{ "cells": [ { "cell_type": "markdown", "metadata": {}, "source": [ "# 080 Curve Fitting with Scikit Learn\n", "\n", "> COM6018\n", "\n", "*Copyright © 2023, 2024 Jon Barker, University of Sheffield. All rights reserved*.\n" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "## Introduction\n", "\n", "In this lab we will be using linear regression to fit a model to the atmospheric gas concentration data that we have been working with in this module.\n", "\n", "The lab assumes that you have read and understood the lecture notes, Curve Fitting with Scikit learn. We will be using ideas from these notes in the lab.\n", "\n", "In the cell below, we will import some of the libraries that we will be using in the lab.\n" ] }, { "cell_type": "code", "execution_count": 1, "metadata": {}, "outputs": [], "source": [ "import pandas as pd\n", "import numpy as np\n", "import matplotlib.pyplot as plt" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "## Step 1 - Load the data\n", "\n", "We will start by loading the data from the file `co2.csv` into a Pandas DataFrame. The data is in the same format as the data we used in the previous lab. \n", "\n", "You can use the `read_csv` method. The file contains comment lines that start with '%' so you will need to use the `comment` parameter of the `read_csv` method to ignore these lines.\n", "\n", "Read the data such that the DataFrame columns are called 'year', 'month', 'day' and 'co2'. The csv file also contains columns called 'NB' and 'scale', but we will ignore these. You can use the `drop` method to remove these columns from the DataFrame. Store the DataFrame in a variable called `co2_df`.\n", "\n", "Write the code below.\n" ] }, { "cell_type": "code", "execution_count": 12, "metadata": {}, "outputs": [], "source": [ "# SOLUTION\n", "\n", "co2_df = pd.read_csv('data/co2.csv', comment='%', names=['year','month','day','co2','NB','scale', 'sta'], skipinitialspace=True)\n", "\n", "co2_df = co2_df.drop(['NB', 'scale'], axis=1)\n" ] }, { "cell_type": "code", "execution_count": 13, "metadata": {}, "outputs": [ { "data": { "text/html": [ "
\n", "\n", "\n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", "
yearmonthdayco2sta
0195811NaNmlo
1195812NaNmlo
2195813NaNmlo
3195814NaNmlo
4195815NaNmlo
..................
2440620241027NaNmlo
2440720241028NaNmlo
2440820241029NaNmlo
2440920241030NaNmlo
2441020241031NaNmlo
\n", "

24411 rows × 5 columns

\n", "
" ], "text/plain": [ " year month day co2 sta\n", "0 1958 1 1 NaN mlo\n", "1 1958 1 2 NaN mlo\n", "2 1958 1 3 NaN mlo\n", "3 1958 1 4 NaN mlo\n", "4 1958 1 5 NaN mlo\n", "... ... ... ... ... ...\n", "24406 2024 10 27 NaN mlo\n", "24407 2024 10 28 NaN mlo\n", "24408 2024 10 29 NaN mlo\n", "24409 2024 10 30 NaN mlo\n", "24410 2024 10 31 NaN mlo\n", "\n", "[24411 rows x 5 columns]" ] }, "execution_count": 13, "metadata": {}, "output_type": "execute_result" } ], "source": [ "co2_df" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "## Step 2 - Prepare the data\n", "\n", "We need to clean the data a little before we use it. For some of the rows the `co2` value is missing and the file contains a `NaN` value. We want to drop all of these rows from the DataFrame. We have done this in previous labs, so look back at your notes if you need to.\n", "\n", "We also want to convert the year, month and day columns into a single column which we will call `decimal_year` which is the year plus the fraction of the year. For example, 1st January 2019 would be 2019.0 and 1st July 2019 would be 2019.5. We will later use this column as the x-axis when we plot the data. Think carefully to work out the formula for combining the year, month and day to make this value. \n", "\n", "(Note that you can cheat a little and pretend that each month is 1/12 of a year. This is not exactly correct, but it will be close enough for our purposes.)\n", "\n", "Write the code for these two steps in the cell below and run the test cell to check your code." ] }, { "cell_type": "code", "execution_count": 19, "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "(18056, 6)\n" ] } ], "source": [ "# SOLUTION\n", "\n", "co2_df = co2_df[~co2_df.co2.isna()]\n", "\n", "co2_df['decimal_year'] = co2_df.year + (co2_df.month - 1) / 12 + (co2_df.day - 1) / 365\n", "\n", "print(co2_df.shape)" ] }, { "cell_type": "code", "execution_count": 20, "metadata": { "scrolled": true }, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "All tests passed!\n" ] } ], "source": [ "# TEST\n", "\n", "assert co2_df.shape == (18056, 6), \"The dataframe has the wrong shape\"\n", "assert set(co2_df.columns) == {'year', 'month', 'day', 'co2', 'decimal_year', 'sta'}, \"The dataframe has the wrong columns\"\n", "print('All tests passed!')" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "## Step 3 - Plotting the data\n", "\n", "We will now make a simple plot of the CO2 concentration versus the decimal year. We can use the DataFrame's `plot` method to do this. Set the `x` and `y` parameters to the column names of the DataFrame that you want to plot. You can use the `figsize` parameter to make the plot larger. A size of 10 by 6 should be suitable.\n", "\n", "Write the line of code in the cell below and run it to produce the plot." ] }, { "cell_type": "code", "execution_count": 21, "metadata": {}, "outputs": [ { "data": { "text/plain": [ "" ] }, "execution_count": 21, "metadata": {}, "output_type": "execute_result" }, { "data": { "image/png": "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", "text/plain": [ "
" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "# SOLUTION \n", "\n", "co2_df.plot(x='decimal_year', y='co2', figsize=(10, 6))" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "This should produce the famous 'Keeling curve'. You will see that the carbon dioxide concentration has been increasing steadily since the earliest date in the data and also that there is a seasonal variation, i.e., the curve has a periodic fluctuation with a period of one year.\n", "\n", "We are going to start by using a low-order polynomial to fit the basic curve and we will then introduce periodic terms to model the seasonal fluctuations. Once we have fitted the model, we will use it to try and predict the date that the CO2 concentration will first reach 450 ppm. " ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "## Step 4 - Fitting a polynomial curve using scikit-learn\n", "\n", "### Step 4.1 - Fitting a straight line\n", "\n", "We will start by fitting a straight line to the data using the `LinearRegression` class from `sklearn.linear_model`. \n", "\n", "Import `LinearRegression` from `sklearn.linear_model` and make an instance of the class which you can call `model`. We will then use the `model.fit` method. The first parameter should be the `decimal_year` values (i.e., the x-axis values) and the second parameter should be the `co2` values (i.e., the y-axis values). You can get these from the DataFrame. \n", "\n", "An important detail: the fit method requires that the x-axis values be passed as a 2-D array, i.e., you need to ensure that the shape of the `decimal_year` array is `(n,1)` rather than `(n,)`. You can do this using the NumPy `reshape` method or the `expand_dims` method. See the lecture notes for an example of how to do this.\n", "\n", "After calling the fit method, you can use the model to predict CO2 concentrations by calling the `predict` method and passing the `decimal_year` values. Store the result in a variable called `co2_predicted`.\n", "\n", "Then make a plot of the original `co2` values and the `co2_predicted` values plotted against the date. Plot these on the same axis so that you can compare the straight-line fit with the actual values.\n" ] }, { "cell_type": "code", "execution_count": 22, "metadata": {}, "outputs": [ { "data": { "text/plain": [ "[]" ] }, "execution_count": 22, "metadata": {}, "output_type": "execute_result" }, { "data": { "image/png": "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", "text/plain": [ "
" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "# SOLUTION\n", "\n", "import matplotlib\n", "from sklearn.linear_model import LinearRegression\n", "co2 = co2_df['co2'].array\n", "date = co2_df['decimal_year'].array.reshape(-1,1)\n", "model = LinearRegression()\n", "model.fit(date, co2)\n", "co2_predicted = model.predict(date)\n", "\n", "plt.figure(figsize=(12, 8))\n", "plt.plot(date, co2, label='CO2 concentration')\n", "plt.plot(date, co2_predicted, label='CO2 concentration')" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "If you have done everything correctly, you should see that the straight line is a reasonable fit to the data except that it does not capture the period component and that it is too steep in the 1960s and not steep enough in the 2000s and 2010s. The pace of the CO2 concentration increase has been increasing over time. \n", "\n", "We can make a better fit using a higher-order polynomial with $x^2$ and $x^3$ terms. We will do this in the next step." ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "### Step 4.2 - Fitting a polynomial curve\n", "\n", "We will now fit a polynomial curve to the data. We will use the `PolynomialFeatures` class from `sklearn.preprocessing` to generate the polynomial features. We will then use the `LinearRegression` class to fit the model.\n", "\n", "Import `PolynomialFeatures` and `LinearRegression` from `sklearn.preprocessing` and `sklearn.linear_model` respectively. Make an instance of the `PolynomialFeatures` class and call it `poly`. Set the `degree` parameter to 3. Pass the `decimal_year` values to the `poly.fit_transform` method and store the result in a variable called `X`. (We will use a capital X to emphasise that this is a matrix, but note that, in general, variables starting with capital letters should only be used for the names of classes).\n", "\n", "The `poly.fit_transform` method returns a matrix with 4 columns. The first column is the constant term, the second column is the $x^1$ term, the third column is the $x^2$ term, and the fourth column is the $x^3$ term. We will use this matrix as the input to the `LinearRegression` `fit` method.\n", "\n", "Make an instance of the `LinearRegression` class and call it `model`. Call the `model.fit` method and pass the `X` matrix and the `co2` values, just as you did in the previous step.\n", "\n", "Now we can call `model.predict` and pass the `X` matrix again to get the predicted CO2 concentration values. Store the result in a variable called `co2_predicted`.\n", "\n", "Make a plot of the original `co2` values and the `co2_predicted` values plotted against `date`. Plot these on the same axis so that you can compare the straight-line fit with the actual values.\n" ] }, { "cell_type": "code", "execution_count": 33, "metadata": {}, "outputs": [ { "data": { "text/plain": [ "[]" ] }, "execution_count": 33, "metadata": {}, "output_type": "execute_result" }, { "data": { "image/png": "iVBORw0KGgoAAAANSUhEUgAAA9oAAAKTCAYAAADmN3BXAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjguNCwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy8fJSN1AAAACXBIWXMAAA9hAAAPYQGoP6dpAADaQ0lEQVR4nOzdd3hUVdvv8d/09ARC7x2kC6igWBAUAREVO2LD3ntXEAv23nsBRUXxwQYCioh0kN57DS2kl2n7+SNkksmUTEKAEL6fc3Gdvda+19or532P5GY1k2EYhgAAAAAAQIUwH+kBAAAAAABQlZBoAwAAAABQgUi0AQAAAACoQCTaAAAAAABUIBJtAAAAAAAqEIk2AAAAAAAViEQbAAAAAIAKZD3SAygPr9erHTt2KD4+XiaT6UgPBwAAAABQxRmGoczMTNWrV09mc/g566My0d6xY4caNmx4pIcBAAAAADjGbN26VQ0aNAgbc1Qm2vHx8ZIKfsCEhIQjPBoAAAAAQFWXkZGhhg0b+vLRcI7KRLtwuXhCQgKJNgAAAADgsIlk+zKHoQEAAAAAUIFItAEAAAAAqEAk2gAAAAAAVCASbQAAAAAAKhCJNgAAAAAAFYhEGwAAAACACkSiDQAAAABABSLRBgAAAACgApFoAwAAAABQgUi0AQAAAACoQCTaAAAAAABUIBJtAAAAAAAqEIk2AAAAAAAViEQbAAAAAIAKRKINAAAAAEAFItEGAAAAAKACkWgDAAAAAFCBSLQBAAAAAKhAJNoAAAAAAFQgEm0AAAAAACoQiTYAAAAAABWIRBsAAAAAgApEog0AAAAAQAUi0QYAAAAAoAKRaAMAAAAADrk/lqforrH/KTvffaSHcshZj/QAAAAAAABV341fLZAkNawWo/v7tj7Cozm0mNEGAAAAABw2ezLzj/QQDjkSbQAAAADAYWM2m470EA45Em0AAAAAwCHjdHt133eLfWXDMI7gaA4PEm0AAAAAwCEzdt4W/bBwW7HyVv303/YjOKJDj0QbAAAAAHDI7MtyBtTd/e2iwz+Qw4hEGwAAAABwyJhCbMl2e7yHdyCHEYk2AAAAAOCw25Kac6SHcMiQaAMAAAAADrszX/m7ys5qk2gDAAAAAA4Zk0Jf55WV7z6MIzl8SLQBAAAAAIfEFzM36bUpa0K+91bRm75ItAEAAAAAh8TwCcvDvvdW0Tu1SbQBAAAAAEcEiTYAAAAAABXI7SHRBgAAAACgwpBoAwAAAABQgdxervcCAAAAAKDCsEcbAAAAAIAK9PPinUd6CIcEiTYAAAAAoMw8XkOfzNio1SmZQd/nOj2l9vHG1LUVPaxKgUQbAAAAAFBm7/+9Xk//skJ9X58e9P1zv62MrKMZr0nbF1bgyI4865EeAAAAAADg6JKR59JLk1b7yl6vIbPZ5Bfz56rdpfZzp+VHaco4yZEo3TFfiqtV4WM9EpjRBgAAAACUSa+XpvmVs53uiNqtfuYc9TmuIJm+w/Kj7rWNK3hx2n1VJsmWSLQBAAAAAGW0L9vpV96RlhcQYzIFVMlhtei1Szvrdst43VeYZPd5SjrlrkMxzCOGRBsAAAAAcFDmbUqNODZ+7hu63/a9JOl512VSz7sP0aiOHBJtAAAAAMBB+X7+1oC6kjPaN5zaVPr7JenPpyVJL7gu0/ue8w7H8A47DkMDAAAAAByUxdvSA+q2pub6lR+L/Vn66zlJ0ouuS/VeFU2yJWa0AQAAAAAVbM6GfX7luyw/SNMKkmz1GaF2lz7le/frkp2Hc2iHBYk2AAAAAKBCfTFr04EnQ/dYx+ke2w8FxbNGSj3vUaeGib7Y276uWndoSyTaAAAAAIAyyM4PfpWX12v4nn9bmqLCJPsu648FlWc97Ttd3BTsSPIqhEQbAAAAABCRFTsy1G74pKDvtqcV35Nt6D7r97rLOl6S9LRriHTKnb635qqdZ5NoAwAAAAAiM+Tj2SHfjZ69ueDBMPSA9VvdYf1JkjTSNVTj7Of7xZqZ0QYAAAAAQNqf4wr5bsm2dMkwZEx5SrdZJ0iSnnIN1aeefjqnXR2/2JJp9qa92RU91COKRBsAAAAAjnF5Lo9mb9gnl8db7j5mbdgrTRku07+vSZKGu67WZ55+kqTjGyX5xZbco33Gy9PK/d3KiHu0AQAAAOAY9+C4JZqweIckadPzA8rRg6FHrF9L//4qSXrCdY2+8pwtSepzXC1d0q2hX3SwleN7MvNVM95Rjm9XPsxoAwAAAMAxaFdGnqas2KUNe7J8SbYkrdyZEXEf/TvUkWToMesY3WQtSLLfjr7Fl2RL0iXdGspc4vQzw1CAqpJkS8xoAwAAAMAxqecLf8rlCcx43/pzrd4d0jWiPuLsFj1uHa3rrb8XVAx4Vd9PayYpxxdjCXLEuDdYpl2FMKMNAAAAAMegYEm2VHgHdqD92c4SNYZGOIqS7Edcw6QThmnzvhy/qBObVg/oy+P1/3a7egkRjvroQKINAAAAACjVvuz8YiVDw61fKmbhR5Kkh13X6xtP74A2t57RXPFRtoD6GnH+y8Q/veaECh3rkXZQifbzzz8vk8mku+++W5KUmpqqO+64Q61bt1Z0dLQaNWqkO++8U+np6X7ttmzZogEDBigmJka1atXSAw88ILfbfTBDAQAAAAAcQkWrvQ09Zf1c11onSTLpQdcNGus5U1LB6eUOa1GaabUETzntVv/62glRh2DER065E+158+bpgw8+UMeOHX11O3bs0I4dO/Tyyy9r2bJl+vzzzzVx4kQNGzbMF+PxeDRgwAA5nU7NnDlTX3zxhT7//HM9+eSTB/eTAAAAAAC0ZFua3pq6Vk536Ku68t2eMvXp9Rq67ot5kgyNtH6uq62T5ZVJOu8tfefp5Yvbtj9X15/a1K9dKMFOHq8qypVoZ2VlaciQIfroo49UrVo1X3379u31ww8/aODAgWrevLnOPPNMPfvss/r55599M9Z//PGHVqxYodGjR6tz587q16+fnn76ab3zzjtyOkuu+QcAAAAAlMV5b/+rVyav0TmvT1eeK3hC/c5f68P2kZKe51desj1d21Kz9bT1M11lnSyvYdKUlk9KXYaW6Hedom0WXznabtGxqFyJ9m233aYBAwaoT58+pcamp6crISFBVmvBAeezZs1Shw4dVLt2bV9M3759lZGRoeXLlwftIz8/XxkZGX5/AAAAAAChbdibrZcmrQ767s2pa8O2/XfdXr+yy+3WKOvHGmqdIq9h0gOum+Q4oSDJ7nNcUW6XmedS8UnsHs2TQ36jCk9olz3RHjt2rBYuXKhRo0aVGrt37149/fTTuvHGG311KSkpfkm2JF85JSX46XajRo1SYmKi70/Dhg2DxgEAAAAAikxZuatc7dJyXUUFr0dNZzygy6zT5DFMutd1i37wnqbZG/ZJkt68vLMv9Li6Cb6ru9rUiVeXRtV0LCpTor1161bdddddGjNmjKKiwm9Wz8jI0IABA9S2bVuNGDHiYMaoRx55ROnp6b4/W7duPaj+AAAAAOBYUPKqrUglRh84Kdzjln68QTXW/yi3YdZdrtv1k7enJCnKWrAsPMZuVY9mBTPXybF2bdybLSn4tV7HCmtZghcsWKDdu3erS5cuvjqPx6Pp06fr7bffVn5+viwWizIzM3XOOecoPj5e48ePl81WdJx7nTp1NHfuXL9+d+3a5XsXjMPhkMPhCPoOAAAAAFBg7a7MCunHbjVLbqf0wzBp5QQ5DYvucN2pSd6ia7hiHUX7r+tXi5Yk/bFil2auL5jpnrMhNew3TCZT8aPMq5QyJdq9e/fW0qVL/equvfZatWnTRg899JAsFosyMjLUt29fORwOTZgwIWDmu0ePHnr22We1e/du1apVS5I0efJkJSQkqG3btgf54wAAAADAsefrOVs0YfF2peW4Sg+OQGp6pvTdvdKa3yWLXTfn3qk/vV38YhKK3Y9duIy8MMmWpNUVlPQfjcqUaMfHx6t9+/Z+dbGxsUpOTlb79u2VkZGhs88+Wzk5ORo9erTfwWU1a9aUxWLR2WefrbZt22ro0KF68cUXlZKSoscff1y33XYbs9YAAAAAUA6Pjl9aelCEHHKqyeQbJMtiyRolXTZGf36cHxB3XN0E3/O2/bkB7+McZUo3q5QK/ckXLlyoOXPmSJJatGjh927jxo1q0qSJLBaLfvnlF91yyy3q0aOHYmNjdfXVV2vkyJEVORQAAAAAgKT92U5Vi7VHFBulfH1ke0WnWpZJ1mjpirHaVaO7pKm+mKt6NFav1rXUoUGir65xckzAfvA7zvTPCUuqyqeOH3SiPW3aNN/zGWecISOCNfaNGzfWb7/9drCfBgAAAACU4qN/NujBc9qUGndj91rqvfAOnWReqSwjSnFX/iA1OUUnPfyrX9wT57aVzeJ/rva7Q7powJsz/Or6ta978IM/SpXrHm0AAAAAwNHh3WnrS42JU44e2vuYTjKvVIYRraucDyu/QfeAuCn3nh6QZEtS85pxAXUNDhyQFkrVPAatAIk2AAAAAFRibo9Xy3eky+utmNQ0NdvpV05QlkbbR8mybY7yrfG60vmoFhqtNGb2loC2phDrvaNsloA6szn84vBTWtSQVLDsvKo5dnenAwAAAEAl5nR79ej4pRq3YJskyWo2ad1z/Q+631vHLPA9JylTX9lHqYN5kxRdTc9XG6UlGwoOOft+wTZd17OpX9tIb+OKtQcm3iW9dkknjZmzRRd1bRDx2I8WzGgDAAAAQCX03fytviRbktwVNKM9+8D91slK1zf2Z9XBvEl7jQTpml8V16ToCq/gM+ihx9C+ftEp5GNv7FHqOJLjHLqzd0vVSwq/xPxoRKINAAAAAJXQ/hJLvCtSTaXpG/szOs68RbuNJA3TCKl2O91wWjNfjNvr1aa92X7tit+dXVJ0seXjnkinvqsoEm0AAAAAqIRK2+NcXrWVqrH2p9XKvF07jeqadvLn+u6JayT5J9Lr92Trn7V7/NrWSogK2e9pLWv6ntsWu2P7WMQebQAAAACohEIdPFYeXq9RkLinbdW39qfVxLxL24wausL5mKb37RWynctTNDPdolbgyeLFXdezqVJznBrcpYHs1mN7TpdEGwAAAAAqIUuQTNswDJnKkYHvz3Eq2bVT+nygmph3aYu3pi53Pq7hQ88J2y4r3+17Lu26rliHVcMHtivz2KqiY/ufGQAAAACgkjIHSaj/Xbcv4vZX92jse/buXSd91l9K36IN3jq6xPmktqumOjdKCtvHq5PX+J6fHtQ+4m8f60i0AQAAAKASCjZxPWXlroC6JdvSgrY/sWmyJKmVaauqfzdIytiu3Y7GutT5hFJU8K5mnCPi8TSsXvXuuz5USLQBAAAA4Cjx+cxNAXWrUjID6ro3q67+HeqovWmDvrU/LUvOHql2Bz1d8yXtUTVJUvVYe9Bl6JHcgY3wSLQBAAAA4CgWbC/3G5cdL9PWOfra/qyqmbKUXbOzdM3PSjMl+WKiQhxY9vzgjodopMcOEm0AAAAAOEzcHq/W7c6SUco904Zh6I8VgcvES9qZnut3YFmh2nvnSF9doARTruZ422h4wjNSdDX9s3avL2ZAx7pB+2xWM7bU7yI8Th0HAAAAgMPkzrH/6belKRp1YQddfmKjoDEp6Xka9M4M7crID9vXvqx89Rj1Z0D9qqtM0piLJU++pns66EbXvaq3yxMQN7BTvaD9tquXGHEsgmNGGwAAAAAOk9+WpkiSHvlxaciYd6etC5tkb03NkSStDrI3e+wpKYoad5XkyZdaD9D1rvuVJ4eS4+wBsR3qBybUoXRumBRxLEi0AQAAAKBS8XgDl5Vf37Op73nd7ixJktnsvzf7AvM/Omnh/ZLXJbUfLF3yhZyySZLmbdovt8fri21TJ75M93H3aJZcpp/hWEeiDQAAAACVSJA8Ww2qRfueR8/eLMn/nu0rLFP1iu19mQyvdPyV0oUfSRabXx/7c1y+52AnlYfTtl5CmeKPdSTaAAAAAHCIzVy/V4Pe+devbn+2M2js/E2pfuU2deLlsBVdueU9cJBaYZ49zPKbnrN9IrPJkHHCDdLAtyRz4BVdM9cXHYRWPHFHxSPRBgAAAIBD7IqP5mjx1jS/uuOfnhw0du2BpeGFHh/QVoM6Fx1G5jkw422Wodst4/WEbbQk6T33QJn6vySZQ1zb9fsq33P/DsFPHEfFINEGAAAAgEpsVUqGYuxFF0Z5vF7JMFRt1ijdb/tekvSK6yJ94ri6aJo7iJ3peUGfS/PC4A7lGPWxjUQbAAAAACoxa4lDz7wejzTxYTVb/ZEk6WnXEL3luVBzHusT0LZnixpB+3zugvYRf//CLg3KMFpI3KMNAAAAAIfMP2v36IO/N0Qcn5XvDqirlRDlezbLq0FbX5R2TpMkPea6TmM8BQm2xRw4m92tSTXNWLc3oD4+yhZQF0zfdrVlszA/W1Yk2gAAAABwiAz9ZG6Z4i/9YFZAXWH+bJVbL9ve1/mWmZLJrHvzb9SP3tPC9mcpwxVexY25/iSNnbdVIwa2LVf7Yx2JNgAAAAAcIfuy8pUc5/CVl+/ICIjp1qS65M7XO7Y31dcyXy7DItvFn+jHr+yl9l/yru1IndKihk4JsewcpWMNAAAAAAAcIVtSc0qNqWH3SN9crr6W+co3bLrJdY/U7oKI+k+MjmyJOCoWiTYAAAAAHCFeo+jZMIyA9/HKkUYPltZPVY7h0LWuB/Snt0tA3MBO9QLqJOnibhxkdiSwdBwAAAAAjpAcZ9HhZxl5/gehVVeGptZ+U9qyQnIk6u3kkZq5oaYkacs+/5nwU1sGX+btsFoC6hY9edbBDhulYEYbAAAAAMppT2Z+0JnocM5pV8f3vGFPtu+5+HbqOtqn7+wjVS19hRRTQ7rmZw0ceKHv/eP/W+bXZ3Js6fu1CyXFRB6L8iHRBgAAAIByeOaXFTrh2Sl66IclZWrXvVl13/PwCct9z6YDJ4Q3NqVonOMptTDvkBLqS9dNlOp2ks1SlIlPX7PHr89TW9Ysz4+AQ4REGwAAAADKaNPebH08Y6Mk6bv528rUtl39xKD1hmGotWmLvrePVAPTXuUnNC1Ismu0lCTVKHY6eUl2a2Sp3emtSMgPBxJtAAAAACijjXuz/crb9pd+enihro2qBa0371igb+1Pq5YpTbtjWshx4x9SUiPf+4pY8v3OkMCD1FDxSLQBAAAAoIxK3k/d84W/tHZXZrnaSpI2Tlf0NxcoyZSthd4WmtHzcymuVkCYI8jM9ZXdGwXUhRLn4Dzsw4FEGwAAAAAqwJ+rdpev4arfpNEXyezK0QxPO13pfFReR/BZ73y3N6BuaPcm5fsuDhkSbQAAAACoAGU7e7zAeeZ/pW+vlDz5muTppmGuB5SjqJB7rqsHOV3c5QlMvnFkkWgDAAAAQBkFWfytkrd8lbzruqQrLZP1uu1dyfDoB09P3eq6S/kqSKSLXwFW3K1nNA+oc3vLk+LjUCLRBgAAAIAKYJSY0z7tpb9Cxt5imaBnbJ/JbDJknHCD7nfdLI8svvehZrSHnNQ4oK5eUlQ5R4xDhUQbAAAAAMrIFGRK2+WOZGbZkCYP10O2sZKkt9zna3Lj+2REmJpF2y0BdbXiSbQrG46cAwAAAHDMy3d79OiPy9SvfR31aVu7XH3kujxh35vk1dPWz6R/p0qSnnNdrg89A6XRC/3iHujbOuJvfjXsxLIPFIccM9oAAAAAjnl3fbNIPyzcpuu/nB9R/NBP5gbU5YVJtK1y6zXbu7rSOlWSSVtPGVWQZAfRvVn1sN++tFtD3/OpLWtGNF4cXiTaAAAAAI5phmFo4vIUX3n5jvSw8et2ZwWtz3cXJdq7M/N8zw459Z7tdZ1vmSmXYZEGf6y044aE7L9zw+BXexW6rVeLsO9LCnZtNw4tEm0AAAAAx7SMXLdf+drP5oWNv+Ob/4LW57mKrtk68dmC5eGxytVnthd1lmWh8gyb7jTulzpcFHSPdyFLKZlxo+QY/XX/GVr85Nlh4wqZw30MhwR7tAEAAAAc00wlph93Z+bLMAyZQiSoK3dmBK2PsvkfVFZNGfrc/qI6mTfIaYnR03HDdcvFlxV88yBz36Y1YiOOLUi0uQLscGJGGwAAAECV4vEa8pbhbuml2wKXit/33eKgsTPX7w3Zzzdzt/ie62mvxtmfUifzBu0z4pVy/vd69p6b1bFBkiTJFPQm7kOjbb0ESZLdQvp3uPD/0gAAAACqDK/X0MC3Zuj8d/+VYUSWbA/5eE5A3Y//bQ8ae8VHgbEleXat0jjHCDU379R2I1mXOJ9UVONufjHmw5iJvTukiy7t1lAT7jjl8H30GMfScQAAAABVxu7MfK04sLQ7I8+txGjb4R3A9gUyj75I9UypWuetp6HOR7RTyYp1+KdeMbbDl4rVS4rWCxd1PGzfAzPaAAAAAKoQq6VoSbbT7Q0TWWDcgm0R9/3OX+vCvj/ZvEz64jyZclO1yNtMFzuf1E4lS1JAot0oOSbi7+LoQ6INAAAAoMrwFlsu7vSUnmjf/33wvdjBvDRpdch355jn6jPbi5IzS67Gp2mI8zHtV0LY/tY+2y/ib+PoQqINAAAAoMrwFsutgx1yVtEuO6GhLrP8qXdsb8hhciureX+lnT9G2Yr2xZzcPDloWxuHk1VZ/E8WAAAAQJXhKTajffPoBYf4a4aerzVFz9s+lsVk6Gv3mUo56z25zf77wkcPO+kQjwOVDYk2AAAAgCrD4/E/afzR8UtDxu7Lyo+435InmJvk1WPWMdLUpyRJb7sH6VH3MHlkkbvEGMzmyK/y+uv+MyKOReVFog0AAACgyvCUSIi/nrMlRKTU9ZkpEfdbvFuLPHrJ9qFusP4mSdrc7VG97L5UkklOt1fuYnd4P3Veu7D9PnFuW99zy1pxalojNuIxofIi0QYAAABQZXi8gXdn57k8B93vvE2pkiSHnHrf9rouskyXTBbp/PfU+NyH1LB6wZ7slIw8uYsdwnb1yU3C9ntl90a+5w71Ew96nKgcSLQBAAAAVBleIzDRXrc7q1x9FSbXknTph7MVrxx9YX9BZ1kWyGu2S5eOljpfIUk6rk7BCeMf/7NBrgNLx2vGO0r9hsNqUYNqBUn6uZ3qlmucqHxItAEAAABUGcFmtM99a0a5+nrip2WSCmbEayhdY+1Pq7t5pTKMaK3s84XUpr8v9o8VuyRJczamaktqtiRpT2Zke8B/veNU/XjryTqzTe1yjROVD4k2AAAAgCojWKJdXqtSMiVJ2bs26Hv7CLUzb9YeI0GXO59Qs259Q7a7efTCMn0nMcamLo2qHdRYUblYj/QAAAAAAKCiRJpovzFlbWQd7lqhpLEXKNm8S1u9NTXU9bC+feRKRdstBzFKVHXMaAMAAACoMkqeOh7Ka1PWBNQ90Le1X7mLaY30WT9ZslK0yttQg50jtMmoq8RoW0DbT67uVr4Bo0oi0QYAAABQZSzYtL/cbQd3aeB7PtO8UGPsz0l5aVrgbalLnU9otwqWd5tNgfdin9mmVrm/i6qHRBsAAABAlfHsbyvL3dZsku7u01IXW6bpQ9urijY5pZZn60rnI0pXnC/Obg1Mo0xBkm8cu0i0AQAAAFQJOU73QbWvFe/QXfaf9ZLtQ1lNXo3znCZd9rVyFRVR+77tODUcBUi0AQAAABz1dqTlqu2Tk0K+N0rZu22SV5r4sEx/jpQkves+T/e7btK6fXnq2aKGL+68TvVC9jFp+a4yjhpVFaeOAwAAADjqfTh9Q9j3eS6v76TwPJfH751dLr1ie0+aM1uSNNI1VJ96+kmSfl+aov+2FO37Hhgm0S5p7mO9I45F1cKMNgAAAIAqae6jRYlu4bLyfLdHHUf84auPU44+tb2ogZbZktkmDf7El2RL0vo9Wcp2FiXmfY6L/NCzWvGRLTlH1UOiDQAAAOCoF+wssloJUXIcOLgs50CyvHlfjpwerySpptI01v6MelqWy7DHSUO+lzpc5NdHvttb4jsceobSkWgDAAAAqLQ27c3W61PWKD3HFTbOpOAJcMyB5eK5B5aLr9iRIUlqbErROPsItTdvkmJrynTNL1LzXgHt29dP9J0y/viA4yIe9/mdI19ijqqHRBsAAABApXXGy9P0+pS1evCHxWHjQk00x9gLjqUqnNG+//vFam/aoHH2EWps3q3N3lrSdZOkescHbf/9/K3qUD9RktSgWkzYMXx6TTff831ntw4bi6qNw9AAAAAAVHqTlu/S1tQcNawePNkNtaB7e1quJOnrOZvVuWGSTrUs1VvmVxVnytMybxPN6fG+hiU3D/ndTftytGlfjiT5lqGHcmab2lr2VF/F2i0sMT/GMaMNAAAA4Khw6ot/hXxXMq8tWf5u/jZp6Th9aH5BcaY8zfC002XOx9WlXZuAvro1rhb0GzZL6elTnMNKkg1mtAEAAABUPaMu6OBXvtbyu/TDV7KZpF883XWv6xY5ZQs6Qx7qxm3yZ0SKRBsAAADAUWN3Rp5qJQRem/XHil1+ZbMvKzb0kHWsbrH+LEn6zN1XI91DZRxY3FsjzhHQl2EET7XJsxEplo4DAAAAOGo8//uqoPWbD+yjLtSpYZLkcell2we+JFu9n9RT7qt8SXYofdrWDlrfpEZsmceLYxOJNgAAAICjxi9Ld0YU17q6WRp7hS6yTJfbMOsB141yn3yPIpmXvuHUZrqnT6uA+npJ0WUdLo5RJNoAAAAAKqXsfHdAndPtDajblZHnV66uDOmLgdLaP5Rr2HWj61597zlDa3Zl+cX1bRd85tpmMeuibg386uoEWa4OhMIebQAAAACVUu9X/o4ornhC3si0S1/Ynpe275Kiq2lI2l1aaBTMTue7PX7tRg5qH7LPGJvFrxzrsISIBAIxow0AAADgsPN6Q53tXSSlxEx1MJl5Lp15ICHvZFqnH+3D1dS8S0pqJA2b7EuypcAD02qHmaWOtvsn1o1C3N8NBEOiDQAAAOCw+mr2ZnUa+YcWbU0rV/tt+4sOPvt33V5JUi/zf/rG/qxqmDJk1O0kDZsi1Wjp127a6j0RfyOqxIz2iPPalWusODaRaAMAAAA4rJ74aZky89y677tF5Wr/zC8ri5VMutTylz6yvaIYU77+9nSU6ZpfpfjA/dcrd2aUb8CSGidz4jgiR6INAAAA4LD4belONXn4V185gtXjQaXmOCVJhter1F9H6AXbR7KavPrefZqGue6XHPEVMVyg3Ei0AQAAABwWt45Z6FfeuDe7XP3M3ZgqeVza/uUwXZH7jSTpDfcFesB9kx4b2PGgx1no6fMLDkvr3qx6hfWJYwOnjgMAAACo9C7p1kDfzd8mSYpRnvTNZWqwaYo8hkmPu6/TN57ekqQhJzWOqL9xN/coNebKkxqpS6MktazFDDnKhhltAAAAAIfclBInfpdm3e5Mv/K1pzSVJNVQusban5bWTZHLHKUbXPf5kmxJslsjS3G6NSl9ltpkMqldvcSI+wQK8b8xAAAAAA65e0IcfLZ8R3rQ+tcmr/UrJ0bb1My0Qz/an1RH80YpJlm/dflIf3q7+GK+j2CWGjgcSLQBAAAAHHKZee6g9ROXpQStz8r3j4/dvUDj7CPUyLxHm7y1lX/1JO1JbO8Xc0IEs9TA4UCiDQAAAOCIcXq8Qev/XlN05/XZ5nlK+G6wqpuytMjbTIOdI5Qe07Dc3zylRXK52wKRINEGAAAAUG4uj1cLNqfKFSJhLs1/W9KC1O33PV9pmaz3bK/L5M7TqoSTdbnzce1TovKckX3PajYF1J3ZJvCObaAiHVSi/fzzz8tkMunuu+/21eXl5em2225TcnKy4uLiNHjwYO3a5X/wwZYtWzRgwADFxMSoVq1aeuCBB+R2B19KAgAAAKDyevJ/yzX4vVl6fPyycrWfuzE1oO6Cd2dKMvSgdayesX0mi8mQul6j8a1fVK6iJElOj0fuCC7i/um2UwLqZq3fV66xApEqd6I9b948ffDBB+rY0f+eunvuuUc///yzvv/+e/3999/asWOHLrzwQt97j8ejAQMGyOl0aubMmfriiy/0+eef68knnyz/TwEAAADgiPhm7hZJ0rfzt1ZIf2t2Zcomt161vadbrRMkSS+7LpbOfV0DOzfyxeW7vVoUZDa8pPb1EwPqqsXYKmSsQCjlSrSzsrI0ZMgQffTRR6pWrZqvPj09XZ988oleffVVnXnmmeratas+++wzzZw5U7Nnz5Yk/fHHH1qxYoVGjx6tzp07q1+/fnr66af1zjvvyOl0VsxPBQAAAKDSem9IF50Y4uCyf5au12e2F3ShZYbchlkPuG7U254LJJNJ7esnynRgJbjLYygjzxXR91rX9r8H+8mBbQ9q/EBpypVo33bbbRowYID69OnjV79gwQK5XC6/+jZt2qhRo0aaNWuWJGnWrFnq0KGDatcu2hfRt29fZWRkaPny5UG/l5+fr4yMDL8/AAAAAI6s/dnlmyirEe/Q3E2BS8aVtlXnL7xOPS3LlW04NMz1gL73nOEX0iQ5VpLkdHtVLyk6ou893L+N73nB430UH8WMNg4ta1kbjB07VgsXLtS8efMC3qWkpMhutyspKcmvvnbt2kpJSfHFFE+yC98Xvgtm1KhReuqpp8o6VAAAAACH0PFPTy5Xuw71E1UjzqG9WflFlTsWSV9fquScFKUY1TTM+YCWG00kSW9efrwvzG4pmCt0ur0at2Cbr/6fB3uF/F6U1eJ7To5zlGvMQFmUaUZ769atuuuuuzRmzBhFRUUdqjEFeOSRR5Senu77s3Vrxez/AAAAAHB4tKuX4Hu2mk16/dLORS/X/CF91l/KStEWaxNdkD/Sl2RL0nmd6vmeUzLyCprsyvTrv2H1mJDftlu5bAmHV5lmtBcsWKDdu3erS5cuvjqPx6Pp06fr7bff1qRJk+R0OpWWluY3q71r1y7VqVNHklSnTh3NnTvXr9/CU8kLY0pyOBxyOPiXJwAAAKAyy8hzKSHIsmzDMLR8R9H2T5PJpPrVCpZ9D7FMkfHN5zIZXqlZLw1YMUSZKkqaB3dp4NdXem7BvuyRv6yIeFxdGiXp3I511Tg5dDIOVKQy/dNO7969tXTpUi1atMj3p1u3bhoyZIjv2WazaerUqb42q1ev1pYtW9SjRw9JUo8ePbR06VLt3r3bFzN58mQlJCSobVsOJQAAAACOVnd8/V/Q+uJJtiSZTZLNbOhh69d61vZpQZLd+UppyPd+SfaHQ7vq5Ys7luwuQPv6CWHfm0wmvX1FFz3Qt03YOKCilGlGOz4+Xu3bt/eri42NVXJysq9+2LBhuvfee1W9enUlJCTojjvuUI8ePdS9e3dJ0tlnn622bdtq6NChevHFF5WSkqLHH39ct912G7PWAAAAwFHs7zV7gtbnOD1+ZZM7XzUm3qKbrb9IkibWGqZzBr0i35HiB5zdLviK15KM0q/TBg6rMh+GVprXXntNZrNZgwcPVn5+vvr27at3333X995iseiXX37RLbfcoh49eig2NlZXX321Ro4cWdFDAQAAAHCIpKTnRRT32uQ1emPqWl+5mjKkL89T1NY5choWPei6Sbaal+kck0mjZ28u11g6NUwqVzvgUDnoRHvatGl+5aioKL3zzjt65513QrZp3Lixfvvtt4P9NAAAAIAjYOPebPV6eVpEscWT7MamFH0X+4q0dbuMqERdlXGHZnvb6uID7x//aVm5xpOWU75rxoBDheP3AAAAAJTJbWMWlrlNF9Majbc/qdru7VJSI3mv/UOzvaHPaLrh1KYR9/3vun1lHg9wKJFoAwAAAIiIYRi65rO5WrEzo/TgYvqbZ+sb+7OqbspSerX20rApstQuOphs4vKUgDaPDYj8oORerWuWaTzAoUaiDQAAAEDLtqfrf4u2h43JcXo0bXXwA88KbU3NKSoYhm6w/KJ37W/KYXJpsqer5vcaLcXX9muTmecu97glaWiPxgfVHqhoJNoAAAAAdO5bM3TX2EV65691B9XPhMU7Ch48bq3+9EY9ZvtakvSZu69uct0j2Sr+LuukGHuF9wkcDBJtAAAAAD4vTVqtr8p4+veLg4vuuv592U4pP0sae4Vab/1OXsOkka6hesp9tbwyy+2t+Lu4mtWIrfA+gYNBog0AAADAzxMhTv8ucc21j7fYRda7t2+WPusnrZ2kPMOmW1x36VNPv6I+QnzTOIjLsE2hBgYcIRV+jzYAAACAqskUIk1OirFJklqbtuhT+0tSyj4ppoZudd6nP/P990+f2jL4wWXLd5TtgDWgMiPRBgAAABBg3e5MtagV71dnKHDW+bH+x6l9/USdYV6kt2xvKd6UKyW3kIaMU8vZefpz+gZJ0sZR/eXxGrJagi+qPfetGb7nzg2TKu4HAY4Alo4DAAAACNDn1ekBdcFWd1/ZvbEarB2jT2wvKd6Uq5mettL1U/TLNoc+OJBk92xRQyaTKWSSXdLn154Q8t3ke07zK/9256kR9QkcTiTaAAAAAMrFLK+ipz4q/Xa/LCZD37lP19Wuh6Xoarr96/98cTPW7S1Tv+FOEW9ZO16Xn9jQV25bL6HsAwcOMZaOAwAAAMe4NbsyA+riowJTheIT2rHK1Zu2t6U5BQn1C67L9J5noCSTPIfgZHHgaEKiDQAAABzjrv50bkBdZp47oK7wZPC62qdP7C+rrXmzZI2SLvhAP/wUJ2XmS5KWbEvza9e2buSzzsES/JLK0h9wJLB0HAAAADjGeSO8WsuQ1N60QT85nihIsmNrSdf8JrU7X9MeOMMXd8G7M/3a9T6uVtD+PrqqW0BdsAS/pPOPr6/LT2wUtD1QGZBoAwAAAFWQ12voltEL9O60dWHjnG6vXJ7IEu2t/36v7+xPq7YpTau8DaUbpkoNukqSom2WkO1u69UiaP1ZbWuHPfgslPgom0Zd2EFnta1d5rbA4UCiDQAAAFQxhmGoz6t/6/dlKXpx4uqQcV6voZOem6LUbGfIfg48SP++qeP+uVUxpnxN83TSRc7hUlIjX6zJFPyObUmKCpOE20ucRH7/2a1CxgJHCxJtAAAAoArxeA0NfHuGNuzN9tVt2ZcTNPa+7xdrf44rZF/puS7J45J+uVua/ITMMvSl+ywNc92vc7pWTEJcMkG//cyWFdIvcCSRaAMAAABVyPo9WVq2PcO/bm9W0Njx/20P29fsFRukMRdJCz6XIZOecg3Vk+5r5JFFTw5sWyHjNSLcHw4cTTh1HAAAAKhCzEFWcMeEWbodSgPTbjWf8IBk3i7ZYnWv5w6N93T0vU+Ish3MMH2K3wR22QkNQwcCRxFmtAEAAIAqxBxkr3SsI/L5tbmP9lYX0xr9ZH9SLc3bpfh60nUTNT6nY+mNy6H4iec9W9Y4JN8ADjcSbQAAAKAKsQSZ0n59ytqI29fa/Iu+sT+rGqYMLfU2kW74U6pb/iT75Ys7hX1fPdbue+7brk65vwNUJiTaAAAAwFEgx+nW/xZtLzigLIxgifaUlbtCHohWxNDdtp+kH4bJYXJpsqerLnE+KSXUDYh87dLgyXPj5JiAuguOrx/2q+3qJeiBvq319hXHy2YhPUHVwP8mAwAAAJXcut2Z6v/GP7pr7CLdMnpB2NhQZ4ud/+6/Ids45NRrtnd1t+U7SdLPsYN1k+se5SpKUsE1YMVdcHyDoP30OS7wXutgiX9xJpNJt/VqoXM71gsbBxxNOAwNAAAAqOT6vDrd9zxz/b6wsd4QmXaou7JrKk0f2l/V8eZ18pqsMg94SRZHP3nHLJQk7UjL1bxNqRGNs5ScGjhmkGgDAAAAlVhZr7/yRhi+cMt+HWfarI/tL6u+aZ/SjFitP/1dde12vroXS8o//mejkmIiO2G8aY04v/LZbQNnuIFjAUvHAQAAgErM5Slboh1pYr5rzg8aZx+h+qZ9Wu+tq/OdI7W3VndJUoy96DqwT//dqEVb0yLqs0/bWn7lvVn5kQ0aqGJItAEAAIBKzO31RhT37bwtavLwr/q3lKXlMgxpxms6Z/n9ijXla7qngy5wjtQmo67sBw4jc1j904Q/V+2OaAyxdv8Fsx9d1S2idkBVQ6INAAAAVGKRzmg/9MNSSdITPy0LHeTOl366RZoyQiYZ+sJ9lq51PagMxUqSTmtVU1LBAWWFWtX2Xw5+/9mtQnZf8r7u5DhHRGMHqhoSbQAAAKASc3kim9EuTbLSpS8GSou/kUwWPe66VsPd18qjgmXiN57WzO+E8PM7F5wC3rVxdb9+hnZvEvY7N5zaVHUTo7Tg8T4VMm7gaESiDQAAAFRi701bH1CXkp5Xpj7amLbof44npK1zpKhE6cofNNpzll/MI/3a+JVnrNsrSfpm7hZf0t2werQSSzkY7bEBbTXz4TOZzcYxjUQbAAAAqKTScpz6ZMbGgPqt+3Mi7uOa5JUaZx+hBqa9ciU1k66fqvzGpwXEFV8uLkl7s4pOHq8ZX5A092tfN6JvluwLONaQaAMAAACV1P4c10G0NnSj5WcNz35GcaY8/etpp1P3PSbVaKmtqbmltj61ZQ3f80f/FCT7y3ekH8R4gGMHiTYAAABQSVlCzAx7S7ks2y6XXrJ+oEdt38gkQ2PcvXW16yGluKIlRTYjPrR744C6f9eVcqI5AEkk2gAAAEClNW7B1qD1wycsD9mmujL0lX2ULrZOl0dm5fYepcfc18mtohPB1+7K9GszP8jBZbkuTzlHDYBEGwAAAKiEPv5ng978c13Qd6tSMoPWtzJt1f/sT+gk8yplGNF6wP64PCfeKMl/Zjw5tuigsn8e7KUaQQ4uywuSaJ/dtnYZfgLg2EWiDQAAABwBhhF++fczv64sU39nmP/TD/YRamjeo83eWrrAOVIdT79QVnPg8vNnfyvqu2H1mKD9ndg0OaCuc6OkMo0JOFaRaAMAAACHUZ7LozemrFXbJydp9OzNB9+hYUgz39IntpcVb8rVbO9xGuR8Wk8Pu1BXn9wkaKKdmu0M0pG/pjViA+quO6XpwY8XOAZYSw8BAAAAUFGe+XWFRs/eIkl6/KdlujLIoWMRc+VJv9wtLf5GFpP0jbuXnnRfq7XPD/KFWCpoau3L605UlM1SMZ0BVRyJNgAAAHAYFSbZ4fyzdk9AXbzDqsx8t6/sTd8h83dDpe3zZZgsGuG8Ul94zlbJ/djh7rSOc0SeDpzWqmbEscCxjqXjAAAAwBG0Mz3wTuuhn8wNqDu5RbKOP7BHuqNpvTLfOlXaPl+KStJvnd7WF56+KplkB1P8arCsYok7gIpDog0AAAAcQQs2748ormODJHVvlqxB5hn63j5Sie69Us020g1/6rbZib44S5A92Sc0qeZ73p9TtD/73rNaHcTIAYRCog0AAAAcQS9MXBVRnMnwqPfWd/SG/V05TC5N9nSRhk2Wt1qzUts+3K+N73n5jgzf8229WoRt9+A5rSVJoy7sENEYARRgjzYAAABwmBRftl2oY4OkUtvFK0cDlt2jxqn/SpLedg/SK+6LtTEqQT/M3+oX++HQrgHt6yZG+57fm7be9xxs9ru4W89ooctPaKRqsfZSxwigCIk2AAAAcJi8OnlNQN1JTav7ld0er1+5qWmnPrK9osapO+Q2O3Rv3g2a4D1Z9ZMKkudHxy/1i+99XO2Ab3iL3dk9a8O+Mo2ZJBsoOxJtAAAA4DB5+691AXUuj/8s9/8W7fA9n2peordtbyrRlCMjvq7+7fqWJkws2GOdkesK2j4Yo/QQABWIPdoAAADAQfhw+no1efhX/bJkR+nBQeS5PH7ltFyXJEPDLL/pc9sLSjTlaHdiJ5lu/FunnNrbF5eZ75YRYQadHMesNHA4kWgDAAAAB+G53woOM7v96//CxgXbny1JL01a7Vd25ufqJesHesI2WhaToe/cp+v3bh9J8bVltZhlsxTtq85x+ifpt4c43CzGzkJW4HAi0QYAAAAqyF+rd4d85/J6Q76buW5vwUNmigb9d4Mutk6XxzDpKddQPei+US7ZfLF2S9Gv8DPX79MVJzWSJLWrl6D7zo78uq6vrz8p4lgAZUOiDQAAAFSQaz+bpxXFrs8qLtxe6s9mbpK2L5Q+7KV6WcuUbsToGtdD+szTT5JJJlPRLHaj5Fjf89rdmcrJd0uSzu9c3y+upBollo+f3KJGBD8RgPIg0QYAAADKKdhy8HV7soLGbtqbHbKf0/L+kj7rJ2Xu0L7oJhrkfFr/eDv63l/UtYHvuVmNokTbYbVo2po9kqTE6KJZ76Bj5UA04LAh0QYAAADKqdPIPwLqLCFmlc99a0ZAnVlePWT9RkN3PCO586SWffVlu0+0yajri5n/eB+/JNphLfoV3mo2KS2n4PTxLak5Ycd6QpNqvufPrz0hbCyAg8OpCAAAAEA5Zea5A+osEU5lJSpLb9re1umWJQUVp9wt9X5SHVfvlWbs8sV5S5ws/sA5rfXjf9slScMnLPfVZ+UHjqW45y/sqNa1N2pw1wZqXGz5OYCKR6INAAAAVCBzkBltt8f/ILTWpi36sdo7is3ZqlzDrodcN+rNs56SVLAcvLiacQ6/ct3E6KDfvfzERmHHVS3WrnvPbl3q+AEcPJaOAwAAAJKcbq9uG7NQo2dvPqh+rBb/RNvl8arFY7/7yv3Mc/R73EjF5mzVVm9NDXaO0ATvyb73xWewHx9wXNgDzoprWD14Ag7g8CPRBgAAACT9vmynfl26U4//tCyi+CXb0oLWT1+z16+8eGtBnFle3W/9Vu/Z35DZlSM1PU3zzv5BK4wmfvHFF4pfekLDCEfPXdlAZUKiDQAAAJTQ5OFflZrtDBszdWXwO7PHH9g/XchkkhKUrU9sL+l26/8KKnvcLl05Xg3qFyXS/23ZL0m66av5kqR4h1XxUcFPEq+f5D97XTcxKuxYARxeJNoAAACApD2Z+X7laz+bGza+5CFlhdJzXX7lqLR1+sn+hHpZFivPsOmHJsOlvs9KFqtqJxTtv96RlieP11Ceq2A/d2aYw81evriTX7l6rD1EJIAjgUQbAAAAkPTMryv9you3pYeM3ZeVL2eJA86CWvmz2vx8vpqZU7TNqKHBzhFq1WeY73Wj6jG+5zqJUfp6TmT7w+Mc/svEa8U7QkQCOBLYyAEAAACUwdJt6Rr4duCd2IWGdm8seb3StFHS9BdlkTTL01a3ue5UqhLUoUGiL9ZkMqllrTit3Z2lfLdHvy7dWa4xPdC3TbnaATg0mNEGAAAAyuC+7xcFrb+zd0tJ0k+zV0pjL5emvyhJ+t56roa6HlaqEoK225WRJ0matnqPNu7N9tX/emfPkGMofhD5wE711LZe8L4BHBkk2gAAAEAIeS5PQJ3bE3xvdlaeW81N2/WT/QlpzUTJ4pDOf1/DnUPlDrOQNCOvYC/2h9M3yGYp+vW8Ra24iMb47AXtI4oDcPiQaAMAAKBK83qDJ8aRCJZoX9ilftDYLbPG6Sf7k2pu3ilXbF3puolS58uV4yzq4+sbTgr7veJ7rR1WS0RjNEd4zzaAw4dEGwAAAFXS7sw8fTFzkzo99YfmbUotVx8Tl6UE1EXZ/BNgk7x6rtrP+tj+iuJNuZrjbaPtl0yU6ndRvts/UT+5eY2w31u4JS2icbmKHcRms5BoA5UNiTYAAACqnJT0PJ347FQNn7Bcmflu3f71wnL18/CPSwPqis9yxylHH9pe0xW530iSPnefrSHOR5XiiZckjZiwolzfLY2r2PJ1u4Vf6YHKhv9fCQAAgCpn9oZ9fuVdGfkhIstua2quJB3Yj/2kzrIskGGx6wHXjRrhvkZuWXXZh7MlSd/M3eJrF2rJeTClxRaf0TaxdByodEi0AQAAcMwzjMj2cXu9hr6dv1XnmOfqf/Yn1MK8Q/nRtWW6dqJW1jkvbNuLuzaMeDzndw6faLeqHR9xXwAOP+7RBgAAQJVT1knelTszI4qbsny7HrJ+o1usP0uSZnuPU6cbxkvV6yohanbYtjXi7BGPp3OjpLDva8Y79PcDZyjOwa/zQGXEjDYAAACOaWt3Zar/m/+UHpi9Vx2mXedLsj9y99cQ56OyJdaWJF3crUHY5i1DzEIP6Fg3oC4+ggS6cXKskuMcpcYBOPxItAEAAHBM2JWRF7T+ovdn+ZVHDmrnV3a6vdL2BdIHp6vuvjnKNhy63XmHnnVfKY8sspgLps9LW+4dyguDOwbUse8aOLqRaAMAAKDKCXa39IgJy4PGpue6/Mr92vvPMOfP+1z69BwpY5s2eOvofOfT+sXbQ5LUtXE1X1JcMjmOdN83y7+BqodEGwAAAFVOsET79yB3YgeTEG3VZ9ecILtces76keIn3SN5nFLr/hrkfEZrjYIl4pd0a6Dvburh17ZTwyTf86qUon3fVnPkM9T3ntUq4lgAlROJNgAAAKoct9dbepCkLftyAuocVot61XXqO/tTusL6lwyZpDMfly4do0zF+OIeOqeNb9l4oWhb0a/XL0xc5Xu+u0/LsOMo3k29pOiIxg6g8iLRBgAAQJXjdEeWaPd+dVpg5Ya/pQ9OU2fzBu034nS180HptAcks/+vzsEOIou1Fy0Dn7Z6j+/5ptObhx3HlHtP9z2XYfIbQCXFhhAAAABUOS5PZPuj/eMM3Wj5RfrqW8nwapm3iW523a1tRi1JBaeTlybabglab7OEn99qnBxbNIrIhg6gEiPRBgAAwFEhx+nWnI2pOrl5shzW4AltoZIHnAWzbndR4hyrXL1o+0ADLHMlQ1KnKzR4ztnKV9Hd1/3eKLoC7I3LOgftM1hCHRsi+S6u+BJ08mzg6EeiDQAAgEovO9+tdsMnSZLO61RPb15+fMjY1Gyn3/7oUJ75daUkqblpu963va6W5u1yySLbgBelbsOUP+c3v3i3tygFHtixXtA+g93K9eiA40odC4CqhT3aAAAAqNRWp2T6kmxJmrB4R8jYfLdHA9+aEVG/ZpNJfc3z9JP9SbU0b1eKUU03WkZKJ1wfPGMu3jbERupoW+DsdZ4rsv3ihbo2rlameACVD4k2AAAAKrWP/tkQUJeRF3xp+IsTV2t7Wq5f3Q+3FF3BlefyFDx4PRq8/2N9YH9N8aZczfG20bn5z2m19eBmn+/uE3g1l80S2elm8x/vo4l3n6qmNWJLDwZQqZFoAwAAoNIyDEPjFmwLqO844o+g8cFi29VL9D1n5LmkrD3SV+drQPpYSdLH7n4a4nxUe5Wod4Z08cVGmiAXVzM+8CTyQZ3rR9S2RpxDbeoklPmbACof9mgDAACg0pq8YleZ4q0llnSbTVJUseXci/6dqLOXPyxl7lSuovSg83r97D1ZkrRxVH+Zii0Zd1gtcnnckqQ1xU4c/+TqbhGPJznWrsRoW5l+BgBHP2a0AQAAUGlt2JtdpniH1f/X2zHXdz/wZGiY5Tf1mnWdlLlTqtFKA/NH+pJsSX5Jdsm+zn5terH68KeIn9G6pu955iNnlmn8AKoGEm0AAABUWqHulG5TJz6gLi3HqR3peX51PZonS3kZesf2hp6wjZbN5JHaXSjd8KfWGQ3Cfrtk0l5o6fb0sO3euPR4PX1+e/33xFmlJuUAqiYSbQAAABwRmXkuZee7y9V2VUqm5mzY51fXeeTkwMBdK6SPemmAZa6chkVPuq6WLvpUO/NK30HpCHKCuCRF28L/Cp0YY9PQ7o1VLdYeNg5A1UWiDQAAgMNu2fZ0dRjxh9oNnySPN8S0tSRDod8N/XRu2G+cb54hfdxb2rdOO4zqutT5pL709JVMJu1I85/5/ufBXgHtQ81oX9StYdjvAgCHoQEAAOCwO7fYXddZ+e6QB4Z5wyThTnfw+6ntculJ65e60jpVcklq1kvnrrhUqUrwtct3e/zaNKweE9CPJcRd2XEOfoUGEB4z2gAAADis9mbl+5VnrN0bMvblP9aUqe8Gpj363v5UQZItk3T6Q9KVP+iqPl19MdPX7FF2vid0JwBwkEi0AQAAcFj9uXK3X3n07M0Rt1064uyg9V6voV6W//SL/VF1Mm+QEV1NGjJO6vWoZLboxtOa+WLTc11+bVvXDjxYTZJ6ta4VUHdqyxoRjxXAsatMifZ7772njh07KiEhQQkJCerRo4d+//133/uUlBQNHTpUderUUWxsrLp06aIffvjBr4/U1FQNGTJECQkJSkpK0rBhw5SVlVUxPw0AAAAqvU//3ehXjnVEfjJ3fFSQJeZej6a+d5c+s72kJFO2FnmbyXTTdKllH19IVLHTv9ftyfL75lfDTgz6rTt6t9Al3fxPJr/l9OYRjxXAsatMiXaDBg30/PPPa8GCBZo/f77OPPNMDRo0SMuXL5ckXXXVVVq9erUmTJigpUuX6sILL9Qll1yi//77z9fHkCFDtHz5ck2ePFm//PKLpk+frhtvvLFifyoAAABUWpl5/ieN787MDxEZgey90ujBOmvPF5KkL91n6RLncCmpkV+Yudh+6/emrVf+gf3d7esnqFZCVNCuHVaLzu1Yz6+uU8Ok8o8VwDGjTIn2wIED1b9/f7Vs2VKtWrXSs88+q7i4OM2ePVuSNHPmTN1xxx068cQT1axZMz3++ONKSkrSggULJEkrV67UxIkT9fHHH+ukk05Sz5499dZbb2ns2LHasWNHyO/m5+crIyPD7w8AAAAqj237czR9zZ6IYvfnOP3KS7aFv5c6lC6mNdIHp0kb/lKO4dCdztv0pPtaORX8YLXi/llTsC982fbwv1eaip2HdvmJDRXLQWgAIlDuPdoej0djx45Vdna2evToIUk6+eST9e233yo1NVVer1djx45VXl6ezjjjDEnSrFmzlJSUpG7duvn66dOnj8xms+bMmRPyW6NGjVJiYqLvT8OGXKkAAABQmfR84S9d9elczS5xt3UwbesmHOTXDF1jmahv7U9LGdul5JYa5HxaE7ynSJLsIa7lKq7k8vVQ2tdL9D0/d0GH8g0XwDGnzIn20qVLFRcXJ4fDoZtvvlnjx49X27ZtJUnfffedXC6XkpOT5XA4dNNNN2n8+PFq0aKFpII93LVq+R8qYbVaVb16daWkpIT85iOPPKL09HTfn61bt5Z12AAAADgMlmxLC/s+Lcep+Zv3l/8DeRl62/aWRti+lM3k0aKEXtKNf2mtUbSXetGTZ5W//xKqxdo197HeWvZUX5lMwa/7AoCSypxot27dWosWLdKcOXN0yy236Oqrr9aKFSskSU888YTS0tI0ZcoUzZ8/X/fee68uueQSLV269KAG6XA4fAewFf4BAABA5VPa0uqRv6yIuK+Sd2i3NW2SPjxD51pmy2VYNMJ1lZ603Sc54lU3sWifdYw9+BgeH3BcQN3Q7o1LHUet+CjuzgZQJmX+L4bdbvfNUHft2lXz5s3TG2+8oQcffFBvv/22li1bpnbt2kmSOnXqpH/++UfvvPOO3n//fdWpU0e7d/tf5+B2u5Wamqo6depUwI8DAACAw213Rp7v2WYJP4+zPMSe6JU7M3RciSXlM9YV3q9t6ArLnxpu/VJKdUmJDXXJ7uv1n9FSA6rHSpLMB2abgyXThY5vlBRQN3JQu7DjBYDyOOh7tL1er/Lz85WTk1PQodm/S4vFIq+34FTHHj16KC0tzXc4miT9+eef8nq9Oumkkw52KAAAADgCznv734hjs53uoPW7iiXrhXam5ypWuXrD9o6es30ih8kltTpHumm6/jNaSpJ+XbpTkrQ9LVeS9NOi7SG/bbcEXiPGcnAAh0KZZrQfeeQR9evXT40aNVJmZqa+/vprTZs2TZMmTVKbNm3UokUL3XTTTXr55ZeVnJysn376yXeNlyQdd9xxOuecc3TDDTfo/fffl8vl0u23367LLrtM9erVK+XrAAAAqIxSiiXJD45boi6NqqlFrbigsblOT9D6HWmBibY3Zbkm2B9Xc/NOuQ2zXnRfqkcve1cqMbHj9nh9z5ef2KhkNz6RHJIGABWhTP+12b17t6666iq1bt1avXv31rx58zRp0iSdddZZstls+u2331SzZk0NHDhQHTt21JdffqkvvvhC/fv39/UxZswYtWnTRr1791b//v3Vs2dPffjhhxX+gwEAAODIuOOb/0K+y8ovmtH++oaiFY3WYvdcyzCkhV/qgvlD1dy8UzuN6rrU+YRufOSNgCRbkuZuSvU91wlxJ7YUmGg3SY4J+3MAQHmVaUb7k08+Cfu+ZcuW+uGHH8LGVK9eXV9//XVZPgsAAICjyJ7MwNnpQvnuotnnLo2qqX5StLan5cpmPZBoO7OlX+6VloxVlEma5umke1y3aPYzl8hhDVz6LUkmFSXpaTmukN8umWjffHrzSH4cACgzjk8EAABAhdqb5ZRhGKXuf46yWVQnMUrb03KVne+Rdq+Uvrta2rtaXpNFLzkv1vuec2XIHDLJlqTMvKLk+pQWNULG2Usc1Hb+8fUj/IkAoGzYqAIAAIByMQwj6CFmkrRoa1pEfSw4cKf2fxPelT7sJe1drf2WZF2a95je85wnI8Svq+8N6eJ7vvGrooN26yRGtnS8W+NqirKFTt4B4GAwow0AAIByefmP1Xrnr/VB3zmLLREvtD/bGVAXpXyNtH6uS6x/S25Jzc9Un+UXa58SfTE3ndYsoF3D6mXfX+0olmg/d2GHMrcHgEgxow0AAIByCZVkS5LZHLhsfNmOdN/zr3f2lPas1v/sT+gS69/yGCZ9bLtCxpBxfkm2JJ3XOfB2mlxX4OnlliDfLK740nG3xwgbCwAHg0QbAAAAFS5Yyrt0e0GiPaBjXbXbM1H6sJdam7dpt5GkK12P6pnMc+UxAlu2q5cYUJcUbQuou/Kk0Fd7SQXJf7yjYEFn0xqxEfwUAFA+LB0HAABAhXN6ApeOL9+RIYecui3zTWn8eEnSDE873e26XXsPzGJ7jMhmmlvWjg+om7Jyt54aFL7dvMf7yO01FG1nfzaAQ4cZbQAAAATYk5mv+79frIVb9per/bfztvqed6Tl6qmfl2vF0oX6yf6k2u4cL8kknfGIrnI94kuym9aIlccb+ZLus9rW9isf3yip1DZRNoviHMw1ATi0SLQBAAAQYNTvKzVuwTZd+O7MiNuMHNTO9zxvY6rv+eTn/1TarK/0i/1RHWfeIk9MDWnoeOmMh+Ut9uuow2oOSLSn3nd6yO85StyL/ewFHHAGoHIg0QYAAECAFTsyytymeJK8I/3AtV/5WXrZ9r5es7+nWFO+ZnnaynzzDKl5L0nSLWc097VZlZIZkGg3rxkX8nuuEsvTE4Ps2waAI4F1MwAAAPDj8RpalZIZNsYbZIl3wLLvlKXS99fqIstaeQyTXncP1jue87Uhoa4vJKbEXdbuMiwdn7R8V8SxAHA4kWgDAADAJz3XpZ4v/Flq3OpdgYn432v2HHgydKVlivTR15InXzuN6rrLeZvmGsf5LS+XpHb1E/zKwRJ4ADjakGgDAADAZ8LiHcrMc5caN3r25oC6xVvTlKBsPW/7SP0tcyWPpFbnqP+S87VfBQn1oE71/dr0al1LXRtX04LNBYeulWVGu7hvb+xernYAcCiwRxsAAAA+k5allBqTle/WmDlb/OpevKijrmq4R785HlF/y1w5DYvU9zkZl33jS7IlKTHGfx+1yWTSyc2TfeWd6bnlGveJTauXqx0AHAok2gAAAPCZsW5vqTHb9uf4lU3y6uK8H3Tf9rvUwLRXm721NNj5lNTjNo2eW3TN12mtagbtr1/7oj3bq1OyIh7rJd0aFI3BZIq4HQAcaiTaAAAAKJMHvl/ie05Wuj63vSjTlOEyed362dNd5zqf01KjmXak5er9aet9sQ/2bR20P4et6FfSR8cvjXgc3Zsllx4EAEcAiTYAAEAVtWZXpl6atErpua6I4tNynBHFLd2eLknqYV6u3x2P6HTLEskaLQ18U3e47lCmYiRJ/21Jk81SNNNsswT/1dMwgu/LfuLcthGNBwAqGw5DAwAAqKLOfm26JGlXRr5evrhTqfHb9ke2P9oij+60/qg7LD/JbDK0xltfrW4YJ9VuK33/qy/O7fXKWiy5Lp50Fxfq/LNhPZuGHQerxQFUViTaAAAAVdySbWkRxYVKXLPz3Yp1FPza6Enbpq/tz+ok8ypJ0lj3GRrhvlqrahfMPjesHq2tqQUJ+11jF8liLn1G2x6ivjS9WteSJHVumFSu9gBwqJBoAwAAVHFrdmVp2fZ0ta+fGDYuxApuLd6WppOb15BWT5Tx4806ybxfmUa0HnMN0wTvyXpxcEdfrNfr39YTwXVdTWrElhoTTFKMXStHniOHld2QACoX/qsEAABwDDj3rRmasTb0ieLfzd+qc9+aEfTdw2PnSb8/LH1zqaz5+7XE21TnOp/VBO/J6tQgUZec0LDCx9u+fkLpQZKi7RaZzawhB1C5MKMNAABwjHhj6hr1bFkj6LsHxy0JWt/ctF1v5b8tzdksSdrf8XpdNPdUOVVwH/b/bu/pFx/qYDNJalAtOuS7aJtFuS6Pr3zpCY1CxgJAZceMNgAAwDEiK99TetAB427qrsstU/WL/TG1NW+WYmpIV3yvr6vd4kuyy8JmMYW967p4ki1J63dHfp82AFQ2zGgDAAAcI2rGO4LWL9qa5ldOVJa6zr1L3Ww/S5KmezrotFvGSfF19NKnvwbpoUio+ezzOtUv01h7talVpngAqEyY0QYAADhGnBZi2fj57/zrez7RtFJTYx6VaeXPchoWPeu6Qle7HpIRVzug3cnNkwPqbj2jedBv1EkMnuSHckqQvgHgaEGiDQAAcIx45teVIfdQW+TRPdbv9Y39GdXw7pWqN9e1luf0kedcGTL7ruwqbsz1JwXUDTmpcdD+6ySG3p8tSfee1cqvbC3nlV8AUBmwdBwAAOAosC8rX12fmSJJ2vT8gHL3s3lfTsB1Wg1Me/S67R11M6+RJH3vPk0X3/SNbt6Sq38/mStJslhM2pnun2wH23Md6gTwhmEOQpOk23q10OmtaspkKv91XwBQWfBPhQAAAEeBp35e4XvOdUZ+qFlJ38zb4lfeN+cb/WZ/RN3Ma5RhROsO5+1a3O05yRGnE5tW98V5vYbSc13l/u6pLWuGfW8xm9SpYZI6NkhSQlTZD1sDgMqERBsAAOAo4PEWLfn2hrlCq9DW1Jyg9R/8vaHgIT9L+ulWJf9+sxJMOVrgban+zlH62XuynjqvvSTJYbUozmH1fX/J1nRfP9ed0jTisf98e09ZuOsawDGEpeMAAABHgeJLsnek5apl7fiQsZ/M2Kinf1kR8r12LJLGXSelrpdHZr3tHqQ33RfKI4sk+SXFVkvBs9tr6MEfiu7a7tkyssPKujaupg4NEiOKBYCqghltAACAI+DVyWs0du6W0gMPKD4ffNZr0/X876tCxgZLss0mySSv3mo8Q/q4j5S6Xkqor8vzH9Nr7ot9SXadhCi/dlZzYaLt9atPjLZHNO47zmwRURwAVCUk2gAAAIfZxGU79ebUtXr4x6URt5mweIdf+f2/1yszL/I904+eVl1f2F7QwF3vSl6XdNxArTr/d801jvOLu6O3f2JsNRf8upie4/+tLo2SQn7LYS36FTPWwQJKAMceEm0AAIDD7ObRCyuknz2Z+RHFnWH+T0MWXq7TLEvlNDmkc1+XLvlKS1P9fxUcMbCtrjixkV9d4TLyd6et96sPduJ4oekP9lKzGrE6vVVNdWlULaIxAkBVwj8xAgAAHEHvTlunW05vHjZxDSXYoWguT9ESb4ecesT6ta6x/iG5pJXeRrrDdbumdLtWkmQrcVf1NUEOONueVnCl199r9kQ8rtoJUfrz/jMijgeAqoZEGwAA4Ah6ceJqOawWDesZ+hRvrzf4KeNbUnPUopb/oWgjJiyXJLU1bdLrtnfUyrxdkjQj+SIN236u8lW0t7ocuT0AIAIsHQcAADjCnv5lRdhl4FNX7Q5af93n8wPqvp6zSTdYftFP9ifUyrxdu40kXeV8SK6zRylfdiXFFN1RXZ5ZdEm6pFuDcrUDgGMFiTYAAEAlcMKzU7R8R3rQd7sz8yLrJH27RttG6THb17KbPPrD01Xn5D+v6d5Osh041Kz4qeLb9hfdtd2sRmzEY72/b+uIYwHgWESiDQAAUEl8/u+moPVBtmJLks5qW7uosHy89N7JOsWyXDmGQw+7rteNrntVq0593d2npQqvxl6Vkulr8sXMou99e1OPoN94elC7gLpYO7sPASAc/isJAABwGO3PdoZ8Z7UELuW+/ov5mrJyV9D47Hy3lJch/f6QtPhrSdJibzPd7bpNG426kqSJd58mSZq1fp+v3brdmWpRK167MoqWq9eMdwT9RtMacQF1xa/vAgAE4r+SAAAAh9H6PVkh3+U6PX5lwzCCJtldGxdcmZW3YZb0fs+CJNtklk69X4OdI3xJ9iP92vjaFN+OvXR7upxuryJhDrKN22rhV0gACIcZbQAAgMNk+Y50XfHxnJDvSx4uHuKwcS3avFf3WMfrdst4Kc2QEhtJF36g6fkt5dZcX9xNpzf3PRe/9mvEhBXqUD+xfD8EAKBU/HMkAABABXB7vOo4YpIe/2lpyJjLP5wddia5YwP/5Hfhlv0BMT9dVkfTqo/SXdYfZTEZ2lDvXOmWGVLjk3XVp3MD4gsVny1Pz3XJXSyLP79zvZDtahU7PA0AEBkSbQAAgArw5p/rlJHn1ujZW0Lee52R5w7bh6dEu4vfn1WsZOhiyzR1/nWgGuasUIYRozuct+u/ri9IUaXPTue6/Jelu9xF37KH2XPdolbgHm0AQHgk2gAAABXgj+UpvudsZ/iEOpTiy7sladCBmeYkZeo92+t6yfah5MqWGvfU3dXe0c/ekxVls0iSxi3YFrbvnBL7v/PcReUODZLKNV4AQHAk2gAAAAcpLcfpd23W78tSwkSH5vT4z2h3aVRNp5iXaqLjYfWzzJPTsEh9RkhXT1BebEESPnP9XknS/d8v9mv7+12n+pW7N0v2Kz8+fpnv+fITGpZrvACA4Ei0AQAADsKcDfvUeeRkv7oHxy2REery6zD8ZrRduTph9UsaYx+lOqb9Wu+tq9TLf5N63iOZLZp54LquMXO2BO3ruLoJfuWmNWL9yqt3Ff3DQGmniN90erOy/BgAcMwj0QYAADgIb/25Lmj99/PDL+UOZv3uA1d/7VgkfXC62m4eLUn6yt1HA5zPKbnliRH182j/NkHrP7vmhDKPSZIePqeov4f7Be8bAFCERBsAAOAg7EjLDVr/4A9LImr/1uXH+56nrtgh/f2S9HFvae9q7TaSdI3zAT3hvk55cshiCnKpdRDXndI0aP3prWpG1L4kk8mkz649Qbee0Vw3nsrsNgCUhnu0AQAADsKGvdnlbjv/8T6qEefQHd/8pyamnXrV9p70V8EM+d5G/dR3zSDtV9EScLM5eKJd8pRzS4i4UO0j0at1LfVqXavc7QHgWMKMNgAAwEEIdzVWccGu/EqOtUuGoSstk/Wb/VF1Ma+THInSBR+q25or/ZLsd4d08Wu7/Km+vueMPJffO1OEM9+SdGfvlhHHAgAiQ6INAABwEKwRzhL/vGRHQJ0pM0Uac5GesX2mGFO+/vW0U9o1f8voeImkon6/v7mH+neo69c21mFVnKNgceL+HP9EuywGdqxbehAAoExYOg4AAHAQIj1cfM7GVL/yAPNs6b3bpNz9cpsdejb/En3u6atfjWTFlLjmq1vjakH7TIqxKSvfrb1Z+b6623u1KNP4Yx38OggAFY3/sgIAAByEXJcnorivD1zDlaAsjbR9rvMtM6VcSXU7a8Hxo/TZj/slScaB/1NcqKXg1WLs2rY/V+/+VXTy+akta5Rp/DXjHWWKBwCUjqXjAAAAh9j2AyeTn2JeqkmOh3W+Zabchlk67UHp+inKjGvuizUMKTs/suQ9xm6RJP21eo+v7qRmyWUam62UO7QBAGXHf1kBAABK2JuVr8Vb08rcrlqMLWj9ua/8oeHWLzTGPkp1Tana4K2ji5wjpDMfkyw2eYutP3d7DX07b6uvfE+fViG/V3I5+qXdGpY6xrPa1i41BgBwcEi0AQAASuj2zBQNeudfvTttXdg4o8QG7e9v7qEoW4lfr7Yv0DjTQ7rWOkmS9KX7LA1wPqdFRtFe6uKJ9uZ92fpq1iZfOS4q8p1+CdGlxxa/t/uhc9pE3DcAIHIk2gAAACG8OHF12Pf/rtvnV25RK16fXn2CJMkqt7InPSN9fJaam3dql5Gkq5wP6RXbjbrq9Lb69c6evnYx9qIE+a6xi7QjPc9XvvSE0LPUberE+5V3ZeSHiCwSZbPoljOaq0WtOF3ZvVGp8QCAsiPRBgAACMPl8YZ899APS3zPNkvBgWW1E6PU3LRd4+wjFDvrJcnw6GdPd52d/6Kmezvpo6u66ZF+x6ldvURf254tQh9gFhfmVPCf7+jpV56wOPAKsaDjPqeNptx7uuKjgi91BwAcHE4dBwAAKCbf7X8Q2b4sp+okRgXELd2W7jvkTJI6NUiSvB4lLvpAv9pHKcrkUroRo8SL3tIdY6J9cS1qxQX0ZY7wLu6SOMgMACon/usMAACqvEVb0/TXqt2lxjndXr+DyCT53VFd3OJtaX7lmq4d0ucDVOPfkYoyufS3p6P65r8gdbjIL656rL1sgy+Dfx7sdcj6BgBEjhltAABQpTndXp3/zr+SChLRhtVjgsblOj065YU/lZrt9KsPlmi7PV49/tMySZJJXg2xTNWTad9IqXny2GL1WM7lGuvppeY14+QutvT87j4tK+rHCirUzwYAOLyY0QYAAFXa9DVFd0zvKLbUu6TZG/cFJNmStDcrsO7XpTslSfW1R1/ZRukZ22eye/OkJqfKcussjfWcKcmk9XuytXxHhq/dJ/9sLNPY7+pdemJ+WZjD0gAARwYz2gAAoEpbsbMo0b30w9na8Fz/oHuirSH2Se8LMqM9dcUuXWr5S49bRyvelCunySH7OU9LJ9wgmc2SlvliF2ze73vOzHeHHOfgLg30w8JtfnV9jiv9zutRF3ZQUoxdLYPs/QYAHBnMaAMAgCrt1clr/Mo3jV4QNM4SItH+dr7/nm1l7NQFK+/WC7aPFG/K1XxvKw3yvCCddNOBJNvfpOUpvufkMPuzLz8xcGa6Q4PEIJH+TCaTHu7XRoO7Nig1FgBweDCjDQAAqqw8lyegbvKKXUFjrUGSZEnasCdbM9buVc8WydKS76TfH1AvS7ryDZtedl+sTzz95Q0zd7EjvWi5+iuXdAoZF+xkcwDA0YkZbQAAUCXluz3qMWpqQH23xtWCxpec0b69Vwvf892f/CF9e6U0/kYpL12Lvc00wPmsPvKcK6/MurSb/2z0V8NO9D1vTS1KtM9oXSvkeE2m8l3xBQCofEi0AQDAUeWv1bs15OPZ2rY/J2zc+t3Z2p/jirjfknnutac0kST1N8/WH44HpFW/SGab1OtxXeh8SuuMoqXa9ZKi/do2Ksfp3+W8ShsAUAmRaAMAgKPKtZ/N07/r9unBcUvCxsU5gu+Qm1/scLJCXq+h0bM2+9UlGpn6KOYdvWt/U9VNWVLt9tKNf0mnPyCPLP5j6tnEr5wQZYvgJ/FnEpk2AFQVJNoAAOCo4PUa+nrOFl95V0Ze+HjDiLjv/y3erh//2+4r9zEvkPX9HjrL+6/chllvus+XbvhLqtNBktSqdsEJ3/FRVm16fkBAYh0fFZjkN68ZG3YMJWfUL+xSP+LxAwAqFw5DAwAAR4Wfl+zQo+OX+srr92TLMIyQe5sz8kIvG0/PcSkxpig5Xrqt4AqwBGVpuO0rDbb8I2VLabHNdFXqtVpiNNdtZptvHrtWfJTW7MrS04PaB+3fagmcy3jz8uPD/nwlf4rnLugQNh4AUHkxow0AAI4KY+duDahLzXaGjD/v7X9DvnN6vH5lr2Goj3mBpjge1GDLP/IaJunkO/Tryd9qidFckrRoa5ovfsa6vZKkKJv/EvJw2tUr5aquEpl2WfoGAFQuJNoAAOCoEGwp+K6M/HL1tXJnRlEhJ1Xnbxiuj+2vqJYpTeu9dXWx80np7GfUv3MTX9ibU9dKktbtzvLVbU0NfSBbWQ8RNxdr8P6VXcvWGABQqZBoAwCAo4ItyHLscDPaJd18enPf85yN+woeVkyQ9+0T1TltsjyGSe+7z1V/5yhdf8XlkqSkYsvLC7/15P+W+erOaV8n5PeK/7tA8avCQimelzepUfZTywEAlQeJNgAAOCq0rZcQULd+T1aQSGnd7syAujt7FyW7KTu3Sd9fI303VOacPVrjra/Bzqf0vPsK5cuufh3qSvK/23p/TkGiPXP9Pl9dwwiv8br0hIalxhT/lsPKsnEAOJpxGBoAADgq1EuMCqgbPmG5rj65SUD9Pd8uDqiLtlkkGRpgnqNHN34mmTIlk0Vvu87Vm+4L5VT4K7ma1gh/ang4kSTkMfai5LpWvKPc3wIAHHnMaAMAgKNCvttbetABS7enB9SZsvfoPdvresf+ppJNmVKtttINU/Wy+1K/JDvW7j+b/O6QLpKkPJenTOO9qGuDMsVH2Syaet/pmnb/GYoNcQc4AODowH/FAQBApZee69Ko31eVs7Wh88wzpXduUz/LfrkMi971DNKdN74vk9UhaYdfdOGy8UJxB5LerPyyJdojB7XThj1ZOrNNrYjbNK8ZV6ZvAAAqJxJtAABwxGTnu5Xj9KhmKUulf1+6M+I+t+wrOgm8pvbrOdunOsuyQMqVtjla6MaM67TCaKJbZJM9SPsmyf7LvAtnl1fuzJCr2LVgL17UMew4YuxW/XjrKRGPGwBQdbB0HAAAHDE9Rk3VCc9OKfX08NenrA35zu+qLklP/bxckqELzdM1xfGAzrIskNdsk3o9pvdbf6wVRhNJUm6IpeCmEvdyxUcVzUuc8dI033Mcy7sBACGQaAMAgCPCMAxl5LklSbeOWRA2NiUjL+S7z//d5Ff2ZuzQp7aX9Kr9fSWacqS6nWW+6W/p9AflsBfNnM9avzeicRbfL709Ldf3vG1/6Du0AQDHNhJtAABwROzNKprFnr0hNWTiuizIwWbFNSpc6m0Y0sKv9Ma+m3WmZZHyDatmNrlNun6qVLudJP+7uGslBJ5ibjWbdHGJQ8xCzVwXHz8AAMWx5gkAABwRPV/406+cHeKwsR8Xbg/bz66MPGn/ZumXu6X1fyrBJP3nbaEHXDfqj6tulMxFS8HrV4v2PZtNJhmG4Su/O6SLeh9XK+AO65KnkPvG36JG2HEBAI5dzGgDAIDDbuPe7IDruuzW4L+W2CymoPWSZJZX1Zd+Ir3bXVr/p2SN0rOuKzTYOULrjAYym/3bXtKtaLY6z+WRs9jhZj1b1ghIsiXJagk+rtNa1Qw5LgDAsY1EGwAAHHbB7rlesHl/0NiSh5NJUtMasWph2qZx9hG62/2p5MqRGveUbpmpjzznyhviVxyH1aK2dRMkFdzLnVNsFj3aFnzmOpgTm1SPOBYAcOwh0QYAABXmlyU79PhPS+X2eMPG3fnNfwF193+/OGjsN3O3+JVtcuvXTjM1MepRdTGvU4YRLZ37mnT1z1Jyc1/cA31bB+1vxYFTyt+YskYz1++TJCVG2/z2b5dm7qbUiGMBAMce9mgDAIAK4fZ4dfvXBQl0pwZJurhbwzL3kefyKKrYzLLXayg91+UrdzKt0+vRnyjm382SpMmeLnrCda1md7tKkvz2XA/oUDfstxZuSVPMgSTe6Q7/DwNmk+Q1woYAAOBTphnt9957Tx07dlRCQoISEhLUo0cP/f77734xs2bN0plnnqnY2FglJCTotNNOU25u0VUYqampGjJkiBISEpSUlKRhw4YpKyurYn4aAABQYTLzXJq4LEV5Ie6bLmnxtjTf8+7M/HJ9s+/r0/3K4xZukyRFK0+PW7/Sj/bhaurdLMXU0OZeb+sG131KUbLScwqS8ad+XuFrWy3WXur3ZqwruOKrVZ34sHHPnN/Br/zbnaeW/sMAAI5ZZUq0GzRooOeff14LFizQ/PnzdeaZZ2rQoEFavny5pIIk+5xzztHZZ5+tuXPnat68ebr99ttlNhd9ZsiQIVq+fLkmT56sX375RdOnT9eNN95YsT8VAAA4aLeOWaibRy/wS17DKT7zPHvDvnJ9c/M+/yu+lm5L18nmZZpkf0jXW3+XxWQoo9Vg6fZ52ttkgKSC/dt7swsS+89nbvK1TYgqw8I9I/x0dclV5bUTHMEDAQBQGZeODxw40K/87LPP6r333tPs2bPVrl073XPPPbrzzjv18MMP+2Jaty7aH7Vy5UpNnDhR8+bNU7du3SRJb731lvr376+XX35Z9erVO5ifBQAAVJDJK3bpn7UFs73fzN2iURd2KKWF5PIUJauFbQ9K7n4N2vKcutl/lSRtM2oo8eK3ldC+nyTJ6S5K5vdm5qt5zTi/5sEOUQvlljNahH1f/GeTpOoRzJYDAI5d5T4MzePxaOzYscrOzlaPHj20e/duzZkzR7Vq1dLJJ5+s2rVr6/TTT9eMGTN8bWbNmqWkpCRfki1Jffr0kdls1pw5c0J+Kz8/XxkZGX5/AABA5N6btl5NHv5Vb05dG1H8DV/OL1P/uzLy9M/aPRHFrtvtv2Xs82tPCAxaMUF65yR1S/1VXsOkz91nq2/+C4o/kGRLUo24omQ3x+XRqpTy/35wSovksO+nrNzlVy5LEg8AOPaUOdFeunSp4uLi5HA4dPPNN2v8+PFq27atNmzYIEkaMWKEbrjhBk2cOFFdunRR7969tXZtwV/qKSkpqlWrll9/VqtV1atXV0pKSshvjho1SomJib4/DRuW/XAVAACOZS9MXCVJenXymnK195ZyEthJz03V6NlbwsYUuv3rhX7lM1oX/W5QU/ulb6+UvhsqZe3SOm89XeQcrhHua5StaL92LWsX7av2eg1l5Lp95U4Nk0J+//Qg91/H2sMv8vNwEhoAoAzKnGi3bt1aixYt0pw5c3TLLbfo6quv1ooVK+T1FpzWedNNN+naa6/V8ccfr9dee02tW7fWp59+elCDfOSRR5Senu77s3Xr1oPqDwCAY0nxvdPllZnnLj0oAl6voVUpmb5yy1qFy70NXWyZpimOB6SVP0tmqxY0vl4DnM9podFKkvTPg70C+uvSKEmStCU1R25v0cnh7w7pEnIMr1/a2a9cI84uszn8DLW1lPcAABRX5uu97Ha7WrQo2MfUtWtXzZs3T2+88YZvX3bbtm394o877jht2VLwL9x16tTR7t27/d673W6lpqaqTp06Ib/pcDjkcHDoCAAA5fFlsQPCyuuv1bt1/vH1y9TGMIyAJdZ/rPBfwXZXn5ZS6kZ9ZRulUy3LCirrHS+d97Zmr4xS/urVvtj6Sf4z2pK0bEfBcvGnfl6hBtWiw8YWqhZrV9u6Cb77tPdmOUv9WR48p43+Wh3Z0ngAAMq9R7uQ1+tVfn6+mjRponr16ml1sb8QJWnNmjVq3LixJKlHjx5KS0vTggULfO///PNPeb1enXTSSQc7FAAAEETJ/cWlCXandJyjzP82r/H/bQ+o27a/6MpPq9xqvvoj6d3uOtWyTLmGXa+YrpKGTZHqtJe9xFHfwWadi4+1eN+l6dG8aE92h/qJpcYXzbwDAFC6Mv2t+cgjj6hfv35q1KiRMjMz9fXXX2vatGmaNGmSTCaTHnjgAQ0fPlydOnVS586d9cUXX2jVqlUaN26cpILZ7XPOOUc33HCD3n//fblcLt1+++267LLLOHEcAIBDpOSJ2aU56bkpAXVTVu5Sn7a1y9TP13O26MIuDfzq7NaC5LmTaZ2et32s45YXrHrLqneyBmy8SHnxjXWfxSrDMPTsbyvL9L2yiC32DwfvXBF6mXkha8n7vQAACKNMifbu3bt11VVXaefOnUpMTFTHjh01adIknXXWWZKku+++W3l5ebrnnnuUmpqqTp06afLkyWrevLmvjzFjxuj2229X7969ZTabNXjwYL355psV+1MBAACfwiXShVwer2xhEsf9OYF7usfO26rnB3cMGm+EuIN6d2Z+QN0L/5uv4dbvdLXlD5lNhpz2JNn7P6/9Dc/T5pemSRn5cnm8AfdpV7QFm1N9z/FluW9bUq/WgYepAQBQXJn+Zvnkk09KjXn44Yf97tEuqXr16vr666/L8lkAAFBO2/YHJqx/r94Tcna6tNPFg5mweEfQ+i2pOX77tPct+EmTHQ+onqkgyf3Bc6riznpBfTu3U7X8osPWvp6zxW9ptyTVTYwK+o3WteO1elemX13/DqHPfSm0r9i+7BiHpdT44oYPbFemeADAsYd1UAAAVGEfTd8QUHd9mDuys51lO1385UmrddfYRSHfb9ibLWXslL69Usk/X616plRt9tbSEOcjus91i9yO6pKkWHtRsrt0e3pAP+NvPSXi8d57VqtSxz2oc9HBbiX3gocy97He+v2uU9WkRmxE8QCAY1fZTzYBAABHjS9mbS5TfFmv8Xr7r3V+5cbJMb5l3yZ5Fb/kC2nu81J+hgyTRe+5BuhN9wXKk/9tIiaTSTF2i3KcHu3PdqrkavQ6IWa0XZ7Ag9sc1tJnqIsfolbyZPRQasVHqVZ88HEAAFAcM9oAAByD8t2eoPV5ruD1kfruph6SpFamrfrePlK1/nlUys+Q6nfVmvN/1YvuywKS7EI5zoJvT121W4YiW8IebHu4w1r6rzd5IX5+AAAqAok2AABHGafbq/+27JenHPupC01YFHxf9W9Ld/qVm9UsWiYd6tCz4qzefN1n/U6/2h9VN/MaeWyxUr+XpGGTlZXUOiC+e7PqQftxR3hS+rtDAk8Mrx5rL7Xdwf6DAgAA4ZBoAwBwlBk+YbkueHem3pi6Nmzc1lT/g9Cqxdh8zxkhloi//Mcav/KPt5zsey4tr+9hXq6kL87QHdafZDN59Ienq9Ze9Kd00o2S2aJ1u7N8sa1rx2vRk2cpOS747HbxRPip80IfPtatSXW/fwyQIruKK9J92QAAlAd/ywAAcJT5Zm7B3dNvTl2rjLzAq7gKlbzW6/RWRddSRTqjazEX7V92e/33Q2/YU5A4JylTL1nf1zf2Z2XZv0EpRjXd5LxHN7ruU15M0QngD/2w1Pf8212nKinGf+Y5rtjd1mnFrhgb0LFu2DGOuqBDRD9LcTed3lwdGySGTeIBACgvEm0AAI5iwz6fF/KducQhX3USo33PwRLtZUFO+7aai35VKLlU/ZpP52qQeYamOu7Xxdbp8homubsOU/Q9CzTJe4IkKSfEKebFE/hCS4af7Xsufgd3jRCz3oXiit2DPahzvbCxharH2jXh9p66+uQmEcUDAFAWJNoAABzF5m3aH/Lda5P9l4Fff2pT33OuMzDR3p6WG1DnP6NdLNHet17PZj2hN+zvKtmUqdXeBrrIOVymAa8oMSlZJzYp2Hu9P9sV8nslmYt969HxS8NE+ouyFZ0y/vqlnSNuBwDAocL1XgAAHEGGYWhXRn7I66tK6vL05Ij7Lrl0PNZe9Nd+TpAZ7ckrdgXUWYslvzvT8pSQ7JFmvCbNeFWnWpzKM2x6y32BPvScK5esKgyvFluwHzw1xykp9CnnFaF5zTg92r+NaidERXxVFwAAhxKJNgAAR9Brk9fozT/X6cw2tfTelV1KvQM6NdtZ7m8VWwWur+ds0XPF9ja7PF6NW7AtSJuixPWZN9/RV3W+lVI3SJKmeTrpSfc12mLU9sUUJrqFJ3+nZhWMt/iy83euCDwp/GDdeFrzCu8TAIDyYuk4AABH0Jt/rpMk/blqtx75MfRy6VUpGfpw+vqD+pYlzGzv61PWBNT9emdPSVJNpekN29v6yj6qIMmOq6NF3V/XNa4H/ZLsFy/q6Hsu3B/+69KCa8Se+XWl712nhokH9XMAAFDZkWgDAFBJ/Lhwe8h357z+j577bVXQd3+v2RNQl5Ke51f+6/4zAg5HK27uxtSAunZ14qS5H2mq434NssyUxzBJJ90s3T5PT29sLamov+RYuy7p1tBXTs8t2Ju9ZlfByeTj/yv62RpUiwk5jpKGD2wbcSwAAJUFiTYAAEdIek7g1Vy7M/KCRIb30LglfmWn26vuo6b61TWtESuz2aRzD1yV1a1xNd87wzACDlVrZ9oofdxH+u1+JZhytNjbTOc5n5FxzvNSVIIWbPaP//Cqbn7l609t5td/efVonlzutgAAHCkk2gAAHCHr9mQG1M3dFDizXBqnx/9+609mbAwZO6BDQaJd/ITxralFz3HK0ZPWL/Wz4wlpx0LJkaDHXdfqAudILTeaaub6fQF9vnpJJ3UtlrhL0nF1433P+4P8g0Kk3J7yJ+kAABwpJNoAABwhn/27KaCuPJO/JQ9I+3Fh4KFmhXIOXLO1Mz3Pt7zbYxiSDPUzz9EUxwO6zjpRZnml9hdJt8/TaM9Z8h74lWFLak5AnwMOzJIX57BaFH3g2q3MvMgT7Ri7/2Fwbi+JNgDg6EOiDQDAERIshayItHLt7qyQ7zbuzfY9L9+RLkkyp23SZ7YX9Z79DdUx7ddGb21p6Hjpok+k+Dp+7acH2Q8e6qT0+KiCy00y89y+uqY1YsOOvXDGvZDH6w0RCQBA5UWiDQDAEbI+SEL89ZzN5eorLSeya7+K3dYlrytfmv6yGn5zpnpZFivfsOp194Vad9FkqfmZQdv/vixF3ghnmQsT7Yw8l5rVLEiwnzj3uLBtjm/kvwTdxdJxAMBRiEQbAIAjYOb6vVqVErhHe/aGsu/RlqTRswsS9NIOHttx4DTyk0wr1eC7s6U/n5bZk6cZnnY6x/mC6p8/Umd1bOzX5tVLOvmVg51yHkx8lE2StC01Vxv2FMykt66TELaNUWJO33sQB6kBAHCkkGgDAFCBZm/YpyYP/6qL3psZNu6tqetCvovklO5p95/hV3YemPldF2SWfP7jfXzPfRoaesX2rr51PK0m3m1SbE2Nrve4rnQ9qo1GXfXrELjfumeLGn7l4nvCa8Q5Qo6xcEb7wR+KTkWvlxgV5qfy36PetXE1ndiketh4AAAqIxJtAAAqSGq2U5d9OFuSNH/zfs0Lc4J4mCuttTszv9RvNakRq5tPb+4r108qSGCDrequEeeQPG5p9vvqPWWABltmyGuYNNrdW7p9nh7f0FaFd2LHOawB7aNLHFB23/eLfc/vXHF8yDEmHJjRLs4U7geX/x71H245WVYLv6oAAI4+gX+bAgCAcrnsw1l+5TW7MnVCOWZki+ei+7OdWrPLf4n51PtOlyS1rhPnqyvcyxw0L90yW/r1fmnXUtkkLfY20xOua7XEaK6LrOGXcksFS8AToqzKKHaoWaGTmoW+57pGnL3Uvks6mDu3AQCoLPhnYgAAKsiaXf7LtoNdhVUo2H3UhdKK3Tvd+9W/demBWfJCzWvGlWyilTszJPnPGCcrXS9Z35c+7SvtWipFJWl5l6d0gXOklhjN1bxmrB4qtqz7rLa1Q47p5zt6hnwXytAejUsPKiHSg9YAAKjMSLQBADhEPvh7Q9D67+ZvDdvuo+lF7UrekR3KmDlbJBXMgJvl1ZWWyfrTcZ8utk4vCOhylXTHQm1rfpnvTux6SdH636Idvj4u7dYwZP8x9rIvgmtRK77MbUizAQBVAYk2AAAHye3xakyQa7k6N0wKGv/guCV+5TNa1/QrZ+S5FMqNpzXzPQdbZf3+mLH6n/1xPWP7TImmHHlqd5SGTZHOe0uK9V/mvb/ElWCrUjJCfjfWEfyu7IrW57iCWfVG1WMOy/cAADgU2KMNAMBBGj17s0b8vCKgPliinefyBNQ9dV47ZeS6NfDtGZKkajGh9zZf2KV+0PpqypAm3KGPXV9KZindiJGj7whFdb9eMgdPkpdt90+sL+oaekY72nbwifYPt/QoNaZh9RjNe6yPEqL5FQUAcPRiRhsAgIO0cEta0PrPZ26KqM5uNatDg0Rf+efFOwJiCrUpdg+1ySSZ5dUVlqn6y3GftPBLSdL37tN0Zv4rijr5ppBJdjB1wly9Fey08B9vPbnUPj++qpvvuWvjyA6GqxnvkMN6eGbQAQA4FPjnYgAADpLVHPrKqjkb9vmdzP3876sCYkomldnOwFlvSbrweP/Z7KT9SzXe/qQ6mQv2dHtqtdclWwdrgdE64rH7xnVhhzK36VA/sdSYPm1r64G+rdWQpeAAgGMIM9oAABxCN49eUGqM3RrZX8cvX9yp4CEnVfr5bp0x/XJ1Mm9QhhGt4a6rtfPS331JdriDzYJZsj29TPFS+H9gKO62Xi10Xqd6Ze4fAICjFYk2AAAhzN2Yqrf/XCtPKVdOrd+b7Vd+/8quvmd3BNdVOQ4k2q9d2ilsnFmGtOAL6a2u0oLPZJKhHzw91Tv/FX3h6avrv1rki72uZ9OgfYRKjWvFO0odZ0BfQZaTAwAAlo4DABDSJR/MkiRNWLxDf9xzesi4xVvT/Mp92xXdR52Z5y71O4Uzw+3rFSzFrhZjO9C26PTxzqZ10sdnSjv+K6io1Vbq/7KScltozxfzJUl7s/J98U1qBF+qHSrtj7GzJxoAgIpCog0AQCnW7MrSqpQMv4PIwgk102sEu4+rWHzhXu18t1eStCMtTzWUroes3xTch71Dkj1eOuNh6aSbJItNvYv1Uz8pWnuzCq7sslvKtmjNXMbZ6ecuKPuebgAAjhUk2gAABOEtseR76bb0oIl2ybhw/lm71698Zptaeuvy433lwr3aTrdX8rg0c/QI/ekYowRTbkFA5yFS7+FSfG0Fc3KLGlq8rWCvdVmXdZc10b7ipEZligcA4FjCHm0AAIL48b/tfuVg919L0rgF24LW92pdM6BuS2qO77lNnXh9es0JinUU/Zt34V7t7loi471TdG3Wx0ow5WqRt5lm9fpWOv/doEl2sxqxkqQdaQUJ+WmtAr9dmkjONZt63+lyWM169ZLwe8kBADjWMaMNAEAQo35b6Vd+4n/LNbRHk4C435btDNq+VZ14/bV6j19d8UnjJ85tG9DGkbVN79te0zmWedJeaa+RoBfdl+p7z+naePo5oQd7oN//LSq4f3t1SkZpoQGa1owL3f8BzWvGafUz/UqNAwDgWEeiDQBAEPuynQF1eS6Pomz+h4aFWjmeEGXzPRuGIZPJpOz8ooPREqOL3suZI/37uqL/fUPnWPLkNsz60nO2XncPVoZiSx3rhj3+p57vysgPERnaaS1rlLkNAAAIjqXjAABE6O81ewLq6iVGBY29qGsD33PhFV+FB5VJUq0Eh2QY0or/Se+cKP39gkzuPM30tFV/5yiNdF8VUZIdTMPq0WVuw1VdAABUHBJtAAAilJHrCqirEed///SFXepLkt/e64vfL7gmrH5SUQJcK3ej9OUg6burpPStUmJD6eIvdIXrMa0xGvr1ed0pwe/EDuWq7k1CvmucXJS8n3pgFvvn23uWqX8AABAeS8cBAMeML2dtUna+R7ec0bxc7d+btl4Xd/NPgt/+a51f+YXBHSUV3Y0tSYsO3LO9dHu6EpStN+pOkt4bLxkeyeKQet4tnXK3ZI+R9GvAd6/r2aRM40yKsYV817pOvN4b0kV1EqN0fKNqZeoXAABEhkQbAHBMeHnSal9SfHqrmmpbL/Sd2Gk5RUu8k2Ptvv3aG/Zmh2oiSdr0/ADfs7XEMd670nNkWjRaUx1jVXP/gcPK2pwr9X1WqtYkbL8l94WXpn+HumHf9yvlPQAAODgsHQcAHBOKzzxf/P7MsLErd2b6noMdilYoNcw7S7FE+3jTWjm+6KuXbB+qpilDWy0NpCt/lC4bE5BkN68ZuC+7LIn2M+e391u2DgAADj8SbQDAUWlPZr7v3ujSlIzLdga/E1uSvF5Dl38021c+rq7/zPe63Vm+53/WFh2O9sz57f3iTCaT6mifXre9rfGO4UpKXaJMI1rPuIZoRP0PpRa9g34/WFIdZQ3/1/XTxb596QkNw0QCAIDDgX/yBgAcdQzD0AnPTpEkLXuqr+JKmcF9cNySiPueuX6fX9lm8V8CviU1Wy1qFdw5va/YKeKDuxSdMi5njjTzLf3leFnRJqe8hknfe07Xy+5LtEdJeqtr6MPNgiXaVkv4RLv4MnVbKbEAAODQI9EGABx13MUur96RlqtWtePDxs/esC+gzuXxBiSlHq+hMXM2+9WVTHxXp2TpzDa1/epMJinabim4rmvZD9Lk4VLGNkWbpLne1hrpGqplRjNf/LkdQ++RjrKVPVE2QtzlDQAAjgwSbQDAUcfp9vqe03ICr9wqqXhiXmjW+n06rVVNv7p7vl2k35el+NWd3qqm5m5M9ZXtxZZxbzxwOFq/9nWk7QuliY9IWw8sO09sqNv3nK9fvN0l+c+Kh7uzmhlpAACOfvxtDgA46hSfoX7ut5VhY71BkmxJmrpyV0DdhMU7AupuOLWZX7n4jPNXszerpvar18oR0ke9CpJsW4zU6zHp9nn6xdtDJZPsTg2Two63+F3bknRik+ph46XAE84BAMCRRaINADiqLNyyX8O+mO8rF95RHYrT4w1a/8WszUHrS7KXOIisMHHPzcnWrZb/6S/HfbrYOr3gZcdLpTsWSKc/KNmiddPpzUp2p7v7tAz7vQf6tvYrf3tT91LHOLBTPbWpE69hPUPv/QYAAIcPS8cBAEeVC98NfzVXScGWjUeqZJItSXM3pmpo4mKlj3tQD9oKlpn/522h4294X2p4gl9s96bJ+uDvDX51Z5RYrl5SUozdrxxumXmhaLtFE+8+rdQ4AABweDCjDQA46i3YnBryncdT/kT70X5t/MptTZs0ZNWt0ndXqY43RTuN6rrbeatib/0zIMmWgs+mR5I4F3r2gvalBwEAgEqHRBsAcNQb+sncoPWGYej3ZTt95WtPaVKu/pOVruesH+kX+2Pqbl4pWaP0hvsCnZn/sn7y9lSrOolB23Vvmlyu7xWKtbPwDACAoxGJNgDgiNuTma9xC7Ypz+UpV3tPiOXhE5el6OEfl/rKwwe207ibe/jKRin3Ypm8LunfN/WX415dYf1LZpOhnz3dteeaGXrNfbFyFRW2fWKMTa1qx5XhJykcZ1ud066OBoS5BgwAAFRe/FM5AOCIO+HZKZKkJdvSNHJQ2ZdL57uDH3g2J8i1XMVXbk9bs0e9WtcK0tLQOeZ5unL+o1L6JiWYpCXephrpGqr5Rhu9lBI+wS7umpOb6tHxS0sPLObaU5rq2lM42AwAgKMVM9oAgCMq3100i/1lKSeBlzYDXZK5WFZdePf2ut1Zvrp5xRLxwtPEO5nW6Tv7SL1vf12W9E1SXG094rlZg5xPa75RsGc72m7xtftwaNewY8hxuss0ZgAAcPRjRhsAcETlufxno71eQ+YQ90LPWr8vaH0omXmugDqbpejfmIuvOB85ZpJet72t8y0Fp5rnmxxynHa3dPKdOmF5mr75brEvdmtqru+5tHuxG1aP8T2PHnZSmcYPAACOTsxoAwCOqJL7st+YujZk7NxNoU8XD+b7BdsC6qJtRbPReS6PlJeh3N+f1MPrrvQl2T94TpX5zgVSr0clR5wuOL6+Xx8vTFzle66dEH4Z+dlta+vxAcdp3M091LNljTKNHwAAHJ1ItAEAR9S2/bl+5XCJdo4z9GFpJQ9ES812Bo0rnNG2yKMe+/8nvdVF0XPeUJTJpVmethqQ/6wuHPmzbNUa+tqYTCZVj7UH7a80JpNJ15/aTN2aVC9XewAAcPRh6TgA4IjasCer9CAV7OX+cPqGkO8/+3ejrj+1mSTp+/lb9cC4JUHjTCbpdPNiPWodo9YbC2a8s+Oa6O7UCzXZ21WSKehd1yFWswMAAAQg0QYAHFEb92ZHFDdiwnK/co04u/ZmFc1a70zP8z0/8b9lwTvZtVzH//2gettnSJL2G3Gq1u8JjXOdqcm/hp5Jl/wPVgMAAAiHpeMAgCNq+to9AXVZ+YEndX8zd6tfed5jfXRn75a+cr2kaN9zyQPWaipNz1k/kt7vqeopM+Q0LPrI3V+n578qdb9ZtavFlzpOS5Ap7ZrxjlLbAQCAYw+JNgCgQk1ctlMXvvuvPp2xMWDfdEl/LE/Rsu0ZAfVTV+4q9Tsmk0ndmxbtey5+QFkhh5y61fKT/nLcqyusf0mGV3sa9lUf58t61n2lMhQnt8frN1v91uXHB/1esBltu4W/RgEAQCCWjgMAKtTNoxdKkhZuSdP0tXv0+bUnhoz9IMSea4fV4ldevDXNr/zqJZ0kyW8vtdPt1Y60XNVLipZJXg0yz9QDtm9V31RwJdjO2Laqe8mrWpHfQlvWzvUbQ8aBa8CSY+0a2Kle0DH1a19HH8/Y6FdX8sR0AAAAiUQbAHAITVsduCy8uAWb9/ueo20WdWqYqNkbUv1ODHd7vBr0zr9+7fq2qyNJ6tgg0a8+3+2VZ/3f+p/9CXU0FyTF24wayun5mJr1ukqyWmVa4z+mlyat9j3vC3FSuSTd37d1QKL9+mWdw/58AADg2MSaNwBAhclxBu6tjtTi4Wcr98De6md/XeGr/3LW5oBYh7Xgr69Yh1XDejaVJLUybVWdX4bK8tV56mjeqCwjSi+6LtW4k35Uq7Ouk9Va8G/L5T3TLMpm0YLH+/jVndqyZvk6AwAAVRoz2gCAChPsSq3dmXmqFR8VUF/ywDO71exbIp5d7L7sqasC92tbi+2N7pyYoxesH+oiy9+ybDJkmK360tlLb7ov1D4lanLXZn5tTSr/6eHVYsp3lzYAADi2MKMNAKgwvy7ZGVA3aXnwg80e/iEwKT+rbe1Sv9GuXkLBQ166NHWk+v01QJdap8liMpTT4lztuvJvDXdfq30qWFbeolacX/vjGyXJWs5Lsc1cpg0AACJAog0AqBD7Q+xv/nD6+qD1vxRLyu/p00qSdNNpBbPP1WJsvndGiYPL1+xIleZ8IL15vPTPK7J68zTX21oX5o/QS4mP6ss1/ou1TCXWisc6rFr2VN+gYyKPBgAAFYGl4wCAChHs7mtJ2pqaG1Dn8vjfc107wXHg/y5YYp5b7DTvokTb0ADzHD1g/Vb6/cAseXJL5fd6UpeMNksyaeG/m9StcbVSxxplswStP65uQqltC312zQkRxwIAgGMLiTYAIKTlO9L15czNsllNuvPMlqqVELjXupC7lDuzC3m9hlo+9rtf3YVdGkiSqscW7IHOc3mV43Qrxl7w19SJppV61Pa1OpsPzI7H1pJ6PSIdf5UcFqukX3197cnK9z1vev7/7d13dBTV2wfw7+4m2dRNCKQQktAJhN4JvYYSEURFKYIFK9jbi6KiiGBFVESs4A8RBASlBpAOoYcSeq8JIUAK6bs77x+bTHZ2Z7aEDSTk+zmHc3bu3HtnhnEjT+69z41z6J6KzRzR2uG6xfdKREREZImBNhERKYr7Zqv4ee6OCzYDV73FKLWSxfsuWZV5FGUR9/bQQOumRr7eiOu3CuCtP4VX095HO+1OAEC2oIV391eh6vgioPW16gcAzl/Pceg+5ERW9bZbZ9Kgxrh0M9dqazEiIiKiYgy0iYjIJQoN8iPaMXWqSo7lMpMXU6lUyNcbEYybKFgyDri0BO0EI/SCGn8aeiKp3nP4tEcfl9xv4nt90HLSWqfbPRZTyyXXJyIionsXA20iIpJldHAqeLGbOfLJ0BLOXJdMA7fUqZ5ZIJ6fhVfdFuJpzUp4XzRNAV9taIvP9I/gjBCGc4+7JsgGgCoWU785Qk1ERESuwqzjRESViFLCMjlL91+2KhMsU4Cb+WTlUcVz09aeUDz32UPNAX2+KZP49BZ42W0JvFX52I8GwJPxeK7wVZwRwmze69yn2luVxTqwVZi5ke1rOlWfiIiISAkDbSKiSmLiv4fR5IN4bD+d5lD9zSeuWZX9s/+KYv0LFmujBzQNFT8vSZRvp4YRNc7/C3zXBlj1FpCThhzfWni24BW8W+VLILKDQ/caXJS13Fx4Ffvrrc091DrcqfpEREREShhoExFVErO3nwMADP9pJ5IzrLfcsrRUJqg+m5YtW9dgFJBlNlpezdcDj7aNFI/TzDKBmwjoqd6HFR7jgSXPAOkXAN9Q4L5p2H//asQb20FvkVtN56m82slyuzAAOHQ5XbF+sVUvd8HTXWrjwPuxUHMTbSIiInIRrtEmIqqEYqasd3rrKwDQKASj565LA/CWkVVgUJhm3lp1HG+7z0c79XFTgdYf6PwK0P45wMMbmjPXAQCFRqNkqvqvNvatrh/sZ1X2bly0rUcBYNo325F6RERERM5goE1EVEkV6I3itlqW/j0gP9X7q7Un8FKv+lblm45Lp5mH+XtaJ1O7ehj4bxIWa017aBvUWqg7PAdV51cA70CxmpvGdE96g4BLN0tG3sMCvBSfRe45GlW3Dr6JiIiI7gROHSciqgS2n7Jelx1/OEWx/kt/JjrV/0fLj0iOH2odAUNRoB2uuoYv3WcCMzsBJ1ZBL6gxT98D2c/tgSr2I0mQDZSMmhsFARdvlqz7ru7v6fD9PNutDrRuGqeegYiIiMhVOKJNRFSB5RYY4OVhP6Ac/vNOq7Kv1p7AwObW2byV1mEruZFtva1X4zAdbqRexgduczBCsw4eKgMAIK/B/RhwqCvOCGF4pFqEbH/Fs9ONRgHT150Uy1Uqx9dQ7zxzw4knICIiInItjmgTEVVQKw8lo9H7q/HzljOlaq8UUD8/d6/0uHtdm/08OHO75NgHuVBvmoquq/vgCbd4eKgM2GpojLRhq3C2xwxxqy6l9d5qVfGIdsl0cV+tc78XvpxuP9kbERERUVlhoE1EVEG9PN80vfvjFcr7V5fGsZQsyfGbsVEW5zMlx8UBuwcK8YRmFTZrXwE2TYWq4BYOGmtjRMF4jCx8FwuuBGNOUeZzW4oDbb1REEfrn+pc2267R9uWjJB3qVfNbn0iIiKissJAm4iogio0lCQbc3a6tzPUahVGxdQUjz9ZeUx6HkYMUW/Geu3r+MD9f6iqygIC6wIPz8Y3dX7CNmNTAMDn8ccxf/dFu9fzK9rGK+1WPhbvvQQAqOLtbred+Xru9wcykzgRERHdPQy0iYjuAT2+2Fim/X94f2Px8+YTRRnGBQE48i9WefwfvvL4AeGqNKQIVXCszSRg7E6g8QPoGhXk9LUCzILq/KLNtEP9lTOOFwsP8Dbrw8Pp6xIRERG5CpOhERFVMIIgYMycPQ7Xf+wX60RojnpnQEMAlonIBODkOmD9JCB5P6LUQIbgjZn6+3Gz6RP49L4Opb4eIL8eu16wr912netXw4I9F6Gw9JuIiIjojmGgTURUwVy6mYv/jqU6VPf89WxsOSnd2mvVy13Qf/oWAEDihZtoGVlFsb3luXaqo3jD/S/gj+MAAMHdB9/m9sHP+jhkwgcru97+lG257OI1q3rL1JS6r1l1eLpr0KSG7rbvgYiIiOh2cOo4EVE5sHDPRQz8ditSMvLs1nXXOP6jO6/QaFWm8yqZmv3jZtsZywN9iqZgX96Lf/2/xF/aSWinPg5otEDMOEyu/ye+0g9FJnwAAA1D/Ry+N2c48swqlQp9okNQ3YFp5kRERERliYE2EVE58Oaigzh0OQMfrzhit2784RSH+5WbRm0wS6K2Kkna16nUW5Lj2vpzwPwRwE890Sx/LwoFDebqewEv7wf6TsbP+6T11Q7O2/7v9W4O1SMiIiKqiJwKtGfOnIlmzZpBp9NBp9MhJiYGq1atsqonCAL69+8PlUqFpUuXSs5duHABcXFx8Pb2RnBwMN58803o9frbeggionvFjewCu3W2n06TLc8rNFiVWU7DHtQiDDov5VVDHy47DACopUrGdPfvoJ7VGTi2HFCpscWnD3oUfIkJ+qcAXRgEQVDsx566QfbXXBMRERFVVE6t0Q4PD8fUqVNRv359CIKAOXPmYNCgQUhMTETjxiUZab/++mvZNXYGgwFxcXEIDQ3F9u3bkZycjFGjRsHd3R2ffPLJ7T8NEVEFt/30ddzILiiZsi3jwo1c2fJZm87g5d71LUqlwXDjMJ3NjNynTx7DVLe/8ZBmM9xURdPOowcDPd7BnJWZuHTdtDb8VOotXDLbTkuJv5f9bbnsWTau8233QURERHQnOTWiPXDgQAwYMAD169dHgwYNMHnyZPj6+mLHjh1inf379+PLL7/Er7/+atV+zZo1OHLkCObOnYsWLVqgf//+mDRpEmbMmIGCAvujOERE96LsfOmsnoV7bO81fTQ5U7ZcLvDdfe6m5FgFhandWVeBlW9hg/Y1POq2EW4qI9LDewLPbgaGzgGCojC4ZQ2xeu+vNuG79ackXfRrHGrVbVzT6jafxRFNw/1vuw8iIiKiO6nUa7QNBgPmz5+P7OxsxMTEAABycnIwfPhwzJgxA6Gh1v/gSkhIQNOmTRESEiKW9e3bF5mZmTh8+LDitfLz85GZmSn5Q0R0rxg0Y5vkuHjvaGct3HvJqmz834ckx32iQyTH/rgFrP0A+KYFsGsWtCo9thuiMSR/IjxHLwKqNxfrWgbNe85Lg/jvhre0ur6bTBKzUJ2n3Wcp3s6rY92qdusSERERlTdOB9qHDh2Cr68vtFotnnvuOSxZsgTR0abtXF599VV07NgRgwYNkm2bkpIiCbIBiMcpKcrJfaZMmQJ/f3/xT0REhLO3TURUblkmINt2SnkN9rydFyRl88a0V+y3QCZgr1XNlB3cFzl4SfM3tmhfBrZ9DRTmIMWvCYYXvIPhhRPQrmt/eLprJG0tlwTVDfIRP9cI8JINquX8OKq13TpznmyHl3rVx/RHrYN3IiIiovLO6X20o6KisH//fmRkZGDRokUYPXo0Nm3ahFOnTmH9+vVITEx0+U2OHz8er732mnicmZnJYJuI7lk7z96QLZ+29gRmWWzH1bFeNcV+FuyWBuV+WjcgPwvYOQtbtV8hQJUNADAGN4a613vo8JsBKJpa/lbfKLv3eSW9ZCuyJzvXVqz39SMt8MqC/eJxkzD7U8FrBHjhtT4N7NYjIiIiKo+cDrQ9PDxQr149AEDr1q2xe/duTJ8+HV5eXjh9+jQCAgIk9R988EF06dIFGzduRGhoKHbt2iU5f/XqVQCQnWpeTKvVQqvVOnurRET3lM0n5Ue6zeUWGODlYRqJPn41Syz3Rh5+qrsf+Po5IPcGAlTAKWMYvtY/iI9HT0CAjyeAFWJ9R7bpyjXLcv5Ex1qK9Qa3rCEJtB3dAoyIiIioonI60LZkNBqRn5+PDz/8EGPGjJGca9q0KaZNm4aBAwcCAGJiYjB58mSkpqYiODgYALB27VrodDpx+jkREQFX0nMRFuAlKXNkZva569loVF0HAMjON8AT+XhMsxbPuS1D1TNFgXfVeng5ORbLjB1hhBpP38grCrRLz9HguUVEwG1dh4iIiKgicCrQHj9+PPr374/IyEhkZWVh3rx52LhxI+Lj4xEaGio7Kh0ZGYnatU1TCmNjYxEdHY3HHnsMn332GVJSUjBhwgSMHTuWI9ZERGbeXnwQ/3tKuv5abbFG+v7mYVbtxL20C3MRePAnbNH+iyBVUQLJKrWBbm8DTR/GP+/Gi20mLT+CRc93dO0DKPDRauxXIiIiIqrgnEqGlpqailGjRiEqKgq9evXC7t27ER8fjz59+jjUXqPRYPny5dBoNIiJicHIkSMxatQofPTRR6W6eSKi8urx33ah1v+tQE6B3mY9QRBky7fITBM/eClDcjyuZz2rOoX5ucCOH4DpzfGe+1wEqTJxwRiEqz2+AsbtAVoMAzRuGNE+UmyTkVsoSZxmK8GanOe61bVb54uHm6NesC8mD27qVN9EREREFZFTI9q//PKLU53L/QOyZs2aWLlypVP9EBFVFIIgYPCMbThQFBRHvx+Pc1PjFOtvPH6t1NdqEOInfvZAIR7RbEDzv18Fck25Ly4J1fCt/gEsNnTBqW7S3SDeHxiNP4oymJ9MvSX5hUDb2oFO3Ueozv6MpIdah+Oh1uFO9UtERERUUd32Gm0iIipxLCVLDLKLCYJgtTVWsTNp2bd3QX0BRmjWYazbUoSpbgC5AHThELq+gR6LAlGo8GNe6yadwv3p6uPiZ3cHt+kqVmAo3b7fRERERPcqp/fRJiIiZVl51lPFlx9MVqyfkVOgeE5pWjkAuEEP7J0NfNsKk91/RZjqBpKFQEwofAJ4aR+ymz6mGGTL+XPXBfuVFDzWoVap2xIRERHdixhoExG50Et/JlqV7ThzXbH+N+tPiZ9f6C5d62y+PZe+aNTYDXo8rNmI9R6vA8teBjIuIs8zCB8Ujkb3/K8w19AHSVfzMG/nebGt0hrqYD/XJKEs3k6MiIiIiEwYaBMRuVBKZp5VWfFaaHNGo4BdZ29Iysb2kCY323CsZP32N2uO4mHNRvzn8QY+d/8RkeprgE8w0HcKEgdvxBxDX+TDAwDw5Zrj+GTlMbGtXNI0AHimax2Hn6vYYx1qOt2GiIiIqLJhoE1EZEfarXxcuplT6vaNw3RWZXN3nsfQWQmSMo1ahS8ebi4ea93UgL4A2PMbRux+AJ+7/4ia6lSkCTosDXoeePkAEPMCBDfpHti1q/lKjn218lPIw6t4yZbb8v7AaKfbEBEREVU2TIZGRGRHm4/XATBlzjYPhB11+EqmVdkfO6xHuTVqFR5sVQNvLDwADxQi7ORcYNd8IPMSQgBcE/zxg/4+zDP0Qr9qdTHYw9vU0CLP2oFL6Q7dl1KCNlssE6UFeLs73QcRERHRvY6BNhGRDeYJyRbtvYTGYTo80am2bN1b+cp7ZmfmFULnWRKUnr52y6qOm1oFlT4PozXxeM5tGaqfL5pa7huK7wruw7eZncXp4dV8PcR2KotIe+/5m/YfDFbxOQDg7X4NHWpbrFfDEKfqExEREVUGnDpORGRDoUGa+fvDZUcU636++pjiufkWWb31Rmm/nsiHasf3wPTm+NB9DqqrbuCKEAgM+AJ4+QC+yOwpBtk6Tze82Ku+2FbrrvyjfKKNqd5qmRHtUTHOrcF+2ew+iIiIiMiEI9pERDZsO5XmUL19F25iTsJ5xfOfrDyGZ7paZ//2Rh5GatbiabcVQLxpivkloRq+1w/CIkNXHGgxEF7u0qzeByf2lRy3jAhQvO4jbSMVz8nNHLe8lj2BZiPrRERERGTCQJuIKp3jKVlQq4D6IX526z4xe7dDfQ75frtV2dpXu6LPtM2y9X2Rg1GatRjjtgKBqqJp5AE1ga5voMdf/uIe2EeSMxAWYDtpma211ra23pJbX61WO7duWynRGhEREVFlxn8hEVGlklOgR9+vTcHvycn9rZJ7mSveu7q0fD1lfsTmpgO7fsRW7dcIUGUDAM4aQ+DfdzwCO4wENO4o/GuFWH3hnksY2jZCPJ71WOvbuidzrSKrlKrd2le74tPVxzltnIiIiEgBA20iqlSu3yoQPxfojTYD7RkbTpf6Oq1rVoGfWfIzHW4BG6YAO2YC+RkIUAGnjdXxrf4BLDPGYHvTWEBjPcJc3d8L4xcfEo/rB/ta1VEypFUNm+ctR8K/HdbSoX7rh/jh59FtHL4PIiIiosqGgTYRVSqCYL9OsZmbTsmW38guQKCP7bXJC5+NgVqtQiAy8aTbKozWrAE25QIAcgPq463Uvlhh7ACjnZyUBkHA8atZ4rHSNHGtmxr5eukI/EOtwm32DQBv9o3C5/HHAQCh/p52ahMRERGRI5h1nIgqlaQrGeJng52o26gwc/zf/Zclx+fSsq3qqLOuAKv+D9u0L2Gc2z/wU+UCwdHAw7PxT8wiLDN2FIPs7lFBCPbTyl7rm/9OSo6VVlD/M66TVVmAt/1EZbWr+YifnfklBBEREREpY6BNRJVGRm4hXvhjn3hcqLe9BrtAYY32RIstvp6aU5IwLVJ1FZ+4/QRMbw7snAkvVQEOGmvj2YJXgee2AY0fQGa+QdJ+9hPtJCPVvz3RVvGelPKeNQzVYWDzMElZdJhOsZ9i5snMtG78XwIRERGRK/BfVURUaVzNzJMc95u+xeG261/vpnju9LVs1FddwjT3Gdjg8RqGu20AjIVAZEcsjv4G9xd8jHhjW1zJzAdg2uqrmFyS7x5RwZJj82BYpTimDdSq6u3o44g616smfm5Sw9/p9kRERERkjYE2EVUallOjr2XlO9y2TpBCErLL+zDL/Sus1b6FBzTboFEJ2GhoDjyxCnhyFXrfNxzFE77fXXLIqnlcszCrMku38vXiZxs7eeGZrnXEz0pT0S2p1SqcmxqHc1PjoHFyay8iIiIiksdkaERUacR94/gItm0CcG4rsOVL4PR69NUARkGF1ca2+F5/P957ejhQsyoAwN9sr+qDlzKsevr8oWYuuidIspx/+qDr+iUiIiIi5zDQJqIKbeK/hzF7+zlseKO7JLGXJUEQoDc6nu0r1WKaOQBMfaAJVv8zF2PdlgKzT5gKVRos1sdgpv5+nBJMWb7b16kq2+f17AIIFsPqnu4ah+8JMI1A2/Jm3yicuJqFbg2CnOqXiIiIiFyHgTYRVWizt58DAPT4YiPOTY1TrGdwIsgGTOuui6lgBA4vxaBdn+NRjyQAgKDxgKrlSKDjS3j9syNK3Vj5c9dFp+7DUqjO9hZcY3vUu63+iYiIiOj2cY02EVVYn6w8KjkuVMgSDgCZeXrZcssR5mJadzXcoMcQ9WYkVn0PWDgaXteTkCNo8ZN+AFKe2AXcNw0IrC1pV1Vmf+0Xe5qC3xYRAfhl6xmx3FaCNTknJ/fnOmoiIiKiCoAj2kRUYf24+YzkePq6k3ijb5RVvYycQrSatFa2j/ScQlSxCI6/WLEf6dtnY4PHckSorwHZADz9cav5U+iyqQFuQoeeWtPU7Is3ciRtt77d0+oaXeoH4dv1p7D/YrqkvFZV5anuctw1/N0oERERUUXAf7URUbkhCALOX89WHGW25387zsuWbz+dptimpXkAnpsObP4Co3cNxMfuvyFCfQ1pgg7nWrwJvJIEfbfxuAnT3tQ5RXthP/u/vZL+vDys11yn5xTIXtveemsiIiIiqpg4ok1E5UJeoQEDpm/BmbRsvBHbAON61rdZXy4Yz8gtLN3FM68AO74H9vwGFNxCkAq4JFTDT/o4LDB0x6GBgwGNGtoCg9gkp8A0Ff1YSqbd7ns2DLZbh4iIiIjuHQy0iahcmLHhFM6kmRKQfbHmhN1A++MVR22eN6e093Qd1RU8o1kOfL0NMBYF6cHReOVSdyw3doC+6Edk8ZRtd01JRzlFQbd5jrWFz8XIXseNU76JiIiIKhX+64+I7jpBEPDt+lNOtfll69lSX2/lg16Y6T4N6zzexKNuG01BdmRHYPhC4PntWGrsLAbZ5swTkaXdyrc637ZWoMP3MLRNuM3zM0e0khx/NKixw30TERER0d3FEW0iKhN5hQZ8v+EU8g1GvNizPny1yj9uJi23Hp1Ou5WPar5al9zLmiNXAQjoqj6I5zTLEL3iCKKLllKvNbRG9ycnw722aTQ68cJNSdsw/5LttFRmQ+Pj/z6EAG/rDOOOeqtfQ5vn+zetLn72ctdgVEytUl+LiIiIiO4sBtpEVCYavrda/Hz5Zi6+G95Kse6v26xHp4fOSsD617vf/o0Y9CjcvxArPJahsbooWZraDZs9e+Cjm31wSgjHXH19dC6qPmuTNJP55rd6SI77RIdg7ZGr0BsFPP37nlLflpe7ddI0S+8OaIQv1x7HvKfbl/o6RERERHTnceo4EZW55QeTnW5z5lq24jmjUTkreWpmnulDYS6w+2fgu9b41uM7NFafR7agxS/6/sDLBzAr8A2cEkzTt5MzcsX2qw+niJ9XvdzFan21RmnBt5McCbSf7loHSRP7omVkFZdck4iIiIjuDAbaRORyBhuB8O36a89F1HlnpaRs8fMlSciW7TwGbP4C+LopsOJ14OY53BB88VXhQ+iU/w1W1XgJ8A9Hr4YhYpv5uy/KXutWvt6qTKOwJddvT7R16jkc3dqLidSIiIiIKh5OHScil9twLNWq7FRqFuoF+912328tOig5bhjqh8Zh/qiO63jcbTWGbV8PCEUj1P6RuNbsaXRZG448mNZ7RwR6AwAeahOOj5YfAQC0qSU/YhwVan2/1Xyl67K9PTTIKTCgdlWf23ouIiIiIrp3cKiEiFxujMza5d5fbXa6n+K9qm1RXU2C57LnsVn7Cp51WwFvIRcIjgaG/AS8tA8DdzYWg2ygJHu3ztNdLLNcl13MvE6xblFBFvdo2ubL38u6rpI+0SH2KxERERFRhcVAm4jKrUlFI87WBHRRH8Tv7lOwSjseOLgA7ioDthui8XjBm8Dz24FmQwGNO1KK12wD+P3JdvCTCZ6dEVOnmmy5n6fjE4QmD25yW/dAREREROUbp44T0V1lK7HZn7suYsqQZiUF+gIMUW/G024r0EhtWldtgAaaJoNx396WSBLqmOoVJSyzXGPdpIa/zXs5fCXD7v16ecgnMXNmLXWwztN+JSIiIiKqsBhoE5FLpWbl2a9kptdXm+xXyk0H9s4Gdv6ArzxMGcyzBS3mG3ri4bEfQ1e9HpL2rLBqNnjGNsmxvfRjCaevO3bTREREREQ2cOo4Edl15totrDiYDEGwn0187ZGrDvdrMAo4m6a8jVcY0oD4d4FpTYB1HwBZybgqBGBq4aOIyf8Wk/SPQVe9nmL7U6m3JMcB3srTxnMK9Ph4xVHxeM2rXRXr/va4cxnGAeDNvlEAgJEdIp1uS0REREQVC0e0iciunl+aRp1/GtXGbiKvd5ckOdzvydQs2fLGqrN42m0F7lPvABKMpsKgRkDHF9FlgQ8KYAqYl43rLNs+r9AAT4t9qs9OGQCVxR7YHw9ugglLTff70p/7JecahChnSI+s6q14TskL3esiNjoEdYN8nW5LRERERBULR7SJyGEHL6U73eaJTrUAAG1qWm+hNWXlMbMjAd3V+/GH+2Ss0L6LwZrtcFMZYajVDRixGHghAWg5QgyyAaBpuPya67/2WO+LbRlkA0C94JKg93p2voNPBLg5uAe25fXrh/g5vH82EREREVVcDLSJyKaMnELx8197LiJfb3C47ZCWNdChTlUAwJ7zN7E08bLk/LWsfHigEA9pNiHe423M9vgMnTSHoRfUWGroiLj8yTgW+z+gfm9ApUKB3ujQdXedveFQPfOQd1jbkind/9e/oc12lonP9k7o7dD1iIiIiKhyYKBNRDaN+X23+PlqZj7G/33I4bafPtRMMvr7yoL9JSezr+OR3AXYon0ZX7jPQpT6EgzuPvhJPwBd87/GK4XjcFiojQdnbhebfLf+pPj5h5GtFa+rNwg4fe2W4nk5MzedFj/XqeZjs67loHRVX618RSIiIiKqlBhoE5FNu8/dlBz/ve+yQk0gOSNXcuyuUaPQYJFALfUYsOxlYFo0Ruf9DyGqdKQIVTClcBhyxh7EZP1IXEHJXtV5haZR7P0X0/HN+lNieai/8hZZyZl5SLyQLh4/2Cpctp6v2d7X5knZOtStqtg3ABTq7SeFIyIiIqLKi8nQiCohQRBgFABNKdYLt5ZZa10sZsp6uasBENBVfRBPaVYB3x8Uzxwy1sIv+gFYYeyA5a/0hF+AcgKyaWtPSI5rWSQkeyO2Ab5YY6pz4GI6BjarLp779MGmsn1GV9fJlus8lbOTA0B4FS+b54mIiIiocmOgTVQJDf9pJ06m3sKmN7vDR+vcj4G6QbanVUsU5iL8zF9Y4zELDdTFI+EqoGEc0OEFDPzhJopXSkeFKgfZALDpxDXJcYC3h+R4XM/6iD98FYcuZwCAZKsuyzXVxeQSpDlCrVZBpQIc2O2MiIiIiCohTh0nqmQy8wqRcOY60m7lY8qqozAYnYsW/9pzyW6dIKRjR/sEYFpjNNn3PhqoL+OW4Inf9H2Bl/YBj/4B1OoEaToyZTeyCyTHz3Wr61C9svTlw80BAG/1i7pj1yQiIiKiioEj2kT3gLxCA9w1aoemgh9PKdm7eu6OC0i6nImlYzs5db1pa0/g1T4NJGV6gxHRqnN4ym0VBqq3w+OAKTt5oW8NfHqzOxYYeiAL3ngisA4A4FRqSbKygc3DbF7vt21nJccDm1eXrRfkp8Xl9FzZc0pi6lRFwpnrTrUBgCGtwtGrYQj8vW1PMyciIiKiyocj2kTlSF6h41tnAcCPm0/jid92Ifr91XhkVoJDbRZa7DG9/2K6U9cEgOn/lWT/htEIHFuJK9N7YaX2HTyo2QIPlQG3glsBD8+G+6sH8bMhDlkoWVNdaDCi91ebxONO9pKPmSVUqxvkg8Zh8vtn+zo5DR4AZo5s5XSbYgyyiYiIiEgOA22icuL3hHNo+N5qLDtwxeE2n6w8hg3Hr8EomPaptjcNXG8wOjT12yH5t4CdPwLftQbmD0Nk5j7oBTWWGTpgcP5HOHHf30DjBwCNmyQZ2fnr2Xhr0UFJV4+0jbB5KfOl1MPaRSrWE2D9/Pamdluu9SYiIiIiul2cOk5UTrz/z2EAwIt/JtqdSq1k7/mbaFc7UPH852uOy5YXGoxwl0kYdkZmL+owpGGU2xpg2vNAninxmKDVYVZ2V/yujxW35qpbzVds0yqyJFP5oz/uQHJGnqRPe0nJzJOO2UreJpec7HmF9dxKPh7cxKn6RERERESWGGgT3UO8PTQ2z/+69axs+ZJ9lzHUYlT54o0c9PyyeHq3gPaqY3jcbTVi1XugUQlAHoDAOkD755EY2B9TfykZpd71Ti/JtGrzONoyyHaE+ZT6rg2CFOtZjujf16y605nFR3ao6dzNERERERFZ4NRxonJo+rqTdusknLZO4GW5BZYl87XO5pYdtJ6u/sG/h6FFAYZqNmCVx3gs0E5Cf81uaFQCthuicTXuN2DcXqD9Mxj9xxFJ22Cdp+S4tNtoFZu9/Zz42c1GwrdBLWpIjusH294yjIiIiIioLDDQJiqHpq07gV1nb9is89L8RKuyz+OP41a+Xra+YGPT5y0n06QF6RfR4fQ32KEdh8/cf0Ij9QXkCh6Yp++J2PxPMbxwAp7eGQyoTT9CsvJKrrn9/3pa9a+2EWg/262O4jk5ttahP2oxKr9o30WFmkREREREZYdTx4nKqSWJl2yuty40GGXLP1l5FJ880NSq/Lq9PaYFATi/Hdj5A3BsOZ7RmPq/JFTDHH0s/jJ0RwZK1l0fvJQh201YgJdVmdIgdP1gXzzX1bk11L6eyj+21BYXyiuU/zsiIiIiIipLDLSJyil7QWJ6TqFs+abj8tPH9QrTxrUowCDNNuCHT4CrSWL5NkNjzDHEYp2xNYxQY9KgxrhwIwc/bZGu8z6blm3zPgFABflIe+1r3azrquSTmgFA83B/6Dxtb6nVv0koViWlAFDux9LkB5rg3SVJ6FyvmmMNiIiIiIhsYKBNVE5dTs8tVTulke65O85LjsOQhpFu6zBMsx5VVLeAqwDcvIDmjwDtnsWIadKAunlEAC7elN6TIAj4Ir4kk3mDEF+UpaVjO9mt8/2IVqg9fiUAoE3NKnZqm4xoXxO9GoYgyE97W/dHRERERAQw0CYqt3advYH9F9PRIiLAqXYFMoH2y/MT8c/+KwAEtFMdw+Nu8YhV74GbqmR6eHjsy0DLkYB38XR1aaDt4abG/c3D8OPmM2JZwpnrksD+xZ71Ze8pwMf2KLSjHEmqZl7n/haOb5MW6u9pvxIRERERkQMYaBOVY4NnbMO5qXFOtbGcUm40Cli9/xwe1mzHE5p4RKtLRrbNp4cf7xAn7qW944x1RvOoED+rQPfSzVysOXJVPK4bJD+ibW+6tzkVAAdnfNsVGejtop6IiIiIiBzHQJuoHDh/3f4651K5cRYnV0zHDu1C0/RwALmCB5YYOiNuzER8vuIW9l9MBwDcytOjio8HAODRH3dIuvnt8bZikP1s1zqYVTSqnXRZmhCtUXXl7bTqBPngzLWS5yzLadoLn4vB5Zu5aFLDv8yuQURERESkhNt7EZUDY+ftc6r+9tPS7bh+e6Kt+FkNI3BiDfDHw8A3LRF1+jdUUd3CJaEaJhcOR4f87/COfgz8azVHraolI77FGbu3nZL2vfvd3ujRMFj2Pn5PkK77tjW1+5tHW0qOX+olP838dvfcBoC2tQIxuGUN+xWJiIiIiMoAR7SJyoHTqc6NaA//aaf4eeaIVoipUxUByMJQzUaM1KwD5pVkHt+haoGf8nthg7EljEW/W/P3Mk3lzi00iPX0RWutR/xc0jdgPfLspildINykhj+2/19PdJy6HgCgdePv+YiIiIjo3sRAm6gMHb6SgQ3HUvFst7ri+mc5cgnMlGTlSddgV8s8BM/lS7BTuxBalelcnsYPnu1GA22exKOfH1PsK19fct1NJ65hSKtwu9dX38aIs9we20RERERE9xoG2kRlZMXBZHFK+BdrTthMamYwOp7+K/FCOrQowP2a7RipWYfma03rpbUq4JCxFn43xOJScH/82bcnUjPzACgH2ubXnbziqFWgvfzFzlZtlMLsB1w0Vduy/wXPdMCivZcwvH2kS/onIiIiIiprDLSJyojlumtBEEq1/vhUahbqBRclGbtxBuf+/AA7tOvF5GbQeACNh+DZ4y0RnxEOQAVcMO13nZFbKNunIJgC7LpBvthy0rQm+3p2gVheTC6ZWLeoYHyz/pRV+ZgutZ1+NjmPtI3AHzsviMdBflp8/nBzl/RNRERERHQncJEkkYOen7sXURNW4ciVzFK1N5+mbYvlllTzdpwFTsQDcx+C8E0rjBL+FZObLaoyBnjtKDBkFqLb9oLleLDSNav7m6Zwvx7bQFL+39FUu/dXvL7bUuMw12T4fu++aMwY3solfRERERER3Q0MtIkcMOzHHViVlIJ8vREDvtlit36eWZKxYtn5etm6iRduSo4f61ATAFAFmXhWswyP7xkCzBsKnFoLFQRsNDTHUwWvo2v+11jl/yjgUw0AoJYZLP9q7QnJ8dyn2qN97UB8P9IUyPpZ7G895vc94ud6wfJ7Ygd4O74ndml4umvQO1o+yzkRERERUUXAqeNEDkg4c11ynFtggJeHRrH+qqRkq7IZG07j/YHRkrLMvEI89EOCWYmAajcTMTdwHtpmbxKTm8EzAGg5EjurDsbji1LNapdQy0Ta649JR6g716+GzvWrScq+HdYSL/6ZiNY1q2Dv+ZKgf+YI+VHlar5lt/81EREREdG9gCPaVGmdTcvGoUsZpWq7+eQ1m+dfXXDAquzXbWetysYvPgSDUYAfcjBKE4/VHv+HBxKfROecddCqCnHIWAvx9d4zTQ/vOxmPLJIGzk91LlkX3b9JqM17WvdaN9lyj6JttnILpKPw9UP8FPt6sWc9m9dyJVfsq01EREREdCdxRJsqrR5fbAQAJIzvKa5ZdtS2U2no29h2YGuXIOBS0hZ86rYeAzUJ8Fblm4rdvKBq8iB+yO6CqYd8gSQVzo30lu2iY92q4mfLZ7BMbKY0FdyjaNsxuenuSl7sWR/fmiVEi66uc7itI1Rma83d5ObEExERERGVYxzRpkrHaBQwf1dJVusv4k/YqC1PKWnY9Vv5GPXrLklZTJ2SYDg1Mw/IzwL2/ArM6op/tO/jEbeN8Fbl47gxHDe7fQzV68eAwTPwx+VgKG+mZWI+2ms5lf3arXyHnqV4RPtMWrZD9c3bFPv84WYOt3W0/6c618ajbSMQESj/SwYiIiIiovKKI9pU6by9+CAW7r0kHi/edwlfDlXePkpupLefwjTtz+OPY/MJ6bTyUH9PAEC06hyuzX8BwWnxQIFpa658wR0rjO0xT98Te4QonO0eBxQFz/mF0ozhliPUY3vUtbr+wYmxaDZxDQCg3eT/xPLqRfcgx11z+79vq1NNfrT8drx3X7T9SkRERERE5RADbap0zINsR8jtRa11kw9Or2bmSY49kY/uOWsw2mM+WqhPA1eKTlStB6H142j/bxDSUbIW2nyE2iiNqxF/OEVy/FqfKKvr6zzlM4J/aWMfasvRaQAY2SFSsT4REREREdnGQJuoFL7feBpv9WtoVZ6VZ9rCq77qEoZr/sNIr+1wP58FqIECQYN4Y1sMfPJdoFYXzN99Eek4ZOMqJZH2jjPX8cnKY5KzGifWLseYreW25K6x7udtmWcjIiIiIiLHMNAmsmPMnD32KwFAYR465/6HtzyWoZ36eFEZYPCPxOdpHbHQ0A3X4Y+BtbsCAMb/bSvIBgxmQ9pvLTqICzdyxOP7mlVXbFezqjfOX8+RlNnK3C0XsFvury3n84ea4c1FB9EnOsTmVmdERERERJUNA22q8JIzcqHzdIePtvT/Ob+6YD+mPdLCqlwQBBy6LL8FmNEomPauTj0K7PsfDIl/4JX8dEAN6AU11hlbo+mgV1GjVX/88M4qp+8px2y7LfMgGwC+Gmp9r8Usg+ymNfxtXkddyu2zHm4TgYfbRJSqLWD/voiIiIiIKioG2lShXbqZg86fboCfpxsOTexrt36+Xn4LqyWJl2UD7TcWHpSt74U8HFs1E9HJS4BLpizjGgCXhaqYr++BBYYeSEUVnG41ACjl9lT5emkytBoBXricngu1Sn5dtRKlXxQUs7y9lpEBDvddGlve6oGUzDw0cvGWYERERERE5QUDbarQdpy5AaBkbbQ9u87ecLjv/+04j8X7zBOnCfi5twZXN/6E+zXb4bc711Ss0uCATwy+vhGDTcbmMJrtmmdrHXWX+tWw5WSaeBxnYzo4AFxON13PMkna7dKopUH7NBuj5a4QEejNLbuIiIiI6J7GQJvKVKHBiBkbTqF7VDBaRAS4vP9Dl9LFzwajYDdB2JWiYNWenAI93luaBADQ4RYGabZjmGYDoreeF781mV4R0HV8EmgxAj8vv4INaVckfQT6eIifH24dbpXt/MBF072/1Ks+alfzRp9o6ZZho2NqYk7CeYfu93bUtAh6a1XzKfNrEhERERHdyxhoU5lJzcoT93L+et1JnJsa53DbtFv5UKtUkmBVjnkgWvedlfh4cBPUC/ZFhzryWbaLR8ABYGDzMCw7cEW2Xn6BAe1UR/Go2wYMUO+Ep6poiy+NFksLWmOBoQd25DXC2S4Di1pY97PrnV7i5wYhfpJz6TkFyCwaha/i7Y4HWoZbtX8nrpFsoO3sLyy+HdbS5nl1Kae2ExERERGRPAbaVGamrpJuR3UlPRdhAV522+UVGtDm43UAgNOfDHBqG6sJRaPQSkH9ksTL4udpQ5vj4KV0afKwW6nAgT/ht2cO/tKeFouPGSOwr9r9GD7mDUz4dDduFZqC5IzcQvh7yWfodtOUTMk2D2YvXM+B1t3+9HKtm3wmb7m15LbENbU9JZ2IiIiIiFzL8YxKRE66fqtAcvzDptMKNaXSbuWLn/MK5ZOXlUaSRVIwN40af4xpDzWM6KY+AOP8x4CvGgFr34fbzdPIFrT4U98Dg/M/Qr+CqUis/gjgHYhJgxuLfQyesQ2AKTu5LW5mwfSWU9dgHlpHO5kUrFZV2+ubE8b3xAMta4jHHLEmIiIiIrqzOKJNZcZy16jfE87jo0FN7LZLzykUPxcajDZqKruVr4evxXZfT/9usR92+kWE7Z+LLdqfUUOVBhQPwNdog92B9+Hx3RHIRskIfPG9d6xbTSw7m5YNAFh+MFnSdfwrXSXH5qPW7y5JQl5hyXO1qRXo1LPZ2hMbAKr7e2HyA01w+EoGOtWrZrMuERERERG5HgNtcpjRKOCb9SfRKrIKujYIsls/OT2vVNf5aPkR8fPWU2m4r1mY033cypMG2oIgIDkjD1oUoI96L94J2wd8vR1qCKihAtIFHxiaDkXVLk8DIY2xZMkhZOOC2P7QxFh4eZimcrtrpBNBjBZpwJe/2BlRodI12e4aaXA8yewZy4K3hxvWvNrN6XY9ouy/VyIiIiIiso2BdiWWrzcorgO2dCO7AN0/3yAm8HIksVnHelVx/GqWU/dUoDfiaHKmeDxuXqJioB39/mrFfiRTzgUBaSd24kO33zBIsx0BqmygeFetWl3w/sXWWJDdAtOiOmBAiGk985+7SoLsQS3C4OdZsg7bMmg+ez1bciy35tqZfa9tCfbTuqQfJYE+Zds/EREREVFlwDXaldSp1CxETViN9/9Jcqj+lJVHxSDbUW6lWBu84tAVh/fEzilQXr+dXaAHstOAhO+BHzoj6M++GO22FgGqbFwWqqKw0+vAS4nA48vxe3Y75MMDL/yxT2xvvuT6a4vkY5a/nLCc3h5lkWEcANzUpfuqzXqsteR45sjWCjVvzw8jW6N7VBDGD2hYJv0TEREREVUmHNGupL5dfwqAad30h/c3trvud9+Fm5LjmCn/IWF8L4XaJvl659ZXC4KAX7aetSrPyiuUjCjbooEBXdUHUXXFbCB5A2A0rffOF9wRb2yDvwzdsd3YGGf6DFTsw3IquOXfjeXotNHiMeWSjwV4O3b/lvo2lu6t3bpmlVL1Y0+/JqHo1yTUfkUiIiIiIrKLgXYlZT7C+t/RVPSODrFZ//Q16fTo5Iw8bD+dJkkMZkkuY7jBKChuZ7X73E0kXc6E1k0NX60brmebspbn642wHCO+eCNHcvxQzWzUu/wvHtBsQYgqHSjexSusJdBiBNr+rUMmfAFY7yvdIMQXJ67eAgCkZOSJCc4cZT6iXVMhI3jnetXQu1EI1h296lTfANCuViB2nbthvyIREREREZULnDp+jxAEAe//k4Sft5xxqH5qVkmisimrjpbqmsN/2mnz/F97LlmVnbuuHMT+WjSaPaRVOP5+oaNYvuzAFau6c3echy9y8IhmAxZ7fIAvrj6N59yWIUSVjjRBh+O1HwOe3w48sxFo97QYZANAN4uEX+b7aI/6dSey8grhjPTckvoTBzaWraNSqfDZQ82sytvXtp9xPDrMue2/iIiIiIjo7nIq0J45cyaaNWsGnU4HnU6HmJgYrFq1CgBw48YNvPjii4iKioKXlxciIyPx0ksvISNDunfxhQsXEBcXB29vbwQHB+PNN9+EXu/c2l+ylnD6On5POI+PVxyF3oEtsXReJVOZLUerXSE5I1e2fM1h+RHdizdysOZICgDgyU61ULOqj3juw2VmGbqNRuDsFjRIeAO7tS/gU/ef0Fp9EoJKAzToj5khHyIm/zvsjnoTCJEPenUW09DNp7ifuHoLBWZ/f99YjH7LGf3rLvFzj4bBivW8PawTz00Z0tRu/0REREREVLE4FWiHh4dj6tSp2Lt3L/bs2YOePXti0KBBOHz4MK5cuYIrV67giy++QFJSEmbPno3Vq1fjqaeeEtsbDAbExcWhoKAA27dvx5w5czB79my8//77Ln+wymb4zyWjy3/uvmizriAIOHol02ad2/Xk7D2y5d9vPGVVtnDPRXT5bAOMAtClfjXUl0kmhvQLwKbPgG9aAHPuw4OarfBSFeCksQY+KRyGq2MSgeHzkeTfFYVwc+iXDcX8vaSB97h5ieLn+5s7v7WYEq1M5vE6Qb4yNaU61q3qsnsgIiIiIqKy59Qa7YEDpQmkJk+ejJkzZ2LHjh146qmnsHjxYvFc3bp1MXnyZIwcORJ6vR5ubm5Ys2YNjhw5gnXr1iEkJAQtWrTApEmT8Pbbb2PixInw8PCQvW5+fj7y8/PF48zMsg0SK7rjKbb/fhbuvYQzFuuQBUFQTIh2o2ittDPMt+gyZ5lRXBAEvLnooHj8ZOfa4mcf5GKAZieGqLcCX5eMahs9/DA/py0WGrohUagHQIUxOtMa8+R000j6ikPJeLyTqS/z9dz9Glsn/Pr6kRZ4YvZuJ5/QeSqVCrHRIVhzxLl12n2iQ/Db423RqDqnkBMRERERVQSlXqNtMBgwf/58ZGdnIyYmRrZORkYGdDod3NxM8XxCQgKaNm2KkJCSxFt9+/ZFZmYmDh8+rHitKVOmwN/fX/wTERFR2tu+J6Vm5kmO9QZBoabJXzIj3qdSbynWf2vRgdLdmAPM10cDQLe6gcCp/4DFT2OP9nl87v4jYjRHAKiA2l2BB35EytMH8I5+DBKF+qZyAO5Fyd32XUgHYEqsllm01to8iH6qS21YahAqM4JuxxuxDZxuAwBBpdgHW6VSoUfDYIT6e5bqmkREREREdGc5HWgfOnQIvr6+0Gq1eO6557BkyRJER0db1UtLS8OkSZPwzDPPiGUpKSmSIBuAeJySkqJ4zfHjxyMjI0P8c/Gi7anRlU1xxuxiBXa21ZLbfqrPtM2K9S3XNLtS8XroeqpLmFtzBdTTmwBzhwCH/oKXqgCnjdXxWeEjwCuHgNHLgOaPYN2pLKt+NBrrZ3rn70P47+hVyS8RagZaZwXX2NnaTI7c36Ej3DUlX7lO9TglnIiIiIjoXuT09l5RUVHYv38/MjIysGjRIowePRqbNm2SBNuZmZmIi4tDdHQ0Jk6ceNs3qdVqodU6PxJYWWQXSKdj5+mtt9Uyt+us41tF5esN+DvxsuJ5yynnyw9ekaxxtik7Df4H5+Jfj9lopj4LFM+o9qoCNHkIg7ZG4IBQF4AKQwqqoF7R6W/+Oyl2oVGrMLhFDfhprf9TXn4wGcsPJksLZeJjy7+/Yv+93k3x1tUywXmgj/zSB3Pme3D/MaaD3fpERERERFTxOB1oe3h4oF49U8jTunVr7N69G9OnT8esWbMAAFlZWejXrx/8/PywZMkSuLuXjIaGhoZi165dkv6uXr0qnqPSeeGPfZLjlYeUZwcYjMrTygsNRsmIKwAcT7EePTb3977LeLB1uHgsF2R/NbQ5XvvLNP3cA4UQDi+F6uAC4OQahBj1CFEDhYIG7g37AS2GAfVjATctDmxZIfbx7pJDWPBsDG5mFyDtVsma8dOfDLB5f5aCfK1/YeOmMDpd10aiMrkm/4ztZPf67jIj70REREREdG+57X20jUajmKgsMzMTsbGx8PDwwL///gtPT+ma0piYGBw6dAipqali2dq1a6HT6WSnn5NjbAXPlnIURm8BU6BtXd/26Ph/x2wn9vpoUGMMaVkDb0RnYJLbr9ilfQGqhaOB4ysBox4HjHXwfuFotMufAQybBzQaCLhZB8PXbpn+G1tiY3QdAH4e1UbxnLtGJZvwTVOKaeByI9oRMtPSre+BW9cTEREREd3rnBrRHj9+PPr374/IyEhkZWVh3rx52LhxI+Lj48UgOycnB3PnzkVmZqaYHTwoKAgajQaxsbGIjo7GY489hs8++wwpKSmYMGECxo4dy6nhZvL1Brw4LxFrjlzF0rGd0CIiQLHu4SsZiufk5NoInE9cvWV1rVmbTkuO72tWHZtOXBOzh285kabYXxjSMKpwIfDdfIy7fkr8r03wqw5V80eBZo9i0FenFduby8w1JTb7aPkRm/V6NVLex7qKt/zU7hoBXg7dgzkviz2xqzuYqMxXZoo7ERERERHdW5z6V39qaipGjRqF5ORk+Pv7o1mzZoiPj0efPn2wceNG7Nxp2su5eGp5sbNnz6JWrVrQaDRYvnw5nn/+ecTExMDHxwejR4/GRx995LonKmdsbZslR28wImrCavF48IxtODc1TrF+3DdbZcvHzNmDn0dbj+7aGqH+X8J5q0B7w/Fr4udfH2+DmDrV0GHKf2JZVn7JCHnS5QzokI1+ml0YrN6GjpojwHrTOcHNG0vyW2GxoQumjXkRwf4+Ra0cC7S1bhr7lQCbf9efPtTM6TZKHmwVjneXJInHsdEhNmqXGNYuEssPJqNXQ+VfCBARERERUcXmVKD9yy+/KJ7r3r07BMH+FOaaNWti5cqVzly2wsotMKDf9M1oGRGArx9t6VCbo8m210Q7at1R+Sndf+66IDn+Z2wnDJqxDQBgGW/+Z9FHz4amYNJqprU+Hzi5Fjn/zsRu7U5oVYUl52p1AVoMhzHqPrz24RYAwAfLjmHmyNZIzcqDo9rXCXRqiryl3o2C0SPKdcGtp7s08A9QGC235KN1w1IH1nITEREREVHFxQWjZWj9sVScv56Dpfuv4O99lxz6RcSfuy/YraPk7X4NbZ4/npKFWZvPSMqam41gL9p7SXLuqTl7ZPtpXbMKVDCineooPnH7CfiiPrBgBNrlboVWVYjjxnDsqPOiaUuux5cDLYZD46UT269KMiVrM592/mpv2/tSLz+YjHPXsyVlG97obrONue+Gt3K4rqPMk6hp3flVIiIiIiIiEy4YLUPmI8Sv/XUAOk939LYzxXjeTutAW2n6+QKLoDy2cQg+XX1Mse9lB65IjkN1jq0rlrh6BF9VXYpM7XyEq4oC5TwAftUx62YrLDV0xlEhEudG3We3q5lm67/H9qhrdX7r2z3w2oID2HXuBgr0RvT6cpN4roq3O2pX87Fqo0Tr5lwgPPuJtnbrhOg8cTk9FwDgV4Z7jRMRERERUcXCYbgy9NXaE5Lj+MPK227ZMnOT/DrmtxcfkhzbCyYts2v3b+rYlmqhuI5nNMuAmZ2AmTHQ7fkO4ao0ZApeWKDvDoz6F3j1MKboR+CoUBOv94lyqN8rRUEqALjJZOMOr+KNqFA/2bbb/6+XQ9co5uw67O4OTDP/cVRrAKa/94fNtjgjIiIiIqLKjSPaZehU6i3J8cFLzmUIL/bZ6uN4obs0wVzSZeu+7I2q/rNfujXWa32sp2unZuUh2M8ThpybGKrZgMHqbeigPgq1SgCuAlC7w1ivD8Ym1cd6Y0vkwwNDa3dFRm7JuuzkTPm11/0ah2J10S8b8goNdrcOA5S3w/IsB1O1G4f520xUR0RERERElRMD7Tvo7PVs5BTo4e0h/9fuzIj3kO+3W5X5e0kDbaNRgLpoFFsQBJy7niOe2/Z/Pa0Ccw8UIum/eeiZvwE4vhqfuReI53Kqt4d360eB6MFQewei664LWPW3aUR91K+7sOVkyXrrPo3kp8e3jAwQA+2G762WrWNJb7Te2xsoXaZwWz5/qBnm7ryAAxfTMaMM1nMTEREREVHlwUD7DirQG7HlZBr6Nraesp2alYdn/7dXtp2fp/VrKjDIB6Bb3+6Bzp9uAACM+3Mfvh9hmt7839FUST1x72ijER+3SIc6aSHiNDvhv98UjGsAnDDWwFJDZ/i2eQQvPCCdqt2xblXxs3mQDQA6L/mRdWfXSQNAsN+d2V/94TYReLhNxB25FhERERER3dsYaN9h/x29KhtovzJ/v2KbrDy94rliCeN7AgC8zLadWnmoZIR8zO/mGcQF4NJeIGkxcPhvjMxKFv9LSBGqILTjCAzYGIYjQk0AKpx7wHo9dM2qyonIGofpZMt9Faa2W20XZubJzrXxxRrpWvfDH/ZVbiAjrll1p+oTERERERHdjru/0LWSOX0tW7Z8++nrt9Wv3mDaOszWlOoGqot4w20BNnm8CvzcE9gxA8hKhqDV4S99NwwveAcd879FVreJOCLUAmB7evZnDzaTLbfcY7qYu0a+v13v9la8htw0ex+t7d8PrXypi+T4o/sb26xPRERERETkShzRvsP2nr+JvEKDYjBarHXNKth7/qbsuVOpWVZlxVPBrfbqvnEGN3fNR7zHH4hSm+2T7e4NRPUHmjwIVb3eeGvCOvFU04lrxM91bGyhFS0zcu1mY3jaVyFAruZre3q4t4fGocRp5vdVN8hH/KVGVTv9ExERERERuRID7bvg4KUMtKsdKB4bjYJVnddjG2D4TzvF4wMX09E8IgAAMOCbrVb1i5OeGQQBIbiB+zQJGKhJAL45gyoAqqiBAkGDTcYW6D30Baii+gMe9veh/uGx1orn5Lbe6tdEecswR7bMkmOQ+fuxJ9jPU3H2ABERERERUVlioH2HvHdfNCYtPwIAOHc9WxJoT1x22Kp+h9pVJccL9lwUA+0CvTQR2nfDWwLZacCRfxCUtBgJ2u2m7bgAQKXGZn1jLDPGIN7QBi/FtYWqaR2H7zuiirfiObmtt97sq7yHtuU+3oDtEfNi+lIE2krJ4oiIiIiIiMoa12jfISG6kunLby06KDn3e8J58fOkwU1wbmqcOEJdbN7OC1Z9+iIHs1uexH0HXwS+aACseA2q89ugVgnYbWyA9wofB14/jlGF47HQ0B2Z8JVNxGaLl4ftKe6WbCVJk9M7Wn4rMHPmI9pd6ldzqF/LX0YQERERERHdKRzRLkPLxnXGwO+2Ynj7SNQN8nWoTd/GJYHn630a4Mu1pozbYf6eAIDD51MwQL0DAzUJ6KneD+3RwpLG1ZsDTR5Ex2UBuAJTQDrJVzpd26MUW2y5Urtagdh17oZ4LDdt3lJ0dR2OJGcCAH4a1cah6+TrHV/TTURERERE5EoMtMtQ03B/HPmoL7zcNYrZwA9eSpccV/UpGfnuHhWML9eegCfy8WTVk8DCuaidtALfe+SXNKjWAGjyENBkCFCtPgDgyrIV4ulrWWZ1YT/xmLn/69/Q4bqOqurrITnu39T+CPtLverjpfmJ+PTBpnaTyBXrVK8aTly9haA7tA83ERERERFRMQbaZUxueyoAYubx6etOSsrFdcyFufA7uxLfuM9BL/U++FzJB64A3irgojEIy40d8K+hI1aNfR6wsaVX28kl2cQ/e6iZ7DppJRob/ZaWm8W67tY1AxVqlujXJBSHP+wruyZcyRuxUahV1Qd9HJiaTkRERERE5EoMtO+SeTsv4MnOtfHfsVSxTIsC4Ogy4PBS4MRq1Cq4hVpFA7iXhGqo0XEYBm0MxkGhDgAVAn08bAbZliIDlROb1arqjXPXcyRlOi/n/vPY9U4vu3Uahvph2QHTZ6V9teU4E2QDpr22R3es5VQbIiIiIiIiV2CgfZecv27aekqLAnRTH8AAzU70dUsEFuSKdQp8amB2RgusMLTHAaEuvg5uiYPCfvH80hc6OXXN9rWVR4/Dq1gH2g+0DHeq/2Cdp906Y7rUxufxxwEA3w5r6VT/REREREREFQED7btAiwI0ubUNWPw19miXw09VFFwLAHThQOPBQOMHcMRYF598v11s98qC/ZJ+Iqsqj1DLUVonrsSZxGl+Wsf+U9K6layxdvZ+iIiIiIiIKgIG2ndKYR56q/ciTrMDvdX74HfSFFz7qYArQiBWGtpjzLOvAzVaA2pTgCtcuFmqS93XrDqWH0yWlNnbFivYImnY9yNaOXStLW/1wN7zNzGweZhzNwlAzUCbiIiIiIjuQQy0y5I+Hzi9Hji8BDi+Cj97ZIqnbmiCoG94P57bF4FEoR4EqDEmoq2kef0QP8Wupz/aQvGc3IZZgT4eMqUlejYKxt+Jl8XjAU2r26xfLCLQGxE21n7LqVPNB2fSstGulv1EaERERERERBUNA+2y9Gs/4Mq+kmO/MFyN7Ifn90XisLoBtIfdkCnoAQBD21ivh/bVuuHA+7Fo/tEaq3PdGgQpX1cm0m5aw9/mrcY1rY5xSLRZx1XiX+2KAr0RPg5ONyciIiIiIqpIGOmUpbo9gaxkIHqwad11eDuEqNXwztiJ/FNpyDfoxaq1qvnIduHv7Y7GYTocvlIyGu7joUGAt/IItSATafduZHubqzu5Xtpdo3Y6izgREREREVFFwWinLHV5HXj1CNB/KhDZQVx7/Xa/hlZVR3aoqdjNiPbSc9kFBpuXDZHJ/u3n6fjvVN7qF+VwXSIiIiIiIpJioF2WPLzF4Npc03B/q+RhOk93xW4ebF3Dqcu+0ruBVVlVX61MTaktb/XA4uc74oXu9Zy6HhEREREREZVgoH2XvBFrHQwrMd8SCwBe7lXfZn1/L+Wg3ZaIQG+0rlmlVG2JiIiIiIjIhIH2XVKzqvyabEcE6+yPThMREREREdHdwUD7Lpo4MBoAcHBirFPt6gX5OlV/VIzy+m8iIiIiIiJyLWYdv4se71Qbj3eq7XS79nWqOlX/nQGNnL4GERERERERlQ5HtCsBT3eN/UpERERERETkEgy0K4gxnU0j34+2jXCo/r/jOpXl7RAREREREZEClSAIwt2+CWdlZmbC398fGRkZ0Ol0d/t27gi9wYiDlzPQtIY/3DWO/X7EaBSgVqvK+M6IiIiIiIjufc7EoVyjXUG4adRoFenc1lsMsomIiIiIiO48Th0nIiIiIiIiciEG2kREREREREQuxECbiIiIiIiIyIUYaBMRERERERG5EANtIiIiIiIiIhdioE1ERERERETkQgy0iYiIiIiIiFyIgTYRERERERGRCzHQJiIiIiIiInIhBtpERERERERELsRAm4iIiIiIiMiFGGgTERERERERuRADbSIiIiIiIiIXYqBNRERERERE5EIMtImIiIiIiIhciIE2ERERERERkQsx0CYiIiIiIiJyIQbaRERERERERC7EQJuIiIiIiIjIhRhoExEREREREbkQA20iIiIiIiIiF2KgTURERERERORCDLSJiIiIiIiIXIiBNhEREREREZELud3tGygNQRAAAJmZmXf5ToiIiIiIiKgyKI4/i+NRWypkoJ2VlQUAiIiIuMt3QkRERERERJVJVlYW/P39bdZRCY6E4+WM0WjElStX4OfnB5VKdbdv567IzMxEREQELl68CJ1Od7dvh+4gvvvKi+++8uK7r7z47isvvvvKi+++/BIEAVlZWQgLC4NabXsVdoUc0Var1QgPD7/bt1Eu6HQ6fgErKb77yovvvvLiu6+8+O4rL777yovvvnyyN5JdjMnQiIiIiIiIiFyIgTYRERERERGRCzHQrqC0Wi0++OADaLXau30rdIfx3VdefPeVF9995cV3X3nx3VdefPf3hgqZDI2IiIiIiIiovOKINhEREREREZELMdAmIiIiIiIiciEG2kREREREREQuxECbiIiIiIiIyIUYaBMRERERERG5EAPtu2jz5s0YOHAgwsLCoFKpsHTpUsn5q1ev4vHHH0dYWBi8vb3Rr18/nDx50qqfhIQE9OzZEz4+PtDpdOjatStyc3PF8zdu3MCIESOg0+kQEBCAp556Crdu3SrrxyMbbvfdnzt3DiqVSvbPwoULxXoXLlxAXFwcvL29ERwcjDfffBN6vf5OPSbJcMX3PiUlBY899hhCQ0Ph4+ODVq1aYfHixZI6/N6XP65496dPn8YDDzyAoKAg6HQ6DB06FFevXpXU4bsvf6ZMmYK2bdvCz88PwcHBGDx4MI4fPy6pk5eXh7Fjx6Jq1arw9fXFgw8+aPVuHfmZvnHjRrRq1QparRb16tXD7Nmzy/rxyAZXvfuXXnoJrVu3hlarRYsWLWSvdfDgQXTp0gWenp6IiIjAZ599VlaPRQ5wxbs/cOAAhg0bhoiICHh5eaFRo0aYPn261bX4vS+fGGjfRdnZ2WjevDlmzJhhdU4QBAwePBhnzpzBP//8g8TERNSsWRO9e/dGdna2WC8hIQH9+vVDbGwsdu3ahd27d2PcuHFQq0te7YgRI3D48GGsXbsWy5cvx+bNm/HMM8/ckWckebf77iMiIpCcnCz58+GHH8LX1xf9+/cHABgMBsTFxaGgoADbt2/HnDlzMHv2bLz//vt39FlJyhXf+1GjRuH48eP4999/cejQIQwZMgRDhw5FYmKiWIff+/Lndt99dnY2YmNjoVKpsH79emzbtg0FBQUYOHAgjEaj2BffffmzadMmjB07Fjt27MDatWtRWFiI2NhYyff61VdfxbJly7Bw4UJs2rQJV65cwZAhQ8TzjvxMP3v2LOLi4tCjRw/s378fr7zyCsaMGYP4+Pg7+rxUwhXvvtiTTz6JRx55RPY6mZmZiI2NRc2aNbF37158/vnnmDhxIn788ccyezayzRXvfu/evQgODsbcuXNx+PBhvPvuuxg/fjy+++47sQ6/9+WYQOUCAGHJkiXi8fHjxwUAQlJSklhmMBiEoKAg4aeffhLL2rdvL0yYMEGx3yNHjggAhN27d4tlq1atElQqlXD58mXXPgSVSmnfvaUWLVoITz75pHi8cuVKQa1WCykpKWLZzJkzBZ1OJ+Tn57v2IahUSvvufXx8hN9//13SV2BgoFiH3/vyrzTvPj4+XlCr1UJGRoZYJz09XVCpVMLatWsFQeC7ryhSU1MFAMKmTZsEQTC9R3d3d2HhwoVinaNHjwoAhISEBEEQHPuZ/tZbbwmNGzeWXOuRRx4R+vbtW9aPRA4qzbs398EHHwjNmze3Kv/++++FKlWqSP7//vbbbwtRUVGufwgqldt998VeeOEFoUePHuIxv/flF0e0y6n8/HwAgKenp1imVquh1WqxdetWAEBqaip27tyJ4OBgdOzYESEhIejWrZt4HjCNeAcEBKBNmzZiWe/evaFWq7Fz58479DTkDEfevaW9e/di//79eOqpp8SyhIQENG3aFCEhIWJZ3759kZmZicOHD5fR3dPtcPTdd+zYEQsWLMCNGzdgNBoxf/585OXloXv37gD4va+IHHn3+fn5UKlU0Gq1Yh1PT0+o1WqxDt99xZCRkQEACAwMBGD6GV5YWIjevXuLdRo2bIjIyEgkJCQAcOxnekJCgqSP4jrFfdDdV5p374iEhAR07doVHh4eYlnfvn1x/Phx3Lx500V3T7fDVe8+IyND7APg9748Y6BdThV/0caPH4+bN2+ioKAAn376KS5duoTk5GQAwJkzZwAAEydOxNNPP43Vq1ejVatW6NWrl7iuLyUlBcHBwZK+3dzcEBgYiJSUlDv7UOQQR969pV9++QWNGjVCx44dxbKUlBTJP8gAiMd89+WTo+/+r7/+QmFhIapWrQqtVotnn30WS5YsQb169QDwe18ROfLuO3ToAB8fH7z99tvIyclBdnY23njjDRgMBrEO3335ZzQa8corr6BTp05o0qQJANN78/DwQEBAgKRuSEiI+N4c+ZmuVCczM1OSu4XujtK+e0fw//nlm6ve/fbt27FgwQLJciB+78svBtrllLu7O/7++2+cOHECgYGB8Pb2xoYNG9C/f39x/XXxmrxnn30WTzzxBFq2bIlp06YhKioKv/766928fboNjrx7c7m5uZg3b55kNJsqJkff/XvvvYf09HSsW7cOe/bswWuvvYahQ4fi0KFDd/Hu6XY48u6DgoKwcOFCLFu2DL6+vvD390d6ejpatWol+7OByqexY8ciKSkJ8+fPv9u3QncY333l5Yp3n5SUhEGDBuGDDz5AbGysC++Oyorb3b4BUta6dWvs378fGRkZKCgoQFBQENq3by9OCaxevToAIDo6WtKuUaNGuHDhAgAgNDQUqampkvN6vR43btxAaGjoHXgKKg17797cokWLkJOTg1GjRknKQ0NDsWvXLklZcSZLvvvyy967P336NL777jskJSWhcePGAIDmzZtjy5YtmDFjBn744Qd+7ysoR773sbGxOH36NNLS0uDm5oaAgACEhoaiTp06APgzv7wbN26cmKAuPDxcLA8NDUVBQQHS09Mlo1tXr14V35sjP9NDQ0OtslVfvXoVOp0OXl5eZfFI5KDbefeOUHr3xefo7nHFuz9y5Ah69eqFZ555BhMmTJCc4/e+/OKvwCsAf39/BAUF4eTJk9izZw8GDRoEAKhVqxbCwsKstgo4ceIEatasCQCIiYlBeno69u7dK55fv349jEYj2rdvf+cegkpF6d2b++WXX3D//fcjKChIUh4TE4NDhw5J/tG9du1a6HQ6q1/OUPmj9O5zcnIAwGoEU6PRiLNc+L2v2Bz53lerVg0BAQFYv349UlNTcf/99wPguy+vBEHAuHHjsGTJEqxfvx61a9eWnG/dujXc3d3x33//iWXHjx/HhQsXEBMTA8Cxn+kxMTGSPorrFPdBd54r3r0jYmJisHnzZhQWFopla9euRVRUFKpUqXL7D0JOc9W7P3z4MHr06IHRo0dj8uTJVtfh974cu8vJ2Cq1rKwsITExUUhMTBQACF999ZWQmJgonD9/XhAEQfjrr7+EDRs2CKdPnxaWLl0q1KxZUxgyZIikj2nTpgk6nU5YuHChcPLkSWHChAmCp6encOrUKbFOv379hJYtWwo7d+4Utm7dKtSvX18YNmzYHX1WknLFuxcEQTh58qSgUqmEVatWWZ3T6/VCkyZNhNjYWGH//v3C6tWrhaCgIGH8+PFl/nyk7HbffUFBgVCvXj2hS5cuws6dO4VTp04JX3zxhaBSqYQVK1aI9fi9L39c8b3/9ddfhYSEBOHUqVPC//73PyEwMFB47bXXJHX47suf559/XvD39xc2btwoJCcni39ycnLEOs8995wQGRkprF+/XtizZ48QExMjxMTEiOcd+Zl+5swZwdvbW3jzzTeFo0ePCjNmzBA0Go2wevXqO/q8VMIV714QTP+/T0xMFJ599lmhQYMG4s+S4izj6enpQkhIiPDYY48JSUlJwvz58wVvb29h1qxZd/R5qYQr3v2hQ4eEoKAgYeTIkZI+UlNTxTr83pdfDLTvog0bNggArP6MHj1aEARBmD59uhAeHi64u7sLkZGRwoQJE2S3ZZoyZYoQHh4ueHt7CzExMcKWLVsk569fvy4MGzZM8PX1FXQ6nfDEE08IWVlZd+IRSYGr3v348eOFiIgIwWAwyF7n3LlzQv/+/QUvLy+hWrVqwuuvvy4UFhaW5aORHa549ydOnBCGDBkiBAcHC97e3kKzZs2stvvi9778ccW7f/vtt4WQkBDB3d1dqF+/vvDll18KRqNRUofvvvyRe+8AhN9++02sk5ubK7zwwgtClSpVBG9vb+GBBx4QkpOTJf048jN9w4YNQosWLQQPDw+hTp06kmvQneeqd9+tWzfZfs6ePSvWOXDggNC5c2dBq9UKNWrUEKZOnXqHnpLkuOLdf/DBB7J91KxZU3Itfu/LJ5UgCIJrx8iJiIiIiIiIKi+u0SYiIiIiIiJyIQbaRERERERERC7EQJuIiIiIiIjIhRhoExEREREREbkQA20iIiIiIiIiF2KgTURERERERORCDLSJiIiIiIiIXIiBNhEREREREZELMdAmIiIiIiIiciEG2kREREREREQuxECbiIiIiIiIyIX+HwfywXZGC/jAAAAAAElFTkSuQmCC", "text/plain": [ "
" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "# SOLUTION\n", "\n", "import matplotlib\n", "from sklearn.linear_model import LinearRegression\n", "from sklearn.preprocessing import PolynomialFeatures\n", "\n", "poly = PolynomialFeatures(degree=3)\n", "co2 = co2_df['co2'].array\n", "date = co2_df['decimal_year'].array.reshape(-1,1)\n", "X = poly.fit_transform(date)\n", "model = LinearRegression()\n", "model.fit(X, co2)\n", "co2_predicted = model.predict(X)\n", "\n", "plt.figure(figsize=(12, 8))\n", "plt.plot(date, co2, label='CO2 concentration')\n", "plt.plot(date, co2_predicted, label='CO2 concentration')" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "You can see that the curve fits much better to the data.\n", "\n", "We can quantify the goodness of the fit by computing the root mean squared error (RMSE) between the predicted CO2 values and the actual values.\n", "\n", "We can do this easily using the `root_mean_squared_error` function from `sklearn.metrics`. Import this function and then call it passing the `co2` values and the `co2_predicted values`.\n", "\n", "Store the result in a variable called `rmse`. Print the value of `rmse`." ] }, { "cell_type": "code", "execution_count": 35, "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "RMSE: 2.26\n" ] } ], "source": [ "# SOLUTION\n", "\n", "from sklearn.metrics import root_mean_squared_error\n", "\n", "rmse = root_mean_squared_error(co2, co2_predicted)\n", "print(f'RMSE: {rmse:.2f}')" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "### Step 4.3 - Finding the best polynomial degree\n", "\n", "In the previous section, we used a polynomial of degree 3. We did this by passing the parameter `degree=3` to the `PolynomialFeatures` constructor. We can try different values of the degree parameter to see which gives the best fit.\n", "\n", "In the cell below, use a loop to try values of `degree` from 1 to 10. For each value of `degree`, compute the RMSE and print it. You should be able to implement this quickly by cutting and pasting the code from the previous cells and then adding a loop around it.\n" ] }, { "cell_type": "code", "execution_count": 37, "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "Degree: 1, RMSE: 4.91\n", "Degree: 2, RMSE: 2.28\n", "Degree: 3, RMSE: 2.26\n", "Degree: 4, RMSE: 2.26\n", "Degree: 5, RMSE: 2.26\n", "Degree: 6, RMSE: 2.26\n", "Degree: 7, RMSE: 2.26\n", "Degree: 8, RMSE: 2.26\n", "Degree: 9, RMSE: 2.26\n", "Degree: 10, RMSE: 2.26\n" ] } ], "source": [ "# SOLUTION\n", "\n", "from sklearn.linear_model import LinearRegression\n", "from sklearn.preprocessing import PolynomialFeatures\n", "\n", "co2 = co2_df['co2'].array\n", "date = co2_df['decimal_year'].array.reshape(-1,1)\n", "\n", "for degree in range(1, 11):\n", " poly = PolynomialFeatures(degree=degree)\n", " X = poly.fit_transform(date)\n", " model = LinearRegression()\n", " model.fit(X, co2)\n", " co2_predicted = model.predict(X)\n", "\n", " rmse = root_mean_squared_error(co2, co2_predicted)\n", " print(f'Degree: {degree}, RMSE: {rmse:.2f}')" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "You should find that the error drops a lot going from degree 1 to degree 2, but then only drops a little bit more going from degree 2 to degree 3. After that, the error does not drop much at all. This suggests that degree 2 or degree 3 (i.e., a 2nd or 3rd order polynomial) is the best choice." ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "### Step 4.4 - A fairer methodology for model evaluation\n", "\n", "In the above, we tested our model by comparing the predictions against the actual values, but we made predictions at the same dates as were used when fitting the model. This is not a fair test because the model has already seen these data points, and so it is not surprising that it fits them well. For our model to be useful, we need it to work well when predicting values for dates that it has not seen before. So to test how well we can do this, we want to fit the data using a subset of the dates (e.g., for dates up to 2000) and then test the model by predicting the values for the dates that were not used for fitting (e.g., from 2000 to 2019).\n", "\n", "The easiest way to do this is to split our `co2` and `date` data into two parts which we might call `co2_train` and `date_train` for the years up to 2000 and `co2_test` and `date_test` for the years from 2000 to 2019. We can then fit the model using the `co2_train` and `date_train` data and then test the model by predicting the CO2 concentration values for the dates in `date_test`. We can then compare these predicted values with the actual values stored in `co2_test`.\n", "\n", "We will first split the data. This can be done using the code below." ] }, { "cell_type": "code", "execution_count": 38, "metadata": {}, "outputs": [], "source": [ "co2 = co2_df['co2'].array\n", "date = co2_df['decimal_year'].array\n", "\n", "date_train = date[date < 2000].reshape(-1,1)\n", "date_test = date[date >= 2000].reshape(-1,1)\n", "co2_train = co2[date < 2000]\n", "co2_test = co2[date >= 2000]" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "Now rewrite the code from the previous section that used a loop to compute the RMSE value for polynomials of order N. But substitute `date` and `co2` with the `date_train`, `date_test` and `co2_train`, `co2_test` values as appropriate. Note that you will also need to make separate `X_train` and `X_test` matrices using the `date_train` and `date_test` values." ] }, { "cell_type": "code", "execution_count": 40, "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "Degree: 1, RMSE: 15.08\n", "Degree: 2, RMSE: 2.66\n", "Degree: 3, RMSE: 16.84\n", "Degree: 4, RMSE: 17.05\n", "Degree: 5, RMSE: 17.27\n", "Degree: 6, RMSE: 17.49\n", "Degree: 7, RMSE: 17.71\n", "Degree: 8, RMSE: 17.94\n", "Degree: 9, RMSE: 18.18\n", "Degree: 10, RMSE: 18.42\n" ] }, { "data": { "image/png": "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", "text/plain": [ "
" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "# SOLUTION\n", "\n", "import matplotlib\n", "from sklearn.linear_model import LinearRegression\n", "from sklearn.preprocessing import PolynomialFeatures\n", "\n", "\n", "plt.plot(date_test, co2_test, label=f'Degree {degree}')\n", "\n", "for degree in range(1, 11):\n", " poly = PolynomialFeatures(degree=degree)\n", " X_train = poly.fit_transform(date_train)\n", " X_test = poly.fit_transform(date_test)\n", " model = LinearRegression()\n", " model.fit(X_train, co2_train)\n", " co2_predicted = model.predict(X_test)\n", " plt.plot(date_test, co2_predicted, label=f'Degree {degree}')\n", " rmse = root_mean_squared_error(co2_test, co2_predicted)\n", " print(f'Degree: {degree}, RMSE: {rmse:.2f}')" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "Using this more rigorous methodology, you should find that the best polynomial order is 2 rather than 3. With order 3 the predictions become quite poor. This is a case of overfitting. The model with order 3 is too complex and fits local variability in the training data that does not reflect the underlying trend.\n" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "### Step 5 - Modeling the periodic component\n", "\n", "We will now improve our model by adding periodic terms. In the lecture notes, we saw examples of how to do this by summing sine and cosine terms. For this to work, we have to assume that we know the period of the periodic component. In this case, we know that the period is one year.\n", "\n", "Our model is going to combine the polynomial terms with the periodic terms. So, it will have the form,\n", "\n", "$y = a + p_1 x + p_2 x^2 + ... + p_n x^{N_\\alpha} + s_1 *sin(1 \\times 2 \\pi x) + c_1 *cos(1 \\times 2 \\pi x) + ... + s_n * sin (N_\\beta \\times 2 \\pi x) + c_n * cos (N_\\beta \\times 2 \\pi x)$\n", "\n", "where $x$ is the date in years and $y$ is the CO2 concentration. In the above equation, $N_\\alpha$ is the order of the polynomial and $N_\\beta$ is the number of periodic terms, and $a$, $p_1$, $p_2$, ..., $p_n$, $s_1$, $c_1$, ..., $s_n$, $c_n$ are the parameters that we need to estimate.\n", "\n", "To fit this model, we will need to store the functions $x$, $x^2$, ..., $x^N$, and $sin(1 \\times 2 \\pi x)$, $cos(1 \\times 2 \\pi x)$, ..., $sin(n \\times 2 \\pi x)$, $cos(n \\times 2 \\pi x)$ in the matrix $X$. We will then use the `LinearRegression` class to fit the model in exactly the same way as we did in the previous section.\n", "\n", "Making `X` with the polynomial terms is easy. We just use the `PolynomialFeatures` class as before. But there is no equivalent function for making the periodic terms -- we need to make those using the NumPy sine and cosine functions and append them to the `X` matrix. See the lecture notes for an example.\n", "\n", "### Step 5.1 - Making the X matrix\n", "\n", "In the cell below, write a function called `make_features` to make the `X` matrix for the model above. The function should take the $x$ values (i.e., the `decimal_year` values) and the order of the polynomial and the number of periodic terms as parameters. Call the parameters `x`, `order_polynomial` and `order_periodic` respectively. It should return the $X$ matrix. \n", "\n", "The function can use the `PolynomialFeatures` class to make the polynomial terms and then append the periodic terms to the matrix. Or it may be clearer to just make both the polynomial and periodic terms using NumPy functions.\n", "\n", "When you have written the function, run the test cell to check that it works." ] }, { "cell_type": "code", "execution_count": 41, "metadata": {}, "outputs": [], "source": [ "# SOLUTION\n", "\n", "def make_features(x, order_polynomial, order_periodic):\n", "\n", " X = np.ones((len(x), 1))\n", " for i in range(1, order_polynomial + 1):\n", " X = np.concatenate([X, x ** i], axis=1)\n", " for i in range(1, order_periodic + 1):\n", " X = np.concatenate([X, np.sin(2 * np.pi * i * x), np.cos(2 * np.pi * i * x)], axis=1)\n", " return X\n" ] }, { "cell_type": "code", "execution_count": 42, "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "All tests passed!\n" ] } ], "source": [ "# TEST \n", "x = np.linspace(0, 1, 10).reshape(-1, 1)\n", "assert make_features(x, 2, 0).shape == (10, 3)\n", "assert make_features(x, 0, 2).shape == (10, 5)\n", "assert make_features(x, 2, 2).shape == (10, 7)\n", "print('All tests passed!')" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "Now use your function to make the feature matrix `X_train` for a model with polynomial order 2 and periodic order 1, i.e., by passing the model the `date_train` values that you used earlier. \n", "\n", "Then make a `LinearRegression` model and call its `fit` method passing `X_train` and `co2_train`.\n", "\n", "Finally, call `model.predict` with the `X_train` values to get the predicted CO2 concentration values. Store the result in a variable called `co2_predicted` and plot the `co2_predicted` and `co2_train` values against the `date_train` values. You should see that the model fits the data well.\n" ] }, { "cell_type": "code", "execution_count": 43, "metadata": {}, "outputs": [ { "data": { "text/plain": [ "[]" ] }, "execution_count": 43, "metadata": {}, "output_type": "execute_result" }, { "data": { "image/png": "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", "text/plain": [ "
" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "# SOLUTION\n", "\n", "import matplotlib.pyplot as plt\n", "from sklearn.linear_model import LinearRegression\n", "\n", "co2 = co2_df['co2'].array\n", "date = co2_df['decimal_year'].array.reshape(-1,1)\n", "X_train = make_features(date_train, order_polynomial=3, order_periodic=3)\n", "model = LinearRegression()\n", "model.fit(X_train, co2_train)\n", "co2_predicted = model.predict(X_train)\n", "\n", "plt.figure(figsize=(12, 8))\n", "plt.plot(date_train, co2_train, label='CO2 concentration')\n", "plt.plot(date_train, co2_predicted, label='CO2 concentration')" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "You can play with the `periodic_order` to see how it affects the fit. In the next section, we will try to find the best value for the `periodic_order` using the same methodology that we used to find the best value for the `polynomial_order`." ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "### Step 5.2 - Finding the best value for the periodic order\n", "\n", "We will now use the same methodology that we used to find the best value for the polynomial order to find the best value for the periodic order. \n", "\n", "We will fix the polynomial order to the value of 2 but we try periodic orders from 0 to 10. In each case, we will fit the model using the `date_train` and `co2_train` values and then test the model by predicting the CO2 concentration values for the `date_test` values. We will compute the RMSE value for each value of periodic order and print it to see which works best.\n", "\n", "You should be able to do all of this using code from the previous sections. \n", "\n", "Implement your solution in the cell below.\n" ] }, { "cell_type": "code", "execution_count": 45, "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "Degree: 0, RMSE: 2.66\n", "Degree: 1, RMSE: 1.27\n", "Degree: 2, RMSE: 1.15\n", "Degree: 3, RMSE: 1.16\n", "Degree: 4, RMSE: 1.16\n", "Degree: 5, RMSE: 1.16\n", "Degree: 6, RMSE: 1.16\n", "Degree: 7, RMSE: 1.16\n", "Degree: 8, RMSE: 1.16\n", "Degree: 9, RMSE: 1.16\n", "Degree: 10, RMSE: 1.16\n" ] } ], "source": [ "# SOLUTION\n", "\n", "import matplotlib\n", "from sklearn.linear_model import LinearRegression\n", "from sklearn.preprocessing import PolynomialFeatures\n", "\n", "\n", "for degree in range(0, 11):\n", " X_train = make_features(date_train, order_polynomial=2, order_periodic=degree)\n", " X_test = make_features(date_test, order_polynomial=2, order_periodic=degree)\n", " model = LinearRegression()\n", " model.fit(X_train, co2_train)\n", " co2_predicted = model.predict(X_test)\n", " rmse = root_mean_squared_error(co2_test, co2_predicted)\n", " print(f'Degree: {degree}, RMSE: {rmse:.2f}')" ] }, { "cell_type": "markdown", "metadata": {}, "source": [] }, { "cell_type": "markdown", "metadata": {}, "source": [ "You should find that the RMSE error with the periodic order set to 0 (i.e. the purely polynomial model) is about 2.57. For the periodic order set to 1 (i.e. adding in just a simple sinusoidal variation), it falls to about 1.21. Increasing to order 2 (i.e., adding a second sinusoidal term), it falls a bit further to about 1.07. After that, the prediction starts to get worse again, i.e., the model starts overfitting the data. \n", "\n", "So, we will conclude that the best model to use is one with polynomial order 2 and periodic order 2." ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "## Step 6 - Using our model to predict the future\n", "\n", "We will now use our model to predict the date when the CO2 concentration will first reach 450 ppm. \n", "\n", "We will use the model of polynomial order 2 and periodic order 2 that we found was the best fit in the previous section. We will re-estimate the model parameters using all the data (i.e., the date and CO2 concentration values from 1958 to 2019). We will then use the model to predict CO2 concentration levels up to the year 2040 and look to see when this value first reaches 450 ppm.\n", "\n", "### Step 6.1 - Re-estimating the model parameters\n", "\n", "We will first re-estimate the model parameters using all the data. This is easy to do. We just need to make the X matrix using all the dates (i.e., passing `date` to our `make_features` function). We then construct a new `LinearRegression` model and finally, we call the `fit` method on the model.\n", "\n", "Implement this in the cell below and run the test function that will check that your model has the correct number of parameters (i.e., 7)" ] }, { "cell_type": "code", "execution_count": 46, "metadata": {}, "outputs": [ { "data": { "text/html": [ "
LinearRegression()
In a Jupyter environment, please rerun this cell to show the HTML representation or trust the notebook.
On GitHub, the HTML representation is unable to render, please try loading this page with nbviewer.org.
" ], "text/plain": [ "LinearRegression()" ] }, "execution_count": 46, "metadata": {}, "output_type": "execute_result" } ], "source": [ "# SOLUTION\n", "\n", "X = make_features(date, order_polynomial=2, order_periodic=2)\n", "model = LinearRegression()\n", "model.fit(X, co2)\n" ] }, { "cell_type": "code", "execution_count": 47, "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "All tests passed!\n" ] } ], "source": [ "# TEST\n", "\n", "assert model.coef_.shape == (7,), \"The model has the wrong number of coefficients\"\n", "print('All tests passed!')" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "### Step 6.2 - Predicting the future\n", "\n", "Now, make a new date axis with dates that run all the way from 2000 to 2040.\n", "\n", "You can do this using `np.linspace` to generate the date values. Choose a resolution of about 0.01 years, i.e., 100 values per year. Call this date axis `date_predict`.\n", "\n", "Then pass this date axis to your `make_features` function to make the corresponding `X` matrix. \n", "\n", "Then pass the `X` matrix to the `model.predict` method to get the predicted CO2 concentration values. Store the result in a variable called `co2_predicted`.\n", "\n", "Finally, plot the `co2_predicted` values against your date axis. \n" ] }, { "cell_type": "code", "execution_count": 48, "metadata": {}, "outputs": [ { "data": { "text/plain": [ "[]" ] }, "execution_count": 48, "metadata": {}, "output_type": "execute_result" }, { "data": { "image/png": "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", "text/plain": [ "
" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "# SOLUTION\n", "\n", "date_predict = np.linspace(2000, 2040, 400).reshape(-1, 1)\n", "X_predict = make_features(date_predict, order_polynomial=2, order_periodic=2)\n", "co2_predicted = model.predict(X_predict)\n", "plt.plot(date_predict, co2_predicted)" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "You can now find the earliest data at which the CO2 concentration is predicted to exceed 450 ppm using something like\n", "\n", "```\n", "np.min(date_predict[co2_predicted > 450])\n", "```\n", "\n", "alternatively, you can ask for the last date at which the CO2 concentration is predicted to fall below 450 ppm using something like\n", "\n", "```\n", "np.max(date_predict[co2_predicted < 450])\n", "```\n", "\n", "These dates are a little different due to the annual fluctuations in the CO2 concentration.\n", "\n", "You can compare your estimate with that made by other data scientists. There are many predictions that can be found with a bit of searching on the Web. For example, the following page has a prediction of 2037.\n", "\n", "https://insight.factset.com/atmospheric-carbon-levels-may-2023-update#:~:text=The%20resulting%2014.3%20tells%20us,global%20warming%20are%20reached%3A%202037.\n", "\n", "What we have to remember is that these are just predictions and that the future is uncertain. Our predictions have been made by modelling the past 60 years of data and extrapolating. This makes the big assumption that growth will continue on the same trajectory as it has in the past. Clearly, though, this is not necessarily the case. For example, as the dangers of CO2 emissions become more apparent, governments are taking action to reduce emissions. This will (hopefully!) change the trajectory of the CO2 concentration and so our model will no longer be valid and our predicted date will be incorrect. If emissions were cut drastically, it could even be the case that concentrations start to fall and that 450 ppm is never reached. " ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "## Conclusion\n", "\n", "In this lab, we have used linear regression to fit a model to the atmospheric CO2 concentration data. We have seen how to use the `PolynomialFeatures` class to make polynomial features and how to use the `LinearRegression` class to fit a model. We have also seen how to use the `mean_squared_error` function to compute the RMSE value for a model. We used the RMSE value to tune the hyperparameter (i.e., finding the polynomial order that produced the smallest error). We then saw how to use the `np.sin` and `np.cos` functions to make periodic features. We used these to make a model that had both polynomial and periodic terms that could fit the data much better. We then tuned the hyperparameters by training on a subset of the data and testing the model's ability to predict the future. Finally, we used our model to predict the date when the CO2 concentration will first reach 450 ppm.\n", "\n", "In this lab, we have used the basic least-means-squares approach to fitting a model. This approach works well, but it can have some drawbacks. When the number of parameters is large, the model can become unstable and overfit the data. There are more sophisticated approaches that can be used to `regularise` the model and prevent overfitting. For example, ridge regression and lasso regression. These are implemented in the `Ridge` and `Lasso` classes in `sklearn.linear_model`. You can read about these in the Scikit Learn documentation." ] } ], "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": "3.12.5" } }, "nbformat": 4, "nbformat_minor": 4 }