HMM.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {},
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "import pandas as pd\n",
    "import os\n",
    "import matplotlib.pyplot as plt\n",
    "import pickle\n",
    "from sklearn.metrics import accuracy_score, f1_score\n",
    "import time\n",
    "import warnings\n",
    "warnings.filterwarnings('ignore')\n",
    "%matplotlib inline"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Base line functions"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "def get_AR_predictions(training, n_periods):\n",
    "    temp_model = AR(training) \n",
    "    fit_model = temp_model.fit()\n",
    "    preds = fit_model.predict(len(training),len(training) + n_periods -1)\n",
    "    return np.array(preds)\n",
    "\n",
    "def get_MA_predictions(training, n_periods):\n",
    "    temp_model = ARMA(training, order=(0, 1))\n",
    "    fit_model = temp_model.fit(disp=False)\n",
    "    preds = fit_model.predict(len(training),len(training) + n_periods - 1)\n",
    "    return np.array(preds)\n",
    "\n",
    "def get_ARMA_predictions(training, n_periods):\n",
    "    temp_model = ARMA(training, order=(2, 1))\n",
    "    fit_model = temp_model.fit(disp=False)\n",
    "    preds = fit_model.predict(len(training),len(training) + n_periods - 1)\n",
    "    return np.array(preds)\n",
    "\n",
    "def get_ARIMA_predections(training, n_periods):\n",
    "    model = pm.auto_arima(training, start_p=1, start_q=1,\n",
    "                      test='adf',       # use adftest to find optimal 'd'\n",
    "                      max_p=3, max_q=3, # maximum p and q\n",
    "                      m=1,              # frequency of series\n",
    "                      d=None,           # let model determine 'd'\n",
    "                      seasonal=False,   # No Seasonality\n",
    "                      start_P=0, \n",
    "                      D=0, \n",
    "                      error_action='ignore',  \n",
    "                      suppress_warnings=True, \n",
    "                      stepwise=True,\n",
    "                      disp = False)\n",
    "    model.fit(training)\n",
    "    preds = model.predict(n_periods=n_periods)\n",
    "    return preds\n",
    "\n",
    "def train_test_split(stock_df, training_days, prediction_days):\n",
    "    training = stock_df.iloc[-(training_days + prediction_days):-prediction_days, :]['Close']\n",
    "    testing = stock_df.iloc[-prediction_days:, :]['Close']\n",
    "    return training, testing \n",
    "\n",
    "def get_complex_future_trend_prediction_baseline(stock_df, predictions):\n",
    "    future_trends ={x:'do somthing' for x in ['week_changes','month_changes','quarter_changes','year_changes']}\n",
    "    indiffrence_ranges = {x:[] for x in ['week_changes','month_changes','quarter_changes','year_changes']}\n",
    "    \n",
    "    first_price = np.array(stock_df['Open'])[-1]\n",
    "    final_vals = predictions\n",
    "    changes = get_frequant_changes(stock_df)\n",
    "    \n",
    "    for k, l in zip(future_trends.keys(), [5,25,75,250]):\n",
    "        temp_change = ((final_vals[l -1] - first_price) / first_price) * 100\n",
    "        if l in [5,25]:\n",
    "            if np.quantile(changes[k], 0.15) <= temp_change <= np.quantile(changes[k], 0.85):\n",
    "                future_trends[k] = 'HOLD'\n",
    "            elif temp_change < np.quantile(changes[k], 0.15):\n",
    "                future_trends[k] = 'HARD SELL'\n",
    "            else:\n",
    "                future_trends[k] = 'HARD BUY'\n",
    "        else: # quarter and year changes\n",
    "            if np.quantile(changes[k], 0.15) <= temp_change <= np.quantile(changes[k], 0.85): #holding range\n",
    "                if l == 75: #quarter\n",
    "                    if temp_change > 5:\n",
    "                        future_trends[k] = 'HARD BUY'\n",
    "                    else:\n",
    "                        future_trends[k] = 'HOLD'\n",
    "                else: #year\n",
    "                    if temp_change > 10:\n",
    "                        future_trends[k] = 'HARD BUY'\n",
    "                    else:\n",
    "                        future_trends[k] = 'HOLD'\n",
    "            elif temp_change < np.quantile(changes[k], 0.15):\n",
    "                future_trends[k] = 'HARD SELL'\n",
    "            else:\n",
    "                future_trends[k] = 'HARD BUY'\n",
    "    return future_trends"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## HMM functions"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "def enumerate_features(data):\n",
    "    \"\"\"\n",
    "    :param data: the data vector in np.array format\n",
    "    :return: a dictionary that enumerates each value into a corresponding number (for the probability tables)\n",
    "    \"\"\"\n",
    "    feature_enumerator = {}\n",
    "    reverse_enumerator = {}\n",
    "    counter = 0\n",
    "    for instance in data:\n",
    "        if instance not in feature_enumerator.keys():\n",
    "            feature_enumerator[instance] = counter\n",
    "            reverse_enumerator[counter] = instance\n",
    "            counter += 1\n",
    "    return feature_enumerator, reverse_enumerator\n",
    "\n",
    "\n",
    "def forward(mat_A, mat_B, vec_pi, data, feature_enumerator):\n",
    "    \"\"\"\n",
    "    :param mat_A: probabilities between hidden variables\n",
    "    :param mat_B: probabilities from hidden to known variables\n",
    "    :param vec_pi: starting probability vector\n",
    "    :param data: the data vector\n",
    "    :param feature_enumerator: enumeration dictionary of the data labels\n",
    "    :return: calculated alpha (the alpha parameter for each iteration) and normalization vector\n",
    "    \"\"\"\n",
    "    # initialize the alpha matrix\n",
    "    size = (np.shape(mat_A)[0], np.shape(data)[0])\n",
    "    mat_alpha = np.zeros(size)\n",
    "    # initialize norm vector\n",
    "    normalization_factor = np.ones((size[1]))\n",
    "    # run the forward algo\n",
    "    for i in range(size[1]):\n",
    "        normalization_factor[i] = 0\n",
    "        for j in range(size[0]):\n",
    "            if i > 0:\n",
    "                mat_alpha[j, i] = mat_B[j, feature_enumerator[data[i]]] * np.sum(mat_alpha[:, i-1] * mat_A[:, j])\n",
    "            else:\n",
    "                # initialization of the first row (dependant on the pi vector and not previous rows)\n",
    "                mat_alpha[:, 0] = vec_pi * mat_B[:, feature_enumerator[data[0]]]\n",
    "        normalization_factor[i] = 1.0/np.sum(mat_alpha[:, i])\n",
    "        mat_alpha[:, i] *= normalization_factor[i]\n",
    "\n",
    "    return mat_alpha, normalization_factor\n",
    "\n",
    "\n",
    "def backward(mat_A, mat_B, data, feature_enumerator, normalization_factor):\n",
    "    \"\"\"\n",
    "    :param mat_A: probabilities between hidden variables\n",
    "    :param mat_B: probabilities from hidden to known variables\n",
    "    :param data: the data vector\n",
    "    :param feature_enumerator: enumeration dictionary of the data labels\n",
    "    :param normalization_factor: the normalization factor from the forward function\n",
    "    :return: calculated beta (the beta parameter for each iteration)\n",
    "    \"\"\"\n",
    "    # initialization of the beta matrix\n",
    "    size = (np.shape(mat_B)[0], np.shape(data)[0])\n",
    "    mat_beta = np.zeros(size)\n",
    "    # run the backward algo\n",
    "    for i in range(size[1]-1, -1, -1):\n",
    "        if i < size[0] - 1:\n",
    "            for j in range(size[0]):\n",
    "                mat_beta[j, i] = np.sum(mat_B[:, feature_enumerator[data[i+1]]] * mat_beta[:, i+1] * mat_A[j, :])\n",
    "                # scale to the same normalization factor in the forward:\n",
    "                mat_beta[j, i] *= normalization_factor[i]\n",
    "        else:\n",
    "            # initialization of the last row\n",
    "            mat_beta[:, i] = normalization_factor[i]\n",
    "    return mat_beta\n",
    "\n",
    "\n",
    "def viterbi(mat_A, mat_B, data, vec_pi, feature_enumerator):\n",
    "    \"\"\"\n",
    "    :param mat_A: probabilities between hidden variables\n",
    "    :param mat_B: probabilities from hidden to known variables\n",
    "    :param vec_pi: starting probability vector\n",
    "    :param data: the data vector\n",
    "    :param feature_enumerator: enumeration dictionary of the data labels\n",
    "    :return:\n",
    "    \"\"\"\n",
    "    size = (np.shape(mat_B)[0], np.shape(data)[0])\n",
    "\n",
    "    path = np.zeros(size[1])\n",
    "    mat_delta = np.zeros(size)\n",
    "    mat_phi = np.zeros(size)\n",
    "\n",
    "    mat_delta[:, 0] = vec_pi * mat_B[:, feature_enumerator[data[0]]]\n",
    "\n",
    "    for i in range(size[1]):\n",
    "        for j in range(size[0]):\n",
    "            mat_delta[j, i] = np.max(mat_delta[:, i-1] * mat_A[:, j]) * mat_B[j, feature_enumerator[data[i]]]\n",
    "            mat_phi[j, i] = np.argmax(mat_delta[:, i-1] * mat_A[:, j])\n",
    "\n",
    "    path[size[1] - 1] = np.argmax(mat_delta[:, size[1]-1])\n",
    "    for i in range(size[1] - 2, -1, -1):\n",
    "        path[i] = mat_phi[int(path[i+1]), i+1]\n",
    "\n",
    "    return path, mat_delta, mat_phi\n",
    "\n",
    "\n",
    "def baum_welch (data, n, feature_enumerator, random = 1, tolerance = 1e-6, max_iter = 10000):\n",
    "    \"\"\"\n",
    "    :param data: the data vector\n",
    "    :param n: amount of hidden variables\n",
    "    :param feature_enumerator: enumeration dictionary of the data labels\n",
    "    \"\"\"\n",
    "\n",
    "    size = np.shape(data)[0]\n",
    "    num_of_values_in_data = int(max(feature_enumerator.values())) + 1\n",
    "    digamma = np.zeros((n, n, size))\n",
    "    np.random.seed(random)\n",
    "\n",
    "    # initialization using dirichlet\n",
    "    vec_pi = np.ndarray.flatten(np.random.dirichlet(np.ones(n), size=1))\n",
    "    mat_a = np.random.dirichlet(np.ones(n), size=n)\n",
    "    mat_b = np.random.dirichlet(np.ones(num_of_values_in_data), size=n)\n",
    "    \n",
    "    #initialize error based on the tolerance\n",
    "    error = tolerance + 10\n",
    "    itter = 0\n",
    "    while ((error > tolerance) & (itter < max_iter)):\n",
    "        former_a = mat_a.copy()\n",
    "        former_b = mat_b.copy()\n",
    "        alpha, norm = forward(mat_a, mat_b, vec_pi, data, feature_enumerator)\n",
    "        beta = backward(mat_a, mat_b, data, feature_enumerator, norm)\n",
    "        for i in range(size - 1):\n",
    "            for j in range(n):\n",
    "                for k in range(n):\n",
    "                    digamma[j, k, i] = alpha[j, i] * mat_a[j, k] * mat_b[k, feature_enumerator[data[i + 1]]] * beta[k, i+1]\n",
    "            digamma[:, :, i] /= np.sum(digamma[:, :, i])\n",
    "\n",
    "        for i in range(n):\n",
    "            for j in range(n):\n",
    "                digamma[i, j, size - 1] = alpha[i, size - 1] * mat_a[i, j]\n",
    "        digamma[:, :, size - 1] /= np.sum(digamma[:, :, size - 1])\n",
    "\n",
    "        for i in range(n):\n",
    "            vec_pi[i] = np.sum(digamma[i, :, 0])\n",
    "            for j in range(n):\n",
    "                mat_a[j, i] = np.sum(digamma[i, j, :size - 1]) / np.sum(digamma[i, :, :size - 1])\n",
    "\n",
    "            for j in range(num_of_values_in_data):\n",
    "                for k in range(size):\n",
    "                    if feature_enumerator[data[k]] == j:\n",
    "                        mat_b[i, j] += np.sum(digamma[i, :, k]) / np.sum(digamma[i, :, :])\n",
    "\n",
    "        error = (np.abs(mat_a - former_a)).max() + (np.abs(mat_b - former_b)).max()\n",
    "        itter += 1\n",
    "    for i in range(n):\n",
    "        mat_b[i, :] /= np.sum(mat_b[i, :])\n",
    "    mat_a = np.transpose(mat_a)\n",
    "    return mat_a, mat_b, vec_pi, alpha\n",
    "\n",
    "\n",
    "def choose_next(path, data, reverse_enumerator, enumerator, a, b, random = 1):\n",
    "    temp_data = data.copy()\n",
    "    temp_path = path.copy()\n",
    "    last_hidden = int(temp_path[np.size(data)-1])\n",
    "    if random == 1:\n",
    "        best_hidden = np.random.choice(len(a[last_hidden, :]), 1, p=a[last_hidden, :])[0]\n",
    "        new_path = np.append(temp_path, [best_hidden])\n",
    "        best_result = np.random.choice(len(enumerator.keys()), 1, p=b[best_hidden, :])[0]\n",
    "        new_data = np.append(temp_data, [reverse_enumerator[best_result]])\n",
    "    else:\n",
    "        best_hidden = [0, 0]\n",
    "        for j in range(len(a[last_hidden, :])):\n",
    "            if a[last_hidden, j] > best_hidden[1]:\n",
    "                best_hidden = [j, a[last_hidden, j]]\n",
    "        new_path = np.append(temp_path, [best_hidden[0]])\n",
    "        best_result = [0, 0]\n",
    "        for j in range(len(enumerator.keys())):\n",
    "            if b[best_hidden[0], j] > best_result[1]:\n",
    "                best_result = [j, b[best_hidden[0], j]]\n",
    "        new_data = np.append(temp_data, [reverse_enumerator[best_result[0]]])\n",
    "    return new_path, new_data\n",
    "\n",
    "\n",
    "def predict(n, k, data, enumerator, reverse_enumerator, prediction = \"simple\", random = 1, seed = 0):\n",
    "    np.random.seed(seed)\n",
    "#     print('baum welcing...')\n",
    "    a, b, pi, alpha = baum_welch(data, k, enumerator)\n",
    "#     print('viterbing...')\n",
    "    path, delta, phi = viterbi(a, b, data, pi, enumerator)\n",
    "    new_data = data.copy()\n",
    "    new_path = path.copy()\n",
    "    last_index = np.size(data)-1\n",
    "    if prediction == \"simple\":\n",
    "        for i in range(n):\n",
    "            new_path, new_data = choose_next(new_path, new_data, reverse_enumerator, enumerator, a, b, random=random)\n",
    "        return new_data\n",
    "    else:\n",
    "        for i in range(n):\n",
    "            new_path, new_data = choose_next(new_path, new_data, reverse_enumerator, enumerator, a, b, random=random)\n",
    "            new_path, delta, phi = viterbi(a, b, new_data, pi, enumerator)\n",
    "        return new_data"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Data functions"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [],
   "source": [
    "def get_ETF_data(data_length):\n",
    "    \"\"\"\n",
    "    :return: DataFrame dictionary containing data of ETF's with 10+ years worth of data key:ETF name, val: dataframe\n",
    "    \"\"\"\n",
    "    df_dict = {}\n",
    "    directory = os.fsencode('ETFs')\n",
    "    counter = 0\n",
    "    for file in os.listdir(directory):\n",
    "        if counter > data_length:\n",
    "            break\n",
    "        filename = os.fsdecode(file)\n",
    "        temp_df = pd.read_csv('ETFs/'+filename)\n",
    "        if len(temp_df) > 3000 and 0 not in temp_df['Open'].values:\n",
    "            del temp_df['OpenInt'] #useless variable\n",
    "            temp_df = symbolise_df(temp_df) # add column of matching symbol\n",
    "            temp_df['Year'] = [int(x[:4])  for x in temp_df['Date']]\n",
    "            temp_df['Two_Symbols'] = [x[0]+x[-1] for x in temp_df['Symbol']]\n",
    "            temp_df = temp_df[temp_df['Year'] >= 2010]\n",
    "            df_dict[filename[0:-7]] = temp_df\n",
    "            counter+= 1\n",
    "    return df_dict\n",
    "\n",
    "def get_stock_data(relevant_stocks):\n",
    "    \"\"\"\n",
    "    :return: DataFrame dictionary containing data of stocks with 10+ years worth of data key:ETF name, val: dataframe\n",
    "    \"\"\"\n",
    "    df_dict = {}\n",
    "    directory = os.fsencode('Stocks')\n",
    "    for file in os.listdir(directory):\n",
    "        filename = os.fsdecode(file)\n",
    "        temp_name = str(filename[0:-7])\n",
    "        if temp_name in relevant_stocks:\n",
    "            temp_df = pd.read_csv('Stocks/'+filename)\n",
    "            if len(temp_df) > 1750:\n",
    "                del temp_df['OpenInt'] #useless variable\n",
    "                temp_df = symbolise_df(temp_df) # add column of matching symbol\n",
    "                temp_df['Year'] = [int(x[:4])  for x in temp_df['Date']]\n",
    "                temp_df['Two_Symbols'] = [x[0]+x[-1] for x in temp_df['Symbol']]\n",
    "                temp_df = temp_df[temp_df['Year'] >= 2010]\n",
    "                df_dict[filename[0:-7]] = temp_df\n",
    "    return df_dict\n",
    "\n",
    "def get_new_fracs_and_bins(stock_df, num_bins):\n",
    "    \"\"\"\n",
    "    :param stock_df:  dataframe containg stock trading data\n",
    "    :param num_bins:  number of symbols to use\n",
    "    :return: column stack array of \"price\" ,\"high- low\" and \"open vs close \"symbols with matching bins\n",
    "    \"\"\"\n",
    "    open_p = np.array(stock_df['Open'])\n",
    "    close_p = np.array(stock_df['Close'])\n",
    "    high_p = np.array(stock_df['High'])\n",
    "    low_p = np.array(stock_df['Low'])\n",
    "\n",
    "    frac_price = (close_p - open_p) / open_p\n",
    "    frac_high_low = (high_p - low_p) / open_p\n",
    "    frac_open_close = np.array([0] + [(open_p[i] - close_p[i - 1]) / close_p[i - 1] for i in range(1, len(open_p))])\n",
    "\n",
    "    labels = ['L', 'A', 'M', 'H'] if num_bins == 4 else ['L', 'M', 'H']\n",
    "    price_symbols, price_bins = pd.qcut(frac_price, num_bins, labels=labels, retbins=True)\n",
    "    high_low_symbols = pd.qcut(frac_high_low, num_bins, labels=labels)\n",
    "    open_close_symbols, open_close_bins = pd.qcut(frac_open_close, num_bins, labels=labels, retbins=True)\n",
    "\n",
    "    return np.column_stack((price_symbols, high_low_symbols, open_close_symbols)), price_bins, open_close_bins\n",
    "\n",
    "\n",
    "def get_symbols(bins_data):\n",
    "    \"\"\"\n",
    "    :param bind_data: column stack array of \"price\" ,\"high- low\" and \"open vs close \"symbols\n",
    "    :return: string containing matching symbols\n",
    "    \"\"\"\n",
    "    string_list = []\n",
    "    for row in bins_data:\n",
    "        temp_string = ''\n",
    "        for col in row:\n",
    "            temp_string += str(col)\n",
    "        string_list.append(temp_string)\n",
    "    return string_list\n",
    "\n",
    "\n",
    "def symbolise_df(stock_df):\n",
    "    \"\"\"\n",
    "    :param stock_df: stock dataframe\n",
    "    :return: adds column of \"Symbol\" to the dataframe\n",
    "    \"\"\"\n",
    "    bins_data, price_bins, open_close_bins = get_new_fracs_and_bins(stock_df, 3)\n",
    "    stock_df['Symbol'] = get_symbols(bins_data)\n",
    "    return stock_df\n",
    "\n",
    "\n",
    "def get_training_symbols(stock_df, num_learning, num_symbols):\n",
    "    \"\"\"\n",
    "    :param df: stock dataframe\n",
    "    :param frac_learning: number of data rows used for learning\n",
    "    :return: training symbols\n",
    "    \"\"\"\n",
    "    if num_symbols == 3:\n",
    "        return stock_df.iloc[-(num_learning+250):-250, :]['Symbol'].values #data for learning\n",
    "    return stock_df.iloc[-(num_learning+250):-250, :]['Two_Symbols'].values #data for learning\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Reverse ploting functions"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [],
   "source": [
    "def get_tricky_fracs(stock_df, labels, frac_type):\n",
    "    open_p = np.array(stock_df['Open'])\n",
    "    close_p = np.array(stock_df['Close'])\n",
    "    \n",
    "    frac_price = (close_p - open_p) / open_p\n",
    "    frac_open_close = np.array([0]+[(open_p[i] - close_p[i-1]) / close_p[i-1] for i in range(1,len(open_p))])\n",
    "\n",
    "    price_symbols = pd.qcut(frac_price, 3, labels = labels)\n",
    "    open_close_symbols = pd.qcut(frac_open_close, 3, labels = labels)\n",
    "\n",
    "    if frac_type == 'price':\n",
    "         return  pd.DataFrame(np.column_stack((frac_price, price_symbols)),columns=['frac','symbol'])\n",
    "    if frac_type == 'open_close':\n",
    "        return  pd.DataFrame(np.column_stack((frac_open_close, open_close_symbols)),columns=['frac','symbol'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [],
   "source": [
    "def get_gaussian_symbol_dic(stock_df, labels, frac_type):\n",
    "    symbols_dic = {}\n",
    "    temp_df = get_tricky_fracs(stock_df, labels, frac_type)\n",
    "    for l in labels:\n",
    "        temp_bin = temp_df[temp_df['symbol'] == l]['frac'].values\n",
    "        symbols_dic[l] = [np.mean(temp_bin), np.std(temp_bin)]\n",
    "    return symbols_dic"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [],
   "source": [
    "def get_series_from_gaussian_symbols(stock_df, predicted):\n",
    "    translated_close_price = []\n",
    "    labels = ['L','M','H']\n",
    "    price_symbol_dic = get_gaussian_symbol_dic(stock_df, labels, 'price')\n",
    "    open_close_symbol_dic = get_gaussian_symbol_dic(stock_df, labels, 'open_close')\n",
    "    \n",
    "    current_val = np.array(stock_df['Open'])[-1] #start predicting from the last known open price\n",
    "    for p in predicted:\n",
    "        p_s, oc_s = p[0], p[-1] #get symbols from prediction\n",
    "        #p_s, oc_s = p[0], p[1] # to work with upgraded method\n",
    "        temp_price_change = np.random.normal(price_symbol_dic[p_s][0], price_symbol_dic[p_s][1]) #sample from relevant distribution\n",
    "        temp_openclose_change = np.random.normal(open_close_symbol_dic[oc_s][0], open_close_symbol_dic[oc_s][1]) #sample from relevant distribution\n",
    "        \n",
    "        first_temp_translation = current_val * (1 + temp_openclose_change)\n",
    "        final_tmep_val = first_temp_translation * (1 + temp_price_change)\n",
    "        translated_close_price.append(final_tmep_val)\n",
    "        current_val = final_tmep_val\n",
    "    return translated_close_price "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [],
   "source": [
    "def translate(stock_df, predictions, num_tryes):\n",
    "    reverse_translations = {}\n",
    "    for i in range(num_tryes):\n",
    "        reverse_translations[i] = np.array(get_series_from_gaussian_symbols(stock_df, predictions))\n",
    "    return sum(reverse_translations.values()) /  num_tryes   "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [],
   "source": [
    "def get_MAPE_score(true_vals, predicted_vals):\n",
    "    return np.mean(np.abs((true_vals - predicted_vals) / true_vals)) * 100"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [],
   "source": [
    "def plot_graphs_and_predictions(k,training_days,predictions_dict, etf_dict):\n",
    "    temp_df = etf_dict[k[0]].iloc[-(training_days + 250):-250, :]\n",
    "    true_vals = etf_dict[k[0]].iloc[-250:,:]\n",
    "    final_vals = translate(temp_df, predictions_dict[k], 100)\n",
    "    translated = pd.DataFrame(final_vals,index = true_vals['Close'].index,columns=['Prediction'])\n",
    "    plt.figure(figsize=(16,8))\n",
    "    plt.plot(temp_df['Close'], label='Training Period', color = 'b')\n",
    "    plt.legend(loc='upper left')\n",
    "    plt.plot(true_vals['Close'], label='True Vlues', color = 'g')\n",
    "    plt.legend(loc='upper left')\n",
    "    plt.plot(translated['Prediction'], label = 'Predicted Values', color = 'black')\n",
    "    plt.legend(loc='upper left')\n",
    "    plt.title(k)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [],
   "source": [
    "def plot_graphs_and_true_vals(etf, etf_dict, training_days, prediction_days):\n",
    "    temp_df = etf_dict[etf].iloc[-(training_days + prediction_days):-prediction_days, :]\n",
    "    true_vals = etf_dict[etf].iloc[-prediction_days:,:]\n",
    "    plt.figure(figsize=(16,8))\n",
    "    plt.plot(temp_df['Close'], label='Close Price history', color = 'b')\n",
    "    plt.plot(true_vals['Close'], label='Close Price history', color = 'g')\n",
    "    plt.title(etf)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Base line Comparison #"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We begin with some basic Time Series models including Auto Regression, Moving Average, ARMA, and ARIMA to predict 10 selected ETFs one year into the future and get some perspective about their performance.\n",
    "\n",
    "We do so by comparing the overall MAPE (Mean Absolute Percentage Erroe) scores of the various ETFs achieved by the different methods:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [],
   "source": [
    "etf_dict = get_ETF_data(200)\n",
    "etfs = ['xlk','jkl', 'xle', 'kce', 'kie','ewz','gld','spy','xlf','fvl']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "from statsmodels.tsa.ar_model import AR\n",
    "from statsmodels.tsa.arima_model import ARMA\n",
    "from statsmodels.tsa.arima_model import ARIMA\n",
    "from statsmodels.tsa.holtwinters import ExponentialSmoothing\n",
    "import pmdarima as pm\n",
    "\n",
    "mape_scores = {}\n",
    "baseline_preds = {}\n",
    "prediction_days = 250\n",
    "for etf in etfs:\n",
    "    for l in [1000]:\n",
    "        temp_df = etf_dict[etf]\n",
    "        temp_train, temp_test = train_test_split(temp_df, l,prediction_days)\n",
    "\n",
    "        AR_preds = get_AR_predictions(temp_train, prediction_days)\n",
    "        baseline_preds[(etf,'AR',l)] = AR_preds\n",
    "        mape_scores[(etf,'AR',l)] = get_MAPE_score(temp_test,AR_preds)\n",
    "        \n",
    "        MA_preds = get_MA_predictions(temp_train, prediction_days)\n",
    "        baseline_preds[(etf,'MA',l)] = MA_preds\n",
    "        mape_scores[(etf,'MA',l)] = get_MAPE_score(temp_test,MA_preds)\n",
    "    \n",
    "        ARMA_preds = get_ARMA_predictions(temp_train, prediction_days)\n",
    "        baseline_preds[(etf,'ARMA',l)] = ARMA_preds\n",
    "        mape_scores[(etf,'ARMA',l)] = get_MAPE_score(temp_test, ARMA_preds) \n",
    "\n",
    "        ARIMA_preds = get_ARIMA_predections(temp_train, prediction_days)\n",
    "        baseline_preds[(etf,'ARIMA',l)] = ARIMA_preds\n",
    "        mape_scores[(etf,'ARIMA',l)] = get_MAPE_score(temp_test, ARIMA_preds)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Method: AR ,Total MAPE score:  102.32480308263949\n",
      "Method: MA ,Total MAPE score:  182.5093396171745\n",
      "Method: ARMA ,Total MAPE score:  97.21843030197299\n",
      "Method: ARIMA ,Total MAPE score:  76.45269914058743\n"
     ]
    }
   ],
   "source": [
    "total_base_scores = {y:0 for y in ['AR','MA','ARMA','ARIMA']}\n",
    "\n",
    "for k,val in mape_scores.items():\n",
    "    total_base_scores[k[1]] += val\n",
    "\n",
    "for k,val in total_base_scores.items():\n",
    "    print('Method:',k,',Total MAPE score: ', val)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Basic Part"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Best Model configuration comparison\n",
    "In these sections, we will try to find the optimal parameters for our model by comparing the total prediction error over 10 ETF's. We do so by comparing the overall MAPE (Mean Absolute Percentage Erroe) scores of the various ETFs achieved by the different models."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Feature engineering"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "final_scores = {}\n",
    "predictions = {}\n",
    "hidden_states = 2\n",
    "for etf in etfs:\n",
    "    for num_symbols in [2,3]:\n",
    "        for training_days in [500,1000,1500]:\n",
    "            #training\n",
    "            start = time.time()\n",
    "            training =  get_training_symbols(etf_dict[etf], training_days, num_symbols)\n",
    "            \n",
    "            temp_enu, temp_reverse_enu = enumerate_features(training)\n",
    "            predicted_data = predict(250, hidden_states, training, temp_enu, temp_reverse_enu) #predict 250 days into the future\n",
    "            predicted = predicted_data[-250:]\n",
    "            end = time.time()\n",
    "            predictions[(etf,num_symbols,training_days)] = predicted\n",
    "            print(num_symbols, 'symbols', training_days,' training days, etf name: ',etf,' training time is: ', end - start)\n",
    "\n",
    "            #prediction and translation       \n",
    "            temp_df = etf_dict[etf].iloc[-(training_days + 250):-250, :]\n",
    "            temp_true_vals = etf_dict[etf].iloc[-250:,:]['Close'].values\n",
    "            final_vals = translate(temp_df, predicted, 100) #number of bootstrap iterations = 100\n",
    "\n",
    "            file_name = open(str(etf) +'_'+ str(num_symbols) +'_'+ str(training_days) +'_days', 'wb')\n",
    "            pickle.dump(predicted, file_name)\n",
    "            #score calculation\n",
    "            temp_score = get_MAPE_score(temp_true_vals, final_vals)\n",
    "            print(num_symbols, 'symbols', training_days,' training days, etf name: ',etf,' score is: ',temp_score)\n",
    "            final_scores[(etf,num_symbols,training_days)] = temp_score\n",
    "        "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "total_symbols_scores = {x:{y:0 for y in [500,1000,1500]} for x in [2,3]}\n",
    "for k,val in final_scores.items():\n",
    "        total_symbols_scores[k[1]][k[2]] += val"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "2 symbols: {500: 142.66098269459795, 1000: 105.59917547451816, 1500: 82.41791321567196}\n",
      "3 symbols: {500: 173.90134249826562, 1000: 102.90933213240382, 1500: 93.3561726123998}\n"
     ]
    }
   ],
   "source": [
    "file_name = open('total_symbols_scores', 'rb')\n",
    "scoress = pickle.load(file_name)\n",
    "for k in scoress.keys():\n",
    "    print(k, \"symbols:\", scoress[k])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## How many hidden states? \n",
    "- over 1000 training days and 3 symbols"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "final_hd_scores = {}\n",
    "hd_predictions = {}\n",
    "training_days = 1000\n",
    "for etf in etfs:\n",
    "    for hidden_state in [2,3,4]:\n",
    "        #training\n",
    "        start = time.time()\n",
    "        training =  get_training_symbols(etf_dict[etf], 1000, 3)\n",
    "\n",
    "        temp_enu, temp_reverse_enu = enumerate_features(training)\n",
    "        predicted_data = predict(250, hidden_state, training, temp_enu, temp_reverse_enu) #predict 250 days into the future\n",
    "        predicted = predicted_data[-250:]\n",
    "        end = time.time()\n",
    "        hd_predictions[(etf,hidden_state)] = predicted\n",
    "        print(hidden_state,'states, etf name: ',etf,' training time is: ', end - start)\n",
    "\n",
    "        #prediction and translation       \n",
    "        temp_df = etf_dict[etf].iloc[-(1000 + 250):-250, :]\n",
    "        temp_true_vals = etf_dict[etf].iloc[-250:,:]['Close'].values\n",
    "        final_vals = translate(temp_df, predicted, 100) #number of bootstrap iterations = 100\n",
    "\n",
    "        file_name = open(str(etf) +'_'+ str(hidden_state), 'wb')\n",
    "        pickle.dump(predicted, file_name)\n",
    "        #score calculation\n",
    "        temp_score = get_MAPE_score(temp_true_vals, final_vals)\n",
    "        print(hidden_state, 'hidden states, etf name: ',etf,' score is: ',temp_score)\n",
    "        final_hd_scores[(etf,hidden_state)] = temp_score"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 134,
   "metadata": {},
   "outputs": [],
   "source": [
    "hd_scores = {x:0 for x in [2,3,4]}\n",
    "for k,val in final_hd_scores.items():\n",
    "        hd_scores[k[1]] += val"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "2 hidden statas: 102.90933213240382\n",
      "3 hidden statas: 118.23450056773022\n",
      "4 hidden statas: 106.10578299226488\n"
     ]
    }
   ],
   "source": [
    "file_name = open('total_hd_scores', 'rb')\n",
    "scoress = pickle.load(file_name)\n",
    "for k in scoress.keys():\n",
    "    print(k, \"hidden statas:\", scoress[k])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## how many bins per feature? (results taken from colab run)\n",
    "- over 1000 training days and 4 hidden states using 2 features"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [],
   "source": [
    "four_bins_scores = {('xlk',500):15.5,\n",
    "                   ('xlk',1000):5.09,\n",
    "                   ('xlk',1500):25.77,\n",
    "                   ('jkl',500):13,\n",
    "                   ('jkl',1000):9,\n",
    "                   ('jkl',1500):14,\n",
    "                   ('xle',500):11.8,\n",
    "                   ('xle',1000):5.6,\n",
    "                   ('xle',1500):11.4,\n",
    "                   ('kce',500):7.02,\n",
    "                   ('kce',1000):22.5,\n",
    "                   ('kce',1500):4.79,\n",
    "                   ('kie',500):3.69,\n",
    "                   ('kie',1000):10.4,\n",
    "                   ('kie',1500):9,\n",
    "                   ('ewz',500):9.28,\n",
    "                   ('ewz',1000):22.19,\n",
    "                   ('ewz',1500):25.21,\n",
    "                   ('gld',500):9.5,\n",
    "                   ('gld',1000):9.91,\n",
    "                   ('gld',1500):14.7,\n",
    "                   ('spy',500):17.7,\n",
    "                   ('spy',1000):13,\n",
    "                   ('spy',1500):12,\n",
    "                   ('xlf',500):9,\n",
    "                   ('xlf',1000):14.6,\n",
    "                   ('xlf',1500):9.6,\n",
    "                   ('fvl',500):9.8,\n",
    "                   ('fvl',1000):4.3,\n",
    "                   ('fvl',1500):6.25}"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "500 training days  106.28999999999999\n",
      "1000 training days  116.58999999999999\n",
      "1500 training days  132.71999999999997\n"
     ]
    }
   ],
   "source": [
    "total_4_bins_scores = {x:0 for x in [500,1000,1500]}\n",
    "for k,val in four_bins_scores.items():\n",
    "    total_4_bins_scores[k[1]] += val\n",
    "for k,val in total_4_bins_scores.items():\n",
    "    print(k, 'training days ', val)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## What kind of prediction technique?\n",
    "Trying to run viterbi in the end of every predicted result"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "final_viterbi_scores = {}\n",
    "viterbi_predictions = {}\n",
    "training_days = 1500\n",
    "for etf in etfs:\n",
    "    for hidden_state in [2,4]:\n",
    "        #training\n",
    "        start = time.time()\n",
    "        training =  get_training_symbols(etf_dict[etf], training_days, 3)\n",
    "\n",
    "        temp_enu, temp_reverse_enu = enumerate_features(training)\n",
    "        predicted_data = predict(250, hidden_state, training, temp_enu, temp_reverse_enu,'advanced') #predict 250 days into the future\n",
    "        predicted = predicted_data[-250:]\n",
    "        end = time.time()\n",
    "        viterbi_predictions[(etf,hidden_state)] = predicted\n",
    "        print(hidden_state,'states, etf name: ',etf,' training time is: ', end - start)\n",
    "\n",
    "        #prediction and translation       \n",
    "        temp_df = etf_dict[etf].iloc[-(training_days + 250):-250, :]\n",
    "        temp_true_vals = etf_dict[etf].iloc[-250:,:]['Close'].values\n",
    "        final_vals = translate(temp_df, predicted, 100) #number of bootstrap iterations = 100\n",
    "\n",
    "        file_name = open('1500_viterbi_' + str(etf) +'_'+ str(hidden_state), 'wb')\n",
    "        pickle.dump(predicted, file_name)\n",
    "        #score calculation\n",
    "        temp_score = get_MAPE_score(temp_true_vals, final_vals)\n",
    "        print(hidden_state, 'hidden states, etf name: ',etf,' score is: ',temp_score)\n",
    "        final_viterbi_scores[(etf,hidden_state)] = temp_score"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "2 hidden statas: 94.43373160939939\n",
      "4 hidden statas: 128.02405417391742\n"
     ]
    }
   ],
   "source": [
    "file_name = open('1500_viterbi_scores', 'rb')\n",
    "viterbi_1500_scores = pickle.load(file_name)\n",
    "for k,val in viterbi_1500_scores.items():\n",
    "    print(k, \"hidden statas:\", viterbi_1500_scores[k])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "After trying all of these combinations and comparing all models we can see that using 1500 training days, 2 hidden states, 2 symbols and normal prediction technique achieved by far the lowest MAPE score of 82.41!\n",
    "\n",
    "Having that figuard out, let's try answering other basic questions:"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# How long to predict?\n",
    "Using the best model configuration (1500 training days + 2 hidden states + 2 symbols + Normal prediction technique), we will try to find out the best time for predicting ETFs prices."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [],
   "source": [
    "file_name = open('total_symbols_predictions', 'rb')\n",
    "best_predictions = pickle.load(file_name)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [],
   "source": [
    "total_errors = {y:np.array([float(0) for x in range(250)]) for y in [1500]}\n",
    "\n",
    "for k,val in best_predictions.items():\n",
    "    if k[1] == 2 and k[2] == 1500: # we are only interested in the best model performance\n",
    "        etf_name = k[0]\n",
    "        training_days = 1500\n",
    "        temp_df = etf_dict[etf_name].iloc[-(training_days + 250):-250, :]\n",
    "        temp_true_vals = np.array(etf_dict[etf_name].iloc[-250:,:]['Close'].values)\n",
    "        final_vals = np.array(translate(temp_df, val, 100)) #number of bootstrap iterations = 100\n",
    "        temp_diff =np.sqrt(np.abs( (temp_true_vals ** 2) - (final_vals ** 2)))\n",
    "        total_errors[training_days] += temp_diff"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [],
   "source": [
    "total_errors[1500] = total_errors[1500] / 10"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Text(0.5, 1.0, 'Overall prediction error per day')"
      ]
     },
     "execution_count": 20,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1152x576 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.figure(figsize=(16,8))\n",
    "plt.plot(total_errors[1500], label='Average prediction error', color = 'red')\n",
    "plt.legend(loc='upper left')\n",
    "plt.title('Overall prediction error per day')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "From the above graph, we can spot a spike in the error made by our model after more than 100 trading days and a surprising minimal error after about 80 days into the future, meaning the optimal prediction period to be in the first quarter of the following year."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Basic Comparison technique\n",
    "In this section, we will try to define a more intuitive comparison technique between our predictions and the actual ETF behavior by examining periodic changes in the closing price of the various ETFs."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 113,
   "metadata": {},
   "outputs": [],
   "source": [
    "def get_frequant_changes(stock_df):\n",
    "    changes = {x:[] for x in ['year_changes','quarter_changes','month_changes','week_changes']}\n",
    "    temp_counter = 0\n",
    "\n",
    "    #set initial parameters\n",
    "    for index, row in stock_df.iterrows():\n",
    "        starting_day = int(row['Date'][-2:])\n",
    "        starting_month = int(row['Date'][6:7])\n",
    "        starting_year = int(row['Date'][:4])\n",
    "        year_price,qurter_price, month_price,week_price = row['Close'],row['Close'],row['Close'],row['Close'] \n",
    "        break\n",
    "    flag = True\n",
    "    q_count = 0\n",
    "    for index, row in stock_df.iterrows():\n",
    "        current_day = float(row['Date'][-2:])\n",
    "        current_month = float(row['Date'][6:7])\n",
    "        current_year = float(row['Date'][:4])\n",
    "        current_price = float(row['Close'])\n",
    "        if flag:\n",
    "            previous_day,previous_month,previous_year = current_day, current_month, current_year\n",
    "            flag = False\n",
    "            continue\n",
    "\n",
    "        if (previous_day + 1) != current_day: #week break\n",
    "            if current_day != 1:\n",
    "                changes['week_changes'] += [((current_price - week_price) / week_price) * 100]\n",
    "                week_price = current_price\n",
    "\n",
    "        if previous_month  != current_month: #month break\n",
    "            changes['month_changes'] += [((current_price - month_price) / month_price) * 100]\n",
    "            month_price = current_price\n",
    "            q_count += 1\n",
    "            if q_count == 3: #end of quarter\n",
    "                changes['quarter_changes'] += [((current_price - qurter_price) / qurter_price) * 100]\n",
    "                qurter_price = current_price\n",
    "                q_count = 0\n",
    "\n",
    "        if previous_year != current_year: #year break\n",
    "            changes['year_changes'] += [((current_price - year_price) / year_price) * 100]\n",
    "            year_price = current_price\n",
    "\n",
    "        previous_day,previous_month,previous_year = current_day, current_month, current_year\n",
    "    return changes\n",
    "    "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 114,
   "metadata": {},
   "outputs": [],
   "source": [
    "def get_complex_future_trend_prediction(stock_df, predictions):\n",
    "    future_trends ={x:'do somthing' for x in ['week_changes','month_changes','quarter_changes','year_changes']}\n",
    "    indiffrence_ranges = {x:[] for x in ['week_changes','month_changes','quarter_changes','year_changes']}\n",
    "    \n",
    "    first_price = np.array(stock_df['Open'])[-1]\n",
    "    final_vals = translate(stock_df, predictions, 100)\n",
    "    changes = get_frequant_changes(stock_df)\n",
    "    \n",
    "    for k, l in zip(future_trends.keys(), [5,25,75,250]):\n",
    "        temp_change = ((final_vals[l -1] - first_price) / first_price) * 100\n",
    "        if l in [5,25]:\n",
    "            if np.quantile(changes[k], 0.15) <= temp_change <= np.quantile(changes[k], 0.85):\n",
    "                future_trends[k] = 'HOLD'\n",
    "            elif temp_change < np.quantile(changes[k], 0.15):\n",
    "                future_trends[k] = 'HARD SELL'\n",
    "            else:\n",
    "                future_trends[k] = 'HARD BUY'\n",
    "        else: # quarter and year changes\n",
    "            if np.quantile(changes[k], 0.15) <= temp_change <= np.quantile(changes[k], 0.85): #holding range\n",
    "                if l == 75: #quarter\n",
    "                    if temp_change > 5:\n",
    "                        future_trends[k] = 'HARD BUY'\n",
    "                    else:\n",
    "                        future_trends[k] = 'HOLD'\n",
    "                else: #year\n",
    "                    if temp_change > 10:\n",
    "                        future_trends[k] = 'HARD BUY'\n",
    "                    else:\n",
    "                        future_trends[k] = 'HOLD'\n",
    "            elif temp_change < np.quantile(changes[k], 0.15):\n",
    "                future_trends[k] = 'HARD SELL'\n",
    "            else:\n",
    "                future_trends[k] = 'HARD BUY'\n",
    "    return future_trends"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 115,
   "metadata": {},
   "outputs": [],
   "source": [
    "def get_true_trends(etf, etf_dict, training_days, prediction_days):\n",
    "    future_trends ={x:'do somthing' for x in ['week_changes','month_changes','quarter_changes','year_changes']}\n",
    "    indiffrence_ranges = {x:[] for x in ['week_changes','month_changes','quarter_changes','year_changes']}\n",
    "    \n",
    "    temp_df = etf_dict[etf].iloc[-(training_days + prediction_days):-prediction_days, :]\n",
    "    true_vals_df = etf_dict[etf].iloc[-prediction_days:, :]\n",
    "    first_price = np.array(temp_df['Open'])[-1] #start predicting from the last known open price\n",
    "    changes = get_frequant_changes(temp_df)\n",
    "    true_vals = true_vals_df['Close'].values\n",
    "    for k, l in zip(future_trends.keys(), [5,25,75,250]):\n",
    "        indiffrence_ranges[k] = np.quantile(changes[k], [0.15,0.85])\n",
    "        true_change = ((true_vals[l - 1] - first_price ) / first_price) * 100\n",
    "        if l in [5,25]:\n",
    "            if np.quantile(changes[k], 0.15) <= true_change <= np.quantile(changes[k], 0.85):\n",
    "                future_trends[k] = 'HOLD'\n",
    "            elif true_change < np.quantile(changes[k], 0.15):\n",
    "                future_trends[k] = 'HARD SELL'\n",
    "            else:\n",
    "                future_trends[k] = 'HARD BUY'\n",
    "        else: # quarter and year changes\n",
    "            if np.quantile(changes[k], 0.15) <= true_change <= np.quantile(changes[k], 0.85): #holding range\n",
    "                if l == 75: #quarter\n",
    "                    if true_change > 5:\n",
    "                        future_trends[k] = 'HARD BUY'\n",
    "                    else:\n",
    "                        future_trends[k] = 'HOLD'\n",
    "                else: #year\n",
    "                    if true_change > 10:\n",
    "                        future_trends[k] = 'HARD BUY'\n",
    "                    else:\n",
    "                        future_trends[k] = 'HOLD'\n",
    "            elif true_change < np.quantile(changes[k], 0.15):\n",
    "                future_trends[k] = 'HARD SELL'\n",
    "            else:\n",
    "                future_trends[k] = 'HARD BUY'\n",
    "    return future_trends, indiffrence_ranges"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 118,
   "metadata": {},
   "outputs": [],
   "source": [
    "def get_f1_acc_scores(suggested_preds, stock_dic):\n",
    "    predicted_trends = {x:[] for x in ['week_changes','month_changes','quarter_changes','year_changes']}\n",
    "    true_trends =  {x:[] for x in ['week_changes','month_changes','quarter_changes','year_changes']}\n",
    "    training_days = 1500\n",
    "    prediction_days = 250\n",
    "    for k,val in suggested_preds.items():\n",
    "        temp_df = stock_dic[k].iloc[-(training_days + prediction_days):-prediction_days, :]\n",
    "        temp_predicted_changes = get_complex_future_trend_prediction(temp_df,val)\n",
    "        temp_true_trends, indiffrence_ranges = get_true_trends(k, stock_dic, training_days, prediction_days)\n",
    "        for tk in true_trends.keys():\n",
    "            predicted_trends[tk] += [temp_predicted_changes[tk]]\n",
    "            true_trends[tk] += [temp_true_trends[tk]]\n",
    "    for k in true_trends.keys():\n",
    "        print(k,'accuracy is: ', accuracy_score(true_trends[k], predicted_trends[k]),' f1 score is:',f1_score(true_trends[k], predicted_trends[k],average='macro'))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Before inspecting our model with that method let's use it to analyze our baseline models:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 246,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "('AR', 'week_changes') accuracy score is: 0.5 F1 score is: 0.3333333333333333\n",
      "('AR', 'month_changes') accuracy score is: 0.1 F1 score is: 0.06060606060606061\n",
      "('AR', 'quarter_changes') accuracy score is: 0.3 F1 score is: 0.2222222222222222\n",
      "('AR', 'year_changes') accuracy score is: 0.3 F1 score is: 0.3\n",
      "('MA', 'week_changes') accuracy score is: 0.2 F1 score is: 0.16666666666666666\n",
      "('MA', 'month_changes') accuracy score is: 0.1 F1 score is: 0.06060606060606061\n",
      "('MA', 'quarter_changes') accuracy score is: 0.1 F1 score is: 0.06666666666666667\n",
      "('MA', 'year_changes') accuracy score is: 0.3 F1 score is: 0.3\n",
      "('ARMA', 'week_changes') accuracy score is: 0.5 F1 score is: 0.2380952380952381\n",
      "('ARMA', 'month_changes') accuracy score is: 0.1 F1 score is: 0.06060606060606061\n",
      "('ARMA', 'quarter_changes') accuracy score is: 0.3 F1 score is: 0.19999999999999998\n",
      "('ARMA', 'year_changes') accuracy score is: 0.3 F1 score is: 0.3\n",
      "('ARIMA', 'week_changes') accuracy score is: 0.5 F1 score is: 0.3333333333333333\n",
      "('ARIMA', 'month_changes') accuracy score is: 0.1 F1 score is: 0.06060606060606061\n",
      "('ARIMA', 'quarter_changes') accuracy score is: 0.4 F1 score is: 0.375\n",
      "('ARIMA', 'year_changes') accuracy score is: 0.3 F1 score is: 0.20454545454545456\n"
     ]
    }
   ],
   "source": [
    "base_f1_acc_scores = {}\n",
    "\n",
    "prediction_days = 250\n",
    "for method in ['AR','MA','ARMA','ARIMA']:\n",
    "    for training_days in [1000]:\n",
    "        predicted_trends = {x:[] for x in ['week_changes','month_changes','quarter_changes','year_changes']}\n",
    "        true_trends =  {x:[] for x in ['week_changes','month_changes','quarter_changes','year_changes']}\n",
    "        for k,val in baseline_preds.items():\n",
    "            if k[1] == method and k[2] == training_days:\n",
    "                temp_df = etf_dict[k[0]].iloc[-(training_days + prediction_days):-prediction_days, :]\n",
    "                temp_predicted_changes = get_complex_future_trend_prediction_baseline(temp_df,val)\n",
    "                temp_true_trends, indiffrence_ranges = get_true_trends(k[0], etf_dict, training_days, prediction_days)\n",
    "                for tk in true_trends.keys():\n",
    "                    predicted_trends[tk] += [temp_predicted_changes[tk]]\n",
    "                    true_trends[tk] += [temp_true_trends[tk]]\n",
    "\n",
    "        for m in true_trends.keys():\n",
    "            base_f1_acc_scores[(method,m)] = 'accuracy score is: ' + str(accuracy_score(true_trends[m], predicted_trends[m])) +' F1 score is: '+ str(f1_score(true_trends[m], predicted_trends[m],average='macro'))\n",
    "\n",
    "for k in base_f1_acc_scores.keys():\n",
    "    print(k, base_f1_acc_scores[k])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Not a great success, so let's move on to our best model and see if it performed any better:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 249,
   "metadata": {},
   "outputs": [],
   "source": [
    "file_name = open('total_symbols_predictions', 'rb')\n",
    "suggested_preds = pickle.load(file_name)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 252,
   "metadata": {},
   "outputs": [],
   "source": [
    "predicted_trends = {x:[] for x in ['week_changes','month_changes','quarter_changes','year_changes']}\n",
    "true_trends =  {x:[] for x in ['week_changes','month_changes','quarter_changes','year_changes']}\n",
    "\n",
    "training_days = 1500\n",
    "prediction_days = 250\n",
    "for k,val in suggested_preds.items():\n",
    "    if k[1] == 2 and k[2] == 1500:\n",
    "        temp_df = etf_dict[k[0]].iloc[-(training_days + prediction_days):-prediction_days, :]\n",
    "        temp_predicted_changes = get_complex_future_trend_prediction(temp_df,val)\n",
    "        temp_true_trends, indiffrence_ranges = get_true_trends(k[0], etf_dict, training_days, prediction_days)\n",
    "        for tk in true_trends.keys():\n",
    "            predicted_trends[tk] += [temp_predicted_changes[tk]]\n",
    "            true_trends[tk] += [temp_true_trends[tk]]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 253,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "week_changes accuracy score is:  0.4  f1 score is: 0.2424242424242424\n",
      "month_changes accuracy score is:  0.2  f1 score is: 0.14074074074074072\n",
      "quarter_changes accuracy score is:  0.8  f1 score is: 0.7916666666666665\n",
      "year_changes accuracy score is:  0.5  f1 score is: 0.2564102564102564\n"
     ]
    }
   ],
   "source": [
    "for k in true_trends.keys():\n",
    "    print(k,'accuracy score is: ', accuracy_score(true_trends[k], predicted_trends[k]),' f1 score is:',f1_score(true_trends[k], predicted_trends[k],average='macro'))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Examining our model performance, we can spot that quarterly changes in the ETF's are the easiest to predict with 80% accuracy and 0.79 F1 score!\n",
    "\n",
    "Moreover, that finding approves the results of our \"best prediction period\" to be around 80 trading days when 75 trading days make for a quarter of a year (full trading year are composed of 252 trading days)."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Advanced Part"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "In this part, we will try to predict the prices of the \"XLK\" (technology sector fund) by predicting the prices of it's composing stocks including Apple, Microsoft, IBM and more."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "At first, let's have a look on our best prediction yet, regarding the XLK fund:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 219,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAA6IAAAHiCAYAAADyP3HCAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADl0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uIDMuMC4yLCBodHRwOi8vbWF0cGxvdGxpYi5vcmcvOIA7rQAAIABJREFUeJzs3Xd0VVX+/vH3TiEJSQikEaSEKi2QELogICCKMDoWcPgNCBYQy6CMooyDZYkzX8QyOBYURbCjICo4FmwMCiNICdIJTQgljZACSchNzu+Pm3tJTICUm9yU57WW67R9zvncOGvNetz77G0sy0JERERERESkuni4uwARERERERGpXxRERUREREREpFopiIqIiIiIiEi1UhAVERERERGRaqUgKiIiIiIiItVKQVRERERERESqlYKoiIjUSsaY/zPG3F+B+w4ZY4YX7j9hjHnX9dXVfMaY7saYde6uQ0RE6icFURERqXWMMWHALcBrhcdDjDGrXfyOJ4wxT5Sx7QxjzHZjTKYx5qAxZkY53tPMGLPCGHPMGGMZY1r/7vpiY8xZY0xWkX88i1wfZozZbYw5Y4z5wRgTWeSajzHmTWNMhjHmhDHmr45rlmX9CpwyxvyhrLWKiIi4ioKoiIjURpOALyzLynZ3IYUM9mDcBLgauNcY86cy3lsAfAXceIE2cy3LCijyTz6AMSYUWA48CgQDG4EPi9z3BNABiASuAB4yxlxd5Pp7wJ1lrFNERMRlFERFRKQ2Ggn8t7QLxpjLjDEpxpiWhcfRxphTxphOF3qgMcbbGPOBMeZjY0yD8hRjWdZcy7I2W5ZlsyxrD/AZMKCM9yZalvUK8Et53lnoBmCHZVlLLcvKwR48o4v81luA2ZZlpVmWtQt4HXuId1gNDDPG+FTg3SIiIhWmICoiIrVRN2CP48CyrNWWZQ0p3F+HfcjuW8YYP+AdYJZlWbvP97DCdp8CucBYy7LOWpb1hGVZT5S3MGOMAS4HdpT33gu42xhz0hizyRhTtOe0K7DVcWBZ1mlgP9DVGNMEuKTo9cL9rkXaHwXygI4urFVEROSiFERFRKQ2agxkXuD6E0AQsAE4Brx8gbaNsA+N3Q/c6hj2WglPYP//10WVfI7Dv7EPrw3HPgR3sTHG0dsaAKT/rn06EFh4jd9dd1wrKhP731NERKTaKIiKiEhtlEbJQOVkWVYesBiIAp6zLMu6wLP6Ad2BORdpd1HGmHuxD4cdZVlWbmWe5VA45De1cNjvF9i/67yh8HIW9iBdVCPs4TKryPHvrxUVCJxyRa0iIiJlpSAqIiK10a/Apee7aIxpDjyOvVfyuYt8A7kK+D/gO2NM04oWZIy5DZgJDLMsK6GizykDC/vkSGAf/htdpAZ/oB3270bTgONFrxfu7yjS/hKgAUWGOYuIiFQHBVEREamNvgAGl3ah8BvNxcBC4HbsYWz2hR5mWdZc4H3sYTT0PM+1jDFDznPtz8A/gSstyzpQyvXVF1oKxhjjCzjCsk/hsePaTcaYAGOMhzFmBDAeWFF4+RMgyhhzY+E9jwG/Fvke9m1gljGmSeEERpOx/20chgDfu6r3VkREpKwUREVEpDZ6G7imcJKh35sGNAUeLRxqeytwqzHm8gs90LKs2dgnLPrWGBNc9JoxpgX2oa7bznP7U0AI8EuRtT5fLXK9JbD2Aq/P5txQ2t2Fxw73AUexD599BphsWdbqwpqTsS/78g/sw5X7AkWXjXkc+7evv2GfZfgZy7K+KnL9z0DROkVERKqFqeTnMCIiIm5hjPknkGRZ1rxqeNd4oKtlWX+rwL0tgKWWZfV3fWUVZ4zpBiyoaXWJiEj9oCAqIiIiIiIi1UpDc0VERERERKRaKYiKiIiIiIhItVIQFRERERERkWqlICoiIiIiIiLVyqs6XxYaGmq1bt26Ol8pIiIiIiIi1WTTpk0plmWFXaxdtQbR1q1bs3Hjxup8pYiIiIiIiFQTY8xvZWmnobkiIiIiIiJSrRRERUREREREpFopiIqIiIiIiEi1qtZvREuTl5dHQkICOTk57i5FXMzX15cWLVrg7e3t7lJERERERKQGcXsQTUhIIDAwkNatW2OMcXc54iKWZZGamkpCQgJt2rRxdzkiIiIiIlKDuH1obk5ODiEhIQqhdYwxhpCQEPV0i4iIiIhICW4PooBCaB2lf68iIiIiIlKaGhFE3Sk1NZWYmBhiYmKIiIigefPmzuOzZ8+W6Rm33nore/bsuWCbl19+mffee88VJTNw4EA6duxIdHQ0AwcOJD4+vlz3l7eWffv2ERMTU94yRURERERESuX2b0TdLSQkhLi4OACeeOIJAgICePDBB4u1sSwLy7Lw8Cg9ty9atOii77nnnnsqX2wRH374ITExMbzyyis8/PDDLF++vEz32Ww2l9ciIiIiIiJSHvW+R/R89u3bR1RUFFOnTiU2Npbjx48zZcoUevXqRdeuXXnyySedbQcOHEhcXBw2m43GjRszc+ZMoqOj6d+/P0lJSQDMmjWLefPmOdvPnDmTPn360LFjR9atWwfA6dOnufHGG4mOjmbcuHH06tXLGZLPZ9CgQezbtw+AX375hcGDB9OzZ09GjhxJYmKi831///vfGTRoEC+99FKxWjZv3kzfvn3p3r07N954I+np6c5nde/enf79+/Pqq6+68C8rIiIiIiL1XY3qEb3/frhI7iq3mBgozFzltnPnThYtWuQMYnPmzCE4OBibzcYVV1zBTTfdRJcuXYrdk56ezuDBg5kzZw5//etfefPNN5k5c2aJZ1uWxYYNG1ixYgVPPvkkX331FS+++CIRERF8/PHHbN26ldjY2IvWuHLlSrp160Zubi733XcfK1asIDQ0lPfee49HH32UBQsWAJCRkcGaNWsAeyh2GD9+PAsWLGDgwIE88sgjzJ49m2effZZJkyaxYMECBgwYwPTp0yv2BxQRERERESlFjQqiNU27du3o3bu38/iDDz5g4cKF2Gw2jh07xs6dO0sEUT8/P0aOHAlAz549+fHHH0t99g033OBsc+jQIQB++uknHn74YQCio6Pp2rXreWu7+eab8fPzo23btrz44ovs2rWLHTt2MHz4cADy8/Np0aKFs/2f/vSnEs9ITU0lJyeHgQMHAjBx4kQmTJhASkoK2dnZDBgwAIAJEybwww8/nP8PJSIiIiIiUg41KohWtOeyqvj7+zv34+PjeeGFF9iwYQONGzdm/PjxpS5N0qBBA+e+p6cnNput1Gf7+PiUaGNZVplrc3wj6pCUlET37t3PG3yL/haHC71PM96KiIiIiEhV0TeiZZSRkUFgYCCNGjXi+PHjfP311y5/x8CBA/noo48A2LZtGzt37izzvV26dOHo0aNs2LABgLNnz7Jjx44L3hMaGoqfn5/zG9V33nmHwYMHExoaiq+vL//73/8AXDbbr4iIiIiICNSwHtGaLDY2li5duhAVFUXbtm2dw1Zd6S9/+Qu33HIL3bt3JzY2lqioKIKCgsp0r4+PD8uWLWPatGlkZmZis9l44IEHLji8F+zh86677iI7O5v27ds7ZwBetGgRd9xxB/7+/owYMaLSv01ERERERMTBlGc4aGX16tXL2rhxY7Fzu3btonPnztVWQ01ms9mw2Wz4+voSHx/PiBEjiI+Px8ur9v73Av37FRERERGpP4wxmyzL6nWxdrU34dRBWVlZDBs2DJvNhmVZvPbaa7U6hIqIiIiIiJRGKacGady4MZs2bXJ3GSIiIiIiUoOMXTqW1YdWkzQjyd2luIyCqIiIiIiISA22dOdSd5fgcpo1V0REREREpBbIzM10dwkuoyAqIiIiIiJSQ60+tNq5fzzruPsKcTEFURERERERkRrq8dWPO/fn/DTHjZW4Vr0PoqmpqcTExBATE0NERATNmzd3Hp89e9Yl78jMzCQkJISsrKxi50ePHs3y5ct54403uP/++13yLhERERERqRuyzmax5rc1TI6djL+3P5uO152JTet9EA0JCSEuLo64uDimTp3K9OnTnccNGjQAwLIsCgoKKvyOwMBAhg4dymeffeY8l5aWxvr167nmmmsq/RtERERERKTuOZZ5DIBBkYNIeziN9Xesd3NFrlPvg+j57Nu3j6ioKKZOnUpsbCxHjhyhcePGzutLlizhjjvuACAxMZEbbriBXr160adPH37++ecSzxs3bhxLlixxHn/88ceMGjUKX1/fYu3Gjx/Pp59+6jwOCAhw7s+ZM4c+ffrQvXt3nnzyScDe2zpy5Eiio6OJiopi2bJlrvkDiIiIiIiIWx3NOArAJYGX4O3pja+X70XuqD1q1PIt9391P3En4lz6zJiIGOZdPa9C9+7cuZNFixbx6quvYrPZzttu2rRpPPTQQ/Tr149Dhw4xevRotm/fXqzNqFGjuPPOO0lLS6NJkyYsWbKEGTNmlLmWL774gsOHD7N+/Xosy+Kaa65h3bp1HDlyhNatW/Pll18CkJ6eXqHfKiIiIiIiNYujR7R5YHM3V+J6NSqI1jTt2rWjd+/eF2337bffsmfPHudxWloa2dnZ+Pn5Oc/5+PgwatQoli9fzujRo9mxYwfDhg0rcy2rVq3iyy+/pEePHgBkZWWxd+9e+vbty8yZM5k5cyZ/+MMfGDBgQDl+oYiIiIiI1ESJWYm89MtLgL1HtK6pUUG0oj2XVcXf39+57+HhgWVZzuOcnBznvmVZbNiwwflN6fmMGzeOZ599luzsbG644Qa8vEr++b28vJzfo+bn5zt7Yi3LYtasWdx+++0l7tm4cSNffPEFM2bMYPTo0TzyyCPl+6EiIiIiIlKjvLH5DX5O+Jnp/aYT6BPo7nJcTt+IlpGHhwdNmjQhPj6egoICPvnkE+e14cOH8/LLLzuP4+JKH148fPhwduzYwauvvsq4ceNKbdO6dWs2bbLPhvXJJ5+Qn58PwFVXXcXChQs5ffo0AAkJCaSkpHD06FECAgKYMGECf/3rX9m8ebNLfq+IiIiIiLjP5hObaR/cnuevet7dpVQJBdFyePrpp7n66qsZNmwYLVq0cJ5/+eWXWbt2Ld27d6dLly68/vrrpd7v6enJ9ddfT0ZGxnmH0N55551888039OnTh7i4OHx8fAC45ppruOmmm+jXrx/dunVj7NixZGVlsXXrVnr37k1MTAxz585Vb6iIiIiISC1nWRYbjm6gR0QPd5dSZUzR4aZVrVevXtbGjRuLndu1axedO3euthqkeunfr4iIiIhI+aw7so4Bbw7g7T++zYToCe4up1yMMZssy+p1sXbqERUREREREalB9p3cB8BlLS9zcyVVR0FURERERESkBsmx2SdG9fP2u0jL2ktBVEREREREpAZxBFFfL183V1J1FERFRERERERqEAVRERERERERqVaOIOrj6ePmSqqOgqiIiIiIiEgNkmPLwdvDG08PT3eXUmUURLGv7xkTE0NUVBRjxozhzJkzFX7W6tWrGT16NAArVqxgzpw552176tQpXnnllXK/44knnuDZZ58t8d7+/fsXO2ez2WjatCnHjx8v17NERERERMR9cmw5dXpYLiiIAuDn50dcXBzbt2+nQYMGvPrqq8WuW5ZFQUFBuZ977bXXMnPmzPNer2gQLc2gQYNISEjg0KFDznPffvstUVFRNGvWzCXvEBERERGRqqcgWg9dfvnl7Nu3j0OHDtG5c2fuvvtuYmNjOXLkCKtWraJ///7ExsYyZswYsrKyAPjqq6/o1KkTAwcOZPny5c5nLV68mHvvvReAxMRErr/+eqKjo4mOjmbdunXMnDmT/fv3ExMTw4wZMwB45pln6N27N927d+fxxx93Pusf//gHHTt2ZPjw4ezZs6dE3R4eHowZM4YPP/zQeW7JkiWMGzcOgNdff53evXsTHR3NjTfeWGqv75AhQ9i4cSMAKSkptG7dGoD8/HxmzJjhrOu1114D4Pjx4wwaNMjZm/zjjz9W+O8uIiIiIiJ29SGIerm7gKLuv/9+4uLiXPrMmJgY5s2bV6a2NpuNL7/8kquvvhqAPXv2sGjRIl555RVSUlJ46qmn+Pbbb/H39+fpp5/m+eef56GHHmLy5Ml8//33tG/fnptvvrnUZ0+bNo3BgwfzySefkJ+fT1ZWFnPmzGH79u3O37xq1Sri4+PZsGEDlmVx7bXXsmbNGvz9/VmyZAlbtmzBZrMRGxtLz549S7xj3LhxTJkyhYcffpjc3Fy++OIL/vWvfwFwww03MHnyZABmzZrFwoUL+ctf/lKmv8vChQsJCgril19+ITc3lwEDBjBixAiWL1/OVVddxd///nfy8/MrNaRZRERERETsFETriezsbGJiYgB7j+jtt9/OsWPHiIyMpF+/fgD8/PPP7Ny5kwEDBgBw9uxZ+vfvz+7du2nTpg0dOnQAYPz48SxYsKDEO77//nvefvttwP5NalBQEGlpacXarFq1ilWrVtGjRw8AsrKyiI+PJzMzk+uvv56GDRsC9iG/penduzdZWVns2bOHXbt20a9fP5o0aQLA9u3bmTVrFqdOnSIrK4urrrqqzH+fVatW8euvv7Js2TIA0tPTiY+Pp3fv3tx2223k5eXxxz/+0fk3FBERERGRilMQrWZl7bl0Ncc3or/n7+/v3LcsiyuvvJIPPvigWJu4uDiMMS6pw7Is/va3v3HnnXcWOz9v3rwyv+NPf/oTS5YsYdeuXc5huQCTJk3i008/JTo6msWLF7N69eoS93p5eTm/hc3JySlW14svvlhqeF2zZg3/+c9/mDBhAjNmzOCWW24pU50iIiIiIlK6+hBE9Y1oGfXr14+1a9eyb98+AM6cOcPevXvp1KkTBw8eZP/+/QAlgqrDsGHDmD9/PmD/5jIjI4PAwEAyMzOdba666irefPNN57enR48eJSkpiUGDBvHJJ5+QnZ1NZmYmK1euPG+d48aN49133+X7778v1nOamZlJs2bNyMvL47333iv13tatW7Np0yYAZ++no6758+eTl5cHwN69ezl9+jS//fYb4eHhTJ48mdtvv53Nmzdf+I8oIiIiIiLntWjLIh7/4fF6EURrVI9oTRYWFsbixYsZN24cubm5ADz11FNceumlLFiwgFGjRhEaGsrAgQPZvn17iftfeOEFpkyZwsKFC/H09GT+/Pn079+fAQMGEBUVxciRI3nmmWfYtWuXcxmWgIAA3n33XWJjY7n55puJiYkhMjKSyy+//Lx1dunShYYNG9KzZ89iPbqzZ8+mb9++REZG0q1bt2IB2OHBBx9k7NixvPPOOwwdOtR5/o477uDQoUPExsZiWRZhYWF8+umnrF69mmeeeQZvb28CAgKcQ49FRERERKT8bltxm3P/yrZXurGSqmcsy6q2l/Xq1ctyzMrqsGvXLjp37lxtNUj10r9fEREREZGLW7lnJdcuOTei8Q+X/oEV41a4saKKMcZssiyr18XaaWiuiIiIiIiIGyWdTioWQgHaNmnrpmqqh4KoiIiIiIiIG/3zx3+WOBcREOGGSqqPgqiIiIiIiIgbbTi6gW7h3fjvpP9y9K9HmRQziam9prq7rCpVIyYrsizLZUugSM1Rnd8fi4iIiIjURuk56Ww8tpF7+9zLoMhBACy6bpGbq6p6ZeoRNcY0NsYsM8bsNsbsMsb0N8YEG2O+McbEF26bVKQAX19fUlNTFVrqGMuySE1Nxde3bk87LSIiIiJSUZZl0fjpxuQV5NGzWU93l1Otytoj+gLwlWVZNxljGgANgUeA7yzLmmOMmQnMBB4ubwEtWrQgISGB5OTk8t4qNZyvry8tWrRwdxkiIiIiIjXS9qRzyz72vERBtBhjTCNgEDAJwLKss8BZY8x1wJDCZm8Bq6lAEPX29qZNmzblvU1ERERERKRW+8eP/wBgbNexXBpyqZurqV5lGZrbFkgGFhljthhj3jDG+ANNLcs6DlC4Da/COkVEREREROqME1knWLpzKTMum8GHN32Ih6lf88iW5dd6AbHAfMuyegCnsQ/DLRNjzBRjzEZjzEYNvxUREREREYF5P8/Dsixu63Gbu0txi7IE0QQgwbKs9YXHy7AH00RjTDOAwm1SaTdblrXAsqxelmX1CgsLc0XNIiIiIiIitdpX+75ieNvhdArt5O5S3OKiQdSyrBPAEWNMx8JTw4CdwApgYuG5icBnVVKhiIiIiIhIHXMi6wSRQZHuLsNtyjpr7l+A9wpnzD0A3Io9xH5kjLkdOAyMqZoSRURERERE6o78gnySzyTTNKCpu0txmzIFUcuy4oBepVwa5tpyRERERERE6rbU7FQKrAKa+tffIFq/pmYSERERERFxsxnfzACo1z2iCqIiIiIiIiLV5EzeGd7e+jYAPZv1dHM17qMgKiIiIiIiUkXyC/JZn7DeefzShpcA+OnWn2gX3M5dZbmdgqiIiIiIiEgVeXrt0/Rb2I//HfkfAP/97b90C+/GgFYD3FyZeymIioiIiIiIVJH/JdgD6JYTWwA4mHaQ9sHt3VlSjVDW5VtERERERESkjOauncu6I+v4NfFXADYd24RlWRw6dYiR7Ue6uTr3UxAVERERERFxgW/2f0OroFa0adKGh799uNi1Hck7+OnwT2TbstUjiobmioiIiIiIVJplWYx4dwSdXu7E7pTdxa7d2fNO1h9dz5NrniSwQSDju493U5U1h4KoiIiIiIhIJaWcSXHu70ja4dzvHNqZy1peBsC3B76lZVBLAn0Cq72+mkZBVEREREREpJIOpB1w7t/zxT0ArJ64mq1TtzK++3gaejcEINw/3C311TQKoiIiIiIiIpVUNIim5aQBEOYfhrenNx7Gg27h3QBo6t/ULfXVNAqiIiIiIiIi5fTDwR9IPZPqPN6ftr9Emya+TZz7IQ1DAPWIOiiIioiIiIiIlENCRgJD3x7KXf+5y3nuQNoBIgIiCGgQ4DzX2Lexc7+BZwMAIgIiqq/QGkxBVEREREREpBw+3vkxANm2bOe5A2kHaB/cniPTjzjP+Xr5OvcdM+n2bd63mqqs2RRERURERERELiIjN4OHvnmIk9knef7n5wFo2agltgIblmWxM3knHYI70Ni3Mb0v6Q2AMcZ5//WdrgegX4t+1V98DaQgKiIiIiIichH/2fsfnln3DLGvxXI4/TAA8Sfj8Z7tzeOrHyf5TDK9LukFwOpJq0l8MLHY/bOvmE36zHT8G/hXe+01kYKoiIiIiIhIKT7f+zkJGQkAHDx1EIDf0n9zXnesFzp7zWzgXG9nQ++GJSYl8vTwpJFPoyqvubZQEBUREREREfmdD7Z9wB8++APD3x7OtsRtzN84v0SbE1knnPv9W/SnR0SP6iyxVvNydwEiIiIiIiI1yaRPJ/HW1rcA2JO6h+6vdgfA03iSb+U721lYAHwz4Rv6Nu9b7JtQuTD1iIqIiIiIiBTalbzLGUJ/78HLHnTuO5ZjAegc2plAn8Aqr60uURAVEREREREptPrQagD+O+m/5D92rvfzo5s+YnLsZOfxJYGXOPeDfIOqrb66QkFURERERETqpK0nttLvjX6knEkp8z1xJ+II9gvm8laX42HOxaVhbYcVm/HWEUQ9jSf+3poJt7wUREVEREREpM6xLIvRH4xm/dH1xJ2IK/N9cYlxxETEOL/3nN5vOsF+wQT7BRPQIMDZrk3jNgDkW/n6NrQCFERFRERERKROycvPo//C/s6lVyzLKtN9tgIbvyb+SkzTGOe55696ntSHUgH7siwOTw19irZN2nJl2ytdWHn9oVlzRURERESkTtmetJ31R9c7j8/knSnTfc+te44cWw4xETGlXi86VLd149bsn7a/coXWY+oRFRERERGROuV41nEAxnQZA0C2LbtM9321/ysARnYYWTWFiZOCqIiIiIiI1CnHMo8BcG+fewHIzitbED2WeYyxXccS2jD0vG08jSfXdLim8kXWcxqaKyIiIiIidYojiLZt0hYo29Bcy7I4mnGUUR1GXbBd3qN5lS9QFERFRERERKTu2JOyh9lrZhPaMJQgH/v6nmUZmpt5NpPTeaeLrQ9aGs2Q6xoamisiIiIiInXG7Stux1Zgo3NoZ/y8/YDiQ3PP5J1h8orJHM88Xuy+zcc3AxAZFFl9xdZj6hEVEREREZE6YXfKbtYeWQvAC1e/gJeHF94e3sWG5n66+1Pe2PIGNsvGousWOc+/v+19gnyCGHXphYfmimsoiIqIiIiISJ2waIs9WL58zcv0aNYDAD9vP7Jt2ViWxdNrnybpdBIAp3JOFbv3RNYJ2jRpU2ytUKk6CqIiIiIiIlJr5dhyAPD18uVQ+iHaNG7DXb3ucl5v6N2Q7LxsTmaf5G/f/c15Pi07rdhzTmafJNgvuHqKFgVRERERERGpvSLnRWIrsJH6UCoH0w7SPrh9sQmF/Lz8WLB5gXMGXYeT2SdLHHcN71otNYsmKxIRERERkVos6XQSJ7NPYlkW+9P2lwicjgmLZn4303muS1gXtiVt4/bPbneeS81OJcQvpHqKFgVRERERERGpnVLOpDj396ft52T2SWKbxRZr4+vlW+K+fs37AfBm3JuAfQ1RDc2tXgqiIiIiIiJSK/2a+Ktz/5v93wDQv0X/Ym1+H0Q7BHcosVZoem46tgKbgmg1UhAVEREREZFaaeuJrc79+Rvn07NZT6LCo4q1GdWh+HIs397yLQENAoqdW75rOQC9L+ldRZXK7ymIioiIiIhIrXMk/Qhz1s4B7DPj7k3dy9A2Q4tNVAQwc+BMfp16rue0sW9jCqyCYm1+OvwT4f7hDIocVPWFC6AgKiIiIiIitdCkzyY51wQ9k3eG3PxcWgW1KtHOw3jQrWk3Z8gMaBBAXkFesTYHTx2kXZN2JUKsVB0FURERERERqVUKrAK2HN8CwB097nCeLy2IOnx3y3cc++sxPEzxCFRgFXAw7SBtmrSpmmKlVAqiIiIiIiJSq7wV9xZpOWm8d8N7dArt5Dx/oSDq5eFFs8BmAEzrO815Pj0nnYSMBNo0VhCtTgqiIiIiIiJSq6zcu5I2jdswLmpcsVlxOwR3KNP9jX0b89ro1wAInhtMvpVPREBEldQqpVMQFRERERGRWsGyLOJT4/np8E8MihyEMQY/bz8ADAb/Bv40YcwKAAAgAElEQVRlflaQT1Cx47CGYS6tVS5MQVRERERERGqFf/38Ly596VKSzyRzeavLAfDzsgfR8n7jeVX7q4odh/kriFYnBVEREREREakVFm5Z6Nwf2GogAN6e3gC0bty6XM9q7Nu42HBc9YhWLy93FyAiIiIiInIxlmVxJP0I13a8lvHdxtMxtCMAiVmJABWabKihd0PnfmjDUNcUKmWiHlEREREREanx0nLSyDybyZDIIYzpOsZ5fnDrwQDc2fPOcj9zcORg576CaPVSj6iIiIiIiNR4h04dAkoOwY0Kj8J63KrQM18Z9QqTYiYxKHJQJauT8lKPqIiIiIiI1Hg/Hf4JgC5hXVz2TF8vX4VQN1EQFRERERGRGu+jHR8RExHj/DZUajcFURERERERqdHm/TyPtUfWMqzNMHeXIi6iICoiIiIiIjXa9K+nAzCi3Qg3VyKuosmKRERERESkxrIV2PDx9OG6TtdxZdsr3V2OuIh6REVEREREpMY6mHaQ3PxcRrYfiTHG3eWIiyiIioiIiIhIjZV4OhGAZgHN3FyJuJKCqIiIiIiI1FjJp5MBCPMPc3Ml4koKoiIiIiIiUq1+O/UbD656kNNnTwPw8c6PmbJyCgVWQYm2yWfsQTTcP7xaa5SqpcmKRERERESk2liWRfsX22MrsDGqwyh6NOvBTUtvAuDDHR+ycfJGOoR0cLa98/M7AQhrqB7RukQ9oiIiIiIiUm2eXvs0tgIbAGk5aSzcvBCApv5NycjN4IPtHzjbHsk44tz38fKp3kKlSimIioiIiIhItVmxZwW+Xr4ApGWnsT15O5cEXsKJB08QFR7F46sf5/VNrwOQkJEAQMeQjm6rV6qGgqiIiIiIiFSZM3lnnPuLtizifwn/44+d/gjAX1f9ld0pu2nXpB0Ac4fPpUtYF+7/+n5SzqQ4g+jSMUurv3CpUgqiIiIiIiJSJVbuWYn/P/3ZcnwLyaeTuW3FbQCE+oUCkJGbwc8JP9Mu2B5ER3YYyeLrFnMm7wyP/fAYNy+7GYAWjVq45wdIlVEQFRERERGRKvHdwe8A+GrfV2w5sQWAJr5NuLv33c42Tf2bMi5qnPM4tlksTXybMH/jfABaBbWisW/jaqxaqoNmzRURERERkSoR7BcMwCPfP+I8t3/afpr4NXEeH3/gOMYY57GnhydD2wzl410fExkUyd6/7C12XeoGBVEREREREakSZ/PPOvcNhgcve9AZQmdfMZvQhqGlhsxhbYbx8a6Pad6oOQ08G1RbvVJ9FERFRERERMTlJn06ibe2vgXAhjs20DG0I418Gjmvzxo067z3Dm87HNC3oXWZgqiIiIiIiLiMZVncvOxmlu48N9Nt7+a9y/WM9sHtGdJ6CIMjB7u6PKkhFERFRERERMRldqfsZunOpQxpPYTVh1ZX6BnGGH6Y+INrC5MaRUFURERERERcZvPxzQD8++p/89jqx2gW0MzNFUlNpCAqIiIiIiIus+HoBny9fOkU2olPbv7E3eVIDaV1REVERERExGW+3v81gyIH4e3p7e5SpAZTEBUREREREZf45egv7Endw7WXXuvuUqSGUxAVERERERGXWBy3mIbeDZkQPcHdpUgNpyAqIiIiIiKVZlkWK/auYES7EcXWCxUpjYKoiIiIiIhU2pYTW0jISOC6jte5uxSpBRRERURERESk0r7e9zUA13S4xs2VSG2gICoiIiIiIpW2LmEdHUM6Eu4f7u5SpBZQEBURERERkUo5m3+WH3/7kYGtBrq7FKklFERFRERERKRSvtn/Dem56dzQ+QZ3lyK1hIKoiIiIiIhUyoc7PqSxb2OGtx3u7lKklihTEDXGHDLGbDPGxBljNhaeCzbGfGOMiS/cNqnaUkVEREREpKY5mHaQJduXMC5qHA08G7i7HKklytMjeoVlWTGWZfUqPJ4JfGdZVgfgu8JjERERERGp42wFNk5mnwTgsz2fkVeQx0MDHnJzVVKbVGZo7nXAW4X7bwF/rHw5IiIiIiJSk8WnxuM925uQuSFk5Gbw/P+ep0NwB1o3bu3u0qQWKWsQtYBVxphNxpgpheeaWpZ1HKBwq3maRURERETquIe+Pdfz+dGOjziScUS9oVJuXmVsN8CyrGPGmHDgG2PM7rK+oDC4TgFo1apVBUoUEREREZGawt/b37k/eeVkAK7vdL27ypFaqkw9opZlHSvcJgGfAH2ARGNMM4DCbdJ57l1gWVYvy7J6hYWFuaZqERERERFxi5PZJ/H18i12LqRhiJuqkdrqokHUGONvjAl07AMjgO3ACmBiYbOJwGdVVaSIiIiIiNQMCRkJDI4c7DyeHDvZjdVIbVWWHtGmwE/GmK3ABuA/lmV9BcwBrjTGxANXFh6LiIiIiEgdZVkWh04dokNwB+e550Y858aKpLa66DeilmUdAKJLOZ8KDKuKokREREREpOY5mnmUzLOZdAnrwtA2Q/n+4PcE+gS6uyyphco6WZGIiIiIiNRz25O2A9A1vCu39biNM3ln3FyR1FYKoiIiIiIiUiYLNi0gsEEgMREx+Hj54OPl4+6SpJZSEBURERERkfNKy07jy31fEuQTxNf7v+bWmFtp5NPI3WVJLacgKiIiIiIi5zX6g9GsO7LOedzrkl5urEbqijKtIyoiIiIiIvXPoVOHioVQgIGtBrqpGqlL1CMqIiIiIiIlZOdl0/nlzgBM6D6BiIAIHuj/AE0Dmrq5MqkLFERFRERERMQpMSuRpgFN2Z+2nxxbDhO6T2DhtQvx9vR2d2lShyiIioiIiIjUczm2HKZ+PpV2Tdrx2OrHWHvbWhKzEgGY1neaQqi4nIKoiIiIiEg9Ziuw8fHOj3lr61vOczuSdhB/Mh6ADsEd3FWa1GEKoiIiIiIi9dSOpB1EzY8qcf5w+mFe2vASY7qMIcg3yA2VSV2nWXNFREREROqpB7950Lkf0CCAl0a+BMBTPz5Fti2bO3ve6a7SpI5TEBURERERqaciAiKc+z6ePtzT5x5aN27tPHdZy8vcUJXUBwqiIiIiIiL11Jm8M8796IhoAF7/w+vOc37eftVek9QPCqIiIiIiIvXQscxjfLTjI/o078Pyscv56KaPAOh9Se9KPXfp0qV4e3szduxYV5QpdZSCqIiIiIhIPfSPNf8AYMPRDVzf+XpCGoYAVHpyos8//xybzcbSpUvJz893ns/NzSUkJIT333+/Us+XukFBVERERESkHsoryANgQvcJJa598f++4Nepv1bouVlZWc79PXv2OPePHTvGyZMnmTRpUoWeK3WLlm8REREREamHDqQdoEtYFxZeu7DEtZEdRlb4uSdPnnTud+vWzdkrmpqaCkBeXl6Fny11h3pERURERETqmZPZJ1l3ZB2DIwfj7ent2mefPMnQoUMBKCgocJ5PTk526XukdlMQFRERERGpB7Lzslkct5hTOadY89sasm3Z/Lnbn13+npMnTxIZGcncuXMBSE9PByAlJcXZxmazufy9UrsoiIqIiIiI1DCWVfLc2bOwfHnxa2+/Df/858Wfd89/7qHhPxty62e3smjLIk7lnAKgWWAzF1UMW7ZsYcyYMSQkJBASEkLbtm0BOHDgAFC8R7To8N2yeOyxx/jzn10fmsV9FERFRERERGoYT0+49VbYuxdycuznHnsMbrwRfvjhXLuJE+Hvf4fMzPM/a1viNl7Z+Irz+HjWcdJz7L2UQT6VmyG3qJdeeolly5YBEBoaSseOHQHYuXMnUHziIsf3omU1e/ZszbZbxyiIioiIiIjUIOnp9l7PxYuhY0fw87P3hD79tP16RkbJe1avPv/zVh86d7GBZwMSTyeSnmsPoo18Grms7uDgYOd+x44d6dSpEwDjx49n165dvP/++0RERADFh+lK/aQgKiIiIiJSg2zbVvLcjTee209JsQfVhUUmu928+fzPO5JxBIDVE1cT3TSaE1knyMjNwM/Lz6UTFRXt5ezSpQteXl706NEDgHvvvZesrCxGjx4N4JzMqLzOnj1b+UKlRlAQFRERERFxozfegP37zx0XGcFaqmXL4KOP4I47zp1btQrON/9PQkYC7Zq0Y3DrwUQERJCYlUh6TjpBvq4blgvnejmbNWvm/D509erVXHbZZXz//fcADB48GKj4ZEXlHdJbmiNHjjBmzBh+++23Sj9LKk5BVERERETETTZsgMmTYepUey/nihWQkFC8TVSUffvcc/bt11/DXXeduz5xIqxbB6NG2Sc0+t//YO5c+/OyzmbxwfYPCGgQAEBYwzCSzySTnpvu0u9DwR5Ehw8fzrFjx/Dy8gKgUaNGdO/e3dmma9euPPvss4V1T2TatGkX7eUsev3ee++tVI1ZWVlERUWxbNkyfij6sa1UOy93FyAiIiIiUl8tXWrfWhZ8/z1cd13x676+9qG6P/8MffrAkiXwyy+Qlma/fu21sGgR+PvDK6/Ar7/C3XdDXBx06ADJrewT/LQLbgdAkG8QGbkZ9iBaBT2ikZGRJc47vgsFaNmypbO39O233wbgqquuYtSoUQBs2LCBM2fOMGTIEOc9RWfYXb58eYXry8rKIjAw0Hnsit5VqTj1iIqIiIiIVKG8PNi1q/RrP/1k36akwJkz5863amXf+vvbt/36gYcHfPfduTbz5sFnn4ExMGWK/dzBg1A4SS0HDsD2pO0AvPGHNwAIbBBI1tksfk38lZaNWrri5zmlpKQQGhpa4vyf/vQnLrvsMt544w1CQ0OdkxgVvc+hb9++XHHFFcybN895LuF3XcS5ubkVqm/v3r3FjpOSkir0HHENBVERERERkSo0bRp06QKJiefOHT5sD4qbNtmPDx6EU6fOXW/WDF58sXjwBAgMhHvvtX8jOm3aufOtW9u3cXH24blg7zXdnrSdvs370sSvCXBultwTWSe4qctNJWrNzs6u0G+02WykpaWVGkQ7duzI2rVruf322wFKBFFHQLSKLJA6ffp05/727fYw/cQTTwCwb9++CtW4e/fuYscKou6lICoiIiIiUgVOn4aZM+HVV+3He/bY1wK94QaIjIR27ey9pbGx9iVZHB2D99wDs2bZA2d0dMnnvvgijBlj7wl1CAqC4GB7EHU4eRL2p+3n0pBLneeKLtfSvem5bzfBPrFQw4YNWbNmTbl/q2P4bGlB9PeMMbz77rvO448//piCgoLzTh60bds2fH19ueaaa4CSgbKs4uPjnfshISEkJiayd+9ekpOTK/Q8qRwFURERERERF8vMhNDQc2t/AnzyCcyebd8WddVV9u3Ro/btCy9A4Son5dK5s/1bUofUtDwSMhKIDDr33Wagz7lvJP1y/cgoXJT03Xff5YorrgBg69at5X63Y3htWYIowJ///Gcsy+Kll15iz549HD16lGeeeQZjDL169SIkJMTZdvv27XTp0oXOnTsDsOdi0wqfx4kTJ5z7ffv25csvv6Rjx46Eh4drWRg3UBAVEREREXGx9eshJ8e+7+lpn3SoyGePTtdfDy0LP9V87jl7z6anZ8Xe2a2bvRfU4aOvjlJgFRDZ+FwQLdoj2rZlWzp27AjAhAkTnOcrEsrKG0QdmjVrBtgnDjp48CCxsbFcffXVpKWlkZ+fD9iDaFRUFAEBAbRo0aLCPaJFez6feuqpYtcq+kypOAVREREREREXO3zYvl2zBo4fty+psmABbN0KI0bYZ7fNyIAPPoCAgHP3BVViItsuXX53Isg+1DUyKJITJ06QnZ1NYIPAYk2K9hI6FJ08qCwyMzNZuHAhUP4g6mifkpJCcnIyYWFhhIaGUlBQwKlTp3j11Vc5duwYMTExAHTo0KHC34g6guicOXPo0aMHY8eOdV77/URG55OUlMSpoh/zSoUpiIqIiIiIuNjhw/ZvOPv1g7AwiImxrxfavbt9HdBu3ewTD/n4FA+ilVF05ZSgICCtLTc1/DdLn19Ks2bNmDx5Mj5ePgB0Cu5U+kMo37ImK1euZPDgwc6lWCoaRFNTU0lJSSEsLMw5LHf+/Pnce++9DB8+nCmF0wK3b9++UkH0xhtv5OGHHwbgmWee4brC9XKKfj96PuvXr6dp06bcdtttFXq/FKcgKiIiIiLiQgkJ9hlxL7kEvL0v3r5oEC2y5Ga5OZZ8gcJvRTOS+OzvM3n9ldcB+Pzzz4luGs0t0bdwq7nV2XapYzHTQmUNohs3buTaa69ly5YtznNFv+0sC0f7pKQkkpOTCQ0NpWXhWOVHH32U/Px8brnlFvwL17Fp3749ycnJzm9by8pms3Hw4EHCwsKc51q1asWnn35KREREmYLozJkzAXsgrU6WZfHGG28UW0+1LvBydwEiIiIiInXF+vX2XlCAwrl/LsqxVijAm29W/N1Fg2hYGDRs6MmZIouTtmjRAh8vH97641uYIlPuFh2iCueG5ubn5zNgwAAaNWrEhx9+SJMmTYq1K22WW19f33LVHBwcDMC0wrVoQkNDnZMSObR2rE1TZP/w4cNERUWV+T3vvPMOOTk59OnTp8S1Sy+9tExDcx3fkfoX/RdWDfbu3cvkyZP5+OOP+fLLL6v13VVJPaIiIiIiIi7y0kvn9kvJPKVy9IiGhUHXrhV/d9GcGBgIYWHnAt19991HfHw8OTk5xdbrhHNh0MHRI3rs2DHWr1/PN998Q3BwMD///HOx3tKKDpEtytvbm/79+zuPIyIiivVaQvEg6rhW3iVX9u/fD8DEiRNLXIuIiGDt2rV8/vnn570/KyvL+T1teYYuu4Kj93fbtm3V+t6qpiAqIiIiIuICe/fCRx+dOx4woGz3OTrYyjKM90KMga++gqlToUEDCA31cV678sorOXv2LPPnz+f99993nh89ejT79u3jgQceYO7cuYwfP97ZI3rgwIFiz//b3/5GaGgov/zyC1DxZVR+b926dc79rl27YowpVmPz5s2d+xUJojabjc8//5ywsDA8PErGn2HDhgHw6aefnvcZjr9Fx44dSUtLo6CgoMzvryzHb61rkyQpiIqIiIiIuMBbb8HZs+cC5fDhZbvP0b5p08rXcNVVMH++fT8kBLp338fhw4cZNWoUQUFBHDhwgPHjxwP24LVy5UqaNGnCs88+y4wZM2jRogWJiYl4eHgwZMgQAHbs2AHA6tWrAdi8eTMHDhxg0aJFlS/4d7oWdgmPGzeONWvWcOTIkWLhsSJB9O6772br1q3n/cZyypQpREREXHDZmsTERACioqKwLKtaQ6HjPwycPn2arKysantvVVMQFRERERFxgf37oV07OHLEvkyLn1/Z7mvZEp5+Gi7QIVchL74Iy5e3c07+07RpU5KSkpxhb+TIkSXuccxiW3T4bvv27YuFwZycHPr27QuAj4+91/XJJ5/k119/rXCtX3/9NTNnziSgyMxNl19+OS1atCjWrujkRmX1+uv2yZoc65KWpmXLlhd8piP4durUqdhxdSj6rgsNH65tNFmRiIiIiIgL7N8P7dvbezbL07tpDDz0kOvrufTS4sfh4eHOGWcnTpxIgwYNStzz+1lvP/vsMxo0aFBsKOr999/v3H/44YeZMWNGsQBZESNGjGDEiBEXbefl5UVwcHC5gmDDhg2LTdpUmvDwcI4fP+48ttlseHmdi0qO9zm+Z929ezcdO3Yscw2VcfDgQQA8PDzYuXNntbyzOqhHVERERESkEj74wB4mN26ETudfntPtwsPDSUpKcq7XWZrfB1FHz+f5zJo1q9IhtLzCwsLKHEQLCgrIyckBcK5Fer5nOnpEN27ciLe3N3fffTcDBw4kJyeHlJQUPDw8GDx4MB4eHsWWrKlKOTk5vPnmm4wYMYK2bduya9euanlvdVCPqIiIiIhIJfy//3du/5573FfHxYSHh/P111+TnZ3tHIL7e4MGDeKBBx7AGEPPnj1pWti126xZs2I9hgCxsbF4V3aGpQooaxBdunQpO3bsoKCggHnz5nHfffedt23z5s05fvw4ubm5zm9i5xd+bPvuu+9y+PBhQkJCCAgIoH379tU2g+3BgwfJzs7mlltuYcmSJQqiIiIiIiJS3CuvQIcO7q7i/MLDwzl9+jTAeYNoUFAQzz77bInzO3fuZOzYsXzzzTfOcytXrqyaQi8iPDzcuabnhRRdH/V8PcAO0dHR5Ofns337dtLS0opdmzx5MmD/ZhXsM+fGx8fzww8/EB0dzcqVKzHGcMstt5T3p1yUY7bedu3a8dprrxEYGOjyd7iLhuaKiIiIiFRQ0flv/vhH99VRFuHh4c79Zs2alevexo0bs2rVKv773/86z11yySUuq608LtYjeuzYMeeMv0XvuZAePXoA8MUXXzB9+vRS29x1110AXHrppWzbto2hQ4cSEhLCpEmTSl2f1BXi4+MBaNOmDZdccomCqIiIiIiIQFycffvii1DObFftioax1q1bV+gZHWpAl6+fnx/Jycl8VHTR1iLmzp1bLDADtG3b9oLPbNWqFQCPPfZYsfNFv5F1zODbrVu3EvcHBQVdvPAKePnll+natWux/4hQVyiIioiIiIhUUK9e9m0V5RCXKhpmIiMjK/SMiIiIEs+qbt27dwdg3rx5Ja5t2rSJF154odi5sLAw2rVrd8FnljaDMIC/v79z3/Gbb7rpphLt0tPTycvLu3Dh5ZSbm8u+ffu4+eabMca49Nk1gYKoiIiIiEgl9enj7gourmh4LBqwysMYw08//cTGjRtdVVa5TZw4kV69epW6JMvMmTNLtI1zdFtXQNGhsI6/n7+/P0uXLi0xTLZJkyacOHGiwu/6vZSUFODiw4prKwVREREREZEKcEzC+vzzUE1LSlaKY2jpHXfcUannDBgwgJYtW7qipArx8PBgyJAh7N69u9j6poBzqRbHDLkNGzYs97es//rXv3jrrbeYPXs2s2bNcp5v3Lixc/+mm27i0KFDAAwfPhyA06dPu3RWW8d3sHVxWC5o1lwRERERkQr58kv79vrr3VtHWTVq1IiMjIw6MeFN8+bNyc3N5dSpUwQHB1NQUIAxhm3btjF16lTmzp2Lp6cnDz74YLmffd999xUbCvvoo4/y3nvvlRgeGxwcTH5+PqtXr+bbb78FzvViuoIjiKpHVEREREREADhyBOLjYfp0qOC8P25RF0IonAtnu3fvJjw8nJtvvpnffvuN9PR0oqOjadCgAc8991y5ZgdetmwZjzzySInA+eSTT7J///5S7/Hw8OCKK65g2bJlgIJoeSiIioiIiIiU0//9n307bJh766ivHOFswoQJJCcns2zZMrZu3QrY1wStiBtvvJF//OMf5b7PGMO1114LKIiWh4KoiIiIiEg57dkDPXvCqFHurqR+cnw3eeDAAee5PxYu5NqlS5dqr8fb25ugoCCXB1FPT0+aNGnismfWJAqiIiIiIlKr5OSAry+8/LL7aoiPh86d3ff++q5oL+Gjjz5a7FpVrel5MaGhoS6dNTc5OZmQkBA8POpmZNNkRSIiIiJSq8THQ24u3HsvtG8Pl18O778PZ8/C0KHQqVPVvj8ry/6N6KWXVu175PzCw8Px8PDg8ccfd65t6m7du3d36bI2ycnJdXbGXFAQFREREZFaxLLsQ2Idrr4a3nsPJk+2HzdoAGfOgKdn1dWwZYt9Gxtbde+QC/P29sZms2GM4a233nJ3OQAMHDiQTz75hGPHjpV7yZjSJCcn19nvQ0FDc0VERESkFjl6FPLyip8rXM4RsPeKzplTsWfbbPDqq7B27YXbbdpk3xYNxFL9HLPbDhkyhObNm/P888+zc+dOt9UzcOBAANZe7H9AZZCbm8uuXbsUREVEREREaoLSQuCuXfZtcLB9+9JLZX9eZiZMm2Z/7rx5cNdd8Je/FG/z/vv2dg5Hj4KfH9SQEaH1XmRkJAkJCUyfPp3Obvxwt0ePHjRs2JCffvqp0s/69NNPSU1N5corr3RBZTWTgqiIiIiInFdyMnh7w7ffursSuz177Nunnjp3btMmezBNToYpUyA11T6EtyxeeAFefNE+tHf+fPu5AwfO3b92Lfz5z/DAA+fuSUmB0NDK/xapW7y9venXr59LgujRo0cBGDNmTKWfVVMpiIqIiIjIee3aZR+yev/97q7ELiXFPmPusGHQqpX93K5d0Lo1eHjYJy/Ky4PTp8v2vPXr7dstW+wBtH17SE+HxET7+WPH7NvU1OI1hIS45OdIHTNw4EDi4uLYvXs3mZmZjB07liNHjpT7OUlJSXh7e9OoUaMqqLJmUBAVERERqYdOn4bx48EYeOed0ts89BAMHmzf37Gj7L2MVcnRG+ntDR99dO58t272rWN47smTF39WQQFs3Gj/OzRtaj/32GP27f799m16un1bNA+kpqpHVEo3depUPD09Wbx4Md9//z1Lly5lypQp5X6OY8Zcx3ewdZGCqIiIiEg9tHixfbZZKD7s9NAhuOUW+8yzzzxT/B4XrkxRbl99ZR8+e+jQuRAYG3tuv107+7Y8QfTHH+HECfvMuxs22P/p0cN+7ehRe/CeNct+7FiaMj9fQ3Pl/Jo1a0ZERAQnTpzAZrMBsHXr1nI/JykpqU5PVAQKoiIiIiL1Sl4e/Pvf8PXX584VXXfzkUfsPaQrV5a8t+jstA4pKfZvM6vS/2fvvsOkKLY+jv+aDJLTkpNkCQooKooSTCiKiPGq6FVUzFnwmiOiYlYMyMWsr1lUDICiJCVKzpJhlyg5bb9/nOnbM7uzYWYnbPh+nmef6q7u6allXOHsqTp18KB0xhnSW29J48dLlStbf8mStqfoPfdI555rfZEEouPHW0b4nHNsmu/RR0t169q1FSvsy5uie9hh0siRUokS9p5MzUVWatasqdTUVKUFfjDWr1+vAxlLPUtyXVepqalhn0EgCgAAgELlp5+kW26xQLNfP+mii/x1kJK/NUrGyrGSBZwHD4beX6OGVLNmfMf8+uuh5966TsmC0iFDLFCU/EA0eE1nVubMsTWh5cuHPhp6V34AACAASURBVE+SBg3ys6ySfd9XX+2f//vfuR8/ipaaNWvq+++/18CBA//X99RTT+mGG274X5bUdV1dfvnlSklJ0dy5czM9Y+XKlapfv37CxpwMBKIAAABFxNy50ocf+uctWliRn1Wr/ADUq6sSLsuZlmYBat260s6dcR+uJJse++yz0oknSmeeaX3PP5/1/V6m8oILcn72/PlSmzahfVktydu7Vypd2j/v0CHn5wOe+++/X6+++qoWLlwoSXr33Xf13nvvSZLGjRsXcu8///yjjRs3qnnz5gkfZyIRiAIAABQRXbpIgX/7SrIgtE0bC0IXLbK9MqdPtwq5lSqFZgBLl5aGDfPXlWac+uoV9Ym1pUttiuy//iWNHm2BaXa1X+rU8dd55lSsNDXV7s9ozRrpu+/889q17fvds8fOH344su8BRUvGSrclSpT43/HEiRM1e/ZsjRw5UkcccYTq1q2rKVOmhNy/dOlSSVKzZs3iP9gkIhAFAAAoItLTQ8+rVpXat7fj11+3Yj3eesx160KnxO7bJ/3zjwWrkrR1q7R/v3/dqzIbaytXWtuiRe5fM3KktePHZ31Perq0bZs/FTdY3brSKafY8U032XYxgdhAo0b5lXWBcF5++WX9/vvvmjVrlsaOHavqQZWtrrvuOh155JFauHChjj76aLVt21YLFiwIeb233UujRo0SOeyEIxAFAABIENeVvvrKnwabaF7QKUkvvWRFelq2lJo0seDNW/vZqJFUrpzty+mpUCH0WZs3h2YGg9eNxtLatdZ6RYRyo21bm6KbYcZjiJ07LRgNF4hKVpRozx6bBhwciNarl/txoGiqXr26unTpovbt26t79+6qHOY/sg0bNqhly5Zq3ry5Zs2apalBC5+9IkcUKwIAAIDWrLEiP3lZGzl6tNSnjzR0aOzGFYlDh/zjG2+Uihe3gOvWW21f0Xnz7Frwv3/fesuu//Zb6LPGj5eeeMI/96rLxlo0gWixYtLxx0t//pn1Pdu2WVulStb3lCljzypTxqoDS/EvzITCp0oW/5F16NBBrVu3liTdfvvtkqQdO3ZowIABkhSSSS2MCEQBAABysHSp9J//2LYnd94Z/XNmzvSflwy7dlmF2M8+C+1v2tTayZMtOA1O4Fx1lfTcc/4+mp6JE6198EFrN2yIz5hXr7bxlCsX2esaNJDWr/fPd+yw4kRpabYNzdat1p9VRjRYmTL+sVeVF8itrl27SsockPbs2VP9+/dXxYoVtSMw5/19bxG2pLJlyyZukElAIAoAAJAF15W++EI66STpnXes77ffLIjJuN4yN7xdGtasid0YI7Frl03H7ds3tN/bpuSPP6Tq1cNXjg0XiBYvLg0eLFWsGJ+M6KFD0pgxUufOkb+2dm37nPbutfN//Us64ggLUBs39rPSBKKIt8cee0zjxo3TWWed9b++ChUqyHEclSlTRtddd50WLVqkHTt2aEtuNsAtJAhEAQBAkbdpU+Ys5cGDtlayb9/Q9Y/z51sw8uKLkb+PlzX0ppsm2s6doXtmerwZgHv3hk7LDZahEKj277e1paVLSykp8QlEX3rJspf9+0f+2tq1rfX+zL3CRV5g+sEH1mb8vsLxAtFy5UKDUiA3SpQooW7duukwb7NbSVWDfqPRt29f7d+/X8OHD//f9i5FAYEoAAAosPbti81zWrSQmjWzDKhn3TrbXzMr2RXCyUpqqrXeesNE27VLCvq38P8EB2MNG4Z/bfHi/nFKirWtWvnn8QhE5861IPniiyN/rReILl5sbVZrOzPuIxqOF3ySDUVeXHTRRTrttNMkSc8888z/+jt37qwaNWpo2bJlmhdYqP3pp58mZYyJRCAKAAAKpLlzLUAYPTpvz1m82N8Tc/ZsP7jdvNnaSpVCCxT9+afUqZO0e3fk7+UFa1u2RDe1Ny/S023M4TKiQdsc6qGHcn6WN023ZUtra9WKTyCalhZ+n8/cqF/f2tNOs7W54QLwZ56RcrMMz/szIxBFXpx00kkaM2aMXNdVv379Qq6lpKRo3bp1mj9/vm677Tadd955SRpl4hCIAgCAAsV1pY8+krx/x733Xt6e5639lKSjjpKuvtoqqnrTZ7/5JjSI6dRJat7cpvJu325rGF1XevRRW4PYs2doZtWzb589t1o1W/u4fXvexh2pPXtsXOECsmDZ7Rhx3nnSoEF+EN6smbUpKbEtVnTggI01LS378WSnSRP/+NtvpTlzMt+TXcXcYF7gTSCKeKlVq5YmT56svXv36ogjjkj2cBKCQBQAABQo48bZVM1Fi+x81668PW/OnNC9Id97zwKU3r3tPNwOCp07SytXWtGbM86QfvhBeuABWz86dqxflCiYNy03sFuDNm2ygNRbf7prl3T99X4111jz/pzyEoh++qn05JNWgVayNbSSBaLbtsVmqvQ//1iW8vrr8xaIBlfZffxxa6+4wlovSG3QIHfPIhBFvKWkpGhTYM4+gSgAAEA+c+CAVT8NNmWKFRaK1pIlluXMiheIPv649Pbbdnz++dZ6QeOYMaGv+eSTzM/xpq56gejatRYY1a1rQerbb0uvvSYNGSK98IL0yy/RfDdZ8zKwGavfZpSbbVL++cdabz2pt2Y0uKhTNFxX6tLFAvThwy3rXKFC9M875hhr9+614HPkSKt6PH++7Znas2funuP9mZUqFf1YgOzU9hY1S//bW7SwIxAFAAAFxvTpoWsRn3/eMosvvRTd8w4ckJYt86eYhuNN37z3XunKK+24Vq3Q6qkjRvjHrVtLb72V+TneuM8+2wKbl17ypxW/+KI0a5Ydb94s3Xqr1K1bdN9TVrzgMTdVYnPiFf7x1mF26WLthx/m7bkzZ2bOJnfsGP3zpk71p3B7/7Y/4QSr9BvJv/W9QLQY/3JGnBzj/dZEUsVY/JAWAPw4AQCAAuPHH61dsMC+brzRKrd++WV0z5swwbYhOe64rO8JLuTjcZzQ6bxeMaOuXaXLLrP1ktu2hb7GC0RbtZLOPVf6/HP/WnC2dcoUvz8vmd6McsqIfvdd+AA6nAkTpP/+1w/GjzjCvq/p0/M2xvXrrb3vPr9vwIC8PfO22yzovOmm6J/h/TcQ7r8FIBZ6BtLzj3vzyIsAAlEAAFAgTJ4svfqqBXstW9pX8eJS+/aZp4TefbdUubKfZczKL7/YM047zdZ7ZnTttVm/tlq10PNHHpF+/dXf0mTJktDrXiCakuJnEKXM04IDuzdIsmA7FqZO9ffizCoQPeMM6aqrcve85s0z7+1Zt27ep+Z661hPOsnvy2sW8vjj7c/01FOjf8ahQ9YGb2EDxFKVKlW0a9cuDR48ONlDSRgCUQAAUCBcdJGtyRw6NLS/Th1bSzh8uJ1PmCA9/bRlAL2+rKxYYdNLy5WzdYMff+xfmzpVeuWVrF+bsXBNixbWetN8ly71r+3dKz34oE2LLVcudO/KSy/N+j1mzsx+/Ll16aXSmjV2HK9Zf3XqxC4QbdrUCkK9/HLexxUL3hrYww9P7jhQuJUrV06O4yR7GAlDIAoAAPK9jRulVausYmvnzqHXvCzVwIHWBmfT5s/P/rl//+1Xfi1TRmrXzr925JHZZ8CGDLGs6w8/2JYm55xj/V6mNLj67Zgxth71nnvsvGlT/1q4acHvvy+VLBmbjOj8+dLy5f55TsWKouUFonnZHzW4su+UKdINN8RmbHnVu7f0xRf+5wcg73IdiDqOU9xxnJmO44wOnDd2HGeq4zhLHMf52HEc6ogBAFCE7djhT2EM5+BB6amnrBhPpLwpth06ZL7mBZLeFE4vI9m4sb92M6OePa3gzsSJoVt4BO89mVOF1HbtLNg89VQLkEuXtn4v0AteIzpligWWd91l58HTejt2lO64w6brvv66BdyXXGKZy1jsNfrdd6HBYbwyorVq2We8ZUv0z8jtFjOJ5jhSnz6sEQViKZKM6C2Sgn8v95Sk51zXbSZpq6RcrioAAACFzf79FuDcemvW99x1l2UOo1kC5U0r9YLOYNdea0GQF6Smp9s+o8ce6+936Vm40LZjGTvW9qiU/Eym5Aef3jTbaJQpY0FpcBCZmmqBb8mSdu44VvF39GjLuj7zjPT779I11/iVaCtWzDz+aKxbZ9OBvSDKG0OsedusZBX8u272v6iQLBB1HNtHFEDhlqtA1HGcepLOlPRW4NyR1F3Sp4FbRknqE48BAgCA/M+r9DpyZPjrEyda4CVlLuKTGxs2WFurVuZrxYvbVide4Ldhg91XvrwFclu3Svv22bVXXsmcke3bN/Q8LU2aNi3yMQarVMmyuHfeaWNIS5Nq1Ai955ZbpDPPzPoZFSrkPRAdM8YqCtepIy1eLH3zTd6elx0vi+llNTO65prsM4ovvCA9+qgFzUVomRxQZOU2I/q8pLsleRM7qkna5rquV1R8jaS6MR4bAAAoICZNsrZUKdsCJHh9pCT93/9Z27ixNGeOZccisWGDrccM3rszWOXKNhV27FgLhGrVskBu504rKnTmmdJvv0lvvpnze1WvbkFsXlSqZFvNPPusZTZXr/b33sytChX8vT+jMXWqVcJdscI+l8aNpbPOiv55OckpEPW2htm7N/O1ceP8bHpWrwdQuOQYiDqOc5akVNd1g3eGCvd7qrB/pTiOc43jONMcx5mW5s2BAQAAhYpXIXbrVtv38fzzQ6+vXm3bmtx+u2Ukvf0ic8vLcmbFC0Q//NDOTz/dAjkvqBk71rZ92bfPis5ceGFk7x+pjOswZ8/OnBHNzTPykhH94gv/OKf1rrGQXSC6f79/nHEN6bZtUo8e8RsXgPwpNxnRLpLOdhznb0kfyabkPi+psuM43gSLepLCFux2XfcN13U7ua7bqUak/wcGAAAFQsbqtFOmhGY916yR6tWT2ra18zlzInv+0qW2T2VWKle2qrRLlti+nO3aZZ3VPOccadSoyN4/Ut5U4sAe9ZIiD0TzOjX3+++lk0+Wfv45dFuaeMkuEA3uyxiIen9WAIqWHANR13UHu65bz3XdRpIukjTOdd1/SRovqV/gtv6SvorbKAEAQL41YYI0ebKttfzvf60y7q5doWsx1661QNSrSrtyZe6f//fftt7y9NOzvscL8qZOtfeR/OI5webOtfWHXoXbeOnXz9ZD3n2335fIqblr10p//WVTc3v0kJo3j+45kfAC0W3bMlf7zS4Qzbj3aKtWsR8bgPwnL/uI3iPpdsdxlsrWjI6IzZAAAEBBsXOnv/5z+HCpf3+/4uyff1o7ZYoFRo0aSbVrW3GhVavCP2/VKtsCJJi3l2aXLlmP48gjrd23zw9Ew2VEGzf2j4cOta1N4mHYMJuO2r6935fIjOiYMdaecUZ0r4+GF4hedpllqIPt3u0fB/+CYurU0Gm5o0blvPcrgMIhokDUdd1fXNc9K3C83HXdY1zXbeq67vmu6+6LzxABAEB+c/CgrbmsUEF6+WXplFP8QMvLaN1/v7U//GDtwIGWJaxb1wLO1FTpvff8KbzTpkkNG2ZeX5qaam1KStbjOeII/7hp09D23HMtSH7xRavI6rnrrvgFasWKWea1alW/L5pAdOfOrAs77d5tW9esXh3av3+/ZaXr15fatInsPfMi496fweMOzojee69//thjoa9J5HgBJBfb8gIAgIhNnWpVaD0PPeQfN29uWa5ly+x8507bF9ILxJo2lX791aatjhplWdIePfyCR19+aa8pX96CrXnzrD+7qa2lSknHHCP98YfUJ7ChXOfO9pyMAVIiBW9XEunUXG8vzb17w++r+f770htv2PHrr/v9y5bZWtnnnkvsNigZM9DLl0uHH27HwYHowoXSZ59Jl1/uX/cwLRcoOvIyNRcAABRRs2aFnh93XOh5+/Z+JtMLKj3XX28Z0blz7fz7763dtMm/xwtKL7xQevppC+hy2lLl+++tME/Dhn5fMoPQjCLNiHrB55494a9PnWptxum73hrMli0je7+8KlMmNPBt2tSyosuW+YGoFzh7Y96xw/Y4/ekny6CHC7gBFE5kRAEAQMRmzJCqVZNefdWCjYyZt5o1/WxmxkDUW0PqTSn12uBd3rzAZfRoazOuGw2natX8uQ3Iq69KN9yQfdXfcLxpxFkFot4622nT7B4viPP2cK1SJfKx5oXjZC6wNGKEbefj/ULAm0K9c6e1W7bYf0c9e4ZWGAZQ+JERBQAAEZswwYoHXXBB+D05vexfmzYWVAYHorVrW+tlTMMFol6g4j3nhhtiN/ZEGzhQSk+PPNsXnBF94glpyJDQ61612SVLQosweYFo8PrURMn4C4MZM6z1frFQvboFrMGBaDLGCSD5CEQBAEBE5s+3qbPdumV9T3CGNGNGtGpVqWRJ/3zFCmu9wNR7zerVFpwOGWIFkYoaLxCdP1/6z3+kwYP9a7Nn+2tnJWnjRv84WRlRyYo0BatYMfS8fHn7Cg5EkzFOAMlHIAoAAHJt507ptNOsONDFF2d9n1f5tly5zIGo40i1avnnGzZIr70m/fijP3Vz1y7ppZcsYM1YRbeo8ALRxYv9vv377c/P264mHC8QzbiFSiJ8+60/9VrKHJiWK5c5ECUjChRNBKIAACBXpk+XmjSR1qyxiq3ZbadSvrxl8fbtszWDGQsNedNzTz/dAqbBg61wzX33Wf/OnTbltFkze8+iyAtEN2zw+5YvD72nf3//2NsuZetWW6tZIgmVQLp2lT74wD8Pnm4t2bjKl/en6m7dSiAKFFUEogAAIEczZkidOllg8dBDUr9+Ob8mJUU6dEhauTLrQDQlRTr1VGn7djs/9lhrd+601zVqFKvvoODxihUFT7sNDkSHD5eefNI/99bapqZGvlVMvAQH0ZJUvLifEd2zx74IRIGiiUAUAADk6O23rf3iC+nBB3P3Gm/6bbi9PL3go359fypn+fJSgwZ2fM890syZRTsQDZcRDQ5KDz/cAnpvD9GGDa2I1Pr1fqCfDM2b+8cbNlgWNJgXiCazqBKA5CMQBQCgiEhPt6Bl27bIXztxoq0N7dMn968JXseYcb2iNzWzaVN/W5OKFTOvKQzeE7SoCReI/vvf/rFXKbd+fb/vnXfs/uA1uIlWvrw0Zowdb9iQOSg+7DALRL39TglEgaKJQBQAgCJiwgTpuusiCyY9GzdK9epF9pqmTf3quNWqhV6rVMnaevX8QDTj9F2JjKiUeXqrx8sen366tGyZ1KqV/ZIh2RlRyc+Ar1njV87t2tXaSpUsG0ogChRtSVjGDgAAEuWPP2yrlUmT/ODk11+l1q2l33/PXRCQnm5rQyNdd+g4tjVHamrm9xk6VGrfXure3cYmZZ6+KxGIStKmTfZn6RUjkizj6AX5jmMFnSpVkj77zPqSHYgGb9tSvrz99+P9oqFWLfvFBoEoULQRiAIAUMhs2iSde66tzfv++/D3LFhg02179875eVu3SgcPZl8lNyulS1ubMSNaubJ0ww123K6dBVLDhtn5X39Zn8TUXE+tWpbp9ITLHntZZkm6/PL4jSs32rSxz3jbNvvvpnp1/1pKilVIXrXKzjP+twGgaGBqLgAAhcScOdKll0o1ali2M1wQ6m2PImVej5mV1FRro6nE6gWi2WW9KlSwqaUnn2znbdtKs2bZli75pfprMgQHohn//DIWAJL8QLRBA3+6c7IUKyY98IAdZ8xqe+tXZ84MPQdQtJARBQCgENixQzrmGGnv3uzvu+IKm6r70UdW/XbAAOnvv6VSpbJ+zWuvWRvNdE8vEA0XOGWnfXv7KsqKFbM/v337pDJlMl/LyAtE88tU1yuvtMz73XeH9nuZ9VmzLBua3X97AAovMqIAABQC111nQeinn9rejBs2SPfea0Hkc89JJ5xg9zVuLD3/vB1Pn27TPTdutGmS8+Zlfu6cOdJLL9nx8cdHPq4bb7Q20kJHMF5WNDgQLV8+/NrZ/BaIVq4svfFG5vEcfri1s2ZlDrABFB1kRAEAyOfeekuaPNnW2V15pdSyZej13bulDz6wf9T37m0ZpjJlpMcf9+8ZMEDavNkyacFrCSXphx/suhRaEEeSVq609vXXo8tcXXutPbt48chfCwtEt23zM8uSZb/D8T7XSLPPiRa8z+jatckbB4DkIhAFACAfW7LEDxIlqza7ZYtVo/VMn27t//1f1sHiYYf5VWkzZqGCn5/RunXW9uoV2bg9jkMQmhdeRjQ4EM2KlyX1igDlV45ja5hfekm65JJkjwZAshCIAgCQj02bZu0zz0jPPmtTadevDw1EvaIvHTvm/f0OHQoNHNets8Ahmoq5yLvgQLRhw+yLEHm/LCgIlYa7dLEvAEUXgSgAAPnQrl2hW3TceKNtaXLqqTbFNtiSJbZvYyTVR0uUkI480g90PZs3h1aqnTTJzr09K5FY5cpZW7q0FZXKTtWqlh1v3DjuwwKAPKNYEQAA+dDixaHnpUv7RV+2bPH7DxywCrjNmlnmMrf27pWmTpX69AntT0vzj1eskH76STrnnMjGjtiJZGquJHXoEJotB4D8ikAUAIB86JZb/OOff7a2WjVrgzOigwZJmzZJRx0V2fOLF7fCRV98IX3zjd8fHIh6wfBll0X2bMROuKq5AFAYEIgCAJDPbN0q/fabf96mjbVeIDpihG3PIlkQecIJ0quvRv9+NWqEvrdnyRJrmzaN/tnIm0gzogBQUBCIAgCQT4wdK/XtK/3yS2i/Fyh6a0YnTZL69ZMOHrTps1275m0NZ+fO0rBhdvzPP37/0qVWaZdCRclDIAqgsKJYEQAA+YDrSj172vEXX1g7apQFg8UCvzYOXgO6cKH03HMWjMYiY3nZZdLtt/uB6ObN0n//a8+OZO0pYotAFEBhRSAKAEA+MHJk5r7LL8/c16OHZU4PO0x67TXri8U2GBUqWOsFotdcI23fHlq5F4nn7f1KIAqgsGFqLgAA+cDIkVKjRv750KHh7/v8c5uKu2qVTct98kmpefO8v3/p0vblBaLeGlSyocnlTYvetSu54wCAWCMjCgBAPrB0qdSrl3TccVLt2tKZZ4a/r2JF6c03pRYt7Pzkk2M3hooV/UC0Xj1rR4yI3fMROW9vWO8XAwBQWBCIAgCQZGvWWKDRtKl09dU53x+cAe3YMXbjCA5Et2yRjj02NtlWRI9AFEBhRSAKAEAS/PabVLOmdPjh0p13Wl/79rl//fTp0urVeauWm1FwILp5s1SnTuyejeh07y6ddZY0ZEiyRwIAsUUgCgBAAqSn23rOUqWk226zdZ7Bune3qbm51aGDfcVSpUr+PqKbN0tt28b2+Yhc2bK2VywAFDYEogAAJMBXX0n33WfH4bKYl12W2PGEU7++7WHqutKmTVK1askeEQCgsKJqLgAACfD2234AetttFuSNGWPn110n9e+fvLF5Gje26b5z50q7d0utWyd7RACAwopAFACAONu9W/rxR+mGG/y+Dz+UTjtN2rNHevXV/LFNird9TLt21nbqlLShAAAKOQJRAADiaOpU6euvpf37pZ49/f5TTrG2TJn8EYRK0okn+seOI7VqlbyxAAAKNwJRAADi5McfbQuUiy+2886dpRUrpCVLkjuurDRtKr3yih27rlS6dHLHAwAovChWBABAHPz1l029laRixSwYrV7dvvKz449P9ggAAEUBgSgAAHEwerS1VatKy5dLFSokdzy51aaNtWXLJnccAIDCjUAUAIA4WLbM2rlzbX/OgqJECZtS3KBBskcCACjMCEQBAIiDuXOl7t2l2rWTPZLIeYWUAACIF4oVAQAQY+np0rx5/jRXAAAQikAUAIAYW7xY2rWLQBQAgKwQiAIAEAOua5nQb7+1/TdLl5a6dUv2qAAAyJ8IRAEAyOD116Wbb7bAMrdOO83Wg551lp137277cgIAgMwoVgQAQJD0dOm66+z40kulY47J+TX790s//RTa16RJ7McGAEBhQUYUAFDkHTok7dsnbdokbdjg92cMLjNautQynzNmhPZ37y498UTsxwkAQGFBIAoAKHIOHAg9v+ACqUwZqUYNadQov3/NGunuu0ODU8+bb0rNmknjx0svvBB67f33pYoVYz9uAAAKCwJRAECR4brSI49IlSpJkyZZ3wsvSJ9/7t/z5Zf+8XvvSU8/Ld1xh50vXy49/rhVxL3rLv++MWOsfeUVafBgqVat+H4fAAAUdKwRBQAUCVu3SgMHSh9/bOdnnSU99ph0662h9/3xh1SqlNS2rTR9uvW5rrXHHSelplrwun279M470uWXS9u2SUceKV19tb0WAABkj4woAKBIuPlmPwgdNswC0xtukKpWtb5p0yxTKlkQWqeO/9oKFSzoTE2187Q0a3v1sq9SpaRPPyUIBQAgt8iIAgAKJdeVfvlF2rLFzn/7zb/2739L8+dLb71l16+5RurYUWreXPrzT+nYY6Xdu/37ixeX+vfP/B7VqtlUXtclCAUAIBJkRAEABc7MmdLKldnfM3SoVa/t18++gu+vVEm66Sb/vFEja2++2dorr5Rq1vSvr1lj7eWXZ36fkiUJQgEAiBSBKACgQNm6VerQwYLHJk2kZcvC3/faa6HnZcuGnter5x97geill9oWLh072vRcz5Qp1p57rrR4sR03aBDtdwAAAAhEAQD52oIFftEgKTTwXLFCuu22zK/ZuNEyoM88Y4WEFi2yqba7d0v//GP3VKni39+woX9crZq17dv7fd6a0KOOkpo2lR580K+UCwAAIscaUQBAvvXHH1LnznbsVa79+2//epUq4TOi339v7Qkn2DRcrwhRcFbUcfxjLyMa7IgjpCeflEaPliZOtAypF7A+9FAU3wwAAPgfMqIAgHxp7Fg/CJUs+zlrlrWSZTovuMDPVgb77DObtnvMMdm/x4svWnXccPt+Oo40aJDUrZudB2dIAQBA3hCIAgDyHdeVbrkltK9JE5saO2GCbadSqZJUvbq0ebOUnh5675IlZvt1rAAAIABJREFUdm9w1jOcm26S1q6VimXzt2H16tZmXGMKAACiRyAKAMh30tKkefNsv89Zs0KvjR4t1ahhx9WrWxC6bZt/PT3dpu82bhybsXjPISMKAEDssEYUAJDveFutHH641K5d5uveHp9eQJqWJlWtaserV0v79oVf9xmN3r2l8eOlrl1j8zwAAEBGFACQD61aZW2DBja99vPPLSh9803r9yrfent9rl/vv/bll+01XbrEZiyOI518cvbTdwEAQGT4axUAkO94GVGvSu2550pLl0p9+tj5nj3Wtmhh7YIF/munTbMiRUcemZixAgCAyBGIAgCS7sABP7iUpL/+smxn5cqh91WrJg0YIP38s53Xry+VL2/rSR95xLZ7WbHC9voEAAD5F2tEAQBJd+GFtlfn/Pm21nPSJOnoozNXvXUc6Y03Qs/bt5deecXOf/3V1ojGqlARAACIDwJRAEBSuK60Y4dlML/4wvpGj7ZtUpYskW6+OXfP6dbNglhJ2r7dquYSiAIAkL8xNRcAkBTDhtleoMFrOefOlcaNkypWlK6/PnfPOekk/3jmTGsJRAEAyN/IiAIAkuLBB0PP27WTPvhAWrdO6tEj91VqO3b0j9PTrSUQBQAgfyMjCgBICNf1jzdvlnbtkgYOtD1Aly6VOne2IFSSrroq98+tUiW0qFHx4lK9erEZMwAAiA8CUQBA3L34omU4L77Y9vwcPdr6+/WTqle3PUJPPNG//+KLI3v+1q1W3EiSbrlFKsF8HwAA8jXHDf4VdZx16tTJnTZtWsLeDwCQP9SvL61ZE9pXoYK0dq21kgWTVatKpUtLe/dG/h6ffCLdcIPtQVquXN7HDAAAIuc4znTXdTvldB8ZUQBA3Awfbnt+HjhgRYWqVfOvPfmkH4RKNsX2999tL9BoXHCBTfMlCAUAIP9j8hIAIC5c19aAep5+2rZqefBB6d13pUsvzfyaLl0SNz4AAJA8BKIAgJgaOtQymyef7Pd16yb961+2b+iOHVLv3kkbHgAAyAdYIwoAiCnHydz30UfShRcmfiwAACCxWCMKAEi4JUvC97dsmdhxAACA/I2puQCAPEtLs6q1U6dKZcrYvqDVqkm7d9tWLe3aJXuEAAAgPyEQBQDk2dCh0jPP2PH550t169pxmTLS5Zcnb1wAACB/YmouACDPvv3WPx4wIHnjAAAABQOBKAAgT1zXpuJecYUFpKeckuwRAQCA/I5AFACQJ5s3SwcOSEcdJfXqlezRAACAgoBAFACQJ+vWWVu7dnLHAQAACg4CUQBAnsyda22dOskdBwAAKDgIRAEAYe3fL3XtKp1xRtb3bNokXX+9HTdqlJBhAQCAQiDHQNRxnDKO4/zhOM5sx3HmOY7zcKC/seM4Ux3HWeI4zseO45SK/3ABAIkyb57022/SmDHS7bdLixeHXj94UKpRQ9q+Xerf39+yBQAAICe5yYjuk9Tddd32ko6UdLrjOMdKekrSc67rNpO0VdJV8RsmACDRli/3j597TmrRQvrsM2nmTOv77Tf/epcuiR0bAAAo2HIMRF2zM3BaMvDlSuou6dNA/yhJfeIyQgBAzIwdK73xhm25kpWdO60K7rJldv7ww/61fv2kDh2kLVukP/6wvnr1pN694zdmAABQ+ORqjajjOMUdx5klKVXST5KWSdrmuu7BwC1rJIWdlOU4zjWO40xzHGdaWlpaLMYMAIjAvHnSW2/Z8bnnStdeK02cGP5e15UqV5aOPlp6+mkpJUV64AEpNTX0vmrVpClTpMaNpdWrpVq14vs9AACAwiVXgajruodc1z1SUj1Jx0hqFe62LF77huu6nVzX7VSjRo3oRwoAiNiBA1KbNtKAAdJdd0k7dli/l+3MaM4c6dAhafZsK0R01lnWX6OGdNJJofd++aV0+OHxGzsAACi8Iqqa67ruNkm/SDpWUmXHcUoELtWTtC62QwOAwu+vv6R77pHWr8/bc7zqteedJ+3d6/d/951//Mwz/vHatZb9vOgi6fvv/f5Jk6y9+mq7f8gQ/9oHH0hPPCGNHOn3sWULAACIRm6q5tZwHKdy4LispJ6SFkgaL6lf4Lb+kr6K1yABoLAaNkwaOtTfAiUaaWnSmWdKr70mff659O67/rXJk6WSJaX//Mfvq1RJWrNGWrBA+vhjqVcv/9rixVK5ctLrr0t33CFVr+5fq1NHGjxYuuIKy5BKVMoFAADRyU1GtLak8Y7j/CXpT0k/ua47WtI9km53HGeppGqSRsRvmABQOM2ZY+2PP0rp6RZITp6c+b5Nm2yblHA++cQKB738sgWd3rTbP/+UnnrKAs/HHrO1ouPG2X6fI0dKd9/tP+PgQcukzpwpNWsmFcvhb4fSpa0lIwoAAKKRm6q5f7mue5Truu1c123juu4jgf7lruse47puU9d1z3ddd1/8hwsAhcf+/dL8+VL58tLu3bZFynnnSccfbwWAdu2yVrIMZNu24Z+zZo0FoAMHWmC4dq0FtsccY9fLl7e2dWupWzebXrt3r/Ttt/4zUlOlCy+UfvnFquLmZM0aa9u3j+pbBwAARVxEa0QBALEzbpwFhNddZ+d33ulf69TJgsqGDW07FckPSjNat06qXduymHXqSO+9J512ml3r0iV0nahkU3E/+MAq4958s/UNHSr98IN0/vnS8OE5j33ECOnii6UTTsj99wsAAOBx3Ow2k4uxTp06udOmTUvY+wFAfjRunDRhgvTmmxaILlsmXXppaIYyWIkSNnVWsixqyZKh13v2tOzp5MlS/fp+tnLWrJwzln/8IXXu7J+/+KJ0003RfV8AAACO40x3XbdTTveREQWABOvRQ3r4YctkPvCA7ds5erS0datN0b3tttD7vSBUklauDL126JC0aJFUr56dP/ustdWr527abMY1nt27R/a9AAAARINAFAAS6I03/ONrr7V1nZ7KlaWyZaXbb8/69amp/nF6uq3rXLPGWkm64AJpzx5p+fLcjaduXZuW+9ln9pwjjsj99wIAABAtAlEAMTVlivT111blFaEOHvTXgc6fb2sxS5XKfF/duhZQvvee33fttdampfl9X31lAWSjRlLfvn5/mTJShQq5G5PjSHfdZa9nKxYAAJAoJZI9AACFw4ED0llnWbVWSTr22PDbkBRlf/wh7dhhe3e2apX1fY5j90hS8eJWBff8821vz+BAdPRoaydNynm7FQAAgPyEf7oAiIkxYywIvf12K64zZUroNFJI06db26VL7l9z0UXSHXfYmk/JzzSnp1s13AsusIq5AAAABQmBKICYGDtWKldOGjLEjiU/8IKZNcv2A81YICg3ypWzr7Q0ac4c295lwwapd+/YjxMAACDeCEQB5MmuXTYtd9kyqWlT21rk8MPt2gcfWNGcv/+2PSqLsiFDpLfftmJAjhPdM+rVk5YskU4+WZo5U6pVy/byBAAAKGgIRAHkSc+eUrVqVnynSRPrq11bKl3aiu0cfrj1n3566PrGwmr+/NDtVjyDB1tbtWr0z+7RQ/rmG2nLFjt/+WVbQwoAAFDQEIgCyJMpU6wAz/Llfia0WDHbisTjutZ+/XXix5dImzZZxjO7LOXevdE//8QT/eM//5TOOy/6ZwEAACQTgSiAXDlwQFq3LrTv0KHQ81NO8Y+rVPGPS5e2du7c+Iwtv/j7b2s//VTats3vP3DAP3766eif36KFf9yxY/TPAQAASDYCUQC50qWL1KCBtHq137d5c+g93bv7xxdcYO3zz0sLFkgdOlhbUG3aJP33vxZg/+c/Up8+0ocf2rYqM2bYPcGB+rPP+plgL0AdOVJq3Tr6MTRr5h9Hu84UAAAgP2AfUQA5WrTIpoJKFox2725TTB96yPquukoaNMgKFXkefNAqunbqZOctW0oTJ9rxI49IDRtK/fsn7FvIkyVLpObN/fMnnrD2q6+s/fRT6bXX/MBTkh57TGrfXurXz14vhQaS0ahQQbrxRunMM/P2HAAAgGQjIwogR7//Hno+bpw0aZJ06ql2fsklVjE3WLFifhAqSa1aSStXWpXdBx+UrrjC9sLM70aNCg1CszJwoLRmTWjxoAULLDi9+mo7z2sgKkkvvWSFnwAAAAoyAlEggZYvlyZMsKmcvXoVnDWTS5datnPlSjuvWNG/VrGi1KZNzs9o2dLaqVP9voULYzfGvDhwQPrnHzvOWPH2lVesnTDBtkzZssUfd+3aoVOVn3jCihV5SpWywHH9evszqlEjft8DAABAQUIgCiTQFVdIJ51k0zW//1565plkjyhno0fbHpiNGtm03JEjpb/+8q9v2ybVrJnzc9q2tXbECL8vP6wZPXDAMrctW1ohoZIl/TWdkgWaV11lFWuPPNLWiLZoIY0fb/fVq+cH6JIVbCpTxo43brT1pF27SrNns64TAADAQyAKJNCiRdauWGHtmDHSvn3JG09OZs+2dZ6STa2VLJhu2NDG/uGHuQ+uWrSwgOyDD/y+eASi+/bZOsr16zNfW7xYmj49tG/aNAus16+X7r7b+mbNsnbXLmnDBn9/1GAnn2wZT8kC9LPPtuPLL5dWrZJKlLDs786d0mWX2VRlAAAAGP5pBMRZero0eLD03HNSaqoFcp6NG6XHH5feekvavdvvP3AgNMuWDCtXSn37StWrS+++K73xRuj1006TLroosmfef7+1Rx0l1anjB+SxNGWKTaf9178yX2vRwl+3OnKkBdFdu9r5hx/69910k/TUUxZwS6HTbbPy5pu2T2q7djYFt2NHW0cr5W7qMgAAQFFCIArE2axZNrX19tvtvHt3ywR6gc+jj0oDBljg5E0JffRRmwp7003S/PnJGLV0zz22pvWNN6RLL5VSUvL+zJ497c/jq68sWEtLy/szM/L27Bw/3rZT6dUrNMiUrIDQoEF2fPCgTbu96CL7pUHJklZ0aNAg256ma1fprLNyft+aNf3sseRnkEuUsOAUAAAAPrZvAeLon38sMybZdh7TplkwVru2rUmcPt1fJ/rll/blujbtVZJefln65pvQNYuJcOiQ9NNPFoCee25sn92+vbU1a1qGONa2b/eP69a1dvVq+7P3PPxw6HtPmGCt40ilS/vBbMuW9lkEV8LNLS/4bNhQKlcu8tcDAAAUZmREgTh69FFr77jDitZ88YUFoZ5w23ns329VdT0rV1pmL5FmzLDqsL16xe89atSwNZRz5uTtObt3S5984u/huW2btR06+PfMnSsNG+afP/xw1s+rX98/vvVWqWzZ6MZ1xhk2pfeFF6J7PQAAQGFGIArE0YgR0oUXZl0d9/DDM/dNnmwZyeef9/fvnDkzfmMMZ+xYa3v2jN977NxpbZ8+FkR+/71935F65hn7M+7Y0abWehnRjz+2gPPnn+08JUX6/HM7rlVLeugh6a67bE1psP/7P//4nHMiH4+nZUsLgM88M/pnAAAAFFZMzQXiZNcuaetWfypqOE2bZu774Qdrq1b1s3PhKsDG04oVlrGM576XJUtau22b9O23tr6yTRvbQuX003N+/axZ9ueyebOdz5xpgZ8XiDZuLD3wgAW5AwbYGtDu3aV586TmzW3tZjhHHCHt2SOtXWsBKwAAAGKPjCgQJ17wWKdO1vc0bCiNGyddconfN3q0tVWr+vtzbtwYnzGGM2iQFSjKbtyx8MorUrVqth5z9WrrmzvXprSGM3GiVbz9+GM7v/xymzr80Uf+Pb//boFohQr+uk7Hse+ne3c7b9066yDUU6ZM+Gw1AAAAYoNAFIgTLxANXhMaTrduthWKJJUv76+ZrFrVAqLKlW0vy1j47jur3rtrl9+XmiotXGjHjz1m25ZI9r7xlJJiAeKOHdL112d/76xZ0lVXWXGniy6S3n/fMpaSjf/ss61K7dNPW4a0UqX4jh0AAAB5w9RcIE5+/dXanAJRSbrsMitclJ4unXCC9VWpYm1KSuwyouefb8V9unXztxo58kgLmtPTbT9Tz4IFsXnP7JxxhgXcW7b4fQ0ahN6zb5/tOypJN9wgjRpl1XyDNWtm+7P27WsVhr37AQAAkD+REQVi5K+/bI3jSSdJy5ZJQ4faFh4tWuT8WseRjjtOOvpov69qVWsbNrSCOrt3Rz6miRNtT0zJ9uz0nrFihX+Pl7ldv94yr7fcYudXXRX5+0WqbFmr0Nu6tU2zveIKf82nZ+JE//jqq/0iR8EaNbLXe9OJ69WL14gBAAAQCwSiQIyceaYVwpkwwTJ2O3ZIw4dLpUrl/hmlSklvvy316GHrJyXp3nttDWX79vbM3Bo/3rKrd9xh51995V/zAtGMe2nu22cFkvbssWm6idCwof25ffutZTZ37ZL27vWvL19u7ezZlr3973/tfOBA6ZFH7Dglxfb/fPBBO4+m+i4AAAASh0AUiIFVq/zMo2QZzDJlQjOcuXXllbbliFds56STpJtukpYuDd1aJCeffmqtF2yOHWtBZrt2tlZ0/37bo9TjVeutVcvGXiwJ/3eoXt3a4KzoypU2llat7Lx/f6uE++qr0uDB0gcfSOedZ9eaNLE2muwxAAAAEodAFIiBxx6zjNzs2dK551pf27Y5V2fNLW8f0quuyjx1NSveXqBTp9oWKWlpNmX1ppukxYttv9JVq/z7vUA0JSU2Y46Gt57WC+rff9/+bOvW9bd7CVaihHTxxX7QfOKJtqfoSy8lZrwAAACIDoEokAeHDtm+k2++aVm5du3sS7L+WClVyqboStKkSTnfv2aNtGiRHe/ZY0V8Nm+26b7e9ihz5viB6Ikn5m67mXjztkx5/33LenpFicLttxpO6dK2nUubNvEZHwAAAGKDQBTIg9NPl+bPt+Obb7b2xBOtPfvs2L7XffdZQOrtoxnOnj1Wgfb33+3c28bkt9+sv1o1CzSrVrXn3H67VLGirUn1NGoU23FHonFja196yYo9dexo56NGJW9MAAAAiD0CUSAPZsywNZULF0qdO1tfjx62rtGbohsrZcvatNr33/fXfe7ZI73zjm294r13tWq2xrRkSX9d5cGDlv2sWtUq9Pbu7Qerd94pdejgv0+5crEddyTKlvWPX3zR1nr27WtrWwEAAFB4sI8oEKG9e6V//1uaPt2yjE8/nXmLlox7YcZKv37Ss8/aFisVKtjXsGEWXFataus+JWnECKsw+847UvPmoWOXpAsu8LOM991nlWolK1KUbBs32vTaW26R1q3z91UFAABA4UEgCkTot9+kDz/0z2vVStx7d+xoBXo++ii0f8wY234l2LHH2nYoxx3nB6he8SRvyqtkQWz58rZmNFbFlfKiZk2pa1f/3NvGBgAAAIUHU3OBCM2c6R+ffrp9JUrJkv46ymCffpp5+qq3D+hdd0nHH2/7kz76qPWlpEinnSaNHOnf36aN1LJlfMYdqeBCTwSiAAAAhU8+yH8ABcuPP9p2IRMnWtYx0Zo2lZYssWJIf/8trV4tbd1qRZO6d7dptxdfbEWIJFurGm696pgxCR12REqWlAYMsGrEyfgzBgAAQHyREQUi8OKLtj/nvfcmL0Dy1nzWry/NmiX93//Z+fbtNk342mv9ILQge+EFK/rEGlEAAIDCh0AUyKVdu6yATsWKtu1Jsnj7lKam2vrO6tX9aykpyRlTPJQtG7+iTwAAAEguAlEglxYvtnbECKlKleSNw8vEeusogwPR1q0TPx4AAAAgUqwRBXJp9mxrg7dDSYbWraW5c/1xBBfz6ds3OWMCAAAAIkEgCuSC60qPPy7VqJH8QFQKrSobvPdn1aqJHwsAAAAQKQJRIAczZlgAunSpNGxYaOCXX/z8s+0ZCgAAABQEBKJANr7+WjrnHKlFCzs//vjkjicrPXokewQAAABA7lGsCMjC5s0WhErSokXSqadKnTold0wAAABAYUAgCoSxb5/Uq5cdly4tXXON9NFHUvHiyR0XAAAAUBgQiCbYH39IGzcmexR557rJHkF8jRtnn9Wzz0p790qvv57cLVsAAACAwoRANM727ZMGDpQeeURyHKlzZ+n662P/PgcP+sHh1q3S4MEWSMXD779LxYpJ550Xn+fnBz/9ZEWJ4vFZAQAAAEUdgWicvf22NHy49OCDft+aNbF9j88+k8qWldq1k9aulR57TBoyxILeJUvy9uwvvpA+/DC0r18/az//PG/Pjtbo0fb9SdLKlRboHzgQ2/dYuFBq2TJ/VsgFAAAACjoC0ThyXenVV6Vy5SyDePPNUsOG0qZN/j1z50qHDuXtfZ5+2jKic+faHpeTJ/vX5s/P27P79pUuuUTavt3Ov/nGn1qcjPWSBw9KvXtbxtd1pYsvtkB/5kz7c4zFlOE//5QmTvQr5QIAAACILQLROJoyxYLD556Tdu+WXnjBArvUVNv38bTTpLZtpdtui/49UlOlqVOlxx+3oHf37uwD0c8+s/fOjf37/ePJk6Vdu6Szz7bzU0+1wG/fvujHHo116/zjRYv873X7dqlECQvK82L9eumYY6R//iEQBQAAAOKFQDQOHnjAgsvevaVq1aSLLrLKq5JUs6a0c6d0yinSjz9a3+uvR/9egwZZe8opNkX18sstK7punVS/vjR7tj1//Xq7r18/uzc3pk/3j1eskDZssONGjSyglkKzu4mwebN/3KqVf3zqqdbmJRCdN08aMMA/v+SS6J8FAAAAIGsEojE2aZL06KPS889b0DRhglSxon89JSXza/bvt2xmpHbtkt59V+rZU+rY0fpGjrT1jbVrW6D28cfSdddJd99t9+fW3r22ZUmFCnYeHIi+9ppUo4Ydp6ZGPu5I7Nlja1737rXz4EA0nJIlQ8/T06Vzz5Xq1rViUb//Hv51w4dLbdpI334r3XqrNGMGGVEAAAAgXghEY2DOHKlLF2nMGAtoPFWrSq1bh96bVXBzww3SqlWZ+xcutCm8U6ZII0bYdiKeyZNtzeQdd9gaVMlax7Hj4Pf+9tvQabo5raX86y+bVvzIIzbmJUukxYvtWq1aUuPGdvzKK9k/J69efVW6/37pqaekN97IOZubnh56vmyZ9OWX/pTehx/O/Jo1ayz47NTJplE/9ZR01FGxGT8AAACAzAhEY+C55ywTet99oRVmvWAtWNu2WT8neCqs5957bQrvccdJV18t3XmnNG2aXVu61Np27cI/78wzpQYNpGOPtS1dgoPknIoYLVtmbc+eFqB9+aX0739bX61aFqj17St98EHoWtJY++cfax96SLr22szXH3nEP27RQkpLC62ge+WVofdnXNO6YoUF9Pv2WVB9661SqVIxGToAAACALBCIRsl1LUv4ww82HVayQPLgQZsKKvnTV4NVqCCdcYZ0113SVVfZVNHBg+1acEY0PV3q1cu2T8m4VvHoo6Vhw2zrkuLFw0/3lSyIXLnSD9beftu/5r1nVpYvt7ZJE6lHj9Br1atbe+mlNnX2ppuyf1ZeeNV6s9KwoX984YX25+ZlP1evtuq3wdLS/OMPPrDv74IL7Dx4zSkAAACA+CEQjdJHH1lRoNNPt/MXX/SvtWljbcYAzvPdd9LQodJbb9mU3scft/5bb/UzgLNnS99/b8f33GMVYt9/X3r5Zeu74w7bS7NWrZy3UQnOwn7+ua0fDTcNONjixVKdOrb1zBVXWCtZsF2ihB336WOBoDdlNx5WrMj+ep060mGHSeXLW9ZYsgBUkn76ydp58yzolGyq8+bN0pFHSv/6V+izvPWwAAAAAOKLQDRKb73lH597bmhW8K67LGN58825e5a3plPy11x6wd3s2Tb1tnlzy4z26RP62lq1cn5+rVq2vvLuu+31559vU2+zWyc6Z44fwDqOjefpp0PXTjqOTfv95RebuhsPs2dnDui9AFOyMS5dapnO+vWtb+VKa73sZ6NGtt/or7/a+dSp9txg3jUAAAAA8UcgGoVly6Tx46Ubb7SspVfxdsECC2gqVLDtWyJZa/jpp9b+8ou1S5ZY27Rp6H1169p0WG8tam6nkw4YYEV4HEdq2dK2kMkq23jggGURg9ee1q1r61ODg2ZJKlvWWm86ciylplpQ6WWdJcsm9+zpn6ekWKBdpozUrJllRydNsmvbtlkVXW+M3p/VjBn+6++914LSrl1jP34AAAAA4RGIRmHIEAt8Bg2yIMnLSrZsGX1Ac955lrVbtEhau9YqxTZo4E+JDVamjBUKevtty3RGygvkPv88/PXJk60A0bHH5vysLVuszWl6cDS8AP/UU+3PR/KnPf/4o01VDlaqlHTSSfZLAskC0cqV/eC5enWpUiVp9Gj/Nb16ZV3sCQAAAEB8EIjmQnq6TVUdP94q444aZQFo3bqxfZ8WLWzt5qBBdn7ZZVnfW6qUVYT1sn2RaNZMOvlk2+90587M13/6yQLL4MxjVoYMsdZbnxmJ7dszT5H17NxpW9Wcd54Fih9+KP35pz/99pRTMhdxkqQjjrCMdXq6H4h6HMcyzFOn+n1s0wIAAAAkHoFoFt591wLD+++3vSfbtZO6d7fCQgcOWOXaWGvSxNZt/v67nT/wQOzfw/Pww1YYqVUr6ZprbC/Nb7+VnnzSMqWtWkkVK+b8nFatLGMZvGVKbt1/vxUN8qYlB1u1yoJRLxNasqRtI5OTJk0sm7tuXeZAVPIDZ8kyv+EyzgAAAADiq0SyB5AfTZ0qXX65HT/2mN9/zz02DXfJkuyzldHypvj+/bdtRRLP/SxPOMEyhGvWSG++advOeNvQSOGzjVmpWDHnKrzheGs1b7pJ6tcv9NrmzdZ6W8Xklrd36yefSGPGZA5ee/a0APjRR3NX6AkAAABA7BGIhvHII5aB+/ln6euvLUs5cGDmwkGxFhwYtWgR3/cqVsy+t9697fzgwdDrkaybrFTJ33YmEl6xpA0bLPBcscIPHDdtsjbSQLRTJ5uufMcddj5tWuZ7Hn7YtqRp1CjyMQMAAADIO6bmBtmzR7ruOqvMes01VnjomWdsrWK8g1ApNBDN7dZb1vAOAAANVklEQVQveeEV/pGk+fNDr0USiFasGBqILliQ/dYwkk2bXbfOigtJFjgefbRlMSU/I1qtWu7H4d0fnF395JPM9ziOTeEFAAAAkBwEokFGjJBef1064wy/YFAiBQddkQZg0WjUSPrqK6vOO3269VWoYG1wkJqTihVtPeehQ7busnVrv+KtJO3da3urzprl982bZ623L+o771i7cKG1Tz9tbaQZUUmqU8faYsUyT/kFAAAAkHwEokGmTLEg5rvvpHr1Ev/+xYpJt98uffNN4t7z7LOlYcMsS3jEEbZ9zGef+dVpc6NhQ2tnzLCKtZIfWEq2xcwzz1hRI48XiHbrZq2XQV2wwKblLl5s59EUE6pd2z/OuO8pAAAAgOQjEA0yc6bUsWNyx/Dss9JZZyX2Pc87z/Yu/e03C+L69o3s9X37WsA3erQVP5KkP/6QUlPt2FunmZZmX5IFouXLS23bhu5BOnKk/UJAkrp0ie778QLR9PToXg8AAAAgvghEA1xXWr5cat482SNJjtq1pSpVonttpUrSYYfZOlGvAJFk2U3JihF57rvPAta5c20Kb7Fi/jTkU06xbWB++MHOhw+PbjwtW0b3OgAAAACJQSAasHmzrWWMZEoqfOXKSc8/b9Nwy5e3vh07rN2wQTrmGDt+4w2r1Dt1qk0FlvxA1FsvOmmStdFWtW3XzvYnvfHG6F4PAAAAIL4IRANWr7aWQDQ6XpVbSbr0Umt37rR2/XqpfXupRNBmQbt2SR062LFXkOiMM6ydNUuqWdMPaKMxc6b00kvRvx4AAABA/BCIBhCI5k1wkNm2rbU7dkj791vxoTp1pBo1Ql/j7RnaoIFV6W3UyPZvTU+XGjdOyLABAAAAJEGJnG8pGqpUsamh0U4HLeoOHfKPvWB++3apdGk7btjQP/a0b2/t88/btGjHsYB15UoCUQAAAKAwIxANOPFE+0J0wgWiv/7q9zVqZF9//+33lS1rbfBeoV4g2qRJnAYKAAAAIOkIRBET3j6gkgWWZcpIY8f6fY0a2d6i115rgeYpp4R/Tp061pIRBQAAAAovAlHEXJUqVmho0ya/r149W//53XfZv9bbA5RAFAAAACi8KFaEmCtXTqpQwY4rVJC2brUgNDfIiAIAAACFHxlRxJzj+IFojRpS5cq5f+3FF0sHDxKIAgAAAIUZGVHERJ8+1v71l7W1allbpUpkz2nUSLr/fgtmAQAAABROZEQREx99ZNu11Kxp53XrWlu1avLGBAAAACB/yjEj6jhOfcdxxjuOs8BxnHmO49wS6K/qOM5PjuMsCbQR5r5QmJQu7Qehkr/Ws1Kl5IwHAAAAQP6Vm6m5ByXd4bpuK0nHSrrBcZzWkgZJGuu6bjNJYwPngCSpVClrvSm6AAAAAODJcWqu67rrJa0PHO9wHGeBpLqSzpF0cuC2UZJ+kXRPXEaJAueSS6R586RHH032SAAAAADkNxGtEXUcp5GkoyRNlZQSCFLluu56x3FqZvNSFDFNm0off5zsUQAAAADIj3JdNddxnPKSPpN0q+u6/0Twumscx5nmOM60tLS0aMYIAAAAAChEchWIOo5TUhaEvu+67ueB7o2O49QOXK8tKTXca13XfcN13U6u63aqUaNGLMYMAAAAACjAclM115E0QtIC13WHBV36WlL/wHF/SV/FfngAAAAAgMImN2tEu0i6TNIcx3FmBfrulTRE0ieO41wlaZWk8+MzRAAAAABAYZKbqrm/S3KyuNwjtsMBAAAAABR2uS5WBAAAAABALBCIAgAAAAASikAUAAAAAJBQBKIAAAAAgIQiEAUAAAAAJBSBKAAAAAAgoQhEAQAAAAAJRSAKAAAAAEgoAlEAAAAAQEIRiAIAAAAAEopAFAAAAACQUASiAAAAAICEIhAFAAAAACSU47pu4t7McdIkrUzYG+Yv1SVtSvYgEDN8noULn2fhwudZuPB5Fi58noULn2fhEqvPs6HrujVyuimhgWhR5jjONNd1OyV7HIgNPs/Chc+zcOHzLFz4PAsXPs/Chc+zcEn058nUXAAAAABAQhGIAgAAAAASikA0cd5I9gAQU3yehQufZ+HC51m48HkWLnyehQufZ+GS0M+TNaIAAAAAgIQiIwoAAAAASCgC0TxwHOdtx3FSHceZG9T3kOM4ax3HmRX46hV0bbDjOEsdx1nkOM5pQf2nB/qWOo4zKNHfB0wWn+fHQZ/l347jzAr0N3IcZ0/QteFBr+noOM6cwOf5ouM4TjK+n6LMcZz6juOMdxxngeM48xzHuSXQX9VxnJ8cx1kSaKsE+p3AZ7XUcZy/HMfpEPSs/oH7lziO0z9Z31NRls3n+bTjOAsDn9kXjuNUDvTz85mPZfN58vdnAZTN58nfnwWQ4zhlHMf5w3Gc2YHP8+FAf2PHcaYG/i782HGcUoH+0oHzpYHrjYKeFfbnFomTzef5fuBzmevYv39LBvpPdhxne9DP5wNBz4r9/29d1+Uryi9JXSV1kDQ3qO8hSXeGube1pNmSSktqLGmZpOKBr2WSmkgqFbindbK/t6L4Fe7zzHD9WUkPBI4bZXPfH5KOk+RI+l7SGcn+3oral6TakjoEjitIWhz4GRwqaVCgf5CkpwLHvQKflSPpWElTA/1VJS0PtFUCx1WS/f0Vta9sPs9TJZUI9D8V9Hny85mPv7L5PPn7swB+ZfV5ZriHvz8LyFfgz7584LikpKmBvxc/kXRRoH+4pIGB4+slDQ8cXyTp48Bx2J/bZH9/Re0rm8+zV+CaI+nDoM/zZEmjwzwnLv+/JSOaB67rTpC0JZe3nyPpI9d197muu0LSUknHBL6Wuq673HXd/ZI+CtyLBMvu8wz8VvYC2Q9rlhzHqS2pouu6k137yX1HUp9YjxXZc113veu6MwLHOyQtkFRX9rM1KnDbKPmfzTmS3nHNFEmVA5/laZJ+cl13i+u6WyX9JOn0BH4rUNafp+u6P7quezBw2xRJ9bJ7Dj+f+UM2P59Z4e/PfCynz5O/PwuWwN+DOwOnJQNfrqTukj4N9Gf8+9P7e/VTST0Cn3lWP7dIoKw+T9d1vwtcc2W/AMr270/F6f+3BKLxcWNgqtjb3tQ/2f+UVwfdsybQl1U/8pcTJW10XXdJUF9jx3FmOo7zq+M4Jwb66so+Qw+fZ5IFpgkdJfstYIrruusl+8eTpJqB2/j5LCAyfJ7B/i3LoHj4+SwAwnye/P1ZgGXx88nfnwWM4zjFA1OpU2W/gF0maVvQL/6CP5v//RwGrm+XVE38fOYbGT9P13WnBl0rKekySWOCXnJcYCrv947jHBHoi8vnSSAae69JOlzSkZLWy6ajSJb6zsjNph/5y8UK/W3uekkNXNc9StLtkj5wHKei+DzzFcdxykv6TNKtruv+k92tYfr4+cxnsvo8Hcf5j6SDkt4PdPHzWQCE+Tz5+7MAy+b/t/z9WcC4rnvIdd0jZVmyYyS1CndboOXnM5/L+Hk6jtMm6PKrkia4rvtb4HyGpIau67aX9JKkLwP9cfk8CURjzHXdjYEPPF3Sm/KnIayRVD/o1nqS1mXTj3zCcZwSkvpK+tjrC0w12Rw4ni77bWFz2ecZPL2BzzNJAr/l++z/27mfEJviKIDj3+NfWaEolpS9laaULDRZSKlRJkWizMRe2UwNCysLeyKJZqFMGklhR1lYyMwsJiEbivwpNqNj8fvdMTRvZmOuefl+Nm+69za9907n/t7vd8/vANcz81Y9/K6WfzVlYO/rcfNziesQT6I0kNoLHKolRuZnF5grno6f3Wue/HT87GKZ+Ql4RNlTuLbGE36PzUwe1vNrKNuczM8lZlY89wBExBCwgbIg1FzzpSnlzcwxYGVErGeR4ulE9C9rfuRW+4GmA+socLB2F9sMbKXUZD8FttZuZKsoG71H23zPWtBuYDIzZ0qGImJDRCyvf2+hxPNlLff8GhE9dY/EYeD2v3jT/7P63V8CJjLzwqxTo0DT+fYIv2IzChyOogf4XGN5D+iNiHW1TLC3HlOLOsUzIvYAp4F9mflt1nHzcwmbJ56On11onvstOH52nRqfpgP5akoMJ4CHQF+97M/xsxlX+4AHdVGwU96qRR3iORkRxyl9MPrr4l9z/caaf0TEdspc8QOLdL9dsfAl6iQiblC6S62PiLfAELArIrZRHle/Ak4AZOaLiBgBxiklZCcz80f9P6coP26XA5cz80XLH0XMHc/MvERJtj+bLOwEhiNiGvgBDGRm0+hoELgCrKbsWbuL2raDsufhed0XAXAGOA+MRMQx4A1woJ4bo3SQmwK+AUcBMvNjRJyl3IABhmfFWe3pFM+LlI6M9+u4+SQzBzA/l7pO8ex3/OxKc8azPk1x/Ow+m4CrdbFgGTCSmXciYhy4GRHngGeUxQfq67WImKI8CT0I8+etWtUpntPAa+BxHT9vZeYwZTFhsJ7/TumUnMD0Ytxvo1YySZIkSZLUCktzJUmSJEmtciIqSZIkSWqVE1FJkiRJUquciEqSJEmSWuVEVJIkSZLUKieikiRJkqRWORGVJEmSJLXKiagkSZIkqVU/AV0Ud4lm/CLLAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<Figure size 1152x576 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plot_graphs_and_predictions(('xlk',2,1500),1500,best_predictions, etf_dict)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Now, we will try to recreate the XLK fund by plotting the composing stocks separately with regards to their percentage in the fund: (Source https://www.etf.com/XLK#overview)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [],
   "source": [
    "xlk_stocks = {'msft':0.20, 'aapl':0.17, 'v':0.06, 'ma':0.05, 'intc':0.04, 'csco':0.04, 'adbe':0.03, 'crm':0.03, 'orcl':0.03,  'acn':0.03 ,'ibm': 0.02, 'txn': 0.02, 'avgo':0.02, 'nvda':0.02, 'qcom': 0.02, 'fis':0.015, 'adp':0.01}  "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [],
   "source": [
    "stock_dict = get_stock_data(xlk_stocks.keys())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [],
   "source": [
    "strange_vals = []\n",
    "for i in range(1500):\n",
    "    temp_price = sum([stock_dict[s].iloc[-(1750- i),:]['Close'] * xlk_stocks[s] for s in xlk_stocks.keys()])\n",
    "    strange_vals.append(temp_price)\n",
    "\n",
    "strange_preds = []\n",
    "for i in range(250):\n",
    "    temp_price = sum([stock_dict[s].iloc[-(250 - i),:]['Close'] * xlk_stocks[s] for s in xlk_stocks.keys()])\n",
    "    strange_preds.append(temp_price)\n",
    "    "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.legend.Legend at 0x248db6d6eb8>"
      ]
     },
     "execution_count": 24,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1152x576 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "temp_df = etf_dict['xlk'].iloc[-(1500 + 250):-250, :]\n",
    "true_vals = etf_dict['xlk'].iloc[-250:,:]\n",
    "stranger_things = pd.DataFrame(strange_vals,index = temp_df['Close'].index,columns=['Prediction'])\n",
    "stranger_preds =  pd.DataFrame(strange_preds,index = true_vals['Close'].index,columns=['Prediction'])\n",
    "plt.figure(figsize=(16,8))\n",
    "plt.plot(stranger_things['Prediction'], label = 'Training Period Translation', color = 'purple')\n",
    "plt.plot(stranger_preds['Prediction'], label = 'Prediction Period Translation', color = 'orange')\n",
    "plt.plot(temp_df['Close'], label='True Training Period Values', color = 'b')\n",
    "plt.plot(true_vals['Close'], label='True Prediction Period Values', color = 'g')\n",
    "plt.title('Naive Translation Comparison')\n",
    "plt.legend(loc='upper left')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "As we can see from the above figure, simply plotting the stocks according to their percentage is not enough.\n",
    "Doing so, we only calculate the NAV (Net Asset Value) of the fund without taking into account the TER (Total Expense Ratio), the total amount of cash in the fund and more.\n",
    "unfortunately, in this method, we can spot a clear and growing difference between our \"translation\" and the original fund price.\n",
    "\n",
    "To account for this mismatch, we will use linear regression as sort of  \"damage control\" to make our \"translation\" closer to the true value of the fund:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.legend.Legend at 0x248db8fdf60>"
      ]
     },
     "execution_count": 25,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAA6wAAAHiCAYAAADs9kFxAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADl0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uIDMuMC4yLCBodHRwOi8vbWF0cGxvdGxpYi5vcmcvOIA7rQAAIABJREFUeJzs3Xd41FX69/H3Se8EQpUOioIQAlIU7Ao2sLuuLCpiL+vqqo+4rn0tu7p2XdcVy4qKv1VcXbEgiIINBCkqIE16C4H0TJJJzvPHd2aYJJNkkkwyKZ/XdeVy5lvvmYxcuec+5z7GWouIiIiIiIhIcxMR7gBEREREREREAlHCKiIiIiIiIs2SElYRERERERFplpSwioiIiIiISLOkhFVERERERESaJSWsIiIiIiIi0iwpYRURkTozxvzJGPNSuONoLMaYbcaY4+t57vHGmJ9DHJIEYIzpZ4zJD3ccIiLSeJSwioi0QcaYTcaY3caYRL9tVxhjvgjmfGvtQ9baK0Ic08/GmHzPT5kxxuX3/E+hvFeoGGOijDHWGNPHu81a+4W19vBGuNfBnnvlV/o5L9T3aq4qf5Fgrd1orU0KY0giItLIosIdgIiIhE0U8AfgoXAHAuCf5HkS5xnW2mqruMaYKGutuylia06CTdCMMZHW2rLattVyjbC8x8aYCABrbXlT31tERJoXVVhFRNquR4FbjTGpgXYaY54yxmw1xuQaY5YaY47x23evMWaG5/EnxpgbKp27whhzrufxYcaYz4wx+4wxvxhjflOfYD0V4AXGmKeNMfuAPxtjDjHGzDfGZBlj9hpjXjfGtPM7Z5sx5o/GmB+NMTnGmLeMMbGefZ2NMR8ZY7I9sS2o5r5HGWO+8xy303P/aM9u7zne6vB5xpiTjTGb/M4/3Bjzpef8H40xZ/jtm+G53sfGmDxjzLfGmL71fH9mGGOe8/w+CoBjqtmW6tme6am032GMMdW9xwHuE+fZv9MYs90Y87gxJsazb50x5lS/Y2M872265/lYv/dyuTHmWL9jvzLGPGCM+RYoAHpVuu9bwEHAx573+o/eqnOla9zvuUeBMea/xpg0z+891xizyBjTy+/4QcaYuZ4Y17SlarWISEuhhFVEpO1aAnwB3FrN/u+BDKAD8CbwH2NMXIDj3gQu8j4xxgwCegOzjTPk+DPPMZ09xz1vjKnvkNkxwGqgE/BXwAB/AboBg4B+wF2VzvkNMM6z7wjgYs/224CNnmt1DXCelxunEt0RGAucClzt2edNuA631iZZa9/1P9GTyH0IzPbc52bgbWPMwX6HTfLcuwOwBXiglvegJpOA+4Bk4Ntqtj0PJOC8HycClwOX+F2j8ntc2d3ACCAdGIbzntzh2fcWfp8F4DRgh7V2pTGmJ/ABcA/Oa50GzDLGpPkdfzEwFUgBtvnf1Fp7EbADOM3zXj9ezXvwW89r7gEcBnwDvOi55wY8v2djTDLOZ/PfOJ/N3wEvGmMOrea6IiISBkpYRUTatruB3xtjOlXeYa2dYa3Nsta6rbV/B2KBQH/MvwdkGGN6e57/DphlrS0GJgCbrLWveK7zA/AucH49491irf2HtbbMWltkrV1rrZ1nrS2x1u4BngCOq3TOk9baXdbaLJzkMcOzvRSnYtfLc/6XgW5orf3eWrvIE/9GnOSn8j2qMxaIAR611pZaa+cCH+MkVV7vWGuXWGtLgTf84gvIU530/znEb/d71tpvrbXlnve/wjagHCeBn2atzfO8nic4kMRDpfc4QAi/A+611mZ63vP7/c5/Ezjb74uNSZ5t4CTFH1hrP/XE9wmwAucLAK+XrbWrPe9VfYciT/fMbd0PfAqstdbO91zvPzhJNsCZnn3/9vxulwL/pf6fTRERaQRKWEVE2jBr7U84Sdy0yvuMMbcYY1Z7htJmA+1wqoyVr5GHU0H0JmG/xUm8wKm0jvZPsHASnq71DHlrpRi7GmP+zzM0NRd4NUCMu/weFwLeOaCPAJuBecaYDcaY2wLd0DhDmmcbY3Z57nF/gHtU5yCcBND6bdsMdA8ivoCstamVftb57d4a4BT/bZ2BSE8M1cUT6Br+ulV3vrV2DU4V8wxjTBLOFxbehLU3cFGlz8KROO9RsPcOxm6/x0UBnnvf397A2ErxXOh5fSIi0kwoYRURkXuAK/FLWowzX/V2nGpce2ttKpCDMwQ3kLdwkpGjgHhgvmf7VuDLSglWkrX22nrGais9/ytQDAyx1qYAU2qIseKFrM211t5sre0DnA3cbowJVDn9J/ATcLDnHnf73aNyPJXtAHp654h69AK2BxNjPQSKx3/bHqAMJ1mrLp7aXtPOWs73Dgs+B1hurd3k2b4VeKXSZyHRWvtoHe5d2/662ArMC/DZvKHWM0VEpMkoYRURaeOsteuBt4Eb/TYn48zdzASijDF348wrrM5HOEnM/cDbft1dPwQGGGMuNsZEe35GGmMGhij8ZJwGPTmeOZLVzcetwhgz0RjT35NM5uAkcoE66CZ79hd44vbOX8XTcTcLZz5oIN/gvI+3eF77icDpwP8FG2coeYYdvwM8ZIxJ8jR4uhmYUYfLvAXcbYzp6BlKflel89/Cmbt6FQeqqwCvA+cYY8YZYyI9zZtOMMb4V1hrs5vq3+u6+gA43Bgzye+zOUpzWEVEmhclrCIiAk6imej3/FOcuZZrcYZ8uqhhuKZnvuQs4GT8khTPcOHxOMOEd+AMf/0rznzYULgHGIWTUH6AMz82WIcCnwP5wNfAU9barwIcdwtwKZCHU219O0AMb3qGlZ7rv8PzvkwEzgL2Ak8Dk6y1a+sQZwWm6jqsN9Z+VgXXASXAr8CXwGs4jYeCdR/O3NMfgZXAIuBh705r7Tachl5H4peYeyqt5+AkuJk4DaZuoW5/izwE3Od5r2+qw3lVWGtzgFOAyThV412e1xGqz6aIiISAqTitRkRERERERKR5UIVVREREREREmiUlrCIiIiIiItIsKWEVERERERGRZkkJq4iIiIiIiDRLSlhFRERERESkWYoKdwCBdOzY0fbp0yfcYYiIiIiIiEiILV26dK+1tlMwxzbLhLVPnz4sWbIk3GGIiIiIiIhIiBljNgd7rIYEi4iIiIiISLOkhFVERERERESaJSWsIiIiIiIi0iw1yzmsgZSWlrJt2zZcLle4Q5FGEBcXR48ePYiOjg53KCIiIiIi0ky0mIR127ZtJCcn06dPH4wx4Q5HQshaS1ZWFtu2baNv377hDkdERERERJqJFjMk2OVykZaWpmS1FTLGkJaWpuq5iIiIiIhU0GISVkDJaium362IiIiIiFTWohLWcIuMjCQjI4PBgwczceJEsrOzwx1SUF544QX+/e9/N/g6mzZtYvDgwVW233333cydO7fB1xcREREREfHXYuawNgfx8fEsX74cgEsvvZTnnnuOO++8s9Hu53a7iYpq+K/ommuuCUE01bv//vsb9foiIiIiItI2qcJaT0cddRTbt2/3PX/00UcZOXIk6enp3HPPPb7tDzzwAIcddhjjxo3joosu4rHHHgPg+++/Jz09naOOOorbbrvNV7l89dVXueCCC5g4cSLjx4+v9toFBQWcccYZDB06lMGDB/P2228DMG3aNAYNGkR6ejq33norAPfee6/vvsuXL+fII48kPT2dc845h/379wNw/PHHc/vttzNq1CgGDBjAwoULg34vpkyZwjvvvANAnz59uOeeexg+fDhDhgxhzZo1vninTp3KyJEjGTZsGO+//34d33EREREREWlrWm6F9fjjq277zW/guuugsBBOP73q/ilTnJ+9e+H88yvu++KLoG9dVlbGvHnzuPzyywGYM2cO69atY/HixVhrOfPMM1mwYAEJCQm8++67LFu2DLfbzfDhwzniiCMAuOyyy3jxxRcZM2YM06ZNq3D9b7/9lpUrV9KhQ4dqr52ZmclBBx3E7NmzAcjJyWHfvn289957rFmzBmNMwCHLl1xyCc888wzHHXccd999N/fddx9PPvkk4FR0Fy9ezEcffcR9991X72G+HTt25IcffuD555/nscce46WXXuLBBx/kxBNP5OWXXyY7O5tRo0Zx8sknk5iYWK97iIiIiIhI66cKax0UFRWRkZFBWloa+/btY9y4cYCTsM6ZM4dhw4YxfPhw1qxZw7p16/jqq68466yziI+PJzk5mYkTJwKQnZ1NXl4eY8aMAWDSpEkV7jNu3Dg6dOhQ47WHDBnC3Llzuf3221m4cCHt2rUjJSWFuLg4rrjiCmbNmkVCQkKF6+bk5JCdnc1xxx0HOMOaFyxY4Nt/7rnnAnDEEUewadOmer9Pga4zZ84cHnnkETIyMjj++ONxuVxs2bKl3vcQEREREZHWr+VWWGuqiCYk1Ly/Y8c6VVS9vHNYc3JymDBhAs899xw33ngj1lruuOMOrr766grHP/HEEwGvY62t8T7+Vcfqrg2wdOlSPvroI+644w7Gjx/P3XffzeLFi5k3bx4zZ87k2Wef5fPPPw/69cXGxgJOcym32x30ecFcx1rLu+++y6GHHlrv64qIiIiISNuiCms9tGvXjqeffprHHnuM0tJSTjnlFF5++WXy8/MB2L59O3v27OHoo4/mf//7Hy6Xi/z8fN/w3fbt25OcnMx3330HwMyZM6u9V3XX3rFjBwkJCUyePJlbb72VH374gfz8fHJycjj99NN58sknfQ2i/ONu3769b37q66+/7qu2NrZTTjmFZ555xpesL1u2rEnuKyIiIiIiLVfLrbCG2bBhwxg6dCgzZ87k4osvZvXq1Rx11FEAJCUlMWPGDEaOHMmZZ57J0KFD6d27NyNGjKBdu3YATJ8+nSuvvJLExESOP/543/bKxo8fH/Da69ev57bbbiMiIoLo6Gj+8Y9/kJeXx1lnnYXL5cJaG7DC+9prr3HNNddQWFhIv379eOWVV+r0un/55Rd69Ojhe15dFbmyu+66i5tuuon09HSstfTp04cPP/ywTvcWEREREZG2xdQ2PDUcRowYYZcsWVJh2+rVqxk4cGCYIqq//Px8kpKSKCws5Nhjj+XFF19k+PDhvu0AjzzyCDt37uSpp54Kc7Th1VJ/xyIiIiIiEjxjzFJr7YhgjlWFtZFdddVVrFq1CpfLxaWXXsrw4cMBmD17Ng8//DBut5vevXvz6quvhjdQERERERGRZkYJayN78803A26/8MILufDCC5s4GhERERERaZUyM+GJJ2DzZnjjjXBHEzJquiQiIiIiItKSzZ4NffrAI4+A2w2lpeGOKGSUsIqIiIiIiLQ0mzbBihXO45Ej4be/hVWr4O23ITo6rKGFkhJWERERERGRlmLtWrjsMjjkELjxRmdb584wfTocdlh4Y2sESlhFRERERESau59/hosugoEDYeZMuP76VjVXtTpKWIOQlZVFRkYGGRkZdO3ale7du/uel5SUhPx+69evJyMjo8ZjNm7cyMyZM33PFy1axM033xyS+/fo0YMhQ4b4XmOorisiIiIiInXkXYZ0/nz48EO47TZnOPCTT0KPHmENrSmoS3AQ0tLSWL58OQD33nsvSUlJ3HrrrRWOsdZirSUiomm+A/AmrL/97W8BGD16NKNHjw7Z9RcuXEhqamq1+91uN1FRUdU+D/Y8EREREREJ4Ouv4S9/gXPPhSuvhCuugEmToEOHcEfWpFRhbYD169czePBgrrnmGoYPH87OnTu56qqrGDFiBIcffjj333+/79gePXpw7733MmzYMNLT01m7di0An3/+OUOHDiUjI4Phw4dTUFBQ4R4bNmzgmGOOYdiwYRxxxBEsWrQIgGnTpjF//nwyMjJ4+umnmTt3LmeffTYAe/fu5cwzzyQ9PZ0xY8bw008/AfDnP/+Zyy+/nOOOO45+/frx3HPP1en1Hn300dx5550ce+yxPPvss0yePJlbbrmFE044gT/96U813vfqq69m3LhxXHbZZfV7s0VEREREWjtrYd48OOEEOPpoWLoUvAWxuLg2l6xCC62w3ve/n1m1Izek1xx0UAr3TDy8zuetWrWKV155hRdeeAGARx55hA4dOuB2uznhhBM4//zzGTRoEABdunRh2bJlPP300zz++OO88MILPProo7z44ouMHj2a/Px84uLiKly/W7dufPbZZ8TFxbFmzRouvfRSFi1axCOPPMKzzz7Lf//7XwDmzp3rO+euu+5i9OjRfPDBB8yZM4cpU6awZMkSANauXcu8efPIzs5m4MCBXHPNNURGRlZ5Xcccc4xv+9SpU7nRM6E7NzeXBQsWADB58mQ2bNjAvHnziIiI4Nprr632vsuWLWPBggVVXp+IiIiIiHhcdRW89BIcdJAz5PfKKyEhIdxRhVWLTFibk/79+zNy5Ejf87feeovp06fjdrvZsWMHq1at8iWs5557LgBHHHEEH330EQBjx47lpptuYtKkSZx33nkkJSVVuH5xcTE33HADK1asICoqig0bNtQa01dffcXs2bMBGD9+PFOmTPFVbidMmEBMTAydO3emQ4cOZGZm0rVr1yrXqG5IsHcIstcFF1zgGwZd033POussJasiIiIiIv7Ky2HWLDjxRKd6esEFMGIETJkCsbHhjq5ZaJEJa30qoY0lMTHR93jdunU89dRTLF68mNTUVCZPnozL5fLtj/V86CIjI3G73YAzXPbMM89k9uzZjBw5ki+++AJjjO+cv//97/Ts2ZMZM2ZQWlpaJaENxHonZgd4Huv3wfePoz6vt/Lzmu5b+TwRERERkTbL7Ya33oKHH4bVq+Hvf4c//hHGjw93ZM2O5rCGUG5uLsnJyaSkpLBz504+/fTTWs/ZsGED6enp3HHHHQwbNoxffvmlwv6cnBy6deuGMYbXXnvNlwQmJyeTl5cX8JrHHnssb3haXM+dO5cePXo0ScIYrvuKiIiIiLQI1jpDfg89FC65BKKinCVq/vCHcEfWbLXICmtzNXz4cAYNGsTgwYPp168fY8eOrfWcxx57jIULFxIREUF6ejrjx49ny5Ytvv033HAD559/Pm+99RYnn3yyr0I6bNgwysrKGDp0KJdffrlv2DHA/fffz2WXXUZ6ejpJSUm88sordX4t/nNYhw0bFtQ1QnFfEREREZFWx+12klNj4P33neG/TzwBEyYcaKokAZnKwzibgxEjRlhvsx6v1atXM3DgwDBFJE1Bv2MRERERaVXy8uCFF+Cpp+DLL6F/f8jNheRkJ3lto4wxS621I4I5VhVWERERERGRUMrOhmeecTr97tsH48ZBcbGzLyUlvLG1MEpYRUREREREQqWoCAYMgMxMmDgR7rwTRo8Od1QtlhJWERERERGRhtixA957D66/HuLj4aGHnOVpMjLCHVmL16ISVmtthSVfpPVojnOpRURERERqtGkT/O1vMH06lJXBqac681SvuCLckbUaLaYlVVxcHFlZWUpsWiFrLVlZWcTFxYU7FBERERGR2u3ZA1OnwiGHOMvUTJkCa9c6yaqEVIupsPbo0YNt27aRmZkZ7lCkEcTFxdGjR49whyEiIiIiUr3CQkhIcIb9zpkD110Ht90G+ju20bSYhDU6Opq+ffuGOwwREREREWlrli6FBx+EX36BH390lqXZuBFiYsIdWavXYoYEi4iIiIiINKmvv4bTTnMaKM2fD+efDyUlzj4lq02ixVRYRUREREREmswnnzjJaseOTtff666Ddu3CHVWbo4RVRERERETEWvjoI8jNhYsugpNPhhdegMmTITEx3NG1WRoSLCIiIiIibVd5Obz7LhxxBEyYAE895SSvUVFw9dVKVsOs1oTVGNPTGDPfGLPaGPOzMeYPnu0djDGfGWPWef7bvprzL/Ucs84Yc2moX4CIiIiIiEi9zJsHQ4Y4c1MLCuDVV2HhQjAm3JGJRzAVVjdwi7V2IHAkcL0xZhAwDZhnrT0EmOd5XoExpgNwDzAaGAXcU11iKyIiIiIi0uhKSiAnx3kcGQkRETBzJqxaBZdeCtHR4Y1PKqg1YbXW7rTW/uB5nAesBroDZwGveQ57DTg7wOmnAJ9Za/dZa/cDnwGnhiJwERERERGRoBUVwbPPQv/+cPfdzrbjjoMVK+DCC53kVZqdOs1hNcb0AYYBi4Au1tqd4CS1QOcAp3QHtvo93+bZJiIiIiIi0vjy8uDRR6FvX/j976F3b2euKjhDfyPU1qc5C7pLsDEmCXgXuMlam2uCG9cd6CBbzfWvAq4C6NWrV7BhiYiIiIiIVO/WW+HFF52uv2+/DcceqzmqLUhQXycYY6JxktU3rLWzPJt3G2O6efZ3A/YEOHUb0NPveQ9gR6B7WGtftNaOsNaO6NSpU7Dxi4iIiIiIHJCZCXfeCcuXO89vvx2+/RY++8wZAqxktUUJpkuwAaYDq621j/vt+gDwdv29FHg/wOmfAuONMe09zZbGe7aJiIiIiIiEzs6dcMst0KcPPPwwfP65s71fPzjyyLCGJvUXTIV1LHAxcKIxZrnn53TgEWCcMWYdMM7zHGPMCGPMSwDW2n3AA8D3np/7PdtERERERERCY9o0Z47qU0/BeefBzz/DH/8Y7qgkBGqdw2qt/YrAc1EBTgpw/BLgCr/nLwMv1zdAERERERGRKn791ammGgOxsXDJJc7w3/79wx2ZhJBaYomIiIiISMvx008waRIcfDB89JGz7b77nMZKSlZbHSWsIiIiIiLS/C1dCueeC0OGwAcfOPNVjzgi3FFJIwt6WRsREREREZGwcLvhnHMgNxfuvhtuvBHS0sIdlTQBJawiIiIiItK8WAvz5zvDfF97zZmjOmsWHHIItGsX7uikCWlIsIiIiIiINA/WwuzZMHYsnHQSLFgAa9c6+0aMULLaBilhFRERERFp5e794GfeXbot3GHUbM8eZ07qhAmwfTs8/zxs3OjMWZU2S0OCRURERERasdU7c3n1m00AnHdEj/AGU5nb7ayZOnQodOrkdP79/e9h8mSIjg53dNIMKGEVEREREWnFVu/MBSA14UACaK2l3EJkhAlPUCUl8Prr8PDDsHs3bN4MHTrA//1feOKRZktDgkVEREREWrE8lxuAUne5b9vv31pGxv1zmj4Ylwuee86ppF5xBaSmOolramrTxyItghJWEREREZFWLLeoFIBiv4T1w5U7yXO5Wb8nr2mDWbMGbrgBevWCjz+G77+Hs8+GCKUlEpiGBIuIiIiItGJ5xU6F1V1uKSxxkxATRUpcFLkuNz9tz+XgzsmNd/PsbHj2Wdi7F558EjIyYMUKp5GSCdNwZGlR9FWGiIiIiEgrlucq9T3em1cCQFpSLACbsgooL7ehv+nevfDnP0Pv3nDXXbBpE5SVOfvS05WsStCUsIqIiIiItGK5RW7f470FxXyzfi+/7i0A4Mm56zjnH9+E9obvv+8kqg89BOPHww8/wH//C5GRob2PtAkaEiwiIiIi0orlukqJijC4yy0f/7iTfy38tcL+FVuzG36TLVsgPx8GDXLWUj3/fLj9due5SAOowioiIiIi0orlutz07ZgIUCVZbbD1651uv/37w803O9t69IDXXlOyKiGhhFVEREREpBXLyi9mQNcQN1ZatQomT4ZDD4U33oBrr4WXXgrtPURQwioiIiIi0uqUlVse+mg12/YXkplXTPfUeNrFRzf8wtbToOnDD515qbfcAr/+Ck8/DT17Nvz6IpUoYRURERERaWXW78nnxQUbuXj6Yord5XRMimHOzcfW/4LffAOnn+5UUwGuvx42b4a//Q26dg1N0CIBKGEVEREREWllCkuczsDebsCdkmPpkhJXt4tYC59/DieeCGPHwuLFUOIsi0NiIqSlhTJkkYCUsIqIiIiItDLZRaUVnnf0rLv6xhWjueCIHvTsEM/Abik1X2TqVDjpJFizBh5/3KmoTp3aWCGLBKRlbUREREREWpmcwooJa4/2CQCMPbgjYw/uCMDt76xk9c7cAweVlztrqJ58MiQnwznnwKhRcNllEFfH6qxIiKjCKiIiIiLSiqzemcuWfYUVtnVPja9yXIR/JvDmmzBkCJx7Lvz73862M890uv8qWZUwUoVVRERERBqsoNiZM5kYqz8vw8lay2lPLayyPSaqap3q5pMH8Nbirc6T3/0ODj/cSVwvuKCxwxQJmv5FEREREZEGG/7AZxS7y9n0yBnhDqVNy3W5fY8jIwy/P/FgXKXlFQ8qL4eICDqnxPHHbV/zeI+xuN+dRdTZZ1Uqu4qEnz6RIiIiItJgxe7y2g+SRrevoMT3+OFzhnDTyQOYdtphzoaCAqd5Ur9+sG0bALGTJwFQfMZEJavSLOlTKSIiIhJC/1mylWc/X4e1NtyhhEVbfd3NRVZ+se/xOcO7Ow9ycuDBB6F3b7jlFjj4YMjLAyA2ORHQFw7SfGlIsIiIiEiIlLjLue2dlQCcObQ7vdISwhxR09uyr5DeaYnhDqPNyvJUWD/8/dFER0ZAfj707w9ZWXDGGXDnnXDUUb7jY6MjASh2l4UlXpHaqMIqIiIiEiID/vyx7/HZz3/dJquNt/1nZbhDaNOufn0pAB3+N8vZkJQE994LP/wAH35YIVkFp0lWSlwU7rK291mVlkEJq4iIiEgIZOYVV3i+r6CkwnzClmz1zlz+s2RrjcckxjiVupyi0hqPk8ZTtmmz73G3G66ELVucJzfcAMOGBTznzKEHsfLeU+jZoe2NBpCWQQmriIiISAN9sGIHIx+cW2X7pqwC3GXlfLBiB2Xllp+25/DQR6tbXOX1tKcWcts7Kykvrz1uY5ogIKlo92644goyh40C4C/5yzFr1kCvXmEOTKThlLCKiIiINMDe/GL++PbyCtv+fMZAADZmFvDpz7u58a1lPD1vHRdPX8SLCzayN7/6ymtRSRm5ruZTpVy2Zb/v8d784mqPc3ma9uSqwtp0XC7nv9HR7PnsC5Zf+nsAut9wpdNYSaQVUMIqIiIi0gDXvfED7kqVx0O7JtMxKZZZP2zHXe4kcrN/3Mn+QieZe+DDVdVe76S/f0H6vXMaL+A6sNZyzvPf+J7vyHEFPM5dVk6Z5z3IVsLa+JYtg/PPh6OPBmuhQwdGXfQU18Q5w367t48Pc4AioaOEVURERKQBXKVOd9WYqAP0XeIMAAAgAElEQVR/VsVFR3Le8O4s2byP/Z55rOv35Pv2f7BiB+6ywMuIVJcUNrUpryzmkpcXV9i2I7uowvNvNuxlX0GJr7qaHBdFYUkZ2/YXNlmcbcq338KECTB8OMydC6efDiUVq/VpiTH075QUpgBFQk8Jq4iIiEgDdEiMIb1HO9b+5TRW3D2euyYMYkTv9vRKS6C0zLLWL1H1tzuv+uG1ABOeWdgY4Qbti18yWbhub4Vt/kOCt+4rZNK/FnHZK4sp9iTt0ZHOn5ZTXvm+6QJtKz74AMaMge++c9ZU3bwZ7r8fYmP55KddvsNG9e1AZIQmEkvroYRVREREpAEKi8tI8HTIbZcQzeVH98UYQ8/2TtfVNxdt8R3bJSWWM9K7AXDy37+ktFKV1b8Z00/bcxs79IC27S+kz7TZAfcVljiJ6dLN+znmb/MBWLEtx1dhPemwzgBEKWFqOGvh449hlmd5mlNOgWeecRLVP/0J2rXzHXrNjKW+x2MP7tjUkYo0KiWsIiIiIg1QUOImMSaqyvY+aYlVtj149hD+cNIhABSVlvHRjzsr7K+8DE44mi99uyGr2n3ehHXxr/sqbP9xWw7gJEsZPVPpnBLXeAG2duXl8N57MGKEM+T38ced7bGxzvI0iQc+V8/NX8+Zz37le371sf2YNEqdgaV1UcIqIiIi0gBFJWXEeyqs/nqlJZDRMxWAnh3iWfPAqZw8qAt90hI5bkAnABau28uCtZm+5WK2V5ojumlvQSNHX1VibNXk+5UpIwEoLHYDsCunYpzeCl9cdATJcVHkN6Muxy3KZ59Bejqcey7k5sL06fD551UO+8+SrSxcl8mjn/7Cym05REYYrji6L3ecPpAIVbellVHCKiIiIlIP2YUlfLk2s9oKK8BlY/sATtEsLtpJamOiInht6iiOOaQj7yzdxiUvL+bJeesA2JxVsVlRQXFZ472AasRGVf3z8ITDOtMxKYYCT4V1ZzWNoeJjokiMiSLfk9hKEEpLId8zz9nbQOmNN2D1apg6FWJiqpxy2zsruXj6gYZYZeVWnYGl1VLCKiIiIlIP1874gUtfXszu3GISYqtWWAH6dXS6tU4+sneVfZ2SYn2P/7VgI+Xllifnrq1wjMvd9Amr/7zaPmkJzLpuDAAJMVEUlXgqrLkujhvQiU2PnMHGh073HT+id3uS4qLIdylhrZXLBc8/76yX+uCDzrbTT4eVK2HSJIgK/CVIdbq1U8IqrVPd/k8QERERacOstRjjDLnckHmg+291FdYhPdox5+ZjOaRz1WVGUuKjfY+LSsv47tcsNmRWHAJcXBp46ZvG5PK754UjezG8V3sAEmIifRXWrPwSDva8pogIw8yrjmR/QQmJsVEkxUaxN7+EPtNmc9eEQVx+dN8mfw3NWkEB/POf8NhjsHMnHHUUnHiis88Y56cG/o25/HVPVcIqrZMqrCIiIiJBuPLfS+j3p49YuS2b1TtzK8xbra7CCjCgS7IvyfXXzi9hBViw1llCZqhn3itAcRgqrP739B8enBATSZEnYS0ocZPkN9f1yH5pnDbE6X6cHBdFiadK++9vNzV+wHWQ5ypl0cbqm0o1ieuvh1tugYEDnfmpX38N48YFfbqrmi8xDkpVoytpnZSwioiIiAThs1W7sRbOfPZrTntqIXFRB5LUzsl1Txa8CWu3dnFERxq+Wp8JwPBefglrGCqsxe4D9/TOuwWnGVOBZ0iws5RP4KqyfyLbPqHq/MtwOuvZr7nwxe8oaMo5tnv3wl13OXNSAaZNg2++gXnz4IQTaq2oVuadH5wUG8XFfkPNOyQ2r/daJFRqTViNMS8bY/YYY37y2/a2MWa552eTMWZ5NeduMsb86DluSSgDFxEREQmnuOgDf0adnXFQnc/3JqztE2LonBznW3d18pG9uWvCICBMFVa/JNm/whofHUlhcRkl7nJKyspJqqaqPP7wrr7H7ROiAx4TLhs9XZd35QZuGhVSu3bBbbdBnz7wl7/A3LnO9sMOc4YB16C83LJgbWbA4b/ehPWBsw/n9yce7NseqIov0hoEU2F9FTjVf4O19kJrbYa1NgN4F5hVw/kneI4dUf8wRURERJoZT4LwzEXDiIqs/6C1Xh0SaJ94ILHrlBzLb0b0ACpWO5tKhSHBfkl5h8QYduQU+daKra7C2rdjInd7Eu5Ay/2Ei3fpIIBd1XQ5Dpn/9/+cRPXxx+Hss+Gnn+D3vw/6dG/36PeWba+yz9vQKik2mo5JsVx6VG9m33h0qCIXaXZq/dfVWrsA2Bdon3G+yvkN8FaI4xIRERFp1rbvL+KMId2YOLTu1VWAPh0TATh7WHff0NmYyAiSY6OI9Qw3dpWGYw7rgSTZXXYgyTtz6EHkudy8v9xJopICrNfqNfXovgztmUp+GJblqY5/x+VGSVi3bgVvRbS8HCZPhl9+gRkz4PDDg7rE/DV7ePbzdbzw5QYAlmzeX+WYvGJnjduk2CgiIgz3nTWYww9qF5rXINIMNXQO6zHAbmvtumr2W2COMWapMeaqBt5LREREJGySKyVoe/OLSY6r/4ILR/Ruzw93jePUwV19w4NT4qMwxhAdaTAGZi3bzk/bcxoUd135J8mFJQcej+rbgYSYSN79YRtQc6MpgKTYyKadK1oL/zVtd+YUhe7Cq1bBxRc7FdUvv3S2PfoovPSSs2RNkFylZVz26vc8Nmetb+jyO0u28fOOir9/b4W1IZ89kZakoQnrRdRcXR1rrR0OnAZcb4w5troDjTFXGWOWGGOWZGZmNjAsERERkdApL7fkl1RNvip3+q0rb6Mcb4U1Oc65njGG2KgINmYWMOGZr9iV46qwPmpj8q+w+idFUZERDO2RytrdznI+iTVUWMGpADan9Vj9k+dfdufXcGSQli+HCy6AwYPhvffg5pud+alQ50ZKABsrLWn01/OGkBgbyT++2FBhu3/TJZG2oN4JqzEmCjgXeLu6Y6y1Ozz/3QO8B4yq4dgXrbUjrLUjOnXqVN+wREREREIur9iNtc6QXX9pSaHpzOptTpToV7X0X77kyIfn8dicX0Jyr9oUl5bTMSmWFyYPZ0J6twr7uqTE+h77d0kOJCk22pdcNQcFfl84rNrRwKp1SQmccgrMmQN33gmbNjnrqnbtWuup1flw5Q4ABndP4ZUpI7lwZC+OG9CJBWszKfH7EsGXsKrCKm1EQyqsJwNrrLXbAu00xiQaY5K9j4HxwE+BjhURERFpzvJczrzBdpW63nZIjA10eJ15K63+c0Yr+25jwJYiIVfsLiMhJpJTB3er0nk21W+ZmoRaGiolxUaSX+zm170FrNud1yix1oV3eHN6j3Zsyiqs0ISpVtbCF1/AlCngdkNMjFNV3bwZHngAOnZsUGx//L/lPO+ppL5zzRhOOKwzACcN7EKuy82AP3/MHbN+BCDPpQqrtC3BLGvzFvAtcKgxZpsx5nLPrt9SaTiwMeYgY8xHnqddgK+MMSuAxcBsa+0noQtdREREpGnkFjlJQmqlIcBpIVr78qDUeOBAMhJIShNV1ApLyoiPDpyMemM4sl8HhvZMDXiMV1x0JK7SMk547AvGPbGAv32yJizL9Hh5hwT37ZhIWbllf2FJ7SdZC598Ascc46yZ+umnsH69s2/MGEit+T0Ixuqducz6wWlkNe20wyqsfTshvRtREc6XBm8t3kJpWTn5xW6iI02FJYdEWrNa/+Wz1l5UzfYpAbbtAE73PN4IDG1gfCIiIiL18uO2HJ7+fB2rduTy9bQTG3StXG+FNb5yhTXUCWtptcds2x/CRkE1yCkqrXZuboQneRrZp0Ot14mLjqwwH/b5LzbQOy2BC0f2Ck2gdeStsPZJc7ozZ+YXk5ZUQ4V8506YOBGWLoWePeG552DqVIiLC2lcH/+0C4ALR/TkqmP6VdhnjGHGFaP57YvfOSFluygodpMUG6V1V6XN0FczIiIi0urkF7uZ+OxXfLZqN9uzG57oeSuffzj5EM4Y0o0TDnX6bTS06ZKXN2H1n/P55IUZdEo+kFCFtLNtDXKKSkkJweuKC1ClXRuKZkd1kOsq5YrXvmfb/kLfe9unYwIAe3KLq55QVgY//+w87tIFDjoIpk93qqrXXRfyZBXgm/V7Gdozlb+en+77QsDfkf3SePPK0QBs3V9Ivsut+avSpujTLiIiIq1OqJdTyS1yKp892yfw3O+G4yot47uNWb61VBuqfUI0KXFR3HrKob5tZw/rzszvt5CZV8yoPh1YvGkfhSVulm3JJi46giN6117lrI+colIGdw+csHrXh63cfCqQ+OiqxzRV0u313g/bmbt6D3NX7+Gmkw8B4NAuKQAs2bSPYwd4Gn2WljrrpT78MGRlOU2UkpPhgw8aPcYd2UWMObjmObC9PVXh177ZhMVpaCXSVqjCKiIiIq2O//qhAGV1abATgHdIsLfyGBcdyfGHdm7QNf0ZY1h57ylcclSfCtsvGuUMnx03qAsAWfkl/O6lRZz3j29Ddu/KahoSfMlRvZkypg9Tj+5b63UCVVj3F1Q/5LkxeL9oAHhy7jq6tYtjQJckhnRvxz8XbGRPZg48/7yzXurUqZCUBP/6FySG5ouIYGTX8H57dU+NZ0z/NH7Yst+psNayBq5Ia6KEVURERFqdwkprpjZ0PVDvkODkJh6KeVZGdzY9cgb9OjkJVFZBEI2CGmBfQQmFJWXVJlCJsVHce+bhta7BChAfoItwdlHTJqy7cl0VnvfrlEhUZAR/PS+dYnc5X81fBtdf7wz9nT3bma967rkQ0TR/Ipe4yyksKavSzCuQ9B6p5BSVBpXgirQmSlhFRESk1SmqVGF9YcGGBl1vU1YB8dGRRAcxFLYxeJs77SuoOO/Sf33OYLhKy9i0t6Da/Te+tQyAyABzKesq1m+d1utP6E9MZATZwXTmDaHdlRLWw9Pi4OGH6f7MowDsP6g3fP89fPMNnH46NHEjo5yiwMslBdIuPprSMsuWrAI6JYd+Lq1Ic6WEVURERFqdykOC//HFBqyt37BgV2kZHyzfwcmeYbnhkOZZ7zUr/0DC987SbQz488es3pkb9HVufGsZxz/2RbWJbmaekxAfXcucymB4K6xx0RHcdsphXDqmd3BLyYTQ/sKKFd0T/3QN/OlPJK/5CWMgp7AERoxo8kTVy5ewBlEx9R5TUFJWoRmXSGunhFVERERancoJK8C6PfXrUJvncuMut4zs076hYdVbaqKTrOT4DaldtDELgP+t2BHUNcrKLXNW7QYqdiP2FxMVwXEDOtW6xmow4jzrhCbEOMOHUxNicJWW4ypturVYKyfIgwb3gaVLiXjvPdrFR1dJaJtaTpETX10SVkAJq7QpSlhFRESk1SkqrZqQ1aUS6c/bcTgxJnyLKyTFRGFMxQSswDNPN9h5oa9+s+nAuQES1l05Ln7cnkPnECVD3gprvKf5UppnWPPe/ADLyYTa1q2wbh3ZhaX8buCB5DvlPzNh+HAA2ifENPmc2sqyPQlzakLt6/n6J6yh+h2JtARKWEVERKTVCVRh3ZB5YO5mfrGbq19fwvpqqq6bsyoeCwTVaKixREQYkmOj2JVzINnbmePMz8wJMulavjXb99i/wrp6Zy4/bsvhxpnL6nS92kRFeCusTsLazbPW7I5sV7XnNNiGDXDlldC/P+W33kp2YQntuwYe3twuPrrJ59RW5m0K1SWl9gTUP2H1zmkWaQu0DquIiIi0OoXFBxLWYw7pyNZ9hfzf91tZvjWbly4ZwcK1mXz6826Wbcnms5uP8zW9+fTnXby4YCNLN+/nnomDuGxsX1/ymxTGhBWcJXX8mwjt8iasQQ5r9U/O/Cusl7y82Dd3FZz1X0Oh3DNnuGs7p0FQd0/Cuj27EAjxGrKrV8NDD8Gbb0J0NFx5JTuuvYnyGWtJTYjmoXOGkFhpKZgOiTF8vmYPizZmMbpfWmjjCdKuHBcRBjol1S1hbepu1SLhpAqriIiItDreJPM3I3rw4NlDGN6rPbtyXSxYm8mWfQUs81Qb9+QVM/rhub7z7nzvR5Zu3g/Ay1//CvgNCQ7z2pcpcdEVlmkJtsJaWOImM6+YH7fn0K+jszxOfrGbPXkutu4rrJCs3nTyIZw+pFtI4j38oBRuP/UwnrwwAziQsDZKhfXtt2HWLLj5Zvj1V3juOWZsdT4Dw3q1Z9LoXpyVUTERP6RLEgAXvvhdjZe21vK3T9bw0/ackIe9I9tFl5Q4ooLoPl0xYdWyNtJ2KGEVERGRVifPVUp8dCR/O38ovdISOPqQA8NCXaXlFZZ2cZUe6Jh7WNcU3+Ot+4p46KPV7POsfRruCmu7+Gh251RM9pJjo8guqnlY67D7P2Pkg3PJLizl0jF9ACgoLuPsZ7/mmL/Nr3BsKJv5GGO49vj+pHmqh/ExkbSLj/ZVhhvku+9g4kR47z3n+c03w+bN8Nhj0M1JuPfkuuieGs8RvQM3yzqsa3JQt8ouLOX5LzYw4Zmv6t1purL/LtvOXz5cxa7cIl8Fujb+VVVVWKUt0addREREWp19hSUV5vn5L9OSU1RapfFPsbuM2KjIKsu9vLhgo+9xOOewAqTER5FXqVnS6H5pLPo1q9pz1u/Jp9jvNQ3r5TQg2ptfzI4AiaN3+ZzG0j4huv6NjqyFBQvgL3+BuXOhQwc47zxnX7t2VQ7PL3bXmNidObQ7N7+9ggGeSmt19vhVoP+7fDuLNu4j11XKU78dVu91eW96ezkA/TolBp04R/itjZsUxgZgIk1NFVYRERFpdfYXVExYO6fE+aqHTsJaUqGauG1/EQC5rlLGD+rCe9eNqXLNcCesA7pUTGy6p8Zz+EEp5LnclJUHrvxt3V9Y4flBnmG5C9ZmBjw+Lalxm/mkJsTUv9HRxRfD8cfDjz/Co486FdUpU6o9PL/YXWNVPDLCcMaQblTz1vnsyTuQ2H+2ajczv9/KRz/u8iWdDbExs4Bu7eLrfJ5/8irS2ilhFRERkVZnX0EJ7St1Un3/+rEA5HoqrBPTD+Jv56UDB+aB5haVkhIfzbBe7ZlVKWlNjAnvHNbfjOhZ4XlGz1RSPc2icqupWu7JrVhFTfXMg5y3Zg9REYYrju7LucO7E+OpFKY1cvfZ1IRo31IutSovhw8+gCLnywQmToRnnnHmqN56KyTVXBktKHbX+iVDbFQExe6a14Xdk+tUWEf17cBX6/b6ts9euZNcV8M7KmuJGpGaKWEVERGRVmX9nnxWbMshpdJw0BRPsrYzx0VhSRkdk2PonZYAwLcbsvjDzGVkF5WS4mloM7xXe1+joEmjewXVGKcx9WhfsRKXEh/la8RTXeMlb7Ll5f8aTh3clT9PGMTjv8nghMM6AU0xJDim1jm3lJXBzJkwdCicdRa88Yaz/cIL4YYbID64imRtFVaA2OgIikvLazwm0zN8/PzhPch1OUOyR/Vxuhxv21cUVCw1cddW4hVp4zQAXkRERFqVZz5fB1ScewhOhTQywrDR03CpY1IsCZ65gI9++ovvuJT4A38eGc/Iy+MHdGrMkINiTMVhoAkxUb4Ka3XzQuf/sqfa651/RA/f4ycuzGDt7nzf8j6NpV18NNkF1VQly8vh9ded5WnWroWBA2HGDCdRrYf8YnetnZ1joyIrzPENJKeolJjICMYf3oX/966zbUiPdizetI89eS4GkVLj+ZW5SitWdH87smc1R1b15hWjK6yhK9IWqMIqIiIirYo3Ibjz9IEVthtjSImLYsOefMBZ+zI+wDDfVL/lQ7w5Ykp881hG5OtpJ3KOZ53UxJhI2sU7Q3hzikpxl5Uz5+ddvk62O7KL+GFLNod0dobOpveo2JioS8qB7rQJMVFk9Ext9PjbJ8SQV+zGXeaXJHo77xoD//wnJCTAO+/ATz/B734HUfWrrxQUl5EUW/PvLZghwfkuJ/FNTYjxVdxPPKwzULWCHQz/4dvHDujk66IcjDEHd2T84V3rfE+RlkwVVhEREWlVNmQWMG5QF4YGSMCS4qJYtTMXcCqsgRLWnh0SfI8NTsaa0kzWveyeGk9HT2OkxNgDQ4KzC0v454KNPPrpL/zrkhEM7p7CzO+3AnDL+EMZe3AaUREV6xT+CWtT8XbtzS92k2pL4cUX4fnn4euvoXNnZ85qWtqBbwrqqbzceoYE11ZhjaDYXY61tkoF2yu/2E2SJ+63rjySwlI3fdKc9Wz9GzIFq8ivwnrVMf3qfL5IW6OEVURERFqN0jJnjdVxg7oE3F9WdmC+YMfkGOKiqiY03nmtcCBviolqPoPSCkqchCfBL2HNLSpl2Zb9AOzMKeL+D39mq2d+Zbv4aJL9Eu60xBiyCkpo38jDfwPxJn55TzxN6pN/hb17nc6/+/c7CWvHjjVfIEgFJe4K96tObHQk1kJpmSUmKnDCmudy+yq1vfw+GylxUWTm1b3C6k1Y//G74RXWBxaRwJrPv74iIiIiDbQ5qxB3ueXgToE7yJb6NbhJSwxcYe3R/kBSct+Zh9M7LaFKw6Nw8q7beWiXZOKinT/lit3lvsZLv+4t8CWrgC+p9Xr/hrG8etnIaiuKjSnZ7SR4+U89ByNHwldfwfz5cOihIb2PtxNxanzNXY9jo7zvX/XDgvOLS0kO0LypQ2IM+4LteOynyPOFQ1yYu06LtBSqsIqIiEirsSHTmZ/av3PghNV/7mRMVIRvvidAUmwUd08YRFz0gUTi+EM78+VtnRsp2vq5+MjejO6bxqFdk33zdUvKyikodh7vyK7Yuda/iRQ4Cbl/Ut7odu+Gzz6DyZNJSnPm0ea//iacclSj3XLGd5sBfE2pquNNWEtqaLyUX+ymc3LV4dP1XVPWW2GNj1bCKhIMJawiIiLSaqz3NFTq3ykx4H7vEiKf3XwsULHz7le3n0BqQuOuQxoKxhgO9VRZoz3L1JS6rS8Rqpx8Va6wNplt2+DRR515qm43nHQSSbFO4pffu3+j3bag2M0/F2wEqPX3GesZEl5Tp+CC4jKSOlb9k7l9QrRvyZu6cClhFakTDQkWERGRVmNDZj5dU+IqzNn05/bMYe2QWDWRaQnJamWREYbICIO7vJxCz7zNkrLyCj2LEmOauD6xaxdcfTX06+c0VJo0CVatgm7dfL+XvEZcmmXt7jzf49rm6cb6DamuTp7LTWKAIcHtE2LYX1BKTmEp5UGspborx8Uz89b5KuEJGhIsEhQlrCIiItJqbMgsoH/nwNVVgEfOG0L31PgWmZxWJzrSUFJWTqFnbuTX67PwG+lMREQTzVUt8QyPNQb+8x+48kpYvx6mT4dDDgEOdAnOc9V97mew3li0xfe4tnVlYyJrnsNaVm7JLSoNWKVOTYhhe3YRQ++fw5Nz19Ya1x2zVvL3z9ay+Nd9ABWGnotI9TQkWERERFqN/QUl9OvYvtr9Z2V056yM7hW2Pf6boXQNwxIvoRIdGeEMCS6peT3RRrN8OTz0kFNZ/fJL6NIFtm6FxKpfHCR5KpX5rtBXWFftyOX0pxdW2FZr0yVvhbU0cIV1R3YRJWXl9EmrOue3Q+KBJPajn3bxx/E1N47yVnHnrd4NELDhl4hUpQqriIiItBoFxe46D7U8d3gPxhzccpcXiY6MoLDE7Zuf22QWLYKJE2HYMPj0UzjmGGeuKgRMVuHAMNiCYjfWWj7+cSdlIYr75x05vsd3TRjEs5OG1bocUW1zWH/dWwBA345VX8/Bfo29tu8vYvnW7BrvleAZmr0jx1m7VXNYRYKjhFVERERajYKSwPMNW7PoSONb0qbJvPMOHHkkfPMNPPAAbN4MDz4I0TUPwTXGEBcdQXZRKf3/9BHXvvEDbyzaHJKQ9njWRO3fKZHzh/dgQvpBtZ5T27I23iZefQM08Urvkep7XFRaxtnPfc07S7dVOa683LJgbWaViqqGBIsEp239iy4iIiKtVlm5xVVa3uaa2URHRlSbsHrnaDaYtTBnDpSVwemnOz9PPAFXXAFJgZcQqk58dCT/W7EDb2E1M6/unXYD2ZPrIiUuinm3HB/0Ob4Ka4AhwXvyXNz/4SqSY6MCLmvTrV3Vbbf+ZwXnH9Gjwrb5v+zh8teWVDk2sqnmFou0cKqwioiISKtQ4OmSm9TGKqwxNSSsB1ezHm3Qysvh/fdh1Cg49VRnmRqAhAS46aY6J6vgJKz7Cw/EGx2ipHp3bjGd6zgXuaYuwZ/+7Mw1rW6otTEm4FDhyvIbsSOySFughFVERERahULfciFtK2H1r7BWrqi+OnVk/S88Zw5kZMDZZ8O+ffCvfzlzVRuo8tDY2uaZBmtnTlGdm2fVNCS41JPETp8yotrz/3v92FrvkRugwdRxAzoFG6JIm9e2/kUXERGRVstbYU2MbWNDgqMMebnOa4+JiqCkzEm0RvXtEHAoa41KS52fhATIzXWaKM2YARdeCFGh+bOxcsJabkPTdGlTViET0rvV6Zyami55vwQY3Tet2vMDLXdT5TqFznI/h3VNZkJ6N04b0o3uqfF1ilOkLVPCKiIiIq1CgWfoZVussBZ6kvWoSGde5I0nHcKVx/QN/iIuF7z6KjzyCEydCnffDeee6/xEhHZAXuXuuKFYjie7sIScotKghuj681VYS6vGkOsqJTk2qta5pgO7pbB6Z261+3OKSomPjuSTm46tU2wi4tCQYBEREWkVCjxDgttchTUygtIyp0oZ5UmuBnVLITmu9uofBQXw5JPQvz9cey107QpHHeXsi4gIebIKVbvjFpaUMfyBz7j3g5/rfc0t+woB6NWh6nqpNaluDuuvewt45etN5AUx/3TG5aN4uYZhw4t+3UdyXNv6EkUklJSwioiISKvgrbAmtrkK64EKYJQnwQy6A+2VV8LNN8OAATB3Lnz7LYwb1xhh+ngrrD07OMNiC0vc7Cso4dVvNtX7mlkFzrDbjsmxdTrPO+e3csL66c+7gr5GWlIsJx7WhUuP6k1qQsUvCdbuzmPlthzfkgLLO+4AACAASURBVDsiUndKWEVERKRV2O+ZK9g+ISbMkTQt/y673iHB1ear+/bBvffCr786z6dNg4ULYf58OOkkMI2/1Ip3Duv4QV3pk5bgq4w3RI6n63BqEHNK/UVFRhAZYao0XeqSUrfEF5zfQ0mlxHfdbmcd18lH9qrz9UTEoYRVREREWoVsb9KSWLekpaXzT1i9jyMqZ6y7d8Ptt0Pv3nDfffDJJ8729HQ4+uimChUAb2QdEmNIiIki1xV4SZ7a5BSWsnBdJuDMYQVIrceXFbFREbhKyyusB+tdl7VPWvBDjGOiqiasm/cVADDttIF1jktEHEpYRUREpEXLKSzl+Efns2BdJlERhuQ2uA6rl3cOa4V09bbboE8feOwxmDABVq505quGye5cJzEc1C2FpNgo9ubXb7js1TOWcPH0xeS6Ssn2dPRNqcdc0dioCKZ/9SsjH5xLVn4xV7y2hBXbcgB477ral63xiomKwF1uKfdbt/Wb9VmkJca0ubWBRUJJ//eIiIhIi/btxiw2ZRWyKauQjkmxmCYY1tqc+M9hHdAlmXV78mlfmAt0djYWFMBFFznDfwcMCE+Qfv7fqYeyYO1ejj+0E28t3sKGzfm+fdbaoH9/Kz1JZb7LTXZhKclxUURF1r0W4yxt4yS8/1uxg7mrd/v2VV6Cpybe9WS/2ZBF77QEjIGv1u/lkqN61zkmETlACauIiIi0aGV+Fa32CW1rODBUHBJ8z8AYzv9yAUNHngVffw2jR8NzzzXJ3NRgDevVnmG92gPOsGBvwySAHTmuoNcoLfWsN5vrKiW7sKRKw6NgRfkl/D/vOLA8TYQ5sOxNMLyV7snTFwEw67oxAJxwWOd6xSUiDg0JFhERkRatoOTA0iPtE9tWwyWARL/hpklHjeSEN56FP/zBma8KzSpZrazy7+uTn4Lvzutdymd3bjFfrs3ksK4p9Yqh0G8d2P8s3eZ7nBATVadqfeXkNivf07k4se4NnETkACWsIiIi0uSem7+e8/7xTUiutc+vQtcjyOpca5LkN2o1+pZbYNMm+PvfnTVVm7kOlZokZQa5/Mv/VuzwPV65NZv9haWclXFQvWI49pCOQNWlgOoyHBgODAn28s7NTUtqe1+iiIRSrQmrMeZlY8weY8xPftvuNcZsN8Ys///s3Xd8VfX9x/HXybzZG8iAsPcGNyiiOKni3nvWalur/YnV2lq3Vuu27lVr3W1dOLDKUhFkiGxCAllk79zk5t7z++Pce5OQAST3Zr6fjwePnHXP+QYSks/9fL+fj/vPSW289gTDMLYYhrHdMIyFvhy4iIiI9F4PfbaF1VmlFFTYO32vpgHrkAOo6tqrLVkC11wDLheRkTbv4aB77oKkpG4c2IEJDW7+q2h+eS1b8iv3+bob3lzj3c4srgEgKbJjmcx7T5/EMxdM5+ub5zQ7Ht7JgPWrzQWANe1ZRDpufzKsrwAntHL8b6ZpTnX/+WTvk4ZhBAJPAScC44HzDMMY35nBioiISN/y467STt+jaVZuaEJEp+/XY5kmfP45HHkkHHUU/PvfkJFBVJPKuL2t4NR091pWj3+vzeX4R5fs83VNk6FZxVbrmI4GhuEhQZw4KZnB8eE8fcF0b6a26dro/RG8V8GnLzbuISo0CFvwgQW+ItLcPgNW0zSXACUduPfBwHbTNDNM06wH/gWc2oH7iIiISB9imo2BQEVtQztX7p+88lrv9kHD4jt9vx5p926rgNLxx8POnfD449bU35Eje3XLlImpMSS0Emh6Ciq1JTI0iAsPHQLAqizrTQ9frF8+aVIyR422MtR2h3MfVzfXtL3QQUPjfDYmkf6uM2tYrzcMY717ynBcK+dTgd1N9rPdx0RERKQfK6pqnMLbtGBSTX3DAWe1APLL7YweGMnb1xy23xVmewWnE7ZutbYHDYLoaHjuOdi+HW64AcKszzXa1rsrI7eWGW06zbs19gZXs2JTALFhvvl7CA+x7psad2DTyz2vAzjY/cZJR76eRaS5jgaszwAjgKlAHvBwK9e0Nielze9awzCuNgxjlWEYqwoLCzs4LBEREenJHvliK7e8t967f+eHG9mcX4Fpmoy/4zMWNjm3Lx+vz+OI+78ip6yWo0YneYOEXs/hgNdegwkTrOm/tbUQHAxffglXXQWhzddqRtp6b4YVYP5kawpu0zWg7RVfMk2T+gaXu39qo470YG3NhBSr2vDv5h1Yz9rosMZ/h6mDrVxO/T4yxSKybx36zjZNc49pmk7TNF3A81jTf/eWDQxusp8G5LZyneeez5mmOdM0zZlJvahYgIiIiOyfPRV2Hl+8zVuMxuO/a3O9v9g3bSvSnvIaBze/s46cslocTpO0A8yG9Uh1dfDsszB6NFxyCdhs8OSTLQLUvUX18oD118eMZPnCuQxPbFx/7Kmw25q6ButrxRYcwFx3j9OHz5ris/EMjg8n8/6TvVOD91fTTPeUwTEAHDy0j7yJItKNOvQ/nGEYyaZp5rl3TwM2tHLZD8AowzCGATnAucD5HRqliIiI9HqH3Lu4zXO1TXphFlTYqa53MswdwLz9w26mDI5lzKAoAKrrGpjyl8+bvf6kScl+GHEXW74crr3WWqv6xBNw8sn71UO1twfrhmGQGhvWLMNaXNX2lGDP2lJbUCBPnT+d4uq6HvF3EN1kSvKAKBsfXj+LEQP6cBEwkS6yz4DVMIw3gTlAomEY2cCfgDmGYUzFmuKbCVzjvjYFeME0zZNM02wwDON64DMgEHjJNM2f/fJZiIiISI/W3lo+w4DaJgVuzn3+OzIKq1l2y9GkxYXzf+5pwpn3nwzA/7YUtLhHUlTHWpp0q4oKePppawrwH/8IRx8NK1bAoYfuV6DqERkaxPjk6F4//bRplV17Q9sFjzwZ1tDgAMJCAkkL6f5gFSB6r0z3pLSYbhqJSN+yz4DVNM3zWjn8YhvX5gInNdn/BGjR8kZERET6F88Uz7sWTKS8pp7Zo5I49anlALjM5hnWjEKrTcnqrFIGRjf2GLU7nNiCA9lVUtOFI/eDkhKryu9jj0FZGZx+utWyxjDgsMM6dMsPb5jVavGQ3qRpld0dBdXsKq5pta9u0wxrT+KrNbQi0py+s0RERMTvzvr7twCkxYZx/dxRTBkc6z1XZW+gpr5lRi2v3N6stUhFrQOAnNLaFtf2Gm+9BenpcOedMGcO/PADvPfeAWVUWxMYYBAQ0LtD1uAmU4JfWr6TIx/6n3f/xWU7ySyy3sjILLbesAgN1q+xIv2BvtNFRETEr55bssObFW0tY1Ze6/AGpi9fehC3nTQOsNrV2B2N01xLa9wBa1kvC1izs62eqQATJ8L8+bB+PXzwAcyc2a1D60lC2shQVtU1cNdHG5nz1695YvE2LnlpJdDzMqwe6a18jYtIxylgFREREb+695PNAJw4cRAjkiK9x4e6f7Ffsq2QYnffzUhbEFcdOZyRAyLdAWtjhvVvX1g9SfdUtF1BtkfJyIBrroERI2DhQuvYhAnw5pswaVL3jq0HCglqPUNcXdfYq/dh99cA9MwM6893Hs9nvz2yu4ch0qf0vO90ERER6RMcThcL3OtUAWakxzU7//mNR/HnX4ynrMbBtzuKAQgLtrJmg6Jt5FfYmxVjWvRzPlV1Dd6pwT3Wli1WW5rRo+GVV+Dyy+H++7t7VD1ecBsZ1kp7Q6vHbcE9L8MaERrUI8cl0pspYBURERG/2FlUzdrdZd79Q4cnNDsfEhTAseMHArA5vwKAsBB3wBpjY0+FvVkxJoD12WUtAtanzp/u87F3yosvwjvvwK9/bWVZn3kGhg7t7lH1eG1NCW6aYW2qwdl25WkR6TsUsIqIiIhfeCoDA7z3y8OZmNqyzUdyTBjBgQZb91QBzTOsBZV1VNc3D1a25ldS2SSAOXlyMidP7uYerCtXwqmnwmefWfsLF1prVh95BFJTu3VovUnToktNVbUSsJ48OZmpTQp3iUjfpYBVRERE/KKoylqX+vmNR7aYDuwRGGCQFhdOiXsNa7g7wzowxobTZZJdYhVYuvm40QD8+cONzV4f1p3TL5cuheOPh0MOgWXLoMDdHzY+HgYM6L5x9VLtFV3a21PnT/dm40Wkb1PAKiIiIn5R7M6wJkaGtnvd4PjGqqqe9X+psVb/1W8zrLWtx4wbSGiTDJyng4utuwrvnHceHHkkrF0LDz5oZVQvuqh7xtJHhLSVYW1jDauI9A8KWEVERMQviqvqCTAgNiy43evS3QFrSGCAN2A9fEQiyTE2PliTA1iBbJSt8T4u9/LFiNAgP4y8FaYJH38M9VYmmBNOgMceg5074fe/h6iorhlHH2bbx5TgaUM0BVikP1LAKiIiIn6RW1bLgCgbAQGttyvxGOIOWJ1mYxEdW3AgaXFh3v2w4ECiwxqD04OGWlOMo/wdsDqd8PbbMHWq1T/17bet45dcYhVVClfPTV+JtLX8tzRN0xuw/m7e6K4ekoj0AApYRURExKdq6htYs6uUjXkVjEved+ZxUIw1/dfpal71NbJJMBoWEkiQO/BdeOJY0hMiWlzjU04nvPaa1Tf1nHOgrs7aP/dc/zxPiLa1zMQ7nFbAGhxoEBceAsApU1K6emgi0o26aB6NiIiI9Bf3fbKZ17/LAuCYcfsuPhQfEdLqcc9038AAg2hbENV1VoubhIgQNja4AIhsJcjpFNMEw4CAAKvKb2iolVU9/XQIVJEff4pqNWB1UWVvICI0iAkp0Txx3jTmuVshiUj/oAyriIiI+NTu0hrv9sHDEtq50uLJnO3Nkz2NCQvGMAxS3IWYpg6Opa7BCl7DfVUptqYGHn8cJk6EsjIraF20yCqqdNZZCla7QFSTKcGeisH1DS6q6xqIDA3CMAx+MSXFu85ZRPoHBawiIiLiU57MaJQtiEOGxe/z+oTI9gNWT9Gmv541hXevPYxRA6Ooc2dYQ9so1LPfKivhgQdg6FD4zW8gKQkKC61zgwZZgat0Cc8a1ilpMfzplPGAlWGtdAesItI/6btfREREfKqwoo6Dh8bz/CUz9ysbFhve+rTepoEvQHpChHftap3DE7B2IttWWAhjxkBpqdVP9bbbYPbsjt9POsWTLa91OL0Z1romGVYR6Z+UYRURERGfKqi0MzDGRsw+2tl4eILOvYsJe4KU4MCWv654pgS31buz7cEVwDvvWNtJSfC738HKldb0XwWr3SosuEnA6v53dThdVNU1tFpBWET6B333i4iIiM+YpsmeijqOjQo9oNe9dOlMhiVGNjsW5s64tRawnn9IOj/uKmNEUsT+PSAnBx56CJ57zqoAfPTRkJgIt99+QOMU/xkcF865Bw3mosPS2VVsrYOudxddGhyv9kEi/ZUCVhEREfGZyroGah1OBkbbDuh1c8e2rPw6ZlAUESGBXDtnRItzZ85I48wZafu+cV4e3HknvPyyFaheeCHceqsVrEqPEhBgcP8ZkwHIK7MD4Giw2tr4vd+uiPRY+u4XERERn/loXR4AA6IPLMPamoOGxrPhzuMxOlL4qKEBgoLA4YA33oDLLoNbboFhwzo9LvG/0GArq25vcFJV1+Bdzywi/Y+++0VERMRn/vDBTwAMiDqwDGtbDjhYXb8e7r3Xak2zaBEMGQK5uRAV5ZPxSNfwrH9et7uMmnrnfq+HFpG+R0WXRERExCccTpd3e1CMbwLW/fbDD7BgAUyZAh9/DFOnWlOAQcFqL+TpzXv3x5uAtitJi0jfpwyriIiI+IRn3eHp01IZlrifxZB84Z//hAsugNhY+POf4YYbIH7f/V+l59o7QFWGVaT/UsAqIiIiPrG71Krsul/FkDrDNOHLLyEgAI45BubPhwcfhGuugeho/z5busTefVcVsIr0X5oSLCIiIj6xp8LKsCbHhvnnAaYJH34Ihx4Kxx1nBalgBam//72C1T5k77XLse4pwiLS/yhgFREREZ8oqqoDIDHSD8HFZ5/BtGlwyilQUADPPgv//a/vnyM90t4ZVxHpPxSwioiIiE8UVtZhCw7wXXDR0AB1VhDMnj1gt8Orr8LWrXD11RDa+dY50nO9c+1hnD4tFYCkSP1bi/RXClhFRETEJ4qq6kmKCu1Y39Sm6urguedg9Gh44gnr2AUXwM8/w8UXQ7DWM/YHBw2N5+Gzp7DtnhOJUZVgkX5LAauIiIj4RGFlHYmdyYTV1MDjj8OIEVYBpcREmDzZOhcYaP2RfsUwDIID9euqSH+mBQEiIiLiE2W19QyI6kT/1Ysugvffh9mz4eWX4dhjobPZWhER6dX0lpWIiIj4RKW9gSjbAbwXXloKf/kL5ORY+7feCt98A0uWwLx5ClZFREQZVhEREfGN/Q5YCwvhb3+DJ5+EykpISYErr4SZM/0/SBER6VUUsIqIiEinmaZJpd1BlK2d4jimCf/3f/DUU1bF37PPhj/8oXGdqoiIyF40JVhEREQ6ze5w4XCarWdYCwqsj4ZhbZ99NmzaBP/6l4JVERFplwJWERER6bRKuwOgeYZ161a47DJIS4P1661jr7xi/RkzpsvHKCIivY8CVhEREem0CnsDANG2IPjpJzj3XBg7Ft56C667DpKSrAtVSElERA6A1rCKiIhIp5XX1gMQTQMccUTjetUbb4SBA7t5dCIi0lspYBUREZHOWbaMHR+sgOAJDEtLhHfftSr+xsd398hERKSX05RgEREROXCmCV9+CXPmwOzZbNyaS3hQAEPiw+G44xSsioiITyjDKiIiIgcmM9Nao/r995CaCo89Rmb4QQyvbSAgQGtURUTEd5RhFRERkX1zOiEjw9oeNAiCg+HZZ2HHDvj1rym1O0mICO3eMYqISJ+jDKuIiIi0raEB3nwT7r0Xampg2zaw2WDp0maXldTUMzwpspsGKSIifZUyrCIiItJSXR08/7zVL/Xii62M6kMPQWBgi0vX7i5jd0ktseHBrdxIRESk4/YZsBqG8ZJhGAWGYWxocuwhwzA2G4ax3jCMDwzDiG3jtZmGYfxkGMZawzBW+XLgIiIi4keLF8PVV0NCAvznP7B2LZx9dqsB64KnlgMQZVPAKiIivrU/GdZXgBP2OvYFMNE0zcnAVuDWdl5/tGmaU03TnNmxIYqIiIjfVVZif/Cv8PDD1v6JJ8LXX1uFlU45BQJa/5XBNE3vdn55bRcMVERE+pN9BqymaS4BSvY69rlpmg3u3e+AND+MTURERPytrAzuuot3513I2JJx7Fi90TpuGHDUUdbHdpTWOLzb5x08xJ8jFRGRfsgXa1gvBz5t45wJfG4YxmrDMK5u7yaGYVxtGMYqwzBWFRYW+mBYIiIi0q4334T0dLjjDlZPmQ3ApVMvoMLu2McLGxVX1QHw+HnTmDYkzi/DFBGR/qtTAathGLcBDcAbbVxyhGma04ETgV8ZhnFkW/cyTfM50zRnmqY5MykpqTPDEhERkbbk5kJOjrU9ejQcfzysWUPY/JMA2F1Sy7Pf7Njv23kyrPHhIT4fqoiISIcDVsMwLgHmAxeYTRewNGGaZq77YwHwAXBwR58nIiIinZCVBdddB8OHw+23A1A+fjK/P+s2ykdPoKquMav6U04FpmmyJb+y3WyraZqc9/x3AMRFqOCSiIj4XocCVsMwTgBuAU4xTbOmjWsiDMOI8mwDxwEbWrtWRERE/GTbNrj8chg5El54AS69FP74RwCeWLyNd1Zn8/6abKrqGhg5IJKzZ6bxc04532WUcPyjS7jl3fVt3rqm3onTZb1nHacMq4iI+MH+tLV5E/gWGGMYRrZhGFcATwJRwBfuljV/d1+bYhjGJ+6XDgSWGYaxDlgJfGya5iK/fBYiIiLSusces9aq/upXkJEBf/+7lWUFthdWAVawWVXnJDI0iBFJkRRX13PVa1Y3upyytiv/Ns2+KmAVERF/CNrXBaZpntfK4RfbuDYXOMm9nQFM6dToRERE5MCsWgX33AM33ghHHgl33GFlVAcObHaZaZpsL7ACVrvDSZXdQZQtiPSECACq6qxmAEEBbVcJrqht8G6HhbTszyoiItJZvqgSLCIiIt1t+XKrd+pBB1n9U3fvto4PGNAiWAX4039/JrvUyp5W1TVQVddAREgQ6Qnhza6rqXd6t+sbXNzz8UYKKu1AY4b10XOm+uETEhERUcAqIiLS+519NsyaZWVX77vPKrB0wQVtXr5iexGvfZvl3a+uc1JlbyDSFsS45Giev3gmm+86gVOnprA5v5INOeUArMsu4/mlO7n6tdUAVNRaAeuwxAg/fnIiItKfKWAVERHpbUwTPvsMnO7s59y58Le/QWYmLFwI0dFtvrTC7uD8F74HwDPbd1tBJbnldhIirHWo88YPxBYcSHiItXJo/hPLKK9xUO2eJrx2d5n3XgDRYaoQLCIi/qGAVUREpLdwueC992D6dDjhBPjgA+v4tdfCb38LEfvOdGYWVXu3Hz9vGgkRIXy0Pg+A0QOjml1rC278NeGmd9by3o853n2ny/SuYY227bMkhoiISIcoYBUREenpnE74xz9g4kQ480yorYVXX4VTTz2g29Q3uHhzpbW29b/XH8H8ySnYghuLJY0aGNnieo8vNxXw4bpc735OaS3f7ywmKjSIGGVYRUTET/SWqIiISE9lmmC45+3edReEhsJbb8EZZ0DggVflfXd1Nm+u3AXAqAFWNtUzrTctLoyJKTHNrrc7XLRlR1EVX28p5PTpqQQF6v1vERHxD/2EERER6Wlqa+GJJ6ypv1VVVnD61Vewdq1VYKkDwSpAhrvv6nu/PMzbhiYpMhSAB86YTMBeLWzsDmuN7P+dMIa/ntW8U92W/Epq6p0kx4R1aCwiIiL7QwGriIhIT1FZCQ89BMOGwa9/DZGRUFBgnUtNhYDO/djOq7AzLDGCGenx3mOvX3kIFxwyhOlD4lpc7wlgRw+I4swZad7jMWHB/JhVCqjgkoiI+JemBIuIiPQEubkwaRKUlMC8eXD77XDkkb59RFktKbG2ZsdSY8O457RJrV7/x/njSI6xMWdMEgAfXj+LPRV2nluSwdJtRQDEKmAVERE/UoZVRESkuxQWwn//a22npFjVfr/7Dj7/3OfBKkBemf2ApvAOiLLxh5PGedeoTkqL4djxA5k9KpFa93Th2HAFrCIi4j8KWEVERLpabi7cdBMMHQrnngtlVl9T7rkHDjnEL490OF3sqbSTEtv5NadHjx3g3Y4NC+n0/URERNqigFVERKSr5ObCr34Fw4fDY49Z1X5Xr4bYWL8/ek+FHdOElBjbvi/ehwkp0d5ttbQRERF/0hpWERERf3M6rcq+VVXw0ktwySVwyy1W4NpFcsvsACT7IMNqGAYf3TCLj9bnkRanKsEiIuI/ClhFRET8ZcMGuPdeqKuD996D0aOtLGtcy4q8vrB2dxlT0mIwDKPFuV0lNYBVZMkXJqbGMDE1Zt8XioiIdIKmBIuIiPjaqlVw2mlW1d8PP4RRo8Dlss75KVj9eksBC55azhvf72r1/IaccsJDAhmWGOGX54uIiPiDAlYRERFfevVVOOgg+PpruOMOyMqC++/vdA/VfdlTYU35XbGjqNXzP+4qZWJKDIEBLbOvIiIiPZUCVhERkc4wTVi8GJYts/bnz4f77oPMTLjzToiP75JhVNQ2APB9Rgl55bXNzu0qrmF9djlzxw1o7aUiIiI9lgJWERGRjjBN+PhjOPxwOPZYeOAB63hCAixcCDH+W9+5ZGuhN5Na3+CirKaefHeGtbi6nmMe/qbZ9T/uKgXg6DEKWEVEpHdR0SUREZED9dlnVlC6di2kp8Mzz8Cll3bZ4y9+aSUAX910FKc/s4KyGkez8zX1zmb7GYVVBBgwNDG8y8YoIiLiC8qwioiI7I+GBnC4A8OdO6G6Gl5+GbZtg2uvBVvn+5t6rNheRH2Dq9VzTpfp3Z778DfNgtX0hMaAdLe7KnB+uZ3Hv9pOgGEQGhToszGKiIh0BQWsIiIi7amvhxdfhLFj4fnnrWNXXAGbNllZ1eBgnz7u+4xizn/he578alur53PLals9DvDsRTNYdsvRhAQF8MqKTKBxOvCCaak+HaeIiEhX0JRgERGR1tTWWoHqgw/C7t0wc6bVngZ8HqQ2tWJHMQDbCqpaPf9zbjkAIwdEsr2givCQQN699nDGDIryVgAeEh9OTqkV2O5w3+cvp07w25hFRET8RRlWERGR1px9NtxwAwwZAosWwcqVMG+eXx9Z3+DiP2tzAMgrt1Npd/DvNTmYZuM04I/W5xETFuwNQA8fkcj4lOhm7WriI0JY9HM+Owqr2FpQRWpsGOEheo9aRER6H/30EhERASgrg6eegmuugcREuPVWuPlmOPJIMLqmd+nzSzPILK4h2hbEzqJqrv3HapZvL2ZgtI3DRiSwu6SGj3/K46rZwzl8RCKvX3EwI5IiW9wnLNhaq3rKE8uIDQ9hymD/VSwWERHxJ2VYRUSkfysshNtus6r93n671aoGrHY1Rx3VZcEqwNrdZQxLjODXx4yivNbB8u3W9OAvN+0BYFNeBaYJJ09KBmD2qCRSYsNa3KfCbhViqq53klNWy/QhcV30GYiIiPiWAlYREemfXC4rgzp0KNx3Hxx3HPz4I1xySbcNaXtBFeOSo0hPiPAeiwoN8q5b3VlUDcDQxIhWX+9R3qRycEJECKep4JKIiPRSClhFRKR/KSmxPgYEQGYmnH46bNgA77wD06Z127DsDidZxdWMHBDF6IGN03yPmzCIrXuswkk7i6pJiAghJqz9ok83HTeG8JBA1t4xj69umkNCZKhfxy4iIuIvWsMqIiL9w7ZtcP/98MYbsH49jB4Nb79tBa49wM6ialymVf13SHxjP9Wpg2N478dsVu4s4efcCsYMitrnvU6enMzJk5P9OVwREZEuoYBVRET6tp9/hnvvhX/9C0JC4KqrIMod9PWQYPXJr7bxxUZrneqoAZEYhsEDZ0wC4BdTAl3j/wAAIABJREFUUnjwsy289cNuNudXcMWs4d05VBERkS6lgFVERPqusjI46CArML3pJvjd72DQoO4eVQt//XwrACGBAQxPstannnPQEO/5MQOj+HB9Lg6nyVRV/BURkX6kZ7y1LCIi4isrVsDChdZ2bKy1NjUrCx58sEcGq06XSYBhFUd695eHERoU2OKaEUmR1De4AJicFtvVQxQREek2ClhFRKT3M0346iuYOxeOOAJefBFyc61zJ58MCQndO752PLckA5cJvzl2VJvB6KlTUwBrfWtyjK0rhyciItKtNCVYRER6t4wMuPBC+PZbSE6GRx6Bq6+GiPZbv/QUDyzaDEBSO5V8Dx+ZyM77TsJlgtGFfWFFRES6mwJWERHpfVwuyMmBwYNh4EBwOODpp+Gyy8DWezKQ1XUN3u3Q4PYnPRmGQaBiVRER6WcUsIqISO/R0ABvvWVV/XU4YNMmK5P6ww/dPbIO2ZxfCUByjI3DRyR282hERER6Hq1hFRGRnq++3lqXOnasNf3XMODOO7t7VJ22aEMehgEf3TALW3DLYksiIiL9nTKsIiLS8334IVx5JcyYAR98AKec0mN6qB6o9dllpMWFU1pTz4vLdnLY8AQS2lm/KiIi0p8pYBURkZ6nqgqefRbCwuC662DBAvjiCzjmGCu72ktV2h2c8uRyRg2IxN7gxGXCPadN6u5hiYiI9Fi98+1pERHpm8rL4Z57YOhQuPlmWLLEOh4YCMce2+uC1Qq7A5fLpNLuoMLuYOm2IgC2FVSxu6SWwfFhDEvsHdWMRUREuoMyrCIi0jP84x9w/fVW0Dp/Ptx2Gxx6aHePqsNq651M/vPn/GJKCplF1fyUU97imq9vProbRiYiItJ7KGAVEZHuk5dnrUUdONDKqs6bB3/4A0yb1t0j67SSmnoAPlyX2+Y1gQG9K2MsIiLS1TQlWEREut6uXVY2ddgw+MtfrGOzZsE77/SJYBWgvMbRbP+6OSOa7StYFRER2bf9yrAahvESMB8oME1zovtYPPAWMBTIBM42TbO0lddeAtzu3r3bNM1XOz9sERHplbZvh/vvh1dftdajXnIJ3Hhjd4/KLyrsVsB6xvQ0zpqZxkFD4zlm3EACAwwWPLWcyFBNchIREdmX/c2wvgKcsNexhcBi0zRHAYvd+824g9o/AYcABwN/MgwjrsOjFRGR3u2++6y1qtdeawWvzz8PI0d296j8orzWClgvO2Iohw5PIDDAYEZ6HFPSYrhi1jBev+Lgbh6hiIhIz7dfAatpmkuAkr0Onwp4sqWvAgtaeenxwBemaZa4s69f0DLwFRGRvmrNGjjzTPjhB2v/rrsgMxOeeAKGDOnWofnbXR9tBCAmLLjZccMw+OP88UxOi+2OYYmIiPQqnVnDOtA0zTwA98cBrVyTCuxusp/tPtaCYRhXG4axyjCMVYWFhZ0YloiIdLtvv4WTT4bp0+HLL2HHDut4SgoMGtS9Y+sCpmmSXVoLQPReAauIiIjsP38XXWqtooTZ2oWmaT5nmuZM0zRnJiUl+XlYIiLiN2ecAYcfDitXWj1Vs7Lg3HO7e1Rdqq7B5d2O0lpVERGRDutMwLrHMIxkAPfHglauyQYGN9lPA9qu7y8iIr2PacJXX4HLHaTNng2PPGJN/f3DHyAmxu9DyC6tYfpdX7A6a+/VK93DE7DOn5xMgKoBi4iIdFhnAtb/Ape4ty8B/tPKNZ8BxxmGEecutnSc+5iIiPR2Lhe8/z7MmAHHHAOffmod/+1vrcq/ERFdMoyvNu9h1gP/o6S6npeXZ3bJM/elzuEE4LARCd08EhERkd5tvwJWwzDeBL4FxhiGkW0YxhXA/cA8wzC2AfPc+xiGMdMwjBcATNMsAe4CfnD/+Yv7mIiI9FZOJ/zznzBpkjX9t7ISXnoJ5s3r8qFkFlVz+SurvPt7Kuze7UUb8jn96eWYZqsrUfzK7rAyrKFBgV3+bBERkb5kvxbWmKZ5Xhunjmnl2lXAlU32XwJe6tDoRESk53G5rKm+kZFW4HrWWRDUPes0315l1fWLtgUxPiWaTXmVmKaJYRjc/M46quoauOmddWDCI+dM7bJx1TVYGVZbsL9LRYiIiPRt+kkqIiLts9vh6afhsMOs7eBg+PprWL8ezjuvy4LVjMIq7O6pth7ZpbUMiQ9n/Z+PZ974QZTXOiiprgdg9MBIAN7/MYf31+SQWVTd6n2/zyhm9O2fNsvOtsc0zX1mbZVhFRER8Q0FrCIi0rrqaqt40rBh8KtfgWFAfr51buhQCOi6HyHPLdnB3Ie/4fklGQDkl9tZvGkPO4uqSY0NA2B4krVmNsMdmO7dTmZ1Vmmr937wsy3UN7hYs6v18yXV9dgdTkzTxOF0cc6z33H9m2t4b3U2X29prd4g2JVhFRER8QnV2hcRkZaysqxiSsXFMHeuNfV3zhwraO1i9Q0u7v1kMwCPLt5GSU3z4kpnzUgDYEh8OGBVDD5oaDzltQ4ATpo0iE9+yieruGWG1TRNNuVVAFBW42hxvtLuYPpdX3jvcey4AazMtEoxfLw+D4AVC+eS4g6aPercGVZbsDKsIiIinaG3fkVExFJcDJ+5C7kPGQIXXwzLl8PixXD00d0SrAJszq/wbjtdZotKwCMHWFN/PZnWnNJaAMprHcyfnMzTF8xgcHwYmcU1Le79c24FNfVWNjS3vHFK8MvLd3LHfzbw1WYrg/rJT1Zm+ctNLTOqizftaXHMM3U5NEg/ZkVERDpDGVYRkf4uPx8efhieecYKSvPyrIJKjzzS3SMD4JQnl7d6fOygKDbnVzI+JRqwspkJESHklFmBZ3mNg9hwa1rw8MRIbya1qR2FVd7t3DIr0N1dUsOdH24EID4iZJ/j+ymnvMWxxinByrCKiIh0ht76FRHpr3Jz4frrrfWojzwCCxbAd99ZwWoP4am2C/CPKw7xbj94xmTuOW0i04fEMiM9zns8LS6MN1fu4pZ311NcXU+Mex3r4SMS2FZQRX5588JK2e5s7JS0GHYWVVNWU8/sB//nPb9oQ36LMZ08ObnZfl55y2JNdd6iS/oxKyIi0hnKsIpIl/vlP1YzOS2WOWOSGJcc3d3D6X9cLqtgUnExPP+8NfX3lltg5MjuHlkLnum9M9PjmDUqkY1/OZ46h4s4d+bz/euOaHb99PQ41mWX85a73U1SZCgAk1JjAMgoqmJQjM17/e6SGhIjQ5mUFsN/1uaSv1el4M35lc32d9x7EoEBBh+v/xiAY8YOILOVtbHKsIqIiPiGAlYR6VJFVXV8uiGfTzfk88Cizbx+xcHMHpXU3cPqHzZuhHvvtab9vv46TJoEOTmQmNjdI2vTbnfA+n8njAUgPCSI8HZm6V4xaxgFFXV8/JNVECkpygpOI23Wj7vquuZtcbbsqWRoQjhjBkVTad/FCY8ubXc8gQHWOt4b5o4kPCSIoqo6Vuwo9vZ+BdiYW8FtH2wAFLCKiIh0luYqiUiX2pxnZawS3BmyNbvKunM4/cOaNXDmmTBhAvz735CaCp4+oj04WAWr4i/A4PiwfVxpSYsL56kLphMSaP14S4qyMqwRoZ6AtcF77YodRazZVcZhIxI4Y3oqyU0yrwAPnjm52T3mjR/oPXfTcWP45ZwRJMfYqHU4vRWJAT5xB8sLpqZ4pySLiIhIxyjDKiJdasseK2D9/MYjOf7RpeSV13bziPq455+Hq6+G6Gi4/Xb4zW96fJC6IaeckQMisQUHsrukluBAgwFRtn2/sInQ4ADqnS4SI603RiLdAWtVk4D1FXe14VOmpBAeEsT/bp7D2D8uAuCb388hPSGClJgw4iKCibYFk+ieXtyUp51NbpmdWHfqd1NeBaMGRPLoudMO7BMXERGRFpRhFekFXC6zu4fgM1nF1UTZgoiPCCEl1uat6GqaprcVCFif87/X5OBwurprqL2TacLXX8PKldb+/Plw991WX9W77urxwerKnSXMf2IZz36TAcDu0hpSY8O8U3H31/kHDwHaz7BmFddw7LiBjBoYBVjTd+87fRLHjB1AekIEALNGJTIhJYbB8eGEhbSc3uvJyn6xcQ+f/JSH02Xy465SJqfFHtB4RUREpHUKWEV6uNp6JxP//BlPLN7W3UPxiaziGtITwjEMg+QYG3nuViLv/5jD2D8uYpe7V+Z/1+Xy27fW8tKynd053N7DNOHTT2HWLKtn6gMPWMeTk+G22yC2dwRQ763OBqCg0nojY3dJDWlx4Qd8n1tOGMuaP84jymZNyQ13ryWtdvdcNU2TrJJqhiY0v/d5Bw/hxUsP2u/neHq//u3LrVz3xo+s2VVKaY2Do8dqXbaIiIgvKGAV6eE+35hPTb2TZ5dkdPdQfGJXSQ3p8Vb2KiU2zNsSZPmOIgAW/Wyt/yuqqgMgp0xThvdp0SKYORNOOgmys+Gpp+CNN7p7VB2yMrMEsDKhW/IrWZ9dzqS0mAO+T0CA4a0k7NmPCAn0ZlgLKuuwO1ykJxx4MNxUYmRos16tH6zJISjA4MjRClhFRER8QQGrSA+3x91mIy6idxVvWZ9dxs3vrGs2zdfpMskurWGIO0hIiQmjqq6BCruDOPf6v+0FVQDeIjaaEtwGpxMa3NNbN26Eigp46SXYtg2uuw5sB7bmsycoqLCzs8hqEZNZXMO5z30LwLHjBvjk/hGhQd6ANdP9HM/U344KCDA4fVqqd/+N73dx8LB4om296/tVRESkp1LAKtKDVdod3PvJZgDKaxz7uLpn+ef3u3h3dTYvLM3ANE0ufXklV776Aw6nSXq8FbAmx1pB1QOfbibfnWktqLQyq9nudiaeNa7iVl9vBaZjxzZmUX/1K9i0CS67DELa6fnSgzldJgffuxiw1oWu3V1GaY2DO+aPZ0Z6vE+eERka5C26lFViTT3vbIYVYHhSZLP9Y8YNbONKEREROVCqEizSgz351XbvdoW9gdyyWm9V0p7shaUZ3tYe76zOJi0unK+3FHrPezKsyTHW5/LG97u85woqrIA1q9jKgO1yf+z37HYrUH3gAdi1C6ZPh5QU61xoy+q1vYnLZfLYl1sBiLYFcfKkZF5YtpPAAIOLD0v32XOSokL5aH0eJ03KI6u4msAAwyffTymxzbPZs0b27MJWIiIivYkyrCI9WIW9MasaGGD0mgJEd3+8iQq7O5NVXMNv31rb7PywRGsaZmorwYKn2E6Wu/hSdmktzj5UJbnDTjnFyqSmpsInn8CqVTBvXnePyiceXbyNx91vzqy49RjmuqcAzx6VSFCg735MzZ+cDMCDizazOa+SYYkRBPvg/gkRzd8wGDUgso0rRURE5EApYBXpweoaGtdv/mJyMq99m8V1b6xmU15FN46qfQ99trnd8+EhgQyKtjJSnpYjTRVX11NSXU9xdT3DEyNocJnk9sfCS+XlVja1vNzaX7gQvvoKli+HE08E48DavPRknsrAYE3bPXRYAvedPonHfNzH9MJD00mMDCExMpT1OeVMSj3wYk6tGT0okiNHJ3H+IUP4vxPGEHCALXhERESkbQpYRXoou8NJjnsd59hBURw7fiD1Thef/JTPiY8tpbS6nse+3EZ9Q88qSvTU/3Z4t8clR3u3z5yRBlhVVQ13sNW0t+ZFh6Zz5ykTME1Yus2aPnzYiAQAznhmRbPiTX1acTHccQekp1tB6iefWMfnzrXa1fShQNWjdq9/24AAg/MOHkJMmG8LFxmGwWEjElmVVUphZR2Hu7++Ois0KJDXLj+Ye0+bxHVzRvrkniIiImLRGlaRHmjptkIuenElYE2LfP7imd42Lx6vrMjkscXbiAkL4tIjhnXHMNu18MSxnDwpmRU7ipgzZgCb8ip4d3U2Fx3a+prEuxZMZN3uMgB+8y9rCvH0IXG88f0uCirrWJVZyqxRfXhtoNMJt94KTz8N1dVw+ulW/9Tp07t7ZD7X4HRxwmNLiQwN4v1fHk6Ve/p4V0iMtIpSxYYHe99EERERkZ5LAatID/Rck56rg+PDsQUHegsUeSS6p9M+ungbFx6a7tO1fh3ldJkYBpwzczDXHjUCgHPihwAwICqUj26YxYSU6GavefXyg6mttzJsI/Za+zdtSKx3e38SixmFVQxLjPBmcHuF8nKIiYHAQNiwARYssALXCRO6e2R+szGvwtu+KK/CTr3TRWx4MH/6xXi/P9szDT0lJqx3fZ2IiIj0U93/G66INNPgdLFmV5l3P97dnzRwr3VxLnchorIaB7k+bv3yQ2YJ/2xSuXd/ldbUY5rNpwJ7GIbBxNSYFkHCUaOTOGHiIMBav/j4eY3rFlPjGoP08loHt76/ngl3LGr12RtzK5j78De82EsKU7FjB1x1lVXpNyvLOvbhh/CPf/TpYBVgVWapd9uTVX/4rCmcNs3/Gc/pQ+IACA3Wjz8REZHeQBlWkR4mr9xOVV0DhgGmSZvr+IqbTBEuqq7ztorxhbP+/i0AZ89MIygwANM0WfDUci45fCinT287qPBMW06M7HiblVOmpDBv3EBqHU5CgwK9xx/9citb91hZOafLbBHA7yyy2t8s3lTAlbOHd/j5frdxI9x7L7z5JgQHw5VXNralCQxs/7V9RF55YxGtJ9zVgcentHyTwx8OHZ7AXadOYOZQ3/R2FREREf/SW8wiPUxBpRX0jRkYBbQsSONRWFXv3S5psu1LGe4gsKzGwbrscn739rp2r88usQKRQTGd6wsaFhJIfERIs2OeYBXgzg9/prJJyx+AnDKrDc63GcV86u4B2+MUFcG0afDBB3DjjbBzJzz5JAwa1N0j87sGp4vfvbWWF5ZmUFbT+G+3Ka+CAVGhLaa8+9NFhw1tdRaAiIiI9DwKWEV6mIIKa3qvZ5psehuZ06ZFmIqr61q9prOeW5JBQYWdm96xAtV9Lfn7ObcCw4Axg3wXDDx6ztQWx177Not/rdzd7NjOohrv9oOfbfHZ8zvtu+/gzjut7cREK7OalQV//SskJ3fv2LrQ4s0FvL8mh7s/3sQ7q7ObzRx48vy+V1hKREREfENTgkV6iJyyWt5auYuwEOvb8sJD0zl23MAWRYo88ssb160WV/suw1rRJHP58fo8ggMD+GpzAQBhwe1PWV2zu5RhiRFEhvruv5YF01L5ZmshW/IruXzWMG52B897r0HMKq5m2pBYZo9K4smvtlFT30B4SDf9F2ea8M03cPfdsHgxJCTAdddBUpJV/bcfWr69qNn++ORoLjh0CEeNTiLK5tv2NSIiItJ3KMMq0kP8e00Oj3+1nQcWbSYwwCA+PKTVIkUeP+WUE20LwhYcwIOLtniL13TWkq1WD9RLDx9KrcPJmysbiy+1F7Cu3V3GN1sLOXbcQJ+Mo6m/nTOVT34zm7ljB3iPldc0nxKcVVzD0IQIZqbH4TLhH99l+Xwc+2XbNpg92+qZumGDlUnNzLSC1X7K6TL5/Oc9zY7VOJzMn5yiYFVERETapYBVpAd4cNFmHmoyjfWIkYkEBLQMVD+6YRYPnTmZo8dYwY9hGJwyJQWAV7/N9MlYlm8vJtoWxJWzG3u7XjFrGFfMGkZlXYO3OvHe7vpoIzFhwVx8WOt9Vn0hPiKEn/58HJGhQZQ2CVjtDie55bWkJ4Qze1Qi45OjWbypwG/jaMHlgjz3utmkJKiosNam7twJN90EkZHtv76PW59dRn5F80rWG3PLu2k0IiIi0ptoSrBID/D01zsACA0K4LFzpzIjvfUKphNTY5iYGkOULYj/bSmkvNbBXQsm8t6PORR3svBSbb2TsJBANudXMD4lmtTYxiI4aXFh2IIDqW9wkVNWy+D45utqdxZVszqrlFtOGEtanO+qFbcmyhZMTFgwZTWNn292aQ2mibcH68TUaP63pdCv4wDA6YR33oF77rEW+K5dC7GxsG7d/jWO7Se2uQtmjR0Uxeb8SgAePWdaey8RERERAZRhFekRQgKtb8UJKdGcMDGZpKj2q+w2DQpDgwI5ceIgMoqq2nlF2x75fAtDF37MuDsWsXVPJVvyKxk7KBrDMEiMtCr1JsfYGO2uWrwxr8L7WtM0Wbu7jKP/+jUAp0xN6dAYDlRcRDClTQLWOz/cCEB6QgQAI5IiKayso7zW0errO83hgFdegXHj4LzzrAzrLbc0nlewSlVdAw1OFwDbC6sICQrglhPHAvDtrXM5eXL/KTglIiIiHaeAVaQHMAyYN34gz108c7+u37vlS3pCOLll9jan67bncXcfTICl24qoqXeSFmdlV588fzopMTamDYlj7KAobMEBvLA0w3v9vZ9sYsFTywG45YSxzbKy/hQfEUqJu9CU3eFk6TaroM9Qd0XlgdE2oHmvWp96+2247DJrqu9778FPP8EFF0BA//0v1eUyeWFpBuU1DrJLa5j4p8+809zXZ5cxakAkR48ZQOb9J3dpCxsRERHp3frvb1ciPYTd4aSuwcXUwbEkRu5f/9K9A9aB0TacLrPNasF2h5MXl+2kpr6h2fG9e5luzLWyp55xHDo8gRW3HsPAaBsRoUGcNi2VTXnWlE7TNHllRSYAFx2azi/njNivsfvCoOhQ75rIqrrGzyk23Pp7CQ+xikPV1Lfew/aAVVfD3/4GL79s7Z99NixaBKtXW1V/+3Gg6vFzbgV3f7yJX/3zR15engnAv9fmUF3XwOqsUmaNTOzeAYqIiEivpDWsIt2swj1tNTps/6ul2vaq1jvAPYV4T4W92XRiu8PJhD99RkhgALUOJyGBBhcdNpS6BichgQH8kFnS7D7v/ZgNQEJk84DYIynKRnV9AxtzKzjp8aUALDxxLFfOGtbq9f4yKCaMPRV13PXRRm+Rp7+eNcV73tPOptMBa3k5PPWUFawWFcH551uZ1eBgOP74zt27j6mss76Ol20vYq27YvWeijr+/s0OHE6TqYNju3N4IiIi0kspLSDSzTx9T6NtHX//aIB7CmxBZfNKrLlltThdJrUOK3BblVWK3eFkzO2LePjzrXyxsYCo0CAmpcY0e11CROuZ3oiQQEwT3l6123vskGHxBAV27X8l8eFWcP/isp3eDGtkaGMQH+7ert4ro3xAXn0V0tPhttvgoINg2TJ4442O36+PK2tStTko0ODRc6YC8Ow31hTyQTG2bhmXiIiI9G4KWEW6macwUMwBZFgBLjx0CL8/fgwAKe41gf9ek9vsmj0VzddwZhZVe9u9vP5dFjuLqhgzKIoPb5hF5v0ne69rK8MaHhrkvm9jYDw5reszZzHhjX9XhZXW5xgR2hjwR7gzrLUHmmHNz4fiYms7ORnmzoVVq+CTT+CIIzo36D6uaRGsOaOTWDAtlRMmDKLeXXhJ61ZFRESkIxSwinQzT8C1v+tXPe5eMIlfHT0SsLJXp05N4dMNeazOKiW/3Aoo9864ltc6vNN+y2sdfJdRQnKTQkkPnzWF5BgbCRGtB6yeLKZnyuc/rzqEwFb6xfrbqVNSOe/gIQD8uMsaS9OA1bOGtbpuPzOsu3fDr38Nw4bBffdZx447Dt5/H2bM8N3A+yjTNLnvk83e/cPd61UnpER7j+2r8rWIiIhIaxSwinSznDIrqOxshd3Tp6fhcJqc8cwKfv/uOqAxE3rc+IHMHpVIea2jxbrVlNjGqZpnzEjj21uPaXOKr2dtaF65nXMPGszhI7qnkE5AgMGvjraKPP2w0/p8IlsJWLcV7KPVT0YGXH01jBgBzzxjVfq99lr/DLoXq7A7WLK1kOKqOoYu/Jj/rM1pdj6ruMY7NfujG2Zx1ow0wOob7NEdb2yIiIhI76eAVaSb5ZbVEhYcSGz4gU0J3tsRIxK45sjhQGO13yVbi0iKCuXZi2YwJS2W0hoHlfbmWcfBTXq67otnqi3g7cvaXVJjw4i2BXkDcE+QCo3Z1ueWZDSrItzCbbfBa69ZQev27fDCCzBypF/H3Rs9/b8dXPzSSm/13xeW7mx2Pru01rs9MTUGw92HdkKqlWE9cnRS1wxURERE+hwFrCLdLK+8luRYm/eX/I4KCgzg1pPGcdO80RRX11NV18Cy7UWcOSMNwzCarZG9ad5oFkxN4ZdzRnDG9LT9fkZEk8JGYwZ1b8BqGAbjU6JpcPeebZphDQ1q/K/t5MeXsny71aeVtWvhrLNgwwa+3lLAPaf8FnbuhCeftAosSatKqq1p6++stoptOZwuauudrM+2pmNnl9YA8K+rD232ugFRNt686lCeuWB6F45WRERE+hK1tRHpZhtzKxjlw2zlsKQIAH7KLgfwrkdtGrDOGpXIDceMOuB7N10n2t0ZVoDxyTF8l2FlWJuOrWnwn1VcwwUvfE/mz3+Hjz6C6Gg4+2wuXZ0FwA2nTScaaY+neJfnY4PL5NSnlrF1TxX/u3kO2aW1BAYYzEyPa/Haw0YkdOlYRUREpG9RhlWkG2QVV1Nb7yS3rJbM4ppWf9HvqJnp8QB8uWkP0Lju1MT0XjMuuWMhWtNiTIltVBLuSuObFPUJ3mvd7SHD4pvtO779Htdf7oKsLCvL6uaZPi1t82RQPYqq6ti6x1ofnFdeS3F1HfERIV3e3khERET6vg5nWA3DGAO81eTQcOAO0zQfbXLNHOA/gGfB0/umaf6lo88U6Qsq7A6OeuhrZqTHMS45ipCgAI6bMMhn9x8UY2PUgEgWbcgHGqfxHjd+EGt2lbHwxLHYggPbu0WbEiKt9bB2h7PTU5h9YXxrgbdpwrJlPHHeQRx872Lv4ZnXv05ieCiLY602PLbgAOwOF9sKqjh0uLKAbWlwuthV0jxgbdpztbTaQWm1g9gDbMskIiIisj86HLCaprkFmApgGEYgkAN80MqlS03TnN/R54j0NZ//bGU+V2eVUlhZx9S0WIYlRvj0GbNHJfHScut9Ik+GNS4ihPvPmNzpex/vw+C6s0YOiARgcHwYuFzw4Ydw992wahUJX3zR7NpyewPl9gZM08o0O5zWx9LqeqRtWSU13r8rgD/9Yjx3frjRu1+v/VE+AAAgAElEQVRSU09ZbT1x4d2fcRcREZG+x1drWI8BdpimmeWj+4n0WT/uKvVu7yqpYfqQWJ8/Y8rgxnYiESEdy6b2BiFBAfzz8oMY9v3XMHUq/PQTDB8Ozz9P4JFH8tG4WmzBARz7yBLva3aX1BIbEYzTXaypRAFru1ZnWV+v958+iaKqOi45bCjFVfVE2YK479PNvLtqN3aHiyEJ+19tWkRERGR/+SpgPRd4s41zhxmGsQ7IBW42TfNnHz1TpFf6OaecQ4fHk1FYTUFlHSmd7L/amvgma03DQ/t2bbXDh0TDUb+BuDh4/XU491wIsj7niakhOJyuZtc/9b/tXD+3sXXNmt1lLN9eRGlNPSdPSu4RU517ii837uH/3l0PwPwpKd5KzDcfPwaA+z7dzDp3ca+mb5KIiIiI+EqnK2QYhhECnAK808rpH4F00zSnAE8A/27nPlcbhrHKMIxVhYWFnR2WSI9kmibbC6oYOyiaGe5CS+l+yEw1nZ4ZGdrHMqx2O/z97zB3LjgcYLPB0qWwYQNceKE3WPVoWowpPiKED9bmkFFU7T22bncZF7zwPdf/cw0/qwBTM5+610FD87ZBrYnVlGARERHxA1+UdDwR+NE0zT17nzBNs8I0zSr39idAsGEYia3dxDTN50zTnGma5sykJDWZl76pvNZBdb2TtLgwbp8/nkfPmcqCaak+f05Ckwq+njWsvV51NTz6KIwYAb/8JdTWwh73fzsjR0LAvv87W3jiWOobXPzpPxsACN9runRhVR1VdQ0MXfgxn/6U5/NPwZeKquq475NN1De49n1xB+1dHXhv6+44jlT3DIGkyFC/jUNERET6L18ErOfRxnRgwzAGGe75dYZhHOx+XrEPninSK2WX1gKQFhdGamwYC6alEhrk+wxo0wzr3kFZr7R9OwwdCjfeCGPGwOLFsGIFpKUd0G1On5bKjPQ4MoutQGyae/3wL+eMAOCyl3/gmtdXAfDc0gzfjd8P/vLhRp5dksGy7f6ZkVJT38D3O0uYmR7H4puOavWamPDGtcAjB0b6ZRwiIiLSv3UqYDUMIxyYB7zf5Ni1hmFc6949E9jgXsP6OHCu6SnRKdIPrcsuAyA11r8Fapq2rYnpre1GSkrgq6+s7eHD4eyzYdky69jcuXAAa01nj0okNCiAoMAABkQ1ZgLvP30yfztnCpcfMcx7bPl26z215Bibbz4PP3C5TFbsKAKgqNI/RaNufGstAFfMGsaIpLaDUc8/w6gBClhFRETE9zo1V9A0zRogYa9jf2+y/STwZGeeIdKXvLYii+FJEYxNjvL7sx4/bxojkyJ7XxGhPXvgkUfg6achOBhyciAsDJ56qsO3fO3yg3EnAqmud3qPx0eEcNq0NFp7H21gdM8NWL/fWUJRlRWoNl2P60s7CquJCQtm3viB7V737EUzeP/HHO/UYBERERFf6iOL20R6PpfLZGdxNZcePrRZISB/OWVKit+f4VO5ufDAA/Dcc1BfD+ecA3/4gxWsdpJhGAS64/bbTx7HVcXVTE6L9U6Xbi2o78lTqXeVWEFqUIBBph8CVofTRWZRNVcdOZygfXytTk6LZXKa71sziYiIiIACVpEuU1BZR32Di8Hx6lfZjGla80qzs62s6kUXwcKFMHq0Xx43emAU3/z+6BbHQwIDqHe6OHNGGu+uzvZrMaPOyi2zYxhwxMhEdvohYN1dUkODy2x3KrCIiIhIV/B/mkdEANhVYhX6GaKA1bJ5M1xyiVXxF+Dgg62g9aWX/Basticuwlrre/Fh6cSEBffIgHV7QRWf/5xPXnktSZGhjB4YyZY9lWzJr/Tpc/LL7QCkxPbcadEiIiLSPyhgFekiCljd1q2zpvuOH///7d13eFVF/sfx96SSBmkk9ITepSoiooAoWLGsu7oqtrXX1V2xrIuirsq6ru23dhdxbYi9AIoNlCYgSi8m9IQUIL1nfn+cy01CQkjITe5N+LyeJw/nzpl7zlyGE/LNfGcGZs+GyEhnlBUgvvb5ko3p+G7OdPw2IYEE+vtRXOa99eHyi0u5c9YvbM+suq3M+Ce/59o3VpCUnkeHyBC6uUZAJzy1gLyiUo/dPzXbCVjb+fA8XhERETk6KGAVaSLb9+ZjDEf34jTPPw+DB8PcuXDPPbB1Kzz2WL1W/G0sj51/DP+98lgSYsIIDvDz6gjr7BU7eX/lTp79ZrO7bO6aVPfx8m37GNUjhr7tW7vLtmZ6LjXYHbD68ErJIiIicnTQHFaRJrJzbz4d2oQQFHCU/Z5owQJo3doJVE8/HaZNg1tucUZWfUhIkD9je8cBEOhvKCnzXsD64xZny5oUV2ouwHcb06rUuejYLsSGV2zRk5yRR/8ObTxy/z1ZhUS0CiA0SP9FiIiIiHcdZT85i3jP2t3ZdI0N83Yzmoa1MG8ejB4NJ5/srP4LkJgI99/vc8HqwYK8PMK6emcWAEuSMknLcYLWrZl57j11T+sXT+foUEKC/PnbmX0B2LQnt8H3LSu3/Lx9H68v3kbnqKM8dV1ERER8gn59LtIEduzNZ+OeHH5/bGdvN6XxzZ3rBKXLl0OnTvDss3D11d5uVb0E+vt5bYR1f34xu7MKuWBoJ95fuZMrXvuJ68d0Z3tmPuP6xDG+bzxjerd11//T6G58+PMuPvtlN3lFpbRv04o/je5W7/sWlpQx7onv2O0a1Y1opf8eRERExPv0E4lIIyortzz02TriWjupm0O7+PbI4hErK3Pmofr5OYHq3r3w8ssweTIEBXm7dfUWFOBsceMNUz9ZC8CE/vHMX7+HdSnZ3Pr2zwAkxIRy5jHtq73n+G4xvPpDMkk/JANw+RHs9Ttvbao7WAWOvtR1ERER8Un6iUSkEW1Oy2HGoq1Mn7sRoOXtwVpSAq+/7qz4+/77Ttmdd8LGjfCnPzXLYBWcPVmLvJASnF9cyserdgPQJSaU0/pVXTU5MabmlPI/juhS5fWs5Tvqfe9fXWnID57Tnz+P78Uj5w6s9zVEREREPE0Bq0gj2lxpXmFwgB8xYc0zgKumqAhefNHZL/WKKyAkBKKinHMhIRDQvJM3ggK8kxKckVPsPu4cFcoj5w3k1nE93GVdYmr+hUf3tuG8cfVx7tfv/lT/gHXTnhwGdmzD5Sckctv4noe8l4iIiEhTUsAq0ojWpWS7jxNiQjE+sH2LR0ycCNdf7+yb+umn8PPPMH68t1vlMUH+1RddSssu5PcvLnav4NsY0nOLAIhvHUxYcABBAX5cenyC+/yhRlgBRvdsy8r7T+XS47uw7aD9WytblryXL1an8FylLXMmPrWAhZsz6BUf4YFPISIiIuI5ClhFGtGKbfvcx29cPcKLLWmg7Gx44gnIc+31OWUKzJ8PixfDWWf5xD6qnhToClhXbt9HebkF4NUfk1mWvJc5a1Kw1tb5WtZaPv81hbyi0sPWzXAFrK9MPtZdFlNp65row4zQR4cFkRgTRlZBCQs2pVc7/86y7fz+xcXc+OZKnvhyE4UlZVhr2ZCaA0CfdgpYRURExLcoYBVpJMWl5fyyYz8XH9eZZfeeQnzrVt5uUv3t3QtTp0JCAvz1r84KwOCMsJ5ySosLVA8ICvBjc1ou5/9nEf/6ypl/vD7FCer+t2Q7J//zuzoHrXPWpHLTWyt56LN1h617IGCNjagITP39nL/jQZ3qtsfqgI5OvQc/Xesu27E3n22ZeTw1fzPhwRXp2gXFZezcV+B+Hd+mGf4bFRERkRZNAatII1m7O4ui0nJG92xLXHMLVktKnFHUhASYNg3GjnVW/73gAm+3rElUXmH35QXJzihkpfTu7Xvzq6R7H4q1lme+dlJvZy3fwZa0nFrrH5jDGhMWXKX85/tP5d3rRtap7cd3i+F3wzqxLTOfnMISktJzGT39W877zyJSswu5aWwPpl9wDAB5xaWs3O5kARzTqQ2n9o2v7dIiIiIiTU4Bq0gjWbVjPwDDEqK83JJ6yHUtEhUQAEuXwtlnw+rV8MEHMGyYd9vWhGLDK0Y4i8vKmfrJWtJyiqrU+WDlrsNeZ9f+Ajak5vCH4Z0pt7BmV+1BbkZuEZGhgdW2lIkKC6JVoH+d23/JiC6UllteWZjMzMXbANib5wTDveLDCQ12rrU/v4QXvk8iJiyIj24cRUhQ3e8hIiIi0hQUsIo0kpSsQoID/IiLCD58ZW9LSoLrroOOHSE11Un1/eoreOstGDDA261rcgdWyA0N8qdDm1buoO/hcwfw97P6MTwhiu82ptV6jczcInca8ZjebYGKoPFQMnKLiA1v+L+XIV2iOLVfPM98s5lPf9ld5dyIbjGEBTlpwbNX7GR9Sja3j++Jn1/LTO8WERGR5k0Bq0gjSc8pom1EsG+vDLxhA1x+ubM9zYwZcMklFfNSAwO92jRv6hAZAjhbyyRUWpl34oB2XHViV0b1iOW39Dy+WJ1S4/uttQx7eD7XzFwOwDGdI/H3M+6AdWlSJnfN/qXaPFgnYPXM1kd/Ht8LayEzr5g/jujCgI6tmX39SMKDAwh1jaR+vjqFuIhgLhmRcJiriYiIiHiHAlaRRnIgYPVZqakwcCC89x7ceiskJ8N//uNsVXOU69++NQB/PrWXO36/alRX9+jngYWNbnxzJet2V0/zzcitOpLavnUr2oQEMnvFTlKyCrjhzZXMWr6T39Jzq9Rz/s14Zr5z5RV/zx3ckc9uGc3wxGgAQl0jrOk5RSTGhml0VURERHyWAlaRRpKeU0RbD6R3etSyZTB9unPcrh28/jps2wZPPgkdOni3bT4krnUrtj52JhMHtOPAIOjoXrHu8yf1inXPM127O4uSsqp7tm7e46QCB/ob1j44AT/X6GpqdiG3v7OKINeiTkuT97rfU1pWTmp2IfEe+iWHn5/hntP7kBATytAukVXOHZjDCjSPlHURERE5ailgFWkE1lpSsgqIa+0jwcCCBTBhAowY4QSs+1z7w/7xj9C2rXfb5uNOds0/7RwV6i4LDvDnhyljAfjr7F+Z+omzhUxWfgl3zf7FHYj+OGUcYa5tZDq4toxZmryX1OxCAFZt309WfgkAm9NyKSwpd4/eesJ1J3fn+7+OJcC/6rf6A3NYAUrL6r6nrIiIiEhTU8Aq0gi2ZeaTXVhK/w6eCz6OyMaNcNJJcPLJsGqVE6wmJ0NUM1q52MuuHd2NhXeNpUdceJXy2Epbz7y1dDtZ+SU89+1mZi3fydNfbyYyNLBKSvhnt47mv1ce635tDLy3YieDpn0JwPJtzi8RBnWuOhraGCqvBmxRwCoiIiK+K+DwVUSkvg5saTO4CYKPasrLISMD4uIgOhrS0uDZZ+HqqyEkpOnb08z5+Rk6R4fWWN6udSuKy8rZm1fsDjwP6BUXUWXBreiwIMb2juOly4aREBPGKwuTeG/FTvf5r9btITEmlMSY6vfytIjgAE7pE0dmXjFTz+7f6PcTEREROVIKWEUawaod+wkN8qdXfMThK3tKWRnMng2PPAKhobB4sZPuu359xcq/4lGL7xlHcVk5vf82t9q50T1ja3gHnNa/HQA94ytGbItKy1i3O4vxfeObZFVpPz/Dq1cce/iKIiIiIl6mlGARDysoLuPbjWkc06kN/k2x+mpJibN4Ur9+cNFFzuubbsK9WpCC1UZjjCE4oCK9dkilxY2uGJVY63u7xVYErGnZRWTkFtOujWdWCBYRERFpKRSwinjY/5ZsY1tmPleO6to0N5w5E664wkn3fe89WLMGLrsM/PR4N5Wusc5ere9dN5JLRnThvCEdiWhV+z623dpW7O+6MdVZVTi+tQJWERERkcqUEiziYR+t2sWQLpFMcKV+elx+Prz8srNf6kUXOSv9tmsHZ5yh0VQvee/6kWzfm0+Avx+PnDewTu/p1jacv53Zl4c/X8/6FGcv13hfWVVaRERExEdoCEbEg15ftJW1u7MZ1b3m+YsNkp0Njz8OiYlw++0wZ45THhICZ56pYNWLYsODGdql/isvj+8bD8DK7c4KwRphFREREalKAauIh+QVlbr34xyW6OFtY/77XydQvftuGDYMFi505q1Ks9YpKoQgfz9+3JIJVKQWi4iIiIhDAauIh6zd7aR13jquB2N6tW34BdPSYL+zPQ5RUc5eqj/95Iysnnhiw68vXhfg70dibCjFZeV0jg4hNEizNEREREQqU8Aq4iFbM/IA+N2wzg3bmmTnTiflNzER/v1vp+zcc+HDD2H48IY3VHxK97bOasE945pwCyQRERGRZkK/zhfxgJKycqbP2whAh8gjnIeYlOTMUZ0xw9lT9bLLnAWVpEXrEecKWCvtyyoiIiIiDgWsIh7w6g/JZOQWMbJbDAH+R5i4cOed8MUXcNVVMGWKM8IqLZ5GWEVEREQOTSnBIh6wLTOfmLAg3rpmRN3f9OuvcPHFsGWL8/qJJ5xR1uefV7B6FDm+WwxDu0QyqkeMt5siIiIi4nMUsIp4QH5xKRGtAuo2d3XZMpg0CQYNgs8+cwJXgO7doWPHxm2o+Jx2bVrxwY2jaN8mxNtNEREREfE5SgkW8YC8orLDr/BqrROofvqps+rvAw/ALbdAdHSTtFFEREREpLlRwCrNwue/ptC3fQTd2vrmwjT5xaWEBftXP2GtsxXNcceBMXDMMTBqFNxwA7Ru3fQNFRERERFpRpQSLD5vwaZ0bnprJbe9s+qQdYpKy9ibV9yEraoqr/igEVZr4ZNPYMQI52vRIqf84YedBZUUrIqIiIiIHJYCVvF5m/bkALB6VxalZeU11pn26TqGPvQVWfklTdk0t/wi1whrWRnMmgWDBzvpvxkZ8OKLMGyYV9olIiIiItKcKWAVn5dZaeQ0NbuwxjqLfssE4PXFW5ugRdXlHxhhLShw0n2LimDmTNi0Ca69FoKDvdIuEREREZHmTAGr+Ly9uZUC1qyaA9ao0EAAPli5s0na5FZUxO7/e4Vd+wsIC/SD8HD48UdYuxYuuwwCNE1cRERERORIKWAVn5eZV0SAn7NdTMohAtYDgWyTzWPNz4enn4bu3bll8T6nDelZzrk+fcC/hgWYRERERESkXhocsBpjthpjVhtjVhljltdw3hhjnjHGbDHG/GqMGdrQe8rRJSO3mP4dnEWK3v1pR7XzpWXlpOUUAZBbVEp5uW3cBq1fD4mJcPvt0L07W7v2AyA6RgspiYiIiIh4kqdGWMdaawdba4fXcO50oKfr61rgeQ/dU44CZeWW5Iw8+rRrTbfYMH7YksHa3VlV6ixJ2ktpuWVIl0jKLeQVl3q+Ifv2Oam+AD17wllnwfff89b0N8gsgdE9Y7nvzH6ev6+IiIiIyFGsKVKCJwEzrWMJEGmMad8E95UWYPWuLLIKSjihRwwf3jgKfz/DnNWpVer844v1RIYGcs6gDgDkFHowYE1Lg3vugYQEOP98KC6GgAC2PPY0OwYM594PVwMwoX87woM1X1VERERExJM8EbBa4EtjzApjzLU1nO8IVM7j3OkqEzmshZvSATixRyxtQgPxM/Dct1v43lWeXVjCupRsrh7VlbiIVgD8/eO1PPP15obdOCUF/vxnJ/X38cfhzDNh/nwICmLe2lTGP7mAG99c6a7eMTKkYfcTEREREZFqPBGwjrLWDsVJ/b3JGHPSQedNDe+pNsnQGHOtMWa5MWZ5enq6B5olLcHipEz6d2hNTLizLcx5Q5zfdXy8ahcA2zLyAegZH0HrEGeEc/76PTz51aYq1ykvt3Wb22pddTZtgmefhT/8wZmz+vbbMHAgqVmFXPfGCsAZ/T0gvnWrI/+QIiIiIiJSowYHrNba3a4/04APgeMOqrIT6FzpdSdgdw3XeclaO9xaO7xt27YNbZa0EFvScunXvmIxo4fOHUCfdhH8lpYLQHJmHgCJsaFEtAqs8t6v1+/hopcWsyE1m4teXsLo6d8e+kYbN8IVV8CddzqvTzoJtm2D//4Xevd2V/t2Y1qVtwX4GXrFh9OtbVgDPqWIiIiIiNSkQZPujDFhgJ+1Nsd1fBow7aBqnwA3G2PeAUYAWdbalIbcV1q+x+ZsYFlyJmk5RXStFAwGB/gzsnsM7yzbQXm5JTWrAHBScg8eP736dWfR6olPLXSXlZaVE+Bf6fc0v/4K//gHzJoFrVrBbbc55cZAx4rM9Z378pm7JpUV2/bRvk0rPrn5RJ77ZjNXndiVhBgFqyIiIiIijaGhq8TEAx8aYw5c6y1r7VxjzPUA1toXgC+AM4AtQD5wZQPvKS1cYUkZL3z/m/t114MCwh5x4RSUlLE7q4DMvGKC/P0IDw7AGMMdp/aqlg5c2dbMPHrERTgvnnnGCVDDw2HKFGfOalxcje+7491fWLZ1LwC/H96JthHBPDhpQAM/qYiIiIiI1KZBAau1NgkYVEP5C5WOLXBTQ+4jRwdrLS8uSKLsoLmmwxOjq7zu6Qo4b/jfSgpKyogOC8L1SxNuHtuDgZ3acMe7q9iXX8LQLpGs3pVFSZlzzS3fL6dH/3jo1w8mTIAHHoBbboFo5x7pOUUs3JzOMZ0iCQ7w4/znF3HLuB7uYBVg8sjERvobEBERERGRyrQPhzSZnfvy6RQVesjzUz9Zy8zF29yvh3aJpHvbcNpGBFepNywhiqtP7MqrPyQDEOhfsa6Xn59hbO84Fk4ZR6C/ITjAn7KyclLnfM2oH4rZ/fhTMCAKZsxw5qZOnVrl2lf8dxlrd2cD8OzFQ0jPKeLvH691n59x5bEM6NjmiP8ORERERESk7ppiH1YRliZlcuLj37pX9z1YWk4h/1uyDX+/iuDzgxtH8c8Lqw3g4+9nuP+sfrxw6VAA9+hpZeHBAQQH+MO8efiPOoEOZ59GSGkRa86+iGcvuovSsvIq9Wf8mEyvv81xB6sAi37LdB+fPagD66dNZEzvmlOGRURERETE8xSwSpPYkJoDwJKkvTWeX5K0l3ILM69yFpk+e1CHw17ztH7taj5RVlaxPc3330NaGubFFykICOaDgtb867tkliVXtOPvH6/hgU/XUVzqBLFP/t4Jkt9ett1d57LjEwgJ8j9sm0RERERExHMUsEqTODBwWlZeXuP53fud1X4HdY5k/h0nM/2CYw5/TT/DK5OHM+e20U5BaSm88QYMGACff+6U3Xefs2XNtddyTqUgeEt6rvv4h80ZgJPuO/v6kZwxsL373KDOkSy77xSO61p1Hq2IiIiIiDQ+BazSJLIKSgAodaXvZuWX8NHPu8jKd8pTswqJCA4gPDiAHnHhdR7NHN8vnr7RwfDSS9CrF0yeDEFBEBLiVAgLg0Bnf9ZnLh7CXROdPVVX78wCYEtaDkkZeVx7UjfG9I5jeGI0rQL9mTapP+P6xPHWn0YQF9HKY38PIiIiIiJSd1p0SRpdek4RT3zpbDWzOS2Xi19awuIkZ37oNaO7cuHwzsxYtPXIbzBuHCxaBMceC08/DWed5eyjWoMbx/Rg5bb9/LR1L8kZeZzxzA8AxB20sNPkkYlaDVhERERExMsUsEqj2ZCaza1v/8ymPRXpt6t3ZVWp8+W6PeQXlwHQKSqkbhfOyYHXXoPrr4fgYLjrLggNhfHjDxmoVnZ8t2jmr9/D2Ce+c5dN6H+I+bAiIiIiIuI1SgmWRvPByl1VgtU/j+9V5fw5gzqwLTOfNbuy6BIdyoc3jqr9gvv2wbRpkJAAt98O8+Y55ZMmwamn1ilYBRjRNabK6z8M70zn6ENvtyMiIiIiIt6hgFUaTVKlhY0+umkUkwZXXfl3dM9YANalZDOwY5tq+626FRXBvfc6gerUqTB6NCxdCuecc0Tt6tehNaGV5sgeqwWVRERERER8klKCpdEkpee5j/u2j6CwuOoKwR1dKcAlZZbI0MDqFygocBZPCgqC+fPh9NOdwHVQ9b1Z68Pfz7Do7nFkF5TSOiSANiE13FtERERERLxOAas0iulzN5CUkccVJyQyaXAHggP8CfSrOqDfKbIiDTc6LKjixNatMH06zJrlbEkTEwMLFzrzVT0kMjSIyNCgw1cUERERERGvUcAqHldebvnPd78BcGq/eIZ0iQKcfVMPmHXdyCqLLEWFBsGmTfDYY85eqn5+cOWVUOYsyOTJYFVERERERJoHBazicdv35ruPhyVE1VjnONe80Y6RIezaX0B0YQ70G+qk/950E/zlL9CpU5O0V0REREREfJMCVvG4dSnZAHx684m0CvSvcm7hXWMJDnSlBi9fzhSzlU/6HsPxx/aCl1+GM86A+PimbrKIiIiIiPggBazicet2Z+PvZ+gZH17tXOfoUPjhB3jkEZg7l3NiYzknKQkiWjkpwCIiIiIiIi7a1kY8bn1KNt3bhlUbXWXdOhgzxtmWZsUKePRR+O03iIjwSjtFRERERMS3aYRVPG5dSrZ7jirWwr59EB0NbdrA9u3w1FNwzTUQGlr7hURERERE5KimgFU8al9eMSlZhfRrFwGzZ8PDDzvB6jffQMeOsGWLswKwiIiIiIjIYShyEI/6Zl0KAAP+fgdceCEUFsIVVzgjraBgVURERERE6kwjrOIRCzal882GNOYt3cKA1K2MzNoO774LF1wA/v6Hv4CIiIiIiMhBFLBKwxQUwCuvMHlXN1dBANPHdcXvtZUaTRURERERkQZRRCFHJicHpk+HxETKb73NXdwxMoRRl56tYFVERERERBpMUYXU3yuvQEICTJkCgwfzy8dfA3D/Wf34+OZR+PkZLzdQRERERERaAgWsUjfp6ZCb6xyHhTl7qS5dCvPm8VVQe/z9DL8b2onY8GDvtlNERERERFoMBaxSu9274Y47IDER/u//nLKLL4aPP4bjjgNg/vo9jOgaTZvQQO+1U0REREREWhwtuiQ127rVmaP66qtQVgaXXAKTJlWrVlhSxqY9uZw5sEPTt1FERERERFo0BaxSs+uvh2++gSuvdOaqdutWrYq1luVb9wGQGBva1C0UEREREZEWTinB4lizBi69FJHUCgAAABFISURBVHbscF4/9RQkJcGLL9YYrAK889MOLn11KQBdohWwioiIiIiIZylgPdqtWAHnnw8DBzrzUn/+2Snv0wc6dar1rV+vT3Mfd40Na8xWioiIiIjIUUgpwUer8nI491z49FOIjISpU+HWWyE6us6XKCgpBSA8OIDI0KDGaqmIiIiIiBylFLAeTayFVatgyBDw84MePeDRR+HGG6F163peyrJudzaDO0fy0uRhjdRgERERERE5mikl+GhgLXz2GZxwAgwdWpH2++STcPfd9Q5WAVKzC9mXX8J5QzoSF9HKww0WERERERFRwNqylZfD7NlOkHr22ZCaCs8/D337NvjS63ZnA9CvQ/2DXRERERERkbpQSnBLlpUFV10F7dvDjBnwxz9CYKBHLr3WFbD2ba+AVUREREREGocC1pakuBhmzoQ5c5yR1ago+PFH6NcP/P09eqs1u7JIjAklPFj/hEREREREpHEoJbglKCiA555zFlG65hrYvh0yMpxzAwd6PFgtLi1ncVImxybWfUVhERERERGR+lLA2tytXg1du8Itt0BCAsybB8uWQdu2Dbrsvrxi/vXlRp77ZjPgjKje88FqSsrKWb0ri5zCUsb1ifPEJxAREREREamR8jmbo/37YdMmOO446N0bxo+Ha6+Fk07yyOWLSss4//lFJGfkAVBWDrNX7mDH3gKGJ0RRUlYOQP8ObTxyPxERERERkZpohNULMnOLKCguI6ewhHH/+o4V2/bV7Y3p6XDffc5I6gUXQGkpBAXB//7nsWAV4NedWSRn5PH3s/rRKz6cf8/fxI69BQD8c95G7v5gNQAdo0I8dk8REREREZGDKWD1gmEPz+e8//zI8m37SErP48mvNtb+hpQUuPNOSEyERx+FCRPg008hwPMD5B+v2sWFLywGYNLgDjx78VD3ufF940jNLgSgQ5tW+PsZj99fRERERETkAAWsTSCvqJTfPb+I9SnZ7rINqTlkF5QA0LpVINZa97nVO7PIKyqtuMCqVfD0086o6tq1MGsWDB5c5/uXl1vu+3A1d8xaddi6S5IyAbh5bA9iwoPp3S6C9dMm8vLk4bxw6TB3vTm3eW5EV0REREREpCYKWJvAim37WL5tH6c/vZC0nEJ3+a79TpptfnEZQx76iimzf6WwpIyzn/uBy+59C+6/36k4cSL89hvMnElZ7z7M+DGZfXnFdb7/gs3pvLl0Ox+s3MWL3//Gl2tTOePphWTkFlUJlHOLSklKz2N4QhR/mdDbXR4S5M+p/eIJ8PfjrxN6M7pnLG1CPbOfq4iIiIiIyKEccU6pMaYzMBNoB5QDL1lrnz6ozhjgYyDZVfSBtXbakd6zuTKVMmdf/D7JfTx9rpMK/P2mdADeXb6Dmz9+Fjqey8rAGCjb4L7AvNxg4nfsZ/f+Ah74dB3JGXk8OGnAYe+dW1TK1a8vJ9Df0Ld9ax6ds8F9bvjD8wF4efJwNqfluNtz4bBOh7zeTWN7cNPYHnX74CIiIiIiIg3QkEmQpcCd1tqVxpgIYIUx5itr7bqD6i201p7VgPs0e/vzS9zHZeW2lpowuuO57uPMe6fyy4Y9JMSEcd0bKwA4rV88AHuyiw5738KSMn5K3ktZuWXqpP6c0D2WO2atIi27yD0XFeCamcvdx52jQ5g8MrFOn0tERERERKQxHXHAaq1NAVJcxznGmPVAR+DggPWotz+/In03I9cJNDtFhbBzX0Gt75vw1AIycovp1jbMXfbluj0AbErLqfW9pWXl9Ll/LgD+foYLhnYiLDiAT24+EWstV874ibG941iwKZ2vN6QBsPTeU4hv3ar+H1BERERERKQReGQOqzEmERgCLK3h9EhjzC/GmDnGmP6euJ835RWVsnJ7HbehwVnw6P6P17pfb3XtbRq9I7lKvdE9Y6u9NyPXCXST0p339IgLd59L2V9YZf7pAT9uyeCBT9YyZNpX7rKusWGEBVf8bsIYw4wrj+PyExK5fXwvd3lcRHCdP5eIiIiIiEhja3DAaowJB94HbrfWZh90eiWQYK0dBDwLfFTLda41xiw3xixPT09vaLMazd0frObqGT9RUFxWp/qVR1HHxxjW7Hb+iqL3V/2MY3vHuY+vPalbtes8eE5/3vrTCM4e1IHzh3SkoKSMLNcqw68sTGLO6hT25RVzyStLmbFoKzmVVhluV8uo6cBObWjXuhV927fGGG1TIyIiIiIivqNBAasxJhAnWH3TWvvBweettdnW2lzX8RdAoDGm+lCic/4la+1wa+3wtm3bNqRZjeqSEV3Yl1/Cl+tSD1vXWssfXnL2NA0K8GNo6ib3uegzTgWcUc03/zSCySMT3OfuPaOv+/imsd359x8GcfkJicS1bsWzFw/hlL7OPNaNqTm8tXQ7D3++nhveXMmQhypGVV+6bBjv3zASgDG9a//7/O6vY/jwxhMO+3lERERERESaUkNWCTbAq8B6a+2Th6jTDthjrbXGmONwAuTMI72nLxjaJQp/P8OWtNxa6xUWFDH/tY9IyXLSeL+4dTTZmQPh9V8AiIwIcdcd1cOJ4W8Y050+7SKqXOeOU3vj71d15LNDpDNi+oeXllS7741junPXxD7u12sfnEBokH+tbW0VWPt5ERERERERb2jIKsGjgMuA1caYVa6ye4EuANbaF4DfATcYY0qBAuAiW9PEy2YkKMCPTlEhrNqxn4c/W8cdp/UiNKjSX2NxMcycyaNzNvF6z5PdxR0iW9EttiPgBKzRYdX3MZ1SKdB87YrhbNqTWy1YBejfoQ2hQf7kV0pLTvrHGWzfm0+X6NAqdSvPXRUREREREWlOGrJK8A9ArZMerbXPAc8d6T18VWJMGN9vSmfh5gx6xoeTEBPG8d1iwFo48UT46SeSr/63u35kaKA7qA0O8KOotJzI0CAADhW9j+sTz7g+8TWeCwrwo2d8BL/s2E+bkEAeOncAfn6GxNiwGuuLiIiIiIg0Rx5ZJfhoc3KvijmhU95fzUUvLWHOLzt58LN1bL35LzBnDm3GneSuM+e20e7jMwa2B6BNSPUR1vp4eNIAukSH8vmtJ3LOoA4NupaIiIiIiIgvUr7oEbhweCe27d7H6ytT3GU3vO2k+i5u145zunTg0183AnBc12jiIipW6X3sgoGcObC9e4uaI02QHtipDQvuGnuEn0BERERERMT3aYS1vgoLiXhkGg9eU3OwuCE1h+lznWB1dM9YZl03sso81OAAf8b3i6+00FGzntIrIiIiIiLSaBSw1ldQEHz0EUyYwOJz2jFlYh/CKq3CGxVakerbNiL4kJcJ8D8QxGrvUxERERERkZooJbi+/Pxg2TIICaE9cANwUq9YznzmBwB+/vtp7NpfwONzNlTZT/VgMWFBXD4ygd8f27lp2i0iIiIiItLMKGA9EiEhVV4emKMa4Er97RgZwjMXD6n1EsYYHpw0oHHaJyIiIiIi0gIoYPWA2PAgLj6uMxcO12ipiIiIiIiIpyhg9QBjDI+ef4y3myEiIiIiItKiaNElERERERER8UkKWEVERERERMQnKWAVERERERERn6SAVURERERERHySAlYRERERERHxSQpYRURERERExCcpYBURERERERGfpIBVREREREREfJICVhEREREREfFJClhFRERERETEJylgFREREREREZ+kgFVERERERER8kgJWERERERER8UkKWEVERERERMQnKWAVERERERERn6SAVURERERERHySAlYRERERERHxSQpYRURERERExCcZa62321CNMSYd2ObtdtQiFsjwdiPEY9SfLYv6s2VRf7Ys6s+WRf3Zsqg/WxZf788Ea23bulT0yYDV1xljlltrh3u7HeIZ6s+WRf3Zsqg/Wxb1Z8ui/mxZ1J8tS0vqT6UEi4iIiIiIiE9SwCoiIiIiIiI+SQHrkXnJ2w0Qj1J/tizqz5ZF/dmyqD9bFvVny6L+bFlaTH9qDquIiIiIiIj4JI2wioiIiIiIiE9SwFpPxpiJxpiNxpgtxpi7vd0eqZ0xprMx5ltjzHpjzFpjzG2u8geMMbuMMatcX2dUes89rv7daIyZ4L3WS02MMVuNMatd/bbcVRZtjPnKGLPZ9WeUq9wYY55x9eevxpih3m29VGaM6V3pGVxljMk2xtyu57N5Mca8ZoxJM8asqVRW72fSGHO5q/5mY8zl3vgsR7tD9OU/jTEbXP31oTEm0lWeaIwpqPScvlDpPcNc36e3uPrbeOPzHO0O0Z/1/v6qn319wyH6891KfbnVGLPKVd6ynk9rrb7q+AX4A78B3YAg4Begn7fbpa9a+6w9MNR1HAFsAvoBDwB/qaF+P1e/BgNdXf3t7+3Poa8qfbQViD2obDpwt+v4buBx1/EZwBzAAMcDS73dfn0dsl/9gVQgQc9n8/oCTgKGAmsqldXrmQSigSTXn1Gu4yhvf7aj7esQfXkaEOA6frxSXyZWrnfQdZYBI139PAc43duf7Wj8OkR/1uv7q3729Z2vmvrzoPP/Av7uOm5Rz6dGWOvnOGCLtTbJWlsMvANM8nKbpBbW2hRr7UrXcQ6wHuhYy1smAe9Ya4ustcnAFpx+F982CXjddfw6cG6l8pnWsQSINMa090YD5bBOAX6z1m6rpY6eTx9krV0A7D2ouL7P5ATgK2vtXmvtPuArYGLjt14qq6kvrbVfWmtLXS+XAJ1qu4arP1tbaxdb56fjmVT0vzShQzybh3Ko76/62ddH1NafrlHS3wNv13aN5vp8KmCtn47Ajkqvd1J78CM+xBiTCAwBlrqKbnalOL12IF0N9XFzYIEvjTErjDHXusrirbUp4PySAohzlas/m4+LqPofrZ7P5q2+z6T6tnm4CmdE5oCuxpifjTHfG2NGu8o64vTfAepL31Of7696NpuH0cAea+3mSmUt5vlUwFo/NeV4a5nlZsAYEw68D9xurc0Gnge6A4OBFJw0ClAfNwejrLVDgdOBm4wxJ9VSV/3ZDBhjgoBzgPdcRXo+W65D9aH61scZY+4DSoE3XUUpQBdr7RDgDuAtY0xr1Je+rr7fX9WfzcPFVP2lb4t6PhWw1s9OoHOl152A3V5qi9SRMSYQJ1h901r7AYC1do+1tsxaWw68TEVaofrYx1lrd7v+TAM+xOm7PQdSfV1/prmqqz+bh9OBldbaPaDns4Wo7zOpvvVhrkWwzgIucaUR4kodzXQdr8CZ59gLpy8rpw2rL33IEXx/1bPp44wxAcD5wLsHylra86mAtX5+AnoaY7q6RgQuAj7xcpukFq6c/leB9dbaJyuVV57HeB5wYMW1T4CLjDHBxpiuQE+cyeniA4wxYcaYiAPHOIuBrMHptwOril4OfOw6/gSY7FqZ9Hgg60CaoviUKr8Z1vPZItT3mZwHnGaMiXKlKJ7mKhMvM8ZMBKYA51hr8yuVtzXG+LuOu+E8j0mu/swxxhzv+j94MhX9L152BN9f9bOv7xsPbLDWulN9W9rzGeDtBjQn1tpSY8zNOP+J+gOvWWvXerlZUrtRwGXA6gNLfQP3AhcbYwbjpEFsBa4DsNauNcbMAtbhpD7dZK0ta/JWy6HEAx+6VmAPAN6y1s41xvwEzDLGXA1sBy501f8CZ1XSLUA+cGXTN1lqY4wJBU7F9Qy6TNfz2XwYY94GxgCxxpidwFTgMerxTFpr9xpjHsL54RhgmrW2rovFiIccoi/vwVk59ivX994l1trrcVYsnWaMKQXKgOsr9dkNwAwgBGfOa+V5r9JEDtGfY+r7/VU/+/qGmvrTWvsq1deAgBb2fBpXZoeIiIiIiIiIT1FKsIiIiIiIiPgkBawiIiIiIiLikxSwioiIiIiIiE9SwCoiIiIiIiI+SQGriIiIiIiI+CQFrCIiIiIiIuKTFLCKiIiIiIiIT1LAKiIiIiIiIj7p/wE99zDctykw1wAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<Figure size 1152x576 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "from sklearn.linear_model import LinearRegression\n",
    "diffs = np.array(strange_vals) - np.array(etf_dict['xlk'].iloc[-(1500 + 250):-250, :]['Close'].values)\n",
    "model = LinearRegression().fit(np.array(list(range(1500))).reshape((-1,1)),diffs)\n",
    "intercept = model.intercept_\n",
    "slope = model.coef_\n",
    "plt.figure(figsize=(16,8))\n",
    "x = np.linspace(0, 1750, 1000)\n",
    "plt.plot(x, intercept + slope * x, '--r', label = 'Reggression Line')\n",
    "plt.plot(diffs, label = 'Translation Error')\n",
    "plt.title('Naive Translation Error over time')\n",
    "plt.legend(loc='upper left')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Now, let's try to translate the actual values of our stocks in the prediction period using the suggested method:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.legend.Legend at 0x1be961bc080>"
      ]
     },
     "execution_count": 36,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAA6gAAAHiCAYAAADlHeELAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADl0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uIDMuMC4yLCBodHRwOi8vbWF0cGxvdGxpYi5vcmcvOIA7rQAAIABJREFUeJzs3Xd0VVX6//H3TkIJJBCa0qRYqCG5hCT03hFRUGkq2BV07IUvo8Iw2EZHEWV+DBZgRgaiIogOMIiCIL0YkCogSG8hhJZAyv79cW4uCekk4Sbwea3lOufss/c+zzn3spZP9j77GmstIiIiIiIiIt7m4+0AREREREREREAJqoiIiIiIiBQRSlBFRERERESkSFCCKiIiIiIiIkWCElQREREREREpEpSgioiIiIiISJGgBFVEpAgwxrQ1xmz3dhzXAmPMSGPMJ96OwxuMMX7GGGuMqZNNne3GmLZXLqqizxjTwRiz2dtxiIhcC5SgiohcQcaYPcaYLpeWW2uXWmvreyOmSxljRhtjEo0xZ4wxJ40xy40xLb0dV0Gx1r5hrX24IPt0J71n3P8lGGOS0xwX2cTGGPO5MWZ02jJrbX1r7dJCul5PY8xSY8xpY8xRY8xiY8ythXGtgmStXWytbeztOERErgVKUEVErmHGGL8sTkVZawOAysAi4MsrF9VFxhhfb1w3r9xJb4D7mT0OrEg9ziyxyea5X7WMMQOBKOAzoAZQDfgL0MebceXkWvysRES8SQmqiEgR4J5CuD/N8R5jzAvGmI3GmDhjTJQxpnSa872NMdFpRjhD0pwbYYzZ5R6l2mKM6Zvm3P3GmGXGmPeNMSeA0dnFZa1NAqYBNYwxVXJ5/TBjzC/u63/pjn1smvMvGWMOGWMOGmMedk85vdl9boox5v8ZY+YaY84CHY0xpYwx7xpj9hpjjhhjJhpj/N31KxtjvnPHccI9OufjPveyMeaAO47txpjO7vLRxpjP08TTxxiz2d3HYmNMw9x+DrmVZmrtcGPMTmCbu/wjY8x+Y8wpY8waY0yrNG3GGmOmu0c5TxtjNhljwtKcH+l+hqeMMduMMR3c5S2NMSvd93PIGDPeGFMik5iGAwOA1NHfWe7y/Wn6Ku1uf8j9LN8zxpR0n+vifj4vGWOOuWMZksX9+wB/B0ZZaydba09Za5OttYustY+l1jHGvGaM+cM4o6tTjDHl3Odudj+/+93xnTDGPGKMaW6M+dV9rx+kud7Dxpglxph/uD+3rcaYjpec3+p+rruMMQ+nOZd6XyONMYeBj1PLcvHsC+R5iYhcy5SgiogUXf2BHkBdIAS4H5wEEGcU6jGgEvBPYI4xppS73S6gLVAeZ4Tqc2NMtTT9Ngd+B64DXs8uAPf/XA8BYoDYnK7vrj8LmAJUBKYDaRPkHsBzQBfgZqB9Jpcd7I4rEPgZeBuoB7jcbWoAr7nrPg/sB6oA1wMjAWuMqQ88CURYawOB7sCeTO6vnjvGZ9x9zAW+TU0q3DL9HC5THyACaOI+XuXusyLwFfBlms8R4A7g30AQMA8Y7467Mc7zD7PWlgN6AnvdbZKAp3FGv1u7Y3/s0kCstf/AGdFMHf3te2kdnOcc7o6xqbu//0tzvibgD1THGTn+f6lJ5SUauet8ldlDcXsYuBfoANwEVAA+uKROuPvcvTjPYgTQCQgG7jXGtE5TtxXOHwIqA38FZhljgtznjgC3AuWAR4APTZo/srjvKwCoBQxPG0AOz76gnpeIyDVLCaqISNE13lp70Fp7AvgWJ0ED53+o/2mtXeUehZoKnAdaAFhrv3S3S7HWRgE7gMg0/R601n5orU2y1sZnce3+xpiTQLz7ene5R1Nzun4LwM8de6K19mtgddp+gcnW2s3W2nM4CfSlvrHWLrPWprj7fQR41lp7wlp7GngDGOium4gzVbS2+3pLrbUWSAZKAY2MMSWstXustbsyudYA4L/W2u+ttYnAuzgJRKs0dbL6HC7HG9ba2NTnbq39t/u+koC/4SRMN6ep/5O19n/W2mScRDX12klAaaCxMcbPWrvbWvu7u8817s8myV02icz/EJAb9wCjrbXHrLVHgTHAfWnOJwBj3c9+Ds7nVS+Tfiq5t4dyuNa77ns5jfPHhsHu0ddUf7XWnrfWzgUuAJ+7Y9uP88eMpmnqHgI+dMf2H5w/yvQEsNZ+a6393Tp+BH7A+aNOqiT3fV/I5N9Ils++AJ+XiMg1SwmqiEjRdTjN/jmcER2A2sDz7mmNJ92J5A04ozIYY4aYi9NvT+KMLlVO09e+XFz7C2ttEM6o5CagWZpz2V2/OnDAnSRmdr3qlxxnFkvasipAGWBdmmvNd5cDvAPsBBYYY343xowAsNbuxBkVHQ0cNcbMMMZUz+Ra1YE/Ug/cSfE+nFHaVFl9Dpcj3f26p3tuM8bE4YxQlyX9Z3Xptcu649yOM3o8Buf+phtjqrr7bGCM+a8x5rAx5pS7Tto+86IaaZ6Pez/tsznuTp7TxpjZ84lJ019W0n0W7v2SXPyssdYeSXM+HmckNO1x2mvvv+R7+AcX/430Nsasck8VPgl0I/0zOmKtvZBZkNk9ewrueYmIXLOUoIqIFD/7gNettUFp/itjrZ1ujKkNfIwzvbWSO8ncBJg07W0mfWbKWnscZzrj6DTThLO8Ps6oVQ1jTNrr3ZBm/xDONMfMzmUW33GcxKNxmmuVdy9GhLX2tLX2eWvtjcBtwHPG/a6ptfY/1to2OAm1xZkqfKmD7vMAuOO+ATiQ7YO5fJ57c78T+RxwJ84U3grAGdJ/Vll3ZO3n1trWOFOPfYE33af+ifOZ3+yegvpaNn3m9F04RJrngzPl9XKezRacZ31nNnXSfRbua10Ajl3G9SD99yy1v4PGeX/5K5zndb3738gC8vBvJJtnX1DPS0TkmqUEVUTkyivhXkwl9b+8rhL6MfC4e4EYY4wpa4y51RgTiDPCZnH/T70x5gGcEdTLZq3dBvwPeCkX11+BM732SeMsDHQ76acXfwE8YIxpaIwpw8V3SbO6dor7eu8bY65z31MNY0x3935v4yygY4BT7msnG2PqG2M6ud/nTMBJcpMzucQXwK3GmM7GWUjoeZxpl8vz+pwuQyDOdNHjQAmc0d6yuWnofn4d3fcXT/r7CwTigLPGWfApw/unaRwBbszm/HTgNeMsRlUFeBX4PJv6mXJ/js/j/KFjqDGmnHEWRWprjJmY5lrPGWPquL9LrwPT3W0vRzVjTOr3cCDOu6vzcaZ+l8T5N5JsjOkNdM5tpzk8+wJ5XiIi1zIlqCIiV95cLv6PbTw5rKR7KWvtWpz3Mj/CmRa6E/fCPdbaLTirpa7AST6aAMsKIOZ3gEeNMdflcP0LQD/gIeAkzmI23+EkfVhrUxf6WeRut8Ld//lsrv2yu+5K95TVhUDqb8be4j4+4+7rH9baxThJyFs4yd9hnAWhRl7asXu65r3Ah+66twG3ZTW9s4DNdce+A2cBp1Nk/45mWqVw3llNvb8KwCvuc88DQ4HTOKOpUdn08wkQaoyJNcZktoDRX4ANwK/ARpxFnd7MpF6OrLUzcBbAegRntPQwzjTZb9xVPnbHuhTnfdHTOIs9Xa7lQGMgdbXqO93v/54EnsVZzOsEcBfOdzS3snv2Bfa8RESuVSb96xkiIiIFyxizCphorZ2cybmGONNRS6VZhEkkX4zzszH3Wms7eDsWERHJG42giohIgTLGtDfGVHVPrRyK85Mb89Oc72uMKWmMqYDzXui3Sk5FREQElKCKiEjBq48zzTEOZ7rpXdbatFNXH8N5/28Xzrt7w654hCIiIlIkaYqviIiIiIiIFAkaQRUREREREZEiQQmqiIiIiIiIFAl5/e29QlG5cmVbp04db4chIiIiIiIihWDdunXHrbVVcqpXJBLUOnXqsHbtWm+HISIiIiIiIoXAGPNHburlOMXXGPOZMeaoMWZTmrKKxpjvjTE73NsK7nJjjBlvjNlpjNlojAm7/FsQERERERGRa0lu3kGdAvS4pGwE8IO19hbgB/cxQE/gFvd/jwL/r2DCFBERERERkatdjgmqtXYJcOKS4tuBqe79qcAdacr/ZR0rgSBjTLWCClZERERERESuXpf7Dur1qT+6bq09ZIy5zl1eA9iXpt5+d9mhS9pjjHkUZ5SVWrVqZbhAYmIi+/fvJyEh4TJDFCl6SpcuTc2aNSlRooS3QxERERERKXIKepEkk0mZzayitXYSMAkgPDw8Q539+/cTGBhInTp1MCazbkWKF2stMTEx7N+/n7p163o7HBERERGRIudyfwf1SOrUXff2qLt8P3BDmno1gYOXc4GEhAQqVaqk5FSuGsYYKlWqpFkBIiIiIiJZuNwEdQ4w1L0/FPgmTfkQ92q+LYC41KnAl0PJqVxt9J0WEREREclabn5mZjqwAqhvjNlvjHkIeAvoaozZAXR1HwPMBX4HdgIfA8MLJeorICYmBpfLhcvlomrVqtSoUcNzfOHChQK5xty5cz19BgQEUL9+fVwuFw888ECu+7hw4QIdO3bMsd6QIUPYtWtXfsIFnJFtPz8/XC4XwcHBDBo0KM8jgnmNZeLEibzwwgvpyrZt28aNN96YoW6DBg3YuHFjnvoSEREREZGiIcd3UK21g7I41TmTuhZ4Ir9BFQWVKlUiOjoagNGjRxMQEJAhsbHWYq3Fx+fyBqJ79epFr169AGjTpg0fffQRLpcrQ72kpCT8/DL/qEqWLMmiRYtyvNa//vWvy4oxM0FBQURHR2Ot5a677uKzzz5j+PDc/S0iOTm5QGJp0KAB5cuXZ9WqVTRv3hyA6OhofHx8CAkJyXf/IiIiIiJy5V3uFN9r1s6dOwkODubxxx8nLCyMffv2ERQU5Dk/Y8YMHn74YQCOHDlCv379CA8PJzIykpUrV+b6OhMnTmTw4MHceuut9O7dm9jYWDp27EhYWBghISHMmzcPcEY0K1euDMD8+fPp2rUrffv2pX79+ulGYlu0aMGmTZs89V988UVCQkJo3bo1x48fB2D79u1EREQQGRnJK6+84uk3K8YY2rZty86dOwH47LPPiIyMJDQ0lCeffBJrred6//d//0dkZCTr16/3xAIwdepUgoODCQ4O5tVXX/X0/c9//pN69erRoUMHVq9enen1Bw0axIwZM9I9+8GDBwPw9ddf07x5c1wuFz169PDcY1oDBw7ku+++8xwHBAR49l9//XUiIiIICQnhjTfeAODUqVN0796d0NBQgoODmT17drbPR0RERERE8qagV/EtFM88A+7BzALjcsG4cZfXdsuWLUyePJmJEyeSlJSUZb2nnnqKl156iRYtWrBnzx569+7tScxyY8WKFfzyyy8EBQVx4cIFvv32WwICAjhy5Aht27alZ8+eGdqsX7+ezZs3U6VKFSIiIli7di3h4eHp6sTExNCtWzfeeecdnnrqKaZMmcILL7zAE088wciRI+nbty/jcvFwEhMTmT9/PgMGDGDDhg189913LF++HD8/Px588EG+/PJL+vTpQ0xMDC1btuTNN99M1/6PP/5g9OjRrFu3joCAADp16sT8+fNp2LAhb731FuvWraNs2bK0adOG9u3bZ7j+wIEDad26NX//+9/x8fHhiy++YMGCBQB06NCBvn37Yozho48+4v333+f111/P1XOfM2cOhw8fZvXq1aSkpNCjRw+6dOnCjh07qFevHv/73/+w1nL69Olc9SciIiIiIrlTLBLUouamm24iIiIix3oLFy5k+/btnuPY2Fji4+Px9/fP1XV69OjhGZ211vLCCy+wbNkyfH19+eOPPzh58iSlS5dO16ZVq1ZUrVoVgNDQUPbs2ZMhQQ0ICKBr164ANGvWjDVr1mCtZe3atdxxxx0ADB48mLFjx2Ya18mTJz1TkTt27Mh9993HuHHjWL16teda8fHx3HLLLQD4+/vTp0+fDP2sXLmSrl27UrFiRcAZEV2yZAknT55MV3733Xdz9OjRDO1r1apF7dq1Wbp0KaVKlaJy5crcfPPNgJP89u/fnyNHjpCQkEDjxo2zfdZpLViwgLlz57J06VIAzpw5w2+//YbL5eKVV14hICCAPn360LJly1z3KSIiIiIiOSsWCerljnQWlrJly3r2fXx8cF69daRdMMhay+rVqylZsmS+rzN58mTOnTvHL7/8gp+fH1WrViUhISFDglqqVCnPvq+vb6YjvFnVye0Ks6nvoKZlreWRRx5h1KhR6coTEhIoU6ZMpv2kfW6Xym0sqdN8S5UqxaBBF1+XHjZsGGPGjKFbt27Mnz8/0xFhPz8/UlJSAGc0OHXfWsuoUaMYOnRohjZr1qzhv//9L88++yz9+vXjpZdeylWcIiIiIiKSM72Dmk8+Pj5UqFCBHTt2kJKSwqxZszznunTpwoQJEzzHlyZ1eREXF8f111+Pn58f8+fP58iRI/mK+1LGGMLCwpgzZw5Aunc7c6Nr167MmDGDmJgYAI4fP86+ffuybdOyZUsWLlxIbGwsiYmJREVF0b59e1q1asX3339PbGws58+fZ+bMmVn2cffddzN79mxmzpxJ//79PeVxcXHUqFEDay1Tp07NtG2dOnVYt24dADNnzvQkzN27d+eTTz7h3LlzAOzdu5eYmBj2799PYGAgQ4cO5ZlnnmH9+vW5f0AiIiIiIpKjYjGCWtS9/fbb9OjRg1q1atGoUSPOnz8PwIQJExg2bBiTJ08mKSmJjh07pktY82Lo0KH06dOHiIgIwsPDqVu3bkHeAgAfffQR9913H2+88QY9evSgfPnyuW7rcrkYOXIknTp1wlpLyZIlmTRpElWqVMmyTe3atXnttddo27YtAHfccQfdu3cH4OWXX6Z58+bUqFEj2+nU1113HSEhIZw/f54aNWp4yseMGUPv3r254YYbiIiIIDY2NkPbYcOGcccddzBv3jx69uyJr68vAH369GH79u2e1YHLlSvH9OnT2bBhAyNHjsTPz89zfyIiIiIiUnBMdtMsr5Tw8HC7du3adGVbt26lYcOGXoro2nT27FnKlCmDMYYpU6Ywb948oqKivB3WVUffbRERERG51hhj1llrw3OqpxFU8VixYgXPP/88KSkpVKpUiSlTpng7JBERERERuYYoQRWPLl26sGHDBm+HISIiIiIiubHnP7D8Huh3BEpf5+1oCoQWSRIRERERESmOtrztbOO2eDeOAqQEVUREREREpLixKXByo7N/5nfvxlKAlKCKiIiIiIgUNwe+u7i/6iFITvBeLAVI76CKiIiIiIgUNyfcv4JSow+UqgiJZ8C3tHdjKgAaQc1CTEwMLpcLl8tF1apVqVGjhuf4woULBXadhQsXUr58eVwuFw0bNuT111/PV3+vvPIK48aNA+DPf/4zixYtyrLu+vXrmT9/vud41qxZvPPOO/m6PkBSUhK+vr64XC6Cg4MZMGAA8fHxeerjgQceYPv27bmu/8knn/DMM8+kK9u5cye1a9fm0p9SCg4OZv369XnqS0RERESkSDm5Eco1gPbfQIvJULqytyMqEBpBzUKlSpWIjo4GYPTo0QQEBPDCCy+kq2OtxVqLj0/+8vyOHTsye/Zszpw5Q0hICL179yY0NNRzPikpCT+/vH9UOSW769evZ9OmTfTo0QOAvn375vkaWQkMDCQ6OhprLQMHDuTjjz/mqaeeylXb5ORkJk+enO8Ybr75Zq677jqWL19O69atAdi8eTOJiYmEhYXlu38REREREa85+wcE3OztKAqcRlDzaOfOnQQHB/P4448TFhbGvn37CAoK8pyfMWMGDz/8MABHjhyhX79+hIeHExkZycqVK7PtOyAggLCwMHbt2sUnn3zCwIED6d27Nz179gTgrbfeIjIykpCQEMaMGeNpN2bMGOrXr0/Xrl3ZsWOHp/zee+9l9uzZAKxatYqWLVsSGhpK8+bNOXv2LGPGjGHatGm4XC6++uqrdCOHu3fvpmPHjoSEhNC1a1f279/v6fPpp5+mVatW3HjjjcyaNSvbezLG0LZtW3bu3AnA1KlTiYyMxOVyMXz4cFJSUkhKSiIoKIhXXnmFyMhIVq9eTZs2bTx/IPj8889p0qQJwcHBjBw50tP3J598Qr169ejQoUOWz3bQoEHMmDHDczx9+nQGDRoEwDfffEPz5s1p2rQp3bp14+jRoxnap32GqZ9Rqsw+j9OnT9OzZ09CQ0MJDg7mq6++yvb5iIiIiIhclvMnoFQlb0dR4IrFCOoz858h+nB0gfbpqupiXI9xl9V2y5YtTJ48mYkTJ5KUlJRlvaeeeoqXXnqJFi1asGfPHnr37s2mTZuyrH/s2DFWr17N66+/ztKlS1mxYgXR0dFUqFCBuXPnsnfvXlatWoW1ll69erF8+XL8/PyYOXMm0dHRXLhwAZfLRcuWLdP1m5CQwMCBA5k5cyZhYWHExcVRunRpXnvtNTZt2uSZEvzJJ5942gwfPpyHH36Ye+65h0mTJvHMM894kq2jR4+ybNkyfv31V/r375/tyGtiYiLz58/n9ttvZ9OmTcyaNcsT96OPPsqMGTPo378/cXFxhIWFMXbs2HTt9+/fzyuvvMLatWspX748Xbp04bvvvsPlcvHXv/6V9evXExgYSLt27WjRokWG6w8YMICIiAjGjRuHr68vUVFRfPvttwC0a9eOPn36YIxh4sSJ/P3vf+ftt9/O8l7Syurz2LdvH3Xq1GHevHkAxMXF5ao/EREREZE8uRALJSt4O4oCVywS1KLmpptuIiIiIsd6CxcuTPceZWxsLPHx8fj7+6ert2jRIpo2bYqPjw+vvvoq9evXZ+nSpXTr1o0KFZwv3YIFC5g3bx5NmzYF4MyZM/z2228cP36cO++8E39/f/z9/bntttsyxLF161Zq1arlmdZavnz5HGNftWoV333nrAw2ZMgQXn31Vc+5O+64A2MMISEhHDhwINP2p0+fxuVyAdC+fXvuv/9+JkyYwJo1awgPDwcgPj6eG264AYCSJUtmmuiuWrWKTp06UbmyM6d+8ODBLFmyhISEBDp37kylSs5fjfr378/evXsztK9Rowb16tVj8eLFlC9fnsDAQBo0aADA3r176d+/P4cPH+b8+fPUq1cvx+eSKqvPo3nz5owYMYIRI0Zw2223eaYWi4iIiIgUmJRESDoNJSt6O5ICVywS1Msd6SwsZcuW9ez7+PikW4QnIeHi8s7WWlavXk3JkiWz7S/1HdTsrmOt5ZVXXuGhhx5KV+fdd9/FGJNt/9baHOvkRalSpdL1nZnUd1AvjePBBx/kr3/9a7rypKQk/P39M40xq/6BXN9T6jTf8uXLe6b3AjzxxBOMHDmSXr16sXDhQt56660Mbf38/EhJSQGcd2NTR8yz+jwA1q5dy9y5c3nxxRfp3bt3umnJIiIiIiL5diHW2V6FI6h6BzWffHx8qFChAjt27CAlJSXdO5ldunRhwoQJnuNLE7a86N69O59++ilnz54FnKmvx48fp127dnz99dckJCRw6tQpz6hnWo0bN+aPP/7wrFx76tQpkpOTCQwM5PTp05ler0WLFnzxxReA8w5ou3btLjv2VF26dOGLL77g+PHjgLNScmajnpfGsWjRImJiYkhKSmLGjBm0b9+eFi1a8MMPP3DixAkuXLiQ7bued911F99++y1ffvklAwYM8JTHxcVRo0YNrLVMnTo107Z16tRh3bp1gLPKcXJyMpD153HgwAECAgK47777eO6557JdLVhERERE5LKkJqilNIIqmXj77bfp0aMHtWrVolGjRpw/fx6ACRMmMGzYMCZPnkxSUhIdO3ZMl7DmRa9evdi2bZvnPcvAwED+85//EBkZSd++fQkNDaVOnTqZJpKlSpVi+vTpDBs2jISEBPz9/fnxxx/p1KkT77zzDk2bNuXPf/5zujYfffQRDz30EG+++SbXX399gayq26RJE0aNGkWXLl1ISUmhRIkSTJw4kerVq2fZpmbNmowZM4YOHTpgreW2227j1ltvBZyf1GnRogXVq1f3TBvOTKVKlWjWrBlxcXHUqlXLUz569Gj69u1LzZo1iYyM5NChQxnaPvbYY9x+++18//33dOvWzTN6nNXnsWXLFkaMGIGPjw8lS5Zk4sSJl/WsRERERESydD7G2V6FI6gmuymUV0p4eLhdu3ZturKtW7fSsGFDL0UkUnj03RYRERGRy5aSCPOaQtxm6L0dyuV+HRVvMsass9ZmParkpim+IiIiIiIixcX+2U5yChBwk3djKQRKUEVERERERIqLQwucrX8N8PH1biyFQAmqiIiIiIhIcXFmF1QIg9t+83YkhUIJqoiIiIiISHFxdh8E3gJ+ZbwdSaFQgioiIiIiIlIcWAvx+6FMTW9HUmiUoIqIiIiIiBQH52MgOUEJ6rXq8OHDDBw4kJtuuolGjRrRq1cvfvvtys31jo6OZu7cuXlu16FDBy792Z7U8vr16xMaGkrr1q3Zvn17vuILCAgA4ODBg9x1113Z1h03bhznzp3zHPfq1YuTJ0/m6/rg/JZpjRo1cLlcBAcHM2fOnDy1X7t2LU899VSe2tSpU4fjx4/nqY2IiIiISL4lnXa2Jcp7N45CpAQ1C9Za+vbtS4cOHdi1axdbtmzhjTfe4MiRI7lqn5ycnKG/lJSUPMVwuQlqdqZNm8aGDRsYOnQoL774Yobzl8adG9WrV+err77Kts6lCercuXMJCgrK87Uy8+yzzxIdHc2XX37Jgw8+mOvnnJSURHh4OOPHjy+QOEREREREClVygrP19fduHIVICWoWFi1aRIkSJXj88cc9ZS6Xi7Zt22Kt5cUXXyQ4OJgmTZoQFRUFwOLFi+nYsSODBw+mSZMm7Nmzh4YNGzJ8+HDCwsLYt28fCxYsoGXLloSFhXH33Xdz5swZANasWUOrVq0IDQ0lMjKSuLg4XnvtNaKionC5XERFRXH27FkefPBBIiIiaNq0Kd988w0A8fHxDBw4kJCQEAYMGEB8fHyO99euXTt27twJOCOCY8aMoU2bNnz55Zfs2rWLHj160KxZM9q2bcu2bdsA2L17Ny1btiQiIoJXX33V09eePXsIDg4GnAT3hRdeoEmTJoSEhPDhhx8yfvx4Dh48SMeOHemVLDKRAAAgAElEQVTYsaPnmqmjkO+99x7BwcEEBwczbtw4T58NGzbkkUceoXHjxnTr1i3H+2rYsCF+fn4cP36cY8eOceeddxIREUFERATLli0DnBHXRx99lG7dujFkyBAWL15M7969AThx4gR33HEHISEhtGjRgo0bNwIQExNDt27daNq0KY899hjW2hyfr4iIiIhIgUt2//+wb2nvxlGI/LwdQO48A0QXcJ8uYFyWZzdt2kSzZs0yPff1118THR3Nhg0bOH78OBEREbRr1w6A1atXs2nTJurWrcuePXvYvn07kydP5h//+AfHjx9n7NixLFy4kLJly/L222/z3nvvMWLECAYMGEBUVBQRERGcOnWKMmXKMGbMGNauXctHH30EwMiRI+nUqROfffYZJ0+eJDIyki5duvDPf/6TMmXKsHHjRjZu3EhYWFiOd//tt9/SpEkTz3Hp0qX5+eefAejcuTMTJ07klltuYdWqVQwfPpwff/yRp59+mmHDhjFkyBAmTJiQab+TJk1i9+7d/PLLL/j5+XHixAkqVqzIe++9x6JFi6hcuXK6+uvWrWPy5MmsWrUKay3Nmzenffv2VKhQgR07djB9+nQ+/vhj+vfvz8yZM7n33nuzvKdVq1bh4+NDlSpVuOeee3j22Wdp06YNe/fupXv37mzdutVzzZ9//hl/f38WL17saT9q1CiaNm3K7Nmz+fHHHxkyZAjR0dH85S9/oU2bNrz22mv897//ZdKkSTk+XxERERGRAncNjKAWkwS1aPn5558ZNGgQvr6+XH/99bRv3541a9ZQrlw5IiMjqVu3rqdu7dq1adGiBQArV65ky5YttG7dGoALFy7QsmVLtm/fTrVq1YiIiACgXLlymV53wYIFzJkzh3fffReAhIQE9u7dy5IlSzzvUYaEhBASEpJl7Pfccw/+/v7UqVOHDz/80FM+YMAAAM6cOcPy5cu5++67PefOnz8PwLJly5g5cyYA9913Hy+//HKG/hcuXMjjjz+On5/z1apYsWKWsYDzLPv27UvZsmUB6NevH0uXLqVPnz7UrVsXl8sFQLNmzdizZ0+mfbz//vt8/vnnBAYGEhUVhTGGhQsXsmXLFk+dU6dOcfq0M2e/T58++Ptn/Ef9888/e+6vU6dOxMTEEBcXx5IlS/j6668BuPXWW6lQoUK29yQiIiIiUig8CapGUL0s65HOwtK4ceMs36vMbopnaqKV2bG1lq5duzJ9+vR0dTZu3IgxJseYrLXMnDmT+vXrZziXm/bgvIMaHh6eZdwpKSkEBQURHZ35iHVO17HW5jqW1PpZKVWqlGff19c3yym+zz77LC+88EK6spSUFFasWJFpInrpZ5RdLKn3kpd7EhEREREpFJ4pvlfvCKreQc1Cp06dOH/+PB9//LGnbM2aNfz000+0a9eOqKgokpOTOXbsGEuWLCEyMjLHPlu0aMGyZcs8736eO3eO3377jQYNGnDw4EHWrFkDwOnTp0lKSiIwMNAz6gfQvXt3PvzwQ08i9csvvwDO+6TTpk0DnKnJqe9OXo5y5cpRt25dvvzyS8BJ2jZs2ABA69atmTFjBoDnepfq1q0bEydOJCkpCXDe6wQy3Euqdu3aMXv2bM6dO8fZs2eZNWsWbdu2vez408aROjUayDLhvjSW1PtavHgxlStXply5cunK582bR2xsbL7jExERERHJs2tgBFUJahaMMcyaNYvvv/+em266icaNGzN69GiqV69O3759CQkJITQ0lE6dOvG3v/2NqlWr5thnlSpVmDJlCoMGDfIsxLNt2zZKlixJVFQUf/rTnwgNDaVr164kJCTQsWNHtmzZ4lkk6dVXXyUxMZGQkBCCg4M9CxUNGzaMM2fOEBISwt/+9rdcJcvZmTZtGp9++imhoaE0btzYsxjTBx98wIQJE4iIiCAuLi7Ttg8//DC1atXyPJ///Oc/ADz66KP07NnTs0hSqrCwMO6//34iIyNp3rw5Dz/8ME2bNs1X/ADjx49n7dq1hISE0KhRIyZOnJhjm9GjR3vajBgxgqlTpwLOu6lLliwhLCyMBQsWUKtWrXzHJyIiIiKSZ9fACKopCiuShoeH20t/t3Pr1q00bNjQSxGJFB59t0VERETksuz8BFY/ArfvhbI3eDuaPDHGrLPWZnzX8BIaQRURERERESkOroFVfJWgioiIiIiIFAfXwO+gKkEVEREREREpDrRIkoiIiIiIiBQJyfFg/MCnmPxa6GVQgioiIiIiIlIcJJ29qt8/hXwmqMaYp40xm4wxm40xz7jLKhpjvjfG7HBvKxRMqCIiIiIiIteoXZ/Bb+OhTE1vR1KoLjtBNcYEA48AkUAo0NsYcwswAvjBWnsL8IP7uNjZs2cPwcHB6cpGjx7Nu+++66WI0lu8eDHLly/Pc7s6depw/PjxTMubNGlCaGgo3bp14/Dhw5cdW9pnt3btWp566qls67/xxhvpjlu1anXZ107r/vvvp27durhcLsLCwlixYkWe2s+ZM4e33norT20CAgLyVF9EREREJEf758Cqh5z9gLrejaWQ5WcEtSGw0lp7zlqbBPwE9AVuB6a660wF7shfiFefpKSkdMfJycl57uNyE9TsLFq0iA0bNhAeHp4haYTLizM8PJzx48dnW+fSaxXkfb3zzjtER0fz1ltv8dhjj+W6XVJSEn369GHEiGL59xURERERuZocmu9sr+8Ikf/0biyFLD8J6iagnTGmkjGmDNALuAG43lp7CMC9vS7/YRY9HTp04OWXXyYyMpJ69eqxdOlSABISEnjggQdo0qQJTZs2ZdGiRQBMmTKFu+++m9tuu41u3bqxePFiOnbsyODBg2nSpAkAn3/+OZGRkbhcLh577DFPQjh//nzCwsIIDQ2lc+fO7Nmzh4kTJ/L+++/jcrlYunQpx44d48477yQiIoKIiAiWLVsGQExMDN26daNp06Y89thjWGtzvLd27dqxc+dOwBkRfO2112jevDkrVqxg3bp1tG/fnmbNmtG9e3cOHToEwLp16wgNDaVly5ZMmDDB09fixYvp3bs3AGfOnPE8m5CQEGbOnMmIESOIj4/H5XJxzz33eK4JYK3lxRdfJDg4mCZNmhAVFeXps0OHDtx11100aNCAe+65J8f7SntPu3btokePHjRr1oy2bduybds2wBlxfe655+jYsSMvv/wyU6ZM4cknnwTgjz/+oHPnzoSEhNC5c2f27t0LwO7du2nZsiURERG8+uqrOT5bEREREZE8SUmG3Z9DhabQ+cerforvZS//ZK3daox5G/geOANsAJKyb3WRMeZR4FGAWrVqZV953TMQG325oWaugguajctXF0lJSaxevZq5c+fyl7/8hYULF3qSs19//ZVt27bRrVs3fvvtNwBWrFjBxo0bqVixIosXL2b16tVs2rSJunXrsnXrVqKioli2bBklSpRg+PDhTJs2jZ49e/LII4+wZMkS6taty4kTJ6hYsSKPP/44AQEBvPDCCwAMHjyYZ599ljZt2rB37166d+/O1q1b+ctf/kKbNm147bXX+O9//8ukSZNyvK/vvvvOkzSfPXuW4OBgxowZQ2JiIu3bt+ebb76hSpUqREVF8ec//5nPPvuMBx54gA8//JD27dvz4osvZtrvX//6V8qXL8+vv/4KQGxsLHfeeScfffQR0dEZP9+vv/6a6OhoNmzYwPHjx4mIiKBdu3YA/PLLL2zevJnq1avTunVrli1bRps2bbK8p2+//dZzT48++igTJ07klltuYdWqVQwfPpwff/wRgN9++42FCxfi6+vLlClTPO2ffPJJhgwZwtChQ/nss8946qmnmD17Nk8//TTDhg1jyJAh6RJzEREREZF8S0mGFfdC0mkoH5xz/atAvtYnttZ+CnwKYIx5A9gPHDHGVLPWHjLGVAOOZtF2EjAJIDw8POdhvSvMGJNjeb9+/QBo1qwZe/bsAeDnn3/mT3/6EwANGjSgdu3angS1a9euVKxY0dM+MjKSunWdOeQ//PAD69atIyIiAoD4+Hiuu+46Vq5cSbt27Tz10rZPa+HChWzZssVzfOrUKU6fPs2SJUv4+uuvAbj11lupUCHrNas6duyIr68vISEhjB07FgBfX1/uvPNOALZv386mTZvo2rUr4Ez5rVatGnFxcZw8eZL27dsDcN999zFv3rxMY5wxY4bnOLtYwHmWgwYNwtfXl+uvv5727duzZs0aypUrR2RkJDVrOn89crlc7NmzJ9ME9cUXX2Ts2LFUqVKFTz/9lDNnzrB8+XLuvvtuT53z58979u+++258fX0z9LNixQrPc7zvvvt46aWXAFi2bBkzZ870lL/88svZ3pOIiIiISK4d/A7+cP//cz4H14qLfCWoxpjrrLVHjTG1gH5AS6AuMBR4y739Jt9ReuHDqFSpErGxsenKTpw44UkUAUqVKgU4SVzqe6XZTTUtW7ZslsfWWoYOHcqbb76Zrs6cOXOyTJbTSklJYcWKFfj7Z1x2OjftwXkHtXLlyunKSpcu7UnYrLU0btw4w2JDJ0+ezNU1rLW5jiW1flZSnz2kf/6Xeuedd7jrrrs8x6dOnSIoKCjTEVvI+BllJe195OWeRERERERy7fhKZ9vuGyiV+UDV1Sa/v4M60xizBfgWeMJaG4uTmHY1xuwAurqPi52AgACqVavGDz/8ADjJ6fz587OdRgrOu47Tpk0DnOmie/fupX79+jler3Pnznz11VccPXrUc70//viDli1b8tNPP7F7925POUBgYCCnT5/2tO/WrRsfffSR5zg1AUsbz7x58zIk3XlRv359jh075klQExMT2bx5M0FBQZQvX56ff/4ZwHO9S10aY2osJUqUIDExMUP9du3aERUVRXJyMseOHWPJkiVERkZedvwA5cqVo27dunz55ZeAkwRv2LAhx3atWrXyjP5OmzbN8z1o3bp1unIRERERkQITswoqhkPNPt6O5IrJV4JqrW1rrW1krQ211v7gLoux1na21t7i3p4omFCvvH/961+MHTsWl8tFp06dGDVqFDfddFO2bYYPH05ycjJNmjRhwIABTJkyJd1oX1YaNWrE2LFj6datGyEhIXTt2pVDhw5RpUoVJk2aRL9+/QgNDWXAgAEA3HbbbcyaNcuzSNL48eNZu3YtISEhNGrUiIkTJwIwatQolixZQlhYGAsWLMj5fd9slCxZkq+++oqXX36Z0NBQXC6XZ8XdyZMn88QTT9CyZctMR3EBXnnlFWJjYwkODiY0NNSzgNSjjz5KSEiIZ5GkVH379iUkJITQ0FA6derE3/72N6pWrXrZ8aeaNm0an376KaGhoTRu3Jhvvsl5kH/8+PFMnjyZkJAQ/v3vf/PBBx8A8MEHHzBhwgQiIiKIi4vLd2wiIiIiIoDz/mnMWqiUvwGa4sbkZlXXwhYeHm7Xrl2brmzr1q00bNjQSxGJFB59t0VEREQkRxtehc1jocVUuHGIt6PJN2PMOmtteE718jvFV0RERERERAragW/Avzrc0M/bkVxRSlBFRERERESKkgtxcHIT3PwolAjwdjRXlBJUERERERERb4g/7Lxreqk9nwMWqvW44iF5W5FOUIvC+7EiBUnfaREREZFr3PL7YPUwOPQ9zKoGv3+asc6Bb6FcQ6jc/MrH52X5+h3UwlS6dGliYmKoVKmSfmdSrgrWWmJiYihdurS3QxERERERb9nzubPd6fzqBrEb059POgfHlkOdwVc2riKiyCaoNWvWZP/+/Rw7dszboYgUmNKlS1OzZk1vhyEiIiIiV9LRpXB6J9x4f8ZzyfHpj7ePh6TTUKv/FQmtqCmyCWqJEiWoW7eut8MQERERERHJnxVD4exu8K+Wvty3NMQfSF92+jdn9d6qna5cfEVIkX4HVUREREREpNg7u9vZLu7pbEtVdrbXdYD4g+nrxh/KmMheQ5SgioiIiIiIFJYLcRnLmo2HgRcgoC6cu2QENf4QlFaCKiIiIiIiIgXt1NaMZSUrgE8JZyrvhROQnOCUn/kdTm4A/6pXNsYiRAmqiIiIiIhIQUlJBJty8Thui7NNndYLTnIKToIKzqhp4hn4eaBzHBRS+HEWUUpQRURERERECoJNgZnXwdI7L5ad3AQ+paDbCrjxAaestHuE1L+Gsz13AH6fAifWQOQ/od6TVzTsokQJqoiIiIiISEE49jMknoT9sy+WHfkBKreAwJuh+adw+x4IauycK1vb3W4pxKyCUlXgpkfAmCseelFRZH9mRkREREREpFiJWe1sS1WCn24HHz84uRFcf3PKjbmYlAKUqw9V2sCGkc5xpebXdHIKSlBFREREREQKxolfnK1PSTgw52J59V6Z1zcGAus5I68Ajf9cuPEVA0pQRURERERE8iP5AuyYALHuBDX+sLP1C4Q6g6F8o6zbpi6eVKIc1OhduHEWA0pQRURERERELpe1sOR2ODQ/baGzCXsPbn44+/alKjnb0lWv+em9oEWSRERERERELt/545ckp2mkJp/Z8SSo1xVcTMWYElQREREREZHLdWpr+mOTZpJq2t8+zYp1j7aWqVVwMRVjmuIrIiIiIiKSF0cWg02ChGOwfHD6c2VugLO7nf3cjKAGhTjbuvcWaIjFlRJUERERERGRvPiho7OtecfFslKV4HwMVOsGx1dC2VrOCr05qRwJd5+GEgGFE2sxowRVREREREQkty6cvLif+pumNz0M+2Y6+0FNIHJi3vpUcuqhd1BFRERERERy6/APF/cvxDqLG0VMBJvslJUo7524rhJKUEVERERERHIj+Tz8Nv7icfxB8K8JPr6QeMopK6vFjvJDCaqIiIiIiEhOks/DvKZwdAmUKOeUndqecaXegJuufGxXESWoIiIiIiIiOTm+3PlJmZsegrBxTtm5fRkTVP9qVz62q4gSVBERERERubYc+Qm+qgj75+S+zaHvwfhC2HsQFHyxPNA9YtpqOtR/GoxSrPzQ0xMRERERkWvH7/+ClUOdBY5OrM99u8MLoHILZ3pvxfCL5eUaOts6A6HZuIKN9RqkBFVERERERK4dvzwHZ/9w9lPO567NoQVwYh3UuN05NgYavuDsl29c8DFew5SgioiIiIjItcGmOCOnjV+BEkGQdC537Y4udbb1/3SxzPU2dF8NFUIKPs5rmBJUERERERG5NiSecpLUkhXArwwk5zJBjT/gLH7kW/pimfGBShGFE+c1TAmqiIiIiIhcGy7EOttSFcHXP/cjqOcOgH+NwotLPJSgioiIiIjIteHCCWdbsmLeR1DLKEG9EpSgioiIiIjIteH0LmdbsgL4lsndCGpKIpzZDWXrFm5sAihBFRERERGRa8HOSbBsgLPveQc1Pud2Jzc6I62VWxRufAIoQRURERERkavdyU2w+jFn3/hCufrOCGpupvjGRjvbtL99KoVGCaqIiIiIiFx9Dv8AS+92Vu09sdYpa/k59DsCPiWcEdRLp/geXghn96YvO3fQ2ZapWfgxC37eDkBERERERKTA/djF2Saegrit4FMSag8AH3cKlHYE9dR2wMCPXaFUZbjz2MV+Eg5BqUrgW+qKhn+tUoIqIiIiIiJXrwuxcGAOlGtwMTkFZwQ18bSzCNJ3DS6Wnz/u/KxM6qq98QehdLUrG/M1TFN8RURERETk6pJ4+uL+2qfg1Da44a70dUpVcX52ZkbJjO1np5nOe+4g+CtBvVKUoIqIiIiIyNUl9Z1TgNh1zrbhc+nr+FfPvo/kC3Bqh9NXpeYFG59kKV8JqjHmWWPMZmPMJmPMdGNMaWNMXWPMKmPMDmNMlDEmkz9JiIiIiIiIFJLUlXcB4g9BzdvBr2z6OqlTeNMKHnVxPzEOji0BLNQdUihhSkaXnaAaY2oATwHh1tpgwBcYCLwNvG+tvQWIBR4qiEBFRERERERy5dT29Mf+mSSjJSulP67/DFTvefH4QqzTj09JCLix4GOUTOV3iq8f4G+M8QPKAIeATsBX7vNTgTvyeQ0REREREZHci9sMJs2CSJmNlgbekv64gstZwTfVhZNOghp4M/j4Fk6cksFlJ6jW2gPAu8BenMQ0DlgHnLTWJrmr7Qcy+TaIiIiIiIgUgmPL4djPcH2ni2VBIRnrla4Mgy1UbuUc+5SC0lUunk88CWf/gLIaPb2S8jPFtwJwO1AXqA6UBXpmUtVm0f5RY8xaY8zaY8eOZVZFREREREQkbw7OA+MLTd++WHZd+6zr13/a2ZatBX6BFxPbjaPg5Aat4HuF5WeKbxdgt7X2mLU2EfgaaAUEuaf8AtQEDmbW2Fo7yVobbq0Nr1KlSmZVRERERERE8ubYEggKTb9Kb4nArOvX7g99D0GVVmAMtJzqlMesdLZKUK+o/CSoe4EWxpgyxhgDdAa2AIuA1B8ZGgp8k78QRUREREREcuHsPji6FGrcBr5lct/Ov2qa/UveUPQpUTCxSa7k5x3UVTiLIa0HfnX3NQl4GXjOGLMTqAR8WgBxioiIiIiIZO/oT4CFG/qBr79TVjE8b30YA63+c/HYL6DAwpOc+eVcJWvW2lHAqEuKfwci89OviIiIiIhInp1YD76loXwjZ+Xd7msyrtabG0HB7m0TuGV4wcYo2cpXgioiIiIiIuJ1Kcmw+XU48I3z/qmPO82plMfR01RBTaDPbihb2xlRlStGCaqIiIiIiBRvx5fBr+6JndW6F0yfAXUKph/Jk/wskiQiIiIiIuJ9h3+4uB8U6r04JN80gioiIiIiIsXb6d+cn5Vp8DzUHujtaCQflKCKiIiIiEjxdnYvlGsADZ/zdiSST5riKyIiIiIixVNKEuyfA8eXQ5ma3o5GCoASVBERERERKZ42vwFLbnf2fUt7NxYpEEpQRURERESkeNr/zcX9xq94Lw4pMEpQRURERESkeLoQ42x7bYSyN3g3FikQSlBFRERERKT4SUmGc/udkdOgJt6ORgqIElQRERERESmafp8KCzvCsRVgbfpzCUfAJkOZGt6JTQqFElQRERERESl6YtbAyvvh6GL4vhVsfDX9+XP7na1W772qKEEVEREREZGiISURVj0Cx1dB3GanrMUUZ7v5ddg46mLdo4udrb9GUK8mSlBFRERERKRo2PYe7PoEFrSA2GjAQO1BcFcs1B4Im8bA/m+dur9PhvKNICjEqyFLwVKCKiIiIiIi3nd0CUT/38Xj7R8475f6loSSQdDy31CqMuz90nkf9dwBqNoVfHy9F7MUOCWoIiIiIiJy5Vl7ceGj8zGw/nnw8Utfp0rbi/s+fnB9J9jzb5hREpJOg3+1KxdvkXQO+M7bQRQoJagiIiIiInJlJZ2D6T6wfZxzvLA9nFjrvIOaqsHzEDkpfbuqnZ2tTXK2pa/1BPUJ4DZgs7cDKTB+OVcREREREREpQAlHnO365+DELxcXRCp9/cVzrjfBp0T6dlW7pD/2K1O4cRZ5q9zbfUBjbwZSYJSgioiIiIjIlXXhxMX9Pf92tg2eg1uGQfxB533US5NTgIAbL+7f9DBU61G4cRZpe4Gt7v1vgavjWShBFRERERGRK+t8zMX9JqOdd0lvfNB5zzTwZriuXdZtXW87SWyzcYUeZtE2J83+P4CxQAUvxVJwlKCKiIiIiMiVEbsB9n/jLHAE0GMdVAzLWx+NXir4uIqlDUBF4GMgASjt3XAKiBJUERERERG5Mr5vDUlnLx6Xqem9WIq9LUAToJ+3AylQWsVXREREREQK37kD6ZNTgJIVvRPLVeE4UNXbQRQ4jaCKiIiIiEjh2zfL2XZfDTYZDn2f8XdPJQ9OcDW8c3opfSNERERERKRwndwE65+BoFCoGA7GQOUW3o6qGLNALM47qFcXTfEVEREREZHCtX+2M2raerqTnEo+nQaSUYIqIiIiIiKSV0d/ggouKN/Q25FcJVJ/R1YJqoiIiIiISO5ZC7G/OFN7pYCkJqhX3zuoSlBFRERERKTwxB+A8zFQoam3I7mKzHdvr75VfJWgioiIiIhI4YmNdrYVXN6N46rxM/Bn936wNwMpFEpQRURERESkcFgL655x9oOaeDeWq8asNPsBXouisChBFRERERGRwnFyA5zZBaUqQ4lAb0dzldjs3i7xahSFRQmqiIiIiIgUjn2znW2Pdd6N46qyD+gLtPV2IIVCCaqIiIiIiBS88zGw5Q24vjOUreXtaK4SFtgL3ODtQAqNElQRERERESl4B76DlERwventSK4iccAZlKCKiIiIiIjkxbb3oVx9qNjM25FcReLc24pejaIwKUEVEREREZGClXTWWSCpzr1glHIUnHj3trRXoyhM+raIiIiIiEjBOrPb2Qbc7N04rjoJ7q2/V6MoTEpQRURERESk4CSehv3u1XsDb/JuLFedq38E1c/bAYiIiIiIyFXi8EL4sevF48B63ovlqqQRVBERERERkdw58N3F/VuGQ8ny/5+9O4+zsfz/OP66DbNgxr7vIolszZeiFElFaV+0KZXST2lT6du+LyrVV0JKWqSiBYlShCJjKQlZxr4z9m3G3L8/PufMOWcWZjlnzizv5+Mxj+u+r3Mv1xkej/r4XNfnCt9YiqSin0FVgCoiIiIiInm3PxGWv+U7j20cvrEUWcqgioiIiIiInFjSosDz6OrhGUeR5g1QlUHNwHGcJo7jLPL72es4zn2O41R0HOdHx3FWeNoKwRywiIiIiIgUQN7KvXFNrY1UGBB83im+yqBm4Lructd1W7mu2wo4HTgIfA08CkxzXbcxMM1zLiIiIiIiRdmBNVAyFhreYuela4ZzNEWUpvhm13nAKtd11wKXAh95+j8CLgvSO0REREREpKDanwhlG0DTAdB9KZQ/LdwjKoJUJCm7rgPGeI6rua67GcDTVs3sBsdx+jiOk+A4TsL27duDNAwREREREQmLA54A1XGg3CnhHk0RpQzqCTmOEwn0AL7MyX2u6w53XTfedd34KlWq5HUYIiIiIiISLjPoSv8AACAASURBVKkpsHcZlGkQ7pEUcYeACKBkuAcSMsHIoF4ELHBdd6vnfKvjODUAPO22ILxDREREREQKqgUPgHsMytQP90iKuEMU5em9EJwAtSe+6b0A3wG9PMe9gG+D8A4RERERESlIUg7Bxu8te7r+K+urd114x1RkucBbwJtA0d6+J0+5YcdxSgPnA3f6db8MfOE4zm3AOuDqvLxDREREREQKoMVPwdLXwCkBbiqcOhBiqoV7VEXUp8B9nuO64RxIyOUpQHVd9yBQKV3fTqyqr4iIiIiIFEXHjsLSQXbsplrb5J7wjafI+9nTng2MCOdAQq7orq4VEREREZHQ+Ps5wIVS5aHxndD4boipEe5RFVG7gQ+BtsCvYR5L6ClAFRERERGRrCX9BVGVYMVQqHsVVGgFG76BuFPggrlQKi7cIyzCtuPbtfOMcA4k3yhAFRERERGRzKWmwOSWvvN1X8LFy2D/amjUR8FpyH3taSOBN8I5kHwTjCq+IiIiIiJSFG2fFXjuHoMj2+HYQSirPU9D7w9Pm4jtf1r0KUAVEREREZGMtv0K0zrZccky1u5fBTN62HFs4/CMq1iZA3QDaoZ7IPlGAaqIiIiIiAQ6sA5+Osd3fuUOOGeCHe+cCzUugupdwzO2YmMRsITisvbUSwGqiIiIiIgE2rPEd1yiFEREQ62LfX3NHoMSxWPKafg86mnPD+so8puKJImIiIiISKCDG609b3rmU3krtsnX4RQ/qdj03l4UtwyqAlQREREREfFJOQB/3GHHVdpbBtXrzI9hVwKULB2esRUJQ4DqQA/gZSwIrZvumlnAHuC8/B1aAaAAVUREREREfHYv9h37B6cADW60H8mDfp72FGAZsBd4Ld01I4EKwJX5OK6CQWtQRURERESKq8RPYPk7cHSPr2//GmvrXR+WIRVNE4BXANevb5mnPZbJ9TOATkDxy1QrgyoiIiIiUhztXw2/32TH8++Fy7dATDU4sNr62g0P39iKnFuBnUDLTD7bnO58DLAWeCTUgyqQlEEVERERESmOdi8JPJ99DbgurPsSyjTw7X0qQbDT096Srr8ZsCld3wygLNAnxGMqmBSgioiIiIgUR/tXWRtVydoDa+HIDkhaBCffHb5xFTnb/Y63etpXsEzpaWQMUDcDDYDiuY2PAlQRERERkeJo+ywoGQtXbIeGt0JqMhzdZZ/F1Azv2IqUJZn0nQpcB9TEAlT/tambgRr5MK6CSQGqiIiIiEhxc2AdrB8HDW4Gx4HSteHwFjjsyfBFVgrv+Aq1J4DBfuf/eNpafn0xnrYmcBCr5AuQBMwDqoVygAWaAlQRERERkeJm51xrT7rV2pha4KbCHk8wFVUxPOMq9P4Gngfu9+tbBJQDfgHae/q81Xm9mepNwGF828o0Cu0wCzBV8RURERERKW52LbA9TsudZudRla3dt8LaSAWouTPT08ZhxY7qApOBzkBjLEidCpzpuc4/QJ3u+fxG4IH8GW4BpABVRERERKS4ObjR1plGRNp5ZAVr0wonKUDNnQRPWx4416//WU8bCVzs198EC8m+wAooVQZGA05IR1mQKUAVERERESluDm+FaL91jpHlrd3wrW0vU6pceMZVaB0GhgCeqdMkpfu8Wxb3VccCVu+es60pzsEpKEAVERERESl+Dm+FMnV9594MKkCT/uCoVE32uVi2dK5f3z5P2xe4lOMXPfL/7OWgjqww0t88EREREZHiJqsMKkCVs/J/PIXaJgKDU39nARec4H5vxeQKQNdgDarQUgZVRERERKQ4cVPhyPbAALVknO847pT8H1Ohc9TTHgNGHee6ytl4ljdArZKXARUZClBFRERERIqT1BRo8WxgprREhO+4TL38H1OhcxIWSl0DvJrus5pYVhV8wefxeK+pEZyhFXKa4isiIiIiUpxEREKzx6Bqx8D+Vq/AuZO1/vSEUoANwBp8gSj4MqBn+PXVzsbzDnranP/DwOqk1fSf3J/kY8k5vreg0t8+ERERERGBUx+GmheGexSFwAK/Y2+14w+BVM/xucAi4B+OXxzJ6yIgFngoxyO5YfwNvP3H2yRsSjjxxYWEAlQREREREZFs+8rveDvQGLgFW48KFrS2BJpm83n1gb3AaTkaxb87/2XOhjkAzFg7A9d1c3R/QaUAVUREREREipmjwBfArhzccwy4HHjNr28VvkJIuz1tdqb15t0DUx6gdKnSAAycNpCDyQdPcEfhoABVRERERESKkY3AHcC1wMgc3DcP+AZoATzu6VtOxkq9jfI6wBNyXZfZ62dzU4ubmN5rOkO6DSEyIjLk780PquIrIiIiIiLFyLnASs/xvhzcNwVwgJ+BPzx9+8m4zjT0GdRpidPYfXg3Lau15Jz653BO/XNC/s78ogyqiIiIiIgUEy6QCJzuOc/JtNjJwH+wbWHa+vV7940dim07E9oQK2FTAl0/7kqpEqXoULdDSN8VDgpQRURERESkmDiErSW9Gpuaeyib9/0AzAUu85xXAmp5jht72ruAscEZ5nF8sPADSkWUYt3962hRrUXI35ffFKCKiIiIiEgx4S2KVBEoTfYzqLM97f1+feOA1kD+ZTGTDiUxfP5wbmpxE9XLVs+39+YnrUEVEREREZFiwhugViBnAeoGLGMa7dfXjsA9UUNv24FtHHOP0blB53x9b35SBlVERERERIoJ/wxqDNkPUDfim9IbPnuP7AUgLiouzCMJHQWoIiIiIiJSDEwFOnmOczrFVwFqflGAKiIiIiIiRVAqsNNzvBy4wO+zBliAmp0iSUeAVUDDoI4uNxSgioiIiIiIFEp9sUq9R7EKvAC3AHOAcmQ/gzofC1LDv6VLcQhQVSRJRERERESKoOGedj/wD1AKGIEvBMosQF2DBbVl/foWe9r4kIwyJxSgioiIiIiIFDqu3/EGYBhwKoHhj3+A+hcQCTTFMqWz/K7bBDhAjVAN9oT+3PInr//+OhVjKgIKUEVERERERAqRNX7H/YHdwMPprimLFT9aArT065+d7rrNQFXCFTptO7CNVsNaBfRFRkSGZSz5QWtQRURERESkiJnjd7zQ0z6Q7hpvRrR5Jvef6Xe8iXBlTxOTEqk2qFpY3h0ueQpQHccp7zjOV47jLHMcZ6njOGc6jlPRcZwfHcdZ4WkrBGuwIiIiIiIiJ7bE73gPcCUQle6amse5fw5WBdgFVhOuAHX4fFtHGxsZy9PnPE3pUqV556J3wjKW/JLXDOpbwA+u656C5cWXAo8C01zXbQxM85yLiIiIiIjkk+XpzjPbwzSzANU/dNmLBapLge5BGlf2DZg6gJdnv0zXk7qyd+Benjr3KQ48doB+bfvl+1jyU64DVMdx4oCOwEgA13WPuq67G7gU+Mhz2UfAZXkdpIiIiIiISPbsBH4mMCitncl16QsN3Qt08zvfAyR4jq8M2uiy48dVPzLo90EA9GzeM1/fHW55yaA2BLYDHzqOs9BxnPcdxykDVHNddzOAp60ahHGKiIiIiIhkw+fALgKzoW0zua418H9+56cDVfzOdwPLsD1T828d6Po967np65s4pfIpTLp+Er1a9sq3dxcEeQlQSwJtgKGu67YGDpCD6byO4/RxHCfBcZyE7du352EYIiIiIiIiXjOxjGkPv74zM7muJPA/oIvfeWW/z3djU4WbYNvM5I/Bcwaz/eB2xl41lm6Nu+E4+ffugiAvAeoGYIPrunM9519hAetWx3FqAHjabZnd7LrucNd1413Xja9SpUpml4iIiIiIiOTAAWAycB5Qxq//eNuyDPC0zYCKfv2bsQxq3WAO8ISWbF9Cy2otaVGtRb6+t6DIdYDquu4WYL3jOE08XecB/wDfAd48dC/g2zyNUEREREREJFt+wYob3QSUzuY9XYFkrOZrCWCVp78ntk9q/lXwXbB5AVNWTeGUyqfk2zsLmrzuNnsP8KnjOJFY/eVbsT/VLxzHuQ1YB1ydx3eIiIiIiIhkQwIWjpwBRHv6BmR9eRr/sKh6us9Ct/501a5V1C9fn4gSEWw/sJ2rv7TQ6fyG54fsnQVdngJU13UXAfGZfHReXp4rIiIiIiKSfS4wDJgCnIJvem9qLp5VGhgO9PGcH87z6DKzctdKGr/TmIFnDeTF815k1KJRrE5azcxbZ3JW3bNC8s7CIK/7oIqIiIiIiITZXKAvtm/p6X79DrkrcORfLKljHsaV0U+rf6LzR53pMcaKOL07710Afl7zM40qNirWwSkoQBURERERkUJvqt9xMIoLXYitY/0LyPt02wFTB1D3zbqs37Oeft/345c1v7B0x1IA9hzZQ58Jffhh5Q90bdg1z+8q7BSgioiIiIhIIbcUm5p7PcEpgRMDjAZOC8KzYPDcwazfu54XZr7A8p3LKeGUoFxUORbduQiAEQtGAPDUuU8F5X2FmQJUEREREREp5NZhhZE+BeqFdSQHjh7gUPKhtPMNezeQkpoCwLD5wwB468K32PnwTlpWb8ldp98FQKWYSlQtUzX/B1zAKEAVEREREZFCai/wBPAbUCvMY4Fjqcco+1JZun7im6o7e91sAKJLRqf11Y6rTUSJCABaVm8JQFxUXD6OtOBSgCoiIiIiIoXUvcDznuPo410YUiPmj2Dexnm8MPMFAGatm8XqpNXsO7KPnxN/pkypMizuu5jW1VsDUC6qXNq9NcraPqvdGnfL/4EXQHndB1VERERERCRM/vQ7fjrf3pqYlEhKagqNKzUmMSmRPhP7BHxeM7YmJ719EgClS5WmR5MeNKrYiOm3TOejRR9xTv1z0q7tfnJ3hnYfyi2tbsm38Rdkjuu64R4D8fHxbkJCQriHISIiIiIihUpVrKBR/kzxXbN7DfdPuZ9vln0DwG+9f6P9B+2Pe09kRCSze88mvmZ8yMdXkDmOM9913RP+EjTFV0RERERECqAU4C6gOjAJSJ9YOwxsB+4gP4LT5GPJnP3h2WnBKZAWnDau2JgypcpkuOe5Ts+R2D+x2AenOaEAVURERERECpDfgD3ANGAYsBW4mIxTeDd62to5fsPhlMPc8d0dzN80P9v3TEucxoa9G7i55c0s7ruY5lWbA9CzeU/m3j6XqJJRADSr0iztnra12lIztmaOx1ecKUAVEREREZECYibQAQtGV3n6vBnLZ7G9Sb3We9o6OXrD+KXjiXkhhvcXvs/IhSOzfd+XS74kNjKWYRcPo3nV5rzb7V2ql63Oq+e/SoWYChxMPghA5wad0+6pHZfz4Lm4U5EkEREREREpAFYAl3mOBwP1gSjgEmAzcC3QG2gOtAFGeK49KdtvcF2X6766Lu08KiIqW/dNWD6BDxZ9wPWnXZ+2XczZ9c5m84Ob0645nHIYgHva3sP5Dc9n+c7lNK3cNNtjE6MMqoiIiIiIhMFPwD7P8QHgv0CS3+drsCC1BLYO9VsgEl8WdSbQzXNN9qzds5bk1GTuaXsPFWMqsvvI7mzd98PKHwB4vtPzWV7ToloLABpVbMQlTS7hofYP4ThOtscmRgGqiIiIiIjks53A+cDVnvMuwJdkLIR0n99xeWz671RgEDbFt2W237jvyD56jusJwHXNr6NWbC2SDiUFXJN8LJlUNzXDvYu2LuKsumfRoEKDLJ8/vdd01vRfo6A0jxSgioiIiIhIPtvhaadg03XnZHLNVqyKr79OwFJggOe8RrbelpKaQrv32zFnwxzKRpalTY02lI8uz+7DvgxqqptK9AvR9Pu+X8C9CZsS+H3975xZ+8zjvqNCTAXqla+XrfFI1hSgioiIiIhIPtvpd7zQ0/YAZuHLqlbN5L5O6c4zu8bHmw2dsnIKS3csBeDvvn8TXTKaCjEV0gLUvhP7Ej88nlQ3laEJQ3FdXyZ3xpoZuLg8eOaD2flikkcKUEVEREREJJ8kA2sJDFC/Ad4HxmFTeD/DtzY1vfT7iWa9hcvT058m4tkIko8lM3v9bAB2PrwzLctZPro8f279k7sm3sV7899j4ZaFaff+tPqntOM1u9dQLqoc1cpWy9Y3lLxRgCoiIiIiIvnkAqyoUQ/P+SrgUuA2fBuMlATKZnF/KSDGc7wIODvLNz0z4xkAkg4n8d3y72hfpz0VYyqmfV4nzranGTZ/WFqfd1uYrp90TcuiJu5OpH75+tn4bhIMClBFRERERCQf7AN+SddXKRfPWef5ybpA0uZ9vu1fXv/tdZZsX0LvVr0DrulYr2OG+wZfMDjteM+RPSQdSmLOhjk0qtgoF+OU3FCAKiIiIiIi+WCGp30CX2Aal4vnVAbqZPnpsh3LqPmGb+rv0IShNK3clFta3RJwXecGnXm0w6Np599c+w3NqzZPO9+6fytfL/uanYd2ct8Z9yH5QwGqiIiIiIiEmHcqb01sv9P1wAoguFuypKSmcP+U+wFoWKEhAPuO7uOceucQUSIi4NqSJUryUpeX+Ofuf7iy6ZW0r9OemrG+wHbL/i38tv43KsZUpH2d9kEdp2RNAaqIiIiIiITYeCAV+BCIwtaRBn/a7Kd/fcoPK3/glS6vMP6a8Wn9Z9bJeouYplWa8tU1X1GlTBVio2K5rfVtAPz35/8ycuFI4mvGU8JR2JRf9JsWEREREZEQ+xk4Fegakqdv2reJLqO7cMu3t9C0clMGtB9A+ejyaZ93PSn773296+tEl4xOq/zbpFKToI9XsqYAVUREREREQsgF5gPtQvaGF2e+yLTEaQDc/Z+7cRyHctHl0j6vXrZ6tp9VLroc/dv1TzuPLhkdvIHKCSlAFRERERGRENoEbAdah+wNs9bNomO9jvx000/0je8LQGxkLABXNL0ix8/zz7h2qNMhOIOUbCl54ktERERERERya6GnbROSpy/bsYzF2xbzRMcnOK/heWn9ESUi2PrQ1oC9T7Orc4POTOw5kaZVmqYVW5L8oQBVRERERERCJBW4wXOc9b6luZWYlEjTIU2JKRmTljn1V7VM1Vw/u/vJ3fMyNMklTfEVEREREZEQmQfsBaoBZYP+9JdnvQzAPW3voVrZakF/vuQ/BagiIiIiIhIiIz3t4pA8ffLKyXQ9qSsvnvdiSJ4v+U8BqoiIiIiIhEAiMAK4EaiS7buOpBzBdV0AXNdl7e61mV534OgB1u9dT8e6HYkoEZH34UqBoABVRERERERCYIynfSbbd7w7712iX4jmw0Uf4rou/b7vR/236nPBJxfw69pfA6594/c3AGhcqXGwBiwFgAJUEREREREJMhd4E7gAyF4V3DGLx/B/3/8fAPM3zWf6mum8m/AuAFNXTeWcUecw9u+xadc/Of1JAE6udHIwBy5hpiq+IiIiIiISZNuBHcBFWV4xLGEYlzS5hJqxNRnyxxD6Te4HQN1ydXk34V1GLBhBCacEux/ZzeSVk7ntu9u4btx11nd4NwDNqjSjZbXgVweW8FEGVUREREREgmyVpz0p00837t3IXZPuotYbtVi4eWFacNq7VW827N0AQHJqMmfWPpPYqFiuaXYNSY8k0axKM26fcDt9JvYBYEi3ITiOE/JvI/lHAaqIiIiIiATJduBm4A7PecYANSU1hamrpqadtxneBoBx14xjSPchnFX3LAAm9JzA5Bsmp11XskRJHu7wMHuP7AVg9GWj6VivY0i+hYSP462QFU7x8fFuQkJCuIchIiIiIiJ58hjwkt95Mv6rCmevm83tE25n2Y5lAFQtU5V72t7DufXPTQtMkw4lsXn/Zk6tcmqGpx89dpST3j6JlNQUNj+4OYTfQ4LNcZz5ruvGn+g6rUEVEREREZEg2ENgcPoc/uFGSmoKV3xxBXsO70nrW33vaspElgl4SoWYClSIqZDpGyIjIvn0ik/T1qBK0aMAVUREREREguD3dOfVAs6mrprKtgPbGHfNOGrG1mTKyikZgtPs0LTeok0BqoiIiIiIBEGip60BbMY/QD2ccpiB0wZSJ64O3Rt3J6pkFGfUPiMcg5QCTgGqiIiIiIgEQSIQBTTHAlRfqPHO3Hf4a+tffHfdd0SVjArT+KQwUBVfEREREREJgtVAPWAYcA1wLgB7Du/hxVkv0q1xNy5pckn4hieFgjKoIiIiIiISBIlAQ6ABMDat9+O/Pmb34d08c+4z4RqYFCJ5ClAdx1kD7AOOASmu68Y7jlMR+xtZH1gDXOO6blLehikiIiIiIgVbItAuoGffkX28OPNFzqh9BvE1T7jDiEhQpvh2cl23ld+eNo8C01zXbQxM85yLiIiIiEiR9TSQhGVPjeu69Jvcj837NzP4gsHhGpgUMqFYg3op8JHn+CPgshC8Q0REREREwmoT8CRwGBgJwPTEOAZMHcCBowcY8/cYRv85moFnDaRd7XbHe5BImryuQXWBqY7juMAw13WHA9Vc190M4LruZsdxquZ1kCIiIiIiUtA8DHwKPAfApn030Gn0XQCUjSzL0zOeplqZajze8fHwDVEKnbwGqB1c193kCUJ/dBxnWXZvdBynD9AHoG7dunkchoiIiIiI5J9ELDj1+WlVk7Tjp2c8DcDoy0dTulTpfByXFHZ5muLruu4mT7sN+BpoC2x1HKcGgKfdlsW9w13XjXddN75KlSp5GYaIiIiIiITMQmAD8B2wxdN3v6ftCbwHbGXG2jVUiqlEwwoN0+48p945+TlQKQJyHaA6jlPGcZxY7zHQFfgb+5vby3NZL+DbvA5SRERERETC4SjQBqiDlZq5wdO/ELgJ+Ay4E9etwtTVU+lYryP3tr0XgL/7/k1UyagwjFkKs7xM8a0GfO04jvc5n7mu+4PjOPOALxzHuQ1YB1yd92GKiIiIiEj+G5nuPBELWtfjrdh74OgBRi4cyYa9G3ih8wvc1OImrmh6BXXK1cnnsUpRkOsA1XXd1UDLTPp3AuflZVAiIiIiIhJOLlYu5v10/YnAWMDl720HOW2ok/ZJ21ptubHFjTiOo+BUci0U28yIiIiIiEih9ieBwelW4BnP8c2kpEZw7VdvB9zxSIdHKOEovJC8yWsVXxERERERKXIS044Op9Rm874DfLgomWc7Wd/9UyL5Z/shxl0zjhbVWlA7rjbRJaPDNFYpShSgioiIiIhIOpMAePP3u3nil3c5kGyVeb0B6tdLD/HJ5Z9wRdMrwjVAKaKUgxcRERERET9r8RZHemDquxxI9n0ya521tePaKTiVkFAGVURERESkWHKBV7CqvPcDsZ7+fzNcefi/h3lzzpt0/XggpUvB51c9T0ypmPwbqhQbyqCKiIiIiBRL/wADgaeAGsBhAA4c/RuASz+H5zo9x8geI4kqGcWjZz3KoRTYecgq9oqEgjKoIiIiIiLF0nK/4wPAzcAXbNn/PbXjYMJy+Pa6xwPueLXLq8zZOIe4qLj8HKgUIwpQRURERESKpZWethawEfgN1z1A3XLT+GYZ/HvPigx3DOgwID8HKMWQpviKiIiIiBQSBw/Crl0Z+3/8Efbt850nJsJLL8GxY1k9KRUYx5GUCvQY04av/okm6dA+dh5aRqkIl1IRF9OoYqMQfAOR41OAKiIiIiJSSHToAJUqwXPPwaWXwu+/w7Bh0LUr9O/vu+7JJ+Gxx+DTT7N60mLgD574JYkJ/05gxa7DlI3cy86DNu23WplTQ/1VRDKlKb4iIiIiIgXQnj3Wlitn7eHDsGiRHT/5pLXffee7fulSy5g+/TR88on1vf8+3HQTOE76p88FYNxSO6sQ3YxSEUs4mLwAgLio+sH8KiLZpgyqiIiIiEgBs38/lC8PF11k5x9/DF9+efx75syB6Gh4/nlf38yZ0KoVLFkCmzZBv342/TfV/Zu9R2B1Eux+ZDeNKrYHYM+RPwCoGNM4FF9L5ISUQRURERERCZNZs+DMMyEiIrB/1Chrf//dgtWbb8547/r1sGABNG8OCQlw7bWQkmKfXXABjBgBkyfDnXfCmDH2nCFDYMcOeGfkEnYdgvLR5SkXXY64qLoAHE5ZAkCl0gpQJTwUoIqIiIiI5LOEBBg0CMaOhTffhKuugtGjLdv53XcwY4bv2h9+8B07Drz2Ghw6BLVr2w9Aw4awbBmMHGnPPOMM6+/Tx9ao/vabBaYAa9fCvqMr2HbAYe19awGIi6oHQHxNq8AUGVE9pN9fJCuO67rhHgPx8fFuQkJCuIchIiIiIhJSrmsFjX76ydcXF2eZz4MHA6/t2xeGDoXu3WHSJJu627EjnH12zt75wAPw7rtQooQFtgBrkkqzYlcMXRruBGDr/llUK3u257gV1couzO1XFMmU4zjzXdeNP9F1WoMqIiIiIpJPNm3yBactWsDVV1vxo6pVra9KFejRA3r2tMASLDgFuzanwSnAJZfAkSO+4BTnGHFRhyhVolbaNRVjGqYdR5S4PecvEQkSTfEVEREREckHa9bYfqUAs2dD+/aBn+/aZdnUkp7/Q/ff1xSgfv3cvdc73RcsyP128m4qxEDtuNPT+ktFVEk7rlz6rNy9SCQIFKCKiIiIiITYwoXQpo0dly0LrVtnvKZixcDzsmV9x2+8AZGRuXt3TIzv+LrrIDm5Ip06beSXX6L8rirld9wkdy8SCQJN8RURERERCaHt26FbN995jx6BQWNW/Pcuvf/+vI1hxAiIjYVmzaBBA4c//qiJ61ZKd9XFwKNAdN5eJpIHClBFREREREJo0iTYssWCzObN4cUXs39vtWpwzjl5H8Ptt8OePdCkiQWpJ59s284EmgC8lPeXieSBqviKiIiIiITAkiUwYIDtRVqpEmzbZpV0c8p1A7OpIoVRdqv4ag2qiIiIiEgIXH89/PWXHT/5ZO6CU1BwKsWLAlQRERERkRDwTqF9+224557wjkWksNAaVBERERGRIBs+HFavhgsvhH79wj0akcJDAaqIiIiISBAtWgR33mnHd9+tKboiOaEAVUREREQkSFJSfBnTDz+ESy4J73hEChsFqCIiIiIiQfLllzB7NlxzDdxyS7hHI1L4KEAVERERkSKjlquVwQAAIABJREFUVy+bXnvokJ0fOeL7yQ8//QQRETB6dP68T6SoUYAqIiIiIkXCsGEWGA4fDqVLw8qVUK0aREdD+fKwYUNo33/sGHz/PVx+OURFhfZdIkWVAlQRERERyZWUFEhNDfcoTHIy3HVXYN/HH8OePXZ8+DCcdJJdl13jxsGMGbBxI7RpA61aWRDq9ddfsGmT73zuXNiyxab3ikjuKEAVERERkVypWhWuuircozALFlj7zju+vs8+s3b8eGjSBI4ehXnzsve8H3+073buuVC7NixcCH/+6XsPQMuW9uO1fLm1rVvn+muIFHsKUEVEREQkTXIy9O5tW6P07Rv42b//Wgbxjz9sCm1SEnz9tU2lDRfXhcWLYdkyOz/vPJg0yY5XroSaNW3K7ddfW19iYvaeO3y4tY0bW/t//2ftL79Yu26dtTt2+O5ZvRpKlIB69XL3XUREAaqIiIiIeIwYAc8/b9ujALz3nm8Kb2KiZSEffRSGDvUVIfJel97OnTB9emjH67q2z2iLFr6KuQ0aQLdu8NZbdu4NLL1B49q1J37uX3/BV1/B449bMP7ZZ/a8unVtnevKlYFB6Jo10Lat/e7q1oVSpYL1DUWKHwWoIiIiIsXcgQMwZgz06QPPPmuFhR56yD7zTludONHaQYNg1KjA+//9F3btgu7dYf166+vZEzp1CswwBtvbb2cMjqOjrb33Xnv3ww/beenSULEi/Pe/cPDg8Z/7zTeWQe7Xz4or9explXmrVrUsqTerCva7eukl39ThG24IzncTKa4UoIqIiIgUc8OGwfXX+85LlvRlHt96yzKVw4bZea1aliX0N2GCBbfffw+vvWZ9f/5p7YwZoRmz61pg2KWLBchPPQXTpgVeU6mSfRevK66w1pshzsrChXDKKRZ8+nv1VWje3Hfepg3ExcH8+XbeoAE891zuvo+IGAWoIiIiInm0enVgddfcGDQIZs8OznhyKn1GcehQqF8f4uMtMP3xR1iyBN58E1atsqzqmWfatR07WjtunLWlSsELL8C2bXae3aJEObV5M2zdCpdeahnNp5+Gzp2Pf8/w4XDGGfY9jld9eMOGjEE4WEZ48WIYORL+9z/4z39sKvPixTBggP1uHCdPX0uk2FOAKiIiIpJLa9da5u6kkyx4ya1//rEAxxvs5bekJN/xvn1wySV27K2C682eNm1q+3tGR9uU4Hvv9U399UpMtLWbXkuXhmbM3sC3Vavs3+M4VgBq1Sr7RwWwrXI+/dSmKI8fb5nZFSuscm9Weve2DHNcnN139Kitz1VwKpJ3ClBFREREspCcDH//nbF/7FibOlq/vq3ZBJg50wKUzAoGncjkydamptp60Py2YwfUqQO7d0PZsr7+xo2tf/x4O2/UyPdZvXo2/Tc2NjCY81bLnTnTtmkJRYC6bZsFxzVqwOmn5+xeb0Dr/XMdPRpuvNGmA195pU0D3rPH1queSFyc77h+/ZyNQ0QypwBVREREipy337bqq3l17bVw2mnw22+BU3ivuy7rKb3pt2bJDu8WKZD9bVCCaccOqFIFypXL+FmdOr7jrLZP2bDB2pNPtjYuDtq1s3Wcq1bBkSPBHe/kybbNyzvvQExMzu499VRrL7/cfu/esXt98421F1104mf5/74UoIoEhwJUERERKTIGDYLq1aF/fwuQ8mL6dF82sEMHaNjQgl5vX6VKsH+/7/rffrNtRgCmToV77oFFi2ytYokSFuj5B6L+li3zBVqrVmX8fN26wG1dgm3HDqhcOfPPvGs13303sOCQv6uvtvayy6w95xxbi9q0qd2/YkVwxrl8ue1DunKlVdXt0SPnzyhTxnd86aU2RTu9Ro2yN2Xbv4iSfyAvIrmnAFVERESKhHvvtXWcW7f6+ryVZHPj++8tyPJO9Vy3zoJebyXY11+3YMcbyJx5JvTqZVNkL7jAAtM337RA1XUtU3f//ZkX51m2zPbuBPj9d6uK26CBVcA9dswyl1dfbf3vvJP775SVbdssg5oZb2DctGnW93/yiU2L9U4Pbtgw8J5gFEoaNMgysp072z8EVK6c+/1GvX9m//5r7X33WcGj++6zKcOvvpq959Sq5TuOjMzdWEQkkOO6brjHQHx8vJuQkBDuYYiIiEghtXq1TS/1Tru98UZbN1mpkhUy8havcV3bp7NOneMXtDl61KaC1qplayi3b894zW+/WVB66JAFZ9WrW/9ll8G332a8vlo1C55nzbKMrJd3eu3rr8MPP1hQWrasFd8B278zfZXdtWszrzKbG65rRY/69888MOvf36ZMZ+edX39tAfykSRZwHz0KzZpZ5jUva1GTk+33kJLi67v0Ut903Jzatcv2Q33vPQt0t261LHdOJSb6gvEC8L/UIgWa4zjzXdeNP9F1yqCKiIhIofb559CypU2R3bDBAsaPP7b9KNevhyFD7LqDB+Hiiy0bec89x3/mtGk21fa++zJfU9q4sW8KcUyMLzgF2xszvQ8/hL/+sqA4/V6d/hVyBw+2oG7XLt9+m+mDU/BV182refNsvejRo1CzZubXvPaajT07AfHll9t0Xm82ODLSssrLlll14NzatcuC0//9z9fnLdyUGxUr2pTlzz6zf2jITXAKWf/ORCT38hygOo4T4TjOQsdxJnrOGziOM9dxnBWO44x1HEcTHkREROS4tm3L3RpL17UqunXr2trEWrUsGwi23yX4gtH337dpu2CZs8yKHCUm2rYp3sq8XbvaGkX/bVi++MLWlmYV1HjXY/bpY4Hs+PFwyy1QtaptR7Nkie/ajRstkK5QwaaYegv4AIwalfHZsbHQurVlKPNq716rWutdR1ujRubXRUZaoajs8q/0C/aPB5B5NeTs2rHD2sqV4eef4auvch9UejkO9Oxpf0a5FRVllYqD9Q8GIhKcDGp/wH/SxivAm67rNgaSgNuC8A4REREppO64w36yMnq0TX/1BnY5sWmTTR294w6ITzdxzH9/zCNHrMAOwDXXWHC6c6fv84MHoXt3m655ySUwZ45lx8qUsUCofHkLIsGuK1066zE1bWoB1Tvv2BrHyy/3fVarlgWl+/dbcD17to1t2jRfkSJv9dg2bSy47dXLgtjZs20q8amnZqw8mxu//24ZZm/ho9jYvD8zM97KvpkVfwLLjE6blvUU2eRkmxYN9jvq1MkC64Liyy8t0BWR4MhTgOo4Tm2gO/C+59wBOgNfeS75CLgsL+8QERGRwmvNGstcvv++BaKLFgV+Pn++BWBgWcEtW3L2fG820j8Y9SpdGkaOtOOlS2HoUMuqXnWV9T3zDDz0kK1fvf56X3bV6+OPA88TEix7d7zg1KtSpcyL5tSqZYFmbKxdM3u29Tdp4rvmm28sEHUcC25HjYLHH4f27a2venVbM5nbNY8pKVbE6cIL7XzpUnjiCejSJXfPOxFv4O3/DwL+7r7b3v377xk/W7IEHnsM7rrLzitVCs0YRaTgyGsGdTDwMOCtR1cJ2O26rncJ+wagVmY3Oo7Tx3GcBMdxErZnVnlARERECr2JE33HvXpZgOgfhE6fbu3nn1ubWXGh4/nrL2v9p8b6q13b2nfesYCuc2ebagu2BvH11y3j+e23cOedtkbV6/TTA5/VsGH2th45Hv+qr0lJVnyoTp3AoDcy0taFZqVaNZsOnds1nZ9+atvggO3j2aiRTWkOVRXa8uUtsPZO003vK09aI7MMa/PmVr3XK6utcESk6Mh1gOo4zsXANtd15/t3Z3Jppv++57rucNd1413Xja+SVV1zERERKbQ2bLAKsGDrJi+4wKazeoNSsKxprVo27bZWLfj11+w/33WtAFKrVr6gM7369a0dPdoq4z73XMZrjx61rOTgwbaNi1e5ctkfS3adcoq13v1SIedrIL17b/pvp5NdKSn2O2jd2jKa3gA/lCIirCjR2rUWbKZf++vdl3TNGl+f62a+zlYBqkjRl5cMagegh+M4a4DPsam9g4HyjuN4t3GuDWzK0whFRESkUHnuOcuY1alje35OmQILFsB339m+lT17WlC4cKEVNjr3XLv+P/+xiqpHjtiU1s2b7XnHjtkU2LZt4fBh33u2bLGgpnfvrMfSuLEVJkpJsWI9JUoEVtwF65s61Vdcac4cX0Y32G6+2TKFjz3mWzPrXaOZXScKUF0XBg4MDLa9HnzQ3v/00xY0BmurmhOpVMmmTA8YYH8f/B04YO2gQfa7B5sK7p/N/r//s3/w8P4ZiUjRlesA1XXdga7r1nZdtz5wHfCz67o3AL8AntUd9AJyOFlHRERECqutW+HJJ33nzz5rlXDBppDefrsdr15tGTWABx6wtmtXCzjvvhteeMH2GAVYt86KDc2bB2PHWt8ff/i2+PBmJTPjOPDiixbYPPyw9VWoYEV5Fi+24Hfv3sAqte3awbXX5vpXcFwlS9pUYcexvU8h51uVeNdh+lcW9jdjBrz8MvTokfGzUaOgY0crBJWf/NeOdu9uRaIWLLDiVLt3W//evfDWW3a8enXg/RdfHDg9WkSKrlDsg/oI8IDjOCuxNakjQ/AOERERKYC822089BBccYVl8vzdcou17dqBtwSFN1C76y4LHr/80s7XrrVs4MqVvvu9Qe1Iv/+7OF6ACjZ9ePfuwICtc2db31iihG+KaX576il7//nn5+y+8uWt3b3b9hz1ruH08p9CfZvfXgp79lgQeMklFiDnp/SZ2tmzbY1vmTL2Zzx4sGXId+2yz73TfXfssOJU3oJOIlL0lTzxJSfmuu50YLrneDXQNhjPFRERkcLj118t83nWWfDaa5lf461Wu3evBVfgC1Adx6bh+gdYEycGVnfdscN+vv3WsnJvvGFTiU8kKirHXyfk2rWzqcc5DRb9A1Tv9OA1a3zB3DPP+K794AMrBFW+vGWiIf+m9fpr186X/YaMGdJKlezvwSbPwrC1a63SccWKqtwrUtyEIoMqIiIixcjUqTYFs3Nny4a9807W15Yr59vDctYsK1zkv66wWTPfcWws9O1rQegZZ9ha0h07YNw4m0o8dqyt6SzMcpPJ9BZv8q7RBVvLGx/v247FPyvrrY7rzT7Xq5fzd+bVHXfYVjZe//wT+HnVqhagerPqa9dagav8zvSKSPgpQBUREZFcmzvXqvNOmgRXX21rRTPbk9TfSy9Z+/vvvuypV/Pm1pYuDf/7H2zcaEWTnnjCrh0zxoKwypUtIC6OSpWyqbFLl/r6vPvBgk2HnTDBgnqwwPXzz2H5cjtv1Cj/xupVtqytR/ZKH6A2b+4LUF3XMsLhCKRFJPwUoIqIiBRzrmvrMwcMsOOc+Phja1etsuAxO9MxTzrJNy23Ro3Az7zrSStX9hVJAjj77MBgtl274p1dK18+MCj13yv09NNtSvO0abbFC9ha4L//tgrA4Zwym5Bg7c8/B/bXqGF/vkeO2PTvtWsVoIoUVwpQRUREipG5cy3b+dZbcNFFFsw8+aRl3AYNsqI9X3+d/ectW2bFbRo2zP49JUpYASXIuB7yP/+xNaxjx1og6xUbG1jFNaeVb4ua8uV9a3jT8xamKl3aig49/LBlJEeNglNPza8RZu700wPPv/4ahg2zf2xo0MD6Fiywgk7ePWxFpHgJSpEkERERKbjmzIHhw+H77zPfO3PBAlsDuG2bnQ8ZYlMyO3WybVGOZ9my3E219a41TU0N7C9TBmbO9J2/954v+H31VdsHddQoq8xbnHkLJaXXq1dgVeK4uMDAvl+/0I4rO0aMsDWpAJdd5uv3Bs/jx1sbjqnIIhJ+ClBFREQKOdfNerrrTz/5CuacdZYFeHv2ZLzujTcgJsYKGE2bZj/jxkGXLlad9+KLM97TvLmtEW2bi9r93qDzRHtb3nmn7zg2Fj78EIYODSysVBxlFaBmlnX0zzZfdFFIhpMjt91mGd30+7R6A9L//c9a/71pRaT4UIAqIiJSSN1xh637PHoUGjeG774LnBbruvB//2fHCQm+6ZXbt1ugWqqUBXrvvWcFjiIjA5/vrbYLti2IdwomWDXdJUvsfv8gMrs6d4ZPPw3MoGVXcQ9OwRegRkTAsWN2PG6c7XGanv8/AsTEhH5sJ+I48PzzGfsjI6F3b9saBwL/volI8aE1qCIiIoXMlCm21+X779tUXNe1qqhTpgRe9+efVlV3xIjAtX9VqliRourVLdB59NGMwam3uI5XYmLguXev0m+/tUA3pxwHrr/e1klKznkD1NhY6NnTjq+4IvM/C++ffWxs/owtL4YOhRkzbFp6Cf1fqkixpAyqiIhIIXL4sG0j4vXLL1b5tnTpwCDywAG4/XY7zsm0zokT4Y8/YPZsm+br9e+/gWtNn3jCsnFt2uTue0jeeAPUMmVg9GhbY5yVyEhYt86XaS3IIiOhY8dwj0JEwkkBqoiISCGwb59lS/3/5/2qq6BpUzuuV88ClSuusO1ZXnwR5s+H7t1PvM7TX/fu9nPwIKxcCS1bWv+qVb5r1q2z4kgvvGDbwUj+q1DBd1yypP3dOB7vtj4iIgWdJk+IiIgUcMOHWzXWSy6BefN8/Xff7Ttu2NCq8LZvD5s3W4B69tmWEc2N0qWhRQvbAgTsmWDTiV96yY67d8/dsyXvvBnUwpAVFRHJCQWoIiIiBdiPP/qKEE2aZO0//9hxp06+6/wLGl1/vbXpq6TmRp8+lpHdssXOp02zokqgKqvhVKWKtXv3hnccIiLBpgBVRESkABs92tbl3XWXnQ8ebNN6u3ULvK53b/joIzuePh1uvRUefDA4Y6he3ZdB9a5z7dJFRWzCybv29+DB8I5DRCTYtAZVREQkzFwXUlKsAuuSJbB0KTRpAlu3wief2FrToUNhyJCsg0LHgZtvhqQkuO8+q8yb1d6oOVWjhlVWBQtQS5aEH34IzrMld2rXtrZ58/COQ0Qk2BSgioiIhNH+/XDeeTZtd9Qo6N8fNm70fV6ihG/PyOxkLPv3t+xpXFzwxli7NuzaBbt3W4Bat27GbWgk/61ff+LiSCIihY0m54iIiITR88/bti7791umdONGaNfO9/mkSZZNzYlgBqcAbdtae9NN8PPPgXuqSvjUru0rliQiUlQogyoiIpLPXBdGjrR1nYMHw7XX2r6lEyfCiBG2f+nhw5alLFUq3KP1BajeisBXXx2+sYiISNGmAFVERCQfJSdDzZqwY4ev75574OSTLZPq3bolOjo848tMbKytjW3WzM7PPju84xERkaJLAaqIiEg++vVXX3D6ww9WcKhDBzsvyPuKnnoqnHIKLFtmVX1FRERCQQGqiIhIPvruO8uObtgAlSqFezQ5k5CgbU1ERCS0FKCKiIjkg9RU2590zBjbQ7SwBacAZcrYj4iISKioiq+IiEg2TZ0KN94YuA3MiQwdams3b7jBtpPZvh0eeih0YxQRESnMlEEVERHJQmqq7TUJUK0aPP44zJsHMTFWbTcr69fD++/btjFDh9oep//8Y59deCGcc07oxy4iIlIYKYMqIiICJCbCgAEwf76dJyVB1apQv7799OsHf/5pn82ebVuvvPSSnbsu/PijVegdMgTq1oVnn7UAdfFi3ztefRXGjcvPbyUiIlK4OK7rhnsMxMfHuwkJCeEehoiIFEM7dljwOXasnUdG2jYqrVrB669nvL5pU1i61Hd+8CA89pjtZ+pVoYIFuF6TJsGWLXDzzVa1V0REpLhxHGe+67rxJ7pO/5kUEZFiJTUVPvsMFiyAqCj46SerTguW9RwzBqZNsx+w83ffhZkzwXHgzjvhvvt8z3v0UXj77cB3TJwI27bBoEHw5JPQtWv+fDcREZHCThlUEREpMvr2heHD4Ywz4OWXLROa3nPPWdCYmb//tv0+q1WzYkY9esC338Lq1XDBBRaMNm7sW0PqOFC7tq05vfJK3/TdY8eghBbRiIiIpFEGVUREiqytW2HGDBg1Cr75xqblJifDe+/Z57/9Bt262fTdqCjffQcPwrBh0LGjBZzLlkHr1nDuufa8U0+1oLNpUwtQ27Sx+xo2hBUr7Dglxfc817Xg9IMP4NZbrYhSXJyCUxERkdxSgCoiIoXKvHk2ZXb3bjt/4w1Yt84CTrAqubVrWxXdNWugSRPfvcOH2xYxn3xiQam/1q19x/feC7/+Cp06ZXx/yZJWLGnGDLsObOsZgOefD8Y3FBERKb40xVdERAqVtm0t8GzTBqZMyfj5v/9a5rR9e1sL2r2777PLL7dpvN5s6PFs22ZVfLOSlAQVK0L16rB5c46/hoiISLGS3Sm+moQkIiKFxqFDsHAh3H47/PCDL4Ppr2FDaNTIjv/919f/998WsHbokL13HS84BavUO3YszJqVveeJiIjIiSlAFRGRAst1Ye9e3/kff9ga0LZt7fz11+Grr3zTfQEiIqByZQtUf/4Zliyx4PSzz+zep54K3viuuQZOOil4zxMRESnuFKCKiEiB9cgjNo123DirjPviixATA+edZ5+XLGnVc8uVs7Wpc+dav+NYBd6JE6F5c+jc2Qon/ec/0KBB+L6PiIiIHJ+KJImISIGyYAGcdZZN5/V6+21YvhymTrUtYmJjM94Xn25VS9++MHiwHW/fbsHrXXeFbtwiIiKSdwpQRUQkrDZvtq1ZypSx80ceseDUcWw/0vPOg08/taq6MTG2lUt2nHxy4Pnhw1Y4SURERAouTfEVEZGwmTABataEsmXhv/+FLl3gp58sCD12DDZtguee813/zTdQqlT2n790KTz4oB1HR9v2NCIiIlJwKYMqIiL5bvhwC0S3bPH1vfiitSefbNV5HcfOGzSA996zAknnn5+z95xyiu1NWrEi3H23rVUVERGRgkv7oIqISL749Vd47DG4+GIYONDXP2QIXHABtGsHjz4K999vlXhFRESk6MjuPqjKoIqISMglJ9tU24QEmD3b13/WWZbZBMumltR/lURERIo1/a+AiIiEzOrVlh1dudLOX33V1oK2agUffQRPPOG7VsGpiIiI6H8HREQkZF55xRecXnEFDBjg++zss8MzJhERESm4VMVXRERC5o8/rDLv5MkwZky4RyMiIiIFXa4DVMdxoh3H+cNxnD8dx1niOM4znv4GjuPMdRxnheM4Yx3HiQzecEVEpCA4cAD+/ff412zeDIsW2XTeCy+ESP3XQERERE4gLxnUI0Bn13VbAq2ACx3HOQN4BXjTdd3GQBJwW96HKSIiBcnZZ0OTJrYdzMaNkL4g/Bdf2P6mAO3b5//4REREpHDKdYDqmv2e01KeHxfoDHzl6f8IuCxPIxQRkZA7fBjmzIFjx7K+xnXtZ+tWWLjQ+t55B2rXhtat4brrrBJvUhJce6193rIlXH556McvIiIiRUOe1qA6jhPhOM4iYBvwI7AK2O26bornkg1ArbwNUUREQmHhQrj1VsuAPvYYnHkmvP561tfffz9UqwZnnGHnI0bAbZ45Mn/+CWPHQo0a8Pjj1nfxxfD556H9DiIiIlK0OG76eVm5eYjjlAe+Bp4EPnRdt5Gnvw7wveu6p2VyTx+gD0DdunVPX7t2bZ7HISJSVCQlwQsvwEUXwXnn5f45Bw7Am2/CsmUWOJ5yivXv3QsNG8LOnYHXX345jB8Po0ZZNtUbgG7e7JuyC1b4aOpUcBwrhNSuXeBzypeHXbvscxERERHHcea7rht/ouuCUsXXdd3dwHTgDKC84zje7WtqA5uyuGe467rxruvGV6lSJRjDEBEpMj780LKZXbpYhjMxEWbMyHjdwYOwfn3mz9i5E+rVs71GP/0UHnrI+nfvtnWhO3fCzTf7rj/lFPj6a/jgA8us3n47rFgBKSnw8cd2zZQpcOSIVeX1Bp9t28K2bXD0KFx/vfW1aKHgVERERHIuL1V8q3gypziOEwN0AZYCvwBXeS7rBXyb10GKiBQnrguffeY7r13bsp3nngv9+8O8efDgg7Zu9OKLoW5dSE7O+Jw//rAg9LXX4Kab4McfYdAgqFABliyxaz74wDKlK1bASy9BdLQvawowYQI89RQ88ohN3+3Sxarxlky3i3aVKlCqFDRrZudt2gT1VyIiIiLFRK6n+DqO0wIrghSBBbpfuK77rOM4DYHPgYrAQuBG13WPHO9Z8fHxbkJCQq7GISJSVMyaBXfeCatWWZbygw8suFy6NPPrS5SA1FTfvR06BH7+8sswcKBNFx4xAh5+2PorVoRevSxDeuqpgfesWGFrSVu0gEsv9fWfcQZ8+y1UrXr877B5MwwbZgFtTEz2v7uIiIgUbdmd4huUNah5pQBVRIq7xYstKPTq2xeGDLHM6Ny59tn48dC7d+b3v/46PPCA73z4cJvSW78+/PUXbNpkQWbFijZVuFy5E4+pQwf47Tc7njzZ9jIVERERyY3sBqglT3SBiEheffWVBVv//gutWtk6x8jIcI+q4HBd39rNJ56AO+6wab2OY7+ns8+2z2691bZxWb7cAs4ff7TiRMuWwcqVvufNmWOZWLBsKFiBo3XrcjauH3+09adxcXn7fiIiIiLZpQBVREJq5Uro2dPWUEZEwPffw5gxNsVUzLx58PffMHJk1hlSr4EDrT140NaYnnuuFSlaudKm+86caZnUihVtqnD58rkfV+nSub9XREREJDcUoIpISKxcCb//Djt2WBZu6lQr5lO7tk09TUqCrl1tKurTT+ctkCoMvKsp0le2nTPH9h8F2+Ilu0qXtuAUoEkTm4L7+ONW6Ais6m5R/52KiIhI0aMAVURC4okn4PPP7bhePfsB297kq6/g/vt91zqO7dVZlPXoYRnkceOs9Xr1Vd9xhQq5e/btt8Mnn/iC02HD4MYbcz9WERERkXAJyj6oIlJ8TZgAZ50Fv/4a2J+Y6Du+5Rbf8cUX+45jYqwq7PjxvgxjAajbliO7dkFCAtx9t03RveceW2v76KO2Z+ixY1boaOJEq4Lbty/s3++73/t7Gj0692PwrlEFe3efPrl/loiIiEgauq1KAAAa9ElEQVQ4qYqviOTawYM2ZTcpyddXogR06gTTptkU09694cUXISrKPj961NZR9u5tU34/+wzuusu2Uhk/Hj76CBYtKvhblOzaZYWf3n7bpjFnpWpVm97cqpWv7/LL7bv+9ptVyn32Wcs458XAgVY46aOP8vYcERERkVDIbhVfZVBFCoC//rKqqz//DG3a2JrMgi452abwJiVBv36+/tRUC04BXnnFtj/xBqdgVWlffx2aNYPYWLjgAuufNAn++1/LAH78cf59j+NZutSmHh88aFNo/QPRt96CJ5+EsmXtz+zKK2HwYN/nV19t7bZtgcEp2HrRceOgY0c779Yt72N96SUFpyIiIlL4KYMqEkZHjlhA178/jBgR+Nn8+Rb4FEQ7d1oWcOZMWz+6Zw8sWGDHCQnw4INWmbZ168D1lllp394KKnldeil8803wx/3yy1bxtnPnwP7Nm2367Z13+r5D48ZWfGjrVt91vXpZUaeUFDj/fMtYrlwZ+B3nzoUqVaxqcXKyBaFz5tj1/fvbFOdatWwf0sOH7Xuedlrwv6uIiIhIQZLdDKoCVJEwOXoUTj3V1iCmptpx6dIWHHnVqGGZuJtugvh4yy7eeCOc8//t3Xl8VNXdx/HPMQSkKpUIqMgqUgGtEMEFAcUHFYoKaBFwpQpaUfsCtYp2EaWllVYQqYJFKWJFDZtIEZ8KKIgW5QmbBSKGxLAoAiK7KE1ynj9+dzqTkIQsk5lJ+L5fr3nNzLl37pzJ4c7lN+d3zrnMeh6vuCL29V661N47Lw+uvtrqd+ON4e3e2xjLk04q/TG3bLFxlAcPWlC7dq0Ff9HkvaUfg/VqTplikxLdeaf9bVesgMxMa4+rrgq/bvBgWL/e0nELe+IJ60UtycGD8OyzMGgQ1Ktn6bwjRti2ceMsaBURERGp7hSgiiSwzZtt7OHWrdbTlpcHo0bBTTdZ79ywYTBhQsHX7Nhhy4hMmhQuW70a2raNXb29t0By5UqbFKnDUb9iSi8314L2ceMs1Tc9Hdq3r9gx166FVq2gRg3YudPGgxZ25ZWwYEHRr//Tn2xN0aQk6xUeOza8bcwYa6fjyjhQYsWK8N9t504LWkVERESqO41BFUlgDz0Ee/ZYMPbZZ5CTAzffbOmlyck2qVBhr79uqaLt2lkqKsBrr8W02qxYAR9+aIFbNINTsCDyBz+A5s3teej4Tz5pY3M3bSrdcby3cbHZ2ZY6m5wMn34a7pG9/35o0QL69rWJmBYsKPhZQj2/PXpYO4XSd0N/c7DyBx4oe3AKlrY9ZoyN01VwKiIiIlKQelBF4qBlSws0Z8wofp/rr4dVqyx4BTj3XBvvOHgw/OUvcN55NgvuvHkVr08oqEtJKXp7Xp7NWPvYYzbedPNmaNy44u9blJyccJC6ZImlM4fk51sQH2nHDut5btMGfvlLWLjQ0mk7dAinS99wg40fveYaS9Xt2NHK33jDxon262d/6xNPtLYpTn6+pTh36VK+4FRERETkWKUeVJEEE1qKZd06CzTPPbfk/WfPtrGPYL1ua9fapDpNm1pZ69bh7RU1bhyccoqNlQx5/33rgfz6a1sCZehQC07BlpapLM2a2cREdeoUDE7hyOVcRo2CU0+13sjnnrO/zaBBti093T7T8OH2Q8DgwUfW/brrLDgFG/taUnAKFpRedpmCUxEREZHKov9miVSy9HTr9UtJsWVXevWC44+3mWqPpnZt6+FbscLGUoIFcGA9hJ9/boHXzp2lr09uLtx7LzzyiD0/cACeesoev/12eL/bb7egeOZMWL7cAsYzz4SuXY/sxYy2Bg1g7lw46yx7nppq91lZ4X327bMxuWBpwCNHhreFendr17blVy69FL76ysoqM7gWERERkYqpEe8KiFRnX35pvXQhoaBwzpwj18YsTo3gLF2yxMacXnONPR861ALIyZNtDdF164pP0Y00Y0Z4AqZRoyAtzerZqhXMn2/LpFxwgaXOgqXBZmTY53jhhdj1Hl52mc2q6729/znnWM/zxRfDoUPw5z/bfm++aUH/vn3wzTeW9jxwoKX93n+/BdNvvRUeW1rZwbWIiIiIlJ/GoIpUgq+/tmB08mQLMGfNsvTc/v1t++HDNnlPNNxzD0ycCI8+WvTkSpG8t3Th1avt+Rln2ORM48ZZ7+TZZ9vkQffcY4HgaaeFex5Hj4aHH45Oncvq8GGbQCkvz4LyxYvhvvts265dpQvMp061FOCuXSuzpiIiIiJSFI1BFYkD7y2I+sUvLDgFGD/eevh697Zxo+PHRy84BesN7d3bAsjly60OANOnW/AGsGyZpQSPGWPBaffuVv7FFzYj7ymnWOprv37w8ssWnIJNxhRy0UXRq3NZ1axpwTLAJZdYLy9Y+nFpglOwXlUFpyIiIiKJTT2oIlHivc2su3atPW/SxMaIxiIldvduGx9aq5al5rZoYemwrVvDSy8dGVxmZdk+hev/5ZfWqwqWTty/v40Dzc6GgwetFzNeli2znugxY+z52Wfb8jEiIiIikvjUgyoSY0uXhoNTgBEjYjdes25d66Xdvt0CzY0brTwjIzyRUMjTT1swu3JluGzSJLtv2BAuvxx+9CMYMMDGa2ZkWE9rPINTsKVhnnoqvERMaLIoEREREak+FKCKREFamo3lPOEE63nMy4M77ohtHSLTVydOtHGkAAsWwA9/CO+9Z4HrsGFWnpoK335rAe2dd4Zf+847BQPtmjUtcE0U48bZONr77493TUREREQk2pTiK1JB3luPZE6OrR3apUt86rFrF9SrF64T2PNdu6BTJ/jgg/jUS0RERESktCm+WmZGpIL+/ncLTp99Nn7BKdhER9Om2Uy1IaFJkh54ID51EhEREREpCwWoIuXgvY0x7dwZhg+3MZu33hrvWtnan5EmTAivYSoiIiIikugUoIqU0datNr50wYJw2V//CnXqxK9OxRkyJN41EBEREREpPQWocZSba7O8HjhgM6omJcU3RTQacnNhyxZo3jzeNakc335ra4R+8YVNHNSlC9x9t9bXFBERERGJBs3iGye7dtmamUlJNsPq5ZdDt26Qnx+999izx2ZyHTIEPvnEnp98si0dsmpVxY69bZutSRlp/nxITrYJg/bvr9jxy2PXLnjxRUu/9d7Ghu7ZE933mD3bgtN//MPuX39dwamIiIiISLQoQI0D7y1ozMiw56H1Jf/zH/j8cwt+nnnG7ivi0UdtuZHnn4e2baFRI9i717b94Q8F992/Hz77rPTHHjYM+vaFN96w5xMnwtVXh7dnZlas7uXRsaMtl5KebvW67TYYORLeeguWLKn48SdPtnGmp5wCPXtW/HgiIiIiIlKQAtQYmjIF+vSBSy6BGTNg1Cjr4du719aoBEsf7dXLAsBevWDz5vK91yuvWGAaCoS7dYODB+Hcc237Rx/Bb35js74CXHEFnH22pegeTWamrZUJ1qMI4TU3f/97u9+4sXz1Lov9+2290ch6gQWmP/2pPX76abjmmqJ7ObdutR8BFi2CHTuKfo+8PHjhBbjhBhg82Mr69LHUbBERERERiS6tg1qJ9u+35T1SU6FZM7j22nAK75NPwkMPhQOdnTuhQYOij7NkCVx6acGyAwfgwQctTficc2DxYnuvOnWsJ7ZRIzjpJFi2DOrXh+++g9Wr4aKLrLfz3nvDx8rJsfoBrF8PrVsX/5m8hxYtLHBu3NieL1wILVtaj+2vf23ve/75sHx55QVy330HtWtbavTUqXDjjfDhhyW/ZufOguuEtmtnqc9gS7Pk5Bz5muuugzlz7HGfPnD99dZzXLt21D6KiIiIiEi1p3VQ42jrVhtXmpoKWVnh8pNPDo+JvO++gsFb/fpw1llF9zy+/HLBAHXhQrjyyiP3e/VV6zFt29Z6BJ95xo4LcPzx1jsLFmiNHWsTGS1cWLB3sU8f2LCh+M+WlWVpyE8/DXXrws9+ZsEpWFB6wgnw299aau3Pf269j5Vh5Uq7f+89aNLkyO09esC6dTZhU8jy5eHU3N69w8EpwKZN8P33ULMm7N5tEx+tX2/HePRR620OpWKLiIiIiEjlUKJilKWnW89inToWzHXqFN42YoTdN21qgVxhs2fb2Mk1ayw4ysy0tNvJk62XECwFN5S++vDDlsrbpYsFvJmZ1ovarZttb9++6Do2bGiB8DvvWJCak2M9hfXr2zjUr78u/vMtWmT33bvDgAFw+un2PDnZUmnBAtRWrWzCon/966h/snJ5//2iy9u0sfvUVPjJT+zxRx/ZxFChTvrsbEvt7dCh4ERPM2ZAv342xnTGDAtOwYJwBaciIiIiIpVPKb5RtGuXBaQbNliv4oUXWgA5Z471wH3wgd06dbLex9IYPx6GDrUAae9eS2Pt2tVmqL35Zgu8Qv74R+sxnDjRnuflHT3FdsMGmDvXUn4/+cQmGnrxRRg0qOj9r7zSelAzM+298/MtmOvVq2Da6/Tp0L+/PY72P7HvvrPe2uRk6/FMT7e/93PP2e3NNy2ATU2Fw4chJcX2z8214H/RIvscixfDZZdZ/Zo0sR7mmTPD79OmjU0m1bt3dOsvIiIiInKsKW2KrwLUMvLeUkBr17a02Vmz4NAhG5M5ZYoFb+++a4FPtKSlWW/l2LHWQ5qUBF99FR5PWdhrr9lkSD/+cdneJz/fxqiuX2/v2b27vVcoyP3kE0sf/t3vLOAuycaN4dTfsv4Ty821v2vPnjaetbDbb4eXXrIe54YN4aabLA26UyfrDU5LKzi+FyzV+K677DOsWGHHyMqyJXEABg60Y4SMGAGPP162eouIiIiISNEUoFaCgwdtbOIrr9jzZs2OnFjn2mutRzKavvrKUmlr17ZgeM6cyuvV27LFeg4PHLDnQ4ZYGvDDD9vzpCTYvt3SYI/mjjssaN+3r+hAszhTp1pabUqKBbqRvc3e21je66+3Y5dWdrZN7vTYYzBhgqUxHzpkPzJAOPgGm3RqzRr7rCIiIiIiUnGlDVA1BrWUVqywXsxXXrFA55xzLO32lltsptzsbJuw58UXo//ep55q94cOQefOlZty2rhxwfVXJ04MB6dg4zZLE5yCzbALsG2b3R88WHBZmOLMm2f333wDb79t42tDv1/s3m0Bb1l7h5s3t97SkSPDY2xDwSnYbMjff2+fdf58BaciIiIiIvGgWXxLYf16S33Ny7Me1NAYz8KaN6+c948cZ1pZs+JG6trVejBfeunIbZ07l/44DRva/ZYtFtTXqWPjUl991coPH7bgMyXFxnoed5yV/fOfNgZ21ixbVzUjwwLVrKxwanEoNbe0nIPRo209U4Dhw4/cp2ZN20dEREREROJDAWopTJtm9+vWhWeJjbW5c60nslWr2LzflCnW23jrrTYpUVqaBY633FL6Y7Rta8vtTJxoa7zm59v42KFDLeCfOzcccF9xhfW4Ll1q68f26mU/DCxbZtuzs+Hjj8M/Dpx3Xtk/U2pq+PHdd5f99SIiIiIiUrkUoJbC0qVwwQXxC07BxrbGWuPGNtNtyF13le319epZb+ysWQV7gefNswB1zRp7XquWzap7wgk2MVOtWrZUzvTpFqB27myzH48da/uPHl32HlQo2MPdqFHZXy8iIiIiIpVLY1CPIj8fVq2ysZdSdqHxqjNnWg9pgwawY4eVrVtn67xecok9P3jQZua96ioLVkNB6IMP2pjf2bPt+dVXl68uxx1n6cJpaVBDP82IiIiIiCQcBahHsWmTpbi2bx/vmlRNkRMqdeliEz7t2AF9+8Ibb1iqbmTqLYTTiK+91vbr0cMmRcrNtRl8W7cuf31atYJ+/cr/ehERERERqTwKUI+ieXMbE9m/f7xrUjVFzpTbuLH1oKanW9ov2FqpN99c8DWhdOYLLoAZM+wYoVl7O3YsuL6piIiIiIhUH0p0LIXIIEvK5vDh8OOUFAtQFy0Kl/Xvb72oOTm2tMv+/bbea2GhSZE6darU6oqIiIiISByVO0B1zjUGXgZOA/KBSd77Z5xzKUAa0AzIAfp573dXvKpSFXkfftyiBZx2Wvj55s3WqwrQtGnJx+nUycaw9u0b/TqKiIiIiEhiqEiyZC7woPe+NXAxcK9zrg3wCLDIe98SWBQ8l2PUQw/BTTfZEjFt2xacCTm0Tmpp1K0L775rkyqJiIiIiEj1VO4eVO/9NmBb8Hi/cy4DOAPoDXQNdpsKLAaGV6iWUmXVqxdeRxYKrl+alBT7+oiIiIiISOKKynQzzrlmQCrwMXBqELyGgtgG0XgPqR7atbNZdCPXVxUREREREYEoTJLknDsRmAUM897vc86V9nV3AXcBNGnSpKLVkCqiZk1bh1RERERERKSwCvWgOueSseB0mvd+dlC83Tl3erD9dGBHUa/13k/y3nfw3neoX79+RaohIiIiIiIi1UC5A1RnXaWTgQzv/diITXOBgcHjgcCb5a+eiIiIiIiIHCsqkuLbCbgV+LdzbnVQ9ivgSWC6c24QsBm4oWJVFBERERERkWNBRWbx/QAobsBpt/IeV0RERERERI5NUZnFV0RERERERKSiFKCKiIiIiIhIQlCAKiIiIiIiIglBAaqIiIiIiIgkBAWoIiIiIiIikhAUoIqIiIiIiEhCUIAqIiIiIiIiCUEBqoiIiIiIiCQEBagiIiIiIiKSEBSgioiIiIiISEJQgCoiIiIiIiIJQQGqiIiIiIiIJAQFqCIiIiIiIpIQFKCKiIiIiIhIQnDe+3jXAefcTmBTvOsRJ/WAr+NdCYkatWf1ovasXtSe1Yvas3pRe1Yvas/qJVrt2dR7X/9oOyVEgHosc86le+87xLseEh1qz+pF7Vm9qD2rF7Vn9aL2rF7UntVLrNtTKb4iIiIiIiKSEBSgioiIiIiISEJQgBp/k+JdAYkqtWf1ovasXtSe1Yvas3pRe1Yvas/qJabtqTGoIiIiIiIikhDUgyoiIiIiIiIJQQFqJXDO/c05t8M5tzai7HHn3BfOudXBrWfEtkedcxudcxucc90jynsEZRudc4/E+nOIKaY90yLaMsc5tzoob+acOxSx7fmI17R3zv07aM/xzjkXj89zrHPONXbOveecy3DOrXPODQ3KU5xzC5xzmcF93aDcBe210Tn3iXPu/IhjDQz2z3TODYzXZzqWldCef3bOfRq02RvOuZODcp2jCayE9tQ1tAoqoT11Da2CnHPHO+eWO+fWBO35RFDe3Dn3cXAtTHPO1QzKawXPNwbbm0Ucq8jzVmKnhPacFrTLWmf/B04Oyrs65/ZGnJ+PRRwrut+33nvdonwDLgXOB9ZGlD0O/LKIfdsAa4BaQHMgC0gKblnAmUDNYJ828f5sx+KtqPYstH0M8FjwuFkJ+y0HOgIOeBv4Sbw/27F4A04Hzg8enwR8FpyHfwIeCcofAUYHj3sG7eWAi4GPg/IUIDu4rxs8rhvvz3es3Upoz6uAGkH56Ij21DmawLcS2lPX0Cp4K649C+2ja2gVuQV/+xODx8nAx8F1cTowICh/HhgSPL4HeD54PABICx4Xed7G+/Mda7cS2rNnsM0Br0W0Z1dgXhHHifr3rXpQK4H3/n3gm1Lu3ht43Xv/vff+c2AjcGFw2+i9z/beHwZeD/aVGCupPYNfcPthJ3CxnHOnA3W898u8nc0vA32iXVc5Ou/9Nu/9yuDxfiADOAM7v6YGu00l3D69gZe9+Qg4OWjP7sAC7/033vvdwAKgRww/ilB8e3rv3/He5wa7fQQ0Kuk4OkcTQwnnZ3F0DU1gR2tPXUOrluA6eCB4mhzcPPA/wMygvPD1M3RdnQl0C9q8uPNWYqi49vTezw+2eeyHoRKvn1TC960C1Ni6L0g3+1sofRD7ot4Ssc/WoKy4ckksXYDt3vvMiLLmzrlVzrklzrkuQdkZWBuGqD0TQJBulIr9aniq934b2H+qgAbBbjpHq4hC7RnpDqzHJUTnaBVQRHvqGlqFFXN+6hpaxTjnkoKU7B3YD7NZwJ6IHwQj2+a/52GwfS9wCjo/E0bh9vTefxyxLRm4FfjfiJd0DFKC33bOnROURb09FaDGzkSgBdAO2IaltIB1nxfmSyiXxHIjBX/53QY08d6nAg8Arzrn6qD2TDjOuROBWcAw7/2+knYtokznaIIprj2dc78GcoFpQZHO0SqgiPbUNbQKK+H7VtfQKsZ7n+e9b4f1ql0ItC5qt+Be52eCK9yezrlzIzZPAN733i8Nnq8Emnrv2wJ/AeYE5VFvTwWoMeK93x78I8gHXiCcyrAVaByxayPgyxLKJUE452oA1wNpobIgXWVX8HgF9svij7D2jEyRUHvGUfCr4Cxgmvd+dlC8PUgjC6WT7QjKdY4muGLaE2cTV10D3BykKukcrQKKak9dQ6uuEs5PXUOrMO/9HmAxNmbx5KA9oWDb/Pc8DLb/EBsypfMzwUS0Zw8A59wIoD72Q1Fon32hlGDv/Xwg2TlXj0poTwWoMRL6j2/gOiA0I+xcYEAw01lzoCWW7/1/QMtgZrSa2ODyubGssxzVFcCn3vv/ph055+o755KCx2di7ZkdpIzud85dHIy/uA14Mx6VPtYFf//JQIb3fmzEprlAaCbegYTbZy5wmzMXA3uD9vwncJVzrm6QbnhVUCYxVFx7Oud6AMOBXt77byPKdY4msBLaU9fQKqiE71vQNbTKCdonNCN6bawNM4D3gL7BboWvn6Hral/g3eDHwuLOW4mhYtrzU+fcYGyejRuDHwVD+58WnH845y7E4shdVML3bY2j7yJl5Zx7DZvpqp5zbiswAujqnGuHdXnnAD8H8N6vc85NB9ZjaWj3eu/zguPch/2HNwn4m/d+XYw/ilB0e3rvJ2MnYOGJHS4FRjrncoE84G7vfWiCpSHAS0BtbDzc20g8dMLGVPw7GHcB8CvgSWC6c24QsBm4Idg2H5vRbiPwLXA7gPf+G+fc77AvZoCREW0tsVNce47HZohcEFxPP/Le343O0URXXHveqGtolVRkewa9L7qGVj2nA1ODHxGOA6Z77+c559YDrzvnfg+swn6UILj/u3NuI9ZzOgBKPm8lpoprz1xgE7AsuH7O9t6PxH5kGBJsP4TN3OyB3Gh/37og60lEREREREQkrpTiKyIiIiIiIglBAaqIiIiIiIgkBAWoIiIiIiIikhAUoIqIiIiIiEhCUIAqIiIiIiIiCUEBqoiIiIiIiCQEBagiIiIiIiKSEBSgioiIiIiISEL4f2OEE6W1aYGiAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<Figure size 1152x576 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "corrected_preds = [x - (intercept + (i + 1500) * slope) for i,x in enumerate(strange_preds)]\n",
    "corrected =  pd.DataFrame(corrected_preds,index = true_vals['Close'].index,columns=['Prediction'])\n",
    "plt.figure(figsize=(16,8))\n",
    "plt.plot(temp_df['Close'], label='True Training Period Values', color = 'b')\n",
    "plt.plot(true_vals['Close'], label='True Prediction Period Values', color = 'g')\n",
    "plt.plot(corrected['Prediction'], label='Corrected Prediction Period', color = 'yellow')\n",
    "plt.plot(stranger_preds['Prediction'], label='Unorrected Prediction Period', color = 'orange')\n",
    "plt.title('Linear Reggression Translation Comparison')\n",
    "plt.legend(loc='upper left')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Predicting the stocks separately\n",
    "We will use our best model configuration in order to predict each individual stock price:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "stock_scores = {}\n",
    "stock_predictions = {}\n",
    "hidden_states = 2\n",
    "num_symbols = 2\n",
    "training_days = 1500\n",
    "for stock in xlk_stocks.keys():\n",
    "    #training\n",
    "    start = time.time()\n",
    "    training =  get_training_symbols(stock_dict[stock], training_days, num_symbols)\n",
    "    temp_enu, temp_reverse_enu = enumerate_features(training)\n",
    "    predicted_data = predict(250, hidden_states, training, temp_enu, temp_reverse_enu) #predict 250 days into the future\n",
    "    predicted = predicted_data[-250:]\n",
    "    end = time.time()\n",
    "    stock_predictions[stock] = predicted\n",
    "    print(stock,' training time is: ', end - start)\n",
    "\n",
    "    #prediction and translation       \n",
    "    temp_df = stock_dict[stock].iloc[-(training_days + 250):-250, :]\n",
    "    temp_true_vals = stock_dict[stock].iloc[-250:,:]['Close'].values\n",
    "    final_vals = translate(temp_df, predicted, 100) #number of bootstrap iterations = 100\n",
    "    \n",
    "    #score calculation\n",
    "    temp_score = get_MAPE_score(temp_true_vals, final_vals)\n",
    "    print(stock,' score is: ',temp_score)\n",
    "    stock_scores[stock] = temp_score"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "metadata": {},
   "outputs": [],
   "source": [
    "file_name = open('stock_scores', 'rb')\n",
    "stock_scores =  pickle.load(file_name)\n",
    "file_name = open('stock_predictions', 'rb')\n",
    "stock_predictions = pickle.load(file_name)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "metadata": {},
   "outputs": [],
   "source": [
    "stock_translation = {}\n",
    "for k, val in stock_predictions.items():\n",
    "    temp_df = stock_dict[k].iloc[-(1500 + 250):-250, :]\n",
    "    stock_translation[k] = translate(temp_df, val, 100)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "metadata": {},
   "outputs": [],
   "source": [
    "stock_preds = []\n",
    "for i in range(250):\n",
    "    temp_price = sum([stock_translation[s][-(250 - i)] * xlk_stocks[s] for s in xlk_stocks.keys()])\n",
    "    stock_preds.append(temp_price)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Finally, using linear regression to compensate for \"translation error\" we will plot all our predictions together to recreate the XLK fund:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.legend.Legend at 0x1be95c6d5c0>"
      ]
     },
     "execution_count": 43,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1152x576 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "true_vals = etf_dict['xlk'].iloc[-250:,:]\n",
    "temp_df = etf_dict['xlk'].iloc[-(1500 + 250):-250, :]\n",
    "corrected_stock_preds = [x - (intercept + (i + 1500) * slope) for i,x in enumerate(stock_preds)]\n",
    "corrected_stock_preds_df =  pd.DataFrame(corrected_stock_preds,index = true_vals['Close'].index,columns=['Prediction'])\n",
    "\n",
    "original_vals = translate(temp_df, best_predictions[('xlk',2,1500)], 100)\n",
    "translated = pd.DataFrame(original_vals,index = true_vals['Close'].index,columns=['Prediction'])\n",
    "\n",
    "plt.figure(figsize=(16,8))\n",
    "plt.plot(temp_df['Close'], label='True Training Period Values', color = 'b')\n",
    "plt.plot(true_vals['Close'], label='True Prediction Period Values', color = 'g')\n",
    "plt.plot(corrected_stock_preds_df['Prediction'], label='Upgraded Method Predictions', color = 'r')\n",
    "plt.plot(translated['Prediction'], label='Old Method Predictions', color = 'black')\n",
    "plt.title('Prediction Methods Comparison')\n",
    "plt.legend(loc='upper left')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We can see that the new method resembles the true value of the fund better than our previous attempt.\n",
    "\n",
    "Now, let's compare the errors of the two methods and their MAPE scores to get a better understanding:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 141,
   "metadata": {},
   "outputs": [],
   "source": [
    "corrected_stock_preds_lis = [x[0] for x in corrected_stock_preds]\n",
    "prediction_error_corrected = np.array(etf_dict['xlk'].iloc[-250:, :]['Close'].values) -  np.array(corrected_stock_preds_lis) \n",
    "prediction_error_original = np.array(etf_dict['xlk'].iloc[-250:, :]['Close'].values) - original_vals"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 221,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.legend.Legend at 0x13d84311630>"
      ]
     },
     "execution_count": 221,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1152x576 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.figure(figsize=(16,8))\n",
    "plt.plot(prediction_error_corrected, label ='Upgraded Method Errors',  color = 'red')\n",
    "plt.plot(prediction_error_original, label ='Old Method Errors', color = 'black')\n",
    "plt.title('Prediction Error Comparison')\n",
    "plt.legend(loc='upper left')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 46,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Original MAPE score:  15.407698149158271\n",
      "Upgraded method MAPE score:  10.61374890445389\n"
     ]
    }
   ],
   "source": [
    "xlk_true_vals = etf_dict['xlk'].iloc[-250:,:]['Close'].values\n",
    "print('Original MAPE score: ', get_MAPE_score(xlk_true_vals, original_vals))\n",
    "print('Upgraded method MAPE score: ', get_MAPE_score(xlk_true_vals, corrected_stock_preds))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "It's visible that the prediction error of our new method is consistently lower than that of our best model yet, and combined with lower MAPE score ( 10.6 against  15.4) we proved the new method to be superior."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Creative Part"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "In this part, we will try to use the leading financial institutions' recommendations for promising stocks over the years to further improve our prediction method.\n",
    "Moreover, we will analyze the predictions made by leading financial investors and commentators for \"10 best stocks\" for the year 2017 and compare them to our best predictions. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 144,
   "metadata": {},
   "outputs": [],
   "source": [
    "best_for_2016 = ['pfe','bac','hon','aig','ddaiy','tel','brs','kmt','kkr','tjx'] #http://money.com/money/4161047/great-stocks-buy-2016/\n",
    "best_for_2015 = ['abm','ne','yhoo','googl','amba','psec','pzzi','aapl','odfl','fb'] #https://www.nasdaq.com/articles/10-best-stocks-buy-2015-2015-01-05\n",
    "best_for_2014 = ['msft','exel','rmti','crm','ftk','fdx','nke','nfg','twtr','ibm'] #https://www.nasdaq.com/articles/10-stocks-watch-2014-2013-11-19\n",
    "best_for_2013 = ['ddaiy','myl','intc','qcom','shw','gldd','fmx','two','vale','nflx'] #https://investorplace.com/best-stocks-for-2013/\n",
    "best_for_2012 = ['cat','fdx','hsy','cof','mako','msft','arco','tkc','aa','std'] #https://www.forbes.com/sites/tomtaulli/2011/12/27/experts-top-stock-picks-for-2012/#3ff2733f29af\n",
    "best_dic = {2016:best_for_2016, 2015:best_for_2015, 2014:best_for_2014, 2013:best_for_2013, 2012:best_for_2012}"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 65,
   "metadata": {},
   "outputs": [],
   "source": [
    "all_tickers = best_for_2016 + best_for_2015 + best_for_2014 + best_for_2013 + best_for_2012\n",
    "unique_bests = set([b for b in all_tickers])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 96,
   "metadata": {},
   "outputs": [],
   "source": [
    "best_stocks_dict = get_stock_data(unique_bests)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Let's focus on that group of stocks, including recurring promising stocks and other well known brands e.g Netflix, Caterpiller and more:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 97,
   "metadata": {},
   "outputs": [],
   "source": [
    "target_stocks = ['fdx','msft','bac','aapl','fmx','qcom','nflx','cat','pfe','aapl']"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Now, let's add for each stock data frame a new symbol representing the predicted behavior over the years where 'U' stands for Up and 'R' stands for Regular:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 98,
   "metadata": {},
   "outputs": [],
   "source": [
    "for s in target_stocks:\n",
    "    temp_df = best_stocks_dict[s]\n",
    "    best_years = [k for k in best_dic.keys() if s in best_dic[k]]\n",
    "    temp_df['New Char'] = ['U' if x in best_years else 'R' for x in temp_df['Year']]\n",
    "    temp_df['Two_Symbols'] = temp_df['Two_Symbols'] + temp_df['New Char']   "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Finally, we will predict each stock price:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "upgradeed_predictions = {}\n",
    "upgradeed_scores = {}\n",
    "hidden_states = 2\n",
    "num_symbols = 2\n",
    "training_days = 1500\n",
    "for stock in target_stocks:\n",
    "    #training\n",
    "    start = time.time()\n",
    "    training =  get_training_symbols(best_stocks_dict[stock], training_days, num_symbols)\n",
    "    temp_enu, temp_reverse_enu = enumerate_features(training)\n",
    "    predicted_data = predict(250, hidden_states, training, temp_enu, temp_reverse_enu) #predict 250 days into the future\n",
    "    predicted = predicted_data[-250:]\n",
    "    end = time.time()\n",
    "    upgradeed_predictions[stock] = predicted\n",
    "    print(stock,' training time is: ', end - start)\n",
    "\n",
    "    #prediction and translation       \n",
    "    temp_df = best_stocks_dict[stock].iloc[-(training_days + 250):-250, :]\n",
    "    temp_true_vals = best_stocks_dict[stock].iloc[-250:,:]['Close'].values\n",
    "    final_vals = translate(temp_df, predicted, 100) #number of bootstrap iterations = 100\n",
    "    \n",
    "    #score calculation\n",
    "    temp_score = get_MAPE_score(temp_true_vals, final_vals)\n",
    "    print(stock,' score is: ',temp_score)\n",
    "    upgradeed_scores[stock] = temp_score"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 103,
   "metadata": {},
   "outputs": [],
   "source": [
    "file_name = open('upgraded_scores', 'rb')\n",
    "upgraded_scores =  pickle.load(file_name)\n",
    "file_name = open('upgraded_predictions', 'rb')\n",
    "upgraded_predictions =  pickle.load(file_name)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Now, let's compare these results to our normal predictions, using MAPE, accuracy and F1 measure:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 109,
   "metadata": {},
   "outputs": [],
   "source": [
    "file_name = open('regular_scores', 'rb')\n",
    "regular_scores =  pickle.load(file_name)\n",
    "file_name = open('regular_predictions', 'rb')\n",
    "regular_predictions =  pickle.load(file_name)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 111,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "{'regular': 143.36113401811053, 'upgraded': 136.16127465393544}\n"
     ]
    }
   ],
   "source": [
    "comparison_scores = {'regular':0, 'upgraded':0}\n",
    "for k in regular_scores.keys():\n",
    "    comparison_scores['regular'] += regular_scores[k]\n",
    "    comparison_scores['upgraded'] += upgradeed_scores[k]\n",
    "print(comparison_scores)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 119,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "week_changes accuracy is:  0.5555555555555556  f1 score is: 0.2380952380952381\n",
      "month_changes accuracy is:  0.5555555555555556  f1 score is: 0.2380952380952381\n",
      "quarter_changes accuracy is:  0.4444444444444444  f1 score is: 0.3148148148148148\n",
      "year_changes accuracy is:  0.3333333333333333  f1 score is: 0.26666666666666666\n"
     ]
    }
   ],
   "source": [
    "get_f1_acc_scores(regular_predictions, best_stocks_dict)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 120,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "week_changes accuracy is:  0.6666666666666666  f1 score is: 0.26666666666666666\n",
      "month_changes accuracy is:  0.6666666666666666  f1 score is: 0.5897435897435898\n",
      "quarter_changes accuracy is:  0.5555555555555556  f1 score is: 0.41269841269841273\n",
      "year_changes accuracy is:  0.4444444444444444  f1 score is: 0.3151515151515151\n"
     ]
    }
   ],
   "source": [
    "get_f1_acc_scores(upgradeed_predictions, best_stocks_dict)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Examining the results of our comparison we can spot a slight improvement in the MAPE measure in favor of our improved method and overall better performance using our intuitive periodic comparison method where the monthly changes demonstrate the highest improvement (0.23 to 0.58 increase in F1 measure)."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## \"Top 10 Stocks for 2017\" recommendations analysis "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Why bother with computational learning? let's just take the word of professionals and have a look at their recommendations for the best stocks for the year 2017.\n",
    "\n",
    "For the sake of simplicity, in this section, we consider a stock recommendation to be correct if it's total growth over the year is greater than 10%"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 146,
   "metadata": {},
   "outputs": [],
   "source": [
    "best_for_2017_cnvc = ['abt','googl','bdx','cvs','xom','jnj','low','msft','sfm','utx'] #https://www.cnbc.com/2016/12/29/top-10-stocks-for-2017-commentary.html\n",
    "best_for_2017_investorplace = ['lrcx','vz','pru','bac','unh','wmt','hd','googl','lmt','has'] #https://investorplace.com/2016/12/10-best-stocks-to-buy-2017/\n",
    "best_for_2017_money = ['schw','nov','unp','prblx','cvs','gild','intc','agco','sci','ttc','mtn','uovey'] #http://money.com/money/4616745/investing-stock-picks-from-pros/"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 147,
   "metadata": {},
   "outputs": [],
   "source": [
    "all_recs = best_for_2017_cnvc + best_for_2017_investorplace + best_for_2017_money\n",
    "unique_recs = set([rec for rec in all_recs])\n",
    "rec_stocks_dict = get_stock_data(unique_recs)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 170,
   "metadata": {},
   "outputs": [],
   "source": [
    "percent_changes = {}\n",
    "for k in rec_stocks_dict.keys():\n",
    "    percent_changes[k] = (float(rec_stocks_dict[k].iloc[-1]['Close'] - rec_stocks_dict[k].iloc[-218]['Close'] ) / rec_stocks_dict[k].iloc[-218]['Close']) * 100"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 184,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Overall predictions accuracy for stocks with 10% growth: 0.6666666666666666\n"
     ]
    }
   ],
   "source": [
    "print('Overall predictions accuracy for stocks with 10% growth:', sum([1 for x in percent_changes.values() if x >= 10]) / len(unique_recs))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 175,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Average stock price change in precents:  18.559615755524433\n"
     ]
    }
   ],
   "source": [
    "print('Average stock price change in precents: ', sum(percent_changes.values()) / len(unique_recs))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 185,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "best recommendation for 2017 is: lrcx (Lam Research Corporation) with  95.15385339230262 % growth\n"
     ]
    }
   ],
   "source": [
    "print('best recommendation for 2017 is:',[k for k in percent_changes.keys() if percent_changes[k] == max(percent_changes.values())][0],'(Lam Research Corporation) with ',max(percent_changes.values()), '% growth' ) "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 186,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Worst recommendation for 2017 is: vz (Verizon Communications Inc) with  -13.775216138328519 % lost\n"
     ]
    }
   ],
   "source": [
    "print('Worst recommendation for 2017 is:',[k for k in percent_changes.keys() if percent_changes[k] == min(percent_changes.values())][0],'(Verizon Communications Inc) with ',min(percent_changes.values()), '% lost' ) "
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.7.3"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}