{
"cells": [
{
"cell_type": "markdown",
"metadata": {},
"source": [
"# Ejercicio de Regresión Lineal"
]
},
{
"cell_type": "markdown",
"metadata": {
"ExecuteTime": {
"end_time": "2018-05-05T08:19:33.323030Z",
"start_time": "2018-05-05T08:19:33.308960Z"
}
},
"source": [
"En este notebook intentaremos predecir cuántas veces será compartido en Redes Sociales un artículo de Machine Learning segun algunas de sus características"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Predicción: ¿Cuántas veces será compartido un artículo del Blog?"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Vamos a comenzar por Importar y Visualizar los datos"
]
},
{
"cell_type": "code",
"execution_count": 20,
"metadata": {
"ExecuteTime": {
"end_time": "2018-05-06T15:28:50.037886Z",
"start_time": "2018-05-06T15:28:50.028212Z"
}
},
"outputs": [],
"source": [
"# Imports necesarios\n",
"import numpy as np\n",
"import pandas as pd\n",
"import seaborn as sb\n",
"import matplotlib.pyplot as plt\n",
"%matplotlib inline\n",
"from mpl_toolkits.mplot3d import Axes3D\n",
"from matplotlib import cm\n",
"plt.rcParams['figure.figsize'] = (16, 9)\n",
"plt.style.use('ggplot')\n",
"from sklearn import linear_model\n",
"from sklearn.metrics import mean_squared_error, r2_score"
]
},
{
"cell_type": "code",
"execution_count": 21,
"metadata": {
"ExecuteTime": {
"end_time": "2018-05-06T15:28:50.051323Z",
"start_time": "2018-05-06T15:28:50.040878Z"
}
},
"outputs": [],
"source": [
"#cargamos los datos de entrada\n",
"data = pd.read_csv(\"./articulos_ml.csv\")"
]
},
{
"cell_type": "code",
"execution_count": 22,
"metadata": {
"ExecuteTime": {
"end_time": "2018-05-06T15:28:50.064402Z",
"start_time": "2018-05-06T15:28:50.055301Z"
}
},
"outputs": [
{
"data": {
"text/plain": [
"(161, 8)"
]
},
"execution_count": 22,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"#veamos cuantas dimensiones y registros contiene\n",
"data.shape"
]
},
{
"cell_type": "code",
"execution_count": 23,
"metadata": {
"ExecuteTime": {
"end_time": "2018-05-06T15:28:50.093837Z",
"start_time": "2018-05-06T15:28:50.068610Z"
}
},
"outputs": [
{
"data": {
"text/html": [
"
\n",
"\n",
"
\n",
" \n",
"
\n",
"
\n",
"
Title
\n",
"
url
\n",
"
Word count
\n",
"
# of Links
\n",
"
# of comments
\n",
"
# Images video
\n",
"
Elapsed days
\n",
"
# Shares
\n",
"
\n",
" \n",
" \n",
"
\n",
"
0
\n",
"
What is Machine Learning and how do we use it ...
\n",
"
https://blog.signals.network/what-is-machine-l...
\n",
"
1888
\n",
"
1
\n",
"
2.0
\n",
"
2
\n",
"
34
\n",
"
200000
\n",
"
\n",
"
\n",
"
1
\n",
"
10 Companies Using Machine Learning in Cool Ways
\n",
"
NaN
\n",
"
1742
\n",
"
9
\n",
"
NaN
\n",
"
9
\n",
"
5
\n",
"
25000
\n",
"
\n",
"
\n",
"
2
\n",
"
How Artificial Intelligence Is Revolutionizing...
\n",
"
NaN
\n",
"
962
\n",
"
6
\n",
"
0.0
\n",
"
1
\n",
"
10
\n",
"
42000
\n",
"
\n",
"
\n",
"
3
\n",
"
Dbrain and the Blockchain of Artificial Intell...
\n",
"
NaN
\n",
"
1221
\n",
"
3
\n",
"
NaN
\n",
"
2
\n",
"
68
\n",
"
200000
\n",
"
\n",
"
\n",
"
4
\n",
"
Nasa finds entire solar system filled with eig...
\n",
"
NaN
\n",
"
2039
\n",
"
1
\n",
"
104.0
\n",
"
4
\n",
"
131
\n",
"
200000
\n",
"
\n",
" \n",
"
\n",
"
"
],
"text/plain": [
" Title \\\n",
"0 What is Machine Learning and how do we use it ... \n",
"1 10 Companies Using Machine Learning in Cool Ways \n",
"2 How Artificial Intelligence Is Revolutionizing... \n",
"3 Dbrain and the Blockchain of Artificial Intell... \n",
"4 Nasa finds entire solar system filled with eig... \n",
"\n",
" url Word count # of Links \\\n",
"0 https://blog.signals.network/what-is-machine-l... 1888 1 \n",
"1 NaN 1742 9 \n",
"2 NaN 962 6 \n",
"3 NaN 1221 3 \n",
"4 NaN 2039 1 \n",
"\n",
" # of comments # Images video Elapsed days # Shares \n",
"0 2.0 2 34 200000 \n",
"1 NaN 9 5 25000 \n",
"2 0.0 1 10 42000 \n",
"3 NaN 2 68 200000 \n",
"4 104.0 4 131 200000 "
]
},
"execution_count": 23,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"#son 161 registros con 8 columnas. Veamos los primeros registros para tener una idea\n",
"data.head()"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Vemos que algunos campos (por ejemplo en comentarios) tienen valores nulos.\n",
"\n",
"En nuestro caso la columna Shares será nuestra salida, es decir nuestro valor \"Y\", el valor que queremos predecir"
]
},
{
"cell_type": "code",
"execution_count": 24,
"metadata": {
"ExecuteTime": {
"end_time": "2018-05-06T15:28:50.126775Z",
"start_time": "2018-05-06T15:28:50.097200Z"
}
},
"outputs": [
{
"data": {
"text/html": [
"
\n",
"\n",
"
\n",
" \n",
"
\n",
"
\n",
"
Word count
\n",
"
# of Links
\n",
"
# of comments
\n",
"
# Images video
\n",
"
Elapsed days
\n",
"
# Shares
\n",
"
\n",
" \n",
" \n",
"
\n",
"
count
\n",
"
161.000000
\n",
"
161.000000
\n",
"
129.000000
\n",
"
161.000000
\n",
"
161.000000
\n",
"
161.000000
\n",
"
\n",
"
\n",
"
mean
\n",
"
1808.260870
\n",
"
9.739130
\n",
"
8.782946
\n",
"
3.670807
\n",
"
98.124224
\n",
"
27948.347826
\n",
"
\n",
"
\n",
"
std
\n",
"
1141.919385
\n",
"
47.271625
\n",
"
13.142822
\n",
"
3.418290
\n",
"
114.337535
\n",
"
43408.006839
\n",
"
\n",
"
\n",
"
min
\n",
"
250.000000
\n",
"
0.000000
\n",
"
0.000000
\n",
"
1.000000
\n",
"
1.000000
\n",
"
0.000000
\n",
"
\n",
"
\n",
"
25%
\n",
"
990.000000
\n",
"
3.000000
\n",
"
2.000000
\n",
"
1.000000
\n",
"
31.000000
\n",
"
2800.000000
\n",
"
\n",
"
\n",
"
50%
\n",
"
1674.000000
\n",
"
5.000000
\n",
"
6.000000
\n",
"
3.000000
\n",
"
62.000000
\n",
"
16458.000000
\n",
"
\n",
"
\n",
"
75%
\n",
"
2369.000000
\n",
"
7.000000
\n",
"
12.000000
\n",
"
5.000000
\n",
"
124.000000
\n",
"
35691.000000
\n",
"
\n",
"
\n",
"
max
\n",
"
8401.000000
\n",
"
600.000000
\n",
"
104.000000
\n",
"
22.000000
\n",
"
1002.000000
\n",
"
350000.000000
\n",
"
\n",
" \n",
"
\n",
"
"
],
"text/plain": [
" Word count # of Links # of comments # Images video Elapsed days \\\n",
"count 161.000000 161.000000 129.000000 161.000000 161.000000 \n",
"mean 1808.260870 9.739130 8.782946 3.670807 98.124224 \n",
"std 1141.919385 47.271625 13.142822 3.418290 114.337535 \n",
"min 250.000000 0.000000 0.000000 1.000000 1.000000 \n",
"25% 990.000000 3.000000 2.000000 1.000000 31.000000 \n",
"50% 1674.000000 5.000000 6.000000 3.000000 62.000000 \n",
"75% 2369.000000 7.000000 12.000000 5.000000 124.000000 \n",
"max 8401.000000 600.000000 104.000000 22.000000 1002.000000 \n",
"\n",
" # Shares \n",
"count 161.000000 \n",
"mean 27948.347826 \n",
"std 43408.006839 \n",
"min 0.000000 \n",
"25% 2800.000000 \n",
"50% 16458.000000 \n",
"75% 35691.000000 \n",
"max 350000.000000 "
]
},
"execution_count": 24,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"# Ahora veamos algunas estadísticas de nuestros datos\n",
"data.describe()"
]
},
{
"cell_type": "markdown",
"metadata": {
"ExecuteTime": {
"end_time": "2018-05-05T08:36:49.122754Z",
"start_time": "2018-05-05T08:36:49.108512Z"
}
},
"source": [
"De aqui observamos que por ejemplo la media de Cantidad de palabras es 1808\n",
"\n",
"Hay un artíclo mínimo con 250 palabras y el máximo contiene 8401.\n",
"\n",
"Y en cuanto a las salidas, vemos mínimo de 0 veces compartido y máximo de 350000 (eso es mucho!)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Visualización General"
]
},
{
"cell_type": "code",
"execution_count": 39,
"metadata": {
"ExecuteTime": {
"end_time": "2018-05-06T15:50:03.185268Z",
"start_time": "2018-05-06T15:50:02.485816Z"
}
},
"outputs": [
{
"data": {
"image/png": "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\n",
"text/plain": [
"
"
]
},
"metadata": {},
"output_type": "display_data"
}
],
"source": [
"#vamos a Visualizar los datos de entrada\n",
"colores=['orange','blue']\n",
"tamanios=[30,60]\n",
"\n",
"f1 = data['Word count'].values\n",
"f2 = data['# Shares'].values\n",
"\n",
"# Vamos a pintar en 2 colores los puntos por debajo de la media de Cantidad de Palabras\n",
"asignar=[]\n",
"for index, row in data.iterrows():\n",
" if(row['Word count']>1808):\n",
" asignar.append(colores[0])\n",
" else:\n",
" asignar.append(colores[1])\n",
" \n",
"plt.scatter(f1, f2, c=asignar, s=tamanios[0])\n",
"plt.show()"
]
},
{
"cell_type": "code",
"execution_count": 26,
"metadata": {
"ExecuteTime": {
"end_time": "2018-05-06T15:28:50.963199Z",
"start_time": "2018-05-06T15:28:50.509749Z"
}
},
"outputs": [
{
"data": {
"image/png": "iVBORw0KGgoAAAANSUhEUgAAA7UAAAIMCAYAAAAq64s8AAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADl0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uIDIuMi4yLCBodHRwOi8vbWF0cGxvdGxpYi5vcmcvhp/UCwAAIABJREFUeJzs3Xt8VPWB///3mVtuk8tcEsJN5KqC0kSiRXQBIbXtWltWrdauv24RKxW/ILht1WrrtgpibQW5WFsv2G7tatcq7brVtilFduHBNghBbgoIyt2QTEgyk8vczu8PJJomSIAkZ07m9Xw8fDycD2cm7+FDJnnP53POGKZpmgIAAAAAwIYcVgcAAAAAAOBMUWoBAAAAALZFqQUAAAAA2BalFgAAAABgW5RaAAAAAIBtUWoBAAAAALZFqQUAAAAA2BalFgAAAABgW5RaAAAAAIBtUWoBAAAAALZFqQUAAAAA2JarKwe9+uqrWrVqlQzD0ODBgzVr1iwdO3ZMixcvVjgc1tChQzV79my5XC7FYjEtW7ZMe/bsUW5urubOnauioiJJ0iuvvKJVq1bJ4XBo+vTpKikpkSRVVVVpxYoVSiaTmjp1qqZNm9al8IcOHTrDp42eFAwGVVNTY3UM9CLmPD0x7+mHOU9PzHv6Yc7TU6rN+4ABA7p87ClXakOhkF577TUtXLhQP/nJT5RMJrVu3Tr96le/0tVXX60lS5YoJydHq1atkiStWrVKOTk5Wrp0qa6++mo9//zzkqQDBw5o3bp1euyxx3TffffpmWeeUTKZVDKZ1DPPPKPvfve7WrRokdauXasDBw6c4VMHAAAAAKSTLm0/TiaTikajSiQSikajKigo0LZt2zR+/HhJ0uTJk1VZWSlJ2rBhgyZPnixJGj9+vLZu3SrTNFVZWakJEybI7XarqKhIxcXF2r17t3bv3q3i4mL169dPLpdLEyZMaHssAAAAAAA+ySm3H/v9fl1zzTW6/fbb5fF49KlPfUrDhg1Tdna2nE5n2zGhUEjS8ZXdQCAgSXI6ncrOzlZjY6NCoZBGjhzZ7nFP3OfE8Sf+f9euXZ1mqaioUEVFhSRp4cKFCgaDZ/Kc0cNcLhdzk2aY8/TEvKcf5jw9Me/phzlPT3ae91OW2nA4rMrKSi1fvlzZ2dl67LHHVFVVddLjTdPsMGYYRqfjn3R8Z8rLy1VeXt52O5X2fOMjqbYfHz2POU9PzHv6Yc7TE/Oefpjz9JRq834659SestRu2bJFRUVFysvLkyR9+tOf1jvvvKOmpiYlEgk5nU6FQiH5/X5Jx1daa2trFQgElEgk1NTUJK/X2zZ+wsfv8/Hx2tpa+Xy+Lj8BAAAAAED6OuU5tcFgULt27VJra6tM09SWLVs0aNAgjRkzRuvXr5ckrV69WmVlZZKkcePGafXq1ZKk9evXa8yYMTIMQ2VlZVq3bp1isZiqq6t1+PBhjRgxQsOHD9fhw4dVXV2teDyudevWtT0WAAAAAACf5JQrtSNHjtT48eN19913y+l06txzz1V5ebkuvvhiLV68WC+88IKGDh2qKVOmSJKmTJmiZcuWafbs2fJ6vZo7d64kafDgwbrssst01113yeFwaMaMGXI4jnfqW265RfPnz1cymdSVV16pwYMH9+BTBgAAAAD0FYZ5spNdbYDPqU1NqbYfHz2POU9PzHv6Yc7TE/Oefpjz9JRq896tn1MLAAAAAECqotQCAAAAAGyLUgsAAAAAsC1KLQAAAADAtii1AAAAAADbotQCAAAAAGyLUgsAAAAAsC1KLQAAAADAtii1QDepqzO0Y4dL0ajVSQAAAID04bI6AGB3pindd1+eKioydeyYof79k5o3r1HTprVYHQ0AAADo81ipBc7SypWZevHFbB086FIk4tTu3W498kieGhoMq6MBAAAAfR6lFjhLr76apZaW9t9K+/a59Oc/Z1iUCAAAAGnDNJV94BkFNl2vwKbrlX3gmeNbCdMI24+Bs5Sfn+wwlpWVVHFxx3EAAACgO3nfe0ze/T+VI9ksSXI3VskRDSk87NsWJ+s9rNQCZ+nOO8MaODDebmz06LgmTOCKUQAAAOhZmTWvtRVaSXIkm5VV+0cLE/U+VmqBszRkSELPPhvST36Sq2PHHBo+PK7vfa9BBqfUAgAAoIc5kh0vTmokmjs5su+i1ALd4MIL41qxos7qGAAAAEgz8axhcjXvbT+WPdyiNNZg+zEAAAAA2NSx8xepNf/TSrgCSrj8as2/VMfOf8zqWL2KlVoAAAAAsKmkJ6Da0pflbN4vSUpkDbY4Ue+j1AIAAACAzaVjmT2B7ccAAAAAANui1AIAAAAAbItSCwAAAACwLUotAAAAAMC2KLUAAAAAANui1AIAAAAAbItSCwAAAACwLUotAAAAAMC2KLUAAAAAANui1AIAAAAAbItSCwAAAACwLUotAAAAAMC2KLUAAAAAANui1AIAAAAAbItSCwAAAACwLUotAAAAAMC2KLUAAAAAANui1AIAAAAAbItSCwAAAACwLUotAAAAAMC2KLUAAAAAANui1AIAAAAAbItSCwAAAACwLUotAAAAAMC2KLUAAAAAANui1AIAAAAAbItSCwAAAACwLUotAAAAAMC2KLUAAAAAANtyneqAQ4cOadGiRW23q6urdcMNN2jSpElatGiRjh49qsLCQs2bN09er1emaWrFihXatGmTMjIyNGvWLA0bNkyStHr1ar388suSpGuvvVaTJ0+WJO3Zs0fLly9XNBpVaWmppk+fLsMweuDpAgAAAAD6klOu1A4YMECPPvqoHn30UT3yyCPyeDy69NJLtXLlSl100UVasmSJLrroIq1cuVKStGnTJh05ckRLlizRbbfdpqefflqSFA6H9dJLL2nBggVasGCBXnrpJYXDYUnSU089pZkzZ2rJkiU6cuSIqqqqevApAwAAAAD6itPafrxlyxYVFxersLBQlZWVmjRpkiRp0qRJqqyslCRt2LBBEydOlGEYGjVqlCKRiOrq6lRVVaWxY8fK6/XK6/Vq7NixqqqqUl1dnZqbmzVq1CgZhqGJEye2PRYAAAAAAJ/ktErt2rVrdfnll0uS6uvr5fP5JEk+n08NDQ2SpFAopGAw2HafQCCgUCikUCikQCDQNu73+zsdP3E8AAAAAACncspzak+Ix+N688039dWvfvUTjzNNs8PYyc6PNQyj0+NPpqKiQhUVFZKkhQsXtivPSB0ul4u5STPMeXpi3tMPc56emPf0w5ynJzvPe5dL7aZNmzR06FAVFBRIkvLz81VXVyefz6e6ujrl5eVJOr7SWlNT03a/2tpa+Xw++f1+bd++vW08FApp9OjRCgQCqq2tbXe83+/vNEN5ebnKy8vbbn/86yB1BINB5ibNMOfpiXlPP8x5emLe0w9znp5Sbd4HDBjQ5WO7vP3441uPJamsrExvvPGGJOmNN97QJZdc0ja+Zs0amaapnTt3Kjs7Wz6fTyUlJdq8ebPC4bDC4bA2b96skpIS+Xw+ZWVlaefOnTJNU2vWrFFZWVmXnwAAAAAAIH11aaW2tbVVb731lm677ba2sWnTpmnRokVatWqVgsGg7rrrLklSaWmpNm7cqDlz5sjj8WjWrFmSJK/Xq+uuu0733nuvJOn666+X1+uVJN1666164oknFI1GVVJSotLS0m59kgAAAACAvskwT+ek1hRz6NAhqyOgE6m2dQE9jzlPT8x7+mHO0xPznn6Y8/SUavPeI9uPAQAAAABINZRaAAAAAIBtUWoBAAAAALZFqQUAAAAA2BalFgAAAABgW5RaAAAAAIBtUWoBAAAAALZFqQVwSh984NDBg7xcAAAAIPW4rA4AIHU1NBiaOdOnt992K5GQhg2L68kn61RcnLQ6GgAAACCJlVoAn+Duu/O1Zk2mqqudqq11qrIyQ3PnFlgdCwAAAGhDqQVwUm+/7e4wtnevS7GYBWEAAACATlBqAZxURobZ6ZjTaUEYAAAAoBOUWgAndfXVzcrJ+ej8WY8nqcmTW+XglQMAAAApggtFATip2bMj8npN/dd/Zck0pUmTWjVnTtjqWAAAAEAbSi2ATzR9epOmT2+yOgYAAADQKTYRAgAAAABsi1ILAAAAALAtSi0AAAAAwLYotQAAAAAA26LUAgAAAABsi1ILAAAAALAtSi0AAAAAwLYotQAAAAAA26LUAgAAAABsi1ILAAAAALAtSi0AAAAAwLYotQAAAAAA26LUAgAAAABsi1ILAAAAALAtSi0AAAAAwLYotQAAAAAA26LUAgAAAABsi1ILAAAAALAtSi0AAAAAwLYotQAAAAAA26LUAgAAAABsi1ILAAAAALAtSi0AAAAAwLYotQAAAAAA26LUAgAAAABsi1ILAAAAALAtSi0AAAAAwLYotQAAAAAA26LUAgAAAABsi1ILAAAAALAtSi0AAAAAwLYotQAAAAAA26LUAgAAAABsi1ILAAAAALAtV1cOikQievLJJ7V//34ZhqHbb79dAwYM0KJFi3T06FEVFhZq3rx58nq9Mk1TK1as0KZNm5SRkaFZs2Zp2LBhkqTVq1fr5ZdfliRde+21mjx5siRpz549Wr58uaLRqEpLSzV9+nQZhtEzz7iP2rnTpZ/9LEfJpHTbbRFdcEHc6kgAAAAA0OO6tFK7YsUKlZSUaPHixXr00Uc1cOBArVy5UhdddJGWLFmiiy66SCtXrpQkbdq0SUeOHNGSJUt022236emnn5YkhcNhvfTSS1qwYIEWLFigl156SeFwWJL01FNPaebMmVqyZImOHDmiqqqqHnq6fdOrr2bqxhsDeuGFHP3mNzm66aaAfvvbLKtjAQAAAECPO2WpbWpq0o4dOzRlyhRJksvlUk5OjiorKzVp0iRJ0qRJk1RZWSlJ2rBhgyZOnCjDMDRq1ChFIhHV1dWpqqpKY8eOldfrldfr1dixY1VVVaW6ujo1Nzdr1KhRMgxDEydObHssdM3Pfpaj6mpn2+2jR516+ukcCxMBAAAAQO845fbj6upq5eXl6YknntD777+vYcOG6etf/7rq6+vl8/kkST6fTw0NDZKkUCikYDDYdv9AIKBQKKRQKKRAINA27vf7Ox0/cTy6LhRydhirq+N0aQAAAAB93ylLbSKR0N69e3XLLbdo5MiRWrFiRdtW486Yptlh7GTnxxqG0enxJ1NRUaGKigpJ0sKFC9uV53Q2aJBD773Xccyqvx+Xy8XcpJnenvO335aWLnUqK0uaNy+hgQN77UvjY/heTz/MeXpi3tMPc56e7Dzvpyy1gUBAgUBAI0eOlCSNHz9eK1euVH5+vurq6uTz+VRXV6e8vLy242tqatruX1tbK5/PJ7/fr+3bt7eNh0IhjR49WoFAQLW1te2O9/v9nWYpLy9XeXl52+2Pf510du+9Ls2e7deePcdXbM89N6G7765TTU3MkjzBYJC5STO9Oecvvpilhx/O09Gjx/+9v/yyqccfP6bLLov2ytfHR/heTz/MeXpi3tMPc56eUm3eBwwY0OVjT7lHtaCgQIFAQIcOHZIkbdmyRYMGDVJZWZneeOMNSdIbb7yhSy65RJJUVlamNWvWyDRN7dy5U9nZ2fL5fCopKdHmzZsVDocVDoe1efNmlZSUyOfzKSsrSzt37pRpmlqzZo3KysrO5HmnrZKSuF5//ageeaReDz9crz/+8aguucSaQgv0JNOUnnoqp63QStLBgy795Ce5FqYCAACAlbr0kT633HKLlixZong8rqKiIs2aNUumaWrRokVatWqVgsGg7rrrLklSaWmpNm7cqDlz5sjj8WjWrFmSJK/Xq+uuu0733nuvJOn666+X1+uVJN1666164oknFI1GVVJSotLS0p54rn1aTo6pm29usjoG0KOamw0dO9bxvbhQiHPIAQAA0pVhns5JrSnmxOoxUkuqbV1Az+utOTdN6XOfC2rrVk+78SuuaNWLL9ae5F7oKXyvpx/mPD0x7+mHOU9PqTbv3br9GABShWFI8+Y1auDA+IcjpoYOjet736u3NBcAAACs06XtxwCQKj73uVaNG1ejX/86W1lZSd10U7Nyc2274QQAepWz+YDydn1XrtZDSrryFT5nlloDU62OBQBnhVILwHYKC5O6886w1TEAwF6SMfm2fl2eyI62Iec7exX61H8onnOehcEA4Oyw/RgAACANZNZWyB3Z3W7MFf1A3n1PWJQIALoHpRYAACAdJJokdfzIPyPZ2vtZAKAbUWoBAADSQGvwc4pnDW03lnTmKdL/ZosSAUD3oNQCAACkAdOVo/qRCxX1jlXC00+x7FFqPOcORf1XWB0NAM4KF4oCAABIE1H/Farx/UFGvEGmM0dy8KsgAPvjlQwAACCdGIZMd77VKQCg21BqAQAA+ghX/ZvKf/eHcrZ+oKS7QJGBM9Tc/8tWxwKAHkWpBQAA6AOMeET+t++Uq3nv8YHW/XLueVAx7wWK515obTgA6EFcKAoAAKAPyKz+vZwnCu2HnLFaeQ88ZVEiAOgdlFoAAIC+wHBIMjoMm0bHsVRgJJqkZNzqGAD6AEotAABAH9BS9EXFs4a1G0u4CxUZNNOiRJ1zRnYpsHGaCv9vogorpyj33fmSaVodC4CNcU4tAABAH2A6s1Q3epnyd/9QztYjSrgLFBn0DcW9F1gd7SNmUr7ts+SJbG8bch58VonMc9Q08P+zMBgAO6PUAgAA9BHx3LGqLX2p2x/XiNUr64OVMl1Zai78ouTMPKPHcYe3fnQhqw85ki3KrPkDpRbAGaPUAgAAnC4zoYzav8jZclgtRVcr6QlanajHGEf+rMI3Z8nVsk+mHPLue0KhC59RInv4aT+W6ciQDLek5r/7Is7uCQsgLXFOLQAAwGkwYscU3HiNfNtmqmD3dxV88/PKOvRrq2P1DNOUc+v35WrZJ0kylJS7aZfyd33/jB4unj1KsZzz2o0lXfmKDPjaWUcFkL4otQAAAKch790H5WncLIcZlSS5Wg/Ju/9JKdFicbLuZyQiUusHHcadrUfO8AENhS56Vs3Bf1Qs53y15l2s+mH3qzV41VkmBZDO2H4MAABwGv7+nFBJcrYelqtpt+K5F1qQqOeYzmzJXSA1H2w3nnQXnPljuv2qu5DPzrUbI1YvOdzH/00AKYaVWqQc05Q++MCh5uZTHwsAQG9Lun2djPmVyBxoQZoeZjiUHDpDCddHzznuKVbjObMtDIXe5Gg5osCm61VYOUWFlVNUsP0OKRm1OhbQDiu1SCnr13v0wAN5+uADp3JyTF11VYu+//0GpejnxgMA0lDDuXfLHd7edp5p0pGtluBVMjspu31BcuQdqneMUs6hZ2U6MhUePEuJ7KFWx0Iv8e24Qxn169tuO1sOKun2qWHkQxamAtqj1CJltLZK3/lOvt591y1JOnpU+vd/z9bo0TF9+css26JvC4cNffvb+dq+3S2Px9SUKa26555G3tABUlDCO0o1pS/L+/5SOWMhNRVNU2vws1bH6lGx/FIdy19qdQz0MiMW6rDd3lBSnoaNFiUCOkeptbnmZumhh/K0datb2dmmZs4Ma/Jke24JWb/eo7172/+TbG526NVXMym16PNmzfLpL3/56HMf333XpawsU3Pnhi1MBeBkkhn91TBqgdUxgJ5leGQa7g7DpkGFQGrhnFqb+8Y3/HruuRxt2JChNWsydeedPq1d67E61hnJzTWVmWl2GPfY8+kAXVZfb2j79va/ILS2OvSXv2RYlAgAAMl0eRXNK5Wpj7YNJR05ain8goWpgI54m8XG9u1z6q233NLHXmhqapz62c9ydPnl9lutLS2N6fzz49q48aMWW1iY0O23N1qYCuh5pnn8v47j7D0GUp2j9bByDj4n03CpaeB0JT1BqyMB3erYBY8r6SqQp7FKpsOtluA1igy+zepYQDuUWhs7dsyhSKTjL70tLfb8RdgwpF/8olb33Zev995zKSfH1De/GdbFF8etjgb0qIICUyNHxnXkyEcvyW63qSuuaLUwFYBTyTj6R+Xvul+u6CFJUvYHv9Wx8xcpWnCZxcmAbuTIUMN5C61OAXwiSq2NjR4d05AhCb3zzke7yF0uU5ddZr9V2hP8flM//ekxq2MAve6nP63T3Lk+7drlkstlauLEVn3nO+xSAFJZ7r7H2wqtJLla9it3749UW/qKhakAIP1Qam3M5ZIWLqzXfffl6eBBl3Jykho/Pqo5c7iwDGA3Pp+pX/wi1LYNmaseAynOTMoRre0w3NlYn5NoUe77i+Ru3CzTla+Gc7+lRM5Iq1MBSGOUWpu79NKo/vjHGu3d65TPl5Tf38mJeQBsgzIL2IThUNITkFoPtBtOevwWBeoejmhIRiKiROagk74g+bfOUEbd6rYrergbq1Q79gU+uxaAZbj6cR/gcEjDhycotAAA9KLGc2YrntG/7XY8Y5Aah3zbwkRnIdkq39ZvKLjhKgXf/EcFN35BzsjuDoe5GrfL07BRH6+7rpYDyn1vUe9lBYC/w0otAADAGWgt/Lxqcscq58CzksOlyMAZSmYUWR3rjOS9+5Aya16ToeNvkDsbQ/K9fadqLn613XHO1gMyEg0d7u+I1/VKTgDoDKUWAADgDCUzB6pxxPesjnHWPA2b2grtCc6WA3K0HpZU2DYWLbhMicwhcrW83zZmyqUW3xW9FRUAOmD7MQAAQJozHRmdjHlkOnPaj7ly1ThkrmKZQ2XKoYQ7qObCz6lp0K29FRUAOmClFgAAoIcYsTrl7/yu3E27lXRmqan/P6u5/41Wx+qgqf9Ncod3yJGolySZciiaVybTnd/h2Ob+N6il6Bq5G99SInPg8YtKAYCFKLUAAAA9xL/1FmXU/63ttqvpXSXdfrUGP2Nhqo6ai6+X5FD2kf+QklFF88apcdi9Jz3edGYpWvDp3gsIAJ+AUgsAANADXJFdcoffaTfmjB9TzqFfplyplaTm4mvVXHyt1TEA4LRxTi0AAEAPMJItkhnr+AdmvPfDAEAfRqkFAADoATHvGMWzzm03lnRkqiV4lTWBAKCPotQCAAD0BMOhuguWqjVvnOKeYsWyhqlpwNfUNODrVicDgD6Fc2oBAAB6SMJ7vmov/r2MeINMR5bkcFsdCQD6HEotAABADzNdeVZHAIA+i1ILAAAApItEi/Le/aE8jVUyHRlq6neDmgfcZHUq4KxQagEAAIA04dt+hzJrX5fx4W1X5G3JcKi5/42W5gLOBheKAgAAANKAIxqSu3FzW6GVJGe8QdlH/tOyTEB3oNQCAAAAacBINsswox3/oLMxwEYotQAAAEAaSGQMUDzznHZjppyK5pVZlAjoHpRaAAAAIB0Yho5dsEStueOUcBcqnjFIzUXXqHHYvVYnA84KF4oCAAAA0kQie5hqx/1ejtZqmc4sma5cqyMBZ61LpfaOO+5QZmamHA6HnE6nFi5cqHA4rEWLFuno0aMqLCzUvHnz5PV6ZZqmVqxYoU2bNikjI0OzZs3SsGHDJEmrV6/Wyy+/LEm69tprNXnyZEnSnj17tHz5ckWjUZWWlmr69OkyDONkcQAAAACchWRGkdURgG7T5ZXaBx54QHl5H31w+MqVK3XRRRdp2rRpWrlypVauXKmbb75ZmzZt0pEjR7RkyRLt2rVLTz/9tBYsWKBwOKyXXnpJCxculCTdc889Kisrk9fr1VNPPaWZM2dq5MiRevjhh1VVVaXS0tLuf7YAAAAAgD7ljM+prays1KRJkyRJkyZNUmVlpSRpw4YNmjhxogzD0KhRoxSJRFRXV6eqqiqNHTtWXq9XXq9XY8eOVVVVlerq6tTc3KxRo0bJMAxNnDix7bEAAEAaME2rEwAAbKzLK7Xz58+XJH3mM59ReXm56uvr5fP5JEk+n08NDQ2SpFAopGAw2Ha/QCCgUCikUCikQCDQNu73+zsdP3E8AADo2zyhtcp7b6EcrUeV9PgVHjxLLUVfsDoWAMBmulRqH3zwQfn9ftXX1+uhhx7SgAEDTnqs2cm7rSc7P9YwjE6PP5mKigpVVFRIkhYuXNiuPCN1uFwu5ibNMOfpiXlPP9065621cv/tOzKa3vvw9n4V7Pk3xQdfIeWO6J6vgW7B93rPM977dzn3PiPFm2TmjlLi4mWSp8CyPMx5erLzvHep1Pr9fklSfn6+LrnkEu3evVv5+fmqq6uTz+dTXV1d2/m2gUBANTU1bfetra2Vz+eT3+/X9u3b28ZDoZBGjx6tQCCg2tradsef+Hp/r7y8XOXl5W23P/51kDqCwSBzk2aY8/TEvKef7pzznPeXy3Oi0H7I0XJY0S2PqGHUw93yNdA9+F7vWRlHX1fBO9+SI37s+ED9ZsUbD6i29CXLMjHn6SnV5v2TFlL/3inPqW1paVFzc3Pb/7/11ls655xzVFZWpjfeeEOS9MYbb+iSSy6RJJWVlWnNmjUyTVM7d+5Udna2fD6fSkpKtHnzZoXDYYXDYW3evFklJSXy+XzKysrSzp07ZZqm1qxZo7IyPgAaAIA+zXCe5A/O+HIfgC3lHPp3OU8U2g+5IjvkbNrTtQcwTSkZ64FkgH2ccqW2vr5eP/7xjyVJiURCV1xxhUpKSjR8+HAtWrRIq1atUjAY1F133SVJKi0t1caNGzVnzhx5PB7NmjVLkuT1enXdddfp3nuPf7jz9ddfL6/XK0m69dZb9cQTTygajaqkpIQrHwMA0Mc19f+Kcg79Uq6W99vG4p7+igz+poWpAAuYHQupkYzKSDSd8q7e9x5XVvVKGYmIEpkDdWzkI0p4R/VESiClGebpnNSaYg4dOmR1BHQi1bYuoOcx5+mJeU8/3T3n7mOVytu7QM4PLxTVeM5stQY/022Pj+7B93rPytn/lHL3LJDDjLaNRXNGq6bsj5Jx8p0LmR/8TvnvfFvOZORj97vgw/udbCdE1zDn6SnV5v10th93+erHAAAA3SlWcIlqS1+xOgZgqcigW+Vs2afM0F+lRPPxFddRP/rEQitJ2R/8tl2hlSRX02656zcqVnBJT0YGUg6lFgAAALCKYahh5INqSLbKSLTIdOd36W5mZ6XXcEoGv94j/XA1BgAAYB+mKSXjVqcAup8jo8uFVpKaBnxdCVf7j/2JZ49SLK+ku5PN5xCRAAAgAElEQVQBKY+3cgAAgC3k7P+5so78Ro54oxIZxaof+aDiuWOtjgVYojUwWQ3D7lPO4V/LSISVyDxHx877kWQYVkcDeh2lFgAApDxP6A3lvrdYjkS9JMnVekC+HXN0tOxPksNjcTrAGs0DvqrmAV+1OgZgObYfAwCAlJdz6N/bCu0JrqY9ygj91aJEAIBUQakFAACpr7OPKDEcMtl0BgBpj1ILAABSXnjQrUq4/e3GYtkjFPVPsigRACBV8PYmAABIebH8S9Qw/N+UfXDF8QtFZQ5U/cgFkoNfZQAg3fGTAAAA2EJz8XVqLr7O6hgAgBTD9mMAAAAAgG1RagEAAAAAtkWpBQAAAADYFqUWAAAAAGBblFoAAAAAgG1RagEAAAAAtkWpBQAAAADYFqUWAAAAAGBbLqsDAACA1OVs3q/8d74tV8t+OTILlBOYpsjgb1gdCwCANpRaAADQOTMp37Zb5QlvPX67Rcpt3KN41mC1Bj9nbTYAAD7E9mMAANApd+NmuZp2txtzJBqUfeg/LEoEAEBHlFoAANA5MynJ7OwPejsJAAAnRakFAACdiuWVKp49ot1Y0pmrpuKvWJQIAICOKLUAAKBzhkN1Y55SS8EVimcOUTL/IjWeO0+tRf9odTIAANpwoSgAAHBSiawhCpW8KEkKBoOK1NRYnAgAgPZYqQUAAAAA2BalFgAAAABgW5RaAAAAAIBtUWoBAAAAALZFqQUAAAAA2BalFgAAAABgW5RaAAAAAIBtUWoBAAAAALZFqQUAAAAA2BalFgAAAABgWy6rAwAAgJNIRpVZ8ycZibBaCq+W6cq1OhEAACmHUgsAQApyNu+Xb+t0uSO7JMWVeH+Jjo1coGhgstXRAABIKWw/BnBSiYRkmlanANJT3q775YnskKG4DEmulveVt/dhvikBAPg7lFoAHRw65NCNN/p1+eVFuvLKQs2fn8vv0UAvc7Ue6jDmjB6VET9mQRoAAFIX248BtGOa0syZPm3cmNE2tn+/U/36JXXrrRELkwHpJenK72QsT6aT82oBAPg4VmoBtLN3r1N79rR/v6ulxaE//jHTokRAemocMltxT3Hb7aQzX039rpccvB8NAMDH8ZMRQDsul+To5O0uw2D/MdCbov5Jqv3Ui/LuXy4j0aLIgJsV811udSwAAFIOpRZAO+eck9CoUXGtX+9sG8vJSeif/qnZwlRAekrkjFD9+YusjgEAQEqj1ALo4KmnQvrWtwr07rsuZWaa+tKXmnXTTZRaAAAApB5KLYAO/H5Tzz5bZ3UMAAAA4JS4UBQAAAAAwLYotQAAAAAA26LUAgAAAABsi1ILAAAAALCtLl8oKplM6p577pHf79c999yj6upqLV68WOFwWEOHDtXs2bPlcrkUi8W0bNky7dmzR7m5uZo7d66KiookSa+88opWrVolh8Oh6dOnq6SkRJJUVVWlFStWKJlMaurUqZo2bVrPPFsAAAAAQJ/S5ZXaP/zhDxo4cGDb7V/96le6+uqrtWTJEuXk5GjVqlWSpFWrViknJ0dLly7V1Vdfreeff16SdODAAa1bt06PPfaY7rvvPj3zzDNKJpNKJpN65pln9N3vfleLFi3S2rVrdeDAgW5+mgAAIGUl43K0HJSSUauTAABsqEultra2Vhs3btTUqVMlSaZpatu2bRo/frwkafLkyaqsrJQkbdiwQZMnT5YkjR8/Xlu3bpVpmqqsrNSECRPkdrtVVFSk4uJi7d69W7t371ZxcbH69esnl8ulCRMmtD0WAADo27IO/0aFGz6jwg2fV2FluXL2/dTqSNYzTWV+8Ip8W25V/tv/KmfT+1YnAoCU1qXtx88995xuvvlmNTc3S5IaGxuVnZ0tp9MpSfL7/QqFQpKkUCikQCAgSXI6ncrOzlZjY6NCoZBGjhzZ9pgfv8+J40/8/65du7rhqQEAgFTmaD2i3L0/kit6WJLkjNfKu2+5WgsuVzxvrMXprJO36z5lH3lRjmSLJCmj7n8UuvBZxXMvtDgZAKSmU5baN998U/n5+Ro2bJi2bdt2ygc0TbPDmGEYnY5/0vGdqaioUEVFhSRp4cKFCgaDp8yD3udyuZibNMOcpyfmPf1095w7tv+srdCe4IzXKRB6UYlhU7rt69hKNCTXsb+2FVpJcrUeVPDQUsUv/60lkfheTz/MeXqy87yfstS+88472rBhgzZt2qRoNKrm5mY999xzampqUiKRkNPpVCgUkt/vl3R8pbW2tlaBQECJREJNTU3yer1t4yd8/D4fH6+trZXP5+s0S3l5ucrLy9tu19TUnNmzRo8KBoPMTZphztNTb817Milt3+6S12vq3HMTPf71cHLdPefZsUzly5Ch9m9wN8Uz1Jimrymu8A4FWmo7jMci1aq16O+E1/j0w5ynp1Sb9wEDBnT52FOeU/vVr35VTz75pJYvX665c+fqwgsv1Jw5czRmzBitX79ekrR69WqVlZVJksaNG6fVq1dLktavX68xY8bIMAyVlZVp3bp1isViqq6u1uHDhzVixAgNHz5chw8fVnV1teLxuNatW9f2WACA9LZtm0uf/WxQ//RPQX3hC0HdeKNf4XDnu3lgP83FNyiePaLdWDzzHEXOud2iRNaLZw9XMrN/p+MAgM51+SN9/t4///M/a/HixXrhhRc0dOhQTZlyfJvQlClTtGzZMs2ePVter1dz586VJA0ePFiXXXaZ7rrrLjkcDs2YMUMOx/FOfcstt2j+/PlKJpO68sorNXjw4G54agAAOzNN6dvfLtD27R5JUlOT9L//69T99+dp8eJ6i9OhO5jOLIUu+oXydv+bnK0fKOn2q2HYPUp6Cq2OZh2HR41D5ipv7yNytexX0pGluHe0GoZ/z+pkAJCyDPNkJ7vawKFDh6yOgE6k2tYF9DzmPD319LxXVzt01VWFOnrU2W587NioXnuNf29W4Hu99xixemXW/lmJjH6KFlwhneR6I72BeU8/zHl6SrV5P53tx2e8UgsAQE/yek1lZXV837WzMaCvMd35ai6+3uoYAGALXfqcWgAAelt2tql/+IdWeTwflVifL6F/+ZeIhakAAECqYaUWAJCyFi6s17nnxrV6dYY8HunrX4+ovLzV6lgAACCFUGoBACnL4ZBmzYpo1ixWZwEAQOfYfgwAAAAAsC1KLQAAAADAtii1AAAAAADbotQCAAAAAGyLUgsA3SQSMbRvn1PJpNVJ+r5EQtq3z6mmJsPqKAAAwGKUWgA4S6Ypfe97eSovL9Q//mNQV11VqDVrPFbH6rNefz1DV111/O/6M58Jav78XKsjAQAAC1FqAeAs/eY3Wfr1r7O1b59LdXVO7djh1n335au52epkfU9jo6Ef/CBfb7/tVl2dU++959Zzz+XotdcyrI6GU0lGlXX4ReW+O1+u8DtWpwEA9CGUWgA4S3/4Q6ZaWtq/nL73nkvr1lG0utuf/pShffuc7caamhx6+eVsixKhK4xYvYIbv6SCd76t3P1PKFB1vbx7fmR1LABAH0GpBYCzlJXVcSwz05TPx8m13c3vTyojw+wwnpPD33Uqy937I3nCb8lQQpLkjIeU/cF/yhENWZwMANAXUGoB4CzdcUej+vWLtxsbPTqm0tKYRYn6rokTozrvvPZ/1/37JzR7dtiiROgKV/OeDmPO1kNyRd62IA0AoK9xWR0AAOzuooviWrr0mJYu9SocdmjYsLgefLBeBhfm7XZOp/SrX4X0ve/l6f33XcrPT2revEYNH56wOho+QSJjYMcxT7Hi2SMsSAMA6GsotQDQDS6/PKrLL2crZW8IBJJ64oljVsfAaWgcdo88DZvkanpbhqSk06uW4OeVzCiyOhoAoA+g1AIA8AmOHnVo2TKvqqsdmjatWVdd1coq/GlKeoKqufh3yjn4rJzNe9XU73rFfJdbHQsA0EdQagEAOIm9e526+eaA3nvv+I/LiopM3XBDk+bPb7A4mf2YLq/CQ+ZYHQMA0AdxoSgAAE7ikUfy2gqtdPzjg/70p0zV1vLj0+5c9Rvk3/wVBTd8Xr6tt8jRctDqSACAM8RKLQAAJ9FZea2pcWjfPqcCAT5GyK6cLQfk3z5LrtYPi2z4Lbma39fRca9JDo+14QAAp423mgEAltqzx6lvfStfd9xRoL/9zW11nHaGDo13GBswINnhY4VgL973l31UaD/kiuxSZvV/W5QIAHA2WKkFAFjmf/7Ho3nzCnT48PEfR3/9a6buvLNRM2dGLE523H33NWjbNre2bXMpFnOouDiuGTPCys42rY6Gs2AkOn6usaGEHPE6C9IAAM4WpRYAYJnHH89tK7SSVF/v0AsvZGvGjIhcKfATKj/f1O9/X6M//zlD77/v1Je+1KLiYrYd212k/z8rs3aVHIn6trF4xiA195tmYSoAPc2IR+R9f5HckXcUzxyk8LnfVtLjtzoWukEK/MoAAEhXx451PAumvt6hcNhQQUFqrIY6ndLnPtd6Wvf5v//zaOlSryIRQxdcENP99zeyuptCYr7LFB48U1kfvCxHvE4JT5Eah8yV6eaXW6DPMhPyv3WTMhrebBvKOPY31Vy8UqYr18Jg6A6UWgCAZfr3T2jHjvbn0RYWJpSfb98CuHmzS7ffXqAPPjj+I/Zvf8vQu++69OKLIYuT4ePC596p8DnflCMWUtJTJBlOqyMB6EGZR/8gd+OWdmPupreVc+Aphc+9y6JU6C5cKAoAcFZqax16+22X4mdw7aSHHqrXeefF5HQeL7HnnBPX3Xc3yDC6OWQvWrYst63QnrBli0c7d/I+cspxZCiZ0Z9CC6QBV+QdOcxox/GmvRakQXfjJywA4Iwkk9KddxZo7VqPGhsdGjAgofvvb9DUqV3fqjtkSEKvvXZUr72WqXDYoS9+sVl5efZdpZWkpqaOjTwcNlRT49CoURYEAgCouWiacg79Qs7YR7tmko4sNRd90cJU6C6s1AIAzsjSpQ797ndZOnzYpXDYoZ073frBD/LUenqnnyojQ5o2rUU339xk+0IrSRMntsrtbv88zj03rnHjOq4QAAB6RyJnhJr636y4p//x2+6gmou+pNZAucXJ0B1YqQUAnJGKCodisfarku++69I11wTlcknDhsX10EP1KXPBp97yjW9EtGOHS2vXZrStYH//+/XKyLA6GQCkt8Zhdysy6Ba5GzYp7h2tROYgqyOhm1BqAQBnxOvtWFYNQ9q2zSNJ2rzZowMHnHrllVpbnyN7uhwOafHieoVCDh096tCIEXE5OWUTAFJC0lOo1uBVVsdAN2P7MQDgjHznO0n169f+6lCm2b69btvm1pYt7a9unC78/qTOO49CCwBAT6PUAgDOSGmpqZ/9rE5Tpzbr059u1aBBHS9/3NpqqKEhjZZpAQBAr6PUAgDO2CWXxPTLX9bp5ZdrNXduozIzk+3+fOjQuC69lAskAQCAnsM5tQCAbvGVrzTr7bfdqqjIUDjsUHFxQj/4QYM8HquTAQCAvoxSCwDoFoYh/eAHDbrnHkPHjhkqLk6m1QWiAACANSi1AIBulZVlKisrvT7GBwAAWIdSCwAA0pqjtVq5e38kZ+thxXLOU/jcf5XpyrE6FgCgiyi1AAAgbRmxOgU23yB30y5JUmbdanka3lRt6SuSwfU0AcAOeLUGAABpy7v/ybZCe4K7cYsyav5sUSIAwOmi1AIAgLTlbN7fYcxhtsr1d0UXAJC6KLUAACBttRR9QUlHVruxhLtQLUVfsigRAOB0UWoBAEDaagl+Xs1FX1TCXShJimcMVGTgdCWyBlucDADQVVwoCgB6SChkaMsWt0aMiGvgwKTVcQB0xjBUf/5jCrcclLNpl2K5Y2W6/VanAgCcBkotAPSAxx/36vnns3XwoEuFhQlNndqiH/+4XoZhdTIAnUlkDlQic6DVMQAAZ4DtxwDQzfbtc+rZZ3N08ODx9w2PHnXqd7/L0uuvZ1icDLaUaJERj5zZfU1TuXsWqrCyXIX/9w/ybblFRuxY9+YDAMBilFoA6Ga/+12Wamqc7caamx169dWsk9wD6EQyqoLt/09FlVeqsHKy/Ju/Ike09rQeImffcuUceEruyA65m/coq/aP8m2f2UOBAQCwBqUWALrZyJExeTwdz6EdPDhhQRrYVd7uf1NW9Uq5WvbJ1XpImXX/o4Id/++0HiOztkKOZEu7MXdklxzRo90ZFQAAS1FqAaCbfeYzrRo7NtZubOTImL75zbBFiWBHnoaNMmS2G3NFdsmIn86/I7OLYwAA2NcpLxQVjUb1wAMPKB6PK5FIaPz48brhhhtUXV2txYsXKxwOa+jQoZo9e7ZcLpdisZiWLVumPXv2KDc3V3PnzlVRUZEk6ZVXXtGqVavkcDg0ffp0lZSUSJKqqqq0YsUKJZNJTZ06VdOmTevZZw0APcjplH7965B+8hOv3nnHrf79E/rOdxpVUECZwGkwnB3HHC6ZnY2fRKt/ityNW+QwW9vGYtkjlfQUdkdCAABSwilLrdvt1gMPPKDMzEzF43F9//vfV0lJiV599VVdffXVuvzyy/Xzn/9cq1at0lVXXaVVq1YpJydHS5cu1dq1a/X8889r3rx5OnDggNatW6fHHntMdXV1evDBB/X4449Lkp555hndf//9CgQCuvfee1VWVqZBgwb1+JMHgJ6Sk2Pq+99vtDoGbKw5+Fm5Iu/IkWyWJJkyFM29WHJ2/dzs8JA5MhKNyqz9i4xkVPGcEao7f1FPRQYAwBKn3H5sGIYyMzMlSYlEQolEQoZhaNu2bRo/frwkafLkyaqsrJQkbdiwQZMnT5YkjR8/Xlu3bpVpmqqsrNSECRPkdrtVVFSk4uJi7d69W7t371ZxcbH69esnl8ulCRMmtD0WAADpKnLObDUOma1obomi3jFq6n+zjl1wmoXUMNQ4/H4dvfSvqh6/VqGLfsFnsAIA+pwufU5tMpnU3XffrSNHjuizn/2s+vXrp+zsbDmdx7dA+f1+hUIhSVIoFFIgEJAkOZ1OZWdnq7GxUaFQSCNHjmx7zI/f58TxJ/5/165d3fPsAACwK8NQZMidigy50+okAACktC6VWofDoUcffVSRSEQ//vGPdfDgwZMea5odzxkzDKPT8U86vjMVFRWqqKiQJC1cuFDBYLAr8dHLXC4Xc5NmmPP0xLynH+Y8PTHv6Yc5T092nvculdoTcnJyNHr0aO3atUtNTU1KJBJyOp0KhULy+49vZwoEAqqtrVUgEFAikVBTU5O8Xm/b+Akfv8/Hx2tra+Xz+Tr9+uXl5SovL2+7XVNTczrx0UuCwSBzk2aY8/TEvKcf5jw9Me/phzlPT6k27wMGDOjysac8p7ahoUGRSETS8Sshb9myRQMHDtSYMWO0fv16SdLq1atVVlYmSRo3bpxWr14tSVq/fr3GjBkjwzBUVlamdevWKRaLqbq6WocPH9aIESM0fPhwHT58WNXV1YrH41q3bl3bYwEAAAAA8ElOuVJbV1en5cuXK5lMyjRNXXbZZRo3bpwGDRqkxYsX64UXXtDQoUM1ZcoUSdKUKVO0bNkyzZ49W16vV3PnzpUkDR48WJdddpnuuusuORwOzZgxQw7H8U59yy23aP78+Uomk7ryyis1ePDgHnzKPW/fPqfuvTdfBw86lZeX1C23RDRtWovVsQAAAACgzzHMk53sagOHDh2yOkIH8bj0+c8HtX27p20sEEjo2WdDKiuLWZis96Ta1gX0POY8PTHv6Yc5T0/Me/phztNTqs17t24/xun53//1aNeu9gvgtbVOPfWU16JEAAAAANB3UWq7WWuroWSy49Wb43ELwgAAAABAH0ep7WaTJ7dq6ND2DTYvL6mbb26yKBEAAAAA9F2U2m6WkSE9/nidxo1r1YABcZ13XkyzZzfqyitbrY4G9Fk7drj0ta/59IUvBHXrrT4dPMhLG4AUZpoy4o2SmbQ6CQD0Caf1ObXompKSuH7/+1q1tkoej2R03I0MoJvU1Dg0Y4ZP77/vbhvbs8el//7vo8rKsjAYAHQi84PfybtvuRyxWiXdfkUG3abm/l+2OhYA2BrLGT0oI4NCC/S0J5/MaVdoJWnXLpd++9tsixIBQOccLQeV9+6D8kS2yRU9Ik9ku/L2LpCz+X2rowGArVFqAdhabW3Hl7Fk0lB1dWq9vB054lBDA+9yAeks59Av5YoebjfmjFYr58AzFiUCgL4htX7rA4DT9LWvNamgINFurLg4rq98JTUuzvbuu05dc01An/1soaZOLdSsWQWKpcdHVgP4O6az8x0kpjOnl5MAQN9CqQVga6WlMc2cGdawYTEVFCQ0cmRM//qvjRowwPoLsJimNGeOTxs3ZqimxqlDh1z6r//K0sMP51odDYAFIgO+pljmkHZj8cxzFBk0w6JEANA3cKEo9Jh9+5x6/fUMjRkT14QJUc4vRo+ZMyeib3wjoupqpwYMSMjtPvV9esOhQ07t3+9sN5ZMGtqwwWNRIgBWMt0+1Y35qfLfXSBHtEZJj1/1Q+9W0hO0OhoA2BqlFj1i4cJc/cd/ZKumxqmcnKQuvjiqX/4yJA+/y6OHZGVJQ4YkTn1gL8rKSsrjMTuM830ApK947qdUW/Ki1TEAoE9h+zG63f790osvHi+0khSJOLR2bYaefppzhpBe/H5TpaUxGcZHxTY/P6kbb0yN830BAAD6AlZq0e1WrTJUXX2yLZcRa0IBFlm2rE4//GGeNm/2yOMxdeONTfryl5utjgUAANBnUGrR7T71qeOrUfX17TcCpNrWUKA3ZGRI8+c3WB0DAACgz2L7MbpdSYmpyy5rldv90ZbL88+Pac6cRgtTAYBNmaay9/9c/s03ybflX+SpW2t1IgAAUgortWlsxw6XnnzSq1hM+vrXI7r00u778Myf/7xOv/lNlv7610wNHRrXrFlh5ed3vGAOAOCT5e+8W1lH/lMOMypJ8jRU6dioR9Ra+DmLkwEAkBootWnq9dczdO+9BW3nvr7xRobuuiusGTO655xXp1O66aZm3XQT5w4CwJky4mFl1K1pK7SS5IzVKOfgM5Ra4Cxk1P5FOQefk5IxtQSmqmnQ/8/enYc3VaV/AP/eJXu6JE3LviNqGbAg7gqO8hsVUXEZZ9x3cVARcVdGZ1wQd1wAFRUHHB0REUTGDVnd2QWRpex726RtkmbPvb8/MrTElK1NcpPm+3ke/8jx5ualp7m97z3nvOdmcO9BouzF6cc56vXXrXHFnGpqJLz/vgkqB1OJiDKGEKmFEE2sli020kZEh8e49xMU/nYXjK65MNYsQv6m0cjfMErrsIioGThSm6N+X8QJANxuEcEgYDRqEBAREdWTa5ehYOPjkIJ7IEQTZ9CELT00iIqoZbDsnAwpUl3/WlRDMLrmwx0NABJvgoiyEZPaHNWmTRTr1+vi2kpKFCa0REQaEyJe2NcOh+zfXN+mQgIEEapoRijvD3B3f0LDCImym6AkznQQFD8ExQeVSS1RVuL04xz1xBO16NEjDElSIQgqOneO4NFHue0IEZHWTBUzIe2X0AKAgCgCtv6o6vspXGVTocpWjaIjyn4Rc/eEtqihLVSdXYNoiCgZOFKbo7p1i+KLLyrx5ZdGBAICBg8OwGzmgloiot8rL5fw5JP5qKqSUFwcxaOPutGlSyr33Rb+91/8NVnROxCxJN6ME2Uqy/aJMFVMhxANIGLuipqjn8uIxLG2x9OQgnuhq1sDKCFETF1Qc/SzWodFRM3ApDaHGQzAhRcGtA6DiChj1dQIuOEGOzZtaliusWmTjNmzq2C1puZBoL/kIli3T4Ds31TfFtUXo6790JR8HlEqmPZ8DOuW5yFFvQAAnW89xHA1nH2maxwZoMr5cPaZBslXDjHqQ9j6B0Dg5EWibMZvMBER0QFMmmSJS2gBoLxcxpQp5pR9pipb4Dr2VQQLTkLE2BnBvD6o7fZPRKxHp+wziZLNtPej+oR2H7luHSTf5gO8I/2i5u4I5/VmQkvUAnCkloiI6AAqKxu72RVQUZHam+BIfllGjGgRNZmqJDQJqgJBjWgQDBG1dHw0RUSHJRwGli3TYfNm6dAHE7UQ113nQ1FR/PrZ4uIorr2W+8QSHUzAcQ4UMb6ScMTctdEiTamkd32LwjW3I3/9IxADu9L62USUPhypJaJD+v57PR5+uABbt0qwWFQcd1wYb73lgsmkdWREqXX00RHccYcX771nhtMpwuFQcP31dSkuFEWU/XztboQU3ANj1VcQlACipo6oOfpFQBDSFoN183Ow7ngHYjS2u4PR+Q1cf3gTkbzeaYuBiNJDUFU1a0ve7trFJ26ZyOFwoKqqSuswKEmiUeBPfyrG2rX7rytUcf31dXjqqdiNQi71ud8vQFEAiyVrL51Jk0v9HgwCe/dKaN06Cr1e62i0k0t9Tg2a1e+qCqgRQNQd+tgkEqJ+FC8eCDmwJa7dbz8L1b2npDWWbMTvem7KtH5v27btYR/LkVoiOqhNm2Ts2vX7KccCfv01vTcoWvP7Bdx5ZyFWrdJBVYEePSIYN64aBQVMbnOBwQB07MjRWaIjJgiAkP6/F2JoL4RITUK7FK5OeyxElHpcU0tEB1VUpMBiSSz4kWv7Gj/wQAE+/9yEHTtk7NwpY948I0aMKNQ6LCIiakTU0B6KviSx3dhOg2iIKNWY1BLRQdntCk45JQRZbkhiS0qiuP1270He1fKsWpU40rB2rQ5RDt4REWUeUYa34x2IGGJJrAoZYUspars/rnFgRJQKnH5MRIc0dmwNevSI4Ntv9TCbVfztb16ceGJY67AaFQoBo0YVYOlSPQRBxWmnhfDoo25IzSzavH9Sv3+byEeDREQZyd/6UgTtZ8K4dzoUfTECxeenfW1v0ihBSIGdiBraApLx0McT5RgmtUR0SJIE3HmnF3feqXUkh3b//YWYNs0EVY1V2NywIXYD889/upt13jPPDKK8XEYoFMtiJUnFKT/ApRYAACAASURBVKeE0lnIk5po6VIdli7V4Y9/DOKoozi0TpRLFH0RfB1u0TqMZjHveAeWne9CClUhqnfA1+YK1HX8m9ZhEWUUJrVE1GIoCrBkib4+oQWASETAd98Zmn3uhx/2QKcD5s83QFWBk04KYdSo5iXKlFrRKHDLLTZ8/70BHo+IV1+NYvBgP55+mv1GRNlB8m1C3tZXIIUrAQCivxbW7eMRtP8REesxGkdHlDmY1BJRi6Ik1rRqtO1ICQJw//0e3H+/p/kno7SYPt2IefOMCIViDzlcLgmffmrCNdf4UFoa0Tg6IqJDM+96rz6h3UcKu2De9S+4ezytUVREmYerwYioxRBFoGfP36/1VdGnT0iTeEhb+ye0+9TUSPjqK65HI6LsoOhbobG9Bhqr7EyUyzhSS0QtytixsX0JV6/WQRSBvn1DGD26VuOoclNNjYAJE6zYtk3CeecFcMEFgbSuQS4tjWDmTBVAw4eaTArKyjKzyBkR0e/52l4F8573ofOV17eFTV1R1+5GDaMiyjxMaomoRbFYVEycWF2/1U5zqx5T01RVifjzn4uwfn2sUNeXXxoxZ04Ar7xSk7YYbryxDrNnG/HLLzoAAiRJRb9+IQwYEExbDEREzaHKVrh6TUH+xn9CClYgqi+Gu9soqLoCrUMjyihMaomoRWIyq63nnsurT2gBIBgUMX++AVu3SujUKT0ViM1mFdOmOfHGGxasW6fDSScFcfXVPlasJqKsEjV1RPUf3tY6DKKMxqSWiIiSbteuxKcKTqeEdevktCW1QGzkfuRIb9o+j4iIiNKPhaKIiCjpSksT1622aRNB375cz0pERETJxaSWiIiSbsQIL048MQiDIbafksMRxRVX+OBwJGF/JSIiIqL9cPoxERElncmk4uOPnZg714B162RccEEAHTumb9oxERER5Q4mtURElBKiCAwcGMTAgaw2THRAShjGilnQ+crha3UpopZuWkdERJR1mNQSJdGWLRKefTYPTqeI0tIw7rvPC7O5sW3TiYgo1wkRN4pW/hU6zxoICMO8+z3Utb0O3i73aB0aEbVkShgQZLSk7QCY1BIlye7dIq680o6tW2PbmHz7rRG//KLHtGnOlnTNICKiJMnb/Cz0npX1r6WwE+Y9H6Ku/Q1QdXYNIyOilshQNQd5W1+EGKqCorPB2+E2BFpdrHVYScFCUURJMnZsXn1Cu88vv+jw44+6A7yDiIhymezbmNAmBXdC512rQTRE1JKJoSoUbHgEes9KyMGd0HtXo6D8cUi+zVqHlhRMaomSpKoq8evk84nYsoUTIoiIKFHU0DahTdG3QsTMdbVElFzm3f+GHNwR1yaFK2DZ8aZGESUXk1qiJPnTnwLQ6eK3K2nbNoJzzmGRHCIiSuTp+iDC5h71r6OiBX7HuVAMrTSMiohaIlU4wMxBQZ/eQFLkkENIVVVVGDduHGpqaiAIAgYOHIhBgwbB6/XipZdeQmVlJYqLi3H33XfDarVCVVVMmjQJy5cvh8FgwLBhw9C1a1cAwPz58zF9+nQAwCWXXIIzzzwTALBp0yaMGzcOoVAIffr0wQ033ACBixApy1x+uR/ff6/HwoVGOJ0iOnSI4tZbPbDbuS8nERElUvTFqOo7E5Ydb0H2b4Gv1aUI2fprHRYRtUC+NlfCvOvf0AW21LdFDO3g7TBUu6CS6JBJrSRJuOaaa9C1a1f4/X48+OCD6N27N+bPn49evXphyJAhmDFjBmbMmIGrr74ay5cvx549e/DKK69gw4YNeOuttzB69Gh4vV5MmzYNY8aMAQA8+OCD6NevH6xWKyZOnIihQ4fiqKOOwtNPP40VK1agT58+Kf/HEyWTIAAvv1yLXbs82LFDRs+eYVgsrHycDq+9ZsFnn5kQCAjo3j2C556rgc3Gnz0RZT5Vzoe380itwyCiFk7VFaL62FdRsGk0xFAlFL0dnk53QzEmLoPIRoecfmyz2epHWk0mE9q1aweXy4XFixdjwIABAIABAwZg8eLFAIAlS5agf//+EAQBPXr0QF1dHaqrq7FixQr07t0bVqsVVqsVvXv3xooVK1BdXQ2/348ePXpAEAT079+//lxE2ahtWwUnnhjK+YTW4xEwdKgNAwYU4//+z4HnnrNCTcGPZMoUE155JQ+rVumxYYMOn39uwq232pL/QURERERZLFLQF84+01B50gI4+3yCkL3lzAw5ogo2FRUV2Lx5M7p3747a2lrYbLEbR5vNBrfbDQBwuVxwOBz17ykqKoLL5YLL5UJRUVF9u91ub7R93/FElN1uu82G+fON9a83b5ZhNqu4/fa6pH7Op5+aUVcX/3xu3Toddu6U0K5dNKmfRURERESZ57CT2kAggBdeeAHXX389zGbzAY9TGxmKOdD6WEEQGj3+QObMmYM5c+YAAMaMGROXPFPmkGWZfZNjft/nVVXA+vXxlxe/X8S8eXl47DFT0j/791RVRF6eDfw1TC1+13MP+zw3sd/TLLAH0m/PAIG9UNpfDLX9ZUj3hvfs89yUzf1+WEltJBLBCy+8gDPOOAMnnXQSAKCgoADV1dWw2Wyorq5Gfn4+gNhIa1VVVf17nU4nbDYb7HY71qxZU9/ucrlQWlqKoqIiOJ3OuOPt9sY3HB84cCAGDhxY/3r/z6HM4XA42Dc55vd9XlUlIhotTjguFAqjqsqZ0N4c/ftb8NNPeQgGG0Zru3QJIT/fCf4apha/67mHfZ6b2O/pI/m3omjllZD+V8xH2D0b/m2fo/boZ9MaB/s8N2Vav7dte/jrfQ+5plZVVbz++uto164dBg8eXN/er18/LFiwAACwYMECnHDCCfXtCxcuhKqqWL9+PcxmM2w2G8rKyrBy5Up4vV54vV6sXLkSZWVlsNlsMJlMWL9+PVRVxcKFC9GvX78j/TcTUQZxOBR06xaOa9PrVfTvH0r6Z912Wx1uuMGHo44Ko337CM44I4A33qhO+ucQERGlWt6mZyDvV51WVAIwOL+BGKzQLiiiLHDIkdp169Zh4cKF6NixI+677z4AwBVXXIEhQ4bgpZdewty5c+FwODByZKxyX58+fbBs2TIMHz4cer0ew4YNAwBYrVZceumleOihhwAAl112GaxWKwDg5ptvxvjx4xEKhVBWVsbKx0QtwBtv1OCuuwpRXi7/L6ENYuRIT9I/RxCAv//djVGjAEUBJCnpH0FERJQWUjhxlEwKVUEKbINiKNEgIqLsIKhHsqg1w+zatUvrEKgRmTZ1gVLvYH2uKLHEk1tPtzz8ruce9nluYr+nT/76R2Dd9W5cW8TYGZX9voIqW9IWB/s8N2Vavyd1+jERUXOIIhNaIiKiw+Hp+iCC+cdDEQ0AgIi+Dbztb05rQkuUjY5oSx8iIiIiIkoNVc6Ds88M6F3zIfu3IlB8PqcdEx0GJrVERERERJlCEBEqOgvJL61I1HIxqSUiaiZVBRYt0mPOHCNOOimEc88NsGAVERERUZowqSUiagZVBYYNK8ScOUb4fCLee0/BCSeE8d57Tuh0WkdHRERE1PKxUBQRUTMsXqzD3LmxhBYAgkERP/6ox9SpZo0jIyIiIsoNTGqJiJphwQIjvN74S2kkIuCHH/QaRURERESUW5jUEhE1wymnBGE2K3FtkqSiXz+W+CAiIiJKBya1RNRi+P0C3n3XjNGj87BpU3oqNZ12WghnnBGE0RhLbGU5ltBeeaUvLZ9PRETZS6rbCNPuqZB8W7QOhSirsVAUEbUIe/aIuOqqIqxfL0NRBHz4oRnDhnkxdGhdSj9XEIC33qrGF18YMWeOASecEMJll/lZJIqIiA5MVVGwbiSMVXMgRVyI6hzwFw+Cu8fTWkdGlJU4UktELcLjj+dj7VodFEUAAFRVSZg82QK/X0j5Z4siMGhQAC++WIsrrmBCS0REB6evXghTxWeQIi4AgBSugnnvDOhql2ocGVF2YlJLlCKKAnz6qRF33FGIcePSk1zlsl27EqcbV1YK2LKFG8YSEVFmMVV8ClGJX6YiRt0wVXyqUURE2Y3Tj4lS5G9/K8TXXxsRDIqYMUPFrFkmTJvmhNWqah1ai1RSoiS0ORwKOnSIpuXznU4R77xjhs8n4KabfGjfPj2fm06BAODzibDbE3/WRER0+MLWnlAhQkDD9VSFDqG8P2gYFVH24kgtUQqsWiVj0aJYQgsAqipg1So9xo+3ahxZyzVqlBvduoXrXxcWRnHZZf60PERYskTG4MEOjB2bjzffzMNFFxVh5kxjyj83XVQVeOyxfJx9dgkGDizG+ec7sGoVn4kSETWVr82VCFt7xbWF83shUDJEo4iIshvvSoiaKBoFXn7Ziu+/N0CvV3HjjXUYODAIAFi2TI/a2sRnRhs28CuXKh07RjFrVhXeesuC3bslXH11HcrKImn57DFjCrBtW0Pf7tkjY/x4Ky68MAChBcw6/9e/zJgyxVz/kGbvXgkjRtiwbBlnHRARNYlkhLNsKqxbX4XsW4+IpRTeTrcDIosyEDUF77CJmuieewrxyScmRCKxrGXVKh2efroGgwcHccYZQdjtUbhcDes5BUFFnz7cuzSVCgpU3HOPN+2f63QmPsBwOkUEAoDJlPZwku6rrxpmHeyzaZOEZcui6NJFo6CIiLKcKlvh6faQ1mEQtQicfkzUBG63gB9+0NcntADgckn4178sAICuXaO44AI/Cgtj6yqNRgUnnxzCTTeldnsZ0kZRUeIa06IiBcYWMgNZp0sckTUaAYtFg2CIiIiIfodJLVETeDxio9WM/f6Gr9To0W5MnerEiBEeTJhQjalTnTAY0hklpcuDD9aiY8eGqc6tW0cwbJg3K6YeRw+jntWNN9bBbo8/8Jhjwigt5fRjIiIi0h6nHxM1Qdu2UbRrF4XTGb9dTM+e4d+9jqBnT086QyMN9OsXwezZVXjnHTMCAQE33FCHdu0yu0LwzJlGTJhghcslwuFQMHKkp35N+O8NGBDCk0/W4t13LfD5BHTrFsHo0bUAitIbNBEREVEjBFVVs/ZR+65du7QOgRrhcDhQVVWldRgpt2qVjJEjC7FtmwyjUUWvXmFMnFgNkylrv1JNlit93lJs3Srh4ouLsHdvw3PNdu1iiXlx8eEn4+z33MM+z03s99zDPs9Nmdbvbdu2PexjOVJL1AQ7d4pYs0aHsWNrIIpAfr6Kdu1a3r6k1DJNnGiJS2gBYOdOGZMnmzUptEVERETUHExqiY7QU0/lYdo0MyoqJBQUKDj99ABef71G67CIDpssNz6bQK9PcyBEREREScBCUURHYNMmCR9+GEtoAaC2VsTXXxsxfXoLKXNLOWHo0Dq0axe/h2+nThFcfTWrcxMREVH2YVJLdAQ+/9yYUBwqFBLxzTdMail7tGmj4Pnna3DCCUF07x7GyScH8fLL1bDZcm89OBEREWU/Tj8mOgI9e4ZhMilxW/cAsX1pibJJ//4h9O/v1DoMIiIiombjSC3RERgwIIS+fcMQhIYRrWOPDWHoUBbXISIiIiLSAkdqiY6AIABTpjgxcaIVy5bp0LlzBMOHe5Gfn9nTNteskbFnj4STTgrBYsmsWKurBSxZokf37hF06cIRbyIiIiI6MkxqiY6QwQDccUd2jMz6/cCNN9qxYoUeHo+ATp2iuP9+Ny66KKB1aACA11+3YNIkC3bskGG3R3HaaUGMHx/bJomIiIiI6HDw1pGoBXvuuTwsXGiA2y1CVQVs2SLj+efzEMiAnLaiQsRbb8USWgBwuSR88YUJH3xg1jgyIiIiIsomTGqJWrDVq/UAhLi2HTskrFmj0yag/cyZY8Tu3fGTRcJhAfPmGTSKiIiIiOJEAyhYey8cS/4Ex9LzYdn6GqBm1jImIoDTj4latLw8JaGtsFBFmzbar13t1i0Ms1mBzxf/bK11a+1jo9y1caOEpUv1OOWUEDp04O8iEeU22293wlj13/rH43LdekCQUdfxNk3jIvo9jtQStWD33ONBu3aR+teyrOLUU4No0yYx2U23E08Mo2/fMICGJ75du0Zw113ZsV6ZWhZVBUaMKMCQIQ7cfbcNgwc7MGpUvtZhERFpRoh4oPOsjJvvJSo+GKv+q1lMRAfCkVqiFqy0NIL333fipZfy4HaLOOOMAG66yad1WABilaT/9S8nXn3VipUr9WjVKor77vOguFj7hJtyz9y5BsyebaqfOVBVJWH6dDP+/Gcfjjsucoh3ExG1QGoEgpp4/WusjUhrTGqJWrju3aMYN65G6zAaZTQC993HkVnS3qxZpoSp8LW1Ij791ITjjvNoFBURkXZUnQ0RU1dIob0NbZAQKjhRw6gOQVUAgRNRcxGTWiIiynmlpSEIggmq2jDRTq9X8Yc/hDWMiohIW9U9J6BwzR2Q/ZsAQYdQwYlwdx2ldVgJDM5vkLflRYihKig6OzydbkeweLDWYVEaMaklIqKcd801PkyfbsaqVfr6tt69Q7jgggzY/4qISCOKvhiusg8hROqgijIgZt4OBWKoEgXrH4Yc3BFrCO5AwYbH4LQeh6ipg7bBUdpwfJ6IiHKeyQR89JETt9/uwZ/+5Mfdd7vxn/+4IPPRL1FSiYFdMFR9BTG4R+tQ6AiosiUjE1oAMO+c3JDQ/o8c2gPLjokaRURa4J9rIiIiAHl5Kh5+mOtniVIlf/0omKr+Cym0F1F9K/iLL4D7qH9qHRZlO7HxdEYVpDQHQlriSC0RERERpZSu5ieY906rLzokhfbCvGcqZPcKjSOjbOdrcw0ixo5xbRFDO9R1uFWjiEgLTGqJiIiIKKXMe6dBjMbPhBCjbpj3TtcoImopFL0dNce8jGDBiQibuiKY3w81Rz8HxdBG69AojTj9mIiIiIhSKmw5GipECGjYi1yFjIilh4ZRUUsRKjwRzj6faB0GaYgjtURERESUUr42VyFsKY1rC1tL4Wv9Z40iIqKWhCO1RERERJRakgnOsg+Rt+V5yP4tiJi7wdP5noytqEtE2YVJLRHRIXi9AhYuNKB16yj69AlDELSOiIiaTVUhe3+FGKlBqOAEJldpoOoK4T7qSa3DIKIWiEktEdFBzJxpxDPP5GPrVgkWi4JevSKYPNkFi0XVOjQiaiIh4oZ91fXQeVdDiAYQMXdB7VFPI2Q7VevQiIioCbimlojoAIJB4MUX87B1qwxAQF2dhB9/NOCpp/K0Do2ImiG//FEYan+CGK2DgCh0vnLklz8GqMqh30xERBmHSS0RNVtFBfD003n4+9/zsXlzy9nsfMMGGbt2Jf571q3TaRANESWLrq48oU0K7oEU2KlBNERE1FycfkxEzbJ0qYzhw3XYskUPAPjsMxMeesiNyy/3axxZ87VuraCgQIHPF//8Lz+fozlE2UyRLAltqmyForNpEA0RETUXR2qJqFmeeaYAW7Y0VE6qqJAwcaIFSgvI+xwOBf37B6HXN/xj2rWL4J57PBpGRUTN5e04DFFdSf1rRTAgaP8jVNmqYVRERNRUhxypHT9+PJYtW4aCggK88MILAACv14uXXnoJlZWVKC4uxt133w2r1QpVVTFp0iQsX74cBoMBw4YNQ9euXQEA8+fPx/Tp0wEAl1xyCc4880wAwKZNmzBu3DiEQiH06dMHN9xwAwSWFiXKGk5n4rMxl0tEXZ2AvLzsL6b0wgu16Ns3jDlzDMjPV3HXXR506xbVOiwiaoaQfQCcvd5B3rbxEBQ/AkXnwNf2aq3DIiKiJjrkSO2ZZ56Jhx9+OK5txowZ6NWrF1555RX06tULM2bMAAAsX74ce/bswSuvvIJbb70Vb731FoBYEjxt2jSMHj0ao0ePxrRp0+D1egEAEydOxNChQ/HKK69gz549WLFiRbL/jUSUQg5HYoJXVKTAas3+hBYABAG4+mof3n23Gq+8UsOElqiFiOT3QfUfJsLV+z342l0D7tVFRJS9DpnUlpaWwmqNn46zePFiDBgwAAAwYMAALF68GACwZMkS9O/fH4IgoEePHqirq0N1dTVWrFiB3r17w2q1wmq1onfv3lixYgWqq6vh9/vRo0cPCIKA/v3715+LiLLDI4+40a1bQwLbpk0Ew4Z5eX9IRERERGnRpEJRtbW1sNlixRRsNhvcbjcAwOVyweFw1B9XVFQEl8sFl8uFoqKi+na73d5o+77jiSh79O4dwXffhfHSSwHU1Qm47ro6tG3bAhbUEhEREVFWSGr1Y1VNnG54oPWxgiA0evzBzJkzB3PmzAEAjBkzJi6BpswhyzL7JsfIsozHHzf+75VB01goffhdzz3s89zEfs897PPclM393qSktqCgANXV1bDZbKiurkZ+fj6A2EhrVVVV/XFOpxM2mw12ux1r1qypb3e5XCgtLUVRURGcTmfc8Xa7/YCfO3DgQAwcOLD+9f6fRZnD4XCwb3IM+zw3sd9zD/s8N7Hfcw/7PDdlWr+3bdv2sI9t0pY+/fr1w4IFCwAACxYswAknnFDfvnDhQqiqivXr18NsNsNms6GsrAwrV66E1+uF1+vFypUrUVZWBpvNBpPJhPXr10NVVSxcuBD9+vVrSkhEREQZz7j3UxQtG4LixQNhW30TxFCl1iERERFlvUOO1I4dOxZr1qyBx+PBbbfdhssvvxxDhgzBSy+9hLlz58LhcGDkyJEAgD59+mDZsmUYPnw49Ho9hg0bBgCwWq249NJL8dBDDwEALrvssvriUzfffDPGjx+PUCiEsrIy9OnTJ1X/ViIiIs3oan5EQfkoSOHYDCVd3W+QgrtR1Xc2K+8SERE1g6Ae6cLWDLJr1y6tQ6BGZNrUBUq9bO/zSAR49VUrfv5ZD6tVxfDhHvTqFdE6rIyX7f2ebrZVN8Pk/DyuTRGNcB43FeGC4zWK6siwz3MT+z33sM9zU6b1+5FMP05qoSgiomw0bJgNX3xhRDQaGy1bvlyHt95yoayMiS0lkRpOaBKUCISoT4NgiIiIWo4mraklouwRDgMPPZSPs88uxtlnF+Phh/MRYa5Wb+dOET/9pK9PaAFg924Zr76ap2FU1BL5W10MRbTEtUXMXREqPFmjiIiIiFoGjtQStXCPPFKADz4wQ1FiSduGDTKiUQHPPFOrcWSZoaJCgtuduJ7R4+EzP0quQKsh8Nath6lqNoRIHRRDa9Qc9SQg6rQOjYiIKKsxqSVq4X76SV+f0AJANCrgp5/0GkaUWUpLw+jYMYry8oYkVhBU9OkT0jAqyjY+n4Bx46xYs0bGUUdFcOedXuTlJZas8Ha9H97OIyEofqgyZwMQERElA5NaohausVJwipL+ODKVwQA89JAbo0fnY8sWGfn5Co4/Pox77vFoHRpliXAY+MtfirBsWexh0VdfAd9+a8D06VUwGht5gyhDFZnQEhERJQuTWqIWrnfvMDZu3H96o4rjjkssWJPLzj03iDPPrMQvv+hRXBxFly5RrUOiLPLJJyb88kv8FOJVq3R4/30zbryRRaCIiIhSjUktUQv33HO1iEaBlStjo0hlZSE8+2yNxlFlHqMROPFETjmmI7d6tQ6RSPy6bEUR8OuvXCtLRESUDkxqiVo4k0nFhAk19VOORdY/Ikqqiy7y46OPzHC7G75cZrOCCy/0axgVEbVIShDGqi8BVUHAcQ4gmbSOiCgjMKklyhFMZolS4/jjwxgyxI///teIqioJRUVRDBwYQP/+HPknouSRvGthX3MbZN8mACoipi6oPvYVRPLLtA6NSHNMaomIiJrp6adrMXSoF4sX69G3bwjdunFdNhElV2H536Hzbah/rfNvREH5P+DsO0PDqIgyA5NaIiKiJOjcOYrOnTnlmIhSQwrubrxNVQEhcb91olzCCYlERERERBkuKhcmtCm6Qia0RGBSS5QWGzdKmDDBgkWL9I3uG0tERER0MHXtb0JUV1T/OirbUNfmGg0jIsocnH5MlGJPPJGPjz4ywemUYDYrOP74ECZPdkGv1zoyIiIiyhaBVhcjYuoK646JgKqgrv2NCBf00zosoozApJYohbZvl/Dxx7GEFgB8PhHffWfApEkWDB1ap3F0RERElE0i+cehpvQ1rcMgyjicfkyUQvPnG1BZKcW1KYqAn3/OzGHabdsk3HSTDYMHO3DddTasXy8d+k1ERERERBriSC1RCvXuHUJ+vgK3O/75UZcuEY0iOjCfT8C119qxYYOuvq28XIdZs6pgtysaRkZEREREdGAcqSVKoeOOi+DUU4PQ6xuqQ5WWhnDnnV4No2rce++ZsGFD/HOuLVtkvPmmRaOIiIiIiIgOjSO1RCn25pvVmDbNhHnzjOjWLYLbbvMiLy/zSiBPmWIBkLgtgNMpIhqNbYMn84pBRERERBmGt6hEKSZJwF/+4sdf/uLXOpQD2r1bRG1t4sQNWVawZ4+I/v1LoCjAsceG8fLLNRmZlO8TiQDl5TJKShROm6acI3t+Qd7m5yBFahExdoC7+z+h6B1ah0VERJRSTGqJCC6XCL8/cZTWYFAxb54Rqhr7f9u2yRgxAnj77ep0h3hYvv7agNGj87Fzp4SCAgWnnRbEiy/WQuRCC8oBUmAn7KtvgRzcAQDQu5dC9m1EVd/PALEZf+5VFbL3V0CQgaLTkhQtERFR8vBWj4hwzDERdOgQjWsTBBUGA+oT2n3WrNEhknl1ruD3A48/no/163WoqxOxa5eMmTPNeOcds9ahEaWFdesr9QntPrq6tTA45zT5nJJ/KxxLz4Nj+cVwLL8I8rwBEEOVzQ2ViIgoqZjUEhEkCRg9uhbHHBOGxaKgpCSK884LoFWraMKxoggIiYO6mluyRI8tW+JHo0IhAfPnGzWKiCi9xEhtQpughiGFKpp8zsJ190LvXQVR8UGMeiG6fkLB2nubEyYREVHScfoxtRjbt0v4+GMTunSJ4PzzAyxqdIROPjmEr76qxLp1Mux2Ba1bK3j88XyUl+sQDseyWEFQcfzxIUgZuH2tw6HAYlHh8cRn3BZL5q7/JUomX+vLIe9LGgAAIABJREFUoXfNgxRtqK4eMbSDv2Rw006oRiH5tyU0y/4tTYyQiIgoNXjbTy3CuHEWvPWWFRUVEvR6FW+8EcYHHzhRUMCE5khIElBa2jC3eNQoNwBg0SI9FEXA8ceH8MQTiaNBmeDYYyPo3TuM774z1Le1bh3B8OEeDaMiSp9g0Vnwtb0WxsrZkMI1iBpK4Ol0F1SdvYlnFKFKidP3G2sjIiLSEpNaynput4ApUyyoqIgNH4ZCAlau1GPMmDw8/bRb4+iymygCjz6aPT/Dd9914ckn87B2rQ5Wq4I77/SiZ88MXABMlCKebo/A22kExNBeRI0dAFHX9JMJAvzF50PavgOi4gMAqLoC+FtflqRoiYiIkoNJLWW9det02LUrcT7spk3NuJmjrGQ2qxg9OnuScK2tXy9j6VIdTj45hC5dEtdPU3ZSZQuicteknMvb5V5Eje1hqpgJQIR09C2oM56ZlHMTERElC5NaynqdO0dQUhLF7t3xv86NFTnSktst4LPPTMjPV3DOOQHomHOTRlQVuOuuAsyda0R1tYSioijOP9/PmQ3UKH+bv8Lf5q8AAIfDAVRVaRwRERFRPFY/pqxXXBxLEs1mpb6te/cwHnggc27Q58wx4JxzinHffYW4/XYbzjuvGDt38utH2pg3z4DPPzehujo2w8HplDBjhhnLlvFJC6WPGNwDnecXQAlpHQoREWU5jtRSi/Dkk24MHBjExx+b0K5dFLfd5oXNlhlFohQFePbZfGzbFvu6RSICfvtNh1GjCjBpUrXG0VEumjXLCJ8v/qGK2y1i1iwj+vYNaxQV5Qw1isLfhsNQ8wOESC2ixvZwd3kAweJBWkdGRERZikltC6AowBdfGLFggQGnnx7E+ecHIObYIKAgAH/8YxB//GNQ61DiuN0CJkywYvPmxDW/ja0DJkqHnj3DEEUVitKw/ZFer6JXLya0lHqW7W/CVPEZBMSKuIm+cuRvehqV9j8Ckknj6Igo1whhFwQlAsVQonUo1AxMarOcogDXX2/Dd98ZEAiI+OgjEz74IIQpU1wZuZdoqrlcAh58sBAbN8owmxVccYUPV17p1ySW9esl3HST/YAFq7jdEGnlqqt8mD7djJUr9f9rUdG7dwgXXhjQNC7KDYaab+sT2n1k/xboa5ciZD9do6iIKNcIUR9sa/4G2bMaghpFxNwV1aUToBhaaR0aNQGT2iz31VeG+oQWAIJBET/8oMfs2cacu0FVVeDGG+1YvLhhn9KNG3UoKlJwzjlBVFWJePLJPOzYIaN16ygeftiNtm2Vg5yxeZ54ouCACW2bNhGMGMH9U1sCl0vEU0/lYetWGcXFUTz8sAcdOmRWkbLfM5mAqVOdeO01KzZskNGzZxh/+1sdZP5FoDRQpbxG23gj2UAM7oFl+5sQ1DDq2t+CqKmj1iERtTgF6x+E0Tmn/rVUW4nC34bDVfahhlFRU/EWJsstWtSQ0O4TColYtMiQc0nthg0y1q2LTyJra0U8+WQ+pk6NYOlSPSorG4avV6/WYdasKuTlpWbEtLIycQ64Xq/gvPMCuPLKOnzyiRmTJ1vw17/6cOaZmTVtmg5PKARccYUdq1fr69t+/VWHmTOrMmZN94FYrSoefJAPVtJJDDlh3fYqpOAu+Iv+D4FWlwJCjq0VAeDpdBd07iWQg7sBACqAUH4ZIpajtA0sQ+hd36Fw3d2QgzsBAMbK2XB3ewyBVhdpHBlRy6Lzrklok/2bgKifSyGyUO79NW1hzjgjCKMxfrRRr1dwxhnaJUlLlwoYPToPs2YZEU3jgFUwKCASSWzftEnGF1+YEpLMDRtkTJpkSVk8DkfiKHDnzlFcd50XI0fa8P77FsyaZcJtt9nw3HPWlMVBqfPpp0asXRv/IGXjRh3efJP9SfHE4F4ULb8Y1h0TYaqcjcJ196Pwt+Fah6WJiPVYuP7wNvxF5yBYeAq87W+F6w/vaB1Wxsjb+nx9QgsAcmgvrNsnaBgRUcukivrERkEPCBzzy0ZMarPcn/4UxGmnNSS2BoOCU04J4fzzkzNKu2CBHpddVoRzz3XglltsjY4+7u/hh/MxaJCMcePyMHx4IS67rAj+NC1p7dkzjC5dfp/VqgD2FcMRfvf/BGzfnrqFx4884kbnzg3xFBdHcf31dXj55Xzs3NlwwfR4RMycaU7bz+lQysslXHONDeee68AVV9jxyy+8uB/Itm0yIpHf/14BO3fm4IJ2Oqi8zc9C599Y/1pUQzBUL4Tk36phVNqJ5B2H6l7vwFk2DZ7ujwGSUeuQMoYYciW2hV2A0shTWyJqMr/jfCiiuf61ChlB26mAyO3tshHvVrOcKALvvluNL74wYOFCA04/PYRBg5JT/XjNGhkjRthQURG7QV+1Cti+XcJ//1vV6Pk3b5bw2Wcm1NTEbvJDIRE//6zHxIlWDB/ubX5AhyCKwLhx1bj33kLs3Cmhrk6A233g5EIQVFRXC1DVWPXkZDv22Ag++6wS77xjQXW1iBtuqEO3blFMnWpOOLa2VoDTKaF9e23XYnq9Am66yY7y8oYL+pYtMj79tArFxQ0jzxs3SnjhhTxUV4soKwvjySe1iFZ7l13mx+TJlrhp7Xl5Ci6/3KdhVJQOkQgwfrwVP/6oh9WqYsQID0pLD5x0SKG9iW1hJ+S6ckRNnVIZKmUZRe8A/OWJbSJv2YiSqa7T7YBkhLHiM0CNIlR4Cjxd79c6LGoiXiFbAFEEBg0KYtCg5E45HjfOWp/Q7rN+vYxvv9Wjf/9QwvHff2+A0/n7JFLAypXpe+J11FFRzJzpRF2dgGXLZNx8sx1e7/4xNYzcqqqARYuM+O47PU4/PfHfkww2m4p77olP6Nu3j2LFivjjSkqiaNNG++JCH3xgQnl5/GVh2zYZb7xhwahRsfWXW7dKuOqqImzfHjtu0SIDfvtNxbvvpjta7XXsGNsTefJkC3bvFlFSouD88wMp+32izHH77YX4/HMTotHY9WTZMh3eeceF3r0bT2zD1p4wuubFtUUM7REq6JPyWCm71HZ9CPbfhkMOxEbxI4Z28HS8S+OoiFqmuvY3oa79TVqHQUnA6cd0QIFA4vBlKCTA7W781+b440MoLExMzHr0SP+UKYtFxemnh3HeeYH6mPR6Bb+fguz1ivjkk/QWA3jiiVr07BmCThcrJNShQwR33+3JiC2YampEJE7TRlyfv/hiXn1CC8QeDvz8s4Bff83NZ2S33VaHr76qxCefOPHFF5V49FG31iFRiu3cKeKnnwz1CS0A7N4t4+WXE6v67uPpNALBgpOhiLHrTVTfCnVtr4Gqs6c8XsoukYJ+qDx+NtxdHoC70whU9f0MweJztA6LiCij5eZdKB2Wyy7zYdEiPerqGrKtzp0jOPvsxkeEjzkmgjPPDOLLL03w+wUIgopevcK4/fbUTT2uqxPwj3/kY+1aHcxmBUOHenHWWbFRMkEAxo6txZo1dfj+ez3cbgFjx+bH3YgKgoqjjkpv0l1SouC//63C118b4XSKGDzYj8LC1FbKnTjRgo8/NsHnE9C5cxTPP1+DkpLEQlZXXeXDhx+asXt3w6XBbo+tBd4nlvjGq60VsG2bhJ49c3PNl9WqoqwsrHUYlCYVFRJqaw/+8CeBZIKzbBr01d9C9m1AoPg8KIY2KYySspmqs8HbKTcLiRERNQWTWjqg884L4pdffPjsMyM8HhElJVH8/e9umEwHTsBee60GP/6ow3/+E0ZpaRjXXlsHUwoHQm+4wYbvvmsoMLJunQ6vv+7CySc3JBilpRGUlkYQCgFz5piwcmVDtbtjj43g2mvTv/5RloHzzkvPlkszZhjxwgtWeDyxhxMbN+pw4412zJpVlbCWuG1bBQ884MHrr1vhdIqw2RRcc01d3FpBnU5BfAEuoF07Faedlvwpt6EQ8PrrVixZokdJSRT33ONBmzap21uY6HCUlobRsWMU5eUNSawgqCgrO8R3QBAQsp+BkP2MFEdIWU8JQYjWQZULU1P0gYiohWFSSwf1wAMejBjhQW2tiOJi5YB/W1UVmDvXgNmzjTj5ZAHPPFMDY4qLWZaXy1i1Kr4ce2WlhDfesOLkk6sTjtfrgf/8x4kXXsjDxo0yOnaM4N57PTCbM3s/0eaaOtVcn9DuU14uY/16GUcfnTiy+uc/+3HJJf76pFa335LoQAD/2ws4/hehTRsV+fnJ/zneeKMdCxYYoCixz/vxRwOmTatC69ZMbEk7ej3QqVMEW7bIiEQASYotv7j3Xu77S82kqsjbNBrGqi8gRr2IGtqipvuTiHDtNRHRQTGppUMyGNDoVNX9DR9eiM8/N8LvFzFtmop333Xgo4+csFhSlzDW1gqNboPT2FrgffLzVfzzn6lb8+jxCPj8cyNsNgVnnRXMiHWyaiNdoCg46B7CktR4n2/fLqOqKvEfJafgSrJqlYylS/X1CS0AbN4s48UX8/Dss7XNOvdPP+nx7LN5cDpFOBwKHnjAjRNO4PRhOjyTJpnx7beG+u2cotFY5XAdd4Gg5lBVWLa+BsuOdyCqsZk8UqgCtnUjUdnvq7htRmTvbygo/wfE0F4ociE8nYYjVHSWVpGTGoXetQBSqBIBxzlQdYVaR0SUc1goiprtt99kzJ1rgN8f+3WKRmMVjydMsKT0c3v3DqNLl/jMTK9XceaZya0Cfbi++sqAP/2pGHffXYhbbrFh0CAHKiq0/4pddJEfFkt8gtqtWwTHHnvk61/btYvCbk9MdouLmxzeAW3cKDe6RrG5P9PKShF33VWIH380YMMGHX74wYDhw21wubTvK8oOX39tRDAY//uyaZOMVauY1VLTiKEqFC2/CPlbnqtPaPeRfZuhr11S/1qI+mD7dSgMNd9C59sAg3sxCtfdD8m3pdlxCBEPCtbdh6LlF8O26jrIntXNPmdLJ4RdcCy7APbVN8O2biSKl54L496ZWodFlHN4F0fN9vPPetTUJG7ls3Ztam/wdDrgmWdilYRttijat4/g0kt9uOWWukO/OckUBXj22Xxs2yYDEBAOi1i9Wo9RowrSHsvv/fWvftx2mxdHHx1Ghw4RnHZaAG++Wd2kZVpms4pLL/WhoKAhse3cOYKnnkp+gaj+/UNo0yb+vIKg4uSTD/3QQlWBBQv0GDMmDz//rIsbrX77bUtc9WYgtm3RpEmJ+wcT1dbGitFde60dTzyRB49HaHRmgsGAFr+UgVKnYN29MLiXQkDiFBpV1EGVGq5Ppr0fQ/ZvijtGDu2GZfuE5gWhqrD/cjUsu9+HofZnmJxzYF99Y1KS5ZasoPwf0HtWQlRjf5vkwHbkbR0LKJz9Q5lL8m1C3sbRMO/8FxBtZNpjFuL0Yzqo5ctj0z09HhFdu0bw2GNuFBTE37iddloQdnsULlf8frC9e6f+gn7iiSF8+WUVtm2TYLMpKVnXeTj27BEbHUHcsSMD5h8DGDnSi5EjvVCU2L7GzT3X6acHMXmyBW3aRHHbbXU4+mg7qqqSE+s+druCoiIFu3c3FKUSBMDjOfg/IBqNFRD74QcDfD4RkyaZcdZZQYwfXwNBiK0Lbkwo1PRiLFu3SpgwwYJwWMBNN8UX1qLs5fcDf/lLUf3a/W++MeKHHwy46y4Pli/Xobq64ft99NHhtFdSz3RicC90ntUI5/WEYmitdTgZTfZtPuD/i1iOQTivd0ODEoCAxL91gtq8v7m6mh+h8/4aH1dwJ6xbxwId32vWuZtNCUPn/RWKzo6oqaO2sfyO7N+S0CYF90AKbEXU3D39AREdgmXb67BunwApXAUVIsy7JsPV699QjNl9neZILR3Qxo0Sbr3VjrlzTVi82IAPP7TgmmuKEtZodu8exbnnBpCXFxu90+tVnHBCCLfckrqtfPYnCECnTlHNElogloA19vn5+ZlV0Ki5CS0AvPCCFffdV4gfftDj++8N2LYtNZeRQGDfFikNyaaiCJg37+AVyD791IiFC2MJLQB4vRLmzjXgxx9jMwduvtmH1q3jk482bSK47rqmjfAvWKDHpZcWYcoUK/7zHwv++tcivP9+evc+ptR4/30zfv01fsbJ6tU6uFwiHn/cjRNOCKJnzxAuvNCHSZNcGkWZmfLLH4dj6SAUrb4WxUsHIX/Do1qHlNFUKfGaoQgG+IvOh6vXv+IqIPtb/RkRY3xiF9UVoa79zc2KQQrthagkjtiIEW333ta7FqJ4yf+haMUlcCw7H/ZfrgMUbZYZNSYqJ66fVXSFUPStNIiG6OCEqA+WXe9BCsdGIgQo0NetRf6mJzSOrPmY1NIBvfxyHnbtih/MX7NGxk8/JU4rfu65Wrz3nhO33OLBxIkRTJvmTOlWPpnGaAQuuMAHq7Vh6libNhGMGJGexH5/igI880weBg924IILHHjxRWujxaKa4rPPjHjjDQvKy3XYs0fGihV6jBhhQyj5u/lAUYRGi1lFDjEYNm+eEeFw/KXN65XwzTexZLh9+ygee8yNXr1CaNs2gl69QvjnP2vRtm3THkCMHZsXt6+v0ylh0iRL0n7mpJ3fftPFFSoDYjUDfvtNh0su8WPGDCe++qoKEybUpHyv6Wwiu1fCvPsDyKE9AGLJknnPR9DV/KxNQKoCY8VnKFxzByxbX4MQTf82bofib3UpolJe/WtFNMPb4W+o7vUmFL097lhVV4iaHmMQyitDxNAWIWtPuLs8iIi1tFkxBIvOSkiWFdEIf/F5h3yvafd/ULRsCBxLzkXB2rshRJK0DEiJoKD8H9D5NkBUgpDCLhhc3yBv0zPJOX8SeDqPRMTQrv61IpoRcJwLVc47yLuItCH5N0MM7U1sD+7SIJrk4vRjOiCvN3E6pt8vYO9eCUDiNKd+/cLo1y8Mh8OQ9Kmo2eCBB7w4/vgwpk41w2JRcMcdXnTrdpASwyny2GP5mDIlNhUWAH79VYbfL+CRR5q/3ci0aSbU1cVPqd68Wcb8+RH07dvs08cxm1UcdVQEO3c2XKYEQcXxx4fg9wOTJ1vwyy86nH56EH/+s79+neMJJ4Qwc6apvjItAJhMCk48sSHzvvDCAC64IIBwOLY2uznbQNbUJD4brKkREQggpx7stEQXX+zDrFkmeL0NfVxQoODii1vG+qNUMe+dBjEaP7onRt0w752G2sIT0x5P4ZphMDq/hqgEoAIwVc6Cs2xaRiUddR1ugaJzwLR3KqCq8JcMhr/t1Qc8PmQfgCr7AEAJAUIzL2L/o8r5cHd5AHlbX4IY2A1VV4ig/Y8ItLoM1oO8z7h3BgrKH4cYjVWl13tXQQpWwHXcv5sdk+xbDzGwI65NgAq9Z2Wzz50skfwyOI/7CNZtr0KM1MLX6mIEiwdpHRZRo6LGTlD0DoiBbXHtir5Eo4iSJ2OS2hUrVmDSpElQFAVnn302hgwZonVITVJRIWL8eCuqqwVcfbUvq7cJGTTIjwULDAgEGm7oOnWK4uyzM2faT6YZODCIgQO1/fksWmSoT2gBIBgUMW+eMSlJbWNbluh0KiwWoKZGwOTJFtTVCbj22jq0a9f8qdevvVaNO++0YcMGGZIE9OsXwkMPuXH55Q4sX66DqgqYPduE2bNNmDLFBUEA/vIXH6ZPN2HpUj2iUQE6XWw6/O/7RRBi+402V0lJFOvXx/9gHA4l5fs0p1ptrYCpU2NZ+eWX+xPW0ueCU08NY8gQP7780ojKSgklJVEMHuxHnz7Ze11Ph4ilB1RIcUWPVIgIm3ukPRbZuwbG6oUQldhiegGA3rsa1q2vwdPtobTHczD+1hfD3/riI3uTeOQXMTFUCfPuD6HoCuFrdSmw39TnQKshCBQPguwrR9TQGqrOfpAzxZh3f1Cf0O6j866GGNgFxdj2iOPbn6IrAmQrEIof+d2/cFYmiJo7ofaY5w9+kBKCsWIm5MAO+Ftfjqix3cGPJ0oBVbbC1+oyWHa+AylSAwAIm7rB3XWUxpE1X0YktYqi4O2338aoUaNQVFSEhx56CP369UP79u21Du2I/PqrjJtvtv+vAi7w9dcmDB3qxV13pX8KajJcemkAy5b58M03RtTWimjdOop773XDas29m9tsoaqNFzxK1vTgW27x4qef9HA6G0Zre/SIwGQScP75xdiyJfa7P22aCaNGuXHxxQeoynSYbDYV773ngt8f2ztXrwfeeMNSn9ACQDgs4Kef9Fi4UI8BA0LQ64EPP3Ti/ffNWLpUjzPOCOKSS/xJWU/cmH/8w42hQyVs2iRDVQV07BjB/fe7kzFwopmff9ZjxIgCbN0aS9bffdeCsWNrsvohXVMIQqzC+h13eLFypQ59+oSS8rCmpfO1vhzmXf+G3ruqvi1iORa+tlelPRa9eznESOK+1rK/PO2xZALj3k+Qv2k05OAuqBBh2fE2XL3eRdTUqeEgUX9EU5kFNXFNiKCEISjNn+atGFohWHAiTJX/rX9IEtWVwNvxjmafO53EUBXsK6+Erm4tBERh2TUF3g63o67DTVqHRjnI2+UeBIrOgmXXvxE1tEFd+5uh6rTfraO5MiKpLS8vR+vWrdGqVWxR/amnnorFixdnXVI7ZkxefUILALW1IqZNM+HWW71ZOQ1REIDRo92orfWgokJCly6RRreyoMwhCLEqrFu3xndUU/akbcyJJ4bx1FO1ePttCzweER07RjBmTC3uvbcEW7Y0ZI179sgYP96Kiy4KJCWZ3P/788svDQntPj6fiCVLYkktEEt+r7/eh+uvT/3auWOPjeDzz6vw8ccm+P2xUU2bLbsf/Dz1VF59QgsAW7bo8NRT+Zgxw6lhVNrp0CGKDh3Sv5Qga4kGOI/7EHlbnofs24yIqRM8Xe6NGxFMl2DhaYjqiiCFG353VQgI5/VJeyyaU6OwbnsN8v/WzglQoPOtR/6GR1Hd+19NPm3Qdhr0tUsh7LcsKWLqhKipW7NDBoCaY19DxPwy9LU/Q5VM8Ha4HeHCE5Jy7nTJ3/gE9HUNlaWl0F5Ydk2Cr+2VjRYJI0q1SH4f1Oa3rOtgRqQoLpcLRUVF9a+LioqwYcMGDSNqmv23d2hoE1FRIaFTp+y9ISooUFFQwK0qssVLL9Vg2DAB69bJEMVY0vX88zVJO/8FF8TWo+6voiLxOJdLhNcrJL0q9YABQcyebYqbYl1QoGDgwOaNCjeHxaLi2mszr/hMU1VWJl7LKioyY3sqyg6qrgDuo7Svphk1d4a/5CKY9k6HFKmBIhoRziuDt5mVgrORGKqEFEoseCE1UjTmSHg7jYAU3A199XcQFT8ixg6oOfqlpKzzBQCIMrxd7knOuTQiBbYntInBvZB8GxHJ+4MGERG1PBmR1KqNlAkVGrkYzpkzB3PmzAEAjBkzBg6HI+WxHYlOnSQsXx7f1rq1gN69bY2uRWypZFnOuL7JJQ4H8NVXQHW1AkEACgtFAEWHfF9zlJQkfl+Li0V07lyU9Gm/Q4cCX36pYtEioK5OQGGhigsvVHH22YnbKlDTFBeL2Lo1vq2kREz4XvO7nnuyss8dE6DU/A3qjk+g2vpAaDsYDiEHN39Q8iEYHUA4PrGVza0O2aeH7PfiSVCjfkQjPgiGItiSEW8LIlnbAr+fBW9shcL2fYAMnfaZld91arZs7veMSGqLiorgdDZMDXI6nbDZEi+JAwcOxMCBA+tfV2VYid2HHxaxdq0dGzboEI0KaNMmgltu8aC2NrcqZTocjozrm1z2/+3dXWxUZR7H8d9MWyhloHQKbZcuKC31ohiE2sZKFGiKmhAuTFPJoolBoqxbkABLotwovpCYSIVkW0MTCYtcEIyhssZETERArMRibdUSkFaUBkvH6WCZ6Yudl2cvGmZ56QDDCjNn5vu5mpycSZ7D7/wP/Z/zzHPuRBSvvTZFHR228PT7nJygVqy4KI/n9pz7774rffHFGB0/PkaVlUO6775AUq64fbs891y6zp7NDD+dzc0N6Nln++R2X7nYFrWefKyb+VQpb9XIx97kfafw+Jylcgz+SymBC5KkwNhp8vx1nQI3yDSq3L1WPD9ur5T8fyrb06rUwTOSpFDKBPVnL5a3zy8pPv+9rFvr+H/EW+5Tp978YnNx0dQWFhaqu7tbLpdLTqdTTU1NWrNmTayHFbX8/JA+/titDz8cJ7c7RU88MaDcXBYVQeIrKzP66KNe7dgxXj6fTStW9GvGjNs35d5mk+bPH9b8+bfhBbnQkiVDKi72a/v2kRd5PP+8TwUF1v0JBYAR/dP+rj8yyzX+3L9lUifKN/0fCo3Ni/WwEl4wY4bcc/+j8V3blfLHefX/5W/yZ82L9bCAhGIzo839jYGWlhbt2rVLoVBIFRUVqqqquuF3fv3V+i8KTkTxdpcHtx+ZJydyTz5knpzIPfmQeXKKt9wt96RWkkpKSlRSUhLrYQAAAAAALCQJV0oAAAAAACQKmloAAAAAgGXR1AIAAAAALIumFgAAAABgWTS1AAAAAADLoqkFAAAAAFgWTS0AAAAAwLJoagEAAAAAlkVTCwAAAACwLJpaAAAAAIBl0dQCAAAAACyLphYAAAAAYFk0tQAAAAAAy6KpBQAAAABYFk0tAAAAAMCyaGoBAAAAAJZFUwsAAAAAsCybMcbEehAAAAAAANwKntTiT/fSSy/Fegi4w8g8OZF78iHz5ETuyYfMk5OVc6epBQAAAABYFk0tAAAAAMCyUjZt2rQp1oNA4ikoKIj1EHCHkXlyIvfkQ+bJidyTD5knJ6vmzkJRAAAAAADLYvoxAAAAAMCyUmM9AFjPqlWrlJ6eLrvdrpSUFL355pvy+XzaunWrfvvtN02ZMkXr1q2Tw+GQMUY7d+7Ut99+q7Fjx6qmpsay0xqSzTvvvKOWlhZlZmaqtrZWkm4p50OHDmnfvn2SpKqqKi1cuDBWh4QbGC3z999/X5999pkmTpwoSVq2bJlKSkokSY2NjTp48KDsdrueeeYZzZkzR5LU2tqqnTt3KhQKqbKyUo8//nhsDgg35Ha7VV/tzEiPAAAGvUlEQVRfr99//102m02LFi3S4sWLqfUEFyl36j1xDQ8P65VXXlEgEFAwGFR5ebmWLl0ql8ulbdu2yefzacaMGXrhhReUmpoqv9+vuro6/fTTT5owYYLWrl2rnJwcSZHPBcSfSLnX19frxIkTysjIkDTyt/3dd99t7Wu8AaJUU1Nj+vr6rti2e/du09jYaIwxprGx0ezevdsYY8w333xjNm/ebEKhkDl16pTZuHHjHR8vbk17e7vp7Ow069evD2+LNmev12tWrVplvF7vFZ8Rn0bLfO/evWb//v3X7NvV1WU2bNhghoeHTU9Pj1m9erUJBoMmGAya1atXm/Pnzxu/3282bNhgurq67uRhIAoej8d0dnYaY4wZGBgwa9asMV1dXdR6gouUO/WeuEKhkBkcHDTGGOP3+83GjRvNqVOnTG1trTl69KgxxpiGhgZz4MABY4wxn3zyiWloaDDGGHP06FHz9ttvG2MinwuIT5Fyr6urM1999dU1+1v5Gs/0Y/wpmpubtWDBAknSggUL1NzcLEk6fvy45s+fL5vNpnvuuUf9/f26cOFCLIeKm1RcXCyHw3HFtmhzbm1t1ezZs+VwOORwODR79my1trbe8WPBzRkt80iam5s1b948paWlKScnR3l5eero6FBHR4fy8vKUm5ur1NRUzZs3L3yeIP5kZWWF78KPGzdO+fn58ng81HqCi5R7JNS79dlsNqWnp0uSgsGggsGgbDab2tvbVV5eLklauHDhFbV+6UlceXm5fvjhBxljIp4LiE+Rco/Eytd4ph/jlmzevFmS9Mgjj2jRokXq6+tTVlaWpJH/LC9evChJ8ng8mjx5cvh72dnZ8ng84X1hLdHm7PF4lJ2dHd7udDqv+4cT4tOBAwd05MgRFRQU6Omnn5bD4ZDH41FRUVF4n8uzvTzz7OxsnT59+o6PGdFzuVw6c+aMZs6cSa0nkctzP3nyJPWewEKhkF588UWdP39ejz32mHJzc5WRkaGUlBRJV+Z6eU2npKQoIyNDXq/3uucC4tPVuRcVFenTTz/Vnj179MEHH+jee+/VU089pbS0NEtf42lqEbXXX39dTqdTfX19euONNzR16tSI+5pRFte+3h0iWFM0OZO/tTz66KOqrq6WJO3du1fvvfeeampqRs1couatamhoSLW1tVq+fHn4N1ajodYTy9W5U++JzW6366233lJ/f7+2bNmic+fORdw3UraRzgXEr6tzP3v2rJ588klNmjRJgUBADQ0N2r9/v6qrqy19jWf6MaLmdDolSZmZmSorK1NHR4cyMzPD04ovXLgQXmQiOztbbrc7/N3e3l6e0lpYtDk7nU719vaGt/OU3nomTZoku90uu92uyspKdXZ2ShrJ/OpsnU7nNdup+fgXCARUW1urhx9+WA888IAkaj0ZjJY79Z4cxo8fr+LiYp0+fVoDAwMKBoOS/perdGXmwWBQAwMDcjgcEc8FxL9Lube2tiorK0s2m01paWmqqKgITyG38jWephZRGRoa0uDgYPjzd999p+nTp6u0tFSHDx+WJB0+fFhlZWWSpNLSUh05ckTGGP3444/KyMiIuyLAzYs25zlz5qitrU0+n08+n09tbW2skmgxl/8G/uuvv9a0adMkjWTe1NQkv98vl8ul7u5uzZw5U4WFheru7pbL5VIgEFBTU5NKS0tjNXzcgDFG27dvV35+vpYsWRLeTq0ntki5U++J6+LFi+rv75c0siLu999/r/z8fM2aNUvHjh2TNLK67aX87r//fh06dEiSdOzYMc2aNUs2my3iuYD4FCn3S7V+6XfSl9e6Va/xNsM8AkShp6dHW7ZskTRy5+6hhx5SVVWVvF6vtm7dKrfbrcmTJ2v9+vXh1z/s2LFDbW1tGjNmjGpqalRYWBjjo8DN2LZtm06cOCGv16vMzEwtXbpUZWVlUed88OBBNTY2ShpZAr6ioiKWh4XrGC3z9vZ2/fzzz7LZbJoyZYpWrlwZvjG1b98+ff7557Lb7Vq+fLnmzp0rSWppadGuXbsUCoVUUVGhqqqqWB4WruPkyZN6+eWXNX369PBUsmXLlqmoqIhaT2CRcv/yyy+p9wT1yy+/qL6+XqFQSMYYPfjgg6qurlZPT881r/RJS0vT8PCw6urqdObMGTkcDq1du1a5ubmSIp8LiD+Rcn/11VfDayXcddddWrlypdLT0y19jaepBQAAAABYFtOPAQAAAACWRVMLAAAAALAsmloAAAAAgGXR1AIAAAAALIumFgAAAABgWTS1AAAAAADLoqkFAAAAAFgWTS0AAAAAwLL+C53B4WYPInKmAAAAAElFTkSuQmCC\n",
"text/plain": [
"
"
]
},
"metadata": {},
"output_type": "display_data"
}
],
"source": [
"# Vamos a RECORTAR los datos en la zona donde se concentran más los puntos\n",
"# esto es en el eje X: entre 0 y 3.500\n",
"# y en el eje Y: entre 0 y 80.000\n",
"filtered_data = data[(data['Word count'] <= 3500) & (data['# Shares'] <= 80000)]\n",
"\n",
"f1 = filtered_data['Word count'].values\n",
"f2 = filtered_data['# Shares'].values\n",
"\n",
"# Vamos a pintar en colores los puntos por debajo y por encima de la media de Cantidad de Palabras\n",
"asignar=[]\n",
"for index, row in filtered_data.iterrows():\n",
" if(row['Word count']>1808):\n",
" asignar.append(colores[0])\n",
" else:\n",
" asignar.append(colores[1])\n",
" \n",
"plt.scatter(f1, f2, c=asignar, s=tamanios[0])\n",
"plt.show()"
]
},
{
"cell_type": "code",
"execution_count": 27,
"metadata": {
"ExecuteTime": {
"end_time": "2018-05-06T15:28:51.024042Z",
"start_time": "2018-05-06T15:28:50.967169Z"
}
},
"outputs": [
{
"data": {
"text/html": [
"
\n",
"\n",
"
\n",
" \n",
"
\n",
"
\n",
"
Word count
\n",
"
# of Links
\n",
"
# of comments
\n",
"
# Images video
\n",
"
Elapsed days
\n",
"
# Shares
\n",
"
\n",
" \n",
" \n",
"
\n",
"
count
\n",
"
148.000000
\n",
"
148.000000
\n",
"
121.000000
\n",
"
148.000000
\n",
"
148.000000
\n",
"
148.000000
\n",
"
\n",
"
\n",
"
mean
\n",
"
1640.209459
\n",
"
5.743243
\n",
"
7.256198
\n",
"
3.331081
\n",
"
91.554054
\n",
"
20545.648649
\n",
"
\n",
"
\n",
"
std
\n",
"
821.975365
\n",
"
6.064418
\n",
"
6.346297
\n",
"
2.706476
\n",
"
91.143923
\n",
"
19933.865031
\n",
"
\n",
"
\n",
"
min
\n",
"
250.000000
\n",
"
0.000000
\n",
"
0.000000
\n",
"
1.000000
\n",
"
1.000000
\n",
"
0.000000
\n",
"
\n",
"
\n",
"
25%
\n",
"
971.000000
\n",
"
3.000000
\n",
"
2.000000
\n",
"
1.000000
\n",
"
28.750000
\n",
"
2750.000000
\n",
"
\n",
"
\n",
"
50%
\n",
"
1536.000000
\n",
"
5.000000
\n",
"
6.000000
\n",
"
3.000000
\n",
"
60.000000
\n",
"
15836.000000
\n",
"
\n",
"
\n",
"
75%
\n",
"
2335.750000
\n",
"
7.000000
\n",
"
11.000000
\n",
"
4.000000
\n",
"
110.500000
\n",
"
34177.500000
\n",
"
\n",
"
\n",
"
max
\n",
"
3485.000000
\n",
"
49.000000
\n",
"
30.000000
\n",
"
22.000000
\n",
"
349.000000
\n",
"
77000.000000
\n",
"
\n",
" \n",
"
\n",
"
"
],
"text/plain": [
" Word count # of Links # of comments # Images video Elapsed days \\\n",
"count 148.000000 148.000000 121.000000 148.000000 148.000000 \n",
"mean 1640.209459 5.743243 7.256198 3.331081 91.554054 \n",
"std 821.975365 6.064418 6.346297 2.706476 91.143923 \n",
"min 250.000000 0.000000 0.000000 1.000000 1.000000 \n",
"25% 971.000000 3.000000 2.000000 1.000000 28.750000 \n",
"50% 1536.000000 5.000000 6.000000 3.000000 60.000000 \n",
"75% 2335.750000 7.000000 11.000000 4.000000 110.500000 \n",
"max 3485.000000 49.000000 30.000000 22.000000 349.000000 \n",
"\n",
" # Shares \n",
"count 148.000000 \n",
"mean 20545.648649 \n",
"std 19933.865031 \n",
"min 0.000000 \n",
"25% 2750.000000 \n",
"50% 15836.000000 \n",
"75% 34177.500000 \n",
"max 77000.000000 "
]
},
"execution_count": 27,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"# Veamos como cambian los valores una vez filtrados\n",
"filtered_data.describe()"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Regresión Lineal Simple (1 variable)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Vamos a intentar primero una Regresión Lineal con 1 sóla variable"
]
},
{
"cell_type": "code",
"execution_count": 28,
"metadata": {
"ExecuteTime": {
"end_time": "2018-05-06T15:28:51.036535Z",
"start_time": "2018-05-06T15:28:51.027426Z"
}
},
"outputs": [],
"source": [
"# Asignamos nuestra variable de entrada X para entrenamiento y las etiquetas Y.\n",
"dataX =filtered_data[[\"Word count\"]]\n",
"X_train = np.array(dataX)\n",
"y_train = filtered_data['# Shares'].values"
]
},
{
"cell_type": "code",
"execution_count": 29,
"metadata": {
"ExecuteTime": {
"end_time": "2018-05-06T15:28:51.054663Z",
"start_time": "2018-05-06T15:28:51.042421Z"
},
"scrolled": true
},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"Coefficients: \n",
" [5.69765366]\n",
"Independent term: \n",
" 11200.303223074163\n",
"Mean squared error: 372888728.34\n",
"Variance score: 0.06\n"
]
}
],
"source": [
"# Creamos el objeto de Regresión Linear\n",
"regr = linear_model.LinearRegression()\n",
"\n",
"# Entrenamos nuestro modelo\n",
"regr.fit(X_train, y_train)\n",
"\n",
"# Hacemos las predicciones que en definitiva una línea (en este caso, al ser 2D)\n",
"y_pred = regr.predict(X_train)\n",
"\n",
"# Veamos los coeficienetes obtenidos, En nuestro caso, serán la Tangente\n",
"print('Coefficients: \\n', regr.coef_)\n",
"# Este es el valor donde corta el eje Y (en X=0)\n",
"print('Independent term: \\n', regr.intercept_)\n",
"# Error Cuadrado Medio\n",
"print(\"Mean squared error: %.2f\" % mean_squared_error(y_train, y_pred))\n",
"# Puntaje de Varianza. El mejor puntaje es un 1.0\n",
"print('Variance score: %.2f' % r2_score(y_train, y_pred))"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Visualizamos la Recta que obtuvimos"
]
},
{
"cell_type": "code",
"execution_count": 30,
"metadata": {
"ExecuteTime": {
"end_time": "2018-05-06T15:28:51.432242Z",
"start_time": "2018-05-06T15:28:51.060915Z"
}
},
"outputs": [
{
"data": {
"image/png": "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\n",
"text/plain": [
"