Vob4lqG5a7pVRvgAeQRttwWahE9EgsL9VLBQ5Z/pdI9FDPYcfu7bKBxhBBXcWZvQPQvF0c+EGPsDkZv4GCrGE/JVRu5qXFhGTMhNyONi9WJZSrT/D3FXaD+3pU9uF9OCUFugv633r9XQIpTF7B0DXNACvXO9TqfyfnUsuezciOF/nBJxlbm67IOUBKsiBvkE5FGyN3IZr7zvtdfrjU7nRiR8zF64Zivvu5UVXGjxzizhWyIL1rrkGLsy7e/X/RAJxn1oCHaXogRfy/rv0u6HhXqdCSmglkCC853AyYXrB3Y3zvjqvDLM2+wAZE3c0sdbJSUsHZnOKRCTtT0S7l5C2+5Bx5wQXeUAyisL10GKgh0RYzc/Cum6k+rWtLa/v/9eFTGHmyDmeHM071by2mpHH8iNmW0QQzcVCuN4EXiw6hhyeovRYAx3REzYIiin0k1I6XIxFXZpQetnFmayE8p/cyQSEPZFc9b3gAsr0CwqYFdGa+nFTvdMf1Y2fsvsfLMUcJAf74UUzcf57+uR8H0BcH4PynmEl/FiL+MFVFS4Zf0od1yLpZKOgsJUyP3622huXQIJ30cUn1emnEwYD4DNaBihrqSFUEFyTD+aQ05Ec0u2ve1WyAhzeQ/KuQTie3ZEc2JVhVa+Tu9HAuiLufbb28+36iJ/UTYW0fz8U3LhzGgOfDzflhXKW5sHQHFsFZ+X+105t0rh/v2QYWNNNP916ZnTBR1Dhp3znOapFCz+3r+eoDUeaG5kfLkV8b79kdd56d2pcrRORfPdIKTEuorceuTnnqWEsoA2KnW8Tg9Cc/9qTn/DwnM3QWFElfMrIRltBn//PZGSYGFaCxMxNOef7GPoWCQP5g17m6PQuW4VPP6+2btuj5SQawN/REqTWZEidf18GXoyFlrq9xP6gfFp2llmQAzc8khTeR8NhvwRWhDqnO4qNOK11vfOPQaFnlRa2JzGCTTCFrZADPj6PiEd1WIZDQnHmZfG6TRi6E+g9S3yDAmGl6FJ/WAkJEznE9IsPWivTJN8EZ4cyn8/38qARoz9XUizeQJaJGZBTNMDePhFqxOEv/vOfjwt8Be0yE2eu6ar7UeLjPKB+OKIXPie8TrNLNfPUZIBydG8Ed9yDjGjUyPB9g5adD9GC26z7ecy1/7NkcWtUmIrGjG0B5MTtpC7dz7BY9n2nwZZPuZFFqtTaFhqzwLurNr3c8dZWMP2yJK6pP9+Bk8c10K97oGE+U2Rlf8W7xOTUTEBa4HuLkiJdQASnAchpuzndGRyywiMfbxc5yIFXta3FkThJ5Xnvwnw/hsj74cRXs7T/BkjqL61dVv6ABK8xnpZnwV28vMr4GtDC++dhQVsgNbBXdAauBTyWlsG2KNi+y/k7XKkj6m50Hyyj9MdSi6Rb4Wx2gdZ6U5ASoxFkSJ2LuQuXDUB8Qivz8uREWN2NBd+x/+fn1yoAN17qeQZ+gPQOr2If6ZH3hovUTFOnzZaKr2sN6KxvhfO7/i7/4IKYXi5MtTuAZB/RoHOut6HW/X+us7ffTWk4D4TKbXmQHzgBnTcjrWKt05eGbUimqeep8XE7t5O+/vxGd4PMkVeacUrX+UpDvC2zraK34dGOO7aiA8uZfUt9IE6PQAmhAU8X7Z90Xo4ppV+5cffQsqx6b0+Nqcjz7ol4jEr5e5r8sxVkZfxS5TMLVOor/l9HJyZ609n0NgqdhtkOCrNV1KjUqfQLlujHYCWybX1PTQUulsjb7huy4rm+8mK7ea/50bzwC8oGXLbpKw7k0tlgOaAY3P/b+Pt3+3YQmvVK05jGh87Q9B6/ShSdG5MD5Pw1vGZqA+Pz5cdZiiyTByOtNXrAtf4f2uj+KSq1sV9UTb9c4H7/dz6yFozX+66Kovkjj7RXoC0cueipEHLIfepHajoIoY0n3fTYJDO8smtUt4Dvzc/oHdBi+EAZK08wJ81Ha3nAMm2bn0CLWrD/PgIxESd2mL7z4jyVZyOtMeH0bBSH4kY9MoWBrTYrIqEpKfweHzERJ9Vpg/QkaHNkhctjCygy/nxfTTylaxUZpJs8pyTvR4zZdTFiIE4Hgn6ZbTJ+X7VD03EV9Fwv90bMWaz+Th7rkxZc+2ffV8FXJs7vsz78DmttL/TORRZ/ZbxPnBj7r8zaY1J2h0JIbugEJmtkWC7LK1b1rZFlrNlkMCwO7LSPEzOTbKrPtUJ3VWQoL0gCnE4DQmn/SjJhNNx/Gd5dCZHIVNreh1sQAv5PibA++/q/XFP4E00Jyzo42FUD+j2qA/QUUjIdvzIdlKYDjGbsza7pwTtYbnjY9EctaX/XsvHapER7S5UZFoaSVvXRELNd/33lGiN/Q45y3oJmvl+9T00V6/u7b8osqh9t1jWbmgOxRlANE89BlyZ+/96PLdGK22PlOOH++fHNJQvr1QpZ/H51GSpLNTpGWj+WwitK8sg740TaF2pXYsHAJ1b6LNkrC1bI/Hdbfz4KsRbreC/v4XCj+YvPrdMnXZ2D5pPp+/q+m7K/Ci+84f/PhspDyrXBVrfT0KeT7MgJdwziI98loZRbj1aC2mrzQOg0O49toB31q+atFfLuQ8Q7zs/Eux3Rl5aayJBdgofF89SQqnRrK8U2xvxVYs3+6+ze70+N6URgn804gNXRgaNjN5xVA9prUWpAx28f7KQxr1R3q9ZvY4fR3LGVIiPL+OtPBrJUEvTCX+PeIvSSfgL7T89WmeOQ0a4edC6/zRac0b6+1cJld0F8eYXIH5gNhqhchcDR3fWPybkZ6I8dFL++EDolKGkYVHMM8ndWhQKA3pBZJnuj6yL7wIP+3+VYkqbTF4LImFhFhTWkGnYV6BaZm1DCpxBPiAyK0B/HzQtbZHnNGbyAZzl1BhMIwldTzw15vLvLCHgt7z8a+KJOZvVWYFGZ9ukZhP9/EjrnY+lLZ0ws0BzbRSjvLy3z0vIInR+1T6AwoSuQVrZ5ZHV8wp8e7vu3rvEMxZE1qnZkLLsWZ+U+/vE3GXYEA0vnz5osfiut82BXu6vbD/XHc0mz8iHmHwXuCrXZuvk/uuq/TssUk3G1yJezo1arUuns6HTmQO5CWYZq3fL0+6uzeg4r8yJmI4hSBC5DSm1jqai62mT956XnGLQ6W/d1T2F/760miFX9myXmkeRMmNm/13JmjAh3t/Lug9i7nZFFprfoHmsJ/NVj/pAoYxZ7odVkMIhywVzCy0mn0SM5dFI4b4yUuLeS8NDbX0KSSRL0ByFlMzHIAXEEki5kVnsB1FNAZExtdlOT9sii9WdaP2aAQniVeK0+9DYEvVMpNCcEa2nB+eu26OFchpStlyA1qnbkOAw3PtBVe+02i2VhbG6nI/PzZCAvB7iLdal2i4FtXsA0Jij+6Cw2N3JeTghJUF2zWpU2AnO793d331JtD5fSS7cBPEru1d5f2//LSnweYV2rDynoHlpWz++GVcW+u8qVvS88eFk759HoXlgbn/OT4GtWihjWzwAqNkCTkfPj6ORYJtXuPTJXbMIJecWOioLrvNyLuGfs9A8+ENyHiCU4C0LfWtJZNDLjw3LlXcg3fMU+fq8DCkFLnPaayGl5D1UVMDSBqVOoU4fQOvdg2gd2R+tr+fjfCW5UOKSZd4d8SWjCv3XcvU+FdU8igx5Pj6NjNEzIwPhDUgZk1fwdhrSm7tmcO74GODPaMxOhta/u5HH+flVx2y7PhO9AJPSBy1+Z/kAzgSxYnz5jtnEk59QKjxjI7SoT+PP+76ffwXYJXddV0LCTsCe+WsL5ZwFCbhV4zTzA3eol2knZEm7GI81bqFe8xPahj5450aM1yHIDW0KWnA5zcrt9fk6jbjtgUhRcEJnZemkbY6m821ShyNmNwt1aNWqviyN/blXRoL9gkggW7RkH1iJBsN6E3CRHy+KLKtr+O8B3dEq0F2FHHNdvA8tOGOpliguv/gchwSkI5DQOZmPty+3n6OkEErHxTvz1Dk39/8PgR9WaP/5kcA6urM683KOzj23cl4dJCysghiQ3RHz+JjXSatKsh38/in9PW738/viOQHK9oNCWZdHgv1UyEK5ERLor6fidpne/ochK9oApOC6ymk+WqVPtfP9C3Q3c7rTImH3Wj9/DxWtH+3qA0g5/jBaV8Ygd9OnEJN3fat16rR/BnxIQ/g4BikNKiXzLbz7WcDHNJjNBRFTv2dn95ToVyf5+x+I1r79kaLrp/j2qxXLOxx5f7xFY6vxbCeBk6qUk47z3zn+vheg3bU2orHFZyVPOtpkqczR/LbX44bIwn0M8lZ7hhZCJajRA4BG4mpD3i/H0sjHVFS4frkFa8XyTg38Hngnd+4a4LwWx1KWA+04JLQfhASw/NjYyutkcFf9qvgfsi6fCGziv+8ALqtYvnzuo0ORsNUP8VZ7edstiNbAe8nxSWX6U6696vQAaIsFnMaOSucir7yjKCSd9bZ6iBIKvkI5d/K2X9nLMwopSM5FvHAr3r9ZuNjdyGC4DwVeByk2j+6KPh3nlJ1p5Bfa2dt/WaTcOBKFTk5Zsny1K3UKdboqjXDWw73tMkPED1oZr07rCi/r7UjBVVTQboNkoy6VJYWyboPm1WH+zlsixfnxPhZKG3e83bdBculG3i4bonVgL6/vhcnN11TkgdrxmagPnxQ/iBE+EeVV6Cxrc7/ipFGS9jqIQcgG+TZI+3kuOSa5Gxp9kfB6ObB5sXy575HNyl6CvtFIhjgNYpj3QALDVfTAQpl7xh5oscgWnqOoHlPcQeHg33Mg4XZz/30mHgtdkuYgmmyT2t3zK5a1H1p0z6HBjB7i9Tx3GfpIw7tyjt52KC9HZv1aCgkk01fpp3Sy/Wru/z5IsbFwmXL6/2vSUATsDryd66NHIWFuqhb6UL5Os3rsiwTu0/z38RQUW13QyxSZq3jdrZJ/TuF5Q3vQ9zdFVoppEROW5eo4ATixRZpLIAZpjP+eHlkqHgNOaaXP+vWbI6bwGeTavwoSIK6jZFJbOs6hyyNhbiv/PRuyiB4I1eKUJ9D7rwz8koY3xGr+nO/RYr6KuvsAspzfgZiu7ZzGIohRGl313YvXIe+RHwBn585dSG678yo04cvtdndDQnfmdr8lJa3fTWieBTye+30KmlMfoFqW+uK7b4QEkJNpeAPOSC6fQsW2OgLfjQsJoteiOepxKir16Cgo1GKpLNTpwSjkKhN0dkNKyZuotv1o7R4ASCmYjfWhiKkfgNbQM/0ZmYFjM6TUqJT/Iff7RMSz5XMzPUBOedLV2EL8ZOZBsyCNndWeQZ4pA3PXZq73pb12kHCZ8WtLI15yI/+9aAt9NMt9djMS6o5BBpcZkdCd5b/ampLhl7TZA4CaLOCFci5HY/vx55DC4CgaHp1boJwVpcNEaHir3ofvHoXW1Uf9ecuidWFTKhrNkCB7Ta4fnIzv1OXnNkHrYad9i45zyvfRfHITrghCgvJZSCBfB63bQyu2/07UoNQptP09yCPh+dz/xwOP+PFF3pfLJEveKHd8AA2D4f7eZ5fJyujt9GO6UZQVynq535PtWDSr1/MOiB86E4WTdBvqn2uvvsiI+3+4ogl5Kh2J5vEpimWZ2J+JXoBJ7eMdLxvQS+Q7pX9vh9zcS7sz5Wivhrb0XNt/90fxxIfkrunKopx3uz0OTdpb5v7PZxNfhZKWKjoyH4siK0UWvjA98Ce06FadbPN0DwNuyv3eD8Xrzk2Le9U7nX3RAnMZEsRmRKFC99ORGS/L2F9B821Ss7rfxt+l7AScZxaXQcqiUUjbfw5ahM+kgltzjt5auEIMMXBP4UoHWoz9pIvtV3PHc1NCq4zn0fDjrD9dTS42HXf17EH7742YxEvQYm6IcXiI3I4KZdrfy3sxssr9nYYHSd6leRvvH1OW7VM5+vMC/6ChNBmJFuW7yVnXWqC7DIqlPg5nNpG1eoce0Nwcz3/itK7z/tafXAb4rujScUFfFFlQlkeMXEtJhyfU++f6+Qt0TAy4O57ot8V67VEfoON8Mq2/9w2IaR6GQjEuIid0tPjuO6A5JbOg3oQUGkeSY5gr0szyEhxGQyH7jI+3LSrQKVoiRyDhOB8mYrS4VbjX4UloLhyJGM1z0Dozpmy90nH9G+71dw5SIGfttSq+q1or709Nlspinfq5a3GP0ty50rs/0CYPABo5mTbHE8wiweNI5GVxARKaB/tYKLtTR35sbYrW6RmRYP80MshsR8lEjv78LZC1dwkUZvZTpHRaDRliHvYyL4nWle5c7/M83iAkfB1JI4fWqd5ulXd/8fv3A37ixzM7/WPQeld55w9omwdArRbwAr3Js/6L5IFtkSB+g7fZXGjeKu1hRcNb9Vg0t96c9Uu0DX22e9uyZdqOr3oOLIMUL5lnXV7BtSlSzpUdB8fQEOgPQPNKZjSbOXddmTCJtil1vE4vwsMhvd/n19C8obDUeMB3lPTjDekYxnQr4o2n9DYbW6FODfG6xyCj+ZdJ95H3Wtb+s1E9BHtVZNB4EPhW7vyWtJgDqd2fiV6ASemDFt8L/PgQHzRL0LDmboIWokrbsBWesQZiEtdv8l8pd2afFI9CVpSrKFg7aGw/2K13BQ1vh77Z9UiD+jwKd5gBWShb3fmlL42swmcBl/jx3Mj9tBKTjBQtmZVvaaT9nAFlL/4xioWdno5bR5VVatS2TWrx2d5WTyOl2PHI0pq54J7XQll3REL9gYgZ6YeYqLE+8VYOk/G+dSNdb7+6NRW2X3Wao5Dla08/dz4eKtBCGWemwRDMR84VFFkZdvM+l7dWlK3TTXL9c3Xgt3TcRWcLKlrUmjzjO8jlNuvDcwOLVS1rE7qjkDJqbwreT5Tc0jh3PABZq5+nwTAv4XW9TJmy5vq94SERKEfBQsha/Tw5z5+efnr6/l3QnRMxtwc3+a8luq32gUIbbeZlWwYxNbsixn4InoS4B+/8Le/n+3q7ZaF3p5PzgKvSV9Fc90MkwN2EFMjDkHfZThXo5Bnl29F8ug+a835ERyNBKwqdLKxnVzyXDhJk1qak91eTcu6F1qqFva12oODW3gLd2iyVBZr3olwSl/m5C/Dkcy2WtzYPADoK9YvQSJQ6n/ejH6Gdb1bOXVcpqbvfsy+NsLBLfXxOj4SRM8v0L8SD7YSUbut5X50RCUu/R4anp2nsXjYz3QjgNLau74eUgysgj5Vd0Pp6EFKQLlHhXYsC8lJovjvA++5cXg9dhgZ11va5PlC7BwA1WcAL9G7yOlwTKY72QvP1T+jIU3bL/9BRWbK596lsR6GdvE9UVvDTcW7ZCa2nUyPj3t5IafQUDW+tUXQhgNNRATs5Uo6dhwTuAd7+x9NQ+FRV5tem1CmUdQrE756fO3cjHn7p9VM2rLl/rqzHorlzAOJ5DkOKrftpeF+N7G5MFNp/BSQ7ZHlwVkTz9/rNrq9Qt2sDT2Tv4GU8DMkvGxTH4tflM9EL8E3+FDs9WlwPzf2+C1nEJvcO9GJXE0SF545BmspFS15fjCP8IQ1hbh3EjGzhv0u7Xvo7HeCD9HY0oZ+AFoclkVb6aWA1v76sgLgrzmB62a7GvSe8Pq8CfkWFZF5+71A06e6IFsPZULx3ZqVdlwKT3FWZm7T/gdS0TSpivIb48UwoRGY2xOjsgnto0FFxUkVQ2BcJNfN6GQ9FAn3LWzlR0/ardFx81kIKo2XQ4pi996Vl+3+O1kDvnwd6H50CLWTL58p8ROGeKu0/GimapvLfJ6K8LZMjweQntLjtGh3H8P7eH2YoW9Yu6OaF3WW9zKu2UkansRu+e4yPswdpuN/O2gK9m4Az/Hg1pNyYCVkTWvbUqfP9aSgImjK//v9dVHDBb2cfQIzrT2jsLrISEhIPIZenqGx/oiGA9EFC01lIUXCUl/ESYJsq5Sz+jwSF6fy9b/e54A467rxSJadG5vkxHClhDkUK6WcRQ95q6M12/u67o/n/TnJekcV2LFHOc5DAcSdi6kcjYWtPKuaUKtCt1VLp115GY90+n4aAcAutb+teuwcAspyu4vT28LZfGVnTM0VslRDMvEv31GjOGoAE2qeQcqfDzghd9S+nsYzfvwVa8zfyfj8SWWzXpoKxCCkcXkJ5fk5HAv1jaE2eGVltv08LoVdep5civm1j5FV3NI1QjJb6ae45tXkAFMZAbRZwGjlljkRC6LneRhsgj42MdllBOc8DZQrsvZHCIEuQuisVvbUK5X0Eeb28iNaEhZD3ylk0cqxV2X56D3/3mb0u9sE3UyDHp5YsX+1KnUJZV0UKxwFo/s97VVYymNDwLDOkxJkFrfdHIJ7zWDQHrFd8t5I0F0RKh1WRLJR5Qa6YHVdp90If+BbyAsuUhVN5uQ9opV9NqM9EL8A39UPHXRq2Q5rDKZBWbR8f4PfR0NDNSgkrC50seMXBQHkBOa9NHurHB6FQhoHIjesyZFEd5sdlaa+MmIKrkRv/ZD4xfgdp6ientW0s53Wa99JglM6h4Q0zddkyFuvVJ4OT0GIzG2IQj0ZC7a3AMSXptXWbVK/Pu2koN06hESayFL5jR3f9pvgfclfdw/vA1YhBnAcJIbNWrNO2bL+aq9c50GK7K7JWzYeUJftUKWdhHMyKmJlvoeR4WyIhaXMULnF2SXp5N+kTEUM3FC26DyPvjbtpbO83G86gVOirRaEpv9gfSyOMqGriyQFd/DdT1brN3bsFEgwHojE6EClOX6L1nD0nAH+gEZJ0KLmdCyrQadv7IyH2dB+XTT2dqJ7Qs/Y+4NdNhSyUSyFL8HHeZ9emiRdgxXrNZ/4f5f2/D/KKKLWrVBOaK/mYnR6tL3ejdfZgYO8W++nCyOqdz3Nwo39XyXifL+eCyEI7pY/1LPns93pQzovxXTSQV+EVPp5moeKuYrTJUplve+RFlfdIuIKK+YSKz6VGDwCnd3yhjDMiJf/RtCCAF/rA5LnjdVCujmGIpyoV1ob4kP38eCE0r2yF1tYNkEW1kjCTo70FWv/P99+zI2XGji3Qyr/3bUjhuDtSEGQeMCfT2g4tbfEAoGYLeIHeOkiJu77/Ho3mvY26o9MVfeRFdAFSDo7wfnoRra1T+fKOwXMSeZ/4NYXk/hXq1ZAx9Aivw+NQHqST0XpdKeyeNip1vKw/QoqXO73fDkF8y3Et0MuSEPdFPPRVuf/uxPO1Va3TXPs/jhTZz6E1Zjs0Vy/TQlmL6+rMaL7ZBjgNNxQW+snXylMj+2TMUaANMDNDi9YXaMG9B2nrD0dxoPellB4wsz4ppfEVaW+KtHS/A8amlMb788yPMybqb6mTRjazfimlz/2+G5HgdT5abOf0Mi+MvEye9XumTCl9VKGco1Bc25RIYAQJdQug7R3/XeW9c3TnQIzdyymlY/zcxYhR2qEiLcvqyMxmQAvvKsj6+wENpuF32bPK0kWu1p8jJuRA4DO0wI9ESf0eyD+/BM0v+4qZHY4sPbsgb6C5kZC4DvBeSunIsmXN0d8LLeo/RwqOJ5GQMFlK6T8V6GR9qw9SNvRBi8WeyALyMlIWHJVSes7vmSal9LcuaObb6UA0ji5F+RV+klK6xMzmA8anlN6sUNY83XmRALIBSpj6tl+2IfCPlNJ3qtBFfbQfcmFeDAniWUKrn6WUHilLr0lZJ0sp/a+z/6vS89/bIoXO8yhh4id+vn9K6TMzG4oUaM9UKbfT2ADVwX+RV8WqaJzNlVL6WUVa+Xo4DDEyxyFG5ISU0lNV6fjv2t6/MFZPR3PgtcCrKaXP/XzflNIXZjYFEqJ+291aUHcfaEJ/Z2RV/TUap7ujMK+PW6VvZnugMfQE8D7yVjgAKal/nlI6roVy7obi5m9GSpjdUPtvhtruwB6Udx0k0F6Ftg39AZoT3m2B1j7Io+RNtKa8gbyV5gbGpZQOqUIvR3cLZCQ5JqX0vP9eCQkgpfmJrJ/6XD0G5bz6I7IqPpVSOsuvWzil9ItWyur3z4WMBm8gYfcO5Ak6ruT9llJKXs6LkfLlA+CvSFD8OKV0gZkNSin9tyTNvimlL/y4DxKS30wpXePnpkOK/d+nlP5c4XWLz9kTGXpeQGPqYyQwjPDndbuuFMp6Lsod8z+0/r+EvF+XBv6UUnq5Qtny88n6yFPpiJTSC2Y2P1ISHF72/c1saErpX16fCfEPP0QC+BVICJsG8Sgfli2n08731d2Qx+MfkbfSP5BQ+hck2P++At1sHs6SD7+JxtJ2wMUppZfNbEXgP2XqtkBvDsRHbIs8bc5PKb1tZqNRHbxRugIa9A21+5QppQvN7F2UW+8VXw8fSCn9uipdpz1VSukDM5sMeWf8GNXt+Wi8/qkEjYEppU/MrC/ypNslpXSamT2M6vNeM5sKKfTfaqGMfZAS7zdO/yDEA8yAFJOlx2p+vjCzNZAH+VFmdh/apexOMxuMvGlfrEB3SaRw2hJtkXqOf1+cUnrXr7kbGR8qtZW//9rIM+lcMxuHFORPmNmGaL56rQrNHO3d0Pr0MvBPtP71Rcb5k3oyD04wpK+BduWb9EHM+kx+fBmaYEAd4x7gyML1rbiH74wmm+MRo5Btu5NZiYf6/51abWhkt81c7o5Ai+KZ+IBBE3Lm1tSTOPIlkAversgiOAUl9+buhu5cSCGzX+5cS678fu8BwP1+vAKytG+J3JHz2cW7SsDalm1Si9cC63ufOgF5Vkzh7bYXuR0VytJHC/jRSPGwnfehb6PFo5JXTfZM71u1bb+aoz8GKR/2QcLHrWiB26mH/WlfJHD1Q1bVY70+ituwddX+6+faf2fgp7n/dkQJwmatoe/v4n1/EzomYc08xYYgS3uXVkY67lG+FbIALI0YsQ2b0PwRuVjgimVeEIVzZbuW3Elut4YqY6HYDmgufAvfgpryCcLa/v5IQfgDZJ2/BFkr82UfgixaVV1ca+kDzeo/uwcxs5f2sK8u7/U2FOVXOBF5WCwL7N9K+yOPnx84zUPQWtsfJVQc1WqfKjxjLbTLzqM0PCur9tHZaPAAtyJlPshrr+W+n7tvc++zx6H1pnSoQLHtqclSiSz9nYVdZYkRn6Wx20a383+hb9biAUBHb9UlkfV/AZSEdRNk1HmSJjt4VayPddHaN6OP18OQUmsz4OQqfcDLuiha659zOnMj6//WndV7xfrdAsX+L+2/q4RwTOd9cQnkmbuq1+fzaC6a2Y8X7UF91uoBkO8P1GsB7+P1eJN/z43Wgu9Tcetlp7cQztN7Pe6ODDu3eB+bl1wOoAp0D8dzhnlZ70F85FxIGD/A66XsZgEzI750JFI4jUI7wb3i/WF2NP4r7VRYaP9d8XAI4F08h5SPrSo56mZFStEZkEEw2xr7RWR0mQWtWZUSRdNYgzdGYZ1z+zPOQIbeVjxq5qeRFHxhpGj/MZpbMq/wS2kh/12uv06L76CI5v9tEZ89A7ld0L7un4legG/SBzFb+a1M50caum383OSIeZib1hQamcB4EY2dT5ZFC0iWrXioD8xOE7why8HefrwcErg38d/LI836llXL11l5/XgZ5LGxb9WBR9eM0hxoYq88oRfozOsTxJDcuRW8LpbLnevKTbQt26Q2ec6OiDnI8qB8B1lvhnZW/yVoLuDvezeyhB3l5yu5itKm7Vdz9BfyfnSb018ZKZ6Oo0UXXKc7E2KSp8qdmx15m5Teo5tGvpjpEQN6MRIWMkFxN3qY1BIpTB5ETMJD+NaGuf+HoLjQLpkxtIjdASzlv7fwcboZ8jTqg8eponnlvu5oliz/ZIgRv77k9Z0KPnRUEOzh46tshvK2vD/yxtk79/tyGnlk9kcCeRaCNMx/Lzsh+0Czflw852P58rJ9v5PnzOj1vD9SaFzp5zcnl6+HCvOhj/lp0FpyO76zBppfVitLs9n/TepgZaSQGNPCu8/kn9OQwHWF96kDyDG2Vd69WTmRO/szNNbwsjk68vlS1qCRq+Y+PJwLKYqWqlCu0Sh/QIecEYVrsoS5+5ekOTR7LyTUHILm5nuRp85wnBGvWn9O7xbEl52EvChGIS+gm3HlZqsfr7/5kIL8iFz/X42OoSnd7lbnx3Mh74zZkeLtacS7zEMLwnIXfWo75F01hOqGh62Af+FJYb2sN3kdPEULuYRo8Ld9kcCcbWv7MA0F9FS0qITyvrUungMPKTQzIXJDSoY1IEEwSxS5NY2wnj2Q18pAFObaSlLPudB69QoSmAd4fT6IBPLHqb6l7dT+3o96/z8E8S6nIG+45VAem6p5pQ4GPkLeCaB16i0k1D+Hyy8V6LVLqTO7j8uxNBK774cUBtuiebWl3FfeV89H6/tffYzOgXifbbz9qvDoQxGf/xaN/EdXIm/geZACpVL7F+hnc+KJNJSaC6AQ7L7F677On4legG/iBy1a30Ha63nQArFTK52CjotN5mWxp0+UA9DC/BDSLvdDjF6nmtUCvVN8sGzoAyRL4rccFT0f6GRhLjxvOXzLtbL1QDlGaS56LiyOQAvP/Llzi9HCDjXUuE1qsS69ve7Gk1fl/juJGhL6IEb8ErRYzlrx3n60efvVHJ2dvN//hVx25h7Q6+8T+Dq5c2tSMqkVHZnPXZDycXa0IB+GFC89tqb579GImd8DCSEXO/2huf7RnVJjhH82QczBIiiO+D3gHr/mPOC7frwPnkS1Slk7qZ9ZyAkMnc0d+f8Q43m0zwP5XA19ctdMTUmrbbveP0d/mtyYPwc4LPffT3DFMbJmrlyCXm19oNAWI2iyvS4VhZkuaJ+M5sPLkRAyP7KIXtQKfaf7LX/vbVG4xJ7IZfryCjTyOZCWRAqmfLLD/Lq1Bo18Q2VzFU3rbTvYy/aG988zceG2SjmbPSv7+O+1kbK7lLWSNlkqnfbuiLcYRScMMZoXz6UbJTdt9gBAluUsN9eayPNrLf89ZbHcFfv/7GhdXg9Z/TMB9BzgrIrlNBr5A9ajkSR7O2+zKl4VZfvUsBbrdDDyAngQVwggAWk4LezWQ5s8AKjZAo6UFuchY1g/H1c35f6/KHteqx80j76KK1qR4ukcHxvrVaS1CDI0zYoMms/huV68rR6iOg+Y9Z2l0Zz5CA0D3HLIuDeqCk2/t3alTo72YX7/TjTyvy2KvEJa8k51GtsAF/rxVkhZthBSdLakhETz03PAXrlzByBlTCVvPTrKOlujfD3T+Rg7FxlPbsCTvfemz0QvwDfhQ8cFexvEBO+BGPGRSJv4gh+3yjDughb0TX2QX4Jc5e+j43aew8qUFzFee/vkNQUNK2UrmuQ8k7glXxW6OzC7LdAvwyhVtiwU6sOQNf0EtBXnjZTceq1Aq/ZtUps842hyyivkwVA2m3anVkoaAuJQGkxUZUUcNW6/2s2zFkGW0NICZ2dl9u+NEIOwL1JyXFq8ppP7831ye7Rw7e8LxRxokTyG1pi6fB8f7XNIPyQk3YaYyDPwhKk+jrtcjJHn2BmIKZ4OMXS3IyFnIySAnk1Ha32lxHlontoKMTh5l+/seACysJXJAN4HuceeixjNo4p1iYTcByg3/7Xl/ekoFGTM1okobOohGjt1nJa7p1u36Xb0Ab9ub2SpO4Hclt40ts8e4vVQad97v3dZxOgfjeaDWb3dTiMn1HU1rprQ3AaF+m2KmNAh/pwdaWEbVu9XNyIl0Klo3S4m4d6WijuMIGXZMGT9zITmkxEDWnrr7Vxf6oPc13cnlxjX+0B2zfyIPygl3NFeS+UVKITxdjT++xbeZzs0N5batYQaPQDoyIsMRYqHq5xmfxRKeDoVdnvp5DmzOf2jkWA+2sfZRcC5LZR1XbQ7wbKIxzqXhhdgqaTTLfSpNeiBII7nuvH2vrIndUrNHgC59r+GGizguTobjHiIbOevixFfPQblKli0h/1qbu8Dt6K5sD9SQlb1qlnc338TtIbMibzTHkcKvoWRsqhHuzSiOeUxp7lH2THfCa3alDpNaC+B8nTsiBRS67ZAo1/h93pofc3W0wuQt1VP6mAaxKdchea/kd5WrXhTXu6f1RFfciLyUpsTKcq3IbdZQivPmFifiV6Ab9oHabnWRkznXmjhnIEKGvUmNLdCsZ6LIgv1rmjhXJBcyAjl3RkXQy6MU3oZH/DjLfFYuxbK2AdpZ49DzNFBiKHPM+VbI4ZpSJVBQveM0rZoAWkptjL3nNmRNvhC4NgWadS+TWqTZ+ziE/peiDHPZ1ruVgCnuZWyaK2pwsjXvv0qXYcf9CRuvtOtN33MjkJKiAMr0jWkINwBV4ggZuxcf+aUPWzzbZBy9DYkiM/uY+0pnNkrSWdpJHSPRsJWllF/I8QszeTjdrZW65tyOYAepuscQPn5ajkaWwM+hwTRo2h4mG2BXMm7tVa38/3pKLgdi5Q3dyIBsj9iHLdv9o4Tog8g5UJWZxsgZdGMaO68j1yOGjRH300XIY1dPGcI8ClStpyNPBYzN+KWM6ojJuw5pCh9BvhWZ21Qkt7ewDW5fnEyufUPMf2PU8FrD4VZ/dDbfUGvw68w3d2Vk0Y2ffN+eaz31fvJ7dbi12yFGOiquZBqsVQCG+WOD6CRV2p/lPtkGRru+ZuiuaG0AYWaPADouP7tjtbopb0f7IcMRn1oQZFXeM5cwHikKLsSCbbZDh3T5q7raq3OK4PnobE2HY2MPH9Fc+uX82qJclXpU5t7n6oaimp05PnW9ed065XWGT3/rtUDIEe/xxbwYv17vzwaON1/H4rm6crCchfPXBWteS9RfVv7RWiETE/jbfQDJFOsjJJGXkXJnBpd9V8/3hcp9sf08J1rUep0Uc4VkUHyeXLhxxXp9fVxvzxa/w9C6+xaSHlQuk7JheAX5wqvi+uRsqxSeHDhnc9ACrzMc2k3NGctVrin1yg1UgrFRs8qr+ME/iAS3nb3hWIBxMB/i4pxf3zVarS2T5a7I0bpGQrxZMV7OisvsijP6JPZzn7+GF8wqjLa+9JIqrYgjbjHZ7zM+YSbWyIFShnhY6PccVlGqVvms7PB2aS+rbP/urjuSWraJrXk89ZASphKlkrKWSm3QXkhSnuXeN+agxq2X6XB1HUafpA7XpgKiWgpsfVmV32ji/Y4Clk/Z0RWoCyHzXH03OqxNrKuTY8sC6eihXIJYIkyZc1dszLyUuiLGPBsu8DpUEjao+S2SizTp4rX0vMcQPn2zQSCvkiw3dbf4Qbvb1kywrLbBNf+/mg+ypIsr4QYr2xbP0OC7iVl+1U7+gDynlgPCYmDkPJhGh8P1yPr6o+Q8DQlEsxbUWosgJjNtZHAuCVilK+lpFDXhObiaK2ZDSmg1kIW+1epIDDxVQFkGTRus755Lr72oHXleSqMXW/76ZFi70qUs+gBtLYOrUBnORpWyaHIijgAKQPP9PGTuYxv5n22paTZ1GOp3DF3vCE5o4CPhfO9T62B5siW5kNq8ABAa8q9aE76PuIvlkHC5+G0sKVrgf6cSMG9OxKOD0Yu6KfRkR/qSqmRCe6G+IoT0S5lo7z80yHBs3Renhb7VJdzanE85d+NgoKjrg81egBkdU0PLOA01rw+yNp9MPLUmhUpTb7bWT111087e1bu92w01p2y5T3fP6f7eN/J33k9FJI5l/eVUl6wXb0bHfmjlkOwm9BtSanTrE6blLMfjS1ky+Yq2ix3zy2I/zkRGR+H+Vg7lWqeZaPpPgS/Pw1DRaX+6seL+fe5uILff++He1v31s9EL8A34YMYxeV9ctwVMUT3Iktg007ZBa188splEKPYB7ltZ4mnTgWOb6GcS6IYtTvRgn4KDXfGSm7yiEHeAnlRLIEW258i7d9qSOh4GFkalkQLdCnmizYzSnTuJp9ZdCZHDGoV5nsvJCAcCPwSWUH7A1O00E75yWdA4b8++e/i+RK0y1gpH6OcAipfzgNRotxvIybpW35+vqrtk70P3YcfbIUsAd1a2OgoLJ+OFp8lySk3cu0/pZe77GIxEAk1P/AxcQZy42zJ8ld8Lo3EY6P892ik4Kuk1MhdOx2K+Z8PKd9OQwLoSAqa+op9oI4cQPnxeBNSwK2J5ti9kEDyE3IWZapnLK/t/Z3eRkjInhcxhpchhfaMuWsq5b+psw8U2mg2xHRldK5AORUOBXb3cwvRguceUpbsj9an76O5YACaU/ZtZSw43cWc7kOIWVzb6ZZ2P6ejsnQnr7+pvQ73RsqIp3CGHgmRVdeVE5B3zhFe1vWQB9AJFelk3hObZ+3r73+kl/kCNHcNRvNYS/Nr7rglSyUN44Ihq/xF3i4PIW+QlZGRJzN+jCSnNCw7Dgr9t7IHgPfD+XJ9+xg/fh5Y3Y/np4cGCMQzHIlyqpxDI6/aMZRU7iPeZhc0322Ctl0FKQc3IJdEc2L2KToK9N2FtKxGz3NL9NgDgDZZwP3+q5BAuCLinxZCCqQTqBjeTdeetVkC3Ww+G1h8ly7oXormu8FoXb0QGV4287G5EcpZVHYXrSr5rxahpAGKNih1qBYuX2VnpbWRd9ZxXge7I/nhScS/XIG2tG2lz3cagp97n6loTXG2J5IDszn8HCrkfvq6fyZ6AXr7xwfJiWjx3BstEhsj196qyoKpkNvuPIgpehJ5guyHLJ/3+OeS3D1lJ7XjnP4qaNvBsxFz/0P/v/TgQIvjTkhxsx4SVmZESojf+zOexuMTUYxpt6EYTBhGaWe6dpMfQjdb5Rbo1b5Nau73tkiQW5eO1p68S//obmi21Urp942hhu1XaS38oGqS21q33kRM6Pne9jcjV77ZkcKtFeEwz8SvjpjbKb3c52VtQ+vJpzbycTQGKV+y7QIvIcfQFftiyXqoMwfQtxHTvYL30bURY78tDWGklQW9tvcvtNVJSPhYGFnrTkWKt5k7u2dC9IECrSUR4703Er4X9np9Brislf7U5HmD0Lx0MfA6Yu76NytPC7S3oJFhfrGqNL1fPeJt9SJaCxZCHmpn0diKuyc7Vs2DFO7PojlrirLlpKOSdRHkSXE4je1Mf4Ti/1fOXdet4ryz9yn0jUqWSjoKCkcgL527/HggGvsX0EgiXZpmJ+3WsgeA9/dsrE+D5qRM8TQ7MkL1KKdG7lnDkTHqXuDXFEJau3oHxJNdn/u9orf5QzSSZJ6P1qiyhoza+xSthbTMWrK8bfEAoGYLeLGcSEm0FuLP10G86XKdPatEect41m6NeM9un4Hmuffw8BqnebyX91g0B85QdRxQvwGqLUqd3P3dhctvhdbDwWVpIznlZDTPLI08oDZFYX3X08IGBE63aQh+7v9taCEEHylxX/b2Xhp58ffx5+3SSlm/bp+JXoDe/kFa+oyR/4V35AFlBnGBzqJIC7kW0vheiSyLw5EFeCNk/WlJ+ECC991IODgKuUwuTEV3Y6S4WQYJ3FvQiE+/Hbd4ImZ5pYp0a2eUCvRrcZNvQreWbVL9nsG5462Q2+3SaMvUbEuzbH/sIYgp6TQWmgljpaxl+1U6KhdqCz9gwmy9uZKPgXuRYnIkLTI0OZp7oczvZ3vdzuXtdSO5ZHFUV0As7f1mGJoLnkdMbss5YKg/B9A6Xo9ZOMdo5NK7UVc0JuL77+F98ySkLJ4HCSGX0oPtcevsA97XH0WM3Iw+Vg9Fwk3eS7CswNQhVw2N+TXPhO6O59gpSbMDjWbv5/2q9NaDhX41hoZXypNI8Dyos2dVbKtimVeiYp6eXL3diowDM3vfOhTNf3PQ8LQp2061WyppCLZ9ES+Rz/F0J7nkuBX6Z+0eAHRUqO2M1oxpkEAw1uv0SWrIfdCk/WdABq/9O7umcP1USAk4u//eERnIXkR8wMJIwV859r/OPkUbQlqKfZGaPQBy99diAS/01SO9DvdDa9bRiHd9lp7N/bV51ubuWcXH1kXI63ERlCQ9S2xdKgkttNcARY1KHdoXLr99dp2Pgev83L4oFPlir9cqW2VvlDuuLQS/yXOm9vY+mUa41Kmt9tWv42eiF6A3fci5snXy/5EoQUzpycbvmwwx2Q8gxm1ZH3ir+f9zIoErz4C2kgV3NNLW/4Xcjg8V7l8E2M+PF6IRnz7Cy38/FYTZHN3aGaXiddTgJt/Ns1reJtXvn9b7zlL+ewufyLLdavrgLo2IobiPEgsnE8BK6ffvRIvbr9JgVtoSfkDNW28Wy537vaLXQSuu4XMihiNTXJ1OY2vkPYEf+fHqLbbPmjQs/TvQYE53IucqX2ZsNXnvHuUAoqPwNad/b+ftMYf/N5oW8wjU/f4FugMRY7Co/94SeB+FpVRlvNvSB9Dccgu5PA/Iq+QAcoktq7w7XeSqadbW3dGm41zdLLGvFelWLO9UGW1k9V4HrbM/pcI2kYVyNo3vp8Bsd9X3m9x7PB1345oRMctH02L+B2q0VCIr6pvIgDENyp1yCh0VbXdTTeCq3QMA9zpA63q2vmyJ5vupkbC4Ft14PHbXB5qdzz2vtKcS4n2ORjzEhUjptr3Xxw+9zVZrsf1r61O0OUyKmj0ACvfVagFHhrez/Dgz5uyHeMBKOwo1KUvtnrV+7xrIULaL/x7ttMvuptS2/Fc5urUodWhTuLxfO97fbX9/TqaM3Q3NX9dTMfyKNoTg07l30uw0lKgXAEfn/utViUKbvt/ELkBv+dCFKxsdGfOpq3QOxMyORm5wGyAFxhzIuvSoD5TTKLlFWCfPyDMw/ZEgX3XP47yG+1wkHM+DmNstkOfKhlRUbNAGRqnJM3rsJk8bt0lFiqERaMK+BSmQVkSug/f4NecB3/XjfegiuRMTyErZ5Lktb79KzeEH5FyXqWHrzWbtX+gDX1p8urq+C/pDkGfWeYhROBHf3tD/v4WcoNxdmxWfj+aRx9F2huchJq9TJW1X5cwd15YDyGk8hBRbDyG38T0Qw1A57KaN7/+VekfMzZ005oHHaGGbtLr6QPG89/l76GgR2oAWFEV0nFu6y1UzBVLwVJkLd0HWuk3ICa909FTbsaux6tcdjjPD3qfuQUrSuZBy8wDvD6UU2U3qtLMwwfxaMLoE3fy62sfrdJfcuenQWjBTxXaq3VKZa4ONkXAwN/JMOAOtrZXK6LTakdRyIxphsHchgWB/H1ObITfsHiV0dtrd5eoahNaesuM/y3OwIzCdn9uGiltZtqNP0aYwqfyYyfWHA/24Rx4A1GwBz7VrH7TuXVv4f3r/r2oIcts9awvPWxV5hH/bx1PZ3V/aYoCiDUod2hQun6M/BikvnvQy/QTxGPN6XVRRlrclBB/JZ9vRUZbKj7V+KLfSFa32pa/rZ6IXoDd96NqVLRv0U1FSqPEOuqkfT+ODK7/10kpI6N8j/6wu6HUao1osZ4vvb8jdfG206GTx6d9FrmGVBAXawCg1eUYdbvJt2SbVr5/c33c7xHBsgCbhGRCT8zByRb88d0/Z5E61WCm76lvd9cluaLYt/CDXHj3eepNu3LkLz5utlXLm+sJ3gDP896WIgbidAhNVgf7GPpbm8H6/OWKY/prVdQVateYAQsJhtqhvDZzvx1mizIEoCWel5GttfP+8MLgvEpD3QUzoIUihcStwysTqAwVa2yNBYH7ELGaM8k24krQH9Vprrhq/Z2fvT6sipu5ocrmOaITfdemp5u2xrj//JG+b6dHavT0SnvajWkhLrWGCfl1eUFjS234BpITfBIUfPEn1XdXamdi7L1pLslwn8zi97yIhfADVFFm1ewAg76YzvW/uifiVp5AyexBSblRWvjfpq7Xl6urkGcsjPmtMhXva0qcy2tQYJlW8lvp3wKrNAk5Hfu9MH1c34KFG/vt0cvNExbqdIJ61ueetBfyJimFY1G+Aql2pQ5vC5Zs8ZzmUV21zP97Lz09ZgUbbQvCRzHMeWq++4t3k42zm3O9e76nx5btM7AL0pg/du7JtS0lXNp9Mf+qd+DaUPXkbH3jr+jPmIbcAddXxchNvpzGqueNVKblY0nHhmQttsTS7T4xPIw3lPLSezLBuRqldW+WWifurtE2q96ERaDI/mUZYz0Zo4Z0JMWKzlewDtVopi/VDF7GvueNS26/SpvAD2rT1Jm1IPFW4d2HEhM6AmIZT/PyyuFWhu/b3/+fCFYLI8vc0UkL8nUZukeFUT+y6KDXmAEKL9nlo0e6HFFs35f6/iBYy6bfx/ddAjOu0SLl1N7J63oIUjwORIFEptKMdfcCvyVz2j0bM8pZonB1GLu6/wvvXnqum+B5ovA9HAtN9aC09Dlnyh3qdd6fUWMT756w0BMPM4j8cKUxmrfjutYcJ0lFQugUpQk5Ca8goxDDfjCtNKpS13ZbKbYAL/XgrtJXpQshyX5oHoL0eAP3R+neN1+/dTvs8tM72JIFtW3J1NSn/Mkigq7LzT1v6VI5+rWFStM8DoG1J6JHAeasfz4+UZ9cBL1Dds2aieNbm6E1Thi5tNEDl2qkWpQ5tCpfv4nmrIB4z30fLemi1JQS/cM+BKPfb3vmxkxsjQ9Da0KOccF+3z0QvwNf5Q5uSuSDFwps0GMOrkaD8baczEjFN++Xu6UpIaFuW6hzdLMRiPdyDBAmiL+Ka9hbruBZGye9v51a57UjmtDISDvsigSybhKfzyeZRcgtuN32gditlgX7tsa/UHH7gNDeipq03aVPiqSbP2c3b7UBvr5H+/T06MhVdLuyIKXwOKSGmQIzoHGjM34IY5uIuHWWYhVpzANFgwAcj5vUc/30xEhTHoETMi1asx7a8v1/3bRQadpC3zQBvr7uQBe9acpaart6/nX3Ar1nM2yqz8qzptNYvXFc1XKq2XDV0VOiM9vfth+ax27xvnIFvk4nWw+48IBZHQuUmTmNONMc+7nWwMErqV9qNm5rDBJvQPwG4INdOx+Px+TTyQpVllGu3VPLV/CnreT/NksZegIwdTfONdEO7zqSWh6N5KcujsxgSQtdEa8A8SNGxaAvlnGC5unLPGUzFXfXq7lO0KUyqWLfU6wFQqwWcjnPvUMT33AAskquXFWiRX3EatXnW1v2hfQao2pU6tClcvsRz10SKiGm760+5eyZECP5OSCbdBvFV25MLt0dKjQepmCy/N3wmegG+zh/akMzF712bxvZjk3kHPwcxCCciS+OwkrTakqW6MPGs65PsskiAOxfX0FMyk3KOVlsYJep3k59QyZymQzspzOdlP83reCS5LQ27oVG7lbJJH6gl9pU2hR/Qkfk8iR5uvUkb3bkLz5kXKSCn8D51NlJ0TU0Fy5qX4WlcGPBzuyAFwQ3++zQKDEMJurXmAGoyroYji9/p/vtQZKWv6iLbrvfPK2GuQArO/ZCgexNi9A7DBfBWPj3tA4W+3wd5fByFGKU5/dxGwOYtlK3WXDVN6G+DrJ23ef+a3dv/KeDiCnQWwZUpiFHMh3SujOaDq6ggcNKGMEG+Kiid5OWa3cfY+rhnUMV2apul0vviPmhtGYCUe/cg7607qtRpgW6dSS2XRQLybWj9nNnH5xZoTnmTinnFmjyjx7m62vFpR5+iozdF7SEtuePaPACo2QJeqIPNEX82OVoDDqTFXU9ok2dtO/sXNRugcrRrVepQc7h81fcoeW1bQvCL/RnxZtm4WsLH6z5I+ToMeXK0vHPP1/kz0QvwdfzQRle23DNGowXxFdwtGnluXEk1gaYdMar5CX0epzMKMRynoJCRk8jFlFZ891oZJep3k58gyZy8XCsjITFTdM2NGKayZa3dSpmvAz+uJfaVNoUfFJ7R4603aaM7Nx2F0KkQo3So18v3/Py+dPQGKmOtWMjfeWpvk3vQeH0PWf3Pxr2WKtRl3TmA8uFyNwMHI8+iWWlsO1Z5Pmnj+6+D5uTM2vltNLfsQCP84hk6CjVVw09a7gPFc0gAPxfNJ6siZvkQepZorrZcNU1or42Y4+kR43UqmsOXwNeyMjQRQ3w+Et7WQWN3JBq7t6D5aTmqeVLUGibo/+etn7sjhdbSNJRlc/nYqLrbQ+2WSmQEmd6Pb0Fz84lI8BqGDCinUmH3B9rsAYDWp1FIwb4zUrz+yPtCpYSOTWj3OFdXOz7t6FM0xnztIS20zwOgVgs4ubAnb/ezkHJ0Y+QFkvHYQyqWs62etTX2q3YZoGpX6tCRV609XL5N9Vt3CH6ep5gC8dfrohCfzBh7K7CvH29DD3OMfJ0/E70AX7cPbUzm0uRZayHN5KL++3pgi5L3tiVGlVzcnU/oJwI/RwtaH8R8XEVF9yXawCg5rVrd5PPX0OZkTogB+ZG//wbI62M+ygvJtVsp8+9EjbGv0MHyXVv4QeEZA+nh1ptMuMRT30LMzDTIdfghxNCdQc4CXIHeVH7vTciSfCJSkl3tY2vV3LVl+n+tOYAKtK9CzNuKSBG5EFLOnUCLiULrfn+/bhOkyBqLFA0LI3f2tZEr+tHArlXp1tUHyFm1vW0eQsqm55GVdkFkuVynSrmcXu25aor1Q4NJzkIORiPlcRWlxqV+z2A0T12IPD42o7HLxF1U9wCoLUywWEfIUvYdpNQ5wOv6ULR+t7qla22WSqcx3uvxaCQwT474gZuQ59LwiuVrmwdAk2eNQPzKBSgssbJSo9jvqClXVzs+bexTJ1BTmFST8tbmAUDNFnAkcF7uY34ZGmESL6AwiSxfUdUE7G3xrG1Df2pX/qu2KXVoY7h8m+q4thD8At1DEI9yEwo12dfnq7vxVAqTwmeiF+Dr9GECJHNp8swxiGn8Ga4VLUubmrNU+4S+CxJgNwGO8PMvI8E70/xVsqrSBkbJ6dbtJj9BkjkhJiGLBd6BRijRTsAJZejTBitl8d2oKfa1eA31hR/UuvUm7XXnzmvUl0IM2IhcfeyJvJhOKVPWTp4xDFkq8pb+m+i4s0SZeaXuHEDF9t8fKXXv8b470vtqjxJY1fX+BZrLIevfrWhO/JWPhwXJbedZsl5r6QM+NmdAiqcsFGwTp7kDcL2fm4OSCr0mz9iImnLVNHn31b28U6I16zwa82GVBJQLIY+cNf33qUj4ugcZILbxeqoU1pGj3+MwQafzbdwC7WU+xo+fpzGPzk/1pKZts1Qi4eVkZPlfGoW1boqUJ9dTIpdYse1pU1LLbp7diidl23J11fhetfepQn8aSk1hUk6vrTtgUZMFHClKL0WK0h8hvsqQomNtNC8+Q4kkpgW6bfGsbUO/alf+q3YkdW5LuHyb6rVtuYoKdXCX9/W7EH8xrbflIs3q7Zv6megF+Lp8mADJXLp49gbk4onLTMB+XZ0xqmfjDLH/XtEn9oeQoDO7LxxDWhkY1MgoOb1a3eQLtOuO+ytaf3ZDC9r6iKm/gmoWtXZZKWuPfaWxUNYafkDNW2/SxsRTdBTq5vF2urpwzTyd3dPKx+v7BlrQ0lNvDqB8+x+JmM39vG8djTzinqXmWM+evH+x/EiRcSuaE2cH5qraVu3oAz6PPIWY5FWR0uUW/+98Oq4NZee+WnPVNKG/F1Lin40Y/LmQN9yNdFxry5Z3FWTpvQgxcougXQqy/B8tMbXUECaYu2bvHI1pUAjqQ4i5nR1Z21tVvtRmqURW0/n9eEevx+3R/HoNEm4ewneHaaGsJ9AGD4BOntXqDg215upq16fuPkWbwqSc5oQIQe2xBZyvKkpPQ/PUEWju2wYlY6zkrUmbPGvb0KfaZYCqXalDRw+w2sPl21W/1BuCX5QrlkR86oFIpjgW8QT53VC+MVu6dlk3E7sAX4cPbUrm0mJZOl2QaVOMKlrMb8JdaxFTszFijB73iec5PPyiAt22MErUv1XuBEnm5HW6KRLuZkO5HC7zSXj9kjTaYqWkTbGvOfq1hR/Qxq03aWPiKW+bi5FnwaU0krHeQE4B19OPv/+OVMz9UKAxmhpyAOXo3QWc5cfDEYO0HxIaKoWfTYj3b0JzFcSMlt5+sV19gI7rwM7IOj8NsorejBRzpRNvdvKMHueqcTpzIsElW2NPxxVDyEPlR368eg/KugbwNr4Wet89lx7sKEAPwwSdRv/c8c7I0juN94GxiLl/kuqCQu2WSjRfjvcy7o+U5t9HQu5uaG25ngoCKG30AGjHh5pzdbWpjG3pU1l7UXNIC+3zAGhXEvq8ovRJ7xPXOq1DqJiAljZ71tbYr9plgGpHUue2hMu3qV7bEoJfeMYY5AWZ1cv1yCC1E3DUxK6DiVLvE7sAX5cPNSdzaUP52pmlui/SdF7i9H6NFBC3IjfBg/CcFRVo1s4oOd263eTbGfc3F65k8gnsabSg/T1X/uG44FiBbruslLXFvtLG8ANq3HqTNrpzF56zENLK7+W/F/Q2ux64rCe0O3le3qOl1ZCpnuQAyhRlfdA8dW3h/+n9vx4l9Wvn+zehuU5+XpnQfQApCbJEvqvQYJi2RHP/tEhgnCN3Tysx8D3OVZOjNQQJheehdeZE4Lrc/7fkafagr66KBNFvo7m8pR0wqCFM0P+fwr/70Vi7t0Tz9dRIcbIWuXCmkuVrm6USMcjXe1/a1Mu6m7f7XlRQFNFGD4B2fGhTrq6ay1h7n6JNYVL5PpD7XYsHQJ4+7dmtp6goXRnxWkMr0mmLZ22b+1idBqjzqT+p8/m0IVy+TXXZrhD8eYGl/XgvxLOegoxxUyMZ4CXg0tw9k4SnRvbpRyDDlmix2MLMtkLWys1Q50sppU8nVsHMzFJK483MkDZ1KPKmeAMxIof4peellN6qSj+l9IWZXYsWxqHIff+vZvYF8GFK6YEWaP7czFZF2cM3Q5nE50IuZ88gJuq6lNKfKpKeDDHZfcxsMhRy8g/E2N6L4rbvTSn9M1eW1IyQmS2CGPhd0WRwDhIw3jWzc9Hku1dK6fmKZcTM1kRM/D5m9g/ELO2EMv4/DPzRzGb297/O7+mTUhrfHe2U0hNm1g/1ze+klF4zs++jUIlrUkrvVijntsAPUkqfoTHwRkrpIDPbAyVJ3QbVeem5wvvrF2bWB1l6fgAklLfmYeQedwdKvvZZRboJMRmzAp8A7yLF1NKonx0K/D6l9FF2X2ft7/9l42r6lNJvzew4xNBdZWbTIBf50Smlj8uW08v6ZVuaWV/g/3AvEDNbJKX0GrCTmY1IKb1XvKenSCl94jStq/fvhsaPzexT4Ggzmx14PqX0g+7o5trfEFNzG9DXzPZPKV1kZkug8J7TUkp/aaVsJcre4/dvQvPBKtfX2QfMbAgStC/29WlP4EGfo88EvkBz6+iU0tt+T6l3L16XUvrEzN4GjjezLVJKt5vZXmi72FPL0M7+Tyl9aGYnI8bu1JTSUWZ2qZndhJQwH6eU3sk9u9W++riZ9Uc74OyVUnq4zH1N6ntm4CgzOw+5UK9oZs+llK4rPK+rd98I+B/yRrodzfVvIUHnC6R42Sql9EL5NwQzmyal9DcfV08gBnkDxNyeiSyiI4AHvXxfVKGfUnrSx/v+NELHFkkpjTOzv+Tn1BK0svn/bpSrZXnk/fUzJDRtgryK/laljO2Amc2JPDEe8lNno3c/FjjWzGZDio0/ppQ+zO6ra14pWca29CngA+BqM9sZraXLmtlDSFB6y8zuBXZLKb1fsbx5HuBGxK9m/OtOZvZd4LCUUqU+amabAc86j3Mj8DfU5zf1Mn+AFBFXpZSeqUI7Q0rpETPbHdjazEYi3vKslNK/KpRzIcSP7JBS+reZ/Qr4FCkhE+KpngKeSyn9t5Vy1gEz61tog9eA3yC+7RxkhFsArdVVeLVLaWzAcAPaqeN9Gvk1+iH+cruy729mZ6Mw8Wv89/vAHi5nHAr8DjjDzI4HSs9V7UJK6SEzuwStc++juj0BKbjeQmUeieSXUnA+YglgJTMb4PdvnFL6o5kdCNycUlrLzO5IKb3u99TGA/UWTLKKDTPrl1L6PHfqQ2ABM5sspXSbmS2PlAbLTUylBnRYQI8H/ppS2sYF52WBD1JK25rZlCmlj1rtxM5c3pP99vffH2n0Wy13bYxSjuZDZvZfZAGcGTHL15nZ1SgE46W8UqMzmNniSCt/MZpgv40E7xPN7D3gHaRpH1C1jGa2JNLSHpBSetXP/QnV7yvefqch69r5uXcrLdR2sfiWVhSZ2UC00C5jZocC/0JWX1ywXwT1/0vK0vR7s/53B/DblNLbLsgMB/6NFDInp5R+VqGs6wCbm9ntLnD/AU3wTyOmdDI/fiGldIjf05XwnRdo1kFC3EFoAZ7PzAallG4ys8eqKjWg0ZZmtj2yetyHFFibAWv5eH0uJ9BaXUqNQjl6tKD5GB6MXCUPgu4Xytx/hyNL30ve3w81s+uQBfCklNK/e1K2MpiYC3rNfeDfyNpzNfAxyvtxIarLjEn8Q1UlgZkNzCmB9kVeim8hAaQ/8IgzkM9nSo3uaOf7h5kt7PQuA3Yws1NSSvua2bLIWvVg8Z5W4fPC4pnwX4Zero02RmvUE8hSOYqGVfA+lDi4LEYAc7oQ9hDwR6TYGY6EZpAS4s2yBM3sfOBVM7sVWQJ/lFI6ywXekcgF/69mtndVYTGPlNLzvi7shfLTXOF//adkOb8N3JVSegMJRD9LKZ1qZs8jpvtnZvYhUmhNNIEug6+di6aU7jIpsl9BioIz0fx1EhpzZ6eUrvJ7JoagUGufMrP+KaXPUkqXm9knKDR4YzTuD0XzzTWIp6ik1IAO88MVSJB9BdXlwyhM5CAU2vWrsjTNbG1kJDnBzD5H4/RmFDL2F9RmR6WUrq9a3iblb0lRmrv/l2a2NZrzNkAhLfshBdmK+FzTSt3WhYLyqU4DVO1KHTObCvH3x/rvHZHMNi3qq+8jL4gT8srHiQFf73+eUvo1UuithsblKBQmsyoyTByXKR/KwtvrXuBzFMo3BnlYHpVSusDMFjezoZOyUgOYtENRaJMrW43lm+Axqk53GTTx9CiePEdzFRTWkt8etKcx7z1xk29rMickxJyE3MLm8D51NAqf+B5iRK6sqW7XAv5EizuKUGPsK20MP6DGrTdprzv3zPguHMgr4Wnk3v5HtNjPgjysWnK/ndgfus4BVJyvjkLWmkVy/WIFvkb7ybepjmrrA3SMqZ8MhYVt53SvQwzzdcAmuevKJt5sS66a3LW7IUvygT6uRvr39wp9ZYJnaadNYYJZmyEm+xqfQ+5GgsN5KIF21Xw3bU3s3ckz10S7S01bhSZtTJTahj7Qti2t21DW2voUbQqTcjpt3QGLmpPQl3jel/kcetDHas3/04a+VXv+K2oOl6YN4fJtqsu2hOA3ec6UyEP7PJ/DrkQeTDdM7Dr4OnwmOY+NCeHKVlM5+6aGO/duiPF6AFmp10MTxI+An6Ya3TlTSp+Z2etoIn67JppPuPb7EDN7E/h76qF1OrXuJn++H/7XzKZAGtSHEbNwOprU90UL2XMtFu//kKLsQsQYPocUZXOhSf7PKaXHuytrGaQWrJR598Mkjfp5wN5mdnpKaT/33lgJODK5x0kJmm0NP0gp/dDM3kWWxJWRR83SSEh6FjEgT+XK0ln7t82d28zWQHPIO2b2AbImrUsjS/tRKN/AjSmlv1arga8HOhu3mQeM1+tmyO3yQiQ0ruIeMD9DbfWNRZ19wMyGAzebwgT/nFJ61sweQAqiPigUcXvglymlH2b3VZhPlkJheO+hfr8VEhgGIObscpRT5OdentJzlZnNi5SRGyFBZDxSGH4XKbi/7Ec9XQuqwtoQJmhmh6M1+oOU0psmt+k1kWCTeUMejcIvS8/3E8tSmVJ62MzGppQ+KFnOtnoA1A2T+/olwDoppbdMHp9/RcqXT1GOiUEoV8vFfs8EtX62o09Z+0Ja2ukB0DYLeHfIeOtW2z3V4FlbN3KyRR/kNfNhSukIL+8/zOwqNBZ+2Aqv5nRqC5d2erWHy7cDqX0h+MXnfGRmjyM+4L9I0XFbSukxmIQ9NRyTlGJjQrqy9RRpIsaoJrmI1+omXpVRKkmzkpu8tSHur5NyfWBmZyDt9xQppV/68wciBcxvuitrxeeVXnwLzEdtsa+5Z7cz/OCnaCLfH7Xb+shC9NvC+3Wl1HrV2uDOnWOU104KvbkFxU72RYme1jCzi5C3wp3dlbU3wcymSCllrupPAC8AxyCvlx8h4XaMmf26TuHr64a6+4AzmQcjK9ApZnYFmpcPRuP1U5RM9OzuaBXKWXuumvyzXRj/E+oL30EJF3c3hbr8JKV0b5Xy1glrX5jg00hoWdTM7kSC8aIo7vu3KJRlv1RSUZzDv4BfAoc5E78GyqL/FlJArYYEuscq0u0WFZQaU6SU/uOCzPiU0rVm9jGyBG/iZR2OlC9P1V3OFlFbrq42oh19qvYwqQy5+qkzBHVJZOF+zsxuQ4nBxyBvsk9R35oHbW39YtUyTwikHoa11Il2G6DyqFupk9oQLt8OpDaE4HfynI/M7FEkx3yRU2rUlqutt8Im/Fw9ceGM5T+QVf0lxMT9jEYylzPq1vpWLN+XMaqmWLWNUiNG9biU0qNmNj+KUf3DxCrn1xFdDWivy8dQ3N/DZnYqWhiXRML9Wyj85rNUs0XJlQjXAf9MKe1fJ+1W4Vr5V2nEvh6IhJGDgDtSSqViX61jgsShKLRrQRST/Jq/+3LA31JK42oq+yrIq+bqlNJD3V3v95yNtmrc0X+viBbEPkiZ9TtUB8cDH1VdGFxpehzagu/3wOtIsJseadQNeDeldEAVul93uLJoIJpT7wYW/v/27i3UjvIMA/D7GiWbqiQWRAUVRMiFWpSCF72JUtBoFA9oPECFVCFETSVBjVQUNZUgohhbIYpHTK14YW8sVis1CTF4RDygpFRLLyyttYqi4iHaz4v3H9d2Z0fXnjUz/xze50aTlT2ZJLPXmvm+7/3/iLiH5AvQGOZ10FZsX07apWi7Oq+B9P51CdT1XA3gKWir2J3p9XGLGidD20wXa9VcDU0sbIO2n1sMFfnGWqtmluNfmr5+A4BLoe/9YlX8hRGxcrw/cT3S3+MyaJpoATRB8hz0d/oYdB0viIgVJY49D1pwdB3UDf8F1B2/CBr5L/WgkB4Kik7lE6n4ej5a0KksJgAi4s8k/wg9KBcTAGdAkybnRsScH5TrRvI4AFdA0bHbYrRW19cA/lQU4HKq+ppKD9gPpGNcCI3xb4e+Zz+Bik9zeiiYZQJgbUQsn/b6gVCh8KAy3wMkj0/neij0fbsaup8qOuCPd+GzhdOmRXM3NEiuBfDTiDgvPU9cDsWdigZUZe8rJJdgVNR5vILj7QVN190M7bAz1j1g09J9arFW0fb0c5X/25OcH5nXgWybQUxsMOMoWwm1rFI9BN/3MBqZFnNKUxrnAHivzINChedR+erXzBQ/CI05TkEjfuOcZ+3j3LHrorbrImITyY1QAe3RiNiWfv/sNzZVoCagAF1Dj0DRiHtJ3gkViOan106PiPfznGVz6rwG0vvXVdDf6ZFQgWTntNfHPdYU9Pl3NLUbxGPQeh/vQaP3LwN4NyLuG+c8Z0xqHAvFWZalG/jfQiPzRwD4qihqZL7+a4sJpvfXl0j+Evr3fwN68Cpd1EjHbXOnsrYJgLpFxNb0OXIDVOAH9HD3RBuKGkA11xRrikmlY9c+AdBUB7xuMWGsZRKzNKD2ALCTaXcukitQcQOqECXi0j9wvMrj8nWIGiL4u/l9XNSYofcTG9Qo24vQDUwxynZ7+nFAlepFUO5rc6bT/Dajmv5/ObTYzBnQh8/l0M3nOmiUa+Kq51BRGfgi93d/6tqcBmBDXVV/fnfngRxFDUZE8LvZ16VQMeIv0EP9w1D2dawxUU6LH5DcAsUPToBuEl+G4gc/grbPzb1K9TwofnQw1I0rxrlPhR5ytgJ4PSoY507diRug7sSrJB8E8GREPJRe70tRY+YE1HqMtjVbAI3PL4eysNuynWgGTV4DZY5F8mdQJ2kK+ky8ESrsf4gx16qZ+TrJRdCCvksj4qJpv2ZRTOvWt+H6J7kQu8YEHwJwfVQcE5z555/wWK3rVNYxAdC0NBGwElpf49kYc/enXOZ6TVE7EP0cauYUkZaboAm7Q6H3glUxQUSiiQmApjrgfbObBtQ7UANqPrRO39jxIJsbkj+OCiP49sN6X9gA2j/Kxl0zqv8neQ40UXAmNFWwH4DPoz0Z1c6isvDnQePiRe6v9txj7g9halT4rYhYSy1MuAoavT4FKuw8OeZxNqBj8QM2OM69uxvlvkk3mhdAW0H+BIryrIFWQP89tKhltjxxTnVdAzOKCaXeT9INbtH1vwNaX2rstWpmOd750G431wBYD+Bvoa3nNkFratw913NsCmuKCbKmnDO1ptT+OTuVs0wAHANNALwC3a/cBk0AbIi5rymSBTXFuSQiVqUft+6BeZJritVHWhqPoKbf60SouLESNXXA+6RLDSizqgyisAF826X6FbSg1b+gUba7SO6Tc5SNHc6odhkrzv21EWvIvnIUP7gK6vY+Ai3utREqcnQifkCNc98C4OqI2FLD8Vt/o1yFWSagFkOjx9dExEd5zy6vtl8DLLFWzSzHOAraYeSp9Hl6JDQdRQCfRsTFVZ1v1TiKCR4TGWOCXdPEBEBOdRWl2oDkAVCk5QKoqbe0bFEj1wSAO+Dj6WIDyqwKgylsAO0cZaOybYdD+4e/gFFG9a9QRnUpgH9H+a1HbTfYosWcqlb8mdLNx01QAWINtCjgzOzrWDuV9CF+kGOcu883ykC+Caguaes1QC0menikLS3H+PXTO7XzoO/75dAaNg9GxGvptQMi4t2ZX9M2zBwT7KqqJwCseWViUp4A6Ia+NKDMyhhUYQNo3yhbHzKq1l5VZ1/7ED9owzh33wxhAspGqAW5j4IWH90BdW4XANg+vQjflWJBV86zbaqYALBmlS00egKgG/rQgDKbxOAKG0D+UbY+ZlStHZrIvjp+YLPp8wTU0JE8BMBURPyd5DIAlwH4DVTMWgMt0H02gB0ubA1TmQkA6wZPAHRLHxpQZmUNsrCRW98zqpZHk9lXxw/MhiEVMm+Etob8AMAfADwPbe16LoCF0OfXtoj4b6bTtEzaHDWyyXkCoJvcgLKhcmEjE2dUrUo5sq+OH5j1Wypg3gngpIj4B8mHoUV334aKmStI/g5aB2p9+hpP7Jj1iCcAuskNKBsiFzYyc0bVJpUz++r4gVl/kTwJwLVQR/afAN4EsBnAgQA+g3Y/+U9EXJbpFM2sAZ4A6CY3oGxoXNhoEWdUba6cfTWzOpE8DsAVUAH+1ojYRHIjgC8APFqMn7u4adZvngDoJjegbEj2zH0CNsqouqhhc5Gyr2dB2dePSb4C7VRwJYD3AewHdVp/7aKGmZUREVtJTkGxttfST+8N4BkXNcyGIyKeTjv5FRMALmp0QET8L/3X79HWe57YMOswZ1/NrAkkj4e2ST8MwLMRsTrn+ZhZHp4AMLO28sSGWYdFxGaSe2KUfX2V5H1Q9vVeZ1/NrAoRsYXkvgCWFEUNP9iYDY8nAMysrTyxYdYDzr6aWZO8zaeZmZm1iQsbZj3h1a/NzMzMzGyIXNgw6xFnX83MzMzMbGhc2DAzMzMzMzOzztoj9wmYmZmZmZmZmZXlwoaZmZmZmZmZdZYLG2ZmZmZmZmbWWS5smJmZmZmZmVlnubBhZmZmZmZmZp3lwoaZmZmZmZmZddY3uLp2tW72fPAAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<Figure size 1080x1080 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Visualizing the correlation\n",
    "fig = plt.figure(figsize=(15,15));\n",
    "ax = fig.add_subplot(111);\n",
    "plot_corr(correlation, xnames =correlation.columns, ax = ax, );"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "id": "7e812238",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1008x432 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# have a look at feature distributions\n",
    "fig, ax = plt.subplots(figsize=(14, 6))\n",
    "sns.histplot(df_train['load_shortfall_3h'])\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "1ae54d51",
   "metadata": {},
   "source": [
    "Not all machine learning models can interpreted the *datetime* data type. For this reason, we will convert the **time** column into a numeric by further breaking it down to different column to represent the year, month, day, hour, minute and second."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "id": "d8fe8c77",
   "metadata": {},
   "outputs": [],
   "source": [
    "#converting the time column to datetime data type\n",
    "df_train['time'] = pd.to_datetime(df_train['time'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "id": "b22115f8",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Users\\PAULAB~1\\AppData\\Local\\Temp/ipykernel_21216/3067928506.py:3: FutureWarning: Series.dt.weekofyear and Series.dt.week have been deprecated. Please use Series.dt.isocalendar().week instead.\n",
      "  df_train['Week_of_Year'] = df_train['time'].dt.weekofyear\n"
     ]
    },
    {
     "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>time</th>\n",
       "      <th>Madrid_wind_speed</th>\n",
       "      <th>Valencia_wind_deg</th>\n",
       "      <th>Bilbao_rain_1h</th>\n",
       "      <th>Valencia_wind_speed</th>\n",
       "      <th>Seville_humidity</th>\n",
       "      <th>Madrid_humidity</th>\n",
       "      <th>Bilbao_clouds_all</th>\n",
       "      <th>Bilbao_wind_speed</th>\n",
       "      <th>Seville_clouds_all</th>\n",
       "      <th>...</th>\n",
       "      <th>load_shortfall_3h</th>\n",
       "      <th>Day_of_Week</th>\n",
       "      <th>Week_of_Year</th>\n",
       "      <th>Day_of_Year</th>\n",
       "      <th>Month_of_Year</th>\n",
       "      <th>Year</th>\n",
       "      <th>Day_of_Month</th>\n",
       "      <th>Hour_of_Day</th>\n",
       "      <th>Hour_of_Year</th>\n",
       "      <th>Hour_of_Week</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>2015-01-01 03:00:00</td>\n",
       "      <td>0.666667</td>\n",
       "      <td>level_5</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.666667</td>\n",
       "      <td>74.333333</td>\n",
       "      <td>64.000000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>6715.666667</td>\n",
       "      <td>3</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>2015</td>\n",
       "      <td>1</td>\n",
       "      <td>3</td>\n",
       "      <td>27</td>\n",
       "      <td>75</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>2015-01-01 06:00:00</td>\n",
       "      <td>0.333333</td>\n",
       "      <td>level_10</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.666667</td>\n",
       "      <td>78.333333</td>\n",
       "      <td>64.666667</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>4171.666667</td>\n",
       "      <td>3</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>2015</td>\n",
       "      <td>1</td>\n",
       "      <td>6</td>\n",
       "      <td>30</td>\n",
       "      <td>78</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>2015-01-01 09:00:00</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>level_9</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>71.333333</td>\n",
       "      <td>64.333333</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>4274.666667</td>\n",
       "      <td>3</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>2015</td>\n",
       "      <td>1</td>\n",
       "      <td>9</td>\n",
       "      <td>33</td>\n",
       "      <td>81</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>2015-01-01 12:00:00</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>level_8</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>65.333333</td>\n",
       "      <td>56.333333</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>5075.666667</td>\n",
       "      <td>3</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>2015</td>\n",
       "      <td>1</td>\n",
       "      <td>12</td>\n",
       "      <td>36</td>\n",
       "      <td>84</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>2015-01-01 15:00:00</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>level_7</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>59.000000</td>\n",
       "      <td>57.000000</td>\n",
       "      <td>2.0</td>\n",
       "      <td>0.333333</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>6620.666667</td>\n",
       "      <td>3</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>2015</td>\n",
       "      <td>1</td>\n",
       "      <td>15</td>\n",
       "      <td>39</td>\n",
       "      <td>87</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 57 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "                 time  Madrid_wind_speed Valencia_wind_deg  Bilbao_rain_1h  \\\n",
       "0 2015-01-01 03:00:00           0.666667           level_5             0.0   \n",
       "1 2015-01-01 06:00:00           0.333333          level_10             0.0   \n",
       "2 2015-01-01 09:00:00           1.000000           level_9             0.0   \n",
       "3 2015-01-01 12:00:00           1.000000           level_8             0.0   \n",
       "4 2015-01-01 15:00:00           1.000000           level_7             0.0   \n",
       "\n",
       "   Valencia_wind_speed  Seville_humidity  Madrid_humidity  Bilbao_clouds_all  \\\n",
       "0             0.666667         74.333333        64.000000                0.0   \n",
       "1             1.666667         78.333333        64.666667                0.0   \n",
       "2             1.000000         71.333333        64.333333                0.0   \n",
       "3             1.000000         65.333333        56.333333                0.0   \n",
       "4             1.000000         59.000000        57.000000                2.0   \n",
       "\n",
       "   Bilbao_wind_speed  Seville_clouds_all  ...  load_shortfall_3h  Day_of_Week  \\\n",
       "0           1.000000                 0.0  ...        6715.666667            3   \n",
       "1           1.000000                 0.0  ...        4171.666667            3   \n",
       "2           1.000000                 0.0  ...        4274.666667            3   \n",
       "3           1.000000                 0.0  ...        5075.666667            3   \n",
       "4           0.333333                 0.0  ...        6620.666667            3   \n",
       "\n",
       "   Week_of_Year  Day_of_Year  Month_of_Year  Year Day_of_Month  Hour_of_Day  \\\n",
       "0             1            1              1  2015            1            3   \n",
       "1             1            1              1  2015            1            6   \n",
       "2             1            1              1  2015            1            9   \n",
       "3             1            1              1  2015            1           12   \n",
       "4             1            1              1  2015            1           15   \n",
       "\n",
       "   Hour_of_Year  Hour_of_Week  \n",
       "0            27            75  \n",
       "1            30            78  \n",
       "2            33            81  \n",
       "3            36            84  \n",
       "4            39            87  \n",
       "\n",
       "[5 rows x 57 columns]"
      ]
     },
     "execution_count": 22,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#creating the temporal features \n",
    "df_train['Day_of_Week'] = df_train['time'].dt.dayofweek\n",
    "df_train['Week_of_Year'] = df_train['time'].dt.weekofyear\n",
    "df_train['Day_of_Year'] = df_train['time'].dt.dayofyear\n",
    "df_train['Month_of_Year'] = pd.DatetimeIndex(df_train['time']).month #Actual Month\n",
    "df_train['Year'] = pd.DatetimeIndex(df_train['time']).year #Actual Year\n",
    "df_train['Day_of_Month'] = pd.DatetimeIndex(df_train['time']).day #Day of month\n",
    "df_train['Hour_of_Day'] = pd.DatetimeIndex(df_train['time']).hour #Hour of day\n",
    "df_train['Hour_of_Year'] = (df_train['time'].dt.dayofyear )* 24 + df_train['time'].dt.hour #Hour of year -1\n",
    "df_train['Hour_of_Week'] = (df_train['time'].dt.dayofweek ) * 24 +  df_train['time'].dt.hour #Hour of week\n",
    "#df_train = df_train.drop('time', axis=1)\n",
    "df_train.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "id": "a95b34f2",
   "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>Valencia_wind_deg_level_10</th>\n",
       "      <th>Valencia_wind_deg_level_2</th>\n",
       "      <th>Valencia_wind_deg_level_3</th>\n",
       "      <th>Valencia_wind_deg_level_4</th>\n",
       "      <th>Valencia_wind_deg_level_5</th>\n",
       "      <th>Valencia_wind_deg_level_6</th>\n",
       "      <th>Valencia_wind_deg_level_7</th>\n",
       "      <th>Valencia_wind_deg_level_8</th>\n",
       "      <th>Valencia_wind_deg_level_9</th>\n",
       "      <th>Seville_pressure_sp10</th>\n",
       "      <th>...</th>\n",
       "      <th>Seville_pressure_sp23</th>\n",
       "      <th>Seville_pressure_sp24</th>\n",
       "      <th>Seville_pressure_sp25</th>\n",
       "      <th>Seville_pressure_sp3</th>\n",
       "      <th>Seville_pressure_sp4</th>\n",
       "      <th>Seville_pressure_sp5</th>\n",
       "      <th>Seville_pressure_sp6</th>\n",
       "      <th>Seville_pressure_sp7</th>\n",
       "      <th>Seville_pressure_sp8</th>\n",
       "      <th>Seville_pressure_sp9</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 33 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "   Valencia_wind_deg_level_10  Valencia_wind_deg_level_2  \\\n",
       "0                           0                          0   \n",
       "1                           1                          0   \n",
       "2                           0                          0   \n",
       "3                           0                          0   \n",
       "4                           0                          0   \n",
       "\n",
       "   Valencia_wind_deg_level_3  Valencia_wind_deg_level_4  \\\n",
       "0                          0                          0   \n",
       "1                          0                          0   \n",
       "2                          0                          0   \n",
       "3                          0                          0   \n",
       "4                          0                          0   \n",
       "\n",
       "   Valencia_wind_deg_level_5  Valencia_wind_deg_level_6  \\\n",
       "0                          1                          0   \n",
       "1                          0                          0   \n",
       "2                          0                          0   \n",
       "3                          0                          0   \n",
       "4                          0                          0   \n",
       "\n",
       "   Valencia_wind_deg_level_7  Valencia_wind_deg_level_8  \\\n",
       "0                          0                          0   \n",
       "1                          0                          0   \n",
       "2                          0                          0   \n",
       "3                          0                          1   \n",
       "4                          1                          0   \n",
       "\n",
       "   Valencia_wind_deg_level_9  Seville_pressure_sp10  ...  \\\n",
       "0                          0                      0  ...   \n",
       "1                          0                      0  ...   \n",
       "2                          1                      0  ...   \n",
       "3                          0                      0  ...   \n",
       "4                          0                      0  ...   \n",
       "\n",
       "   Seville_pressure_sp23  Seville_pressure_sp24  Seville_pressure_sp25  \\\n",
       "0                      0                      0                      1   \n",
       "1                      0                      0                      1   \n",
       "2                      0                      0                      1   \n",
       "3                      0                      0                      1   \n",
       "4                      0                      0                      1   \n",
       "\n",
       "   Seville_pressure_sp3  Seville_pressure_sp4  Seville_pressure_sp5  \\\n",
       "0                     0                     0                     0   \n",
       "1                     0                     0                     0   \n",
       "2                     0                     0                     0   \n",
       "3                     0                     0                     0   \n",
       "4                     0                     0                     0   \n",
       "\n",
       "   Seville_pressure_sp6  Seville_pressure_sp7  Seville_pressure_sp8  \\\n",
       "0                     0                     0                     0   \n",
       "1                     0                     0                     0   \n",
       "2                     0                     0                     0   \n",
       "3                     0                     0                     0   \n",
       "4                     0                     0                     0   \n",
       "\n",
       "   Seville_pressure_sp9  \n",
       "0                     0  \n",
       "1                     0  \n",
       "2                     0  \n",
       "3                     0  \n",
       "4                     0  \n",
       "\n",
       "[5 rows x 33 columns]"
      ]
     },
     "execution_count": 23,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#selecting and creating the dummy variables for the categorical features\n",
    "cat_var = df_train[['Valencia_wind_deg', 'Seville_pressure']]\n",
    "cat_var_dum = pd.get_dummies(cat_var, drop_first=True)\n",
    "cat_var_dum.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "id": "88af8c28",
   "metadata": {},
   "outputs": [],
   "source": [
    "time = df_train[['time']]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "id": "9055d9c4",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "2068"
      ]
     },
     "execution_count": 25,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#count the number of null values in each column\n",
    "df_train['Valencia_pressure'].isnull().sum()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "id": "fff7d729",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1008x432 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "#histogram of the feature with the missing value (Valencia_pressure)\n",
    "fig, ax = plt.subplots(figsize=(14, 6))\n",
    "sns.histplot(df_train['Valencia_pressure'], ax=ax);"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "id": "18985e81",
   "metadata": {},
   "outputs": [],
   "source": [
    "#Now let's copy the training data to a new dataframe\n",
    "df2 = df_train.copy()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "id": "d0e99687",
   "metadata": {},
   "outputs": [],
   "source": [
    "#dropping the column with Object data type - a technique to improve our model\n",
    "df2 = df2.drop(['time', 'Valencia_wind_deg', 'Seville_pressure'], axis=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "id": "ce6b11a6",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Creating a dataframe of our response or target variable\n",
    "response_y = df2['load_shortfall_3h']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "id": "e0546460",
   "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>Madrid_wind_speed</th>\n",
       "      <th>Bilbao_rain_1h</th>\n",
       "      <th>Valencia_wind_speed</th>\n",
       "      <th>Seville_humidity</th>\n",
       "      <th>Madrid_humidity</th>\n",
       "      <th>Bilbao_clouds_all</th>\n",
       "      <th>Bilbao_wind_speed</th>\n",
       "      <th>Seville_clouds_all</th>\n",
       "      <th>Bilbao_wind_deg</th>\n",
       "      <th>Barcelona_wind_speed</th>\n",
       "      <th>...</th>\n",
       "      <th>Madrid_temp_min</th>\n",
       "      <th>Day_of_Week</th>\n",
       "      <th>Week_of_Year</th>\n",
       "      <th>Day_of_Year</th>\n",
       "      <th>Month_of_Year</th>\n",
       "      <th>Year</th>\n",
       "      <th>Day_of_Month</th>\n",
       "      <th>Hour_of_Day</th>\n",
       "      <th>Hour_of_Year</th>\n",
       "      <th>Hour_of_Week</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0.666667</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.666667</td>\n",
       "      <td>74.333333</td>\n",
       "      <td>64.000000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>223.333333</td>\n",
       "      <td>6.333333</td>\n",
       "      <td>...</td>\n",
       "      <td>265.938000</td>\n",
       "      <td>3</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>2015</td>\n",
       "      <td>1</td>\n",
       "      <td>3</td>\n",
       "      <td>27</td>\n",
       "      <td>75</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>0.333333</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.666667</td>\n",
       "      <td>78.333333</td>\n",
       "      <td>64.666667</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>221.000000</td>\n",
       "      <td>4.000000</td>\n",
       "      <td>...</td>\n",
       "      <td>266.386667</td>\n",
       "      <td>3</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>2015</td>\n",
       "      <td>1</td>\n",
       "      <td>6</td>\n",
       "      <td>30</td>\n",
       "      <td>78</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>71.333333</td>\n",
       "      <td>64.333333</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>214.333333</td>\n",
       "      <td>2.000000</td>\n",
       "      <td>...</td>\n",
       "      <td>272.708667</td>\n",
       "      <td>3</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>2015</td>\n",
       "      <td>1</td>\n",
       "      <td>9</td>\n",
       "      <td>33</td>\n",
       "      <td>81</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>65.333333</td>\n",
       "      <td>56.333333</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>199.666667</td>\n",
       "      <td>2.333333</td>\n",
       "      <td>...</td>\n",
       "      <td>281.895219</td>\n",
       "      <td>3</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>2015</td>\n",
       "      <td>1</td>\n",
       "      <td>12</td>\n",
       "      <td>36</td>\n",
       "      <td>84</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>59.000000</td>\n",
       "      <td>57.000000</td>\n",
       "      <td>2.0</td>\n",
       "      <td>0.333333</td>\n",
       "      <td>0.0</td>\n",
       "      <td>185.000000</td>\n",
       "      <td>4.333333</td>\n",
       "      <td>...</td>\n",
       "      <td>280.678437</td>\n",
       "      <td>3</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>2015</td>\n",
       "      <td>1</td>\n",
       "      <td>15</td>\n",
       "      <td>39</td>\n",
       "      <td>87</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 53 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "   Madrid_wind_speed  Bilbao_rain_1h  Valencia_wind_speed  Seville_humidity  \\\n",
       "0           0.666667             0.0             0.666667         74.333333   \n",
       "1           0.333333             0.0             1.666667         78.333333   \n",
       "2           1.000000             0.0             1.000000         71.333333   \n",
       "3           1.000000             0.0             1.000000         65.333333   \n",
       "4           1.000000             0.0             1.000000         59.000000   \n",
       "\n",
       "   Madrid_humidity  Bilbao_clouds_all  Bilbao_wind_speed  Seville_clouds_all  \\\n",
       "0        64.000000                0.0           1.000000                 0.0   \n",
       "1        64.666667                0.0           1.000000                 0.0   \n",
       "2        64.333333                0.0           1.000000                 0.0   \n",
       "3        56.333333                0.0           1.000000                 0.0   \n",
       "4        57.000000                2.0           0.333333                 0.0   \n",
       "\n",
       "   Bilbao_wind_deg  Barcelona_wind_speed  ...  Madrid_temp_min  Day_of_Week  \\\n",
       "0       223.333333              6.333333  ...       265.938000            3   \n",
       "1       221.000000              4.000000  ...       266.386667            3   \n",
       "2       214.333333              2.000000  ...       272.708667            3   \n",
       "3       199.666667              2.333333  ...       281.895219            3   \n",
       "4       185.000000              4.333333  ...       280.678437            3   \n",
       "\n",
       "   Week_of_Year  Day_of_Year  Month_of_Year  Year  Day_of_Month  Hour_of_Day  \\\n",
       "0             1            1              1  2015             1            3   \n",
       "1             1            1              1  2015             1            6   \n",
       "2             1            1              1  2015             1            9   \n",
       "3             1            1              1  2015             1           12   \n",
       "4             1            1              1  2015             1           15   \n",
       "\n",
       "   Hour_of_Year  Hour_of_Week  \n",
       "0            27            75  \n",
       "1            30            78  \n",
       "2            33            81  \n",
       "3            36            84  \n",
       "4            39            87  \n",
       "\n",
       "[5 rows x 53 columns]"
      ]
     },
     "execution_count": 30,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df2 = df2.drop('load_shortfall_3h', axis=1)\n",
    "df2.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "51ab9d2b",
   "metadata": {},
   "source": [
    "Here the StandarScaler is used to resize the features of the dataframe"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "id": "8e35a3e8",
   "metadata": {},
   "outputs": [],
   "source": [
    "# create scaler object\n",
    "scaler = StandardScaler()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "id": "98929fd4",
   "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>Madrid_wind_speed</th>\n",
       "      <th>Bilbao_rain_1h</th>\n",
       "      <th>Valencia_wind_speed</th>\n",
       "      <th>Seville_humidity</th>\n",
       "      <th>Madrid_humidity</th>\n",
       "      <th>Bilbao_clouds_all</th>\n",
       "      <th>Bilbao_wind_speed</th>\n",
       "      <th>Seville_clouds_all</th>\n",
       "      <th>Bilbao_wind_deg</th>\n",
       "      <th>Barcelona_wind_speed</th>\n",
       "      <th>...</th>\n",
       "      <th>Madrid_temp_min</th>\n",
       "      <th>Day_of_Week</th>\n",
       "      <th>Week_of_Year</th>\n",
       "      <th>Day_of_Year</th>\n",
       "      <th>Month_of_Year</th>\n",
       "      <th>Year</th>\n",
       "      <th>Day_of_Month</th>\n",
       "      <th>Hour_of_Day</th>\n",
       "      <th>Hour_of_Year</th>\n",
       "      <th>Hour_of_Week</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>-0.950708</td>\n",
       "      <td>-0.362123</td>\n",
       "      <td>-0.796169</td>\n",
       "      <td>0.516117</td>\n",
       "      <td>0.270621</td>\n",
       "      <td>-1.335491</td>\n",
       "      <td>-0.501451</td>\n",
       "      <td>-0.565065</td>\n",
       "      <td>0.630823</td>\n",
       "      <td>1.932284</td>\n",
       "      <td>...</td>\n",
       "      <td>-2.309892</td>\n",
       "      <td>-0.00274</td>\n",
       "      <td>-1.709703</td>\n",
       "      <td>-1.728991</td>\n",
       "      <td>-1.602429</td>\n",
       "      <td>-1.226179</td>\n",
       "      <td>-1.675368</td>\n",
       "      <td>-1.090901</td>\n",
       "      <td>-1.731947</td>\n",
       "      <td>-0.157423</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>-1.130863</td>\n",
       "      <td>-0.362123</td>\n",
       "      <td>-0.381412</td>\n",
       "      <td>0.692953</td>\n",
       "      <td>0.298017</td>\n",
       "      <td>-1.335491</td>\n",
       "      <td>-0.501451</td>\n",
       "      <td>-0.565065</td>\n",
       "      <td>0.607959</td>\n",
       "      <td>0.630270</td>\n",
       "      <td>...</td>\n",
       "      <td>-2.261154</td>\n",
       "      <td>-0.00274</td>\n",
       "      <td>-1.709703</td>\n",
       "      <td>-1.728991</td>\n",
       "      <td>-1.602429</td>\n",
       "      <td>-1.226179</td>\n",
       "      <td>-1.675368</td>\n",
       "      <td>-0.654451</td>\n",
       "      <td>-1.730761</td>\n",
       "      <td>-0.095526</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>-0.770554</td>\n",
       "      <td>-0.362123</td>\n",
       "      <td>-0.657917</td>\n",
       "      <td>0.383491</td>\n",
       "      <td>0.284319</td>\n",
       "      <td>-1.335491</td>\n",
       "      <td>-0.501451</td>\n",
       "      <td>-0.565065</td>\n",
       "      <td>0.542632</td>\n",
       "      <td>-0.485743</td>\n",
       "      <td>...</td>\n",
       "      <td>-1.574407</td>\n",
       "      <td>-0.00274</td>\n",
       "      <td>-1.709703</td>\n",
       "      <td>-1.728991</td>\n",
       "      <td>-1.602429</td>\n",
       "      <td>-1.226179</td>\n",
       "      <td>-1.675368</td>\n",
       "      <td>-0.218001</td>\n",
       "      <td>-1.729575</td>\n",
       "      <td>-0.033629</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>-0.770554</td>\n",
       "      <td>-0.362123</td>\n",
       "      <td>-0.657917</td>\n",
       "      <td>0.118238</td>\n",
       "      <td>-0.044439</td>\n",
       "      <td>-1.335491</td>\n",
       "      <td>-0.501451</td>\n",
       "      <td>-0.565065</td>\n",
       "      <td>0.398912</td>\n",
       "      <td>-0.299741</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.576488</td>\n",
       "      <td>-0.00274</td>\n",
       "      <td>-1.709703</td>\n",
       "      <td>-1.728991</td>\n",
       "      <td>-1.602429</td>\n",
       "      <td>-1.226179</td>\n",
       "      <td>-1.675368</td>\n",
       "      <td>0.218449</td>\n",
       "      <td>-1.728389</td>\n",
       "      <td>0.028268</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>-0.770554</td>\n",
       "      <td>-0.362123</td>\n",
       "      <td>-0.657917</td>\n",
       "      <td>-0.161751</td>\n",
       "      <td>-0.017043</td>\n",
       "      <td>-1.274045</td>\n",
       "      <td>-0.894581</td>\n",
       "      <td>-0.565065</td>\n",
       "      <td>0.255192</td>\n",
       "      <td>0.816272</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.708665</td>\n",
       "      <td>-0.00274</td>\n",
       "      <td>-1.709703</td>\n",
       "      <td>-1.728991</td>\n",
       "      <td>-1.602429</td>\n",
       "      <td>-1.226179</td>\n",
       "      <td>-1.675368</td>\n",
       "      <td>0.654899</td>\n",
       "      <td>-1.727204</td>\n",
       "      <td>0.090165</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 53 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "   Madrid_wind_speed  Bilbao_rain_1h  Valencia_wind_speed  Seville_humidity  \\\n",
       "0          -0.950708       -0.362123            -0.796169          0.516117   \n",
       "1          -1.130863       -0.362123            -0.381412          0.692953   \n",
       "2          -0.770554       -0.362123            -0.657917          0.383491   \n",
       "3          -0.770554       -0.362123            -0.657917          0.118238   \n",
       "4          -0.770554       -0.362123            -0.657917         -0.161751   \n",
       "\n",
       "   Madrid_humidity  Bilbao_clouds_all  Bilbao_wind_speed  Seville_clouds_all  \\\n",
       "0         0.270621          -1.335491          -0.501451           -0.565065   \n",
       "1         0.298017          -1.335491          -0.501451           -0.565065   \n",
       "2         0.284319          -1.335491          -0.501451           -0.565065   \n",
       "3        -0.044439          -1.335491          -0.501451           -0.565065   \n",
       "4        -0.017043          -1.274045          -0.894581           -0.565065   \n",
       "\n",
       "   Bilbao_wind_deg  Barcelona_wind_speed  ...  Madrid_temp_min  Day_of_Week  \\\n",
       "0         0.630823              1.932284  ...        -2.309892     -0.00274   \n",
       "1         0.607959              0.630270  ...        -2.261154     -0.00274   \n",
       "2         0.542632             -0.485743  ...        -1.574407     -0.00274   \n",
       "3         0.398912             -0.299741  ...        -0.576488     -0.00274   \n",
       "4         0.255192              0.816272  ...        -0.708665     -0.00274   \n",
       "\n",
       "   Week_of_Year  Day_of_Year  Month_of_Year      Year  Day_of_Month  \\\n",
       "0     -1.709703    -1.728991      -1.602429 -1.226179     -1.675368   \n",
       "1     -1.709703    -1.728991      -1.602429 -1.226179     -1.675368   \n",
       "2     -1.709703    -1.728991      -1.602429 -1.226179     -1.675368   \n",
       "3     -1.709703    -1.728991      -1.602429 -1.226179     -1.675368   \n",
       "4     -1.709703    -1.728991      -1.602429 -1.226179     -1.675368   \n",
       "\n",
       "   Hour_of_Day  Hour_of_Year  Hour_of_Week  \n",
       "0    -1.090901     -1.731947     -0.157423  \n",
       "1    -0.654451     -1.730761     -0.095526  \n",
       "2    -0.218001     -1.729575     -0.033629  \n",
       "3     0.218449     -1.728389      0.028268  \n",
       "4     0.654899     -1.727204      0.090165  \n",
       "\n",
       "[5 rows x 53 columns]"
      ]
     },
     "execution_count": 32,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# convert the scaled predictor values into a dataframe\n",
    "df2 = pd.DataFrame(scaler.fit_transform(df2),columns = df2.columns)\n",
    "df2.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "36a9dbab",
   "metadata": {},
   "source": [
    "Here we use the K-nearest neighbours (KNN) Algorithm to Impute the missing valuse, KNN calculates the mean of the neighbours based on the number of neighbours specified in side pramenter here we are using (5) as the number of neighbours"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "id": "30003ed8",
   "metadata": {},
   "outputs": [],
   "source": [
    "# feature engineering on existing features\n",
    "# This also imputes missing values of the features\n",
    "imputer = KNNImputer(n_neighbors=6)\n",
    "df2 = pd.DataFrame(imputer.fit_transform(df2),columns = df2.columns)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "id": "8b39ae4b",
   "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>Madrid_wind_speed</th>\n",
       "      <th>Bilbao_rain_1h</th>\n",
       "      <th>Valencia_wind_speed</th>\n",
       "      <th>Seville_humidity</th>\n",
       "      <th>Madrid_humidity</th>\n",
       "      <th>Bilbao_clouds_all</th>\n",
       "      <th>Bilbao_wind_speed</th>\n",
       "      <th>Seville_clouds_all</th>\n",
       "      <th>Bilbao_wind_deg</th>\n",
       "      <th>Barcelona_wind_speed</th>\n",
       "      <th>...</th>\n",
       "      <th>Madrid_temp_min</th>\n",
       "      <th>Day_of_Week</th>\n",
       "      <th>Week_of_Year</th>\n",
       "      <th>Day_of_Year</th>\n",
       "      <th>Month_of_Year</th>\n",
       "      <th>Year</th>\n",
       "      <th>Day_of_Month</th>\n",
       "      <th>Hour_of_Day</th>\n",
       "      <th>Hour_of_Year</th>\n",
       "      <th>Hour_of_Week</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>-0.950708</td>\n",
       "      <td>-0.362123</td>\n",
       "      <td>-0.796169</td>\n",
       "      <td>0.516117</td>\n",
       "      <td>0.270621</td>\n",
       "      <td>-1.335491</td>\n",
       "      <td>-0.501451</td>\n",
       "      <td>-0.565065</td>\n",
       "      <td>0.630823</td>\n",
       "      <td>1.932284</td>\n",
       "      <td>...</td>\n",
       "      <td>-2.309892</td>\n",
       "      <td>-0.00274</td>\n",
       "      <td>-1.709703</td>\n",
       "      <td>-1.728991</td>\n",
       "      <td>-1.602429</td>\n",
       "      <td>-1.226179</td>\n",
       "      <td>-1.675368</td>\n",
       "      <td>-1.090901</td>\n",
       "      <td>-1.731947</td>\n",
       "      <td>-0.157423</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>-1.130863</td>\n",
       "      <td>-0.362123</td>\n",
       "      <td>-0.381412</td>\n",
       "      <td>0.692953</td>\n",
       "      <td>0.298017</td>\n",
       "      <td>-1.335491</td>\n",
       "      <td>-0.501451</td>\n",
       "      <td>-0.565065</td>\n",
       "      <td>0.607959</td>\n",
       "      <td>0.630270</td>\n",
       "      <td>...</td>\n",
       "      <td>-2.261154</td>\n",
       "      <td>-0.00274</td>\n",
       "      <td>-1.709703</td>\n",
       "      <td>-1.728991</td>\n",
       "      <td>-1.602429</td>\n",
       "      <td>-1.226179</td>\n",
       "      <td>-1.675368</td>\n",
       "      <td>-0.654451</td>\n",
       "      <td>-1.730761</td>\n",
       "      <td>-0.095526</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>-0.770554</td>\n",
       "      <td>-0.362123</td>\n",
       "      <td>-0.657917</td>\n",
       "      <td>0.383491</td>\n",
       "      <td>0.284319</td>\n",
       "      <td>-1.335491</td>\n",
       "      <td>-0.501451</td>\n",
       "      <td>-0.565065</td>\n",
       "      <td>0.542632</td>\n",
       "      <td>-0.485743</td>\n",
       "      <td>...</td>\n",
       "      <td>-1.574407</td>\n",
       "      <td>-0.00274</td>\n",
       "      <td>-1.709703</td>\n",
       "      <td>-1.728991</td>\n",
       "      <td>-1.602429</td>\n",
       "      <td>-1.226179</td>\n",
       "      <td>-1.675368</td>\n",
       "      <td>-0.218001</td>\n",
       "      <td>-1.729575</td>\n",
       "      <td>-0.033629</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>-0.770554</td>\n",
       "      <td>-0.362123</td>\n",
       "      <td>-0.657917</td>\n",
       "      <td>0.118238</td>\n",
       "      <td>-0.044439</td>\n",
       "      <td>-1.335491</td>\n",
       "      <td>-0.501451</td>\n",
       "      <td>-0.565065</td>\n",
       "      <td>0.398912</td>\n",
       "      <td>-0.299741</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.576488</td>\n",
       "      <td>-0.00274</td>\n",
       "      <td>-1.709703</td>\n",
       "      <td>-1.728991</td>\n",
       "      <td>-1.602429</td>\n",
       "      <td>-1.226179</td>\n",
       "      <td>-1.675368</td>\n",
       "      <td>0.218449</td>\n",
       "      <td>-1.728389</td>\n",
       "      <td>0.028268</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>-0.770554</td>\n",
       "      <td>-0.362123</td>\n",
       "      <td>-0.657917</td>\n",
       "      <td>-0.161751</td>\n",
       "      <td>-0.017043</td>\n",
       "      <td>-1.274045</td>\n",
       "      <td>-0.894581</td>\n",
       "      <td>-0.565065</td>\n",
       "      <td>0.255192</td>\n",
       "      <td>0.816272</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.708665</td>\n",
       "      <td>-0.00274</td>\n",
       "      <td>-1.709703</td>\n",
       "      <td>-1.728991</td>\n",
       "      <td>-1.602429</td>\n",
       "      <td>-1.226179</td>\n",
       "      <td>-1.675368</td>\n",
       "      <td>0.654899</td>\n",
       "      <td>-1.727204</td>\n",
       "      <td>0.090165</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 53 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "   Madrid_wind_speed  Bilbao_rain_1h  Valencia_wind_speed  Seville_humidity  \\\n",
       "0          -0.950708       -0.362123            -0.796169          0.516117   \n",
       "1          -1.130863       -0.362123            -0.381412          0.692953   \n",
       "2          -0.770554       -0.362123            -0.657917          0.383491   \n",
       "3          -0.770554       -0.362123            -0.657917          0.118238   \n",
       "4          -0.770554       -0.362123            -0.657917         -0.161751   \n",
       "\n",
       "   Madrid_humidity  Bilbao_clouds_all  Bilbao_wind_speed  Seville_clouds_all  \\\n",
       "0         0.270621          -1.335491          -0.501451           -0.565065   \n",
       "1         0.298017          -1.335491          -0.501451           -0.565065   \n",
       "2         0.284319          -1.335491          -0.501451           -0.565065   \n",
       "3        -0.044439          -1.335491          -0.501451           -0.565065   \n",
       "4        -0.017043          -1.274045          -0.894581           -0.565065   \n",
       "\n",
       "   Bilbao_wind_deg  Barcelona_wind_speed  ...  Madrid_temp_min  Day_of_Week  \\\n",
       "0         0.630823              1.932284  ...        -2.309892     -0.00274   \n",
       "1         0.607959              0.630270  ...        -2.261154     -0.00274   \n",
       "2         0.542632             -0.485743  ...        -1.574407     -0.00274   \n",
       "3         0.398912             -0.299741  ...        -0.576488     -0.00274   \n",
       "4         0.255192              0.816272  ...        -0.708665     -0.00274   \n",
       "\n",
       "   Week_of_Year  Day_of_Year  Month_of_Year      Year  Day_of_Month  \\\n",
       "0     -1.709703    -1.728991      -1.602429 -1.226179     -1.675368   \n",
       "1     -1.709703    -1.728991      -1.602429 -1.226179     -1.675368   \n",
       "2     -1.709703    -1.728991      -1.602429 -1.226179     -1.675368   \n",
       "3     -1.709703    -1.728991      -1.602429 -1.226179     -1.675368   \n",
       "4     -1.709703    -1.728991      -1.602429 -1.226179     -1.675368   \n",
       "\n",
       "   Hour_of_Day  Hour_of_Year  Hour_of_Week  \n",
       "0    -1.090901     -1.731947     -0.157423  \n",
       "1    -0.654451     -1.730761     -0.095526  \n",
       "2    -0.218001     -1.729575     -0.033629  \n",
       "3     0.218449     -1.728389      0.028268  \n",
       "4     0.654899     -1.727204      0.090165  \n",
       "\n",
       "[5 rows x 53 columns]"
      ]
     },
     "execution_count": 34,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df2.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "id": "bed4b300",
   "metadata": {},
   "outputs": [],
   "source": [
    "#df = pd.DataFrame(scaler.fit_transform(df), columns = df.columns)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "id": "eaf6ad67",
   "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>Madrid_wind_speed</th>\n",
       "      <th>Bilbao_rain_1h</th>\n",
       "      <th>Valencia_wind_speed</th>\n",
       "      <th>Seville_humidity</th>\n",
       "      <th>Madrid_humidity</th>\n",
       "      <th>Bilbao_clouds_all</th>\n",
       "      <th>Bilbao_wind_speed</th>\n",
       "      <th>Seville_clouds_all</th>\n",
       "      <th>Bilbao_wind_deg</th>\n",
       "      <th>Barcelona_wind_speed</th>\n",
       "      <th>...</th>\n",
       "      <th>Seville_pressure_sp23</th>\n",
       "      <th>Seville_pressure_sp24</th>\n",
       "      <th>Seville_pressure_sp25</th>\n",
       "      <th>Seville_pressure_sp3</th>\n",
       "      <th>Seville_pressure_sp4</th>\n",
       "      <th>Seville_pressure_sp5</th>\n",
       "      <th>Seville_pressure_sp6</th>\n",
       "      <th>Seville_pressure_sp7</th>\n",
       "      <th>Seville_pressure_sp8</th>\n",
       "      <th>Seville_pressure_sp9</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>-0.950708</td>\n",
       "      <td>-0.362123</td>\n",
       "      <td>-0.796169</td>\n",
       "      <td>0.516117</td>\n",
       "      <td>0.270621</td>\n",
       "      <td>-1.335491</td>\n",
       "      <td>-0.501451</td>\n",
       "      <td>-0.565065</td>\n",
       "      <td>0.630823</td>\n",
       "      <td>1.932284</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>-1.130863</td>\n",
       "      <td>-0.362123</td>\n",
       "      <td>-0.381412</td>\n",
       "      <td>0.692953</td>\n",
       "      <td>0.298017</td>\n",
       "      <td>-1.335491</td>\n",
       "      <td>-0.501451</td>\n",
       "      <td>-0.565065</td>\n",
       "      <td>0.607959</td>\n",
       "      <td>0.630270</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>-0.770554</td>\n",
       "      <td>-0.362123</td>\n",
       "      <td>-0.657917</td>\n",
       "      <td>0.383491</td>\n",
       "      <td>0.284319</td>\n",
       "      <td>-1.335491</td>\n",
       "      <td>-0.501451</td>\n",
       "      <td>-0.565065</td>\n",
       "      <td>0.542632</td>\n",
       "      <td>-0.485743</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>-0.770554</td>\n",
       "      <td>-0.362123</td>\n",
       "      <td>-0.657917</td>\n",
       "      <td>0.118238</td>\n",
       "      <td>-0.044439</td>\n",
       "      <td>-1.335491</td>\n",
       "      <td>-0.501451</td>\n",
       "      <td>-0.565065</td>\n",
       "      <td>0.398912</td>\n",
       "      <td>-0.299741</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>-0.770554</td>\n",
       "      <td>-0.362123</td>\n",
       "      <td>-0.657917</td>\n",
       "      <td>-0.161751</td>\n",
       "      <td>-0.017043</td>\n",
       "      <td>-1.274045</td>\n",
       "      <td>-0.894581</td>\n",
       "      <td>-0.565065</td>\n",
       "      <td>0.255192</td>\n",
       "      <td>0.816272</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 86 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "   Madrid_wind_speed  Bilbao_rain_1h  Valencia_wind_speed  Seville_humidity  \\\n",
       "0          -0.950708       -0.362123            -0.796169          0.516117   \n",
       "1          -1.130863       -0.362123            -0.381412          0.692953   \n",
       "2          -0.770554       -0.362123            -0.657917          0.383491   \n",
       "3          -0.770554       -0.362123            -0.657917          0.118238   \n",
       "4          -0.770554       -0.362123            -0.657917         -0.161751   \n",
       "\n",
       "   Madrid_humidity  Bilbao_clouds_all  Bilbao_wind_speed  Seville_clouds_all  \\\n",
       "0         0.270621          -1.335491          -0.501451           -0.565065   \n",
       "1         0.298017          -1.335491          -0.501451           -0.565065   \n",
       "2         0.284319          -1.335491          -0.501451           -0.565065   \n",
       "3        -0.044439          -1.335491          -0.501451           -0.565065   \n",
       "4        -0.017043          -1.274045          -0.894581           -0.565065   \n",
       "\n",
       "   Bilbao_wind_deg  Barcelona_wind_speed  ...  Seville_pressure_sp23  \\\n",
       "0         0.630823              1.932284  ...                      0   \n",
       "1         0.607959              0.630270  ...                      0   \n",
       "2         0.542632             -0.485743  ...                      0   \n",
       "3         0.398912             -0.299741  ...                      0   \n",
       "4         0.255192              0.816272  ...                      0   \n",
       "\n",
       "   Seville_pressure_sp24  Seville_pressure_sp25  Seville_pressure_sp3  \\\n",
       "0                      0                      1                     0   \n",
       "1                      0                      1                     0   \n",
       "2                      0                      1                     0   \n",
       "3                      0                      1                     0   \n",
       "4                      0                      1                     0   \n",
       "\n",
       "   Seville_pressure_sp4  Seville_pressure_sp5  Seville_pressure_sp6  \\\n",
       "0                     0                     0                     0   \n",
       "1                     0                     0                     0   \n",
       "2                     0                     0                     0   \n",
       "3                     0                     0                     0   \n",
       "4                     0                     0                     0   \n",
       "\n",
       "   Seville_pressure_sp7  Seville_pressure_sp8  Seville_pressure_sp9  \n",
       "0                     0                     0                     0  \n",
       "1                     0                     0                     0  \n",
       "2                     0                     0                     0  \n",
       "3                     0                     0                     0  \n",
       "4                     0                     0                     0  \n",
       "\n",
       "[5 rows x 86 columns]"
      ]
     },
     "execution_count": 36,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#cancatenating the df2 with the dummy categorical df2\n",
    "df2 = pd.concat([df2, cat_var_dum], axis=1)\n",
    "df2.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "id": "7d8f47fa",
   "metadata": {},
   "outputs": [],
   "source": [
    "#Rearranging the columns\n",
    "column_titles = [i for i in df2.columns]\n",
    "df2 = df2.reindex(columns = column_titles)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "id": "53ffb82c",
   "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>Madrid_wind_speed</th>\n",
       "      <th>Bilbao_rain_1h</th>\n",
       "      <th>Valencia_wind_speed</th>\n",
       "      <th>Seville_humidity</th>\n",
       "      <th>Madrid_humidity</th>\n",
       "      <th>Bilbao_clouds_all</th>\n",
       "      <th>Bilbao_wind_speed</th>\n",
       "      <th>Seville_clouds_all</th>\n",
       "      <th>Bilbao_wind_deg</th>\n",
       "      <th>Barcelona_wind_speed</th>\n",
       "      <th>...</th>\n",
       "      <th>Seville_pressure_sp23</th>\n",
       "      <th>Seville_pressure_sp24</th>\n",
       "      <th>Seville_pressure_sp25</th>\n",
       "      <th>Seville_pressure_sp3</th>\n",
       "      <th>Seville_pressure_sp4</th>\n",
       "      <th>Seville_pressure_sp5</th>\n",
       "      <th>Seville_pressure_sp6</th>\n",
       "      <th>Seville_pressure_sp7</th>\n",
       "      <th>Seville_pressure_sp8</th>\n",
       "      <th>Seville_pressure_sp9</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>-0.950708</td>\n",
       "      <td>-0.362123</td>\n",
       "      <td>-0.796169</td>\n",
       "      <td>0.516117</td>\n",
       "      <td>0.270621</td>\n",
       "      <td>-1.335491</td>\n",
       "      <td>-0.501451</td>\n",
       "      <td>-0.565065</td>\n",
       "      <td>0.630823</td>\n",
       "      <td>1.932284</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>-1.130863</td>\n",
       "      <td>-0.362123</td>\n",
       "      <td>-0.381412</td>\n",
       "      <td>0.692953</td>\n",
       "      <td>0.298017</td>\n",
       "      <td>-1.335491</td>\n",
       "      <td>-0.501451</td>\n",
       "      <td>-0.565065</td>\n",
       "      <td>0.607959</td>\n",
       "      <td>0.630270</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>-0.770554</td>\n",
       "      <td>-0.362123</td>\n",
       "      <td>-0.657917</td>\n",
       "      <td>0.383491</td>\n",
       "      <td>0.284319</td>\n",
       "      <td>-1.335491</td>\n",
       "      <td>-0.501451</td>\n",
       "      <td>-0.565065</td>\n",
       "      <td>0.542632</td>\n",
       "      <td>-0.485743</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>-0.770554</td>\n",
       "      <td>-0.362123</td>\n",
       "      <td>-0.657917</td>\n",
       "      <td>0.118238</td>\n",
       "      <td>-0.044439</td>\n",
       "      <td>-1.335491</td>\n",
       "      <td>-0.501451</td>\n",
       "      <td>-0.565065</td>\n",
       "      <td>0.398912</td>\n",
       "      <td>-0.299741</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>-0.770554</td>\n",
       "      <td>-0.362123</td>\n",
       "      <td>-0.657917</td>\n",
       "      <td>-0.161751</td>\n",
       "      <td>-0.017043</td>\n",
       "      <td>-1.274045</td>\n",
       "      <td>-0.894581</td>\n",
       "      <td>-0.565065</td>\n",
       "      <td>0.255192</td>\n",
       "      <td>0.816272</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 86 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "   Madrid_wind_speed  Bilbao_rain_1h  Valencia_wind_speed  Seville_humidity  \\\n",
       "0          -0.950708       -0.362123            -0.796169          0.516117   \n",
       "1          -1.130863       -0.362123            -0.381412          0.692953   \n",
       "2          -0.770554       -0.362123            -0.657917          0.383491   \n",
       "3          -0.770554       -0.362123            -0.657917          0.118238   \n",
       "4          -0.770554       -0.362123            -0.657917         -0.161751   \n",
       "\n",
       "   Madrid_humidity  Bilbao_clouds_all  Bilbao_wind_speed  Seville_clouds_all  \\\n",
       "0         0.270621          -1.335491          -0.501451           -0.565065   \n",
       "1         0.298017          -1.335491          -0.501451           -0.565065   \n",
       "2         0.284319          -1.335491          -0.501451           -0.565065   \n",
       "3        -0.044439          -1.335491          -0.501451           -0.565065   \n",
       "4        -0.017043          -1.274045          -0.894581           -0.565065   \n",
       "\n",
       "   Bilbao_wind_deg  Barcelona_wind_speed  ...  Seville_pressure_sp23  \\\n",
       "0         0.630823              1.932284  ...                      0   \n",
       "1         0.607959              0.630270  ...                      0   \n",
       "2         0.542632             -0.485743  ...                      0   \n",
       "3         0.398912             -0.299741  ...                      0   \n",
       "4         0.255192              0.816272  ...                      0   \n",
       "\n",
       "   Seville_pressure_sp24  Seville_pressure_sp25  Seville_pressure_sp3  \\\n",
       "0                      0                      1                     0   \n",
       "1                      0                      1                     0   \n",
       "2                      0                      1                     0   \n",
       "3                      0                      1                     0   \n",
       "4                      0                      1                     0   \n",
       "\n",
       "   Seville_pressure_sp4  Seville_pressure_sp5  Seville_pressure_sp6  \\\n",
       "0                     0                     0                     0   \n",
       "1                     0                     0                     0   \n",
       "2                     0                     0                     0   \n",
       "3                     0                     0                     0   \n",
       "4                     0                     0                     0   \n",
       "\n",
       "   Seville_pressure_sp7  Seville_pressure_sp8  Seville_pressure_sp9  \n",
       "0                     0                     0                     0  \n",
       "1                     0                     0                     0  \n",
       "2                     0                     0                     0  \n",
       "3                     0                     0                     0  \n",
       "4                     0                     0                     0  \n",
       "\n",
       "[5 rows x 86 columns]"
      ]
     },
     "execution_count": 38,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df2.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "id": "eec4eaac",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Drop All temp_max and temp_min of all the cities to avoid Multicollinearity\n",
    "df2 = df2.drop(['Seville_temp_max','Valencia_temp_max','Barcelona_temp_max','Madrid_temp_max','Bilbao_temp_max','Seville_temp_min','Valencia_temp_min','Barcelona_temp_min','Madrid_temp_min','Bilbao_temp_min'],axis=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "id": "6f010cc4",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Now, let's drop the redundant time variables\n",
    "df2 = df2.drop(['Week_of_Year','Month_of_Year', 'Day_of_Week', 'Year', 'Hour_of_Year'], axis=1)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "e9f5adad",
   "metadata": {},
   "source": [
    "### Cleaning the test dataframe and handling the missing values"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "id": "ab2f5bc4",
   "metadata": {},
   "outputs": [],
   "source": [
    "#copy the test df to a new df\n",
    "test_df = df_test.copy()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "id": "279779d1",
   "metadata": {},
   "outputs": [],
   "source": [
    "test_df['time'] = pd.to_datetime(test_df['time'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "id": "1f10bf77",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Users\\PAULAB~1\\AppData\\Local\\Temp/ipykernel_21216/858496451.py:2: FutureWarning: Series.dt.weekofyear and Series.dt.week have been deprecated. Please use Series.dt.isocalendar().week instead.\n",
      "  test_df['Week_of_Year'] = test_df['time'].dt.weekofyear\n"
     ]
    },
    {
     "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>time</th>\n",
       "      <th>Madrid_wind_speed</th>\n",
       "      <th>Valencia_wind_deg</th>\n",
       "      <th>Bilbao_rain_1h</th>\n",
       "      <th>Valencia_wind_speed</th>\n",
       "      <th>Seville_humidity</th>\n",
       "      <th>Madrid_humidity</th>\n",
       "      <th>Bilbao_clouds_all</th>\n",
       "      <th>Bilbao_wind_speed</th>\n",
       "      <th>Seville_clouds_all</th>\n",
       "      <th>...</th>\n",
       "      <th>Madrid_temp_min</th>\n",
       "      <th>Day_of_Week</th>\n",
       "      <th>Week_of_Year</th>\n",
       "      <th>Day_of_Year</th>\n",
       "      <th>Month_of_Year</th>\n",
       "      <th>Year</th>\n",
       "      <th>Day_of_Month</th>\n",
       "      <th>Hour_of_Day</th>\n",
       "      <th>Hour_of_Year</th>\n",
       "      <th>Hour_of_Week</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>8763</th>\n",
       "      <td>2018-01-01 00:00:00</td>\n",
       "      <td>5.000000</td>\n",
       "      <td>level_8</td>\n",
       "      <td>0.0</td>\n",
       "      <td>5.000000</td>\n",
       "      <td>87.000000</td>\n",
       "      <td>71.333333</td>\n",
       "      <td>20.000000</td>\n",
       "      <td>3.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>...</td>\n",
       "      <td>279.150000</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>2018</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>24</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8764</th>\n",
       "      <td>2018-01-01 03:00:00</td>\n",
       "      <td>4.666667</td>\n",
       "      <td>level_8</td>\n",
       "      <td>0.0</td>\n",
       "      <td>5.333333</td>\n",
       "      <td>89.000000</td>\n",
       "      <td>78.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>3.666667</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>...</td>\n",
       "      <td>278.150000</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>2018</td>\n",
       "      <td>1</td>\n",
       "      <td>3</td>\n",
       "      <td>27</td>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8765</th>\n",
       "      <td>2018-01-01 06:00:00</td>\n",
       "      <td>2.333333</td>\n",
       "      <td>level_7</td>\n",
       "      <td>0.0</td>\n",
       "      <td>5.000000</td>\n",
       "      <td>89.000000</td>\n",
       "      <td>89.666667</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>2.333333</td>\n",
       "      <td>6.666667</td>\n",
       "      <td>...</td>\n",
       "      <td>276.150000</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>2018</td>\n",
       "      <td>1</td>\n",
       "      <td>6</td>\n",
       "      <td>30</td>\n",
       "      <td>6</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8766</th>\n",
       "      <td>2018-01-01 09:00:00</td>\n",
       "      <td>2.666667</td>\n",
       "      <td>level_7</td>\n",
       "      <td>0.0</td>\n",
       "      <td>5.333333</td>\n",
       "      <td>93.333333</td>\n",
       "      <td>82.666667</td>\n",
       "      <td>26.666667</td>\n",
       "      <td>5.666667</td>\n",
       "      <td>6.666667</td>\n",
       "      <td>...</td>\n",
       "      <td>274.483333</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>2018</td>\n",
       "      <td>1</td>\n",
       "      <td>9</td>\n",
       "      <td>33</td>\n",
       "      <td>9</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8767</th>\n",
       "      <td>2018-01-01 12:00:00</td>\n",
       "      <td>4.000000</td>\n",
       "      <td>level_7</td>\n",
       "      <td>0.0</td>\n",
       "      <td>8.666667</td>\n",
       "      <td>65.333333</td>\n",
       "      <td>64.000000</td>\n",
       "      <td>26.666667</td>\n",
       "      <td>10.666667</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>...</td>\n",
       "      <td>280.150000</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>2018</td>\n",
       "      <td>1</td>\n",
       "      <td>12</td>\n",
       "      <td>36</td>\n",
       "      <td>12</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 56 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "                    time  Madrid_wind_speed Valencia_wind_deg  Bilbao_rain_1h  \\\n",
       "8763 2018-01-01 00:00:00           5.000000           level_8             0.0   \n",
       "8764 2018-01-01 03:00:00           4.666667           level_8             0.0   \n",
       "8765 2018-01-01 06:00:00           2.333333           level_7             0.0   \n",
       "8766 2018-01-01 09:00:00           2.666667           level_7             0.0   \n",
       "8767 2018-01-01 12:00:00           4.000000           level_7             0.0   \n",
       "\n",
       "      Valencia_wind_speed  Seville_humidity  Madrid_humidity  \\\n",
       "8763             5.000000         87.000000        71.333333   \n",
       "8764             5.333333         89.000000        78.000000   \n",
       "8765             5.000000         89.000000        89.666667   \n",
       "8766             5.333333         93.333333        82.666667   \n",
       "8767             8.666667         65.333333        64.000000   \n",
       "\n",
       "      Bilbao_clouds_all  Bilbao_wind_speed  Seville_clouds_all  ...  \\\n",
       "8763          20.000000           3.000000            0.000000  ...   \n",
       "8764           0.000000           3.666667            0.000000  ...   \n",
       "8765           0.000000           2.333333            6.666667  ...   \n",
       "8766          26.666667           5.666667            6.666667  ...   \n",
       "8767          26.666667          10.666667            0.000000  ...   \n",
       "\n",
       "      Madrid_temp_min  Day_of_Week  Week_of_Year  Day_of_Year  Month_of_Year  \\\n",
       "8763       279.150000            0             1            1              1   \n",
       "8764       278.150000            0             1            1              1   \n",
       "8765       276.150000            0             1            1              1   \n",
       "8766       274.483333            0             1            1              1   \n",
       "8767       280.150000            0             1            1              1   \n",
       "\n",
       "      Year Day_of_Month  Hour_of_Day  Hour_of_Year  Hour_of_Week  \n",
       "8763  2018            1            0            24             0  \n",
       "8764  2018            1            3            27             3  \n",
       "8765  2018            1            6            30             6  \n",
       "8766  2018            1            9            33             9  \n",
       "8767  2018            1           12            36            12  \n",
       "\n",
       "[5 rows x 56 columns]"
      ]
     },
     "execution_count": 43,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "test_df['Day_of_Week'] = test_df['time'].dt.dayofweek\n",
    "test_df['Week_of_Year'] = test_df['time'].dt.weekofyear\n",
    "test_df['Day_of_Year'] = test_df['time'].dt.dayofyear\n",
    "test_df['Month_of_Year'] = pd.DatetimeIndex(test_df['time']).month #Actual Month\n",
    "test_df['Year'] = pd.DatetimeIndex(test_df['time']).year #Actual Year\n",
    "test_df['Day_of_Month'] = pd.DatetimeIndex(test_df['time']).day #Day of month\n",
    "test_df['Hour_of_Day'] = pd.DatetimeIndex(test_df['time']).hour #Hour of day\n",
    "test_df['Hour_of_Year'] = (test_df['time'].dt.dayofyear )* 24 + test_df['time'].dt.hour #Hour of year -1\n",
    "test_df['Hour_of_Week'] = (test_df['time'].dt.dayofweek ) * 24 +  test_df['time'].dt.hour #Hour of week\n",
    "#df_train = df_train.drop('time', axis=1)\n",
    "test_df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 44,
   "id": "61499903",
   "metadata": {},
   "outputs": [],
   "source": [
    "test_time = test_df[['time']]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 45,
   "id": "9023bf53",
   "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>time</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>8763</th>\n",
       "      <td>2018-01-01 00:00:00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8764</th>\n",
       "      <td>2018-01-01 03:00:00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8765</th>\n",
       "      <td>2018-01-01 06:00:00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8766</th>\n",
       "      <td>2018-01-01 09:00:00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8767</th>\n",
       "      <td>2018-01-01 12:00:00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11678</th>\n",
       "      <td>2018-12-31 09:00:00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11679</th>\n",
       "      <td>2018-12-31 12:00:00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11680</th>\n",
       "      <td>2018-12-31 15:00:00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11681</th>\n",
       "      <td>2018-12-31 18:00:00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11682</th>\n",
       "      <td>2018-12-31 21:00:00</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>2920 rows × 1 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "                     time\n",
       "8763  2018-01-01 00:00:00\n",
       "8764  2018-01-01 03:00:00\n",
       "8765  2018-01-01 06:00:00\n",
       "8766  2018-01-01 09:00:00\n",
       "8767  2018-01-01 12:00:00\n",
       "...                   ...\n",
       "11678 2018-12-31 09:00:00\n",
       "11679 2018-12-31 12:00:00\n",
       "11680 2018-12-31 15:00:00\n",
       "11681 2018-12-31 18:00:00\n",
       "11682 2018-12-31 21:00:00\n",
       "\n",
       "[2920 rows x 1 columns]"
      ]
     },
     "execution_count": 45,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "test_time"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 46,
   "id": "c11043ba",
   "metadata": {},
   "outputs": [],
   "source": [
    "test_time = test_time.reset_index().drop([\"index\"], axis=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 47,
   "id": "dccf3a57",
   "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>time</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>2018-01-01 00:00:00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>2018-01-01 03:00:00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>2018-01-01 06:00:00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>2018-01-01 09:00:00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>2018-01-01 12:00:00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2915</th>\n",
       "      <td>2018-12-31 09:00:00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2916</th>\n",
       "      <td>2018-12-31 12:00:00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2917</th>\n",
       "      <td>2018-12-31 15:00:00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2918</th>\n",
       "      <td>2018-12-31 18:00:00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2919</th>\n",
       "      <td>2018-12-31 21:00:00</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>2920 rows × 1 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "                    time\n",
       "0    2018-01-01 00:00:00\n",
       "1    2018-01-01 03:00:00\n",
       "2    2018-01-01 06:00:00\n",
       "3    2018-01-01 09:00:00\n",
       "4    2018-01-01 12:00:00\n",
       "...                  ...\n",
       "2915 2018-12-31 09:00:00\n",
       "2916 2018-12-31 12:00:00\n",
       "2917 2018-12-31 15:00:00\n",
       "2918 2018-12-31 18:00:00\n",
       "2919 2018-12-31 21:00:00\n",
       "\n",
       "[2920 rows x 1 columns]"
      ]
     },
     "execution_count": 47,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "test_time"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 48,
   "id": "d9855e21",
   "metadata": {},
   "outputs": [],
   "source": [
    "#drop the categorical columns \n",
    "test_df = test_df.drop(['time', 'Valencia_wind_deg', 'Seville_pressure' ], axis=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 49,
   "id": "2f3ef3ad",
   "metadata": {},
   "outputs": [],
   "source": [
    "#sacling the dataFrame using the StandarScalar\n",
    "test_df = pd.DataFrame(scaler.fit_transform(test_df), columns = test_df.columns)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 50,
   "id": "9d5932aa",
   "metadata": {},
   "outputs": [],
   "source": [
    "#Imputing the missing values using Knn Imputer\n",
    "test_df = pd.DataFrame(imputer.fit_transform(test_df), columns = test_df.columns)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 51,
   "id": "bb9f9892",
   "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>Madrid_wind_speed</th>\n",
       "      <th>Bilbao_rain_1h</th>\n",
       "      <th>Valencia_wind_speed</th>\n",
       "      <th>Seville_humidity</th>\n",
       "      <th>Madrid_humidity</th>\n",
       "      <th>Bilbao_clouds_all</th>\n",
       "      <th>Bilbao_wind_speed</th>\n",
       "      <th>Seville_clouds_all</th>\n",
       "      <th>Bilbao_wind_deg</th>\n",
       "      <th>Barcelona_wind_speed</th>\n",
       "      <th>...</th>\n",
       "      <th>Madrid_temp_min</th>\n",
       "      <th>Day_of_Week</th>\n",
       "      <th>Week_of_Year</th>\n",
       "      <th>Day_of_Year</th>\n",
       "      <th>Month_of_Year</th>\n",
       "      <th>Year</th>\n",
       "      <th>Day_of_Month</th>\n",
       "      <th>Hour_of_Day</th>\n",
       "      <th>Hour_of_Year</th>\n",
       "      <th>Hour_of_Week</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1.432590</td>\n",
       "      <td>-0.440268</td>\n",
       "      <td>0.995600</td>\n",
       "      <td>0.964514</td>\n",
       "      <td>0.360022</td>\n",
       "      <td>-0.766227</td>\n",
       "      <td>0.433023</td>\n",
       "      <td>-0.612116</td>\n",
       "      <td>0.314013</td>\n",
       "      <td>0.911814</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.855293</td>\n",
       "      <td>-1.493348</td>\n",
       "      <td>-1.690057</td>\n",
       "      <td>-1.727312</td>\n",
       "      <td>-1.602745</td>\n",
       "      <td>0.0</td>\n",
       "      <td>-1.673503</td>\n",
       "      <td>-1.527525</td>\n",
       "      <td>-1.731458</td>\n",
       "      <td>-1.694498</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1.244747</td>\n",
       "      <td>-0.440268</td>\n",
       "      <td>1.162600</td>\n",
       "      <td>1.061565</td>\n",
       "      <td>0.636255</td>\n",
       "      <td>-1.422372</td>\n",
       "      <td>0.835964</td>\n",
       "      <td>-0.612116</td>\n",
       "      <td>-0.197584</td>\n",
       "      <td>1.322013</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.969818</td>\n",
       "      <td>-1.493348</td>\n",
       "      <td>-1.690057</td>\n",
       "      <td>-1.727312</td>\n",
       "      <td>-1.602745</td>\n",
       "      <td>0.0</td>\n",
       "      <td>-1.673503</td>\n",
       "      <td>-1.091089</td>\n",
       "      <td>-1.730271</td>\n",
       "      <td>-1.632732</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>-0.070152</td>\n",
       "      <td>-0.440268</td>\n",
       "      <td>0.995600</td>\n",
       "      <td>1.061565</td>\n",
       "      <td>1.119662</td>\n",
       "      <td>-1.422372</td>\n",
       "      <td>0.030083</td>\n",
       "      <td>-0.348454</td>\n",
       "      <td>-0.334010</td>\n",
       "      <td>0.911814</td>\n",
       "      <td>...</td>\n",
       "      <td>-1.198870</td>\n",
       "      <td>-1.493348</td>\n",
       "      <td>-1.690057</td>\n",
       "      <td>-1.727312</td>\n",
       "      <td>-1.602745</td>\n",
       "      <td>0.0</td>\n",
       "      <td>-1.673503</td>\n",
       "      <td>-0.654654</td>\n",
       "      <td>-1.729085</td>\n",
       "      <td>-1.570966</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>0.117691</td>\n",
       "      <td>-0.440268</td>\n",
       "      <td>1.162600</td>\n",
       "      <td>1.271842</td>\n",
       "      <td>0.829617</td>\n",
       "      <td>-0.547512</td>\n",
       "      <td>2.044786</td>\n",
       "      <td>-0.348454</td>\n",
       "      <td>0.348120</td>\n",
       "      <td>-0.113683</td>\n",
       "      <td>...</td>\n",
       "      <td>-1.389746</td>\n",
       "      <td>-1.493348</td>\n",
       "      <td>-1.690057</td>\n",
       "      <td>-1.727312</td>\n",
       "      <td>-1.602745</td>\n",
       "      <td>0.0</td>\n",
       "      <td>-1.673503</td>\n",
       "      <td>-0.218218</td>\n",
       "      <td>-1.727899</td>\n",
       "      <td>-1.509200</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>0.869062</td>\n",
       "      <td>-0.440268</td>\n",
       "      <td>2.832609</td>\n",
       "      <td>-0.086869</td>\n",
       "      <td>0.056166</td>\n",
       "      <td>-0.547512</td>\n",
       "      <td>5.066840</td>\n",
       "      <td>-0.612116</td>\n",
       "      <td>0.723291</td>\n",
       "      <td>1.322013</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.740767</td>\n",
       "      <td>-1.493348</td>\n",
       "      <td>-1.690057</td>\n",
       "      <td>-1.727312</td>\n",
       "      <td>-1.602745</td>\n",
       "      <td>0.0</td>\n",
       "      <td>-1.673503</td>\n",
       "      <td>0.218218</td>\n",
       "      <td>-1.726712</td>\n",
       "      <td>-1.447435</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 53 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "   Madrid_wind_speed  Bilbao_rain_1h  Valencia_wind_speed  Seville_humidity  \\\n",
       "0           1.432590       -0.440268             0.995600          0.964514   \n",
       "1           1.244747       -0.440268             1.162600          1.061565   \n",
       "2          -0.070152       -0.440268             0.995600          1.061565   \n",
       "3           0.117691       -0.440268             1.162600          1.271842   \n",
       "4           0.869062       -0.440268             2.832609         -0.086869   \n",
       "\n",
       "   Madrid_humidity  Bilbao_clouds_all  Bilbao_wind_speed  Seville_clouds_all  \\\n",
       "0         0.360022          -0.766227           0.433023           -0.612116   \n",
       "1         0.636255          -1.422372           0.835964           -0.612116   \n",
       "2         1.119662          -1.422372           0.030083           -0.348454   \n",
       "3         0.829617          -0.547512           2.044786           -0.348454   \n",
       "4         0.056166          -0.547512           5.066840           -0.612116   \n",
       "\n",
       "   Bilbao_wind_deg  Barcelona_wind_speed  ...  Madrid_temp_min  Day_of_Week  \\\n",
       "0         0.314013              0.911814  ...        -0.855293    -1.493348   \n",
       "1        -0.197584              1.322013  ...        -0.969818    -1.493348   \n",
       "2        -0.334010              0.911814  ...        -1.198870    -1.493348   \n",
       "3         0.348120             -0.113683  ...        -1.389746    -1.493348   \n",
       "4         0.723291              1.322013  ...        -0.740767    -1.493348   \n",
       "\n",
       "   Week_of_Year  Day_of_Year  Month_of_Year  Year  Day_of_Month  Hour_of_Day  \\\n",
       "0     -1.690057    -1.727312      -1.602745   0.0     -1.673503    -1.527525   \n",
       "1     -1.690057    -1.727312      -1.602745   0.0     -1.673503    -1.091089   \n",
       "2     -1.690057    -1.727312      -1.602745   0.0     -1.673503    -0.654654   \n",
       "3     -1.690057    -1.727312      -1.602745   0.0     -1.673503    -0.218218   \n",
       "4     -1.690057    -1.727312      -1.602745   0.0     -1.673503     0.218218   \n",
       "\n",
       "   Hour_of_Year  Hour_of_Week  \n",
       "0     -1.731458     -1.694498  \n",
       "1     -1.730271     -1.632732  \n",
       "2     -1.729085     -1.570966  \n",
       "3     -1.727899     -1.509200  \n",
       "4     -1.726712     -1.447435  \n",
       "\n",
       "[5 rows x 53 columns]"
      ]
     },
     "execution_count": 51,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "test_df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 52,
   "id": "5c70b720",
   "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>Valencia_wind_deg_level_10</th>\n",
       "      <th>Valencia_wind_deg_level_2</th>\n",
       "      <th>Valencia_wind_deg_level_3</th>\n",
       "      <th>Valencia_wind_deg_level_4</th>\n",
       "      <th>Valencia_wind_deg_level_5</th>\n",
       "      <th>Valencia_wind_deg_level_6</th>\n",
       "      <th>Valencia_wind_deg_level_7</th>\n",
       "      <th>Valencia_wind_deg_level_8</th>\n",
       "      <th>Valencia_wind_deg_level_9</th>\n",
       "      <th>Seville_pressure_sp10</th>\n",
       "      <th>...</th>\n",
       "      <th>Seville_pressure_sp23</th>\n",
       "      <th>Seville_pressure_sp24</th>\n",
       "      <th>Seville_pressure_sp25</th>\n",
       "      <th>Seville_pressure_sp3</th>\n",
       "      <th>Seville_pressure_sp4</th>\n",
       "      <th>Seville_pressure_sp5</th>\n",
       "      <th>Seville_pressure_sp6</th>\n",
       "      <th>Seville_pressure_sp7</th>\n",
       "      <th>Seville_pressure_sp8</th>\n",
       "      <th>Seville_pressure_sp9</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>8763</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8764</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8765</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8766</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8767</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 33 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "      Valencia_wind_deg_level_10  Valencia_wind_deg_level_2  \\\n",
       "8763                           0                          0   \n",
       "8764                           0                          0   \n",
       "8765                           0                          0   \n",
       "8766                           0                          0   \n",
       "8767                           0                          0   \n",
       "\n",
       "      Valencia_wind_deg_level_3  Valencia_wind_deg_level_4  \\\n",
       "8763                          0                          0   \n",
       "8764                          0                          0   \n",
       "8765                          0                          0   \n",
       "8766                          0                          0   \n",
       "8767                          0                          0   \n",
       "\n",
       "      Valencia_wind_deg_level_5  Valencia_wind_deg_level_6  \\\n",
       "8763                          0                          0   \n",
       "8764                          0                          0   \n",
       "8765                          0                          0   \n",
       "8766                          0                          0   \n",
       "8767                          0                          0   \n",
       "\n",
       "      Valencia_wind_deg_level_7  Valencia_wind_deg_level_8  \\\n",
       "8763                          0                          1   \n",
       "8764                          0                          1   \n",
       "8765                          1                          0   \n",
       "8766                          1                          0   \n",
       "8767                          1                          0   \n",
       "\n",
       "      Valencia_wind_deg_level_9  Seville_pressure_sp10  ...  \\\n",
       "8763                          0                      0  ...   \n",
       "8764                          0                      0  ...   \n",
       "8765                          0                      0  ...   \n",
       "8766                          0                      0  ...   \n",
       "8767                          0                      0  ...   \n",
       "\n",
       "      Seville_pressure_sp23  Seville_pressure_sp24  Seville_pressure_sp25  \\\n",
       "8763                      0                      0                      1   \n",
       "8764                      0                      0                      1   \n",
       "8765                      0                      0                      1   \n",
       "8766                      0                      0                      1   \n",
       "8767                      0                      0                      1   \n",
       "\n",
       "      Seville_pressure_sp3  Seville_pressure_sp4  Seville_pressure_sp5  \\\n",
       "8763                     0                     0                     0   \n",
       "8764                     0                     0                     0   \n",
       "8765                     0                     0                     0   \n",
       "8766                     0                     0                     0   \n",
       "8767                     0                     0                     0   \n",
       "\n",
       "      Seville_pressure_sp6  Seville_pressure_sp7  Seville_pressure_sp8  \\\n",
       "8763                     0                     0                     0   \n",
       "8764                     0                     0                     0   \n",
       "8765                     0                     0                     0   \n",
       "8766                     0                     0                     0   \n",
       "8767                     0                     0                     0   \n",
       "\n",
       "      Seville_pressure_sp9  \n",
       "8763                     0  \n",
       "8764                     0  \n",
       "8765                     0  \n",
       "8766                     0  \n",
       "8767                     0  \n",
       "\n",
       "[5 rows x 33 columns]"
      ]
     },
     "execution_count": 52,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#creating dummies for test categorical variables\n",
    "cat_var = df_test[['Valencia_wind_deg', 'Seville_pressure']]\n",
    "test_dummies = pd.get_dummies(cat_var, drop_first=True)\n",
    "test_dummies.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "a46bf08d",
   "metadata": {},
   "source": [
    "Resetting the index to start from zero not (8763), which will make an additional column called index, so the index column is being drop here: this is done to make the work of concatenation easrie with out having issues with NaN"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 53,
   "id": "41db3c93",
   "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>Valencia_wind_deg_level_10</th>\n",
       "      <th>Valencia_wind_deg_level_2</th>\n",
       "      <th>Valencia_wind_deg_level_3</th>\n",
       "      <th>Valencia_wind_deg_level_4</th>\n",
       "      <th>Valencia_wind_deg_level_5</th>\n",
       "      <th>Valencia_wind_deg_level_6</th>\n",
       "      <th>Valencia_wind_deg_level_7</th>\n",
       "      <th>Valencia_wind_deg_level_8</th>\n",
       "      <th>Valencia_wind_deg_level_9</th>\n",
       "      <th>Seville_pressure_sp10</th>\n",
       "      <th>...</th>\n",
       "      <th>Seville_pressure_sp23</th>\n",
       "      <th>Seville_pressure_sp24</th>\n",
       "      <th>Seville_pressure_sp25</th>\n",
       "      <th>Seville_pressure_sp3</th>\n",
       "      <th>Seville_pressure_sp4</th>\n",
       "      <th>Seville_pressure_sp5</th>\n",
       "      <th>Seville_pressure_sp6</th>\n",
       "      <th>Seville_pressure_sp7</th>\n",
       "      <th>Seville_pressure_sp8</th>\n",
       "      <th>Seville_pressure_sp9</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 33 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "   Valencia_wind_deg_level_10  Valencia_wind_deg_level_2  \\\n",
       "0                           0                          0   \n",
       "1                           0                          0   \n",
       "2                           0                          0   \n",
       "3                           0                          0   \n",
       "4                           0                          0   \n",
       "\n",
       "   Valencia_wind_deg_level_3  Valencia_wind_deg_level_4  \\\n",
       "0                          0                          0   \n",
       "1                          0                          0   \n",
       "2                          0                          0   \n",
       "3                          0                          0   \n",
       "4                          0                          0   \n",
       "\n",
       "   Valencia_wind_deg_level_5  Valencia_wind_deg_level_6  \\\n",
       "0                          0                          0   \n",
       "1                          0                          0   \n",
       "2                          0                          0   \n",
       "3                          0                          0   \n",
       "4                          0                          0   \n",
       "\n",
       "   Valencia_wind_deg_level_7  Valencia_wind_deg_level_8  \\\n",
       "0                          0                          1   \n",
       "1                          0                          1   \n",
       "2                          1                          0   \n",
       "3                          1                          0   \n",
       "4                          1                          0   \n",
       "\n",
       "   Valencia_wind_deg_level_9  Seville_pressure_sp10  ...  \\\n",
       "0                          0                      0  ...   \n",
       "1                          0                      0  ...   \n",
       "2                          0                      0  ...   \n",
       "3                          0                      0  ...   \n",
       "4                          0                      0  ...   \n",
       "\n",
       "   Seville_pressure_sp23  Seville_pressure_sp24  Seville_pressure_sp25  \\\n",
       "0                      0                      0                      1   \n",
       "1                      0                      0                      1   \n",
       "2                      0                      0                      1   \n",
       "3                      0                      0                      1   \n",
       "4                      0                      0                      1   \n",
       "\n",
       "   Seville_pressure_sp3  Seville_pressure_sp4  Seville_pressure_sp5  \\\n",
       "0                     0                     0                     0   \n",
       "1                     0                     0                     0   \n",
       "2                     0                     0                     0   \n",
       "3                     0                     0                     0   \n",
       "4                     0                     0                     0   \n",
       "\n",
       "   Seville_pressure_sp6  Seville_pressure_sp7  Seville_pressure_sp8  \\\n",
       "0                     0                     0                     0   \n",
       "1                     0                     0                     0   \n",
       "2                     0                     0                     0   \n",
       "3                     0                     0                     0   \n",
       "4                     0                     0                     0   \n",
       "\n",
       "   Seville_pressure_sp9  \n",
       "0                     0  \n",
       "1                     0  \n",
       "2                     0  \n",
       "3                     0  \n",
       "4                     0  \n",
       "\n",
       "[5 rows x 33 columns]"
      ]
     },
     "execution_count": 53,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#resetting the index and droping the index column\n",
    "test_dummies = test_dummies.reset_index().drop([\"index\"], axis=1)\n",
    "test_dummies.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 54,
   "id": "ad700743",
   "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>Madrid_wind_speed</th>\n",
       "      <th>Bilbao_rain_1h</th>\n",
       "      <th>Valencia_wind_speed</th>\n",
       "      <th>Seville_humidity</th>\n",
       "      <th>Madrid_humidity</th>\n",
       "      <th>Bilbao_clouds_all</th>\n",
       "      <th>Bilbao_wind_speed</th>\n",
       "      <th>Seville_clouds_all</th>\n",
       "      <th>Bilbao_wind_deg</th>\n",
       "      <th>Barcelona_wind_speed</th>\n",
       "      <th>...</th>\n",
       "      <th>Seville_pressure_sp23</th>\n",
       "      <th>Seville_pressure_sp24</th>\n",
       "      <th>Seville_pressure_sp25</th>\n",
       "      <th>Seville_pressure_sp3</th>\n",
       "      <th>Seville_pressure_sp4</th>\n",
       "      <th>Seville_pressure_sp5</th>\n",
       "      <th>Seville_pressure_sp6</th>\n",
       "      <th>Seville_pressure_sp7</th>\n",
       "      <th>Seville_pressure_sp8</th>\n",
       "      <th>Seville_pressure_sp9</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1.432590</td>\n",
       "      <td>-0.440268</td>\n",
       "      <td>0.995600</td>\n",
       "      <td>0.964514</td>\n",
       "      <td>0.360022</td>\n",
       "      <td>-0.766227</td>\n",
       "      <td>0.433023</td>\n",
       "      <td>-0.612116</td>\n",
       "      <td>0.314013</td>\n",
       "      <td>0.911814</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1.244747</td>\n",
       "      <td>-0.440268</td>\n",
       "      <td>1.162600</td>\n",
       "      <td>1.061565</td>\n",
       "      <td>0.636255</td>\n",
       "      <td>-1.422372</td>\n",
       "      <td>0.835964</td>\n",
       "      <td>-0.612116</td>\n",
       "      <td>-0.197584</td>\n",
       "      <td>1.322013</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>-0.070152</td>\n",
       "      <td>-0.440268</td>\n",
       "      <td>0.995600</td>\n",
       "      <td>1.061565</td>\n",
       "      <td>1.119662</td>\n",
       "      <td>-1.422372</td>\n",
       "      <td>0.030083</td>\n",
       "      <td>-0.348454</td>\n",
       "      <td>-0.334010</td>\n",
       "      <td>0.911814</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>0.117691</td>\n",
       "      <td>-0.440268</td>\n",
       "      <td>1.162600</td>\n",
       "      <td>1.271842</td>\n",
       "      <td>0.829617</td>\n",
       "      <td>-0.547512</td>\n",
       "      <td>2.044786</td>\n",
       "      <td>-0.348454</td>\n",
       "      <td>0.348120</td>\n",
       "      <td>-0.113683</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>0.869062</td>\n",
       "      <td>-0.440268</td>\n",
       "      <td>2.832609</td>\n",
       "      <td>-0.086869</td>\n",
       "      <td>0.056166</td>\n",
       "      <td>-0.547512</td>\n",
       "      <td>5.066840</td>\n",
       "      <td>-0.612116</td>\n",
       "      <td>0.723291</td>\n",
       "      <td>1.322013</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 86 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "   Madrid_wind_speed  Bilbao_rain_1h  Valencia_wind_speed  Seville_humidity  \\\n",
       "0           1.432590       -0.440268             0.995600          0.964514   \n",
       "1           1.244747       -0.440268             1.162600          1.061565   \n",
       "2          -0.070152       -0.440268             0.995600          1.061565   \n",
       "3           0.117691       -0.440268             1.162600          1.271842   \n",
       "4           0.869062       -0.440268             2.832609         -0.086869   \n",
       "\n",
       "   Madrid_humidity  Bilbao_clouds_all  Bilbao_wind_speed  Seville_clouds_all  \\\n",
       "0         0.360022          -0.766227           0.433023           -0.612116   \n",
       "1         0.636255          -1.422372           0.835964           -0.612116   \n",
       "2         1.119662          -1.422372           0.030083           -0.348454   \n",
       "3         0.829617          -0.547512           2.044786           -0.348454   \n",
       "4         0.056166          -0.547512           5.066840           -0.612116   \n",
       "\n",
       "   Bilbao_wind_deg  Barcelona_wind_speed  ...  Seville_pressure_sp23  \\\n",
       "0         0.314013              0.911814  ...                      0   \n",
       "1        -0.197584              1.322013  ...                      0   \n",
       "2        -0.334010              0.911814  ...                      0   \n",
       "3         0.348120             -0.113683  ...                      0   \n",
       "4         0.723291              1.322013  ...                      0   \n",
       "\n",
       "   Seville_pressure_sp24  Seville_pressure_sp25  Seville_pressure_sp3  \\\n",
       "0                      0                      1                     0   \n",
       "1                      0                      1                     0   \n",
       "2                      0                      1                     0   \n",
       "3                      0                      1                     0   \n",
       "4                      0                      1                     0   \n",
       "\n",
       "   Seville_pressure_sp4  Seville_pressure_sp5  Seville_pressure_sp6  \\\n",
       "0                     0                     0                     0   \n",
       "1                     0                     0                     0   \n",
       "2                     0                     0                     0   \n",
       "3                     0                     0                     0   \n",
       "4                     0                     0                     0   \n",
       "\n",
       "   Seville_pressure_sp7  Seville_pressure_sp8  Seville_pressure_sp9  \n",
       "0                     0                     0                     0  \n",
       "1                     0                     0                     0  \n",
       "2                     0                     0                     0  \n",
       "3                     0                     0                     0  \n",
       "4                     0                     0                     0  \n",
       "\n",
       "[5 rows x 86 columns]"
      ]
     },
     "execution_count": 54,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#concatenating the test dataframe and the test_dummies dataframe\n",
    "test_df = pd.concat([test_df, test_dummies], axis=1)\n",
    "test_df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 55,
   "id": "e4b589af",
   "metadata": {},
   "outputs": [],
   "source": [
    "# DROPPING REDUNDANT TIME VARIABLES\n",
    "test_df = test_df.drop(['Week_of_Year','Month_of_Year', 'Day_of_Week', 'Year', 'Hour_of_Year'], axis=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 56,
   "id": "78d28ec1",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Drop All temp_max and temp_min all the cities to avoid Multicollinearity\n",
    "test_df = test_df.drop(['Seville_temp_max','Valencia_temp_max','Barcelona_temp_max','Madrid_temp_max','Bilbao_temp_max','Seville_temp_min','Valencia_temp_min','Barcelona_temp_min','Madrid_temp_min','Bilbao_temp_min'],axis=1)\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "600712b5",
   "metadata": {},
   "source": [
    "<a id=\"five\"></a>\n",
    "## 3. Modelling\n",
    "<a class=\"anchor\" id=\"1.1\"></a>\n",
    "<a href=#cont>Back to Table of Contents</a>\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "c536aa8b",
   "metadata": {},
   "source": [
    "#### A Brief Introduction of XGBoost Machine Learning Model\n",
    "\n",
    "The XGBoost library implements the gradient boosting decision tree algorithm.\n",
    "\n",
    "This algorithm is also known as gradient boosting, multiple additive regression trees, stochastic gradient boosting or gradient boosting machines.\n",
    "\n",
    "**Boosting is an ensemble technique** where new models are added to correct the errors made by existing models. Models are added sequentially until no further improvements can be made. A popular example is the AdaBoost algorithm that weights data points that are hard to predict.\n",
    "\n",
    "Gradient boosting is an approach where new models are created that **predict the residuals or errors** of prior models and then added together to make the final prediction. It is called gradient boosting because it uses a gradient descent algorithm to minimize the loss when adding new models.\n",
    "\n",
    "This approach supports both regression and classification predictive modeling problems.\n",
    "\n",
    "**To use the XGBoost, kindly install the library using ..... \"pip install xgboost\" ..... on Conda Prompt**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 57,
   "id": "abd71384",
   "metadata": {},
   "outputs": [],
   "source": [
    "import xgboost as xgb"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 58,
   "id": "10e894cc",
   "metadata": {},
   "outputs": [],
   "source": [
    "data_dmatrix = xgb.DMatrix(data=df2,label=response_y)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 59,
   "id": "46e9d2c2",
   "metadata": {},
   "outputs": [],
   "source": [
    "#creating the features and target variables\n",
    "X = df2\n",
    "y = response_y"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 60,
   "id": "e72becb9",
   "metadata": {},
   "outputs": [],
   "source": [
    "# split data\n",
    "x_train, x_test, y_train, y_test = train_test_split(X, y, test_size=0.2, shuffle=False, random_state=123)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 61,
   "id": "e60be808",
   "metadata": {},
   "outputs": [],
   "source": [
    "# create one or more ML models\n",
    "\n",
    "# Create linear model\n",
    "lm = LinearRegression()\n",
    "\n",
    "#create Xgboost\n",
    "xg_reg = xgb.XGBRegressor(objective ='reg:squarederror', colsample_bytree = 0.7, learning_rate = 0.1,\n",
    "                max_depth = 5, subsample = 0.7, alpha = 20, n_estimators = 100)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 62,
   "id": "e0601380",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "XGBRegressor(alpha=20, base_score=0.5, booster='gbtree', callbacks=None,\n",
       "             colsample_bylevel=1, colsample_bynode=1, colsample_bytree=0.7,\n",
       "             early_stopping_rounds=None, enable_categorical=False,\n",
       "             eval_metric=None, gamma=0, gpu_id=-1, grow_policy='depthwise',\n",
       "             importance_type=None, interaction_constraints='',\n",
       "             learning_rate=0.1, max_bin=256, max_cat_to_onehot=4,\n",
       "             max_delta_step=0, max_depth=5, max_leaves=0, min_child_weight=1,\n",
       "             missing=nan, monotone_constraints='()', n_estimators=100, n_jobs=0,\n",
       "             num_parallel_tree=1, predictor='auto', random_state=0,\n",
       "             reg_alpha=20, ...)"
      ]
     },
     "execution_count": 62,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Fit the Ml model\n",
    "lm.fit(x_train, y_train)\n",
    "\n",
    "xg_reg.fit(x_train, y_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 63,
   "id": "a9537bc9",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "linear Model : 22543762.95953021\n",
      "Xgboost Model : 4906199.818129918\n"
     ]
    }
   ],
   "source": [
    "# evaluate our selected ML models\n",
    "train_lm = lm.predict(x_train)\n",
    "xgboost = xg_reg.predict(x_train)\n",
    "\n",
    "print('linear Model :', metrics.mean_squared_error(y_train, train_lm))\n",
    "print('Xgboost Model :', metrics.mean_squared_error(y_train, xgboost))"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "5b012e8c",
   "metadata": {},
   "source": [
    "<a id=\"six\"></a>\n",
    "## 4. Model Performance\n",
    "<a class=\"anchor\" id=\"1.1\"></a>\n",
    "<a href=#cont>Back to Table of Contents</a>\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 64,
   "id": "3d9ee38a",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "linear Model : 23085374.82402318\n",
      "Xgboost:  13886754.58592376\n"
     ]
    }
   ],
   "source": [
    "# Compare model performance\n",
    "test_lm = lm.predict(x_test)\n",
    "test_xgboost = xg_reg.predict(x_test)\n",
    "\n",
    "print('linear Model :', metrics.mean_squared_error(y_test, test_lm))\n",
    "print(\"Xgboost: \", metrics.mean_squared_error(y_test, test_xgboost))\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 65,
   "id": "72d951ca",
   "metadata": {},
   "outputs": [],
   "source": [
    "new_prediction = xg_reg.predict(test_df)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 66,
   "id": "a45c61d6",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([ 6832.12  ,  2390.1204,  2771.9375, ..., 13759.705 , 14030.099 ,\n",
       "       15509.206 ], dtype=float32)"
      ]
     },
     "execution_count": 66,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#test_df = test_df.drop('load_shortfall_3h', axis=1)\n",
    "new_prediction"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 67,
   "id": "20afaad9",
   "metadata": {},
   "outputs": [],
   "source": [
    "test_df['load_shortfall_3h'] = new_prediction"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 68,
   "id": "83023db3",
   "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>time</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>2018-01-01 00:00:00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>2018-01-01 03:00:00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>2018-01-01 06:00:00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>2018-01-01 09:00:00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>2018-01-01 12:00:00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2915</th>\n",
       "      <td>2018-12-31 09:00:00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2916</th>\n",
       "      <td>2018-12-31 12:00:00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2917</th>\n",
       "      <td>2018-12-31 15:00:00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2918</th>\n",
       "      <td>2018-12-31 18:00:00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2919</th>\n",
       "      <td>2018-12-31 21:00:00</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>2920 rows × 1 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "                    time\n",
       "0    2018-01-01 00:00:00\n",
       "1    2018-01-01 03:00:00\n",
       "2    2018-01-01 06:00:00\n",
       "3    2018-01-01 09:00:00\n",
       "4    2018-01-01 12:00:00\n",
       "...                  ...\n",
       "2915 2018-12-31 09:00:00\n",
       "2916 2018-12-31 12:00:00\n",
       "2917 2018-12-31 15:00:00\n",
       "2918 2018-12-31 18:00:00\n",
       "2919 2018-12-31 21:00:00\n",
       "\n",
       "[2920 rows x 1 columns]"
      ]
     },
     "execution_count": 68,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data = test_time\n",
    "data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 69,
   "id": "45e86988",
   "metadata": {},
   "outputs": [],
   "source": [
    "data = pd.concat([data, test_df[['load_shortfall_3h']]], axis=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 70,
   "id": "f30dfa40",
   "metadata": {},
   "outputs": [],
   "source": [
    "data.to_csv('Team JM3 Model Sub', header=True, index=False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 71,
   "id": "755322e3",
   "metadata": {},
   "outputs": [],
   "source": [
    "import pickle\n",
    "\n",
    "model_save_path = \"new_test_run.pkl\"\n",
    "with open(model_save_path,'wb') as file:\n",
    "    pickle.dump(xg_reg,file)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "71bb8698",
   "metadata": {},
   "source": [
    "<a id=\"seven\"></a>\n",
    "## Conclusion\n",
    "<a class=\"anchor\" id=\"1.1\"></a>\n",
    "<a href=#cont>Back to Table of Contents</a>\n",
    "\n",
    "\n",
    "Our model solution produced results that were very close to the original targeted variable in the train set. It took care of outlier and made sure the results were of above average quality. We are therefore confident that our solution can successfully be used to predict the renewable energy shortfall of Spain.\n",
    "\n",
    "We however recommend that further testing be done on the other machine learning models that we did not cover due to time limitations so establish if there could be a better solution. We also recommend that the API be developed further such that it can take in arbitrary data that can be used o make future predictions"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "f2c7a12a",
   "metadata": {},
   "source": [
    " <a id=\"app\"></a>\n",
    " ## Appendix\n",
    "\n",
    "<a href=#cont>Back to Table of Contents</a>\n",
    "\n",
    "1. Problem statement: https://www.kaggle.com/c/spain-electricity-shortfall-challenge/overview\n",
    "2. What is Exploratory Data Analysis?: https://towardsdatascience.com/exploratory-data-analysis-8fc1cb20fd15\n",
    "3. Statistics How To: https://www.statisticshowto.com/univariate/\n",
    "4. Multivariate analysis: an overview:https://s4be.cochrane.org/blog/2021/09/09/multivariate-analysis-an-overview/#:~:text=Multivariate%20analysis%20is%20defined%20as,and%20their%20structure%20are%20important\n",
    "5. :K-Nearest Neighbor(KNN) Algorithm for Machine Learning: https://www.javatpoint.com/k-nearest-neighbor-algorithm-for-machine-learning"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3 (ipykernel)",
   "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": "3.9.7"
  },
  "latex_envs": {
   "LaTeX_envs_menu_present": true,
   "autoclose": false,
   "autocomplete": true,
   "bibliofile": "biblio.bib",
   "cite_by": "apalike",
   "current_citInitial": 1,
   "eqLabelWithNumbers": true,
   "eqNumInitial": 1,
   "hotkeys": {
    "equation": "Ctrl-E",
    "itemize": "Ctrl-I"
   },
   "labels_anchors": false,
   "latex_user_defs": false,
   "report_style_numbering": false,
   "user_envs_cfg": false
  },
  "toc": {
   "base_numbering": 1,
   "nav_menu": {},
   "number_sections": true,
   "sideBar": true,
   "skip_h1_title": false,
   "title_cell": "Table of Contents",
   "title_sidebar": "Contents",
   "toc_cell": false,
   "toc_position": {},
   "toc_section_display": true,
   "toc_window_display": false
  },
  "varInspector": {
   "cols": {
    "lenName": 16,
    "lenType": 16,
    "lenVar": 40
   },
   "kernels_config": {
    "python": {
     "delete_cmd_postfix": "",
     "delete_cmd_prefix": "del ",
     "library": "var_list.py",
     "varRefreshCmd": "print(var_dic_list())"
    },
    "r": {
     "delete_cmd_postfix": ") ",
     "delete_cmd_prefix": "rm(",
     "library": "var_list.r",
     "varRefreshCmd": "cat(var_dic_list()) "
    }
   },
   "types_to_exclude": [
    "module",
    "function",
    "builtin_function_or_method",
    "instance",
    "_Feature"
   ],
   "window_display": false
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}